From 26c5032eaa64090b2a01973b0c6ea9e7f6a80fa7 Mon Sep 17 00:00:00 2001
From: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Date: Tue, 4 Jul 2006 14:16:28 +1000
Subject: [POWERPC] Add briq support to CHRP

The support for Briq machines has been floating around as patches for
ages. This cleans it up and adds it once for all.

Some of this is based on initial code provided by Karsten Jeppesen
<karsten@jeppesens.com> and mostly rewritten from scratch by me.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 include/asm-powerpc/processor.h | 1 +
 1 file changed, 1 insertion(+)

(limited to 'include')

diff --git a/include/asm-powerpc/processor.h b/include/asm-powerpc/processor.h
index 22e54a2a6604..6cb6fb19e57f 100644
--- a/include/asm-powerpc/processor.h
+++ b/include/asm-powerpc/processor.h
@@ -32,6 +32,7 @@
 #define _CHRP_Motorola	0x04	/* motorola chrp, the cobra */
 #define _CHRP_IBM	0x05	/* IBM chrp, the longtrail and longtrail 2 */
 #define _CHRP_Pegasos	0x06	/* Genesi/bplan's Pegasos and Pegasos2 */
+#define _CHRP_briq	0x07	/* TotalImpact's briQ */
 
 #if defined(__KERNEL__) && defined(CONFIG_PPC32)
 
-- 
cgit v1.2.3


From b5a1a9abe1a54ba40a9612001920f98bbdd0c56f Mon Sep 17 00:00:00 2001
From: Jeremy Kerr <jk@ozlabs.org>
Date: Tue, 4 Jul 2006 16:46:44 +1000
Subject: [POWERPC] Use const qualifiers for prom parsing utilites

The of_bus callbacks map and get_flags can be constified, as they don't
alter the range or addr arguments. of_dump_addr and of_read_addr can
also be constified.

Built for 32- and 64-bit powerpc

Signed-off-by: Jeremy Kerr <jk@ozlabs.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 arch/powerpc/kernel/prom_parse.c | 22 ++++++++++++----------
 include/asm-powerpc/prom.h       |  2 +-
 2 files changed, 13 insertions(+), 11 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/kernel/prom_parse.c b/arch/powerpc/kernel/prom_parse.c
index 21009b1f7869..e9960170667b 100644
--- a/arch/powerpc/kernel/prom_parse.c
+++ b/arch/powerpc/kernel/prom_parse.c
@@ -27,7 +27,7 @@
 
 /* Debug utility */
 #ifdef DEBUG
-static void of_dump_addr(const char *s, u32 *addr, int na)
+static void of_dump_addr(const char *s, const u32 *addr, int na)
 {
 	printk("%s", s);
 	while(na--)
@@ -35,7 +35,7 @@ static void of_dump_addr(const char *s, u32 *addr, int na)
 	printk("\n");
 }
 #else
-static void of_dump_addr(const char *s, u32 *addr, int na) { }
+static void of_dump_addr(const char *s, const u32 *addr, int na) { }
 #endif
 
 
@@ -46,9 +46,10 @@ struct of_bus {
 	int		(*match)(struct device_node *parent);
 	void		(*count_cells)(struct device_node *child,
 				       int *addrc, int *sizec);
-	u64		(*map)(u32 *addr, u32 *range, int na, int ns, int pna);
+	u64		(*map)(u32 *addr, const u32 *range,
+				int na, int ns, int pna);
 	int		(*translate)(u32 *addr, u64 offset, int na);
-	unsigned int	(*get_flags)(u32 *addr);
+	unsigned int	(*get_flags)(const u32 *addr);
 };
 
 
@@ -65,7 +66,8 @@ static void of_bus_default_count_cells(struct device_node *dev,
 		*sizec = prom_n_size_cells(dev);
 }
 
-static u64 of_bus_default_map(u32 *addr, u32 *range, int na, int ns, int pna)
+static u64 of_bus_default_map(u32 *addr, const u32 *range,
+		int na, int ns, int pna)
 {
 	u64 cp, s, da;
 
@@ -93,7 +95,7 @@ static int of_bus_default_translate(u32 *addr, u64 offset, int na)
 	return 0;
 }
 
-static unsigned int of_bus_default_get_flags(u32 *addr)
+static unsigned int of_bus_default_get_flags(const u32 *addr)
 {
 	return IORESOURCE_MEM;
 }
@@ -118,7 +120,7 @@ static void of_bus_pci_count_cells(struct device_node *np,
 		*sizec = 2;
 }
 
-static u64 of_bus_pci_map(u32 *addr, u32 *range, int na, int ns, int pna)
+static u64 of_bus_pci_map(u32 *addr, const u32 *range, int na, int ns, int pna)
 {
 	u64 cp, s, da;
 
@@ -143,7 +145,7 @@ static int of_bus_pci_translate(u32 *addr, u64 offset, int na)
 	return of_bus_default_translate(addr + 1, offset, na - 1);
 }
 
-static unsigned int of_bus_pci_get_flags(u32 *addr)
+static unsigned int of_bus_pci_get_flags(const u32 *addr)
 {
 	unsigned int flags = 0;
 	u32 w = addr[0];
@@ -178,7 +180,7 @@ static void of_bus_isa_count_cells(struct device_node *child,
 		*sizec = 1;
 }
 
-static u64 of_bus_isa_map(u32 *addr, u32 *range, int na, int ns, int pna)
+static u64 of_bus_isa_map(u32 *addr, const u32 *range, int na, int ns, int pna)
 {
 	u64 cp, s, da;
 
@@ -203,7 +205,7 @@ static int of_bus_isa_translate(u32 *addr, u64 offset, int na)
 	return of_bus_default_translate(addr + 1, offset, na - 1);
 }
 
-static unsigned int of_bus_isa_get_flags(u32 *addr)
+static unsigned int of_bus_isa_get_flags(const u32 *addr)
 {
 	unsigned int flags = 0;
 	u32 w = addr[0];
diff --git a/include/asm-powerpc/prom.h b/include/asm-powerpc/prom.h
index b095a285c84b..56f6ea0c76de 100644
--- a/include/asm-powerpc/prom.h
+++ b/include/asm-powerpc/prom.h
@@ -198,7 +198,7 @@ extern int release_OF_resource(struct device_node* node, int index);
 
 
 /* Helper to read a big number */
-static inline u64 of_read_number(u32 *cell, int size)
+static inline u64 of_read_number(const u32 *cell, int size)
 {
 	u64 r = 0;
 	while (size--)
-- 
cgit v1.2.3


From dac411e7aa92d23dadbcb8721845ab88577294c7 Mon Sep 17 00:00:00 2001
From: Michael Ellerman <michael@ellerman.id.au>
Date: Thu, 13 Jul 2006 17:52:04 +1000
Subject: [POWERPC] iseries: Move e2a()/strne2a() into their only caller

The ASCII -> EBCDIC functions, e2a() and strne2a() are now only used in
dt.c, so move them in there.

Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 arch/powerpc/lib/Makefile           |   1 -
 arch/powerpc/lib/e2a.c              | 116 ------------------------------------
 arch/powerpc/platforms/iseries/dt.c |  98 +++++++++++++++++++++++++++++-
 include/asm-powerpc/system.h        |   5 --
 4 files changed, 97 insertions(+), 123 deletions(-)
 delete mode 100644 arch/powerpc/lib/e2a.c

(limited to 'include')

diff --git a/arch/powerpc/lib/Makefile b/arch/powerpc/lib/Makefile
index ff7096458249..336dd191f768 100644
--- a/arch/powerpc/lib/Makefile
+++ b/arch/powerpc/lib/Makefile
@@ -14,7 +14,6 @@ endif
 obj-$(CONFIG_PPC64)	+= checksum_64.o copypage_64.o copyuser_64.o \
 			   memcpy_64.o usercopy_64.o mem_64.o string.o \
 			   strcase.o
-obj-$(CONFIG_PPC_ISERIES) += e2a.o
 obj-$(CONFIG_XMON)	+= sstep.o
 
 ifeq ($(CONFIG_PPC64),y)
diff --git a/arch/powerpc/lib/e2a.c b/arch/powerpc/lib/e2a.c
deleted file mode 100644
index 4b72ed8fd50e..000000000000
--- a/arch/powerpc/lib/e2a.c
+++ /dev/null
@@ -1,116 +0,0 @@
-/*
- *  EBCDIC to ASCII conversion
- *
- * This function moved here from arch/powerpc/platforms/iseries/viopath.c 
- *
- * (C) Copyright 2000-2004 IBM Corporation
- *
- * This program is free software;  you can redistribute it and/or
- * modify it under the terms of the GNU General Public License as
- * published by the Free Software Foundation; either version 2 of the
- * License, or (at your option) anyu later version.
- *
- * This program is distributed in the hope that it will be useful, but
- * WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
- * General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software Foundation,
- * Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
- *
- */
-
-#include <linux/module.h>
-
-unsigned char e2a(unsigned char x)
-{
-	switch (x) {
-	case 0xF0:
-		return '0';
-	case 0xF1:
-		return '1';
-	case 0xF2:
-		return '2';
-	case 0xF3:
-		return '3';
-	case 0xF4:
-		return '4';
-	case 0xF5:
-		return '5';
-	case 0xF6:
-		return '6';
-	case 0xF7:
-		return '7';
-	case 0xF8:
-		return '8';
-	case 0xF9:
-		return '9';
-	case 0xC1:
-		return 'A';
-	case 0xC2:
-		return 'B';
-	case 0xC3:
-		return 'C';
-	case 0xC4:
-		return 'D';
-	case 0xC5:
-		return 'E';
-	case 0xC6:
-		return 'F';
-	case 0xC7:
-		return 'G';
-	case 0xC8:
-		return 'H';
-	case 0xC9:
-		return 'I';
-	case 0xD1:
-		return 'J';
-	case 0xD2:
-		return 'K';
-	case 0xD3:
-		return 'L';
-	case 0xD4:
-		return 'M';
-	case 0xD5:
-		return 'N';
-	case 0xD6:
-		return 'O';
-	case 0xD7:
-		return 'P';
-	case 0xD8:
-		return 'Q';
-	case 0xD9:
-		return 'R';
-	case 0xE2:
-		return 'S';
-	case 0xE3:
-		return 'T';
-	case 0xE4:
-		return 'U';
-	case 0xE5:
-		return 'V';
-	case 0xE6:
-		return 'W';
-	case 0xE7:
-		return 'X';
-	case 0xE8:
-		return 'Y';
-	case 0xE9:
-		return 'Z';
-	}
-	return ' ';
-}
-EXPORT_SYMBOL(e2a);
-
-unsigned char* strne2a(unsigned char *dest, const unsigned char *src, size_t n)
-{
-	int i;
-
-	n = strnlen(src, n);
-
-	for (i = 0; i < n; i++)
-		dest[i] = e2a(src[i]);
-
-	return dest;
-}
diff --git a/arch/powerpc/platforms/iseries/dt.c b/arch/powerpc/platforms/iseries/dt.c
index d194140c1ebf..39c676ab9d6d 100644
--- a/arch/powerpc/platforms/iseries/dt.c
+++ b/arch/powerpc/platforms/iseries/dt.c
@@ -1,5 +1,6 @@
 /*
- *    Copyright (c) 2005-2006 Michael Ellerman, IBM Corporation
+ *    Copyright (C) 2005-2006 Michael Ellerman, IBM Corporation
+ *    Copyright (C) 2000-2004, IBM Corporation
  *
  *    Description:
  *      This file contains all the routines to build a flattened device
@@ -76,6 +77,101 @@ static char __initdata device_type_pci[] = "pci";
 static char __initdata device_type_vdevice[] = "vdevice";
 static char __initdata device_type_vscsi[] = "vscsi";
 
+
+/* EBCDIC to ASCII conversion routines */
+
+unsigned char e2a(unsigned char x)
+{
+	switch (x) {
+	case 0xF0:
+		return '0';
+	case 0xF1:
+		return '1';
+	case 0xF2:
+		return '2';
+	case 0xF3:
+		return '3';
+	case 0xF4:
+		return '4';
+	case 0xF5:
+		return '5';
+	case 0xF6:
+		return '6';
+	case 0xF7:
+		return '7';
+	case 0xF8:
+		return '8';
+	case 0xF9:
+		return '9';
+	case 0xC1:
+		return 'A';
+	case 0xC2:
+		return 'B';
+	case 0xC3:
+		return 'C';
+	case 0xC4:
+		return 'D';
+	case 0xC5:
+		return 'E';
+	case 0xC6:
+		return 'F';
+	case 0xC7:
+		return 'G';
+	case 0xC8:
+		return 'H';
+	case 0xC9:
+		return 'I';
+	case 0xD1:
+		return 'J';
+	case 0xD2:
+		return 'K';
+	case 0xD3:
+		return 'L';
+	case 0xD4:
+		return 'M';
+	case 0xD5:
+		return 'N';
+	case 0xD6:
+		return 'O';
+	case 0xD7:
+		return 'P';
+	case 0xD8:
+		return 'Q';
+	case 0xD9:
+		return 'R';
+	case 0xE2:
+		return 'S';
+	case 0xE3:
+		return 'T';
+	case 0xE4:
+		return 'U';
+	case 0xE5:
+		return 'V';
+	case 0xE6:
+		return 'W';
+	case 0xE7:
+		return 'X';
+	case 0xE8:
+		return 'Y';
+	case 0xE9:
+		return 'Z';
+	}
+	return ' ';
+}
+EXPORT_SYMBOL(e2a);
+
+unsigned char* strne2a(unsigned char *dest, const unsigned char *src, size_t n)
+{
+	int i;
+
+	n = strnlen(src, n);
+
+	for (i = 0; i < n; i++)
+		dest[i] = e2a(src[i]);
+
+	return dest;
+}
+
 static struct iseries_flat_dt * __init dt_init(void)
 {
 	struct iseries_flat_dt *dt;
diff --git a/include/asm-powerpc/system.h b/include/asm-powerpc/system.h
index d075725bf444..5deb7bc7bb1f 100644
--- a/include/asm-powerpc/system.h
+++ b/include/asm-powerpc/system.h
@@ -169,11 +169,6 @@ extern u32 booke_wdt_enabled;
 extern u32 booke_wdt_period;
 #endif /* CONFIG_BOOKE_WDT */
 
-/* EBCDIC -> ASCII conversion for [0-9A-Z] on iSeries */
-extern unsigned char e2a(unsigned char);
-extern unsigned char* strne2a(unsigned char *dest,
-		const unsigned char *src, size_t n);
-
 struct device_node;
 extern void note_scsi_host(struct device_node *, void *);
 
-- 
cgit v1.2.3


From c59acae85409fdf5d7574e90009c8410daf38938 Mon Sep 17 00:00:00 2001
From: Michael Ellerman <michael@ellerman.id.au>
Date: Thu, 13 Jul 2006 17:52:09 +1000
Subject: [POWERPC] iseries: Make ItExtVpdPanel private to iSeries

No one outside platforms/iseries needs ItExtVpdPanel anymore, so move
it in there. It used to be needed by lparcfg, and so was exported, but
isn't needed anymore, so unexport it.

Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 arch/powerpc/kernel/lparcfg.c                     |  1 -
 arch/powerpc/platforms/iseries/dt.c               |  2 +-
 arch/powerpc/platforms/iseries/it_exp_vpd_panel.h | 51 +++++++++++++++++++++++
 arch/powerpc/platforms/iseries/lpardata.c         |  3 +-
 include/asm-powerpc/iseries/it_exp_vpd_panel.h    | 51 -----------------------
 5 files changed, 53 insertions(+), 55 deletions(-)
 create mode 100644 arch/powerpc/platforms/iseries/it_exp_vpd_panel.h
 delete mode 100644 include/asm-powerpc/iseries/it_exp_vpd_panel.h

(limited to 'include')

diff --git a/arch/powerpc/kernel/lparcfg.c b/arch/powerpc/kernel/lparcfg.c
index 23f34daa044a..2d94b372d49b 100644
--- a/arch/powerpc/kernel/lparcfg.c
+++ b/arch/powerpc/kernel/lparcfg.c
@@ -32,7 +32,6 @@
 #include <asm/rtas.h>
 #include <asm/system.h>
 #include <asm/time.h>
-#include <asm/iseries/it_exp_vpd_panel.h>
 #include <asm/prom.h>
 #include <asm/vdso_datapage.h>
 
diff --git a/arch/powerpc/platforms/iseries/dt.c b/arch/powerpc/platforms/iseries/dt.c
index c5f59a8b9ef3..aa582262aefd 100644
--- a/arch/powerpc/platforms/iseries/dt.c
+++ b/arch/powerpc/platforms/iseries/dt.c
@@ -34,13 +34,13 @@
 #include <asm/iseries/hv_types.h>
 #include <asm/iseries/hv_lp_config.h>
 #include <asm/iseries/hv_call_xm.h>
-#include <asm/iseries/it_exp_vpd_panel.h>
 #include <asm/udbg.h>
 
 #include "processor_vpd.h"
 #include "call_hpt.h"
 #include "call_pci.h"
 #include "pci.h"
+#include "it_exp_vpd_panel.h"
 
 #ifdef DEBUG
 #define DBG(fmt...) udbg_printf(fmt)
diff --git a/arch/powerpc/platforms/iseries/it_exp_vpd_panel.h b/arch/powerpc/platforms/iseries/it_exp_vpd_panel.h
new file mode 100644
index 000000000000..6de9097b7f57
--- /dev/null
+++ b/arch/powerpc/platforms/iseries/it_exp_vpd_panel.h
@@ -0,0 +1,51 @@
+/*
+ * Copyright (C) 2002  Dave Boutcher IBM Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
+ */
+#ifndef _PLATFORMS_ISERIES_IT_EXT_VPD_PANEL_H
+#define _PLATFORMS_ISERIES_IT_EXT_VPD_PANEL_H
+
+/*
+ *	This struct maps the panel information
+ *
+ * Warning:
+ *	This data must match the architecture for the panel information
+ */
+
+#include <asm/types.h>
+
+struct ItExtVpdPanel {
+	/* Definition of the Extended Vpd On Panel Data Area */
+	char	systemSerial[8];
+	char	mfgID[4];
+	char	reserved1[24];
+	char	machineType[4];
+	char	systemID[6];
+	char	somUniqueCnt[4];
+	char	serialNumberCount;
+	char	reserved2[7];
+	u16	bbu3;
+	u16	bbu2;
+	u16	bbu1;
+	char	xLocationLabel[8];
+	u8	xRsvd1[6];
+	u16	xFrameId;
+	u8	xRsvd2[48];
+};
+
+extern struct ItExtVpdPanel	xItExtVpdPanel;
+
+#endif /* _PLATFORMS_ISERIES_IT_EXT_VPD_PANEL_H */
diff --git a/arch/powerpc/platforms/iseries/lpardata.c b/arch/powerpc/platforms/iseries/lpardata.c
index a7769445d6c7..13e9bd132254 100644
--- a/arch/powerpc/platforms/iseries/lpardata.c
+++ b/arch/powerpc/platforms/iseries/lpardata.c
@@ -18,7 +18,6 @@
 #include <asm/iseries/it_lp_reg_save.h>
 #include <asm/paca.h>
 #include <asm/iseries/lpar_map.h>
-#include <asm/iseries/it_exp_vpd_panel.h>
 #include <asm/iseries/it_lp_queue.h>
 
 #include "naca.h"
@@ -27,6 +26,7 @@
 #include "ipl_parms.h"
 #include "processor_vpd.h"
 #include "release_data.h"
+#include "it_exp_vpd_panel.h"
 
 /* The HvReleaseData is the root of the information shared between
  * the hypervisor and Linux.
@@ -134,7 +134,6 @@ struct ItIplParmsReal xItIplParmsReal __attribute__((__section__(".data")));
 
 /* May be filled in by the hypervisor so cannot end up in the BSS */
 struct ItExtVpdPanel xItExtVpdPanel __attribute__((__section__(".data")));
-EXPORT_SYMBOL(xItExtVpdPanel);
 
 #define maxPhysicalProcessors 32
 
diff --git a/include/asm-powerpc/iseries/it_exp_vpd_panel.h b/include/asm-powerpc/iseries/it_exp_vpd_panel.h
deleted file mode 100644
index 304a609ae21a..000000000000
--- a/include/asm-powerpc/iseries/it_exp_vpd_panel.h
+++ /dev/null
@@ -1,51 +0,0 @@
-/*
- * Copyright (C) 2002  Dave Boutcher IBM Corporation
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
- */
-#ifndef _ASM_POWERPC_ISERIES_IT_EXT_VPD_PANEL_H
-#define _ASM_POWERPC_ISERIES_IT_EXT_VPD_PANEL_H
-
-/*
- *	This struct maps the panel information
- *
- * Warning:
- *	This data must match the architecture for the panel information
- */
-
-#include <asm/types.h>
-
-struct ItExtVpdPanel {
-	/* Definition of the Extended Vpd On Panel Data Area */
-	char	systemSerial[8];
-	char	mfgID[4];
-	char	reserved1[24];
-	char	machineType[4];
-	char	systemID[6];
-	char	somUniqueCnt[4];
-	char	serialNumberCount;
-	char	reserved2[7];
-	u16	bbu3;
-	u16	bbu2;
-	u16	bbu1;
-	char	xLocationLabel[8];
-	u8	xRsvd1[6];
-	u16	xFrameId;
-	u8	xRsvd2[48];
-};
-
-extern struct ItExtVpdPanel	xItExtVpdPanel;
-
-#endif /* _ASM_POWERPC_ISERIES_IT_EXT_VPD_PANEL_H */
-- 
cgit v1.2.3


From a2ced11b6af59854cc2a2791dccd8b6c0da2f733 Mon Sep 17 00:00:00 2001
From: Michael Ellerman <michael@ellerman.id.au>
Date: Thu, 13 Jul 2006 17:52:12 +1000
Subject: [POWERPC] iseries: Make HvLpConfig_get(Primary)LpIndex functions

HvLpConfig_get(Primary)LpIndex are currently static inlines that return
fields from the itLpNaca, if we make them real functions we can make the
itLpNaca private to iSeries.

Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 arch/powerpc/platforms/iseries/hvlpconfig.c | 13 +++++++++++++
 arch/powerpc/platforms/iseries/setup.c      |  1 +
 include/asm-powerpc/iseries/hv_lp_config.h  | 13 ++-----------
 3 files changed, 16 insertions(+), 11 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/platforms/iseries/hvlpconfig.c b/arch/powerpc/platforms/iseries/hvlpconfig.c
index 663a1affb4bb..cfcedaae6ea2 100644
--- a/arch/powerpc/platforms/iseries/hvlpconfig.c
+++ b/arch/powerpc/platforms/iseries/hvlpconfig.c
@@ -18,9 +18,22 @@
 
 #include <linux/module.h>
 #include <asm/iseries/hv_lp_config.h>
+#include <asm/iseries/it_lp_naca.h>
 
 HvLpIndex HvLpConfig_getLpIndex_outline(void)
 {
 	return HvLpConfig_getLpIndex();
 }
 EXPORT_SYMBOL(HvLpConfig_getLpIndex_outline);
+
+HvLpIndex HvLpConfig_getLpIndex(void)
+{
+	return itLpNaca.xLpIndex;
+}
+EXPORT_SYMBOL(HvLpConfig_getLpIndex);
+
+HvLpIndex HvLpConfig_getPrimaryLpIndex(void)
+{
+	return itLpNaca.xPrimaryLpIndex;
+}
+EXPORT_SYMBOL_GPL(HvLpConfig_getPrimaryLpIndex);
diff --git a/arch/powerpc/platforms/iseries/setup.c b/arch/powerpc/platforms/iseries/setup.c
index c9605d773a77..c34299b18d8b 100644
--- a/arch/powerpc/platforms/iseries/setup.c
+++ b/arch/powerpc/platforms/iseries/setup.c
@@ -48,6 +48,7 @@
 #include <asm/iseries/hv_call_event.h>
 #include <asm/iseries/hv_call_xm.h>
 #include <asm/iseries/it_lp_queue.h>
+#include <asm/iseries/it_lp_naca.h>
 #include <asm/iseries/mf.h>
 #include <asm/iseries/hv_lp_event.h>
 #include <asm/iseries/lpar_map.h>
diff --git a/include/asm-powerpc/iseries/hv_lp_config.h b/include/asm-powerpc/iseries/hv_lp_config.h
index df8b20739719..a006fd1e4a2c 100644
--- a/include/asm-powerpc/iseries/hv_lp_config.h
+++ b/include/asm-powerpc/iseries/hv_lp_config.h
@@ -25,7 +25,6 @@
 
 #include <asm/iseries/hv_call_sc.h>
 #include <asm/iseries/hv_types.h>
-#include <asm/iseries/it_lp_naca.h>
 
 enum {
 	HvCallCfg_Cur	= 0,
@@ -44,16 +43,8 @@ enum {
 #define HvCallCfgGetHostingLpIndex			HvCallCfg + 32
 
 extern HvLpIndex HvLpConfig_getLpIndex_outline(void);
-
-static inline HvLpIndex	HvLpConfig_getLpIndex(void)
-{
-	return itLpNaca.xLpIndex;
-}
-
-static inline HvLpIndex	HvLpConfig_getPrimaryLpIndex(void)
-{
-	return itLpNaca.xPrimaryLpIndex;
-}
+extern HvLpIndex HvLpConfig_getLpIndex(void);
+extern HvLpIndex HvLpConfig_getPrimaryLpIndex(void);
 
 static inline u64 HvLpConfig_getMsChunks(void)
 {
-- 
cgit v1.2.3


From 06a36db1d712242a00cb30aaebdd088b4be28082 Mon Sep 17 00:00:00 2001
From: Michael Ellerman <michael@ellerman.id.au>
Date: Thu, 13 Jul 2006 17:52:17 +1000
Subject: [POWERPC] iseries: Move ItLpNaca into platforms/iseries

Move ItLpNaca into platforms/iseries now that it's not used elsewhere.

Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 arch/powerpc/kernel/setup_64.c              |  1 -
 arch/powerpc/platforms/iseries/hvlpconfig.c |  2 +-
 arch/powerpc/platforms/iseries/it_lp_naca.h | 80 +++++++++++++++++++++++++++++
 arch/powerpc/platforms/iseries/lpardata.c   |  3 +-
 arch/powerpc/platforms/iseries/lpevents.c   |  2 +-
 arch/powerpc/platforms/iseries/setup.c      |  2 +-
 include/asm-powerpc/iseries/it_lp_naca.h    | 80 -----------------------------
 7 files changed, 84 insertions(+), 86 deletions(-)
 create mode 100644 arch/powerpc/platforms/iseries/it_lp_naca.h
 delete mode 100644 include/asm-powerpc/iseries/it_lp_naca.h

(limited to 'include')

diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index fd1785e4c9bb..e2447aef3a8f 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -56,7 +56,6 @@
 #include <asm/page.h>
 #include <asm/mmu.h>
 #include <asm/lmb.h>
-#include <asm/iseries/it_lp_naca.h>
 #include <asm/firmware.h>
 #include <asm/xmon.h>
 #include <asm/udbg.h>
diff --git a/arch/powerpc/platforms/iseries/hvlpconfig.c b/arch/powerpc/platforms/iseries/hvlpconfig.c
index cfcedaae6ea2..f0475f0b1853 100644
--- a/arch/powerpc/platforms/iseries/hvlpconfig.c
+++ b/arch/powerpc/platforms/iseries/hvlpconfig.c
@@ -18,7 +18,7 @@
 
 #include <linux/module.h>
 #include <asm/iseries/hv_lp_config.h>
-#include <asm/iseries/it_lp_naca.h>
+#include "it_lp_naca.h"
 
 HvLpIndex HvLpConfig_getLpIndex_outline(void)
 {
diff --git a/arch/powerpc/platforms/iseries/it_lp_naca.h b/arch/powerpc/platforms/iseries/it_lp_naca.h
new file mode 100644
index 000000000000..9bbf58986819
--- /dev/null
+++ b/arch/powerpc/platforms/iseries/it_lp_naca.h
@@ -0,0 +1,80 @@
+/*
+ * Copyright (C) 2001  Mike Corrigan IBM Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
+ */
+#ifndef _PLATFORMS_ISERIES_IT_LP_NACA_H
+#define _PLATFORMS_ISERIES_IT_LP_NACA_H
+
+#include <linux/types.h>
+
+/*
+ *	This control block contains the data that is shared between the
+ *	hypervisor (PLIC) and the OS.
+ */
+
+struct ItLpNaca {
+// CACHE_LINE_1 0x0000 - 0x007F Contains read-only data
+	u32	xDesc;			// Eye catcher			x00-x03
+	u16	xSize;			// Size of this class		x04-x05
+	u16	xIntHdlrOffset;		// Offset to IntHdlr array	x06-x07
+	u8	xMaxIntHdlrEntries;	// Number of entries in array	x08-x08
+	u8	xPrimaryLpIndex;	// LP Index of Primary		x09-x09
+	u8	xServiceLpIndex;	// LP Ind of Service Focal Pointx0A-x0A
+	u8	xLpIndex;		// LP Index			x0B-x0B
+	u16	xMaxLpQueues;		// Number of allocated queues	x0C-x0D
+	u16	xLpQueueOffset;		// Offset to start of LP queues	x0E-x0F
+	u8	xPirEnvironMode;	// Piranha or hardware		x10-x10
+	u8	xPirConsoleMode;	// Piranha console indicator	x11-x11
+	u8	xPirDasdMode;		// Piranha dasd indicator	x12-x12
+	u8	xRsvd1_0[5];		// Reserved for Piranha related	x13-x17
+	u8	flags;			// flags, see below		x18-x1F
+	u8	xSpVpdFormat;		// VPD areas are in CSP format	...
+	u8	xIntProcRatio;		// Ratio of int procs to procs	...
+	u8	xRsvd1_2[5];		// Reserved			...
+	u16	xRsvd1_3;		// Reserved			x20-x21
+	u16	xPlicVrmIndex;		// VRM index of PLIC		x22-x23
+	u16	xMinSupportedSlicVrmInd;// Min supported OS VRM index	x24-x25
+	u16	xMinCompatableSlicVrmInd;// Min compatible OS VRM index x26-x27
+	u64	xLoadAreaAddr;		// ER address of load area	x28-x2F
+	u32	xLoadAreaChunks;	// Chunks for the load area	x30-x33
+	u32	xPaseSysCallCRMask;	// Mask used to test CR before  x34-x37
+					// doing an ASR switch on PASE
+					// system call.
+	u64	xSlicSegmentTablePtr;	// Pointer to Slic seg table.   x38-x3f
+	u8	xRsvd1_4[64];		//				x40-x7F
+
+// CACHE_LINE_2 0x0080 - 0x00FF Contains local read-write data
+	u8	xRsvd2_0[128];		// Reserved			x00-x7F
+
+// CACHE_LINE_3-6 0x0100 - 0x02FF Contains LP Queue indicators
+// NB: Padding required to keep xInterrruptHdlr at x300 which is required
+// for v4r4 PLIC.
+	u8	xOldLpQueue[128];	// LP Queue needed for v4r4	100-17F
+	u8	xRsvd3_0[384];		// Reserved			180-2FF
+
+// CACHE_LINE_7-8 0x0300 - 0x03FF Contains the address of the OS interrupt
+//  handlers
+	u64	xInterruptHdlr[32];	// Interrupt handlers		300-x3FF
+};
+
+extern struct ItLpNaca		itLpNaca;
+
+#define ITLPNACA_LPAR		0x80	/* Is LPAR installed on the system */
+#define ITLPNACA_PARTITIONED	0x40	/* Is the system partitioned */
+#define ITLPNACA_HWSYNCEDTBS	0x20	/* Hardware synced TBs */
+#define ITLPNACA_HMTINT		0x10	/* Utilize MHT for interrupts */
+
+#endif /* _PLATFORMS_ISERIES_IT_LP_NACA_H */
diff --git a/arch/powerpc/platforms/iseries/lpardata.c b/arch/powerpc/platforms/iseries/lpardata.c
index 13e9bd132254..8162049bb04d 100644
--- a/arch/powerpc/platforms/iseries/lpardata.c
+++ b/arch/powerpc/platforms/iseries/lpardata.c
@@ -13,7 +13,6 @@
 #include <asm/processor.h>
 #include <asm/ptrace.h>
 #include <asm/abs_addr.h>
-#include <asm/iseries/it_lp_naca.h>
 #include <asm/lppaca.h>
 #include <asm/iseries/it_lp_reg_save.h>
 #include <asm/paca.h>
@@ -27,6 +26,7 @@
 #include "processor_vpd.h"
 #include "release_data.h"
 #include "it_exp_vpd_panel.h"
+#include "it_lp_naca.h"
 
 /* The HvReleaseData is the root of the information shared between
  * the hypervisor and Linux.
@@ -127,7 +127,6 @@ struct ItLpNaca itLpNaca = {
 		(u64)instruction_access_slb_iSeries /* 0x480 I-SLB */
 	}
 };
-EXPORT_SYMBOL(itLpNaca);
 
 /* May be filled in by the hypervisor so cannot end up in the BSS */
 struct ItIplParmsReal xItIplParmsReal __attribute__((__section__(".data")));
diff --git a/arch/powerpc/platforms/iseries/lpevents.c b/arch/powerpc/platforms/iseries/lpevents.c
index 2a9f81ea27d6..98c1c2440aad 100644
--- a/arch/powerpc/platforms/iseries/lpevents.c
+++ b/arch/powerpc/platforms/iseries/lpevents.c
@@ -20,7 +20,7 @@
 #include <asm/iseries/it_lp_queue.h>
 #include <asm/iseries/hv_lp_event.h>
 #include <asm/iseries/hv_call_event.h>
-#include <asm/iseries/it_lp_naca.h>
+#include "it_lp_naca.h"
 
 /*
  * The LpQueue is used to pass event data from the hypervisor to
diff --git a/arch/powerpc/platforms/iseries/setup.c b/arch/powerpc/platforms/iseries/setup.c
index c34299b18d8b..7f1953066ff8 100644
--- a/arch/powerpc/platforms/iseries/setup.c
+++ b/arch/powerpc/platforms/iseries/setup.c
@@ -48,7 +48,6 @@
 #include <asm/iseries/hv_call_event.h>
 #include <asm/iseries/hv_call_xm.h>
 #include <asm/iseries/it_lp_queue.h>
-#include <asm/iseries/it_lp_naca.h>
 #include <asm/iseries/mf.h>
 #include <asm/iseries/hv_lp_event.h>
 #include <asm/iseries/lpar_map.h>
@@ -60,6 +59,7 @@
 #include "irq.h"
 #include "vpd_areas.h"
 #include "processor_vpd.h"
+#include "it_lp_naca.h"
 #include "main_store.h"
 #include "call_sm.h"
 #include "call_hpt.h"
diff --git a/include/asm-powerpc/iseries/it_lp_naca.h b/include/asm-powerpc/iseries/it_lp_naca.h
deleted file mode 100644
index 4fdcf052927f..000000000000
--- a/include/asm-powerpc/iseries/it_lp_naca.h
+++ /dev/null
@@ -1,80 +0,0 @@
-/*
- * Copyright (C) 2001  Mike Corrigan IBM Corporation
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
- */
-#ifndef _ASM_POWERPC_ISERIES_IT_LP_NACA_H
-#define _ASM_POWERPC_ISERIES_IT_LP_NACA_H
-
-#include <linux/types.h>
-
-/*
- *	This control block contains the data that is shared between the
- *	hypervisor (PLIC) and the OS.
- */
-
-struct ItLpNaca {
-// CACHE_LINE_1 0x0000 - 0x007F Contains read-only data
-	u32	xDesc;			// Eye catcher			x00-x03
-	u16	xSize;			// Size of this class		x04-x05
-	u16	xIntHdlrOffset;		// Offset to IntHdlr array	x06-x07
-	u8	xMaxIntHdlrEntries;	// Number of entries in array	x08-x08
-	u8	xPrimaryLpIndex;	// LP Index of Primary		x09-x09
-	u8	xServiceLpIndex;	// LP Ind of Service Focal Pointx0A-x0A
-	u8	xLpIndex;		// LP Index			x0B-x0B
-	u16	xMaxLpQueues;		// Number of allocated queues	x0C-x0D
-	u16	xLpQueueOffset;		// Offset to start of LP queues	x0E-x0F
-	u8	xPirEnvironMode;	// Piranha or hardware		x10-x10
-	u8	xPirConsoleMode;	// Piranha console indicator	x11-x11
-	u8	xPirDasdMode;		// Piranha dasd indicator	x12-x12
-	u8	xRsvd1_0[5];		// Reserved for Piranha related	x13-x17
-	u8	flags;			// flags, see below		x18-x1F
-	u8	xSpVpdFormat;		// VPD areas are in CSP format	...
-	u8	xIntProcRatio;		// Ratio of int procs to procs	...
-	u8	xRsvd1_2[5];		// Reserved			...
-	u16	xRsvd1_3;		// Reserved			x20-x21
-	u16	xPlicVrmIndex;		// VRM index of PLIC		x22-x23
-	u16	xMinSupportedSlicVrmInd;// Min supported OS VRM index	x24-x25
-	u16	xMinCompatableSlicVrmInd;// Min compatible OS VRM index x26-x27
-	u64	xLoadAreaAddr;		// ER address of load area	x28-x2F
-	u32	xLoadAreaChunks;	// Chunks for the load area	x30-x33
-	u32	xPaseSysCallCRMask;	// Mask used to test CR before  x34-x37
-					// doing an ASR switch on PASE
-					// system call.
-	u64	xSlicSegmentTablePtr;	// Pointer to Slic seg table.   x38-x3f
-	u8	xRsvd1_4[64];		//				x40-x7F
-
-// CACHE_LINE_2 0x0080 - 0x00FF Contains local read-write data
-	u8	xRsvd2_0[128];		// Reserved			x00-x7F
-
-// CACHE_LINE_3-6 0x0100 - 0x02FF Contains LP Queue indicators
-// NB: Padding required to keep xInterrruptHdlr at x300 which is required
-// for v4r4 PLIC.
-	u8	xOldLpQueue[128];	// LP Queue needed for v4r4	100-17F
-	u8	xRsvd3_0[384];		// Reserved			180-2FF
-
-// CACHE_LINE_7-8 0x0300 - 0x03FF Contains the address of the OS interrupt
-//  handlers
-	u64	xInterruptHdlr[32];	// Interrupt handlers		300-x3FF
-};
-
-extern struct ItLpNaca		itLpNaca;
-
-#define ITLPNACA_LPAR		0x80	/* Is LPAR installed on the system */
-#define ITLPNACA_PARTITIONED	0x40	/* Is the system partitioned */
-#define ITLPNACA_HWSYNCEDTBS	0x20	/* Hardware synced TBs */
-#define ITLPNACA_HMTINT		0x10	/* Utilize MHT for interrupts */
-
-#endif /* _ASM_POWERPC_ISERIES_IT_LP_NACA_H */
-- 
cgit v1.2.3


From 8bff05b052db7a4cfaaf0eee7f8145600548e9c9 Mon Sep 17 00:00:00 2001
From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Thu, 13 Jul 2006 18:51:22 +1000
Subject: [POWERPC] iseries: A new iSeries console

This driver uses the hvc_console.c infrastructure that is used by the
pSeries virtual and RTAS consoles.  This will allow us to make viocons.c
obsolete and is another step along the way to a combined kernel (as
viocons could not coexist with CONFIG_VT).

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 arch/powerpc/platforms/iseries/Kconfig |   8 +-
 arch/powerpc/platforms/iseries/dt.c    |   3 +-
 drivers/char/Kconfig                   |   7 +
 drivers/char/Makefile                  |   1 +
 drivers/char/hvc_iseries.c             | 593 +++++++++++++++++++++++++++++++++
 drivers/char/viocons.c                 |  31 +-
 include/asm-powerpc/iseries/vio.h      |  28 ++
 7 files changed, 638 insertions(+), 33 deletions(-)
 create mode 100644 drivers/char/hvc_iseries.c

(limited to 'include')

diff --git a/arch/powerpc/platforms/iseries/Kconfig b/arch/powerpc/platforms/iseries/Kconfig
index 3d957a30c8c2..887b68804e6d 100644
--- a/arch/powerpc/platforms/iseries/Kconfig
+++ b/arch/powerpc/platforms/iseries/Kconfig
@@ -3,13 +3,17 @@ menu "iSeries device drivers"
 	depends on PPC_ISERIES
 
 config VIOCONS
-	tristate "iSeries Virtual Console Support"
+	tristate "iSeries Virtual Console Support (Obsolete)"
+	help
+	  This is the old virtual console driver for legacy iSeries.
+	  You should use the iSeries Hypervisor Virtual Console
+	  support instead.
 
 config VIODASD
 	tristate "iSeries Virtual I/O disk support"
 	help
 	  If you are running on an iSeries system and you want to use
- 	  virtual disks created and managed by OS/400, say Y.
+	  virtual disks created and managed by OS/400, say Y.
 
 config VIOCD
 	tristate "iSeries Virtual I/O CD support"
diff --git a/arch/powerpc/platforms/iseries/dt.c b/arch/powerpc/platforms/iseries/dt.c
index d194140c1ebf..894b534669d0 100644
--- a/arch/powerpc/platforms/iseries/dt.c
+++ b/arch/powerpc/platforms/iseries/dt.c
@@ -298,7 +298,8 @@ static void __init dt_vdevices(struct iseries_flat_dt *dt)
 	dt_prop_u32(dt, "#address-cells", 1);
 	dt_prop_u32(dt, "#size-cells", 0);
 
-	dt_do_vdevice(dt, "vty", reg, -1, device_type_serial, NULL, 1);
+	dt_do_vdevice(dt, "vty", reg, -1, device_type_serial,
+			"IBM,iSeries-vty", 1);
 	reg++;
 
 	dt_do_vdevice(dt, "v-scsi", reg, -1, device_type_vscsi,
diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig
index 11de59ff4229..a7ef542afbc2 100644
--- a/drivers/char/Kconfig
+++ b/drivers/char/Kconfig
@@ -610,6 +610,13 @@ config HVC_CONSOLE
 	  console. This driver allows each pSeries partition to have a console
 	  which is accessed via the HMC.
 
+config HVC_ISERIES
+	bool "iSeries Hypervisor Virtual Console support"
+	depends on PPC_ISERIES && !VIOCONS
+	select HVC_DRIVER
+	help
+	  iSeries machines support a hypervisor virtual console.
+
 config HVC_RTAS
 	bool "IBM RTAS Console support"
 	depends on PPC_RTAS
diff --git a/drivers/char/Makefile b/drivers/char/Makefile
index 7a7ee5721279..8c6dfc621520 100644
--- a/drivers/char/Makefile
+++ b/drivers/char/Makefile
@@ -42,6 +42,7 @@ obj-$(CONFIG_AMIGA_BUILTIN_SERIAL) += amiserial.o
 obj-$(CONFIG_SX)		+= sx.o generic_serial.o
 obj-$(CONFIG_RIO)		+= rio/ generic_serial.o
 obj-$(CONFIG_HVC_CONSOLE)	+= hvc_vio.o hvsi.o
+obj-$(CONFIG_HVC_ISERIES)	+= hvc_iseries.o
 obj-$(CONFIG_HVC_RTAS)		+= hvc_rtas.o
 obj-$(CONFIG_HVC_DRIVER)	+= hvc_console.o
 obj-$(CONFIG_RAW_DRIVER)	+= raw.o
diff --git a/drivers/char/hvc_iseries.c b/drivers/char/hvc_iseries.c
new file mode 100644
index 000000000000..256afc8e5838
--- /dev/null
+++ b/drivers/char/hvc_iseries.c
@@ -0,0 +1,593 @@
+/*
+ * iSeries vio driver interface to hvc_console.c
+ *
+ * This code is based heavily on hvc_vio.c and viocons.c
+ *
+ * Copyright (C) 2006 Stephen Rothwell, IBM Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
+ */
+#include <stdarg.h>
+#include <linux/types.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/spinlock.h>
+#include <linux/console.h>
+
+#include <asm/hvconsole.h>
+#include <asm/vio.h>
+#include <asm/prom.h>
+#include <asm/iseries/vio.h>
+#include <asm/iseries/hv_call.h>
+#include <asm/iseries/hv_lp_config.h>
+#include <asm/iseries/hv_lp_event.h>
+
+#include "hvc_console.h"
+
+#define VTTY_PORTS 10
+
+static DEFINE_SPINLOCK(consolelock);
+static DEFINE_SPINLOCK(consoleloglock);
+
+static const char hvc_driver_name[] = "hvc_console";
+
+#define IN_BUF_SIZE	200
+
+/*
+ * Our port information.
+ */
+static struct port_info {
+	HvLpIndex lp;
+	u64 seq;	/* sequence number of last HV send */
+	u64 ack;	/* last ack from HV */
+	struct hvc_struct *hp;
+	int in_start;
+	int in_end;
+	unsigned char in_buf[IN_BUF_SIZE];
+} port_info[VTTY_PORTS] = {
+	[ 0 ... VTTY_PORTS - 1 ] = {
+		.lp = HvLpIndexInvalid
+	}
+};
+
+#define viochar_is_console(pi)	((pi) == &port_info[0])
+
+static struct vio_device_id hvc_driver_table[] __devinitdata = {
+	{"serial", "IBM,iSeries-vty"},
+	{ "", "" }
+};
+MODULE_DEVICE_TABLE(vio, hvc_driver_table);
+
+static void hvlog(char *fmt, ...)
+{
+	int i;
+	unsigned long flags;
+	va_list args;
+	static char buf[256];
+
+	spin_lock_irqsave(&consoleloglock, flags);
+	va_start(args, fmt);
+	i = vscnprintf(buf, sizeof(buf) - 1, fmt, args);
+	va_end(args);
+	buf[i++] = '\r';
+	HvCall_writeLogBuffer(buf, i);
+	spin_unlock_irqrestore(&consoleloglock, flags);
+}
+
+/*
+ * Initialize the common fields in a charLpEvent
+ */
+static void init_data_event(struct viocharlpevent *viochar, HvLpIndex lp)
+{
+	struct HvLpEvent *hev = &viochar->event;
+
+	memset(viochar, 0, sizeof(struct viocharlpevent));
+
+	hev->flags = HV_LP_EVENT_VALID | HV_LP_EVENT_DEFERRED_ACK |
+		HV_LP_EVENT_INT;
+	hev->xType = HvLpEvent_Type_VirtualIo;
+	hev->xSubtype = viomajorsubtype_chario | viochardata;
+	hev->xSourceLp = HvLpConfig_getLpIndex();
+	hev->xTargetLp = lp;
+	hev->xSizeMinus1 = sizeof(struct viocharlpevent);
+	hev->xSourceInstanceId = viopath_sourceinst(lp);
+	hev->xTargetInstanceId = viopath_targetinst(lp);
+}
+
+static int get_chars(uint32_t vtermno, char *buf, int count)
+{
+	struct port_info *pi;
+	int n = 0;
+	unsigned long flags;
+
+	if (vtermno >= VTTY_PORTS)
+		return -EINVAL;
+	if (count == 0)
+		return 0;
+
+	pi = &port_info[vtermno];
+	spin_lock_irqsave(&consolelock, flags);
+
+	if (pi->in_end == 0)
+		goto done;
+
+	n = pi->in_end - pi->in_start;
+	if (n > count)
+		n = count;
+	memcpy(buf, &pi->in_buf[pi->in_start], n);
+	pi->in_start += n;
+	if (pi->in_start == pi->in_end) {
+		pi->in_start = 0;
+		pi->in_end = 0;
+	}
+done:
+	spin_unlock_irqrestore(&consolelock, flags);
+	return n;
+}
+
+static int put_chars(uint32_t vtermno, const char *buf, int count)
+{
+	struct viocharlpevent *viochar;
+	struct port_info *pi;
+	HvLpEvent_Rc hvrc;
+	unsigned long flags;
+	int sent = 0;
+
+	if (vtermno >= VTTY_PORTS)
+		return -EINVAL;
+
+	pi = &port_info[vtermno];
+
+	spin_lock_irqsave(&consolelock, flags);
+
+	if (viochar_is_console(pi) && !viopath_isactive(pi->lp)) {
+		spin_lock_irqsave(&consoleloglock, flags);
+		HvCall_writeLogBuffer(buf, count);
+		spin_unlock_irqrestore(&consoleloglock, flags);
+		sent = count;
+		goto done;
+	}
+
+	viochar = vio_get_event_buffer(viomajorsubtype_chario);
+	if (viochar == NULL) {
+		hvlog("\n\rviocons: Can't get viochar buffer.");
+		goto done;
+	}
+
+	while ((count > 0) && ((pi->seq - pi->ack) < VIOCHAR_WINDOW)) {
+		int len;
+
+		len = (count > VIOCHAR_MAX_DATA) ? VIOCHAR_MAX_DATA : count;
+
+		if (viochar_is_console(pi)) {
+			spin_lock_irqsave(&consoleloglock, flags);
+			HvCall_writeLogBuffer(buf, len);
+			spin_unlock_irqrestore(&consoleloglock, flags);
+		}
+
+		init_data_event(viochar, pi->lp);
+
+		viochar->len = len;
+		viochar->event.xCorrelationToken = pi->seq++;
+		viochar->event.xSizeMinus1 =
+			offsetof(struct viocharlpevent, data) + len;
+
+		memcpy(viochar->data, buf, len);
+
+		hvrc = HvCallEvent_signalLpEvent(&viochar->event);
+		if (hvrc)
+			hvlog("\n\rerror sending event! return code %d\n\r",
+				(int)hvrc);
+		sent += len;
+		count -= len;
+		buf += len;
+	}
+
+	vio_free_event_buffer(viomajorsubtype_chario, viochar);
+done:
+	spin_unlock_irqrestore(&consolelock, flags);
+	return sent;
+}
+
+static struct hv_ops hvc_get_put_ops = {
+	.get_chars = get_chars,
+	.put_chars = put_chars,
+};
+
+static int __devinit hvc_vio_probe(struct vio_dev *vdev,
+			const struct vio_device_id *id)
+{
+	struct hvc_struct *hp;
+	struct port_info *pi;
+
+	/* probed with invalid parameters. */
+	if (!vdev || !id)
+		return -EPERM;
+
+	if (vdev->unit_address >= VTTY_PORTS)
+		return -ENODEV;
+
+	pi = &port_info[vdev->unit_address];
+
+	hp = hvc_alloc(vdev->unit_address, vdev->irq, &hvc_get_put_ops);
+	if (IS_ERR(hp))
+		return PTR_ERR(hp);
+	pi->hp = hp;
+	dev_set_drvdata(&vdev->dev, pi);
+
+	return 0;
+}
+
+static int __devexit hvc_vio_remove(struct vio_dev *vdev)
+{
+	struct port_info *pi = dev_get_drvdata(&vdev->dev);
+	struct hvc_struct *hp = pi->hp;
+
+	return hvc_remove(hp);
+}
+
+static struct vio_driver hvc_vio_driver = {
+	.id_table	= hvc_driver_table,
+	.probe		= hvc_vio_probe,
+	.remove		= hvc_vio_remove,
+	.driver		= {
+		.name	= hvc_driver_name,
+		.owner	= THIS_MODULE,
+	}
+};
+
+static void hvc_open_event(struct HvLpEvent *event)
+{
+	unsigned long flags;
+	struct viocharlpevent *cevent = (struct viocharlpevent *)event;
+	u8 port = cevent->virtual_device;
+	struct port_info *pi;
+	int reject = 0;
+
+	if (hvlpevent_is_ack(event)) {
+		if (port >= VTTY_PORTS)
+			return;
+
+		spin_lock_irqsave(&consolelock, flags);
+
+		pi = &port_info[port];
+		if (event->xRc == HvLpEvent_Rc_Good) {
+			pi->seq = pi->ack = 0;
+			/*
+			 * This line allows connections from the primary
+			 * partition but once one is connected from the
+			 * primary partition nothing short of a reboot
+			 * of linux will allow access from the hosting
+			 * partition again without a required iSeries fix.
+			 */
+			pi->lp = event->xTargetLp;
+		}
+
+		spin_unlock_irqrestore(&consolelock, flags);
+		if (event->xRc != HvLpEvent_Rc_Good)
+			printk(KERN_WARNING
+			       "hvc: handle_open_event: event->xRc == (%d).\n",
+			       event->xRc);
+
+		if (event->xCorrelationToken != 0) {
+			atomic_t *aptr= (atomic_t *)event->xCorrelationToken;
+			atomic_set(aptr, 1);
+		} else
+			printk(KERN_WARNING
+			       "hvc: weird...got open ack without atomic\n");
+		return;
+	}
+
+	/* This had better require an ack, otherwise complain */
+	if (!hvlpevent_need_ack(event)) {
+		printk(KERN_WARNING "hvc: viocharopen without ack bit!\n");
+		return;
+	}
+
+	spin_lock_irqsave(&consolelock, flags);
+
+	/* Make sure this is a good virtual tty */
+	if (port >= VTTY_PORTS) {
+		event->xRc = HvLpEvent_Rc_SubtypeError;
+		cevent->subtype_result_code = viorc_openRejected;
+		/*
+		 * Flag state here since we can't printk while holding
+		 * the consolelock spinlock.
+		 */
+		reject = 1;
+	} else {
+		pi = &port_info[port];
+		if ((pi->lp != HvLpIndexInvalid) &&
+				(pi->lp != event->xSourceLp)) {
+			/*
+			 * If this is tty is already connected to a different
+			 * partition, fail.
+			 */
+			event->xRc = HvLpEvent_Rc_SubtypeError;
+			cevent->subtype_result_code = viorc_openRejected;
+			reject = 2;
+		} else {
+			pi->lp = event->xSourceLp;
+			event->xRc = HvLpEvent_Rc_Good;
+			cevent->subtype_result_code = viorc_good;
+			pi->seq = pi->ack = 0;
+		}
+	}
+
+	spin_unlock_irqrestore(&consolelock, flags);
+
+	if (reject == 1)
+		printk(KERN_WARNING "hvc: open rejected: bad virtual tty.\n");
+	else if (reject == 2)
+		printk(KERN_WARNING "hvc: open rejected: console in exclusive "
+				"use by another partition.\n");
+
+	/* Return the acknowledgement */
+	HvCallEvent_ackLpEvent(event);
+}
+
+/*
+ * Handle a close charLpEvent.  This should ONLY be an Interrupt because the
+ * virtual console should never actually issue a close event to the hypervisor
+ * because the virtual console never goes away.  A close event coming from the
+ * hypervisor simply means that there are no client consoles connected to the
+ * virtual console.
+ */
+static void hvc_close_event(struct HvLpEvent *event)
+{
+	unsigned long flags;
+	struct viocharlpevent *cevent = (struct viocharlpevent *)event;
+	u8 port = cevent->virtual_device;
+
+	if (!hvlpevent_is_int(event)) {
+		printk(KERN_WARNING
+			"hvc: got unexpected close acknowlegement\n");
+		return;
+	}
+
+	if (port >= VTTY_PORTS) {
+		printk(KERN_WARNING
+			"hvc: close message from invalid virtual device.\n");
+		return;
+	}
+
+	/* For closes, just mark the console partition invalid */
+	spin_lock_irqsave(&consolelock, flags);
+
+	if (port_info[port].lp == event->xSourceLp)
+		port_info[port].lp = HvLpIndexInvalid;
+
+	spin_unlock_irqrestore(&consolelock, flags);
+}
+
+static void hvc_data_event(struct HvLpEvent *event)
+{
+	unsigned long flags;
+	struct viocharlpevent *cevent = (struct viocharlpevent *)event;
+	struct port_info *pi;
+	int n;
+	u8 port = cevent->virtual_device;
+
+	if (port >= VTTY_PORTS) {
+		printk(KERN_WARNING "hvc: data on invalid virtual device %d\n",
+				port);
+		return;
+	}
+	if (cevent->len == 0)
+		return;
+
+	/*
+	 * Change 05/01/2003 - Ryan Arnold: If a partition other than
+	 * the current exclusive partition tries to send us data
+	 * events then just drop them on the floor because we don't
+	 * want his stinking data.  He isn't authorized to receive
+	 * data because he wasn't the first one to get the console,
+	 * therefore he shouldn't be allowed to send data either.
+	 * This will work without an iSeries fix.
+	 */
+	pi = &port_info[port];
+	if (pi->lp != event->xSourceLp)
+		return;
+
+	spin_lock_irqsave(&consolelock, flags);
+
+	n = IN_BUF_SIZE - pi->in_end;
+	if (n > cevent->len)
+		n = cevent->len;
+	if (n > 0) {
+		memcpy(&pi->in_buf[pi->in_end], cevent->data, n);
+		pi->in_end += n;
+	}
+	spin_unlock_irqrestore(&consolelock, flags);
+	if (n == 0)
+		printk(KERN_WARNING "hvc: input buffer overflow\n");
+}
+
+static void hvc_ack_event(struct HvLpEvent *event)
+{
+	struct viocharlpevent *cevent = (struct viocharlpevent *)event;
+	unsigned long flags;
+	u8 port = cevent->virtual_device;
+
+	if (port >= VTTY_PORTS) {
+		printk(KERN_WARNING "hvc: data on invalid virtual device\n");
+		return;
+	}
+
+	spin_lock_irqsave(&consolelock, flags);
+	port_info[port].ack = event->xCorrelationToken;
+	spin_unlock_irqrestore(&consolelock, flags);
+}
+
+static void hvc_config_event(struct HvLpEvent *event)
+{
+	struct viocharlpevent *cevent = (struct viocharlpevent *)event;
+
+	if (cevent->data[0] == 0x01)
+		printk(KERN_INFO "hvc: window resized to %d: %d: %d: %d\n",
+		       cevent->data[1], cevent->data[2],
+		       cevent->data[3], cevent->data[4]);
+	else
+		printk(KERN_WARNING "hvc: unknown config event\n");
+}
+
+static void hvc_handle_event(struct HvLpEvent *event)
+{
+	int charminor;
+
+	if (event == NULL)
+		return;
+
+	charminor = event->xSubtype & VIOMINOR_SUBTYPE_MASK;
+	switch (charminor) {
+	case viocharopen:
+		hvc_open_event(event);
+		break;
+	case viocharclose:
+		hvc_close_event(event);
+		break;
+	case viochardata:
+		hvc_data_event(event);
+		break;
+	case viocharack:
+		hvc_ack_event(event);
+		break;
+	case viocharconfig:
+		hvc_config_event(event);
+		break;
+	default:
+		if (hvlpevent_is_int(event) && hvlpevent_need_ack(event)) {
+			event->xRc = HvLpEvent_Rc_InvalidSubtype;
+			HvCallEvent_ackLpEvent(event);
+		}
+	}
+}
+
+static int send_open(HvLpIndex remoteLp, void *sem)
+{
+	return HvCallEvent_signalLpEventFast(remoteLp,
+			HvLpEvent_Type_VirtualIo,
+			viomajorsubtype_chario | viocharopen,
+			HvLpEvent_AckInd_DoAck, HvLpEvent_AckType_ImmediateAck,
+			viopath_sourceinst(remoteLp),
+			viopath_targetinst(remoteLp),
+			(u64)(unsigned long)sem, VIOVERSION << 16,
+			0, 0, 0, 0);
+}
+
+static int hvc_vio_init(void)
+{
+	atomic_t wait_flag;
+	int rc;
+
+	/* +2 for fudge */
+	rc = viopath_open(HvLpConfig_getPrimaryLpIndex(),
+			viomajorsubtype_chario, VIOCHAR_WINDOW + 2);
+	if (rc)
+		printk(KERN_WARNING "hvc: error opening to primary %d\n", rc);
+
+	if (viopath_hostLp == HvLpIndexInvalid)
+		vio_set_hostlp();
+
+	/*
+	 * And if the primary is not the same as the hosting LP, open to the
+	 * hosting lp
+	 */
+	if ((viopath_hostLp != HvLpIndexInvalid) &&
+	    (viopath_hostLp != HvLpConfig_getPrimaryLpIndex())) {
+		printk(KERN_INFO "hvc: open path to hosting (%d)\n",
+				viopath_hostLp);
+		rc = viopath_open(viopath_hostLp, viomajorsubtype_chario,
+				VIOCHAR_WINDOW + 2);	/* +2 for fudge */
+		if (rc)
+			printk(KERN_WARNING
+				"error opening to partition %d: %d\n",
+				viopath_hostLp, rc);
+	}
+
+	if (vio_setHandler(viomajorsubtype_chario, hvc_handle_event) < 0)
+		printk(KERN_WARNING
+			"hvc: error seting handler for console events!\n");
+
+	/*
+	 * First, try to open the console to the hosting lp.
+	 * Wait on a semaphore for the response.
+	 */
+	atomic_set(&wait_flag, 0);
+	if ((viopath_isactive(viopath_hostLp)) &&
+	    (send_open(viopath_hostLp, &wait_flag) == 0)) {
+		printk(KERN_INFO "hvc: hosting partition %d\n", viopath_hostLp);
+		while (atomic_read(&wait_flag) == 0)
+			mb();
+		atomic_set(&wait_flag, 0);
+	}
+
+	/*
+	 * If we don't have an active console, try the primary
+	 */
+	if ((!viopath_isactive(port_info[0].lp)) &&
+	    (viopath_isactive(HvLpConfig_getPrimaryLpIndex())) &&
+	    (send_open(HvLpConfig_getPrimaryLpIndex(), &wait_flag) == 0)) {
+		printk(KERN_INFO "hvc: opening console to primary partition\n");
+		while (atomic_read(&wait_flag) == 0)
+			mb();
+	}
+
+	/* Register as a vio device to receive callbacks */
+	rc = vio_register_driver(&hvc_vio_driver);
+
+	return rc;
+}
+module_init(hvc_vio_init); /* after drivers/char/hvc_console.c */
+
+static void hvc_vio_exit(void)
+{
+	vio_unregister_driver(&hvc_vio_driver);
+}
+module_exit(hvc_vio_exit);
+
+/* the device tree order defines our numbering */
+static int hvc_find_vtys(void)
+{
+	struct device_node *vty;
+	int num_found = 0;
+
+	for (vty = of_find_node_by_name(NULL, "vty"); vty != NULL;
+			vty = of_find_node_by_name(vty, "vty")) {
+		uint32_t *vtermno;
+
+		/* We have statically defined space for only a certain number
+		 * of console adapters.
+		 */
+		if ((num_found >= MAX_NR_HVC_CONSOLES) ||
+				(num_found >= VTTY_PORTS))
+			break;
+
+		vtermno = (uint32_t *)get_property(vty, "reg", NULL);
+		if (!vtermno)
+			continue;
+
+		if (!device_is_compatible(vty, "IBM,iSeries-vty"))
+			continue;
+
+		if (num_found == 0)
+			add_preferred_console("hvc", 0, NULL);
+		hvc_instantiate(*vtermno, num_found, &hvc_get_put_ops);
+		++num_found;
+	}
+
+	return num_found;
+}
+console_initcall(hvc_find_vtys);
diff --git a/drivers/char/viocons.c b/drivers/char/viocons.c
index 766f7864c6c6..f3efeaf2826e 100644
--- a/drivers/char/viocons.c
+++ b/drivers/char/viocons.c
@@ -43,7 +43,6 @@
 #include <linux/sysrq.h>
 
 #include <asm/iseries/vio.h>
-
 #include <asm/iseries/hv_lp_event.h>
 #include <asm/iseries/hv_call_event.h>
 #include <asm/iseries/hv_lp_config.h>
@@ -67,35 +66,6 @@ static int vio_sysrq_pressed;
 extern int sysrq_enabled;
 #endif
 
-/*
- * The structure of the events that flow between us and OS/400.  You can't
- * mess with this unless the OS/400 side changes too
- */
-struct viocharlpevent {
-	struct HvLpEvent event;
-	u32 reserved;
-	u16 version;
-	u16 subtype_result_code;
-	u8 virtual_device;
-	u8 len;
-	u8 data[VIOCHAR_MAX_DATA];
-};
-
-#define VIOCHAR_WINDOW		10
-#define VIOCHAR_HIGHWATERMARK	3
-
-enum viocharsubtype {
-	viocharopen = 0x0001,
-	viocharclose = 0x0002,
-	viochardata = 0x0003,
-	viocharack = 0x0004,
-	viocharconfig = 0x0005
-};
-
-enum viochar_rc {
-	viochar_rc_ebusy = 1
-};
-
 #define VIOCHAR_NUM_BUF		16
 
 /*
@@ -1183,6 +1153,7 @@ static int __init viocons_init(void)
 		port_info[i].magic = VIOTTY_MAGIC;
 	}
 	HvCall_setLogBufferFormatAndCodepage(HvCall_LogBuffer_ASCII, 437);
+	add_preferred_console("viocons", 0, NULL);
 	register_console(&viocons_early);
 	return 0;
 }
diff --git a/include/asm-powerpc/iseries/vio.h b/include/asm-powerpc/iseries/vio.h
index 72a97d37aac3..7a95d296abd1 100644
--- a/include/asm-powerpc/iseries/vio.h
+++ b/include/asm-powerpc/iseries/vio.h
@@ -122,6 +122,34 @@ enum viorc {
 	viorc_openRejected = 0x0301
 };
 
+/*
+ * The structure of the events that flow between us and OS/400 for chario
+ * events.  You can't mess with this unless the OS/400 side changes too.
+ */
+struct viocharlpevent {
+	struct HvLpEvent event;
+	u32 reserved;
+	u16 version;
+	u16 subtype_result_code;
+	u8 virtual_device;
+	u8 len;
+	u8 data[VIOCHAR_MAX_DATA];
+};
+
+#define VIOCHAR_WINDOW		10
+
+enum viocharsubtype {
+	viocharopen = 0x0001,
+	viocharclose = 0x0002,
+	viochardata = 0x0003,
+	viocharack = 0x0004,
+	viocharconfig = 0x0005
+};
+
+enum viochar_rc {
+	viochar_rc_ebusy = 1
+};
+
 struct device;
 
 extern struct device *iSeries_vio_dev;
-- 
cgit v1.2.3


From 54f5cd8afa1c9c9f8b152a946b0a7e0ecdef1631 Mon Sep 17 00:00:00 2001
From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Thu, 13 Jul 2006 18:56:56 +1000
Subject: [POWERPC] iseries: Remove unnecessary include of
 iseries/hv_lp_event.h

Also remove unnecessary reference to struct HvLpEvent.

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 arch/powerpc/kernel/asm-offsets.c         | 1 -
 include/asm-powerpc/iseries/it_lp_queue.h | 2 --
 2 files changed, 3 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c
index 7ee84968087b..ac0631958b20 100644
--- a/arch/powerpc/kernel/asm-offsets.c
+++ b/arch/powerpc/kernel/asm-offsets.c
@@ -40,7 +40,6 @@
 #ifdef CONFIG_PPC64
 #include <asm/paca.h>
 #include <asm/lppaca.h>
-#include <asm/iseries/hv_lp_event.h>
 #include <asm/cache.h>
 #include <asm/compat.h>
 #endif
diff --git a/include/asm-powerpc/iseries/it_lp_queue.h b/include/asm-powerpc/iseries/it_lp_queue.h
index 284c5a7db3ac..3f6814769295 100644
--- a/include/asm-powerpc/iseries/it_lp_queue.h
+++ b/include/asm-powerpc/iseries/it_lp_queue.h
@@ -27,8 +27,6 @@
 #include <asm/types.h>
 #include <asm/ptrace.h>
 
-struct HvLpEvent;
-
 #define IT_LP_MAX_QUEUES	8
 
 #define IT_LP_NOT_USED		0	/* Queue will not be used by PLIC */
-- 
cgit v1.2.3


From ca652c9396fa052815518e2b2ce2ebee6d9fb861 Mon Sep 17 00:00:00 2001
From: Michael Ellerman <michael@ellerman.id.au>
Date: Fri, 14 Jul 2006 14:25:33 +1000
Subject: [POWERPC] iseries: Move iommu_table_cb into platforms/iseries

Although we pass the address of an iommu_table_cb to HvCallXm_getTceTableParms,
we don't actually need the structure definition anywhere except in the
iseries iommu code, so move the struct in there.

Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 arch/powerpc/platforms/iseries/iommu.c   | 17 +++++++++++++++++
 include/asm-powerpc/iseries/hv_call_xm.h | 17 -----------------
 2 files changed, 17 insertions(+), 17 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/platforms/iseries/iommu.c b/arch/powerpc/platforms/iseries/iommu.c
index e3bd2015f2c9..2c3dbcd4613c 100644
--- a/arch/powerpc/platforms/iseries/iommu.c
+++ b/arch/powerpc/platforms/iseries/iommu.c
@@ -87,6 +87,23 @@ static void tce_free_iSeries(struct iommu_table *tbl, long index, long npages)
 	}
 }
 
+/*
+ * Structure passed to HvCallXm_getTceTableParms
+ */
+struct iommu_table_cb {
+	unsigned long	itc_busno;	/* Bus number for this tce table */
+	unsigned long	itc_start;	/* Will be NULL for secondary */
+	unsigned long	itc_totalsize;	/* Size (in pages) of whole table */
+	unsigned long	itc_offset;	/* Index into real tce table of the
+					   start of our section */
+	unsigned long	itc_size;	/* Size (in pages) of our section */
+	unsigned long	itc_index;	/* Index of this tce table */
+	unsigned short	itc_maxtables;	/* Max num of tables for partition */
+	unsigned char	itc_virtbus;	/* Flag to indicate virtual bus */
+	unsigned char	itc_slotno;	/* IOA Tce Slot Index */
+	unsigned char	itc_rsvd[4];
+};
+
 /*
  * Call Hv with the architected data structure to get TCE table info.
  * info. Put the returned data into the Linux representation of the
diff --git a/include/asm-powerpc/iseries/hv_call_xm.h b/include/asm-powerpc/iseries/hv_call_xm.h
index ca9202cb01ed..392ac3f54df0 100644
--- a/include/asm-powerpc/iseries/hv_call_xm.h
+++ b/include/asm-powerpc/iseries/hv_call_xm.h
@@ -16,23 +16,6 @@
 #define HvCallXmSetTce			HvCallXm + 11
 #define HvCallXmSetTces			HvCallXm + 13
 
-/*
- * Structure passed to HvCallXm_getTceTableParms
- */
-struct iommu_table_cb {
-	unsigned long	itc_busno;	/* Bus number for this tce table */
-	unsigned long	itc_start;	/* Will be NULL for secondary */
-	unsigned long	itc_totalsize;	/* Size (in pages) of whole table */
-	unsigned long	itc_offset;	/* Index into real tce table of the
-					   start of our section */
-	unsigned long	itc_size;	/* Size (in pages) of our section */
-	unsigned long	itc_index;	/* Index of this tce table */
-	unsigned short	itc_maxtables;	/* Max num of tables for partition */
-	unsigned char	itc_virtbus;	/* Flag to indicate virtual bus */
-	unsigned char	itc_slotno;	/* IOA Tce Slot Index */
-	unsigned char	itc_rsvd[4];
-};
-
 static inline void HvCallXm_getTceTableParms(u64 cb)
 {
 	HvCall1(HvCallXmGetTceTableParms, cb);
-- 
cgit v1.2.3


From 804af2cf6e7af31d2e664b54e657dddd9b531dbd Mon Sep 17 00:00:00 2001
From: Hugh Dickins <hugh@veritas.com>
Date: Wed, 26 Jul 2006 21:39:49 +0100
Subject: [AGPGART] remove private page protection map

AGP keeps its own copy of the protection_map, upcoming DRM changes will
also require access to this map from modules.

Signed-off-by: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
Signed-off-by: Dave Jones <davej@redhat.com>
---
 drivers/char/agp/frontend.c | 27 ++-------------------------
 include/linux/mm.h          |  1 +
 mm/mmap.c                   |  7 +++++++
 3 files changed, 10 insertions(+), 25 deletions(-)

(limited to 'include')

diff --git a/drivers/char/agp/frontend.c b/drivers/char/agp/frontend.c
index d9c5a9142ad1..0f2ed2aa2d81 100644
--- a/drivers/char/agp/frontend.c
+++ b/drivers/char/agp/frontend.c
@@ -151,35 +151,12 @@ static void agp_add_seg_to_client(struct agp_client *client,
 	client->segments = seg;
 }
 
-/* Originally taken from linux/mm/mmap.c from the array
- * protection_map.
- * The original really should be exported to modules, or
- * some routine which does the conversion for you
- */
-
-static const pgprot_t my_protect_map[16] =
-{
-	__P000, __P001, __P010, __P011, __P100, __P101, __P110, __P111,
-	__S000, __S001, __S010, __S011, __S100, __S101, __S110, __S111
-};
-
 static pgprot_t agp_convert_mmap_flags(int prot)
 {
-#define _trans(x,bit1,bit2) \
-((bit1==bit2)?(x&bit1):(x&bit1)?bit2:0)
-
 	unsigned long prot_bits;
-	pgprot_t temp;
-
-	prot_bits = _trans(prot, PROT_READ, VM_READ) |
-	    _trans(prot, PROT_WRITE, VM_WRITE) |
-	    _trans(prot, PROT_EXEC, VM_EXEC);
-
-	prot_bits |= VM_SHARED;
 
-	temp = my_protect_map[prot_bits & 0x0000000f];
-
-	return temp;
+	prot_bits = calc_vm_prot_bits(prot) | VM_SHARED;
+	return vm_get_page_prot(prot_bits);
 }
 
 static int agp_create_segment(struct agp_client *client, struct agp_region *region)
diff --git a/include/linux/mm.h b/include/linux/mm.h
index 990957e0929f..4fba4560699b 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -1012,6 +1012,7 @@ static inline unsigned long vma_pages(struct vm_area_struct *vma)
 	return (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;
 }
 
+pgprot_t vm_get_page_prot(unsigned long vm_flags);
 struct vm_area_struct *find_extend_vma(struct mm_struct *, unsigned long addr);
 struct page *vmalloc_to_page(void *addr);
 unsigned long vmalloc_to_pfn(void *addr);
diff --git a/mm/mmap.c b/mm/mmap.c
index c1868ecdbc5f..c7ed061f4507 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -60,6 +60,13 @@ pgprot_t protection_map[16] = {
 	__S000, __S001, __S010, __S011, __S100, __S101, __S110, __S111
 };
 
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+	return protection_map[vm_flags &
+				(VM_READ|VM_WRITE|VM_EXEC|VM_SHARED)];
+}
+EXPORT_SYMBOL(vm_get_page_prot);
+
 int sysctl_overcommit_memory = OVERCOMMIT_GUESS;  /* heuristic overcommit */
 int sysctl_overcommit_ratio = 50;	/* default is 50% */
 int sysctl_max_map_count __read_mostly = DEFAULT_MAX_MAP_COUNT;
-- 
cgit v1.2.3


From d8e2be90d301a0381e9b2528fe2835cf2992bca3 Mon Sep 17 00:00:00 2001
From: Daniel Drake <dsd@gentoo.org>
Date: Tue, 18 Jul 2006 21:30:34 +0100
Subject: [PATCH] ieee80211: small ERP handling additions

This adds a flag to the ieee80211_network structure which indicates whether
the stored erp_value is valid (a check against 0 is not enough, since an ERP
of 0 is valid and very meaningful).

I also added the ERP IE bit-definitions to ieee80211.h.

This is needed by some upcoming softmac patches.

Signed-off-by: Daniel Drake <dsd@gentoo.org>
Acked-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 include/net/ieee80211.h      | 7 +++++++
 net/ieee80211/ieee80211_rx.c | 1 +
 2 files changed, 8 insertions(+)

(limited to 'include')

diff --git a/include/net/ieee80211.h b/include/net/ieee80211.h
index ecc42864b001..aa007f49bf29 100644
--- a/include/net/ieee80211.h
+++ b/include/net/ieee80211.h
@@ -240,6 +240,11 @@ struct ieee80211_snap_hdr {
 #define WLAN_CAPABILITY_SHORT_SLOT_TIME (1<<10)
 #define WLAN_CAPABILITY_DSSS_OFDM (1<<13)
 
+/* 802.11g ERP information element */
+#define WLAN_ERP_NON_ERP_PRESENT (1<<0)
+#define WLAN_ERP_USE_PROTECTION (1<<1)
+#define WLAN_ERP_BARKER_PREAMBLE (1<<2)
+
 /* Status codes */
 enum ieee80211_statuscode {
 	WLAN_STATUS_SUCCESS = 0,
@@ -747,6 +752,8 @@ struct ieee80211_txb {
 #define NETWORK_HAS_IBSS_DFS            (1<<8)
 #define NETWORK_HAS_TPC_REPORT          (1<<9)
 
+#define NETWORK_HAS_ERP_VALUE           (1<<10)
+
 #define QOS_QUEUE_NUM                   4
 #define QOS_OUI_LEN                     3
 #define QOS_OUI_TYPE                    2
diff --git a/net/ieee80211/ieee80211_rx.c b/net/ieee80211/ieee80211_rx.c
index 72d4d4e04d42..f73fc164b9ef 100644
--- a/net/ieee80211/ieee80211_rx.c
+++ b/net/ieee80211/ieee80211_rx.c
@@ -1166,6 +1166,7 @@ static int ieee80211_parse_info_param(struct ieee80211_info_element
 
 		case MFIE_TYPE_ERP_INFO:
 			network->erp_value = info_element->data[0];
+			network->flags |= NETWORK_HAS_ERP_VALUE;
 			IEEE80211_DEBUG_MGMT("MFIE_TYPE_ERP_SET: %d\n",
 					     network->erp_value);
 			break;
-- 
cgit v1.2.3


From 5acd0c4153be25269d7cb9a4b09fd6db571c5cc1 Mon Sep 17 00:00:00 2001
From: Daniel Drake <dsd@gentoo.org>
Date: Tue, 18 Jul 2006 21:33:27 +0100
Subject: [PATCH] softmac: ERP handling and driver-level notifications

This patch implements ERP handling in softmac so that the drivers can support
protection and preambles properly.

I added a new struct, ieee80211softmac_bss_info, which is used for
BSS-dependent variables like these.

A new hook has been added (bssinfo_change), which allows the drivers to be
notified when anything in bssinfo changes.

I modified the txrates_change API to match the bssinfo_change API. The
existing one is a little messy and the usefulness of providing the old rates
is questionable (and can be implemented at driver level if really necessary).
No drivers are using this API (yet), so this should be safe.

Signed-off-by: Daniel Drake <dsd@gentoo.org>
Acked-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 include/net/ieee80211softmac.h                  | 52 +++++++++++++++--
 net/ieee80211/softmac/ieee80211softmac_assoc.c  | 21 +++++--
 net/ieee80211/softmac/ieee80211softmac_io.c     | 14 +++++
 net/ieee80211/softmac/ieee80211softmac_module.c | 78 ++++++++++++++++++-------
 net/ieee80211/softmac/ieee80211softmac_priv.h   |  8 ++-
 5 files changed, 142 insertions(+), 31 deletions(-)

(limited to 'include')

diff --git a/include/net/ieee80211softmac.h b/include/net/ieee80211softmac.h
index 00ad810eb883..c27d03433c21 100644
--- a/include/net/ieee80211softmac.h
+++ b/include/net/ieee80211softmac.h
@@ -86,9 +86,6 @@ struct ieee80211softmac_assoc_info {
 	
 	/* BSSID we're trying to associate to */
 	char bssid[ETH_ALEN];
-
-	/* Rates supported by the network */
-	struct ieee80211softmac_ratesinfo supported_rates;
 	
 	/* some flags.
 	 * static_essid is valid if the essid is constant,
@@ -103,6 +100,7 @@ struct ieee80211softmac_assoc_info {
 	 * bssfixed is used for SIOCSIWAP.
 	 */
 	u8 static_essid:1,
+	   short_preamble_available:1,
 	   associating:1,
 	   assoc_wait:1,
 	   bssvalid:1,
@@ -115,6 +113,19 @@ struct ieee80211softmac_assoc_info {
 	struct work_struct timeout;
 };
 
+struct ieee80211softmac_bss_info {
+	/* Rates supported by the network */
+	struct ieee80211softmac_ratesinfo supported_rates;
+
+	/* This indicates whether frames can currently be transmitted with
+	 * short preamble (only use this variable during TX at CCK rates) */
+	u8 short_preamble:1;
+
+	/* This indicates whether protection (e.g. self-CTS) should be used
+	 * when transmitting with OFDM modulation */
+	u8 use_protection:1;
+};
+
 enum {
 	IEEE80211SOFTMAC_AUTH_OPEN_REQUEST	= 1,
 	IEEE80211SOFTMAC_AUTH_OPEN_RESPONSE	= 2,
@@ -157,6 +168,10 @@ struct ieee80211softmac_txrates {
 #define IEEE80211SOFTMAC_TXRATECHG_MCAST		(1 << 2) /* mcast_rate */
 #define IEEE80211SOFTMAC_TXRATECHG_MGT_MCAST		(1 << 3) /* mgt_mcast_rate */
 
+#define IEEE80211SOFTMAC_BSSINFOCHG_RATES		(1 << 0) /* supported_rates */
+#define IEEE80211SOFTMAC_BSSINFOCHG_SHORT_PREAMBLE	(1 << 1) /* short_preamble */
+#define IEEE80211SOFTMAC_BSSINFOCHG_PROTECTION		(1 << 2) /* use_protection */
+
 struct ieee80211softmac_device {
 	/* 802.11 structure for data stuff */
 	struct ieee80211_device *ieee;
@@ -200,10 +215,16 @@ struct ieee80211softmac_device {
 	 * The driver just needs to read them.
 	 */
 	struct ieee80211softmac_txrates txrates;
-	/* If the driver needs to do stuff on TX rate changes, assign this callback. */
+
+	/* If the driver needs to do stuff on TX rate changes, assign this
+	 * callback. See IEEE80211SOFTMAC_TXRATECHG for change flags. */
 	void (*txrates_change)(struct net_device *dev,
-			       u32 changes, /* see IEEE80211SOFTMAC_TXRATECHG flags */
-			       const struct ieee80211softmac_txrates *rates_before_change);
+			       u32 changes);
+
+	/* If the driver needs to do stuff when BSS properties change, assign
+	 * this callback. see IEEE80211SOFTMAC_BSSINFOCHG for change flags. */
+	void (*bssinfo_change)(struct net_device *dev,
+			       u32 changes);
 
 	/* private stuff follows */
 	/* this lock protects this structure */
@@ -216,6 +237,7 @@ struct ieee80211softmac_device {
 	
 	struct ieee80211softmac_scaninfo *scaninfo;
 	struct ieee80211softmac_assoc_info associnfo;
+	struct ieee80211softmac_bss_info bssinfo;
 
 	struct list_head auth_queue;
 	struct list_head events;
@@ -279,6 +301,24 @@ static inline u8 ieee80211softmac_suggest_txrate(struct ieee80211softmac_device
 		return txrates->mcast_rate;
 }
 
+/* Helper function which advises you when it is safe to transmit with short
+ * preamble.
+ * You should only call this function when transmitting at CCK rates. */
+static inline int ieee80211softmac_short_preamble_ok(struct ieee80211softmac_device *mac,
+						    int is_multicast,
+						    int is_mgt)
+{
+	return (is_multicast && is_mgt) ? 0 : mac->bssinfo.short_preamble;
+}
+
+/* Helper function which advises you whether protection (e.g. self-CTS) is
+ * needed. 1 = protection needed, 0 = no protection needed
+ * Only use this function when transmitting with OFDM modulation. */
+static inline int ieee80211softmac_protection_needed(struct ieee80211softmac_device *mac)
+{
+	return mac->bssinfo.use_protection;
+}
+
 /* Start the SoftMAC. Call this after you initialized the device
  * and it is ready to run.
  */
diff --git a/net/ieee80211/softmac/ieee80211softmac_assoc.c b/net/ieee80211/softmac/ieee80211softmac_assoc.c
index 44215ce64d4e..589f6d2c548a 100644
--- a/net/ieee80211/softmac/ieee80211softmac_assoc.c
+++ b/net/ieee80211/softmac/ieee80211softmac_assoc.c
@@ -96,7 +96,7 @@ ieee80211softmac_disassoc(struct ieee80211softmac_device *mac)
 	mac->associated = 0;
 	mac->associnfo.bssvalid = 0;
 	mac->associnfo.associating = 0;
-	ieee80211softmac_init_txrates(mac);
+	ieee80211softmac_init_bss(mac);
 	ieee80211softmac_call_events_locked(mac, IEEE80211SOFTMAC_EVENT_DISASSOCIATED, NULL);
 	spin_unlock_irqrestore(&mac->lock, flags);
 }
@@ -334,11 +334,19 @@ ieee80211softmac_associated(struct ieee80211softmac_device *mac,
 	struct ieee80211_assoc_response * resp,
 	struct ieee80211softmac_network *net)
 {
+	u16 cap = le16_to_cpu(resp->capability);
+	u8 erp_value = net->erp_value;
+
 	mac->associnfo.associating = 0;
-	mac->associnfo.supported_rates = net->supported_rates;
+	mac->bssinfo.supported_rates = net->supported_rates;
 	ieee80211softmac_recalc_txrates(mac);
 
 	mac->associated = 1;
+
+	mac->associnfo.short_preamble_available =
+		(cap & WLAN_CAPABILITY_SHORT_PREAMBLE) != 0;
+	ieee80211softmac_process_erp(mac, erp_value);
+
 	if (mac->set_bssid_filter)
 		mac->set_bssid_filter(mac->dev, net->bssid);
 	memcpy(mac->ieee->bssid, net->bssid, ETH_ALEN);
@@ -351,9 +359,9 @@ ieee80211softmac_associated(struct ieee80211softmac_device *mac,
 int
 ieee80211softmac_handle_assoc_response(struct net_device * dev,
 				       struct ieee80211_assoc_response * resp,
-				       struct ieee80211_network * _ieee80211_network_do_not_use)
+				       struct ieee80211_network * _ieee80211_network)
 {
-	/* NOTE: the network parameter has to be ignored by
+	/* NOTE: the network parameter has to be mostly ignored by
 	 *       this code because it is the ieee80211's pointer
 	 *       to the struct, not ours (we made a copy)
 	 */
@@ -385,6 +393,11 @@ ieee80211softmac_handle_assoc_response(struct net_device * dev,
 	/* now that we know it was for us, we can cancel the timeout */
 	cancel_delayed_work(&mac->associnfo.timeout);
 
+	/* if the association response included an ERP IE, update our saved
+	 * copy */
+	if (_ieee80211_network->flags & NETWORK_HAS_ERP_VALUE)
+		network->erp_value = _ieee80211_network->erp_value;
+
 	switch (status) {
 		case 0:
 			dprintk(KERN_INFO PFX "associated!\n");
diff --git a/net/ieee80211/softmac/ieee80211softmac_io.c b/net/ieee80211/softmac/ieee80211softmac_io.c
index 6ae5a1dc7956..82bfddbf33a2 100644
--- a/net/ieee80211/softmac/ieee80211softmac_io.c
+++ b/net/ieee80211/softmac/ieee80211softmac_io.c
@@ -467,3 +467,17 @@ ieee80211softmac_send_mgt_frame(struct ieee80211softmac_device *mac,
 	kfree(pkt);
 	return 0;
 }
+
+/* Beacon handling */
+int ieee80211softmac_handle_beacon(struct net_device *dev,
+	struct ieee80211_beacon *beacon,
+	struct ieee80211_network *network)
+{
+	struct ieee80211softmac_device *mac = ieee80211_priv(dev);
+
+	if (mac->associated && memcmp(network->bssid, mac->associnfo.bssid, ETH_ALEN) == 0)
+		ieee80211softmac_process_erp(mac, network->erp_value);
+
+	return 0;
+}
+
diff --git a/net/ieee80211/softmac/ieee80211softmac_module.c b/net/ieee80211/softmac/ieee80211softmac_module.c
index 4b2e57d12418..c275646b2269 100644
--- a/net/ieee80211/softmac/ieee80211softmac_module.c
+++ b/net/ieee80211/softmac/ieee80211softmac_module.c
@@ -44,6 +44,7 @@ struct net_device *alloc_ieee80211softmac(int sizeof_priv)
 	softmac->ieee->handle_assoc_response = ieee80211softmac_handle_assoc_response;
 	softmac->ieee->handle_reassoc_request = ieee80211softmac_handle_reassoc_req;
 	softmac->ieee->handle_disassoc = ieee80211softmac_handle_disassoc;
+ 	softmac->ieee->handle_beacon = ieee80211softmac_handle_beacon;
 	softmac->scaninfo = NULL;
 
 	softmac->associnfo.scan_retry = IEEE80211SOFTMAC_ASSOC_SCAN_RETRY_LIMIT;
@@ -209,35 +210,59 @@ static u8 highest_supported_rate(struct ieee80211softmac_device *mac,
 	return user_rate;
 }
 
+void ieee80211softmac_process_erp(struct ieee80211softmac_device *mac,
+	u8 erp_value)
+{
+ 	int use_protection;
+	int short_preamble;
+ 	u32 changes = 0;
+
+	/* Barker preamble mode */
+	short_preamble = ((erp_value & WLAN_ERP_BARKER_PREAMBLE) == 0
+			  && mac->associnfo.short_preamble_available) ? 1 : 0;
+
+	/* Protection needed? */
+	use_protection = (erp_value & WLAN_ERP_USE_PROTECTION) != 0;
+
+	if (mac->bssinfo.short_preamble != short_preamble) {
+		changes |= IEEE80211SOFTMAC_BSSINFOCHG_SHORT_PREAMBLE;
+		mac->bssinfo.short_preamble = short_preamble;
+	}
+
+	if (mac->bssinfo.use_protection != use_protection) {
+		changes |= IEEE80211SOFTMAC_BSSINFOCHG_PROTECTION;
+		mac->bssinfo.use_protection = use_protection;
+	}
+
+	if (mac->bssinfo_change && changes)
+		mac->bssinfo_change(mac->dev, changes);
+}
+
 void ieee80211softmac_recalc_txrates(struct ieee80211softmac_device *mac)
 {
 	struct ieee80211softmac_txrates *txrates = &mac->txrates;
-	struct ieee80211softmac_txrates oldrates;
 	u32 change = 0;
 
-	if (mac->txrates_change)
-		oldrates = mac->txrates;
-
 	change |= IEEE80211SOFTMAC_TXRATECHG_DEFAULT;
-	txrates->default_rate = highest_supported_rate(mac, &mac->associnfo.supported_rates, 0);
+	txrates->default_rate = highest_supported_rate(mac, &mac->bssinfo.supported_rates, 0);
 
 	change |= IEEE80211SOFTMAC_TXRATECHG_DEFAULT_FBACK;
 	txrates->default_fallback = lower_rate(mac, txrates->default_rate);
 
 	change |= IEEE80211SOFTMAC_TXRATECHG_MCAST;
-	txrates->mcast_rate = highest_supported_rate(mac, &mac->associnfo.supported_rates, 1);
+	txrates->mcast_rate = highest_supported_rate(mac, &mac->bssinfo.supported_rates, 1);
 
 	if (mac->txrates_change)
-		mac->txrates_change(mac->dev, change, &oldrates);
+		mac->txrates_change(mac->dev, change);
 
 }
 
-void ieee80211softmac_init_txrates(struct ieee80211softmac_device *mac)
+void ieee80211softmac_init_bss(struct ieee80211softmac_device *mac)
 {
 	struct ieee80211_device *ieee = mac->ieee;
 	u32 change = 0;
 	struct ieee80211softmac_txrates *txrates = &mac->txrates;
-	struct ieee80211softmac_txrates oldrates;
+	struct ieee80211softmac_bss_info *bssinfo = &mac->bssinfo;
 
 	/* TODO: We need some kind of state machine to lower the default rates
 	 *       if we loose too many packets.
@@ -245,8 +270,6 @@ void ieee80211softmac_init_txrates(struct ieee80211softmac_device *mac)
 	/* Change the default txrate to the highest possible value.
 	 * The txrate machine will lower it, if it is too high.
 	 */
-	if (mac->txrates_change)
-		oldrates = mac->txrates;
 	/* FIXME: We don't correctly handle backing down to lower
 	   rates, so 801.11g devices start off at 11M for now. People
 	   can manually change it if they really need to, but 11M is
@@ -272,7 +295,23 @@ void ieee80211softmac_init_txrates(struct ieee80211softmac_device *mac)
 	change |= IEEE80211SOFTMAC_TXRATECHG_MGT_MCAST;
 
 	if (mac->txrates_change)
-		mac->txrates_change(mac->dev, change, &oldrates);
+		mac->txrates_change(mac->dev, change);
+
+	change = 0;
+
+	bssinfo->supported_rates.count = 0;
+	memset(bssinfo->supported_rates.rates, 0,
+		sizeof(bssinfo->supported_rates.rates));
+	change |= IEEE80211SOFTMAC_BSSINFOCHG_RATES;
+
+	bssinfo->short_preamble = 0;
+	change |= IEEE80211SOFTMAC_BSSINFOCHG_SHORT_PREAMBLE;
+
+	bssinfo->use_protection = 0;
+	change |= IEEE80211SOFTMAC_BSSINFOCHG_PROTECTION;
+
+	if (mac->bssinfo_change)
+		mac->bssinfo_change(mac->dev, change);
 
 	mac->running = 1;
 }
@@ -282,7 +321,7 @@ void ieee80211softmac_start(struct net_device *dev)
 	struct ieee80211softmac_device *mac = ieee80211_priv(dev);
 
 	ieee80211softmac_start_check_rates(mac);
-	ieee80211softmac_init_txrates(mac);
+	ieee80211softmac_init_bss(mac);
 }
 EXPORT_SYMBOL_GPL(ieee80211softmac_start);
 
@@ -335,7 +374,6 @@ u8 ieee80211softmac_lower_rate_delta(struct ieee80211softmac_device *mac, u8 rat
 static void ieee80211softmac_add_txrates_badness(struct ieee80211softmac_device *mac,
 						 int amount)
 {
-	struct ieee80211softmac_txrates oldrates;
 	u8 default_rate = mac->txrates.default_rate;
 	u8 default_fallback = mac->txrates.default_fallback;
 	u32 changes = 0;
@@ -348,8 +386,6 @@ printk("badness %d\n", mac->txrate_badness);
 	mac->txrate_badness += amount;
 	if (mac->txrate_badness <= -1000) {
 		/* Very small badness. Try a faster bitrate. */
-		if (mac->txrates_change)
-			memcpy(&oldrates, &mac->txrates, sizeof(oldrates));
 		default_rate = raise_rate(mac, default_rate);
 		changes |= IEEE80211SOFTMAC_TXRATECHG_DEFAULT;
 		default_fallback = get_fallback_rate(mac, default_rate);
@@ -358,8 +394,6 @@ printk("badness %d\n", mac->txrate_badness);
 printk("Bitrate raised to %u\n", default_rate);
 	} else if (mac->txrate_badness >= 10000) {
 		/* Very high badness. Try a slower bitrate. */
-		if (mac->txrates_change)
-			memcpy(&oldrates, &mac->txrates, sizeof(oldrates));
 		default_rate = lower_rate(mac, default_rate);
 		changes |= IEEE80211SOFTMAC_TXRATECHG_DEFAULT;
 		default_fallback = get_fallback_rate(mac, default_rate);
@@ -372,7 +406,7 @@ printk("Bitrate lowered to %u\n", default_rate);
 	mac->txrates.default_fallback = default_fallback;
 
 	if (changes && mac->txrates_change)
-		mac->txrates_change(mac->dev, changes, &oldrates);
+		mac->txrates_change(mac->dev, changes);
 }
 
 void ieee80211softmac_fragment_lost(struct net_device *dev,
@@ -416,7 +450,11 @@ ieee80211softmac_create_network(struct ieee80211softmac_device *mac,
 	memcpy(&softnet->supported_rates.rates[softnet->supported_rates.count], net->rates_ex, net->rates_ex_len);
 	softnet->supported_rates.count += net->rates_ex_len;
 	sort(softnet->supported_rates.rates, softnet->supported_rates.count, sizeof(softnet->supported_rates.rates[0]), rate_cmp, NULL);
-	
+
+	/* we save the ERP value because it is needed at association time, and
+	 * many AP's do not include an ERP IE in the association response. */
+	softnet->erp_value = net->erp_value;
+
 	softnet->capabilities = net->capability;
 	return softnet;
 }
diff --git a/net/ieee80211/softmac/ieee80211softmac_priv.h b/net/ieee80211/softmac/ieee80211softmac_priv.h
index fa1f8e3acfc0..0642e090b8a7 100644
--- a/net/ieee80211/softmac/ieee80211softmac_priv.h
+++ b/net/ieee80211/softmac/ieee80211softmac_priv.h
@@ -116,9 +116,11 @@ ieee80211softmac_get_network_by_essid(struct ieee80211softmac_device *mac,
 	struct ieee80211softmac_essid *essid);
 
 /* Rates related */
+void ieee80211softmac_process_erp(struct ieee80211softmac_device *mac,
+	u8 erp_value);
 int ieee80211softmac_ratesinfo_rate_supported(struct ieee80211softmac_ratesinfo *ri, u8 rate);
 u8 ieee80211softmac_lower_rate_delta(struct ieee80211softmac_device *mac, u8 rate, int delta);
-void ieee80211softmac_init_txrates(struct ieee80211softmac_device *mac);
+void ieee80211softmac_init_bss(struct ieee80211softmac_device *mac);
 void ieee80211softmac_recalc_txrates(struct ieee80211softmac_device *mac);
 static inline u8 lower_rate(struct ieee80211softmac_device *mac, u8 rate) {
 	return ieee80211softmac_lower_rate_delta(mac, rate, 1);
@@ -133,6 +135,9 @@ static inline u8 get_fallback_rate(struct ieee80211softmac_device *mac, u8 rate)
 /*** prototypes from _io.c */
 int ieee80211softmac_send_mgt_frame(struct ieee80211softmac_device *mac,
 	void* ptrarg, u32 type, u32 arg);
+int ieee80211softmac_handle_beacon(struct net_device *dev,
+	struct ieee80211_beacon *beacon,
+	struct ieee80211_network *network);
 
 /*** prototypes from _auth.c */
 /* do these have to go into the public header? */
@@ -189,6 +194,7 @@ struct ieee80211softmac_network {
 	    authenticated:1,
 	    auth_desynced_once:1;
 
+	u8 erp_value;				/* Saved ERP value */
 	u16 capabilities;			/* Capabilities bitfield */
 	u8 challenge_len;			/* Auth Challenge length */
 	char *challenge;			/* Challenge Text */
-- 
cgit v1.2.3


From d7712ac254a4ae2e9c927e29e37b8c7ac334e6ad Mon Sep 17 00:00:00 2001
From: Daniel Drake <dsd@gentoo.org>
Date: Tue, 18 Jul 2006 21:34:56 +0100
Subject: [PATCH] softmac: export highest_supported_rate function

zd1211 needs this functionality, no point duplicating it.

Signed-off-by: Daniel Drake <dsd@gentoo.org>
Acked-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 include/net/ieee80211softmac.h                  |  8 ++++++++
 net/ieee80211/softmac/ieee80211softmac_module.c | 16 +++++-----------
 2 files changed, 13 insertions(+), 11 deletions(-)

(limited to 'include')

diff --git a/include/net/ieee80211softmac.h b/include/net/ieee80211softmac.h
index c27d03433c21..425b3a57ac74 100644
--- a/include/net/ieee80211softmac.h
+++ b/include/net/ieee80211softmac.h
@@ -279,6 +279,14 @@ extern void ieee80211softmac_fragment_lost(struct net_device *dev,
  * Note that the rates need to be sorted. */
 extern void ieee80211softmac_set_rates(struct net_device *dev, u8 count, u8 *rates);
 
+/* Finds the highest rate which is:
+ *  1. Present in ri (optionally a basic rate)
+ *  2. Supported by the device
+ *  3. Less than or equal to the user-defined rate
+ */
+extern u8 ieee80211softmac_highest_supported_rate(struct ieee80211softmac_device *mac,
+	struct ieee80211softmac_ratesinfo *ri, int basic_only);
+
 /* Helper function which advises you the rate at which a frame should be
  * transmitted at. */
 static inline u8 ieee80211softmac_suggest_txrate(struct ieee80211softmac_device *mac,
diff --git a/net/ieee80211/softmac/ieee80211softmac_module.c b/net/ieee80211/softmac/ieee80211softmac_module.c
index c275646b2269..addea1cf73ae 100644
--- a/net/ieee80211/softmac/ieee80211softmac_module.c
+++ b/net/ieee80211/softmac/ieee80211softmac_module.c
@@ -179,21 +179,14 @@ int ieee80211softmac_ratesinfo_rate_supported(struct ieee80211softmac_ratesinfo
 	return 0;
 }
 
-/* Finds the highest rate which is:
- *  1. Present in ri (optionally a basic rate)
- *  2. Supported by the device
- *  3. Less than or equal to the user-defined rate
- */
-static u8 highest_supported_rate(struct ieee80211softmac_device *mac,
+u8 ieee80211softmac_highest_supported_rate(struct ieee80211softmac_device *mac,
 	struct ieee80211softmac_ratesinfo *ri, int basic_only)
 {
 	u8 user_rate = mac->txrates.user_rate;
 	int i;
 
-	if (ri->count == 0) {
-		dprintk(KERN_ERR PFX "empty ratesinfo?\n");
+	if (ri->count == 0)
 		return IEEE80211_CCK_RATE_1MB;
-	}
 
 	for (i = ri->count - 1; i >= 0; i--) {
 		u8 rate = ri->rates[i];
@@ -209,6 +202,7 @@ static u8 highest_supported_rate(struct ieee80211softmac_device *mac,
 	/* If we haven't found a suitable rate by now, just trust the user */
 	return user_rate;
 }
+EXPORT_SYMBOL_GPL(ieee80211softmac_highest_supported_rate);
 
 void ieee80211softmac_process_erp(struct ieee80211softmac_device *mac,
 	u8 erp_value)
@@ -244,13 +238,13 @@ void ieee80211softmac_recalc_txrates(struct ieee80211softmac_device *mac)
 	u32 change = 0;
 
 	change |= IEEE80211SOFTMAC_TXRATECHG_DEFAULT;
-	txrates->default_rate = highest_supported_rate(mac, &mac->bssinfo.supported_rates, 0);
+	txrates->default_rate = ieee80211softmac_highest_supported_rate(mac, &mac->bssinfo.supported_rates, 0);
 
 	change |= IEEE80211SOFTMAC_TXRATECHG_DEFAULT_FBACK;
 	txrates->default_fallback = lower_rate(mac, txrates->default_rate);
 
 	change |= IEEE80211SOFTMAC_TXRATECHG_MCAST;
-	txrates->mcast_rate = highest_supported_rate(mac, &mac->bssinfo.supported_rates, 1);
+	txrates->mcast_rate = ieee80211softmac_highest_supported_rate(mac, &mac->bssinfo.supported_rates, 1);
 
 	if (mac->txrates_change)
 		mac->txrates_change(mac->dev, change);
-- 
cgit v1.2.3


From f2060f039e8a8bc83b10e6d0f8fb440425560569 Mon Sep 17 00:00:00 2001
From: Daniel Drake <dsd@gentoo.org>
Date: Tue, 18 Jul 2006 21:38:05 +0100
Subject: [PATCH] ieee80211: Make ieee80211_rx_any usable

ieee80211_rx_any is new to 2.6.18-rc1, even though it appears this function
was never completed:

http://lists.sipsolutions.net/pipermail/softmac-dev/2006-February/000103.html

This patch changes ieee80211_rx_any to always claim the skb, which avoids
further driver complexity and the possibility of leaking management frames.
It also exports the function so that people can actually use it.

Signed-off-by: Daniel Drake <dsd@gentoo.org>
Acked-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 include/net/ieee80211.h      |  2 ++
 net/ieee80211/ieee80211_rx.c | 38 ++++++++++++++++++++++++++++----------
 2 files changed, 30 insertions(+), 10 deletions(-)

(limited to 'include')

diff --git a/include/net/ieee80211.h b/include/net/ieee80211.h
index aa007f49bf29..b174ebb277a9 100644
--- a/include/net/ieee80211.h
+++ b/include/net/ieee80211.h
@@ -1259,6 +1259,8 @@ extern int ieee80211_tx_frame(struct ieee80211_device *ieee,
 			      int total_len, int encrypt_mpdu);
 
 /* ieee80211_rx.c */
+extern void ieee80211_rx_any(struct ieee80211_device *ieee,
+		     struct sk_buff *skb, struct ieee80211_rx_stats *stats);
 extern int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb,
 			struct ieee80211_rx_stats *rx_stats);
 /* make sure to set stats->len */
diff --git a/net/ieee80211/ieee80211_rx.c b/net/ieee80211/ieee80211_rx.c
index f73fc164b9ef..d60358d702d7 100644
--- a/net/ieee80211/ieee80211_rx.c
+++ b/net/ieee80211/ieee80211_rx.c
@@ -779,33 +779,44 @@ int ieee80211_rx(struct ieee80211_device *ieee, struct sk_buff *skb,
 	return 0;
 }
 
-/* Filter out unrelated packets, call ieee80211_rx[_mgt] */
-int ieee80211_rx_any(struct ieee80211_device *ieee,
+/* Filter out unrelated packets, call ieee80211_rx[_mgt]
+ * This function takes over the skb, it should not be used again after calling
+ * this function. */
+void ieee80211_rx_any(struct ieee80211_device *ieee,
 		     struct sk_buff *skb, struct ieee80211_rx_stats *stats)
 {
 	struct ieee80211_hdr_4addr *hdr;
 	int is_packet_for_us;
 	u16 fc;
 
-	if (ieee->iw_mode == IW_MODE_MONITOR)
-		return ieee80211_rx(ieee, skb, stats) ? 0 : -EINVAL;
+	if (ieee->iw_mode == IW_MODE_MONITOR) {
+		if (!ieee80211_rx(ieee, skb, stats))
+			dev_kfree_skb_irq(skb);
+		return;
+	}
+
+	if (skb->len < sizeof(struct ieee80211_hdr))
+		goto drop_free;
 
 	hdr = (struct ieee80211_hdr_4addr *)skb->data;
 	fc = le16_to_cpu(hdr->frame_ctl);
 
 	if ((fc & IEEE80211_FCTL_VERS) != 0)
-		return -EINVAL;
+		goto drop_free;
 		
 	switch (fc & IEEE80211_FCTL_FTYPE) {
 	case IEEE80211_FTYPE_MGMT:
+		if (skb->len < sizeof(struct ieee80211_hdr_3addr))
+			goto drop_free;
 		ieee80211_rx_mgt(ieee, hdr, stats);
-		return 0;
+		dev_kfree_skb_irq(skb);
+		return;
 	case IEEE80211_FTYPE_DATA:
 		break;
 	case IEEE80211_FTYPE_CTL:
-		return 0;
+		return;
 	default:
-		return -EINVAL;
+		return;
 	}
 
 	is_packet_for_us = 0;
@@ -849,8 +860,14 @@ int ieee80211_rx_any(struct ieee80211_device *ieee,
 	}
 
 	if (is_packet_for_us)
-		return (ieee80211_rx(ieee, skb, stats) ? 0 : -EINVAL);
-	return 0;
+		if (!ieee80211_rx(ieee, skb, stats))
+			dev_kfree_skb_irq(skb);
+	return;
+
+drop_free:
+	dev_kfree_skb_irq(skb);
+	ieee->stats.rx_dropped++;
+	return;
 }
 
 #define MGMT_FRAME_FIXED_PART_LENGTH		0x24
@@ -1730,5 +1747,6 @@ void ieee80211_rx_mgt(struct ieee80211_device *ieee,
 	}
 }
 
+EXPORT_SYMBOL_GPL(ieee80211_rx_any);
 EXPORT_SYMBOL(ieee80211_rx_mgt);
 EXPORT_SYMBOL(ieee80211_rx);
-- 
cgit v1.2.3


From 1c57e86d75cf162bdadb3a5fe0cd3f65aa1a9ca3 Mon Sep 17 00:00:00 2001
From: Erich Chen <erich@areca.com.tw>
Date: Wed, 12 Jul 2006 08:59:32 -0700
Subject: [SCSI] arcmsr: initial driver, version 1.20.00.13

arcmsr is a driver for the Areca Raid controller, a host based RAID
subsystem that speaks SCSI at the firmware level.

This patch is quite a clean up over the initial submission with
contributions from:

Randy Dunlap <rdunlap@xenotime.net>
Christoph Hellwig <hch@lst.de>
Matthew Wilcox <matthew@wil.cx>
Adrian Bunk <bunk@stusta.de>

Signed-off-by: Erich Chen <erich@areca.com.tw>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 Documentation/scsi/ChangeLog.arcmsr |   56 ++
 Documentation/scsi/arcmsr_spec.txt  |  574 ++++++++++++++
 drivers/scsi/Kconfig                |   14 +
 drivers/scsi/Makefile               |    1 +
 drivers/scsi/arcmsr/Makefile        |    6 +
 drivers/scsi/arcmsr/arcmsr.h        |  472 +++++++++++
 drivers/scsi/arcmsr/arcmsr_attr.c   |  392 +++++++++
 drivers/scsi/arcmsr/arcmsr_hba.c    | 1496 +++++++++++++++++++++++++++++++++++
 include/linux/pci_ids.h             |   17 +
 9 files changed, 3028 insertions(+)
 create mode 100644 Documentation/scsi/ChangeLog.arcmsr
 create mode 100644 Documentation/scsi/arcmsr_spec.txt
 create mode 100644 drivers/scsi/arcmsr/Makefile
 create mode 100644 drivers/scsi/arcmsr/arcmsr.h
 create mode 100644 drivers/scsi/arcmsr/arcmsr_attr.c
 create mode 100644 drivers/scsi/arcmsr/arcmsr_hba.c

(limited to 'include')

diff --git a/Documentation/scsi/ChangeLog.arcmsr b/Documentation/scsi/ChangeLog.arcmsr
new file mode 100644
index 000000000000..162c47fdf45f
--- /dev/null
+++ b/Documentation/scsi/ChangeLog.arcmsr
@@ -0,0 +1,56 @@
+**************************************************************************
+** History
+**
+**   REV#         DATE             NAME         DESCRIPTION
+** 1.00.00.00    3/31/2004       Erich Chen     First release
+** 1.10.00.04    7/28/2004       Erich Chen     modify for ioctl
+** 1.10.00.06    8/28/2004       Erich Chen     modify for 2.6.x
+** 1.10.00.08    9/28/2004       Erich Chen     modify for x86_64
+** 1.10.00.10   10/10/2004       Erich Chen     bug fix for SMP & ioctl
+** 1.20.00.00   11/29/2004       Erich Chen     bug fix with arcmsr_bus_reset when PHY error
+** 1.20.00.02   12/09/2004       Erich Chen     bug fix with over 2T bytes RAID Volume
+** 1.20.00.04    1/09/2005       Erich Chen     fits for Debian linux kernel version 2.2.xx
+** 1.20.00.05    2/20/2005       Erich Chen     cleanly as look like a Linux driver at 2.6.x
+**                                              thanks for peoples kindness comment
+**						Kornel Wieliczek
+**						Christoph Hellwig
+**						Adrian Bunk
+**						Andrew Morton
+**						Christoph Hellwig
+**						James Bottomley
+**						Arjan van de Ven
+** 1.20.00.06    3/12/2005       Erich Chen     fix with arcmsr_pci_unmap_dma "unsigned long" cast,
+**						modify PCCB POOL allocated by "dma_alloc_coherent"
+**						(Kornel Wieliczek's comment)
+** 1.20.00.07    3/23/2005       Erich Chen     bug fix with arcmsr_scsi_host_template_init
+**						occur segmentation fault,
+**						if RAID adapter does not on PCI slot
+**						and modprobe/rmmod this driver twice.
+**						bug fix enormous stack usage (Adrian Bunk's comment)
+** 1.20.00.08    6/23/2005       Erich Chen     bug fix with abort command,
+**						in case of heavy loading when sata cable
+**						working on low quality connection
+** 1.20.00.09    9/12/2005       Erich Chen     bug fix with abort command handling, firmware version check
+**						and firmware update notify for hardware bug fix
+** 1.20.00.10    9/23/2005       Erich Chen     enhance sysfs function for change driver's max tag Q number.
+**						add DMA_64BIT_MASK for backward compatible with all 2.6.x
+**						add some useful message for abort command
+**						add ioctl code 'ARCMSR_IOCTL_FLUSH_ADAPTER_CACHE'
+**						customer can send this command for sync raid volume data
+** 1.20.00.11    9/29/2005       Erich Chen     by comment of Arjan van de Ven fix incorrect msleep redefine
+**						cast off sizeof(dma_addr_t) condition for 64bit pci_set_dma_mask
+** 1.20.00.12    9/30/2005       Erich Chen     bug fix with 64bit platform's ccbs using if over 4G system memory
+**						change 64bit pci_set_consistent_dma_mask into 32bit
+**						increcct adapter count if adapter initialize fail.
+**						miss edit at arcmsr_build_ccb....
+**						psge += sizeof(struct _SG64ENTRY *) =>
+**						psge += sizeof(struct _SG64ENTRY)
+**						64 bits sg entry would be incorrectly calculated
+**						thanks Kornel Wieliczek give me kindly notify
+**						and detail description
+** 1.20.00.13   11/15/2005       Erich Chen     scheduling pending ccb with FIFO
+**						change the architecture of arcmsr command queue list
+**						for linux standard list
+**						enable usage of pci message signal interrupt
+**						follow Randy.Danlup kindness suggestion cleanup this code
+**************************************************************************
\ No newline at end of file
diff --git a/Documentation/scsi/arcmsr_spec.txt b/Documentation/scsi/arcmsr_spec.txt
new file mode 100644
index 000000000000..5e0042340fd3
--- /dev/null
+++ b/Documentation/scsi/arcmsr_spec.txt
@@ -0,0 +1,574 @@
+*******************************************************************************
+**                            ARECA FIRMWARE SPEC
+*******************************************************************************
+**	Usage of IOP331 adapter
+**	(All In/Out is in IOP331's view)
+**	1. Message 0 --> InitThread message and retrun code
+**	2. Doorbell is used for RS-232 emulation
+**		inDoorBell :    bit0 -- data in ready
+**			(DRIVER DATA WRITE OK)
+**				bit1 -- data out has been read
+**			(DRIVER DATA READ OK)
+**		outDooeBell:    bit0 -- data out ready
+**			(IOP331 DATA WRITE OK)
+**				bit1 -- data in has been read
+**			(IOP331 DATA READ OK)
+**	3. Index Memory Usage
+**	offset 0xf00 : for RS232 out (request buffer)
+**	offset 0xe00 : for RS232 in  (scratch buffer)
+**	offset 0xa00 : for inbound message code message_rwbuffer
+**			(driver send to IOP331)
+**	offset 0xa00 : for outbound message code message_rwbuffer
+**			(IOP331 send to driver)
+**	4. RS-232 emulation
+**		Currently 128 byte buffer is used
+**			1st uint32_t : Data length (1--124)
+**			Byte 4--127  : Max 124 bytes of data
+**	5. PostQ
+**	All SCSI Command must be sent through postQ:
+**	(inbound queue port)	Request frame must be 32 bytes aligned
+**	#bit27--bit31 => flag for post ccb
+**	#bit0--bit26  => real address (bit27--bit31) of post arcmsr_cdb
+**		bit31 :
+**			0 : 256 bytes frame
+**			1 : 512 bytes frame
+**		bit30 :
+**			0 : normal request
+**			1 : BIOS request
+**		bit29 : reserved
+**		bit28 : reserved
+**		bit27 : reserved
+**  ---------------------------------------------------------------------------
+**	(outbount queue port)	Request reply
+**	#bit27--bit31
+**		=> flag for reply
+**	#bit0--bit26
+**		=> real address (bit27--bit31) of reply arcmsr_cdb
+**			bit31 : must be 0 (for this type of reply)
+**			bit30 : reserved for BIOS handshake
+**			bit29 : reserved
+**			bit28 :
+**			0 : no error, ignore AdapStatus/DevStatus/SenseData
+**			1 : Error, error code in AdapStatus/DevStatus/SenseData
+**			bit27 : reserved
+**	6. BIOS request
+**		All BIOS request is the same with request from PostQ
+**		Except :
+**			Request frame is sent from configuration space
+**		offset: 0x78 : Request Frame (bit30 == 1)
+**		offset: 0x18 : writeonly to generate
+**					IRQ to IOP331
+**		Completion of request:
+**			(bit30 == 0, bit28==err flag)
+**	7. Definition of SGL entry (structure)
+**	8. Message1 Out - Diag Status Code (????)
+**	9. Message0 message code :
+**		0x00 : NOP
+**		0x01 : Get Config
+**		->offset 0xa00 :for outbound message code message_rwbuffer
+**		(IOP331 send to driver)
+**		Signature             0x87974060(4)
+**		Request len           0x00000200(4)
+**		numbers of queue      0x00000100(4)
+**		SDRAM Size            0x00000100(4)-->256 MB
+**		IDE Channels          0x00000008(4)
+**		vendor                40 bytes char
+**		model                  8 bytes char
+**		FirmVer               16 bytes char
+**		Device Map            16 bytes char
+**		FirmwareVersion DWORD <== Added for checking of
+**						new firmware capability
+**		0x02 : Set Config
+**		->offset 0xa00 :for inbound message code message_rwbuffer
+**		(driver send to IOP331)
+**		Signature             0x87974063(4)
+**		UPPER32 of Request Frame  (4)-->Driver Only
+**		0x03 : Reset (Abort all queued Command)
+**		0x04 : Stop Background Activity
+**		0x05 : Flush Cache
+**		0x06 : Start Background Activity
+**			(re-start if background is halted)
+**		0x07 : Check If Host Command Pending
+**			(Novell May Need This Function)
+**		0x08 : Set controller time
+**		->offset 0xa00 : for inbound message code message_rwbuffer
+**		(driver to IOP331)
+**		byte 0 : 0xaa <-- signature
+**		byte 1 : 0x55 <-- signature
+**		byte 2 : year (04)
+**		byte 3 : month (1..12)
+**		byte 4 : date (1..31)
+**		byte 5 : hour (0..23)
+**		byte 6 : minute (0..59)
+**		byte 7 : second (0..59)
+*******************************************************************************
+*******************************************************************************
+**      	RS-232 Interface for Areca Raid Controller
+**      The low level command interface is exclusive with VT100 terminal
+**  --------------------------------------------------------------------
+**      1. Sequence of command execution
+**  --------------------------------------------------------------------
+**    	(A) Header : 3 bytes sequence (0x5E, 0x01, 0x61)
+**    	(B) Command block : variable length of data including length,
+**    		command code, data and checksum byte
+**    	(C) Return data : variable length of data
+**  --------------------------------------------------------------------
+**    2. Command block
+**  --------------------------------------------------------------------
+**    	(A) 1st byte : command block length (low byte)
+**    	(B) 2nd byte : command block length (high byte)
+**                note ..command block length shouldn't > 2040 bytes,
+**    		length excludes these two bytes
+**    	(C) 3rd byte : command code
+**    	(D) 4th and following bytes : variable length data bytes
+**    		depends on command code
+**    	(E) last byte : checksum byte (sum of 1st byte until last data byte)
+**  --------------------------------------------------------------------
+**    3. Command code and associated data
+**  --------------------------------------------------------------------
+**    	The following are command code defined in raid controller Command
+**    	code 0x10--0x1? are used for system level management,
+**    	no password checking is needed and should be implemented in separate
+**    	well controlled utility and not for end user access.
+**    	Command code 0x20--0x?? always check the password,
+**    	password must be entered to enable these command.
+**    	enum
+**    	{
+**    		GUI_SET_SERIAL=0x10,
+**    		GUI_SET_VENDOR,
+**    		GUI_SET_MODEL,
+**    		GUI_IDENTIFY,
+**    		GUI_CHECK_PASSWORD,
+**    		GUI_LOGOUT,
+**    		GUI_HTTP,
+**    		GUI_SET_ETHERNET_ADDR,
+**    		GUI_SET_LOGO,
+**    		GUI_POLL_EVENT,
+**    		GUI_GET_EVENT,
+**    		GUI_GET_HW_MONITOR,
+**    		//    GUI_QUICK_CREATE=0x20, (function removed)
+**    		GUI_GET_INFO_R=0x20,
+**    		GUI_GET_INFO_V,
+**    		GUI_GET_INFO_P,
+**    		GUI_GET_INFO_S,
+**    		GUI_CLEAR_EVENT,
+**    		GUI_MUTE_BEEPER=0x30,
+**    		GUI_BEEPER_SETTING,
+**    		GUI_SET_PASSWORD,
+**    		GUI_HOST_INTERFACE_MODE,
+**    		GUI_REBUILD_PRIORITY,
+**    		GUI_MAX_ATA_MODE,
+**    		GUI_RESET_CONTROLLER,
+**    		GUI_COM_PORT_SETTING,
+**    		GUI_NO_OPERATION,
+**    		GUI_DHCP_IP,
+**    		GUI_CREATE_PASS_THROUGH=0x40,
+**    		GUI_MODIFY_PASS_THROUGH,
+**    		GUI_DELETE_PASS_THROUGH,
+**    		GUI_IDENTIFY_DEVICE,
+**    		GUI_CREATE_RAIDSET=0x50,
+**    		GUI_DELETE_RAIDSET,
+**    		GUI_EXPAND_RAIDSET,
+**    		GUI_ACTIVATE_RAIDSET,
+**    		GUI_CREATE_HOT_SPARE,
+**    		GUI_DELETE_HOT_SPARE,
+**    		GUI_CREATE_VOLUME=0x60,
+**    		GUI_MODIFY_VOLUME,
+**    		GUI_DELETE_VOLUME,
+**    		GUI_START_CHECK_VOLUME,
+**    		GUI_STOP_CHECK_VOLUME
+**    	};
+**    Command description :
+**    	GUI_SET_SERIAL : Set the controller serial#
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x10
+**    		byte 3          : password length (should be 0x0f)
+**    		byte 4-0x13     : should be "ArEcATecHnoLogY"
+**    		byte 0x14--0x23 : Serial number string (must be 16 bytes)
+**      GUI_SET_VENDOR : Set vendor string for the controller
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x11
+**    		byte 3          : password length (should be 0x08)
+**    		byte 4-0x13     : should be "ArEcAvAr"
+**    		byte 0x14--0x3B : vendor string (must be 40 bytes)
+**      GUI_SET_MODEL : Set the model name of the controller
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x12
+**    		byte 3          : password length (should be 0x08)
+**    		byte 4-0x13     : should be "ArEcAvAr"
+**    		byte 0x14--0x1B : model string (must be 8 bytes)
+**      GUI_IDENTIFY : Identify device
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x13
+**    		                  return "Areca RAID Subsystem "
+**      GUI_CHECK_PASSWORD : Verify password
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x14
+**    		byte 3          : password length
+**    		byte 4-0x??     : user password to be checked
+**      GUI_LOGOUT : Logout GUI (force password checking on next command)
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x15
+**      GUI_HTTP : HTTP interface (reserved for Http proxy service)(0x16)
+**
+**      GUI_SET_ETHERNET_ADDR : Set the ethernet MAC address
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x17
+**    		byte 3          : password length (should be 0x08)
+**    		byte 4-0x13     : should be "ArEcAvAr"
+**    		byte 0x14--0x19 : Ethernet MAC address (must be 6 bytes)
+**      GUI_SET_LOGO : Set logo in HTTP
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x18
+**    		byte 3          : Page# (0/1/2/3) (0xff --> clear OEM logo)
+**    		byte 4/5/6/7    : 0x55/0xaa/0xa5/0x5a
+**    		byte 8          : TITLE.JPG data (each page must be 2000 bytes)
+**    		                  note page0 1st 2 byte must be
+**    					actual length of the JPG file
+**      GUI_POLL_EVENT : Poll If Event Log Changed
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x19
+**      GUI_GET_EVENT : Read Event
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x1a
+**    		byte 3          : Event Page (0:1st page/1/2/3:last page)
+**      GUI_GET_HW_MONITOR : Get HW monitor data
+**    		byte 0,1        : length
+**    		byte 2 			: command code 0x1b
+**    		byte 3 			: # of FANs(example 2)
+**    		byte 4 			: # of Voltage sensor(example 3)
+**    		byte 5 			: # of temperature sensor(example 2)
+**    		byte 6 			: # of power
+**    		byte 7/8        : Fan#0 (RPM)
+**    		byte 9/10       : Fan#1
+**    		byte 11/12 		: Voltage#0 original value in *1000
+**    		byte 13/14 		: Voltage#0 value
+**    		byte 15/16 		: Voltage#1 org
+**    		byte 17/18 		: Voltage#1
+**    		byte 19/20 		: Voltage#2 org
+**    		byte 21/22 		: Voltage#2
+**    		byte 23 		: Temp#0
+**    		byte 24 		: Temp#1
+**    		byte 25 		: Power indicator (bit0 : power#0,
+**    						 bit1 : power#1)
+**    		byte 26 		: UPS indicator
+**      GUI_QUICK_CREATE : Quick create raid/volume set
+**    	    byte 0,1        : length
+**    	    byte 2          : command code 0x20
+**    	    byte 3/4/5/6    : raw capacity
+**    	    byte 7 			: raid level
+**    	    byte 8 			: stripe size
+**    	    byte 9 			: spare
+**    	    byte 10/11/12/13: device mask (the devices to create raid/volume)
+**    		This function is removed, application like
+**    		to implement quick create function
+**    	need to use GUI_CREATE_RAIDSET and GUI_CREATE_VOLUMESET function.
+**      GUI_GET_INFO_R : Get Raid Set Information
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x20
+**    		byte 3          : raidset#
+**    	typedef struct sGUI_RAIDSET
+**    	{
+**    		BYTE grsRaidSetName[16];
+**    		DWORD grsCapacity;
+**    		DWORD grsCapacityX;
+**    		DWORD grsFailMask;
+**    		BYTE grsDevArray[32];
+**    		BYTE grsMemberDevices;
+**    		BYTE grsNewMemberDevices;
+**    		BYTE grsRaidState;
+**    		BYTE grsVolumes;
+**    		BYTE grsVolumeList[16];
+**    		BYTE grsRes1;
+**    		BYTE grsRes2;
+**    		BYTE grsRes3;
+**    		BYTE grsFreeSegments;
+**    		DWORD grsRawStripes[8];
+**    		DWORD grsRes4;
+**    		DWORD grsRes5; //     Total to 128 bytes
+**    		DWORD grsRes6; //     Total to 128 bytes
+**    	} sGUI_RAIDSET, *pGUI_RAIDSET;
+**      GUI_GET_INFO_V : Get Volume Set Information
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x21
+**    		byte 3          : volumeset#
+**    	typedef struct sGUI_VOLUMESET
+**    	{
+**    		BYTE gvsVolumeName[16]; //     16
+**    		DWORD gvsCapacity;
+**    		DWORD gvsCapacityX;
+**    		DWORD gvsFailMask;
+**    		DWORD gvsStripeSize;
+**    		DWORD gvsNewFailMask;
+**    		DWORD gvsNewStripeSize;
+**    		DWORD gvsVolumeStatus;
+**    		DWORD gvsProgress; //     32
+**    		sSCSI_ATTR gvsScsi;
+**    		BYTE gvsMemberDisks;
+**    		BYTE gvsRaidLevel; //     8
+**    		BYTE gvsNewMemberDisks;
+**    		BYTE gvsNewRaidLevel;
+**    		BYTE gvsRaidSetNumber;
+**    		BYTE gvsRes0; //     4
+**    		BYTE gvsRes1[4]; //     64 bytes
+**    	} sGUI_VOLUMESET, *pGUI_VOLUMESET;
+**      GUI_GET_INFO_P : Get Physical Drive Information
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x22
+**    		byte 3          : drive # (from 0 to max-channels - 1)
+**    	typedef struct sGUI_PHY_DRV
+**    	{
+**    		BYTE gpdModelName[40];
+**    		BYTE gpdSerialNumber[20];
+**    		BYTE gpdFirmRev[8];
+**    		DWORD gpdCapacity;
+**    		DWORD gpdCapacityX; //     Reserved for expansion
+**    		BYTE gpdDeviceState;
+**    		BYTE gpdPioMode;
+**    		BYTE gpdCurrentUdmaMode;
+**    		BYTE gpdUdmaMode;
+**    		BYTE gpdDriveSelect;
+**    		BYTE gpdRaidNumber; //     0xff if not belongs to a raid set
+**    		sSCSI_ATTR gpdScsi;
+**    		BYTE gpdReserved[40]; //     Total to 128 bytes
+**    	} sGUI_PHY_DRV, *pGUI_PHY_DRV;
+**    	GUI_GET_INFO_S : Get System Information
+**      	byte 0,1        : length
+**      	byte 2          : command code 0x23
+**    	typedef struct sCOM_ATTR
+**    	{
+**    		BYTE comBaudRate;
+**    		BYTE comDataBits;
+**    		BYTE comStopBits;
+**    		BYTE comParity;
+**    		BYTE comFlowControl;
+**    	} sCOM_ATTR, *pCOM_ATTR;
+**    	typedef struct sSYSTEM_INFO
+**    	{
+**    		BYTE gsiVendorName[40];
+**    		BYTE gsiSerialNumber[16];
+**    		BYTE gsiFirmVersion[16];
+**    		BYTE gsiBootVersion[16];
+**    		BYTE gsiMbVersion[16];
+**    		BYTE gsiModelName[8];
+**    		BYTE gsiLocalIp[4];
+**    		BYTE gsiCurrentIp[4];
+**    		DWORD gsiTimeTick;
+**    		DWORD gsiCpuSpeed;
+**    		DWORD gsiICache;
+**    		DWORD gsiDCache;
+**    		DWORD gsiScache;
+**    		DWORD gsiMemorySize;
+**    		DWORD gsiMemorySpeed;
+**    		DWORD gsiEvents;
+**    		BYTE gsiMacAddress[6];
+**    		BYTE gsiDhcp;
+**    		BYTE gsiBeeper;
+**    		BYTE gsiChannelUsage;
+**    		BYTE gsiMaxAtaMode;
+**    		BYTE gsiSdramEcc; //     1:if ECC enabled
+**    		BYTE gsiRebuildPriority;
+**    		sCOM_ATTR gsiComA; //     5 bytes
+**    		sCOM_ATTR gsiComB; //     5 bytes
+**    		BYTE gsiIdeChannels;
+**    		BYTE gsiScsiHostChannels;
+**    		BYTE gsiIdeHostChannels;
+**    		BYTE gsiMaxVolumeSet;
+**    		BYTE gsiMaxRaidSet;
+**    		BYTE gsiEtherPort; //     1:if ether net port supported
+**    		BYTE gsiRaid6Engine; //     1:Raid6 engine supported
+**    		BYTE gsiRes[75];
+**    	} sSYSTEM_INFO, *pSYSTEM_INFO;
+**    	GUI_CLEAR_EVENT : Clear System Event
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x24
+**      GUI_MUTE_BEEPER : Mute current beeper
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x30
+**      GUI_BEEPER_SETTING : Disable beeper
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x31
+**    		byte 3          : 0->disable, 1->enable
+**      GUI_SET_PASSWORD : Change password
+**    		byte 0,1        : length
+**    		byte 2 			: command code 0x32
+**    		byte 3 			: pass word length ( must <= 15 )
+**    		byte 4 			: password (must be alpha-numerical)
+**    	GUI_HOST_INTERFACE_MODE : Set host interface mode
+**    		byte 0,1        : length
+**    		byte 2 			: command code 0x33
+**    		byte 3 			: 0->Independent, 1->cluster
+**      GUI_REBUILD_PRIORITY : Set rebuild priority
+**    		byte 0,1        : length
+**    		byte 2 			: command code 0x34
+**    		byte 3 			: 0/1/2/3 (low->high)
+**      GUI_MAX_ATA_MODE : Set maximum ATA mode to be used
+**    		byte 0,1        : length
+**    		byte 2 			: command code 0x35
+**    		byte 3 			: 0/1/2/3 (133/100/66/33)
+**      GUI_RESET_CONTROLLER : Reset Controller
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x36
+**                            *Response with VT100 screen (discard it)
+**      GUI_COM_PORT_SETTING : COM port setting
+**    		byte 0,1        : length
+**    		byte 2 			: command code 0x37
+**    		byte 3 			: 0->COMA (term port),
+**    					  1->COMB (debug port)
+**    		byte 4 			: 0/1/2/3/4/5/6/7
+**    			(1200/2400/4800/9600/19200/38400/57600/115200)
+**    		byte 5 			: data bit
+**    					(0:7 bit, 1:8 bit : must be 8 bit)
+**    		byte 6 			: stop bit (0:1, 1:2 stop bits)
+**    		byte 7 			: parity (0:none, 1:off, 2:even)
+**    		byte 8 			: flow control
+**    			(0:none, 1:xon/xoff, 2:hardware => must use none)
+**      GUI_NO_OPERATION : No operation
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x38
+**      GUI_DHCP_IP : Set DHCP option and local IP address
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x39
+**    		byte 3          : 0:dhcp disabled, 1:dhcp enabled
+**    		byte 4/5/6/7    : IP address
+**      GUI_CREATE_PASS_THROUGH : Create pass through disk
+**    		byte 0,1        : length
+**    		byte 2 			: command code 0x40
+**    		byte 3 			: device #
+**    		byte 4 			: scsi channel (0/1)
+**    		byte 5 			: scsi id (0-->15)
+**    		byte 6 			: scsi lun (0-->7)
+**    		byte 7 			: tagged queue (1 : enabled)
+**    		byte 8 			: cache mode (1 : enabled)
+**    		byte 9 			: max speed (0/1/2/3/4,
+**    			async/20/40/80/160 for scsi)
+**    			(0/1/2/3/4, 33/66/100/133/150 for ide  )
+**      GUI_MODIFY_PASS_THROUGH : Modify pass through disk
+**    		byte 0,1        : length
+**    		byte 2 			: command code 0x41
+**    		byte 3 			: device #
+**    		byte 4 			: scsi channel (0/1)
+**    		byte 5 			: scsi id (0-->15)
+**    		byte 6 			: scsi lun (0-->7)
+**    		byte 7 			: tagged queue (1 : enabled)
+**    		byte 8 			: cache mode (1 : enabled)
+**    		byte 9 			: max speed (0/1/2/3/4,
+**    					async/20/40/80/160 for scsi)
+**    			(0/1/2/3/4, 33/66/100/133/150 for ide  )
+**      GUI_DELETE_PASS_THROUGH : Delete pass through disk
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x42
+**    		byte 3          : device# to be deleted
+**      GUI_IDENTIFY_DEVICE : Identify Device
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x43
+**    		byte 3          : Flash Method
+**    				(0:flash selected, 1:flash not selected)
+**    		byte 4/5/6/7    : IDE device mask to be flashed
+**                           note .... no response data available
+**    	GUI_CREATE_RAIDSET : Create Raid Set
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x50
+**    		byte 3/4/5/6    : device mask
+**    		byte 7-22       : raidset name (if byte 7 == 0:use default)
+**      GUI_DELETE_RAIDSET : Delete Raid Set
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x51
+**    		byte 3          : raidset#
+**    	GUI_EXPAND_RAIDSET : Expand Raid Set
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x52
+**    		byte 3          : raidset#
+**    		byte 4/5/6/7    : device mask for expansion
+**    		byte 8/9/10     : (8:0 no change, 1 change, 0xff:terminate,
+**    				9:new raid level,
+**    				10:new stripe size
+**    				0/1/2/3/4/5->4/8/16/32/64/128K )
+**    		byte 11/12/13   : repeat for each volume in the raidset
+**      GUI_ACTIVATE_RAIDSET : Activate incomplete raid set
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x53
+**    		byte 3          : raidset#
+**      GUI_CREATE_HOT_SPARE : Create hot spare disk
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x54
+**    		byte 3/4/5/6    : device mask for hot spare creation
+**    	GUI_DELETE_HOT_SPARE : Delete hot spare disk
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x55
+**    		byte 3/4/5/6    : device mask for hot spare deletion
+**    	GUI_CREATE_VOLUME : Create volume set
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x60
+**    		byte 3          : raidset#
+**    		byte 4-19       : volume set name
+**    				(if byte4 == 0, use default)
+**    		byte 20-27      : volume capacity (blocks)
+**    		byte 28 		: raid level
+**    		byte 29 		: stripe size
+**    				(0/1/2/3/4/5->4/8/16/32/64/128K)
+**    		byte 30 		: channel
+**    		byte 31 		: ID
+**    		byte 32 		: LUN
+**    		byte 33 		: 1 enable tag
+**    		byte 34 		: 1 enable cache
+**    		byte 35 		: speed
+**    		(0/1/2/3/4->async/20/40/80/160 for scsi)
+**    		(0/1/2/3/4->33/66/100/133/150 for IDE  )
+**    		byte 36 		: 1 to select quick init
+**
+**    	GUI_MODIFY_VOLUME : Modify volume Set
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x61
+**    		byte 3          : volumeset#
+**    		byte 4-19       : new volume set name
+**    		(if byte4 == 0, not change)
+**    		byte 20-27      : new volume capacity (reserved)
+**    		byte 28 		: new raid level
+**    		byte 29 		: new stripe size
+**    		(0/1/2/3/4/5->4/8/16/32/64/128K)
+**    		byte 30 		: new channel
+**    		byte 31 		: new ID
+**    		byte 32 		: new LUN
+**    		byte 33 		: 1 enable tag
+**    		byte 34 		: 1 enable cache
+**    		byte 35 		: speed
+**    		(0/1/2/3/4->async/20/40/80/160 for scsi)
+**    		(0/1/2/3/4->33/66/100/133/150 for IDE  )
+**    	GUI_DELETE_VOLUME : Delete volume set
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x62
+**    		byte 3          : volumeset#
+**    	GUI_START_CHECK_VOLUME : Start volume consistency check
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x63
+**    		byte 3          : volumeset#
+**    	GUI_STOP_CHECK_VOLUME : Stop volume consistency check
+**    		byte 0,1        : length
+**    		byte 2          : command code 0x64
+** ---------------------------------------------------------------------
+**    4. Returned data
+** ---------------------------------------------------------------------
+**    	(A) Header          : 3 bytes sequence (0x5E, 0x01, 0x61)
+**    	(B) Length          : 2 bytes
+**    			(low byte 1st, excludes length and checksum byte)
+**    	(C) status or data  :
+**           <1> If length == 1 ==> 1 byte status code
+**    		#define GUI_OK                    0x41
+**    		#define GUI_RAIDSET_NOT_NORMAL    0x42
+**    		#define GUI_VOLUMESET_NOT_NORMAL  0x43
+**    		#define GUI_NO_RAIDSET            0x44
+**    		#define GUI_NO_VOLUMESET          0x45
+**    		#define GUI_NO_PHYSICAL_DRIVE     0x46
+**    		#define GUI_PARAMETER_ERROR       0x47
+**    		#define GUI_UNSUPPORTED_COMMAND   0x48
+**    		#define GUI_DISK_CONFIG_CHANGED   0x49
+**    		#define GUI_INVALID_PASSWORD      0x4a
+**    		#define GUI_NO_DISK_SPACE         0x4b
+**    		#define GUI_CHECKSUM_ERROR        0x4c
+**    		#define GUI_PASSWORD_REQUIRED     0x4d
+**           <2> If length > 1 ==>
+**    		data block returned from controller
+**    		and the contents depends on the command code
+**    	(E) Checksum        : checksum of length and status or data byte
+**************************************************************************
diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig
index 96a81cd17617..d61662c1a0ee 100644
--- a/drivers/scsi/Kconfig
+++ b/drivers/scsi/Kconfig
@@ -469,6 +469,20 @@ config SCSI_IN2000
 	  To compile this driver as a module, choose M here: the
 	  module will be called in2000.
 
+config SCSI_ARCMSR
+	tristate "ARECA ARC11X0[PCI-X]/ARC12X0[PCI-EXPRESS] SATA-RAID support"
+	depends on PCI && SCSI
+	help
+	  This driver supports all of ARECA's SATA RAID controller cards.
+	  This is an ARECA-maintained driver by Erich Chen.
+	  If you have any problems, please mail to: < erich@areca.com.tw >
+	  Areca supports Linux RAID config tools.
+
+	  < http://www.areca.com.tw >
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called arcmsr (modprobe arcmsr).
+
 source "drivers/scsi/megaraid/Kconfig.megaraid"
 
 config SCSI_SATA
diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile
index ebd0cf00bf3e..b2de9bfdfdcd 100644
--- a/drivers/scsi/Makefile
+++ b/drivers/scsi/Makefile
@@ -59,6 +59,7 @@ obj-$(CONFIG_SCSI_PSI240I)	+= psi240i.o
 obj-$(CONFIG_SCSI_BUSLOGIC)	+= BusLogic.o
 obj-$(CONFIG_SCSI_DPT_I2O)	+= dpt_i2o.o
 obj-$(CONFIG_SCSI_U14_34F)	+= u14-34f.o
+obj-$(CONFIG_SCSI_ARCMSR)	+= arcmsr/
 obj-$(CONFIG_SCSI_ULTRASTOR)	+= ultrastor.o
 obj-$(CONFIG_SCSI_AHA152X)	+= aha152x.o
 obj-$(CONFIG_SCSI_AHA1542)	+= aha1542.o
diff --git a/drivers/scsi/arcmsr/Makefile b/drivers/scsi/arcmsr/Makefile
new file mode 100644
index 000000000000..721aced39168
--- /dev/null
+++ b/drivers/scsi/arcmsr/Makefile
@@ -0,0 +1,6 @@
+# File: drivers/arcmsr/Makefile
+# Makefile for the ARECA PCI-X PCI-EXPRESS SATA RAID controllers SCSI driver.
+
+arcmsr-objs := arcmsr_attr.o arcmsr_hba.o
+
+obj-$(CONFIG_SCSI_ARCMSR) := arcmsr.o
diff --git a/drivers/scsi/arcmsr/arcmsr.h b/drivers/scsi/arcmsr/arcmsr.h
new file mode 100644
index 000000000000..aff96db9ccf6
--- /dev/null
+++ b/drivers/scsi/arcmsr/arcmsr.h
@@ -0,0 +1,472 @@
+/*
+*******************************************************************************
+**        O.S   : Linux
+**   FILE NAME  : arcmsr.h
+**        BY    : Erich Chen
+**   Description: SCSI RAID Device Driver for
+**                ARECA RAID Host adapter
+*******************************************************************************
+** Copyright (C) 2002 - 2005, Areca Technology Corporation All rights reserved.
+**
+**     Web site: www.areca.com.tw
+**       E-mail: erich@areca.com.tw
+**
+** This program is free software; you can redistribute it and/or modify
+** it under the terms of the GNU General Public License version 2 as
+** published by the Free Software Foundation.
+** This program is distributed in the hope that it will be useful,
+** but WITHOUT ANY WARRANTY; without even the implied warranty of
+** MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+** GNU General Public License for more details.
+*******************************************************************************
+** Redistribution and use in source and binary forms, with or without
+** modification, are permitted provided that the following conditions
+** are met:
+** 1. Redistributions of source code must retain the above copyright
+**    notice, this list of conditions and the following disclaimer.
+** 2. Redistributions in binary form must reproduce the above copyright
+**    notice, this list of conditions and the following disclaimer in the
+**    documentation and/or other materials provided with the distribution.
+** 3. The name of the author may not be used to endorse or promote products
+**    derived from this software without specific prior written permission.
+**
+** THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
+** IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
+** OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
+** IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
+** INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES(INCLUDING, BUT
+** NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+** DATA, OR PROFITS; OR BUSINESS INTERRUPTION)HOWEVER CAUSED AND ON ANY
+** THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+**(INCLUDING NEGLIGENCE OR OTHERWISE)ARISING IN ANY WAY OUT OF THE USE OF
+** THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*******************************************************************************
+*/
+#include <linux/interrupt.h>
+
+struct class_device_attribute;
+
+#define ARCMSR_MAX_OUTSTANDING_CMD 						256
+#define ARCMSR_MAX_FREECCB_NUM							288
+#define ARCMSR_DRIVER_VERSION				"Driver Version 1.20.00.13"
+#define ARCMSR_SCSI_INITIATOR_ID						255
+#define ARCMSR_MAX_XFER_SECTORS							512
+#define ARCMSR_MAX_TARGETID							 17
+#define ARCMSR_MAX_TARGETLUN							  8
+#define ARCMSR_MAX_CMD_PERLUN				 ARCMSR_MAX_OUTSTANDING_CMD
+#define ARCMSR_MAX_QBUFFER						       4096
+#define ARCMSR_MAX_SG_ENTRIES							 38
+
+/*
+*******************************************************************************
+**        split 64bits dma addressing
+*******************************************************************************
+*/
+#define dma_addr_hi32(addr)               (uint32_t) ((addr>>16)>>16)
+#define dma_addr_lo32(addr)               (uint32_t) (addr & 0xffffffff)
+/*
+*******************************************************************************
+**        MESSAGE CONTROL CODE
+*******************************************************************************
+*/
+struct CMD_MESSAGE
+{
+      uint32_t HeaderLength;
+      uint8_t  Signature[8];
+      uint32_t Timeout;
+      uint32_t ControlCode;
+      uint32_t ReturnCode;
+      uint32_t Length;
+};
+/*
+*******************************************************************************
+**        IOP Message Transfer Data for user space
+*******************************************************************************
+*/
+struct CMD_MESSAGE_FIELD
+{
+    struct CMD_MESSAGE			cmdmessage;
+    uint8_t				messagedatabuffer[1032];
+};
+/* IOP message transfer */
+#define ARCMSR_MESSAGE_FAIL             0x0001
+/* DeviceType */
+#define ARECA_SATA_RAID				0x90000000
+/* FunctionCode */
+#define FUNCTION_READ_RQBUFFER			0x0801
+#define FUNCTION_WRITE_WQBUFFER			0x0802
+#define FUNCTION_CLEAR_RQBUFFER			0x0803
+#define FUNCTION_CLEAR_WQBUFFER			0x0804
+#define FUNCTION_CLEAR_ALLQBUFFER		0x0805
+#define FUNCTION_RETURN_CODE_3F			0x0806
+#define FUNCTION_SAY_HELLO			0x0807
+#define FUNCTION_SAY_GOODBYE			0x0808
+#define FUNCTION_FLUSH_ADAPTER_CACHE		0x0809
+/* ARECA IO CONTROL CODE*/
+#define ARCMSR_MESSAGE_READ_RQBUFFER       \
+	ARECA_SATA_RAID | FUNCTION_READ_RQBUFFER
+#define ARCMSR_MESSAGE_WRITE_WQBUFFER      \
+	ARECA_SATA_RAID | FUNCTION_WRITE_WQBUFFER
+#define ARCMSR_MESSAGE_CLEAR_RQBUFFER      \
+	ARECA_SATA_RAID | FUNCTION_CLEAR_RQBUFFER
+#define ARCMSR_MESSAGE_CLEAR_WQBUFFER      \
+	ARECA_SATA_RAID | FUNCTION_CLEAR_WQBUFFER
+#define ARCMSR_MESSAGE_CLEAR_ALLQBUFFER    \
+	ARECA_SATA_RAID | FUNCTION_CLEAR_ALLQBUFFER
+#define ARCMSR_MESSAGE_RETURN_CODE_3F      \
+	ARECA_SATA_RAID | FUNCTION_RETURN_CODE_3F
+#define ARCMSR_MESSAGE_SAY_HELLO           \
+	ARECA_SATA_RAID | FUNCTION_SAY_HELLO
+#define ARCMSR_MESSAGE_SAY_GOODBYE         \
+	ARECA_SATA_RAID | FUNCTION_SAY_GOODBYE
+#define ARCMSR_MESSAGE_FLUSH_ADAPTER_CACHE \
+	ARECA_SATA_RAID | FUNCTION_FLUSH_ADAPTER_CACHE
+/* ARECA IOCTL ReturnCode */
+#define ARCMSR_MESSAGE_RETURNCODE_OK              0x00000001
+#define ARCMSR_MESSAGE_RETURNCODE_ERROR           0x00000006
+#define ARCMSR_MESSAGE_RETURNCODE_3F              0x0000003F
+/*
+*************************************************************
+**   structure for holding DMA address data
+*************************************************************
+*/
+#define IS_SG64_ADDR                0x01000000 /* bit24 */
+struct  SG32ENTRY
+{
+	uint32_t					length;
+	uint32_t					address;
+};
+struct  SG64ENTRY
+{
+ 	uint32_t					length;
+ 	uint32_t					address;
+ 	uint32_t					addresshigh;
+};
+struct SGENTRY_UNION
+{
+	union
+	{
+		struct SG32ENTRY            sg32entry;
+		struct SG64ENTRY            sg64entry;
+	}u;
+};
+/*
+********************************************************************
+**      Q Buffer of IOP Message Transfer
+********************************************************************
+*/
+struct QBUFFER
+{
+	uint32_t      data_len;
+	uint8_t       data[124];
+};
+/*
+*******************************************************************************
+**      FIRMWARE INFO
+*******************************************************************************
+*/
+struct FIRMWARE_INFO
+{
+	uint32_t      signature;                /*0, 00-03*/
+	uint32_t      request_len;              /*1, 04-07*/
+	uint32_t      numbers_queue;            /*2, 08-11*/
+	uint32_t      sdram_size;               /*3, 12-15*/
+	uint32_t      ide_channels;             /*4, 16-19*/
+	char          vendor[40];               /*5, 20-59*/
+	char          model[8];                 /*15, 60-67*/
+	char          firmware_ver[16];         /*17, 68-83*/
+	char          device_map[16];           /*21, 84-99*/
+};
+/* signature of set and get firmware config */
+#define ARCMSR_SIGNATURE_GET_CONFIG                   0x87974060
+#define ARCMSR_SIGNATURE_SET_CONFIG                   0x87974063
+/* message code of inbound message register */
+#define ARCMSR_INBOUND_MESG0_NOP                      0x00000000
+#define ARCMSR_INBOUND_MESG0_GET_CONFIG               0x00000001
+#define ARCMSR_INBOUND_MESG0_SET_CONFIG               0x00000002
+#define ARCMSR_INBOUND_MESG0_ABORT_CMD                0x00000003
+#define ARCMSR_INBOUND_MESG0_STOP_BGRB                0x00000004
+#define ARCMSR_INBOUND_MESG0_FLUSH_CACHE              0x00000005
+#define ARCMSR_INBOUND_MESG0_START_BGRB               0x00000006
+#define ARCMSR_INBOUND_MESG0_CHK331PENDING            0x00000007
+#define ARCMSR_INBOUND_MESG0_SYNC_TIMER               0x00000008
+/* doorbell interrupt generator */
+#define ARCMSR_INBOUND_DRIVER_DATA_WRITE_OK           0x00000001
+#define ARCMSR_INBOUND_DRIVER_DATA_READ_OK            0x00000002
+#define ARCMSR_OUTBOUND_IOP331_DATA_WRITE_OK          0x00000001
+#define ARCMSR_OUTBOUND_IOP331_DATA_READ_OK           0x00000002
+/* ccb areca cdb flag */
+#define ARCMSR_CCBPOST_FLAG_SGL_BSIZE                 0x80000000
+#define ARCMSR_CCBPOST_FLAG_IAM_BIOS                  0x40000000
+#define ARCMSR_CCBREPLY_FLAG_IAM_BIOS                 0x40000000
+#define ARCMSR_CCBREPLY_FLAG_ERROR                    0x10000000
+/* outbound firmware ok */
+#define ARCMSR_OUTBOUND_MESG1_FIRMWARE_OK             0x80000000
+/*
+*******************************************************************************
+**    ARECA SCSI COMMAND DESCRIPTOR BLOCK size 0x1F8 (504)
+*******************************************************************************
+*/
+struct ARCMSR_CDB
+{
+	uint8_t							Bus;
+	uint8_t							TargetID;
+	uint8_t							LUN;
+	uint8_t							Function;
+
+	uint8_t							CdbLength;
+	uint8_t							sgcount;
+	uint8_t							Flags;
+#define ARCMSR_CDB_FLAG_SGL_BSIZE          0x01
+#define ARCMSR_CDB_FLAG_BIOS               0x02
+#define ARCMSR_CDB_FLAG_WRITE              0x04
+#define ARCMSR_CDB_FLAG_SIMPLEQ            0x00
+#define ARCMSR_CDB_FLAG_HEADQ              0x08
+#define ARCMSR_CDB_FLAG_ORDEREDQ           0x10
+	uint8_t							Reserved1;
+
+	uint32_t						Context;
+	uint32_t						DataLength;
+
+	uint8_t							Cdb[16];
+
+	uint8_t							DeviceStatus;
+#define ARCMSR_DEV_CHECK_CONDITION          0x02
+#define ARCMSR_DEV_SELECT_TIMEOUT			0xF0
+#define ARCMSR_DEV_ABORTED				0xF1
+#define ARCMSR_DEV_INIT_FAIL				0xF2
+	uint8_t							SenseData[15];
+
+	union
+	{
+		struct SG32ENTRY                sg32entry[ARCMSR_MAX_SG_ENTRIES];
+		struct SG64ENTRY                sg64entry[ARCMSR_MAX_SG_ENTRIES];
+	} u;
+};
+/*
+*******************************************************************************
+**     Messaging Unit (MU) of the Intel R 80331 I/O processor (80331)
+*******************************************************************************
+*/
+struct MessageUnit
+{
+	uint32_t	resrved0[4];			/*0000 000F*/
+	uint32_t	inbound_msgaddr0;		/*0010 0013*/
+	uint32_t	inbound_msgaddr1;		/*0014 0017*/
+	uint32_t	outbound_msgaddr0;		/*0018 001B*/
+	uint32_t	outbound_msgaddr1;		/*001C 001F*/
+	uint32_t	inbound_doorbell;		/*0020 0023*/
+	uint32_t	inbound_intstatus;		/*0024 0027*/
+	uint32_t	inbound_intmask;		/*0028 002B*/
+	uint32_t	outbound_doorbell;		/*002C 002F*/
+	uint32_t	outbound_intstatus;		/*0030 0033*/
+	uint32_t	outbound_intmask;		/*0034 0037*/
+	uint32_t	reserved1[2];			/*0038 003F*/
+	uint32_t	inbound_queueport;		/*0040 0043*/
+	uint32_t	outbound_queueport;     	/*0044 0047*/
+	uint32_t	reserved2[2];			/*0048 004F*/
+	uint32_t	reserved3[492];			/*0050 07FF 492*/
+	uint32_t	reserved4[128];			/*0800 09FF 128*/
+	uint32_t	message_rwbuffer[256];		/*0a00 0DFF 256*/
+	uint32_t	message_wbuffer[32];		/*0E00 0E7F  32*/
+	uint32_t	reserved5[32];			/*0E80 0EFF  32*/
+	uint32_t	message_rbuffer[32];		/*0F00 0F7F  32*/
+	uint32_t	reserved6[32];			/*0F80 0FFF  32*/
+};
+/*
+*******************************************************************************
+**                 Adapter Control Block
+*******************************************************************************
+*/
+struct AdapterControlBlock
+{
+	struct pci_dev *		pdev;
+	struct Scsi_Host *		host;
+	unsigned long			vir2phy_offset;
+	/* Offset is used in making arc cdb physical to virtual calculations */
+	uint32_t			outbound_int_enable;
+
+	struct MessageUnit __iomem *		pmu;
+	/* message unit ATU inbound base address0 */
+
+	uint32_t			acb_flags;
+#define ACB_F_SCSISTOPADAPTER         0x0001
+#define ACB_F_MSG_STOP_BGRB           0x0002
+	/* stop RAID background rebuild */
+#define ACB_F_MSG_START_BGRB          0x0004
+	/* stop RAID background rebuild */
+#define ACB_F_IOPDATA_OVERFLOW        0x0008
+	/* iop message data rqbuffer overflow */
+#define ACB_F_MESSAGE_WQBUFFER_CLEARED  0x0010
+	/* message clear wqbuffer */
+#define ACB_F_MESSAGE_RQBUFFER_CLEARED  0x0020
+	/* message clear rqbuffer */
+#define ACB_F_MESSAGE_WQBUFFER_READED   0x0040
+#define ACB_F_BUS_RESET               0x0080
+#define ACB_F_IOP_INITED              0x0100
+	/* iop init */
+
+	struct CommandControlBlock *			pccb_pool[ARCMSR_MAX_FREECCB_NUM];
+	/* used for memory free */
+	struct list_head		ccb_free_list;
+	/* head of free ccb list */
+	atomic_t			ccboutstandingcount;
+
+	void *				dma_coherent;
+	/* dma_coherent used for memory free */
+	dma_addr_t			dma_coherent_handle;
+	/* dma_coherent_handle used for memory free */
+
+	uint8_t				rqbuffer[ARCMSR_MAX_QBUFFER];
+	/* data collection buffer for read from 80331 */
+	int32_t				rqbuf_firstindex;
+	/* first of read buffer  */
+	int32_t				rqbuf_lastindex;
+	/* last of read buffer   */
+	uint8_t				wqbuffer[ARCMSR_MAX_QBUFFER];
+	/* data collection buffer for write to 80331  */
+	int32_t				wqbuf_firstindex;
+	/* first of write buffer */
+	int32_t				wqbuf_lastindex;
+	/* last of write buffer  */
+	uint8_t				devstate[ARCMSR_MAX_TARGETID][ARCMSR_MAX_TARGETLUN];
+	/* id0 ..... id15, lun0...lun7 */
+#define ARECA_RAID_GONE               0x55
+#define ARECA_RAID_GOOD               0xaa
+	uint32_t			num_resets;
+	uint32_t			num_aborts;
+	uint32_t			firm_request_len;
+	uint32_t			firm_numbers_queue;
+	uint32_t			firm_sdram_size;
+	uint32_t			firm_hd_channels;
+	char				firm_model[12];
+	char				firm_version[20];
+};/* HW_DEVICE_EXTENSION */
+/*
+*******************************************************************************
+**                   Command Control Block
+**             this CCB length must be 32 bytes boundary
+*******************************************************************************
+*/
+struct CommandControlBlock
+{
+	struct ARCMSR_CDB		arcmsr_cdb;
+	/*
+	** 0-503 (size of CDB=504):
+	** arcmsr messenger scsi command descriptor size 504 bytes
+	*/
+	uint32_t			cdb_shifted_phyaddr;
+	/* 504-507 */
+	uint32_t			reserved1;
+	/* 508-511 */
+#if BITS_PER_LONG == 64
+	/*  ======================512+64 bytes========================  */
+	struct list_head		list;
+	/* 512-527 16 bytes next/prev ptrs for ccb lists */
+	struct scsi_cmnd *		pcmd;
+	/* 528-535 8 bytes pointer of linux scsi command */
+	struct AdapterControlBlock *	acb;
+	/* 536-543 8 bytes pointer of acb */
+
+	uint16_t			ccb_flags;
+	/* 544-545 */
+	#define		CCB_FLAG_READ			0x0000
+	#define		CCB_FLAG_WRITE			0x0001
+	#define		CCB_FLAG_ERROR			0x0002
+	#define		CCB_FLAG_FLUSHCACHE		0x0004
+	#define		CCB_FLAG_MASTER_ABORTED		0x0008
+	uint16_t			startdone;
+	/* 546-547 */
+	#define		ARCMSR_CCB_DONE			0x0000
+	#define		ARCMSR_CCB_START		0x55AA
+	#define		ARCMSR_CCB_ABORTED		0xAA55
+	#define		ARCMSR_CCB_ILLEGAL		0xFFFF
+	uint32_t			reserved2[7];
+	/* 548-551 552-555 556-559 560-563 564-567 568-571 572-575 */
+#else
+	/*  ======================512+32 bytes========================  */
+	struct list_head		list;
+	/* 512-519 8 bytes next/prev ptrs for ccb lists */
+	struct scsi_cmnd *		pcmd;
+	/* 520-523 4 bytes pointer of linux scsi command */
+	struct AdapterControlBlock *	acb;
+	/* 524-527 4 bytes pointer of acb */
+
+	uint16_t			ccb_flags;
+	/* 528-529 */
+	#define		CCB_FLAG_READ			0x0000
+	#define		CCB_FLAG_WRITE			0x0001
+	#define		CCB_FLAG_ERROR			0x0002
+	#define		CCB_FLAG_FLUSHCACHE		0x0004
+	#define		CCB_FLAG_MASTER_ABORTED		0x0008
+	uint16_t			startdone;
+	/* 530-531 */
+	#define		ARCMSR_CCB_DONE			0x0000
+	#define		ARCMSR_CCB_START		0x55AA
+	#define		ARCMSR_CCB_ABORTED		0xAA55
+	#define		ARCMSR_CCB_ILLEGAL		0xFFFF
+	uint32_t			reserved2[3];
+	/* 532-535 536-539 540-543 */
+#endif
+	/*  ==========================================================  */
+};
+/*
+*******************************************************************************
+**    ARECA SCSI sense data
+*******************************************************************************
+*/
+struct SENSE_DATA
+{
+	uint8_t				ErrorCode:7;
+#define SCSI_SENSE_CURRENT_ERRORS	0x70
+#define SCSI_SENSE_DEFERRED_ERRORS	0x71
+	uint8_t				Valid:1;
+	uint8_t				SegmentNumber;
+	uint8_t				SenseKey:4;
+	uint8_t				Reserved:1;
+	uint8_t				IncorrectLength:1;
+	uint8_t				EndOfMedia:1;
+	uint8_t				FileMark:1;
+	uint8_t				Information[4];
+	uint8_t				AdditionalSenseLength;
+	uint8_t				CommandSpecificInformation[4];
+	uint8_t				AdditionalSenseCode;
+	uint8_t				AdditionalSenseCodeQualifier;
+	uint8_t				FieldReplaceableUnitCode;
+	uint8_t				SenseKeySpecific[3];
+};
+/*
+*******************************************************************************
+**  Outbound Interrupt Status Register - OISR
+*******************************************************************************
+*/
+#define     ARCMSR_MU_OUTBOUND_INTERRUPT_STATUS_REG                 0x30
+#define     ARCMSR_MU_OUTBOUND_PCI_INT                              0x10
+#define     ARCMSR_MU_OUTBOUND_POSTQUEUE_INT                        0x08
+#define     ARCMSR_MU_OUTBOUND_DOORBELL_INT                         0x04
+#define     ARCMSR_MU_OUTBOUND_MESSAGE1_INT                         0x02
+#define     ARCMSR_MU_OUTBOUND_MESSAGE0_INT                         0x01
+#define     ARCMSR_MU_OUTBOUND_HANDLE_INT                 \
+                    (ARCMSR_MU_OUTBOUND_MESSAGE0_INT      \
+                     |ARCMSR_MU_OUTBOUND_MESSAGE1_INT     \
+                     |ARCMSR_MU_OUTBOUND_DOORBELL_INT     \
+                     |ARCMSR_MU_OUTBOUND_POSTQUEUE_INT    \
+                     |ARCMSR_MU_OUTBOUND_PCI_INT)
+/*
+*******************************************************************************
+**  Outbound Interrupt Mask Register - OIMR
+*******************************************************************************
+*/
+#define     ARCMSR_MU_OUTBOUND_INTERRUPT_MASK_REG                   0x34
+#define     ARCMSR_MU_OUTBOUND_PCI_INTMASKENABLE                    0x10
+#define     ARCMSR_MU_OUTBOUND_POSTQUEUE_INTMASKENABLE              0x08
+#define     ARCMSR_MU_OUTBOUND_DOORBELL_INTMASKENABLE               0x04
+#define     ARCMSR_MU_OUTBOUND_MESSAGE1_INTMASKENABLE               0x02
+#define     ARCMSR_MU_OUTBOUND_MESSAGE0_INTMASKENABLE               0x01
+#define     ARCMSR_MU_OUTBOUND_ALL_INTMASKENABLE                    0x1F
+
+extern void arcmsr_post_Qbuffer(struct AdapterControlBlock *acb);
+extern struct class_device_attribute *arcmsr_host_attrs[];
+extern int arcmsr_alloc_sysfs_attr(struct AdapterControlBlock *acb);
+void arcmsr_free_sysfs_attr(struct AdapterControlBlock *acb);
+
diff --git a/drivers/scsi/arcmsr/arcmsr_attr.c b/drivers/scsi/arcmsr/arcmsr_attr.c
new file mode 100644
index 000000000000..0459f4194d7c
--- /dev/null
+++ b/drivers/scsi/arcmsr/arcmsr_attr.c
@@ -0,0 +1,392 @@
+/*
+*******************************************************************************
+**        O.S   : Linux
+**   FILE NAME  : arcmsr_attr.c
+**        BY    : Erich Chen
+**   Description: attributes exported to sysfs and device host
+*******************************************************************************
+** Copyright (C) 2002 - 2005, Areca Technology Corporation All rights reserved
+**
+**     Web site: www.areca.com.tw
+**       E-mail: erich@areca.com.tw
+**
+** This program is free software; you can redistribute it and/or modify
+** it under the terms of the GNU General Public License version 2 as
+** published by the Free Software Foundation.
+** This program is distributed in the hope that it will be useful,
+** but WITHOUT ANY WARRANTY; without even the implied warranty of
+** MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+** GNU General Public License for more details.
+*******************************************************************************
+** Redistribution and use in source and binary forms, with or without
+** modification, are permitted provided that the following conditions
+** are met:
+** 1. Redistributions of source code must retain the above copyright
+**    notice, this list of conditions and the following disclaimer.
+** 2. Redistributions in binary form must reproduce the above copyright
+**    notice, this list of conditions and the following disclaimer in the
+**    documentation and/or other materials provided with the distribution.
+** 3. The name of the author may not be used to endorse or promote products
+**    derived from this software without specific prior written permission.
+**
+** THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
+** IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
+** OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
+** IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
+** INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES(INCLUDING,BUT
+** NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+** DATA, OR PROFITS; OR BUSINESS INTERRUPTION)HOWEVER CAUSED AND ON ANY
+** THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+** (INCLUDING NEGLIGENCE OR OTHERWISE)ARISING IN ANY WAY OUT OF THE USE OF
+** THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*******************************************************************************
+** For history of changes, see Documentation/scsi/ChangeLog.arcmsr
+**     Firmware Specification, see Documentation/scsi/arcmsr_spec.txt
+*******************************************************************************
+*/
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/errno.h>
+#include <linux/delay.h>
+#include <linux/pci.h>
+
+#include <scsi/scsi_cmnd.h>
+#include <scsi/scsi_device.h>
+#include <scsi/scsi_host.h>
+#include <scsi/scsi_transport.h>
+#include "arcmsr.h"
+
+struct class_device_attribute *arcmsr_host_attrs[];
+
+static ssize_t
+arcmsr_sysfs_iop_message_read(struct kobject *kobj, char *buf, loff_t off,
+    size_t count)
+{
+	struct class_device *cdev = container_of(kobj,struct class_device,kobj);
+	struct Scsi_Host *host = class_to_shost(cdev);
+	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+	struct MessageUnit __iomem *reg = acb->pmu;
+	uint8_t *pQbuffer,*ptmpQbuffer;
+	int32_t allxfer_len = 0;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EACCES;
+
+	/* do message unit read. */
+	ptmpQbuffer = (uint8_t *)buf;
+	while ((acb->rqbuf_firstindex != acb->rqbuf_lastindex)
+		&& (allxfer_len < 1031)) {
+		pQbuffer = &acb->rqbuffer[acb->rqbuf_firstindex];
+		memcpy(ptmpQbuffer, pQbuffer, 1);
+		acb->rqbuf_firstindex++;
+		acb->rqbuf_firstindex %= ARCMSR_MAX_QBUFFER;
+		ptmpQbuffer++;
+		allxfer_len++;
+	}
+	if (acb->acb_flags & ACB_F_IOPDATA_OVERFLOW) {
+		struct QBUFFER __iomem * prbuffer = (struct QBUFFER __iomem *)
+					&reg->message_rbuffer;
+		uint8_t __iomem * iop_data = (uint8_t __iomem *)prbuffer->data;
+		int32_t iop_len;
+
+		acb->acb_flags &= ~ACB_F_IOPDATA_OVERFLOW;
+		iop_len = readl(&prbuffer->data_len);
+		while (iop_len > 0) {
+			acb->rqbuffer[acb->rqbuf_lastindex] = readb(iop_data);
+			acb->rqbuf_lastindex++;
+			acb->rqbuf_lastindex %= ARCMSR_MAX_QBUFFER;
+			iop_data++;
+			iop_len--;
+		}
+		writel(ARCMSR_INBOUND_DRIVER_DATA_READ_OK,
+				&reg->inbound_doorbell);
+	}
+	return (allxfer_len);
+}
+
+static ssize_t
+arcmsr_sysfs_iop_message_write(struct kobject *kobj, char *buf, loff_t off,
+    size_t count)
+{
+	struct class_device *cdev = container_of(kobj,struct class_device,kobj);
+	struct Scsi_Host *host = class_to_shost(cdev);
+	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+	int32_t my_empty_len, user_len, wqbuf_firstindex, wqbuf_lastindex;
+	uint8_t *pQbuffer, *ptmpuserbuffer;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EACCES;
+	if (count > 1032)
+		return -EINVAL;
+	/* do message unit write. */
+	ptmpuserbuffer = (uint8_t *)buf;
+	user_len = (int32_t)count;
+	wqbuf_lastindex = acb->wqbuf_lastindex;
+	wqbuf_firstindex = acb->wqbuf_firstindex;
+	if (wqbuf_lastindex != wqbuf_firstindex) {
+		arcmsr_post_Qbuffer(acb);
+		return 0;	/*need retry*/
+	} else {
+		my_empty_len = (wqbuf_firstindex-wqbuf_lastindex - 1)
+				&(ARCMSR_MAX_QBUFFER - 1);
+		if (my_empty_len >= user_len) {
+			while (user_len > 0) {
+				pQbuffer =
+				&acb->wqbuffer[acb->wqbuf_lastindex];
+				memcpy(pQbuffer, ptmpuserbuffer, 1);
+				acb->wqbuf_lastindex++;
+				acb->wqbuf_lastindex %= ARCMSR_MAX_QBUFFER;
+				ptmpuserbuffer++;
+				user_len--;
+			}
+			if (acb->acb_flags & ACB_F_MESSAGE_WQBUFFER_CLEARED) {
+				acb->acb_flags &=
+					~ACB_F_MESSAGE_WQBUFFER_CLEARED;
+				arcmsr_post_Qbuffer(acb);
+			}
+			return count;
+		} else {
+			return 0;	/*need retry*/
+		}
+	}
+}
+
+static ssize_t
+arcmsr_sysfs_iop_message_clear(struct kobject *kobj, char *buf, loff_t off,
+    size_t count)
+{
+	struct class_device *cdev = container_of(kobj,struct class_device,kobj);
+	struct Scsi_Host *host = class_to_shost(cdev);
+	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+	struct MessageUnit __iomem *reg = acb->pmu;
+	uint8_t *pQbuffer;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EACCES;
+
+	if (acb->acb_flags & ACB_F_IOPDATA_OVERFLOW) {
+		acb->acb_flags &= ~ACB_F_IOPDATA_OVERFLOW;
+		writel(ARCMSR_INBOUND_DRIVER_DATA_READ_OK
+				, &reg->inbound_doorbell);
+	}
+	acb->acb_flags |=
+		(ACB_F_MESSAGE_WQBUFFER_CLEARED
+		| ACB_F_MESSAGE_RQBUFFER_CLEARED
+		| ACB_F_MESSAGE_WQBUFFER_READED);
+	acb->rqbuf_firstindex = 0;
+	acb->rqbuf_lastindex = 0;
+	acb->wqbuf_firstindex = 0;
+	acb->wqbuf_lastindex = 0;
+	pQbuffer = acb->rqbuffer;
+	memset(pQbuffer, 0, sizeof (struct QBUFFER));
+	pQbuffer = acb->wqbuffer;
+	memset(pQbuffer, 0, sizeof (struct QBUFFER));
+	return 1;
+}
+
+static struct bin_attribute arcmsr_sysfs_message_read_attr = {
+	.attr = {
+		.name = "mu_read",
+		.mode = S_IRUSR ,
+		.owner = THIS_MODULE,
+	},
+	.size = 1032,
+	.read = arcmsr_sysfs_iop_message_read,
+};
+
+static struct bin_attribute arcmsr_sysfs_message_write_attr = {
+	.attr = {
+		.name = "mu_write",
+		.mode = S_IWUSR,
+		.owner = THIS_MODULE,
+	},
+	.size = 1032,
+	.write = arcmsr_sysfs_iop_message_write,
+};
+
+static struct bin_attribute arcmsr_sysfs_message_clear_attr = {
+	.attr = {
+		.name = "mu_clear",
+		.mode = S_IWUSR,
+		.owner = THIS_MODULE,
+	},
+	.size = 1,
+	.write = arcmsr_sysfs_iop_message_clear,
+};
+
+int arcmsr_alloc_sysfs_attr(struct AdapterControlBlock *acb)
+{
+	struct Scsi_Host *host = acb->host;
+	int error;
+
+	error = sysfs_create_bin_file(&host->shost_classdev.kobj,
+				&arcmsr_sysfs_message_read_attr);
+	if (error) {
+		printk(KERN_ERR "arcmsr: alloc sysfs mu_read failed\n");
+		goto error_bin_file_message_read;
+	}
+	error = sysfs_create_bin_file(&host->shost_classdev.kobj,
+				&arcmsr_sysfs_message_write_attr);
+	if (error) {
+		printk(KERN_ERR "arcmsr: alloc sysfs mu_write failed\n");
+		goto error_bin_file_message_write;
+	}
+	error = sysfs_create_bin_file(&host->shost_classdev.kobj,
+				&arcmsr_sysfs_message_clear_attr);
+	if (error) {
+		printk(KERN_ERR "arcmsr: alloc sysfs mu_clear failed\n");
+		goto error_bin_file_message_clear;
+	}
+	return 0;
+error_bin_file_message_clear:
+	error = sysfs_remove_bin_file(&host->shost_classdev.kobj,
+				&arcmsr_sysfs_message_write_attr);
+	if (error)
+		printk(KERN_ERR "arcmsr: sysfs_remove_bin_file mu_write failed\n");
+error_bin_file_message_write:
+	error = sysfs_remove_bin_file(&host->shost_classdev.kobj,
+				&arcmsr_sysfs_message_read_attr);
+	if (error)
+		printk(KERN_ERR "arcmsr: sysfs_remove_bin_file mu_read failed\n");
+error_bin_file_message_read:
+	return error;
+}
+
+void
+arcmsr_free_sysfs_attr(struct AdapterControlBlock *acb) {
+	struct Scsi_Host *host = acb->host;
+	int error;
+
+	error = sysfs_remove_bin_file(&host->shost_classdev.kobj,
+				&arcmsr_sysfs_message_clear_attr);
+	if (error)
+		printk(KERN_ERR "arcmsr: free sysfs mu_clear failed\n");
+	error = sysfs_remove_bin_file(&host->shost_classdev.kobj,
+				&arcmsr_sysfs_message_write_attr);
+	if (error)
+		printk(KERN_ERR "arcmsr: free sysfs mu_write failed\n");
+	error = sysfs_remove_bin_file(&host->shost_classdev.kobj,
+				&arcmsr_sysfs_message_read_attr);
+	if (error)
+		printk(KERN_ERR "arcmsr: free sysfss mu_read failed\n");
+}
+
+
+static ssize_t
+arcmsr_attr_host_driver_version(struct class_device *cdev, char *buf) {
+	return snprintf(buf, PAGE_SIZE,
+			"ARCMSR: %s\n",
+			ARCMSR_DRIVER_VERSION);
+}
+
+static ssize_t
+arcmsr_attr_host_driver_posted_cmd(struct class_device *cdev, char *buf) {
+	struct Scsi_Host *host = class_to_shost(cdev);
+	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+	return snprintf(buf, PAGE_SIZE,
+			"Current commands posted:     %4d\n",
+			atomic_read(&acb->ccboutstandingcount));
+}
+
+static ssize_t
+arcmsr_attr_host_driver_reset(struct class_device *cdev, char *buf) {
+	struct Scsi_Host *host = class_to_shost(cdev);
+	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+	return snprintf(buf, PAGE_SIZE,
+			"SCSI Host Resets:            %4d\n",
+			acb->num_resets);
+}
+
+static ssize_t
+arcmsr_attr_host_driver_abort(struct class_device *cdev, char *buf) {
+	struct Scsi_Host *host = class_to_shost(cdev);
+	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+	return snprintf(buf, PAGE_SIZE,
+			"SCSI Aborts/Timeouts:        %4d\n",
+			acb->num_aborts);
+}
+
+static ssize_t
+arcmsr_attr_host_fw_model(struct class_device *cdev, char *buf) {
+    struct Scsi_Host *host = class_to_shost(cdev);
+	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+	return snprintf(buf, PAGE_SIZE,
+			"Adapter Model: %s\n",
+			acb->firm_model);
+}
+
+static ssize_t
+arcmsr_attr_host_fw_version(struct class_device *cdev, char *buf) {
+	struct Scsi_Host *host = class_to_shost(cdev);
+	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+
+	return snprintf(buf, PAGE_SIZE,
+			"Firmware Version:  %s\n",
+			acb->firm_version);
+}
+
+static ssize_t
+arcmsr_attr_host_fw_request_len(struct class_device *cdev, char *buf) {
+	struct Scsi_Host *host = class_to_shost(cdev);
+	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+
+	return snprintf(buf, PAGE_SIZE,
+			"Reguest Lenth: %4d\n",
+			acb->firm_request_len);
+}
+
+static ssize_t
+arcmsr_attr_host_fw_numbers_queue(struct class_device *cdev, char *buf) {
+	struct Scsi_Host *host = class_to_shost(cdev);
+	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+
+	return snprintf(buf, PAGE_SIZE,
+			"Numbers of Queue: %4d\n",
+			acb->firm_numbers_queue);
+}
+
+static ssize_t
+arcmsr_attr_host_fw_sdram_size(struct class_device *cdev, char *buf) {
+	struct Scsi_Host *host = class_to_shost(cdev);
+	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+
+	return snprintf(buf, PAGE_SIZE,
+			"SDRAM Size: %4d\n",
+			acb->firm_sdram_size);
+}
+
+static ssize_t
+arcmsr_attr_host_fw_hd_channels(struct class_device *cdev, char *buf) {
+	struct Scsi_Host *host = class_to_shost(cdev);
+	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+
+	return snprintf(buf, PAGE_SIZE,
+			"Hard Disk Channels: %4d\n",
+			acb->firm_hd_channels);
+}
+
+static CLASS_DEVICE_ATTR(host_driver_version, S_IRUGO, arcmsr_attr_host_driver_version, NULL);
+static CLASS_DEVICE_ATTR(host_driver_posted_cmd, S_IRUGO, arcmsr_attr_host_driver_posted_cmd, NULL);
+static CLASS_DEVICE_ATTR(host_driver_reset, S_IRUGO, arcmsr_attr_host_driver_reset, NULL);
+static CLASS_DEVICE_ATTR(host_driver_abort, S_IRUGO, arcmsr_attr_host_driver_abort, NULL);
+static CLASS_DEVICE_ATTR(host_fw_model, S_IRUGO, arcmsr_attr_host_fw_model, NULL);
+static CLASS_DEVICE_ATTR(host_fw_version, S_IRUGO, arcmsr_attr_host_fw_version, NULL);
+static CLASS_DEVICE_ATTR(host_fw_request_len, S_IRUGO, arcmsr_attr_host_fw_request_len, NULL);
+static CLASS_DEVICE_ATTR(host_fw_numbers_queue, S_IRUGO, arcmsr_attr_host_fw_numbers_queue, NULL);
+static CLASS_DEVICE_ATTR(host_fw_sdram_size, S_IRUGO, arcmsr_attr_host_fw_sdram_size, NULL);
+static CLASS_DEVICE_ATTR(host_fw_hd_channels, S_IRUGO, arcmsr_attr_host_fw_hd_channels, NULL);
+
+struct class_device_attribute *arcmsr_host_attrs[] = {
+	&class_device_attr_host_driver_version,
+	&class_device_attr_host_driver_posted_cmd,
+	&class_device_attr_host_driver_reset,
+	&class_device_attr_host_driver_abort,
+	&class_device_attr_host_fw_model,
+	&class_device_attr_host_fw_version,
+	&class_device_attr_host_fw_request_len,
+	&class_device_attr_host_fw_numbers_queue,
+	&class_device_attr_host_fw_sdram_size,
+	&class_device_attr_host_fw_hd_channels,
+	NULL,
+};
diff --git a/drivers/scsi/arcmsr/arcmsr_hba.c b/drivers/scsi/arcmsr/arcmsr_hba.c
new file mode 100644
index 000000000000..475f978ff8f0
--- /dev/null
+++ b/drivers/scsi/arcmsr/arcmsr_hba.c
@@ -0,0 +1,1496 @@
+/*
+*******************************************************************************
+**        O.S   : Linux
+**   FILE NAME  : arcmsr_hba.c
+**        BY    : Erich Chen
+**   Description: SCSI RAID Device Driver for
+**                ARECA RAID Host adapter
+*******************************************************************************
+** Copyright (C) 2002 - 2005, Areca Technology Corporation All rights reserved
+**
+**     Web site: www.areca.com.tw
+**       E-mail: erich@areca.com.tw
+**
+** This program is free software; you can redistribute it and/or modify
+** it under the terms of the GNU General Public License version 2 as
+** published by the Free Software Foundation.
+** This program is distributed in the hope that it will be useful,
+** but WITHOUT ANY WARRANTY; without even the implied warranty of
+** MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+** GNU General Public License for more details.
+*******************************************************************************
+** Redistribution and use in source and binary forms, with or without
+** modification, are permitted provided that the following conditions
+** are met:
+** 1. Redistributions of source code must retain the above copyright
+**    notice, this list of conditions and the following disclaimer.
+** 2. Redistributions in binary form must reproduce the above copyright
+**    notice, this list of conditions and the following disclaimer in the
+**    documentation and/or other materials provided with the distribution.
+** 3. The name of the author may not be used to endorse or promote products
+**    derived from this software without specific prior written permission.
+**
+** THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
+** IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
+** OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
+** IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
+** INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES(INCLUDING,BUT
+** NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+** DATA, OR PROFITS; OR BUSINESS INTERRUPTION)HOWEVER CAUSED AND ON ANY
+** THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+** (INCLUDING NEGLIGENCE OR OTHERWISE)ARISING IN ANY WAY OUT OF THE USE OF
+** THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*******************************************************************************
+** For history of changes, see Documentation/scsi/ChangeLog.arcmsr
+**     Firmware Specification, see Documentation/scsi/arcmsr_spec.txt
+*******************************************************************************
+*/
+#include <linux/module.h>
+#include <linux/reboot.h>
+#include <linux/spinlock.h>
+#include <linux/pci_ids.h>
+#include <linux/interrupt.h>
+#include <linux/moduleparam.h>
+#include <linux/errno.h>
+#include <linux/types.h>
+#include <linux/delay.h>
+#include <linux/dma-mapping.h>
+#include <linux/timer.h>
+#include <linux/pci.h>
+#include <asm/dma.h>
+#include <asm/io.h>
+#include <asm/system.h>
+#include <asm/uaccess.h>
+#include <scsi/scsi_host.h>
+#include <scsi/scsi.h>
+#include <scsi/scsi_cmnd.h>
+#include <scsi/scsi_tcq.h>
+#include <scsi/scsi_device.h>
+#include <scsi/scsi_transport.h>
+#include <scsi/scsicam.h>
+#include "arcmsr.h"
+
+MODULE_AUTHOR("Erich Chen <erich@areca.com.tw>");
+MODULE_DESCRIPTION("ARECA (ARC11xx/12xx) SATA RAID HOST Adapter");
+MODULE_LICENSE("Dual BSD/GPL");
+MODULE_VERSION(ARCMSR_DRIVER_VERSION);
+
+static int arcmsr_iop_message_xfer(struct AdapterControlBlock *acb, struct scsi_cmnd *cmd);
+static int arcmsr_abort(struct scsi_cmnd *);
+static int arcmsr_bus_reset(struct scsi_cmnd *);
+static int arcmsr_bios_param(struct scsi_device *sdev,
+				struct block_device *bdev, sector_t capacity, int *info);
+static int arcmsr_queue_command(struct scsi_cmnd * cmd,
+				void (*done) (struct scsi_cmnd *));
+static int arcmsr_probe(struct pci_dev *pdev,
+				const struct pci_device_id *id);
+static void arcmsr_remove(struct pci_dev *pdev);
+static void arcmsr_shutdown(struct pci_dev *pdev);
+static void arcmsr_iop_init(struct AdapterControlBlock *acb);
+static void arcmsr_free_ccb_pool(struct AdapterControlBlock *acb);
+static void arcmsr_stop_adapter_bgrb(struct AdapterControlBlock *acb);
+static void arcmsr_flush_adapter_cache(struct AdapterControlBlock *acb);
+static uint8_t arcmsr_wait_msgint_ready(struct AdapterControlBlock *acb);
+static const char *arcmsr_info(struct Scsi_Host *);
+static irqreturn_t arcmsr_interrupt(struct AdapterControlBlock *acb);
+
+static int arcmsr_adjust_disk_queue_depth(struct scsi_device *sdev, int queue_depth)
+{
+	if (queue_depth > ARCMSR_MAX_CMD_PERLUN)
+		queue_depth = ARCMSR_MAX_CMD_PERLUN;
+	scsi_adjust_queue_depth(sdev, MSG_ORDERED_TAG, queue_depth);
+	return queue_depth;
+}
+
+static struct scsi_host_template arcmsr_scsi_host_template = {
+	.module			= THIS_MODULE,
+	.name			= "ARCMSR ARECA SATA RAID HOST Adapter" ARCMSR_DRIVER_VERSION,
+	.info			= arcmsr_info,
+	.queuecommand		= arcmsr_queue_command,
+	.eh_abort_handler	= arcmsr_abort,
+	.eh_bus_reset_handler	= arcmsr_bus_reset,
+	.bios_param		= arcmsr_bios_param,
+	.change_queue_depth	= arcmsr_adjust_disk_queue_depth,
+	.can_queue		= ARCMSR_MAX_OUTSTANDING_CMD,
+	.this_id		= ARCMSR_SCSI_INITIATOR_ID,
+	.sg_tablesize		= ARCMSR_MAX_SG_ENTRIES,
+	.max_sectors    	= ARCMSR_MAX_XFER_SECTORS,
+	.cmd_per_lun		= ARCMSR_MAX_CMD_PERLUN,
+	.use_clustering		= ENABLE_CLUSTERING,
+	.shost_attrs		= arcmsr_host_attrs,
+};
+
+static struct pci_device_id arcmsr_device_id_table[] = {
+	{PCI_DEVICE(PCI_VENDOR_ID_ARECA, PCI_DEVICE_ID_ARECA_1110)},
+	{PCI_DEVICE(PCI_VENDOR_ID_ARECA, PCI_DEVICE_ID_ARECA_1120)},
+	{PCI_DEVICE(PCI_VENDOR_ID_ARECA, PCI_DEVICE_ID_ARECA_1130)},
+	{PCI_DEVICE(PCI_VENDOR_ID_ARECA, PCI_DEVICE_ID_ARECA_1160)},
+	{PCI_DEVICE(PCI_VENDOR_ID_ARECA, PCI_DEVICE_ID_ARECA_1170)},
+	{PCI_DEVICE(PCI_VENDOR_ID_ARECA, PCI_DEVICE_ID_ARECA_1210)},
+	{PCI_DEVICE(PCI_VENDOR_ID_ARECA, PCI_DEVICE_ID_ARECA_1220)},
+	{PCI_DEVICE(PCI_VENDOR_ID_ARECA, PCI_DEVICE_ID_ARECA_1230)},
+	{PCI_DEVICE(PCI_VENDOR_ID_ARECA, PCI_DEVICE_ID_ARECA_1260)},
+	{PCI_DEVICE(PCI_VENDOR_ID_ARECA, PCI_DEVICE_ID_ARECA_1270)},
+	{PCI_DEVICE(PCI_VENDOR_ID_ARECA, PCI_DEVICE_ID_ARECA_1280)},
+	{PCI_DEVICE(PCI_VENDOR_ID_ARECA, PCI_DEVICE_ID_ARECA_1380)},
+	{PCI_DEVICE(PCI_VENDOR_ID_ARECA, PCI_DEVICE_ID_ARECA_1381)},
+	{PCI_DEVICE(PCI_VENDOR_ID_ARECA, PCI_DEVICE_ID_ARECA_1680)},
+	{PCI_DEVICE(PCI_VENDOR_ID_ARECA, PCI_DEVICE_ID_ARECA_1681)},
+	{0, 0}, /* Terminating entry */
+};
+MODULE_DEVICE_TABLE(pci, arcmsr_device_id_table);
+static struct pci_driver arcmsr_pci_driver = {
+	.name			= "arcmsr",
+	.id_table		= arcmsr_device_id_table,
+	.probe			= arcmsr_probe,
+	.remove			= arcmsr_remove,
+	.shutdown		= arcmsr_shutdown
+};
+
+static irqreturn_t arcmsr_do_interrupt(int irq, void *dev_id,
+	struct pt_regs *regs)
+{
+	irqreturn_t handle_state;
+	struct AdapterControlBlock *acb;
+	unsigned long flags;
+
+	acb = (struct AdapterControlBlock *)dev_id;
+
+	spin_lock_irqsave(acb->host->host_lock, flags);
+	handle_state = arcmsr_interrupt(acb);
+	spin_unlock_irqrestore(acb->host->host_lock, flags);
+	return handle_state;
+}
+
+static int arcmsr_bios_param(struct scsi_device *sdev,
+		struct block_device *bdev, sector_t capacity, int *geom)
+{
+	int ret, heads, sectors, cylinders, total_capacity;
+	unsigned char *buffer;/* return copy of block device's partition table */
+
+	buffer = scsi_bios_ptable(bdev);
+	if (buffer) {
+		ret = scsi_partsize(buffer, capacity, &geom[2], &geom[0], &geom[1]);
+		kfree(buffer);
+		if (ret != -1)
+			return ret;
+	}
+	total_capacity = capacity;
+	heads = 64;
+	sectors = 32;
+	cylinders = total_capacity / (heads * sectors);
+	if (cylinders > 1024) {
+		heads = 255;
+		sectors = 63;
+		cylinders = total_capacity / (heads * sectors);
+	}
+	geom[0] = heads;
+	geom[1] = sectors;
+	geom[2] = cylinders;
+	return 0;
+}
+
+static int arcmsr_alloc_ccb_pool(struct AdapterControlBlock *acb)
+{
+	struct pci_dev *pdev = acb->pdev;
+	struct MessageUnit __iomem *reg = acb->pmu;
+	u32 ccb_phyaddr_hi32;
+	void *dma_coherent;
+	dma_addr_t dma_coherent_handle, dma_addr;
+	struct CommandControlBlock *ccb_tmp;
+	int i, j;
+
+	dma_coherent = dma_alloc_coherent(&pdev->dev,
+			ARCMSR_MAX_FREECCB_NUM *
+			sizeof (struct CommandControlBlock) + 0x20,
+			&dma_coherent_handle, GFP_KERNEL);
+	if (!dma_coherent)
+		return -ENOMEM;
+
+	acb->dma_coherent = dma_coherent;
+	acb->dma_coherent_handle = dma_coherent_handle;
+
+	if (((unsigned long)dma_coherent & 0x1F)) {
+		dma_coherent = dma_coherent +
+			(0x20 - ((unsigned long)dma_coherent & 0x1F));
+		dma_coherent_handle = dma_coherent_handle +
+			(0x20 - ((unsigned long)dma_coherent_handle & 0x1F));
+	}
+
+	dma_addr = dma_coherent_handle;
+	ccb_tmp = (struct CommandControlBlock *)dma_coherent;
+	for (i = 0; i < ARCMSR_MAX_FREECCB_NUM; i++) {
+		ccb_tmp->cdb_shifted_phyaddr = dma_addr >> 5;
+		ccb_tmp->acb = acb;
+		acb->pccb_pool[i] = ccb_tmp;
+		list_add_tail(&ccb_tmp->list, &acb->ccb_free_list);
+		dma_addr = dma_addr + sizeof (struct CommandControlBlock);
+		ccb_tmp++;
+	}
+
+	acb->vir2phy_offset = (unsigned long)ccb_tmp -
+			      (unsigned long)dma_addr;
+	for (i = 0; i < ARCMSR_MAX_TARGETID; i++)
+		for (j = 0; j < ARCMSR_MAX_TARGETLUN; j++)
+			acb->devstate[i][j] = ARECA_RAID_GOOD;
+
+	/*
+	** here we need to tell iop 331 our ccb_tmp.HighPart
+	** if ccb_tmp.HighPart is not zero
+	*/
+	ccb_phyaddr_hi32 = (uint32_t) ((dma_coherent_handle >> 16) >> 16);
+	if (ccb_phyaddr_hi32 != 0) {
+		writel(ARCMSR_SIGNATURE_SET_CONFIG, &reg->message_rwbuffer[0]);
+		writel(ccb_phyaddr_hi32, &reg->message_rwbuffer[1]);
+		writel(ARCMSR_INBOUND_MESG0_SET_CONFIG, &reg->inbound_msgaddr0);
+		if (arcmsr_wait_msgint_ready(acb))
+			printk(KERN_NOTICE "arcmsr%d: "
+			       "'set ccb high part physical address' timeout\n",
+				acb->host->host_no);
+	}
+
+	writel(readl(&reg->outbound_intmask) |
+			ARCMSR_MU_OUTBOUND_ALL_INTMASKENABLE,
+	       &reg->outbound_intmask);
+	return 0;
+}
+
+static int arcmsr_probe(struct pci_dev *pdev,
+	const struct pci_device_id *id)
+{
+	struct Scsi_Host *host;
+	struct AdapterControlBlock *acb;
+	uint8_t bus, dev_fun;
+	int error;
+
+	error = pci_enable_device(pdev);
+	if (error)
+		goto out;
+	pci_set_master(pdev);
+
+	host = scsi_host_alloc(&arcmsr_scsi_host_template,
+			sizeof(struct AdapterControlBlock));
+	if (!host) {
+		error = -ENOMEM;
+		goto out_disable_device;
+	}
+	acb = (struct AdapterControlBlock *)host->hostdata;
+	memset(acb, 0, sizeof (struct AdapterControlBlock));
+
+	error = pci_set_dma_mask(pdev, DMA_64BIT_MASK);
+	if (error) {
+		error = pci_set_dma_mask(pdev, DMA_32BIT_MASK);
+		if (error) {
+			printk(KERN_WARNING
+			       "scsi%d: No suitable DMA mask available\n",
+			       host->host_no);
+			goto out_host_put;
+		}
+	}
+	bus = pdev->bus->number;
+	dev_fun = pdev->devfn;
+	acb->host = host;
+	acb->pdev = pdev;
+	host->max_sectors = ARCMSR_MAX_XFER_SECTORS;
+	host->max_lun = ARCMSR_MAX_TARGETLUN;
+	host->max_id = ARCMSR_MAX_TARGETID;/*16:8*/
+	host->max_cmd_len = 16;    /*this is issue of 64bit LBA, over 2T byte*/
+	host->sg_tablesize = ARCMSR_MAX_SG_ENTRIES;
+	host->can_queue = ARCMSR_MAX_FREECCB_NUM; /* max simultaneous cmds */
+	host->cmd_per_lun = ARCMSR_MAX_CMD_PERLUN;
+	host->this_id = ARCMSR_SCSI_INITIATOR_ID;
+	host->unique_id = (bus << 8) | dev_fun;
+	host->irq = pdev->irq;
+	error = pci_request_regions(pdev, "arcmsr");
+	if (error)
+		goto out_host_put;
+
+	acb->pmu = ioremap(pci_resource_start(pdev, 0),
+			   pci_resource_len(pdev, 0));
+	if (!acb->pmu) {
+		printk(KERN_NOTICE "arcmsr%d: memory"
+			" mapping region fail \n", acb->host->host_no);
+		goto out_release_regions;
+	}
+	acb->acb_flags |= (ACB_F_MESSAGE_WQBUFFER_CLEARED |
+			   ACB_F_MESSAGE_RQBUFFER_CLEARED |
+			   ACB_F_MESSAGE_WQBUFFER_READED);
+	acb->acb_flags &= ~ACB_F_SCSISTOPADAPTER;
+	INIT_LIST_HEAD(&acb->ccb_free_list);
+
+	error = arcmsr_alloc_ccb_pool(acb);
+	if (error)
+		goto out_iounmap;
+
+	error = request_irq(pdev->irq, arcmsr_do_interrupt,
+			SA_INTERRUPT | SA_SHIRQ, "arcmsr", acb);
+	if (error)
+		goto out_free_ccb_pool;
+
+	arcmsr_iop_init(acb);
+	pci_set_drvdata(pdev, host);
+
+	error = scsi_add_host(host, &pdev->dev);
+	if (error)
+		goto out_free_irq;
+
+	error = arcmsr_alloc_sysfs_attr(acb);
+	if (error)
+		goto out_free_sysfs;
+
+	scsi_scan_host(host);
+	return 0;
+ out_free_sysfs:
+ out_free_irq:
+	free_irq(pdev->irq, acb);
+ out_free_ccb_pool:
+	arcmsr_free_ccb_pool(acb);
+ out_iounmap:
+	iounmap(acb->pmu);
+ out_release_regions:
+	pci_release_regions(pdev);
+ out_host_put:
+	scsi_host_put(host);
+ out_disable_device:
+	pci_disable_device(pdev);
+ out:
+	return error;
+}
+
+static void arcmsr_abort_allcmd(struct AdapterControlBlock *acb)
+{
+	struct MessageUnit __iomem *reg = acb->pmu;
+
+	writel(ARCMSR_INBOUND_MESG0_ABORT_CMD, &reg->inbound_msgaddr0);
+	if (arcmsr_wait_msgint_ready(acb))
+		printk(KERN_NOTICE
+			"arcmsr%d: wait 'abort all outstanding command' timeout \n"
+			, acb->host->host_no);
+}
+
+static void arcmsr_pci_unmap_dma(struct CommandControlBlock *ccb)
+{
+	struct AdapterControlBlock *acb = ccb->acb;
+	struct scsi_cmnd *pcmd = ccb->pcmd;
+
+	if (pcmd->use_sg != 0) {
+		struct scatterlist *sl;
+
+		sl = (struct scatterlist *)pcmd->request_buffer;
+		pci_unmap_sg(acb->pdev, sl, pcmd->use_sg, pcmd->sc_data_direction);
+	}
+	else if (pcmd->request_bufflen != 0)
+		pci_unmap_single(acb->pdev,
+			pcmd->SCp.dma_handle,
+			pcmd->request_bufflen, pcmd->sc_data_direction);
+}
+
+static void arcmsr_ccb_complete(struct CommandControlBlock *ccb, int stand_flag)
+{
+	struct AdapterControlBlock *acb = ccb->acb;
+	struct scsi_cmnd *pcmd = ccb->pcmd;
+
+	arcmsr_pci_unmap_dma(ccb);
+	if (stand_flag == 1)
+		atomic_dec(&acb->ccboutstandingcount);
+	ccb->startdone = ARCMSR_CCB_DONE;
+	ccb->ccb_flags = 0;
+	list_add_tail(&ccb->list, &acb->ccb_free_list);
+	pcmd->scsi_done(pcmd);
+}
+
+static void arcmsr_remove(struct pci_dev *pdev)
+{
+	struct Scsi_Host *host = pci_get_drvdata(pdev);
+	struct AdapterControlBlock *acb =
+		(struct AdapterControlBlock *) host->hostdata;
+	struct MessageUnit __iomem *reg = acb->pmu;
+	int poll_count = 0;
+
+	arcmsr_free_sysfs_attr(acb);
+	scsi_remove_host(host);
+	arcmsr_stop_adapter_bgrb(acb);
+	arcmsr_flush_adapter_cache(acb);
+	writel(readl(&reg->outbound_intmask) |
+		ARCMSR_MU_OUTBOUND_ALL_INTMASKENABLE,
+		&reg->outbound_intmask);
+	acb->acb_flags |= ACB_F_SCSISTOPADAPTER;
+	acb->acb_flags &= ~ACB_F_IOP_INITED;
+
+	for (poll_count = 0; poll_count < 256; poll_count++) {
+		if (!atomic_read(&acb->ccboutstandingcount))
+			break;
+		arcmsr_interrupt(acb);
+		msleep(25);
+	}
+
+	if (atomic_read(&acb->ccboutstandingcount)) {
+		int i;
+
+		arcmsr_abort_allcmd(acb);
+		for (i = 0; i < ARCMSR_MAX_OUTSTANDING_CMD; i++)
+			readl(&reg->outbound_queueport);
+		for (i = 0; i < ARCMSR_MAX_FREECCB_NUM; i++) {
+			struct CommandControlBlock *ccb = acb->pccb_pool[i];
+			if (ccb->startdone == ARCMSR_CCB_START) {
+				ccb->startdone = ARCMSR_CCB_ABORTED;
+				ccb->pcmd->result = DID_ABORT << 16;
+				arcmsr_ccb_complete(ccb, 1);
+			}
+		}
+	}
+
+	free_irq(pdev->irq, acb);
+	iounmap(acb->pmu);
+	arcmsr_free_ccb_pool(acb);
+	pci_release_regions(pdev);
+
+	scsi_host_put(host);
+
+	pci_disable_device(pdev);
+	pci_set_drvdata(pdev, NULL);
+}
+
+static void arcmsr_shutdown(struct pci_dev *pdev)
+{
+	struct Scsi_Host *host = pci_get_drvdata(pdev);
+	struct AdapterControlBlock *acb =
+		(struct AdapterControlBlock *)host->hostdata;
+
+	arcmsr_stop_adapter_bgrb(acb);
+	arcmsr_flush_adapter_cache(acb);
+}
+
+static int arcmsr_module_init(void)
+{
+	int error = 0;
+
+	error = pci_register_driver(&arcmsr_pci_driver);
+	return error;
+}
+
+static void arcmsr_module_exit(void)
+{
+	pci_unregister_driver(&arcmsr_pci_driver);
+}
+module_init(arcmsr_module_init);
+module_exit(arcmsr_module_exit);
+
+static u32 arcmsr_disable_outbound_ints(struct AdapterControlBlock *acb)
+{
+	struct MessageUnit __iomem *reg = acb->pmu;
+	u32 orig_mask = readl(&reg->outbound_intmask);
+
+	writel(orig_mask | ARCMSR_MU_OUTBOUND_ALL_INTMASKENABLE,
+			&reg->outbound_intmask);
+	return orig_mask;
+}
+
+static void arcmsr_enable_outbound_ints(struct AdapterControlBlock *acb,
+		u32 orig_mask)
+{
+	struct MessageUnit __iomem *reg = acb->pmu;
+	u32 mask;
+
+	mask = orig_mask & ~(ARCMSR_MU_OUTBOUND_POSTQUEUE_INTMASKENABLE |
+			     ARCMSR_MU_OUTBOUND_DOORBELL_INTMASKENABLE);
+	writel(mask, &reg->outbound_intmask);
+}
+
+static void arcmsr_flush_adapter_cache(struct AdapterControlBlock *acb)
+{
+	struct MessageUnit __iomem *reg=acb->pmu;
+
+	writel(ARCMSR_INBOUND_MESG0_FLUSH_CACHE, &reg->inbound_msgaddr0);
+	if (arcmsr_wait_msgint_ready(acb))
+		printk(KERN_NOTICE
+			"arcmsr%d: wait 'flush adapter cache' timeout \n"
+			, acb->host->host_no);
+}
+
+static void arcmsr_report_sense_info(struct CommandControlBlock *ccb)
+{
+	struct scsi_cmnd *pcmd = ccb->pcmd;
+	struct SENSE_DATA *sensebuffer = (struct SENSE_DATA *)pcmd->sense_buffer;
+
+	pcmd->result = DID_OK << 16;
+	if (sensebuffer) {
+		int sense_data_length =
+			sizeof (struct SENSE_DATA) < sizeof (pcmd->sense_buffer)
+			? sizeof (struct SENSE_DATA) : sizeof (pcmd->sense_buffer);
+		memset(sensebuffer, 0, sizeof (pcmd->sense_buffer));
+		memcpy(sensebuffer, ccb->arcmsr_cdb.SenseData, sense_data_length);
+		sensebuffer->ErrorCode = SCSI_SENSE_CURRENT_ERRORS;
+		sensebuffer->Valid = 1;
+	}
+}
+
+static uint8_t arcmsr_wait_msgint_ready(struct AdapterControlBlock *acb)
+{
+	struct MessageUnit __iomem *reg = acb->pmu;
+	uint32_t Index;
+	uint8_t Retries = 0x00;
+
+	do {
+		for (Index = 0; Index < 100; Index++) {
+			if (readl(&reg->outbound_intstatus)
+				& ARCMSR_MU_OUTBOUND_MESSAGE0_INT) {
+				writel(ARCMSR_MU_OUTBOUND_MESSAGE0_INT
+					, &reg->outbound_intstatus);
+				return 0x00;
+			}
+			msleep_interruptible(10);
+		}/*max 1 seconds*/
+	} while (Retries++ < 20);/*max 20 sec*/
+	return 0xff;
+}
+
+static void arcmsr_build_ccb(struct AdapterControlBlock *acb,
+	struct CommandControlBlock *ccb, struct scsi_cmnd *pcmd)
+{
+	struct ARCMSR_CDB *arcmsr_cdb = (struct ARCMSR_CDB *)&ccb->arcmsr_cdb;
+	int8_t *psge = (int8_t *)&arcmsr_cdb->u;
+	uint32_t address_lo, address_hi;
+	int arccdbsize = 0x30;
+
+	ccb->pcmd = pcmd;
+	memset(arcmsr_cdb, 0, sizeof (struct ARCMSR_CDB));
+	arcmsr_cdb->Bus = 0;
+	arcmsr_cdb->TargetID = pcmd->device->id;
+	arcmsr_cdb->LUN = pcmd->device->lun;
+	arcmsr_cdb->Function = 1;
+	arcmsr_cdb->CdbLength = (uint8_t)pcmd->cmd_len;
+	arcmsr_cdb->Context = (unsigned long)arcmsr_cdb;
+	memcpy(arcmsr_cdb->Cdb, pcmd->cmnd, pcmd->cmd_len);
+	if (pcmd->use_sg) {
+		int length, sgcount, i, cdb_sgcount = 0;
+		struct scatterlist *sl;
+
+		/* Get Scatter Gather List from scsiport. */
+		sl = (struct scatterlist *) pcmd->request_buffer;
+		sgcount = pci_map_sg(acb->pdev, sl, pcmd->use_sg,
+				pcmd->sc_data_direction);
+		/* map stor port SG list to our iop SG List. */
+		for (i = 0; i < sgcount; i++) {
+			/* Get the physical address of the current data pointer */
+			length = cpu_to_le32(sg_dma_len(sl));
+			address_lo = cpu_to_le32(dma_addr_lo32(sg_dma_address(sl)));
+			address_hi = cpu_to_le32(dma_addr_hi32(sg_dma_address(sl)));
+			if (address_hi == 0) {
+				struct SG32ENTRY *pdma_sg = (struct SG32ENTRY *)psge;
+
+				pdma_sg->address = address_lo;
+				pdma_sg->length = length;
+				psge += sizeof (struct SG32ENTRY);
+				arccdbsize += sizeof (struct SG32ENTRY);
+			} else {
+				struct SG64ENTRY *pdma_sg = (struct SG64ENTRY *)psge;
+
+				pdma_sg->addresshigh = address_hi;
+				pdma_sg->address = address_lo;
+				pdma_sg->length = length|IS_SG64_ADDR;
+				psge += sizeof (struct SG64ENTRY);
+				arccdbsize += sizeof (struct SG64ENTRY);
+			}
+			sl++;
+			cdb_sgcount++;
+		}
+		arcmsr_cdb->sgcount = (uint8_t)cdb_sgcount;
+		arcmsr_cdb->DataLength = pcmd->request_bufflen;
+		if ( arccdbsize > 256)
+			arcmsr_cdb->Flags |= ARCMSR_CDB_FLAG_SGL_BSIZE;
+	} else if (pcmd->request_bufflen) {
+		dma_addr_t dma_addr;
+		dma_addr = pci_map_single(acb->pdev, pcmd->request_buffer,
+				pcmd->request_bufflen, pcmd->sc_data_direction);
+		pcmd->SCp.dma_handle = dma_addr;
+		address_lo = cpu_to_le32(dma_addr_lo32(dma_addr));
+		address_hi = cpu_to_le32(dma_addr_hi32(dma_addr));
+		if (address_hi == 0) {
+			struct  SG32ENTRY *pdma_sg = (struct SG32ENTRY *)psge;
+			pdma_sg->address = address_lo;
+			pdma_sg->length = pcmd->request_bufflen;
+		} else {
+			struct SG64ENTRY *pdma_sg = (struct SG64ENTRY *)psge;
+			pdma_sg->addresshigh = address_hi;
+			pdma_sg->address = address_lo;
+			pdma_sg->length = pcmd->request_bufflen|IS_SG64_ADDR;
+		}
+		arcmsr_cdb->sgcount = 1;
+		arcmsr_cdb->DataLength = pcmd->request_bufflen;
+	}
+	if (pcmd->sc_data_direction == DMA_TO_DEVICE ) {
+		arcmsr_cdb->Flags |= ARCMSR_CDB_FLAG_WRITE;
+		ccb->ccb_flags |= CCB_FLAG_WRITE;
+	}
+}
+
+static void arcmsr_post_ccb(struct AdapterControlBlock *acb, struct CommandControlBlock *ccb)
+{
+	struct MessageUnit __iomem *reg = acb->pmu;
+	uint32_t cdb_shifted_phyaddr = ccb->cdb_shifted_phyaddr;
+	struct ARCMSR_CDB *arcmsr_cdb = (struct ARCMSR_CDB *)&ccb->arcmsr_cdb;
+
+	atomic_inc(&acb->ccboutstandingcount);
+	ccb->startdone = ARCMSR_CCB_START;
+	if (arcmsr_cdb->Flags & ARCMSR_CDB_FLAG_SGL_BSIZE)
+		writel(cdb_shifted_phyaddr | ARCMSR_CCBPOST_FLAG_SGL_BSIZE,
+			&reg->inbound_queueport);
+	else
+		writel(cdb_shifted_phyaddr, &reg->inbound_queueport);
+}
+
+void arcmsr_post_Qbuffer(struct AdapterControlBlock *acb)
+{
+	struct MessageUnit __iomem *reg = acb->pmu;
+	struct QBUFFER __iomem *pwbuffer = (struct QBUFFER __iomem *) &reg->message_wbuffer;
+	uint8_t __iomem *iop_data = (uint8_t __iomem *) pwbuffer->data;
+	int32_t allxfer_len = 0;
+
+	if (acb->acb_flags & ACB_F_MESSAGE_WQBUFFER_READED) {
+		acb->acb_flags &= (~ACB_F_MESSAGE_WQBUFFER_READED);
+		while ((acb->wqbuf_firstindex != acb->wqbuf_lastindex)
+			&& (allxfer_len < 124)) {
+			writeb(acb->wqbuffer[acb->wqbuf_firstindex], iop_data);
+			acb->wqbuf_firstindex++;
+			acb->wqbuf_firstindex %= ARCMSR_MAX_QBUFFER;
+			iop_data++;
+			allxfer_len++;
+		}
+		writel(allxfer_len, &pwbuffer->data_len);
+		writel(ARCMSR_INBOUND_DRIVER_DATA_WRITE_OK
+			, &reg->inbound_doorbell);
+	}
+}
+
+static void arcmsr_stop_adapter_bgrb(struct AdapterControlBlock *acb)
+{
+	struct MessageUnit __iomem *reg = acb->pmu;
+
+	acb->acb_flags &= ~ACB_F_MSG_START_BGRB;
+	writel(ARCMSR_INBOUND_MESG0_STOP_BGRB, &reg->inbound_msgaddr0);
+	if (arcmsr_wait_msgint_ready(acb))
+		printk(KERN_NOTICE
+			"arcmsr%d: wait 'stop adapter background rebulid' timeout \n"
+			, acb->host->host_no);
+}
+
+static void arcmsr_free_ccb_pool(struct AdapterControlBlock *acb)
+{
+	dma_free_coherent(&acb->pdev->dev,
+		ARCMSR_MAX_FREECCB_NUM * sizeof (struct CommandControlBlock) + 0x20,
+		acb->dma_coherent,
+		acb->dma_coherent_handle);
+}
+
+static irqreturn_t arcmsr_interrupt(struct AdapterControlBlock *acb)
+{
+	struct MessageUnit __iomem *reg = acb->pmu;
+	struct CommandControlBlock *ccb;
+	uint32_t flag_ccb, outbound_intstatus, outbound_doorbell;
+
+	outbound_intstatus = readl(&reg->outbound_intstatus)
+		& acb->outbound_int_enable;
+	writel(outbound_intstatus, &reg->outbound_intstatus);
+	if (outbound_intstatus & ARCMSR_MU_OUTBOUND_DOORBELL_INT) {
+		outbound_doorbell = readl(&reg->outbound_doorbell);
+		writel(outbound_doorbell, &reg->outbound_doorbell);
+		if (outbound_doorbell & ARCMSR_OUTBOUND_IOP331_DATA_WRITE_OK) {
+			struct QBUFFER __iomem * prbuffer =
+				(struct QBUFFER __iomem *) &reg->message_rbuffer;
+			uint8_t __iomem * iop_data = (uint8_t __iomem *)prbuffer->data;
+			int32_t my_empty_len, iop_len, rqbuf_firstindex, rqbuf_lastindex;
+
+			rqbuf_lastindex = acb->rqbuf_lastindex;
+			rqbuf_firstindex = acb->rqbuf_firstindex;
+			iop_len = readl(&prbuffer->data_len);
+			my_empty_len = (rqbuf_firstindex - rqbuf_lastindex - 1)
+					&(ARCMSR_MAX_QBUFFER - 1);
+			if (my_empty_len >= iop_len) {
+				while (iop_len > 0) {
+					acb->rqbuffer[acb->rqbuf_lastindex] = readb(iop_data);
+					acb->rqbuf_lastindex++;
+					acb->rqbuf_lastindex %= ARCMSR_MAX_QBUFFER;
+					iop_data++;
+					iop_len--;
+				}
+				writel(ARCMSR_INBOUND_DRIVER_DATA_READ_OK,
+					&reg->inbound_doorbell);
+			} else
+				acb->acb_flags |= ACB_F_IOPDATA_OVERFLOW;
+		}
+		if (outbound_doorbell & ARCMSR_OUTBOUND_IOP331_DATA_READ_OK) {
+			acb->acb_flags |= ACB_F_MESSAGE_WQBUFFER_READED;
+			if (acb->wqbuf_firstindex != acb->wqbuf_lastindex) {
+				struct QBUFFER __iomem * pwbuffer =
+						(struct QBUFFER __iomem *) &reg->message_wbuffer;
+				uint8_t __iomem * iop_data = (uint8_t __iomem *) pwbuffer->data;
+				int32_t allxfer_len = 0;
+
+				acb->acb_flags &= (~ACB_F_MESSAGE_WQBUFFER_READED);
+				while ((acb->wqbuf_firstindex != acb->wqbuf_lastindex)
+					&& (allxfer_len < 124)) {
+					writeb(acb->wqbuffer[acb->wqbuf_firstindex], iop_data);
+					acb->wqbuf_firstindex++;
+					acb->wqbuf_firstindex %= ARCMSR_MAX_QBUFFER;
+					iop_data++;
+					allxfer_len++;
+				}
+				writel(allxfer_len, &pwbuffer->data_len);
+				writel(ARCMSR_INBOUND_DRIVER_DATA_WRITE_OK,
+					&reg->inbound_doorbell);
+			}
+			if (acb->wqbuf_firstindex == acb->wqbuf_lastindex)
+				acb->acb_flags |= ACB_F_MESSAGE_WQBUFFER_CLEARED;
+		}
+	}
+	if (outbound_intstatus & ARCMSR_MU_OUTBOUND_POSTQUEUE_INT) {
+		int id, lun;
+		/*
+		****************************************************************
+		**               areca cdb command done
+		****************************************************************
+		*/
+		while (1) {
+			if ((flag_ccb = readl(&reg->outbound_queueport)) == 0xFFFFFFFF)
+				break;/*chip FIFO no ccb for completion already*/
+			/* check if command done with no error*/
+			ccb = (struct CommandControlBlock *)(acb->vir2phy_offset +
+				(flag_ccb << 5));
+			if ((ccb->acb != acb) || (ccb->startdone != ARCMSR_CCB_START)) {
+				if (ccb->startdone == ARCMSR_CCB_ABORTED) {
+					struct scsi_cmnd *abortcmd=ccb->pcmd;
+					if (abortcmd) {
+					abortcmd->result |= DID_ABORT >> 16;
+					arcmsr_ccb_complete(ccb, 1);
+					printk(KERN_NOTICE
+						"arcmsr%d: ccb='0x%p' isr got aborted command \n"
+						, acb->host->host_no, ccb);
+					}
+					continue;
+				}
+				printk(KERN_NOTICE
+					"arcmsr%d: isr get an illegal ccb command done acb='0x%p'"
+					"ccb='0x%p' ccbacb='0x%p' startdone = 0x%x"
+					" ccboutstandingcount=%d \n"
+					, acb->host->host_no
+					, acb
+					, ccb
+					, ccb->acb
+					, ccb->startdone
+					, atomic_read(&acb->ccboutstandingcount));
+				continue;
+			}
+			id = ccb->pcmd->device->id;
+			lun = ccb->pcmd->device->lun;
+			if (!(flag_ccb & ARCMSR_CCBREPLY_FLAG_ERROR)) {
+				if (acb->devstate[id][lun] == ARECA_RAID_GONE)
+					acb->devstate[id][lun] = ARECA_RAID_GOOD;
+				ccb->pcmd->result = DID_OK << 16;
+				arcmsr_ccb_complete(ccb, 1);
+			} else {
+				switch(ccb->arcmsr_cdb.DeviceStatus) {
+				case ARCMSR_DEV_SELECT_TIMEOUT: {
+						acb->devstate[id][lun] = ARECA_RAID_GONE;
+						ccb->pcmd->result = DID_TIME_OUT << 16;
+						arcmsr_ccb_complete(ccb, 1);
+					}
+					break;
+				case ARCMSR_DEV_ABORTED:
+				case ARCMSR_DEV_INIT_FAIL: {
+						acb->devstate[id][lun] = ARECA_RAID_GONE;
+						ccb->pcmd->result = DID_BAD_TARGET << 16;
+						arcmsr_ccb_complete(ccb, 1);
+					}
+					break;
+				case ARCMSR_DEV_CHECK_CONDITION: {
+						acb->devstate[id][lun] = ARECA_RAID_GOOD;
+						arcmsr_report_sense_info(ccb);
+						arcmsr_ccb_complete(ccb, 1);
+					}
+					break;
+				default:
+					printk(KERN_NOTICE
+						"arcmsr%d: scsi id=%d lun=%d"
+						" isr get command error done,"
+						"but got unknown DeviceStatus = 0x%x \n"
+						, acb->host->host_no
+						, id
+						, lun
+						, ccb->arcmsr_cdb.DeviceStatus);
+						acb->devstate[id][lun] = ARECA_RAID_GONE;
+						ccb->pcmd->result = DID_NO_CONNECT << 16;
+						arcmsr_ccb_complete(ccb, 1);
+					break;
+				}
+			}
+		}/*drain reply FIFO*/
+	}
+	if (!(outbound_intstatus & ARCMSR_MU_OUTBOUND_HANDLE_INT))
+		return IRQ_NONE;
+	return IRQ_HANDLED;
+}
+
+static void arcmsr_iop_parking(struct AdapterControlBlock *acb)
+{
+	if (acb) {
+		/* stop adapter background rebuild */
+		if (acb->acb_flags & ACB_F_MSG_START_BGRB) {
+			acb->acb_flags &= ~ACB_F_MSG_START_BGRB;
+			arcmsr_stop_adapter_bgrb(acb);
+			arcmsr_flush_adapter_cache(acb);
+		}
+	}
+}
+
+static int arcmsr_iop_message_xfer(struct AdapterControlBlock *acb, struct scsi_cmnd *cmd)
+{
+	struct MessageUnit __iomem *reg = acb->pmu;
+	struct CMD_MESSAGE_FIELD *pcmdmessagefld;
+	int retvalue = 0, transfer_len = 0;
+	char *buffer;
+	uint32_t controlcode = (uint32_t ) cmd->cmnd[5] << 24 |
+						(uint32_t ) cmd->cmnd[6] << 16 |
+						(uint32_t ) cmd->cmnd[7] << 8  |
+						(uint32_t ) cmd->cmnd[8];
+					/* 4 bytes: Areca io control code */
+	if (cmd->use_sg) {
+		struct scatterlist *sg = (struct scatterlist *)cmd->request_buffer;
+
+		buffer = kmap_atomic(sg->page, KM_IRQ0) + sg->offset;
+		if (cmd->use_sg > 1) {
+			retvalue = ARCMSR_MESSAGE_FAIL;
+			goto message_out;
+		}
+		transfer_len += sg->length;
+	} else {
+		buffer = cmd->request_buffer;
+		transfer_len = cmd->request_bufflen;
+	}
+	if (transfer_len > sizeof(struct CMD_MESSAGE_FIELD)) {
+		retvalue = ARCMSR_MESSAGE_FAIL;
+		goto message_out;
+	}
+	pcmdmessagefld = (struct CMD_MESSAGE_FIELD *) buffer;
+	switch(controlcode) {
+	case ARCMSR_MESSAGE_READ_RQBUFFER: {
+			unsigned long *ver_addr;
+			dma_addr_t buf_handle;
+			uint8_t *pQbuffer, *ptmpQbuffer;
+			int32_t allxfer_len = 0;
+
+			ver_addr = pci_alloc_consistent(acb->pdev, 1032, &buf_handle);
+			if (!ver_addr) {
+				retvalue = ARCMSR_MESSAGE_FAIL;
+				goto message_out;
+			}
+			ptmpQbuffer = (uint8_t *) ver_addr;
+			while ((acb->rqbuf_firstindex != acb->rqbuf_lastindex)
+				&& (allxfer_len < 1031)) {
+				pQbuffer = &acb->rqbuffer[acb->rqbuf_firstindex];
+				memcpy(ptmpQbuffer, pQbuffer, 1);
+				acb->rqbuf_firstindex++;
+				acb->rqbuf_firstindex %= ARCMSR_MAX_QBUFFER;
+				ptmpQbuffer++;
+				allxfer_len++;
+			}
+			if (acb->acb_flags & ACB_F_IOPDATA_OVERFLOW) {
+				struct QBUFFER __iomem * prbuffer = (struct QBUFFER __iomem *)
+							&reg->message_rbuffer;
+				uint8_t __iomem * iop_data = (uint8_t __iomem *)prbuffer->data;
+				int32_t iop_len;
+
+				acb->acb_flags &= ~ACB_F_IOPDATA_OVERFLOW;
+				iop_len = readl(&prbuffer->data_len);
+				while (iop_len > 0) {
+					acb->rqbuffer[acb->rqbuf_lastindex] = readb(iop_data);
+					acb->rqbuf_lastindex++;
+					acb->rqbuf_lastindex %= ARCMSR_MAX_QBUFFER;
+					iop_data++;
+					iop_len--;
+				}
+				writel(ARCMSR_INBOUND_DRIVER_DATA_READ_OK,
+						&reg->inbound_doorbell);
+			}
+			memcpy(pcmdmessagefld->messagedatabuffer,
+				(uint8_t *)ver_addr, allxfer_len);
+			pcmdmessagefld->cmdmessage.Length = allxfer_len;
+			pcmdmessagefld->cmdmessage.ReturnCode = ARCMSR_MESSAGE_RETURNCODE_OK;
+			pci_free_consistent(acb->pdev, 1032, ver_addr, buf_handle);
+		}
+		break;
+	case ARCMSR_MESSAGE_WRITE_WQBUFFER: {
+			unsigned long *ver_addr;
+			dma_addr_t buf_handle;
+			int32_t my_empty_len, user_len, wqbuf_firstindex, wqbuf_lastindex;
+			uint8_t *pQbuffer, *ptmpuserbuffer;
+
+			ver_addr = pci_alloc_consistent(acb->pdev, 1032, &buf_handle);
+			if (!ver_addr) {
+				retvalue = ARCMSR_MESSAGE_FAIL;
+				goto message_out;
+			}
+			ptmpuserbuffer = (uint8_t *)ver_addr;
+			user_len = pcmdmessagefld->cmdmessage.Length;
+			memcpy(ptmpuserbuffer, pcmdmessagefld->messagedatabuffer, user_len);
+			wqbuf_lastindex = acb->wqbuf_lastindex;
+			wqbuf_firstindex = acb->wqbuf_firstindex;
+			if (wqbuf_lastindex != wqbuf_firstindex) {
+				struct SENSE_DATA *sensebuffer =
+					(struct SENSE_DATA *)cmd->sense_buffer;
+				arcmsr_post_Qbuffer(acb);
+				/* has error report sensedata */
+				sensebuffer->ErrorCode = 0x70;
+				sensebuffer->SenseKey = ILLEGAL_REQUEST;
+				sensebuffer->AdditionalSenseLength = 0x0A;
+				sensebuffer->AdditionalSenseCode = 0x20;
+				sensebuffer->Valid = 1;
+				retvalue = ARCMSR_MESSAGE_FAIL;
+			} else {
+				my_empty_len = (wqbuf_firstindex-wqbuf_lastindex - 1)
+						&(ARCMSR_MAX_QBUFFER - 1);
+				if (my_empty_len >= user_len) {
+					while (user_len > 0) {
+						pQbuffer =
+						&acb->wqbuffer[acb->wqbuf_lastindex];
+						memcpy(pQbuffer, ptmpuserbuffer, 1);
+						acb->wqbuf_lastindex++;
+						acb->wqbuf_lastindex %= ARCMSR_MAX_QBUFFER;
+						ptmpuserbuffer++;
+						user_len--;
+					}
+					if (acb->acb_flags & ACB_F_MESSAGE_WQBUFFER_CLEARED) {
+						acb->acb_flags &=
+							~ACB_F_MESSAGE_WQBUFFER_CLEARED;
+						arcmsr_post_Qbuffer(acb);
+					}
+				} else {
+					/* has error report sensedata */
+					struct SENSE_DATA *sensebuffer =
+						(struct SENSE_DATA *)cmd->sense_buffer;
+					sensebuffer->ErrorCode = 0x70;
+					sensebuffer->SenseKey = ILLEGAL_REQUEST;
+					sensebuffer->AdditionalSenseLength = 0x0A;
+					sensebuffer->AdditionalSenseCode = 0x20;
+					sensebuffer->Valid = 1;
+					retvalue = ARCMSR_MESSAGE_FAIL;
+				}
+			}
+			pci_free_consistent(acb->pdev, 1032, ver_addr, buf_handle);
+		}
+		break;
+	case ARCMSR_MESSAGE_CLEAR_RQBUFFER: {
+			uint8_t *pQbuffer = acb->rqbuffer;
+
+			if (acb->acb_flags & ACB_F_IOPDATA_OVERFLOW) {
+				acb->acb_flags &= ~ACB_F_IOPDATA_OVERFLOW;
+				writel(ARCMSR_INBOUND_DRIVER_DATA_READ_OK,
+					&reg->inbound_doorbell);
+			}
+			acb->acb_flags |= ACB_F_MESSAGE_RQBUFFER_CLEARED;
+			acb->rqbuf_firstindex = 0;
+			acb->rqbuf_lastindex = 0;
+			memset(pQbuffer, 0, ARCMSR_MAX_QBUFFER);
+			pcmdmessagefld->cmdmessage.ReturnCode =
+				ARCMSR_MESSAGE_RETURNCODE_OK;
+		}
+		break;
+	case ARCMSR_MESSAGE_CLEAR_WQBUFFER: {
+			uint8_t *pQbuffer = acb->wqbuffer;
+
+			if (acb->acb_flags & ACB_F_IOPDATA_OVERFLOW) {
+				acb->acb_flags &= ~ACB_F_IOPDATA_OVERFLOW;
+				writel(ARCMSR_INBOUND_DRIVER_DATA_READ_OK
+						, &reg->inbound_doorbell);
+			}
+			acb->acb_flags |=
+				(ACB_F_MESSAGE_WQBUFFER_CLEARED |
+					ACB_F_MESSAGE_WQBUFFER_READED);
+			acb->wqbuf_firstindex = 0;
+			acb->wqbuf_lastindex = 0;
+			memset(pQbuffer, 0, ARCMSR_MAX_QBUFFER);
+			pcmdmessagefld->cmdmessage.ReturnCode =
+				ARCMSR_MESSAGE_RETURNCODE_OK;
+		}
+		break;
+	case ARCMSR_MESSAGE_CLEAR_ALLQBUFFER: {
+			uint8_t *pQbuffer;
+
+			if (acb->acb_flags & ACB_F_IOPDATA_OVERFLOW) {
+				acb->acb_flags &= ~ACB_F_IOPDATA_OVERFLOW;
+				writel(ARCMSR_INBOUND_DRIVER_DATA_READ_OK
+						, &reg->inbound_doorbell);
+			}
+			acb->acb_flags |=
+				(ACB_F_MESSAGE_WQBUFFER_CLEARED
+				| ACB_F_MESSAGE_RQBUFFER_CLEARED
+				| ACB_F_MESSAGE_WQBUFFER_READED);
+			acb->rqbuf_firstindex = 0;
+			acb->rqbuf_lastindex = 0;
+			acb->wqbuf_firstindex = 0;
+			acb->wqbuf_lastindex = 0;
+			pQbuffer = acb->rqbuffer;
+			memset(pQbuffer, 0, sizeof (struct QBUFFER));
+			pQbuffer = acb->wqbuffer;
+			memset(pQbuffer, 0, sizeof (struct QBUFFER));
+			pcmdmessagefld->cmdmessage.ReturnCode = ARCMSR_MESSAGE_RETURNCODE_OK;
+		}
+		break;
+	case ARCMSR_MESSAGE_RETURN_CODE_3F: {
+			pcmdmessagefld->cmdmessage.ReturnCode = ARCMSR_MESSAGE_RETURNCODE_3F;
+		}
+		break;
+	case ARCMSR_MESSAGE_SAY_HELLO: {
+			int8_t * hello_string = "Hello! I am ARCMSR";
+
+			memcpy(pcmdmessagefld->messagedatabuffer, hello_string
+				, (int16_t)strlen(hello_string));
+			pcmdmessagefld->cmdmessage.ReturnCode = ARCMSR_MESSAGE_RETURNCODE_OK;
+		}
+		break;
+	case ARCMSR_MESSAGE_SAY_GOODBYE:
+		arcmsr_iop_parking(acb);
+		break;
+	case ARCMSR_MESSAGE_FLUSH_ADAPTER_CACHE:
+		arcmsr_flush_adapter_cache(acb);
+		break;
+	default:
+		retvalue = ARCMSR_MESSAGE_FAIL;
+	}
+ message_out:
+	if (cmd->use_sg) {
+		struct scatterlist *sg;
+
+		sg = (struct scatterlist *) cmd->request_buffer;
+		kunmap_atomic(buffer - sg->offset, KM_IRQ0);
+	}
+	return retvalue;
+}
+
+static struct CommandControlBlock *arcmsr_get_freeccb(struct AdapterControlBlock *acb)
+{
+	struct list_head *head = &acb->ccb_free_list;
+	struct CommandControlBlock *ccb = NULL;
+
+	if (!list_empty(head)) {
+		ccb = list_entry(head->next, struct CommandControlBlock, list);
+		list_del(head->next);
+	}
+	return ccb;
+}
+
+static void arcmsr_handle_virtual_command(struct AdapterControlBlock *acb,
+		struct scsi_cmnd *cmd)
+{
+	switch (cmd->cmnd[0]) {
+	case INQUIRY: {
+		unsigned char inqdata[36];
+		char *buffer;
+
+		if (cmd->device->lun) {
+			cmd->result = (DID_TIME_OUT << 16);
+			cmd->scsi_done(cmd);
+			return;
+		}
+		inqdata[0] = TYPE_PROCESSOR;
+		/* Periph Qualifier & Periph Dev Type */
+		inqdata[1] = 0;
+		/* rem media bit & Dev Type Modifier */
+		inqdata[2] = 0;
+		/* ISO,ECMA,& ANSI versions */
+		inqdata[4] = 31;
+		/* length of additional data */
+		strncpy(&inqdata[8], "Areca   ", 8);
+		/* Vendor Identification */
+		strncpy(&inqdata[16], "RAID controller ", 16);
+		/* Product Identification */
+		strncpy(&inqdata[32], "R001", 4); /* Product Revision */
+		if (cmd->use_sg) {
+			struct scatterlist *sg;
+
+			sg = (struct scatterlist *) cmd->request_buffer;
+			buffer = kmap_atomic(sg->page, KM_IRQ0) + sg->offset;
+		} else {
+			buffer = cmd->request_buffer;
+		}
+		memcpy(buffer, inqdata, sizeof(inqdata));
+		if (cmd->use_sg) {
+			struct scatterlist *sg;
+
+			sg = (struct scatterlist *) cmd->request_buffer;
+			kunmap_atomic(buffer - sg->offset, KM_IRQ0);
+		}
+		cmd->scsi_done(cmd);
+	}
+	break;
+	case WRITE_BUFFER:
+	case READ_BUFFER: {
+		if (arcmsr_iop_message_xfer(acb, cmd))
+			cmd->result = (DID_ERROR << 16);
+		cmd->scsi_done(cmd);
+	}
+	break;
+	default:
+		cmd->scsi_done(cmd);
+	}
+}
+
+static int arcmsr_queue_command(struct scsi_cmnd *cmd,
+	void (* done)(struct scsi_cmnd *))
+{
+	struct Scsi_Host *host = cmd->device->host;
+	struct AdapterControlBlock *acb =
+		(struct AdapterControlBlock *) host->hostdata;
+	struct CommandControlBlock *ccb;
+	int target = cmd->device->id;
+	int lun = cmd->device->lun;
+
+	cmd->scsi_done = done;
+	cmd->host_scribble = NULL;
+	cmd->result = 0;
+	if (acb->acb_flags & ACB_F_BUS_RESET) {
+		printk(KERN_NOTICE "arcmsr%d: bus reset"
+			" and return busy \n"
+			, acb->host->host_no);
+		return SCSI_MLQUEUE_HOST_BUSY;
+	}
+	if(target == 16) {
+		/* virtual device for iop message transfer */
+		arcmsr_handle_virtual_command(acb, cmd);
+		return 0;
+	}
+	if (acb->devstate[target][lun] == ARECA_RAID_GONE) {
+		uint8_t block_cmd;
+
+		block_cmd = cmd->cmnd[0] & 0x0f;
+		if (block_cmd == 0x08 || block_cmd == 0x0a) {
+			printk(KERN_NOTICE
+				"arcmsr%d: block 'read/write'"
+				"command with gone raid volume"
+				" Cmd=%2x, TargetId=%d, Lun=%d \n"
+				, acb->host->host_no
+				, cmd->cmnd[0]
+				, target, lun);
+			cmd->result = (DID_NO_CONNECT << 16);
+			cmd->scsi_done(cmd);
+			return 0;
+		}
+	}
+	if (atomic_read(&acb->ccboutstandingcount) >=
+			ARCMSR_MAX_OUTSTANDING_CMD)
+		return SCSI_MLQUEUE_HOST_BUSY;
+
+	ccb = arcmsr_get_freeccb(acb);
+	if (!ccb)
+		return SCSI_MLQUEUE_HOST_BUSY;
+	arcmsr_build_ccb(acb, ccb, cmd);
+	arcmsr_post_ccb(acb, ccb);
+	return 0;
+}
+
+static void arcmsr_get_firmware_spec(struct AdapterControlBlock *acb)
+{
+	struct MessageUnit __iomem *reg = acb->pmu;
+	char *acb_firm_model = acb->firm_model;
+	char *acb_firm_version = acb->firm_version;
+	char __iomem *iop_firm_model = (char __iomem *) &reg->message_rwbuffer[15];
+	char __iomem *iop_firm_version = (char __iomem *) &reg->message_rwbuffer[17];
+	int count;
+
+	writel(ARCMSR_INBOUND_MESG0_GET_CONFIG, &reg->inbound_msgaddr0);
+	if (arcmsr_wait_msgint_ready(acb))
+		printk(KERN_NOTICE
+			"arcmsr%d: wait "
+			"'get adapter firmware miscellaneous data' timeout \n"
+			, acb->host->host_no);
+	count = 8;
+	while (count) {
+		*acb_firm_model = readb(iop_firm_model);
+		acb_firm_model++;
+		iop_firm_model++;
+		count--;
+	}
+	count = 16;
+	while (count) {
+		*acb_firm_version = readb(iop_firm_version);
+		acb_firm_version++;
+		iop_firm_version++;
+		count--;
+	}
+	printk(KERN_INFO
+		"ARECA RAID ADAPTER%d: FIRMWARE VERSION %s \n"
+		, acb->host->host_no
+		, acb->firm_version);
+	acb->firm_request_len = readl(&reg->message_rwbuffer[1]);
+	acb->firm_numbers_queue = readl(&reg->message_rwbuffer[2]);
+	acb->firm_sdram_size = readl(&reg->message_rwbuffer[3]);
+	acb->firm_hd_channels = readl(&reg->message_rwbuffer[4]);
+}
+
+static void arcmsr_polling_ccbdone(struct AdapterControlBlock *acb,
+	struct CommandControlBlock *poll_ccb)
+{
+	struct MessageUnit __iomem *reg = acb->pmu;
+	struct CommandControlBlock *ccb;
+	uint32_t flag_ccb, outbound_intstatus, poll_ccb_done = 0, poll_count = 0;
+	int id, lun;
+
+ polling_ccb_retry:
+	poll_count++;
+	outbound_intstatus = readl(&reg->outbound_intstatus)
+					& acb->outbound_int_enable;
+	writel(outbound_intstatus, &reg->outbound_intstatus);/*clear interrupt*/
+	while (1) {
+		if ((flag_ccb = readl(&reg->outbound_queueport)) == 0xFFFFFFFF) {
+			if (poll_ccb_done)
+				break;
+			else {
+				msleep(25);
+				if (poll_count > 100)
+					break;
+				goto polling_ccb_retry;
+			}
+		}
+		ccb = (struct CommandControlBlock *)
+			(acb->vir2phy_offset + (flag_ccb << 5));
+		if ((ccb->acb != acb) ||
+			(ccb->startdone != ARCMSR_CCB_START)) {
+			if ((ccb->startdone == ARCMSR_CCB_ABORTED) ||
+				(ccb == poll_ccb)) {
+				printk(KERN_NOTICE
+					"arcmsr%d: scsi id=%d lun=%d ccb='0x%p'"
+					" poll command abort successfully \n"
+					, acb->host->host_no
+					, ccb->pcmd->device->id
+					, ccb->pcmd->device->lun
+					, ccb);
+				ccb->pcmd->result = DID_ABORT << 16;
+				arcmsr_ccb_complete(ccb, 1);
+				poll_ccb_done = 1;
+				continue;
+			}
+			printk(KERN_NOTICE
+				"arcmsr%d: polling get an illegal ccb"
+				" command done ccb='0x%p'"
+				"ccboutstandingcount=%d \n"
+				, acb->host->host_no
+				, ccb
+				, atomic_read(&acb->ccboutstandingcount));
+			continue;
+		}
+		id = ccb->pcmd->device->id;
+		lun = ccb->pcmd->device->lun;
+		if (!(flag_ccb & ARCMSR_CCBREPLY_FLAG_ERROR)) {
+			if (acb->devstate[id][lun] == ARECA_RAID_GONE)
+				acb->devstate[id][lun] = ARECA_RAID_GOOD;
+			ccb->pcmd->result = DID_OK << 16;
+			arcmsr_ccb_complete(ccb, 1);
+		} else {
+			switch(ccb->arcmsr_cdb.DeviceStatus) {
+			case ARCMSR_DEV_SELECT_TIMEOUT: {
+					acb->devstate[id][lun] = ARECA_RAID_GONE;
+					ccb->pcmd->result = DID_TIME_OUT << 16;
+					arcmsr_ccb_complete(ccb, 1);
+				}
+				break;
+			case ARCMSR_DEV_ABORTED:
+			case ARCMSR_DEV_INIT_FAIL: {
+					acb->devstate[id][lun] = ARECA_RAID_GONE;
+					ccb->pcmd->result = DID_BAD_TARGET << 16;
+					arcmsr_ccb_complete(ccb, 1);
+				}
+				break;
+			case ARCMSR_DEV_CHECK_CONDITION: {
+					acb->devstate[id][lun] = ARECA_RAID_GOOD;
+					arcmsr_report_sense_info(ccb);
+					arcmsr_ccb_complete(ccb, 1);
+				}
+				break;
+			default:
+				printk(KERN_NOTICE
+					"arcmsr%d: scsi id=%d lun=%d"
+					" polling and getting command error done"
+					"but got unknown DeviceStatus = 0x%x \n"
+					, acb->host->host_no
+					, id
+					, lun
+					, ccb->arcmsr_cdb.DeviceStatus);
+				acb->devstate[id][lun] = ARECA_RAID_GONE;
+				ccb->pcmd->result = DID_BAD_TARGET << 16;
+				arcmsr_ccb_complete(ccb, 1);
+				break;
+			}
+		}
+	}
+}
+
+static void arcmsr_iop_init(struct AdapterControlBlock *acb)
+{
+	struct MessageUnit __iomem *reg = acb->pmu;
+	uint32_t intmask_org, mask, outbound_doorbell, firmware_state = 0;
+
+	do {
+		firmware_state = readl(&reg->outbound_msgaddr1);
+	} while (!(firmware_state & ARCMSR_OUTBOUND_MESG1_FIRMWARE_OK));
+	intmask_org = readl(&reg->outbound_intmask)
+			| ARCMSR_MU_OUTBOUND_MESSAGE0_INTMASKENABLE;
+	arcmsr_get_firmware_spec(acb);
+
+	acb->acb_flags |= ACB_F_MSG_START_BGRB;
+	writel(ARCMSR_INBOUND_MESG0_START_BGRB, &reg->inbound_msgaddr0);
+	if (arcmsr_wait_msgint_ready(acb)) {
+		printk(KERN_NOTICE "arcmsr%d: "
+			"wait 'start adapter background rebulid' timeout\n",
+			acb->host->host_no);
+	}
+
+	outbound_doorbell = readl(&reg->outbound_doorbell);
+	writel(outbound_doorbell, &reg->outbound_doorbell);
+	writel(ARCMSR_INBOUND_DRIVER_DATA_READ_OK, &reg->inbound_doorbell);
+	mask = ~(ARCMSR_MU_OUTBOUND_POSTQUEUE_INTMASKENABLE
+			| ARCMSR_MU_OUTBOUND_DOORBELL_INTMASKENABLE);
+	writel(intmask_org & mask, &reg->outbound_intmask);
+	acb->outbound_int_enable = ~(intmask_org & mask) & 0x000000ff;
+	acb->acb_flags |= ACB_F_IOP_INITED;
+}
+
+static void arcmsr_iop_reset(struct AdapterControlBlock *acb)
+{
+	struct MessageUnit __iomem *reg = acb->pmu;
+	struct CommandControlBlock *ccb;
+	uint32_t intmask_org;
+	int i = 0;
+
+	if (atomic_read(&acb->ccboutstandingcount) != 0) {
+		/* talk to iop 331 outstanding command aborted */
+		arcmsr_abort_allcmd(acb);
+		/* wait for 3 sec for all command aborted*/
+		msleep_interruptible(3000);
+		/* disable all outbound interrupt */
+		intmask_org = arcmsr_disable_outbound_ints(acb);
+		/* clear all outbound posted Q */
+		for (i = 0; i < ARCMSR_MAX_OUTSTANDING_CMD; i++)
+			readl(&reg->outbound_queueport);
+		for (i = 0; i < ARCMSR_MAX_FREECCB_NUM; i++) {
+			ccb = acb->pccb_pool[i];
+			if ((ccb->startdone == ARCMSR_CCB_START) ||
+				(ccb->startdone == ARCMSR_CCB_ABORTED)) {
+				ccb->startdone = ARCMSR_CCB_ABORTED;
+				ccb->pcmd->result = DID_ABORT << 16;
+				arcmsr_ccb_complete(ccb, 1);
+			}
+		}
+		/* enable all outbound interrupt */
+		arcmsr_enable_outbound_ints(acb, intmask_org);
+	}
+	atomic_set(&acb->ccboutstandingcount, 0);
+}
+
+static int arcmsr_bus_reset(struct scsi_cmnd *cmd)
+{
+	struct AdapterControlBlock *acb =
+		(struct AdapterControlBlock *)cmd->device->host->hostdata;
+	int i;
+
+	acb->num_resets++;
+	acb->acb_flags |= ACB_F_BUS_RESET;
+	for (i = 0; i < 400; i++) {
+		if (!atomic_read(&acb->ccboutstandingcount))
+			break;
+		arcmsr_interrupt(acb);
+		msleep(25);
+	}
+	arcmsr_iop_reset(acb);
+	acb->acb_flags &= ~ACB_F_BUS_RESET;
+	return SUCCESS;
+}
+
+static void arcmsr_abort_one_cmd(struct AdapterControlBlock *acb,
+		struct CommandControlBlock *ccb)
+{
+	u32 intmask;
+
+	ccb->startdone = ARCMSR_CCB_ABORTED;
+
+	/*
+	** Wait for 3 sec for all command done.
+	*/
+	msleep_interruptible(3000);
+
+	intmask = arcmsr_disable_outbound_ints(acb);
+	arcmsr_polling_ccbdone(acb, ccb);
+	arcmsr_enable_outbound_ints(acb, intmask);
+}
+
+static int arcmsr_abort(struct scsi_cmnd *cmd)
+{
+	struct AdapterControlBlock *acb =
+		(struct AdapterControlBlock *)cmd->device->host->hostdata;
+	int i = 0;
+
+	printk(KERN_NOTICE
+		"arcmsr%d: abort device command of scsi id=%d lun=%d \n",
+		acb->host->host_no, cmd->device->id, cmd->device->lun);
+	acb->num_aborts++;
+
+	/*
+	************************************************
+	** the all interrupt service routine is locked
+	** we need to handle it as soon as possible and exit
+	************************************************
+	*/
+	if (!atomic_read(&acb->ccboutstandingcount))
+		return SUCCESS;
+
+	for (i = 0; i < ARCMSR_MAX_FREECCB_NUM; i++) {
+		struct CommandControlBlock *ccb = acb->pccb_pool[i];
+		if (ccb->startdone == ARCMSR_CCB_START && ccb->pcmd == cmd) {
+			arcmsr_abort_one_cmd(acb, ccb);
+			break;
+		}
+	}
+
+	return SUCCESS;
+}
+
+static const char *arcmsr_info(struct Scsi_Host *host)
+{
+	struct AdapterControlBlock *acb =
+		(struct AdapterControlBlock *) host->hostdata;
+	static char buf[256];
+	char *type;
+	int raid6 = 1;
+
+	switch (acb->pdev->device) {
+	case PCI_DEVICE_ID_ARECA_1110:
+	case PCI_DEVICE_ID_ARECA_1210:
+		raid6 = 0;
+		/*FALLTHRU*/
+	case PCI_DEVICE_ID_ARECA_1120:
+	case PCI_DEVICE_ID_ARECA_1130:
+	case PCI_DEVICE_ID_ARECA_1160:
+	case PCI_DEVICE_ID_ARECA_1170:
+	case PCI_DEVICE_ID_ARECA_1220:
+	case PCI_DEVICE_ID_ARECA_1230:
+	case PCI_DEVICE_ID_ARECA_1260:
+	case PCI_DEVICE_ID_ARECA_1270:
+	case PCI_DEVICE_ID_ARECA_1280:
+		type = "SATA";
+		break;
+	case PCI_DEVICE_ID_ARECA_1380:
+	case PCI_DEVICE_ID_ARECA_1381:
+	case PCI_DEVICE_ID_ARECA_1680:
+	case PCI_DEVICE_ID_ARECA_1681:
+		type = "SAS";
+		break;
+	default:
+		type = "X-TYPE";
+		break;
+	}
+	sprintf(buf, "Areca %s Host Adapter RAID Controller%s\n        %s",
+			type, raid6 ? "( RAID6 capable)" : "",
+			ARCMSR_DRIVER_VERSION);
+	return buf;
+}
+
+
diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h
index c09396d2c77b..df7b62676d87 100644
--- a/include/linux/pci_ids.h
+++ b/include/linux/pci_ids.h
@@ -2004,6 +2004,23 @@
 #define PCI_DEVICE_ID_ALTIMA_AC9100	0x03ea
 #define PCI_DEVICE_ID_ALTIMA_AC1003	0x03eb
 
+#define PCI_VENDOR_ID_ARECA		0x17d3
+#define PCI_DEVICE_ID_ARECA_1110	0x1110
+#define PCI_DEVICE_ID_ARECA_1120	0x1120
+#define PCI_DEVICE_ID_ARECA_1130	0x1130
+#define PCI_DEVICE_ID_ARECA_1160	0x1160
+#define PCI_DEVICE_ID_ARECA_1170	0x1170
+#define PCI_DEVICE_ID_ARECA_1210	0x1210
+#define PCI_DEVICE_ID_ARECA_1220	0x1220
+#define PCI_DEVICE_ID_ARECA_1230	0x1230
+#define PCI_DEVICE_ID_ARECA_1260	0x1260
+#define PCI_DEVICE_ID_ARECA_1270	0x1270
+#define PCI_DEVICE_ID_ARECA_1280	0x1280
+#define PCI_DEVICE_ID_ARECA_1380	0x1380
+#define PCI_DEVICE_ID_ARECA_1381	0x1381
+#define PCI_DEVICE_ID_ARECA_1680	0x1680
+#define PCI_DEVICE_ID_ARECA_1681	0x1681
+
 #define PCI_VENDOR_ID_S2IO		0x17d5
 #define	PCI_DEVICE_ID_S2IO_WIN		0x5731
 #define	PCI_DEVICE_ID_S2IO_UNI		0x5831
-- 
cgit v1.2.3


From 3c5100c1c40cc5e27b4da4a736994c76d93392a0 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Wed, 26 Jul 2006 16:58:33 +0900
Subject: [PATCH] libata: cosmetic changes to PM functions

Unify pm_message_t argument to the new-style @mesg.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/scsi/libata-core.c | 12 ++++++------
 drivers/scsi/libata-scsi.c |  8 ++++----
 include/linux/libata.h     |  6 +++---
 3 files changed, 13 insertions(+), 13 deletions(-)

(limited to 'include')

diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c
index 386e5f21e191..66feebdb01c3 100644
--- a/drivers/scsi/libata-core.c
+++ b/drivers/scsi/libata-core.c
@@ -5767,11 +5767,11 @@ int pci_test_config_bits(struct pci_dev *pdev, const struct pci_bits *bits)
 	return (tmp == bits->val) ? 1 : 0;
 }
 
-void ata_pci_device_do_suspend(struct pci_dev *pdev, pm_message_t state)
+void ata_pci_device_do_suspend(struct pci_dev *pdev, pm_message_t mesg)
 {
 	pci_save_state(pdev);
 
-	if (state.event == PM_EVENT_SUSPEND) {
+	if (mesg.event == PM_EVENT_SUSPEND) {
 		pci_disable_device(pdev);
 		pci_set_power_state(pdev, PCI_D3hot);
 	}
@@ -5785,24 +5785,24 @@ void ata_pci_device_do_resume(struct pci_dev *pdev)
 	pci_set_master(pdev);
 }
 
-int ata_pci_device_suspend(struct pci_dev *pdev, pm_message_t state)
+int ata_pci_device_suspend(struct pci_dev *pdev, pm_message_t mesg)
 {
 	struct ata_host_set *host_set = dev_get_drvdata(&pdev->dev);
 	int rc = 0;
 
-	rc = ata_host_set_suspend(host_set, state);
+	rc = ata_host_set_suspend(host_set, mesg);
 	if (rc)
 		return rc;
 
 	if (host_set->next) {
-		rc = ata_host_set_suspend(host_set->next, state);
+		rc = ata_host_set_suspend(host_set->next, mesg);
 		if (rc) {
 			ata_host_set_resume(host_set);
 			return rc;
 		}
 	}
 
-	ata_pci_device_do_suspend(pdev, state);
+	ata_pci_device_do_suspend(pdev, mesg);
 
 	return 0;
 }
diff --git a/drivers/scsi/libata-scsi.c b/drivers/scsi/libata-scsi.c
index 7ced41ecde86..1638e57028cb 100644
--- a/drivers/scsi/libata-scsi.c
+++ b/drivers/scsi/libata-scsi.c
@@ -400,7 +400,7 @@ void ata_dump_status(unsigned id, struct ata_taskfile *tf)
 /**
  *	ata_scsi_device_suspend - suspend ATA device associated with sdev
  *	@sdev: the SCSI device to suspend
- *	@state: target power management state
+ *	@mesg: target power management message
  *
  *	Request suspend EH action on the ATA device associated with
  *	@sdev and wait for the operation to complete.
@@ -411,7 +411,7 @@ void ata_dump_status(unsigned id, struct ata_taskfile *tf)
  *	RETURNS:
  *	0 on success, -errno otherwise.
  */
-int ata_scsi_device_suspend(struct scsi_device *sdev, pm_message_t state)
+int ata_scsi_device_suspend(struct scsi_device *sdev, pm_message_t mesg)
 {
 	struct ata_port *ap = ata_shost_to_port(sdev->host);
 	struct ata_device *dev = ata_scsi_find_dev(ap, sdev);
@@ -438,7 +438,7 @@ int ata_scsi_device_suspend(struct scsi_device *sdev, pm_message_t state)
 
 	/* request suspend */
 	action = ATA_EH_SUSPEND;
-	if (state.event != PM_EVENT_SUSPEND)
+	if (mesg.event != PM_EVENT_SUSPEND)
 		action |= ATA_EH_PM_FREEZE;
 	ap->eh_info.dev_action[dev->devno] |= action;
 	ap->eh_info.flags |= ATA_EHI_QUIET;
@@ -463,7 +463,7 @@ int ata_scsi_device_suspend(struct scsi_device *sdev, pm_message_t state)
 	spin_unlock_irqrestore(ap->lock, flags);
  out:
 	if (rc == 0)
-		sdev->sdev_gendev.power.power_state = state;
+		sdev->sdev_gendev.power.power_state = mesg;
 	return rc;
 }
 
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 66c3100c2b94..b9416708bba2 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -676,9 +676,9 @@ extern void ata_std_ports(struct ata_ioports *ioaddr);
 extern int ata_pci_init_one (struct pci_dev *pdev, struct ata_port_info **port_info,
 			     unsigned int n_ports);
 extern void ata_pci_remove_one (struct pci_dev *pdev);
-extern void ata_pci_device_do_suspend(struct pci_dev *pdev, pm_message_t state);
+extern void ata_pci_device_do_suspend(struct pci_dev *pdev, pm_message_t mesg);
 extern void ata_pci_device_do_resume(struct pci_dev *pdev);
-extern int ata_pci_device_suspend(struct pci_dev *pdev, pm_message_t state);
+extern int ata_pci_device_suspend(struct pci_dev *pdev, pm_message_t mesg);
 extern int ata_pci_device_resume(struct pci_dev *pdev);
 extern int ata_pci_clear_simplex(struct pci_dev *pdev);
 #endif /* CONFIG_PCI */
@@ -697,7 +697,7 @@ extern int sata_scr_write_flush(struct ata_port *ap, int reg, u32 val);
 extern int ata_port_online(struct ata_port *ap);
 extern int ata_port_offline(struct ata_port *ap);
 extern int ata_scsi_device_resume(struct scsi_device *);
-extern int ata_scsi_device_suspend(struct scsi_device *, pm_message_t state);
+extern int ata_scsi_device_suspend(struct scsi_device *, pm_message_t mesg);
 extern int ata_host_set_suspend(struct ata_host_set *host_set,
 				pm_message_t mesg);
 extern void ata_host_set_resume(struct ata_host_set *host_set);
-- 
cgit v1.2.3


From a7f67bdf2c9f24509b8e81e0f35573b611987c80 Mon Sep 17 00:00:00 2001
From: Jeremy Kerr <jk@ozlabs.org>
Date: Wed, 12 Jul 2006 15:35:54 +1000
Subject: [POWERPC] Constify & voidify get_property()

Now that get_property() returns a void *, there's no need to cast its
return value. Also, treat the return value as const, so we can
constify get_property later.

powerpc core changes.

Signed-off-by: Jeremy Kerr <jk@ozlabs.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 arch/powerpc/kernel/btext.c            | 24 ++++++------
 arch/powerpc/kernel/ibmebus.c          |  6 +--
 arch/powerpc/kernel/legacy_serial.c    | 35 ++++++++---------
 arch/powerpc/kernel/lparcfg.c          | 11 +++---
 arch/powerpc/kernel/machine_kexec_64.c | 10 ++---
 arch/powerpc/kernel/pci_32.c           | 39 ++++++++++---------
 arch/powerpc/kernel/pci_64.c           | 28 +++++++-------
 arch/powerpc/kernel/pci_dn.c           | 13 ++++---
 arch/powerpc/kernel/prom.c             | 23 ++++++------
 arch/powerpc/kernel/prom_parse.c       | 68 +++++++++++++++++-----------------
 arch/powerpc/kernel/rtas-proc.c        | 25 +++++++------
 arch/powerpc/kernel/rtas.c             | 28 +++++++-------
 arch/powerpc/kernel/rtas_pci.c         | 22 +++++------
 arch/powerpc/kernel/setup-common.c     | 19 ++++------
 arch/powerpc/kernel/setup_64.c         | 14 +++----
 arch/powerpc/kernel/sysfs.c            |  5 +--
 arch/powerpc/kernel/time.c             |  4 +-
 arch/powerpc/kernel/vio.c              | 16 ++++----
 arch/powerpc/mm/numa.c                 | 31 +++++++---------
 arch/powerpc/sysdev/fsl_soc.c          | 30 +++++++--------
 arch/powerpc/sysdev/mmio_nvram.c       |  4 +-
 include/asm-powerpc/ibmebus.h          |  2 +-
 include/asm-powerpc/prom.h             | 16 ++++----
 include/asm-powerpc/vio.h              |  4 +-
 24 files changed, 239 insertions(+), 238 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/kernel/btext.c b/arch/powerpc/kernel/btext.c
index f4e5e14ee2b6..995fcef156fd 100644
--- a/arch/powerpc/kernel/btext.c
+++ b/arch/powerpc/kernel/btext.c
@@ -158,35 +158,35 @@ int btext_initialize(struct device_node *np)
 {
 	unsigned int width, height, depth, pitch;
 	unsigned long address = 0;
-	u32 *prop;
+	const u32 *prop;
 
-	prop = (u32 *)get_property(np, "linux,bootx-width", NULL);
+	prop = get_property(np, "linux,bootx-width", NULL);
 	if (prop == NULL)
-		prop = (u32 *)get_property(np, "width", NULL);
+		prop = get_property(np, "width", NULL);
 	if (prop == NULL)
 		return -EINVAL;
 	width = *prop;
-	prop = (u32 *)get_property(np, "linux,bootx-height", NULL);
+	prop = get_property(np, "linux,bootx-height", NULL);
 	if (prop == NULL)
-		prop = (u32 *)get_property(np, "height", NULL);
+		prop = get_property(np, "height", NULL);
 	if (prop == NULL)
 		return -EINVAL;
 	height = *prop;
-	prop = (u32 *)get_property(np, "linux,bootx-depth", NULL);
+	prop = get_property(np, "linux,bootx-depth", NULL);
 	if (prop == NULL)
-		prop = (u32 *)get_property(np, "depth", NULL);
+		prop = get_property(np, "depth", NULL);
 	if (prop == NULL)
 		return -EINVAL;
 	depth = *prop;
 	pitch = width * ((depth + 7) / 8);
-	prop = (u32 *)get_property(np, "linux,bootx-linebytes", NULL);
+	prop = get_property(np, "linux,bootx-linebytes", NULL);
 	if (prop == NULL)
-		prop = (u32 *)get_property(np, "linebytes", NULL);
+		prop = get_property(np, "linebytes", NULL);
 	if (prop)
 		pitch = *prop;
 	if (pitch == 1)
 		pitch = 0x1000;
-	prop = (u32 *)get_property(np, "address", NULL);
+	prop = get_property(np, "address", NULL);
 	if (prop)
 		address = *prop;
 
@@ -214,11 +214,11 @@ int btext_initialize(struct device_node *np)
 
 int __init btext_find_display(int allow_nonstdout)
 {
-	char *name;
+	const char *name;
 	struct device_node *np = NULL; 
 	int rc = -ENODEV;
 
-	name = (char *)get_property(of_chosen, "linux,stdout-path", NULL);
+	name = get_property(of_chosen, "linux,stdout-path", NULL);
 	if (name != NULL) {
 		np = of_find_node_by_path(name);
 		if (np != NULL) {
diff --git a/arch/powerpc/kernel/ibmebus.c b/arch/powerpc/kernel/ibmebus.c
index 97ddc02a3d42..d9a0b087fa7f 100644
--- a/arch/powerpc/kernel/ibmebus.c
+++ b/arch/powerpc/kernel/ibmebus.c
@@ -167,7 +167,7 @@ static DEVICE_ATTR(name, S_IRUSR | S_IRGRP | S_IROTH, ibmebusdev_show_name,
 		   NULL);
 
 static struct ibmebus_dev* __devinit ibmebus_register_device_common(
-	struct ibmebus_dev *dev, char *name)
+	struct ibmebus_dev *dev, const char *name)
 {
 	int err = 0;
 
@@ -194,10 +194,10 @@ static struct ibmebus_dev* __devinit ibmebus_register_device_node(
 	struct device_node *dn)
 {
 	struct ibmebus_dev *dev;
-	char *loc_code;
+	const char *loc_code;
 	int length;
 
-	loc_code = (char *)get_property(dn, "ibm,loc-code", NULL);
+	loc_code = get_property(dn, "ibm,loc-code", NULL);
 	if (!loc_code) {
                 printk(KERN_WARNING "%s: node %s missing 'ibm,loc-code'\n",
 		       __FUNCTION__, dn->name ? dn->name : "<unknown>");
diff --git a/arch/powerpc/kernel/legacy_serial.c b/arch/powerpc/kernel/legacy_serial.c
index 359ab89748e0..ee1e0b8c7f1f 100644
--- a/arch/powerpc/kernel/legacy_serial.c
+++ b/arch/powerpc/kernel/legacy_serial.c
@@ -39,16 +39,17 @@ static int __init add_legacy_port(struct device_node *np, int want_index,
 				  phys_addr_t taddr, unsigned long irq,
 				  upf_t flags, int irq_check_parent)
 {
-	u32 *clk, *spd, clock = BASE_BAUD * 16;
+	const u32 *clk, *spd;
+	u32 clock = BASE_BAUD * 16;
 	int index;
 
 	/* get clock freq. if present */
-	clk = (u32 *)get_property(np, "clock-frequency", NULL);
+	clk = get_property(np, "clock-frequency", NULL);
 	if (clk && *clk)
 		clock = *clk;
 
 	/* get default speed if present */
-	spd = (u32 *)get_property(np, "current-speed", NULL);
+	spd = get_property(np, "current-speed", NULL);
 
 	/* If we have a location index, then try to use it */
 	if (want_index >= 0 && want_index < MAX_LEGACY_SERIAL_PORTS)
@@ -113,7 +114,7 @@ static int __init add_legacy_soc_port(struct device_node *np,
 				      struct device_node *soc_dev)
 {
 	u64 addr;
-	u32 *addrp;
+	const u32 *addrp;
 	upf_t flags = UPF_BOOT_AUTOCONF | UPF_SKIP_TEST | UPF_SHARE_IRQ;
 
 	/* We only support ports that have a clock frequency properly
@@ -140,15 +141,15 @@ static int __init add_legacy_soc_port(struct device_node *np,
 static int __init add_legacy_isa_port(struct device_node *np,
 				      struct device_node *isa_brg)
 {
-	u32 *reg;
-	char *typep;
+	const u32 *reg;
+	const char *typep;
 	int index = -1;
 	u64 taddr;
 
 	DBG(" -> add_legacy_isa_port(%s)\n", np->full_name);
 
 	/* Get the ISA port number */
-	reg = (u32 *)get_property(np, "reg", NULL);
+	reg = get_property(np, "reg", NULL);
 	if (reg == NULL)
 		return -1;
 
@@ -159,7 +160,7 @@ static int __init add_legacy_isa_port(struct device_node *np,
 	/* Now look for an "ibm,aix-loc" property that gives us ordering
 	 * if any...
 	 */
-	typep = (char *)get_property(np, "ibm,aix-loc", NULL);
+	typep = get_property(np, "ibm,aix-loc", NULL);
 
 	/* If we have a location index, then use it */
 	if (typep && *typep == 'S')
@@ -184,7 +185,7 @@ static int __init add_legacy_pci_port(struct device_node *np,
 				      struct device_node *pci_dev)
 {
 	u64 addr, base;
-	u32 *addrp;
+	const u32 *addrp;
 	unsigned int flags;
 	int iotype, index = -1, lindex = 0;
 
@@ -223,7 +224,7 @@ static int __init add_legacy_pci_port(struct device_node *np,
 	 * we get to their "reg" property
 	 */
 	if (np != pci_dev) {
-		u32 *reg = (u32 *)get_property(np, "reg", NULL);
+		const u32 *reg = get_property(np, "reg", NULL);
 		if (reg && (*reg < 4))
 			index = lindex = *reg;
 	}
@@ -281,13 +282,13 @@ static void __init setup_legacy_serial_console(int console)
 void __init find_legacy_serial_ports(void)
 {
 	struct device_node *np, *stdout = NULL;
-	char *path;
+	const char *path;
 	int index;
 
 	DBG(" -> find_legacy_serial_port()\n");
 
 	/* Now find out if one of these is out firmware console */
-	path = (char *)get_property(of_chosen, "linux,stdout-path", NULL);
+	path = get_property(of_chosen, "linux,stdout-path", NULL);
 	if (path != NULL) {
 		stdout = of_find_node_by_path(path);
 		if (stdout)
@@ -487,8 +488,8 @@ static int __init check_legacy_serial_console(void)
 {
 	struct device_node *prom_stdout = NULL;
 	int speed = 0, offset = 0;
-	char *name;
-	u32 *spd;
+	const char *name;
+	const u32 *spd;
 
 	DBG(" -> check_legacy_serial_console()\n");
 
@@ -509,7 +510,7 @@ static int __init check_legacy_serial_console(void)
 	}
 	/* We are getting a weird phandle from OF ... */
 	/* ... So use the full path instead */
-	name = (char *)get_property(of_chosen, "linux,stdout-path", NULL);
+	name = get_property(of_chosen, "linux,stdout-path", NULL);
 	if (name == NULL) {
 		DBG(" no linux,stdout-path !\n");
 		return -ENODEV;
@@ -521,12 +522,12 @@ static int __init check_legacy_serial_console(void)
 	}
 	DBG("stdout is %s\n", prom_stdout->full_name);
 
-	name = (char *)get_property(prom_stdout, "name", NULL);
+	name = get_property(prom_stdout, "name", NULL);
 	if (!name) {
 		DBG(" stdout package has no name !\n");
 		goto not_found;
 	}
-	spd = (u32 *)get_property(prom_stdout, "current-speed", NULL);
+	spd = get_property(prom_stdout, "current-speed", NULL);
 	if (spd)
 		speed = *spd;
 
diff --git a/arch/powerpc/kernel/lparcfg.c b/arch/powerpc/kernel/lparcfg.c
index 2d94b372d49b..3ce3a2d56fa8 100644
--- a/arch/powerpc/kernel/lparcfg.c
+++ b/arch/powerpc/kernel/lparcfg.c
@@ -309,12 +309,11 @@ static int pseries_lparcfg_data(struct seq_file *m, void *v)
 	int partition_potential_processors;
 	int partition_active_processors;
 	struct device_node *rtas_node;
-	int *lrdrp = NULL;
+	const int *lrdrp = NULL;
 
 	rtas_node = find_path_device("/rtas");
 	if (rtas_node)
-		lrdrp = (int *)get_property(rtas_node, "ibm,lrdr-capacity",
-		                            NULL);
+		lrdrp = get_property(rtas_node, "ibm,lrdr-capacity", NULL);
 
 	if (lrdrp == NULL) {
 		partition_potential_processors = vdso_data->processorCount;
@@ -519,7 +518,8 @@ static int lparcfg_data(struct seq_file *m, void *v)
 	const char *model = "";
 	const char *system_id = "";
 	const char *tmp;
-	unsigned int *lp_index_ptr, lp_index = 0;
+	const unsigned int *lp_index_ptr;
+	unsigned int lp_index = 0;
 
 	seq_printf(m, "%s %s \n", MODULE_NAME, MODULE_VERS);
 
@@ -539,8 +539,7 @@ static int lparcfg_data(struct seq_file *m, void *v)
 			if (firmware_has_feature(FW_FEATURE_ISERIES))
 				system_id += 4;
 		}
-		lp_index_ptr = (unsigned int *)
-			get_property(rootdn, "ibm,partition-no", NULL);
+		lp_index_ptr = get_property(rootdn, "ibm,partition-no", NULL);
 		if (lp_index_ptr)
 			lp_index = *lp_index_ptr;
 	}
diff --git a/arch/powerpc/kernel/machine_kexec_64.c b/arch/powerpc/kernel/machine_kexec_64.c
index b438d45a068c..4efdaa9d3f43 100644
--- a/arch/powerpc/kernel/machine_kexec_64.c
+++ b/arch/powerpc/kernel/machine_kexec_64.c
@@ -33,8 +33,8 @@ int default_machine_kexec_prepare(struct kimage *image)
 	unsigned long begin, end;	/* limits of segment */
 	unsigned long low, high;	/* limits of blocked memory range */
 	struct device_node *node;
-	unsigned long *basep;
-	unsigned int *sizep;
+	const unsigned long *basep;
+	const unsigned int *sizep;
 
 	if (!ppc_md.hpte_clear_all)
 		return -ENOENT;
@@ -74,10 +74,8 @@ int default_machine_kexec_prepare(struct kimage *image)
 	/* We also should not overwrite the tce tables */
 	for (node = of_find_node_by_type(NULL, "pci"); node != NULL;
 			node = of_find_node_by_type(node, "pci")) {
-		basep = (unsigned long *)get_property(node, "linux,tce-base",
-							NULL);
-		sizep = (unsigned int *)get_property(node, "linux,tce-size",
-							NULL);
+		basep = get_property(node, "linux,tce-base", NULL);
+		sizep = get_property(node, "linux,tce-size", NULL);
 		if (basep == NULL || sizep == NULL)
 			continue;
 
diff --git a/arch/powerpc/kernel/pci_32.c b/arch/powerpc/kernel/pci_32.c
index 898dae8ab6d9..3f6bd36e9e14 100644
--- a/arch/powerpc/kernel/pci_32.c
+++ b/arch/powerpc/kernel/pci_32.c
@@ -633,12 +633,12 @@ pcibios_alloc_controller(void)
 static void
 make_one_node_map(struct device_node* node, u8 pci_bus)
 {
-	int *bus_range;
+	const int *bus_range;
 	int len;
 
 	if (pci_bus >= pci_bus_count)
 		return;
-	bus_range = (int *) get_property(node, "bus-range", &len);
+	bus_range = get_property(node, "bus-range", &len);
 	if (bus_range == NULL || len < 2 * sizeof(int)) {
 		printk(KERN_WARNING "Can't get bus-range for %s, "
 		       "assuming it starts at 0\n", node->full_name);
@@ -648,13 +648,13 @@ make_one_node_map(struct device_node* node, u8 pci_bus)
 
 	for (node=node->child; node != 0;node = node->sibling) {
 		struct pci_dev* dev;
-		unsigned int *class_code, *reg;
+		const unsigned int *class_code, *reg;
 	
-		class_code = (unsigned int *) get_property(node, "class-code", NULL);
+		class_code = get_property(node, "class-code", NULL);
 		if (!class_code || ((*class_code >> 8) != PCI_CLASS_BRIDGE_PCI &&
 			(*class_code >> 8) != PCI_CLASS_BRIDGE_CARDBUS))
 			continue;
-		reg = (unsigned int *)get_property(node, "reg", NULL);
+		reg = get_property(node, "reg", NULL);
 		if (!reg)
 			continue;
 		dev = pci_find_slot(pci_bus, ((reg[0] >> 8) & 0xff));
@@ -669,7 +669,7 @@ pcibios_make_OF_bus_map(void)
 {
 	int i;
 	struct pci_controller* hose;
-	u8* of_prop_map;
+	struct property *map_prop;
 
 	pci_to_OF_bus_map = (u8*)kmalloc(pci_bus_count, GFP_KERNEL);
 	if (!pci_to_OF_bus_map) {
@@ -691,9 +691,12 @@ pcibios_make_OF_bus_map(void)
 			continue;
 		make_one_node_map(node, hose->first_busno);
 	}
-	of_prop_map = get_property(find_path_device("/"), "pci-OF-bus-map", NULL);
-	if (of_prop_map)
-		memcpy(of_prop_map, pci_to_OF_bus_map, pci_bus_count);
+	map_prop = of_find_property(find_path_device("/"),
+			"pci-OF-bus-map", NULL);
+	if (map_prop) {
+		BUG_ON(pci_bus_count > map_prop->length);
+		memcpy(map_prop->value, pci_to_OF_bus_map, pci_bus_count);
+	}
 #ifdef DEBUG
 	printk("PCI->OF bus map:\n");
 	for (i=0; i<pci_bus_count; i++) {
@@ -712,7 +715,7 @@ scan_OF_pci_childs(struct device_node* node, pci_OF_scan_iterator filter, void*
 	struct device_node* sub_node;
 
 	for (; node != 0;node = node->sibling) {
-		unsigned int *class_code;
+		const unsigned int *class_code;
 	
 		if (filter(node, data))
 			return node;
@@ -722,7 +725,7 @@ scan_OF_pci_childs(struct device_node* node, pci_OF_scan_iterator filter, void*
 		 * a fake root for all functions of a multi-function device,
 		 * we go down them as well.
 		 */
-		class_code = (unsigned int *) get_property(node, "class-code", NULL);
+		class_code = get_property(node, "class-code", NULL);
 		if ((!class_code || ((*class_code >> 8) != PCI_CLASS_BRIDGE_PCI &&
 			(*class_code >> 8) != PCI_CLASS_BRIDGE_CARDBUS)) &&
 			strcmp(node->name, "multifunc-device"))
@@ -737,10 +740,10 @@ scan_OF_pci_childs(struct device_node* node, pci_OF_scan_iterator filter, void*
 static int
 scan_OF_pci_childs_iterator(struct device_node* node, void* data)
 {
-	unsigned int *reg;
+	const unsigned int *reg;
 	u8* fdata = (u8*)data;
 	
-	reg = (unsigned int *) get_property(node, "reg", NULL);
+	reg = get_property(node, "reg", NULL);
 	if (reg && ((reg[0] >> 8) & 0xff) == fdata[1]
 		&& ((reg[0] >> 16) & 0xff) == fdata[0])
 		return 1;
@@ -841,7 +844,7 @@ find_OF_pci_device_filter(struct device_node* node, void* data)
 int
 pci_device_from_OF_node(struct device_node* node, u8* bus, u8* devfn)
 {
-	unsigned int *reg;
+	const unsigned int *reg;
 	struct pci_controller* hose;
 	struct pci_dev* dev = NULL;
 	
@@ -854,7 +857,7 @@ pci_device_from_OF_node(struct device_node* node, u8* bus, u8* devfn)
 	if (!scan_OF_pci_childs(((struct device_node*)hose->arch_data)->child,
 			find_OF_pci_device_filter, (void *)node))
 		return -ENODEV;
-	reg = (unsigned int *) get_property(node, "reg", NULL);
+	reg = get_property(node, "reg", NULL);
 	if (!reg)
 		return -ENODEV;
 	*bus = (reg[0] >> 16) & 0xff;
@@ -885,8 +888,8 @@ pci_process_bridge_OF_ranges(struct pci_controller *hose,
 			   struct device_node *dev, int primary)
 {
 	static unsigned int static_lc_ranges[256] __initdata;
-	unsigned int *dt_ranges, *lc_ranges, *ranges, *prev;
-	unsigned int size;
+	const unsigned int *dt_ranges;
+	unsigned int *lc_ranges, *ranges, *prev, size;
 	int rlen = 0, orig_rlen;
 	int memno = 0;
 	struct resource *res;
@@ -897,7 +900,7 @@ pci_process_bridge_OF_ranges(struct pci_controller *hose,
 	 * that can have more than 3 ranges, fortunately using contiguous
 	 * addresses -- BenH
 	 */
-	dt_ranges = (unsigned int *) get_property(dev, "ranges", &rlen);
+	dt_ranges = get_property(dev, "ranges", &rlen);
 	if (!dt_ranges)
 		return;
 	/* Sanity check, though hopefully that never happens */
diff --git a/arch/powerpc/kernel/pci_64.c b/arch/powerpc/kernel/pci_64.c
index 1d85fcba51e4..e795a7e2a38e 100644
--- a/arch/powerpc/kernel/pci_64.c
+++ b/arch/powerpc/kernel/pci_64.c
@@ -246,10 +246,10 @@ static void __init pcibios_claim_of_setup(void)
 #ifdef CONFIG_PPC_MULTIPLATFORM
 static u32 get_int_prop(struct device_node *np, const char *name, u32 def)
 {
-	u32 *prop;
+	const u32 *prop;
 	int len;
 
-	prop = (u32 *) get_property(np, name, &len);
+	prop = get_property(np, name, &len);
 	if (prop && len >= 4)
 		return *prop;
 	return def;
@@ -278,10 +278,11 @@ static void pci_parse_of_addrs(struct device_node *node, struct pci_dev *dev)
 	u64 base, size;
 	unsigned int flags;
 	struct resource *res;
-	u32 *addrs, i;
+	const u32 *addrs;
+	u32 i;
 	int proplen;
 
-	addrs = (u32 *) get_property(node, "assigned-addresses", &proplen);
+	addrs = get_property(node, "assigned-addresses", &proplen);
 	if (!addrs)
 		return;
 	DBG("    parse addresses (%d bytes) @ %p\n", proplen, addrs);
@@ -381,7 +382,7 @@ void __devinit of_scan_bus(struct device_node *node,
 				  struct pci_bus *bus)
 {
 	struct device_node *child = NULL;
-	u32 *reg;
+	const u32 *reg;
 	int reglen, devfn;
 	struct pci_dev *dev;
 
@@ -389,7 +390,7 @@ void __devinit of_scan_bus(struct device_node *node,
 
 	while ((child = of_get_next_child(node, child)) != NULL) {
 		DBG("  * %s\n", child->full_name);
-		reg = (u32 *) get_property(child, "reg", &reglen);
+		reg = get_property(child, "reg", &reglen);
 		if (reg == NULL || reglen < 20)
 			continue;
 		devfn = (reg[0] >> 8) & 0xff;
@@ -413,7 +414,7 @@ void __devinit of_scan_pci_bridge(struct device_node *node,
 			 	struct pci_dev *dev)
 {
 	struct pci_bus *bus;
-	u32 *busrange, *ranges;
+	const u32 *busrange, *ranges;
 	int len, i, mode;
 	struct resource *res;
 	unsigned int flags;
@@ -422,13 +423,13 @@ void __devinit of_scan_pci_bridge(struct device_node *node,
 	DBG("of_scan_pci_bridge(%s)\n", node->full_name);
 
 	/* parse bus-range property */
-	busrange = (u32 *) get_property(node, "bus-range", &len);
+	busrange = get_property(node, "bus-range", &len);
 	if (busrange == NULL || len != 8) {
 		printk(KERN_DEBUG "Can't get bus-range for PCI-PCI bridge %s\n",
 		       node->full_name);
 		return;
 	}
-	ranges = (u32 *) get_property(node, "ranges", &len);
+	ranges = get_property(node, "ranges", &len);
 	if (ranges == NULL) {
 		printk(KERN_DEBUG "Can't get ranges for PCI-PCI bridge %s\n",
 		       node->full_name);
@@ -892,13 +893,13 @@ static void __devinit pci_process_ISA_OF_ranges(struct device_node *isa_node,
 		unsigned int size;
 	};
 
-	struct isa_range *range;
+	const struct isa_range *range;
 	unsigned long pci_addr;
 	unsigned int isa_addr;
 	unsigned int size;
 	int rlen = 0;
 
-	range = (struct isa_range *) get_property(isa_node, "ranges", &rlen);
+	range = get_property(isa_node, "ranges", &rlen);
 	if (range == NULL || (rlen < sizeof(struct isa_range))) {
 		printk(KERN_ERR "no ISA ranges or unexpected isa range size,"
 		       "mapping 64k\n");
@@ -939,7 +940,8 @@ static void __devinit pci_process_ISA_OF_ranges(struct device_node *isa_node,
 void __devinit pci_process_bridge_OF_ranges(struct pci_controller *hose,
 					    struct device_node *dev, int prim)
 {
-	unsigned int *ranges, pci_space;
+	const unsigned int *ranges;
+	unsigned int pci_space;
 	unsigned long size;
 	int rlen = 0;
 	int memno = 0;
@@ -957,7 +959,7 @@ void __devinit pci_process_bridge_OF_ranges(struct pci_controller *hose,
 	 *			(size depending on dev->n_addr_cells)
 	 *   cells 4+5 or 5+6:	the size of the range
 	 */
-	ranges = (unsigned int *) get_property(dev, "ranges", &rlen);
+	ranges = get_property(dev, "ranges", &rlen);
 	if (ranges == NULL)
 		return;
 	hose->io_base_phys = 0;
diff --git a/arch/powerpc/kernel/pci_dn.c b/arch/powerpc/kernel/pci_dn.c
index 1c18953514c3..68df018dae0e 100644
--- a/arch/powerpc/kernel/pci_dn.c
+++ b/arch/powerpc/kernel/pci_dn.c
@@ -40,8 +40,8 @@
 static void * __devinit update_dn_pci_info(struct device_node *dn, void *data)
 {
 	struct pci_controller *phb = data;
-	int *type = (int *)get_property(dn, "ibm,pci-config-space-type", NULL);
-	u32 *regs;
+	const int *type = get_property(dn, "ibm,pci-config-space-type", NULL);
+	const u32 *regs;
 	struct pci_dn *pdn;
 
 	if (mem_init_done)
@@ -54,14 +54,14 @@ static void * __devinit update_dn_pci_info(struct device_node *dn, void *data)
 	dn->data = pdn;
 	pdn->node = dn;
 	pdn->phb = phb;
-	regs = (u32 *)get_property(dn, "reg", NULL);
+	regs = get_property(dn, "reg", NULL);
 	if (regs) {
 		/* First register entry is addr (00BBSS00)  */
 		pdn->busno = (regs[0] >> 16) & 0xff;
 		pdn->devfn = (regs[0] >> 8) & 0xff;
 	}
 	if (firmware_has_feature(FW_FEATURE_ISERIES)) {
-		u32 *busp = (u32 *)get_property(dn, "linux,subbus", NULL);
+		const u32 *busp = get_property(dn, "linux,subbus", NULL);
 		if (busp)
 			pdn->bussubno = *busp;
 	}
@@ -96,10 +96,11 @@ void *traverse_pci_devices(struct device_node *start, traverse_func pre,
 
 	/* We started with a phb, iterate all childs */
 	for (dn = start->child; dn; dn = nextdn) {
-		u32 *classp, class;
+		const u32 *classp;
+		u32 class;
 
 		nextdn = NULL;
-		classp = (u32 *)get_property(dn, "class-code", NULL);
+		classp = get_property(dn, "class-code", NULL);
 		class = classp ? *classp : 0;
 
 		if (pre && ((ret = pre(dn, data)) != NULL))
diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c
index a1787ffb6319..2a3d84a39cb5 100644
--- a/arch/powerpc/kernel/prom.c
+++ b/arch/powerpc/kernel/prom.c
@@ -942,11 +942,11 @@ void __init early_init_devtree(void *params)
 int
 prom_n_addr_cells(struct device_node* np)
 {
-	int* ip;
+	const int *ip;
 	do {
 		if (np->parent)
 			np = np->parent;
-		ip = (int *) get_property(np, "#address-cells", NULL);
+		ip = get_property(np, "#address-cells", NULL);
 		if (ip != NULL)
 			return *ip;
 	} while (np->parent);
@@ -958,11 +958,11 @@ EXPORT_SYMBOL(prom_n_addr_cells);
 int
 prom_n_size_cells(struct device_node* np)
 {
-	int* ip;
+	const int* ip;
 	do {
 		if (np->parent)
 			np = np->parent;
-		ip = (int *) get_property(np, "#size-cells", NULL);
+		ip = get_property(np, "#size-cells", NULL);
 		if (ip != NULL)
 			return *ip;
 	} while (np->parent);
@@ -1034,7 +1034,7 @@ int device_is_compatible(struct device_node *device, const char *compat)
 	const char* cp;
 	int cplen, l;
 
-	cp = (char *) get_property(device, "compatible", &cplen);
+	cp = get_property(device, "compatible", &cplen);
 	if (cp == NULL)
 		return 0;
 	while (cplen > 0) {
@@ -1449,7 +1449,7 @@ static int of_finish_dynamic_node(struct device_node *node)
 {
 	struct device_node *parent = of_get_parent(node);
 	int err = 0;
-	phandle *ibm_phandle;
+	const phandle *ibm_phandle;
 
 	node->name = get_property(node, "name", NULL);
 	node->type = get_property(node, "device_type", NULL);
@@ -1466,8 +1466,7 @@ static int of_finish_dynamic_node(struct device_node *node)
 		return -ENODEV;
 
 	/* fix up new node's linux_phandle field */
-	if ((ibm_phandle = (unsigned int *)get_property(node,
-							"ibm,phandle", NULL)))
+	if ((ibm_phandle = get_property(node, "ibm,phandle", NULL)))
 		node->linux_phandle = *ibm_phandle;
 
 out:
@@ -1658,16 +1657,16 @@ struct device_node *of_get_cpu_node(int cpu, unsigned int *thread)
 	hardid = get_hard_smp_processor_id(cpu);
 
 	for_each_node_by_type(np, "cpu") {
-		u32 *intserv;
+		const u32 *intserv;
 		unsigned int plen, t;
 
 		/* Check for ibm,ppc-interrupt-server#s. If it doesn't exist
 		 * fallback to "reg" property and assume no threads
 		 */
-		intserv = (u32 *)get_property(np, "ibm,ppc-interrupt-server#s",
-					      &plen);
+		intserv = get_property(np, "ibm,ppc-interrupt-server#s",
+				&plen);
 		if (intserv == NULL) {
-			u32 *reg = (u32 *)get_property(np, "reg", NULL);
+			const u32 *reg = get_property(np, "reg", NULL);
 			if (reg == NULL)
 				continue;
 			if (*reg == hardid) {
diff --git a/arch/powerpc/kernel/prom_parse.c b/arch/powerpc/kernel/prom_parse.c
index e9960170667b..cdcd5d665468 100644
--- a/arch/powerpc/kernel/prom_parse.c
+++ b/arch/powerpc/kernel/prom_parse.c
@@ -270,7 +270,7 @@ static int of_translate_one(struct device_node *parent, struct of_bus *bus,
 			    struct of_bus *pbus, u32 *addr,
 			    int na, int ns, int pna)
 {
-	u32 *ranges;
+	const u32 *ranges;
 	unsigned int rlen;
 	int rone;
 	u64 offset = OF_BAD_ADDR;
@@ -287,7 +287,7 @@ static int of_translate_one(struct device_node *parent, struct of_bus *bus,
 	 * to translate addresses that aren't supposed to be translated in
 	 * the first place. --BenH.
 	 */
-	ranges = (u32 *)get_property(parent, "ranges", &rlen);
+	ranges = get_property(parent, "ranges", &rlen);
 	if (ranges == NULL || rlen == 0) {
 		offset = of_read_number(addr, na);
 		memset(addr, 0, pna * 4);
@@ -330,7 +330,7 @@ static int of_translate_one(struct device_node *parent, struct of_bus *bus,
  * that can be mapped to a cpu physical address). This is not really specified
  * that way, but this is traditionally the way IBM at least do things
  */
-u64 of_translate_address(struct device_node *dev, u32 *in_addr)
+u64 of_translate_address(struct device_node *dev, const u32 *in_addr)
 {
 	struct device_node *parent = NULL;
 	struct of_bus *bus, *pbus;
@@ -407,10 +407,10 @@ u64 of_translate_address(struct device_node *dev, u32 *in_addr)
 }
 EXPORT_SYMBOL(of_translate_address);
 
-u32 *of_get_address(struct device_node *dev, int index, u64 *size,
+const u32 *of_get_address(struct device_node *dev, int index, u64 *size,
 		    unsigned int *flags)
 {
-	u32 *prop;
+	const u32 *prop;
 	unsigned int psize;
 	struct device_node *parent;
 	struct of_bus *bus;
@@ -427,7 +427,7 @@ u32 *of_get_address(struct device_node *dev, int index, u64 *size,
 		return NULL;
 
 	/* Get "reg" or "assigned-addresses" property */
-	prop = (u32 *)get_property(dev, bus->addresses, &psize);
+	prop = get_property(dev, bus->addresses, &psize);
 	if (prop == NULL)
 		return NULL;
 	psize /= 4;
@@ -445,10 +445,10 @@ u32 *of_get_address(struct device_node *dev, int index, u64 *size,
 }
 EXPORT_SYMBOL(of_get_address);
 
-u32 *of_get_pci_address(struct device_node *dev, int bar_no, u64 *size,
+const u32 *of_get_pci_address(struct device_node *dev, int bar_no, u64 *size,
 			unsigned int *flags)
 {
-	u32 *prop;
+	const u32 *prop;
 	unsigned int psize;
 	struct device_node *parent;
 	struct of_bus *bus;
@@ -469,7 +469,7 @@ u32 *of_get_pci_address(struct device_node *dev, int bar_no, u64 *size,
 		return NULL;
 
 	/* Get "reg" or "assigned-addresses" property */
-	prop = (u32 *)get_property(dev, bus->addresses, &psize);
+	prop = get_property(dev, bus->addresses, &psize);
 	if (prop == NULL)
 		return NULL;
 	psize /= 4;
@@ -487,7 +487,7 @@ u32 *of_get_pci_address(struct device_node *dev, int bar_no, u64 *size,
 }
 EXPORT_SYMBOL(of_get_pci_address);
 
-static int __of_address_to_resource(struct device_node *dev, u32 *addrp,
+static int __of_address_to_resource(struct device_node *dev, const u32 *addrp,
 				    u64 size, unsigned int flags,
 				    struct resource *r)
 {
@@ -518,7 +518,7 @@ static int __of_address_to_resource(struct device_node *dev, u32 *addrp,
 int of_address_to_resource(struct device_node *dev, int index,
 			   struct resource *r)
 {
-	u32		*addrp;
+	const u32	*addrp;
 	u64		size;
 	unsigned int	flags;
 
@@ -532,7 +532,7 @@ EXPORT_SYMBOL_GPL(of_address_to_resource);
 int of_pci_address_to_resource(struct device_node *dev, int bar,
 			       struct resource *r)
 {
-	u32		*addrp;
+	const u32	*addrp;
 	u64		size;
 	unsigned int	flags;
 
@@ -543,13 +543,14 @@ int of_pci_address_to_resource(struct device_node *dev, int bar,
 }
 EXPORT_SYMBOL_GPL(of_pci_address_to_resource);
 
-void of_parse_dma_window(struct device_node *dn, unsigned char *dma_window_prop,
+void of_parse_dma_window(struct device_node *dn, const void *dma_window_prop,
 		unsigned long *busno, unsigned long *phys, unsigned long *size)
 {
-	u32 *dma_window, cells;
-	unsigned char *prop;
+	const u32 *dma_window;
+	u32 cells;
+	const unsigned char *prop;
 
-	dma_window = (u32 *)dma_window_prop;
+	dma_window = dma_window_prop;
 
 	/* busno is always one cell */
 	*busno = *(dma_window++);
@@ -578,13 +579,13 @@ static struct device_node *of_irq_dflt_pic;
 static struct device_node *of_irq_find_parent(struct device_node *child)
 {
 	struct device_node *p;
-	phandle *parp;
+	const phandle *parp;
 
 	if (!of_node_get(child))
 		return NULL;
 
 	do {
-		parp = (phandle *)get_property(child, "interrupt-parent", NULL);
+		parp = get_property(child, "interrupt-parent", NULL);
 		if (parp == NULL)
 			p = of_get_parent(child);
 		else {
@@ -646,11 +647,11 @@ void of_irq_map_init(unsigned int flags)
 
 }
 
-int of_irq_map_raw(struct device_node *parent, u32 *intspec, u32 *addr,
-		   struct of_irq *out_irq)
+int of_irq_map_raw(struct device_node *parent, const u32 *intspec,
+		const u32 *addr, struct of_irq *out_irq)
 {
 	struct device_node *ipar, *tnode, *old = NULL, *newpar = NULL;
-	u32 *tmp, *imap, *imask;
+	const u32 *tmp, *imap, *imask;
 	u32 intsize = 1, addrsize, newintsize = 0, newaddrsize = 0;
 	int imaplen, match, i;
 
@@ -661,7 +662,7 @@ int of_irq_map_raw(struct device_node *parent, u32 *intspec, u32 *addr,
 	 * is none, we are nice and just walk up the tree
 	 */
 	do {
-		tmp = (u32 *)get_property(ipar, "#interrupt-cells", NULL);
+		tmp = get_property(ipar, "#interrupt-cells", NULL);
 		if (tmp != NULL) {
 			intsize = *tmp;
 			break;
@@ -682,7 +683,7 @@ int of_irq_map_raw(struct device_node *parent, u32 *intspec, u32 *addr,
 	 */
 	old = of_node_get(ipar);
 	do {
-		tmp = (u32 *)get_property(old, "#address-cells", NULL);
+		tmp = get_property(old, "#address-cells", NULL);
 		tnode = of_get_parent(old);
 		of_node_put(old);
 		old = tnode;
@@ -709,7 +710,7 @@ int of_irq_map_raw(struct device_node *parent, u32 *intspec, u32 *addr,
 		}
 
 		/* Now look for an interrupt-map */
-		imap = (u32 *)get_property(ipar, "interrupt-map", &imaplen);
+		imap = get_property(ipar, "interrupt-map", &imaplen);
 		/* No interrupt map, check for an interrupt parent */
 		if (imap == NULL) {
 			DBG(" -> no map, getting parent\n");
@@ -719,7 +720,7 @@ int of_irq_map_raw(struct device_node *parent, u32 *intspec, u32 *addr,
 		imaplen /= sizeof(u32);
 
 		/* Look for a mask */
-		imask = (u32 *)get_property(ipar, "interrupt-map-mask", NULL);
+		imask = get_property(ipar, "interrupt-map-mask", NULL);
 
 		/* If we were passed no "reg" property and we attempt to parse
 		 * an interrupt-map, then #address-cells must be 0.
@@ -766,14 +767,14 @@ int of_irq_map_raw(struct device_node *parent, u32 *intspec, u32 *addr,
 			/* Get #interrupt-cells and #address-cells of new
 			 * parent
 			 */
-			tmp = (u32 *)get_property(newpar, "#interrupt-cells",
+			tmp = get_property(newpar, "#interrupt-cells",
 						  NULL);
 			if (tmp == NULL) {
 				DBG(" -> parent lacks #interrupt-cells !\n");
 				goto fail;
 			}
 			newintsize = *tmp;
-			tmp = (u32 *)get_property(newpar, "#address-cells",
+			tmp = get_property(newpar, "#address-cells",
 						  NULL);
 			newaddrsize = (tmp == NULL) ? 0 : *tmp;
 
@@ -819,14 +820,14 @@ EXPORT_SYMBOL_GPL(of_irq_map_raw);
 static int of_irq_map_oldworld(struct device_node *device, int index,
 			       struct of_irq *out_irq)
 {
-	u32 *ints;
+	const u32 *ints;
 	int intlen;
 
 	/*
 	 * Old machines just have a list of interrupt numbers
 	 * and no interrupt-controller nodes.
 	 */
-	ints = (u32 *) get_property(device, "AAPL,interrupts", &intlen);
+	ints = get_property(device, "AAPL,interrupts", &intlen);
 	if (ints == NULL)
 		return -EINVAL;
 	intlen /= sizeof(u32);
@@ -851,7 +852,8 @@ static int of_irq_map_oldworld(struct device_node *device, int index,
 int of_irq_map_one(struct device_node *device, int index, struct of_irq *out_irq)
 {
 	struct device_node *p;
-	u32 *intspec, *tmp, intsize, intlen, *addr;
+	const u32 *intspec, *tmp, *addr;
+	u32 intsize, intlen;
 	int res;
 
 	DBG("of_irq_map_one: dev=%s, index=%d\n", device->full_name, index);
@@ -861,13 +863,13 @@ int of_irq_map_one(struct device_node *device, int index, struct of_irq *out_irq
 		return of_irq_map_oldworld(device, index, out_irq);
 
 	/* Get the interrupts property */
-	intspec = (u32 *)get_property(device, "interrupts", &intlen);
+	intspec = get_property(device, "interrupts", &intlen);
 	if (intspec == NULL)
 		return -EINVAL;
 	intlen /= sizeof(u32);
 
 	/* Get the reg property (if any) */
-	addr = (u32 *)get_property(device, "reg", NULL);
+	addr = get_property(device, "reg", NULL);
 
 	/* Look for the interrupt parent. */
 	p = of_irq_find_parent(device);
@@ -875,7 +877,7 @@ int of_irq_map_one(struct device_node *device, int index, struct of_irq *out_irq
 		return -EINVAL;
 
 	/* Get size of interrupt specifier */
-	tmp = (u32 *)get_property(p, "#interrupt-cells", NULL);
+	tmp = get_property(p, "#interrupt-cells", NULL);
 	if (tmp == NULL) {
 		of_node_put(p);
 		return -EINVAL;
diff --git a/arch/powerpc/kernel/rtas-proc.c b/arch/powerpc/kernel/rtas-proc.c
index 9c9ad1fa9cce..2fe82abf1c52 100644
--- a/arch/powerpc/kernel/rtas-proc.c
+++ b/arch/powerpc/kernel/rtas-proc.c
@@ -246,12 +246,12 @@ struct file_operations ppc_rtas_rmo_buf_ops = {
 
 static int ppc_rtas_find_all_sensors(void);
 static void ppc_rtas_process_sensor(struct seq_file *m,
-	struct individual_sensor *s, int state, int error, char *loc);
+	struct individual_sensor *s, int state, int error, const char *loc);
 static char *ppc_rtas_process_error(int error);
 static void get_location_code(struct seq_file *m,
-	struct individual_sensor *s, char *loc);
-static void check_location_string(struct seq_file *m, char *c);
-static void check_location(struct seq_file *m, char *c);
+	struct individual_sensor *s, const char *loc);
+static void check_location_string(struct seq_file *m, const char *c);
+static void check_location(struct seq_file *m, const char *c);
 
 static int __init proc_rtas_init(void)
 {
@@ -446,11 +446,11 @@ static int ppc_rtas_sensors_show(struct seq_file *m, void *v)
 	for (i=0; i<sensors.quant; i++) {
 		struct individual_sensor *p = &sensors.sensor[i];
 		char rstr[64];
-		char *loc;
+		const char *loc;
 		int llen, offs;
 
 		sprintf (rstr, SENSOR_PREFIX"%04d", p->token);
-		loc = (char *) get_property(rtas_node, rstr, &llen);
+		loc = get_property(rtas_node, rstr, &llen);
 
 		/* A sensor may have multiple instances */
 		for (j = 0, offs = 0; j <= p->quant; j++) {
@@ -474,10 +474,10 @@ static int ppc_rtas_sensors_show(struct seq_file *m, void *v)
 
 static int ppc_rtas_find_all_sensors(void)
 {
-	unsigned int *utmp;
+	const unsigned int *utmp;
 	int len, i;
 
-	utmp = (unsigned int *) get_property(rtas_node, "rtas-sensors", &len);
+	utmp = get_property(rtas_node, "rtas-sensors", &len);
 	if (utmp == NULL) {
 		printk (KERN_ERR "error: could not get rtas-sensors\n");
 		return 1;
@@ -530,7 +530,7 @@ static char *ppc_rtas_process_error(int error)
  */
 
 static void ppc_rtas_process_sensor(struct seq_file *m,
-	struct individual_sensor *s, int state, int error, char *loc)
+	struct individual_sensor *s, int state, int error, const char *loc)
 {
 	/* Defined return vales */
 	const char * key_switch[]        = { "Off\t", "Normal\t", "Secure\t", 
@@ -682,7 +682,7 @@ static void ppc_rtas_process_sensor(struct seq_file *m,
 
 /* ****************************************************************** */
 
-static void check_location(struct seq_file *m, char *c)
+static void check_location(struct seq_file *m, const char *c)
 {
 	switch (c[0]) {
 		case LOC_PLANAR:
@@ -719,7 +719,7 @@ static void check_location(struct seq_file *m, char *c)
  * ${LETTER}${NUMBER}[[-/]${LETTER}${NUMBER} [ ... ] ]
  * the '.' may be an abbrevation
  */
-static void check_location_string(struct seq_file *m, char *c)
+static void check_location_string(struct seq_file *m, const char *c)
 {
 	while (*c) {
 		if (isalpha(*c) || *c == '.')
@@ -733,7 +733,8 @@ static void check_location_string(struct seq_file *m, char *c)
 
 /* ****************************************************************** */
 
-static void get_location_code(struct seq_file *m, struct individual_sensor *s, char *loc)
+static void get_location_code(struct seq_file *m, struct individual_sensor *s,
+		const char *loc)
 {
 	if (!loc || !*loc) {
 		seq_printf(m, "---");/* does not have a location */
diff --git a/arch/powerpc/kernel/rtas.c b/arch/powerpc/kernel/rtas.c
index 4a4cb5598402..10e10be324c9 100644
--- a/arch/powerpc/kernel/rtas.c
+++ b/arch/powerpc/kernel/rtas.c
@@ -177,10 +177,12 @@ void __init udbg_init_rtas_console(void)
 void rtas_progress(char *s, unsigned short hex)
 {
 	struct device_node *root;
-	int width, *p;
+	int width;
+	const int *p;
 	char *os;
 	static int display_character, set_indicator;
-	static int display_width, display_lines, *row_width, form_feed;
+	static int display_width, display_lines, form_feed;
+	const static int *row_width;
 	static DEFINE_SPINLOCK(progress_lock);
 	static int current_line;
 	static int pending_newline = 0;  /* did last write end with unprinted newline? */
@@ -191,16 +193,16 @@ void rtas_progress(char *s, unsigned short hex)
 	if (display_width == 0) {
 		display_width = 0x10;
 		if ((root = find_path_device("/rtas"))) {
-			if ((p = (unsigned int *)get_property(root,
+			if ((p = get_property(root,
 					"ibm,display-line-length", NULL)))
 				display_width = *p;
-			if ((p = (unsigned int *)get_property(root,
+			if ((p = get_property(root,
 					"ibm,form-feed", NULL)))
 				form_feed = *p;
-			if ((p = (unsigned int *)get_property(root,
+			if ((p = get_property(root,
 					"ibm,display-number-of-lines", NULL)))
 				display_lines = *p;
-			row_width = (unsigned int *)get_property(root,
+			row_width = get_property(root,
 					"ibm,display-truncation-length", NULL);
 		}
 		display_character = rtas_token("display-character");
@@ -293,10 +295,10 @@ EXPORT_SYMBOL(rtas_progress);		/* needed by rtas_flash module */
 
 int rtas_token(const char *service)
 {
-	int *tokp;
+	const int *tokp;
 	if (rtas.dev == NULL)
 		return RTAS_UNKNOWN_SERVICE;
-	tokp = (int *) get_property(rtas.dev, service, NULL);
+	tokp = get_property(rtas.dev, service, NULL);
 	return tokp ? *tokp : RTAS_UNKNOWN_SERVICE;
 }
 EXPORT_SYMBOL(rtas_token);
@@ -824,15 +826,15 @@ void __init rtas_initialize(void)
 	 */
 	rtas.dev = of_find_node_by_name(NULL, "rtas");
 	if (rtas.dev) {
-		u32 *basep, *entryp;
-		u32 *sizep;
+		const u32 *basep, *entryp, *sizep;
 
-		basep = (u32 *)get_property(rtas.dev, "linux,rtas-base", NULL);
-		sizep = (u32 *)get_property(rtas.dev, "rtas-size", NULL);
+		basep = get_property(rtas.dev, "linux,rtas-base", NULL);
+		sizep = get_property(rtas.dev, "rtas-size", NULL);
 		if (basep != NULL && sizep != NULL) {
 			rtas.base = *basep;
 			rtas.size = *sizep;
-			entryp = (u32 *)get_property(rtas.dev, "linux,rtas-entry", NULL);
+			entryp = get_property(rtas.dev,
+					"linux,rtas-entry", NULL);
 			if (entryp == NULL) /* Ugh */
 				rtas.entry = rtas.base;
 			else
diff --git a/arch/powerpc/kernel/rtas_pci.c b/arch/powerpc/kernel/rtas_pci.c
index cda022657324..5a798ac6aecf 100644
--- a/arch/powerpc/kernel/rtas_pci.c
+++ b/arch/powerpc/kernel/rtas_pci.c
@@ -57,7 +57,7 @@ static inline int config_access_valid(struct pci_dn *dn, int where)
 
 static int of_device_available(struct device_node * dn)
 {
-        char * status;
+        const char *status;
 
         status = get_property(dn, "status", NULL);
 
@@ -178,7 +178,7 @@ struct pci_ops rtas_pci_ops = {
 
 int is_python(struct device_node *dev)
 {
-	char *model = (char *)get_property(dev, "model", NULL);
+	const char *model = get_property(dev, "model", NULL);
 
 	if (model && strstr(model, "Python"))
 		return 1;
@@ -234,7 +234,7 @@ void __init init_pci_config_tokens (void)
 unsigned long __devinit get_phb_buid (struct device_node *phb)
 {
 	int addr_cells;
-	unsigned int *buid_vals;
+	const unsigned int *buid_vals;
 	unsigned int len;
 	unsigned long buid;
 
@@ -247,7 +247,7 @@ unsigned long __devinit get_phb_buid (struct device_node *phb)
 	if (phb->parent->parent)
 		return 0;
 
-	buid_vals = (unsigned int *) get_property(phb, "reg", &len);
+	buid_vals = get_property(phb, "reg", &len);
 	if (buid_vals == NULL)
 		return 0;
 
@@ -264,10 +264,10 @@ unsigned long __devinit get_phb_buid (struct device_node *phb)
 static int phb_set_bus_ranges(struct device_node *dev,
 			      struct pci_controller *phb)
 {
-	int *bus_range;
+	const int *bus_range;
 	unsigned int len;
 
-	bus_range = (int *) get_property(dev, "bus-range", &len);
+	bus_range = get_property(dev, "bus-range", &len);
 	if (bus_range == NULL || len < 2 * sizeof(int)) {
 		return 1;
  	}
@@ -325,15 +325,15 @@ unsigned long __init find_and_init_phbs(void)
 	 * in chosen.
 	 */
 	if (of_chosen) {
-		int *prop;
+		const int *prop;
 
-		prop = (int *)get_property(of_chosen, "linux,pci-probe-only",
-					   NULL);
+		prop = get_property(of_chosen,
+				"linux,pci-probe-only", NULL);
 		if (prop)
 			pci_probe_only = *prop;
 
-		prop = (int *)get_property(of_chosen,
-					   "linux,pci-assign-all-buses", NULL);
+		prop = get_property(of_chosen,
+				"linux,pci-assign-all-buses", NULL);
 		if (prop)
 			pci_assign_all_buses = *prop;
 	}
diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c
index c6d7b98af7d5..aaf1727b3570 100644
--- a/arch/powerpc/kernel/setup-common.c
+++ b/arch/powerpc/kernel/setup-common.c
@@ -304,16 +304,15 @@ struct seq_operations cpuinfo_op = {
 void __init check_for_initrd(void)
 {
 #ifdef CONFIG_BLK_DEV_INITRD
-	unsigned long *prop;
+	const unsigned long *prop;
 
 	DBG(" -> check_for_initrd()\n");
 
 	if (of_chosen) {
-		prop = (unsigned long *)get_property(of_chosen,
-				"linux,initrd-start", NULL);
+		prop = get_property(of_chosen, "linux,initrd-start", NULL);
 		if (prop != NULL) {
 			initrd_start = (unsigned long)__va(*prop);
-			prop = (unsigned long *)get_property(of_chosen,
+			prop = get_property(of_chosen,
 					"linux,initrd-end", NULL);
 			if (prop != NULL) {
 				initrd_end = (unsigned long)__va(*prop);
@@ -366,15 +365,14 @@ void __init smp_setup_cpu_maps(void)
 	int cpu = 0;
 
 	while ((dn = of_find_node_by_type(dn, "cpu")) && cpu < NR_CPUS) {
-		int *intserv;
+		const int *intserv;
 		int j, len = sizeof(u32), nthreads = 1;
 
-		intserv = (int *)get_property(dn, "ibm,ppc-interrupt-server#s",
-					      &len);
+		intserv = get_property(dn, "ibm,ppc-interrupt-server#s", &len);
 		if (intserv)
 			nthreads = len / sizeof(int);
 		else {
-			intserv = (int *) get_property(dn, "reg", NULL);
+			intserv = get_property(dn, "reg", NULL);
 			if (!intserv)
 				intserv = &cpu;	/* assume logical == phys */
 		}
@@ -395,13 +393,12 @@ void __init smp_setup_cpu_maps(void)
 	if (machine_is(pseries) && firmware_has_feature(FW_FEATURE_LPAR) &&
 	    (dn = of_find_node_by_path("/rtas"))) {
 		int num_addr_cell, num_size_cell, maxcpus;
-		unsigned int *ireg;
+		const unsigned int *ireg;
 
 		num_addr_cell = prom_n_addr_cells(dn);
 		num_size_cell = prom_n_size_cells(dn);
 
-		ireg = (unsigned int *)
-			get_property(dn, "ibm,lrdr-capacity", NULL);
+		ireg = get_property(dn, "ibm,lrdr-capacity", NULL);
 
 		if (!ireg)
 			goto out;
diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index e2447aef3a8f..77efe19ccd2c 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -106,7 +106,7 @@ static int smt_enabled_cmdline;
 static void check_smt_enabled(void)
 {
 	struct device_node *dn;
-	char *smt_option;
+	const char *smt_option;
 
 	/* Allow the command line to overrule the OF option */
 	if (smt_enabled_cmdline)
@@ -115,7 +115,7 @@ static void check_smt_enabled(void)
 	dn = of_find_node_by_path("/options");
 
 	if (dn) {
-		smt_option = (char *)get_property(dn, "ibm,smt-enabled", NULL);
+		smt_option = get_property(dn, "ibm,smt-enabled", NULL);
 
                 if (smt_option) {
 			if (!strcmp(smt_option, "on"))
@@ -292,7 +292,7 @@ static void __init initialize_cache_info(void)
 		 */
 
 		if ( num_cpus == 1 ) {
-			u32 *sizep, *lsizep;
+			const u32 *sizep, *lsizep;
 			u32 size, lsize;
 			const char *dc, *ic;
 
@@ -307,10 +307,10 @@ static void __init initialize_cache_info(void)
 
 			size = 0;
 			lsize = cur_cpu_spec->dcache_bsize;
-			sizep = (u32 *)get_property(np, "d-cache-size", NULL);
+			sizep = get_property(np, "d-cache-size", NULL);
 			if (sizep != NULL)
 				size = *sizep;
-			lsizep = (u32 *) get_property(np, dc, NULL);
+			lsizep = get_property(np, dc, NULL);
 			if (lsizep != NULL)
 				lsize = *lsizep;
 			if (sizep == 0 || lsizep == 0)
@@ -324,10 +324,10 @@ static void __init initialize_cache_info(void)
 
 			size = 0;
 			lsize = cur_cpu_spec->icache_bsize;
-			sizep = (u32 *)get_property(np, "i-cache-size", NULL);
+			sizep = get_property(np, "i-cache-size", NULL);
 			if (sizep != NULL)
 				size = *sizep;
-			lsizep = (u32 *)get_property(np, ic, NULL);
+			lsizep = get_property(np, ic, NULL);
 			if (lsizep != NULL)
 				lsize = *lsizep;
 			if (sizep == 0 || lsizep == 0)
diff --git a/arch/powerpc/kernel/sysfs.c b/arch/powerpc/kernel/sysfs.c
index 010435095550..1d724aef438a 100644
--- a/arch/powerpc/kernel/sysfs.c
+++ b/arch/powerpc/kernel/sysfs.c
@@ -60,7 +60,7 @@ static int smt_snooze_cmdline;
 static int __init smt_setup(void)
 {
 	struct device_node *options;
-	unsigned int *val;
+	const unsigned int *val;
 	unsigned int cpu;
 
 	if (!cpu_has_feature(CPU_FTR_SMT))
@@ -70,8 +70,7 @@ static int __init smt_setup(void)
 	if (!options)
 		return -ENODEV;
 
-	val = (unsigned int *)get_property(options, "ibm,smt-snooze-delay",
-					   NULL);
+	val = get_property(options, "ibm,smt-snooze-delay", NULL);
 	if (!smt_snooze_cmdline && val) {
 		for_each_possible_cpu(cpu)
 			per_cpu(smt_snooze_delay, cpu) = *val;
diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c
index 774c0a3c5019..8d4ccf061a4d 100644
--- a/arch/powerpc/kernel/time.c
+++ b/arch/powerpc/kernel/time.c
@@ -859,14 +859,14 @@ EXPORT_SYMBOL(do_settimeofday);
 static int __init get_freq(char *name, int cells, unsigned long *val)
 {
 	struct device_node *cpu;
-	unsigned int *fp;
+	const unsigned int *fp;
 	int found = 0;
 
 	/* The cpu node should have timebase and clock frequency properties */
 	cpu = of_find_node_by_type(NULL, "cpu");
 
 	if (cpu) {
-		fp = (unsigned int *)get_property(cpu, name, NULL);
+		fp = get_property(cpu, name, NULL);
 		if (fp) {
 			found = 1;
 			*val = 0;
diff --git a/arch/powerpc/kernel/vio.c b/arch/powerpc/kernel/vio.c
index fad8580f9081..cb87e71eec66 100644
--- a/arch/powerpc/kernel/vio.c
+++ b/arch/powerpc/kernel/vio.c
@@ -77,7 +77,7 @@ static struct iommu_table *vio_build_iommu_table(struct vio_dev *dev)
 	} else
 #endif
 	{
-		unsigned char *dma_window;
+		const unsigned char *dma_window;
 		struct iommu_table *tbl;
 		unsigned long offset, size;
 
@@ -217,7 +217,7 @@ static void __devinit vio_dev_release(struct device *dev)
 struct vio_dev * __devinit vio_register_device_node(struct device_node *of_node)
 {
 	struct vio_dev *viodev;
-	unsigned int *unit_address;
+	const unsigned int *unit_address;
 
 	/* we need the 'device_type' property, in order to match with drivers */
 	if (of_node->type == NULL) {
@@ -227,7 +227,7 @@ struct vio_dev * __devinit vio_register_device_node(struct device_node *of_node)
 		return NULL;
 	}
 
-	unit_address = (unsigned int *)get_property(of_node, "reg", NULL);
+	unit_address = get_property(of_node, "reg", NULL);
 	if (unit_address == NULL) {
 		printk(KERN_WARNING "%s: node %s missing 'reg'\n",
 				__FUNCTION__,
@@ -249,7 +249,7 @@ struct vio_dev * __devinit vio_register_device_node(struct device_node *of_node)
 	viodev->type = of_node->type;
 	viodev->unit_address = *unit_address;
 	if (firmware_has_feature(FW_FEATURE_ISERIES)) {
-		unit_address = (unsigned int *)get_property(of_node,
+		unit_address = get_property(of_node,
 				"linux,unit_address", NULL);
 		if (unit_address != NULL)
 			viodev->unit_address = *unit_address;
@@ -423,7 +423,7 @@ static int vio_hotplug(struct device *dev, char **envp, int num_envp,
 {
 	const struct vio_dev *vio_dev = to_vio_dev(dev);
 	struct device_node *dn = dev->platform_data;
-	char *cp;
+	const char *cp;
 	int length;
 
 	if (!num_envp)
@@ -431,7 +431,7 @@ static int vio_hotplug(struct device *dev, char **envp, int num_envp,
 
 	if (!dn)
 		return -ENODEV;
-	cp = (char *)get_property(dn, "compatible", &length);
+	cp = get_property(dn, "compatible", &length);
 	if (!cp)
 		return -ENODEV;
 
@@ -493,11 +493,11 @@ static struct vio_dev *vio_find_name(const char *kobj_name)
  */
 struct vio_dev *vio_find_node(struct device_node *vnode)
 {
-	uint32_t *unit_address;
+	const uint32_t *unit_address;
 	char kobj_name[BUS_ID_SIZE];
 
 	/* construct the kobject name from the device node */
-	unit_address = (uint32_t *)get_property(vnode, "reg", NULL);
+	unit_address = get_property(vnode, "reg", NULL);
 	if (!unit_address)
 		return NULL;
 	snprintf(kobj_name, BUS_ID_SIZE, "%x", *unit_address);
diff --git a/arch/powerpc/mm/numa.c b/arch/powerpc/mm/numa.c
index fbe23933f731..6c0f1c7d83e5 100644
--- a/arch/powerpc/mm/numa.c
+++ b/arch/powerpc/mm/numa.c
@@ -159,12 +159,12 @@ static struct device_node * __cpuinit find_cpu_node(unsigned int cpu)
 {
 	unsigned int hw_cpuid = get_hard_smp_processor_id(cpu);
 	struct device_node *cpu_node = NULL;
-	unsigned int *interrupt_server, *reg;
+	const unsigned int *interrupt_server, *reg;
 	int len;
 
 	while ((cpu_node = of_find_node_by_type(cpu_node, "cpu")) != NULL) {
 		/* Try interrupt server first */
-		interrupt_server = (unsigned int *)get_property(cpu_node,
+		interrupt_server = get_property(cpu_node,
 					"ibm,ppc-interrupt-server#s", &len);
 
 		len = len / sizeof(u32);
@@ -175,8 +175,7 @@ static struct device_node * __cpuinit find_cpu_node(unsigned int cpu)
 					return cpu_node;
 			}
 		} else {
-			reg = (unsigned int *)get_property(cpu_node,
-							   "reg", &len);
+			reg = get_property(cpu_node, "reg", &len);
 			if (reg && (len > 0) && (reg[0] == hw_cpuid))
 				return cpu_node;
 		}
@@ -186,9 +185,9 @@ static struct device_node * __cpuinit find_cpu_node(unsigned int cpu)
 }
 
 /* must hold reference to node during call */
-static int *of_get_associativity(struct device_node *dev)
+static const int *of_get_associativity(struct device_node *dev)
 {
-	return (unsigned int *)get_property(dev, "ibm,associativity", NULL);
+	return get_property(dev, "ibm,associativity", NULL);
 }
 
 /* Returns nid in the range [0..MAX_NUMNODES-1], or -1 if no useful numa
@@ -197,7 +196,7 @@ static int *of_get_associativity(struct device_node *dev)
 static int of_node_to_nid_single(struct device_node *device)
 {
 	int nid = -1;
-	unsigned int *tmp;
+	const unsigned int *tmp;
 
 	if (min_common_depth == -1)
 		goto out;
@@ -255,7 +254,7 @@ EXPORT_SYMBOL_GPL(of_node_to_nid);
 static int __init find_min_common_depth(void)
 {
 	int depth;
-	unsigned int *ref_points;
+	const unsigned int *ref_points;
 	struct device_node *rtas_root;
 	unsigned int len;
 
@@ -270,7 +269,7 @@ static int __init find_min_common_depth(void)
 	 * configuration (should be all 0's) and the second is for a normal
 	 * NUMA configuration.
 	 */
-	ref_points = (unsigned int *)get_property(rtas_root,
+	ref_points = get_property(rtas_root,
 			"ibm,associativity-reference-points", &len);
 
 	if ((len >= 1) && ref_points) {
@@ -297,7 +296,7 @@ static void __init get_n_mem_cells(int *n_addr_cells, int *n_size_cells)
 	of_node_put(memory);
 }
 
-static unsigned long __devinit read_n_cells(int n, unsigned int **buf)
+static unsigned long __devinit read_n_cells(int n, const unsigned int **buf)
 {
 	unsigned long result = 0;
 
@@ -435,15 +434,13 @@ static int __init parse_numa_properties(void)
 		unsigned long size;
 		int nid;
 		int ranges;
-		unsigned int *memcell_buf;
+		const unsigned int *memcell_buf;
 		unsigned int len;
 
-		memcell_buf = (unsigned int *)get_property(memory,
+		memcell_buf = get_property(memory,
 			"linux,usable-memory", &len);
 		if (!memcell_buf || len <= 0)
-			memcell_buf =
-				(unsigned int *)get_property(memory, "reg",
-					&len);
+			memcell_buf = get_property(memory, "reg", &len);
 		if (!memcell_buf || len <= 0)
 			continue;
 
@@ -787,10 +784,10 @@ int hot_add_scn_to_nid(unsigned long scn_addr)
 	while ((memory = of_find_node_by_type(memory, "memory")) != NULL) {
 		unsigned long start, size;
 		int ranges;
-		unsigned int *memcell_buf;
+		const unsigned int *memcell_buf;
 		unsigned int len;
 
-		memcell_buf = (unsigned int *)get_property(memory, "reg", &len);
+		memcell_buf = get_property(memory, "reg", &len);
 		if (!memcell_buf || len <= 0)
 			continue;
 
diff --git a/arch/powerpc/sysdev/fsl_soc.c b/arch/powerpc/sysdev/fsl_soc.c
index e983972132d8..07c47e8309ed 100644
--- a/arch/powerpc/sysdev/fsl_soc.c
+++ b/arch/powerpc/sysdev/fsl_soc.c
@@ -41,7 +41,7 @@ phys_addr_t get_immrbase(void)
 	soc = of_find_node_by_type(NULL, "soc");
 	if (soc) {
 		unsigned int size;
-		void *prop = get_property(soc, "reg", &size);
+		const void *prop = get_property(soc, "reg", &size);
 		immrbase = of_translate_address(soc, prop);
 		of_node_put(soc);
 	};
@@ -86,8 +86,8 @@ static int __init gfar_mdio_of_init(void)
 
 		while ((child = of_get_next_child(np, child)) != NULL) {
 			if (child->n_intrs) {
-				u32 *id =
-				    (u32 *) get_property(child, "reg", NULL);
+				const u32 *id =
+					get_property(child, "reg", NULL);
 				mdio_data.irq[*id] = child->intrs[0].line;
 			}
 		}
@@ -127,10 +127,10 @@ static int __init gfar_of_init(void)
 		struct resource r[4];
 		struct device_node *phy, *mdio;
 		struct gianfar_platform_data gfar_data;
-		unsigned int *id;
-		char *model;
-		void *mac_addr;
-		phandle *ph;
+		const unsigned int *id;
+		const char *model;
+		const void *mac_addr;
+		const phandle *ph;
 
 		memset(r, 0, sizeof(r));
 		memset(&gfar_data, 0, sizeof(gfar_data));
@@ -188,7 +188,7 @@ static int __init gfar_of_init(void)
 			    FSL_GIANFAR_DEV_HAS_VLAN |
 			    FSL_GIANFAR_DEV_HAS_EXTENDED_HASH;
 
-		ph = (phandle *) get_property(np, "phy-handle", NULL);
+		ph = get_property(np, "phy-handle", NULL);
 		phy = of_find_node_by_phandle(*ph);
 
 		if (phy == NULL) {
@@ -198,7 +198,7 @@ static int __init gfar_of_init(void)
 
 		mdio = of_get_parent(phy);
 
-		id = (u32 *) get_property(phy, "reg", NULL);
+		id = get_property(phy, "reg", NULL);
 		ret = of_address_to_resource(mdio, 0, &res);
 		if (ret) {
 			of_node_put(phy);
@@ -242,7 +242,7 @@ static int __init fsl_i2c_of_init(void)
 	     i++) {
 		struct resource r[2];
 		struct fsl_i2c_platform_data i2c_data;
-		unsigned char *flags = NULL;
+		const unsigned char *flags = NULL;
 
 		memset(&r, 0, sizeof(r));
 		memset(&i2c_data, 0, sizeof(i2c_data));
@@ -294,7 +294,7 @@ static int __init mpc83xx_wdt_init(void)
 	struct resource r;
 	struct device_node *soc, *np;
 	struct platform_device *dev;
-	unsigned int *freq;
+	const unsigned int *freq;
 	int ret;
 
 	np = of_find_compatible_node(NULL, "watchdog", "mpc83xx_wdt");
@@ -311,7 +311,7 @@ static int __init mpc83xx_wdt_init(void)
 		goto nosoc;
 	}
 
-	freq = (unsigned int *)get_property(soc, "bus-frequency", NULL);
+	freq = get_property(soc, "bus-frequency", NULL);
 	if (!freq) {
 		ret = -ENODEV;
 		goto err;
@@ -351,7 +351,7 @@ nodev:
 arch_initcall(mpc83xx_wdt_init);
 #endif
 
-static enum fsl_usb2_phy_modes determine_usb_phy(char * phy_type)
+static enum fsl_usb2_phy_modes determine_usb_phy(const char *phy_type)
 {
 	if (!phy_type)
 		return FSL_USB2_PHY_NONE;
@@ -379,7 +379,7 @@ static int __init fsl_usb_of_init(void)
 	     i++) {
 		struct resource r[2];
 		struct fsl_usb2_platform_data usb_data;
-		unsigned char *prop = NULL;
+		const unsigned char *prop = NULL;
 
 		memset(&r, 0, sizeof(r));
 		memset(&usb_data, 0, sizeof(usb_data));
@@ -428,7 +428,7 @@ static int __init fsl_usb_of_init(void)
 	     i++) {
 		struct resource r[2];
 		struct fsl_usb2_platform_data usb_data;
-		unsigned char *prop = NULL;
+		const unsigned char *prop = NULL;
 
 		memset(&r, 0, sizeof(r));
 		memset(&usb_data, 0, sizeof(usb_data));
diff --git a/arch/powerpc/sysdev/mmio_nvram.c b/arch/powerpc/sysdev/mmio_nvram.c
index 615350d46b52..ff23f5a4d4b9 100644
--- a/arch/powerpc/sysdev/mmio_nvram.c
+++ b/arch/powerpc/sysdev/mmio_nvram.c
@@ -80,7 +80,7 @@ static ssize_t mmio_nvram_get_size(void)
 int __init mmio_nvram_init(void)
 {
 	struct device_node *nvram_node;
-	unsigned long *buffer;
+	const unsigned long *buffer;
 	int proplen;
 	unsigned long nvram_addr;
 	int ret;
@@ -91,7 +91,7 @@ int __init mmio_nvram_init(void)
 		goto out;
 
 	ret = -EIO;
-	buffer = (unsigned long *)get_property(nvram_node, "reg", &proplen);
+	buffer = get_property(nvram_node, "reg", &proplen);
 	if (proplen != 2*sizeof(unsigned long))
 		goto out;
 
diff --git a/include/asm-powerpc/ibmebus.h b/include/asm-powerpc/ibmebus.h
index 7a42723d107c..7ab195a27888 100644
--- a/include/asm-powerpc/ibmebus.h
+++ b/include/asm-powerpc/ibmebus.h
@@ -48,7 +48,7 @@ extern struct dma_mapping_ops ibmebus_dma_ops;
 extern struct bus_type ibmebus_bus_type;
 
 struct ibmebus_dev {	
-	char *name;
+	const char *name;
 	struct of_device ofdev;
 };
 
diff --git a/include/asm-powerpc/prom.h b/include/asm-powerpc/prom.h
index 56f6ea0c76de..abdf1be66e97 100644
--- a/include/asm-powerpc/prom.h
+++ b/include/asm-powerpc/prom.h
@@ -72,8 +72,8 @@ struct property {
 };
 
 struct device_node {
-	char	*name;
-	char	*type;
+	const char *name;
+	const char *type;
 	phandle	node;
 	phandle linux_phandle;
 	char	*full_name;
@@ -209,15 +209,15 @@ static inline u64 of_read_number(const u32 *cell, int size)
 /* Translate an OF address block into a CPU physical address
  */
 #define OF_BAD_ADDR	((u64)-1)
-extern u64 of_translate_address(struct device_node *np, u32 *addr);
+extern u64 of_translate_address(struct device_node *np, const u32 *addr);
 
 /* Extract an address from a device, returns the region size and
  * the address space flags too. The PCI version uses a BAR number
  * instead of an absolute index
  */
-extern u32 *of_get_address(struct device_node *dev, int index,
+extern const u32 *of_get_address(struct device_node *dev, int index,
 			   u64 *size, unsigned int *flags);
-extern u32 *of_get_pci_address(struct device_node *dev, int bar_no,
+extern const u32 *of_get_pci_address(struct device_node *dev, int bar_no,
 			       u64 *size, unsigned int *flags);
 
 /* Get an address as a resource. Note that if your address is
@@ -234,7 +234,7 @@ extern int of_pci_address_to_resource(struct device_node *dev, int bar,
 /* Parse the ibm,dma-window property of an OF node into the busno, phys and
  * size parameters.
  */
-void of_parse_dma_window(struct device_node *dn, unsigned char *dma_window_prop,
+void of_parse_dma_window(struct device_node *dn, const void *dma_window_prop,
 		unsigned long *busno, unsigned long *phys, unsigned long *size);
 
 extern void kdump_move_device_tree(void);
@@ -288,8 +288,8 @@ extern void of_irq_map_init(unsigned int flags);
  *
  */
 
-extern int of_irq_map_raw(struct device_node *parent, u32 *intspec, u32 *addr,
-			  struct of_irq *out_irq);
+extern int of_irq_map_raw(struct device_node *parent, const u32 *intspec,
+			  const u32 *addr, struct of_irq *out_irq);
 
 
 /***
diff --git a/include/asm-powerpc/vio.h b/include/asm-powerpc/vio.h
index dc9bd101ca14..4b51d42e1419 100644
--- a/include/asm-powerpc/vio.h
+++ b/include/asm-powerpc/vio.h
@@ -46,8 +46,8 @@ struct iommu_table;
  */
 struct vio_dev {
 	struct iommu_table *iommu_table;     /* vio_map_* uses this */
-	char *name;
-	char *type;
+	const char *name;
+	const char *type;
 	uint32_t unit_address;
 	unsigned int irq;
 	struct device dev;
-- 
cgit v1.2.3


From c61c27d58af61e5b78257019b173732c29ce0c64 Mon Sep 17 00:00:00 2001
From: Jeremy Kerr <jk@ozlabs.org>
Date: Wed, 12 Jul 2006 15:39:54 +1000
Subject: [POWERPC] cell: Constify & voidify get_property()

Now that get_property() returns a void *, there's no need to cast its
return value. Also, treat the return value as const, so we can
constify get_property later.

cell platform changes.

Built for cell_defconfig

Signed-off-by: Jeremy Kerr <jk@ozlabs.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 arch/powerpc/platforms/cell/cbe_regs.c   |  8 +++-----
 arch/powerpc/platforms/cell/interrupt.c  |  5 ++---
 arch/powerpc/platforms/cell/iommu.c      | 22 ++++++++++++----------
 arch/powerpc/platforms/cell/spider-pic.c | 10 +++++-----
 arch/powerpc/platforms/cell/spu_base.c   | 18 +++++++++---------
 include/asm-powerpc/spu.h                |  2 +-
 6 files changed, 32 insertions(+), 33 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/platforms/cell/cbe_regs.c b/arch/powerpc/platforms/cell/cbe_regs.c
index ce696c1cca75..3f3859d12e00 100644
--- a/arch/powerpc/platforms/cell/cbe_regs.c
+++ b/arch/powerpc/platforms/cell/cbe_regs.c
@@ -97,7 +97,7 @@ void __init cbe_regs_init(void)
 		struct cbe_regs_map *map = &cbe_regs_maps[cbe_regs_map_count++];
 
 		/* That hack must die die die ! */
-		struct address_prop {
+		const struct address_prop {
 			unsigned long address;
 			unsigned int len;
 		} __attribute__((packed)) *prop;
@@ -114,13 +114,11 @@ void __init cbe_regs_init(void)
 			if (cbe_thread_map[i].cpu_node == cpu)
 				cbe_thread_map[i].regs = map;
 
-		prop = (struct address_prop *)get_property(cpu, "pervasive",
-							   NULL);
+		prop = get_property(cpu, "pervasive", NULL);
 		if (prop != NULL)
 			map->pmd_regs = ioremap(prop->address, prop->len);
 
-		prop = (struct address_prop *)get_property(cpu, "iic",
-							   NULL);
+		prop = get_property(cpu, "iic", NULL);
 		if (prop != NULL)
 			map->iic_regs = ioremap(prop->address, prop->len);
 	}
diff --git a/arch/powerpc/platforms/cell/interrupt.c b/arch/powerpc/platforms/cell/interrupt.c
index 9d5da7896892..b26b496f6548 100644
--- a/arch/powerpc/platforms/cell/interrupt.c
+++ b/arch/powerpc/platforms/cell/interrupt.c
@@ -250,15 +250,14 @@ static int __init setup_iic(void)
 	struct resource r0, r1;
 	struct irq_host *host;
 	int found = 0;
- 	u32 *np;
+ 	const u32 *np;
 
 	for (dn = NULL;
 	     (dn = of_find_node_by_name(dn,"interrupt-controller")) != NULL;) {
 		if (!device_is_compatible(dn,
 				     "IBM,CBEA-Internal-Interrupt-Controller"))
 			continue;
- 		np = (u32 *)get_property(dn, "ibm,interrupt-server-ranges",
-					 NULL);
+		np = get_property(dn, "ibm,interrupt-server-ranges", NULL);
  		if (np == NULL) {
 			printk(KERN_WARNING "IIC: CPU association not found\n");
 			of_node_put(dn);
diff --git a/arch/powerpc/platforms/cell/iommu.c b/arch/powerpc/platforms/cell/iommu.c
index a35004e14c69..d2b20eba5b87 100644
--- a/arch/powerpc/platforms/cell/iommu.c
+++ b/arch/powerpc/platforms/cell/iommu.c
@@ -308,15 +308,16 @@ static void cell_do_map_iommu(struct cell_iommu *iommu,
 
 static void iommu_devnode_setup(struct device_node *d)
 {
-	unsigned int *ioid;
-	unsigned long *dma_window, map_start, map_size, token;
+	const unsigned int *ioid;
+	unsigned long map_start, map_size, token;
+	const unsigned long *dma_window;
 	struct cell_iommu *iommu;
 
-	ioid = (unsigned int *)get_property(d, "ioid", NULL);
+	ioid = get_property(d, "ioid", NULL);
 	if (!ioid)
 		pr_debug("No ioid entry found !\n");
 
-	dma_window = (unsigned long *)get_property(d, "ibm,dma-window", NULL);
+	dma_window = get_property(d, "ibm,dma-window", NULL);
 	if (!dma_window)
 		pr_debug("No ibm,dma-window entry found !\n");
 
@@ -371,8 +372,9 @@ static int cell_map_iommu_hardcoded(int num_nodes)
 
 static int cell_map_iommu(void)
 {
-	unsigned int num_nodes = 0, *node_id;
-	unsigned long *base, *mmio_base;
+	unsigned int num_nodes = 0;
+	const unsigned int *node_id;
+	const unsigned long *base, *mmio_base;
 	struct device_node *dn;
 	struct cell_iommu *iommu = NULL;
 
@@ -381,7 +383,7 @@ static int cell_map_iommu(void)
 	for(dn = of_find_node_by_type(NULL, "cpu");
 	    dn;
 	    dn = of_find_node_by_type(dn, "cpu")) {
-		node_id = (unsigned int *)get_property(dn, "node-id", NULL);
+		node_id = get_property(dn, "node-id", NULL);
 
 		if (num_nodes < *node_id)
 			num_nodes = *node_id;
@@ -396,9 +398,9 @@ static int cell_map_iommu(void)
 	    dn;
 	    dn = of_find_node_by_type(dn, "cpu")) {
 
-		node_id = (unsigned int *)get_property(dn, "node-id", NULL);
-		base = (unsigned long *)get_property(dn, "ioc-cache", NULL);
-		mmio_base = (unsigned long *)get_property(dn, "ioc-translation", NULL);
+		node_id = get_property(dn, "node-id", NULL);
+		base = get_property(dn, "ioc-cache", NULL);
+		mmio_base = get_property(dn, "ioc-translation", NULL);
 
 		if (!base || !mmio_base || !node_id)
 			return cell_map_iommu_hardcoded(num_nodes);
diff --git a/arch/powerpc/platforms/cell/spider-pic.c b/arch/powerpc/platforms/cell/spider-pic.c
index ae7ef88f1a37..ab4c252a4d9b 100644
--- a/arch/powerpc/platforms/cell/spider-pic.c
+++ b/arch/powerpc/platforms/cell/spider-pic.c
@@ -230,7 +230,7 @@ static void spider_irq_cascade(unsigned int irq, struct irq_desc *desc,
 static unsigned int __init spider_find_cascade_and_node(struct spider_pic *pic)
 {
 	unsigned int virq;
-	u32 *imap, *tmp;
+	const u32 *imap, *tmp;
 	int imaplen, intsize, unit;
 	struct device_node *iic;
 	struct irq_host *iic_host;
@@ -248,25 +248,25 @@ static unsigned int __init spider_find_cascade_and_node(struct spider_pic *pic)
 #endif
 
 	/* Now do the horrible hacks */
-	tmp = (u32 *)get_property(pic->of_node, "#interrupt-cells", NULL);
+	tmp = get_property(pic->of_node, "#interrupt-cells", NULL);
 	if (tmp == NULL)
 		return NO_IRQ;
 	intsize = *tmp;
-	imap = (u32 *)get_property(pic->of_node, "interrupt-map", &imaplen);
+	imap = get_property(pic->of_node, "interrupt-map", &imaplen);
 	if (imap == NULL || imaplen < (intsize + 1))
 		return NO_IRQ;
 	iic = of_find_node_by_phandle(imap[intsize]);
 	if (iic == NULL)
 		return NO_IRQ;
 	imap += intsize + 1;
-	tmp = (u32 *)get_property(iic, "#interrupt-cells", NULL);
+	tmp = get_property(iic, "#interrupt-cells", NULL);
 	if (tmp == NULL)
 		return NO_IRQ;
 	intsize = *tmp;
 	/* Assume unit is last entry of interrupt specifier */
 	unit = imap[intsize - 1];
 	/* Ok, we have a unit, now let's try to get the node */
-	tmp = (u32 *)get_property(iic, "ibm,interrupt-server-ranges", NULL);
+	tmp = get_property(iic, "ibm,interrupt-server-ranges", NULL);
 	if (tmp == NULL) {
 		of_node_put(iic);
 		return NO_IRQ;
diff --git a/arch/powerpc/platforms/cell/spu_base.c b/arch/powerpc/platforms/cell/spu_base.c
index 5d2313a6c82b..86d55675e1d2 100644
--- a/arch/powerpc/platforms/cell/spu_base.c
+++ b/arch/powerpc/platforms/cell/spu_base.c
@@ -488,10 +488,10 @@ int spu_irq_class_1_bottom(struct spu *spu)
 
 static int __init find_spu_node_id(struct device_node *spe)
 {
-	unsigned int *id;
+	const unsigned int *id;
 	struct device_node *cpu;
 	cpu = spe->parent->parent;
-	id = (unsigned int *)get_property(cpu, "node-id", NULL);
+	id = get_property(cpu, "node-id", NULL);
 	return id ? *id : 0;
 }
 
@@ -500,7 +500,7 @@ static int __init cell_spuprop_present(struct spu *spu, struct device_node *spe,
 {
 	static DEFINE_MUTEX(add_spumem_mutex);
 
-	struct address_prop {
+	const struct address_prop {
 		unsigned long address;
 		unsigned int len;
 	} __attribute__((packed)) *p;
@@ -511,7 +511,7 @@ static int __init cell_spuprop_present(struct spu *spu, struct device_node *spe,
 	struct zone *zone;
 	int ret;
 
-	p = (void*)get_property(spe, prop, &proplen);
+	p = get_property(spe, prop, &proplen);
 	WARN_ON(proplen != sizeof (*p));
 
 	start_pfn = p->address >> PAGE_SHIFT;
@@ -531,12 +531,12 @@ static int __init cell_spuprop_present(struct spu *spu, struct device_node *spe,
 static void __iomem * __init map_spe_prop(struct spu *spu,
 		struct device_node *n, const char *name)
 {
-	struct address_prop {
+	const struct address_prop {
 		unsigned long address;
 		unsigned int len;
 	} __attribute__((packed)) *prop;
 
-	void *p;
+	const void *p;
 	int proplen;
 	void* ret = NULL;
 	int err = 0;
@@ -570,14 +570,14 @@ static int __init spu_map_interrupts(struct spu *spu, struct device_node *np)
 {
 	struct irq_host *host;
 	unsigned int isrc;
-	u32 *tmp;
+	const u32 *tmp;
 
 	host = iic_get_irq_host(spu->node);
 	if (host == NULL)
 		return -ENODEV;
 
 	/* Get the interrupt source from the device-tree */
-	tmp = (u32 *)get_property(np, "isrc", NULL);
+	tmp = get_property(np, "isrc", NULL);
 	if (!tmp)
 		return -ENODEV;
 	spu->isrc = isrc = tmp[0];
@@ -593,7 +593,7 @@ static int __init spu_map_interrupts(struct spu *spu, struct device_node *np)
 
 static int __init spu_map_device(struct spu *spu, struct device_node *node)
 {
-	char *prop;
+	const char *prop;
 	int ret;
 
 	ret = -ENODEV;
diff --git a/include/asm-powerpc/spu.h b/include/asm-powerpc/spu.h
index c02d105d8294..b42b53c40f5d 100644
--- a/include/asm-powerpc/spu.h
+++ b/include/asm-powerpc/spu.h
@@ -106,7 +106,7 @@ struct spu_context;
 struct spu_runqueue;
 
 struct spu {
-	char *name;
+	const char *name;
 	unsigned long local_store_phys;
 	u8 *local_store;
 	unsigned long problem_phys;
-- 
cgit v1.2.3


From 018a3d1db7cdb6127656c1622ee1d2302e16436d Mon Sep 17 00:00:00 2001
From: Jeremy Kerr <jk@ozlabs.org>
Date: Wed, 12 Jul 2006 15:40:29 +1000
Subject: [POWERPC] powermac: Constify & voidify get_property()

Now that get_property() returns a void *, there's no need to cast its
return value. Also, treat the return value as const, so we can
constify get_property later.

powermac platform & macintosh driver changes.

Built for pmac32_defconfig, g5_defconfig

Signed-off-by: Jeremy Kerr <jk@ozlabs.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 arch/powerpc/platforms/powermac/backlight.c  |  3 ++-
 arch/powerpc/platforms/powermac/cpufreq_32.c | 23 ++++++++---------
 arch/powerpc/platforms/powermac/cpufreq_64.c | 27 ++++++++++----------
 arch/powerpc/platforms/powermac/feature.c    | 30 +++++++++++-----------
 arch/powerpc/platforms/powermac/low_i2c.c    | 24 +++++++++---------
 arch/powerpc/platforms/powermac/pci.c        | 37 +++++++++++++++-------------
 arch/powerpc/platforms/powermac/pfunc_base.c |  2 +-
 arch/powerpc/platforms/powermac/pfunc_core.c |  5 ++--
 arch/powerpc/platforms/powermac/setup.c      | 18 ++++++--------
 arch/powerpc/platforms/powermac/smp.c        |  7 +++---
 arch/powerpc/platforms/powermac/udbg_scc.c   | 10 ++++----
 drivers/i2c/busses/i2c-powermac.c            |  3 ++-
 drivers/ide/ppc/pmac.c                       |  6 ++---
 drivers/macintosh/macio_asic.c               | 10 +++++---
 drivers/macintosh/macio_sysfs.c              |  8 +++---
 drivers/macintosh/smu.c                      | 19 +++++++-------
 drivers/macintosh/therm_adt746x.c            |  8 +++---
 drivers/macintosh/therm_pm72.c               | 14 +++++------
 drivers/macintosh/therm_windtunnel.c         |  4 +--
 drivers/macintosh/via-cuda.c                 |  4 +--
 drivers/macintosh/via-pmu-led.c              |  2 +-
 drivers/macintosh/via-pmu.c                  | 10 ++++----
 drivers/macintosh/windfarm_pm81.c            |  4 +--
 drivers/macintosh/windfarm_pm91.c            |  2 +-
 drivers/macintosh/windfarm_smu_controls.c    | 13 +++++-----
 drivers/macintosh/windfarm_smu_sat.c         |  8 +++---
 drivers/macintosh/windfarm_smu_sensors.c     | 12 ++++-----
 drivers/serial/pmac_zilog.c                  |  9 ++++---
 include/asm-powerpc/smu.h                    |  2 +-
 29 files changed, 166 insertions(+), 158 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/platforms/powermac/backlight.c b/arch/powerpc/platforms/powermac/backlight.c
index 69f65e215a5c..205b4a392862 100644
--- a/arch/powerpc/platforms/powermac/backlight.c
+++ b/arch/powerpc/platforms/powermac/backlight.c
@@ -38,7 +38,8 @@ int pmac_has_backlight_type(const char *type)
 	struct device_node* bk_node = find_devices("backlight");
 
 	if (bk_node) {
-		char *prop = get_property(bk_node, "backlight-control", NULL);
+		const char *prop = get_property(bk_node,
+				"backlight-control", NULL);
 		if (prop && strncmp(prop, type, strlen(type)) == 0)
 			return 1;
 	}
diff --git a/arch/powerpc/platforms/powermac/cpufreq_32.c b/arch/powerpc/platforms/powermac/cpufreq_32.c
index 62926248bdb8..c2b6b4134f68 100644
--- a/arch/powerpc/platforms/powermac/cpufreq_32.c
+++ b/arch/powerpc/platforms/powermac/cpufreq_32.c
@@ -421,7 +421,7 @@ static int pmac_cpufreq_cpu_init(struct cpufreq_policy *policy)
 
 static u32 read_gpio(struct device_node *np)
 {
-	u32 *reg = (u32 *)get_property(np, "reg", NULL);
+	const u32 *reg = get_property(np, "reg", NULL);
 	u32 offset;
 
 	if (reg == NULL)
@@ -497,7 +497,7 @@ static int pmac_cpufreq_init_MacRISC3(struct device_node *cpunode)
 								"frequency-gpio");
 	struct device_node *slew_done_gpio_np = of_find_node_by_name(NULL,
 								     "slewing-done");
-	u32 *value;
+	const u32 *value;
 
 	/*
 	 * Check to see if it's GPIO driven or PMU only
@@ -519,15 +519,15 @@ static int pmac_cpufreq_init_MacRISC3(struct device_node *cpunode)
 	 */
 	if (frequency_gpio && slew_done_gpio) {
 		int lenp, rc;
-		u32 *freqs, *ratio;
+		const u32 *freqs, *ratio;
 
-		freqs = (u32 *)get_property(cpunode, "bus-frequencies", &lenp);
+		freqs = get_property(cpunode, "bus-frequencies", &lenp);
 		lenp /= sizeof(u32);
 		if (freqs == NULL || lenp != 2) {
 			printk(KERN_ERR "cpufreq: bus-frequencies incorrect or missing\n");
 			return 1;
 		}
-		ratio = (u32 *)get_property(cpunode, "processor-to-bus-ratio*2", NULL);
+		ratio = get_property(cpunode, "processor-to-bus-ratio*2", NULL);
 		if (ratio == NULL) {
 			printk(KERN_ERR "cpufreq: processor-to-bus-ratio*2 missing\n");
 			return 1;
@@ -562,7 +562,7 @@ static int pmac_cpufreq_init_MacRISC3(struct device_node *cpunode)
 	/* If we use the PMU, look for the min & max frequencies in the
 	 * device-tree
 	 */
-	value = (u32 *)get_property(cpunode, "min-clock-frequency", NULL);
+	value = get_property(cpunode, "min-clock-frequency", NULL);
 	if (!value)
 		return 1;
 	low_freq = (*value) / 1000;
@@ -571,7 +571,7 @@ static int pmac_cpufreq_init_MacRISC3(struct device_node *cpunode)
 	if (low_freq < 100000)
 		low_freq *= 10;
 
-	value = (u32 *)get_property(cpunode, "max-clock-frequency", NULL);
+	value = get_property(cpunode, "max-clock-frequency", NULL);
 	if (!value)
 		return 1;
 	hi_freq = (*value) / 1000;
@@ -611,13 +611,14 @@ static int pmac_cpufreq_init_7447A(struct device_node *cpunode)
 static int pmac_cpufreq_init_750FX(struct device_node *cpunode)
 {
 	struct device_node *volt_gpio_np;
-	u32 pvr, *value;
+	u32 pvr;
+	const u32 *value;
 
 	if (get_property(cpunode, "dynamic-power-step", NULL) == NULL)
 		return 1;
 
 	hi_freq = cur_freq;
-	value = (u32 *)get_property(cpunode, "reduced-clock-frequency", NULL);
+	value = get_property(cpunode, "reduced-clock-frequency", NULL);
 	if (!value)
 		return 1;
 	low_freq = (*value) / 1000;
@@ -650,7 +651,7 @@ static int pmac_cpufreq_init_750FX(struct device_node *cpunode)
 static int __init pmac_cpufreq_setup(void)
 {
 	struct device_node	*cpunode;
-	u32			*value;
+	const u32		*value;
 
 	if (strstr(cmd_line, "nocpufreq"))
 		return 0;
@@ -661,7 +662,7 @@ static int __init pmac_cpufreq_setup(void)
 		goto out;
 
 	/* Get current cpu clock freq */
-	value = (u32 *)get_property(cpunode, "clock-frequency", NULL);
+	value = get_property(cpunode, "clock-frequency", NULL);
 	if (!value)
 		goto out;
 	cur_freq = (*value) / 1000;
diff --git a/arch/powerpc/platforms/powermac/cpufreq_64.c b/arch/powerpc/platforms/powermac/cpufreq_64.c
index a6a84ac5433e..c364c89adb4e 100644
--- a/arch/powerpc/platforms/powermac/cpufreq_64.c
+++ b/arch/powerpc/platforms/powermac/cpufreq_64.c
@@ -89,7 +89,7 @@ static DEFINE_MUTEX(g5_switch_mutex);
 
 #ifdef CONFIG_PPC_SMU
 
-static u32 *g5_pmode_data;
+static const u32 *g5_pmode_data;
 static int g5_pmode_max;
 
 static struct smu_sdbp_fvt *g5_fvt_table;	/* table of op. points */
@@ -391,7 +391,8 @@ static int __init g5_neo2_cpufreq_init(struct device_node *cpus)
 	unsigned int psize, ssize;
 	unsigned long max_freq;
 	char *freq_method, *volt_method;
-	u32 *valp, pvr_hi;
+	const u32 *valp;
+	u32 pvr_hi;
 	int use_volts_vdnap = 0;
 	int use_volts_smu = 0;
 	int rc = -ENODEV;
@@ -409,8 +410,7 @@ static int __init g5_neo2_cpufreq_init(struct device_node *cpus)
 	/* Get first CPU node */
 	for (cpunode = NULL;
 	     (cpunode = of_get_next_child(cpus, cpunode)) != NULL;) {
-		u32 *reg =
-			(u32 *)get_property(cpunode, "reg", NULL);
+		const u32 *reg = get_property(cpunode, "reg", NULL);
 		if (reg == NULL || (*reg) != 0)
 			continue;
 		if (!strcmp(cpunode->type, "cpu"))
@@ -422,7 +422,7 @@ static int __init g5_neo2_cpufreq_init(struct device_node *cpus)
 	}
 
 	/* Check 970FX for now */
-	valp = (u32 *)get_property(cpunode, "cpu-version", NULL);
+	valp = get_property(cpunode, "cpu-version", NULL);
 	if (!valp) {
 		DBG("No cpu-version property !\n");
 		goto bail_noprops;
@@ -434,7 +434,7 @@ static int __init g5_neo2_cpufreq_init(struct device_node *cpus)
 	}
 
 	/* Look for the powertune data in the device-tree */
-	g5_pmode_data = (u32 *)get_property(cpunode, "power-mode-data",&psize);
+	g5_pmode_data = get_property(cpunode, "power-mode-data",&psize);
 	if (!g5_pmode_data) {
 		DBG("No power-mode-data !\n");
 		goto bail_noprops;
@@ -442,7 +442,7 @@ static int __init g5_neo2_cpufreq_init(struct device_node *cpus)
 	g5_pmode_max = psize / sizeof(u32) - 1;
 
 	if (use_volts_smu) {
-		struct smu_sdbp_header *shdr;
+		const struct smu_sdbp_header *shdr;
 
 		/* Look for the FVT table */
 		shdr = smu_get_sdb_partition(SMU_SDB_FVT_ID, NULL);
@@ -493,7 +493,7 @@ static int __init g5_neo2_cpufreq_init(struct device_node *cpus)
 	 * half freq in this version. So far, I haven't yet seen a machine
 	 * supporting anything else.
 	 */
-	valp = (u32 *)get_property(cpunode, "clock-frequency", NULL);
+	valp = get_property(cpunode, "clock-frequency", NULL);
 	if (!valp)
 		return -ENODEV;
 	max_freq = (*valp)/1000;
@@ -541,8 +541,8 @@ static int __init g5_neo2_cpufreq_init(struct device_node *cpus)
 static int __init g5_pm72_cpufreq_init(struct device_node *cpus)
 {
 	struct device_node *cpuid = NULL, *hwclock = NULL, *cpunode = NULL;
-	u8 *eeprom = NULL;
-	u32 *valp;
+	const u8 *eeprom = NULL;
+	const u32 *valp;
 	u64 max_freq, min_freq, ih, il;
 	int has_volt = 1, rc = 0;
 
@@ -563,7 +563,7 @@ static int __init g5_pm72_cpufreq_init(struct device_node *cpus)
 	/* Lookup the cpuid eeprom node */
         cpuid = of_find_node_by_path("/u3@0,f8000000/i2c@f8001000/cpuid@a0");
 	if (cpuid != NULL)
-		eeprom = (u8 *)get_property(cpuid, "cpuid", NULL);
+		eeprom = get_property(cpuid, "cpuid", NULL);
 	if (eeprom == NULL) {
 		printk(KERN_ERR "cpufreq: Can't find cpuid EEPROM !\n");
 		rc = -ENODEV;
@@ -573,7 +573,8 @@ static int __init g5_pm72_cpufreq_init(struct device_node *cpus)
 	/* Lookup the i2c hwclock */
 	for (hwclock = NULL;
 	     (hwclock = of_find_node_by_name(hwclock, "i2c-hwclock")) != NULL;){
-		char *loc = get_property(hwclock, "hwctrl-location", NULL);
+		const char *loc = get_property(hwclock,
+				"hwctrl-location", NULL);
 		if (loc == NULL)
 			continue;
 		if (strcmp(loc, "CPU CLOCK"))
@@ -637,7 +638,7 @@ static int __init g5_pm72_cpufreq_init(struct device_node *cpus)
 	 */
 
 	/* Get max frequency from device-tree */
-	valp = (u32 *)get_property(cpunode, "clock-frequency", NULL);
+	valp = get_property(cpunode, "clock-frequency", NULL);
 	if (!valp) {
 		printk(KERN_ERR "cpufreq: Can't find CPU frequency !\n");
 		rc = -ENODEV;
diff --git a/arch/powerpc/platforms/powermac/feature.c b/arch/powerpc/platforms/powermac/feature.c
index f8313bf9a9f7..13fcaf5b1796 100644
--- a/arch/powerpc/platforms/powermac/feature.c
+++ b/arch/powerpc/platforms/powermac/feature.c
@@ -1058,8 +1058,8 @@ core99_reset_cpu(struct device_node *node, long param, long value)
 	if (np == NULL)
 		return -ENODEV;
 	for (np = np->child; np != NULL; np = np->sibling) {
-		u32 *num = (u32 *)get_property(np, "reg", NULL);
-		u32 *rst = (u32 *)get_property(np, "soft-reset", NULL);
+		u32 *num = get_property(np, "reg", NULL);
+		u32 *rst = get_property(np, "soft-reset", NULL);
 		if (num == NULL || rst == NULL)
 			continue;
 		if (param == *num) {
@@ -1087,7 +1087,7 @@ core99_usb_enable(struct device_node *node, long param, long value)
 {
 	struct macio_chip *macio;
 	unsigned long flags;
-	char *prop;
+	const char *prop;
 	int number;
 	u32 reg;
 
@@ -1096,7 +1096,7 @@ core99_usb_enable(struct device_node *node, long param, long value)
 	    macio->type != macio_intrepid)
 		return -ENODEV;
 
-	prop = (char *)get_property(node, "AAPL,clock-id", NULL);
+	prop = get_property(node, "AAPL,clock-id", NULL);
 	if (!prop)
 		return -ENODEV;
 	if (strncmp(prop, "usb0u048", 8) == 0)
@@ -1507,8 +1507,8 @@ static long g5_reset_cpu(struct device_node *node, long param, long value)
 	if (np == NULL)
 		return -ENODEV;
 	for (np = np->child; np != NULL; np = np->sibling) {
-		u32 *num = (u32 *)get_property(np, "reg", NULL);
-		u32 *rst = (u32 *)get_property(np, "soft-reset", NULL);
+		const u32 *num = get_property(np, "reg", NULL);
+		const u32 *rst = get_property(np, "soft-reset", NULL);
 		if (num == NULL || rst == NULL)
 			continue;
 		if (param == *num) {
@@ -2408,7 +2408,7 @@ static int __init probe_motherboard(void)
 	 */
 	dt = find_devices("device-tree");
 	if (dt != NULL)
-		model = (const char *) get_property(dt, "model", NULL);
+		model = get_property(dt, "model", NULL);
 	for(i=0; model && i<(sizeof(pmac_mb_defs)/sizeof(struct pmac_mb_def)); i++) {
 	    if (strcmp(model, pmac_mb_defs[i].model_string) == 0) {
 		pmac_mb = pmac_mb_defs[i];
@@ -2536,7 +2536,7 @@ found:
  */
 static void __init probe_uninorth(void)
 {
-	u32 *addrp;
+	const u32 *addrp;
 	phys_addr_t address;
 	unsigned long actrl;
 
@@ -2555,7 +2555,7 @@ static void __init probe_uninorth(void)
 	if (uninorth_node == NULL)
 		return;
 
-	addrp = (u32 *)get_property(uninorth_node, "reg", NULL);
+	addrp = get_property(uninorth_node, "reg", NULL);
 	if (addrp == NULL)
 		return;
 	address = of_translate_address(uninorth_node, addrp);
@@ -2596,7 +2596,7 @@ static void __init probe_one_macio(const char *name, const char *compat, int typ
 	struct device_node*	node;
 	int			i;
 	volatile u32 __iomem	*base;
-	u32			*addrp, *revp;
+	const u32		*addrp, *revp;
 	phys_addr_t		addr;
 	u64			size;
 
@@ -2639,7 +2639,7 @@ static void __init probe_one_macio(const char *name, const char *compat, int typ
 		return;
 	}
 	if (type == macio_keylargo || type == macio_keylargo2) {
-		u32 *did = (u32 *)get_property(node, "device-id", NULL);
+		const u32 *did = get_property(node, "device-id", NULL);
 		if (*did == 0x00000025)
 			type = macio_pangea;
 		if (*did == 0x0000003e)
@@ -2652,7 +2652,7 @@ static void __init probe_one_macio(const char *name, const char *compat, int typ
 	macio_chips[i].base	= base;
 	macio_chips[i].flags	= MACIO_FLAG_SCCB_ON | MACIO_FLAG_SCCB_ON;
 	macio_chips[i].name	= macio_names[type];
-	revp = (u32 *)get_property(node, "revision-id", NULL);
+	revp = get_property(node, "revision-id", NULL);
 	if (revp)
 		macio_chips[i].rev = *revp;
 	printk(KERN_INFO "Found a %s mac-io controller, rev: %d, mapped at 0x%p\n",
@@ -2695,15 +2695,15 @@ static void __init
 initial_serial_shutdown(struct device_node *np)
 {
 	int len;
-	struct slot_names_prop {
+	const struct slot_names_prop {
 		int	count;
 		char	name[1];
 	} *slots;
-	char *conn;
+	const char *conn;
 	int port_type = PMAC_SCC_ASYNC;
 	int modem = 0;
 
-	slots = (struct slot_names_prop *)get_property(np, "slot-names", &len);
+	slots = get_property(np, "slot-names", &len);
 	conn = get_property(np, "AAPL,connector", &len);
 	if (conn && (strcmp(conn, "infrared") == 0))
 		port_type = PMAC_SCC_IRDA;
diff --git a/arch/powerpc/platforms/powermac/low_i2c.c b/arch/powerpc/platforms/powermac/low_i2c.c
index 8677f50c2586..c2c7cf75dd5f 100644
--- a/arch/powerpc/platforms/powermac/low_i2c.c
+++ b/arch/powerpc/platforms/powermac/low_i2c.c
@@ -477,7 +477,8 @@ static int kw_i2c_xfer(struct pmac_i2c_bus *bus, u8 addrdir, int subsize,
 static struct pmac_i2c_host_kw *__init kw_i2c_host_init(struct device_node *np)
 {
 	struct pmac_i2c_host_kw *host;
-	u32			*psteps, *prate, *addrp, steps;
+	const u32		*psteps, *prate, *addrp;
+	u32			steps;
 
 	host = kzalloc(sizeof(struct pmac_i2c_host_kw), GFP_KERNEL);
 	if (host == NULL) {
@@ -490,7 +491,7 @@ static struct pmac_i2c_host_kw *__init kw_i2c_host_init(struct device_node *np)
 	 * on all i2c keywest nodes so far ... we would have to fallback
 	 * to macio parsing if that wasn't the case
 	 */
-	addrp = (u32 *)get_property(np, "AAPL,address", NULL);
+	addrp = get_property(np, "AAPL,address", NULL);
 	if (addrp == NULL) {
 		printk(KERN_ERR "low_i2c: Can't find address for %s\n",
 		       np->full_name);
@@ -504,13 +505,13 @@ static struct pmac_i2c_host_kw *__init kw_i2c_host_init(struct device_node *np)
 	host->timeout_timer.function = kw_i2c_timeout;
 	host->timeout_timer.data = (unsigned long)host;
 
-	psteps = (u32 *)get_property(np, "AAPL,address-step", NULL);
+	psteps = get_property(np, "AAPL,address-step", NULL);
 	steps = psteps ? (*psteps) : 0x10;
 	for (host->bsteps = 0; (steps & 0x01) == 0; host->bsteps++)
 		steps >>= 1;
 	/* Select interface rate */
 	host->speed = KW_I2C_MODE_25KHZ;
-	prate = (u32 *)get_property(np, "AAPL,i2c-rate", NULL);
+	prate = get_property(np, "AAPL,i2c-rate", NULL);
 	if (prate) switch(*prate) {
 	case 100:
 		host->speed = KW_I2C_MODE_100KHZ;
@@ -618,8 +619,8 @@ static void __init kw_i2c_probe(void)
 		} else {
 			for (child = NULL;
 			     (child = of_get_next_child(np, child)) != NULL;) {
-				u32 *reg =
-					(u32 *)get_property(child, "reg", NULL);
+				const u32 *reg = get_property(child,
+						"reg", NULL);
 				if (reg == NULL)
 					continue;
 				kw_i2c_add(host, np, child, *reg);
@@ -881,7 +882,7 @@ static void __init smu_i2c_probe(void)
 {
 	struct device_node *controller, *busnode;
 	struct pmac_i2c_bus *bus;
-	u32 *reg;
+	const u32 *reg;
 	int sz;
 
 	if (!smu_present())
@@ -904,7 +905,7 @@ static void __init smu_i2c_probe(void)
 		if (strcmp(busnode->type, "i2c") &&
 		    strcmp(busnode->type, "i2c-bus"))
 			continue;
-		reg = (u32 *)get_property(busnode, "reg", NULL);
+		reg = get_property(busnode, "reg", NULL);
 		if (reg == NULL)
 			continue;
 
@@ -948,9 +949,8 @@ struct pmac_i2c_bus *pmac_i2c_find_bus(struct device_node *node)
 		list_for_each_entry(bus, &pmac_i2c_busses, link) {
 			if (p == bus->busnode) {
 				if (prev && bus->flags & pmac_i2c_multibus) {
-					u32 *reg;
-					reg = (u32 *)get_property(prev, "reg",
-								  NULL);
+					const u32 *reg;
+					reg = get_property(prev, "reg", NULL);
 					if (!reg)
 						continue;
 					if (((*reg) >> 8) != bus->channel)
@@ -971,7 +971,7 @@ EXPORT_SYMBOL_GPL(pmac_i2c_find_bus);
 
 u8 pmac_i2c_get_dev_addr(struct device_node *device)
 {
-	u32 *reg = (u32 *)get_property(device, "reg", NULL);
+	const u32 *reg = get_property(device, "reg", NULL);
 
 	if (reg == NULL)
 		return 0;
diff --git a/arch/powerpc/platforms/powermac/pci.c b/arch/powerpc/platforms/powermac/pci.c
index 556b349797e8..787ffd999bc2 100644
--- a/arch/powerpc/platforms/powermac/pci.c
+++ b/arch/powerpc/platforms/powermac/pci.c
@@ -69,16 +69,16 @@ struct device_node *k2_skiplist[2];
 static int __init fixup_one_level_bus_range(struct device_node *node, int higher)
 {
 	for (; node != 0;node = node->sibling) {
-		int * bus_range;
-		unsigned int *class_code;
+		const int * bus_range;
+		const unsigned int *class_code;
 		int len;
 
 		/* For PCI<->PCI bridges or CardBus bridges, we go down */
-		class_code = (unsigned int *) get_property(node, "class-code", NULL);
+		class_code = get_property(node, "class-code", NULL);
 		if (!class_code || ((*class_code >> 8) != PCI_CLASS_BRIDGE_PCI &&
 			(*class_code >> 8) != PCI_CLASS_BRIDGE_CARDBUS))
 			continue;
-		bus_range = (int *) get_property(node, "bus-range", &len);
+		bus_range = get_property(node, "bus-range", &len);
 		if (bus_range != NULL && len > 2 * sizeof(int)) {
 			if (bus_range[1] > higher)
 				higher = bus_range[1];
@@ -96,13 +96,15 @@ static int __init fixup_one_level_bus_range(struct device_node *node, int higher
  */
 static void __init fixup_bus_range(struct device_node *bridge)
 {
-	int * bus_range;
-	int len;
+	int *bus_range, len;
+	struct property *prop;
 
 	/* Lookup the "bus-range" property for the hose */
-	bus_range = (int *) get_property(bridge, "bus-range", &len);
-	if (bus_range == NULL || len < 2 * sizeof(int))
+	prop = of_find_property(bridge, "bus-range", &len);
+	if (prop == NULL || prop->length < 2 * sizeof(int))
 		return;
+
+	bus_range = (int *)prop->value;
 	bus_range[1] = fixup_one_level_bus_range(bridge->child, bus_range[1]);
 }
 
@@ -240,7 +242,7 @@ static struct pci_ops macrisc_pci_ops =
 static int chaos_validate_dev(struct pci_bus *bus, int devfn, int offset)
 {
 	struct device_node *np;
-	u32 *vendor, *device;
+	const u32 *vendor, *device;
 
 	if (offset >= 0x100)
 		return  PCIBIOS_BAD_REGISTER_NUMBER;
@@ -248,8 +250,8 @@ static int chaos_validate_dev(struct pci_bus *bus, int devfn, int offset)
 	if (np == NULL)
 		return PCIBIOS_DEVICE_NOT_FOUND;
 
-	vendor = (u32 *)get_property(np, "vendor-id", NULL);
-	device = (u32 *)get_property(np, "device-id", NULL);
+	vendor = get_property(np, "vendor-id", NULL);
+	device = get_property(np, "device-id", NULL);
 	if (vendor == NULL || device == NULL)
 		return PCIBIOS_DEVICE_NOT_FOUND;
 
@@ -689,20 +691,21 @@ static void __init fixup_nec_usb2(void)
 
 	for (nec = NULL; (nec = of_find_node_by_name(nec, "usb")) != NULL;) {
 		struct pci_controller *hose;
-		u32 data, *prop;
+		u32 data;
+		const u32 *prop;
 		u8 bus, devfn;
 
-		prop = (u32 *)get_property(nec, "vendor-id", NULL);
+		prop = get_property(nec, "vendor-id", NULL);
 		if (prop == NULL)
 			continue;
 		if (0x1033 != *prop)
 			continue;
-		prop = (u32 *)get_property(nec, "device-id", NULL);
+		prop = get_property(nec, "device-id", NULL);
 		if (prop == NULL)
 			continue;
 		if (0x0035 != *prop)
 			continue;
-		prop = (u32 *)get_property(nec, "reg", NULL);
+		prop = get_property(nec, "reg", NULL);
 		if (prop == NULL)
 			continue;
 		devfn = (prop[0] >> 8) & 0xff;
@@ -901,7 +904,7 @@ static int __init add_bridge(struct device_node *dev)
 	struct pci_controller *hose;
 	struct resource rsrc;
 	char *disp_name;
-	int *bus_range;
+	const int *bus_range;
 	int primary = 1, has_address = 0;
 
 	DBG("Adding PCI host bridge %s\n", dev->full_name);
@@ -910,7 +913,7 @@ static int __init add_bridge(struct device_node *dev)
 	has_address = (of_address_to_resource(dev, 0, &rsrc) == 0);
 
 	/* Get bus range if any */
-	bus_range = (int *) get_property(dev, "bus-range", &len);
+	bus_range = get_property(dev, "bus-range", &len);
 	if (bus_range == NULL || len < 2 * sizeof(int)) {
 		printk(KERN_WARNING "Can't get bus-range for %s, assume"
 		       " bus 0\n", dev->full_name);
diff --git a/arch/powerpc/platforms/powermac/pfunc_base.c b/arch/powerpc/platforms/powermac/pfunc_base.c
index 6d66359ec8c8..829dacec96e5 100644
--- a/arch/powerpc/platforms/powermac/pfunc_base.c
+++ b/arch/powerpc/platforms/powermac/pfunc_base.c
@@ -114,7 +114,7 @@ static void macio_gpio_init_one(struct macio_chip *macio)
 	 * we just create them all
 	 */
 	for (gp = NULL; (gp = of_get_next_child(gparent, gp)) != NULL;) {
-		u32 *reg = (u32 *)get_property(gp, "reg", NULL);
+		const u32 *reg = get_property(gp, "reg", NULL);
 		unsigned long offset;
 		if (reg == NULL)
 			continue;
diff --git a/arch/powerpc/platforms/powermac/pfunc_core.c b/arch/powerpc/platforms/powermac/pfunc_core.c
index b117adbf9571..7651f278615a 100644
--- a/arch/powerpc/platforms/powermac/pfunc_core.c
+++ b/arch/powerpc/platforms/powermac/pfunc_core.c
@@ -813,14 +813,15 @@ struct pmf_function *__pmf_find_function(struct device_node *target,
 	struct pmf_device *dev;
 	struct pmf_function *func, *result = NULL;
 	char fname[64];
-	u32 *prop, ph;
+	const u32 *prop;
+	u32 ph;
 
 	/*
 	 * Look for a "platform-*" function reference. If we can't find
 	 * one, then we fallback to a direct call attempt
 	 */
 	snprintf(fname, 63, "platform-%s", name);
-	prop = (u32 *)get_property(target, fname, NULL);
+	prop = get_property(target, fname, NULL);
 	if (prop == NULL)
 		goto find_it;
 	ph = *prop;
diff --git a/arch/powerpc/platforms/powermac/setup.c b/arch/powerpc/platforms/powermac/setup.c
index 31a9da769fa2..824a618396ab 100644
--- a/arch/powerpc/platforms/powermac/setup.c
+++ b/arch/powerpc/platforms/powermac/setup.c
@@ -116,7 +116,7 @@ extern struct smp_ops_t core99_smp_ops;
 static void pmac_show_cpuinfo(struct seq_file *m)
 {
 	struct device_node *np;
-	char *pp;
+	const char *pp;
 	int plen;
 	int mbmodel;
 	unsigned int mbflags;
@@ -134,12 +134,12 @@ static void pmac_show_cpuinfo(struct seq_file *m)
 	seq_printf(m, "machine\t\t: ");
 	np = of_find_node_by_path("/");
 	if (np != NULL) {
-		pp = (char *) get_property(np, "model", NULL);
+		pp = get_property(np, "model", NULL);
 		if (pp != NULL)
 			seq_printf(m, "%s\n", pp);
 		else
 			seq_printf(m, "PowerMac\n");
-		pp = (char *) get_property(np, "compatible", &plen);
+		pp = get_property(np, "compatible", &plen);
 		if (pp != NULL) {
 			seq_printf(m, "motherboard\t:");
 			while (plen > 0) {
@@ -163,10 +163,8 @@ static void pmac_show_cpuinfo(struct seq_file *m)
 	if (np == NULL)
 		np = of_find_node_by_type(NULL, "cache");
 	if (np != NULL) {
-		unsigned int *ic = (unsigned int *)
-			get_property(np, "i-cache-size", NULL);
-		unsigned int *dc = (unsigned int *)
-			get_property(np, "d-cache-size", NULL);
+		const unsigned int *ic = get_property(np, "i-cache-size", NULL);
+		const unsigned int *dc = get_property(np, "d-cache-size", NULL);
 		seq_printf(m, "L2 cache\t:");
 		has_l2cache = 1;
 		if (get_property(np, "cache-unified", NULL) != 0 && dc) {
@@ -254,7 +252,7 @@ static void __init l2cr_init(void)
 		if (np == 0)
 			np = find_type_devices("cpu");
 		if (np != 0) {
-			unsigned int *l2cr = (unsigned int *)
+			const unsigned int *l2cr =
 				get_property(np, "l2cr-value", NULL);
 			if (l2cr != 0) {
 				ppc_override_l2cr = 1;
@@ -277,7 +275,7 @@ static void __init l2cr_init(void)
 static void __init pmac_setup_arch(void)
 {
 	struct device_node *cpu, *ic;
-	int *fp;
+	const int *fp;
 	unsigned long pvr;
 
 	pvr = PVR_VER(mfspr(SPRN_PVR));
@@ -287,7 +285,7 @@ static void __init pmac_setup_arch(void)
 	loops_per_jiffy = 50000000 / HZ;
 	cpu = of_find_node_by_type(NULL, "cpu");
 	if (cpu != NULL) {
-		fp = (int *) get_property(cpu, "clock-frequency", NULL);
+		fp = get_property(cpu, "clock-frequency", NULL);
 		if (fp != NULL) {
 			if (pvr >= 0x30 && pvr < 0x80)
 				/* PPC970 etc. */
diff --git a/arch/powerpc/platforms/powermac/smp.c b/arch/powerpc/platforms/powermac/smp.c
index 827b7121ffb8..653eeb64d1e2 100644
--- a/arch/powerpc/platforms/powermac/smp.c
+++ b/arch/powerpc/platforms/powermac/smp.c
@@ -548,7 +548,7 @@ static void __init smp_core99_setup_i2c_hwsync(int ncpus)
 	struct device_node *cc = NULL;	
 	struct device_node *p;
 	const char *name = NULL;
-	u32 *reg;
+	const u32 *reg;
 	int ok;
 
 	/* Look for the clock chip */
@@ -562,7 +562,7 @@ static void __init smp_core99_setup_i2c_hwsync(int ncpus)
 		pmac_tb_clock_chip_host = pmac_i2c_find_bus(cc);
 		if (pmac_tb_clock_chip_host == NULL)
 			continue;
-		reg = (u32 *)get_property(cc, "reg", NULL);
+		reg = get_property(cc, "reg", NULL);
 		if (reg == NULL)
 			continue;
 		switch (*reg) {
@@ -707,8 +707,7 @@ static void __init smp_core99_setup(int ncpus)
 		core99_tb_gpio = KL_GPIO_TB_ENABLE;	/* default value */
 		cpu = of_find_node_by_type(NULL, "cpu");
 		if (cpu != NULL) {
-			tbprop = (u32 *)get_property(cpu, "timebase-enable",
-						     NULL);
+			tbprop = get_property(cpu, "timebase-enable", NULL);
 			if (tbprop)
 				core99_tb_gpio = *tbprop;
 			of_node_put(cpu);
diff --git a/arch/powerpc/platforms/powermac/udbg_scc.c b/arch/powerpc/platforms/powermac/udbg_scc.c
index 37e5b1eff911..ce1a235855f7 100644
--- a/arch/powerpc/platforms/powermac/udbg_scc.c
+++ b/arch/powerpc/platforms/powermac/udbg_scc.c
@@ -68,11 +68,11 @@ static unsigned char scc_inittab[] = {
 
 void udbg_scc_init(int force_scc)
 {
-	u32 *reg;
+	const u32 *reg;
 	unsigned long addr;
 	struct device_node *stdout = NULL, *escc = NULL, *macio = NULL;
 	struct device_node *ch, *ch_def = NULL, *ch_a = NULL;
-	char *path;
+	const char *path;
 	int i, x;
 
 	escc = of_find_node_by_name(NULL, "escc");
@@ -81,7 +81,7 @@ void udbg_scc_init(int force_scc)
 	macio = of_get_parent(escc);
 	if (macio == NULL)
 		goto bail;
-	path = (char *)get_property(of_chosen, "linux,stdout-path", NULL);
+	path = get_property(of_chosen, "linux,stdout-path", NULL);
 	if (path != NULL)
 		stdout = of_find_node_by_path(path);
 	for (ch = NULL; (ch = of_get_next_child(escc, ch)) != NULL;) {
@@ -96,13 +96,13 @@ void udbg_scc_init(int force_scc)
 	ch = ch_def ? ch_def : ch_a;
 
 	/* Get address within mac-io ASIC */
-	reg = (u32 *)get_property(escc, "reg", NULL);
+	reg = get_property(escc, "reg", NULL);
 	if (reg == NULL)
 		goto bail;
 	addr = reg[0];
 
 	/* Get address of mac-io PCI itself */
-	reg = (u32 *)get_property(macio, "assigned-addresses", NULL);
+	reg = get_property(macio, "assigned-addresses", NULL);
 	if (reg == NULL)
 		goto bail;
 	addr += reg[2];
diff --git a/drivers/i2c/busses/i2c-powermac.c b/drivers/i2c/busses/i2c-powermac.c
index 2a0b3be7cdd0..e8a6de5a1517 100644
--- a/drivers/i2c/busses/i2c-powermac.c
+++ b/drivers/i2c/busses/i2c-powermac.c
@@ -209,7 +209,8 @@ static int i2c_powermac_probe(struct device *dev)
 	struct pmac_i2c_bus *bus = dev->platform_data;
 	struct device_node *parent = NULL;
 	struct i2c_adapter *adapter;
-	char name[32], *basename;
+	char name[32];
+	const char *basename;
 	int rc;
 
 	if (bus == NULL)
diff --git a/drivers/ide/ppc/pmac.c b/drivers/ide/ppc/pmac.c
index ebf961f1718d..fa46856e8068 100644
--- a/drivers/ide/ppc/pmac.c
+++ b/drivers/ide/ppc/pmac.c
@@ -1154,7 +1154,7 @@ static int
 pmac_ide_setup_device(pmac_ide_hwif_t *pmif, ide_hwif_t *hwif)
 {
 	struct device_node *np = pmif->node;
-	int *bidp;
+	const int *bidp;
 
 	pmif->cable_80 = 0;
 	pmif->broken_dma = pmif->broken_dma_warn = 0;
@@ -1176,14 +1176,14 @@ pmac_ide_setup_device(pmac_ide_hwif_t *pmif, ide_hwif_t *hwif)
 		pmif->broken_dma = 1;
 	}
 
-	bidp = (int *)get_property(np, "AAPL,bus-id", NULL);
+	bidp = get_property(np, "AAPL,bus-id", NULL);
 	pmif->aapl_bus_id =  bidp ? *bidp : 0;
 
 	/* Get cable type from device-tree */
 	if (pmif->kind == controller_kl_ata4 || pmif->kind == controller_un_ata6
 	    || pmif->kind == controller_k2_ata6
 	    || pmif->kind == controller_sh_ata6) {
-		char* cable = get_property(np, "cable-type", NULL);
+		const char* cable = get_property(np, "cable-type", NULL);
 		if (cable && !strncmp(cable, "80-", 3))
 			pmif->cable_80 = 1;
 	}
diff --git a/drivers/macintosh/macio_asic.c b/drivers/macintosh/macio_asic.c
index 80c0c665b5f6..7817cf286d0c 100644
--- a/drivers/macintosh/macio_asic.c
+++ b/drivers/macintosh/macio_asic.c
@@ -139,7 +139,9 @@ static int macio_uevent(struct device *dev, char **envp, int num_envp,
 {
 	struct macio_dev * macio_dev;
 	struct of_device * of;
-	char *scratch, *compat, *compat2;
+	char *scratch;
+	const char *compat, *compat2;
+
 	int i = 0;
 	int length, cplen, cplen2, seen = 0;
 
@@ -173,7 +175,7 @@ static int macio_uevent(struct device *dev, char **envp, int num_envp,
          * it's not really legal to split it out with commas. We split it
          * up using a number of environment variables instead. */
 
-	compat = (char *) get_property(of->node, "compatible", &cplen);
+	compat = get_property(of->node, "compatible", &cplen);
 	compat2 = compat;
 	cplen2= cplen;
 	while (compat && cplen > 0) {
@@ -454,7 +456,7 @@ static struct macio_dev * macio_add_one_device(struct macio_chip *chip,
 					       struct resource *parent_res)
 {
 	struct macio_dev *dev;
-	u32 *reg;
+	const u32 *reg;
 	
 	if (np == NULL)
 		return NULL;
@@ -489,7 +491,7 @@ static struct macio_dev * macio_add_one_device(struct macio_chip *chip,
 #endif
 			MAX_NODE_NAME_SIZE, np->name);
 	} else {
-		reg = (u32 *)get_property(np, "reg", NULL);
+		reg = get_property(np, "reg", NULL);
 		sprintf(dev->ofdev.dev.bus_id, "%1d.%08x:%.*s",
 			chip->lbus.index,
 			reg ? *reg : 0, MAX_NODE_NAME_SIZE, np->name);
diff --git a/drivers/macintosh/macio_sysfs.c b/drivers/macintosh/macio_sysfs.c
index cae24a13526a..8566bdfdd4b8 100644
--- a/drivers/macintosh/macio_sysfs.c
+++ b/drivers/macintosh/macio_sysfs.c
@@ -16,12 +16,12 @@ static ssize_t
 compatible_show (struct device *dev, struct device_attribute *attr, char *buf)
 {
 	struct of_device *of;
-	char *compat;
+	const char *compat;
 	int cplen;
 	int length = 0;
 
 	of = &to_macio_device (dev)->ofdev;
-	compat = (char *) get_property(of->node, "compatible", &cplen);
+	compat = get_property(of->node, "compatible", &cplen);
 	if (!compat) {
 		*buf = '\0';
 		return 0;
@@ -42,12 +42,12 @@ static ssize_t modalias_show (struct device *dev, struct device_attribute *attr,
 			      char *buf)
 {
 	struct of_device *of;
-	char *compat;
+	const char *compat;
 	int cplen;
 	int length;
 
 	of = &to_macio_device (dev)->ofdev;
-	compat = (char *) get_property (of->node, "compatible", &cplen);
+	compat = get_property(of->node, "compatible", &cplen);
 	if (!compat) compat = "", cplen = 1;
 	length = sprintf (buf, "of:N%sT%s", of->node->name, of->node->type);
 	buf += length;
diff --git a/drivers/macintosh/smu.c b/drivers/macintosh/smu.c
index f139a74696fe..6f358600536e 100644
--- a/drivers/macintosh/smu.c
+++ b/drivers/macintosh/smu.c
@@ -447,7 +447,7 @@ EXPORT_SYMBOL(smu_present);
 int __init smu_init (void)
 {
 	struct device_node *np;
-	u32 *data;
+	const u32 *data;
 
         np = of_find_node_by_type(NULL, "smu");
         if (np == NULL)
@@ -483,7 +483,7 @@ int __init smu_init (void)
 		printk(KERN_ERR "SMU: Can't find doorbell GPIO !\n");
 		goto fail;
 	}
-	data = (u32 *)get_property(np, "reg", NULL);
+	data = get_property(np, "reg", NULL);
 	if (data == NULL) {
 		of_node_put(np);
 		printk(KERN_ERR "SMU: Can't find doorbell GPIO address !\n");
@@ -506,7 +506,7 @@ int __init smu_init (void)
 		np = of_find_node_by_name(NULL, "smu-interrupt");
 		if (np == NULL)
 			break;
-		data = (u32 *)get_property(np, "reg", NULL);
+		data = get_property(np, "reg", NULL);
 		if (data == NULL) {
 			of_node_put(np);
 			break;
@@ -959,11 +959,11 @@ static struct smu_sdbp_header *smu_create_sdb_partition(int id)
 /* Note: Only allowed to return error code in pointers (using ERR_PTR)
  * when interruptible is 1
  */
-struct smu_sdbp_header *__smu_get_sdb_partition(int id, unsigned int *size,
-						int interruptible)
+const struct smu_sdbp_header *__smu_get_sdb_partition(int id,
+		unsigned int *size, int interruptible)
 {
 	char pname[32];
-	struct smu_sdbp_header *part;
+	const struct smu_sdbp_header *part;
 
 	if (!smu)
 		return NULL;
@@ -980,8 +980,7 @@ struct smu_sdbp_header *__smu_get_sdb_partition(int id, unsigned int *size,
 	} else
 		mutex_lock(&smu_part_access);
 
-	part = (struct smu_sdbp_header *)get_property(smu->of_node,
-						      pname, size);
+	part = get_property(smu->of_node, pname, size);
 	if (part == NULL) {
 		DPRINTK("trying to extract from SMU ...\n");
 		part = smu_create_sdb_partition(id);
@@ -992,7 +991,7 @@ struct smu_sdbp_header *__smu_get_sdb_partition(int id, unsigned int *size,
 	return part;
 }
 
-struct smu_sdbp_header *smu_get_sdb_partition(int id, unsigned int *size)
+const struct smu_sdbp_header *smu_get_sdb_partition(int id, unsigned int *size)
 {
 	return __smu_get_sdb_partition(id, size, 0);
 }
@@ -1071,7 +1070,7 @@ static ssize_t smu_write(struct file *file, const char __user *buf,
 		pp->mode = smu_file_events;
 		return 0;
 	} else if (hdr.cmdtype == SMU_CMDTYPE_GET_PARTITION) {
-		struct smu_sdbp_header *part;
+		const struct smu_sdbp_header *part;
 		part = __smu_get_sdb_partition(hdr.cmd, NULL, 1);
 		if (part == NULL)
 			return -EINVAL;
diff --git a/drivers/macintosh/therm_adt746x.c b/drivers/macintosh/therm_adt746x.c
index 7f86478bdd36..a0f30d0853ea 100644
--- a/drivers/macintosh/therm_adt746x.c
+++ b/drivers/macintosh/therm_adt746x.c
@@ -47,7 +47,7 @@ static u8 FAN_SPD_SET[2] = {0x30, 0x31};
 
 static u8 default_limits_local[3] = {70, 50, 70};    /* local, sensor1, sensor2 */
 static u8 default_limits_chip[3] = {80, 65, 80};    /* local, sensor1, sensor2 */
-static char *sensor_location[3] = {NULL, NULL, NULL};
+static const char *sensor_location[3] = {NULL, NULL, NULL};
 
 static int limit_adjust = 0;
 static int fan_speed = -1;
@@ -553,7 +553,7 @@ static int __init
 thermostat_init(void)
 {
 	struct device_node* np;
-	u32 *prop;
+	const u32 *prop;
 	int i = 0, offset = 0;
 	
 	np = of_find_node_by_name(NULL, "fan");
@@ -566,13 +566,13 @@ thermostat_init(void)
 	else
 		return -ENODEV;
 
-	prop = (u32 *)get_property(np, "hwsensor-params-version", NULL);
+	prop = get_property(np, "hwsensor-params-version", NULL);
 	printk(KERN_INFO "adt746x: version %d (%ssupported)\n", *prop,
 			 (*prop == 1)?"":"un");
 	if (*prop != 1)
 		return -ENODEV;
 
-	prop = (u32 *)get_property(np, "reg", NULL);
+	prop = get_property(np, "reg", NULL);
 	if (!prop)
 		return -ENODEV;
 
diff --git a/drivers/macintosh/therm_pm72.c b/drivers/macintosh/therm_pm72.c
index 20bf67244e2c..d00c0c37a12e 100644
--- a/drivers/macintosh/therm_pm72.c
+++ b/drivers/macintosh/therm_pm72.c
@@ -660,7 +660,7 @@ static int read_eeprom(int cpu, struct mpu_data *out)
 {
 	struct device_node *np;
 	char nodename[64];
-	u8 *data;
+	const u8 *data;
 	int len;
 
 	/* prom.c routine for finding a node by path is a bit brain dead
@@ -673,7 +673,7 @@ static int read_eeprom(int cpu, struct mpu_data *out)
 		printk(KERN_ERR "therm_pm72: Failed to retrieve cpuid node from device-tree\n");
 		return -ENODEV;
 	}
-	data = (u8 *)get_property(np, "cpuid", &len);
+	data = get_property(np, "cpuid", &len);
 	if (data == NULL) {
 		printk(KERN_ERR "therm_pm72: Failed to retrieve cpuid property from device-tree\n");
 		of_node_put(np);
@@ -1336,7 +1336,7 @@ static int init_backside_state(struct backside_pid_state *state)
 	 */
 	u3 = of_find_node_by_path("/u3@0,f8000000");
 	if (u3 != NULL) {
-		u32 *vers = (u32 *)get_property(u3, "device-rev", NULL);
+		const u32 *vers = get_property(u3, "device-rev", NULL);
 		if (vers)
 			if (((*vers) & 0x3f) < 0x34)
 				u3h = 0;
@@ -2111,8 +2111,8 @@ static void fcu_lookup_fans(struct device_node *fcu_node)
 
 	while ((np = of_get_next_child(fcu_node, np)) != NULL) {
 		int type = -1;
-		char *loc;
-		u32 *reg;
+		const char *loc;
+		const u32 *reg;
 
 		DBG(" control: %s, type: %s\n", np->name, np->type);
 
@@ -2128,8 +2128,8 @@ static void fcu_lookup_fans(struct device_node *fcu_node)
 			continue;
 
 		/* Lookup for a matching location */
-		loc = (char *)get_property(np, "location", NULL);
-		reg = (u32 *)get_property(np, "reg", NULL);
+		loc = get_property(np, "location", NULL);
+		reg = get_property(np, "reg", NULL);
 		if (loc == NULL || reg == NULL)
 			continue;
 		DBG(" matching location: %s, reg: 0x%08x\n", loc, *reg);
diff --git a/drivers/macintosh/therm_windtunnel.c b/drivers/macintosh/therm_windtunnel.c
index c7d1c290cb0c..738faab1b22c 100644
--- a/drivers/macintosh/therm_windtunnel.c
+++ b/drivers/macintosh/therm_windtunnel.c
@@ -484,14 +484,14 @@ struct apple_thermal_info {
 static int __init
 g4fan_init( void )
 {
-	struct apple_thermal_info *info;
+	const struct apple_thermal_info *info;
 	struct device_node *np;
 
 	init_MUTEX( &x.lock );
 
 	if( !(np=of_find_node_by_name(NULL, "power-mgt")) )
 		return -ENODEV;
-	info = (struct apple_thermal_info*)get_property(np, "thermal-info", NULL);
+	info = get_property(np, "thermal-info", NULL);
 	of_node_put(np);
 
 	if( !info || !machine_is_compatible("PowerMac3,6") )
diff --git a/drivers/macintosh/via-cuda.c b/drivers/macintosh/via-cuda.c
index 69d5452fd22f..7512d1c15207 100644
--- a/drivers/macintosh/via-cuda.c
+++ b/drivers/macintosh/via-cuda.c
@@ -123,7 +123,7 @@ int __init find_via_cuda(void)
 {
     struct adb_request req;
     phys_addr_t taddr;
-    u32 *reg;
+    const u32 *reg;
     int err;
 
     if (vias != 0)
@@ -132,7 +132,7 @@ int __init find_via_cuda(void)
     if (vias == 0)
 	return 0;
 
-    reg = (u32 *)get_property(vias, "reg", NULL);
+    reg = get_property(vias, "reg", NULL);
     if (reg == NULL) {
 	    printk(KERN_ERR "via-cuda: No \"reg\" property !\n");
 	    goto fail;
diff --git a/drivers/macintosh/via-pmu-led.c b/drivers/macintosh/via-pmu-led.c
index 5189d5454b1f..179af10105d9 100644
--- a/drivers/macintosh/via-pmu-led.c
+++ b/drivers/macintosh/via-pmu-led.c
@@ -120,7 +120,7 @@ static int __init via_pmu_led_init(void)
 	dt = of_find_node_by_path("/");
 	if (dt == NULL)
 		return -ENODEV;
-	model = (const char *)get_property(dt, "model", NULL);
+	model = get_property(dt, "model", NULL);
 	if (model == NULL)
 		return -ENODEV;
 	if (strncmp(model, "PowerBook", strlen("PowerBook")) != 0 &&
diff --git a/drivers/macintosh/via-pmu.c b/drivers/macintosh/via-pmu.c
index 06ca80bfd6b9..80e88b4a1d07 100644
--- a/drivers/macintosh/via-pmu.c
+++ b/drivers/macintosh/via-pmu.c
@@ -287,7 +287,7 @@ static char *pbook_type[] = {
 int __init find_via_pmu(void)
 {
 	u64 taddr;
-	u32 *reg;
+	const u32 *reg;
 
 	if (via != 0)
 		return 1;
@@ -295,7 +295,7 @@ int __init find_via_pmu(void)
 	if (vias == NULL)
 		return 0;
 
-	reg = (u32 *)get_property(vias, "reg", NULL);
+	reg = get_property(vias, "reg", NULL);
 	if (reg == NULL) {
 		printk(KERN_ERR "via-pmu: No \"reg\" property !\n");
 		goto fail;
@@ -337,7 +337,7 @@ int __init find_via_pmu(void)
 		
 		gpiop = of_find_node_by_name(NULL, "gpio");
 		if (gpiop) {
-			reg = (u32 *)get_property(gpiop, "reg", NULL);
+			reg = get_property(gpiop, "reg", NULL);
 			if (reg)
 				gaddr = of_translate_address(gpiop, reg);
 			if (gaddr != OF_BAD_ADDR)
@@ -486,9 +486,9 @@ static int __init via_pmu_dev_init(void)
 		pmu_batteries[1].flags |= PMU_BATT_TYPE_SMART;
 	} else {
 		struct device_node* prim = find_devices("power-mgt");
-		u32 *prim_info = NULL;
+		const u32 *prim_info = NULL;
 		if (prim)
-			prim_info = (u32 *)get_property(prim, "prim-info", NULL);
+			prim_info = get_property(prim, "prim-info", NULL);
 		if (prim_info) {
 			/* Other stuffs here yet unknown */
 			pmu_battery_count = (prim_info[6] >> 16) & 0xff;
diff --git a/drivers/macintosh/windfarm_pm81.c b/drivers/macintosh/windfarm_pm81.c
index f1df6efcbe68..2ff546e4c92f 100644
--- a/drivers/macintosh/windfarm_pm81.c
+++ b/drivers/macintosh/windfarm_pm81.c
@@ -396,7 +396,7 @@ static void wf_smu_sys_fans_tick(struct wf_smu_sys_fans_state *st)
 static void wf_smu_create_cpu_fans(void)
 {
 	struct wf_cpu_pid_param pid_param;
-	struct smu_sdbp_header *hdr;
+	const struct smu_sdbp_header *hdr;
 	struct smu_sdbp_cpupiddata *piddata;
 	struct smu_sdbp_fvt *fvt;
 	s32 tmax, tdelta, maxpow, powadj;
@@ -702,7 +702,7 @@ static struct notifier_block wf_smu_events = {
 
 static int wf_init_pm(void)
 {
-	struct smu_sdbp_header *hdr;
+	const struct smu_sdbp_header *hdr;
 
 	hdr = smu_get_sdb_partition(SMU_SDB_SENSORTREE_ID, NULL);
 	if (hdr != 0) {
diff --git a/drivers/macintosh/windfarm_pm91.c b/drivers/macintosh/windfarm_pm91.c
index 0d6372e96d32..59e9ffe37c39 100644
--- a/drivers/macintosh/windfarm_pm91.c
+++ b/drivers/macintosh/windfarm_pm91.c
@@ -144,7 +144,7 @@ static struct wf_smu_slots_fans_state *wf_smu_slots_fans;
 static void wf_smu_create_cpu_fans(void)
 {
 	struct wf_cpu_pid_param pid_param;
-	struct smu_sdbp_header *hdr;
+	const struct smu_sdbp_header *hdr;
 	struct smu_sdbp_cpupiddata *piddata;
 	struct smu_sdbp_fvt *fvt;
 	s32 tmax, tdelta, maxpow, powadj;
diff --git a/drivers/macintosh/windfarm_smu_controls.c b/drivers/macintosh/windfarm_smu_controls.c
index a9e88edc0c72..bff1f372f188 100644
--- a/drivers/macintosh/windfarm_smu_controls.c
+++ b/drivers/macintosh/windfarm_smu_controls.c
@@ -159,14 +159,15 @@ static struct smu_fan_control *smu_fan_create(struct device_node *node,
 					      int pwm_fan)
 {
 	struct smu_fan_control *fct;
-	s32 *v; u32 *reg;
-	char *l;
+	const s32 *v;
+	const u32 *reg;
+	const char *l;
 
 	fct = kmalloc(sizeof(struct smu_fan_control), GFP_KERNEL);
 	if (fct == NULL)
 		return NULL;
 	fct->ctrl.ops = &smu_fan_ops;
-	l = (char *)get_property(node, "location", NULL);
+	l = get_property(node, "location", NULL);
 	if (l == NULL)
 		goto fail;
 
@@ -223,17 +224,17 @@ static struct smu_fan_control *smu_fan_create(struct device_node *node,
 		goto fail;
 
 	/* Get min & max values*/
-	v = (s32 *)get_property(node, "min-value", NULL);
+	v = get_property(node, "min-value", NULL);
 	if (v == NULL)
 		goto fail;
 	fct->min = *v;
-	v = (s32 *)get_property(node, "max-value", NULL);
+	v = get_property(node, "max-value", NULL);
 	if (v == NULL)
 		goto fail;
 	fct->max = *v;
 
 	/* Get "reg" value */
-	reg = (u32 *)get_property(node, "reg", NULL);
+	reg = get_property(node, "reg", NULL);
 	if (reg == NULL)
 		goto fail;
 	fct->reg = *reg;
diff --git a/drivers/macintosh/windfarm_smu_sat.c b/drivers/macintosh/windfarm_smu_sat.c
index e295a07a1ebc..aceb61d9fbc8 100644
--- a/drivers/macintosh/windfarm_smu_sat.c
+++ b/drivers/macintosh/windfarm_smu_sat.c
@@ -233,15 +233,15 @@ static void wf_sat_create(struct i2c_adapter *adapter, struct device_node *dev)
 {
 	struct wf_sat *sat;
 	struct wf_sat_sensor *sens;
-	u32 *reg;
-	char *loc, *type;
+	const u32 *reg;
+	const char *loc, *type;
 	u8 addr, chip, core;
 	struct device_node *child;
 	int shift, cpu, index;
 	char *name;
 	int vsens[2], isens[2];
 
-	reg = (u32 *) get_property(dev, "reg", NULL);
+	reg = get_property(dev, "reg", NULL);
 	if (reg == NULL)
 		return;
 	addr = *reg;
@@ -268,7 +268,7 @@ static void wf_sat_create(struct i2c_adapter *adapter, struct device_node *dev)
 	isens[0] = isens[1] = -1;
 	child = NULL;
 	while ((child = of_get_next_child(dev, child)) != NULL) {
-		reg = (u32 *) get_property(child, "reg", NULL);
+		reg = get_property(child, "reg", NULL);
 		type = get_property(child, "device_type", NULL);
 		loc = get_property(child, "location", NULL);
 		if (reg == NULL || loc == NULL)
diff --git a/drivers/macintosh/windfarm_smu_sensors.c b/drivers/macintosh/windfarm_smu_sensors.c
index bed25dcf8a1e..defe9922ebd1 100644
--- a/drivers/macintosh/windfarm_smu_sensors.c
+++ b/drivers/macintosh/windfarm_smu_sensors.c
@@ -198,14 +198,14 @@ static struct wf_sensor_ops smu_slotspow_ops = {
 static struct smu_ad_sensor *smu_ads_create(struct device_node *node)
 {
 	struct smu_ad_sensor *ads;
-	char *c, *l;
-	u32 *v;
+	const char *c, *l;
+	const u32 *v;
 
 	ads = kmalloc(sizeof(struct smu_ad_sensor), GFP_KERNEL);
 	if (ads == NULL)
 		return NULL;
-	c = (char *)get_property(node, "device_type", NULL);
-	l = (char *)get_property(node, "location", NULL);
+	c = get_property(node, "device_type", NULL);
+	l = get_property(node, "location", NULL);
 	if (c == NULL || l == NULL)
 		goto fail;
 
@@ -255,7 +255,7 @@ static struct smu_ad_sensor *smu_ads_create(struct device_node *node)
 	} else
 		goto fail;
 
-	v = (u32 *)get_property(node, "reg", NULL);
+	v = get_property(node, "reg", NULL);
 	if (v == NULL)
 		goto fail;
 	ads->reg = *v;
@@ -382,7 +382,7 @@ smu_cpu_power_create(struct wf_sensor *volts, struct wf_sensor *amps)
 
 static void smu_fetch_param_partitions(void)
 {
-	struct smu_sdbp_header *hdr;
+	const struct smu_sdbp_header *hdr;
 
 	/* Get CPU voltage/current/power calibration data */
 	hdr = smu_get_sdb_partition(SMU_SDB_CPUVCP_ID, NULL);
diff --git a/drivers/serial/pmac_zilog.c b/drivers/serial/pmac_zilog.c
index bfd2a22759eb..a3b99caf80e6 100644
--- a/drivers/serial/pmac_zilog.c
+++ b/drivers/serial/pmac_zilog.c
@@ -1400,8 +1400,8 @@ static struct uart_ops pmz_pops = {
 static int __init pmz_init_port(struct uart_pmac_port *uap)
 {
 	struct device_node *np = uap->node;
-	char *conn;
-	struct slot_names_prop {
+	const char *conn;
+	const struct slot_names_prop {
 		int	count;
 		char	name[1];
 	} *slots;
@@ -1458,7 +1458,7 @@ no_dma:
 		uap->flags |= PMACZILOG_FLAG_IS_IRDA;
 	uap->port_type = PMAC_SCC_ASYNC;
 	/* 1999 Powerbook G3 has slot-names property instead */
-	slots = (struct slot_names_prop *)get_property(np, "slot-names", &len);
+	slots = get_property(np, "slot-names", &len);
 	if (slots && slots->count > 0) {
 		if (strcmp(slots->name, "IrDA") == 0)
 			uap->flags |= PMACZILOG_FLAG_IS_IRDA;
@@ -1470,7 +1470,8 @@ no_dma:
 	if (ZS_IS_INTMODEM(uap)) {
 		struct device_node* i2c_modem = find_devices("i2c-modem");
 		if (i2c_modem) {
-			char* mid = get_property(i2c_modem, "modem-id", NULL);
+			const char* mid =
+				get_property(i2c_modem, "modem-id", NULL);
 			if (mid) switch(*mid) {
 			case 0x04 :
 			case 0x05 :
diff --git a/include/asm-powerpc/smu.h b/include/asm-powerpc/smu.h
index 51e65fc46a03..e49f644ca63a 100644
--- a/include/asm-powerpc/smu.h
+++ b/include/asm-powerpc/smu.h
@@ -517,7 +517,7 @@ struct smu_sdbp_cpupiddata {
  * This returns the pointer to an SMU "sdb" partition data or NULL
  * if not found. The data format is described below
  */
-extern struct smu_sdbp_header *smu_get_sdb_partition(int id,
+extern const struct smu_sdbp_header *smu_get_sdb_partition(int id,
 					unsigned int *size);
 
 /* Get "sdb" partition data from an SMU satellite */
-- 
cgit v1.2.3


From 931b261f442e779b0656d9b04c7ffe4939ef8c0a Mon Sep 17 00:00:00 2001
From: Jeremy Kerr <jk@ozlabs.org>
Date: Wed, 12 Jul 2006 15:42:06 +1000
Subject: [POWERPC] Make get_property() return a const void *

Previous changes have treated the return values of get_property as
const, so now we can make the actual change to get_property(). There
shouldn't be a need to cast the return values anymore.

We will now get compiler warnings when property values are assigned to
a non-const variable.

If properties need to be updated, there's still the of_find_property
function.

Built for cell_defconfig, chrp32_defconfig, g5_defconfig,
iseries_defconfig, maple_defconfig, pmac32_defconfig, ppc64_defconfig
and pseries_defconfig.

Signed-off-by: Jeremy Kerr <jk@ozlabs.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 arch/powerpc/kernel/prom.c | 2 +-
 include/asm-powerpc/prom.h | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c
index 2a3d84a39cb5..bf2005b2feb6 100644
--- a/arch/powerpc/kernel/prom.c
+++ b/arch/powerpc/kernel/prom.c
@@ -1527,7 +1527,7 @@ struct property *of_find_property(struct device_node *np, const char *name,
  * Find a property with a given name for a given node
  * and return the value.
  */
-void *get_property(struct device_node *np, const char *name, int *lenp)
+const void *get_property(struct device_node *np, const char *name, int *lenp)
 {
 	struct property *pp = of_find_property(np,name,lenp);
 	return pp ? pp->value : NULL;
diff --git a/include/asm-powerpc/prom.h b/include/asm-powerpc/prom.h
index abdf1be66e97..31bfea4686a6 100644
--- a/include/asm-powerpc/prom.h
+++ b/include/asm-powerpc/prom.h
@@ -160,7 +160,7 @@ extern void unflatten_device_tree(void);
 extern void early_init_devtree(void *);
 extern int device_is_compatible(struct device_node *device, const char *);
 extern int machine_is_compatible(const char *compat);
-extern void *get_property(struct device_node *node, const char *name,
+extern const void *get_property(struct device_node *node, const char *name,
 		int *lenp);
 extern void print_properties(struct device_node *node);
 extern int prom_n_addr_cells(struct device_node* np);
-- 
cgit v1.2.3


From cb18bd40030c879cd93fef02fd579f74dbab473d Mon Sep 17 00:00:00 2001
From: Mike Kravetz <kravetz@us.ibm.com>
Date: Thu, 20 Jul 2006 23:39:51 -0700
Subject: [POWERPC] Instrument Hypervisor Calls: merge headers

Move all the Hypervisor call definitions to to a single header file.

Signed-off-by: Mike Kravetz <kravetz@us.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 drivers/net/ibmveth.h        | 10 ----------
 include/asm-powerpc/hvcall.h |  8 ++++++++
 2 files changed, 8 insertions(+), 10 deletions(-)

(limited to 'include')

diff --git a/drivers/net/ibmveth.h b/drivers/net/ibmveth.h
index 8385bf836507..149191cef2f0 100644
--- a/drivers/net/ibmveth.h
+++ b/drivers/net/ibmveth.h
@@ -41,16 +41,6 @@
 #define IbmVethMcastRemoveFilter     0x2UL
 #define IbmVethMcastClearFilterTable 0x3UL
 
-/* hcall numbers */
-#define H_VIO_SIGNAL             0x104
-#define H_REGISTER_LOGICAL_LAN   0x114
-#define H_FREE_LOGICAL_LAN       0x118
-#define H_ADD_LOGICAL_LAN_BUFFER 0x11C
-#define H_SEND_LOGICAL_LAN       0x120
-#define H_MULTICAST_CTRL         0x130
-#define H_CHANGE_LOGICAL_LAN_MAC 0x14C
-#define H_FREE_LOGICAL_LAN_BUFFER 0x1D4
-
 /* hcall macros */
 #define h_register_logical_lan(ua, buflst, rxq, fltlst, mac) \
   plpar_hcall_norets(H_REGISTER_LOGICAL_LAN, ua, buflst, rxq, fltlst, mac)
diff --git a/include/asm-powerpc/hvcall.h b/include/asm-powerpc/hvcall.h
index 0d3c4e85711a..f07ae50cbc2c 100644
--- a/include/asm-powerpc/hvcall.h
+++ b/include/asm-powerpc/hvcall.h
@@ -164,9 +164,15 @@
 #define H_VIO_SIGNAL		0x104
 #define H_SEND_CRQ		0x108
 #define H_COPY_RDMA		0x110
+#define H_REGISTER_LOGICAL_LAN	0x114
+#define H_FREE_LOGICAL_LAN	0x118
+#define H_ADD_LOGICAL_LAN_BUFFER 0x11C
+#define H_SEND_LOGICAL_LAN	0x120
+#define H_MULTICAST_CTRL	0x130
 #define H_SET_XDABR		0x134
 #define H_STUFF_TCE		0x138
 #define H_PUT_TCE_INDIRECT	0x13C
+#define H_CHANGE_LOGICAL_LAN_MAC 0x14C
 #define H_VTERM_PARTNER_INFO	0x150
 #define H_REGISTER_VTERM	0x154
 #define H_FREE_VTERM		0x158
@@ -196,11 +202,13 @@
 #define H_GET_HCA_INFO          0x1B8
 #define H_GET_PERF_COUNT        0x1BC
 #define H_MANAGE_TRACE          0x1C0
+#define H_FREE_LOGICAL_LAN_BUFFER 0x1D4
 #define H_QUERY_INT_STATE       0x1E4
 #define H_POLL_PENDING		0x1D8
 #define H_JOIN			0x298
 #define H_VASI_STATE            0x2A4
 #define H_ENABLE_CRQ		0x2B0
+#define MAX_HCALL_OPCODES	(H_ENABLE_CRQ >> 2)
 
 #ifndef __ASSEMBLY__
 
-- 
cgit v1.2.3


From b9377ffc3a03cde558d76349a262a1adbb6d3112 Mon Sep 17 00:00:00 2001
From: Anton Blanchard <anton@samba.org>
Date: Wed, 19 Jul 2006 08:01:28 +1000
Subject: [POWERPC] clean up pseries hcall interfaces

Our pseries hcall interfaces are out of control:

	plpar_hcall_norets
	plpar_hcall
	plpar_hcall_8arg_2ret
	plpar_hcall_4out
	plpar_hcall_7arg_7ret
	plpar_hcall_9arg_9ret

Create 3 interfaces to cover all cases:

	plpar_hcall_norets:	7 arguments no returns
	plpar_hcall:		6 arguments 4 returns
	plpar_hcall9:		9 arguments 9 returns

There are only 2 cases in the kernel that need plpar_hcall9, hopefully
we can keep it that way.

Pass in a buffer to stash return parameters so we avoid the &dummy1,
&dummy2 madness.

Signed-off-by: Anton Blanchard <anton@samba.org>
--
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 arch/powerpc/kernel/lparcfg.c                   |  18 ++-
 arch/powerpc/kernel/rtas.c                      |  11 +-
 arch/powerpc/platforms/pseries/hvCall.S         | 207 ++++--------------------
 arch/powerpc/platforms/pseries/hvconsole.c      |   5 +-
 arch/powerpc/platforms/pseries/lpar.c           |  12 +-
 arch/powerpc/platforms/pseries/plpar_wrappers.h |  97 +++++++++--
 arch/powerpc/platforms/pseries/xics.c           |  22 +--
 drivers/net/ibmveth.c                           |   3 +-
 drivers/net/ibmveth.h                           |  17 +-
 include/asm-powerpc/hvcall.h                    | 105 +++---------
 10 files changed, 184 insertions(+), 313 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/kernel/lparcfg.c b/arch/powerpc/kernel/lparcfg.c
index 3ce3a2d56fa8..41c05dcd68f4 100644
--- a/arch/powerpc/kernel/lparcfg.c
+++ b/arch/powerpc/kernel/lparcfg.c
@@ -182,8 +182,14 @@ static unsigned int h_get_ppp(unsigned long *entitled,
 			      unsigned long *resource)
 {
 	unsigned long rc;
-	rc = plpar_hcall_4out(H_GET_PPP, 0, 0, 0, 0, entitled, unallocated,
-			      aggregation, resource);
+	unsigned long retbuf[PLPAR_HCALL_BUFSIZE];
+
+	rc = plpar_hcall(H_GET_PPP, retbuf);
+
+	*entitled = retbuf[0];
+	*unallocated = retbuf[1];
+	*aggregation = retbuf[2];
+	*resource = retbuf[3];
 
 	log_plpar_hcall_return(rc, "H_GET_PPP");
 
@@ -193,8 +199,12 @@ static unsigned int h_get_ppp(unsigned long *entitled,
 static void h_pic(unsigned long *pool_idle_time, unsigned long *num_procs)
 {
 	unsigned long rc;
-	unsigned long dummy;
-	rc = plpar_hcall(H_PIC, 0, 0, 0, 0, pool_idle_time, num_procs, &dummy);
+	unsigned long retbuf[PLPAR_HCALL_BUFSIZE];
+
+	rc = plpar_hcall(H_PIC, retbuf);
+
+	*pool_idle_time = retbuf[0];
+	*num_procs = retbuf[1];
 
 	if (rc != H_AUTHORITY)
 		log_plpar_hcall_return(rc, "H_PIC");
diff --git a/arch/powerpc/kernel/rtas.c b/arch/powerpc/kernel/rtas.c
index 10e10be324c9..14353b8789dd 100644
--- a/arch/powerpc/kernel/rtas.c
+++ b/arch/powerpc/kernel/rtas.c
@@ -668,15 +668,14 @@ static int rtas_ibm_suspend_me(struct rtas_args *args)
 	int i;
 	long state;
 	long rc;
-	unsigned long dummy;
-
+	unsigned long retbuf[PLPAR_HCALL_BUFSIZE];
 	struct rtas_suspend_me_data data;
 
 	/* Make sure the state is valid */
-	rc = plpar_hcall(H_VASI_STATE,
-			 ((u64)args->args[0] << 32) | args->args[1],
-			 0, 0, 0,
-			 &state, &dummy, &dummy);
+	rc = plpar_hcall(H_VASI_STATE, retbuf,
+			 ((u64)args->args[0] << 32) | args->args[1]);
+
+	state = retbuf[0];
 
 	if (rc) {
 		printk(KERN_ERR "rtas_ibm_suspend_me: vasi_state returned %ld\n",rc);
diff --git a/arch/powerpc/platforms/pseries/hvCall.S b/arch/powerpc/platforms/pseries/hvCall.S
index c9ff547f9d25..9a99b056bd27 100644
--- a/arch/powerpc/platforms/pseries/hvCall.S
+++ b/arch/powerpc/platforms/pseries/hvCall.S
@@ -1,7 +1,6 @@
 /*
  * This file contains the generic code to perform a call to the
  * pSeries LPAR hypervisor.
- * NOTE: this file will go away when we move to inline this work.
  *
  * This program is free software; you can redistribute it and/or
  * modify it under the terms of the GNU General Public License
@@ -16,42 +15,6 @@
 
 	.text
 
-/* long plpar_hcall(unsigned long opcode,		R3
-			unsigned long arg1,		R4
-			unsigned long arg2,		R5
-			unsigned long arg3,		R6
-			unsigned long arg4,		R7
-			unsigned long *out1,		R8
-			unsigned long *out2,		R9
-			unsigned long *out3);		R10
- */
-_GLOBAL(plpar_hcall)
-	HMT_MEDIUM
-
-	mfcr	r0
-
-	std	r8,STK_PARM(r8)(r1)	/* Save out ptrs */
-	std	r9,STK_PARM(r9)(r1)
-	std	r10,STK_PARM(r10)(r1)
-
-	stw	r0,8(r1)
-
-	HVSC				/* invoke the hypervisor */
-
-	lwz	r0,8(r1)
-
-	ld	r8,STK_PARM(r8)(r1)	/* Fetch r4-r6 ret args */
-	ld	r9,STK_PARM(r9)(r1)
-	ld	r10,STK_PARM(r10)(r1)
-	std	r4,0(r8)
-	std	r5,0(r9)
-	std	r6,0(r10)
-
-	mtcrf	0xff,r0
-	blr				/* return r3 = status */
-
-
-/* Simple interface with no output values (other than status) */
 _GLOBAL(plpar_hcall_norets)
 	HMT_MEDIUM
 
@@ -64,164 +27,64 @@ _GLOBAL(plpar_hcall_norets)
 	mtcrf	0xff,r0
 	blr				/* return r3 = status */
 
-
-/* long plpar_hcall_8arg_2ret(unsigned long opcode,	R3
-			unsigned long arg1,		R4
-			unsigned long arg2,		R5
-			unsigned long arg3,		R6
-			unsigned long arg4,		R7
-			unsigned long arg5,		R8
-			unsigned long arg6,		R9
-			unsigned long arg7,		R10
-			unsigned long arg8,		112(R1)
-			unsigned long *out1);		120(R1)
- */
-_GLOBAL(plpar_hcall_8arg_2ret)
+_GLOBAL(plpar_hcall)
 	HMT_MEDIUM
 
 	mfcr	r0
-	ld	r11,STK_PARM(r11)(r1)	/* put arg8 in R11 */
 	stw	r0,8(r1)
 
-	HVSC				/* invoke the hypervisor */
-
-	lwz	r0,8(r1)
-	ld	r10,STK_PARM(r12)(r1)	/* Fetch r4 ret arg */
-	std	r4,0(r10)
-	mtcrf	0xff,r0
-	blr				/* return r3 = status */
-
-
-/* long plpar_hcall_4out(unsigned long opcode,		R3
-		 	unsigned long arg1,		R4
-		 	unsigned long arg2,		R5
-		 	unsigned long arg3,		R6
-		 	unsigned long arg4,		R7
-		 	unsigned long *out1,		R8
-		 	unsigned long *out2,		R9
-		 	unsigned long *out3,		R10
-		 	unsigned long *out4);		112(R1)
- */
-_GLOBAL(plpar_hcall_4out)
-	HMT_MEDIUM
-
-	mfcr	r0
-	stw	r0,8(r1)
+	std     r4,STK_PARM(r4)(r1)     /* Save ret buffer */
 
-	std	r8,STK_PARM(r8)(r1)	/* Save out ptrs */
-	std	r9,STK_PARM(r9)(r1)
-	std	r10,STK_PARM(r10)(r1)
+	mr	r4,r5
+	mr	r5,r6
+	mr	r6,r7
+	mr	r7,r8
+	mr	r8,r9
+	mr	r9,r10
 
 	HVSC				/* invoke the hypervisor */
 
-	lwz	r0,8(r1)
-
-	ld	r8,STK_PARM(r8)(r1)	/* Fetch r4-r7 ret args */
-	ld	r9,STK_PARM(r9)(r1)
-	ld	r10,STK_PARM(r10)(r1)
-	ld	r11,STK_PARM(r11)(r1)
-	std	r4,0(r8)
-	std	r5,0(r9)
-	std	r6,0(r10)
-	std	r7,0(r11)
-
-	mtcrf	0xff,r0
-	blr				/* return r3 = status */
-
-/* plpar_hcall_7arg_7ret(unsigned long opcode,		R3
-			 unsigned long arg1,		R4
-			 unsigned long arg2,		R5
-			 unsigned long arg3,		R6
-			 unsigned long arg4,		R7
-			 unsigned long arg5,		R8
-			 unsigned long arg6,		R9
-			 unsigned long arg7,		R10
-			 unsigned long *out1,		112(R1)
-			 unsigned long *out2,		110(R1)
-			 unsigned long *out3,		108(R1)
-			 unsigned long *out4,		106(R1)
-			 unsigned long *out5,		104(R1)
-			 unsigned long *out6,		102(R1)
-			 unsigned long *out7);		100(R1)
-*/
-_GLOBAL(plpar_hcall_7arg_7ret)
-	HMT_MEDIUM
-
-	mfcr	r0
-	stw	r0,8(r1)
-
-	HVSC				/* invoke the hypervisor */
+	ld	r12,STK_PARM(r4)(r1)
+	std	r4,  0(r12)
+	std	r5,  8(r12)
+	std	r6, 16(r12)
+	std	r7, 24(r12)
 
 	lwz	r0,8(r1)
-
-	ld	r11,STK_PARM(r11)(r1)	/* Fetch r4 ret arg */
-	std	r4,0(r11)
-	ld	r11,STK_PARM(r12)(r1)	/* Fetch r5 ret arg */
-	std	r5,0(r11)
-	ld	r11,STK_PARM(r13)(r1)	/* Fetch r6 ret arg */
-	std	r6,0(r11)
-	ld	r11,STK_PARM(r14)(r1)	/* Fetch r7 ret arg */
-	std	r7,0(r11)
-	ld	r11,STK_PARM(r15)(r1)	/* Fetch r8 ret arg */
-	std	r8,0(r11)
-	ld	r11,STK_PARM(r16)(r1)	/* Fetch r9 ret arg */
-	std	r9,0(r11)
-	ld	r11,STK_PARM(r17)(r1)	/* Fetch r10 ret arg */
-	std	r10,0(r11)
-
 	mtcrf	0xff,r0
 
 	blr				/* return r3 = status */
 
-/* plpar_hcall_9arg_9ret(unsigned long opcode,		R3
-			 unsigned long arg1,		R4
-			 unsigned long arg2,		R5
-			 unsigned long arg3,		R6
-			 unsigned long arg4,		R7
-			 unsigned long arg5,		R8
-			 unsigned long arg6,		R9
-			 unsigned long arg7,		R10
-			 unsigned long arg8,		112(R1)
-			 unsigned long arg9,		110(R1)
-			 unsigned long *out1,		108(R1)
-			 unsigned long *out2,		106(R1)
-			 unsigned long *out3,		104(R1)
-			 unsigned long *out4,		102(R1)
-			 unsigned long *out5,		100(R1)
-			 unsigned long *out6,		 98(R1)
-			 unsigned long *out7);		 96(R1)
-			 unsigned long *out8,		 94(R1)
-		         unsigned long *out9,            92(R1)
-*/
-_GLOBAL(plpar_hcall_9arg_9ret)
+_GLOBAL(plpar_hcall9)
 	HMT_MEDIUM
 
 	mfcr	r0
 	stw	r0,8(r1)
 
-	ld	r11,STK_PARM(r11)(r1)	 /* put arg8 in R11 */
-	ld	r12,STK_PARM(r12)(r1)    /* put arg9 in R12 */
+	std     r4,STK_PARM(r4)(r1)     /* Save ret buffer */
+
+	mr	r4,r5
+	mr	r5,r6
+	mr	r6,r7
+	mr	r7,r8
+	mr	r8,r9
+	mr	r9,r10
+	ld	r10,STK_PARM(r11)(r1)	 /* put arg7 in R10 */
+	ld	r11,STK_PARM(r12)(r1)	 /* put arg8 in R11 */
+	ld	r12,STK_PARM(r13)(r1)    /* put arg9 in R12 */
 
 	HVSC				/* invoke the hypervisor */
 
-	ld	r0,STK_PARM(r13)(r1)	/* Fetch r4 ret arg */
-	stdx	r4,r0,r0
-	ld	r0,STK_PARM(r14)(r1)	/* Fetch r5 ret arg */
-	stdx	r5,r0,r0
-	ld	r0,STK_PARM(r15)(r1)	/* Fetch r6 ret arg */
-	stdx	r6,r0,r0
-	ld	r0,STK_PARM(r16)(r1)	/* Fetch r7 ret arg */
-	stdx	r7,r0,r0
-	ld	r0,STK_PARM(r17)(r1)	/* Fetch r8 ret arg */
-	stdx	r8,r0,r0
-	ld	r0,STK_PARM(r18)(r1)	/* Fetch r9 ret arg */
-	stdx	r9,r0,r0
-	ld	r0,STK_PARM(r19)(r1)	/* Fetch r10 ret arg */
-	stdx	r10,r0,r0
-	ld	r0,STK_PARM(r20)(r1)	/* Fetch r11 ret arg */
-	stdx	r11,r0,r0
-	ld	r0,STK_PARM(r21)(r1)	/* Fetch r12 ret arg */
-	stdx	r12,r0,r0
+	ld	r12,STK_PARM(r4)(r1)
+	std	r4,  0(r12)
+	std	r5,  8(r12)
+	std	r6, 16(r12)
+	std	r7, 24(r12)
+	std	r8, 32(r12)
+	std	r9, 40(r12)
+	std	r10,48(r12)
+	std	r11,56(r12)
+	std	r12,64(r12)
 
 	lwz	r0,8(r1)
 	mtcrf	0xff,r0
diff --git a/arch/powerpc/platforms/pseries/hvconsole.c b/arch/powerpc/platforms/pseries/hvconsole.c
index a72a987f1d4d..3f6a89b09816 100644
--- a/arch/powerpc/platforms/pseries/hvconsole.c
+++ b/arch/powerpc/platforms/pseries/hvconsole.c
@@ -27,6 +27,7 @@
 #include <linux/module.h>
 #include <asm/hvcall.h>
 #include <asm/hvconsole.h>
+#include "plpar_wrappers.h"
 
 /**
  * hvc_get_chars - retrieve characters from firmware for denoted vterm adatper
@@ -40,9 +41,9 @@ int hvc_get_chars(uint32_t vtermno, char *buf, int count)
 {
 	unsigned long got;
 
-	if (plpar_hcall(H_GET_TERM_CHAR, vtermno, 0, 0, 0, &got,
-		(unsigned long *)buf, (unsigned long *)buf+1) == H_SUCCESS)
+	if (plpar_get_term_char(vtermno, &got, buf) == H_SUCCESS)
 		return got;
+
 	return 0;
 }
 
diff --git a/arch/powerpc/platforms/pseries/lpar.c b/arch/powerpc/platforms/pseries/lpar.c
index 4cb7ff227f72..6cbf14266d5e 100644
--- a/arch/powerpc/platforms/pseries/lpar.c
+++ b/arch/powerpc/platforms/pseries/lpar.c
@@ -48,13 +48,11 @@
 #define DBG_LOW(fmt...) do { } while(0)
 #endif
 
-/* in pSeries_hvCall.S */
+/* in hvCall.S */
 EXPORT_SYMBOL(plpar_hcall);
-EXPORT_SYMBOL(plpar_hcall_4out);
+EXPORT_SYMBOL(plpar_hcall9);
 EXPORT_SYMBOL(plpar_hcall_norets);
-EXPORT_SYMBOL(plpar_hcall_8arg_2ret);
-EXPORT_SYMBOL(plpar_hcall_7arg_7ret);
-EXPORT_SYMBOL(plpar_hcall_9arg_9ret);
+
 extern void pSeries_find_serial_port(void);
 
 
@@ -277,7 +275,6 @@ long pSeries_lpar_hpte_insert(unsigned long hpte_group,
 	unsigned long flags;
 	unsigned long slot;
 	unsigned long hpte_v, hpte_r;
-	unsigned long dummy0, dummy1;
 
 	if (!(vflags & HPTE_V_BOLTED))
 		DBG_LOW("hpte_insert(group=%lx, va=%016lx, pa=%016lx, "
@@ -302,8 +299,7 @@ long pSeries_lpar_hpte_insert(unsigned long hpte_group,
 	if (rflags & (_PAGE_GUARDED|_PAGE_NO_CACHE))
 		hpte_r &= ~_PAGE_COHERENT;
 
-	lpar_rc = plpar_hcall(H_ENTER, flags, hpte_group, hpte_v,
-			      hpte_r, &slot, &dummy0, &dummy1);
+	lpar_rc = plpar_pte_enter(flags, hpte_group, hpte_v, hpte_r, &slot);
 	if (unlikely(lpar_rc == H_PTEG_FULL)) {
 		if (!(vflags & HPTE_V_BOLTED))
 			DBG_LOW(" full\n");
diff --git a/arch/powerpc/platforms/pseries/plpar_wrappers.h b/arch/powerpc/platforms/pseries/plpar_wrappers.h
index 3bd1b3e06003..ebd15de7597e 100644
--- a/arch/powerpc/platforms/pseries/plpar_wrappers.h
+++ b/arch/powerpc/platforms/pseries/plpar_wrappers.h
@@ -5,20 +5,17 @@
 
 static inline long poll_pending(void)
 {
-	unsigned long dummy;
-	return plpar_hcall(H_POLL_PENDING, 0, 0, 0, 0, &dummy, &dummy, &dummy);
+	return plpar_hcall_norets(H_POLL_PENDING);
 }
 
 static inline long prod_processor(void)
 {
-	plpar_hcall_norets(H_PROD);
-	return 0;
+	return plpar_hcall_norets(H_PROD);
 }
 
 static inline long cede_processor(void)
 {
-	plpar_hcall_norets(H_CEDE);
-	return 0;
+	return plpar_hcall_norets(H_CEDE);
 }
 
 static inline long vpa_call(unsigned long flags, unsigned long cpu,
@@ -42,21 +39,47 @@ static inline long register_vpa(unsigned long cpu, unsigned long vpa)
 
 extern void vpa_init(int cpu);
 
+static inline long plpar_pte_enter(unsigned long flags,
+		unsigned long hpte_group, unsigned long hpte_v,
+		unsigned long hpte_r, unsigned long *slot)
+{
+	long rc;
+	unsigned long retbuf[PLPAR_HCALL_BUFSIZE];
+
+	rc = plpar_hcall(H_ENTER, retbuf, flags, hpte_group, hpte_v, hpte_r);
+
+	*slot = retbuf[0];
+
+	return rc;
+}
+
 static inline long plpar_pte_remove(unsigned long flags, unsigned long ptex,
 		unsigned long avpn, unsigned long *old_pteh_ret,
 		unsigned long *old_ptel_ret)
 {
-	unsigned long dummy;
-	return plpar_hcall(H_REMOVE, flags, ptex, avpn, 0, old_pteh_ret,
-			old_ptel_ret, &dummy);
+	long rc;
+	unsigned long retbuf[PLPAR_HCALL_BUFSIZE];
+
+	rc = plpar_hcall(H_REMOVE, retbuf, flags, ptex, avpn);
+
+	*old_pteh_ret = retbuf[0];
+	*old_ptel_ret = retbuf[1];
+
+	return rc;
 }
 
 static inline long plpar_pte_read(unsigned long flags, unsigned long ptex,
 		unsigned long *old_pteh_ret, unsigned long *old_ptel_ret)
 {
-	unsigned long dummy;
-	return plpar_hcall(H_READ, flags, ptex, 0, 0, old_pteh_ret,
-			old_ptel_ret, &dummy);
+	long rc;
+	unsigned long retbuf[PLPAR_HCALL_BUFSIZE];
+
+	rc = plpar_hcall(H_READ, retbuf, flags, ptex);
+
+	*old_pteh_ret = retbuf[0];
+	*old_ptel_ret = retbuf[1];
+
+	return rc;
 }
 
 static inline long plpar_pte_protect(unsigned long flags, unsigned long ptex,
@@ -68,9 +91,14 @@ static inline long plpar_pte_protect(unsigned long flags, unsigned long ptex,
 static inline long plpar_tce_get(unsigned long liobn, unsigned long ioba,
 		unsigned long *tce_ret)
 {
-	unsigned long dummy;
-	return plpar_hcall(H_GET_TCE, liobn, ioba, 0, 0, tce_ret, &dummy,
-			&dummy);
+	long rc;
+	unsigned long retbuf[PLPAR_HCALL_BUFSIZE];
+
+	rc = plpar_hcall(H_GET_TCE, retbuf, liobn, ioba);
+
+	*tce_ret = retbuf[0];
+
+	return rc;
 }
 
 static inline long plpar_tce_put(unsigned long liobn, unsigned long ioba,
@@ -94,9 +122,17 @@ static inline long plpar_tce_stuff(unsigned long liobn, unsigned long ioba,
 static inline long plpar_get_term_char(unsigned long termno,
 		unsigned long *len_ret, char *buf_ret)
 {
+	long rc;
+	unsigned long retbuf[PLPAR_HCALL_BUFSIZE];
 	unsigned long *lbuf = (unsigned long *)buf_ret;	/* TODO: alignment? */
-	return plpar_hcall(H_GET_TERM_CHAR, termno, 0, 0, 0, len_ret,
-			lbuf + 0, lbuf + 1);
+
+	rc = plpar_hcall(H_GET_TERM_CHAR, retbuf, termno);
+
+	*len_ret = retbuf[0];
+	lbuf[0] = retbuf[1];
+	lbuf[1] = retbuf[2];
+
+	return rc;
 }
 
 static inline long plpar_put_term_char(unsigned long termno, unsigned long len,
@@ -107,4 +143,31 @@ static inline long plpar_put_term_char(unsigned long termno, unsigned long len,
 			lbuf[1]);
 }
 
+static inline long plpar_eoi(unsigned long xirr)
+{
+	return plpar_hcall_norets(H_EOI, xirr);
+}
+
+static inline long plpar_cppr(unsigned long cppr)
+{
+	return plpar_hcall_norets(H_CPPR, cppr);
+}
+
+static inline long plpar_ipi(unsigned long servernum, unsigned long mfrr)
+{
+	return plpar_hcall_norets(H_IPI, servernum, mfrr);
+}
+
+static inline long plpar_xirr(unsigned long *xirr_ret)
+{
+	long rc;
+	unsigned long retbuf[PLPAR_HCALL_BUFSIZE];
+
+	rc = plpar_hcall(H_XIRR, retbuf);
+
+	*xirr_ret = retbuf[0];
+
+	return rc;
+}
+
 #endif /* _PSERIES_PLPAR_WRAPPERS_H */
diff --git a/arch/powerpc/platforms/pseries/xics.c b/arch/powerpc/platforms/pseries/xics.c
index 1eab4688be17..c88ec63129f3 100644
--- a/arch/powerpc/platforms/pseries/xics.c
+++ b/arch/powerpc/platforms/pseries/xics.c
@@ -34,6 +34,7 @@
 #include <asm/i8259.h>
 
 #include "xics.h"
+#include "plpar_wrappers.h"
 
 #define XICS_IPI		2
 #define XICS_IRQ_SPURIOUS	0
@@ -110,27 +111,6 @@ static inline void direct_qirr_info(int n_cpu, u8 value)
 /* LPAR low level accessors */
 
 
-static inline long plpar_eoi(unsigned long xirr)
-{
-	return plpar_hcall_norets(H_EOI, xirr);
-}
-
-static inline long plpar_cppr(unsigned long cppr)
-{
-	return plpar_hcall_norets(H_CPPR, cppr);
-}
-
-static inline long plpar_ipi(unsigned long servernum, unsigned long mfrr)
-{
-	return plpar_hcall_norets(H_IPI, servernum, mfrr);
-}
-
-static inline long plpar_xirr(unsigned long *xirr_ret)
-{
-	unsigned long dummy;
-	return plpar_hcall(H_XIRR, 0, 0, 0, 0, xirr_ret, &dummy, &dummy);
-}
-
 static inline unsigned int lpar_xirr_info_get(int n_cpu)
 {
 	unsigned long lpar_rc;
diff --git a/drivers/net/ibmveth.c b/drivers/net/ibmveth.c
index 0464e78f733a..e56eac88b809 100644
--- a/drivers/net/ibmveth.c
+++ b/drivers/net/ibmveth.c
@@ -702,7 +702,8 @@ static int ibmveth_start_xmit(struct sk_buff *skb, struct net_device *netdev)
 					     desc[3].desc,
 					     desc[4].desc,
 					     desc[5].desc,
-					     correlator);
+					     correlator,
+					     &correlator);
 	} while ((lpar_rc == H_BUSY) && (retry_count--));
 
 	if(lpar_rc != H_SUCCESS && lpar_rc != H_DROPPED) {
diff --git a/drivers/net/ibmveth.h b/drivers/net/ibmveth.h
index 149191cef2f0..f5b25bff1540 100644
--- a/drivers/net/ibmveth.h
+++ b/drivers/net/ibmveth.h
@@ -51,8 +51,21 @@
 #define h_add_logical_lan_buffer(ua, buf) \
   plpar_hcall_norets(H_ADD_LOGICAL_LAN_BUFFER, ua, buf)
 
-#define h_send_logical_lan(ua, buf1, buf2, buf3, buf4, buf5, buf6, correlator) \
-  plpar_hcall_8arg_2ret(H_SEND_LOGICAL_LAN, ua, buf1, buf2, buf3, buf4, buf5, buf6, correlator, &correlator)
+static inline long h_send_logical_lan(unsigned long unit_address,
+		unsigned long desc1, unsigned long desc2, unsigned long desc3,
+		unsigned long desc4, unsigned long desc5, unsigned long desc6,
+		unsigned long corellator_in, unsigned long *corellator_out)
+{
+	long rc;
+	unsigned long retbuf[PLPAR_HCALL9_BUFSIZE];
+
+	rc = plpar_hcall9(H_SEND_LOGICAL_LAN, retbuf, unit_address, desc1,
+			desc2, desc3, desc4, desc5, desc6, corellator_in);
+
+	*corellator_out = retbuf[0];
+
+	return rc;
+}
 
 #define h_multicast_ctrl(ua, cmd, mac) \
   plpar_hcall_norets(H_MULTICAST_CTRL, ua, cmd, mac)
diff --git a/include/asm-powerpc/hvcall.h b/include/asm-powerpc/hvcall.h
index f07ae50cbc2c..63ce1ac8c1f4 100644
--- a/include/asm-powerpc/hvcall.h
+++ b/include/asm-powerpc/hvcall.h
@@ -212,94 +212,39 @@
 
 #ifndef __ASSEMBLY__
 
-/* plpar_hcall() -- Generic call interface using above opcodes
+/**
+ * plpar_hcall_norets: - Make a pseries hypervisor call with no return arguments
+ * @opcode: The hypervisor call to make.
  *
- * The actual call interface is a hypervisor call instruction with
- * the opcode in R3 and input args in R4-R7.
- * Status is returned in R3 with variable output values in R4-R11.
- * Only H_PTE_READ with H_READ_4 uses R6-R11 so we ignore it for now
- * and return only two out args which MUST ALWAYS BE PROVIDED.
- */
-long plpar_hcall(unsigned long opcode,
-		 unsigned long arg1,
-		 unsigned long arg2,
-		 unsigned long arg3,
-		 unsigned long arg4,
-		 unsigned long *out1,
-		 unsigned long *out2,
-		 unsigned long *out3);
-
-/* Same as plpar_hcall but for those opcodes that return no values
- * other than status.  Slightly more efficient.
+ * This call supports up to 7 arguments and only returns the status of
+ * the hcall. Use this version where possible, its slightly faster than
+ * the other plpar_hcalls.
  */
 long plpar_hcall_norets(unsigned long opcode, ...);
 
-/*
- * Special hcall interface for ibmveth support.
- * Takes 8 input parms. Returns a rc and stores the
- * R4 return value in *out1.
- */
-long plpar_hcall_8arg_2ret(unsigned long opcode,
-			   unsigned long arg1,
-			   unsigned long arg2,
-			   unsigned long arg3,
-			   unsigned long arg4,
-			   unsigned long arg5,
-			   unsigned long arg6,
-			   unsigned long arg7,
-			   unsigned long arg8,
-			   unsigned long *out1);
-
-/* plpar_hcall_4out()
+/**
+ * plpar_hcall: - Make a pseries hypervisor call
+ * @opcode: The hypervisor call to make.
+ * @retbuf: Buffer to store up to 4 return arguments in.
  *
- * same as plpar_hcall except with 4 output arguments.
+ * This call supports up to 6 arguments and 4 return arguments. Use
+ * PLPAR_HCALL_BUFSIZE to size the return argument buffer.
  *
+ * Used for all but the craziest of phyp interfaces (see plpar_hcall9)
  */
-long plpar_hcall_4out(unsigned long opcode,
-		      unsigned long arg1,
-		      unsigned long arg2,
-		      unsigned long arg3,
-		      unsigned long arg4,
-		      unsigned long *out1,
-		      unsigned long *out2,
-		      unsigned long *out3,
-		      unsigned long *out4);
+#define PLPAR_HCALL_BUFSIZE 4
+long plpar_hcall(unsigned long opcode, unsigned long *retbuf, ...);
 
-long plpar_hcall_7arg_7ret(unsigned long opcode,
-			   unsigned long arg1,
-			   unsigned long arg2,
-			   unsigned long arg3,
-			   unsigned long arg4,
-			   unsigned long arg5,
-			   unsigned long arg6,
-			   unsigned long arg7,
-			   unsigned long *out1,
-			   unsigned long *out2,
-			   unsigned long *out3,
-			   unsigned long *out4,
-			   unsigned long *out5,
-			   unsigned long *out6,
-			   unsigned long *out7);
-
-long plpar_hcall_9arg_9ret(unsigned long opcode,
-			   unsigned long arg1,
-			   unsigned long arg2,
-			   unsigned long arg3,
-			   unsigned long arg4,
-			   unsigned long arg5,
-			   unsigned long arg6,
-			   unsigned long arg7,
-			   unsigned long arg8,
-			   unsigned long arg9,
-			   unsigned long *out1,
-			   unsigned long *out2,
-			   unsigned long *out3,
-			   unsigned long *out4,
-			   unsigned long *out5,
-			   unsigned long *out6,
-			   unsigned long *out7,
-			   unsigned long *out8,
-			   unsigned long *out9);
+/**
+ * plpar_hcall9: - Make a pseries hypervisor call with up to 9 return arguments
+ * @opcode: The hypervisor call to make.
+ * @retbuf: Buffer to store up to 9 return arguments in.
+ *
+ * This call supports up to 9 arguments and 9 return arguments. Use
+ * PLPAR_HCALL9_BUFSIZE to size the return argument buffer.
+ */
+#define PLPAR_HCALL9_BUFSIZE 9
+long plpar_hcall9(unsigned long opcode, unsigned long *retbuf, ...);
 
 #endif /* __ASSEMBLY__ */
 #endif /* __KERNEL__ */
-- 
cgit v1.2.3


From 9e5c50fa8686ede7c37b939a0b950df50346eb3d Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@lst.de>
Date: Fri, 4 Aug 2006 17:18:30 +0200
Subject: [SCSI] remove SCSI_STATE_ #defines

These aren't used anymore since the field in scsi_cmnd where it was
stored has been removed.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 include/scsi/scsi_cmnd.h | 14 --------------
 1 file changed, 14 deletions(-)

(limited to 'include')

diff --git a/include/scsi/scsi_cmnd.h b/include/scsi/scsi_cmnd.h
index 58e6444eebee..be117f812deb 100644
--- a/include/scsi/scsi_cmnd.h
+++ b/include/scsi/scsi_cmnd.h
@@ -118,20 +118,6 @@ struct scsi_cmnd {
 	unsigned long pid;	/* Process ID, starts at 0. Unique per host. */
 };
 
-/*
- * These are the values that scsi_cmd->state can take.
- */
-#define SCSI_STATE_TIMEOUT         0x1000
-#define SCSI_STATE_FINISHED        0x1001
-#define SCSI_STATE_FAILED          0x1002
-#define SCSI_STATE_QUEUED          0x1003
-#define SCSI_STATE_UNUSED          0x1006
-#define SCSI_STATE_DISCONNECTING   0x1008
-#define SCSI_STATE_INITIALIZING    0x1009
-#define SCSI_STATE_BHQUEUE         0x100a
-#define SCSI_STATE_MLQUEUE         0x100b
-
-
 extern struct scsi_cmnd *scsi_get_command(struct scsi_device *, gfp_t);
 extern void scsi_put_command(struct scsi_cmnd *);
 extern void scsi_io_completion(struct scsi_cmnd *, unsigned int);
-- 
cgit v1.2.3


From 4ff36718ede26ee2da73f2dae94d71e2b06845fc Mon Sep 17 00:00:00 2001
From: Matthew Wilcox <matthew@wil.cx>
Date: Tue, 4 Jul 2006 12:15:20 -0600
Subject: [SCSI] Improve inquiry printing

 - Replace scsi_device_types array API with scsi_device_type function API.
   Gets rid of a lot of common code, as well as being easier to use.
 - Add the new device types in SPC4 r05a, and rename some of the older ones.
 - Reformat the printing of inquiry data; now fits on one line and
   includes PQ.

I think I've addressed all the feedback from the previous versions.  My
current test box prints:

scsi 2:0:1:0: Direct access     HP 18.2G ATLAS10K3_18_SCA HP05 PQ: 0 ANSI: 2

Signed-off-by: Matthew Wilcox <matthew@wil.cx>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 drivers/block/cciss_scsi.c | 14 ++++------
 drivers/scsi/fcal.c        |  3 +--
 drivers/scsi/g_NCR5380.c   |  3 +--
 drivers/scsi/megaraid.c    |  4 +--
 drivers/scsi/scsi.c        | 36 ++++++++++++++++++-------
 drivers/scsi/scsi_proc.c   |  4 +--
 drivers/scsi/scsi_scan.c   | 66 ++++++----------------------------------------
 include/scsi/scsi.h        | 10 +++----
 8 files changed, 46 insertions(+), 94 deletions(-)

(limited to 'include')

diff --git a/drivers/block/cciss_scsi.c b/drivers/block/cciss_scsi.c
index afdff32f6724..05f79d7393f7 100644
--- a/drivers/block/cciss_scsi.c
+++ b/drivers/block/cciss_scsi.c
@@ -251,10 +251,6 @@ scsi_cmd_stack_free(int ctlr)
 	stk->pool = NULL;
 }
 
-/* scsi_device_types comes from scsi.h */
-#define DEVICETYPE(n) (n<0 || n>MAX_SCSI_DEVICE_CODE) ? \
-	"Unknown" : scsi_device_types[n]
-
 #if 0
 static int xmargin=8;
 static int amargin=60;
@@ -389,7 +385,7 @@ cciss_scsi_add_entry(int ctlr, int hostno,
 	   time anyway (the scsi layer's inquiries will show that info) */
 	if (hostno != -1)
 		printk("cciss%d: %s device c%db%dt%dl%d added.\n", 
-			ctlr, DEVICETYPE(sd->devtype), hostno, 
+			ctlr, scsi_device_type(sd->devtype), hostno,
 			sd->bus, sd->target, sd->lun);
 	return 0;
 }
@@ -407,7 +403,7 @@ cciss_scsi_remove_entry(int ctlr, int hostno, int entry)
 		ccissscsi[ctlr].dev[i] = ccissscsi[ctlr].dev[i+1];
 	ccissscsi[ctlr].ndevices--;
 	printk("cciss%d: %s device c%db%dt%dl%d removed.\n",
-		ctlr, DEVICETYPE(sd.devtype), hostno, 
+		ctlr, scsi_device_type(sd.devtype), hostno,
 			sd.bus, sd.target, sd.lun);
 }
 
@@ -458,7 +454,7 @@ adjust_cciss_scsi_table(int ctlr, int hostno,
 		if (found == 0) { /* device no longer present. */ 
 			changes++;
 			/* printk("cciss%d: %s device c%db%dt%dl%d removed.\n",
-				ctlr, DEVICETYPE(csd->devtype), hostno, 
+				ctlr, scsi_device_type(csd->devtype), hostno,
 					csd->bus, csd->target, csd->lun); */
 			cciss_scsi_remove_entry(ctlr, hostno, i);
 			/* note, i not incremented */
@@ -468,7 +464,7 @@ adjust_cciss_scsi_table(int ctlr, int hostno,
 			printk("cciss%d: device c%db%dt%dl%d type changed "
 				"(device type now %s).\n",
 				ctlr, hostno, csd->bus, csd->target, csd->lun,
-					DEVICETYPE(csd->devtype));
+					scsi_device_type(csd->devtype));
 			csd->devtype = sd[j].devtype;
 			i++;	/* so just move along. */
 		} else 		/* device is same as it ever was, */
@@ -1098,7 +1094,7 @@ cciss_update_non_disk_devices(int cntl_num, int hostno)
 			if (ncurrent >= CCISS_MAX_SCSI_DEVS_PER_HBA) {
 				printk(KERN_INFO "cciss%d: %s ignored, "
 					"too many devices.\n", cntl_num,
-					DEVICETYPE(devtype));
+					scsi_device_type(devtype));
 				break;
 			}
 			memcpy(&currentsd[ncurrent].scsi3addr[0], 
diff --git a/drivers/scsi/fcal.c b/drivers/scsi/fcal.c
index 7f891023aa15..c4e16c0775de 100644
--- a/drivers/scsi/fcal.c
+++ b/drivers/scsi/fcal.c
@@ -248,8 +248,7 @@ int fcal_proc_info (struct Scsi_Host *host, char *buffer, char **start, off_t of
 				if (scd->id == target) {
 					SPRINTF ("  [AL-PA: %02x, Id: %02d, Port WWN: %08x%08x, Node WWN: %08x%08x]  ",
 						alpa, target, u1[0], u1[1], u2[0], u2[1]);
-					SPRINTF ("%s ", (scd->type < MAX_SCSI_DEVICE_CODE) ?
-						scsi_device_types[(short) scd->type] : "Unknown device");
+					SPRINTF ("%s ", scsi_device_type(scd->type));
 
 					for (j = 0; (j < 8) && (scd->vendor[j] >= 0x20); j++)
 						SPRINTF ("%c", scd->vendor[j]);
diff --git a/drivers/scsi/g_NCR5380.c b/drivers/scsi/g_NCR5380.c
index 67f1100f3103..cdd893bb4e28 100644
--- a/drivers/scsi/g_NCR5380.c
+++ b/drivers/scsi/g_NCR5380.c
@@ -811,7 +811,6 @@ static int generic_NCR5380_proc_info(struct Scsi_Host *scsi_ptr, char *buffer, c
 	struct NCR5380_hostdata *hostdata;
 #ifdef NCR5380_STATS
 	struct scsi_device *dev;
-	extern const char *const scsi_device_types[MAX_SCSI_DEVICE_CODE];
 #endif
 
 	NCR5380_setup(scsi_ptr);
@@ -851,7 +850,7 @@ static int generic_NCR5380_proc_info(struct Scsi_Host *scsi_ptr, char *buffer, c
 		long tr = hostdata->time_read[dev->id] / HZ;
 		long tw = hostdata->time_write[dev->id] / HZ;
 
-		PRINTP("  T:%d %s " ANDP dev->id ANDP(dev->type < MAX_SCSI_DEVICE_CODE) ? scsi_device_types[(int) dev->type] : "Unknown");
+		PRINTP("  T:%d %s " ANDP dev->id ANDP scsi_device_type(dev->type));
 		for (i = 0; i < 8; i++)
 			if (dev->vendor[i] >= 0x20)
 				*(buffer + (len++)) = dev->vendor[i];
diff --git a/drivers/scsi/megaraid.c b/drivers/scsi/megaraid.c
index 76edbb639d37..ccb0055ac73a 100644
--- a/drivers/scsi/megaraid.c
+++ b/drivers/scsi/megaraid.c
@@ -2822,9 +2822,7 @@ mega_print_inquiry(char *page, char *scsi_inq)
 
 	i = scsi_inq[0] & 0x1f;
 
-	len += sprintf(page+len, "  Type:   %s ",
-		i < MAX_SCSI_DEVICE_CODE ? scsi_device_types[i] :
-		   "Unknown          ");
+	len += sprintf(page+len, "  Type:   %s ", scsi_device_type(i));
 
 	len += sprintf(page+len,
 	"                 ANSI SCSI revision: %02x", scsi_inq[2] & 0x07);
diff --git a/drivers/scsi/scsi.c b/drivers/scsi/scsi.c
index b332caddd5b3..94df671d776a 100644
--- a/drivers/scsi/scsi.c
+++ b/drivers/scsi/scsi.c
@@ -96,24 +96,40 @@ unsigned int scsi_logging_level;
 EXPORT_SYMBOL(scsi_logging_level);
 #endif
 
-const char *const scsi_device_types[MAX_SCSI_DEVICE_CODE] = {
-	"Direct-Access    ",
-	"Sequential-Access",
+static const char *const scsi_device_types[] = {
+	"Direct access    ",
+	"Sequential access",
 	"Printer          ",
 	"Processor        ",
 	"WORM             ",
-	"CD-ROM           ",
+	"CD/DVD           ",
 	"Scanner          ",
-	"Optical Device   ",
-	"Medium Changer   ",
+	"Optical memory   ",
+	"Media changer    ",
 	"Communications   ",
-	"Unknown          ",
-	"Unknown          ",
+	"ASC IT8          ",
+	"ASC IT8          ",
 	"RAID             ",
 	"Enclosure        ",
-	"Direct-Access-RBC",
+	"Direct access RBC",
+	"Optical card     ",
+	"Bridge controller",
+	"Object storage   ",
+	"Automation/Drive ",
 };
-EXPORT_SYMBOL(scsi_device_types);
+
+const char * scsi_device_type(unsigned type)
+{
+	if (type == 0x1e)
+		return "Well-known LUN   ";
+	if (type == 0x1f)
+		return "No Device        ";
+	if (type > ARRAY_SIZE(scsi_device_types))
+		return "Unknown          ";
+	return scsi_device_types[type];
+}
+
+EXPORT_SYMBOL(scsi_device_type);
 
 struct scsi_host_cmd_pool {
 	kmem_cache_t	*slab;
diff --git a/drivers/scsi/scsi_proc.c b/drivers/scsi/scsi_proc.c
index 55200e4fdf11..524a5f7a5193 100644
--- a/drivers/scsi/scsi_proc.c
+++ b/drivers/scsi/scsi_proc.c
@@ -178,9 +178,7 @@ static int proc_print_scsidevice(struct device *dev, void *data)
 
 	seq_printf(s, "\n");
 
-	seq_printf(s, "  Type:   %s ",
-		     sdev->type < MAX_SCSI_DEVICE_CODE ?
-	       scsi_device_types[(int) sdev->type] : "Unknown          ");
+	seq_printf(s, "  Type:   %s ", scsi_device_type(sdev->type));
 	seq_printf(s, "               ANSI"
 		     " SCSI revision: %02x", (sdev->scsi_level - 1) ?
 		     sdev->scsi_level - 1 : 1);
diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c
index 1bd92b9b46d9..180399406510 100644
--- a/drivers/scsi/scsi_scan.c
+++ b/drivers/scsi/scsi_scan.c
@@ -133,59 +133,6 @@ static void scsi_unlock_floptical(struct scsi_device *sdev,
 			 SCSI_TIMEOUT, 3);
 }
 
-/**
- * print_inquiry - printk the inquiry information
- * @inq_result:	printk this SCSI INQUIRY
- *
- * Description:
- *     printk the vendor, model, and other information found in the
- *     INQUIRY data in @inq_result.
- *
- * Notes:
- *     Remove this, and replace with a hotplug event that logs any
- *     relevant information.
- **/
-static void print_inquiry(unsigned char *inq_result)
-{
-	int i;
-
-	printk(KERN_NOTICE "  Vendor: ");
-	for (i = 8; i < 16; i++)
-		if (inq_result[i] >= 0x20 && i < inq_result[4] + 5)
-			printk("%c", inq_result[i]);
-		else
-			printk(" ");
-
-	printk("  Model: ");
-	for (i = 16; i < 32; i++)
-		if (inq_result[i] >= 0x20 && i < inq_result[4] + 5)
-			printk("%c", inq_result[i]);
-		else
-			printk(" ");
-
-	printk("  Rev: ");
-	for (i = 32; i < 36; i++)
-		if (inq_result[i] >= 0x20 && i < inq_result[4] + 5)
-			printk("%c", inq_result[i]);
-		else
-			printk(" ");
-
-	printk("\n");
-
-	i = inq_result[0] & 0x1f;
-
-	printk(KERN_NOTICE "  Type:   %s ",
-	       i <
-	       MAX_SCSI_DEVICE_CODE ? scsi_device_types[i] :
-	       "Unknown          ");
-	printk("                 ANSI SCSI revision: %02x",
-	       inq_result[2] & 0x07);
-	if ((inq_result[2] & 0x07) == 1 && (inq_result[3] & 0x0f) == 1)
-		printk(" CCS\n");
-	else
-		printk("\n");
-}
-
 /**
  * scsi_alloc_sdev - allocate and setup a scsi_Device
  *
@@ -653,9 +600,8 @@ static int scsi_add_lun(struct scsi_device *sdev, char *inq_result, int *bflags)
 	if (*bflags & BLIST_ISROM) {
 		/*
 		 * It would be better to modify sdev->type, and set
-		 * sdev->removable, but then the print_inquiry() output
-		 * would not show TYPE_ROM; if print_inquiry() is removed
-		 * the issue goes away.
+		 * sdev->removable; this can now be done since
+		 * print_inquiry has gone away.
 		 */
 		inq_result[0] = TYPE_ROM;
 		inq_result[1] |= 0x80;	/* removable */
@@ -684,8 +630,6 @@ static int scsi_add_lun(struct scsi_device *sdev, char *inq_result, int *bflags)
 		printk(KERN_INFO "scsi: unknown device type %d\n", sdev->type);
 	}
 
-	print_inquiry(inq_result);
-
 	/*
 	 * For a peripheral qualifier (PQ) value of 1 (001b), the SCSI
 	 * spec says: The device server is capable of supporting the
@@ -715,6 +659,12 @@ static int scsi_add_lun(struct scsi_device *sdev, char *inq_result, int *bflags)
 	if (inq_result[7] & 0x10)
 		sdev->sdtr = 1;
 
+	sdev_printk(KERN_NOTICE "scsi", sdev, "%s %.8s %.16s %.4s PQ: %d "
+			"ANSI: %d%s\n", scsi_device_type(sdev->type),
+			sdev->vendor, sdev->model, sdev->rev,
+			sdev->inq_periph_qual, inq_result[2] & 0x07,
+			(inq_result[3] & 0x0f) == 1 ? " CCS" : "");
+
 	/*
 	 * End sysfs code.
 	 */
diff --git a/include/scsi/scsi.h b/include/scsi/scsi.h
index c60b8ff2f5e4..1bc675201413 100644
--- a/include/scsi/scsi.h
+++ b/include/scsi/scsi.h
@@ -24,13 +24,6 @@
 extern const unsigned char scsi_command_size[8];
 #define COMMAND_SIZE(opcode) scsi_command_size[((opcode) >> 5) & 7]
 
-/*
- *	SCSI device types
- */
-
-#define MAX_SCSI_DEVICE_CODE 15
-extern const char *const scsi_device_types[MAX_SCSI_DEVICE_CODE];
-
 /*
  * Special value for scanning to specify scanning or rescanning of all
  * possible channels, (target) ids, or luns on a given shost.
@@ -225,6 +218,9 @@ static inline int scsi_status_is_good(int status)
 #define TYPE_RBC	    0x0e
 #define TYPE_NO_LUN         0x7f
 
+/* Returns a human-readable name for the device */
+extern const char * scsi_device_type(unsigned type);
+
 /*
  * standard mode-select header prepended to all mode-select commands
  */
-- 
cgit v1.2.3


From 40681b95a4ef798bc38c92e0d9b8c06bbdd34409 Mon Sep 17 00:00:00 2001
From: Michael Ellerman <michael@ellerman.id.au>
Date: Wed, 2 Aug 2006 11:13:50 +1000
Subject: [POWERPC] Make doc comments extractable

We don't have much in the way of doc comments, but some of those we do have
don't work because they start with "/***" or "/*", not "/**" which is what
kernel-doc requires.

Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 arch/powerpc/kernel/crash_dump.c |  2 +-
 arch/powerpc/sysdev/i8259.c      |  2 +-
 include/asm-powerpc/irq.h        | 24 ++++++++++++------------
 include/asm-powerpc/prom.h       |  8 ++++----
 4 files changed, 18 insertions(+), 18 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/kernel/crash_dump.c b/arch/powerpc/kernel/crash_dump.c
index 371973be8d71..2f6f5a7bc69e 100644
--- a/arch/powerpc/kernel/crash_dump.c
+++ b/arch/powerpc/kernel/crash_dump.c
@@ -80,7 +80,7 @@ static int __init parse_savemaxmem(char *p)
 }
 __setup("savemaxmem=", parse_savemaxmem);
 
-/*
+/**
  * copy_oldmem_page - copy one page from "oldmem"
  * @pfn: page frame number to be copied
  * @buf: target memory address for the copy; this can be in kernel address
diff --git a/arch/powerpc/sysdev/i8259.c b/arch/powerpc/sysdev/i8259.c
index 9855820b9548..26a6a3becd66 100644
--- a/arch/powerpc/sysdev/i8259.c
+++ b/arch/powerpc/sysdev/i8259.c
@@ -224,7 +224,7 @@ static struct irq_host_ops i8259_host_ops = {
 	.xlate = i8259_host_xlate,
 };
 
-/****
+/**
  * i8259_init - Initialize the legacy controller
  * @node: device node of the legacy PIC (can be NULL, but then, it will match
  *        all interrupts, so beware)
diff --git a/include/asm-powerpc/irq.h b/include/asm-powerpc/irq.h
index d903a62959da..4da41efb1319 100644
--- a/include/asm-powerpc/irq.h
+++ b/include/asm-powerpc/irq.h
@@ -137,7 +137,7 @@ struct irq_map_entry {
 extern struct irq_map_entry irq_map[NR_IRQS];
 
 
-/***
+/**
  * irq_alloc_host - Allocate a new irq_host data structure
  * @node: device-tree node of the interrupt controller
  * @revmap_type: type of reverse mapping to use
@@ -159,14 +159,14 @@ extern struct irq_host *irq_alloc_host(unsigned int revmap_type,
 				       irq_hw_number_t inval_irq);
 
 
-/***
+/**
  * irq_find_host - Locates a host for a given device node
  * @node: device-tree node of the interrupt controller
  */
 extern struct irq_host *irq_find_host(struct device_node *node);
 
 
-/***
+/**
  * irq_set_default_host - Set a "default" host
  * @host: default host pointer
  *
@@ -178,7 +178,7 @@ extern struct irq_host *irq_find_host(struct device_node *node);
 extern void irq_set_default_host(struct irq_host *host);
 
 
-/***
+/**
  * irq_set_virq_count - Set the maximum number of virt irqs
  * @count: number of linux virtual irqs, capped with NR_IRQS
  *
@@ -188,7 +188,7 @@ extern void irq_set_default_host(struct irq_host *host);
 extern void irq_set_virq_count(unsigned int count);
 
 
-/***
+/**
  * irq_create_mapping - Map a hardware interrupt into linux virq space
  * @host: host owning this hardware interrupt or NULL for default host
  * @hwirq: hardware irq number in that host space
@@ -202,13 +202,13 @@ extern unsigned int irq_create_mapping(struct irq_host *host,
 				       irq_hw_number_t hwirq);
 
 
-/***
+/**
  * irq_dispose_mapping - Unmap an interrupt
  * @virq: linux virq number of the interrupt to unmap
  */
 extern void irq_dispose_mapping(unsigned int virq);
 
-/***
+/**
  * irq_find_mapping - Find a linux virq from an hw irq number.
  * @host: host owning this hardware interrupt
  * @hwirq: hardware irq number in that host space
@@ -221,7 +221,7 @@ extern unsigned int irq_find_mapping(struct irq_host *host,
 				     irq_hw_number_t hwirq);
 
 
-/***
+/**
  * irq_radix_revmap - Find a linux virq from a hw irq number.
  * @host: host owning this hardware interrupt
  * @hwirq: hardware irq number in that host space
@@ -232,7 +232,7 @@ extern unsigned int irq_find_mapping(struct irq_host *host,
 extern unsigned int irq_radix_revmap(struct irq_host *host,
 				     irq_hw_number_t hwirq);
 
-/***
+/**
  * irq_linear_revmap - Find a linux virq from a hw irq number.
  * @host: host owning this hardware interrupt
  * @hwirq: hardware irq number in that host space
@@ -247,7 +247,7 @@ extern unsigned int irq_linear_revmap(struct irq_host *host,
 
 
 
-/***
+/**
  * irq_alloc_virt - Allocate virtual irq numbers
  * @host: host owning these new virtual irqs
  * @count: number of consecutive numbers to allocate
@@ -261,7 +261,7 @@ extern unsigned int irq_alloc_virt(struct irq_host *host,
 				   unsigned int count,
 				   unsigned int hint);
 
-/***
+/**
  * irq_free_virt - Free virtual irq numbers
  * @virq: virtual irq number of the first interrupt to free
  * @count: number of interrupts to free
@@ -300,7 +300,7 @@ extern unsigned int irq_of_parse_and_map(struct device_node *dev, int index);
 
 /* -- End OF helpers -- */
 
-/***
+/**
  * irq_early_init - Init irq remapping subsystem
  */
 extern void irq_early_init(void);
diff --git a/include/asm-powerpc/prom.h b/include/asm-powerpc/prom.h
index 31bfea4686a6..7a457bd462a2 100644
--- a/include/asm-powerpc/prom.h
+++ b/include/asm-powerpc/prom.h
@@ -259,7 +259,7 @@ struct of_irq {
 	u32 specifier[OF_MAX_IRQ_SPEC];	/* Specifier copy */
 };
 
-/***
+/**
  * of_irq_map_init - Initialize the irq remapper
  * @flags:	flags defining workarounds to enable
  *
@@ -272,7 +272,7 @@ struct of_irq {
 
 extern void of_irq_map_init(unsigned int flags);
 
-/***
+/**
  * of_irq_map_raw - Low level interrupt tree parsing
  * @parent:	the device interrupt parent
  * @intspec:	interrupt specifier ("interrupts" property of the device)
@@ -292,7 +292,7 @@ extern int of_irq_map_raw(struct device_node *parent, const u32 *intspec,
 			  const u32 *addr, struct of_irq *out_irq);
 
 
-/***
+/**
  * of_irq_map_one - Resolve an interrupt for a device
  * @device:	the device whose interrupt is to be resolved
  * @index:     	index of the interrupt to resolve
@@ -305,7 +305,7 @@ extern int of_irq_map_raw(struct device_node *parent, const u32 *intspec,
 extern int of_irq_map_one(struct device_node *device, int index,
 			  struct of_irq *out_irq);
 
-/***
+/**
  * of_irq_map_pci - Resolve the interrupt for a PCI device
  * @pdev:	the device whose interrupt is to be resolved
  * @out_irq:	structure of_irq filled by this function
-- 
cgit v1.2.3


From 2f6093c84730b4bad65bcd0f2f904a5769b1dfc5 Mon Sep 17 00:00:00 2001
From: Michael Neuling <mikey@neuling.org>
Date: Mon, 7 Aug 2006 16:19:19 +1000
Subject: [POWERPC] Implement SLB shadow buffer

This adds a shadow buffer for the SLBs and regsiters it with PHYP.
Only the bolted SLB entries (top 3) are shadowed.

The SLB shadow buffer tells the hypervisor what the kernel needs to
have in the SLB for the kernel to be able to function.  The hypervisor
can use this information to speed up partition context switches.

Signed-off-by: Michael Neuling <mikey@neuling.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 arch/powerpc/kernel/asm-offsets.c               |  2 ++
 arch/powerpc/kernel/entry_64.S                  | 13 +++++++++
 arch/powerpc/kernel/paca.c                      | 15 +++++++++-
 arch/powerpc/mm/slb.c                           | 37 ++++++++++++++++++++++---
 arch/powerpc/platforms/pseries/lpar.c           | 24 +++++++++++++---
 arch/powerpc/platforms/pseries/plpar_wrappers.h | 10 +++++++
 arch/powerpc/platforms/pseries/setup.c          | 12 ++++++--
 include/asm-powerpc/lppaca.h                    | 19 +++++++++++++
 include/asm-powerpc/paca.h                      |  3 ++
 9 files changed, 124 insertions(+), 11 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c
index ac0631958b20..2ef7ea860379 100644
--- a/arch/powerpc/kernel/asm-offsets.c
+++ b/arch/powerpc/kernel/asm-offsets.c
@@ -135,11 +135,13 @@ int main(void)
 	DEFINE(PACA_STARTPURR, offsetof(struct paca_struct, startpurr));
 	DEFINE(PACA_USER_TIME, offsetof(struct paca_struct, user_time));
 	DEFINE(PACA_SYSTEM_TIME, offsetof(struct paca_struct, system_time));
+	DEFINE(PACA_SLBSHADOWPTR, offsetof(struct paca_struct, slb_shadow_ptr));
 
 	DEFINE(LPPACASRR0, offsetof(struct lppaca, saved_srr0));
 	DEFINE(LPPACASRR1, offsetof(struct lppaca, saved_srr1));
 	DEFINE(LPPACAANYINT, offsetof(struct lppaca, int_dword.any_int));
 	DEFINE(LPPACADECRINT, offsetof(struct lppaca, int_dword.fields.decr_int));
+	DEFINE(SLBSHADOW_SAVEAREA, offsetof(struct slb_shadow, save_area));
 #endif /* CONFIG_PPC64 */
 
 	/* RTAS */
diff --git a/arch/powerpc/kernel/entry_64.S b/arch/powerpc/kernel/entry_64.S
index 54d9f5cdaab4..5baea498ea64 100644
--- a/arch/powerpc/kernel/entry_64.S
+++ b/arch/powerpc/kernel/entry_64.S
@@ -323,6 +323,11 @@ _GLOBAL(ret_from_fork)
  * The code which creates the new task context is in 'copy_thread'
  * in arch/powerpc/kernel/process.c 
  */
+#define SHADOW_SLB_BOLTED_STACK_ESID \
+		(SLBSHADOW_SAVEAREA + 0x10*(SLB_NUM_BOLTED-1))
+#define SHADOW_SLB_BOLTED_STACK_VSID \
+		(SLBSHADOW_SAVEAREA + 0x10*(SLB_NUM_BOLTED-1) + 8)
+
 	.align	7
 _GLOBAL(_switch)
 	mflr	r0
@@ -375,6 +380,14 @@ BEGIN_FTR_SECTION
 	ld	r7,KSP_VSID(r4)	/* Get new stack's VSID */
 	oris	r0,r6,(SLB_ESID_V)@h
 	ori	r0,r0,(SLB_NUM_BOLTED-1)@l
+
+	/* Update the last bolted SLB */
+	ld	r9,PACA_SLBSHADOWPTR(r13)
+ 	li	r12,0
+  	std	r12,SHADOW_SLB_BOLTED_STACK_ESID(r9) /* Clear ESID */
+	std	r7,SHADOW_SLB_BOLTED_STACK_VSID(r9)  /* Save VSID */
+ 	std	r0,SHADOW_SLB_BOLTED_STACK_ESID(r9)  /* Save ESID */
+
 	slbie	r6
 	slbie	r6		/* Workaround POWER5 < DD2.1 issue */
 	slbmte	r7,r0
diff --git a/arch/powerpc/kernel/paca.c b/arch/powerpc/kernel/paca.c
index c68741fed14b..55f1a25085cd 100644
--- a/arch/powerpc/kernel/paca.c
+++ b/arch/powerpc/kernel/paca.c
@@ -17,6 +17,7 @@
 #include <asm/lppaca.h>
 #include <asm/iseries/it_lp_reg_save.h>
 #include <asm/paca.h>
+#include <asm/mmu.h>
 
 
 /* This symbol is provided by the linker - let it fill in the paca
@@ -45,6 +46,17 @@ struct lppaca lppaca[] = {
 	},
 };
 
+/*
+ * 3 persistent SLBs are registered here.  The buffer will be zero
+ * initially, hence will all be invaild until we actually write them.
+ */
+struct slb_shadow slb_shadow[] __cacheline_aligned = {
+	[0 ... (NR_CPUS-1)] = {
+		.persistent = SLB_NUM_BOLTED,
+		.buffer_length = sizeof(struct slb_shadow),
+	},
+};
+
 /* The Paca is an array with one entry per processor.  Each contains an
  * lppaca, which contains the information shared between the
  * hypervisor and Linux.
@@ -59,7 +71,8 @@ struct lppaca lppaca[] = {
 	.lock_token = 0x8000,						    \
 	.paca_index = (number),		/* Paca Index */		    \
 	.kernel_toc = (unsigned long)(&__toc_start) + 0x8000UL,		    \
-	.hw_cpu_id = 0xffff,
+	.hw_cpu_id = 0xffff,						    \
+	.slb_shadow_ptr = &slb_shadow[number],
 
 #ifdef CONFIG_PPC_ISERIES
 #define PACA_INIT_ISERIES(number)					    \
diff --git a/arch/powerpc/mm/slb.c b/arch/powerpc/mm/slb.c
index de0c8842415c..d3733912adb4 100644
--- a/arch/powerpc/mm/slb.c
+++ b/arch/powerpc/mm/slb.c
@@ -22,6 +22,8 @@
 #include <asm/paca.h>
 #include <asm/cputable.h>
 #include <asm/cacheflush.h>
+#include <asm/smp.h>
+#include <linux/compiler.h>
 
 #ifdef DEBUG
 #define DBG(fmt...) udbg_printf(fmt)
@@ -50,9 +52,32 @@ static inline unsigned long mk_vsid_data(unsigned long ea, unsigned long flags)
 	return (get_kernel_vsid(ea) << SLB_VSID_SHIFT) | flags;
 }
 
-static inline void create_slbe(unsigned long ea, unsigned long flags,
-			       unsigned long entry)
+static inline void slb_shadow_update(unsigned long esid, unsigned long vsid,
+				     unsigned long entry)
 {
+	/*
+	 * Clear the ESID first so the entry is not valid while we are
+	 * updating it.
+	 */
+	get_slb_shadow()->save_area[entry].esid = 0;
+	barrier();
+	get_slb_shadow()->save_area[entry].vsid = vsid;
+	barrier();
+	get_slb_shadow()->save_area[entry].esid = esid;
+
+}
+
+static inline void create_shadowed_slbe(unsigned long ea, unsigned long flags,
+					unsigned long entry)
+{
+	/*
+	 * Updating the shadow buffer before writing the SLB ensures
+	 * we don't get a stale entry here if we get preempted by PHYP
+	 * between these two statements.
+	 */
+	slb_shadow_update(mk_esid_data(ea, entry), mk_vsid_data(ea, flags),
+			  entry);
+
 	asm volatile("slbmte  %0,%1" :
 		     : "r" (mk_vsid_data(ea, flags)),
 		       "r" (mk_esid_data(ea, entry))
@@ -77,6 +102,10 @@ void slb_flush_and_rebolt(void)
 	if ((ksp_esid_data & ESID_MASK) == PAGE_OFFSET)
 		ksp_esid_data &= ~SLB_ESID_V;
 
+	/* Only third entry (stack) may change here so only resave that */
+	slb_shadow_update(ksp_esid_data,
+			  mk_vsid_data(ksp_esid_data, lflags), 2);
+
 	/* We need to do this all in asm, so we're sure we don't touch
 	 * the stack between the slbia and rebolting it. */
 	asm volatile("isync\n"
@@ -209,9 +238,9 @@ void slb_initialize(void)
 	asm volatile("isync":::"memory");
 	asm volatile("slbmte  %0,%0"::"r" (0) : "memory");
 	asm volatile("isync; slbia; isync":::"memory");
-	create_slbe(PAGE_OFFSET, lflags, 0);
+	create_shadowed_slbe(PAGE_OFFSET, lflags, 0);
 
-	create_slbe(VMALLOC_START, vflags, 1);
+	create_shadowed_slbe(VMALLOC_START, vflags, 1);
 
 	/* We don't bolt the stack for the time being - we're in boot,
 	 * so the stack is in the bolted segment.  By the time it goes
diff --git a/arch/powerpc/platforms/pseries/lpar.c b/arch/powerpc/platforms/pseries/lpar.c
index 6cbf14266d5e..1820a0b0a8c6 100644
--- a/arch/powerpc/platforms/pseries/lpar.c
+++ b/arch/powerpc/platforms/pseries/lpar.c
@@ -252,18 +252,34 @@ out:
 void vpa_init(int cpu)
 {
 	int hwcpu = get_hard_smp_processor_id(cpu);
-	unsigned long vpa = __pa(&lppaca[cpu]);
+	unsigned long addr;
 	long ret;
 
 	if (cpu_has_feature(CPU_FTR_ALTIVEC))
 		lppaca[cpu].vmxregs_in_use = 1;
 
-	ret = register_vpa(hwcpu, vpa);
+	addr = __pa(&lppaca[cpu]);
+	ret = register_vpa(hwcpu, addr);
 
-	if (ret)
+	if (ret) {
 		printk(KERN_ERR "WARNING: vpa_init: VPA registration for "
 				"cpu %d (hw %d) of area %lx returns %ld\n",
-				cpu, hwcpu, vpa, ret);
+				cpu, hwcpu, addr, ret);
+		return;
+	}
+	/*
+	 * PAPR says this feature is SLB-Buffer but firmware never
+	 * reports that.  All SPLPAR support SLB shadow buffer.
+	 */
+	addr = __pa(&slb_shadow[cpu]);
+	if (firmware_has_feature(FW_FEATURE_SPLPAR)) {
+		ret = register_slb_shadow(hwcpu, addr);
+		if (ret)
+			printk(KERN_ERR
+			       "WARNING: vpa_init: SLB shadow buffer "
+			       "registration for cpu %d (hw %d) of area %lx "
+			       "returns %ld\n", cpu, hwcpu, addr, ret);
+	}
 }
 
 long pSeries_lpar_hpte_insert(unsigned long hpte_group,
diff --git a/arch/powerpc/platforms/pseries/plpar_wrappers.h b/arch/powerpc/platforms/pseries/plpar_wrappers.h
index ebd15de7597e..3eb7b294d92f 100644
--- a/arch/powerpc/platforms/pseries/plpar_wrappers.h
+++ b/arch/powerpc/platforms/pseries/plpar_wrappers.h
@@ -37,6 +37,16 @@ static inline long register_vpa(unsigned long cpu, unsigned long vpa)
 	return vpa_call(0x1, cpu, vpa);
 }
 
+static inline long unregister_slb_shadow(unsigned long cpu, unsigned long vpa)
+{
+	return vpa_call(0x7, cpu, vpa);
+}
+
+static inline long register_slb_shadow(unsigned long cpu, unsigned long vpa)
+{
+	return vpa_call(0x3, cpu, vpa);
+}
+
 extern void vpa_init(int cpu);
 
 static inline long plpar_pte_enter(unsigned long flags,
diff --git a/arch/powerpc/platforms/pseries/setup.c b/arch/powerpc/platforms/pseries/setup.c
index de214d86ff44..6ebeecfd6bcb 100644
--- a/arch/powerpc/platforms/pseries/setup.c
+++ b/arch/powerpc/platforms/pseries/setup.c
@@ -234,9 +234,17 @@ static void pseries_kexec_cpu_down_xics(int crash_shutdown, int secondary)
 {
 	/* Don't risk a hypervisor call if we're crashing */
 	if (firmware_has_feature(FW_FEATURE_SPLPAR) && !crash_shutdown) {
-		unsigned long vpa = __pa(get_lppaca());
+		unsigned long addr;
 
-		if (unregister_vpa(hard_smp_processor_id(), vpa)) {
+		addr = __pa(get_slb_shadow());
+		if (unregister_slb_shadow(hard_smp_processor_id(), addr))
+			printk("SLB shadow buffer deregistration of "
+			       "cpu %u (hw_cpu_id %d) failed\n",
+			       smp_processor_id(),
+			       hard_smp_processor_id());
+
+		addr = __pa(get_lppaca());
+		if (unregister_vpa(hard_smp_processor_id(), addr)) {
 			printk("VPA deregistration of cpu %u (hw_cpu_id %d) "
 					"failed\n", smp_processor_id(),
 					hard_smp_processor_id());
diff --git a/include/asm-powerpc/lppaca.h b/include/asm-powerpc/lppaca.h
index 4dc514aabfe7..942bb450baff 100644
--- a/include/asm-powerpc/lppaca.h
+++ b/include/asm-powerpc/lppaca.h
@@ -27,7 +27,9 @@
 //
 //
 //----------------------------------------------------------------------------
+#include <linux/cache.h>
 #include <asm/types.h>
+#include <asm/mmu.h>
 
 /* The Hypervisor barfs if the lppaca crosses a page boundary.  A 1k
  * alignment is sufficient to prevent this */
@@ -133,5 +135,22 @@ struct lppaca {
 
 extern struct lppaca lppaca[];
 
+/*
+ * SLB shadow buffer structure as defined in the PAPR.  The save_area
+ * contains adjacent ESID and VSID pairs for each shadowed SLB.  The
+ * ESID is stored in the lower 64bits, then the VSID.
+ */
+struct slb_shadow {
+	u32	persistent;		// Number of persistent SLBs	x00-x03
+	u32	buffer_length;		// Total shadow buffer length	x04-x07
+	u64	reserved;		// Alignment			x08-x0f
+	struct	{
+		u64     esid;
+		u64	vsid;
+	} save_area[SLB_NUM_BOLTED];	//				x10-x40
+} ____cacheline_aligned;
+
+extern struct slb_shadow slb_shadow[];
+
 #endif /* __KERNEL__ */
 #endif /* _ASM_POWERPC_LPPACA_H */
diff --git a/include/asm-powerpc/paca.h b/include/asm-powerpc/paca.h
index 2d4585f06209..7ffa2512524e 100644
--- a/include/asm-powerpc/paca.h
+++ b/include/asm-powerpc/paca.h
@@ -23,6 +23,7 @@
 register struct paca_struct *local_paca asm("r13");
 #define get_paca()	local_paca
 #define get_lppaca()	(get_paca()->lppaca_ptr)
+#define get_slb_shadow()	(get_paca()->slb_shadow_ptr)
 
 struct task_struct;
 
@@ -98,6 +99,8 @@ struct paca_struct {
 	u64 user_time;			/* accumulated usermode TB ticks */
 	u64 system_time;		/* accumulated system TB ticks */
 	u64 startpurr;			/* PURR/TB value snapshot */
+
+	struct slb_shadow *slb_shadow_ptr;
 };
 
 extern struct paca_struct paca[];
-- 
cgit v1.2.3


From 5cf13911b1e72707b6f0eb39b2d819ec6e343d76 Mon Sep 17 00:00:00 2001
From: Michael Neuling <mikey@neuling.org>
Date: Mon, 7 Aug 2006 17:34:50 +1000
Subject: [POWERPC] Update lppaca offset comments

Update offset comments.  No functional change.

Signed-off-by: Michael Neuling <mikey@neuling.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 include/asm-powerpc/lppaca.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/include/asm-powerpc/lppaca.h b/include/asm-powerpc/lppaca.h
index 942bb450baff..821ea0c512b4 100644
--- a/include/asm-powerpc/lppaca.h
+++ b/include/asm-powerpc/lppaca.h
@@ -116,7 +116,7 @@ struct lppaca {
 
 
 //=============================================================================
-// CACHE_LINE_3 0x0100 - 0x007F: This line is shared with other processors
+// CACHE_LINE_3 0x0100 - 0x017F: This line is shared with other processors
 //=============================================================================
 	// This is the yield_count.  An "odd" value (low bit on) means that
 	// the processor is yielded (either because of an OS yield or a PLIC
@@ -128,7 +128,7 @@ struct lppaca {
 	u8	reserved6[124];		// Reserved                     x04-x7F
 
 //=============================================================================
-// CACHE_LINE_4-5 0x0100 - 0x01FF Contains PMC interrupt data
+// CACHE_LINE_4-5 0x0180 - 0x027F Contains PMC interrupt data
 //=============================================================================
 	u8	pmc_save_area[256];	// PMC interrupt Area           x00-xFF
 } __attribute__((__aligned__(0x400)));
-- 
cgit v1.2.3


From b03732f006bd1ecee32587ec8235c41af5ad905f Mon Sep 17 00:00:00 2001
From: Brian King <brking@us.ibm.com>
Date: Mon, 7 Aug 2006 14:27:10 -0500
Subject: [PATCH] libata: Add ata_host_set_init

Add ata_host_set_init in preparation for SAS attached SATA.

Signed-off-by: Brian King <brking@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/scsi/libata-core.c | 28 ++++++++++++++++++++++++----
 include/linux/libata.h     |  2 ++
 2 files changed, 26 insertions(+), 4 deletions(-)

(limited to 'include')

diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c
index 351544d3653a..886440b128a5 100644
--- a/drivers/scsi/libata-core.c
+++ b/drivers/scsi/libata-core.c
@@ -5350,6 +5350,28 @@ err_out:
 	return NULL;
 }
 
+/**
+ *	ata_sas_host_init - Initialize a host_set struct
+ *	@host_set:	host_set to initialize
+ *	@dev:		device host_set is attached to
+ *	@flags:	host_set flags
+ *	@ops:		port_ops
+ *
+ *	LOCKING:
+ *	PCI/etc. bus probe sem.
+ *
+ */
+
+void ata_host_set_init(struct ata_host_set *host_set,
+		       struct device *dev, unsigned long flags,
+		       const struct ata_port_operations *ops)
+{
+	spin_lock_init(&host_set->lock);
+	host_set->dev = dev;
+	host_set->flags = flags;
+	host_set->ops = ops;
+}
+
 /**
  *	ata_device_add - Register hardware device with ATA and SCSI layers
  *	@ent: Probe information describing hardware device to be registered
@@ -5381,15 +5403,12 @@ int ata_device_add(const struct ata_probe_ent *ent)
 			   (ent->n_ports * sizeof(void *)), GFP_KERNEL);
 	if (!host_set)
 		return 0;
-	spin_lock_init(&host_set->lock);
 
-	host_set->dev = dev;
+	ata_host_set_init(host_set, dev, ent->host_set_flags, ent->port_ops);
 	host_set->n_ports = ent->n_ports;
 	host_set->irq = ent->irq;
 	host_set->mmio_base = ent->mmio_base;
 	host_set->private_data = ent->private_data;
-	host_set->ops = ent->port_ops;
-	host_set->flags = ent->host_set_flags;
 
 	/* register each port bound to this device */
 	for (i = 0; i < ent->n_ports; i++) {
@@ -5908,6 +5927,7 @@ EXPORT_SYMBOL_GPL(sata_deb_timing_hotplug);
 EXPORT_SYMBOL_GPL(sata_deb_timing_long);
 EXPORT_SYMBOL_GPL(ata_std_bios_param);
 EXPORT_SYMBOL_GPL(ata_std_ports);
+EXPORT_SYMBOL_GPL(ata_host_set_init);
 EXPORT_SYMBOL_GPL(ata_device_add);
 EXPORT_SYMBOL_GPL(ata_port_detach);
 EXPORT_SYMBOL_GPL(ata_host_set_remove);
diff --git a/include/linux/libata.h b/include/linux/libata.h
index b9416708bba2..be15ef5d8d85 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -684,6 +684,8 @@ extern int ata_pci_clear_simplex(struct pci_dev *pdev);
 #endif /* CONFIG_PCI */
 extern int ata_device_add(const struct ata_probe_ent *ent);
 extern void ata_port_detach(struct ata_port *ap);
+extern void ata_host_set_init(struct ata_host_set *, struct device *,
+			      unsigned long, const struct ata_port_operations *);
 extern void ata_host_set_remove(struct ata_host_set *host_set);
 extern int ata_scsi_detect(struct scsi_host_template *sht);
 extern int ata_scsi_ioctl(struct scsi_device *dev, int cmd, void __user *arg);
-- 
cgit v1.2.3


From 80289167fd3ebaeb7b2641e69cbec44b61165fe7 Mon Sep 17 00:00:00 2001
From: Brian King <brking@us.ibm.com>
Date: Mon, 7 Aug 2006 14:27:31 -0500
Subject: [PATCH] libata: Add support for SATA attachment to SAS adapters

The following patch enhances libata to allow SAS device drivers
to utilize libata to talk to SATA devices. It introduces some
new APIs which allow libata to be used without allocating a
virtual scsi host.

New APIs:

ata_sas_port_alloc - Allocate an ata_port
ata_sas_port_init - Initialize an ata_port (probe device, etc)
ata_sas_port_destroy - Free an ata_port allocated by ata_sas_port_alloc
ata_sas_slave_configure - configure scsi device
ata_sas_queuecmd - queue a scsi command, similar to ata_scsi_queuecomand

These new APIs can be used either directly by a SAS LLDD or could be used
by the SAS transport class.

Possible usage for a SAS LLDD would be:

scsi_scan_host
	target_alloc
		ata_sas_port_alloc
	slave_alloc
		ata_sas_port_init
	slave_configure
		ata_sas_slave_configure

Commands received by the LLDD for SATA devices would call ata_sas_queuecmd.

Device teardown would occur with:

slave_destroy
	port_disable
target_destroy
	ata_sas_port_destroy

Signed-off-by: Brian King <brking@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/scsi/libata-core.c |   2 +-
 drivers/scsi/libata-scsi.c | 149 +++++++++++++++++++++++++++++++++++++++++++++
 drivers/scsi/libata.h      |   1 +
 include/linux/libata.h     |   9 +++
 4 files changed, 160 insertions(+), 1 deletion(-)

(limited to 'include')

diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c
index dc35cccd2898..83d93fc0751b 100644
--- a/drivers/scsi/libata-core.c
+++ b/drivers/scsi/libata-core.c
@@ -1528,7 +1528,7 @@ err_out_nosup:
  *	Zero on success, negative errno otherwise.
  */
 
-static int ata_bus_probe(struct ata_port *ap)
+int ata_bus_probe(struct ata_port *ap)
 {
 	unsigned int classes[ATA_MAX_DEVICES];
 	int tries[ATA_MAX_DEVICES];
diff --git a/drivers/scsi/libata-scsi.c b/drivers/scsi/libata-scsi.c
index 1638e57028cb..37ec02661433 100644
--- a/drivers/scsi/libata-scsi.c
+++ b/drivers/scsi/libata-scsi.c
@@ -3158,3 +3158,152 @@ void ata_scsi_dev_rescan(void *data)
 			scsi_rescan_device(&(dev->sdev->sdev_gendev));
 	}
 }
+
+/**
+ *	ata_sas_port_alloc - Allocate port for a SAS attached SATA device
+ *	@pdev: PCI device that the scsi device is attached to
+ *	@port_info: Information from low-level host driver
+ *	@host: SCSI host that the scsi device is attached to
+ *
+ *	LOCKING:
+ *	PCI/etc. bus probe sem.
+ *
+ *	RETURNS:
+ *	ata_port pointer on success / NULL on failure.
+ */
+
+struct ata_port *ata_sas_port_alloc(struct ata_host_set *host_set,
+				    struct ata_port_info *port_info,
+				    struct Scsi_Host *host)
+{
+	struct ata_port *ap = kzalloc(sizeof(*ap), GFP_KERNEL);
+	struct ata_probe_ent *ent;
+
+	if (!ap)
+		return NULL;
+
+	ent = ata_probe_ent_alloc(host_set->dev, port_info);
+	if (!ent) {
+		kfree(ap);
+		return NULL;
+	}
+
+	ata_port_init(ap, host_set, ent, 0);
+	ap->lock = host->host_lock;
+	kfree(ent);
+	return ap;
+}
+EXPORT_SYMBOL_GPL(ata_sas_port_alloc);
+
+/**
+ *	ata_sas_port_start - Set port up for dma.
+ *	@ap: Port to initialize
+ *
+ *	Called just after data structures for each port are
+ *	initialized.  Allocates DMA pad.
+ *
+ *	May be used as the port_start() entry in ata_port_operations.
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ */
+int ata_sas_port_start(struct ata_port *ap)
+{
+	return ata_pad_alloc(ap, ap->dev);
+}
+EXPORT_SYMBOL_GPL(ata_sas_port_start);
+
+/**
+ *	ata_port_stop - Undo ata_sas_port_start()
+ *	@ap: Port to shut down
+ *
+ *	Frees the DMA pad.
+ *
+ *	May be used as the port_stop() entry in ata_port_operations.
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ */
+
+void ata_sas_port_stop(struct ata_port *ap)
+{
+	ata_pad_free(ap, ap->dev);
+}
+EXPORT_SYMBOL_GPL(ata_sas_port_stop);
+
+/**
+ *	ata_sas_port_init - Initialize a SATA device
+ *	@ap: SATA port to initialize
+ *
+ *	LOCKING:
+ *	PCI/etc. bus probe sem.
+ *
+ *	RETURNS:
+ *	Zero on success, non-zero on error.
+ */
+
+int ata_sas_port_init(struct ata_port *ap)
+{
+	int rc = ap->ops->port_start(ap);
+
+	if (!rc)
+		rc = ata_bus_probe(ap);
+
+	return rc;
+}
+EXPORT_SYMBOL_GPL(ata_sas_port_init);
+
+/**
+ *	ata_sas_port_destroy - Destroy a SATA port allocated by ata_sas_port_alloc
+ *	@ap: SATA port to destroy
+ *
+ */
+
+void ata_sas_port_destroy(struct ata_port *ap)
+{
+	ap->ops->port_stop(ap);
+	kfree(ap);
+}
+EXPORT_SYMBOL_GPL(ata_sas_port_destroy);
+
+/**
+ *	ata_sas_slave_configure - Default slave_config routine for libata devices
+ *	@sdev: SCSI device to configure
+ *	@ap: ATA port to which SCSI device is attached
+ *
+ *	RETURNS:
+ *	Zero.
+ */
+
+int ata_sas_slave_configure(struct scsi_device *sdev, struct ata_port *ap)
+{
+	ata_scsi_sdev_config(sdev);
+	ata_scsi_dev_config(sdev, ap->device);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ata_sas_slave_configure);
+
+/**
+ *	ata_sas_queuecmd - Issue SCSI cdb to libata-managed device
+ *	@cmd: SCSI command to be sent
+ *	@done: Completion function, called when command is complete
+ *	@ap:	ATA port to which the command is being sent
+ *
+ *	RETURNS:
+ *	Zero.
+ */
+
+int ata_sas_queuecmd(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *),
+		     struct ata_port *ap)
+{
+	ata_scsi_dump_cdb(ap, cmd);
+
+	if (likely(ata_scsi_dev_enabled(ap->device)))
+		__ata_scsi_queuecmd(cmd, done, ap->device);
+	else {
+		cmd->result = (DID_BAD_TARGET << 16);
+		done(cmd);
+	}
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ata_sas_queuecmd);
diff --git a/drivers/scsi/libata.h b/drivers/scsi/libata.h
index 0b7a37c2785d..d4a4f82360ec 100644
--- a/drivers/scsi/libata.h
+++ b/drivers/scsi/libata.h
@@ -111,6 +111,7 @@ extern void ata_scsi_rbuf_fill(struct ata_scsi_args *args,
                                            u8 *rbuf, unsigned int buflen));
 extern void ata_schedule_scsi_eh(struct Scsi_Host *shost);
 extern void ata_scsi_dev_rescan(void *data);
+extern int ata_bus_probe(struct ata_port *ap);
 
 /* libata-eh.c */
 extern enum scsi_eh_timer_return ata_scsi_timed_out(struct scsi_cmnd *cmd);
diff --git a/include/linux/libata.h b/include/linux/libata.h
index be15ef5d8d85..cf5eb1da3e32 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -691,6 +691,15 @@ extern int ata_scsi_detect(struct scsi_host_template *sht);
 extern int ata_scsi_ioctl(struct scsi_device *dev, int cmd, void __user *arg);
 extern int ata_scsi_queuecmd(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *));
 extern int ata_scsi_release(struct Scsi_Host *host);
+extern void ata_sas_port_destroy(struct ata_port *);
+extern struct ata_port *ata_sas_port_alloc(struct ata_host_set *,
+					   struct ata_port_info *, struct Scsi_Host *);
+extern int ata_sas_port_init(struct ata_port *);
+extern int ata_sas_port_start(struct ata_port *ap);
+extern void ata_sas_port_stop(struct ata_port *ap);
+extern int ata_sas_slave_configure(struct scsi_device *, struct ata_port *);
+extern int ata_sas_queuecmd(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *),
+			    struct ata_port *ap);
 extern unsigned int ata_host_intr(struct ata_port *ap, struct ata_queued_cmd *qc);
 extern int sata_scr_valid(struct ata_port *ap);
 extern int sata_scr_read(struct ata_port *ap, int reg, u32 *val);
-- 
cgit v1.2.3


From 2ec7df0457b710d9201f211dbccdbecf0ad38b7e Mon Sep 17 00:00:00 2001
From: Alan Cox <alan@lxorguk.ukuu.org.uk>
Date: Thu, 10 Aug 2006 16:59:10 +0900
Subject: [PATCH] libata: rework legacy handling to remove much of the cruft

Kill host_set->next
Fix simplex support
Allow per platform setting of IDE legacy bases

Some of this can be tidied further later on, in particular all the
legacy port gunge belongs as a PCI quirk/PCI header decode to understand
the special legacy IDE rules in the PCI spec.

Longer term Jeff also wants to move the request_irq/free_irq out of core
which will make this even cleaner.

tj: folded in three followup patches - ata_piix-fix, broken-arch-fix
and fix-new-legacy-handling, and separated per-dev xfermask into
separate patch preceding this one.  Folded in fixes are...

* ata_piix-fix: fix build failure due to host_set->next removal
* broken-arch-fix: add missing include/asm-*/libata-portmap.h
* fix-new-legacy-handling:
	* In ata_pci_init_legacy_port(), probe_num was incorrectly
          incremented during initialization of the secondary port and
          probe_ent->n_ports was incorrectly fixed to 1.

	* Both legacy ports ended up having the same hard_port_no.

	* When printing port information, both legacy ports printed
	  the first irq.

Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/scsi/ata_piix.c              |   2 -
 drivers/scsi/libata-bmdma.c          | 135 +++++++++++++++++------------------
 drivers/scsi/libata-core.c           |  59 +++++++++------
 include/asm-alpha/libata-portmap.h   |   1 +
 include/asm-generic/libata-portmap.h |  12 ++++
 include/asm-i386/libata-portmap.h    |   1 +
 include/asm-ia64/libata-portmap.h    |   1 +
 include/asm-powerpc/libata-portmap.h |   1 +
 include/asm-sparc/libata-portmap.h   |   1 +
 include/asm-sparc64/libata-portmap.h |   1 +
 include/asm-x86_64/libata-portmap.h  |   1 +
 include/linux/libata.h               |   5 +-
 12 files changed, 123 insertions(+), 97 deletions(-)
 create mode 100644 include/asm-alpha/libata-portmap.h
 create mode 100644 include/asm-generic/libata-portmap.h
 create mode 100644 include/asm-i386/libata-portmap.h
 create mode 100644 include/asm-ia64/libata-portmap.h
 create mode 100644 include/asm-powerpc/libata-portmap.h
 create mode 100644 include/asm-sparc/libata-portmap.h
 create mode 100644 include/asm-sparc64/libata-portmap.h
 create mode 100644 include/asm-x86_64/libata-portmap.h

(limited to 'include')

diff --git a/drivers/scsi/ata_piix.c b/drivers/scsi/ata_piix.c
index 5e8afc876980..40ecab5793e3 100644
--- a/drivers/scsi/ata_piix.c
+++ b/drivers/scsi/ata_piix.c
@@ -932,8 +932,6 @@ static int piix_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 
 static void piix_host_stop(struct ata_host_set *host_set)
 {
-	if (host_set->next == NULL)
-		kfree(host_set->private_data);
 	ata_host_stop(host_set);
 }
 
diff --git a/drivers/scsi/libata-bmdma.c b/drivers/scsi/libata-bmdma.c
index 3268cf5375ea..e694f6075c3b 100644
--- a/drivers/scsi/libata-bmdma.c
+++ b/drivers/scsi/libata-bmdma.c
@@ -854,7 +854,7 @@ ata_pci_init_native_mode(struct pci_dev *pdev, struct ata_port_info **port, int
 		if (bmdma) {
 			bmdma += 8;
 			if(inb(bmdma + 2) & 0x80)
-			probe_ent->host_set_flags |= ATA_HOST_SIMPLEX;
+				probe_ent->host_set_flags |= ATA_HOST_SIMPLEX;
 			probe_ent->port[p].bmdma_addr = bmdma;
 		}
 		ata_std_ports(&probe_ent->port[p]);
@@ -867,44 +867,55 @@ ata_pci_init_native_mode(struct pci_dev *pdev, struct ata_port_info **port, int
 
 
 static struct ata_probe_ent *ata_pci_init_legacy_port(struct pci_dev *pdev,
-				struct ata_port_info *port, int port_num)
+				struct ata_port_info **port, int port_mask)
 {
 	struct ata_probe_ent *probe_ent;
-	unsigned long bmdma;
+	unsigned long bmdma = pci_resource_start(pdev, 4);
 
-	probe_ent = ata_probe_ent_alloc(pci_dev_to_dev(pdev), port);
+	int port_num = 0;
+
+	probe_ent = ata_probe_ent_alloc(pci_dev_to_dev(pdev), port[0]);
 	if (!probe_ent)
 		return NULL;
 
 	probe_ent->legacy_mode = 1;
-	probe_ent->n_ports = 1;
-	probe_ent->hard_port_no = port_num;
-	probe_ent->private_data = port->private_data;
-
-	switch(port_num)
-	{
-		case 0:
-			probe_ent->irq = 14;
-			probe_ent->port[0].cmd_addr = 0x1f0;
-			probe_ent->port[0].altstatus_addr =
-			probe_ent->port[0].ctl_addr = 0x3f6;
-			break;
-		case 1:
+	probe_ent->hard_port_no = 0;
+	probe_ent->private_data = port[0]->private_data;
+
+	if (port_mask & ATA_PORT_PRIMARY) {
+		probe_ent->irq = 14;
+		probe_ent->port[port_num].cmd_addr = ATA_PRIMARY_CMD;
+		probe_ent->port[port_num].altstatus_addr =
+		probe_ent->port[port_num].ctl_addr = ATA_PRIMARY_CTL;
+		if (bmdma) {
+			probe_ent->port[0].bmdma_addr = bmdma;
+			if (inb(bmdma + 2) & 0x80)
+				probe_ent->host_set_flags |= ATA_HOST_SIMPLEX;
+		}
+		ata_std_ports(&probe_ent->port[port_num]);
+		port_num ++;
+	}
+	if (port_mask & ATA_PORT_SECONDARY) {
+		if (port_num == 1)
+			probe_ent->irq2 = 15;
+		else {
+			/* Secondary only. IRQ 15 only and "first" port is port 1 */
 			probe_ent->irq = 15;
-			probe_ent->port[0].cmd_addr = 0x170;
-			probe_ent->port[0].altstatus_addr =
-			probe_ent->port[0].ctl_addr = 0x376;
-			break;
+			probe_ent->hard_port_no = 1;
+		}
+		probe_ent->port[port_num].cmd_addr = ATA_SECONDARY_CMD;
+		probe_ent->port[port_num].altstatus_addr =
+		probe_ent->port[port_num].ctl_addr = ATA_SECONDARY_CTL;
+		if (bmdma) {
+			probe_ent->port[port_num].bmdma_addr = bmdma + 8;
+			if (inb(bmdma + 10) & 0x80)
+				probe_ent->host_set_flags |= ATA_HOST_SIMPLEX;
+		}
+		ata_std_ports(&probe_ent->port[port_num]);
+		port_num ++;
 	}
 
-	bmdma = pci_resource_start(pdev, 4);
-	if (bmdma != 0) {
-		bmdma += 8 * port_num;
-		probe_ent->port[0].bmdma_addr = bmdma;
-		if (inb(bmdma + 2) & 0x80)
-			probe_ent->host_set_flags |= ATA_HOST_SIMPLEX;
-	}
-	ata_std_ports(&probe_ent->port[0]);
+	probe_ent->n_ports = port_num;
 
 	return probe_ent;
 }
@@ -924,6 +935,10 @@ static struct ata_probe_ent *ata_pci_init_legacy_port(struct pci_dev *pdev,
  *	regions, sets the dma mask, enables bus master mode, and calls
  *	ata_device_add()
  *
+ *	ASSUMPTION:
+ *	Nobody makes a single channel controller that appears solely as
+ *	the secondary legacy port on PCI.
+ *
  *	LOCKING:
  *	Inherited from PCI layer (may sleep).
  *
@@ -934,7 +949,7 @@ static struct ata_probe_ent *ata_pci_init_legacy_port(struct pci_dev *pdev,
 int ata_pci_init_one (struct pci_dev *pdev, struct ata_port_info **port_info,
 		      unsigned int n_ports)
 {
-	struct ata_probe_ent *probe_ent = NULL, *probe_ent2 = NULL;
+	struct ata_probe_ent *probe_ent = NULL;
 	struct ata_port_info *port[2];
 	u8 tmp8, mask;
 	unsigned int legacy_mode = 0;
@@ -983,35 +998,34 @@ int ata_pci_init_one (struct pci_dev *pdev, struct ata_port_info **port_info,
 		goto err_out;
 	}
 
-	/* FIXME: Should use platform specific mappers for legacy port ranges */
 	if (legacy_mode) {
-		if (!request_region(0x1f0, 8, "libata")) {
+		if (!request_region(ATA_PRIMARY_CMD, 8, "libata")) {
 			struct resource *conflict, res;
-			res.start = 0x1f0;
-			res.end = 0x1f0 + 8 - 1;
+			res.start = ATA_PRIMARY_CMD;
+			res.end = ATA_PRIMARY_CMD + 8 - 1;
 			conflict = ____request_resource(&ioport_resource, &res);
 			if (!strcmp(conflict->name, "libata"))
-				legacy_mode |= (1 << 0);
+				legacy_mode |= ATA_PORT_PRIMARY;
 			else {
 				disable_dev_on_err = 0;
-				printk(KERN_WARNING "ata: 0x1f0 IDE port busy\n");
+				printk(KERN_WARNING "ata: 0x%0X IDE port busy\n", ATA_PRIMARY_CMD);
 			}
 		} else
-			legacy_mode |= (1 << 0);
+			legacy_mode |= ATA_PORT_PRIMARY;
 
-		if (!request_region(0x170, 8, "libata")) {
+		if (!request_region(ATA_SECONDARY_CMD, 8, "libata")) {
 			struct resource *conflict, res;
-			res.start = 0x170;
-			res.end = 0x170 + 8 - 1;
+			res.start = ATA_SECONDARY_CMD;
+			res.end = ATA_SECONDARY_CMD + 8 - 1;
 			conflict = ____request_resource(&ioport_resource, &res);
 			if (!strcmp(conflict->name, "libata"))
-				legacy_mode |= (1 << 1);
+				legacy_mode |= ATA_PORT_SECONDARY;
 			else {
 				disable_dev_on_err = 0;
-				printk(KERN_WARNING "ata: 0x170 IDE port busy\n");
+				printk(KERN_WARNING "ata: 0x%X IDE port busy\n", ATA_SECONDARY_CMD);
 			}
 		} else
-			legacy_mode |= (1 << 1);
+			legacy_mode |= ATA_PORT_SECONDARY;
 	}
 
 	/* we have legacy mode, but all ports are unavailable */
@@ -1029,17 +1043,14 @@ int ata_pci_init_one (struct pci_dev *pdev, struct ata_port_info **port_info,
 		goto err_out_regions;
 
 	if (legacy_mode) {
-		if (legacy_mode & (1 << 0))
-			probe_ent = ata_pci_init_legacy_port(pdev, port[0], 0);
-		if (legacy_mode & (1 << 1))
-			probe_ent2 = ata_pci_init_legacy_port(pdev, port[1], 1);
+		probe_ent = ata_pci_init_legacy_port(pdev, port, legacy_mode);
 	} else {
 		if (n_ports == 2)
 			probe_ent = ata_pci_init_native_mode(pdev, port, ATA_PORT_PRIMARY | ATA_PORT_SECONDARY);
 		else
 			probe_ent = ata_pci_init_native_mode(pdev, port, ATA_PORT_PRIMARY);
 	}
-	if (!probe_ent && !probe_ent2) {
+	if (!probe_ent) {
 		rc = -ENOMEM;
 		goto err_out_regions;
 	}
@@ -1047,35 +1058,17 @@ int ata_pci_init_one (struct pci_dev *pdev, struct ata_port_info **port_info,
 	pci_set_master(pdev);
 
 	/* FIXME: check ata_device_add return */
-	if (legacy_mode) {
-		struct device *dev = &pdev->dev;
-		struct ata_host_set *host_set = NULL;
-
-		if (legacy_mode & (1 << 0)) {
-			ata_device_add(probe_ent);
-			host_set = dev_get_drvdata(dev);
-		}
-
-		if (legacy_mode & (1 << 1)) {
-			ata_device_add(probe_ent2);
-			if (host_set) {
-				host_set->next = dev_get_drvdata(dev);
-				dev_set_drvdata(dev, host_set);
-			}
-		}
-	} else
-		ata_device_add(probe_ent);
+	ata_device_add(probe_ent);
 
 	kfree(probe_ent);
-	kfree(probe_ent2);
 
 	return 0;
 
 err_out_regions:
-	if (legacy_mode & (1 << 0))
-		release_region(0x1f0, 8);
-	if (legacy_mode & (1 << 1))
-		release_region(0x170, 8);
+	if (legacy_mode & ATA_PORT_PRIMARY)
+		release_region(ATA_PRIMARY_CMD, 8);
+	if (legacy_mode & ATA_PORT_SECONDARY)
+		release_region(ATA_SECONDARY_CMD, 8);
 	pci_release_regions(pdev);
 err_out:
 	if (disable_dev_on_err)
diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c
index 4e6c2e8ac0f6..3634279d896d 100644
--- a/drivers/scsi/libata-core.c
+++ b/drivers/scsi/libata-core.c
@@ -5223,8 +5223,9 @@ void ata_port_init(struct ata_port *ap, struct ata_host_set *host_set,
 	ap->host_set = host_set;
 	ap->dev = ent->dev;
 	ap->port_no = port_no;
-	ap->hard_port_no =
-		ent->legacy_mode ? ent->hard_port_no : port_no;
+	ap->hard_port_no = port_no;
+	if (ent->legacy_mode)
+		ap->hard_port_no += ent->hard_port_no;
 	ap->pio_mask = ent->pio_mask;
 	ap->mwdma_mask = ent->mwdma_mask;
 	ap->udma_mask = ent->udma_mask;
@@ -5400,6 +5401,7 @@ int ata_device_add(const struct ata_probe_ent *ent)
 	ata_host_set_init(host_set, dev, ent->host_set_flags, ent->port_ops);
 	host_set->n_ports = ent->n_ports;
 	host_set->irq = ent->irq;
+	host_set->irq2 = ent->irq2;
 	host_set->mmio_base = ent->mmio_base;
 	host_set->private_data = ent->private_data;
 
@@ -5407,11 +5409,16 @@ int ata_device_add(const struct ata_probe_ent *ent)
 	for (i = 0; i < host_set->n_ports; i++) {
 		struct ata_port *ap;
 		unsigned long xfer_mode_mask;
+		int irq_line = ent->irq;
 
 		ap = ata_port_add(ent, host_set, i);
 		if (!ap)
 			goto err_out;
 
+		/* Report the secondary IRQ for second channel legacy */
+		if (i == 1 && ent->irq2)
+			irq_line = ent->irq2;
+
 		host_set->ports[i] = ap;
 		xfer_mode_mask =(ap->udma_mask << ATA_SHIFT_UDMA) |
 				(ap->mwdma_mask << ATA_SHIFT_MWDMA) |
@@ -5419,20 +5426,20 @@ int ata_device_add(const struct ata_probe_ent *ent)
 
 		/* print per-port info to dmesg */
 		ata_port_printk(ap, KERN_INFO, "%cATA max %s cmd 0x%lX "
-				"ctl 0x%lX bmdma 0x%lX irq %lu\n",
+				"ctl 0x%lX bmdma 0x%lX irq %d\n",
 				ap->flags & ATA_FLAG_SATA ? 'S' : 'P',
 				ata_mode_string(xfer_mode_mask),
 				ap->ioaddr.cmd_addr,
 				ap->ioaddr.ctl_addr,
 				ap->ioaddr.bmdma_addr,
-				ent->irq);
+				irq_line);
 
 		ata_chk_status(ap);
 		host_set->ops->irq_clear(ap);
 		ata_eh_freeze_port(ap);	/* freeze port before requesting IRQ */
 	}
 
-	/* obtain irq, that is shared between channels */
+	/* obtain irq, that may be shared between channels */
 	rc = request_irq(ent->irq, ent->port_ops->irq_handler, ent->irq_flags,
 			 DRV_NAME, host_set);
 	if (rc) {
@@ -5441,6 +5448,21 @@ int ata_device_add(const struct ata_probe_ent *ent)
 		goto err_out;
 	}
 
+	/* do we have a second IRQ for the other channel, eg legacy mode */
+	if (ent->irq2) {
+		/* We will get weird core code crashes later if this is true
+		   so trap it now */
+		BUG_ON(ent->irq == ent->irq2);
+
+		rc = request_irq(ent->irq2, ent->port_ops->irq_handler, ent->irq_flags,
+			 DRV_NAME, host_set);
+		if (rc) {
+			dev_printk(KERN_ERR, dev, "irq %lu request failed: %d\n",
+				   ent->irq2, rc);
+			goto err_out_free_irq;
+		}
+	}
+
 	/* perform each probe synchronously */
 	DPRINTK("probe begin\n");
 	for (i = 0; i < host_set->n_ports; i++) {
@@ -5514,6 +5536,8 @@ int ata_device_add(const struct ata_probe_ent *ent)
 	VPRINTK("EXIT, returning %u\n", ent->n_ports);
 	return ent->n_ports; /* success */
 
+err_out_free_irq:
+	free_irq(ent->irq, host_set);
 err_out:
 	for (i = 0; i < host_set->n_ports; i++) {
 		struct ata_port *ap = host_set->ports[i];
@@ -5605,6 +5629,8 @@ void ata_host_set_remove(struct ata_host_set *host_set)
 		ata_port_detach(host_set->ports[i]);
 
 	free_irq(host_set->irq, host_set);
+	if (host_set->irq2)
+		free_irq(host_set->irq2, host_set);
 
 	for (i = 0; i < host_set->n_ports; i++) {
 		struct ata_port *ap = host_set->ports[i];
@@ -5614,10 +5640,11 @@ void ata_host_set_remove(struct ata_host_set *host_set)
 		if ((ap->flags & ATA_FLAG_NO_LEGACY) == 0) {
 			struct ata_ioports *ioaddr = &ap->ioaddr;
 
-			if (ioaddr->cmd_addr == 0x1f0)
-				release_region(0x1f0, 8);
-			else if (ioaddr->cmd_addr == 0x170)
-				release_region(0x170, 8);
+			/* FIXME: Add -ac IDE pci mods to remove these special cases */
+			if (ioaddr->cmd_addr == ATA_PRIMARY_CMD)
+				release_region(ATA_PRIMARY_CMD, 8);
+			else if (ioaddr->cmd_addr == ATA_SECONDARY_CMD)
+				release_region(ATA_SECONDARY_CMD, 8);
 		}
 
 		scsi_host_put(ap->host);
@@ -5735,11 +5762,8 @@ void ata_pci_remove_one (struct pci_dev *pdev)
 {
 	struct device *dev = pci_dev_to_dev(pdev);
 	struct ata_host_set *host_set = dev_get_drvdata(dev);
-	struct ata_host_set *host_set2 = host_set->next;
 
 	ata_host_set_remove(host_set);
-	if (host_set2)
-		ata_host_set_remove(host_set2);
 
 	pci_release_regions(pdev);
 	pci_disable_device(pdev);
@@ -5807,14 +5831,6 @@ int ata_pci_device_suspend(struct pci_dev *pdev, pm_message_t mesg)
 	if (rc)
 		return rc;
 
-	if (host_set->next) {
-		rc = ata_host_set_suspend(host_set->next, mesg);
-		if (rc) {
-			ata_host_set_resume(host_set);
-			return rc;
-		}
-	}
-
 	ata_pci_device_do_suspend(pdev, mesg);
 
 	return 0;
@@ -5826,9 +5842,6 @@ int ata_pci_device_resume(struct pci_dev *pdev)
 
 	ata_pci_device_do_resume(pdev);
 	ata_host_set_resume(host_set);
-	if (host_set->next)
-		ata_host_set_resume(host_set->next);
-
 	return 0;
 }
 #endif /* CONFIG_PCI */
diff --git a/include/asm-alpha/libata-portmap.h b/include/asm-alpha/libata-portmap.h
new file mode 100644
index 000000000000..75484ef0c743
--- /dev/null
+++ b/include/asm-alpha/libata-portmap.h
@@ -0,0 +1 @@
+#include <asm-generic/libata-portmap.h>
diff --git a/include/asm-generic/libata-portmap.h b/include/asm-generic/libata-portmap.h
new file mode 100644
index 000000000000..9202fd02d5be
--- /dev/null
+++ b/include/asm-generic/libata-portmap.h
@@ -0,0 +1,12 @@
+#ifndef __ASM_GENERIC_LIBATA_PORTMAP_H
+#define __ASM_GENERIC_LIBATA_PORTMAP_H
+
+#define ATA_PRIMARY_CMD		0x1F0
+#define ATA_PRIMARY_CTL		0x3F6
+#define ATA_PRIMARY_IRQ		14
+
+#define ATA_SECONDARY_CMD	0x170
+#define ATA_SECONDARY_CTL	0x376
+#define ATA_SECONDARY_IRQ	15
+
+#endif
diff --git a/include/asm-i386/libata-portmap.h b/include/asm-i386/libata-portmap.h
new file mode 100644
index 000000000000..75484ef0c743
--- /dev/null
+++ b/include/asm-i386/libata-portmap.h
@@ -0,0 +1 @@
+#include <asm-generic/libata-portmap.h>
diff --git a/include/asm-ia64/libata-portmap.h b/include/asm-ia64/libata-portmap.h
new file mode 100644
index 000000000000..75484ef0c743
--- /dev/null
+++ b/include/asm-ia64/libata-portmap.h
@@ -0,0 +1 @@
+#include <asm-generic/libata-portmap.h>
diff --git a/include/asm-powerpc/libata-portmap.h b/include/asm-powerpc/libata-portmap.h
new file mode 100644
index 000000000000..75484ef0c743
--- /dev/null
+++ b/include/asm-powerpc/libata-portmap.h
@@ -0,0 +1 @@
+#include <asm-generic/libata-portmap.h>
diff --git a/include/asm-sparc/libata-portmap.h b/include/asm-sparc/libata-portmap.h
new file mode 100644
index 000000000000..75484ef0c743
--- /dev/null
+++ b/include/asm-sparc/libata-portmap.h
@@ -0,0 +1 @@
+#include <asm-generic/libata-portmap.h>
diff --git a/include/asm-sparc64/libata-portmap.h b/include/asm-sparc64/libata-portmap.h
new file mode 100644
index 000000000000..75484ef0c743
--- /dev/null
+++ b/include/asm-sparc64/libata-portmap.h
@@ -0,0 +1 @@
+#include <asm-generic/libata-portmap.h>
diff --git a/include/asm-x86_64/libata-portmap.h b/include/asm-x86_64/libata-portmap.h
new file mode 100644
index 000000000000..75484ef0c743
--- /dev/null
+++ b/include/asm-x86_64/libata-portmap.h
@@ -0,0 +1 @@
+#include <asm-generic/libata-portmap.h>
diff --git a/include/linux/libata.h b/include/linux/libata.h
index cf5eb1da3e32..4504776570e4 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -36,6 +36,8 @@
 #include <linux/workqueue.h>
 #include <scsi/scsi_host.h>
 
+#include <asm/libata-portmap.h>
+
 /*
  * compile-time options: to be removed as soon as all the drivers are
  * converted to the new debugging mechanism
@@ -356,6 +358,7 @@ struct ata_probe_ent {
 	unsigned int		udma_mask;
 	unsigned int		legacy_mode;
 	unsigned long		irq;
+	unsigned long		irq2;
 	unsigned int		irq_flags;
 	unsigned long		host_flags;
 	unsigned long		host_set_flags;
@@ -367,6 +370,7 @@ struct ata_host_set {
 	spinlock_t		lock;
 	struct device 		*dev;
 	unsigned long		irq;
+	unsigned long		irq2;
 	void __iomem		*mmio_base;
 	unsigned int		n_ports;
 	void			*private_data;
@@ -374,7 +378,6 @@ struct ata_host_set {
 	unsigned long		flags;
 	int			simplex_claimed;	/* Keep seperate in case we
 							   ever need to do this locked */
-	struct ata_host_set	*next;		/* for legacy mode */
 	struct ata_port		*ports[0];
 };
 
-- 
cgit v1.2.3


From dd5b06c490de72440ec39f814de99a714a45a1a9 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Thu, 10 Aug 2006 16:59:12 +0900
Subject: [PATCH] libata: implement dummy port

Implement dummy port which can be requested by setting appropriate bit
in probe_ent->dummy_port_mask.  The dummy port is used as placeholder
for stolen legacy port.  This allows libata to guarantee that
index_of(ap) == ap->port_no == actual_device_port_no, and thus to
remove error-prone ap->hard_port_no.

As it's used only when one port of a legacy controller is reserved by
some other entity (e.g. IDE), the focus is on keeping the added *code*
complexity at minimum, so dummy port allocates all libata core
resources and acts as a normal port.  It just has all dummy port_ops.

This patch only implements dummy port.  The following patch will make
libata use it for stolen legacy ports.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/scsi/libata-core.c | 61 ++++++++++++++++++++++++++++++++++++++--------
 include/linux/libata.h     |  8 ++++++
 2 files changed, 59 insertions(+), 10 deletions(-)

(limited to 'include')

diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c
index 3634279d896d..f2e7e2f13db9 100644
--- a/drivers/scsi/libata-core.c
+++ b/drivers/scsi/libata-core.c
@@ -5311,7 +5311,6 @@ static struct ata_port * ata_port_add(const struct ata_probe_ent *ent,
 {
 	struct Scsi_Host *shost;
 	struct ata_port *ap;
-	int rc;
 
 	DPRINTK("ENTER\n");
 
@@ -5333,15 +5332,7 @@ static struct ata_port * ata_port_add(const struct ata_probe_ent *ent,
 	ata_port_init(ap, host_set, ent, port_no);
 	ata_port_init_shost(ap, shost);
 
-	rc = ap->ops->port_start(ap);
-	if (rc)
-		goto err_out;
-
 	return ap;
-
-err_out:
-	scsi_host_put(shost);
-	return NULL;
 }
 
 /**
@@ -5415,11 +5406,27 @@ int ata_device_add(const struct ata_probe_ent *ent)
 		if (!ap)
 			goto err_out;
 
+		host_set->ports[i] = ap;
+
+		/* dummy? */
+		if (ent->dummy_port_mask & (1 << i)) {
+			ata_port_printk(ap, KERN_INFO, "DUMMY\n");
+			ap->ops = &ata_dummy_port_ops;
+			continue;
+		}
+
+		/* start port */
+		rc = ap->ops->port_start(ap);
+		if (rc) {
+			host_set->ports[i] = NULL;
+			scsi_host_put(ap->host);
+			goto err_out;
+		}
+
 		/* Report the secondary IRQ for second channel legacy */
 		if (i == 1 && ent->irq2)
 			irq_line = ent->irq2;
 
-		host_set->ports[i] = ap;
 		xfer_mode_mask =(ap->udma_mask << ATA_SHIFT_UDMA) |
 				(ap->mwdma_mask << ATA_SHIFT_MWDMA) |
 				(ap->pio_mask << ATA_SHIFT_PIO);
@@ -5940,6 +5947,39 @@ u32 ata_wait_register(void __iomem *reg, u32 mask, u32 val,
 	return tmp;
 }
 
+/*
+ * Dummy port_ops
+ */
+static void ata_dummy_noret(struct ata_port *ap)	{ }
+static int ata_dummy_ret0(struct ata_port *ap)		{ return 0; }
+static void ata_dummy_qc_noret(struct ata_queued_cmd *qc) { }
+
+static u8 ata_dummy_check_status(struct ata_port *ap)
+{
+	return ATA_DRDY;
+}
+
+static unsigned int ata_dummy_qc_issue(struct ata_queued_cmd *qc)
+{
+	return AC_ERR_SYSTEM;
+}
+
+const struct ata_port_operations ata_dummy_port_ops = {
+	.port_disable		= ata_port_disable,
+	.check_status		= ata_dummy_check_status,
+	.check_altstatus	= ata_dummy_check_status,
+	.dev_select		= ata_noop_dev_select,
+	.qc_prep		= ata_noop_qc_prep,
+	.qc_issue		= ata_dummy_qc_issue,
+	.freeze			= ata_dummy_noret,
+	.thaw			= ata_dummy_noret,
+	.error_handler		= ata_dummy_noret,
+	.post_internal_cmd	= ata_dummy_qc_noret,
+	.irq_clear		= ata_dummy_noret,
+	.port_start		= ata_dummy_ret0,
+	.port_stop		= ata_dummy_noret,
+};
+
 /*
  * libata is essentially a library of internal helper functions for
  * low-level ATA host controller drivers.  As such, the API/ABI is
@@ -5950,6 +5990,7 @@ u32 ata_wait_register(void __iomem *reg, u32 mask, u32 val,
 EXPORT_SYMBOL_GPL(sata_deb_timing_normal);
 EXPORT_SYMBOL_GPL(sata_deb_timing_hotplug);
 EXPORT_SYMBOL_GPL(sata_deb_timing_long);
+EXPORT_SYMBOL_GPL(ata_dummy_port_ops);
 EXPORT_SYMBOL_GPL(ata_std_bios_param);
 EXPORT_SYMBOL_GPL(ata_std_ports);
 EXPORT_SYMBOL_GPL(ata_host_set_init);
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 4504776570e4..30bfe8f1666e 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -353,6 +353,7 @@ struct ata_probe_ent {
 	struct ata_ioports	port[ATA_MAX_PORTS];
 	unsigned int		n_ports;
 	unsigned int		hard_port_no;
+	unsigned int		dummy_port_mask;
 	unsigned int		pio_mask;
 	unsigned int		mwdma_mask;
 	unsigned int		udma_mask;
@@ -652,6 +653,8 @@ extern const unsigned long sata_deb_timing_normal[];
 extern const unsigned long sata_deb_timing_hotplug[];
 extern const unsigned long sata_deb_timing_long[];
 
+extern const struct ata_port_operations ata_dummy_port_ops;
+
 static inline const unsigned long *
 sata_ehc_deb_timing(struct ata_eh_context *ehc)
 {
@@ -661,6 +664,11 @@ sata_ehc_deb_timing(struct ata_eh_context *ehc)
 		return sata_deb_timing_normal;
 }
 
+static inline int ata_port_is_dummy(struct ata_port *ap)
+{
+	return ap->ops == &ata_dummy_port_ops;
+}
+
 extern void ata_port_probe(struct ata_port *);
 extern void __sata_phy_reset(struct ata_port *ap);
 extern void sata_phy_reset(struct ata_port *ap);
-- 
cgit v1.2.3


From 4852ba24f647199be797545226c6d325db231937 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Thu, 10 Aug 2006 16:59:18 +0900
Subject: [PATCH] libata: kill unused hard_port_no and legacy_mode

Kill unused probe_ent/ap->hard_port_no and probe_ent->legacy_mode.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/scsi/libata-core.c | 1 -
 include/linux/libata.h     | 3 ---
 2 files changed, 4 deletions(-)

(limited to 'include')

diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c
index 2ef86b4e7f6d..3f963f206d4a 100644
--- a/drivers/scsi/libata-core.c
+++ b/drivers/scsi/libata-core.c
@@ -5223,7 +5223,6 @@ void ata_port_init(struct ata_port *ap, struct ata_host_set *host_set,
 	ap->host_set = host_set;
 	ap->dev = ent->dev;
 	ap->port_no = port_no;
-	ap->hard_port_no = port_no;
 	ap->pio_mask = ent->pio_mask;
 	ap->mwdma_mask = ent->mwdma_mask;
 	ap->udma_mask = ent->udma_mask;
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 30bfe8f1666e..ed749f778697 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -352,12 +352,10 @@ struct ata_probe_ent {
 	struct scsi_host_template *sht;
 	struct ata_ioports	port[ATA_MAX_PORTS];
 	unsigned int		n_ports;
-	unsigned int		hard_port_no;
 	unsigned int		dummy_port_mask;
 	unsigned int		pio_mask;
 	unsigned int		mwdma_mask;
 	unsigned int		udma_mask;
-	unsigned int		legacy_mode;
 	unsigned long		irq;
 	unsigned long		irq2;
 	unsigned int		irq_flags;
@@ -509,7 +507,6 @@ struct ata_port {
 	unsigned int		pflags; /* ATA_PFLAG_xxx */
 	unsigned int		id;	/* unique id req'd by scsi midlyr */
 	unsigned int		port_no; /* unique port #; from zero */
-	unsigned int		hard_port_no;	/* hardware port #; from zero */
 
 	struct ata_prd		*prd;	 /* our SG list */
 	dma_addr_t		prd_dma; /* and its DMA mapping */
-- 
cgit v1.2.3


From 8b881b0410de0f72a43e814393abf3a4cb29ebb4 Mon Sep 17 00:00:00 2001
From: Jeff Garzik <jeff@garzik.org>
Date: Sun, 11 Jun 2006 09:59:27 -0400
Subject: [ATA] Increase lba48 max-sectors from 200 to 256.

Also, moved ATA_MAX_SECTORS and ATA_MAX_SECTORS_LBA48 from
linux/libata.h to linux/ata.h, now that they truly reflect the standard
(well... mostly; note TODO comment).

This changes the performance profile (and potential bug profile)
for a bunch of drivers, so be wary.
---
 include/linux/ata.h    | 2 ++
 include/linux/libata.h | 2 --
 2 files changed, 2 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/include/linux/ata.h b/include/linux/ata.h
index 3671af869696..8d708a3d505b 100644
--- a/include/linux/ata.h
+++ b/include/linux/ata.h
@@ -40,6 +40,8 @@ enum {
 	ATA_MAX_DEVICES		= 2,	/* per bus/port */
 	ATA_MAX_PRD		= 256,	/* we could make these 256/256 */
 	ATA_SECT_SIZE		= 512,
+	ATA_MAX_SECTORS		= 256,
+	ATA_MAX_SECTORS_LBA48	= 65535,/* TODO: 65536? */
 
 	ATA_ID_WORDS		= 256,
 	ATA_ID_SERNO_OFS	= 10,
diff --git a/include/linux/libata.h b/include/linux/libata.h
index ed749f778697..060da736b3a8 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -114,8 +114,6 @@ enum {
 	/* tag ATA_MAX_QUEUE - 1 is reserved for internal commands */
 	ATA_MAX_QUEUE		= 32,
 	ATA_TAG_INTERNAL	= ATA_MAX_QUEUE - 1,
-	ATA_MAX_SECTORS		= 200,	/* FIXME */
-	ATA_MAX_SECTORS_LBA48	= 65535,
 	ATA_MAX_BUS		= 2,
 	ATA_DEF_BUSY_WAIT	= 10000,
 	ATA_SHORT_PAUSE		= (HZ >> 6) + 1,
-- 
cgit v1.2.3


From b352e57dc3bb5033996adaa67c2f69b795eddd39 Mon Sep 17 00:00:00 2001
From: Alan Cox <alan@lxorguk.ukuu.org.uk>
Date: Thu, 10 Aug 2006 18:52:12 +0100
Subject: [PATCH] libata: Add CompactFlash support

The CFA world has some additional rules and drive modes we need to support for
newer expansion cards and on embedded boxes

Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/ata/libata-core.c | 62 +++++++++++++++++++++++++++++++++++++++++------
 include/linux/ata.h       | 20 ++++++++++++++-
 include/linux/libata.h    |  4 +--
 3 files changed, 75 insertions(+), 11 deletions(-)

(limited to 'include')

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 0ac0b519cf2d..9092416a6301 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -386,9 +386,13 @@ static const char *ata_mode_string(unsigned int xfer_mask)
 		"PIO2",
 		"PIO3",
 		"PIO4",
+		"PIO5",
+		"PIO6",
 		"MWDMA0",
 		"MWDMA1",
 		"MWDMA2",
+		"MWDMA3",
+		"MWDMA4",
 		"UDMA/16",
 		"UDMA/25",
 		"UDMA/33",
@@ -875,6 +879,23 @@ static unsigned int ata_id_xfermask(const u16 *id)
 
 	mwdma_mask = id[ATA_ID_MWDMA_MODES] & 0x07;
 
+	if (ata_id_is_cfa(id)) {
+		/*
+		 *	Process compact flash extended modes
+		 */
+		int pio = id[163] & 0x7;
+		int dma = (id[163] >> 3) & 7;
+
+		if (pio)
+			pio_mask |= (1 << 5);
+		if (pio > 1)
+			pio_mask |= (1 << 6);
+		if (dma)
+			mwdma_mask |= (1 << 3);
+		if (dma > 1)
+			mwdma_mask |= (1 << 4);
+	}
+
 	udma_mask = 0;
 	if (id[ATA_ID_FIELD_VALID] & (1 << 2))
 		udma_mask = id[ATA_ID_UDMA_MODES] & 0xff;
@@ -1356,6 +1377,7 @@ int ata_dev_configure(struct ata_device *dev, int print_info)
 	struct ata_port *ap = dev->ap;
 	const u16 *id = dev->id;
 	unsigned int xfer_mask;
+	char revbuf[7];		/* XYZ-99\0 */
 	int rc;
 
 	if (!ata_dev_enabled(dev) && ata_msg_info(ap)) {
@@ -1399,6 +1421,15 @@ int ata_dev_configure(struct ata_device *dev, int print_info)
 
 	/* ATA-specific feature tests */
 	if (dev->class == ATA_DEV_ATA) {
+		if (ata_id_is_cfa(id)) {
+			if (id[162] & 1) /* CPRM may make this media unusable */
+				ata_dev_printk(dev, KERN_WARNING, "ata%u: device %u  supports DRM functions and may not be fully accessable.\n",
+					ap->id, dev->devno);
+			snprintf(revbuf, 7, "CFA");
+		}
+		else
+			snprintf(revbuf, 7, "ATA-%d",  ata_id_major_version(id));
+
 		dev->n_sectors = ata_id_n_sectors(id);
 
 		if (ata_id_has_lba(id)) {
@@ -1417,9 +1448,9 @@ int ata_dev_configure(struct ata_device *dev, int print_info)
 
 			/* print device info to dmesg */
 			if (ata_msg_drv(ap) && print_info)
-				ata_dev_printk(dev, KERN_INFO, "ATA-%d, "
+				ata_dev_printk(dev, KERN_INFO, "%s, "
 					"max %s, %Lu sectors: %s %s\n",
-					ata_id_major_version(id),
+					revbuf,
 					ata_mode_string(xfer_mask),
 					(unsigned long long)dev->n_sectors,
 					lba_desc, ncq_desc);
@@ -1440,9 +1471,9 @@ int ata_dev_configure(struct ata_device *dev, int print_info)
 
 			/* print device info to dmesg */
 			if (ata_msg_drv(ap) && print_info)
-				ata_dev_printk(dev, KERN_INFO, "ATA-%d, "
+				ata_dev_printk(dev, KERN_INFO, "%s, "
 					"max %s, %Lu sectors: CHS %u/%u/%u\n",
-					ata_id_major_version(id),
+					revbuf,
 					ata_mode_string(xfer_mask),
 					(unsigned long long)dev->n_sectors,
 					dev->cylinders, dev->heads,
@@ -1900,10 +1931,11 @@ int sata_set_spd(struct ata_port *ap)
  * drivers/ide/ide-timing.h and was originally written by Vojtech Pavlik
  */
 /*
- * PIO 0-5, MWDMA 0-2 and UDMA 0-6 timings (in nanoseconds).
+ * PIO 0-4, MWDMA 0-2 and UDMA 0-6 timings (in nanoseconds).
  * These were taken from ATA/ATAPI-6 standard, rev 0a, except
- * for PIO 5, which is a nonstandard extension and UDMA6, which
- * is currently supported only by Maxtor drives.
+ * for UDMA6, which is currently supported only by Maxtor drives.
+ *
+ * For PIO 5/6 MWDMA 3/4 see the CFA specification 3.0.
  */
 
 static const struct ata_timing ata_timing[] = {
@@ -1913,6 +1945,8 @@ static const struct ata_timing ata_timing[] = {
 	{ XFER_UDMA_4,     0,   0,   0,   0,   0,   0,   0,  30 },
 	{ XFER_UDMA_3,     0,   0,   0,   0,   0,   0,   0,  45 },
 
+	{ XFER_MW_DMA_4,  25,   0,   0,   0,  55,  20,  80,   0 },
+	{ XFER_MW_DMA_3,  25,   0,   0,   0,  65,  25, 100,   0 },
 	{ XFER_UDMA_2,     0,   0,   0,   0,   0,   0,   0,  60 },
 	{ XFER_UDMA_1,     0,   0,   0,   0,   0,   0,   0,  80 },
 	{ XFER_UDMA_0,     0,   0,   0,   0,   0,   0,   0, 120 },
@@ -1927,7 +1961,8 @@ static const struct ata_timing ata_timing[] = {
 	{ XFER_SW_DMA_1,  90,   0,   0,   0, 240, 240, 480,   0 },
 	{ XFER_SW_DMA_0, 120,   0,   0,   0, 480, 480, 960,   0 },
 
-/*	{ XFER_PIO_5,     20,  50,  30, 100,  50,  30, 100,   0 }, */
+	{ XFER_PIO_6,     10,  55,  20,  80,  55,  20,  80,   0 },
+	{ XFER_PIO_5,     15,  65,  25, 100,  65,  25, 100,   0 },
 	{ XFER_PIO_4,     25,  70,  25, 120,  70,  25, 120,   0 },
 	{ XFER_PIO_3,     30,  80,  70, 180,  80,  70, 180,   0 },
 
@@ -3062,6 +3097,17 @@ static void ata_dev_xfermask(struct ata_device *dev)
 				       dev->mwdma_mask, dev->udma_mask);
 	xfer_mask &= ata_id_xfermask(dev->id);
 
+	/*
+	 *	CFA Advanced TrueIDE timings are not allowed on a shared
+	 *	cable
+	 */
+	if (ata_dev_pair(dev)) {
+		/* No PIO5 or PIO6 */
+		xfer_mask &= ~(0x03 << (ATA_SHIFT_PIO + 5));
+		/* No MWDMA3 or MWDMA 4 */
+		xfer_mask &= ~(0x03 << (ATA_SHIFT_MWDMA + 3));
+	}
+
 	if (ata_dma_blacklisted(dev)) {
 		xfer_mask &= ~(ATA_MASK_MWDMA | ATA_MASK_UDMA);
 		ata_dev_printk(dev, KERN_WARNING,
diff --git a/include/linux/ata.h b/include/linux/ata.h
index 8d708a3d505b..991b858acc30 100644
--- a/include/linux/ata.h
+++ b/include/linux/ata.h
@@ -170,12 +170,16 @@ enum {
 	XFER_UDMA_2		= 0x42,
 	XFER_UDMA_1		= 0x41,
 	XFER_UDMA_0		= 0x40,
+	XFER_MW_DMA_4		= 0x24,	/* CFA only */
+	XFER_MW_DMA_3		= 0x23,	/* CFA only */
 	XFER_MW_DMA_2		= 0x22,
 	XFER_MW_DMA_1		= 0x21,
 	XFER_MW_DMA_0		= 0x20,
 	XFER_SW_DMA_2		= 0x12,
 	XFER_SW_DMA_1		= 0x11,
 	XFER_SW_DMA_0		= 0x10,
+	XFER_PIO_6		= 0x0E,	/* CFA only */
+	XFER_PIO_5		= 0x0D,	/* CFA only */
 	XFER_PIO_4		= 0x0C,
 	XFER_PIO_3		= 0x0B,
 	XFER_PIO_2		= 0x0A,
@@ -274,7 +278,6 @@ struct ata_taskfile {
 };
 
 #define ata_id_is_ata(id)	(((id)[0] & (1 << 15)) == 0)
-#define ata_id_is_cfa(id)	((id)[0] == 0x848A)
 #define ata_id_is_sata(id)	((id)[93] == 0)
 #define ata_id_rahead_enabled(id) ((id)[85] & (1 << 6))
 #define ata_id_wcache_enabled(id) ((id)[85] & (1 << 5))
@@ -306,6 +309,9 @@ static inline unsigned int ata_id_major_version(const u16 *id)
 {
 	unsigned int mver;
 
+	if (id[ATA_ID_MAJOR_VER] == 0xFFFF)
+		return 0;
+
 	for (mver = 14; mver >= 1; mver--)
 		if (id[ATA_ID_MAJOR_VER] & (1 << mver))
 			break;
@@ -324,6 +330,18 @@ static inline int ata_id_current_chs_valid(const u16 *id)
 		id[56];    /* sectors in current translation */
 }
 
+static inline int ata_id_is_cfa(const u16 *id)
+{
+	u16 v = id[0];
+	if (v == 0x848A)	/* Standard CF */
+		return 1;
+	/* Could be CF hiding as standard ATA */
+	if (ata_id_major_version(id) >= 3 &&  id[82] != 0xFFFF &&
+			(id[82] & ( 1 << 2)))
+		return 1;
+	return 0;
+}
+
 static inline int atapi_cdb_len(const u16 *dev_id)
 {
 	u16 tmp = dev_id[0] & 0x3;
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 060da736b3a8..806682603ac5 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -225,8 +225,8 @@ enum {
 	/* encoding various smaller bitmaps into a single
 	 * unsigned int bitmap
 	 */
-	ATA_BITS_PIO		= 5,
-	ATA_BITS_MWDMA		= 3,
+	ATA_BITS_PIO		= 7,
+	ATA_BITS_MWDMA		= 5,
 	ATA_BITS_UDMA		= 8,
 
 	ATA_SHIFT_PIO		= 0,
-- 
cgit v1.2.3


From 016131b8fffa1085b4ad165ab228116fdc278ebe Mon Sep 17 00:00:00 2001
From: James Smart <James.Smart@Emulex.Com>
Date: Mon, 14 Aug 2006 08:20:25 -0400
Subject: [SCSI] fc transport: convert fc_host symbolic_name attribute to a
 dynamic attribute

Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 drivers/scsi/scsi_transport_fc.c | 4 ++--
 include/scsi/scsi_transport_fc.h | 3 ++-
 2 files changed, 4 insertions(+), 3 deletions(-)

(limited to 'include')

diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c
index b03aa85108e5..c1c5cdffca38 100644
--- a/drivers/scsi/scsi_transport_fc.c
+++ b/drivers/scsi/scsi_transport_fc.c
@@ -815,7 +815,6 @@ fc_private_host_rd_attr_cast(node_name, "0x%llx\n", 20, unsigned long long);
 fc_private_host_rd_attr_cast(port_name, "0x%llx\n", 20, unsigned long long);
 fc_private_host_rd_attr_cast(permanent_port_name, "0x%llx\n", 20,
 			     unsigned long long);
-fc_private_host_rd_attr(symbolic_name, "%s\n", (FC_SYMBOLIC_NAME_SIZE +1));
 fc_private_host_rd_attr(maxframe_size, "%u bytes\n", 20);
 fc_private_host_rd_attr(serial_number, "%s\n", (FC_SERIAL_NUMBER_SIZE +1));
 
@@ -858,6 +857,7 @@ fc_host_rd_attr(port_id, "0x%06x\n", 20);
 fc_host_rd_enum_attr(port_type, FC_PORTTYPE_MAX_NAMELEN);
 fc_host_rd_enum_attr(port_state, FC_PORTSTATE_MAX_NAMELEN);
 fc_host_rd_attr_cast(fabric_name, "0x%llx\n", 20, unsigned long long);
+fc_host_rd_attr(symbolic_name, "%s\n", FC_SYMBOLIC_NAME_SIZE + 1);
 
 
 /* Private Host Attributes */
@@ -1223,7 +1223,6 @@ fc_attach_transport(struct fc_function_template *ft)
 	SETUP_HOST_ATTRIBUTE_RD(permanent_port_name);
 	SETUP_HOST_ATTRIBUTE_RD(supported_classes);
 	SETUP_HOST_ATTRIBUTE_RD(supported_fc4s);
-	SETUP_HOST_ATTRIBUTE_RD(symbolic_name);
 	SETUP_HOST_ATTRIBUTE_RD(supported_speeds);
 	SETUP_HOST_ATTRIBUTE_RD(maxframe_size);
 	SETUP_HOST_ATTRIBUTE_RD(serial_number);
@@ -1234,6 +1233,7 @@ fc_attach_transport(struct fc_function_template *ft)
 	SETUP_HOST_ATTRIBUTE_RD(active_fc4s);
 	SETUP_HOST_ATTRIBUTE_RD(speed);
 	SETUP_HOST_ATTRIBUTE_RD(fabric_name);
+	SETUP_HOST_ATTRIBUTE_RD(symbolic_name);
 
 	/* Transport-managed attributes */
 	SETUP_PRIVATE_HOST_ATTRIBUTE_RW(tgtid_bind_type);
diff --git a/include/scsi/scsi_transport_fc.h b/include/scsi/scsi_transport_fc.h
index 6d28b0317657..b7f62b85f0b3 100644
--- a/include/scsi/scsi_transport_fc.h
+++ b/include/scsi/scsi_transport_fc.h
@@ -409,6 +409,7 @@ struct fc_function_template {
 	void	(*get_host_active_fc4s)(struct Scsi_Host *);
 	void	(*get_host_speed)(struct Scsi_Host *);
 	void	(*get_host_fabric_name)(struct Scsi_Host *);
+	void	(*get_host_symbolic_name)(struct Scsi_Host *);
 
 	struct fc_host_statistics * (*get_fc_host_stats)(struct Scsi_Host *);
 	void	(*reset_fc_host_stats)(struct Scsi_Host *);
@@ -445,7 +446,6 @@ struct fc_function_template {
 	unsigned long	show_host_permanent_port_name:1;
 	unsigned long	show_host_supported_classes:1;
 	unsigned long	show_host_supported_fc4s:1;
-	unsigned long	show_host_symbolic_name:1;
 	unsigned long	show_host_supported_speeds:1;
 	unsigned long	show_host_maxframe_size:1;
 	unsigned long	show_host_serial_number:1;
@@ -456,6 +456,7 @@ struct fc_function_template {
 	unsigned long	show_host_active_fc4s:1;
 	unsigned long	show_host_speed:1;
 	unsigned long	show_host_fabric_name:1;
+	unsigned long	show_host_symbolic_name:1;
 };
 
 
-- 
cgit v1.2.3


From b8d08210126a7b769b857720a59721a453a57a1e Mon Sep 17 00:00:00 2001
From: James Smart <James.Smart@Emulex.Com>
Date: Thu, 17 Aug 2006 08:00:43 -0400
Subject: [SCSI] fc transport: add fc_host system_hostname attribute and
 u64_to_wwn()

This patch updates the fc transport for the following:

- Addition of a new attribute "system_hostname" which can be
  used to set the fully qualified hostname that the fc_host
  is attached to. The fc_host can then register this string
  as the FDMI-based host name attribute.
  Note: for NPIV, a fc_host could be associated with a system which
    is not the local system.

- Add the inline function u64_to_wwn(), which is the inverse of the
  existing wwn_to_u64() function.

- Slight reorg, just to keep dynamic attributes with each other, etc

Signed-off-by: James Smart <James.Smart@emulex.com>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 drivers/scsi/scsi_transport_fc.c | 30 ++++++++++++++++++++++++++++--
 include/scsi/scsi_transport_fc.h | 38 +++++++++++++++++++++++++++-----------
 2 files changed, 55 insertions(+), 13 deletions(-)

(limited to 'include')

diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c
index c1c5cdffca38..79d31ca2b741 100644
--- a/drivers/scsi/scsi_transport_fc.c
+++ b/drivers/scsi/scsi_transport_fc.c
@@ -301,8 +301,6 @@ static int fc_host_setup(struct transport_container *tc, struct device *dev,
 	fc_host->supported_classes = FC_COS_UNSPECIFIED;
 	memset(fc_host->supported_fc4s, 0,
 		sizeof(fc_host->supported_fc4s));
-	memset(fc_host->symbolic_name, 0,
-		sizeof(fc_host->symbolic_name));
 	fc_host->supported_speeds = FC_PORTSPEED_UNKNOWN;
 	fc_host->maxframe_size = -1;
 	memset(fc_host->serial_number, 0,
@@ -315,6 +313,8 @@ static int fc_host_setup(struct transport_container *tc, struct device *dev,
 		sizeof(fc_host->active_fc4s));
 	fc_host->speed = FC_PORTSPEED_UNKNOWN;
 	fc_host->fabric_name = -1;
+	memset(fc_host->symbolic_name, 0, sizeof(fc_host->symbolic_name));
+	memset(fc_host->system_hostname, 0, sizeof(fc_host->system_hostname));
 
 	fc_host->tgtid_bind_type = FC_TGTID_BIND_BY_WWPN;
 
@@ -688,6 +688,25 @@ store_fc_host_##field(struct class_device *cdev, const char *buf,	\
 	return count;							\
 }
 
+#define fc_host_store_str_function(field, slen)				\
+static ssize_t								\
+store_fc_host_##field(struct class_device *cdev, const char *buf,	\
+			   size_t count)				\
+{									\
+	struct Scsi_Host *shost = transport_class_to_shost(cdev);	\
+	struct fc_internal *i = to_fc_internal(shost->transportt);	\
+	unsigned int cnt=count;						\
+									\
+	/* count may include a LF at end of string */			\
+	if (buf[cnt-1] == '\n')						\
+		cnt--;							\
+	if (cnt > ((slen) - 1))						\
+		return -EINVAL;						\
+	memcpy(fc_host_##field(shost), buf, cnt);			\
+	i->f->set_host_##field(shost);					\
+	return count;							\
+}
+
 #define fc_host_rd_attr(field, format_string, sz)			\
 	fc_host_show_function(field, format_string, sz, )		\
 static FC_CLASS_DEVICE_ATTR(host, field, S_IRUGO,			\
@@ -859,6 +878,12 @@ fc_host_rd_enum_attr(port_state, FC_PORTSTATE_MAX_NAMELEN);
 fc_host_rd_attr_cast(fabric_name, "0x%llx\n", 20, unsigned long long);
 fc_host_rd_attr(symbolic_name, "%s\n", FC_SYMBOLIC_NAME_SIZE + 1);
 
+fc_private_host_show_function(system_hostname, "%s\n",
+		FC_SYMBOLIC_NAME_SIZE + 1, )
+fc_host_store_str_function(system_hostname, FC_SYMBOLIC_NAME_SIZE)
+static FC_CLASS_DEVICE_ATTR(host, system_hostname, S_IRUGO | S_IWUSR,
+		show_fc_host_system_hostname, store_fc_host_system_hostname);
+
 
 /* Private Host Attributes */
 
@@ -1234,6 +1259,7 @@ fc_attach_transport(struct fc_function_template *ft)
 	SETUP_HOST_ATTRIBUTE_RD(speed);
 	SETUP_HOST_ATTRIBUTE_RD(fabric_name);
 	SETUP_HOST_ATTRIBUTE_RD(symbolic_name);
+	SETUP_HOST_ATTRIBUTE_RW(system_hostname);
 
 	/* Transport-managed attributes */
 	SETUP_PRIVATE_HOST_ATTRIBUTE_RW(tgtid_bind_type);
diff --git a/include/scsi/scsi_transport_fc.h b/include/scsi/scsi_transport_fc.h
index b7f62b85f0b3..c74be5dabfeb 100644
--- a/include/scsi/scsi_transport_fc.h
+++ b/include/scsi/scsi_transport_fc.h
@@ -312,7 +312,6 @@ struct fc_host_attrs {
 	u64 permanent_port_name;
 	u32 supported_classes;
 	u8  supported_fc4s[FC_FC4_LIST_SIZE];
-	char symbolic_name[FC_SYMBOLIC_NAME_SIZE];
 	u32 supported_speeds;
 	u32 maxframe_size;
 	char serial_number[FC_SERIAL_NUMBER_SIZE];
@@ -324,6 +323,8 @@ struct fc_host_attrs {
 	u8  active_fc4s[FC_FC4_LIST_SIZE];
 	u32 speed;
 	u64 fabric_name;
+	char symbolic_name[FC_SYMBOLIC_NAME_SIZE];
+	char system_hostname[FC_SYMBOLIC_NAME_SIZE];
 
 	/* Private (Transport-managed) Attributes */
 	enum fc_tgtid_binding_type  tgtid_bind_type;
@@ -354,8 +355,6 @@ struct fc_host_attrs {
 	(((struct fc_host_attrs *)(x)->shost_data)->supported_classes)
 #define fc_host_supported_fc4s(x)	\
 	(((struct fc_host_attrs *)(x)->shost_data)->supported_fc4s)
-#define fc_host_symbolic_name(x)	\
-	(((struct fc_host_attrs *)(x)->shost_data)->symbolic_name)
 #define fc_host_supported_speeds(x)	\
 	(((struct fc_host_attrs *)(x)->shost_data)->supported_speeds)
 #define fc_host_maxframe_size(x)	\
@@ -374,6 +373,10 @@ struct fc_host_attrs {
 	(((struct fc_host_attrs *)(x)->shost_data)->speed)
 #define fc_host_fabric_name(x)	\
 	(((struct fc_host_attrs *)(x)->shost_data)->fabric_name)
+#define fc_host_symbolic_name(x)	\
+	(((struct fc_host_attrs *)(x)->shost_data)->symbolic_name)
+#define fc_host_system_hostname(x)	\
+	(((struct fc_host_attrs *)(x)->shost_data)->system_hostname)
 #define fc_host_tgtid_bind_type(x) \
 	(((struct fc_host_attrs *)(x)->shost_data)->tgtid_bind_type)
 #define fc_host_rports(x) \
@@ -410,6 +413,7 @@ struct fc_function_template {
 	void	(*get_host_speed)(struct Scsi_Host *);
 	void	(*get_host_fabric_name)(struct Scsi_Host *);
 	void	(*get_host_symbolic_name)(struct Scsi_Host *);
+	void	(*set_host_system_hostname)(struct Scsi_Host *);
 
 	struct fc_host_statistics * (*get_fc_host_stats)(struct Scsi_Host *);
 	void	(*reset_fc_host_stats)(struct Scsi_Host *);
@@ -457,6 +461,7 @@ struct fc_function_template {
 	unsigned long	show_host_speed:1;
 	unsigned long	show_host_fabric_name:1;
 	unsigned long	show_host_symbolic_name:1;
+	unsigned long	show_host_system_hostname:1;
 };
 
 
@@ -492,6 +497,25 @@ fc_remote_port_chkready(struct fc_rport *rport)
 	return result;
 }
 
+static inline u64 wwn_to_u64(u8 *wwn)
+{
+	return (u64)wwn[0] << 56 | (u64)wwn[1] << 48 |
+	    (u64)wwn[2] << 40 | (u64)wwn[3] << 32 |
+	    (u64)wwn[4] << 24 | (u64)wwn[5] << 16 |
+	    (u64)wwn[6] <<  8 | (u64)wwn[7];
+}
+
+static inline void u64_to_wwn(u64 inm, u8 *wwn)
+{
+	wwn[0] = (inm >> 56) & 0xff;
+	wwn[1] = (inm >> 48) & 0xff;
+	wwn[2] = (inm >> 40) & 0xff;
+	wwn[3] = (inm >> 32) & 0xff;
+	wwn[4] = (inm >> 24) & 0xff;
+	wwn[5] = (inm >> 16) & 0xff;
+	wwn[6] = (inm >> 8) & 0xff;
+	wwn[7] = inm & 0xff;
+}
 
 struct scsi_transport_template *fc_attach_transport(
 			struct fc_function_template *);
@@ -503,12 +527,4 @@ void fc_remote_port_delete(struct fc_rport  *rport);
 void fc_remote_port_rolechg(struct fc_rport  *rport, u32 roles);
 int scsi_is_fc_rport(const struct device *);
 
-static inline u64 wwn_to_u64(u8 *wwn)
-{
-	return (u64)wwn[0] << 56 | (u64)wwn[1] << 48 |
-	    (u64)wwn[2] << 40 | (u64)wwn[3] << 32 |
-	    (u64)wwn[4] << 24 | (u64)wwn[5] << 16 |
-	    (u64)wwn[6] <<  8 | (u64)wwn[7];
-}
-
 #endif /* SCSI_TRANSPORT_FC_H */
-- 
cgit v1.2.3


From cca3974e48607c3775dc73b544a5700b2e37c21a Mon Sep 17 00:00:00 2001
From: Jeff Garzik <jeff@garzik.org>
Date: Thu, 24 Aug 2006 03:19:22 -0400
Subject: libata: Grand renaming.

The biggest change is that ata_host_set is renamed to ata_host.

* ata_host_set			=> ata_host
* ata_probe_ent->host_flags	=> ata_probe_ent->port_flags
* ata_probe_ent->host_set_flags	=> ata_probe_ent->_host_flags
* ata_host_stats		=> ata_port_stats
* ata_port->host		=> ata_port->scsi_host
* ata_port->host_set		=> ata_port->host
* ata_port_info->host_flags	=> ata_port_info->flags
* ata_(.*)host_set(.*)\(\)	=> ata_\1host\2()

The leading underscore in ata_probe_ent->_host_flags is to avoid
reusing ->host_flags for different purpose.  Currently, the only user
of the field is libata-bmdma.c and probe_ent itself is scheduled to be
removed.

ata_port->host is reused for different purpose but this field is used
inside libata core proper and of different type.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/ata/ahci.c         |  98 +++++++--------
 drivers/ata/ata_piix.c     |  56 ++++-----
 drivers/ata/libata-bmdma.c |  34 +++---
 drivers/ata/libata-core.c  | 292 ++++++++++++++++++++++-----------------------
 drivers/ata/libata-eh.c    |  16 +--
 drivers/ata/libata-scsi.c  |  78 ++++++------
 drivers/ata/libata.h       |   2 +-
 drivers/ata/pdc_adma.c     |  46 +++----
 drivers/ata/sata_mv.c      |  98 ++++++++-------
 drivers/ata/sata_nv.c      |  58 ++++-----
 drivers/ata/sata_promise.c |  54 ++++-----
 drivers/ata/sata_qstor.c   |  44 +++----
 drivers/ata/sata_sil.c     |  47 ++++----
 drivers/ata/sata_sil24.c   |  59 +++++----
 drivers/ata/sata_sis.c     |  18 +--
 drivers/ata/sata_svw.c     |  10 +-
 drivers/ata/sata_sx4.c     |  64 +++++-----
 drivers/ata/sata_uli.c     |   8 +-
 drivers/ata/sata_via.c     |   4 +-
 drivers/ata/sata_vsc.c     |  16 +--
 include/linux/libata.h     |  39 +++---
 21 files changed, 568 insertions(+), 573 deletions(-)

(limited to 'include')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 813031c01fba..3f1106fdaed1 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -277,7 +277,7 @@ static const struct ata_port_info ahci_port_info[] = {
 	/* board_ahci */
 	{
 		.sht		= &ahci_sht,
-		.host_flags	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
+		.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
 				  ATA_FLAG_MMIO | ATA_FLAG_PIO_DMA |
 				  ATA_FLAG_SKIP_D2H_BSY,
 		.pio_mask	= 0x1f, /* pio0-4 */
@@ -287,7 +287,7 @@ static const struct ata_port_info ahci_port_info[] = {
 	/* board_ahci_vt8251 */
 	{
 		.sht		= &ahci_sht,
-		.host_flags	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
+		.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
 				  ATA_FLAG_MMIO | ATA_FLAG_PIO_DMA |
 				  ATA_FLAG_SKIP_D2H_BSY |
 				  AHCI_FLAG_RESET_NEEDS_CLO | AHCI_FLAG_NO_NCQ,
@@ -709,7 +709,7 @@ static void ahci_fill_cmd_slot(struct ahci_port_priv *pp, unsigned int tag,
 static int ahci_clo(struct ata_port *ap)
 {
 	void __iomem *port_mmio = (void __iomem *) ap->ioaddr.cmd_addr;
-	struct ahci_host_priv *hpriv = ap->host_set->private_data;
+	struct ahci_host_priv *hpriv = ap->host->private_data;
 	u32 tmp;
 
 	if (!(hpriv->cap & HOST_CAP_CLO))
@@ -741,7 +741,7 @@ static int ahci_prereset(struct ata_port *ap)
 static int ahci_softreset(struct ata_port *ap, unsigned int *class)
 {
 	struct ahci_port_priv *pp = ap->private_data;
-	void __iomem *mmio = ap->host_set->mmio_base;
+	void __iomem *mmio = ap->host->mmio_base;
 	void __iomem *port_mmio = ahci_port_base(mmio, ap->port_no);
 	const u32 cmd_fis_len = 5; /* five dwords */
 	const char *reason = NULL;
@@ -850,7 +850,7 @@ static int ahci_hardreset(struct ata_port *ap, unsigned int *class)
 	struct ahci_port_priv *pp = ap->private_data;
 	u8 *d2h_fis = pp->rx_fis + RX_FIS_D2H_REG;
 	struct ata_taskfile tf;
-	void __iomem *mmio = ap->host_set->mmio_base;
+	void __iomem *mmio = ap->host->mmio_base;
 	void __iomem *port_mmio = ahci_port_base(mmio, ap->port_no);
 	int rc;
 
@@ -1039,7 +1039,7 @@ static void ahci_error_intr(struct ata_port *ap, u32 irq_stat)
 
 static void ahci_host_intr(struct ata_port *ap)
 {
-	void __iomem *mmio = ap->host_set->mmio_base;
+	void __iomem *mmio = ap->host->mmio_base;
 	void __iomem *port_mmio = ahci_port_base(mmio, ap->port_no);
 	struct ata_eh_info *ehi = &ap->eh_info;
 	u32 status, qc_active;
@@ -1091,7 +1091,7 @@ static void ahci_irq_clear(struct ata_port *ap)
 
 static irqreturn_t ahci_interrupt(int irq, void *dev_instance, struct pt_regs *regs)
 {
-	struct ata_host_set *host_set = dev_instance;
+	struct ata_host *host = dev_instance;
 	struct ahci_host_priv *hpriv;
 	unsigned int i, handled = 0;
 	void __iomem *mmio;
@@ -1099,8 +1099,8 @@ static irqreturn_t ahci_interrupt(int irq, void *dev_instance, struct pt_regs *r
 
 	VPRINTK("ENTER\n");
 
-	hpriv = host_set->private_data;
-	mmio = host_set->mmio_base;
+	hpriv = host->private_data;
+	mmio = host->mmio_base;
 
 	/* sigh.  0xffffffff is a valid return from h/w */
 	irq_stat = readl(mmio + HOST_IRQ_STAT);
@@ -1108,22 +1108,22 @@ static irqreturn_t ahci_interrupt(int irq, void *dev_instance, struct pt_regs *r
 	if (!irq_stat)
 		return IRQ_NONE;
 
-        spin_lock(&host_set->lock);
+        spin_lock(&host->lock);
 
-        for (i = 0; i < host_set->n_ports; i++) {
+        for (i = 0; i < host->n_ports; i++) {
 		struct ata_port *ap;
 
 		if (!(irq_stat & (1 << i)))
 			continue;
 
-		ap = host_set->ports[i];
+		ap = host->ports[i];
 		if (ap) {
 			ahci_host_intr(ap);
 			VPRINTK("port %u\n", i);
 		} else {
 			VPRINTK("port %u (no irq)\n", i);
 			if (ata_ratelimit())
-				dev_printk(KERN_WARNING, host_set->dev,
+				dev_printk(KERN_WARNING, host->dev,
 					"interrupt on disabled port %u\n", i);
 		}
 
@@ -1135,7 +1135,7 @@ static irqreturn_t ahci_interrupt(int irq, void *dev_instance, struct pt_regs *r
 		handled = 1;
 	}
 
-	spin_unlock(&host_set->lock);
+	spin_unlock(&host->lock);
 
 	VPRINTK("EXIT\n");
 
@@ -1157,7 +1157,7 @@ static unsigned int ahci_qc_issue(struct ata_queued_cmd *qc)
 
 static void ahci_freeze(struct ata_port *ap)
 {
-	void __iomem *mmio = ap->host_set->mmio_base;
+	void __iomem *mmio = ap->host->mmio_base;
 	void __iomem *port_mmio = ahci_port_base(mmio, ap->port_no);
 
 	/* turn IRQ off */
@@ -1166,7 +1166,7 @@ static void ahci_freeze(struct ata_port *ap)
 
 static void ahci_thaw(struct ata_port *ap)
 {
-	void __iomem *mmio = ap->host_set->mmio_base;
+	void __iomem *mmio = ap->host->mmio_base;
 	void __iomem *port_mmio = ahci_port_base(mmio, ap->port_no);
 	u32 tmp;
 
@@ -1181,7 +1181,7 @@ static void ahci_thaw(struct ata_port *ap)
 
 static void ahci_error_handler(struct ata_port *ap)
 {
-	void __iomem *mmio = ap->host_set->mmio_base;
+	void __iomem *mmio = ap->host->mmio_base;
 	void __iomem *port_mmio = ahci_port_base(mmio, ap->port_no);
 
 	if (!(ap->pflags & ATA_PFLAG_FROZEN)) {
@@ -1198,7 +1198,7 @@ static void ahci_error_handler(struct ata_port *ap)
 static void ahci_post_internal_cmd(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
-	void __iomem *mmio = ap->host_set->mmio_base;
+	void __iomem *mmio = ap->host->mmio_base;
 	void __iomem *port_mmio = ahci_port_base(mmio, ap->port_no);
 
 	if (qc->flags & ATA_QCFLAG_FAILED)
@@ -1213,9 +1213,9 @@ static void ahci_post_internal_cmd(struct ata_queued_cmd *qc)
 
 static int ahci_port_suspend(struct ata_port *ap, pm_message_t mesg)
 {
-	struct ahci_host_priv *hpriv = ap->host_set->private_data;
+	struct ahci_host_priv *hpriv = ap->host->private_data;
 	struct ahci_port_priv *pp = ap->private_data;
-	void __iomem *mmio = ap->host_set->mmio_base;
+	void __iomem *mmio = ap->host->mmio_base;
 	void __iomem *port_mmio = ahci_port_base(mmio, ap->port_no);
 	const char *emsg = NULL;
 	int rc;
@@ -1233,8 +1233,8 @@ static int ahci_port_suspend(struct ata_port *ap, pm_message_t mesg)
 static int ahci_port_resume(struct ata_port *ap)
 {
 	struct ahci_port_priv *pp = ap->private_data;
-	struct ahci_host_priv *hpriv = ap->host_set->private_data;
-	void __iomem *mmio = ap->host_set->mmio_base;
+	struct ahci_host_priv *hpriv = ap->host->private_data;
+	void __iomem *mmio = ap->host->mmio_base;
 	void __iomem *port_mmio = ahci_port_base(mmio, ap->port_no);
 
 	ahci_init_port(port_mmio, hpriv->cap, pp->cmd_slot_dma, pp->rx_fis_dma);
@@ -1244,8 +1244,8 @@ static int ahci_port_resume(struct ata_port *ap)
 
 static int ahci_pci_device_suspend(struct pci_dev *pdev, pm_message_t mesg)
 {
-	struct ata_host_set *host_set = dev_get_drvdata(&pdev->dev);
-	void __iomem *mmio = host_set->mmio_base;
+	struct ata_host *host = dev_get_drvdata(&pdev->dev);
+	void __iomem *mmio = host->mmio_base;
 	u32 ctl;
 
 	if (mesg.event == PM_EVENT_SUSPEND) {
@@ -1264,9 +1264,9 @@ static int ahci_pci_device_suspend(struct pci_dev *pdev, pm_message_t mesg)
 
 static int ahci_pci_device_resume(struct pci_dev *pdev)
 {
-	struct ata_host_set *host_set = dev_get_drvdata(&pdev->dev);
-	struct ahci_host_priv *hpriv = host_set->private_data;
-	void __iomem *mmio = host_set->mmio_base;
+	struct ata_host *host = dev_get_drvdata(&pdev->dev);
+	struct ahci_host_priv *hpriv = host->private_data;
+	void __iomem *mmio = host->mmio_base;
 	int rc;
 
 	ata_pci_device_do_resume(pdev);
@@ -1276,20 +1276,20 @@ static int ahci_pci_device_resume(struct pci_dev *pdev)
 		if (rc)
 			return rc;
 
-		ahci_init_controller(mmio, pdev, host_set->n_ports, hpriv->cap);
+		ahci_init_controller(mmio, pdev, host->n_ports, hpriv->cap);
 	}
 
-	ata_host_set_resume(host_set);
+	ata_host_resume(host);
 
 	return 0;
 }
 
 static int ahci_port_start(struct ata_port *ap)
 {
-	struct device *dev = ap->host_set->dev;
-	struct ahci_host_priv *hpriv = ap->host_set->private_data;
+	struct device *dev = ap->host->dev;
+	struct ahci_host_priv *hpriv = ap->host->private_data;
 	struct ahci_port_priv *pp;
-	void __iomem *mmio = ap->host_set->mmio_base;
+	void __iomem *mmio = ap->host->mmio_base;
 	void __iomem *port_mmio = ahci_port_base(mmio, ap->port_no);
 	void *mem;
 	dma_addr_t mem_dma;
@@ -1350,10 +1350,10 @@ static int ahci_port_start(struct ata_port *ap)
 
 static void ahci_port_stop(struct ata_port *ap)
 {
-	struct device *dev = ap->host_set->dev;
-	struct ahci_host_priv *hpriv = ap->host_set->private_data;
+	struct device *dev = ap->host->dev;
+	struct ahci_host_priv *hpriv = ap->host->private_data;
 	struct ahci_port_priv *pp = ap->private_data;
-	void __iomem *mmio = ap->host_set->mmio_base;
+	void __iomem *mmio = ap->host->mmio_base;
 	void __iomem *port_mmio = ahci_port_base(mmio, ap->port_no);
 	const char *emsg = NULL;
 	int rc;
@@ -1581,7 +1581,7 @@ static int ahci_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 	memset(hpriv, 0, sizeof(*hpriv));
 
 	probe_ent->sht		= ahci_port_info[board_idx].sht;
-	probe_ent->host_flags	= ahci_port_info[board_idx].host_flags;
+	probe_ent->port_flags	= ahci_port_info[board_idx].flags;
 	probe_ent->pio_mask	= ahci_port_info[board_idx].pio_mask;
 	probe_ent->udma_mask	= ahci_port_info[board_idx].udma_mask;
 	probe_ent->port_ops	= ahci_port_info[board_idx].port_ops;
@@ -1599,9 +1599,9 @@ static int ahci_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (rc)
 		goto err_out_hpriv;
 
-	if (!(probe_ent->host_flags & AHCI_FLAG_NO_NCQ) &&
+	if (!(probe_ent->port_flags & AHCI_FLAG_NO_NCQ) &&
 	    (hpriv->cap & HOST_CAP_NCQ))
-		probe_ent->host_flags |= ATA_FLAG_NCQ;
+		probe_ent->port_flags |= ATA_FLAG_NCQ;
 
 	ahci_print_info(probe_ent);
 
@@ -1632,27 +1632,27 @@ err_out:
 static void ahci_remove_one (struct pci_dev *pdev)
 {
 	struct device *dev = pci_dev_to_dev(pdev);
-	struct ata_host_set *host_set = dev_get_drvdata(dev);
-	struct ahci_host_priv *hpriv = host_set->private_data;
+	struct ata_host *host = dev_get_drvdata(dev);
+	struct ahci_host_priv *hpriv = host->private_data;
 	unsigned int i;
 	int have_msi;
 
-	for (i = 0; i < host_set->n_ports; i++)
-		ata_port_detach(host_set->ports[i]);
+	for (i = 0; i < host->n_ports; i++)
+		ata_port_detach(host->ports[i]);
 
 	have_msi = hpriv->flags & AHCI_FLAG_MSI;
-	free_irq(host_set->irq, host_set);
+	free_irq(host->irq, host);
 
-	for (i = 0; i < host_set->n_ports; i++) {
-		struct ata_port *ap = host_set->ports[i];
+	for (i = 0; i < host->n_ports; i++) {
+		struct ata_port *ap = host->ports[i];
 
-		ata_scsi_release(ap->host);
-		scsi_host_put(ap->host);
+		ata_scsi_release(ap->scsi_host);
+		scsi_host_put(ap->scsi_host);
 	}
 
 	kfree(hpriv);
-	pci_iounmap(pdev, host_set->mmio_base);
-	kfree(host_set);
+	pci_iounmap(pdev, host->mmio_base);
+	kfree(host);
 
 	if (have_msi)
 		pci_disable_msi(pdev);
diff --git a/drivers/ata/ata_piix.c b/drivers/ata/ata_piix.c
index 0ca4c3b78dc5..22b2dba90b9a 100644
--- a/drivers/ata/ata_piix.c
+++ b/drivers/ata/ata_piix.c
@@ -151,7 +151,7 @@ struct piix_host_priv {
 
 static int piix_init_one (struct pci_dev *pdev,
 				    const struct pci_device_id *ent);
-static void piix_host_stop(struct ata_host_set *host_set);
+static void piix_host_stop(struct ata_host *host);
 static void piix_set_piomode (struct ata_port *ap, struct ata_device *adev);
 static void piix_set_dmamode (struct ata_port *ap, struct ata_device *adev);
 static void piix_pata_error_handler(struct ata_port *ap);
@@ -362,7 +362,7 @@ static struct ata_port_info piix_port_info[] = {
 	/* piix4_pata */
 	{
 		.sht		= &piix_sht,
-		.host_flags	= ATA_FLAG_SLAVE_POSS,
+		.flags		= ATA_FLAG_SLAVE_POSS,
 		.pio_mask	= 0x1f,	/* pio0-4 */
 #if 0
 		.mwdma_mask	= 0x06, /* mwdma1-2 */
@@ -376,7 +376,7 @@ static struct ata_port_info piix_port_info[] = {
 	/* ich5_pata */
 	{
 		.sht		= &piix_sht,
-		.host_flags	= ATA_FLAG_SLAVE_POSS | PIIX_FLAG_CHECKINTR,
+		.flags		= ATA_FLAG_SLAVE_POSS | PIIX_FLAG_CHECKINTR,
 		.pio_mask	= 0x1f,	/* pio0-4 */
 #if 0
 		.mwdma_mask	= 0x06, /* mwdma1-2 */
@@ -390,7 +390,7 @@ static struct ata_port_info piix_port_info[] = {
 	/* ich5_sata */
 	{
 		.sht		= &piix_sht,
-		.host_flags	= ATA_FLAG_SATA | PIIX_FLAG_CHECKINTR |
+		.flags		= ATA_FLAG_SATA | PIIX_FLAG_CHECKINTR |
 				  PIIX_FLAG_IGNORE_PCS,
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma0-2 */
@@ -401,7 +401,7 @@ static struct ata_port_info piix_port_info[] = {
 	/* i6300esb_sata */
 	{
 		.sht		= &piix_sht,
-		.host_flags	= ATA_FLAG_SATA |
+		.flags		= ATA_FLAG_SATA |
 				  PIIX_FLAG_CHECKINTR | PIIX_FLAG_IGNORE_PCS,
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma0-2 */
@@ -412,7 +412,7 @@ static struct ata_port_info piix_port_info[] = {
 	/* ich6_sata */
 	{
 		.sht		= &piix_sht,
-		.host_flags	= ATA_FLAG_SATA |
+		.flags		= ATA_FLAG_SATA |
 				  PIIX_FLAG_CHECKINTR | PIIX_FLAG_SCR,
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma0-2 */
@@ -423,7 +423,7 @@ static struct ata_port_info piix_port_info[] = {
 	/* ich6_sata_ahci */
 	{
 		.sht		= &piix_sht,
-		.host_flags	= ATA_FLAG_SATA |
+		.flags		= ATA_FLAG_SATA |
 				  PIIX_FLAG_CHECKINTR | PIIX_FLAG_SCR |
 				  PIIX_FLAG_AHCI,
 		.pio_mask	= 0x1f,	/* pio0-4 */
@@ -435,7 +435,7 @@ static struct ata_port_info piix_port_info[] = {
 	/* ich6m_sata_ahci */
 	{
 		.sht		= &piix_sht,
-		.host_flags	= ATA_FLAG_SATA |
+		.flags		= ATA_FLAG_SATA |
 				  PIIX_FLAG_CHECKINTR | PIIX_FLAG_SCR |
 				  PIIX_FLAG_AHCI,
 		.pio_mask	= 0x1f,	/* pio0-4 */
@@ -447,7 +447,7 @@ static struct ata_port_info piix_port_info[] = {
 	/* ich8_sata_ahci */
 	{
 		.sht		= &piix_sht,
-		.host_flags	= ATA_FLAG_SATA |
+		.flags		= ATA_FLAG_SATA |
 				  PIIX_FLAG_CHECKINTR | PIIX_FLAG_SCR |
 				  PIIX_FLAG_AHCI,
 		.pio_mask	= 0x1f,	/* pio0-4 */
@@ -485,7 +485,7 @@ MODULE_PARM_DESC(force_pcs, "force honoring or ignoring PCS to work around "
  */
 static void piix_pata_cbl_detect(struct ata_port *ap)
 {
-	struct pci_dev *pdev = to_pci_dev(ap->host_set->dev);
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
 	u8 tmp, mask;
 
 	/* no 80c support in host controller? */
@@ -517,7 +517,7 @@ cbl40:
  */
 static int piix_pata_prereset(struct ata_port *ap)
 {
-	struct pci_dev *pdev = to_pci_dev(ap->host_set->dev);
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
 
 	if (!pci_test_config_bits(pdev, &piix_enable_bits[ap->port_no])) {
 		ata_port_printk(ap, KERN_INFO, "port disabled. ignoring.\n");
@@ -551,8 +551,8 @@ static void piix_pata_error_handler(struct ata_port *ap)
  */
 static unsigned int piix_sata_present_mask(struct ata_port *ap)
 {
-	struct pci_dev *pdev = to_pci_dev(ap->host_set->dev);
-	struct piix_host_priv *hpriv = ap->host_set->private_data;
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	struct piix_host_priv *hpriv = ap->host->private_data;
 	const unsigned int *map = hpriv->map;
 	int base = 2 * ap->port_no;
 	unsigned int present_mask = 0;
@@ -631,7 +631,7 @@ static void piix_sata_error_handler(struct ata_port *ap)
 static void piix_set_piomode (struct ata_port *ap, struct ata_device *adev)
 {
 	unsigned int pio	= adev->pio_mode - XFER_PIO_0;
-	struct pci_dev *dev	= to_pci_dev(ap->host_set->dev);
+	struct pci_dev *dev	= to_pci_dev(ap->host->dev);
 	unsigned int is_slave	= (adev->devno != 0);
 	unsigned int master_port= ap->port_no ? 0x42 : 0x40;
 	unsigned int slave_port	= 0x44;
@@ -683,7 +683,7 @@ static void piix_set_piomode (struct ata_port *ap, struct ata_device *adev)
 static void piix_set_dmamode (struct ata_port *ap, struct ata_device *adev)
 {
 	unsigned int udma	= adev->dma_mode; /* FIXME: MWDMA too */
-	struct pci_dev *dev	= to_pci_dev(ap->host_set->dev);
+	struct pci_dev *dev	= to_pci_dev(ap->host->dev);
 	u8 maslave		= ap->port_no ? 0x42 : 0x40;
 	u8 speed		= udma;
 	unsigned int drive_dn	= (ap->port_no ? 2 : 0) + adev->devno;
@@ -835,13 +835,13 @@ static void __devinit piix_init_pcs(struct pci_dev *pdev,
 	if (force_pcs == 1) {
 		dev_printk(KERN_INFO, &pdev->dev,
 			   "force ignoring PCS (0x%x)\n", new_pcs);
-		pinfo[0].host_flags |= PIIX_FLAG_IGNORE_PCS;
-		pinfo[1].host_flags |= PIIX_FLAG_IGNORE_PCS;
+		pinfo[0].flags |= PIIX_FLAG_IGNORE_PCS;
+		pinfo[1].flags |= PIIX_FLAG_IGNORE_PCS;
 	} else if (force_pcs == 2) {
 		dev_printk(KERN_INFO, &pdev->dev,
 			   "force honoring PCS (0x%x)\n", new_pcs);
-		pinfo[0].host_flags &= ~PIIX_FLAG_IGNORE_PCS;
-		pinfo[1].host_flags &= ~PIIX_FLAG_IGNORE_PCS;
+		pinfo[0].flags &= ~PIIX_FLAG_IGNORE_PCS;
+		pinfo[1].flags &= ~PIIX_FLAG_IGNORE_PCS;
 	}
 }
 
@@ -881,7 +881,7 @@ static void __devinit piix_init_sata_map(struct pci_dev *pdev,
 		default:
 			printk(" P%d", map[i]);
 			if (i & 1)
-				pinfo[i / 2].host_flags |= ATA_FLAG_SLAVE_POSS;
+				pinfo[i / 2].flags |= ATA_FLAG_SLAVE_POSS;
 			break;
 		}
 	}
@@ -916,7 +916,7 @@ static int piix_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 	struct ata_port_info port_info[2];
 	struct ata_port_info *ppinfo[2] = { &port_info[0], &port_info[1] };
 	struct piix_host_priv *hpriv;
-	unsigned long host_flags;
+	unsigned long port_flags;
 
 	if (!printed_version++)
 		dev_printk(KERN_DEBUG, &pdev->dev,
@@ -935,9 +935,9 @@ static int piix_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 	port_info[0].private_data = hpriv;
 	port_info[1].private_data = hpriv;
 
-	host_flags = port_info[0].host_flags;
+	port_flags = port_info[0].flags;
 
-	if (host_flags & PIIX_FLAG_AHCI) {
+	if (port_flags & PIIX_FLAG_AHCI) {
 		u8 tmp;
 		pci_read_config_byte(pdev, PIIX_SCC, &tmp);
 		if (tmp == PIIX_AHCI_DEVICE) {
@@ -948,7 +948,7 @@ static int piix_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 	}
 
 	/* Initialize SATA map */
-	if (host_flags & ATA_FLAG_SATA) {
+	if (port_flags & ATA_FLAG_SATA) {
 		piix_init_sata_map(pdev, port_info,
 				   piix_map_db_table[ent->driver_data]);
 		piix_init_pcs(pdev, port_info,
@@ -961,7 +961,7 @@ static int piix_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 	 * MSI is disabled (and it is disabled, as we don't use
 	 * message-signalled interrupts currently).
 	 */
-	if (host_flags & PIIX_FLAG_CHECKINTR)
+	if (port_flags & PIIX_FLAG_CHECKINTR)
 		pci_intx(pdev, 1);
 
 	if (piix_check_450nx_errata(pdev)) {
@@ -976,11 +976,11 @@ static int piix_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 	return ata_pci_init_one(pdev, ppinfo, 2);
 }
 
-static void piix_host_stop(struct ata_host_set *host_set)
+static void piix_host_stop(struct ata_host *host)
 {
-	struct piix_host_priv *hpriv = host_set->private_data;
+	struct piix_host_priv *hpriv = host->private_data;
 
-	ata_host_stop(host_set);
+	ata_host_stop(host);
 
 	kfree(hpriv);
 }
diff --git a/drivers/ata/libata-bmdma.c b/drivers/ata/libata-bmdma.c
index 158f62dbf21b..760502859821 100644
--- a/drivers/ata/libata-bmdma.c
+++ b/drivers/ata/libata-bmdma.c
@@ -193,7 +193,7 @@ void ata_tf_load(struct ata_port *ap, const struct ata_taskfile *tf)
  *	synchronization with interrupt handler / other threads.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 static void ata_exec_command_pio(struct ata_port *ap, const struct ata_taskfile *tf)
@@ -216,7 +216,7 @@ static void ata_exec_command_pio(struct ata_port *ap, const struct ata_taskfile
  *	FIXME: missing write posting for 400nS delay enforcement
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 static void ata_exec_command_mmio(struct ata_port *ap, const struct ata_taskfile *tf)
@@ -237,7 +237,7 @@ static void ata_exec_command_mmio(struct ata_port *ap, const struct ata_taskfile
  *	synchronization with interrupt handler / other threads.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 void ata_exec_command(struct ata_port *ap, const struct ata_taskfile *tf)
 {
@@ -422,7 +422,7 @@ u8 ata_altstatus(struct ata_port *ap)
  *	@qc: Info associated with this ATA transaction.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 static void ata_bmdma_setup_mmio (struct ata_queued_cmd *qc)
@@ -452,7 +452,7 @@ static void ata_bmdma_setup_mmio (struct ata_queued_cmd *qc)
  *	@qc: Info associated with this ATA transaction.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 static void ata_bmdma_start_mmio (struct ata_queued_cmd *qc)
@@ -483,7 +483,7 @@ static void ata_bmdma_start_mmio (struct ata_queued_cmd *qc)
  *	@qc: Info associated with this ATA transaction.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 static void ata_bmdma_setup_pio (struct ata_queued_cmd *qc)
@@ -511,7 +511,7 @@ static void ata_bmdma_setup_pio (struct ata_queued_cmd *qc)
  *	@qc: Info associated with this ATA transaction.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 static void ata_bmdma_start_pio (struct ata_queued_cmd *qc)
@@ -535,7 +535,7 @@ static void ata_bmdma_start_pio (struct ata_queued_cmd *qc)
  *	May be used as the bmdma_start() entry in ata_port_operations.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 void ata_bmdma_start(struct ata_queued_cmd *qc)
 {
@@ -557,7 +557,7 @@ void ata_bmdma_start(struct ata_queued_cmd *qc)
  *	May be used as the bmdma_setup() entry in ata_port_operations.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 void ata_bmdma_setup(struct ata_queued_cmd *qc)
 {
@@ -577,7 +577,7 @@ void ata_bmdma_setup(struct ata_queued_cmd *qc)
  *	May be used as the irq_clear() entry in ata_port_operations.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 void ata_bmdma_irq_clear(struct ata_port *ap)
@@ -605,7 +605,7 @@ void ata_bmdma_irq_clear(struct ata_port *ap)
  *	May be used as the bmdma_status() entry in ata_port_operations.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 u8 ata_bmdma_status(struct ata_port *ap)
@@ -629,7 +629,7 @@ u8 ata_bmdma_status(struct ata_port *ap)
  *	May be used as the bmdma_stop() entry in ata_port_operations.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 void ata_bmdma_stop(struct ata_queued_cmd *qc)
@@ -838,7 +838,7 @@ ata_pci_init_native_mode(struct pci_dev *pdev, struct ata_port_info **port, int
 		bmdma = pci_resource_start(pdev, 4);
 		if (bmdma) {
 			if (inb(bmdma + 2) & 0x80)
-				probe_ent->host_set_flags |= ATA_HOST_SIMPLEX;
+				probe_ent->_host_flags |= ATA_HOST_SIMPLEX;
 			probe_ent->port[p].bmdma_addr = bmdma;
 		}
 		ata_std_ports(&probe_ent->port[p]);
@@ -854,7 +854,7 @@ ata_pci_init_native_mode(struct pci_dev *pdev, struct ata_port_info **port, int
 		if (bmdma) {
 			bmdma += 8;
 			if(inb(bmdma + 2) & 0x80)
-				probe_ent->host_set_flags |= ATA_HOST_SIMPLEX;
+				probe_ent->_host_flags |= ATA_HOST_SIMPLEX;
 			probe_ent->port[p].bmdma_addr = bmdma;
 		}
 		ata_std_ports(&probe_ent->port[p]);
@@ -887,7 +887,7 @@ static struct ata_probe_ent *ata_pci_init_legacy_port(struct pci_dev *pdev,
 		if (bmdma) {
 			probe_ent->port[0].bmdma_addr = bmdma;
 			if (inb(bmdma + 2) & 0x80)
-				probe_ent->host_set_flags |= ATA_HOST_SIMPLEX;
+				probe_ent->_host_flags |= ATA_HOST_SIMPLEX;
 		}
 		ata_std_ports(&probe_ent->port[0]);
 	} else
@@ -904,7 +904,7 @@ static struct ata_probe_ent *ata_pci_init_legacy_port(struct pci_dev *pdev,
 		if (bmdma) {
 			probe_ent->port[1].bmdma_addr = bmdma + 8;
 			if (inb(bmdma + 10) & 0x80)
-				probe_ent->host_set_flags |= ATA_HOST_SIMPLEX;
+				probe_ent->_host_flags |= ATA_HOST_SIMPLEX;
 		}
 		ata_std_ports(&probe_ent->port[1]);
 	} else
@@ -957,7 +957,7 @@ int ata_pci_init_one (struct pci_dev *pdev, struct ata_port_info **port_info,
 	else
 		port[1] = port[0];
 
-	if ((port[0]->host_flags & ATA_FLAG_NO_LEGACY) == 0
+	if ((port[0]->flags & ATA_FLAG_NO_LEGACY) == 0
 	    && (pdev->class >> 8) == PCI_CLASS_STORAGE_IDE) {
 		/* TODO: What if one channel is in native mode ... */
 		pci_read_config_byte(pdev, PCI_CLASS_PROG, &tmp8);
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 9092416a6301..1c9315401f7a 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -1335,7 +1335,7 @@ static void ata_dev_config_ncq(struct ata_device *dev,
 	}
 
 	if (ap->flags & ATA_FLAG_NCQ) {
-		hdepth = min(ap->host->can_queue, ATA_MAX_QUEUE - 1);
+		hdepth = min(ap->scsi_host->can_queue, ATA_MAX_QUEUE - 1);
 		dev->flags |= ATA_DFLAG_NCQ;
 	}
 
@@ -1349,12 +1349,13 @@ static void ata_set_port_max_cmd_len(struct ata_port *ap)
 {
 	int i;
 
-	if (ap->host) {
-		ap->host->max_cmd_len = 0;
+	if (ap->scsi_host) {
+		unsigned int len = 0;
+
 		for (i = 0; i < ATA_MAX_DEVICES; i++)
-			ap->host->max_cmd_len = max_t(unsigned int,
-						      ap->host->max_cmd_len,
-						      ap->device[i].cdb_len);
+			len = max(len, ap->device[i].cdb_len);
+
+		ap->scsi_host->max_cmd_len = len;
 	}
 }
 
@@ -1662,7 +1663,7 @@ int ata_bus_probe(struct ata_port *ap)
  *	Modify @ap data structure such that the system
  *	thinks that the entire port is enabled.
  *
- *	LOCKING: host_set lock, or some other form of
+ *	LOCKING: host lock, or some other form of
  *	serialization.
  */
 
@@ -1800,7 +1801,7 @@ struct ata_device *ata_dev_pair(struct ata_device *adev)
  *	never attempt to probe or communicate with devices
  *	on this port.
  *
- *	LOCKING: host_set lock, or some other form of
+ *	LOCKING: host lock, or some other form of
  *	serialization.
  */
 
@@ -2258,8 +2259,8 @@ int ata_set_mode(struct ata_port *ap, struct ata_device **r_failed_dev)
 	/* Record simplex status. If we selected DMA then the other
 	 * host channels are not permitted to do so.
 	 */
-	if (used_dma && (ap->host_set->flags & ATA_HOST_SIMPLEX))
-		ap->host_set->simplex_claimed = 1;
+	if (used_dma && (ap->host->flags & ATA_HOST_SIMPLEX))
+		ap->host->simplex_claimed = 1;
 
 	/* step5: chip specific finalisation */
 	if (ap->ops->post_set_mode)
@@ -2281,7 +2282,7 @@ int ata_set_mode(struct ata_port *ap, struct ata_device **r_failed_dev)
  *	other threads.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 static inline void ata_tf_to_host(struct ata_port *ap,
@@ -2445,7 +2446,7 @@ static unsigned int ata_bus_softreset(struct ata_port *ap,
  *
  *	LOCKING:
  *	PCI/etc. bus probe sem.
- *	Obtains host_set lock.
+ *	Obtains host lock.
  *
  *	SIDE EFFECTS:
  *	Sets ATA_FLAG_DISABLED if bus reset fails.
@@ -3080,7 +3081,7 @@ static int ata_dma_blacklisted(const struct ata_device *dev)
 static void ata_dev_xfermask(struct ata_device *dev)
 {
 	struct ata_port *ap = dev->ap;
-	struct ata_host_set *hs = ap->host_set;
+	struct ata_host *host = ap->host;
 	unsigned long xfer_mask;
 
 	/* controller modes available */
@@ -3114,7 +3115,7 @@ static void ata_dev_xfermask(struct ata_device *dev)
 			       "device is on DMA blacklist, disabling DMA\n");
 	}
 
-	if ((hs->flags & ATA_HOST_SIMPLEX) && hs->simplex_claimed) {
+	if ((host->flags & ATA_HOST_SIMPLEX) && host->simplex_claimed) {
 		xfer_mask &= ~(ATA_MASK_MWDMA | ATA_MASK_UDMA);
 		ata_dev_printk(dev, KERN_WARNING, "simplex DMA is claimed by "
 			       "other device, disabling DMA\n");
@@ -3207,7 +3208,7 @@ static unsigned int ata_dev_init_params(struct ata_device *dev,
  *	Unmap all mapped DMA memory associated with this command.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 static void ata_sg_clean(struct ata_queued_cmd *qc)
@@ -3267,7 +3268,7 @@ static void ata_sg_clean(struct ata_queued_cmd *qc)
  *	associated with the current disk command.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  */
 static void ata_fill_sg(struct ata_queued_cmd *qc)
@@ -3319,7 +3320,7 @@ static void ata_fill_sg(struct ata_queued_cmd *qc)
  *	supplied PACKET command.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS: 0 when ATAPI DMA can be used
  *               nonzero otherwise
@@ -3341,7 +3342,7 @@ int ata_check_atapi_dma(struct ata_queued_cmd *qc)
  *	Prepare ATA taskfile for submission.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 void ata_qc_prep(struct ata_queued_cmd *qc)
 {
@@ -3363,7 +3364,7 @@ void ata_noop_qc_prep(struct ata_queued_cmd *qc) { }
  *	to point to a single memory buffer, @buf of byte length @buflen.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 void ata_sg_init_one(struct ata_queued_cmd *qc, void *buf, unsigned int buflen)
@@ -3394,7 +3395,7 @@ void ata_sg_init_one(struct ata_queued_cmd *qc, void *buf, unsigned int buflen)
  *	elements.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 void ata_sg_init(struct ata_queued_cmd *qc, struct scatterlist *sg,
@@ -3413,7 +3414,7 @@ void ata_sg_init(struct ata_queued_cmd *qc, struct scatterlist *sg,
  *	DMA-map the memory buffer associated with queued_cmd @qc.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	Zero on success, negative on error.
@@ -3482,7 +3483,7 @@ skip_map:
  *	DMA-map the scatter-gather table associated with queued_cmd @qc.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	Zero on success, negative on error.
@@ -3991,7 +3992,7 @@ static inline int ata_hsm_ok_in_wq(struct ata_port *ap, struct ata_queued_cmd *q
  *	Finish @qc which is running on standard HSM.
  *
  *	LOCKING:
- *	If @in_wq is zero, spin_lock_irqsave(host_set lock).
+ *	If @in_wq is zero, spin_lock_irqsave(host lock).
  *	Otherwise, none on entry and grabs host lock.
  */
 static void ata_hsm_qc_complete(struct ata_queued_cmd *qc, int in_wq)
@@ -4003,8 +4004,8 @@ static void ata_hsm_qc_complete(struct ata_queued_cmd *qc, int in_wq)
 		if (in_wq) {
 			spin_lock_irqsave(ap->lock, flags);
 
-			/* EH might have kicked in while host_set lock
-			 * is released.
+			/* EH might have kicked in while host lock is
+			 * released.
 			 */
 			qc = ata_qc_from_tag(ap, qc->tag);
 			if (qc) {
@@ -4369,7 +4370,7 @@ struct ata_queued_cmd *ata_qc_new_init(struct ata_device *dev)
  *	in case something prevents using it.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 void ata_qc_free(struct ata_queued_cmd *qc)
 {
@@ -4422,7 +4423,7 @@ void __ata_qc_complete(struct ata_queued_cmd *qc)
  *	command has completed, with either an ok or not-ok status.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 void ata_qc_complete(struct ata_queued_cmd *qc)
 {
@@ -4485,7 +4486,7 @@ void ata_qc_complete(struct ata_queued_cmd *qc)
  *	and commands are completed accordingly.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	Number of completed commands on success, -errno otherwise.
@@ -4556,7 +4557,7 @@ static inline int ata_should_dma_map(struct ata_queued_cmd *qc)
  *	writing the taskfile to hardware, starting the command.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 void ata_qc_issue(struct ata_queued_cmd *qc)
 {
@@ -4617,7 +4618,7 @@ err:
  *	May be used as the qc_issue() entry in ata_port_operations.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	Zero on success, AC_ERR_* mask on failure
@@ -4746,7 +4747,7 @@ unsigned int ata_qc_issue_prot(struct ata_queued_cmd *qc)
  *	handled via polling with interrupts disabled (nIEN bit).
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	One if interrupt was handled, zero if not (shared irq).
@@ -4833,14 +4834,14 @@ idle_irq:
 /**
  *	ata_interrupt - Default ATA host interrupt handler
  *	@irq: irq line (unused)
- *	@dev_instance: pointer to our ata_host_set information structure
+ *	@dev_instance: pointer to our ata_host information structure
  *	@regs: unused
  *
  *	Default interrupt handler for PCI IDE devices.  Calls
  *	ata_host_intr() for each port that is not disabled.
  *
  *	LOCKING:
- *	Obtains host_set lock during operation.
+ *	Obtains host lock during operation.
  *
  *	RETURNS:
  *	IRQ_NONE or IRQ_HANDLED.
@@ -4848,18 +4849,18 @@ idle_irq:
 
 irqreturn_t ata_interrupt (int irq, void *dev_instance, struct pt_regs *regs)
 {
-	struct ata_host_set *host_set = dev_instance;
+	struct ata_host *host = dev_instance;
 	unsigned int i;
 	unsigned int handled = 0;
 	unsigned long flags;
 
 	/* TODO: make _irqsave conditional on x86 PCI IDE legacy mode */
-	spin_lock_irqsave(&host_set->lock, flags);
+	spin_lock_irqsave(&host->lock, flags);
 
-	for (i = 0; i < host_set->n_ports; i++) {
+	for (i = 0; i < host->n_ports; i++) {
 		struct ata_port *ap;
 
-		ap = host_set->ports[i];
+		ap = host->ports[i];
 		if (ap &&
 		    !(ap->flags & ATA_FLAG_DISABLED)) {
 			struct ata_queued_cmd *qc;
@@ -4871,7 +4872,7 @@ irqreturn_t ata_interrupt (int irq, void *dev_instance, struct pt_regs *regs)
 		}
 	}
 
-	spin_unlock_irqrestore(&host_set->lock, flags);
+	spin_unlock_irqrestore(&host->lock, flags);
 
 	return IRQ_RETVAL(handled);
 }
@@ -5036,15 +5037,15 @@ int ata_flush_cache(struct ata_device *dev)
 	return 0;
 }
 
-static int ata_host_set_request_pm(struct ata_host_set *host_set,
-				   pm_message_t mesg, unsigned int action,
-				   unsigned int ehi_flags, int wait)
+static int ata_host_request_pm(struct ata_host *host, pm_message_t mesg,
+			       unsigned int action, unsigned int ehi_flags,
+			       int wait)
 {
 	unsigned long flags;
 	int i, rc;
 
-	for (i = 0; i < host_set->n_ports; i++) {
-		struct ata_port *ap = host_set->ports[i];
+	for (i = 0; i < host->n_ports; i++) {
+		struct ata_port *ap = host->ports[i];
 
 		/* Previous resume operation might still be in
 		 * progress.  Wait for PM_PENDING to clear.
@@ -5084,11 +5085,11 @@ static int ata_host_set_request_pm(struct ata_host_set *host_set,
 }
 
 /**
- *	ata_host_set_suspend - suspend host_set
- *	@host_set: host_set to suspend
+ *	ata_host_suspend - suspend host
+ *	@host: host to suspend
  *	@mesg: PM message
  *
- *	Suspend @host_set.  Actual operation is performed by EH.  This
+ *	Suspend @host.  Actual operation is performed by EH.  This
  *	function requests EH to perform PM operations and waits for EH
  *	to finish.
  *
@@ -5098,11 +5099,11 @@ static int ata_host_set_request_pm(struct ata_host_set *host_set,
  *	RETURNS:
  *	0 on success, -errno on failure.
  */
-int ata_host_set_suspend(struct ata_host_set *host_set, pm_message_t mesg)
+int ata_host_suspend(struct ata_host *host, pm_message_t mesg)
 {
 	int i, j, rc;
 
-	rc = ata_host_set_request_pm(host_set, mesg, 0, ATA_EHI_QUIET, 1);
+	rc = ata_host_request_pm(host, mesg, 0, ATA_EHI_QUIET, 1);
 	if (rc)
 		goto fail;
 
@@ -5110,8 +5111,8 @@ int ata_host_set_suspend(struct ata_host_set *host_set, pm_message_t mesg)
 	 * This happens if hotplug occurs between completion of device
 	 * suspension and here.
 	 */
-	for (i = 0; i < host_set->n_ports; i++) {
-		struct ata_port *ap = host_set->ports[i];
+	for (i = 0; i < host->n_ports; i++) {
+		struct ata_port *ap = host->ports[i];
 
 		for (j = 0; j < ATA_MAX_DEVICES; j++) {
 			struct ata_device *dev = &ap->device[j];
@@ -5126,30 +5127,30 @@ int ata_host_set_suspend(struct ata_host_set *host_set, pm_message_t mesg)
 		}
 	}
 
-	host_set->dev->power.power_state = mesg;
+	host->dev->power.power_state = mesg;
 	return 0;
 
  fail:
-	ata_host_set_resume(host_set);
+	ata_host_resume(host);
 	return rc;
 }
 
 /**
- *	ata_host_set_resume - resume host_set
- *	@host_set: host_set to resume
+ *	ata_host_resume - resume host
+ *	@host: host to resume
  *
- *	Resume @host_set.  Actual operation is performed by EH.  This
+ *	Resume @host.  Actual operation is performed by EH.  This
  *	function requests EH to perform PM operations and returns.
  *	Note that all resume operations are performed parallely.
  *
  *	LOCKING:
  *	Kernel thread context (may sleep).
  */
-void ata_host_set_resume(struct ata_host_set *host_set)
+void ata_host_resume(struct ata_host *host)
 {
-	ata_host_set_request_pm(host_set, PMSG_ON, ATA_EH_SOFTRESET,
-				ATA_EHI_NO_AUTOPSY | ATA_EHI_QUIET, 0);
-	host_set->dev->power.power_state = PMSG_ON;
+	ata_host_request_pm(host, PMSG_ON, ATA_EH_SOFTRESET,
+			    ATA_EHI_NO_AUTOPSY | ATA_EHI_QUIET, 0);
+	host->dev->power.power_state = PMSG_ON;
 }
 
 /**
@@ -5206,10 +5207,10 @@ void ata_port_stop (struct ata_port *ap)
 	ata_pad_free(ap, dev);
 }
 
-void ata_host_stop (struct ata_host_set *host_set)
+void ata_host_stop (struct ata_host *host)
 {
-	if (host_set->mmio_base)
-		iounmap(host_set->mmio_base);
+	if (host->mmio_base)
+		iounmap(host->mmio_base);
 }
 
 /**
@@ -5231,7 +5232,7 @@ void ata_dev_init(struct ata_device *dev)
 
 	/* High bits of dev->flags are used to record warm plug
 	 * requests which occur asynchronously.  Synchronize using
-	 * host_set lock.
+	 * host lock.
 	 */
 	spin_lock_irqsave(ap->lock, flags);
 	dev->flags &= ~ATA_DFLAG_INIT_MASK;
@@ -5247,7 +5248,7 @@ void ata_dev_init(struct ata_device *dev)
 /**
  *	ata_port_init - Initialize an ata_port structure
  *	@ap: Structure to initialize
- *	@host_set: Collection of hosts to which @ap belongs
+ *	@host: Collection of hosts to which @ap belongs
  *	@ent: Probe information provided by low-level driver
  *	@port_no: Port number associated with this ata_port
  *
@@ -5256,22 +5257,22 @@ void ata_dev_init(struct ata_device *dev)
  *	LOCKING:
  *	Inherited from caller.
  */
-void ata_port_init(struct ata_port *ap, struct ata_host_set *host_set,
+void ata_port_init(struct ata_port *ap, struct ata_host *host,
 		   const struct ata_probe_ent *ent, unsigned int port_no)
 {
 	unsigned int i;
 
-	ap->lock = &host_set->lock;
+	ap->lock = &host->lock;
 	ap->flags = ATA_FLAG_DISABLED;
 	ap->id = ata_unique_id++;
 	ap->ctl = ATA_DEVCTL_OBS;
-	ap->host_set = host_set;
+	ap->host = host;
 	ap->dev = ent->dev;
 	ap->port_no = port_no;
 	ap->pio_mask = ent->pio_mask;
 	ap->mwdma_mask = ent->mwdma_mask;
 	ap->udma_mask = ent->udma_mask;
-	ap->flags |= ent->host_flags;
+	ap->flags |= ent->port_flags;
 	ap->ops = ent->port_ops;
 	ap->hw_sata_spd_limit = UINT_MAX;
 	ap->active_tag = ATA_TAG_POISON;
@@ -5324,7 +5325,7 @@ void ata_port_init(struct ata_port *ap, struct ata_host_set *host_set,
  */
 static void ata_port_init_shost(struct ata_port *ap, struct Scsi_Host *shost)
 {
-	ap->host = shost;
+	ap->scsi_host = shost;
 
 	shost->unique_id = ap->id;
 	shost->max_id = 16;
@@ -5336,7 +5337,7 @@ static void ata_port_init_shost(struct ata_port *ap, struct Scsi_Host *shost)
 /**
  *	ata_port_add - Attach low-level ATA driver to system
  *	@ent: Information provided by low-level driver
- *	@host_set: Collections of ports to which we add
+ *	@host: Collections of ports to which we add
  *	@port_no: Port number associated with this host
  *
  *	Attach low-level ATA driver to system.
@@ -5348,7 +5349,7 @@ static void ata_port_init_shost(struct ata_port *ap, struct Scsi_Host *shost)
  *	New ata_port on success, for NULL on error.
  */
 static struct ata_port * ata_port_add(const struct ata_probe_ent *ent,
-				      struct ata_host_set *host_set,
+				      struct ata_host *host,
 				      unsigned int port_no)
 {
 	struct Scsi_Host *shost;
@@ -5357,7 +5358,7 @@ static struct ata_port * ata_port_add(const struct ata_probe_ent *ent,
 	DPRINTK("ENTER\n");
 
 	if (!ent->port_ops->error_handler &&
-	    !(ent->host_flags & (ATA_FLAG_SATA_RESET | ATA_FLAG_SRST))) {
+	    !(ent->port_flags & (ATA_FLAG_SATA_RESET | ATA_FLAG_SRST))) {
 		printk(KERN_ERR "ata%u: no reset mechanism available\n",
 		       port_no);
 		return NULL;
@@ -5371,32 +5372,31 @@ static struct ata_port * ata_port_add(const struct ata_probe_ent *ent,
 
 	ap = ata_shost_to_port(shost);
 
-	ata_port_init(ap, host_set, ent, port_no);
+	ata_port_init(ap, host, ent, port_no);
 	ata_port_init_shost(ap, shost);
 
 	return ap;
 }
 
 /**
- *	ata_sas_host_init - Initialize a host_set struct
- *	@host_set:	host_set to initialize
- *	@dev:		device host_set is attached to
- *	@flags:	host_set flags
- *	@ops:		port_ops
+ *	ata_sas_host_init - Initialize a host struct
+ *	@host:	host to initialize
+ *	@dev:	device host is attached to
+ *	@flags:	host flags
+ *	@ops:	port_ops
  *
  *	LOCKING:
  *	PCI/etc. bus probe sem.
  *
  */
 
-void ata_host_set_init(struct ata_host_set *host_set,
-		       struct device *dev, unsigned long flags,
-		       const struct ata_port_operations *ops)
+void ata_host_init(struct ata_host *host, struct device *dev,
+		   unsigned long flags, const struct ata_port_operations *ops)
 {
-	spin_lock_init(&host_set->lock);
-	host_set->dev = dev;
-	host_set->flags = flags;
-	host_set->ops = ops;
+	spin_lock_init(&host->lock);
+	host->dev = dev;
+	host->flags = flags;
+	host->ops = ops;
 }
 
 /**
@@ -5421,34 +5421,34 @@ int ata_device_add(const struct ata_probe_ent *ent)
 {
 	unsigned int i;
 	struct device *dev = ent->dev;
-	struct ata_host_set *host_set;
+	struct ata_host *host;
 	int rc;
 
 	DPRINTK("ENTER\n");
 	/* alloc a container for our list of ATA ports (buses) */
-	host_set = kzalloc(sizeof(struct ata_host_set) +
-			   (ent->n_ports * sizeof(void *)), GFP_KERNEL);
-	if (!host_set)
+	host = kzalloc(sizeof(struct ata_host) +
+		       (ent->n_ports * sizeof(void *)), GFP_KERNEL);
+	if (!host)
 		return 0;
 
-	ata_host_set_init(host_set, dev, ent->host_set_flags, ent->port_ops);
-	host_set->n_ports = ent->n_ports;
-	host_set->irq = ent->irq;
-	host_set->irq2 = ent->irq2;
-	host_set->mmio_base = ent->mmio_base;
-	host_set->private_data = ent->private_data;
+	ata_host_init(host, dev, ent->_host_flags, ent->port_ops);
+	host->n_ports = ent->n_ports;
+	host->irq = ent->irq;
+	host->irq2 = ent->irq2;
+	host->mmio_base = ent->mmio_base;
+	host->private_data = ent->private_data;
 
 	/* register each port bound to this device */
-	for (i = 0; i < host_set->n_ports; i++) {
+	for (i = 0; i < host->n_ports; i++) {
 		struct ata_port *ap;
 		unsigned long xfer_mode_mask;
 		int irq_line = ent->irq;
 
-		ap = ata_port_add(ent, host_set, i);
+		ap = ata_port_add(ent, host, i);
 		if (!ap)
 			goto err_out;
 
-		host_set->ports[i] = ap;
+		host->ports[i] = ap;
 
 		/* dummy? */
 		if (ent->dummy_port_mask & (1 << i)) {
@@ -5460,8 +5460,8 @@ int ata_device_add(const struct ata_probe_ent *ent)
 		/* start port */
 		rc = ap->ops->port_start(ap);
 		if (rc) {
-			host_set->ports[i] = NULL;
-			scsi_host_put(ap->host);
+			host->ports[i] = NULL;
+			scsi_host_put(ap->scsi_host);
 			goto err_out;
 		}
 
@@ -5484,13 +5484,13 @@ int ata_device_add(const struct ata_probe_ent *ent)
 				irq_line);
 
 		ata_chk_status(ap);
-		host_set->ops->irq_clear(ap);
+		host->ops->irq_clear(ap);
 		ata_eh_freeze_port(ap);	/* freeze port before requesting IRQ */
 	}
 
 	/* obtain irq, that may be shared between channels */
 	rc = request_irq(ent->irq, ent->port_ops->irq_handler, ent->irq_flags,
-			 DRV_NAME, host_set);
+			 DRV_NAME, host);
 	if (rc) {
 		dev_printk(KERN_ERR, dev, "irq %lu request failed: %d\n",
 			   ent->irq, rc);
@@ -5504,7 +5504,7 @@ int ata_device_add(const struct ata_probe_ent *ent)
 		BUG_ON(ent->irq == ent->irq2);
 
 		rc = request_irq(ent->irq2, ent->port_ops->irq_handler, ent->irq_flags,
-			 DRV_NAME, host_set);
+			 DRV_NAME, host);
 		if (rc) {
 			dev_printk(KERN_ERR, dev, "irq %lu request failed: %d\n",
 				   ent->irq2, rc);
@@ -5514,8 +5514,8 @@ int ata_device_add(const struct ata_probe_ent *ent)
 
 	/* perform each probe synchronously */
 	DPRINTK("probe begin\n");
-	for (i = 0; i < host_set->n_ports; i++) {
-		struct ata_port *ap = host_set->ports[i];
+	for (i = 0; i < host->n_ports; i++) {
+		struct ata_port *ap = host->ports[i];
 		u32 scontrol;
 		int rc;
 
@@ -5526,7 +5526,7 @@ int ata_device_add(const struct ata_probe_ent *ent)
 		}
 		ap->sata_spd_limit = ap->hw_sata_spd_limit;
 
-		rc = scsi_add_host(ap->host, dev);
+		rc = scsi_add_host(ap->scsi_host, dev);
 		if (rc) {
 			ata_port_printk(ap, KERN_ERR, "scsi_add_host failed\n");
 			/* FIXME: do something useful here */
@@ -5574,29 +5574,29 @@ int ata_device_add(const struct ata_probe_ent *ent)
 
 	/* probes are done, now scan each port's disk(s) */
 	DPRINTK("host probe begin\n");
-	for (i = 0; i < host_set->n_ports; i++) {
-		struct ata_port *ap = host_set->ports[i];
+	for (i = 0; i < host->n_ports; i++) {
+		struct ata_port *ap = host->ports[i];
 
 		ata_scsi_scan_host(ap);
 	}
 
-	dev_set_drvdata(dev, host_set);
+	dev_set_drvdata(dev, host);
 
 	VPRINTK("EXIT, returning %u\n", ent->n_ports);
 	return ent->n_ports; /* success */
 
 err_out_free_irq:
-	free_irq(ent->irq, host_set);
+	free_irq(ent->irq, host);
 err_out:
-	for (i = 0; i < host_set->n_ports; i++) {
-		struct ata_port *ap = host_set->ports[i];
+	for (i = 0; i < host->n_ports; i++) {
+		struct ata_port *ap = host->ports[i];
 		if (ap) {
 			ap->ops->port_stop(ap);
-			scsi_host_put(ap->host);
+			scsi_host_put(ap->scsi_host);
 		}
 	}
 
-	kfree(host_set);
+	kfree(host);
 	VPRINTK("EXIT, returning 0\n");
 	return 0;
 }
@@ -5656,12 +5656,12 @@ void ata_port_detach(struct ata_port *ap)
 
  skip_eh:
 	/* remove the associated SCSI host */
-	scsi_remove_host(ap->host);
+	scsi_remove_host(ap->scsi_host);
 }
 
 /**
- *	ata_host_set_remove - PCI layer callback for device removal
- *	@host_set: ATA host set that was removed
+ *	ata_host_remove - PCI layer callback for device removal
+ *	@host: ATA host set that was removed
  *
  *	Unregister all objects associated with this host set. Free those
  *	objects.
@@ -5670,21 +5670,21 @@ void ata_port_detach(struct ata_port *ap)
  *	Inherited from calling layer (may sleep).
  */
 
-void ata_host_set_remove(struct ata_host_set *host_set)
+void ata_host_remove(struct ata_host *host)
 {
 	unsigned int i;
 
-	for (i = 0; i < host_set->n_ports; i++)
-		ata_port_detach(host_set->ports[i]);
+	for (i = 0; i < host->n_ports; i++)
+		ata_port_detach(host->ports[i]);
 
-	free_irq(host_set->irq, host_set);
-	if (host_set->irq2)
-		free_irq(host_set->irq2, host_set);
+	free_irq(host->irq, host);
+	if (host->irq2)
+		free_irq(host->irq2, host);
 
-	for (i = 0; i < host_set->n_ports; i++) {
-		struct ata_port *ap = host_set->ports[i];
+	for (i = 0; i < host->n_ports; i++) {
+		struct ata_port *ap = host->ports[i];
 
-		ata_scsi_release(ap->host);
+		ata_scsi_release(ap->scsi_host);
 
 		if ((ap->flags & ATA_FLAG_NO_LEGACY) == 0) {
 			struct ata_ioports *ioaddr = &ap->ioaddr;
@@ -5696,13 +5696,13 @@ void ata_host_set_remove(struct ata_host_set *host_set)
 				release_region(ATA_SECONDARY_CMD, 8);
 		}
 
-		scsi_host_put(ap->host);
+		scsi_host_put(ap->scsi_host);
 	}
 
-	if (host_set->ops->host_stop)
-		host_set->ops->host_stop(host_set);
+	if (host->ops->host_stop)
+		host->ops->host_stop(host);
 
-	kfree(host_set);
+	kfree(host);
 }
 
 /**
@@ -5719,9 +5719,9 @@ void ata_host_set_remove(struct ata_host_set *host_set)
  *	One.
  */
 
-int ata_scsi_release(struct Scsi_Host *host)
+int ata_scsi_release(struct Scsi_Host *shost)
 {
-	struct ata_port *ap = ata_shost_to_port(host);
+	struct ata_port *ap = ata_shost_to_port(shost);
 
 	DPRINTK("ENTER\n");
 
@@ -5748,7 +5748,7 @@ ata_probe_ent_alloc(struct device *dev, const struct ata_port_info *port)
 	probe_ent->dev = dev;
 
 	probe_ent->sht = port->sht;
-	probe_ent->host_flags = port->host_flags;
+	probe_ent->port_flags = port->flags;
 	probe_ent->pio_mask = port->pio_mask;
 	probe_ent->mwdma_mask = port->mwdma_mask;
 	probe_ent->udma_mask = port->udma_mask;
@@ -5786,11 +5786,11 @@ void ata_std_ports(struct ata_ioports *ioaddr)
 
 #ifdef CONFIG_PCI
 
-void ata_pci_host_stop (struct ata_host_set *host_set)
+void ata_pci_host_stop (struct ata_host *host)
 {
-	struct pci_dev *pdev = to_pci_dev(host_set->dev);
+	struct pci_dev *pdev = to_pci_dev(host->dev);
 
-	pci_iounmap(pdev, host_set->mmio_base);
+	pci_iounmap(pdev, host->mmio_base);
 }
 
 /**
@@ -5810,9 +5810,9 @@ void ata_pci_host_stop (struct ata_host_set *host_set)
 void ata_pci_remove_one (struct pci_dev *pdev)
 {
 	struct device *dev = pci_dev_to_dev(pdev);
-	struct ata_host_set *host_set = dev_get_drvdata(dev);
+	struct ata_host *host = dev_get_drvdata(dev);
 
-	ata_host_set_remove(host_set);
+	ata_host_remove(host);
 
 	pci_release_regions(pdev);
 	pci_disable_device(pdev);
@@ -5873,10 +5873,10 @@ void ata_pci_device_do_resume(struct pci_dev *pdev)
 
 int ata_pci_device_suspend(struct pci_dev *pdev, pm_message_t mesg)
 {
-	struct ata_host_set *host_set = dev_get_drvdata(&pdev->dev);
+	struct ata_host *host = dev_get_drvdata(&pdev->dev);
 	int rc = 0;
 
-	rc = ata_host_set_suspend(host_set, mesg);
+	rc = ata_host_suspend(host, mesg);
 	if (rc)
 		return rc;
 
@@ -5887,10 +5887,10 @@ int ata_pci_device_suspend(struct pci_dev *pdev, pm_message_t mesg)
 
 int ata_pci_device_resume(struct pci_dev *pdev)
 {
-	struct ata_host_set *host_set = dev_get_drvdata(&pdev->dev);
+	struct ata_host *host = dev_get_drvdata(&pdev->dev);
 
 	ata_pci_device_do_resume(pdev);
-	ata_host_set_resume(host_set);
+	ata_host_resume(host);
 	return 0;
 }
 #endif /* CONFIG_PCI */
@@ -6035,10 +6035,10 @@ EXPORT_SYMBOL_GPL(sata_deb_timing_long);
 EXPORT_SYMBOL_GPL(ata_dummy_port_ops);
 EXPORT_SYMBOL_GPL(ata_std_bios_param);
 EXPORT_SYMBOL_GPL(ata_std_ports);
-EXPORT_SYMBOL_GPL(ata_host_set_init);
+EXPORT_SYMBOL_GPL(ata_host_init);
 EXPORT_SYMBOL_GPL(ata_device_add);
 EXPORT_SYMBOL_GPL(ata_port_detach);
-EXPORT_SYMBOL_GPL(ata_host_set_remove);
+EXPORT_SYMBOL_GPL(ata_host_remove);
 EXPORT_SYMBOL_GPL(ata_sg_init);
 EXPORT_SYMBOL_GPL(ata_sg_init_one);
 EXPORT_SYMBOL_GPL(ata_hsm_move);
@@ -6105,8 +6105,8 @@ EXPORT_SYMBOL_GPL(sata_scr_write);
 EXPORT_SYMBOL_GPL(sata_scr_write_flush);
 EXPORT_SYMBOL_GPL(ata_port_online);
 EXPORT_SYMBOL_GPL(ata_port_offline);
-EXPORT_SYMBOL_GPL(ata_host_set_suspend);
-EXPORT_SYMBOL_GPL(ata_host_set_resume);
+EXPORT_SYMBOL_GPL(ata_host_suspend);
+EXPORT_SYMBOL_GPL(ata_host_resume);
 EXPORT_SYMBOL_GPL(ata_id_string);
 EXPORT_SYMBOL_GPL(ata_id_c_string);
 EXPORT_SYMBOL_GPL(ata_scsi_simulate);
diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c
index 2c476eee463f..b1b510493c2d 100644
--- a/drivers/ata/libata-eh.c
+++ b/drivers/ata/libata-eh.c
@@ -200,7 +200,7 @@ void ata_scsi_error(struct Scsi_Host *host)
 	/* synchronize with port task */
 	ata_port_flush_task(ap);
 
-	/* synchronize with host_set lock and sort out timeouts */
+	/* synchronize with host lock and sort out timeouts */
 
 	/* For new EH, all qcs are finished in one of three ways -
 	 * normal completion, error completion, and SCSI timeout.
@@ -377,7 +377,7 @@ void ata_port_wait_eh(struct ata_port *ap)
 	spin_unlock_irqrestore(ap->lock, flags);
 
 	/* make sure SCSI EH is complete */
-	if (scsi_host_in_recovery(ap->host)) {
+	if (scsi_host_in_recovery(ap->scsi_host)) {
 		msleep(10);
 		goto retry;
 	}
@@ -486,7 +486,7 @@ void ata_eng_timeout(struct ata_port *ap)
  *	other commands are drained.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 void ata_qc_schedule_eh(struct ata_queued_cmd *qc)
 {
@@ -513,14 +513,14 @@ void ata_qc_schedule_eh(struct ata_queued_cmd *qc)
  *	all commands are drained.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 void ata_port_schedule_eh(struct ata_port *ap)
 {
 	WARN_ON(!ap->ops->error_handler);
 
 	ap->pflags |= ATA_PFLAG_EH_PENDING;
-	scsi_schedule_eh(ap->host);
+	scsi_schedule_eh(ap->scsi_host);
 
 	DPRINTK("port EH scheduled\n");
 }
@@ -532,7 +532,7 @@ void ata_port_schedule_eh(struct ata_port *ap)
  *	Abort all active qc's of @ap and schedule EH.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	Number of aborted qc's.
@@ -575,7 +575,7 @@ int ata_port_abort(struct ata_port *ap)
  *	is frozen.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 static void __ata_port_freeze(struct ata_port *ap)
 {
@@ -596,7 +596,7 @@ static void __ata_port_freeze(struct ata_port *ap)
  *	Abort and freeze @ap.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	Number of aborted commands.
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index d168e3413661..3986ec8741b4 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -321,7 +321,7 @@ int ata_scsi_ioctl(struct scsi_device *scsidev, int cmd, void __user *arg)
  *	current command.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	Command allocated, or %NULL if none available.
@@ -537,7 +537,7 @@ int ata_scsi_device_resume(struct scsi_device *sdev)
  *	format sense blocks.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 void ata_to_sense_error(unsigned id, u8 drv_stat, u8 drv_err, u8 *sk, u8 *asc,
 			u8 *ascq, int verbose)
@@ -649,7 +649,7 @@ void ata_to_sense_error(unsigned id, u8 drv_stat, u8 drv_err, u8 *sk, u8 *asc,
  *	block. Clear sense key, ASC & ASCQ if there is no error.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 void ata_gen_ata_desc_sense(struct ata_queued_cmd *qc)
 {
@@ -918,7 +918,7 @@ int ata_scsi_change_queue_depth(struct scsi_device *sdev, int queue_depth)
  *	[See SAT revision 5 at www.t10.org]
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	Zero on success, non-zero on error.
@@ -986,7 +986,7 @@ invalid_fld:
  *	FLUSH CACHE EXT.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	Zero on success, non-zero on error.
@@ -1109,7 +1109,7 @@ static void scsi_16_lba_len(const u8 *scsicmd, u64 *plba, u32 *plen)
  *	Converts SCSI VERIFY command to an ATA READ VERIFY command.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	Zero on success, non-zero on error.
@@ -1233,7 +1233,7 @@ nothing_to_do:
  *	%WRITE_16 are currently supported.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	Zero on success, non-zero on error.
@@ -1467,7 +1467,7 @@ static void ata_scsi_qc_complete(struct ata_queued_cmd *qc)
  *	issued to @dev.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	1 if deferring is needed, 0 otherwise.
@@ -1510,7 +1510,7 @@ static int ata_scmd_need_defer(struct ata_device *dev, int is_io)
  *	termination.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	0 on success, SCSI_ML_QUEUE_DEVICE_BUSY if the command
@@ -1589,7 +1589,7 @@ defer:
  *	Maps buffer contained within SCSI command @cmd.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	Length of response buffer.
@@ -1623,7 +1623,7 @@ static unsigned int ata_scsi_rbuf_get(struct scsi_cmnd *cmd, u8 **buf_out)
  *	Unmaps response buffer contained within @cmd.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 static inline void ata_scsi_rbuf_put(struct scsi_cmnd *cmd, u8 *buf)
@@ -1649,7 +1649,7 @@ static inline void ata_scsi_rbuf_put(struct scsi_cmnd *cmd, u8 *buf)
  *	and sense buffer are assumed to be set).
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 void ata_scsi_rbuf_fill(struct ata_scsi_args *args,
@@ -1680,7 +1680,7 @@ void ata_scsi_rbuf_fill(struct ata_scsi_args *args,
  *	with non-VPD INQUIRY command output.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 unsigned int ata_scsiop_inq_std(struct ata_scsi_args *args, u8 *rbuf,
@@ -1736,7 +1736,7 @@ unsigned int ata_scsiop_inq_std(struct ata_scsi_args *args, u8 *rbuf,
  *	Returns list of inquiry VPD pages available.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 unsigned int ata_scsiop_inq_00(struct ata_scsi_args *args, u8 *rbuf,
@@ -1764,7 +1764,7 @@ unsigned int ata_scsiop_inq_00(struct ata_scsi_args *args, u8 *rbuf,
  *	Returns ATA device serial number.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 unsigned int ata_scsiop_inq_80(struct ata_scsi_args *args, u8 *rbuf,
@@ -1797,7 +1797,7 @@ unsigned int ata_scsiop_inq_80(struct ata_scsi_args *args, u8 *rbuf,
  *	   name ("ATA     "), model and serial numbers.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 unsigned int ata_scsiop_inq_83(struct ata_scsi_args *args, u8 *rbuf,
@@ -1849,7 +1849,7 @@ unsigned int ata_scsiop_inq_83(struct ata_scsi_args *args, u8 *rbuf,
  *	that the caller should successfully complete this SCSI command.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 unsigned int ata_scsiop_noop(struct ata_scsi_args *args, u8 *rbuf,
@@ -1990,7 +1990,7 @@ static int ata_dev_supports_fua(u16 *id)
  *	descriptor for other device types.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 unsigned int ata_scsiop_mode_sense(struct ata_scsi_args *args, u8 *rbuf,
@@ -2129,7 +2129,7 @@ saving_not_supp:
  *	Simulate READ CAPACITY commands.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 unsigned int ata_scsiop_read_cap(struct ata_scsi_args *args, u8 *rbuf,
@@ -2204,7 +2204,7 @@ unsigned int ata_scsiop_read_cap(struct ata_scsi_args *args, u8 *rbuf,
  *	Simulate REPORT LUNS command.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 unsigned int ata_scsiop_report_luns(struct ata_scsi_args *args, u8 *rbuf,
@@ -2256,7 +2256,7 @@ void ata_scsi_set_sense(struct scsi_cmnd *cmd, u8 sk, u8 asc, u8 ascq)
  *	and the specified additional sense codes.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 void ata_scsi_badcmd(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *), u8 asc, u8 ascq)
@@ -2421,7 +2421,7 @@ static void atapi_qc_complete(struct ata_queued_cmd *qc)
  *	@scsicmd: SCSI CDB associated with this PACKET command
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	Zero on success, non-zero on failure.
@@ -2500,7 +2500,7 @@ static struct ata_device * __ata_scsi_find_dev(struct ata_port *ap,
  *	Determine if commands should be sent to the specified device.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	0 if commands are not allowed / 1 if commands are allowed
@@ -2534,7 +2534,7 @@ static int ata_scsi_dev_enabled(struct ata_device *dev)
  *	SCSI command to be sent.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	Associated ATA device, or %NULL if not found.
@@ -2808,7 +2808,7 @@ static inline int __ata_scsi_queuecmd(struct scsi_cmnd *cmd,
  *	ATA and ATAPI devices appearing as SCSI devices.
  *
  *	LOCKING:
- *	Releases scsi-layer-held lock, and obtains host_set lock.
+ *	Releases scsi-layer-held lock, and obtains host lock.
  *
  *	RETURNS:
  *	Return value from __ata_scsi_queuecmd() if @cmd can be queued,
@@ -2852,7 +2852,7 @@ int ata_scsi_queuecmd(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
  *	that can be handled internally.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 void ata_scsi_simulate(struct ata_device *dev, struct scsi_cmnd *cmd,
@@ -2944,7 +2944,7 @@ void ata_scsi_scan_host(struct ata_port *ap)
 		if (!ata_dev_enabled(dev) || dev->sdev)
 			continue;
 
-		sdev = __scsi_add_device(ap->host, 0, i, 0, NULL);
+		sdev = __scsi_add_device(ap->scsi_host, 0, i, 0, NULL);
 		if (!IS_ERR(sdev)) {
 			dev->sdev = sdev;
 			scsi_device_put(sdev);
@@ -2958,11 +2958,11 @@ void ata_scsi_scan_host(struct ata_port *ap)
  *
  *	This function is called from ata_eh_hotplug() and responsible
  *	for taking the SCSI device attached to @dev offline.  This
- *	function is called with host_set lock which protects dev->sdev
+ *	function is called with host lock which protects dev->sdev
  *	against clearing.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  *
  *	RETURNS:
  *	1 if attached SCSI device exists, 0 otherwise.
@@ -2998,16 +2998,16 @@ static void ata_scsi_remove_dev(struct ata_device *dev)
 	 * be removed if there is __scsi_device_get() interface which
 	 * increments reference counts regardless of device state.
 	 */
-	mutex_lock(&ap->host->scan_mutex);
+	mutex_lock(&ap->scsi_host->scan_mutex);
 	spin_lock_irqsave(ap->lock, flags);
 
-	/* clearing dev->sdev is protected by host_set lock */
+	/* clearing dev->sdev is protected by host lock */
 	sdev = dev->sdev;
 	dev->sdev = NULL;
 
 	if (sdev) {
 		/* If user initiated unplug races with us, sdev can go
-		 * away underneath us after the host_set lock and
+		 * away underneath us after the host lock and
 		 * scan_mutex are released.  Hold onto it.
 		 */
 		if (scsi_device_get(sdev) == 0) {
@@ -3024,7 +3024,7 @@ static void ata_scsi_remove_dev(struct ata_device *dev)
 	}
 
 	spin_unlock_irqrestore(ap->lock, flags);
-	mutex_unlock(&ap->host->scan_mutex);
+	mutex_unlock(&ap->scsi_host->scan_mutex);
 
 	if (sdev) {
 		ata_dev_printk(dev, KERN_INFO, "detaching (SCSI %s)\n",
@@ -3176,7 +3176,7 @@ void ata_scsi_dev_rescan(void *data)
  *	ata_sas_port_alloc - Allocate port for a SAS attached SATA device
  *	@pdev: PCI device that the scsi device is attached to
  *	@port_info: Information from low-level host driver
- *	@host: SCSI host that the scsi device is attached to
+ *	@shost: SCSI host that the scsi device is attached to
  *
  *	LOCKING:
  *	PCI/etc. bus probe sem.
@@ -3185,9 +3185,9 @@ void ata_scsi_dev_rescan(void *data)
  *	ata_port pointer on success / NULL on failure.
  */
 
-struct ata_port *ata_sas_port_alloc(struct ata_host_set *host_set,
+struct ata_port *ata_sas_port_alloc(struct ata_host *host,
 				    struct ata_port_info *port_info,
-				    struct Scsi_Host *host)
+				    struct Scsi_Host *shost)
 {
 	struct ata_port *ap = kzalloc(sizeof(*ap), GFP_KERNEL);
 	struct ata_probe_ent *ent;
@@ -3195,14 +3195,14 @@ struct ata_port *ata_sas_port_alloc(struct ata_host_set *host_set,
 	if (!ap)
 		return NULL;
 
-	ent = ata_probe_ent_alloc(host_set->dev, port_info);
+	ent = ata_probe_ent_alloc(host->dev, port_info);
 	if (!ent) {
 		kfree(ap);
 		return NULL;
 	}
 
-	ata_port_init(ap, host_set, ent, 0);
-	ap->lock = host->host_lock;
+	ata_port_init(ap, host, ent, 0);
+	ap->lock = shost->host_lock;
 	kfree(ent);
 	return ap;
 }
diff --git a/drivers/ata/libata.h b/drivers/ata/libata.h
index d4a4f82360ec..a5ecb71390a9 100644
--- a/drivers/ata/libata.h
+++ b/drivers/ata/libata.h
@@ -69,7 +69,7 @@ extern int ata_flush_cache(struct ata_device *dev);
 extern void ata_dev_init(struct ata_device *dev);
 extern int ata_task_ioctl(struct scsi_device *scsidev, void __user *arg);
 extern int ata_cmd_ioctl(struct scsi_device *scsidev, void __user *arg);
-extern void ata_port_init(struct ata_port *ap, struct ata_host_set *host_set,
+extern void ata_port_init(struct ata_port *ap, struct ata_host *host,
 			  const struct ata_probe_ent *ent, unsigned int port_no);
 extern struct ata_probe_ent *ata_probe_ent_alloc(struct device *dev,
 						 const struct ata_port_info *port);
diff --git a/drivers/ata/pdc_adma.c b/drivers/ata/pdc_adma.c
index 61d2aa697b4d..912211ada816 100644
--- a/drivers/ata/pdc_adma.c
+++ b/drivers/ata/pdc_adma.c
@@ -127,7 +127,7 @@ static int adma_ata_init_one (struct pci_dev *pdev,
 static irqreturn_t adma_intr (int irq, void *dev_instance,
 				struct pt_regs *regs);
 static int adma_port_start(struct ata_port *ap);
-static void adma_host_stop(struct ata_host_set *host_set);
+static void adma_host_stop(struct ata_host *host);
 static void adma_port_stop(struct ata_port *ap);
 static void adma_phy_reset(struct ata_port *ap);
 static void adma_qc_prep(struct ata_queued_cmd *qc);
@@ -182,7 +182,7 @@ static struct ata_port_info adma_port_info[] = {
 	/* board_1841_idx */
 	{
 		.sht		= &adma_ata_sht,
-		.host_flags	= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST |
+		.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST |
 				  ATA_FLAG_NO_LEGACY | ATA_FLAG_MMIO |
 				  ATA_FLAG_PIO_POLLING,
 		.pio_mask	= 0x10, /* pio4 */
@@ -237,7 +237,7 @@ static void adma_reset_engine(void __iomem *chan)
 static void adma_reinit_engine(struct ata_port *ap)
 {
 	struct adma_port_priv *pp = ap->private_data;
-	void __iomem *mmio_base = ap->host_set->mmio_base;
+	void __iomem *mmio_base = ap->host->mmio_base;
 	void __iomem *chan = ADMA_REGS(mmio_base, ap->port_no);
 
 	/* mask/clear ATA interrupts */
@@ -265,7 +265,7 @@ static void adma_reinit_engine(struct ata_port *ap)
 
 static inline void adma_enter_reg_mode(struct ata_port *ap)
 {
-	void __iomem *chan = ADMA_REGS(ap->host_set->mmio_base, ap->port_no);
+	void __iomem *chan = ADMA_REGS(ap->host->mmio_base, ap->port_no);
 
 	writew(aPIOMD4, chan + ADMA_CONTROL);
 	readb(chan + ADMA_STATUS);	/* flush */
@@ -412,7 +412,7 @@ static void adma_qc_prep(struct ata_queued_cmd *qc)
 static inline void adma_packet_start(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
-	void __iomem *chan = ADMA_REGS(ap->host_set->mmio_base, ap->port_no);
+	void __iomem *chan = ADMA_REGS(ap->host->mmio_base, ap->port_no);
 
 	VPRINTK("ENTER, ap %p\n", ap);
 
@@ -442,13 +442,13 @@ static unsigned int adma_qc_issue(struct ata_queued_cmd *qc)
 	return ata_qc_issue_prot(qc);
 }
 
-static inline unsigned int adma_intr_pkt(struct ata_host_set *host_set)
+static inline unsigned int adma_intr_pkt(struct ata_host *host)
 {
 	unsigned int handled = 0, port_no;
-	u8 __iomem *mmio_base = host_set->mmio_base;
+	u8 __iomem *mmio_base = host->mmio_base;
 
-	for (port_no = 0; port_no < host_set->n_ports; ++port_no) {
-		struct ata_port *ap = host_set->ports[port_no];
+	for (port_no = 0; port_no < host->n_ports; ++port_no) {
+		struct ata_port *ap = host->ports[port_no];
 		struct adma_port_priv *pp;
 		struct ata_queued_cmd *qc;
 		void __iomem *chan = ADMA_REGS(mmio_base, port_no);
@@ -476,13 +476,13 @@ static inline unsigned int adma_intr_pkt(struct ata_host_set *host_set)
 	return handled;
 }
 
-static inline unsigned int adma_intr_mmio(struct ata_host_set *host_set)
+static inline unsigned int adma_intr_mmio(struct ata_host *host)
 {
 	unsigned int handled = 0, port_no;
 
-	for (port_no = 0; port_no < host_set->n_ports; ++port_no) {
+	for (port_no = 0; port_no < host->n_ports; ++port_no) {
 		struct ata_port *ap;
-		ap = host_set->ports[port_no];
+		ap = host->ports[port_no];
 		if (ap && (!(ap->flags & ATA_FLAG_DISABLED))) {
 			struct ata_queued_cmd *qc;
 			struct adma_port_priv *pp = ap->private_data;
@@ -511,14 +511,14 @@ static inline unsigned int adma_intr_mmio(struct ata_host_set *host_set)
 
 static irqreturn_t adma_intr(int irq, void *dev_instance, struct pt_regs *regs)
 {
-	struct ata_host_set *host_set = dev_instance;
+	struct ata_host *host = dev_instance;
 	unsigned int handled = 0;
 
 	VPRINTK("ENTER\n");
 
-	spin_lock(&host_set->lock);
-	handled  = adma_intr_pkt(host_set) | adma_intr_mmio(host_set);
-	spin_unlock(&host_set->lock);
+	spin_lock(&host->lock);
+	handled  = adma_intr_pkt(host) | adma_intr_mmio(host);
+	spin_unlock(&host->lock);
 
 	VPRINTK("EXIT\n");
 
@@ -544,7 +544,7 @@ static void adma_ata_setup_port(struct ata_ioports *port, unsigned long base)
 
 static int adma_port_start(struct ata_port *ap)
 {
-	struct device *dev = ap->host_set->dev;
+	struct device *dev = ap->host->dev;
 	struct adma_port_priv *pp;
 	int rc;
 
@@ -582,10 +582,10 @@ err_out:
 
 static void adma_port_stop(struct ata_port *ap)
 {
-	struct device *dev = ap->host_set->dev;
+	struct device *dev = ap->host->dev;
 	struct adma_port_priv *pp = ap->private_data;
 
-	adma_reset_engine(ADMA_REGS(ap->host_set->mmio_base, ap->port_no));
+	adma_reset_engine(ADMA_REGS(ap->host->mmio_base, ap->port_no));
 	if (pp != NULL) {
 		ap->private_data = NULL;
 		if (pp->pkt != NULL)
@@ -596,14 +596,14 @@ static void adma_port_stop(struct ata_port *ap)
 	ata_port_stop(ap);
 }
 
-static void adma_host_stop(struct ata_host_set *host_set)
+static void adma_host_stop(struct ata_host *host)
 {
 	unsigned int port_no;
 
 	for (port_no = 0; port_no < ADMA_PORTS; ++port_no)
-		adma_reset_engine(ADMA_REGS(host_set->mmio_base, port_no));
+		adma_reset_engine(ADMA_REGS(host->mmio_base, port_no));
 
-	ata_pci_host_stop(host_set);
+	ata_pci_host_stop(host);
 }
 
 static void adma_host_init(unsigned int chip_id,
@@ -684,7 +684,7 @@ static int adma_ata_init_one(struct pci_dev *pdev,
 	INIT_LIST_HEAD(&probe_ent->node);
 
 	probe_ent->sht		= adma_port_info[board_idx].sht;
-	probe_ent->host_flags	= adma_port_info[board_idx].host_flags;
+	probe_ent->port_flags	= adma_port_info[board_idx].flags;
 	probe_ent->pio_mask	= adma_port_info[board_idx].pio_mask;
 	probe_ent->mwdma_mask	= adma_port_info[board_idx].mwdma_mask;
 	probe_ent->udma_mask	= adma_port_info[board_idx].udma_mask;
diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c
index a2915a56accd..34f1939b44c9 100644
--- a/drivers/ata/sata_mv.c
+++ b/drivers/ata/sata_mv.c
@@ -342,7 +342,7 @@ static u32 mv5_scr_read(struct ata_port *ap, unsigned int sc_reg_in);
 static void mv5_scr_write(struct ata_port *ap, unsigned int sc_reg_in, u32 val);
 static void mv_phy_reset(struct ata_port *ap);
 static void __mv_phy_reset(struct ata_port *ap, int can_sleep);
-static void mv_host_stop(struct ata_host_set *host_set);
+static void mv_host_stop(struct ata_host *host);
 static int mv_port_start(struct ata_port *ap);
 static void mv_port_stop(struct ata_port *ap);
 static void mv_qc_prep(struct ata_queued_cmd *qc);
@@ -480,35 +480,35 @@ static const struct ata_port_operations mv_iie_ops = {
 static const struct ata_port_info mv_port_info[] = {
 	{  /* chip_504x */
 		.sht		= &mv_sht,
-		.host_flags	= MV_COMMON_FLAGS,
+		.flags		= MV_COMMON_FLAGS,
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.udma_mask	= 0x7f,	/* udma0-6 */
 		.port_ops	= &mv5_ops,
 	},
 	{  /* chip_508x */
 		.sht		= &mv_sht,
-		.host_flags	= (MV_COMMON_FLAGS | MV_FLAG_DUAL_HC),
+		.flags		= (MV_COMMON_FLAGS | MV_FLAG_DUAL_HC),
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.udma_mask	= 0x7f,	/* udma0-6 */
 		.port_ops	= &mv5_ops,
 	},
 	{  /* chip_5080 */
 		.sht		= &mv_sht,
-		.host_flags	= (MV_COMMON_FLAGS | MV_FLAG_DUAL_HC),
+		.flags		= (MV_COMMON_FLAGS | MV_FLAG_DUAL_HC),
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.udma_mask	= 0x7f,	/* udma0-6 */
 		.port_ops	= &mv5_ops,
 	},
 	{  /* chip_604x */
 		.sht		= &mv_sht,
-		.host_flags	= (MV_COMMON_FLAGS | MV_6XXX_FLAGS),
+		.flags		= (MV_COMMON_FLAGS | MV_6XXX_FLAGS),
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.udma_mask	= 0x7f,	/* udma0-6 */
 		.port_ops	= &mv6_ops,
 	},
 	{  /* chip_608x */
 		.sht		= &mv_sht,
-		.host_flags	= (MV_COMMON_FLAGS | MV_6XXX_FLAGS |
+		.flags		= (MV_COMMON_FLAGS | MV_6XXX_FLAGS |
 				   MV_FLAG_DUAL_HC),
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.udma_mask	= 0x7f,	/* udma0-6 */
@@ -516,14 +516,14 @@ static const struct ata_port_info mv_port_info[] = {
 	},
 	{  /* chip_6042 */
 		.sht		= &mv_sht,
-		.host_flags	= (MV_COMMON_FLAGS | MV_6XXX_FLAGS),
+		.flags		= (MV_COMMON_FLAGS | MV_6XXX_FLAGS),
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.udma_mask	= 0x7f,	/* udma0-6 */
 		.port_ops	= &mv_iie_ops,
 	},
 	{  /* chip_7042 */
 		.sht		= &mv_sht,
-		.host_flags	= (MV_COMMON_FLAGS | MV_6XXX_FLAGS |
+		.flags		= (MV_COMMON_FLAGS | MV_6XXX_FLAGS |
 				   MV_FLAG_DUAL_HC),
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.udma_mask	= 0x7f,	/* udma0-6 */
@@ -618,12 +618,12 @@ static inline void __iomem *mv_port_base(void __iomem *base, unsigned int port)
 
 static inline void __iomem *mv_ap_base(struct ata_port *ap)
 {
-	return mv_port_base(ap->host_set->mmio_base, ap->port_no);
+	return mv_port_base(ap->host->mmio_base, ap->port_no);
 }
 
-static inline int mv_get_hc_count(unsigned long host_flags)
+static inline int mv_get_hc_count(unsigned long port_flags)
 {
-	return ((host_flags & MV_FLAG_DUAL_HC) ? 2 : 1);
+	return ((port_flags & MV_FLAG_DUAL_HC) ? 2 : 1);
 }
 
 static void mv_irq_clear(struct ata_port *ap)
@@ -809,7 +809,7 @@ static void mv_scr_write(struct ata_port *ap, unsigned int sc_reg_in, u32 val)
 
 /**
  *      mv_host_stop - Host specific cleanup/stop routine.
- *      @host_set: host data structure
+ *      @host: host data structure
  *
  *      Disable ints, cleanup host memory, call general purpose
  *      host_stop.
@@ -817,10 +817,10 @@ static void mv_scr_write(struct ata_port *ap, unsigned int sc_reg_in, u32 val)
  *      LOCKING:
  *      Inherited from caller.
  */
-static void mv_host_stop(struct ata_host_set *host_set)
+static void mv_host_stop(struct ata_host *host)
 {
-	struct mv_host_priv *hpriv = host_set->private_data;
-	struct pci_dev *pdev = to_pci_dev(host_set->dev);
+	struct mv_host_priv *hpriv = host->private_data;
+	struct pci_dev *pdev = to_pci_dev(host->dev);
 
 	if (hpriv->hp_flags & MV_HP_FLAG_MSI) {
 		pci_disable_msi(pdev);
@@ -828,7 +828,7 @@ static void mv_host_stop(struct ata_host_set *host_set)
 		pci_intx(pdev, 0);
 	}
 	kfree(hpriv);
-	ata_host_stop(host_set);
+	ata_host_stop(host);
 }
 
 static inline void mv_priv_free(struct mv_port_priv *pp, struct device *dev)
@@ -875,8 +875,8 @@ static void mv_edma_cfg(struct mv_host_priv *hpriv, void __iomem *port_mmio)
  */
 static int mv_port_start(struct ata_port *ap)
 {
-	struct device *dev = ap->host_set->dev;
-	struct mv_host_priv *hpriv = ap->host_set->private_data;
+	struct device *dev = ap->host->dev;
+	struct mv_host_priv *hpriv = ap->host->private_data;
 	struct mv_port_priv *pp;
 	void __iomem *port_mmio = mv_ap_base(ap);
 	void *mem;
@@ -965,17 +965,17 @@ err_out:
  *      Stop DMA, cleanup port memory.
  *
  *      LOCKING:
- *      This routine uses the host_set lock to protect the DMA stop.
+ *      This routine uses the host lock to protect the DMA stop.
  */
 static void mv_port_stop(struct ata_port *ap)
 {
-	struct device *dev = ap->host_set->dev;
+	struct device *dev = ap->host->dev;
 	struct mv_port_priv *pp = ap->private_data;
 	unsigned long flags;
 
-	spin_lock_irqsave(&ap->host_set->lock, flags);
+	spin_lock_irqsave(&ap->host->lock, flags);
 	mv_stop_dma(ap);
-	spin_unlock_irqrestore(&ap->host_set->lock, flags);
+	spin_unlock_irqrestore(&ap->host->lock, flags);
 
 	ap->private_data = NULL;
 	ata_pad_free(ap, dev);
@@ -1330,7 +1330,7 @@ static void mv_err_intr(struct ata_port *ap, int reset_allowed)
 
 /**
  *      mv_host_intr - Handle all interrupts on the given host controller
- *      @host_set: host specific structure
+ *      @host: host specific structure
  *      @relevant: port error bits relevant to this host controller
  *      @hc: which host controller we're to look at
  *
@@ -1344,10 +1344,9 @@ static void mv_err_intr(struct ata_port *ap, int reset_allowed)
  *      LOCKING:
  *      Inherited from caller.
  */
-static void mv_host_intr(struct ata_host_set *host_set, u32 relevant,
-			 unsigned int hc)
+static void mv_host_intr(struct ata_host *host, u32 relevant, unsigned int hc)
 {
-	void __iomem *mmio = host_set->mmio_base;
+	void __iomem *mmio = host->mmio_base;
 	void __iomem *hc_mmio = mv_hc_base(mmio, hc);
 	struct ata_queued_cmd *qc;
 	u32 hc_irq_cause;
@@ -1371,7 +1370,7 @@ static void mv_host_intr(struct ata_host_set *host_set, u32 relevant,
 
 	for (port = port0; port < port0 + MV_PORTS_PER_HC; port++) {
 		u8 ata_status = 0;
-		struct ata_port *ap = host_set->ports[port];
+		struct ata_port *ap = host->ports[port];
 		struct mv_port_priv *pp = ap->private_data;
 
 		hard_port = mv_hardport_from_port(port); /* range 0..3 */
@@ -1444,15 +1443,15 @@ static void mv_host_intr(struct ata_host_set *host_set, u32 relevant,
  *      reported here.
  *
  *      LOCKING:
- *      This routine holds the host_set lock while processing pending
+ *      This routine holds the host lock while processing pending
  *      interrupts.
  */
 static irqreturn_t mv_interrupt(int irq, void *dev_instance,
 				struct pt_regs *regs)
 {
-	struct ata_host_set *host_set = dev_instance;
+	struct ata_host *host = dev_instance;
 	unsigned int hc, handled = 0, n_hcs;
-	void __iomem *mmio = host_set->mmio_base;
+	void __iomem *mmio = host->mmio_base;
 	struct mv_host_priv *hpriv;
 	u32 irq_stat;
 
@@ -1465,18 +1464,18 @@ static irqreturn_t mv_interrupt(int irq, void *dev_instance,
 		return IRQ_NONE;
 	}
 
-	n_hcs = mv_get_hc_count(host_set->ports[0]->flags);
-	spin_lock(&host_set->lock);
+	n_hcs = mv_get_hc_count(host->ports[0]->flags);
+	spin_lock(&host->lock);
 
 	for (hc = 0; hc < n_hcs; hc++) {
 		u32 relevant = irq_stat & (HC0_IRQ_PEND << (hc * HC_SHIFT));
 		if (relevant) {
-			mv_host_intr(host_set, relevant, hc);
+			mv_host_intr(host, relevant, hc);
 			handled++;
 		}
 	}
 
-	hpriv = host_set->private_data;
+	hpriv = host->private_data;
 	if (IS_60XX(hpriv)) {
 		/* deal with the interrupt coalescing bits */
 		if (irq_stat & (TRAN_LO_DONE | TRAN_HI_DONE | PORTS_0_7_COAL_DONE)) {
@@ -1491,12 +1490,12 @@ static irqreturn_t mv_interrupt(int irq, void *dev_instance,
 		       readl(mmio + PCI_IRQ_CAUSE_OFS));
 
 		DPRINTK("All regs @ PCI error\n");
-		mv_dump_all_regs(mmio, -1, to_pci_dev(host_set->dev));
+		mv_dump_all_regs(mmio, -1, to_pci_dev(host->dev));
 
 		writelfl(0, mmio + PCI_IRQ_CAUSE_OFS);
 		handled++;
 	}
-	spin_unlock(&host_set->lock);
+	spin_unlock(&host->lock);
 
 	return IRQ_RETVAL(handled);
 }
@@ -1528,7 +1527,7 @@ static unsigned int mv5_scr_offset(unsigned int sc_reg_in)
 
 static u32 mv5_scr_read(struct ata_port *ap, unsigned int sc_reg_in)
 {
-	void __iomem *mmio = mv5_phy_base(ap->host_set->mmio_base, ap->port_no);
+	void __iomem *mmio = mv5_phy_base(ap->host->mmio_base, ap->port_no);
 	unsigned int ofs = mv5_scr_offset(sc_reg_in);
 
 	if (ofs != 0xffffffffU)
@@ -1539,7 +1538,7 @@ static u32 mv5_scr_read(struct ata_port *ap, unsigned int sc_reg_in)
 
 static void mv5_scr_write(struct ata_port *ap, unsigned int sc_reg_in, u32 val)
 {
-	void __iomem *mmio = mv5_phy_base(ap->host_set->mmio_base, ap->port_no);
+	void __iomem *mmio = mv5_phy_base(ap->host->mmio_base, ap->port_no);
 	unsigned int ofs = mv5_scr_offset(sc_reg_in);
 
 	if (ofs != 0xffffffffU)
@@ -1904,8 +1903,8 @@ static void mv_channel_reset(struct mv_host_priv *hpriv, void __iomem *mmio,
 
 static void mv_stop_and_reset(struct ata_port *ap)
 {
-	struct mv_host_priv *hpriv = ap->host_set->private_data;
-	void __iomem *mmio = ap->host_set->mmio_base;
+	struct mv_host_priv *hpriv = ap->host->private_data;
+	void __iomem *mmio = ap->host->mmio_base;
 
 	mv_stop_dma(ap);
 
@@ -1936,7 +1935,7 @@ static inline void __msleep(unsigned int msec, int can_sleep)
 static void __mv_phy_reset(struct ata_port *ap, int can_sleep)
 {
 	struct mv_port_priv *pp	= ap->private_data;
-	struct mv_host_priv *hpriv = ap->host_set->private_data;
+	struct mv_host_priv *hpriv = ap->host->private_data;
 	void __iomem *port_mmio = mv_ap_base(ap);
 	struct ata_taskfile tf;
 	struct ata_device *dev = &ap->device[0];
@@ -2035,7 +2034,7 @@ static void mv_phy_reset(struct ata_port *ap)
  *      chip/bus, fail the command, and move on.
  *
  *      LOCKING:
- *      This routine holds the host_set lock while failing the command.
+ *      This routine holds the host lock while failing the command.
  */
 static void mv_eng_timeout(struct ata_port *ap)
 {
@@ -2044,18 +2043,17 @@ static void mv_eng_timeout(struct ata_port *ap)
 
 	ata_port_printk(ap, KERN_ERR, "Entering mv_eng_timeout\n");
 	DPRINTK("All regs @ start of eng_timeout\n");
-	mv_dump_all_regs(ap->host_set->mmio_base, ap->port_no,
-			 to_pci_dev(ap->host_set->dev));
+	mv_dump_all_regs(ap->host->mmio_base, ap->port_no,
+			 to_pci_dev(ap->host->dev));
 
 	qc = ata_qc_from_tag(ap, ap->active_tag);
         printk(KERN_ERR "mmio_base %p ap %p qc %p scsi_cmnd %p &cmnd %p\n",
-	       ap->host_set->mmio_base, ap, qc, qc->scsicmd,
-	       &qc->scsicmd->cmnd);
+	       ap->host->mmio_base, ap, qc, qc->scsicmd, &qc->scsicmd->cmnd);
 
-	spin_lock_irqsave(&ap->host_set->lock, flags);
+	spin_lock_irqsave(&ap->host->lock, flags);
 	mv_err_intr(ap, 0);
 	mv_stop_and_reset(ap);
-	spin_unlock_irqrestore(&ap->host_set->lock, flags);
+	spin_unlock_irqrestore(&ap->host->lock, flags);
 
 	WARN_ON(!(qc->flags & ATA_QCFLAG_ACTIVE));
 	if (qc->flags & ATA_QCFLAG_ACTIVE) {
@@ -2236,7 +2234,7 @@ static int mv_init_host(struct pci_dev *pdev, struct ata_probe_ent *probe_ent,
 	if (rc)
 		goto done;
 
-	n_hc = mv_get_hc_count(probe_ent->host_flags);
+	n_hc = mv_get_hc_count(probe_ent->port_flags);
 	probe_ent->n_ports = MV_PORTS_PER_HC * n_hc;
 
 	for (port = 0; port < probe_ent->n_ports; port++)
@@ -2389,7 +2387,7 @@ static int mv_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	memset(hpriv, 0, sizeof(*hpriv));
 
 	probe_ent->sht = mv_port_info[board_idx].sht;
-	probe_ent->host_flags = mv_port_info[board_idx].host_flags;
+	probe_ent->port_flags = mv_port_info[board_idx].flags;
 	probe_ent->pio_mask = mv_port_info[board_idx].pio_mask;
 	probe_ent->udma_mask = mv_port_info[board_idx].udma_mask;
 	probe_ent->port_ops = mv_port_info[board_idx].port_ops;
diff --git a/drivers/ata/sata_nv.c b/drivers/ata/sata_nv.c
index be46df75ab5a..27c22feebf30 100644
--- a/drivers/ata/sata_nv.c
+++ b/drivers/ata/sata_nv.c
@@ -81,7 +81,7 @@ enum {
 };
 
 static int nv_init_one (struct pci_dev *pdev, const struct pci_device_id *ent);
-static void nv_ck804_host_stop(struct ata_host_set *host_set);
+static void nv_ck804_host_stop(struct ata_host *host);
 static irqreturn_t nv_generic_interrupt(int irq, void *dev_instance,
 					struct pt_regs *regs);
 static irqreturn_t nv_nf2_interrupt(int irq, void *dev_instance,
@@ -257,7 +257,7 @@ static struct ata_port_info nv_port_info[] = {
 	/* generic */
 	{
 		.sht		= &nv_sht,
-		.host_flags	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
+		.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
 		.pio_mask	= NV_PIO_MASK,
 		.mwdma_mask	= NV_MWDMA_MASK,
 		.udma_mask	= NV_UDMA_MASK,
@@ -266,7 +266,7 @@ static struct ata_port_info nv_port_info[] = {
 	/* nforce2/3 */
 	{
 		.sht		= &nv_sht,
-		.host_flags	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
+		.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
 		.pio_mask	= NV_PIO_MASK,
 		.mwdma_mask	= NV_MWDMA_MASK,
 		.udma_mask	= NV_UDMA_MASK,
@@ -275,7 +275,7 @@ static struct ata_port_info nv_port_info[] = {
 	/* ck804 */
 	{
 		.sht		= &nv_sht,
-		.host_flags	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
+		.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
 		.pio_mask	= NV_PIO_MASK,
 		.mwdma_mask	= NV_MWDMA_MASK,
 		.udma_mask	= NV_UDMA_MASK,
@@ -292,17 +292,17 @@ MODULE_VERSION(DRV_VERSION);
 static irqreturn_t nv_generic_interrupt(int irq, void *dev_instance,
 					struct pt_regs *regs)
 {
-	struct ata_host_set *host_set = dev_instance;
+	struct ata_host *host = dev_instance;
 	unsigned int i;
 	unsigned int handled = 0;
 	unsigned long flags;
 
-	spin_lock_irqsave(&host_set->lock, flags);
+	spin_lock_irqsave(&host->lock, flags);
 
-	for (i = 0; i < host_set->n_ports; i++) {
+	for (i = 0; i < host->n_ports; i++) {
 		struct ata_port *ap;
 
-		ap = host_set->ports[i];
+		ap = host->ports[i];
 		if (ap &&
 		    !(ap->flags & ATA_FLAG_DISABLED)) {
 			struct ata_queued_cmd *qc;
@@ -318,7 +318,7 @@ static irqreturn_t nv_generic_interrupt(int irq, void *dev_instance,
 
 	}
 
-	spin_unlock_irqrestore(&host_set->lock, flags);
+	spin_unlock_irqrestore(&host->lock, flags);
 
 	return IRQ_RETVAL(handled);
 }
@@ -354,12 +354,12 @@ static int nv_host_intr(struct ata_port *ap, u8 irq_stat)
 	return 1;
 }
 
-static irqreturn_t nv_do_interrupt(struct ata_host_set *host_set, u8 irq_stat)
+static irqreturn_t nv_do_interrupt(struct ata_host *host, u8 irq_stat)
 {
 	int i, handled = 0;
 
-	for (i = 0; i < host_set->n_ports; i++) {
-		struct ata_port *ap = host_set->ports[i];
+	for (i = 0; i < host->n_ports; i++) {
+		struct ata_port *ap = host->ports[i];
 
 		if (ap && !(ap->flags & ATA_FLAG_DISABLED))
 			handled += nv_host_intr(ap, irq_stat);
@@ -373,14 +373,14 @@ static irqreturn_t nv_do_interrupt(struct ata_host_set *host_set, u8 irq_stat)
 static irqreturn_t nv_nf2_interrupt(int irq, void *dev_instance,
 				    struct pt_regs *regs)
 {
-	struct ata_host_set *host_set = dev_instance;
+	struct ata_host *host = dev_instance;
 	u8 irq_stat;
 	irqreturn_t ret;
 
-	spin_lock(&host_set->lock);
-	irq_stat = inb(host_set->ports[0]->ioaddr.scr_addr + NV_INT_STATUS);
-	ret = nv_do_interrupt(host_set, irq_stat);
-	spin_unlock(&host_set->lock);
+	spin_lock(&host->lock);
+	irq_stat = inb(host->ports[0]->ioaddr.scr_addr + NV_INT_STATUS);
+	ret = nv_do_interrupt(host, irq_stat);
+	spin_unlock(&host->lock);
 
 	return ret;
 }
@@ -388,14 +388,14 @@ static irqreturn_t nv_nf2_interrupt(int irq, void *dev_instance,
 static irqreturn_t nv_ck804_interrupt(int irq, void *dev_instance,
 				      struct pt_regs *regs)
 {
-	struct ata_host_set *host_set = dev_instance;
+	struct ata_host *host = dev_instance;
 	u8 irq_stat;
 	irqreturn_t ret;
 
-	spin_lock(&host_set->lock);
-	irq_stat = readb(host_set->mmio_base + NV_INT_STATUS_CK804);
-	ret = nv_do_interrupt(host_set, irq_stat);
-	spin_unlock(&host_set->lock);
+	spin_lock(&host->lock);
+	irq_stat = readb(host->mmio_base + NV_INT_STATUS_CK804);
+	ret = nv_do_interrupt(host, irq_stat);
+	spin_unlock(&host->lock);
 
 	return ret;
 }
@@ -418,7 +418,7 @@ static void nv_scr_write (struct ata_port *ap, unsigned int sc_reg, u32 val)
 
 static void nv_nf2_freeze(struct ata_port *ap)
 {
-	unsigned long scr_addr = ap->host_set->ports[0]->ioaddr.scr_addr;
+	unsigned long scr_addr = ap->host->ports[0]->ioaddr.scr_addr;
 	int shift = ap->port_no * NV_INT_PORT_SHIFT;
 	u8 mask;
 
@@ -429,7 +429,7 @@ static void nv_nf2_freeze(struct ata_port *ap)
 
 static void nv_nf2_thaw(struct ata_port *ap)
 {
-	unsigned long scr_addr = ap->host_set->ports[0]->ioaddr.scr_addr;
+	unsigned long scr_addr = ap->host->ports[0]->ioaddr.scr_addr;
 	int shift = ap->port_no * NV_INT_PORT_SHIFT;
 	u8 mask;
 
@@ -442,7 +442,7 @@ static void nv_nf2_thaw(struct ata_port *ap)
 
 static void nv_ck804_freeze(struct ata_port *ap)
 {
-	void __iomem *mmio_base = ap->host_set->mmio_base;
+	void __iomem *mmio_base = ap->host->mmio_base;
 	int shift = ap->port_no * NV_INT_PORT_SHIFT;
 	u8 mask;
 
@@ -453,7 +453,7 @@ static void nv_ck804_freeze(struct ata_port *ap)
 
 static void nv_ck804_thaw(struct ata_port *ap)
 {
-	void __iomem *mmio_base = ap->host_set->mmio_base;
+	void __iomem *mmio_base = ap->host->mmio_base;
 	int shift = ap->port_no * NV_INT_PORT_SHIFT;
 	u8 mask;
 
@@ -568,9 +568,9 @@ err_out:
 	return rc;
 }
 
-static void nv_ck804_host_stop(struct ata_host_set *host_set)
+static void nv_ck804_host_stop(struct ata_host *host)
 {
-	struct pci_dev *pdev = to_pci_dev(host_set->dev);
+	struct pci_dev *pdev = to_pci_dev(host->dev);
 	u8 regval;
 
 	/* disable SATA space for CK804 */
@@ -578,7 +578,7 @@ static void nv_ck804_host_stop(struct ata_host_set *host_set)
 	regval &= ~NV_MCP_SATA_CFG_20_SATA_SPACE_EN;
 	pci_write_config_byte(pdev, NV_MCP_SATA_CFG_20, regval);
 
-	ata_pci_host_stop(host_set);
+	ata_pci_host_stop(host);
 }
 
 static int __init nv_init(void)
diff --git a/drivers/ata/sata_promise.c b/drivers/ata/sata_promise.c
index a5b3a7db7a9f..d627812ea73d 100644
--- a/drivers/ata/sata_promise.c
+++ b/drivers/ata/sata_promise.c
@@ -104,7 +104,7 @@ static void pdc_tf_load_mmio(struct ata_port *ap, const struct ata_taskfile *tf)
 static void pdc_exec_command_mmio(struct ata_port *ap, const struct ata_taskfile *tf);
 static void pdc_irq_clear(struct ata_port *ap);
 static unsigned int pdc_qc_issue_prot(struct ata_queued_cmd *qc);
-static void pdc_host_stop(struct ata_host_set *host_set);
+static void pdc_host_stop(struct ata_host *host);
 
 
 static struct scsi_host_template pdc_ata_sht = {
@@ -175,7 +175,7 @@ static const struct ata_port_info pdc_port_info[] = {
 	/* board_2037x */
 	{
 		.sht		= &pdc_ata_sht,
-		.host_flags	= PDC_COMMON_FLAGS | ATA_FLAG_SATA,
+		.flags		= PDC_COMMON_FLAGS | ATA_FLAG_SATA,
 		.pio_mask	= 0x1f, /* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma0-2 */
 		.udma_mask	= 0x7f, /* udma0-6 ; FIXME */
@@ -185,7 +185,7 @@ static const struct ata_port_info pdc_port_info[] = {
 	/* board_20319 */
 	{
 		.sht		= &pdc_ata_sht,
-		.host_flags	= PDC_COMMON_FLAGS | ATA_FLAG_SATA,
+		.flags		= PDC_COMMON_FLAGS | ATA_FLAG_SATA,
 		.pio_mask	= 0x1f, /* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma0-2 */
 		.udma_mask	= 0x7f, /* udma0-6 ; FIXME */
@@ -195,7 +195,7 @@ static const struct ata_port_info pdc_port_info[] = {
 	/* board_20619 */
 	{
 		.sht		= &pdc_ata_sht,
-		.host_flags	= PDC_COMMON_FLAGS | ATA_FLAG_SLAVE_POSS,
+		.flags		= PDC_COMMON_FLAGS | ATA_FLAG_SLAVE_POSS,
 		.pio_mask	= 0x1f, /* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma0-2 */
 		.udma_mask	= 0x7f, /* udma0-6 ; FIXME */
@@ -205,7 +205,7 @@ static const struct ata_port_info pdc_port_info[] = {
 	/* board_20771 */
 	{
 		.sht		= &pdc_ata_sht,
-		.host_flags	= PDC_COMMON_FLAGS | ATA_FLAG_SATA,
+		.flags		= PDC_COMMON_FLAGS | ATA_FLAG_SATA,
 		.pio_mask	= 0x1f, /* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma0-2 */
 		.udma_mask	= 0x7f, /* udma0-6 ; FIXME */
@@ -215,7 +215,7 @@ static const struct ata_port_info pdc_port_info[] = {
 	/* board_2057x */
 	{
 		.sht		= &pdc_ata_sht,
-		.host_flags	= PDC_COMMON_FLAGS | ATA_FLAG_SATA,
+		.flags		= PDC_COMMON_FLAGS | ATA_FLAG_SATA,
 		.pio_mask	= 0x1f, /* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma0-2 */
 		.udma_mask	= 0x7f, /* udma0-6 ; FIXME */
@@ -225,7 +225,7 @@ static const struct ata_port_info pdc_port_info[] = {
 	/* board_40518 */
 	{
 		.sht		= &pdc_ata_sht,
-		.host_flags	= PDC_COMMON_FLAGS | ATA_FLAG_SATA,
+		.flags		= PDC_COMMON_FLAGS | ATA_FLAG_SATA,
 		.pio_mask	= 0x1f, /* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma0-2 */
 		.udma_mask	= 0x7f, /* udma0-6 ; FIXME */
@@ -292,7 +292,7 @@ static struct pci_driver pdc_ata_pci_driver = {
 
 static int pdc_port_start(struct ata_port *ap)
 {
-	struct device *dev = ap->host_set->dev;
+	struct device *dev = ap->host->dev;
 	struct pdc_port_priv *pp;
 	int rc;
 
@@ -326,7 +326,7 @@ err_out:
 
 static void pdc_port_stop(struct ata_port *ap)
 {
-	struct device *dev = ap->host_set->dev;
+	struct device *dev = ap->host->dev;
 	struct pdc_port_priv *pp = ap->private_data;
 
 	ap->private_data = NULL;
@@ -336,11 +336,11 @@ static void pdc_port_stop(struct ata_port *ap)
 }
 
 
-static void pdc_host_stop(struct ata_host_set *host_set)
+static void pdc_host_stop(struct ata_host *host)
 {
-	struct pdc_host_priv *hp = host_set->private_data;
+	struct pdc_host_priv *hp = host->private_data;
 
-	ata_pci_host_stop(host_set);
+	ata_pci_host_stop(host);
 
 	kfree(hp);
 }
@@ -443,14 +443,14 @@ static void pdc_qc_prep(struct ata_queued_cmd *qc)
 
 static void pdc_eng_timeout(struct ata_port *ap)
 {
-	struct ata_host_set *host_set = ap->host_set;
+	struct ata_host *host = ap->host;
 	u8 drv_stat;
 	struct ata_queued_cmd *qc;
 	unsigned long flags;
 
 	DPRINTK("ENTER\n");
 
-	spin_lock_irqsave(&host_set->lock, flags);
+	spin_lock_irqsave(&host->lock, flags);
 
 	qc = ata_qc_from_tag(ap, ap->active_tag);
 
@@ -473,7 +473,7 @@ static void pdc_eng_timeout(struct ata_port *ap)
 		break;
 	}
 
-	spin_unlock_irqrestore(&host_set->lock, flags);
+	spin_unlock_irqrestore(&host->lock, flags);
 	ata_eh_qc_complete(qc);
 	DPRINTK("EXIT\n");
 }
@@ -509,15 +509,15 @@ static inline unsigned int pdc_host_intr( struct ata_port *ap,
 
 static void pdc_irq_clear(struct ata_port *ap)
 {
-	struct ata_host_set *host_set = ap->host_set;
-	void __iomem *mmio = host_set->mmio_base;
+	struct ata_host *host = ap->host;
+	void __iomem *mmio = host->mmio_base;
 
 	readl(mmio + PDC_INT_SEQMASK);
 }
 
 static irqreturn_t pdc_interrupt (int irq, void *dev_instance, struct pt_regs *regs)
 {
-	struct ata_host_set *host_set = dev_instance;
+	struct ata_host *host = dev_instance;
 	struct ata_port *ap;
 	u32 mask = 0;
 	unsigned int i, tmp;
@@ -526,12 +526,12 @@ static irqreturn_t pdc_interrupt (int irq, void *dev_instance, struct pt_regs *r
 
 	VPRINTK("ENTER\n");
 
-	if (!host_set || !host_set->mmio_base) {
+	if (!host || !host->mmio_base) {
 		VPRINTK("QUICK EXIT\n");
 		return IRQ_NONE;
 	}
 
-	mmio_base = host_set->mmio_base;
+	mmio_base = host->mmio_base;
 
 	/* reading should also clear interrupts */
 	mask = readl(mmio_base + PDC_INT_SEQMASK);
@@ -541,7 +541,7 @@ static irqreturn_t pdc_interrupt (int irq, void *dev_instance, struct pt_regs *r
 		return IRQ_NONE;
 	}
 
-	spin_lock(&host_set->lock);
+	spin_lock(&host->lock);
 
 	mask &= 0xffff;		/* only 16 tags possible */
 	if (!mask) {
@@ -551,9 +551,9 @@ static irqreturn_t pdc_interrupt (int irq, void *dev_instance, struct pt_regs *r
 
 	writel(mask, mmio_base + PDC_INT_SEQMASK);
 
-	for (i = 0; i < host_set->n_ports; i++) {
+	for (i = 0; i < host->n_ports; i++) {
 		VPRINTK("port %u\n", i);
-		ap = host_set->ports[i];
+		ap = host->ports[i];
 		tmp = mask & (1 << (i + 1));
 		if (tmp && ap &&
 		    !(ap->flags & ATA_FLAG_DISABLED)) {
@@ -568,7 +568,7 @@ static irqreturn_t pdc_interrupt (int irq, void *dev_instance, struct pt_regs *r
 	VPRINTK("EXIT\n");
 
 done_irq:
-	spin_unlock(&host_set->lock);
+	spin_unlock(&host->lock);
 	return IRQ_RETVAL(handled);
 }
 
@@ -581,8 +581,8 @@ static inline void pdc_packet_start(struct ata_queued_cmd *qc)
 
 	VPRINTK("ENTER, ap %p\n", ap);
 
-	writel(0x00000001, ap->host_set->mmio_base + (seq * 4));
-	readl(ap->host_set->mmio_base + (seq * 4));	/* flush */
+	writel(0x00000001, ap->host->mmio_base + (seq * 4));
+	readl(ap->host->mmio_base + (seq * 4));	/* flush */
 
 	pp->pkt[2] = seq;
 	wmb();			/* flush PRD, pkt writes */
@@ -743,7 +743,7 @@ static int pdc_ata_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 	probe_ent->private_data = hp;
 
 	probe_ent->sht		= pdc_port_info[board_idx].sht;
-	probe_ent->host_flags	= pdc_port_info[board_idx].host_flags;
+	probe_ent->port_flags	= pdc_port_info[board_idx].flags;
 	probe_ent->pio_mask	= pdc_port_info[board_idx].pio_mask;
 	probe_ent->mwdma_mask	= pdc_port_info[board_idx].mwdma_mask;
 	probe_ent->udma_mask	= pdc_port_info[board_idx].udma_mask;
diff --git a/drivers/ata/sata_qstor.c b/drivers/ata/sata_qstor.c
index 71bd6712b377..fa29dfe2a7b5 100644
--- a/drivers/ata/sata_qstor.c
+++ b/drivers/ata/sata_qstor.c
@@ -116,7 +116,7 @@ static void qs_scr_write (struct ata_port *ap, unsigned int sc_reg, u32 val);
 static int qs_ata_init_one (struct pci_dev *pdev, const struct pci_device_id *ent);
 static irqreturn_t qs_intr (int irq, void *dev_instance, struct pt_regs *regs);
 static int qs_port_start(struct ata_port *ap);
-static void qs_host_stop(struct ata_host_set *host_set);
+static void qs_host_stop(struct ata_host *host);
 static void qs_port_stop(struct ata_port *ap);
 static void qs_phy_reset(struct ata_port *ap);
 static void qs_qc_prep(struct ata_queued_cmd *qc);
@@ -174,7 +174,7 @@ static const struct ata_port_info qs_port_info[] = {
 	/* board_2068_idx */
 	{
 		.sht		= &qs_ata_sht,
-		.host_flags	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
+		.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
 				  ATA_FLAG_SATA_RESET |
 				  //FIXME ATA_FLAG_SRST |
 				  ATA_FLAG_MMIO | ATA_FLAG_PIO_POLLING,
@@ -220,7 +220,7 @@ static void qs_irq_clear(struct ata_port *ap)
 
 static inline void qs_enter_reg_mode(struct ata_port *ap)
 {
-	u8 __iomem *chan = ap->host_set->mmio_base + (ap->port_no * 0x4000);
+	u8 __iomem *chan = ap->host->mmio_base + (ap->port_no * 0x4000);
 
 	writeb(QS_CTR0_REG, chan + QS_CCT_CTR0);
 	readb(chan + QS_CCT_CTR0);        /* flush */
@@ -228,7 +228,7 @@ static inline void qs_enter_reg_mode(struct ata_port *ap)
 
 static inline void qs_reset_channel_logic(struct ata_port *ap)
 {
-	u8 __iomem *chan = ap->host_set->mmio_base + (ap->port_no * 0x4000);
+	u8 __iomem *chan = ap->host->mmio_base + (ap->port_no * 0x4000);
 
 	writeb(QS_CTR1_RCHN, chan + QS_CCT_CTR1);
 	readb(chan + QS_CCT_CTR0);        /* flush */
@@ -342,7 +342,7 @@ static void qs_qc_prep(struct ata_queued_cmd *qc)
 static inline void qs_packet_start(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
-	u8 __iomem *chan = ap->host_set->mmio_base + (ap->port_no * 0x4000);
+	u8 __iomem *chan = ap->host->mmio_base + (ap->port_no * 0x4000);
 
 	VPRINTK("ENTER, ap %p\n", ap);
 
@@ -375,11 +375,11 @@ static unsigned int qs_qc_issue(struct ata_queued_cmd *qc)
 	return ata_qc_issue_prot(qc);
 }
 
-static inline unsigned int qs_intr_pkt(struct ata_host_set *host_set)
+static inline unsigned int qs_intr_pkt(struct ata_host *host)
 {
 	unsigned int handled = 0;
 	u8 sFFE;
-	u8 __iomem *mmio_base = host_set->mmio_base;
+	u8 __iomem *mmio_base = host->mmio_base;
 
 	do {
 		u32 sff0 = readl(mmio_base + QS_HST_SFF);
@@ -391,7 +391,7 @@ static inline unsigned int qs_intr_pkt(struct ata_host_set *host_set)
 			u8 sDST = sff0 >> 16;	/* dev status */
 			u8 sHST = sff1 & 0x3f;	/* host status */
 			unsigned int port_no = (sff1 >> 8) & 0x03;
-			struct ata_port *ap = host_set->ports[port_no];
+			struct ata_port *ap = host->ports[port_no];
 
 			DPRINTK("SFF=%08x%08x: sCHAN=%u sHST=%d sDST=%02x\n",
 					sff1, sff0, port_no, sHST, sDST);
@@ -421,13 +421,13 @@ static inline unsigned int qs_intr_pkt(struct ata_host_set *host_set)
 	return handled;
 }
 
-static inline unsigned int qs_intr_mmio(struct ata_host_set *host_set)
+static inline unsigned int qs_intr_mmio(struct ata_host *host)
 {
 	unsigned int handled = 0, port_no;
 
-	for (port_no = 0; port_no < host_set->n_ports; ++port_no) {
+	for (port_no = 0; port_no < host->n_ports; ++port_no) {
 		struct ata_port *ap;
-		ap = host_set->ports[port_no];
+		ap = host->ports[port_no];
 		if (ap &&
 		    !(ap->flags & ATA_FLAG_DISABLED)) {
 			struct ata_queued_cmd *qc;
@@ -457,14 +457,14 @@ static inline unsigned int qs_intr_mmio(struct ata_host_set *host_set)
 
 static irqreturn_t qs_intr(int irq, void *dev_instance, struct pt_regs *regs)
 {
-	struct ata_host_set *host_set = dev_instance;
+	struct ata_host *host = dev_instance;
 	unsigned int handled = 0;
 
 	VPRINTK("ENTER\n");
 
-	spin_lock(&host_set->lock);
-	handled  = qs_intr_pkt(host_set) | qs_intr_mmio(host_set);
-	spin_unlock(&host_set->lock);
+	spin_lock(&host->lock);
+	handled  = qs_intr_pkt(host) | qs_intr_mmio(host);
+	spin_unlock(&host->lock);
 
 	VPRINTK("EXIT\n");
 
@@ -491,9 +491,9 @@ static void qs_ata_setup_port(struct ata_ioports *port, unsigned long base)
 
 static int qs_port_start(struct ata_port *ap)
 {
-	struct device *dev = ap->host_set->dev;
+	struct device *dev = ap->host->dev;
 	struct qs_port_priv *pp;
-	void __iomem *mmio_base = ap->host_set->mmio_base;
+	void __iomem *mmio_base = ap->host->mmio_base;
 	void __iomem *chan = mmio_base + (ap->port_no * 0x4000);
 	u64 addr;
 	int rc;
@@ -530,7 +530,7 @@ err_out:
 
 static void qs_port_stop(struct ata_port *ap)
 {
-	struct device *dev = ap->host_set->dev;
+	struct device *dev = ap->host->dev;
 	struct qs_port_priv *pp = ap->private_data;
 
 	if (pp != NULL) {
@@ -543,10 +543,10 @@ static void qs_port_stop(struct ata_port *ap)
 	ata_port_stop(ap);
 }
 
-static void qs_host_stop(struct ata_host_set *host_set)
+static void qs_host_stop(struct ata_host *host)
 {
-	void __iomem *mmio_base = host_set->mmio_base;
-	struct pci_dev *pdev = to_pci_dev(host_set->dev);
+	void __iomem *mmio_base = host->mmio_base;
+	struct pci_dev *pdev = to_pci_dev(host->dev);
 
 	writeb(0, mmio_base + QS_HCT_CTRL); /* disable host interrupts */
 	writeb(QS_CNFG3_GSRST, mmio_base + QS_HCF_CNFG3); /* global reset */
@@ -673,7 +673,7 @@ static int qs_ata_init_one(struct pci_dev *pdev,
 	INIT_LIST_HEAD(&probe_ent->node);
 
 	probe_ent->sht		= qs_port_info[board_idx].sht;
-	probe_ent->host_flags	= qs_port_info[board_idx].host_flags;
+	probe_ent->port_flags	= qs_port_info[board_idx].flags;
 	probe_ent->pio_mask	= qs_port_info[board_idx].pio_mask;
 	probe_ent->mwdma_mask	= qs_port_info[board_idx].mwdma_mask;
 	probe_ent->udma_mask	= qs_port_info[board_idx].udma_mask;
diff --git a/drivers/ata/sata_sil.c b/drivers/ata/sata_sil.c
index f17b3ae4dd37..c63dbabc0cd9 100644
--- a/drivers/ata/sata_sil.c
+++ b/drivers/ata/sata_sil.c
@@ -56,7 +56,7 @@ enum {
 	SIL_FLAG_RERR_ON_DMA_ACT = (1 << 29),
 	SIL_FLAG_MOD15WRITE	= (1 << 30),
 
-	SIL_DFL_HOST_FLAGS	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
+	SIL_DFL_PORT_FLAGS	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
 				  ATA_FLAG_MMIO | ATA_FLAG_HRST_TO_RESUME,
 
 	/*
@@ -218,7 +218,7 @@ static const struct ata_port_info sil_port_info[] = {
 	/* sil_3112 */
 	{
 		.sht		= &sil_sht,
-		.host_flags	= SIL_DFL_HOST_FLAGS | SIL_FLAG_MOD15WRITE,
+		.flags		= SIL_DFL_PORT_FLAGS | SIL_FLAG_MOD15WRITE,
 		.pio_mask	= 0x1f,			/* pio0-4 */
 		.mwdma_mask	= 0x07,			/* mwdma0-2 */
 		.udma_mask	= 0x3f,			/* udma0-5 */
@@ -227,7 +227,7 @@ static const struct ata_port_info sil_port_info[] = {
 	/* sil_3112_no_sata_irq */
 	{
 		.sht		= &sil_sht,
-		.host_flags	= SIL_DFL_HOST_FLAGS | SIL_FLAG_MOD15WRITE |
+		.flags		= SIL_DFL_PORT_FLAGS | SIL_FLAG_MOD15WRITE |
 				  SIL_FLAG_NO_SATA_IRQ,
 		.pio_mask	= 0x1f,			/* pio0-4 */
 		.mwdma_mask	= 0x07,			/* mwdma0-2 */
@@ -237,7 +237,7 @@ static const struct ata_port_info sil_port_info[] = {
 	/* sil_3512 */
 	{
 		.sht		= &sil_sht,
-		.host_flags	= SIL_DFL_HOST_FLAGS | SIL_FLAG_RERR_ON_DMA_ACT,
+		.flags		= SIL_DFL_PORT_FLAGS | SIL_FLAG_RERR_ON_DMA_ACT,
 		.pio_mask	= 0x1f,			/* pio0-4 */
 		.mwdma_mask	= 0x07,			/* mwdma0-2 */
 		.udma_mask	= 0x3f,			/* udma0-5 */
@@ -246,7 +246,7 @@ static const struct ata_port_info sil_port_info[] = {
 	/* sil_3114 */
 	{
 		.sht		= &sil_sht,
-		.host_flags	= SIL_DFL_HOST_FLAGS | SIL_FLAG_RERR_ON_DMA_ACT,
+		.flags		= SIL_DFL_PORT_FLAGS | SIL_FLAG_RERR_ON_DMA_ACT,
 		.pio_mask	= 0x1f,			/* pio0-4 */
 		.mwdma_mask	= 0x07,			/* mwdma0-2 */
 		.udma_mask	= 0x3f,			/* udma0-5 */
@@ -295,10 +295,9 @@ static unsigned char sil_get_device_cache_line(struct pci_dev *pdev)
 
 static void sil_post_set_mode (struct ata_port *ap)
 {
-	struct ata_host_set *host_set = ap->host_set;
+	struct ata_host *host = ap->host;
 	struct ata_device *dev;
-	void __iomem *addr =
-		host_set->mmio_base + sil_port[ap->port_no].xfer_mode;
+	void __iomem *addr = host->mmio_base + sil_port[ap->port_no].xfer_mode;
 	u32 tmp, dev_mode[2];
 	unsigned int i;
 
@@ -440,15 +439,15 @@ static void sil_host_intr(struct ata_port *ap, u32 bmdma2)
 static irqreturn_t sil_interrupt(int irq, void *dev_instance,
 				 struct pt_regs *regs)
 {
-	struct ata_host_set *host_set = dev_instance;
-	void __iomem *mmio_base = host_set->mmio_base;
+	struct ata_host *host = dev_instance;
+	void __iomem *mmio_base = host->mmio_base;
 	int handled = 0;
 	int i;
 
-	spin_lock(&host_set->lock);
+	spin_lock(&host->lock);
 
-	for (i = 0; i < host_set->n_ports; i++) {
-		struct ata_port *ap = host_set->ports[i];
+	for (i = 0; i < host->n_ports; i++) {
+		struct ata_port *ap = host->ports[i];
 		u32 bmdma2 = readl(mmio_base + sil_port[ap->port_no].bmdma2);
 
 		if (unlikely(!ap || ap->flags & ATA_FLAG_DISABLED))
@@ -466,14 +465,14 @@ static irqreturn_t sil_interrupt(int irq, void *dev_instance,
 		handled = 1;
 	}
 
-	spin_unlock(&host_set->lock);
+	spin_unlock(&host->lock);
 
 	return IRQ_RETVAL(handled);
 }
 
 static void sil_freeze(struct ata_port *ap)
 {
-	void __iomem *mmio_base = ap->host_set->mmio_base;
+	void __iomem *mmio_base = ap->host->mmio_base;
 	u32 tmp;
 
 	/* global IRQ mask doesn't block SATA IRQ, turn off explicitly */
@@ -488,7 +487,7 @@ static void sil_freeze(struct ata_port *ap)
 
 static void sil_thaw(struct ata_port *ap)
 {
-	void __iomem *mmio_base = ap->host_set->mmio_base;
+	void __iomem *mmio_base = ap->host->mmio_base;
 	u32 tmp;
 
 	/* clear IRQ */
@@ -567,7 +566,7 @@ static void sil_dev_config(struct ata_port *ap, struct ata_device *dev)
 }
 
 static void sil_init_controller(struct pci_dev *pdev,
-				int n_ports, unsigned long host_flags,
+				int n_ports, unsigned long port_flags,
 				void __iomem *mmio_base)
 {
 	u8 cls;
@@ -587,7 +586,7 @@ static void sil_init_controller(struct pci_dev *pdev,
 			   "cache line size not set.  Driver may not function\n");
 
 	/* Apply R_ERR on DMA activate FIS errata workaround */
-	if (host_flags & SIL_FLAG_RERR_ON_DMA_ACT) {
+	if (port_flags & SIL_FLAG_RERR_ON_DMA_ACT) {
 		int cnt;
 
 		for (i = 0, cnt = 0; i < n_ports; i++) {
@@ -658,7 +657,7 @@ static int sil_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 	probe_ent->udma_mask = sil_port_info[ent->driver_data].udma_mask;
        	probe_ent->irq = pdev->irq;
        	probe_ent->irq_flags = IRQF_SHARED;
-	probe_ent->host_flags = sil_port_info[ent->driver_data].host_flags;
+	probe_ent->port_flags = sil_port_info[ent->driver_data].flags;
 
 	mmio_base = pci_iomap(pdev, 5, 0);
 	if (mmio_base == NULL) {
@@ -679,7 +678,7 @@ static int sil_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 		ata_std_ports(&probe_ent->port[i]);
 	}
 
-	sil_init_controller(pdev, probe_ent->n_ports, probe_ent->host_flags,
+	sil_init_controller(pdev, probe_ent->n_ports, probe_ent->port_flags,
 			    mmio_base);
 
 	pci_set_master(pdev);
@@ -703,12 +702,12 @@ err_out:
 #ifdef CONFIG_PM
 static int sil_pci_device_resume(struct pci_dev *pdev)
 {
-	struct ata_host_set *host_set = dev_get_drvdata(&pdev->dev);
+	struct ata_host *host = dev_get_drvdata(&pdev->dev);
 
 	ata_pci_device_do_resume(pdev);
-	sil_init_controller(pdev, host_set->n_ports, host_set->ports[0]->flags,
-			    host_set->mmio_base);
-	ata_host_set_resume(host_set);
+	sil_init_controller(pdev, host->n_ports, host->ports[0]->flags,
+			    host->mmio_base);
+	ata_host_resume(host);
 
 	return 0;
 }
diff --git a/drivers/ata/sata_sil24.c b/drivers/ata/sata_sil24.c
index 2d7cf3264587..39cb07baebae 100644
--- a/drivers/ata/sata_sil24.c
+++ b/drivers/ata/sata_sil24.c
@@ -316,7 +316,7 @@ struct sil24_port_priv {
 	struct ata_taskfile tf;			/* Cached taskfile registers */
 };
 
-/* ap->host_set->private_data */
+/* ap->host->private_data */
 struct sil24_host_priv {
 	void __iomem *host_base;	/* global controller control (128 bytes @BAR0) */
 	void __iomem *port_base;	/* port registers (4 * 8192 bytes @BAR2) */
@@ -337,7 +337,7 @@ static void sil24_error_handler(struct ata_port *ap);
 static void sil24_post_internal_cmd(struct ata_queued_cmd *qc);
 static int sil24_port_start(struct ata_port *ap);
 static void sil24_port_stop(struct ata_port *ap);
-static void sil24_host_stop(struct ata_host_set *host_set);
+static void sil24_host_stop(struct ata_host *host);
 static int sil24_init_one(struct pci_dev *pdev, const struct pci_device_id *ent);
 #ifdef CONFIG_PM
 static int sil24_pci_device_resume(struct pci_dev *pdev);
@@ -415,7 +415,7 @@ static const struct ata_port_operations sil24_ops = {
 };
 
 /*
- * Use bits 30-31 of host_flags to encode available port numbers.
+ * Use bits 30-31 of port_flags to encode available port numbers.
  * Current maxium is 4.
  */
 #define SIL24_NPORTS2FLAG(nports)	((((unsigned)(nports) - 1) & 0x3) << 30)
@@ -425,7 +425,7 @@ static struct ata_port_info sil24_port_info[] = {
 	/* sil_3124 */
 	{
 		.sht		= &sil24_sht,
-		.host_flags	= SIL24_COMMON_FLAGS | SIL24_NPORTS2FLAG(4) |
+		.flags		= SIL24_COMMON_FLAGS | SIL24_NPORTS2FLAG(4) |
 				  SIL24_FLAG_PCIX_IRQ_WOC,
 		.pio_mask	= 0x1f,			/* pio0-4 */
 		.mwdma_mask	= 0x07,			/* mwdma0-2 */
@@ -435,7 +435,7 @@ static struct ata_port_info sil24_port_info[] = {
 	/* sil_3132 */
 	{
 		.sht		= &sil24_sht,
-		.host_flags	= SIL24_COMMON_FLAGS | SIL24_NPORTS2FLAG(2),
+		.flags		= SIL24_COMMON_FLAGS | SIL24_NPORTS2FLAG(2),
 		.pio_mask	= 0x1f,			/* pio0-4 */
 		.mwdma_mask	= 0x07,			/* mwdma0-2 */
 		.udma_mask	= 0x3f,			/* udma0-5 */
@@ -444,7 +444,7 @@ static struct ata_port_info sil24_port_info[] = {
 	/* sil_3131/sil_3531 */
 	{
 		.sht		= &sil24_sht,
-		.host_flags	= SIL24_COMMON_FLAGS | SIL24_NPORTS2FLAG(1),
+		.flags		= SIL24_COMMON_FLAGS | SIL24_NPORTS2FLAG(1),
 		.pio_mask	= 0x1f,			/* pio0-4 */
 		.mwdma_mask	= 0x07,			/* mwdma0-2 */
 		.udma_mask	= 0x3f,			/* udma0-5 */
@@ -871,8 +871,8 @@ static inline void sil24_host_intr(struct ata_port *ap)
 
 static irqreturn_t sil24_interrupt(int irq, void *dev_instance, struct pt_regs *regs)
 {
-	struct ata_host_set *host_set = dev_instance;
-	struct sil24_host_priv *hpriv = host_set->private_data;
+	struct ata_host *host = dev_instance;
+	struct sil24_host_priv *hpriv = host->private_data;
 	unsigned handled = 0;
 	u32 status;
 	int i;
@@ -888,20 +888,20 @@ static irqreturn_t sil24_interrupt(int irq, void *dev_instance, struct pt_regs *
 	if (!(status & IRQ_STAT_4PORTS))
 		goto out;
 
-	spin_lock(&host_set->lock);
+	spin_lock(&host->lock);
 
-	for (i = 0; i < host_set->n_ports; i++)
+	for (i = 0; i < host->n_ports; i++)
 		if (status & (1 << i)) {
-			struct ata_port *ap = host_set->ports[i];
+			struct ata_port *ap = host->ports[i];
 			if (ap && !(ap->flags & ATA_FLAG_DISABLED)) {
-				sil24_host_intr(host_set->ports[i]);
+				sil24_host_intr(host->ports[i]);
 				handled++;
 			} else
 				printk(KERN_ERR DRV_NAME
 				       ": interrupt from disabled port %d\n", i);
 		}
 
-	spin_unlock(&host_set->lock);
+	spin_unlock(&host->lock);
  out:
 	return IRQ_RETVAL(handled);
 }
@@ -941,7 +941,7 @@ static inline void sil24_cblk_free(struct sil24_port_priv *pp, struct device *de
 
 static int sil24_port_start(struct ata_port *ap)
 {
-	struct device *dev = ap->host_set->dev;
+	struct device *dev = ap->host->dev;
 	struct sil24_port_priv *pp;
 	union sil24_cmd_block *cb;
 	size_t cb_size = sizeof(*cb) * SIL24_MAX_CMDS;
@@ -980,7 +980,7 @@ err_out:
 
 static void sil24_port_stop(struct ata_port *ap)
 {
-	struct device *dev = ap->host_set->dev;
+	struct device *dev = ap->host->dev;
 	struct sil24_port_priv *pp = ap->private_data;
 
 	sil24_cblk_free(pp, dev);
@@ -988,10 +988,10 @@ static void sil24_port_stop(struct ata_port *ap)
 	kfree(pp);
 }
 
-static void sil24_host_stop(struct ata_host_set *host_set)
+static void sil24_host_stop(struct ata_host *host)
 {
-	struct sil24_host_priv *hpriv = host_set->private_data;
-	struct pci_dev *pdev = to_pci_dev(host_set->dev);
+	struct sil24_host_priv *hpriv = host->private_data;
+	struct pci_dev *pdev = to_pci_dev(host->dev);
 
 	pci_iounmap(pdev, hpriv->host_base);
 	pci_iounmap(pdev, hpriv->port_base);
@@ -999,7 +999,7 @@ static void sil24_host_stop(struct ata_host_set *host_set)
 }
 
 static void sil24_init_controller(struct pci_dev *pdev, int n_ports,
-				  unsigned long host_flags,
+				  unsigned long port_flags,
 				  void __iomem *host_base,
 				  void __iomem *port_base)
 {
@@ -1032,7 +1032,7 @@ static void sil24_init_controller(struct pci_dev *pdev, int n_ports,
 		}
 
 		/* Configure IRQ WoC */
-		if (host_flags & SIL24_FLAG_PCIX_IRQ_WOC)
+		if (port_flags & SIL24_FLAG_PCIX_IRQ_WOC)
 			writel(PORT_CS_IRQ_WOC, port + PORT_CTRL_STAT);
 		else
 			writel(PORT_CS_IRQ_WOC, port + PORT_CTRL_CLR);
@@ -1101,12 +1101,12 @@ static int sil24_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	INIT_LIST_HEAD(&probe_ent->node);
 
 	probe_ent->sht		= pinfo->sht;
-	probe_ent->host_flags	= pinfo->host_flags;
+	probe_ent->port_flags	= pinfo->flags;
 	probe_ent->pio_mask	= pinfo->pio_mask;
 	probe_ent->mwdma_mask	= pinfo->mwdma_mask;
 	probe_ent->udma_mask	= pinfo->udma_mask;
 	probe_ent->port_ops	= pinfo->port_ops;
-	probe_ent->n_ports	= SIL24_FLAG2NPORTS(pinfo->host_flags);
+	probe_ent->n_ports	= SIL24_FLAG2NPORTS(pinfo->flags);
 
 	probe_ent->irq = pdev->irq;
 	probe_ent->irq_flags = IRQF_SHARED;
@@ -1144,14 +1144,14 @@ static int sil24_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	}
 
 	/* Apply workaround for completion IRQ loss on PCI-X errata */
-	if (probe_ent->host_flags & SIL24_FLAG_PCIX_IRQ_WOC) {
+	if (probe_ent->port_flags & SIL24_FLAG_PCIX_IRQ_WOC) {
 		tmp = readl(host_base + HOST_CTRL);
 		if (tmp & (HOST_CTRL_TRDY | HOST_CTRL_STOP | HOST_CTRL_DEVSEL))
 			dev_printk(KERN_INFO, &pdev->dev,
 				   "Applying completion IRQ loss on PCI-X "
 				   "errata fix\n");
 		else
-			probe_ent->host_flags &= ~SIL24_FLAG_PCIX_IRQ_WOC;
+			probe_ent->port_flags &= ~SIL24_FLAG_PCIX_IRQ_WOC;
 	}
 
 	for (i = 0; i < probe_ent->n_ports; i++) {
@@ -1164,7 +1164,7 @@ static int sil24_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 		ata_std_ports(&probe_ent->port[i]);
 	}
 
-	sil24_init_controller(pdev, probe_ent->n_ports, probe_ent->host_flags,
+	sil24_init_controller(pdev, probe_ent->n_ports, probe_ent->port_flags,
 			      host_base, port_base);
 
 	pci_set_master(pdev);
@@ -1191,19 +1191,18 @@ static int sil24_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 #ifdef CONFIG_PM
 static int sil24_pci_device_resume(struct pci_dev *pdev)
 {
-	struct ata_host_set *host_set = dev_get_drvdata(&pdev->dev);
-	struct sil24_host_priv *hpriv = host_set->private_data;
+	struct ata_host *host = dev_get_drvdata(&pdev->dev);
+	struct sil24_host_priv *hpriv = host->private_data;
 
 	ata_pci_device_do_resume(pdev);
 
 	if (pdev->dev.power.power_state.event == PM_EVENT_SUSPEND)
 		writel(HOST_CTRL_GLOBAL_RST, hpriv->host_base + HOST_CTRL);
 
-	sil24_init_controller(pdev, host_set->n_ports,
-			      host_set->ports[0]->flags,
+	sil24_init_controller(pdev, host->n_ports, host->ports[0]->flags,
 			      hpriv->host_base, hpriv->port_base);
 
-	ata_host_set_resume(host_set);
+	ata_host_resume(host);
 
 	return 0;
 }
diff --git a/drivers/ata/sata_sis.c b/drivers/ata/sata_sis.c
index ac24f66897f6..9b17375d8056 100644
--- a/drivers/ata/sata_sis.c
+++ b/drivers/ata/sata_sis.c
@@ -128,7 +128,7 @@ static const struct ata_port_operations sis_ops = {
 
 static struct ata_port_info sis_port_info = {
 	.sht		= &sis_sht,
-	.host_flags	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
+	.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
 	.pio_mask	= 0x1f,
 	.mwdma_mask	= 0x7,
 	.udma_mask	= 0x7f,
@@ -158,7 +158,7 @@ static unsigned int get_scr_cfg_addr(unsigned int port_no, unsigned int sc_reg,
 
 static u32 sis_scr_cfg_read (struct ata_port *ap, unsigned int sc_reg)
 {
-	struct pci_dev *pdev = to_pci_dev(ap->host_set->dev);
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
 	unsigned int cfg_addr = get_scr_cfg_addr(ap->port_no, sc_reg, pdev->device);
 	u32 val, val2 = 0;
 	u8 pmr;
@@ -178,7 +178,7 @@ static u32 sis_scr_cfg_read (struct ata_port *ap, unsigned int sc_reg)
 
 static void sis_scr_cfg_write (struct ata_port *ap, unsigned int scr, u32 val)
 {
-	struct pci_dev *pdev = to_pci_dev(ap->host_set->dev);
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
 	unsigned int cfg_addr = get_scr_cfg_addr(ap->port_no, scr, pdev->device);
 	u8 pmr;
 
@@ -195,7 +195,7 @@ static void sis_scr_cfg_write (struct ata_port *ap, unsigned int scr, u32 val)
 
 static u32 sis_scr_read (struct ata_port *ap, unsigned int sc_reg)
 {
-	struct pci_dev *pdev = to_pci_dev(ap->host_set->dev);
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
 	u32 val, val2 = 0;
 	u8 pmr;
 
@@ -217,7 +217,7 @@ static u32 sis_scr_read (struct ata_port *ap, unsigned int sc_reg)
 
 static void sis_scr_write (struct ata_port *ap, unsigned int sc_reg, u32 val)
 {
-	struct pci_dev *pdev = to_pci_dev(ap->host_set->dev);
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
 	u8 pmr;
 
 	if (sc_reg > SCR_CONTROL)
@@ -275,17 +275,17 @@ static int sis_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 	/* check and see if the SCRs are in IO space or PCI cfg space */
 	pci_read_config_dword(pdev, SIS_GENCTL, &genctl);
 	if ((genctl & GENCTL_IOMAPPED_SCR) == 0)
-		probe_ent->host_flags |= SIS_FLAG_CFGSCR;
+		probe_ent->port_flags |= SIS_FLAG_CFGSCR;
 
 	/* if hardware thinks SCRs are in IO space, but there are
 	 * no IO resources assigned, change to PCI cfg space.
 	 */
-	if ((!(probe_ent->host_flags & SIS_FLAG_CFGSCR)) &&
+	if ((!(probe_ent->port_flags & SIS_FLAG_CFGSCR)) &&
 	    ((pci_resource_start(pdev, SIS_SCR_PCI_BAR) == 0) ||
 	     (pci_resource_len(pdev, SIS_SCR_PCI_BAR) < 128))) {
 		genctl &= ~GENCTL_IOMAPPED_SCR;
 		pci_write_config_dword(pdev, SIS_GENCTL, genctl);
-		probe_ent->host_flags |= SIS_FLAG_CFGSCR;
+		probe_ent->port_flags |= SIS_FLAG_CFGSCR;
 	}
 
 	pci_read_config_byte(pdev, SIS_PMR, &pmr);
@@ -306,7 +306,7 @@ static int sis_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 		port2_start = 0x20;
 	}
 
-	if (!(probe_ent->host_flags & SIS_FLAG_CFGSCR)) {
+	if (!(probe_ent->port_flags & SIS_FLAG_CFGSCR)) {
 		probe_ent->port[0].scr_addr =
 			pci_resource_start(pdev, SIS_SCR_PCI_BAR);
 		probe_ent->port[1].scr_addr =
diff --git a/drivers/ata/sata_svw.c b/drivers/ata/sata_svw.c
index baf259a966d0..2a7e3495cf16 100644
--- a/drivers/ata/sata_svw.c
+++ b/drivers/ata/sata_svw.c
@@ -169,7 +169,7 @@ static void k2_sata_tf_read(struct ata_port *ap, struct ata_taskfile *tf)
  *	@qc: Info associated with this ATA transaction.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 static void k2_bmdma_setup_mmio (struct ata_queued_cmd *qc)
@@ -199,7 +199,7 @@ static void k2_bmdma_setup_mmio (struct ata_queued_cmd *qc)
  *	@qc: Info associated with this ATA transaction.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host_set lock)
+ *	spin_lock_irqsave(host lock)
  */
 
 static void k2_bmdma_start_mmio (struct ata_queued_cmd *qc)
@@ -261,12 +261,12 @@ static int k2_sata_proc_info(struct Scsi_Host *shost, char *page, char **start,
 		return 0;
 
 	/* Find the OF node for the PCI device proper */
-	np = pci_device_to_OF_node(to_pci_dev(ap->host_set->dev));
+	np = pci_device_to_OF_node(to_pci_dev(ap->host->dev));
 	if (np == NULL)
 		return 0;
 
 	/* Match it to a port node */
-	index = (ap == ap->host_set->ports[0]) ? 0 : 1;
+	index = (ap == ap->host->ports[0]) ? 0 : 1;
 	for (np = np->child; np != NULL; np = np->sibling) {
 		u32 *reg = (u32 *)get_property(np, "reg", NULL);
 		if (!reg)
@@ -423,7 +423,7 @@ static int k2_sata_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 	writel(0x0, mmio_base + K2_SATA_SIM_OFFSET);
 
 	probe_ent->sht = &k2_sata_sht;
-	probe_ent->host_flags = ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
+	probe_ent->port_flags = ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
 				ATA_FLAG_MMIO;
 	probe_ent->port_ops = &k2_sata_ops;
 	probe_ent->n_ports = 4;
diff --git a/drivers/ata/sata_sx4.c b/drivers/ata/sata_sx4.c
index 0da83cba5c12..091867e10ea3 100644
--- a/drivers/ata/sata_sx4.c
+++ b/drivers/ata/sata_sx4.c
@@ -160,7 +160,7 @@ static void pdc_port_stop(struct ata_port *ap);
 static void pdc20621_qc_prep(struct ata_queued_cmd *qc);
 static void pdc_tf_load_mmio(struct ata_port *ap, const struct ata_taskfile *tf);
 static void pdc_exec_command_mmio(struct ata_port *ap, const struct ata_taskfile *tf);
-static void pdc20621_host_stop(struct ata_host_set *host_set);
+static void pdc20621_host_stop(struct ata_host *host);
 static unsigned int pdc20621_dimm_init(struct ata_probe_ent *pe);
 static int pdc20621_detect_dimm(struct ata_probe_ent *pe);
 static unsigned int pdc20621_i2c_read(struct ata_probe_ent *pe,
@@ -218,7 +218,7 @@ static const struct ata_port_info pdc_port_info[] = {
 	/* board_20621 */
 	{
 		.sht		= &pdc_sata_sht,
-		.host_flags	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
+		.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
 				  ATA_FLAG_SRST | ATA_FLAG_MMIO |
 				  ATA_FLAG_NO_ATAPI | ATA_FLAG_PIO_POLLING,
 		.pio_mask	= 0x1f, /* pio0-4 */
@@ -244,21 +244,21 @@ static struct pci_driver pdc_sata_pci_driver = {
 };
 
 
-static void pdc20621_host_stop(struct ata_host_set *host_set)
+static void pdc20621_host_stop(struct ata_host *host)
 {
-	struct pci_dev *pdev = to_pci_dev(host_set->dev);
-	struct pdc_host_priv *hpriv = host_set->private_data;
+	struct pci_dev *pdev = to_pci_dev(host->dev);
+	struct pdc_host_priv *hpriv = host->private_data;
 	void __iomem *dimm_mmio = hpriv->dimm_mmio;
 
 	pci_iounmap(pdev, dimm_mmio);
 	kfree(hpriv);
 
-	pci_iounmap(pdev, host_set->mmio_base);
+	pci_iounmap(pdev, host->mmio_base);
 }
 
 static int pdc_port_start(struct ata_port *ap)
 {
-	struct device *dev = ap->host_set->dev;
+	struct device *dev = ap->host->dev;
 	struct pdc_port_priv *pp;
 	int rc;
 
@@ -293,7 +293,7 @@ err_out:
 
 static void pdc_port_stop(struct ata_port *ap)
 {
-	struct device *dev = ap->host_set->dev;
+	struct device *dev = ap->host->dev;
 	struct pdc_port_priv *pp = ap->private_data;
 
 	ap->private_data = NULL;
@@ -453,8 +453,8 @@ static void pdc20621_dma_prep(struct ata_queued_cmd *qc)
 	struct scatterlist *sg;
 	struct ata_port *ap = qc->ap;
 	struct pdc_port_priv *pp = ap->private_data;
-	void __iomem *mmio = ap->host_set->mmio_base;
-	struct pdc_host_priv *hpriv = ap->host_set->private_data;
+	void __iomem *mmio = ap->host->mmio_base;
+	struct pdc_host_priv *hpriv = ap->host->private_data;
 	void __iomem *dimm_mmio = hpriv->dimm_mmio;
 	unsigned int portno = ap->port_no;
 	unsigned int i, idx, total_len = 0, sgt_len;
@@ -514,8 +514,8 @@ static void pdc20621_nodata_prep(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
 	struct pdc_port_priv *pp = ap->private_data;
-	void __iomem *mmio = ap->host_set->mmio_base;
-	struct pdc_host_priv *hpriv = ap->host_set->private_data;
+	void __iomem *mmio = ap->host->mmio_base;
+	struct pdc_host_priv *hpriv = ap->host->private_data;
 	void __iomem *dimm_mmio = hpriv->dimm_mmio;
 	unsigned int portno = ap->port_no;
 	unsigned int i;
@@ -565,8 +565,8 @@ static void __pdc20621_push_hdma(struct ata_queued_cmd *qc,
 				 u32 pkt_ofs)
 {
 	struct ata_port *ap = qc->ap;
-	struct ata_host_set *host_set = ap->host_set;
-	void __iomem *mmio = host_set->mmio_base;
+	struct ata_host *host = ap->host;
+	void __iomem *mmio = host->mmio_base;
 
 	/* hard-code chip #0 */
 	mmio += PDC_CHIP0_OFS;
@@ -583,7 +583,7 @@ static void pdc20621_push_hdma(struct ata_queued_cmd *qc,
 				u32 pkt_ofs)
 {
 	struct ata_port *ap = qc->ap;
-	struct pdc_host_priv *pp = ap->host_set->private_data;
+	struct pdc_host_priv *pp = ap->host->private_data;
 	unsigned int idx = pp->hdma_prod & PDC_HDMA_Q_MASK;
 
 	if (!pp->doing_hdma) {
@@ -601,7 +601,7 @@ static void pdc20621_push_hdma(struct ata_queued_cmd *qc,
 static void pdc20621_pop_hdma(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
-	struct pdc_host_priv *pp = ap->host_set->private_data;
+	struct pdc_host_priv *pp = ap->host->private_data;
 	unsigned int idx = pp->hdma_cons & PDC_HDMA_Q_MASK;
 
 	/* if nothing on queue, we're done */
@@ -620,7 +620,7 @@ static void pdc20621_dump_hdma(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
 	unsigned int port_no = ap->port_no;
-	struct pdc_host_priv *hpriv = ap->host_set->private_data;
+	struct pdc_host_priv *hpriv = ap->host->private_data;
 	void *dimm_mmio = hpriv->dimm_mmio;
 
 	dimm_mmio += (port_no * PDC_DIMM_WINDOW_STEP);
@@ -638,9 +638,9 @@ static inline void pdc20621_dump_hdma(struct ata_queued_cmd *qc) { }
 static void pdc20621_packet_start(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
-	struct ata_host_set *host_set = ap->host_set;
+	struct ata_host *host = ap->host;
 	unsigned int port_no = ap->port_no;
-	void __iomem *mmio = host_set->mmio_base;
+	void __iomem *mmio = host->mmio_base;
 	unsigned int rw = (qc->tf.flags & ATA_TFLAG_WRITE);
 	u8 seq = (u8) (port_no + 1);
 	unsigned int port_ofs;
@@ -781,8 +781,8 @@ static inline unsigned int pdc20621_host_intr( struct ata_port *ap,
 
 static void pdc20621_irq_clear(struct ata_port *ap)
 {
-	struct ata_host_set *host_set = ap->host_set;
-	void __iomem *mmio = host_set->mmio_base;
+	struct ata_host *host = ap->host;
+	void __iomem *mmio = host->mmio_base;
 
 	mmio += PDC_CHIP0_OFS;
 
@@ -791,7 +791,7 @@ static void pdc20621_irq_clear(struct ata_port *ap)
 
 static irqreturn_t pdc20621_interrupt (int irq, void *dev_instance, struct pt_regs *regs)
 {
-	struct ata_host_set *host_set = dev_instance;
+	struct ata_host *host = dev_instance;
 	struct ata_port *ap;
 	u32 mask = 0;
 	unsigned int i, tmp, port_no;
@@ -800,12 +800,12 @@ static irqreturn_t pdc20621_interrupt (int irq, void *dev_instance, struct pt_re
 
 	VPRINTK("ENTER\n");
 
-	if (!host_set || !host_set->mmio_base) {
+	if (!host || !host->mmio_base) {
 		VPRINTK("QUICK EXIT\n");
 		return IRQ_NONE;
 	}
 
-	mmio_base = host_set->mmio_base;
+	mmio_base = host->mmio_base;
 
 	/* reading should also clear interrupts */
 	mmio_base += PDC_CHIP0_OFS;
@@ -822,16 +822,16 @@ static irqreturn_t pdc20621_interrupt (int irq, void *dev_instance, struct pt_re
 		return IRQ_NONE;
 	}
 
-        spin_lock(&host_set->lock);
+        spin_lock(&host->lock);
 
         for (i = 1; i < 9; i++) {
 		port_no = i - 1;
 		if (port_no > 3)
 			port_no -= 4;
-		if (port_no >= host_set->n_ports)
+		if (port_no >= host->n_ports)
 			ap = NULL;
 		else
-			ap = host_set->ports[port_no];
+			ap = host->ports[port_no];
 		tmp = mask & (1 << i);
 		VPRINTK("seq %u, port_no %u, ap %p, tmp %x\n", i, port_no, ap, tmp);
 		if (tmp && ap &&
@@ -845,7 +845,7 @@ static irqreturn_t pdc20621_interrupt (int irq, void *dev_instance, struct pt_re
 		}
 	}
 
-        spin_unlock(&host_set->lock);
+        spin_unlock(&host->lock);
 
 	VPRINTK("mask == 0x%x\n", mask);
 
@@ -857,13 +857,13 @@ static irqreturn_t pdc20621_interrupt (int irq, void *dev_instance, struct pt_re
 static void pdc_eng_timeout(struct ata_port *ap)
 {
 	u8 drv_stat;
-	struct ata_host_set *host_set = ap->host_set;
+	struct ata_host *host = ap->host;
 	struct ata_queued_cmd *qc;
 	unsigned long flags;
 
 	DPRINTK("ENTER\n");
 
-	spin_lock_irqsave(&host_set->lock, flags);
+	spin_lock_irqsave(&host->lock, flags);
 
 	qc = ata_qc_from_tag(ap, ap->active_tag);
 
@@ -885,7 +885,7 @@ static void pdc_eng_timeout(struct ata_port *ap)
 		break;
 	}
 
-	spin_unlock_irqrestore(&host_set->lock, flags);
+	spin_unlock_irqrestore(&host->lock, flags);
 	ata_eh_qc_complete(qc);
 	DPRINTK("EXIT\n");
 }
@@ -1429,7 +1429,7 @@ static int pdc_sata_init_one (struct pci_dev *pdev, const struct pci_device_id *
 	hpriv->dimm_mmio = dimm_mmio;
 
 	probe_ent->sht		= pdc_port_info[board_idx].sht;
-	probe_ent->host_flags	= pdc_port_info[board_idx].host_flags;
+	probe_ent->port_flags	= pdc_port_info[board_idx].flags;
 	probe_ent->pio_mask	= pdc_port_info[board_idx].pio_mask;
 	probe_ent->mwdma_mask	= pdc_port_info[board_idx].mwdma_mask;
 	probe_ent->udma_mask	= pdc_port_info[board_idx].udma_mask;
diff --git a/drivers/ata/sata_uli.c b/drivers/ata/sata_uli.c
index 654aae2b25c5..8fc6e800011a 100644
--- a/drivers/ata/sata_uli.c
+++ b/drivers/ata/sata_uli.c
@@ -128,7 +128,7 @@ static const struct ata_port_operations uli_ops = {
 
 static struct ata_port_info uli_port_info = {
 	.sht            = &uli_sht,
-	.host_flags     = ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
+	.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
 	.pio_mask       = 0x1f,		/* pio0-4 */
 	.udma_mask      = 0x7f,		/* udma0-6 */
 	.port_ops       = &uli_ops,
@@ -143,13 +143,13 @@ MODULE_VERSION(DRV_VERSION);
 
 static unsigned int get_scr_cfg_addr(struct ata_port *ap, unsigned int sc_reg)
 {
-	struct uli_priv *hpriv = ap->host_set->private_data;
+	struct uli_priv *hpriv = ap->host->private_data;
 	return hpriv->scr_cfg_addr[ap->port_no] + (4 * sc_reg);
 }
 
 static u32 uli_scr_cfg_read (struct ata_port *ap, unsigned int sc_reg)
 {
-	struct pci_dev *pdev = to_pci_dev(ap->host_set->dev);
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
 	unsigned int cfg_addr = get_scr_cfg_addr(ap, sc_reg);
 	u32 val;
 
@@ -159,7 +159,7 @@ static u32 uli_scr_cfg_read (struct ata_port *ap, unsigned int sc_reg)
 
 static void uli_scr_cfg_write (struct ata_port *ap, unsigned int scr, u32 val)
 {
-	struct pci_dev *pdev = to_pci_dev(ap->host_set->dev);
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
 	unsigned int cfg_addr = get_scr_cfg_addr(ap, scr);
 
 	pci_write_config_dword(pdev, cfg_addr, val);
diff --git a/drivers/ata/sata_via.c b/drivers/ata/sata_via.c
index a0699a1728d4..7f087aef99de 100644
--- a/drivers/ata/sata_via.c
+++ b/drivers/ata/sata_via.c
@@ -176,7 +176,7 @@ static const struct ata_port_operations vt6421_sata_ops = {
 
 static struct ata_port_info vt6420_port_info = {
 	.sht		= &svia_sht,
-	.host_flags	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
+	.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
 	.pio_mask	= 0x1f,
 	.mwdma_mask	= 0x07,
 	.udma_mask	= 0x7f,
@@ -346,7 +346,7 @@ static struct ata_probe_ent *vt6421_init_probe_ent(struct pci_dev *pdev)
 	INIT_LIST_HEAD(&probe_ent->node);
 
 	probe_ent->sht		= &svia_sht;
-	probe_ent->host_flags	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY;
+	probe_ent->port_flags	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY;
 	probe_ent->port_ops	= &vt6421_sata_ops;
 	probe_ent->n_ports	= N_PORTS;
 	probe_ent->irq		= pdev->irq;
diff --git a/drivers/ata/sata_vsc.c b/drivers/ata/sata_vsc.c
index 4c69a705a483..d0d92f33de54 100644
--- a/drivers/ata/sata_vsc.c
+++ b/drivers/ata/sata_vsc.c
@@ -123,7 +123,7 @@ static void vsc_intr_mask_update(struct ata_port *ap, u8 ctl)
 	void __iomem *mask_addr;
 	u8 mask;
 
-	mask_addr = ap->host_set->mmio_base +
+	mask_addr = ap->host->mmio_base +
 		VSC_SATA_INT_MASK_OFFSET + ap->port_no;
 	mask = readb(mask_addr);
 	if (ctl & ATA_NIEN)
@@ -206,20 +206,20 @@ static void vsc_sata_tf_read(struct ata_port *ap, struct ata_taskfile *tf)
 static irqreturn_t vsc_sata_interrupt (int irq, void *dev_instance,
 				       struct pt_regs *regs)
 {
-	struct ata_host_set *host_set = dev_instance;
+	struct ata_host *host = dev_instance;
 	unsigned int i;
 	unsigned int handled = 0;
 	u32 int_status;
 
-	spin_lock(&host_set->lock);
+	spin_lock(&host->lock);
 
-	int_status = readl(host_set->mmio_base + VSC_SATA_INT_STAT_OFFSET);
+	int_status = readl(host->mmio_base + VSC_SATA_INT_STAT_OFFSET);
 
-	for (i = 0; i < host_set->n_ports; i++) {
+	for (i = 0; i < host->n_ports; i++) {
 		if (int_status & ((u32) 0xFF << (8 * i))) {
 			struct ata_port *ap;
 
-			ap = host_set->ports[i];
+			ap = host->ports[i];
 
 			if (is_vsc_sata_int_err(i, int_status)) {
 				u32 err_status;
@@ -259,7 +259,7 @@ static irqreturn_t vsc_sata_interrupt (int irq, void *dev_instance,
 		}
 	}
 
-	spin_unlock(&host_set->lock);
+	spin_unlock(&host->lock);
 
 	return IRQ_RETVAL(handled);
 }
@@ -395,7 +395,7 @@ static int __devinit vsc_sata_init_one (struct pci_dev *pdev, const struct pci_d
 	pci_write_config_byte(pdev, PCI_CACHE_LINE_SIZE, 0x80);
 
 	probe_ent->sht = &vsc_sata_sht;
-	probe_ent->host_flags = ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
+	probe_ent->port_flags = ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
 				ATA_FLAG_MMIO;
 	probe_ent->port_ops = &vsc_sata_ops;
 	probe_ent->n_ports = 4;
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 806682603ac5..563885cb0995 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -197,7 +197,7 @@ enum {
 	ATA_QCFLAG_EH_SCHEDULED = (1 << 18), /* EH scheduled (obsolete) */
 
 	/* host set flags */
-	ATA_HOST_SIMPLEX	= (1 << 0),	/* Host is simplex, one DMA channel per host_set only */
+	ATA_HOST_SIMPLEX	= (1 << 0),	/* Host is simplex, one DMA channel per host only */
 	
 	/* various lengths of time */
 	ATA_TMOUT_BOOT		= 30 * HZ,	/* heuristic */
@@ -357,13 +357,13 @@ struct ata_probe_ent {
 	unsigned long		irq;
 	unsigned long		irq2;
 	unsigned int		irq_flags;
-	unsigned long		host_flags;
-	unsigned long		host_set_flags;
+	unsigned long		port_flags;
+	unsigned long		_host_flags;
 	void __iomem		*mmio_base;
 	void			*private_data;
 };
 
-struct ata_host_set {
+struct ata_host {
 	spinlock_t		lock;
 	struct device 		*dev;
 	unsigned long		irq;
@@ -420,7 +420,7 @@ struct ata_queued_cmd {
 	void			*private_data;
 };
 
-struct ata_host_stats {
+struct ata_port_stats {
 	unsigned long		unhandled_irq;
 	unsigned long		idle_irq;
 	unsigned long		rw_reqbuf;
@@ -498,7 +498,7 @@ struct ata_eh_context {
 };
 
 struct ata_port {
-	struct Scsi_Host	*host;	/* our co-allocated scsi host */
+	struct Scsi_Host	*scsi_host; /* our co-allocated scsi host */
 	const struct ata_port_operations *ops;
 	spinlock_t		*lock;
 	unsigned long		flags;	/* ATA_FLAG_xxx */
@@ -523,7 +523,7 @@ struct ata_port {
 	unsigned int		hw_sata_spd_limit;
 	unsigned int		sata_spd_limit;	/* SATA PHY speed limit */
 
-	/* record runtime error info, protected by host_set lock */
+	/* record runtime error info, protected by host lock */
 	struct ata_eh_info	eh_info;
 	/* EH context owned by EH */
 	struct ata_eh_context	eh_context;
@@ -537,8 +537,8 @@ struct ata_port {
 	unsigned int		active_tag;
 	u32			sactive;
 
-	struct ata_host_stats	stats;
-	struct ata_host_set	*host_set;
+	struct ata_port_stats	stats;
+	struct ata_host		*host;
 	struct device 		*dev;
 
 	struct work_struct	port_task;
@@ -614,7 +614,7 @@ struct ata_port_operations {
 	int (*port_start) (struct ata_port *ap);
 	void (*port_stop) (struct ata_port *ap);
 
-	void (*host_stop) (struct ata_host_set *host_set);
+	void (*host_stop) (struct ata_host *host);
 
 	void (*bmdma_stop) (struct ata_queued_cmd *qc);
 	u8   (*bmdma_status) (struct ata_port *ap);
@@ -622,7 +622,7 @@ struct ata_port_operations {
 
 struct ata_port_info {
 	struct scsi_host_template	*sht;
-	unsigned long		host_flags;
+	unsigned long		flags;
 	unsigned long		pio_mask;
 	unsigned long		mwdma_mask;
 	unsigned long		udma_mask;
@@ -690,15 +690,15 @@ extern int ata_pci_clear_simplex(struct pci_dev *pdev);
 #endif /* CONFIG_PCI */
 extern int ata_device_add(const struct ata_probe_ent *ent);
 extern void ata_port_detach(struct ata_port *ap);
-extern void ata_host_set_init(struct ata_host_set *, struct device *,
-			      unsigned long, const struct ata_port_operations *);
-extern void ata_host_set_remove(struct ata_host_set *host_set);
+extern void ata_host_init(struct ata_host *, struct device *,
+			  unsigned long, const struct ata_port_operations *);
+extern void ata_host_remove(struct ata_host *host);
 extern int ata_scsi_detect(struct scsi_host_template *sht);
 extern int ata_scsi_ioctl(struct scsi_device *dev, int cmd, void __user *arg);
 extern int ata_scsi_queuecmd(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *));
 extern int ata_scsi_release(struct Scsi_Host *host);
 extern void ata_sas_port_destroy(struct ata_port *);
-extern struct ata_port *ata_sas_port_alloc(struct ata_host_set *,
+extern struct ata_port *ata_sas_port_alloc(struct ata_host *,
 					   struct ata_port_info *, struct Scsi_Host *);
 extern int ata_sas_port_init(struct ata_port *);
 extern int ata_sas_port_start(struct ata_port *ap);
@@ -715,9 +715,8 @@ extern int ata_port_online(struct ata_port *ap);
 extern int ata_port_offline(struct ata_port *ap);
 extern int ata_scsi_device_resume(struct scsi_device *);
 extern int ata_scsi_device_suspend(struct scsi_device *, pm_message_t mesg);
-extern int ata_host_set_suspend(struct ata_host_set *host_set,
-				pm_message_t mesg);
-extern void ata_host_set_resume(struct ata_host_set *host_set);
+extern int ata_host_suspend(struct ata_host *host, pm_message_t mesg);
+extern void ata_host_resume(struct ata_host *host);
 extern int ata_ratelimit(void);
 extern unsigned int ata_busy_sleep(struct ata_port *ap,
 				   unsigned long timeout_pat,
@@ -742,7 +741,7 @@ extern u8 ata_altstatus(struct ata_port *ap);
 extern void ata_exec_command(struct ata_port *ap, const struct ata_taskfile *tf);
 extern int ata_port_start (struct ata_port *ap);
 extern void ata_port_stop (struct ata_port *ap);
-extern void ata_host_stop (struct ata_host_set *host_set);
+extern void ata_host_stop (struct ata_host *host);
 extern irqreturn_t ata_interrupt (int irq, void *dev_instance, struct pt_regs *regs);
 extern void ata_mmio_data_xfer(struct ata_device *adev, unsigned char *buf,
 			       unsigned int buflen, int write_data);
@@ -828,7 +827,7 @@ struct pci_bits {
 	unsigned long		val;
 };
 
-extern void ata_pci_host_stop (struct ata_host_set *host_set);
+extern void ata_pci_host_stop (struct ata_host *host);
 extern struct ata_probe_ent *
 ata_pci_init_native_mode(struct pci_dev *pdev, struct ata_port_info **port, int portmask);
 extern int pci_test_config_bits(struct pci_dev *pdev, const struct pci_bits *bits);
-- 
cgit v1.2.3


From f39b7a55a84e34e3074b168e30dc73b66e85261d Mon Sep 17 00:00:00 2001
From: Olof Johansson <olof@lixom.net>
Date: Fri, 11 Aug 2006 00:07:08 -0500
Subject: [POWERPC] Cleanup CPU inits

Cleanup CPU inits a bit more, Geoff Levand already did some earlier.

* Move CPU state save to cpu_setup, since cpu_setup is only ever done
  on cpu 0 on 64-bit and save is never done more than once.
* Rename __restore_cpu_setup to __restore_cpu_ppc970 and add
  function pointers to the cputable to use instead. Powermac always
  has 970 so no need to check there.
* Rename __970_cpu_preinit to __cpu_preinit_ppc970 and check PVR before
  calling it instead of in it, it's too early to use cputable.
* Rename pSeries_secondary_smp_init to generic_secondary_smp_init since
  everyone but powermac and iSeries use it.

Signed-off-by: Olof Johansson <olof@lixom.net>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 arch/powerpc/kernel/asm-offsets.c      |  1 +
 arch/powerpc/kernel/cpu_setup_ppc970.S | 99 +++++++++-------------------------
 arch/powerpc/kernel/cputable.c         |  6 +++
 arch/powerpc/kernel/head_64.S          | 52 +++++++++++-------
 arch/powerpc/platforms/cell/smp.c      |  4 +-
 arch/powerpc/platforms/pseries/smp.c   |  4 +-
 include/asm-powerpc/cputable.h         |  3 ++
 7 files changed, 71 insertions(+), 98 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c
index a2f95e467a75..c53acd2a6dfc 100644
--- a/arch/powerpc/kernel/asm-offsets.c
+++ b/arch/powerpc/kernel/asm-offsets.c
@@ -246,6 +246,7 @@ int main(void)
 	DEFINE(CPU_SPEC_PVR_VALUE, offsetof(struct cpu_spec, pvr_value));
 	DEFINE(CPU_SPEC_FEATURES, offsetof(struct cpu_spec, cpu_features));
 	DEFINE(CPU_SPEC_SETUP, offsetof(struct cpu_spec, cpu_setup));
+	DEFINE(CPU_SPEC_RESTORE, offsetof(struct cpu_spec, cpu_restore));
 
 #ifndef CONFIG_PPC64
 	DEFINE(pbe_address, offsetof(struct pbe, address));
diff --git a/arch/powerpc/kernel/cpu_setup_ppc970.S b/arch/powerpc/kernel/cpu_setup_ppc970.S
index f69af2c5d7b3..f619932794e8 100644
--- a/arch/powerpc/kernel/cpu_setup_ppc970.S
+++ b/arch/powerpc/kernel/cpu_setup_ppc970.S
@@ -16,27 +16,12 @@
 #include <asm/asm-offsets.h>
 #include <asm/cache.h>
 
-_GLOBAL(__970_cpu_preinit)
-	/*
-	 * Do nothing if not running in HV mode
-	 */
+_GLOBAL(__cpu_preinit_ppc970)
+	/* Do nothing if not running in HV mode */
 	mfmsr	r0
 	rldicl.	r0,r0,4,63
 	beqlr
 
-	/*
-	 * Deal only with PPC970 and PPC970FX.
-	 */
-	mfspr	r0,SPRN_PVR
-	srwi	r0,r0,16
-	cmpwi	r0,0x39
-	beq	1f
-	cmpwi	r0,0x3c
-	beq	1f
-	cmpwi	r0,0x44
-	bnelr
-1:
-
 	/* Make sure HID4:rm_ci is off before MMU is turned off, that large
 	 * pages are enabled with HID4:61 and clear HID5:DCBZ_size and
 	 * HID5:DCBZ32_ill
@@ -72,21 +57,6 @@ _GLOBAL(__970_cpu_preinit)
 	isync
 	blr
 
-_GLOBAL(__setup_cpu_ppc970)
-	mfspr	r0,SPRN_HID0
-	li	r11,5			/* clear DOZE and SLEEP */
-	rldimi	r0,r11,52,8		/* set NAP and DPM */
-	mtspr	SPRN_HID0,r0
-	mfspr	r0,SPRN_HID0
-	mfspr	r0,SPRN_HID0
-	mfspr	r0,SPRN_HID0
-	mfspr	r0,SPRN_HID0
-	mfspr	r0,SPRN_HID0
-	mfspr	r0,SPRN_HID0
-	sync
-	isync
-	blr
-
 /* Definitions for the table use to save CPU states */
 #define CS_HID0		0
 #define CS_HID1		8
@@ -101,33 +71,28 @@ cpu_state_storage:
 	.balign	L1_CACHE_BYTES,0
 	.text
 
-/* Called in normal context to backup CPU 0 state. This
- * does not include cache settings. This function is also
- * called for machine sleep. This does not include the MMU
- * setup, BATs, etc... but rather the "special" registers
- * like HID0, HID1, HID4, etc...
- */
-_GLOBAL(__save_cpu_setup)
-	/* Some CR fields are volatile, we back it up all */
-	mfcr	r7
-
-	/* Get storage ptr */
-	LOAD_REG_IMMEDIATE(r5,cpu_state_storage)
 
-	/* We only deal with 970 for now */
-	mfspr	r0,SPRN_PVR
-	srwi	r0,r0,16
-	cmpwi	r0,0x39
-	beq	1f
-	cmpwi	r0,0x3c
-	beq	1f
-	cmpwi	r0,0x44
-	bne	2f
-
-1:	/* skip if not running in HV mode */
+_GLOBAL(__setup_cpu_ppc970)
+	/* Do nothing if not running in HV mode */
 	mfmsr	r0
 	rldicl.	r0,r0,4,63
-	beq	2f
+	beqlr
+
+	mfspr	r0,SPRN_HID0
+	li	r11,5			/* clear DOZE and SLEEP */
+	rldimi	r0,r11,52,8		/* set NAP and DPM */
+	mtspr	SPRN_HID0,r0
+	mfspr	r0,SPRN_HID0
+	mfspr	r0,SPRN_HID0
+	mfspr	r0,SPRN_HID0
+	mfspr	r0,SPRN_HID0
+	mfspr	r0,SPRN_HID0
+	mfspr	r0,SPRN_HID0
+	sync
+	isync
+
+	/* Save away cpu state */
+	LOAD_REG_IMMEDIATE(r5,cpu_state_storage)
 
 	/* Save HID0,1,4 and 5 */
 	mfspr	r3,SPRN_HID0
@@ -139,35 +104,19 @@ _GLOBAL(__save_cpu_setup)
 	mfspr	r3,SPRN_HID5
 	std	r3,CS_HID5(r5)
 
-2:
-	mtcr	r7
 	blr
 
 /* Called with no MMU context (typically MSR:IR/DR off) to
  * restore CPU state as backed up by the previous
  * function. This does not include cache setting
  */
-_GLOBAL(__restore_cpu_setup)
-	/* Get storage ptr (FIXME when using anton reloc as we
-	 * are running with translation disabled here
-	 */
-	LOAD_REG_IMMEDIATE(r5,cpu_state_storage)
-
-	/* We only deal with 970 for now */
-	mfspr	r0,SPRN_PVR
-	srwi	r0,r0,16
-	cmpwi	r0,0x39
-	beq	1f
-	cmpwi	r0,0x3c
-	beq	1f
-	cmpwi	r0,0x44
-	bnelr
-
-1:	/* skip if not running in HV mode */
+_GLOBAL(__restore_cpu_ppc970)
+	/* Do nothing if not running in HV mode */
 	mfmsr	r0
 	rldicl.	r0,r0,4,63
 	beqlr
 
+	LOAD_REG_IMMEDIATE(r5,cpu_state_storage)
 	/* Before accessing memory, we make sure rm_ci is clear */
 	li	r0,0
 	mfspr	r3,SPRN_HID4
diff --git a/arch/powerpc/kernel/cputable.c b/arch/powerpc/kernel/cputable.c
index 272e43622fd6..306da4cd37a0 100644
--- a/arch/powerpc/kernel/cputable.c
+++ b/arch/powerpc/kernel/cputable.c
@@ -39,7 +39,10 @@ extern void __setup_cpu_7400(unsigned long offset, struct cpu_spec* spec);
 extern void __setup_cpu_7410(unsigned long offset, struct cpu_spec* spec);
 extern void __setup_cpu_745x(unsigned long offset, struct cpu_spec* spec);
 #endif /* CONFIG_PPC32 */
+#ifdef CONFIG_PPC64
 extern void __setup_cpu_ppc970(unsigned long offset, struct cpu_spec* spec);
+extern void __restore_cpu_ppc970(void);
+#endif /* CONFIG_PPC64 */
 
 /* This table only contains "desktop" CPUs, it need to be filled with embedded
  * ones as well...
@@ -184,6 +187,7 @@ struct cpu_spec	cpu_specs[] = {
 		.dcache_bsize		= 128,
 		.num_pmcs		= 8,
 		.cpu_setup		= __setup_cpu_ppc970,
+		.cpu_restore		= __restore_cpu_ppc970,
 		.oprofile_cpu_type	= "ppc64/970",
 		.oprofile_type		= PPC_OPROFILE_POWER4,
 		.platform		= "ppc970",
@@ -199,6 +203,7 @@ struct cpu_spec	cpu_specs[] = {
 		.dcache_bsize		= 128,
 		.num_pmcs		= 8,
 		.cpu_setup		= __setup_cpu_ppc970,
+		.cpu_restore		= __restore_cpu_ppc970,
 		.oprofile_cpu_type	= "ppc64/970",
 		.oprofile_type		= PPC_OPROFILE_POWER4,
 		.platform		= "ppc970",
@@ -214,6 +219,7 @@ struct cpu_spec	cpu_specs[] = {
 		.dcache_bsize		= 128,
 		.num_pmcs		= 8,
 		.cpu_setup		= __setup_cpu_ppc970,
+		.cpu_restore		= __restore_cpu_ppc970,
 		.oprofile_cpu_type	= "ppc64/970",
 		.oprofile_type		= PPC_OPROFILE_POWER4,
 		.platform		= "ppc970",
diff --git a/arch/powerpc/kernel/head_64.S b/arch/powerpc/kernel/head_64.S
index 6ff3cf506088..e9963d9f335a 100644
--- a/arch/powerpc/kernel/head_64.S
+++ b/arch/powerpc/kernel/head_64.S
@@ -132,7 +132,7 @@ _GLOBAL(__secondary_hold)
 	bne	100b
 
 #if defined(CONFIG_SMP) || defined(CONFIG_KEXEC)
-	LOAD_REG_IMMEDIATE(r4, .pSeries_secondary_smp_init)
+	LOAD_REG_IMMEDIATE(r4, .generic_secondary_smp_init)
 	mtctr	r4
 	mr	r3,r24
 	bctr
@@ -1484,19 +1484,17 @@ fwnmi_data_area:
         . = 0x8000
 
 /*
- * On pSeries, secondary processors spin in the following code.
+ * On pSeries and most other platforms, secondary processors spin
+ * in the following code.
  * At entry, r3 = this processor's number (physical cpu id)
  */
-_GLOBAL(pSeries_secondary_smp_init)
+_GLOBAL(generic_secondary_smp_init)
 	mr	r24,r3
 	
 	/* turn on 64-bit mode */
 	bl	.enable_64b_mode
 	isync
 
-	/* Copy some CPU settings from CPU 0 */
-	bl	.__restore_cpu_setup
-
 	/* Set up a paca value for this processor. Since we have the
 	 * physical cpu id in r24, we need to search the pacas to find
 	 * which logical id maps to our physical one.
@@ -1522,15 +1520,28 @@ _GLOBAL(pSeries_secondary_smp_init)
 					/* start.			 */
 	sync
 
-	/* Create a temp kernel stack for use before relocation is on.	*/
+#ifndef CONFIG_SMP
+	b	3b			/* Never go on non-SMP		 */
+#else
+	cmpwi	0,r23,0
+	beq	3b			/* Loop until told to go	 */
+
+	/* See if we need to call a cpu state restore handler */
+	LOAD_REG_IMMEDIATE(r23, cur_cpu_spec)
+	ld	r23,0(r23)
+	ld	r23,CPU_SPEC_RESTORE(r23)
+	cmpdi	0,r23,0
+	beq	4f
+	ld	r23,0(r23)
+	mtctr	r23
+	bctrl
+
+4:	/* Create a temp kernel stack for use before relocation is on.	*/
 	ld	r1,PACAEMERGSP(r13)
 	subi	r1,r1,STACK_FRAME_OVERHEAD
 
-	cmpwi	0,r23,0
-#ifdef CONFIG_SMP
-	bne	.__secondary_start
+	b	.__secondary_start
 #endif
-	b 	3b			/* Loop until told to go	 */
 
 #ifdef CONFIG_PPC_ISERIES
 _STATIC(__start_initialization_iSeries)
@@ -1611,7 +1622,16 @@ _GLOBAL(__start_initialization_multiplatform)
 	bl	.enable_64b_mode
 
 	/* Setup some critical 970 SPRs before switching MMU off */
-	bl	.__970_cpu_preinit
+	mfspr	r0,SPRN_PVR
+	srwi	r0,r0,16
+	cmpwi	r0,0x39		/* 970 */
+	beq	1f
+	cmpwi	r0,0x3c		/* 970FX */
+	beq	1f
+	cmpwi	r0,0x44		/* 970MP */
+	bne	2f
+1:	bl	.__cpu_preinit_ppc970
+2:
 
 	/* Switch off MMU if not already */
 	LOAD_REG_IMMEDIATE(r4, .__after_prom_start - KERNELBASE)
@@ -1782,7 +1802,7 @@ _GLOBAL(pmac_secondary_start)
 	isync
 
 	/* Copy some CPU settings from CPU 0 */
-	bl	.__restore_cpu_setup
+	bl	.__restore_cpu_ppc970
 
 	/* pSeries do that early though I don't think we really need it */
 	mfmsr	r3
@@ -1932,12 +1952,6 @@ _STATIC(start_here_multiplatform)
 	mr	r5,r26
 	bl	.identify_cpu
 
-	/* Save some low level config HIDs of CPU0 to be copied to
-	 * other CPUs later on, or used for suspend/resume
-	 */
-	bl	.__save_cpu_setup
-	sync
-
 	/* Do very early kernel initializations, including initial hash table,
 	 * stab and slb setup before we turn on relocation.	*/
 
diff --git a/arch/powerpc/platforms/cell/smp.c b/arch/powerpc/platforms/cell/smp.c
index 46aef0640742..1c0acbad7425 100644
--- a/arch/powerpc/platforms/cell/smp.c
+++ b/arch/powerpc/platforms/cell/smp.c
@@ -57,7 +57,7 @@
  */
 static cpumask_t of_spin_map;
 
-extern void pSeries_secondary_smp_init(unsigned long);
+extern void generic_secondary_smp_init(unsigned long);
 
 /**
  * smp_startup_cpu() - start the given cpu
@@ -74,7 +74,7 @@ static inline int __devinit smp_startup_cpu(unsigned int lcpu)
 {
 	int status;
 	unsigned long start_here = __pa((u32)*((unsigned long *)
-					       pSeries_secondary_smp_init));
+					       generic_secondary_smp_init));
 	unsigned int pcpu;
 	int start_cpu;
 
diff --git a/arch/powerpc/platforms/pseries/smp.c b/arch/powerpc/platforms/pseries/smp.c
index f39dad8b99e0..c6624b8a0e77 100644
--- a/arch/powerpc/platforms/pseries/smp.c
+++ b/arch/powerpc/platforms/pseries/smp.c
@@ -62,7 +62,7 @@
  */
 static cpumask_t of_spin_map;
 
-extern void pSeries_secondary_smp_init(unsigned long);
+extern void generic_secondary_smp_init(unsigned long);
 
 #ifdef CONFIG_HOTPLUG_CPU
 
@@ -270,7 +270,7 @@ static inline int __devinit smp_startup_cpu(unsigned int lcpu)
 {
 	int status;
 	unsigned long start_here = __pa((u32)*((unsigned long *)
-					       pSeries_secondary_smp_init));
+					       generic_secondary_smp_init));
 	unsigned int pcpu;
 	int start_cpu;
 
diff --git a/include/asm-powerpc/cputable.h b/include/asm-powerpc/cputable.h
index 1ba3c9983614..748bc1805da9 100644
--- a/include/asm-powerpc/cputable.h
+++ b/include/asm-powerpc/cputable.h
@@ -36,6 +36,7 @@
 struct cpu_spec;
 
 typedef	void (*cpu_setup_t)(unsigned long offset, struct cpu_spec* spec);
+typedef	void (*cpu_restore_t)(void);
 
 enum powerpc_oprofile_type {
 	PPC_OPROFILE_INVALID = 0,
@@ -65,6 +66,8 @@ struct cpu_spec {
 	 * BHT, SPD, etc... from head.S before branching to identify_machine
 	 */
 	cpu_setup_t	cpu_setup;
+	/* Used to restore cpu setup on secondary processors and at resume */
+	cpu_restore_t	cpu_restore;
 
 	/* Used by oprofile userspace to select the right counters */
 	char		*oprofile_cpu_type;
-- 
cgit v1.2.3


From 869d7f381e8c32de85ddfa9621125fb10a885f87 Mon Sep 17 00:00:00 2001
From: Jon Loeliger <jdl@freescale.com>
Date: Tue, 15 Aug 2006 16:19:02 -0500
Subject: [POWERPC] Allow MPC8641 HPCN to build with CONFIG_PCI disabled too.

Signed-off-by: Jon Loeliger <jdl@freescale.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 arch/powerpc/platforms/86xx/mpc86xx_hpcn.c | 9 +++++++--
 arch/powerpc/platforms/86xx/pci.c          | 3 ++-
 include/asm-powerpc/mpc86xx.h              | 2 --
 3 files changed, 9 insertions(+), 5 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/platforms/86xx/mpc86xx_hpcn.c b/arch/powerpc/platforms/86xx/mpc86xx_hpcn.c
index 4a33d95e7ad7..496cc7c3a54c 100644
--- a/arch/powerpc/platforms/86xx/mpc86xx_hpcn.c
+++ b/arch/powerpc/platforms/86xx/mpc86xx_hpcn.c
@@ -52,6 +52,7 @@ unsigned long pci_dram_offset = 0;
 #endif
 
 
+#ifdef CONFIG_PCI
 static void mpc86xx_8259_cascade(unsigned int irq, struct irq_desc *desc,
 				 struct pt_regs *regs)
 {
@@ -60,14 +61,18 @@ static void mpc86xx_8259_cascade(unsigned int irq, struct irq_desc *desc,
 		generic_handle_irq(cascade_irq, regs);
 	desc->chip->eoi(irq);
 }
+#endif	/* CONFIG_PCI */
 
 void __init
 mpc86xx_hpcn_init_irq(void)
 {
 	struct mpic *mpic1;
-	struct device_node *np, *cascade_node = NULL;
-	int cascade_irq;
+	struct device_node *np;
 	phys_addr_t openpic_paddr;
+#ifdef CONFIG_PCI
+	struct device_node *cascade_node = NULL;
+	int cascade_irq;
+#endif
 
 	np = of_find_node_by_type(NULL, "open-pic");
 	if (np == NULL)
diff --git a/arch/powerpc/platforms/86xx/pci.c b/arch/powerpc/platforms/86xx/pci.c
index d7050c1108ff..481e18ed5be9 100644
--- a/arch/powerpc/platforms/86xx/pci.c
+++ b/arch/powerpc/platforms/86xx/pci.c
@@ -188,7 +188,8 @@ int __init add_bridge(struct device_node *dev)
 
 	printk(KERN_INFO "Found MPC86xx PCIE host bridge at 0x%08lx. "
 	       "Firmware bus number: %d->%d\n",
-		rsrc.start, hose->first_busno, hose->last_busno);
+	       (unsigned long) rsrc.start,
+	       hose->first_busno, hose->last_busno);
 
 	DBG(" ->Hose at 0x%p, cfg_addr=0x%p,cfg_data=0x%p\n",
 		hose, hose->cfg_addr, hose->cfg_data);
diff --git a/include/asm-powerpc/mpc86xx.h b/include/asm-powerpc/mpc86xx.h
index f260382739fa..2d6ad859df7f 100644
--- a/include/asm-powerpc/mpc86xx.h
+++ b/include/asm-powerpc/mpc86xx.h
@@ -23,8 +23,6 @@
 #define _ISA_MEM_BASE   isa_mem_base
 #ifdef CONFIG_PCI
 #define PCI_DRAM_OFFSET pci_dram_offset
-#else
-#define PCI_DRAM_OFFSET 0
 #endif
 
 #define CPU0_BOOT_RELEASE 0x01000000
-- 
cgit v1.2.3


From 6f3d5d3cc4b1447578ae8484166bbc34a64150c5 Mon Sep 17 00:00:00 2001
From: Michael Ellerman <michael@ellerman.id.au>
Date: Wed, 16 Aug 2006 22:04:14 +1000
Subject: [POWERPC] Add a helper for calculating RTAS "config_addr" parameters

Several RTAS calls take a "config_addr" parameter, which is a particular
way of specifying a PCI busno, devfn and register number into a 32-bit word.
Currently these are open-coded, and I'll be adding another soon, replace
them with a helper that encapsulates the logic. Be more strict about masking
the busno too, just in case.

Booted on P5 LPAR.

Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 arch/powerpc/kernel/rtas_pci.c |  6 ++----
 include/asm-powerpc/rtas.h     | 16 ++++++++++++++++
 2 files changed, 18 insertions(+), 4 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/kernel/rtas_pci.c b/arch/powerpc/kernel/rtas_pci.c
index 5a798ac6aecf..b4a0de79c060 100644
--- a/arch/powerpc/kernel/rtas_pci.c
+++ b/arch/powerpc/kernel/rtas_pci.c
@@ -81,8 +81,7 @@ int rtas_read_config(struct pci_dn *pdn, int where, int size, u32 *val)
 	if (!config_access_valid(pdn, where))
 		return PCIBIOS_BAD_REGISTER_NUMBER;
 
-	addr = ((where & 0xf00) << 20) | (pdn->busno << 16) |
-		(pdn->devfn << 8) | (where & 0xff);
+	addr = rtas_config_addr(pdn->busno, pdn->devfn, where);
 	buid = pdn->phb->buid;
 	if (buid) {
 		ret = rtas_call(ibm_read_pci_config, 4, 2, &returnval,
@@ -134,8 +133,7 @@ int rtas_write_config(struct pci_dn *pdn, int where, int size, u32 val)
 	if (!config_access_valid(pdn, where))
 		return PCIBIOS_BAD_REGISTER_NUMBER;
 
-	addr = ((where & 0xf00) << 20) | (pdn->busno << 16) |
-		(pdn->devfn << 8) | (where & 0xff);
+	addr = rtas_config_addr(pdn->busno, pdn->devfn, where);
 	buid = pdn->phb->buid;
 	if (buid) {
 		ret = rtas_call(ibm_write_pci_config, 5, 1, NULL, addr,
diff --git a/include/asm-powerpc/rtas.h b/include/asm-powerpc/rtas.h
index 82a27e9a041f..d34f9e1f242c 100644
--- a/include/asm-powerpc/rtas.h
+++ b/include/asm-powerpc/rtas.h
@@ -230,5 +230,21 @@ extern unsigned long rtas_rmo_buf;
 
 #define GLOBAL_INTERRUPT_QUEUE 9005
 
+/**
+ * rtas_config_addr - Format a busno, devfn and reg for RTAS.
+ * @busno: The bus number.
+ * @devfn: The device and function number as encoded by PCI_DEVFN().
+ * @reg: The register number.
+ *
+ * This function encodes the given busno, devfn and register number as
+ * required for RTAS calls that take a "config_addr" parameter.
+ * See PAPR requirement 7.3.4-1 for more info.
+ */
+static inline u32 rtas_config_addr(int busno, int devfn, int reg)
+{
+	return ((reg & 0xf00) << 20) | ((busno & 0xff) << 16) |
+			(devfn << 8) | (reg & 0xff);
+}
+
 #endif /* __KERNEL__ */
 #endif /* _POWERPC_RTAS_H */
-- 
cgit v1.2.3


From f4ad7b5807385ad1fed0347d966e51a797cd1013 Mon Sep 17 00:00:00 2001
From: James Bottomley <James.Bottomley@steeleye.com>
Date: Fri, 25 Aug 2006 13:48:18 -0500
Subject: [SCSI] scsi_transport_sas: remove local_attached flag

This flag denotes local attachment of the phy.  There are two problems
with it:

1) It's actually redundant ... you can get the same information simply
by seeing whether a host is the phys parent
2) we condition a lot of phy parameters on it on the false assumption
that we can only control local phys.  I'm wiring up phy resets in the
aic94xx now, and it will be able to reset non-local phys as well.

I fixed 2) by moving the local check into the reset and stats function
of the mptsas, since that seems to be the only HBA that can't
(currently) control non-local phys.

Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 drivers/message/fusion/mptsas.c   | 11 ++++++++---
 drivers/scsi/scsi_transport_sas.c | 10 ++--------
 include/scsi/scsi_transport_sas.h |  5 ++---
 3 files changed, 12 insertions(+), 14 deletions(-)

(limited to 'include')

diff --git a/drivers/message/fusion/mptsas.c b/drivers/message/fusion/mptsas.c
index dfdd1e445768..b752a479f6db 100644
--- a/drivers/message/fusion/mptsas.c
+++ b/drivers/message/fusion/mptsas.c
@@ -852,6 +852,10 @@ static int mptsas_get_linkerrors(struct sas_phy *phy)
 	dma_addr_t dma_handle;
 	int error;
 
+	/* FIXME: only have link errors on local phys */
+	if (!scsi_is_sas_phy_local(phy))
+		return -EINVAL;
+
 	hdr.PageVersion = MPI_SASPHY1_PAGEVERSION;
 	hdr.ExtPageLength = 0;
 	hdr.PageNumber = 1 /* page number 1*/;
@@ -924,6 +928,10 @@ static int mptsas_phy_reset(struct sas_phy *phy, int hard_reset)
 	unsigned long timeleft;
 	int error = -ERESTARTSYS;
 
+	/* FIXME: fusion doesn't allow non-local phy reset */
+	if (!scsi_is_sas_phy_local(phy))
+		return -EINVAL;
+
 	/* not implemented for expanders */
 	if (phy->identify.target_port_protocols & SAS_PROTOCOL_SMP)
 		return -ENXIO;
@@ -1570,9 +1578,6 @@ static int mptsas_probe_one_phy(struct device *dev,
 
 	if (!phy_info->phy) {
 
-		if (local)
-			phy->local_attached = 1;
-
 		error = sas_phy_add(phy);
 		if (error) {
 			sas_phy_free(phy);
diff --git a/drivers/scsi/scsi_transport_sas.c b/drivers/scsi/scsi_transport_sas.c
index 5a625c3fddae..d518c1207fb4 100644
--- a/drivers/scsi/scsi_transport_sas.c
+++ b/drivers/scsi/scsi_transport_sas.c
@@ -266,9 +266,6 @@ show_sas_phy_##field(struct class_device *cdev, char *buf)		\
 	struct sas_internal *i = to_sas_internal(shost->transportt);	\
 	int error;							\
 									\
-	if (!phy->local_attached)					\
-		return -EINVAL;						\
-									\
 	error = i->f->get_linkerrors ? i->f->get_linkerrors(phy) : 0;	\
 	if (error)							\
 		return error;						\
@@ -299,9 +296,6 @@ static ssize_t do_sas_phy_reset(struct class_device *cdev,
 	struct sas_internal *i = to_sas_internal(shost->transportt);
 	int error;
 
-	if (!phy->local_attached)
-		return -EINVAL;
-
 	error = i->f->phy_reset(phy, hard_reset);
 	if (error)
 		return error;
@@ -849,7 +843,7 @@ show_sas_rphy_enclosure_identifier(struct class_device *cdev, char *buf)
 	 * Only devices behind an expander are supported, because the
 	 * enclosure identifier is a SMP feature.
 	 */
-	if (phy->local_attached)
+	if (scsi_is_sas_phy_local(phy))
 		return -EINVAL;
 
 	error = i->f->get_enclosure_identifier(rphy, &identifier);
@@ -870,7 +864,7 @@ show_sas_rphy_bay_identifier(struct class_device *cdev, char *buf)
 	struct sas_internal *i = to_sas_internal(shost->transportt);
 	int val;
 
-	if (phy->local_attached)
+	if (scsi_is_sas_phy_local(phy))
 		return -EINVAL;
 
 	val = i->f->get_bay_identifier(rphy);
diff --git a/include/scsi/scsi_transport_sas.h b/include/scsi/scsi_transport_sas.h
index 6cc2314098cf..eeb2200de855 100644
--- a/include/scsi/scsi_transport_sas.h
+++ b/include/scsi/scsi_transport_sas.h
@@ -57,9 +57,6 @@ struct sas_phy {
 	enum sas_linkrate	maximum_linkrate_hw;
 	enum sas_linkrate	maximum_linkrate;
 
-	/* internal state */
-	unsigned int		local_attached : 1;
-
 	/* link error statistics */
 	u32			invalid_dword_count;
 	u32			running_disparity_error_count;
@@ -196,4 +193,6 @@ scsi_is_sas_expander_device(struct device *dev)
 		rphy->identify.device_type == SAS_EDGE_EXPANDER_DEVICE;
 }
 
+#define scsi_is_sas_phy_local(phy)	scsi_is_host_device((phy)->dev.parent)
+
 #endif /* SCSI_TRANSPORT_SAS_H */
-- 
cgit v1.2.3


From 2908d778ab3e244900c310974e1fc1c69066e450 Mon Sep 17 00:00:00 2001
From: James Bottomley <James.Bottomley@SteelEye.com>
Date: Tue, 29 Aug 2006 09:22:51 -0500
Subject: [SCSI] aic94xx: new driver

This is the end point of the separate aic94xx driver based on the
original driver and transport class from Luben Tuikov
<ltuikov@yahoo.com>

The log of the separate development is:

Alexis Bruemmer:
  o aic94xx: fix hotplug/unplug for expanderless systems
  o aic94xx: disable split completion timer/setting by default
  o aic94xx: wide port off expander support
  o aic94xx: remove various inline functions
  o aic94xx: use bitops
  o aic94xx: remove queue comment
  o aic94xx: remove sas_common.c
  o aic94xx: sas remove depot's
  o aic94xx: use available list_for_each_entry_safe_reverse()
  o aic94xx: sas header file merge

James Bottomley:
  o aic94xx: fix TF_TMF_NO_CTX processing
  o aic94xx: convert to request_firmware interface
  o aic94xx: fix hotplug/unplug
  o aic94xx: add link error counts to the expander phys
  o aic94xx: add transport class phy reset capability
  o aic94xx: remove local_attached flag
  o Remove README
  o Fixup Makefile variable for libsas rename
  o Rename sas->libsas
  o aic94xx: correct return code for sas_discover_event
  o aic94xx: use parent backlink port
  o aic94xx: remove channel abstraction
  o aic94xx: fix routing algorithms
  o aic94xx: add backlink port
  o aic94xx: fix cascaded expander properties
  o aic94xx: fix sleep under lock
  o aic94xx: fix panic on module removal in complex topology
  o aic94xx: make use of the new sas_port
  o rename sas_port to asd_sas_port
  o Fix for eh_strategy_handler move
  o aic94xx: move entirely over to correct transport class formulation
  o remove last vestages of sas_rphy_alloc()
  o update for eh_timed_out move
  o Preliminary expander support for aic94xx
  o sas: remove event thread
  o minor warning cleanups
  o remove last vestiges of id mapping arrays
  o Further updates
  o Convert aic94xx over entirely to the transport class end device and
  o update aic94xx/sas to use the new sas transport class end device
  o [PATCH] aic94xx: attaching to the sas transport class
  o Add missing completion removal from prior patch
  o [PATCH] aic94xx: attaching to the sas transport class
  o Build fixes from akpm

Jeff Garzik:
  o [scsi aic94xx] Remove ->owner from PCI info table

Luben Tuikov:
  o initial aic94xx driver

Mike Anderson:
  o aic94xx: fix panic on module insertion
  o aic94xx: stub out SATA_DEV case
  o aic94xx: compile warning cleanups
  o aic94xx: sas_alloc_task
  o aic94xx: ref count update
  o aic94xx nexus loss time value
  o [PATCH] aic94xx: driver assertion in non-x86 BIOS env

Randy Dunlap:
  o libsas: externs not needed

Robert Tarte:
  o aic94xx: sequence patch - fixes SATA support

Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 Documentation/scsi/libsas.txt          |  484 +++++++
 drivers/scsi/Kconfig                   |    5 +-
 drivers/scsi/Makefile                  |    2 +
 drivers/scsi/aic94xx/Kconfig           |   41 +
 drivers/scsi/aic94xx/Makefile          |   39 +
 drivers/scsi/aic94xx/aic94xx.h         |  114 ++
 drivers/scsi/aic94xx/aic94xx_dev.c     |  353 +++++
 drivers/scsi/aic94xx/aic94xx_dump.c    |  959 +++++++++++++
 drivers/scsi/aic94xx/aic94xx_dump.h    |   52 +
 drivers/scsi/aic94xx/aic94xx_hwi.c     | 1376 ++++++++++++++++++
 drivers/scsi/aic94xx/aic94xx_hwi.h     |  397 ++++++
 drivers/scsi/aic94xx/aic94xx_init.c    |  860 ++++++++++++
 drivers/scsi/aic94xx/aic94xx_reg.c     |  332 +++++
 drivers/scsi/aic94xx/aic94xx_reg.h     |  302 ++++
 drivers/scsi/aic94xx/aic94xx_reg_def.h | 2398 ++++++++++++++++++++++++++++++++
 drivers/scsi/aic94xx/aic94xx_sas.h     |  785 +++++++++++
 drivers/scsi/aic94xx/aic94xx_scb.c     |  732 ++++++++++
 drivers/scsi/aic94xx/aic94xx_sds.c     | 1136 +++++++++++++++
 drivers/scsi/aic94xx/aic94xx_seq.c     | 1401 +++++++++++++++++++
 drivers/scsi/aic94xx/aic94xx_seq.h     |   70 +
 drivers/scsi/aic94xx/aic94xx_task.c    |  642 +++++++++
 drivers/scsi/aic94xx/aic94xx_tmf.c     |  636 +++++++++
 drivers/scsi/libsas/Kconfig            |   39 +
 drivers/scsi/libsas/Makefile           |   36 +
 drivers/scsi/libsas/sas_discover.c     |  749 ++++++++++
 drivers/scsi/libsas/sas_dump.c         |   76 +
 drivers/scsi/libsas/sas_dump.h         |   42 +
 drivers/scsi/libsas/sas_event.c        |   75 +
 drivers/scsi/libsas/sas_expander.c     | 1862 +++++++++++++++++++++++++
 drivers/scsi/libsas/sas_init.c         |  227 +++
 drivers/scsi/libsas/sas_internal.h     |  146 ++
 drivers/scsi/libsas/sas_phy.c          |  157 +++
 drivers/scsi/libsas/sas_port.c         |  279 ++++
 drivers/scsi/libsas/sas_scsi_host.c    |  786 +++++++++++
 include/scsi/libsas.h                  |  627 +++++++++
 include/scsi/sas.h                     |  644 +++++++++
 include/scsi/scsi.h                    |    6 +
 37 files changed, 18866 insertions(+), 1 deletion(-)
 create mode 100644 Documentation/scsi/libsas.txt
 create mode 100644 drivers/scsi/aic94xx/Kconfig
 create mode 100644 drivers/scsi/aic94xx/Makefile
 create mode 100644 drivers/scsi/aic94xx/aic94xx.h
 create mode 100644 drivers/scsi/aic94xx/aic94xx_dev.c
 create mode 100644 drivers/scsi/aic94xx/aic94xx_dump.c
 create mode 100644 drivers/scsi/aic94xx/aic94xx_dump.h
 create mode 100644 drivers/scsi/aic94xx/aic94xx_hwi.c
 create mode 100644 drivers/scsi/aic94xx/aic94xx_hwi.h
 create mode 100644 drivers/scsi/aic94xx/aic94xx_init.c
 create mode 100644 drivers/scsi/aic94xx/aic94xx_reg.c
 create mode 100644 drivers/scsi/aic94xx/aic94xx_reg.h
 create mode 100644 drivers/scsi/aic94xx/aic94xx_reg_def.h
 create mode 100644 drivers/scsi/aic94xx/aic94xx_sas.h
 create mode 100644 drivers/scsi/aic94xx/aic94xx_scb.c
 create mode 100644 drivers/scsi/aic94xx/aic94xx_sds.c
 create mode 100644 drivers/scsi/aic94xx/aic94xx_seq.c
 create mode 100644 drivers/scsi/aic94xx/aic94xx_seq.h
 create mode 100644 drivers/scsi/aic94xx/aic94xx_task.c
 create mode 100644 drivers/scsi/aic94xx/aic94xx_tmf.c
 create mode 100644 drivers/scsi/libsas/Kconfig
 create mode 100644 drivers/scsi/libsas/Makefile
 create mode 100644 drivers/scsi/libsas/sas_discover.c
 create mode 100644 drivers/scsi/libsas/sas_dump.c
 create mode 100644 drivers/scsi/libsas/sas_dump.h
 create mode 100644 drivers/scsi/libsas/sas_event.c
 create mode 100644 drivers/scsi/libsas/sas_expander.c
 create mode 100644 drivers/scsi/libsas/sas_init.c
 create mode 100644 drivers/scsi/libsas/sas_internal.h
 create mode 100644 drivers/scsi/libsas/sas_phy.c
 create mode 100644 drivers/scsi/libsas/sas_port.c
 create mode 100644 drivers/scsi/libsas/sas_scsi_host.c
 create mode 100644 include/scsi/libsas.h
 create mode 100644 include/scsi/sas.h

(limited to 'include')

diff --git a/Documentation/scsi/libsas.txt b/Documentation/scsi/libsas.txt
new file mode 100644
index 000000000000..9e2078b2a615
--- /dev/null
+++ b/Documentation/scsi/libsas.txt
@@ -0,0 +1,484 @@
+SAS Layer
+---------
+
+The SAS Layer is a management infrastructure which manages
+SAS LLDDs.  It sits between SCSI Core and SAS LLDDs.  The
+layout is as follows: while SCSI Core is concerned with
+SAM/SPC issues, and a SAS LLDD+sequencer is concerned with
+phy/OOB/link management, the SAS layer is concerned with:
+
+      * SAS Phy/Port/HA event management (LLDD generates,
+        SAS Layer processes),
+      * SAS Port management (creation/destruction),
+      * SAS Domain discovery and revalidation,
+      * SAS Domain device management,
+      * SCSI Host registration/unregistration,
+      * Device registration with SCSI Core (SAS) or libata
+        (SATA), and
+      * Expander management and exporting expander control
+        to user space.
+
+A SAS LLDD is a PCI device driver.  It is concerned with
+phy/OOB management, and vendor specific tasks and generates
+events to the SAS layer.
+
+The SAS Layer does most SAS tasks as outlined in the SAS 1.1
+spec.
+
+The sas_ha_struct describes the SAS LLDD to the SAS layer.
+Most of it is used by the SAS Layer but a few fields need to
+be initialized by the LLDDs.
+
+After initializing your hardware, from the probe() function
+you call sas_register_ha(). It will register your LLDD with
+the SCSI subsystem, creating a SCSI host and it will
+register your SAS driver with the sysfs SAS tree it creates.
+It will then return.  Then you enable your phys to actually
+start OOB (at which point your driver will start calling the
+notify_* event callbacks).
+
+Structure descriptions:
+
+struct sas_phy --------------------
+Normally this is statically embedded to your driver's
+phy structure:
+	struct my_phy {
+	       blah;
+	       struct sas_phy sas_phy;
+	       bleh;
+	};
+And then all the phys are an array of my_phy in your HA
+struct (shown below).
+
+Then as you go along and initialize your phys you also
+initialize the sas_phy struct, along with your own
+phy structure.
+
+In general, the phys are managed by the LLDD and the ports
+are managed by the SAS layer.  So the phys are initialized
+and updated by the LLDD and the ports are initialized and
+updated by the SAS layer.
+
+There is a scheme where the LLDD can RW certain fields,
+and the SAS layer can only read such ones, and vice versa.
+The idea is to avoid unnecessary locking.
+
+enabled -- must be set (0/1)
+id -- must be set [0,MAX_PHYS)
+class, proto, type, role, oob_mode, linkrate -- must be set
+oob_mode --  you set this when OOB has finished and then notify
+the SAS Layer.
+
+sas_addr -- this normally points to an array holding the sas
+address of the phy, possibly somewhere in your my_phy
+struct.
+
+attached_sas_addr -- set this when you (LLDD) receive an
+IDENTIFY frame or a FIS frame, _before_ notifying the SAS
+layer.  The idea is that sometimes the LLDD may want to fake
+or provide a different SAS address on that phy/port and this
+allows it to do this.  At best you should copy the sas
+address from the IDENTIFY frame or maybe generate a SAS
+address for SATA directly attached devices.  The Discover
+process may later change this.
+
+frame_rcvd -- this is where you copy the IDENTIFY/FIS frame
+when you get it; you lock, copy, set frame_rcvd_size and
+unlock the lock, and then call the event.  It is a pointer
+since there's no way to know your hw frame size _exactly_,
+so you define the actual array in your phy struct and let
+this pointer point to it.  You copy the frame from your
+DMAable memory to that area holding the lock.
+
+sas_prim -- this is where primitives go when they're
+received.  See sas.h. Grab the lock, set the primitive,
+release the lock, notify.
+
+port -- this points to the sas_port if the phy belongs
+to a port -- the LLDD only reads this. It points to the
+sas_port this phy is part of.  Set by the SAS Layer.
+
+ha -- may be set; the SAS layer sets it anyway.
+
+lldd_phy -- you should set this to point to your phy so you
+can find your way around faster when the SAS layer calls one
+of your callbacks and passes you a phy.  If the sas_phy is
+embedded you can also use container_of -- whatever you
+prefer.
+
+
+struct sas_port --------------------
+The LLDD doesn't set any fields of this struct -- it only
+reads them.  They should be self explanatory.
+
+phy_mask is 32 bit, this should be enough for now, as I
+haven't heard of a HA having more than 8 phys.
+
+lldd_port -- I haven't found use for that -- maybe other
+LLDD who wish to have internal port representation can make
+use of this.
+
+
+struct sas_ha_struct --------------------
+It normally is statically declared in your own LLDD
+structure describing your adapter:
+struct my_sas_ha {
+       blah;
+       struct sas_ha_struct sas_ha;
+       struct my_phy phys[MAX_PHYS];
+       struct sas_port sas_ports[MAX_PHYS]; /* (1) */
+       bleh;
+};
+
+(1) If your LLDD doesn't have its own port representation.
+
+What needs to be initialized (sample function given below).
+
+pcidev
+sas_addr -- since the SAS layer doesn't want to mess with
+	 memory allocation, etc, this points to statically
+	 allocated array somewhere (say in your host adapter
+	 structure) and holds the SAS address of the host
+	 adapter as given by you or the manufacturer, etc.
+sas_port
+sas_phy -- an array of pointers to structures. (see
+	note above on sas_addr).
+	These must be set.  See more notes below.
+num_phys -- the number of phys present in the sas_phy array,
+	 and the number of ports present in the sas_port
+	 array.  There can be a maximum num_phys ports (one per
+	 port) so we drop the num_ports, and only use
+	 num_phys.
+
+The event interface:
+
+	/* LLDD calls these to notify the class of an event. */
+	void (*notify_ha_event)(struct sas_ha_struct *, enum ha_event);
+	void (*notify_port_event)(struct sas_phy *, enum port_event);
+	void (*notify_phy_event)(struct sas_phy *, enum phy_event);
+
+When sas_register_ha() returns, those are set and can be
+called by the LLDD to notify the SAS layer of such events
+the SAS layer.
+
+The port notification:
+
+	/* The class calls these to notify the LLDD of an event. */
+	void (*lldd_port_formed)(struct sas_phy *);
+	void (*lldd_port_deformed)(struct sas_phy *);
+
+If the LLDD wants notification when a port has been formed
+or deformed it sets those to a function satisfying the type.
+
+A SAS LLDD should also implement at least one of the Task
+Management Functions (TMFs) described in SAM:
+
+	/* Task Management Functions. Must be called from process context. */
+	int (*lldd_abort_task)(struct sas_task *);
+	int (*lldd_abort_task_set)(struct domain_device *, u8 *lun);
+	int (*lldd_clear_aca)(struct domain_device *, u8 *lun);
+	int (*lldd_clear_task_set)(struct domain_device *, u8 *lun);
+	int (*lldd_I_T_nexus_reset)(struct domain_device *);
+	int (*lldd_lu_reset)(struct domain_device *, u8 *lun);
+	int (*lldd_query_task)(struct sas_task *);
+
+For more information please read SAM from T10.org.
+
+Port and Adapter management:
+
+	/* Port and Adapter management */
+	int (*lldd_clear_nexus_port)(struct sas_port *);
+	int (*lldd_clear_nexus_ha)(struct sas_ha_struct *);
+
+A SAS LLDD should implement at least one of those.
+
+Phy management:
+
+	/* Phy management */
+	int (*lldd_control_phy)(struct sas_phy *, enum phy_func);
+
+lldd_ha -- set this to point to your HA struct. You can also
+use container_of if you embedded it as shown above.
+
+A sample initialization and registration function
+can look like this (called last thing from probe())
+*but* before you enable the phys to do OOB:
+
+static int register_sas_ha(struct my_sas_ha *my_ha)
+{
+	int i;
+	static struct sas_phy   *sas_phys[MAX_PHYS];
+	static struct sas_port  *sas_ports[MAX_PHYS];
+
+	my_ha->sas_ha.sas_addr = &my_ha->sas_addr[0];
+
+	for (i = 0; i < MAX_PHYS; i++) {
+		sas_phys[i] = &my_ha->phys[i].sas_phy;
+		sas_ports[i] = &my_ha->sas_ports[i];
+	}
+
+	my_ha->sas_ha.sas_phy  = sas_phys;
+	my_ha->sas_ha.sas_port = sas_ports;
+	my_ha->sas_ha.num_phys = MAX_PHYS;
+
+	my_ha->sas_ha.lldd_port_formed = my_port_formed;
+
+	my_ha->sas_ha.lldd_dev_found = my_dev_found;
+	my_ha->sas_ha.lldd_dev_gone = my_dev_gone;
+
+	my_ha->sas_ha.lldd_max_execute_num = lldd_max_execute_num; (1)
+
+	my_ha->sas_ha.lldd_queue_size = ha_can_queue;
+	my_ha->sas_ha.lldd_execute_task = my_execute_task;
+
+	my_ha->sas_ha.lldd_abort_task     = my_abort_task;
+	my_ha->sas_ha.lldd_abort_task_set = my_abort_task_set;
+	my_ha->sas_ha.lldd_clear_aca      = my_clear_aca;
+	my_ha->sas_ha.lldd_clear_task_set = my_clear_task_set;
+	my_ha->sas_ha.lldd_I_T_nexus_reset= NULL; (2)
+	my_ha->sas_ha.lldd_lu_reset       = my_lu_reset;
+	my_ha->sas_ha.lldd_query_task     = my_query_task;
+
+	my_ha->sas_ha.lldd_clear_nexus_port = my_clear_nexus_port;
+	my_ha->sas_ha.lldd_clear_nexus_ha = my_clear_nexus_ha;
+
+	my_ha->sas_ha.lldd_control_phy = my_control_phy;
+
+	return sas_register_ha(&my_ha->sas_ha);
+}
+
+(1) This is normally a LLDD parameter, something of the
+lines of a task collector.  What it tells the SAS Layer is
+whether the SAS layer should run in Direct Mode (default:
+value 0 or 1) or Task Collector Mode (value greater than 1).
+
+In Direct Mode, the SAS Layer calls Execute Task as soon as
+it has a command to send to the SDS, _and_ this is a single
+command, i.e. not linked.
+
+Some hardware (e.g. aic94xx) has the capability to DMA more
+than one task at a time (interrupt) from host memory.  Task
+Collector Mode is an optional feature for HAs which support
+this in their hardware.  (Again, it is completely optional
+even if your hardware supports it.)
+
+In Task Collector Mode, the SAS Layer would do _natural_
+coalescing of tasks and at the appropriate moment it would
+call your driver to DMA more than one task in a single HA
+interrupt. DMBS may want to use this by insmod/modprobe
+setting the lldd_max_execute_num to something greater than
+1.
+
+(2) SAS 1.1 does not define I_T Nexus Reset TMF.
+
+Events
+------
+
+Events are _the only way_ a SAS LLDD notifies the SAS layer
+of anything.  There is no other method or way a LLDD to tell
+the SAS layer of anything happening internally or in the SAS
+domain.
+
+Phy events:
+	PHYE_LOSS_OF_SIGNAL, (C)
+	PHYE_OOB_DONE,
+	PHYE_OOB_ERROR,      (C)
+	PHYE_SPINUP_HOLD.
+
+Port events, passed on a _phy_:
+	PORTE_BYTES_DMAED,      (M)
+	PORTE_BROADCAST_RCVD,   (E)
+	PORTE_LINK_RESET_ERR,   (C)
+	PORTE_TIMER_EVENT,      (C)
+	PORTE_HARD_RESET.
+
+Host Adapter event:
+	HAE_RESET
+
+A SAS LLDD should be able to generate
+	- at least one event from group C (choice),
+	- events marked M (mandatory) are mandatory (only one),
+	- events marked E (expander) if it wants the SAS layer
+	  to handle domain revalidation (only one such).
+	- Unmarked events are optional.
+
+Meaning:
+
+HAE_RESET -- when your HA got internal error and was reset.
+
+PORTE_BYTES_DMAED -- on receiving an IDENTIFY/FIS frame
+PORTE_BROADCAST_RCVD -- on receiving a primitive
+PORTE_LINK_RESET_ERR -- timer expired, loss of signal, loss
+of DWS, etc. (*)
+PORTE_TIMER_EVENT -- DWS reset timeout timer expired (*)
+PORTE_HARD_RESET -- Hard Reset primitive received.
+
+PHYE_LOSS_OF_SIGNAL -- the device is gone (*)
+PHYE_OOB_DONE -- OOB went fine and oob_mode is valid
+PHYE_OOB_ERROR -- Error while doing OOB, the device probably
+got disconnected. (*)
+PHYE_SPINUP_HOLD -- SATA is present, COMWAKE not sent.
+
+(*) should set/clear the appropriate fields in the phy,
+    or alternatively call the inlined sas_phy_disconnected()
+    which is just a helper, from their tasklet.
+
+The Execute Command SCSI RPC:
+
+	int (*lldd_execute_task)(struct sas_task *, int num,
+				 unsigned long gfp_flags);
+
+Used to queue a task to the SAS LLDD.  @task is the tasks to
+be executed.  @num should be the number of tasks being
+queued at this function call (they are linked listed via
+task::list), @gfp_mask should be the gfp_mask defining the
+context of the caller.
+
+This function should implement the Execute Command SCSI RPC,
+or if you're sending a SCSI Task as linked commands, you
+should also use this function.
+
+That is, when lldd_execute_task() is called, the command(s)
+go out on the transport *immediately*.  There is *no*
+queuing of any sort and at any level in a SAS LLDD.
+
+The use of task::list is two-fold, one for linked commands,
+the other discussed below.
+
+It is possible to queue up more than one task at a time, by
+initializing the list element of struct sas_task, and
+passing the number of tasks enlisted in this manner in num.
+
+Returns: -SAS_QUEUE_FULL, -ENOMEM, nothing was queued;
+	 0, the task(s) were queued.
+
+If you want to pass num > 1, then either
+A) you're the only caller of this function and keep track
+   of what you've queued to the LLDD, or
+B) you know what you're doing and have a strategy of
+   retrying.
+
+As opposed to queuing one task at a time (function call),
+batch queuing of tasks, by having num > 1, greatly
+simplifies LLDD code, sequencer code, and _hardware design_,
+and has some performance advantages in certain situations
+(DBMS).
+
+The LLDD advertises if it can take more than one command at
+a time at lldd_execute_task(), by setting the
+lldd_max_execute_num parameter (controlled by "collector"
+module parameter in aic94xx SAS LLDD).
+
+You should leave this to the default 1, unless you know what
+you're doing.
+
+This is a function of the LLDD, to which the SAS layer can
+cater to.
+
+int lldd_queue_size
+	The host adapter's queue size.  This is the maximum
+number of commands the lldd can have pending to domain
+devices on behalf of all upper layers submitting through
+lldd_execute_task().
+
+You really want to set this to something (much) larger than
+1.
+
+This _really_ has absolutely nothing to do with queuing.
+There is no queuing in SAS LLDDs.
+
+struct sas_task {
+	dev -- the device this task is destined to
+	list -- must be initialized (INIT_LIST_HEAD)
+	task_proto -- _one_ of enum sas_proto
+	scatter -- pointer to scatter gather list array
+	num_scatter -- number of elements in scatter
+	total_xfer_len -- total number of bytes expected to be transfered
+	data_dir -- PCI_DMA_...
+	task_done -- callback when the task has finished execution
+};
+
+When an external entity, entity other than the LLDD or the
+SAS Layer, wants to work with a struct domain_device, it
+_must_ call kobject_get() when getting a handle on the
+device and kobject_put() when it is done with the device.
+
+This does two things:
+     A) implements proper kfree() for the device;
+     B) increments/decrements the kref for all players:
+     domain_device
+	all domain_device's ... (if past an expander)
+	    port
+		host adapter
+		     pci device
+			 and up the ladder, etc.
+
+DISCOVERY
+---------
+
+The sysfs tree has the following purposes:
+    a) It shows you the physical layout of the SAS domain at
+       the current time, i.e. how the domain looks in the
+       physical world right now.
+    b) Shows some device parameters _at_discovery_time_.
+
+This is a link to the tree(1) program, very useful in
+viewing the SAS domain:
+ftp://mama.indstate.edu/linux/tree/
+I expect user space applications to actually create a
+graphical interface of this.
+
+That is, the sysfs domain tree doesn't show or keep state if
+you e.g., change the meaning of the READY LED MEANING
+setting, but it does show you the current connection status
+of the domain device.
+
+Keeping internal device state changes is responsibility of
+upper layers (Command set drivers) and user space.
+
+When a device or devices are unplugged from the domain, this
+is reflected in the sysfs tree immediately, and the device(s)
+removed from the system.
+
+The structure domain_device describes any device in the SAS
+domain.  It is completely managed by the SAS layer.  A task
+points to a domain device, this is how the SAS LLDD knows
+where to send the task(s) to.  A SAS LLDD only reads the
+contents of the domain_device structure, but it never creates
+or destroys one.
+
+Expander management from User Space
+-----------------------------------
+
+In each expander directory in sysfs, there is a file called
+"smp_portal".  It is a binary sysfs attribute file, which
+implements an SMP portal (Note: this is *NOT* an SMP port),
+to which user space applications can send SMP requests and
+receive SMP responses.
+
+Functionality is deceptively simple:
+
+1. Build the SMP frame you want to send. The format and layout
+   is described in the SAS spec.  Leave the CRC field equal 0.
+open(2)
+2. Open the expander's SMP portal sysfs file in RW mode.
+write(2)
+3. Write the frame you built in 1.
+read(2)
+4. Read the amount of data you expect to receive for the frame you built.
+   If you receive different amount of data you expected to receive,
+   then there was some kind of error.
+close(2)
+All this process is shown in detail in the function do_smp_func()
+and its callers, in the file "expander_conf.c".
+
+The kernel functionality is implemented in the file
+"sas_expander.c".
+
+The program "expander_conf.c" implements this. It takes one
+argument, the sysfs file name of the SMP portal to the
+expander, and gives expander information, including routing
+tables.
+
+The SMP portal gives you complete control of the expander,
+so please be careful.
diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig
index d61662c1a0ee..7de5fdfdab67 100644
--- a/drivers/scsi/Kconfig
+++ b/drivers/scsi/Kconfig
@@ -209,7 +209,7 @@ config SCSI_LOGGING
 	  there should be no noticeable performance impact as long as you have
 	  logging turned off.
 
-menu "SCSI Transport Attributes"
+menu "SCSI Transports"
 	depends on SCSI
 
 config SCSI_SPI_ATTRS
@@ -242,6 +242,8 @@ config SCSI_SAS_ATTRS
 	  If you wish to export transport-specific information about
 	  each attached SAS device to sysfs, say Y.
 
+source "drivers/scsi/libsas/Kconfig"
+
 endmenu
 
 menu "SCSI low-level drivers"
@@ -431,6 +433,7 @@ config SCSI_AIC7XXX_OLD
 	  module will be called aic7xxx_old.
 
 source "drivers/scsi/aic7xxx/Kconfig.aic79xx"
+source "drivers/scsi/aic94xx/Kconfig"
 
 # All the I2O code and drivers do not seem to be 64bit safe.
 config SCSI_DPT_I2O
diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile
index b2de9bfdfdcd..83da70decdd1 100644
--- a/drivers/scsi/Makefile
+++ b/drivers/scsi/Makefile
@@ -32,6 +32,7 @@ obj-$(CONFIG_SCSI_SPI_ATTRS)	+= scsi_transport_spi.o
 obj-$(CONFIG_SCSI_FC_ATTRS) 	+= scsi_transport_fc.o
 obj-$(CONFIG_SCSI_ISCSI_ATTRS)	+= scsi_transport_iscsi.o
 obj-$(CONFIG_SCSI_SAS_ATTRS)	+= scsi_transport_sas.o
+obj-$(CONFIG_SCSI_SAS_LIBSAS)	+= libsas/
 
 obj-$(CONFIG_ISCSI_TCP) 	+= libiscsi.o	iscsi_tcp.o
 obj-$(CONFIG_INFINIBAND_ISER) 	+= libiscsi.o
@@ -68,6 +69,7 @@ obj-$(CONFIG_SCSI_AIC7XXX)	+= aic7xxx/
 obj-$(CONFIG_SCSI_AIC79XX)	+= aic7xxx/
 obj-$(CONFIG_SCSI_AACRAID)	+= aacraid/
 obj-$(CONFIG_SCSI_AIC7XXX_OLD)	+= aic7xxx_old.o
+obj-$(CONFIG_SCSI_AIC94XX)	+= aic94xx/
 obj-$(CONFIG_SCSI_IPS)		+= ips.o
 obj-$(CONFIG_SCSI_FD_MCS)	+= fd_mcs.o
 obj-$(CONFIG_SCSI_FUTURE_DOMAIN)+= fdomain.o
diff --git a/drivers/scsi/aic94xx/Kconfig b/drivers/scsi/aic94xx/Kconfig
new file mode 100644
index 000000000000..0ed391d8ee84
--- /dev/null
+++ b/drivers/scsi/aic94xx/Kconfig
@@ -0,0 +1,41 @@
+#
+# Kernel configuration file for aic94xx SAS/SATA driver.
+#
+# Copyright (c) 2005 Adaptec, Inc.  All rights reserved.
+# Copyright (c) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+#
+# This file is licensed under GPLv2.
+#
+# This file is part of the aic94xx driver.
+#
+# The aic94xx driver is free software; you can redistribute it and/or
+# modify it under the terms of the GNU General Public License as
+# published by the Free Software Foundation; version 2 of the
+# License.
+#
+# The aic94xx driver is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with Aic94xx Driver; if not, write to the Free Software
+# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+#
+#
+
+config SCSI_AIC94XX
+	tristate "Adaptec AIC94xx SAS/SATA support"
+	depends on PCI
+	select SCSI_SAS_LIBSAS
+	help
+		This driver supports Adaptec's SAS/SATA 3Gb/s 64 bit PCI-X
+		AIC94xx chip based host adapters.
+
+config AIC94XX_DEBUG
+	bool "Compile in debug mode"
+	default y
+	depends on SCSI_AIC94XX
+	help
+		Compiles the aic94xx driver in debug mode.  In debug mode,
+		the driver prints some messages to the console.
diff --git a/drivers/scsi/aic94xx/Makefile b/drivers/scsi/aic94xx/Makefile
new file mode 100644
index 000000000000..e6b70123940c
--- /dev/null
+++ b/drivers/scsi/aic94xx/Makefile
@@ -0,0 +1,39 @@
+#
+# Makefile for Adaptec aic94xx SAS/SATA driver.
+#
+# Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+# Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+#
+# This file is licensed under GPLv2.
+#
+# This file is part of the the aic94xx driver.
+#
+# The aic94xx driver is free software; you can redistribute it and/or
+# modify it under the terms of the GNU General Public License as
+# published by the Free Software Foundation; version 2 of the
+# License.
+#
+# The aic94xx driver is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with the aic94xx driver; if not, write to the Free Software
+# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+
+ifeq ($(CONFIG_AIC94XX_DEBUG),y)
+	EXTRA_CFLAGS += -DASD_DEBUG -DASD_ENTER_EXIT
+endif
+
+obj-$(CONFIG_SCSI_AIC94XX) += aic94xx.o
+aic94xx-y += aic94xx_init.o \
+	     aic94xx_hwi.o  \
+	     aic94xx_reg.o  \
+	     aic94xx_sds.o  \
+	     aic94xx_seq.o  \
+	     aic94xx_dump.o \
+	     aic94xx_scb.o  \
+	     aic94xx_dev.o  \
+	     aic94xx_tmf.o  \
+	     aic94xx_task.o
diff --git a/drivers/scsi/aic94xx/aic94xx.h b/drivers/scsi/aic94xx/aic94xx.h
new file mode 100644
index 000000000000..cb7caf1c9ce1
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx.h
@@ -0,0 +1,114 @@
+/*
+ * Aic94xx SAS/SATA driver header file.
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ * $Id: //depot/aic94xx/aic94xx.h#31 $
+ */
+
+#ifndef _AIC94XX_H_
+#define _AIC94XX_H_
+
+#include <linux/slab.h>
+#include <linux/ctype.h>
+#include <scsi/libsas.h>
+
+#define ASD_DRIVER_NAME		"aic94xx"
+#define ASD_DRIVER_DESCRIPTION	"Adaptec aic94xx SAS/SATA driver"
+
+#define asd_printk(fmt, ...)	printk(KERN_NOTICE ASD_DRIVER_NAME ": " fmt, ## __VA_ARGS__)
+
+#ifdef ASD_ENTER_EXIT
+#define ENTER  printk(KERN_NOTICE "%s: ENTER %s\n", ASD_DRIVER_NAME, \
+		__FUNCTION__)
+#define EXIT   printk(KERN_NOTICE "%s: --EXIT %s\n", ASD_DRIVER_NAME, \
+		__FUNCTION__)
+#else
+#define ENTER
+#define EXIT
+#endif
+
+#ifdef ASD_DEBUG
+#define ASD_DPRINTK asd_printk
+#else
+#define ASD_DPRINTK(fmt, ...)
+#endif
+
+/* 2*ITNL timeout + 1 second */
+#define AIC94XX_SCB_TIMEOUT  (5*HZ)
+
+extern kmem_cache_t *asd_dma_token_cache;
+extern kmem_cache_t *asd_ascb_cache;
+extern char sas_addr_str[2*SAS_ADDR_SIZE + 1];
+
+static inline void asd_stringify_sas_addr(char *p, const u8 *sas_addr)
+{
+	int i;
+	for (i = 0; i < SAS_ADDR_SIZE; i++, p += 2)
+		snprintf(p, 3, "%02X", sas_addr[i]);
+	*p = '\0';
+}
+
+static inline void asd_destringify_sas_addr(u8 *sas_addr, const char *p)
+{
+	int i;
+	for (i = 0; i < SAS_ADDR_SIZE; i++) {
+		u8 h, l;
+		if (!*p)
+			break;
+		h = isdigit(*p) ? *p-'0' : *p-'A'+10;
+		p++;
+		l = isdigit(*p) ? *p-'0' : *p-'A'+10;
+		p++;
+		sas_addr[i] = (h<<4) | l;
+	}
+}
+
+struct asd_ha_struct;
+struct asd_ascb;
+
+int  asd_read_ocm(struct asd_ha_struct *asd_ha);
+int  asd_read_flash(struct asd_ha_struct *asd_ha);
+
+int  asd_dev_found(struct domain_device *dev);
+void asd_dev_gone(struct domain_device *dev);
+
+void asd_invalidate_edb(struct asd_ascb *ascb, int edb_id);
+
+int  asd_execute_task(struct sas_task *, int num, unsigned long gfp_flags);
+
+/* ---------- TMFs ---------- */
+int  asd_abort_task(struct sas_task *);
+int  asd_abort_task_set(struct domain_device *, u8 *lun);
+int  asd_clear_aca(struct domain_device *, u8 *lun);
+int  asd_clear_task_set(struct domain_device *, u8 *lun);
+int  asd_lu_reset(struct domain_device *, u8 *lun);
+int  asd_query_task(struct sas_task *);
+
+/* ---------- Adapter and Port management ---------- */
+int  asd_clear_nexus_port(struct asd_sas_port *port);
+int  asd_clear_nexus_ha(struct sas_ha_struct *sas_ha);
+
+/* ---------- Phy Management ---------- */
+int  asd_control_phy(struct asd_sas_phy *phy, enum phy_func func);
+
+#endif
diff --git a/drivers/scsi/aic94xx/aic94xx_dev.c b/drivers/scsi/aic94xx/aic94xx_dev.c
new file mode 100644
index 000000000000..6f8901b748f7
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx_dev.c
@@ -0,0 +1,353 @@
+/*
+ * Aic94xx SAS/SATA DDB management
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ * $Id: //depot/aic94xx/aic94xx_dev.c#21 $
+ */
+
+#include "aic94xx.h"
+#include "aic94xx_hwi.h"
+#include "aic94xx_reg.h"
+#include "aic94xx_sas.h"
+
+#define FIND_FREE_DDB(_ha) find_first_zero_bit((_ha)->hw_prof.ddb_bitmap, \
+					       (_ha)->hw_prof.max_ddbs)
+#define SET_DDB(_ddb, _ha) set_bit(_ddb, (_ha)->hw_prof.ddb_bitmap)
+#define CLEAR_DDB(_ddb, _ha) clear_bit(_ddb, (_ha)->hw_prof.ddb_bitmap)
+
+static inline int asd_get_ddb(struct asd_ha_struct *asd_ha)
+{
+	unsigned long flags;
+	int ddb, i;
+
+	spin_lock_irqsave(&asd_ha->hw_prof.ddb_lock, flags);
+	ddb = FIND_FREE_DDB(asd_ha);
+	if (ddb >= asd_ha->hw_prof.max_ddbs) {
+		ddb = -ENOMEM;
+		spin_unlock_irqrestore(&asd_ha->hw_prof.ddb_lock, flags);
+		goto out;
+	}
+	SET_DDB(ddb, asd_ha);
+	spin_unlock_irqrestore(&asd_ha->hw_prof.ddb_lock, flags);
+
+	for (i = 0; i < sizeof(struct asd_ddb_ssp_smp_target_port); i+= 4)
+		asd_ddbsite_write_dword(asd_ha, ddb, i, 0);
+out:
+	return ddb;
+}
+
+#define INIT_CONN_TAG   offsetof(struct asd_ddb_ssp_smp_target_port, init_conn_tag)
+#define DEST_SAS_ADDR   offsetof(struct asd_ddb_ssp_smp_target_port, dest_sas_addr)
+#define SEND_QUEUE_HEAD offsetof(struct asd_ddb_ssp_smp_target_port, send_queue_head)
+#define DDB_TYPE        offsetof(struct asd_ddb_ssp_smp_target_port, ddb_type)
+#define CONN_MASK       offsetof(struct asd_ddb_ssp_smp_target_port, conn_mask)
+#define DDB_TARG_FLAGS  offsetof(struct asd_ddb_ssp_smp_target_port, flags)
+#define DDB_TARG_FLAGS2 offsetof(struct asd_ddb_stp_sata_target_port, flags2)
+#define EXEC_QUEUE_TAIL offsetof(struct asd_ddb_ssp_smp_target_port, exec_queue_tail)
+#define SEND_QUEUE_TAIL offsetof(struct asd_ddb_ssp_smp_target_port, send_queue_tail)
+#define SISTER_DDB      offsetof(struct asd_ddb_ssp_smp_target_port, sister_ddb)
+#define MAX_CCONN       offsetof(struct asd_ddb_ssp_smp_target_port, max_concurrent_conn)
+#define NUM_CTX         offsetof(struct asd_ddb_ssp_smp_target_port, num_contexts)
+#define ATA_CMD_SCBPTR  offsetof(struct asd_ddb_stp_sata_target_port, ata_cmd_scbptr)
+#define SATA_TAG_ALLOC_MASK offsetof(struct asd_ddb_stp_sata_target_port, sata_tag_alloc_mask)
+#define NUM_SATA_TAGS   offsetof(struct asd_ddb_stp_sata_target_port, num_sata_tags)
+#define SATA_STATUS     offsetof(struct asd_ddb_stp_sata_target_port, sata_status)
+#define NCQ_DATA_SCB_PTR offsetof(struct asd_ddb_stp_sata_target_port, ncq_data_scb_ptr)
+#define ITNL_TIMEOUT    offsetof(struct asd_ddb_ssp_smp_target_port, itnl_timeout)
+
+static inline void asd_free_ddb(struct asd_ha_struct *asd_ha, int ddb)
+{
+	unsigned long flags;
+
+	if (!ddb || ddb >= 0xFFFF)
+		return;
+	asd_ddbsite_write_byte(asd_ha, ddb, DDB_TYPE, DDB_TYPE_UNUSED);
+	spin_lock_irqsave(&asd_ha->hw_prof.ddb_lock, flags);
+	CLEAR_DDB(ddb, asd_ha);
+	spin_unlock_irqrestore(&asd_ha->hw_prof.ddb_lock, flags);
+}
+
+static inline void asd_set_ddb_type(struct domain_device *dev)
+{
+	struct asd_ha_struct *asd_ha = dev->port->ha->lldd_ha;
+	int ddb = (int) (unsigned long) dev->lldd_dev;
+
+	if (dev->dev_type == SATA_PM_PORT)
+		asd_ddbsite_write_byte(asd_ha,ddb, DDB_TYPE, DDB_TYPE_PM_PORT);
+	else if (dev->tproto)
+		asd_ddbsite_write_byte(asd_ha,ddb, DDB_TYPE, DDB_TYPE_TARGET);
+	else
+		asd_ddbsite_write_byte(asd_ha,ddb,DDB_TYPE,DDB_TYPE_INITIATOR);
+}
+
+static int asd_init_sata_tag_ddb(struct domain_device *dev)
+{
+	struct asd_ha_struct *asd_ha = dev->port->ha->lldd_ha;
+	int ddb, i;
+
+	ddb = asd_get_ddb(asd_ha);
+	if (ddb < 0)
+		return ddb;
+
+	for (i = 0; i < sizeof(struct asd_ddb_sata_tag); i += 2)
+		asd_ddbsite_write_word(asd_ha, ddb, i, 0xFFFF);
+
+	asd_ddbsite_write_word(asd_ha, (int) (unsigned long) dev->lldd_dev,
+			       SISTER_DDB, ddb);
+	return 0;
+}
+
+static inline int asd_init_sata(struct domain_device *dev)
+{
+	struct asd_ha_struct *asd_ha = dev->port->ha->lldd_ha;
+	int ddb = (int) (unsigned long) dev->lldd_dev;
+	u32 qdepth = 0;
+	int res = 0;
+
+	asd_ddbsite_write_word(asd_ha, ddb, ATA_CMD_SCBPTR, 0xFFFF);
+	if ((dev->dev_type == SATA_DEV || dev->dev_type == SATA_PM_PORT) &&
+	    dev->sata_dev.identify_device &&
+	    dev->sata_dev.identify_device[10] != 0) {
+		u16 w75 = le16_to_cpu(dev->sata_dev.identify_device[75]);
+		u16 w76 = le16_to_cpu(dev->sata_dev.identify_device[76]);
+
+		if (w76 & 0x100) /* NCQ? */
+			qdepth = (w75 & 0x1F) + 1;
+		asd_ddbsite_write_dword(asd_ha, ddb, SATA_TAG_ALLOC_MASK,
+					(1<<qdepth)-1);
+		asd_ddbsite_write_byte(asd_ha, ddb, NUM_SATA_TAGS, qdepth);
+	}
+	if (dev->dev_type == SATA_DEV || dev->dev_type == SATA_PM ||
+	    dev->dev_type == SATA_PM_PORT) {
+		struct dev_to_host_fis *fis = (struct dev_to_host_fis *)
+			dev->frame_rcvd;
+		asd_ddbsite_write_byte(asd_ha, ddb, SATA_STATUS, fis->status);
+	}
+	asd_ddbsite_write_word(asd_ha, ddb, NCQ_DATA_SCB_PTR, 0xFFFF);
+	if (qdepth > 0)
+		res = asd_init_sata_tag_ddb(dev);
+	return res;
+}
+
+static int asd_init_target_ddb(struct domain_device *dev)
+{
+	int ddb, i;
+	struct asd_ha_struct *asd_ha = dev->port->ha->lldd_ha;
+	u8 flags = 0;
+
+	ddb = asd_get_ddb(asd_ha);
+	if (ddb < 0)
+		return ddb;
+
+	dev->lldd_dev = (void *) (unsigned long) ddb;
+
+	asd_ddbsite_write_byte(asd_ha, ddb, 0, DDB_TP_CONN_TYPE);
+	asd_ddbsite_write_byte(asd_ha, ddb, 1, 0);
+	asd_ddbsite_write_word(asd_ha, ddb, INIT_CONN_TAG, 0xFFFF);
+	for (i = 0; i < SAS_ADDR_SIZE; i++)
+		asd_ddbsite_write_byte(asd_ha, ddb, DEST_SAS_ADDR+i,
+				       dev->sas_addr[i]);
+	asd_ddbsite_write_word(asd_ha, ddb, SEND_QUEUE_HEAD, 0xFFFF);
+	asd_set_ddb_type(dev);
+	asd_ddbsite_write_byte(asd_ha, ddb, CONN_MASK, dev->port->phy_mask);
+	if (dev->port->oob_mode != SATA_OOB_MODE) {
+		flags |= OPEN_REQUIRED;
+		if ((dev->dev_type == SATA_DEV) ||
+		    (dev->tproto & SAS_PROTO_STP)) {
+			struct smp_resp *rps_resp = &dev->sata_dev.rps_resp;
+			if (rps_resp->frame_type == SMP_RESPONSE &&
+			    rps_resp->function == SMP_REPORT_PHY_SATA &&
+			    rps_resp->result == SMP_RESP_FUNC_ACC) {
+				if (rps_resp->rps.affil_valid)
+					flags |= STP_AFFIL_POL;
+				if (rps_resp->rps.affil_supp)
+					flags |= SUPPORTS_AFFIL;
+			}
+		} else {
+			flags |= CONCURRENT_CONN_SUPP;
+			if (!dev->parent &&
+			    (dev->dev_type == EDGE_DEV ||
+			     dev->dev_type == FANOUT_DEV))
+				asd_ddbsite_write_byte(asd_ha, ddb, MAX_CCONN,
+						       4);
+			else
+				asd_ddbsite_write_byte(asd_ha, ddb, MAX_CCONN,
+						       dev->pathways);
+			asd_ddbsite_write_byte(asd_ha, ddb, NUM_CTX, 1);
+		}
+	}
+	if (dev->dev_type == SATA_PM)
+		flags |= SATA_MULTIPORT;
+	asd_ddbsite_write_byte(asd_ha, ddb, DDB_TARG_FLAGS, flags);
+
+	flags = 0;
+	if (dev->tproto & SAS_PROTO_STP)
+		flags |= STP_CL_POL_NO_TX;
+	asd_ddbsite_write_byte(asd_ha, ddb, DDB_TARG_FLAGS2, flags);
+
+	asd_ddbsite_write_word(asd_ha, ddb, EXEC_QUEUE_TAIL, 0xFFFF);
+	asd_ddbsite_write_word(asd_ha, ddb, SEND_QUEUE_TAIL, 0xFFFF);
+	asd_ddbsite_write_word(asd_ha, ddb, SISTER_DDB, 0xFFFF);
+
+	if (dev->dev_type == SATA_DEV || (dev->tproto & SAS_PROTO_STP)) {
+		i = asd_init_sata(dev);
+		if (i < 0) {
+			asd_free_ddb(asd_ha, ddb);
+			return i;
+		}
+	}
+
+	if (dev->dev_type == SAS_END_DEV) {
+		struct sas_end_device *rdev = rphy_to_end_device(dev->rphy);
+		if (rdev->I_T_nexus_loss_timeout > 0)
+			asd_ddbsite_write_word(asd_ha, ddb, ITNL_TIMEOUT,
+					       min(rdev->I_T_nexus_loss_timeout,
+						   (u16)ITNL_TIMEOUT_CONST));
+		else
+			asd_ddbsite_write_word(asd_ha, ddb, ITNL_TIMEOUT,
+					       (u16)ITNL_TIMEOUT_CONST);
+	}
+	return 0;
+}
+
+static int asd_init_sata_pm_table_ddb(struct domain_device *dev)
+{
+	struct asd_ha_struct *asd_ha = dev->port->ha->lldd_ha;
+	int ddb, i;
+
+	ddb = asd_get_ddb(asd_ha);
+	if (ddb < 0)
+		return ddb;
+
+	for (i = 0; i < 32; i += 2)
+		asd_ddbsite_write_word(asd_ha, ddb, i, 0xFFFF);
+
+	asd_ddbsite_write_word(asd_ha, (int) (unsigned long) dev->lldd_dev,
+			       SISTER_DDB, ddb);
+
+	return 0;
+}
+
+#define PM_PORT_FLAGS offsetof(struct asd_ddb_sata_pm_port, pm_port_flags)
+#define PARENT_DDB    offsetof(struct asd_ddb_sata_pm_port, parent_ddb)
+
+/**
+ * asd_init_sata_pm_port_ddb -- SATA Port Multiplier Port
+ * dev: pointer to domain device
+ *
+ * For SATA Port Multiplier Ports we need to allocate one SATA Port
+ * Multiplier Port DDB and depending on whether the target on it
+ * supports SATA II NCQ, one SATA Tag DDB.
+ */
+static int asd_init_sata_pm_port_ddb(struct domain_device *dev)
+{
+	int ddb, i, parent_ddb, pmtable_ddb;
+	struct asd_ha_struct *asd_ha = dev->port->ha->lldd_ha;
+	u8  flags;
+
+	ddb = asd_get_ddb(asd_ha);
+	if (ddb < 0)
+		return ddb;
+
+	asd_set_ddb_type(dev);
+	flags = (dev->sata_dev.port_no << 4) | PM_PORT_SET;
+	asd_ddbsite_write_byte(asd_ha, ddb, PM_PORT_FLAGS, flags);
+	asd_ddbsite_write_word(asd_ha, ddb, SISTER_DDB, 0xFFFF);
+	asd_ddbsite_write_word(asd_ha, ddb, ATA_CMD_SCBPTR, 0xFFFF);
+	asd_init_sata(dev);
+
+	parent_ddb = (int) (unsigned long) dev->parent->lldd_dev;
+	asd_ddbsite_write_word(asd_ha, ddb, PARENT_DDB, parent_ddb);
+	pmtable_ddb = asd_ddbsite_read_word(asd_ha, parent_ddb, SISTER_DDB);
+	asd_ddbsite_write_word(asd_ha, pmtable_ddb, dev->sata_dev.port_no,ddb);
+
+	if (asd_ddbsite_read_byte(asd_ha, ddb, NUM_SATA_TAGS) > 0) {
+		i = asd_init_sata_tag_ddb(dev);
+		if (i < 0) {
+			asd_free_ddb(asd_ha, ddb);
+			return i;
+		}
+	}
+	return 0;
+}
+
+static int asd_init_initiator_ddb(struct domain_device *dev)
+{
+	return -ENODEV;
+}
+
+/**
+ * asd_init_sata_pm_ddb -- SATA Port Multiplier
+ * dev: pointer to domain device
+ *
+ * For STP and direct-attached SATA Port Multipliers we need
+ * one target port DDB entry and one SATA PM table DDB entry.
+ */
+static int asd_init_sata_pm_ddb(struct domain_device *dev)
+{
+	int res = 0;
+
+	res = asd_init_target_ddb(dev);
+	if (res)
+		goto out;
+	res = asd_init_sata_pm_table_ddb(dev);
+	if (res)
+		asd_free_ddb(dev->port->ha->lldd_ha,
+			     (int) (unsigned long) dev->lldd_dev);
+out:
+	return res;
+}
+
+int asd_dev_found(struct domain_device *dev)
+{
+	int res = 0;
+
+	switch (dev->dev_type) {
+	case SATA_PM:
+		res = asd_init_sata_pm_ddb(dev);
+		break;
+	case SATA_PM_PORT:
+		res = asd_init_sata_pm_port_ddb(dev);
+		break;
+	default:
+		if (dev->tproto)
+			res = asd_init_target_ddb(dev);
+		else
+			res = asd_init_initiator_ddb(dev);
+	}
+	return res;
+}
+
+void asd_dev_gone(struct domain_device *dev)
+{
+	int ddb, sister_ddb;
+	struct asd_ha_struct *asd_ha = dev->port->ha->lldd_ha;
+
+	ddb = (int) (unsigned long) dev->lldd_dev;
+	sister_ddb = asd_ddbsite_read_word(asd_ha, ddb, SISTER_DDB);
+
+	if (sister_ddb != 0xFFFF)
+		asd_free_ddb(asd_ha, sister_ddb);
+	asd_free_ddb(asd_ha, ddb);
+	dev->lldd_dev = NULL;
+}
diff --git a/drivers/scsi/aic94xx/aic94xx_dump.c b/drivers/scsi/aic94xx/aic94xx_dump.c
new file mode 100644
index 000000000000..e6ade5996d95
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx_dump.c
@@ -0,0 +1,959 @@
+/*
+ * Aic94xx SAS/SATA driver dump interface.
+ *
+ * Copyright (C) 2004 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2004 David Chaw <david_chaw@adaptec.com>
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ * 2005/07/14/LT  Complete overhaul of this file.  Update pages, register
+ * locations, names, etc.  Make use of macros.  Print more information.
+ * Print all cseq and lseq mip and mdp.
+ *
+ */
+
+#include "linux/pci.h"
+#include "aic94xx.h"
+#include "aic94xx_reg.h"
+#include "aic94xx_reg_def.h"
+#include "aic94xx_sas.h"
+
+#include "aic94xx_dump.h"
+
+#ifdef ASD_DEBUG
+
+#define MD(x)	    (1 << (x))
+#define MODE_COMMON (1 << 31)
+#define MODE_0_7    (0xFF)
+
+static const struct lseq_cio_regs {
+	char	*name;
+	u32	offs;
+	u8	width;
+	u32	mode;
+} LSEQmCIOREGS[] = {
+	{"LmMnSCBPTR",    0x20, 16, MD(0)|MD(1)|MD(2)|MD(3)|MD(4) },
+	{"LmMnDDBPTR",    0x22, 16, MD(0)|MD(1)|MD(2)|MD(3)|MD(4) },
+	{"LmREQMBX",      0x30, 32, MODE_COMMON },
+	{"LmRSPMBX",      0x34, 32, MODE_COMMON },
+	{"LmMnINT",       0x38, 32, MODE_0_7 },
+	{"LmMnINTEN",     0x3C, 32, MODE_0_7 },
+	{"LmXMTPRIMD",    0x40, 32, MODE_COMMON },
+	{"LmXMTPRIMCS",   0x44,  8, MODE_COMMON },
+	{"LmCONSTAT",     0x45,  8, MODE_COMMON },
+	{"LmMnDMAERRS",   0x46,  8, MD(0)|MD(1) },
+	{"LmMnSGDMAERRS", 0x47,  8, MD(0)|MD(1) },
+	{"LmMnEXPHDRP",   0x48,  8, MD(0) },
+	{"LmMnSASAALIGN", 0x48,  8, MD(1) },
+	{"LmMnMSKHDRP",   0x49,  8, MD(0) },
+	{"LmMnSTPALIGN",  0x49,  8, MD(1) },
+	{"LmMnRCVHDRP",   0x4A,  8, MD(0) },
+	{"LmMnXMTHDRP",   0x4A,  8, MD(1) },
+	{"LmALIGNMODE",   0x4B,  8, MD(1) },
+	{"LmMnEXPRCVCNT", 0x4C, 32, MD(0) },
+	{"LmMnXMTCNT",    0x4C, 32, MD(1) },
+	{"LmMnCURRTAG",   0x54, 16, MD(0) },
+	{"LmMnPREVTAG",   0x56, 16, MD(0) },
+	{"LmMnACKOFS",    0x58,  8, MD(1) },
+	{"LmMnXFRLVL",    0x59,  8, MD(0)|MD(1) },
+	{"LmMnSGDMACTL",  0x5A,  8, MD(0)|MD(1) },
+	{"LmMnSGDMASTAT", 0x5B,  8, MD(0)|MD(1) },
+	{"LmMnDDMACTL",   0x5C,  8, MD(0)|MD(1) },
+	{"LmMnDDMASTAT",  0x5D,  8, MD(0)|MD(1) },
+	{"LmMnDDMAMODE",  0x5E, 16, MD(0)|MD(1) },
+	{"LmMnPIPECTL",   0x61,  8, MD(0)|MD(1) },
+	{"LmMnACTSCB",    0x62, 16, MD(0)|MD(1) },
+	{"LmMnSGBHADR",   0x64,  8, MD(0)|MD(1) },
+	{"LmMnSGBADR",    0x65,  8, MD(0)|MD(1) },
+	{"LmMnSGDCNT",    0x66,  8, MD(0)|MD(1) },
+	{"LmMnSGDMADR",   0x68, 32, MD(0)|MD(1) },
+	{"LmMnSGDMADR",   0x6C, 32, MD(0)|MD(1) },
+	{"LmMnXFRCNT",    0x70, 32, MD(0)|MD(1) },
+	{"LmMnXMTCRC",    0x74, 32, MD(1) },
+	{"LmCURRTAG",     0x74, 16, MD(0) },
+	{"LmPREVTAG",     0x76, 16, MD(0) },
+	{"LmMnDPSEL",     0x7B,  8, MD(0)|MD(1) },
+	{"LmDPTHSTAT",    0x7C,  8, MODE_COMMON },
+	{"LmMnHOLDLVL",   0x7D,  8, MD(0) },
+	{"LmMnSATAFS",    0x7E,  8, MD(1) },
+	{"LmMnCMPLTSTAT", 0x7F,  8, MD(0)|MD(1) },
+	{"LmPRMSTAT0",    0x80, 32, MODE_COMMON },
+	{"LmPRMSTAT1",    0x84, 32, MODE_COMMON },
+	{"LmGPRMINT",     0x88,  8, MODE_COMMON },
+        {"LmMnCURRSCB",   0x8A, 16, MD(0) },
+	{"LmPRMICODE",    0x8C, 32, MODE_COMMON },
+	{"LmMnRCVCNT",    0x90, 16, MD(0) },
+	{"LmMnBUFSTAT",   0x92, 16, MD(0) },
+	{"LmMnXMTHDRSIZE",0x92,  8, MD(1) },
+	{"LmMnXMTSIZE",   0x93,  8, MD(1) },
+	{"LmMnTGTXFRCNT", 0x94, 32, MD(0) },
+	{"LmMnEXPROFS",   0x98, 32, MD(0) },
+	{"LmMnXMTROFS",   0x98, 32, MD(1) },
+	{"LmMnRCVROFS",   0x9C, 32, MD(0) },
+	{"LmCONCTL",      0xA0, 16, MODE_COMMON },
+	{"LmBITLTIMER",   0xA2, 16, MODE_COMMON },
+	{"LmWWNLOW",      0xA8, 32, MODE_COMMON },
+	{"LmWWNHIGH",     0xAC, 32, MODE_COMMON },
+	{"LmMnFRMERR",    0xB0, 32, MD(0) },
+	{"LmMnFRMERREN",  0xB4, 32, MD(0) },
+	{"LmAWTIMER",     0xB8, 16, MODE_COMMON },
+	{"LmAWTCTL",      0xBA,  8, MODE_COMMON },
+	{"LmMnHDRCMPS",   0xC0, 32, MD(0) },
+	{"LmMnXMTSTAT",   0xC4,  8, MD(1) },
+	{"LmHWTSTATEN",   0xC5,  8, MODE_COMMON },
+	{"LmMnRRDYRC",    0xC6,  8, MD(0) },
+        {"LmMnRRDYTC",    0xC6,  8, MD(1) },
+	{"LmHWTSTAT",     0xC7,  8, MODE_COMMON },
+	{"LmMnDATABUFADR",0xC8, 16, MD(0)|MD(1) },
+	{"LmDWSSTATUS",   0xCB,  8, MODE_COMMON },
+	{"LmMnACTSTAT",   0xCE, 16, MD(0)|MD(1) },
+	{"LmMnREQSCB",    0xD2, 16, MD(0)|MD(1) },
+	{"LmXXXPRIM",     0xD4, 32, MODE_COMMON },
+	{"LmRCVASTAT",    0xD9,  8, MODE_COMMON },
+	{"LmINTDIS1",     0xDA,  8, MODE_COMMON },
+	{"LmPSTORESEL",   0xDB,  8, MODE_COMMON },
+	{"LmPSTORE",      0xDC, 32, MODE_COMMON },
+	{"LmPRIMSTAT0EN", 0xE0, 32, MODE_COMMON },
+	{"LmPRIMSTAT1EN", 0xE4, 32, MODE_COMMON },
+	{"LmDONETCTL",    0xF2, 16, MODE_COMMON },
+	{NULL, 0, 0, 0 }
+};
+/*
+static struct lseq_cio_regs LSEQmOOBREGS[] = {
+   {"OOB_BFLTR"        ,0x100, 8, MD(5)},
+   {"OOB_INIT_MIN"     ,0x102,16, MD(5)},
+   {"OOB_INIT_MAX"     ,0x104,16, MD(5)},
+   {"OOB_INIT_NEG"     ,0x106,16, MD(5)},
+   {"OOB_SAS_MIN"      ,0x108,16, MD(5)},
+   {"OOB_SAS_MAX"      ,0x10A,16, MD(5)},
+   {"OOB_SAS_NEG"      ,0x10C,16, MD(5)},
+   {"OOB_WAKE_MIN"     ,0x10E,16, MD(5)},
+   {"OOB_WAKE_MAX"     ,0x110,16, MD(5)},
+   {"OOB_WAKE_NEG"     ,0x112,16, MD(5)},
+   {"OOB_IDLE_MAX"     ,0x114,16, MD(5)},
+   {"OOB_BURST_MAX"    ,0x116,16, MD(5)},
+   {"OOB_XMIT_BURST"   ,0x118, 8, MD(5)},
+   {"OOB_SEND_PAIRS"   ,0x119, 8, MD(5)},
+   {"OOB_INIT_IDLE"    ,0x11A, 8, MD(5)},
+   {"OOB_INIT_NEGO"    ,0x11C, 8, MD(5)},
+   {"OOB_SAS_IDLE"     ,0x11E, 8, MD(5)},
+   {"OOB_SAS_NEGO"     ,0x120, 8, MD(5)},
+   {"OOB_WAKE_IDLE"    ,0x122, 8, MD(5)},
+   {"OOB_WAKE_NEGO"    ,0x124, 8, MD(5)},
+   {"OOB_DATA_KBITS"   ,0x126, 8, MD(5)},
+   {"OOB_BURST_DATA"   ,0x128,32, MD(5)},
+   {"OOB_ALIGN_0_DATA" ,0x12C,32, MD(5)},
+   {"OOB_ALIGN_1_DATA" ,0x130,32, MD(5)},
+   {"OOB_SYNC_DATA"    ,0x134,32, MD(5)},
+   {"OOB_D10_2_DATA"   ,0x138,32, MD(5)},
+   {"OOB_PHY_RST_CNT"  ,0x13C,32, MD(5)},
+   {"OOB_SIG_GEN"      ,0x140, 8, MD(5)},
+   {"OOB_XMIT"         ,0x141, 8, MD(5)},
+   {"FUNCTION_MAKS"    ,0x142, 8, MD(5)},
+   {"OOB_MODE"         ,0x143, 8, MD(5)},
+   {"CURRENT_STATUS"   ,0x144, 8, MD(5)},
+   {"SPEED_MASK"       ,0x145, 8, MD(5)},
+   {"PRIM_COUNT"       ,0x146, 8, MD(5)},
+   {"OOB_SIGNALS"      ,0x148, 8, MD(5)},
+   {"OOB_DATA_DET"     ,0x149, 8, MD(5)},
+   {"OOB_TIME_OUT"     ,0x14C, 8, MD(5)},
+   {"OOB_TIMER_ENABLE" ,0x14D, 8, MD(5)},
+   {"OOB_STATUS"       ,0x14E, 8, MD(5)},
+   {"HOT_PLUG_DELAY"   ,0x150, 8, MD(5)},
+   {"RCD_DELAY"        ,0x151, 8, MD(5)},
+   {"COMSAS_TIMER"     ,0x152, 8, MD(5)},
+   {"SNTT_DELAY"       ,0x153, 8, MD(5)},
+   {"SPD_CHNG_DELAY"   ,0x154, 8, MD(5)},
+   {"SNLT_DELAY"       ,0x155, 8, MD(5)},
+   {"SNWT_DELAY"       ,0x156, 8, MD(5)},
+   {"ALIGN_DELAY"      ,0x157, 8, MD(5)},
+   {"INT_ENABLE_0"     ,0x158, 8, MD(5)},
+   {"INT_ENABLE_1"     ,0x159, 8, MD(5)},
+   {"INT_ENABLE_2"     ,0x15A, 8, MD(5)},
+   {"INT_ENABLE_3"     ,0x15B, 8, MD(5)},
+   {"OOB_TEST_REG"     ,0x15C, 8, MD(5)},
+   {"PHY_CONTROL_0"    ,0x160, 8, MD(5)},
+   {"PHY_CONTROL_1"    ,0x161, 8, MD(5)},
+   {"PHY_CONTROL_2"    ,0x162, 8, MD(5)},
+   {"PHY_CONTROL_3"    ,0x163, 8, MD(5)},
+   {"PHY_OOB_CAL_TX"   ,0x164, 8, MD(5)},
+   {"PHY_OOB_CAL_RX"   ,0x165, 8, MD(5)},
+   {"OOB_PHY_CAL_TX"   ,0x166, 8, MD(5)},
+   {"OOB_PHY_CAL_RX"   ,0x167, 8, MD(5)},
+   {"PHY_CONTROL_4"    ,0x168, 8, MD(5)},
+   {"PHY_TEST"         ,0x169, 8, MD(5)},
+   {"PHY_PWR_CTL"      ,0x16A, 8, MD(5)},
+   {"PHY_PWR_DELAY"    ,0x16B, 8, MD(5)},
+   {"OOB_SM_CON"       ,0x16C, 8, MD(5)},
+   {"ADDR_TRAP_1"      ,0x16D, 8, MD(5)},
+   {"ADDR_NEXT_1"      ,0x16E, 8, MD(5)},
+   {"NEXT_ST_1"        ,0x16F, 8, MD(5)},
+   {"OOB_SM_STATE"     ,0x170, 8, MD(5)},
+   {"ADDR_TRAP_2"      ,0x171, 8, MD(5)},
+   {"ADDR_NEXT_2"      ,0x172, 8, MD(5)},
+   {"NEXT_ST_2"        ,0x173, 8, MD(5)},
+   {NULL, 0, 0, 0 }
+};
+*/
+#define STR_8BIT   "   %30s[0x%04x]:0x%02x\n"
+#define STR_16BIT  "   %30s[0x%04x]:0x%04x\n"
+#define STR_32BIT  "   %30s[0x%04x]:0x%08x\n"
+#define STR_64BIT  "   %30s[0x%04x]:0x%llx\n"
+
+#define PRINT_REG_8bit(_ha, _n, _r) asd_printk(STR_8BIT, #_n, _n,      \
+					     asd_read_reg_byte(_ha, _r))
+#define PRINT_REG_16bit(_ha, _n, _r) asd_printk(STR_16BIT, #_n, _n,     \
+					      asd_read_reg_word(_ha, _r))
+#define PRINT_REG_32bit(_ha, _n, _r) asd_printk(STR_32BIT, #_n, _n,      \
+					      asd_read_reg_dword(_ha, _r))
+
+#define PRINT_CREG_8bit(_ha, _n) asd_printk(STR_8BIT, #_n, _n,      \
+					     asd_read_reg_byte(_ha, C##_n))
+#define PRINT_CREG_16bit(_ha, _n) asd_printk(STR_16BIT, #_n, _n,     \
+					      asd_read_reg_word(_ha, C##_n))
+#define PRINT_CREG_32bit(_ha, _n) asd_printk(STR_32BIT, #_n, _n,      \
+					      asd_read_reg_dword(_ha, C##_n))
+
+#define MSTR_8BIT   "   Mode:%02d %30s[0x%04x]:0x%02x\n"
+#define MSTR_16BIT  "   Mode:%02d %30s[0x%04x]:0x%04x\n"
+#define MSTR_32BIT  "   Mode:%02d %30s[0x%04x]:0x%08x\n"
+
+#define PRINT_MREG_8bit(_ha, _m, _n, _r) asd_printk(MSTR_8BIT, _m, #_n, _n,   \
+					     asd_read_reg_byte(_ha, _r))
+#define PRINT_MREG_16bit(_ha, _m, _n, _r) asd_printk(MSTR_16BIT, _m, #_n, _n, \
+					      asd_read_reg_word(_ha, _r))
+#define PRINT_MREG_32bit(_ha, _m, _n, _r) asd_printk(MSTR_32BIT, _m, #_n, _n, \
+					      asd_read_reg_dword(_ha, _r))
+
+/* can also be used for MD when the register is mode aware already */
+#define PRINT_MIS_byte(_ha, _n) asd_printk(STR_8BIT, #_n,CSEQ_##_n-CMAPPEDSCR,\
+                                           asd_read_reg_byte(_ha, CSEQ_##_n))
+#define PRINT_MIS_word(_ha, _n) asd_printk(STR_16BIT,#_n,CSEQ_##_n-CMAPPEDSCR,\
+                                           asd_read_reg_word(_ha, CSEQ_##_n))
+#define PRINT_MIS_dword(_ha, _n)                      \
+        asd_printk(STR_32BIT,#_n,CSEQ_##_n-CMAPPEDSCR,\
+                   asd_read_reg_dword(_ha, CSEQ_##_n))
+#define PRINT_MIS_qword(_ha, _n)                                       \
+        asd_printk(STR_64BIT, #_n,CSEQ_##_n-CMAPPEDSCR,                \
+                   (unsigned long long)(((u64)asd_read_reg_dword(_ha, CSEQ_##_n))     \
+                 | (((u64)asd_read_reg_dword(_ha, (CSEQ_##_n)+4))<<32)))
+
+#define CMDP_REG(_n, _m) (_m*(CSEQ_PAGE_SIZE*2)+CSEQ_##_n)
+#define PRINT_CMDP_word(_ha, _n) \
+asd_printk("%20s 0x%04x 0x%04x 0x%04x 0x%04x 0x%04x 0x%04x 0x%04x 0x%04x\n", \
+	#_n, \
+	asd_read_reg_word(_ha, CMDP_REG(_n, 0)), \
+	asd_read_reg_word(_ha, CMDP_REG(_n, 1)), \
+	asd_read_reg_word(_ha, CMDP_REG(_n, 2)), \
+	asd_read_reg_word(_ha, CMDP_REG(_n, 3)), \
+	asd_read_reg_word(_ha, CMDP_REG(_n, 4)), \
+	asd_read_reg_word(_ha, CMDP_REG(_n, 5)), \
+	asd_read_reg_word(_ha, CMDP_REG(_n, 6)), \
+	asd_read_reg_word(_ha, CMDP_REG(_n, 7)))
+
+#define PRINT_CMDP_byte(_ha, _n) \
+asd_printk("%20s 0x%04x 0x%04x 0x%04x 0x%04x 0x%04x 0x%04x 0x%04x 0x%04x\n", \
+	#_n, \
+	asd_read_reg_byte(_ha, CMDP_REG(_n, 0)), \
+	asd_read_reg_byte(_ha, CMDP_REG(_n, 1)), \
+	asd_read_reg_byte(_ha, CMDP_REG(_n, 2)), \
+	asd_read_reg_byte(_ha, CMDP_REG(_n, 3)), \
+	asd_read_reg_byte(_ha, CMDP_REG(_n, 4)), \
+	asd_read_reg_byte(_ha, CMDP_REG(_n, 5)), \
+	asd_read_reg_byte(_ha, CMDP_REG(_n, 6)), \
+	asd_read_reg_byte(_ha, CMDP_REG(_n, 7)))
+
+static void asd_dump_cseq_state(struct asd_ha_struct *asd_ha)
+{
+	int mode;
+
+	asd_printk("CSEQ STATE\n");
+
+	asd_printk("ARP2 REGISTERS\n");
+
+	PRINT_CREG_32bit(asd_ha, ARP2CTL);
+	PRINT_CREG_32bit(asd_ha, ARP2INT);
+	PRINT_CREG_32bit(asd_ha, ARP2INTEN);
+	PRINT_CREG_8bit(asd_ha, MODEPTR);
+	PRINT_CREG_8bit(asd_ha, ALTMODE);
+	PRINT_CREG_8bit(asd_ha, FLAG);
+	PRINT_CREG_8bit(asd_ha, ARP2INTCTL);
+	PRINT_CREG_16bit(asd_ha, STACK);
+	PRINT_CREG_16bit(asd_ha, PRGMCNT);
+	PRINT_CREG_16bit(asd_ha, ACCUM);
+	PRINT_CREG_16bit(asd_ha, SINDEX);
+	PRINT_CREG_16bit(asd_ha, DINDEX);
+	PRINT_CREG_8bit(asd_ha, SINDIR);
+	PRINT_CREG_8bit(asd_ha, DINDIR);
+	PRINT_CREG_8bit(asd_ha, JUMLDIR);
+	PRINT_CREG_8bit(asd_ha, ARP2HALTCODE);
+	PRINT_CREG_16bit(asd_ha, CURRADDR);
+	PRINT_CREG_16bit(asd_ha, LASTADDR);
+	PRINT_CREG_16bit(asd_ha, NXTLADDR);
+
+	asd_printk("IOP REGISTERS\n");
+
+	PRINT_REG_32bit(asd_ha, BISTCTL1, CBISTCTL);
+	PRINT_CREG_32bit(asd_ha, MAPPEDSCR);
+
+	asd_printk("CIO REGISTERS\n");
+
+	for (mode = 0; mode < 9; mode++)
+		PRINT_MREG_16bit(asd_ha, mode, MnSCBPTR, CMnSCBPTR(mode));
+	PRINT_MREG_16bit(asd_ha, 15, MnSCBPTR, CMnSCBPTR(15));
+
+	for (mode = 0; mode < 9; mode++)
+		PRINT_MREG_16bit(asd_ha, mode, MnDDBPTR, CMnDDBPTR(mode));
+	PRINT_MREG_16bit(asd_ha, 15, MnDDBPTR, CMnDDBPTR(15));
+
+	for (mode = 0; mode < 8; mode++)
+		PRINT_MREG_32bit(asd_ha, mode, MnREQMBX, CMnREQMBX(mode));
+	for (mode = 0; mode < 8; mode++)
+		PRINT_MREG_32bit(asd_ha, mode, MnRSPMBX, CMnRSPMBX(mode));
+	for (mode = 0; mode < 8; mode++)
+		PRINT_MREG_32bit(asd_ha, mode, MnINT, CMnINT(mode));
+	for (mode = 0; mode < 8; mode++)
+		PRINT_MREG_32bit(asd_ha, mode, MnINTEN, CMnINTEN(mode));
+
+	PRINT_CREG_8bit(asd_ha, SCRATCHPAGE);
+	for (mode = 0; mode < 8; mode++)
+		PRINT_MREG_8bit(asd_ha, mode, MnSCRATCHPAGE,
+				CMnSCRATCHPAGE(mode));
+
+	PRINT_REG_32bit(asd_ha, CLINKCON, CLINKCON);
+	PRINT_REG_8bit(asd_ha, CCONMSK, CCONMSK);
+	PRINT_REG_8bit(asd_ha, CCONEXIST, CCONEXIST);
+	PRINT_REG_16bit(asd_ha, CCONMODE, CCONMODE);
+	PRINT_REG_32bit(asd_ha, CTIMERCALC, CTIMERCALC);
+	PRINT_REG_8bit(asd_ha, CINTDIS, CINTDIS);
+
+	asd_printk("SCRATCH MEMORY\n");
+
+	asd_printk("MIP 4 >>>>>\n");
+	PRINT_MIS_word(asd_ha, Q_EXE_HEAD);
+	PRINT_MIS_word(asd_ha, Q_EXE_TAIL);
+	PRINT_MIS_word(asd_ha, Q_DONE_HEAD);
+	PRINT_MIS_word(asd_ha, Q_DONE_TAIL);
+	PRINT_MIS_word(asd_ha, Q_SEND_HEAD);
+	PRINT_MIS_word(asd_ha, Q_SEND_TAIL);
+	PRINT_MIS_word(asd_ha, Q_DMA2CHIM_HEAD);
+	PRINT_MIS_word(asd_ha, Q_DMA2CHIM_TAIL);
+	PRINT_MIS_word(asd_ha, Q_COPY_HEAD);
+	PRINT_MIS_word(asd_ha, Q_COPY_TAIL);
+	PRINT_MIS_word(asd_ha, REG0);
+	PRINT_MIS_word(asd_ha, REG1);
+	PRINT_MIS_dword(asd_ha, REG2);
+	PRINT_MIS_byte(asd_ha, LINK_CTL_Q_MAP);
+	PRINT_MIS_byte(asd_ha, MAX_CSEQ_MODE);
+	PRINT_MIS_byte(asd_ha, FREE_LIST_HACK_COUNT);
+
+	asd_printk("MIP 5 >>>>\n");
+	PRINT_MIS_qword(asd_ha, EST_NEXUS_REQ_QUEUE);
+	PRINT_MIS_qword(asd_ha, EST_NEXUS_REQ_COUNT);
+	PRINT_MIS_word(asd_ha, Q_EST_NEXUS_HEAD);
+	PRINT_MIS_word(asd_ha, Q_EST_NEXUS_TAIL);
+	PRINT_MIS_word(asd_ha, NEED_EST_NEXUS_SCB);
+	PRINT_MIS_byte(asd_ha, EST_NEXUS_REQ_HEAD);
+	PRINT_MIS_byte(asd_ha, EST_NEXUS_REQ_TAIL);
+	PRINT_MIS_byte(asd_ha, EST_NEXUS_SCB_OFFSET);
+
+	asd_printk("MIP 6 >>>>\n");
+	PRINT_MIS_word(asd_ha, INT_ROUT_RET_ADDR0);
+	PRINT_MIS_word(asd_ha, INT_ROUT_RET_ADDR1);
+	PRINT_MIS_word(asd_ha, INT_ROUT_SCBPTR);
+	PRINT_MIS_byte(asd_ha, INT_ROUT_MODE);
+	PRINT_MIS_byte(asd_ha, ISR_SCRATCH_FLAGS);
+	PRINT_MIS_word(asd_ha, ISR_SAVE_SINDEX);
+	PRINT_MIS_word(asd_ha, ISR_SAVE_DINDEX);
+	PRINT_MIS_word(asd_ha, Q_MONIRTT_HEAD);
+	PRINT_MIS_word(asd_ha, Q_MONIRTT_TAIL);
+	PRINT_MIS_byte(asd_ha, FREE_SCB_MASK);
+	PRINT_MIS_word(asd_ha, BUILTIN_FREE_SCB_HEAD);
+	PRINT_MIS_word(asd_ha, BUILTIN_FREE_SCB_TAIL);
+	PRINT_MIS_word(asd_ha, EXTENDED_FREE_SCB_HEAD);
+	PRINT_MIS_word(asd_ha, EXTENDED_FREE_SCB_TAIL);
+
+	asd_printk("MIP 7 >>>>\n");
+	PRINT_MIS_qword(asd_ha, EMPTY_REQ_QUEUE);
+	PRINT_MIS_qword(asd_ha, EMPTY_REQ_COUNT);
+	PRINT_MIS_word(asd_ha, Q_EMPTY_HEAD);
+	PRINT_MIS_word(asd_ha, Q_EMPTY_TAIL);
+	PRINT_MIS_word(asd_ha, NEED_EMPTY_SCB);
+	PRINT_MIS_byte(asd_ha, EMPTY_REQ_HEAD);
+	PRINT_MIS_byte(asd_ha, EMPTY_REQ_TAIL);
+	PRINT_MIS_byte(asd_ha, EMPTY_SCB_OFFSET);
+	PRINT_MIS_word(asd_ha, PRIMITIVE_DATA);
+	PRINT_MIS_dword(asd_ha, TIMEOUT_CONST);
+
+	asd_printk("MDP 0 >>>>\n");
+	asd_printk("%-20s %6s %6s %6s %6s %6s %6s %6s %6s\n",
+		   "Mode: ", "0", "1", "2", "3", "4", "5", "6", "7");
+	PRINT_CMDP_word(asd_ha, LRM_SAVE_SINDEX);
+	PRINT_CMDP_word(asd_ha, LRM_SAVE_SCBPTR);
+	PRINT_CMDP_word(asd_ha, Q_LINK_HEAD);
+	PRINT_CMDP_word(asd_ha, Q_LINK_TAIL);
+	PRINT_CMDP_byte(asd_ha, LRM_SAVE_SCRPAGE);
+
+	asd_printk("MDP 0 Mode 8 >>>>\n");
+	PRINT_MIS_word(asd_ha, RET_ADDR);
+	PRINT_MIS_word(asd_ha, RET_SCBPTR);
+	PRINT_MIS_word(asd_ha, SAVE_SCBPTR);
+	PRINT_MIS_word(asd_ha, EMPTY_TRANS_CTX);
+	PRINT_MIS_word(asd_ha, RESP_LEN);
+	PRINT_MIS_word(asd_ha, TMF_SCBPTR);
+	PRINT_MIS_word(asd_ha, GLOBAL_PREV_SCB);
+	PRINT_MIS_word(asd_ha, GLOBAL_HEAD);
+	PRINT_MIS_word(asd_ha, CLEAR_LU_HEAD);
+	PRINT_MIS_byte(asd_ha, TMF_OPCODE);
+	PRINT_MIS_byte(asd_ha, SCRATCH_FLAGS);
+	PRINT_MIS_word(asd_ha, HSB_SITE);
+	PRINT_MIS_word(asd_ha, FIRST_INV_SCB_SITE);
+	PRINT_MIS_word(asd_ha, FIRST_INV_DDB_SITE);
+
+	asd_printk("MDP 1 Mode 8 >>>>\n");
+	PRINT_MIS_qword(asd_ha, LUN_TO_CLEAR);
+	PRINT_MIS_qword(asd_ha, LUN_TO_CHECK);
+
+	asd_printk("MDP 2 Mode 8 >>>>\n");
+	PRINT_MIS_qword(asd_ha, HQ_NEW_POINTER);
+	PRINT_MIS_qword(asd_ha, HQ_DONE_BASE);
+	PRINT_MIS_dword(asd_ha, HQ_DONE_POINTER);
+	PRINT_MIS_byte(asd_ha, HQ_DONE_PASS);
+}
+
+#define PRINT_LREG_8bit(_h, _lseq, _n) \
+        asd_printk(STR_8BIT, #_n, _n, asd_read_reg_byte(_h, Lm##_n(_lseq)))
+#define PRINT_LREG_16bit(_h, _lseq, _n) \
+        asd_printk(STR_16BIT, #_n, _n, asd_read_reg_word(_h, Lm##_n(_lseq)))
+#define PRINT_LREG_32bit(_h, _lseq, _n) \
+        asd_printk(STR_32BIT, #_n, _n, asd_read_reg_dword(_h, Lm##_n(_lseq)))
+
+#define PRINT_LMIP_byte(_h, _lseq, _n)                              \
+	asd_printk(STR_8BIT, #_n, LmSEQ_##_n(_lseq)-LmSCRATCH(_lseq), \
+		   asd_read_reg_byte(_h, LmSEQ_##_n(_lseq)))
+#define PRINT_LMIP_word(_h, _lseq, _n)                              \
+	asd_printk(STR_16BIT, #_n, LmSEQ_##_n(_lseq)-LmSCRATCH(_lseq), \
+		   asd_read_reg_word(_h, LmSEQ_##_n(_lseq)))
+#define PRINT_LMIP_dword(_h, _lseq, _n)                             \
+	asd_printk(STR_32BIT, #_n, LmSEQ_##_n(_lseq)-LmSCRATCH(_lseq), \
+		   asd_read_reg_dword(_h, LmSEQ_##_n(_lseq)))
+#define PRINT_LMIP_qword(_h, _lseq, _n)                                \
+	asd_printk(STR_64BIT, #_n, LmSEQ_##_n(_lseq)-LmSCRATCH(_lseq), \
+		 (unsigned long long)(((unsigned long long) \
+		 asd_read_reg_dword(_h, LmSEQ_##_n(_lseq))) \
+	          | (((unsigned long long) \
+		 asd_read_reg_dword(_h, LmSEQ_##_n(_lseq)+4))<<32)))
+
+static void asd_print_lseq_cio_reg(struct asd_ha_struct *asd_ha,
+				   u32 lseq_cio_addr, int i)
+{
+	switch (LSEQmCIOREGS[i].width) {
+	case 8:
+		asd_printk("%20s[0x%x]: 0x%02x\n", LSEQmCIOREGS[i].name,
+			   LSEQmCIOREGS[i].offs,
+			   asd_read_reg_byte(asd_ha, lseq_cio_addr +
+					     LSEQmCIOREGS[i].offs));
+
+		break;
+	case 16:
+		asd_printk("%20s[0x%x]: 0x%04x\n", LSEQmCIOREGS[i].name,
+			   LSEQmCIOREGS[i].offs,
+			   asd_read_reg_word(asd_ha, lseq_cio_addr +
+					     LSEQmCIOREGS[i].offs));
+
+		break;
+	case 32:
+		asd_printk("%20s[0x%x]: 0x%08x\n", LSEQmCIOREGS[i].name,
+			   LSEQmCIOREGS[i].offs,
+			   asd_read_reg_dword(asd_ha, lseq_cio_addr +
+					      LSEQmCIOREGS[i].offs));
+		break;
+	}
+}
+
+static void asd_dump_lseq_state(struct asd_ha_struct *asd_ha, int lseq)
+{
+	u32 moffs;
+	int mode;
+
+	asd_printk("LSEQ %d STATE\n", lseq);
+
+	asd_printk("LSEQ%d: ARP2 REGISTERS\n", lseq);
+	PRINT_LREG_32bit(asd_ha, lseq, ARP2CTL);
+	PRINT_LREG_32bit(asd_ha, lseq, ARP2INT);
+	PRINT_LREG_32bit(asd_ha, lseq, ARP2INTEN);
+	PRINT_LREG_8bit(asd_ha, lseq, MODEPTR);
+	PRINT_LREG_8bit(asd_ha, lseq, ALTMODE);
+	PRINT_LREG_8bit(asd_ha, lseq, FLAG);
+	PRINT_LREG_8bit(asd_ha, lseq, ARP2INTCTL);
+	PRINT_LREG_16bit(asd_ha, lseq, STACK);
+	PRINT_LREG_16bit(asd_ha, lseq, PRGMCNT);
+	PRINT_LREG_16bit(asd_ha, lseq, ACCUM);
+	PRINT_LREG_16bit(asd_ha, lseq, SINDEX);
+	PRINT_LREG_16bit(asd_ha, lseq, DINDEX);
+	PRINT_LREG_8bit(asd_ha, lseq, SINDIR);
+	PRINT_LREG_8bit(asd_ha, lseq, DINDIR);
+	PRINT_LREG_8bit(asd_ha, lseq, JUMLDIR);
+	PRINT_LREG_8bit(asd_ha, lseq, ARP2HALTCODE);
+	PRINT_LREG_16bit(asd_ha, lseq, CURRADDR);
+	PRINT_LREG_16bit(asd_ha, lseq, LASTADDR);
+	PRINT_LREG_16bit(asd_ha, lseq, NXTLADDR);
+
+	asd_printk("LSEQ%d: IOP REGISTERS\n", lseq);
+
+	PRINT_LREG_32bit(asd_ha, lseq, MODECTL);
+	PRINT_LREG_32bit(asd_ha, lseq, DBGMODE);
+	PRINT_LREG_32bit(asd_ha, lseq, CONTROL);
+	PRINT_REG_32bit(asd_ha, BISTCTL0, LmBISTCTL0(lseq));
+	PRINT_REG_32bit(asd_ha, BISTCTL1, LmBISTCTL1(lseq));
+
+	asd_printk("LSEQ%d: CIO REGISTERS\n", lseq);
+	asd_printk("Mode common:\n");
+
+	for (mode = 0; mode < 8; mode++) {
+		u32 lseq_cio_addr = LmSEQ_PHY_BASE(mode, lseq);
+		int i;
+
+		for (i = 0; LSEQmCIOREGS[i].name; i++)
+			if (LSEQmCIOREGS[i].mode == MODE_COMMON)
+				asd_print_lseq_cio_reg(asd_ha,lseq_cio_addr,i);
+	}
+
+	asd_printk("Mode unique:\n");
+	for (mode = 0; mode < 8; mode++) {
+		u32 lseq_cio_addr = LmSEQ_PHY_BASE(mode, lseq);
+		int i;
+
+		asd_printk("Mode %d\n", mode);
+		for  (i = 0; LSEQmCIOREGS[i].name; i++) {
+			if (!(LSEQmCIOREGS[i].mode & (1 << mode)))
+				continue;
+			asd_print_lseq_cio_reg(asd_ha, lseq_cio_addr, i);
+		}
+	}
+
+	asd_printk("SCRATCH MEMORY\n");
+
+	asd_printk("LSEQ%d MIP 0 >>>>\n", lseq);
+	PRINT_LMIP_word(asd_ha, lseq, Q_TGTXFR_HEAD);
+	PRINT_LMIP_word(asd_ha, lseq, Q_TGTXFR_TAIL);
+	PRINT_LMIP_byte(asd_ha, lseq, LINK_NUMBER);
+	PRINT_LMIP_byte(asd_ha, lseq, SCRATCH_FLAGS);
+	PRINT_LMIP_qword(asd_ha, lseq, CONNECTION_STATE);
+	PRINT_LMIP_word(asd_ha, lseq, CONCTL);
+	PRINT_LMIP_byte(asd_ha, lseq, CONSTAT);
+	PRINT_LMIP_byte(asd_ha, lseq, CONNECTION_MODES);
+	PRINT_LMIP_word(asd_ha, lseq, REG1_ISR);
+	PRINT_LMIP_word(asd_ha, lseq, REG2_ISR);
+	PRINT_LMIP_word(asd_ha, lseq, REG3_ISR);
+	PRINT_LMIP_qword(asd_ha, lseq,REG0_ISR);
+
+	asd_printk("LSEQ%d MIP 1 >>>>\n", lseq);
+	PRINT_LMIP_word(asd_ha, lseq, EST_NEXUS_SCBPTR0);
+	PRINT_LMIP_word(asd_ha, lseq, EST_NEXUS_SCBPTR1);
+	PRINT_LMIP_word(asd_ha, lseq, EST_NEXUS_SCBPTR2);
+	PRINT_LMIP_word(asd_ha, lseq, EST_NEXUS_SCBPTR3);
+	PRINT_LMIP_byte(asd_ha, lseq, EST_NEXUS_SCB_OPCODE0);
+	PRINT_LMIP_byte(asd_ha, lseq, EST_NEXUS_SCB_OPCODE1);
+	PRINT_LMIP_byte(asd_ha, lseq, EST_NEXUS_SCB_OPCODE2);
+	PRINT_LMIP_byte(asd_ha, lseq, EST_NEXUS_SCB_OPCODE3);
+	PRINT_LMIP_byte(asd_ha, lseq, EST_NEXUS_SCB_HEAD);
+	PRINT_LMIP_byte(asd_ha, lseq, EST_NEXUS_SCB_TAIL);
+	PRINT_LMIP_byte(asd_ha, lseq, EST_NEXUS_BUF_AVAIL);
+	PRINT_LMIP_dword(asd_ha, lseq, TIMEOUT_CONST);
+	PRINT_LMIP_word(asd_ha, lseq, ISR_SAVE_SINDEX);
+	PRINT_LMIP_word(asd_ha, lseq, ISR_SAVE_DINDEX);
+
+	asd_printk("LSEQ%d MIP 2 >>>>\n", lseq);
+	PRINT_LMIP_word(asd_ha, lseq, EMPTY_SCB_PTR0);
+	PRINT_LMIP_word(asd_ha, lseq, EMPTY_SCB_PTR1);
+	PRINT_LMIP_word(asd_ha, lseq, EMPTY_SCB_PTR2);
+	PRINT_LMIP_word(asd_ha, lseq, EMPTY_SCB_PTR3);
+	PRINT_LMIP_byte(asd_ha, lseq, EMPTY_SCB_OPCD0);
+	PRINT_LMIP_byte(asd_ha, lseq, EMPTY_SCB_OPCD1);
+	PRINT_LMIP_byte(asd_ha, lseq, EMPTY_SCB_OPCD2);
+	PRINT_LMIP_byte(asd_ha, lseq, EMPTY_SCB_OPCD3);
+	PRINT_LMIP_byte(asd_ha, lseq, EMPTY_SCB_HEAD);
+	PRINT_LMIP_byte(asd_ha, lseq, EMPTY_SCB_TAIL);
+	PRINT_LMIP_byte(asd_ha, lseq, EMPTY_BUFS_AVAIL);
+
+	asd_printk("LSEQ%d MIP 3 >>>>\n", lseq);
+	PRINT_LMIP_dword(asd_ha, lseq, DEV_PRES_TMR_TOUT_CONST);
+	PRINT_LMIP_dword(asd_ha, lseq, SATA_INTERLOCK_TIMEOUT);
+	PRINT_LMIP_dword(asd_ha, lseq, SRST_ASSERT_TIMEOUT);
+	PRINT_LMIP_dword(asd_ha, lseq, RCV_FIS_TIMEOUT);
+	PRINT_LMIP_dword(asd_ha, lseq, ONE_MILLISEC_TIMEOUT);
+	PRINT_LMIP_dword(asd_ha, lseq, TEN_MS_COMINIT_TIMEOUT);
+	PRINT_LMIP_dword(asd_ha, lseq, SMP_RCV_TIMEOUT);
+
+	for (mode = 0; mode < 3; mode++) {
+		asd_printk("LSEQ%d MDP 0 MODE %d >>>>\n", lseq, mode);
+		moffs = mode * LSEQ_MODE_SCRATCH_SIZE;
+
+		asd_printk(STR_16BIT, "RET_ADDR", 0,
+			   asd_read_reg_word(asd_ha, LmSEQ_RET_ADDR(lseq)
+					     + moffs));
+		asd_printk(STR_16BIT, "REG0_MODE", 2,
+			   asd_read_reg_word(asd_ha, LmSEQ_REG0_MODE(lseq)
+					     + moffs));
+		asd_printk(STR_16BIT, "MODE_FLAGS", 4,
+			   asd_read_reg_word(asd_ha, LmSEQ_MODE_FLAGS(lseq)
+					     + moffs));
+		asd_printk(STR_16BIT, "RET_ADDR2", 0x6,
+			   asd_read_reg_word(asd_ha, LmSEQ_RET_ADDR2(lseq)
+					     + moffs));
+		asd_printk(STR_16BIT, "RET_ADDR1", 0x8,
+			   asd_read_reg_word(asd_ha, LmSEQ_RET_ADDR1(lseq)
+					     + moffs));
+		asd_printk(STR_8BIT, "OPCODE_TO_CSEQ", 0xB,
+			   asd_read_reg_byte(asd_ha, LmSEQ_OPCODE_TO_CSEQ(lseq)
+					     + moffs));
+		asd_printk(STR_16BIT, "DATA_TO_CSEQ", 0xC,
+			   asd_read_reg_word(asd_ha, LmSEQ_DATA_TO_CSEQ(lseq)
+					     + moffs));
+	}
+
+	asd_printk("LSEQ%d MDP 0 MODE 5 >>>>\n", lseq);
+	moffs = LSEQ_MODE5_PAGE0_OFFSET;
+	asd_printk(STR_16BIT, "RET_ADDR", 0,
+		   asd_read_reg_word(asd_ha, LmSEQ_RET_ADDR(lseq) + moffs));
+	asd_printk(STR_16BIT, "REG0_MODE", 2,
+		   asd_read_reg_word(asd_ha, LmSEQ_REG0_MODE(lseq) + moffs));
+	asd_printk(STR_16BIT, "MODE_FLAGS", 4,
+		   asd_read_reg_word(asd_ha, LmSEQ_MODE_FLAGS(lseq) + moffs));
+	asd_printk(STR_16BIT, "RET_ADDR2", 0x6,
+		   asd_read_reg_word(asd_ha, LmSEQ_RET_ADDR2(lseq) + moffs));
+	asd_printk(STR_16BIT, "RET_ADDR1", 0x8,
+		   asd_read_reg_word(asd_ha, LmSEQ_RET_ADDR1(lseq) + moffs));
+	asd_printk(STR_8BIT, "OPCODE_TO_CSEQ", 0xB,
+	   asd_read_reg_byte(asd_ha, LmSEQ_OPCODE_TO_CSEQ(lseq) + moffs));
+	asd_printk(STR_16BIT, "DATA_TO_CSEQ", 0xC,
+	   asd_read_reg_word(asd_ha, LmSEQ_DATA_TO_CSEQ(lseq) + moffs));
+
+	asd_printk("LSEQ%d MDP 0 MODE 0 >>>>\n", lseq);
+	PRINT_LMIP_word(asd_ha, lseq, FIRST_INV_DDB_SITE);
+	PRINT_LMIP_word(asd_ha, lseq, EMPTY_TRANS_CTX);
+	PRINT_LMIP_word(asd_ha, lseq, RESP_LEN);
+	PRINT_LMIP_word(asd_ha, lseq, FIRST_INV_SCB_SITE);
+	PRINT_LMIP_dword(asd_ha, lseq, INTEN_SAVE);
+	PRINT_LMIP_byte(asd_ha, lseq, LINK_RST_FRM_LEN);
+	PRINT_LMIP_byte(asd_ha, lseq, LINK_RST_PROTOCOL);
+	PRINT_LMIP_byte(asd_ha, lseq, RESP_STATUS);
+	PRINT_LMIP_byte(asd_ha, lseq, LAST_LOADED_SGE);
+	PRINT_LMIP_byte(asd_ha, lseq, SAVE_SCBPTR);
+
+	asd_printk("LSEQ%d MDP 0 MODE 1 >>>>\n", lseq);
+	PRINT_LMIP_word(asd_ha, lseq, Q_XMIT_HEAD);
+	PRINT_LMIP_word(asd_ha, lseq, M1_EMPTY_TRANS_CTX);
+	PRINT_LMIP_word(asd_ha, lseq, INI_CONN_TAG);
+	PRINT_LMIP_byte(asd_ha, lseq, FAILED_OPEN_STATUS);
+	PRINT_LMIP_byte(asd_ha, lseq, XMIT_REQUEST_TYPE);
+	PRINT_LMIP_byte(asd_ha, lseq, M1_RESP_STATUS);
+	PRINT_LMIP_byte(asd_ha, lseq, M1_LAST_LOADED_SGE);
+	PRINT_LMIP_word(asd_ha, lseq, M1_SAVE_SCBPTR);
+
+	asd_printk("LSEQ%d MDP 0 MODE 2 >>>>\n", lseq);
+	PRINT_LMIP_word(asd_ha, lseq, PORT_COUNTER);
+	PRINT_LMIP_word(asd_ha, lseq, PM_TABLE_PTR);
+	PRINT_LMIP_word(asd_ha, lseq, SATA_INTERLOCK_TMR_SAVE);
+	PRINT_LMIP_word(asd_ha, lseq, IP_BITL);
+	PRINT_LMIP_word(asd_ha, lseq, COPY_SMP_CONN_TAG);
+	PRINT_LMIP_byte(asd_ha, lseq, P0M2_OFFS1AH);
+
+	asd_printk("LSEQ%d MDP 0 MODE 4/5 >>>>\n", lseq);
+	PRINT_LMIP_byte(asd_ha, lseq, SAVED_OOB_STATUS);
+	PRINT_LMIP_byte(asd_ha, lseq, SAVED_OOB_MODE);
+	PRINT_LMIP_word(asd_ha, lseq, Q_LINK_HEAD);
+	PRINT_LMIP_byte(asd_ha, lseq, LINK_RST_ERR);
+	PRINT_LMIP_byte(asd_ha, lseq, SAVED_OOB_SIGNALS);
+	PRINT_LMIP_byte(asd_ha, lseq, SAS_RESET_MODE);
+	PRINT_LMIP_byte(asd_ha, lseq, LINK_RESET_RETRY_COUNT);
+	PRINT_LMIP_byte(asd_ha, lseq, NUM_LINK_RESET_RETRIES);
+	PRINT_LMIP_word(asd_ha, lseq, OOB_INT_ENABLES);
+	PRINT_LMIP_word(asd_ha, lseq, NOTIFY_TIMER_TIMEOUT);
+	PRINT_LMIP_word(asd_ha, lseq, NOTIFY_TIMER_DOWN_COUNT);
+
+	asd_printk("LSEQ%d MDP 1 MODE 0 >>>>\n", lseq);
+	PRINT_LMIP_qword(asd_ha, lseq, SG_LIST_PTR_ADDR0);
+	PRINT_LMIP_qword(asd_ha, lseq, SG_LIST_PTR_ADDR1);
+
+	asd_printk("LSEQ%d MDP 1 MODE 1 >>>>\n", lseq);
+	PRINT_LMIP_qword(asd_ha, lseq, M1_SG_LIST_PTR_ADDR0);
+	PRINT_LMIP_qword(asd_ha, lseq, M1_SG_LIST_PTR_ADDR1);
+
+	asd_printk("LSEQ%d MDP 1 MODE 2 >>>>\n", lseq);
+	PRINT_LMIP_dword(asd_ha, lseq, INVALID_DWORD_COUNT);
+	PRINT_LMIP_dword(asd_ha, lseq, DISPARITY_ERROR_COUNT);
+	PRINT_LMIP_dword(asd_ha, lseq, LOSS_OF_SYNC_COUNT);
+
+	asd_printk("LSEQ%d MDP 1 MODE 4/5 >>>>\n", lseq);
+	PRINT_LMIP_dword(asd_ha, lseq, FRAME_TYPE_MASK);
+	PRINT_LMIP_dword(asd_ha, lseq, HASHED_SRC_ADDR_MASK_PRINT);
+	PRINT_LMIP_byte(asd_ha, lseq, NUM_FILL_BYTES_MASK);
+	PRINT_LMIP_word(asd_ha, lseq, TAG_MASK);
+	PRINT_LMIP_word(asd_ha, lseq, TARGET_PORT_XFER_TAG);
+	PRINT_LMIP_dword(asd_ha, lseq, DATA_OFFSET);
+
+	asd_printk("LSEQ%d MDP 2 MODE 0 >>>>\n", lseq);
+	PRINT_LMIP_dword(asd_ha, lseq, SMP_RCV_TIMER_TERM_TS);
+	PRINT_LMIP_byte(asd_ha, lseq, DEVICE_BITS);
+	PRINT_LMIP_word(asd_ha, lseq, SDB_DDB);
+	PRINT_LMIP_word(asd_ha, lseq, SDB_NUM_TAGS);
+	PRINT_LMIP_word(asd_ha, lseq, SDB_CURR_TAG);
+
+	asd_printk("LSEQ%d MDP 2 MODE 1 >>>>\n", lseq);
+	PRINT_LMIP_qword(asd_ha, lseq, TX_ID_ADDR_FRAME);
+	PRINT_LMIP_dword(asd_ha, lseq, OPEN_TIMER_TERM_TS);
+	PRINT_LMIP_dword(asd_ha, lseq, SRST_AS_TIMER_TERM_TS);
+	PRINT_LMIP_dword(asd_ha, lseq, LAST_LOADED_SG_EL);
+
+	asd_printk("LSEQ%d MDP 2 MODE 2 >>>>\n", lseq);
+	PRINT_LMIP_dword(asd_ha, lseq, CLOSE_TIMER_TERM_TS);
+	PRINT_LMIP_dword(asd_ha, lseq, BREAK_TIMER_TERM_TS);
+	PRINT_LMIP_dword(asd_ha, lseq, DWS_RESET_TIMER_TERM_TS);
+	PRINT_LMIP_dword(asd_ha, lseq, SATA_INTERLOCK_TIMER_TERM_TS);
+	PRINT_LMIP_dword(asd_ha, lseq, MCTL_TIMER_TERM_TS);
+
+	asd_printk("LSEQ%d MDP 2 MODE 4/5 >>>>\n", lseq);
+	PRINT_LMIP_dword(asd_ha, lseq, COMINIT_TIMER_TERM_TS);
+	PRINT_LMIP_dword(asd_ha, lseq, RCV_ID_TIMER_TERM_TS);
+	PRINT_LMIP_dword(asd_ha, lseq, RCV_FIS_TIMER_TERM_TS);
+	PRINT_LMIP_dword(asd_ha, lseq, DEV_PRES_TIMER_TERM_TS);
+}
+
+/**
+ * asd_dump_ddb_site -- dump a CSEQ DDB site
+ * @asd_ha: pointer to host adapter structure
+ * @site_no: site number of interest
+ */
+void asd_dump_target_ddb(struct asd_ha_struct *asd_ha, u16 site_no)
+{
+	if (site_no >= asd_ha->hw_prof.max_ddbs)
+		return;
+
+#define DDB_FIELDB(__name)                                        \
+	asd_ddbsite_read_byte(asd_ha, site_no,                    \
+			      offsetof(struct asd_ddb_ssp_smp_target_port, __name))
+#define DDB2_FIELDB(__name)                                       \
+	asd_ddbsite_read_byte(asd_ha, site_no,                    \
+			      offsetof(struct asd_ddb_stp_sata_target_port, __name))
+#define DDB_FIELDW(__name)                                        \
+	asd_ddbsite_read_word(asd_ha, site_no,                    \
+			      offsetof(struct asd_ddb_ssp_smp_target_port, __name))
+
+#define DDB_FIELDD(__name)                                         \
+	asd_ddbsite_read_dword(asd_ha, site_no,                    \
+			       offsetof(struct asd_ddb_ssp_smp_target_port, __name))
+
+	asd_printk("DDB: 0x%02x\n", site_no);
+	asd_printk("conn_type: 0x%02x\n", DDB_FIELDB(conn_type));
+	asd_printk("conn_rate: 0x%02x\n", DDB_FIELDB(conn_rate));
+	asd_printk("init_conn_tag: 0x%04x\n", be16_to_cpu(DDB_FIELDW(init_conn_tag)));
+	asd_printk("send_queue_head: 0x%04x\n", be16_to_cpu(DDB_FIELDW(send_queue_head)));
+	asd_printk("sq_suspended: 0x%02x\n", DDB_FIELDB(sq_suspended));
+	asd_printk("DDB Type: 0x%02x\n", DDB_FIELDB(ddb_type));
+	asd_printk("AWT Default: 0x%04x\n", DDB_FIELDW(awt_def));
+	asd_printk("compat_features: 0x%02x\n", DDB_FIELDB(compat_features));
+	asd_printk("Pathway Blocked Count: 0x%02x\n",
+		   DDB_FIELDB(pathway_blocked_count));
+	asd_printk("arb_wait_time: 0x%04x\n", DDB_FIELDW(arb_wait_time));
+	asd_printk("more_compat_features: 0x%08x\n",
+		   DDB_FIELDD(more_compat_features));
+	asd_printk("Conn Mask: 0x%02x\n", DDB_FIELDB(conn_mask));
+	asd_printk("flags: 0x%02x\n", DDB_FIELDB(flags));
+	asd_printk("flags2: 0x%02x\n", DDB2_FIELDB(flags2));
+	asd_printk("ExecQ Tail: 0x%04x\n",DDB_FIELDW(exec_queue_tail));
+	asd_printk("SendQ Tail: 0x%04x\n",DDB_FIELDW(send_queue_tail));
+	asd_printk("Active Task Count: 0x%04x\n",
+		   DDB_FIELDW(active_task_count));
+	asd_printk("ITNL Reason: 0x%02x\n", DDB_FIELDB(itnl_reason));
+	asd_printk("ITNL Timeout Const: 0x%04x\n", DDB_FIELDW(itnl_timeout));
+	asd_printk("ITNL timestamp: 0x%08x\n", DDB_FIELDD(itnl_timestamp));
+}
+
+void asd_dump_ddb_0(struct asd_ha_struct *asd_ha)
+{
+#define DDB0_FIELDB(__name)                                  \
+	asd_ddbsite_read_byte(asd_ha, 0,                     \
+			      offsetof(struct asd_ddb_seq_shared, __name))
+#define DDB0_FIELDW(__name)                                  \
+	asd_ddbsite_read_word(asd_ha, 0,                     \
+			      offsetof(struct asd_ddb_seq_shared, __name))
+
+#define DDB0_FIELDD(__name)                                  \
+	asd_ddbsite_read_dword(asd_ha,0 ,                    \
+			       offsetof(struct asd_ddb_seq_shared, __name))
+
+#define DDB0_FIELDA(__name, _o)                              \
+	asd_ddbsite_read_byte(asd_ha, 0,                     \
+			      offsetof(struct asd_ddb_seq_shared, __name)+_o)
+
+
+	asd_printk("DDB: 0\n");
+	asd_printk("q_free_ddb_head:%04x\n", DDB0_FIELDW(q_free_ddb_head));
+	asd_printk("q_free_ddb_tail:%04x\n", DDB0_FIELDW(q_free_ddb_tail));
+	asd_printk("q_free_ddb_cnt:%04x\n",  DDB0_FIELDW(q_free_ddb_cnt));
+	asd_printk("q_used_ddb_head:%04x\n", DDB0_FIELDW(q_used_ddb_head));
+	asd_printk("q_used_ddb_tail:%04x\n", DDB0_FIELDW(q_used_ddb_tail));
+	asd_printk("shared_mem_lock:%04x\n", DDB0_FIELDW(shared_mem_lock));
+	asd_printk("smp_conn_tag:%04x\n",    DDB0_FIELDW(smp_conn_tag));
+	asd_printk("est_nexus_buf_cnt:%04x\n", DDB0_FIELDW(est_nexus_buf_cnt));
+	asd_printk("est_nexus_buf_thresh:%04x\n",
+		   DDB0_FIELDW(est_nexus_buf_thresh));
+	asd_printk("conn_not_active:%02x\n", DDB0_FIELDB(conn_not_active));
+	asd_printk("phy_is_up:%02x\n",       DDB0_FIELDB(phy_is_up));
+	asd_printk("port_map_by_links:%02x %02x %02x %02x "
+		   "%02x %02x %02x %02x\n",
+		   DDB0_FIELDA(port_map_by_links, 0),
+		   DDB0_FIELDA(port_map_by_links, 1),
+		   DDB0_FIELDA(port_map_by_links, 2),
+		   DDB0_FIELDA(port_map_by_links, 3),
+		   DDB0_FIELDA(port_map_by_links, 4),
+		   DDB0_FIELDA(port_map_by_links, 5),
+		   DDB0_FIELDA(port_map_by_links, 6),
+		   DDB0_FIELDA(port_map_by_links, 7));
+}
+
+static void asd_dump_scb_site(struct asd_ha_struct *asd_ha, u16 site_no)
+{
+
+#define SCB_FIELDB(__name)                                                 \
+	asd_scbsite_read_byte(asd_ha, site_no, sizeof(struct scb_header)   \
+			      + offsetof(struct initiate_ssp_task, __name))
+#define SCB_FIELDW(__name)                                                 \
+	asd_scbsite_read_word(asd_ha, site_no, sizeof(struct scb_header)   \
+			      + offsetof(struct initiate_ssp_task, __name))
+#define SCB_FIELDD(__name)                                                 \
+	asd_scbsite_read_dword(asd_ha, site_no, sizeof(struct scb_header)  \
+			       + offsetof(struct initiate_ssp_task, __name))
+
+	asd_printk("Total Xfer Len: 0x%08x.\n", SCB_FIELDD(total_xfer_len));
+	asd_printk("Frame Type: 0x%02x.\n", SCB_FIELDB(ssp_frame.frame_type));
+	asd_printk("Tag: 0x%04x.\n", SCB_FIELDW(ssp_frame.tag));
+	asd_printk("Target Port Xfer Tag: 0x%04x.\n",
+		   SCB_FIELDW(ssp_frame.tptt));
+	asd_printk("Data Offset: 0x%08x.\n", SCB_FIELDW(ssp_frame.data_offs));
+	asd_printk("Retry Count: 0x%02x.\n", SCB_FIELDB(retry_count));
+}
+
+/**
+ * asd_dump_scb_sites -- dump currently used CSEQ SCB sites
+ * @asd_ha: pointer to host adapter struct
+ */
+void asd_dump_scb_sites(struct asd_ha_struct *asd_ha)
+{
+	u16	site_no;
+
+	for (site_no = 0; site_no < asd_ha->hw_prof.max_scbs; site_no++) {
+		u8 opcode;
+
+		if (!SCB_SITE_VALID(site_no))
+			continue;
+
+		/* We are only interested in SCB sites currently used.
+		 */
+		opcode = asd_scbsite_read_byte(asd_ha, site_no,
+					       offsetof(struct scb_header,
+							opcode));
+		if (opcode == 0xFF)
+			continue;
+
+		asd_printk("\nSCB: 0x%x\n", site_no);
+		asd_dump_scb_site(asd_ha, site_no);
+	}
+}
+
+/**
+ * ads_dump_seq_state -- dump CSEQ and LSEQ states
+ * @asd_ha: pointer to host adapter structure
+ * @lseq_mask: mask of LSEQs of interest
+ */
+void asd_dump_seq_state(struct asd_ha_struct *asd_ha, u8 lseq_mask)
+{
+	int lseq;
+
+	asd_dump_cseq_state(asd_ha);
+
+	if (lseq_mask != 0)
+		for_each_sequencer(lseq_mask, lseq_mask, lseq)
+			asd_dump_lseq_state(asd_ha, lseq);
+}
+
+void asd_dump_frame_rcvd(struct asd_phy *phy,
+			 struct done_list_struct *dl)
+{
+	unsigned long flags;
+	int i;
+
+	switch ((dl->status_block[1] & 0x70) >> 3) {
+	case SAS_PROTO_STP:
+		ASD_DPRINTK("STP proto device-to-host FIS:\n");
+		break;
+	default:
+	case SAS_PROTO_SSP:
+		ASD_DPRINTK("SAS proto IDENTIFY:\n");
+		break;
+	}
+	spin_lock_irqsave(&phy->sas_phy.frame_rcvd_lock, flags);
+	for (i = 0; i < phy->sas_phy.frame_rcvd_size; i+=4)
+		ASD_DPRINTK("%02x: %02x %02x %02x %02x\n",
+			    i,
+			    phy->frame_rcvd[i],
+			    phy->frame_rcvd[i+1],
+			    phy->frame_rcvd[i+2],
+			    phy->frame_rcvd[i+3]);
+	spin_unlock_irqrestore(&phy->sas_phy.frame_rcvd_lock, flags);
+}
+
+static inline void asd_dump_scb(struct asd_ascb *ascb, int ind)
+{
+	asd_printk("scb%d: vaddr: 0x%p, dma_handle: 0x%llx, next: 0x%llx, "
+		   "index:%d, opcode:0x%02x\n",
+		   ind, ascb->dma_scb.vaddr,
+		   (unsigned long long)ascb->dma_scb.dma_handle,
+		   (unsigned long long)
+		   le64_to_cpu(ascb->scb->header.next_scb),
+		   le16_to_cpu(ascb->scb->header.index),
+		   ascb->scb->header.opcode);
+}
+
+void asd_dump_scb_list(struct asd_ascb *ascb, int num)
+{
+	int i = 0;
+
+	asd_printk("dumping %d scbs:\n", num);
+
+	asd_dump_scb(ascb, i++);
+	--num;
+
+	if (num > 0 && !list_empty(&ascb->list)) {
+		struct list_head *el;
+
+		list_for_each(el, &ascb->list) {
+			struct asd_ascb *s = list_entry(el, struct asd_ascb,
+							list);
+			asd_dump_scb(s, i++);
+			if (--num <= 0)
+				break;
+		}
+	}
+}
+
+#endif /* ASD_DEBUG */
diff --git a/drivers/scsi/aic94xx/aic94xx_dump.h b/drivers/scsi/aic94xx/aic94xx_dump.h
new file mode 100644
index 000000000000..0c388e7da6bb
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx_dump.h
@@ -0,0 +1,52 @@
+/*
+ * Aic94xx SAS/SATA driver dump header file.
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#ifndef _AIC94XX_DUMP_H_
+#define _AIC94XX_DUMP_H_
+
+#ifdef ASD_DEBUG
+
+void asd_dump_ddb_0(struct asd_ha_struct *asd_ha);
+void asd_dump_target_ddb(struct asd_ha_struct *asd_ha, u16 site_no);
+void asd_dump_scb_sites(struct asd_ha_struct *asd_ha);
+void asd_dump_seq_state(struct asd_ha_struct *asd_ha, u8 lseq_mask);
+void asd_dump_frame_rcvd(struct asd_phy *phy,
+			 struct done_list_struct *dl);
+void asd_dump_scb_list(struct asd_ascb *ascb, int num);
+#else /* ASD_DEBUG */
+
+static inline void asd_dump_ddb_0(struct asd_ha_struct *asd_ha) { }
+static inline void asd_dump_target_ddb(struct asd_ha_struct *asd_ha,
+				     u16 site_no) { }
+static inline void asd_dump_scb_sites(struct asd_ha_struct *asd_ha) { }
+static inline void asd_dump_seq_state(struct asd_ha_struct *asd_ha,
+				      u8 lseq_mask) { }
+static inline void asd_dump_frame_rcvd(struct asd_phy *phy,
+				       struct done_list_struct *dl) { }
+static inline void asd_dump_scb_list(struct asd_ascb *ascb, int num) { }
+#endif /* ASD_DEBUG */
+
+#endif /* _AIC94XX_DUMP_H_ */
diff --git a/drivers/scsi/aic94xx/aic94xx_hwi.c b/drivers/scsi/aic94xx/aic94xx_hwi.c
new file mode 100644
index 000000000000..075cea85b56b
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx_hwi.c
@@ -0,0 +1,1376 @@
+/*
+ * Aic94xx SAS/SATA driver hardware interface.
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#include <linux/pci.h>
+#include <linux/delay.h>
+#include <linux/module.h>
+
+#include "aic94xx.h"
+#include "aic94xx_reg.h"
+#include "aic94xx_hwi.h"
+#include "aic94xx_seq.h"
+#include "aic94xx_dump.h"
+
+u32 MBAR0_SWB_SIZE;
+
+/* ---------- Initialization ---------- */
+
+static void asd_get_user_sas_addr(struct asd_ha_struct *asd_ha)
+{
+	extern char sas_addr_str[];
+	/* If the user has specified a WWN it overrides other settings
+	 */
+	if (sas_addr_str[0] != '\0')
+		asd_destringify_sas_addr(asd_ha->hw_prof.sas_addr,
+					 sas_addr_str);
+	else if (asd_ha->hw_prof.sas_addr[0] != 0)
+		asd_stringify_sas_addr(sas_addr_str, asd_ha->hw_prof.sas_addr);
+}
+
+static void asd_propagate_sas_addr(struct asd_ha_struct *asd_ha)
+{
+	int i;
+
+	for (i = 0; i < ASD_MAX_PHYS; i++) {
+		if (asd_ha->hw_prof.phy_desc[i].sas_addr[0] == 0)
+			continue;
+		/* Set a phy's address only if it has none.
+		 */
+		ASD_DPRINTK("setting phy%d addr to %llx\n", i,
+			    SAS_ADDR(asd_ha->hw_prof.sas_addr));
+		memcpy(asd_ha->hw_prof.phy_desc[i].sas_addr,
+		       asd_ha->hw_prof.sas_addr, SAS_ADDR_SIZE);
+	}
+}
+
+/* ---------- PHY initialization ---------- */
+
+static void asd_init_phy_identify(struct asd_phy *phy)
+{
+	phy->identify_frame = phy->id_frm_tok->vaddr;
+
+	memset(phy->identify_frame, 0, sizeof(*phy->identify_frame));
+
+	phy->identify_frame->dev_type = SAS_END_DEV;
+	if (phy->sas_phy.role & PHY_ROLE_INITIATOR)
+		phy->identify_frame->initiator_bits = phy->sas_phy.iproto;
+	if (phy->sas_phy.role & PHY_ROLE_TARGET)
+		phy->identify_frame->target_bits = phy->sas_phy.tproto;
+	memcpy(phy->identify_frame->sas_addr, phy->phy_desc->sas_addr,
+	       SAS_ADDR_SIZE);
+	phy->identify_frame->phy_id = phy->sas_phy.id;
+}
+
+static int asd_init_phy(struct asd_phy *phy)
+{
+	struct asd_ha_struct *asd_ha = phy->sas_phy.ha->lldd_ha;
+	struct asd_sas_phy *sas_phy = &phy->sas_phy;
+
+	sas_phy->enabled = 1;
+	sas_phy->class = SAS;
+	sas_phy->iproto = SAS_PROTO_ALL;
+	sas_phy->tproto = 0;
+	sas_phy->type = PHY_TYPE_PHYSICAL;
+	sas_phy->role = PHY_ROLE_INITIATOR;
+	sas_phy->oob_mode = OOB_NOT_CONNECTED;
+	sas_phy->linkrate = PHY_LINKRATE_NONE;
+
+	phy->id_frm_tok = asd_alloc_coherent(asd_ha,
+					     sizeof(*phy->identify_frame),
+					     GFP_KERNEL);
+	if (!phy->id_frm_tok) {
+		asd_printk("no mem for IDENTIFY for phy%d\n", sas_phy->id);
+		return -ENOMEM;
+	} else
+		asd_init_phy_identify(phy);
+
+	memset(phy->frame_rcvd, 0, sizeof(phy->frame_rcvd));
+
+	return 0;
+}
+
+static int asd_init_phys(struct asd_ha_struct *asd_ha)
+{
+	u8 i;
+	u8 phy_mask = asd_ha->hw_prof.enabled_phys;
+
+	for (i = 0; i < ASD_MAX_PHYS; i++) {
+		struct asd_phy *phy = &asd_ha->phys[i];
+
+		phy->phy_desc = &asd_ha->hw_prof.phy_desc[i];
+
+		phy->sas_phy.enabled = 0;
+		phy->sas_phy.id = i;
+		phy->sas_phy.sas_addr = &phy->phy_desc->sas_addr[0];
+		phy->sas_phy.frame_rcvd = &phy->frame_rcvd[0];
+		phy->sas_phy.ha = &asd_ha->sas_ha;
+		phy->sas_phy.lldd_phy = phy;
+	}
+
+	/* Now enable and initialize only the enabled phys. */
+	for_each_phy(phy_mask, phy_mask, i) {
+		int err = asd_init_phy(&asd_ha->phys[i]);
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
+/* ---------- Sliding windows ---------- */
+
+static int asd_init_sw(struct asd_ha_struct *asd_ha)
+{
+	struct pci_dev *pcidev = asd_ha->pcidev;
+	int err;
+	u32 v;
+
+	/* Unlock MBARs */
+	err = pci_read_config_dword(pcidev, PCI_CONF_MBAR_KEY, &v);
+	if (err) {
+		asd_printk("couldn't access conf. space of %s\n",
+			   pci_name(pcidev));
+		goto Err;
+	}
+	if (v)
+		err = pci_write_config_dword(pcidev, PCI_CONF_MBAR_KEY, v);
+	if (err) {
+		asd_printk("couldn't write to MBAR_KEY of %s\n",
+			   pci_name(pcidev));
+		goto Err;
+	}
+
+	/* Set sliding windows A, B and C to point to proper internal
+	 * memory regions.
+	 */
+	pci_write_config_dword(pcidev, PCI_CONF_MBAR0_SWA, REG_BASE_ADDR);
+	pci_write_config_dword(pcidev, PCI_CONF_MBAR0_SWB,
+			       REG_BASE_ADDR_CSEQCIO);
+	pci_write_config_dword(pcidev, PCI_CONF_MBAR0_SWC, REG_BASE_ADDR_EXSI);
+	asd_ha->io_handle[0].swa_base = REG_BASE_ADDR;
+	asd_ha->io_handle[0].swb_base = REG_BASE_ADDR_CSEQCIO;
+	asd_ha->io_handle[0].swc_base = REG_BASE_ADDR_EXSI;
+	MBAR0_SWB_SIZE = asd_ha->io_handle[0].len - 0x80;
+	if (!asd_ha->iospace) {
+		/* MBAR1 will point to OCM (On Chip Memory) */
+		pci_write_config_dword(pcidev, PCI_CONF_MBAR1, OCM_BASE_ADDR);
+		asd_ha->io_handle[1].swa_base = OCM_BASE_ADDR;
+	}
+	spin_lock_init(&asd_ha->iolock);
+Err:
+	return err;
+}
+
+/* ---------- SCB initialization ---------- */
+
+/**
+ * asd_init_scbs - manually allocate the first SCB.
+ * @asd_ha: pointer to host adapter structure
+ *
+ * This allocates the very first SCB which would be sent to the
+ * sequencer for execution.  Its bus address is written to
+ * CSEQ_Q_NEW_POINTER, mode page 2, mode 8.  Since the bus address of
+ * the _next_ scb to be DMA-ed to the host adapter is read from the last
+ * SCB DMA-ed to the host adapter, we have to always stay one step
+ * ahead of the sequencer and keep one SCB already allocated.
+ */
+static int asd_init_scbs(struct asd_ha_struct *asd_ha)
+{
+	struct asd_seq_data *seq = &asd_ha->seq;
+	int bitmap_bytes;
+
+	/* allocate the index array and bitmap */
+	asd_ha->seq.tc_index_bitmap_bits = asd_ha->hw_prof.max_scbs;
+	asd_ha->seq.tc_index_array = kzalloc(asd_ha->seq.tc_index_bitmap_bits*
+					     sizeof(void *), GFP_KERNEL);
+	if (!asd_ha->seq.tc_index_array)
+		return -ENOMEM;
+
+	bitmap_bytes = (asd_ha->seq.tc_index_bitmap_bits+7)/8;
+	bitmap_bytes = BITS_TO_LONGS(bitmap_bytes*8)*sizeof(unsigned long);
+	asd_ha->seq.tc_index_bitmap = kzalloc(bitmap_bytes, GFP_KERNEL);
+	if (!asd_ha->seq.tc_index_bitmap)
+		return -ENOMEM;
+
+	spin_lock_init(&seq->tc_index_lock);
+
+	seq->next_scb.size = sizeof(struct scb);
+	seq->next_scb.vaddr = dma_pool_alloc(asd_ha->scb_pool, GFP_KERNEL,
+					     &seq->next_scb.dma_handle);
+	if (!seq->next_scb.vaddr) {
+		kfree(asd_ha->seq.tc_index_bitmap);
+		kfree(asd_ha->seq.tc_index_array);
+		asd_ha->seq.tc_index_bitmap = NULL;
+		asd_ha->seq.tc_index_array = NULL;
+		return -ENOMEM;
+	}
+
+	seq->pending = 0;
+	spin_lock_init(&seq->pend_q_lock);
+	INIT_LIST_HEAD(&seq->pend_q);
+
+	return 0;
+}
+
+static inline void asd_get_max_scb_ddb(struct asd_ha_struct *asd_ha)
+{
+	asd_ha->hw_prof.max_scbs = asd_get_cmdctx_size(asd_ha)/ASD_SCB_SIZE;
+	asd_ha->hw_prof.max_ddbs = asd_get_devctx_size(asd_ha)/ASD_DDB_SIZE;
+	ASD_DPRINTK("max_scbs:%d, max_ddbs:%d\n",
+		    asd_ha->hw_prof.max_scbs,
+		    asd_ha->hw_prof.max_ddbs);
+}
+
+/* ---------- Done List initialization ---------- */
+
+static void asd_dl_tasklet_handler(unsigned long);
+
+static int asd_init_dl(struct asd_ha_struct *asd_ha)
+{
+	asd_ha->seq.actual_dl
+		= asd_alloc_coherent(asd_ha,
+			     ASD_DL_SIZE * sizeof(struct done_list_struct),
+				     GFP_KERNEL);
+	if (!asd_ha->seq.actual_dl)
+		return -ENOMEM;
+	asd_ha->seq.dl = asd_ha->seq.actual_dl->vaddr;
+	asd_ha->seq.dl_toggle = ASD_DEF_DL_TOGGLE;
+	asd_ha->seq.dl_next = 0;
+	tasklet_init(&asd_ha->seq.dl_tasklet, asd_dl_tasklet_handler,
+		     (unsigned long) asd_ha);
+
+	return 0;
+}
+
+/* ---------- EDB and ESCB init ---------- */
+
+static int asd_alloc_edbs(struct asd_ha_struct *asd_ha, unsigned int gfp_flags)
+{
+	struct asd_seq_data *seq = &asd_ha->seq;
+	int i;
+
+	seq->edb_arr = kmalloc(seq->num_edbs*sizeof(*seq->edb_arr), gfp_flags);
+	if (!seq->edb_arr)
+		return -ENOMEM;
+
+	for (i = 0; i < seq->num_edbs; i++) {
+		seq->edb_arr[i] = asd_alloc_coherent(asd_ha, ASD_EDB_SIZE,
+						     gfp_flags);
+		if (!seq->edb_arr[i])
+			goto Err_unroll;
+		memset(seq->edb_arr[i]->vaddr, 0, ASD_EDB_SIZE);
+	}
+
+	ASD_DPRINTK("num_edbs:%d\n", seq->num_edbs);
+
+	return 0;
+
+Err_unroll:
+	for (i-- ; i >= 0; i--)
+		asd_free_coherent(asd_ha, seq->edb_arr[i]);
+	kfree(seq->edb_arr);
+	seq->edb_arr = NULL;
+
+	return -ENOMEM;
+}
+
+static int asd_alloc_escbs(struct asd_ha_struct *asd_ha,
+			   unsigned int gfp_flags)
+{
+	struct asd_seq_data *seq = &asd_ha->seq;
+	struct asd_ascb *escb;
+	int i, escbs;
+
+	seq->escb_arr = kmalloc(seq->num_escbs*sizeof(*seq->escb_arr),
+				gfp_flags);
+	if (!seq->escb_arr)
+		return -ENOMEM;
+
+	escbs = seq->num_escbs;
+	escb = asd_ascb_alloc_list(asd_ha, &escbs, gfp_flags);
+	if (!escb) {
+		asd_printk("couldn't allocate list of escbs\n");
+		goto Err;
+	}
+	seq->num_escbs -= escbs;  /* subtract what was not allocated */
+	ASD_DPRINTK("num_escbs:%d\n", seq->num_escbs);
+
+	for (i = 0; i < seq->num_escbs; i++, escb = list_entry(escb->list.next,
+							       struct asd_ascb,
+							       list)) {
+		seq->escb_arr[i] = escb;
+		escb->scb->header.opcode = EMPTY_SCB;
+	}
+
+	return 0;
+Err:
+	kfree(seq->escb_arr);
+	seq->escb_arr = NULL;
+	return -ENOMEM;
+
+}
+
+static void asd_assign_edbs2escbs(struct asd_ha_struct *asd_ha)
+{
+	struct asd_seq_data *seq = &asd_ha->seq;
+	int i, k, z = 0;
+
+	for (i = 0; i < seq->num_escbs; i++) {
+		struct asd_ascb *ascb = seq->escb_arr[i];
+		struct empty_scb *escb = &ascb->scb->escb;
+
+		ascb->edb_index = z;
+
+		escb->num_valid = ASD_EDBS_PER_SCB;
+
+		for (k = 0; k < ASD_EDBS_PER_SCB; k++) {
+			struct sg_el *eb = &escb->eb[k];
+			struct asd_dma_tok *edb = seq->edb_arr[z++];
+
+			memset(eb, 0, sizeof(*eb));
+			eb->bus_addr = cpu_to_le64(((u64) edb->dma_handle));
+			eb->size = cpu_to_le32(((u32) edb->size));
+		}
+	}
+}
+
+/**
+ * asd_init_escbs -- allocate and initialize empty scbs
+ * @asd_ha: pointer to host adapter structure
+ *
+ * An empty SCB has sg_elements of ASD_EDBS_PER_SCB (7) buffers.
+ * They transport sense data, etc.
+ */
+static int asd_init_escbs(struct asd_ha_struct *asd_ha)
+{
+	struct asd_seq_data *seq = &asd_ha->seq;
+	int err = 0;
+
+	/* Allocate two empty data buffers (edb) per sequencer. */
+	int edbs = 2*(1+asd_ha->hw_prof.num_phys);
+
+	seq->num_escbs = (edbs+ASD_EDBS_PER_SCB-1)/ASD_EDBS_PER_SCB;
+	seq->num_edbs = seq->num_escbs * ASD_EDBS_PER_SCB;
+
+	err = asd_alloc_edbs(asd_ha, GFP_KERNEL);
+	if (err) {
+		asd_printk("couldn't allocate edbs\n");
+		return err;
+	}
+
+	err = asd_alloc_escbs(asd_ha, GFP_KERNEL);
+	if (err) {
+		asd_printk("couldn't allocate escbs\n");
+		return err;
+	}
+
+	asd_assign_edbs2escbs(asd_ha);
+	/* In order to insure that normal SCBs do not overfill sequencer
+	 * memory and leave no space for escbs (halting condition),
+	 * we increment pending here by the number of escbs.  However,
+	 * escbs are never pending.
+	 */
+	seq->pending   = seq->num_escbs;
+	seq->can_queue = 1 + (asd_ha->hw_prof.max_scbs - seq->pending)/2;
+
+	return 0;
+}
+
+/* ---------- HW initialization ---------- */
+
+/**
+ * asd_chip_hardrst -- hard reset the chip
+ * @asd_ha: pointer to host adapter structure
+ *
+ * This takes 16 cycles and is synchronous to CFCLK, which runs
+ * at 200 MHz, so this should take at most 80 nanoseconds.
+ */
+int asd_chip_hardrst(struct asd_ha_struct *asd_ha)
+{
+	int i;
+	int count = 100;
+	u32 reg;
+
+	for (i = 0 ; i < 4 ; i++) {
+		asd_write_reg_dword(asd_ha, COMBIST, HARDRST);
+	}
+
+	do {
+		udelay(1);
+		reg = asd_read_reg_dword(asd_ha, CHIMINT);
+		if (reg & HARDRSTDET) {
+			asd_write_reg_dword(asd_ha, CHIMINT,
+					    HARDRSTDET|PORRSTDET);
+			return 0;
+		}
+	} while (--count > 0);
+
+	return -ENODEV;
+}
+
+/**
+ * asd_init_chip -- initialize the chip
+ * @asd_ha: pointer to host adapter structure
+ *
+ * Hard resets the chip, disables HA interrupts, downloads the sequnecer
+ * microcode and starts the sequencers.  The caller has to explicitly
+ * enable HA interrupts with asd_enable_ints(asd_ha).
+ */
+static int asd_init_chip(struct asd_ha_struct *asd_ha)
+{
+	int err;
+
+	err = asd_chip_hardrst(asd_ha);
+	if (err) {
+		asd_printk("couldn't hard reset %s\n",
+			    pci_name(asd_ha->pcidev));
+		goto out;
+	}
+
+	asd_disable_ints(asd_ha);
+
+	err = asd_init_seqs(asd_ha);
+	if (err) {
+		asd_printk("couldn't init seqs for %s\n",
+			   pci_name(asd_ha->pcidev));
+		goto out;
+	}
+
+	err = asd_start_seqs(asd_ha);
+	if (err) {
+		asd_printk("coudln't start seqs for %s\n",
+			   pci_name(asd_ha->pcidev));
+		goto out;
+	}
+out:
+	return err;
+}
+
+#define MAX_DEVS ((OCM_MAX_SIZE) / (ASD_DDB_SIZE))
+
+static int max_devs = 0;
+module_param_named(max_devs, max_devs, int, S_IRUGO);
+MODULE_PARM_DESC(max_devs, "\n"
+	"\tMaximum number of SAS devices to support (not LUs).\n"
+	"\tDefault: 2176, Maximum: 65663.\n");
+
+static int max_cmnds = 0;
+module_param_named(max_cmnds, max_cmnds, int, S_IRUGO);
+MODULE_PARM_DESC(max_cmnds, "\n"
+	"\tMaximum number of commands queuable.\n"
+	"\tDefault: 512, Maximum: 66047.\n");
+
+static void asd_extend_devctx_ocm(struct asd_ha_struct *asd_ha)
+{
+	unsigned long dma_addr = OCM_BASE_ADDR;
+	u32 d;
+
+	dma_addr -= asd_ha->hw_prof.max_ddbs * ASD_DDB_SIZE;
+	asd_write_reg_addr(asd_ha, DEVCTXBASE, (dma_addr_t) dma_addr);
+	d = asd_read_reg_dword(asd_ha, CTXDOMAIN);
+	d |= 4;
+	asd_write_reg_dword(asd_ha, CTXDOMAIN, d);
+	asd_ha->hw_prof.max_ddbs += MAX_DEVS;
+}
+
+static int asd_extend_devctx(struct asd_ha_struct *asd_ha)
+{
+	dma_addr_t dma_handle;
+	unsigned long dma_addr;
+	u32 d;
+	int size;
+
+	asd_extend_devctx_ocm(asd_ha);
+
+	asd_ha->hw_prof.ddb_ext = NULL;
+	if (max_devs <= asd_ha->hw_prof.max_ddbs || max_devs > 0xFFFF) {
+		max_devs = asd_ha->hw_prof.max_ddbs;
+		return 0;
+	}
+
+	size = (max_devs - asd_ha->hw_prof.max_ddbs + 1) * ASD_DDB_SIZE;
+
+	asd_ha->hw_prof.ddb_ext = asd_alloc_coherent(asd_ha, size, GFP_KERNEL);
+	if (!asd_ha->hw_prof.ddb_ext) {
+		asd_printk("couldn't allocate memory for %d devices\n",
+			   max_devs);
+		max_devs = asd_ha->hw_prof.max_ddbs;
+		return -ENOMEM;
+	}
+	dma_handle = asd_ha->hw_prof.ddb_ext->dma_handle;
+	dma_addr = ALIGN((unsigned long) dma_handle, ASD_DDB_SIZE);
+	dma_addr -= asd_ha->hw_prof.max_ddbs * ASD_DDB_SIZE;
+	dma_handle = (dma_addr_t) dma_addr;
+	asd_write_reg_addr(asd_ha, DEVCTXBASE, dma_handle);
+	d = asd_read_reg_dword(asd_ha, CTXDOMAIN);
+	d &= ~4;
+	asd_write_reg_dword(asd_ha, CTXDOMAIN, d);
+
+	asd_ha->hw_prof.max_ddbs = max_devs;
+
+	return 0;
+}
+
+static int asd_extend_cmdctx(struct asd_ha_struct *asd_ha)
+{
+	dma_addr_t dma_handle;
+	unsigned long dma_addr;
+	u32 d;
+	int size;
+
+	asd_ha->hw_prof.scb_ext = NULL;
+	if (max_cmnds <= asd_ha->hw_prof.max_scbs || max_cmnds > 0xFFFF) {
+		max_cmnds = asd_ha->hw_prof.max_scbs;
+		return 0;
+	}
+
+	size = (max_cmnds - asd_ha->hw_prof.max_scbs + 1) * ASD_SCB_SIZE;
+
+	asd_ha->hw_prof.scb_ext = asd_alloc_coherent(asd_ha, size, GFP_KERNEL);
+	if (!asd_ha->hw_prof.scb_ext) {
+		asd_printk("couldn't allocate memory for %d commands\n",
+			   max_cmnds);
+		max_cmnds = asd_ha->hw_prof.max_scbs;
+		return -ENOMEM;
+	}
+	dma_handle = asd_ha->hw_prof.scb_ext->dma_handle;
+	dma_addr = ALIGN((unsigned long) dma_handle, ASD_SCB_SIZE);
+	dma_addr -= asd_ha->hw_prof.max_scbs * ASD_SCB_SIZE;
+	dma_handle = (dma_addr_t) dma_addr;
+	asd_write_reg_addr(asd_ha, CMDCTXBASE, dma_handle);
+	d = asd_read_reg_dword(asd_ha, CTXDOMAIN);
+	d &= ~1;
+	asd_write_reg_dword(asd_ha, CTXDOMAIN, d);
+
+	asd_ha->hw_prof.max_scbs = max_cmnds;
+
+	return 0;
+}
+
+/**
+ * asd_init_ctxmem -- initialize context memory
+ * asd_ha: pointer to host adapter structure
+ *
+ * This function sets the maximum number of SCBs and
+ * DDBs which can be used by the sequencer.  This is normally
+ * 512 and 128 respectively.  If support for more SCBs or more DDBs
+ * is required then CMDCTXBASE, DEVCTXBASE and CTXDOMAIN are
+ * initialized here to extend context memory to point to host memory,
+ * thus allowing unlimited support for SCBs and DDBs -- only limited
+ * by host memory.
+ */
+static int asd_init_ctxmem(struct asd_ha_struct *asd_ha)
+{
+	int bitmap_bytes;
+
+	asd_get_max_scb_ddb(asd_ha);
+	asd_extend_devctx(asd_ha);
+	asd_extend_cmdctx(asd_ha);
+
+	/* The kernel wants bitmaps to be unsigned long sized. */
+	bitmap_bytes = (asd_ha->hw_prof.max_ddbs+7)/8;
+	bitmap_bytes = BITS_TO_LONGS(bitmap_bytes*8)*sizeof(unsigned long);
+	asd_ha->hw_prof.ddb_bitmap = kzalloc(bitmap_bytes, GFP_KERNEL);
+	if (!asd_ha->hw_prof.ddb_bitmap)
+		return -ENOMEM;
+	spin_lock_init(&asd_ha->hw_prof.ddb_lock);
+
+	return 0;
+}
+
+int asd_init_hw(struct asd_ha_struct *asd_ha)
+{
+	int err;
+	u32 v;
+
+	err = asd_init_sw(asd_ha);
+	if (err)
+		return err;
+
+	err = pci_read_config_dword(asd_ha->pcidev, PCIC_HSTPCIX_CNTRL, &v);
+	if (err) {
+		asd_printk("couldn't read PCIC_HSTPCIX_CNTRL of %s\n",
+			   pci_name(asd_ha->pcidev));
+		return err;
+	}
+	pci_write_config_dword(asd_ha->pcidev, PCIC_HSTPCIX_CNTRL,
+					v | SC_TMR_DIS);
+	if (err) {
+		asd_printk("couldn't disable split completion timer of %s\n",
+			   pci_name(asd_ha->pcidev));
+		return err;
+	}
+
+	err = asd_read_ocm(asd_ha);
+	if (err) {
+		asd_printk("couldn't read ocm(%d)\n", err);
+		/* While suspicios, it is not an error that we
+		 * couldn't read the OCM. */
+	}
+
+	err = asd_read_flash(asd_ha);
+	if (err) {
+		asd_printk("couldn't read flash(%d)\n", err);
+		/* While suspicios, it is not an error that we
+		 * couldn't read FLASH memory.
+		 */
+	}
+
+	asd_init_ctxmem(asd_ha);
+
+	asd_get_user_sas_addr(asd_ha);
+	if (!asd_ha->hw_prof.sas_addr[0]) {
+		asd_printk("No SAS Address provided for %s\n",
+			   pci_name(asd_ha->pcidev));
+		err = -ENODEV;
+		goto Out;
+	}
+
+	asd_propagate_sas_addr(asd_ha);
+
+	err = asd_init_phys(asd_ha);
+	if (err) {
+		asd_printk("couldn't initialize phys for %s\n",
+			    pci_name(asd_ha->pcidev));
+		goto Out;
+	}
+
+	err = asd_init_scbs(asd_ha);
+	if (err) {
+		asd_printk("couldn't initialize scbs for %s\n",
+			    pci_name(asd_ha->pcidev));
+		goto Out;
+	}
+
+	err = asd_init_dl(asd_ha);
+	if (err) {
+		asd_printk("couldn't initialize the done list:%d\n",
+			    err);
+		goto Out;
+	}
+
+	err = asd_init_escbs(asd_ha);
+	if (err) {
+		asd_printk("couldn't initialize escbs\n");
+		goto Out;
+	}
+
+	err = asd_init_chip(asd_ha);
+	if (err) {
+		asd_printk("couldn't init the chip\n");
+		goto Out;
+	}
+Out:
+	return err;
+}
+
+/* ---------- Chip reset ---------- */
+
+/**
+ * asd_chip_reset -- reset the host adapter, etc
+ * @asd_ha: pointer to host adapter structure of interest
+ *
+ * Called from the ISR.  Hard reset the chip.  Let everything
+ * timeout.  This should be no different than hot-unplugging the
+ * host adapter.  Once everything times out we'll init the chip with
+ * a call to asd_init_chip() and enable interrupts with asd_enable_ints().
+ * XXX finish.
+ */
+static void asd_chip_reset(struct asd_ha_struct *asd_ha)
+{
+	struct sas_ha_struct *sas_ha = &asd_ha->sas_ha;
+
+	ASD_DPRINTK("chip reset for %s\n", pci_name(asd_ha->pcidev));
+	asd_chip_hardrst(asd_ha);
+	sas_ha->notify_ha_event(sas_ha, HAE_RESET);
+}
+
+/* ---------- Done List Routines ---------- */
+
+static void asd_dl_tasklet_handler(unsigned long data)
+{
+	struct asd_ha_struct *asd_ha = (struct asd_ha_struct *) data;
+	struct asd_seq_data *seq = &asd_ha->seq;
+	unsigned long flags;
+
+	while (1) {
+		struct done_list_struct *dl = &seq->dl[seq->dl_next];
+		struct asd_ascb *ascb;
+
+		if ((dl->toggle & DL_TOGGLE_MASK) != seq->dl_toggle)
+			break;
+
+		/* find the aSCB */
+		spin_lock_irqsave(&seq->tc_index_lock, flags);
+		ascb = asd_tc_index_find(seq, (int)le16_to_cpu(dl->index));
+		spin_unlock_irqrestore(&seq->tc_index_lock, flags);
+		if (unlikely(!ascb)) {
+			ASD_DPRINTK("BUG:sequencer:dl:no ascb?!\n");
+			goto next_1;
+		} else if (ascb->scb->header.opcode == EMPTY_SCB) {
+			goto out;
+		} else if (!ascb->uldd_timer && !del_timer(&ascb->timer)) {
+			goto next_1;
+		}
+		spin_lock_irqsave(&seq->pend_q_lock, flags);
+		list_del_init(&ascb->list);
+		seq->pending--;
+		spin_unlock_irqrestore(&seq->pend_q_lock, flags);
+	out:
+		ascb->tasklet_complete(ascb, dl);
+
+	next_1:
+		seq->dl_next = (seq->dl_next + 1) & (ASD_DL_SIZE-1);
+		if (!seq->dl_next)
+			seq->dl_toggle ^= DL_TOGGLE_MASK;
+	}
+}
+
+/* ---------- Interrupt Service Routines ---------- */
+
+/**
+ * asd_process_donelist_isr -- schedule processing of done list entries
+ * @asd_ha: pointer to host adapter structure
+ */
+static inline void asd_process_donelist_isr(struct asd_ha_struct *asd_ha)
+{
+	tasklet_schedule(&asd_ha->seq.dl_tasklet);
+}
+
+/**
+ * asd_com_sas_isr -- process device communication interrupt (COMINT)
+ * @asd_ha: pointer to host adapter structure
+ */
+static inline void asd_com_sas_isr(struct asd_ha_struct *asd_ha)
+{
+	u32 comstat = asd_read_reg_dword(asd_ha, COMSTAT);
+
+	/* clear COMSTAT int */
+	asd_write_reg_dword(asd_ha, COMSTAT, 0xFFFFFFFF);
+
+	if (comstat & CSBUFPERR) {
+		asd_printk("%s: command/status buffer dma parity error\n",
+			   pci_name(asd_ha->pcidev));
+	} else if (comstat & CSERR) {
+		int i;
+		u32 dmaerr = asd_read_reg_dword(asd_ha, DMAERR);
+		dmaerr &= 0xFF;
+		asd_printk("%s: command/status dma error, DMAERR: 0x%02x, "
+			   "CSDMAADR: 0x%04x, CSDMAADR+4: 0x%04x\n",
+			   pci_name(asd_ha->pcidev),
+			   dmaerr,
+			   asd_read_reg_dword(asd_ha, CSDMAADR),
+			   asd_read_reg_dword(asd_ha, CSDMAADR+4));
+		asd_printk("CSBUFFER:\n");
+		for (i = 0; i < 8; i++) {
+			asd_printk("%08x %08x %08x %08x\n",
+				   asd_read_reg_dword(asd_ha, CSBUFFER),
+				   asd_read_reg_dword(asd_ha, CSBUFFER+4),
+				   asd_read_reg_dword(asd_ha, CSBUFFER+8),
+				   asd_read_reg_dword(asd_ha, CSBUFFER+12));
+		}
+		asd_dump_seq_state(asd_ha, 0);
+	} else if (comstat & OVLYERR) {
+		u32 dmaerr = asd_read_reg_dword(asd_ha, DMAERR);
+		dmaerr = (dmaerr >> 8) & 0xFF;
+		asd_printk("%s: overlay dma error:0x%x\n",
+			   pci_name(asd_ha->pcidev),
+			   dmaerr);
+	}
+	asd_chip_reset(asd_ha);
+}
+
+static inline void asd_arp2_err(struct asd_ha_struct *asd_ha, u32 dchstatus)
+{
+	static const char *halt_code[256] = {
+		"UNEXPECTED_INTERRUPT0",
+		"UNEXPECTED_INTERRUPT1",
+		"UNEXPECTED_INTERRUPT2",
+		"UNEXPECTED_INTERRUPT3",
+		"UNEXPECTED_INTERRUPT4",
+		"UNEXPECTED_INTERRUPT5",
+		"UNEXPECTED_INTERRUPT6",
+		"UNEXPECTED_INTERRUPT7",
+		"UNEXPECTED_INTERRUPT8",
+		"UNEXPECTED_INTERRUPT9",
+		"UNEXPECTED_INTERRUPT10",
+		[11 ... 19] = "unknown[11,19]",
+		"NO_FREE_SCB_AVAILABLE",
+		"INVALID_SCB_OPCODE",
+		"INVALID_MBX_OPCODE",
+		"INVALID_ATA_STATE",
+		"ATA_QUEUE_FULL",
+		"ATA_TAG_TABLE_FAULT",
+		"ATA_TAG_MASK_FAULT",
+		"BAD_LINK_QUEUE_STATE",
+		"DMA2CHIM_QUEUE_ERROR",
+		"EMPTY_SCB_LIST_FULL",
+		"unknown[30]",
+		"IN_USE_SCB_ON_FREE_LIST",
+		"BAD_OPEN_WAIT_STATE",
+		"INVALID_STP_AFFILIATION",
+		"unknown[34]",
+		"EXEC_QUEUE_ERROR",
+		"TOO_MANY_EMPTIES_NEEDED",
+		"EMPTY_REQ_QUEUE_ERROR",
+		"Q_MONIRTT_MGMT_ERROR",
+		"TARGET_MODE_FLOW_ERROR",
+		"DEVICE_QUEUE_NOT_FOUND",
+		"START_IRTT_TIMER_ERROR",
+		"ABORT_TASK_ILLEGAL_REQ",
+		[43 ... 255] = "unknown[43,255]"
+	};
+
+	if (dchstatus & CSEQINT) {
+		u32 arp2int = asd_read_reg_dword(asd_ha, CARP2INT);
+
+		if (arp2int & (ARP2WAITTO|ARP2ILLOPC|ARP2PERR|ARP2CIOPERR)) {
+			asd_printk("%s: CSEQ arp2int:0x%x\n",
+				   pci_name(asd_ha->pcidev),
+				   arp2int);
+		} else if (arp2int & ARP2HALTC)
+			asd_printk("%s: CSEQ halted: %s\n",
+				   pci_name(asd_ha->pcidev),
+				   halt_code[(arp2int>>16)&0xFF]);
+		else
+			asd_printk("%s: CARP2INT:0x%x\n",
+				   pci_name(asd_ha->pcidev),
+				   arp2int);
+	}
+	if (dchstatus & LSEQINT_MASK) {
+		int lseq;
+		u8  lseq_mask = dchstatus & LSEQINT_MASK;
+
+		for_each_sequencer(lseq_mask, lseq_mask, lseq) {
+			u32 arp2int = asd_read_reg_dword(asd_ha,
+							 LmARP2INT(lseq));
+			if (arp2int & (ARP2WAITTO | ARP2ILLOPC | ARP2PERR
+				       | ARP2CIOPERR)) {
+				asd_printk("%s: LSEQ%d arp2int:0x%x\n",
+					   pci_name(asd_ha->pcidev),
+					   lseq, arp2int);
+				/* XXX we should only do lseq reset */
+			} else if (arp2int & ARP2HALTC)
+				asd_printk("%s: LSEQ%d halted: %s\n",
+					   pci_name(asd_ha->pcidev),
+					   lseq,halt_code[(arp2int>>16)&0xFF]);
+			else
+				asd_printk("%s: LSEQ%d ARP2INT:0x%x\n",
+					   pci_name(asd_ha->pcidev), lseq,
+					   arp2int);
+		}
+	}
+	asd_chip_reset(asd_ha);
+}
+
+/**
+ * asd_dch_sas_isr -- process device channel interrupt (DEVINT)
+ * @asd_ha: pointer to host adapter structure
+ */
+static inline void asd_dch_sas_isr(struct asd_ha_struct *asd_ha)
+{
+	u32 dchstatus = asd_read_reg_dword(asd_ha, DCHSTATUS);
+
+	if (dchstatus & CFIFTOERR) {
+		asd_printk("%s: CFIFTOERR\n", pci_name(asd_ha->pcidev));
+		asd_chip_reset(asd_ha);
+	} else
+		asd_arp2_err(asd_ha, dchstatus);
+}
+
+/**
+ * ads_rbi_exsi_isr -- process external system interface interrupt (INITERR)
+ * @asd_ha: pointer to host adapter structure
+ */
+static inline void asd_rbi_exsi_isr(struct asd_ha_struct *asd_ha)
+{
+	u32 stat0r = asd_read_reg_dword(asd_ha, ASISTAT0R);
+
+	if (!(stat0r & ASIERR)) {
+		asd_printk("hmm, EXSI interrupted but no error?\n");
+		return;
+	}
+
+	if (stat0r & ASIFMTERR) {
+		asd_printk("ASI SEEPROM format error for %s\n",
+			   pci_name(asd_ha->pcidev));
+	} else if (stat0r & ASISEECHKERR) {
+		u32 stat1r = asd_read_reg_dword(asd_ha, ASISTAT1R);
+		asd_printk("ASI SEEPROM checksum 0x%x error for %s\n",
+			   stat1r & CHECKSUM_MASK,
+			   pci_name(asd_ha->pcidev));
+	} else {
+		u32 statr = asd_read_reg_dword(asd_ha, ASIERRSTATR);
+
+		if (!(statr & CPI2ASIMSTERR_MASK)) {
+			ASD_DPRINTK("hmm, ASIERR?\n");
+			return;
+		} else {
+			u32 addr = asd_read_reg_dword(asd_ha, ASIERRADDR);
+			u32 data = asd_read_reg_dword(asd_ha, ASIERRDATAR);
+
+			asd_printk("%s: CPI2 xfer err: addr: 0x%x, wdata: 0x%x, "
+				   "count: 0x%x, byteen: 0x%x, targerr: 0x%x "
+				   "master id: 0x%x, master err: 0x%x\n",
+				   pci_name(asd_ha->pcidev),
+				   addr, data,
+				   (statr & CPI2ASIBYTECNT_MASK) >> 16,
+				   (statr & CPI2ASIBYTEEN_MASK) >> 12,
+				   (statr & CPI2ASITARGERR_MASK) >> 8,
+				   (statr & CPI2ASITARGMID_MASK) >> 4,
+				   (statr & CPI2ASIMSTERR_MASK));
+		}
+	}
+	asd_chip_reset(asd_ha);
+}
+
+/**
+ * asd_hst_pcix_isr -- process host interface interrupts
+ * @asd_ha: pointer to host adapter structure
+ *
+ * Asserted on PCIX errors: target abort, etc.
+ */
+static inline void asd_hst_pcix_isr(struct asd_ha_struct *asd_ha)
+{
+	u16 status;
+	u32 pcix_status;
+	u32 ecc_status;
+
+	pci_read_config_word(asd_ha->pcidev, PCI_STATUS, &status);
+	pci_read_config_dword(asd_ha->pcidev, PCIX_STATUS, &pcix_status);
+	pci_read_config_dword(asd_ha->pcidev, ECC_CTRL_STAT, &ecc_status);
+
+	if (status & PCI_STATUS_DETECTED_PARITY)
+		asd_printk("parity error for %s\n", pci_name(asd_ha->pcidev));
+	else if (status & PCI_STATUS_REC_MASTER_ABORT)
+		asd_printk("master abort for %s\n", pci_name(asd_ha->pcidev));
+	else if (status & PCI_STATUS_REC_TARGET_ABORT)
+		asd_printk("target abort for %s\n", pci_name(asd_ha->pcidev));
+	else if (status & PCI_STATUS_PARITY)
+		asd_printk("data parity for %s\n", pci_name(asd_ha->pcidev));
+	else if (pcix_status & RCV_SCE) {
+		asd_printk("received split completion error for %s\n",
+			   pci_name(asd_ha->pcidev));
+		pci_write_config_dword(asd_ha->pcidev,PCIX_STATUS,pcix_status);
+		/* XXX: Abort task? */
+		return;
+	} else if (pcix_status & UNEXP_SC) {
+		asd_printk("unexpected split completion for %s\n",
+			   pci_name(asd_ha->pcidev));
+		pci_write_config_dword(asd_ha->pcidev,PCIX_STATUS,pcix_status);
+		/* ignore */
+		return;
+	} else if (pcix_status & SC_DISCARD)
+		asd_printk("split completion discarded for %s\n",
+			   pci_name(asd_ha->pcidev));
+	else if (ecc_status & UNCOR_ECCERR)
+		asd_printk("uncorrectable ECC error for %s\n",
+			   pci_name(asd_ha->pcidev));
+	asd_chip_reset(asd_ha);
+}
+
+/**
+ * asd_hw_isr -- host adapter interrupt service routine
+ * @irq: ignored
+ * @dev_id: pointer to host adapter structure
+ * @regs: ignored
+ *
+ * The ISR processes done list entries and level 3 error handling.
+ */
+irqreturn_t asd_hw_isr(int irq, void *dev_id, struct pt_regs *regs)
+{
+	struct asd_ha_struct *asd_ha = dev_id;
+	u32 chimint = asd_read_reg_dword(asd_ha, CHIMINT);
+
+	if (!chimint)
+		return IRQ_NONE;
+
+	asd_write_reg_dword(asd_ha, CHIMINT, chimint);
+	(void) asd_read_reg_dword(asd_ha, CHIMINT);
+
+	if (chimint & DLAVAIL)
+		asd_process_donelist_isr(asd_ha);
+	if (chimint & COMINT)
+		asd_com_sas_isr(asd_ha);
+	if (chimint & DEVINT)
+		asd_dch_sas_isr(asd_ha);
+	if (chimint & INITERR)
+		asd_rbi_exsi_isr(asd_ha);
+	if (chimint & HOSTERR)
+		asd_hst_pcix_isr(asd_ha);
+
+	return IRQ_HANDLED;
+}
+
+/* ---------- SCB handling ---------- */
+
+static inline struct asd_ascb *asd_ascb_alloc(struct asd_ha_struct *asd_ha,
+					      unsigned int gfp_flags)
+{
+	extern kmem_cache_t *asd_ascb_cache;
+	struct asd_seq_data *seq = &asd_ha->seq;
+	struct asd_ascb *ascb;
+	unsigned long flags;
+
+	ascb = kmem_cache_alloc(asd_ascb_cache, gfp_flags);
+
+	if (ascb) {
+		memset(ascb, 0, sizeof(*ascb));
+		ascb->dma_scb.size = sizeof(struct scb);
+		ascb->dma_scb.vaddr = dma_pool_alloc(asd_ha->scb_pool,
+						     gfp_flags,
+						    &ascb->dma_scb.dma_handle);
+		if (!ascb->dma_scb.vaddr) {
+			kmem_cache_free(asd_ascb_cache, ascb);
+			return NULL;
+		}
+		memset(ascb->dma_scb.vaddr, 0, sizeof(struct scb));
+		asd_init_ascb(asd_ha, ascb);
+
+		spin_lock_irqsave(&seq->tc_index_lock, flags);
+		ascb->tc_index = asd_tc_index_get(seq, ascb);
+		spin_unlock_irqrestore(&seq->tc_index_lock, flags);
+		if (ascb->tc_index == -1)
+			goto undo;
+
+		ascb->scb->header.index = cpu_to_le16((u16)ascb->tc_index);
+	}
+
+	return ascb;
+undo:
+	dma_pool_free(asd_ha->scb_pool, ascb->dma_scb.vaddr,
+		      ascb->dma_scb.dma_handle);
+	kmem_cache_free(asd_ascb_cache, ascb);
+	ASD_DPRINTK("no index for ascb\n");
+	return NULL;
+}
+
+/**
+ * asd_ascb_alloc_list -- allocate a list of aSCBs
+ * @asd_ha: pointer to host adapter structure
+ * @num: pointer to integer number of aSCBs
+ * @gfp_flags: GFP_ flags.
+ *
+ * This is the only function which is used to allocate aSCBs.
+ * It can allocate one or many. If more than one, then they form
+ * a linked list in two ways: by their list field of the ascb struct
+ * and by the next_scb field of the scb_header.
+ *
+ * Returns NULL if no memory was available, else pointer to a list
+ * of ascbs.  When this function returns, @num would be the number
+ * of SCBs which were not able to be allocated, 0 if all requested
+ * were able to be allocated.
+ */
+struct asd_ascb *asd_ascb_alloc_list(struct asd_ha_struct
+				     *asd_ha, int *num,
+				     unsigned int gfp_flags)
+{
+	struct asd_ascb *first = NULL;
+
+	for ( ; *num > 0; --*num) {
+		struct asd_ascb *ascb = asd_ascb_alloc(asd_ha, gfp_flags);
+
+		if (!ascb)
+			break;
+		else if (!first)
+			first = ascb;
+		else {
+			struct asd_ascb *last = list_entry(first->list.prev,
+							   struct asd_ascb,
+							   list);
+			list_add_tail(&ascb->list, &first->list);
+			last->scb->header.next_scb =
+				cpu_to_le64(((u64)ascb->dma_scb.dma_handle));
+		}
+	}
+
+	return first;
+}
+
+/**
+ * asd_swap_head_scb -- swap the head scb
+ * @asd_ha: pointer to host adapter structure
+ * @ascb: pointer to the head of an ascb list
+ *
+ * The sequencer knows the DMA address of the next SCB to be DMAed to
+ * the host adapter, from initialization or from the last list DMAed.
+ * seq->next_scb keeps the address of this SCB.  The sequencer will
+ * DMA to the host adapter this list of SCBs.  But the head (first
+ * element) of this list is not known to the sequencer.  Here we swap
+ * the head of the list with the known SCB (memcpy()).
+ * Only one memcpy() is required per list so it is in our interest
+ * to keep the list of SCB as long as possible so that the ratio
+ * of number of memcpy calls to the number of SCB DMA-ed is as small
+ * as possible.
+ *
+ * LOCKING: called with the pending list lock held.
+ */
+static inline void asd_swap_head_scb(struct asd_ha_struct *asd_ha,
+				     struct asd_ascb *ascb)
+{
+	struct asd_seq_data *seq = &asd_ha->seq;
+	struct asd_ascb *last = list_entry(ascb->list.prev,
+					   struct asd_ascb,
+					   list);
+	struct asd_dma_tok t = ascb->dma_scb;
+
+	memcpy(seq->next_scb.vaddr, ascb->scb, sizeof(*ascb->scb));
+	ascb->dma_scb = seq->next_scb;
+	ascb->scb = ascb->dma_scb.vaddr;
+	seq->next_scb = t;
+	last->scb->header.next_scb =
+		cpu_to_le64(((u64)seq->next_scb.dma_handle));
+}
+
+/**
+ * asd_start_timers -- (add and) start timers of SCBs
+ * @list: pointer to struct list_head of the scbs
+ * @to: timeout in jiffies
+ *
+ * If an SCB in the @list has no timer function, assign the default
+ * one,  then start the timer of the SCB.  This function is
+ * intended to be called from asd_post_ascb_list(), just prior to
+ * posting the SCBs to the sequencer.
+ */
+static inline void asd_start_scb_timers(struct list_head *list)
+{
+	struct asd_ascb *ascb;
+	list_for_each_entry(ascb, list, list) {
+		if (!ascb->uldd_timer) {
+			ascb->timer.data = (unsigned long) ascb;
+			ascb->timer.function = asd_ascb_timedout;
+			ascb->timer.expires = jiffies + AIC94XX_SCB_TIMEOUT;
+			add_timer(&ascb->timer);
+		}
+	}
+}
+
+/**
+ * asd_post_ascb_list -- post a list of 1 or more aSCBs to the host adapter
+ * @asd_ha: pointer to a host adapter structure
+ * @ascb: pointer to the first aSCB in the list
+ * @num: number of aSCBs in the list (to be posted)
+ *
+ * See queueing comment in asd_post_escb_list().
+ *
+ * Additional note on queuing: In order to minimize the ratio of memcpy()
+ * to the number of ascbs sent, we try to batch-send as many ascbs as possible
+ * in one go.
+ * Two cases are possible:
+ *    A) can_queue >= num,
+ *    B) can_queue < num.
+ * Case A: we can send the whole batch at once.  Increment "pending"
+ * in the beginning of this function, when it is checked, in order to
+ * eliminate races when this function is called by multiple processes.
+ * Case B: should never happen if the managing layer considers
+ * lldd_queue_size.
+ */
+int asd_post_ascb_list(struct asd_ha_struct *asd_ha, struct asd_ascb *ascb,
+		       int num)
+{
+	unsigned long flags;
+	LIST_HEAD(list);
+	int can_queue;
+
+	spin_lock_irqsave(&asd_ha->seq.pend_q_lock, flags);
+	can_queue = asd_ha->hw_prof.max_scbs - asd_ha->seq.pending;
+	if (can_queue >= num)
+		asd_ha->seq.pending += num;
+	else
+		can_queue = 0;
+
+	if (!can_queue) {
+		spin_unlock_irqrestore(&asd_ha->seq.pend_q_lock, flags);
+		asd_printk("%s: scb queue full\n", pci_name(asd_ha->pcidev));
+		return -SAS_QUEUE_FULL;
+	}
+
+	asd_swap_head_scb(asd_ha, ascb);
+
+	__list_add(&list, ascb->list.prev, &ascb->list);
+
+	asd_start_scb_timers(&list);
+
+	asd_ha->seq.scbpro += num;
+	list_splice_init(&list, asd_ha->seq.pend_q.prev);
+	asd_write_reg_dword(asd_ha, SCBPRO, (u32)asd_ha->seq.scbpro);
+	spin_unlock_irqrestore(&asd_ha->seq.pend_q_lock, flags);
+
+	return 0;
+}
+
+/**
+ * asd_post_escb_list -- post a list of 1 or more empty scb
+ * @asd_ha: pointer to a host adapter structure
+ * @ascb: pointer to the first empty SCB in the list
+ * @num: number of aSCBs in the list (to be posted)
+ *
+ * This is essentially the same as asd_post_ascb_list, but we do not
+ * increment pending, add those to the pending list or get indexes.
+ * See asd_init_escbs() and asd_init_post_escbs().
+ *
+ * Since sending a list of ascbs is a superset of sending a single
+ * ascb, this function exists to generalize this.  More specifically,
+ * when sending a list of those, we want to do only a _single_
+ * memcpy() at swap head, as opposed to for each ascb sent (in the
+ * case of sending them one by one).  That is, we want to minimize the
+ * ratio of memcpy() operations to the number of ascbs sent.  The same
+ * logic applies to asd_post_ascb_list().
+ */
+int asd_post_escb_list(struct asd_ha_struct *asd_ha, struct asd_ascb *ascb,
+		       int num)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&asd_ha->seq.pend_q_lock, flags);
+	asd_swap_head_scb(asd_ha, ascb);
+	asd_ha->seq.scbpro += num;
+	asd_write_reg_dword(asd_ha, SCBPRO, (u32)asd_ha->seq.scbpro);
+	spin_unlock_irqrestore(&asd_ha->seq.pend_q_lock, flags);
+
+	return 0;
+}
+
+/* ---------- LED ---------- */
+
+/**
+ * asd_turn_led -- turn on/off an LED
+ * @asd_ha: pointer to host adapter structure
+ * @phy_id: the PHY id whose LED we want to manupulate
+ * @op: 1 to turn on, 0 to turn off
+ */
+void asd_turn_led(struct asd_ha_struct *asd_ha, int phy_id, int op)
+{
+	if (phy_id < ASD_MAX_PHYS) {
+		u32 v = asd_read_reg_dword(asd_ha, LmCONTROL(phy_id));
+		if (op)
+			v |= LEDPOL;
+		else
+			v &= ~LEDPOL;
+		asd_write_reg_dword(asd_ha, LmCONTROL(phy_id), v);
+	}
+}
+
+/**
+ * asd_control_led -- enable/disable an LED on the board
+ * @asd_ha: pointer to host adapter structure
+ * @phy_id: integer, the phy id
+ * @op: integer, 1 to enable, 0 to disable the LED
+ *
+ * First we output enable the LED, then we set the source
+ * to be an external module.
+ */
+void asd_control_led(struct asd_ha_struct *asd_ha, int phy_id, int op)
+{
+	if (phy_id < ASD_MAX_PHYS) {
+		u32 v;
+
+		v = asd_read_reg_dword(asd_ha, GPIOOER);
+		if (op)
+			v |= (1 << phy_id);
+		else
+			v &= ~(1 << phy_id);
+		asd_write_reg_dword(asd_ha, GPIOOER, v);
+
+		v = asd_read_reg_dword(asd_ha, GPIOCNFGR);
+		if (op)
+			v |= (1 << phy_id);
+		else
+			v &= ~(1 << phy_id);
+		asd_write_reg_dword(asd_ha, GPIOCNFGR, v);
+	}
+}
+
+/* ---------- PHY enable ---------- */
+
+static int asd_enable_phy(struct asd_ha_struct *asd_ha, int phy_id)
+{
+	struct asd_phy *phy = &asd_ha->phys[phy_id];
+
+	asd_write_reg_byte(asd_ha, LmSEQ_OOB_REG(phy_id, INT_ENABLE_2), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_OOB_REG(phy_id, HOT_PLUG_DELAY),
+			   HOTPLUG_DELAY_TIMEOUT);
+
+	/* Get defaults from manuf. sector */
+	/* XXX we need defaults for those in case MS is broken. */
+	asd_write_reg_byte(asd_ha, LmSEQ_OOB_REG(phy_id, PHY_CONTROL_0),
+			   phy->phy_desc->phy_control_0);
+	asd_write_reg_byte(asd_ha, LmSEQ_OOB_REG(phy_id, PHY_CONTROL_1),
+			   phy->phy_desc->phy_control_1);
+	asd_write_reg_byte(asd_ha, LmSEQ_OOB_REG(phy_id, PHY_CONTROL_2),
+			   phy->phy_desc->phy_control_2);
+	asd_write_reg_byte(asd_ha, LmSEQ_OOB_REG(phy_id, PHY_CONTROL_3),
+			   phy->phy_desc->phy_control_3);
+
+	asd_write_reg_dword(asd_ha, LmSEQ_TEN_MS_COMINIT_TIMEOUT(phy_id),
+			    ASD_COMINIT_TIMEOUT);
+
+	asd_write_reg_addr(asd_ha, LmSEQ_TX_ID_ADDR_FRAME(phy_id),
+			   phy->id_frm_tok->dma_handle);
+
+	asd_control_led(asd_ha, phy_id, 1);
+
+	return 0;
+}
+
+int asd_enable_phys(struct asd_ha_struct *asd_ha, const u8 phy_mask)
+{
+	u8  phy_m;
+	u8  i;
+	int num = 0, k;
+	struct asd_ascb *ascb;
+	struct asd_ascb *ascb_list;
+
+	if (!phy_mask) {
+		asd_printk("%s called with phy_mask of 0!?\n", __FUNCTION__);
+		return 0;
+	}
+
+	for_each_phy(phy_mask, phy_m, i) {
+		num++;
+		asd_enable_phy(asd_ha, i);
+	}
+
+	k = num;
+	ascb_list = asd_ascb_alloc_list(asd_ha, &k, GFP_KERNEL);
+	if (!ascb_list) {
+		asd_printk("no memory for control phy ascb list\n");
+		return -ENOMEM;
+	}
+	num -= k;
+
+	ascb = ascb_list;
+	for_each_phy(phy_mask, phy_m, i) {
+		asd_build_control_phy(ascb, i, ENABLE_PHY);
+		ascb = list_entry(ascb->list.next, struct asd_ascb, list);
+	}
+	ASD_DPRINTK("posting %d control phy scbs\n", num);
+	k = asd_post_ascb_list(asd_ha, ascb_list, num);
+	if (k)
+		asd_ascb_free_list(ascb_list);
+
+	return k;
+}
diff --git a/drivers/scsi/aic94xx/aic94xx_hwi.h b/drivers/scsi/aic94xx/aic94xx_hwi.h
new file mode 100644
index 000000000000..c7d505388fed
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx_hwi.h
@@ -0,0 +1,397 @@
+/*
+ * Aic94xx SAS/SATA driver hardware interface header file.
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#ifndef _AIC94XX_HWI_H_
+#define _AIC94XX_HWI_H_
+
+#include <linux/interrupt.h>
+#include <linux/pci.h>
+#include <linux/dma-mapping.h>
+
+#include <scsi/libsas.h>
+
+#include "aic94xx.h"
+#include "aic94xx_sas.h"
+
+/* Define ASD_MAX_PHYS to the maximum phys ever. Currently 8. */
+#define ASD_MAX_PHYS       8
+#define ASD_PCBA_SN_SIZE   12
+
+/* Those are to be further named properly, the "RAZORx" part, and
+ * subsequently included in include/linux/pci_ids.h.
+ */
+#define PCI_DEVICE_ID_ADAPTEC2_RAZOR10 0x410
+#define PCI_DEVICE_ID_ADAPTEC2_RAZOR12 0x412
+#define PCI_DEVICE_ID_ADAPTEC2_RAZOR1E 0x41E
+#define PCI_DEVICE_ID_ADAPTEC2_RAZOR30 0x430
+#define PCI_DEVICE_ID_ADAPTEC2_RAZOR32 0x432
+#define PCI_DEVICE_ID_ADAPTEC2_RAZOR3E 0x43E
+#define PCI_DEVICE_ID_ADAPTEC2_RAZOR3F 0x43F
+
+struct asd_ha_addrspace {
+	void __iomem  *addr;
+	unsigned long  start;       /* pci resource start */
+	unsigned long  len;         /* pci resource len */
+	unsigned long  flags;       /* pci resource flags */
+
+	/* addresses internal to the host adapter */
+	u32 swa_base; /* mmspace 1 (MBAR1) uses this only */
+	u32 swb_base;
+	u32 swc_base;
+};
+
+struct bios_struct {
+	int    present;
+	u8     maj;
+	u8     min;
+	u32    bld;
+};
+
+struct unit_element_struct {
+	u16    num;
+	u16    size;
+	void   *area;
+};
+
+struct flash_struct {
+	u32    bar;
+	int    present;
+	int    wide;
+	u8     manuf;
+	u8     dev_id;
+	u8     sec_prot;
+
+	u32    dir_offs;
+};
+
+struct asd_phy_desc {
+	/* From CTRL-A settings, then set to what is appropriate */
+	u8     sas_addr[SAS_ADDR_SIZE];
+	u8     max_sas_lrate;
+	u8     min_sas_lrate;
+	u8     max_sata_lrate;
+	u8     min_sata_lrate;
+	u8     flags;
+#define ASD_CRC_DIS  1
+#define ASD_SATA_SPINUP_HOLD 2
+
+	u8     phy_control_0; /* mode 5 reg 0x160 */
+	u8     phy_control_1; /* mode 5 reg 0x161 */
+	u8     phy_control_2; /* mode 5 reg 0x162 */
+	u8     phy_control_3; /* mode 5 reg 0x163 */
+};
+
+struct asd_dma_tok {
+	void *vaddr;
+	dma_addr_t dma_handle;
+	size_t size;
+};
+
+struct hw_profile {
+	struct bios_struct bios;
+	struct unit_element_struct ue;
+	struct flash_struct flash;
+
+	u8     sas_addr[SAS_ADDR_SIZE];
+	char   pcba_sn[ASD_PCBA_SN_SIZE+1];
+
+	u8     enabled_phys;	  /* mask of enabled phys */
+	struct asd_phy_desc phy_desc[ASD_MAX_PHYS];
+	u32    max_scbs;	  /* absolute sequencer scb queue size */
+	struct asd_dma_tok *scb_ext;
+	u32    max_ddbs;
+	struct asd_dma_tok *ddb_ext;
+
+	spinlock_t ddb_lock;
+	void  *ddb_bitmap;
+
+	int    num_phys;	  /* ENABLEABLE */
+	int    max_phys;	  /* REPORTED + ENABLEABLE */
+
+	unsigned addr_range;	  /* max # of addrs; max # of possible ports */
+	unsigned port_name_base;
+	unsigned dev_name_base;
+	unsigned sata_name_base;
+};
+
+struct asd_ascb {
+	struct list_head list;
+	struct asd_ha_struct *ha;
+
+	struct scb *scb;	  /* equals dma_scb->vaddr */
+	struct asd_dma_tok dma_scb;
+	struct asd_dma_tok *sg_arr;
+
+	void (*tasklet_complete)(struct asd_ascb *, struct done_list_struct *);
+	u8     uldd_timer:1;
+
+	/* internally generated command */
+	struct timer_list timer;
+	struct completion completion;
+	u8        tag_valid:1;
+	__be16    tag;		  /* error recovery only */
+
+	/* If this is an Empty SCB, index of first edb in seq->edb_arr. */
+	int    edb_index;
+
+	/* Used by the timer timeout function. */
+	int    tc_index;
+
+	void   *uldd_task;
+};
+
+#define ASD_DL_SIZE_BITS   0x8
+#define ASD_DL_SIZE        (1<<(2+ASD_DL_SIZE_BITS))
+#define ASD_DEF_DL_TOGGLE  0x01
+
+struct asd_seq_data {
+	spinlock_t pend_q_lock;
+	u16    scbpro;
+	int    pending;
+	struct list_head pend_q;
+	int    can_queue;	  /* per adapter */
+	struct asd_dma_tok next_scb; /* next scb to be delivered to CSEQ */
+
+	spinlock_t tc_index_lock;
+	void **tc_index_array;
+	void *tc_index_bitmap;
+	int   tc_index_bitmap_bits;
+
+	struct tasklet_struct dl_tasklet;
+	struct done_list_struct *dl; /* array of done list entries, equals */
+	struct asd_dma_tok *actual_dl; /* actual_dl->vaddr */
+	int    dl_toggle;
+	int    dl_next;
+
+	int    num_edbs;
+	struct asd_dma_tok **edb_arr;
+	int    num_escbs;
+	struct asd_ascb **escb_arr; /* array of pointers to escbs */
+};
+
+/* This is the Host Adapter structure.  It describes the hardware
+ * SAS adapter.
+ */
+struct asd_ha_struct {
+	struct pci_dev   *pcidev;
+	const char       *name;
+
+	struct sas_ha_struct sas_ha;
+
+	u8                revision_id;
+
+	int               iospace;
+	spinlock_t        iolock;
+	struct asd_ha_addrspace io_handle[2];
+
+	struct hw_profile hw_prof;
+
+	struct asd_phy    phys[ASD_MAX_PHYS];
+	struct asd_sas_port   ports[ASD_MAX_PHYS];
+
+	struct dma_pool  *scb_pool;
+
+	struct asd_seq_data  seq; /* sequencer related */
+};
+
+/* ---------- Common macros ---------- */
+
+#define ASD_BUSADDR_LO(__dma_handle) ((u32)(__dma_handle))
+#define ASD_BUSADDR_HI(__dma_handle) (((sizeof(dma_addr_t))==8)     \
+                                    ? ((u32)((__dma_handle) >> 32)) \
+                                    : ((u32)0))
+
+#define dev_to_asd_ha(__dev)  pci_get_drvdata(to_pci_dev(__dev))
+#define SCB_SITE_VALID(__site_no) (((__site_no) & 0xF0FF) != 0x00FF   \
+				 && ((__site_no) & 0xF0FF) > 0x001F)
+/* For each bit set in __lseq_mask, set __lseq to equal the bit
+ * position of the set bit and execute the statement following.
+ * __mc is the temporary mask, used as a mask "counter".
+ */
+#define for_each_sequencer(__lseq_mask, __mc, __lseq)                        \
+	for ((__mc)=(__lseq_mask),(__lseq)=0;(__mc)!=0;(__lseq++),(__mc)>>=1)\
+		if (((__mc) & 1))
+#define for_each_phy(__lseq_mask, __mc, __lseq)                              \
+	for ((__mc)=(__lseq_mask),(__lseq)=0;(__mc)!=0;(__lseq++),(__mc)>>=1)\
+		if (((__mc) & 1))
+
+#define PHY_ENABLED(_HA, _I) ((_HA)->hw_prof.enabled_phys & (1<<(_I)))
+
+/* ---------- DMA allocs ---------- */
+
+static inline struct asd_dma_tok *asd_dmatok_alloc(unsigned int flags)
+{
+	return kmem_cache_alloc(asd_dma_token_cache, flags);
+}
+
+static inline void asd_dmatok_free(struct asd_dma_tok *token)
+{
+	kmem_cache_free(asd_dma_token_cache, token);
+}
+
+static inline struct asd_dma_tok *asd_alloc_coherent(struct asd_ha_struct *
+						     asd_ha, size_t size,
+						     unsigned int flags)
+{
+	struct asd_dma_tok *token = asd_dmatok_alloc(flags);
+	if (token) {
+		token->size = size;
+		token->vaddr = dma_alloc_coherent(&asd_ha->pcidev->dev,
+						  token->size,
+						  &token->dma_handle,
+						  flags);
+		if (!token->vaddr) {
+			asd_dmatok_free(token);
+			token = NULL;
+		}
+	}
+	return token;
+}
+
+static inline void asd_free_coherent(struct asd_ha_struct *asd_ha,
+				     struct asd_dma_tok *token)
+{
+	if (token) {
+		dma_free_coherent(&asd_ha->pcidev->dev, token->size,
+				  token->vaddr, token->dma_handle);
+		asd_dmatok_free(token);
+	}
+}
+
+static inline void asd_init_ascb(struct asd_ha_struct *asd_ha,
+				 struct asd_ascb *ascb)
+{
+	INIT_LIST_HEAD(&ascb->list);
+	ascb->scb = ascb->dma_scb.vaddr;
+	ascb->ha = asd_ha;
+	ascb->timer.function = NULL;
+	init_timer(&ascb->timer);
+	ascb->tc_index = -1;
+	init_completion(&ascb->completion);
+}
+
+/* Must be called with the tc_index_lock held!
+ */
+static inline void asd_tc_index_release(struct asd_seq_data *seq, int index)
+{
+	seq->tc_index_array[index] = NULL;
+	clear_bit(index, seq->tc_index_bitmap);
+}
+
+/* Must be called with the tc_index_lock held!
+ */
+static inline int asd_tc_index_get(struct asd_seq_data *seq, void *ptr)
+{
+	int index;
+
+	index = find_first_zero_bit(seq->tc_index_bitmap,
+				    seq->tc_index_bitmap_bits);
+	if (index == seq->tc_index_bitmap_bits)
+		return -1;
+
+	seq->tc_index_array[index] = ptr;
+	set_bit(index, seq->tc_index_bitmap);
+
+	return index;
+}
+
+/* Must be called with the tc_index_lock held!
+ */
+static inline void *asd_tc_index_find(struct asd_seq_data *seq, int index)
+{
+	return seq->tc_index_array[index];
+}
+
+/**
+ * asd_ascb_free -- free a single aSCB after is has completed
+ * @ascb: pointer to the aSCB of interest
+ *
+ * This frees an aSCB after it has been executed/completed by
+ * the sequencer.
+ */
+static inline void asd_ascb_free(struct asd_ascb *ascb)
+{
+	if (ascb) {
+		struct asd_ha_struct *asd_ha = ascb->ha;
+		unsigned long flags;
+
+		BUG_ON(!list_empty(&ascb->list));
+		spin_lock_irqsave(&ascb->ha->seq.tc_index_lock, flags);
+		asd_tc_index_release(&ascb->ha->seq, ascb->tc_index);
+		spin_unlock_irqrestore(&ascb->ha->seq.tc_index_lock, flags);
+		dma_pool_free(asd_ha->scb_pool, ascb->dma_scb.vaddr,
+			      ascb->dma_scb.dma_handle);
+		kmem_cache_free(asd_ascb_cache, ascb);
+	}
+}
+
+/**
+ * asd_ascb_list_free -- free a list of ascbs
+ * @ascb_list: a list of ascbs
+ *
+ * This function will free a list of ascbs allocated by asd_ascb_alloc_list.
+ * It is used when say the scb queueing function returned QUEUE_FULL,
+ * and we do not need the ascbs any more.
+ */
+static inline void asd_ascb_free_list(struct asd_ascb *ascb_list)
+{
+	LIST_HEAD(list);
+	struct list_head *n, *pos;
+
+	__list_add(&list, ascb_list->list.prev, &ascb_list->list);
+	list_for_each_safe(pos, n, &list) {
+		list_del_init(pos);
+		asd_ascb_free(list_entry(pos, struct asd_ascb, list));
+	}
+}
+
+/* ---------- Function declarations ---------- */
+
+int  asd_init_hw(struct asd_ha_struct *asd_ha);
+irqreturn_t asd_hw_isr(int irq, void *dev_id, struct pt_regs *regs);
+
+
+struct asd_ascb *asd_ascb_alloc_list(struct asd_ha_struct
+				     *asd_ha, int *num,
+				     unsigned int gfp_mask);
+
+int  asd_post_ascb_list(struct asd_ha_struct *asd_ha, struct asd_ascb *ascb,
+			int num);
+int  asd_post_escb_list(struct asd_ha_struct *asd_ha, struct asd_ascb *ascb,
+			int num);
+
+int  asd_init_post_escbs(struct asd_ha_struct *asd_ha);
+void asd_build_control_phy(struct asd_ascb *ascb, int phy_id, u8 subfunc);
+void asd_control_led(struct asd_ha_struct *asd_ha, int phy_id, int op);
+void asd_turn_led(struct asd_ha_struct *asd_ha, int phy_id, int op);
+int  asd_enable_phys(struct asd_ha_struct *asd_ha, const u8 phy_mask);
+void asd_build_initiate_link_adm_task(struct asd_ascb *ascb, int phy_id,
+				      u8 subfunc);
+
+void asd_ascb_timedout(unsigned long data);
+int  asd_chip_hardrst(struct asd_ha_struct *asd_ha);
+
+#endif
diff --git a/drivers/scsi/aic94xx/aic94xx_init.c b/drivers/scsi/aic94xx/aic94xx_init.c
new file mode 100644
index 000000000000..3ec2e46f80c6
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx_init.c
@@ -0,0 +1,860 @@
+/*
+ * Aic94xx SAS/SATA driver initialization.
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#include <linux/config.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/pci.h>
+#include <linux/delay.h>
+
+#include <scsi/scsi_host.h>
+
+#include "aic94xx.h"
+#include "aic94xx_reg.h"
+#include "aic94xx_hwi.h"
+#include "aic94xx_seq.h"
+
+/* The format is "version.release.patchlevel" */
+#define ASD_DRIVER_VERSION "1.0.2"
+
+static int use_msi = 0;
+module_param_named(use_msi, use_msi, int, S_IRUGO);
+MODULE_PARM_DESC(use_msi, "\n"
+	"\tEnable(1) or disable(0) using PCI MSI.\n"
+	"\tDefault: 0");
+
+static int lldd_max_execute_num = 0;
+module_param_named(collector, lldd_max_execute_num, int, S_IRUGO);
+MODULE_PARM_DESC(collector, "\n"
+	"\tIf greater than one, tells the SAS Layer to run in Task Collector\n"
+	"\tMode.  If 1 or 0, tells the SAS Layer to run in Direct Mode.\n"
+	"\tThe aic94xx SAS LLDD supports both modes.\n"
+	"\tDefault: 0 (Direct Mode).\n");
+
+char sas_addr_str[2*SAS_ADDR_SIZE + 1] = "";
+
+static struct scsi_transport_template *aic94xx_transport_template;
+
+static struct scsi_host_template aic94xx_sht = {
+	.module			= THIS_MODULE,
+	/* .name is initialized */
+	.name			= "aic94xx",
+	.queuecommand		= sas_queuecommand,
+	.target_alloc		= sas_target_alloc,
+	.slave_configure	= sas_slave_configure,
+	.slave_destroy		= sas_slave_destroy,
+	.change_queue_depth	= sas_change_queue_depth,
+	.change_queue_type	= sas_change_queue_type,
+	.bios_param		= sas_bios_param,
+	.can_queue		= 1,
+	.cmd_per_lun		= 1,
+	.this_id		= -1,
+	.sg_tablesize		= SG_ALL,
+	.max_sectors		= SCSI_DEFAULT_MAX_SECTORS,
+	.use_clustering		= ENABLE_CLUSTERING,
+};
+
+static int __devinit asd_map_memio(struct asd_ha_struct *asd_ha)
+{
+	int err, i;
+	struct asd_ha_addrspace *io_handle;
+
+	asd_ha->iospace = 0;
+	for (i = 0; i < 3; i += 2) {
+		io_handle = &asd_ha->io_handle[i==0?0:1];
+		io_handle->start = pci_resource_start(asd_ha->pcidev, i);
+		io_handle->len   = pci_resource_len(asd_ha->pcidev, i);
+		io_handle->flags = pci_resource_flags(asd_ha->pcidev, i);
+		err = -ENODEV;
+		if (!io_handle->start || !io_handle->len) {
+			asd_printk("MBAR%d start or length for %s is 0.\n",
+				   i==0?0:1, pci_name(asd_ha->pcidev));
+			goto Err;
+		}
+		err = pci_request_region(asd_ha->pcidev, i, ASD_DRIVER_NAME);
+		if (err) {
+			asd_printk("couldn't reserve memory region for %s\n",
+				   pci_name(asd_ha->pcidev));
+			goto Err;
+		}
+		if (io_handle->flags & IORESOURCE_CACHEABLE)
+			io_handle->addr = ioremap(io_handle->start,
+						  io_handle->len);
+		else
+			io_handle->addr = ioremap_nocache(io_handle->start,
+							  io_handle->len);
+		if (!io_handle->addr) {
+			asd_printk("couldn't map MBAR%d of %s\n", i==0?0:1,
+				   pci_name(asd_ha->pcidev));
+			goto Err_unreq;
+		}
+	}
+
+	return 0;
+Err_unreq:
+	pci_release_region(asd_ha->pcidev, i);
+Err:
+	if (i > 0) {
+		io_handle = &asd_ha->io_handle[0];
+		iounmap(io_handle->addr);
+		pci_release_region(asd_ha->pcidev, 0);
+	}
+	return err;
+}
+
+static void __devexit asd_unmap_memio(struct asd_ha_struct *asd_ha)
+{
+	struct asd_ha_addrspace *io_handle;
+
+	io_handle = &asd_ha->io_handle[1];
+	iounmap(io_handle->addr);
+	pci_release_region(asd_ha->pcidev, 2);
+
+	io_handle = &asd_ha->io_handle[0];
+	iounmap(io_handle->addr);
+	pci_release_region(asd_ha->pcidev, 0);
+}
+
+static int __devinit asd_map_ioport(struct asd_ha_struct *asd_ha)
+{
+	int i = PCI_IOBAR_OFFSET, err;
+	struct asd_ha_addrspace *io_handle = &asd_ha->io_handle[0];
+
+	asd_ha->iospace = 1;
+	io_handle->start = pci_resource_start(asd_ha->pcidev, i);
+	io_handle->len   = pci_resource_len(asd_ha->pcidev, i);
+	io_handle->flags = pci_resource_flags(asd_ha->pcidev, i);
+	io_handle->addr  = (void __iomem *) io_handle->start;
+	if (!io_handle->start || !io_handle->len) {
+		asd_printk("couldn't get IO ports for %s\n",
+			   pci_name(asd_ha->pcidev));
+		return -ENODEV;
+	}
+	err = pci_request_region(asd_ha->pcidev, i, ASD_DRIVER_NAME);
+	if (err) {
+		asd_printk("couldn't reserve io space for %s\n",
+			   pci_name(asd_ha->pcidev));
+	}
+
+	return err;
+}
+
+static void __devexit asd_unmap_ioport(struct asd_ha_struct *asd_ha)
+{
+	pci_release_region(asd_ha->pcidev, PCI_IOBAR_OFFSET);
+}
+
+static int __devinit asd_map_ha(struct asd_ha_struct *asd_ha)
+{
+	int err;
+	u16 cmd_reg;
+
+	err = pci_read_config_word(asd_ha->pcidev, PCI_COMMAND, &cmd_reg);
+	if (err) {
+		asd_printk("couldn't read command register of %s\n",
+			   pci_name(asd_ha->pcidev));
+		goto Err;
+	}
+
+	err = -ENODEV;
+	if (cmd_reg & PCI_COMMAND_MEMORY) {
+		if ((err = asd_map_memio(asd_ha)))
+			goto Err;
+	} else if (cmd_reg & PCI_COMMAND_IO) {
+		if ((err = asd_map_ioport(asd_ha)))
+			goto Err;
+		asd_printk("%s ioport mapped -- upgrade your hardware\n",
+			   pci_name(asd_ha->pcidev));
+	} else {
+		asd_printk("no proper device access to %s\n",
+			   pci_name(asd_ha->pcidev));
+		goto Err;
+	}
+
+	return 0;
+Err:
+	return err;
+}
+
+static void __devexit asd_unmap_ha(struct asd_ha_struct *asd_ha)
+{
+	if (asd_ha->iospace)
+		asd_unmap_ioport(asd_ha);
+	else
+		asd_unmap_memio(asd_ha);
+}
+
+static const char *asd_dev_rev[30] = {
+	[0] = "A0",
+	[1] = "A1",
+	[8] = "B0",
+};
+
+static int __devinit asd_common_setup(struct asd_ha_struct *asd_ha)
+{
+	int err, i;
+
+	err = pci_read_config_byte(asd_ha->pcidev, PCI_REVISION_ID,
+				   &asd_ha->revision_id);
+	if (err) {
+		asd_printk("couldn't read REVISION ID register of %s\n",
+			   pci_name(asd_ha->pcidev));
+		goto Err;
+	}
+	err = -ENODEV;
+	if (asd_ha->revision_id < AIC9410_DEV_REV_B0) {
+		asd_printk("%s is revision %s (%X), which is not supported\n",
+			   pci_name(asd_ha->pcidev),
+			   asd_dev_rev[asd_ha->revision_id],
+			   asd_ha->revision_id);
+		goto Err;
+	}
+	/* Provide some sane default values. */
+	asd_ha->hw_prof.max_scbs = 512;
+	asd_ha->hw_prof.max_ddbs = 128;
+	asd_ha->hw_prof.num_phys = ASD_MAX_PHYS;
+	/* All phys are enabled, by default. */
+	asd_ha->hw_prof.enabled_phys = 0xFF;
+	for (i = 0; i < ASD_MAX_PHYS; i++) {
+		asd_ha->hw_prof.phy_desc[i].max_sas_lrate = PHY_LINKRATE_3;
+		asd_ha->hw_prof.phy_desc[i].min_sas_lrate = PHY_LINKRATE_1_5;
+		asd_ha->hw_prof.phy_desc[i].max_sata_lrate= PHY_LINKRATE_1_5;
+		asd_ha->hw_prof.phy_desc[i].min_sata_lrate= PHY_LINKRATE_1_5;
+	}
+
+	return 0;
+Err:
+	return err;
+}
+
+static int __devinit asd_aic9410_setup(struct asd_ha_struct *asd_ha)
+{
+	int err = asd_common_setup(asd_ha);
+
+	if (err)
+		return err;
+
+	asd_ha->hw_prof.addr_range = 8;
+	asd_ha->hw_prof.port_name_base = 0;
+	asd_ha->hw_prof.dev_name_base = 8;
+	asd_ha->hw_prof.sata_name_base = 16;
+
+	return 0;
+}
+
+static int __devinit asd_aic9405_setup(struct asd_ha_struct *asd_ha)
+{
+	int err = asd_common_setup(asd_ha);
+
+	if (err)
+		return err;
+
+	asd_ha->hw_prof.addr_range = 4;
+	asd_ha->hw_prof.port_name_base = 0;
+	asd_ha->hw_prof.dev_name_base = 4;
+	asd_ha->hw_prof.sata_name_base = 8;
+
+	return 0;
+}
+
+static ssize_t asd_show_dev_rev(struct device *dev,
+				struct device_attribute *attr, char *buf)
+{
+	struct asd_ha_struct *asd_ha = dev_to_asd_ha(dev);
+	return snprintf(buf, PAGE_SIZE, "%s\n",
+			asd_dev_rev[asd_ha->revision_id]);
+}
+static DEVICE_ATTR(revision, S_IRUGO, asd_show_dev_rev, NULL);
+
+static ssize_t asd_show_dev_bios_build(struct device *dev,
+				       struct device_attribute *attr,char *buf)
+{
+	struct asd_ha_struct *asd_ha = dev_to_asd_ha(dev);
+	return snprintf(buf, PAGE_SIZE, "%d\n", asd_ha->hw_prof.bios.bld);
+}
+static DEVICE_ATTR(bios_build, S_IRUGO, asd_show_dev_bios_build, NULL);
+
+static ssize_t asd_show_dev_pcba_sn(struct device *dev,
+				    struct device_attribute *attr, char *buf)
+{
+	struct asd_ha_struct *asd_ha = dev_to_asd_ha(dev);
+	return snprintf(buf, PAGE_SIZE, "%s\n", asd_ha->hw_prof.pcba_sn);
+}
+static DEVICE_ATTR(pcba_sn, S_IRUGO, asd_show_dev_pcba_sn, NULL);
+
+static void asd_create_dev_attrs(struct asd_ha_struct *asd_ha)
+{
+	device_create_file(&asd_ha->pcidev->dev, &dev_attr_revision);
+	device_create_file(&asd_ha->pcidev->dev, &dev_attr_bios_build);
+	device_create_file(&asd_ha->pcidev->dev, &dev_attr_pcba_sn);
+}
+
+static void asd_remove_dev_attrs(struct asd_ha_struct *asd_ha)
+{
+	device_remove_file(&asd_ha->pcidev->dev, &dev_attr_revision);
+	device_remove_file(&asd_ha->pcidev->dev, &dev_attr_bios_build);
+	device_remove_file(&asd_ha->pcidev->dev, &dev_attr_pcba_sn);
+}
+
+/* The first entry, 0, is used for dynamic ids, the rest for devices
+ * we know about.
+ */
+static struct asd_pcidev_struct {
+	const char * name;
+	int (*setup)(struct asd_ha_struct *asd_ha);
+} asd_pcidev_data[] = {
+	/* Id 0 is used for dynamic ids. */
+	{ .name  = "Adaptec AIC-94xx SAS/SATA Host Adapter",
+	  .setup = asd_aic9410_setup
+	},
+	{ .name  = "Adaptec AIC-9410W SAS/SATA Host Adapter",
+	  .setup = asd_aic9410_setup
+	},
+	{ .name  = "Adaptec AIC-9405W SAS/SATA Host Adapter",
+	  .setup = asd_aic9405_setup
+	},
+};
+
+static inline int asd_create_ha_caches(struct asd_ha_struct *asd_ha)
+{
+	asd_ha->scb_pool = dma_pool_create(ASD_DRIVER_NAME "_scb_pool",
+					   &asd_ha->pcidev->dev,
+					   sizeof(struct scb),
+					   8, 0);
+	if (!asd_ha->scb_pool) {
+		asd_printk("couldn't create scb pool\n");
+		return -ENOMEM;
+	}
+
+	return 0;
+}
+
+/**
+ * asd_free_edbs -- free empty data buffers
+ * asd_ha: pointer to host adapter structure
+ */
+static inline void asd_free_edbs(struct asd_ha_struct *asd_ha)
+{
+	struct asd_seq_data *seq = &asd_ha->seq;
+	int i;
+
+	for (i = 0; i < seq->num_edbs; i++)
+		asd_free_coherent(asd_ha, seq->edb_arr[i]);
+	kfree(seq->edb_arr);
+	seq->edb_arr = NULL;
+}
+
+static inline void asd_free_escbs(struct asd_ha_struct *asd_ha)
+{
+	struct asd_seq_data *seq = &asd_ha->seq;
+	int i;
+
+	for (i = 0; i < seq->num_escbs; i++) {
+		if (!list_empty(&seq->escb_arr[i]->list))
+			list_del_init(&seq->escb_arr[i]->list);
+
+		asd_ascb_free(seq->escb_arr[i]);
+	}
+	kfree(seq->escb_arr);
+	seq->escb_arr = NULL;
+}
+
+static inline void asd_destroy_ha_caches(struct asd_ha_struct *asd_ha)
+{
+	int i;
+
+	if (asd_ha->hw_prof.ddb_ext)
+		asd_free_coherent(asd_ha, asd_ha->hw_prof.ddb_ext);
+	if (asd_ha->hw_prof.scb_ext)
+		asd_free_coherent(asd_ha, asd_ha->hw_prof.scb_ext);
+
+	if (asd_ha->hw_prof.ddb_bitmap)
+		kfree(asd_ha->hw_prof.ddb_bitmap);
+	asd_ha->hw_prof.ddb_bitmap = NULL;
+
+	for (i = 0; i < ASD_MAX_PHYS; i++) {
+		struct asd_phy *phy = &asd_ha->phys[i];
+
+		asd_free_coherent(asd_ha, phy->id_frm_tok);
+	}
+	if (asd_ha->seq.escb_arr)
+		asd_free_escbs(asd_ha);
+	if (asd_ha->seq.edb_arr)
+		asd_free_edbs(asd_ha);
+	if (asd_ha->hw_prof.ue.area) {
+		kfree(asd_ha->hw_prof.ue.area);
+		asd_ha->hw_prof.ue.area = NULL;
+	}
+	if (asd_ha->seq.tc_index_array) {
+		kfree(asd_ha->seq.tc_index_array);
+		kfree(asd_ha->seq.tc_index_bitmap);
+		asd_ha->seq.tc_index_array = NULL;
+		asd_ha->seq.tc_index_bitmap = NULL;
+	}
+	if (asd_ha->seq.actual_dl) {
+			asd_free_coherent(asd_ha, asd_ha->seq.actual_dl);
+			asd_ha->seq.actual_dl = NULL;
+			asd_ha->seq.dl = NULL;
+	}
+	if (asd_ha->seq.next_scb.vaddr) {
+		dma_pool_free(asd_ha->scb_pool, asd_ha->seq.next_scb.vaddr,
+			      asd_ha->seq.next_scb.dma_handle);
+		asd_ha->seq.next_scb.vaddr = NULL;
+	}
+	dma_pool_destroy(asd_ha->scb_pool);
+	asd_ha->scb_pool = NULL;
+}
+
+kmem_cache_t *asd_dma_token_cache;
+kmem_cache_t *asd_ascb_cache;
+
+static int asd_create_global_caches(void)
+{
+	if (!asd_dma_token_cache) {
+		asd_dma_token_cache
+			= kmem_cache_create(ASD_DRIVER_NAME "_dma_token",
+					    sizeof(struct asd_dma_tok),
+					    0,
+					    SLAB_HWCACHE_ALIGN,
+					    NULL, NULL);
+		if (!asd_dma_token_cache) {
+			asd_printk("couldn't create dma token cache\n");
+			return -ENOMEM;
+		}
+	}
+
+	if (!asd_ascb_cache) {
+		asd_ascb_cache = kmem_cache_create(ASD_DRIVER_NAME "_ascb",
+						   sizeof(struct asd_ascb),
+						   0,
+						   SLAB_HWCACHE_ALIGN,
+						   NULL, NULL);
+		if (!asd_ascb_cache) {
+			asd_printk("couldn't create ascb cache\n");
+			goto Err;
+		}
+	}
+
+	return 0;
+Err:
+	kmem_cache_destroy(asd_dma_token_cache);
+	asd_dma_token_cache = NULL;
+	return -ENOMEM;
+}
+
+static void asd_destroy_global_caches(void)
+{
+	if (asd_dma_token_cache)
+		kmem_cache_destroy(asd_dma_token_cache);
+	asd_dma_token_cache = NULL;
+
+	if (asd_ascb_cache)
+		kmem_cache_destroy(asd_ascb_cache);
+	asd_ascb_cache = NULL;
+}
+
+static int asd_register_sas_ha(struct asd_ha_struct *asd_ha)
+{
+	int i;
+	struct asd_sas_phy   **sas_phys =
+		kmalloc(ASD_MAX_PHYS * sizeof(struct asd_sas_phy), GFP_KERNEL);
+	struct asd_sas_port  **sas_ports =
+		kmalloc(ASD_MAX_PHYS * sizeof(struct asd_sas_port), GFP_KERNEL);
+
+	if (!sas_phys || !sas_ports) {
+		kfree(sas_phys);
+		kfree(sas_ports);
+		return -ENOMEM;
+	}
+
+	asd_ha->sas_ha.sas_ha_name = (char *) asd_ha->name;
+	asd_ha->sas_ha.lldd_module = THIS_MODULE;
+	asd_ha->sas_ha.sas_addr = &asd_ha->hw_prof.sas_addr[0];
+
+	for (i = 0; i < ASD_MAX_PHYS; i++) {
+		sas_phys[i] = &asd_ha->phys[i].sas_phy;
+		sas_ports[i] = &asd_ha->ports[i];
+	}
+
+	asd_ha->sas_ha.sas_phy = sas_phys;
+	asd_ha->sas_ha.sas_port= sas_ports;
+	asd_ha->sas_ha.num_phys= ASD_MAX_PHYS;
+
+	asd_ha->sas_ha.lldd_queue_size = asd_ha->seq.can_queue;
+
+	return sas_register_ha(&asd_ha->sas_ha);
+}
+
+static int asd_unregister_sas_ha(struct asd_ha_struct *asd_ha)
+{
+	int err;
+
+	err = sas_unregister_ha(&asd_ha->sas_ha);
+
+	sas_remove_host(asd_ha->sas_ha.core.shost);
+	scsi_remove_host(asd_ha->sas_ha.core.shost);
+	scsi_host_put(asd_ha->sas_ha.core.shost);
+
+	kfree(asd_ha->sas_ha.sas_phy);
+	kfree(asd_ha->sas_ha.sas_port);
+
+	return err;
+}
+
+static int __devinit asd_pci_probe(struct pci_dev *dev,
+				   const struct pci_device_id *id)
+{
+	struct asd_pcidev_struct *asd_dev;
+	unsigned asd_id = (unsigned) id->driver_data;
+	struct asd_ha_struct *asd_ha;
+	struct Scsi_Host *shost;
+	int err;
+
+	if (asd_id >= ARRAY_SIZE(asd_pcidev_data)) {
+		asd_printk("wrong driver_data in PCI table\n");
+		return -ENODEV;
+	}
+
+	if ((err = pci_enable_device(dev))) {
+		asd_printk("couldn't enable device %s\n", pci_name(dev));
+		return err;
+	}
+
+	pci_set_master(dev);
+
+	err = -ENOMEM;
+
+	shost = scsi_host_alloc(&aic94xx_sht, sizeof(void *));
+	if (!shost)
+		goto Err;
+
+	asd_dev = &asd_pcidev_data[asd_id];
+
+	asd_ha = kzalloc(sizeof(*asd_ha), GFP_KERNEL);
+	if (!asd_ha) {
+		asd_printk("out of memory\n");
+		goto Err;
+	}
+	asd_ha->pcidev = dev;
+	asd_ha->sas_ha.pcidev = asd_ha->pcidev;
+	asd_ha->sas_ha.lldd_ha = asd_ha;
+
+	asd_ha->name = asd_dev->name;
+	asd_printk("found %s, device %s\n", asd_ha->name, pci_name(dev));
+
+	SHOST_TO_SAS_HA(shost) = &asd_ha->sas_ha;
+	asd_ha->sas_ha.core.shost = shost;
+	shost->transportt = aic94xx_transport_template;
+	shost->max_id = ~0;
+	shost->max_lun = ~0;
+	shost->max_cmd_len = 16;
+
+	err = scsi_add_host(shost, &dev->dev);
+	if (err) {
+		scsi_host_put(shost);
+		goto Err_free;
+	}
+
+
+
+	err = asd_dev->setup(asd_ha);
+	if (err)
+		goto Err_free;
+
+	err = -ENODEV;
+	if (!pci_set_dma_mask(dev, DMA_64BIT_MASK)
+	    && !pci_set_consistent_dma_mask(dev, DMA_64BIT_MASK))
+		;
+	else if (!pci_set_dma_mask(dev, DMA_32BIT_MASK)
+		 && !pci_set_consistent_dma_mask(dev, DMA_32BIT_MASK))
+		;
+	else {
+		asd_printk("no suitable DMA mask for %s\n", pci_name(dev));
+		goto Err_free;
+	}
+
+	pci_set_drvdata(dev, asd_ha);
+
+	err = asd_map_ha(asd_ha);
+	if (err)
+		goto Err_free;
+
+	err = asd_create_ha_caches(asd_ha);
+        if (err)
+		goto Err_unmap;
+
+	err = asd_init_hw(asd_ha);
+	if (err)
+		goto Err_free_cache;
+
+	asd_printk("device %s: SAS addr %llx, PCBA SN %s, %d phys, %d enabled "
+		   "phys, flash %s, BIOS %s%d\n",
+		   pci_name(dev), SAS_ADDR(asd_ha->hw_prof.sas_addr),
+		   asd_ha->hw_prof.pcba_sn, asd_ha->hw_prof.max_phys,
+		   asd_ha->hw_prof.num_phys,
+		   asd_ha->hw_prof.flash.present ? "present" : "not present",
+		   asd_ha->hw_prof.bios.present ? "build " : "not present",
+		   asd_ha->hw_prof.bios.bld);
+
+	if (use_msi)
+		pci_enable_msi(asd_ha->pcidev);
+
+	err = request_irq(asd_ha->pcidev->irq, asd_hw_isr, SA_SHIRQ,
+			  ASD_DRIVER_NAME, asd_ha);
+	if (err) {
+		asd_printk("couldn't get irq %d for %s\n",
+			   asd_ha->pcidev->irq, pci_name(asd_ha->pcidev));
+		goto Err_irq;
+	}
+	asd_enable_ints(asd_ha);
+
+	err = asd_init_post_escbs(asd_ha);
+	if (err) {
+		asd_printk("couldn't post escbs for %s\n",
+			   pci_name(asd_ha->pcidev));
+		goto Err_escbs;
+	}
+	ASD_DPRINTK("escbs posted\n");
+
+	asd_create_dev_attrs(asd_ha);
+
+	err = asd_register_sas_ha(asd_ha);
+	if (err)
+		goto Err_reg_sas;
+
+	err = asd_enable_phys(asd_ha, asd_ha->hw_prof.enabled_phys);
+	if (err) {
+		asd_printk("coudln't enable phys, err:%d\n", err);
+		goto Err_en_phys;
+	}
+	ASD_DPRINTK("enabled phys\n");
+	/* give the phy enabling interrupt event time to come in (1s
+	 * is empirically about all it takes) */
+	ssleep(1);
+	/* Wait for discovery to finish */
+	scsi_flush_work(asd_ha->sas_ha.core.shost);
+
+	return 0;
+Err_en_phys:
+	asd_unregister_sas_ha(asd_ha);
+Err_reg_sas:
+	asd_remove_dev_attrs(asd_ha);
+Err_escbs:
+	asd_disable_ints(asd_ha);
+	free_irq(dev->irq, asd_ha);
+Err_irq:
+	if (use_msi)
+		pci_disable_msi(dev);
+	asd_chip_hardrst(asd_ha);
+Err_free_cache:
+	asd_destroy_ha_caches(asd_ha);
+Err_unmap:
+	asd_unmap_ha(asd_ha);
+Err_free:
+	kfree(asd_ha);
+	scsi_remove_host(shost);
+Err:
+	pci_disable_device(dev);
+	return err;
+}
+
+static void asd_free_queues(struct asd_ha_struct *asd_ha)
+{
+	unsigned long flags;
+	LIST_HEAD(pending);
+	struct list_head *n, *pos;
+
+	spin_lock_irqsave(&asd_ha->seq.pend_q_lock, flags);
+	asd_ha->seq.pending = 0;
+	list_splice_init(&asd_ha->seq.pend_q, &pending);
+	spin_unlock_irqrestore(&asd_ha->seq.pend_q_lock, flags);
+
+	if (!list_empty(&pending))
+		ASD_DPRINTK("Uh-oh! Pending is not empty!\n");
+
+	list_for_each_safe(pos, n, &pending) {
+		struct asd_ascb *ascb = list_entry(pos, struct asd_ascb, list);
+		list_del_init(pos);
+		ASD_DPRINTK("freeing from pending\n");
+		asd_ascb_free(ascb);
+	}
+}
+
+static void asd_turn_off_leds(struct asd_ha_struct *asd_ha)
+{
+	u8 phy_mask = asd_ha->hw_prof.enabled_phys;
+	u8 i;
+
+	for_each_phy(phy_mask, phy_mask, i) {
+		asd_turn_led(asd_ha, i, 0);
+		asd_control_led(asd_ha, i, 0);
+	}
+}
+
+static void __devexit asd_pci_remove(struct pci_dev *dev)
+{
+	struct asd_ha_struct *asd_ha = pci_get_drvdata(dev);
+
+	if (!asd_ha)
+		return;
+
+	asd_unregister_sas_ha(asd_ha);
+
+	asd_disable_ints(asd_ha);
+
+	asd_remove_dev_attrs(asd_ha);
+
+	/* XXX more here as needed */
+
+	free_irq(dev->irq, asd_ha);
+	if (use_msi)
+		pci_disable_msi(asd_ha->pcidev);
+	asd_turn_off_leds(asd_ha);
+	asd_chip_hardrst(asd_ha);
+	asd_free_queues(asd_ha);
+	asd_destroy_ha_caches(asd_ha);
+	asd_unmap_ha(asd_ha);
+	kfree(asd_ha);
+	pci_disable_device(dev);
+	return;
+}
+
+static ssize_t asd_version_show(struct device_driver *driver, char *buf)
+{
+	return snprintf(buf, PAGE_SIZE, "%s\n", ASD_DRIVER_VERSION);
+}
+static DRIVER_ATTR(version, S_IRUGO, asd_version_show, NULL);
+
+static void asd_create_driver_attrs(struct device_driver *driver)
+{
+	driver_create_file(driver, &driver_attr_version);
+}
+
+static void asd_remove_driver_attrs(struct device_driver *driver)
+{
+	driver_remove_file(driver, &driver_attr_version);
+}
+
+static struct sas_domain_function_template aic94xx_transport_functions = {
+	.lldd_port_formed	= asd_update_port_links,
+
+	.lldd_dev_found		= asd_dev_found,
+	.lldd_dev_gone		= asd_dev_gone,
+
+	.lldd_execute_task	= asd_execute_task,
+
+	.lldd_abort_task	= asd_abort_task,
+	.lldd_abort_task_set	= asd_abort_task_set,
+	.lldd_clear_aca		= asd_clear_aca,
+	.lldd_clear_task_set	= asd_clear_task_set,
+	.lldd_I_T_nexus_reset	= NULL,
+	.lldd_lu_reset		= asd_lu_reset,
+	.lldd_query_task	= asd_query_task,
+
+	.lldd_clear_nexus_port	= asd_clear_nexus_port,
+	.lldd_clear_nexus_ha	= asd_clear_nexus_ha,
+
+	.lldd_control_phy	= asd_control_phy,
+};
+
+static const struct pci_device_id aic94xx_pci_table[] __devinitdata = {
+	{PCI_DEVICE(PCI_VENDOR_ID_ADAPTEC2, PCI_DEVICE_ID_ADAPTEC2_RAZOR10),
+	 0, 0, 1},
+	{PCI_DEVICE(PCI_VENDOR_ID_ADAPTEC2, PCI_DEVICE_ID_ADAPTEC2_RAZOR12),
+	 0, 0, 1},
+	{PCI_DEVICE(PCI_VENDOR_ID_ADAPTEC2, PCI_DEVICE_ID_ADAPTEC2_RAZOR1E),
+	 0, 0, 1},
+	{PCI_DEVICE(PCI_VENDOR_ID_ADAPTEC2, PCI_DEVICE_ID_ADAPTEC2_RAZOR30),
+	 0, 0, 2},
+	{PCI_DEVICE(PCI_VENDOR_ID_ADAPTEC2, PCI_DEVICE_ID_ADAPTEC2_RAZOR32),
+	 0, 0, 2},
+	{PCI_DEVICE(PCI_VENDOR_ID_ADAPTEC2, PCI_DEVICE_ID_ADAPTEC2_RAZOR3E),
+	 0, 0, 2},
+	{PCI_DEVICE(PCI_VENDOR_ID_ADAPTEC2, PCI_DEVICE_ID_ADAPTEC2_RAZOR3F),
+	 0, 0, 2},
+	{}
+};
+
+MODULE_DEVICE_TABLE(pci, aic94xx_pci_table);
+
+static struct pci_driver aic94xx_pci_driver = {
+	.name		= ASD_DRIVER_NAME,
+	.id_table	= aic94xx_pci_table,
+	.probe		= asd_pci_probe,
+	.remove		= __devexit_p(asd_pci_remove),
+};
+
+static int __init aic94xx_init(void)
+{
+	int err;
+
+
+	asd_printk("%s version %s loaded\n", ASD_DRIVER_DESCRIPTION,
+		   ASD_DRIVER_VERSION);
+
+	err = asd_create_global_caches();
+	if (err)
+		return err;
+
+	aic94xx_transport_template =
+		sas_domain_attach_transport(&aic94xx_transport_functions);
+	if (err)
+		goto out_destroy_caches;
+
+	err = pci_register_driver(&aic94xx_pci_driver);
+	if (err)
+		goto out_release_transport;
+
+	asd_create_driver_attrs(&aic94xx_pci_driver.driver);
+
+	return err;
+
+ out_release_transport:
+	sas_release_transport(aic94xx_transport_template);
+ out_destroy_caches:
+	asd_destroy_global_caches();
+
+	return err;
+}
+
+static void __exit aic94xx_exit(void)
+{
+	asd_remove_driver_attrs(&aic94xx_pci_driver.driver);
+	pci_unregister_driver(&aic94xx_pci_driver);
+	sas_release_transport(aic94xx_transport_template);
+	asd_destroy_global_caches();
+	asd_printk("%s version %s unloaded\n", ASD_DRIVER_DESCRIPTION,
+		   ASD_DRIVER_VERSION);
+}
+
+module_init(aic94xx_init);
+module_exit(aic94xx_exit);
+
+MODULE_AUTHOR("Luben Tuikov <luben_tuikov@adaptec.com>");
+MODULE_DESCRIPTION(ASD_DRIVER_DESCRIPTION);
+MODULE_LICENSE("GPL v2");
+MODULE_VERSION(ASD_DRIVER_VERSION);
diff --git a/drivers/scsi/aic94xx/aic94xx_reg.c b/drivers/scsi/aic94xx/aic94xx_reg.c
new file mode 100644
index 000000000000..f210dac3203d
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx_reg.c
@@ -0,0 +1,332 @@
+/*
+ * Aic94xx SAS/SATA driver register access.
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#include <linux/pci.h>
+#include "aic94xx_reg.h"
+#include "aic94xx.h"
+
+/* Writing to device address space.
+ * Offset comes before value to remind that the operation of
+ * this function is *offs = val.
+ */
+static inline void asd_write_byte(struct asd_ha_struct *asd_ha,
+				  unsigned long offs, u8 val)
+{
+	if (unlikely(asd_ha->iospace))
+		outb(val,
+		     (unsigned long)asd_ha->io_handle[0].addr + (offs & 0xFF));
+	else
+		writeb(val, asd_ha->io_handle[0].addr + offs);
+	wmb();
+}
+
+static inline void asd_write_word(struct asd_ha_struct *asd_ha,
+				  unsigned long offs, u16 val)
+{
+	if (unlikely(asd_ha->iospace))
+		outw(val,
+		     (unsigned long)asd_ha->io_handle[0].addr + (offs & 0xFF));
+	else
+		writew(val, asd_ha->io_handle[0].addr + offs);
+	wmb();
+}
+
+static inline void asd_write_dword(struct asd_ha_struct *asd_ha,
+				   unsigned long offs, u32 val)
+{
+	if (unlikely(asd_ha->iospace))
+		outl(val,
+		     (unsigned long)asd_ha->io_handle[0].addr + (offs & 0xFF));
+	else
+		writel(val, asd_ha->io_handle[0].addr + offs);
+	wmb();
+}
+
+/* Reading from device address space.
+ */
+static inline u8 asd_read_byte(struct asd_ha_struct *asd_ha,
+			       unsigned long offs)
+{
+	u8 val;
+	if (unlikely(asd_ha->iospace))
+		val = inb((unsigned long) asd_ha->io_handle[0].addr
+			  + (offs & 0xFF));
+	else
+		val = readb(asd_ha->io_handle[0].addr + offs);
+	rmb();
+	return val;
+}
+
+static inline u16 asd_read_word(struct asd_ha_struct *asd_ha,
+				unsigned long offs)
+{
+	u16 val;
+	if (unlikely(asd_ha->iospace))
+		val = inw((unsigned long)asd_ha->io_handle[0].addr
+			  + (offs & 0xFF));
+	else
+		val = readw(asd_ha->io_handle[0].addr + offs);
+	rmb();
+	return val;
+}
+
+static inline u32 asd_read_dword(struct asd_ha_struct *asd_ha,
+				 unsigned long offs)
+{
+	u32 val;
+	if (unlikely(asd_ha->iospace))
+		val = inl((unsigned long) asd_ha->io_handle[0].addr
+			  + (offs & 0xFF));
+	else
+		val = readl(asd_ha->io_handle[0].addr + offs);
+	rmb();
+	return val;
+}
+
+static inline u32 asd_mem_offs_swa(void)
+{
+	return 0;
+}
+
+static inline u32 asd_mem_offs_swc(void)
+{
+	return asd_mem_offs_swa() + MBAR0_SWA_SIZE;
+}
+
+static inline u32 asd_mem_offs_swb(void)
+{
+	return asd_mem_offs_swc() + MBAR0_SWC_SIZE + 0x20;
+}
+
+/* We know that the register wanted is in the range
+ * of the sliding window.
+ */
+#define ASD_READ_SW(ww, type, ord)                                     \
+static inline type asd_read_##ww##_##ord (struct asd_ha_struct *asd_ha,\
+					  u32 reg)                     \
+{                                                                      \
+	struct asd_ha_addrspace *io_handle = &asd_ha->io_handle[0];    \
+	u32 map_offs=(reg - io_handle-> ww##_base )+asd_mem_offs_##ww ();\
+	return asd_read_##ord (asd_ha, (unsigned long) map_offs);      \
+}
+
+#define ASD_WRITE_SW(ww, type, ord)                                    \
+static inline void asd_write_##ww##_##ord (struct asd_ha_struct *asd_ha,\
+				  u32 reg, type val)                   \
+{                                                                      \
+	struct asd_ha_addrspace *io_handle = &asd_ha->io_handle[0];    \
+	u32 map_offs=(reg - io_handle-> ww##_base )+asd_mem_offs_##ww ();\
+	asd_write_##ord (asd_ha, (unsigned long) map_offs, val);       \
+}
+
+ASD_READ_SW(swa, u8,  byte);
+ASD_READ_SW(swa, u16, word);
+ASD_READ_SW(swa, u32, dword);
+
+ASD_READ_SW(swb, u8,  byte);
+ASD_READ_SW(swb, u16, word);
+ASD_READ_SW(swb, u32, dword);
+
+ASD_READ_SW(swc, u8,  byte);
+ASD_READ_SW(swc, u16, word);
+ASD_READ_SW(swc, u32, dword);
+
+ASD_WRITE_SW(swa, u8,  byte);
+ASD_WRITE_SW(swa, u16, word);
+ASD_WRITE_SW(swa, u32, dword);
+
+ASD_WRITE_SW(swb, u8,  byte);
+ASD_WRITE_SW(swb, u16, word);
+ASD_WRITE_SW(swb, u32, dword);
+
+ASD_WRITE_SW(swc, u8,  byte);
+ASD_WRITE_SW(swc, u16, word);
+ASD_WRITE_SW(swc, u32, dword);
+
+/*
+ * A word about sliding windows:
+ * MBAR0 is divided into sliding windows A, C and B, in that order.
+ * SWA starts at offset 0 of MBAR0, up to 0x57, with size 0x58 bytes.
+ * SWC starts at offset 0x58 of MBAR0, up to 0x60, with size 0x8 bytes.
+ * From 0x60 to 0x7F, we have a copy of PCI config space 0x60-0x7F.
+ * SWB starts at offset 0x80 of MBAR0 and extends to the end of MBAR0.
+ * See asd_init_sw() in aic94xx_hwi.c
+ *
+ * We map the most common registers we'd access of the internal 4GB
+ * host adapter memory space.  If a register/internal memory location
+ * is wanted which is not mapped, we slide SWB, by paging it,
+ * see asd_move_swb() in aic94xx_reg.c.
+ */
+
+/**
+ * asd_move_swb -- move sliding window B
+ * @asd_ha: pointer to host adapter structure
+ * @reg: register desired to be within range of the new window
+ */
+static inline void asd_move_swb(struct asd_ha_struct *asd_ha, u32 reg)
+{
+	u32 base = reg & ~(MBAR0_SWB_SIZE-1);
+	pci_write_config_dword(asd_ha->pcidev, PCI_CONF_MBAR0_SWB, base);
+	asd_ha->io_handle[0].swb_base = base;
+}
+
+static void __asd_write_reg_byte(struct asd_ha_struct *asd_ha, u32 reg, u8 val)
+{
+	struct asd_ha_addrspace *io_handle=&asd_ha->io_handle[0];
+	BUG_ON(reg >= 0xC0000000 || reg < ALL_BASE_ADDR);
+	if (io_handle->swa_base <= reg
+	    && reg < io_handle->swa_base + MBAR0_SWA_SIZE)
+		asd_write_swa_byte (asd_ha, reg,val);
+	else if (io_handle->swb_base <= reg
+		 && reg < io_handle->swb_base + MBAR0_SWB_SIZE)
+		asd_write_swb_byte (asd_ha, reg, val);
+	else if (io_handle->swc_base <= reg
+		 && reg < io_handle->swc_base + MBAR0_SWC_SIZE)
+		asd_write_swc_byte (asd_ha, reg, val);
+	else {
+		/* Ok, we have to move SWB */
+		asd_move_swb(asd_ha, reg);
+		asd_write_swb_byte (asd_ha, reg, val);
+	}
+}
+
+#define ASD_WRITE_REG(type, ord)                                  \
+void asd_write_reg_##ord (struct asd_ha_struct *asd_ha, u32 reg, type val)\
+{                                                                 \
+	struct asd_ha_addrspace *io_handle=&asd_ha->io_handle[0]; \
+	unsigned long flags;                                      \
+	BUG_ON(reg >= 0xC0000000 || reg < ALL_BASE_ADDR);         \
+	spin_lock_irqsave(&asd_ha->iolock, flags);                \
+	if (io_handle->swa_base <= reg                            \
+	    && reg < io_handle->swa_base + MBAR0_SWA_SIZE)        \
+		asd_write_swa_##ord (asd_ha, reg,val);            \
+	else if (io_handle->swb_base <= reg                       \
+		 && reg < io_handle->swb_base + MBAR0_SWB_SIZE)   \
+		asd_write_swb_##ord (asd_ha, reg, val);           \
+	else if (io_handle->swc_base <= reg                       \
+		 && reg < io_handle->swc_base + MBAR0_SWC_SIZE)   \
+		asd_write_swc_##ord (asd_ha, reg, val);           \
+	else {                                                    \
+		/* Ok, we have to move SWB */                     \
+		asd_move_swb(asd_ha, reg);                        \
+		asd_write_swb_##ord (asd_ha, reg, val);           \
+	}                                                         \
+	spin_unlock_irqrestore(&asd_ha->iolock, flags);           \
+}
+
+ASD_WRITE_REG(u8, byte);
+ASD_WRITE_REG(u16,word);
+ASD_WRITE_REG(u32,dword);
+
+static u8 __asd_read_reg_byte(struct asd_ha_struct *asd_ha, u32 reg)
+{
+	struct asd_ha_addrspace *io_handle=&asd_ha->io_handle[0];
+	u8 val;
+	BUG_ON(reg >= 0xC0000000 || reg < ALL_BASE_ADDR);
+	if (io_handle->swa_base <= reg
+	    && reg < io_handle->swa_base + MBAR0_SWA_SIZE)
+		val = asd_read_swa_byte (asd_ha, reg);
+	else if (io_handle->swb_base <= reg
+		 && reg < io_handle->swb_base + MBAR0_SWB_SIZE)
+		val = asd_read_swb_byte (asd_ha, reg);
+	else if (io_handle->swc_base <= reg
+		 && reg < io_handle->swc_base + MBAR0_SWC_SIZE)
+		val = asd_read_swc_byte (asd_ha, reg);
+	else {
+		/* Ok, we have to move SWB */
+		asd_move_swb(asd_ha, reg);
+		val = asd_read_swb_byte (asd_ha, reg);
+	}
+	return val;
+}
+
+#define ASD_READ_REG(type, ord)                                   \
+type asd_read_reg_##ord (struct asd_ha_struct *asd_ha, u32 reg)   \
+{                                                                 \
+	struct asd_ha_addrspace *io_handle=&asd_ha->io_handle[0]; \
+	type val;                                                 \
+	unsigned long flags;                                      \
+	BUG_ON(reg >= 0xC0000000 || reg < ALL_BASE_ADDR);         \
+	spin_lock_irqsave(&asd_ha->iolock, flags);                \
+	if (io_handle->swa_base <= reg                            \
+	    && reg < io_handle->swa_base + MBAR0_SWA_SIZE)        \
+		val = asd_read_swa_##ord (asd_ha, reg);           \
+	else if (io_handle->swb_base <= reg                       \
+		 && reg < io_handle->swb_base + MBAR0_SWB_SIZE)   \
+		val = asd_read_swb_##ord (asd_ha, reg);           \
+	else if (io_handle->swc_base <= reg                       \
+		 && reg < io_handle->swc_base + MBAR0_SWC_SIZE)   \
+		val = asd_read_swc_##ord (asd_ha, reg);           \
+	else {                                                    \
+		/* Ok, we have to move SWB */                     \
+		asd_move_swb(asd_ha, reg);                        \
+		val = asd_read_swb_##ord (asd_ha, reg);           \
+	}                                                         \
+	spin_unlock_irqrestore(&asd_ha->iolock, flags);           \
+	return val;                                               \
+}
+
+ASD_READ_REG(u8, byte);
+ASD_READ_REG(u16,word);
+ASD_READ_REG(u32,dword);
+
+/**
+ * asd_read_reg_string -- read a string of bytes from io space memory
+ * @asd_ha: pointer to host adapter structure
+ * @dst: pointer to a destination buffer where data will be written to
+ * @offs: start offset (register) to read from
+ * @count: number of bytes to read
+ */
+void asd_read_reg_string(struct asd_ha_struct *asd_ha, void *dst,
+			 u32 offs, int count)
+{
+	u8 *p = dst;
+	unsigned long flags;
+
+	spin_lock_irqsave(&asd_ha->iolock, flags);
+	for ( ; count > 0; count--, offs++, p++)
+		*p = __asd_read_reg_byte(asd_ha, offs);
+	spin_unlock_irqrestore(&asd_ha->iolock, flags);
+}
+
+/**
+ * asd_write_reg_string -- write a string of bytes to io space memory
+ * @asd_ha: pointer to host adapter structure
+ * @src: pointer to source buffer where data will be read from
+ * @offs: start offset (register) to write to
+ * @count: number of bytes to write
+ */
+void asd_write_reg_string(struct asd_ha_struct *asd_ha, void *src,
+			  u32 offs, int count)
+{
+	u8 *p = src;
+	unsigned long flags;
+
+	spin_lock_irqsave(&asd_ha->iolock, flags);
+	for ( ; count > 0; count--, offs++, p++)
+		__asd_write_reg_byte(asd_ha, offs, *p);
+	spin_unlock_irqrestore(&asd_ha->iolock, flags);
+}
diff --git a/drivers/scsi/aic94xx/aic94xx_reg.h b/drivers/scsi/aic94xx/aic94xx_reg.h
new file mode 100644
index 000000000000..2279307fd27e
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx_reg.h
@@ -0,0 +1,302 @@
+/*
+ * Aic94xx SAS/SATA driver hardware registers definitions.
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#ifndef _AIC94XX_REG_H_
+#define _AIC94XX_REG_H_
+
+#include <asm/io.h>
+#include "aic94xx_hwi.h"
+
+/* Values */
+#define AIC9410_DEV_REV_B0            0x8
+
+/* MBAR0, SWA, SWB, SWC, internal memory space addresses */
+#define REG_BASE_ADDR                 0xB8000000
+#define REG_BASE_ADDR_CSEQCIO         0xB8002000
+#define REG_BASE_ADDR_EXSI            0xB8042800
+
+#define MBAR0_SWA_SIZE                0x58
+extern  u32    MBAR0_SWB_SIZE;
+#define MBAR0_SWC_SIZE                0x8
+
+/* MBAR1, points to On Chip Memory */
+#define OCM_BASE_ADDR                 0xA0000000
+#define OCM_MAX_SIZE                  0x20000
+
+/* Smallest address possible to reference */
+#define ALL_BASE_ADDR                 OCM_BASE_ADDR
+
+/* PCI configuration space registers */
+#define PCI_IOBAR_OFFSET              4
+
+#define PCI_CONF_MBAR1                0x6C
+#define PCI_CONF_MBAR0_SWA            0x70
+#define PCI_CONF_MBAR0_SWB            0x74
+#define PCI_CONF_MBAR0_SWC            0x78
+#define PCI_CONF_MBAR_KEY             0x7C
+#define PCI_CONF_FLSH_BAR             0xB8
+
+#include "aic94xx_reg_def.h"
+
+u8  asd_read_reg_byte(struct asd_ha_struct *asd_ha, u32 reg);
+u16 asd_read_reg_word(struct asd_ha_struct *asd_ha, u32 reg);
+u32 asd_read_reg_dword(struct asd_ha_struct *asd_ha, u32 reg);
+
+void asd_write_reg_byte(struct asd_ha_struct *asd_ha, u32 reg, u8 val);
+void asd_write_reg_word(struct asd_ha_struct *asd_ha, u32 reg, u16 val);
+void asd_write_reg_dword(struct asd_ha_struct *asd_ha, u32 reg, u32 val);
+
+void asd_read_reg_string(struct asd_ha_struct *asd_ha, void *dst,
+			 u32 offs, int count);
+void asd_write_reg_string(struct asd_ha_struct *asd_ha, void *src,
+			  u32 offs, int count);
+
+#define ASD_READ_OCM(type, ord, S)                                    \
+static inline type asd_read_ocm_##ord (struct asd_ha_struct *asd_ha,  \
+					 u32 offs)                    \
+{                                                                     \
+	struct asd_ha_addrspace *io_handle = &asd_ha->io_handle[1];   \
+	type val = read##S (io_handle->addr + (unsigned long) offs);  \
+	rmb();                                                        \
+	return val;                                                   \
+}
+
+ASD_READ_OCM(u8, byte, b);
+ASD_READ_OCM(u16,word, w);
+ASD_READ_OCM(u32,dword,l);
+
+#define ASD_WRITE_OCM(type, ord, S)                                    \
+static inline void asd_write_ocm_##ord (struct asd_ha_struct *asd_ha,  \
+					 u32 offs, type val)          \
+{                                                                     \
+	struct asd_ha_addrspace *io_handle = &asd_ha->io_handle[1];   \
+	write##S (val, io_handle->addr + (unsigned long) offs);       \
+	return;                                                       \
+}
+
+ASD_WRITE_OCM(u8, byte, b);
+ASD_WRITE_OCM(u16,word, w);
+ASD_WRITE_OCM(u32,dword,l);
+
+#define ASD_DDBSITE_READ(type, ord)                                        \
+static inline type asd_ddbsite_read_##ord (struct asd_ha_struct *asd_ha,   \
+					   u16 ddb_site_no,                \
+					   u16 offs)                       \
+{                                                                          \
+	asd_write_reg_word(asd_ha, ALTCIOADR, MnDDB_SITE + offs);          \
+	asd_write_reg_word(asd_ha, ADDBPTR, ddb_site_no);                  \
+	return asd_read_reg_##ord (asd_ha, CTXACCESS);                     \
+}
+
+ASD_DDBSITE_READ(u32, dword);
+ASD_DDBSITE_READ(u16, word);
+
+static inline u8 asd_ddbsite_read_byte(struct asd_ha_struct *asd_ha,
+				       u16 ddb_site_no,
+				       u16 offs)
+{
+	if (offs & 1)
+		return asd_ddbsite_read_word(asd_ha, ddb_site_no,
+					     offs & ~1) >> 8;
+	else
+		return asd_ddbsite_read_word(asd_ha, ddb_site_no,
+					     offs) & 0xFF;
+}
+
+
+#define ASD_DDBSITE_WRITE(type, ord)                                       \
+static inline void asd_ddbsite_write_##ord (struct asd_ha_struct *asd_ha,  \
+					u16 ddb_site_no,                   \
+					u16 offs, type val)                \
+{                                                                          \
+	asd_write_reg_word(asd_ha, ALTCIOADR, MnDDB_SITE + offs);          \
+	asd_write_reg_word(asd_ha, ADDBPTR, ddb_site_no);                  \
+	asd_write_reg_##ord (asd_ha, CTXACCESS, val);                      \
+}
+
+ASD_DDBSITE_WRITE(u32, dword);
+ASD_DDBSITE_WRITE(u16, word);
+
+static inline void asd_ddbsite_write_byte(struct asd_ha_struct *asd_ha,
+					  u16 ddb_site_no,
+					  u16 offs, u8 val)
+{
+	u16 base = offs & ~1;
+	u16 rval = asd_ddbsite_read_word(asd_ha, ddb_site_no, base);
+	if (offs & 1)
+		rval = (val << 8) | (rval & 0xFF);
+	else
+		rval = (rval & 0xFF00) | val;
+	asd_ddbsite_write_word(asd_ha, ddb_site_no, base, rval);
+}
+
+
+#define ASD_SCBSITE_READ(type, ord)                                        \
+static inline type asd_scbsite_read_##ord (struct asd_ha_struct *asd_ha,   \
+					   u16 scb_site_no,                \
+					   u16 offs)                       \
+{                                                                          \
+	asd_write_reg_word(asd_ha, ALTCIOADR, MnSCB_SITE + offs);          \
+	asd_write_reg_word(asd_ha, ASCBPTR, scb_site_no);                  \
+	return asd_read_reg_##ord (asd_ha, CTXACCESS);                     \
+}
+
+ASD_SCBSITE_READ(u32, dword);
+ASD_SCBSITE_READ(u16, word);
+
+static inline u8 asd_scbsite_read_byte(struct asd_ha_struct *asd_ha,
+				       u16 scb_site_no,
+				       u16 offs)
+{
+	if (offs & 1)
+		return asd_scbsite_read_word(asd_ha, scb_site_no,
+					     offs & ~1) >> 8;
+	else
+		return asd_scbsite_read_word(asd_ha, scb_site_no,
+					     offs) & 0xFF;
+}
+
+
+#define ASD_SCBSITE_WRITE(type, ord)                                       \
+static inline void asd_scbsite_write_##ord (struct asd_ha_struct *asd_ha,  \
+					u16 scb_site_no,                   \
+					u16 offs, type val)                \
+{                                                                          \
+	asd_write_reg_word(asd_ha, ALTCIOADR, MnSCB_SITE + offs);          \
+	asd_write_reg_word(asd_ha, ASCBPTR, scb_site_no);                  \
+	asd_write_reg_##ord (asd_ha, CTXACCESS, val);                      \
+}
+
+ASD_SCBSITE_WRITE(u32, dword);
+ASD_SCBSITE_WRITE(u16, word);
+
+static inline void asd_scbsite_write_byte(struct asd_ha_struct *asd_ha,
+					  u16 scb_site_no,
+					  u16 offs, u8 val)
+{
+	u16 base = offs & ~1;
+	u16 rval = asd_scbsite_read_word(asd_ha, scb_site_no, base);
+	if (offs & 1)
+		rval = (val << 8) | (rval & 0xFF);
+	else
+		rval = (rval & 0xFF00) | val;
+	asd_scbsite_write_word(asd_ha, scb_site_no, base, rval);
+}
+
+/**
+ * asd_ddbsite_update_word -- atomically update a word in a ddb site
+ * @asd_ha: pointer to host adapter structure
+ * @ddb_site_no: the DDB site number
+ * @offs: the offset into the DDB
+ * @oldval: old value found in that offset
+ * @newval: the new value to replace it
+ *
+ * This function is used when the sequencers are running and we need to
+ * update a DDB site atomically without expensive pausing and upausing
+ * of the sequencers and accessing the DDB site through the CIO bus.
+ *
+ * Return 0 on success; -EFAULT on parity error; -EAGAIN if the old value
+ * is different than the current value at that offset.
+ */
+static inline int asd_ddbsite_update_word(struct asd_ha_struct *asd_ha,
+					  u16 ddb_site_no, u16 offs,
+					  u16 oldval, u16 newval)
+{
+	u8  done;
+	u16 oval = asd_ddbsite_read_word(asd_ha, ddb_site_no, offs);
+	if (oval != oldval)
+		return -EAGAIN;
+	asd_write_reg_word(asd_ha, AOLDDATA, oldval);
+	asd_write_reg_word(asd_ha, ANEWDATA, newval);
+	do {
+		done = asd_read_reg_byte(asd_ha, ATOMICSTATCTL);
+	} while (!(done & ATOMICDONE));
+	if (done & ATOMICERR)
+		return -EFAULT;	  /* parity error */
+	else if (done & ATOMICWIN)
+		return 0;	  /* success */
+	else
+		return -EAGAIN;	  /* oldval different than current value */
+}
+
+static inline int asd_ddbsite_update_byte(struct asd_ha_struct *asd_ha,
+					  u16 ddb_site_no, u16 offs,
+					  u8 _oldval, u8 _newval)
+{
+	u16 base = offs & ~1;
+	u16 oval;
+	u16 nval = asd_ddbsite_read_word(asd_ha, ddb_site_no, base);
+	if (offs & 1) {
+		if ((nval >> 8) != _oldval)
+			return -EAGAIN;
+		nval = (_newval << 8) | (nval & 0xFF);
+		oval = (_oldval << 8) | (nval & 0xFF);
+	} else {
+		if ((nval & 0xFF) != _oldval)
+			return -EAGAIN;
+		nval = (nval & 0xFF00) | _newval;
+		oval = (nval & 0xFF00) | _oldval;
+	}
+	return asd_ddbsite_update_word(asd_ha, ddb_site_no, base, oval, nval);
+}
+
+static inline void asd_write_reg_addr(struct asd_ha_struct *asd_ha, u32 reg,
+				      dma_addr_t dma_handle)
+{
+	asd_write_reg_dword(asd_ha, reg,   ASD_BUSADDR_LO(dma_handle));
+	asd_write_reg_dword(asd_ha, reg+4, ASD_BUSADDR_HI(dma_handle));
+}
+
+static inline u32 asd_get_cmdctx_size(struct asd_ha_struct *asd_ha)
+{
+	/* DCHREVISION returns 0, possibly broken */
+	u32 ctxmemsize = asd_read_reg_dword(asd_ha, LmMnINT(0,0)) & CTXMEMSIZE;
+	return ctxmemsize ? 65536 : 32768;
+}
+
+static inline u32 asd_get_devctx_size(struct asd_ha_struct *asd_ha)
+{
+	u32 ctxmemsize = asd_read_reg_dword(asd_ha, LmMnINT(0,0)) & CTXMEMSIZE;
+	return ctxmemsize ? 8192 : 4096;
+}
+
+static inline void asd_disable_ints(struct asd_ha_struct *asd_ha)
+{
+	asd_write_reg_dword(asd_ha, CHIMINTEN, RST_CHIMINTEN);
+}
+
+static inline void asd_enable_ints(struct asd_ha_struct *asd_ha)
+{
+	/* Enable COM SAS interrupt on errors, COMSTAT */
+	asd_write_reg_dword(asd_ha, COMSTATEN,
+			    EN_CSBUFPERR | EN_CSERR | EN_OVLYERR);
+	/* Enable DCH SAS CFIFTOERR */
+	asd_write_reg_dword(asd_ha, DCHSTATUS, EN_CFIFTOERR);
+	/* Enable Host Device interrupts */
+	asd_write_reg_dword(asd_ha, CHIMINTEN, SET_CHIMINTEN);
+}
+
+#endif
diff --git a/drivers/scsi/aic94xx/aic94xx_reg_def.h b/drivers/scsi/aic94xx/aic94xx_reg_def.h
new file mode 100644
index 000000000000..b79f45f3ad47
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx_reg_def.h
@@ -0,0 +1,2398 @@
+/*
+ * Aic94xx SAS/SATA driver hardware registers defintions.
+ *
+ * Copyright (C) 2004 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2004 David Chaw <david_chaw@adaptec.com>
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * Luben Tuikov: Some register value updates to make it work with the window
+ * agnostic register r/w functions.  Some register corrections, sizes,
+ * etc.
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ * $Id: //depot/aic94xx/aic94xx_reg_def.h#27 $
+ *
+ */
+
+#ifndef _ADP94XX_REG_DEF_H_
+#define _ADP94XX_REG_DEF_H_
+
+/*
+ * Common definitions.
+ */
+#define CSEQ_MODE_PAGE_SIZE	0x200		/* CSEQ mode page size */
+#define LmSEQ_MODE_PAGE_SIZE	0x200		/* LmSEQ mode page size */
+#define LmSEQ_HOST_REG_SIZE   	0x4000		/* LmSEQ Host Register size */
+
+/********************* COM_SAS registers definition *************************/
+
+/* The base is REG_BASE_ADDR, defined in aic94xx_reg.h.
+ */
+
+/*
+ * CHIM Registers, Address Range : (0x00-0xFF)
+ */
+#define COMBIST		(REG_BASE_ADDR + 0x00)
+
+/* bits 31:24 */
+#define		L7BLKRST		0x80000000
+#define		L6BLKRST		0x40000000
+#define		L5BLKRST		0x20000000
+#define		L4BLKRST		0x10000000
+#define		L3BLKRST		0x08000000
+#define		L2BLKRST		0x04000000
+#define		L1BLKRST		0x02000000
+#define		L0BLKRST		0x01000000
+#define		LmBLKRST		0xFF000000
+#define LmBLKRST_COMBIST(phyid)		(1 << (24 + phyid))
+
+#define		OCMBLKRST		0x00400000
+#define		CTXMEMBLKRST		0x00200000
+#define		CSEQBLKRST		0x00100000
+#define		EXSIBLKRST		0x00040000
+#define		DPIBLKRST		0x00020000
+#define		DFIFBLKRST		0x00010000
+#define		HARDRST			0x00000200
+#define		COMBLKRST		0x00000100
+#define		FRCDFPERR		0x00000080
+#define		FRCCIOPERR		0x00000020
+#define		FRCBISTERR		0x00000010
+#define		COMBISTEN		0x00000004
+#define		COMBISTDONE		0x00000002	/* ro */
+#define 	COMBISTFAIL		0x00000001	/* ro */
+
+#define COMSTAT		(REG_BASE_ADDR + 0x04)
+
+#define		REQMBXREAD		0x00000040
+#define 	RSPMBXAVAIL		0x00000020
+#define 	CSBUFPERR		0x00000008
+#define		OVLYERR			0x00000004
+#define 	CSERR			0x00000002
+#define		OVLYDMADONE		0x00000001
+
+#define		COMSTAT_MASK		(REQMBXREAD | RSPMBXAVAIL | \
+					 CSBUFPERR | OVLYERR | CSERR |\
+					 OVLYDMADONE)
+
+#define COMSTATEN	(REG_BASE_ADDR + 0x08)
+
+#define		EN_REQMBXREAD		0x00000040
+#define		EN_RSPMBXAVAIL		0x00000020
+#define		EN_CSBUFPERR		0x00000008
+#define		EN_OVLYERR		0x00000004
+#define		EN_CSERR		0x00000002
+#define		EN_OVLYDONE		0x00000001
+
+#define SCBPRO		(REG_BASE_ADDR + 0x0C)
+
+#define		SCBCONS_MASK		0xFFFF0000
+#define		SCBPRO_MASK		0x0000FFFF
+
+#define CHIMREQMBX	(REG_BASE_ADDR + 0x10)
+
+#define CHIMRSPMBX	(REG_BASE_ADDR + 0x14)
+
+#define CHIMINT		(REG_BASE_ADDR + 0x18)
+
+#define		EXT_INT0		0x00000800
+#define		EXT_INT1		0x00000400
+#define		PORRSTDET		0x00000200
+#define		HARDRSTDET		0x00000100
+#define		DLAVAILQ		0x00000080	/* ro */
+#define		HOSTERR			0x00000040
+#define		INITERR			0x00000020
+#define		DEVINT			0x00000010
+#define		COMINT			0x00000008
+#define		DEVTIMER2		0x00000004
+#define		DEVTIMER1		0x00000002
+#define		DLAVAIL			0x00000001
+
+#define		CHIMINT_MASK		(HOSTERR | INITERR | DEVINT | COMINT |\
+					 DEVTIMER2 | DEVTIMER1 | DLAVAIL)
+
+#define 	DEVEXCEPT_MASK		(HOSTERR | INITERR | DEVINT | COMINT)
+
+#define CHIMINTEN	(REG_BASE_ADDR + 0x1C)
+
+#define		RST_EN_EXT_INT1		0x01000000
+#define		RST_EN_EXT_INT0		0x00800000
+#define		RST_EN_HOSTERR		0x00400000
+#define		RST_EN_INITERR		0x00200000
+#define		RST_EN_DEVINT		0x00100000
+#define		RST_EN_COMINT		0x00080000
+#define		RST_EN_DEVTIMER2	0x00040000
+#define		RST_EN_DEVTIMER1	0x00020000
+#define		RST_EN_DLAVAIL		0x00010000
+#define		SET_EN_EXT_INT1		0x00000100
+#define		SET_EN_EXT_INT0		0x00000080
+#define		SET_EN_HOSTERR		0x00000040
+#define		SET_EN_INITERR		0x00000020
+#define		SET_EN_DEVINT		0x00000010
+#define		SET_EN_COMINT		0x00000008
+#define		SET_EN_DEVTIMER2	0x00000004
+#define		SET_EN_DEVTIMER1	0x00000002
+#define		SET_EN_DLAVAIL		0x00000001
+
+#define		RST_CHIMINTEN		(RST_EN_HOSTERR | RST_EN_INITERR | \
+					 RST_EN_DEVINT | RST_EN_COMINT | \
+					 RST_EN_DEVTIMER2 | RST_EN_DEVTIMER1 |\
+					 RST_EN_DLAVAIL)
+
+#define		SET_CHIMINTEN		(SET_EN_HOSTERR | SET_EN_INITERR |\
+					 SET_EN_DEVINT | SET_EN_COMINT |\
+					 SET_EN_DLAVAIL)
+
+#define OVLYDMACTL	(REG_BASE_ADDR + 0x20)
+
+#define		OVLYADR_MASK		0x07FF0000
+#define		OVLYLSEQ_MASK		0x0000FF00
+#define		OVLYCSEQ		0x00000080
+#define		OVLYHALTERR		0x00000040
+#define		PIOCMODE		0x00000020
+#define		RESETOVLYDMA		0x00000008	/* wo */
+#define		STARTOVLYDMA		0x00000004
+#define		STOPOVLYDMA		0x00000002	/* wo */
+#define		OVLYDMAACT		0x00000001	/* ro */
+
+#define OVLYDMACNT	(REG_BASE_ADDR + 0x24)
+
+#define		OVLYDOMAIN1		0x20000000	/* ro */
+#define		OVLYDOMAIN0		0x10000000
+#define		OVLYBUFADR_MASK		0x007F0000
+#define		OVLYDMACNT_MASK		0x00003FFF
+
+#define OVLYDMAADR	(REG_BASE_ADDR + 0x28)
+
+#define DMAERR		(REG_BASE_ADDR + 0x30)
+
+#define		OVLYERRSTAT_MASK	0x0000FF00	/* ro */
+#define		CSERRSTAT_MASK		0x000000FF	/* ro */
+
+#define SPIODATA	(REG_BASE_ADDR + 0x34)
+
+/* 0x38 - 0x3C are reserved  */
+
+#define T1CNTRLR	(REG_BASE_ADDR + 0x40)
+
+#define		T1DONE			0x00010000	/* ro */
+#define		TIMER64			0x00000400
+#define		T1ENABLE		0x00000200
+#define		T1RELOAD		0x00000100
+#define		T1PRESCALER_MASK	0x00000003
+
+#define	T1CMPR		(REG_BASE_ADDR + 0x44)
+
+#define T1CNTR		(REG_BASE_ADDR + 0x48)
+
+#define T2CNTRLR	(REG_BASE_ADDR + 0x4C)
+
+#define		T2DONE			0x00010000	/* ro */
+#define		T2ENABLE		0x00000200
+#define		T2RELOAD		0x00000100
+#define		T2PRESCALER_MASK	0x00000003
+
+#define	T2CMPR		(REG_BASE_ADDR + 0x50)
+
+#define T2CNTR		(REG_BASE_ADDR + 0x54)
+
+/* 0x58h - 0xFCh are reserved */
+
+/*
+ * DCH_SAS Registers, Address Range : (0x800-0xFFF)
+ */
+#define CMDCTXBASE	(REG_BASE_ADDR + 0x800)
+
+#define DEVCTXBASE	(REG_BASE_ADDR + 0x808)
+
+#define CTXDOMAIN	(REG_BASE_ADDR + 0x810)
+
+#define		DEVCTXDOMAIN1		0x00000008	/* ro */
+#define		DEVCTXDOMAIN0		0x00000004
+#define		CMDCTXDOMAIN1		0x00000002	/* ro */
+#define		CMDCTXDOMAIN0		0x00000001
+
+#define DCHCTL		(REG_BASE_ADDR + 0x814)
+
+#define		OCMBISTREPAIR		0x00080000
+#define		OCMBISTEN		0x00040000
+#define		OCMBISTDN		0x00020000	/* ro */
+#define		OCMBISTFAIL		0x00010000	/* ro */
+#define		DDBBISTEN		0x00004000
+#define		DDBBISTDN		0x00002000	/* ro */
+#define		DDBBISTFAIL		0x00001000	/* ro */
+#define		SCBBISTEN		0x00000400
+#define		SCBBISTDN		0x00000200	/* ro */
+#define		SCBBISTFAIL		0x00000100	/* ro */
+
+#define		MEMSEL_MASK		0x000000E0
+#define		MEMSEL_CCM_LSEQ		0x00000000
+#define		MEMSEL_CCM_IOP		0x00000020
+#define		MEMSEL_CCM_SASCTL	0x00000040
+#define		MEMSEL_DCM_LSEQ		0x00000060
+#define		MEMSEL_DCM_IOP		0x00000080
+#define		MEMSEL_OCM		0x000000A0
+
+#define		FRCERR			0x00000010
+#define		AUTORLS			0x00000001
+
+#define DCHREVISION	(REG_BASE_ADDR + 0x818)
+
+#define		DCHREVISION_MASK	0x000000FF
+
+#define DCHSTATUS	(REG_BASE_ADDR + 0x81C)
+
+#define		EN_CFIFTOERR		0x00020000
+#define		CFIFTOERR		0x00000200
+#define		CSEQINT			0x00000100	/* ro */
+#define		LSEQ7INT		0x00000080	/* ro */
+#define		LSEQ6INT		0x00000040	/* ro */
+#define		LSEQ5INT		0x00000020	/* ro */
+#define		LSEQ4INT		0x00000010	/* ro */
+#define		LSEQ3INT		0x00000008	/* ro */
+#define		LSEQ2INT		0x00000004	/* ro */
+#define		LSEQ1INT		0x00000002	/* ro */
+#define		LSEQ0INT		0x00000001	/* ro */
+
+#define		LSEQINT_MASK		(LSEQ7INT | LSEQ6INT | LSEQ5INT |\
+					 LSEQ4INT | LSEQ3INT | LSEQ2INT	|\
+					 LSEQ1INT | LSEQ0INT)
+
+#define DCHDFIFDEBUG	(REG_BASE_ADDR + 0x820)
+#define		ENFAIRMST		0x00FF0000
+#define		DISWRMST9		0x00000200
+#define		DISWRMST8		0x00000100
+#define		DISRDMST		0x000000FF
+
+#define ATOMICSTATCTL	(REG_BASE_ADDR + 0x824)
+/* 8 bit wide */
+#define		AUTOINC			0x80
+#define		ATOMICERR		0x04
+#define		ATOMICWIN		0x02
+#define		ATOMICDONE		0x01
+
+
+#define ALTCIOADR	(REG_BASE_ADDR + 0x828)
+/* 16 bit; bits 8:0 define CIO addr space of CSEQ */
+
+#define ASCBPTR		(REG_BASE_ADDR + 0x82C)
+/* 16 bit wide */
+
+#define ADDBPTR		(REG_BASE_ADDR + 0x82E)
+/* 16 bit wide */
+
+#define ANEWDATA	(REG_BASE_ADDR + 0x830)
+/* 16 bit */
+
+#define AOLDDATA	(REG_BASE_ADDR + 0x834)
+/* 16 bit */
+
+#define CTXACCESS	(REG_BASE_ADDR + 0x838)
+/* 32 bit */
+
+/* 0x83Ch - 0xFFCh are reserved */
+
+/*
+ * ARP2 External Processor Registers, Address Range : (0x00-0x1F)
+ */
+#define ARP2CTL		0x00
+
+#define		FRCSCRPERR		0x00040000
+#define		FRCARP2PERR		0x00020000
+#define		FRCARP2ILLOPC		0x00010000
+#define		ENWAITTO		0x00008000
+#define		PERRORDIS		0x00004000
+#define		FAILDIS			0x00002000
+#define		CIOPERRDIS		0x00001000
+#define		BREAKEN3		0x00000800
+#define		BREAKEN2		0x00000400
+#define		BREAKEN1		0x00000200
+#define		BREAKEN0		0x00000100
+#define		EPAUSE			0x00000008
+#define		PAUSED			0x00000004	/* ro */
+#define		STEP			0x00000002
+#define		ARP2RESET		0x00000001	/* wo */
+
+#define ARP2INT		0x04
+
+#define		HALTCODE_MASK		0x00FF0000	/* ro */
+#define		ARP2WAITTO		0x00000100
+#define		ARP2HALTC		0x00000080
+#define		ARP2ILLOPC		0x00000040
+#define		ARP2PERR		0x00000020
+#define		ARP2CIOPERR		0x00000010
+#define		ARP2BREAK3		0x00000008
+#define		ARP2BREAK2		0x00000004
+#define		ARP2BREAK1		0x00000002
+#define		ARP2BREAK0		0x00000001
+
+#define ARP2INTEN	0x08
+
+#define		EN_ARP2WAITTO		0x00000100
+#define		EN_ARP2HALTC		0x00000080
+#define		EN_ARP2ILLOPC		0x00000040
+#define		EN_ARP2PERR		0x00000020
+#define		EN_ARP2CIOPERR		0x00000010
+#define		EN_ARP2BREAK3		0x00000008
+#define		EN_ARP2BREAK2		0x00000004
+#define		EN_ARP2BREAK1		0x00000002
+#define		EN_ARP2BREAK0		0x00000001
+
+#define ARP2BREAKADR01	0x0C
+
+#define		BREAKADR1_MASK		0x0FFF0000
+#define		BREAKADR0_MASK		0x00000FFF
+
+#define	ARP2BREAKADR23	0x10
+
+#define		BREAKADR3_MASK		0x0FFF0000
+#define		BREAKADR2_MASK		0x00000FFF
+
+/* 0x14h - 0x1Ch are reserved */
+
+/*
+ * ARP2 Registers, Address Range : (0x00-0x1F)
+ * The definitions have the same address offset for CSEQ and LmSEQ
+ * CIO Bus Registers.
+ */
+#define MODEPTR		0x00
+
+#define		DSTMODE			0xF0
+#define		SRCMODE			0x0F
+
+#define ALTMODE		0x01
+
+#define		ALTDMODE		0xF0
+#define		ALTSMODE		0x0F
+
+#define ATOMICXCHG	0x02
+
+#define FLAG		0x04
+
+#define		INTCODE_MASK		0xF0
+#define		ALTMODEV2		0x04
+#define		CARRY_INT		0x02
+#define		CARRY			0x01
+
+#define ARP2INTCTL	0x05
+
+#define 	PAUSEDIS		0x80
+#define		RSTINTCTL		0x40
+#define		POPALTMODE		0x08
+#define		ALTMODEV		0x04
+#define		INTMASK			0x02
+#define		IRET			0x01
+
+#define STACK		0x06
+
+#define FUNCTION1	0x07
+
+#define PRGMCNT		0x08
+
+#define ACCUM		0x0A
+
+#define SINDEX		0x0C
+
+#define DINDEX		0x0E
+
+#define ALLONES		0x10
+
+#define ALLZEROS	0x11
+
+#define SINDIR		0x12
+
+#define DINDIR		0x13
+
+#define JUMLDIR		0x14
+
+#define ARP2HALTCODE	0x15
+
+#define CURRADDR	0x16
+
+#define LASTADDR	0x18
+
+#define NXTLADDR	0x1A
+
+#define DBGPORTPTR	0x1C
+
+#define DBGPORT		0x1D
+
+/*
+ * CIO Registers.
+ * The definitions have the same address offset for CSEQ and LmSEQ
+ * CIO Bus Registers.
+ */
+#define MnSCBPTR      	0x20
+
+#define MnDDBPTR      	0x22
+
+#define SCRATCHPAGE	0x24
+
+#define MnSCRATCHPAGE	0x25
+
+#define SCRATCHPAGESV	0x26
+
+#define MnSCRATCHPAGESV	0x27
+
+#define MnDMAERRS	0x46
+
+#define MnSGDMAERRS	0x47
+
+#define MnSGBUF		0x53
+
+#define MnSGDMASTAT	0x5b
+
+#define MnDDMACTL	0x5c	/* RAZOR.rspec.fm rev 1.5 is wrong */
+
+#define MnDDMASTAT	0x5d	/* RAZOR.rspec.fm rev 1.5 is wrong */
+
+#define MnDDMAMODE	0x5e	/* RAZOR.rspec.fm rev 1.5 is wrong */
+
+#define MnDMAENG	0x60
+
+#define MnPIPECTL	0x61
+
+#define MnSGBADR	0x65
+
+#define MnSCB_SITE	0x100
+
+#define MnDDB_SITE	0x180
+
+/*
+ * The common definitions below have the same address offset for both
+ * CSEQ and LmSEQ.
+ */
+#define BISTCTL0	0x4C
+
+#define BISTCTL1	0x50
+
+#define MAPPEDSCR	0x800
+
+/*
+ * CSEQ Host Register, Address Range : (0x000-0xFFC)
+ */
+#define CSEQ_HOST_REG_BASE_ADR		0xB8001000
+
+#define CARP2CTL			(CSEQ_HOST_REG_BASE_ADR	+ ARP2CTL)
+
+#define CARP2INT			(CSEQ_HOST_REG_BASE_ADR	+ ARP2INT)
+
+#define CARP2INTEN			(CSEQ_HOST_REG_BASE_ADR	+ ARP2INTEN)
+
+#define CARP2BREAKADR01			(CSEQ_HOST_REG_BASE_ADR+ARP2BREAKADR01)
+
+#define CARP2BREAKADR23			(CSEQ_HOST_REG_BASE_ADR+ARP2BREAKADR23)
+
+#define CBISTCTL			(CSEQ_HOST_REG_BASE_ADR	+ BISTCTL1)
+
+#define		CSEQRAMBISTEN		0x00000040
+#define		CSEQRAMBISTDN		0x00000020	/* ro */
+#define		CSEQRAMBISTFAIL		0x00000010	/* ro */
+#define		CSEQSCRBISTEN		0x00000004
+#define		CSEQSCRBISTDN		0x00000002	/* ro */
+#define		CSEQSCRBISTFAIL		0x00000001	/* ro */
+
+#define CMAPPEDSCR			(CSEQ_HOST_REG_BASE_ADR	+ MAPPEDSCR)
+
+/*
+ * CSEQ CIO Bus Registers, Address Range : (0x0000-0x1FFC)
+ * 16 modes, each mode is 512 bytes.
+ * Unless specified, the register should valid for all modes.
+ */
+#define CSEQ_CIO_REG_BASE_ADR		REG_BASE_ADDR_CSEQCIO
+
+#define CSEQm_CIO_REG(Mode, Reg) \
+		(CSEQ_CIO_REG_BASE_ADR  + \
+		((u32) (Mode) * CSEQ_MODE_PAGE_SIZE) + (u32) (Reg))
+
+#define CMODEPTR	(CSEQ_CIO_REG_BASE_ADR + MODEPTR)
+
+#define CALTMODE	(CSEQ_CIO_REG_BASE_ADR + ALTMODE)
+
+#define CATOMICXCHG	(CSEQ_CIO_REG_BASE_ADR + ATOMICXCHG)
+
+#define CFLAG		(CSEQ_CIO_REG_BASE_ADR + FLAG)
+
+#define CARP2INTCTL	(CSEQ_CIO_REG_BASE_ADR + ARP2INTCTL)
+
+#define CSTACK		(CSEQ_CIO_REG_BASE_ADR + STACK)
+
+#define CFUNCTION1	(CSEQ_CIO_REG_BASE_ADR + FUNCTION1)
+
+#define CPRGMCNT	(CSEQ_CIO_REG_BASE_ADR + PRGMCNT)
+
+#define CACCUM		(CSEQ_CIO_REG_BASE_ADR + ACCUM)
+
+#define CSINDEX		(CSEQ_CIO_REG_BASE_ADR + SINDEX)
+
+#define CDINDEX		(CSEQ_CIO_REG_BASE_ADR + DINDEX)
+
+#define CALLONES	(CSEQ_CIO_REG_BASE_ADR + ALLONES)
+
+#define CALLZEROS	(CSEQ_CIO_REG_BASE_ADR + ALLZEROS)
+
+#define CSINDIR		(CSEQ_CIO_REG_BASE_ADR + SINDIR)
+
+#define CDINDIR		(CSEQ_CIO_REG_BASE_ADR + DINDIR)
+
+#define CJUMLDIR	(CSEQ_CIO_REG_BASE_ADR + JUMLDIR)
+
+#define CARP2HALTCODE	(CSEQ_CIO_REG_BASE_ADR + ARP2HALTCODE)
+
+#define CCURRADDR	(CSEQ_CIO_REG_BASE_ADR + CURRADDR)
+
+#define CLASTADDR	(CSEQ_CIO_REG_BASE_ADR + LASTADDR)
+
+#define CNXTLADDR	(CSEQ_CIO_REG_BASE_ADR + NXTLADDR)
+
+#define CDBGPORTPTR	(CSEQ_CIO_REG_BASE_ADR + DBGPORTPTR)
+
+#define CDBGPORT	(CSEQ_CIO_REG_BASE_ADR + DBGPORT)
+
+#define CSCRATCHPAGE	(CSEQ_CIO_REG_BASE_ADR + SCRATCHPAGE)
+
+#define CMnSCBPTR(Mode)       CSEQm_CIO_REG(Mode, MnSCBPTR)
+
+#define CMnDDBPTR(Mode)       CSEQm_CIO_REG(Mode, MnDDBPTR)
+
+#define CMnSCRATCHPAGE(Mode)		CSEQm_CIO_REG(Mode, MnSCRATCHPAGE)
+
+#define CLINKCON	(CSEQ_CIO_REG_BASE_ADR + 0x28)
+
+#define	CCIOAACESS	(CSEQ_CIO_REG_BASE_ADR + 0x2C)
+
+/* mode 0-7 */
+#define MnREQMBX 0x30
+#define CMnREQMBX(Mode)			CSEQm_CIO_REG(Mode, 0x30)
+
+/* mode 8 */
+#define CSEQCON				CSEQm_CIO_REG(8, 0x30)
+
+/* mode 0-7 */
+#define MnRSPMBX 0x34
+#define CMnRSPMBX(Mode)			CSEQm_CIO_REG(Mode, 0x34)
+
+/* mode 8 */
+#define CSEQCOMCTL			CSEQm_CIO_REG(8, 0x34)
+
+/* mode 8 */
+#define CSEQCOMSTAT			CSEQm_CIO_REG(8, 0x35)
+
+/* mode 8 */
+#define CSEQCOMINTEN			CSEQm_CIO_REG(8, 0x36)
+
+/* mode 8 */
+#define CSEQCOMDMACTL			CSEQm_CIO_REG(8, 0x37)
+
+#define		CSHALTERR		0x10
+#define		RESETCSDMA		0x08		/* wo */
+#define		STARTCSDMA		0x04
+#define		STOPCSDMA		0x02		/* wo */
+#define		CSDMAACT		0x01		/* ro */
+
+/* mode 0-7 */
+#define MnINT 0x38
+#define CMnINT(Mode)			CSEQm_CIO_REG(Mode, 0x38)
+
+#define		CMnREQMBXE		0x02
+#define		CMnRSPMBXF		0x01
+#define		CMnINT_MASK		0x00000003
+
+/* mode 8 */
+#define CSEQREQMBX			CSEQm_CIO_REG(8, 0x38)
+
+/* mode 0-7 */
+#define MnINTEN 0x3C
+#define CMnINTEN(Mode)			CSEQm_CIO_REG(Mode, 0x3C)
+
+#define		EN_CMnRSPMBXF		0x01
+
+/* mode 8 */
+#define CSEQRSPMBX			CSEQm_CIO_REG(8, 0x3C)
+
+/* mode 8 */
+#define CSDMAADR			CSEQm_CIO_REG(8, 0x40)
+
+/* mode 8 */
+#define CSDMACNT			CSEQm_CIO_REG(8, 0x48)
+
+/* mode 8 */
+#define CSEQDLCTL			CSEQm_CIO_REG(8, 0x4D)
+
+#define		DONELISTEND		0x10
+#define 	DONELISTSIZE_MASK	0x0F
+#define		DONELISTSIZE_8ELEM	0x01
+#define		DONELISTSIZE_16ELEM	0x02
+#define		DONELISTSIZE_32ELEM	0x03
+#define		DONELISTSIZE_64ELEM	0x04
+#define		DONELISTSIZE_128ELEM	0x05
+#define		DONELISTSIZE_256ELEM	0x06
+#define		DONELISTSIZE_512ELEM	0x07
+#define		DONELISTSIZE_1024ELEM	0x08
+#define		DONELISTSIZE_2048ELEM	0x09
+#define		DONELISTSIZE_4096ELEM	0x0A
+#define		DONELISTSIZE_8192ELEM	0x0B
+#define		DONELISTSIZE_16384ELEM	0x0C
+
+/* mode 8 */
+#define CSEQDLOFFS			CSEQm_CIO_REG(8, 0x4E)
+
+/* mode 11 */
+#define CM11INTVEC0			CSEQm_CIO_REG(11, 0x50)
+
+/* mode 11 */
+#define CM11INTVEC1			CSEQm_CIO_REG(11, 0x52)
+
+/* mode 11 */
+#define CM11INTVEC2			CSEQm_CIO_REG(11, 0x54)
+
+#define	CCONMSK	  			(CSEQ_CIO_REG_BASE_ADR + 0x60)
+
+#define	CCONEXIST			(CSEQ_CIO_REG_BASE_ADR + 0x61)
+
+#define	CCONMODE			(CSEQ_CIO_REG_BASE_ADR + 0x62)
+
+#define CTIMERCALC			(CSEQ_CIO_REG_BASE_ADR + 0x64)
+
+#define CINTDIS				(CSEQ_CIO_REG_BASE_ADR + 0x68)
+
+/* mode 8, 32x32 bits, 128 bytes of mapped buffer */
+#define CSBUFFER			CSEQm_CIO_REG(8, 0x80)
+
+#define	CSCRATCH			(CSEQ_CIO_REG_BASE_ADR + 0x1C0)
+
+/* mode 0-8 */
+#define CMnSCRATCH(Mode)		CSEQm_CIO_REG(Mode, 0x1E0)
+
+/*
+ * CSEQ Mapped Instruction RAM Page, Address Range : (0x0000-0x1FFC)
+ */
+#define CSEQ_RAM_REG_BASE_ADR		0xB8004000
+
+/*
+ * The common definitions below have the same address offset for all the Link
+ * sequencers.
+ */
+#define MODECTL		0x40
+
+#define DBGMODE		0x44
+
+#define CONTROL		0x48
+#define LEDTIMER		0x00010000
+#define LEDTIMERS_10us		0x00000000
+#define LEDTIMERS_1ms		0x00000800
+#define LEDTIMERS_100ms		0x00001000
+#define LEDMODE_TXRX		0x00000000
+#define LEDMODE_CONNECTED	0x00000200
+#define LEDPOL			0x00000100
+
+#define LSEQRAM		0x1000
+
+/*
+ * LmSEQ Host Registers, Address Range : (0x0000-0x3FFC)
+ */
+#define LSEQ0_HOST_REG_BASE_ADR		0xB8020000
+#define LSEQ1_HOST_REG_BASE_ADR		0xB8024000
+#define LSEQ2_HOST_REG_BASE_ADR		0xB8028000
+#define LSEQ3_HOST_REG_BASE_ADR		0xB802C000
+#define LSEQ4_HOST_REG_BASE_ADR		0xB8030000
+#define LSEQ5_HOST_REG_BASE_ADR		0xB8034000
+#define LSEQ6_HOST_REG_BASE_ADR		0xB8038000
+#define LSEQ7_HOST_REG_BASE_ADR		0xB803C000
+
+#define LmARP2CTL(LinkNum)		(LSEQ0_HOST_REG_BASE_ADR +	  \
+					((LinkNum)*LmSEQ_HOST_REG_SIZE) + \
+					ARP2CTL)
+
+#define LmARP2INT(LinkNum)		(LSEQ0_HOST_REG_BASE_ADR +	  \
+					((LinkNum)*LmSEQ_HOST_REG_SIZE) + \
+					ARP2INT)
+
+#define LmARP2INTEN(LinkNum)		(LSEQ0_HOST_REG_BASE_ADR +	  \
+					((LinkNum)*LmSEQ_HOST_REG_SIZE) + \
+					ARP2INTEN)
+
+#define LmDBGMODE(LinkNum)		(LSEQ0_HOST_REG_BASE_ADR +	  \
+					((LinkNum)*LmSEQ_HOST_REG_SIZE) + \
+					DBGMODE)
+
+#define LmCONTROL(LinkNum)		(LSEQ0_HOST_REG_BASE_ADR +	  \
+					((LinkNum)*LmSEQ_HOST_REG_SIZE) + \
+					CONTROL)
+
+#define LmARP2BREAKADR01(LinkNum)	(LSEQ0_HOST_REG_BASE_ADR +	  \
+					((LinkNum)*LmSEQ_HOST_REG_SIZE) + \
+					ARP2BREAKADR01)
+
+#define LmARP2BREAKADR23(LinkNum)	(LSEQ0_HOST_REG_BASE_ADR +	  \
+					((LinkNum)*LmSEQ_HOST_REG_SIZE) + \
+					ARP2BREAKADR23)
+
+#define LmMODECTL(LinkNum)		(LSEQ0_HOST_REG_BASE_ADR +	  \
+					((LinkNum)*LmSEQ_HOST_REG_SIZE) + \
+					MODECTL)
+
+#define		LmAUTODISCI		0x08000000
+#define		LmDSBLBITLT		0x04000000
+#define		LmDSBLANTT		0x02000000
+#define		LmDSBLCRTT		0x01000000
+#define		LmDSBLCONT		0x00000100
+#define		LmPRIMODE		0x00000080
+#define		LmDSBLHOLD		0x00000040
+#define		LmDISACK		0x00000020
+#define		LmBLIND48		0x00000010
+#define		LmRCVMODE_MASK		0x0000000C
+#define		LmRCVMODE_PLD		0x00000000
+#define		LmRCVMODE_HPC		0x00000004
+
+#define LmDBGMODE(LinkNum)		(LSEQ0_HOST_REG_BASE_ADR +	  \
+					((LinkNum)*LmSEQ_HOST_REG_SIZE) + \
+					DBGMODE)
+
+#define		LmFRCPERR		0x80000000
+#define		LmMEMSEL_MASK		0x30000000
+#define		LmFRCRBPERR		0x00000000
+#define		LmFRCTBPERR		0x10000000
+#define		LmFRCSGBPERR		0x20000000
+#define		LmFRCARBPERR		0x30000000
+#define		LmRCVIDW		0x00080000
+#define		LmINVDWERR		0x00040000
+#define		LmRCVDISP		0x00004000
+#define		LmDISPERR		0x00002000
+#define		LmDSBLDSCR		0x00000800
+#define		LmDSBLSCR		0x00000400
+#define		LmFRCNAK		0x00000200
+#define		LmFRCROFS		0x00000100
+#define		LmFRCCRC		0x00000080
+#define		LmFRMTYPE_MASK		0x00000070
+#define		LmSG_DATA		0x00000000
+#define		LmSG_COMMAND		0x00000010
+#define		LmSG_TASK		0x00000020
+#define		LmSG_TGTXFER		0x00000030
+#define		LmSG_RESPONSE		0x00000040
+#define		LmSG_IDENADDR		0x00000050
+#define		LmSG_OPENADDR		0x00000060
+#define		LmDISCRCGEN		0x00000008
+#define		LmDISCRCCHK		0x00000004
+#define		LmSSXMTFRM		0x00000002
+#define		LmSSRCVFRM		0x00000001
+
+#define LmCONTROL(LinkNum)		(LSEQ0_HOST_REG_BASE_ADR +	  \
+					((LinkNum)*LmSEQ_HOST_REG_SIZE) + \
+					CONTROL)
+
+#define		LmSTEPXMTFRM		0x00000002
+#define		LmSTEPRCVFRM		0x00000001
+
+#define LmBISTCTL0(LinkNum)		(LSEQ0_HOST_REG_BASE_ADR +	  \
+					((LinkNum)*LmSEQ_HOST_REG_SIZE) + \
+					BISTCTL0)
+
+#define		ARBBISTEN		0x40000000
+#define		ARBBISTDN		0x20000000	/* ro */
+#define		ARBBISTFAIL		0x10000000	/* ro */
+#define		TBBISTEN		0x00000400
+#define		TBBISTDN		0x00000200	/* ro */
+#define		TBBISTFAIL		0x00000100	/* ro */
+#define		RBBISTEN		0x00000040
+#define		RBBISTDN		0x00000020	/* ro */
+#define		RBBISTFAIL		0x00000010	/* ro */
+#define		SGBISTEN		0x00000004
+#define		SGBISTDN		0x00000002	/* ro */
+#define		SGBISTFAIL		0x00000001	/* ro */
+
+#define LmBISTCTL1(LinkNum)		(LSEQ0_HOST_REG_BASE_ADR +	 \
+					((LinkNum)*LmSEQ_HOST_REG_SIZE) +\
+					BISTCTL1)
+
+#define		LmRAMPAGE1		0x00000200
+#define		LmRAMPAGE0		0x00000100
+#define		LmIMEMBISTEN		0x00000040
+#define		LmIMEMBISTDN		0x00000020	/* ro */
+#define		LmIMEMBISTFAIL		0x00000010	/* ro */
+#define		LmSCRBISTEN		0x00000004
+#define		LmSCRBISTDN		0x00000002	/* ro */
+#define		LmSCRBISTFAIL		0x00000001	/* ro */
+#define		LmRAMPAGE		(LmRAMPAGE1 + LmRAMPAGE0)
+#define		LmRAMPAGE_LSHIFT	0x8
+
+#define LmSCRATCH(LinkNum)		(LSEQ0_HOST_REG_BASE_ADR +	   \
+					((LinkNum) * LmSEQ_HOST_REG_SIZE) +\
+					MAPPEDSCR)
+
+#define LmSEQRAM(LinkNum)		(LSEQ0_HOST_REG_BASE_ADR +	   \
+					((LinkNum) * LmSEQ_HOST_REG_SIZE) +\
+					LSEQRAM)
+
+/*
+ * LmSEQ CIO Bus Register, Address Range : (0x0000-0xFFC)
+ * 8 modes, each mode is 512 bytes.
+ * Unless specified, the register should valid for all modes.
+ */
+#define LmSEQ_CIOBUS_REG_BASE		0x2000
+
+#define  LmSEQ_PHY_BASE(Mode, LinkNum) \
+		(LSEQ0_HOST_REG_BASE_ADR + \
+		(LmSEQ_HOST_REG_SIZE * (u32) (LinkNum)) + \
+		LmSEQ_CIOBUS_REG_BASE + \
+		((u32) (Mode) * LmSEQ_MODE_PAGE_SIZE))
+
+#define  LmSEQ_PHY_REG(Mode, LinkNum, Reg) \
+                 (LmSEQ_PHY_BASE(Mode, LinkNum) + (u32) (Reg))
+
+#define LmMODEPTR(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, MODEPTR)
+
+#define LmALTMODE(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, ALTMODE)
+
+#define LmATOMICXCHG(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, ATOMICXCHG)
+
+#define LmFLAG(LinkNum)			LmSEQ_PHY_REG(0, LinkNum, FLAG)
+
+#define LmARP2INTCTL(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, ARP2INTCTL)
+
+#define LmSTACK(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, STACK)
+
+#define LmFUNCTION1(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, FUNCTION1)
+
+#define LmPRGMCNT(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, PRGMCNT)
+
+#define LmACCUM(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, ACCUM)
+
+#define LmSINDEX(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, SINDEX)
+
+#define LmDINDEX(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, DINDEX)
+
+#define LmALLONES(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, ALLONES)
+
+#define LmALLZEROS(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, ALLZEROS)
+
+#define LmSINDIR(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, SINDIR)
+
+#define LmDINDIR(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, DINDIR)
+
+#define LmJUMLDIR(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, JUMLDIR)
+
+#define LmARP2HALTCODE(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, ARP2HALTCODE)
+
+#define LmCURRADDR(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, CURRADDR)
+
+#define LmLASTADDR(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, LASTADDR)
+
+#define LmNXTLADDR(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, NXTLADDR)
+
+#define LmDBGPORTPTR(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, DBGPORTPTR)
+
+#define LmDBGPORT(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, DBGPORT)
+
+#define LmSCRATCHPAGE(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, SCRATCHPAGE)
+
+#define LmMnSCRATCHPAGE(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 	\
+						      MnSCRATCHPAGE)
+
+#define LmTIMERCALC(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0x28)
+
+#define LmREQMBX(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0x30)
+
+#define LmRSPMBX(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0x34)
+
+#define LmMnINT(LinkNum, Mode)		LmSEQ_PHY_REG(Mode, LinkNum, 0x38)
+
+#define		CTXMEMSIZE		0x80000000	/* ro */
+#define		LmACKREQ		0x08000000
+#define		LmNAKREQ		0x04000000
+#define		LmMnXMTERR		0x02000000
+#define		LmM5OOBSVC		0x01000000
+#define		LmHWTINT		0x00800000
+#define		LmMnCTXDONE		0x00100000
+#define		LmM2REQMBXF		0x00080000
+#define		LmM2RSPMBXE		0x00040000
+#define		LmMnDMAERR		0x00020000
+#define		LmRCVPRIM		0x00010000
+#define		LmRCVERR		0x00008000
+#define		LmADDRRCV		0x00004000
+#define		LmMnHDRMISS		0x00002000
+#define		LmMnWAITSCB		0x00001000
+#define		LmMnRLSSCB		0x00000800
+#define		LmMnSAVECTX		0x00000400
+#define		LmMnFETCHSG		0x00000200
+#define		LmMnLOADCTX		0x00000100
+#define		LmMnCFGICL		0x00000080
+#define		LmMnCFGSATA		0x00000040
+#define		LmMnCFGEXPSATA		0x00000020
+#define		LmMnCFGCMPLT		0x00000010
+#define		LmMnCFGRBUF		0x00000008
+#define		LmMnSAVETTR		0x00000004
+#define		LmMnCFGRDAT		0x00000002
+#define		LmMnCFGHDR		0x00000001
+
+#define LmMnINTEN(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0x3C)
+
+#define		EN_LmACKREQ		0x08000000
+#define		EN_LmNAKREQ		0x04000000
+#define		EN_LmMnXMTERR		0x02000000
+#define		EN_LmM5OOBSVC		0x01000000
+#define		EN_LmHWTINT		0x00800000
+#define		EN_LmMnCTXDONE		0x00100000
+#define		EN_LmM2REQMBXF		0x00080000
+#define		EN_LmM2RSPMBXE		0x00040000
+#define		EN_LmMnDMAERR		0x00020000
+#define		EN_LmRCVPRIM		0x00010000
+#define		EN_LmRCVERR		0x00008000
+#define		EN_LmADDRRCV		0x00004000
+#define		EN_LmMnHDRMISS		0x00002000
+#define		EN_LmMnWAITSCB		0x00001000
+#define		EN_LmMnRLSSCB		0x00000800
+#define		EN_LmMnSAVECTX		0x00000400
+#define		EN_LmMnFETCHSG		0x00000200
+#define		EN_LmMnLOADCTX		0x00000100
+#define		EN_LmMnCFGICL		0x00000080
+#define		EN_LmMnCFGSATA		0x00000040
+#define		EN_LmMnCFGEXPSATA	0x00000020
+#define		EN_LmMnCFGCMPLT		0x00000010
+#define		EN_LmMnCFGRBUF		0x00000008
+#define		EN_LmMnSAVETTR		0x00000004
+#define		EN_LmMnCFGRDAT		0x00000002
+#define		EN_LmMnCFGHDR		0x00000001
+
+#define		LmM0INTEN_MASK		(EN_LmMnCFGCMPLT | EN_LmMnCFGRBUF | \
+					 EN_LmMnSAVETTR | EN_LmMnCFGRDAT | \
+					 EN_LmMnCFGHDR | EN_LmRCVERR | \
+					 EN_LmADDRRCV | EN_LmMnHDRMISS | \
+					 EN_LmMnRLSSCB | EN_LmMnSAVECTX | \
+					 EN_LmMnFETCHSG | EN_LmMnLOADCTX | \
+					 EN_LmHWTINT | EN_LmMnCTXDONE | \
+					 EN_LmRCVPRIM | EN_LmMnCFGSATA | \
+					 EN_LmMnCFGEXPSATA | EN_LmMnDMAERR)
+
+#define		LmM1INTEN_MASK		(EN_LmMnCFGCMPLT | EN_LmADDRRCV | \
+					 EN_LmMnRLSSCB | EN_LmMnSAVECTX | \
+					 EN_LmMnFETCHSG | EN_LmMnLOADCTX | \
+					 EN_LmMnXMTERR | EN_LmHWTINT | \
+					 EN_LmMnCTXDONE | EN_LmRCVPRIM | \
+					 EN_LmRCVERR | EN_LmMnDMAERR)
+
+#define		LmM2INTEN_MASK		(EN_LmADDRRCV | EN_LmHWTINT | \
+					 EN_LmM2REQMBXF | EN_LmRCVPRIM | \
+					 EN_LmRCVERR)
+
+#define		LmM5INTEN_MASK		(EN_LmADDRRCV | EN_LmM5OOBSVC | \
+					 EN_LmHWTINT | EN_LmRCVPRIM | \
+					 EN_LmRCVERR)
+
+#define LmXMTPRIMD(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0x40)
+
+#define LmXMTPRIMCS(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0x44)
+
+#define LmCONSTAT(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0x45)
+
+#define LmMnDMAERRS(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0x46)
+
+#define LmMnSGDMAERRS(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0x47)
+
+#define LmM0EXPHDRP(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0x48)
+
+#define LmM1SASALIGN(LinkNum)		LmSEQ_PHY_REG(1, LinkNum, 0x48)
+#define SAS_ALIGN_DEFAULT		0xFF
+
+#define LmM0MSKHDRP(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0x49)
+
+#define LmM1STPALIGN(LinkNum)		LmSEQ_PHY_REG(1, LinkNum, 0x49)
+#define STP_ALIGN_DEFAULT		0x1F
+
+#define LmM0RCVHDRP(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0x4A)
+
+#define LmM1XMTHDRP(LinkNum)		LmSEQ_PHY_REG(1, LinkNum, 0x4A)
+
+#define LmM0ICLADR(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0x4B)
+
+#define LmM1ALIGNMODE(LinkNum)		LmSEQ_PHY_REG(1, LinkNum, 0x4B)
+
+#define		LmDISALIGN		0x20
+#define		LmROTSTPALIGN		0x10
+#define		LmSTPALIGN		0x08
+#define		LmROTNOTIFY		0x04
+#define		LmDUALALIGN		0x02
+#define		LmROTALIGN		0x01
+
+#define LmM0EXPRCVNT(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0x4C)
+
+#define LmM1XMTCNT(LinkNum)		LmSEQ_PHY_REG(1, LinkNum, 0x4C)
+
+#define LmMnBUFSTAT(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0x4E)
+
+#define		LmMnBUFPERR		0x01
+
+/* mode 0-1 */
+#define LmMnXFRLVL(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0x59)
+
+#define		LmMnXFRLVL_128		0x05
+#define		LmMnXFRLVL_256		0x04
+#define		LmMnXFRLVL_512		0x03
+#define		LmMnXFRLVL_1024		0x02
+#define		LmMnXFRLVL_1536		0x01
+#define		LmMnXFRLVL_2048		0x00
+
+ /* mode 0-1 */
+#define LmMnSGDMACTL(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0x5A)
+
+#define 	LmMnRESETSG		0x04
+#define 	LmMnSTOPSG		0x02
+#define 	LmMnSTARTSG		0x01
+
+/* mode 0-1 */
+#define LmMnSGDMASTAT(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0x5B)
+
+/* mode 0-1 */
+#define LmMnDDMACTL(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0x5C)
+
+#define 	LmMnFLUSH		0x40		/* wo */
+#define 	LmMnRLSRTRY		0x20		/* wo */
+#define 	LmMnDISCARD		0x10		/* wo */
+#define 	LmMnRESETDAT		0x08		/* wo */
+#define 	LmMnSUSDAT		0x04		/* wo */
+#define 	LmMnSTOPDAT		0x02		/* wo */
+#define 	LmMnSTARTDAT		0x01		/* wo */
+
+/* mode 0-1 */
+#define LmMnDDMASTAT(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0x5D)
+
+#define		LmMnDPEMPTY		0x80
+#define		LmMnFLUSHING		0x40
+#define		LmMnDDMAREQ		0x20
+#define		LmMnHDMAREQ		0x10
+#define		LmMnDATFREE		0x08
+#define		LmMnDATSUS		0x04
+#define		LmMnDATACT		0x02
+#define		LmMnDATEN		0x01
+
+/* mode 0-1 */
+#define LmMnDDMAMODE(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0x5E)
+
+#define 	LmMnDMATYPE_NORMAL		0x0000
+#define 	LmMnDMATYPE_HOST_ONLY_TX	0x0001
+#define 	LmMnDMATYPE_DEVICE_ONLY_TX	0x0002
+#define 	LmMnDMATYPE_INVALID		0x0003
+#define 	LmMnDMATYPE_MASK	0x0003
+
+#define 	LmMnDMAWRAP		0x0004
+#define 	LmMnBITBUCKET		0x0008
+#define 	LmMnDISHDR		0x0010
+#define 	LmMnSTPCRC		0x0020
+#define 	LmXTEST			0x0040
+#define 	LmMnDISCRC		0x0080
+#define 	LmMnENINTLK		0x0100
+#define 	LmMnADDRFRM		0x0400
+#define 	LmMnENXMTCRC		0x0800
+
+/* mode 0-1 */
+#define LmMnXFRCNT(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0x70)
+
+/* mode 0-1 */
+#define LmMnDPSEL(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0x7B)
+#define 	LmMnDPSEL_MASK		0x07
+#define 	LmMnEOLPRE		0x40
+#define 	LmMnEOSPRE		0x80
+
+/* Registers used in conjunction with LmMnDPSEL and LmMnDPACC registers */
+/* Receive Mode n = 0 */
+#define LmMnHRADDR			0x00
+#define LmMnHBYTECNT			0x01
+#define LmMnHREWIND			0x02
+#define LmMnDWADDR			0x03
+#define LmMnDSPACECNT			0x04
+#define LmMnDFRMSIZE			0x05
+
+/* Registers used in conjunction with LmMnDPSEL and LmMnDPACC registers */
+/* Transmit Mode n = 1 */
+#define LmMnHWADDR			0x00
+#define LmMnHSPACECNT			0x01
+/* #define LmMnHREWIND			0x02 */
+#define LmMnDRADDR			0x03
+#define LmMnDBYTECNT			0x04
+/* #define LmMnDFRMSIZE			0x05 */
+
+/* mode 0-1 */
+#define LmMnDPACC(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0x78)
+#define 	LmMnDPACC_MASK		0x00FFFFFF
+
+/* mode 0-1 */
+#define LmMnHOLDLVL(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0x7D)
+
+#define LmPRMSTAT0(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0x80)
+#define LmPRMSTAT0BYTE0			0x80
+#define LmPRMSTAT0BYTE1			0x81
+#define LmPRMSTAT0BYTE2			0x82
+#define LmPRMSTAT0BYTE3			0x83
+
+#define		LmFRAMERCVD		0x80000000
+#define		LmXFRRDYRCVD		0x40000000
+#define		LmUNKNOWNP		0x20000000
+#define		LmBREAK			0x10000000
+#define		LmDONE			0x08000000
+#define		LmOPENACPT		0x04000000
+#define		LmOPENRJCT		0x02000000
+#define		LmOPENRTRY		0x01000000
+#define		LmCLOSERV1		0x00800000
+#define		LmCLOSERV0		0x00400000
+#define		LmCLOSENORM		0x00200000
+#define		LmCLOSECLAF		0x00100000
+#define		LmNOTIFYRV2		0x00080000
+#define		LmNOTIFYRV1		0x00040000
+#define		LmNOTIFYRV0		0x00020000
+#define		LmNOTIFYSPIN		0x00010000
+#define		LmBROADRV4		0x00008000
+#define		LmBROADRV3		0x00004000
+#define		LmBROADRV2		0x00002000
+#define		LmBROADRV1		0x00001000
+#define		LmBROADSES		0x00000800
+#define		LmBROADRVCH1		0x00000400
+#define		LmBROADRVCH0		0x00000200
+#define		LmBROADCH		0x00000100
+#define		LmAIPRVWP		0x00000080
+#define		LmAIPWP			0x00000040
+#define		LmAIPWD			0x00000020
+#define		LmAIPWC			0x00000010
+#define		LmAIPRV2		0x00000008
+#define		LmAIPRV1		0x00000004
+#define		LmAIPRV0		0x00000002
+#define		LmAIPNRML		0x00000001
+
+#define		LmBROADCAST_MASK	(LmBROADCH | LmBROADRVCH0 | \
+					 LmBROADRVCH1)
+
+#define LmPRMSTAT1(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0x84)
+#define LmPRMSTAT1BYTE0			0x84
+#define LmPRMSTAT1BYTE1			0x85
+#define LmPRMSTAT1BYTE2			0x86
+#define LmPRMSTAT1BYTE3			0x87
+
+#define		LmFRMRCVDSTAT		0x80000000
+#define		LmBREAK_DET		0x04000000
+#define		LmCLOSE_DET		0x02000000
+#define		LmDONE_DET		0x01000000
+#define		LmXRDY			0x00040000
+#define 	LmSYNCSRST		0x00020000
+#define 	LmSYNC			0x00010000
+#define 	LmXHOLD			0x00008000
+#define 	LmRRDY			0x00004000
+#define 	LmHOLD			0x00002000
+#define 	LmROK			0x00001000
+#define 	LmRIP			0x00000800
+#define 	LmCRBLK			0x00000400
+#define 	LmACK			0x00000200
+#define 	LmNAK			0x00000100
+#define 	LmHARDRST		0x00000080
+#define 	LmERROR			0x00000040
+#define 	LmRERR			0x00000020
+#define 	LmPMREQP		0x00000010
+#define 	LmPMREQS		0x00000008
+#define 	LmPMACK			0x00000004
+#define 	LmPMNAK			0x00000002
+#define 	LmDMAT			0x00000001
+
+/* mode 1 */
+#define	LmMnSATAFS(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0x7E)
+#define	LmMnXMTSIZE(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0x93)
+
+/* mode 0 */
+#define LmMnFRMERR(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0xB0)
+
+#define		LmACRCERR		0x00000800
+#define		LmPHYOVRN		0x00000400
+#define		LmOBOVRN		0x00000200
+#define 	LmMnZERODATA		0x00000100
+#define		LmSATAINTLK		0x00000080
+#define		LmMnCRCERR		0x00000020
+#define		LmRRDYOVRN		0x00000010
+#define		LmMISSSOAF		0x00000008
+#define		LmMISSSOF		0x00000004
+#define		LmMISSEOAF		0x00000002
+#define		LmMISSEOF		0x00000001
+
+#define LmFRMERREN(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0xB4)
+
+#define 	EN_LmACRCERR		0x00000800
+#define 	EN_LmPHYOVRN		0x00000400
+#define 	EN_LmOBOVRN		0x00000200
+#define 	EN_LmMnZERODATA		0x00000100
+#define 	EN_LmSATAINTLK		0x00000080
+#define 	EN_LmFRMBAD		0x00000040
+#define 	EN_LmMnCRCERR		0x00000020
+#define 	EN_LmRRDYOVRN		0x00000010
+#define 	EN_LmMISSSOAF		0x00000008
+#define 	EN_LmMISSSOF		0x00000004
+#define 	EN_LmMISSEOAF		0x00000002
+#define 	EN_LmMISSEOF		0x00000001
+
+#define 	LmFRMERREN_MASK  	(EN_LmSATAINTLK | EN_LmMnCRCERR | \
+					 EN_LmRRDYOVRN | EN_LmMISSSOF | \
+					 EN_LmMISSEOAF | EN_LmMISSEOF | \
+					 EN_LmACRCERR | LmPHYOVRN | \
+					 EN_LmOBOVRN | EN_LmMnZERODATA)
+
+#define LmHWTSTATEN(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0xC5)
+
+#define		EN_LmDONETO		0x80
+#define		EN_LmINVDISP		0x40
+#define		EN_LmINVDW		0x20
+#define		EN_LmDWSEVENT		0x08
+#define		EN_LmCRTTTO		0x04
+#define		EN_LmANTTTO		0x02
+#define		EN_LmBITLTTO		0x01
+
+#define		LmHWTSTATEN_MASK	(EN_LmINVDISP | EN_LmINVDW | \
+					 EN_LmDWSEVENT | EN_LmCRTTTO | \
+					 EN_LmANTTTO | EN_LmDONETO | \
+					 EN_LmBITLTTO)
+
+#define LmHWTSTAT(LinkNum) 		LmSEQ_PHY_REG(0, LinkNum, 0xC7)
+
+#define		LmDONETO		0x80
+#define		LmINVDISP		0x40
+#define		LmINVDW			0x20
+#define		LmDWSEVENT		0x08
+#define		LmCRTTTO		0x04
+#define		LmANTTTO		0x02
+#define		LmBITLTTO		0x01
+
+#define LmMnDATABUFADR(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0xC8)
+#define		LmDATABUFADR_MASK	0x0FFF
+
+#define LmMnDATABUF(LinkNum, Mode)	LmSEQ_PHY_REG(Mode, LinkNum, 0xCA)
+
+#define	LmPRIMSTAT0EN(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0xE0)
+
+#define 	EN_LmUNKNOWNP 		0x20000000
+#define 	EN_LmBREAK		0x10000000
+#define 	EN_LmDONE		0x08000000
+#define 	EN_LmOPENACPT		0x04000000
+#define 	EN_LmOPENRJCT		0x02000000
+#define 	EN_LmOPENRTRY		0x01000000
+#define 	EN_LmCLOSERV1		0x00800000
+#define 	EN_LmCLOSERV0		0x00400000
+#define 	EN_LmCLOSENORM		0x00200000
+#define 	EN_LmCLOSECLAF		0x00100000
+#define 	EN_LmNOTIFYRV2		0x00080000
+#define 	EN_LmNOTIFYRV1		0x00040000
+#define 	EN_LmNOTIFYRV0		0x00020000
+#define 	EN_LmNOTIFYSPIN		0x00010000
+#define 	EN_LmBROADRV4		0x00008000
+#define 	EN_LmBROADRV3		0x00004000
+#define 	EN_LmBROADRV2		0x00002000
+#define 	EN_LmBROADRV1		0x00001000
+#define 	EN_LmBROADRV0		0x00000800
+#define 	EN_LmBROADRVCH1		0x00000400
+#define 	EN_LmBROADRVCH0		0x00000200
+#define 	EN_LmBROADCH		0x00000100
+#define 	EN_LmAIPRVWP		0x00000080
+#define 	EN_LmAIPWP		0x00000040
+#define 	EN_LmAIPWD		0x00000020
+#define 	EN_LmAIPWC		0x00000010
+#define 	EN_LmAIPRV2		0x00000008
+#define 	EN_LmAIPRV1		0x00000004
+#define 	EN_LmAIPRV0		0x00000002
+#define 	EN_LmAIPNRML		0x00000001
+
+#define		LmPRIMSTAT0EN_MASK	(EN_LmBREAK | \
+					 EN_LmDONE | EN_LmOPENACPT | \
+					 EN_LmOPENRJCT | EN_LmOPENRTRY | \
+					 EN_LmCLOSERV1 | EN_LmCLOSERV0 | \
+					 EN_LmCLOSENORM | EN_LmCLOSECLAF | \
+					 EN_LmBROADRV4 | EN_LmBROADRV3 | \
+					 EN_LmBROADRV2 | EN_LmBROADRV1 | \
+					 EN_LmBROADRV0 | EN_LmBROADRVCH1 | \
+					 EN_LmBROADRVCH0 | EN_LmBROADCH | \
+					 EN_LmAIPRVWP | EN_LmAIPWP | \
+					 EN_LmAIPWD | EN_LmAIPWC | \
+					 EN_LmAIPRV2 | EN_LmAIPRV1 | \
+					 EN_LmAIPRV0 | EN_LmAIPNRML)
+
+#define LmPRIMSTAT1EN(LinkNum)		LmSEQ_PHY_REG(0, LinkNum, 0xE4)
+
+#define		EN_LmXRDY		0x00040000
+#define		EN_LmSYNCSRST		0x00020000
+#define		EN_LmSYNC		0x00010000
+#define 	EN_LmXHOLD		0x00008000
+#define 	EN_LmRRDY		0x00004000
+#define 	EN_LmHOLD		0x00002000
+#define 	EN_LmROK		0x00001000
+#define 	EN_LmRIP		0x00000800
+#define 	EN_LmCRBLK		0x00000400
+#define 	EN_LmACK		0x00000200
+#define 	EN_LmNAK		0x00000100
+#define 	EN_LmHARDRST		0x00000080
+#define 	EN_LmERROR		0x00000040
+#define 	EN_LmRERR		0x00000020
+#define 	EN_LmPMREQP		0x00000010
+#define 	EN_LmPMREQS		0x00000008
+#define 	EN_LmPMACK		0x00000004
+#define 	EN_LmPMNAK		0x00000002
+#define 	EN_LmDMAT		0x00000001
+
+#define LmPRIMSTAT1EN_MASK		(EN_LmHARDRST | \
+					 EN_LmSYNCSRST | \
+					 EN_LmPMREQP | EN_LmPMREQS | \
+					 EN_LmPMACK | EN_LmPMNAK)
+
+#define LmSMSTATE(LinkNum) 		LmSEQ_PHY_REG(0, LinkNum, 0xE8)
+
+#define LmSMSTATEBRK(LinkNum) 		LmSEQ_PHY_REG(0, LinkNum, 0xEC)
+
+#define LmSMDBGCTL(LinkNum) 		LmSEQ_PHY_REG(0, LinkNum, 0xF0)
+
+
+/*
+ * LmSEQ CIO Bus Mode 3 Register.
+ * Mode 3: Configuration and Setup, IOP Context SCB.
+ */
+#define LmM3SATATIMER(LinkNum) 		LmSEQ_PHY_REG(3, LinkNum, 0x48)
+
+#define LmM3INTVEC0(LinkNum)		LmSEQ_PHY_REG(3, LinkNum, 0x90)
+
+#define LmM3INTVEC1(LinkNum)		LmSEQ_PHY_REG(3, LinkNum, 0x92)
+
+#define LmM3INTVEC2(LinkNum)		LmSEQ_PHY_REG(3, LinkNum, 0x94)
+
+#define LmM3INTVEC3(LinkNum)		LmSEQ_PHY_REG(3, LinkNum, 0x96)
+
+#define LmM3INTVEC4(LinkNum)		LmSEQ_PHY_REG(3, LinkNum, 0x98)
+
+#define LmM3INTVEC5(LinkNum)		LmSEQ_PHY_REG(3, LinkNum, 0x9A)
+
+#define LmM3INTVEC6(LinkNum)		LmSEQ_PHY_REG(3, LinkNum, 0x9C)
+
+#define LmM3INTVEC7(LinkNum)		LmSEQ_PHY_REG(3, LinkNum, 0x9E)
+
+#define LmM3INTVEC8(LinkNum)		LmSEQ_PHY_REG(3, LinkNum, 0xA4)
+
+#define LmM3INTVEC9(LinkNum)		LmSEQ_PHY_REG(3, LinkNum, 0xA6)
+
+#define LmM3INTVEC10(LinkNum)		LmSEQ_PHY_REG(3, LinkNum, 0xB0)
+
+#define LmM3FRMGAP(LinkNum)		LmSEQ_PHY_REG(3, LinkNum, 0xB4)
+
+#define LmBITL_TIMER(LinkNum) 		LmSEQ_PHY_REG(0, LinkNum, 0xA2)
+
+#define LmWWN(LinkNum) 			LmSEQ_PHY_REG(0, LinkNum, 0xA8)
+
+
+/*
+ * LmSEQ CIO Bus Mode 5 Registers.
+ * Mode 5: Phy/OOB Control and Status.
+ */
+#define LmSEQ_OOB_REG(phy_id, reg)	LmSEQ_PHY_REG(5, (phy_id), (reg))
+
+#define OOB_BFLTR	0x100
+
+#define		BFLTR_THR_MASK		0xF0
+#define		BFLTR_TC_MASK		0x0F
+
+#define OOB_INIT_MIN	0x102
+
+#define OOB_INIT_MAX	0x104
+
+#define OOB_INIT_NEG	0x106
+
+#define	OOB_SAS_MIN	0x108
+
+#define OOB_SAS_MAX	0x10A
+
+#define OOB_SAS_NEG	0x10C
+
+#define OOB_WAKE_MIN	0x10E
+
+#define OOB_WAKE_MAX	0x110
+
+#define OOB_WAKE_NEG	0x112
+
+#define OOB_IDLE_MAX	0x114
+
+#define OOB_BURST_MAX	0x116
+
+#define OOB_DATA_KBITS	0x126
+
+#define OOB_ALIGN_0_DATA	0x12C
+
+#define OOB_ALIGN_1_DATA	0x130
+
+#define D10_2_DATA_k		0x00
+#define SYNC_DATA_k		0x02
+#define ALIGN_1_DATA_k		0x04
+#define ALIGN_0_DATA_k		0x08
+#define BURST_DATA_k		0x10
+
+#define OOB_PHY_RESET_COUNT	0x13C
+
+#define OOB_SIG_GEN	0x140
+
+#define		START_OOB		0x80
+#define		START_DWS		0x40
+#define		ALIGN_CNT3		0x30
+#define 	ALIGN_CNT2		0x20
+#define 	ALIGN_CNT1		0x10
+#define 	ALIGN_CNT4		0x00
+#define		STOP_DWS		0x08
+#define		SEND_COMSAS		0x04
+#define		SEND_COMINIT		0x02
+#define		SEND_COMWAKE		0x01
+
+#define OOB_XMIT	0x141
+
+#define		TX_ENABLE		0x80
+#define		XMIT_OOB_BURST		0x10
+#define		XMIT_D10_2		0x08
+#define		XMIT_SYNC		0x04
+#define		XMIT_ALIGN_1		0x02
+#define		XMIT_ALIGN_0		0x01
+
+#define FUNCTION_MASK	0x142
+
+#define		SAS_MODE_DIS		0x80
+#define		SATA_MODE_DIS		0x40
+#define		SPINUP_HOLD_DIS		0x20
+#define		HOT_PLUG_DIS		0x10
+#define		SATA_PS_DIS		0x08
+#define		FUNCTION_MASK_DEFAULT	(SPINUP_HOLD_DIS | SATA_PS_DIS)
+
+#define OOB_MODE	0x143
+
+#define		SAS_MODE		0x80
+#define		SATA_MODE		0x40
+#define		SLOW_CLK		0x20
+#define		FORCE_XMIT_15		0x08
+#define		PHY_SPEED_60		0x04
+#define		PHY_SPEED_30		0x02
+#define		PHY_SPEED_15		0x01
+
+#define	CURRENT_STATUS	0x144
+
+#define		CURRENT_OOB_DONE	0x80
+#define		CURRENT_LOSS_OF_SIGNAL	0x40
+#define		CURRENT_SPINUP_HOLD	0x20
+#define		CURRENT_HOT_PLUG_CNCT	0x10
+#define		CURRENT_GTO_TIMEOUT	0x08
+#define		CURRENT_OOB_TIMEOUT	0x04
+#define		CURRENT_DEVICE_PRESENT	0x02
+#define		CURRENT_OOB_ERROR	0x01
+
+#define 	CURRENT_OOB1_ERROR	(CURRENT_HOT_PLUG_CNCT | \
+					 CURRENT_GTO_TIMEOUT)
+
+#define 	CURRENT_OOB2_ERROR	(CURRENT_HOT_PLUG_CNCT | \
+					 CURRENT_OOB_ERROR)
+
+#define		DEVICE_ADDED_W_CNT	(CURRENT_OOB_DONE | \
+					 CURRENT_HOT_PLUG_CNCT | \
+					 CURRENT_DEVICE_PRESENT)
+
+#define		DEVICE_ADDED_WO_CNT	(CURRENT_OOB_DONE | \
+					 CURRENT_DEVICE_PRESENT)
+
+#define 	DEVICE_REMOVED		CURRENT_LOSS_OF_SIGNAL
+
+#define		CURRENT_PHY_MASK	(CURRENT_OOB_DONE | \
+					 CURRENT_LOSS_OF_SIGNAL | \
+					 CURRENT_SPINUP_HOLD | \
+					 CURRENT_HOT_PLUG_CNCT | \
+					 CURRENT_GTO_TIMEOUT | \
+					 CURRENT_DEVICE_PRESENT | \
+					 CURRENT_OOB_ERROR )
+
+#define		CURRENT_ERR_MASK	(CURRENT_LOSS_OF_SIGNAL | \
+					 CURRENT_GTO_TIMEOUT | \
+					 CURRENT_OOB_TIMEOUT | \
+					 CURRENT_OOB_ERROR )
+
+#define SPEED_MASK	0x145
+
+#define		SATA_SPEED_30_DIS	0x10
+#define		SATA_SPEED_15_DIS	0x08
+#define		SAS_SPEED_60_DIS	0x04
+#define		SAS_SPEED_30_DIS	0x02
+#define		SAS_SPEED_15_DIS	0x01
+#define		SAS_SPEED_MASK_DEFAULT	0x00
+
+#define OOB_TIMER_ENABLE	0x14D
+
+#define		HOT_PLUG_EN		0x80
+#define		RCD_EN			0x40
+#define 	COMTIMER_EN		0x20
+#define		SNTT_EN			0x10
+#define		SNLT_EN			0x04
+#define		SNWT_EN			0x02
+#define		ALIGN_EN		0x01
+
+#define OOB_STATUS		0x14E
+
+#define		OOB_DONE		0x80
+#define		LOSS_OF_SIGNAL		0x40		/* ro */
+#define		SPINUP_HOLD		0x20
+#define		HOT_PLUG_CNCT		0x10		/* ro */
+#define		GTO_TIMEOUT		0x08		/* ro */
+#define		OOB_TIMEOUT		0x04		/* ro */
+#define		DEVICE_PRESENT		0x02		/* ro */
+#define		OOB_ERROR		0x01		/* ro */
+
+#define		OOB_STATUS_ERROR_MASK	(LOSS_OF_SIGNAL | GTO_TIMEOUT | \
+					 OOB_TIMEOUT | OOB_ERROR)
+
+#define OOB_STATUS_CLEAR	0x14F
+
+#define		OOB_DONE_CLR		0x80
+#define		LOSS_OF_SIGNAL_CLR 	0x40
+#define		SPINUP_HOLD_CLR		0x20
+#define		HOT_PLUG_CNCT_CLR     	0x10
+#define		GTO_TIMEOUT_CLR		0x08
+#define		OOB_TIMEOUT_CLR		0x04
+#define		OOB_ERROR_CLR		0x01
+
+#define HOT_PLUG_DELAY		0x150
+/* In 5 ms units. 20 = 100 ms. */
+#define	HOTPLUG_DELAY_TIMEOUT		20
+
+
+#define INT_ENABLE_2		0x15A
+
+#define		OOB_DONE_EN		0x80
+#define		LOSS_OF_SIGNAL_EN	0x40
+#define		SPINUP_HOLD_EN		0x20
+#define		HOT_PLUG_CNCT_EN	0x10
+#define		GTO_TIMEOUT_EN		0x08
+#define		OOB_TIMEOUT_EN		0x04
+#define		DEVICE_PRESENT_EN	0x02
+#define		OOB_ERROR_EN		0x01
+
+#define PHY_CONTROL_0		0x160
+
+#define		PHY_LOWPWREN_TX		0x80
+#define		PHY_LOWPWREN_RX		0x40
+#define		SPARE_REG_160_B5	0x20
+#define		OFFSET_CANCEL_RX	0x10
+
+/* bits 3:2 */
+#define		PHY_RXCOMCENTER_60V	0x00
+#define		PHY_RXCOMCENTER_70V	0x04
+#define		PHY_RXCOMCENTER_80V	0x08
+#define		PHY_RXCOMCENTER_90V	0x0C
+#define 	PHY_RXCOMCENTER_MASK	0x0C
+
+#define		PHY_RESET		0x02
+#define		SAS_DEFAULT_SEL		0x01
+
+#define PHY_CONTROL_1		0x161
+
+/* bits 2:0 */
+#define		SATA_PHY_DETLEVEL_50mv	0x00
+#define		SATA_PHY_DETLEVEL_75mv	0x01
+#define		SATA_PHY_DETLEVEL_100mv	0x02
+#define		SATA_PHY_DETLEVEL_125mv	0x03
+#define		SATA_PHY_DETLEVEL_150mv	0x04
+#define		SATA_PHY_DETLEVEL_175mv	0x05
+#define		SATA_PHY_DETLEVEL_200mv	0x06
+#define		SATA_PHY_DETLEVEL_225mv	0x07
+#define		SATA_PHY_DETLEVEL_MASK	0x07
+
+/* bits 5:3 */
+#define		SAS_PHY_DETLEVEL_50mv	0x00
+#define		SAS_PHY_DETLEVEL_75mv	0x08
+#define		SAS_PHY_DETLEVEL_100mv	0x10
+#define		SAS_PHY_DETLEVEL_125mv	0x11
+#define		SAS_PHY_DETLEVEL_150mv	0x20
+#define		SAS_PHY_DETLEVEL_175mv	0x21
+#define		SAS_PHY_DETLEVEL_200mv	0x30
+#define		SAS_PHY_DETLEVEL_225mv	0x31
+#define		SAS_PHY_DETLEVEL_MASK	0x38
+
+#define PHY_CONTROL_2		0x162
+
+/* bits 7:5 */
+#define 	SATA_PHY_DRV_400mv	0x00
+#define 	SATA_PHY_DRV_450mv	0x20
+#define 	SATA_PHY_DRV_500mv	0x40
+#define 	SATA_PHY_DRV_550mv	0x60
+#define 	SATA_PHY_DRV_600mv	0x80
+#define 	SATA_PHY_DRV_650mv	0xA0
+#define 	SATA_PHY_DRV_725mv	0xC0
+#define 	SATA_PHY_DRV_800mv	0xE0
+#define		SATA_PHY_DRV_MASK	0xE0
+
+/* bits 4:3 */
+#define 	SATA_PREEMP_0		0x00
+#define 	SATA_PREEMP_1		0x08
+#define 	SATA_PREEMP_2		0x10
+#define 	SATA_PREEMP_3		0x18
+#define 	SATA_PREEMP_MASK	0x18
+
+#define 	SATA_CMSH1P5		0x04
+
+/* bits 1:0 */
+#define 	SATA_SLEW_0		0x00
+#define 	SATA_SLEW_1		0x01
+#define 	SATA_SLEW_2		0x02
+#define 	SATA_SLEW_3		0x03
+#define 	SATA_SLEW_MASK		0x03
+
+#define PHY_CONTROL_3		0x163
+
+/* bits 7:5 */
+#define 	SAS_PHY_DRV_400mv	0x00
+#define 	SAS_PHY_DRV_450mv	0x20
+#define 	SAS_PHY_DRV_500mv	0x40
+#define 	SAS_PHY_DRV_550mv	0x60
+#define 	SAS_PHY_DRV_600mv	0x80
+#define 	SAS_PHY_DRV_650mv	0xA0
+#define 	SAS_PHY_DRV_725mv	0xC0
+#define 	SAS_PHY_DRV_800mv	0xE0
+#define		SAS_PHY_DRV_MASK	0xE0
+
+/* bits 4:3 */
+#define 	SAS_PREEMP_0		0x00
+#define 	SAS_PREEMP_1		0x08
+#define 	SAS_PREEMP_2		0x10
+#define 	SAS_PREEMP_3		0x18
+#define 	SAS_PREEMP_MASK		0x18
+
+#define 	SAS_CMSH1P5		0x04
+
+/* bits 1:0 */
+#define 	SAS_SLEW_0		0x00
+#define 	SAS_SLEW_1		0x01
+#define 	SAS_SLEW_2		0x02
+#define 	SAS_SLEW_3		0x03
+#define 	SAS_SLEW_MASK		0x03
+
+#define PHY_CONTROL_4		0x168
+
+#define		PHY_DONE_CAL_TX		0x80
+#define		PHY_DONE_CAL_RX		0x40
+#define		RX_TERM_LOAD_DIS	0x20
+#define		TX_TERM_LOAD_DIS	0x10
+#define		AUTO_TERM_CAL_DIS	0x08
+#define		PHY_SIGDET_FLTR_EN	0x04
+#define		OSC_FREQ		0x02
+#define		PHY_START_CAL		0x01
+
+/*
+ * HST_PCIX2 Registers, Addresss Range: (0x00-0xFC)
+ */
+#define PCIX_REG_BASE_ADR		0xB8040000
+
+#define PCIC_VENDOR_ID	0x00
+
+#define PCIC_DEVICE_ID	0x02
+
+#define PCIC_COMMAND	0x04
+
+#define		INT_DIS			0x0400
+#define		FBB_EN			0x0200		/* ro */
+#define		SERR_EN			0x0100
+#define		STEP_EN			0x0080		/* ro */
+#define		PERR_EN			0x0040
+#define		VGA_EN			0x0020		/* ro */
+#define		MWI_EN			0x0010
+#define		SPC_EN			0x0008
+#define		MST_EN			0x0004
+#define		MEM_EN			0x0002
+#define		IO_EN			0x0001
+
+#define	PCIC_STATUS	0x06
+
+#define		PERR_DET		0x8000
+#define		SERR_GEN		0x4000
+#define		MABT_DET		0x2000
+#define		TABT_DET		0x1000
+#define		TABT_GEN		0x0800
+#define		DPERR_DET		0x0100
+#define		CAP_LIST		0x0010
+#define		INT_STAT		0x0008
+
+#define	PCIC_DEVREV_ID	0x08
+
+#define	PCIC_CLASS_CODE	0x09
+
+#define	PCIC_CACHELINE_SIZE	0x0C
+
+#define	PCIC_MBAR0	0x10
+
+#define 	PCIC_MBAR0_OFFSET	0
+
+#define	PCIC_MBAR1	0x18
+
+#define 	PCIC_MBAR1_OFFSET	2
+
+#define	PCIC_IOBAR	0x20
+
+#define 	PCIC_IOBAR_OFFSET	4
+
+#define	PCIC_SUBVENDOR_ID	0x2C
+
+#define PCIC_SUBSYTEM_ID	0x2E
+
+#define PCIX_STATUS		0x44
+#define 	RCV_SCE		0x20000000
+#define 	UNEXP_SC	0x00080000
+#define 	SC_DISCARD	0x00040000
+
+#define ECC_CTRL_STAT		0x48
+#define 	UNCOR_ECCERR	0x00000008
+
+#define PCIC_PM_CSR		0x5C
+
+#define		PWR_STATE_D0		0
+#define		PWR_STATE_D1		1	/* not supported */
+#define		PWR_STATE_D2		2 	/* not supported */
+#define		PWR_STATE_D3		3
+
+#define PCIC_BASE1	0x6C	/* internal use only */
+
+#define		BASE1_RSVD		0xFFFFFFF8
+
+#define PCIC_BASEA	0x70	/* internal use only */
+
+#define		BASEA_RSVD		0xFFFFFFC0
+#define 	BASEA_START		0
+
+#define PCIC_BASEB	0x74	/* internal use only */
+
+#define		BASEB_RSVD		0xFFFFFF80
+#define		BASEB_IOMAP_MASK	0x7F
+#define 	BASEB_START		0x80
+
+#define PCIC_BASEC	0x78	/* internal use only */
+
+#define		BASEC_RSVD		0xFFFFFFFC
+#define 	BASEC_MASK		0x03
+#define 	BASEC_START		0x58
+
+#define PCIC_MBAR_KEY	0x7C	/* internal use only */
+
+#define 	MBAR_KEY_MASK		0xFFFFFFFF
+
+#define PCIC_HSTPCIX_CNTRL	0xA0
+
+#define 	REWIND_DIS		0x0800
+#define		SC_TMR_DIS		0x04000000
+
+#define PCIC_MBAR0_MASK	0xA8
+#define		PCIC_MBAR0_SIZE_MASK 	0x1FFFE000
+#define		PCIC_MBAR0_SIZE_SHIFT 	13
+#define		PCIC_MBAR0_SIZE(val)	\
+		    (((val) & PCIC_MBAR0_SIZE_MASK) >> PCIC_MBAR0_SIZE_SHIFT)
+
+#define PCIC_FLASH_MBAR	0xB8
+
+#define PCIC_INTRPT_STAT 0xD4
+
+#define PCIC_TP_CTRL	0xFC
+
+/*
+ * EXSI Registers, Addresss Range: (0x00-0xFC)
+ */
+#define EXSI_REG_BASE_ADR		REG_BASE_ADDR_EXSI
+
+#define	EXSICNFGR	(EXSI_REG_BASE_ADR + 0x00)
+
+#define		OCMINITIALIZED		0x80000000
+#define		ASIEN			0x00400000
+#define		HCMODE			0x00200000
+#define		PCIDEF			0x00100000
+#define		COMSTOCK		0x00080000
+#define		SEEPROMEND		0x00040000
+#define		MSTTIMEN		0x00020000
+#define		XREGEX			0x00000200
+#define		NVRAMW			0x00000100
+#define		NVRAMEX			0x00000080
+#define		SRAMW			0x00000040
+#define		SRAMEX			0x00000020
+#define		FLASHW			0x00000010
+#define		FLASHEX			0x00000008
+#define		SEEPROMCFG		0x00000004
+#define		SEEPROMTYP		0x00000002
+#define		SEEPROMEX		0x00000001
+
+
+#define EXSICNTRLR	(EXSI_REG_BASE_ADR + 0x04)
+
+#define		MODINT_EN		0x00000001
+
+
+#define PMSTATR		(EXSI_REG_BASE_ADR + 0x10)
+
+#define		FLASHRST		0x00000002
+#define		FLASHRDY		0x00000001
+
+
+#define FLCNFGR		(EXSI_REG_BASE_ADR + 0x14)
+
+#define		FLWEH_MASK		0x30000000
+#define		FLWESU_MASK		0x0C000000
+#define		FLWEPW_MASK		0x03F00000
+#define		FLOEH_MASK		0x000C0000
+#define 	FLOESU_MASK		0x00030000
+#define 	FLOEPW_MASK		0x0000FC00
+#define 	FLCSH_MASK		0x00000300
+#define 	FLCSSU_MASK		0x000000C0
+#define 	FLCSPW_MASK		0x0000003F
+
+#define SRCNFGR		(EXSI_REG_BASE_ADR + 0x18)
+
+#define		SRWEH_MASK		0x30000000
+#define		SRWESU_MASK		0x0C000000
+#define		SRWEPW_MASK		0x03F00000
+
+#define		SROEH_MASK		0x000C0000
+#define 	SROESU_MASK		0x00030000
+#define 	SROEPW_MASK		0x0000FC00
+#define		SRCSH_MASK		0x00000300
+#define		SRCSSU_MASK		0x000000C0
+#define		SRCSPW_MASK		0x0000003F
+
+#define NVCNFGR		(EXSI_REG_BASE_ADR + 0x1C)
+
+#define 	NVWEH_MASK		0x30000000
+#define 	NVWESU_MASK		0x0C000000
+#define 	NVWEPW_MASK		0x03F00000
+#define 	NVOEH_MASK		0x000C0000
+#define 	NVOESU_MASK		0x00030000
+#define 	NVOEPW_MASK		0x0000FC00
+#define 	NVCSH_MASK		0x00000300
+#define 	NVCSSU_MASK		0x000000C0
+#define 	NVCSPW_MASK		0x0000003F
+
+#define XRCNFGR		(EXSI_REG_BASE_ADR + 0x20)
+
+#define 	XRWEH_MASK		0x30000000
+#define 	XRWESU_MASK		0x0C000000
+#define 	XRWEPW_MASK		0x03F00000
+#define 	XROEH_MASK		0x000C0000
+#define 	XROESU_MASK		0x00030000
+#define 	XROEPW_MASK		0x0000FC00
+#define 	XRCSH_MASK		0x00000300
+#define 	XRCSSU_MASK		0x000000C0
+#define		XRCSPW_MASK		0x0000003F
+
+#define XREGADDR	(EXSI_REG_BASE_ADR + 0x24)
+
+#define 	XRADDRINCEN		0x80000000
+#define 	XREGADD_MASK		0x007FFFFF
+
+
+#define XREGDATAR	(EXSI_REG_BASE_ADR + 0x28)
+
+#define		XREGDATA_MASK 		0x0000FFFF
+
+#define GPIOOER		(EXSI_REG_BASE_ADR + 0x40)
+
+#define GPIOODENR	(EXSI_REG_BASE_ADR + 0x44)
+
+#define GPIOINVR	(EXSI_REG_BASE_ADR + 0x48)
+
+#define GPIODATAOR	(EXSI_REG_BASE_ADR + 0x4C)
+
+#define GPIODATAIR	(EXSI_REG_BASE_ADR + 0x50)
+
+#define GPIOCNFGR	(EXSI_REG_BASE_ADR + 0x54)
+
+#define		GPIO_EXTSRC		0x00000001
+
+#define SCNTRLR		(EXSI_REG_BASE_ADR + 0xA0)
+
+#define 	SXFERDONE		0x00000100
+#define 	SXFERCNT_MASK		0x000000E0
+#define 	SCMDTYP_MASK		0x0000001C
+#define 	SXFERSTART		0x00000002
+#define 	SXFEREN			0x00000001
+
+#define	SRATER		(EXSI_REG_BASE_ADR + 0xA4)
+
+#define	SADDRR		(EXSI_REG_BASE_ADR + 0xA8)
+
+#define 	SADDR_MASK		0x0000FFFF
+
+#define SDATAOR		(EXSI_REG_BASE_ADR + 0xAC)
+
+#define	SDATAOR0	(EXSI_REG_BASE_ADR + 0xAC)
+#define SDATAOR1	(EXSI_REG_BASE_ADR + 0xAD)
+#define SDATAOR2	(EXSI_REG_BASE_ADR + 0xAE)
+#define SDATAOR3	(EXSI_REG_BASE_ADR + 0xAF)
+
+#define SDATAIR		(EXSI_REG_BASE_ADR + 0xB0)
+
+#define SDATAIR0	(EXSI_REG_BASE_ADR + 0xB0)
+#define SDATAIR1	(EXSI_REG_BASE_ADR + 0xB1)
+#define SDATAIR2	(EXSI_REG_BASE_ADR + 0xB2)
+#define SDATAIR3	(EXSI_REG_BASE_ADR + 0xB3)
+
+#define ASISTAT0R	(EXSI_REG_BASE_ADR + 0xD0)
+#define 	ASIFMTERR		0x00000400
+#define 	ASISEECHKERR		0x00000200
+#define 	ASIERR			0x00000100
+
+#define ASISTAT1R	(EXSI_REG_BASE_ADR + 0xD4)
+#define 	CHECKSUM_MASK		0x0000FFFF
+
+#define ASIERRADDR	(EXSI_REG_BASE_ADR + 0xD8)
+#define ASIERRDATAR	(EXSI_REG_BASE_ADR + 0xDC)
+#define ASIERRSTATR	(EXSI_REG_BASE_ADR + 0xE0)
+#define 	CPI2ASIBYTECNT_MASK	0x00070000
+#define 	CPI2ASIBYTEEN_MASK      0x0000F000
+#define 	CPI2ASITARGERR_MASK	0x00000F00
+#define 	CPI2ASITARGMID_MASK	0x000000F0
+#define 	CPI2ASIMSTERR_MASK	0x0000000F
+
+/*
+ * XSRAM, External SRAM (DWord and any BE pattern accessible)
+ */
+#define XSRAM_REG_BASE_ADDR             0xB8100000
+#define XSRAM_SIZE                        0x100000
+
+/*
+ * NVRAM Registers, Address Range: (0x00000 - 0x3FFFF).
+ */
+#define		NVRAM_REG_BASE_ADR	0xBF800000
+#define		NVRAM_MAX_BASE_ADR	0x003FFFFF
+
+/* OCM base address */
+#define		OCM_BASE_ADDR		0xA0000000
+#define		OCM_MAX_SIZE		0x20000
+
+/*
+ * Sequencers (Central and Link) Scratch RAM page definitions.
+ */
+
+/*
+ * The Central Management Sequencer (CSEQ) Scratch Memory is a 1024
+ * byte memory.  It is dword accessible and has byte parity
+ * protection. The CSEQ accesses it in 32 byte windows, either as mode
+ * dependent or mode independent memory. Each mode has 96 bytes,
+ * (three 32 byte pages 0-2, not contiguous), leaving 128 bytes of
+ * Mode Independent memory (four 32 byte pages 3-7). Note that mode
+ * dependent scratch memory, Mode 8, page 0-3 overlaps mode
+ * independent scratch memory, pages 0-3.
+ * - 896 bytes of mode dependent scratch, 96 bytes per Modes 0-7, and
+ * 128 bytes in mode 8,
+ * - 259 bytes of mode independent scratch, common to modes 0-15.
+ *
+ * Sequencer scratch RAM is 1024 bytes.  This scratch memory is
+ * divided into mode dependent and mode independent scratch with this
+ * memory further subdivided into pages of size 32 bytes. There are 5
+ * pages (160 bytes) of mode independent scratch and 3 pages of
+ * dependent scratch memory for modes 0-7 (768 bytes). Mode 8 pages
+ * 0-2 dependent scratch overlap with pages 0-2 of mode independent
+ * scratch memory.
+ *
+ * The host accesses this scratch in a different manner from the
+ * central sequencer. The sequencer has to use CSEQ registers CSCRPAGE
+ * and CMnSCRPAGE to access the scratch memory. A flat mapping of the
+ * scratch memory is avaliable for software convenience and to prevent
+ * corruption while the sequencer is running. This memory is mapped
+ * onto addresses 800h - BFFh, total of 400h bytes.
+ *
+ * These addresses are mapped as follows:
+ *
+ *        800h-83Fh   Mode Dependent Scratch Mode 0 Pages 0-1
+ *        840h-87Fh   Mode Dependent Scratch Mode 1 Pages 0-1
+ *        880h-8BFh   Mode Dependent Scratch Mode 2 Pages 0-1
+ *        8C0h-8FFh   Mode Dependent Scratch Mode 3 Pages 0-1
+ *        900h-93Fh   Mode Dependent Scratch Mode 4 Pages 0-1
+ *        940h-97Fh   Mode Dependent Scratch Mode 5 Pages 0-1
+ *        980h-9BFh   Mode Dependent Scratch Mode 6 Pages 0-1
+ *        9C0h-9FFh   Mode Dependent Scratch Mode 7 Pages 0-1
+ *        A00h-A5Fh   Mode Dependent Scratch Mode 8 Pages 0-2
+ *                    Mode Independent Scratch Pages 0-2
+ *        A60h-A7Fh   Mode Dependent Scratch Mode 8 Page 3
+ *                    Mode Independent Scratch Page 3
+ *        A80h-AFFh   Mode Independent Scratch Pages 4-7
+ *        B00h-B1Fh   Mode Dependent Scratch Mode 0 Page 2
+ *        B20h-B3Fh   Mode Dependent Scratch Mode 1 Page 2
+ *        B40h-B5Fh   Mode Dependent Scratch Mode 2 Page 2
+ *        B60h-B7Fh   Mode Dependent Scratch Mode 3 Page 2
+ *        B80h-B9Fh   Mode Dependent Scratch Mode 4 Page 2
+ *        BA0h-BBFh   Mode Dependent Scratch Mode 5 Page 2
+ *        BC0h-BDFh   Mode Dependent Scratch Mode 6 Page 2
+ *        BE0h-BFFh   Mode Dependent Scratch Mode 7 Page 2
+ */
+
+/* General macros */
+#define CSEQ_PAGE_SIZE			32  /* Scratch page size (in bytes) */
+
+/* All macros start with offsets from base + 0x800 (CMAPPEDSCR).
+ * Mode dependent scratch page 0, mode 0.
+ * For modes 1-7 you have to do arithmetic. */
+#define CSEQ_LRM_SAVE_SINDEX		(CMAPPEDSCR + 0x0000)
+#define CSEQ_LRM_SAVE_SCBPTR		(CMAPPEDSCR + 0x0002)
+#define CSEQ_Q_LINK_HEAD		(CMAPPEDSCR + 0x0004)
+#define CSEQ_Q_LINK_TAIL		(CMAPPEDSCR + 0x0006)
+#define CSEQ_LRM_SAVE_SCRPAGE		(CMAPPEDSCR + 0x0008)
+
+/* Mode dependent scratch page 0 mode 8 macros. */
+#define CSEQ_RET_ADDR			(CMAPPEDSCR + 0x0200)
+#define CSEQ_RET_SCBPTR			(CMAPPEDSCR + 0x0202)
+#define CSEQ_SAVE_SCBPTR		(CMAPPEDSCR + 0x0204)
+#define CSEQ_EMPTY_TRANS_CTX		(CMAPPEDSCR + 0x0206)
+#define CSEQ_RESP_LEN			(CMAPPEDSCR + 0x0208)
+#define CSEQ_TMF_SCBPTR			(CMAPPEDSCR + 0x020A)
+#define CSEQ_GLOBAL_PREV_SCB		(CMAPPEDSCR + 0x020C)
+#define CSEQ_GLOBAL_HEAD		(CMAPPEDSCR + 0x020E)
+#define CSEQ_CLEAR_LU_HEAD		(CMAPPEDSCR + 0x0210)
+#define CSEQ_TMF_OPCODE			(CMAPPEDSCR + 0x0212)
+#define CSEQ_SCRATCH_FLAGS		(CMAPPEDSCR + 0x0213)
+#define CSEQ_HSB_SITE                   (CMAPPEDSCR + 0x021A)
+#define CSEQ_FIRST_INV_SCB_SITE		(CMAPPEDSCR + 0x021C)
+#define CSEQ_FIRST_INV_DDB_SITE		(CMAPPEDSCR + 0x021E)
+
+/* Mode dependent scratch page 1 mode 8 macros. */
+#define CSEQ_LUN_TO_CLEAR		(CMAPPEDSCR + 0x0220)
+#define CSEQ_LUN_TO_CHECK		(CMAPPEDSCR + 0x0228)
+
+/* Mode dependent scratch page 2 mode 8 macros */
+#define CSEQ_HQ_NEW_POINTER		(CMAPPEDSCR + 0x0240)
+#define CSEQ_HQ_DONE_BASE		(CMAPPEDSCR + 0x0248)
+#define CSEQ_HQ_DONE_POINTER		(CMAPPEDSCR + 0x0250)
+#define CSEQ_HQ_DONE_PASS		(CMAPPEDSCR + 0x0254)
+
+/* Mode independent scratch page 4 macros. */
+#define CSEQ_Q_EXE_HEAD			(CMAPPEDSCR + 0x0280)
+#define CSEQ_Q_EXE_TAIL			(CMAPPEDSCR + 0x0282)
+#define CSEQ_Q_DONE_HEAD                (CMAPPEDSCR + 0x0284)
+#define CSEQ_Q_DONE_TAIL                (CMAPPEDSCR + 0x0286)
+#define CSEQ_Q_SEND_HEAD		(CMAPPEDSCR + 0x0288)
+#define CSEQ_Q_SEND_TAIL		(CMAPPEDSCR + 0x028A)
+#define CSEQ_Q_DMA2CHIM_HEAD		(CMAPPEDSCR + 0x028C)
+#define CSEQ_Q_DMA2CHIM_TAIL		(CMAPPEDSCR + 0x028E)
+#define CSEQ_Q_COPY_HEAD		(CMAPPEDSCR + 0x0290)
+#define CSEQ_Q_COPY_TAIL		(CMAPPEDSCR + 0x0292)
+#define CSEQ_REG0			(CMAPPEDSCR + 0x0294)
+#define CSEQ_REG1			(CMAPPEDSCR + 0x0296)
+#define CSEQ_REG2			(CMAPPEDSCR + 0x0298)
+#define CSEQ_LINK_CTL_Q_MAP		(CMAPPEDSCR + 0x029C)
+#define CSEQ_MAX_CSEQ_MODE		(CMAPPEDSCR + 0x029D)
+#define CSEQ_FREE_LIST_HACK_COUNT	(CMAPPEDSCR + 0x029E)
+
+/* Mode independent scratch page 5 macros. */
+#define CSEQ_EST_NEXUS_REQ_QUEUE	(CMAPPEDSCR + 0x02A0)
+#define CSEQ_EST_NEXUS_REQ_COUNT	(CMAPPEDSCR + 0x02A8)
+#define CSEQ_Q_EST_NEXUS_HEAD		(CMAPPEDSCR + 0x02B0)
+#define CSEQ_Q_EST_NEXUS_TAIL		(CMAPPEDSCR + 0x02B2)
+#define CSEQ_NEED_EST_NEXUS_SCB		(CMAPPEDSCR + 0x02B4)
+#define CSEQ_EST_NEXUS_REQ_HEAD		(CMAPPEDSCR + 0x02B6)
+#define CSEQ_EST_NEXUS_REQ_TAIL		(CMAPPEDSCR + 0x02B7)
+#define CSEQ_EST_NEXUS_SCB_OFFSET	(CMAPPEDSCR + 0x02B8)
+
+/* Mode independent scratch page 6 macros. */
+#define CSEQ_INT_ROUT_RET_ADDR0		(CMAPPEDSCR + 0x02C0)
+#define CSEQ_INT_ROUT_RET_ADDR1		(CMAPPEDSCR + 0x02C2)
+#define CSEQ_INT_ROUT_SCBPTR		(CMAPPEDSCR + 0x02C4)
+#define CSEQ_INT_ROUT_MODE		(CMAPPEDSCR + 0x02C6)
+#define CSEQ_ISR_SCRATCH_FLAGS		(CMAPPEDSCR + 0x02C7)
+#define CSEQ_ISR_SAVE_SINDEX		(CMAPPEDSCR + 0x02C8)
+#define CSEQ_ISR_SAVE_DINDEX		(CMAPPEDSCR + 0x02CA)
+#define CSEQ_Q_MONIRTT_HEAD		(CMAPPEDSCR + 0x02D0)
+#define CSEQ_Q_MONIRTT_TAIL		(CMAPPEDSCR + 0x02D2)
+#define CSEQ_FREE_SCB_MASK		(CMAPPEDSCR + 0x02D5)
+#define CSEQ_BUILTIN_FREE_SCB_HEAD	(CMAPPEDSCR + 0x02D6)
+#define CSEQ_BUILTIN_FREE_SCB_TAIL	(CMAPPEDSCR + 0x02D8)
+#define CSEQ_EXTENDED_FREE_SCB_HEAD	(CMAPPEDSCR + 0x02DA)
+#define CSEQ_EXTENDED_FREE_SCB_TAIL	(CMAPPEDSCR + 0x02DC)
+
+/* Mode independent scratch page 7 macros. */
+#define CSEQ_EMPTY_REQ_QUEUE		(CMAPPEDSCR + 0x02E0)
+#define CSEQ_EMPTY_REQ_COUNT		(CMAPPEDSCR + 0x02E8)
+#define CSEQ_Q_EMPTY_HEAD		(CMAPPEDSCR + 0x02F0)
+#define CSEQ_Q_EMPTY_TAIL		(CMAPPEDSCR + 0x02F2)
+#define CSEQ_NEED_EMPTY_SCB		(CMAPPEDSCR + 0x02F4)
+#define CSEQ_EMPTY_REQ_HEAD		(CMAPPEDSCR + 0x02F6)
+#define CSEQ_EMPTY_REQ_TAIL		(CMAPPEDSCR + 0x02F7)
+#define CSEQ_EMPTY_SCB_OFFSET		(CMAPPEDSCR + 0x02F8)
+#define CSEQ_PRIMITIVE_DATA		(CMAPPEDSCR + 0x02FA)
+#define CSEQ_TIMEOUT_CONST		(CMAPPEDSCR + 0x02FC)
+
+/***************************************************************************
+* Link m Sequencer scratch RAM is 512 bytes.
+* This scratch memory is divided into mode dependent and mode
+* independent scratch with this memory further subdivided into
+* pages of size 32 bytes. There are 4 pages (128 bytes) of
+* mode independent scratch and 4 pages of dependent scratch
+* memory for modes 0-2 (384 bytes).
+*
+* The host accesses this scratch in a different manner from the
+* link sequencer. The sequencer has to use LSEQ registers
+* LmSCRPAGE and LmMnSCRPAGE to access the scratch memory. A flat
+* mapping of the scratch memory is avaliable for software
+* convenience and to prevent corruption while the sequencer is
+* running. This memory is mapped onto addresses 800h - 9FFh.
+*
+* These addresses are mapped as follows:
+*
+*        800h-85Fh   Mode Dependent Scratch Mode 0 Pages 0-2
+*        860h-87Fh   Mode Dependent Scratch Mode 0 Page 3
+*                    Mode Dependent Scratch Mode 5 Page 0
+*        880h-8DFh   Mode Dependent Scratch Mode 1 Pages 0-2
+*        8E0h-8FFh   Mode Dependent Scratch Mode 1 Page 3
+*                    Mode Dependent Scratch Mode 5 Page 1
+*        900h-95Fh   Mode Dependent Scratch Mode 2 Pages 0-2
+*        960h-97Fh   Mode Dependent Scratch Mode 2 Page 3
+*                    Mode Dependent Scratch Mode 5 Page 2
+*        980h-9DFh   Mode Independent Scratch Pages 0-3
+*        9E0h-9FFh   Mode Independent Scratch Page 3
+*                    Mode Dependent Scratch Mode 5 Page 3
+*
+****************************************************************************/
+/* General macros */
+#define LSEQ_MODE_SCRATCH_SIZE		0x80 /* Size of scratch RAM per mode */
+#define LSEQ_PAGE_SIZE			0x20 /* Scratch page size (in bytes) */
+#define LSEQ_MODE5_PAGE0_OFFSET 	0x60
+
+/* Common mode dependent scratch page 0 macros for modes 0,1,2, and 5 */
+/* Indexed using LSEQ_MODE_SCRATCH_SIZE * mode, for modes 0,1,2. */
+#define LmSEQ_RET_ADDR(LinkNum)		(LmSCRATCH(LinkNum) + 0x0000)
+#define LmSEQ_REG0_MODE(LinkNum)	(LmSCRATCH(LinkNum) + 0x0002)
+#define LmSEQ_MODE_FLAGS(LinkNum)	(LmSCRATCH(LinkNum) + 0x0004)
+
+/* Mode flag macros (byte 0) */
+#define		SAS_SAVECTX_OCCURRED		0x80
+#define		SAS_OOBSVC_OCCURRED		0x40
+#define		SAS_OOB_DEVICE_PRESENT		0x20
+#define		SAS_CFGHDR_OCCURRED		0x10
+#define		SAS_RCV_INTS_ARE_DISABLED	0x08
+#define		SAS_OOB_HOT_PLUG_CNCT		0x04
+#define		SAS_AWAIT_OPEN_CONNECTION	0x02
+#define		SAS_CFGCMPLT_OCCURRED		0x01
+
+/* Mode flag macros (byte 1) */
+#define		SAS_RLSSCB_OCCURRED		0x80
+#define		SAS_FORCED_HEADER_MISS		0x40
+
+#define LmSEQ_RET_ADDR2(LinkNum)	(LmSCRATCH(LinkNum) + 0x0006)
+#define LmSEQ_RET_ADDR1(LinkNum)	(LmSCRATCH(LinkNum) + 0x0008)
+#define LmSEQ_OPCODE_TO_CSEQ(LinkNum)	(LmSCRATCH(LinkNum) + 0x000B)
+#define LmSEQ_DATA_TO_CSEQ(LinkNum)	(LmSCRATCH(LinkNum) + 0x000C)
+
+/* Mode dependent scratch page 0 macros for mode 0 (non-common) */
+/* Absolute offsets */
+#define LmSEQ_FIRST_INV_DDB_SITE(LinkNum)	(LmSCRATCH(LinkNum) + 0x000E)
+#define LmSEQ_EMPTY_TRANS_CTX(LinkNum)		(LmSCRATCH(LinkNum) + 0x0010)
+#define LmSEQ_RESP_LEN(LinkNum)			(LmSCRATCH(LinkNum) + 0x0012)
+#define LmSEQ_FIRST_INV_SCB_SITE(LinkNum)	(LmSCRATCH(LinkNum) + 0x0014)
+#define LmSEQ_INTEN_SAVE(LinkNum)		(LmSCRATCH(LinkNum) + 0x0016)
+#define LmSEQ_LINK_RST_FRM_LEN(LinkNum)		(LmSCRATCH(LinkNum) + 0x001A)
+#define LmSEQ_LINK_RST_PROTOCOL(LinkNum)	(LmSCRATCH(LinkNum) + 0x001B)
+#define LmSEQ_RESP_STATUS(LinkNum)		(LmSCRATCH(LinkNum) + 0x001C)
+#define LmSEQ_LAST_LOADED_SGE(LinkNum)		(LmSCRATCH(LinkNum) + 0x001D)
+#define LmSEQ_SAVE_SCBPTR(LinkNum)		(LmSCRATCH(LinkNum) + 0x001E)
+
+/* Mode dependent scratch page 0 macros for mode 1 (non-common) */
+/* Absolute offsets */
+#define LmSEQ_Q_XMIT_HEAD(LinkNum)		(LmSCRATCH(LinkNum) + 0x008E)
+#define LmSEQ_M1_EMPTY_TRANS_CTX(LinkNum)	(LmSCRATCH(LinkNum) + 0x0090)
+#define LmSEQ_INI_CONN_TAG(LinkNum)		(LmSCRATCH(LinkNum) + 0x0092)
+#define LmSEQ_FAILED_OPEN_STATUS(LinkNum)	(LmSCRATCH(LinkNum) + 0x009A)
+#define LmSEQ_XMIT_REQUEST_TYPE(LinkNum)	(LmSCRATCH(LinkNum) + 0x009B)
+#define LmSEQ_M1_RESP_STATUS(LinkNum)		(LmSCRATCH(LinkNum) + 0x009C)
+#define LmSEQ_M1_LAST_LOADED_SGE(LinkNum)	(LmSCRATCH(LinkNum) + 0x009D)
+#define LmSEQ_M1_SAVE_SCBPTR(LinkNum)		(LmSCRATCH(LinkNum) + 0x009E)
+
+/* Mode dependent scratch page 0 macros for mode 2 (non-common) */
+#define LmSEQ_PORT_COUNTER(LinkNum)		(LmSCRATCH(LinkNum) + 0x010E)
+#define LmSEQ_PM_TABLE_PTR(LinkNum)		(LmSCRATCH(LinkNum) + 0x0110)
+#define LmSEQ_SATA_INTERLOCK_TMR_SAVE(LinkNum)	(LmSCRATCH(LinkNum) + 0x0112)
+#define LmSEQ_IP_BITL(LinkNum)			(LmSCRATCH(LinkNum) + 0x0114)
+#define LmSEQ_COPY_SMP_CONN_TAG(LinkNum)	(LmSCRATCH(LinkNum) + 0x0116)
+#define LmSEQ_P0M2_OFFS1AH(LinkNum)		(LmSCRATCH(LinkNum) + 0x011A)
+
+/* Mode dependent scratch page 0 macros for modes 4/5 (non-common) */
+/* Absolute offsets */
+#define LmSEQ_SAVED_OOB_STATUS(LinkNum)		(LmSCRATCH(LinkNum) + 0x006E)
+#define LmSEQ_SAVED_OOB_MODE(LinkNum)		(LmSCRATCH(LinkNum) + 0x006F)
+#define LmSEQ_Q_LINK_HEAD(LinkNum)		(LmSCRATCH(LinkNum) + 0x0070)
+#define LmSEQ_LINK_RST_ERR(LinkNum)		(LmSCRATCH(LinkNum) + 0x0072)
+#define LmSEQ_SAVED_OOB_SIGNALS(LinkNum)	(LmSCRATCH(LinkNum) + 0x0073)
+#define LmSEQ_SAS_RESET_MODE(LinkNum)		(LmSCRATCH(LinkNum) + 0x0074)
+#define LmSEQ_LINK_RESET_RETRY_COUNT(LinkNum)	(LmSCRATCH(LinkNum) + 0x0075)
+#define LmSEQ_NUM_LINK_RESET_RETRIES(LinkNum)	(LmSCRATCH(LinkNum) + 0x0076)
+#define LmSEQ_OOB_INT_ENABLES(LinkNum)		(LmSCRATCH(LinkNum) + 0x007A)
+#define LmSEQ_NOTIFY_TIMER_TIMEOUT(LinkNum)	(LmSCRATCH(LinkNum) + 0x007C)
+#define LmSEQ_NOTIFY_TIMER_DOWN_COUNT(LinkNum)	(LmSCRATCH(LinkNum) + 0x007E)
+
+/* Mode dependent scratch page 1, mode 0 and mode 1 */
+#define LmSEQ_SG_LIST_PTR_ADDR0(LinkNum)        (LmSCRATCH(LinkNum) + 0x0020)
+#define LmSEQ_SG_LIST_PTR_ADDR1(LinkNum)        (LmSCRATCH(LinkNum) + 0x0030)
+#define LmSEQ_M1_SG_LIST_PTR_ADDR0(LinkNum)     (LmSCRATCH(LinkNum) + 0x00A0)
+#define LmSEQ_M1_SG_LIST_PTR_ADDR1(LinkNum)     (LmSCRATCH(LinkNum) + 0x00B0)
+
+/* Mode dependent scratch page 1 macros for mode 2 */
+/* Absolute offsets */
+#define LmSEQ_INVALID_DWORD_COUNT(LinkNum)	(LmSCRATCH(LinkNum) + 0x0120)
+#define LmSEQ_DISPARITY_ERROR_COUNT(LinkNum) 	(LmSCRATCH(LinkNum) + 0x0124)
+#define LmSEQ_LOSS_OF_SYNC_COUNT(LinkNum)	(LmSCRATCH(LinkNum) + 0x0128)
+
+/* Mode dependent scratch page 1 macros for mode 4/5 */
+#define LmSEQ_FRAME_TYPE_MASK(LinkNum)	      (LmSCRATCH(LinkNum) + 0x00E0)
+#define LmSEQ_HASHED_DEST_ADDR_MASK(LinkNum)  (LmSCRATCH(LinkNum) + 0x00E1)
+#define LmSEQ_HASHED_SRC_ADDR_MASK_PRINT(LinkNum) (LmSCRATCH(LinkNum) + 0x00E4)
+#define LmSEQ_HASHED_SRC_ADDR_MASK(LinkNum)   (LmSCRATCH(LinkNum) + 0x00E5)
+#define LmSEQ_NUM_FILL_BYTES_MASK(LinkNum)    (LmSCRATCH(LinkNum) + 0x00EB)
+#define LmSEQ_TAG_MASK(LinkNum)		      (LmSCRATCH(LinkNum) + 0x00F0)
+#define LmSEQ_TARGET_PORT_XFER_TAG(LinkNum)   (LmSCRATCH(LinkNum) + 0x00F2)
+#define LmSEQ_DATA_OFFSET(LinkNum)	      (LmSCRATCH(LinkNum) + 0x00F4)
+
+/* Mode dependent scratch page 2 macros for mode 0 */
+/* Absolute offsets */
+#define LmSEQ_SMP_RCV_TIMER_TERM_TS(LinkNum)	(LmSCRATCH(LinkNum) + 0x0040)
+#define LmSEQ_DEVICE_BITS(LinkNum)		(LmSCRATCH(LinkNum) + 0x005B)
+#define LmSEQ_SDB_DDB(LinkNum)			(LmSCRATCH(LinkNum) + 0x005C)
+#define LmSEQ_SDB_NUM_TAGS(LinkNum)		(LmSCRATCH(LinkNum) + 0x005E)
+#define LmSEQ_SDB_CURR_TAG(LinkNum)		(LmSCRATCH(LinkNum) + 0x005F)
+
+/* Mode dependent scratch page 2 macros for mode 1 */
+/* Absolute offsets */
+/* byte 0 bits 1-0 are domain select. */
+#define LmSEQ_TX_ID_ADDR_FRAME(LinkNum)		(LmSCRATCH(LinkNum) + 0x00C0)
+#define LmSEQ_OPEN_TIMER_TERM_TS(LinkNum)	(LmSCRATCH(LinkNum) + 0x00C8)
+#define LmSEQ_SRST_AS_TIMER_TERM_TS(LinkNum)	(LmSCRATCH(LinkNum) + 0x00CC)
+#define LmSEQ_LAST_LOADED_SG_EL(LinkNum)	(LmSCRATCH(LinkNum) + 0x00D4)
+
+/* Mode dependent scratch page 2 macros for mode 2 */
+/* Absolute offsets */
+#define LmSEQ_STP_SHUTDOWN_TIMER_TERM_TS(LinkNum) (LmSCRATCH(LinkNum) + 0x0140)
+#define LmSEQ_CLOSE_TIMER_TERM_TS(LinkNum)	(LmSCRATCH(LinkNum) + 0x0144)
+#define LmSEQ_BREAK_TIMER_TERM_TS(LinkNum)	(LmSCRATCH(LinkNum) + 0x0148)
+#define LmSEQ_DWS_RESET_TIMER_TERM_TS(LinkNum)	(LmSCRATCH(LinkNum) + 0x014C)
+#define LmSEQ_SATA_INTERLOCK_TIMER_TERM_TS(LinkNum) \
+						(LmSCRATCH(LinkNum) + 0x0150)
+#define LmSEQ_MCTL_TIMER_TERM_TS(LinkNum)	(LmSCRATCH(LinkNum) + 0x0154)
+
+/* Mode dependent scratch page 2 macros for mode 5 */
+#define LmSEQ_COMINIT_TIMER_TERM_TS(LinkNum)	(LmSCRATCH(LinkNum) + 0x0160)
+#define LmSEQ_RCV_ID_TIMER_TERM_TS(LinkNum)	(LmSCRATCH(LinkNum) + 0x0164)
+#define LmSEQ_RCV_FIS_TIMER_TERM_TS(LinkNum)	(LmSCRATCH(LinkNum) + 0x0168)
+#define LmSEQ_DEV_PRES_TIMER_TERM_TS(LinkNum)	(LmSCRATCH(LinkNum) + 0x016C)
+
+/* Mode dependent scratch page 3 macros for modes 0 and 1 */
+/* None defined */
+
+/* Mode dependent scratch page 3 macros for modes 2 and 5 */
+/* None defined */
+
+/* Mode Independent Scratch page 0 macros. */
+#define LmSEQ_Q_TGTXFR_HEAD(LinkNum)	(LmSCRATCH(LinkNum) + 0x0180)
+#define LmSEQ_Q_TGTXFR_TAIL(LinkNum)	(LmSCRATCH(LinkNum) + 0x0182)
+#define LmSEQ_LINK_NUMBER(LinkNum)	(LmSCRATCH(LinkNum) + 0x0186)
+#define LmSEQ_SCRATCH_FLAGS(LinkNum)	(LmSCRATCH(LinkNum) + 0x0187)
+/*
+ * Currently only bit 0, SAS_DWSAQD, is used.
+ */
+#define		SAS_DWSAQD			0x01  /*
+						       * DWSSTATUS: DWSAQD
+						       * bit las read in ISR.
+						       */
+#define  LmSEQ_CONNECTION_STATE(LinkNum) (LmSCRATCH(LinkNum) + 0x0188)
+/* Connection states (byte 0) */
+#define		SAS_WE_OPENED_CS		0x01
+#define		SAS_DEVICE_OPENED_CS		0x02
+#define		SAS_WE_SENT_DONE_CS		0x04
+#define		SAS_DEVICE_SENT_DONE_CS		0x08
+#define		SAS_WE_SENT_CLOSE_CS		0x10
+#define		SAS_DEVICE_SENT_CLOSE_CS	0x20
+#define		SAS_WE_SENT_BREAK_CS		0x40
+#define		SAS_DEVICE_SENT_BREAK_CS	0x80
+/* Connection states (byte 1) */
+#define		SAS_OPN_TIMEOUT_OR_OPN_RJCT_CS	0x01
+#define		SAS_AIP_RECEIVED_CS		0x02
+#define		SAS_CREDIT_TIMEOUT_OCCURRED_CS	0x04
+#define		SAS_ACKNAK_TIMEOUT_OCCURRED_CS	0x08
+#define		SAS_SMPRSP_TIMEOUT_OCCURRED_CS	0x10
+#define		SAS_DONE_TIMEOUT_OCCURRED_CS	0x20
+/* Connection states (byte 2) */
+#define		SAS_SMP_RESPONSE_RECEIVED_CS	0x01
+#define		SAS_INTLK_TIMEOUT_OCCURRED_CS	0x02
+#define		SAS_DEVICE_SENT_DMAT_CS		0x04
+#define		SAS_DEVICE_SENT_SYNCSRST_CS	0x08
+#define		SAS_CLEARING_AFFILIATION_CS	0x20
+#define		SAS_RXTASK_ACTIVE_CS		0x40
+#define		SAS_TXTASK_ACTIVE_CS		0x80
+/* Connection states (byte 3) */
+#define		SAS_PHY_LOSS_OF_SIGNAL_CS	0x01
+#define		SAS_DWS_TIMER_EXPIRED_CS	0x02
+#define		SAS_LINK_RESET_NOT_COMPLETE_CS	0x04
+#define		SAS_PHY_DISABLED_CS		0x08
+#define		SAS_LINK_CTL_TASK_ACTIVE_CS	0x10
+#define		SAS_PHY_EVENT_TASK_ACTIVE_CS	0x20
+#define		SAS_DEVICE_SENT_ID_FRAME_CS	0x40
+#define		SAS_DEVICE_SENT_REG_FIS_CS	0x40
+#define		SAS_DEVICE_SENT_HARD_RESET_CS	0x80
+#define  	SAS_PHY_IS_DOWN_FLAGS	(SAS_PHY_LOSS_OF_SIGNAL_CS|\
+					 SAS_DWS_TIMER_EXPIRED_CS |\
+					 SAS_LINK_RESET_NOT_COMPLETE_CS|\
+					 SAS_PHY_DISABLED_CS)
+
+#define		SAS_LINK_CTL_PHY_EVENT_FLAGS   (SAS_LINK_CTL_TASK_ACTIVE_CS |\
+						SAS_PHY_EVENT_TASK_ACTIVE_CS |\
+						SAS_DEVICE_SENT_ID_FRAME_CS  |\
+						SAS_DEVICE_SENT_HARD_RESET_CS)
+
+#define LmSEQ_CONCTL(LinkNum)		(LmSCRATCH(LinkNum) + 0x018C)
+#define LmSEQ_CONSTAT(LinkNum)		(LmSCRATCH(LinkNum) + 0x018E)
+#define LmSEQ_CONNECTION_MODES(LinkNum)	(LmSCRATCH(LinkNum) + 0x018F)
+#define LmSEQ_REG1_ISR(LinkNum)		(LmSCRATCH(LinkNum) + 0x0192)
+#define LmSEQ_REG2_ISR(LinkNum)		(LmSCRATCH(LinkNum) + 0x0194)
+#define LmSEQ_REG3_ISR(LinkNum)		(LmSCRATCH(LinkNum) + 0x0196)
+#define LmSEQ_REG0_ISR(LinkNum)		(LmSCRATCH(LinkNum) + 0x0198)
+
+/* Mode independent scratch page 1 macros. */
+#define LmSEQ_EST_NEXUS_SCBPTR0(LinkNum)	(LmSCRATCH(LinkNum) + 0x01A0)
+#define LmSEQ_EST_NEXUS_SCBPTR1(LinkNum)	(LmSCRATCH(LinkNum) + 0x01A2)
+#define LmSEQ_EST_NEXUS_SCBPTR2(LinkNum)	(LmSCRATCH(LinkNum) + 0x01A4)
+#define LmSEQ_EST_NEXUS_SCBPTR3(LinkNum)	(LmSCRATCH(LinkNum) + 0x01A6)
+#define LmSEQ_EST_NEXUS_SCB_OPCODE0(LinkNum)	(LmSCRATCH(LinkNum) + 0x01A8)
+#define LmSEQ_EST_NEXUS_SCB_OPCODE1(LinkNum)	(LmSCRATCH(LinkNum) + 0x01A9)
+#define LmSEQ_EST_NEXUS_SCB_OPCODE2(LinkNum)	(LmSCRATCH(LinkNum) + 0x01AA)
+#define LmSEQ_EST_NEXUS_SCB_OPCODE3(LinkNum)	(LmSCRATCH(LinkNum) + 0x01AB)
+#define LmSEQ_EST_NEXUS_SCB_HEAD(LinkNum)	(LmSCRATCH(LinkNum) + 0x01AC)
+#define LmSEQ_EST_NEXUS_SCB_TAIL(LinkNum)	(LmSCRATCH(LinkNum) + 0x01AD)
+#define LmSEQ_EST_NEXUS_BUF_AVAIL(LinkNum)	(LmSCRATCH(LinkNum) + 0x01AE)
+#define LmSEQ_TIMEOUT_CONST(LinkNum)		(LmSCRATCH(LinkNum) + 0x01B8)
+#define LmSEQ_ISR_SAVE_SINDEX(LinkNum)	        (LmSCRATCH(LinkNum) + 0x01BC)
+#define LmSEQ_ISR_SAVE_DINDEX(LinkNum)	        (LmSCRATCH(LinkNum) + 0x01BE)
+
+/* Mode independent scratch page 2 macros. */
+#define LmSEQ_EMPTY_SCB_PTR0(LinkNum)	(LmSCRATCH(LinkNum) + 0x01C0)
+#define LmSEQ_EMPTY_SCB_PTR1(LinkNum)	(LmSCRATCH(LinkNum) + 0x01C2)
+#define LmSEQ_EMPTY_SCB_PTR2(LinkNum)	(LmSCRATCH(LinkNum) + 0x01C4)
+#define LmSEQ_EMPTY_SCB_PTR3(LinkNum)	(LmSCRATCH(LinkNum) + 0x01C6)
+#define LmSEQ_EMPTY_SCB_OPCD0(LinkNum)	(LmSCRATCH(LinkNum) + 0x01C8)
+#define LmSEQ_EMPTY_SCB_OPCD1(LinkNum)	(LmSCRATCH(LinkNum) + 0x01C9)
+#define LmSEQ_EMPTY_SCB_OPCD2(LinkNum)	(LmSCRATCH(LinkNum) + 0x01CA)
+#define LmSEQ_EMPTY_SCB_OPCD3(LinkNum)	(LmSCRATCH(LinkNum) + 0x01CB)
+#define LmSEQ_EMPTY_SCB_HEAD(LinkNum)	(LmSCRATCH(LinkNum) + 0x01CC)
+#define LmSEQ_EMPTY_SCB_TAIL(LinkNum)	(LmSCRATCH(LinkNum) + 0x01CD)
+#define LmSEQ_EMPTY_BUFS_AVAIL(LinkNum)	(LmSCRATCH(LinkNum) + 0x01CE)
+#define LmSEQ_ATA_SCR_REGS(LinkNum)	(LmSCRATCH(LinkNum) + 0x01D4)
+
+/* Mode independent scratch page 3 macros. */
+#define LmSEQ_DEV_PRES_TMR_TOUT_CONST(LinkNum)	(LmSCRATCH(LinkNum) + 0x01E0)
+#define LmSEQ_SATA_INTERLOCK_TIMEOUT(LinkNum)	(LmSCRATCH(LinkNum) + 0x01E4)
+#define LmSEQ_STP_SHUTDOWN_TIMEOUT(LinkNum)	(LmSCRATCH(LinkNum) + 0x01E8)
+#define LmSEQ_SRST_ASSERT_TIMEOUT(LinkNum)	(LmSCRATCH(LinkNum) + 0x01EC)
+#define LmSEQ_RCV_FIS_TIMEOUT(LinkNum)		(LmSCRATCH(LinkNum) + 0x01F0)
+#define LmSEQ_ONE_MILLISEC_TIMEOUT(LinkNum)	(LmSCRATCH(LinkNum) + 0x01F4)
+#define LmSEQ_TEN_MS_COMINIT_TIMEOUT(LinkNum)	(LmSCRATCH(LinkNum) + 0x01F8)
+#define LmSEQ_SMP_RCV_TIMEOUT(LinkNum)		(LmSCRATCH(LinkNum) + 0x01FC)
+
+#endif
diff --git a/drivers/scsi/aic94xx/aic94xx_sas.h b/drivers/scsi/aic94xx/aic94xx_sas.h
new file mode 100644
index 000000000000..64d231712345
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx_sas.h
@@ -0,0 +1,785 @@
+/*
+ * Aic94xx SAS/SATA driver SAS definitions and hardware interface header file.
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#ifndef _AIC94XX_SAS_H_
+#define _AIC94XX_SAS_H_
+
+#include <scsi/libsas.h>
+
+/* ---------- DDBs ---------- */
+/* DDBs are device descriptor blocks which describe a device in the
+ * domain that this sequencer can maintain low-level connections for
+ * us.  They are be 64 bytes.
+ */
+
+struct asd_ddb_ssp_smp_target_port {
+	u8     conn_type;	  /* byte 0 */
+#define DDB_TP_CONN_TYPE 0x81	  /* Initiator port and addr frame type 0x01 */
+
+	u8     conn_rate;
+	__be16 init_conn_tag;
+	u8     dest_sas_addr[8];  /* bytes 4-11 */
+
+	__le16 send_queue_head;
+	u8     sq_suspended;
+	u8     ddb_type;	  /* DDB_TYPE_TARGET */
+#define DDB_TYPE_UNUSED    0xFF
+#define DDB_TYPE_TARGET    0xFE
+#define DDB_TYPE_INITIATOR 0xFD
+#define DDB_TYPE_PM_PORT   0xFC
+
+	__le16 _r_a;
+	__be16 awt_def;
+
+	u8     compat_features;	  /* byte 20 */
+	u8     pathway_blocked_count;
+	__be16 arb_wait_time;
+	__be32 more_compat_features; /* byte 24 */
+
+	u8     conn_mask;
+	u8     flags;	  /* concurrent conn:2,2 and open:0(1) */
+#define CONCURRENT_CONN_SUPP 0x04
+#define OPEN_REQUIRED        0x01
+
+	u16    _r_b;
+	__le16 exec_queue_tail;
+	__le16 send_queue_tail;
+	__le16 sister_ddb;
+
+	__le16 _r_c;
+
+	u8     max_concurrent_conn;
+	u8     num_concurrent_conn;
+	u8     num_contexts;
+
+	u8     _r_d;
+
+	__le16 active_task_count;
+
+	u8     _r_e[9];
+
+	u8     itnl_reason;	  /* I_T nexus loss reason */
+
+	__le16 _r_f;
+
+	__le16 itnl_timeout;
+#define ITNL_TIMEOUT_CONST 0x7D0 /* 2 seconds */
+
+	__le32 itnl_timestamp;
+} __attribute__ ((packed));
+
+struct asd_ddb_stp_sata_target_port {
+	u8     conn_type;	  /* byte 0 */
+	u8     conn_rate;
+	__be16 init_conn_tag;
+	u8     dest_sas_addr[8];  /* bytes 4-11 */
+
+	__le16 send_queue_head;
+	u8     sq_suspended;
+	u8     ddb_type;	  /* DDB_TYPE_TARGET */
+
+	__le16 _r_a;
+
+	__be16 awt_def;
+	u8     compat_features;	  /* byte 20 */
+	u8     pathway_blocked_count;
+	__be16 arb_wait_time;
+	__be32 more_compat_features; /* byte 24 */
+
+	u8     conn_mask;
+	u8     flags;	  /* concurrent conn:2,2 and open:0(1) */
+#define SATA_MULTIPORT     0x80
+#define SUPPORTS_AFFIL     0x40
+#define STP_AFFIL_POL      0x20
+
+	u8     _r_b;
+	u8     flags2;		  /* STP close policy:0 */
+#define STP_CL_POL_NO_TX    0x00
+#define STP_CL_POL_BTW_CMDS 0x01
+
+	__le16 exec_queue_tail;
+	__le16 send_queue_tail;
+	__le16 sister_ddb;
+	__le16 ata_cmd_scbptr;
+	__le32 sata_tag_alloc_mask;
+	__le16 active_task_count;
+	__le16 _r_c;
+	__le32 sata_sactive;
+	u8     num_sata_tags;
+	u8     sata_status;
+	u8     sata_ending_status;
+	u8     itnl_reason;	  /* I_T nexus loss reason */
+	__le16 ncq_data_scb_ptr;
+	__le16 itnl_timeout;
+	__le32 itnl_timestamp;
+} __attribute__ ((packed));
+
+/* This struct asd_ddb_init_port, describes the device descriptor block
+ * of an initiator port (when the sequencer is operating in target mode).
+ * Bytes [0,11] and [20,27] are from the OPEN address frame.
+ * The sequencer allocates an initiator port DDB entry.
+ */
+struct asd_ddb_init_port {
+	u8     conn_type;	  /* byte 0 */
+	u8     conn_rate;
+	__be16 init_conn_tag;     /* BE */
+	u8     dest_sas_addr[8];
+	__le16 send_queue_head;   /* LE, byte 12 */
+	u8     sq_suspended;
+	u8     ddb_type;	  /* DDB_TYPE_INITIATOR */
+	__le16 _r_a;
+	__be16 awt_def;		  /* BE */
+	u8     compat_features;
+	u8     pathway_blocked_count;
+	__be16 arb_wait_time;	  /* BE */
+	__be32 more_compat_features; /* BE */
+	u8     conn_mask;
+	u8     flags;		  /* == 5 */
+	u16    _r_b;
+	__le16 exec_queue_tail;	  /* execution queue tail */
+	__le16 send_queue_tail;
+	__le16 sister_ddb;
+	__le16 init_resp_timeout; /* initiator response timeout */
+	__le32 _r_c;
+	__le16 active_tasks;	  /* active task count */
+	__le16 init_list;	  /* initiator list link pointer */
+	__le32 _r_d;
+	u8     max_conn_to[3]; /* from Conn-Disc mode page, in us, LE */
+	u8     itnl_reason;	  /* I_T nexus loss reason */
+	__le16 bus_inact_to; /* from Conn-Disc mode page, in 100 us, LE */
+	__le16 itnl_to;		  /* from the Protocol Specific Port Ctrl MP */
+	__le32 itnl_timestamp;
+} __attribute__ ((packed));
+
+/* This struct asd_ddb_sata_tag, describes a look-up table to be used
+ * by the sequencers.  SATA II, IDENTIFY DEVICE data, word 76, bit 8:
+ * NCQ support.  This table is used by the sequencers to find the
+ * corresponding SCB, given a SATA II tag value.
+ */
+struct asd_ddb_sata_tag {
+	__le16 scb_pointer[32];
+} __attribute__ ((packed));
+
+/* This struct asd_ddb_sata_pm_table, describes a port number to
+ * connection handle look-up table.  SATA targets attached to a port
+ * multiplier require a 4-bit port number value.  There is one DDB
+ * entry of this type for each SATA port multiplier (sister DDB).
+ * Given a SATA PM port number, this table gives us the SATA PM Port
+ * DDB of the SATA port multiplier port (i.e. the SATA target
+ * discovered on the port).
+ */
+struct asd_ddb_sata_pm_table {
+	__le16 ddb_pointer[16];
+	__le16 _r_a[16];
+} __attribute__ ((packed));
+
+/* This struct asd_ddb_sata_pm_port, describes the SATA port multiplier
+ * port format DDB.
+ */
+struct asd_ddb_sata_pm_port {
+	u8     _r_a[15];
+	u8     ddb_type;
+	u8     _r_b[13];
+	u8     pm_port_flags;
+#define PM_PORT_MASK  0xF0
+#define PM_PORT_SET   0x02
+	u8     _r_c[6];
+	__le16 sister_ddb;
+	__le16 ata_cmd_scbptr;
+	__le32 sata_tag_alloc_mask;
+	__le16 active_task_count;
+	__le16 parent_ddb;
+	__le32 sata_sactive;
+	u8     num_sata_tags;
+	u8     sata_status;
+	u8     sata_ending_status;
+	u8     _r_d[9];
+} __attribute__ ((packed));
+
+/* This struct asd_ddb_seq_shared, describes a DDB shared by the
+ * central and link sequencers.  port_map_by_links is indexed phy
+ * number [0,7]; each byte is a bit mask of all the phys that are in
+ * the same port as the indexed phy.
+ */
+struct asd_ddb_seq_shared {
+	__le16 q_free_ddb_head;
+	__le16 q_free_ddb_tail;
+	__le16 q_free_ddb_cnt;
+	__le16 q_used_ddb_head;
+	__le16 q_used_ddb_tail;
+	__le16 shared_mem_lock;
+	__le16 smp_conn_tag;
+	__le16 est_nexus_buf_cnt;
+	__le16 est_nexus_buf_thresh;
+	u32    _r_a;
+	u8     settable_max_contexts;
+	u8     _r_b[23];
+	u8     conn_not_active;
+	u8     phy_is_up;
+	u8     _r_c[8];
+	u8     port_map_by_links[8];
+} __attribute__ ((packed));
+
+/* ---------- SG Element ---------- */
+
+/* This struct sg_el, describes the hardware scatter gather buffer
+ * element.  All entries are little endian.  In an SCB, there are 2 of
+ * this, plus one more, called a link element of this indicating a
+ * sublist if needed.
+ *
+ * A link element has only the bus address set and the flags (DS) bit
+ * valid.  The bus address points to the start of the sublist.
+ *
+ * If a sublist is needed, then that sublist should also include the 2
+ * sg_el embedded in the SCB, in which case next_sg_offset is 32,
+ * since sizeof(sg_el) = 16; EOS should be 1 and EOL 0 in this case.
+ */
+struct sg_el {
+	__le64 bus_addr;
+	__le32 size;
+	__le16 _r;
+	u8     next_sg_offs;
+	u8     flags;
+#define ASD_SG_EL_DS_MASK   0x30
+#define ASD_SG_EL_DS_OCM    0x10
+#define ASD_SG_EL_DS_HM     0x00
+#define ASD_SG_EL_LIST_MASK 0xC0
+#define ASD_SG_EL_LIST_EOL  0x40
+#define ASD_SG_EL_LIST_EOS  0x80
+} __attribute__ ((packed));
+
+/* ---------- SCBs ---------- */
+
+/* An SCB (sequencer control block) is comprised of a common header
+ * and a task part, for a total of 128 bytes.  All fields are in LE
+ * order, unless otherwise noted.
+ */
+
+/* This struct scb_header, defines the SCB header format.
+ */
+struct scb_header {
+	__le64 next_scb;
+	__le16 index;		  /* transaction context */
+	u8     opcode;
+} __attribute__ ((packed));
+
+/* SCB opcodes: Execution queue
+ */
+#define INITIATE_SSP_TASK       0x00
+#define INITIATE_LONG_SSP_TASK  0x01
+#define INITIATE_BIDIR_SSP_TASK 0x02
+#define ABORT_TASK              0x03
+#define INITIATE_SSP_TMF        0x04
+#define SSP_TARG_GET_DATA       0x05
+#define SSP_TARG_GET_DATA_GOOD  0x06
+#define SSP_TARG_SEND_RESP      0x07
+#define QUERY_SSP_TASK          0x08
+#define INITIATE_ATA_TASK       0x09
+#define INITIATE_ATAPI_TASK     0x0a
+#define CONTROL_ATA_DEV         0x0b
+#define INITIATE_SMP_TASK       0x0c
+#define SMP_TARG_SEND_RESP      0x0f
+
+/* SCB opcodes: Send Queue
+ */
+#define SSP_TARG_SEND_DATA      0x40
+#define SSP_TARG_SEND_DATA_GOOD 0x41
+
+/* SCB opcodes: Link Queue
+ */
+#define CONTROL_PHY             0x80
+#define SEND_PRIMITIVE          0x81
+#define INITIATE_LINK_ADM_TASK  0x82
+
+/* SCB opcodes: other
+ */
+#define EMPTY_SCB               0xc0
+#define INITIATE_SEQ_ADM_TASK   0xc1
+#define EST_ICL_TARG_WINDOW     0xc2
+#define COPY_MEM                0xc3
+#define CLEAR_NEXUS             0xc4
+#define INITIATE_DDB_ADM_TASK   0xc6
+#define ESTABLISH_NEXUS_ESCB    0xd0
+
+#define LUN_SIZE                8
+
+/* See SAS spec, task IU
+ */
+struct ssp_task_iu {
+	u8     lun[LUN_SIZE];	  /* BE */
+	u16    _r_a;
+	u8     tmf;
+	u8     _r_b;
+	__be16 tag;		  /* BE */
+	u8     _r_c[14];
+} __attribute__ ((packed));
+
+/* See SAS spec, command IU
+ */
+struct ssp_command_iu {
+	u8     lun[LUN_SIZE];
+	u8     _r_a;
+	u8     efb_prio_attr;	  /* enable first burst, task prio & attr */
+#define EFB_MASK        0x80
+#define TASK_PRIO_MASK	0x78
+#define TASK_ATTR_MASK  0x07
+
+	u8    _r_b;
+	u8     add_cdb_len;	  /* in dwords, since bit 0,1 are reserved */
+	union {
+		u8     cdb[16];
+		struct {
+			__le64 long_cdb_addr;	  /* bus address, LE */
+			__le32 long_cdb_size;	  /* LE */
+			u8     _r_c[3];
+			u8     eol_ds;		  /* eol:6,6, ds:5,4 */
+		} long_cdb;	  /* sequencer extension */
+	};
+} __attribute__ ((packed));
+
+struct xfer_rdy_iu {
+	__be32 requested_offset;  /* BE */
+	__be32 write_data_len;	  /* BE */
+	__be32 _r_a;
+} __attribute__ ((packed));
+
+/* ---------- SCB tasks ---------- */
+
+/* This is both ssp_task and long_ssp_task
+ */
+struct initiate_ssp_task {
+	u8     proto_conn_rate;	  /* proto:6,4, conn_rate:3,0 */
+	__le32 total_xfer_len;
+	struct ssp_frame_hdr  ssp_frame;
+	struct ssp_command_iu ssp_cmd;
+	__le16 sister_scb;	  /* 0xFFFF */
+	__le16 conn_handle;	  /* index to DDB for the intended target */
+	u8     data_dir;	  /* :1,0 */
+#define DATA_DIR_NONE   0x00
+#define DATA_DIR_IN     0x01
+#define DATA_DIR_OUT    0x02
+#define DATA_DIR_BYRECIPIENT 0x03
+
+	u8     _r_a;
+	u8     retry_count;
+	u8     _r_b[5];
+	struct sg_el sg_element[3]; /* 2 real and 1 link */
+} __attribute__ ((packed));
+
+/* This defines both ata_task and atapi_task.
+ * ata: C bit of FIS should be 1,
+ * atapi: C bit of FIS should be 1, and command register should be 0xA0,
+ * to indicate a packet command.
+ */
+struct initiate_ata_task {
+	u8     proto_conn_rate;
+	__le32 total_xfer_len;
+	struct host_to_dev_fis fis;
+	__le32 data_offs;
+	u8     atapi_packet[16];
+	u8     _r_a[12];
+	__le16 sister_scb;
+	__le16 conn_handle;
+	u8     ata_flags;	  /* CSMI:6,6, DTM:4,4, QT:3,3, data dir:1,0 */
+#define CSMI_TASK           0x40
+#define DATA_XFER_MODE_DMA  0x10
+#define ATA_Q_TYPE_MASK     0x08
+#define	ATA_Q_TYPE_UNTAGGED 0x00
+#define ATA_Q_TYPE_NCQ      0x08
+
+	u8     _r_b;
+	u8     retry_count;
+	u8     _r_c;
+	u8     flags;
+#define STP_AFFIL_POLICY   0x20
+#define SET_AFFIL_POLICY   0x10
+#define RET_PARTIAL_SGLIST 0x02
+
+	u8     _r_d[3];
+	struct sg_el sg_element[3];
+} __attribute__ ((packed));
+
+struct initiate_smp_task {
+	u8     proto_conn_rate;
+	u8     _r_a[40];
+	struct sg_el smp_req;
+	__le16 sister_scb;
+	__le16 conn_handle;
+	u8     _r_c[8];
+	struct sg_el smp_resp;
+	u8     _r_d[32];
+} __attribute__ ((packed));
+
+struct control_phy {
+	u8     phy_id;
+	u8     sub_func;
+#define DISABLE_PHY            0x00
+#define ENABLE_PHY             0x01
+#define RELEASE_SPINUP_HOLD    0x02
+#define ENABLE_PHY_NO_SAS_OOB  0x03
+#define ENABLE_PHY_NO_SATA_OOB 0x04
+#define PHY_NO_OP              0x05
+#define EXECUTE_HARD_RESET     0x81
+
+	u8     func_mask;
+	u8     speed_mask;
+	u8     hot_plug_delay;
+	u8     port_type;
+	u8     flags;
+#define DEV_PRES_TIMER_OVERRIDE_ENABLE 0x01
+#define DISABLE_PHY_IF_OOB_FAILS       0x02
+
+	__le32 timeout_override;
+	u8     link_reset_retries;
+	u8     _r_a[47];
+	__le16 conn_handle;
+	u8     _r_b[56];
+} __attribute__ ((packed));
+
+struct control_ata_dev {
+	u8     proto_conn_rate;
+	__le32 _r_a;
+	struct host_to_dev_fis fis;
+	u8     _r_b[32];
+	__le16 sister_scb;
+	__le16 conn_handle;
+	u8     ata_flags;	  /* 0 */
+	u8     _r_c[55];
+} __attribute__ ((packed));
+
+struct empty_scb {
+	u8     num_valid;
+	__le32 _r_a;
+#define ASD_EDBS_PER_SCB 7
+/* header+data+CRC+DMA suffix data */
+#define ASD_EDB_SIZE (24+1024+4+16)
+	struct sg_el eb[ASD_EDBS_PER_SCB];
+#define ELEMENT_NOT_VALID  0xC0
+} __attribute__ ((packed));
+
+struct initiate_link_adm {
+	u8     phy_id;
+	u8     sub_func;
+#define GET_LINK_ERROR_COUNT      0x00
+#define RESET_LINK_ERROR_COUNT    0x01
+#define ENABLE_NOTIFY_SPINUP_INTS 0x02
+
+	u8     _r_a[57];
+	__le16 conn_handle;
+	u8     _r_b[56];
+} __attribute__ ((packed));
+
+struct copy_memory {
+	u8     _r_a;
+	__le16 xfer_len;
+	__le16 _r_b;
+	__le64 src_busaddr;
+	u8     src_ds;		  /* See definition of sg_el */
+	u8     _r_c[45];
+	__le16 conn_handle;
+	__le64 _r_d;
+	__le64 dest_busaddr;
+	u8     dest_ds;		  /* See definition of sg_el */
+	u8     _r_e[39];
+} __attribute__ ((packed));
+
+struct abort_task {
+	u8     proto_conn_rate;
+	__le32 _r_a;
+	struct ssp_frame_hdr ssp_frame;
+	struct ssp_task_iu ssp_task;
+	__le16 sister_scb;
+	__le16 conn_handle;
+	u8     flags;	  /* ovrd_itnl_timer:3,3, suspend_data_trans:2,2 */
+#define SUSPEND_DATA_TRANS 0x04
+
+	u8     _r_b;
+	u8     retry_count;
+	u8     _r_c[5];
+	__le16 index;  /* Transaction context of task to be queried */
+	__le16 itnl_to;
+	u8     _r_d[44];
+} __attribute__ ((packed));
+
+struct clear_nexus {
+	u8     nexus;
+#define NEXUS_ADAPTER  0x00
+#define NEXUS_PORT     0x01
+#define NEXUS_I_T      0x02
+#define NEXUS_I_T_L    0x03
+#define NEXUS_TAG      0x04
+#define NEXUS_TRANS_CX 0x05
+#define NEXUS_SATA_TAG 0x06
+#define NEXUS_T_L      0x07
+#define NEXUS_L        0x08
+#define NEXUS_T_TAG    0x09
+
+	__le32 _r_a;
+	u8     flags;
+#define SUSPEND_TX     0x80
+#define RESUME_TX      0x40
+#define SEND_Q         0x04
+#define EXEC_Q         0x02
+#define NOTINQ         0x01
+
+	u8     _r_b[3];
+	u8     conn_mask;
+	u8     _r_c[19];
+	struct ssp_task_iu ssp_task; /* LUN and TAG */
+	__le16 _r_d;
+	__le16 conn_handle;
+	__le64 _r_e;
+	__le16 index;  /* Transaction context of task to be cleared */
+	__le16 context;		  /* Clear nexus context */
+	u8     _r_f[44];
+} __attribute__ ((packed));
+
+struct initiate_ssp_tmf {
+	u8     proto_conn_rate;
+	__le32 _r_a;
+	struct ssp_frame_hdr ssp_frame;
+	struct ssp_task_iu ssp_task;
+	__le16 sister_scb;
+	__le16 conn_handle;
+	u8     flags;	  /* itnl override and suspend data tx */
+#define OVERRIDE_ITNL_TIMER  8
+
+	u8     _r_b;
+	u8     retry_count;
+	u8     _r_c[5];
+	__le16 index;  /* Transaction context of task to be queried */
+	__le16 itnl_to;
+	u8     _r_d[44];
+} __attribute__ ((packed));
+
+/* Transmits an arbitrary primitive on the link.
+ * Used for NOTIFY and BROADCAST.
+ */
+struct send_prim {
+	u8     phy_id;
+	u8     wait_transmit; 	  /* :0,0 */
+	u8     xmit_flags;
+#define XMTPSIZE_MASK      0xF0
+#define XMTPSIZE_SINGLE    0x10
+#define XMTPSIZE_REPEATED  0x20
+#define XMTPSIZE_CONT      0x20
+#define XMTPSIZE_TRIPLE    0x30
+#define XMTPSIZE_REDUNDANT 0x60
+#define XMTPSIZE_INF       0
+
+#define XMTCONTEN          0x04
+#define XMTPFRM            0x02	  /* Transmit at the next frame boundary */
+#define XMTPIMM            0x01	  /* Transmit immediately */
+
+	__le16 _r_a;
+	u8     prim[4];		  /* K, D0, D1, D2 */
+	u8     _r_b[50];
+	__le16 conn_handle;
+	u8     _r_c[56];
+} __attribute__ ((packed));
+
+/* This describes both SSP Target Get Data and SSP Target Get Data And
+ * Send Good Response SCBs.  Used when the sequencer is operating in
+ * target mode...
+ */
+struct ssp_targ_get_data {
+	u8     proto_conn_rate;
+	__le32 total_xfer_len;
+	struct ssp_frame_hdr ssp_frame;
+	struct xfer_rdy_iu  xfer_rdy;
+	u8     lun[LUN_SIZE];
+	__le64 _r_a;
+	__le16 sister_scb;
+	__le16 conn_handle;
+	u8     data_dir;	  /* 01b */
+	u8     _r_b;
+	u8     retry_count;
+	u8     _r_c[5];
+	struct sg_el sg_element[3];
+} __attribute__ ((packed));
+
+/* ---------- The actual SCB struct ---------- */
+
+struct scb {
+	struct scb_header header;
+	union {
+		struct initiate_ssp_task ssp_task;
+		struct initiate_ata_task ata_task;
+		struct initiate_smp_task smp_task;
+		struct control_phy       control_phy;
+		struct control_ata_dev   control_ata_dev;
+		struct empty_scb         escb;
+		struct initiate_link_adm link_adm;
+		struct copy_memory       cp_mem;
+		struct abort_task        abort_task;
+		struct clear_nexus       clear_nexus;
+		struct initiate_ssp_tmf  ssp_tmf;
+	};
+} __attribute__ ((packed));
+
+/* ---------- Done List ---------- */
+/* The done list entry opcode field is defined below.
+ * The mnemonic encoding and meaning is as follows:
+ * TC - Task Complete, status was received and acknowledged
+ * TF - Task Failed, indicates an error prior to receiving acknowledgment
+ *   for the command:
+ *   - no conn,
+ *   - NACK or R_ERR received in response to this command,
+ *   - credit blocked or not available, or in the case of SMP request,
+ *   - no SMP response was received.
+ *   In these four cases it is known that the target didn't receive the
+ *   command.
+ * TI - Task Interrupted, error after the command was acknowledged.  It is
+ *   known that the command was received by the target.
+ * TU - Task Unacked, command was transmitted but neither ACK (R_OK) nor NAK
+ *   (R_ERR) was received due to loss of signal, broken connection, loss of
+ *   dword sync or other reason.  The application client should send the
+ *   appropriate task query.
+ * TA - Task Aborted, see TF.
+ * _RESP - The completion includes an empty buffer containing status.
+ * TO - Timeout.
+ */
+#define TC_NO_ERROR             0x00
+#define TC_UNDERRUN             0x01
+#define TC_OVERRUN              0x02
+#define TF_OPEN_TO              0x03
+#define TF_OPEN_REJECT          0x04
+#define TI_BREAK                0x05
+#define TI_PROTO_ERR            0x06
+#define TC_SSP_RESP             0x07
+#define TI_PHY_DOWN             0x08
+#define TF_PHY_DOWN             0x09
+#define TC_LINK_ADM_RESP        0x0a
+#define TC_CSMI                 0x0b
+#define TC_ATA_RESP             0x0c
+#define TU_PHY_DOWN             0x0d
+#define TU_BREAK                0x0e
+#define TI_SATA_TO              0x0f
+#define TI_NAK                  0x10
+#define TC_CONTROL_PHY          0x11
+#define TF_BREAK                0x12
+#define TC_RESUME               0x13
+#define TI_ACK_NAK_TO           0x14
+#define TF_SMPRSP_TO            0x15
+#define TF_SMP_XMIT_RCV_ERR     0x16
+#define TC_PARTIAL_SG_LIST      0x17
+#define TU_ACK_NAK_TO           0x18
+#define TU_SATA_TO              0x19
+#define TF_NAK_RECV             0x1a
+#define TA_I_T_NEXUS_LOSS       0x1b
+#define TC_ATA_R_ERR_RECV       0x1c
+#define TF_TMF_NO_CTX           0x1d
+#define TA_ON_REQ               0x1e
+#define TF_TMF_NO_TAG           0x1f
+#define TF_TMF_TAG_FREE         0x20
+#define TF_TMF_TASK_DONE        0x21
+#define TF_TMF_NO_CONN_HANDLE   0x22
+#define TC_TASK_CLEARED         0x23
+#define TI_SYNCS_RECV           0x24
+#define TU_SYNCS_RECV           0x25
+#define TF_IRTT_TO              0x26
+#define TF_NO_SMP_CONN          0x27
+#define TF_IU_SHORT             0x28
+#define TF_DATA_OFFS_ERR        0x29
+#define TF_INV_CONN_HANDLE      0x2a
+#define TF_REQUESTED_N_PENDING  0x2b
+
+/* 0xc1 - 0xc7: empty buffer received,
+   0xd1 - 0xd7: establish nexus empty buffer received
+*/
+/* This is the ESCB mask */
+#define ESCB_RECVD              0xC0
+
+
+/* This struct done_list_struct defines the done list entry.
+ * All fields are LE.
+ */
+struct done_list_struct {
+	__le16 index;		  /* aka transaction context */
+	u8     opcode;
+	u8     status_block[4];
+	u8     toggle;		  /* bit 0 */
+#define DL_TOGGLE_MASK     0x01
+} __attribute__ ((packed));
+
+/* ---------- PHYS ---------- */
+
+struct asd_phy {
+	struct asd_sas_phy        sas_phy;
+	struct asd_phy_desc   *phy_desc; /* hw profile */
+
+	struct sas_identify_frame *identify_frame;
+	struct asd_dma_tok  *id_frm_tok;
+
+	u8         frame_rcvd[ASD_EDB_SIZE];
+};
+
+
+#define ASD_SCB_SIZE sizeof(struct scb)
+#define ASD_DDB_SIZE sizeof(struct asd_ddb_ssp_smp_target_port)
+
+/* Define this to 0 if you do not want NOTIFY (ENABLE SPINIP) sent.
+ * Default: 0x10 (it's a mask)
+ */
+#define ASD_NOTIFY_ENABLE_SPINUP  0x10
+
+/* If enabled, set this to the interval between transmission
+ * of NOTIFY (ENABLE SPINUP). In units of 200 us.
+ */
+#define ASD_NOTIFY_TIMEOUT        2500
+
+/* Initial delay after OOB, before we transmit NOTIFY (ENABLE SPINUP).
+ * If 0, transmit immediately. In milliseconds.
+ */
+#define ASD_NOTIFY_DOWN_COUNT     0
+
+/* Device present timer timeout constant, 10 ms. */
+#define ASD_DEV_PRESENT_TIMEOUT   0x2710
+
+#define ASD_SATA_INTERLOCK_TIMEOUT 0
+
+/* How long to wait before shutting down an STP connection, unless
+ * an STP target sent frame(s). 50 usec.
+ * IGNORED by the sequencer (i.e. value 0 always).
+ */
+#define ASD_STP_SHUTDOWN_TIMEOUT  0x0
+
+/* ATA soft reset timer timeout. 5 usec. */
+#define ASD_SRST_ASSERT_TIMEOUT   0x05
+
+/* 31 sec */
+#define ASD_RCV_FIS_TIMEOUT       0x01D905C0
+
+#define ASD_ONE_MILLISEC_TIMEOUT  0x03e8
+
+/* COMINIT timer */
+#define ASD_TEN_MILLISEC_TIMEOUT  0x2710
+#define ASD_COMINIT_TIMEOUT ASD_TEN_MILLISEC_TIMEOUT
+
+/* 1 sec */
+#define ASD_SMP_RCV_TIMEOUT       0x000F4240
+
+#endif
diff --git a/drivers/scsi/aic94xx/aic94xx_scb.c b/drivers/scsi/aic94xx/aic94xx_scb.c
new file mode 100644
index 000000000000..fc1b7438a913
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx_scb.c
@@ -0,0 +1,732 @@
+/*
+ * Aic94xx SAS/SATA driver SCB management.
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#include <linux/pci.h>
+
+#include "aic94xx.h"
+#include "aic94xx_reg.h"
+#include "aic94xx_hwi.h"
+#include "aic94xx_seq.h"
+
+#include "aic94xx_dump.h"
+
+/* ---------- EMPTY SCB ---------- */
+
+#define DL_PHY_MASK      7
+#define BYTES_DMAED      0
+#define PRIMITIVE_RECVD  0x08
+#define PHY_EVENT        0x10
+#define LINK_RESET_ERROR 0x18
+#define TIMER_EVENT      0x20
+#define REQ_TASK_ABORT   0xF0
+#define REQ_DEVICE_RESET 0xF1
+#define SIGNAL_NCQ_ERROR 0xF2
+#define CLEAR_NCQ_ERROR  0xF3
+
+#define PHY_EVENTS_STATUS (CURRENT_LOSS_OF_SIGNAL | CURRENT_OOB_DONE   \
+			   | CURRENT_SPINUP_HOLD | CURRENT_GTO_TIMEOUT \
+			   | CURRENT_OOB_ERROR)
+
+static inline void get_lrate_mode(struct asd_phy *phy, u8 oob_mode)
+{
+	switch (oob_mode & 7) {
+	case PHY_SPEED_60:
+		/* FIXME: sas transport class doesn't have this */
+		phy->sas_phy.linkrate = PHY_LINKRATE_6;
+		phy->sas_phy.phy->negotiated_linkrate = SAS_LINK_RATE_6_0_GBPS;
+		break;
+	case PHY_SPEED_30:
+		phy->sas_phy.linkrate = PHY_LINKRATE_3;
+		phy->sas_phy.phy->negotiated_linkrate = SAS_LINK_RATE_3_0_GBPS;
+		break;
+	case PHY_SPEED_15:
+		phy->sas_phy.linkrate = PHY_LINKRATE_1_5;
+		phy->sas_phy.phy->negotiated_linkrate = SAS_LINK_RATE_1_5_GBPS;
+		break;
+	}
+	if (oob_mode & SAS_MODE)
+		phy->sas_phy.oob_mode = SAS_OOB_MODE;
+	else if (oob_mode & SATA_MODE)
+		phy->sas_phy.oob_mode = SATA_OOB_MODE;
+}
+
+static inline void asd_phy_event_tasklet(struct asd_ascb *ascb,
+					 struct done_list_struct *dl)
+{
+	struct asd_ha_struct *asd_ha = ascb->ha;
+	struct sas_ha_struct *sas_ha = &asd_ha->sas_ha;
+	int phy_id = dl->status_block[0] & DL_PHY_MASK;
+	struct asd_phy *phy = &asd_ha->phys[phy_id];
+
+	u8 oob_status = dl->status_block[1] & PHY_EVENTS_STATUS;
+	u8 oob_mode   = dl->status_block[2];
+
+	switch (oob_status) {
+	case CURRENT_LOSS_OF_SIGNAL:
+		/* directly attached device was removed */
+		ASD_DPRINTK("phy%d: device unplugged\n", phy_id);
+		asd_turn_led(asd_ha, phy_id, 0);
+		sas_phy_disconnected(&phy->sas_phy);
+		sas_ha->notify_phy_event(&phy->sas_phy, PHYE_LOSS_OF_SIGNAL);
+		break;
+	case CURRENT_OOB_DONE:
+		/* hot plugged device */
+		asd_turn_led(asd_ha, phy_id, 1);
+		get_lrate_mode(phy, oob_mode);
+		ASD_DPRINTK("phy%d device plugged: lrate:0x%x, proto:0x%x\n",
+			    phy_id, phy->sas_phy.linkrate, phy->sas_phy.iproto);
+		sas_ha->notify_phy_event(&phy->sas_phy, PHYE_OOB_DONE);
+		break;
+	case CURRENT_SPINUP_HOLD:
+		/* hot plug SATA, no COMWAKE sent */
+		asd_turn_led(asd_ha, phy_id, 1);
+		sas_ha->notify_phy_event(&phy->sas_phy, PHYE_SPINUP_HOLD);
+		break;
+	case CURRENT_GTO_TIMEOUT:
+	case CURRENT_OOB_ERROR:
+		ASD_DPRINTK("phy%d error while OOB: oob status:0x%x\n", phy_id,
+			    dl->status_block[1]);
+		asd_turn_led(asd_ha, phy_id, 0);
+		sas_phy_disconnected(&phy->sas_phy);
+		sas_ha->notify_phy_event(&phy->sas_phy, PHYE_OOB_ERROR);
+		break;
+	}
+}
+
+/* If phys are enabled sparsely, this will do the right thing. */
+static inline unsigned ord_phy(struct asd_ha_struct *asd_ha,
+			       struct asd_phy *phy)
+{
+	u8 enabled_mask = asd_ha->hw_prof.enabled_phys;
+	int i, k = 0;
+
+	for_each_phy(enabled_mask, enabled_mask, i) {
+		if (&asd_ha->phys[i] == phy)
+			return k;
+		k++;
+	}
+	return 0;
+}
+
+/**
+ * asd_get_attached_sas_addr -- extract/generate attached SAS address
+ * phy: pointer to asd_phy
+ * sas_addr: pointer to buffer where the SAS address is to be written
+ *
+ * This function extracts the SAS address from an IDENTIFY frame
+ * received.  If OOB is SATA, then a SAS address is generated from the
+ * HA tables.
+ *
+ * LOCKING: the frame_rcvd_lock needs to be held since this parses the frame
+ * buffer.
+ */
+static inline void asd_get_attached_sas_addr(struct asd_phy *phy, u8 *sas_addr)
+{
+	if (phy->sas_phy.frame_rcvd[0] == 0x34
+	    && phy->sas_phy.oob_mode == SATA_OOB_MODE) {
+		struct asd_ha_struct *asd_ha = phy->sas_phy.ha->lldd_ha;
+		/* FIS device-to-host */
+		u64 addr = be64_to_cpu(*(__be64 *)phy->phy_desc->sas_addr);
+
+		addr += asd_ha->hw_prof.sata_name_base + ord_phy(asd_ha, phy);
+		*(__be64 *)sas_addr = cpu_to_be64(addr);
+	} else {
+		struct sas_identify_frame *idframe =
+			(void *) phy->sas_phy.frame_rcvd;
+		memcpy(sas_addr, idframe->sas_addr, SAS_ADDR_SIZE);
+	}
+}
+
+static inline void asd_bytes_dmaed_tasklet(struct asd_ascb *ascb,
+					   struct done_list_struct *dl,
+					   int edb_id, int phy_id)
+{
+	unsigned long flags;
+	int edb_el = edb_id + ascb->edb_index;
+	struct asd_dma_tok *edb = ascb->ha->seq.edb_arr[edb_el];
+	struct asd_phy *phy = &ascb->ha->phys[phy_id];
+	struct sas_ha_struct *sas_ha = phy->sas_phy.ha;
+	u16 size = ((dl->status_block[3] & 7) << 8) | dl->status_block[2];
+
+	size = min(size, (u16) sizeof(phy->frame_rcvd));
+
+	spin_lock_irqsave(&phy->sas_phy.frame_rcvd_lock, flags);
+	memcpy(phy->sas_phy.frame_rcvd, edb->vaddr, size);
+	phy->sas_phy.frame_rcvd_size = size;
+	asd_get_attached_sas_addr(phy, phy->sas_phy.attached_sas_addr);
+	spin_unlock_irqrestore(&phy->sas_phy.frame_rcvd_lock, flags);
+	asd_dump_frame_rcvd(phy, dl);
+	sas_ha->notify_port_event(&phy->sas_phy, PORTE_BYTES_DMAED);
+}
+
+static inline void asd_link_reset_err_tasklet(struct asd_ascb *ascb,
+					      struct done_list_struct *dl,
+					      int phy_id)
+{
+	struct asd_ha_struct *asd_ha = ascb->ha;
+	struct sas_ha_struct *sas_ha = &asd_ha->sas_ha;
+	struct asd_sas_phy *sas_phy = sas_ha->sas_phy[phy_id];
+	u8 lr_error = dl->status_block[1];
+	u8 retries_left = dl->status_block[2];
+
+	switch (lr_error) {
+	case 0:
+		ASD_DPRINTK("phy%d: Receive ID timer expired\n", phy_id);
+		break;
+	case 1:
+		ASD_DPRINTK("phy%d: Loss of signal\n", phy_id);
+		break;
+	case 2:
+		ASD_DPRINTK("phy%d: Loss of dword sync\n", phy_id);
+		break;
+	case 3:
+		ASD_DPRINTK("phy%d: Receive FIS timeout\n", phy_id);
+		break;
+	default:
+		ASD_DPRINTK("phy%d: unknown link reset error code: 0x%x\n",
+			    phy_id, lr_error);
+		break;
+	}
+
+	asd_turn_led(asd_ha, phy_id, 0);
+	sas_phy_disconnected(sas_phy);
+	sas_ha->notify_port_event(sas_phy, PORTE_LINK_RESET_ERR);
+
+	if (retries_left == 0) {
+		int num = 1;
+		struct asd_ascb *cp = asd_ascb_alloc_list(ascb->ha, &num,
+							  GFP_ATOMIC);
+		if (!cp) {
+			asd_printk("%s: out of memory\n", __FUNCTION__);
+			goto out;
+		}
+		ASD_DPRINTK("phy%d: retries:0 performing link reset seq\n",
+			    phy_id);
+		asd_build_control_phy(cp, phy_id, ENABLE_PHY);
+		if (asd_post_ascb_list(ascb->ha, cp, 1) != 0)
+			asd_ascb_free(cp);
+	}
+out:
+	;
+}
+
+static inline void asd_primitive_rcvd_tasklet(struct asd_ascb *ascb,
+					      struct done_list_struct *dl,
+					      int phy_id)
+{
+	unsigned long flags;
+	struct sas_ha_struct *sas_ha = &ascb->ha->sas_ha;
+	struct asd_sas_phy *sas_phy = sas_ha->sas_phy[phy_id];
+	u8  reg  = dl->status_block[1];
+	u32 cont = dl->status_block[2] << ((reg & 3)*8);
+
+	reg &= ~3;
+	switch (reg) {
+	case LmPRMSTAT0BYTE0:
+		switch (cont) {
+		case LmBROADCH:
+		case LmBROADRVCH0:
+		case LmBROADRVCH1:
+		case LmBROADSES:
+			ASD_DPRINTK("phy%d: BROADCAST change received:%d\n",
+				    phy_id, cont);
+			spin_lock_irqsave(&sas_phy->sas_prim_lock, flags);
+			sas_phy->sas_prim = ffs(cont);
+			spin_unlock_irqrestore(&sas_phy->sas_prim_lock, flags);
+			sas_ha->notify_port_event(sas_phy,PORTE_BROADCAST_RCVD);
+			break;
+
+		case LmUNKNOWNP:
+			ASD_DPRINTK("phy%d: unknown BREAK\n", phy_id);
+			break;
+
+		default:
+			ASD_DPRINTK("phy%d: primitive reg:0x%x, cont:0x%04x\n",
+				    phy_id, reg, cont);
+			break;
+		}
+		break;
+	case LmPRMSTAT1BYTE0:
+		switch (cont) {
+		case LmHARDRST:
+			ASD_DPRINTK("phy%d: HARD_RESET primitive rcvd\n",
+				    phy_id);
+			/* The sequencer disables all phys on that port.
+			 * We have to re-enable the phys ourselves. */
+			sas_ha->notify_port_event(sas_phy, PORTE_HARD_RESET);
+			break;
+
+		default:
+			ASD_DPRINTK("phy%d: primitive reg:0x%x, cont:0x%04x\n",
+				    phy_id, reg, cont);
+			break;
+		}
+		break;
+	default:
+		ASD_DPRINTK("unknown primitive register:0x%x\n",
+			    dl->status_block[1]);
+		break;
+	}
+}
+
+/**
+ * asd_invalidate_edb -- invalidate an EDB and if necessary post the ESCB
+ * @ascb: pointer to Empty SCB
+ * @edb_id: index [0,6] to the empty data buffer which is to be invalidated
+ *
+ * After an EDB has been invalidated, if all EDBs in this ESCB have been
+ * invalidated, the ESCB is posted back to the sequencer.
+ * Context is tasklet/IRQ.
+ */
+void asd_invalidate_edb(struct asd_ascb *ascb, int edb_id)
+{
+	struct asd_seq_data *seq = &ascb->ha->seq;
+	struct empty_scb *escb = &ascb->scb->escb;
+	struct sg_el     *eb   = &escb->eb[edb_id];
+	struct asd_dma_tok *edb = seq->edb_arr[ascb->edb_index + edb_id];
+
+	memset(edb->vaddr, 0, ASD_EDB_SIZE);
+	eb->flags |= ELEMENT_NOT_VALID;
+	escb->num_valid--;
+
+	if (escb->num_valid == 0) {
+		int i;
+		/* ASD_DPRINTK("reposting escb: vaddr: 0x%p, "
+			    "dma_handle: 0x%08llx, next: 0x%08llx, "
+			    "index:%d, opcode:0x%02x\n",
+			    ascb->dma_scb.vaddr,
+			    (u64)ascb->dma_scb.dma_handle,
+			    le64_to_cpu(ascb->scb->header.next_scb),
+			    le16_to_cpu(ascb->scb->header.index),
+			    ascb->scb->header.opcode);
+		*/
+		escb->num_valid = ASD_EDBS_PER_SCB;
+		for (i = 0; i < ASD_EDBS_PER_SCB; i++)
+			escb->eb[i].flags = 0;
+		if (!list_empty(&ascb->list))
+			list_del_init(&ascb->list);
+		i = asd_post_escb_list(ascb->ha, ascb, 1);
+		if (i)
+			asd_printk("couldn't post escb, err:%d\n", i);
+	}
+}
+
+static void escb_tasklet_complete(struct asd_ascb *ascb,
+				  struct done_list_struct *dl)
+{
+	struct asd_ha_struct *asd_ha = ascb->ha;
+	struct sas_ha_struct *sas_ha = &asd_ha->sas_ha;
+	int edb = (dl->opcode & DL_PHY_MASK) - 1; /* [0xc1,0xc7] -> [0,6] */
+	u8  sb_opcode = dl->status_block[0];
+	int phy_id = sb_opcode & DL_PHY_MASK;
+	struct asd_sas_phy *sas_phy = sas_ha->sas_phy[phy_id];
+
+	if (edb > 6 || edb < 0) {
+		ASD_DPRINTK("edb is 0x%x! dl->opcode is 0x%x\n",
+			    edb, dl->opcode);
+		ASD_DPRINTK("sb_opcode : 0x%x, phy_id: 0x%x\n",
+			    sb_opcode, phy_id);
+		ASD_DPRINTK("escb: vaddr: 0x%p, "
+			    "dma_handle: 0x%llx, next: 0x%llx, "
+			    "index:%d, opcode:0x%02x\n",
+			    ascb->dma_scb.vaddr,
+			    (unsigned long long)ascb->dma_scb.dma_handle,
+			    (unsigned long long)
+			    le64_to_cpu(ascb->scb->header.next_scb),
+			    le16_to_cpu(ascb->scb->header.index),
+			    ascb->scb->header.opcode);
+	}
+
+	sb_opcode &= ~DL_PHY_MASK;
+
+	switch (sb_opcode) {
+	case BYTES_DMAED:
+		ASD_DPRINTK("%s: phy%d: BYTES_DMAED\n", __FUNCTION__, phy_id);
+		asd_bytes_dmaed_tasklet(ascb, dl, edb, phy_id);
+		break;
+	case PRIMITIVE_RECVD:
+		ASD_DPRINTK("%s: phy%d: PRIMITIVE_RECVD\n", __FUNCTION__,
+			    phy_id);
+		asd_primitive_rcvd_tasklet(ascb, dl, phy_id);
+		break;
+	case PHY_EVENT:
+		ASD_DPRINTK("%s: phy%d: PHY_EVENT\n", __FUNCTION__, phy_id);
+		asd_phy_event_tasklet(ascb, dl);
+		break;
+	case LINK_RESET_ERROR:
+		ASD_DPRINTK("%s: phy%d: LINK_RESET_ERROR\n", __FUNCTION__,
+			    phy_id);
+		asd_link_reset_err_tasklet(ascb, dl, phy_id);
+		break;
+	case TIMER_EVENT:
+		ASD_DPRINTK("%s: phy%d: TIMER_EVENT, lost dw sync\n",
+			    __FUNCTION__, phy_id);
+		asd_turn_led(asd_ha, phy_id, 0);
+		/* the device is gone */
+		sas_phy_disconnected(sas_phy);
+		sas_ha->notify_port_event(sas_phy, PORTE_TIMER_EVENT);
+		break;
+	case REQ_TASK_ABORT:
+		ASD_DPRINTK("%s: phy%d: REQ_TASK_ABORT\n", __FUNCTION__,
+			    phy_id);
+		break;
+	case REQ_DEVICE_RESET:
+		ASD_DPRINTK("%s: phy%d: REQ_DEVICE_RESET\n", __FUNCTION__,
+			    phy_id);
+		break;
+	case SIGNAL_NCQ_ERROR:
+		ASD_DPRINTK("%s: phy%d: SIGNAL_NCQ_ERROR\n", __FUNCTION__,
+			    phy_id);
+		break;
+	case CLEAR_NCQ_ERROR:
+		ASD_DPRINTK("%s: phy%d: CLEAR_NCQ_ERROR\n", __FUNCTION__,
+			    phy_id);
+		break;
+	default:
+		ASD_DPRINTK("%s: phy%d: unknown event:0x%x\n", __FUNCTION__,
+			    phy_id, sb_opcode);
+		ASD_DPRINTK("edb is 0x%x! dl->opcode is 0x%x\n",
+			    edb, dl->opcode);
+		ASD_DPRINTK("sb_opcode : 0x%x, phy_id: 0x%x\n",
+			    sb_opcode, phy_id);
+		ASD_DPRINTK("escb: vaddr: 0x%p, "
+			    "dma_handle: 0x%llx, next: 0x%llx, "
+			    "index:%d, opcode:0x%02x\n",
+			    ascb->dma_scb.vaddr,
+			    (unsigned long long)ascb->dma_scb.dma_handle,
+			    (unsigned long long)
+			    le64_to_cpu(ascb->scb->header.next_scb),
+			    le16_to_cpu(ascb->scb->header.index),
+			    ascb->scb->header.opcode);
+
+		break;
+	}
+
+	asd_invalidate_edb(ascb, edb);
+}
+
+int asd_init_post_escbs(struct asd_ha_struct *asd_ha)
+{
+	struct asd_seq_data *seq = &asd_ha->seq;
+	int i;
+
+	for (i = 0; i < seq->num_escbs; i++)
+		seq->escb_arr[i]->tasklet_complete = escb_tasklet_complete;
+
+	ASD_DPRINTK("posting %d escbs\n", i);
+	return asd_post_escb_list(asd_ha, seq->escb_arr[0], seq->num_escbs);
+}
+
+/* ---------- CONTROL PHY ---------- */
+
+#define CONTROL_PHY_STATUS (CURRENT_DEVICE_PRESENT | CURRENT_OOB_DONE   \
+			    | CURRENT_SPINUP_HOLD | CURRENT_GTO_TIMEOUT \
+			    | CURRENT_OOB_ERROR)
+
+/**
+ * control_phy_tasklet_complete -- tasklet complete for CONTROL PHY ascb
+ * @ascb: pointer to an ascb
+ * @dl: pointer to the done list entry
+ *
+ * This function completes a CONTROL PHY scb and frees the ascb.
+ * A note on LEDs:
+ *  - an LED blinks if there is IO though it,
+ *  - if a device is connected to the LED, it is lit,
+ *  - if no device is connected to the LED, is is dimmed (off).
+ */
+static void control_phy_tasklet_complete(struct asd_ascb *ascb,
+					 struct done_list_struct *dl)
+{
+	struct asd_ha_struct *asd_ha = ascb->ha;
+	struct scb *scb = ascb->scb;
+	struct control_phy *control_phy = &scb->control_phy;
+	u8 phy_id = control_phy->phy_id;
+	struct asd_phy *phy = &ascb->ha->phys[phy_id];
+
+	u8 status     = dl->status_block[0];
+	u8 oob_status = dl->status_block[1];
+	u8 oob_mode   = dl->status_block[2];
+	/* u8 oob_signals= dl->status_block[3]; */
+
+	if (status != 0) {
+		ASD_DPRINTK("%s: phy%d status block opcode:0x%x\n",
+			    __FUNCTION__, phy_id, status);
+		goto out;
+	}
+
+	switch (control_phy->sub_func) {
+	case DISABLE_PHY:
+		asd_ha->hw_prof.enabled_phys &= ~(1 << phy_id);
+		asd_turn_led(asd_ha, phy_id, 0);
+		asd_control_led(asd_ha, phy_id, 0);
+		ASD_DPRINTK("%s: disable phy%d\n", __FUNCTION__, phy_id);
+		break;
+
+	case ENABLE_PHY:
+		asd_control_led(asd_ha, phy_id, 1);
+		if (oob_status & CURRENT_OOB_DONE) {
+			asd_ha->hw_prof.enabled_phys |= (1 << phy_id);
+			get_lrate_mode(phy, oob_mode);
+			asd_turn_led(asd_ha, phy_id, 1);
+			ASD_DPRINTK("%s: phy%d, lrate:0x%x, proto:0x%x\n",
+				    __FUNCTION__, phy_id,phy->sas_phy.linkrate,
+				    phy->sas_phy.iproto);
+		} else if (oob_status & CURRENT_SPINUP_HOLD) {
+			asd_ha->hw_prof.enabled_phys |= (1 << phy_id);
+			asd_turn_led(asd_ha, phy_id, 1);
+			ASD_DPRINTK("%s: phy%d, spinup hold\n", __FUNCTION__,
+				    phy_id);
+		} else if (oob_status & CURRENT_ERR_MASK) {
+			asd_turn_led(asd_ha, phy_id, 0);
+			ASD_DPRINTK("%s: phy%d: error: oob status:0x%02x\n",
+				    __FUNCTION__, phy_id, oob_status);
+		} else if (oob_status & (CURRENT_HOT_PLUG_CNCT
+					 | CURRENT_DEVICE_PRESENT))  {
+			asd_ha->hw_prof.enabled_phys |= (1 << phy_id);
+			asd_turn_led(asd_ha, phy_id, 1);
+			ASD_DPRINTK("%s: phy%d: hot plug or device present\n",
+				    __FUNCTION__, phy_id);
+		} else {
+			asd_ha->hw_prof.enabled_phys |= (1 << phy_id);
+			asd_turn_led(asd_ha, phy_id, 0);
+			ASD_DPRINTK("%s: phy%d: no device present: "
+				    "oob_status:0x%x\n",
+				    __FUNCTION__, phy_id, oob_status);
+		}
+		break;
+	case RELEASE_SPINUP_HOLD:
+	case PHY_NO_OP:
+	case EXECUTE_HARD_RESET:
+		ASD_DPRINTK("%s: phy%d: sub_func:0x%x\n", __FUNCTION__,
+			    phy_id, control_phy->sub_func);
+		/* XXX finish */
+		break;
+	default:
+		ASD_DPRINTK("%s: phy%d: sub_func:0x%x?\n", __FUNCTION__,
+			    phy_id, control_phy->sub_func);
+		break;
+	}
+out:
+	asd_ascb_free(ascb);
+}
+
+static inline void set_speed_mask(u8 *speed_mask, struct asd_phy_desc *pd)
+{
+	/* disable all speeds, then enable defaults */
+	*speed_mask = SAS_SPEED_60_DIS | SAS_SPEED_30_DIS | SAS_SPEED_15_DIS
+		| SATA_SPEED_30_DIS | SATA_SPEED_15_DIS;
+
+	switch (pd->max_sas_lrate) {
+	case PHY_LINKRATE_6:
+		*speed_mask &= ~SAS_SPEED_60_DIS;
+	default:
+	case PHY_LINKRATE_3:
+		*speed_mask &= ~SAS_SPEED_30_DIS;
+	case PHY_LINKRATE_1_5:
+		*speed_mask &= ~SAS_SPEED_15_DIS;
+	}
+
+	switch (pd->min_sas_lrate) {
+	case PHY_LINKRATE_6:
+		*speed_mask |= SAS_SPEED_30_DIS;
+	case PHY_LINKRATE_3:
+		*speed_mask |= SAS_SPEED_15_DIS;
+	default:
+	case PHY_LINKRATE_1_5:
+		/* nothing to do */
+		;
+	}
+
+	switch (pd->max_sata_lrate) {
+	case PHY_LINKRATE_3:
+		*speed_mask &= ~SATA_SPEED_30_DIS;
+	default:
+	case PHY_LINKRATE_1_5:
+		*speed_mask &= ~SATA_SPEED_15_DIS;
+	}
+
+	switch (pd->min_sata_lrate) {
+	case PHY_LINKRATE_3:
+		*speed_mask |= SATA_SPEED_15_DIS;
+	default:
+	case PHY_LINKRATE_1_5:
+		/* nothing to do */
+		;
+	}
+}
+
+/**
+ * asd_build_control_phy -- build a CONTROL PHY SCB
+ * @ascb: pointer to an ascb
+ * @phy_id: phy id to control, integer
+ * @subfunc: subfunction, what to actually to do the phy
+ *
+ * This function builds a CONTROL PHY scb.  No allocation of any kind
+ * is performed. @ascb is allocated with the list function.
+ * The caller can override the ascb->tasklet_complete to point
+ * to its own callback function.  It must call asd_ascb_free()
+ * at its tasklet complete function.
+ * See the default implementation.
+ */
+void asd_build_control_phy(struct asd_ascb *ascb, int phy_id, u8 subfunc)
+{
+	struct asd_phy *phy = &ascb->ha->phys[phy_id];
+	struct scb *scb = ascb->scb;
+	struct control_phy *control_phy = &scb->control_phy;
+
+	scb->header.opcode = CONTROL_PHY;
+	control_phy->phy_id = (u8) phy_id;
+	control_phy->sub_func = subfunc;
+
+	switch (subfunc) {
+	case EXECUTE_HARD_RESET:  /* 0x81 */
+	case ENABLE_PHY:          /* 0x01 */
+		/* decide hot plug delay */
+		control_phy->hot_plug_delay = HOTPLUG_DELAY_TIMEOUT;
+
+		/* decide speed mask */
+		set_speed_mask(&control_phy->speed_mask, phy->phy_desc);
+
+		/* initiator port settings are in the hi nibble */
+		if (phy->sas_phy.role == PHY_ROLE_INITIATOR)
+			control_phy->port_type = SAS_PROTO_ALL << 4;
+		else if (phy->sas_phy.role == PHY_ROLE_TARGET)
+			control_phy->port_type = SAS_PROTO_ALL;
+		else
+			control_phy->port_type =
+				(SAS_PROTO_ALL << 4) | SAS_PROTO_ALL;
+
+		/* link reset retries, this should be nominal */
+		control_phy->link_reset_retries = 10;
+
+	case RELEASE_SPINUP_HOLD: /* 0x02 */
+		/* decide the func_mask */
+		control_phy->func_mask = FUNCTION_MASK_DEFAULT;
+		if (phy->phy_desc->flags & ASD_SATA_SPINUP_HOLD)
+			control_phy->func_mask &= ~SPINUP_HOLD_DIS;
+		else
+			control_phy->func_mask |= SPINUP_HOLD_DIS;
+	}
+
+	control_phy->conn_handle = cpu_to_le16(0xFFFF);
+
+	ascb->tasklet_complete = control_phy_tasklet_complete;
+}
+
+/* ---------- INITIATE LINK ADM TASK ---------- */
+
+static void link_adm_tasklet_complete(struct asd_ascb *ascb,
+				      struct done_list_struct *dl)
+{
+	u8 opcode = dl->opcode;
+	struct initiate_link_adm *link_adm = &ascb->scb->link_adm;
+	u8 phy_id = link_adm->phy_id;
+
+	if (opcode != TC_NO_ERROR) {
+		asd_printk("phy%d: link adm task 0x%x completed with error "
+			   "0x%x\n", phy_id, link_adm->sub_func, opcode);
+	}
+	ASD_DPRINTK("phy%d: link adm task 0x%x: 0x%x\n",
+		    phy_id, link_adm->sub_func, opcode);
+
+	asd_ascb_free(ascb);
+}
+
+void asd_build_initiate_link_adm_task(struct asd_ascb *ascb, int phy_id,
+				      u8 subfunc)
+{
+	struct scb *scb = ascb->scb;
+	struct initiate_link_adm *link_adm = &scb->link_adm;
+
+	scb->header.opcode = INITIATE_LINK_ADM_TASK;
+
+	link_adm->phy_id = phy_id;
+	link_adm->sub_func = subfunc;
+	link_adm->conn_handle = cpu_to_le16(0xFFFF);
+
+	ascb->tasklet_complete = link_adm_tasklet_complete;
+}
+
+/* ---------- SCB timer ---------- */
+
+/**
+ * asd_ascb_timedout -- called when a pending SCB's timer has expired
+ * @data: unsigned long, a pointer to the ascb in question
+ *
+ * This is the default timeout function which does the most necessary.
+ * Upper layers can implement their own timeout function, say to free
+ * resources they have with this SCB, and then call this one at the
+ * end of their timeout function.  To do this, one should initialize
+ * the ascb->timer.{function, data, expires} prior to calling the post
+ * funcion.  The timer is started by the post function.
+ */
+void asd_ascb_timedout(unsigned long data)
+{
+	struct asd_ascb *ascb = (void *) data;
+	struct asd_seq_data *seq = &ascb->ha->seq;
+	unsigned long flags;
+
+	ASD_DPRINTK("scb:0x%x timed out\n", ascb->scb->header.opcode);
+
+	spin_lock_irqsave(&seq->pend_q_lock, flags);
+	seq->pending--;
+	list_del_init(&ascb->list);
+	spin_unlock_irqrestore(&seq->pend_q_lock, flags);
+
+	asd_ascb_free(ascb);
+}
+
+/* ---------- CONTROL PHY ---------- */
+
+/* Given the spec value, return a driver value. */
+static const int phy_func_table[] = {
+	[PHY_FUNC_NOP]        = PHY_NO_OP,
+	[PHY_FUNC_LINK_RESET] = ENABLE_PHY,
+	[PHY_FUNC_HARD_RESET] = EXECUTE_HARD_RESET,
+	[PHY_FUNC_DISABLE]    = DISABLE_PHY,
+	[PHY_FUNC_RELEASE_SPINUP_HOLD] = RELEASE_SPINUP_HOLD,
+};
+
+int asd_control_phy(struct asd_sas_phy *phy, enum phy_func func)
+{
+	struct asd_ha_struct *asd_ha = phy->ha->lldd_ha;
+	struct asd_ascb *ascb;
+	int res = 1;
+
+	if (func == PHY_FUNC_CLEAR_ERROR_LOG)
+		return -ENOSYS;
+
+	ascb = asd_ascb_alloc_list(asd_ha, &res, GFP_KERNEL);
+	if (!ascb)
+		return -ENOMEM;
+
+	asd_build_control_phy(ascb, phy->id, phy_func_table[func]);
+	res = asd_post_ascb_list(asd_ha, ascb , 1);
+	if (res)
+		asd_ascb_free(ascb);
+
+	return res;
+}
diff --git a/drivers/scsi/aic94xx/aic94xx_sds.c b/drivers/scsi/aic94xx/aic94xx_sds.c
new file mode 100644
index 000000000000..eec1e0db0e0f
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx_sds.c
@@ -0,0 +1,1136 @@
+/*
+ * Aic94xx SAS/SATA driver access to shared data structures and memory
+ * maps.
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#include <linux/pci.h>
+#include <linux/delay.h>
+
+#include "aic94xx.h"
+#include "aic94xx_reg.h"
+
+/* ---------- OCM stuff ---------- */
+
+struct asd_ocm_dir_ent {
+	u8 type;
+	u8 offs[3];
+	u8 _r1;
+	u8 size[3];
+} __attribute__ ((packed));
+
+struct asd_ocm_dir {
+	char sig[2];
+	u8   _r1[2];
+	u8   major;          /* 0 */
+	u8   minor;          /* 0 */
+	u8   _r2;
+	u8   num_de;
+	struct asd_ocm_dir_ent entry[15];
+} __attribute__ ((packed));
+
+#define	OCM_DE_OCM_DIR			0x00
+#define	OCM_DE_WIN_DRVR			0x01
+#define	OCM_DE_BIOS_CHIM		0x02
+#define	OCM_DE_RAID_ENGN		0x03
+#define	OCM_DE_BIOS_INTL		0x04
+#define	OCM_DE_BIOS_CHIM_OSM		0x05
+#define	OCM_DE_BIOS_CHIM_DYNAMIC	0x06
+#define	OCM_DE_ADDC2C_RES0		0x07
+#define	OCM_DE_ADDC2C_RES1		0x08
+#define	OCM_DE_ADDC2C_RES2		0x09
+#define	OCM_DE_ADDC2C_RES3		0x0A
+
+#define OCM_INIT_DIR_ENTRIES	5
+/***************************************************************************
+*  OCM dircetory default
+***************************************************************************/
+static struct asd_ocm_dir OCMDirInit =
+{
+	.sig = {0x4D, 0x4F},	/* signature */
+	.num_de = OCM_INIT_DIR_ENTRIES,	/* no. of directory entries */
+};
+
+/***************************************************************************
+*  OCM dircetory Entries default
+***************************************************************************/
+static struct asd_ocm_dir_ent OCMDirEntriesInit[OCM_INIT_DIR_ENTRIES] =
+{
+	{
+		.type = (OCM_DE_ADDC2C_RES0),	/* Entry type  */
+		.offs = {128},			/* Offset */
+		.size = {0, 4},			/* size */
+	},
+	{
+		.type = (OCM_DE_ADDC2C_RES1),	/* Entry type  */
+		.offs = {128, 4},		/* Offset */
+		.size = {0, 4},			/* size */
+	},
+	{
+		.type = (OCM_DE_ADDC2C_RES2),	/* Entry type  */
+		.offs = {128, 8},		/* Offset */
+		.size = {0, 4},			/* size */
+	},
+	{
+		.type = (OCM_DE_ADDC2C_RES3),	/* Entry type  */
+		.offs = {128, 12},		/* Offset */
+		.size = {0, 4},			/* size */
+	},
+	{
+		.type = (OCM_DE_WIN_DRVR),	/* Entry type  */
+		.offs = {128, 16},		/* Offset */
+		.size = {128, 235, 1},		/* size */
+	},
+};
+
+struct asd_bios_chim_struct {
+	char sig[4];
+	u8   major;          /* 1 */
+	u8   minor;          /* 0 */
+	u8   bios_major;
+	u8   bios_minor;
+	__le32  bios_build;
+	u8   flags;
+	u8   pci_slot;
+	__le16  ue_num;
+	__le16  ue_size;
+	u8  _r[14];
+	/* The unit element array is right here.
+	 */
+} __attribute__ ((packed));
+
+/**
+ * asd_read_ocm_seg - read an on chip memory (OCM) segment
+ * @asd_ha: pointer to the host adapter structure
+ * @buffer: where to write the read data
+ * @offs: offset into OCM where to read from
+ * @size: how many bytes to read
+ *
+ * Return the number of bytes not read. Return 0 on success.
+ */
+static int asd_read_ocm_seg(struct asd_ha_struct *asd_ha, void *buffer,
+			    u32 offs, int size)
+{
+	u8 *p = buffer;
+	if (unlikely(asd_ha->iospace))
+		asd_read_reg_string(asd_ha, buffer, offs+OCM_BASE_ADDR, size);
+	else {
+		for ( ; size > 0; size--, offs++, p++)
+			*p = asd_read_ocm_byte(asd_ha, offs);
+	}
+	return size;
+}
+
+static int asd_read_ocm_dir(struct asd_ha_struct *asd_ha,
+			    struct asd_ocm_dir *dir, u32 offs)
+{
+	int err = asd_read_ocm_seg(asd_ha, dir, offs, sizeof(*dir));
+	if (err) {
+		ASD_DPRINTK("couldn't read ocm segment\n");
+		return err;
+	}
+
+	if (dir->sig[0] != 'M' || dir->sig[1] != 'O') {
+		ASD_DPRINTK("no valid dir signature(%c%c) at start of OCM\n",
+			    dir->sig[0], dir->sig[1]);
+		return -ENOENT;
+	}
+	if (dir->major != 0) {
+		asd_printk("unsupported major version of ocm dir:0x%x\n",
+			   dir->major);
+		return -ENOENT;
+	}
+	dir->num_de &= 0xf;
+	return 0;
+}
+
+/**
+ * asd_write_ocm_seg - write an on chip memory (OCM) segment
+ * @asd_ha: pointer to the host adapter structure
+ * @buffer: where to read the write data
+ * @offs: offset into OCM to write to
+ * @size: how many bytes to write
+ *
+ * Return the number of bytes not written. Return 0 on success.
+ */
+static void asd_write_ocm_seg(struct asd_ha_struct *asd_ha, void *buffer,
+			    u32 offs, int size)
+{
+	u8 *p = buffer;
+	if (unlikely(asd_ha->iospace))
+		asd_write_reg_string(asd_ha, buffer, offs+OCM_BASE_ADDR, size);
+	else {
+		for ( ; size > 0; size--, offs++, p++)
+			asd_write_ocm_byte(asd_ha, offs, *p);
+	}
+	return;
+}
+
+#define THREE_TO_NUM(X) ((X)[0] | ((X)[1] << 8) | ((X)[2] << 16))
+
+static int asd_find_dir_entry(struct asd_ocm_dir *dir, u8 type,
+			      u32 *offs, u32 *size)
+{
+	int i;
+	struct asd_ocm_dir_ent *ent;
+
+	for (i = 0; i < dir->num_de; i++) {
+		if (dir->entry[i].type == type)
+			break;
+	}
+	if (i >= dir->num_de)
+		return -ENOENT;
+	ent = &dir->entry[i];
+	*offs = (u32) THREE_TO_NUM(ent->offs);
+	*size = (u32) THREE_TO_NUM(ent->size);
+	return 0;
+}
+
+#define OCM_BIOS_CHIM_DE  2
+#define BC_BIOS_PRESENT   1
+
+static int asd_get_bios_chim(struct asd_ha_struct *asd_ha,
+			     struct asd_ocm_dir *dir)
+{
+	int err;
+	struct asd_bios_chim_struct *bc_struct;
+	u32 offs, size;
+
+	err = asd_find_dir_entry(dir, OCM_BIOS_CHIM_DE, &offs, &size);
+	if (err) {
+		ASD_DPRINTK("couldn't find BIOS_CHIM dir ent\n");
+		goto out;
+	}
+	err = -ENOMEM;
+	bc_struct = kmalloc(sizeof(*bc_struct), GFP_KERNEL);
+	if (!bc_struct) {
+		asd_printk("no memory for bios_chim struct\n");
+		goto out;
+	}
+	err = asd_read_ocm_seg(asd_ha, (void *)bc_struct, offs,
+			       sizeof(*bc_struct));
+	if (err) {
+		ASD_DPRINTK("couldn't read ocm segment\n");
+		goto out2;
+	}
+	if (strncmp(bc_struct->sig, "SOIB", 4)
+	    && strncmp(bc_struct->sig, "IPSA", 4)) {
+		ASD_DPRINTK("BIOS_CHIM entry has no valid sig(%c%c%c%c)\n",
+			    bc_struct->sig[0], bc_struct->sig[1],
+			    bc_struct->sig[2], bc_struct->sig[3]);
+		err = -ENOENT;
+		goto out2;
+	}
+	if (bc_struct->major != 1) {
+		asd_printk("BIOS_CHIM unsupported major version:0x%x\n",
+			   bc_struct->major);
+		err = -ENOENT;
+		goto out2;
+	}
+	if (bc_struct->flags & BC_BIOS_PRESENT) {
+		asd_ha->hw_prof.bios.present = 1;
+		asd_ha->hw_prof.bios.maj = bc_struct->bios_major;
+		asd_ha->hw_prof.bios.min = bc_struct->bios_minor;
+		asd_ha->hw_prof.bios.bld = le32_to_cpu(bc_struct->bios_build);
+		ASD_DPRINTK("BIOS present (%d,%d), %d\n",
+			    asd_ha->hw_prof.bios.maj,
+			    asd_ha->hw_prof.bios.min,
+			    asd_ha->hw_prof.bios.bld);
+	}
+	asd_ha->hw_prof.ue.num = le16_to_cpu(bc_struct->ue_num);
+	asd_ha->hw_prof.ue.size= le16_to_cpu(bc_struct->ue_size);
+	ASD_DPRINTK("ue num:%d, ue size:%d\n", asd_ha->hw_prof.ue.num,
+		    asd_ha->hw_prof.ue.size);
+	size = asd_ha->hw_prof.ue.num * asd_ha->hw_prof.ue.size;
+	if (size > 0) {
+		err = -ENOMEM;
+		asd_ha->hw_prof.ue.area = kmalloc(size, GFP_KERNEL);
+		if (!asd_ha->hw_prof.ue.area)
+			goto out2;
+		err = asd_read_ocm_seg(asd_ha, (void *)asd_ha->hw_prof.ue.area,
+				       offs + sizeof(*bc_struct), size);
+		if (err) {
+			kfree(asd_ha->hw_prof.ue.area);
+			asd_ha->hw_prof.ue.area = NULL;
+			asd_ha->hw_prof.ue.num  = 0;
+			asd_ha->hw_prof.ue.size = 0;
+			ASD_DPRINTK("couldn't read ue entries(%d)\n", err);
+		}
+	}
+out2:
+	kfree(bc_struct);
+out:
+	return err;
+}
+
+static void
+asd_hwi_initialize_ocm_dir (struct asd_ha_struct *asd_ha)
+{
+	int i;
+
+	/* Zero OCM */
+	for (i = 0; i < OCM_MAX_SIZE; i += 4)
+		asd_write_ocm_dword(asd_ha, i, 0);
+
+	/* Write Dir */
+	asd_write_ocm_seg(asd_ha, &OCMDirInit, 0,
+			  sizeof(struct asd_ocm_dir));
+
+	/* Write Dir Entries */
+	for (i = 0; i < OCM_INIT_DIR_ENTRIES; i++)
+		asd_write_ocm_seg(asd_ha, &OCMDirEntriesInit[i],
+				  sizeof(struct asd_ocm_dir) +
+				  (i * sizeof(struct asd_ocm_dir_ent))
+				  , sizeof(struct asd_ocm_dir_ent));
+
+}
+
+static int
+asd_hwi_check_ocm_access (struct asd_ha_struct *asd_ha)
+{
+	struct pci_dev *pcidev = asd_ha->pcidev;
+	u32 reg;
+	int err = 0;
+	u32 v;
+
+	/* check if OCM has been initialized by BIOS */
+	reg = asd_read_reg_dword(asd_ha, EXSICNFGR);
+
+	if (!(reg & OCMINITIALIZED)) {
+		err = pci_read_config_dword(pcidev, PCIC_INTRPT_STAT, &v);
+		if (err) {
+			asd_printk("couldn't access PCIC_INTRPT_STAT of %s\n",
+					pci_name(pcidev));
+			goto out;
+		}
+
+		printk(KERN_INFO "OCM is not initialized by BIOS,"
+		       "reinitialize it and ignore it, current IntrptStatus"
+		       "is 0x%x\n", v);
+
+		if (v)
+			err = pci_write_config_dword(pcidev,
+						     PCIC_INTRPT_STAT, v);
+		if (err) {
+			asd_printk("couldn't write PCIC_INTRPT_STAT of %s\n",
+					pci_name(pcidev));
+			goto out;
+		}
+
+		asd_hwi_initialize_ocm_dir(asd_ha);
+
+	}
+out:
+	return err;
+}
+
+/**
+ * asd_read_ocm - read on chip memory (OCM)
+ * @asd_ha: pointer to the host adapter structure
+ */
+int asd_read_ocm(struct asd_ha_struct *asd_ha)
+{
+	int err;
+	struct asd_ocm_dir *dir;
+
+	if (asd_hwi_check_ocm_access(asd_ha))
+		return -1;
+
+	dir = kmalloc(sizeof(*dir), GFP_KERNEL);
+	if (!dir) {
+		asd_printk("no memory for ocm dir\n");
+		return -ENOMEM;
+	}
+
+	err = asd_read_ocm_dir(asd_ha, dir, 0);
+	if (err)
+		goto out;
+
+	err = asd_get_bios_chim(asd_ha, dir);
+out:
+	kfree(dir);
+	return err;
+}
+
+/* ---------- FLASH stuff ---------- */
+
+#define FLASH_RESET			0xF0
+#define FLASH_MANUF_AMD                 1
+
+#define FLASH_SIZE                      0x200000
+#define FLASH_DIR_COOKIE                "*** ADAPTEC FLASH DIRECTORY *** "
+#define FLASH_NEXT_ENTRY_OFFS		0x2000
+#define FLASH_MAX_DIR_ENTRIES		32
+
+#define FLASH_DE_TYPE_MASK              0x3FFFFFFF
+#define FLASH_DE_MS                     0x120
+#define FLASH_DE_CTRL_A_USER            0xE0
+
+struct asd_flash_de {
+	__le32   type;
+	__le32   offs;
+	__le32   pad_size;
+	__le32   image_size;
+	__le32   chksum;
+	u8       _r[12];
+	u8       version[32];
+} __attribute__ ((packed));
+
+struct asd_flash_dir {
+	u8    cookie[32];
+	__le32   rev;		  /* 2 */
+	__le32   chksum;
+	__le32   chksum_antidote;
+	__le32   bld;
+	u8    bld_id[32];	  /* build id data */
+	u8    ver_data[32];	  /* date and time of build */
+	__le32   ae_mask;
+	__le32   v_mask;
+	__le32   oc_mask;
+	u8    _r[20];
+	struct asd_flash_de dir_entry[FLASH_MAX_DIR_ENTRIES];
+} __attribute__ ((packed));
+
+struct asd_manuf_sec {
+	char  sig[2];		  /* 'S', 'M' */
+	u16   offs_next;
+	u8    maj;           /* 0 */
+	u8    min;           /* 0 */
+	u16   chksum;
+	u16   size;
+	u8    _r[6];
+	u8    sas_addr[SAS_ADDR_SIZE];
+	u8    pcba_sn[ASD_PCBA_SN_SIZE];
+	/* Here start the other segments */
+	u8    linked_list[0];
+} __attribute__ ((packed));
+
+struct asd_manuf_phy_desc {
+	u8    state;         /* low 4 bits */
+#define MS_PHY_STATE_ENABLEABLE 0
+#define MS_PHY_STATE_REPORTED   1
+#define MS_PHY_STATE_HIDDEN     2
+	u8    phy_id;
+	u16   _r;
+	u8    phy_control_0; /* mode 5 reg 0x160 */
+	u8    phy_control_1; /* mode 5 reg 0x161 */
+	u8    phy_control_2; /* mode 5 reg 0x162 */
+	u8    phy_control_3; /* mode 5 reg 0x163 */
+} __attribute__ ((packed));
+
+struct asd_manuf_phy_param {
+	char  sig[2];		  /* 'P', 'M' */
+	u16   next;
+	u8    maj;           /* 0 */
+	u8    min;           /* 2 */
+	u8    num_phy_desc;  /* 8 */
+	u8    phy_desc_size; /* 8 */
+	u8    _r[3];
+	u8    usage_model_id;
+	u32   _r2;
+	struct asd_manuf_phy_desc phy_desc[ASD_MAX_PHYS];
+} __attribute__ ((packed));
+
+#if 0
+static const char *asd_sb_type[] = {
+	"unknown",
+	"SGPIO",
+	[2 ... 0x7F] = "unknown",
+	[0x80] = "ADPT_I2C",
+	[0x81 ... 0xFF] = "VENDOR_UNIQUExx"
+};
+#endif
+
+struct asd_ms_sb_desc {
+	u8    type;
+	u8    node_desc_index;
+	u8    conn_desc_index;
+	u8    _recvd[0];
+} __attribute__ ((packed));
+
+#if 0
+static const char *asd_conn_type[] = {
+	[0 ... 7] = "unknown",
+	"SFF8470",
+	"SFF8482",
+	"SFF8484",
+	[0x80] = "PCIX_DAUGHTER0",
+	[0x81] = "SAS_DAUGHTER0",
+	[0x82 ... 0xFF] = "VENDOR_UNIQUExx"
+};
+
+static const char *asd_conn_location[] = {
+	"unknown",
+	"internal",
+	"external",
+	"board_to_board",
+};
+#endif
+
+struct asd_ms_conn_desc {
+	u8    type;
+	u8    location;
+	u8    num_sideband_desc;
+	u8    size_sideband_desc;
+	u32   _resvd;
+	u8    name[16];
+	struct asd_ms_sb_desc sb_desc[0];
+} __attribute__ ((packed));
+
+struct asd_nd_phy_desc {
+	u8    vp_attch_type;
+	u8    attch_specific[0];
+} __attribute__ ((packed));
+
+#if 0
+static const char *asd_node_type[] = {
+	"IOP",
+	"IO_CONTROLLER",
+	"EXPANDER",
+	"PORT_MULTIPLIER",
+	"PORT_MULTIPLEXER",
+	"MULTI_DROP_I2C_BUS",
+};
+#endif
+
+struct asd_ms_node_desc {
+	u8    type;
+	u8    num_phy_desc;
+	u8    size_phy_desc;
+	u8    _resvd;
+	u8    name[16];
+	struct asd_nd_phy_desc phy_desc[0];
+} __attribute__ ((packed));
+
+struct asd_ms_conn_map {
+	char  sig[2];		  /* 'M', 'C' */
+	__le16 next;
+	u8    maj;		  /* 0 */
+	u8    min;		  /* 0 */
+	__le16 cm_size;		  /* size of this struct */
+	u8    num_conn;
+	u8    conn_size;
+	u8    num_nodes;
+	u8    usage_model_id;
+	u32   _resvd;
+	struct asd_ms_conn_desc conn_desc[0];
+	struct asd_ms_node_desc node_desc[0];
+} __attribute__ ((packed));
+
+struct asd_ctrla_phy_entry {
+	u8    sas_addr[SAS_ADDR_SIZE];
+	u8    sas_link_rates;  /* max in hi bits, min in low bits */
+	u8    flags;
+	u8    sata_link_rates;
+	u8    _r[5];
+} __attribute__ ((packed));
+
+struct asd_ctrla_phy_settings {
+	u8    id0;		  /* P'h'y */
+	u8    _r;
+	u16   next;
+	u8    num_phys;	      /* number of PHYs in the PCI function */
+	u8    _r2[3];
+	struct asd_ctrla_phy_entry phy_ent[ASD_MAX_PHYS];
+} __attribute__ ((packed));
+
+struct asd_ll_el {
+	u8   id0;
+	u8   id1;
+	__le16  next;
+	u8   something_here[0];
+} __attribute__ ((packed));
+
+static int asd_poll_flash(struct asd_ha_struct *asd_ha)
+{
+	int c;
+	u8 d;
+
+	for (c = 5000; c > 0; c--) {
+		d  = asd_read_reg_byte(asd_ha, asd_ha->hw_prof.flash.bar);
+		d ^= asd_read_reg_byte(asd_ha, asd_ha->hw_prof.flash.bar);
+		if (!d)
+			return 0;
+		udelay(5);
+	}
+	return -ENOENT;
+}
+
+static int asd_reset_flash(struct asd_ha_struct *asd_ha)
+{
+	int err;
+
+	err = asd_poll_flash(asd_ha);
+	if (err)
+		return err;
+	asd_write_reg_byte(asd_ha, asd_ha->hw_prof.flash.bar, FLASH_RESET);
+	err = asd_poll_flash(asd_ha);
+
+	return err;
+}
+
+static inline int asd_read_flash_seg(struct asd_ha_struct *asd_ha,
+				     void *buffer, u32 offs, int size)
+{
+	asd_read_reg_string(asd_ha, buffer, asd_ha->hw_prof.flash.bar+offs,
+			    size);
+	return 0;
+}
+
+/**
+ * asd_find_flash_dir - finds and reads the flash directory
+ * @asd_ha: pointer to the host adapter structure
+ * @flash_dir: pointer to flash directory structure
+ *
+ * If found, the flash directory segment will be copied to
+ * @flash_dir.  Return 1 if found, 0 if not.
+ */
+static int asd_find_flash_dir(struct asd_ha_struct *asd_ha,
+			      struct asd_flash_dir *flash_dir)
+{
+	u32 v;
+	for (v = 0; v < FLASH_SIZE; v += FLASH_NEXT_ENTRY_OFFS) {
+		asd_read_flash_seg(asd_ha, flash_dir, v,
+				   sizeof(FLASH_DIR_COOKIE)-1);
+		if (memcmp(flash_dir->cookie, FLASH_DIR_COOKIE,
+			   sizeof(FLASH_DIR_COOKIE)-1) == 0) {
+			asd_ha->hw_prof.flash.dir_offs = v;
+			asd_read_flash_seg(asd_ha, flash_dir, v,
+					   sizeof(*flash_dir));
+			return 1;
+		}
+	}
+	return 0;
+}
+
+static int asd_flash_getid(struct asd_ha_struct *asd_ha)
+{
+	int err = 0;
+	u32 reg, inc;
+
+	reg = asd_read_reg_dword(asd_ha, EXSICNFGR);
+
+	if (!(reg & FLASHEX)) {
+		ASD_DPRINTK("flash doesn't exist\n");
+		return -ENOENT;
+	}
+	if (pci_read_config_dword(asd_ha->pcidev, PCI_CONF_FLSH_BAR,
+				  &asd_ha->hw_prof.flash.bar)) {
+		asd_printk("couldn't read PCI_CONF_FLSH_BAR of %s\n",
+			   pci_name(asd_ha->pcidev));
+		return -ENOENT;
+	}
+	asd_ha->hw_prof.flash.present = 1;
+	asd_ha->hw_prof.flash.wide = reg & FLASHW ? 1 : 0;
+	err = asd_reset_flash(asd_ha);
+	if (err) {
+		ASD_DPRINTK("couldn't reset flash(%d)\n", err);
+		return err;
+	}
+	/* Get flash info. This would most likely be AMD Am29LV family flash.
+	 * First try the sequence for word mode.  It is the same as for
+	 * 008B (byte mode only), 160B (word mode) and 800D (word mode).
+	 */
+	reg = asd_ha->hw_prof.flash.bar;
+	inc = asd_ha->hw_prof.flash.wide ? 2 : 1;
+	asd_write_reg_byte(asd_ha, reg + 0x555, 0xAA);
+	asd_write_reg_byte(asd_ha, reg + 0x2AA, 0x55);
+	asd_write_reg_byte(asd_ha, reg + 0x555, 0x90);
+	asd_ha->hw_prof.flash.manuf = asd_read_reg_byte(asd_ha, reg);
+	asd_ha->hw_prof.flash.dev_id= asd_read_reg_byte(asd_ha,reg+inc);
+	asd_ha->hw_prof.flash.sec_prot = asd_read_reg_byte(asd_ha,reg+inc+inc);
+	/* Get out of autoselect mode. */
+	err = asd_reset_flash(asd_ha);
+
+	if (asd_ha->hw_prof.flash.manuf == FLASH_MANUF_AMD) {
+		ASD_DPRINTK("0Found FLASH(%d) manuf:%d, dev_id:0x%x, "
+			    "sec_prot:%d\n",
+			    asd_ha->hw_prof.flash.wide ? 16 : 8,
+			    asd_ha->hw_prof.flash.manuf,
+			    asd_ha->hw_prof.flash.dev_id,
+			    asd_ha->hw_prof.flash.sec_prot);
+		return 0;
+	}
+
+	/* Ok, try the sequence for byte mode of 160B and 800D.
+	 * We may actually never need this.
+	 */
+	asd_write_reg_byte(asd_ha, reg + 0xAAA, 0xAA);
+	asd_write_reg_byte(asd_ha, reg + 0x555, 0x55);
+	asd_write_reg_byte(asd_ha, reg + 0xAAA, 0x90);
+	asd_ha->hw_prof.flash.manuf = asd_read_reg_byte(asd_ha, reg);
+	asd_ha->hw_prof.flash.dev_id = asd_read_reg_byte(asd_ha, reg + 2);
+	asd_ha->hw_prof.flash.sec_prot = asd_read_reg_byte(asd_ha, reg + 4);
+	err = asd_reset_flash(asd_ha);
+
+	if (asd_ha->hw_prof.flash.manuf == FLASH_MANUF_AMD) {
+		ASD_DPRINTK("1Found FLASH(%d) manuf:%d, dev_id:0x%x, "
+			    "sec_prot:%d\n",
+			    asd_ha->hw_prof.flash.wide ? 16 : 8,
+			    asd_ha->hw_prof.flash.manuf,
+			    asd_ha->hw_prof.flash.dev_id,
+			    asd_ha->hw_prof.flash.sec_prot);
+		return 0;
+	}
+
+	return -ENOENT;
+}
+
+static u16 asd_calc_flash_chksum(u16 *p, int size)
+{
+	u16 chksum = 0;
+
+	while (size-- > 0)
+		chksum += *p++;
+
+	return chksum;
+}
+
+
+static int asd_find_flash_de(struct asd_flash_dir *flash_dir, u32 entry_type,
+			     u32 *offs, u32 *size)
+{
+	int i;
+	struct asd_flash_de *de;
+
+	for (i = 0; i < FLASH_MAX_DIR_ENTRIES; i++) {
+		u32 type = le32_to_cpu(flash_dir->dir_entry[i].type);
+
+		type &= FLASH_DE_TYPE_MASK;
+		if (type == entry_type)
+			break;
+	}
+	if (i >= FLASH_MAX_DIR_ENTRIES)
+		return -ENOENT;
+	de = &flash_dir->dir_entry[i];
+	*offs = le32_to_cpu(de->offs);
+	*size = le32_to_cpu(de->pad_size);
+	return 0;
+}
+
+static int asd_validate_ms(struct asd_manuf_sec *ms)
+{
+	if (ms->sig[0] != 'S' || ms->sig[1] != 'M') {
+		ASD_DPRINTK("manuf sec: no valid sig(%c%c)\n",
+			    ms->sig[0], ms->sig[1]);
+		return -ENOENT;
+	}
+	if (ms->maj != 0) {
+		asd_printk("unsupported manuf. sector. major version:%x\n",
+			   ms->maj);
+		return -ENOENT;
+	}
+	ms->offs_next = le16_to_cpu((__force __le16) ms->offs_next);
+	ms->chksum = le16_to_cpu((__force __le16) ms->chksum);
+	ms->size = le16_to_cpu((__force __le16) ms->size);
+
+	if (asd_calc_flash_chksum((u16 *)ms, ms->size/2)) {
+		asd_printk("failed manuf sector checksum\n");
+	}
+
+	return 0;
+}
+
+static int asd_ms_get_sas_addr(struct asd_ha_struct *asd_ha,
+			       struct asd_manuf_sec *ms)
+{
+	memcpy(asd_ha->hw_prof.sas_addr, ms->sas_addr, SAS_ADDR_SIZE);
+	return 0;
+}
+
+static int asd_ms_get_pcba_sn(struct asd_ha_struct *asd_ha,
+			      struct asd_manuf_sec *ms)
+{
+	memcpy(asd_ha->hw_prof.pcba_sn, ms->pcba_sn, ASD_PCBA_SN_SIZE);
+	asd_ha->hw_prof.pcba_sn[ASD_PCBA_SN_SIZE] = '\0';
+	return 0;
+}
+
+/**
+ * asd_find_ll_by_id - find a linked list entry by its id
+ * @start: void pointer to the first element in the linked list
+ * @id0: the first byte of the id  (offs 0)
+ * @id1: the second byte of the id (offs 1)
+ *
+ * @start has to be the _base_ element start, since the
+ * linked list entries's offset is from this pointer.
+ * Some linked list entries use only the first id, in which case
+ * you can pass 0xFF for the second.
+ */
+static void *asd_find_ll_by_id(void * const start, const u8 id0, const u8 id1)
+{
+	struct asd_ll_el *el = start;
+
+	do {
+		switch (id1) {
+		default:
+			if (el->id1 == id1)
+		case 0xFF:
+				if (el->id0 == id0)
+					return el;
+		}
+		el = start + le16_to_cpu(el->next);
+	} while (el != start);
+
+	return NULL;
+}
+
+/**
+ * asd_ms_get_phy_params - get phy parameters from the manufacturing sector
+ * @asd_ha: pointer to the host adapter structure
+ * @manuf_sec: pointer to the manufacturing sector
+ *
+ * The manufacturing sector contans also the linked list of sub-segments,
+ * since when it was read, its size was taken from the flash directory,
+ * not from the structure size.
+ *
+ * HIDDEN phys do not count in the total count.  REPORTED phys cannot
+ * be enabled but are reported and counted towards the total.
+ * ENEBLEABLE phys are enabled by default and count towards the total.
+ * The absolute total phy number is ASD_MAX_PHYS.  hw_prof->num_phys
+ * merely specifies the number of phys the host adapter decided to
+ * report.  E.g., it is possible for phys 0, 1 and 2 to be HIDDEN,
+ * phys 3, 4 and 5 to be REPORTED and phys 6 and 7 to be ENEBLEABLE.
+ * In this case ASD_MAX_PHYS is 8, hw_prof->num_phys is 5, and only 2
+ * are actually enabled (enabled by default, max number of phys
+ * enableable in this case).
+ */
+static int asd_ms_get_phy_params(struct asd_ha_struct *asd_ha,
+				 struct asd_manuf_sec *manuf_sec)
+{
+	int i;
+	int en_phys = 0;
+	int rep_phys = 0;
+	struct asd_manuf_phy_param *phy_param;
+	struct asd_manuf_phy_param dflt_phy_param;
+
+	phy_param = asd_find_ll_by_id(manuf_sec, 'P', 'M');
+	if (!phy_param) {
+		ASD_DPRINTK("ms: no phy parameters found\n");
+		ASD_DPRINTK("ms: Creating default phy parameters\n");
+		dflt_phy_param.sig[0] = 'P';
+		dflt_phy_param.sig[1] = 'M';
+		dflt_phy_param.maj = 0;
+		dflt_phy_param.min = 2;
+		dflt_phy_param.num_phy_desc = 8;
+		dflt_phy_param.phy_desc_size = sizeof(struct asd_manuf_phy_desc);
+		for (i =0; i < ASD_MAX_PHYS; i++) {
+			dflt_phy_param.phy_desc[i].state = 0;
+			dflt_phy_param.phy_desc[i].phy_id = i;
+			dflt_phy_param.phy_desc[i].phy_control_0 = 0xf6;
+			dflt_phy_param.phy_desc[i].phy_control_1 = 0x10;
+			dflt_phy_param.phy_desc[i].phy_control_2 = 0x43;
+			dflt_phy_param.phy_desc[i].phy_control_3 = 0xeb;
+		}
+
+		phy_param = &dflt_phy_param;
+
+	}
+
+	if (phy_param->maj != 0) {
+		asd_printk("unsupported manuf. phy param major version:0x%x\n",
+			   phy_param->maj);
+		return -ENOENT;
+	}
+
+	ASD_DPRINTK("ms: num_phy_desc: %d\n", phy_param->num_phy_desc);
+	asd_ha->hw_prof.enabled_phys = 0;
+	for (i = 0; i < phy_param->num_phy_desc; i++) {
+		struct asd_manuf_phy_desc *pd = &phy_param->phy_desc[i];
+		switch (pd->state & 0xF) {
+		case MS_PHY_STATE_HIDDEN:
+			ASD_DPRINTK("ms: phy%d: HIDDEN\n", i);
+			continue;
+		case MS_PHY_STATE_REPORTED:
+			ASD_DPRINTK("ms: phy%d: REPORTED\n", i);
+			asd_ha->hw_prof.enabled_phys &= ~(1 << i);
+			rep_phys++;
+			continue;
+		case MS_PHY_STATE_ENABLEABLE:
+			ASD_DPRINTK("ms: phy%d: ENEBLEABLE\n", i);
+			asd_ha->hw_prof.enabled_phys |= (1 << i);
+			en_phys++;
+			break;
+		}
+		asd_ha->hw_prof.phy_desc[i].phy_control_0 = pd->phy_control_0;
+		asd_ha->hw_prof.phy_desc[i].phy_control_1 = pd->phy_control_1;
+		asd_ha->hw_prof.phy_desc[i].phy_control_2 = pd->phy_control_2;
+		asd_ha->hw_prof.phy_desc[i].phy_control_3 = pd->phy_control_3;
+	}
+	asd_ha->hw_prof.max_phys = rep_phys + en_phys;
+	asd_ha->hw_prof.num_phys = en_phys;
+	ASD_DPRINTK("ms: max_phys:0x%x, num_phys:0x%x\n",
+		    asd_ha->hw_prof.max_phys, asd_ha->hw_prof.num_phys);
+	ASD_DPRINTK("ms: enabled_phys:0x%x\n", asd_ha->hw_prof.enabled_phys);
+	return 0;
+}
+
+static int asd_ms_get_connector_map(struct asd_ha_struct *asd_ha,
+				    struct asd_manuf_sec *manuf_sec)
+{
+	struct asd_ms_conn_map *cm;
+
+	cm = asd_find_ll_by_id(manuf_sec, 'M', 'C');
+	if (!cm) {
+		ASD_DPRINTK("ms: no connector map found\n");
+		return 0;
+	}
+
+	if (cm->maj != 0) {
+		ASD_DPRINTK("ms: unsupported: connector map major version 0x%x"
+			    "\n", cm->maj);
+		return -ENOENT;
+	}
+
+	/* XXX */
+
+	return 0;
+}
+
+
+/**
+ * asd_process_ms - find and extract information from the manufacturing sector
+ * @asd_ha: pointer to the host adapter structure
+ * @flash_dir: pointer to the flash directory
+ */
+static int asd_process_ms(struct asd_ha_struct *asd_ha,
+			  struct asd_flash_dir *flash_dir)
+{
+	int err;
+	struct asd_manuf_sec *manuf_sec;
+	u32 offs, size;
+
+	err = asd_find_flash_de(flash_dir, FLASH_DE_MS, &offs, &size);
+	if (err) {
+		ASD_DPRINTK("Couldn't find the manuf. sector\n");
+		goto out;
+	}
+
+	if (size == 0)
+		goto out;
+
+	err = -ENOMEM;
+	manuf_sec = kmalloc(size, GFP_KERNEL);
+	if (!manuf_sec) {
+		ASD_DPRINTK("no mem for manuf sector\n");
+		goto out;
+	}
+
+	err = asd_read_flash_seg(asd_ha, (void *)manuf_sec, offs, size);
+	if (err) {
+		ASD_DPRINTK("couldn't read manuf sector at 0x%x, size 0x%x\n",
+			    offs, size);
+		goto out2;
+	}
+
+	err = asd_validate_ms(manuf_sec);
+	if (err) {
+		ASD_DPRINTK("couldn't validate manuf sector\n");
+		goto out2;
+	}
+
+	err = asd_ms_get_sas_addr(asd_ha, manuf_sec);
+	if (err) {
+		ASD_DPRINTK("couldn't read the SAS_ADDR\n");
+		goto out2;
+	}
+	ASD_DPRINTK("manuf sect SAS_ADDR %llx\n",
+		    SAS_ADDR(asd_ha->hw_prof.sas_addr));
+
+	err = asd_ms_get_pcba_sn(asd_ha, manuf_sec);
+	if (err) {
+		ASD_DPRINTK("couldn't read the PCBA SN\n");
+		goto out2;
+	}
+	ASD_DPRINTK("manuf sect PCBA SN %s\n", asd_ha->hw_prof.pcba_sn);
+
+	err = asd_ms_get_phy_params(asd_ha, manuf_sec);
+	if (err) {
+		ASD_DPRINTK("ms: couldn't get phy parameters\n");
+		goto out2;
+	}
+
+	err = asd_ms_get_connector_map(asd_ha, manuf_sec);
+	if (err) {
+		ASD_DPRINTK("ms: couldn't get connector map\n");
+		goto out2;
+	}
+
+out2:
+	kfree(manuf_sec);
+out:
+	return err;
+}
+
+static int asd_process_ctrla_phy_settings(struct asd_ha_struct *asd_ha,
+					  struct asd_ctrla_phy_settings *ps)
+{
+	int i;
+	for (i = 0; i < ps->num_phys; i++) {
+		struct asd_ctrla_phy_entry *pe = &ps->phy_ent[i];
+
+		if (!PHY_ENABLED(asd_ha, i))
+			continue;
+		if (*(u64 *)pe->sas_addr == 0) {
+			asd_ha->hw_prof.enabled_phys &= ~(1 << i);
+			continue;
+		}
+		/* This is the SAS address which should be sent in IDENTIFY. */
+		memcpy(asd_ha->hw_prof.phy_desc[i].sas_addr, pe->sas_addr,
+		       SAS_ADDR_SIZE);
+		asd_ha->hw_prof.phy_desc[i].max_sas_lrate =
+			(pe->sas_link_rates & 0xF0) >> 4;
+		asd_ha->hw_prof.phy_desc[i].min_sas_lrate =
+			(pe->sas_link_rates & 0x0F);
+		asd_ha->hw_prof.phy_desc[i].max_sata_lrate =
+			(pe->sata_link_rates & 0xF0) >> 4;
+		asd_ha->hw_prof.phy_desc[i].min_sata_lrate =
+			(pe->sata_link_rates & 0x0F);
+		asd_ha->hw_prof.phy_desc[i].flags = pe->flags;
+		ASD_DPRINTK("ctrla: phy%d: sas_addr: %llx, sas rate:0x%x-0x%x,"
+			    " sata rate:0x%x-0x%x, flags:0x%x\n",
+			    i,
+			    SAS_ADDR(asd_ha->hw_prof.phy_desc[i].sas_addr),
+			    asd_ha->hw_prof.phy_desc[i].max_sas_lrate,
+			    asd_ha->hw_prof.phy_desc[i].min_sas_lrate,
+			    asd_ha->hw_prof.phy_desc[i].max_sata_lrate,
+			    asd_ha->hw_prof.phy_desc[i].min_sata_lrate,
+			    asd_ha->hw_prof.phy_desc[i].flags);
+	}
+
+	return 0;
+}
+
+/**
+ * asd_process_ctrl_a_user - process CTRL-A user settings
+ * @asd_ha: pointer to the host adapter structure
+ * @flash_dir: pointer to the flash directory
+ */
+static int asd_process_ctrl_a_user(struct asd_ha_struct *asd_ha,
+				   struct asd_flash_dir *flash_dir)
+{
+	int err, i;
+	u32 offs, size;
+	struct asd_ll_el *el;
+	struct asd_ctrla_phy_settings *ps;
+	struct asd_ctrla_phy_settings dflt_ps;
+
+	err = asd_find_flash_de(flash_dir, FLASH_DE_CTRL_A_USER, &offs, &size);
+	if (err) {
+		ASD_DPRINTK("couldn't find CTRL-A user settings section\n");
+		ASD_DPRINTK("Creating default CTRL-A user settings section\n");
+
+		dflt_ps.id0 = 'h';
+		dflt_ps.num_phys = 8;
+		for (i =0; i < ASD_MAX_PHYS; i++) {
+			memcpy(dflt_ps.phy_ent[i].sas_addr,
+			       asd_ha->hw_prof.sas_addr, SAS_ADDR_SIZE);
+			dflt_ps.phy_ent[i].sas_link_rates = 0x98;
+			dflt_ps.phy_ent[i].flags = 0x0;
+			dflt_ps.phy_ent[i].sata_link_rates = 0x0;
+		}
+
+		size = sizeof(struct asd_ctrla_phy_settings);
+		ps = &dflt_ps;
+	}
+
+	if (size == 0)
+		goto out;
+
+	err = -ENOMEM;
+	el = kmalloc(size, GFP_KERNEL);
+	if (!el) {
+		ASD_DPRINTK("no mem for ctrla user settings section\n");
+		goto out;
+	}
+
+	err = asd_read_flash_seg(asd_ha, (void *)el, offs, size);
+	if (err) {
+		ASD_DPRINTK("couldn't read ctrla phy settings section\n");
+		goto out2;
+	}
+
+	err = -ENOENT;
+	ps = asd_find_ll_by_id(el, 'h', 0xFF);
+	if (!ps) {
+		ASD_DPRINTK("couldn't find ctrla phy settings struct\n");
+		goto out2;
+	}
+
+	err = asd_process_ctrla_phy_settings(asd_ha, ps);
+	if (err) {
+		ASD_DPRINTK("couldn't process ctrla phy settings\n");
+		goto out2;
+	}
+out2:
+	kfree(el);
+out:
+	return err;
+}
+
+/**
+ * asd_read_flash - read flash memory
+ * @asd_ha: pointer to the host adapter structure
+ */
+int asd_read_flash(struct asd_ha_struct *asd_ha)
+{
+	int err;
+	struct asd_flash_dir *flash_dir;
+
+	err = asd_flash_getid(asd_ha);
+	if (err)
+		return err;
+
+	flash_dir = kmalloc(sizeof(*flash_dir), GFP_KERNEL);
+	if (!flash_dir)
+		return -ENOMEM;
+
+	err = -ENOENT;
+	if (!asd_find_flash_dir(asd_ha, flash_dir)) {
+		ASD_DPRINTK("couldn't find flash directory\n");
+		goto out;
+	}
+
+	if (le32_to_cpu(flash_dir->rev) != 2) {
+		asd_printk("unsupported flash dir version:0x%x\n",
+			   le32_to_cpu(flash_dir->rev));
+		goto out;
+	}
+
+	err = asd_process_ms(asd_ha, flash_dir);
+	if (err) {
+		ASD_DPRINTK("couldn't process manuf sector settings\n");
+		goto out;
+	}
+
+	err = asd_process_ctrl_a_user(asd_ha, flash_dir);
+	if (err) {
+		ASD_DPRINTK("couldn't process CTRL-A user settings\n");
+		goto out;
+	}
+
+out:
+	kfree(flash_dir);
+	return err;
+}
diff --git a/drivers/scsi/aic94xx/aic94xx_seq.c b/drivers/scsi/aic94xx/aic94xx_seq.c
new file mode 100644
index 000000000000..9050c6f3f6bd
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx_seq.c
@@ -0,0 +1,1401 @@
+/*
+ * Aic94xx SAS/SATA driver sequencer interface.
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * Parts of this code adapted from David Chaw's adp94xx_seq.c.
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#include <linux/delay.h>
+#include <linux/pci.h>
+#include <linux/firmware.h>
+#include "aic94xx_reg.h"
+#include "aic94xx_hwi.h"
+
+#include "aic94xx_seq.h"
+#include "aic94xx_dump.h"
+
+/* It takes no more than 0.05 us for an instruction
+ * to complete. So waiting for 1 us should be more than
+ * plenty.
+ */
+#define PAUSE_DELAY 1
+#define PAUSE_TRIES 1000
+
+static const struct firmware *sequencer_fw;
+static const char *sequencer_version;
+static u16 cseq_vecs[CSEQ_NUM_VECS], lseq_vecs[LSEQ_NUM_VECS], mode2_task,
+	cseq_idle_loop, lseq_idle_loop;
+static u8 *cseq_code, *lseq_code;
+static u32 cseq_code_size, lseq_code_size;
+
+static u16 first_scb_site_no = 0xFFFF;
+static u16 last_scb_site_no;
+
+/* ---------- Pause/Unpause CSEQ/LSEQ ---------- */
+
+/**
+ * asd_pause_cseq - pause the central sequencer
+ * @asd_ha: pointer to host adapter structure
+ *
+ * Return 0 on success, negative on failure.
+ */
+int asd_pause_cseq(struct asd_ha_struct *asd_ha)
+{
+	int	count = PAUSE_TRIES;
+	u32	arp2ctl;
+
+	arp2ctl = asd_read_reg_dword(asd_ha, CARP2CTL);
+	if (arp2ctl & PAUSED)
+		return 0;
+
+	asd_write_reg_dword(asd_ha, CARP2CTL, arp2ctl | EPAUSE);
+	do {
+		arp2ctl = asd_read_reg_dword(asd_ha, CARP2CTL);
+		if (arp2ctl & PAUSED)
+			return 0;
+		udelay(PAUSE_DELAY);
+	} while (--count > 0);
+
+	ASD_DPRINTK("couldn't pause CSEQ\n");
+	return -1;
+}
+
+/**
+ * asd_unpause_cseq - unpause the central sequencer.
+ * @asd_ha: pointer to host adapter structure.
+ *
+ * Return 0 on success, negative on error.
+ */
+int asd_unpause_cseq(struct asd_ha_struct *asd_ha)
+{
+	u32	arp2ctl;
+	int	count = PAUSE_TRIES;
+
+	arp2ctl = asd_read_reg_dword(asd_ha, CARP2CTL);
+	if (!(arp2ctl & PAUSED))
+		return 0;
+
+	asd_write_reg_dword(asd_ha, CARP2CTL, arp2ctl & ~EPAUSE);
+	do {
+		arp2ctl = asd_read_reg_dword(asd_ha, CARP2CTL);
+		if (!(arp2ctl & PAUSED))
+			return 0;
+		udelay(PAUSE_DELAY);
+	} while (--count > 0);
+
+	ASD_DPRINTK("couldn't unpause the CSEQ\n");
+	return -1;
+}
+
+/**
+ * asd_seq_pause_lseq - pause a link sequencer
+ * @asd_ha: pointer to a host adapter structure
+ * @lseq: link sequencer of interest
+ *
+ * Return 0 on success, negative on error.
+ */
+static inline int asd_seq_pause_lseq(struct asd_ha_struct *asd_ha, int lseq)
+{
+	u32    arp2ctl;
+	int    count = PAUSE_TRIES;
+
+	arp2ctl = asd_read_reg_dword(asd_ha, LmARP2CTL(lseq));
+	if (arp2ctl & PAUSED)
+		return 0;
+
+	asd_write_reg_dword(asd_ha, LmARP2CTL(lseq), arp2ctl | EPAUSE);
+	do {
+		arp2ctl = asd_read_reg_dword(asd_ha, LmARP2CTL(lseq));
+		if (arp2ctl & PAUSED)
+			return 0;
+		udelay(PAUSE_DELAY);
+	} while (--count > 0);
+
+	ASD_DPRINTK("couldn't pause LSEQ %d\n", lseq);
+	return -1;
+}
+
+/**
+ * asd_pause_lseq - pause the link sequencer(s)
+ * @asd_ha: pointer to host adapter structure
+ * @lseq_mask: mask of link sequencers of interest
+ *
+ * Return 0 on success, negative on failure.
+ */
+int asd_pause_lseq(struct asd_ha_struct *asd_ha, u8 lseq_mask)
+{
+	int lseq;
+	int err = 0;
+
+	for_each_sequencer(lseq_mask, lseq_mask, lseq) {
+		err = asd_seq_pause_lseq(asd_ha, lseq);
+		if (err)
+			return err;
+	}
+
+	return err;
+}
+
+/**
+ * asd_seq_unpause_lseq - unpause a link sequencer
+ * @asd_ha: pointer to host adapter structure
+ * @lseq: link sequencer of interest
+ *
+ * Return 0 on success, negative on error.
+ */
+static inline int asd_seq_unpause_lseq(struct asd_ha_struct *asd_ha, int lseq)
+{
+	u32 arp2ctl;
+	int count = PAUSE_TRIES;
+
+	arp2ctl = asd_read_reg_dword(asd_ha, LmARP2CTL(lseq));
+	if (!(arp2ctl & PAUSED))
+		return 0;
+
+	asd_write_reg_dword(asd_ha, LmARP2CTL(lseq), arp2ctl & ~EPAUSE);
+	do {
+		arp2ctl = asd_read_reg_dword(asd_ha, LmARP2CTL(lseq));
+		if (!(arp2ctl & PAUSED))
+			return 0;
+		udelay(PAUSE_DELAY);
+	} while (--count > 0);
+
+	ASD_DPRINTK("couldn't unpause LSEQ %d\n", lseq);
+	return 0;
+}
+
+
+/**
+ * asd_unpause_lseq - unpause the link sequencer(s)
+ * @asd_ha: pointer to host adapter structure
+ * @lseq_mask: mask of link sequencers of interest
+ *
+ * Return 0 on success, negative on failure.
+ */
+int asd_unpause_lseq(struct asd_ha_struct *asd_ha, u8 lseq_mask)
+{
+	int lseq;
+	int err = 0;
+
+	for_each_sequencer(lseq_mask, lseq_mask, lseq) {
+		err = asd_seq_unpause_lseq(asd_ha, lseq);
+		if (err)
+			return err;
+	}
+
+	return err;
+}
+
+/* ---------- Downloading CSEQ/LSEQ microcode ---------- */
+
+static int asd_verify_cseq(struct asd_ha_struct *asd_ha, const u8 *_prog,
+			   u32 size)
+{
+	u32 addr = CSEQ_RAM_REG_BASE_ADR;
+	const u32 *prog = (u32 *) _prog;
+	u32 i;
+
+	for (i = 0; i < size; i += 4, prog++, addr += 4) {
+		u32 val = asd_read_reg_dword(asd_ha, addr);
+
+		if (le32_to_cpu(*prog) != val) {
+			asd_printk("%s: cseq verify failed at %u "
+				   "read:0x%x, wanted:0x%x\n",
+				   pci_name(asd_ha->pcidev),
+				   i, val, le32_to_cpu(*prog));
+			return -1;
+		}
+	}
+	ASD_DPRINTK("verified %d bytes, passed\n", size);
+	return 0;
+}
+
+/**
+ * asd_verify_lseq - verify the microcode of a link sequencer
+ * @asd_ha: pointer to host adapter structure
+ * @_prog: pointer to the microcode
+ * @size: size of the microcode in bytes
+ * @lseq: link sequencer of interest
+ *
+ * The link sequencer code is accessed in 4 KB pages, which are selected
+ * by setting LmRAMPAGE (bits 8 and 9) of the LmBISTCTL1 register.
+ * The 10 KB LSEQm instruction code is mapped, page at a time, at
+ * LmSEQRAM address.
+ */
+static int asd_verify_lseq(struct asd_ha_struct *asd_ha, const u8 *_prog,
+			   u32 size, int lseq)
+{
+#define LSEQ_CODEPAGE_SIZE 4096
+	int pages =  (size + LSEQ_CODEPAGE_SIZE - 1) / LSEQ_CODEPAGE_SIZE;
+	u32 page;
+	const u32 *prog = (u32 *) _prog;
+
+	for (page = 0; page < pages; page++) {
+		u32 i;
+
+		asd_write_reg_dword(asd_ha, LmBISTCTL1(lseq),
+				    page << LmRAMPAGE_LSHIFT);
+		for (i = 0; size > 0 && i < LSEQ_CODEPAGE_SIZE;
+		     i += 4, prog++, size-=4) {
+
+			u32 val = asd_read_reg_dword(asd_ha, LmSEQRAM(lseq)+i);
+
+			if (le32_to_cpu(*prog) != val) {
+				asd_printk("%s: LSEQ%d verify failed "
+					   "page:%d, offs:%d\n",
+					   pci_name(asd_ha->pcidev),
+					   lseq, page, i);
+				return -1;
+			}
+		}
+	}
+	ASD_DPRINTK("LSEQ%d verified %d bytes, passed\n", lseq,
+		    (int)((u8 *)prog-_prog));
+	return 0;
+}
+
+/**
+ * asd_verify_seq -- verify CSEQ/LSEQ microcode
+ * @asd_ha: pointer to host adapter structure
+ * @prog: pointer to microcode
+ * @size: size of the microcode
+ * @lseq_mask: if 0, verify CSEQ microcode, else mask of LSEQs of interest
+ *
+ * Return 0 if microcode is correct, negative on mismatch.
+ */
+static int asd_verify_seq(struct asd_ha_struct *asd_ha, const u8 *prog,
+			      u32 size, u8 lseq_mask)
+{
+	if (lseq_mask == 0)
+		return asd_verify_cseq(asd_ha, prog, size);
+	else {
+		int lseq, err;
+
+		for_each_sequencer(lseq_mask, lseq_mask, lseq) {
+			err = asd_verify_lseq(asd_ha, prog, size, lseq);
+			if (err)
+				return err;
+		}
+	}
+
+	return 0;
+}
+#define ASD_DMA_MODE_DOWNLOAD
+#ifdef ASD_DMA_MODE_DOWNLOAD
+/* This is the size of the CSEQ Mapped instruction page */
+#define MAX_DMA_OVLY_COUNT ((1U << 14)-1)
+static int asd_download_seq(struct asd_ha_struct *asd_ha,
+			    const u8 * const prog, u32 size, u8 lseq_mask)
+{
+	u32 comstaten;
+	u32 reg;
+	int page;
+	const int pages = (size + MAX_DMA_OVLY_COUNT - 1) / MAX_DMA_OVLY_COUNT;
+	struct asd_dma_tok *token;
+	int err = 0;
+
+	if (size % 4) {
+		asd_printk("sequencer program not multiple of 4\n");
+		return -1;
+	}
+
+	asd_pause_cseq(asd_ha);
+	asd_pause_lseq(asd_ha, 0xFF);
+
+	/* save, disable and clear interrupts */
+	comstaten = asd_read_reg_dword(asd_ha, COMSTATEN);
+	asd_write_reg_dword(asd_ha, COMSTATEN, 0);
+	asd_write_reg_dword(asd_ha, COMSTAT, COMSTAT_MASK);
+
+	asd_write_reg_dword(asd_ha, CHIMINTEN, RST_CHIMINTEN);
+	asd_write_reg_dword(asd_ha, CHIMINT, CHIMINT_MASK);
+
+	token = asd_alloc_coherent(asd_ha, MAX_DMA_OVLY_COUNT, GFP_KERNEL);
+	if (!token) {
+		asd_printk("out of memory for dma SEQ download\n");
+		err = -ENOMEM;
+		goto out;
+	}
+	ASD_DPRINTK("dma-ing %d bytes\n", size);
+
+	for (page = 0; page < pages; page++) {
+		int i;
+		u32 left = min(size-page*MAX_DMA_OVLY_COUNT,
+			       (u32)MAX_DMA_OVLY_COUNT);
+
+		memcpy(token->vaddr, prog + page*MAX_DMA_OVLY_COUNT, left);
+		asd_write_reg_addr(asd_ha, OVLYDMAADR, token->dma_handle);
+		asd_write_reg_dword(asd_ha, OVLYDMACNT, left);
+		reg = !page ? RESETOVLYDMA : 0;
+		reg |= (STARTOVLYDMA | OVLYHALTERR);
+		reg |= (lseq_mask ? (((u32)lseq_mask) << 8) : OVLYCSEQ);
+		/* Start DMA. */
+		asd_write_reg_dword(asd_ha, OVLYDMACTL, reg);
+
+		for (i = PAUSE_TRIES*100; i > 0; i--) {
+			u32 dmadone = asd_read_reg_dword(asd_ha, OVLYDMACTL);
+			if (!(dmadone & OVLYDMAACT))
+				break;
+			udelay(PAUSE_DELAY);
+		}
+	}
+
+	reg = asd_read_reg_dword(asd_ha, COMSTAT);
+	if (!(reg & OVLYDMADONE) || (reg & OVLYERR)
+	    || (asd_read_reg_dword(asd_ha, CHIMINT) & DEVEXCEPT_MASK)){
+		asd_printk("%s: error DMA-ing sequencer code\n",
+			   pci_name(asd_ha->pcidev));
+		err = -ENODEV;
+	}
+
+	asd_free_coherent(asd_ha, token);
+ out:
+	asd_write_reg_dword(asd_ha, COMSTATEN, comstaten);
+
+	return err ? : asd_verify_seq(asd_ha, prog, size, lseq_mask);
+}
+#else /* ASD_DMA_MODE_DOWNLOAD */
+static int asd_download_seq(struct asd_ha_struct *asd_ha, const u8 *_prog,
+			    u32 size, u8 lseq_mask)
+{
+	int i;
+	u32 reg = 0;
+	const u32 *prog = (u32 *) _prog;
+
+	if (size % 4) {
+		asd_printk("sequencer program not multiple of 4\n");
+		return -1;
+	}
+
+	asd_pause_cseq(asd_ha);
+	asd_pause_lseq(asd_ha, 0xFF);
+
+	reg |= (lseq_mask ? (((u32)lseq_mask) << 8) : OVLYCSEQ);
+	reg |= PIOCMODE;
+
+	asd_write_reg_dword(asd_ha, OVLYDMACNT, size);
+	asd_write_reg_dword(asd_ha, OVLYDMACTL, reg);
+
+	ASD_DPRINTK("downloading %s sequencer%s in PIO mode...\n",
+		    lseq_mask ? "LSEQ" : "CSEQ", lseq_mask ? "s" : "");
+
+	for (i = 0; i < size; i += 4, prog++)
+		asd_write_reg_dword(asd_ha, SPIODATA, *prog);
+
+	reg = (reg & ~PIOCMODE) | OVLYHALTERR;
+	asd_write_reg_dword(asd_ha, OVLYDMACTL, reg);
+
+	return asd_verify_seq(asd_ha, _prog, size, lseq_mask);
+}
+#endif /* ASD_DMA_MODE_DOWNLOAD */
+
+/**
+ * asd_seq_download_seqs - download the sequencer microcode
+ * @asd_ha: pointer to host adapter structure
+ *
+ * Download the central and link sequencer microcode.
+ */
+static int asd_seq_download_seqs(struct asd_ha_struct *asd_ha)
+{
+	int 	err;
+
+	if (!asd_ha->hw_prof.enabled_phys) {
+		asd_printk("%s: no enabled phys!\n", pci_name(asd_ha->pcidev));
+		return -ENODEV;
+	}
+
+	/* Download the CSEQ */
+	ASD_DPRINTK("downloading CSEQ...\n");
+	err = asd_download_seq(asd_ha, cseq_code, cseq_code_size, 0);
+	if (err) {
+		asd_printk("CSEQ download failed:%d\n", err);
+		return err;
+	}
+
+	/* Download the Link Sequencers code. All of the Link Sequencers
+	 * microcode can be downloaded at the same time.
+	 */
+	ASD_DPRINTK("downloading LSEQs...\n");
+	err = asd_download_seq(asd_ha, lseq_code, lseq_code_size,
+			       asd_ha->hw_prof.enabled_phys);
+	if (err) {
+		/* Try it one at a time */
+		u8 lseq;
+		u8 lseq_mask = asd_ha->hw_prof.enabled_phys;
+
+		for_each_sequencer(lseq_mask, lseq_mask, lseq) {
+			err = asd_download_seq(asd_ha, lseq_code,
+					       lseq_code_size, 1<<lseq);
+			if (err)
+				break;
+		}
+	}
+	if (err)
+		asd_printk("LSEQs download failed:%d\n", err);
+
+	return err;
+}
+
+/* ---------- Initializing the chip, chip memory, etc. ---------- */
+
+/**
+ * asd_init_cseq_mip - initialize CSEQ mode independent pages 4-7
+ * @asd_ha: pointer to host adapter structure
+ */
+static void asd_init_cseq_mip(struct asd_ha_struct *asd_ha)
+{
+	/* CSEQ Mode Independent, page 4 setup. */
+	asd_write_reg_word(asd_ha, CSEQ_Q_EXE_HEAD, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_Q_EXE_TAIL, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_Q_DONE_HEAD, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_Q_DONE_TAIL, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_Q_SEND_HEAD, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_Q_SEND_TAIL, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_Q_DMA2CHIM_HEAD, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_Q_DMA2CHIM_TAIL, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_Q_COPY_HEAD, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_Q_COPY_TAIL, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_REG0, 0);
+	asd_write_reg_word(asd_ha, CSEQ_REG1, 0);
+	asd_write_reg_dword(asd_ha, CSEQ_REG2, 0);
+	asd_write_reg_byte(asd_ha, CSEQ_LINK_CTL_Q_MAP, 0);
+	{
+		u8 con = asd_read_reg_byte(asd_ha, CCONEXIST);
+		u8 val = hweight8(con);
+		asd_write_reg_byte(asd_ha, CSEQ_MAX_CSEQ_MODE, (val<<4)|val);
+	}
+	asd_write_reg_word(asd_ha, CSEQ_FREE_LIST_HACK_COUNT, 0);
+
+	/* CSEQ Mode independent, page 5 setup. */
+	asd_write_reg_dword(asd_ha, CSEQ_EST_NEXUS_REQ_QUEUE, 0);
+	asd_write_reg_dword(asd_ha, CSEQ_EST_NEXUS_REQ_QUEUE+4, 0);
+	asd_write_reg_dword(asd_ha, CSEQ_EST_NEXUS_REQ_COUNT, 0);
+	asd_write_reg_dword(asd_ha, CSEQ_EST_NEXUS_REQ_COUNT+4, 0);
+	asd_write_reg_word(asd_ha, CSEQ_Q_EST_NEXUS_HEAD, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_Q_EST_NEXUS_TAIL, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_NEED_EST_NEXUS_SCB, 0);
+	asd_write_reg_byte(asd_ha, CSEQ_EST_NEXUS_REQ_HEAD, 0);
+	asd_write_reg_byte(asd_ha, CSEQ_EST_NEXUS_REQ_TAIL, 0);
+	asd_write_reg_byte(asd_ha, CSEQ_EST_NEXUS_SCB_OFFSET, 0);
+
+	/* CSEQ Mode independent, page 6 setup. */
+	asd_write_reg_word(asd_ha, CSEQ_INT_ROUT_RET_ADDR0, 0);
+	asd_write_reg_word(asd_ha, CSEQ_INT_ROUT_RET_ADDR1, 0);
+	asd_write_reg_word(asd_ha, CSEQ_INT_ROUT_SCBPTR, 0);
+	asd_write_reg_byte(asd_ha, CSEQ_INT_ROUT_MODE, 0);
+	asd_write_reg_byte(asd_ha, CSEQ_ISR_SCRATCH_FLAGS, 0);
+	asd_write_reg_word(asd_ha, CSEQ_ISR_SAVE_SINDEX, 0);
+	asd_write_reg_word(asd_ha, CSEQ_ISR_SAVE_DINDEX, 0);
+	asd_write_reg_word(asd_ha, CSEQ_Q_MONIRTT_HEAD, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_Q_MONIRTT_TAIL, 0xFFFF);
+	/* Calculate the free scb mask. */
+	{
+		u16 cmdctx = asd_get_cmdctx_size(asd_ha);
+		cmdctx = (~((cmdctx/128)-1)) >> 8;
+		asd_write_reg_byte(asd_ha, CSEQ_FREE_SCB_MASK, (u8)cmdctx);
+	}
+	asd_write_reg_word(asd_ha, CSEQ_BUILTIN_FREE_SCB_HEAD,
+			   first_scb_site_no);
+	asd_write_reg_word(asd_ha, CSEQ_BUILTIN_FREE_SCB_TAIL,
+			   last_scb_site_no);
+	asd_write_reg_word(asd_ha, CSEQ_EXTENDED_FREE_SCB_HEAD, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_EXTENDED_FREE_SCB_TAIL, 0xFFFF);
+
+	/* CSEQ Mode independent, page 7 setup. */
+	asd_write_reg_dword(asd_ha, CSEQ_EMPTY_REQ_QUEUE, 0);
+	asd_write_reg_dword(asd_ha, CSEQ_EMPTY_REQ_QUEUE+4, 0);
+	asd_write_reg_dword(asd_ha, CSEQ_EMPTY_REQ_COUNT, 0);
+	asd_write_reg_dword(asd_ha, CSEQ_EMPTY_REQ_COUNT+4, 0);
+	asd_write_reg_word(asd_ha, CSEQ_Q_EMPTY_HEAD, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_Q_EMPTY_TAIL, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_NEED_EMPTY_SCB, 0);
+	asd_write_reg_byte(asd_ha, CSEQ_EMPTY_REQ_HEAD, 0);
+	asd_write_reg_byte(asd_ha, CSEQ_EMPTY_REQ_TAIL, 0);
+	asd_write_reg_byte(asd_ha, CSEQ_EMPTY_SCB_OFFSET, 0);
+	asd_write_reg_word(asd_ha, CSEQ_PRIMITIVE_DATA, 0);
+	asd_write_reg_dword(asd_ha, CSEQ_TIMEOUT_CONST, 0);
+}
+
+/**
+ * asd_init_cseq_mdp - initialize CSEQ Mode dependent pages
+ * @asd_ha: pointer to host adapter structure
+ */
+static void asd_init_cseq_mdp(struct asd_ha_struct *asd_ha)
+{
+	int	i;
+	int	moffs;
+
+	moffs = CSEQ_PAGE_SIZE * 2;
+
+	/* CSEQ Mode dependent, modes 0-7, page 0 setup. */
+	for (i = 0; i < 8; i++) {
+		asd_write_reg_word(asd_ha, i*moffs+CSEQ_LRM_SAVE_SINDEX, 0);
+		asd_write_reg_word(asd_ha, i*moffs+CSEQ_LRM_SAVE_SCBPTR, 0);
+		asd_write_reg_word(asd_ha, i*moffs+CSEQ_Q_LINK_HEAD, 0xFFFF);
+		asd_write_reg_word(asd_ha, i*moffs+CSEQ_Q_LINK_TAIL, 0xFFFF);
+		asd_write_reg_byte(asd_ha, i*moffs+CSEQ_LRM_SAVE_SCRPAGE, 0);
+	}
+
+	/* CSEQ Mode dependent, mode 0-7, page 1 and 2 shall be ignored. */
+
+	/* CSEQ Mode dependent, mode 8, page 0 setup. */
+	asd_write_reg_word(asd_ha, CSEQ_RET_ADDR, 0xFFFF);
+	asd_write_reg_word(asd_ha, CSEQ_RET_SCBPTR, 0);
+	asd_write_reg_word(asd_ha, CSEQ_SAVE_SCBPTR, 0);
+	asd_write_reg_word(asd_ha, CSEQ_EMPTY_TRANS_CTX, 0);
+	asd_write_reg_word(asd_ha, CSEQ_RESP_LEN, 0);
+	asd_write_reg_word(asd_ha, CSEQ_TMF_SCBPTR, 0);
+	asd_write_reg_word(asd_ha, CSEQ_GLOBAL_PREV_SCB, 0);
+	asd_write_reg_word(asd_ha, CSEQ_GLOBAL_HEAD, 0);
+	asd_write_reg_word(asd_ha, CSEQ_CLEAR_LU_HEAD, 0);
+	asd_write_reg_byte(asd_ha, CSEQ_TMF_OPCODE, 0);
+	asd_write_reg_byte(asd_ha, CSEQ_SCRATCH_FLAGS, 0);
+	asd_write_reg_word(asd_ha, CSEQ_HSB_SITE, 0);
+	asd_write_reg_word(asd_ha, CSEQ_FIRST_INV_SCB_SITE,
+			   (u16)last_scb_site_no+1);
+	asd_write_reg_word(asd_ha, CSEQ_FIRST_INV_DDB_SITE,
+			   (u16)asd_ha->hw_prof.max_ddbs);
+
+	/* CSEQ Mode dependent, mode 8, page 1 setup. */
+	asd_write_reg_dword(asd_ha, CSEQ_LUN_TO_CLEAR, 0);
+	asd_write_reg_dword(asd_ha, CSEQ_LUN_TO_CLEAR + 4, 0);
+	asd_write_reg_dword(asd_ha, CSEQ_LUN_TO_CHECK, 0);
+	asd_write_reg_dword(asd_ha, CSEQ_LUN_TO_CHECK + 4, 0);
+
+	/* CSEQ Mode dependent, mode 8, page 2 setup. */
+	/* Tell the sequencer the bus address of the first SCB. */
+	asd_write_reg_addr(asd_ha, CSEQ_HQ_NEW_POINTER,
+			   asd_ha->seq.next_scb.dma_handle);
+	ASD_DPRINTK("First SCB dma_handle: 0x%llx\n",
+		    (unsigned long long)asd_ha->seq.next_scb.dma_handle);
+
+	/* Tell the sequencer the first Done List entry address. */
+	asd_write_reg_addr(asd_ha, CSEQ_HQ_DONE_BASE,
+			   asd_ha->seq.actual_dl->dma_handle);
+
+	/* Initialize the Q_DONE_POINTER with the least significant
+	 * 4 bytes of the first Done List address. */
+	asd_write_reg_dword(asd_ha, CSEQ_HQ_DONE_POINTER,
+			    ASD_BUSADDR_LO(asd_ha->seq.actual_dl->dma_handle));
+
+	asd_write_reg_byte(asd_ha, CSEQ_HQ_DONE_PASS, ASD_DEF_DL_TOGGLE);
+
+	/* CSEQ Mode dependent, mode 8, page 3 shall be ignored. */
+}
+
+/**
+ * asd_init_cseq_scratch -- setup and init CSEQ
+ * @asd_ha: pointer to host adapter structure
+ *
+ * Setup and initialize Central sequencers. Initialiaze the mode
+ * independent and dependent scratch page to the default settings.
+ */
+static void asd_init_cseq_scratch(struct asd_ha_struct *asd_ha)
+{
+	asd_init_cseq_mip(asd_ha);
+	asd_init_cseq_mdp(asd_ha);
+}
+
+/**
+ * asd_init_lseq_mip -- initialize LSEQ Mode independent pages 0-3
+ * @asd_ha: pointer to host adapter structure
+ */
+static void asd_init_lseq_mip(struct asd_ha_struct *asd_ha, u8 lseq)
+{
+	int i;
+
+	/* LSEQ Mode independent page 0 setup. */
+	asd_write_reg_word(asd_ha, LmSEQ_Q_TGTXFR_HEAD(lseq), 0xFFFF);
+	asd_write_reg_word(asd_ha, LmSEQ_Q_TGTXFR_TAIL(lseq), 0xFFFF);
+	asd_write_reg_byte(asd_ha, LmSEQ_LINK_NUMBER(lseq), lseq);
+	asd_write_reg_byte(asd_ha, LmSEQ_SCRATCH_FLAGS(lseq),
+			   ASD_NOTIFY_ENABLE_SPINUP);
+	asd_write_reg_dword(asd_ha, LmSEQ_CONNECTION_STATE(lseq),0x08000000);
+	asd_write_reg_word(asd_ha, LmSEQ_CONCTL(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_CONSTAT(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_CONNECTION_MODES(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_REG1_ISR(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_REG2_ISR(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_REG3_ISR(lseq), 0);
+	asd_write_reg_dword(asd_ha, LmSEQ_REG0_ISR(lseq), 0);
+	asd_write_reg_dword(asd_ha, LmSEQ_REG0_ISR(lseq)+4, 0);
+
+	/* LSEQ Mode independent page 1 setup. */
+	asd_write_reg_word(asd_ha, LmSEQ_EST_NEXUS_SCBPTR0(lseq), 0xFFFF);
+	asd_write_reg_word(asd_ha, LmSEQ_EST_NEXUS_SCBPTR1(lseq), 0xFFFF);
+	asd_write_reg_word(asd_ha, LmSEQ_EST_NEXUS_SCBPTR2(lseq), 0xFFFF);
+	asd_write_reg_word(asd_ha, LmSEQ_EST_NEXUS_SCBPTR3(lseq), 0xFFFF);
+	asd_write_reg_byte(asd_ha, LmSEQ_EST_NEXUS_SCB_OPCODE0(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_EST_NEXUS_SCB_OPCODE1(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_EST_NEXUS_SCB_OPCODE2(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_EST_NEXUS_SCB_OPCODE3(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_EST_NEXUS_SCB_HEAD(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_EST_NEXUS_SCB_TAIL(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_EST_NEXUS_BUF_AVAIL(lseq), 0);
+	asd_write_reg_dword(asd_ha, LmSEQ_TIMEOUT_CONST(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_ISR_SAVE_SINDEX(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_ISR_SAVE_DINDEX(lseq), 0);
+
+	/* LSEQ Mode Independent page 2 setup. */
+	asd_write_reg_word(asd_ha, LmSEQ_EMPTY_SCB_PTR0(lseq), 0xFFFF);
+	asd_write_reg_word(asd_ha, LmSEQ_EMPTY_SCB_PTR1(lseq), 0xFFFF);
+	asd_write_reg_word(asd_ha, LmSEQ_EMPTY_SCB_PTR2(lseq), 0xFFFF);
+	asd_write_reg_word(asd_ha, LmSEQ_EMPTY_SCB_PTR3(lseq), 0xFFFF);
+	asd_write_reg_byte(asd_ha, LmSEQ_EMPTY_SCB_OPCD0(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_EMPTY_SCB_OPCD1(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_EMPTY_SCB_OPCD2(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_EMPTY_SCB_OPCD3(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_EMPTY_SCB_HEAD(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_EMPTY_SCB_TAIL(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_EMPTY_BUFS_AVAIL(lseq), 0);
+	for (i = 0; i < 12; i += 4)
+		asd_write_reg_dword(asd_ha, LmSEQ_ATA_SCR_REGS(lseq) + i, 0);
+
+	/* LSEQ Mode Independent page 3 setup. */
+
+	/* Device present timer timeout */
+	asd_write_reg_dword(asd_ha, LmSEQ_DEV_PRES_TMR_TOUT_CONST(lseq),
+			    ASD_DEV_PRESENT_TIMEOUT);
+
+	/* SATA interlock timer disabled */
+	asd_write_reg_dword(asd_ha, LmSEQ_SATA_INTERLOCK_TIMEOUT(lseq),
+			    ASD_SATA_INTERLOCK_TIMEOUT);
+
+	/* STP shutdown timer timeout constant, IGNORED by the sequencer,
+	 * always 0. */
+	asd_write_reg_dword(asd_ha, LmSEQ_STP_SHUTDOWN_TIMEOUT(lseq),
+			    ASD_STP_SHUTDOWN_TIMEOUT);
+
+	asd_write_reg_dword(asd_ha, LmSEQ_SRST_ASSERT_TIMEOUT(lseq),
+			    ASD_SRST_ASSERT_TIMEOUT);
+
+	asd_write_reg_dword(asd_ha, LmSEQ_RCV_FIS_TIMEOUT(lseq),
+			    ASD_RCV_FIS_TIMEOUT);
+
+	asd_write_reg_dword(asd_ha, LmSEQ_ONE_MILLISEC_TIMEOUT(lseq),
+			    ASD_ONE_MILLISEC_TIMEOUT);
+
+	/* COM_INIT timer */
+	asd_write_reg_dword(asd_ha, LmSEQ_TEN_MS_COMINIT_TIMEOUT(lseq),
+			    ASD_TEN_MILLISEC_TIMEOUT);
+
+	asd_write_reg_dword(asd_ha, LmSEQ_SMP_RCV_TIMEOUT(lseq),
+			    ASD_SMP_RCV_TIMEOUT);
+}
+
+/**
+ * asd_init_lseq_mdp -- initialize LSEQ mode dependent pages.
+ * @asd_ha: pointer to host adapter structure
+ */
+static void asd_init_lseq_mdp(struct asd_ha_struct *asd_ha,  int lseq)
+{
+	int    i;
+	u32    moffs;
+	u16 ret_addr[] = {
+		0xFFFF,		  /* mode 0 */
+		0xFFFF,		  /* mode 1 */
+		mode2_task,	  /* mode 2 */
+		0,
+		0xFFFF,		  /* mode 4/5 */
+		0xFFFF,		  /* mode 4/5 */
+	};
+
+	/*
+	 * Mode 0,1,2 and 4/5 have common field on page 0 for the first
+	 * 14 bytes.
+	 */
+	for (i = 0; i < 3; i++) {
+		moffs = i * LSEQ_MODE_SCRATCH_SIZE;
+		asd_write_reg_word(asd_ha, LmSEQ_RET_ADDR(lseq)+moffs,
+				   ret_addr[i]);
+		asd_write_reg_word(asd_ha, LmSEQ_REG0_MODE(lseq)+moffs, 0);
+		asd_write_reg_word(asd_ha, LmSEQ_MODE_FLAGS(lseq)+moffs, 0);
+		asd_write_reg_word(asd_ha, LmSEQ_RET_ADDR2(lseq)+moffs,0xFFFF);
+		asd_write_reg_word(asd_ha, LmSEQ_RET_ADDR1(lseq)+moffs,0xFFFF);
+		asd_write_reg_byte(asd_ha, LmSEQ_OPCODE_TO_CSEQ(lseq)+moffs,0);
+		asd_write_reg_word(asd_ha, LmSEQ_DATA_TO_CSEQ(lseq)+moffs,0);
+	}
+	/*
+	 *  Mode 5 page 0 overlaps the same scratch page with Mode 0 page 3.
+	 */
+	asd_write_reg_word(asd_ha,
+			 LmSEQ_RET_ADDR(lseq)+LSEQ_MODE5_PAGE0_OFFSET,
+			   ret_addr[5]);
+	asd_write_reg_word(asd_ha,
+			 LmSEQ_REG0_MODE(lseq)+LSEQ_MODE5_PAGE0_OFFSET,0);
+	asd_write_reg_word(asd_ha,
+			 LmSEQ_MODE_FLAGS(lseq)+LSEQ_MODE5_PAGE0_OFFSET, 0);
+	asd_write_reg_word(asd_ha,
+			 LmSEQ_RET_ADDR2(lseq)+LSEQ_MODE5_PAGE0_OFFSET,0xFFFF);
+	asd_write_reg_word(asd_ha,
+			 LmSEQ_RET_ADDR1(lseq)+LSEQ_MODE5_PAGE0_OFFSET,0xFFFF);
+	asd_write_reg_byte(asd_ha,
+		         LmSEQ_OPCODE_TO_CSEQ(lseq)+LSEQ_MODE5_PAGE0_OFFSET,0);
+	asd_write_reg_word(asd_ha,
+		         LmSEQ_DATA_TO_CSEQ(lseq)+LSEQ_MODE5_PAGE0_OFFSET, 0);
+
+	/* LSEQ Mode dependent 0, page 0 setup. */
+	asd_write_reg_word(asd_ha, LmSEQ_FIRST_INV_DDB_SITE(lseq),
+			   (u16)asd_ha->hw_prof.max_ddbs);
+	asd_write_reg_word(asd_ha, LmSEQ_EMPTY_TRANS_CTX(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_RESP_LEN(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_FIRST_INV_SCB_SITE(lseq),
+			   (u16)last_scb_site_no+1);
+	asd_write_reg_word(asd_ha, LmSEQ_INTEN_SAVE(lseq),
+			    (u16) LmM0INTEN_MASK & 0xFFFF0000 >> 16);
+	asd_write_reg_word(asd_ha, LmSEQ_INTEN_SAVE(lseq) + 2,
+			    (u16) LmM0INTEN_MASK & 0xFFFF);
+	asd_write_reg_byte(asd_ha, LmSEQ_LINK_RST_FRM_LEN(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_LINK_RST_PROTOCOL(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_RESP_STATUS(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_LAST_LOADED_SGE(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_SAVE_SCBPTR(lseq), 0);
+
+	/* LSEQ mode dependent, mode 1, page 0 setup. */
+	asd_write_reg_word(asd_ha, LmSEQ_Q_XMIT_HEAD(lseq), 0xFFFF);
+	asd_write_reg_word(asd_ha, LmSEQ_M1_EMPTY_TRANS_CTX(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_INI_CONN_TAG(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_FAILED_OPEN_STATUS(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_XMIT_REQUEST_TYPE(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_M1_RESP_STATUS(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_M1_LAST_LOADED_SGE(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_M1_SAVE_SCBPTR(lseq), 0);
+
+	/* LSEQ Mode dependent mode 2, page 0 setup */
+	asd_write_reg_word(asd_ha, LmSEQ_PORT_COUNTER(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_PM_TABLE_PTR(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_SATA_INTERLOCK_TMR_SAVE(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_IP_BITL(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_COPY_SMP_CONN_TAG(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_P0M2_OFFS1AH(lseq), 0);
+
+	/* LSEQ Mode dependent, mode 4/5, page 0 setup. */
+	asd_write_reg_byte(asd_ha, LmSEQ_SAVED_OOB_STATUS(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_SAVED_OOB_MODE(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_Q_LINK_HEAD(lseq), 0xFFFF);
+	asd_write_reg_byte(asd_ha, LmSEQ_LINK_RST_ERR(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_SAVED_OOB_SIGNALS(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_SAS_RESET_MODE(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_LINK_RESET_RETRY_COUNT(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_NUM_LINK_RESET_RETRIES(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_OOB_INT_ENABLES(lseq), 0);
+	/*
+	 * Set the desired interval between transmissions of the NOTIFY
+	 * (ENABLE SPINUP) primitive.  Must be initilized to val - 1.
+	 */
+	asd_write_reg_word(asd_ha, LmSEQ_NOTIFY_TIMER_TIMEOUT(lseq),
+			   ASD_NOTIFY_TIMEOUT - 1);
+	/* No delay for the first NOTIFY to be sent to the attached target. */
+	asd_write_reg_word(asd_ha, LmSEQ_NOTIFY_TIMER_DOWN_COUNT(lseq),
+			   ASD_NOTIFY_DOWN_COUNT);
+
+	/* LSEQ Mode dependent, mode 0 and 1, page 1 setup. */
+	for (i = 0; i < 2; i++)	{
+		int j;
+		/* Start from Page 1 of Mode 0 and 1. */
+		moffs = LSEQ_PAGE_SIZE + i*LSEQ_MODE_SCRATCH_SIZE;
+		/* All the fields of page 1 can be intialized to 0. */
+		for (j = 0; j < LSEQ_PAGE_SIZE; j += 4)
+			asd_write_reg_dword(asd_ha, LmSCRATCH(lseq)+moffs+j,0);
+	}
+
+	/* LSEQ Mode dependent, mode 2, page 1 setup. */
+	asd_write_reg_dword(asd_ha, LmSEQ_INVALID_DWORD_COUNT(lseq), 0);
+	asd_write_reg_dword(asd_ha, LmSEQ_DISPARITY_ERROR_COUNT(lseq), 0);
+	asd_write_reg_dword(asd_ha, LmSEQ_LOSS_OF_SYNC_COUNT(lseq), 0);
+
+	/* LSEQ Mode dependent, mode 4/5, page 1. */
+	for (i = 0; i < LSEQ_PAGE_SIZE; i+=4)
+		asd_write_reg_dword(asd_ha, LmSEQ_FRAME_TYPE_MASK(lseq)+i, 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_FRAME_TYPE_MASK(lseq), 0xFF);
+	asd_write_reg_byte(asd_ha, LmSEQ_HASHED_DEST_ADDR_MASK(lseq), 0xFF);
+	asd_write_reg_byte(asd_ha, LmSEQ_HASHED_DEST_ADDR_MASK(lseq)+1,0xFF);
+	asd_write_reg_byte(asd_ha, LmSEQ_HASHED_DEST_ADDR_MASK(lseq)+2,0xFF);
+	asd_write_reg_byte(asd_ha, LmSEQ_HASHED_SRC_ADDR_MASK(lseq), 0xFF);
+	asd_write_reg_byte(asd_ha, LmSEQ_HASHED_SRC_ADDR_MASK(lseq)+1, 0xFF);
+	asd_write_reg_byte(asd_ha, LmSEQ_HASHED_SRC_ADDR_MASK(lseq)+2, 0xFF);
+	asd_write_reg_dword(asd_ha, LmSEQ_DATA_OFFSET(lseq), 0xFFFFFFFF);
+
+	/* LSEQ Mode dependent, mode 0, page 2 setup. */
+	asd_write_reg_dword(asd_ha, LmSEQ_SMP_RCV_TIMER_TERM_TS(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_DEVICE_BITS(lseq), 0);
+	asd_write_reg_word(asd_ha, LmSEQ_SDB_DDB(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_SDB_NUM_TAGS(lseq), 0);
+	asd_write_reg_byte(asd_ha, LmSEQ_SDB_CURR_TAG(lseq), 0);
+
+	/* LSEQ Mode Dependent 1, page 2 setup. */
+	asd_write_reg_dword(asd_ha, LmSEQ_TX_ID_ADDR_FRAME(lseq), 0);
+	asd_write_reg_dword(asd_ha, LmSEQ_TX_ID_ADDR_FRAME(lseq)+4, 0);
+	asd_write_reg_dword(asd_ha, LmSEQ_OPEN_TIMER_TERM_TS(lseq), 0);
+	asd_write_reg_dword(asd_ha, LmSEQ_SRST_AS_TIMER_TERM_TS(lseq), 0);
+	asd_write_reg_dword(asd_ha, LmSEQ_LAST_LOADED_SG_EL(lseq), 0);
+
+	/* LSEQ Mode Dependent 2, page 2 setup. */
+	/* The LmSEQ_STP_SHUTDOWN_TIMER_TERM_TS is IGNORED by the sequencer,
+	 * i.e. always 0. */
+	asd_write_reg_dword(asd_ha, LmSEQ_STP_SHUTDOWN_TIMER_TERM_TS(lseq),0);
+	asd_write_reg_dword(asd_ha, LmSEQ_CLOSE_TIMER_TERM_TS(lseq), 0);
+	asd_write_reg_dword(asd_ha, LmSEQ_BREAK_TIMER_TERM_TS(lseq), 0);
+	asd_write_reg_dword(asd_ha, LmSEQ_DWS_RESET_TIMER_TERM_TS(lseq), 0);
+	asd_write_reg_dword(asd_ha,LmSEQ_SATA_INTERLOCK_TIMER_TERM_TS(lseq),0);
+	asd_write_reg_dword(asd_ha, LmSEQ_MCTL_TIMER_TERM_TS(lseq), 0);
+
+	/* LSEQ Mode Dependent 4/5, page 2 setup. */
+	asd_write_reg_dword(asd_ha, LmSEQ_COMINIT_TIMER_TERM_TS(lseq), 0);
+	asd_write_reg_dword(asd_ha, LmSEQ_RCV_ID_TIMER_TERM_TS(lseq), 0);
+	asd_write_reg_dword(asd_ha, LmSEQ_RCV_FIS_TIMER_TERM_TS(lseq), 0);
+	asd_write_reg_dword(asd_ha, LmSEQ_DEV_PRES_TIMER_TERM_TS(lseq),	0);
+}
+
+/**
+ * asd_init_lseq_scratch -- setup and init link sequencers
+ * @asd_ha: pointer to host adapter struct
+ */
+static void asd_init_lseq_scratch(struct asd_ha_struct *asd_ha)
+{
+	u8 lseq;
+	u8 lseq_mask;
+
+	lseq_mask = asd_ha->hw_prof.enabled_phys;
+	for_each_sequencer(lseq_mask, lseq_mask, lseq) {
+		asd_init_lseq_mip(asd_ha, lseq);
+		asd_init_lseq_mdp(asd_ha, lseq);
+	}
+}
+
+/**
+ * asd_init_scb_sites -- initialize sequencer SCB sites (memory).
+ * @asd_ha: pointer to host adapter structure
+ *
+ * This should be done before initializing common CSEQ and LSEQ
+ * scratch since those areas depend on some computed values here,
+ * last_scb_site_no, etc.
+ */
+static void asd_init_scb_sites(struct asd_ha_struct *asd_ha)
+{
+	u16	site_no;
+	u16     max_scbs = 0;
+
+	for (site_no = asd_ha->hw_prof.max_scbs-1;
+	     site_no != (u16) -1;
+	     site_no--) {
+		u16	i;
+
+		/* Initialize all fields in the SCB site to 0. */
+		for (i = 0; i < ASD_SCB_SIZE; i += 4)
+			asd_scbsite_write_dword(asd_ha, site_no, i, 0);
+
+		/* Workaround needed by SEQ to fix a SATA issue is to exclude
+		 * certain SCB sites from the free list. */
+		if (!SCB_SITE_VALID(site_no))
+			continue;
+
+		if (last_scb_site_no == 0)
+			last_scb_site_no = site_no;
+
+		/* For every SCB site, we need to initialize the
+		 * following fields: Q_NEXT, SCB_OPCODE, SCB_FLAGS,
+		 * and SG Element Flag. */
+
+		/* Q_NEXT field of the last SCB is invalidated. */
+		asd_scbsite_write_word(asd_ha, site_no, 0, first_scb_site_no);
+
+		/* Initialize SCB Site Opcode field to invalid. */
+		asd_scbsite_write_byte(asd_ha, site_no,
+				       offsetof(struct scb_header, opcode),
+				       0xFF);
+
+		/* Initialize SCB Site Flags field to mean a response
+		 * frame has been received.  This means inadvertent
+		 * frames received to be dropped. */
+		asd_scbsite_write_byte(asd_ha, site_no, 0x49, 0x01);
+
+		first_scb_site_no = site_no;
+		max_scbs++;
+	}
+	asd_ha->hw_prof.max_scbs = max_scbs;
+	ASD_DPRINTK("max_scbs:%d\n", asd_ha->hw_prof.max_scbs);
+	ASD_DPRINTK("first_scb_site_no:0x%x\n", first_scb_site_no);
+	ASD_DPRINTK("last_scb_site_no:0x%x\n", last_scb_site_no);
+}
+
+/**
+ * asd_init_cseq_cio - initialize CSEQ CIO registers
+ * @asd_ha: pointer to host adapter structure
+ */
+static void asd_init_cseq_cio(struct asd_ha_struct *asd_ha)
+{
+	int i;
+
+	asd_write_reg_byte(asd_ha, CSEQCOMINTEN, 0);
+	asd_write_reg_byte(asd_ha, CSEQDLCTL, ASD_DL_SIZE_BITS);
+	asd_write_reg_byte(asd_ha, CSEQDLOFFS, 0);
+	asd_write_reg_byte(asd_ha, CSEQDLOFFS+1, 0);
+	asd_ha->seq.scbpro = 0;
+	asd_write_reg_dword(asd_ha, SCBPRO, 0);
+	asd_write_reg_dword(asd_ha, CSEQCON, 0);
+
+	/* Intialize CSEQ Mode 11 Interrupt Vectors.
+	 * The addresses are 16 bit wide and in dword units.
+	 * The values of their macros are in byte units.
+	 * Thus we have to divide by 4. */
+	asd_write_reg_word(asd_ha, CM11INTVEC0, cseq_vecs[0]);
+	asd_write_reg_word(asd_ha, CM11INTVEC1, cseq_vecs[1]);
+	asd_write_reg_word(asd_ha, CM11INTVEC2, cseq_vecs[2]);
+
+	/* Enable ARP2HALTC (ARP2 Halted from Halt Code Write). */
+	asd_write_reg_byte(asd_ha, CARP2INTEN, EN_ARP2HALTC);
+
+	/* Initialize CSEQ Scratch Page to 0x04. */
+	asd_write_reg_byte(asd_ha, CSCRATCHPAGE, 0x04);
+
+	/* Initialize CSEQ Mode[0-8] Dependent registers. */
+	/* Initialize Scratch Page to 0. */
+	for (i = 0; i < 9; i++)
+		asd_write_reg_byte(asd_ha, CMnSCRATCHPAGE(i), 0);
+
+	/* Reset the ARP2 Program Count. */
+	asd_write_reg_word(asd_ha, CPRGMCNT, cseq_idle_loop);
+
+	for (i = 0; i < 8; i++) {
+		/* Intialize Mode n Link m Interrupt Enable. */
+		asd_write_reg_dword(asd_ha, CMnINTEN(i), EN_CMnRSPMBXF);
+		/* Initialize Mode n Request Mailbox. */
+		asd_write_reg_dword(asd_ha, CMnREQMBX(i), 0);
+	}
+}
+
+/**
+ * asd_init_lseq_cio -- initialize LmSEQ CIO registers
+ * @asd_ha: pointer to host adapter structure
+ */
+static void asd_init_lseq_cio(struct asd_ha_struct *asd_ha, int lseq)
+{
+	u8  *sas_addr;
+	int  i;
+
+	/* Enable ARP2HALTC (ARP2 Halted from Halt Code Write). */
+	asd_write_reg_dword(asd_ha, LmARP2INTEN(lseq), EN_ARP2HALTC);
+
+	asd_write_reg_byte(asd_ha, LmSCRATCHPAGE(lseq), 0);
+
+	/* Initialize Mode 0,1, and 2 SCRATCHPAGE to 0. */
+	for (i = 0; i < 3; i++)
+		asd_write_reg_byte(asd_ha, LmMnSCRATCHPAGE(lseq, i), 0);
+
+	/* Initialize Mode 5 SCRATCHPAGE to 0. */
+	asd_write_reg_byte(asd_ha, LmMnSCRATCHPAGE(lseq, 5), 0);
+
+	asd_write_reg_dword(asd_ha, LmRSPMBX(lseq), 0);
+	/* Initialize Mode 0,1,2 and 5 Interrupt Enable and
+	 * Interrupt registers. */
+	asd_write_reg_dword(asd_ha, LmMnINTEN(lseq, 0), LmM0INTEN_MASK);
+	asd_write_reg_dword(asd_ha, LmMnINT(lseq, 0), 0xFFFFFFFF);
+	/* Mode 1 */
+	asd_write_reg_dword(asd_ha, LmMnINTEN(lseq, 1), LmM1INTEN_MASK);
+	asd_write_reg_dword(asd_ha, LmMnINT(lseq, 1), 0xFFFFFFFF);
+	/* Mode 2 */
+	asd_write_reg_dword(asd_ha, LmMnINTEN(lseq, 2), LmM2INTEN_MASK);
+	asd_write_reg_dword(asd_ha, LmMnINT(lseq, 2), 0xFFFFFFFF);
+	/* Mode 5 */
+	asd_write_reg_dword(asd_ha, LmMnINTEN(lseq, 5), LmM5INTEN_MASK);
+	asd_write_reg_dword(asd_ha, LmMnINT(lseq, 5), 0xFFFFFFFF);
+
+	/* Enable HW Timer status. */
+	asd_write_reg_byte(asd_ha, LmHWTSTATEN(lseq), LmHWTSTATEN_MASK);
+
+	/* Enable Primitive Status 0 and 1. */
+	asd_write_reg_dword(asd_ha, LmPRIMSTAT0EN(lseq), LmPRIMSTAT0EN_MASK);
+	asd_write_reg_dword(asd_ha, LmPRIMSTAT1EN(lseq), LmPRIMSTAT1EN_MASK);
+
+	/* Enable Frame Error. */
+	asd_write_reg_dword(asd_ha, LmFRMERREN(lseq), LmFRMERREN_MASK);
+	asd_write_reg_byte(asd_ha, LmMnHOLDLVL(lseq, 0), 0x50);
+
+	/* Initialize Mode 0 Transfer Level to 512. */
+	asd_write_reg_byte(asd_ha,  LmMnXFRLVL(lseq, 0), LmMnXFRLVL_512);
+	/* Initialize Mode 1 Transfer Level to 256. */
+	asd_write_reg_byte(asd_ha, LmMnXFRLVL(lseq, 1), LmMnXFRLVL_256);
+
+	/* Initialize Program Count. */
+	asd_write_reg_word(asd_ha, LmPRGMCNT(lseq), lseq_idle_loop);
+
+	/* Enable Blind SG Move. */
+	asd_write_reg_dword(asd_ha, LmMODECTL(lseq), LmBLIND48);
+	asd_write_reg_word(asd_ha, LmM3SATATIMER(lseq),
+			   ASD_SATA_INTERLOCK_TIMEOUT);
+
+	(void) asd_read_reg_dword(asd_ha, LmREQMBX(lseq));
+
+	/* Clear Primitive Status 0 and 1. */
+	asd_write_reg_dword(asd_ha, LmPRMSTAT0(lseq), 0xFFFFFFFF);
+	asd_write_reg_dword(asd_ha, LmPRMSTAT1(lseq), 0xFFFFFFFF);
+
+	/* Clear HW Timer status. */
+	asd_write_reg_byte(asd_ha, LmHWTSTAT(lseq), 0xFF);
+
+	/* Clear DMA Errors for Mode 0 and 1. */
+	asd_write_reg_byte(asd_ha, LmMnDMAERRS(lseq, 0), 0xFF);
+	asd_write_reg_byte(asd_ha, LmMnDMAERRS(lseq, 1), 0xFF);
+
+	/* Clear SG DMA Errors for Mode 0 and 1. */
+	asd_write_reg_byte(asd_ha, LmMnSGDMAERRS(lseq, 0), 0xFF);
+	asd_write_reg_byte(asd_ha, LmMnSGDMAERRS(lseq, 1), 0xFF);
+
+	/* Clear Mode 0 Buffer Parity Error. */
+	asd_write_reg_byte(asd_ha, LmMnBUFSTAT(lseq, 0), LmMnBUFPERR);
+
+	/* Clear Mode 0 Frame Error register. */
+	asd_write_reg_dword(asd_ha, LmMnFRMERR(lseq, 0), 0xFFFFFFFF);
+
+	/* Reset LSEQ external interrupt arbiter. */
+	asd_write_reg_byte(asd_ha, LmARP2INTCTL(lseq), RSTINTCTL);
+
+	/* Set the Phy SAS for the LmSEQ WWN. */
+	sas_addr = asd_ha->phys[lseq].phy_desc->sas_addr;
+	for (i = 0; i < SAS_ADDR_SIZE; i++)
+		asd_write_reg_byte(asd_ha, LmWWN(lseq) + i, sas_addr[i]);
+
+	/* Set the Transmit Size to 1024 bytes, 0 = 256 Dwords. */
+	asd_write_reg_byte(asd_ha, LmMnXMTSIZE(lseq, 1), 0);
+
+	/* Set the Bus Inactivity Time Limit Timer. */
+	asd_write_reg_word(asd_ha, LmBITL_TIMER(lseq), 9);
+
+	/* Enable SATA Port Multiplier. */
+	asd_write_reg_byte(asd_ha, LmMnSATAFS(lseq, 1), 0x80);
+
+	/* Initialize Interrupt Vector[0-10] address in Mode 3.
+	 * See the comment on CSEQ_INT_* */
+	asd_write_reg_word(asd_ha, LmM3INTVEC0(lseq), lseq_vecs[0]);
+	asd_write_reg_word(asd_ha, LmM3INTVEC1(lseq), lseq_vecs[1]);
+	asd_write_reg_word(asd_ha, LmM3INTVEC2(lseq), lseq_vecs[2]);
+	asd_write_reg_word(asd_ha, LmM3INTVEC3(lseq), lseq_vecs[3]);
+	asd_write_reg_word(asd_ha, LmM3INTVEC4(lseq), lseq_vecs[4]);
+	asd_write_reg_word(asd_ha, LmM3INTVEC5(lseq), lseq_vecs[5]);
+	asd_write_reg_word(asd_ha, LmM3INTVEC6(lseq), lseq_vecs[6]);
+	asd_write_reg_word(asd_ha, LmM3INTVEC7(lseq), lseq_vecs[7]);
+	asd_write_reg_word(asd_ha, LmM3INTVEC8(lseq), lseq_vecs[8]);
+	asd_write_reg_word(asd_ha, LmM3INTVEC9(lseq), lseq_vecs[9]);
+	asd_write_reg_word(asd_ha, LmM3INTVEC10(lseq), lseq_vecs[10]);
+	/*
+	 * Program the Link LED control, applicable only for
+	 * Chip Rev. B or later.
+	 */
+	asd_write_reg_dword(asd_ha, LmCONTROL(lseq),
+			    (LEDTIMER | LEDMODE_TXRX | LEDTIMERS_100ms));
+
+	/* Set the Align Rate for SAS and STP mode. */
+	asd_write_reg_byte(asd_ha, LmM1SASALIGN(lseq), SAS_ALIGN_DEFAULT);
+	asd_write_reg_byte(asd_ha, LmM1STPALIGN(lseq), STP_ALIGN_DEFAULT);
+}
+
+
+/**
+ * asd_post_init_cseq -- clear CSEQ Mode n Int. status and Response mailbox
+ * @asd_ha: pointer to host adapter struct
+ */
+static void asd_post_init_cseq(struct asd_ha_struct *asd_ha)
+{
+	int i;
+
+	for (i = 0; i < 8; i++)
+		asd_write_reg_dword(asd_ha, CMnINT(i), 0xFFFFFFFF);
+	for (i = 0; i < 8; i++)
+		asd_read_reg_dword(asd_ha, CMnRSPMBX(i));
+	/* Reset the external interrupt arbiter. */
+	asd_write_reg_byte(asd_ha, CARP2INTCTL, RSTINTCTL);
+}
+
+/**
+ * asd_init_ddb_0 -- initialize DDB 0
+ * @asd_ha: pointer to host adapter structure
+ *
+ * Initialize DDB site 0 which is used internally by the sequencer.
+ */
+static void asd_init_ddb_0(struct asd_ha_struct *asd_ha)
+{
+	int	i;
+
+	/* Zero out the DDB explicitly */
+	for (i = 0; i < sizeof(struct asd_ddb_seq_shared); i+=4)
+		asd_ddbsite_write_dword(asd_ha, 0, i, 0);
+
+	asd_ddbsite_write_word(asd_ha, 0,
+		 offsetof(struct asd_ddb_seq_shared, q_free_ddb_head), 0);
+	asd_ddbsite_write_word(asd_ha, 0,
+		 offsetof(struct asd_ddb_seq_shared, q_free_ddb_tail),
+			       asd_ha->hw_prof.max_ddbs-1);
+	asd_ddbsite_write_word(asd_ha, 0,
+		 offsetof(struct asd_ddb_seq_shared, q_free_ddb_cnt), 0);
+	asd_ddbsite_write_word(asd_ha, 0,
+		 offsetof(struct asd_ddb_seq_shared, q_used_ddb_head), 0xFFFF);
+	asd_ddbsite_write_word(asd_ha, 0,
+		 offsetof(struct asd_ddb_seq_shared, q_used_ddb_tail), 0xFFFF);
+	asd_ddbsite_write_word(asd_ha, 0,
+		 offsetof(struct asd_ddb_seq_shared, shared_mem_lock), 0);
+	asd_ddbsite_write_word(asd_ha, 0,
+		 offsetof(struct asd_ddb_seq_shared, smp_conn_tag), 0);
+	asd_ddbsite_write_word(asd_ha, 0,
+		 offsetof(struct asd_ddb_seq_shared, est_nexus_buf_cnt), 0);
+	asd_ddbsite_write_word(asd_ha, 0,
+		 offsetof(struct asd_ddb_seq_shared, est_nexus_buf_thresh),
+			       asd_ha->hw_prof.num_phys * 2);
+	asd_ddbsite_write_byte(asd_ha, 0,
+		 offsetof(struct asd_ddb_seq_shared, settable_max_contexts),0);
+	asd_ddbsite_write_byte(asd_ha, 0,
+	       offsetof(struct asd_ddb_seq_shared, conn_not_active), 0xFF);
+	asd_ddbsite_write_byte(asd_ha, 0,
+	       offsetof(struct asd_ddb_seq_shared, phy_is_up), 0x00);
+	/* DDB 0 is reserved */
+	set_bit(0, asd_ha->hw_prof.ddb_bitmap);
+}
+
+/**
+ * asd_seq_setup_seqs -- setup and initialize central and link sequencers
+ * @asd_ha: pointer to host adapter structure
+ */
+static void asd_seq_setup_seqs(struct asd_ha_struct *asd_ha)
+{
+	int 		lseq;
+	u8		lseq_mask;
+
+	/* Initialize SCB sites. Done first to compute some values which
+	 * the rest of the init code depends on. */
+	asd_init_scb_sites(asd_ha);
+
+	/* Initialize CSEQ Scratch RAM registers. */
+	asd_init_cseq_scratch(asd_ha);
+
+	/* Initialize LmSEQ Scratch RAM registers. */
+	asd_init_lseq_scratch(asd_ha);
+
+	/* Initialize CSEQ CIO registers. */
+	asd_init_cseq_cio(asd_ha);
+
+	asd_init_ddb_0(asd_ha);
+
+	/* Initialize LmSEQ CIO registers. */
+	lseq_mask = asd_ha->hw_prof.enabled_phys;
+	for_each_sequencer(lseq_mask, lseq_mask, lseq)
+		asd_init_lseq_cio(asd_ha, lseq);
+	asd_post_init_cseq(asd_ha);
+}
+
+
+/**
+ * asd_seq_start_cseq -- start the central sequencer, CSEQ
+ * @asd_ha: pointer to host adapter structure
+ */
+static int asd_seq_start_cseq(struct asd_ha_struct *asd_ha)
+{
+	/* Reset the ARP2 instruction to location zero. */
+	asd_write_reg_word(asd_ha, CPRGMCNT, cseq_idle_loop);
+
+	/* Unpause the CSEQ  */
+	return asd_unpause_cseq(asd_ha);
+}
+
+/**
+ * asd_seq_start_lseq -- start a link sequencer
+ * @asd_ha: pointer to host adapter structure
+ * @lseq: the link sequencer of interest
+ */
+static int asd_seq_start_lseq(struct asd_ha_struct *asd_ha, int lseq)
+{
+	/* Reset the ARP2 instruction to location zero. */
+	asd_write_reg_word(asd_ha, LmPRGMCNT(lseq), lseq_idle_loop);
+
+	/* Unpause the LmSEQ  */
+	return asd_seq_unpause_lseq(asd_ha, lseq);
+}
+
+static int asd_request_firmware(struct asd_ha_struct *asd_ha)
+{
+	int err, i;
+	struct sequencer_file_header header, *hdr_ptr;
+	u32 csum = 0;
+	u16 *ptr_cseq_vecs, *ptr_lseq_vecs;
+
+	if (sequencer_fw)
+		/* already loaded */
+		return 0;
+
+	err = request_firmware(&sequencer_fw,
+			       SAS_RAZOR_SEQUENCER_FW_FILE,
+			       &asd_ha->pcidev->dev);
+	if (err)
+		return err;
+
+	hdr_ptr = (struct sequencer_file_header *)sequencer_fw->data;
+
+	header.csum = le32_to_cpu(hdr_ptr->csum);
+	header.major = le32_to_cpu(hdr_ptr->major);
+	header.minor = le32_to_cpu(hdr_ptr->minor);
+	sequencer_version = hdr_ptr->version;
+	header.cseq_table_offset = le32_to_cpu(hdr_ptr->cseq_table_offset);
+	header.cseq_table_size = le32_to_cpu(hdr_ptr->cseq_table_size);
+	header.lseq_table_offset = le32_to_cpu(hdr_ptr->lseq_table_offset);
+	header.lseq_table_size = le32_to_cpu(hdr_ptr->lseq_table_size);
+	header.cseq_code_offset = le32_to_cpu(hdr_ptr->cseq_code_offset);
+	header.cseq_code_size = le32_to_cpu(hdr_ptr->cseq_code_size);
+	header.lseq_code_offset = le32_to_cpu(hdr_ptr->lseq_code_offset);
+	header.lseq_code_size = le32_to_cpu(hdr_ptr->lseq_code_size);
+	header.mode2_task = le16_to_cpu(hdr_ptr->mode2_task);
+	header.cseq_idle_loop = le16_to_cpu(hdr_ptr->cseq_idle_loop);
+	header.lseq_idle_loop = le16_to_cpu(hdr_ptr->lseq_idle_loop);
+
+	for (i = sizeof(header.csum); i < sequencer_fw->size; i++)
+		csum += sequencer_fw->data[i];
+
+	if (csum != header.csum) {
+		asd_printk("Firmware file checksum mismatch\n");
+		return -EINVAL;
+	}
+
+	if (header.cseq_table_size != CSEQ_NUM_VECS ||
+	    header.lseq_table_size != LSEQ_NUM_VECS) {
+		asd_printk("Firmware file table size mismatch\n");
+		return -EINVAL;
+	}
+
+	ptr_cseq_vecs = (u16 *)&sequencer_fw->data[header.cseq_table_offset];
+	ptr_lseq_vecs = (u16 *)&sequencer_fw->data[header.lseq_table_offset];
+	mode2_task = header.mode2_task;
+	cseq_idle_loop = header.cseq_idle_loop;
+	lseq_idle_loop = header.lseq_idle_loop;
+
+	for (i = 0; i < CSEQ_NUM_VECS; i++)
+		cseq_vecs[i] = le16_to_cpu(ptr_cseq_vecs[i]);
+
+	for (i = 0; i < LSEQ_NUM_VECS; i++)
+		lseq_vecs[i] = le16_to_cpu(ptr_lseq_vecs[i]);
+
+	cseq_code = &sequencer_fw->data[header.cseq_code_offset];
+	cseq_code_size = header.cseq_code_size;
+	lseq_code = &sequencer_fw->data[header.lseq_code_offset];
+	lseq_code_size = header.lseq_code_size;
+
+	return 0;
+}
+
+int asd_init_seqs(struct asd_ha_struct *asd_ha)
+{
+	int err;
+
+	err = asd_request_firmware(asd_ha);
+
+	if (err) {
+		asd_printk("Failed to load sequencer firmware file %s, error %d\n",
+			   SAS_RAZOR_SEQUENCER_FW_FILE, err);
+		return err;
+	}
+
+	asd_printk("using sequencer %s\n", sequencer_version);
+	err = asd_seq_download_seqs(asd_ha);
+	if (err) {
+		asd_printk("couldn't download sequencers for %s\n",
+			   pci_name(asd_ha->pcidev));
+		return err;
+	}
+
+	asd_seq_setup_seqs(asd_ha);
+
+	return 0;
+}
+
+int asd_start_seqs(struct asd_ha_struct *asd_ha)
+{
+	int err;
+	u8  lseq_mask;
+	int lseq;
+
+	err = asd_seq_start_cseq(asd_ha);
+	if (err) {
+		asd_printk("couldn't start CSEQ for %s\n",
+			   pci_name(asd_ha->pcidev));
+		return err;
+	}
+
+	lseq_mask = asd_ha->hw_prof.enabled_phys;
+	for_each_sequencer(lseq_mask, lseq_mask, lseq) {
+		err = asd_seq_start_lseq(asd_ha, lseq);
+		if (err) {
+			asd_printk("coudln't start LSEQ %d for %s\n", lseq,
+				   pci_name(asd_ha->pcidev));
+			return err;
+		}
+	}
+
+	return 0;
+}
+
+/**
+ * asd_update_port_links -- update port_map_by_links and phy_is_up
+ * @sas_phy: pointer to the phy which has been added to a port
+ *
+ * 1) When a link reset has completed and we got BYTES DMAED with a
+ * valid frame we call this function for that phy, to indicate that
+ * the phy is up, i.e. we update the phy_is_up in DDB 0.  The
+ * sequencer checks phy_is_up when pending SCBs are to be sent, and
+ * when an open address frame has been received.
+ *
+ * 2) When we know of ports, we call this function to update the map
+ * of phys participaing in that port, i.e. we update the
+ * port_map_by_links in DDB 0.  When a HARD_RESET primitive has been
+ * received, the sequencer disables all phys in that port.
+ * port_map_by_links is also used as the conn_mask byte in the
+ * initiator/target port DDB.
+ */
+void asd_update_port_links(struct asd_sas_phy *sas_phy)
+{
+	struct asd_ha_struct *asd_ha = sas_phy->ha->lldd_ha;
+	const u8 phy_mask = (u8) sas_phy->port->phy_mask;
+	u8  phy_is_up;
+	u8  mask;
+	int i, err;
+
+	for_each_phy(phy_mask, mask, i)
+		asd_ddbsite_write_byte(asd_ha, 0,
+				       offsetof(struct asd_ddb_seq_shared,
+						port_map_by_links)+i,phy_mask);
+
+	for (i = 0; i < 12; i++) {
+		phy_is_up = asd_ddbsite_read_byte(asd_ha, 0,
+			  offsetof(struct asd_ddb_seq_shared, phy_is_up));
+		err = asd_ddbsite_update_byte(asd_ha, 0,
+				offsetof(struct asd_ddb_seq_shared, phy_is_up),
+				phy_is_up,
+				phy_is_up | phy_mask);
+		if (!err)
+			break;
+		else if (err == -EFAULT) {
+			asd_printk("phy_is_up: parity error in DDB 0\n");
+			break;
+		}
+	}
+
+	if (err)
+		asd_printk("couldn't update DDB 0:error:%d\n", err);
+}
diff --git a/drivers/scsi/aic94xx/aic94xx_seq.h b/drivers/scsi/aic94xx/aic94xx_seq.h
new file mode 100644
index 000000000000..42281c36153b
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx_seq.h
@@ -0,0 +1,70 @@
+/*
+ * Aic94xx SAS/SATA driver sequencer interface header file.
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#ifndef _AIC94XX_SEQ_H_
+#define _AIC94XX_SEQ_H_
+
+#define CSEQ_NUM_VECS	3
+#define LSEQ_NUM_VECS	11
+
+#define SAS_RAZOR_SEQUENCER_FW_FILE "aic94xx-seq.fw"
+
+/* Note:  All quantites in the sequencer file are little endian */
+struct sequencer_file_header {
+	/* Checksum of the entire contents of the sequencer excluding
+	 * these four bytes */
+	u32	csum;
+	/* numeric major version */
+	u32	major;
+	/* numeric minor version */
+	u32	minor;
+	/* version string printed by driver */
+	char	version[16];
+	u32	cseq_table_offset;
+	u32	cseq_table_size;
+	u32	lseq_table_offset;
+	u32	lseq_table_size;
+	u32	cseq_code_offset;
+	u32	cseq_code_size;
+	u32	lseq_code_offset;
+	u32	lseq_code_size;
+	u16	mode2_task;
+	u16	cseq_idle_loop;
+	u16	lseq_idle_loop;
+} __attribute__((packed));
+
+#ifdef __KERNEL__
+int asd_pause_cseq(struct asd_ha_struct *asd_ha);
+int asd_unpause_cseq(struct asd_ha_struct *asd_ha);
+int asd_pause_lseq(struct asd_ha_struct *asd_ha, u8 lseq_mask);
+int asd_unpause_lseq(struct asd_ha_struct *asd_ha, u8 lseq_mask);
+int asd_init_seqs(struct asd_ha_struct *asd_ha);
+int asd_start_seqs(struct asd_ha_struct *asd_ha);
+
+void asd_update_port_links(struct asd_sas_phy *phy);
+#endif
+
+#endif
diff --git a/drivers/scsi/aic94xx/aic94xx_task.c b/drivers/scsi/aic94xx/aic94xx_task.c
new file mode 100644
index 000000000000..285e70dae933
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx_task.c
@@ -0,0 +1,642 @@
+/*
+ * Aic94xx SAS/SATA Tasks
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#include <linux/spinlock.h>
+#include "aic94xx.h"
+#include "aic94xx_sas.h"
+#include "aic94xx_hwi.h"
+
+static void asd_unbuild_ata_ascb(struct asd_ascb *a);
+static void asd_unbuild_smp_ascb(struct asd_ascb *a);
+static void asd_unbuild_ssp_ascb(struct asd_ascb *a);
+
+static inline void asd_can_dequeue(struct asd_ha_struct *asd_ha, int num)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&asd_ha->seq.pend_q_lock, flags);
+	asd_ha->seq.can_queue += num;
+	spin_unlock_irqrestore(&asd_ha->seq.pend_q_lock, flags);
+}
+
+/* PCI_DMA_... to our direction translation.
+ */
+static const u8 data_dir_flags[] = {
+	[PCI_DMA_BIDIRECTIONAL] = DATA_DIR_BYRECIPIENT,	/* UNSPECIFIED */
+	[PCI_DMA_TODEVICE]      = DATA_DIR_OUT, /* OUTBOUND */
+	[PCI_DMA_FROMDEVICE]    = DATA_DIR_IN, /* INBOUND */
+	[PCI_DMA_NONE]          = DATA_DIR_NONE, /* NO TRANSFER */
+};
+
+static inline int asd_map_scatterlist(struct sas_task *task,
+				      struct sg_el *sg_arr,
+				      unsigned long gfp_flags)
+{
+	struct asd_ascb *ascb = task->lldd_task;
+	struct asd_ha_struct *asd_ha = ascb->ha;
+	struct scatterlist *sc;
+	int num_sg, res;
+
+	if (task->data_dir == PCI_DMA_NONE)
+		return 0;
+
+	if (task->num_scatter == 0) {
+		void *p = task->scatter;
+		dma_addr_t dma = pci_map_single(asd_ha->pcidev, p,
+						task->total_xfer_len,
+						task->data_dir);
+		sg_arr[0].bus_addr = cpu_to_le64((u64)dma);
+		sg_arr[0].size = cpu_to_le32(task->total_xfer_len);
+		sg_arr[0].flags |= ASD_SG_EL_LIST_EOL;
+		return 0;
+	}
+
+	num_sg = pci_map_sg(asd_ha->pcidev, task->scatter, task->num_scatter,
+			    task->data_dir);
+	if (num_sg == 0)
+		return -ENOMEM;
+
+	if (num_sg > 3) {
+		int i;
+
+		ascb->sg_arr = asd_alloc_coherent(asd_ha,
+						  num_sg*sizeof(struct sg_el),
+						  gfp_flags);
+		if (!ascb->sg_arr) {
+			res = -ENOMEM;
+			goto err_unmap;
+		}
+		for (sc = task->scatter, i = 0; i < num_sg; i++, sc++) {
+			struct sg_el *sg =
+				&((struct sg_el *)ascb->sg_arr->vaddr)[i];
+			sg->bus_addr = cpu_to_le64((u64)sg_dma_address(sc));
+			sg->size = cpu_to_le32((u32)sg_dma_len(sc));
+			if (i == num_sg-1)
+				sg->flags |= ASD_SG_EL_LIST_EOL;
+		}
+
+		for (sc = task->scatter, i = 0; i < 2; i++, sc++) {
+			sg_arr[i].bus_addr =
+				cpu_to_le64((u64)sg_dma_address(sc));
+			sg_arr[i].size = cpu_to_le32((u32)sg_dma_len(sc));
+		}
+		sg_arr[1].next_sg_offs = 2 * sizeof(*sg_arr);
+		sg_arr[1].flags |= ASD_SG_EL_LIST_EOS;
+
+		memset(&sg_arr[2], 0, sizeof(*sg_arr));
+		sg_arr[2].bus_addr=cpu_to_le64((u64)ascb->sg_arr->dma_handle);
+	} else {
+		int i;
+		for (sc = task->scatter, i = 0; i < num_sg; i++, sc++) {
+			sg_arr[i].bus_addr =
+				cpu_to_le64((u64)sg_dma_address(sc));
+			sg_arr[i].size = cpu_to_le32((u32)sg_dma_len(sc));
+		}
+		sg_arr[i-1].flags |= ASD_SG_EL_LIST_EOL;
+	}
+
+	return 0;
+err_unmap:
+	pci_unmap_sg(asd_ha->pcidev, task->scatter, task->num_scatter,
+		     task->data_dir);
+	return res;
+}
+
+static inline void asd_unmap_scatterlist(struct asd_ascb *ascb)
+{
+	struct asd_ha_struct *asd_ha = ascb->ha;
+	struct sas_task *task = ascb->uldd_task;
+
+	if (task->data_dir == PCI_DMA_NONE)
+		return;
+
+	if (task->num_scatter == 0) {
+		dma_addr_t dma = (dma_addr_t)
+		       le64_to_cpu(ascb->scb->ssp_task.sg_element[0].bus_addr);
+		pci_unmap_single(ascb->ha->pcidev, dma, task->total_xfer_len,
+				 task->data_dir);
+		return;
+	}
+
+	asd_free_coherent(asd_ha, ascb->sg_arr);
+	pci_unmap_sg(asd_ha->pcidev, task->scatter, task->num_scatter,
+		     task->data_dir);
+}
+
+/* ---------- Task complete tasklet ---------- */
+
+static void asd_get_response_tasklet(struct asd_ascb *ascb,
+				     struct done_list_struct *dl)
+{
+	struct asd_ha_struct *asd_ha = ascb->ha;
+	struct sas_task *task = ascb->uldd_task;
+	struct task_status_struct *ts = &task->task_status;
+	unsigned long flags;
+	struct tc_resp_sb_struct {
+		__le16 index_escb;
+		u8     len_lsb;
+		u8     flags;
+	} __attribute__ ((packed)) *resp_sb = (void *) dl->status_block;
+
+/* 	int  size   = ((resp_sb->flags & 7) << 8) | resp_sb->len_lsb; */
+	int  edb_id = ((resp_sb->flags & 0x70) >> 4)-1;
+	struct asd_ascb *escb;
+	struct asd_dma_tok *edb;
+	void *r;
+
+	spin_lock_irqsave(&asd_ha->seq.tc_index_lock, flags);
+	escb = asd_tc_index_find(&asd_ha->seq,
+				 (int)le16_to_cpu(resp_sb->index_escb));
+	spin_unlock_irqrestore(&asd_ha->seq.tc_index_lock, flags);
+
+	if (!escb) {
+		ASD_DPRINTK("Uh-oh! No escb for this dl?!\n");
+		return;
+	}
+
+	ts->buf_valid_size = 0;
+	edb = asd_ha->seq.edb_arr[edb_id + escb->edb_index];
+	r = edb->vaddr;
+	if (task->task_proto == SAS_PROTO_SSP) {
+		struct ssp_response_iu *iu =
+			r + 16 + sizeof(struct ssp_frame_hdr);
+
+		ts->residual = le32_to_cpu(*(__le32 *)r);
+		ts->resp = SAS_TASK_COMPLETE;
+		if (iu->datapres == 0)
+			ts->stat = iu->status;
+		else if (iu->datapres == 1)
+			ts->stat = iu->resp_data[3];
+		else if (iu->datapres == 2) {
+			ts->stat = SAM_CHECK_COND;
+			ts->buf_valid_size = min((u32) SAS_STATUS_BUF_SIZE,
+					 be32_to_cpu(iu->sense_data_len));
+			memcpy(ts->buf, iu->sense_data, ts->buf_valid_size);
+			if (iu->status != SAM_CHECK_COND) {
+				ASD_DPRINTK("device %llx sent sense data, but "
+					    "stat(0x%x) is not CHECK_CONDITION"
+					    "\n",
+					    SAS_ADDR(task->dev->sas_addr),
+					    ts->stat);
+			}
+		}
+	}  else {
+		struct ata_task_resp *resp = (void *) &ts->buf[0];
+
+		ts->residual = le32_to_cpu(*(__le32 *)r);
+
+		if (SAS_STATUS_BUF_SIZE >= sizeof(*resp)) {
+			resp->frame_len = le16_to_cpu(*(__le16 *)(r+6));
+			memcpy(&resp->ending_fis[0], r+16, 24);
+			ts->buf_valid_size = sizeof(*resp);
+		}
+	}
+
+	asd_invalidate_edb(escb, edb_id);
+}
+
+static void asd_task_tasklet_complete(struct asd_ascb *ascb,
+				      struct done_list_struct *dl)
+{
+	struct sas_task *task = ascb->uldd_task;
+	struct task_status_struct *ts = &task->task_status;
+	unsigned long flags;
+	u8 opcode = dl->opcode;
+
+	asd_can_dequeue(ascb->ha, 1);
+
+Again:
+	switch (opcode) {
+	case TC_NO_ERROR:
+		ts->resp = SAS_TASK_COMPLETE;
+		ts->stat = SAM_GOOD;
+		break;
+	case TC_UNDERRUN:
+		ts->resp = SAS_TASK_COMPLETE;
+		ts->stat = SAS_DATA_UNDERRUN;
+		ts->residual = le32_to_cpu(*(__le32 *)dl->status_block);
+		break;
+	case TC_OVERRUN:
+		ts->resp = SAS_TASK_COMPLETE;
+		ts->stat = SAS_DATA_OVERRUN;
+		ts->residual = 0;
+		break;
+	case TC_SSP_RESP:
+	case TC_ATA_RESP:
+		ts->resp = SAS_TASK_COMPLETE;
+		ts->stat = SAS_PROTO_RESPONSE;
+		asd_get_response_tasklet(ascb, dl);
+		break;
+	case TF_OPEN_REJECT:
+		ts->resp = SAS_TASK_UNDELIVERED;
+		ts->stat = SAS_OPEN_REJECT;
+		if (dl->status_block[1] & 2)
+			ts->open_rej_reason = 1 + dl->status_block[2];
+		else if (dl->status_block[1] & 1)
+			ts->open_rej_reason = (dl->status_block[2] >> 4)+10;
+		else
+			ts->open_rej_reason = SAS_OREJ_UNKNOWN;
+		break;
+	case TF_OPEN_TO:
+		ts->resp = SAS_TASK_UNDELIVERED;
+		ts->stat = SAS_OPEN_TO;
+		break;
+	case TF_PHY_DOWN:
+	case TU_PHY_DOWN:
+		ts->resp = SAS_TASK_UNDELIVERED;
+		ts->stat = SAS_PHY_DOWN;
+		break;
+	case TI_PHY_DOWN:
+		ts->resp = SAS_TASK_COMPLETE;
+		ts->stat = SAS_PHY_DOWN;
+		break;
+	case TI_BREAK:
+	case TI_PROTO_ERR:
+	case TI_NAK:
+	case TI_ACK_NAK_TO:
+	case TF_SMP_XMIT_RCV_ERR:
+	case TC_ATA_R_ERR_RECV:
+		ts->resp = SAS_TASK_COMPLETE;
+		ts->stat = SAS_INTERRUPTED;
+		break;
+	case TF_BREAK:
+	case TU_BREAK:
+	case TU_ACK_NAK_TO:
+	case TF_SMPRSP_TO:
+		ts->resp = SAS_TASK_UNDELIVERED;
+		ts->stat = SAS_DEV_NO_RESPONSE;
+		break;
+	case TF_NAK_RECV:
+		ts->resp = SAS_TASK_COMPLETE;
+		ts->stat = SAS_NAK_R_ERR;
+		break;
+	case TA_I_T_NEXUS_LOSS:
+		opcode = dl->status_block[0];
+		goto Again;
+		break;
+	case TF_INV_CONN_HANDLE:
+		ts->resp = SAS_TASK_UNDELIVERED;
+		ts->stat = SAS_DEVICE_UNKNOWN;
+		break;
+	case TF_REQUESTED_N_PENDING:
+		ts->resp = SAS_TASK_UNDELIVERED;
+		ts->stat = SAS_PENDING;
+		break;
+	case TC_TASK_CLEARED:
+	case TA_ON_REQ:
+		ts->resp = SAS_TASK_COMPLETE;
+		ts->stat = SAS_ABORTED_TASK;
+		break;
+
+	case TF_NO_SMP_CONN:
+	case TF_TMF_NO_CTX:
+	case TF_TMF_NO_TAG:
+	case TF_TMF_TAG_FREE:
+	case TF_TMF_TASK_DONE:
+	case TF_TMF_NO_CONN_HANDLE:
+	case TF_IRTT_TO:
+	case TF_IU_SHORT:
+	case TF_DATA_OFFS_ERR:
+		ts->resp = SAS_TASK_UNDELIVERED;
+		ts->stat = SAS_DEV_NO_RESPONSE;
+		break;
+
+	case TC_LINK_ADM_RESP:
+	case TC_CONTROL_PHY:
+	case TC_RESUME:
+	case TC_PARTIAL_SG_LIST:
+	default:
+		ASD_DPRINTK("%s: dl opcode: 0x%x?\n", __FUNCTION__, opcode);
+		break;
+	}
+
+	switch (task->task_proto) {
+	case SATA_PROTO:
+	case SAS_PROTO_STP:
+		asd_unbuild_ata_ascb(ascb);
+		break;
+	case SAS_PROTO_SMP:
+		asd_unbuild_smp_ascb(ascb);
+		break;
+	case SAS_PROTO_SSP:
+		asd_unbuild_ssp_ascb(ascb);
+	default:
+		break;
+	}
+
+	spin_lock_irqsave(&task->task_state_lock, flags);
+	task->task_state_flags &= ~SAS_TASK_STATE_PENDING;
+	task->task_state_flags |= SAS_TASK_STATE_DONE;
+	if (unlikely((task->task_state_flags & SAS_TASK_STATE_ABORTED))) {
+		spin_unlock_irqrestore(&task->task_state_lock, flags);
+		ASD_DPRINTK("task 0x%p done with opcode 0x%x resp 0x%x "
+			    "stat 0x%x but aborted by upper layer!\n",
+			    task, opcode, ts->resp, ts->stat);
+		complete(&ascb->completion);
+	} else {
+		spin_unlock_irqrestore(&task->task_state_lock, flags);
+		task->lldd_task = NULL;
+		asd_ascb_free(ascb);
+		mb();
+		task->task_done(task);
+	}
+}
+
+/* ---------- ATA ---------- */
+
+static int asd_build_ata_ascb(struct asd_ascb *ascb, struct sas_task *task,
+			      unsigned long gfp_flags)
+{
+	struct domain_device *dev = task->dev;
+	struct scb *scb;
+	u8     flags;
+	int    res = 0;
+
+	scb = ascb->scb;
+
+	if (unlikely(task->ata_task.device_control_reg_update))
+		scb->header.opcode = CONTROL_ATA_DEV;
+	else if (dev->sata_dev.command_set == ATA_COMMAND_SET)
+		scb->header.opcode = INITIATE_ATA_TASK;
+	else
+		scb->header.opcode = INITIATE_ATAPI_TASK;
+
+	scb->ata_task.proto_conn_rate = (1 << 5); /* STP */
+	if (dev->port->oob_mode == SAS_OOB_MODE)
+		scb->ata_task.proto_conn_rate |= dev->linkrate;
+
+	scb->ata_task.total_xfer_len = cpu_to_le32(task->total_xfer_len);
+	scb->ata_task.fis = task->ata_task.fis;
+	scb->ata_task.fis.fis_type = 0x27;
+	if (likely(!task->ata_task.device_control_reg_update))
+		scb->ata_task.fis.flags |= 0x80; /* C=1: update ATA cmd reg */
+	scb->ata_task.fis.flags &= 0xF0; /* PM_PORT field shall be 0 */
+	if (dev->sata_dev.command_set == ATAPI_COMMAND_SET)
+		memcpy(scb->ata_task.atapi_packet, task->ata_task.atapi_packet,
+		       16);
+	scb->ata_task.sister_scb = cpu_to_le16(0xFFFF);
+	scb->ata_task.conn_handle = cpu_to_le16(
+		(u16)(unsigned long)dev->lldd_dev);
+
+	if (likely(!task->ata_task.device_control_reg_update)) {
+		flags = 0;
+		if (task->ata_task.dma_xfer)
+			flags |= DATA_XFER_MODE_DMA;
+		if (task->ata_task.use_ncq &&
+		    dev->sata_dev.command_set != ATAPI_COMMAND_SET)
+			flags |= ATA_Q_TYPE_NCQ;
+		flags |= data_dir_flags[task->data_dir];
+		scb->ata_task.ata_flags = flags;
+
+		scb->ata_task.retry_count = task->ata_task.retry_count;
+
+		flags = 0;
+		if (task->ata_task.set_affil_pol)
+			flags |= SET_AFFIL_POLICY;
+		if (task->ata_task.stp_affil_pol)
+			flags |= STP_AFFIL_POLICY;
+		scb->ata_task.flags = flags;
+	}
+	ascb->tasklet_complete = asd_task_tasklet_complete;
+
+	if (likely(!task->ata_task.device_control_reg_update))
+		res = asd_map_scatterlist(task, scb->ata_task.sg_element,
+					  gfp_flags);
+
+	return res;
+}
+
+static void asd_unbuild_ata_ascb(struct asd_ascb *a)
+{
+	asd_unmap_scatterlist(a);
+}
+
+/* ---------- SMP ---------- */
+
+static int asd_build_smp_ascb(struct asd_ascb *ascb, struct sas_task *task,
+			      unsigned long gfp_flags)
+{
+	struct asd_ha_struct *asd_ha = ascb->ha;
+	struct domain_device *dev = task->dev;
+	struct scb *scb;
+
+	pci_map_sg(asd_ha->pcidev, &task->smp_task.smp_req, 1,
+		   PCI_DMA_FROMDEVICE);
+	pci_map_sg(asd_ha->pcidev, &task->smp_task.smp_resp, 1,
+		   PCI_DMA_FROMDEVICE);
+
+	scb = ascb->scb;
+
+	scb->header.opcode = INITIATE_SMP_TASK;
+
+	scb->smp_task.proto_conn_rate = dev->linkrate;
+
+	scb->smp_task.smp_req.bus_addr =
+		cpu_to_le64((u64)sg_dma_address(&task->smp_task.smp_req));
+	scb->smp_task.smp_req.size =
+		cpu_to_le32((u32)sg_dma_len(&task->smp_task.smp_req)-4);
+
+	scb->smp_task.smp_resp.bus_addr =
+		cpu_to_le64((u64)sg_dma_address(&task->smp_task.smp_resp));
+	scb->smp_task.smp_resp.size =
+		cpu_to_le32((u32)sg_dma_len(&task->smp_task.smp_resp)-4);
+
+	scb->smp_task.sister_scb = cpu_to_le16(0xFFFF);
+	scb->smp_task.conn_handle = cpu_to_le16((u16)
+						(unsigned long)dev->lldd_dev);
+
+	ascb->tasklet_complete = asd_task_tasklet_complete;
+
+	return 0;
+}
+
+static void asd_unbuild_smp_ascb(struct asd_ascb *a)
+{
+	struct sas_task *task = a->uldd_task;
+
+	BUG_ON(!task);
+	pci_unmap_sg(a->ha->pcidev, &task->smp_task.smp_req, 1,
+		     PCI_DMA_FROMDEVICE);
+	pci_unmap_sg(a->ha->pcidev, &task->smp_task.smp_resp, 1,
+		     PCI_DMA_FROMDEVICE);
+}
+
+/* ---------- SSP ---------- */
+
+static int asd_build_ssp_ascb(struct asd_ascb *ascb, struct sas_task *task,
+			      unsigned long gfp_flags)
+{
+	struct domain_device *dev = task->dev;
+	struct scb *scb;
+	int    res = 0;
+
+	scb = ascb->scb;
+
+	scb->header.opcode = INITIATE_SSP_TASK;
+
+	scb->ssp_task.proto_conn_rate  = (1 << 4); /* SSP */
+	scb->ssp_task.proto_conn_rate |= dev->linkrate;
+	scb->ssp_task.total_xfer_len = cpu_to_le32(task->total_xfer_len);
+	scb->ssp_task.ssp_frame.frame_type = SSP_DATA;
+	memcpy(scb->ssp_task.ssp_frame.hashed_dest_addr, dev->hashed_sas_addr,
+	       HASHED_SAS_ADDR_SIZE);
+	memcpy(scb->ssp_task.ssp_frame.hashed_src_addr,
+	       dev->port->ha->hashed_sas_addr, HASHED_SAS_ADDR_SIZE);
+	scb->ssp_task.ssp_frame.tptt = cpu_to_be16(0xFFFF);
+
+	memcpy(scb->ssp_task.ssp_cmd.lun, task->ssp_task.LUN, 8);
+	if (task->ssp_task.enable_first_burst)
+		scb->ssp_task.ssp_cmd.efb_prio_attr |= EFB_MASK;
+	scb->ssp_task.ssp_cmd.efb_prio_attr |= (task->ssp_task.task_prio << 3);
+	scb->ssp_task.ssp_cmd.efb_prio_attr |= (task->ssp_task.task_attr & 7);
+	memcpy(scb->ssp_task.ssp_cmd.cdb, task->ssp_task.cdb, 16);
+
+	scb->ssp_task.sister_scb = cpu_to_le16(0xFFFF);
+	scb->ssp_task.conn_handle = cpu_to_le16(
+		(u16)(unsigned long)dev->lldd_dev);
+	scb->ssp_task.data_dir = data_dir_flags[task->data_dir];
+	scb->ssp_task.retry_count = scb->ssp_task.retry_count;
+
+	ascb->tasklet_complete = asd_task_tasklet_complete;
+
+	res = asd_map_scatterlist(task, scb->ssp_task.sg_element, gfp_flags);
+
+	return res;
+}
+
+static void asd_unbuild_ssp_ascb(struct asd_ascb *a)
+{
+	asd_unmap_scatterlist(a);
+}
+
+/* ---------- Execute Task ---------- */
+
+static inline int asd_can_queue(struct asd_ha_struct *asd_ha, int num)
+{
+	int res = 0;
+	unsigned long flags;
+
+	spin_lock_irqsave(&asd_ha->seq.pend_q_lock, flags);
+	if ((asd_ha->seq.can_queue - num) < 0)
+		res = -SAS_QUEUE_FULL;
+	else
+		asd_ha->seq.can_queue -= num;
+	spin_unlock_irqrestore(&asd_ha->seq.pend_q_lock, flags);
+
+	return res;
+}
+
+int asd_execute_task(struct sas_task *task, const int num,
+		     unsigned long gfp_flags)
+{
+	int res = 0;
+	LIST_HEAD(alist);
+	struct sas_task *t = task;
+	struct asd_ascb *ascb = NULL, *a;
+	struct asd_ha_struct *asd_ha = task->dev->port->ha->lldd_ha;
+
+	res = asd_can_queue(asd_ha, num);
+	if (res)
+		return res;
+
+	res = num;
+	ascb = asd_ascb_alloc_list(asd_ha, &res, gfp_flags);
+	if (res) {
+		res = -ENOMEM;
+		goto out_err;
+	}
+
+	__list_add(&alist, ascb->list.prev, &ascb->list);
+	list_for_each_entry(a, &alist, list) {
+		a->uldd_task = t;
+		t->lldd_task = a;
+		t = list_entry(t->list.next, struct sas_task, list);
+	}
+	list_for_each_entry(a, &alist, list) {
+		t = a->uldd_task;
+		a->uldd_timer = 1;
+		if (t->task_proto & SAS_PROTO_STP)
+			t->task_proto = SAS_PROTO_STP;
+		switch (t->task_proto) {
+		case SATA_PROTO:
+		case SAS_PROTO_STP:
+			res = asd_build_ata_ascb(a, t, gfp_flags);
+			break;
+		case SAS_PROTO_SMP:
+			res = asd_build_smp_ascb(a, t, gfp_flags);
+			break;
+		case SAS_PROTO_SSP:
+			res = asd_build_ssp_ascb(a, t, gfp_flags);
+			break;
+		default:
+			asd_printk("unknown sas_task proto: 0x%x\n",
+				   t->task_proto);
+			res = -ENOMEM;
+			break;
+		}
+		if (res)
+			goto out_err_unmap;
+	}
+	list_del_init(&alist);
+
+	res = asd_post_ascb_list(asd_ha, ascb, num);
+	if (unlikely(res)) {
+		a = NULL;
+		__list_add(&alist, ascb->list.prev, &ascb->list);
+		goto out_err_unmap;
+	}
+
+	return 0;
+out_err_unmap:
+	{
+		struct asd_ascb *b = a;
+		list_for_each_entry(a, &alist, list) {
+			if (a == b)
+				break;
+			t = a->uldd_task;
+			switch (t->task_proto) {
+			case SATA_PROTO:
+			case SAS_PROTO_STP:
+				asd_unbuild_ata_ascb(a);
+				break;
+			case SAS_PROTO_SMP:
+				asd_unbuild_smp_ascb(a);
+				break;
+			case SAS_PROTO_SSP:
+				asd_unbuild_ssp_ascb(a);
+			default:
+				break;
+			}
+			t->lldd_task = NULL;
+		}
+	}
+	list_del_init(&alist);
+out_err:
+	if (ascb)
+		asd_ascb_free_list(ascb);
+	asd_can_dequeue(asd_ha, num);
+	return res;
+}
diff --git a/drivers/scsi/aic94xx/aic94xx_tmf.c b/drivers/scsi/aic94xx/aic94xx_tmf.c
new file mode 100644
index 000000000000..61234384503b
--- /dev/null
+++ b/drivers/scsi/aic94xx/aic94xx_tmf.c
@@ -0,0 +1,636 @@
+/*
+ * Aic94xx Task Management Functions
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This file is part of the aic94xx driver.
+ *
+ * The aic94xx driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; version 2 of the
+ * License.
+ *
+ * The aic94xx driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with the aic94xx driver; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#include <linux/spinlock.h>
+#include "aic94xx.h"
+#include "aic94xx_sas.h"
+#include "aic94xx_hwi.h"
+
+/* ---------- Internal enqueue ---------- */
+
+static int asd_enqueue_internal(struct asd_ascb *ascb,
+		void (*tasklet_complete)(struct asd_ascb *,
+					 struct done_list_struct *),
+				void (*timed_out)(unsigned long))
+{
+	int res;
+
+	ascb->tasklet_complete = tasklet_complete;
+	ascb->uldd_timer = 1;
+
+	ascb->timer.data = (unsigned long) ascb;
+	ascb->timer.function = timed_out;
+	ascb->timer.expires = jiffies + AIC94XX_SCB_TIMEOUT;
+
+	add_timer(&ascb->timer);
+
+	res = asd_post_ascb_list(ascb->ha, ascb, 1);
+	if (unlikely(res))
+		del_timer(&ascb->timer);
+	return res;
+}
+
+static inline void asd_timedout_common(unsigned long data)
+{
+	struct asd_ascb *ascb = (void *) data;
+	struct asd_seq_data *seq = &ascb->ha->seq;
+        unsigned long flags;
+
+	spin_lock_irqsave(&seq->pend_q_lock, flags);
+        seq->pending--;
+        list_del_init(&ascb->list);
+        spin_unlock_irqrestore(&seq->pend_q_lock, flags);
+}
+
+/* ---------- CLEAR NEXUS ---------- */
+
+static void asd_clear_nexus_tasklet_complete(struct asd_ascb *ascb,
+					     struct done_list_struct *dl)
+{
+	ASD_DPRINTK("%s: here\n", __FUNCTION__);
+	if (!del_timer(&ascb->timer)) {
+		ASD_DPRINTK("%s: couldn't delete timer\n", __FUNCTION__);
+		return;
+	}
+	ASD_DPRINTK("%s: opcode: 0x%x\n", __FUNCTION__, dl->opcode);
+	ascb->uldd_task = (void *) (unsigned long) dl->opcode;
+	complete(&ascb->completion);
+}
+
+static void asd_clear_nexus_timedout(unsigned long data)
+{
+	struct asd_ascb *ascb = (void *) data;
+
+	ASD_DPRINTK("%s: here\n", __FUNCTION__);
+	asd_timedout_common(data);
+	ascb->uldd_task = (void *) TMF_RESP_FUNC_FAILED;
+	complete(&ascb->completion);
+}
+
+#define CLEAR_NEXUS_PRE         \
+	ASD_DPRINTK("%s: PRE\n", __FUNCTION__); \
+        res = 1;                \
+	ascb = asd_ascb_alloc_list(asd_ha, &res, GFP_KERNEL); \
+	if (!ascb)              \
+		return -ENOMEM; \
+                                \
+	scb = ascb->scb;        \
+	scb->header.opcode = CLEAR_NEXUS
+
+#define CLEAR_NEXUS_POST        \
+	ASD_DPRINTK("%s: POST\n", __FUNCTION__); \
+	res = asd_enqueue_internal(ascb, asd_clear_nexus_tasklet_complete, \
+				   asd_clear_nexus_timedout);              \
+	if (res)                \
+		goto out_err;   \
+	ASD_DPRINTK("%s: clear nexus posted, waiting...\n", __FUNCTION__); \
+	wait_for_completion(&ascb->completion); \
+	res = (int) (unsigned long) ascb->uldd_task; \
+	if (res == TC_NO_ERROR) \
+		res = TMF_RESP_FUNC_COMPLETE;   \
+out_err:                        \
+	asd_ascb_free(ascb);    \
+	return res
+
+int asd_clear_nexus_ha(struct sas_ha_struct *sas_ha)
+{
+	struct asd_ha_struct *asd_ha = sas_ha->lldd_ha;
+	struct asd_ascb *ascb;
+	struct scb *scb;
+	int res;
+
+	CLEAR_NEXUS_PRE;
+	scb->clear_nexus.nexus = NEXUS_ADAPTER;
+	CLEAR_NEXUS_POST;
+}
+
+int asd_clear_nexus_port(struct asd_sas_port *port)
+{
+	struct asd_ha_struct *asd_ha = port->ha->lldd_ha;
+	struct asd_ascb *ascb;
+	struct scb *scb;
+	int res;
+
+	CLEAR_NEXUS_PRE;
+	scb->clear_nexus.nexus = NEXUS_PORT;
+	scb->clear_nexus.conn_mask = port->phy_mask;
+	CLEAR_NEXUS_POST;
+}
+
+#if 0
+static int asd_clear_nexus_I_T(struct domain_device *dev)
+{
+	struct asd_ha_struct *asd_ha = dev->port->ha->lldd_ha;
+	struct asd_ascb *ascb;
+	struct scb *scb;
+	int res;
+
+	CLEAR_NEXUS_PRE;
+	scb->clear_nexus.nexus = NEXUS_I_T;
+	scb->clear_nexus.flags = SEND_Q | EXEC_Q | NOTINQ;
+	if (dev->tproto)
+		scb->clear_nexus.flags |= SUSPEND_TX;
+	scb->clear_nexus.conn_handle = cpu_to_le16((u16)(unsigned long)
+						   dev->lldd_dev);
+	CLEAR_NEXUS_POST;
+}
+#endif
+
+static int asd_clear_nexus_I_T_L(struct domain_device *dev, u8 *lun)
+{
+	struct asd_ha_struct *asd_ha = dev->port->ha->lldd_ha;
+	struct asd_ascb *ascb;
+	struct scb *scb;
+	int res;
+
+	CLEAR_NEXUS_PRE;
+	scb->clear_nexus.nexus = NEXUS_I_T_L;
+	scb->clear_nexus.flags = SEND_Q | EXEC_Q | NOTINQ;
+	if (dev->tproto)
+		scb->clear_nexus.flags |= SUSPEND_TX;
+	memcpy(scb->clear_nexus.ssp_task.lun, lun, 8);
+	scb->clear_nexus.conn_handle = cpu_to_le16((u16)(unsigned long)
+						   dev->lldd_dev);
+	CLEAR_NEXUS_POST;
+}
+
+static int asd_clear_nexus_tag(struct sas_task *task)
+{
+	struct asd_ha_struct *asd_ha = task->dev->port->ha->lldd_ha;
+	struct asd_ascb *tascb = task->lldd_task;
+	struct asd_ascb *ascb;
+	struct scb *scb;
+	int res;
+
+	CLEAR_NEXUS_PRE;
+	scb->clear_nexus.nexus = NEXUS_TAG;
+	memcpy(scb->clear_nexus.ssp_task.lun, task->ssp_task.LUN, 8);
+	scb->clear_nexus.ssp_task.tag = tascb->tag;
+	if (task->dev->tproto)
+		scb->clear_nexus.conn_handle = cpu_to_le16((u16)(unsigned long)
+							  task->dev->lldd_dev);
+	CLEAR_NEXUS_POST;
+}
+
+static int asd_clear_nexus_index(struct sas_task *task)
+{
+	struct asd_ha_struct *asd_ha = task->dev->port->ha->lldd_ha;
+	struct asd_ascb *tascb = task->lldd_task;
+	struct asd_ascb *ascb;
+	struct scb *scb;
+	int res;
+
+	CLEAR_NEXUS_PRE;
+	scb->clear_nexus.nexus = NEXUS_TRANS_CX;
+	if (task->dev->tproto)
+		scb->clear_nexus.conn_handle = cpu_to_le16((u16)(unsigned long)
+							  task->dev->lldd_dev);
+	scb->clear_nexus.index = cpu_to_le16(tascb->tc_index);
+	CLEAR_NEXUS_POST;
+}
+
+/* ---------- TMFs ---------- */
+
+static void asd_tmf_timedout(unsigned long data)
+{
+	struct asd_ascb *ascb = (void *) data;
+
+	ASD_DPRINTK("tmf timed out\n");
+	asd_timedout_common(data);
+	ascb->uldd_task = (void *) TMF_RESP_FUNC_FAILED;
+	complete(&ascb->completion);
+}
+
+static int asd_get_tmf_resp_tasklet(struct asd_ascb *ascb,
+				    struct done_list_struct *dl)
+{
+	struct asd_ha_struct *asd_ha = ascb->ha;
+	unsigned long flags;
+	struct tc_resp_sb_struct {
+		__le16 index_escb;
+		u8     len_lsb;
+		u8     flags;
+	} __attribute__ ((packed)) *resp_sb = (void *) dl->status_block;
+
+	int  edb_id = ((resp_sb->flags & 0x70) >> 4)-1;
+	struct asd_ascb *escb;
+	struct asd_dma_tok *edb;
+	struct ssp_frame_hdr *fh;
+	struct ssp_response_iu   *ru;
+	int res = TMF_RESP_FUNC_FAILED;
+
+	ASD_DPRINTK("tmf resp tasklet\n");
+
+	spin_lock_irqsave(&asd_ha->seq.tc_index_lock, flags);
+	escb = asd_tc_index_find(&asd_ha->seq,
+				 (int)le16_to_cpu(resp_sb->index_escb));
+	spin_unlock_irqrestore(&asd_ha->seq.tc_index_lock, flags);
+
+	if (!escb) {
+		ASD_DPRINTK("Uh-oh! No escb for this dl?!\n");
+		return res;
+	}
+
+	edb = asd_ha->seq.edb_arr[edb_id + escb->edb_index];
+	ascb->tag = *(__be16 *)(edb->vaddr+4);
+	fh = edb->vaddr + 16;
+	ru = edb->vaddr + 16 + sizeof(*fh);
+	res = ru->status;
+	if (ru->datapres == 1)	  /* Response data present */
+		res = ru->resp_data[3];
+#if 0
+	ascb->tag = fh->tag;
+#endif
+	ascb->tag_valid = 1;
+
+	asd_invalidate_edb(escb, edb_id);
+	return res;
+}
+
+static void asd_tmf_tasklet_complete(struct asd_ascb *ascb,
+				     struct done_list_struct *dl)
+{
+	if (!del_timer(&ascb->timer))
+		return;
+
+	ASD_DPRINTK("tmf tasklet complete\n");
+
+	if (dl->opcode == TC_SSP_RESP)
+		ascb->uldd_task = (void *) (unsigned long)
+			asd_get_tmf_resp_tasklet(ascb, dl);
+	else
+		ascb->uldd_task = (void *) 0xFF00 + (unsigned long) dl->opcode;
+
+	complete(&ascb->completion);
+}
+
+static inline int asd_clear_nexus(struct sas_task *task)
+{
+	int res = TMF_RESP_FUNC_FAILED;
+	struct asd_ascb *tascb = task->lldd_task;
+	unsigned long flags;
+
+	ASD_DPRINTK("task not done, clearing nexus\n");
+	if (tascb->tag_valid)
+		res = asd_clear_nexus_tag(task);
+	else
+		res = asd_clear_nexus_index(task);
+	wait_for_completion_timeout(&tascb->completion,
+				    AIC94XX_SCB_TIMEOUT);
+	ASD_DPRINTK("came back from clear nexus\n");
+	spin_lock_irqsave(&task->task_state_lock, flags);
+	if (task->task_state_flags & SAS_TASK_STATE_DONE)
+		res = TMF_RESP_FUNC_COMPLETE;
+	spin_unlock_irqrestore(&task->task_state_lock, flags);
+
+	return res;
+}
+
+/**
+ * asd_abort_task -- ABORT TASK TMF
+ * @task: the task to be aborted
+ *
+ * Before calling ABORT TASK the task state flags should be ORed with
+ * SAS_TASK_STATE_ABORTED (unless SAS_TASK_STATE_DONE is set) under
+ * the task_state_lock IRQ spinlock, then ABORT TASK *must* be called.
+ *
+ * Implements the ABORT TASK TMF, I_T_L_Q nexus.
+ * Returns: SAS TMF responses (see sas_task.h),
+ *          -ENOMEM,
+ *          -SAS_QUEUE_FULL.
+ *
+ * When ABORT TASK returns, the caller of ABORT TASK checks first the
+ * task->task_state_flags, and then the return value of ABORT TASK.
+ *
+ * If the task has task state bit SAS_TASK_STATE_DONE set, then the
+ * task was completed successfully prior to it being aborted.  The
+ * caller of ABORT TASK has responsibility to call task->task_done()
+ * xor free the task, depending on their framework.  The return code
+ * is TMF_RESP_FUNC_FAILED in this case.
+ *
+ * Else the SAS_TASK_STATE_DONE bit is not set,
+ * 	If the return code is TMF_RESP_FUNC_COMPLETE, then
+ * 		the task was aborted successfully.  The caller of
+ * 		ABORT TASK has responsibility to call task->task_done()
+ *              to finish the task, xor free the task depending on their
+ *		framework.
+ *	else
+ * 		the ABORT TASK returned some kind of error. The task
+ *              was _not_ cancelled.  Nothing can be assumed.
+ *		The caller of ABORT TASK may wish to retry.
+ */
+int asd_abort_task(struct sas_task *task)
+{
+	struct asd_ascb *tascb = task->lldd_task;
+	struct asd_ha_struct *asd_ha = tascb->ha;
+	int res = 1;
+	unsigned long flags;
+	struct asd_ascb *ascb = NULL;
+	struct scb *scb;
+
+	spin_lock_irqsave(&task->task_state_lock, flags);
+	if (task->task_state_flags & SAS_TASK_STATE_DONE) {
+		spin_unlock_irqrestore(&task->task_state_lock, flags);
+		res = TMF_RESP_FUNC_COMPLETE;
+		ASD_DPRINTK("%s: task 0x%p done\n", __FUNCTION__, task);
+		goto out_done;
+	}
+	spin_unlock_irqrestore(&task->task_state_lock, flags);
+
+	ascb = asd_ascb_alloc_list(asd_ha, &res, GFP_KERNEL);
+	if (!ascb)
+		return -ENOMEM;
+	scb = ascb->scb;
+
+	scb->header.opcode = ABORT_TASK;
+
+	switch (task->task_proto) {
+	case SATA_PROTO:
+	case SAS_PROTO_STP:
+		scb->abort_task.proto_conn_rate = (1 << 5); /* STP */
+		break;
+	case SAS_PROTO_SSP:
+		scb->abort_task.proto_conn_rate  = (1 << 4); /* SSP */
+		scb->abort_task.proto_conn_rate |= task->dev->linkrate;
+		break;
+	case SAS_PROTO_SMP:
+		break;
+	default:
+		break;
+	}
+
+	if (task->task_proto == SAS_PROTO_SSP) {
+		scb->abort_task.ssp_frame.frame_type = SSP_TASK;
+		memcpy(scb->abort_task.ssp_frame.hashed_dest_addr,
+		       task->dev->hashed_sas_addr, HASHED_SAS_ADDR_SIZE);
+		memcpy(scb->abort_task.ssp_frame.hashed_src_addr,
+		       task->dev->port->ha->hashed_sas_addr,
+		       HASHED_SAS_ADDR_SIZE);
+		scb->abort_task.ssp_frame.tptt = cpu_to_be16(0xFFFF);
+
+		memcpy(scb->abort_task.ssp_task.lun, task->ssp_task.LUN, 8);
+		scb->abort_task.ssp_task.tmf = TMF_ABORT_TASK;
+		scb->abort_task.ssp_task.tag = cpu_to_be16(0xFFFF);
+	}
+
+	scb->abort_task.sister_scb = cpu_to_le16(0xFFFF);
+	scb->abort_task.conn_handle = cpu_to_le16(
+		(u16)(unsigned long)task->dev->lldd_dev);
+	scb->abort_task.retry_count = 1;
+	scb->abort_task.index = cpu_to_le16((u16)tascb->tc_index);
+	scb->abort_task.itnl_to = cpu_to_le16(ITNL_TIMEOUT_CONST);
+
+	res = asd_enqueue_internal(ascb, asd_tmf_tasklet_complete,
+				   asd_tmf_timedout);
+	if (res)
+		goto out;
+	wait_for_completion(&ascb->completion);
+	ASD_DPRINTK("tmf came back\n");
+
+	res = (int) (unsigned long) ascb->uldd_task;
+	tascb->tag = ascb->tag;
+	tascb->tag_valid = ascb->tag_valid;
+
+	spin_lock_irqsave(&task->task_state_lock, flags);
+	if (task->task_state_flags & SAS_TASK_STATE_DONE) {
+		spin_unlock_irqrestore(&task->task_state_lock, flags);
+		res = TMF_RESP_FUNC_COMPLETE;
+		ASD_DPRINTK("%s: task 0x%p done\n", __FUNCTION__, task);
+		goto out_done;
+	}
+	spin_unlock_irqrestore(&task->task_state_lock, flags);
+
+	switch (res) {
+	/* The task to be aborted has been sent to the device.
+	 * We got a Response IU for the ABORT TASK TMF. */
+	case TC_NO_ERROR + 0xFF00:
+	case TMF_RESP_FUNC_COMPLETE:
+	case TMF_RESP_FUNC_FAILED:
+		res = asd_clear_nexus(task);
+		break;
+	case TMF_RESP_INVALID_FRAME:
+	case TMF_RESP_OVERLAPPED_TAG:
+	case TMF_RESP_FUNC_ESUPP:
+	case TMF_RESP_NO_LUN:
+		goto out_done; break;
+	}
+	/* In the following we assume that the managing layer
+	 * will _never_ make a mistake, when issuing ABORT TASK.
+	 */
+	switch (res) {
+	default:
+		res = asd_clear_nexus(task);
+		/* fallthrough */
+	case TC_NO_ERROR + 0xFF00:
+	case TMF_RESP_FUNC_COMPLETE:
+		break;
+	/* The task hasn't been sent to the device xor we never got
+	 * a (sane) Response IU for the ABORT TASK TMF.
+	 */
+	case TF_NAK_RECV + 0xFF00:
+		res = TMF_RESP_INVALID_FRAME;
+		break;
+	case TF_TMF_TASK_DONE + 0xFF00:	/* done but not reported yet */
+		res = TMF_RESP_FUNC_FAILED;
+		wait_for_completion_timeout(&tascb->completion,
+					    AIC94XX_SCB_TIMEOUT);
+		spin_lock_irqsave(&task->task_state_lock, flags);
+		if (task->task_state_flags & SAS_TASK_STATE_DONE)
+			res = TMF_RESP_FUNC_COMPLETE;
+		spin_unlock_irqrestore(&task->task_state_lock, flags);
+		goto out_done;
+	case TF_TMF_NO_TAG + 0xFF00:
+	case TF_TMF_TAG_FREE + 0xFF00: /* the tag is in the free list */
+	case TF_TMF_NO_CONN_HANDLE + 0xFF00: /* no such device */
+		res = TMF_RESP_FUNC_COMPLETE;
+		goto out_done;
+	case TF_TMF_NO_CTX + 0xFF00: /* not in seq, or proto != SSP */
+		res = TMF_RESP_FUNC_ESUPP;
+		goto out;
+	}
+out_done:
+	if (res == TMF_RESP_FUNC_COMPLETE) {
+		task->lldd_task = NULL;
+		mb();
+		asd_ascb_free(tascb);
+	}
+out:
+	asd_ascb_free(ascb);
+	ASD_DPRINTK("task 0x%p aborted, res: 0x%x\n", task, res);
+	return res;
+}
+
+/**
+ * asd_initiate_ssp_tmf -- send a TMF to an I_T_L or I_T_L_Q nexus
+ * @dev: pointer to struct domain_device of interest
+ * @lun: pointer to u8[8] which is the LUN
+ * @tmf: the TMF to be performed (see sas_task.h or the SAS spec)
+ * @index: the transaction context of the task to be queried if QT TMF
+ *
+ * This function is used to send ABORT TASK SET, CLEAR ACA,
+ * CLEAR TASK SET, LU RESET and QUERY TASK TMFs.
+ *
+ * No SCBs should be queued to the I_T_L nexus when this SCB is
+ * pending.
+ *
+ * Returns: TMF response code (see sas_task.h or the SAS spec)
+ */
+static int asd_initiate_ssp_tmf(struct domain_device *dev, u8 *lun,
+				int tmf, int index)
+{
+	struct asd_ha_struct *asd_ha = dev->port->ha->lldd_ha;
+	struct asd_ascb *ascb;
+	int res = 1;
+	struct scb *scb;
+
+	if (!(dev->tproto & SAS_PROTO_SSP))
+		return TMF_RESP_FUNC_ESUPP;
+
+	ascb = asd_ascb_alloc_list(asd_ha, &res, GFP_KERNEL);
+	if (!ascb)
+		return -ENOMEM;
+	scb = ascb->scb;
+
+	if (tmf == TMF_QUERY_TASK)
+		scb->header.opcode = QUERY_SSP_TASK;
+	else
+		scb->header.opcode = INITIATE_SSP_TMF;
+
+	scb->ssp_tmf.proto_conn_rate  = (1 << 4); /* SSP */
+	scb->ssp_tmf.proto_conn_rate |= dev->linkrate;
+	/* SSP frame header */
+	scb->ssp_tmf.ssp_frame.frame_type = SSP_TASK;
+	memcpy(scb->ssp_tmf.ssp_frame.hashed_dest_addr,
+	       dev->hashed_sas_addr, HASHED_SAS_ADDR_SIZE);
+	memcpy(scb->ssp_tmf.ssp_frame.hashed_src_addr,
+	       dev->port->ha->hashed_sas_addr, HASHED_SAS_ADDR_SIZE);
+	scb->ssp_tmf.ssp_frame.tptt = cpu_to_be16(0xFFFF);
+	/* SSP Task IU */
+	memcpy(scb->ssp_tmf.ssp_task.lun, lun, 8);
+	scb->ssp_tmf.ssp_task.tmf = tmf;
+
+	scb->ssp_tmf.sister_scb = cpu_to_le16(0xFFFF);
+	scb->ssp_tmf.conn_handle= cpu_to_le16((u16)(unsigned long)
+					      dev->lldd_dev);
+	scb->ssp_tmf.retry_count = 1;
+	scb->ssp_tmf.itnl_to = cpu_to_le16(ITNL_TIMEOUT_CONST);
+	if (tmf == TMF_QUERY_TASK)
+		scb->ssp_tmf.index = cpu_to_le16(index);
+
+	res = asd_enqueue_internal(ascb, asd_tmf_tasklet_complete,
+				   asd_tmf_timedout);
+	if (res)
+		goto out_err;
+	wait_for_completion(&ascb->completion);
+	res = (int) (unsigned long) ascb->uldd_task;
+
+	switch (res) {
+	case TC_NO_ERROR + 0xFF00:
+		res = TMF_RESP_FUNC_COMPLETE;
+		break;
+	case TF_NAK_RECV + 0xFF00:
+		res = TMF_RESP_INVALID_FRAME;
+		break;
+	case TF_TMF_TASK_DONE + 0xFF00:
+		res = TMF_RESP_FUNC_FAILED;
+		break;
+	case TF_TMF_NO_TAG + 0xFF00:
+	case TF_TMF_TAG_FREE + 0xFF00: /* the tag is in the free list */
+	case TF_TMF_NO_CONN_HANDLE + 0xFF00: /* no such device */
+		res = TMF_RESP_FUNC_COMPLETE;
+		break;
+	case TF_TMF_NO_CTX + 0xFF00: /* not in seq, or proto != SSP */
+		res = TMF_RESP_FUNC_ESUPP;
+		break;
+	default:
+		ASD_DPRINTK("%s: converting result 0x%x to TMF_RESP_FUNC_FAILED\n",
+			    __FUNCTION__, res);
+		res = TMF_RESP_FUNC_FAILED;
+		break;
+	}
+out_err:
+	asd_ascb_free(ascb);
+	return res;
+}
+
+int asd_abort_task_set(struct domain_device *dev, u8 *lun)
+{
+	int res = asd_initiate_ssp_tmf(dev, lun, TMF_ABORT_TASK_SET, 0);
+
+	if (res == TMF_RESP_FUNC_COMPLETE)
+		asd_clear_nexus_I_T_L(dev, lun);
+	return res;
+}
+
+int asd_clear_aca(struct domain_device *dev, u8 *lun)
+{
+	int res = asd_initiate_ssp_tmf(dev, lun, TMF_CLEAR_ACA, 0);
+
+	if (res == TMF_RESP_FUNC_COMPLETE)
+		asd_clear_nexus_I_T_L(dev, lun);
+	return res;
+}
+
+int asd_clear_task_set(struct domain_device *dev, u8 *lun)
+{
+	int res = asd_initiate_ssp_tmf(dev, lun, TMF_CLEAR_TASK_SET, 0);
+
+	if (res == TMF_RESP_FUNC_COMPLETE)
+		asd_clear_nexus_I_T_L(dev, lun);
+	return res;
+}
+
+int asd_lu_reset(struct domain_device *dev, u8 *lun)
+{
+	int res = asd_initiate_ssp_tmf(dev, lun, TMF_LU_RESET, 0);
+
+	if (res == TMF_RESP_FUNC_COMPLETE)
+		asd_clear_nexus_I_T_L(dev, lun);
+	return res;
+}
+
+/**
+ * asd_query_task -- send a QUERY TASK TMF to an I_T_L_Q nexus
+ * task: pointer to sas_task struct of interest
+ *
+ * Returns: TMF_RESP_FUNC_COMPLETE if the task is not in the task set,
+ * or TMF_RESP_FUNC_SUCC if the task is in the task set.
+ *
+ * Normally the management layer sets the task to aborted state,
+ * and then calls query task and then abort task.
+ */
+int asd_query_task(struct sas_task *task)
+{
+	struct asd_ascb *ascb = task->lldd_task;
+	int index;
+
+	if (ascb) {
+		index = ascb->tc_index;
+		return asd_initiate_ssp_tmf(task->dev, task->ssp_task.LUN,
+					    TMF_QUERY_TASK, index);
+	}
+	return TMF_RESP_FUNC_COMPLETE;
+}
diff --git a/drivers/scsi/libsas/Kconfig b/drivers/scsi/libsas/Kconfig
new file mode 100644
index 000000000000..aafdc92f8312
--- /dev/null
+++ b/drivers/scsi/libsas/Kconfig
@@ -0,0 +1,39 @@
+#
+# Kernel configuration file for the SAS Class
+#
+# Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+# Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+#
+# This file is licensed under GPLv2.
+#
+# This program is free software; you can redistribute it and/or
+# modify it under the terms of the GNU General Public License as
+# published by the Free Software Foundation; version 2 of the
+# License.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
+# USA
+#
+
+config SCSI_SAS_LIBSAS
+	tristate "SAS Domain Transport Attributes"
+	depends on SCSI
+	select SCSI_SAS_ATTRS
+	help
+	  This provides transport specific helpers for SAS drivers which
+	  use the domain device construct (like the aic94xxx).
+
+config SCSI_SAS_LIBSAS_DEBUG
+	bool "Compile the SAS Domain Transport Attributes in debug mode"
+	default y
+	depends on SCSI_SAS_LIBSAS
+	help
+		Compiles the SAS Layer in debug mode.  In debug mode, the
+		SAS Layer prints diagnostic and debug messages.
diff --git a/drivers/scsi/libsas/Makefile b/drivers/scsi/libsas/Makefile
new file mode 100644
index 000000000000..44d972a3b4bd
--- /dev/null
+++ b/drivers/scsi/libsas/Makefile
@@ -0,0 +1,36 @@
+#
+# Kernel Makefile for the libsas helpers
+#
+# Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+# Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+#
+# This file is licensed under GPLv2.
+#
+# This program is free software; you can redistribute it and/or
+# modify it under the terms of the GNU General Public License as
+# published by the Free Software Foundation; version 2 of the
+# License.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
+# USA
+
+ifeq ($(CONFIG_SCSI_SAS_LIBSAS_DEBUG),y)
+	EXTRA_CFLAGS += -DSAS_DEBUG
+endif
+
+obj-$(CONFIG_SCSI_SAS_LIBSAS) += libsas.o
+libsas-y +=  sas_init.o     \
+		sas_phy.o      \
+		sas_port.o     \
+		sas_event.o    \
+		sas_dump.o     \
+		sas_discover.o \
+		sas_expander.o \
+		sas_scsi_host.o
diff --git a/drivers/scsi/libsas/sas_discover.c b/drivers/scsi/libsas/sas_discover.c
new file mode 100644
index 000000000000..d977bd492d8d
--- /dev/null
+++ b/drivers/scsi/libsas/sas_discover.c
@@ -0,0 +1,749 @@
+/*
+ * Serial Attached SCSI (SAS) Discover process
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#include <linux/pci.h>
+#include <linux/scatterlist.h>
+#include <scsi/scsi_host.h>
+#include <scsi/scsi_eh.h>
+#include "sas_internal.h"
+
+#include <scsi/scsi_transport.h>
+#include <scsi/scsi_transport_sas.h>
+#include "../scsi_sas_internal.h"
+
+/* ---------- Basic task processing for discovery purposes ---------- */
+
+void sas_init_dev(struct domain_device *dev)
+{
+        INIT_LIST_HEAD(&dev->siblings);
+        INIT_LIST_HEAD(&dev->dev_list_node);
+        switch (dev->dev_type) {
+        case SAS_END_DEV:
+                break;
+        case EDGE_DEV:
+        case FANOUT_DEV:
+                INIT_LIST_HEAD(&dev->ex_dev.children);
+                break;
+        case SATA_DEV:
+        case SATA_PM:
+        case SATA_PM_PORT:
+                INIT_LIST_HEAD(&dev->sata_dev.children);
+                break;
+        default:
+                break;
+        }
+}
+
+static void sas_task_timedout(unsigned long _task)
+{
+	struct sas_task *task = (void *) _task;
+	unsigned long flags;
+
+	spin_lock_irqsave(&task->task_state_lock, flags);
+	if (!(task->task_state_flags & SAS_TASK_STATE_DONE))
+		task->task_state_flags |= SAS_TASK_STATE_ABORTED;
+	spin_unlock_irqrestore(&task->task_state_lock, flags);
+
+	complete(&task->completion);
+}
+
+static void sas_disc_task_done(struct sas_task *task)
+{
+	if (!del_timer(&task->timer))
+		return;
+	complete(&task->completion);
+}
+
+#define SAS_DEV_TIMEOUT 10
+
+/**
+ * sas_execute_task -- Basic task processing for discovery
+ * @task: the task to be executed
+ * @buffer: pointer to buffer to do I/O
+ * @size: size of @buffer
+ * @pci_dma_dir: PCI_DMA_...
+ */
+static int sas_execute_task(struct sas_task *task, void *buffer, int size,
+			    int pci_dma_dir)
+{
+	int res = 0;
+	struct scatterlist *scatter = NULL;
+	struct task_status_struct *ts = &task->task_status;
+	int num_scatter = 0;
+	int retries = 0;
+	struct sas_internal *i =
+		to_sas_internal(task->dev->port->ha->core.shost->transportt);
+
+	if (pci_dma_dir != PCI_DMA_NONE) {
+		scatter = kzalloc(sizeof(*scatter), GFP_KERNEL);
+		if (!scatter)
+			goto out;
+
+		sg_init_one(scatter, buffer, size);
+		num_scatter = 1;
+	}
+
+	task->task_proto = task->dev->tproto;
+	task->scatter = scatter;
+	task->num_scatter = num_scatter;
+	task->total_xfer_len = size;
+	task->data_dir = pci_dma_dir;
+	task->task_done = sas_disc_task_done;
+
+	for (retries = 0; retries < 5; retries++) {
+		task->task_state_flags = SAS_TASK_STATE_PENDING;
+		init_completion(&task->completion);
+
+		task->timer.data = (unsigned long) task;
+		task->timer.function = sas_task_timedout;
+		task->timer.expires = jiffies + SAS_DEV_TIMEOUT*HZ;
+		add_timer(&task->timer);
+
+		res = i->dft->lldd_execute_task(task, 1, GFP_KERNEL);
+		if (res) {
+			del_timer(&task->timer);
+			SAS_DPRINTK("executing SAS discovery task failed:%d\n",
+				    res);
+			goto ex_err;
+		}
+		wait_for_completion(&task->completion);
+		res = -ETASK;
+		if (task->task_state_flags & SAS_TASK_STATE_ABORTED) {
+			int res2;
+			SAS_DPRINTK("task aborted, flags:0x%x\n",
+				    task->task_state_flags);
+			res2 = i->dft->lldd_abort_task(task);
+			SAS_DPRINTK("came back from abort task\n");
+			if (!(task->task_state_flags & SAS_TASK_STATE_DONE)) {
+				if (res2 == TMF_RESP_FUNC_COMPLETE)
+					continue; /* Retry the task */
+				else
+					goto ex_err;
+			}
+		}
+		if (task->task_status.stat == SAM_BUSY ||
+			   task->task_status.stat == SAM_TASK_SET_FULL ||
+			   task->task_status.stat == SAS_QUEUE_FULL) {
+			SAS_DPRINTK("task: q busy, sleeping...\n");
+			schedule_timeout_interruptible(HZ);
+		} else if (task->task_status.stat == SAM_CHECK_COND) {
+			struct scsi_sense_hdr shdr;
+
+			if (!scsi_normalize_sense(ts->buf, ts->buf_valid_size,
+						  &shdr)) {
+				SAS_DPRINTK("couldn't normalize sense\n");
+				continue;
+			}
+			if ((shdr.sense_key == 6 && shdr.asc == 0x29) ||
+			    (shdr.sense_key == 2 && shdr.asc == 4 &&
+			     shdr.ascq == 1)) {
+				SAS_DPRINTK("device %016llx LUN: %016llx "
+					    "powering up or not ready yet, "
+					    "sleeping...\n",
+					    SAS_ADDR(task->dev->sas_addr),
+					    SAS_ADDR(task->ssp_task.LUN));
+
+				schedule_timeout_interruptible(5*HZ);
+			} else if (shdr.sense_key == 1) {
+				res = 0;
+				break;
+			} else if (shdr.sense_key == 5) {
+				break;
+			} else {
+				SAS_DPRINTK("dev %016llx LUN: %016llx "
+					    "sense key:0x%x ASC:0x%x ASCQ:0x%x"
+					    "\n",
+					    SAS_ADDR(task->dev->sas_addr),
+					    SAS_ADDR(task->ssp_task.LUN),
+					    shdr.sense_key,
+					    shdr.asc, shdr.ascq);
+			}
+		} else if (task->task_status.resp != SAS_TASK_COMPLETE ||
+			   task->task_status.stat != SAM_GOOD) {
+			SAS_DPRINTK("task finished with resp:0x%x, "
+				    "stat:0x%x\n",
+				    task->task_status.resp,
+				    task->task_status.stat);
+			goto ex_err;
+		} else {
+			res = 0;
+			break;
+		}
+	}
+ex_err:
+	if (pci_dma_dir != PCI_DMA_NONE)
+		kfree(scatter);
+out:
+	return res;
+}
+
+/* ---------- Domain device discovery ---------- */
+
+/**
+ * sas_get_port_device -- Discover devices which caused port creation
+ * @port: pointer to struct sas_port of interest
+ *
+ * Devices directly attached to a HA port, have no parent.  This is
+ * how we know they are (domain) "root" devices.  All other devices
+ * do, and should have their "parent" pointer set appropriately as
+ * soon as a child device is discovered.
+ */
+static int sas_get_port_device(struct asd_sas_port *port)
+{
+	unsigned long flags;
+	struct asd_sas_phy *phy;
+	struct sas_rphy *rphy;
+	struct domain_device *dev;
+
+	dev = kzalloc(sizeof(*dev), GFP_KERNEL);
+	if (!dev)
+		return -ENOMEM;
+
+	spin_lock_irqsave(&port->phy_list_lock, flags);
+	if (list_empty(&port->phy_list)) {
+		spin_unlock_irqrestore(&port->phy_list_lock, flags);
+		kfree(dev);
+		return -ENODEV;
+	}
+	phy = container_of(port->phy_list.next, struct asd_sas_phy, port_phy_el);
+	spin_lock(&phy->frame_rcvd_lock);
+	memcpy(dev->frame_rcvd, phy->frame_rcvd, min(sizeof(dev->frame_rcvd),
+					     (size_t)phy->frame_rcvd_size));
+	spin_unlock(&phy->frame_rcvd_lock);
+	spin_unlock_irqrestore(&port->phy_list_lock, flags);
+
+	if (dev->frame_rcvd[0] == 0x34 && port->oob_mode == SATA_OOB_MODE) {
+		struct dev_to_host_fis *fis =
+			(struct dev_to_host_fis *) dev->frame_rcvd;
+		if (fis->interrupt_reason == 1 && fis->lbal == 1 &&
+		    fis->byte_count_low==0x69 && fis->byte_count_high == 0x96
+		    && (fis->device & ~0x10) == 0)
+			dev->dev_type = SATA_PM;
+		else
+			dev->dev_type = SATA_DEV;
+		dev->tproto = SATA_PROTO;
+	} else {
+		struct sas_identify_frame *id =
+			(struct sas_identify_frame *) dev->frame_rcvd;
+		dev->dev_type = id->dev_type;
+		dev->iproto = id->initiator_bits;
+		dev->tproto = id->target_bits;
+	}
+
+	sas_init_dev(dev);
+
+	switch (dev->dev_type) {
+	case SAS_END_DEV:
+		rphy = sas_end_device_alloc(port->port);
+		break;
+	case EDGE_DEV:
+		rphy = sas_expander_alloc(port->port,
+					  SAS_EDGE_EXPANDER_DEVICE);
+		break;
+	case FANOUT_DEV:
+		rphy = sas_expander_alloc(port->port,
+					  SAS_FANOUT_EXPANDER_DEVICE);
+		break;
+	case SATA_DEV:
+	default:
+		printk("ERROR: Unidentified device type %d\n", dev->dev_type);
+		rphy = NULL;
+		break;
+	}
+
+	if (!rphy) {
+		kfree(dev);
+		return -ENODEV;
+	}
+	rphy->identify.phy_identifier = phy->phy->identify.phy_identifier;
+	memcpy(dev->sas_addr, port->attached_sas_addr, SAS_ADDR_SIZE);
+	sas_fill_in_rphy(dev, rphy);
+	sas_hash_addr(dev->hashed_sas_addr, dev->sas_addr);
+	port->port_dev = dev;
+	dev->port = port;
+	dev->linkrate = port->linkrate;
+	dev->min_linkrate = port->linkrate;
+	dev->max_linkrate = port->linkrate;
+	dev->pathways = port->num_phys;
+	memset(port->disc.fanout_sas_addr, 0, SAS_ADDR_SIZE);
+	memset(port->disc.eeds_a, 0, SAS_ADDR_SIZE);
+	memset(port->disc.eeds_b, 0, SAS_ADDR_SIZE);
+	port->disc.max_level = 0;
+
+	dev->rphy = rphy;
+	spin_lock(&port->dev_list_lock);
+	list_add_tail(&dev->dev_list_node, &port->dev_list);
+	spin_unlock(&port->dev_list_lock);
+
+	return 0;
+}
+
+/* ---------- Discover and Revalidate ---------- */
+
+/* ---------- SATA ---------- */
+
+static void sas_get_ata_command_set(struct domain_device *dev)
+{
+	struct dev_to_host_fis *fis =
+		(struct dev_to_host_fis *) dev->frame_rcvd;
+
+	if ((fis->sector_count == 1 && /* ATA */
+	     fis->lbal         == 1 &&
+	     fis->lbam         == 0 &&
+	     fis->lbah         == 0 &&
+	     fis->device       == 0)
+	    ||
+	    (fis->sector_count == 0 && /* CE-ATA (mATA) */
+	     fis->lbal         == 0 &&
+	     fis->lbam         == 0xCE &&
+	     fis->lbah         == 0xAA &&
+	     (fis->device & ~0x10) == 0))
+
+		dev->sata_dev.command_set = ATA_COMMAND_SET;
+
+	else if ((fis->interrupt_reason == 1 &&	/* ATAPI */
+		  fis->lbal             == 1 &&
+		  fis->byte_count_low   == 0x14 &&
+		  fis->byte_count_high  == 0xEB &&
+		  (fis->device & ~0x10) == 0))
+
+		dev->sata_dev.command_set = ATAPI_COMMAND_SET;
+
+	else if ((fis->sector_count == 1 && /* SEMB */
+		  fis->lbal         == 1 &&
+		  fis->lbam         == 0x3C &&
+		  fis->lbah         == 0xC3 &&
+		  fis->device       == 0)
+		||
+		 (fis->interrupt_reason == 1 &&	/* SATA PM */
+		  fis->lbal             == 1 &&
+		  fis->byte_count_low   == 0x69 &&
+		  fis->byte_count_high  == 0x96 &&
+		  (fis->device & ~0x10) == 0))
+
+		/* Treat it as a superset? */
+		dev->sata_dev.command_set = ATAPI_COMMAND_SET;
+}
+
+/**
+ * sas_issue_ata_cmd -- Basic SATA command processing for discovery
+ * @dev: the device to send the command to
+ * @command: the command register
+ * @features: the features register
+ * @buffer: pointer to buffer to do I/O
+ * @size: size of @buffer
+ * @pci_dma_dir: PCI_DMA_...
+ */
+static int sas_issue_ata_cmd(struct domain_device *dev, u8 command,
+			     u8 features, void *buffer, int size,
+			     int pci_dma_dir)
+{
+	int res = 0;
+	struct sas_task *task;
+	struct dev_to_host_fis *d2h_fis = (struct dev_to_host_fis *)
+		&dev->frame_rcvd[0];
+
+	res = -ENOMEM;
+	task = sas_alloc_task(GFP_KERNEL);
+	if (!task)
+		goto out;
+
+	task->dev = dev;
+
+	task->ata_task.fis.command = command;
+	task->ata_task.fis.features = features;
+	task->ata_task.fis.device = d2h_fis->device;
+	task->ata_task.retry_count = 1;
+
+	res = sas_execute_task(task, buffer, size, pci_dma_dir);
+
+	sas_free_task(task);
+out:
+	return res;
+}
+
+static void sas_sata_propagate_sas_addr(struct domain_device *dev)
+{
+	unsigned long flags;
+	struct asd_sas_port *port = dev->port;
+	struct asd_sas_phy  *phy;
+
+	BUG_ON(dev->parent);
+
+	memcpy(port->attached_sas_addr, dev->sas_addr, SAS_ADDR_SIZE);
+	spin_lock_irqsave(&port->phy_list_lock, flags);
+	list_for_each_entry(phy, &port->phy_list, port_phy_el)
+		memcpy(phy->attached_sas_addr, dev->sas_addr, SAS_ADDR_SIZE);
+	spin_unlock_irqrestore(&port->phy_list_lock, flags);
+}
+
+#define ATA_IDENTIFY_DEV         0xEC
+#define ATA_IDENTIFY_PACKET_DEV  0xA1
+#define ATA_SET_FEATURES         0xEF
+#define ATA_FEATURE_PUP_STBY_SPIN_UP 0x07
+
+/**
+ * sas_discover_sata_dev -- discover a STP/SATA device (SATA_DEV)
+ * @dev: STP/SATA device of interest (ATA/ATAPI)
+ *
+ * The LLDD has already been notified of this device, so that we can
+ * send FISes to it.  Here we try to get IDENTIFY DEVICE or IDENTIFY
+ * PACKET DEVICE, if ATAPI device, so that the LLDD can fine-tune its
+ * performance for this device.
+ */
+static int sas_discover_sata_dev(struct domain_device *dev)
+{
+	int     res;
+	__le16  *identify_x;
+	u8      command;
+
+	identify_x = kzalloc(512, GFP_KERNEL);
+	if (!identify_x)
+		return -ENOMEM;
+
+	if (dev->sata_dev.command_set == ATA_COMMAND_SET) {
+		dev->sata_dev.identify_device = identify_x;
+		command = ATA_IDENTIFY_DEV;
+	} else {
+		dev->sata_dev.identify_packet_device = identify_x;
+		command = ATA_IDENTIFY_PACKET_DEV;
+	}
+
+	res = sas_issue_ata_cmd(dev, command, 0, identify_x, 512,
+				PCI_DMA_FROMDEVICE);
+	if (res)
+		goto out_err;
+
+	/* lives on the media? */
+	if (le16_to_cpu(identify_x[0]) & 4) {
+		/* incomplete response */
+		SAS_DPRINTK("sending SET FEATURE/PUP_STBY_SPIN_UP to "
+			    "dev %llx\n", SAS_ADDR(dev->sas_addr));
+		if (!le16_to_cpu(identify_x[83] & (1<<6)))
+			goto cont1;
+		res = sas_issue_ata_cmd(dev, ATA_SET_FEATURES,
+					ATA_FEATURE_PUP_STBY_SPIN_UP,
+					NULL, 0, PCI_DMA_NONE);
+		if (res)
+			goto cont1;
+
+		schedule_timeout_interruptible(5*HZ); /* More time? */
+		res = sas_issue_ata_cmd(dev, command, 0, identify_x, 512,
+					PCI_DMA_FROMDEVICE);
+		if (res)
+			goto out_err;
+	}
+cont1:
+	/* Get WWN */
+	if (dev->port->oob_mode != SATA_OOB_MODE) {
+		memcpy(dev->sas_addr, dev->sata_dev.rps_resp.rps.stp_sas_addr,
+		       SAS_ADDR_SIZE);
+	} else if (dev->sata_dev.command_set == ATA_COMMAND_SET &&
+		   (le16_to_cpu(dev->sata_dev.identify_device[108]) & 0xF000)
+		   == 0x5000) {
+		int i;
+
+		for (i = 0; i < 4; i++) {
+			dev->sas_addr[2*i] =
+	     (le16_to_cpu(dev->sata_dev.identify_device[108+i]) & 0xFF00) >> 8;
+			dev->sas_addr[2*i+1] =
+	      le16_to_cpu(dev->sata_dev.identify_device[108+i]) & 0x00FF;
+		}
+	}
+	sas_hash_addr(dev->hashed_sas_addr, dev->sas_addr);
+	if (!dev->parent)
+		sas_sata_propagate_sas_addr(dev);
+
+	/* XXX Hint: register this SATA device with SATL.
+	   When this returns, dev->sata_dev->lu is alive and
+	   present.
+	sas_satl_register_dev(dev);
+	*/
+	return 0;
+out_err:
+	dev->sata_dev.identify_packet_device = NULL;
+	dev->sata_dev.identify_device = NULL;
+	kfree(identify_x);
+	return res;
+}
+
+static int sas_discover_sata_pm(struct domain_device *dev)
+{
+	return -ENODEV;
+}
+
+int sas_notify_lldd_dev_found(struct domain_device *dev)
+{
+	int res = 0;
+	struct sas_ha_struct *sas_ha = dev->port->ha;
+	struct Scsi_Host *shost = sas_ha->core.shost;
+	struct sas_internal *i = to_sas_internal(shost->transportt);
+
+	if (i->dft->lldd_dev_found) {
+		res = i->dft->lldd_dev_found(dev);
+		if (res) {
+			printk("sas: driver on pcidev %s cannot handle "
+			       "device %llx, error:%d\n",
+			       pci_name(sas_ha->pcidev),
+			       SAS_ADDR(dev->sas_addr), res);
+		}
+	}
+	return res;
+}
+
+
+void sas_notify_lldd_dev_gone(struct domain_device *dev)
+{
+	struct sas_ha_struct *sas_ha = dev->port->ha;
+	struct Scsi_Host *shost = sas_ha->core.shost;
+	struct sas_internal *i = to_sas_internal(shost->transportt);
+
+	if (i->dft->lldd_dev_gone)
+		i->dft->lldd_dev_gone(dev);
+}
+
+/* ---------- Common/dispatchers ---------- */
+
+/**
+ * sas_discover_sata -- discover an STP/SATA domain device
+ * @dev: pointer to struct domain_device of interest
+ *
+ * First we notify the LLDD of this device, so we can send frames to
+ * it.  Then depending on the type of device we call the appropriate
+ * discover functions.  Once device discover is done, we notify the
+ * LLDD so that it can fine-tune its parameters for the device, by
+ * removing it and then adding it.  That is, the second time around,
+ * the driver would have certain fields, that it is looking at, set.
+ * Finally we initialize the kobj so that the device can be added to
+ * the system at registration time.  Devices directly attached to a HA
+ * port, have no parents.  All other devices do, and should have their
+ * "parent" pointer set appropriately before calling this function.
+ */
+int sas_discover_sata(struct domain_device *dev)
+{
+	int res;
+
+	sas_get_ata_command_set(dev);
+
+	res = sas_notify_lldd_dev_found(dev);
+	if (res)
+		return res;
+
+	switch (dev->dev_type) {
+	case SATA_DEV:
+		res = sas_discover_sata_dev(dev);
+		break;
+	case SATA_PM:
+		res = sas_discover_sata_pm(dev);
+		break;
+	default:
+		break;
+	}
+
+	sas_notify_lldd_dev_gone(dev);
+	if (!res) {
+		sas_notify_lldd_dev_found(dev);
+	}
+	return res;
+}
+
+/**
+ * sas_discover_end_dev -- discover an end device (SSP, etc)
+ * @end: pointer to domain device of interest
+ *
+ * See comment in sas_discover_sata().
+ */
+int sas_discover_end_dev(struct domain_device *dev)
+{
+	int res;
+
+	res = sas_notify_lldd_dev_found(dev);
+	if (res)
+		return res;
+
+	res = sas_rphy_add(dev->rphy);
+	if (res)
+		goto out_err;
+
+	/* do this to get the end device port attributes which will have
+	 * been scanned in sas_rphy_add */
+	sas_notify_lldd_dev_gone(dev);
+	sas_notify_lldd_dev_found(dev);
+
+	return 0;
+
+out_err:
+	sas_notify_lldd_dev_gone(dev);
+	return res;
+}
+
+/* ---------- Device registration and unregistration ---------- */
+
+static inline void sas_unregister_common_dev(struct domain_device *dev)
+{
+	sas_notify_lldd_dev_gone(dev);
+	if (!dev->parent)
+		dev->port->port_dev = NULL;
+	else
+		list_del_init(&dev->siblings);
+	list_del_init(&dev->dev_list_node);
+}
+
+void sas_unregister_dev(struct domain_device *dev)
+{
+	if (dev->rphy) {
+		sas_remove_children(&dev->rphy->dev);
+		sas_rphy_delete(dev->rphy);
+		dev->rphy = NULL;
+	}
+	if (dev->dev_type == EDGE_DEV || dev->dev_type == FANOUT_DEV) {
+		/* remove the phys and ports, everything else should be gone */
+		kfree(dev->ex_dev.ex_phy);
+		dev->ex_dev.ex_phy = NULL;
+	}
+	sas_unregister_common_dev(dev);
+}
+
+void sas_unregister_domain_devices(struct asd_sas_port *port)
+{
+	struct domain_device *dev, *n;
+
+	list_for_each_entry_safe_reverse(dev,n,&port->dev_list,dev_list_node)
+		sas_unregister_dev(dev);
+
+	port->port->rphy = NULL;
+
+}
+
+/* ---------- Discovery and Revalidation ---------- */
+
+/**
+ * sas_discover_domain -- discover the domain
+ * @port: port to the domain of interest
+ *
+ * NOTE: this process _must_ quit (return) as soon as any connection
+ * errors are encountered.  Connection recovery is done elsewhere.
+ * Discover process only interrogates devices in order to discover the
+ * domain.
+ */
+static void sas_discover_domain(void *data)
+{
+	int error = 0;
+	struct asd_sas_port *port = data;
+
+	sas_begin_event(DISCE_DISCOVER_DOMAIN, &port->disc.disc_event_lock,
+			&port->disc.pending);
+
+	if (port->port_dev)
+		return ;
+	else {
+		error = sas_get_port_device(port);
+		if (error)
+			return;
+	}
+
+	SAS_DPRINTK("DOING DISCOVERY on port %d, pid:%d\n", port->id,
+		    current->pid);
+
+	switch (port->port_dev->dev_type) {
+	case SAS_END_DEV:
+		error = sas_discover_end_dev(port->port_dev);
+		break;
+	case EDGE_DEV:
+	case FANOUT_DEV:
+		error = sas_discover_root_expander(port->port_dev);
+		break;
+	case SATA_DEV:
+	case SATA_PM:
+		error = sas_discover_sata(port->port_dev);
+		break;
+	default:
+		SAS_DPRINTK("unhandled device %d\n", port->port_dev->dev_type);
+		break;
+	}
+
+	if (error) {
+		kfree(port->port_dev); /* not kobject_register-ed yet */
+		port->port_dev = NULL;
+	}
+
+	SAS_DPRINTK("DONE DISCOVERY on port %d, pid:%d, result:%d\n", port->id,
+		    current->pid, error);
+}
+
+static void sas_revalidate_domain(void *data)
+{
+	int res = 0;
+	struct asd_sas_port *port = data;
+
+	sas_begin_event(DISCE_REVALIDATE_DOMAIN, &port->disc.disc_event_lock,
+			&port->disc.pending);
+
+	SAS_DPRINTK("REVALIDATING DOMAIN on port %d, pid:%d\n", port->id,
+		    current->pid);
+	if (port->port_dev)
+		res = sas_ex_revalidate_domain(port->port_dev);
+
+	SAS_DPRINTK("done REVALIDATING DOMAIN on port %d, pid:%d, res 0x%x\n",
+		    port->id, current->pid, res);
+}
+
+/* ---------- Events ---------- */
+
+int sas_discover_event(struct asd_sas_port *port, enum discover_event ev)
+{
+	struct sas_discovery *disc;
+
+	if (!port)
+		return 0;
+	disc = &port->disc;
+
+	BUG_ON(ev >= DISC_NUM_EVENTS);
+
+	sas_queue_event(ev, &disc->disc_event_lock, &disc->pending,
+			&disc->disc_work[ev], port->ha->core.shost);
+
+	return 0;
+}
+
+/**
+ * sas_init_disc -- initialize the discovery struct in the port
+ * @port: pointer to struct port
+ *
+ * Called when the ports are being initialized.
+ */
+void sas_init_disc(struct sas_discovery *disc, struct asd_sas_port *port)
+{
+	int i;
+
+	static void (*sas_event_fns[DISC_NUM_EVENTS])(void *) = {
+		[DISCE_DISCOVER_DOMAIN] = sas_discover_domain,
+		[DISCE_REVALIDATE_DOMAIN] = sas_revalidate_domain,
+	};
+
+	spin_lock_init(&disc->disc_event_lock);
+	disc->pending = 0;
+	for (i = 0; i < DISC_NUM_EVENTS; i++)
+		INIT_WORK(&disc->disc_work[i], sas_event_fns[i], port);
+}
diff --git a/drivers/scsi/libsas/sas_dump.c b/drivers/scsi/libsas/sas_dump.c
new file mode 100644
index 000000000000..f1246d2c9bef
--- /dev/null
+++ b/drivers/scsi/libsas/sas_dump.c
@@ -0,0 +1,76 @@
+/*
+ * Serial Attached SCSI (SAS) Dump/Debugging routines
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#include "sas_dump.h"
+
+#ifdef SAS_DEBUG
+
+static const char *sas_hae_str[] = {
+	[0] = "HAE_RESET",
+};
+
+static const char *sas_porte_str[] = {
+	[0] = "PORTE_BYTES_DMAED",
+	[1] = "PORTE_BROADCAST_RCVD",
+	[2] = "PORTE_LINK_RESET_ERR",
+	[3] = "PORTE_TIMER_EVENT",
+	[4] = "PORTE_HARD_RESET",
+};
+
+static const char *sas_phye_str[] = {
+	[0] = "PHYE_LOSS_OF_SIGNAL",
+	[1] = "PHYE_OOB_DONE",
+	[2] = "PHYE_OOB_ERROR",
+	[3] = "PHYE_SPINUP_HOLD",
+};
+
+void sas_dprint_porte(int phyid, enum port_event pe)
+{
+	SAS_DPRINTK("phy%d: port event: %s\n", phyid, sas_porte_str[pe]);
+}
+void sas_dprint_phye(int phyid, enum phy_event pe)
+{
+	SAS_DPRINTK("phy%d: phy event: %s\n", phyid, sas_phye_str[pe]);
+}
+
+void sas_dprint_hae(struct sas_ha_struct *sas_ha, enum ha_event he)
+{
+	SAS_DPRINTK("ha %s: %s event\n", pci_name(sas_ha->pcidev),
+		    sas_hae_str[he]);
+}
+
+void sas_dump_port(struct asd_sas_port *port)
+{
+	SAS_DPRINTK("port%d: class:0x%x\n", port->id, port->class);
+	SAS_DPRINTK("port%d: sas_addr:%llx\n", port->id,
+		    SAS_ADDR(port->sas_addr));
+	SAS_DPRINTK("port%d: attached_sas_addr:%llx\n", port->id,
+		    SAS_ADDR(port->attached_sas_addr));
+	SAS_DPRINTK("port%d: iproto:0x%x\n", port->id, port->iproto);
+	SAS_DPRINTK("port%d: tproto:0x%x\n", port->id, port->tproto);
+	SAS_DPRINTK("port%d: oob_mode:0x%x\n", port->id, port->oob_mode);
+	SAS_DPRINTK("port%d: num_phys:%d\n", port->id, port->num_phys);
+}
+
+#endif /* SAS_DEBUG */
diff --git a/drivers/scsi/libsas/sas_dump.h b/drivers/scsi/libsas/sas_dump.h
new file mode 100644
index 000000000000..47b45d4f5258
--- /dev/null
+++ b/drivers/scsi/libsas/sas_dump.h
@@ -0,0 +1,42 @@
+/*
+ * Serial Attached SCSI (SAS) Dump/Debugging routines header file
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#include "sas_internal.h"
+
+#ifdef SAS_DEBUG
+
+void sas_dprint_porte(int phyid, enum port_event pe);
+void sas_dprint_phye(int phyid, enum phy_event pe);
+void sas_dprint_hae(struct sas_ha_struct *sas_ha, enum ha_event he);
+void sas_dump_port(struct asd_sas_port *port);
+
+#else /* SAS_DEBUG */
+
+static inline void sas_dprint_porte(int phyid, enum port_event pe) { }
+static inline void sas_dprint_phye(int phyid, enum phy_event pe) { }
+static inline void sas_dprint_hae(struct sas_ha_struct *sas_ha,
+				  enum ha_event he) { }
+static inline void sas_dump_port(struct asd_sas_port *port) { }
+
+#endif /* SAS_DEBUG */
diff --git a/drivers/scsi/libsas/sas_event.c b/drivers/scsi/libsas/sas_event.c
new file mode 100644
index 000000000000..19110ed1c89c
--- /dev/null
+++ b/drivers/scsi/libsas/sas_event.c
@@ -0,0 +1,75 @@
+/*
+ * Serial Attached SCSI (SAS) Event processing
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#include <scsi/scsi_host.h>
+#include "sas_internal.h"
+#include "sas_dump.h"
+
+static void notify_ha_event(struct sas_ha_struct *sas_ha, enum ha_event event)
+{
+	BUG_ON(event >= HA_NUM_EVENTS);
+
+	sas_queue_event(event, &sas_ha->event_lock, &sas_ha->pending,
+			&sas_ha->ha_events[event], sas_ha->core.shost);
+}
+
+static void notify_port_event(struct asd_sas_phy *phy, enum port_event event)
+{
+	struct sas_ha_struct *ha = phy->ha;
+
+	BUG_ON(event >= PORT_NUM_EVENTS);
+
+	sas_queue_event(event, &ha->event_lock, &phy->port_events_pending,
+			&phy->port_events[event], ha->core.shost);
+}
+
+static void notify_phy_event(struct asd_sas_phy *phy, enum phy_event event)
+{
+	struct sas_ha_struct *ha = phy->ha;
+
+	BUG_ON(event >= PHY_NUM_EVENTS);
+
+	sas_queue_event(event, &ha->event_lock, &phy->phy_events_pending,
+			&phy->phy_events[event], ha->core.shost);
+}
+
+int sas_init_events(struct sas_ha_struct *sas_ha)
+{
+	static void (*sas_ha_event_fns[HA_NUM_EVENTS])(void *) = {
+		[HAE_RESET] = sas_hae_reset,
+	};
+
+	int i;
+
+	spin_lock_init(&sas_ha->event_lock);
+
+	for (i = 0; i < HA_NUM_EVENTS; i++)
+		INIT_WORK(&sas_ha->ha_events[i], sas_ha_event_fns[i], sas_ha);
+
+	sas_ha->notify_ha_event = notify_ha_event;
+	sas_ha->notify_port_event = notify_port_event;
+	sas_ha->notify_phy_event = notify_phy_event;
+
+	return 0;
+}
diff --git a/drivers/scsi/libsas/sas_expander.c b/drivers/scsi/libsas/sas_expander.c
new file mode 100644
index 000000000000..b653a263f76a
--- /dev/null
+++ b/drivers/scsi/libsas/sas_expander.c
@@ -0,0 +1,1862 @@
+/*
+ * Serial Attached SCSI (SAS) Expander discovery and configuration
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#include <linux/pci.h>
+#include <linux/scatterlist.h>
+
+#include "sas_internal.h"
+
+#include <scsi/scsi_transport.h>
+#include <scsi/scsi_transport_sas.h>
+#include "../scsi_sas_internal.h"
+
+static int sas_discover_expander(struct domain_device *dev);
+static int sas_configure_routing(struct domain_device *dev, u8 *sas_addr);
+static int sas_configure_phy(struct domain_device *dev, int phy_id,
+			     u8 *sas_addr, int include);
+static int sas_disable_routing(struct domain_device *dev,  u8 *sas_addr);
+
+#if 0
+/* FIXME: smp needs to migrate into the sas class */
+static ssize_t smp_portal_read(struct kobject *, char *, loff_t, size_t);
+static ssize_t smp_portal_write(struct kobject *, char *, loff_t, size_t);
+#endif
+
+/* ---------- SMP task management ---------- */
+
+static void smp_task_timedout(unsigned long _task)
+{
+	struct sas_task *task = (void *) _task;
+	unsigned long flags;
+
+	spin_lock_irqsave(&task->task_state_lock, flags);
+	if (!(task->task_state_flags & SAS_TASK_STATE_DONE))
+		task->task_state_flags |= SAS_TASK_STATE_ABORTED;
+	spin_unlock_irqrestore(&task->task_state_lock, flags);
+
+	complete(&task->completion);
+}
+
+static void smp_task_done(struct sas_task *task)
+{
+	if (!del_timer(&task->timer))
+		return;
+	complete(&task->completion);
+}
+
+/* Give it some long enough timeout. In seconds. */
+#define SMP_TIMEOUT 10
+
+static int smp_execute_task(struct domain_device *dev, void *req, int req_size,
+			    void *resp, int resp_size)
+{
+	int res;
+	struct sas_task *task = sas_alloc_task(GFP_KERNEL);
+	struct sas_internal *i =
+		to_sas_internal(dev->port->ha->core.shost->transportt);
+
+	if (!task)
+		return -ENOMEM;
+
+	task->dev = dev;
+	task->task_proto = dev->tproto;
+	sg_init_one(&task->smp_task.smp_req, req, req_size);
+	sg_init_one(&task->smp_task.smp_resp, resp, resp_size);
+
+	task->task_done = smp_task_done;
+
+	task->timer.data = (unsigned long) task;
+	task->timer.function = smp_task_timedout;
+	task->timer.expires = jiffies + SMP_TIMEOUT*HZ;
+	add_timer(&task->timer);
+
+	res = i->dft->lldd_execute_task(task, 1, GFP_KERNEL);
+
+	if (res) {
+		del_timer(&task->timer);
+		SAS_DPRINTK("executing SMP task failed:%d\n", res);
+		goto ex_err;
+	}
+
+	wait_for_completion(&task->completion);
+	res = -ETASK;
+	if ((task->task_state_flags & SAS_TASK_STATE_ABORTED)) {
+		SAS_DPRINTK("smp task timed out or aborted\n");
+		i->dft->lldd_abort_task(task);
+		if (!(task->task_state_flags & SAS_TASK_STATE_DONE)) {
+			SAS_DPRINTK("SMP task aborted and not done\n");
+			goto ex_err;
+		}
+	}
+	if (task->task_status.resp == SAS_TASK_COMPLETE &&
+	    task->task_status.stat == SAM_GOOD)
+		res = 0;
+	else
+		SAS_DPRINTK("%s: task to dev %016llx response: 0x%x "
+			    "status 0x%x\n", __FUNCTION__,
+			    SAS_ADDR(dev->sas_addr),
+			    task->task_status.resp,
+			    task->task_status.stat);
+ex_err:
+	sas_free_task(task);
+	return res;
+}
+
+/* ---------- Allocations ---------- */
+
+static inline void *alloc_smp_req(int size)
+{
+	u8 *p = kzalloc(size, GFP_KERNEL);
+	if (p)
+		p[0] = SMP_REQUEST;
+	return p;
+}
+
+static inline void *alloc_smp_resp(int size)
+{
+	return kzalloc(size, GFP_KERNEL);
+}
+
+/* ---------- Expander configuration ---------- */
+
+static void sas_set_ex_phy(struct domain_device *dev, int phy_id,
+			   void *disc_resp)
+{
+	struct expander_device *ex = &dev->ex_dev;
+	struct ex_phy *phy = &ex->ex_phy[phy_id];
+	struct smp_resp *resp = disc_resp;
+	struct discover_resp *dr = &resp->disc;
+	struct sas_rphy *rphy = dev->rphy;
+	int rediscover = (phy->phy != NULL);
+
+	if (!rediscover) {
+		phy->phy = sas_phy_alloc(&rphy->dev, phy_id);
+
+		/* FIXME: error_handling */
+		BUG_ON(!phy->phy);
+	}
+
+	switch (resp->result) {
+	case SMP_RESP_PHY_VACANT:
+		phy->phy_state = PHY_VACANT;
+		return;
+	default:
+		phy->phy_state = PHY_NOT_PRESENT;
+		return;
+	case SMP_RESP_FUNC_ACC:
+		phy->phy_state = PHY_EMPTY; /* do not know yet */
+		break;
+	}
+
+	phy->phy_id = phy_id;
+	phy->attached_dev_type = dr->attached_dev_type;
+	phy->linkrate = dr->linkrate;
+	phy->attached_sata_host = dr->attached_sata_host;
+	phy->attached_sata_dev  = dr->attached_sata_dev;
+	phy->attached_sata_ps   = dr->attached_sata_ps;
+	phy->attached_iproto = dr->iproto << 1;
+	phy->attached_tproto = dr->tproto << 1;
+	memcpy(phy->attached_sas_addr, dr->attached_sas_addr, SAS_ADDR_SIZE);
+	phy->attached_phy_id = dr->attached_phy_id;
+	phy->phy_change_count = dr->change_count;
+	phy->routing_attr = dr->routing_attr;
+	phy->virtual = dr->virtual;
+	phy->last_da_index = -1;
+
+	phy->phy->identify.initiator_port_protocols = phy->attached_iproto;
+	phy->phy->identify.target_port_protocols = phy->attached_tproto;
+	phy->phy->identify.phy_identifier = phy_id;
+	phy->phy->minimum_linkrate_hw = SAS_LINK_RATE_1_5_GBPS;
+	phy->phy->maximum_linkrate_hw = SAS_LINK_RATE_3_0_GBPS;
+	phy->phy->minimum_linkrate = SAS_LINK_RATE_1_5_GBPS;
+	phy->phy->maximum_linkrate = SAS_LINK_RATE_3_0_GBPS;
+	switch (phy->linkrate) {
+	case PHY_LINKRATE_1_5:
+		phy->phy->negotiated_linkrate = SAS_LINK_RATE_1_5_GBPS;
+		break;
+	case PHY_LINKRATE_3:
+		phy->phy->negotiated_linkrate = SAS_LINK_RATE_3_0_GBPS;
+		break;
+	case PHY_LINKRATE_6:
+		phy->phy->negotiated_linkrate = SAS_LINK_RATE_6_0_GBPS;
+		break;
+	default:
+		phy->phy->negotiated_linkrate = SAS_LINK_RATE_UNKNOWN;
+		break;
+	}
+
+	if (!rediscover)
+		sas_phy_add(phy->phy);
+
+	SAS_DPRINTK("ex %016llx phy%02d:%c attached: %016llx\n",
+		    SAS_ADDR(dev->sas_addr), phy->phy_id,
+		    phy->routing_attr == TABLE_ROUTING ? 'T' :
+		    phy->routing_attr == DIRECT_ROUTING ? 'D' :
+		    phy->routing_attr == SUBTRACTIVE_ROUTING ? 'S' : '?',
+		    SAS_ADDR(phy->attached_sas_addr));
+
+	return;
+}
+
+#define DISCOVER_REQ_SIZE  16
+#define DISCOVER_RESP_SIZE 56
+
+static int sas_ex_phy_discover(struct domain_device *dev, int single)
+{
+	struct expander_device *ex = &dev->ex_dev;
+	int  res = 0;
+	u8   *disc_req;
+	u8   *disc_resp;
+
+	disc_req = alloc_smp_req(DISCOVER_REQ_SIZE);
+	if (!disc_req)
+		return -ENOMEM;
+
+	disc_resp = alloc_smp_req(DISCOVER_RESP_SIZE);
+	if (!disc_resp) {
+		kfree(disc_req);
+		return -ENOMEM;
+	}
+
+	disc_req[1] = SMP_DISCOVER;
+
+	if (0 <= single && single < ex->num_phys) {
+		disc_req[9] = single;
+		res = smp_execute_task(dev, disc_req, DISCOVER_REQ_SIZE,
+				       disc_resp, DISCOVER_RESP_SIZE);
+		if (res)
+			goto out_err;
+		sas_set_ex_phy(dev, single, disc_resp);
+	} else {
+		int i;
+
+		for (i = 0; i < ex->num_phys; i++) {
+			disc_req[9] = i;
+			res = smp_execute_task(dev, disc_req,
+					       DISCOVER_REQ_SIZE, disc_resp,
+					       DISCOVER_RESP_SIZE);
+			if (res)
+				goto out_err;
+			sas_set_ex_phy(dev, i, disc_resp);
+		}
+	}
+out_err:
+	kfree(disc_resp);
+	kfree(disc_req);
+	return res;
+}
+
+static int sas_expander_discover(struct domain_device *dev)
+{
+	struct expander_device *ex = &dev->ex_dev;
+	int res = -ENOMEM;
+
+	ex->ex_phy = kzalloc(sizeof(*ex->ex_phy)*ex->num_phys, GFP_KERNEL);
+	if (!ex->ex_phy)
+		return -ENOMEM;
+
+	res = sas_ex_phy_discover(dev, -1);
+	if (res)
+		goto out_err;
+
+	return 0;
+ out_err:
+	kfree(ex->ex_phy);
+	ex->ex_phy = NULL;
+	return res;
+}
+
+#define MAX_EXPANDER_PHYS 128
+
+static void ex_assign_report_general(struct domain_device *dev,
+					    struct smp_resp *resp)
+{
+	struct report_general_resp *rg = &resp->rg;
+
+	dev->ex_dev.ex_change_count = be16_to_cpu(rg->change_count);
+	dev->ex_dev.max_route_indexes = be16_to_cpu(rg->route_indexes);
+	dev->ex_dev.num_phys = min(rg->num_phys, (u8)MAX_EXPANDER_PHYS);
+	dev->ex_dev.conf_route_table = rg->conf_route_table;
+	dev->ex_dev.configuring = rg->configuring;
+	memcpy(dev->ex_dev.enclosure_logical_id, rg->enclosure_logical_id, 8);
+}
+
+#define RG_REQ_SIZE   8
+#define RG_RESP_SIZE 32
+
+static int sas_ex_general(struct domain_device *dev)
+{
+	u8 *rg_req;
+	struct smp_resp *rg_resp;
+	int res;
+	int i;
+
+	rg_req = alloc_smp_req(RG_REQ_SIZE);
+	if (!rg_req)
+		return -ENOMEM;
+
+	rg_resp = alloc_smp_resp(RG_RESP_SIZE);
+	if (!rg_resp) {
+		kfree(rg_req);
+		return -ENOMEM;
+	}
+
+	rg_req[1] = SMP_REPORT_GENERAL;
+
+	for (i = 0; i < 5; i++) {
+		res = smp_execute_task(dev, rg_req, RG_REQ_SIZE, rg_resp,
+				       RG_RESP_SIZE);
+
+		if (res) {
+			SAS_DPRINTK("RG to ex %016llx failed:0x%x\n",
+				    SAS_ADDR(dev->sas_addr), res);
+			goto out;
+		} else if (rg_resp->result != SMP_RESP_FUNC_ACC) {
+			SAS_DPRINTK("RG:ex %016llx returned SMP result:0x%x\n",
+				    SAS_ADDR(dev->sas_addr), rg_resp->result);
+			res = rg_resp->result;
+			goto out;
+		}
+
+		ex_assign_report_general(dev, rg_resp);
+
+		if (dev->ex_dev.configuring) {
+			SAS_DPRINTK("RG: ex %llx self-configuring...\n",
+				    SAS_ADDR(dev->sas_addr));
+			schedule_timeout_interruptible(5*HZ);
+		} else
+			break;
+	}
+out:
+	kfree(rg_req);
+	kfree(rg_resp);
+	return res;
+}
+
+static void ex_assign_manuf_info(struct domain_device *dev, void
+					*_mi_resp)
+{
+	u8 *mi_resp = _mi_resp;
+	struct sas_rphy *rphy = dev->rphy;
+	struct sas_expander_device *edev = rphy_to_expander_device(rphy);
+
+	memcpy(edev->vendor_id, mi_resp + 12, SAS_EXPANDER_VENDOR_ID_LEN);
+	memcpy(edev->product_id, mi_resp + 20, SAS_EXPANDER_PRODUCT_ID_LEN);
+	memcpy(edev->product_rev, mi_resp + 36,
+	       SAS_EXPANDER_PRODUCT_REV_LEN);
+
+	if (mi_resp[8] & 1) {
+		memcpy(edev->component_vendor_id, mi_resp + 40,
+		       SAS_EXPANDER_COMPONENT_VENDOR_ID_LEN);
+		edev->component_id = mi_resp[48] << 8 | mi_resp[49];
+		edev->component_revision_id = mi_resp[50];
+	}
+}
+
+#define MI_REQ_SIZE   8
+#define MI_RESP_SIZE 64
+
+static int sas_ex_manuf_info(struct domain_device *dev)
+{
+	u8 *mi_req;
+	u8 *mi_resp;
+	int res;
+
+	mi_req = alloc_smp_req(MI_REQ_SIZE);
+	if (!mi_req)
+		return -ENOMEM;
+
+	mi_resp = alloc_smp_resp(MI_RESP_SIZE);
+	if (!mi_resp) {
+		kfree(mi_req);
+		return -ENOMEM;
+	}
+
+	mi_req[1] = SMP_REPORT_MANUF_INFO;
+
+	res = smp_execute_task(dev, mi_req, MI_REQ_SIZE, mi_resp,MI_RESP_SIZE);
+	if (res) {
+		SAS_DPRINTK("MI: ex %016llx failed:0x%x\n",
+			    SAS_ADDR(dev->sas_addr), res);
+		goto out;
+	} else if (mi_resp[2] != SMP_RESP_FUNC_ACC) {
+		SAS_DPRINTK("MI ex %016llx returned SMP result:0x%x\n",
+			    SAS_ADDR(dev->sas_addr), mi_resp[2]);
+		goto out;
+	}
+
+	ex_assign_manuf_info(dev, mi_resp);
+out:
+	kfree(mi_req);
+	kfree(mi_resp);
+	return res;
+}
+
+#define PC_REQ_SIZE  44
+#define PC_RESP_SIZE 8
+
+int sas_smp_phy_control(struct domain_device *dev, int phy_id,
+			enum phy_func phy_func)
+{
+	u8 *pc_req;
+	u8 *pc_resp;
+	int res;
+
+	pc_req = alloc_smp_req(PC_REQ_SIZE);
+	if (!pc_req)
+		return -ENOMEM;
+
+	pc_resp = alloc_smp_resp(PC_RESP_SIZE);
+	if (!pc_resp) {
+		kfree(pc_req);
+		return -ENOMEM;
+	}
+
+	pc_req[1] = SMP_PHY_CONTROL;
+	pc_req[9] = phy_id;
+	pc_req[10]= phy_func;
+
+	res = smp_execute_task(dev, pc_req, PC_REQ_SIZE, pc_resp,PC_RESP_SIZE);
+
+	kfree(pc_resp);
+	kfree(pc_req);
+	return res;
+}
+
+static void sas_ex_disable_phy(struct domain_device *dev, int phy_id)
+{
+	struct expander_device *ex = &dev->ex_dev;
+	struct ex_phy *phy = &ex->ex_phy[phy_id];
+
+	sas_smp_phy_control(dev, phy_id, PHY_FUNC_DISABLE);
+	phy->linkrate = PHY_DISABLED;
+}
+
+static void sas_ex_disable_port(struct domain_device *dev, u8 *sas_addr)
+{
+	struct expander_device *ex = &dev->ex_dev;
+	int i;
+
+	for (i = 0; i < ex->num_phys; i++) {
+		struct ex_phy *phy = &ex->ex_phy[i];
+
+		if (phy->phy_state == PHY_VACANT ||
+		    phy->phy_state == PHY_NOT_PRESENT)
+			continue;
+
+		if (SAS_ADDR(phy->attached_sas_addr) == SAS_ADDR(sas_addr))
+			sas_ex_disable_phy(dev, i);
+	}
+}
+
+static int sas_dev_present_in_domain(struct asd_sas_port *port,
+					    u8 *sas_addr)
+{
+	struct domain_device *dev;
+
+	if (SAS_ADDR(port->sas_addr) == SAS_ADDR(sas_addr))
+		return 1;
+	list_for_each_entry(dev, &port->dev_list, dev_list_node) {
+		if (SAS_ADDR(dev->sas_addr) == SAS_ADDR(sas_addr))
+			return 1;
+	}
+	return 0;
+}
+
+#define RPEL_REQ_SIZE	16
+#define RPEL_RESP_SIZE	32
+int sas_smp_get_phy_events(struct sas_phy *phy)
+{
+	int res;
+	struct sas_rphy *rphy = dev_to_rphy(phy->dev.parent);
+	struct domain_device *dev = sas_find_dev_by_rphy(rphy);
+	u8 *req = alloc_smp_req(RPEL_REQ_SIZE);
+	u8 *resp = kzalloc(RPEL_RESP_SIZE, GFP_KERNEL);
+
+	if (!resp)
+		return -ENOMEM;
+
+	req[1] = SMP_REPORT_PHY_ERR_LOG;
+	req[9] = phy->number;
+
+	res = smp_execute_task(dev, req, RPEL_REQ_SIZE,
+			            resp, RPEL_RESP_SIZE);
+
+	if (!res)
+		goto out;
+
+	phy->invalid_dword_count = scsi_to_u32(&resp[12]);
+	phy->running_disparity_error_count = scsi_to_u32(&resp[16]);
+	phy->loss_of_dword_sync_count = scsi_to_u32(&resp[20]);
+	phy->phy_reset_problem_count = scsi_to_u32(&resp[24]);
+
+ out:
+	kfree(resp);
+	return res;
+
+}
+
+#define RPS_REQ_SIZE  16
+#define RPS_RESP_SIZE 60
+
+static int sas_get_report_phy_sata(struct domain_device *dev,
+					  int phy_id,
+					  struct smp_resp *rps_resp)
+{
+	int res;
+	u8 *rps_req = alloc_smp_req(RPS_REQ_SIZE);
+
+	if (!rps_req)
+		return -ENOMEM;
+
+	rps_req[1] = SMP_REPORT_PHY_SATA;
+	rps_req[9] = phy_id;
+
+	res = smp_execute_task(dev, rps_req, RPS_REQ_SIZE,
+			            rps_resp, RPS_RESP_SIZE);
+
+	kfree(rps_req);
+	return 0;
+}
+
+static void sas_ex_get_linkrate(struct domain_device *parent,
+				       struct domain_device *child,
+				       struct ex_phy *parent_phy)
+{
+	struct expander_device *parent_ex = &parent->ex_dev;
+	struct sas_port *port;
+	int i;
+
+	child->pathways = 0;
+
+	port = parent_phy->port;
+
+	for (i = 0; i < parent_ex->num_phys; i++) {
+		struct ex_phy *phy = &parent_ex->ex_phy[i];
+
+		if (phy->phy_state == PHY_VACANT ||
+		    phy->phy_state == PHY_NOT_PRESENT)
+			continue;
+
+		if (SAS_ADDR(phy->attached_sas_addr) ==
+		    SAS_ADDR(child->sas_addr)) {
+
+			child->min_linkrate = min(parent->min_linkrate,
+						  phy->linkrate);
+			child->max_linkrate = max(parent->max_linkrate,
+						  phy->linkrate);
+			child->pathways++;
+			sas_port_add_phy(port, phy->phy);
+		}
+	}
+	child->linkrate = min(parent_phy->linkrate, child->max_linkrate);
+	child->pathways = min(child->pathways, parent->pathways);
+}
+
+static struct domain_device *sas_ex_discover_end_dev(
+	struct domain_device *parent, int phy_id)
+{
+	struct expander_device *parent_ex = &parent->ex_dev;
+	struct ex_phy *phy = &parent_ex->ex_phy[phy_id];
+	struct domain_device *child = NULL;
+	struct sas_rphy *rphy;
+	int res;
+
+	if (phy->attached_sata_host || phy->attached_sata_ps)
+		return NULL;
+
+	child = kzalloc(sizeof(*child), GFP_KERNEL);
+	if (!child)
+		return NULL;
+
+	child->parent = parent;
+	child->port   = parent->port;
+	child->iproto = phy->attached_iproto;
+	memcpy(child->sas_addr, phy->attached_sas_addr, SAS_ADDR_SIZE);
+	sas_hash_addr(child->hashed_sas_addr, child->sas_addr);
+	phy->port = sas_port_alloc(&parent->rphy->dev, phy_id);
+	BUG_ON(!phy->port);
+	/* FIXME: better error handling*/
+	BUG_ON(sas_port_add(phy->port) != 0);
+	sas_ex_get_linkrate(parent, child, phy);
+
+	if ((phy->attached_tproto & SAS_PROTO_STP) || phy->attached_sata_dev) {
+		child->dev_type = SATA_DEV;
+		if (phy->attached_tproto & SAS_PROTO_STP)
+			child->tproto = phy->attached_tproto;
+		if (phy->attached_sata_dev)
+			child->tproto |= SATA_DEV;
+		res = sas_get_report_phy_sata(parent, phy_id,
+					      &child->sata_dev.rps_resp);
+		if (res) {
+			SAS_DPRINTK("report phy sata to %016llx:0x%x returned "
+				    "0x%x\n", SAS_ADDR(parent->sas_addr),
+				    phy_id, res);
+			kfree(child);
+			return NULL;
+		}
+		memcpy(child->frame_rcvd, &child->sata_dev.rps_resp.rps.fis,
+		       sizeof(struct dev_to_host_fis));
+		sas_init_dev(child);
+		res = sas_discover_sata(child);
+		if (res) {
+			SAS_DPRINTK("sas_discover_sata() for device %16llx at "
+				    "%016llx:0x%x returned 0x%x\n",
+				    SAS_ADDR(child->sas_addr),
+				    SAS_ADDR(parent->sas_addr), phy_id, res);
+			kfree(child);
+			return NULL;
+		}
+	} else if (phy->attached_tproto & SAS_PROTO_SSP) {
+		child->dev_type = SAS_END_DEV;
+		rphy = sas_end_device_alloc(phy->port);
+		/* FIXME: error handling */
+		BUG_ON(!rphy);
+		child->tproto = phy->attached_tproto;
+		sas_init_dev(child);
+
+		child->rphy = rphy;
+		sas_fill_in_rphy(child, rphy);
+
+		spin_lock(&parent->port->dev_list_lock);
+		list_add_tail(&child->dev_list_node, &parent->port->dev_list);
+		spin_unlock(&parent->port->dev_list_lock);
+
+		res = sas_discover_end_dev(child);
+		if (res) {
+			SAS_DPRINTK("sas_discover_end_dev() for device %16llx "
+				    "at %016llx:0x%x returned 0x%x\n",
+				    SAS_ADDR(child->sas_addr),
+				    SAS_ADDR(parent->sas_addr), phy_id, res);
+			/* FIXME: this kfrees list elements without removing them */
+			//kfree(child);
+			return NULL;
+		}
+	} else {
+		SAS_DPRINTK("target proto 0x%x at %016llx:0x%x not handled\n",
+			    phy->attached_tproto, SAS_ADDR(parent->sas_addr),
+			    phy_id);
+	}
+
+	list_add_tail(&child->siblings, &parent_ex->children);
+	return child;
+}
+
+static struct domain_device *sas_ex_discover_expander(
+	struct domain_device *parent, int phy_id)
+{
+	struct sas_expander_device *parent_ex = rphy_to_expander_device(parent->rphy);
+	struct ex_phy *phy = &parent->ex_dev.ex_phy[phy_id];
+	struct domain_device *child = NULL;
+	struct sas_rphy *rphy;
+	struct sas_expander_device *edev;
+	struct asd_sas_port *port;
+	int res;
+
+	if (phy->routing_attr == DIRECT_ROUTING) {
+		SAS_DPRINTK("ex %016llx:0x%x:D <--> ex %016llx:0x%x is not "
+			    "allowed\n",
+			    SAS_ADDR(parent->sas_addr), phy_id,
+			    SAS_ADDR(phy->attached_sas_addr),
+			    phy->attached_phy_id);
+		return NULL;
+	}
+	child = kzalloc(sizeof(*child), GFP_KERNEL);
+	if (!child)
+		return NULL;
+
+	phy->port = sas_port_alloc(&parent->rphy->dev, phy_id);
+	/* FIXME: better error handling */
+	BUG_ON(sas_port_add(phy->port) != 0);
+
+
+	switch (phy->attached_dev_type) {
+	case EDGE_DEV:
+		rphy = sas_expander_alloc(phy->port,
+					  SAS_EDGE_EXPANDER_DEVICE);
+		break;
+	case FANOUT_DEV:
+		rphy = sas_expander_alloc(phy->port,
+					  SAS_FANOUT_EXPANDER_DEVICE);
+		break;
+	default:
+		rphy = NULL;	/* shut gcc up */
+		BUG();
+	}
+	port = parent->port;
+	child->rphy = rphy;
+	edev = rphy_to_expander_device(rphy);
+	child->dev_type = phy->attached_dev_type;
+	child->parent = parent;
+	child->port = port;
+	child->iproto = phy->attached_iproto;
+	child->tproto = phy->attached_tproto;
+	memcpy(child->sas_addr, phy->attached_sas_addr, SAS_ADDR_SIZE);
+	sas_hash_addr(child->hashed_sas_addr, child->sas_addr);
+	sas_ex_get_linkrate(parent, child, phy);
+	edev->level = parent_ex->level + 1;
+	parent->port->disc.max_level = max(parent->port->disc.max_level,
+					   edev->level);
+	sas_init_dev(child);
+	sas_fill_in_rphy(child, rphy);
+	sas_rphy_add(rphy);
+
+	spin_lock(&parent->port->dev_list_lock);
+	list_add_tail(&child->dev_list_node, &parent->port->dev_list);
+	spin_unlock(&parent->port->dev_list_lock);
+
+	res = sas_discover_expander(child);
+	if (res) {
+		kfree(child);
+		return NULL;
+	}
+	list_add_tail(&child->siblings, &parent->ex_dev.children);
+	return child;
+}
+
+static int sas_ex_discover_dev(struct domain_device *dev, int phy_id)
+{
+	struct expander_device *ex = &dev->ex_dev;
+	struct ex_phy *ex_phy = &ex->ex_phy[phy_id];
+	struct domain_device *child = NULL;
+	int res = 0;
+
+	/* Phy state */
+	if (ex_phy->linkrate == PHY_SPINUP_HOLD) {
+		if (!sas_smp_phy_control(dev, phy_id, PHY_FUNC_LINK_RESET))
+			res = sas_ex_phy_discover(dev, phy_id);
+		if (res)
+			return res;
+	}
+
+	/* Parent and domain coherency */
+	if (!dev->parent && (SAS_ADDR(ex_phy->attached_sas_addr) ==
+			     SAS_ADDR(dev->port->sas_addr))) {
+		sas_add_parent_port(dev, phy_id);
+		return 0;
+	}
+	if (dev->parent && (SAS_ADDR(ex_phy->attached_sas_addr) ==
+			    SAS_ADDR(dev->parent->sas_addr))) {
+		sas_add_parent_port(dev, phy_id);
+		if (ex_phy->routing_attr == TABLE_ROUTING)
+			sas_configure_phy(dev, phy_id, dev->port->sas_addr, 1);
+		return 0;
+	}
+
+	if (sas_dev_present_in_domain(dev->port, ex_phy->attached_sas_addr))
+		sas_ex_disable_port(dev, ex_phy->attached_sas_addr);
+
+	if (ex_phy->attached_dev_type == NO_DEVICE) {
+		if (ex_phy->routing_attr == DIRECT_ROUTING) {
+			memset(ex_phy->attached_sas_addr, 0, SAS_ADDR_SIZE);
+			sas_configure_routing(dev, ex_phy->attached_sas_addr);
+		}
+		return 0;
+	} else if (ex_phy->linkrate == PHY_LINKRATE_UNKNOWN)
+		return 0;
+
+	if (ex_phy->attached_dev_type != SAS_END_DEV &&
+	    ex_phy->attached_dev_type != FANOUT_DEV &&
+	    ex_phy->attached_dev_type != EDGE_DEV) {
+		SAS_DPRINTK("unknown device type(0x%x) attached to ex %016llx "
+			    "phy 0x%x\n", ex_phy->attached_dev_type,
+			    SAS_ADDR(dev->sas_addr),
+			    phy_id);
+		return 0;
+	}
+
+	res = sas_configure_routing(dev, ex_phy->attached_sas_addr);
+	if (res) {
+		SAS_DPRINTK("configure routing for dev %016llx "
+			    "reported 0x%x. Forgotten\n",
+			    SAS_ADDR(ex_phy->attached_sas_addr), res);
+		sas_disable_routing(dev, ex_phy->attached_sas_addr);
+		return res;
+	}
+
+	switch (ex_phy->attached_dev_type) {
+	case SAS_END_DEV:
+		child = sas_ex_discover_end_dev(dev, phy_id);
+		break;
+	case FANOUT_DEV:
+		if (SAS_ADDR(dev->port->disc.fanout_sas_addr)) {
+			SAS_DPRINTK("second fanout expander %016llx phy 0x%x "
+				    "attached to ex %016llx phy 0x%x\n",
+				    SAS_ADDR(ex_phy->attached_sas_addr),
+				    ex_phy->attached_phy_id,
+				    SAS_ADDR(dev->sas_addr),
+				    phy_id);
+			sas_ex_disable_phy(dev, phy_id);
+			break;
+		} else
+			memcpy(dev->port->disc.fanout_sas_addr,
+			       ex_phy->attached_sas_addr, SAS_ADDR_SIZE);
+		/* fallthrough */
+	case EDGE_DEV:
+		child = sas_ex_discover_expander(dev, phy_id);
+		break;
+	default:
+		break;
+	}
+
+	if (child) {
+		int i;
+
+		for (i = 0; i < ex->num_phys; i++) {
+			if (ex->ex_phy[i].phy_state == PHY_VACANT ||
+			    ex->ex_phy[i].phy_state == PHY_NOT_PRESENT)
+				continue;
+
+			if (SAS_ADDR(ex->ex_phy[i].attached_sas_addr) ==
+			    SAS_ADDR(child->sas_addr))
+				ex->ex_phy[i].phy_state= PHY_DEVICE_DISCOVERED;
+		}
+	}
+
+	return res;
+}
+
+static int sas_find_sub_addr(struct domain_device *dev, u8 *sub_addr)
+{
+	struct expander_device *ex = &dev->ex_dev;
+	int i;
+
+	for (i = 0; i < ex->num_phys; i++) {
+		struct ex_phy *phy = &ex->ex_phy[i];
+
+		if (phy->phy_state == PHY_VACANT ||
+		    phy->phy_state == PHY_NOT_PRESENT)
+			continue;
+
+		if ((phy->attached_dev_type == EDGE_DEV ||
+		     phy->attached_dev_type == FANOUT_DEV) &&
+		    phy->routing_attr == SUBTRACTIVE_ROUTING) {
+
+			memcpy(sub_addr, phy->attached_sas_addr,SAS_ADDR_SIZE);
+
+			return 1;
+		}
+	}
+	return 0;
+}
+
+static int sas_check_level_subtractive_boundary(struct domain_device *dev)
+{
+	struct expander_device *ex = &dev->ex_dev;
+	struct domain_device *child;
+	u8 sub_addr[8] = {0, };
+
+	list_for_each_entry(child, &ex->children, siblings) {
+		if (child->dev_type != EDGE_DEV &&
+		    child->dev_type != FANOUT_DEV)
+			continue;
+		if (sub_addr[0] == 0) {
+			sas_find_sub_addr(child, sub_addr);
+			continue;
+		} else {
+			u8 s2[8];
+
+			if (sas_find_sub_addr(child, s2) &&
+			    (SAS_ADDR(sub_addr) != SAS_ADDR(s2))) {
+
+				SAS_DPRINTK("ex %016llx->%016llx-?->%016llx "
+					    "diverges from subtractive "
+					    "boundary %016llx\n",
+					    SAS_ADDR(dev->sas_addr),
+					    SAS_ADDR(child->sas_addr),
+					    SAS_ADDR(s2),
+					    SAS_ADDR(sub_addr));
+
+				sas_ex_disable_port(child, s2);
+			}
+		}
+	}
+	return 0;
+}
+/**
+ * sas_ex_discover_devices -- discover devices attached to this expander
+ * dev: pointer to the expander domain device
+ * single: if you want to do a single phy, else set to -1;
+ *
+ * Configure this expander for use with its devices and register the
+ * devices of this expander.
+ */
+static int sas_ex_discover_devices(struct domain_device *dev, int single)
+{
+	struct expander_device *ex = &dev->ex_dev;
+	int i = 0, end = ex->num_phys;
+	int res = 0;
+
+	if (0 <= single && single < end) {
+		i = single;
+		end = i+1;
+	}
+
+	for ( ; i < end; i++) {
+		struct ex_phy *ex_phy = &ex->ex_phy[i];
+
+		if (ex_phy->phy_state == PHY_VACANT ||
+		    ex_phy->phy_state == PHY_NOT_PRESENT ||
+		    ex_phy->phy_state == PHY_DEVICE_DISCOVERED)
+			continue;
+
+		switch (ex_phy->linkrate) {
+		case PHY_DISABLED:
+		case PHY_RESET_PROBLEM:
+		case PHY_PORT_SELECTOR:
+			continue;
+		default:
+			res = sas_ex_discover_dev(dev, i);
+			if (res)
+				break;
+			continue;
+		}
+	}
+
+	if (!res)
+		sas_check_level_subtractive_boundary(dev);
+
+	return res;
+}
+
+static int sas_check_ex_subtractive_boundary(struct domain_device *dev)
+{
+	struct expander_device *ex = &dev->ex_dev;
+	int i;
+	u8  *sub_sas_addr = NULL;
+
+	if (dev->dev_type != EDGE_DEV)
+		return 0;
+
+	for (i = 0; i < ex->num_phys; i++) {
+		struct ex_phy *phy = &ex->ex_phy[i];
+
+		if (phy->phy_state == PHY_VACANT ||
+		    phy->phy_state == PHY_NOT_PRESENT)
+			continue;
+
+		if ((phy->attached_dev_type == FANOUT_DEV ||
+		     phy->attached_dev_type == EDGE_DEV) &&
+		    phy->routing_attr == SUBTRACTIVE_ROUTING) {
+
+			if (!sub_sas_addr)
+				sub_sas_addr = &phy->attached_sas_addr[0];
+			else if (SAS_ADDR(sub_sas_addr) !=
+				 SAS_ADDR(phy->attached_sas_addr)) {
+
+				SAS_DPRINTK("ex %016llx phy 0x%x "
+					    "diverges(%016llx) on subtractive "
+					    "boundary(%016llx). Disabled\n",
+					    SAS_ADDR(dev->sas_addr), i,
+					    SAS_ADDR(phy->attached_sas_addr),
+					    SAS_ADDR(sub_sas_addr));
+				sas_ex_disable_phy(dev, i);
+			}
+		}
+	}
+	return 0;
+}
+
+static void sas_print_parent_topology_bug(struct domain_device *child,
+						 struct ex_phy *parent_phy,
+						 struct ex_phy *child_phy)
+{
+	static const char ra_char[] = {
+		[DIRECT_ROUTING] = 'D',
+		[SUBTRACTIVE_ROUTING] = 'S',
+		[TABLE_ROUTING] = 'T',
+	};
+	static const char *ex_type[] = {
+		[EDGE_DEV] = "edge",
+		[FANOUT_DEV] = "fanout",
+	};
+	struct domain_device *parent = child->parent;
+
+	sas_printk("%s ex %016llx phy 0x%x <--> %s ex %016llx phy 0x%x "
+		   "has %c:%c routing link!\n",
+
+		   ex_type[parent->dev_type],
+		   SAS_ADDR(parent->sas_addr),
+		   parent_phy->phy_id,
+
+		   ex_type[child->dev_type],
+		   SAS_ADDR(child->sas_addr),
+		   child_phy->phy_id,
+
+		   ra_char[parent_phy->routing_attr],
+		   ra_char[child_phy->routing_attr]);
+}
+
+static int sas_check_eeds(struct domain_device *child,
+				 struct ex_phy *parent_phy,
+				 struct ex_phy *child_phy)
+{
+	int res = 0;
+	struct domain_device *parent = child->parent;
+
+	if (SAS_ADDR(parent->port->disc.fanout_sas_addr) != 0) {
+		res = -ENODEV;
+		SAS_DPRINTK("edge ex %016llx phy S:0x%x <--> edge ex %016llx "
+			    "phy S:0x%x, while there is a fanout ex %016llx\n",
+			    SAS_ADDR(parent->sas_addr),
+			    parent_phy->phy_id,
+			    SAS_ADDR(child->sas_addr),
+			    child_phy->phy_id,
+			    SAS_ADDR(parent->port->disc.fanout_sas_addr));
+	} else if (SAS_ADDR(parent->port->disc.eeds_a) == 0) {
+		memcpy(parent->port->disc.eeds_a, parent->sas_addr,
+		       SAS_ADDR_SIZE);
+		memcpy(parent->port->disc.eeds_b, child->sas_addr,
+		       SAS_ADDR_SIZE);
+	} else if (((SAS_ADDR(parent->port->disc.eeds_a) ==
+		    SAS_ADDR(parent->sas_addr)) ||
+		   (SAS_ADDR(parent->port->disc.eeds_a) ==
+		    SAS_ADDR(child->sas_addr)))
+		   &&
+		   ((SAS_ADDR(parent->port->disc.eeds_b) ==
+		     SAS_ADDR(parent->sas_addr)) ||
+		    (SAS_ADDR(parent->port->disc.eeds_b) ==
+		     SAS_ADDR(child->sas_addr))))
+		;
+	else {
+		res = -ENODEV;
+		SAS_DPRINTK("edge ex %016llx phy 0x%x <--> edge ex %016llx "
+			    "phy 0x%x link forms a third EEDS!\n",
+			    SAS_ADDR(parent->sas_addr),
+			    parent_phy->phy_id,
+			    SAS_ADDR(child->sas_addr),
+			    child_phy->phy_id);
+	}
+
+	return res;
+}
+
+/* Here we spill over 80 columns.  It is intentional.
+ */
+static int sas_check_parent_topology(struct domain_device *child)
+{
+	struct expander_device *child_ex = &child->ex_dev;
+	struct expander_device *parent_ex;
+	int i;
+	int res = 0;
+
+	if (!child->parent)
+		return 0;
+
+	if (child->parent->dev_type != EDGE_DEV &&
+	    child->parent->dev_type != FANOUT_DEV)
+		return 0;
+
+	parent_ex = &child->parent->ex_dev;
+
+	for (i = 0; i < parent_ex->num_phys; i++) {
+		struct ex_phy *parent_phy = &parent_ex->ex_phy[i];
+		struct ex_phy *child_phy;
+
+		if (parent_phy->phy_state == PHY_VACANT ||
+		    parent_phy->phy_state == PHY_NOT_PRESENT)
+			continue;
+
+		if (SAS_ADDR(parent_phy->attached_sas_addr) != SAS_ADDR(child->sas_addr))
+			continue;
+
+		child_phy = &child_ex->ex_phy[parent_phy->attached_phy_id];
+
+		switch (child->parent->dev_type) {
+		case EDGE_DEV:
+			if (child->dev_type == FANOUT_DEV) {
+				if (parent_phy->routing_attr != SUBTRACTIVE_ROUTING ||
+				    child_phy->routing_attr != TABLE_ROUTING) {
+					sas_print_parent_topology_bug(child, parent_phy, child_phy);
+					res = -ENODEV;
+				}
+			} else if (parent_phy->routing_attr == SUBTRACTIVE_ROUTING) {
+				if (child_phy->routing_attr == SUBTRACTIVE_ROUTING) {
+					res = sas_check_eeds(child, parent_phy, child_phy);
+				} else if (child_phy->routing_attr != TABLE_ROUTING) {
+					sas_print_parent_topology_bug(child, parent_phy, child_phy);
+					res = -ENODEV;
+				}
+			} else if (parent_phy->routing_attr == TABLE_ROUTING &&
+				   child_phy->routing_attr != SUBTRACTIVE_ROUTING) {
+				sas_print_parent_topology_bug(child, parent_phy, child_phy);
+				res = -ENODEV;
+			}
+			break;
+		case FANOUT_DEV:
+			if (parent_phy->routing_attr != TABLE_ROUTING ||
+			    child_phy->routing_attr != SUBTRACTIVE_ROUTING) {
+				sas_print_parent_topology_bug(child, parent_phy, child_phy);
+				res = -ENODEV;
+			}
+			break;
+		default:
+			break;
+		}
+	}
+
+	return res;
+}
+
+#define RRI_REQ_SIZE  16
+#define RRI_RESP_SIZE 44
+
+static int sas_configure_present(struct domain_device *dev, int phy_id,
+				 u8 *sas_addr, int *index, int *present)
+{
+	int i, res = 0;
+	struct expander_device *ex = &dev->ex_dev;
+	struct ex_phy *phy = &ex->ex_phy[phy_id];
+	u8 *rri_req;
+	u8 *rri_resp;
+
+	*present = 0;
+	*index = 0;
+
+	rri_req = alloc_smp_req(RRI_REQ_SIZE);
+	if (!rri_req)
+		return -ENOMEM;
+
+	rri_resp = alloc_smp_resp(RRI_RESP_SIZE);
+	if (!rri_resp) {
+		kfree(rri_req);
+		return -ENOMEM;
+	}
+
+	rri_req[1] = SMP_REPORT_ROUTE_INFO;
+	rri_req[9] = phy_id;
+
+	for (i = 0; i < ex->max_route_indexes ; i++) {
+		*(__be16 *)(rri_req+6) = cpu_to_be16(i);
+		res = smp_execute_task(dev, rri_req, RRI_REQ_SIZE, rri_resp,
+				       RRI_RESP_SIZE);
+		if (res)
+			goto out;
+		res = rri_resp[2];
+		if (res == SMP_RESP_NO_INDEX) {
+			SAS_DPRINTK("overflow of indexes: dev %016llx "
+				    "phy 0x%x index 0x%x\n",
+				    SAS_ADDR(dev->sas_addr), phy_id, i);
+			goto out;
+		} else if (res != SMP_RESP_FUNC_ACC) {
+			SAS_DPRINTK("%s: dev %016llx phy 0x%x index 0x%x "
+				    "result 0x%x\n", __FUNCTION__,
+				    SAS_ADDR(dev->sas_addr), phy_id, i, res);
+			goto out;
+		}
+		if (SAS_ADDR(sas_addr) != 0) {
+			if (SAS_ADDR(rri_resp+16) == SAS_ADDR(sas_addr)) {
+				*index = i;
+				if ((rri_resp[12] & 0x80) == 0x80)
+					*present = 0;
+				else
+					*present = 1;
+				goto out;
+			} else if (SAS_ADDR(rri_resp+16) == 0) {
+				*index = i;
+				*present = 0;
+				goto out;
+			}
+		} else if (SAS_ADDR(rri_resp+16) == 0 &&
+			   phy->last_da_index < i) {
+			phy->last_da_index = i;
+			*index = i;
+			*present = 0;
+			goto out;
+		}
+	}
+	res = -1;
+out:
+	kfree(rri_req);
+	kfree(rri_resp);
+	return res;
+}
+
+#define CRI_REQ_SIZE  44
+#define CRI_RESP_SIZE  8
+
+static int sas_configure_set(struct domain_device *dev, int phy_id,
+			     u8 *sas_addr, int index, int include)
+{
+	int res;
+	u8 *cri_req;
+	u8 *cri_resp;
+
+	cri_req = alloc_smp_req(CRI_REQ_SIZE);
+	if (!cri_req)
+		return -ENOMEM;
+
+	cri_resp = alloc_smp_resp(CRI_RESP_SIZE);
+	if (!cri_resp) {
+		kfree(cri_req);
+		return -ENOMEM;
+	}
+
+	cri_req[1] = SMP_CONF_ROUTE_INFO;
+	*(__be16 *)(cri_req+6) = cpu_to_be16(index);
+	cri_req[9] = phy_id;
+	if (SAS_ADDR(sas_addr) == 0 || !include)
+		cri_req[12] |= 0x80;
+	memcpy(cri_req+16, sas_addr, SAS_ADDR_SIZE);
+
+	res = smp_execute_task(dev, cri_req, CRI_REQ_SIZE, cri_resp,
+			       CRI_RESP_SIZE);
+	if (res)
+		goto out;
+	res = cri_resp[2];
+	if (res == SMP_RESP_NO_INDEX) {
+		SAS_DPRINTK("overflow of indexes: dev %016llx phy 0x%x "
+			    "index 0x%x\n",
+			    SAS_ADDR(dev->sas_addr), phy_id, index);
+	}
+out:
+	kfree(cri_req);
+	kfree(cri_resp);
+	return res;
+}
+
+static int sas_configure_phy(struct domain_device *dev, int phy_id,
+				    u8 *sas_addr, int include)
+{
+	int index;
+	int present;
+	int res;
+
+	res = sas_configure_present(dev, phy_id, sas_addr, &index, &present);
+	if (res)
+		return res;
+	if (include ^ present)
+		return sas_configure_set(dev, phy_id, sas_addr, index,include);
+
+	return res;
+}
+
+/**
+ * sas_configure_parent -- configure routing table of parent
+ * parent: parent expander
+ * child: child expander
+ * sas_addr: SAS port identifier of device directly attached to child
+ */
+static int sas_configure_parent(struct domain_device *parent,
+				struct domain_device *child,
+				u8 *sas_addr, int include)
+{
+	struct expander_device *ex_parent = &parent->ex_dev;
+	int res = 0;
+	int i;
+
+	if (parent->parent) {
+		res = sas_configure_parent(parent->parent, parent, sas_addr,
+					   include);
+		if (res)
+			return res;
+	}
+
+	if (ex_parent->conf_route_table == 0) {
+		SAS_DPRINTK("ex %016llx has self-configuring routing table\n",
+			    SAS_ADDR(parent->sas_addr));
+		return 0;
+	}
+
+	for (i = 0; i < ex_parent->num_phys; i++) {
+		struct ex_phy *phy = &ex_parent->ex_phy[i];
+
+		if ((phy->routing_attr == TABLE_ROUTING) &&
+		    (SAS_ADDR(phy->attached_sas_addr) ==
+		     SAS_ADDR(child->sas_addr))) {
+			res = sas_configure_phy(parent, i, sas_addr, include);
+			if (res)
+				return res;
+		}
+	}
+
+	return res;
+}
+
+/**
+ * sas_configure_routing -- configure routing
+ * dev: expander device
+ * sas_addr: port identifier of device directly attached to the expander device
+ */
+static int sas_configure_routing(struct domain_device *dev, u8 *sas_addr)
+{
+	if (dev->parent)
+		return sas_configure_parent(dev->parent, dev, sas_addr, 1);
+	return 0;
+}
+
+static int sas_disable_routing(struct domain_device *dev,  u8 *sas_addr)
+{
+	if (dev->parent)
+		return sas_configure_parent(dev->parent, dev, sas_addr, 0);
+	return 0;
+}
+
+#if 0
+#define SMP_BIN_ATTR_NAME "smp_portal"
+
+static void sas_ex_smp_hook(struct domain_device *dev)
+{
+	struct expander_device *ex_dev = &dev->ex_dev;
+	struct bin_attribute *bin_attr = &ex_dev->smp_bin_attr;
+
+	memset(bin_attr, 0, sizeof(*bin_attr));
+
+	bin_attr->attr.name = SMP_BIN_ATTR_NAME;
+	bin_attr->attr.owner = THIS_MODULE;
+	bin_attr->attr.mode = 0600;
+
+	bin_attr->size = 0;
+	bin_attr->private = NULL;
+	bin_attr->read = smp_portal_read;
+	bin_attr->write= smp_portal_write;
+	bin_attr->mmap = NULL;
+
+	ex_dev->smp_portal_pid = -1;
+	init_MUTEX(&ex_dev->smp_sema);
+}
+#endif
+
+/**
+ * sas_discover_expander -- expander discovery
+ * @ex: pointer to expander domain device
+ *
+ * See comment in sas_discover_sata().
+ */
+static int sas_discover_expander(struct domain_device *dev)
+{
+	int res;
+
+	res = sas_notify_lldd_dev_found(dev);
+	if (res)
+		return res;
+
+	res = sas_ex_general(dev);
+	if (res)
+		goto out_err;
+	res = sas_ex_manuf_info(dev);
+	if (res)
+		goto out_err;
+
+	res = sas_expander_discover(dev);
+	if (res) {
+		SAS_DPRINTK("expander %016llx discovery failed(0x%x)\n",
+			    SAS_ADDR(dev->sas_addr), res);
+		goto out_err;
+	}
+
+	sas_check_ex_subtractive_boundary(dev);
+	res = sas_check_parent_topology(dev);
+	if (res)
+		goto out_err;
+	return 0;
+out_err:
+	sas_notify_lldd_dev_gone(dev);
+	return res;
+}
+
+static int sas_ex_level_discovery(struct asd_sas_port *port, const int level)
+{
+	int res = 0;
+	struct domain_device *dev;
+
+	list_for_each_entry(dev, &port->dev_list, dev_list_node) {
+		if (dev->dev_type == EDGE_DEV ||
+		    dev->dev_type == FANOUT_DEV) {
+			struct sas_expander_device *ex =
+				rphy_to_expander_device(dev->rphy);
+
+			if (level == ex->level)
+				res = sas_ex_discover_devices(dev, -1);
+			else if (level > 0)
+				res = sas_ex_discover_devices(port->port_dev, -1);
+
+		}
+	}
+
+	return res;
+}
+
+static int sas_ex_bfs_disc(struct asd_sas_port *port)
+{
+	int res;
+	int level;
+
+	do {
+		level = port->disc.max_level;
+		res = sas_ex_level_discovery(port, level);
+		mb();
+	} while (level < port->disc.max_level);
+
+	return res;
+}
+
+int sas_discover_root_expander(struct domain_device *dev)
+{
+	int res;
+	struct sas_expander_device *ex = rphy_to_expander_device(dev->rphy);
+
+	sas_rphy_add(dev->rphy);
+
+	ex->level = dev->port->disc.max_level; /* 0 */
+	res = sas_discover_expander(dev);
+	if (!res)
+		sas_ex_bfs_disc(dev->port);
+
+	return res;
+}
+
+/* ---------- Domain revalidation ---------- */
+
+static int sas_get_phy_discover(struct domain_device *dev,
+				int phy_id, struct smp_resp *disc_resp)
+{
+	int res;
+	u8 *disc_req;
+
+	disc_req = alloc_smp_req(DISCOVER_REQ_SIZE);
+	if (!disc_req)
+		return -ENOMEM;
+
+	disc_req[1] = SMP_DISCOVER;
+	disc_req[9] = phy_id;
+
+	res = smp_execute_task(dev, disc_req, DISCOVER_REQ_SIZE,
+			       disc_resp, DISCOVER_RESP_SIZE);
+	if (res)
+		goto out;
+	else if (disc_resp->result != SMP_RESP_FUNC_ACC) {
+		res = disc_resp->result;
+		goto out;
+	}
+out:
+	kfree(disc_req);
+	return res;
+}
+
+static int sas_get_phy_change_count(struct domain_device *dev,
+				    int phy_id, int *pcc)
+{
+	int res;
+	struct smp_resp *disc_resp;
+
+	disc_resp = alloc_smp_resp(DISCOVER_RESP_SIZE);
+	if (!disc_resp)
+		return -ENOMEM;
+
+	res = sas_get_phy_discover(dev, phy_id, disc_resp);
+	if (!res)
+		*pcc = disc_resp->disc.change_count;
+
+	kfree(disc_resp);
+	return res;
+}
+
+static int sas_get_phy_attached_sas_addr(struct domain_device *dev,
+					 int phy_id, u8 *attached_sas_addr)
+{
+	int res;
+	struct smp_resp *disc_resp;
+	struct discover_resp *dr;
+
+	disc_resp = alloc_smp_resp(DISCOVER_RESP_SIZE);
+	if (!disc_resp)
+		return -ENOMEM;
+	dr = &disc_resp->disc;
+
+	res = sas_get_phy_discover(dev, phy_id, disc_resp);
+	if (!res) {
+		memcpy(attached_sas_addr,disc_resp->disc.attached_sas_addr,8);
+		if (dr->attached_dev_type == 0)
+			memset(attached_sas_addr, 0, 8);
+	}
+	kfree(disc_resp);
+	return res;
+}
+
+static int sas_find_bcast_phy(struct domain_device *dev, int *phy_id,
+			      int from_phy)
+{
+	struct expander_device *ex = &dev->ex_dev;
+	int res = 0;
+	int i;
+
+	for (i = from_phy; i < ex->num_phys; i++) {
+		int phy_change_count = 0;
+
+		res = sas_get_phy_change_count(dev, i, &phy_change_count);
+		if (res)
+			goto out;
+		else if (phy_change_count != ex->ex_phy[i].phy_change_count) {
+			ex->ex_phy[i].phy_change_count = phy_change_count;
+			*phy_id = i;
+			return 0;
+		}
+	}
+out:
+	return res;
+}
+
+static int sas_get_ex_change_count(struct domain_device *dev, int *ecc)
+{
+	int res;
+	u8  *rg_req;
+	struct smp_resp  *rg_resp;
+
+	rg_req = alloc_smp_req(RG_REQ_SIZE);
+	if (!rg_req)
+		return -ENOMEM;
+
+	rg_resp = alloc_smp_resp(RG_RESP_SIZE);
+	if (!rg_resp) {
+		kfree(rg_req);
+		return -ENOMEM;
+	}
+
+	rg_req[1] = SMP_REPORT_GENERAL;
+
+	res = smp_execute_task(dev, rg_req, RG_REQ_SIZE, rg_resp,
+			       RG_RESP_SIZE);
+	if (res)
+		goto out;
+	if (rg_resp->result != SMP_RESP_FUNC_ACC) {
+		res = rg_resp->result;
+		goto out;
+	}
+
+	*ecc = be16_to_cpu(rg_resp->rg.change_count);
+out:
+	kfree(rg_resp);
+	kfree(rg_req);
+	return res;
+}
+
+static int sas_find_bcast_dev(struct domain_device *dev,
+			      struct domain_device **src_dev)
+{
+	struct expander_device *ex = &dev->ex_dev;
+	int ex_change_count = -1;
+	int res;
+
+	res = sas_get_ex_change_count(dev, &ex_change_count);
+	if (res)
+		goto out;
+	if (ex_change_count != -1 &&
+	    ex_change_count != ex->ex_change_count) {
+		*src_dev = dev;
+		ex->ex_change_count = ex_change_count;
+	} else {
+		struct domain_device *ch;
+
+		list_for_each_entry(ch, &ex->children, siblings) {
+			if (ch->dev_type == EDGE_DEV ||
+			    ch->dev_type == FANOUT_DEV) {
+				res = sas_find_bcast_dev(ch, src_dev);
+				if (src_dev)
+					return res;
+			}
+		}
+	}
+out:
+	return res;
+}
+
+static void sas_unregister_ex_tree(struct domain_device *dev)
+{
+	struct expander_device *ex = &dev->ex_dev;
+	struct domain_device *child, *n;
+
+	list_for_each_entry_safe(child, n, &ex->children, siblings) {
+		if (child->dev_type == EDGE_DEV ||
+		    child->dev_type == FANOUT_DEV)
+			sas_unregister_ex_tree(child);
+		else
+			sas_unregister_dev(child);
+	}
+	sas_unregister_dev(dev);
+}
+
+static void sas_unregister_devs_sas_addr(struct domain_device *parent,
+					 int phy_id)
+{
+	struct expander_device *ex_dev = &parent->ex_dev;
+	struct ex_phy *phy = &ex_dev->ex_phy[phy_id];
+	struct domain_device *child, *n;
+
+	list_for_each_entry_safe(child, n, &ex_dev->children, siblings) {
+		if (SAS_ADDR(child->sas_addr) ==
+		    SAS_ADDR(phy->attached_sas_addr)) {
+			if (child->dev_type == EDGE_DEV ||
+			    child->dev_type == FANOUT_DEV)
+				sas_unregister_ex_tree(child);
+			else
+				sas_unregister_dev(child);
+			break;
+		}
+	}
+	sas_disable_routing(parent, phy->attached_sas_addr);
+	memset(phy->attached_sas_addr, 0, SAS_ADDR_SIZE);
+	sas_port_delete_phy(phy->port, phy->phy);
+	if (phy->port->num_phys == 0)
+		sas_port_delete(phy->port);
+	phy->port = NULL;
+}
+
+static int sas_discover_bfs_by_root_level(struct domain_device *root,
+					  const int level)
+{
+	struct expander_device *ex_root = &root->ex_dev;
+	struct domain_device *child;
+	int res = 0;
+
+	list_for_each_entry(child, &ex_root->children, siblings) {
+		if (child->dev_type == EDGE_DEV ||
+		    child->dev_type == FANOUT_DEV) {
+			struct sas_expander_device *ex =
+				rphy_to_expander_device(child->rphy);
+
+			if (level > ex->level)
+				res = sas_discover_bfs_by_root_level(child,
+								     level);
+			else if (level == ex->level)
+				res = sas_ex_discover_devices(child, -1);
+		}
+	}
+	return res;
+}
+
+static int sas_discover_bfs_by_root(struct domain_device *dev)
+{
+	int res;
+	struct sas_expander_device *ex = rphy_to_expander_device(dev->rphy);
+	int level = ex->level+1;
+
+	res = sas_ex_discover_devices(dev, -1);
+	if (res)
+		goto out;
+	do {
+		res = sas_discover_bfs_by_root_level(dev, level);
+		mb();
+		level += 1;
+	} while (level <= dev->port->disc.max_level);
+out:
+	return res;
+}
+
+static int sas_discover_new(struct domain_device *dev, int phy_id)
+{
+	struct ex_phy *ex_phy = &dev->ex_dev.ex_phy[phy_id];
+	struct domain_device *child;
+	int res;
+
+	SAS_DPRINTK("ex %016llx phy%d new device attached\n",
+		    SAS_ADDR(dev->sas_addr), phy_id);
+	res = sas_ex_phy_discover(dev, phy_id);
+	if (res)
+		goto out;
+	res = sas_ex_discover_devices(dev, phy_id);
+	if (res)
+		goto out;
+	list_for_each_entry(child, &dev->ex_dev.children, siblings) {
+		if (SAS_ADDR(child->sas_addr) ==
+		    SAS_ADDR(ex_phy->attached_sas_addr)) {
+			if (child->dev_type == EDGE_DEV ||
+			    child->dev_type == FANOUT_DEV)
+				res = sas_discover_bfs_by_root(child);
+			break;
+		}
+	}
+out:
+	return res;
+}
+
+static int sas_rediscover_dev(struct domain_device *dev, int phy_id)
+{
+	struct expander_device *ex = &dev->ex_dev;
+	struct ex_phy *phy = &ex->ex_phy[phy_id];
+	u8 attached_sas_addr[8];
+	int res;
+
+	res = sas_get_phy_attached_sas_addr(dev, phy_id, attached_sas_addr);
+	switch (res) {
+	case SMP_RESP_NO_PHY:
+		phy->phy_state = PHY_NOT_PRESENT;
+		sas_unregister_devs_sas_addr(dev, phy_id);
+		goto out; break;
+	case SMP_RESP_PHY_VACANT:
+		phy->phy_state = PHY_VACANT;
+		sas_unregister_devs_sas_addr(dev, phy_id);
+		goto out; break;
+	case SMP_RESP_FUNC_ACC:
+		break;
+	}
+
+	if (SAS_ADDR(attached_sas_addr) == 0) {
+		phy->phy_state = PHY_EMPTY;
+		sas_unregister_devs_sas_addr(dev, phy_id);
+	} else if (SAS_ADDR(attached_sas_addr) ==
+		   SAS_ADDR(phy->attached_sas_addr)) {
+		SAS_DPRINTK("ex %016llx phy 0x%x broadcast flutter\n",
+			    SAS_ADDR(dev->sas_addr), phy_id);
+	} else
+		res = sas_discover_new(dev, phy_id);
+out:
+	return res;
+}
+
+static int sas_rediscover(struct domain_device *dev, const int phy_id)
+{
+	struct expander_device *ex = &dev->ex_dev;
+	struct ex_phy *changed_phy = &ex->ex_phy[phy_id];
+	int res = 0;
+	int i;
+
+	SAS_DPRINTK("ex %016llx phy%d originated BROADCAST(CHANGE)\n",
+		    SAS_ADDR(dev->sas_addr), phy_id);
+
+	if (SAS_ADDR(changed_phy->attached_sas_addr) != 0) {
+		for (i = 0; i < ex->num_phys; i++) {
+			struct ex_phy *phy = &ex->ex_phy[i];
+
+			if (i == phy_id)
+				continue;
+			if (SAS_ADDR(phy->attached_sas_addr) ==
+			    SAS_ADDR(changed_phy->attached_sas_addr)) {
+				SAS_DPRINTK("phy%d part of wide port with "
+					    "phy%d\n", phy_id, i);
+				goto out;
+			}
+		}
+		res = sas_rediscover_dev(dev, phy_id);
+	} else
+		res = sas_discover_new(dev, phy_id);
+out:
+	return res;
+}
+
+/**
+ * sas_revalidate_domain -- revalidate the domain
+ * @port: port to the domain of interest
+ *
+ * NOTE: this process _must_ quit (return) as soon as any connection
+ * errors are encountered.  Connection recovery is done elsewhere.
+ * Discover process only interrogates devices in order to discover the
+ * domain.
+ */
+int sas_ex_revalidate_domain(struct domain_device *port_dev)
+{
+	int res;
+	struct domain_device *dev = NULL;
+
+	res = sas_find_bcast_dev(port_dev, &dev);
+	if (res)
+		goto out;
+	if (dev) {
+		struct expander_device *ex = &dev->ex_dev;
+		int i = 0, phy_id;
+
+		do {
+			phy_id = -1;
+			res = sas_find_bcast_phy(dev, &phy_id, i);
+			if (phy_id == -1)
+				break;
+			res = sas_rediscover(dev, phy_id);
+			i = phy_id + 1;
+		} while (i < ex->num_phys);
+	}
+out:
+	return res;
+}
+
+#if 0
+/* ---------- SMP portal ---------- */
+
+static ssize_t smp_portal_write(struct kobject *kobj, char *buf, loff_t offs,
+				size_t size)
+{
+	struct domain_device *dev = to_dom_device(kobj);
+	struct expander_device *ex = &dev->ex_dev;
+
+	if (offs != 0)
+		return -EFBIG;
+	else if (size == 0)
+		return 0;
+
+	down_interruptible(&ex->smp_sema);
+	if (ex->smp_req)
+		kfree(ex->smp_req);
+	ex->smp_req = kzalloc(size, GFP_USER);
+	if (!ex->smp_req) {
+		up(&ex->smp_sema);
+		return -ENOMEM;
+	}
+	memcpy(ex->smp_req, buf, size);
+	ex->smp_req_size = size;
+	ex->smp_portal_pid = current->pid;
+	up(&ex->smp_sema);
+
+	return size;
+}
+
+static ssize_t smp_portal_read(struct kobject *kobj, char *buf, loff_t offs,
+			       size_t size)
+{
+	struct domain_device *dev = to_dom_device(kobj);
+	struct expander_device *ex = &dev->ex_dev;
+	u8 *smp_resp;
+	int res = -EINVAL;
+
+	/* XXX: sysfs gives us an offset of 0x10 or 0x8 while in fact
+	 *  it should be 0.
+	 */
+
+	down_interruptible(&ex->smp_sema);
+	if (!ex->smp_req || ex->smp_portal_pid != current->pid)
+		goto out;
+
+	res = 0;
+	if (size == 0)
+		goto out;
+
+	res = -ENOMEM;
+	smp_resp = alloc_smp_resp(size);
+	if (!smp_resp)
+		goto out;
+	res = smp_execute_task(dev, ex->smp_req, ex->smp_req_size,
+			       smp_resp, size);
+	if (!res) {
+		memcpy(buf, smp_resp, size);
+		res = size;
+	}
+
+	kfree(smp_resp);
+out:
+	kfree(ex->smp_req);
+	ex->smp_req = NULL;
+	ex->smp_req_size = 0;
+	ex->smp_portal_pid = -1;
+	up(&ex->smp_sema);
+	return res;
+}
+#endif
diff --git a/drivers/scsi/libsas/sas_init.c b/drivers/scsi/libsas/sas_init.c
new file mode 100644
index 000000000000..b961664b8106
--- /dev/null
+++ b/drivers/scsi/libsas/sas_init.c
@@ -0,0 +1,227 @@
+/*
+ * Serial Attached SCSI (SAS) Transport Layer initialization
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
+ * USA
+ *
+ */
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/device.h>
+#include <linux/spinlock.h>
+#include <scsi/scsi_host.h>
+#include <scsi/scsi_device.h>
+#include <scsi/scsi_transport.h>
+#include <scsi/scsi_transport_sas.h>
+
+#include "sas_internal.h"
+
+#include "../scsi_sas_internal.h"
+
+kmem_cache_t *sas_task_cache;
+
+/*------------ SAS addr hash -----------*/
+void sas_hash_addr(u8 *hashed, const u8 *sas_addr)
+{
+        const u32 poly = 0x00DB2777;
+        u32     r = 0;
+        int     i;
+
+        for (i = 0; i < 8; i++) {
+                int b;
+                for (b = 7; b >= 0; b--) {
+                        r <<= 1;
+                        if ((1 << b) & sas_addr[i]) {
+                                if (!(r & 0x01000000))
+                                        r ^= poly;
+                        } else if (r & 0x01000000)
+                                r ^= poly;
+                }
+        }
+
+        hashed[0] = (r >> 16) & 0xFF;
+        hashed[1] = (r >> 8) & 0xFF ;
+        hashed[2] = r & 0xFF;
+}
+
+
+/* ---------- HA events ---------- */
+
+void sas_hae_reset(void *data)
+{
+	struct sas_ha_struct *ha = data;
+
+	sas_begin_event(HAE_RESET, &ha->event_lock,
+			&ha->pending);
+}
+
+int sas_register_ha(struct sas_ha_struct *sas_ha)
+{
+	int error = 0;
+
+	spin_lock_init(&sas_ha->phy_port_lock);
+	sas_hash_addr(sas_ha->hashed_sas_addr, sas_ha->sas_addr);
+
+	if (sas_ha->lldd_queue_size == 0)
+		sas_ha->lldd_queue_size = 1;
+	else if (sas_ha->lldd_queue_size == -1)
+		sas_ha->lldd_queue_size = 128; /* Sanity */
+
+	error = sas_register_phys(sas_ha);
+	if (error) {
+		printk(KERN_NOTICE "couldn't register sas phys:%d\n", error);
+		return error;
+	}
+
+	error = sas_register_ports(sas_ha);
+	if (error) {
+		printk(KERN_NOTICE "couldn't register sas ports:%d\n", error);
+		goto Undo_phys;
+	}
+
+	error = sas_init_events(sas_ha);
+	if (error) {
+		printk(KERN_NOTICE "couldn't start event thread:%d\n", error);
+		goto Undo_ports;
+	}
+
+	if (sas_ha->lldd_max_execute_num > 1) {
+		error = sas_init_queue(sas_ha);
+		if (error) {
+			printk(KERN_NOTICE "couldn't start queue thread:%d, "
+			       "running in direct mode\n", error);
+			sas_ha->lldd_max_execute_num = 1;
+		}
+	}
+
+	return 0;
+
+Undo_ports:
+	sas_unregister_ports(sas_ha);
+Undo_phys:
+
+	return error;
+}
+
+int sas_unregister_ha(struct sas_ha_struct *sas_ha)
+{
+	if (sas_ha->lldd_max_execute_num > 1) {
+		sas_shutdown_queue(sas_ha);
+	}
+
+	sas_unregister_ports(sas_ha);
+
+	return 0;
+}
+
+static int sas_get_linkerrors(struct sas_phy *phy)
+{
+	if (scsi_is_sas_phy_local(phy))
+		/* FIXME: we have no local phy stats
+		 * gathering at this time */
+		return -EINVAL;
+
+	return sas_smp_get_phy_events(phy);
+}
+
+static int sas_phy_reset(struct sas_phy *phy, int hard_reset)
+{
+	int ret;
+	enum phy_func reset_type;
+
+	if (hard_reset)
+		reset_type = PHY_FUNC_HARD_RESET;
+	else
+		reset_type = PHY_FUNC_LINK_RESET;
+
+	if (scsi_is_sas_phy_local(phy)) {
+		struct Scsi_Host *shost = dev_to_shost(phy->dev.parent);
+		struct sas_ha_struct *sas_ha = SHOST_TO_SAS_HA(shost);
+		struct asd_sas_phy *asd_phy = sas_ha->sas_phy[phy->number];
+		struct sas_internal *i =
+			to_sas_internal(sas_ha->core.shost->transportt);
+
+		ret = i->dft->lldd_control_phy(asd_phy, reset_type);
+	} else {
+		struct sas_rphy *rphy = dev_to_rphy(phy->dev.parent);
+		struct domain_device *ddev = sas_find_dev_by_rphy(rphy);
+		ret = sas_smp_phy_control(ddev, phy->number, reset_type);
+	}
+	return ret;
+}
+
+static struct sas_function_template sft = {
+	.phy_reset = sas_phy_reset,
+	.get_linkerrors = sas_get_linkerrors,
+};
+
+struct scsi_transport_template *
+sas_domain_attach_transport(struct sas_domain_function_template *dft)
+{
+	struct scsi_transport_template *stt = sas_attach_transport(&sft);
+	struct sas_internal *i;
+
+	if (!stt)
+		return stt;
+
+	i = to_sas_internal(stt);
+	i->dft = dft;
+	stt->create_work_queue = 1;
+	stt->eh_timed_out = sas_scsi_timed_out;
+	stt->eh_strategy_handler = sas_scsi_recover_host;
+
+	return stt;
+}
+EXPORT_SYMBOL_GPL(sas_domain_attach_transport);
+
+
+void sas_domain_release_transport(struct scsi_transport_template *stt)
+{
+	sas_release_transport(stt);
+}
+EXPORT_SYMBOL_GPL(sas_domain_release_transport);
+
+/* ---------- SAS Class register/unregister ---------- */
+
+static int __init sas_class_init(void)
+{
+	sas_task_cache = kmem_cache_create("sas_task", sizeof(struct sas_task),
+					   0, SLAB_HWCACHE_ALIGN, NULL, NULL);
+	if (!sas_task_cache)
+		return -ENOMEM;
+
+	return 0;
+}
+
+static void __exit sas_class_exit(void)
+{
+	kmem_cache_destroy(sas_task_cache);
+}
+
+MODULE_AUTHOR("Luben Tuikov <luben_tuikov@adaptec.com>");
+MODULE_DESCRIPTION("SAS Transport Layer");
+MODULE_LICENSE("GPL v2");
+
+module_init(sas_class_init);
+module_exit(sas_class_exit);
+
+EXPORT_SYMBOL_GPL(sas_register_ha);
+EXPORT_SYMBOL_GPL(sas_unregister_ha);
diff --git a/drivers/scsi/libsas/sas_internal.h b/drivers/scsi/libsas/sas_internal.h
new file mode 100644
index 000000000000..89c397680846
--- /dev/null
+++ b/drivers/scsi/libsas/sas_internal.h
@@ -0,0 +1,146 @@
+/*
+ * Serial Attached SCSI (SAS) class internal header file
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
+ * USA
+ *
+ */
+
+#ifndef _SAS_INTERNAL_H_
+#define _SAS_INTERNAL_H_
+
+#include <scsi/scsi.h>
+#include <scsi/scsi_host.h>
+#include <scsi/scsi_transport_sas.h>
+#include <scsi/libsas.h>
+
+#define sas_printk(fmt, ...) printk(KERN_NOTICE "sas: " fmt, ## __VA_ARGS__)
+
+#ifdef SAS_DEBUG
+#define SAS_DPRINTK(fmt, ...) printk(KERN_NOTICE "sas: " fmt, ## __VA_ARGS__)
+#else
+#define SAS_DPRINTK(fmt, ...)
+#endif
+
+void sas_scsi_recover_host(struct Scsi_Host *shost);
+
+int sas_show_class(enum sas_class class, char *buf);
+int sas_show_proto(enum sas_proto proto, char *buf);
+int sas_show_linkrate(enum sas_phy_linkrate linkrate, char *buf);
+int sas_show_oob_mode(enum sas_oob_mode oob_mode, char *buf);
+
+int  sas_register_phys(struct sas_ha_struct *sas_ha);
+void sas_unregister_phys(struct sas_ha_struct *sas_ha);
+
+int  sas_register_ports(struct sas_ha_struct *sas_ha);
+void sas_unregister_ports(struct sas_ha_struct *sas_ha);
+
+enum scsi_eh_timer_return sas_scsi_timed_out(struct scsi_cmnd *);
+
+int  sas_init_queue(struct sas_ha_struct *sas_ha);
+int  sas_init_events(struct sas_ha_struct *sas_ha);
+void sas_shutdown_queue(struct sas_ha_struct *sas_ha);
+
+void sas_deform_port(struct asd_sas_phy *phy);
+
+void sas_porte_bytes_dmaed(void *);
+void sas_porte_broadcast_rcvd(void *);
+void sas_porte_link_reset_err(void *);
+void sas_porte_timer_event(void *);
+void sas_porte_hard_reset(void *);
+
+int sas_notify_lldd_dev_found(struct domain_device *);
+void sas_notify_lldd_dev_gone(struct domain_device *);
+
+int sas_smp_phy_control(struct domain_device *dev, int phy_id,
+			enum phy_func phy_func);
+int sas_smp_get_phy_events(struct sas_phy *phy);
+
+struct domain_device *sas_find_dev_by_rphy(struct sas_rphy *rphy);
+
+void sas_hae_reset(void *);
+
+static inline void sas_queue_event(int event, spinlock_t *lock,
+				   unsigned long *pending,
+				   struct work_struct *work,
+				   struct Scsi_Host *shost)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(lock, flags);
+	if (test_bit(event, pending)) {
+		spin_unlock_irqrestore(lock, flags);
+		return;
+	}
+	__set_bit(event, pending);
+	spin_unlock_irqrestore(lock, flags);
+	scsi_queue_work(shost, work);
+}
+
+static inline void sas_begin_event(int event, spinlock_t *lock,
+				   unsigned long *pending)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(lock, flags);
+	__clear_bit(event, pending);
+	spin_unlock_irqrestore(lock, flags);
+}
+
+static inline void sas_fill_in_rphy(struct domain_device *dev,
+				    struct sas_rphy *rphy)
+{
+	rphy->identify.sas_address = SAS_ADDR(dev->sas_addr);
+	rphy->identify.initiator_port_protocols = dev->iproto;
+	rphy->identify.target_port_protocols = dev->tproto;
+	switch (dev->dev_type) {
+	case SATA_DEV:
+		/* FIXME: need sata device type */
+	case SAS_END_DEV:
+		rphy->identify.device_type = SAS_END_DEVICE;
+		break;
+	case EDGE_DEV:
+		rphy->identify.device_type = SAS_EDGE_EXPANDER_DEVICE;
+		break;
+	case FANOUT_DEV:
+		rphy->identify.device_type = SAS_FANOUT_EXPANDER_DEVICE;
+		break;
+	default:
+		rphy->identify.device_type = SAS_PHY_UNUSED;
+		break;
+	}
+}
+
+static inline void sas_add_parent_port(struct domain_device *dev, int phy_id)
+{
+	struct expander_device *ex = &dev->ex_dev;
+	struct ex_phy *ex_phy = &ex->ex_phy[phy_id];
+
+	if (!ex->parent_port) {
+		ex->parent_port = sas_port_alloc(&dev->rphy->dev, phy_id);
+		/* FIXME: error handling */
+		BUG_ON(!ex->parent_port);
+		BUG_ON(sas_port_add(ex->parent_port));
+		sas_port_mark_backlink(ex->parent_port);
+	}
+	sas_port_add_phy(ex->parent_port, ex_phy->phy);
+}
+
+#endif /* _SAS_INTERNAL_H_ */
diff --git a/drivers/scsi/libsas/sas_phy.c b/drivers/scsi/libsas/sas_phy.c
new file mode 100644
index 000000000000..024ab00e70d2
--- /dev/null
+++ b/drivers/scsi/libsas/sas_phy.c
@@ -0,0 +1,157 @@
+/*
+ * Serial Attached SCSI (SAS) Phy class
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#include "sas_internal.h"
+#include <scsi/scsi_host.h>
+#include <scsi/scsi_transport.h>
+#include <scsi/scsi_transport_sas.h>
+#include "../scsi_sas_internal.h"
+
+/* ---------- Phy events ---------- */
+
+static void sas_phye_loss_of_signal(void *data)
+{
+	struct asd_sas_phy *phy = data;
+
+	sas_begin_event(PHYE_LOSS_OF_SIGNAL, &phy->ha->event_lock,
+			&phy->phy_events_pending);
+	phy->error = 0;
+	sas_deform_port(phy);
+}
+
+static void sas_phye_oob_done(void *data)
+{
+	struct asd_sas_phy *phy = data;
+
+	sas_begin_event(PHYE_OOB_DONE, &phy->ha->event_lock,
+			&phy->phy_events_pending);
+	phy->error = 0;
+}
+
+static void sas_phye_oob_error(void *data)
+{
+	struct asd_sas_phy *phy = data;
+	struct sas_ha_struct *sas_ha = phy->ha;
+	struct asd_sas_port *port = phy->port;
+	struct sas_internal *i =
+		to_sas_internal(sas_ha->core.shost->transportt);
+
+	sas_begin_event(PHYE_OOB_ERROR, &phy->ha->event_lock,
+			&phy->phy_events_pending);
+
+	sas_deform_port(phy);
+
+	if (!port && phy->enabled && i->dft->lldd_control_phy) {
+		phy->error++;
+		switch (phy->error) {
+		case 1:
+		case 2:
+			i->dft->lldd_control_phy(phy, PHY_FUNC_HARD_RESET);
+			break;
+		case 3:
+		default:
+			phy->error = 0;
+			phy->enabled = 0;
+			i->dft->lldd_control_phy(phy, PHY_FUNC_DISABLE);
+			break;
+		}
+	}
+}
+
+static void sas_phye_spinup_hold(void *data)
+{
+	struct asd_sas_phy *phy = data;
+	struct sas_ha_struct *sas_ha = phy->ha;
+	struct sas_internal *i =
+		to_sas_internal(sas_ha->core.shost->transportt);
+
+	sas_begin_event(PHYE_SPINUP_HOLD, &phy->ha->event_lock,
+			&phy->phy_events_pending);
+
+	phy->error = 0;
+	i->dft->lldd_control_phy(phy, PHY_FUNC_RELEASE_SPINUP_HOLD);
+}
+
+/* ---------- Phy class registration ---------- */
+
+int sas_register_phys(struct sas_ha_struct *sas_ha)
+{
+	int i;
+
+	static void (*sas_phy_event_fns[PHY_NUM_EVENTS])(void *) = {
+		[PHYE_LOSS_OF_SIGNAL] = sas_phye_loss_of_signal,
+		[PHYE_OOB_DONE] = sas_phye_oob_done,
+		[PHYE_OOB_ERROR] = sas_phye_oob_error,
+		[PHYE_SPINUP_HOLD] = sas_phye_spinup_hold,
+	};
+
+	static void (*sas_port_event_fns[PORT_NUM_EVENTS])(void *) = {
+		[PORTE_BYTES_DMAED] = sas_porte_bytes_dmaed,
+		[PORTE_BROADCAST_RCVD] = sas_porte_broadcast_rcvd,
+		[PORTE_LINK_RESET_ERR] = sas_porte_link_reset_err,
+		[PORTE_TIMER_EVENT] = sas_porte_timer_event,
+		[PORTE_HARD_RESET] = sas_porte_hard_reset,
+	};
+
+	/* Now register the phys. */
+	for (i = 0; i < sas_ha->num_phys; i++) {
+		int k;
+		struct asd_sas_phy *phy = sas_ha->sas_phy[i];
+
+		phy->error = 0;
+		INIT_LIST_HEAD(&phy->port_phy_el);
+		for (k = 0; k < PORT_NUM_EVENTS; k++)
+			INIT_WORK(&phy->port_events[k], sas_port_event_fns[k],
+				  phy);
+
+		for (k = 0; k < PHY_NUM_EVENTS; k++)
+			INIT_WORK(&phy->phy_events[k], sas_phy_event_fns[k],
+				  phy);
+		phy->port = NULL;
+		phy->ha = sas_ha;
+		spin_lock_init(&phy->frame_rcvd_lock);
+		spin_lock_init(&phy->sas_prim_lock);
+		phy->frame_rcvd_size = 0;
+
+		phy->phy = sas_phy_alloc(&sas_ha->core.shost->shost_gendev,
+					 i);
+		if (!phy->phy)
+			return -ENOMEM;
+
+		phy->phy->identify.initiator_port_protocols =
+			phy->iproto;
+		phy->phy->identify.target_port_protocols = phy->tproto;
+		phy->phy->identify.sas_address = SAS_ADDR(sas_ha->sas_addr);
+		phy->phy->identify.phy_identifier = i;
+		phy->phy->minimum_linkrate_hw = SAS_LINK_RATE_1_5_GBPS;
+		phy->phy->maximum_linkrate_hw = SAS_LINK_RATE_3_0_GBPS;
+		phy->phy->minimum_linkrate = SAS_LINK_RATE_1_5_GBPS;
+		phy->phy->maximum_linkrate = SAS_LINK_RATE_3_0_GBPS;
+		phy->phy->negotiated_linkrate = SAS_LINK_RATE_UNKNOWN;
+
+		sas_phy_add(phy->phy);
+	}
+
+	return 0;
+}
diff --git a/drivers/scsi/libsas/sas_port.c b/drivers/scsi/libsas/sas_port.c
new file mode 100644
index 000000000000..253cdcf306a2
--- /dev/null
+++ b/drivers/scsi/libsas/sas_port.c
@@ -0,0 +1,279 @@
+/*
+ * Serial Attached SCSI (SAS) Port class
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#include "sas_internal.h"
+
+#include <scsi/scsi_transport.h>
+#include <scsi/scsi_transport_sas.h>
+#include "../scsi_sas_internal.h"
+
+/**
+ * sas_form_port -- add this phy to a port
+ * @phy: the phy of interest
+ *
+ * This function adds this phy to an existing port, thus creating a wide
+ * port, or it creates a port and adds the phy to the port.
+ */
+static void sas_form_port(struct asd_sas_phy *phy)
+{
+	int i;
+	struct sas_ha_struct *sas_ha = phy->ha;
+	struct asd_sas_port *port = phy->port;
+	struct sas_internal *si =
+		to_sas_internal(sas_ha->core.shost->transportt);
+
+	if (port) {
+		if (memcmp(port->attached_sas_addr, phy->attached_sas_addr,
+			   SAS_ADDR_SIZE) == 0)
+			sas_deform_port(phy);
+		else {
+			SAS_DPRINTK("%s: phy%d belongs to port%d already(%d)!\n",
+				    __FUNCTION__, phy->id, phy->port->id,
+				    phy->port->num_phys);
+			return;
+		}
+	}
+
+	/* find a port */
+	spin_lock(&sas_ha->phy_port_lock);
+	for (i = 0; i < sas_ha->num_phys; i++) {
+		port = sas_ha->sas_port[i];
+		spin_lock(&port->phy_list_lock);
+		if (*(u64 *) port->sas_addr &&
+		    memcmp(port->attached_sas_addr,
+			   phy->attached_sas_addr, SAS_ADDR_SIZE) == 0 &&
+		    port->num_phys > 0) {
+			/* wide port */
+			SAS_DPRINTK("phy%d matched wide port%d\n", phy->id,
+				    port->id);
+			break;
+		} else if (*(u64 *) port->sas_addr == 0 && port->num_phys==0) {
+			memcpy(port->sas_addr, phy->sas_addr, SAS_ADDR_SIZE);
+			break;
+		}
+		spin_unlock(&port->phy_list_lock);
+	}
+
+	if (i >= sas_ha->num_phys) {
+		printk(KERN_NOTICE "%s: couldn't find a free port, bug?\n",
+		       __FUNCTION__);
+		spin_unlock(&sas_ha->phy_port_lock);
+		return;
+	}
+
+	/* add the phy to the port */
+	list_add_tail(&phy->port_phy_el, &port->phy_list);
+	phy->port = port;
+	port->num_phys++;
+	port->phy_mask |= (1U << phy->id);
+
+	if (!port->phy)
+		port->phy = phy->phy;
+
+	SAS_DPRINTK("phy%d added to port%d, phy_mask:0x%x\n", phy->id,
+		    port->id, port->phy_mask);
+
+	if (*(u64 *)port->attached_sas_addr == 0) {
+		port->class = phy->class;
+		memcpy(port->attached_sas_addr, phy->attached_sas_addr,
+		       SAS_ADDR_SIZE);
+		port->iproto = phy->iproto;
+		port->tproto = phy->tproto;
+		port->oob_mode = phy->oob_mode;
+		port->linkrate = phy->linkrate;
+	} else
+		port->linkrate = max(port->linkrate, phy->linkrate);
+	spin_unlock(&port->phy_list_lock);
+	spin_unlock(&sas_ha->phy_port_lock);
+
+	if (!port->port) {
+		port->port = sas_port_alloc(phy->phy->dev.parent, port->id);
+		BUG_ON(!port->port);
+		sas_port_add(port->port);
+	}
+	sas_port_add_phy(port->port, phy->phy);
+
+	if (port->port_dev)
+		port->port_dev->pathways = port->num_phys;
+
+	/* Tell the LLDD about this port formation. */
+	if (si->dft->lldd_port_formed)
+		si->dft->lldd_port_formed(phy);
+
+	sas_discover_event(phy->port, DISCE_DISCOVER_DOMAIN);
+}
+
+/**
+ * sas_deform_port -- remove this phy from the port it belongs to
+ * @phy: the phy of interest
+ *
+ * This is called when the physical link to the other phy has been
+ * lost (on this phy), in Event thread context. We cannot delay here.
+ */
+void sas_deform_port(struct asd_sas_phy *phy)
+{
+	struct sas_ha_struct *sas_ha = phy->ha;
+	struct asd_sas_port *port = phy->port;
+	struct sas_internal *si =
+		to_sas_internal(sas_ha->core.shost->transportt);
+
+	if (!port)
+		return;		  /* done by a phy event */
+
+	if (port->port_dev)
+		port->port_dev->pathways--;
+
+	if (port->num_phys == 1) {
+		sas_unregister_domain_devices(port);
+		sas_port_delete(port->port);
+		port->port = NULL;
+	} else
+		sas_port_delete_phy(port->port, phy->phy);
+
+
+	if (si->dft->lldd_port_deformed)
+		si->dft->lldd_port_deformed(phy);
+
+	spin_lock(&sas_ha->phy_port_lock);
+	spin_lock(&port->phy_list_lock);
+
+	list_del_init(&phy->port_phy_el);
+	phy->port = NULL;
+	port->num_phys--;
+	port->phy_mask &= ~(1U << phy->id);
+
+	if (port->num_phys == 0) {
+		INIT_LIST_HEAD(&port->phy_list);
+		memset(port->sas_addr, 0, SAS_ADDR_SIZE);
+		memset(port->attached_sas_addr, 0, SAS_ADDR_SIZE);
+		port->class = 0;
+		port->iproto = 0;
+		port->tproto = 0;
+		port->oob_mode = 0;
+		port->phy_mask = 0;
+	}
+	spin_unlock(&port->phy_list_lock);
+	spin_unlock(&sas_ha->phy_port_lock);
+
+	return;
+}
+
+/* ---------- SAS port events ---------- */
+
+void sas_porte_bytes_dmaed(void *data)
+{
+	struct asd_sas_phy *phy = data;
+
+	sas_begin_event(PORTE_BYTES_DMAED, &phy->ha->event_lock,
+			&phy->port_events_pending);
+
+	sas_form_port(phy);
+}
+
+void sas_porte_broadcast_rcvd(void *data)
+{
+	unsigned long flags;
+	u32 prim;
+	struct asd_sas_phy *phy = data;
+
+	sas_begin_event(PORTE_BROADCAST_RCVD, &phy->ha->event_lock,
+			&phy->port_events_pending);
+
+	spin_lock_irqsave(&phy->sas_prim_lock, flags);
+	prim = phy->sas_prim;
+	spin_unlock_irqrestore(&phy->sas_prim_lock, flags);
+
+	SAS_DPRINTK("broadcast received: %d\n", prim);
+	sas_discover_event(phy->port, DISCE_REVALIDATE_DOMAIN);
+}
+
+void sas_porte_link_reset_err(void *data)
+{
+	struct asd_sas_phy *phy = data;
+
+	sas_begin_event(PORTE_LINK_RESET_ERR, &phy->ha->event_lock,
+			&phy->port_events_pending);
+
+	sas_deform_port(phy);
+}
+
+void sas_porte_timer_event(void *data)
+{
+	struct asd_sas_phy *phy = data;
+
+	sas_begin_event(PORTE_TIMER_EVENT, &phy->ha->event_lock,
+			&phy->port_events_pending);
+
+	sas_deform_port(phy);
+}
+
+void sas_porte_hard_reset(void *data)
+{
+	struct asd_sas_phy *phy = data;
+
+	sas_begin_event(PORTE_HARD_RESET, &phy->ha->event_lock,
+			&phy->port_events_pending);
+
+	sas_deform_port(phy);
+}
+
+/* ---------- SAS port registration ---------- */
+
+static void sas_init_port(struct asd_sas_port *port,
+			  struct sas_ha_struct *sas_ha, int i)
+{
+	port->id = i;
+	INIT_LIST_HEAD(&port->dev_list);
+	spin_lock_init(&port->phy_list_lock);
+	INIT_LIST_HEAD(&port->phy_list);
+	port->num_phys = 0;
+	port->phy_mask = 0;
+	port->ha = sas_ha;
+
+	spin_lock_init(&port->dev_list_lock);
+}
+
+int sas_register_ports(struct sas_ha_struct *sas_ha)
+{
+	int i;
+
+	/* initialize the ports and discovery */
+	for (i = 0; i < sas_ha->num_phys; i++) {
+		struct asd_sas_port *port = sas_ha->sas_port[i];
+
+		sas_init_port(port, sas_ha, i);
+		sas_init_disc(&port->disc, port);
+	}
+	return 0;
+}
+
+void sas_unregister_ports(struct sas_ha_struct *sas_ha)
+{
+	int i;
+
+	for (i = 0; i < sas_ha->num_phys; i++)
+		if (sas_ha->sas_phy[i]->port)
+			sas_deform_port(sas_ha->sas_phy[i]);
+
+}
diff --git a/drivers/scsi/libsas/sas_scsi_host.c b/drivers/scsi/libsas/sas_scsi_host.c
new file mode 100644
index 000000000000..43e0e4e36934
--- /dev/null
+++ b/drivers/scsi/libsas/sas_scsi_host.c
@@ -0,0 +1,786 @@
+/*
+ * Serial Attached SCSI (SAS) class SCSI Host glue.
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
+ * USA
+ *
+ */
+
+#include "sas_internal.h"
+
+#include <scsi/scsi_host.h>
+#include <scsi/scsi_device.h>
+#include <scsi/scsi_tcq.h>
+#include <scsi/scsi.h>
+#include <scsi/scsi_transport.h>
+#include <scsi/scsi_transport_sas.h>
+#include "../scsi_sas_internal.h"
+
+#include <linux/err.h>
+#include <linux/blkdev.h>
+#include <linux/scatterlist.h>
+
+/* ---------- SCSI Host glue ---------- */
+
+#define TO_SAS_TASK(_scsi_cmd)  ((void *)(_scsi_cmd)->host_scribble)
+#define ASSIGN_SAS_TASK(_sc, _t) do { (_sc)->host_scribble = (void *) _t; } while (0)
+
+static void sas_scsi_task_done(struct sas_task *task)
+{
+	struct task_status_struct *ts = &task->task_status;
+	struct scsi_cmnd *sc = task->uldd_task;
+	unsigned ts_flags = task->task_state_flags;
+	int hs = 0, stat = 0;
+
+	if (unlikely(!sc)) {
+		SAS_DPRINTK("task_done called with non existing SCSI cmnd!\n");
+		list_del_init(&task->list);
+		sas_free_task(task);
+		return;
+	}
+
+	if (ts->resp == SAS_TASK_UNDELIVERED) {
+		/* transport error */
+		hs = DID_NO_CONNECT;
+	} else { /* ts->resp == SAS_TASK_COMPLETE */
+		/* task delivered, what happened afterwards? */
+		switch (ts->stat) {
+		case SAS_DEV_NO_RESPONSE:
+		case SAS_INTERRUPTED:
+		case SAS_PHY_DOWN:
+		case SAS_NAK_R_ERR:
+		case SAS_OPEN_TO:
+			hs = DID_NO_CONNECT;
+			break;
+		case SAS_DATA_UNDERRUN:
+			sc->resid = ts->residual;
+			if (sc->request_bufflen - sc->resid < sc->underflow)
+				hs = DID_ERROR;
+			break;
+		case SAS_DATA_OVERRUN:
+			hs = DID_ERROR;
+			break;
+		case SAS_QUEUE_FULL:
+			hs = DID_SOFT_ERROR; /* retry */
+			break;
+		case SAS_DEVICE_UNKNOWN:
+			hs = DID_BAD_TARGET;
+			break;
+		case SAS_SG_ERR:
+			hs = DID_PARITY;
+			break;
+		case SAS_OPEN_REJECT:
+			if (ts->open_rej_reason == SAS_OREJ_RSVD_RETRY)
+				hs = DID_SOFT_ERROR; /* retry */
+			else
+				hs = DID_ERROR;
+			break;
+		case SAS_PROTO_RESPONSE:
+			SAS_DPRINTK("LLDD:%s sent SAS_PROTO_RESP for an SSP "
+				    "task; please report this\n",
+				    task->dev->port->ha->sas_ha_name);
+			break;
+		case SAS_ABORTED_TASK:
+			hs = DID_ABORT;
+			break;
+		case SAM_CHECK_COND:
+			memcpy(sc->sense_buffer, ts->buf,
+			       max(SCSI_SENSE_BUFFERSIZE, ts->buf_valid_size));
+			stat = SAM_CHECK_COND;
+			break;
+		default:
+			stat = ts->stat;
+			break;
+		}
+	}
+	ASSIGN_SAS_TASK(sc, NULL);
+	sc->result = (hs << 16) | stat;
+	list_del_init(&task->list);
+	sas_free_task(task);
+	/* This is very ugly but this is how SCSI Core works. */
+	if (ts_flags & SAS_TASK_STATE_ABORTED)
+		scsi_finish_command(sc);
+	else
+		sc->scsi_done(sc);
+}
+
+static enum task_attribute sas_scsi_get_task_attr(struct scsi_cmnd *cmd)
+{
+	enum task_attribute ta = TASK_ATTR_SIMPLE;
+	if (cmd->request && blk_rq_tagged(cmd->request)) {
+		if (cmd->device->ordered_tags &&
+		    (cmd->request->flags & REQ_HARDBARRIER))
+			ta = TASK_ATTR_HOQ;
+	}
+	return ta;
+}
+
+static struct sas_task *sas_create_task(struct scsi_cmnd *cmd,
+					       struct domain_device *dev,
+					       unsigned long gfp_flags)
+{
+	struct sas_task *task = sas_alloc_task(gfp_flags);
+	struct scsi_lun lun;
+
+	if (!task)
+		return NULL;
+
+	*(u32 *)cmd->sense_buffer = 0;
+	task->uldd_task = cmd;
+	ASSIGN_SAS_TASK(cmd, task);
+
+	task->dev = dev;
+	task->task_proto = task->dev->tproto; /* BUG_ON(!SSP) */
+
+	task->ssp_task.retry_count = 1;
+	int_to_scsilun(cmd->device->lun, &lun);
+	memcpy(task->ssp_task.LUN, &lun.scsi_lun, 8);
+	task->ssp_task.task_attr = sas_scsi_get_task_attr(cmd);
+	memcpy(task->ssp_task.cdb, cmd->cmnd, 16);
+
+	task->scatter = cmd->request_buffer;
+	task->num_scatter = cmd->use_sg;
+	task->total_xfer_len = cmd->request_bufflen;
+	task->data_dir = cmd->sc_data_direction;
+
+	task->task_done = sas_scsi_task_done;
+
+	return task;
+}
+
+static int sas_queue_up(struct sas_task *task)
+{
+	struct sas_ha_struct *sas_ha = task->dev->port->ha;
+	struct scsi_core *core = &sas_ha->core;
+	unsigned long flags;
+	LIST_HEAD(list);
+
+	spin_lock_irqsave(&core->task_queue_lock, flags);
+	if (sas_ha->lldd_queue_size < core->task_queue_size + 1) {
+		spin_unlock_irqrestore(&core->task_queue_lock, flags);
+		return -SAS_QUEUE_FULL;
+	}
+	list_add_tail(&task->list, &core->task_queue);
+	core->task_queue_size += 1;
+	spin_unlock_irqrestore(&core->task_queue_lock, flags);
+	up(&core->queue_thread_sema);
+
+	return 0;
+}
+
+/**
+ * sas_queuecommand -- Enqueue a command for processing
+ * @parameters: See SCSI Core documentation
+ *
+ * Note: XXX: Remove the host unlock/lock pair when SCSI Core can
+ * call us without holding an IRQ spinlock...
+ */
+int sas_queuecommand(struct scsi_cmnd *cmd,
+		     void (*scsi_done)(struct scsi_cmnd *))
+{
+	int res = 0;
+	struct domain_device *dev = cmd_to_domain_dev(cmd);
+	struct Scsi_Host *host = cmd->device->host;
+	struct sas_internal *i = to_sas_internal(host->transportt);
+
+	spin_unlock_irq(host->host_lock);
+
+	{
+		struct sas_ha_struct *sas_ha = dev->port->ha;
+		struct sas_task *task;
+
+		res = -ENOMEM;
+		task = sas_create_task(cmd, dev, GFP_ATOMIC);
+		if (!task)
+			goto out;
+
+		cmd->scsi_done = scsi_done;
+		/* Queue up, Direct Mode or Task Collector Mode. */
+		if (sas_ha->lldd_max_execute_num < 2)
+			res = i->dft->lldd_execute_task(task, 1, GFP_ATOMIC);
+		else
+			res = sas_queue_up(task);
+
+		/* Examine */
+		if (res) {
+			SAS_DPRINTK("lldd_execute_task returned: %d\n", res);
+			ASSIGN_SAS_TASK(cmd, NULL);
+			sas_free_task(task);
+			if (res == -SAS_QUEUE_FULL) {
+				cmd->result = DID_SOFT_ERROR << 16; /* retry */
+				res = 0;
+				scsi_done(cmd);
+			}
+			goto out;
+		}
+	}
+out:
+	spin_lock_irq(host->host_lock);
+	return res;
+}
+
+static void sas_scsi_clear_queue_lu(struct list_head *error_q, struct scsi_cmnd *my_cmd)
+{
+	struct scsi_cmnd *cmd, *n;
+
+	list_for_each_entry_safe(cmd, n, error_q, eh_entry) {
+		if (cmd == my_cmd)
+			list_del_init(&cmd->eh_entry);
+	}
+}
+
+static void sas_scsi_clear_queue_I_T(struct list_head *error_q,
+				     struct domain_device *dev)
+{
+	struct scsi_cmnd *cmd, *n;
+
+	list_for_each_entry_safe(cmd, n, error_q, eh_entry) {
+		struct domain_device *x = cmd_to_domain_dev(cmd);
+
+		if (x == dev)
+			list_del_init(&cmd->eh_entry);
+	}
+}
+
+static void sas_scsi_clear_queue_port(struct list_head *error_q,
+				      struct asd_sas_port *port)
+{
+	struct scsi_cmnd *cmd, *n;
+
+	list_for_each_entry_safe(cmd, n, error_q, eh_entry) {
+		struct domain_device *dev = cmd_to_domain_dev(cmd);
+		struct asd_sas_port *x = dev->port;
+
+		if (x == port)
+			list_del_init(&cmd->eh_entry);
+	}
+}
+
+enum task_disposition {
+	TASK_IS_DONE,
+	TASK_IS_ABORTED,
+	TASK_IS_AT_LU,
+	TASK_IS_NOT_AT_LU,
+};
+
+static enum task_disposition sas_scsi_find_task(struct sas_task *task)
+{
+	struct sas_ha_struct *ha = task->dev->port->ha;
+	unsigned long flags;
+	int i, res;
+	struct sas_internal *si =
+		to_sas_internal(task->dev->port->ha->core.shost->transportt);
+
+	if (ha->lldd_max_execute_num > 1) {
+		struct scsi_core *core = &ha->core;
+		struct sas_task *t, *n;
+
+		spin_lock_irqsave(&core->task_queue_lock, flags);
+		list_for_each_entry_safe(t, n, &core->task_queue, list) {
+			if (task == t) {
+				list_del_init(&t->list);
+				spin_unlock_irqrestore(&core->task_queue_lock,
+						       flags);
+				SAS_DPRINTK("%s: task 0x%p aborted from "
+					    "task_queue\n",
+					    __FUNCTION__, task);
+				return TASK_IS_ABORTED;
+			}
+		}
+		spin_unlock_irqrestore(&core->task_queue_lock, flags);
+	}
+
+	for (i = 0; i < 5; i++) {
+		SAS_DPRINTK("%s: aborting task 0x%p\n", __FUNCTION__, task);
+		res = si->dft->lldd_abort_task(task);
+
+		spin_lock_irqsave(&task->task_state_lock, flags);
+		if (task->task_state_flags & SAS_TASK_STATE_DONE) {
+			spin_unlock_irqrestore(&task->task_state_lock, flags);
+			SAS_DPRINTK("%s: task 0x%p is done\n", __FUNCTION__,
+				    task);
+			return TASK_IS_DONE;
+		}
+		spin_unlock_irqrestore(&task->task_state_lock, flags);
+
+		if (res == TMF_RESP_FUNC_COMPLETE) {
+			SAS_DPRINTK("%s: task 0x%p is aborted\n",
+				    __FUNCTION__, task);
+			return TASK_IS_ABORTED;
+		} else if (si->dft->lldd_query_task) {
+			SAS_DPRINTK("%s: querying task 0x%p\n",
+				    __FUNCTION__, task);
+			res = si->dft->lldd_query_task(task);
+			if (res == TMF_RESP_FUNC_SUCC) {
+				SAS_DPRINTK("%s: task 0x%p at LU\n",
+					    __FUNCTION__, task);
+				return TASK_IS_AT_LU;
+			} else if (res == TMF_RESP_FUNC_COMPLETE) {
+				SAS_DPRINTK("%s: task 0x%p not at LU\n",
+					    __FUNCTION__, task);
+				return TASK_IS_NOT_AT_LU;
+			}
+		}
+	}
+	return res;
+}
+
+static int sas_recover_lu(struct domain_device *dev, struct scsi_cmnd *cmd)
+{
+	int res = TMF_RESP_FUNC_FAILED;
+	struct scsi_lun lun;
+	struct sas_internal *i =
+		to_sas_internal(dev->port->ha->core.shost->transportt);
+
+	int_to_scsilun(cmd->device->lun, &lun);
+
+	SAS_DPRINTK("eh: device %llx LUN %x has the task\n",
+		    SAS_ADDR(dev->sas_addr),
+		    cmd->device->lun);
+
+	if (i->dft->lldd_abort_task_set)
+		res = i->dft->lldd_abort_task_set(dev, lun.scsi_lun);
+
+	if (res == TMF_RESP_FUNC_FAILED) {
+		if (i->dft->lldd_clear_task_set)
+			res = i->dft->lldd_clear_task_set(dev, lun.scsi_lun);
+	}
+
+	if (res == TMF_RESP_FUNC_FAILED) {
+		if (i->dft->lldd_lu_reset)
+			res = i->dft->lldd_lu_reset(dev, lun.scsi_lun);
+	}
+
+	return res;
+}
+
+static int sas_recover_I_T(struct domain_device *dev)
+{
+	int res = TMF_RESP_FUNC_FAILED;
+	struct sas_internal *i =
+		to_sas_internal(dev->port->ha->core.shost->transportt);
+
+	SAS_DPRINTK("I_T nexus reset for dev %016llx\n",
+		    SAS_ADDR(dev->sas_addr));
+
+	if (i->dft->lldd_I_T_nexus_reset)
+		res = i->dft->lldd_I_T_nexus_reset(dev);
+
+	return res;
+}
+
+void sas_scsi_recover_host(struct Scsi_Host *shost)
+{
+	struct sas_ha_struct *ha = SHOST_TO_SAS_HA(shost);
+	unsigned long flags;
+	LIST_HEAD(error_q);
+	struct scsi_cmnd *cmd, *n;
+	enum task_disposition res = TASK_IS_DONE;
+	int tmf_resp;
+	struct sas_internal *i = to_sas_internal(shost->transportt);
+
+	spin_lock_irqsave(shost->host_lock, flags);
+	list_splice_init(&shost->eh_cmd_q, &error_q);
+	spin_unlock_irqrestore(shost->host_lock, flags);
+
+	SAS_DPRINTK("Enter %s\n", __FUNCTION__);
+
+	/* All tasks on this list were marked SAS_TASK_STATE_ABORTED
+	 * by sas_scsi_timed_out() callback.
+	 */
+Again:
+	SAS_DPRINTK("going over list...\n");
+	list_for_each_entry_safe(cmd, n, &error_q, eh_entry) {
+		struct sas_task *task = TO_SAS_TASK(cmd);
+
+		SAS_DPRINTK("trying to find task 0x%p\n", task);
+		list_del_init(&cmd->eh_entry);
+		res = sas_scsi_find_task(task);
+
+		cmd->eh_eflags = 0;
+		shost->host_failed--;
+
+		switch (res) {
+		case TASK_IS_DONE:
+			SAS_DPRINTK("%s: task 0x%p is done\n", __FUNCTION__,
+				    task);
+			task->task_done(task);
+			continue;
+		case TASK_IS_ABORTED:
+			SAS_DPRINTK("%s: task 0x%p is aborted\n",
+				    __FUNCTION__, task);
+			task->task_done(task);
+			continue;
+		case TASK_IS_AT_LU:
+			SAS_DPRINTK("task 0x%p is at LU: lu recover\n", task);
+			tmf_resp = sas_recover_lu(task->dev, cmd);
+			if (tmf_resp == TMF_RESP_FUNC_COMPLETE) {
+				SAS_DPRINTK("dev %016llx LU %x is "
+					    "recovered\n",
+					    SAS_ADDR(task->dev),
+					    cmd->device->lun);
+				task->task_done(task);
+				sas_scsi_clear_queue_lu(&error_q, cmd);
+				goto Again;
+			}
+			/* fallthrough */
+		case TASK_IS_NOT_AT_LU:
+			SAS_DPRINTK("task 0x%p is not at LU: I_T recover\n",
+				    task);
+			tmf_resp = sas_recover_I_T(task->dev);
+			if (tmf_resp == TMF_RESP_FUNC_COMPLETE) {
+				SAS_DPRINTK("I_T %016llx recovered\n",
+					    SAS_ADDR(task->dev->sas_addr));
+				task->task_done(task);
+				sas_scsi_clear_queue_I_T(&error_q, task->dev);
+				goto Again;
+			}
+			/* Hammer time :-) */
+			if (i->dft->lldd_clear_nexus_port) {
+				struct asd_sas_port *port = task->dev->port;
+				SAS_DPRINTK("clearing nexus for port:%d\n",
+					    port->id);
+				res = i->dft->lldd_clear_nexus_port(port);
+				if (res == TMF_RESP_FUNC_COMPLETE) {
+					SAS_DPRINTK("clear nexus port:%d "
+						    "succeeded\n", port->id);
+					task->task_done(task);
+					sas_scsi_clear_queue_port(&error_q,
+								  port);
+					goto Again;
+				}
+			}
+			if (i->dft->lldd_clear_nexus_ha) {
+				SAS_DPRINTK("clear nexus ha\n");
+				res = i->dft->lldd_clear_nexus_ha(ha);
+				if (res == TMF_RESP_FUNC_COMPLETE) {
+					SAS_DPRINTK("clear nexus ha "
+						    "succeeded\n");
+					task->task_done(task);
+					goto out;
+				}
+			}
+			/* If we are here -- this means that no amount
+			 * of effort could recover from errors.  Quite
+			 * possibly the HA just disappeared.
+			 */
+			SAS_DPRINTK("error from  device %llx, LUN %x "
+				    "couldn't be recovered in any way\n",
+				    SAS_ADDR(task->dev->sas_addr),
+				    cmd->device->lun);
+
+			task->task_done(task);
+			goto clear_q;
+		}
+	}
+out:
+	SAS_DPRINTK("--- Exit %s\n", __FUNCTION__);
+	return;
+clear_q:
+	SAS_DPRINTK("--- Exit %s -- clear_q\n", __FUNCTION__);
+	list_for_each_entry_safe(cmd, n, &error_q, eh_entry) {
+		struct sas_task *task = TO_SAS_TASK(cmd);
+		list_del_init(&cmd->eh_entry);
+		task->task_done(task);
+	}
+}
+
+enum scsi_eh_timer_return sas_scsi_timed_out(struct scsi_cmnd *cmd)
+{
+	struct sas_task *task = TO_SAS_TASK(cmd);
+	unsigned long flags;
+
+	if (!task) {
+		SAS_DPRINTK("command 0x%p, task 0x%p, timed out: EH_HANDLED\n",
+			    cmd, task);
+		return EH_HANDLED;
+	}
+
+	spin_lock_irqsave(&task->task_state_lock, flags);
+	if (task->task_state_flags & SAS_TASK_STATE_DONE) {
+		spin_unlock_irqrestore(&task->task_state_lock, flags);
+		SAS_DPRINTK("command 0x%p, task 0x%p, timed out: EH_HANDLED\n",
+			    cmd, task);
+		return EH_HANDLED;
+	}
+	task->task_state_flags |= SAS_TASK_STATE_ABORTED;
+	spin_unlock_irqrestore(&task->task_state_lock, flags);
+
+	SAS_DPRINTK("command 0x%p, task 0x%p, timed out: EH_NOT_HANDLED\n",
+		    cmd, task);
+
+	return EH_NOT_HANDLED;
+}
+
+struct domain_device *sas_find_dev_by_rphy(struct sas_rphy *rphy)
+{
+	struct Scsi_Host *shost = dev_to_shost(rphy->dev.parent);
+	struct sas_ha_struct *ha = SHOST_TO_SAS_HA(shost);
+	struct domain_device *found_dev = NULL;
+	int i;
+
+	spin_lock(&ha->phy_port_lock);
+	for (i = 0; i < ha->num_phys; i++) {
+		struct asd_sas_port *port = ha->sas_port[i];
+		struct domain_device *dev;
+
+		spin_lock(&port->dev_list_lock);
+		list_for_each_entry(dev, &port->dev_list, dev_list_node) {
+			if (rphy == dev->rphy) {
+				found_dev = dev;
+				spin_unlock(&port->dev_list_lock);
+				goto found;
+			}
+		}
+		spin_unlock(&port->dev_list_lock);
+	}
+ found:
+	spin_unlock(&ha->phy_port_lock);
+
+	return found_dev;
+}
+
+static inline struct domain_device *sas_find_target(struct scsi_target *starget)
+{
+	struct sas_rphy *rphy = dev_to_rphy(starget->dev.parent);
+
+	return sas_find_dev_by_rphy(rphy);
+}
+
+int sas_target_alloc(struct scsi_target *starget)
+{
+	struct domain_device *found_dev = sas_find_target(starget);
+
+	if (!found_dev)
+		return -ENODEV;
+
+	starget->hostdata = found_dev;
+	return 0;
+}
+
+#define SAS_DEF_QD 32
+#define SAS_MAX_QD 64
+
+int sas_slave_configure(struct scsi_device *scsi_dev)
+{
+	struct domain_device *dev = sdev_to_domain_dev(scsi_dev);
+	struct sas_ha_struct *sas_ha;
+
+	BUG_ON(dev->rphy->identify.device_type != SAS_END_DEVICE);
+
+	sas_ha = dev->port->ha;
+
+	sas_read_port_mode_page(scsi_dev);
+
+	if (scsi_dev->tagged_supported) {
+		scsi_set_tag_type(scsi_dev, MSG_SIMPLE_TAG);
+		scsi_activate_tcq(scsi_dev, SAS_DEF_QD);
+	} else {
+		SAS_DPRINTK("device %llx, LUN %x doesn't support "
+			    "TCQ\n", SAS_ADDR(dev->sas_addr),
+			    scsi_dev->lun);
+		scsi_dev->tagged_supported = 0;
+		scsi_set_tag_type(scsi_dev, 0);
+		scsi_deactivate_tcq(scsi_dev, 1);
+	}
+
+	return 0;
+}
+
+void sas_slave_destroy(struct scsi_device *scsi_dev)
+{
+}
+
+int sas_change_queue_depth(struct scsi_device *scsi_dev, int new_depth)
+{
+	int res = min(new_depth, SAS_MAX_QD);
+
+	if (scsi_dev->tagged_supported)
+		scsi_adjust_queue_depth(scsi_dev, scsi_get_tag_type(scsi_dev),
+					res);
+	else {
+		struct domain_device *dev = sdev_to_domain_dev(scsi_dev);
+		sas_printk("device %llx LUN %x queue depth changed to 1\n",
+			   SAS_ADDR(dev->sas_addr),
+			   scsi_dev->lun);
+		scsi_adjust_queue_depth(scsi_dev, 0, 1);
+		res = 1;
+	}
+
+	return res;
+}
+
+int sas_change_queue_type(struct scsi_device *scsi_dev, int qt)
+{
+	if (!scsi_dev->tagged_supported)
+		return 0;
+
+	scsi_deactivate_tcq(scsi_dev, 1);
+
+	scsi_set_tag_type(scsi_dev, qt);
+	scsi_activate_tcq(scsi_dev, scsi_dev->queue_depth);
+
+	return qt;
+}
+
+int sas_bios_param(struct scsi_device *scsi_dev,
+			  struct block_device *bdev,
+			  sector_t capacity, int *hsc)
+{
+	hsc[0] = 255;
+	hsc[1] = 63;
+	sector_div(capacity, 255*63);
+	hsc[2] = capacity;
+
+	return 0;
+}
+
+/* ---------- Task Collector Thread implementation ---------- */
+
+static void sas_queue(struct sas_ha_struct *sas_ha)
+{
+	struct scsi_core *core = &sas_ha->core;
+	unsigned long flags;
+	LIST_HEAD(q);
+	int can_queue;
+	int res;
+	struct sas_internal *i = to_sas_internal(core->shost->transportt);
+
+	spin_lock_irqsave(&core->task_queue_lock, flags);
+	while (!core->queue_thread_kill &&
+	       !list_empty(&core->task_queue)) {
+
+		can_queue = sas_ha->lldd_queue_size - core->task_queue_size;
+		if (can_queue >= 0) {
+			can_queue = core->task_queue_size;
+			list_splice_init(&core->task_queue, &q);
+		} else {
+			struct list_head *a, *n;
+
+			can_queue = sas_ha->lldd_queue_size;
+			list_for_each_safe(a, n, &core->task_queue) {
+				list_move_tail(a, &q);
+				if (--can_queue == 0)
+					break;
+			}
+			can_queue = sas_ha->lldd_queue_size;
+		}
+		core->task_queue_size -= can_queue;
+		spin_unlock_irqrestore(&core->task_queue_lock, flags);
+		{
+			struct sas_task *task = list_entry(q.next,
+							   struct sas_task,
+							   list);
+			list_del_init(&q);
+			res = i->dft->lldd_execute_task(task, can_queue,
+							GFP_KERNEL);
+			if (unlikely(res))
+				__list_add(&q, task->list.prev, &task->list);
+		}
+		spin_lock_irqsave(&core->task_queue_lock, flags);
+		if (res) {
+			list_splice_init(&q, &core->task_queue); /*at head*/
+			core->task_queue_size += can_queue;
+		}
+	}
+	spin_unlock_irqrestore(&core->task_queue_lock, flags);
+}
+
+static DECLARE_COMPLETION(queue_th_comp);
+
+/**
+ * sas_queue_thread -- The Task Collector thread
+ * @_sas_ha: pointer to struct sas_ha
+ */
+static int sas_queue_thread(void *_sas_ha)
+{
+	struct sas_ha_struct *sas_ha = _sas_ha;
+	struct scsi_core *core = &sas_ha->core;
+
+	daemonize("sas_queue_%d", core->shost->host_no);
+	current->flags |= PF_NOFREEZE;
+
+	complete(&queue_th_comp);
+
+	while (1) {
+		down_interruptible(&core->queue_thread_sema);
+		sas_queue(sas_ha);
+		if (core->queue_thread_kill)
+			break;
+	}
+
+	complete(&queue_th_comp);
+
+	return 0;
+}
+
+int sas_init_queue(struct sas_ha_struct *sas_ha)
+{
+	int res;
+	struct scsi_core *core = &sas_ha->core;
+
+	spin_lock_init(&core->task_queue_lock);
+	core->task_queue_size = 0;
+	INIT_LIST_HEAD(&core->task_queue);
+	init_MUTEX_LOCKED(&core->queue_thread_sema);
+
+	res = kernel_thread(sas_queue_thread, sas_ha, 0);
+	if (res >= 0)
+		wait_for_completion(&queue_th_comp);
+
+	return res < 0 ? res : 0;
+}
+
+void sas_shutdown_queue(struct sas_ha_struct *sas_ha)
+{
+	unsigned long flags;
+	struct scsi_core *core = &sas_ha->core;
+	struct sas_task *task, *n;
+
+	init_completion(&queue_th_comp);
+	core->queue_thread_kill = 1;
+	up(&core->queue_thread_sema);
+	wait_for_completion(&queue_th_comp);
+
+	if (!list_empty(&core->task_queue))
+		SAS_DPRINTK("HA: %llx: scsi core task queue is NOT empty!?\n",
+			    SAS_ADDR(sas_ha->sas_addr));
+
+	spin_lock_irqsave(&core->task_queue_lock, flags);
+	list_for_each_entry_safe(task, n, &core->task_queue, list) {
+		struct scsi_cmnd *cmd = task->uldd_task;
+
+		list_del_init(&task->list);
+
+		ASSIGN_SAS_TASK(cmd, NULL);
+		sas_free_task(task);
+		cmd->result = DID_ABORT << 16;
+		cmd->scsi_done(cmd);
+	}
+	spin_unlock_irqrestore(&core->task_queue_lock, flags);
+}
+
+EXPORT_SYMBOL_GPL(sas_queuecommand);
+EXPORT_SYMBOL_GPL(sas_target_alloc);
+EXPORT_SYMBOL_GPL(sas_slave_configure);
+EXPORT_SYMBOL_GPL(sas_slave_destroy);
+EXPORT_SYMBOL_GPL(sas_change_queue_depth);
+EXPORT_SYMBOL_GPL(sas_change_queue_type);
+EXPORT_SYMBOL_GPL(sas_bios_param);
diff --git a/include/scsi/libsas.h b/include/scsi/libsas.h
new file mode 100644
index 000000000000..72acdabe7f80
--- /dev/null
+++ b/include/scsi/libsas.h
@@ -0,0 +1,627 @@
+/*
+ * SAS host prototypes and structures header file
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
+ * USA
+ *
+ */
+
+#ifndef _LIBSAS_H_
+#define _LIBSAS_H_
+
+
+#include <linux/timer.h>
+#include <linux/pci.h>
+#include <scsi/sas.h>
+#include <linux/list.h>
+#include <asm/semaphore.h>
+#include <scsi/scsi_device.h>
+#include <scsi/scsi_cmnd.h>
+#include <scsi/scsi_transport_sas.h>
+
+struct block_device;
+
+enum sas_class {
+	SAS,
+	EXPANDER
+};
+
+enum sas_phy_role {
+	PHY_ROLE_NONE = 0,
+	PHY_ROLE_TARGET = 0x40,
+	PHY_ROLE_INITIATOR = 0x80,
+};
+
+enum sas_phy_type {
+        PHY_TYPE_PHYSICAL,
+        PHY_TYPE_VIRTUAL
+};
+
+/* The events are mnemonically described in sas_dump.c
+ * so when updating/adding events here, please also
+ * update the other file too.
+ */
+enum ha_event {
+	HAE_RESET             = 0U,
+	HA_NUM_EVENTS         = 1,
+};
+
+enum port_event {
+	PORTE_BYTES_DMAED     = 0U,
+	PORTE_BROADCAST_RCVD  = 1,
+	PORTE_LINK_RESET_ERR  = 2,
+	PORTE_TIMER_EVENT     = 3,
+	PORTE_HARD_RESET      = 4,
+	PORT_NUM_EVENTS       = 5,
+};
+
+enum phy_event {
+	PHYE_LOSS_OF_SIGNAL   = 0U,
+	PHYE_OOB_DONE         = 1,
+	PHYE_OOB_ERROR        = 2,
+	PHYE_SPINUP_HOLD      = 3, /* hot plug SATA, no COMWAKE sent */
+	PHY_NUM_EVENTS        = 4,
+};
+
+enum discover_event {
+	DISCE_DISCOVER_DOMAIN   = 0U,
+	DISCE_REVALIDATE_DOMAIN = 1,
+	DISCE_PORT_GONE         = 2,
+	DISC_NUM_EVENTS 	= 3,
+};
+
+/* ---------- Expander Devices ---------- */
+
+#define ETASK 0xFA
+
+#define to_dom_device(_obj) container_of(_obj, struct domain_device, dev_obj)
+#define to_dev_attr(_attr)  container_of(_attr, struct domain_dev_attribute,\
+                                         attr)
+
+enum routing_attribute {
+	DIRECT_ROUTING,
+	SUBTRACTIVE_ROUTING,
+	TABLE_ROUTING,
+};
+
+enum ex_phy_state {
+	PHY_EMPTY,
+	PHY_VACANT,
+	PHY_NOT_PRESENT,
+	PHY_DEVICE_DISCOVERED
+};
+
+struct ex_phy {
+	int    phy_id;
+
+	enum ex_phy_state phy_state;
+
+	enum sas_dev_type attached_dev_type;
+	enum sas_phy_linkrate linkrate;
+
+	u8   attached_sata_host:1;
+	u8   attached_sata_dev:1;
+	u8   attached_sata_ps:1;
+
+	enum sas_proto attached_tproto;
+	enum sas_proto attached_iproto;
+
+	u8   attached_sas_addr[SAS_ADDR_SIZE];
+	u8   attached_phy_id;
+
+	u8   phy_change_count;
+	enum routing_attribute routing_attr;
+	u8   virtual:1;
+
+	int  last_da_index;
+
+	struct sas_phy *phy;
+	struct sas_port *port;
+};
+
+struct expander_device {
+	struct list_head children;
+
+	u16    ex_change_count;
+	u16    max_route_indexes;
+	u8     num_phys;
+	u8     configuring:1;
+	u8     conf_route_table:1;
+	u8     enclosure_logical_id[8];
+
+	struct ex_phy *ex_phy;
+	struct sas_port *parent_port;
+};
+
+/* ---------- SATA device ---------- */
+enum ata_command_set {
+        ATA_COMMAND_SET   = 0,
+        ATAPI_COMMAND_SET = 1,
+};
+
+struct sata_device {
+        enum   ata_command_set command_set;
+        struct smp_resp        rps_resp; /* report_phy_sata_resp */
+        __le16 *identify_device;
+        __le16 *identify_packet_device;
+
+        u8     port_no;        /* port number, if this is a PM (Port) */
+        struct list_head children; /* PM Ports if this is a PM */
+};
+
+/* ---------- Domain device ---------- */
+struct domain_device {
+        enum sas_dev_type dev_type;
+
+        enum sas_phy_linkrate linkrate;
+        enum sas_phy_linkrate min_linkrate;
+        enum sas_phy_linkrate max_linkrate;
+
+        int  pathways;
+
+        struct domain_device *parent;
+        struct list_head siblings; /* devices on the same level */
+        struct asd_sas_port *port;        /* shortcut to root of the tree */
+
+        struct list_head dev_list_node;
+
+        enum sas_proto    iproto;
+        enum sas_proto    tproto;
+
+        struct sas_rphy *rphy;
+
+        u8  sas_addr[SAS_ADDR_SIZE];
+        u8  hashed_sas_addr[HASHED_SAS_ADDR_SIZE];
+
+        u8  frame_rcvd[32];
+
+        union {
+                struct expander_device ex_dev;
+                struct sata_device     sata_dev; /* STP & directly attached */
+        };
+
+        void *lldd_dev;
+};
+
+struct sas_discovery {
+	spinlock_t disc_event_lock;
+	struct work_struct disc_work[DISC_NUM_EVENTS];
+	unsigned long    pending;
+	u8     fanout_sas_addr[8];
+	u8     eeds_a[8];
+	u8     eeds_b[8];
+	int    max_level;
+};
+
+
+/* The port struct is Class:RW, driver:RO */
+struct asd_sas_port {
+/* private: */
+	struct completion port_gone_completion;
+
+	struct sas_discovery disc;
+	struct domain_device *port_dev;
+	spinlock_t dev_list_lock;
+	struct list_head dev_list;
+	enum   sas_phy_linkrate linkrate;
+
+	struct sas_phy *phy;
+	struct work_struct work;
+
+/* public: */
+	int id;
+
+	enum sas_class   class;
+	u8               sas_addr[SAS_ADDR_SIZE];
+	u8               attached_sas_addr[SAS_ADDR_SIZE];
+	enum sas_proto   iproto;
+	enum sas_proto   tproto;
+
+	enum sas_oob_mode oob_mode;
+
+	spinlock_t       phy_list_lock;
+	struct list_head phy_list;
+	int              num_phys;
+	u32              phy_mask;
+
+	struct sas_ha_struct *ha;
+
+	struct sas_port	*port;
+
+	void *lldd_port;	  /* not touched by the sas class code */
+};
+
+/* The phy pretty much is controlled by the LLDD.
+ * The class only reads those fields.
+ */
+struct asd_sas_phy {
+/* private: */
+	/* protected by ha->event_lock */
+	struct work_struct   port_events[PORT_NUM_EVENTS];
+	struct work_struct   phy_events[PHY_NUM_EVENTS];
+
+	unsigned long port_events_pending;
+	unsigned long phy_events_pending;
+
+	int error;
+
+	struct sas_phy *phy;
+
+/* public: */
+	/* The following are class:RO, driver:R/W */
+	int            enabled;	  /* must be set */
+
+	int            id;	  /* must be set */
+	enum sas_class class;
+	enum sas_proto iproto;
+	enum sas_proto tproto;
+
+	enum sas_phy_type  type;
+	enum sas_phy_role  role;
+	enum sas_oob_mode  oob_mode;
+	enum sas_phy_linkrate linkrate;
+
+	u8   *sas_addr;		  /* must be set */
+	u8   attached_sas_addr[SAS_ADDR_SIZE]; /* class:RO, driver: R/W */
+
+	spinlock_t     frame_rcvd_lock;
+	u8             *frame_rcvd; /* must be set */
+	int            frame_rcvd_size;
+
+	spinlock_t     sas_prim_lock;
+	u32            sas_prim;
+
+	struct list_head port_phy_el; /* driver:RO */
+	struct asd_sas_port      *port; /* Class:RW, driver: RO */
+
+	struct sas_ha_struct *ha; /* may be set; the class sets it anyway */
+
+	void *lldd_phy;		  /* not touched by the sas_class_code */
+};
+
+struct scsi_core {
+	struct Scsi_Host *shost;
+
+	spinlock_t        task_queue_lock;
+	struct list_head  task_queue;
+	int               task_queue_size;
+
+	struct semaphore  queue_thread_sema;
+	int               queue_thread_kill;
+};
+
+struct sas_ha_struct {
+/* private: */
+	spinlock_t       event_lock;
+	struct work_struct ha_events[HA_NUM_EVENTS];
+	unsigned long	 pending;
+
+	struct scsi_core core;
+
+/* public: */
+	char *sas_ha_name;
+	struct pci_dev *pcidev;	  /* should be set */
+	struct module *lldd_module; /* should be set */
+
+	u8 *sas_addr;		  /* must be set */
+	u8 hashed_sas_addr[HASHED_SAS_ADDR_SIZE];
+
+	spinlock_t      phy_port_lock;
+	struct asd_sas_phy  **sas_phy; /* array of valid pointers, must be set */
+	struct asd_sas_port **sas_port; /* array of valid pointers, must be set */
+	int             num_phys; /* must be set, gt 0, static */
+
+	/* The class calls this to send a task for execution. */
+	int lldd_max_execute_num;
+	int lldd_queue_size;
+
+	/* LLDD calls these to notify the class of an event. */
+	void (*notify_ha_event)(struct sas_ha_struct *, enum ha_event);
+	void (*notify_port_event)(struct asd_sas_phy *, enum port_event);
+	void (*notify_phy_event)(struct asd_sas_phy *, enum phy_event);
+
+	void *lldd_ha;		  /* not touched by sas class code */
+};
+
+#define SHOST_TO_SAS_HA(_shost) (*(struct sas_ha_struct **)(_shost)->hostdata)
+
+static inline struct domain_device *
+starget_to_domain_dev(struct scsi_target *starget) {
+	return starget->hostdata;
+}
+
+static inline struct domain_device *
+sdev_to_domain_dev(struct scsi_device *sdev) {
+	return starget_to_domain_dev(sdev->sdev_target);
+}
+
+static inline struct domain_device *
+cmd_to_domain_dev(struct scsi_cmnd *cmd)
+{
+	return sdev_to_domain_dev(cmd->device);
+}
+
+void sas_hash_addr(u8 *hashed, const u8 *sas_addr);
+
+/* Before calling a notify event, LLDD should use this function
+ * when the link is severed (possibly from its tasklet).
+ * The idea is that the Class only reads those, while the LLDD,
+ * can R/W these (thus avoiding a race).
+ */
+static inline void sas_phy_disconnected(struct asd_sas_phy *phy)
+{
+	phy->oob_mode = OOB_NOT_CONNECTED;
+	phy->linkrate = PHY_LINKRATE_NONE;
+}
+
+/* ---------- Tasks ---------- */
+/*
+      service_response |  SAS_TASK_COMPLETE  |  SAS_TASK_UNDELIVERED |
+  exec_status          |                     |                       |
+  ---------------------+---------------------+-----------------------+
+       SAM_...         |         X           |                       |
+       DEV_NO_RESPONSE |         X           |           X           |
+       INTERRUPTED     |         X           |                       |
+       QUEUE_FULL      |                     |           X           |
+       DEVICE_UNKNOWN  |                     |           X           |
+       SG_ERR          |                     |           X           |
+  ---------------------+---------------------+-----------------------+
+ */
+
+enum service_response {
+	SAS_TASK_COMPLETE,
+	SAS_TASK_UNDELIVERED = -1,
+};
+
+enum exec_status {
+	SAM_GOOD         = 0,
+	SAM_CHECK_COND   = 2,
+	SAM_COND_MET     = 4,
+	SAM_BUSY         = 8,
+	SAM_INTERMEDIATE = 0x10,
+	SAM_IM_COND_MET  = 0x12,
+	SAM_RESV_CONFLICT= 0x14,
+	SAM_TASK_SET_FULL= 0x28,
+	SAM_ACA_ACTIVE   = 0x30,
+	SAM_TASK_ABORTED = 0x40,
+
+	SAS_DEV_NO_RESPONSE = 0x80,
+	SAS_DATA_UNDERRUN,
+	SAS_DATA_OVERRUN,
+	SAS_INTERRUPTED,
+	SAS_QUEUE_FULL,
+	SAS_DEVICE_UNKNOWN,
+	SAS_SG_ERR,
+	SAS_OPEN_REJECT,
+	SAS_OPEN_TO,
+	SAS_PROTO_RESPONSE,
+	SAS_PHY_DOWN,
+	SAS_NAK_R_ERR,
+	SAS_PENDING,
+	SAS_ABORTED_TASK,
+};
+
+/* When a task finishes with a response, the LLDD examines the
+ * response:
+ * 	- For an ATA task task_status_struct::stat is set to
+ * SAS_PROTO_RESPONSE, and the task_status_struct::buf is set to the
+ * contents of struct ata_task_resp.
+ * 	- For SSP tasks, if no data is present or status/TMF response
+ * is valid, task_status_struct::stat is set.  If data is present
+ * (SENSE data), the LLDD copies up to SAS_STATUS_BUF_SIZE, sets
+ * task_status_struct::buf_valid_size, and task_status_struct::stat is
+ * set to SAM_CHECK_COND.
+ *
+ * "buf" has format SCSI Sense for SSP task, or struct ata_task_resp
+ * for ATA task.
+ *
+ * "frame_len" is the total frame length, which could be more or less
+ * than actually copied.
+ *
+ * Tasks ending with response, always set the residual field.
+ */
+struct ata_task_resp {
+	u16  frame_len;
+	u8   ending_fis[24];	  /* dev to host or data-in */
+	u32  sstatus;
+	u32  serror;
+	u32  scontrol;
+	u32  sactive;
+};
+
+#define SAS_STATUS_BUF_SIZE 96
+
+struct task_status_struct {
+	enum service_response resp;
+	enum exec_status      stat;
+	int  buf_valid_size;
+
+	u8   buf[SAS_STATUS_BUF_SIZE];
+
+	u32  residual;
+	enum sas_open_rej_reason open_rej_reason;
+};
+
+/* ATA and ATAPI task queuable to a SAS LLDD.
+ */
+struct sas_ata_task {
+	struct host_to_dev_fis fis;
+	u8     atapi_packet[16];  /* 0 if not ATAPI task */
+
+	u8     retry_count;	  /* hardware retry, should be > 0 */
+
+	u8     dma_xfer:1;	  /* PIO:0 or DMA:1 */
+	u8     use_ncq:1;
+	u8     set_affil_pol:1;
+	u8     stp_affil_pol:1;
+
+	u8     device_control_reg_update:1;
+};
+
+struct sas_smp_task {
+	struct scatterlist smp_req;
+	struct scatterlist smp_resp;
+};
+
+enum task_attribute {
+	TASK_ATTR_SIMPLE = 0,
+	TASK_ATTR_HOQ    = 1,
+	TASK_ATTR_ORDERED= 2,
+	TASK_ATTR_ACA    = 4,
+};
+
+struct sas_ssp_task {
+	u8     retry_count;	  /* hardware retry, should be > 0 */
+
+	u8     LUN[8];
+	u8     enable_first_burst:1;
+	enum   task_attribute task_attr;
+	u8     task_prio;
+	u8     cdb[16];
+};
+
+struct sas_task {
+	struct domain_device *dev;
+	struct list_head      list;
+
+	spinlock_t   task_state_lock;
+	unsigned     task_state_flags;
+
+	enum   sas_proto      task_proto;
+
+	/* Used by the discovery code. */
+	struct timer_list     timer;
+	struct completion     completion;
+
+	union {
+		struct sas_ata_task ata_task;
+		struct sas_smp_task smp_task;
+		struct sas_ssp_task ssp_task;
+	};
+
+	struct scatterlist *scatter;
+	int    num_scatter;
+	u32    total_xfer_len;
+	u8     data_dir:2;	  /* Use PCI_DMA_... */
+
+	struct task_status_struct task_status;
+	void   (*task_done)(struct sas_task *);
+
+	void   *lldd_task;	  /* for use by LLDDs */
+	void   *uldd_task;
+};
+
+
+
+#define SAS_TASK_STATE_PENDING  1
+#define SAS_TASK_STATE_DONE     2
+#define SAS_TASK_STATE_ABORTED  4
+
+static inline struct sas_task *sas_alloc_task(unsigned long flags)
+{
+	extern kmem_cache_t *sas_task_cache;
+	struct sas_task *task = kmem_cache_alloc(sas_task_cache, flags);
+
+	if (task) {
+		memset(task, 0, sizeof(*task));
+		INIT_LIST_HEAD(&task->list);
+		spin_lock_init(&task->task_state_lock);
+		task->task_state_flags = SAS_TASK_STATE_PENDING;
+		init_timer(&task->timer);
+		init_completion(&task->completion);
+	}
+
+	return task;
+}
+
+static inline void sas_free_task(struct sas_task *task)
+{
+	if (task) {
+		extern kmem_cache_t *sas_task_cache;
+		BUG_ON(!list_empty(&task->list));
+		kmem_cache_free(sas_task_cache, task);
+	}
+}
+
+struct sas_domain_function_template {
+	/* The class calls these to notify the LLDD of an event. */
+	void (*lldd_port_formed)(struct asd_sas_phy *);
+	void (*lldd_port_deformed)(struct asd_sas_phy *);
+
+	/* The class calls these when a device is found or gone. */
+	int  (*lldd_dev_found)(struct domain_device *);
+	void (*lldd_dev_gone)(struct domain_device *);
+
+	int (*lldd_execute_task)(struct sas_task *, int num,
+				 unsigned long gfp_flags);
+
+	/* Task Management Functions. Must be called from process context. */
+	int (*lldd_abort_task)(struct sas_task *);
+	int (*lldd_abort_task_set)(struct domain_device *, u8 *lun);
+	int (*lldd_clear_aca)(struct domain_device *, u8 *lun);
+	int (*lldd_clear_task_set)(struct domain_device *, u8 *lun);
+	int (*lldd_I_T_nexus_reset)(struct domain_device *);
+	int (*lldd_lu_reset)(struct domain_device *, u8 *lun);
+	int (*lldd_query_task)(struct sas_task *);
+
+	/* Port and Adapter management */
+	int (*lldd_clear_nexus_port)(struct asd_sas_port *);
+	int (*lldd_clear_nexus_ha)(struct sas_ha_struct *);
+
+	/* Phy management */
+	int (*lldd_control_phy)(struct asd_sas_phy *, enum phy_func);
+};
+
+extern int sas_register_ha(struct sas_ha_struct *);
+extern int sas_unregister_ha(struct sas_ha_struct *);
+
+extern int sas_queuecommand(struct scsi_cmnd *,
+		     void (*scsi_done)(struct scsi_cmnd *));
+extern int sas_target_alloc(struct scsi_target *);
+extern int sas_slave_alloc(struct scsi_device *);
+extern int sas_slave_configure(struct scsi_device *);
+extern void sas_slave_destroy(struct scsi_device *);
+extern int sas_change_queue_depth(struct scsi_device *, int new_depth);
+extern int sas_change_queue_type(struct scsi_device *, int qt);
+extern int sas_bios_param(struct scsi_device *,
+			  struct block_device *,
+			  sector_t capacity, int *hsc);
+extern struct scsi_transport_template *
+sas_domain_attach_transport(struct sas_domain_function_template *);
+extern void sas_domain_release_transport(struct scsi_transport_template *);
+
+int  sas_discover_root_expander(struct domain_device *);
+
+void sas_init_ex_attr(void);
+
+int  sas_ex_revalidate_domain(struct domain_device *);
+
+void sas_unregister_domain_devices(struct asd_sas_port *port);
+void sas_init_disc(struct sas_discovery *disc, struct asd_sas_port *);
+int  sas_discover_event(struct asd_sas_port *, enum discover_event ev);
+
+int  sas_discover_sata(struct domain_device *);
+int  sas_discover_end_dev(struct domain_device *);
+
+void sas_unregister_dev(struct domain_device *);
+
+void sas_init_dev(struct domain_device *);
+
+#endif /* _SASLIB_H_ */
diff --git a/include/scsi/sas.h b/include/scsi/sas.h
new file mode 100644
index 000000000000..752853a113dc
--- /dev/null
+++ b/include/scsi/sas.h
@@ -0,0 +1,644 @@
+/*
+ * SAS structures and definitions header file
+ *
+ * Copyright (C) 2005 Adaptec, Inc.  All rights reserved.
+ * Copyright (C) 2005 Luben Tuikov <luben_tuikov@adaptec.com>
+ *
+ * This file is licensed under GPLv2.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
+ * USA
+ *
+ */
+
+#ifndef _SAS_H_
+#define _SAS_H_
+
+#include <linux/types.h>
+#include <asm/byteorder.h>
+
+#define SAS_ADDR_SIZE        8
+#define HASHED_SAS_ADDR_SIZE 3
+#define SAS_ADDR(_sa) ((unsigned long long) be64_to_cpu(*(__be64 *)(_sa)))
+
+#define SMP_REQUEST             0x40
+#define SMP_RESPONSE            0x41
+
+#define SSP_DATA                0x01
+#define SSP_XFER_RDY            0x05
+#define SSP_COMMAND             0x06
+#define SSP_RESPONSE            0x07
+#define SSP_TASK                0x16
+
+#define SMP_REPORT_GENERAL       0x00
+#define SMP_REPORT_MANUF_INFO    0x01
+#define SMP_READ_GPIO_REG        0x02
+#define SMP_DISCOVER             0x10
+#define SMP_REPORT_PHY_ERR_LOG   0x11
+#define SMP_REPORT_PHY_SATA      0x12
+#define SMP_REPORT_ROUTE_INFO    0x13
+#define SMP_WRITE_GPIO_REG       0x82
+#define SMP_CONF_ROUTE_INFO      0x90
+#define SMP_PHY_CONTROL          0x91
+#define SMP_PHY_TEST_FUNCTION    0x92
+
+#define SMP_RESP_FUNC_ACC        0x00
+#define SMP_RESP_FUNC_UNK        0x01
+#define SMP_RESP_FUNC_FAILED     0x02
+#define SMP_RESP_INV_FRM_LEN     0x03
+#define SMP_RESP_NO_PHY          0x10
+#define SMP_RESP_NO_INDEX        0x11
+#define SMP_RESP_PHY_NO_SATA     0x12
+#define SMP_RESP_PHY_UNK_OP      0x13
+#define SMP_RESP_PHY_UNK_TESTF   0x14
+#define SMP_RESP_PHY_TEST_INPROG 0x15
+#define SMP_RESP_PHY_VACANT      0x16
+
+/* SAM TMFs */
+#define TMF_ABORT_TASK      0x01
+#define TMF_ABORT_TASK_SET  0x02
+#define TMF_CLEAR_TASK_SET  0x04
+#define TMF_LU_RESET        0x08
+#define TMF_CLEAR_ACA       0x40
+#define TMF_QUERY_TASK      0x80
+
+/* SAS TMF responses */
+#define TMF_RESP_FUNC_COMPLETE   0x00
+#define TMF_RESP_INVALID_FRAME   0x02
+#define TMF_RESP_FUNC_ESUPP      0x04
+#define TMF_RESP_FUNC_FAILED     0x05
+#define TMF_RESP_FUNC_SUCC       0x08
+#define TMF_RESP_NO_LUN          0x09
+#define TMF_RESP_OVERLAPPED_TAG  0x0A
+
+enum sas_oob_mode {
+	OOB_NOT_CONNECTED,
+	SATA_OOB_MODE,
+	SAS_OOB_MODE
+};
+
+/* See sas_discover.c if you plan on changing these.
+ */
+enum sas_dev_type {
+	NO_DEVICE   = 0,	  /* protocol */
+	SAS_END_DEV = 1,	  /* protocol */
+	EDGE_DEV    = 2,	  /* protocol */
+	FANOUT_DEV  = 3,	  /* protocol */
+	SAS_HA      = 4,
+	SATA_DEV    = 5,
+	SATA_PM     = 7,
+	SATA_PM_PORT= 8,
+};
+
+enum sas_phy_linkrate {
+	PHY_LINKRATE_NONE = 0,
+	PHY_LINKRATE_UNKNOWN = 0,
+	PHY_DISABLED,
+	PHY_RESET_PROBLEM,
+	PHY_SPINUP_HOLD,
+	PHY_PORT_SELECTOR,
+	PHY_LINKRATE_1_5 = 0x08,
+	PHY_LINKRATE_G1  = PHY_LINKRATE_1_5,
+	PHY_LINKRATE_3   = 0x09,
+	PHY_LINKRATE_G2  = PHY_LINKRATE_3,
+	PHY_LINKRATE_6   = 0x0A,
+};
+
+/* Partly from IDENTIFY address frame. */
+enum sas_proto {
+	SATA_PROTO    = 1,
+	SAS_PROTO_SMP = 2,	  /* protocol */
+	SAS_PROTO_STP = 4,	  /* protocol */
+	SAS_PROTO_SSP = 8,	  /* protocol */
+	SAS_PROTO_ALL = 0xE,
+};
+
+/* From the spec; local phys only */
+enum phy_func {
+	PHY_FUNC_NOP,
+	PHY_FUNC_LINK_RESET,		  /* Enables the phy */
+	PHY_FUNC_HARD_RESET,
+	PHY_FUNC_DISABLE,
+	PHY_FUNC_CLEAR_ERROR_LOG = 5,
+	PHY_FUNC_CLEAR_AFFIL,
+	PHY_FUNC_TX_SATA_PS_SIGNAL,
+	PHY_FUNC_RELEASE_SPINUP_HOLD = 0x10, /* LOCAL PORT ONLY! */
+};
+
+/* SAS LLDD would need to report only _very_few_ of those, like BROADCAST.
+ * Most of those are here for completeness.
+ */
+enum sas_prim {
+	SAS_PRIM_AIP_NORMAL = 1,
+	SAS_PRIM_AIP_R0     = 2,
+	SAS_PRIM_AIP_R1     = 3,
+	SAS_PRIM_AIP_R2     = 4,
+	SAS_PRIM_AIP_WC     = 5,
+	SAS_PRIM_AIP_WD     = 6,
+	SAS_PRIM_AIP_WP     = 7,
+	SAS_PRIM_AIP_RWP    = 8,
+
+	SAS_PRIM_BC_CH      = 9,
+	SAS_PRIM_BC_RCH0    = 10,
+	SAS_PRIM_BC_RCH1    = 11,
+	SAS_PRIM_BC_R0      = 12,
+	SAS_PRIM_BC_R1      = 13,
+	SAS_PRIM_BC_R2      = 14,
+	SAS_PRIM_BC_R3      = 15,
+	SAS_PRIM_BC_R4      = 16,
+
+	SAS_PRIM_NOTIFY_ENSP= 17,
+	SAS_PRIM_NOTIFY_R0  = 18,
+	SAS_PRIM_NOTIFY_R1  = 19,
+	SAS_PRIM_NOTIFY_R2  = 20,
+
+	SAS_PRIM_CLOSE_CLAF = 21,
+	SAS_PRIM_CLOSE_NORM = 22,
+	SAS_PRIM_CLOSE_R0   = 23,
+	SAS_PRIM_CLOSE_R1   = 24,
+
+	SAS_PRIM_OPEN_RTRY  = 25,
+	SAS_PRIM_OPEN_RJCT  = 26,
+	SAS_PRIM_OPEN_ACPT  = 27,
+
+	SAS_PRIM_DONE       = 28,
+	SAS_PRIM_BREAK      = 29,
+
+	SATA_PRIM_DMAT      = 33,
+	SATA_PRIM_PMNAK     = 34,
+	SATA_PRIM_PMACK     = 35,
+	SATA_PRIM_PMREQ_S   = 36,
+	SATA_PRIM_PMREQ_P   = 37,
+	SATA_SATA_R_ERR     = 38,
+};
+
+enum sas_open_rej_reason {
+	/* Abandon open */
+	SAS_OREJ_UNKNOWN   = 0,
+	SAS_OREJ_BAD_DEST  = 1,
+	SAS_OREJ_CONN_RATE = 2,
+	SAS_OREJ_EPROTO    = 3,
+	SAS_OREJ_RESV_AB0  = 4,
+	SAS_OREJ_RESV_AB1  = 5,
+	SAS_OREJ_RESV_AB2  = 6,
+	SAS_OREJ_RESV_AB3  = 7,
+	SAS_OREJ_WRONG_DEST= 8,
+	SAS_OREJ_STP_NORES = 9,
+
+	/* Retry open */
+	SAS_OREJ_NO_DEST   = 10,
+	SAS_OREJ_PATH_BLOCKED = 11,
+	SAS_OREJ_RSVD_CONT0 = 12,
+	SAS_OREJ_RSVD_CONT1 = 13,
+	SAS_OREJ_RSVD_INIT0 = 14,
+	SAS_OREJ_RSVD_INIT1 = 15,
+	SAS_OREJ_RSVD_STOP0 = 16,
+	SAS_OREJ_RSVD_STOP1 = 17,
+	SAS_OREJ_RSVD_RETRY = 18,
+};
+
+struct  dev_to_host_fis {
+	u8     fis_type;	  /* 0x34 */
+	u8     flags;
+	u8     status;
+	u8     error;
+
+	u8     lbal;
+	union { u8 lbam; u8 byte_count_low; };
+	union { u8 lbah; u8 byte_count_high; };
+	u8     device;
+
+	u8     lbal_exp;
+	u8     lbam_exp;
+	u8     lbah_exp;
+	u8     _r_a;
+
+	union { u8  sector_count; u8 interrupt_reason; };
+	u8     sector_count_exp;
+	u8     _r_b;
+	u8     _r_c;
+
+	u32    _r_d;
+} __attribute__ ((packed));
+
+struct host_to_dev_fis {
+	u8     fis_type;	  /* 0x27 */
+	u8     flags;
+	u8     command;
+	u8     features;
+
+	u8     lbal;
+	union { u8 lbam; u8 byte_count_low; };
+	union { u8 lbah; u8 byte_count_high; };
+	u8     device;
+
+	u8     lbal_exp;
+	u8     lbam_exp;
+	u8     lbah_exp;
+	u8     features_exp;
+
+	union { u8  sector_count; u8 interrupt_reason; };
+	u8     sector_count_exp;
+	u8     _r_a;
+	u8     control;
+
+	u32    _r_b;
+} __attribute__ ((packed));
+
+/* Prefer to have code clarity over header file clarity.
+ */
+#ifdef __LITTLE_ENDIAN_BITFIELD
+struct sas_identify_frame {
+	/* Byte 0 */
+	u8  frame_type:4;
+	u8  dev_type:3;
+	u8  _un0:1;
+
+	/* Byte 1 */
+	u8  _un1;
+
+	/* Byte 2 */
+	union {
+		struct {
+			u8  _un20:1;
+			u8  smp_iport:1;
+			u8  stp_iport:1;
+			u8  ssp_iport:1;
+			u8  _un247:4;
+		};
+		u8 initiator_bits;
+	};
+
+	/* Byte 3 */
+	union {
+		struct {
+			u8  _un30:1;
+			u8 smp_tport:1;
+			u8 stp_tport:1;
+			u8 ssp_tport:1;
+			u8 _un347:4;
+		};
+		u8 target_bits;
+	};
+
+	/* Byte 4 - 11 */
+	u8 _un4_11[8];
+
+	/* Byte 12 - 19 */
+	u8 sas_addr[SAS_ADDR_SIZE];
+
+	/* Byte 20 */
+	u8 phy_id;
+
+	u8 _un21_27[7];
+
+	__be32 crc;
+} __attribute__ ((packed));
+
+struct ssp_frame_hdr {
+	u8     frame_type;
+	u8     hashed_dest_addr[HASHED_SAS_ADDR_SIZE];
+	u8     _r_a;
+	u8     hashed_src_addr[HASHED_SAS_ADDR_SIZE];
+	__be16 _r_b;
+
+	u8     changing_data_ptr:1;
+	u8     retransmit:1;
+	u8     retry_data_frames:1;
+	u8     _r_c:5;
+
+	u8     num_fill_bytes:2;
+	u8     _r_d:6;
+
+	u32    _r_e;
+	__be16 tag;
+	__be16 tptt;
+	__be32 data_offs;
+} __attribute__ ((packed));
+
+struct ssp_response_iu {
+	u8     _r_a[10];
+
+	u8     datapres:2;
+	u8     _r_b:6;
+
+	u8     status;
+
+	u32    _r_c;
+
+	__be32 sense_data_len;
+	__be32 response_data_len;
+
+	u8     resp_data[0];
+	u8     sense_data[0];
+} __attribute__ ((packed));
+
+/* ---------- SMP ---------- */
+
+struct report_general_resp {
+	__be16  change_count;
+	__be16  route_indexes;
+	u8      _r_a;
+	u8      num_phys;
+
+	u8      conf_route_table:1;
+	u8      configuring:1;
+	u8      _r_b:6;
+
+	u8      _r_c;
+
+	u8      enclosure_logical_id[8];
+
+	u8      _r_d[12];
+} __attribute__ ((packed));
+
+struct discover_resp {
+	u8    _r_a[5];
+
+	u8    phy_id;
+	__be16 _r_b;
+
+	u8    _r_c:4;
+	u8    attached_dev_type:3;
+	u8    _r_d:1;
+
+	u8    linkrate:4;
+	u8    _r_e:4;
+
+	u8    attached_sata_host:1;
+	u8    iproto:3;
+	u8    _r_f:4;
+
+	u8    attached_sata_dev:1;
+	u8    tproto:3;
+	u8    _r_g:3;
+	u8    attached_sata_ps:1;
+
+	u8    sas_addr[8];
+	u8    attached_sas_addr[8];
+	u8    attached_phy_id;
+
+	u8    _r_h[7];
+
+	u8    hmin_linkrate:4;
+	u8    pmin_linkrate:4;
+	u8    hmax_linkrate:4;
+	u8    pmax_linkrate:4;
+
+	u8    change_count;
+
+	u8    pptv:4;
+	u8    _r_i:3;
+	u8    virtual:1;
+
+	u8    routing_attr:4;
+	u8    _r_j:4;
+
+	u8    conn_type;
+	u8    conn_el_index;
+	u8    conn_phy_link;
+
+	u8    _r_k[8];
+} __attribute__ ((packed));
+
+struct report_phy_sata_resp {
+	u8    _r_a[5];
+
+	u8    phy_id;
+	u8    _r_b;
+
+	u8    affil_valid:1;
+	u8    affil_supp:1;
+	u8    _r_c:6;
+
+	u32    _r_d;
+
+	u8    stp_sas_addr[8];
+
+	struct dev_to_host_fis fis;
+
+	u32   _r_e;
+
+	u8    affil_stp_ini_addr[8];
+
+	__be32 crc;
+} __attribute__ ((packed));
+
+struct smp_resp {
+	u8    frame_type;
+	u8    function;
+	u8    result;
+	u8    reserved;
+	union {
+		struct report_general_resp  rg;
+		struct discover_resp        disc;
+		struct report_phy_sata_resp rps;
+	};
+} __attribute__ ((packed));
+
+#elif defined(__BIG_ENDIAN_BITFIELD)
+struct sas_identify_frame {
+	/* Byte 0 */
+	u8  _un0:1;
+	u8  dev_type:3;
+	u8  frame_type:4;
+
+	/* Byte 1 */
+	u8  _un1;
+
+	/* Byte 2 */
+	union {
+		struct {
+			u8  _un247:4;
+			u8  ssp_iport:1;
+			u8  stp_iport:1;
+			u8  smp_iport:1;
+			u8  _un20:1;
+		};
+		u8 initiator_bits;
+	};
+
+	/* Byte 3 */
+	union {
+		struct {
+			u8 _un347:4;
+			u8 ssp_tport:1;
+			u8 stp_tport:1;
+			u8 smp_tport:1;
+			u8 _un30:1;
+		};
+		u8 target_bits;
+	};
+
+	/* Byte 4 - 11 */
+	u8 _un4_11[8];
+
+	/* Byte 12 - 19 */
+	u8 sas_addr[SAS_ADDR_SIZE];
+
+	/* Byte 20 */
+	u8 phy_id;
+
+	u8 _un21_27[7];
+
+	__be32 crc;
+} __attribute__ ((packed));
+
+struct ssp_frame_hdr {
+	u8     frame_type;
+	u8     hashed_dest_addr[HASHED_SAS_ADDR_SIZE];
+	u8     _r_a;
+	u8     hashed_src_addr[HASHED_SAS_ADDR_SIZE];
+	__be16 _r_b;
+
+	u8     _r_c:5;
+	u8     retry_data_frames:1;
+	u8     retransmit:1;
+	u8     changing_data_ptr:1;
+
+	u8     _r_d:6;
+	u8     num_fill_bytes:2;
+
+	u32    _r_e;
+	__be16 tag;
+	__be16 tptt;
+	__be32 data_offs;
+} __attribute__ ((packed));
+
+struct ssp_response_iu {
+	u8     _r_a[10];
+
+	u8     _r_b:6;
+	u8     datapres:2;
+
+	u8     status;
+
+	u32    _r_c;
+
+	__be32 sense_data_len;
+	__be32 response_data_len;
+
+	u8     resp_data[0];
+	u8     sense_data[0];
+} __attribute__ ((packed));
+
+/* ---------- SMP ---------- */
+
+struct report_general_resp {
+	__be16  change_count;
+	__be16  route_indexes;
+	u8      _r_a;
+	u8      num_phys;
+
+	u8      _r_b:6;
+	u8      configuring:1;
+	u8      conf_route_table:1;
+
+	u8      _r_c;
+
+	u8      enclosure_logical_id[8];
+
+	u8      _r_d[12];
+} __attribute__ ((packed));
+
+struct discover_resp {
+	u8    _r_a[5];
+
+	u8    phy_id;
+	__be16 _r_b;
+
+	u8    _r_d:1;
+	u8    attached_dev_type:3;
+	u8    _r_c:4;
+
+	u8    _r_e:4;
+	u8    linkrate:4;
+
+	u8    _r_f:4;
+	u8    iproto:3;
+	u8    attached_sata_host:1;
+
+	u8    attached_sata_ps:1;
+	u8    _r_g:3;
+	u8    tproto:3;
+	u8    attached_sata_dev:1;
+
+	u8    sas_addr[8];
+	u8    attached_sas_addr[8];
+	u8    attached_phy_id;
+
+	u8    _r_h[7];
+
+	u8    pmin_linkrate:4;
+	u8    hmin_linkrate:4;
+	u8    pmax_linkrate:4;
+	u8    hmax_linkrate:4;
+
+	u8    change_count;
+
+	u8    virtual:1;
+	u8    _r_i:3;
+	u8    pptv:4;
+
+	u8    _r_j:4;
+	u8    routing_attr:4;
+
+	u8    conn_type;
+	u8    conn_el_index;
+	u8    conn_phy_link;
+
+	u8    _r_k[8];
+} __attribute__ ((packed));
+
+struct report_phy_sata_resp {
+	u8    _r_a[5];
+
+	u8    phy_id;
+	u8    _r_b;
+
+	u8    _r_c:6;
+	u8    affil_supp:1;
+	u8    affil_valid:1;
+
+	u32   _r_d;
+
+	u8    stp_sas_addr[8];
+
+	struct dev_to_host_fis fis;
+
+	u32   _r_e;
+
+	u8    affil_stp_ini_addr[8];
+
+	__be32 crc;
+} __attribute__ ((packed));
+
+struct smp_resp {
+	u8    frame_type;
+	u8    function;
+	u8    result;
+	u8    reserved;
+	union {
+		struct report_general_resp  rg;
+		struct discover_resp        disc;
+		struct report_phy_sata_resp rps;
+	};
+} __attribute__ ((packed));
+
+#else
+#error "Bitfield order not defined!"
+#endif
+
+#endif /* _SAS_H_ */
diff --git a/include/scsi/scsi.h b/include/scsi/scsi.h
index 1bc675201413..84a6d5fe0920 100644
--- a/include/scsi/scsi.h
+++ b/include/scsi/scsi.h
@@ -429,4 +429,10 @@ struct scsi_lun {
 /* Used to obtain the PCI location of a device */
 #define SCSI_IOCTL_GET_PCI		0x5387
 
+/* Pull a u32 out of a SCSI message (using BE SCSI conventions) */
+static inline u32 scsi_to_u32(u8 *ptr)
+{
+	return (ptr[0]<<24) + (ptr[1]<<16) + (ptr[2]<<8) + ptr[3];
+}
+
 #endif /* _SCSI_SCSI_H */
-- 
cgit v1.2.3


From 669a5db411d85a14f86cd92bc16bf7ab5b8aa235 Mon Sep 17 00:00:00 2001
From: Jeff Garzik <jeff@garzik.org>
Date: Tue, 29 Aug 2006 18:12:40 -0400
Subject: [libata] Add a bunch of PATA drivers.

The vast majority of drivers and changes are from Alan Cox.  Albert Lee
contributed and maintains pata_pdc2027x.  Adrian Bunk, Andrew Morton,
and Tejun Heo contributed various minor fixes and updates.

Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/ata/Kconfig             |  334 +++++++++++
 drivers/ata/Makefile            |   41 ++
 drivers/ata/ata_generic.c       |  252 ++++++++
 drivers/ata/ata_piix.c          |  436 ++++++++++----
 drivers/ata/pata_ali.c          |  679 +++++++++++++++++++++
 drivers/ata/pata_amd.c          |  707 ++++++++++++++++++++++
 drivers/ata/pata_artop.c        |  518 ++++++++++++++++
 drivers/ata/pata_atiixp.c       |  306 ++++++++++
 drivers/ata/pata_cmd64x.c       |  505 ++++++++++++++++
 drivers/ata/pata_cs5520.c       |  336 +++++++++++
 drivers/ata/pata_cs5530.c       |  387 ++++++++++++
 drivers/ata/pata_cs5535.c       |  291 +++++++++
 drivers/ata/pata_cypress.c      |  227 +++++++
 drivers/ata/pata_efar.c         |  342 +++++++++++
 drivers/ata/pata_hpt366.c       |  478 +++++++++++++++
 drivers/ata/pata_hpt37x.c       | 1257 +++++++++++++++++++++++++++++++++++++++
 drivers/ata/pata_hpt3x2n.c      |  597 +++++++++++++++++++
 drivers/ata/pata_hpt3x3.c       |  226 +++++++
 drivers/ata/pata_isapnp.c       |  156 +++++
 drivers/ata/pata_it8172.c       |  288 +++++++++
 drivers/ata/pata_it821x.c       |  847 ++++++++++++++++++++++++++
 drivers/ata/pata_jmicron.c      |  266 +++++++++
 drivers/ata/pata_legacy.c       |  949 +++++++++++++++++++++++++++++
 drivers/ata/pata_mpiix.c        |  313 ++++++++++
 drivers/ata/pata_netcell.c      |  175 ++++++
 drivers/ata/pata_ns87410.c      |  236 ++++++++
 drivers/ata/pata_oldpiix.c      |  339 +++++++++++
 drivers/ata/pata_opti.c         |  292 +++++++++
 drivers/ata/pata_optidma.c      |  547 +++++++++++++++++
 drivers/ata/pata_pcmcia.c       |  393 ++++++++++++
 drivers/ata/pata_pdc2027x.c     |  869 +++++++++++++++++++++++++++
 drivers/ata/pata_pdc202xx_old.c |  423 +++++++++++++
 drivers/ata/pata_qdi.c          |  403 +++++++++++++
 drivers/ata/pata_radisys.c      |  335 +++++++++++
 drivers/ata/pata_rz1000.c       |  205 +++++++
 drivers/ata/pata_sc1200.c       |  287 +++++++++
 drivers/ata/pata_serverworks.c  |  587 ++++++++++++++++++
 drivers/ata/pata_sil680.c       |  381 ++++++++++++
 drivers/ata/pata_sis.c          | 1030 ++++++++++++++++++++++++++++++++
 drivers/ata/pata_sl82c105.c     |  388 ++++++++++++
 drivers/ata/pata_triflex.c      |  285 +++++++++
 drivers/ata/pata_via.c          |  568 ++++++++++++++++++
 include/linux/libata.h          |    2 +-
 43 files changed, 18372 insertions(+), 111 deletions(-)
 create mode 100644 drivers/ata/ata_generic.c
 create mode 100644 drivers/ata/pata_ali.c
 create mode 100644 drivers/ata/pata_amd.c
 create mode 100644 drivers/ata/pata_artop.c
 create mode 100644 drivers/ata/pata_atiixp.c
 create mode 100644 drivers/ata/pata_cmd64x.c
 create mode 100644 drivers/ata/pata_cs5520.c
 create mode 100644 drivers/ata/pata_cs5530.c
 create mode 100644 drivers/ata/pata_cs5535.c
 create mode 100644 drivers/ata/pata_cypress.c
 create mode 100644 drivers/ata/pata_efar.c
 create mode 100644 drivers/ata/pata_hpt366.c
 create mode 100644 drivers/ata/pata_hpt37x.c
 create mode 100644 drivers/ata/pata_hpt3x2n.c
 create mode 100644 drivers/ata/pata_hpt3x3.c
 create mode 100644 drivers/ata/pata_isapnp.c
 create mode 100644 drivers/ata/pata_it8172.c
 create mode 100644 drivers/ata/pata_it821x.c
 create mode 100644 drivers/ata/pata_jmicron.c
 create mode 100644 drivers/ata/pata_legacy.c
 create mode 100644 drivers/ata/pata_mpiix.c
 create mode 100644 drivers/ata/pata_netcell.c
 create mode 100644 drivers/ata/pata_ns87410.c
 create mode 100644 drivers/ata/pata_oldpiix.c
 create mode 100644 drivers/ata/pata_opti.c
 create mode 100644 drivers/ata/pata_optidma.c
 create mode 100644 drivers/ata/pata_pcmcia.c
 create mode 100644 drivers/ata/pata_pdc2027x.c
 create mode 100644 drivers/ata/pata_pdc202xx_old.c
 create mode 100644 drivers/ata/pata_qdi.c
 create mode 100644 drivers/ata/pata_radisys.c
 create mode 100644 drivers/ata/pata_rz1000.c
 create mode 100644 drivers/ata/pata_sc1200.c
 create mode 100644 drivers/ata/pata_serverworks.c
 create mode 100644 drivers/ata/pata_sil680.c
 create mode 100644 drivers/ata/pata_sis.c
 create mode 100644 drivers/ata/pata_sl82c105.c
 create mode 100644 drivers/ata/pata_triflex.c
 create mode 100644 drivers/ata/pata_via.c

(limited to 'include')

diff --git a/drivers/ata/Kconfig b/drivers/ata/Kconfig
index 13027d56b7f6..cbda988692e7 100644
--- a/drivers/ata/Kconfig
+++ b/drivers/ata/Kconfig
@@ -145,6 +145,340 @@ config SATA_INTEL_COMBINED
 	depends on IDE=y && !BLK_DEV_IDE_SATA && (SATA_AHCI || ATA_PIIX)
 	default y
 
+config PATA_ALI
+	tristate "ALi PATA support (Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for the ALi ATA interfaces
+	  found on the many ALi chipsets.
+
+	  If unsure, say N.
+
+config PATA_AMD
+	tristate "AMD/NVidia PATA support (Experimental)"
+	depends on PCI
+	help
+	  This option enables support for the AMD and NVidia PATA
+	  interfaces found on the chipsets for Athlon/Athlon64.
+
+	  If unsure, say N.
+
+config PATA_ARTOP
+	tristate "ARTOP 6210/6260 PATA support (Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for ARTOP PATA controllers.
+
+	  If unsure, say N.
+
+config PATA_ATIIXP
+	tristate "ATI PATA support (Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for the ATI ATA interfaces
+	  found on the many ATI chipsets.
+
+	  If unsure, say N.
+
+config PATA_CMD64X
+	tristate "CMD64x PATA support (Very Experimental)"
+	depends on PCI&& EXPERIMENTAL
+	help
+	  This option enables support for the CMD64x series chips
+	  except for the CMD640.
+
+	  If unsure, say N.
+
+config PATA_CS5520
+	tristate "CS5510/5520 PATA support"
+	depends on PCI
+	help
+	  This option enables support for the Cyrix 5510/5520
+	  companion chip used with the MediaGX/Geode processor family.
+
+	  If unsure, say N.
+
+config PATA_CS5530
+	tristate "CS5530 PATA support (Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for the Cyrix/NatSemi/AMD CS5530
+	  companion chip used with the MediaGX/Geode processor family.
+
+	  If unsure, say N.
+
+config PATA_CS5535
+	tristate "CS5535 PATA support (Experimental)"
+	depends on PCI && X86 && !X86_64 && EXPERIMENTAL
+	help
+	  This option enables support for the NatSemi/AMD CS5535
+	  companion chip used with the Geode processor family.
+
+	  If unsure, say N.
+
+config PATA_CYPRESS
+	tristate "Cypress CY82C693 PATA support (Very Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for the Cypress/Contaq CY82C693
+	  chipset found in some Alpha systems
+
+	  If unsure, say N.
+
+config PATA_EFAR
+	tristate "EFAR SLC90E66 support"
+	depends on PCI
+	help
+	  This option enables support for the EFAR SLC90E66
+	  IDE controller found on some older machines.
+
+	  If unsure, say N.
+
+config ATA_GENERIC
+	tristate "Generic ATA support"
+	depends on PCI
+	help
+	  This option enables support for generic BIOS configured
+	  ATA controllers via the new ATA layer
+
+	  If unsure, say N.
+
+config PATA_HPT366
+	tristate "HPT 366/368 PATA support (Very Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for the HPT 366 and 368
+	  PATA controllers via the new ATA layer.
+
+	  If unsure, say N.
+
+config PATA_HPT37X
+	tristate "HPT 370/370A/371/372/374/302 PATA support (Very Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for the majority of the later HPT
+	  PATA controllers via the new ATA layer.
+
+	  If unsure, say N.
+
+config PATA_HPT3X2N
+	tristate "HPT 372N/302N PATA support (Very Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for the N variant HPT PATA
+	  controllers via the new ATA layer
+
+	  If unsure, say N.
+
+config PATA_HPT3X3
+	tristate "HPT 343/363 PATA support (Experimental)"
+	depends on PCI
+	help
+	  This option enables support for the HPT 343/363
+	  PATA controllers via the new ATA layer
+
+	  If unsure, say N.
+
+config PATA_ISAPNP
+	tristate "ISA Plug and Play PATA support (Very Experimental)"
+	depends on EXPERIMENTAL && ISAPNP
+	help
+	  This option enables support for ISA plug & play ATA
+	  controllers such as those found on old soundcards.
+
+	  If unsure, say N.
+
+config PATA_IT8172
+	tristate "IT8172 PATA support (Very Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for the ITE 8172 PATA controller
+	  via the new ATA layer.
+
+	  If unsure, say N.
+
+config PATA_IT821X
+	tristate "IT821x PATA support (Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for the ITE 8211 and 8212
+	  PATA controllers via the new ATA layer, including RAID
+	  mode.
+
+	  If unsure, say N.
+
+config PATA_LEGACY
+	tristate "Legacy ISA PATA support (Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for ISA/VLB bus legacy PATA
+	  ports and allows them to be accessed via the new ATA layer.
+
+	  If unsure, say N.
+
+config PATA_TRIFLEX
+	tristate "Compaq Triflex PATA support"
+	depends on PCI
+	help
+	  Enable support for the Compaq 'Triflex' IDE controller as found
+	  on many Compaq Pentium-Pro systems, via the new ATA layer.
+
+	  If unsure, say N.
+
+config PATA_MPIIX
+	tristate "Intel PATA MPIIX support"
+	depends on PCI
+	help
+	  This option enables support for MPIIX PATA support.
+
+	  If unsure, say N.
+
+config PATA_OLDPIIX
+	tristate "Intel PATA old PIIX support (Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for old(?) PIIX PATA support.
+
+	  If unsure, say N.
+
+config PATA_NETCELL
+	tristate "NETCELL Revolution RAID support"
+	depends on PCI
+	help
+	  This option enables support for the Netcell Revolution RAID
+	  PATA controller.
+
+	  If unsure, say N.
+
+config PATA_NS87410
+	tristate "Nat Semi NS87410 PATA support (Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for the National Semiconductor
+	  NS87410 PCI-IDE controller.
+
+	  If unsure, say N.
+
+config PATA_OPTI
+	tristate "OPTI621/6215 PATA support (Very Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables full PIO support for the early Opti ATA
+	  controllers found on some old motherboards.
+
+	  If unsure, say N.
+
+config PATA_OPTIDMA
+	tristate "OPTI FireStar PATA support (Veyr Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables DMA/PIO support for the later OPTi
+	  controllers found on some old motherboards and in some
+	  latops
+
+	  If unsure, say N.
+
+config PATA_PCMCIA
+	tristate "PCMCIA PATA support"
+	depends on PCMCIA
+	help
+	  This option enables support for PCMCIA ATA interfaces, including
+	  compact flash card adapters via the new ATA layer.
+
+	  If unsure, say N.
+
+config PATA_PDC_OLD
+	tristate "Older Promise PATA controller support (Very Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for the Promise 20246, 20262, 20263,
+	  20265 and 20267 adapters.
+
+	  If unsure, say N.
+
+config PATA_QDI
+	tristate "QDI VLB PATA support"
+	help
+	  Support for QDI 6500 and 6580 PATA controllers on VESA local bus.
+
+config PATA_RADISYS
+	tristate "RADISYS 82600 PATA support (Very experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for the RADISYS 82600
+	  PATA controllers via the new ATA layer
+
+	  If unsure, say N.
+
+config PATA_RZ1000
+	tristate "PC Tech RZ1000 PATA support"
+	depends on PCI
+	help
+	  This option enables basic support for the PC Tech RZ1000/1
+	  PATA controllers via the new ATA layer
+
+	  If unsure, say N.
+
+config PATA_SC1200
+	tristate "SC1200 PATA support (Raving Lunatic)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for the NatSemi/AMD SC1200 SoC
+	  companion chip used with the Geode processor family.
+
+	  If unsure, say N.
+
+config PATA_SERVERWORKS
+	tristate "SERVERWORKS OSB4/CSB5/CSB6/HT1000 PATA support (Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for the Serverworks OSB4/CSB5/CSB6 and
+	  HT1000 PATA controllers, via the new ATA layer.
+
+	  If unsure, say N.
+
+config PATA_PDC2027X
+	tristate "Promise PATA 2027x support"
+	depends on PCI
+	help
+	  This option enables support for Promise PATA pdc20268 to pdc20277 host adapters.
+
+	  If unsure, say N.
+
+config PATA_SIL680
+	tristate "CMD / Silicon Image 680 PATA support"
+	depends on PCI
+	help
+	  This option enables support for CMD / Silicon Image 680 PATA.
+
+	  If unsure, say N.
+
+config PATA_SIS
+	tristate "SiS PATA support (Experimental)"
+	depends on PCI && EXPERIMENTAL
+	help
+	  This option enables support for SiS PATA controllers
+
+	  If unsure, say N.
+
+config PATA_VIA
+	tristate "VIA PATA support"
+	depends on PCI
+	help
+	  This option enables support for the VIA PATA interfaces
+	  found on the many VIA chipsets.
+
+	  If unsure, say N.
+
+config PATA_WINBOND
+	tristate "Winbond SL82C105 PATA support"
+	depends on PCI
+	help
+	  This option enables support for SL82C105 PATA devices found in the
+	  Netwinder and some other systems
+
+	  If unsure, say N.
+
 endif
 endmenu
 
diff --git a/drivers/ata/Makefile b/drivers/ata/Makefile
index e260e3fe65c8..b183a04aeaf7 100644
--- a/drivers/ata/Makefile
+++ b/drivers/ata/Makefile
@@ -17,5 +17,46 @@ obj-$(CONFIG_SATA_ULI)		+= sata_uli.o
 obj-$(CONFIG_SATA_MV)		+= sata_mv.o
 obj-$(CONFIG_PDC_ADMA)		+= pdc_adma.o
 
+obj-$(CONFIG_PATA_ALI)		+= pata_ali.o
+obj-$(CONFIG_PATA_AMD)		+= pata_amd.o
+obj-$(CONFIG_PATA_ARTOP)	+= pata_artop.o
+obj-$(CONFIG_PATA_ATIIXP)	+= pata_atiixp.o
+obj-$(CONFIG_PATA_CMD64X)	+= pata_cmd64x.o
+obj-$(CONFIG_PATA_CS5520)	+= pata_cs5520.o
+obj-$(CONFIG_PATA_CS5530)	+= pata_cs5530.o
+obj-$(CONFIG_PATA_CS5535)	+= pata_cs5535.o
+obj-$(CONFIG_PATA_CYPRESS)	+= pata_cypress.o
+obj-$(CONFIG_PATA_EFAR)		+= pata_efar.o
+obj-$(CONFIG_PATA_HPT366)	+= pata_hpt366.o
+obj-$(CONFIG_PATA_HPT37X)	+= pata_hpt37x.o
+obj-$(CONFIG_PATA_HPT3X2N)	+= pata_hpt3x2n.o
+obj-$(CONFIG_PATA_HPT3X3)	+= pata_hpt3x3.o
+obj-$(CONFIG_PATA_ISAPNP)	+= pata_isapnp.o
+obj-$(CONFIG_PATA_IT8172)	+= pata_it8172.o
+obj-$(CONFIG_PATA_IT821X)	+= pata_it821x.o
+obj-$(CONFIG_PATA_NETCELL)	+= pata_netcell.o
+obj-$(CONFIG_PATA_NS87410)	+= pata_ns87410.o
+obj-$(CONFIG_PATA_OPTI)		+= pata_opti.o
+obj-$(CONFIG_PATA_OPTIDMA)	+= pata_optidma.o
+obj-$(CONFIG_PATA_MPIIX)	+= pata_mpiix.o
+obj-$(CONFIG_PATA_OLDPIIX)	+= pata_oldpiix.o
+obj-$(CONFIG_PATA_PCMCIA)	+= pata_pcmcia.o
+obj-$(CONFIG_PATA_PDC2027X)	+= pata_pdc2027x.o
+obj-$(CONFIG_PATA_PDC_OLD)	+= pata_pdc202xx_old.o
+obj-$(CONFIG_PATA_QDI)		+= pata_qdi.o
+obj-$(CONFIG_PATA_RADISYS)	+= pata_radisys.o
+obj-$(CONFIG_PATA_RZ1000)	+= pata_rz1000.o
+obj-$(CONFIG_PATA_SC1200)	+= pata_sc1200.o
+obj-$(CONFIG_PATA_SERVERWORKS)	+= pata_serverworks.o
+obj-$(CONFIG_PATA_SIL680)	+= pata_sil680.o
+obj-$(CONFIG_PATA_VIA)		+= pata_via.o
+obj-$(CONFIG_PATA_WINBOND)	+= pata_sl82c105.o
+obj-$(CONFIG_PATA_SIS)		+= pata_sis.o
+obj-$(CONFIG_PATA_TRIFLEX)	+= pata_triflex.o
+# Should be last but one libata driver
+obj-$(CONFIG_ATA_GENERIC)	+= ata_generic.o
+# Should be last libata driver
+obj-$(CONFIG_PATA_LEGACY)	+= pata_legacy.o
+
 libata-objs	:= libata-core.o libata-scsi.o libata-sff.o libata-eh.o
 
diff --git a/drivers/ata/ata_generic.c b/drivers/ata/ata_generic.c
new file mode 100644
index 000000000000..20b191dcf7ed
--- /dev/null
+++ b/drivers/ata/ata_generic.c
@@ -0,0 +1,252 @@
+/*
+ *  ata_generic.c - Generic PATA/SATA controller driver.
+ *  Copyright 2005 Red Hat Inc <alan@redhat.com>, all rights reserved.
+ *
+ *  Elements from ide/pci/generic.c 
+ *	    Copyright (C) 2001-2002	Andre Hedrick <andre@linux-ide.org>
+ *	    Portions (C) Copyright 2002  Red Hat Inc <alan@redhat.com>
+ *
+ *  May be copied or modified under the terms of the GNU General Public License
+ *  
+ *  Driver for PCI IDE interfaces implementing the standard bus mastering
+ *  interface functionality. This assumes the BIOS did the drive set up and
+ *  tuning for us. By default we do not grab all IDE class devices as they
+ *  may have other drivers or need fixups to avoid problems. Instead we keep
+ *  a default list of stuff without documentation/driver that appears to
+ *  work.  
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "ata_generic"
+#define DRV_VERSION "0.2.6"
+
+/*
+ *	A generic parallel ATA driver using libata
+ */
+
+/**
+ *	generic_pre_reset		-	probe begin
+ *	@ap: ATA port
+ *
+ *	Set up cable type and use generic probe init
+ */
+ 
+static int generic_pre_reset(struct ata_port *ap)
+{
+	ap->cbl = ATA_CBL_PATA80;
+	return ata_std_prereset(ap);
+}
+
+
+/**
+ *	generic_error_handler - Probe specified port on PATA host controller
+ *	@ap: Port to probe
+ *	@classes:
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+ 
+static void generic_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, generic_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	generic_set_mode	-	mode setting
+ *	@ap: interface to set up
+ *
+ *	Use a non standard set_mode function. We don't want to be tuned.
+ *	The BIOS configured everything. Our job is not to fiddle. We
+ *	read the dma enabled bits from the PCI configuration of the device
+ *	and respect them. 
+ */
+ 
+static void generic_set_mode(struct ata_port *ap)
+{
+	int dma_enabled = 0;
+	int i;
+
+	/* Bits 5 and 6 indicate if DMA is active on master/slave */
+	if (ap->ioaddr.bmdma_addr)
+		dma_enabled = inb(ap->ioaddr.bmdma_addr + ATA_DMA_CMD);
+	
+	for (i = 0; i < ATA_MAX_DEVICES; i++) {
+		struct ata_device *dev = &ap->device[i];
+		if (ata_dev_enabled(dev)) {
+			/* We don't really care */
+			dev->pio_mode = XFER_PIO_0;
+			dev->dma_mode = XFER_MW_DMA_0;
+			/* We do need the right mode information for DMA or PIO 
+			   and this comes from the current configuration flags */
+			if (dma_enabled & (1 << (5 + i))) {
+				dev->xfer_mode = XFER_MW_DMA_0;
+				dev->xfer_shift = ATA_SHIFT_MWDMA;
+				dev->flags &= ~ATA_DFLAG_PIO;
+			} else {
+				dev->xfer_mode = XFER_PIO_0;
+				dev->xfer_shift = ATA_SHIFT_PIO;
+				dev->flags |= ATA_DFLAG_PIO;
+			}
+		}
+	}
+}
+
+static struct scsi_host_template generic_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations generic_port_ops = {
+	.set_mode	= generic_set_mode,
+	
+	.port_disable	= ata_port_disable,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+	
+	.data_xfer	= ata_pio_data_xfer,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= generic_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+	
+static int all_generic_ide;		/* Set to claim all devices */
+
+/**
+ *	ata_generic_init		-	attach generic IDE
+ *	@dev: PCI device found
+ *	@id: match entry
+ *
+ *	Called each time a matching IDE interface is found. We check if the
+ *	interface is one we wish to claim and if so we perform any chip 
+ *	specific hacks then let the ATA layer do the heavy lifting.
+ */
+ 
+static int ata_generic_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	u16 command;
+	static struct ata_port_info info = {
+		.sht = &generic_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x3f,
+		.port_ops = &generic_port_ops
+	};
+	static struct ata_port_info *port_info[2] = { &info, &info };
+		
+	/* Don't use the generic entry unless instructed to do so */
+	if (id->driver_data == 1 && all_generic_ide == 0)
+		return -ENODEV;
+
+	/* Devices that need care */
+	if (dev->vendor == PCI_VENDOR_ID_UMC &&
+	    dev->device == PCI_DEVICE_ID_UMC_UM8886A &&
+	    (!(PCI_FUNC(dev->devfn) & 1)))
+		return -ENODEV;
+
+	if (dev->vendor == PCI_VENDOR_ID_OPTI &&
+	    dev->device == PCI_DEVICE_ID_OPTI_82C558 &&
+	    (!(PCI_FUNC(dev->devfn) & 1)))
+		return -ENODEV;
+
+	/* Don't re-enable devices in generic mode or we will break some
+	   motherboards with disabled and unused IDE controllers */
+	pci_read_config_word(dev, PCI_COMMAND, &command);
+	if (!(command & PCI_COMMAND_IO))
+		return -ENODEV;
+		
+	if (dev->vendor == PCI_VENDOR_ID_AL)
+	    	ata_pci_clear_simplex(dev);
+
+	return ata_pci_init_one(dev, port_info, 2);
+}
+
+static struct pci_device_id ata_generic[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_PCTECH, PCI_DEVICE_ID_PCTECH_SAMURAI_IDE), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_HOLTEK, PCI_DEVICE_ID_HOLTEK_6565), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_UMC,    PCI_DEVICE_ID_UMC_UM8673F), }, 
+	{ PCI_DEVICE(PCI_VENDOR_ID_UMC,    PCI_DEVICE_ID_UMC_UM8886A), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_UMC,    PCI_DEVICE_ID_UMC_UM8886BF), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_HINT,   PCI_DEVICE_ID_HINT_VXPROII_IDE), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_VIA,    PCI_DEVICE_ID_VIA_82C561), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_OPTI,   PCI_DEVICE_ID_OPTI_82C558), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_TOSHIBA,PCI_DEVICE_ID_TOSHIBA_PICCOLO), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_TOSHIBA,PCI_DEVICE_ID_TOSHIBA_PICCOLO_1), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_TOSHIBA,PCI_DEVICE_ID_TOSHIBA_PICCOLO_2),  },
+	/* Must come last. If you add entries adjust this table appropriately */
+	{ PCI_ANY_ID,		PCI_ANY_ID,			   PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_STORAGE_IDE << 8, 0xFFFFFF00UL, 1},
+	{ 0, },
+};
+
+static struct pci_driver ata_generic_pci_driver = {
+	.name 		= DRV_NAME,
+	.id_table	= ata_generic,
+	.probe 		= ata_generic_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init ata_generic_init(void)
+{
+	return pci_module_init(&ata_generic_pci_driver);
+}
+
+
+static void __exit ata_generic_exit(void)
+{
+	pci_unregister_driver(&ata_generic_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for generic ATA");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, ata_generic);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(ata_generic_init);
+module_exit(ata_generic_exit);
+
+module_param(all_generic_ide, int, 0);
diff --git a/drivers/ata/ata_piix.c b/drivers/ata/ata_piix.c
index 22b2dba90b9a..12b3a42fb356 100644
--- a/drivers/ata/ata_piix.c
+++ b/drivers/ata/ata_piix.c
@@ -93,7 +93,7 @@
 #include <linux/libata.h>
 
 #define DRV_NAME	"ata_piix"
-#define DRV_VERSION	"2.00"
+#define DRV_VERSION	"2.00ac6"
 
 enum {
 	PIIX_IOCFG		= 0x54, /* IDE I/O configuration register */
@@ -116,15 +116,18 @@ enum {
 	PIIX_80C_SEC		= (1 << 7) | (1 << 6),
 
 	/* controller IDs */
-	piix4_pata		= 0,
-	ich5_pata		= 1,
-	ich5_sata		= 2,
-	esb_sata		= 3,
-	ich6_sata		= 4,
-	ich6_sata_ahci		= 5,
-	ich6m_sata_ahci		= 6,
-	ich8_sata_ahci		= 7,
-
+	piix_pata_33		= 0,	/* PIIX3 or 4 at 33Mhz */
+	ich_pata_33		= 1,	/* ICH up to UDMA 33 only */
+	ich_pata_66		= 2,	/* ICH up to 66 Mhz */
+	ich_pata_100		= 3,	/* ICH up to UDMA 100 */
+	ich_pata_133		= 4,	/* ICH up to UDMA 133 */
+	ich5_sata		= 5,
+	esb_sata		= 6,
+	ich6_sata		= 7,
+	ich6_sata_ahci		= 8,
+	ich6m_sata_ahci		= 9,
+	ich8_sata_ahci		= 10,
+	
 	/* constants for mapping table */
 	P0			= 0,  /* port 0 */
 	P1			= 1,  /* port 1 */
@@ -152,19 +155,54 @@ struct piix_host_priv {
 static int piix_init_one (struct pci_dev *pdev,
 				    const struct pci_device_id *ent);
 static void piix_host_stop(struct ata_host *host);
-static void piix_set_piomode (struct ata_port *ap, struct ata_device *adev);
-static void piix_set_dmamode (struct ata_port *ap, struct ata_device *adev);
 static void piix_pata_error_handler(struct ata_port *ap);
+static void ich_pata_error_handler(struct ata_port *ap);
 static void piix_sata_error_handler(struct ata_port *ap);
+static void piix_set_piomode (struct ata_port *ap, struct ata_device *adev);
+static void piix_set_dmamode (struct ata_port *ap, struct ata_device *adev);
+static void ich_set_dmamode (struct ata_port *ap, struct ata_device *adev);
 
 static unsigned int in_module_init = 1;
 
 static const struct pci_device_id piix_pci_tbl[] = {
 #ifdef ATA_ENABLE_PATA
-	{ 0x8086, 0x7111, PCI_ANY_ID, PCI_ANY_ID, 0, 0, piix4_pata },
-	{ 0x8086, 0x24db, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich5_pata },
-	{ 0x8086, 0x25a2, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich5_pata },
-	{ 0x8086, 0x27df, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich5_pata },
+	/* Intel PIIX4 for the 430TX/440BX/MX chipset: UDMA 33 */
+	/* Also PIIX4E (fn3 rev 2) and PIIX4M (fn3 rev 3) */
+	{ 0x8086, 0x7111, PCI_ANY_ID, PCI_ANY_ID, 0, 0, piix_pata_33 },
+	{ 0x8086, 0x24db, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich_pata_100 },
+	{ 0x8086, 0x25a2, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich_pata_100 },
+	/* Intel PIIX4 */
+	{ 0x8086, 0x7199, PCI_ANY_ID, PCI_ANY_ID, 0, 0, piix_pata_33 },
+	/* Intel PIIX4 */
+	{ 0x8086, 0x7601, PCI_ANY_ID, PCI_ANY_ID, 0, 0, piix_pata_33 },
+	/* Intel PIIX */
+	{ 0x8086, 0x84CA, PCI_ANY_ID, PCI_ANY_ID, 0, 0, piix_pata_33 },
+	/* Intel ICH (i810, i815, i840) UDMA 66*/
+	{ 0x8086, 0x2411, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich_pata_66 },
+	/* Intel ICH0 : UDMA 33*/
+	{ 0x8086, 0x2421, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich_pata_33 },
+	/* Intel ICH2M */
+	{ 0x8086, 0x244A, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich_pata_100 },
+	/* Intel ICH2 (i810E2, i845, 850, 860) UDMA 100 */
+	{ 0x8086, 0x244B, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich_pata_100 },
+	/*  Intel ICH3M */
+	{ 0x8086, 0x248A, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich_pata_100 },
+	/* Intel ICH3 (E7500/1) UDMA 100 */
+	{ 0x8086, 0x248B, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich_pata_100 },
+	/* Intel ICH4 (i845GV, i845E, i852, i855) UDMA 100 */
+	{ 0x8086, 0x24CA, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich_pata_100 },
+	{ 0x8086, 0x24CB, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich_pata_100 },
+	/* Intel ICH5 */
+	{ 0x8086, 0x24DB, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich_pata_133 },
+	/* C-ICH (i810E2) */
+	{ 0x8086, 0x245B, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich_pata_100 },
+	/* ESB (855GME/875P + 6300ESB) UDMA 100  */	
+	{ 0x8086, 0x25A2, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich_pata_100 },
+	/* ICH6 (and 6) (i915) UDMA 100 */
+	{ 0x8086, 0x266F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich_pata_100 },
+	/* ICH7/7-R (i945, i975) UDMA 100*/
+	{ 0x8086, 0x27DF, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich_pata_133 },
+	{ 0x8086, 0x269E, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ich_pata_100 },
 #endif
 
 	/* NOTE: The following PCI ids must be kept in sync with the
@@ -263,6 +301,39 @@ static const struct ata_port_operations piix_pata_ops = {
 	.host_stop		= piix_host_stop,
 };
 
+static const struct ata_port_operations ich_pata_ops = {
+	.port_disable		= ata_port_disable,
+	.set_piomode		= piix_set_piomode,
+	.set_dmamode		= ich_set_dmamode,
+	.mode_filter		= ata_pci_default_filter,
+
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.check_status		= ata_check_status,
+	.exec_command		= ata_exec_command,
+	.dev_select		= ata_std_dev_select,
+
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= ata_qc_issue_prot,
+	.data_xfer		= ata_pio_data_xfer,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= ich_pata_error_handler,
+	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+
+	.irq_handler		= ata_interrupt,
+	.irq_clear		= ata_bmdma_irq_clear,
+
+	.port_start		= ata_port_start,
+	.port_stop		= ata_port_stop,
+	.host_stop		= ata_host_stop,
+};
+
 static const struct ata_port_operations piix_sata_ops = {
 	.port_disable		= ata_port_disable,
 
@@ -359,35 +430,56 @@ static const struct piix_map_db *piix_map_db_table[] = {
 };
 
 static struct ata_port_info piix_port_info[] = {
-	/* piix4_pata */
+	/* piix_pata_33: 0:  PIIX3 or 4 at 33MHz */
 	{
 		.sht		= &piix_sht,
-		.flags		= ATA_FLAG_SLAVE_POSS,
+		.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
 		.pio_mask	= 0x1f,	/* pio0-4 */
-#if 0
-		.mwdma_mask	= 0x06, /* mwdma1-2 */
-#else
-		.mwdma_mask	= 0x00, /* mwdma broken */
-#endif
+		.mwdma_mask	= 0x06, /* mwdma1-2 ?? CHECK 0 should be ok but slow */
 		.udma_mask	= ATA_UDMA_MASK_40C,
 		.port_ops	= &piix_pata_ops,
 	},
 
-	/* ich5_pata */
+	/* ich_pata_33: 1 	ICH0 - ICH at 33Mhz*/
+	{
+		.sht		= &piix_sht,
+		.flags		= ATA_FLAG_SRST | ATA_FLAG_SLAVE_POSS,
+		.pio_mask 	= 0x1f,	/* pio 0-4 */
+		.mwdma_mask	= 0x06, /* Check: maybe 0x07  */
+		.udma_mask	= ATA_UDMA2, /* UDMA33 */
+		.port_ops	= &ich_pata_ops,
+	},
+	/* ich_pata_66: 2 	ICH controllers up to 66MHz */
 	{
 		.sht		= &piix_sht,
-		.flags		= ATA_FLAG_SLAVE_POSS | PIIX_FLAG_CHECKINTR,
+		.flags		= ATA_FLAG_SRST | ATA_FLAG_SLAVE_POSS,
+		.pio_mask 	= 0x1f,	/* pio 0-4 */
+		.mwdma_mask	= 0x06, /* MWDMA0 is broken on chip */
+		.udma_mask	= ATA_UDMA4,
+		.port_ops	= &ich_pata_ops,
+	},
+	
+	/* ich_pata_100: 3 */
+	{
+		.sht		= &piix_sht,
+		.flags		= ATA_FLAG_SRST | ATA_FLAG_SLAVE_POSS | PIIX_FLAG_CHECKINTR,
 		.pio_mask	= 0x1f,	/* pio0-4 */
-#if 0
 		.mwdma_mask	= 0x06, /* mwdma1-2 */
-#else
-		.mwdma_mask	= 0x00, /* mwdma broken */
-#endif
-		.udma_mask	= 0x3f, /* udma0-5 */
-		.port_ops	= &piix_pata_ops,
+		.udma_mask	= ATA_UDMA5, /* udma0-5 */
+		.port_ops	= &ich_pata_ops,
 	},
 
-	/* ich5_sata */
+	/* ich_pata_133: 4 	ICH with full UDMA6 */
+	{
+		.sht		= &piix_sht,
+		.flags		= ATA_FLAG_SRST | ATA_FLAG_SLAVE_POSS | PIIX_FLAG_CHECKINTR,
+		.pio_mask 	= 0x1f,	/* pio 0-4 */
+		.mwdma_mask	= 0x06, /* Check: maybe 0x07  */
+		.udma_mask	= ATA_UDMA6, /* UDMA133 */
+		.port_ops	= &ich_pata_ops,
+	},
+
+	/* ich5_sata: 5 */
 	{
 		.sht		= &piix_sht,
 		.flags		= ATA_FLAG_SATA | PIIX_FLAG_CHECKINTR |
@@ -398,7 +490,7 @@ static struct ata_port_info piix_port_info[] = {
 		.port_ops	= &piix_sata_ops,
 	},
 
-	/* i6300esb_sata */
+	/* i6300esb_sata: 6 */
 	{
 		.sht		= &piix_sht,
 		.flags		= ATA_FLAG_SATA |
@@ -409,7 +501,7 @@ static struct ata_port_info piix_port_info[] = {
 		.port_ops	= &piix_sata_ops,
 	},
 
-	/* ich6_sata */
+	/* ich6_sata: 7 */
 	{
 		.sht		= &piix_sht,
 		.flags		= ATA_FLAG_SATA |
@@ -420,7 +512,7 @@ static struct ata_port_info piix_port_info[] = {
 		.port_ops	= &piix_sata_ops,
 	},
 
-	/* ich6_sata_ahci */
+	/* ich6_sata_ahci:8 */
 	{
 		.sht		= &piix_sht,
 		.flags		= ATA_FLAG_SATA |
@@ -432,7 +524,7 @@ static struct ata_port_info piix_port_info[] = {
 		.port_ops	= &piix_sata_ops,
 	},
 
-	/* ich6m_sata_ahci */
+	/* ich6m_sata_ahci: 9 */
 	{
 		.sht		= &piix_sht,
 		.flags		= ATA_FLAG_SATA |
@@ -455,6 +547,7 @@ static struct ata_port_info piix_port_info[] = {
 		.udma_mask	= 0x7f,	/* udma0-6 */
 		.port_ops	= &piix_sata_ops,
 	},
+
 };
 
 static struct pci_bits piix_enable_bits[] = {
@@ -483,7 +576,8 @@ MODULE_PARM_DESC(force_pcs, "force honoring or ignoring PCS to work around "
  *	LOCKING:
  *	None (inherited from caller).
  */
-static void piix_pata_cbl_detect(struct ata_port *ap)
+
+static void ich_pata_cbl_detect(struct ata_port *ap)
 {
 	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
 	u8 tmp, mask;
@@ -503,14 +597,12 @@ static void piix_pata_cbl_detect(struct ata_port *ap)
 
 cbl40:
 	ap->cbl = ATA_CBL_PATA40;
-	ap->udma_mask &= ATA_UDMA_MASK_40C;
 }
 
 /**
  *	piix_pata_prereset - prereset for PATA host controller
  *	@ap: Target port
  *
- *	Prereset including cable detection.
  *
  *	LOCKING:
  *	None (inherited from caller).
@@ -524,9 +616,7 @@ static int piix_pata_prereset(struct ata_port *ap)
 		ap->eh_context.i.action &= ~ATA_EH_RESET_MASK;
 		return 0;
 	}
-
-	piix_pata_cbl_detect(ap);
-
+	ap->cbl = ATA_CBL_PATA40;
 	return ata_std_prereset(ap);
 }
 
@@ -536,6 +626,36 @@ static void piix_pata_error_handler(struct ata_port *ap)
 			   ata_std_postreset);
 }
 
+
+/**
+ *	ich_pata_prereset - prereset for PATA host controller
+ *	@ap: Target port
+ *
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+static int ich_pata_prereset(struct ata_port *ap)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+
+	if (!pci_test_config_bits(pdev, &piix_enable_bits[ap->port_no])) {
+		ata_port_printk(ap, KERN_INFO, "port disabled. ignoring.\n");
+		ap->eh_context.i.action &= ~ATA_EH_RESET_MASK;
+		return 0;
+	}
+
+	ich_pata_cbl_detect(ap);
+
+	return ata_std_prereset(ap);
+}
+
+static void ich_pata_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, ich_pata_prereset, ata_std_softreset, NULL,
+			   ata_std_postreset);
+}
+
 /**
  *	piix_sata_present_mask - determine present mask for SATA host controller
  *	@ap: Target port
@@ -637,6 +757,13 @@ static void piix_set_piomode (struct ata_port *ap, struct ata_device *adev)
 	unsigned int slave_port	= 0x44;
 	u16 master_data;
 	u8 slave_data;
+	u8 udma_enable;
+	int control = 0;
+	
+	/*
+	 *	See Intel Document 298600-004 for the timing programing rules
+	 *	for ICH controllers.
+	 */
 
 	static const	 /* ISP  RTC */
 	u8 timings[][2]	= { { 0, 0 },
@@ -645,20 +772,30 @@ static void piix_set_piomode (struct ata_port *ap, struct ata_device *adev)
 			    { 2, 1 },
 			    { 2, 3 }, };
 
+	if (pio >= 2)
+		control |= 1;	/* TIME1 enable */
+	if (ata_pio_need_iordy(adev))
+		control |= 2;	/* IE enable */
+
+	/* Intel specifies that the PPE functionality is for disk only */	   
+	if (adev->class == ATA_DEV_ATA)
+		control |= 4;	/* PPE enable */
+
 	pci_read_config_word(dev, master_port, &master_data);
 	if (is_slave) {
+		/* Enable SITRE (seperate slave timing register) */
 		master_data |= 0x4000;
-		/* enable PPE, IE and TIME */
-		master_data |= 0x0070;
+		/* enable PPE1, IE1 and TIME1 as needed */
+		master_data |= (control << 4);
 		pci_read_config_byte(dev, slave_port, &slave_data);
 		slave_data &= (ap->port_no ? 0x0f : 0xf0);
-		slave_data |=
-			(timings[pio][0] << 2) |
-			(timings[pio][1] << (ap->port_no ? 4 : 0));
+		/* Load the timing nibble for this slave */
+		slave_data |= ((timings[pio][0] << 2) | timings[pio][1]) << (ap->port_no ? 4 : 0);
 	} else {
+		/* Master keeps the bits in a different format */
 		master_data &= 0xccf8;
-		/* enable PPE, IE and TIME */
-		master_data |= 0x0007;
+		/* Enable PPE, IE and TIME as appropriate */
+		master_data |= control;
 		master_data |=
 			(timings[pio][0] << 12) |
 			(timings[pio][1] << 8);
@@ -666,13 +803,23 @@ static void piix_set_piomode (struct ata_port *ap, struct ata_device *adev)
 	pci_write_config_word(dev, master_port, master_data);
 	if (is_slave)
 		pci_write_config_byte(dev, slave_port, slave_data);
+
+	/* Ensure the UDMA bit is off - it will be turned back on if
+	   UDMA is selected */
+	   
+	if (ap->udma_mask) {
+		pci_read_config_byte(dev, 0x48, &udma_enable);
+		udma_enable &= ~(1 << (2 * ap->port_no + adev->devno));
+		pci_write_config_byte(dev, 0x48, udma_enable);
+	}
 }
 
 /**
- *	piix_set_dmamode - Initialize host controller PATA PIO timings
+ *	do_pata_set_dmamode - Initialize host controller PATA PIO timings
  *	@ap: Port whose timings we are configuring
- *	@adev: um
+ *	@adev: Drive in question
  *	@udma: udma mode, 0 - 6
+ *	@is_ich: set if the chip is an ICH device
  *
  *	Set UDMA mode for device, in host controller PCI config space.
  *
@@ -680,70 +827,140 @@ static void piix_set_piomode (struct ata_port *ap, struct ata_device *adev)
  *	None (inherited from caller).
  */
 
-static void piix_set_dmamode (struct ata_port *ap, struct ata_device *adev)
+static void do_pata_set_dmamode (struct ata_port *ap, struct ata_device *adev, int isich)
 {
-	unsigned int udma	= adev->dma_mode; /* FIXME: MWDMA too */
 	struct pci_dev *dev	= to_pci_dev(ap->host->dev);
-	u8 maslave		= ap->port_no ? 0x42 : 0x40;
-	u8 speed		= udma;
-	unsigned int drive_dn	= (ap->port_no ? 2 : 0) + adev->devno;
-	int a_speed		= 3 << (drive_dn * 4);
-	int u_flag		= 1 << drive_dn;
-	int v_flag		= 0x01 << drive_dn;
-	int w_flag		= 0x10 << drive_dn;
-	int u_speed		= 0;
-	int			sitre;
-	u16			reg4042, reg4a;
-	u8			reg48, reg54, reg55;
-
-	pci_read_config_word(dev, maslave, &reg4042);
-	DPRINTK("reg4042 = 0x%04x\n", reg4042);
-	sitre = (reg4042 & 0x4000) ? 1 : 0;
-	pci_read_config_byte(dev, 0x48, &reg48);
-	pci_read_config_word(dev, 0x4a, &reg4a);
-	pci_read_config_byte(dev, 0x54, &reg54);
-	pci_read_config_byte(dev, 0x55, &reg55);
-
-	switch(speed) {
-		case XFER_UDMA_4:
-		case XFER_UDMA_2:	u_speed = 2 << (drive_dn * 4); break;
-		case XFER_UDMA_6:
-		case XFER_UDMA_5:
-		case XFER_UDMA_3:
-		case XFER_UDMA_1:	u_speed = 1 << (drive_dn * 4); break;
-		case XFER_UDMA_0:	u_speed = 0 << (drive_dn * 4); break;
-		case XFER_MW_DMA_2:
-		case XFER_MW_DMA_1:	break;
-		default:
-			BUG();
-			return;
-	}
+	u8 master_port		= ap->port_no ? 0x42 : 0x40;
+	u16 master_data;
+	u8 speed		= adev->dma_mode;
+	int devid		= adev->devno + 2 * ap->port_no;
+	u8 udma_enable;
+	
+	static const	 /* ISP  RTC */
+	u8 timings[][2]	= { { 0, 0 },
+			    { 0, 0 },
+			    { 1, 0 },
+			    { 2, 1 },
+			    { 2, 3 }, };
+
+	pci_read_config_word(dev, master_port, &master_data);
+	pci_read_config_byte(dev, 0x48, &udma_enable);
 
 	if (speed >= XFER_UDMA_0) {
-		if (!(reg48 & u_flag))
-			pci_write_config_byte(dev, 0x48, reg48 | u_flag);
-		if (speed == XFER_UDMA_5) {
-			pci_write_config_byte(dev, 0x55, (u8) reg55|w_flag);
-		} else {
-			pci_write_config_byte(dev, 0x55, (u8) reg55 & ~w_flag);
+		unsigned int udma = adev->dma_mode - XFER_UDMA_0;
+		u16 udma_timing;
+		u16 ideconf;
+		int u_clock, u_speed;
+		
+		/*
+	 	 * UDMA is handled by a combination of clock switching and
+		 * selection of dividers 
+		 *  
+		 * Handy rule: Odd modes are UDMATIMx 01, even are 02
+		 *	       except UDMA0 which is 00 
+		 */
+		u_speed = min(2 - (udma & 1), udma);
+		if (udma == 5)
+			u_clock = 0x1000;	/* 100Mhz */
+		else if (udma > 2)
+			u_clock = 1;		/* 66Mhz */
+		else
+			u_clock = 0;		/* 33Mhz */
+			
+		udma_enable |= (1 << devid);
+			
+		/* Load the CT/RP selection */
+		pci_read_config_word(dev, 0x4A, &udma_timing);
+		udma_timing &= ~(3 << (4 * devid));
+		udma_timing |= u_speed << (4 * devid);
+		pci_write_config_word(dev, 0x4A, udma_timing);
+
+		if (isich) {		
+			/* Select a 33/66/100Mhz clock */
+			pci_read_config_word(dev, 0x54, &ideconf);
+			ideconf &= ~(0x1001 << devid);
+			ideconf |= u_clock << devid;
+			/* For ICH or later we should set bit 10 for better
+			   performance (WR_PingPong_En) */
+			pci_write_config_word(dev, 0x54, ideconf);
 		}
-		if ((reg4a & a_speed) != u_speed)
-			pci_write_config_word(dev, 0x4a, (reg4a & ~a_speed) | u_speed);
-		if (speed > XFER_UDMA_2) {
-			if (!(reg54 & v_flag))
-				pci_write_config_byte(dev, 0x54, reg54 | v_flag);
-		} else
-			pci_write_config_byte(dev, 0x54, reg54 & ~v_flag);
 	} else {
-		if (reg48 & u_flag)
-			pci_write_config_byte(dev, 0x48, reg48 & ~u_flag);
-		if (reg4a & a_speed)
-			pci_write_config_word(dev, 0x4a, reg4a & ~a_speed);
-		if (reg54 & v_flag)
-			pci_write_config_byte(dev, 0x54, reg54 & ~v_flag);
-		if (reg55 & w_flag)
-			pci_write_config_byte(dev, 0x55, (u8) reg55 & ~w_flag);
+		/*
+		 * MWDMA is driven by the PIO timings. We must also enable
+		 * IORDY unconditionally along with TIME1. PPE has already
+		 * been set when the PIO timing was set.
+		 */
+		unsigned int mwdma	= adev->dma_mode - XFER_MW_DMA_0;
+		unsigned int control;
+		u8 slave_data;
+		const unsigned int needed_pio[3] = {
+			XFER_PIO_0, XFER_PIO_3, XFER_PIO_4
+		};
+		int pio = needed_pio[mwdma] - XFER_PIO_0;
+		
+		control = 3;	/* IORDY|TIME1 */
+		
+		/* If the drive MWDMA is faster than it can do PIO then
+		   we must force PIO into PIO0 */
+		   
+		if (adev->pio_mode < needed_pio[mwdma])
+			/* Enable DMA timing only */
+			control |= 8;	/* PIO cycles in PIO0 */
+
+		if (adev->devno) {	/* Slave */
+			master_data &= 0xFF4F;  /* Mask out IORDY|TIME1|DMAONLY */
+			master_data |= control << 4;
+			pci_read_config_byte(dev, 0x44, &slave_data);
+			slave_data &= (0x0F + 0xE1 * ap->port_no);
+			/* Load the matching timing */
+			slave_data |= ((timings[pio][0] << 2) | timings[pio][1]) << (ap->port_no ? 4 : 0);
+			pci_write_config_byte(dev, 0x44, slave_data);
+		} else { 	/* Master */
+			master_data &= 0xCCF4;	/* Mask out IORDY|TIME1|DMAONLY 
+						   and master timing bits */
+			master_data |= control;
+			master_data |=
+				(timings[pio][0] << 12) |
+				(timings[pio][1] << 8);
+		}
+		udma_enable &= ~(1 << devid);
+		pci_write_config_word(dev, master_port, master_data);
 	}
+	/* Don't scribble on 0x48 if the controller does not support UDMA */
+	if (ap->udma_mask)
+		pci_write_config_byte(dev, 0x48, udma_enable);
+}
+
+/**
+ *	piix_set_dmamode - Initialize host controller PATA DMA timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: um
+ *
+ *	Set MW/UDMA mode for device, in host controller PCI config space.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void piix_set_dmamode (struct ata_port *ap, struct ata_device *adev)
+{
+	do_pata_set_dmamode(ap, adev, 0);
+}
+
+/**
+ *	ich_set_dmamode - Initialize host controller PATA DMA timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: um
+ *
+ *	Set MW/UDMA mode for device, in host controller PCI config space.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void ich_set_dmamode (struct ata_port *ap, struct ata_device *adev)
+{
+	do_pata_set_dmamode(ap, adev, 1);
 }
 
 #define AHCI_PCI_BAR 5
@@ -872,7 +1089,7 @@ static void __devinit piix_init_sata_map(struct pci_dev *pdev,
 
 		case IDE:
 			WARN_ON((i & 1) || map[i + 1] != IDE);
-			pinfo[i / 2] = piix_port_info[ich5_pata];
+			pinfo[i / 2] = piix_port_info[ich_pata_100];
 			pinfo[i / 2].private_data = hpriv;
 			i++;
 			printk(" IDE IDE");
@@ -1007,4 +1224,3 @@ static void __exit piix_exit(void)
 
 module_init(piix_init);
 module_exit(piix_exit);
-
diff --git a/drivers/ata/pata_ali.c b/drivers/ata/pata_ali.c
new file mode 100644
index 000000000000..0e0415ecfce7
--- /dev/null
+++ b/drivers/ata/pata_ali.c
@@ -0,0 +1,679 @@
+/*
+ * pata_ali.c 	- ALI 15x3 PATA for new ATA layer
+ *			  (C) 2005 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ * based in part upon
+ * linux/drivers/ide/pci/alim15x3.c		Version 0.17	2003/01/02
+ *
+ *  Copyright (C) 1998-2000 Michel Aubry, Maintainer
+ *  Copyright (C) 1998-2000 Andrzej Krzysztofowicz, Maintainer
+ *  Copyright (C) 1999-2000 CJ, cjtsai@ali.com.tw, Maintainer
+ *
+ *  Copyright (C) 1998-2000 Andre Hedrick (andre@linux-ide.org)
+ *  May be copied or modified under the terms of the GNU General Public License
+ *  Copyright (C) 2002 Alan Cox <alan@redhat.com>
+ *  ALi (now ULi M5228) support by Clear Zhang <Clear.Zhang@ali.com.tw>
+ *
+ *  Documentation
+ *	Chipset documentation available under NDA only
+ *
+ *  TODO/CHECK
+ *	Cannot have ATAPI on both master & slave for rev < c2 (???) but
+ *	otherwise should do atapi DMA.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+#include <linux/dmi.h>
+
+#define DRV_NAME "pata_ali"
+#define DRV_VERSION "0.6.5"
+
+/*
+ *	Cable special cases
+ */
+
+static struct dmi_system_id cable_dmi_table[] = {
+	{
+		.ident = "HP Pavilion N5430",
+		.matches = {
+			DMI_MATCH(DMI_BOARD_VENDOR, "Hewlett-Packard"),
+			DMI_MATCH(DMI_BOARD_NAME, "OmniBook N32N-736"),
+		},
+	},
+	{ }
+};
+
+static int ali_cable_override(struct pci_dev *pdev)
+{
+	/* Fujitsu P2000 */
+	if (pdev->subsystem_vendor == 0x10CF && pdev->subsystem_device == 0x10AF)
+	   	return 1;
+	/* Systems by DMI */
+	if (dmi_check_system(cable_dmi_table))
+		return 1;
+	return 0;
+}
+
+/**
+ *	ali_c2_cable_detect	-	cable detection
+ *	@ap: ATA port
+ *
+ *	Perform cable detection for C2 and later revisions
+ */
+
+static int ali_c2_cable_detect(struct ata_port *ap)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u8 ata66;
+
+	/* Certain laptops use short but suitable cables and don't
+	   implement the detect logic */
+
+	if (ali_cable_override(pdev))
+		return ATA_CBL_PATA80;
+
+	/* Host view cable detect 0x4A bit 0 primary bit 1 secondary
+	   Bit set for 40 pin */
+	pci_read_config_byte(pdev, 0x4A, &ata66);
+	if (ata66 & (1 << ap->port_no))
+		return ATA_CBL_PATA40;
+	else
+		return ATA_CBL_PATA80;
+}
+
+/**
+ *	ali_early_error_handler	-	reset for eary chip
+ *	@ap: ATA port
+ *
+ *	Handle the reset callback for the later chips with cable detect
+ */
+
+static int ali_c2_pre_reset(struct ata_port *ap)
+{
+	ap->cbl = ali_c2_cable_detect(ap);
+	return ata_std_prereset(ap);
+}
+
+static void ali_c2_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, ali_c2_pre_reset,
+			       ata_std_softreset, NULL,
+			       ata_std_postreset);
+}
+
+/**
+ *	ali_early_cable_detect	-	cable detection
+ *	@ap: ATA port
+ *
+ *	Perform cable detection for older chipsets. This turns out to be
+ *	rather easy to implement
+ */
+
+static int ali_early_cable_detect(struct ata_port *ap)
+{
+	return ATA_CBL_PATA40;
+}
+
+/**
+ *	ali_early_probe_init	-	reset for early chip
+ *	@ap: ATA port
+ *
+ *	Handle the reset callback for the early (pre cable detect) chips.
+ */
+
+static int ali_early_pre_reset(struct ata_port *ap)
+{
+	ap->cbl = ali_early_cable_detect(ap);
+	return ata_std_prereset(ap);
+}
+
+static void ali_early_error_handler(struct ata_port *ap)
+{
+	return ata_bmdma_drive_eh(ap, ali_early_pre_reset,
+				     ata_std_softreset, NULL,
+				     ata_std_postreset);
+}
+
+/**
+ *	ali_20_filter		-	filter for earlier ALI DMA
+ *	@ap: ALi ATA port
+ *	@adev: attached device
+ *
+ *	Ensure that we do not do DMA on CD devices. We may be able to
+ *	fix that later on. Also ensure we do not do UDMA on WDC drives
+ */
+
+static unsigned long ali_20_filter(const struct ata_port *ap, struct ata_device *adev, unsigned long mask)
+{
+	char model_num[40];
+	/* No DMA on anything but a disk for now */
+	if (adev->class != ATA_DEV_ATA)
+		mask &= ~(ATA_MASK_MWDMA | ATA_MASK_UDMA);
+	ata_id_string(adev->id, model_num, ATA_ID_PROD_OFS, sizeof(model_num));
+	if (strstr(model_num, "WDC"))
+		return mask &= ~ATA_MASK_UDMA;
+	return ata_pci_default_filter(ap, adev, mask);
+}
+
+/**
+ *	ali_fifo_control	-	FIFO manager
+ *	@ap: ALi channel to control
+ *	@adev: device for FIFO control
+ *	@on: 0 for off 1 for on
+ *
+ *	Enable or disable the FIFO on a given device. Because of the way the
+ *	ALi FIFO works it provides a boost on ATA disk but can be confused by
+ *	ATAPI and we must therefore manage it.
+ */
+
+static void ali_fifo_control(struct ata_port *ap, struct ata_device *adev, int on)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	int pio_fifo = 0x54 + ap->port_no;
+	u8 fifo;
+	int shift = 4 * adev->devno;
+
+	/* ATA - FIFO on set nibble to 0x05, ATAPI - FIFO off, set nibble to
+	   0x00. Not all the docs agree but the behaviour we now use is the
+	   one stated in the BIOS Programming Guide */
+	   
+	pci_read_config_byte(pdev, pio_fifo, &fifo);
+	fifo &= ~(0x0F << shift);
+	if (on)
+		fifo |= (on << shift);
+	pci_write_config_byte(pdev, pio_fifo, fifo);
+}
+
+/**
+ *	ali_program_modes	-	load mode registers
+ *	@ap: ALi channel to load
+ *	@adev: Device the timing is for
+ *	@cmd: Command timing
+ *	@data: Data timing
+ *	@ultra: UDMA timing or zero for off
+ *
+ *	Loads the timing registers for cmd/data and disable UDMA if
+ *	ultra is zero. If ultra is set then load and enable the UDMA
+ *	timing but do not touch the command/data timing.
+ */
+
+static void ali_program_modes(struct ata_port *ap, struct ata_device *adev, struct ata_timing *t, u8 ultra)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	int cas = 0x58 + 4 * ap->port_no;	/* Command timing */
+	int cbt = 0x59 + 4 * ap->port_no;	/* Command timing */
+	int drwt = 0x5A + 4 * ap->port_no + adev->devno; /* R/W timing */
+	int udmat = 0x56 + ap->port_no;	/* UDMA timing */
+	int shift = 4 * adev->devno;
+	u8 udma;
+
+	if (t != NULL) {
+		t->setup = FIT(t->setup, 1, 8) & 7;
+		t->act8b = FIT(t->act8b, 1, 8) & 7;
+		t->rec8b = FIT(t->rec8b, 1, 16) & 15;
+		t->active = FIT(t->active, 1, 8) & 7;
+		t->recover = FIT(t->recover, 1, 16) & 15;
+
+		pci_write_config_byte(pdev, cas, t->setup);
+		pci_write_config_byte(pdev, cbt, (t->act8b << 4) | t->rec8b);
+		pci_write_config_byte(pdev, drwt, (t->active << 4) | t->recover);
+	}
+
+	/* Set up the UDMA enable */
+	pci_read_config_byte(pdev, udmat, &udma);
+	udma &= ~(0x0F << shift);
+	udma |= ultra << shift;
+	pci_write_config_byte(pdev, udmat, udma);
+}
+
+/**
+ *	ali_set_piomode	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Program the ALi registers for PIO mode. FIXME: add timings for
+ *	PIO5.
+ */
+
+static void ali_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct ata_device *pair = ata_dev_pair(adev);
+	struct ata_timing t;
+	unsigned long T =  1000000000 / 33333;	/* PCI clock based */
+
+	ata_timing_compute(adev, adev->pio_mode, &t, T, 1);
+	if (pair) {
+		struct ata_timing p;
+		ata_timing_compute(pair, pair->pio_mode, &p, T, 1);
+		ata_timing_merge(&p, &t, &t, ATA_TIMING_SETUP|ATA_TIMING_8BIT);
+		if (pair->dma_mode) {
+			ata_timing_compute(pair, pair->dma_mode, &p, T, 1);
+			ata_timing_merge(&p, &t, &t, ATA_TIMING_SETUP|ATA_TIMING_8BIT);
+		}
+	}
+
+	/* PIO FIFO is only permitted on ATA disk */
+	if (adev->class != ATA_DEV_ATA)
+		ali_fifo_control(ap, adev, 0x00);
+	ali_program_modes(ap, adev, &t, 0);
+	if (adev->class == ATA_DEV_ATA)
+		ali_fifo_control(ap, adev, 0x05);
+
+}
+
+/**
+ *	ali_set_dmamode	-	set initial DMA mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	FIXME: MWDMA timings
+ */
+
+static void ali_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	static u8 udma_timing[7] = { 0xC, 0xB, 0xA, 0x9, 0x8, 0xF, 0xD };
+	struct ata_device *pair = ata_dev_pair(adev);
+	struct ata_timing t;
+	unsigned long T =  1000000000 / 33333;	/* PCI clock based */
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+
+
+	if (adev->class == ATA_DEV_ATA)
+		ali_fifo_control(ap, adev, 0x08);
+
+	if (adev->dma_mode >= XFER_UDMA_0) {
+		ali_program_modes(ap, adev, NULL, udma_timing[adev->dma_mode - XFER_UDMA_0]);
+		if (adev->dma_mode >= XFER_UDMA_3) {
+			u8 reg4b;
+			pci_read_config_byte(pdev, 0x4B, &reg4b);
+			reg4b |= 1;
+			pci_write_config_byte(pdev, 0x4B, reg4b);
+		}
+	} else {
+		ata_timing_compute(adev, adev->dma_mode, &t, T, 1);
+		if (pair) {
+			struct ata_timing p;
+			ata_timing_compute(pair, pair->pio_mode, &p, T, 1);
+			ata_timing_merge(&p, &t, &t, ATA_TIMING_SETUP|ATA_TIMING_8BIT);
+			if (pair->dma_mode) {
+				ata_timing_compute(pair, pair->dma_mode, &p, T, 1);
+				ata_timing_merge(&p, &t, &t, ATA_TIMING_SETUP|ATA_TIMING_8BIT);
+			}
+		}
+		ali_program_modes(ap, adev, &t, 0);
+	}
+}
+
+/**
+ *	ali_lock_sectors	-	Keep older devices to 255 sector mode
+ *	@ap: ATA port
+ *	@adev: Device
+ *
+ *	Called during the bus probe for each device that is found. We use
+ *	this call to lock the sector count of the device to 255 or less on
+ *	older ALi controllers. If we didn't do this then large I/O's would
+ *	require LBA48 commands which the older ALi requires are issued by
+ *	slower PIO methods
+ */
+
+static void ali_lock_sectors(struct ata_port *ap, struct ata_device *adev)
+{
+	adev->max_sectors = 255;
+}
+
+static struct scsi_host_template ali_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	/* Keep LBA28 counts so large I/O's don't turn LBA48 and PIO
+	   with older controllers. Not locked so will grow on C5 or later */
+	.max_sectors		= 255,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+/*
+ *	Port operations for PIO only ALi
+ */
+
+static struct ata_port_operations ali_early_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= ali_set_piomode,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= ali_early_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+/*
+ *	Port operations for DMA capable ALi without cable
+ *	detect
+ */
+static struct ata_port_operations ali_20_port_ops = {
+	.port_disable	= ata_port_disable,
+
+	.set_piomode	= ali_set_piomode,
+	.set_dmamode	= ali_set_dmamode,
+	.mode_filter	= ali_20_filter,
+
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+	.dev_config	= ali_lock_sectors,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= ali_early_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+/*
+ *	Port operations for DMA capable ALi with cable detect
+ */
+static struct ata_port_operations ali_c2_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= ali_set_piomode,
+	.set_dmamode	= ali_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+	.dev_config	= ali_lock_sectors,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= ali_c2_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+/*
+ *	Port operations for DMA capable ALi with cable detect and LBA48
+ */
+static struct ata_port_operations ali_c5_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= ali_set_piomode,
+	.set_dmamode	= ali_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= ali_c2_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+/**
+ *	ali_init_one		-	discovery callback
+ *	@pdev: PCI device ID
+ *	@id: PCI table info
+ *
+ *	An ALi IDE interface has been discovered. Figure out what revision
+ *	and perform configuration work before handing it to the ATA layer
+ */
+
+static int ali_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info_early = {
+		.sht = &ali_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.port_ops = &ali_early_port_ops
+	};
+	/* Revision 0x20 added DMA */
+	static struct ata_port_info info_20 = {
+		.sht = &ali_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST | ATA_FLAG_PIO_LBA48,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.port_ops = &ali_20_port_ops
+	};
+	/* Revision 0x20 with support logic added UDMA */
+	static struct ata_port_info info_20_udma = {
+		.sht = &ali_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST | ATA_FLAG_PIO_LBA48,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07, 
+		.udma_mask = 0x07,	/* UDMA33 */
+		.port_ops = &ali_20_port_ops
+	};
+	/* Revision 0xC2 adds UDMA66 */
+	static struct ata_port_info info_c2 = {
+		.sht = &ali_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST | ATA_FLAG_PIO_LBA48,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x1f,
+		.port_ops = &ali_c2_port_ops
+	};
+	/* Revision 0xC3 is UDMA100 */
+	static struct ata_port_info info_c3 = {
+		.sht = &ali_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST | ATA_FLAG_PIO_LBA48,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x3f,
+		.port_ops = &ali_c2_port_ops
+	};
+	/* Revision 0xC4 is UDMA133 */
+	static struct ata_port_info info_c4 = {
+		.sht = &ali_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST | ATA_FLAG_PIO_LBA48,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x7f,
+		.port_ops = &ali_c2_port_ops
+	};
+	/* Revision 0xC5 is UDMA133 with LBA48 DMA */
+	static struct ata_port_info info_c5 = {
+		.sht = &ali_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x7f,
+		.port_ops = &ali_c5_port_ops
+	};
+
+	static struct ata_port_info *port_info[2];
+	u8 rev, tmp;
+	struct pci_dev *north, *isa_bridge;
+
+	pci_read_config_byte(pdev, PCI_REVISION_ID, &rev);
+
+	/*
+	 * The chipset revision selects the driver operations and
+	 * mode data.
+	 */
+
+	if (rev < 0x20) {
+		port_info[0] = port_info[1] = &info_early;
+	} else if (rev < 0xC2) {
+		/* 1543-E/F, 1543C-C, 1543C-D, 1543C-E */
+		pci_read_config_byte(pdev, 0x4B, &tmp);
+		/* Clear CD-ROM DMA write bit */
+		tmp &= 0x7F;
+		pci_write_config_byte(pdev, 0x4B, tmp);
+        	port_info[0] = port_info[1] = &info_20;
+	} else if (rev == 0xC2) {
+        	port_info[0] = port_info[1] = &info_c2;
+	} else if (rev == 0xC3) {
+        	port_info[0] = port_info[1] = &info_c3;
+	} else if (rev == 0xC4) {
+        	port_info[0] = port_info[1] = &info_c4;
+	} else
+        	port_info[0] = port_info[1] = &info_c5;
+
+	if (rev >= 0xC2) {
+		/* Enable cable detection logic */
+		pci_read_config_byte(pdev, 0x4B, &tmp);
+		pci_write_config_byte(pdev, 0x4B, tmp | 0x08);
+	}
+
+	north = pci_get_slot(pdev->bus, PCI_DEVFN(0,0));
+	isa_bridge = pci_get_device(PCI_VENDOR_ID_AL, PCI_DEVICE_ID_AL_M1533, NULL);
+
+	if (north && north->vendor == PCI_VENDOR_ID_AL) {
+		/* Configure the ALi bridge logic. For non ALi rely on BIOS.
+		   Set the south bridge enable bit */
+		pci_read_config_byte(isa_bridge, 0x79, &tmp);
+		if (rev == 0xC2)
+			pci_write_config_byte(isa_bridge, 0x79, tmp | 0x04);
+		else if (rev > 0xC2)
+			pci_write_config_byte(isa_bridge, 0x79, tmp | 0x02);
+	}
+
+	if (rev >= 0x20) {
+		if (rev < 0xC2) {
+			/* Are we paired with a UDMA capable chip */
+			pci_read_config_byte(isa_bridge, 0x5E, &tmp);
+			if ((tmp & 0x1E) == 0x12)
+		        	port_info[0] = port_info[1] = &info_20_udma;
+		}
+		/*
+		 * CD_ROM DMA on (0x53 bit 0). Enable this even if we want
+		 * to use PIO. 0x53 bit 1 (rev 20 only) - enable FIFO control
+		 * via 0x54/55.
+		 */
+		pci_read_config_byte(pdev, 0x53, &tmp);
+		if (rev <= 0x20)
+			tmp &= ~0x02;
+		if (rev == 0xc7)
+			tmp |= 0x03;
+		else
+			tmp |= 0x01;	/* CD_ROM enable for DMA */
+		pci_write_config_byte(pdev, 0x53, tmp);
+	}
+
+	pci_dev_put(isa_bridge);
+	pci_dev_put(north);
+
+	ata_pci_clear_simplex(pdev);
+	return ata_pci_init_one(pdev, port_info, 2);
+}
+
+static struct pci_device_id ali[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_AL, PCI_DEVICE_ID_AL_M5228), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_AL, PCI_DEVICE_ID_AL_M5229), },
+	{ 0, },
+};
+
+static struct pci_driver ali_pci_driver = {
+	.name 		= DRV_NAME,
+	.id_table	= ali,
+	.probe 		= ali_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init ali_init(void)
+{
+	return pci_register_driver(&ali_pci_driver);
+}
+
+
+static void __exit ali_exit(void)
+{
+	pci_unregister_driver(&ali_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for ALi PATA");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, ali);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(ali_init);
+module_exit(ali_exit);
diff --git a/drivers/ata/pata_amd.c b/drivers/ata/pata_amd.c
new file mode 100644
index 000000000000..6ab568428a14
--- /dev/null
+++ b/drivers/ata/pata_amd.c
@@ -0,0 +1,707 @@
+/*
+ * pata_amd.c 	- AMD PATA for new ATA layer
+ *			  (C) 2005-2006 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ *  Based on pata-sil680. Errata information is taken from data sheets
+ *  and the amd74xx.c driver by Vojtech Pavlik. Nvidia SATA devices are
+ *  claimed by sata-nv.c.
+ *
+ *  TODO:
+ *	Variable system clock when/if it makes sense
+ *	Power management on ports
+ *
+ *
+ *  Documentation publically available.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "pata_amd"
+#define DRV_VERSION "0.2.2"
+
+/**
+ *	timing_setup		-	shared timing computation and load
+ *	@ap: ATA port being set up
+ *	@adev: drive being configured
+ *	@offset: port offset
+ *	@speed: target speed
+ *	@clock: clock multiplier (number of times 33MHz for this part)
+ *
+ *	Perform the actual timing set up for Nvidia or AMD PATA devices.
+ *	The actual devices vary so they all call into this helper function
+ *	providing the clock multipler and offset (because AMD and Nvidia put
+ *	the ports at different locations).
+ */
+
+static void timing_setup(struct ata_port *ap, struct ata_device *adev, int offset, int speed, int clock)
+{
+	static const unsigned char amd_cyc2udma[] = {
+		6, 6, 5, 4, 0, 1, 1, 2, 2, 3, 3, 3, 3, 3, 3, 7
+	};
+
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	struct ata_device *peer = ata_dev_pair(adev);
+	int dn = ap->port_no * 2 + adev->devno;
+	struct ata_timing at, apeer;
+	int T, UT;
+	const int amd_clock = 33333;	/* KHz. */
+	u8 t;
+
+	T = 1000000000 / amd_clock;
+	UT = T / min_t(int, max_t(int, clock, 1), 2);
+
+	if (ata_timing_compute(adev, speed, &at, T, UT) < 0) {
+		dev_printk(KERN_ERR, &pdev->dev, "unknown mode %d.\n", speed);
+		return;
+	}
+
+	if (peer) {
+		/* This may be over conservative */
+		if (peer->dma_mode) {
+			ata_timing_compute(peer, peer->dma_mode, &apeer, T, UT);
+			ata_timing_merge(&apeer, &at, &at, ATA_TIMING_8BIT);
+		}
+		ata_timing_compute(peer, peer->pio_mode, &apeer, T, UT);
+		ata_timing_merge(&apeer, &at, &at, ATA_TIMING_8BIT);
+	}
+
+	if (speed == XFER_UDMA_5 && amd_clock <= 33333) at.udma = 1;
+	if (speed == XFER_UDMA_6 && amd_clock <= 33333) at.udma = 15;
+
+	/*
+	 *	Now do the setup work
+	 */
+
+	/* Configure the address set up timing */
+	pci_read_config_byte(pdev, offset + 0x0C, &t);
+	t = (t & ~(3 << ((3 - dn) << 1))) | ((FIT(at.setup, 1, 4) - 1) << ((3 - dn) << 1));
+	pci_write_config_byte(pdev, offset + 0x0C , t);
+
+	/* Configure the 8bit I/O timing */
+	pci_write_config_byte(pdev, offset + 0x0E + (1 - (dn >> 1)),
+		((FIT(at.act8b, 1, 16) - 1) << 4) | (FIT(at.rec8b, 1, 16) - 1));
+
+	/* Drive timing */
+	pci_write_config_byte(pdev, offset + 0x08 + (3 - dn),
+		((FIT(at.active, 1, 16) - 1) << 4) | (FIT(at.recover, 1, 16) - 1));
+
+	switch (clock) {
+		case 1:
+		t = at.udma ? (0xc0 | (FIT(at.udma, 2, 5) - 2)) : 0x03;
+		break;
+
+		case 2:
+		t = at.udma ? (0xc0 | amd_cyc2udma[FIT(at.udma, 2, 10)]) : 0x03;
+		break;
+
+		case 3:
+		t = at.udma ? (0xc0 | amd_cyc2udma[FIT(at.udma, 1, 10)]) : 0x03;
+		break;
+
+		case 4:
+		t = at.udma ? (0xc0 | amd_cyc2udma[FIT(at.udma, 1, 15)]) : 0x03;
+		break;
+
+		default:
+			return;
+	}
+
+	/* UDMA timing */
+	pci_write_config_byte(pdev, offset + 0x10 + (3 - dn), t);
+}
+
+/**
+ *	amd_probe_init		-	cable detection
+ *	@ap: ATA port
+ *
+ *	Perform cable detection. The BIOS stores this in PCI config
+ *	space for us.
+ */
+
+static int amd_pre_reset(struct ata_port *ap)
+{
+	static const u32 bitmask[2] = {0x03, 0xC0};
+	static const struct pci_bits amd_enable_bits[] = {
+		{ 0x40, 1, 0x02, 0x02 },
+		{ 0x40, 1, 0x01, 0x01 }
+	};
+
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u8 ata66;
+	
+	if (!pci_test_config_bits(pdev, &amd_enable_bits[ap->port_no])) {
+		ata_port_disable(ap);
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+
+	pci_read_config_byte(pdev, 0x42, &ata66);
+	if (ata66 & bitmask[ap->port_no])
+		ap->cbl = ATA_CBL_PATA80;
+	else
+		ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+
+}
+
+static void amd_error_handler(struct ata_port *ap)
+{
+	return ata_bmdma_drive_eh(ap, amd_pre_reset,
+				      ata_std_softreset, NULL,
+				      ata_std_postreset);
+}
+
+static int amd_early_pre_reset(struct ata_port *ap)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	static struct pci_bits amd_enable_bits[] = {
+		{ 0x40, 1, 0x02, 0x02 },
+		{ 0x40, 1, 0x01, 0x01 }
+	};
+
+	if (!pci_test_config_bits(pdev, &amd_enable_bits[ap->port_no])) {
+		ata_port_disable(ap);
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+	/* No host side cable detection */
+	ap->cbl = ATA_CBL_PATA80;
+	return ata_std_prereset(ap);
+
+}
+
+static void amd_early_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, amd_early_pre_reset,
+			       ata_std_softreset, NULL,
+			       ata_std_postreset);
+}
+
+/**
+ *	amd33_set_piomode	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Program the AMD registers for PIO mode.
+ */
+
+static void amd33_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	timing_setup(ap, adev, 0x40, adev->pio_mode, 1);
+}
+
+static void amd66_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	timing_setup(ap, adev, 0x40, adev->pio_mode, 2);
+}
+
+static void amd100_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	timing_setup(ap, adev, 0x40, adev->pio_mode, 3);
+}
+
+static void amd133_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	timing_setup(ap, adev, 0x40, adev->pio_mode, 4);
+}
+
+/**
+ *	amd33_set_dmamode	-	set initial DMA mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Program the MWDMA/UDMA modes for the AMD and Nvidia
+ *	chipset.
+ */
+
+static void amd33_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	timing_setup(ap, adev, 0x40, adev->dma_mode, 1);
+}
+
+static void amd66_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	timing_setup(ap, adev, 0x40, adev->dma_mode, 2);
+}
+
+static void amd100_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	timing_setup(ap, adev, 0x40, adev->dma_mode, 3);
+}
+
+static void amd133_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	timing_setup(ap, adev, 0x40, adev->dma_mode, 4);
+}
+
+
+/**
+ *	nv_probe_init	-	cable detection
+ *	@ap: ATA port
+ *
+ *	Perform cable detection. The BIOS stores this in PCI config
+ *	space for us.
+ */
+
+static int nv_pre_reset(struct ata_port *ap) {
+	static const u8 bitmask[2] = {0x03, 0xC0};
+
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u8 ata66;
+	u16 udma;
+
+	pci_read_config_byte(pdev, 0x52, &ata66);
+	if (ata66 & bitmask[ap->port_no])
+		ap->cbl = ATA_CBL_PATA80;
+	else
+		ap->cbl = ATA_CBL_PATA40;
+
+	/* We now have to double check because the Nvidia boxes BIOS
+	   doesn't always set the cable bits but does set mode bits */
+
+	pci_read_config_word(pdev, 0x62 - 2 * ap->port_no, &udma);
+	if ((udma & 0xC4) == 0xC4 || (udma & 0xC400) == 0xC400)
+		ap->cbl = ATA_CBL_PATA80;
+	return ata_std_prereset(ap);
+}
+
+static void nv_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, nv_pre_reset,
+			       ata_std_softreset, NULL,
+			       ata_std_postreset);
+}
+/**
+ *	nv100_set_piomode	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Program the AMD registers for PIO mode.
+ */
+
+static void nv100_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	timing_setup(ap, adev, 0x50, adev->pio_mode, 3);
+}
+
+static void nv133_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	timing_setup(ap, adev, 0x50, adev->pio_mode, 4);
+}
+
+/**
+ *	nv100_set_dmamode	-	set initial DMA mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Program the MWDMA/UDMA modes for the AMD and Nvidia
+ *	chipset.
+ */
+
+static void nv100_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	timing_setup(ap, adev, 0x50, adev->dma_mode, 3);
+}
+
+static void nv133_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	timing_setup(ap, adev, 0x50, adev->dma_mode, 4);
+}
+
+static struct scsi_host_template amd_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations amd33_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= amd33_set_piomode,
+	.set_dmamode	= amd33_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= amd_early_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static struct ata_port_operations amd66_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= amd66_set_piomode,
+	.set_dmamode	= amd66_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= amd_early_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static struct ata_port_operations amd100_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= amd100_set_piomode,
+	.set_dmamode	= amd100_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= amd_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static struct ata_port_operations amd133_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= amd133_set_piomode,
+	.set_dmamode	= amd133_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= amd_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static struct ata_port_operations nv100_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= nv100_set_piomode,
+	.set_dmamode	= nv100_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= nv_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static struct ata_port_operations nv133_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= nv133_set_piomode,
+	.set_dmamode	= nv133_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= nv_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info[10] = {
+		{	/* 0: AMD 7401 */
+			.sht = &amd_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,	/* No SWDMA */
+			.udma_mask = 0x07,	/* UDMA 33 */
+			.port_ops = &amd33_port_ops
+		},
+		{	/* 1: Early AMD7409 - no swdma */
+			.sht = &amd_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = 0x1f,	/* UDMA 66 */
+			.port_ops = &amd66_port_ops
+		},
+		{	/* 2: AMD 7409, no swdma errata */
+			.sht = &amd_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = 0x1f,	/* UDMA 66 */
+			.port_ops = &amd66_port_ops
+		},
+		{	/* 3: AMD 7411 */
+			.sht = &amd_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = 0x3f,	/* UDMA 100 */
+			.port_ops = &amd100_port_ops
+		},
+		{	/* 4: AMD 7441 */
+			.sht = &amd_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = 0x3f,	/* UDMA 100 */
+			.port_ops = &amd100_port_ops
+		},
+		{	/* 5: AMD 8111*/
+			.sht = &amd_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = 0x7f,	/* UDMA 133, no swdma */
+			.port_ops = &amd133_port_ops
+		},
+		{	/* 6: AMD 8111 UDMA 100 (Serenade) */
+			.sht = &amd_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = 0x3f,	/* UDMA 100, no swdma */
+			.port_ops = &amd133_port_ops
+		},
+		{	/* 7: Nvidia Nforce */
+			.sht = &amd_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = 0x3f,	/* UDMA 100 */
+			.port_ops = &nv100_port_ops
+		},
+		{	/* 8: Nvidia Nforce2 and later */
+			.sht = &amd_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = 0x7f,	/* UDMA 133, no swdma */
+			.port_ops = &nv133_port_ops
+		},
+		{	/* 9: AMD CS5536 (Geode companion) */
+			.sht = &amd_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = 0x3f,	/* UDMA 100 */
+			.port_ops = &amd100_port_ops
+		}
+	};
+	static struct ata_port_info *port_info[2];
+	static int printed_version;
+	int type = id->driver_data;
+	u8 rev;
+	u8 fifo;
+
+	if (!printed_version++)
+		dev_printk(KERN_DEBUG, &pdev->dev, "version " DRV_VERSION "\n");
+
+	pci_read_config_byte(pdev, PCI_REVISION_ID, &rev);
+	pci_read_config_byte(pdev, 0x41, &fifo);
+
+	/* Check for AMD7409 without swdma errata and if found adjust type */
+	if (type == 1 && rev > 0x7)
+		type = 2;
+
+	/* Check for AMD7411 */
+	if (type == 3)
+		/* FIFO is broken */
+		pci_write_config_byte(pdev, 0x41, fifo & 0x0F);
+	else
+		pci_write_config_byte(pdev, 0x41, fifo | 0xF0);
+
+	/* Serenade ? */
+	if (type == 5 && pdev->subsystem_vendor == PCI_VENDOR_ID_AMD &&
+			 pdev->subsystem_device == PCI_DEVICE_ID_AMD_SERENADE)
+		type = 6;	/* UDMA 100 only */
+
+	if (type < 3)
+		ata_pci_clear_simplex(pdev);
+
+	/* And fire it up */
+
+	port_info[0] = port_info[1] = &info[type];
+	return ata_pci_init_one(pdev, port_info, 2);
+}
+
+static const struct pci_device_id amd[] = {
+	{ PCI_VENDOR_ID_AMD,	PCI_DEVICE_ID_AMD_COBRA_7401,		PCI_ANY_ID, PCI_ANY_ID, 0, 0,  0 },
+	{ PCI_VENDOR_ID_AMD,	PCI_DEVICE_ID_AMD_VIPER_7409,		PCI_ANY_ID, PCI_ANY_ID, 0, 0,  1 },
+	{ PCI_VENDOR_ID_AMD,	PCI_DEVICE_ID_AMD_VIPER_7411,		PCI_ANY_ID, PCI_ANY_ID, 0, 0,  3 },
+	{ PCI_VENDOR_ID_AMD,	PCI_DEVICE_ID_AMD_OPUS_7441,		PCI_ANY_ID, PCI_ANY_ID, 0, 0,  4 },
+	{ PCI_VENDOR_ID_AMD,	PCI_DEVICE_ID_AMD_8111_IDE,		PCI_ANY_ID, PCI_ANY_ID, 0, 0,  5 },
+	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_IDE,	PCI_ANY_ID, PCI_ANY_ID, 0, 0,  7 },
+	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE2_IDE,	PCI_ANY_ID, PCI_ANY_ID, 0, 0,  8 },
+	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE2S_IDE,	PCI_ANY_ID, PCI_ANY_ID, 0, 0,  8 },
+	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE3_IDE,	PCI_ANY_ID, PCI_ANY_ID, 0, 0,  8 },
+	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE3S_IDE,	PCI_ANY_ID, PCI_ANY_ID, 0, 0, 8 },
+	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_CK804_IDE,	PCI_ANY_ID, PCI_ANY_ID, 0, 0, 8 },
+	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP04_IDE,	PCI_ANY_ID, PCI_ANY_ID, 0, 0, 8 },
+	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP51_IDE,	PCI_ANY_ID, PCI_ANY_ID, 0, 0, 8 },
+	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_IDE,	PCI_ANY_ID, PCI_ANY_ID, 0, 0, 8 },
+	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP61_IDE,	PCI_ANY_ID, PCI_ANY_ID, 0, 0, 8 },
+	{ PCI_VENDOR_ID_AMD,	PCI_DEVICE_ID_AMD_CS5536_IDE,		PCI_ANY_ID, PCI_ANY_ID, 0, 0, 9 },
+	{ 0, },
+};
+
+static struct pci_driver amd_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= amd,
+	.probe 		= amd_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init amd_init(void)
+{
+	return pci_register_driver(&amd_pci_driver);
+}
+
+static void __exit amd_exit(void)
+{
+	pci_unregister_driver(&amd_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for AMD PATA IDE");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, amd);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(amd_init);
+module_exit(amd_exit);
diff --git a/drivers/ata/pata_artop.c b/drivers/ata/pata_artop.c
new file mode 100644
index 000000000000..d6ef3bf1bac7
--- /dev/null
+++ b/drivers/ata/pata_artop.c
@@ -0,0 +1,518 @@
+/*
+ *    pata_artop.c - ARTOP ATA controller driver
+ *
+ *	(C) 2006 Red Hat <alan@redhat.com>
+ *
+ *    Based in part on drivers/ide/pci/aec62xx.c
+ *	Copyright (C) 1999-2002	Andre Hedrick <andre@linux-ide.org>
+ *	865/865R fixes for Macintosh card version from a patch to the old
+ *		driver by Thibaut VARENE <varenet@parisc-linux.org>
+ *	When setting the PCI latency we must set 0x80 or higher for burst
+ *		performance Alessandro Zummo <alessandro.zummo@towertech.it>
+ *
+ *	TODO
+ *	850 serialization once the core supports it
+ *	Investigate no_dsc on 850R
+ *	Clock detect
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+#include <linux/ata.h>
+
+#define DRV_NAME	"pata_artop"
+#define DRV_VERSION	"0.4.1"
+
+/*
+ *	The ARTOP has 33 Mhz and "over clocked" timing tables. Until we
+ *	get PCI bus speed functionality we leave this as 0. Its a variable
+ *	for when we get the functionality and also for folks wanting to
+ *	test stuff.
+ */
+
+static int clock = 0;
+
+static int artop6210_pre_reset(struct ata_port *ap)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	const struct pci_bits artop_enable_bits[] = {
+		{ 0x4AU, 1U, 0x02UL, 0x02UL },	/* port 0 */
+		{ 0x4AU, 1U, 0x04UL, 0x04UL },	/* port 1 */
+	};
+
+	if (!pci_test_config_bits(pdev, &artop_enable_bits[ap->port_no])) {
+		ata_port_disable(ap);
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	artop6210_error_handler - Probe specified port on PATA host controller
+ *	@ap: Port to probe
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void artop6210_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, artop6210_pre_reset,
+				    ata_std_softreset, NULL,
+				    ata_std_postreset);
+}
+
+/**
+ *	artop6260_pre_reset	-	check for 40/80 pin
+ *	@ap: Port
+ *
+ *	The ARTOP hardware reports the cable detect bits in register 0x49.
+ *	Nothing complicated needed here.
+ */
+
+static int artop6260_pre_reset(struct ata_port *ap)
+{
+	static const struct pci_bits artop_enable_bits[] = {
+		{ 0x4AU, 1U, 0x02UL, 0x02UL },	/* port 0 */
+		{ 0x4AU, 1U, 0x04UL, 0x04UL },	/* port 1 */
+	};
+
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u8 tmp;
+
+	/* Odd numbered device ids are the units with enable bits (the -R cards) */
+	if (pdev->device % 1 && !pci_test_config_bits(pdev, &artop_enable_bits[ap->port_no])) {
+		ata_port_disable(ap);
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+	pci_read_config_byte(pdev, 0x49, &tmp);
+	if (tmp & (1 >> ap->port_no))
+		ap->cbl = ATA_CBL_PATA40;
+	else
+		ap->cbl = ATA_CBL_PATA80;
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	artop6260_error_handler - Probe specified port on PATA host controller
+ *	@ap: Port to probe
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void artop6260_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, artop6260_pre_reset,
+				    ata_std_softreset, NULL,
+				    ata_std_postreset);
+}
+
+/**
+ *	artop6210_load_piomode - Load a set of PATA PIO timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: Device
+ *	@pio: PIO mode
+ *
+ *	Set PIO mode for device, in host controller PCI config space. This
+ *	is used both to set PIO timings in PIO mode and also to set the
+ *	matching PIO clocking for UDMA, as well as the MWDMA timings.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void artop6210_load_piomode(struct ata_port *ap, struct ata_device *adev, unsigned int pio)
+{
+	struct pci_dev *pdev	= to_pci_dev(ap->host->dev);
+	int dn = adev->devno + 2 * ap->port_no;
+	const u16 timing[2][5] = {
+		{ 0x0000, 0x000A, 0x0008, 0x0303, 0x0301 },
+		{ 0x0700, 0x070A, 0x0708, 0x0403, 0x0401 }
+
+	};
+	/* Load the PIO timing active/recovery bits */
+	pci_write_config_word(pdev, 0x40 + 2 * dn, timing[clock][pio]);
+}
+
+/**
+ *	artop6210_set_piomode - Initialize host controller PATA PIO timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: Device we are configuring
+ *
+ *	Set PIO mode for device, in host controller PCI config space. For
+ *	ARTOP we must also clear the UDMA bits if we are not doing UDMA. In
+ *	the event UDMA is used the later call to set_dmamode will set the
+ *	bits as required.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void artop6210_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev	= to_pci_dev(ap->host->dev);
+	int dn = adev->devno + 2 * ap->port_no;
+	u8 ultra;
+
+	artop6210_load_piomode(ap, adev, adev->pio_mode - XFER_PIO_0);
+
+	/* Clear the UDMA mode bits (set_dmamode will redo this if needed) */
+	pci_read_config_byte(pdev, 0x54, &ultra);
+	ultra &= ~(3 << (2 * dn));
+	pci_write_config_byte(pdev, 0x54, ultra);
+}
+
+/**
+ *	artop6260_load_piomode - Initialize host controller PATA PIO timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: Device we are configuring
+ *	@pio: PIO mode
+ *
+ *	Set PIO mode for device, in host controller PCI config space. The
+ *	ARTOP6260 and relatives store the timing data differently.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void artop6260_load_piomode (struct ata_port *ap, struct ata_device *adev, unsigned int pio)
+{
+	struct pci_dev *pdev	= to_pci_dev(ap->host->dev);
+	int dn = adev->devno + 2 * ap->port_no;
+	const u8 timing[2][5] = {
+		{ 0x00, 0x0A, 0x08, 0x33, 0x31 },
+		{ 0x70, 0x7A, 0x78, 0x43, 0x41 }
+
+	};
+	/* Load the PIO timing active/recovery bits */
+	pci_write_config_byte(pdev, 0x40 + dn, timing[clock][pio]);
+}
+
+/**
+ *	artop6260_set_piomode - Initialize host controller PATA PIO timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: Device we are configuring
+ *
+ *	Set PIO mode for device, in host controller PCI config space. For
+ *	ARTOP we must also clear the UDMA bits if we are not doing UDMA. In
+ *	the event UDMA is used the later call to set_dmamode will set the
+ *	bits as required.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void artop6260_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev	= to_pci_dev(ap->host->dev);
+	u8 ultra;
+
+	artop6260_load_piomode(ap, adev, adev->pio_mode - XFER_PIO_0);
+
+	/* Clear the UDMA mode bits (set_dmamode will redo this if needed) */
+	pci_read_config_byte(pdev, 0x44 + ap->port_no, &ultra);
+	ultra &= ~(7 << (4  * adev->devno));	/* One nibble per drive */
+	pci_write_config_byte(pdev, 0x44 + ap->port_no, ultra);
+}
+
+/**
+ *	artop6210_set_dmamode - Initialize host controller PATA PIO timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: um
+ *
+ *	Set DMA mode for device, in host controller PCI config space.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void artop6210_set_dmamode (struct ata_port *ap, struct ata_device *adev)
+{
+	unsigned int pio;
+	struct pci_dev *pdev	= to_pci_dev(ap->host->dev);
+	int dn = adev->devno + 2 * ap->port_no;
+	u8 ultra;
+
+	if (adev->dma_mode == XFER_MW_DMA_0)
+		pio = 1;
+	else
+		pio = 4;
+
+	/* Load the PIO timing active/recovery bits */
+	artop6210_load_piomode(ap, adev, pio);
+
+	pci_read_config_byte(pdev, 0x54, &ultra);
+	ultra &= ~(3 << (2 * dn));
+
+	/* Add ultra DMA bits if in UDMA mode */
+	if (adev->dma_mode >= XFER_UDMA_0) {
+		u8 mode = (adev->dma_mode - XFER_UDMA_0) + 1 - clock;
+		if (mode == 0)
+			mode = 1;
+		ultra |= (mode << (2 * dn));
+	}
+	pci_write_config_byte(pdev, 0x54, ultra);
+}
+
+/**
+ *	artop6260_set_dmamode - Initialize host controller PATA PIO timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: Device we are configuring
+ *
+ *	Set DMA mode for device, in host controller PCI config space. The
+ *	ARTOP6260 and relatives store the timing data differently.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void artop6260_set_dmamode (struct ata_port *ap, struct ata_device *adev)
+{
+	unsigned int pio	= adev->pio_mode - XFER_PIO_0;
+	struct pci_dev *pdev	= to_pci_dev(ap->host->dev);
+	u8 ultra;
+
+	if (adev->dma_mode == XFER_MW_DMA_0)
+		pio = 1;
+	else
+		pio = 4;
+
+	/* Load the PIO timing active/recovery bits */
+	artop6260_load_piomode(ap, adev, pio);
+
+	/* Add ultra DMA bits if in UDMA mode */
+	pci_read_config_byte(pdev, 0x44 + ap->port_no, &ultra);
+	ultra &= ~(7 << (4  * adev->devno));	/* One nibble per drive */
+	if (adev->dma_mode >= XFER_UDMA_0) {
+		u8 mode = adev->dma_mode - XFER_UDMA_0 + 1 - clock;
+		if (mode == 0)
+			mode = 1;
+		ultra |= (mode << (4 * adev->devno));
+	}
+	pci_write_config_byte(pdev, 0x44 + ap->port_no, ultra);
+}
+
+static struct scsi_host_template artop_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static const struct ata_port_operations artop6210_ops = {
+	.port_disable		= ata_port_disable,
+	.set_piomode		= artop6210_set_piomode,
+	.set_dmamode		= artop6210_set_dmamode,
+	.mode_filter		= ata_pci_default_filter,
+
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.check_status		= ata_check_status,
+	.exec_command		= ata_exec_command,
+	.dev_select		= ata_std_dev_select,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= artop6210_error_handler,
+	.post_internal_cmd 	= ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= ata_qc_issue_prot,
+	.eng_timeout		= ata_eng_timeout,
+	.data_xfer		= ata_pio_data_xfer,
+
+	.irq_handler		= ata_interrupt,
+	.irq_clear		= ata_bmdma_irq_clear,
+
+	.port_start		= ata_port_start,
+	.port_stop		= ata_port_stop,
+	.host_stop		= ata_host_stop,
+};
+
+static const struct ata_port_operations artop6260_ops = {
+	.port_disable		= ata_port_disable,
+	.set_piomode		= artop6260_set_piomode,
+	.set_dmamode		= artop6260_set_dmamode,
+
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.check_status		= ata_check_status,
+	.exec_command		= ata_exec_command,
+	.dev_select		= ata_std_dev_select,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= artop6260_error_handler,
+	.post_internal_cmd 	= ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= ata_qc_issue_prot,
+	.data_xfer		= ata_pio_data_xfer,
+
+	.eng_timeout		= ata_eng_timeout,
+
+	.irq_handler		= ata_interrupt,
+	.irq_clear		= ata_bmdma_irq_clear,
+
+	.port_start		= ata_port_start,
+	.port_stop		= ata_port_stop,
+	.host_stop		= ata_host_stop,
+};
+
+
+/**
+ *	artop_init_one - Register ARTOP ATA PCI device with kernel services
+ *	@pdev: PCI device to register
+ *	@ent: Entry in artop_pci_tbl matching with @pdev
+ *
+ *	Called from kernel PCI layer.
+ *
+ *	LOCKING:
+ *	Inherited from PCI layer (may sleep).
+ *
+ *	RETURNS:
+ *	Zero on success, or -ERRNO value.
+ */
+
+static int artop_init_one (struct pci_dev *pdev, const struct pci_device_id *id)
+{
+	static int printed_version;
+	static struct ata_port_info info_6210 = {
+		.sht		= &artop_sht,
+		.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask	= 0x1f,	/* pio0-4 */
+		.mwdma_mask	= 0x07, /* mwdma0-2 */
+		.udma_mask 	= ATA_UDMA2,
+		.port_ops	= &artop6210_ops,
+	};
+	static struct ata_port_info info_626x = {
+		.sht		= &artop_sht,
+		.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask	= 0x1f,	/* pio0-4 */
+		.mwdma_mask	= 0x07, /* mwdma0-2 */
+		.udma_mask 	= ATA_UDMA4,
+		.port_ops	= &artop6260_ops,
+	};
+	static struct ata_port_info info_626x_fast = {
+		.sht		= &artop_sht,
+		.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask	= 0x1f,	/* pio0-4 */
+		.mwdma_mask	= 0x07, /* mwdma0-2 */
+		.udma_mask 	= ATA_UDMA5,
+		.port_ops	= &artop6260_ops,
+	};
+	struct ata_port_info *port_info[2];
+	struct ata_port_info *info;
+	int ports = 2;
+
+	if (!printed_version++)
+		dev_printk(KERN_DEBUG, &pdev->dev,
+			   "version " DRV_VERSION "\n");
+
+	if (id->driver_data == 0) {	/* 6210 variant */
+		info = &info_6210;
+		/* BIOS may have left us in UDMA, clear it before libata probe */
+		pci_write_config_byte(pdev, 0x54, 0);
+		/* For the moment (also lacks dsc) */
+		printk(KERN_WARNING "ARTOP 6210 requires serialize functionality not yet supported by libata.\n");
+		printk(KERN_WARNING "Secondary ATA ports will not be activated.\n");
+		ports = 1;
+	}
+	else if (id->driver_data == 1)	/* 6260 */
+		info = &info_626x;
+	else if (id->driver_data == 2)	{ /* 6260 or 6260 + fast */
+		unsigned long io = pci_resource_start(pdev, 4);
+		u8 reg;
+
+		info = &info_626x;
+		if (inb(io) & 0x10)
+			info = &info_626x_fast;
+		/* Mac systems come up with some registers not set as we
+		   will need them */
+
+		/* Clear reset & test bits */
+		pci_read_config_byte(pdev, 0x49, &reg);
+		pci_write_config_byte(pdev, 0x49, reg & ~ 0x30);
+
+		/* PCI latency must be > 0x80 for burst mode, tweak it
+		 * if required.
+		 */
+		pci_read_config_byte(pdev, PCI_LATENCY_TIMER, &reg);
+		if (reg <= 0x80)
+			pci_write_config_byte(pdev, PCI_LATENCY_TIMER, 0x90);
+
+		/* Enable IRQ output and burst mode */
+		pci_read_config_byte(pdev, 0x4a, &reg);
+		pci_write_config_byte(pdev, 0x4a, (reg & ~0x01) | 0x80);
+
+	}
+	port_info[0] = port_info[1] = info;
+	return ata_pci_init_one(pdev, port_info, ports);
+}
+
+static const struct pci_device_id artop_pci_tbl[] = {
+	{ 0x1191, 0x0005, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0},
+	{ 0x1191, 0x0006, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 1},
+	{ 0x1191, 0x0007, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 1},
+	{ 0x1191, 0x0008, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 2},
+	{ 0x1191, 0x0009, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 2},
+	{ }	/* terminate list */
+};
+
+static struct pci_driver artop_pci_driver = {
+	.name			= DRV_NAME,
+	.id_table		= artop_pci_tbl,
+	.probe			= artop_init_one,
+	.remove			= ata_pci_remove_one,
+};
+
+static int __init artop_init(void)
+{
+	return pci_register_driver(&artop_pci_driver);
+}
+
+static void __exit artop_exit(void)
+{
+	pci_unregister_driver(&artop_pci_driver);
+}
+
+
+module_init(artop_init);
+module_exit(artop_exit);
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("SCSI low-level driver for ARTOP PATA");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, artop_pci_tbl);
+MODULE_VERSION(DRV_VERSION);
+
diff --git a/drivers/ata/pata_atiixp.c b/drivers/ata/pata_atiixp.c
new file mode 100644
index 000000000000..3f78a1e54a75
--- /dev/null
+++ b/drivers/ata/pata_atiixp.c
@@ -0,0 +1,306 @@
+/*
+ * pata_atiixp.c 	- ATI PATA for new ATA layer
+ *			  (C) 2005 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ * Based on
+ *
+ *  linux/drivers/ide/pci/atiixp.c	Version 0.01-bart2	Feb. 26, 2004
+ *
+ *  Copyright (C) 2003 ATI Inc. <hyu@ati.com>
+ *  Copyright (C) 2004 Bartlomiej Zolnierkiewicz
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "pata_atiixp"
+#define DRV_VERSION "0.4.2"
+
+enum {
+	ATIIXP_IDE_PIO_TIMING	= 0x40,
+	ATIIXP_IDE_MWDMA_TIMING	= 0x44,
+	ATIIXP_IDE_PIO_CONTROL	= 0x48,
+	ATIIXP_IDE_PIO_MODE	= 0x4a,
+	ATIIXP_IDE_UDMA_CONTROL	= 0x54,
+	ATIIXP_IDE_UDMA_MODE 	= 0x56
+};
+
+static int atiixp_pre_reset(struct ata_port *ap)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	static struct pci_bits atiixp_enable_bits[] = {
+		{ 0x48, 1, 0x01, 0x00 },
+		{ 0x48, 1, 0x08, 0x00 }
+	};
+
+	if (!pci_test_config_bits(pdev, &atiixp_enable_bits[ap->port_no])) {
+		ata_port_disable(ap);
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+	ap->cbl = ATA_CBL_PATA80;
+	return ata_std_prereset(ap);
+}
+
+static void atiixp_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, atiixp_pre_reset, ata_std_softreset, NULL,   ata_std_postreset);
+}
+
+/**
+ *	atiixp_set_pio_timing	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Called by both the pio and dma setup functions to set the controller
+ *	timings for PIO transfers. We must load both the mode number and
+ *	timing values into the controller.
+ */
+
+static void atiixp_set_pio_timing(struct ata_port *ap, struct ata_device *adev, int pio)
+{
+	static u8 pio_timings[5] = { 0x5D, 0x47, 0x34, 0x22, 0x20 };
+
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	int dn = 2 * ap->port_no + adev->devno;
+
+	/* Check this is correct - the order is odd in both drivers */
+	int timing_shift = (16 * ap->port_no) + 8 * (adev->devno ^ 1);
+	u16 pio_mode_data, pio_timing_data;
+
+	pci_read_config_word(pdev, ATIIXP_IDE_PIO_MODE, &pio_mode_data);
+	pio_mode_data &= ~(0x7 << (4 * dn));
+	pio_mode_data |= pio << (4 * dn);
+	pci_write_config_word(pdev, ATIIXP_IDE_PIO_MODE, pio_mode_data);
+
+	pci_read_config_word(pdev, ATIIXP_IDE_PIO_TIMING, &pio_timing_data);
+	pio_mode_data &= ~(0xFF << timing_shift);
+	pio_mode_data |= (pio_timings[pio] << timing_shift);
+	pci_write_config_word(pdev, ATIIXP_IDE_PIO_TIMING, pio_timing_data);
+}
+
+/**
+ *	atiixp_set_piomode	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Called to do the PIO mode setup. We use a shared helper for this
+ *	as the DMA setup must also adjust the PIO timing information.
+ */
+
+static void atiixp_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	atiixp_set_pio_timing(ap, adev, adev->pio_mode - XFER_PIO_0);
+}
+
+/**
+ *	atiixp_set_dmamode	-	set initial DMA mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Called to do the DMA mode setup. We use timing tables for most
+ *	modes but must tune an appropriate PIO mode to match.
+ */
+
+static void atiixp_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	static u8 mwdma_timings[5] = { 0x77, 0x21, 0x20 };
+
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	int dma = adev->dma_mode;
+	int dn = 2 * ap->port_no + adev->devno;
+	int wanted_pio;
+
+	if (adev->dma_mode >= XFER_UDMA_0) {
+		u16 udma_mode_data;
+
+		dma -= XFER_UDMA_0;
+
+		pci_read_config_word(pdev, ATIIXP_IDE_UDMA_MODE, &udma_mode_data);
+		udma_mode_data &= ~(0x7 << (4 * dn));
+		udma_mode_data |= dma << (4 * dn);
+		pci_write_config_word(pdev, ATIIXP_IDE_UDMA_MODE, udma_mode_data);
+	} else {
+		u16 mwdma_timing_data;
+		/* Check this is correct - the order is odd in both drivers */
+		int timing_shift = (16 * ap->port_no) + 8 * (adev->devno ^ 1);
+
+		dma -= XFER_MW_DMA_0;
+
+		pci_read_config_word(pdev, ATIIXP_IDE_MWDMA_TIMING, &mwdma_timing_data);
+		mwdma_timing_data &= ~(0xFF << timing_shift);
+		mwdma_timing_data |= (mwdma_timings[dma] << timing_shift);
+		pci_write_config_word(pdev, ATIIXP_IDE_MWDMA_TIMING, mwdma_timing_data);
+	}
+	/*
+	 *	We must now look at the PIO mode situation. We may need to
+	 *	adjust the PIO mode to keep the timings acceptable
+	 */
+	 if (adev->dma_mode >= XFER_MW_DMA_2)
+	 	wanted_pio = 4;
+	else if (adev->dma_mode == XFER_MW_DMA_1)
+		wanted_pio = 3;
+	else if (adev->dma_mode == XFER_MW_DMA_0)
+		wanted_pio = 0;
+	else BUG();
+
+	if (adev->pio_mode != wanted_pio)
+		atiixp_set_pio_timing(ap, adev, wanted_pio);
+}
+
+/**
+ *	atiixp_bmdma_start	-	DMA start callback
+ *	@qc: Command in progress
+ *
+ *	When DMA begins we need to ensure that the UDMA control
+ *	register for the channel is correctly set.
+ */
+
+static void atiixp_bmdma_start(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct ata_device *adev = qc->dev;
+
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	int dn = (2 * ap->port_no) + adev->devno;
+	u16 tmp16;
+
+	pci_read_config_word(pdev, ATIIXP_IDE_UDMA_CONTROL, &tmp16);
+	if (adev->dma_mode >= XFER_UDMA_0)
+		tmp16 |= (1 << dn);
+	else
+		tmp16 &= ~(1 << dn);
+	pci_write_config_word(pdev, ATIIXP_IDE_UDMA_CONTROL, tmp16);
+	ata_bmdma_start(qc);
+}
+
+/**
+ *	atiixp_dma_stop	-	DMA stop callback
+ *	@qc: Command in progress
+ *
+ *	DMA has completed. Clear the UDMA flag as the next operations will
+ *	be PIO ones not UDMA data transfer.
+ */
+
+static void atiixp_bmdma_stop(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	int dn = (2 * ap->port_no) + qc->dev->devno;
+	u16 tmp16;
+
+	pci_read_config_word(pdev, ATIIXP_IDE_UDMA_CONTROL, &tmp16);
+	tmp16 &= ~(1 << dn);
+	pci_write_config_word(pdev, ATIIXP_IDE_UDMA_CONTROL, tmp16);
+	ata_bmdma_stop(qc);
+}
+
+static struct scsi_host_template atiixp_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations atiixp_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= atiixp_set_piomode,
+	.set_dmamode	= atiixp_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= atiixp_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= atiixp_bmdma_start,
+	.bmdma_stop	= atiixp_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static int atiixp_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info = {
+		.sht = &atiixp_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x06,	/* No MWDMA0 support */
+		.udma_mask = 0x3F,
+		.port_ops = &atiixp_port_ops
+	};
+	static struct ata_port_info *port_info[2] = { &info, &info };
+	return ata_pci_init_one(dev, port_info, 2);
+}
+
+static struct pci_device_id atiixp[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_ATI, PCI_DEVICE_ID_ATI_IXP200_IDE), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_ATI, PCI_DEVICE_ID_ATI_IXP300_IDE), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_ATI, PCI_DEVICE_ID_ATI_IXP400_IDE), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_ATI, PCI_DEVICE_ID_ATI_IXP600_IDE), },
+	{ 0, },
+};
+
+static struct pci_driver atiixp_pci_driver = {
+	.name 		= DRV_NAME,
+	.id_table	= atiixp,
+	.probe 		= atiixp_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init atiixp_init(void)
+{
+	return pci_register_driver(&atiixp_pci_driver);
+}
+
+
+static void __exit atiixp_exit(void)
+{
+	pci_unregister_driver(&atiixp_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for ATI IXP200/300/400");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, atiixp);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(atiixp_init);
+module_exit(atiixp_exit);
diff --git a/drivers/ata/pata_cmd64x.c b/drivers/ata/pata_cmd64x.c
new file mode 100644
index 000000000000..03284611a727
--- /dev/null
+++ b/drivers/ata/pata_cmd64x.c
@@ -0,0 +1,505 @@
+/*
+ * pata_cmd64x.c 	- ATI PATA for new ATA layer
+ *			  (C) 2005 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ * Based upon
+ * linux/drivers/ide/pci/cmd64x.c		Version 1.30	Sept 10, 2002
+ *
+ * cmd64x.c: Enable interrupts at initialization time on Ultra/PCI machines.
+ *           Note, this driver is not used at all on other systems because
+ *           there the "BIOS" has done all of the following already.
+ *           Due to massive hardware bugs, UltraDMA is only supported
+ *           on the 646U2 and not on the 646U.
+ *
+ * Copyright (C) 1998		Eddie C. Dost  (ecd@skynet.be)
+ * Copyright (C) 1998		David S. Miller (davem@redhat.com)
+ *
+ * Copyright (C) 1999-2002	Andre Hedrick <andre@linux-ide.org>
+ *
+ * TODO
+ *	Testing work
+ */
+ 
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "pata_cmd64x"
+#define DRV_VERSION "0.2.1"
+
+/*
+ * CMD64x specific registers definition.
+ */
+ 
+enum {
+	CFR 		= 0x50,
+		CFR_INTR_CH0  = 0x02,
+	CNTRL 		= 0x51,
+		CNTRL_DIS_RA0 = 0x40,
+		CNTRL_DIS_RA1 = 0x80,
+		CNTRL_ENA_2ND = 0x08,
+	CMDTIM 		= 0x52,
+	ARTTIM0 	= 0x53,
+	DRWTIM0 	= 0x54,
+	ARTTIM1 	= 0x55,
+	DRWTIM1 	= 0x56,
+	ARTTIM23 	= 0x57,
+		ARTTIM23_DIS_RA2  = 0x04,
+		ARTTIM23_DIS_RA3  = 0x08,
+		ARTTIM23_INTR_CH1 = 0x10,
+	ARTTIM2 	= 0x57,
+	ARTTIM3 	= 0x57,
+	DRWTIM23	= 0x58,
+	DRWTIM2 	= 0x58,
+	BRST 		= 0x59,
+	DRWTIM3 	= 0x5b,
+	BMIDECR0	= 0x70,
+	MRDMODE		= 0x71,
+		MRDMODE_INTR_CH0 = 0x04,
+		MRDMODE_INTR_CH1 = 0x08,
+		MRDMODE_BLK_CH0  = 0x10,
+		MRDMODE_BLK_CH1	 = 0x20,
+	BMIDESR0	= 0x72,
+	UDIDETCR0	= 0x73,
+	DTPR0		= 0x74,
+	BMIDECR1	= 0x78,
+	BMIDECSR	= 0x79,
+	BMIDESR1	= 0x7A,
+	UDIDETCR1	= 0x7B,
+	DTPR1		= 0x7C
+};
+
+static int cmd64x_pre_reset(struct ata_port *ap)
+{
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+static int cmd648_pre_reset(struct ata_port *ap)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u8 r;
+
+	/* Check cable detect bits */
+	pci_read_config_byte(pdev, BMIDECSR, &r);
+	if (r & (1 << ap->port_no))
+		ap->cbl = ATA_CBL_PATA80;
+	else	
+		ap->cbl = ATA_CBL_PATA40;
+		
+	return ata_std_prereset(ap);
+}
+
+static void cmd64x_error_handler(struct ata_port *ap)
+{
+	return ata_bmdma_drive_eh(ap, cmd64x_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+static void cmd648_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, cmd648_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	cmd64x_set_piomode	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Called to do the PIO mode setup.
+ */
+ 
+static void cmd64x_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	struct ata_timing t;
+	const unsigned long T = 1000000 / 33;
+	const u8 setup_data[] = { 0x40, 0x40, 0x40, 0x80, 0x00 };
+	
+	u8 reg;
+	
+	/* Port layout is not logical so use a table */
+	const u8 arttim_port[2][2] = { 
+		{ ARTTIM0, ARTTIM1 },
+		{ ARTTIM23, ARTTIM23 }
+	};
+	const u8 drwtim_port[2][2] = {
+		{ DRWTIM0, DRWTIM1 },
+		{ DRWTIM2, DRWTIM3 }
+	};
+	
+	int arttim = arttim_port[ap->port_no][adev->devno];
+	int drwtim = drwtim_port[ap->port_no][adev->devno];
+	
+	
+	if (ata_timing_compute(adev, adev->pio_mode, &t, T, 0) < 0) {
+		printk(KERN_ERR DRV_NAME ": mode computation failed.\n");
+		return;
+	}
+	if (ap->port_no) {
+		/* Slave has shared address setup */
+		struct ata_device *pair = ata_dev_pair(adev);
+		
+		if (pair) {
+			struct ata_timing tp;
+			ata_timing_compute(pair, pair->pio_mode, &tp, T, 0);
+			ata_timing_merge(&t, &tp, &t, ATA_TIMING_SETUP);
+		}
+	}
+	
+	printk(KERN_DEBUG DRV_NAME ": active %d recovery %d setup %d.\n",
+		t.active, t.recover, t.setup);
+	if (t.recover > 16) {
+		t.active += t.recover - 16;
+		t.recover = 16;
+	}
+	if (t.active > 16)
+		t.active = 16;
+		
+	/* Now convert the clocks into values we can actually stuff into
+	   the chip */
+	   
+	if (t.recover > 1)
+		t.recover--;
+	else
+		t.recover = 15;
+	
+	if (t.setup > 4)
+		t.setup = 0xC0;
+	else
+		t.setup = setup_data[t.setup];
+		
+	t.active &= 0x0F;	/* 0 = 16 */
+	
+	/* Load setup timing */
+	pci_read_config_byte(pdev, arttim, &reg);
+	reg &= 0x3F;
+	reg |= t.setup;
+	pci_write_config_byte(pdev, arttim, reg);
+	
+	/* Load active/recovery */
+	pci_write_config_byte(pdev, drwtim, (t.active << 4) | t.recover);	
+}
+
+/**
+ *	cmd64x_set_dmamode	-	set initial DMA mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Called to do the DMA mode setup.
+ */
+ 
+static void cmd64x_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	static const u8 udma_data[] = {
+		0x31, 0x21, 0x11, 0x25, 0x15, 0x05
+	};
+	static const u8 mwdma_data[] = { 
+		0x30, 0x20, 0x10
+	};
+	
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u8 regU, regD;
+
+	int pciU = UDIDETCR0 + 8 * ap->port_no;
+	int pciD = BMIDESR0 + 8 * ap->port_no;
+	int shift = 2 * adev->devno;
+	
+	pci_read_config_byte(pdev, pciD, &regD);
+	pci_read_config_byte(pdev, pciU, &regU);
+
+	regD &= ~(0x20 << shift);
+	regU &= ~(0x35 << shift);
+	
+	if (adev->dma_mode >= XFER_UDMA_0)
+		regU |= udma_data[adev->dma_mode - XFER_UDMA_0] << shift;
+	else
+		regD |= mwdma_data[adev->dma_mode - XFER_MW_DMA_0] << shift;
+
+	regD |= 0x20 << adev->devno;
+
+	pci_write_config_byte(pdev, pciU, regU);
+	pci_write_config_byte(pdev, pciD, regD);
+}
+
+/**
+ *	cmd648_dma_stop	-	DMA stop callback
+ *	@qc: Command in progress
+ *
+ *	DMA has completed.
+ */
+
+static void cmd648_bmdma_stop(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u8 dma_intr;
+	int dma_reg = ap->port_no ? ARTTIM23_INTR_CH1 : CFR_INTR_CH0;
+	int dma_mask = ap->port_no ? ARTTIM2 : CFR;
+	
+	ata_bmdma_stop(qc);
+	
+	pci_read_config_byte(pdev, dma_reg, &dma_intr);
+	pci_write_config_byte(pdev, dma_reg, dma_intr | dma_mask);
+}
+		
+/**
+ *	cmd646r1_dma_stop	-	DMA stop callback
+ *	@qc: Command in progress
+ *
+ *	Stub for now while investigating the r1 quirk in the old driver.
+ */
+
+static void cmd646r1_bmdma_stop(struct ata_queued_cmd *qc)
+{
+	ata_bmdma_stop(qc);
+}
+		
+static struct scsi_host_template cmd64x_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations cmd64x_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= cmd64x_set_piomode,
+	.set_dmamode	= cmd64x_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= cmd64x_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+	
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+static struct ata_port_operations cmd646r1_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= cmd64x_set_piomode,
+	.set_dmamode	= cmd64x_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= cmd64x_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= cmd646r1_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+	
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+static struct ata_port_operations cmd648_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= cmd64x_set_piomode,
+	.set_dmamode	= cmd64x_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= cmd648_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= cmd648_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+	
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+	
+static int cmd64x_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+	u32 class_rev;
+	
+	static struct ata_port_info cmd_info[6] = {
+		{	/* CMD 643 - no UDMA */
+			.sht = &cmd64x_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.port_ops = &cmd64x_port_ops
+		},
+		{	/* CMD 646 with broken UDMA */
+			.sht = &cmd64x_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.port_ops = &cmd64x_port_ops
+		},
+		{	/* CMD 646 with working UDMA */
+			.sht = &cmd64x_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = ATA_UDMA1,
+			.port_ops = &cmd64x_port_ops
+		},
+		{	/* CMD 646 rev 1  */
+			.sht = &cmd64x_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.port_ops = &cmd646r1_port_ops
+		},
+		{	/* CMD 648 */
+			.sht = &cmd64x_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = ATA_UDMA2,
+			.port_ops = &cmd648_port_ops
+		},
+		{	/* CMD 649 */
+			.sht = &cmd64x_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = ATA_UDMA3,
+			.port_ops = &cmd648_port_ops
+		}
+	};
+	static struct ata_port_info *port_info[2], *info;
+	u8 mrdmode;
+	
+	info = &cmd_info[id->driver_data];
+	
+	pci_read_config_dword(pdev, PCI_CLASS_REVISION, &class_rev);
+	class_rev &= 0xFF;
+	
+	if (id->driver_data == 0)	/* 643 */
+		ata_pci_clear_simplex(pdev);
+	
+	if (pdev->device == PCI_DEVICE_ID_CMD_646) {
+		/* Does UDMA work ? */
+		if (class_rev > 4)
+			info = &cmd_info[2];
+		/* Early rev with other problems ? */
+		else if (class_rev == 1)
+			info = &cmd_info[3];
+	}
+
+	pci_write_config_byte(pdev, PCI_LATENCY_TIMER, 64);
+	pci_read_config_byte(pdev, MRDMODE, &mrdmode);
+	mrdmode &= ~ 0x30;	/* IRQ set up */
+	mrdmode |= 0x02;	/* Memory read line enable */
+	pci_write_config_byte(pdev, MRDMODE, mrdmode);
+	
+	/* Force PIO 0 here.. */
+	
+	/* PPC specific fixup copied from old driver */
+#ifdef CONFIG_PPC
+	pci_write_config_byte(pdev, UDIDETCR0, 0xF0);
+#endif
+	
+	port_info[0] = port_info[1] = info;
+	return ata_pci_init_one(pdev, port_info, 2);
+}
+
+static struct pci_device_id cmd64x[] = {
+	{ PCI_VENDOR_ID_CMD, PCI_DEVICE_ID_CMD_643, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0},
+	{ PCI_VENDOR_ID_CMD, PCI_DEVICE_ID_CMD_646, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 1},
+	{ PCI_VENDOR_ID_CMD, PCI_DEVICE_ID_CMD_648, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 4},
+	{ PCI_VENDOR_ID_CMD, PCI_DEVICE_ID_CMD_649, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 5},
+	{ 0, },
+};
+
+static struct pci_driver cmd64x_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= cmd64x,
+	.probe 		= cmd64x_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init cmd64x_init(void)
+{
+	return pci_register_driver(&cmd64x_pci_driver);
+}
+
+
+static void __exit cmd64x_exit(void)
+{
+	pci_unregister_driver(&cmd64x_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for CMD64x series PATA controllers");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, cmd64x);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(cmd64x_init);
+module_exit(cmd64x_exit);
diff --git a/drivers/ata/pata_cs5520.c b/drivers/ata/pata_cs5520.c
new file mode 100644
index 000000000000..792ce4828510
--- /dev/null
+++ b/drivers/ata/pata_cs5520.c
@@ -0,0 +1,336 @@
+/*
+ *	IDE tuning and bus mastering support for the CS5510/CS5520
+ *	chipsets
+ *
+ *	The CS5510/CS5520 are slightly unusual devices. Unlike the
+ *	typical IDE controllers they do bus mastering with the drive in
+ *	PIO mode and smarter silicon.
+ *
+ *	The practical upshot of this is that we must always tune the
+ *	drive for the right PIO mode. We must also ignore all the blacklists
+ *	and the drive bus mastering DMA information. Also to confuse matters
+ *	further we can do DMA on PIO only drives.
+ *
+ *	DMA on the 5510 also requires we disable_hlt() during DMA on early
+ *	revisions.
+ *
+ *	*** This driver is strictly experimental ***
+ *
+ *	(c) Copyright Red Hat Inc 2002
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * Documentation:
+ *	Not publically available.
+ */
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME	"pata_cs5520"
+#define DRV_VERSION	"0.6.2"
+
+struct pio_clocks
+{
+	int address;
+	int assert;
+	int recovery;
+};
+
+static const struct pio_clocks cs5520_pio_clocks[]={
+	{3, 6, 11},
+	{2, 5, 6},
+	{1, 4, 3},
+	{1, 3, 2},
+	{1, 2, 1}
+};
+
+/**
+ *	cs5520_set_timings	-	program PIO timings
+ *	@ap: ATA port
+ *	@adev: ATA device
+ *
+ *	Program the PIO mode timings for the controller according to the pio
+ *	clocking table.
+ */
+
+static void cs5520_set_timings(struct ata_port *ap, struct ata_device *adev, int pio)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	int slave = adev->devno;
+
+	pio -= XFER_PIO_0;
+
+	/* Channel command timing */
+	pci_write_config_byte(pdev, 0x62 + ap->port_no,
+				(cs5520_pio_clocks[pio].recovery << 4) |
+				(cs5520_pio_clocks[pio].assert));
+	/* FIXME: should these use address ? */
+	/* Read command timing */
+	pci_write_config_byte(pdev, 0x64 +  4*ap->port_no + slave,
+				(cs5520_pio_clocks[pio].recovery << 4) |
+				(cs5520_pio_clocks[pio].assert));
+	/* Write command timing */
+	pci_write_config_byte(pdev, 0x66 +  4*ap->port_no + slave,
+				(cs5520_pio_clocks[pio].recovery << 4) |
+				(cs5520_pio_clocks[pio].assert));
+}
+
+/**
+ *	cs5520_enable_dma	-	turn on DMA bits
+ *
+ *	Turn on the DMA bits for this disk. Needed because the BIOS probably
+ *	has not done the work for us. Belongs in the core SATA code.
+ */
+
+static void cs5520_enable_dma(struct ata_port *ap, struct ata_device *adev)
+{
+	/* Set the DMA enable/disable flag */
+	u8 reg = inb(ap->ioaddr.bmdma_addr + 0x02);
+	reg |= 1<<(adev->devno + 5);
+	outb(reg, ap->ioaddr.bmdma_addr + 0x02);
+}
+
+/**
+ *	cs5520_set_dmamode	-	program DMA timings
+ *	@ap: ATA port
+ *	@adev: ATA device
+ *
+ *	Program the DMA mode timings for the controller according to the pio
+ *	clocking table. Note that this device sets the DMA timings to PIO
+ *	mode values. This may seem bizarre but the 5520 architecture talks
+ *	PIO mode to the disk and DMA mode to the controller so the underlying
+ *	transfers are PIO timed.
+ */
+
+static void cs5520_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	static const int dma_xlate[3] = { XFER_PIO_0, XFER_PIO_3, XFER_PIO_4 };
+	cs5520_set_timings(ap, adev, dma_xlate[adev->dma_mode]);
+	cs5520_enable_dma(ap, adev);
+}
+
+/**
+ *	cs5520_set_piomode	-	program PIO timings
+ *	@ap: ATA port
+ *	@adev: ATA device
+ *
+ *	Program the PIO mode timings for the controller according to the pio
+ *	clocking table. We know pio_mode will equal dma_mode because of the
+ *	CS5520 architecture. At least once we turned DMA on and wrote a
+ *	mode setter.
+ */
+
+static void cs5520_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	cs5520_set_timings(ap, adev, adev->pio_mode);
+}
+
+
+static int cs5520_pre_reset(struct ata_port *ap)
+{
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+static void cs5520_error_handler(struct ata_port *ap)
+{
+	return ata_bmdma_drive_eh(ap, cs5520_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+static struct scsi_host_template cs5520_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations cs5520_port_ops = {
+	.port_disable		= ata_port_disable,
+	.set_piomode		= cs5520_set_piomode,
+	.set_dmamode		= cs5520_set_dmamode,
+
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.check_status		= ata_check_status,
+	.exec_command		= ata_exec_command,
+	.dev_select		= ata_std_dev_select,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= cs5520_error_handler,
+	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= ata_qc_issue_prot,
+	.data_xfer		= ata_pio_data_xfer,
+
+	.eng_timeout		= ata_eng_timeout,
+
+	.irq_handler		= ata_interrupt,
+	.irq_clear		= ata_bmdma_irq_clear,
+
+	.port_start		= ata_port_start,
+	.port_stop		= ata_port_stop,
+	.host_stop		= ata_host_stop,
+};
+
+static int __devinit cs5520_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	u8 pcicfg;
+	static struct ata_probe_ent probe[2];
+	int ports = 0;
+
+	/* IDE port enable bits */
+	pci_read_config_byte(dev, 0x60, &pcicfg);
+
+	/* Check if the ATA ports are enabled */
+	if ((pcicfg & 3) == 0)
+		return -ENODEV;
+
+	if ((pcicfg & 0x40) == 0) {
+		printk(KERN_WARNING DRV_NAME ": DMA mode disabled. Enabling.\n");
+		pci_write_config_byte(dev, 0x60, pcicfg | 0x40);
+	}
+
+	/* Perform set up for DMA */
+	if (pci_enable_device_bars(dev, 1<<2)) {
+		printk(KERN_ERR DRV_NAME ": unable to configure BAR2.\n");
+		return -ENODEV;
+	}
+	pci_set_master(dev);
+	if (pci_set_dma_mask(dev, DMA_32BIT_MASK)) {
+		printk(KERN_ERR DRV_NAME ": unable to configure DMA mask.\n");
+		return -ENODEV;
+	}
+	if (pci_set_consistent_dma_mask(dev, DMA_32BIT_MASK)) {
+		printk(KERN_ERR DRV_NAME ": unable to configure consistent DMA mask.\n");
+		return -ENODEV;
+	}
+
+	/* We have to do our own plumbing as the PCI setup for this
+	   chipset is non-standard so we can't punt to the libata code */
+
+	INIT_LIST_HEAD(&probe[0].node);
+	probe[0].dev = pci_dev_to_dev(dev);
+	probe[0].port_ops = &cs5520_port_ops;
+	probe[0].sht = &cs5520_sht;
+	probe[0].pio_mask = 0x1F;
+	probe[0].mwdma_mask = id->driver_data;
+	probe[0].irq = 14;
+	probe[0].irq_flags = 0;
+	probe[0].port_flags = ATA_FLAG_SLAVE_POSS|ATA_FLAG_SRST;
+	probe[0].n_ports = 1;
+	probe[0].port[0].cmd_addr = 0x1F0;
+	probe[0].port[0].ctl_addr = 0x3F6;
+	probe[0].port[0].altstatus_addr = 0x3F6;
+	probe[0].port[0].bmdma_addr = pci_resource_start(dev, 2);
+
+	/* The secondary lurks at different addresses but is otherwise
+	   the same beastie */
+
+	probe[1] = probe[0];
+	INIT_LIST_HEAD(&probe[1].node);
+	probe[1].irq = 15;
+	probe[1].port[0].cmd_addr = 0x170;
+	probe[1].port[0].ctl_addr = 0x376;
+	probe[1].port[0].altstatus_addr = 0x376;
+	probe[1].port[0].bmdma_addr = pci_resource_start(dev, 2) + 8;
+
+	/* Let libata fill in the port details */
+	ata_std_ports(&probe[0].port[0]);
+	ata_std_ports(&probe[1].port[0]);
+
+	/* Now add the ports that are active */
+	if (pcicfg & 1)
+		ports += ata_device_add(&probe[0]);
+	if (pcicfg & 2)
+		ports += ata_device_add(&probe[1]);
+	if (ports)
+		return 0;
+	return -ENODEV;
+}
+
+/**
+ *	cs5520_remove_one	-	device unload
+ *	@pdev: PCI device being removed
+ *
+ *	Handle an unplug/unload event for a PCI device. Unload the
+ *	PCI driver but do not use the default handler as we manage
+ *	resources ourself and *MUST NOT* disable the device as it has
+ *	other functions.
+ */
+
+static void __devexit cs5520_remove_one(struct pci_dev *pdev)
+{
+	struct device *dev = pci_dev_to_dev(pdev);
+	struct ata_host *host = dev_get_drvdata(dev);
+
+	ata_host_remove(host);
+	dev_set_drvdata(dev, NULL);
+}
+
+/* For now keep DMA off. We can set it for all but A rev CS5510 once the
+   core ATA code can handle it */
+
+static struct pci_device_id pata_cs5520[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_CYRIX, PCI_DEVICE_ID_CYRIX_5510), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_CYRIX, PCI_DEVICE_ID_CYRIX_5520), },
+	{ 0, },
+};
+
+static struct pci_driver cs5520_pci_driver = {
+	.name 		= DRV_NAME,
+	.id_table	= pata_cs5520,
+	.probe 		= cs5520_init_one,
+	.remove		= cs5520_remove_one
+};
+
+
+static int __init cs5520_init(void)
+{
+	return pci_register_driver(&cs5520_pci_driver);
+}
+
+static void __exit cs5520_exit(void)
+{
+	pci_unregister_driver(&cs5520_pci_driver);
+}
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for Cyrix CS5510/5520");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, pata_cs5520);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(cs5520_init);
+module_exit(cs5520_exit);
+
diff --git a/drivers/ata/pata_cs5530.c b/drivers/ata/pata_cs5530.c
new file mode 100644
index 000000000000..f3d8a3bc1e78
--- /dev/null
+++ b/drivers/ata/pata_cs5530.c
@@ -0,0 +1,387 @@
+/*
+ * pata-cs5530.c 	- CS5530 PATA for new ATA layer
+ *			  (C) 2005 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ * based upon cs5530.c by Mark Lord.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ *
+ * Loosely based on the piix & svwks drivers.
+ *
+ * Documentation:
+ *	Available from AMD web site.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+#include <linux/dmi.h>
+
+#define DRV_NAME	"pata_cs5530"
+#define DRV_VERSION	"0.6"
+
+/**
+ *	cs5530_set_piomode		-	PIO setup
+ *	@ap: ATA interface
+ *	@adev: device on the interface
+ *
+ *	Set our PIO requirements. This is fairly simple on the CS5530
+ *	chips.
+ */
+
+static void cs5530_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	static const unsigned int cs5530_pio_timings[2][5] = {
+		{0x00009172, 0x00012171, 0x00020080, 0x00032010, 0x00040010},
+		{0xd1329172, 0x71212171, 0x30200080, 0x20102010, 0x00100010}
+	};
+	unsigned long base = ( ap->ioaddr.bmdma_addr & ~0x0F) + 0x20 + 0x10 * ap->port_no;
+	u32 tuning;
+	int format;
+
+	/* Find out which table to use */
+	tuning = inl(base + 0x04);
+	format = (tuning & 0x80000000UL) ? 1 : 0;
+
+	/* Now load the right timing register */
+	if (adev->devno)
+		base += 0x08;
+
+	outl(cs5530_pio_timings[format][adev->pio_mode - XFER_PIO_0], base);
+}
+
+/**
+ *	cs5530_set_dmamode		-	DMA timing setup
+ *	@ap: ATA interface
+ *	@adev: Device being configured
+ *
+ *	We cannot mix MWDMA and UDMA without reloading timings each switch
+ *	master to slave. We track the last DMA setup in order to minimise
+ *	reloads.
+ */
+
+static void cs5530_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	unsigned long base = ( ap->ioaddr.bmdma_addr & ~0x0F) + 0x20 + 0x10 * ap->port_no;
+	u32 tuning, timing = 0;
+	u8 reg;
+
+	/* Find out which table to use */
+	tuning = inl(base + 0x04);
+
+	switch(adev->dma_mode) {
+		case XFER_UDMA_0:
+			timing  = 0x00921250;break;
+		case XFER_UDMA_1:
+			timing  = 0x00911140;break;
+		case XFER_UDMA_2:
+			timing  = 0x00911030;break;
+		case XFER_MW_DMA_0:
+			timing  = 0x00077771;break;
+		case XFER_MW_DMA_1:
+			timing  = 0x00012121;break;
+		case XFER_MW_DMA_2:
+			timing  = 0x00002020;break;
+		default:
+			BUG();
+	}
+	/* Merge in the PIO format bit */
+	timing |= (tuning & 0x80000000UL);
+	if (adev->devno == 0) /* Master */
+		outl(timing, base + 0x04);
+	else {
+		if (timing & 0x00100000)
+			tuning |= 0x00100000;	/* UDMA for both */
+		else
+			tuning &= ~0x00100000;	/* MWDMA for both */
+		outl(tuning, base + 0x04);
+		outl(timing, base + 0x0C);
+	}
+
+	/* Set the DMA capable bit in the BMDMA area */
+	reg = inb(ap->ioaddr.bmdma_addr + ATA_DMA_STATUS);
+	reg |= (1 << (5 + adev->devno));
+	outb(reg, ap->ioaddr.bmdma_addr + ATA_DMA_STATUS);
+
+	/* Remember the last DMA setup we did */
+
+	ap->private_data = adev;
+}
+
+/**
+ *	cs5530_qc_issue_prot	-	command issue
+ *	@qc: command pending
+ *
+ *	Called when the libata layer is about to issue a command. We wrap
+ *	this interface so that we can load the correct ATA timings if
+ *	neccessary.  Specifically we have a problem that there is only
+ *	one MWDMA/UDMA bit.
+ */
+
+static unsigned int cs5530_qc_issue_prot(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct ata_device *adev = qc->dev;
+	struct ata_device *prev = ap->private_data;
+
+	/* See if the DMA settings could be wrong */
+	if (adev->dma_mode != 0 && adev != prev && prev != NULL) {
+		/* Maybe, but do the channels match MWDMA/UDMA ? */
+		if ((adev->dma_mode >= XFER_UDMA_0 && prev->dma_mode < XFER_UDMA_0) ||
+		    (adev->dma_mode < XFER_UDMA_0 && prev->dma_mode >= XFER_UDMA_0))
+		    	/* Switch the mode bits */
+		    	cs5530_set_dmamode(ap, adev);
+	}
+
+	return ata_qc_issue_prot(qc);
+}
+
+static int cs5530_pre_reset(struct ata_port *ap)
+{
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+static void cs5530_error_handler(struct ata_port *ap)
+{
+	return ata_bmdma_drive_eh(ap, cs5530_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+
+static struct scsi_host_template cs5530_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations cs5530_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= cs5530_set_piomode,
+	.set_dmamode	= cs5530_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= cs5530_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= cs5530_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static struct dmi_system_id palmax_dmi_table[] = {
+	{
+		.ident = "Palmax PD1100",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "Cyrix"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "Caddis"),
+		},
+	},
+	{ }
+};
+
+static int cs5530_is_palmax(void)
+{
+	if (dmi_check_system(palmax_dmi_table)) {
+		printk(KERN_INFO "Palmax PD1100: Disabling DMA on docking port.\n");
+		return 1;
+	}
+	return 0;
+}
+
+/**
+ *	cs5530_init_one		-	Initialise a CS5530
+ *	@dev: PCI device
+ *	@id: Entry in match table
+ *
+ *	Install a driver for the newly found CS5530 companion chip. Most of
+ *	this is just housekeeping. We have to set the chip up correctly and
+ *	turn off various bits of emulation magic.
+ */
+
+static int cs5530_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	int compiler_warning_pointless_fix;
+	struct pci_dev *master_0 = NULL, *cs5530_0 = NULL;
+	static struct ata_port_info info = {
+		.sht = &cs5530_sht,
+		.flags = ATA_FLAG_SLAVE_POSS|ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x07,
+		.port_ops = &cs5530_port_ops
+	};
+	/* The docking connector doesn't do UDMA, and it seems not MWDMA */
+	static struct ata_port_info info_palmax_secondary = {
+		.sht = &cs5530_sht,
+		.flags = ATA_FLAG_SLAVE_POSS|ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.port_ops = &cs5530_port_ops
+	};
+	static struct ata_port_info *port_info[2] = { &info, &info };
+
+	dev = NULL;
+	while ((dev = pci_get_device(PCI_VENDOR_ID_CYRIX, PCI_ANY_ID, dev)) != NULL) {
+		switch (dev->device) {
+			case PCI_DEVICE_ID_CYRIX_PCI_MASTER:
+				master_0 = pci_dev_get(dev);
+				break;
+			case PCI_DEVICE_ID_CYRIX_5530_LEGACY:
+				cs5530_0 = pci_dev_get(dev);
+				break;
+		}
+	}
+	if (!master_0) {
+		printk(KERN_ERR DRV_NAME ": unable to locate PCI MASTER function\n");
+		goto fail_put;
+	}
+	if (!cs5530_0) {
+		printk(KERN_ERR DRV_NAME ": unable to locate CS5530 LEGACY function\n");
+		goto fail_put;
+	}
+
+	pci_set_master(cs5530_0);
+	compiler_warning_pointless_fix = pci_set_mwi(cs5530_0);
+
+	/*
+	 * Set PCI CacheLineSize to 16-bytes:
+	 * --> Write 0x04 into 8-bit PCI CACHELINESIZE reg of function 0 of the cs5530
+	 *
+	 * Note: This value is constant because the 5530 is only a Geode companion
+	 */
+
+	pci_write_config_byte(cs5530_0, PCI_CACHE_LINE_SIZE, 0x04);
+
+	/*
+	 * Disable trapping of UDMA register accesses (Win98 hack):
+	 * --> Write 0x5006 into 16-bit reg at offset 0xd0 of function 0 of the cs5530
+	 */
+
+	pci_write_config_word(cs5530_0, 0xd0, 0x5006);
+
+	/*
+	 * Bit-1 at 0x40 enables MemoryWriteAndInvalidate on internal X-bus:
+	 * The other settings are what is necessary to get the register
+	 * into a sane state for IDE DMA operation.
+	 */
+
+	pci_write_config_byte(master_0, 0x40, 0x1e);
+
+	/*
+	 * Set max PCI burst size (16-bytes seems to work best):
+	 *	   16bytes: set bit-1 at 0x41 (reg value of 0x16)
+	 *	all others: clear bit-1 at 0x41, and do:
+	 *	  128bytes: OR 0x00 at 0x41
+	 *	  256bytes: OR 0x04 at 0x41
+	 *	  512bytes: OR 0x08 at 0x41
+	 *	 1024bytes: OR 0x0c at 0x41
+	 */
+
+	pci_write_config_byte(master_0, 0x41, 0x14);
+
+	/*
+	 * These settings are necessary to get the chip
+	 * into a sane state for IDE DMA operation.
+	 */
+
+	pci_write_config_byte(master_0, 0x42, 0x00);
+	pci_write_config_byte(master_0, 0x43, 0xc1);
+
+	pci_dev_put(master_0);
+	pci_dev_put(cs5530_0);
+
+	if (cs5530_is_palmax())
+		port_info[1] = &info_palmax_secondary;
+
+	/* Now kick off ATA set up */
+	return ata_pci_init_one(dev, port_info, 2);
+
+fail_put:
+	if (master_0)
+		pci_dev_put(master_0);
+	if (cs5530_0)
+		pci_dev_put(cs5530_0);
+	return -ENODEV;
+}
+
+static struct pci_device_id cs5530[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_CYRIX, PCI_DEVICE_ID_CYRIX_5530_IDE), },
+	{ 0, },
+};
+
+static struct pci_driver cs5530_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= cs5530,
+	.probe 		= cs5530_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init cs5530_init(void)
+{
+	return pci_register_driver(&cs5530_pci_driver);
+}
+
+
+static void __exit cs5530_exit(void)
+{
+	pci_unregister_driver(&cs5530_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for the Cyrix/NS/AMD 5530");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, cs5530);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(cs5530_init);
+module_exit(cs5530_exit);
diff --git a/drivers/ata/pata_cs5535.c b/drivers/ata/pata_cs5535.c
new file mode 100644
index 000000000000..69d6b4258724
--- /dev/null
+++ b/drivers/ata/pata_cs5535.c
@@ -0,0 +1,291 @@
+/*
+ * pata-cs5535.c 	- CS5535 PATA for new ATA layer
+ *			  (C) 2005-2006 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ * based upon cs5535.c from AMD <Jens.Altmann@amd.com> as cleaned up and
+ * made readable and Linux style by Wolfgang Zuleger <wolfgang.zuleger@gmx.de
+ * and Alexander Kiausch <alex.kiausch@t-online.de>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ *
+ * Loosely based on the piix & svwks drivers.
+ *
+ * Documentation:
+ *	Available from AMD web site.
+ * TODO
+ *	Review errata to see if serializing is neccessary
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+#include <asm/msr.h>
+
+#define DRV_NAME	"cs5535"
+#define DRV_VERSION	"0.2.10"
+
+/*
+ *	The Geode (Aka Athlon GX now) uses an internal MSR based
+ *	bus system for control. Demented but there you go.
+ */
+
+#define MSR_ATAC_BASE    	0x51300000
+#define ATAC_GLD_MSR_CAP 	(MSR_ATAC_BASE+0)
+#define ATAC_GLD_MSR_CONFIG    (MSR_ATAC_BASE+0x01)
+#define ATAC_GLD_MSR_SMI       (MSR_ATAC_BASE+0x02)
+#define ATAC_GLD_MSR_ERROR     (MSR_ATAC_BASE+0x03)
+#define ATAC_GLD_MSR_PM        (MSR_ATAC_BASE+0x04)
+#define ATAC_GLD_MSR_DIAG      (MSR_ATAC_BASE+0x05)
+#define ATAC_IO_BAR            (MSR_ATAC_BASE+0x08)
+#define ATAC_RESET             (MSR_ATAC_BASE+0x10)
+#define ATAC_CH0D0_PIO         (MSR_ATAC_BASE+0x20)
+#define ATAC_CH0D0_DMA         (MSR_ATAC_BASE+0x21)
+#define ATAC_CH0D1_PIO         (MSR_ATAC_BASE+0x22)
+#define ATAC_CH0D1_DMA         (MSR_ATAC_BASE+0x23)
+#define ATAC_PCI_ABRTERR       (MSR_ATAC_BASE+0x24)
+
+#define ATAC_BM0_CMD_PRIM      0x00
+#define ATAC_BM0_STS_PRIM      0x02
+#define ATAC_BM0_PRD           0x04
+
+#define CS5535_CABLE_DETECT    0x48
+
+#define CS5535_BAD_PIO(timings) ( (timings&~0x80000000UL)==0x00009172 )
+
+/**
+ *	cs5535_pre_reset	-	detect cable type
+ *	@ap: Port to detect on
+ *
+ *	Perform cable detection for ATA66 capable cable. Return a libata
+ *	cable type.
+ */
+
+static int cs5535_pre_reset(struct ata_port *ap)
+{
+	u8 cable;
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+
+	pci_read_config_byte(pdev, CS5535_CABLE_DETECT, &cable);
+	if (cable & 1)
+		ap->cbl = ATA_CBL_PATA80;
+	else
+		ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	cs5535_error_handler		-	reset/probe
+ *	@ap: Port to reset
+ *
+ *	Reset and configure a port
+ */
+
+static void cs5535_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, cs5535_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	cs5535_set_piomode		-	PIO setup
+ *	@ap: ATA interface
+ *	@adev: device on the interface
+ *
+ *	Set our PIO requirements. The CS5535 is pretty clean about all this
+ */
+
+static void cs5535_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	static const u16 pio_timings[5] = {
+		0xF7F4, 0x53F3, 0x13F1, 0x5131, 0x1131
+	};
+	static const u16 pio_cmd_timings[5] = {
+		0xF7F4, 0x53F3, 0x13F1, 0x5131, 0x1131
+	};
+	u32 reg, dummy;
+	struct ata_device *pair = ata_dev_pair(adev);
+
+	int mode = adev->pio_mode - XFER_PIO_0;
+	int cmdmode = mode;
+
+	/* Command timing has to be for the lowest of the pair of devices */
+	if (pair) {
+		int pairmode = pair->pio_mode - XFER_PIO_0;
+		cmdmode = min(mode, pairmode);
+		/* Write the other drive timing register if it changed */
+		if (cmdmode < pairmode)
+			wrmsr(ATAC_CH0D0_PIO + 2 * pair->devno,
+				pio_cmd_timings[cmdmode] << 16 | pio_timings[pairmode], 0);
+	}
+	/* Write the drive timing register */
+	wrmsr(ATAC_CH0D0_PIO + 2 * adev->devno,
+		pio_cmd_timings[cmdmode] << 16 | pio_timings[mode], 0);
+
+	/* Set the PIO "format 1" bit in the DMA timing register */
+	rdmsr(ATAC_CH0D0_DMA + 2 * adev->devno, reg, dummy);
+	wrmsr(ATAC_CH0D0_DMA + 2 * adev->devno, reg | 0x80000000UL, 0);
+}
+
+/**
+ *	cs5535_set_dmamode		-	DMA timing setup
+ *	@ap: ATA interface
+ *	@adev: Device being configured
+ *
+ */
+
+static void cs5535_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	static const u32 udma_timings[5] = {
+		0x7F7436A1, 0x7F733481, 0x7F723261, 0x7F713161, 0x7F703061
+	};
+	static const u32 mwdma_timings[3] = {
+		0x7F0FFFF3, 0x7F035352, 0x7F024241
+	};
+	u32 reg, dummy;
+	int mode = adev->dma_mode;
+
+	rdmsr(ATAC_CH0D0_DMA + 2 * adev->devno, reg, dummy);
+	reg &= 0x80000000UL;
+	if (mode >= XFER_UDMA_0)
+		reg |= udma_timings[mode - XFER_UDMA_0];
+	else
+		reg |= mwdma_timings[mode - XFER_MW_DMA_0];
+	wrmsr(ATAC_CH0D0_DMA + 2 * adev->devno, reg, 0);
+}
+
+static struct scsi_host_template cs5535_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations cs5535_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= cs5535_set_piomode,
+	.set_dmamode	= cs5535_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= cs5535_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+/**
+ *	cs5535_init_one		-	Initialise a CS5530
+ *	@dev: PCI device
+ *	@id: Entry in match table
+ *
+ *	Install a driver for the newly found CS5530 companion chip. Most of
+ *	this is just housekeeping. We have to set the chip up correctly and
+ *	turn off various bits of emulation magic.
+ */
+
+static int cs5535_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info = {
+		.sht = &cs5535_sht,
+		.flags = ATA_FLAG_SLAVE_POSS|ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x1f,
+		.port_ops = &cs5535_port_ops
+	};
+	struct ata_port_info *ports[1] = { &info };
+
+	u32 timings, dummy;
+
+	/* Check the BIOS set the initial timing clock. If not set the
+	   timings for PIO0 */
+	rdmsr(ATAC_CH0D0_PIO, timings, dummy);
+	if (CS5535_BAD_PIO(timings))
+		wrmsr(ATAC_CH0D0_PIO, 0xF7F4F7F4UL, 0);
+	rdmsr(ATAC_CH0D1_PIO, timings, dummy);
+	if (CS5535_BAD_PIO(timings))
+		wrmsr(ATAC_CH0D1_PIO, 0xF7F4F7F4UL, 0);
+	return ata_pci_init_one(dev, ports, 1);
+}
+
+static struct pci_device_id cs5535[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_NS, 0x002D), },
+	{ 0, },
+};
+
+static struct pci_driver cs5535_pci_driver = {
+	.name		= DRV_NAME,
+	.id_table	= cs5535,
+	.probe 		= cs5535_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init cs5535_init(void)
+{
+	return pci_register_driver(&cs5535_pci_driver);
+}
+
+
+static void __exit cs5535_exit(void)
+{
+	pci_unregister_driver(&cs5535_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox, Jens Altmann, Wolfgan Zuleger, Alexander Kiausch");
+MODULE_DESCRIPTION("low-level driver for the NS/AMD 5530");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, cs5535);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(cs5535_init);
+module_exit(cs5535_exit);
diff --git a/drivers/ata/pata_cypress.c b/drivers/ata/pata_cypress.c
new file mode 100644
index 000000000000..322043d8cf7f
--- /dev/null
+++ b/drivers/ata/pata_cypress.c
@@ -0,0 +1,227 @@
+/*
+ * pata_cypress.c 	- Cypress PATA for new ATA layer
+ *			  (C) 2006 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ * Based heavily on
+ * linux/drivers/ide/pci/cy82c693.c		Version 0.40	Sep. 10, 2002
+ *
+ */
+ 
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "pata_cypress"
+#define DRV_VERSION "0.1.2"
+
+/* here are the offset definitions for the registers */
+
+enum {
+	CY82_IDE_CMDREG		= 0x04,
+	CY82_IDE_ADDRSETUP	= 0x48,
+	CY82_IDE_MASTER_IOR	= 0x4C,
+	CY82_IDE_MASTER_IOW	= 0x4D,
+	CY82_IDE_SLAVE_IOR	= 0x4E,
+	CY82_IDE_SLAVE_IOW	= 0x4F,
+	CY82_IDE_MASTER_8BIT	= 0x50,
+	CY82_IDE_SLAVE_8BIT	= 0x51,
+
+	CY82_INDEX_PORT		= 0x22,
+	CY82_DATA_PORT		= 0x23,
+
+	CY82_INDEX_CTRLREG1	= 0x01,
+	CY82_INDEX_CHANNEL0	= 0x30,
+	CY82_INDEX_CHANNEL1	= 0x31,
+	CY82_INDEX_TIMEOUT	= 0x32
+};
+
+static int cy82c693_pre_reset(struct ata_port *ap)
+{
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+static void cy82c693_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, cy82c693_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	cy82c693_set_piomode	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Called to do the PIO mode setup.
+ */
+ 
+static void cy82c693_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	struct ata_timing t;
+	const unsigned long T = 1000000 / 33;
+	short time_16, time_8;
+	u32 addr;
+	
+	if (ata_timing_compute(adev, adev->pio_mode, &t, T, 1) < 0) {
+		printk(KERN_ERR DRV_NAME ": mome computation failed.\n");
+		return;
+	}
+
+	time_16 = FIT(t.recover, 0, 15) | (FIT(t.active, 0, 15) << 4);
+	time_8 = FIT(t.act8b, 0, 15) | (FIT(t.rec8b, 0, 15) << 4);
+	
+	if (adev->devno == 0) {
+		pci_read_config_dword(pdev, CY82_IDE_ADDRSETUP, &addr);
+		
+		addr &= ~0x0F;	/* Mask bits */
+		addr |= FIT(t.setup, 0, 15);
+		
+		pci_write_config_dword(pdev, CY82_IDE_ADDRSETUP, addr);
+		pci_write_config_byte(pdev, CY82_IDE_MASTER_IOR, time_16);
+		pci_write_config_byte(pdev, CY82_IDE_MASTER_IOW, time_16);
+		pci_write_config_byte(pdev, CY82_IDE_MASTER_8BIT, time_8);
+	} else {
+		pci_read_config_dword(pdev, CY82_IDE_ADDRSETUP, &addr);
+		
+		addr &= ~0xF0;	/* Mask bits */
+		addr |= (FIT(t.setup, 0, 15) << 4);
+
+		pci_write_config_dword(pdev, CY82_IDE_ADDRSETUP, addr);
+		pci_write_config_byte(pdev, CY82_IDE_SLAVE_IOR, time_16);
+		pci_write_config_byte(pdev, CY82_IDE_SLAVE_IOW, time_16);
+		pci_write_config_byte(pdev, CY82_IDE_SLAVE_8BIT, time_8);
+	}
+}
+
+/**
+ *	cy82c693_set_dmamode	-	set initial DMA mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Called to do the DMA mode setup.
+ */
+ 
+static void cy82c693_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	int reg = CY82_INDEX_CHANNEL0 + ap->port_no;
+	
+	/* Be afraid, be very afraid. Magic registers  in low I/O space */
+	outb(reg, 0x22);
+	outb(adev->dma_mode - XFER_MW_DMA_0, 0x23);
+	
+	/* 0x50 gives the best behaviour on the Alpha's using this chip */
+	outb(CY82_INDEX_TIMEOUT, 0x22);
+	outb(0x50, 0x23);
+}
+
+static struct scsi_host_template cy82c693_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations cy82c693_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= cy82c693_set_piomode,
+	.set_dmamode	= cy82c693_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= cy82c693_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+	
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+static int cy82c693_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info = {
+		.sht = &cy82c693_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.port_ops = &cy82c693_port_ops
+	};
+	static struct ata_port_info *port_info[1] = { &info };
+	
+	/* Devfn 1 is the ATA primary. The secondary is magic and on devfn2. For the
+	   moment we don't handle the secondary. FIXME */
+	   
+	if (PCI_FUNC(pdev->devfn) != 1)
+		return -ENODEV;
+	
+	return ata_pci_init_one(pdev, port_info, 1);
+}
+
+static struct pci_device_id cy82c693[] = {
+	{ PCI_VENDOR_ID_CONTAQ, PCI_DEVICE_ID_CONTAQ_82C693, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0},
+	{ 0, },
+};
+
+static struct pci_driver cy82c693_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= cy82c693,
+	.probe 		= cy82c693_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init cy82c693_init(void)
+{
+	return pci_register_driver(&cy82c693_pci_driver);
+}
+
+
+static void __exit cy82c693_exit(void)
+{
+	pci_unregister_driver(&cy82c693_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for the CY82C693 PATA controller");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, cy82c693);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(cy82c693_init);
+module_exit(cy82c693_exit);
diff --git a/drivers/ata/pata_efar.c b/drivers/ata/pata_efar.c
new file mode 100644
index 000000000000..c30bc181304f
--- /dev/null
+++ b/drivers/ata/pata_efar.c
@@ -0,0 +1,342 @@
+/*
+ *    pata_efar.c - EFAR PIIX clone controller driver
+ *
+ *	(C) 2005 Red Hat <alan@redhat.com>
+ *
+ *    Some parts based on ata_piix.c by Jeff Garzik and others.
+ *
+ *    The EFAR is a PIIX4 clone with UDMA66 support. Unlike the later
+ *    Intel ICH controllers the EFAR widened the UDMA mode register bits
+ *    and doesn't require the funky clock selection.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+#include <linux/ata.h>
+
+#define DRV_NAME	"pata_efar"
+#define DRV_VERSION	"0.4.1"
+
+/**
+ *	efar_pre_reset	-	check for 40/80 pin
+ *	@ap: Port
+ *
+ *	Perform cable detection for the EFAR ATA interface. This is
+ *	different to the PIIX arrangement
+ */
+
+static int efar_pre_reset(struct ata_port *ap)
+{
+	static const struct pci_bits efar_enable_bits[] = {
+		{ 0x41U, 1U, 0x80UL, 0x80UL },	/* port 0 */
+		{ 0x43U, 1U, 0x80UL, 0x80UL },	/* port 1 */
+	};
+
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u8 tmp;
+
+	if (!pci_test_config_bits(pdev, &efar_enable_bits[ap->port_no])) {
+		ata_port_disable(ap);
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+	pci_read_config_byte(pdev, 0x47, &tmp);
+	if (tmp & (2 >> ap->port_no))
+		ap->cbl = ATA_CBL_PATA40;
+	else
+		ap->cbl = ATA_CBL_PATA80;
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	efar_probe_reset - Probe specified port on PATA host controller
+ *	@ap: Port to probe
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void efar_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, efar_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	efar_set_piomode - Initialize host controller PATA PIO timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: um
+ *
+ *	Set PIO mode for device, in host controller PCI config space.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void efar_set_piomode (struct ata_port *ap, struct ata_device *adev)
+{
+	unsigned int pio	= adev->pio_mode - XFER_PIO_0;
+	struct pci_dev *dev	= to_pci_dev(ap->host->dev);
+	unsigned int idetm_port= ap->port_no ? 0x42 : 0x40;
+	u16 idetm_data;
+	int control = 0;
+
+	/*
+	 *	See Intel Document 298600-004 for the timing programing rules
+	 *	for PIIX/ICH. The EFAR is a clone so very similar
+	 */
+
+	static const	 /* ISP  RTC */
+	u8 timings[][2]	= { { 0, 0 },
+			    { 0, 0 },
+			    { 1, 0 },
+			    { 2, 1 },
+			    { 2, 3 }, };
+
+	if (pio > 2)
+		control |= 1;	/* TIME1 enable */
+	if (ata_pio_need_iordy(adev))	/* PIO 3/4 require IORDY */
+		control |= 2;	/* IE enable */
+	/* Intel specifies that the PPE functionality is for disk only */
+	if (adev->class == ATA_DEV_ATA)
+		control |= 4;	/* PPE enable */
+
+	pci_read_config_word(dev, idetm_port, &idetm_data);
+
+	/* Enable PPE, IE and TIME as appropriate */
+
+	if (adev->devno == 0) {
+		idetm_data &= 0xCCF0;
+		idetm_data |= control;
+		idetm_data |= (timings[pio][0] << 12) |
+			(timings[pio][1] << 8);
+	} else {
+		int shift = 4 * ap->port_no;
+		u8 slave_data;
+
+		idetm_data &= 0xCC0F;
+		idetm_data |= (control << 4);
+
+		/* Slave timing in seperate register */
+		pci_read_config_byte(dev, 0x44, &slave_data);
+		slave_data &= 0x0F << shift;
+		slave_data |= ((timings[pio][0] << 2) | timings[pio][1]) << shift;
+		pci_write_config_byte(dev, 0x44, slave_data);
+	}
+
+	idetm_data |= 0x4000;	/* Ensure SITRE is enabled */
+	pci_write_config_word(dev, idetm_port, idetm_data);
+}
+
+/**
+ *	efar_set_dmamode - Initialize host controller PATA DMA timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: Device to program
+ *
+ *	Set UDMA/MWDMA mode for device, in host controller PCI config space.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void efar_set_dmamode (struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *dev	= to_pci_dev(ap->host->dev);
+	u8 master_port		= ap->port_no ? 0x42 : 0x40;
+	u16 master_data;
+	u8 speed		= adev->dma_mode;
+	int devid		= adev->devno + 2 * ap->port_no;
+	u8 udma_enable;
+
+	static const	 /* ISP  RTC */
+	u8 timings[][2]	= { { 0, 0 },
+			    { 0, 0 },
+			    { 1, 0 },
+			    { 2, 1 },
+			    { 2, 3 }, };
+
+	pci_read_config_word(dev, master_port, &master_data);
+	pci_read_config_byte(dev, 0x48, &udma_enable);
+
+	if (speed >= XFER_UDMA_0) {
+		unsigned int udma	= adev->dma_mode - XFER_UDMA_0;
+		u16 udma_timing;
+
+		udma_enable |= (1 << devid);
+
+		/* Load the UDMA mode number */
+		pci_read_config_word(dev, 0x4A, &udma_timing);
+		udma_timing &= ~(7 << (4 * devid));
+		udma_timing |= udma << (4 * devid);
+		pci_write_config_word(dev, 0x4A, udma_timing);
+	} else {
+		/*
+		 * MWDMA is driven by the PIO timings. We must also enable
+		 * IORDY unconditionally along with TIME1. PPE has already
+		 * been set when the PIO timing was set.
+		 */
+		unsigned int mwdma	= adev->dma_mode - XFER_MW_DMA_0;
+		unsigned int control;
+		u8 slave_data;
+		const unsigned int needed_pio[3] = {
+			XFER_PIO_0, XFER_PIO_3, XFER_PIO_4
+		};
+		int pio = needed_pio[mwdma] - XFER_PIO_0;
+
+		control = 3;	/* IORDY|TIME1 */
+
+		/* If the drive MWDMA is faster than it can do PIO then
+		   we must force PIO into PIO0 */
+
+		if (adev->pio_mode < needed_pio[mwdma])
+			/* Enable DMA timing only */
+			control |= 8;	/* PIO cycles in PIO0 */
+
+		if (adev->devno) {	/* Slave */
+			master_data &= 0xFF4F;  /* Mask out IORDY|TIME1|DMAONLY */
+			master_data |= control << 4;
+			pci_read_config_byte(dev, 0x44, &slave_data);
+			slave_data &= (0x0F + 0xE1 * ap->port_no);
+			/* Load the matching timing */
+			slave_data |= ((timings[pio][0] << 2) | timings[pio][1]) << (ap->port_no ? 4 : 0);
+			pci_write_config_byte(dev, 0x44, slave_data);
+		} else { 	/* Master */
+			master_data &= 0xCCF4;	/* Mask out IORDY|TIME1|DMAONLY
+						   and master timing bits */
+			master_data |= control;
+			master_data |=
+				(timings[pio][0] << 12) |
+				(timings[pio][1] << 8);
+		}
+		udma_enable &= ~(1 << devid);
+		pci_write_config_word(dev, master_port, master_data);
+	}
+	pci_write_config_byte(dev, 0x48, udma_enable);
+}
+
+static struct scsi_host_template efar_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static const struct ata_port_operations efar_ops = {
+	.port_disable		= ata_port_disable,
+	.set_piomode		= efar_set_piomode,
+	.set_dmamode		= efar_set_dmamode,
+	.mode_filter		= ata_pci_default_filter,
+
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.check_status		= ata_check_status,
+	.exec_command		= ata_exec_command,
+	.dev_select		= ata_std_dev_select,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= efar_error_handler,
+	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= ata_qc_issue_prot,
+	.data_xfer		= ata_pio_data_xfer,
+
+	.eng_timeout		= ata_eng_timeout,
+
+	.irq_handler		= ata_interrupt,
+	.irq_clear		= ata_bmdma_irq_clear,
+
+	.port_start		= ata_port_start,
+	.port_stop		= ata_port_stop,
+	.host_stop		= ata_host_stop,
+};
+
+
+/**
+ *	efar_init_one - Register EFAR ATA PCI device with kernel services
+ *	@pdev: PCI device to register
+ *	@ent: Entry in efar_pci_tbl matching with @pdev
+ *
+ *	Called from kernel PCI layer.
+ *
+ *	LOCKING:
+ *	Inherited from PCI layer (may sleep).
+ *
+ *	RETURNS:
+ *	Zero on success, or -ERRNO value.
+ */
+
+static int efar_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+	static int printed_version;
+	static struct ata_port_info info = {
+		.sht		= &efar_sht,
+		.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask	= 0x1f,	/* pio0-4 */
+		.mwdma_mask	= 0x07, /* mwdma1-2 */
+		.udma_mask 	= 0x0f, /* UDMA 66 */
+		.port_ops	= &efar_ops,
+	};
+	static struct ata_port_info *port_info[2] = { &info, &info };
+
+	if (!printed_version++)
+		dev_printk(KERN_DEBUG, &pdev->dev,
+			   "version " DRV_VERSION "\n");
+
+	return ata_pci_init_one(pdev, port_info, 2);
+}
+
+static const struct pci_device_id efar_pci_tbl[] = {
+	{ 0x1055, 0x9130, PCI_ANY_ID, PCI_ANY_ID, },
+	{ }	/* terminate list */
+};
+
+static struct pci_driver efar_pci_driver = {
+	.name			= DRV_NAME,
+	.id_table		= efar_pci_tbl,
+	.probe			= efar_init_one,
+	.remove			= ata_pci_remove_one,
+};
+
+static int __init efar_init(void)
+{
+	return pci_register_driver(&efar_pci_driver);
+}
+
+static void __exit efar_exit(void)
+{
+	pci_unregister_driver(&efar_pci_driver);
+}
+
+
+module_init(efar_init);
+module_exit(efar_exit);
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("SCSI low-level driver for EFAR PIIX clones");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, efar_pci_tbl);
+MODULE_VERSION(DRV_VERSION);
+
diff --git a/drivers/ata/pata_hpt366.c b/drivers/ata/pata_hpt366.c
new file mode 100644
index 000000000000..e54a0c51681d
--- /dev/null
+++ b/drivers/ata/pata_hpt366.c
@@ -0,0 +1,478 @@
+/*
+ * Libata driver for the highpoint 366 and 368 UDMA66 ATA controllers.
+ *
+ * This driver is heavily based upon:
+ *
+ * linux/drivers/ide/pci/hpt366.c		Version 0.36	April 25, 2003
+ *
+ * Copyright (C) 1999-2003		Andre Hedrick <andre@linux-ide.org>
+ * Portions Copyright (C) 2001	        Sun Microsystems, Inc.
+ * Portions Copyright (C) 2003		Red Hat Inc
+ *
+ *
+ * TODO
+ *	Maybe PLL mode
+ *	Look into engine reset on timeout errors. Should not be
+ *		required.
+ */
+
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME	"pata_hpt366"
+#define DRV_VERSION	"0.5"
+
+struct hpt_clock {
+	u8	xfer_speed;
+	u32	timing;
+};
+
+/* key for bus clock timings
+ * bit
+ * 0:3    data_high_time. inactive time of DIOW_/DIOR_ for PIO and MW
+ *        DMA. cycles = value + 1
+ * 4:8    data_low_time. active time of DIOW_/DIOR_ for PIO and MW
+ *        DMA. cycles = value + 1
+ * 9:12   cmd_high_time. inactive time of DIOW_/DIOR_ during task file
+ *        register access.
+ * 13:17  cmd_low_time. active time of DIOW_/DIOR_ during task file
+ *        register access.
+ * 18:21  udma_cycle_time. clock freq and clock cycles for UDMA xfer.
+ *        during task file register access.
+ * 22:24  pre_high_time. time to initialize 1st cycle for PIO and MW DMA
+ *        xfer.
+ * 25:27  cmd_pre_high_time. time to initialize 1st PIO cycle for task
+ *        register access.
+ * 28     UDMA enable
+ * 29     DMA enable
+ * 30     PIO_MST enable. if set, the chip is in bus master mode during
+ *        PIO.
+ * 31     FIFO enable.
+ */
+
+static const struct hpt_clock hpt366_40[] = {
+	{	XFER_UDMA_4,	0x900fd943	},
+	{	XFER_UDMA_3,	0x900ad943	},
+	{	XFER_UDMA_2,	0x900bd943	},
+	{	XFER_UDMA_1,	0x9008d943	},
+	{	XFER_UDMA_0,	0x9008d943	},
+
+	{	XFER_MW_DMA_2,	0xa008d943	},
+	{	XFER_MW_DMA_1,	0xa010d955	},
+	{	XFER_MW_DMA_0,	0xa010d9fc	},
+
+	{	XFER_PIO_4,	0xc008d963	},
+	{	XFER_PIO_3,	0xc010d974	},
+	{	XFER_PIO_2,	0xc010d997	},
+	{	XFER_PIO_1,	0xc010d9c7	},
+	{	XFER_PIO_0,	0xc018d9d9	},
+	{	0,		0x0120d9d9	}
+};
+
+static const struct hpt_clock hpt366_33[] = {
+	{	XFER_UDMA_4,	0x90c9a731	},
+	{	XFER_UDMA_3,	0x90cfa731	},
+	{	XFER_UDMA_2,	0x90caa731	},
+	{	XFER_UDMA_1,	0x90cba731	},
+	{	XFER_UDMA_0,	0x90c8a731	},
+
+	{	XFER_MW_DMA_2,	0xa0c8a731	},
+	{	XFER_MW_DMA_1,	0xa0c8a732	},	/* 0xa0c8a733 */
+	{	XFER_MW_DMA_0,	0xa0c8a797	},
+
+	{	XFER_PIO_4,	0xc0c8a731	},
+	{	XFER_PIO_3,	0xc0c8a742	},
+	{	XFER_PIO_2,	0xc0d0a753	},
+	{	XFER_PIO_1,	0xc0d0a7a3	},	/* 0xc0d0a793 */
+	{	XFER_PIO_0,	0xc0d0a7aa	},	/* 0xc0d0a7a7 */
+	{	0,		0x0120a7a7	}
+};
+
+static const struct hpt_clock hpt366_25[] = {
+	{	XFER_UDMA_4,	0x90c98521	},
+	{	XFER_UDMA_3,	0x90cf8521	},
+	{	XFER_UDMA_2,	0x90cf8521	},
+	{	XFER_UDMA_1,	0x90cb8521	},
+	{	XFER_UDMA_0,	0x90cb8521	},
+
+	{	XFER_MW_DMA_2,	0xa0ca8521	},
+	{	XFER_MW_DMA_1,	0xa0ca8532	},
+	{	XFER_MW_DMA_0,	0xa0ca8575	},
+
+	{	XFER_PIO_4,	0xc0ca8521	},
+	{	XFER_PIO_3,	0xc0ca8532	},
+	{	XFER_PIO_2,	0xc0ca8542	},
+	{	XFER_PIO_1,	0xc0d08572	},
+	{	XFER_PIO_0,	0xc0d08585	},
+	{	0,		0x01208585	}
+};
+
+static const char *bad_ata33[] = {
+	"Maxtor 92720U8", "Maxtor 92040U6", "Maxtor 91360U4", "Maxtor 91020U3", "Maxtor 90845U3", "Maxtor 90650U2",
+	"Maxtor 91360D8", "Maxtor 91190D7", "Maxtor 91020D6", "Maxtor 90845D5", "Maxtor 90680D4", "Maxtor 90510D3", "Maxtor 90340D2",
+	"Maxtor 91152D8", "Maxtor 91008D7", "Maxtor 90845D6", "Maxtor 90840D6", "Maxtor 90720D5", "Maxtor 90648D5", "Maxtor 90576D4",
+	"Maxtor 90510D4",
+	"Maxtor 90432D3", "Maxtor 90288D2", "Maxtor 90256D2",
+	"Maxtor 91000D8", "Maxtor 90910D8", "Maxtor 90875D7", "Maxtor 90840D7", "Maxtor 90750D6", "Maxtor 90625D5", "Maxtor 90500D4",
+	"Maxtor 91728D8", "Maxtor 91512D7", "Maxtor 91303D6", "Maxtor 91080D5", "Maxtor 90845D4", "Maxtor 90680D4", "Maxtor 90648D3", "Maxtor 90432D2",
+	NULL
+};
+
+static const char *bad_ata66_4[] = {
+	"IBM-DTLA-307075",
+	"IBM-DTLA-307060",
+	"IBM-DTLA-307045",
+	"IBM-DTLA-307030",
+	"IBM-DTLA-307020",
+	"IBM-DTLA-307015",
+	"IBM-DTLA-305040",
+	"IBM-DTLA-305030",
+	"IBM-DTLA-305020",
+	"IC35L010AVER07-0",
+	"IC35L020AVER07-0",
+	"IC35L030AVER07-0",
+	"IC35L040AVER07-0",
+	"IC35L060AVER07-0",
+	"WDC AC310200R",
+	NULL
+};
+
+static const char *bad_ata66_3[] = {
+	"WDC AC310200R",
+	NULL
+};
+
+static int hpt_dma_blacklisted(const struct ata_device *dev, char *modestr, const char *list[])
+{
+	unsigned char model_num[40];
+	char *s;
+	unsigned int len;
+	int i = 0;
+
+	ata_id_string(dev->id, model_num, ATA_ID_PROD_OFS, sizeof(model_num));
+	s = &model_num[0];
+	len = strnlen(s, sizeof(model_num));
+
+	/* ATAPI specifies that empty space is blank-filled; remove blanks */
+	while ((len > 0) && (s[len - 1] == ' ')) {
+		len--;
+		s[len] = 0;
+	}
+
+	while(list[i] != NULL) {
+		if (!strncmp(list[i], s, len)) {
+			printk(KERN_WARNING DRV_NAME ": %s is not supported for %s.\n", 
+				modestr, list[i]);
+			return 1;
+		}
+		i++;
+	}
+	return 0;
+}
+
+/**
+ *	hpt366_filter	-	mode selection filter
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Block UDMA on devices that cause trouble with this controller.
+ */
+ 
+static unsigned long hpt366_filter(const struct ata_port *ap, struct ata_device *adev, unsigned long mask)
+{
+	if (adev->class == ATA_DEV_ATA) {
+		if (hpt_dma_blacklisted(adev, "UDMA",  bad_ata33))
+			mask &= ~ATA_MASK_UDMA;
+		if (hpt_dma_blacklisted(adev, "UDMA3", bad_ata66_3))
+			mask &= ~(0x07 << ATA_SHIFT_UDMA);
+		if (hpt_dma_blacklisted(adev, "UDMA4", bad_ata66_4))
+			mask &= ~(0x0F << ATA_SHIFT_UDMA);
+	}
+	return ata_pci_default_filter(ap, adev, mask);
+}
+
+/**
+ *	hpt36x_find_mode	-	reset the hpt36x bus
+ *	@ap: ATA port
+ *	@speed: transfer mode
+ *
+ *	Return the 32bit register programming information for this channel
+ *	that matches the speed provided.
+ */
+ 
+static u32 hpt36x_find_mode(struct ata_port *ap, int speed)
+{
+	struct hpt_clock *clocks = ap->host->private_data;
+	
+	while(clocks->xfer_speed) {
+		if (clocks->xfer_speed == speed)
+			return clocks->timing;
+		clocks++;
+	}
+	BUG();
+	return 0xffffffffU;	/* silence compiler warning */
+}
+	
+static int hpt36x_pre_reset(struct ata_port *ap)
+{
+	u8 ata66;
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	
+	pci_read_config_byte(pdev, 0x5A, &ata66);
+	if (ata66 & (1 << ap->port_no))
+		ap->cbl = ATA_CBL_PATA40;
+	else
+		ap->cbl = ATA_CBL_PATA80;
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	hpt36x_error_handler	-	reset the hpt36x bus
+ *	@ap: ATA port to reset
+ *
+ *	Perform the reset handling for the 366/368
+ */
+ 
+static void hpt36x_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, hpt36x_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	hpt366_set_piomode		-	PIO setup
+ *	@ap: ATA interface
+ *	@adev: device on the interface
+ *
+ *	Perform PIO mode setup. 
+ */
+ 
+static void hpt366_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u32 addr1, addr2;
+	u32 reg;
+	u32 mode;
+	u8 fast;
+
+	addr1 = 0x40 + 4 * (adev->devno + 2 * ap->port_no);
+	addr2 = 0x51 + 4 * ap->port_no;
+	
+	/* Fast interrupt prediction disable, hold off interrupt disable */
+	pci_read_config_byte(pdev, addr2, &fast);
+	if (fast & 0x80) {
+		fast &= ~0x80;
+		pci_write_config_byte(pdev, addr2, fast);
+	}
+	
+	pci_read_config_dword(pdev, addr1, &reg);
+	mode = hpt36x_find_mode(ap, adev->pio_mode);
+	mode &= ~0x8000000;	/* No FIFO in PIO */
+	mode &= ~0x30070000;	/* Leave config bits alone */
+	reg &= 0x30070000;	/* Strip timing bits */
+	pci_write_config_dword(pdev, addr1, reg | mode);
+}
+
+/**
+ *	hpt366_set_dmamode		-	DMA timing setup
+ *	@ap: ATA interface
+ *	@adev: Device being configured
+ *
+ *	Set up the channel for MWDMA or UDMA modes. Much the same as with
+ *	PIO, load the mode number and then set MWDMA or UDMA flag.
+ */
+ 
+static void hpt366_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u32 addr1, addr2;
+	u32 reg;
+	u32 mode;
+	u8 fast;
+
+	addr1 = 0x40 + 4 * (adev->devno + 2 * ap->port_no);
+	addr2 = 0x51 + 4 * ap->port_no;
+	
+	/* Fast interrupt prediction disable, hold off interrupt disable */
+	pci_read_config_byte(pdev, addr2, &fast);
+	if (fast & 0x80) {
+		fast &= ~0x80;
+		pci_write_config_byte(pdev, addr2, fast);
+	}
+	
+	pci_read_config_dword(pdev, addr1, &reg);
+	mode = hpt36x_find_mode(ap, adev->dma_mode);
+	mode |= 0x8000000;	/* FIFO in MWDMA or UDMA */
+	mode &= ~0xC0000000;	/* Leave config bits alone */
+	reg &= 0xC0000000;	/* Strip timing bits */
+	pci_write_config_dword(pdev, addr1, reg | mode);
+}
+
+static struct scsi_host_template hpt36x_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+/*
+ *	Configuration for HPT366/68
+ */
+ 
+static struct ata_port_operations hpt366_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= hpt366_set_piomode,
+	.set_dmamode	= hpt366_set_dmamode,
+	.mode_filter	= hpt366_filter,
+	
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= hpt36x_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+/**
+ *	hpt36x_init_one		-	Initialise an HPT366/368
+ *	@dev: PCI device
+ *	@id: Entry in match table
+ *
+ *	Initialise an HPT36x device. There are some interesting complications
+ *	here. Firstly the chip may report 366 and be one of several variants.
+ *	Secondly all the timings depend on the clock for the chip which we must
+ *	detect and look up
+ *
+ *	This is the known chip mappings. It may be missing a couple of later
+ *	releases.
+ *
+ *	Chip version		PCI		Rev	Notes
+ *	HPT366			4 (HPT366)	0	UDMA66
+ *	HPT366			4 (HPT366)	1	UDMA66
+ *	HPT368			4 (HPT366)	2	UDMA66
+ *	HPT37x/30x		4 (HPT366)	3+	Other driver
+ *
+ */
+ 
+static int hpt36x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info_hpt366 = {
+		.sht = &hpt36x_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x1f,
+		.port_ops = &hpt366_port_ops
+	};
+	struct ata_port_info *port_info[2] = {&info_hpt366, &info_hpt366};
+
+	u32 class_rev;
+	u32 reg1;
+	u8 drive_fast;
+
+	pci_read_config_dword(dev, PCI_CLASS_REVISION, &class_rev);
+	class_rev &= 0xFF;
+	
+	/* May be a later chip in disguise. Check */
+	/* Newer chips are not in the HPT36x driver. Ignore them */
+	if (class_rev > 2)
+			return -ENODEV;
+
+	pci_write_config_byte(dev, PCI_CACHE_LINE_SIZE, (L1_CACHE_BYTES / 4));
+	pci_write_config_byte(dev, PCI_LATENCY_TIMER, 0x78);
+	pci_write_config_byte(dev, PCI_MIN_GNT, 0x08);
+	pci_write_config_byte(dev, PCI_MAX_LAT, 0x08);
+
+	pci_read_config_byte(dev, 0x51, &drive_fast);
+	if (drive_fast & 0x80)
+		pci_write_config_byte(dev, 0x51, drive_fast & ~0x80);
+
+	pci_read_config_dword(dev, 0x40,  &reg1);
+	
+	/* PCI clocking determines the ATA timing values to use */
+	/* info_hpt366 is safe against re-entry so we can scribble on it */
+	switch(reg1 & 0x700) {
+		case 5:
+			info_hpt366.private_data = &hpt366_40;
+			break;
+		case 9:
+			info_hpt366.private_data = &hpt366_25;
+			break;
+		default:
+			info_hpt366.private_data = &hpt366_33;
+			break;
+	}
+	/* Now kick off ATA set up */
+	return ata_pci_init_one(dev, port_info, 2);
+}
+
+static struct pci_device_id hpt36x[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_TTI, PCI_DEVICE_ID_TTI_HPT366), },
+	{ 0, },
+};
+
+static struct pci_driver hpt36x_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= hpt36x,
+	.probe 		= hpt36x_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init hpt36x_init(void)
+{
+	return pci_register_driver(&hpt36x_pci_driver);
+}
+
+
+static void __exit hpt36x_exit(void)
+{
+	pci_unregister_driver(&hpt36x_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for the Highpoint HPT366/368");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, hpt36x);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(hpt36x_init);
+module_exit(hpt36x_exit);
diff --git a/drivers/ata/pata_hpt37x.c b/drivers/ata/pata_hpt37x.c
new file mode 100644
index 000000000000..7c3da53f1e0c
--- /dev/null
+++ b/drivers/ata/pata_hpt37x.c
@@ -0,0 +1,1257 @@
+/*
+ * Libata driver for the highpoint 37x and 30x UDMA66 ATA controllers.
+ *
+ * This driver is heavily based upon:
+ *
+ * linux/drivers/ide/pci/hpt366.c		Version 0.36	April 25, 2003
+ *
+ * Copyright (C) 1999-2003		Andre Hedrick <andre@linux-ide.org>
+ * Portions Copyright (C) 2001	        Sun Microsystems, Inc.
+ * Portions Copyright (C) 2003		Red Hat Inc
+ *
+ * TODO
+ *	PLL mode
+ *	Look into engine reset on timeout errors. Should not be
+ *		required.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME	"pata_hpt37x"
+#define DRV_VERSION	"0.5"
+
+struct hpt_clock {
+	u8	xfer_speed;
+	u32	timing;
+};
+
+struct hpt_chip {
+	const char *name;
+	unsigned int base;
+	struct hpt_clock const *clocks[4];
+};
+
+/* key for bus clock timings
+ * bit
+ * 0:3    data_high_time. inactive time of DIOW_/DIOR_ for PIO and MW
+ *        DMA. cycles = value + 1
+ * 4:8    data_low_time. active time of DIOW_/DIOR_ for PIO and MW
+ *        DMA. cycles = value + 1
+ * 9:12   cmd_high_time. inactive time of DIOW_/DIOR_ during task file
+ *        register access.
+ * 13:17  cmd_low_time. active time of DIOW_/DIOR_ during task file
+ *        register access.
+ * 18:21  udma_cycle_time. clock freq and clock cycles for UDMA xfer.
+ *        during task file register access.
+ * 22:24  pre_high_time. time to initialize 1st cycle for PIO and MW DMA
+ *        xfer.
+ * 25:27  cmd_pre_high_time. time to initialize 1st PIO cycle for task
+ *        register access.
+ * 28     UDMA enable
+ * 29     DMA enable
+ * 30     PIO_MST enable. if set, the chip is in bus master mode during
+ *        PIO.
+ * 31     FIFO enable.
+ */
+
+/* from highpoint documentation. these are old values */
+static const struct hpt_clock hpt370_timings_33[] = {
+/*	{	XFER_UDMA_5,	0x1A85F442,	0x16454e31	}, */
+	{	XFER_UDMA_5,	0x16454e31	},
+	{	XFER_UDMA_4,	0x16454e31	},
+	{	XFER_UDMA_3,	0x166d4e31	},
+	{	XFER_UDMA_2,	0x16494e31	},
+	{	XFER_UDMA_1,	0x164d4e31	},
+	{	XFER_UDMA_0,	0x16514e31	},
+
+	{	XFER_MW_DMA_2,	0x26514e21	},
+	{	XFER_MW_DMA_1,	0x26514e33	},
+	{	XFER_MW_DMA_0,	0x26514e97	},
+
+	{	XFER_PIO_4,	0x06514e21	},
+	{	XFER_PIO_3,	0x06514e22	},
+	{	XFER_PIO_2,	0x06514e33	},
+	{	XFER_PIO_1,	0x06914e43	},
+	{	XFER_PIO_0,	0x06914e57	},
+	{	0,		0x06514e57	}
+};
+
+static const struct hpt_clock hpt370_timings_66[] = {
+	{	XFER_UDMA_5,	0x14846231	},
+	{	XFER_UDMA_4,	0x14886231	},
+	{	XFER_UDMA_3,	0x148c6231	},
+	{	XFER_UDMA_2,	0x148c6231	},
+	{	XFER_UDMA_1,	0x14906231	},
+	{	XFER_UDMA_0,	0x14986231	},
+
+	{	XFER_MW_DMA_2,	0x26514e21	},
+	{	XFER_MW_DMA_1,	0x26514e33	},
+	{	XFER_MW_DMA_0,	0x26514e97	},
+
+	{	XFER_PIO_4,	0x06514e21	},
+	{	XFER_PIO_3,	0x06514e22	},
+	{	XFER_PIO_2,	0x06514e33	},
+	{	XFER_PIO_1,	0x06914e43	},
+	{	XFER_PIO_0,	0x06914e57	},
+	{	0,		0x06514e57	}
+};
+
+/* these are the current (4 sep 2001) timings from highpoint */
+static const struct hpt_clock hpt370a_timings_33[] = {
+	{	XFER_UDMA_5,	0x12446231	},
+	{	XFER_UDMA_4,	0x12446231	},
+	{	XFER_UDMA_3,	0x126c6231	},
+	{	XFER_UDMA_2,	0x12486231	},
+	{	XFER_UDMA_1,	0x124c6233	},
+	{	XFER_UDMA_0,	0x12506297	},
+
+	{	XFER_MW_DMA_2,	0x22406c31	},
+	{	XFER_MW_DMA_1,	0x22406c33	},
+	{	XFER_MW_DMA_0,	0x22406c97	},
+
+	{	XFER_PIO_4,	0x06414e31	},
+	{	XFER_PIO_3,	0x06414e42	},
+	{	XFER_PIO_2,	0x06414e53	},
+	{	XFER_PIO_1,	0x06814e93	},
+	{	XFER_PIO_0,	0x06814ea7	},
+	{	0,		0x06814ea7	}
+};
+
+/* 2x 33MHz timings */
+static const struct hpt_clock hpt370a_timings_66[] = {
+	{	XFER_UDMA_5,	0x1488e673	},
+	{	XFER_UDMA_4,	0x1488e673	},
+	{	XFER_UDMA_3,	0x1498e673	},
+	{	XFER_UDMA_2,	0x1490e673	},
+	{	XFER_UDMA_1,	0x1498e677	},
+	{	XFER_UDMA_0,	0x14a0e73f	},
+
+	{	XFER_MW_DMA_2,	0x2480fa73	},
+	{	XFER_MW_DMA_1,	0x2480fa77	}, 
+	{	XFER_MW_DMA_0,	0x2480fb3f	},
+
+	{	XFER_PIO_4,	0x0c82be73	},
+	{	XFER_PIO_3,	0x0c82be95	},
+	{	XFER_PIO_2,	0x0c82beb7	},
+	{	XFER_PIO_1,	0x0d02bf37	},
+	{	XFER_PIO_0,	0x0d02bf5f	},
+	{	0,		0x0d02bf5f	}
+};
+
+static const struct hpt_clock hpt370a_timings_50[] = {
+	{	XFER_UDMA_5,	0x12848242	},
+	{	XFER_UDMA_4,	0x12ac8242	},
+	{	XFER_UDMA_3,	0x128c8242	},
+	{	XFER_UDMA_2,	0x120c8242	},
+	{	XFER_UDMA_1,	0x12148254	},
+	{	XFER_UDMA_0,	0x121882ea	},
+
+	{	XFER_MW_DMA_2,	0x22808242	},
+	{	XFER_MW_DMA_1,	0x22808254	},
+	{	XFER_MW_DMA_0,	0x228082ea	},
+
+	{	XFER_PIO_4,	0x0a81f442	},
+	{	XFER_PIO_3,	0x0a81f443	},
+	{	XFER_PIO_2,	0x0a81f454	},
+	{	XFER_PIO_1,	0x0ac1f465	},
+	{	XFER_PIO_0,	0x0ac1f48a	},
+	{	0,		0x0ac1f48a	}
+};
+
+static const struct hpt_clock hpt372_timings_33[] = {
+	{	XFER_UDMA_6,	0x1c81dc62	},
+	{	XFER_UDMA_5,	0x1c6ddc62	},
+	{	XFER_UDMA_4,	0x1c8ddc62	},
+	{	XFER_UDMA_3,	0x1c8edc62	},	/* checkme */
+	{	XFER_UDMA_2,	0x1c91dc62	},
+	{	XFER_UDMA_1,	0x1c9adc62	},	/* checkme */
+	{	XFER_UDMA_0,	0x1c82dc62	},	/* checkme */
+
+	{	XFER_MW_DMA_2,	0x2c829262	},
+	{	XFER_MW_DMA_1,	0x2c829266	},	/* checkme */
+	{	XFER_MW_DMA_0,	0x2c82922e	},	/* checkme */
+
+	{	XFER_PIO_4,	0x0c829c62	},
+	{	XFER_PIO_3,	0x0c829c84	},
+	{	XFER_PIO_2,	0x0c829ca6	},
+	{	XFER_PIO_1,	0x0d029d26	},
+	{	XFER_PIO_0,	0x0d029d5e	},
+	{	0,		0x0d029d5e	}
+};
+
+static const struct hpt_clock hpt372_timings_50[] = {
+	{	XFER_UDMA_5,	0x12848242	},
+	{	XFER_UDMA_4,	0x12ac8242	},
+	{	XFER_UDMA_3,	0x128c8242	},
+	{	XFER_UDMA_2,	0x120c8242	},
+	{	XFER_UDMA_1,	0x12148254	},
+	{	XFER_UDMA_0,	0x121882ea	},
+
+	{	XFER_MW_DMA_2,	0x22808242	},
+	{	XFER_MW_DMA_1,	0x22808254	},
+	{	XFER_MW_DMA_0,	0x228082ea	},
+
+	{	XFER_PIO_4,	0x0a81f442	},
+	{	XFER_PIO_3,	0x0a81f443	},
+	{	XFER_PIO_2,	0x0a81f454	},
+	{	XFER_PIO_1,	0x0ac1f465	},
+	{	XFER_PIO_0,	0x0ac1f48a	},
+	{	0,		0x0a81f443	}
+};
+
+static const struct hpt_clock hpt372_timings_66[] = {
+	{	XFER_UDMA_6,	0x1c869c62	},
+	{	XFER_UDMA_5,	0x1cae9c62	},
+	{	XFER_UDMA_4,	0x1c8a9c62	},
+	{	XFER_UDMA_3,	0x1c8e9c62	},
+	{	XFER_UDMA_2,	0x1c929c62	},
+	{	XFER_UDMA_1,	0x1c9a9c62	},
+	{	XFER_UDMA_0,	0x1c829c62	},
+
+	{	XFER_MW_DMA_2,	0x2c829c62	},
+	{	XFER_MW_DMA_1,	0x2c829c66	},
+	{	XFER_MW_DMA_0,	0x2c829d2e	},
+
+	{	XFER_PIO_4,	0x0c829c62	},
+	{	XFER_PIO_3,	0x0c829c84	},
+	{	XFER_PIO_2,	0x0c829ca6	},
+	{	XFER_PIO_1,	0x0d029d26	},
+	{	XFER_PIO_0,	0x0d029d5e	},
+	{	0,		0x0d029d26	}
+};
+
+static const struct hpt_clock hpt374_timings_33[] = {
+	{	XFER_UDMA_6,	0x12808242	},
+	{	XFER_UDMA_5,	0x12848242	},
+	{	XFER_UDMA_4,	0x12ac8242	},
+	{	XFER_UDMA_3,	0x128c8242	},
+	{	XFER_UDMA_2,	0x120c8242	},
+	{	XFER_UDMA_1,	0x12148254	},
+	{	XFER_UDMA_0,	0x121882ea	},
+
+	{	XFER_MW_DMA_2,	0x22808242	},
+	{	XFER_MW_DMA_1,	0x22808254	},
+	{	XFER_MW_DMA_0,	0x228082ea	},
+
+	{	XFER_PIO_4,	0x0a81f442	},
+	{	XFER_PIO_3,	0x0a81f443	},
+	{	XFER_PIO_2,	0x0a81f454	},
+	{	XFER_PIO_1,	0x0ac1f465	},
+	{	XFER_PIO_0,	0x0ac1f48a	},
+	{	0,		0x06814e93	}
+};
+
+static const struct hpt_chip hpt370 = {
+	"HPT370",
+	48,
+	{
+		hpt370_timings_33,
+		NULL,
+		NULL,
+		hpt370_timings_66
+	}
+};
+
+static const struct hpt_chip hpt370a = {
+	"HPT370A",
+	48,
+	{
+		hpt370a_timings_33,
+		NULL,
+		hpt370a_timings_50,
+		hpt370a_timings_66
+	}
+};
+
+static const struct hpt_chip hpt372 = {
+	"HPT372",
+	55,
+	{
+		hpt372_timings_33,
+		NULL,
+		hpt372_timings_50,
+		hpt372_timings_66
+	}
+};
+
+static const struct hpt_chip hpt302 = {
+	"HPT302",
+	66,
+	{
+		hpt372_timings_33,
+		NULL,
+		hpt372_timings_50,
+		hpt372_timings_66
+	}
+};
+
+static const struct hpt_chip hpt371 = {
+	"HPT371",
+	66,
+	{
+		hpt372_timings_33,
+		NULL,
+		hpt372_timings_50,
+		hpt372_timings_66
+	}
+};
+
+static const struct hpt_chip hpt372a = {
+	"HPT372A",
+	66,
+	{
+		hpt372_timings_33,
+		NULL,
+		hpt372_timings_50,
+		hpt372_timings_66
+	}
+};
+
+static const struct hpt_chip hpt374 = {
+	"HPT374",
+	48,
+	{
+		hpt374_timings_33,
+		NULL,
+		NULL,
+		NULL
+	}
+};
+
+/**
+ *	hpt37x_find_mode	-	reset the hpt37x bus
+ *	@ap: ATA port
+ *	@speed: transfer mode
+ *
+ *	Return the 32bit register programming information for this channel
+ *	that matches the speed provided.
+ */
+ 
+static u32 hpt37x_find_mode(struct ata_port *ap, int speed)
+{
+	struct hpt_clock *clocks = ap->host->private_data;
+	
+	while(clocks->xfer_speed) {
+		if (clocks->xfer_speed == speed)
+			return clocks->timing;
+		clocks++;
+	}
+	BUG();
+	return 0xffffffffU;	/* silence compiler warning */
+}
+
+static int hpt_dma_blacklisted(const struct ata_device *dev, char *modestr, const char *list[])
+{
+	unsigned char model_num[40];
+	char *s;
+	unsigned int len;
+	int i = 0;
+
+	ata_id_string(dev->id, model_num, ATA_ID_PROD_OFS,
+			  sizeof(model_num));
+	s = &model_num[0];
+	len = strnlen(s, sizeof(model_num));
+
+	/* ATAPI specifies that empty space is blank-filled; remove blanks */
+	while ((len > 0) && (s[len - 1] == ' ')) {
+		len--;
+		s[len] = 0;
+	}
+
+	while(list[i] != NULL) {
+		if (!strncmp(list[i], s, len)) {
+			printk(KERN_WARNING DRV_NAME ": %s is not supported for %s.\n", 
+				modestr, list[i]);
+			return 1;
+		}
+		i++;
+	}
+	return 0;
+}
+
+static const char *bad_ata33[] = {
+	"Maxtor 92720U8", "Maxtor 92040U6", "Maxtor 91360U4", "Maxtor 91020U3", "Maxtor 90845U3", "Maxtor 90650U2",
+	"Maxtor 91360D8", "Maxtor 91190D7", "Maxtor 91020D6", "Maxtor 90845D5", "Maxtor 90680D4", "Maxtor 90510D3", "Maxtor 90340D2",
+	"Maxtor 91152D8", "Maxtor 91008D7", "Maxtor 90845D6", "Maxtor 90840D6", "Maxtor 90720D5", "Maxtor 90648D5", "Maxtor 90576D4",
+	"Maxtor 90510D4",
+	"Maxtor 90432D3", "Maxtor 90288D2", "Maxtor 90256D2",
+	"Maxtor 91000D8", "Maxtor 90910D8", "Maxtor 90875D7", "Maxtor 90840D7", "Maxtor 90750D6", "Maxtor 90625D5", "Maxtor 90500D4",
+	"Maxtor 91728D8", "Maxtor 91512D7", "Maxtor 91303D6", "Maxtor 91080D5", "Maxtor 90845D4", "Maxtor 90680D4", "Maxtor 90648D3", "Maxtor 90432D2",
+	NULL
+};
+
+static const char *bad_ata100_5[] = {
+	"IBM-DTLA-307075",
+	"IBM-DTLA-307060",
+	"IBM-DTLA-307045",
+	"IBM-DTLA-307030",
+	"IBM-DTLA-307020",
+	"IBM-DTLA-307015",
+	"IBM-DTLA-305040",
+	"IBM-DTLA-305030",
+	"IBM-DTLA-305020",
+	"IC35L010AVER07-0",
+	"IC35L020AVER07-0",
+	"IC35L030AVER07-0",
+	"IC35L040AVER07-0",
+	"IC35L060AVER07-0",
+	"WDC AC310200R",
+	NULL
+};
+
+/**
+ *	hpt370_filter	-	mode selection filter
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Block UDMA on devices that cause trouble with this controller.
+ */
+ 
+static unsigned long hpt370_filter(const struct ata_port *ap, struct ata_device *adev, unsigned long mask)
+{
+	if (adev->class != ATA_DEV_ATA) {
+		if (hpt_dma_blacklisted(adev, "UDMA", bad_ata33))
+			mask &= ~ATA_MASK_UDMA;
+		if (hpt_dma_blacklisted(adev, "UDMA100", bad_ata100_5))
+			mask &= ~(0x1F << ATA_SHIFT_UDMA);
+	}
+	return ata_pci_default_filter(ap, adev, mask);
+}
+
+/**
+ *	hpt370a_filter	-	mode selection filter
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Block UDMA on devices that cause trouble with this controller.
+ */
+ 
+static unsigned long hpt370a_filter(const struct ata_port *ap, struct ata_device *adev, unsigned long mask)
+{
+	if (adev->class != ATA_DEV_ATA) {
+		if (hpt_dma_blacklisted(adev, "UDMA100", bad_ata100_5))
+			mask &= ~ (0x1F << ATA_SHIFT_UDMA);
+	}
+	return ata_pci_default_filter(ap, adev, mask);
+}
+	
+/**
+ *	hpt37x_pre_reset	-	reset the hpt37x bus
+ *	@ap: ATA port to reset
+ *
+ *	Perform the initial reset handling for the 370/372 and 374 func 0
+ */
+ 
+static int hpt37x_pre_reset(struct ata_port *ap)
+{
+	u8 scr2, ata66;
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	
+	pci_read_config_byte(pdev, 0x5B, &scr2);
+	pci_write_config_byte(pdev, 0x5B, scr2 & ~0x01);
+	/* Cable register now active */
+	pci_read_config_byte(pdev, 0x5A, &ata66);
+	/* Restore state */
+	pci_write_config_byte(pdev, 0x5B, scr2);
+	
+	if (ata66 & (1 << ap->port_no))
+		ap->cbl = ATA_CBL_PATA40;
+	else
+		ap->cbl = ATA_CBL_PATA80;
+
+	/* Reset the state machine */
+	pci_write_config_byte(pdev, 0x50, 0x37); 
+	pci_write_config_byte(pdev, 0x54, 0x37); 
+	udelay(100);
+	
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	hpt37x_error_handler	-	reset the hpt374
+ *	@ap: ATA port to reset
+ *
+ *	Perform probe for HPT37x, except for HPT374 channel 2
+ */
+ 
+static void hpt37x_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, hpt37x_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+static int hpt374_pre_reset(struct ata_port *ap)
+{
+	u16 mcr3, mcr6;
+	u8 ata66;
+	
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	/* Do the extra channel work */
+	pci_read_config_word(pdev, 0x52, &mcr3);
+	pci_read_config_word(pdev, 0x56, &mcr6);
+	/* Set bit 15 of 0x52 to enable TCBLID as input
+	   Set bit 15 of 0x56 to enable FCBLID as input
+	 */
+	pci_write_config_word(pdev, 0x52, mcr3 | 0x8000);
+	pci_write_config_word(pdev, 0x56, mcr6 | 0x8000);
+	pci_read_config_byte(pdev, 0x5A, &ata66);
+	/* Reset TCBLID/FCBLID to output */
+	pci_write_config_word(pdev, 0x52, mcr3);
+	pci_write_config_word(pdev, 0x56, mcr6);
+	
+	if (ata66 & (1 << ap->port_no))
+		ap->cbl = ATA_CBL_PATA40;
+	else
+		ap->cbl = ATA_CBL_PATA80;
+
+	/* Reset the state machine */
+	pci_write_config_byte(pdev, 0x50, 0x37); 
+	pci_write_config_byte(pdev, 0x54, 0x37); 
+	udelay(100);
+	
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	hpt374_error_handler	-	reset the hpt374
+ *	@classes:
+ *
+ *	The 374 cable detect is a little different due to the extra
+ *	channels. The function 0 channels work like usual but function 1
+ *	is special
+ */
+ 
+static void hpt374_error_handler(struct ata_port *ap)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+		
+	if (!(PCI_FUNC(pdev->devfn) & 1))
+		hpt37x_error_handler(ap);
+	else
+		ata_bmdma_drive_eh(ap, hpt374_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	hpt370_set_piomode		-	PIO setup
+ *	@ap: ATA interface
+ *	@adev: device on the interface
+ *
+ *	Perform PIO mode setup. 
+ */
+ 
+static void hpt370_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u32 addr1, addr2;
+	u32 reg;
+	u32 mode;
+	u8 fast;
+
+	addr1 = 0x40 + 4 * (adev->devno + 2 * ap->port_no);
+	addr2 = 0x51 + 4 * ap->port_no;
+	
+	/* Fast interrupt prediction disable, hold off interrupt disable */
+	pci_read_config_byte(pdev, addr2, &fast);
+	fast &= ~0x02;
+	fast |= 0x01;
+	pci_write_config_byte(pdev, addr2, fast);
+	
+	pci_read_config_dword(pdev, addr1, &reg);
+	mode = hpt37x_find_mode(ap, adev->pio_mode);
+	mode &= ~0x8000000;	/* No FIFO in PIO */
+	mode &= ~0x30070000;	/* Leave config bits alone */
+	reg &= 0x30070000;	/* Strip timing bits */
+	pci_write_config_dword(pdev, addr1, reg | mode);
+}
+
+/**
+ *	hpt370_set_dmamode		-	DMA timing setup
+ *	@ap: ATA interface
+ *	@adev: Device being configured
+ *
+ *	Set up the channel for MWDMA or UDMA modes. Much the same as with
+ *	PIO, load the mode number and then set MWDMA or UDMA flag.
+ */
+ 
+static void hpt370_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u32 addr1, addr2;
+	u32 reg;
+	u32 mode;
+	u8 fast;
+
+	addr1 = 0x40 + 4 * (adev->devno + 2 * ap->port_no);
+	addr2 = 0x51 + 4 * ap->port_no;
+	
+	/* Fast interrupt prediction disable, hold off interrupt disable */
+	pci_read_config_byte(pdev, addr2, &fast);
+	fast &= ~0x02;
+	fast |= 0x01;
+	pci_write_config_byte(pdev, addr2, fast);
+	
+	pci_read_config_dword(pdev, addr1, &reg);
+	mode = hpt37x_find_mode(ap, adev->dma_mode);
+	mode |= 0x8000000;	/* FIFO in MWDMA or UDMA */
+	mode &= ~0xC0000000;	/* Leave config bits alone */
+	reg &= 0xC0000000;	/* Strip timing bits */
+	pci_write_config_dword(pdev, addr1, reg | mode);
+}
+
+/**
+ *	hpt370_bmdma_start		-	DMA engine begin
+ *	@qc: ATA command
+ *
+ *	The 370 and 370A want us to reset the DMA engine each time we
+ *	use it. The 372 and later are fine.
+ */
+ 
+static void hpt370_bmdma_start(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	pci_write_config_byte(pdev, 0x50 + 4 * ap->port_no, 0x37);
+	udelay(10);
+	ata_bmdma_start(qc);
+}
+
+/**
+ *	hpt370_bmdma_end		-	DMA engine stop
+ *	@qc: ATA command
+ *
+ *	Work around the HPT370 DMA engine.
+ */
+ 
+static void hpt370_bmdma_stop(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u8 dma_stat = inb(ap->ioaddr.bmdma_addr + 2);
+	u8 dma_cmd;
+	unsigned long bmdma = ap->ioaddr.bmdma_addr;
+	
+	if (dma_stat & 0x01) {
+		udelay(20);
+		dma_stat = inb(bmdma + 2);
+	}
+	if (dma_stat & 0x01) {
+		/* Clear the engine */
+		pci_write_config_byte(pdev, 0x50 + 4 * ap->port_no, 0x37);
+		udelay(10);
+		/* Stop DMA */
+		dma_cmd = inb(bmdma );
+		outb(dma_cmd & 0xFE, bmdma);
+		/* Clear Error */
+		dma_stat = inb(bmdma + 2);
+		outb(dma_stat | 0x06 , bmdma + 2);
+		/* Clear the engine */
+		pci_write_config_byte(pdev, 0x50 + 4 * ap->port_no, 0x37);
+		udelay(10);
+	}
+	ata_bmdma_stop(qc);
+}
+
+/**
+ *	hpt372_set_piomode		-	PIO setup
+ *	@ap: ATA interface
+ *	@adev: device on the interface
+ *
+ *	Perform PIO mode setup. 
+ */
+ 
+static void hpt372_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u32 addr1, addr2;
+	u32 reg;
+	u32 mode;
+	u8 fast;
+
+	addr1 = 0x40 + 4 * (adev->devno + 2 * ap->port_no);
+	addr2 = 0x51 + 4 * ap->port_no;
+	
+	/* Fast interrupt prediction disable, hold off interrupt disable */
+	pci_read_config_byte(pdev, addr2, &fast);
+	fast &= ~0x07;
+	pci_write_config_byte(pdev, addr2, fast);
+	
+	pci_read_config_dword(pdev, addr1, &reg);
+	mode = hpt37x_find_mode(ap, adev->pio_mode);
+	
+	printk("Find mode for %d reports %X\n", adev->pio_mode, mode);
+	mode &= ~0x80000000;	/* No FIFO in PIO */
+	mode &= ~0x30070000;	/* Leave config bits alone */
+	reg &= 0x30070000;	/* Strip timing bits */
+	pci_write_config_dword(pdev, addr1, reg | mode);
+}
+
+/**
+ *	hpt372_set_dmamode		-	DMA timing setup
+ *	@ap: ATA interface
+ *	@adev: Device being configured
+ *
+ *	Set up the channel for MWDMA or UDMA modes. Much the same as with
+ *	PIO, load the mode number and then set MWDMA or UDMA flag.
+ */
+ 
+static void hpt372_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u32 addr1, addr2;
+	u32 reg;
+	u32 mode;
+	u8 fast;
+
+	addr1 = 0x40 + 4 * (adev->devno + 2 * ap->port_no);
+	addr2 = 0x51 + 4 * ap->port_no;
+	
+	/* Fast interrupt prediction disable, hold off interrupt disable */
+	pci_read_config_byte(pdev, addr2, &fast);
+	fast &= ~0x07;
+	pci_write_config_byte(pdev, addr2, fast);
+	
+	pci_read_config_dword(pdev, addr1, &reg);
+	mode = hpt37x_find_mode(ap, adev->dma_mode);
+	printk("Find mode for DMA %d reports %X\n", adev->dma_mode, mode);
+	mode &= ~0xC0000000;	/* Leave config bits alone */
+	mode |= 0x80000000;	/* FIFO in MWDMA or UDMA */
+	reg &= 0xC0000000;	/* Strip timing bits */
+	pci_write_config_dword(pdev, addr1, reg | mode);
+}
+
+/**
+ *	hpt37x_bmdma_end		-	DMA engine stop
+ *	@qc: ATA command
+ *
+ *	Clean up after the HPT372 and later DMA engine
+ */
+ 
+static void hpt37x_bmdma_stop(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	int mscreg = 0x50 + 2 * ap->port_no;
+	u8 bwsr_stat, msc_stat;
+	
+	pci_read_config_byte(pdev, 0x6A, &bwsr_stat);
+	pci_read_config_byte(pdev, mscreg, &msc_stat);
+	if (bwsr_stat & (1 << ap->port_no))
+		pci_write_config_byte(pdev, mscreg, msc_stat | 0x30);
+	ata_bmdma_stop(qc);
+}
+
+
+static struct scsi_host_template hpt37x_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+/*
+ *	Configuration for HPT370
+ */
+ 
+static struct ata_port_operations hpt370_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= hpt370_set_piomode,
+	.set_dmamode	= hpt370_set_dmamode,
+	.mode_filter	= hpt370_filter,
+	
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= hpt37x_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= hpt370_bmdma_start,
+	.bmdma_stop	= hpt370_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+/*
+ *	Configuration for HPT370A. Close to 370 but less filters
+ */
+ 
+static struct ata_port_operations hpt370a_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= hpt370_set_piomode,
+	.set_dmamode	= hpt370_set_dmamode,
+	.mode_filter	= hpt370a_filter,
+	
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= hpt37x_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= hpt370_bmdma_start,
+	.bmdma_stop	= hpt370_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+/*
+ *	Configuration for HPT372, HPT371, HPT302. Slightly different PIO
+ *	and DMA mode setting functionality.
+ */
+ 
+static struct ata_port_operations hpt372_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= hpt372_set_piomode,
+	.set_dmamode	= hpt372_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= hpt37x_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= hpt37x_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+/*
+ *	Configuration for HPT374. Mode setting works like 372 and friends
+ *	but we have a different cable detection procedure.
+ */
+ 
+static struct ata_port_operations hpt374_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= hpt372_set_piomode,
+	.set_dmamode	= hpt372_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= hpt374_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= hpt37x_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+/**
+ *	htp37x_clock_slot	-	Turn timing to PC clock entry
+ *	@freq: Reported frequency timing
+ *	@base: Base timing
+ *
+ *	Turn the timing data intoa clock slot (0 for 33, 1 for 40, 2 for 50
+ *	and 3 for 66Mhz)
+ */
+ 
+static int hpt37x_clock_slot(unsigned int freq, unsigned int base)
+{
+	unsigned int f = (base * freq) / 192;	/* Mhz */
+	if (f < 40)
+		return 0;	/* 33Mhz slot */
+	if (f < 45)
+		return 1;	/* 40Mhz slot */
+	if (f < 55)
+		return 2;	/* 50Mhz slot */
+	return 3;		/* 60Mhz slot */
+}
+
+/**
+ *	hpt37x_calibrate_dpll		-	Calibrate the DPLL loop
+ *	@dev: PCI device 
+ *
+ *	Perform a calibration cycle on the HPT37x DPLL. Returns 1 if this
+ *	succeeds
+ */
+
+static int hpt37x_calibrate_dpll(struct pci_dev *dev)
+{
+	u8 reg5b;
+	u32 reg5c;
+	int tries;
+ 
+	for(tries = 0; tries < 0x5000; tries++) {
+		udelay(50);
+		pci_read_config_byte(dev, 0x5b, &reg5b);
+		if (reg5b & 0x80) {
+			/* See if it stays set */
+			for(tries = 0; tries < 0x1000; tries ++) {
+				pci_read_config_byte(dev, 0x5b, &reg5b);
+				/* Failed ? */
+				if ((reg5b & 0x80) == 0)
+					return 0;
+			}
+			/* Turn off tuning, we have the DPLL set */
+			pci_read_config_dword(dev, 0x5c, &reg5c);
+			pci_write_config_dword(dev, 0x5c, reg5c & ~ 0x100);
+			return 1;
+		}
+	}
+	/* Never went stable */
+	return 0;
+}
+/**
+ *	hpt37x_init_one		-	Initialise an HPT37X/302
+ *	@dev: PCI device
+ *	@id: Entry in match table
+ *
+ *	Initialise an HPT37x device. There are some interesting complications
+ *	here. Firstly the chip may report 366 and be one of several variants.
+ *	Secondly all the timings depend on the clock for the chip which we must
+ *	detect and look up
+ *
+ *	This is the known chip mappings. It may be missing a couple of later
+ *	releases.
+ *
+ *	Chip version		PCI		Rev	Notes
+ *	HPT366			4 (HPT366)	0	Other driver
+ *	HPT366			4 (HPT366)	1	Other driver
+ *	HPT368			4 (HPT366)	2	Other driver
+ *	HPT370			4 (HPT366)	3	UDMA100
+ *	HPT370A			4 (HPT366)	4	UDMA100
+ *	HPT372			4 (HPT366)	5	UDMA133 (1)
+ *	HPT372N			4 (HPT366)	6	Other driver
+ *	HPT372A			5 (HPT372)	1	UDMA133 (1)
+ *	HPT372N			5 (HPT372)	2	Other driver
+ *	HPT302			6 (HPT302)	1	UDMA133
+ *	HPT302N			6 (HPT302)	2	Other driver
+ *	HPT371			7 (HPT371)	*	UDMA133
+ *	HPT374			8 (HPT374)	*	UDMA133 4 channel
+ *	HPT372N			9 (HPT372N)	*	Other driver
+ *
+ *	(1) UDMA133 support depends on the bus clock
+ */
+ 
+static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	/* HPT370 - UDMA100 */
+	static struct ata_port_info info_hpt370 = {
+		.sht = &hpt37x_sht,
+		.flags = ATA_FLAG_SLAVE_POSS|ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x3f,
+		.port_ops = &hpt370_port_ops
+	};
+	/* HPT370A - UDMA100 */
+	static struct ata_port_info info_hpt370a = {
+		.sht = &hpt37x_sht,
+		.flags = ATA_FLAG_SLAVE_POSS|ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x3f,
+		.port_ops = &hpt370a_port_ops
+	};
+	/* HPT371, 372 and friends - UDMA133 */
+	static struct ata_port_info info_hpt372 = {
+		.sht = &hpt37x_sht,
+		.flags = ATA_FLAG_SLAVE_POSS|ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x7f,
+		.port_ops = &hpt372_port_ops
+	};
+	/* HPT371, 372 and friends - UDMA100 at 50MHz clock */
+	static struct ata_port_info info_hpt372_50 = {
+		.sht = &hpt37x_sht,
+		.flags = ATA_FLAG_SLAVE_POSS|ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x3f,
+		.port_ops = &hpt372_port_ops
+	};
+	/* HPT374 - UDMA133 */
+	static struct ata_port_info info_hpt374 = {
+		.sht = &hpt37x_sht,
+		.flags = ATA_FLAG_SLAVE_POSS|ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x7f,
+		.port_ops = &hpt374_port_ops
+	};
+
+	static const int MHz[4] = { 33, 40, 50, 66 };
+
+	struct ata_port_info *port_info[2];
+	struct ata_port_info *port;
+
+	u8 irqmask;
+	u32 class_rev;
+	u32 freq;
+
+	const struct hpt_chip *chip_table;
+	int clock_slot;
+
+	pci_read_config_dword(dev, PCI_CLASS_REVISION, &class_rev);
+	class_rev &= 0xFF;
+	
+	if (dev->device == PCI_DEVICE_ID_TTI_HPT366) {
+		/* May be a later chip in disguise. Check */
+		/* Older chips are in the HPT366 driver. Ignore them */
+		if (class_rev < 3)
+			return -ENODEV;
+		/* N series chips have their own driver. Ignore */
+		if (class_rev == 6)
+			return -ENODEV;
+
+		switch(class_rev) {			
+			case 3:
+				port = &info_hpt370;
+				chip_table = &hpt370;
+				break;
+			case 4:
+				port = &info_hpt370a;
+				chip_table = &hpt370a;
+				break;
+			case 5:
+				port = &info_hpt372;
+				chip_table = &hpt372;
+				break;
+			default:
+				printk(KERN_ERR "pata_hpt37x: Unknown HPT366 subtype please report (%d).\n", class_rev);
+				return -ENODEV;
+		}
+	} else {
+		switch(dev->device) {
+			case PCI_DEVICE_ID_TTI_HPT372:
+				/* 372N if rev >= 2*/
+				if (class_rev >= 2)
+					return -ENODEV;
+				port = &info_hpt372;
+				chip_table = &hpt372a;
+				break;
+			case PCI_DEVICE_ID_TTI_HPT302:
+				/* 302N if rev > 1 */
+				if (class_rev > 1)
+					return -ENODEV;
+				port = &info_hpt372;
+				/* Check this */
+				chip_table = &hpt302;
+				break;
+			case PCI_DEVICE_ID_TTI_HPT371:
+				port = &info_hpt372;
+				chip_table = &hpt371;
+				break;
+			case PCI_DEVICE_ID_TTI_HPT374:
+				chip_table = &hpt374;
+				port = &info_hpt374;
+				break;
+			default:
+				printk(KERN_ERR "pata_hpt37x: PCI table is bogus please report (%d).\n", dev->device);
+				return -ENODEV;
+		}
+	}
+	/* Ok so this is a chip we support */
+
+	pci_write_config_byte(dev, PCI_CACHE_LINE_SIZE, (L1_CACHE_BYTES / 4));
+	pci_write_config_byte(dev, PCI_LATENCY_TIMER, 0x78);
+	pci_write_config_byte(dev, PCI_MIN_GNT, 0x08);
+	pci_write_config_byte(dev, PCI_MAX_LAT, 0x08);
+
+	pci_read_config_byte(dev, 0x5A, &irqmask);
+	irqmask &= ~0x10;
+	pci_write_config_byte(dev, 0x5a, irqmask);
+
+	/*
+	 * default to pci clock. make sure MA15/16 are set to output
+	 * to prevent drives having problems with 40-pin cables. Needed
+	 * for some drives such as IBM-DTLA which will not enter ready
+	 * state on reset when PDIAG is a input.
+	 */
+
+	pci_write_config_byte(dev, 0x5b, 0x23);	
+	
+	pci_read_config_dword(dev, 0x70, &freq);
+	if ((freq >> 12) != 0xABCDE) {
+		int i;
+		u8 sr;
+		u32 total = 0;
+		
+		printk(KERN_WARNING "pata_hpt37x: BIOS has not set timing clocks.\n");
+		
+		/* This is the process the HPT371 BIOS is reported to use */
+		for(i = 0; i < 128; i++) {
+			pci_read_config_byte(dev, 0x78, &sr);
+			total += sr;
+			udelay(15);
+		}
+		freq = total / 128;
+	}
+	freq &= 0x1FF;
+	
+	/*
+	 *	Turn the frequency check into a band and then find a timing
+	 *	table to match it.
+	 */
+	 
+	clock_slot = hpt37x_clock_slot(freq, chip_table->base);
+	if (chip_table->clocks[clock_slot] == NULL) {
+		/*
+		 *	We need to try PLL mode instead
+		 */
+		unsigned int f_low = (MHz[clock_slot] * chip_table->base) / 192;
+		unsigned int f_high = f_low + 2;
+		int adjust;
+		
+		for(adjust = 0; adjust < 8; adjust++) {
+			if (hpt37x_calibrate_dpll(dev))
+				break;
+			/* See if it'll settle at a fractionally different clock */
+			if ((adjust & 3) == 3) {
+				f_low --;
+				f_high ++;
+			}
+			pci_write_config_dword(dev, 0x5C, (f_high << 16) | f_low);
+		}
+		if (adjust == 8) {
+			printk(KERN_WARNING "hpt37x: DPLL did not stabilize.\n");
+			return -ENODEV;
+		}
+		/* Check if this works for all cases */
+		port->private_data = (void *)hpt370_timings_66;
+		
+		printk(KERN_INFO "hpt37x: Bus clock %dMHz, using DPLL.\n", MHz[clock_slot]);
+	} else {
+		port->private_data = (void *)chip_table->clocks[clock_slot];
+		/*
+		 *	Perform a final fixup. The 371 and 372 clock determines
+		 *	if UDMA133 is available.
+		 */
+	 
+		if (clock_slot == 2 && chip_table == &hpt372) {	/* 50Mhz */
+			printk(KERN_WARNING "pata_hpt37x: No UDMA133 support available with 50MHz bus clock.\n");
+			if (port == &info_hpt372)
+				port = &info_hpt372_50;
+			else BUG();
+		}
+		printk(KERN_INFO "hpt37x: %s: Bus clock %dMHz.\n", chip_table->name, MHz[clock_slot]);
+	}
+	port_info[0] = port_info[1] = port;	
+	/* Now kick off ATA set up */
+	return ata_pci_init_one(dev, port_info, 2);
+}
+
+static struct pci_device_id hpt37x[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_TTI, PCI_DEVICE_ID_TTI_HPT366), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_TTI, PCI_DEVICE_ID_TTI_HPT371), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_TTI, PCI_DEVICE_ID_TTI_HPT372), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_TTI, PCI_DEVICE_ID_TTI_HPT374), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_TTI, PCI_DEVICE_ID_TTI_HPT302), },
+	{ 0, },
+};
+
+static struct pci_driver hpt37x_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= hpt37x,
+	.probe 		= hpt37x_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init hpt37x_init(void)
+{
+	return pci_register_driver(&hpt37x_pci_driver);
+}
+
+
+static void __exit hpt37x_exit(void)
+{
+	pci_unregister_driver(&hpt37x_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for the Highpoint HPT37x/30x");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, hpt37x);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(hpt37x_init);
+module_exit(hpt37x_exit);
diff --git a/drivers/ata/pata_hpt3x2n.c b/drivers/ata/pata_hpt3x2n.c
new file mode 100644
index 000000000000..40fcda62c7a2
--- /dev/null
+++ b/drivers/ata/pata_hpt3x2n.c
@@ -0,0 +1,597 @@
+/*
+ * Libata driver for the highpoint 372N and 302N UDMA66 ATA controllers.
+ *
+ * This driver is heavily based upon:
+ *
+ * linux/drivers/ide/pci/hpt366.c		Version 0.36	April 25, 2003
+ *
+ * Copyright (C) 1999-2003		Andre Hedrick <andre@linux-ide.org>
+ * Portions Copyright (C) 2001	        Sun Microsystems, Inc.
+ * Portions Copyright (C) 2003		Red Hat Inc
+ *
+ *
+ * TODO
+ *	371N
+ *	Work out best PLL policy
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME	"pata_hpt3x2n"
+#define DRV_VERSION	"0.3"
+
+enum {
+	HPT_PCI_FAST	=	(1 << 31),
+	PCI66		=	(1 << 1),
+	USE_DPLL	=	(1 << 0)
+};
+
+struct hpt_clock {
+	u8	xfer_speed;
+	u32	timing;
+};
+
+struct hpt_chip {
+	const char *name;
+	struct hpt_clock *clocks[3];
+};
+
+/* key for bus clock timings
+ * bit
+ * 0:3    data_high_time. inactive time of DIOW_/DIOR_ for PIO and MW
+ *        DMA. cycles = value + 1
+ * 4:8    data_low_time. active time of DIOW_/DIOR_ for PIO and MW
+ *        DMA. cycles = value + 1
+ * 9:12   cmd_high_time. inactive time of DIOW_/DIOR_ during task file
+ *        register access.
+ * 13:17  cmd_low_time. active time of DIOW_/DIOR_ during task file
+ *        register access.
+ * 18:21  udma_cycle_time. clock freq and clock cycles for UDMA xfer.
+ *        during task file register access.
+ * 22:24  pre_high_time. time to initialize 1st cycle for PIO and MW DMA
+ *        xfer.
+ * 25:27  cmd_pre_high_time. time to initialize 1st PIO cycle for task
+ *        register access.
+ * 28     UDMA enable
+ * 29     DMA enable
+ * 30     PIO_MST enable. if set, the chip is in bus master mode during
+ *        PIO.
+ * 31     FIFO enable.
+ */
+ 
+/* 66MHz DPLL clocks */
+
+static struct hpt_clock hpt3x2n_clocks[] = {
+	{	XFER_UDMA_7,	0x1c869c62	},
+	{	XFER_UDMA_6,	0x1c869c62	},
+	{	XFER_UDMA_5,	0x1c8a9c62	},
+	{	XFER_UDMA_4,	0x1c8a9c62	},
+	{	XFER_UDMA_3,	0x1c8e9c62	},
+	{	XFER_UDMA_2,	0x1c929c62	},
+	{	XFER_UDMA_1,	0x1c9a9c62	},
+	{	XFER_UDMA_0,	0x1c829c62	},
+
+	{	XFER_MW_DMA_2,	0x2c829c62	},
+	{	XFER_MW_DMA_1,	0x2c829c66	},
+	{	XFER_MW_DMA_0,	0x2c829d2c	},
+
+	{	XFER_PIO_4,	0x0c829c62	},
+	{	XFER_PIO_3,	0x0c829c84	},
+	{	XFER_PIO_2,	0x0c829ca6	},
+	{	XFER_PIO_1,	0x0d029d26	},
+	{	XFER_PIO_0,	0x0d029d5e	},
+	{	0,		0x0d029d5e	}
+};
+
+/**
+ *	hpt3x2n_find_mode	-	reset the hpt3x2n bus
+ *	@ap: ATA port
+ *	@speed: transfer mode
+ *
+ *	Return the 32bit register programming information for this channel
+ *	that matches the speed provided. For the moment the clocks table
+ *	is hard coded but easy to change. This will be needed if we use
+ *	different DPLLs
+ */
+ 
+static u32 hpt3x2n_find_mode(struct ata_port *ap, int speed)
+{
+	struct hpt_clock *clocks = hpt3x2n_clocks;
+	
+	while(clocks->xfer_speed) {
+		if (clocks->xfer_speed == speed)
+			return clocks->timing;
+		clocks++;
+	}
+	BUG();
+	return 0xffffffffU;	/* silence compiler warning */
+}
+
+/**
+ *	hpt3x2n_pre_reset	-	reset the hpt3x2n bus
+ *	@ap: ATA port to reset
+ *
+ *	Perform the initial reset handling for the 3x2n series controllers.
+ *	Reset the hardware and state machine, obtain the cable type.
+ */
+ 
+static int hpt3xn_pre_reset(struct ata_port *ap)
+{
+	u8 scr2, ata66;
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	
+	pci_read_config_byte(pdev, 0x5B, &scr2);
+	pci_write_config_byte(pdev, 0x5B, scr2 & ~0x01);
+	/* Cable register now active */
+	pci_read_config_byte(pdev, 0x5A, &ata66);
+	/* Restore state */
+	pci_write_config_byte(pdev, 0x5B, scr2);
+	
+	if (ata66 & (1 << ap->port_no))
+		ap->cbl = ATA_CBL_PATA40;
+	else
+		ap->cbl = ATA_CBL_PATA80;
+
+	/* Reset the state machine */
+	pci_write_config_byte(pdev, 0x50, 0x37); 
+	pci_write_config_byte(pdev, 0x54, 0x37); 
+	udelay(100);
+
+	return ata_std_prereset(ap);
+}
+	
+/**
+ *	hpt3x2n_error_handler	-	probe the hpt3x2n bus
+ *	@ap: ATA port to reset
+ *
+ *	Perform the probe reset handling for the 3x2N
+ */
+ 
+static void hpt3x2n_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, hpt3xn_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+				     
+/**
+ *	hpt3x2n_set_piomode		-	PIO setup
+ *	@ap: ATA interface
+ *	@adev: device on the interface
+ *
+ *	Perform PIO mode setup. 
+ */
+ 
+static void hpt3x2n_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u32 addr1, addr2;
+	u32 reg;
+	u32 mode;
+	u8 fast;
+
+	addr1 = 0x40 + 4 * (adev->devno + 2 * ap->port_no);
+	addr2 = 0x51 + 4 * ap->port_no;
+	
+	/* Fast interrupt prediction disable, hold off interrupt disable */
+	pci_read_config_byte(pdev, addr2, &fast);
+	fast &= ~0x07;
+	pci_write_config_byte(pdev, addr2, fast);
+	
+	pci_read_config_dword(pdev, addr1, &reg);
+	mode = hpt3x2n_find_mode(ap, adev->pio_mode);
+	mode &= ~0x8000000;	/* No FIFO in PIO */
+	mode &= ~0x30070000;	/* Leave config bits alone */
+	reg &= 0x30070000;	/* Strip timing bits */
+	pci_write_config_dword(pdev, addr1, reg | mode);
+}
+
+/**
+ *	hpt3x2n_set_dmamode		-	DMA timing setup
+ *	@ap: ATA interface
+ *	@adev: Device being configured
+ *
+ *	Set up the channel for MWDMA or UDMA modes. Much the same as with
+ *	PIO, load the mode number and then set MWDMA or UDMA flag.
+ */
+ 
+static void hpt3x2n_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u32 addr1, addr2;
+	u32 reg;
+	u32 mode;
+	u8 fast;
+
+	addr1 = 0x40 + 4 * (adev->devno + 2 * ap->port_no);
+	addr2 = 0x51 + 4 * ap->port_no;
+	
+	/* Fast interrupt prediction disable, hold off interrupt disable */
+	pci_read_config_byte(pdev, addr2, &fast);
+	fast &= ~0x07;
+	pci_write_config_byte(pdev, addr2, fast);
+	
+	pci_read_config_dword(pdev, addr1, &reg);
+	mode = hpt3x2n_find_mode(ap, adev->dma_mode);
+	mode |= 0x8000000;	/* FIFO in MWDMA or UDMA */
+	mode &= ~0xC0000000;	/* Leave config bits alone */
+	reg &= 0xC0000000;	/* Strip timing bits */
+	pci_write_config_dword(pdev, addr1, reg | mode);
+}
+
+/**
+ *	hpt3x2n_bmdma_end		-	DMA engine stop
+ *	@qc: ATA command
+ *
+ *	Clean up after the HPT3x2n and later DMA engine
+ */
+ 
+static void hpt3x2n_bmdma_stop(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	int mscreg = 0x50 + 2 * ap->port_no;
+	u8 bwsr_stat, msc_stat;
+	
+	pci_read_config_byte(pdev, 0x6A, &bwsr_stat);
+	pci_read_config_byte(pdev, mscreg, &msc_stat);
+	if (bwsr_stat & (1 << ap->port_no))
+		pci_write_config_byte(pdev, mscreg, msc_stat | 0x30);
+	ata_bmdma_stop(qc);
+}
+
+/**
+ *	hpt3x2n_set_clock	-	clock control
+ *	@ap: ATA port
+ *	@source: 0x21 or 0x23 for PLL or PCI sourced clock
+ *
+ *	Switch the ATA bus clock between the PLL and PCI clock sources
+ *	while correctly isolating the bus and resetting internal logic
+ *
+ *	We must use the DPLL for
+ *	-	writing
+ *	-	second channel UDMA7 (SATA ports) or higher
+ *	-	66MHz PCI
+ *	
+ *	or we will underclock the device and get reduced performance.
+ */
+ 
+static void hpt3x2n_set_clock(struct ata_port *ap, int source)
+{
+	unsigned long bmdma = ap->ioaddr.bmdma_addr;
+	
+	/* Tristate the bus */
+	outb(0x80, bmdma+0x73);
+	outb(0x80, bmdma+0x77);
+	
+	/* Switch clock and reset channels */
+	outb(source, bmdma+0x7B);
+	outb(0xC0, bmdma+0x79);
+	
+	/* Reset state machines */
+	outb(0x37, bmdma+0x70);
+	outb(0x37, bmdma+0x74);
+	
+	/* Complete reset */
+	outb(0x00, bmdma+0x79);
+	
+	/* Reconnect channels to bus */
+	outb(0x00, bmdma+0x73);
+	outb(0x00, bmdma+0x77);
+}
+
+/* Check if our partner interface is busy */
+
+static int hpt3x2n_pair_idle(struct ata_port *ap)
+{
+	struct ata_host *host = ap->host;
+	struct ata_port *pair = host->ports[ap->port_no ^ 1];
+	
+	if (pair->hsm_task_state == HSM_ST_IDLE)
+		return 1;
+	return 0;
+}
+
+static int hpt3x2n_use_dpll(struct ata_port *ap, int reading)
+{
+	long flags = (long)ap->host->private_data;
+	/* See if we should use the DPLL */
+	if (reading == 0)
+		return USE_DPLL;	/* Needed for write */
+	if (flags & PCI66)
+		return USE_DPLL;	/* Needed at 66Mhz */
+	return 0;	
+}
+
+static unsigned int hpt3x2n_qc_issue_prot(struct ata_queued_cmd *qc)
+{
+	struct ata_taskfile *tf = &qc->tf;
+	struct ata_port *ap = qc->ap;
+	int flags = (long)ap->host->private_data;
+	
+	if (hpt3x2n_pair_idle(ap)) {
+		int dpll = hpt3x2n_use_dpll(ap, (tf->flags & ATA_TFLAG_WRITE));
+		if ((flags & USE_DPLL) != dpll) {
+			if (dpll == 1)
+				hpt3x2n_set_clock(ap, 0x21);
+			else
+				hpt3x2n_set_clock(ap, 0x23);
+		}
+	}
+	return ata_qc_issue_prot(qc);
+}
+
+static struct scsi_host_template hpt3x2n_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+/*
+ *	Configuration for HPT3x2n.
+ */
+ 
+static struct ata_port_operations hpt3x2n_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= hpt3x2n_set_piomode,
+	.set_dmamode	= hpt3x2n_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= hpt3x2n_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= hpt3x2n_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= hpt3x2n_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+/**
+ *	hpt3xn_calibrate_dpll		-	Calibrate the DPLL loop
+ *	@dev: PCI device 
+ *
+ *	Perform a calibration cycle on the HPT3xN DPLL. Returns 1 if this
+ *	succeeds
+ */
+
+static int hpt3xn_calibrate_dpll(struct pci_dev *dev)
+{
+	u8 reg5b;
+	u32 reg5c;
+	int tries;
+ 
+	for(tries = 0; tries < 0x5000; tries++) {
+		udelay(50);
+		pci_read_config_byte(dev, 0x5b, &reg5b);
+		if (reg5b & 0x80) {
+			/* See if it stays set */
+			for(tries = 0; tries < 0x1000; tries ++) {
+				pci_read_config_byte(dev, 0x5b, &reg5b);
+				/* Failed ? */
+				if ((reg5b & 0x80) == 0)
+					return 0;
+			}
+			/* Turn off tuning, we have the DPLL set */
+			pci_read_config_dword(dev, 0x5c, &reg5c);
+			pci_write_config_dword(dev, 0x5c, reg5c & ~ 0x100);
+			return 1;
+		}
+	}
+	/* Never went stable */
+	return 0;
+}
+
+static int hpt3x2n_pci_clock(struct pci_dev *pdev)
+{
+	unsigned long freq;
+	u32 fcnt;
+	
+	pci_read_config_dword(pdev, 0x70/*CHECKME*/, &fcnt);
+	if ((fcnt >> 12) != 0xABCDE) {
+		printk(KERN_WARNING "hpt3xn: BIOS clock data not set.\n");
+		return 33;	/* Not BIOS set */
+	}
+	fcnt &= 0x1FF;
+	
+	freq = (fcnt * 77) / 192;
+	
+	/* Clamp to bands */
+	if (freq < 40)
+		return 33;
+	if (freq < 45)
+		return 40;
+	if (freq < 55)
+		return 50;
+	return 66;
+}
+
+/**
+ *	hpt3x2n_init_one		-	Initialise an HPT37X/302
+ *	@dev: PCI device
+ *	@id: Entry in match table
+ *
+ *	Initialise an HPT3x2n device. There are some interesting complications
+ *	here. Firstly the chip may report 366 and be one of several variants.
+ *	Secondly all the timings depend on the clock for the chip which we must
+ *	detect and look up
+ *
+ *	This is the known chip mappings. It may be missing a couple of later
+ *	releases.
+ *
+ *	Chip version		PCI		Rev	Notes
+ *	HPT372			4 (HPT366)	5	Other driver
+ *	HPT372N			4 (HPT366)	6	UDMA133
+ *	HPT372			5 (HPT372)	1	Other driver
+ *	HPT372N			5 (HPT372)	2	UDMA133
+ *	HPT302			6 (HPT302)	*	Other driver
+ *	HPT302N			6 (HPT302)	> 1	UDMA133
+ *	HPT371			7 (HPT371)	*	Other driver
+ *	HPT371N			7 (HPT371)	> 1	UDMA133
+ *	HPT374			8 (HPT374)	*	Other driver
+ *	HPT372N			9 (HPT372N)	*	UDMA133
+ *
+ *	(1) UDMA133 support depends on the bus clock
+ *
+ *	To pin down		HPT371N
+ */
+ 
+static int hpt3x2n_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	/* HPT372N and friends - UDMA133 */
+	static struct ata_port_info info = {
+		.sht = &hpt3x2n_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x7f,
+		.port_ops = &hpt3x2n_port_ops
+	};
+	struct ata_port_info *port_info[2];
+	struct ata_port_info *port = &info;
+
+	u8 irqmask;
+	u32 class_rev;
+	
+	unsigned int pci_mhz;
+	unsigned int f_low, f_high;
+	int adjust;
+	
+	pci_read_config_dword(dev, PCI_CLASS_REVISION, &class_rev);
+	class_rev &= 0xFF;
+	
+	switch(dev->device) {
+		case PCI_DEVICE_ID_TTI_HPT366:
+			if (class_rev < 6)
+				return -ENODEV;
+			break;
+		case PCI_DEVICE_ID_TTI_HPT372:
+			/* 372N if rev >= 1*/
+			if (class_rev == 0)
+				return -ENODEV;
+			break;
+		case PCI_DEVICE_ID_TTI_HPT302:
+			if (class_rev < 2)
+				return -ENODEV;
+			break;
+		case PCI_DEVICE_ID_TTI_HPT372N:
+			break;
+		default:
+			printk(KERN_ERR "pata_hpt3x2n: PCI table is bogus please report (%d).\n", dev->device);
+			return -ENODEV;
+	}
+
+	/* Ok so this is a chip we support */
+
+	pci_write_config_byte(dev, PCI_CACHE_LINE_SIZE, (L1_CACHE_BYTES / 4));
+	pci_write_config_byte(dev, PCI_LATENCY_TIMER, 0x78);
+	pci_write_config_byte(dev, PCI_MIN_GNT, 0x08);
+	pci_write_config_byte(dev, PCI_MAX_LAT, 0x08);
+
+	pci_read_config_byte(dev, 0x5A, &irqmask);
+	irqmask &= ~0x10;
+	pci_write_config_byte(dev, 0x5a, irqmask);
+
+	/* Tune the PLL. HPT recommend using 75 for SATA, 66 for UDMA133 or
+	   50 for UDMA100. Right now we always use 66 */
+	   
+	pci_mhz = hpt3x2n_pci_clock(dev);
+	
+	f_low = (pci_mhz * 48) / 66;	/* PCI Mhz for 66Mhz DPLL */
+	f_high = f_low + 2;		/* Tolerance */
+	
+	pci_write_config_dword(dev, 0x5C, (f_high << 16) | f_low | 0x100);
+	/* PLL clock */
+	pci_write_config_byte(dev, 0x5B, 0x21);
+	
+	/* Unlike the 37x we don't try jiggling the frequency */
+	for(adjust = 0; adjust < 8; adjust++) {
+		if (hpt3xn_calibrate_dpll(dev))
+			break;
+		pci_write_config_dword(dev, 0x5C, (f_high << 16) | f_low);
+	}
+	if (adjust == 8)
+		printk(KERN_WARNING "hpt3xn: DPLL did not stabilize.\n");
+
+	/* Set our private data up. We only need a few flags so we use
+	   it directly */
+	port->private_data = NULL;
+	if (pci_mhz > 60)
+		port->private_data = (void *)PCI66;
+		
+	/* Now kick off ATA set up */
+	port_info[0] = port_info[1] = port;
+	return ata_pci_init_one(dev, port_info, 2);
+}
+
+static struct pci_device_id hpt3x2n[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_TTI, PCI_DEVICE_ID_TTI_HPT366), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_TTI, PCI_DEVICE_ID_TTI_HPT372), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_TTI, PCI_DEVICE_ID_TTI_HPT302), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_TTI, PCI_DEVICE_ID_TTI_HPT372N), },
+	{ 0, },
+};
+
+static struct pci_driver hpt3x2n_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= hpt3x2n,
+	.probe 		= hpt3x2n_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init hpt3x2n_init(void)
+{
+	return pci_register_driver(&hpt3x2n_pci_driver);
+}
+
+
+static void __exit hpt3x2n_exit(void)
+{
+	pci_unregister_driver(&hpt3x2n_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for the Highpoint HPT3x2n/30x");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, hpt3x2n);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(hpt3x2n_init);
+module_exit(hpt3x2n_exit);
diff --git a/drivers/ata/pata_hpt3x3.c b/drivers/ata/pata_hpt3x3.c
new file mode 100644
index 000000000000..c93406ebf6ed
--- /dev/null
+++ b/drivers/ata/pata_hpt3x3.c
@@ -0,0 +1,226 @@
+/*
+ *	pata_hpt3x3		-	HPT3x3 driver
+ *	(c) Copyright 2005-2006 Red Hat
+ *
+ *	Was pata_hpt34x but the naming was confusing as it supported the
+ *	343 and 363 so it has been renamed.
+ *
+ *	Based on:
+ *	linux/drivers/ide/pci/hpt34x.c		Version 0.40	Sept 10, 2002
+ *	Copyright (C) 1998-2000	Andre Hedrick <andre@linux-ide.org>
+ *
+ *	May be copied or modified under the terms of the GNU General Public
+ *	License
+ */
+ 
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME	"pata_hpt3x3"
+#define DRV_VERSION	"0.4.1"
+
+static int hpt3x3_probe_init(struct ata_port *ap)
+{
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	hpt3x3_probe_reset	-	reset the hpt3x3 bus
+ *	@ap: ATA port to reset
+ *
+ *	Perform the housekeeping when doing an ATA bus reeset. We just
+ *	need to force the cable type.
+ */
+
+static void hpt3x3_error_handler(struct ata_port *ap)
+{
+	return ata_bmdma_drive_eh(ap, hpt3x3_probe_init, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	hpt3x3_set_piomode		-	PIO setup
+ *	@ap: ATA interface
+ *	@adev: device on the interface
+ *
+ *	Set our PIO requirements. This is fairly simple on the HPT3x3 as
+ *	all we have to do is clear the MWDMA and UDMA bits then load the
+ *	mode number.
+ */
+
+static void hpt3x3_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u32 r1, r2;
+	int dn = 2 * ap->port_no + adev->devno;
+
+	pci_read_config_dword(pdev, 0x44, &r1);
+	pci_read_config_dword(pdev, 0x48, &r2);
+	/* Load the PIO timing number */
+	r1 &= ~(7 << (3 * dn));
+	r1 |= (adev->pio_mode - XFER_PIO_0) << (3 * dn);
+	r2 &= ~(0x11 << dn);	/* Clear MWDMA and UDMA bits */
+
+	pci_write_config_dword(pdev, 0x44, r1);
+	pci_write_config_dword(pdev, 0x48, r2);
+}
+
+/**
+ *	hpt3x3_set_dmamode		-	DMA timing setup
+ *	@ap: ATA interface
+ *	@adev: Device being configured
+ *
+ *	Set up the channel for MWDMA or UDMA modes. Much the same as with
+ *	PIO, load the mode number and then set MWDMA or UDMA flag.
+ */
+
+static void hpt3x3_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u32 r1, r2;
+	int dn = 2 * ap->port_no + adev->devno;
+	int mode_num = adev->dma_mode & 0x0F;
+
+	pci_read_config_dword(pdev, 0x44, &r1);
+	pci_read_config_dword(pdev, 0x48, &r2);
+	/* Load the timing number */
+	r1 &= ~(7 << (3 * dn));
+	r1 |= (mode_num << (3 * dn));
+	r2 &= ~(0x11 << dn);	/* Clear MWDMA and UDMA bits */
+
+	if (adev->dma_mode >= XFER_UDMA_0)
+		r2 |= 0x01 << dn;	/* Ultra mode */
+	else
+		r2 |= 0x10 << dn;	/* MWDMA */
+
+	pci_write_config_dword(pdev, 0x44, r1);
+	pci_write_config_dword(pdev, 0x48, r2);
+}
+
+static struct scsi_host_template hpt3x3_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations hpt3x3_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= hpt3x3_set_piomode,
+	.set_dmamode	= hpt3x3_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= hpt3x3_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+/**
+ *	hpt3x3_init_one		-	Initialise an HPT343/363
+ *	@dev: PCI device
+ *	@id: Entry in match table
+ *
+ *	Perform basic initialisation. The chip has a quirk that it won't
+ *	function unless it is at XX00. The old ATA driver touched this up
+ *	but we leave it for pci quirks to do properly.
+ */
+
+static int hpt3x3_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info = {
+		.sht = &hpt3x3_sht,
+		.flags = ATA_FLAG_SLAVE_POSS|ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x07,
+		.port_ops = &hpt3x3_port_ops
+	};
+	static struct ata_port_info *port_info[2] = { &info, &info };
+	u16 cmd;
+
+	/* Initialize the board */
+	pci_write_config_word(dev, 0x80, 0x00);
+	/* Check if it is a 343 or a 363. 363 has COMMAND_MEMORY set */
+	pci_read_config_word(dev, PCI_COMMAND, &cmd);
+	if (cmd & PCI_COMMAND_MEMORY)
+		pci_write_config_byte(dev, PCI_LATENCY_TIMER, 0xF0);
+	else
+		pci_write_config_byte(dev, PCI_LATENCY_TIMER, 0x20);
+
+	/* Now kick off ATA set up */
+	return ata_pci_init_one(dev, port_info, 2);
+}
+
+static struct pci_device_id hpt3x3[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_TTI, PCI_DEVICE_ID_TTI_HPT343), },
+	{ 0, },
+};
+
+static struct pci_driver hpt3x3_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= hpt3x3,
+	.probe 		= hpt3x3_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init hpt3x3_init(void)
+{
+	return pci_register_driver(&hpt3x3_pci_driver);
+}
+
+
+static void __exit hpt3x3_exit(void)
+{
+	pci_unregister_driver(&hpt3x3_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for the Highpoint HPT343/363");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, hpt3x3);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(hpt3x3_init);
+module_exit(hpt3x3_exit);
diff --git a/drivers/ata/pata_isapnp.c b/drivers/ata/pata_isapnp.c
new file mode 100644
index 000000000000..73948c8b7270
--- /dev/null
+++ b/drivers/ata/pata_isapnp.c
@@ -0,0 +1,156 @@
+
+/*
+ *   pata-isapnp.c - ISA PnP PATA controller driver.
+ *   Copyright 2005/2006 Red Hat Inc <alan@redhat.com>, all rights reserved.
+ *
+ *   Based in part on ide-pnp.c by Andrey Panin <pazke@donpac.ru>
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/isapnp.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/ata.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "pata_isapnp"
+#define DRV_VERSION "0.1.5"
+
+static struct scsi_host_template isapnp_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations isapnp_port_ops = {
+	.port_disable	= ata_port_disable,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= ata_bmdma_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+/**
+ *	isapnp_init_one		-	attach an isapnp interface
+ *	@idev: PnP device
+ *	@dev_id: matching detect line
+ *
+ *	Register an ISA bus IDE interface. Such interfaces are PIO 0 and
+ *	non shared IRQ.
+ */
+
+static int isapnp_init_one(struct pnp_dev *idev, const struct pnp_device_id *dev_id)
+{
+	struct ata_probe_ent ae;
+
+	if (pnp_port_valid(idev, 0) == 0)
+		return -ENODEV;
+
+	/* FIXME: Should selected polled PIO here not fail */
+	if (pnp_irq_valid(idev, 0) == 0)
+		return -ENODEV;
+
+	memset(&ae, 0, sizeof(struct ata_probe_ent));
+	INIT_LIST_HEAD(&ae.node);
+	ae.dev = &idev->dev;
+	ae.port_ops = &isapnp_port_ops;
+	ae.sht = &isapnp_sht;
+	ae.n_ports = 1;
+	ae.pio_mask = 1;		/* ISA so PIO 0 cycles */
+	ae.irq = pnp_irq(idev, 0);
+	ae.irq_flags = 0;
+	ae.port_flags = ATA_FLAG_SLAVE_POSS;
+	ae.port[0].cmd_addr = pnp_port_start(idev, 0);
+
+	if (pnp_port_valid(idev, 1) == 0) {
+		ae.port[0].altstatus_addr = pnp_port_start(idev, 1);
+		ae.port[0].ctl_addr = pnp_port_start(idev, 1);
+		ae.port_flags |= ATA_FLAG_SRST;
+	}
+	ata_std_ports(&ae.port[0]);
+
+	if (ata_device_add(&ae) == 0)
+		return -ENODEV;
+	return 0;
+}
+
+/**
+ *	isapnp_remove_one	-	unplug an isapnp interface
+ *	@idev: PnP device
+ *
+ *	Remove a previously configured PnP ATA port. Called only on module
+ *	unload events as the core does not currently deal with ISAPnP docking.
+ */
+
+static void isapnp_remove_one(struct pnp_dev *idev)
+{
+	struct device *dev = &idev->dev;
+	struct ata_host *host = dev_get_drvdata(dev);
+
+	ata_host_remove(host);
+	dev_set_drvdata(dev, NULL);
+}
+
+static struct pnp_device_id isapnp_devices[] = {
+  	/* Generic ESDI/IDE/ATA compatible hard disk controller */
+	{.id = "PNP0600", .driver_data = 0},
+	{.id = ""}
+};
+
+static struct pnp_driver isapnp_driver = {
+	.name		= DRV_NAME,
+	.id_table	= isapnp_devices,
+	.probe		= isapnp_init_one,
+	.remove		= isapnp_remove_one,
+};
+
+static int __init isapnp_init(void)
+{
+	return pnp_register_driver(&isapnp_driver);
+}
+
+static void __exit isapnp_exit(void)
+{
+	pnp_unregister_driver(&isapnp_driver);
+}
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for ISA PnP ATA");
+MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
+
+module_init(isapnp_init);
+module_exit(isapnp_exit);
diff --git a/drivers/ata/pata_it8172.c b/drivers/ata/pata_it8172.c
new file mode 100644
index 000000000000..f8367191fc7b
--- /dev/null
+++ b/drivers/ata/pata_it8172.c
@@ -0,0 +1,288 @@
+/*
+ * pata_it8172.c 	- IT8172 PATA for new ATA layer
+ *			  (C) 2005 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ * Based heavily on
+ *
+ * BRIEF MODULE DESCRIPTION
+ *      IT8172 IDE controller support
+ *
+ * Copyright 2000 MontaVista Software Inc.
+ * Author: MontaVista Software, Inc.
+ *              stevel@mvista.com or source@mvista.com
+ *
+ *  This program is free software; you can redistribute  it and/or modify it
+ *  under  the terms of  the GNU General  Public License as published by the
+ *  Free Software Foundation;  either version 2 of the  License, or (at your
+ *  option) any later version.
+ *
+ *  THIS  SOFTWARE  IS PROVIDED   ``AS  IS'' AND   ANY  EXPRESS OR IMPLIED
+ *  WARRANTIES,   INCLUDING, BUT NOT  LIMITED  TO, THE IMPLIED WARRANTIES OF
+ *  MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN
+ *  NO  EVENT  SHALL   THE AUTHOR  BE    LIABLE FOR ANY   DIRECT, INDIRECT,
+ *  INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+ *  NOT LIMITED   TO, PROCUREMENT OF  SUBSTITUTE GOODS  OR SERVICES; LOSS OF
+ *  USE, DATA,  OR PROFITS; OR  BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
+ *  ANY THEORY OF LIABILITY, WHETHER IN  CONTRACT, STRICT LIABILITY, OR TORT
+ *  (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
+ *  THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ *  You should have received a copy of the  GNU General Public License along
+ *  with this program; if not, write  to the Free Software Foundation, Inc.,
+ *  675 Mass Ave, Cambridge, MA 02139, USA.
+ *
+ *	TODO
+ *	Check for errata
+ *	See if we really need to force native mode
+ *	PIO timings (also lacking in original)
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "pata_it8172"
+#define DRV_VERSION "0.3.1"
+
+static int it8172_pre_reset(struct ata_port *ap)
+{
+	static const struct pci_bits it8172_enable_bits[] = {
+		{ 0x00, 0, 0x00, 0x00 },
+		{ 0x40, 1, 0x00, 0x01 }
+	};
+
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	
+	if (ap->port_no && !pci_test_config_bits(pdev, &it8172_enable_bits[ap->port_no])) {
+		ata_port_disable(ap);
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+static void it8172_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, it8172_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	it8172_set_pio_timing	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Called by both the pio and dma setup functions to set the controller
+ *	timings for PIO transfers. We must load both the mode number and
+ *	timing values into the controller.
+ */
+
+static void it8172_set_pio_timing(struct ata_port *ap, struct ata_device *adev, int pio)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u16 reg40;
+
+	pci_read_config_word(pdev, 0x40, &reg40);
+
+	/*
+	 * FIX! The DIOR/DIOW pulse width and recovery times in port 0x44
+	 * are being left at the default values of 8 PCI clocks (242 nsec
+	 * for a 33 MHz clock). These can be safely shortened at higher
+	 * PIO modes. The DIOR/DIOW pulse width and recovery times only
+	 * apply to PIO modes, not to the DMA modes.
+	 */
+
+	/*
+	 * Enable port 0x44. The IT8172G spec is confused; it calls
+	 * this register the "Slave IDE Timing Register", but in fact,
+	 * it controls timing for both master and slave drives.
+	 */
+
+	reg40 |= 0x4000;
+	if (adev->devno) {
+		reg40 &= 0xC006;
+		if (pio > 1)
+			/* Enable prefetch and IORDY sample-point */
+			reg40 |= 0x0060;
+	} else {
+		reg40 &= 0xC060;
+		if (pio > 1)
+			/* Enable prefetch and IORDY sample-point */
+			reg40 |= 0x0006;
+	}
+	/* Write back the enables */
+	pci_write_config_word(pdev, 0x40, reg40);
+}
+
+/**
+ *	it8172_set_piomode	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Called to do the PIO mode setup. We use a shared helper for this
+ *	as the DMA setup must also adjust the PIO timing information.
+ */
+
+static void it8172_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	it8172_set_pio_timing(ap, adev, adev->pio_mode - XFER_PIO_0);
+}
+
+/**
+ *	it8172_set_dmamode	-	set initial DMA mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Called to do the DMA mode setup. We must tune an appropriate PIO
+ *	mode to match.
+ */
+
+static void it8172_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	int dn = (2 * ap->port_no) + adev->devno;
+	u8 reg48, reg4a;
+	int pio;
+
+	static const int pio_map[] = { 1, 3, 4};
+	/*
+	 * Setting the DMA cycle time to 2 or 3 PCI clocks (60 and 91 nsec
+	 * at 33 MHz PCI clock) seems to cause BadCRC errors during DMA
+	 * transfers on some drives, even though both numbers meet the minimum
+	 * ATAPI-4 spec of 73 and 54 nsec for UDMA 1 and 2 respectively.
+	 * So the faster times are just commented out here. The good news is
+	 * that the slower cycle time has very little affect on transfer
+	 * performance.
+	 */
+
+	pci_read_config_byte(pdev, 0x48, &reg48);
+	pci_read_config_byte(pdev, 0x4A, &reg4a);
+
+ 	reg4a &= ~(3 << (4 * dn));
+
+	if (adev->dma_mode >= XFER_UDMA_0) {
+		reg48 |= 1 << dn;
+#ifdef UDMA_TIMING_SET
+		reg4a |= ((adev->dma_mode - XFER_UDMA_0) << (4 * dn));
+#endif
+	 	pio = 4;
+	} else  {
+		pio = pio_map[adev->dma_mode - XFER_MW_DMA_0];
+		reg48 &= ~ (1 << dn);
+	}
+	pci_write_config_byte(pdev, 0x48, reg48);
+	pci_write_config_byte(pdev, 0x4A, reg4a);
+ 	it8172_set_pio_timing(ap, adev, pio);
+
+}
+
+static struct scsi_host_template it8172_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations it8172_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= it8172_set_piomode,
+	.set_dmamode	= it8172_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= it8172_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static int it8172_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info = {
+		.sht = &it8172_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x06,	/* No MWDMA0 support */
+		.udma_mask = 0x7,
+		.port_ops = &it8172_port_ops
+	};
+	static struct ata_port_info *port_info[2] = { &info, &info };
+
+        if ((!(PCI_FUNC(dev->devfn) & 1) ||
+            (!((dev->class >> 8) == PCI_CLASS_STORAGE_IDE))))
+		return -ENODEV; /* IT8172 is more than an IDE controller */
+
+	return ata_pci_init_one(dev, port_info, 2);
+}
+
+static struct pci_device_id it8172[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_ITE, PCI_DEVICE_ID_ITE_IT8172G), },
+	{ 0, },
+};
+
+static struct pci_driver it8172_pci_driver = {
+	.name 		= DRV_NAME,
+	.id_table	= it8172,
+	.probe 		= it8172_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init it8172_init(void)
+{
+	return pci_register_driver(&it8172_pci_driver);
+}
+
+
+static void __exit it8172_exit(void)
+{
+	pci_unregister_driver(&it8172_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for ITE IT8172");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, it8172);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(it8172_init);
+module_exit(it8172_exit);
diff --git a/drivers/ata/pata_it821x.c b/drivers/ata/pata_it821x.c
new file mode 100644
index 000000000000..c8a7798f8ce5
--- /dev/null
+++ b/drivers/ata/pata_it821x.c
@@ -0,0 +1,847 @@
+/*
+ * ata-it821x.c 	- IT821x PATA for new ATA layer
+ *			  (C) 2005 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ * based upon
+ *
+ * it821x.c
+ * 
+ * linux/drivers/ide/pci/it821x.c		Version 0.09	December 2004
+ *
+ * Copyright (C) 2004		Red Hat <alan@redhat.com>
+ *
+ *  May be copied or modified under the terms of the GNU General Public License
+ *  Based in part on the ITE vendor provided SCSI driver.
+ *
+ *  Documentation available from
+ * 	http://www.ite.com.tw/pc/IT8212F_V04.pdf
+ *  Some other documents are NDA.
+ *
+ *  The ITE8212 isn't exactly a standard IDE controller. It has two
+ *  modes. In pass through mode then it is an IDE controller. In its smart
+ *  mode its actually quite a capable hardware raid controller disguised
+ *  as an IDE controller. Smart mode only understands DMA read/write and
+ *  identify, none of the fancier commands apply. The IT8211 is identical
+ *  in other respects but lacks the raid mode.
+ *
+ *  Errata:
+ *  o	Rev 0x10 also requires master/slave hold the same DMA timings and
+ *	cannot do ATAPI MWDMA.
+ *  o	The identify data for raid volumes lacks CHS info (technically ok)
+ *	but also fails to set the LBA28 and other bits. We fix these in
+ *	the IDE probe quirk code.
+ *  o	If you write LBA48 sized I/O's (ie > 256 sector) in smart mode
+ *	raid then the controller firmware dies
+ *  o	Smart mode without RAID doesn't clear all the necessary identify
+ *	bits to reduce the command set to the one used
+ *
+ *  This has a few impacts on the driver
+ *  - In pass through mode we do all the work you would expect
+ *  - In smart mode the clocking set up is done by the controller generally
+ *    but we must watch the other limits and filter.
+ *  - There are a few extra vendor commands that actually talk to the
+ *    controller but only work PIO with no IRQ.
+ *
+ *  Vendor areas of the identify block in smart mode are used for the
+ *  timing and policy set up. Each HDD in raid mode also has a serial
+ *  block on the disk. The hardware extra commands are get/set chip status,
+ *  rebuild, get rebuild status.
+ *
+ *  In Linux the driver supports pass through mode as if the device was
+ *  just another IDE controller. If the smart mode is running then
+ *  volumes are managed by the controller firmware and each IDE "disk"
+ *  is a raid volume. Even more cute - the controller can do automated
+ *  hotplug and rebuild.
+ *
+ *  The pass through controller itself is a little demented. It has a
+ *  flaw that it has a single set of PIO/MWDMA timings per channel so
+ *  non UDMA devices restrict each others performance. It also has a
+ *  single clock source per channel so mixed UDMA100/133 performance
+ *  isn't perfect and we have to pick a clock. Thankfully none of this
+ *  matters in smart mode. ATAPI DMA is not currently supported.
+ *
+ *  It seems the smart mode is a win for RAID1/RAID10 but otherwise not.
+ *
+ *  TODO
+ *	-	ATAPI and other speed filtering
+ *	-	Command filter in smart mode
+ *	-	RAID configuration ioctls
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+
+#define DRV_NAME "pata_it821x"
+#define DRV_VERSION "0.3.2"
+
+struct it821x_dev
+{
+	unsigned int smart:1,		/* Are we in smart raid mode */
+		timing10:1;		/* Rev 0x10 */
+	u8	clock_mode;		/* 0, ATA_50 or ATA_66 */
+	u8	want[2][2];		/* Mode/Pri log for master slave */
+	/* We need these for switching the clock when DMA goes on/off
+	   The high byte is the 66Mhz timing */
+	u16	pio[2];			/* Cached PIO values */
+	u16	mwdma[2];		/* Cached MWDMA values */
+	u16	udma[2];		/* Cached UDMA values (per drive) */
+	u16	last_device;		/* Master or slave loaded ? */
+};
+
+#define ATA_66		0
+#define ATA_50		1
+#define ATA_ANY		2
+
+#define UDMA_OFF	0
+#define MWDMA_OFF	0
+
+/*
+ *	We allow users to force the card into non raid mode without
+ *	flashing the alternative BIOS. This is also neccessary right now
+ *	for embedded platforms that cannot run a PC BIOS but are using this
+ *	device.
+ */
+
+static int it8212_noraid;
+
+/**
+ *	it821x_pre_reset	-	probe
+ *	@ap: ATA port
+ *
+ *	Set the cable type
+ */
+ 
+static int it821x_pre_reset(struct ata_port *ap)
+{
+	ap->cbl = ATA_CBL_PATA80;
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	it821x_error_handler	-	probe/reset
+ *	@ap: ATA port
+ *
+ *	Set the cable type and trigger a probe
+ */
+ 
+static void it821x_error_handler(struct ata_port *ap)
+{
+	return ata_bmdma_drive_eh(ap, it821x_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	it821x_program	-	program the PIO/MWDMA registers
+ *	@ap: ATA port
+ *	@adev: Device to program
+ *	@timing: Timing value (66Mhz in top 8bits, 50 in the low 8)
+ *
+ *	Program the PIO/MWDMA timing for this channel according to the
+ *	current clock. These share the same register so are managed by
+ *	the DMA start/stop sequence as with the old driver.
+ */
+
+static void it821x_program(struct ata_port *ap, struct ata_device *adev, u16 timing)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	struct it821x_dev *itdev = ap->private_data;
+	int channel = ap->port_no;
+	u8 conf;
+
+	/* Program PIO/MWDMA timing bits */
+	if (itdev->clock_mode == ATA_66)
+		conf = timing >> 8;
+	else
+		conf = timing & 0xFF;
+	pci_write_config_byte(pdev, 0x54 + 4 * channel, conf);
+}
+
+
+/**
+ *	it821x_program_udma	-	program the UDMA registers
+ *	@ap: ATA port
+ *	@adev: ATA device to update
+ *	@timing: Timing bits. Top 8 are for 66Mhz bottom for 50Mhz
+ *
+ *	Program the UDMA timing for this drive according to the
+ *	current clock. Handles the dual clocks and also knows about
+ *	the errata on the 0x10 revision. The UDMA errata is partly handled
+ *	here and partly in start_dma.
+ */
+
+static void it821x_program_udma(struct ata_port *ap, struct ata_device *adev, u16 timing)
+{
+	struct it821x_dev *itdev = ap->private_data;
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	int channel = ap->port_no;
+	int unit = adev->devno;
+	u8 conf;
+
+	/* Program UDMA timing bits */
+	if (itdev->clock_mode == ATA_66)
+		conf = timing >> 8;
+	else
+		conf = timing & 0xFF;
+	if (itdev->timing10 == 0)
+		pci_write_config_byte(pdev, 0x56 + 4 * channel + unit, conf);
+	else {
+		/* Early revision must be programmed for both together */
+		pci_write_config_byte(pdev, 0x56 + 4 * channel, conf);
+		pci_write_config_byte(pdev, 0x56 + 4 * channel + 1, conf);
+	}
+}
+
+/**
+ *	it821x_clock_strategy
+ *	@ap: ATA interface
+ *	@adev: ATA device being updated
+ *
+ *	Select between the 50 and 66Mhz base clocks to get the best
+ *	results for this interface.
+ */
+
+static void it821x_clock_strategy(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	struct it821x_dev *itdev = ap->private_data;
+	u8 unit = adev->devno;
+	struct ata_device *pair = ata_dev_pair(adev);
+
+	int clock, altclock;
+	u8 v;
+	int sel = 0;
+
+	/* Look for the most wanted clocking */
+	if (itdev->want[0][0] > itdev->want[1][0]) {
+		clock = itdev->want[0][1];
+		altclock = itdev->want[1][1];
+	} else {
+		clock = itdev->want[1][1];
+		altclock = itdev->want[0][1];
+	}
+
+	/* Master doesn't care does the slave ? */
+	if (clock == ATA_ANY)
+		clock = altclock;
+
+	/* Nobody cares - keep the same clock */
+	if (clock == ATA_ANY)
+		return;
+	/* No change */
+	if (clock == itdev->clock_mode)
+		return;
+
+	/* Load this into the controller */
+	if (clock == ATA_66)
+		itdev->clock_mode = ATA_66;
+	else {
+		itdev->clock_mode = ATA_50;
+		sel = 1;
+	}
+	pci_read_config_byte(pdev, 0x50, &v);
+	v &= ~(1 << (1 + ap->port_no));
+	v |= sel << (1 + ap->port_no);
+	pci_write_config_byte(pdev, 0x50, v);
+
+	/*
+	 *	Reprogram the UDMA/PIO of the pair drive for the switch
+	 *	MWDMA will be dealt with by the dma switcher
+	 */
+	if (pair && itdev->udma[1-unit] != UDMA_OFF) {
+		it821x_program_udma(ap, pair, itdev->udma[1-unit]);
+		it821x_program(ap, pair, itdev->pio[1-unit]);
+	}
+	/*
+	 *	Reprogram the UDMA/PIO of our drive for the switch.
+	 *	MWDMA will be dealt with by the dma switcher
+	 */
+	if (itdev->udma[unit] != UDMA_OFF) {
+		it821x_program_udma(ap, adev, itdev->udma[unit]);
+		it821x_program(ap, adev, itdev->pio[unit]);
+	}
+}
+
+/**
+ *	it821x_passthru_set_piomode	-	set PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Configure for PIO mode. This is complicated as the register is
+ *	shared by PIO and MWDMA and for both channels.
+ */
+
+static void it821x_passthru_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	/* Spec says 89 ref driver uses 88 */
+	static const u16 pio[]	= { 0xAA88, 0xA382, 0xA181, 0x3332, 0x3121 };
+	static const u8 pio_want[]    = { ATA_66, ATA_66, ATA_66, ATA_66, ATA_ANY };
+
+	struct it821x_dev *itdev = ap->private_data;
+	int unit = adev->devno;
+	int mode_wanted = adev->pio_mode - XFER_PIO_0;
+	
+	/* We prefer 66Mhz clock for PIO 0-3, don't care for PIO4 */
+	itdev->want[unit][1] = pio_want[mode_wanted];
+	itdev->want[unit][0] = 1;	/* PIO is lowest priority */
+	itdev->pio[unit] = pio[mode_wanted];
+	it821x_clock_strategy(ap, adev);
+	it821x_program(ap, adev, itdev->pio[unit]);
+}
+
+/**
+ *	it821x_passthru_set_dmamode	-	set initial DMA mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Set up the DMA modes. The actions taken depend heavily on the mode
+ *	to use. If UDMA is used as is hopefully the usual case then the 
+ *	timing register is private and we need only consider the clock. If
+ *	we are using MWDMA then we have to manage the setting ourself as
+ *	we switch devices and mode.
+ */
+
+static void it821x_passthru_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	static const u16 dma[]	= 	{ 0x8866, 0x3222, 0x3121 };
+	static const u8 mwdma_want[] =  { ATA_ANY, ATA_66, ATA_ANY };
+	static const u16 udma[]	= 	{ 0x4433, 0x4231, 0x3121, 0x2121, 0x1111, 0x2211, 0x1111 };
+	static const u8 udma_want[] =   { ATA_ANY, ATA_50, ATA_ANY, ATA_66, ATA_66, ATA_50, ATA_66 };
+
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	struct it821x_dev *itdev = ap->private_data;
+	int channel = ap->port_no;
+	int unit = adev->devno;
+	u8 conf;
+
+	if (adev->dma_mode >= XFER_UDMA_0) {
+		int mode_wanted = adev->dma_mode - XFER_UDMA_0;
+		
+		itdev->want[unit][1] = udma_want[mode_wanted];
+		itdev->want[unit][0] = 3;	/* UDMA is high priority */
+		itdev->mwdma[unit] = MWDMA_OFF;
+		itdev->udma[unit] = udma[mode_wanted];
+		if (mode_wanted >= 5)
+			itdev->udma[unit] |= 0x8080;	/* UDMA 5/6 select on */
+
+		/* UDMA on. Again revision 0x10 must do the pair */
+		pci_read_config_byte(pdev, 0x50, &conf);
+		if (itdev->timing10)
+			conf &= channel ? 0x9F: 0xE7;
+		else
+			conf &= ~ (1 << (3 + 2 * channel + unit));
+		pci_write_config_byte(pdev, 0x50, conf);
+		it821x_clock_strategy(ap, adev);
+		it821x_program_udma(ap, adev, itdev->udma[unit]);
+	} else {
+		int mode_wanted = adev->dma_mode - XFER_MW_DMA_0;
+		
+		itdev->want[unit][1] = mwdma_want[mode_wanted];
+		itdev->want[unit][0] = 2;	/* MWDMA is low priority */
+		itdev->mwdma[unit] = dma[mode_wanted];
+		itdev->udma[unit] = UDMA_OFF;
+
+		/* UDMA bits off - Revision 0x10 do them in pairs */
+		pci_read_config_byte(pdev, 0x50, &conf);
+		if (itdev->timing10)
+			conf |= channel ? 0x60: 0x18;
+		else
+			conf |= 1 << (3 + 2 * channel + unit);
+		pci_write_config_byte(pdev, 0x50, conf);
+		it821x_clock_strategy(ap, adev);
+	}
+}
+
+/**
+ *	it821x_passthru_dma_start	-	DMA start callback
+ *	@qc: Command in progress
+ *
+ *	Usually drivers set the DMA timing at the point the set_dmamode call
+ *	is made. IT821x however requires we load new timings on the 
+ *	transitions in some cases.
+ */
+
+static void it821x_passthru_bmdma_start(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct ata_device *adev = qc->dev;
+	struct it821x_dev *itdev = ap->private_data;
+	int unit = adev->devno;
+
+	if (itdev->mwdma[unit] != MWDMA_OFF)
+		it821x_program(ap, adev, itdev->mwdma[unit]);
+	else if (itdev->udma[unit] != UDMA_OFF && itdev->timing10)
+		it821x_program_udma(ap, adev, itdev->udma[unit]);
+	ata_bmdma_start(qc);
+}
+
+/**
+ *	it821x_passthru_dma_stop	-	DMA stop callback
+ *	@qc: ATA command
+ *
+ *	We loaded new timings in dma_start, as a result we need to restore
+ *	the PIO timings in dma_stop so that the next command issue gets the
+ *	right clock values.
+ */
+
+static void it821x_passthru_bmdma_stop(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct ata_device *adev = qc->dev;
+	struct it821x_dev *itdev = ap->private_data;
+	int unit = adev->devno;
+
+	ata_bmdma_stop(qc);
+	if (itdev->mwdma[unit] != MWDMA_OFF)
+		it821x_program(ap, adev, itdev->pio[unit]);
+}
+
+
+/**
+ *	it821x_passthru_dev_select	-	Select master/slave
+ *	@ap: ATA port
+ *	@device: Device number (not pointer)
+ *
+ *	Device selection hook. If neccessary perform clock switching
+ */
+ 
+static void it821x_passthru_dev_select(struct ata_port *ap,
+				       unsigned int device)
+{
+	struct it821x_dev *itdev = ap->private_data;
+	if (itdev && device != itdev->last_device) {
+		struct ata_device *adev = &ap->device[device];
+		it821x_program(ap, adev, itdev->pio[adev->devno]);
+		itdev->last_device = device;
+	}
+	ata_std_dev_select(ap, device);
+}
+
+/**
+ *	it821x_smart_qc_issue_prot	-	wrap qc issue prot
+ *	@qc: command
+ *
+ *	Wrap the command issue sequence for the IT821x. We need to
+ *	perform out own device selection timing loads before the
+ *	usual happenings kick off
+ */
+ 
+static unsigned int it821x_smart_qc_issue_prot(struct ata_queued_cmd *qc)
+{
+	switch(qc->tf.command)
+	{
+		/* Commands the firmware supports */
+		case ATA_CMD_READ:
+		case ATA_CMD_READ_EXT:
+		case ATA_CMD_WRITE:
+		case ATA_CMD_WRITE_EXT:
+		case ATA_CMD_PIO_READ:
+		case ATA_CMD_PIO_READ_EXT:
+		case ATA_CMD_PIO_WRITE:
+		case ATA_CMD_PIO_WRITE_EXT:
+		case ATA_CMD_READ_MULTI:
+		case ATA_CMD_READ_MULTI_EXT:
+		case ATA_CMD_WRITE_MULTI:
+		case ATA_CMD_WRITE_MULTI_EXT:
+		case ATA_CMD_ID_ATA:
+		/* Arguably should just no-op this one */
+		case ATA_CMD_SET_FEATURES:
+			return ata_qc_issue_prot(qc);
+	}
+	printk(KERN_DEBUG "it821x: can't process command 0x%02X\n", qc->tf.command);
+	return AC_ERR_INVALID;
+}
+
+/**
+ *	it821x_passthru_qc_issue_prot	-	wrap qc issue prot
+ *	@qc: command
+ *
+ *	Wrap the command issue sequence for the IT821x. We need to
+ *	perform out own device selection timing loads before the
+ *	usual happenings kick off
+ */
+ 
+static unsigned int it821x_passthru_qc_issue_prot(struct ata_queued_cmd *qc)
+{
+	it821x_passthru_dev_select(qc->ap, qc->dev->devno);
+	return ata_qc_issue_prot(qc);
+}
+
+/**
+ *	it821x_smart_set_mode	-	mode setting
+ *	@ap: interface to set up
+ *
+ *	Use a non standard set_mode function. We don't want to be tuned.
+ *	The BIOS configured everything. Our job is not to fiddle. We
+ *	read the dma enabled bits from the PCI configuration of the device
+ *	and respect them. 
+ */
+ 
+static void it821x_smart_set_mode(struct ata_port *ap)
+{
+	int dma_enabled = 0;
+	int i;
+
+	/* Bits 5 and 6 indicate if DMA is active on master/slave */
+	/* It is possible that BMDMA isn't allocated */
+	if (ap->ioaddr.bmdma_addr)
+		dma_enabled = inb(ap->ioaddr.bmdma_addr + ATA_DMA_CMD);
+	
+	for (i = 0; i < ATA_MAX_DEVICES; i++) {
+		struct ata_device *dev = &ap->device[i];
+		if (ata_dev_enabled(dev)) {
+			/* We don't really care */
+			dev->pio_mode = XFER_PIO_0;
+			dev->dma_mode = XFER_MW_DMA_0;
+			/* We do need the right mode information for DMA or PIO 
+			   and this comes from the current configuration flags */
+			if (dma_enabled & (1 << (5 + i))) {
+				dev->xfer_mode = XFER_MW_DMA_0;
+				dev->xfer_shift = ATA_SHIFT_MWDMA;
+				dev->flags &= ~ATA_DFLAG_PIO;
+			} else {
+				dev->xfer_mode = XFER_PIO_0;
+				dev->xfer_shift = ATA_SHIFT_PIO;
+				dev->flags |= ATA_DFLAG_PIO;
+			}
+		}
+	}
+}
+
+/**
+ *	it821x_dev_config	-	Called each device identify
+ *	@ap: ATA port
+ *	@adev: Device that has just been identified
+ *
+ *	Perform the initial setup needed for each device that is chip
+ *	special. In our case we need to lock the sector count to avoid
+ *	blowing the brains out of the firmware with large LBA48 requests
+ *
+ *	FIXME: When FUA appears we need to block FUA too. And SMART and
+ *	basically we need to filter commands for this chip.
+ */
+ 
+static void it821x_dev_config(struct ata_port *ap, struct ata_device *adev)
+{
+	unsigned char model_num[40];
+	char *s;
+	unsigned int len;
+
+	/* This block ought to be a library routine as it is in several
+	   drivers now */
+	   
+	ata_id_string(adev->id, model_num, ATA_ID_PROD_OFS,
+			  sizeof(model_num));
+	s = &model_num[0];
+	len = strnlen(s, sizeof(model_num));
+
+	/* ATAPI specifies that empty space is blank-filled; remove blanks */
+	while ((len > 0) && (s[len - 1] == ' ')) {
+		len--;
+		s[len] = 0;
+	}
+
+	if (adev->max_sectors > 255)
+		adev->max_sectors = 255;
+	
+	if (strstr(model_num, "Integrated Technology Express")) {
+		/* RAID mode */
+		printk(KERN_INFO "IT821x %sRAID%d volume",
+			adev->id[147]?"Bootable ":"",
+			adev->id[129]);
+		if (adev->id[129] != 1)
+			printk("(%dK stripe)", adev->id[146]);
+		printk(".\n");
+	}
+}
+
+
+/**
+ *	it821x_check_atapi_dma	-	ATAPI DMA handler
+ *	@qc: Command we are about to issue
+ *
+ *	Decide if this ATAPI command can be issued by DMA on this
+ *	controller. Return 0 if it can be.
+ */
+ 
+static int it821x_check_atapi_dma(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct it821x_dev *itdev = ap->private_data;
+	
+	/* No ATAPI DMA in smart mode */
+	if (itdev->smart)
+		return -EOPNOTSUPP;
+	/* No ATAPI DMA on rev 10 */
+	if (itdev->timing10)
+		return -EOPNOTSUPP;
+	/* Cool */
+	return 0;
+}
+		
+
+/**
+ *	it821x_port_start	-	port setup
+ *	@ap: ATA port being set up
+ *
+ *	The it821x needs to maintain private data structures and also to
+ *	use the standard PCI interface which lacks support for this
+ *	functionality. We instead set up the private data on the port 
+ *	start hook, and tear it down on port stop
+ */
+ 
+static int it821x_port_start(struct ata_port *ap)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	struct it821x_dev *itdev;
+	u8 conf;
+
+	int ret = ata_port_start(ap);
+	if (ret < 0)
+		return ret;
+		
+	ap->private_data = kmalloc(sizeof(struct it821x_dev), GFP_KERNEL);
+	if (ap->private_data == NULL) {
+		ata_port_stop(ap);
+		return -ENOMEM;
+	}
+	
+	itdev = ap->private_data;
+	memset(itdev, 0, sizeof(struct it821x_dev));
+
+	pci_read_config_byte(pdev, 0x50, &conf);
+
+	if (conf & 1) {
+		itdev->smart = 1;
+		/* Long I/O's although allowed in LBA48 space cause the
+		   onboard firmware to enter the twighlight zone */
+		/* No ATAPI DMA in this mode either */
+	}
+	/* Pull the current clocks from 0x50 */
+	if (conf & (1 << (1 + ap->port_no)))
+		itdev->clock_mode = ATA_50;
+	else
+		itdev->clock_mode = ATA_66;
+
+	itdev->want[0][1] = ATA_ANY;
+	itdev->want[1][1] = ATA_ANY;
+	itdev->last_device = -1;
+
+	pci_read_config_byte(pdev, PCI_REVISION_ID, &conf);
+	if (conf == 0x10) {
+		itdev->timing10 = 1;
+		/* Need to disable ATAPI DMA for this case */
+		if (!itdev->smart)
+			printk(KERN_WARNING DRV_NAME": Revision 0x10, workarounds activated.\n");
+	}
+
+	return 0;
+}
+
+/**
+ *	it821x_port_stop	-	port shutdown
+ *	@ap: ATA port being removed
+ *
+ *	Release the private objects we added in it821x_port_start
+ */
+ 
+static void it821x_port_stop(struct ata_port *ap) {
+	kfree(ap->private_data);
+	ap->private_data = NULL;	/* We want an OOPS if we reuse this
+					   too late! */
+	ata_port_stop(ap);
+}
+
+static struct scsi_host_template it821x_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	/* 255 sectors to begin with. This is locked in smart mode but not
+	   in pass through */
+	.max_sectors		= 255,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations it821x_smart_port_ops = {
+	.set_mode	= it821x_smart_set_mode,
+	.port_disable	= ata_port_disable,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.mode_filter	= ata_pci_default_filter,
+	
+	.check_status 	= ata_check_status,
+	.check_atapi_dma= it821x_check_atapi_dma,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+	.dev_config	= it821x_dev_config,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= it821x_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= it821x_smart_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= it821x_port_start,
+	.port_stop	= it821x_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+static struct ata_port_operations it821x_passthru_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= it821x_passthru_set_piomode,
+	.set_dmamode	= it821x_passthru_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.check_atapi_dma= it821x_check_atapi_dma,
+	.dev_select 	= it821x_passthru_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= it821x_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= it821x_passthru_bmdma_start,
+	.bmdma_stop	= it821x_passthru_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= it821x_passthru_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_handler	= ata_interrupt,
+
+	.port_start	= it821x_port_start,
+	.port_stop	= it821x_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+static void __devinit it821x_disable_raid(struct pci_dev *pdev)
+{
+	/* Reset local CPU, and set BIOS not ready */
+	pci_write_config_byte(pdev, 0x5E, 0x01);
+
+	/* Set to bypass mode, and reset PCI bus */
+	pci_write_config_byte(pdev, 0x50, 0x00);
+	pci_write_config_word(pdev, PCI_COMMAND,
+			      PCI_COMMAND_PARITY | PCI_COMMAND_IO |
+			      PCI_COMMAND_MEMORY | PCI_COMMAND_MASTER);
+	pci_write_config_word(pdev, 0x40, 0xA0F3);
+
+	pci_write_config_dword(pdev,0x4C, 0x02040204);
+	pci_write_config_byte(pdev, 0x42, 0x36);
+	pci_write_config_byte(pdev, PCI_LATENCY_TIMER, 0x20);
+}
+
+	
+static int it821x_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+	u8 conf;
+	
+	static struct ata_port_info info_smart = {
+		.sht = &it821x_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.port_ops = &it821x_smart_port_ops
+	};
+	static struct ata_port_info info_passthru = {
+		.sht = &it821x_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x7f,
+		.port_ops = &it821x_passthru_port_ops
+	};
+	static struct ata_port_info *port_info[2];
+	
+	static char *mode[2] = { "pass through", "smart" };
+
+	/* Force the card into bypass mode if so requested */
+	if (it8212_noraid) {
+		printk(KERN_INFO DRV_NAME ": forcing bypass mode.\n");
+		it821x_disable_raid(pdev);
+	}
+	pci_read_config_byte(pdev, 0x50, &conf);
+	conf &= 1;
+	
+	printk(KERN_INFO DRV_NAME ": controller in %s mode.\n", mode[conf]);
+	if (conf == 0)
+		port_info[0] = port_info[1] = &info_passthru;
+	else
+		port_info[0] = port_info[1] = &info_smart;
+	
+	return ata_pci_init_one(pdev, port_info, 2);
+}
+
+static struct pci_device_id it821x[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_ITE, PCI_DEVICE_ID_ITE_8211), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_ITE, PCI_DEVICE_ID_ITE_8212), },
+	{ 0, },
+};
+
+static struct pci_driver it821x_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= it821x,
+	.probe 		= it821x_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init it821x_init(void)
+{
+	return pci_register_driver(&it821x_pci_driver);
+}
+
+
+static void __exit it821x_exit(void)
+{
+	pci_unregister_driver(&it821x_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for the IT8211/IT8212 IDE RAID controller");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, it821x);
+MODULE_VERSION(DRV_VERSION);
+
+
+module_param_named(noraid, it8212_noraid, int, S_IRUGO);
+MODULE_PARM_DESC(it8212_noraid, "Force card into bypass mode");
+
+module_init(it821x_init);
+module_exit(it821x_exit);
diff --git a/drivers/ata/pata_jmicron.c b/drivers/ata/pata_jmicron.c
new file mode 100644
index 000000000000..6832a643a9eb
--- /dev/null
+++ b/drivers/ata/pata_jmicron.c
@@ -0,0 +1,266 @@
+/*
+ *    pata_jmicron.c - JMicron ATA driver for non AHCI mode. This drives the
+ *			PATA port of the controller. The SATA ports are
+ *			driven by AHCI in the usual configuration although
+ *			this driver can handle other setups if we need it.
+ *
+ *	(c) 2006 Red Hat  <alan@redhat.com>
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+#include <linux/ata.h>
+
+#define DRV_NAME	"pata_jmicron"
+#define DRV_VERSION	"0.1.2"
+
+typedef enum {
+	PORT_PATA0 = 0,
+	PORT_PATA1 = 1,
+	PORT_SATA = 2,
+} port_type;
+
+/**
+ *	jmicron_pre_reset	-	check for 40/80 pin
+ *	@ap: Port
+ *
+ *	Perform the PATA port setup we need.
+
+ *	On the Jmicron 361/363 there is a single PATA port that can be mapped
+ *	either as primary or secondary (or neither). We don't do any policy
+ *	and setup here. We assume that has been done by init_one and the
+ *	BIOS.
+ */
+
+static int jmicron_pre_reset(struct ata_port *ap)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u32 control;
+	u32 control5;
+	int port_mask = 1<< (4 * ap->port_no);
+	int port = ap->port_no;
+	port_type port_map[2];
+
+	/* Check if our port is enabled */
+	pci_read_config_dword(pdev, 0x40, &control);
+	if ((control & port_mask) == 0)
+		return 0;
+
+	/* There are two basic mappings. One has the two SATA ports merged
+	   as master/slave and the secondary as PATA, the other has only the
+	   SATA port mapped */
+	if (control & (1 << 23)) {
+		port_map[0] = PORT_SATA;
+		port_map[1] = PORT_PATA0;
+	} else {
+		port_map[0] = PORT_SATA;
+		port_map[1] = PORT_SATA;
+	}
+
+	/* The 365/366 may have this bit set to map the second PATA port
+	   as the internal primary channel */
+	pci_read_config_dword(pdev, 0x80, &control5);
+	if (control5 & (1<<24))
+		port_map[0] = PORT_PATA1;
+
+	/* The two ports may then be logically swapped by the firmware */
+	if (control & (1 << 22))
+		port = port ^ 1;
+
+	/*
+	 *	Now we know which physical port we are talking about we can
+	 *	actually do our cable checking etc. Thankfully we don't need
+	 *	to do the plumbing for other cases.
+	 */
+	switch (port_map[port])
+	{
+	case PORT_PATA0:
+		if (control & (1 << 5))
+			return 0;
+		if (control & (1 << 3))	/* 40/80 pin primary */
+			ap->cbl = ATA_CBL_PATA40;
+		else
+			ap->cbl = ATA_CBL_PATA80;
+		break;
+	case PORT_PATA1:
+		/* Bit 21 is set if the port is enabled */
+		if ((control5 & (1 << 21)) == 0)
+			return 0;
+		if (control5 & (1 << 19))	/* 40/80 pin secondary */
+			ap->cbl = ATA_CBL_PATA40;
+		else
+			ap->cbl = ATA_CBL_PATA80;
+		break;
+	case PORT_SATA:
+		ap->cbl = ATA_CBL_SATA;
+		break;
+	}
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	jmicron_error_handler - Setup and error handler
+ *	@ap: Port to handle
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void jmicron_error_handler(struct ata_port *ap)
+{
+	return ata_bmdma_drive_eh(ap, jmicron_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/* No PIO or DMA methods needed for this device */
+
+static struct scsi_host_template jmicron_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	/* Special handling needed if you have sector or LBA48 limits */
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	/* Use standard CHS mapping rules */
+	.bios_param		= ata_std_bios_param,
+};
+
+static const struct ata_port_operations jmicron_ops = {
+	.port_disable		= ata_port_disable,
+
+	/* Task file is PCI ATA format, use helpers */
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.check_status		= ata_check_status,
+	.exec_command		= ata_exec_command,
+	.dev_select		= ata_std_dev_select,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= jmicron_error_handler,
+	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+
+	/* BMDMA handling is PCI ATA format, use helpers */
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= ata_qc_issue_prot,
+	.data_xfer		= ata_pio_data_xfer,
+
+	/* Timeout handling. Special recovery hooks here */
+	.eng_timeout		= ata_eng_timeout,
+	.irq_handler		= ata_interrupt,
+	.irq_clear		= ata_bmdma_irq_clear,
+
+	/* Generic PATA PCI ATA helpers */
+	.port_start		= ata_port_start,
+	.port_stop		= ata_port_stop,
+	.host_stop		= ata_host_stop,
+};
+
+
+/**
+ *	jmicron_init_one - Register Jmicron ATA PCI device with kernel services
+ *	@pdev: PCI device to register
+ *	@ent: Entry in jmicron_pci_tbl matching with @pdev
+ *
+ *	Called from kernel PCI layer.
+ *
+ *	LOCKING:
+ *	Inherited from PCI layer (may sleep).
+ *
+ *	RETURNS:
+ *	Zero on success, or -ERRNO value.
+ */
+
+static int jmicron_init_one (struct pci_dev *pdev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info = {
+		.sht		= &jmicron_sht,
+		.flags	= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+
+		.pio_mask	= 0x1f,
+		.mwdma_mask	= 0x07,
+		.udma_mask 	= 0x3f,
+
+		.port_ops	= &jmicron_ops,
+	};
+	struct ata_port_info *port_info[2] = { &info, &info };
+
+	u32 reg;
+
+	if (id->driver_data != 368) {
+		/* Put the controller into AHCI mode in case the AHCI driver
+		   has not yet been loaded. This can be done with either
+		   function present */
+
+		/* FIXME: We may want a way to override this in future */
+		pci_write_config_byte(pdev, 0x41, 0xa1);
+	}
+
+	/* PATA controller is fn 1, AHCI is fn 0 */
+	if (PCI_FUNC(pdev->devfn) != 1)
+		return -ENODEV;
+
+	if ( id->driver_data == 365 || id->driver_data == 366) {
+		/* The 365/66 have two PATA channels, redirect the second */
+		pci_read_config_dword(pdev, 0x80, &reg);
+		reg |= (1 << 24);	/* IDE1 to PATA IDE secondary */
+		pci_write_config_dword(pdev, 0x80, reg);
+	}
+
+	return ata_pci_init_one(pdev, port_info, 2);
+}
+
+static const struct pci_device_id jmicron_pci_tbl[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_JMICRON, PCI_DEVICE_ID_JMICRON_JMB361), 361},
+	{ PCI_DEVICE(PCI_VENDOR_ID_JMICRON, PCI_DEVICE_ID_JMICRON_JMB363), 363},
+	{ PCI_DEVICE(PCI_VENDOR_ID_JMICRON, PCI_DEVICE_ID_JMICRON_JMB365), 365},
+	{ PCI_DEVICE(PCI_VENDOR_ID_JMICRON, PCI_DEVICE_ID_JMICRON_JMB366), 366},
+	{ PCI_DEVICE(PCI_VENDOR_ID_JMICRON, PCI_DEVICE_ID_JMICRON_JMB368), 368},
+	{ }	/* terminate list */
+};
+
+static struct pci_driver jmicron_pci_driver = {
+	.name			= DRV_NAME,
+	.id_table		= jmicron_pci_tbl,
+	.probe			= jmicron_init_one,
+	.remove			= ata_pci_remove_one,
+};
+
+static int __init jmicron_init(void)
+{
+	return pci_register_driver(&jmicron_pci_driver);
+}
+
+static void __exit jmicron_exit(void)
+{
+	pci_unregister_driver(&jmicron_pci_driver);
+}
+
+module_init(jmicron_init);
+module_exit(jmicron_exit);
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("SCSI low-level driver for Jmicron PATA ports");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, jmicron_pci_tbl);
+MODULE_VERSION(DRV_VERSION);
+
diff --git a/drivers/ata/pata_legacy.c b/drivers/ata/pata_legacy.c
new file mode 100644
index 000000000000..eb510660b424
--- /dev/null
+++ b/drivers/ata/pata_legacy.c
@@ -0,0 +1,949 @@
+/*
+ *   pata-legacy.c - Legacy port PATA/SATA controller driver.
+ *   Copyright 2005/2006 Red Hat <alan@redhat.com>, all rights reserved.
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2, or (at your option)
+ *  any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; see the file COPYING.  If not, write to
+ *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ *
+ *   An ATA driver for the legacy ATA ports.
+ *
+ *   Data Sources:
+ *	Opti 82C465/82C611 support: Data sheets at opti-inc.com
+ *	HT6560 series:
+ *	Promise 20230/20620:
+ *		http://www.ryston.cz/petr/vlb/pdc20230b.html
+ *		http://www.ryston.cz/petr/vlb/pdc20230c.html
+ *		http://www.ryston.cz/petr/vlb/pdc20630.html
+ *
+ *  Unsupported but docs exist:
+ *	Appian/Adaptec AIC25VL01/Cirrus Logic PD7220
+ *	Winbond W83759A
+ *
+ *  This driver handles legacy (that is "ISA/VLB side") IDE ports found
+ *  on PC class systems. There are three hybrid devices that are exceptions
+ *  The Cyrix 5510/5520 where a pre SFF ATA device is on the bridge and
+ *  the MPIIX where the tuning is PCI side but the IDE is "ISA side".
+ *
+ *  Specific support is included for the ht6560a/ht6560b/opti82c611a/
+ *  opti82c465mv/promise 20230c/20630
+ *
+ *  Use the autospeed and pio_mask options with:
+ *	Appian ADI/2 aka CLPD7220 or AIC25VL01.
+ *  Use the jumpers, autospeed and set pio_mask to the mode on the jumpers with
+ *	Goldstar GM82C711, PIC-1288A-125, UMC 82C871F, Winbond W83759,
+ *	Winbond W83759A, Promise PDC20230-B
+ *
+ *  For now use autospeed and pio_mask as above with the W83759A. This may
+ *  change.
+ *
+ *  TODO
+ *	Merge existing pata_qdi driver
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/ata.h>
+#include <linux/libata.h>
+#include <linux/platform_device.h>
+
+#define DRV_NAME "pata_legacy"
+#define DRV_VERSION "0.5.3"
+
+#define NR_HOST 6
+
+static int legacy_port[NR_HOST] = { 0x1f0, 0x170, 0x1e8, 0x168, 0x1e0, 0x160 };
+static int legacy_irq[NR_HOST] = { 15, 14, 11, 10, 8, 12 };
+
+struct legacy_data {
+	unsigned long timing;
+	u8 clock[2];
+	u8 last;
+	int fast;
+	struct platform_device *platform_dev;
+
+};
+
+static struct legacy_data legacy_data[NR_HOST];
+static struct ata_host *legacy_host[NR_HOST];
+static int nr_legacy_host;
+
+
+static int probe_all;			/* Set to check all ISA port ranges */
+static int ht6560a;			/* HT 6560A on primary 1, secondary 2, both 3 */
+static int ht6560b;			/* HT 6560A on primary 1, secondary 2, both 3 */
+static int opti82c611a;			/* Opti82c611A on primary 1, secondary 2, both 3 */
+static int opti82c46x;		/* Opti 82c465MV present (pri/sec autodetect) */
+static int autospeed;			/* Chip present which snoops speed changes */
+static int pio_mask = 0x1F;		/* PIO range for autospeed devices */
+
+/**
+ *	legacy_set_mode		-	mode setting
+ *	@ap: IDE interface
+ *
+ *	Use a non standard set_mode function. We don't want to be tuned.
+ *
+ *	The BIOS configured everything. Our job is not to fiddle. Just use
+ *	whatever PIO the hardware is using and leave it at that. When we
+ *	get some kind of nice user driven API for control then we can
+ *	expand on this as per hdparm in the base kernel.
+ */
+
+static void legacy_set_mode(struct ata_port *ap)
+{
+	int i;
+
+	for (i = 0; i < ATA_MAX_DEVICES; i++) {
+		struct ata_device *dev = &ap->device[i];
+		if (ata_dev_enabled(dev)) {
+			dev->pio_mode = XFER_PIO_0;
+			dev->xfer_mode = XFER_PIO_0;
+			dev->xfer_shift = ATA_SHIFT_PIO;
+			dev->flags |= ATA_DFLAG_PIO;
+		}
+	}
+}
+
+static struct scsi_host_template legacy_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+/*
+ *	These ops are used if the user indicates the hardware
+ *	snoops the commands to decide on the mode and handles the
+ *	mode selection "magically" itself. Several legacy controllers
+ *	do this. The mode range can be set if it is not 0x1F by setting
+ *	pio_mask as well.
+ */
+
+static struct ata_port_operations simple_port_ops = {
+	.port_disable	= ata_port_disable,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= ata_bmdma_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer_noirq,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static struct ata_port_operations legacy_port_ops = {
+	.set_mode	= legacy_set_mode,
+
+	.port_disable	= ata_port_disable,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.error_handler	= ata_bmdma_error_handler,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer_noirq,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+/*
+ *	Promise 20230C and 20620 support
+ *
+ *	This controller supports PIO0 to PIO2. We set PIO timings conservatively to
+ *	allow for 50MHz Vesa Local Bus. The 20620 DMA support is weird being DMA to
+ *	controller and PIO'd to the host and not supported.
+ */
+
+static void pdc20230_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	int tries = 5;
+	int pio = adev->pio_mode - XFER_PIO_0;
+	u8 rt;
+	unsigned long flags;
+	
+	/* Safe as UP only. Force I/Os to occur together */
+	
+	local_irq_save(flags);
+	
+	/* Unlock the control interface */
+	do
+	{
+		inb(0x1F5);
+		outb(inb(0x1F2) | 0x80, 0x1F2);
+		inb(0x1F2);
+		inb(0x3F6);
+		inb(0x3F6);
+		inb(0x1F2);
+		inb(0x1F2);
+	}
+	while((inb(0x1F2) & 0x80) && --tries);
+
+	local_irq_restore(flags);
+	
+	outb(inb(0x1F4) & 0x07, 0x1F4);
+
+	rt = inb(0x1F3);
+	rt &= 0x07 << (3 * adev->devno);
+	if (pio)
+		rt |= (1 + 3 * pio) << (3 * adev->devno);
+
+	udelay(100);
+	outb(inb(0x1F2) | 0x01, 0x1F2);
+	udelay(100);
+	inb(0x1F5);
+
+}
+
+static void pdc_data_xfer_vlb(struct ata_device *adev, unsigned char *buf, unsigned int buflen, int write_data)
+{
+	struct ata_port *ap = adev->ap;
+	int slop = buflen & 3;
+	unsigned long flags;
+
+	if (ata_id_has_dword_io(adev->id)) {
+		local_irq_save(flags);
+
+		/* Perform the 32bit I/O synchronization sequence */
+		inb(ap->ioaddr.nsect_addr);
+		inb(ap->ioaddr.nsect_addr);
+		inb(ap->ioaddr.nsect_addr);
+
+		/* Now the data */
+
+		if (write_data)
+			outsl(ap->ioaddr.data_addr, buf, buflen >> 2);
+		else
+			insl(ap->ioaddr.data_addr, buf, buflen >> 2);
+
+		if (unlikely(slop)) {
+			u32 pad;
+			if (write_data) {
+				memcpy(&pad, buf + buflen - slop, slop);
+				outl(le32_to_cpu(pad), ap->ioaddr.data_addr);
+			} else {
+				pad = cpu_to_le16(inl(ap->ioaddr.data_addr));
+				memcpy(buf + buflen - slop, &pad, slop);
+			}
+		}
+		local_irq_restore(flags);
+	}
+	else
+		ata_pio_data_xfer_noirq(adev, buf, buflen, write_data);
+}
+
+static struct ata_port_operations pdc20230_port_ops = {
+	.set_piomode	= pdc20230_set_piomode,
+
+	.port_disable	= ata_port_disable,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.error_handler	= ata_bmdma_error_handler,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= pdc_data_xfer_vlb,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+/*
+ *	Holtek 6560A support
+ *
+ *	This controller supports PIO0 to PIO2 (no IORDY even though higher timings
+ *	can be loaded).
+ */
+
+static void ht6560a_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	u8 active, recover;
+	struct ata_timing t;
+
+	/* Get the timing data in cycles. For now play safe at 50Mhz */
+	ata_timing_compute(adev, adev->pio_mode, &t, 20000, 1000);
+
+	active = FIT(t.active, 2, 15);
+	recover = FIT(t.recover, 4, 15);
+
+	inb(0x3E6);
+	inb(0x3E6);
+	inb(0x3E6);
+	inb(0x3E6);
+
+	outb(recover << 4 | active, ap->ioaddr.device_addr);
+	inb(ap->ioaddr.status_addr);
+}
+
+static struct ata_port_operations ht6560a_port_ops = {
+	.set_piomode	= ht6560a_set_piomode,
+
+	.port_disable	= ata_port_disable,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.error_handler	= ata_bmdma_error_handler,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,	/* Check vlb/noirq */
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+/*
+ *	Holtek 6560B support
+ *
+ *	This controller supports PIO0 to PIO4. We honour the BIOS/jumper FIFO setting
+ *	unless we see an ATAPI device in which case we force it off.
+ *
+ *	FIXME: need to implement 2nd channel support.
+ */
+
+static void ht6560b_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	u8 active, recover;
+	struct ata_timing t;
+
+	/* Get the timing data in cycles. For now play safe at 50Mhz */
+	ata_timing_compute(adev, adev->pio_mode, &t, 20000, 1000);
+
+	active = FIT(t.active, 2, 15);
+	recover = FIT(t.recover, 2, 16);
+	recover &= 0x15;
+
+	inb(0x3E6);
+	inb(0x3E6);
+	inb(0x3E6);
+	inb(0x3E6);
+
+	outb(recover << 4 | active, ap->ioaddr.device_addr);
+
+	if (adev->class != ATA_DEV_ATA) {
+		u8 rconf = inb(0x3E6);
+		if (rconf & 0x24) {
+			rconf &= ~ 0x24;
+			outb(rconf, 0x3E6);
+		}
+	}
+	inb(ap->ioaddr.status_addr);
+}
+
+static struct ata_port_operations ht6560b_port_ops = {
+	.set_piomode	= ht6560b_set_piomode,
+
+	.port_disable	= ata_port_disable,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.error_handler	= ata_bmdma_error_handler,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,	/* FIXME: Check 32bit and noirq */
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+/*
+ *	Opti core chipset helpers
+ */
+ 
+/**
+ *	opti_syscfg	-	read OPTI chipset configuration
+ *	@reg: Configuration register to read
+ *
+ *	Returns the value of an OPTI system board configuration register.
+ */
+
+static u8 opti_syscfg(u8 reg)
+{
+	unsigned long flags;
+	u8 r;
+	
+	/* Uniprocessor chipset and must force cycles adjancent */
+	local_irq_save(flags);
+	outb(reg, 0x22);
+	r = inb(0x24);
+	local_irq_restore(flags);
+	return r;
+}
+
+/*
+ *	Opti 82C611A
+ *
+ *	This controller supports PIO0 to PIO3.
+ */
+
+static void opti82c611a_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	u8 active, recover, setup;
+	struct ata_timing t;
+	struct ata_device *pair = ata_dev_pair(adev);
+	int clock;
+	int khz[4] = { 50000, 40000, 33000, 25000 };
+	u8 rc;
+
+	/* Enter configuration mode */
+	inw(ap->ioaddr.error_addr);
+	inw(ap->ioaddr.error_addr);
+	outb(3, ap->ioaddr.nsect_addr);
+
+	/* Read VLB clock strapping */
+	clock = 1000000000 / khz[inb(ap->ioaddr.lbah_addr) & 0x03];
+
+	/* Get the timing data in cycles */
+	ata_timing_compute(adev, adev->pio_mode, &t, clock, 1000);
+
+	/* Setup timing is shared */
+	if (pair) {
+		struct ata_timing tp;
+		ata_timing_compute(pair, pair->pio_mode, &tp, clock, 1000);
+
+		ata_timing_merge(&t, &tp, &t, ATA_TIMING_SETUP);
+	}
+
+	active = FIT(t.active, 2, 17) - 2;
+	recover = FIT(t.recover, 1, 16) - 1;
+	setup = FIT(t.setup, 1, 4) - 1;
+
+	/* Select the right timing bank for write timing */
+	rc = inb(ap->ioaddr.lbal_addr);
+	rc &= 0x7F;
+	rc |= (adev->devno << 7);
+	outb(rc, ap->ioaddr.lbal_addr);
+
+	/* Write the timings */
+	outb(active << 4 | recover, ap->ioaddr.error_addr);
+
+	/* Select the right bank for read timings, also
+	   load the shared timings for address */
+	rc = inb(ap->ioaddr.device_addr);
+	rc &= 0xC0;
+	rc |= adev->devno;	/* Index select */
+	rc |= (setup << 4) | 0x04;
+	outb(rc, ap->ioaddr.device_addr);
+
+	/* Load the read timings */
+	outb(active << 4 | recover, ap->ioaddr.data_addr);
+
+	/* Ensure the timing register mode is right */
+	rc = inb (ap->ioaddr.lbal_addr);
+	rc &= 0x73;
+	rc |= 0x84;
+	outb(rc, ap->ioaddr.lbal_addr);
+
+	/* Exit command mode */
+	outb(0x83,  ap->ioaddr.nsect_addr);
+}
+
+
+static struct ata_port_operations opti82c611a_port_ops = {
+	.set_piomode	= opti82c611a_set_piomode,
+
+	.port_disable	= ata_port_disable,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.error_handler	= ata_bmdma_error_handler,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+/*
+ *	Opti 82C465MV
+ *
+ *	This controller supports PIO0 to PIO3. Unlike the 611A the MVB
+ *	version is dual channel but doesn't have a lot of unique registers.
+ */
+
+static void opti82c46x_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	u8 active, recover, setup;
+	struct ata_timing t;
+	struct ata_device *pair = ata_dev_pair(adev);
+	int clock;
+	int khz[4] = { 50000, 40000, 33000, 25000 };
+	u8 rc;
+	u8 sysclk;
+
+	/* Get the clock */
+	sysclk = opti_syscfg(0xAC) & 0xC0;	/* BIOS set */
+
+	/* Enter configuration mode */
+	inw(ap->ioaddr.error_addr);
+	inw(ap->ioaddr.error_addr);
+	outb(3, ap->ioaddr.nsect_addr);
+
+	/* Read VLB clock strapping */
+	clock = 1000000000 / khz[sysclk];
+
+	/* Get the timing data in cycles */
+	ata_timing_compute(adev, adev->pio_mode, &t, clock, 1000);
+
+	/* Setup timing is shared */
+	if (pair) {
+		struct ata_timing tp;
+		ata_timing_compute(pair, pair->pio_mode, &tp, clock, 1000);
+
+		ata_timing_merge(&t, &tp, &t, ATA_TIMING_SETUP);
+	}
+
+	active = FIT(t.active, 2, 17) - 2;
+	recover = FIT(t.recover, 1, 16) - 1;
+	setup = FIT(t.setup, 1, 4) - 1;
+
+	/* Select the right timing bank for write timing */
+	rc = inb(ap->ioaddr.lbal_addr);
+	rc &= 0x7F;
+	rc |= (adev->devno << 7);
+	outb(rc, ap->ioaddr.lbal_addr);
+
+	/* Write the timings */
+	outb(active << 4 | recover, ap->ioaddr.error_addr);
+
+	/* Select the right bank for read timings, also
+	   load the shared timings for address */
+	rc = inb(ap->ioaddr.device_addr);
+	rc &= 0xC0;
+	rc |= adev->devno;	/* Index select */
+	rc |= (setup << 4) | 0x04;
+	outb(rc, ap->ioaddr.device_addr);
+
+	/* Load the read timings */
+	outb(active << 4 | recover, ap->ioaddr.data_addr);
+
+	/* Ensure the timing register mode is right */
+	rc = inb (ap->ioaddr.lbal_addr);
+	rc &= 0x73;
+	rc |= 0x84;
+	outb(rc, ap->ioaddr.lbal_addr);
+
+	/* Exit command mode */
+	outb(0x83,  ap->ioaddr.nsect_addr);
+
+	/* We need to know this for quad device on the MVB */
+	ap->host->private_data = ap;
+}
+
+/**
+ *	opt82c465mv_qc_issue_prot	-	command issue
+ *	@qc: command pending
+ *
+ *	Called when the libata layer is about to issue a command. We wrap
+ *	this interface so that we can load the correct ATA timings. The
+ *	MVB has a single set of timing registers and these are shared
+ *	across channels. As there are two registers we really ought to
+ *	track the last two used values as a sort of register window. For
+ *	now we just reload on a channel switch. On the single channel
+ *	setup this condition never fires so we do nothing extra.
+ *
+ *	FIXME: dual channel needs ->serialize support
+ */
+
+static unsigned int opti82c46x_qc_issue_prot(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct ata_device *adev = qc->dev;
+
+	/* If timings are set and for the wrong channel (2nd test is
+	   due to a libata shortcoming and will eventually go I hope) */
+	if (ap->host->private_data != ap->host
+	    && ap->host->private_data != NULL)
+		opti82c46x_set_piomode(ap, adev);
+
+	return ata_qc_issue_prot(qc);
+}
+
+static struct ata_port_operations opti82c46x_port_ops = {
+	.set_piomode	= opti82c46x_set_piomode,
+
+	.port_disable	= ata_port_disable,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.error_handler	= ata_bmdma_error_handler,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= opti82c46x_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+
+/**
+ *	legacy_init_one		-	attach a legacy interface
+ *	@port: port number
+ *	@io: I/O port start
+ *	@ctrl: control port
+ *	@irq: interrupt line
+ *
+ *	Register an ISA bus IDE interface. Such interfaces are PIO and we
+ *	assume do not support IRQ sharing.
+ */
+
+static __init int legacy_init_one(int port, unsigned long io, unsigned long ctrl, int irq)
+{
+	struct legacy_data *ld = &legacy_data[nr_legacy_host];
+	struct ata_probe_ent ae;
+	struct platform_device *pdev;
+	int ret = -EBUSY;
+	struct ata_port_operations *ops = &legacy_port_ops;
+	int pio_modes = pio_mask;
+	u32 mask = (1 << port);
+
+	if (request_region(io, 8, "pata_legacy") == NULL)
+		return -EBUSY;
+	if (request_region(ctrl, 1, "pata_legacy") == NULL)
+		goto fail_io;
+
+	pdev = platform_device_register_simple(DRV_NAME, nr_legacy_host, NULL, 0);
+	if (pdev == NULL)
+		goto fail_dev;
+
+	if (ht6560a & mask) {
+		ops = &ht6560a_port_ops;
+		pio_modes = 0x07;
+	}
+	if (ht6560b & mask) {
+		ops = &ht6560b_port_ops;
+		pio_modes = 0x1F;
+	}
+	if (opti82c611a & mask) {
+		ops = &opti82c611a_port_ops;
+		pio_modes = 0x0F;
+	}
+	if (opti82c46x & mask) {
+		ops = &opti82c46x_port_ops;
+		pio_modes = 0x0F;
+	}
+
+	/* Probe for automatically detectable controllers */
+ 
+	if (io == 0x1F0 && ops == &legacy_port_ops) {
+		unsigned long flags;
+
+		local_irq_save(flags);
+
+		/* Probes */
+		inb(0x1F5);
+		outb(inb(0x1F2) | 0x80, 0x1F2);
+		inb(0x1F2);
+		inb(0x3F6);
+		inb(0x3F6);
+		inb(0x1F2);
+		inb(0x1F2);
+
+		if ((inb(0x1F2) & 0x80) == 0) {
+			/* PDC20230c or 20630 ? */
+			printk(KERN_INFO "PDC20230-C/20630 VLB ATA controller detected.\n");
+				pio_modes = 0x07;
+			ops = &pdc20230_port_ops;
+			udelay(100);
+			inb(0x1F5);
+		} else {
+			outb(0x55, 0x1F2);
+			inb(0x1F2);
+			inb(0x1F2);
+			if (inb(0x1F2) == 0x00) {
+				printk(KERN_INFO "PDC20230-B VLB ATA controller detected.\n");
+			}
+		}
+		local_irq_restore(flags);
+	}
+
+
+	/* Chip does mode setting by command snooping */
+	if (ops == &legacy_port_ops && (autospeed & mask))
+		ops = &simple_port_ops;
+	memset(&ae, 0, sizeof(struct ata_probe_ent));
+	INIT_LIST_HEAD(&ae.node);
+	ae.dev = &pdev->dev;
+	ae.port_ops = ops;
+	ae.sht = &legacy_sht;
+	ae.n_ports = 1;
+	ae.pio_mask = pio_modes;
+	ae.irq = irq;
+	ae.irq_flags = 0;
+	ae.port_flags = ATA_FLAG_SLAVE_POSS|ATA_FLAG_SRST;
+	ae.port[0].cmd_addr = io;
+	ae.port[0].altstatus_addr = ctrl;
+	ae.port[0].ctl_addr =	ctrl;
+	ata_std_ports(&ae.port[0]);
+	ae.private_data = ld;
+
+	ret = ata_device_add(&ae);
+	if (ret == 0) {
+		ret = -ENODEV;
+		goto fail;
+	}
+	legacy_host[nr_legacy_host++] = dev_get_drvdata(&pdev->dev);
+	ld->platform_dev = pdev;
+	return 0;
+
+fail:
+	platform_device_unregister(pdev);
+fail_dev:
+	release_region(ctrl, 1);
+fail_io:
+	release_region(io, 8);
+	return ret;
+}
+
+/**
+ *	legacy_check_special_cases	-	ATA special cases
+ *	@p: PCI device to check
+ *	@master: set this if we find an ATA master
+ *	@master: set this if we find an ATA secondary
+ *
+ *	A small number of vendors implemented early PCI ATA interfaces on bridge logic
+ *	without the ATA interface being PCI visible. Where we have a matching PCI driver
+ *	we must skip the relevant device here. If we don't know about it then the legacy
+ *	driver is the right driver anyway.
+ */
+
+static void legacy_check_special_cases(struct pci_dev *p, int *primary, int *secondary)
+{
+	/* Cyrix CS5510 pre SFF MWDMA ATA on the bridge */
+	if (p->vendor == 0x1078 && p->device == 0x0000) {
+		*primary = *secondary = 1;
+		return;
+	}
+	/* Cyrix CS5520 pre SFF MWDMA ATA on the bridge */
+	if (p->vendor == 0x1078 && p->device == 0x0002) {
+		*primary = *secondary = 1;
+		return;
+	}
+	/* Intel MPIIX - PIO ATA on non PCI side of bridge */
+	if (p->vendor == 0x8086 && p->device == 0x1234) {
+		u16 r;
+		pci_read_config_word(p, 0x6C, &r);
+		if (r & 0x8000) {	/* ATA port enabled */
+			if (r & 0x4000)
+				*secondary = 1;
+			else
+				*primary = 1;
+		}
+		return;
+	}
+}
+
+
+/**
+ *	legacy_init		-	attach legacy interfaces
+ *
+ *	Attach legacy IDE interfaces by scanning the usual IRQ/port suspects.
+ *	Right now we do not scan the ide0 and ide1 address but should do so
+ *	for non PCI systems or systems with no PCI IDE legacy mode devices.
+ *	If you fix that note there are special cases to consider like VLB
+ *	drivers and CS5510/20.
+ */
+
+static __init int legacy_init(void)
+{
+	int i;
+	int ct = 0;
+	int primary = 0;
+	int secondary = 0;
+	int last_port = NR_HOST;
+
+	struct pci_dev *p = NULL;
+
+	for_each_pci_dev(p) {
+		int r;
+		/* Check for any overlap of the system ATA mappings. Native mode controllers
+		   stuck on these addresses or some devices in 'raid' mode won't be found by
+		   the storage class test */
+		for (r = 0; r < 6; r++) {
+			if (pci_resource_start(p, r) == 0x1f0)
+				primary = 1;
+			if (pci_resource_start(p, r) == 0x170)
+				secondary = 1;
+		}
+		/* Check for special cases */
+		legacy_check_special_cases(p, &primary, &secondary);
+
+		/* If PCI bus is present then don't probe for tertiary legacy ports */
+		if (probe_all == 0)
+			last_port = 2;
+	}
+
+	/* If an OPTI 82C46X is present find out where the channels are */	
+	if (opti82c46x) {
+		static const char *optis[4] = {
+			"3/463MV", "5MV",
+			"5MVA", "5MVB"
+		};
+		u8 chans = 1;
+		u8 ctrl = (opti_syscfg(0x30) & 0xC0) >> 6;
+		
+		opti82c46x = 3;	/* Assume master and slave first */
+		printk(KERN_INFO DRV_NAME ": Opti 82C46%s chipset support.\n", optis[ctrl]);
+		if (ctrl == 3)
+			chans = (opti_syscfg(0x3F) & 0x20) ? 2 : 1;
+		ctrl = opti_syscfg(0xAC);
+		/* Check enabled and this port is the 465MV port. On the
+		   MVB we may have two channels */
+		if (ctrl & 8) {
+			if (ctrl & 4)
+				opti82c46x = 2;	/* Slave */
+			else
+				opti82c46x = 1;	/* Master */
+			if (chans == 2)
+				opti82c46x = 3; /* Master and Slave */
+		}	/* Slave only */
+		else if (chans == 1)
+			opti82c46x = 1;
+	}
+
+	for (i = 0; i < last_port; i++) {
+		/* Skip primary if we have seen a PCI one */
+		if (i == 0 && primary == 1)
+			continue;
+		/* Skip secondary if we have seen a PCI one */
+		if (i == 1 && secondary == 1)
+			continue;
+		if (legacy_init_one(i, legacy_port[i],
+				   legacy_port[i] + 0x0206,
+				   legacy_irq[i]) == 0)
+			ct++;
+	}
+	if (ct != 0)
+		return 0;
+	return -ENODEV;
+}
+
+static __exit void legacy_exit(void)
+{
+	int i;
+
+	for (i = 0; i < nr_legacy_host; i++) {
+		struct legacy_data *ld = &legacy_data[i];
+		struct ata_port *ap =legacy_host[i]->ports[0];
+		unsigned long io = ap->ioaddr.cmd_addr;
+		unsigned long ctrl = ap->ioaddr.ctl_addr;
+		ata_host_remove(legacy_host[i]);
+		platform_device_unregister(ld->platform_dev);
+		if (ld->timing)
+			release_region(ld->timing, 2);
+		release_region(io, 8);
+		release_region(ctrl, 1);
+	}
+}
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for legacy ATA");
+MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
+
+module_param(probe_all, int, 0);
+module_param(autospeed, int, 0);
+module_param(ht6560a, int, 0);
+module_param(ht6560b, int, 0);
+module_param(opti82c611a, int, 0);
+module_param(opti82c46x, int, 0);
+module_param(pio_mask, int, 0);
+
+module_init(legacy_init);
+module_exit(legacy_exit);
+
diff --git a/drivers/ata/pata_mpiix.c b/drivers/ata/pata_mpiix.c
new file mode 100644
index 000000000000..1958c4ed09a8
--- /dev/null
+++ b/drivers/ata/pata_mpiix.c
@@ -0,0 +1,313 @@
+/*
+ * pata_mpiix.c 	- Intel MPIIX PATA for new ATA layer
+ *			  (C) 2005-2006 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ * The MPIIX is different enough to the PIIX4 and friends that we give it
+ * a separate driver. The old ide/pci code handles this by just not tuning
+ * MPIIX at all.
+ *
+ * The MPIIX also differs in another important way from the majority of PIIX
+ * devices. The chip is a bridge (pardon the pun) between the old world of
+ * ISA IDE and PCI IDE. Although the ATA timings are PCI configured the actual
+ * IDE controller is not decoded in PCI space and the chip does not claim to
+ * be IDE class PCI. This requires slightly non-standard probe logic compared
+ * with PCI IDE and also that we do not disable the device when our driver is
+ * unloaded (as it has many other functions).
+ *
+ * The driver conciously keeps this logic internally to avoid pushing quirky
+ * PATA history into the clean libata layer.
+ *
+ * Thinkpad specific note: If you boot an MPIIX using thinkpad with a PCMCIA
+ * hard disk present this driver will not detect it. This is not a bug. In this
+ * configuration the secondary port of the MPIIX is disabled and the addresses
+ * are decoded by the PCMCIA bridge and therefore are for a generic IDE driver
+ * to operate.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "pata_mpiix"
+#define DRV_VERSION "0.7.1"
+
+enum {
+	IDETIM = 0x6C,		/* IDE control register */
+	IORDY = (1 << 1),
+	PPE = (1 << 2),
+	FTIM = (1 << 0),
+	ENABLED = (1 << 15),
+	SECONDARY = (1 << 14)
+};
+
+static int mpiix_pre_reset(struct ata_port *ap)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	static const struct pci_bits mpiix_enable_bits[] = {
+		{ 0x6D, 1, 0x80, 0x80 },
+		{ 0x6F, 1, 0x80, 0x80 }
+	};
+
+	if (!pci_test_config_bits(pdev, &mpiix_enable_bits[ap->port_no])) {
+		ata_port_disable(ap);
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	mpiix_error_handler		-	probe reset
+ *	@ap: ATA port
+ *
+ *	Perform the ATA probe and bus reset sequence plus specific handling
+ *	for this hardware. The MPIIX has the enable bits in a different place
+ *	to PIIX4 and friends. As a pure PIO device it has no cable detect
+ */
+
+static void mpiix_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, mpiix_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	mpiix_set_piomode	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Called to do the PIO mode setup. The MPIIX allows us to program the
+ *	IORDY sample point (2-5 clocks), recovery 1-4 clocks and whether
+ *	prefetching or iordy are used.
+ *
+ *	This would get very ugly because we can only program timing for one
+ *	device at a time, the other gets PIO0. Fortunately libata calls
+ *	our qc_issue_prot command before a command is issued so we can
+ *	flip the timings back and forth to reduce the pain.
+ */
+
+static void mpiix_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	int control = 0;
+	int pio = adev->pio_mode - XFER_PIO_0;
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u16 idetim;
+	static const	 /* ISP  RTC */
+	u8 timings[][2]	= { { 0, 0 },
+			    { 0, 0 },
+			    { 1, 0 },
+			    { 2, 1 },
+			    { 2, 3 }, };
+
+	pci_read_config_word(pdev, IDETIM, &idetim);
+	/* Mask the IORDY/TIME/PPE0 bank for this device */
+	if (adev->class == ATA_DEV_ATA)
+		control |= PPE;		/* PPE enable for disk */
+	if (ata_pio_need_iordy(adev))
+		control |= IORDY;	/* IORDY */
+	if (pio > 0)
+		control |= FTIM;	/* This drive is on the fast timing bank */
+
+	/* Mask out timing and clear both TIME bank selects */
+	idetim &= 0xCCEE;
+	idetim &= ~(0x07  << (2 * adev->devno));
+	idetim |= (control << (2 * adev->devno));
+
+	idetim |= (timings[pio][0] << 12) | (timings[pio][1] << 8);
+	pci_write_config_word(pdev, IDETIM, idetim);
+
+	/* We use ap->private_data as a pointer to the device currently
+	   loaded for timing */
+	ap->private_data = adev;
+}
+
+/**
+ *	mpiix_qc_issue_prot	-	command issue
+ *	@qc: command pending
+ *
+ *	Called when the libata layer is about to issue a command. We wrap
+ *	this interface so that we can load the correct ATA timings if
+ *	neccessary. Our logic also clears TIME0/TIME1 for the other device so
+ *	that, even if we get this wrong, cycles to the other device will
+ *	be made PIO0.
+ */
+
+static unsigned int mpiix_qc_issue_prot(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct ata_device *adev = qc->dev;
+
+	/* If modes have been configured and the channel data is not loaded
+	   then load it. We have to check if pio_mode is set as the core code
+	   does not set adev->pio_mode to XFER_PIO_0 while probing as would be
+	   logical */
+
+	if (adev->pio_mode && adev != ap->private_data)
+		mpiix_set_piomode(ap, adev);
+
+	return ata_qc_issue_prot(qc);
+}
+
+static struct scsi_host_template mpiix_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations mpiix_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= mpiix_set_piomode,
+
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= mpiix_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= mpiix_qc_issue_prot,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static int mpiix_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	/* Single threaded by the PCI probe logic */
+	static struct ata_probe_ent probe[2];
+	static int printed_version;
+	u16 idetim;
+	int enabled;
+
+	if (!printed_version++)
+		dev_printk(KERN_DEBUG, &dev->dev, "version " DRV_VERSION "\n");
+
+	/* MPIIX has many functions which can be turned on or off according
+	   to other devices present. Make sure IDE is enabled before we try
+	   and use it */
+
+	pci_read_config_word(dev, IDETIM, &idetim);
+	if (!(idetim & ENABLED))
+		return -ENODEV;
+
+	/* We do our own plumbing to avoid leaking special cases for whacko
+	   ancient hardware into the core code. There are two issues to
+	   worry about.  #1 The chip is a bridge so if in legacy mode and
+	   without BARs set fools the setup.  #2 If you pci_disable_device
+	   the MPIIX your box goes castors up */
+
+	INIT_LIST_HEAD(&probe[0].node);
+	probe[0].dev = pci_dev_to_dev(dev);
+	probe[0].port_ops = &mpiix_port_ops;
+	probe[0].sht = &mpiix_sht;
+	probe[0].pio_mask = 0x1F;
+	probe[0].irq = 14;
+	probe[0].irq_flags = SA_SHIRQ;
+	probe[0].port_flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST;
+	probe[0].n_ports = 1;
+	probe[0].port[0].cmd_addr = 0x1F0;
+	probe[0].port[0].ctl_addr = 0x3F6;
+	probe[0].port[0].altstatus_addr = 0x3F6;
+
+	/* The secondary lurks at different addresses but is otherwise
+	   the same beastie */
+
+	INIT_LIST_HEAD(&probe[1].node);
+	probe[1] = probe[0];
+	probe[1].irq = 15;
+	probe[1].port[0].cmd_addr = 0x170;
+	probe[1].port[0].ctl_addr = 0x376;
+	probe[1].port[0].altstatus_addr = 0x376;
+
+	/* Let libata fill in the port details */
+	ata_std_ports(&probe[0].port[0]);
+	ata_std_ports(&probe[1].port[0]);
+
+	/* Now add the port that is active */
+	enabled = (idetim & SECONDARY) ? 1 : 0;
+
+	if (ata_device_add(&probe[enabled]))
+		return 0;
+	return -ENODEV;
+}
+
+/**
+ *	mpiix_remove_one	-	device unload
+ *	@pdev: PCI device being removed
+ *
+ *	Handle an unplug/unload event for a PCI device. Unload the
+ *	PCI driver but do not use the default handler as we *MUST NOT*
+ *	disable the device as it has other functions.
+ */
+
+static void __devexit mpiix_remove_one(struct pci_dev *pdev)
+{
+	struct device *dev = pci_dev_to_dev(pdev);
+	struct ata_host *host = dev_get_drvdata(dev);
+
+	ata_host_remove(host);
+	dev_set_drvdata(dev, NULL);
+}
+
+
+
+static const struct pci_device_id mpiix[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82371MX), },
+	{ 0, },
+};
+
+static struct pci_driver mpiix_pci_driver = {
+	.name 		= DRV_NAME,
+	.id_table	= mpiix,
+	.probe 		= mpiix_init_one,
+	.remove		= mpiix_remove_one
+};
+
+static int __init mpiix_init(void)
+{
+	return pci_register_driver(&mpiix_pci_driver);
+}
+
+
+static void __exit mpiix_exit(void)
+{
+	pci_unregister_driver(&mpiix_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for Intel MPIIX");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, mpiix);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(mpiix_init);
+module_exit(mpiix_exit);
diff --git a/drivers/ata/pata_netcell.c b/drivers/ata/pata_netcell.c
new file mode 100644
index 000000000000..e3a85778cd4f
--- /dev/null
+++ b/drivers/ata/pata_netcell.c
@@ -0,0 +1,175 @@
+/*
+ *    pata_netcell.c - Netcell PATA driver
+ *
+ *	(c) 2006 Red Hat  <alan@redhat.com>
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+#include <linux/ata.h>
+
+#define DRV_NAME	"pata_netcell"
+#define DRV_VERSION	"0.1.5"
+
+/**
+ *	netcell_probe_init	-	check for 40/80 pin
+ *	@ap: Port
+ *
+ *	Cables are handled by the RAID controller. Report 80 pin.
+ */
+ 
+static int netcell_pre_reset(struct ata_port *ap)
+{
+	ap->cbl = ATA_CBL_PATA80;
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	netcell_probe_reset - Probe specified port on PATA host controller
+ *	@ap: Port to probe
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void netcell_error_handler(struct ata_port *ap)
+{
+	return ata_bmdma_drive_eh(ap, netcell_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/* No PIO or DMA methods needed for this device */
+
+static struct scsi_host_template netcell_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	/* Special handling needed if you have sector or LBA48 limits */
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	/* Use standard CHS mapping rules */
+	.bios_param		= ata_std_bios_param,
+};
+
+static const struct ata_port_operations netcell_ops = {
+	.port_disable		= ata_port_disable,
+
+	/* Task file is PCI ATA format, use helpers */
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.check_status		= ata_check_status,
+	.exec_command		= ata_exec_command,
+	.dev_select		= ata_std_dev_select,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= netcell_error_handler,
+	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+
+	/* BMDMA handling is PCI ATA format, use helpers */
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= ata_qc_issue_prot,
+	.data_xfer		= ata_pio_data_xfer,
+
+	/* Timeout handling. Special recovery hooks here */
+	.eng_timeout		= ata_eng_timeout,
+	.irq_handler		= ata_interrupt,
+	.irq_clear		= ata_bmdma_irq_clear,
+
+	/* Generic PATA PCI ATA helpers */
+	.port_start		= ata_port_start,
+	.port_stop		= ata_port_stop,
+	.host_stop		= ata_host_stop,
+};
+
+
+/**
+ *	netcell_init_one - Register Netcell ATA PCI device with kernel services
+ *	@pdev: PCI device to register
+ *	@ent: Entry in netcell_pci_tbl matching with @pdev
+ *
+ *	Called from kernel PCI layer.
+ *
+ *	LOCKING:
+ *	Inherited from PCI layer (may sleep).
+ *
+ *	RETURNS:
+ *	Zero on success, or -ERRNO value.
+ */
+
+static int netcell_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+	static int printed_version;
+	static struct ata_port_info info = {
+		.sht		= &netcell_sht,
+		.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		/* Actually we don't really care about these as the
+		   firmware deals with it */
+		.pio_mask	= 0x1f,	/* pio0-4 */
+		.mwdma_mask	= 0x07, /* mwdma0-2 */
+		.udma_mask 	= 0x3f, /* UDMA 133 */
+		.port_ops	= &netcell_ops,
+	};
+	static struct ata_port_info *port_info[2] = { &info, &info };
+
+	if (!printed_version++)
+		dev_printk(KERN_DEBUG, &pdev->dev,
+			   "version " DRV_VERSION "\n");
+
+	/* Any chip specific setup/optimisation/messages here */
+	ata_pci_clear_simplex(pdev);
+	
+	/* And let the library code do the work */
+	return ata_pci_init_one(pdev, port_info, 2);
+}
+
+static const struct pci_device_id netcell_pci_tbl[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_NETCELL, PCI_DEVICE_ID_REVOLUTION), },
+	{ }	/* terminate list */
+};
+
+static struct pci_driver netcell_pci_driver = {
+	.name			= DRV_NAME,
+	.id_table		= netcell_pci_tbl,
+	.probe			= netcell_init_one,
+	.remove			= ata_pci_remove_one,
+};
+
+static int __init netcell_init(void)
+{
+	return pci_register_driver(&netcell_pci_driver);
+}
+
+static void __exit netcell_exit(void)
+{
+	pci_unregister_driver(&netcell_pci_driver);
+}
+
+module_init(netcell_init);
+module_exit(netcell_exit);
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("SCSI low-level driver for Netcell PATA RAID");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, netcell_pci_tbl);
+MODULE_VERSION(DRV_VERSION);
+
diff --git a/drivers/ata/pata_ns87410.c b/drivers/ata/pata_ns87410.c
new file mode 100644
index 000000000000..93d6646d2954
--- /dev/null
+++ b/drivers/ata/pata_ns87410.c
@@ -0,0 +1,236 @@
+/*
+ * pata_ns87410.c 	- National Semiconductor 87410 PATA for new ATA layer
+ *			  (C) 2006 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2, or (at your option)
+ *  any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; see the file COPYING.  If not, write to
+ *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "pata_ns87410"
+#define DRV_VERSION "0.4.2"
+
+/**
+ *	ns87410_pre_reset		-	probe begin
+ *	@ap: ATA port
+ *
+ *	Set up cable type and use generic probe init
+ */
+
+static int ns87410_pre_reset(struct ata_port *ap)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	static const struct pci_bits ns87410_enable_bits[] = {
+		{ 0x43, 1, 0x08, 0x08 },
+		{ 0x47, 1, 0x08, 0x08 }
+	};
+
+	if (!pci_test_config_bits(pdev, &ns87410_enable_bits[ap->port_no])) {
+		ata_port_disable(ap);
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	ns87410_error_handler		-	probe reset
+ *	@ap: ATA port
+ *
+ *	Perform the ATA probe and bus reset sequence plus specific handling
+ *	for this hardware. The MPIIX has the enable bits in a different place
+ *	to PIIX4 and friends. As a pure PIO device it has no cable detect
+ */
+
+static void ns87410_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, ns87410_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	ns87410_set_piomode	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Program timing data. This is kept per channel not per device,
+ *	and only affects the data port.
+ */
+
+static void ns87410_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	int port = 0x40 + 4 * ap->port_no;
+	u8 idetcr, idefr;
+	struct ata_timing at;
+
+	static const u8 activebits[15] = {
+		0, 1, 2, 3, 4,
+		5, 5, 6, 6, 6,
+		6, 7, 7, 7, 7
+	};
+
+	static const u8 recoverbits[12] = {
+		0, 1, 2, 3, 4, 5, 6, 6, 7, 7, 7, 7
+	};
+
+	pci_read_config_byte(pdev, port + 3, &idefr);
+
+	if (ata_pio_need_iordy(adev))
+		idefr |= 0x04;	/* IORDY enable */
+	else
+		idefr &= ~0x04;
+
+	if (ata_timing_compute(adev, adev->pio_mode, &at, 30303, 1) < 0) {
+		dev_printk(KERN_ERR, &pdev->dev, "unknown mode %d.\n", adev->pio_mode);
+		return;
+	}
+
+	at.active = FIT(at.active, 2, 16) - 2;
+	at.setup = FIT(at.setup, 1, 4) - 1;
+	at.recover = FIT(at.recover, 1, 12) - 1;
+
+	idetcr = (at.setup << 6) | (recoverbits[at.recover] << 3) | activebits[at.active];
+
+	pci_write_config_byte(pdev, port, idetcr);
+	pci_write_config_byte(pdev, port + 3, idefr);
+	/* We use ap->private_data as a pointer to the device currently
+	   loaded for timing */
+	ap->private_data = adev;
+}
+
+/**
+ *	ns87410_qc_issue_prot	-	command issue
+ *	@qc: command pending
+ *
+ *	Called when the libata layer is about to issue a command. We wrap
+ *	this interface so that we can load the correct ATA timings if
+ *	neccessary.
+ */
+
+static unsigned int ns87410_qc_issue_prot(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct ata_device *adev = qc->dev;
+
+	/* If modes have been configured and the channel data is not loaded
+	   then load it. We have to check if pio_mode is set as the core code
+	   does not set adev->pio_mode to XFER_PIO_0 while probing as would be
+	   logical */
+
+	if (adev->pio_mode && adev != ap->private_data)
+		ns87410_set_piomode(ap, adev);
+
+	return ata_qc_issue_prot(qc);
+}
+
+static struct scsi_host_template ns87410_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations ns87410_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= ns87410_set_piomode,
+
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= ns87410_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ns87410_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static int ns87410_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info = {
+		.sht = &ns87410_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x0F,
+		.port_ops = &ns87410_port_ops
+	};
+	static struct ata_port_info *port_info[2] = {&info, &info};
+	return ata_pci_init_one(dev, port_info, 2);
+}
+
+static const struct pci_device_id ns87410[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_NS, PCI_DEVICE_ID_NS_87410), },
+	{ 0, },
+};
+
+static struct pci_driver ns87410_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= ns87410,
+	.probe 		= ns87410_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init ns87410_init(void)
+{
+	return pci_register_driver(&ns87410_pci_driver);
+}
+
+
+static void __exit ns87410_exit(void)
+{
+	pci_unregister_driver(&ns87410_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for Nat Semi 87410");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, ns87410);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(ns87410_init);
+module_exit(ns87410_exit);
diff --git a/drivers/ata/pata_oldpiix.c b/drivers/ata/pata_oldpiix.c
new file mode 100644
index 000000000000..04c618a2664b
--- /dev/null
+++ b/drivers/ata/pata_oldpiix.c
@@ -0,0 +1,339 @@
+/*
+ *    pata_oldpiix.c - Intel PATA/SATA controllers
+ *
+ *	(C) 2005 Red Hat <alan@redhat.com>
+ *
+ *    Some parts based on ata_piix.c by Jeff Garzik and others.
+ *
+ *    Early PIIX differs significantly from the later PIIX as it lacks
+ *    SITRE and the slave timing registers. This means that you have to
+ *    set timing per channel, or be clever. Libata tells us whenever it
+ *    does drive selection and we use this to reload the timings.
+ *
+ *    Because of these behaviour differences PIIX gets its own driver module.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+#include <linux/ata.h>
+
+#define DRV_NAME	"pata_oldpiix"
+#define DRV_VERSION	"0.5.1"
+
+/**
+ *	oldpiix_pre_reset		-	probe begin
+ *	@ap: ATA port
+ *
+ *	Set up cable type and use generic probe init
+ */
+
+static int oldpiix_pre_reset(struct ata_port *ap)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	static const struct pci_bits oldpiix_enable_bits[] = {
+		{ 0x41U, 1U, 0x80UL, 0x80UL },	/* port 0 */
+		{ 0x43U, 1U, 0x80UL, 0x80UL },	/* port 1 */
+	};
+
+	if (!pci_test_config_bits(pdev, &oldpiix_enable_bits[ap->port_no])) {
+		ata_port_disable(ap);
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	oldpiix_pata_error_handler - Probe specified port on PATA host controller
+ *	@ap: Port to probe
+ *	@classes:
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void oldpiix_pata_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, oldpiix_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	oldpiix_set_piomode - Initialize host controller PATA PIO timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: um
+ *
+ *	Set PIO mode for device, in host controller PCI config space.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void oldpiix_set_piomode (struct ata_port *ap, struct ata_device *adev)
+{
+	unsigned int pio	= adev->pio_mode - XFER_PIO_0;
+	struct pci_dev *dev	= to_pci_dev(ap->host->dev);
+	unsigned int idetm_port= ap->port_no ? 0x42 : 0x40;
+	u16 idetm_data;
+	int control = 0;
+
+	/*
+	 *	See Intel Document 298600-004 for the timing programing rules
+	 *	for PIIX/ICH. Note that the early PIIX does not have the slave
+	 *	timing port at 0x44.
+	 */
+
+	static const	 /* ISP  RTC */
+	u8 timings[][2]	= { { 0, 0 },
+			    { 0, 0 },
+			    { 1, 0 },
+			    { 2, 1 },
+			    { 2, 3 }, };
+
+	if (pio > 2)
+		control |= 1;	/* TIME1 enable */
+	if (ata_pio_need_iordy(adev))
+		control |= 2;	/* IE IORDY */
+
+	/* Intel specifies that the PPE functionality is for disk only */
+	if (adev->class == ATA_DEV_ATA)
+		control |= 4;	/* PPE enable */
+
+	pci_read_config_word(dev, idetm_port, &idetm_data);
+
+	/* Enable PPE, IE and TIME as appropriate. Clear the other
+	   drive timing bits */
+	if (adev->devno == 0) {
+		idetm_data &= 0xCCE0;
+		idetm_data |= control;
+	} else {
+		idetm_data &= 0xCC0E;
+		idetm_data |= (control << 4);
+	}
+	idetm_data |= (timings[pio][0] << 12) |
+			(timings[pio][1] << 8);
+	pci_write_config_word(dev, idetm_port, idetm_data);
+
+	/* Track which port is configured */
+	ap->private_data = adev;
+}
+
+/**
+ *	oldpiix_set_dmamode - Initialize host controller PATA DMA timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: Device to program
+ *	@isich: True if the device is an ICH and has IOCFG registers
+ *
+ *	Set MWDMA mode for device, in host controller PCI config space.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void oldpiix_set_dmamode (struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *dev	= to_pci_dev(ap->host->dev);
+	u8 idetm_port		= ap->port_no ? 0x42 : 0x40;
+	u16 idetm_data;
+
+	static const	 /* ISP  RTC */
+	u8 timings[][2]	= { { 0, 0 },
+			    { 0, 0 },
+			    { 1, 0 },
+			    { 2, 1 },
+			    { 2, 3 }, };
+
+	/*
+	 * MWDMA is driven by the PIO timings. We must also enable
+	 * IORDY unconditionally along with TIME1. PPE has already
+	 * been set when the PIO timing was set.
+	 */
+
+	unsigned int mwdma	= adev->dma_mode - XFER_MW_DMA_0;
+	unsigned int control;
+	const unsigned int needed_pio[3] = {
+		XFER_PIO_0, XFER_PIO_3, XFER_PIO_4
+	};
+	int pio = needed_pio[mwdma] - XFER_PIO_0;
+
+	pci_read_config_word(dev, idetm_port, &idetm_data);
+
+	control = 3;	/* IORDY|TIME0 */
+	/* Intel specifies that the PPE functionality is for disk only */
+	if (adev->class == ATA_DEV_ATA)
+		control |= 4;	/* PPE enable */
+
+	/* If the drive MWDMA is faster than it can do PIO then
+	   we must force PIO into PIO0 */
+
+	if (adev->pio_mode < needed_pio[mwdma])
+		/* Enable DMA timing only */
+		control |= 8;	/* PIO cycles in PIO0 */
+
+	/* Mask out the relevant control and timing bits we will load. Also
+	   clear the other drive TIME register as a precaution */
+	if (adev->devno == 0) {
+		idetm_data &= 0xCCE0;
+		idetm_data |= control;
+	} else {
+		idetm_data &= 0xCC0E;
+		idetm_data |= (control << 4);
+	}
+	idetm_data |= (timings[pio][0] << 12) | (timings[pio][1] << 8);
+	pci_write_config_word(dev, idetm_port, idetm_data);
+
+	/* Track which port is configured */
+	ap->private_data = adev;
+}
+
+/**
+ *	oldpiix_qc_issue_prot	-	command issue
+ *	@qc: command pending
+ *
+ *	Called when the libata layer is about to issue a command. We wrap
+ *	this interface so that we can load the correct ATA timings if
+ *	neccessary. Our logic also clears TIME0/TIME1 for the other device so
+ *	that, even if we get this wrong, cycles to the other device will
+ *	be made PIO0.
+ */
+
+static unsigned int oldpiix_qc_issue_prot(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct ata_device *adev = qc->dev;
+
+	if (adev != ap->private_data) {
+		if (adev->dma_mode)
+			oldpiix_set_dmamode(ap, adev);
+		else if (adev->pio_mode)
+			oldpiix_set_piomode(ap, adev);
+	}
+	return ata_qc_issue_prot(qc);
+}
+
+
+static struct scsi_host_template oldpiix_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static const struct ata_port_operations oldpiix_pata_ops = {
+	.port_disable		= ata_port_disable,
+	.set_piomode		= oldpiix_set_piomode,
+	.set_dmamode		= oldpiix_set_dmamode,
+	.mode_filter		= ata_pci_default_filter,
+
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.check_status		= ata_check_status,
+	.exec_command		= ata_exec_command,
+	.dev_select		= ata_std_dev_select,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= oldpiix_pata_error_handler,
+	.post_internal_cmd 	= ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= oldpiix_qc_issue_prot,
+	.data_xfer		= ata_pio_data_xfer,
+
+	.irq_handler		= ata_interrupt,
+	.irq_clear		= ata_bmdma_irq_clear,
+
+	.port_start		= ata_port_start,
+	.port_stop		= ata_port_stop,
+	.host_stop		= ata_host_stop,
+};
+
+
+/**
+ *	oldpiix_init_one - Register PIIX ATA PCI device with kernel services
+ *	@pdev: PCI device to register
+ *	@ent: Entry in oldpiix_pci_tbl matching with @pdev
+ *
+ *	Called from kernel PCI layer.  We probe for combined mode (sigh),
+ *	and then hand over control to libata, for it to do the rest.
+ *
+ *	LOCKING:
+ *	Inherited from PCI layer (may sleep).
+ *
+ *	RETURNS:
+ *	Zero on success, or -ERRNO value.
+ */
+
+static int oldpiix_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+	static int printed_version;
+	static struct ata_port_info info = {
+		.sht		= &oldpiix_sht,
+		.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask	= 0x1f,	/* pio0-4 */
+		.mwdma_mask	= 0x07, /* mwdma1-2 */
+		.port_ops	= &oldpiix_pata_ops,
+	};
+	static struct ata_port_info *port_info[2] = { &info, &info };
+
+	if (!printed_version++)
+		dev_printk(KERN_DEBUG, &pdev->dev,
+			   "version " DRV_VERSION "\n");
+
+	return ata_pci_init_one(pdev, port_info, 2);
+}
+
+static const struct pci_device_id oldpiix_pci_tbl[] = {
+	{ PCI_DEVICE(0x8086, 0x1230), },
+	{ }	/* terminate list */
+};
+
+static struct pci_driver oldpiix_pci_driver = {
+	.name			= DRV_NAME,
+	.id_table		= oldpiix_pci_tbl,
+	.probe			= oldpiix_init_one,
+	.remove			= ata_pci_remove_one,
+};
+
+static int __init oldpiix_init(void)
+{
+	return pci_register_driver(&oldpiix_pci_driver);
+}
+
+static void __exit oldpiix_exit(void)
+{
+	pci_unregister_driver(&oldpiix_pci_driver);
+}
+
+
+module_init(oldpiix_init);
+module_exit(oldpiix_exit);
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("SCSI low-level driver for early PIIX series controllers");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, oldpiix_pci_tbl);
+MODULE_VERSION(DRV_VERSION);
+
diff --git a/drivers/ata/pata_opti.c b/drivers/ata/pata_opti.c
new file mode 100644
index 000000000000..c3d01325e0e2
--- /dev/null
+++ b/drivers/ata/pata_opti.c
@@ -0,0 +1,292 @@
+/*
+ * pata_opti.c 	- ATI PATA for new ATA layer
+ *			  (C) 2005 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ * Based on
+ *  linux/drivers/ide/pci/opti621.c		Version 0.7	Sept 10, 2002
+ *
+ *  Copyright (C) 1996-1998  Linus Torvalds & authors (see below)
+ *
+ * Authors:
+ * Jaromir Koutek <miri@punknet.cz>,
+ * Jan Harkes <jaharkes@cwi.nl>,
+ * Mark Lord <mlord@pobox.com>
+ * Some parts of code are from ali14xx.c and from rz1000.c.
+ *
+ * Also consulted the FreeBSD prototype driver by Kevin Day to try
+ * and resolve some confusions. Further documentation can be found in
+ * Ralf Brown's interrupt list
+ *
+ * If you have other variants of the Opti range (Viper/Vendetta) please
+ * try this driver with those PCI idents and report back. For the later
+ * chips see the pata_optidma driver
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "pata_opti"
+#define DRV_VERSION "0.2.4"
+
+enum {
+	READ_REG	= 0,	/* index of Read cycle timing register */
+	WRITE_REG 	= 1,	/* index of Write cycle timing register */
+	CNTRL_REG 	= 3,	/* index of Control register */
+	STRAP_REG 	= 5,	/* index of Strap register */
+	MISC_REG 	= 6	/* index of Miscellaneous register */
+};
+
+/**
+ *	opti_pre_reset		-	probe begin
+ *	@ap: ATA port
+ *
+ *	Set up cable type and use generic probe init
+ */
+
+static int opti_pre_reset(struct ata_port *ap)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	static const struct pci_bits opti_enable_bits[] = {
+		{ 0x45, 1, 0x80, 0x00 },
+		{ 0x40, 1, 0x08, 0x00 }
+	};
+
+	if (!pci_test_config_bits(pdev, &opti_enable_bits[ap->port_no])) {
+		ata_port_disable(ap);
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	opti_probe_reset		-	probe reset
+ *	@ap: ATA port
+ *
+ *	Perform the ATA probe and bus reset sequence plus specific handling
+ *	for this hardware. The Opti needs little handling - we have no UDMA66
+ *	capability that needs cable detection. All we must do is check the port
+ *	is enabled.
+ */
+
+static void opti_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, opti_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	opti_write_reg		-	control register setup
+ *	@ap: ATA port
+ *	@value: value
+ *	@reg: control register number
+ *
+ *	The Opti uses magic 'trapdoor' register accesses to do configuration
+ *	rather than using PCI space as other controllers do. The double inw
+ *	on the error register activates configuration mode. We can then write
+ *	the control register
+ */
+
+static void opti_write_reg(struct ata_port *ap, u8 val, int reg)
+{
+	unsigned long regio = ap->ioaddr.cmd_addr;
+
+	/* These 3 unlock the control register access */
+	inw(regio + 1);
+	inw(regio + 1);
+	outb(3, regio + 2);
+
+	/* Do the I/O */
+	outb(val, regio + reg);
+
+	/* Relock */
+	outb(0x83, regio + 2);
+}
+
+#if 0
+/**
+ *	opti_read_reg		-	control register read
+ *	@ap: ATA port
+ *	@reg: control register number
+ *
+ *	The Opti uses magic 'trapdoor' register accesses to do configuration
+ *	rather than using PCI space as other controllers do. The double inw
+ *	on the error register activates configuration mode. We can then read
+ *	the control register
+ */
+
+static u8 opti_read_reg(struct ata_port *ap, int reg)
+{
+	unsigned long regio = ap->ioaddr.cmd_addr;
+	u8 ret;
+	inw(regio + 1);
+	inw(regio + 1);
+	outb(3, regio + 2);
+	ret = inb(regio + reg);
+	outb(0x83, regio + 2);
+}
+#endif
+
+/**
+ *	opti_set_piomode	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Called to do the PIO mode setup. Timing numbers are taken from
+ *	the FreeBSD driver then pre computed to keep the code clean. There
+ *	are two tables depending on the hardware clock speed.
+ */
+
+static void opti_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct ata_device *pair = ata_dev_pair(adev);
+	int clock;
+	int pio = adev->pio_mode - XFER_PIO_0;
+	unsigned long regio = ap->ioaddr.cmd_addr;
+	u8 addr;
+
+	/* Address table precomputed with prefetch off and a DCLK of 2 */
+	static const u8 addr_timing[2][5] = {
+		{ 0x30, 0x20, 0x20, 0x10, 0x10 },
+		{ 0x20, 0x20, 0x10, 0x10, 0x10 }
+	};
+	static const u8 data_rec_timing[2][5] = {
+		{ 0x6B, 0x56, 0x42, 0x32, 0x31 },
+		{ 0x58, 0x44, 0x32, 0x22, 0x21 }
+	};
+
+	outb(0xff, regio + 5);
+	clock = inw(regio + 5) & 1;
+
+	/*
+ 	 *	As with many controllers the address setup time is shared
+ 	 *	and must suit both devices if present.
+	 */
+
+	addr = addr_timing[clock][pio];
+	if (pair) {
+		/* Hardware constraint */
+		u8 pair_addr = addr_timing[clock][pair->pio_mode - XFER_PIO_0];
+		if (pair_addr > addr)
+			addr = pair_addr;
+	}
+
+	/* Commence primary programming sequence */
+	opti_write_reg(ap, adev->devno, MISC_REG);
+	opti_write_reg(ap, data_rec_timing[clock][pio], READ_REG);
+	opti_write_reg(ap, data_rec_timing[clock][pio], WRITE_REG);
+	opti_write_reg(ap, addr, MISC_REG);
+
+	/* Programming sequence complete, override strapping */
+	opti_write_reg(ap, 0x85, CNTRL_REG);
+}
+
+static struct scsi_host_template opti_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations opti_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= opti_set_piomode,
+/*	.set_dmamode	= opti_set_dmamode, */
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= opti_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static int opti_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info = {
+		.sht = &opti_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.port_ops = &opti_port_ops
+	};
+	static struct ata_port_info *port_info[2] = { &info, &info };
+	static int printed_version;
+
+	if (!printed_version++)
+		dev_printk(KERN_DEBUG, &dev->dev, "version " DRV_VERSION "\n");
+
+	return ata_pci_init_one(dev, port_info, 2);
+}
+
+static const struct pci_device_id opti[] = {
+	{ PCI_VENDOR_ID_OPTI, PCI_DEVICE_ID_OPTI_82C621, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0},
+	{ PCI_VENDOR_ID_OPTI, PCI_DEVICE_ID_OPTI_82C825, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 1},
+	{ 0, },
+};
+
+static struct pci_driver opti_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= opti,
+	.probe 		= opti_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init opti_init(void)
+{
+	return pci_register_driver(&opti_pci_driver);
+}
+
+
+static void __exit opti_exit(void)
+{
+	pci_unregister_driver(&opti_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for Opti 621/621X");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, opti);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(opti_init);
+module_exit(opti_exit);
diff --git a/drivers/ata/pata_optidma.c b/drivers/ata/pata_optidma.c
new file mode 100644
index 000000000000..800aea7b9444
--- /dev/null
+++ b/drivers/ata/pata_optidma.c
@@ -0,0 +1,547 @@
+/*
+ * pata_optidma.c 	- Opti DMA PATA for new ATA layer
+ *			  (C) 2006 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ *	The Opti DMA controllers are related to the older PIO PCI controllers
+ *	and indeed the VLB ones. The main differences are that the timing
+ *	numbers are now based off PCI clocks not VLB and differ, and that
+ *	MWDMA is supported.
+ *
+ *	This driver should support Viper-N+, FireStar, FireStar Plus.
+ *
+ *	These devices support virtual DMA for read (aka the CS5520). Later
+ *	chips support UDMA33, but only if the rest of the board logic does,
+ *	so you have to get this right. We don't support the virtual DMA
+ *	but we do handle UDMA.
+ *
+ *	Bits that are worth knowing
+ *		Most control registers are shadowed into I/O registers
+ *		0x1F5 bit 0 tells you if the PCI/VLB clock is 33 or 25Mhz
+ *		Virtual DMA registers *move* between rev 0x02 and rev 0x10
+ *		UDMA requires a 66MHz FSB
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "pata_optidma"
+#define DRV_VERSION "0.2.1"
+
+enum {
+	READ_REG	= 0,	/* index of Read cycle timing register */
+	WRITE_REG 	= 1,	/* index of Write cycle timing register */
+	CNTRL_REG 	= 3,	/* index of Control register */
+	STRAP_REG 	= 5,	/* index of Strap register */
+	MISC_REG 	= 6	/* index of Miscellaneous register */
+};
+
+static int pci_clock;	/* 0 = 33 1 = 25 */
+
+/**
+ *	optidma_pre_reset		-	probe begin
+ *	@ap: ATA port
+ *
+ *	Set up cable type and use generic probe init
+ */
+ 
+static int optidma_pre_reset(struct ata_port *ap)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	static const struct pci_bits optidma_enable_bits = { 
+		0x40, 1, 0x08, 0x00
+	};
+
+	if (ap->port_no && !pci_test_config_bits(pdev, &optidma_enable_bits)) {
+		ata_port_disable(ap);
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	optidma_probe_reset		-	probe reset
+ *	@ap: ATA port
+ *
+ *	Perform the ATA probe and bus reset sequence plus specific handling
+ *	for this hardware. The Opti needs little handling - we have no UDMA66
+ *	capability that needs cable detection. All we must do is check the port
+ *	is enabled.
+ */
+
+static void optidma_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, optidma_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	optidma_unlock		-	unlock control registers
+ *	@ap: ATA port
+ *
+ *	Unlock the control register block for this adapter. Registers must not
+ *	be unlocked in a situation where libata might look at them.
+ */
+ 
+static void optidma_unlock(struct ata_port *ap)
+{
+	unsigned long regio = ap->ioaddr.cmd_addr;
+	
+	/* These 3 unlock the control register access */
+	inw(regio + 1);
+	inw(regio + 1);
+	outb(3, regio + 2);
+}
+
+/**
+ *	optidma_lock		-	issue temporary relock
+ *	@ap: ATA port
+ *
+ *	Re-lock the configuration register settings.
+ */
+ 
+static void optidma_lock(struct ata_port *ap)
+{
+	unsigned long regio = ap->ioaddr.cmd_addr;
+	
+	/* Relock */
+	outb(0x83, regio + 2);
+}
+
+/**
+ *	optidma_set_mode	-	set mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *	@mode: Mode to set
+ *
+ *	Called to do the DMA or PIO mode setup. Timing numbers are all
+ *	pre computed to keep the code clean. There are two tables depending
+ *	on the hardware clock speed.
+ *
+ *	WARNING: While we do this the IDE registers vanish. If we take an
+ *	IRQ here we depend on the host set locking to avoid catastrophe.
+ */
+
+static void optidma_set_mode(struct ata_port *ap, struct ata_device *adev, u8 mode)
+{
+	struct ata_device *pair = ata_dev_pair(adev);
+	int pio = adev->pio_mode - XFER_PIO_0;
+	int dma = adev->dma_mode - XFER_MW_DMA_0;
+	unsigned long regio = ap->ioaddr.cmd_addr;
+	u8 addr;
+
+	/* Address table precomputed with a DCLK of 2 */
+	static const u8 addr_timing[2][5] = {
+		{ 0x30, 0x20, 0x20, 0x10, 0x10 },
+		{ 0x20, 0x20, 0x10, 0x10, 0x10 }
+	};
+	static const u8 data_rec_timing[2][5] = {
+		{ 0x59, 0x46, 0x30, 0x20, 0x20 },
+		{ 0x46, 0x32, 0x20, 0x20, 0x10 }
+	};
+	static const u8 dma_data_rec_timing[2][3] = {
+		{ 0x76, 0x20, 0x20 },
+		{ 0x54, 0x20, 0x10 }
+	};
+
+	/* Switch from IDE to control mode */
+	optidma_unlock(ap);
+	
+
+	/*
+ 	 *	As with many controllers the address setup time is shared
+ 	 *	and must suit both devices if present. FIXME: Check if we
+ 	 *	need to look at slowest of PIO/DMA mode of either device
+	 */
+
+	if (mode >= XFER_MW_DMA_0)
+		addr = 0;
+	else
+		addr = addr_timing[pci_clock][pio];
+	
+	if (pair) {
+		u8 pair_addr;
+		/* Hardware constraint */
+		if (pair->dma_mode)
+			pair_addr = 0;
+		else
+			pair_addr = addr_timing[pci_clock][pair->pio_mode - XFER_PIO_0];
+		if (pair_addr > addr)
+			addr = pair_addr;
+	}
+	
+	/* Commence primary programming sequence */
+	/* First we load the device number into the timing select */
+	outb(adev->devno, regio + MISC_REG);
+	/* Now we load the data timings into read data/write data */
+	if (mode < XFER_MW_DMA_0) {
+		outb(data_rec_timing[pci_clock][pio], regio + READ_REG);
+		outb(data_rec_timing[pci_clock][pio], regio + WRITE_REG);
+	} else if (mode < XFER_UDMA_0) {
+		outb(dma_data_rec_timing[pci_clock][dma], regio + READ_REG);
+		outb(dma_data_rec_timing[pci_clock][dma], regio + WRITE_REG);
+	}
+	/* Finally we load the address setup into the misc register */
+	outb(addr | adev->devno, regio + MISC_REG);
+
+	/* Programming sequence complete, timing 0 dev 0, timing 1 dev 1 */
+	outb(0x85, regio + CNTRL_REG);
+	
+	/* Switch back to IDE mode */
+	optidma_lock(ap);
+	
+	/* Note: at this point our programming is incomplete. We are
+	   not supposed to program PCI 0x43 "things we hacked onto the chip"
+	   until we've done both sets of PIO/DMA timings */
+}
+
+/**
+ *	optiplus_set_mode	-	DMA setup for Firestar Plus
+ *	@ap: ATA port
+ *	@adev: device
+ *	@mode: desired mode
+ *
+ *	The Firestar plus has additional UDMA functionality for UDMA0-2 and
+ *	requires we do some additional work. Because the base work we must do
+ *	is mostly shared we wrap the Firestar setup functionality in this
+ *	one
+ */
+
+static void optiplus_set_mode(struct ata_port *ap, struct ata_device *adev, u8 mode)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u8 udcfg;
+	u8 udslave;
+	int dev2 = 2 * adev->devno;
+	int unit = 2 * ap->port_no + adev->devno;
+	int udma = mode - XFER_UDMA_0;
+	
+	pci_read_config_byte(pdev, 0x44, &udcfg);
+	if (mode <= XFER_UDMA_0) {
+		udcfg &= ~(1 << unit);
+		optidma_set_mode(ap, adev, adev->dma_mode);
+	} else {
+		udcfg |=  (1 << unit);
+		if (ap->port_no) {
+			pci_read_config_byte(pdev, 0x45, &udslave);
+			udslave &= ~(0x03 << dev2);
+			udslave |= (udma << dev2);
+			pci_write_config_byte(pdev, 0x45, udslave);
+		} else {
+			udcfg &= ~(0x30 << dev2);
+			udcfg |= (udma << dev2);
+		}
+	}
+	pci_write_config_byte(pdev, 0x44, udcfg);
+}
+
+/**
+ *	optidma_set_pio_mode	-	PIO setup callback
+ *	@ap: ATA port
+ *	@adev: Device
+ *
+ *	The libata core provides separate functions for handling PIO and
+ *	DMA programming. The architecture of the Firestar makes it easier
+ *	for us to have a common function so we provide wrappers
+ */
+ 
+static void optidma_set_pio_mode(struct ata_port *ap, struct ata_device *adev)
+{
+	optidma_set_mode(ap, adev, adev->pio_mode);
+}
+
+/**
+ *	optidma_set_dma_mode	-	DMA setup callback
+ *	@ap: ATA port
+ *	@adev: Device
+ *
+ *	The libata core provides separate functions for handling PIO and
+ *	DMA programming. The architecture of the Firestar makes it easier
+ *	for us to have a common function so we provide wrappers
+ */
+ 
+static void optidma_set_dma_mode(struct ata_port *ap, struct ata_device *adev)
+{
+	optidma_set_mode(ap, adev, adev->dma_mode);
+}
+
+/**
+ *	optiplus_set_pio_mode	-	PIO setup callback
+ *	@ap: ATA port
+ *	@adev: Device
+ *
+ *	The libata core provides separate functions for handling PIO and
+ *	DMA programming. The architecture of the Firestar makes it easier
+ *	for us to have a common function so we provide wrappers
+ */
+ 
+static void optiplus_set_pio_mode(struct ata_port *ap, struct ata_device *adev)
+{
+	optiplus_set_mode(ap, adev, adev->pio_mode);
+}
+
+/**
+ *	optiplus_set_dma_mode	-	DMA setup callback
+ *	@ap: ATA port
+ *	@adev: Device
+ *
+ *	The libata core provides separate functions for handling PIO and
+ *	DMA programming. The architecture of the Firestar makes it easier
+ *	for us to have a common function so we provide wrappers
+ */
+ 
+static void optiplus_set_dma_mode(struct ata_port *ap, struct ata_device *adev)
+{
+	optiplus_set_mode(ap, adev, adev->dma_mode);
+}
+
+/**
+ *	optidma_make_bits	-	PCI setup helper
+ *	@adev: ATA device
+ *
+ *	Turn the ATA device setup into PCI configuration bits
+ *	for register 0x43 and return the two bits needed.
+ */
+ 
+static u8 optidma_make_bits43(struct ata_device *adev)
+{
+	static const u8 bits43[5] = {
+		0, 0, 0, 1, 2
+	};
+	if (!ata_dev_enabled(adev))
+		return 0;
+	if (adev->dma_mode)
+		return adev->dma_mode - XFER_MW_DMA_0;
+	return bits43[adev->pio_mode - XFER_PIO_0];
+}
+
+/**
+ *	optidma_post_set_mode	-	finalize PCI setup
+ *	@ap: port to set up
+ *
+ *	Finalise the configuration by writing the nibble of extra bits
+ *	of data into the chip.
+ */
+ 
+static void optidma_post_set_mode(struct ata_port *ap)
+{
+	u8 r;
+	int nybble = 4 * ap->port_no;
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	
+	pci_read_config_byte(pdev, 0x43, &r);
+	
+	r &= (0x0F << nybble);
+	r |= (optidma_make_bits43(&ap->device[0]) + 
+	     (optidma_make_bits43(&ap->device[0]) << 2)) << nybble;
+
+	pci_write_config_byte(pdev, 0x43, r);
+}
+
+static struct scsi_host_template optidma_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations optidma_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= optidma_set_pio_mode,
+	.set_dmamode	= optidma_set_dma_mode,
+
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+	.error_handler	= optidma_error_handler,
+	.post_set_mode	= optidma_post_set_mode,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static struct ata_port_operations optiplus_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= optiplus_set_pio_mode,
+	.set_dmamode	= optiplus_set_dma_mode,
+
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+	.error_handler	= optidma_error_handler,
+	.post_set_mode	= optidma_post_set_mode,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+/**
+ *	optiplus_with_udma	-	Look for UDMA capable setup
+ *	@pdev; ATA controller
+ */
+ 
+static int optiplus_with_udma(struct pci_dev *pdev)
+{
+	u8 r;
+	int ret = 0;
+	int ioport = 0x22;
+	struct pci_dev *dev1;
+	
+	/* Find function 1 */
+	dev1 = pci_get_device(0x1045, 0xC701, NULL);
+	if(dev1 == NULL)
+		return 0;
+	
+	/* Rev must be >= 0x10 */
+	pci_read_config_byte(dev1, 0x08, &r);
+	if (r < 0x10)
+		goto done_nomsg;
+	/* Read the chipset system configuration to check our mode */
+	pci_read_config_byte(dev1, 0x5F, &r);
+	ioport |= (r << 8);
+	outb(0x10, ioport);
+	/* Must be 66Mhz sync */
+	if ((inb(ioport + 2) & 1) == 0)
+		goto done;
+
+	/* Check the ATA arbitration/timing is suitable */
+	pci_read_config_byte(pdev, 0x42, &r);
+	if ((r & 0x36) != 0x36)
+		goto done;
+	pci_read_config_byte(dev1, 0x52, &r);
+	if (r & 0x80)	/* IDEDIR disabled */
+		ret = 1;
+done:		
+	printk(KERN_WARNING "UDMA not supported in this configuration.\n");
+done_nomsg:		/* Wrong chip revision */
+	pci_dev_put(dev1);
+	return ret;
+}
+
+static int optidma_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info_82c700 = {
+		.sht = &optidma_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.port_ops = &optidma_port_ops
+	};
+	static struct ata_port_info info_82c700_udma = {
+		.sht = &optidma_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x07,
+		.port_ops = &optiplus_port_ops
+	};
+	static struct ata_port_info *port_info[2];
+	struct ata_port_info *info = &info_82c700;
+	static int printed_version;
+
+	if (!printed_version++)
+		dev_printk(KERN_DEBUG, &dev->dev, "version " DRV_VERSION "\n");
+
+	/* Fixed location chipset magic */
+	inw(0x1F1);
+	inw(0x1F1);
+	pci_clock = inb(0x1F5) & 1;		/* 0 = 33Mhz, 1 = 25Mhz */
+	
+	if (optiplus_with_udma(dev))
+		info = &info_82c700_udma;
+
+	port_info[0] = port_info[1] = info;
+	return ata_pci_init_one(dev, port_info, 2);
+}
+
+static const struct pci_device_id optidma[] = {
+	{ PCI_DEVICE(0x1045, 0xD568), },	/* Opti 82C700 */
+	{ 0, },
+};
+
+static struct pci_driver optidma_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= optidma,
+	.probe 		= optidma_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init optidma_init(void)
+{
+	return pci_register_driver(&optidma_pci_driver);
+}
+
+
+static void __exit optidma_exit(void)
+{
+	pci_unregister_driver(&optidma_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for Opti Firestar/Firestar Plus");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, optidma);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(optidma_init);
+module_exit(optidma_exit);
diff --git a/drivers/ata/pata_pcmcia.c b/drivers/ata/pata_pcmcia.c
new file mode 100644
index 000000000000..2abe0a3bbb92
--- /dev/null
+++ b/drivers/ata/pata_pcmcia.c
@@ -0,0 +1,393 @@
+/*
+ *   pata_pcmcia.c - PCMCIA PATA controller driver.
+ *   Copyright 2005-2006 Red Hat Inc <alan@redhat.com>, all rights reserved.
+ *   PCMCIA ident update Copyright 2006 Marcin Juszkiewicz 
+ *						<openembedded@hrw.one.pl>
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2, or (at your option)
+ *  any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; see the file COPYING.  If not, write to
+ *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ *
+ *   Heavily based upon ide-cs.c
+ *   The initial developer of the original code is David A. Hinds
+ *   <dahinds@users.sourceforge.net>.  Portions created by David A. Hinds
+ *   are Copyright (C) 1999 David A. Hinds.  All Rights Reserved.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/ata.h>
+#include <linux/libata.h>
+
+#include <pcmcia/cs_types.h>
+#include <pcmcia/cs.h>
+#include <pcmcia/cistpl.h>
+#include <pcmcia/ds.h>
+#include <pcmcia/cisreg.h>
+#include <pcmcia/ciscode.h>
+
+
+#define DRV_NAME "pata_pcmcia"
+#define DRV_VERSION "0.2.9"
+
+/*
+ *	Private data structure to glue stuff together
+ */
+
+struct ata_pcmcia_info {
+	struct pcmcia_device *pdev;
+	int		ndev;
+	dev_node_t	node;
+};
+
+static struct scsi_host_template pcmcia_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations pcmcia_port_ops = {
+	.port_disable	= ata_port_disable,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= ata_bmdma_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer_noirq,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+#define CS_CHECK(fn, ret) \
+do { last_fn = (fn); if ((last_ret = (ret)) != 0) goto cs_failed; } while (0)
+
+/**
+ *	pcmcia_init_one		-	attach a PCMCIA interface
+ *	@pdev: pcmcia device
+ *
+ *	Register a PCMCIA IDE interface. Such interfaces are PIO 0 and
+ *	shared IRQ.
+ */
+
+static int pcmcia_init_one(struct pcmcia_device *pdev)
+{
+	struct ata_probe_ent ae;
+	struct ata_pcmcia_info *info;
+	tuple_t tuple;
+	struct {
+		unsigned short buf[128];
+		cisparse_t parse;
+		config_info_t conf;
+		cistpl_cftable_entry_t dflt;
+	} *stk = NULL;
+	cistpl_cftable_entry_t *cfg;
+	int pass, last_ret = 0, last_fn = 0, is_kme = 0, ret = -ENOMEM;
+	unsigned long io_base, ctl_base;
+
+	info = kzalloc(sizeof(*info), GFP_KERNEL);
+	if (info == NULL)
+		return -ENOMEM;
+
+	/* Glue stuff together. FIXME: We may be able to get rid of info with care */
+	info->pdev = pdev;
+	pdev->priv = info;
+
+	/* Set up attributes in order to probe card and get resources */
+	pdev->io.Attributes1 = IO_DATA_PATH_WIDTH_AUTO;
+	pdev->io.Attributes2 = IO_DATA_PATH_WIDTH_8;
+	pdev->io.IOAddrLines = 3;
+	pdev->irq.Attributes = IRQ_TYPE_DYNAMIC_SHARING;
+	pdev->irq.IRQInfo1 = IRQ_LEVEL_ID;
+	pdev->conf.Attributes = CONF_ENABLE_IRQ;
+	pdev->conf.IntType = INT_MEMORY_AND_IO;
+
+	/* Allocate resoure probing structures */
+
+	stk = kzalloc(sizeof(*stk), GFP_KERNEL);
+	if (!stk)
+		goto out1;
+
+	cfg = &stk->parse.cftable_entry;
+
+	/* Tuples we are walking */
+	tuple.TupleData = (cisdata_t *)&stk->buf;
+	tuple.TupleOffset = 0;
+	tuple.TupleDataMax = 255;
+	tuple.Attributes = 0;
+	tuple.DesiredTuple = CISTPL_CONFIG;
+
+	CS_CHECK(GetFirstTuple, pcmcia_get_first_tuple(pdev, &tuple));
+	CS_CHECK(GetTupleData, pcmcia_get_tuple_data(pdev, &tuple));
+	CS_CHECK(ParseTuple, pcmcia_parse_tuple(pdev, &tuple, &stk->parse));
+	pdev->conf.ConfigBase = stk->parse.config.base;
+	pdev->conf.Present = stk->parse.config.rmask[0];
+
+	/* See if we have a manufacturer identifier. Use it to set is_kme for
+	   vendor quirks */
+	tuple.DesiredTuple = CISTPL_MANFID;
+	if (!pcmcia_get_first_tuple(pdev, &tuple) && !pcmcia_get_tuple_data(pdev, &tuple) && !pcmcia_parse_tuple(pdev, &tuple, &stk->parse))
+			is_kme = ((stk->parse.manfid.manf == MANFID_KME) && ((stk->parse.manfid.card == PRODID_KME_KXLC005_A) || (stk->parse.manfid.card == PRODID_KME_KXLC005_B)));
+
+	/* Not sure if this is right... look up the current Vcc */
+	CS_CHECK(GetConfigurationInfo, pcmcia_get_configuration_info(pdev, &stk->conf));
+/*	link->conf.Vcc = stk->conf.Vcc; */
+
+	pass = io_base = ctl_base = 0;
+	tuple.DesiredTuple = CISTPL_CFTABLE_ENTRY;
+	tuple.Attributes = 0;
+	CS_CHECK(GetFirstTuple, pcmcia_get_first_tuple(pdev, &tuple));
+
+	/* Now munch the resources looking for a suitable set */
+	while (1) {
+		if (pcmcia_get_tuple_data(pdev, &tuple) != 0)
+			goto next_entry;
+		if (pcmcia_parse_tuple(pdev, &tuple, &stk->parse) != 0)
+			goto next_entry;
+		/* Check for matching Vcc, unless we're desperate */
+		if (!pass) {
+			if (cfg->vcc.present & (1 << CISTPL_POWER_VNOM)) {
+				if (stk->conf.Vcc != cfg->vcc.param[CISTPL_POWER_VNOM] / 10000)
+					goto next_entry;
+			} else if (stk->dflt.vcc.present & (1 << CISTPL_POWER_VNOM)) {
+				if (stk->conf.Vcc != stk->dflt.vcc.param[CISTPL_POWER_VNOM] / 10000)
+					goto next_entry;
+			}
+		}
+
+		if (cfg->vpp1.present & (1 << CISTPL_POWER_VNOM))
+			pdev->conf.Vpp = cfg->vpp1.param[CISTPL_POWER_VNOM] / 10000;
+		else if (stk->dflt.vpp1.present & (1 << CISTPL_POWER_VNOM))
+			pdev->conf.Vpp = stk->dflt.vpp1.param[CISTPL_POWER_VNOM] / 10000;
+
+		if ((cfg->io.nwin > 0) || (stk->dflt.io.nwin > 0)) {
+			cistpl_io_t *io = (cfg->io.nwin) ? &cfg->io : &stk->dflt.io;
+			pdev->conf.ConfigIndex = cfg->index;
+			pdev->io.BasePort1 = io->win[0].base;
+			pdev->io.IOAddrLines = io->flags & CISTPL_IO_LINES_MASK;
+			if (!(io->flags & CISTPL_IO_16BIT))
+				pdev->io.Attributes1 = IO_DATA_PATH_WIDTH_8;
+			if (io->nwin == 2) {
+				pdev->io.NumPorts1 = 8;
+				pdev->io.BasePort2 = io->win[1].base;
+				pdev->io.NumPorts2 = (is_kme) ? 2 : 1;
+				if (pcmcia_request_io(pdev, &pdev->io) != 0)
+					goto next_entry;
+				io_base = pdev->io.BasePort1;
+				ctl_base = pdev->io.BasePort2;
+			} else if ((io->nwin == 1) && (io->win[0].len >= 16)) {
+				pdev->io.NumPorts1 = io->win[0].len;
+				pdev->io.NumPorts2 = 0;
+				if (pcmcia_request_io(pdev, &pdev->io) != 0)
+					goto next_entry;
+				io_base = pdev->io.BasePort1;
+				ctl_base = pdev->io.BasePort1 + 0x0e;
+			} else goto next_entry;
+			/* If we've got this far, we're done */
+			break;
+		}
+next_entry:
+		if (cfg->flags & CISTPL_CFTABLE_DEFAULT)
+			memcpy(&stk->dflt, cfg, sizeof(stk->dflt));
+		if (pass) {
+			CS_CHECK(GetNextTuple, pcmcia_get_next_tuple(pdev, &tuple));
+		} else if (pcmcia_get_next_tuple(pdev, &tuple) != 0) {
+			CS_CHECK(GetFirstTuple, pcmcia_get_first_tuple(pdev, &tuple));
+			memset(&stk->dflt, 0, sizeof(stk->dflt));
+			pass++;
+		}
+	}
+
+	CS_CHECK(RequestIRQ, pcmcia_request_irq(pdev, &pdev->irq));
+	CS_CHECK(RequestConfiguration, pcmcia_request_configuration(pdev, &pdev->conf));
+
+	/* Success. Disable the IRQ nIEN line, do quirks */
+	outb(0x02, ctl_base);
+	if (is_kme)
+		outb(0x81, ctl_base + 0x01);
+
+	/* FIXME: Could be more ports at base + 0x10 but we only deal with
+	   one right now */
+	if (pdev->io.NumPorts1 >= 0x20)
+		printk(KERN_WARNING DRV_NAME ": second channel not yet supported.\n");
+
+	/*
+ 	 *	Having done the PCMCIA plumbing the ATA side is relatively
+ 	 *	sane.
+	 */
+
+	memset(&ae, 0, sizeof(struct ata_probe_ent));
+	INIT_LIST_HEAD(&ae.node);
+	ae.dev = &pdev->dev;
+	ae.port_ops = &pcmcia_port_ops;
+	ae.sht = &pcmcia_sht;
+	ae.n_ports = 1;
+	ae.pio_mask = 1;		/* ISA so PIO 0 cycles */
+	ae.irq = pdev->irq.AssignedIRQ;
+	ae.irq_flags = SA_SHIRQ;
+	ae.port_flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST;
+	ae.port[0].cmd_addr = io_base;
+	ae.port[0].altstatus_addr = ctl_base;
+	ae.port[0].ctl_addr = ctl_base;
+	ata_std_ports(&ae.port[0]);
+
+	if (ata_device_add(&ae) == 0)
+		goto failed;
+
+	info->ndev = 1;
+	kfree(stk);
+	return 0;
+
+cs_failed:
+	cs_error(pdev, last_fn, last_ret);
+failed:
+	kfree(stk);
+	info->ndev = 0;
+	pcmcia_disable_device(pdev);
+out1:
+	kfree(info);
+	return ret;
+}
+
+/**
+ *	pcmcia_remove_one	-	unplug an pcmcia interface
+ *	@pdev: pcmcia device
+ *
+ *	A PCMCIA ATA device has been unplugged. Perform the needed
+ *	cleanup. Also called on module unload for any active devices.
+ */
+
+static void pcmcia_remove_one(struct pcmcia_device *pdev)
+{
+	struct ata_pcmcia_info *info = pdev->priv;
+	struct device *dev = &pdev->dev;
+
+	if (info != NULL) {
+		/* If we have attached the device to the ATA layer, detach it */
+		if (info->ndev) {
+			struct ata_host *host = dev_get_drvdata(dev);
+			ata_host_remove(host);
+			dev_set_drvdata(dev, NULL);
+		}
+		info->ndev = 0;
+		pdev->priv = NULL;
+	}
+	pcmcia_disable_device(pdev);
+	kfree(info);
+}
+
+static struct pcmcia_device_id pcmcia_devices[] = {
+	PCMCIA_DEVICE_FUNC_ID(4),
+	PCMCIA_DEVICE_MANF_CARD(0x0007, 0x0000),	/* Hitachi */
+	PCMCIA_DEVICE_MANF_CARD(0x0032, 0x0704),
+	PCMCIA_DEVICE_MANF_CARD(0x0045, 0x0401),
+	PCMCIA_DEVICE_MANF_CARD(0x0098, 0x0000),	/* Toshiba */
+	PCMCIA_DEVICE_MANF_CARD(0x00a4, 0x002d),
+	PCMCIA_DEVICE_MANF_CARD(0x00ce, 0x0000),	/* Samsung */
+ 	PCMCIA_DEVICE_MANF_CARD(0x0319, 0x0000),	/* Hitachi */
+	PCMCIA_DEVICE_MANF_CARD(0x2080, 0x0001),
+	PCMCIA_DEVICE_MANF_CARD(0x4e01, 0x0200),	/* Lexar */
+	PCMCIA_DEVICE_PROD_ID123("Caravelle", "PSC-IDE ", "PSC000", 0x8c36137c, 0xd0693ab8, 0x2768a9f0),
+	PCMCIA_DEVICE_PROD_ID123("CDROM", "IDE", "MCD-601p", 0x1b9179ca, 0xede88951, 0x0d902f74),
+	PCMCIA_DEVICE_PROD_ID123("PCMCIA", "IDE CARD", "F1", 0x281f1c5d, 0x1907960c, 0xf7fde8b9),
+	PCMCIA_DEVICE_PROD_ID12("ARGOSY", "CD-ROM", 0x78f308dc, 0x66536591),
+	PCMCIA_DEVICE_PROD_ID12("ARGOSY", "PnPIDE", 0x78f308dc, 0x0c694728),
+	PCMCIA_DEVICE_PROD_ID12("CNF CD-M", "CD-ROM", 0x7d93b852, 0x66536591),
+	PCMCIA_DEVICE_PROD_ID12("Creative Technology Ltd.", "PCMCIA CD-ROM Interface Card", 0xff8c8a45, 0xfe8020c4),
+	PCMCIA_DEVICE_PROD_ID12("Digital Equipment Corporation.", "Digital Mobile Media CD-ROM", 0x17692a66, 0xef1dcbde),
+	PCMCIA_DEVICE_PROD_ID12("EXP", "CD+GAME", 0x6f58c983, 0x63c13aaf),
+	PCMCIA_DEVICE_PROD_ID12("EXP   ", "CD-ROM", 0x0a5c52fd, 0x66536591),
+	PCMCIA_DEVICE_PROD_ID12("EXP   ", "PnPIDE", 0x0a5c52fd, 0x0c694728),
+	PCMCIA_DEVICE_PROD_ID12("FREECOM", "PCCARD-IDE", 0x5714cbf7, 0x48e0ab8e),
+	PCMCIA_DEVICE_PROD_ID12("HITACHI", "FLASH", 0xf4f43949, 0x9eb86aae),
+	PCMCIA_DEVICE_PROD_ID12("HITACHI", "microdrive", 0xf4f43949, 0xa6d76178),
+	PCMCIA_DEVICE_PROD_ID12("IBM", "microdrive", 0xb569a6e5, 0xa6d76178),
+	PCMCIA_DEVICE_PROD_ID12("IBM", "IBM17JSSFP20", 0xb569a6e5, 0xf2508753),
+	PCMCIA_DEVICE_PROD_ID12("IO DATA", "CBIDE2      ", 0x547e66dc, 0x8671043b),
+	PCMCIA_DEVICE_PROD_ID12("IO DATA", "PCIDE", 0x547e66dc, 0x5c5ab149),
+	PCMCIA_DEVICE_PROD_ID12("IO DATA", "PCIDEII", 0x547e66dc, 0xb3662674),
+	PCMCIA_DEVICE_PROD_ID12("LOOKMEET", "CBIDE2      ", 0xe37be2b5, 0x8671043b),
+	PCMCIA_DEVICE_PROD_ID12("M-Systems", "CF500", 0x7ed2ad87, 0x7a13045c),
+	PCMCIA_DEVICE_PROD_ID2("NinjaATA-", 0xebe0bd79),
+	PCMCIA_DEVICE_PROD_ID12("PCMCIA", "CD-ROM", 0x281f1c5d, 0x66536591),
+	PCMCIA_DEVICE_PROD_ID12("PCMCIA", "PnPIDE", 0x281f1c5d, 0x0c694728),
+	PCMCIA_DEVICE_PROD_ID12("SHUTTLE TECHNOLOGY LTD.", "PCCARD-IDE/ATAPI Adapter", 0x4a3f0ba0, 0x322560e1),
+	PCMCIA_DEVICE_PROD_ID12("SEAGATE", "ST1", 0x87c1b330, 0xe1f30883),
+	PCMCIA_DEVICE_PROD_ID12("SAMSUNG", "04/05/06", 0x43d74cb4, 0x6a22777d),
+	PCMCIA_DEVICE_PROD_ID12("SMI VENDOR", "SMI PRODUCT", 0x30896c92, 0x703cc5f6),
+	PCMCIA_DEVICE_PROD_ID12("TOSHIBA", "MK2001MPL", 0xb4585a1a, 0x3489e003),
+	PCMCIA_DEVICE_PROD_ID12("WIT", "IDE16", 0x244e5994, 0x3e232852),
+	PCMCIA_DEVICE_PROD_ID1("STI Flash", 0xe4a13209),
+	PCMCIA_DEVICE_PROD_ID12("STI", "Flash 5.0", 0xbf2df18d, 0x8cb57a0e),
+	PCMCIA_MFC_DEVICE_PROD_ID12(1, "SanDisk", "ConnectPlus", 0x7a954bd9, 0x74be00c6),
+	PCMCIA_DEVICE_NULL,
+};
+
+MODULE_DEVICE_TABLE(pcmcia, pcmcia_devices);
+
+static struct pcmcia_driver pcmcia_driver = {
+	.owner		= THIS_MODULE,
+	.drv = {
+		.name		= DRV_NAME,
+	},
+	.id_table	= pcmcia_devices,
+	.probe		= pcmcia_init_one,
+	.remove		= pcmcia_remove_one,
+};
+
+static int __init pcmcia_init(void)
+{
+	return pcmcia_register_driver(&pcmcia_driver);
+}
+
+static void __exit pcmcia_exit(void)
+{
+	pcmcia_unregister_driver(&pcmcia_driver);
+}
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for PCMCIA ATA");
+MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
+
+module_init(pcmcia_init);
+module_exit(pcmcia_exit);
diff --git a/drivers/ata/pata_pdc2027x.c b/drivers/ata/pata_pdc2027x.c
new file mode 100644
index 000000000000..56b8c1ee2937
--- /dev/null
+++ b/drivers/ata/pata_pdc2027x.c
@@ -0,0 +1,869 @@
+/*
+ *  Promise PATA TX2/TX4/TX2000/133 IDE driver for pdc20268 to pdc20277.
+ *
+ *  This program is free software; you can redistribute it and/or
+ *  modify it under the terms of the GNU General Public License
+ *  as published by the Free Software Foundation; either version
+ *  2 of the License, or (at your option) any later version.
+ *
+ *  Ported to libata by:
+ *  Albert Lee <albertcc@tw.ibm.com> IBM Corporation
+ *
+ *  Copyright (C) 1998-2002		Andre Hedrick <andre@linux-ide.org>
+ *  Portions Copyright (C) 1999 Promise Technology, Inc.
+ *
+ *  Author: Frank Tiernan (frankt@promise.com)
+ *  Released under terms of General Public License
+ *
+ *
+ *  libata documentation is available via 'make {ps|pdf}docs',
+ *  as Documentation/DocBook/libata.*
+ *
+ *  Hardware information only available under NDA.
+ *
+ */
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <scsi/scsi.h>
+#include <scsi/scsi_host.h>
+#include <scsi/scsi_cmnd.h>
+#include <linux/libata.h>
+#include <asm/io.h>
+
+#define DRV_NAME	"pata_pdc2027x"
+#define DRV_VERSION	"0.74-ac3"
+#undef PDC_DEBUG
+
+#ifdef PDC_DEBUG
+#define PDPRINTK(fmt, args...) printk(KERN_ERR "%s: " fmt, __FUNCTION__, ## args)
+#else
+#define PDPRINTK(fmt, args...)
+#endif
+
+enum {
+	PDC_UDMA_100		= 0,
+	PDC_UDMA_133		= 1,
+
+	PDC_100_MHZ		= 100000000,
+	PDC_133_MHZ		= 133333333,
+
+	PDC_SYS_CTL		= 0x1100,
+	PDC_ATA_CTL		= 0x1104,
+	PDC_GLOBAL_CTL		= 0x1108,
+	PDC_CTCR0		= 0x110C,
+	PDC_CTCR1		= 0x1110,
+	PDC_BYTE_COUNT		= 0x1120,
+	PDC_PLL_CTL		= 0x1202,
+};
+
+static int pdc2027x_init_one(struct pci_dev *pdev, const struct pci_device_id *ent);
+static void pdc2027x_remove_one(struct pci_dev *pdev);
+static void pdc2027x_error_handler(struct ata_port *ap);
+static void pdc2027x_set_piomode(struct ata_port *ap, struct ata_device *adev);
+static void pdc2027x_set_dmamode(struct ata_port *ap, struct ata_device *adev);
+static void pdc2027x_post_set_mode(struct ata_port *ap);
+static int pdc2027x_check_atapi_dma(struct ata_queued_cmd *qc);
+
+/*
+ * ATA Timing Tables based on 133MHz controller clock.
+ * These tables are only used when the controller is in 133MHz clock.
+ * If the controller is in 100MHz clock, the ASIC hardware will
+ * set the timing registers automatically when "set feature" command
+ * is issued to the device. However, if the controller clock is 133MHz,
+ * the following tables must be used.
+ */
+static struct pdc2027x_pio_timing {
+	u8 value0, value1, value2;
+} pdc2027x_pio_timing_tbl [] = {
+	{ 0xfb, 0x2b, 0xac }, /* PIO mode 0 */
+	{ 0x46, 0x29, 0xa4 }, /* PIO mode 1 */
+	{ 0x23, 0x26, 0x64 }, /* PIO mode 2 */
+	{ 0x27, 0x0d, 0x35 }, /* PIO mode 3, IORDY on, Prefetch off */
+	{ 0x23, 0x09, 0x25 }, /* PIO mode 4, IORDY on, Prefetch off */
+};
+
+static struct pdc2027x_mdma_timing {
+	u8 value0, value1;
+} pdc2027x_mdma_timing_tbl [] = {
+	{ 0xdf, 0x5f }, /* MDMA mode 0 */
+	{ 0x6b, 0x27 }, /* MDMA mode 1 */
+	{ 0x69, 0x25 }, /* MDMA mode 2 */
+};
+
+static struct pdc2027x_udma_timing {
+	u8 value0, value1, value2;
+} pdc2027x_udma_timing_tbl [] = {
+	{ 0x4a, 0x0f, 0xd5 }, /* UDMA mode 0 */
+	{ 0x3a, 0x0a, 0xd0 }, /* UDMA mode 1 */
+	{ 0x2a, 0x07, 0xcd }, /* UDMA mode 2 */
+	{ 0x1a, 0x05, 0xcd }, /* UDMA mode 3 */
+	{ 0x1a, 0x03, 0xcd }, /* UDMA mode 4 */
+	{ 0x1a, 0x02, 0xcb }, /* UDMA mode 5 */
+	{ 0x1a, 0x01, 0xcb }, /* UDMA mode 6 */
+};
+
+static const struct pci_device_id pdc2027x_pci_tbl[] = {
+	{ PCI_VENDOR_ID_PROMISE, PCI_DEVICE_ID_PROMISE_20268, PCI_ANY_ID, PCI_ANY_ID, 0, 0, PDC_UDMA_100 },
+	{ PCI_VENDOR_ID_PROMISE, PCI_DEVICE_ID_PROMISE_20269, PCI_ANY_ID, PCI_ANY_ID, 0, 0, PDC_UDMA_133 },
+	{ PCI_VENDOR_ID_PROMISE, PCI_DEVICE_ID_PROMISE_20270, PCI_ANY_ID, PCI_ANY_ID, 0, 0, PDC_UDMA_100 },
+	{ PCI_VENDOR_ID_PROMISE, PCI_DEVICE_ID_PROMISE_20271, PCI_ANY_ID, PCI_ANY_ID, 0, 0, PDC_UDMA_133 },
+	{ PCI_VENDOR_ID_PROMISE, PCI_DEVICE_ID_PROMISE_20275, PCI_ANY_ID, PCI_ANY_ID, 0, 0, PDC_UDMA_133 },
+	{ PCI_VENDOR_ID_PROMISE, PCI_DEVICE_ID_PROMISE_20276, PCI_ANY_ID, PCI_ANY_ID, 0, 0, PDC_UDMA_133 },
+	{ PCI_VENDOR_ID_PROMISE, PCI_DEVICE_ID_PROMISE_20277, PCI_ANY_ID, PCI_ANY_ID, 0, 0, PDC_UDMA_133 },
+	{ }	/* terminate list */
+};
+
+static struct pci_driver pdc2027x_pci_driver = {
+	.name			= DRV_NAME,
+	.id_table		= pdc2027x_pci_tbl,
+	.probe			= pdc2027x_init_one,
+	.remove			= __devexit_p(pdc2027x_remove_one),
+};
+
+static struct scsi_host_template pdc2027x_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations pdc2027x_pata100_ops = {
+	.port_disable		= ata_port_disable,
+
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.check_status		= ata_check_status,
+	.exec_command		= ata_exec_command,
+	.dev_select		= ata_std_dev_select,
+
+	.check_atapi_dma	= pdc2027x_check_atapi_dma,
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= ata_qc_issue_prot,
+	.data_xfer		= ata_mmio_data_xfer,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= pdc2027x_error_handler,
+	.post_internal_cmd 	= ata_bmdma_post_internal_cmd,
+
+	.irq_handler		= ata_interrupt,
+	.irq_clear		= ata_bmdma_irq_clear,
+
+	.port_start		= ata_port_start,
+	.port_stop		= ata_port_stop,
+	.host_stop		= ata_pci_host_stop,
+};
+
+static struct ata_port_operations pdc2027x_pata133_ops = {
+	.port_disable		= ata_port_disable,
+	.set_piomode		= pdc2027x_set_piomode,
+	.set_dmamode		= pdc2027x_set_dmamode,
+	.post_set_mode		= pdc2027x_post_set_mode,
+
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.check_status		= ata_check_status,
+	.exec_command		= ata_exec_command,
+	.dev_select		= ata_std_dev_select,
+
+	.check_atapi_dma	= pdc2027x_check_atapi_dma,
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= ata_qc_issue_prot,
+	.data_xfer		= ata_mmio_data_xfer,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= pdc2027x_error_handler,
+	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+
+	.irq_handler		= ata_interrupt,
+	.irq_clear		= ata_bmdma_irq_clear,
+
+	.port_start		= ata_port_start,
+	.port_stop		= ata_port_stop,
+	.host_stop		= ata_pci_host_stop,
+};
+
+static struct ata_port_info pdc2027x_port_info[] = {
+	/* PDC_UDMA_100 */
+	{
+		.sht		= &pdc2027x_sht,
+		.flags		= ATA_FLAG_NO_LEGACY | ATA_FLAG_SLAVE_POSS |
+		                  ATA_FLAG_MMIO,
+		.pio_mask	= 0x1f, /* pio0-4 */
+		.mwdma_mask	= 0x07, /* mwdma0-2 */
+		.udma_mask	= ATA_UDMA5, /* udma0-5 */
+		.port_ops	= &pdc2027x_pata100_ops,
+	},
+	/* PDC_UDMA_133 */
+	{
+		.sht		= &pdc2027x_sht,
+		.flags		= ATA_FLAG_NO_LEGACY | ATA_FLAG_SLAVE_POSS |
+                        	  ATA_FLAG_MMIO,
+		.pio_mask	= 0x1f, /* pio0-4 */
+		.mwdma_mask	= 0x07, /* mwdma0-2 */
+		.udma_mask	= ATA_UDMA6, /* udma0-6 */
+		.port_ops	= &pdc2027x_pata133_ops,
+	},
+};
+
+MODULE_AUTHOR("Andre Hedrick, Frank Tiernan, Albert Lee");
+MODULE_DESCRIPTION("libata driver module for Promise PDC20268 to PDC20277");
+MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
+MODULE_DEVICE_TABLE(pci, pdc2027x_pci_tbl);
+
+/**
+ *	port_mmio - Get the MMIO address of PDC2027x extended registers
+ *	@ap: Port
+ *	@offset: offset from mmio base
+ */
+static inline void* port_mmio(struct ata_port *ap, unsigned int offset)
+{
+	return ap->host->mmio_base + ap->port_no * 0x100 + offset;
+}
+
+/**
+ *	dev_mmio - Get the MMIO address of PDC2027x extended registers
+ *	@ap: Port
+ *	@adev: device
+ *	@offset: offset from mmio base
+ */
+static inline void* dev_mmio(struct ata_port *ap, struct ata_device *adev, unsigned int offset)
+{
+	u8 adj = (adev->devno) ? 0x08 : 0x00;
+	return port_mmio(ap, offset) + adj;
+}
+
+/**
+ *	pdc2027x_pata_cbl_detect - Probe host controller cable detect info
+ *	@ap: Port for which cable detect info is desired
+ *
+ *	Read 80c cable indicator from Promise extended register.
+ *      This register is latched when the system is reset.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+static void pdc2027x_cbl_detect(struct ata_port *ap)
+{
+	u32 cgcr;
+
+	/* check cable detect results */
+	cgcr = readl(port_mmio(ap, PDC_GLOBAL_CTL));
+	if (cgcr & (1 << 26))
+		goto cbl40;
+
+	PDPRINTK("No cable or 80-conductor cable on port %d\n", ap->port_no);
+
+	ap->cbl = ATA_CBL_PATA80;
+	return;
+
+cbl40:
+	printk(KERN_INFO DRV_NAME ": 40-conductor cable detected on port %d\n", ap->port_no);
+	ap->cbl = ATA_CBL_PATA40;
+	ap->udma_mask &= ATA_UDMA_MASK_40C;
+}
+
+/**
+ * pdc2027x_port_enabled - Check PDC ATA control register to see whether the port is enabled.
+ * @ap: Port to check
+ */
+static inline int pdc2027x_port_enabled(struct ata_port *ap)
+{
+	return readb(port_mmio(ap, PDC_ATA_CTL)) & 0x02;
+}
+
+/**
+ *	pdc2027x_prereset - prereset for PATA host controller
+ *	@ap: Target port
+ *
+ *	Probeinit including cable detection.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static int pdc2027x_prereset(struct ata_port *ap)
+{
+	/* Check whether port enabled */
+	if (!pdc2027x_port_enabled(ap)) {
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+	pdc2027x_cbl_detect(ap);
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	pdc2027x_error_handler - Perform reset on PATA port and classify
+ *	@ap: Port to reset
+ *
+ *	Reset PATA phy and classify attached devices.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void pdc2027x_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, pdc2027x_prereset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	pdc2027x_set_piomode - Initialize host controller PATA PIO timings
+ *	@ap: Port to configure
+ *	@adev: um
+ *	@pio: PIO mode, 0 - 4
+ *
+ *	Set PIO mode for device.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void pdc2027x_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	unsigned int pio = adev->pio_mode - XFER_PIO_0;
+	u32 ctcr0, ctcr1;
+
+	PDPRINTK("adev->pio_mode[%X]\n", adev->pio_mode);
+
+	/* Sanity check */
+	if (pio > 4) {
+		printk(KERN_ERR DRV_NAME ": Unknown pio mode [%d] ignored\n", pio);
+		return;
+
+	}
+
+	/* Set the PIO timing registers using value table for 133MHz */
+	PDPRINTK("Set pio regs... \n");
+
+	ctcr0 = readl(dev_mmio(ap, adev, PDC_CTCR0));
+	ctcr0 &= 0xffff0000;
+	ctcr0 |= pdc2027x_pio_timing_tbl[pio].value0 |
+		(pdc2027x_pio_timing_tbl[pio].value1 << 8);
+	writel(ctcr0, dev_mmio(ap, adev, PDC_CTCR0));
+
+	ctcr1 = readl(dev_mmio(ap, adev, PDC_CTCR1));
+	ctcr1 &= 0x00ffffff;
+	ctcr1 |= (pdc2027x_pio_timing_tbl[pio].value2 << 24);
+	writel(ctcr1, dev_mmio(ap, adev, PDC_CTCR1));
+
+	PDPRINTK("Set pio regs done\n");
+
+	PDPRINTK("Set to pio mode[%u] \n", pio);
+}
+
+/**
+ *	pdc2027x_set_dmamode - Initialize host controller PATA UDMA timings
+ *	@ap: Port to configure
+ *	@adev: um
+ *	@udma: udma mode, XFER_UDMA_0 to XFER_UDMA_6
+ *
+ *	Set UDMA mode for device.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+static void pdc2027x_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	unsigned int dma_mode = adev->dma_mode;
+	u32 ctcr0, ctcr1;
+
+	if ((dma_mode >= XFER_UDMA_0) &&
+	   (dma_mode <= XFER_UDMA_6)) {
+		/* Set the UDMA timing registers with value table for 133MHz */
+		unsigned int udma_mode = dma_mode & 0x07;
+
+		if (dma_mode == XFER_UDMA_2) {
+			/*
+			 * Turn off tHOLD.
+			 * If tHOLD is '1', the hardware will add half clock for data hold time.
+			 * This code segment seems to be no effect. tHOLD will be overwritten below.
+			 */
+			ctcr1 = readl(dev_mmio(ap, adev, PDC_CTCR1));
+			writel(ctcr1 & ~(1 << 7), dev_mmio(ap, adev, PDC_CTCR1));
+		}
+
+		PDPRINTK("Set udma regs... \n");
+
+		ctcr1 = readl(dev_mmio(ap, adev, PDC_CTCR1));
+		ctcr1 &= 0xff000000;
+		ctcr1 |= pdc2027x_udma_timing_tbl[udma_mode].value0 |
+			(pdc2027x_udma_timing_tbl[udma_mode].value1 << 8) |
+			(pdc2027x_udma_timing_tbl[udma_mode].value2 << 16);
+		writel(ctcr1, dev_mmio(ap, adev, PDC_CTCR1));
+
+		PDPRINTK("Set udma regs done\n");
+
+		PDPRINTK("Set to udma mode[%u] \n", udma_mode);
+
+	} else  if ((dma_mode >= XFER_MW_DMA_0) &&
+		   (dma_mode <= XFER_MW_DMA_2)) {
+		/* Set the MDMA timing registers with value table for 133MHz */
+		unsigned int mdma_mode = dma_mode & 0x07;
+
+		PDPRINTK("Set mdma regs... \n");
+		ctcr0 = readl(dev_mmio(ap, adev, PDC_CTCR0));
+
+		ctcr0 &= 0x0000ffff;
+		ctcr0 |= (pdc2027x_mdma_timing_tbl[mdma_mode].value0 << 16) |
+			(pdc2027x_mdma_timing_tbl[mdma_mode].value1 << 24);
+
+		writel(ctcr0, dev_mmio(ap, adev, PDC_CTCR0));
+		PDPRINTK("Set mdma regs done\n");
+
+		PDPRINTK("Set to mdma mode[%u] \n", mdma_mode);
+	} else {
+		printk(KERN_ERR DRV_NAME ": Unknown dma mode [%u] ignored\n", dma_mode);
+	}
+}
+
+/**
+ *	pdc2027x_post_set_mode - Set the timing registers back to correct values.
+ *	@ap: Port to configure
+ *
+ *	The pdc2027x hardware will look at "SET FEATURES" and change the timing registers
+ *	automatically. The values set by the hardware might be incorrect, under 133Mhz PLL.
+ *	This function overwrites the possibly incorrect values set by the hardware to be correct.
+ */
+static void pdc2027x_post_set_mode(struct ata_port *ap)
+{
+	int i;
+
+	for (i = 0; i < ATA_MAX_DEVICES; i++) {
+		struct ata_device *dev = &ap->device[i];
+
+		if (ata_dev_enabled(dev)) {
+
+			pdc2027x_set_piomode(ap, dev);
+
+			/*
+			 * Enable prefetch if the device support PIO only.
+			 */
+			if (dev->xfer_shift == ATA_SHIFT_PIO) {
+				u32 ctcr1 = readl(dev_mmio(ap, dev, PDC_CTCR1));
+				ctcr1 |= (1 << 25);
+				writel(ctcr1, dev_mmio(ap, dev, PDC_CTCR1));
+
+				PDPRINTK("Turn on prefetch\n");
+			} else {
+				pdc2027x_set_dmamode(ap, dev);
+			}
+		}
+	}
+}
+
+/**
+ *	pdc2027x_check_atapi_dma - Check whether ATAPI DMA can be supported for this command
+ *	@qc: Metadata associated with taskfile to check
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ *
+ *	RETURNS: 0 when ATAPI DMA can be used
+ *		 1 otherwise
+ */
+static int pdc2027x_check_atapi_dma(struct ata_queued_cmd *qc)
+{
+	struct scsi_cmnd *cmd = qc->scsicmd;
+	u8 *scsicmd = cmd->cmnd;
+	int rc = 1; /* atapi dma off by default */
+
+	/*
+	 * This workaround is from Promise's GPL driver.
+	 * If ATAPI DMA is used for commands not in the
+	 * following white list, say MODE_SENSE and REQUEST_SENSE,
+	 * pdc2027x might hit the irq lost problem.
+	 */
+	switch (scsicmd[0]) {
+	case READ_10:
+	case WRITE_10:
+	case READ_12:
+	case WRITE_12:
+	case READ_6:
+	case WRITE_6:
+	case 0xad: /* READ_DVD_STRUCTURE */
+	case 0xbe: /* READ_CD */
+		/* ATAPI DMA is ok */
+		rc = 0;
+		break;
+	default:
+		;
+	}
+
+	return rc;
+}
+
+/**
+ * pdc_read_counter - Read the ctr counter
+ * @probe_ent: for the port address
+ */
+
+static long pdc_read_counter(struct ata_probe_ent *probe_ent)
+{
+	long counter;
+	int retry = 1;
+	u32 bccrl, bccrh, bccrlv, bccrhv;
+
+retry:
+	bccrl = readl(probe_ent->mmio_base + PDC_BYTE_COUNT) & 0xffff;
+	bccrh = readl(probe_ent->mmio_base + PDC_BYTE_COUNT + 0x100) & 0xffff;
+	rmb();
+
+	/* Read the counter values again for verification */
+	bccrlv = readl(probe_ent->mmio_base + PDC_BYTE_COUNT) & 0xffff;
+	bccrhv = readl(probe_ent->mmio_base + PDC_BYTE_COUNT + 0x100) & 0xffff;
+	rmb();
+
+	counter = (bccrh << 15) | bccrl;
+
+	PDPRINTK("bccrh [%X] bccrl [%X]\n", bccrh,  bccrl);
+	PDPRINTK("bccrhv[%X] bccrlv[%X]\n", bccrhv, bccrlv);
+
+	/*
+	 * The 30-bit decreasing counter are read by 2 pieces.
+	 * Incorrect value may be read when both bccrh and bccrl are changing.
+	 * Ex. When 7900 decrease to 78FF, wrong value 7800 might be read.
+	 */
+	if (retry && !(bccrh == bccrhv && bccrl >= bccrlv)) {
+		retry--;
+		PDPRINTK("rereading counter\n");
+		goto retry;
+	}
+
+	return counter;
+}
+
+/**
+ * adjust_pll - Adjust the PLL input clock in Hz.
+ *
+ * @pdc_controller: controller specific information
+ * @probe_ent: For the port address
+ * @pll_clock: The input of PLL in HZ
+ */
+static void pdc_adjust_pll(struct ata_probe_ent *probe_ent, long pll_clock, unsigned int board_idx)
+{
+
+	u16 pll_ctl;
+	long pll_clock_khz = pll_clock / 1000;
+	long pout_required = board_idx? PDC_133_MHZ:PDC_100_MHZ;
+	long ratio = pout_required / pll_clock_khz;
+	int F, R;
+
+	/* Sanity check */
+	if (unlikely(pll_clock_khz < 5000L || pll_clock_khz > 70000L)) {
+		printk(KERN_ERR DRV_NAME ": Invalid PLL input clock %ldkHz, give up!\n", pll_clock_khz);
+		return;
+	}
+
+#ifdef PDC_DEBUG
+	PDPRINTK("pout_required is %ld\n", pout_required);
+
+	/* Show the current clock value of PLL control register
+	 * (maybe already configured by the firmware)
+	 */
+	pll_ctl = readw(probe_ent->mmio_base + PDC_PLL_CTL);
+
+	PDPRINTK("pll_ctl[%X]\n", pll_ctl);
+#endif
+
+	/*
+	 * Calculate the ratio of F, R and OD
+	 * POUT = (F + 2) / (( R + 2) * NO)
+	 */
+	if (ratio < 8600L) { /* 8.6x */
+		/* Using NO = 0x01, R = 0x0D */
+		R = 0x0d;
+	} else if (ratio < 12900L) { /* 12.9x */
+		/* Using NO = 0x01, R = 0x08 */
+		R = 0x08;
+	} else if (ratio < 16100L) { /* 16.1x */
+		/* Using NO = 0x01, R = 0x06 */
+		R = 0x06;
+	} else if (ratio < 64000L) { /* 64x */
+		R = 0x00;
+	} else {
+		/* Invalid ratio */
+		printk(KERN_ERR DRV_NAME ": Invalid ratio %ld, give up!\n", ratio);
+		return;
+	}
+
+	F = (ratio * (R+2)) / 1000 - 2;
+
+	if (unlikely(F < 0 || F > 127)) {
+		/* Invalid F */
+		printk(KERN_ERR DRV_NAME ": F[%d] invalid!\n", F);
+		return;
+	}
+
+	PDPRINTK("F[%d] R[%d] ratio*1000[%ld]\n", F, R, ratio);
+
+	pll_ctl = (R << 8) | F;
+
+	PDPRINTK("Writing pll_ctl[%X]\n", pll_ctl);
+
+	writew(pll_ctl, probe_ent->mmio_base + PDC_PLL_CTL);
+	readw(probe_ent->mmio_base + PDC_PLL_CTL); /* flush */
+
+	/* Wait the PLL circuit to be stable */
+	mdelay(30);
+
+#ifdef PDC_DEBUG
+	/*
+	 *  Show the current clock value of PLL control register
+	 * (maybe configured by the firmware)
+	 */
+	pll_ctl = readw(probe_ent->mmio_base + PDC_PLL_CTL);
+
+	PDPRINTK("pll_ctl[%X]\n", pll_ctl);
+#endif
+
+	return;
+}
+
+/**
+ * detect_pll_input_clock - Detect the PLL input clock in Hz.
+ * @probe_ent: for the port address
+ * Ex. 16949000 on 33MHz PCI bus for pdc20275.
+ *     Half of the PCI clock.
+ */
+static long pdc_detect_pll_input_clock(struct ata_probe_ent *probe_ent)
+{
+	u32 scr;
+	long start_count, end_count;
+	long pll_clock;
+
+	/* Read current counter value */
+	start_count = pdc_read_counter(probe_ent);
+
+	/* Start the test mode */
+	scr = readl(probe_ent->mmio_base + PDC_SYS_CTL);
+	PDPRINTK("scr[%X]\n", scr);
+	writel(scr | (0x01 << 14), probe_ent->mmio_base + PDC_SYS_CTL);
+	readl(probe_ent->mmio_base + PDC_SYS_CTL); /* flush */
+
+	/* Let the counter run for 100 ms. */
+	mdelay(100);
+
+	/* Read the counter values again */
+	end_count = pdc_read_counter(probe_ent);
+
+	/* Stop the test mode */
+	scr = readl(probe_ent->mmio_base + PDC_SYS_CTL);
+	PDPRINTK("scr[%X]\n", scr);
+	writel(scr & ~(0x01 << 14), probe_ent->mmio_base + PDC_SYS_CTL);
+	readl(probe_ent->mmio_base + PDC_SYS_CTL); /* flush */
+
+	/* calculate the input clock in Hz */
+	pll_clock = (start_count - end_count) * 10;
+
+	PDPRINTK("start[%ld] end[%ld] \n", start_count, end_count);
+	PDPRINTK("PLL input clock[%ld]Hz\n", pll_clock);
+
+	return pll_clock;
+}
+
+/**
+ * pdc_hardware_init - Initialize the hardware.
+ * @pdev: instance of pci_dev found
+ * @pdc_controller: controller specific information
+ * @pe:  for the port address
+ */
+static int pdc_hardware_init(struct pci_dev *pdev, struct ata_probe_ent *pe, unsigned int board_idx)
+{
+	long pll_clock;
+
+	/*
+	 * Detect PLL input clock rate.
+	 * On some system, where PCI bus is running at non-standard clock rate.
+	 * Ex. 25MHz or 40MHz, we have to adjust the cycle_time.
+	 * The pdc20275 controller employs PLL circuit to help correct timing registers setting.
+	 */
+	pll_clock = pdc_detect_pll_input_clock(pe);
+
+	if (pll_clock < 0) /* counter overflow? Try again. */
+		pll_clock = pdc_detect_pll_input_clock(pe);
+
+	dev_printk(KERN_INFO, &pdev->dev, "PLL input clock %ld kHz\n", pll_clock/1000);
+
+	/* Adjust PLL control register */
+	pdc_adjust_pll(pe, pll_clock, board_idx);
+
+	return 0;
+}
+
+/**
+ * pdc_ata_setup_port - setup the mmio address
+ * @port: ata ioports to setup
+ * @base: base address
+ */
+static void pdc_ata_setup_port(struct ata_ioports *port, unsigned long base)
+{
+	port->cmd_addr		=
+	port->data_addr		= base;
+	port->feature_addr	=
+	port->error_addr	= base + 0x05;
+	port->nsect_addr	= base + 0x0a;
+	port->lbal_addr		= base + 0x0f;
+	port->lbam_addr		= base + 0x10;
+	port->lbah_addr		= base + 0x15;
+	port->device_addr	= base + 0x1a;
+	port->command_addr	=
+	port->status_addr	= base + 0x1f;
+	port->altstatus_addr	=
+	port->ctl_addr		= base + 0x81a;
+}
+
+/**
+ * pdc2027x_init_one - PCI probe function
+ * Called when an instance of PCI adapter is inserted.
+ * This function checks whether the hardware is supported,
+ * initialize hardware and register an instance of ata_host to
+ * libata by providing struct ata_probe_ent and ata_device_add().
+ * (implements struct pci_driver.probe() )
+ *
+ * @pdev: instance of pci_dev found
+ * @ent:  matching entry in the id_tbl[]
+ */
+static int __devinit pdc2027x_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+	static int printed_version;
+	unsigned int board_idx = (unsigned int) ent->driver_data;
+
+	struct ata_probe_ent *probe_ent = NULL;
+	unsigned long base;
+	void *mmio_base;
+	int rc;
+
+	if (!printed_version++)
+		dev_printk(KERN_DEBUG, &pdev->dev, "version " DRV_VERSION "\n");
+
+	rc = pci_enable_device(pdev);
+	if (rc)
+		return rc;
+
+	rc = pci_request_regions(pdev, DRV_NAME);
+	if (rc)
+		goto err_out;
+
+	rc = pci_set_dma_mask(pdev, ATA_DMA_MASK);
+	if (rc)
+		goto err_out_regions;
+
+	rc = pci_set_consistent_dma_mask(pdev, ATA_DMA_MASK);
+	if (rc)
+		goto err_out_regions;
+
+	/* Prepare the probe entry */
+	probe_ent = kzalloc(sizeof(*probe_ent), GFP_KERNEL);
+	if (probe_ent == NULL) {
+		rc = -ENOMEM;
+		goto err_out_regions;
+	}
+
+	probe_ent->dev = pci_dev_to_dev(pdev);
+	INIT_LIST_HEAD(&probe_ent->node);
+
+	mmio_base = pci_iomap(pdev, 5, 0);
+	if (!mmio_base) {
+		rc = -ENOMEM;
+		goto err_out_free_ent;
+	}
+
+	base = (unsigned long) mmio_base;
+
+	probe_ent->sht		= pdc2027x_port_info[board_idx].sht;
+	probe_ent->port_flags	= pdc2027x_port_info[board_idx].flags;
+	probe_ent->pio_mask	= pdc2027x_port_info[board_idx].pio_mask;
+	probe_ent->mwdma_mask	= pdc2027x_port_info[board_idx].mwdma_mask;
+	probe_ent->udma_mask	= pdc2027x_port_info[board_idx].udma_mask;
+	probe_ent->port_ops	= pdc2027x_port_info[board_idx].port_ops;
+
+       	probe_ent->irq = pdev->irq;
+       	probe_ent->irq_flags = SA_SHIRQ;
+	probe_ent->mmio_base = mmio_base;
+
+	pdc_ata_setup_port(&probe_ent->port[0], base + 0x17c0);
+	probe_ent->port[0].bmdma_addr = base + 0x1000;
+	pdc_ata_setup_port(&probe_ent->port[1], base + 0x15c0);
+	probe_ent->port[1].bmdma_addr = base + 0x1008;
+
+	probe_ent->n_ports = 2;
+
+	pci_set_master(pdev);
+	//pci_enable_intx(pdev);
+
+	/* initialize adapter */
+	if (pdc_hardware_init(pdev, probe_ent, board_idx) != 0)
+		goto err_out_free_ent;
+
+	ata_device_add(probe_ent);
+	kfree(probe_ent);
+
+	return 0;
+
+err_out_free_ent:
+	kfree(probe_ent);
+err_out_regions:
+	pci_release_regions(pdev);
+err_out:
+	pci_disable_device(pdev);
+	return rc;
+}
+
+/**
+ * pdc2027x_remove_one - Called to remove a single instance of the
+ * adapter.
+ *
+ * @dev: The PCI device to remove.
+ * FIXME: module load/unload not working yet
+ */
+static void __devexit pdc2027x_remove_one(struct pci_dev *pdev)
+{
+	ata_pci_remove_one(pdev);
+}
+
+/**
+ * pdc2027x_init - Called after this module is loaded into the kernel.
+ */
+static int __init pdc2027x_init(void)
+{
+	return pci_module_init(&pdc2027x_pci_driver);
+}
+
+/**
+ * pdc2027x_exit - Called before this module unloaded from the kernel
+ */
+static void __exit pdc2027x_exit(void)
+{
+	pci_unregister_driver(&pdc2027x_pci_driver);
+}
+
+module_init(pdc2027x_init);
+module_exit(pdc2027x_exit);
diff --git a/drivers/ata/pata_pdc202xx_old.c b/drivers/ata/pata_pdc202xx_old.c
new file mode 100644
index 000000000000..6cb52b0e7696
--- /dev/null
+++ b/drivers/ata/pata_pdc202xx_old.c
@@ -0,0 +1,423 @@
+/*
+ * pata_pdc202xx_old.c 	- Promise PDC202xx PATA for new ATA layer
+ *			  (C) 2005 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ * Based in part on linux/drivers/ide/pci/pdc202xx_old.c
+ *
+ * First cut with LBA48/ATAPI
+ *
+ * TODO:
+ *	Channel interlock/reset on both required ?
+ */
+ 
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "pata_pdc202xx_old"
+#define DRV_VERSION "0.2.1"
+
+/**
+ *	pdc2024x_pre_reset		-	probe begin
+ *	@ap: ATA port
+ *
+ *	Set up cable type and use generic probe init
+ */
+ 
+static int pdc2024x_pre_reset(struct ata_port *ap)
+{
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+
+static void pdc2024x_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, pdc2024x_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+
+static int pdc2026x_pre_reset(struct ata_port *ap)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u16 cis;
+	
+	pci_read_config_word(pdev, 0x50, &cis);
+	if (cis & (1 << (10 + ap->port_no)))
+		ap->cbl = ATA_CBL_PATA80;
+	else
+		ap->cbl = ATA_CBL_PATA40;
+
+	return ata_std_prereset(ap);
+}
+
+static void pdc2026x_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, pdc2026x_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	pdc_configure_piomode	-	set chip PIO timing
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *	@pio: PIO mode
+ *
+ *	Called to do the PIO mode setup. Our timing registers are shared
+ *	so a configure_dmamode call will undo any work we do here and vice
+ *	versa
+ */
+ 
+static void pdc_configure_piomode(struct ata_port *ap, struct ata_device *adev, int pio)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	int port = 0x60 + 4 * ap->port_no + 2 * adev->devno;
+	static u16 pio_timing[5] = {
+		0x0913, 0x050C , 0x0308, 0x0206, 0x0104
+	};
+	u8 r_ap, r_bp;
+
+	pci_read_config_byte(pdev, port, &r_ap);
+	pci_read_config_byte(pdev, port + 1, &r_bp);
+	r_ap &= ~0x3F;	/* Preserve ERRDY_EN, SYNC_IN */
+	r_bp &= ~0x07;
+	r_ap |= (pio_timing[pio] >> 8);
+	r_bp |= (pio_timing[pio] & 0xFF);
+	
+	if (ata_pio_need_iordy(adev))
+		r_ap |= 0x20;	/* IORDY enable */
+	if (adev->class == ATA_DEV_ATA)
+		r_ap |= 0x10;	/* FIFO enable */
+	pci_write_config_byte(pdev, port, r_ap);
+	pci_write_config_byte(pdev, port + 1, r_bp);
+}
+
+/**
+ *	pdc_set_piomode	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Called to do the PIO mode setup. Our timing registers are shared
+ *	but we want to set the PIO timing by default.
+ */
+ 
+static void pdc_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	pdc_configure_piomode(ap, adev, adev->pio_mode - XFER_PIO_0);
+}
+
+/**
+ *	pdc_configure_dmamode	-	set DMA mode in chip
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Load DMA cycle times into the chip ready for a DMA transfer
+ *	to occur.
+ */
+ 
+static void pdc_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	int port = 0x60 + 4 * ap->port_no + 2 * adev->devno;
+	static u8 udma_timing[6][2] = {
+		{ 0x60, 0x03 },	/* 33 Mhz Clock */
+		{ 0x40, 0x02 },
+		{ 0x20, 0x01 },
+		{ 0x40, 0x02 },	/* 66 Mhz Clock */
+		{ 0x20, 0x01 },
+		{ 0x20, 0x01 }		
+	};
+	u8 r_bp, r_cp;
+	
+	pci_read_config_byte(pdev, port + 1, &r_bp);
+	pci_read_config_byte(pdev, port + 2, &r_cp);
+	
+	r_bp &= ~0xF0;
+	r_cp &= ~0x0F;
+	
+	if (adev->dma_mode >= XFER_UDMA_0) {
+		int speed = adev->dma_mode - XFER_UDMA_0;
+		r_bp |= udma_timing[speed][0];
+		r_cp |= udma_timing[speed][1];
+		
+	} else {
+		int speed = adev->dma_mode - XFER_MW_DMA_0;
+		r_bp |= 0x60;
+		r_cp |= (5 - speed);
+	}
+	pci_write_config_byte(pdev, port + 1, r_bp);
+	pci_write_config_byte(pdev, port + 2, r_cp);
+	
+}
+
+/**
+ *	pdc2026x_bmdma_start		-	DMA engine begin
+ *	@qc: ATA command
+ *
+ *	In UDMA3 or higher we have to clock switch for the duration of the
+ *	DMA transfer sequence.
+ */
+ 
+static void pdc2026x_bmdma_start(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct ata_device *adev = qc->dev;
+	struct ata_taskfile *tf = &qc->tf;
+	int sel66 = ap->port_no ? 0x08: 0x02;
+
+	unsigned long master = ap->host->ports[0]->ioaddr.bmdma_addr;
+	unsigned long clock = master + 0x11;
+	unsigned long atapi_reg = master + 0x20 + (4 * ap->port_no);
+	
+	u32 len;
+	
+	/* Check we keep host level locking here */
+	if (adev->dma_mode >= XFER_UDMA_2)
+		outb(inb(clock) | sel66, clock);
+	else
+		outb(inb(clock) & ~sel66, clock);
+
+	/* The DMA clocks may have been trashed by a reset. FIXME: make conditional 
+	   and move to qc_issue ? */
+	pdc_set_dmamode(ap, qc->dev);
+
+	/* Cases the state machine will not complete correctly without help */
+	if ((tf->flags & ATA_TFLAG_LBA48) ||  tf->protocol == ATA_PROT_ATAPI_DMA)
+	{
+		if (tf->flags & ATA_TFLAG_LBA48)
+			len = qc->nsect * 512;
+		else
+			len = qc->nbytes;
+		
+		if (tf->flags & ATA_TFLAG_WRITE)
+			len |= 0x06000000;
+		else
+			len |= 0x05000000;
+			
+		outl(len, atapi_reg);
+	}
+	
+	/* Activate DMA */	
+	ata_bmdma_start(qc);
+}
+
+/**
+ *	pdc2026x_bmdma_end		-	DMA engine stop
+ *	@qc: ATA command
+ *
+ *	After a DMA completes we need to put the clock back to 33MHz for
+ *	PIO timings.
+ */
+ 
+static void pdc2026x_bmdma_stop(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct ata_device *adev = qc->dev;
+	struct ata_taskfile *tf = &qc->tf;
+	
+	int sel66 = ap->port_no ? 0x08: 0x02;
+	/* The clock bits are in the same register for both channels */
+	unsigned long master = ap->host->ports[0]->ioaddr.bmdma_addr;
+	unsigned long clock = master + 0x11;
+	unsigned long atapi_reg = master + 0x20 + (4 * ap->port_no);
+	
+	/* Cases the state machine will not complete correctly */
+	if (tf->protocol == ATA_PROT_ATAPI_DMA || ( tf->flags & ATA_TFLAG_LBA48)) {
+		outl(0, atapi_reg);
+		outb(inb(clock) & ~sel66, clock);
+	}
+	/* Check we keep host level locking here */
+	/* Flip back to 33Mhz for PIO */
+	if (adev->dma_mode >= XFER_UDMA_2)
+		outb(inb(clock) & ~sel66, clock);
+
+	ata_bmdma_stop(qc);
+}
+
+/**
+ *	pdc2026x_dev_config	-	device setup hook
+ *	@ap: ATA port
+ *	@adev: newly found device
+ *
+ *	Perform chip specific early setup. We need to lock the transfer
+ *	sizes to 8bit to avoid making the state engine on the 2026x cards
+ *	barf.
+ */
+ 
+static void pdc2026x_dev_config(struct ata_port *ap, struct ata_device *adev)
+{
+	adev->max_sectors = 256;
+}
+
+static struct scsi_host_template pdc_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations pdc2024x_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= pdc_set_piomode,
+	.set_dmamode	= pdc_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= pdc2024x_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+	
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+static struct ata_port_operations pdc2026x_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= pdc_set_piomode,
+	.set_dmamode	= pdc_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+	.dev_config	= pdc2026x_dev_config,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= pdc2026x_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= pdc2026x_bmdma_start,
+	.bmdma_stop	= pdc2026x_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+	
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+static int pdc_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info[3] = {
+		{
+			.sht = &pdc_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = ATA_UDMA2,
+			.port_ops = &pdc2024x_port_ops
+		}, 
+		{
+			.sht = &pdc_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = ATA_UDMA4,
+			.port_ops = &pdc2026x_port_ops
+		},
+		{
+			.sht = &pdc_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = ATA_UDMA5,
+			.port_ops = &pdc2026x_port_ops
+		}
+		
+	};
+	static struct ata_port_info *port_info[2];
+
+	port_info[0] = port_info[1] = &info[id->driver_data];
+	
+	if (dev->device == PCI_DEVICE_ID_PROMISE_20265) {
+		struct pci_dev *bridge = dev->bus->self;
+		/* Don't grab anything behind a Promise I2O RAID */
+		if (bridge && bridge->vendor == PCI_VENDOR_ID_INTEL) {
+			if( bridge->device == PCI_DEVICE_ID_INTEL_I960)
+				return -ENODEV;
+			if( bridge->device == PCI_DEVICE_ID_INTEL_I960RM)
+				return -ENODEV;
+		}
+	}
+	return ata_pci_init_one(dev, port_info, 2);
+}
+
+static struct pci_device_id pdc[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_PROMISE, PCI_DEVICE_ID_PROMISE_20246), 0},
+	{ PCI_DEVICE(PCI_VENDOR_ID_PROMISE, PCI_DEVICE_ID_PROMISE_20262), 1},
+	{ PCI_DEVICE(PCI_VENDOR_ID_PROMISE, PCI_DEVICE_ID_PROMISE_20263), 1},
+	{ PCI_DEVICE(PCI_VENDOR_ID_PROMISE, PCI_DEVICE_ID_PROMISE_20265), 2},
+	{ PCI_DEVICE(PCI_VENDOR_ID_PROMISE, PCI_DEVICE_ID_PROMISE_20267), 2},
+	{ 0, },
+};
+
+static struct pci_driver pdc_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= pdc,
+	.probe 		= pdc_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init pdc_init(void)
+{
+	return pci_register_driver(&pdc_pci_driver);
+}
+
+
+static void __exit pdc_exit(void)
+{
+	pci_unregister_driver(&pdc_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for Promise 2024x and 20262-20267");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, pdc);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(pdc_init);
+module_exit(pdc_exit);
diff --git a/drivers/ata/pata_qdi.c b/drivers/ata/pata_qdi.c
new file mode 100644
index 000000000000..ededb40e084d
--- /dev/null
+++ b/drivers/ata/pata_qdi.c
@@ -0,0 +1,403 @@
+/*
+ *    pata_qdi.c - QDI VLB ATA controllers
+ *	(C) 2006 Red Hat <alan@redhat.com>
+ *
+ * This driver mostly exists as a proof of concept for non PCI devices under
+ * libata. While the QDI6580 was 'neat' in 1993 it is no longer terribly
+ * useful.
+ *
+ * Tuning code written from the documentation at
+ * http://www.ryston.cz/petr/vlb/qd6500.html
+ * http://www.ryston.cz/petr/vlb/qd6580.html
+ *
+ * Probe code based on drivers/ide/legacy/qd65xx.c
+ * Rewritten from the work of Colten Edwards <pje120@cs.usask.ca> by
+ * Samuel Thibault <samuel.thibault@fnac.net>
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+#include <linux/platform_device.h>
+
+#define DRV_NAME "pata_qdi"
+#define DRV_VERSION "0.2.4"
+
+#define NR_HOST 4	/* Two 6580s */
+
+struct qdi_data {
+	unsigned long timing;
+	u8 clock[2];
+	u8 last;
+	int fast;
+	struct platform_device *platform_dev;
+	
+};
+
+static struct ata_host *qdi_host[NR_HOST];
+static struct qdi_data qdi_data[NR_HOST];
+static int nr_qdi_host;
+
+#ifdef MODULE
+static int probe_qdi = 1;
+#else
+static int probe_qdi;
+#endif
+
+static void qdi6500_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct ata_timing t;
+	struct qdi_data *qdi = ap->host->private_data;
+	int active, recovery;
+	u8 timing;
+
+	/* Get the timing data in cycles */
+	ata_timing_compute(adev, adev->pio_mode, &t, 30303, 1000);
+	
+	if (qdi->fast) {
+		active = 8 - FIT(t.active, 1, 8);
+		recovery = 18 - FIT(t.recover, 3, 18);
+	} else {
+		active = 9 - FIT(t.active, 2, 9);
+		recovery = 15 - FIT(t.recover, 0, 15);
+	}
+	timing = (recovery << 4) | active | 0x08;
+	
+	qdi->clock[adev->devno] = timing;
+
+	outb(timing, qdi->timing);	
+}
+
+static void qdi6580_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct ata_timing t;
+	struct qdi_data *qdi = ap->host->private_data;
+	int active, recovery;
+	u8 timing;
+
+	/* Get the timing data in cycles */
+	ata_timing_compute(adev, adev->pio_mode, &t, 30303, 1000);
+	
+	if (qdi->fast) {
+		active = 8 - FIT(t.active, 1, 8);
+		recovery = 18 - FIT(t.recover, 3, 18);
+	} else {
+		active = 9 - FIT(t.active, 2, 9);
+		recovery = 15 - FIT(t.recover, 0, 15);
+	}
+	timing = (recovery << 4) | active | 0x08;
+	
+	qdi->clock[adev->devno] = timing;
+
+	outb(timing, qdi->timing);
+	
+	/* Clear the FIFO */
+	if (adev->class != ATA_DEV_ATA)
+		outb(0x5F, (qdi->timing & 0xFFF0) + 3);
+}
+
+/**
+ *	qdi_qc_issue_prot	-	command issue
+ *	@qc: command pending
+ *
+ *	Called when the libata layer is about to issue a command. We wrap
+ *	this interface so that we can load the correct ATA timings.
+ */
+
+static unsigned int qdi_qc_issue_prot(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct ata_device *adev = qc->dev;
+	struct qdi_data *qdi = ap->host->private_data;
+
+	if (qdi->clock[adev->devno] != qdi->last) {
+		if (adev->pio_mode) {
+			qdi->last = qdi->clock[adev->devno];
+			outb(qdi->clock[adev->devno], qdi->timing);
+		}
+	}
+	return ata_qc_issue_prot(qc);
+}
+
+static void qdi_data_xfer(struct ata_device *adev, unsigned char *buf, unsigned int buflen, int write_data)
+{
+	struct ata_port *ap = adev->ap;
+	int slop = buflen & 3;
+	
+	if (ata_id_has_dword_io(adev->id)) {
+		if (write_data)
+			outsl(ap->ioaddr.data_addr, buf, buflen >> 2);
+		else
+			insl(ap->ioaddr.data_addr, buf, buflen >> 2);
+	
+		if (unlikely(slop)) {
+			u32 pad;
+			if (write_data) {
+				memcpy(&pad, buf + buflen - slop, slop);
+				outl(le32_to_cpu(pad), ap->ioaddr.data_addr);
+			} else {
+				pad = cpu_to_le16(inl(ap->ioaddr.data_addr));
+				memcpy(buf + buflen - slop, &pad, slop);
+			}
+		}
+	} else
+		ata_pio_data_xfer(adev, buf, buflen, write_data);
+}
+
+static struct scsi_host_template qdi_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations qdi6500_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= qdi6500_set_piomode,
+
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= ata_bmdma_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+	
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= qdi_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= qdi_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+	
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+static struct ata_port_operations qdi6580_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= qdi6580_set_piomode,
+
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= ata_bmdma_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+	
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= qdi_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= qdi_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+	
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+/**
+ *	qdi_init_one		-	attach a qdi interface
+ *	@type: Type to display
+ *	@io: I/O port start
+ *	@irq: interrupt line
+ *	@fast: True if on a > 33Mhz VLB
+ *
+ *	Register an ISA bus IDE interface. Such interfaces are PIO and we
+ *	assume do not support IRQ sharing.
+ */
+ 
+static __init int qdi_init_one(unsigned long port, int type, unsigned long io, int irq, int fast)
+{
+	struct ata_probe_ent ae;
+	struct platform_device *pdev;
+	int ret;
+
+	unsigned long ctrl = io + 0x206;
+
+	/*
+	 *	Fill in a probe structure first of all
+	 */
+
+	pdev = platform_device_register_simple(DRV_NAME, nr_qdi_host, NULL, 0);
+	if (pdev == NULL)
+		return -ENOMEM;
+	
+	memset(&ae, 0, sizeof(struct ata_probe_ent));
+	INIT_LIST_HEAD(&ae.node);
+	ae.dev = &pdev->dev;
+	
+	if (type == 6580) {
+		ae.port_ops = &qdi6580_port_ops;
+		ae.pio_mask = 0x1F;
+	} else {
+		ae.port_ops = &qdi6500_port_ops;
+		ae.pio_mask = 0x07;	/* Actually PIO3 !IORDY is possible */
+	}
+	
+	ae.sht = &qdi_sht;
+	ae.n_ports = 1;
+	ae.irq = irq;
+	ae.irq_flags = 0;
+	ae.port_flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST;
+	ae.port[0].cmd_addr = io;
+	ae.port[0].altstatus_addr = ctrl;
+	ae.port[0].ctl_addr =	ctrl;
+	ata_std_ports(&ae.port[0]);
+
+	/*
+	 *	Hook in a private data structure per channel
+	 */
+	ae.private_data = &qdi_data[nr_qdi_host];
+
+	qdi_data[nr_qdi_host].timing = port;
+	qdi_data[nr_qdi_host].fast = fast;
+	qdi_data[nr_qdi_host].platform_dev = pdev;
+
+	printk(KERN_INFO DRV_NAME": qd%d at 0x%lx.\n", type, io);
+	ret = ata_device_add(&ae);
+	if (ret == 0) {
+		platform_device_unregister(pdev);
+		return -ENODEV;
+	}
+		
+	qdi_host[nr_qdi_host++] = dev_get_drvdata(&pdev->dev);
+	return 0;	
+}
+
+/**
+ *	qdi_init		-	attach qdi interfaces
+ *
+ *	Attach qdi IDE interfaces by scanning the ports it may occupy.
+ */
+ 
+static __init int qdi_init(void)
+{
+	unsigned long flags;
+	static const unsigned long qd_port[2] = { 0x30, 0xB0 };
+	static const unsigned long ide_port[2] = { 0x170, 0x1F0 };
+	static const int ide_irq[2] = { 14, 15 };
+	
+	int ct = 0;
+	int i;
+	
+	if (probe_qdi == 0)
+		return -ENODEV;
+	
+	/*
+ 	 *	Check each possible QD65xx base address
+	 */
+	
+	for (i = 0; i < 2; i++) {
+		unsigned long port = qd_port[i];
+		u8 r, res;
+		
+		
+		if (request_region(port, 2, "pata_qdi")) {
+			/* Check for a card */
+			local_irq_save(flags);
+			r = inb_p(port);
+			outb_p(0x19, port);
+			res = inb_p(port);
+			outb_p(r, port);
+			local_irq_restore(flags);
+		
+			/* Fail */
+			if (res == 0x19)
+			{
+				release_region(port, 2);
+				continue;
+			}
+			
+			/* Passes the presence test */
+			r = inb_p(port + 1);	/* Check port agrees with port set */
+			if ((r & 2) >> 1 != i) {
+				release_region(port, 2);
+				continue;
+			}
+
+			/* Check card type */			
+			if ((r & 0xF0) == 0xC0) {
+				/* QD6500: single channel */
+				if (r & 8) {
+					/* Disabled ? */
+					release_region(port, 2);
+					continue;
+				}
+				ct += qdi_init_one(port, 6500, ide_port[r & 0x01], ide_irq[r & 0x01], r & 0x04);
+			}
+			if (((r & 0xF0) == 0xA0) || (r & 0xF0) == 0x50) {
+				/* QD6580: dual channel */
+				if (!request_region(port + 2 , 2, "pata_qdi"))
+				{
+					release_region(port, 2);
+					continue;
+				}
+				res = inb(port + 3);
+				if (res & 1) {
+					/* Single channel mode */
+					ct += qdi_init_one(port, 6580, ide_port[r & 0x01], ide_irq[r & 0x01], r & 0x04);
+				} else {
+					/* Dual channel mode */
+					ct += qdi_init_one(port, 6580, 0x1F0, 14, r & 0x04);
+					ct += qdi_init_one(port + 2, 6580, 0x170, 15, r & 0x04);
+				}
+			}
+		}
+	}
+	if (ct != 0)
+		return 0;
+	return -ENODEV;
+}
+
+static __exit void qdi_exit(void)
+{
+	int i;
+
+	for (i = 0; i < nr_qdi_host; i++) {
+		ata_host_remove(qdi_host[i]);
+		/* Free the control resource. The 6580 dual channel has the resources
+		 * claimed as a pair of 2 byte resources so we need no special cases...
+		 */
+		release_region(qdi_data[i].timing, 2);
+		platform_device_unregister(qdi_data[i].platform_dev);
+	}	
+}
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for qdi ATA");
+MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
+
+module_init(qdi_init);
+module_exit(qdi_exit);
+
+module_param(probe_qdi, int, 0);
+
diff --git a/drivers/ata/pata_radisys.c b/drivers/ata/pata_radisys.c
new file mode 100644
index 000000000000..6f7d0527265f
--- /dev/null
+++ b/drivers/ata/pata_radisys.c
@@ -0,0 +1,335 @@
+/*
+ *    pata_radisys.c - Intel PATA/SATA controllers
+ *
+ *	(C) 2006 Red Hat <alan@redhat.com>
+ *
+ *    Some parts based on ata_piix.c by Jeff Garzik and others.
+ *
+ *    A PIIX relative, this device has a single ATA channel and no
+ *    slave timings, SITRE or PPE. In that sense it is a close relative
+ *    of the original PIIX. It does however support UDMA 33/66 per channel
+ *    although no other modes/timings. Also lacking is 32bit I/O on the ATA
+ *    port.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+#include <linux/ata.h>
+
+#define DRV_NAME	"pata_radisys"
+#define DRV_VERSION	"0.4.1"
+
+/**
+ *	radisys_probe_init		-	probe begin
+ *	@ap: ATA port
+ *
+ *	Set up cable type and use generic probe init
+ */
+ 
+static int radisys_pre_reset(struct ata_port *ap)
+{
+	ap->cbl = ATA_CBL_PATA80;
+	return ata_std_prereset(ap);
+}
+
+
+/**
+ *	radisys_pata_error_handler - Probe specified port on PATA host controller
+ *	@ap: Port to probe
+ *	@classes:
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void radisys_pata_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, radisys_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	radisys_set_piomode - Initialize host controller PATA PIO timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: um
+ *
+ *	Set PIO mode for device, in host controller PCI config space.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void radisys_set_piomode (struct ata_port *ap, struct ata_device *adev)
+{
+	unsigned int pio	= adev->pio_mode - XFER_PIO_0;
+	struct pci_dev *dev	= to_pci_dev(ap->host->dev);
+	u16 idetm_data;
+	int control = 0;
+
+	/*
+	 *	See Intel Document 298600-004 for the timing programing rules
+	 *	for PIIX/ICH. Note that the early PIIX does not have the slave
+	 *	timing port at 0x44. The Radisys is a relative of the PIIX
+	 *	but not the same so be careful.
+	 */
+
+	static const	 /* ISP  RTC */
+	u8 timings[][2]	= { { 0, 0 },	/* Check me */
+			    { 0, 0 },
+			    { 1, 1 },
+			    { 2, 2 },
+			    { 3, 3 }, };
+
+	if (pio > 0)
+		control |= 1;	/* TIME1 enable */
+	if (ata_pio_need_iordy(adev))
+		control |= 2;	/* IE IORDY */
+
+	pci_read_config_word(dev, 0x40, &idetm_data);
+
+	/* Enable IE and TIME as appropriate. Clear the other
+	   drive timing bits */
+	idetm_data &= 0xCCCC;
+	idetm_data |= (control << (4 * adev->devno));
+	idetm_data |= (timings[pio][0] << 12) |
+			(timings[pio][1] << 8);
+	pci_write_config_word(dev, 0x40, idetm_data);
+
+	/* Track which port is configured */
+	ap->private_data = adev;
+}
+
+/**
+ *	radisys_set_dmamode - Initialize host controller PATA DMA timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: Device to program
+ *	@isich: True if the device is an ICH and has IOCFG registers
+ *
+ *	Set MWDMA mode for device, in host controller PCI config space.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void radisys_set_dmamode (struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *dev	= to_pci_dev(ap->host->dev);
+	u16 idetm_data;
+	u8 udma_enable;
+	
+	static const	 /* ISP  RTC */
+	u8 timings[][2]	= { { 0, 0 },
+			    { 0, 0 },
+			    { 1, 1 },
+			    { 2, 2 },
+			    { 3, 3 }, };
+
+	/*
+	 * MWDMA is driven by the PIO timings. We must also enable
+	 * IORDY unconditionally.
+	 */
+
+	pci_read_config_word(dev, 0x40, &idetm_data);
+	pci_read_config_byte(dev, 0x48, &udma_enable);
+
+	if (adev->dma_mode < XFER_UDMA_0) {
+		unsigned int mwdma	= adev->dma_mode - XFER_MW_DMA_0;
+		const unsigned int needed_pio[3] = {
+			XFER_PIO_0, XFER_PIO_3, XFER_PIO_4
+		};
+		int pio = needed_pio[mwdma] - XFER_PIO_0;
+		int control = 3;	/* IORDY|TIME0 */
+
+		/* If the drive MWDMA is faster than it can do PIO then
+		   we must force PIO0 for PIO cycles. */
+
+		if (adev->pio_mode < needed_pio[mwdma])
+			control = 1;
+
+		/* Mask out the relevant control and timing bits we will load. Also
+		   clear the other drive TIME register as a precaution */
+		   
+		idetm_data &= 0xCCCC;
+		idetm_data |= control << (4 * adev->devno);
+		idetm_data |= (timings[pio][0] << 12) | (timings[pio][1] << 8);
+
+		udma_enable &= ~(1 << adev->devno);
+	} else {
+		u8 udma_mode;
+		
+		/* UDMA66 on: UDMA 33 and 66 are switchable via register 0x4A */
+		
+		pci_read_config_byte(dev, 0x4A, &udma_mode);
+		
+		if (adev->xfer_mode == XFER_UDMA_2)
+			udma_mode &= ~ (1 << adev->devno);
+		else /* UDMA 4 */
+			udma_mode |= (1 << adev->devno);
+			
+		pci_write_config_byte(dev, 0x4A, udma_mode);
+		
+		udma_enable |= (1 << adev->devno);
+	}
+	pci_write_config_word(dev, 0x40, idetm_data);
+	pci_write_config_byte(dev, 0x48, udma_enable);
+
+	/* Track which port is configured */
+	ap->private_data = adev;
+}
+
+/**
+ *	radisys_qc_issue_prot	-	command issue
+ *	@qc: command pending
+ *
+ *	Called when the libata layer is about to issue a command. We wrap
+ *	this interface so that we can load the correct ATA timings if
+ *	neccessary. Our logic also clears TIME0/TIME1 for the other device so
+ *	that, even if we get this wrong, cycles to the other device will
+ *	be made PIO0.
+ */
+
+static unsigned int radisys_qc_issue_prot(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct ata_device *adev = qc->dev;
+
+	if (adev != ap->private_data) {
+		/* UDMA timing is not shared */
+		if (adev->dma_mode < XFER_UDMA_0) {
+			if (adev->dma_mode)
+				radisys_set_dmamode(ap, adev);
+			else if (adev->pio_mode)
+				radisys_set_piomode(ap, adev);
+		}
+	}
+	return ata_qc_issue_prot(qc);
+}
+
+
+static struct scsi_host_template radisys_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static const struct ata_port_operations radisys_pata_ops = {
+	.port_disable		= ata_port_disable,
+	.set_piomode		= radisys_set_piomode,
+	.set_dmamode		= radisys_set_dmamode,
+	.mode_filter		= ata_pci_default_filter,
+
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.check_status		= ata_check_status,
+	.exec_command		= ata_exec_command,
+	.dev_select		= ata_std_dev_select,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= radisys_pata_error_handler,
+	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= radisys_qc_issue_prot,
+	.data_xfer		= ata_pio_data_xfer,
+
+	.eng_timeout		= ata_eng_timeout,
+
+	.irq_handler		= ata_interrupt,
+	.irq_clear		= ata_bmdma_irq_clear,
+
+	.port_start		= ata_port_start,
+	.port_stop		= ata_port_stop,
+	.host_stop		= ata_host_stop,
+};
+
+
+/**
+ *	radisys_init_one - Register PIIX ATA PCI device with kernel services
+ *	@pdev: PCI device to register
+ *	@ent: Entry in radisys_pci_tbl matching with @pdev
+ *
+ *	Called from kernel PCI layer.  We probe for combined mode (sigh),
+ *	and then hand over control to libata, for it to do the rest.
+ *
+ *	LOCKING:
+ *	Inherited from PCI layer (may sleep).
+ *
+ *	RETURNS:
+ *	Zero on success, or -ERRNO value.
+ */
+
+static int radisys_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+	static int printed_version;
+	static struct ata_port_info info = {
+		.sht		= &radisys_sht,
+		.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask	= 0x1f,	/* pio0-4 */
+		.mwdma_mask	= 0x07, /* mwdma1-2 */
+		.udma_mask	= 0x14, /* UDMA33/66 only */
+		.port_ops	= &radisys_pata_ops,
+	};
+	static struct ata_port_info *port_info[2] = { &info, &info };
+
+	if (!printed_version++)
+		dev_printk(KERN_DEBUG, &pdev->dev,
+			   "version " DRV_VERSION "\n");
+
+	return ata_pci_init_one(pdev, port_info, 2);
+}
+
+static const struct pci_device_id radisys_pci_tbl[] = {
+	{ 0x1331, 0x8201, PCI_ANY_ID, PCI_ANY_ID, },
+	{ }	/* terminate list */
+};
+
+static struct pci_driver radisys_pci_driver = {
+	.name			= DRV_NAME,
+	.id_table		= radisys_pci_tbl,
+	.probe			= radisys_init_one,
+	.remove			= ata_pci_remove_one,
+};
+
+static int __init radisys_init(void)
+{
+	return pci_register_driver(&radisys_pci_driver);
+}
+
+static void __exit radisys_exit(void)
+{
+	pci_unregister_driver(&radisys_pci_driver);
+}
+
+
+module_init(radisys_init);
+module_exit(radisys_exit);
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("SCSI low-level driver for Radisys R82600 controllers");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, radisys_pci_tbl);
+MODULE_VERSION(DRV_VERSION);
+
diff --git a/drivers/ata/pata_rz1000.c b/drivers/ata/pata_rz1000.c
new file mode 100644
index 000000000000..3c6d84fd4312
--- /dev/null
+++ b/drivers/ata/pata_rz1000.c
@@ -0,0 +1,205 @@
+/*
+ *  RZ1000/1001 driver based upon
+ *
+ *  linux/drivers/ide/pci/rz1000.c	Version 0.06	January 12, 2003
+ *  Copyright (C) 1995-1998  Linus Torvalds & author (see below)
+ *  Principal Author:  mlord@pobox.com (Mark Lord)
+ *
+ *  See linux/MAINTAINERS for address of current maintainer.
+ *
+ *  This file provides support for disabling the buggy read-ahead
+ *  mode of the RZ1000 IDE chipset, commonly used on Intel motherboards.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME	"pata_rz1000"
+#define DRV_VERSION	"0.2.2"
+
+
+/**
+ *	rz1000_prereset		-	probe begin
+ *	@ap: ATA port
+ *
+ *	Set up cable type and use generics
+ */
+
+static int rz1000_prereset(struct ata_port *ap)
+{
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	rz1000_error_handler		-	probe reset
+ *	@ap: ATA port
+ *
+ *	Perform the ATA standard reset sequence
+ */
+
+static void rz1000_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, rz1000_prereset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	rz1000_set_mode		-	mode setting function
+ *	@ap: ATA interface
+ *
+ *	Use a non standard set_mode function. We don't want to be tuned. We
+ *	would prefer to be BIOS generic but for the fact our hardware is
+ *	whacked out.
+ */
+
+static void rz1000_set_mode(struct ata_port *ap)
+{
+	int i;
+
+	for (i = 0; i < ATA_MAX_DEVICES; i++) {
+		struct ata_device *dev = &ap->device[i];
+		if (ata_dev_enabled(dev)) {
+			/* We don't really care */
+			dev->pio_mode = XFER_PIO_0;
+			dev->xfer_mode = XFER_PIO_0;
+			dev->xfer_shift = ATA_SHIFT_PIO;
+			dev->flags |= ATA_DFLAG_PIO;
+		}
+	}
+}
+
+
+static struct scsi_host_template rz1000_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations rz1000_port_ops = {
+	.set_mode	= rz1000_set_mode,
+
+	.port_disable	= ata_port_disable,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.error_handler	= rz1000_error_handler,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= rz1000_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+/**
+ *	rz1000_init_one - Register RZ1000 ATA PCI device with kernel services
+ *	@pdev: PCI device to register
+ *	@ent: Entry in rz1000_pci_tbl matching with @pdev
+ *
+ *	Configure an RZ1000 interface. This doesn't require much special
+ *	handling except that we *MUST* kill the chipset readahead or the
+ *	user may experience data corruption.
+ */
+
+static int rz1000_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+	static int printed_version;
+	struct ata_port_info *port_info[2];
+	u16 reg;
+	static struct ata_port_info info = {
+		.sht = &rz1000_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.port_ops = &rz1000_port_ops
+	};
+
+	if (!printed_version++)
+		printk(KERN_DEBUG DRV_NAME " version " DRV_VERSION "\n");
+
+	/* Be exceptionally paranoid as we must be sure to apply the fix */
+	if (pci_read_config_word(pdev, 0x40, &reg) != 0)
+		goto fail;
+	reg &= 0xDFFF;
+	if (pci_write_config_word(pdev, 0x40, reg) != 0)
+		goto fail;
+	printk(KERN_INFO DRV_NAME ": disabled chipset readahead.\n");
+
+	port_info[0] = &info;
+	port_info[1] = &info;
+	return ata_pci_init_one(pdev, port_info, 2);
+fail:
+	printk(KERN_ERR DRV_NAME ": failed to disable read-ahead on chipset..\n");
+	/* Not safe to use so skip */
+	return -ENODEV;
+}
+
+static struct pci_device_id pata_rz1000[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_PCTECH, PCI_DEVICE_ID_PCTECH_RZ1000), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_PCTECH, PCI_DEVICE_ID_PCTECH_RZ1001), },
+	{ 0, },
+};
+
+static struct pci_driver rz1000_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= pata_rz1000,
+	.probe 		= rz1000_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+
+static int __init rz1000_init(void)
+{
+	return pci_register_driver(&rz1000_pci_driver);
+}
+
+static void __exit rz1000_exit(void)
+{
+	pci_unregister_driver(&rz1000_pci_driver);
+}
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for RZ1000 PCI ATA");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, pata_rz1000);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(rz1000_init);
+module_exit(rz1000_exit);
+
diff --git a/drivers/ata/pata_sc1200.c b/drivers/ata/pata_sc1200.c
new file mode 100644
index 000000000000..9b42b9a52423
--- /dev/null
+++ b/drivers/ata/pata_sc1200.c
@@ -0,0 +1,287 @@
+/*
+ * New ATA layer SC1200 driver		Alan Cox <alan@redhat.com>
+ *
+ * TODO: Mode selection filtering
+ * TODO: Can't enable second channel until ATA core has serialize
+ * TODO: Needs custom DMA cleanup code
+ *
+ * Based very heavily on
+ *
+ * linux/drivers/ide/pci/sc1200.c		Version 0.91	28-Jan-2003
+ *
+ * Copyright (C) 2000-2002		Mark Lord <mlord@pobox.com>
+ * May be copied or modified under the terms of the GNU General Public License
+ *
+ * Development of this chipset driver was funded
+ * by the nice folks at National Semiconductor.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ * 
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME	"sc1200"
+#define DRV_VERSION	"0.2.3"
+
+#define SC1200_REV_A	0x00
+#define SC1200_REV_B1	0x01
+#define SC1200_REV_B3	0x02
+#define SC1200_REV_C1	0x03
+#define SC1200_REV_D1	0x04
+
+/**
+ *	sc1200_clock	-	PCI clock
+ *
+ *	Return the PCI bus clocking for the SC1200 chipset configuration
+ *	in use. We return 0 for 33MHz 1 for 48MHz and 2 for 66Mhz
+ */
+ 
+static int sc1200_clock(void)
+{
+	/* Magic registers that give us the chipset data */
+	u8 chip_id = inb(0x903C);
+	u8 silicon_rev = inb(0x903D);
+	u16 pci_clock;
+	
+	if (chip_id == 0x04 && silicon_rev < SC1200_REV_B1)
+		return 0;	/* 33 MHz mode */
+
+	/* Clock generator configuration 0x901E its 8/9 are the PCI clocking
+	   0/3 is 33Mhz 1 is 48 2 is 66 */
+
+	pci_clock = inw(0x901E);
+	pci_clock >>= 8;
+	pci_clock &= 0x03;
+	if (pci_clock == 3)
+		pci_clock = 0;
+	return pci_clock;
+}
+
+/**
+ *	sc1200_set_piomode		-	PIO setup
+ *	@ap: ATA interface
+ *	@adev: device on the interface
+ *
+ *	Set our PIO requirements. This is fairly simple on the SC1200
+ */
+ 
+static void sc1200_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	static const u32 pio_timings[4][5] = {
+		{0x00009172, 0x00012171, 0x00020080, 0x00032010, 0x00040010},	// format0  33Mhz
+		{0xd1329172, 0x71212171, 0x30200080, 0x20102010, 0x00100010},	// format1, 33Mhz
+		{0xfaa3f4f3, 0xc23232b2, 0x513101c1, 0x31213121, 0x10211021},	// format1, 48Mhz
+		{0xfff4fff4, 0xf35353d3, 0x814102f1, 0x42314231, 0x11311131}	// format1, 66Mhz
+	};
+
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u32 format;
+	unsigned int reg = 0x40 + 0x10 * ap->port_no;
+	int mode = adev->pio_mode - XFER_PIO_0;
+	
+	pci_read_config_dword(pdev, reg + 4, &format);
+	format >>= 31;
+	format += sc1200_clock();
+	pci_write_config_dword(pdev, reg + 8 * adev->devno, 
+				pio_timings[format][mode]);
+}
+
+/**
+ *	sc1200_set_dmamode		-	DMA timing setup
+ *	@ap: ATA interface
+ *	@adev: Device being configured
+ *
+ *	We cannot mix MWDMA and UDMA without reloading timings each switch
+ *	master to slave.
+ */
+ 
+static void sc1200_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	static const u32 udma_timing[3][3] = {
+		{ 0x00921250, 0x00911140, 0x00911030 },
+		{ 0x00932470, 0x00922260, 0x00922140 },
+		{ 0x009436A1, 0x00933481, 0x00923261 }
+	};
+	
+	static const u32 mwdma_timing[3][3] = {
+		{ 0x00077771, 0x00012121, 0x00002020 },
+		{ 0x000BBBB2, 0x00024241, 0x00013131 },
+		{ 0x000FFFF3, 0x00035352, 0x00015151 }
+	};
+	
+	int clock = sc1200_clock();
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	unsigned int reg = 0x40 + 0x10 * ap->port_no;
+	int mode = adev->dma_mode;
+	u32 format;
+
+	if (mode >= XFER_UDMA_0)
+		format = udma_timing[clock][mode - XFER_UDMA_0];
+	else
+		format = mwdma_timing[clock][mode - XFER_MW_DMA_0];
+	
+	if (adev->devno == 0) {
+		u32 timings;
+		
+		pci_read_config_dword(pdev, reg + 4, &timings);
+		timings &= 0x80000000UL;
+		timings |= format;
+		pci_write_config_dword(pdev, reg + 4, timings);
+	} else
+		pci_write_config_dword(pdev, reg + 12, format);
+}
+
+/**
+ *	sc1200_qc_issue_prot	-	command issue
+ *	@qc: command pending
+ *
+ *	Called when the libata layer is about to issue a command. We wrap
+ *	this interface so that we can load the correct ATA timings if
+ *	neccessary.  Specifically we have a problem that there is only
+ *	one MWDMA/UDMA bit.
+ */
+
+static unsigned int sc1200_qc_issue_prot(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct ata_device *adev = qc->dev;
+	struct ata_device *prev = ap->private_data;
+
+	/* See if the DMA settings could be wrong */
+	if (adev->dma_mode != 0 && adev != prev && prev != NULL) {
+		/* Maybe, but do the channels match MWDMA/UDMA ? */
+		if ((adev->dma_mode >= XFER_UDMA_0 && prev->dma_mode < XFER_UDMA_0) ||
+		    (adev->dma_mode < XFER_UDMA_0 && prev->dma_mode >= XFER_UDMA_0))
+		    	/* Switch the mode bits */
+		    	sc1200_set_dmamode(ap, adev);
+	}
+
+	return ata_qc_issue_prot(qc);
+}
+
+static struct scsi_host_template sc1200_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations sc1200_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= sc1200_set_piomode,
+	.set_dmamode	= sc1200_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.error_handler	= ata_bmdma_error_handler,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= sc1200_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+/**
+ *	sc1200_init_one		-	Initialise an SC1200
+ *	@dev: PCI device
+ *	@id: Entry in match table
+ *
+ *	Just throw the needed data at the libata helper and it does all
+ *	our work.
+ */
+ 
+static int sc1200_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info = {
+		.sht = &sc1200_sht,
+		.flags = ATA_FLAG_SLAVE_POSS|ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x07,
+		.port_ops = &sc1200_port_ops
+	};
+	static struct ata_port_info *port_info[2] = { &info, &info };
+	
+	/* Can't enable port 2 yet, see top comments */
+	return ata_pci_init_one(dev, port_info, 1);
+}
+
+static struct pci_device_id sc1200[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_NS, PCI_DEVICE_ID_NS_SCx200_IDE), },
+	{ 0, },
+};
+
+static struct pci_driver sc1200_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= sc1200,
+	.probe 		= sc1200_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init sc1200_init(void)
+{
+	return pci_register_driver(&sc1200_pci_driver);
+}
+
+
+static void __exit sc1200_exit(void)
+{
+	pci_unregister_driver(&sc1200_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox, Mark Lord");
+MODULE_DESCRIPTION("low-level driver for the NS/AMD SC1200");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, sc1200);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(sc1200_init);
+module_exit(sc1200_exit);
diff --git a/drivers/ata/pata_serverworks.c b/drivers/ata/pata_serverworks.c
new file mode 100644
index 000000000000..3ec30036c978
--- /dev/null
+++ b/drivers/ata/pata_serverworks.c
@@ -0,0 +1,587 @@
+/*
+ * ata-serverworks.c 	- Serverworks PATA for new ATA layer
+ *			  (C) 2005 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ * based upon
+ *
+ * serverworks.c
+ * 
+ * Copyright (C) 1998-2000 Michel Aubry
+ * Copyright (C) 1998-2000 Andrzej Krzysztofowicz
+ * Copyright (C) 1998-2000 Andre Hedrick <andre@linux-ide.org>
+ * Portions copyright (c) 2001 Sun Microsystems
+ *
+ *
+ * RCC/ServerWorks IDE driver for Linux
+ *
+ *   OSB4: `Open South Bridge' IDE Interface (fn 1)
+ *         supports UDMA mode 2 (33 MB/s)
+ *
+ *   CSB5: `Champion South Bridge' IDE Interface (fn 1)
+ *         all revisions support UDMA mode 4 (66 MB/s)
+ *         revision A2.0 and up support UDMA mode 5 (100 MB/s)
+ *
+ *         *** The CSB5 does not provide ANY register ***
+ *         *** to detect 80-conductor cable presence. ***
+ *
+ *   CSB6: `Champion South Bridge' IDE Interface (optional: third channel)
+ *
+ * Documentation:
+ *	Available under NDA only. Errata info very hard to get.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "pata_serverworks"
+#define DRV_VERSION "0.3.6"
+
+#define SVWKS_CSB5_REVISION_NEW	0x92 /* min PCI_REVISION_ID for UDMA5 (A2.0) */
+#define SVWKS_CSB6_REVISION	0xa0 /* min PCI_REVISION_ID for UDMA4 (A1.0) */
+
+/* Seagate Barracuda ATA IV Family drives in UDMA mode 5
+ * can overrun their FIFOs when used with the CSB5 */
+
+static const char *csb_bad_ata100[] = {
+	"ST320011A",
+	"ST340016A",
+	"ST360021A",
+	"ST380021A",
+	NULL
+};
+
+/**
+ *	dell_cable	-	Dell serverworks cable detection
+ *	@ap: ATA port to do cable detect
+ *
+ *	Dell hide the 40/80 pin select for their interfaces in the top two
+ *	bits of the subsystem ID. 
+ */
+ 
+static int dell_cable(struct ata_port *ap) {
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	
+	if (pdev->subsystem_device & (1 << (ap->port_no + 14)))
+		return ATA_CBL_PATA80;
+	return ATA_CBL_PATA40;
+}
+
+/**
+ *	sun_cable	-	Sun Cobalt 'Alpine' cable detection
+ *	@ap: ATA port to do cable select
+ *
+ *	Cobalt CSB5 IDE hides the 40/80pin in the top two bits of the
+ *	subsystem ID the same as dell. We could use one function but we may
+ *	need to extend the Dell one in future
+ */
+ 
+static int sun_cable(struct ata_port *ap) {
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	
+	if (pdev->subsystem_device & (1 << (ap->port_no + 14)))
+		return ATA_CBL_PATA80;
+	return ATA_CBL_PATA40;
+}
+
+/**
+ *	osb4_cable	-	OSB4 cable detect
+ *	@ap: ATA port to check
+ *
+ *	The OSB4 isn't UDMA66 capable so this is easy
+ */
+
+static int osb4_cable(struct ata_port *ap) {
+	return ATA_CBL_PATA40;
+}
+
+/**
+ *	csb4_cable	-	CSB5/6 cable detect
+ *	@ap: ATA port to check
+ *
+ *	Serverworks default arrangement is to use the drive side detection
+ *	only.
+ */
+
+static int csb_cable(struct ata_port *ap) {
+	return ATA_CBL_PATA80;
+}
+
+struct sv_cable_table {
+	int device;
+	int subvendor;
+	int (*cable_detect)(struct ata_port *ap);
+};
+
+/*
+ *	Note that we don't copy the old serverworks code because the old
+ *	code contains obvious mistakes
+ */
+ 
+static struct sv_cable_table cable_detect[] = {
+	{ PCI_DEVICE_ID_SERVERWORKS_CSB5IDE, PCI_VENDOR_ID_DELL, dell_cable },
+	{ PCI_DEVICE_ID_SERVERWORKS_CSB6IDE, PCI_VENDOR_ID_DELL, dell_cable },
+	{ PCI_DEVICE_ID_SERVERWORKS_CSB5IDE, PCI_VENDOR_ID_SUN,  sun_cable },
+	{ PCI_DEVICE_ID_SERVERWORKS_OSB4, PCI_ANY_ID, osb4_cable },
+	{ PCI_DEVICE_ID_SERVERWORKS_CSB5IDE, PCI_ANY_ID, csb_cable },
+	{ PCI_DEVICE_ID_SERVERWORKS_CSB6IDE, PCI_ANY_ID, csb_cable },
+	{ PCI_DEVICE_ID_SERVERWORKS_CSB6IDE2, PCI_ANY_ID, csb_cable },
+	{ PCI_DEVICE_ID_SERVERWORKS_HT1000IDE, PCI_ANY_ID, csb_cable },
+	{ }
+};
+
+/**
+ *	serverworks_pre_reset		-	cable detection
+ *	@ap: ATA port
+ *
+ *	Perform cable detection according to the device and subvendor 
+ *	identifications
+ */
+ 
+static int serverworks_pre_reset(struct ata_port *ap) {
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	struct sv_cable_table *cb = cable_detect;
+
+	while(cb->device) {
+		if (cb->device == pdev->device && 
+		    (cb->subvendor == pdev->subsystem_vendor ||
+		      cb->subvendor == PCI_ANY_ID)) {
+			ap->cbl = cb->cable_detect(ap);
+			return ata_std_prereset(ap);
+		}
+		cb++;
+	}
+
+	BUG();
+	return -1;	/* kill compiler warning */
+}
+
+static void serverworks_error_handler(struct ata_port *ap)
+{
+	return ata_bmdma_drive_eh(ap, serverworks_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	serverworks_is_csb	-	Check for CSB or OSB
+ *	@pdev: PCI device to check
+ *
+ *	Returns true if the device being checked is known to be a CSB
+ *	series device.
+ */
+ 
+static u8 serverworks_is_csb(struct pci_dev *pdev)
+{
+	switch (pdev->device) {
+		case PCI_DEVICE_ID_SERVERWORKS_CSB5IDE:
+		case PCI_DEVICE_ID_SERVERWORKS_CSB6IDE:
+		case PCI_DEVICE_ID_SERVERWORKS_CSB6IDE2:
+		case PCI_DEVICE_ID_SERVERWORKS_HT1000IDE:
+			return 1;
+		default:
+			break;
+	}
+	return 0;
+}
+
+/**
+ *	serverworks_osb4_filter	-	mode selection filter
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Filter the offered modes for the device to apply controller
+ *	specific rules. OSB4 requires no UDMA for disks due to a FIFO
+ *	bug we hit.
+ */
+ 
+static unsigned long serverworks_osb4_filter(const struct ata_port *ap, struct ata_device *adev, unsigned long mask)
+{
+	if (adev->class == ATA_DEV_ATA)
+		mask &= ~ATA_MASK_UDMA;
+	return ata_pci_default_filter(ap, adev, mask);
+}
+
+
+/**
+ *	serverworks_csb_filter	-	mode selection filter
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Check the blacklist and disable UDMA5 if matched
+ */
+
+static unsigned long serverworks_csb_filter(const struct ata_port *ap, struct ata_device *adev, unsigned long mask)
+{
+	const char *p;
+	char model_num[40];
+	int len, i;
+
+	/* Disk, UDMA */	
+	if (adev->class != ATA_DEV_ATA)
+		return ata_pci_default_filter(ap, adev, mask);
+
+	/* Actually do need to check */
+	ata_id_string(adev->id, model_num, ATA_ID_PROD_OFS, sizeof(model_num));
+	/* Precuationary - why not do this in the libata core ?? */
+	
+	len = strlen(model_num);
+	while ((len > 0) && (model_num[len - 1] == ' ')) {
+		len--;
+		model_num[len] = 0;
+	}
+
+	for(i = 0; (p = csb_bad_ata100[i]) != NULL; i++) {
+		if (!strncmp(p, model_num, len))
+			mask &= ~(0x1F << ATA_SHIFT_UDMA);
+	}
+	return ata_pci_default_filter(ap, adev, mask);
+}
+
+
+/**
+ *	serverworks_set_piomode	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Program the OSB4/CSB5 timing registers for PIO. The PIO register
+ *	load is done as a simple lookup.
+ */
+static void serverworks_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	static const u8 pio_mode[] = { 0x5d, 0x47, 0x34, 0x22, 0x20 };
+	int offset = 1 + (2 * ap->port_no) - adev->devno;
+	int devbits = (2 * ap->port_no + adev->devno) * 4;
+	u16 csb5_pio;
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	int pio = adev->pio_mode - XFER_PIO_0;
+
+	pci_write_config_byte(pdev, 0x40 + offset, pio_mode[pio]);
+	
+	/* The OSB4 just requires the timing but the CSB series want the
+	   mode number as well */
+	if (serverworks_is_csb(pdev)) {
+		pci_read_config_word(pdev, 0x4A, &csb5_pio);
+		csb5_pio &= ~(0x0F << devbits);
+		pci_write_config_byte(pdev, 0x4A, csb5_pio | (pio << devbits));
+	}
+}
+
+/**
+ *	serverworks_set_dmamode	-	set initial DMA mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Program the MWDMA/UDMA modes for the serverworks OSB4/CSB5
+ *	chipset. The MWDMA mode values are pulled from a lookup table
+ *	while the chipset uses mode number for UDMA.
+ */
+
+static void serverworks_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	static const u8 dma_mode[] = { 0x77, 0x21, 0x20 };
+	int offset = 1 + 2 * ap->port_no - adev->devno;
+	int devbits = (2 * ap->port_no + adev->devno);
+	u8 ultra;
+	u8 ultra_cfg;
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+
+	pci_read_config_byte(pdev, 0x54, &ultra_cfg);
+
+	if (adev->dma_mode >= XFER_UDMA_0) {
+		pci_write_config_byte(pdev, 0x44 + offset,  0x20);
+
+		pci_read_config_byte(pdev, 0x56 + ap->port_no, &ultra);
+		ultra &= ~(0x0F << (ap->port_no * 4));
+		ultra |= (adev->dma_mode - XFER_UDMA_0)
+					<< (ap->port_no * 4);
+		pci_write_config_byte(pdev, 0x56 + ap->port_no, ultra);
+
+		ultra_cfg |=  (1 << devbits);
+	} else {
+		pci_write_config_byte(pdev, 0x44 + offset, 
+			dma_mode[adev->dma_mode - XFER_MW_DMA_0]);
+		ultra_cfg &= ~(1 << devbits);
+	}
+	pci_write_config_byte(pdev, 0x54, ultra_cfg);
+}
+
+static struct scsi_host_template serverworks_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations serverworks_osb4_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= serverworks_set_piomode,
+	.set_dmamode	= serverworks_set_dmamode,
+	.mode_filter	= serverworks_osb4_filter,
+	
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= serverworks_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+	
+	.irq_handler	= ata_interrupt,
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+static struct ata_port_operations serverworks_csb_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= serverworks_set_piomode,
+	.set_dmamode	= serverworks_set_dmamode,
+	.mode_filter	= serverworks_csb_filter,
+	
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= serverworks_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+	
+	.irq_handler	= ata_interrupt,
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+static int serverworks_fixup_osb4(struct pci_dev *pdev)
+{
+	u32 reg;
+	struct pci_dev *isa_dev = pci_get_device(PCI_VENDOR_ID_SERVERWORKS,
+		  PCI_DEVICE_ID_SERVERWORKS_OSB4, NULL);
+	if (isa_dev) {
+		pci_read_config_dword(isa_dev, 0x64, &reg);
+		reg &= ~0x00002000; /* disable 600ns interrupt mask */
+		if (!(reg & 0x00004000))
+			printk(KERN_DEBUG DRV_NAME ": UDMA not BIOS enabled.\n");
+		reg |=  0x00004000; /* enable UDMA/33 support */
+		pci_write_config_dword(isa_dev, 0x64, reg);
+		pci_dev_put(isa_dev);
+		return 0;
+	}
+	printk(KERN_WARNING "ata_serverworks: Unable to find bridge.\n");
+	return -ENODEV;
+}
+
+static int serverworks_fixup_csb(struct pci_dev *pdev)
+{
+	u8 rev;
+	u8 btr;
+	
+	pci_read_config_byte(pdev, PCI_REVISION_ID, &rev);
+
+	/* Third Channel Test */
+	if (!(PCI_FUNC(pdev->devfn) & 1)) {
+		struct pci_dev * findev = NULL;
+		u32 reg4c = 0;
+		findev = pci_get_device(PCI_VENDOR_ID_SERVERWORKS,
+			PCI_DEVICE_ID_SERVERWORKS_CSB5, NULL);
+		if (findev) {
+			pci_read_config_dword(findev, 0x4C, &reg4c);
+			reg4c &= ~0x000007FF;
+			reg4c |=  0x00000040;
+			reg4c |=  0x00000020;
+			pci_write_config_dword(findev, 0x4C, reg4c);
+			pci_dev_put(findev);
+		}
+	} else {
+		struct pci_dev * findev = NULL;
+		u8 reg41 = 0;
+
+		findev = pci_get_device(PCI_VENDOR_ID_SERVERWORKS,
+				PCI_DEVICE_ID_SERVERWORKS_CSB6, NULL);
+		if (findev) {
+			pci_read_config_byte(findev, 0x41, &reg41);
+			reg41 &= ~0x40;
+			pci_write_config_byte(findev, 0x41, reg41);
+			pci_dev_put(findev);
+		}
+	}
+	/* setup the UDMA Control register
+	 *
+	 * 1. clear bit 6 to enable DMA
+	 * 2. enable DMA modes with bits 0-1
+	 * 	00 : legacy
+	 * 	01 : udma2
+	 * 	10 : udma2/udma4
+	 * 	11 : udma2/udma4/udma5
+	 */
+	pci_read_config_byte(pdev, 0x5A, &btr);
+	btr &= ~0x40;
+	if (!(PCI_FUNC(pdev->devfn) & 1))
+		btr |= 0x2;
+	else
+		btr |= (rev >= SVWKS_CSB5_REVISION_NEW) ? 0x3 : 0x2;
+	pci_write_config_byte(pdev, 0x5A, btr);
+	
+	return btr;
+}
+
+static void serverworks_fixup_ht1000(struct pci_dev *pdev)
+{
+	u8 btr;
+	/* Setup HT1000 SouthBridge Controller - Single Channel Only */
+	pci_read_config_byte(pdev, 0x5A, &btr);
+	btr &= ~0x40;
+	btr |= 0x3;
+	pci_write_config_byte(pdev, 0x5A, btr);
+}
+
+
+static int serverworks_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+	int ports = 2;
+	static struct ata_port_info info[4] = {
+		{ /* OSB4 */
+			.sht = &serverworks_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = 0x07,
+			.port_ops = &serverworks_osb4_port_ops
+		}, { /* OSB4 no UDMA */
+			.sht = &serverworks_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = 0x00,
+			.port_ops = &serverworks_osb4_port_ops
+		}, { /* CSB5 */
+			.sht = &serverworks_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = 0x1f,
+			.port_ops = &serverworks_csb_port_ops
+		}, { /* CSB5 - later revisions*/
+			.sht = &serverworks_sht,
+			.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+			.pio_mask = 0x1f,
+			.mwdma_mask = 0x07,
+			.udma_mask = 0x3f,
+			.port_ops = &serverworks_csb_port_ops
+		}
+	};
+	static struct ata_port_info *port_info[2];
+	struct ata_port_info *devinfo = &info[id->driver_data];
+	
+	/* Force master latency timer to 64 PCI clocks */
+	pci_write_config_byte(pdev, PCI_LATENCY_TIMER, 0x40);
+
+	/* OSB4 : South Bridge and IDE */
+	if (pdev->device == PCI_DEVICE_ID_SERVERWORKS_OSB4IDE) {
+		/* Select non UDMA capable OSB4 if we can't do fixups */
+		if ( serverworks_fixup_osb4(pdev) < 0)
+			devinfo = &info[1];
+	}
+	/* setup CSB5/CSB6 : South Bridge and IDE option RAID */
+	else if ((pdev->device == PCI_DEVICE_ID_SERVERWORKS_CSB5IDE) ||
+		 (pdev->device == PCI_DEVICE_ID_SERVERWORKS_CSB6IDE) ||
+		 (pdev->device == PCI_DEVICE_ID_SERVERWORKS_CSB6IDE2)) {
+		 
+		 /* If the returned btr is the newer revision then
+		    select the right info block */
+		 if (serverworks_fixup_csb(pdev) == 3)
+		 	devinfo = &info[3];
+		 	
+		/* Is this the 3rd channel CSB6 IDE ? */
+		if (pdev->device == PCI_DEVICE_ID_SERVERWORKS_CSB6IDE2)
+			ports = 1;
+	}
+	/* setup HT1000E */
+	else if (pdev->device == PCI_DEVICE_ID_SERVERWORKS_HT1000IDE)
+		serverworks_fixup_ht1000(pdev);
+		
+	if (pdev->device == PCI_DEVICE_ID_SERVERWORKS_CSB5IDE)
+		ata_pci_clear_simplex(pdev);
+		
+	port_info[0] = port_info[1] = devinfo;
+	return ata_pci_init_one(pdev, port_info, ports);
+}
+
+static struct pci_device_id serverworks[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_SERVERWORKS, PCI_DEVICE_ID_SERVERWORKS_OSB4IDE), 0},
+	{ PCI_DEVICE(PCI_VENDOR_ID_SERVERWORKS, PCI_DEVICE_ID_SERVERWORKS_CSB5IDE), 2},
+	{ PCI_DEVICE(PCI_VENDOR_ID_SERVERWORKS, PCI_DEVICE_ID_SERVERWORKS_CSB6IDE), 2},
+	{ PCI_DEVICE(PCI_VENDOR_ID_SERVERWORKS, PCI_DEVICE_ID_SERVERWORKS_CSB6IDE2), 2},
+	{ PCI_DEVICE(PCI_VENDOR_ID_SERVERWORKS, PCI_DEVICE_ID_SERVERWORKS_HT1000IDE), 2},
+	{ 0, },
+};
+
+static struct pci_driver serverworks_pci_driver = {
+	.name 		= DRV_NAME,
+	.id_table	= serverworks,
+	.probe 		= serverworks_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init serverworks_init(void)
+{
+	return pci_register_driver(&serverworks_pci_driver);
+}
+
+
+static void __exit serverworks_exit(void)
+{
+	pci_unregister_driver(&serverworks_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for Serverworks OSB4/CSB5/CSB6");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, serverworks);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(serverworks_init);
+module_exit(serverworks_exit);
diff --git a/drivers/ata/pata_sil680.c b/drivers/ata/pata_sil680.c
new file mode 100644
index 000000000000..8f7db9638d0a
--- /dev/null
+++ b/drivers/ata/pata_sil680.c
@@ -0,0 +1,381 @@
+/*
+ * pata_sil680.c 	- SIL680 PATA for new ATA layer
+ *			  (C) 2005 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ * based upon
+ *
+ * linux/drivers/ide/pci/siimage.c		Version 1.07	Nov 30, 2003
+ *
+ * Copyright (C) 2001-2002	Andre Hedrick <andre@linux-ide.org>
+ * Copyright (C) 2003		Red Hat <alan@redhat.com>
+ *
+ *  May be copied or modified under the terms of the GNU General Public License
+ *
+ *  Documentation publically available.
+ *
+ *	If you have strange problems with nVidia chipset systems please
+ *	see the SI support documentation and update your system BIOS
+ *	if neccessary
+ *
+ * TODO
+ *	If we know all our devices are LBA28 (or LBA28 sized)  we could use
+ *	the command fifo mode.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "pata_sil680"
+#define DRV_VERSION "0.3.2"
+
+/**
+ *	sil680_selreg		-	return register base
+ *	@hwif: interface
+ *	@r: config offset
+ *
+ *	Turn a config register offset into the right address in either
+ *	PCI space or MMIO space to access the control register in question
+ *	Thankfully this is a configuration operation so isnt performance
+ *	criticial.
+ */
+
+static unsigned long sil680_selreg(struct ata_port *ap, int r)
+{
+	unsigned long base = 0xA0 + r;
+	base += (ap->port_no << 4);
+	return base;
+}
+
+/**
+ *	sil680_seldev		-	return register base
+ *	@hwif: interface
+ *	@r: config offset
+ *
+ *	Turn a config register offset into the right address in either
+ *	PCI space or MMIO space to access the control register in question
+ *	including accounting for the unit shift.
+ */
+
+static unsigned long sil680_seldev(struct ata_port *ap, struct ata_device *adev, int r)
+{
+	unsigned long base = 0xA0 + r;
+	base += (ap->port_no << 4);
+	base |= adev->devno ? 2 : 0;
+	return base;
+}
+
+
+/**
+ *	sil680_cable_detect	-	cable detection
+ *	@ap: ATA port
+ *
+ *	Perform cable detection. The SIL680 stores this in PCI config
+ *	space for us.
+ */
+
+static int sil680_cable_detect(struct ata_port *ap) {
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	unsigned long addr = sil680_selreg(ap, 0);
+	u8 ata66;
+	pci_read_config_byte(pdev, addr, &ata66);
+	if (ata66 & 1)
+		return ATA_CBL_PATA80;
+	else
+		return ATA_CBL_PATA40;
+}
+
+static int sil680_pre_reset(struct ata_port *ap)
+{
+	ap->cbl = sil680_cable_detect(ap);
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	sil680_bus_reset	-	reset the SIL680 bus
+ *	@ap: ATA port to reset
+ *
+ *	Perform the SIL680 housekeeping when doing an ATA bus reset
+ */
+
+static int sil680_bus_reset(struct ata_port *ap,unsigned int *classes)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	unsigned long addr = sil680_selreg(ap, 0);
+	u8 reset;
+
+	pci_read_config_byte(pdev, addr, &reset);
+	pci_write_config_byte(pdev, addr, reset | 0x03);
+	udelay(25);
+	pci_write_config_byte(pdev, addr, reset);
+	return ata_std_softreset(ap, classes);
+}
+
+static void sil680_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, sil680_pre_reset, sil680_bus_reset, NULL, ata_std_postreset);
+}
+
+/**
+ *	sil680_set_piomode	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Program the SIL680 registers for PIO mode. Note that the task speed
+ *	registers are shared between the devices so we must pick the lowest
+ *	mode for command work.
+ */
+
+static void sil680_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	static u16 speed_p[5] = { 0x328A, 0x2283, 0x1104, 0x10C3, 0x10C1 };
+	static u16 speed_t[5] = { 0x328A, 0x1281, 0x1281, 0x10C3, 0x10C1 };
+
+	unsigned long tfaddr = sil680_selreg(ap, 0x02);
+	unsigned long addr = sil680_seldev(ap, adev, 0x04);
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	int pio = adev->pio_mode - XFER_PIO_0;
+	int lowest_pio = pio;
+	u16 reg;
+
+	struct ata_device *pair = ata_dev_pair(adev);
+
+	if (pair != NULL && adev->pio_mode > pair->pio_mode)
+		lowest_pio = pair->pio_mode - XFER_PIO_0;
+
+	pci_write_config_word(pdev, addr, speed_p[pio]);
+	pci_write_config_word(pdev, tfaddr, speed_t[lowest_pio]);
+
+	pci_read_config_word(pdev, tfaddr-2, &reg);
+	reg &= ~0x0200;			/* Clear IORDY */
+	if (ata_pio_need_iordy(adev))
+		reg |= 0x0200;		/* Enable IORDY */
+	pci_write_config_word(pdev, tfaddr-2, reg);
+}
+
+/**
+ *	sil680_set_dmamode	-	set initial DMA mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Program the MWDMA/UDMA modes for the sil680 k
+ *	chipset. The MWDMA mode values are pulled from a lookup table
+ *	while the chipset uses mode number for UDMA.
+ */
+
+static void sil680_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	static u8 ultra_table[2][7] = {
+		{ 0x0C, 0x07, 0x05, 0x04, 0x02, 0x01, 0xFF },	/* 100MHz */
+		{ 0x0F, 0x0B, 0x07, 0x05, 0x03, 0x02, 0x01 },	/* 133Mhz */
+	};
+	static u16 dma_table[3] = { 0x2208, 0x10C2, 0x10C1 };
+
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	unsigned long ma = sil680_seldev(ap, adev, 0x08);
+	unsigned long ua = sil680_seldev(ap, adev, 0x0C);
+	unsigned long addr_mask = 0x80 + 4 * ap->port_no;
+	int port_shift = adev->devno * 4;
+	u8 scsc, mode;
+	u16 multi, ultra;
+
+	pci_read_config_byte(pdev, 0x8A, &scsc);
+	pci_read_config_byte(pdev, addr_mask, &mode);
+	pci_read_config_word(pdev, ma, &multi);
+	pci_read_config_word(pdev, ua, &ultra);
+
+	/* Mask timing bits */
+	ultra &= ~0x3F;
+	mode &= ~(0x03 << port_shift);
+
+	/* Extract scsc */
+	scsc = (scsc & 0x30) ? 1: 0;
+
+	if (adev->dma_mode >= XFER_UDMA_0) {
+		multi = 0x10C1;
+		ultra |= ultra_table[scsc][adev->dma_mode - XFER_UDMA_0];
+		mode |= (0x03 << port_shift);
+	} else {
+		multi = dma_table[adev->dma_mode - XFER_MW_DMA_0];
+		mode |= (0x02 << port_shift);
+	}
+	pci_write_config_byte(pdev, addr_mask, mode);
+	pci_write_config_word(pdev, ma, multi);
+	pci_write_config_word(pdev, ua, ultra);
+}
+
+static struct scsi_host_template sil680_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations sil680_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= sil680_set_piomode,
+	.set_dmamode	= sil680_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= sil680_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static int sil680_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info = {
+		.sht = &sil680_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x7f,
+		.port_ops = &sil680_port_ops
+	};
+	static struct ata_port_info info_slow = {
+		.sht = &sil680_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x3f,
+		.port_ops = &sil680_port_ops
+	};
+	static struct ata_port_info *port_info[2] = {&info, &info};
+	static int printed_version;
+	u32 class_rev	= 0;
+	u8 tmpbyte	= 0;
+
+	if (!printed_version++)
+		dev_printk(KERN_DEBUG, &pdev->dev, "version " DRV_VERSION "\n");
+
+        pci_read_config_dword(pdev, PCI_CLASS_REVISION, &class_rev);
+        class_rev &= 0xff;
+        /* FIXME: double check */
+	pci_write_config_byte(pdev, PCI_CACHE_LINE_SIZE, (class_rev) ? 1 : 255);
+
+	pci_write_config_byte(pdev, 0x80, 0x00);
+	pci_write_config_byte(pdev, 0x84, 0x00);
+
+	pci_read_config_byte(pdev, 0x8A, &tmpbyte);
+
+	printk(KERN_INFO "sil680: BA5_EN = %d clock = %02X\n",
+			tmpbyte & 1, tmpbyte & 0x30);
+
+	switch(tmpbyte & 0x30) {
+		case 0x00:
+			/* 133 clock attempt to force it on */
+			pci_write_config_byte(pdev, 0x8A, tmpbyte|0x10);
+			break;
+		case 0x30:
+			/* if clocking is disabled */
+			/* 133 clock attempt to force it on */
+			pci_write_config_byte(pdev, 0x8A, tmpbyte & ~0x20);
+			break;
+		case 0x10:
+			/* 133 already */
+			break;
+		case 0x20:
+			/* BIOS set PCI x2 clocking */
+			break;
+	}
+
+	pci_read_config_byte(pdev,   0x8A, &tmpbyte);
+	printk(KERN_INFO "sil680: BA5_EN = %d clock = %02X\n",
+			tmpbyte & 1, tmpbyte & 0x30);
+	if ((tmpbyte & 0x30) == 0)
+		port_info[0] = port_info[1] = &info_slow;
+
+	pci_write_config_byte(pdev,  0xA1, 0x72);
+	pci_write_config_word(pdev,  0xA2, 0x328A);
+	pci_write_config_dword(pdev, 0xA4, 0x62DD62DD);
+	pci_write_config_dword(pdev, 0xA8, 0x43924392);
+	pci_write_config_dword(pdev, 0xAC, 0x40094009);
+	pci_write_config_byte(pdev,  0xB1, 0x72);
+	pci_write_config_word(pdev,  0xB2, 0x328A);
+	pci_write_config_dword(pdev, 0xB4, 0x62DD62DD);
+	pci_write_config_dword(pdev, 0xB8, 0x43924392);
+	pci_write_config_dword(pdev, 0xBC, 0x40094009);
+
+	switch(tmpbyte & 0x30) {
+		case 0x00: printk(KERN_INFO "sil680: 100MHz clock.\n");break;
+		case 0x10: printk(KERN_INFO "sil680: 133MHz clock.\n");break;
+		case 0x20: printk(KERN_INFO "sil680: Using PCI clock.\n");break;
+		/* This last case is _NOT_ ok */
+		case 0x30: printk(KERN_ERR "sil680: Clock disabled ?\n");
+			return -EIO;
+	}
+	return ata_pci_init_one(pdev, port_info, 2);
+}
+
+static const struct pci_device_id sil680[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_CMD, PCI_DEVICE_ID_SII_680), },
+	{ 0, },
+};
+
+static struct pci_driver sil680_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= sil680,
+	.probe 		= sil680_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init sil680_init(void)
+{
+	return pci_register_driver(&sil680_pci_driver);
+}
+
+
+static void __exit sil680_exit(void)
+{
+	pci_unregister_driver(&sil680_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for SI680 PATA");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, sil680);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(sil680_init);
+module_exit(sil680_exit);
diff --git a/drivers/ata/pata_sis.c b/drivers/ata/pata_sis.c
new file mode 100644
index 000000000000..4ad7fb44bfa6
--- /dev/null
+++ b/drivers/ata/pata_sis.c
@@ -0,0 +1,1030 @@
+/*
+ *    pata_sis.c - SiS ATA driver
+ *
+ *	(C) 2005 Red Hat <alan@redhat.com>
+ *
+ *    Based upon linux/drivers/ide/pci/sis5513.c
+ * Copyright (C) 1999-2000	Andre Hedrick <andre@linux-ide.org>
+ * Copyright (C) 2002		Lionel Bouton <Lionel.Bouton@inet6.fr>, Maintainer
+ * Copyright (C) 2003		Vojtech Pavlik <vojtech@suse.cz>
+ * SiS Taiwan		: for direct support and hardware.
+ * Daniela Engert	: for initial ATA100 advices and numerous others.
+ * John Fremlin, Manfred Spraul, Dave Morgan, Peter Kjellerstedt	:
+ *			  for checking code correctness, providing patches.
+ * Original tests and design on the SiS620 chipset.
+ * ATA100 tests and design on the SiS735 chipset.
+ * ATA16/33 support from specs
+ * ATA133 support for SiS961/962 by L.C. Chang <lcchang@sis.com.tw>
+ *
+ *
+ *	TODO
+ *	Check MWDMA on drives that don't support MWDMA speed pio cycles ?
+ *	More Testing
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+#include <linux/ata.h>
+
+#define DRV_NAME	"pata_sis"
+#define DRV_VERSION	"0.4.2"
+
+struct sis_chipset {
+	u16 device;			/* PCI host ID */
+	struct ata_port_info *info;	/* Info block */
+	/* Probably add family, cable detect type etc here to clean
+	   up code later */
+};
+
+/**
+ *	sis_port_base		-	return PCI configuration base for dev
+ *	@adev: device
+ *
+ *	Returns the base of the PCI configuration registers for this port
+ *	number.
+ */
+
+static int sis_port_base(struct ata_device *adev)
+{
+	return  0x40 + (4 * adev->ap->port_no) +  (2 * adev->devno);
+}
+
+/**
+ *	sis_133_pre_reset	-	check for 40/80 pin
+ *	@ap: Port
+ *
+ *	Perform cable detection for the later UDMA133 capable
+ *	SiS chipset.
+ */
+
+static int sis_133_pre_reset(struct ata_port *ap)
+{
+	static const struct pci_bits sis_enable_bits[] = {
+		{ 0x4aU, 1U, 0x02UL, 0x02UL },	/* port 0 */
+		{ 0x4aU, 1U, 0x04UL, 0x04UL },	/* port 1 */
+	};
+	
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u16 tmp;
+
+	if (!pci_test_config_bits(pdev, &sis_enable_bits[ap->port_no])) {
+		ata_port_disable(ap);
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+	/* The top bit of this register is the cable detect bit */
+	pci_read_config_word(pdev, 0x50 + 2 * ap->port_no, &tmp);
+	if (tmp & 0x8000)
+		ap->cbl = ATA_CBL_PATA40;
+	else
+		ap->cbl = ATA_CBL_PATA80;
+
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	sis_error_handler - Probe specified port on PATA host controller
+ *	@ap: Port to probe
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void sis_133_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, sis_133_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+
+/**
+ *	sis_66_pre_reset	-	check for 40/80 pin
+ *	@ap: Port
+ *
+ *	Perform cable detection on the UDMA66, UDMA100 and early UDMA133
+ *	SiS IDE controllers.
+ */
+
+static int sis_66_pre_reset(struct ata_port *ap)
+{
+	static const struct pci_bits sis_enable_bits[] = {
+		{ 0x4aU, 1U, 0x02UL, 0x02UL },	/* port 0 */
+		{ 0x4aU, 1U, 0x04UL, 0x04UL },	/* port 1 */
+	};
+
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u8 tmp;
+
+	if (!pci_test_config_bits(pdev, &sis_enable_bits[ap->port_no])) {
+		ata_port_disable(ap);
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+	/* Older chips keep cable detect in bits 4/5 of reg 0x48 */
+	pci_read_config_byte(pdev, 0x48, &tmp);
+	tmp >>= ap->port_no;
+	if (tmp & 0x10)
+		ap->cbl = ATA_CBL_PATA40;
+	else
+		ap->cbl = ATA_CBL_PATA80;
+
+	return ata_std_prereset(ap);
+}
+
+/**
+ *	sis_66_error_handler - Probe specified port on PATA host controller
+ *	@ap: Port to probe
+ *	@classes:
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void sis_66_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, sis_66_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	sis_old_pre_reset		-	probe begin
+ *	@ap: ATA port
+ *
+ *	Set up cable type and use generic probe init
+ */
+
+static int sis_old_pre_reset(struct ata_port *ap)
+{
+	static const struct pci_bits sis_enable_bits[] = {
+		{ 0x4aU, 1U, 0x02UL, 0x02UL },	/* port 0 */
+		{ 0x4aU, 1U, 0x04UL, 0x04UL },	/* port 1 */
+	};
+	
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+
+	if (!pci_test_config_bits(pdev, &sis_enable_bits[ap->port_no])) {
+		ata_port_disable(ap);
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+
+/**
+ *	sis_old_error_handler - Probe specified port on PATA host controller
+ *	@ap: Port to probe
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void sis_old_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, sis_old_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	sis_set_fifo	-	Set RWP fifo bits for this device
+ *	@ap: Port
+ *	@adev: Device
+ *
+ *	SIS chipsets implement prefetch/postwrite bits for each device
+ *	on both channels. This functionality is not ATAPI compatible and
+ *	must be configured according to the class of device present
+ */
+
+static void sis_set_fifo(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u8 fifoctrl;
+	u8 mask = 0x11;
+
+	mask <<= (2 * ap->port_no);
+	mask <<= adev->devno;
+
+	/* This holds various bits including the FIFO control */
+	pci_read_config_byte(pdev, 0x4B, &fifoctrl);
+	fifoctrl &= ~mask;
+
+	/* Enable for ATA (disk) only */
+	if (adev->class == ATA_DEV_ATA)
+		fifoctrl |= mask;
+	pci_write_config_byte(pdev, 0x4B, fifoctrl);
+}
+
+/**
+ *	sis_old_set_piomode - Initialize host controller PATA PIO timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: Device we are configuring for.
+ *
+ *	Set PIO mode for device, in host controller PCI config space. This
+ *	function handles PIO set up for all chips that are pre ATA100 and
+ *	also early ATA100 devices.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void sis_old_set_piomode (struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev	= to_pci_dev(ap->host->dev);
+	int port = sis_port_base(adev);
+	u8 t1, t2;
+	int speed = adev->pio_mode - XFER_PIO_0;
+
+	const u8 active[]   = { 0x00, 0x07, 0x04, 0x03, 0x01 };
+	const u8 recovery[] = { 0x00, 0x06, 0x04, 0x03, 0x03 };
+
+	sis_set_fifo(ap, adev);
+
+	pci_read_config_byte(pdev, port, &t1);
+	pci_read_config_byte(pdev, port + 1, &t2);
+
+	t1 &= ~0x0F;	/* Clear active/recovery timings */
+	t2 &= ~0x07;
+
+	t1 |= active[speed];
+	t2 |= recovery[speed];
+
+	pci_write_config_byte(pdev, port, t1);
+	pci_write_config_byte(pdev, port + 1, t2);
+}
+
+/**
+ *	sis_100_set_pioode - Initialize host controller PATA PIO timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: Device we are configuring for.
+ *
+ *	Set PIO mode for device, in host controller PCI config space. This
+ *	function handles PIO set up for ATA100 devices and early ATA133.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void sis_100_set_piomode (struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev	= to_pci_dev(ap->host->dev);
+	int port = sis_port_base(adev);
+	int speed = adev->pio_mode - XFER_PIO_0;
+
+	const u8 actrec[] = { 0x00, 0x67, 0x44, 0x33, 0x31 };
+
+	sis_set_fifo(ap, adev);
+
+	pci_write_config_byte(pdev, port, actrec[speed]);
+}
+
+/**
+ *	sis_133_set_pioode - Initialize host controller PATA PIO timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: Device we are configuring for.
+ *
+ *	Set PIO mode for device, in host controller PCI config space. This
+ *	function handles PIO set up for the later ATA133 devices.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void sis_133_set_piomode (struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev	= to_pci_dev(ap->host->dev);
+	int port = 0x40;
+	u32 t1;
+	u32 reg54;
+	int speed = adev->pio_mode - XFER_PIO_0;
+
+	const u32 timing133[] = {
+		0x28269000,	/* Recovery << 24 | Act << 16 | Ini << 12 */
+		0x0C266000,
+		0x04263000,
+		0x0C0A3000,
+		0x05093000
+	};
+	const u32 timing100[] = {
+		0x1E1C6000,	/* Recovery << 24 | Act << 16 | Ini << 12 */
+		0x091C4000,
+		0x031C2000,
+		0x09072000,
+		0x04062000
+	};
+
+	sis_set_fifo(ap, adev);
+
+	/* If bit 14 is set then the registers are mapped at 0x70 not 0x40 */
+	pci_read_config_dword(pdev, 0x54, &reg54);
+	if (reg54 & 0x40000000)
+		port = 0x70;
+	port += 8 * ap->port_no +  4 * adev->devno;
+
+	pci_read_config_dword(pdev, port, &t1);
+	t1 &= 0xC0C00FFF;	/* Mask out timing */
+
+	if (t1 & 0x08)		/* 100 or 133 ? */
+		t1 |= timing133[speed];
+	else
+		t1 |= timing100[speed];
+	pci_write_config_byte(pdev, port, t1);
+}
+
+/**
+ *	sis_old_set_dmamode - Initialize host controller PATA DMA timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: Device to program
+ *
+ *	Set UDMA/MWDMA mode for device, in host controller PCI config space.
+ *	Handles pre UDMA and UDMA33 devices. Supports MWDMA as well unlike
+ *	the old ide/pci driver.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void sis_old_set_dmamode (struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev	= to_pci_dev(ap->host->dev);
+	int speed = adev->dma_mode - XFER_MW_DMA_0;
+	int drive_pci = sis_port_base(adev);
+	u16 timing;
+
+	const u16 mwdma_bits[] = { 0x707, 0x202, 0x202 };
+	const u16 udma_bits[]  = { 0xE000, 0xC000, 0xA000 };
+
+	pci_read_config_word(pdev, drive_pci, &timing);
+
+	if (adev->dma_mode < XFER_UDMA_0) {
+		/* bits 3-0 hold recovery timing bits 8-10 active timing and
+		   the higer bits are dependant on the device */
+		timing &= ~ 0x870F;
+		timing |= mwdma_bits[speed];
+		pci_write_config_word(pdev, drive_pci, timing);
+	} else {
+		/* Bit 15 is UDMA on/off, bit 13-14 are cycle time */
+		speed = adev->dma_mode - XFER_UDMA_0;
+		timing &= ~0x6000;
+		timing |= udma_bits[speed];
+	}
+}
+
+/**
+ *	sis_66_set_dmamode - Initialize host controller PATA DMA timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: Device to program
+ *
+ *	Set UDMA/MWDMA mode for device, in host controller PCI config space.
+ *	Handles UDMA66 and early UDMA100 devices. Supports MWDMA as well unlike
+ *	the old ide/pci driver.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void sis_66_set_dmamode (struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev	= to_pci_dev(ap->host->dev);
+	int speed = adev->dma_mode - XFER_MW_DMA_0;
+	int drive_pci = sis_port_base(adev);
+	u16 timing;
+
+	const u16 mwdma_bits[] = { 0x707, 0x202, 0x202 };
+	const u16 udma_bits[]  = { 0xF000, 0xD000, 0xB000, 0xA000, 0x9000};
+
+	pci_read_config_word(pdev, drive_pci, &timing);
+
+	if (adev->dma_mode < XFER_UDMA_0) {
+		/* bits 3-0 hold recovery timing bits 8-10 active timing and
+		   the higer bits are dependant on the device, bit 15 udma */
+		timing &= ~ 0x870F;
+		timing |= mwdma_bits[speed];
+	} else {
+		/* Bit 15 is UDMA on/off, bit 12-14 are cycle time */
+		speed = adev->dma_mode - XFER_UDMA_0;
+		timing &= ~0x6000;
+		timing |= udma_bits[speed];
+	}
+	pci_write_config_word(pdev, drive_pci, timing);
+}
+
+/**
+ *	sis_100_set_dmamode - Initialize host controller PATA DMA timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: Device to program
+ *
+ *	Set UDMA/MWDMA mode for device, in host controller PCI config space.
+ *	Handles UDMA66 and early UDMA100 devices.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void sis_100_set_dmamode (struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev	= to_pci_dev(ap->host->dev);
+	int speed = adev->dma_mode - XFER_MW_DMA_0;
+	int drive_pci = sis_port_base(adev);
+	u16 timing;
+
+	const u16 udma_bits[]  = { 0x8B00, 0x8700, 0x8500, 0x8300, 0x8200, 0x8100};
+
+	pci_read_config_word(pdev, drive_pci, &timing);
+
+	if (adev->dma_mode < XFER_UDMA_0) {
+		/* NOT SUPPORTED YET: NEED DATA SHEET. DITTO IN OLD DRIVER */
+	} else {
+		/* Bit 15 is UDMA on/off, bit 12-14 are cycle time */
+		speed = adev->dma_mode - XFER_UDMA_0;
+		timing &= ~0x0F00;
+		timing |= udma_bits[speed];
+	}
+	pci_write_config_word(pdev, drive_pci, timing);
+}
+
+/**
+ *	sis_133_early_set_dmamode - Initialize host controller PATA DMA timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: Device to program
+ *
+ *	Set UDMA/MWDMA mode for device, in host controller PCI config space.
+ *	Handles early SiS 961 bridges. Supports MWDMA as well unlike
+ *	the old ide/pci driver.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void sis_133_early_set_dmamode (struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev	= to_pci_dev(ap->host->dev);
+	int speed = adev->dma_mode - XFER_MW_DMA_0;
+	int drive_pci = sis_port_base(adev);
+	u16 timing;
+
+	const u16 udma_bits[]  = { 0x8F00, 0x8A00, 0x8700, 0x8500, 0x8300, 0x8200, 0x8100};
+
+	pci_read_config_word(pdev, drive_pci, &timing);
+
+	if (adev->dma_mode < XFER_UDMA_0) {
+		/* NOT SUPPORTED YET: NEED DATA SHEET. DITTO IN OLD DRIVER */
+	} else {
+		/* Bit 15 is UDMA on/off, bit 12-14 are cycle time */
+		speed = adev->dma_mode - XFER_UDMA_0;
+		timing &= ~0x0F00;
+		timing |= udma_bits[speed];
+	}
+	pci_write_config_word(pdev, drive_pci, timing);
+}
+
+/**
+ *	sis_133_set_dmamode - Initialize host controller PATA DMA timings
+ *	@ap: Port whose timings we are configuring
+ *	@adev: Device to program
+ *
+ *	Set UDMA/MWDMA mode for device, in host controller PCI config space.
+ *	Handles early SiS 961 bridges. Supports MWDMA as well unlike
+ *	the old ide/pci driver.
+ *
+ *	LOCKING:
+ *	None (inherited from caller).
+ */
+
+static void sis_133_set_dmamode (struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev	= to_pci_dev(ap->host->dev);
+	int speed = adev->dma_mode - XFER_MW_DMA_0;
+	int port = 0x40;
+	u32 t1;
+	u32 reg54;
+
+	/* bits 4- cycle time 8 - cvs time */
+	const u32 timing_u100[] = { 0x6B0, 0x470, 0x350, 0x140, 0x120, 0x110, 0x000 };
+	const u32 timing_u133[] = { 0x9F0, 0x6A0, 0x470, 0x250, 0x230, 0x220, 0x210 };
+
+	/* If bit 14 is set then the registers are mapped at 0x70 not 0x40 */
+	pci_read_config_dword(pdev, 0x54, &reg54);
+	if (reg54 & 0x40000000)
+		port = 0x70;
+	port += (8 * ap->port_no) +  (4 * adev->devno);
+
+	pci_read_config_dword(pdev, port, &t1);
+
+	if (adev->dma_mode < XFER_UDMA_0) {
+		t1 &= ~0x00000004;
+		/* FIXME: need data sheet to add MWDMA here. Also lacking on
+		   ide/pci driver */
+	} else {
+		speed = adev->dma_mode - XFER_UDMA_0;
+		/* if & 8 no UDMA133 - need info for ... */
+		t1 &= ~0x00000FF0;
+		t1 |= 0x00000004;
+		if (t1 & 0x08)
+			t1 |= timing_u133[speed];
+		else
+			t1 |= timing_u100[speed];
+	}
+	pci_write_config_dword(pdev, port, t1);
+}
+
+static struct scsi_host_template sis_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static const struct ata_port_operations sis_133_ops = {
+	.port_disable		= ata_port_disable,
+	.set_piomode		= sis_133_set_piomode,
+	.set_dmamode		= sis_133_set_dmamode,
+	.mode_filter		= ata_pci_default_filter,
+
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.check_status		= ata_check_status,
+	.exec_command		= ata_exec_command,
+	.dev_select		= ata_std_dev_select,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= sis_133_error_handler,
+	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= ata_qc_issue_prot,
+	.data_xfer		= ata_pio_data_xfer,
+
+	.eng_timeout		= ata_eng_timeout,
+
+	.irq_handler		= ata_interrupt,
+	.irq_clear		= ata_bmdma_irq_clear,
+
+	.port_start		= ata_port_start,
+	.port_stop		= ata_port_stop,
+	.host_stop		= ata_host_stop,
+};
+
+static const struct ata_port_operations sis_133_early_ops = {
+	.port_disable		= ata_port_disable,
+	.set_piomode		= sis_100_set_piomode,
+	.set_dmamode		= sis_133_early_set_dmamode,
+	.mode_filter		= ata_pci_default_filter,
+
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.check_status		= ata_check_status,
+	.exec_command		= ata_exec_command,
+	.dev_select		= ata_std_dev_select,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= sis_66_error_handler,
+	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+	
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= ata_qc_issue_prot,
+	.data_xfer		= ata_pio_data_xfer,
+
+	.eng_timeout		= ata_eng_timeout,
+
+	.irq_handler		= ata_interrupt,
+	.irq_clear		= ata_bmdma_irq_clear,
+
+	.port_start		= ata_port_start,
+	.port_stop		= ata_port_stop,
+	.host_stop		= ata_host_stop,
+};
+
+static const struct ata_port_operations sis_100_ops = {
+	.port_disable		= ata_port_disable,
+	.set_piomode		= sis_100_set_piomode,
+	.set_dmamode		= sis_100_set_dmamode,
+	.mode_filter		= ata_pci_default_filter,
+
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.check_status		= ata_check_status,
+	.exec_command		= ata_exec_command,
+	.dev_select		= ata_std_dev_select,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= sis_66_error_handler,
+	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+	
+
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= ata_qc_issue_prot,
+	.data_xfer		= ata_pio_data_xfer,
+
+	.eng_timeout		= ata_eng_timeout,
+
+	.irq_handler		= ata_interrupt,
+	.irq_clear		= ata_bmdma_irq_clear,
+
+	.port_start		= ata_port_start,
+	.port_stop		= ata_port_stop,
+	.host_stop		= ata_host_stop,
+};
+
+static const struct ata_port_operations sis_66_ops = {
+	.port_disable		= ata_port_disable,
+	.set_piomode		= sis_old_set_piomode,
+	.set_dmamode		= sis_66_set_dmamode,
+	.mode_filter		= ata_pci_default_filter,
+
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.check_status		= ata_check_status,
+	.exec_command		= ata_exec_command,
+	.dev_select		= ata_std_dev_select,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= sis_66_error_handler,
+	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= ata_qc_issue_prot,
+	.data_xfer		= ata_pio_data_xfer,
+
+	.eng_timeout		= ata_eng_timeout,
+
+	.irq_handler		= ata_interrupt,
+	.irq_clear		= ata_bmdma_irq_clear,
+
+	.port_start		= ata_port_start,
+	.port_stop		= ata_port_stop,
+	.host_stop		= ata_host_stop,
+};
+
+static const struct ata_port_operations sis_old_ops = {
+	.port_disable		= ata_port_disable,
+	.set_piomode		= sis_old_set_piomode,
+	.set_dmamode		= sis_old_set_dmamode,
+	.mode_filter		= ata_pci_default_filter,
+
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.check_status		= ata_check_status,
+	.exec_command		= ata_exec_command,
+	.dev_select		= ata_std_dev_select,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= sis_old_error_handler,
+	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= ata_qc_issue_prot,
+	.data_xfer		= ata_pio_data_xfer,
+
+	.eng_timeout		= ata_eng_timeout,
+
+	.irq_handler		= ata_interrupt,
+	.irq_clear		= ata_bmdma_irq_clear,
+
+	.port_start		= ata_port_start,
+	.port_stop		= ata_port_stop,
+	.host_stop		= ata_host_stop,
+};
+
+static struct ata_port_info sis_info = {
+	.sht		= &sis_sht,
+	.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+	.pio_mask	= 0x1f,	/* pio0-4 */
+	.mwdma_mask	= 0x07,
+	.udma_mask	= 0,
+	.port_ops	= &sis_old_ops,
+};
+static struct ata_port_info sis_info33 = {
+	.sht		= &sis_sht,
+	.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+	.pio_mask	= 0x1f,	/* pio0-4 */
+	.mwdma_mask	= 0x07,
+	.udma_mask	= ATA_UDMA2,	/* UDMA 33 */
+	.port_ops	= &sis_old_ops,
+};
+static struct ata_port_info sis_info66 = {
+	.sht		= &sis_sht,
+	.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+	.pio_mask	= 0x1f,	/* pio0-4 */
+	.udma_mask	= ATA_UDMA4,	/* UDMA 66 */
+	.port_ops	= &sis_66_ops,
+};
+static struct ata_port_info sis_info100 = {
+	.sht		= &sis_sht,
+	.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+	.pio_mask	= 0x1f,	/* pio0-4 */
+	.udma_mask	= ATA_UDMA5,
+	.port_ops	= &sis_100_ops,
+};
+static struct ata_port_info sis_info100_early = {
+	.sht		= &sis_sht,
+	.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+	.udma_mask	= ATA_UDMA5,
+	.pio_mask	= 0x1f,	/* pio0-4 */
+	.port_ops	= &sis_66_ops,
+};
+static struct ata_port_info sis_info133 = {
+	.sht		= &sis_sht,
+	.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+	.pio_mask	= 0x1f,	/* pio0-4 */
+	.udma_mask	= ATA_UDMA6,
+	.port_ops	= &sis_133_ops,
+};
+static struct ata_port_info sis_info133_early = {
+	.sht		= &sis_sht,
+	.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+	.pio_mask	= 0x1f,	/* pio0-4 */
+	.udma_mask	= ATA_UDMA6,
+	.port_ops	= &sis_133_early_ops,
+};
+
+
+static void sis_fixup(struct pci_dev *pdev, struct sis_chipset *sis)
+{
+	u16 regw;
+	u8 reg;
+
+	if (sis->info == &sis_info133) {
+		pci_read_config_word(pdev, 0x50, &regw);
+		if (regw & 0x08)
+			pci_write_config_word(pdev, 0x50, regw & ~0x08);
+		pci_read_config_word(pdev, 0x52, &regw);
+		if (regw & 0x08)
+			pci_write_config_word(pdev, 0x52, regw & ~0x08);
+		return;
+	}
+
+	if (sis->info == &sis_info133_early || sis->info == &sis_info100) {
+		/* Fix up latency */
+		pci_write_config_byte(pdev, PCI_LATENCY_TIMER, 0x80);
+		/* Set compatibility bit */
+		pci_read_config_byte(pdev, 0x49, &reg);
+		if (!(reg & 0x01))
+			pci_write_config_byte(pdev, 0x49, reg | 0x01);
+		return;
+	}
+
+	if (sis->info == &sis_info66 || sis->info == &sis_info100_early) {
+		/* Fix up latency */
+		pci_write_config_byte(pdev, PCI_LATENCY_TIMER, 0x80);
+		/* Set compatibility bit */
+		pci_read_config_byte(pdev, 0x52, &reg);
+		if (!(reg & 0x04))
+			pci_write_config_byte(pdev, 0x52, reg | 0x04);
+		return;
+	}
+
+	if (sis->info == &sis_info33) {
+		pci_read_config_byte(pdev, PCI_CLASS_PROG, &reg);
+		if (( reg & 0x0F ) != 0x00)
+			pci_write_config_byte(pdev, PCI_CLASS_PROG, reg & 0xF0);
+		/* Fall through to ATA16 fixup below */
+	}
+
+	if (sis->info == &sis_info || sis->info == &sis_info33) {
+		/* force per drive recovery and active timings
+		   needed on ATA_33 and below chips */
+		pci_read_config_byte(pdev, 0x52, &reg);
+		if (!(reg & 0x08))
+			pci_write_config_byte(pdev, 0x52, reg|0x08);
+		return;
+	}
+
+	BUG();
+}
+
+/**
+ *	sis_init_one - Register SiS ATA PCI device with kernel services
+ *	@pdev: PCI device to register
+ *	@ent: Entry in sis_pci_tbl matching with @pdev
+ *
+ *	Called from kernel PCI layer.  We probe for combined mode (sigh),
+ *	and then hand over control to libata, for it to do the rest.
+ *
+ *	LOCKING:
+ *	Inherited from PCI layer (may sleep).
+ *
+ *	RETURNS:
+ *	Zero on success, or -ERRNO value.
+ */
+
+static int sis_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+	static int printed_version;
+	static struct ata_port_info *port_info[2];
+	struct ata_port_info *port;
+	struct pci_dev *host = NULL;
+	struct sis_chipset *chipset = NULL;
+
+	static struct sis_chipset sis_chipsets[] = {
+		{ 0x0745, &sis_info100 },
+		{ 0x0735, &sis_info100 },
+		{ 0x0733, &sis_info100 },
+		{ 0x0635, &sis_info100 },
+		{ 0x0633, &sis_info100 },
+
+		{ 0x0730, &sis_info100_early },	/* 100 with ATA 66 layout */
+		{ 0x0550, &sis_info100_early },	/* 100 with ATA 66 layout */
+
+		{ 0x0640, &sis_info66 },
+		{ 0x0630, &sis_info66 },
+		{ 0x0620, &sis_info66 },
+		{ 0x0540, &sis_info66 },
+		{ 0x0530, &sis_info66 },
+
+		{ 0x5600, &sis_info33 },
+		{ 0x5598, &sis_info33 },
+		{ 0x5597, &sis_info33 },
+		{ 0x5591, &sis_info33 },
+		{ 0x5582, &sis_info33 },
+		{ 0x5581, &sis_info33 },
+
+		{ 0x5596, &sis_info },
+		{ 0x5571, &sis_info },
+		{ 0x5517, &sis_info },
+		{ 0x5511, &sis_info },
+
+		{0}
+	};
+	static struct sis_chipset sis133_early = {
+		0x0, &sis_info133_early
+	};
+	static struct sis_chipset sis133 = {
+		0x0, &sis_info133
+	};
+	static struct sis_chipset sis100_early = {
+		0x0, &sis_info100_early
+	};
+	static struct sis_chipset sis100 = {
+		0x0, &sis_info100
+	};
+
+	if (!printed_version++)
+		dev_printk(KERN_DEBUG, &pdev->dev,
+			   "version " DRV_VERSION "\n");
+
+	/* We have to find the bridge first */
+
+	for (chipset = &sis_chipsets[0]; chipset->device; chipset++) {
+		host = pci_get_device(PCI_VENDOR_ID_SI, chipset->device, NULL);
+		if (host != NULL) {
+			if (chipset->device == 0x630) {	/* SIS630 */
+				u8 host_rev;
+				pci_read_config_byte(host, PCI_REVISION_ID, &host_rev);
+				if (host_rev >= 0x30)	/* 630 ET */
+					chipset = &sis100_early;
+			}
+			break;
+		}
+	}
+
+	/* Look for concealed bridges */
+	if (host == NULL) {
+		/* Second check */
+		u32 idemisc;
+		u16 trueid;
+
+		/* Disable ID masking and register remapping then
+		   see what the real ID is */
+
+		pci_read_config_dword(pdev, 0x54, &idemisc);
+		pci_write_config_dword(pdev, 0x54, idemisc & 0x7fffffff);
+		pci_read_config_word(pdev, PCI_DEVICE_ID, &trueid);
+		pci_write_config_dword(pdev, 0x54, idemisc);
+
+		switch(trueid) {
+		case 0x5518:	/* SIS 962/963 */
+			chipset = &sis133;
+			if ((idemisc & 0x40000000) == 0) {
+				pci_write_config_dword(pdev, 0x54, idemisc | 0x40000000);
+				printk(KERN_INFO "SIS5513: Switching to 5513 register mapping\n");
+			}
+			break;
+		case 0x0180:	/* SIS 965/965L */
+			chipset =  &sis133;
+			break;
+		case 0x1180:	/* SIS 966/966L */
+			chipset =  &sis133;
+			break;
+		}
+	}
+
+	/* Further check */
+	if (chipset == NULL) {
+		struct pci_dev *lpc_bridge;
+		u16 trueid;
+		u8 prefctl;
+		u8 idecfg;
+		u8 sbrev;
+
+		/* Try the second unmasking technique */
+		pci_read_config_byte(pdev, 0x4a, &idecfg);
+		pci_write_config_byte(pdev, 0x4a, idecfg | 0x10);
+		pci_read_config_word(pdev, PCI_DEVICE_ID, &trueid);
+		pci_write_config_byte(pdev, 0x4a, idecfg);
+
+		switch(trueid) {
+		case 0x5517:
+			lpc_bridge = pci_get_slot(pdev->bus, 0x10); /* Bus 0 Dev 2 Fn 0 */
+			if (lpc_bridge == NULL)
+				break;
+			pci_read_config_byte(lpc_bridge, PCI_REVISION_ID, &sbrev);
+			pci_read_config_byte(pdev, 0x49, &prefctl);
+			pci_dev_put(lpc_bridge);
+
+			if (sbrev == 0x10 && (prefctl & 0x80)) {
+				chipset = &sis133_early;
+				break;
+			}
+			chipset = &sis100;
+			break;
+		}
+	}
+	pci_dev_put(host);
+
+	/* No chipset info, no support */
+	if (chipset == NULL)
+		return -ENODEV;
+
+	port = chipset->info;
+	port->private_data = chipset;
+
+	sis_fixup(pdev, chipset);
+
+	port_info[0] = port_info[1] = port;
+	return ata_pci_init_one(pdev, port_info, 2);
+}
+
+static const struct pci_device_id sis_pci_tbl[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_SI, 0x5513), },	/* SiS 5513 */
+	{ PCI_DEVICE(PCI_VENDOR_ID_SI, 0x5518), },	/* SiS 5518 */
+	{ }
+};
+
+static struct pci_driver sis_pci_driver = {
+	.name			= DRV_NAME,
+	.id_table		= sis_pci_tbl,
+	.probe			= sis_init_one,
+	.remove			= ata_pci_remove_one,
+};
+
+static int __init sis_init(void)
+{
+	return pci_register_driver(&sis_pci_driver);
+}
+
+static void __exit sis_exit(void)
+{
+	pci_unregister_driver(&sis_pci_driver);
+}
+
+
+module_init(sis_init);
+module_exit(sis_exit);
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("SCSI low-level driver for SiS ATA");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, sis_pci_tbl);
+MODULE_VERSION(DRV_VERSION);
+
diff --git a/drivers/ata/pata_sl82c105.c b/drivers/ata/pata_sl82c105.c
new file mode 100644
index 000000000000..47b290b5604a
--- /dev/null
+++ b/drivers/ata/pata_sl82c105.c
@@ -0,0 +1,388 @@
+/*
+ * pata_sl82c105.c 	- SL82C105 PATA for new ATA layer
+ *			  (C) 2005 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ * Based in part on linux/drivers/ide/pci/sl82c105.c
+ * 		SL82C105/Winbond 553 IDE driver
+ *
+ * and in part on the documentation and errata sheet
+ */
+ 
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "pata_sl82c105"
+#define DRV_VERSION "0.2.2"
+
+enum {
+	/*
+	 * SL82C105 PCI config register 0x40 bits.
+	 */
+	CTRL_IDE_IRQB	=	(1 << 30),
+	CTRL_IDE_IRQA   =	(1 << 28),
+	CTRL_LEGIRQ     =	(1 << 11),
+	CTRL_P1F16      =	(1 << 5),
+	CTRL_P1EN       =	(1 << 4),
+	CTRL_P0F16      =	(1 << 1),
+	CTRL_P0EN       =	(1 << 0)
+};
+
+/**
+ *	sl82c105_pre_reset		-	probe begin
+ *	@ap: ATA port
+ *
+ *	Set up cable type and use generic probe init
+ */
+ 
+static int sl82c105_pre_reset(struct ata_port *ap)
+{
+	static const struct pci_bits sl82c105_enable_bits[] = {
+		{ 0x40, 1, 0x01, 0x01 },
+		{ 0x40, 1, 0x10, 0x10 }
+	};
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+
+	if (ap->port_no && !pci_test_config_bits(pdev, &sl82c105_enable_bits[ap->port_no])) {
+		ata_port_disable(ap);
+		dev_printk(KERN_INFO, &pdev->dev, "port disabled. ignoring.\n");
+		return 0;
+	}
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+
+static void sl82c105_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, sl82c105_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+
+/**
+ *	sl82c105_configure_piomode	-	set chip PIO timing
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *	@pio: PIO mode
+ *
+ *	Called to do the PIO mode setup. Our timing registers are shared
+ *	so a configure_dmamode call will undo any work we do here and vice
+ *	versa
+ */
+ 
+static void sl82c105_configure_piomode(struct ata_port *ap, struct ata_device *adev, int pio)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	static u16 pio_timing[5] = {
+		0x50D, 0x407, 0x304, 0x242, 0x240
+	};
+	u16 dummy;
+	int timing = 0x44 + (8 * ap->port_no) + (4 * adev->devno);
+	
+	pci_write_config_word(pdev, timing, pio_timing[pio]);
+	/* Can we lose this oddity of the old driver */
+	pci_read_config_word(pdev, timing, &dummy);
+}
+
+/**
+ *	sl82c105_set_piomode	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Called to do the PIO mode setup. Our timing registers are shared
+ *	but we want to set the PIO timing by default.
+ */
+ 
+static void sl82c105_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	sl82c105_configure_piomode(ap, adev, adev->pio_mode - XFER_PIO_0);
+}
+
+/**
+ *	sl82c105_configure_dmamode	-	set DMA mode in chip
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Load DMA cycle times into the chip ready for a DMA transfer
+ *	to occur.
+ */
+ 
+static void sl82c105_configure_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	static u16 dma_timing[3] = {
+		0x707, 0x201, 0x200
+	};
+	u16 dummy;
+	int timing = 0x44 + (8 * ap->port_no) + (4 * adev->devno);
+	int dma = adev->dma_mode - XFER_MW_DMA_0;
+	
+	pci_write_config_word(pdev, timing, dma_timing[dma]);
+	/* Can we lose this oddity of the old driver */
+	pci_read_config_word(pdev, timing, &dummy);
+}
+
+/**
+ *	sl82c105_set_dmamode	-	set initial DMA mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Called to do the DMA mode setup. This replaces the PIO timings
+ *	for the device in question. Set appropriate PIO timings not DMA
+ *	timings at this point.
+ */
+ 
+static void sl82c105_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	switch(adev->dma_mode) {
+		case XFER_MW_DMA_0:
+			sl82c105_configure_piomode(ap, adev, 1);
+			break;
+		case XFER_MW_DMA_1:
+			sl82c105_configure_piomode(ap, adev, 3);
+			break;
+		case XFER_MW_DMA_2:
+			sl82c105_configure_piomode(ap, adev, 3);
+			break;
+		default:
+			BUG();
+	}			
+}
+
+/**
+ *	sl82c105_reset_engine	-	Reset the DMA engine
+ *	@ap: ATA interface
+ *
+ *	The sl82c105 has some serious problems with the DMA engine
+ *	when transfers don't run as expected or ATAPI is used. The 
+ *	recommended fix is to reset the engine each use using a chip
+ *	test register.
+ */
+ 
+static void sl82c105_reset_engine(struct ata_port *ap)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u16 val;
+	
+	pci_read_config_word(pdev, 0x7E, &val);
+	pci_write_config_word(pdev, 0x7E, val | 4);
+	pci_write_config_word(pdev, 0x7E, val & ~4);
+}
+
+/**
+ *	sl82c105_bmdma_start		-	DMA engine begin
+ *	@qc: ATA command
+ *
+ *	Reset the DMA engine each use as recommended by the errata
+ *	document. 
+ *
+ *	FIXME: if we switch clock at BMDMA start/end we might get better
+ *	PIO performance on DMA capable devices.
+ */
+ 
+static void sl82c105_bmdma_start(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+
+	sl82c105_reset_engine(ap);
+	
+	/* Set the clocks for DMA */
+	sl82c105_configure_dmamode(ap, qc->dev);
+	/* Activate DMA */	
+	ata_bmdma_start(qc);
+}
+
+/**
+ *	sl82c105_bmdma_end		-	DMA engine stop
+ *	@qc: ATA command
+ *
+ *	Reset the DMA engine each use as recommended by the errata
+ *	document.
+ *
+ *	This function is also called to turn off DMA when a timeout occurs
+ *	during DMA operation. In both cases we need to reset the engine,
+ *	so no actual eng_timeout handler is required.
+ *
+ *	We assume bmdma_stop is always called if bmdma_start as called. If
+ *	not then we may need to wrap qc_issue.
+ */
+ 
+static void sl82c105_bmdma_stop(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+
+	ata_bmdma_stop(qc);
+	sl82c105_reset_engine(ap);
+	
+	/* This will redo the initial setup of the DMA device to matching
+	   PIO timings */
+	sl82c105_set_dmamode(ap, qc->dev);
+}
+
+static struct scsi_host_template sl82c105_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations sl82c105_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= sl82c105_set_piomode,
+	.set_dmamode	= sl82c105_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.error_handler	= sl82c105_error_handler,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= sl82c105_bmdma_start,
+	.bmdma_stop	= sl82c105_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+	
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};	
+
+/**
+ *	sl82c105_bridge_revision	-	find bridge version
+ *	@pdev: PCI device for the ATA function
+ *
+ *	Locates the PCI bridge associated with the ATA function and
+ *	providing it is a Winbond 553 reports the revision. If it cannot
+ *	find a revision or the right device it returns -1
+ */
+ 
+static int sl82c105_bridge_revision(struct pci_dev *pdev)
+{
+	struct pci_dev *bridge;
+	u8 rev;
+
+	/*
+	 * The bridge should be part of the same device, but function 0.
+	 */
+	bridge = pci_get_slot(pdev->bus,
+			       PCI_DEVFN(PCI_SLOT(pdev->devfn), 0));
+	if (!bridge)
+		return -1;
+
+	/*
+	 * Make sure it is a Winbond 553 and is an ISA bridge.
+	 */
+	if (bridge->vendor != PCI_VENDOR_ID_WINBOND ||
+	    bridge->device != PCI_DEVICE_ID_WINBOND_83C553 ||
+	    bridge->class >> 8 != PCI_CLASS_BRIDGE_ISA) {
+	    	pci_dev_put(bridge);
+		return -1;
+	}
+	/*
+	 * We need to find function 0's revision, not function 1
+	 */
+	pci_read_config_byte(bridge, PCI_REVISION_ID, &rev);
+
+	pci_dev_put(bridge);
+	return rev;
+}
+
+	
+static int sl82c105_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info_dma = {
+		.sht = &sl82c105_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.port_ops = &sl82c105_port_ops
+	};
+	static struct ata_port_info info_early = {
+		.sht = &sl82c105_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.port_ops = &sl82c105_port_ops
+	};
+	static struct ata_port_info *port_info[2] = { &info_early, &info_early };
+	u32 val;
+	int rev;
+
+	rev = sl82c105_bridge_revision(dev);
+	
+	if (rev == -1)
+		dev_printk(KERN_WARNING, &dev->dev, "pata_sl82c105: Unable to find bridge, disabling DMA.\n");
+	else if (rev <= 5)
+		dev_printk(KERN_WARNING, &dev->dev, "pata_sl82c105: Early bridge revision, no DMA available.\n");
+	else {
+		port_info[0] = &info_dma;
+		port_info[1] = &info_dma;
+	}
+	
+	pci_read_config_dword(dev, 0x40, &val);
+	val |= CTRL_P0EN | CTRL_P0F16 | CTRL_P1F16;
+	pci_write_config_dword(dev, 0x40, val);
+
+
+	return ata_pci_init_one(dev, port_info, 1); /* For now */
+}
+
+static struct pci_device_id sl82c105[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_WINBOND, PCI_DEVICE_ID_WINBOND_82C105), },
+	{ 0, },
+};
+
+static struct pci_driver sl82c105_pci_driver = {
+	.name 		= DRV_NAME,
+	.id_table	= sl82c105,
+	.probe 		= sl82c105_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init sl82c105_init(void)
+{
+	return pci_register_driver(&sl82c105_pci_driver);
+}
+
+
+static void __exit sl82c105_exit(void)
+{
+	pci_unregister_driver(&sl82c105_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for Sl82c105");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, sl82c105);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(sl82c105_init);
+module_exit(sl82c105_exit);
diff --git a/drivers/ata/pata_triflex.c b/drivers/ata/pata_triflex.c
new file mode 100644
index 000000000000..7dd6ea3a21e4
--- /dev/null
+++ b/drivers/ata/pata_triflex.c
@@ -0,0 +1,285 @@
+/*
+ * pata_triflex.c 	- Compaq PATA for new ATA layer
+ *			  (C) 2005 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ * based upon
+ *
+ * triflex.c
+ *
+ * IDE Chipset driver for the Compaq TriFlex IDE controller.
+ *
+ * Known to work with the Compaq Workstation 5x00 series.
+ *
+ * Copyright (C) 2002 Hewlett-Packard Development Group, L.P.
+ * Author: Torben Mathiasen <torben.mathiasen@hp.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ *
+ * Loosely based on the piix & svwks drivers.
+ *
+ * Documentation:
+ *	Not publically available.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "pata_triflex"
+#define DRV_VERSION "0.2.5"
+
+/**
+ *	triflex_probe_init		-	probe begin
+ *	@ap: ATA port
+ *
+ *	Set up cable type and use generic probe init
+ */
+
+static int triflex_probe_init(struct ata_port *ap)
+{
+	static const struct pci_bits triflex_enable_bits[] = {
+		{ 0x80, 1, 0x01, 0x01 },
+		{ 0x80, 1, 0x02, 0x02 }
+	};
+
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	
+	if (!pci_test_config_bits(pdev, &triflex_enable_bits[ap->port_no])) {
+		ata_port_disable(ap);
+		printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+		return 0;
+	}
+	ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+
+
+static void triflex_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, triflex_probe_init, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	triflex_load_timing		-	timing configuration
+ *	@ap: ATA interface
+ *	@adev: Device on the bus
+ *	@speed: speed to configure
+ *
+ *	The Triflex has one set of timings per device per channel. This
+ *	means we must do some switching. As the PIO and DMA timings don't
+ *	match we have to do some reloading unlike PIIX devices where tuning
+ *	tricks can avoid it.
+ */
+
+static void triflex_load_timing(struct ata_port *ap, struct ata_device *adev, int speed)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u32 timing = 0;
+	u32 triflex_timing, old_triflex_timing;
+	int channel_offset = ap->port_no ? 0x74: 0x70;
+	unsigned int is_slave	= (adev->devno != 0);
+
+
+	pci_read_config_dword(pdev, channel_offset, &old_triflex_timing);
+	triflex_timing = old_triflex_timing;
+
+	switch(speed)
+	{
+		case XFER_MW_DMA_2:
+			timing = 0x0103;break;
+		case XFER_MW_DMA_1:
+			timing = 0x0203;break;
+		case XFER_MW_DMA_0:
+			timing = 0x0808;break;
+		case XFER_SW_DMA_2:
+		case XFER_SW_DMA_1:
+		case XFER_SW_DMA_0:
+			timing = 0x0F0F;break;
+		case XFER_PIO_4:
+			timing = 0x0202;break;
+		case XFER_PIO_3:
+			timing = 0x0204;break;
+		case XFER_PIO_2:
+			timing = 0x0404;break;
+		case XFER_PIO_1:
+			timing = 0x0508;break;
+		case XFER_PIO_0:
+			timing = 0x0808;break;
+		default:
+			BUG();
+	}
+	triflex_timing &= ~ (0xFFFF << (16 * is_slave));
+	triflex_timing |= (timing << (16 * is_slave));
+
+	if (triflex_timing != old_triflex_timing)
+		pci_write_config_dword(pdev, channel_offset, triflex_timing);
+}
+
+/**
+ *	triflex_set_piomode	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	Use the timing loader to set up the PIO mode. We have to do this
+ *	because DMA start/stop will only be called once DMA occurs. If there
+ *	has been no DMA then the PIO timings are still needed.
+ */
+static void triflex_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	triflex_load_timing(ap, adev, adev->pio_mode);
+}
+
+/**
+ *	triflex_dma_start	-	DMA start callback
+ *	@qc: Command in progress
+ *
+ *	Usually drivers set the DMA timing at the point the set_dmamode call
+ *	is made. Triflex however requires we load new timings on the
+ *	transition or keep matching PIO/DMA pairs (ie MWDMA2/PIO4 etc).
+ *	We load the DMA timings just before starting DMA and then restore
+ *	the PIO timing when the DMA is finished.
+ */
+
+static void triflex_bmdma_start(struct ata_queued_cmd *qc)
+{
+	triflex_load_timing(qc->ap, qc->dev, qc->dev->dma_mode);
+	ata_bmdma_start(qc);
+}
+
+/**
+ *	triflex_dma_stop	-	DMA stop callback
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *
+ *	We loaded new timings in dma_start, as a result we need to restore
+ *	the PIO timings in dma_stop so that the next command issue gets the
+ *	right clock values.
+ */
+
+static void triflex_bmdma_stop(struct ata_queued_cmd *qc)
+{
+	ata_bmdma_stop(qc);
+	triflex_load_timing(qc->ap, qc->dev, qc->dev->pio_mode);
+}
+
+static struct scsi_host_template triflex_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations triflex_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= triflex_set_piomode,
+	.mode_filter	= ata_pci_default_filter,
+
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= triflex_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= triflex_bmdma_start,
+	.bmdma_stop	= triflex_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static int triflex_init_one(struct pci_dev *dev, const struct pci_device_id *id)
+{
+	static struct ata_port_info info = {
+		.sht = &triflex_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.port_ops = &triflex_port_ops
+	};
+	static struct ata_port_info *port_info[2] = { &info, &info };
+	static int printed_version;
+
+	if (!printed_version++)
+		dev_printk(KERN_DEBUG, &dev->dev, "version " DRV_VERSION "\n");
+
+	return ata_pci_init_one(dev, port_info, 2);
+}
+
+static const struct pci_device_id triflex[] = {
+	{ PCI_VENDOR_ID_COMPAQ, PCI_DEVICE_ID_COMPAQ_TRIFLEX_IDE,
+				  PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 },
+	{ 0, },
+};
+
+static struct pci_driver triflex_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= triflex,
+	.probe 		= triflex_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init triflex_init(void)
+{
+	return pci_register_driver(&triflex_pci_driver);
+}
+
+
+static void __exit triflex_exit(void)
+{
+	pci_unregister_driver(&triflex_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for Compaq Triflex");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, triflex);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(triflex_init);
+module_exit(triflex_exit);
diff --git a/drivers/ata/pata_via.c b/drivers/ata/pata_via.c
new file mode 100644
index 000000000000..2580e1606d9e
--- /dev/null
+++ b/drivers/ata/pata_via.c
@@ -0,0 +1,568 @@
+/*
+ * pata_via.c 	- VIA PATA for new ATA layer
+ *			  (C) 2005-2006 Red Hat Inc
+ *			  Alan Cox <alan@redhat.com>
+ *
+ *  Documentation
+ *	Most chipset documentation available under NDA only
+ *
+ *  VIA version guide
+ *	VIA VT82C561	-	early design, uses ata_generic currently
+ *	VIA VT82C576	-	MWDMA, 33Mhz
+ *	VIA VT82C586	-	MWDMA, 33Mhz
+ *	VIA VT82C586a	-	Added UDMA to 33Mhz
+ *	VIA VT82C586b	-	UDMA33
+ *	VIA VT82C596a	-	Nonfunctional UDMA66
+ *	VIA VT82C596b	-	Working UDMA66
+ *	VIA VT82C686	-	Nonfunctional UDMA66
+ *	VIA VT82C686a	-	Working UDMA66
+ *	VIA VT82C686b	-	Updated to UDMA100
+ *	VIA VT8231	-	UDMA100
+ *	VIA VT8233	-	UDMA100
+ *	VIA VT8233a	-	UDMA133
+ *	VIA VT8233c	-	UDMA100
+ *	VIA VT8235	-	UDMA133
+ *	VIA VT8237	-	UDMA133
+ *
+ *	Most registers remain compatible across chips. Others start reserved
+ *	and acquire sensible semantics if set to 1 (eg cable detect). A few
+ *	exceptions exist, notably around the FIFO settings.
+ *
+ *	One additional quirk of the VIA design is that like ALi they use few
+ *	PCI IDs for a lot of chips.
+ *
+ *	Based heavily on:
+ *
+ * Version 3.38
+ *
+ * VIA IDE driver for Linux. Supported southbridges:
+ *
+ *   vt82c576, vt82c586, vt82c586a, vt82c586b, vt82c596a, vt82c596b,
+ *   vt82c686, vt82c686a, vt82c686b, vt8231, vt8233, vt8233c, vt8233a,
+ *   vt8235, vt8237
+ *
+ * Copyright (c) 2000-2002 Vojtech Pavlik
+ *
+ * Based on the work of:
+ *	Michel Aubry
+ *	Jeff Garzik
+ *	Andre Hedrick
+
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <scsi/scsi_host.h>
+#include <linux/libata.h>
+
+#define DRV_NAME "pata_via"
+#define DRV_VERSION "0.1.13"
+
+/*
+ *	The following comes directly from Vojtech Pavlik's ide/pci/via82cxxx
+ *	driver.
+ */
+
+enum {
+	VIA_UDMA	= 0x007,
+	VIA_UDMA_NONE	= 0x000,
+	VIA_UDMA_33	= 0x001,
+	VIA_UDMA_66	= 0x002,
+	VIA_UDMA_100	= 0x003,
+	VIA_UDMA_133	= 0x004,
+	VIA_BAD_PREQ	= 0x010, /* Crashes if PREQ# till DDACK# set */
+	VIA_BAD_CLK66	= 0x020, /* 66 MHz clock doesn't work correctly */
+	VIA_SET_FIFO	= 0x040, /* Needs to have FIFO split set */
+	VIA_NO_UNMASK	= 0x080, /* Doesn't work with IRQ unmasking on */
+	VIA_BAD_ID	= 0x100, /* Has wrong vendor ID (0x1107) */
+	VIA_BAD_AST	= 0x200, /* Don't touch Address Setup Timing */
+	VIA_NO_ENABLES	= 0x400, /* Has no enablebits */
+};
+
+/*
+ * VIA SouthBridge chips.
+ */
+
+static const struct via_isa_bridge {
+	const char *name;
+	u16 id;
+	u8 rev_min;
+	u8 rev_max;
+	u16 flags;
+} via_isa_bridges[] = {
+	{ "cx700",	PCI_DEVICE_ID_VIA_CX700,    0x00, 0x2f, VIA_UDMA_133 | VIA_BAD_AST },
+	{ "vt6410",	PCI_DEVICE_ID_VIA_6410,     0x00, 0x2f, VIA_UDMA_133 | VIA_BAD_AST | VIA_NO_ENABLES},
+	{ "vt8237a",	PCI_DEVICE_ID_VIA_8237A,    0x00, 0x2f, VIA_UDMA_133 | VIA_BAD_AST },
+	{ "vt8237",	PCI_DEVICE_ID_VIA_8237,     0x00, 0x2f, VIA_UDMA_133 | VIA_BAD_AST },
+	{ "vt8235",	PCI_DEVICE_ID_VIA_8235,     0x00, 0x2f, VIA_UDMA_133 | VIA_BAD_AST },
+	{ "vt8233a",	PCI_DEVICE_ID_VIA_8233A,    0x00, 0x2f, VIA_UDMA_133 | VIA_BAD_AST },
+	{ "vt8233c",	PCI_DEVICE_ID_VIA_8233C_0,  0x00, 0x2f, VIA_UDMA_100 },
+	{ "vt8233",	PCI_DEVICE_ID_VIA_8233_0,   0x00, 0x2f, VIA_UDMA_100 },
+	{ "vt8231",	PCI_DEVICE_ID_VIA_8231,     0x00, 0x2f, VIA_UDMA_100 },
+	{ "vt82c686b",	PCI_DEVICE_ID_VIA_82C686,   0x40, 0x4f, VIA_UDMA_100 },
+	{ "vt82c686a",	PCI_DEVICE_ID_VIA_82C686,   0x10, 0x2f, VIA_UDMA_66 },
+	{ "vt82c686",	PCI_DEVICE_ID_VIA_82C686,   0x00, 0x0f, VIA_UDMA_33 | VIA_BAD_CLK66 },
+	{ "vt82c596b",	PCI_DEVICE_ID_VIA_82C596,   0x10, 0x2f, VIA_UDMA_66 },
+	{ "vt82c596a",	PCI_DEVICE_ID_VIA_82C596,   0x00, 0x0f, VIA_UDMA_33 | VIA_BAD_CLK66 },
+	{ "vt82c586b",	PCI_DEVICE_ID_VIA_82C586_0, 0x47, 0x4f, VIA_UDMA_33 | VIA_SET_FIFO },
+	{ "vt82c586b",	PCI_DEVICE_ID_VIA_82C586_0, 0x40, 0x46, VIA_UDMA_33 | VIA_SET_FIFO | VIA_BAD_PREQ },
+	{ "vt82c586b",	PCI_DEVICE_ID_VIA_82C586_0, 0x30, 0x3f, VIA_UDMA_33 | VIA_SET_FIFO },
+	{ "vt82c586a",	PCI_DEVICE_ID_VIA_82C586_0, 0x20, 0x2f, VIA_UDMA_33 | VIA_SET_FIFO },
+	{ "vt82c586",	PCI_DEVICE_ID_VIA_82C586_0, 0x00, 0x0f, VIA_UDMA_NONE | VIA_SET_FIFO },
+	{ "vt82c576",	PCI_DEVICE_ID_VIA_82C576,   0x00, 0x2f, VIA_UDMA_NONE | VIA_SET_FIFO | VIA_NO_UNMASK },
+	{ "vt82c576",	PCI_DEVICE_ID_VIA_82C576,   0x00, 0x2f, VIA_UDMA_NONE | VIA_SET_FIFO | VIA_NO_UNMASK | VIA_BAD_ID },
+	{ NULL }
+};
+
+/**
+ *	via_cable_detect	-	cable detection
+ *	@ap: ATA port
+ *
+ *	Perform cable detection. Actually for the VIA case the BIOS
+ *	already did this for us. We read the values provided by the
+ *	BIOS. If you are using an 8235 in a non-PC configuration you
+ *	may need to update this code.
+ *
+ *	Hotplug also impacts on this.
+ */
+
+static int via_cable_detect(struct ata_port *ap) {
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	u32 ata66;
+
+	pci_read_config_dword(pdev, 0x50, &ata66);
+	/* Check both the drive cable reporting bits, we might not have
+	   two drives */
+	if (ata66 & (0x10100000 >> (16 * ap->port_no)))
+		return ATA_CBL_PATA80;
+	else
+		return ATA_CBL_PATA40;
+}
+
+static int via_pre_reset(struct ata_port *ap)
+{
+	const struct via_isa_bridge *config = ap->host->private_data;
+
+	if (!(config->flags & VIA_NO_ENABLES)) {
+		static const struct pci_bits via_enable_bits[] = {
+			{ 0x40, 1, 0x02, 0x02 },
+			{ 0x40, 1, 0x01, 0x01 }
+		};
+
+		struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+		
+		if (!pci_test_config_bits(pdev, &via_enable_bits[ap->port_no])) {
+			ata_port_disable(ap);
+			printk(KERN_INFO "ata%u: port disabled. ignoring.\n", ap->id);
+			return 0;
+		}
+	}
+
+	if ((config->flags & VIA_UDMA) >= VIA_UDMA_66)
+		ap->cbl = via_cable_detect(ap);
+	else
+		ap->cbl = ATA_CBL_PATA40;
+	return ata_std_prereset(ap);
+}
+
+
+/**
+ *	via_error_handler		-	reset for VIA chips
+ *	@ap: ATA port
+ *
+ *	Handle the reset callback for the later chips with cable detect
+ */
+
+static void via_error_handler(struct ata_port *ap)
+{
+	ata_bmdma_drive_eh(ap, via_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
+}
+
+/**
+ *	via_do_set_mode	-	set initial PIO mode data
+ *	@ap: ATA interface
+ *	@adev: ATA device
+ *	@mode: ATA mode being programmed
+ *	@tdiv: Clocks per PCI clock
+ *	@set_ast: Set to program address setup
+ *	@udma_type: UDMA mode/format of registers
+ *
+ *	Program the VIA registers for DMA and PIO modes. Uses the ata timing
+ *	support in order to compute modes.
+ *
+ *	FIXME: Hotplug will require we serialize multiple mode changes
+ *	on the two channels.
+ */
+
+static void via_do_set_mode(struct ata_port *ap, struct ata_device *adev, int mode, int tdiv, int set_ast, int udma_type)
+{
+	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
+	struct ata_device *peer = ata_dev_pair(adev);
+	struct ata_timing t, p;
+	static int via_clock = 33333;	/* Bus clock in kHZ - ought to be tunable one day */
+	unsigned long T =  1000000000 / via_clock;
+	unsigned long UT = T/tdiv;
+	int ut;
+	int offset = 3 - (2*ap->port_no) - adev->devno;
+
+
+	/* Calculate the timing values we require */
+	ata_timing_compute(adev, mode, &t, T, UT);
+
+	/* We share 8bit timing so we must merge the constraints */
+	if (peer) {
+		if (peer->pio_mode) {
+			ata_timing_compute(peer, peer->pio_mode, &p, T, UT);
+			ata_timing_merge(&p, &t, &t, ATA_TIMING_8BIT);
+		}
+	}
+
+	/* Address setup is programmable but breaks on UDMA133 setups */
+	if (set_ast) {
+		u8 setup;	/* 2 bits per drive */
+		int shift = 2 * offset;
+
+		pci_read_config_byte(pdev, 0x4C, &setup);
+		setup &= ~(3 << shift);
+		setup |= FIT(t.setup, 1, 4) << shift;	/* 1,4 or 1,4 - 1  FIXME */
+		pci_write_config_byte(pdev, 0x4C, setup);
+	}
+
+	/* Load the PIO mode bits */
+	pci_write_config_byte(pdev, 0x4F - ap->port_no,
+		((FIT(t.act8b, 1, 16) - 1) << 4) | (FIT(t.rec8b, 1, 16) - 1));
+	pci_write_config_byte(pdev, 0x48 + offset,
+		((FIT(t.active, 1, 16) - 1) << 4) | (FIT(t.recover, 1, 16) - 1));
+
+	/* Load the UDMA bits according to type */
+	switch(udma_type) {
+		default:
+			/* BUG() ? */
+			/* fall through */
+		case 33:
+			ut = t.udma ? (0xe0 | (FIT(t.udma, 2, 5) - 2)) : 0x03;
+			break;
+		case 66:
+			ut = t.udma ? (0xe8 | (FIT(t.udma, 2, 9) - 2)) : 0x0f;
+			break;
+		case 100:
+			ut = t.udma ? (0xe0 | (FIT(t.udma, 2, 9) - 2)) : 0x07;
+			break;
+		case 133:
+			ut = t.udma ? (0xe0 | (FIT(t.udma, 2, 9) - 2)) : 0x07;
+			break;
+	}
+	/* Set UDMA unless device is not UDMA capable */
+	if (udma_type)
+		pci_write_config_byte(pdev, 0x50 + offset, ut);
+}
+
+static void via_set_piomode(struct ata_port *ap, struct ata_device *adev)
+{
+	const struct via_isa_bridge *config = ap->host->private_data;
+	int set_ast = (config->flags & VIA_BAD_AST) ? 0 : 1;
+	int mode = config->flags & VIA_UDMA;
+	static u8 tclock[5] = { 1, 1, 2, 3, 4 };
+	static u8 udma[5] = { 0, 33, 66, 100, 133 };
+
+	via_do_set_mode(ap, adev, adev->pio_mode, tclock[mode], set_ast, udma[mode]);
+}
+
+static void via_set_dmamode(struct ata_port *ap, struct ata_device *adev)
+{
+	const struct via_isa_bridge *config = ap->host->private_data;
+	int set_ast = (config->flags & VIA_BAD_AST) ? 0 : 1;
+	int mode = config->flags & VIA_UDMA;
+	static u8 tclock[5] = { 1, 1, 2, 3, 4 };
+	static u8 udma[5] = { 0, 33, 66, 100, 133 };
+
+	via_do_set_mode(ap, adev, adev->dma_mode, tclock[mode], set_ast, udma[mode]);
+}
+
+static struct scsi_host_template via_sht = {
+	.module			= THIS_MODULE,
+	.name			= DRV_NAME,
+	.ioctl			= ata_scsi_ioctl,
+	.queuecommand		= ata_scsi_queuecmd,
+	.can_queue		= ATA_DEF_QUEUE,
+	.this_id		= ATA_SHT_THIS_ID,
+	.sg_tablesize		= LIBATA_MAX_PRD,
+	.max_sectors		= ATA_MAX_SECTORS,
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
+	.emulated		= ATA_SHT_EMULATED,
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	.proc_name		= DRV_NAME,
+	.dma_boundary		= ATA_DMA_BOUNDARY,
+	.slave_configure	= ata_scsi_slave_config,
+	.bios_param		= ata_std_bios_param,
+};
+
+static struct ata_port_operations via_port_ops = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= via_set_piomode,
+	.set_dmamode	= via_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= via_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+static struct ata_port_operations via_port_ops_noirq = {
+	.port_disable	= ata_port_disable,
+	.set_piomode	= via_set_piomode,
+	.set_dmamode	= via_set_dmamode,
+	.mode_filter	= ata_pci_default_filter,
+
+	.tf_load	= ata_tf_load,
+	.tf_read	= ata_tf_read,
+	.check_status 	= ata_check_status,
+	.exec_command	= ata_exec_command,
+	.dev_select 	= ata_std_dev_select,
+
+	.freeze		= ata_bmdma_freeze,
+	.thaw		= ata_bmdma_thaw,
+	.error_handler	= via_error_handler,
+	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+
+	.bmdma_setup 	= ata_bmdma_setup,
+	.bmdma_start 	= ata_bmdma_start,
+	.bmdma_stop	= ata_bmdma_stop,
+	.bmdma_status 	= ata_bmdma_status,
+
+	.qc_prep 	= ata_qc_prep,
+	.qc_issue	= ata_qc_issue_prot,
+	.eng_timeout	= ata_eng_timeout,
+	.data_xfer	= ata_pio_data_xfer_noirq,
+
+	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
+	.port_start	= ata_port_start,
+	.port_stop	= ata_port_stop,
+	.host_stop	= ata_host_stop
+};
+
+/**
+ *	via_init_one		-	discovery callback
+ *	@pdev: PCI device ID
+ *	@id: PCI table info
+ *
+ *	A VIA IDE interface has been discovered. Figure out what revision
+ *	and perform configuration work before handing it to the ATA layer
+ */
+
+static int via_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+	/* Early VIA without UDMA support */
+	static struct ata_port_info via_mwdma_info = {
+		.sht = &via_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.port_ops = &via_port_ops
+	};
+	/* Ditto with IRQ masking required */
+	static struct ata_port_info via_mwdma_info_borked = {
+		.sht = &via_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.port_ops = &via_port_ops_noirq,
+	};
+	/* VIA UDMA 33 devices (and borked 66) */
+	static struct ata_port_info via_udma33_info = {
+		.sht = &via_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x7,
+		.port_ops = &via_port_ops
+	};
+	/* VIA UDMA 66 devices */
+	static struct ata_port_info via_udma66_info = {
+		.sht = &via_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x1f,
+		.port_ops = &via_port_ops
+	};
+	/* VIA UDMA 100 devices */
+	static struct ata_port_info via_udma100_info = {
+		.sht = &via_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x3f,
+		.port_ops = &via_port_ops
+	};
+	/* UDMA133 with bad AST (All current 133) */
+	static struct ata_port_info via_udma133_info = {
+		.sht = &via_sht,
+		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = 0x7f,	/* FIXME: should check north bridge */
+		.port_ops = &via_port_ops
+	};
+	struct ata_port_info *port_info[2], *type;
+	struct pci_dev *isa = NULL;
+	const struct via_isa_bridge *config;
+	static int printed_version;
+	u8 t;
+	u8 enable;
+	u32 timing;
+
+	if (!printed_version++)
+		dev_printk(KERN_DEBUG, &pdev->dev, "version " DRV_VERSION "\n");
+
+	/* To find out how the IDE will behave and what features we
+	   actually have to look at the bridge not the IDE controller */
+	for (config = via_isa_bridges; config->id; config++)
+		if ((isa = pci_get_device(PCI_VENDOR_ID_VIA +
+			!!(config->flags & VIA_BAD_ID),
+			config->id, NULL))) {
+
+			pci_read_config_byte(isa, PCI_REVISION_ID, &t);
+			if (t >= config->rev_min &&
+			    t <= config->rev_max)
+				break;
+			pci_dev_put(isa);
+		}
+
+	if (!config->id) {
+		printk(KERN_WARNING "via: Unknown VIA SouthBridge, disabling.\n");
+		return -ENODEV;
+	}
+	pci_dev_put(isa);
+
+	/* 0x40 low bits indicate enabled channels */
+	pci_read_config_byte(pdev, 0x40 , &enable);
+	enable &= 3;
+	if (enable == 0) {
+		return -ENODEV;
+	}
+
+	/* Initialise the FIFO for the enabled channels. */
+	if (config->flags & VIA_SET_FIFO) {
+		u8 fifo_setting[4] = {0x00, 0x60, 0x00, 0x20};
+		u8 fifo;
+
+		pci_read_config_byte(pdev, 0x43, &fifo);
+
+		/* Clear PREQ# until DDACK# for errata */
+		if (config->flags & VIA_BAD_PREQ)
+			fifo &= 0x7F;
+		else
+			fifo &= 0x9f;
+		/* Turn on FIFO for enabled channels */
+		fifo |= fifo_setting[enable];
+		pci_write_config_byte(pdev, 0x43, fifo);
+	}
+	/* Clock set up */
+	switch(config->flags & VIA_UDMA) {
+		case VIA_UDMA_NONE:
+			if (config->flags & VIA_NO_UNMASK)
+				type = &via_mwdma_info_borked;
+			else
+				type = &via_mwdma_info;
+			break;
+		case VIA_UDMA_33:
+			type = &via_udma33_info;
+			break;
+		case VIA_UDMA_66:
+			type = &via_udma66_info;
+			/* The 66 MHz devices require we enable the clock */
+			pci_read_config_dword(pdev, 0x50, &timing);
+			timing |= 0x80008;
+			pci_write_config_dword(pdev, 0x50, timing);
+			break;
+		case VIA_UDMA_100:
+			type = &via_udma100_info;
+			break;
+		case VIA_UDMA_133:
+			type = &via_udma133_info;
+			break;
+		default:
+			WARN_ON(1);
+			return -ENODEV;
+	}
+
+	if (config->flags & VIA_BAD_CLK66) {
+		/* Disable the 66MHz clock on problem devices */
+		pci_read_config_dword(pdev, 0x50, &timing);
+		timing &= ~0x80008;
+		pci_write_config_dword(pdev, 0x50, timing);
+	}
+
+	/* We have established the device type, now fire it up */
+	type->private_data = (void *)config;
+
+	port_info[0] = port_info[1] = type;
+	return ata_pci_init_one(pdev, port_info, 2);
+}
+
+static const struct pci_device_id via[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_82C576_1), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_82C586_1), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_6410), },
+	{ PCI_DEVICE(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_SATA_EIDE), },
+	{ 0, },
+};
+
+static struct pci_driver via_pci_driver = {
+        .name 		= DRV_NAME,
+	.id_table	= via,
+	.probe 		= via_init_one,
+	.remove		= ata_pci_remove_one
+};
+
+static int __init via_init(void)
+{
+	return pci_register_driver(&via_pci_driver);
+}
+
+
+static void __exit via_exit(void)
+{
+	pci_unregister_driver(&via_pci_driver);
+}
+
+
+MODULE_AUTHOR("Alan Cox");
+MODULE_DESCRIPTION("low-level driver for VIA PATA");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, via);
+MODULE_VERSION(DRV_VERSION);
+
+module_init(via_init);
+module_exit(via_exit);
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 563885cb0995..a6d818148ae0 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -46,7 +46,7 @@
 #undef ATA_VERBOSE_DEBUG	/* yet more debugging output */
 #undef ATA_IRQ_TRAP		/* define to ack screaming irqs */
 #undef ATA_NDEBUG		/* define to disable quick runtime checks */
-#undef ATA_ENABLE_PATA		/* define to enable PATA support in some
+#define ATA_ENABLE_PATA		/* define to enable PATA support in some
 				 * low-level drivers */
 
 
-- 
cgit v1.2.3


From 187afbed1814ea0851bf30bacbf807217dd7864b Mon Sep 17 00:00:00 2001
From: Jon Masters <jcm@redhat.com>
Date: Mon, 28 Aug 2006 17:08:21 -0500
Subject: [SCSI] MODULE_FIRMWARE for binary firmware(s)

Right now, various kernel modules are being migrated over to use
request_firmware in order to pull in binary firmware blobs from userland
when the module is loaded. This makes sense.

However, there is right now little mechanism in place to automatically
determine which binary firmware blobs must be included with a kernel in
order to satisfy the prerequisites of these drivers. This affects
vendors, but also regular users to a certain extent too.

The attached patch introduces MODULE_FIRMWARE as a mechanism for
advertising that a particular firmware file is to be loaded - it will
then show up via modinfo and could be used e.g. when packaging a kernel.

Signed-off-by: Jon Masters <jcm@redhat.com>

Comments added in line with all the other MODULE_ tag

Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 include/linux/module.h | 5 +++++
 1 file changed, 5 insertions(+)

(limited to 'include')

diff --git a/include/linux/module.h b/include/linux/module.h
index 0dfb794c52d3..d4486cc2e7fe 100644
--- a/include/linux/module.h
+++ b/include/linux/module.h
@@ -156,6 +156,11 @@ extern struct module __this_module;
 */
 #define MODULE_VERSION(_version) MODULE_INFO(version, _version)
 
+/* Optional firmware file (or files) needed by the module
+ * format is simply firmware file name.  Multiple firmware
+ * files require multiple MODULE_FIRMWARE() specifiers */
+#define MODULE_FIRMWARE(_firmware) MODULE_INFO(firmware, _firmware)
+
 /* Given an address, look for it in the exception tables */
 const struct exception_table_entry *search_exception_tables(unsigned long add);
 
-- 
cgit v1.2.3


From 9bec2e38527a9f2497b3d976715c672d08d6160d Mon Sep 17 00:00:00 2001
From: Jeff Garzik <jeff@garzik.org>
Date: Thu, 31 Aug 2006 00:02:15 -0400
Subject: [libata] Trim trailing whitespace.

---
 drivers/ata/ahci.c     | 2 +-
 drivers/ata/pdc_adma.c | 2 +-
 include/linux/ata.h    | 4 ++--
 include/linux/libata.h | 2 +-
 4 files changed, 5 insertions(+), 5 deletions(-)

(limited to 'include')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 3f1106fdaed1..1aabc81d82f1 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -1075,7 +1075,7 @@ static void ahci_host_intr(struct ata_port *ap)
 		return;
 
 	/* ignore interim PIO setup fis interrupts */
-	if (ata_tag_valid(ap->active_tag) && (status & PORT_IRQ_PIOS_FIS)) 
+	if (ata_tag_valid(ap->active_tag) && (status & PORT_IRQ_PIOS_FIS))
 		return;
 
 	if (ata_ratelimit())
diff --git a/drivers/ata/pdc_adma.c b/drivers/ata/pdc_adma.c
index 912211ada816..0e23ecb77bc2 100644
--- a/drivers/ata/pdc_adma.c
+++ b/drivers/ata/pdc_adma.c
@@ -497,7 +497,7 @@ static inline unsigned int adma_intr_mmio(struct ata_host *host)
 					continue;
 				DPRINTK("ata%u: protocol %d (dev_stat 0x%X)\n",
 					ap->id, qc->tf.protocol, status);
-		
+
 				/* complete taskfile transaction */
 				pp->state = adma_state_idle;
 				qc->err_mask |= ac_err_mask(status);
diff --git a/include/linux/ata.h b/include/linux/ata.h
index 991b858acc30..d89441907024 100644
--- a/include/linux/ata.h
+++ b/include/linux/ata.h
@@ -320,8 +320,8 @@ static inline unsigned int ata_id_major_version(const u16 *id)
 
 static inline int ata_id_current_chs_valid(const u16 *id)
 {
-	/* For ATA-1 devices, if the INITIALIZE DEVICE PARAMETERS command 
-	   has not been issued to the device then the values of 
+	/* For ATA-1 devices, if the INITIALIZE DEVICE PARAMETERS command
+	   has not been issued to the device then the values of
 	   id[54] to id[56] are vendor specific. */
 	return (id[53] & 0x01) && /* Current translation valid */
 		id[54] &&  /* cylinders in current translation */
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 563885cb0995..0ddf16c17b93 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -198,7 +198,7 @@ enum {
 
 	/* host set flags */
 	ATA_HOST_SIMPLEX	= (1 << 0),	/* Host is simplex, one DMA channel per host only */
-	
+
 	/* various lengths of time */
 	ATA_TMOUT_BOOT		= 30 * HZ,	/* heuristic */
 	ATA_TMOUT_BOOT_QUICK	= 7 * HZ,	/* heuristic */
-- 
cgit v1.2.3


From 492dfb489658dfe4a755fa29dd0e34e9c8bd8fb8 Mon Sep 17 00:00:00 2001
From: James Bottomley <James.Bottomley@steeleye.com>
Date: Wed, 30 Aug 2006 15:48:45 -0400
Subject: [SCSI] block: add support for shared tag maps

The current block queue implementation already contains most of the
machinery for shared tag maps.  The only remaining pieces are a way to
allocate and destroy a tag map independently of the queues (so that
the maps can be managed on the life cycle of the overseeing entity)

Acked-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 block/ll_rw_blk.c      | 109 +++++++++++++++++++++++++++++++++++++++----------
 include/linux/blkdev.h |   2 +
 2 files changed, 90 insertions(+), 21 deletions(-)

(limited to 'include')

diff --git a/block/ll_rw_blk.c b/block/ll_rw_blk.c
index ddd9253f9d55..556a3d354eab 100644
--- a/block/ll_rw_blk.c
+++ b/block/ll_rw_blk.c
@@ -848,21 +848,18 @@ struct request *blk_queue_find_tag(request_queue_t *q, int tag)
 EXPORT_SYMBOL(blk_queue_find_tag);
 
 /**
- * __blk_queue_free_tags - release tag maintenance info
- * @q:  the request queue for the device
+ * __blk_free_tags - release a given set of tag maintenance info
+ * @bqt:	the tag map to free
  *
- *  Notes:
- *    blk_cleanup_queue() will take care of calling this function, if tagging
- *    has been used. So there's no need to call this directly.
- **/
-static void __blk_queue_free_tags(request_queue_t *q)
+ * Tries to free the specified @bqt@.  Returns true if it was
+ * actually freed and false if there are still references using it
+ */
+static int __blk_free_tags(struct blk_queue_tag *bqt)
 {
-	struct blk_queue_tag *bqt = q->queue_tags;
-
-	if (!bqt)
-		return;
+	int retval;
 
-	if (atomic_dec_and_test(&bqt->refcnt)) {
+	retval = atomic_dec_and_test(&bqt->refcnt);
+	if (retval) {
 		BUG_ON(bqt->busy);
 		BUG_ON(!list_empty(&bqt->busy_list));
 
@@ -873,12 +870,49 @@ static void __blk_queue_free_tags(request_queue_t *q)
 		bqt->tag_map = NULL;
 
 		kfree(bqt);
+
 	}
 
+	return retval;
+}
+
+/**
+ * __blk_queue_free_tags - release tag maintenance info
+ * @q:  the request queue for the device
+ *
+ *  Notes:
+ *    blk_cleanup_queue() will take care of calling this function, if tagging
+ *    has been used. So there's no need to call this directly.
+ **/
+static void __blk_queue_free_tags(request_queue_t *q)
+{
+	struct blk_queue_tag *bqt = q->queue_tags;
+
+	if (!bqt)
+		return;
+
+	__blk_free_tags(bqt);
+
 	q->queue_tags = NULL;
 	q->queue_flags &= ~(1 << QUEUE_FLAG_QUEUED);
 }
 
+
+/**
+ * blk_free_tags - release a given set of tag maintenance info
+ * @bqt:	the tag map to free
+ *
+ * For externally managed @bqt@ frees the map.  Callers of this
+ * function must guarantee to have released all the queues that
+ * might have been using this tag map.
+ */
+void blk_free_tags(struct blk_queue_tag *bqt)
+{
+	if (unlikely(!__blk_free_tags(bqt)))
+		BUG();
+}
+EXPORT_SYMBOL(blk_free_tags);
+
 /**
  * blk_queue_free_tags - release tag maintenance info
  * @q:  the request queue for the device
@@ -901,7 +935,7 @@ init_tag_map(request_queue_t *q, struct blk_queue_tag *tags, int depth)
 	unsigned long *tag_map;
 	int nr_ulongs;
 
-	if (depth > q->nr_requests * 2) {
+	if (q && depth > q->nr_requests * 2) {
 		depth = q->nr_requests * 2;
 		printk(KERN_ERR "%s: adjusted depth to %d\n",
 				__FUNCTION__, depth);
@@ -927,6 +961,38 @@ fail:
 	return -ENOMEM;
 }
 
+static struct blk_queue_tag *__blk_queue_init_tags(struct request_queue *q,
+						   int depth)
+{
+	struct blk_queue_tag *tags;
+
+	tags = kmalloc(sizeof(struct blk_queue_tag), GFP_ATOMIC);
+	if (!tags)
+		goto fail;
+
+	if (init_tag_map(q, tags, depth))
+		goto fail;
+
+	INIT_LIST_HEAD(&tags->busy_list);
+	tags->busy = 0;
+	atomic_set(&tags->refcnt, 1);
+	return tags;
+fail:
+	kfree(tags);
+	return NULL;
+}
+
+/**
+ * blk_init_tags - initialize the tag info for an external tag map
+ * @depth:	the maximum queue depth supported
+ * @tags: the tag to use
+ **/
+struct blk_queue_tag *blk_init_tags(int depth)
+{
+	return __blk_queue_init_tags(NULL, depth);
+}
+EXPORT_SYMBOL(blk_init_tags);
+
 /**
  * blk_queue_init_tags - initialize the queue tag info
  * @q:  the request queue for the device
@@ -941,16 +1007,10 @@ int blk_queue_init_tags(request_queue_t *q, int depth,
 	BUG_ON(tags && q->queue_tags && tags != q->queue_tags);
 
 	if (!tags && !q->queue_tags) {
-		tags = kmalloc(sizeof(struct blk_queue_tag), GFP_ATOMIC);
-		if (!tags)
-			goto fail;
+		tags = __blk_queue_init_tags(q, depth);
 
-		if (init_tag_map(q, tags, depth))
+		if (!tags)
 			goto fail;
-
-		INIT_LIST_HEAD(&tags->busy_list);
-		tags->busy = 0;
-		atomic_set(&tags->refcnt, 1);
 	} else if (q->queue_tags) {
 		if ((rc = blk_queue_resize_tags(q, depth)))
 			return rc;
@@ -1001,6 +1061,13 @@ int blk_queue_resize_tags(request_queue_t *q, int new_depth)
 		return 0;
 	}
 
+	/*
+	 * Currently cannot replace a shared tag map with a new
+	 * one, so error out if this is the case
+	 */
+	if (atomic_read(&bqt->refcnt) != 1)
+		return -EBUSY;
+
 	/*
 	 * save the old state info, so we can copy it back
 	 */
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index aafe82788b4e..427b0d61be6c 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -746,6 +746,8 @@ extern void blk_queue_free_tags(request_queue_t *);
 extern int blk_queue_resize_tags(request_queue_t *, int);
 extern void blk_queue_invalidate_tags(request_queue_t *);
 extern long blk_congestion_wait(int rw, long timeout);
+extern struct blk_queue_tag *blk_init_tags(int);
+extern void blk_free_tags(struct blk_queue_tag *);
 
 extern void blk_rq_bio_prep(request_queue_t *, struct request *, struct bio *);
 extern int blkdev_issue_flush(struct block_device *, sector_t *);
-- 
cgit v1.2.3


From 86e33a296c2c9ed6eece0bfff4ac776f42040504 Mon Sep 17 00:00:00 2001
From: James Bottomley <James.Bottomley@steeleye.com>
Date: Wed, 30 Aug 2006 09:45:51 -0400
Subject: [SCSI] add shared tag map helpers

This patch adds support for sharing tag maps at the host level
(i.e. either every queue [LUN] has its own tag map or there's a single
one for the entire host).  This formulation is primarily intended to
help single issue queue hardware, like the aic7xxx

Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 drivers/scsi/hosts.c     |  3 +++
 include/scsi/scsi_host.h |  7 +++++++
 include/scsi/scsi_tcq.h  | 14 +++++++++++++-
 3 files changed, 23 insertions(+), 1 deletion(-)

(limited to 'include')

diff --git a/drivers/scsi/hosts.c b/drivers/scsi/hosts.c
index f244d4f6597a..68ef1636678d 100644
--- a/drivers/scsi/hosts.c
+++ b/drivers/scsi/hosts.c
@@ -265,6 +265,9 @@ static void scsi_host_dev_release(struct device *dev)
 		destroy_workqueue(shost->work_q);
 
 	scsi_destroy_command_freelist(shost);
+	if (shost->bqt)
+		blk_free_tags(shost->bqt);
+
 	kfree(shost->shost_data);
 
 	if (parent)
diff --git a/include/scsi/scsi_host.h b/include/scsi/scsi_host.h
index b3dd90f3e858..39c6f8cc20c3 100644
--- a/include/scsi/scsi_host.h
+++ b/include/scsi/scsi_host.h
@@ -16,6 +16,7 @@ struct scsi_target;
 struct Scsi_Host;
 struct scsi_host_cmd_pool;
 struct scsi_transport_template;
+struct blk_queue_tags;
 
 
 /*
@@ -465,6 +466,12 @@ struct Scsi_Host {
 	struct scsi_host_template *hostt;
 	struct scsi_transport_template *transportt;
 
+	/*
+	 * area to keep a shared tag map (if needed, will be
+	 * NULL if not)
+	 */
+	struct blk_queue_tag	*bqt;
+
 	/*
 	 * The following two fields are protected with host_lock;
 	 * however, eh routines can safely access during eh processing
diff --git a/include/scsi/scsi_tcq.h b/include/scsi/scsi_tcq.h
index e47e36a4ef49..4eea254b1ce9 100644
--- a/include/scsi/scsi_tcq.h
+++ b/include/scsi/scsi_tcq.h
@@ -4,6 +4,7 @@
 #include <linux/blkdev.h>
 #include <scsi/scsi_cmnd.h>
 #include <scsi/scsi_device.h>
+#include <scsi/scsi_host.h>
 
 
 #define MSG_SIMPLE_TAG	0x20
@@ -66,7 +67,8 @@ static inline void scsi_activate_tcq(struct scsi_device *sdev, int depth)
 		return;
 
 	if (!blk_queue_tagged(sdev->request_queue))
-		blk_queue_init_tags(sdev->request_queue, depth, NULL);
+		blk_queue_init_tags(sdev->request_queue, depth,
+				    sdev->host->bqt);
 
 	scsi_adjust_queue_depth(sdev, scsi_get_tag_type(sdev), depth);
 }
@@ -131,4 +133,14 @@ static inline struct scsi_cmnd *scsi_find_tag(struct scsi_device *sdev, int tag)
 	return sdev->current_cmnd;
 }
 
+/**
+ * scsi_init_shared_tag_map - create a shared tag map
+ * @shost:	the host to share the tag map among all devices
+ * @depth:	the total depth of the map
+ */
+static inline void scsi_init_shared_tag_map(struct Scsi_Host *shost, int depth)
+{
+	shost->bqt = blk_init_tags(depth);
+}
+
 #endif /* _SCSI_SCSI_TCQ_H */
-- 
cgit v1.2.3


From ffd0436ed2e5a741c8d30062b489b989acf0a526 Mon Sep 17 00:00:00 2001
From: Mike Christie <michaelc@cs.wisc.edu>
Date: Thu, 31 Aug 2006 18:09:24 -0400
Subject: [SCSI] libiscsi, iscsi_tcp, iscsi_iser: check that burst lengths are
 valid.

iSCSI RFC states that the first burst length must be smaller than the
max burst length. We currently assume targets will be good, but that may
not be the case, so this patch adds a check.

This patch also moves the unsol data out offset to the lib so the LLDs
do not have to track it.

Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 drivers/infiniband/ulp/iser/iscsi_iser.c | 18 ++++--------------
 drivers/infiniband/ulp/iser/iscsi_iser.h |  1 -
 drivers/scsi/iscsi_tcp.c                 | 30 +++++++-----------------------
 drivers/scsi/iscsi_tcp.h                 |  1 -
 drivers/scsi/libiscsi.c                  | 25 +++++++++++++++++--------
 include/scsi/libiscsi.h                  |  5 +++--
 6 files changed, 31 insertions(+), 49 deletions(-)

(limited to 'include')

diff --git a/drivers/infiniband/ulp/iser/iscsi_iser.c b/drivers/infiniband/ulp/iser/iscsi_iser.c
index 1437d7ee3b19..101e407eaa43 100644
--- a/drivers/infiniband/ulp/iser/iscsi_iser.c
+++ b/drivers/infiniband/ulp/iser/iscsi_iser.c
@@ -141,18 +141,11 @@ iscsi_iser_cmd_init(struct iscsi_cmd_task *ctask)
 
 	if (sc->sc_data_direction == DMA_TO_DEVICE) {
 		BUG_ON(ctask->total_length == 0);
-		/* bytes to be sent via RDMA operations */
-		iser_ctask->rdma_data_count = ctask->total_length -
-					 ctask->imm_count -
-					 ctask->unsol_count;
 
-		debug_scsi("cmd [itt %x total %d imm %d unsol_data %d "
-			   "rdma_data %d]\n",
+		debug_scsi("cmd [itt %x total %d imm %d unsol_data %d\n",
 			   ctask->itt, ctask->total_length, ctask->imm_count,
-			   ctask->unsol_count, iser_ctask->rdma_data_count);
-	} else
-		/* bytes to be sent via RDMA operations */
-		iser_ctask->rdma_data_count = ctask->total_length;
+			   ctask->unsol_count);
+	}
 
 	iser_ctask_rdma_init(iser_ctask);
 }
@@ -196,13 +189,10 @@ iscsi_iser_ctask_xmit_unsol_data(struct iscsi_conn *conn,
 {
 	struct iscsi_data  hdr;
 	int error = 0;
-	struct iscsi_iser_cmd_task *iser_ctask = ctask->dd_data;
 
 	/* Send data-out PDUs while there's still unsolicited data to send */
 	while (ctask->unsol_count > 0) {
-		iscsi_prep_unsolicit_data_pdu(ctask, &hdr,
-					      iser_ctask->rdma_data_count);
-
+		iscsi_prep_unsolicit_data_pdu(ctask, &hdr);
 		debug_scsi("Sending data-out: itt 0x%x, data count %d\n",
 			   hdr.itt, ctask->data_count);
 
diff --git a/drivers/infiniband/ulp/iser/iscsi_iser.h b/drivers/infiniband/ulp/iser/iscsi_iser.h
index 3350ba690cfe..7c3d0c96d889 100644
--- a/drivers/infiniband/ulp/iser/iscsi_iser.h
+++ b/drivers/infiniband/ulp/iser/iscsi_iser.h
@@ -257,7 +257,6 @@ struct iscsi_iser_conn {
 struct iscsi_iser_cmd_task {
 	struct iser_desc             desc;
 	struct iscsi_iser_conn	     *iser_conn;
-	int			     rdma_data_count;/* RDMA bytes           */
 	enum iser_task_status 	     status;
 	int                          command_sent;  /* set if command  sent  */
 	int                          dir[ISER_DIRS_NUM];      /* set if dir use*/
diff --git a/drivers/scsi/iscsi_tcp.c b/drivers/scsi/iscsi_tcp.c
index 058f094f945a..a97a3a4e99eb 100644
--- a/drivers/scsi/iscsi_tcp.c
+++ b/drivers/scsi/iscsi_tcp.c
@@ -1264,19 +1264,6 @@ iscsi_solicit_data_cont(struct iscsi_conn *conn, struct iscsi_cmd_task *ctask,
 			    r2t->data_count);
 }
 
-static void
-iscsi_unsolicit_data_init(struct iscsi_conn *conn, struct iscsi_cmd_task *ctask)
-{
-	struct iscsi_tcp_cmd_task *tcp_ctask = ctask->dd_data;
-	struct iscsi_data_task *dtask;
-
-	dtask = tcp_ctask->dtask = &tcp_ctask->unsol_dtask;
-	iscsi_prep_unsolicit_data_pdu(ctask, &dtask->hdr,
-				      tcp_ctask->r2t_data_count);
-	iscsi_buf_init_iov(&tcp_ctask->headbuf, (char*)&dtask->hdr,
-			   sizeof(struct iscsi_hdr));
-}
-
 /**
  * iscsi_tcp_cmd_init - Initialize iSCSI SCSI_READ or SCSI_WRITE commands
  * @conn: iscsi connection
@@ -1326,14 +1313,11 @@ iscsi_tcp_cmd_init(struct iscsi_cmd_task *ctask)
 		if (ctask->unsol_count)
 			tcp_ctask->xmstate |= XMSTATE_UNS_HDR |
 						XMSTATE_UNS_INIT;
-		tcp_ctask->r2t_data_count = ctask->total_length -
-				    ctask->imm_count -
-				    ctask->unsol_count;
 
-		debug_scsi("cmd [itt 0x%x total %d imm %d imm_data %d "
-			   "r2t_data %d]\n",
+		debug_scsi("cmd [itt 0x%x total %d imm_data %d "
+			   "unsol count %d, unsol offset %d]\n",
 			   ctask->itt, ctask->total_length, ctask->imm_count,
-			   ctask->unsol_count, tcp_ctask->r2t_data_count);
+			   ctask->unsol_count, ctask->unsol_offset);
 	} else
 		tcp_ctask->xmstate = XMSTATE_R_HDR;
 
@@ -1531,8 +1515,10 @@ handle_xmstate_uns_hdr(struct iscsi_conn *conn, struct iscsi_cmd_task *ctask)
 
 	tcp_ctask->xmstate |= XMSTATE_UNS_DATA;
 	if (tcp_ctask->xmstate & XMSTATE_UNS_INIT) {
-		iscsi_unsolicit_data_init(conn, ctask);
-		dtask = tcp_ctask->dtask;
+		dtask = tcp_ctask->dtask = &tcp_ctask->unsol_dtask;
+		iscsi_prep_unsolicit_data_pdu(ctask, &dtask->hdr);
+		iscsi_buf_init_iov(&tcp_ctask->headbuf, (char*)&dtask->hdr,
+				   sizeof(struct iscsi_hdr));
 		if (conn->hdrdgst_en)
 			iscsi_hdr_digest(conn, &tcp_ctask->headbuf,
 					(u8*)dtask->hdrext);
@@ -1720,7 +1706,6 @@ data_out_done:
 	 * Done with this R2T. Check if there are more
 	 * outstanding R2Ts ready to be processed.
 	 */
-	BUG_ON(tcp_ctask->r2t_data_count - r2t->data_length < 0);
 	if (conn->datadgst_en) {
 		rc = iscsi_digest_final_send(conn, ctask, &dtask->digestbuf,
 					    &dtask->digest, 1);
@@ -1732,7 +1717,6 @@ data_out_done:
 		debug_tcp("r2t done dout digest 0x%x\n", dtask->digest);
 	}
 
-	tcp_ctask->r2t_data_count -= r2t->data_length;
 	tcp_ctask->r2t = NULL;
 	spin_lock_bh(&session->lock);
 	__kfifo_put(tcp_ctask->r2tpool.queue, (void*)&r2t, sizeof(void*));
diff --git a/drivers/scsi/iscsi_tcp.h b/drivers/scsi/iscsi_tcp.h
index 6a4ee704e46e..aace8f70dfd7 100644
--- a/drivers/scsi/iscsi_tcp.h
+++ b/drivers/scsi/iscsi_tcp.h
@@ -157,7 +157,6 @@ struct iscsi_tcp_cmd_task {
 	struct scatterlist	*bad_sg;		/* assert statement */
 	int			sg_count;		/* SG's to process  */
 	uint32_t		exp_r2tsn;
-	int			r2t_data_count;		/* R2T Data-Out bytes */
 	int			data_offset;
 	struct iscsi_r2t_info	*r2t;			/* in progress R2T    */
 	struct iscsi_queue	r2tpool;
diff --git a/drivers/scsi/libiscsi.c b/drivers/scsi/libiscsi.c
index 5884cd26d53a..a7c6e70f4ef8 100644
--- a/drivers/scsi/libiscsi.c
+++ b/drivers/scsi/libiscsi.c
@@ -68,8 +68,7 @@ iscsi_check_assign_cmdsn(struct iscsi_session *session, struct iscsi_nopin *hdr)
 EXPORT_SYMBOL_GPL(iscsi_check_assign_cmdsn);
 
 void iscsi_prep_unsolicit_data_pdu(struct iscsi_cmd_task *ctask,
-				   struct iscsi_data *hdr,
-				   int transport_data_cnt)
+				   struct iscsi_data *hdr)
 {
 	struct iscsi_conn *conn = ctask->conn;
 
@@ -82,14 +81,12 @@ void iscsi_prep_unsolicit_data_pdu(struct iscsi_cmd_task *ctask,
 
 	hdr->itt = ctask->hdr->itt;
 	hdr->exp_statsn = cpu_to_be32(conn->exp_statsn);
-
-	hdr->offset = cpu_to_be32(ctask->total_length -
-				  transport_data_cnt -
-				  ctask->unsol_count);
+	hdr->offset = cpu_to_be32(ctask->unsol_offset);
 
 	if (ctask->unsol_count > conn->max_xmit_dlength) {
 		hton24(hdr->dlength, conn->max_xmit_dlength);
 		ctask->data_count = conn->max_xmit_dlength;
+		ctask->unsol_offset += ctask->data_count;
 		hdr->flags = 0;
 	} else {
 		hton24(hdr->dlength, ctask->unsol_count);
@@ -125,6 +122,7 @@ static void iscsi_prep_scsi_cmd_pdu(struct iscsi_cmd_task *ctask)
         memcpy(hdr->cdb, sc->cmnd, sc->cmd_len);
         memset(&hdr->cdb[sc->cmd_len], 0, MAX_COMMAND_SIZE - sc->cmd_len);
 
+	ctask->data_count = 0;
 	if (sc->sc_data_direction == DMA_TO_DEVICE) {
 		hdr->flags |= ISCSI_FLAG_CMD_WRITE;
 		/*
@@ -143,6 +141,7 @@ static void iscsi_prep_scsi_cmd_pdu(struct iscsi_cmd_task *ctask)
 		 */
 		ctask->imm_count = 0;
 		ctask->unsol_count = 0;
+		ctask->unsol_offset = 0;
 		ctask->unsol_datasn = 0;
 
 		if (session->imm_data_en) {
@@ -156,9 +155,12 @@ static void iscsi_prep_scsi_cmd_pdu(struct iscsi_cmd_task *ctask)
 		} else
 			zero_data(ctask->hdr->dlength);
 
-		if (!session->initial_r2t_en)
+		if (!session->initial_r2t_en) {
 			ctask->unsol_count = min(session->first_burst,
 				ctask->total_length) - ctask->imm_count;
+			ctask->unsol_offset = ctask->imm_count;
+		}
+
 		if (!ctask->unsol_count)
 			/* No unsolicit Data-Out's */
 			ctask->hdr->flags |= ISCSI_FLAG_CMD_FINAL;
@@ -1520,11 +1522,18 @@ int iscsi_conn_start(struct iscsi_cls_conn *cls_conn)
 	struct iscsi_conn *conn = cls_conn->dd_data;
 	struct iscsi_session *session = conn->session;
 
-	if (session == NULL) {
+	if (!session) {
 		printk(KERN_ERR "iscsi: can't start unbound connection\n");
 		return -EPERM;
 	}
 
+	if (session->first_burst > session->max_burst) {
+		printk("iscsi: invalid burst lengths: "
+		       "first_burst %d max_burst %d\n",
+		       session->first_burst, session->max_burst);
+		return -EINVAL;
+	}
+
 	spin_lock_bh(&session->lock);
 	conn->c_stage = ISCSI_CONN_STARTED;
 	session->state = ISCSI_STATE_LOGGED_IN;
diff --git a/include/scsi/libiscsi.h b/include/scsi/libiscsi.h
index 41904f611d12..4900650bd081 100644
--- a/include/scsi/libiscsi.h
+++ b/include/scsi/libiscsi.h
@@ -102,6 +102,8 @@ struct iscsi_cmd_task {
 	uint32_t		unsol_datasn;
 	int			imm_count;	/* imm-data (bytes)   */
 	int			unsol_count;	/* unsolicited (bytes)*/
+	/* offset in unsolicited stream (bytes); */
+	int			unsol_offset;
 	int			data_count;	/* remaining Data-Out */
 	struct scsi_cmnd	*sc;		/* associated SCSI cmd*/
 	int			total_length;
@@ -290,8 +292,7 @@ extern int iscsi_conn_get_param(struct iscsi_cls_conn *cls_conn,
 extern int iscsi_check_assign_cmdsn(struct iscsi_session *,
 				    struct iscsi_nopin *);
 extern void iscsi_prep_unsolicit_data_pdu(struct iscsi_cmd_task *,
-					struct iscsi_data *hdr,
-					int transport_data_cnt);
+					struct iscsi_data *hdr);
 extern int iscsi_conn_send_pdu(struct iscsi_cls_conn *, struct iscsi_hdr *,
 				char *, uint32_t);
 extern int iscsi_complete_pdu(struct iscsi_conn *, struct iscsi_hdr *,
-- 
cgit v1.2.3


From 60ecebf5a10e42f5e2d6e07eb9e24bdee8500b81 Mon Sep 17 00:00:00 2001
From: Mike Christie <michaelc@cs.wisc.edu>
Date: Thu, 31 Aug 2006 18:09:25 -0400
Subject: [SCSI] add refcouting around ctask usage in main IO patch

It is possible that a ctask could be completing and getting
cleaned up at the same time, we are finishing up the last
data transfer. This could then result in the data transfer
code using stale or invalid values. This patch adds a refcount
to the ctask. When the count goes to zero then we know the
transmit thread and recv thread or softirq are not touching
it and we can safely release it.

The eh should not need to grab a reference because it only cleans
up a task if it has both the xmit mutex and recv lock (or recv
side suspended).

Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 drivers/scsi/libiscsi.c | 59 +++++++++++++++++++++++++++++++++++++++++--------
 include/scsi/libiscsi.h |  1 +
 2 files changed, 51 insertions(+), 9 deletions(-)

(limited to 'include')

diff --git a/drivers/scsi/libiscsi.c b/drivers/scsi/libiscsi.c
index a7c6e70f4ef8..9584cbc082fe 100644
--- a/drivers/scsi/libiscsi.c
+++ b/drivers/scsi/libiscsi.c
@@ -179,16 +179,15 @@ EXPORT_SYMBOL_GPL(iscsi_prep_scsi_cmd_pdu);
 
 /**
  * iscsi_complete_command - return command back to scsi-ml
- * @session: iscsi session
  * @ctask: iscsi cmd task
  *
  * Must be called with session lock.
  * This function returns the scsi command to scsi-ml and returns
  * the cmd task to the pool of available cmd tasks.
  */
-static void iscsi_complete_command(struct iscsi_session *session,
-				   struct iscsi_cmd_task *ctask)
+static void iscsi_complete_command(struct iscsi_cmd_task *ctask)
 {
+	struct iscsi_session *session = ctask->conn->session;
 	struct scsi_cmnd *sc = ctask->sc;
 
 	ctask->state = ISCSI_TASK_COMPLETED;
@@ -198,6 +197,35 @@ static void iscsi_complete_command(struct iscsi_session *session,
 	sc->scsi_done(sc);
 }
 
+static void __iscsi_get_ctask(struct iscsi_cmd_task *ctask)
+{
+	atomic_inc(&ctask->refcount);
+}
+
+static void iscsi_get_ctask(struct iscsi_cmd_task *ctask)
+{
+	spin_lock_bh(&ctask->conn->session->lock);
+	__iscsi_get_ctask(ctask);
+	spin_unlock_bh(&ctask->conn->session->lock);
+}
+
+static void __iscsi_put_ctask(struct iscsi_cmd_task *ctask)
+{
+	struct iscsi_conn *conn = ctask->conn;
+
+	if (atomic_dec_and_test(&ctask->refcount)) {
+		conn->session->tt->cleanup_cmd_task(conn, ctask);
+		iscsi_complete_command(ctask);
+	}
+}
+
+static void iscsi_put_ctask(struct iscsi_cmd_task *ctask)
+{
+	spin_lock_bh(&ctask->conn->session->lock);
+	__iscsi_put_ctask(ctask);
+	spin_unlock_bh(&ctask->conn->session->lock);
+}
+
 /**
  * iscsi_cmd_rsp - SCSI Command Response processing
  * @conn: iscsi connection
@@ -274,7 +302,7 @@ out:
 		   (long)sc, sc->result, ctask->itt);
 	conn->scsirsp_pdus_cnt++;
 
-	iscsi_complete_command(conn->session, ctask);
+	__iscsi_put_ctask(ctask);
 	return rc;
 }
 
@@ -338,7 +366,7 @@ int __iscsi_complete_pdu(struct iscsi_conn *conn, struct iscsi_hdr *hdr,
 			BUG_ON((void*)ctask != ctask->sc->SCp.ptr);
 			if (hdr->flags & ISCSI_FLAG_DATA_STATUS) {
 				conn->scsirsp_pdus_cnt++;
-				iscsi_complete_command(session, ctask);
+				__iscsi_put_ctask(ctask);
 			}
 			break;
 		case ISCSI_OP_R2T:
@@ -563,7 +591,9 @@ static int iscsi_data_xmit(struct iscsi_conn *conn)
 	BUG_ON(conn->ctask && conn->mtask);
 
 	if (conn->ctask) {
+		iscsi_get_ctask(conn->ctask);
 		rc = tt->xmit_cmd_task(conn, conn->ctask);
+		iscsi_put_ctask(conn->ctask);
 		if (rc)
 			goto again;
 		/* done with this in-progress ctask */
@@ -604,12 +634,19 @@ static int iscsi_data_xmit(struct iscsi_conn *conn)
 					 struct iscsi_cmd_task, running);
 		conn->ctask->state = ISCSI_TASK_RUNNING;
 		list_move_tail(conn->xmitqueue.next, &conn->run_list);
+		__iscsi_get_ctask(conn->ctask);
 		spin_unlock_bh(&conn->session->lock);
 
 		rc = tt->xmit_cmd_task(conn, conn->ctask);
 		if (rc)
 			goto again;
+
 		spin_lock_bh(&conn->session->lock);
+		__iscsi_put_ctask(conn->ctask);
+		if (rc) {
+			spin_unlock_bh(&conn->session->lock);
+			goto again;
+		}
 	}
 	spin_unlock_bh(&conn->session->lock);
 	/* done with this ctask */
@@ -659,6 +696,7 @@ enum {
 	FAILURE_SESSION_FAILED,
 	FAILURE_SESSION_FREED,
 	FAILURE_WINDOW_CLOSED,
+	FAILURE_OOM,
 	FAILURE_SESSION_TERMINATE,
 	FAILURE_SESSION_IN_RECOVERY,
 	FAILURE_SESSION_RECOVERY_TIMEOUT,
@@ -717,10 +755,15 @@ int iscsi_queuecommand(struct scsi_cmnd *sc, void (*done)(struct scsi_cmnd *))
 
 	conn = session->leadconn;
 
-	__kfifo_get(session->cmdpool.queue, (void*)&ctask, sizeof(void*));
+	if (!__kfifo_get(session->cmdpool.queue, (void*)&ctask,
+			 sizeof(void*))) {
+		reason = FAILURE_OOM;
+		goto reject;
+	}
 	sc->SCp.phase = session->age;
 	sc->SCp.ptr = (char *)ctask;
 
+	atomic_set(&ctask->refcount, 1);
 	ctask->state = ISCSI_TASK_PENDING;
 	ctask->mtask = NULL;
 	ctask->conn = conn;
@@ -1057,13 +1100,11 @@ static void fail_command(struct iscsi_conn *conn, struct iscsi_cmd_task *ctask,
 	sc = ctask->sc;
 	if (!sc)
 		return;
-
-	conn->session->tt->cleanup_cmd_task(conn, ctask);
 	iscsi_ctask_mtask_cleanup(ctask);
 
 	sc->result = err;
 	sc->resid = sc->request_bufflen;
-	iscsi_complete_command(conn->session, ctask);
+	__iscsi_put_ctask(ctask);
 }
 
 int iscsi_eh_abort(struct scsi_cmnd *sc)
diff --git a/include/scsi/libiscsi.h b/include/scsi/libiscsi.h
index 4900650bd081..401192e56e50 100644
--- a/include/scsi/libiscsi.h
+++ b/include/scsi/libiscsi.h
@@ -112,6 +112,7 @@ struct iscsi_cmd_task {
 
 	/* state set/tested under session->lock */
 	int			state;
+	atomic_t		refcount;
 	struct list_head	running;	/* running cmd list */
 	void			*dd_data;	/* driver/transport data */
 };
-- 
cgit v1.2.3


From deb81d80ba27da8dfabc29ccb5977db8f4942a0a Mon Sep 17 00:00:00 2001
From: James Bottomley <James.Bottomley@steeleye.com>
Date: Fri, 1 Sep 2006 09:28:48 -0400
Subject: [SCSI] add failure return to scsi_init_shared_tag_map()

And use it in the stex driver.

Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 drivers/scsi/stex.c     | 5 ++---
 include/scsi/scsi_tcq.h | 3 ++-
 2 files changed, 4 insertions(+), 4 deletions(-)

(limited to 'include')

diff --git a/drivers/scsi/stex.c b/drivers/scsi/stex.c
index 15fb99f224ee..3cf3106a29b8 100644
--- a/drivers/scsi/stex.c
+++ b/drivers/scsi/stex.c
@@ -1108,9 +1108,8 @@ stex_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	if (err)
 		goto out_free_irq;
 
-	scsi_init_shared_tag_map(host, ST_CAN_QUEUE);
-	if (host->bqt == NULL) {
-		err = -ENOMEM;
+	err = scsi_init_shared_tag_map(host, ST_CAN_QUEUE);
+	if (err) {
 		printk(KERN_ERR DRV_NAME "(%s): init shared queue failed\n",
 			pci_name(pdev));
 		goto out_free_irq;
diff --git a/include/scsi/scsi_tcq.h b/include/scsi/scsi_tcq.h
index 4eea254b1ce9..d04d05adfa9b 100644
--- a/include/scsi/scsi_tcq.h
+++ b/include/scsi/scsi_tcq.h
@@ -138,9 +138,10 @@ static inline struct scsi_cmnd *scsi_find_tag(struct scsi_device *sdev, int tag)
  * @shost:	the host to share the tag map among all devices
  * @depth:	the total depth of the map
  */
-static inline void scsi_init_shared_tag_map(struct Scsi_Host *shost, int depth)
+static inline int scsi_init_shared_tag_map(struct Scsi_Host *shost, int depth)
 {
 	shost->bqt = blk_init_tags(depth);
+	return shost->bqt ? 0 : -ENOMEM;
 }
 
 #endif /* _SCSI_SCSI_TCQ_H */
-- 
cgit v1.2.3


From 84314fd4740ad73550c76dee4a9578979d84af48 Mon Sep 17 00:00:00 2001
From: James Smart <James.Smart@Emulex.Com>
Date: Fri, 18 Aug 2006 17:30:09 -0400
Subject: [SCSI] SCSI and FC Transport: add netlink support for posting of
 transport events

This patch formally adds support for the posting of FC events via netlink.
It is a followup to the original RFC at:
  http://marc.theaimsgroup.com/?l=linux-scsi&m=114530667923464&w=2
and the initial posting at:
  http://marc.theaimsgroup.com/?l=linux-scsi&m=115507374832500&w=2

The patch has been updated to optimize the send path, per the discussions
in the initial posting.

Per discussions at the Storage Summit and at OLS, we are to use netlink for
async events from transports. Also per discussions, to avoid a netlink
protocol per transport, I've create a single NETLINK_SCSITRANSPORT protocol,
which can then be used by all transports.

This patch:
- Creates new files scsi_netlink.c and scsi_netlink.h, which contains the
  single and shared definitions for the SCSI Transport. It is tied into the
  base SCSI subsystem intialization.
  Contains a single interface routine, scsi_send_transport_event(), for a
  transport to send an event (via multicast to a protocol specific group).
- Creates a new scsi_netlink_fc.h file, which contains the FC netlink event
  messages
- Adds 3 new routines to the fc transport:
   fc_get_event_number() -  to get a FC event #
   fc_host_post_event()  -  to send a simple FC event (32 bits of data)
   fc_host_post_vendor_event() - to send a Vendor unique event, with
                                 arbitrary amounts of data.

   Note: the separation of event number allows for a LLD to send a standard
     event, followed by vendor-specific data for the event.

Note: This patch assumes 2 prior fc transport patches have been installed:
   http://marc.theaimsgroup.com/?l=linux-scsi&m=115555807316329&w=2
   http://marc.theaimsgroup.com/?l=linux-scsi&m=115581614930261&w=2

   Sorry - next time I'll do something like making these individual
   patches of the same posting when I know they'll be posted closely
   together.

Signed-off-by: James Smart <James.Smart@emulex.com>

Tidy up configuration not to make SCSI always select NET

Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 drivers/scsi/Kconfig             |   6 ++
 drivers/scsi/Makefile            |   1 +
 drivers/scsi/scsi.c              |   3 +
 drivers/scsi/scsi_netlink.c      | 199 ++++++++++++++++++++++++++++++++++++++
 drivers/scsi/scsi_priv.h         |  11 +++
 drivers/scsi/scsi_transport_fc.c | 200 ++++++++++++++++++++++++++++++++++++++-
 include/linux/netlink.h          |   2 +
 include/scsi/scsi_netlink.h      |  86 +++++++++++++++++
 include/scsi/scsi_netlink_fc.h   |  71 ++++++++++++++
 include/scsi/scsi_transport_fc.h |  34 +++++++
 10 files changed, 612 insertions(+), 1 deletion(-)
 create mode 100644 drivers/scsi/scsi_netlink.c
 create mode 100644 include/scsi/scsi_netlink.h
 create mode 100644 include/scsi/scsi_netlink_fc.h

(limited to 'include')

diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig
index c8c606589ea6..4d1998d23f0f 100644
--- a/drivers/scsi/Kconfig
+++ b/drivers/scsi/Kconfig
@@ -27,6 +27,11 @@ config SCSI
 	  However, do not compile this as a module if your root file system
 	  (the one containing the directory /) is located on a SCSI device.
 
+config SCSI_NETLINK
+	tristate
+	default	n
+	select NET
+
 config SCSI_PROC_FS
 	bool "legacy /proc/scsi/ support"
 	depends on SCSI && PROC_FS
@@ -222,6 +227,7 @@ config SCSI_SPI_ATTRS
 config SCSI_FC_ATTRS
 	tristate "FiberChannel Transport Attributes"
 	depends on SCSI
+	select SCSI_NETLINK
 	help
 	  If you wish to export transport-specific information about
 	  each attached FiberChannel device to sysfs, say Y.
diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile
index fd9aeb1ba07f..8fc2c594b537 100644
--- a/drivers/scsi/Makefile
+++ b/drivers/scsi/Makefile
@@ -159,6 +159,7 @@ scsi_mod-y			+= scsi.o hosts.o scsi_ioctl.o constants.o \
 				   scsicam.o scsi_error.o scsi_lib.o \
 				   scsi_scan.o scsi_sysfs.o \
 				   scsi_devinfo.o
+scsi_mod-$(CONFIG_SCSI_NETLINK)	+= scsi_netlink.o
 scsi_mod-$(CONFIG_SYSCTL)	+= scsi_sysctl.o
 scsi_mod-$(CONFIG_SCSI_PROC_FS)	+= scsi_proc.o
 
diff --git a/drivers/scsi/scsi.c b/drivers/scsi/scsi.c
index 37843927e47f..eedfd059b82b 100644
--- a/drivers/scsi/scsi.c
+++ b/drivers/scsi/scsi.c
@@ -1118,6 +1118,8 @@ static int __init init_scsi(void)
 	for_each_possible_cpu(i)
 		INIT_LIST_HEAD(&per_cpu(scsi_done_q, i));
 
+	scsi_netlink_init();
+
 	printk(KERN_NOTICE "SCSI subsystem initialized\n");
 	return 0;
 
@@ -1138,6 +1140,7 @@ cleanup_queue:
 
 static void __exit exit_scsi(void)
 {
+	scsi_netlink_exit();
 	scsi_sysfs_unregister();
 	scsi_exit_sysctl();
 	scsi_exit_hosts();
diff --git a/drivers/scsi/scsi_netlink.c b/drivers/scsi/scsi_netlink.c
new file mode 100644
index 000000000000..1b59b27e887f
--- /dev/null
+++ b/drivers/scsi/scsi_netlink.c
@@ -0,0 +1,199 @@
+/*
+ *  scsi_netlink.c  - SCSI Transport Netlink Interface
+ *
+ *  Copyright (C) 2006   James Smart, Emulex Corporation
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ *
+ */
+#include <linux/time.h>
+#include <linux/jiffies.h>
+#include <linux/security.h>
+#include <net/sock.h>
+#include <net/netlink.h>
+
+#include <scsi/scsi_netlink.h>
+#include "scsi_priv.h"
+
+struct sock *scsi_nl_sock = NULL;
+EXPORT_SYMBOL_GPL(scsi_nl_sock);
+
+
+/**
+ * scsi_nl_rcv_msg -
+ *    Receive message handler. Extracts message from a receive buffer.
+ *    Validates message header and calls appropriate transport message handler
+ *
+ * @skb:		socket receive buffer
+ *
+ **/
+static void
+scsi_nl_rcv_msg(struct sk_buff *skb)
+{
+	struct nlmsghdr *nlh;
+	struct scsi_nl_hdr *hdr;
+	uint32_t rlen;
+	int err;
+
+	while (skb->len >= NLMSG_SPACE(0)) {
+		err = 0;
+
+		nlh = (struct nlmsghdr *) skb->data;
+		if ((nlh->nlmsg_len < (sizeof(*nlh) + sizeof(*hdr))) ||
+		    (skb->len < nlh->nlmsg_len)) {
+			printk(KERN_WARNING "%s: discarding partial skb\n",
+				 __FUNCTION__);
+			return;
+		}
+
+		rlen = NLMSG_ALIGN(nlh->nlmsg_len);
+		if (rlen > skb->len)
+			rlen = skb->len;
+
+		if (nlh->nlmsg_type != SCSI_TRANSPORT_MSG) {
+			err = -EBADMSG;
+			goto next_msg;
+		}
+
+		hdr = NLMSG_DATA(nlh);
+		if ((hdr->version != SCSI_NL_VERSION) ||
+		    (hdr->magic != SCSI_NL_MAGIC)) {
+			err = -EPROTOTYPE;
+			goto next_msg;
+		}
+
+		if (security_netlink_recv(skb, CAP_SYS_ADMIN)) {
+			err = -EPERM;
+			goto next_msg;
+		}
+
+		if (nlh->nlmsg_len < (sizeof(*nlh) + hdr->msglen)) {
+			printk(KERN_WARNING "%s: discarding partial message\n",
+				 __FUNCTION__);
+			return;
+		}
+
+		/*
+		 * We currently don't support anyone sending us a message
+		 */
+
+next_msg:
+		if ((err) || (nlh->nlmsg_flags & NLM_F_ACK))
+			netlink_ack(skb, nlh, err);
+
+		skb_pull(skb, rlen);
+	}
+}
+
+
+/**
+ * scsi_nl_rcv_msg -
+ *    Receive handler for a socket. Extracts a received message buffer from
+ *    the socket, and starts message processing.
+ *
+ * @sk:		socket
+ * @len:	unused
+ *
+ **/
+static void
+scsi_nl_rcv(struct sock *sk, int len)
+{
+	struct sk_buff *skb;
+
+	while ((skb = skb_dequeue(&sk->sk_receive_queue))) {
+		scsi_nl_rcv_msg(skb);
+		kfree_skb(skb);
+	}
+}
+
+
+/**
+ * scsi_nl_rcv_event -
+ *    Event handler for a netlink socket.
+ *
+ * @this:		event notifier block
+ * @event:		event type
+ * @ptr:		event payload
+ *
+ **/
+static int
+scsi_nl_rcv_event(struct notifier_block *this, unsigned long event, void *ptr)
+{
+	struct netlink_notify *n = ptr;
+
+	if (n->protocol != NETLINK_SCSITRANSPORT)
+		return NOTIFY_DONE;
+
+	/*
+	 * Currently, we are not tracking PID's, etc. There is nothing
+	 * to handle.
+	 */
+
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block scsi_netlink_notifier = {
+	.notifier_call  = scsi_nl_rcv_event,
+};
+
+
+/**
+ * scsi_netlink_init -
+ *    Called by SCSI subsystem to intialize the SCSI transport netlink
+ *    interface
+ *
+ **/
+void
+scsi_netlink_init(void)
+{
+	int error;
+
+	error = netlink_register_notifier(&scsi_netlink_notifier);
+	if (error) {
+		printk(KERN_ERR "%s: register of event handler failed - %d\n",
+				__FUNCTION__, error);
+		return;
+	}
+
+	scsi_nl_sock = netlink_kernel_create(NETLINK_SCSITRANSPORT,
+				SCSI_NL_GRP_CNT, scsi_nl_rcv, THIS_MODULE);
+	if (!scsi_nl_sock) {
+		printk(KERN_ERR "%s: register of recieve handler failed\n",
+				__FUNCTION__);
+		netlink_unregister_notifier(&scsi_netlink_notifier);
+	}
+
+	return;
+}
+
+
+/**
+ * scsi_netlink_exit -
+ *    Called by SCSI subsystem to disable the SCSI transport netlink
+ *    interface
+ *
+ **/
+void
+scsi_netlink_exit(void)
+{
+	if (scsi_nl_sock) {
+		sock_release(scsi_nl_sock->sk_socket);
+		netlink_unregister_notifier(&scsi_netlink_notifier);
+	}
+
+	return;
+}
+
+
diff --git a/drivers/scsi/scsi_priv.h b/drivers/scsi/scsi_priv.h
index ae24c85aaeea..5d023d44e5e7 100644
--- a/drivers/scsi/scsi_priv.h
+++ b/drivers/scsi/scsi_priv.h
@@ -8,6 +8,7 @@ struct scsi_cmnd;
 struct scsi_device;
 struct scsi_host_template;
 struct Scsi_Host;
+struct scsi_nl_hdr;
 
 
 /*
@@ -110,6 +111,16 @@ extern void __scsi_remove_device(struct scsi_device *);
 
 extern struct bus_type scsi_bus_type;
 
+/* scsi_netlink.c */
+#ifdef CONFIG_SCSI_NETLINK
+extern void scsi_netlink_init(void);
+extern void scsi_netlink_exit(void);
+extern struct sock *scsi_nl_sock;
+#else
+static inline void scsi_netlink_init(void) {}
+static inline void scsi_netlink_exit(void) {}
+#endif
+
 /* 
  * internal scsi timeout functions: for use by mid-layer and transport
  * classes.
diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c
index 79d31ca2b741..05989f130554 100644
--- a/drivers/scsi/scsi_transport_fc.c
+++ b/drivers/scsi/scsi_transport_fc.c
@@ -32,6 +32,9 @@
 #include <scsi/scsi_transport.h>
 #include <scsi/scsi_transport_fc.h>
 #include <scsi/scsi_cmnd.h>
+#include <linux/netlink.h>
+#include <net/netlink.h>
+#include <scsi/scsi_netlink_fc.h>
 #include "scsi_priv.h"
 
 static int fc_queue_work(struct Scsi_Host *, struct work_struct *);
@@ -93,6 +96,29 @@ fc_enum_name_search(port_type, fc_port_type, fc_port_type_names)
 #define FC_PORTTYPE_MAX_NAMELEN		50
 
 
+/* Convert fc_host_event_code values to ascii string name */
+static const struct {
+	enum fc_host_event_code		value;
+	char				*name;
+} fc_host_event_code_names[] = {
+	{ FCH_EVT_LIP,			"lip" },
+	{ FCH_EVT_LINKUP,		"link_up" },
+	{ FCH_EVT_LINKDOWN,		"link_down" },
+	{ FCH_EVT_LIPRESET,		"lip_reset" },
+	{ FCH_EVT_RSCN,			"rscn" },
+	{ FCH_EVT_ADAPTER_CHANGE,	"adapter_chg" },
+	{ FCH_EVT_PORT_UNKNOWN,		"port_unknown" },
+	{ FCH_EVT_PORT_ONLINE,		"port_online" },
+	{ FCH_EVT_PORT_OFFLINE,		"port_offline" },
+	{ FCH_EVT_PORT_FABRIC,		"port_fabric" },
+	{ FCH_EVT_LINK_UNKNOWN,		"link_unknown" },
+	{ FCH_EVT_VENDOR_UNIQUE,	"vendor_unique" },
+};
+fc_enum_name_search(host_event_code, fc_host_event_code,
+		fc_host_event_code_names)
+#define FC_HOST_EVENT_CODE_MAX_NAMELEN	30
+
+
 /* Convert fc_port_state values to ascii string name */
 static struct {
 	enum fc_port_state	value;
@@ -377,10 +403,182 @@ MODULE_PARM_DESC(dev_loss_tmo,
 		 " exceeded, the scsi target is removed. Value should be"
 		 " between 1 and SCSI_DEVICE_BLOCK_MAX_TIMEOUT.");
 
+/**
+ * Netlink Infrastructure
+ **/
+
+static atomic_t fc_event_seq;
+
+/**
+ * fc_get_event_number - Obtain the next sequential FC event number
+ *
+ * Notes:
+ *   We could have inline'd this, but it would have required fc_event_seq to
+ *   be exposed. For now, live with the subroutine call.
+ *   Atomic used to avoid lock/unlock...
+ **/
+u32
+fc_get_event_number(void)
+{
+	return atomic_add_return(1, &fc_event_seq);
+}
+EXPORT_SYMBOL(fc_get_event_number);
+
+
+/**
+ * fc_host_post_event - called to post an even on an fc_host.
+ *
+ * @shost:		host the event occurred on
+ * @event_number:	fc event number obtained from get_fc_event_number()
+ * @event_code:		fc_host event being posted
+ * @event_data:		32bits of data for the event being posted
+ *
+ * Notes:
+ *	This routine assumes no locks are held on entry.
+ **/
+void
+fc_host_post_event(struct Scsi_Host *shost, u32 event_number,
+		enum fc_host_event_code event_code, u32 event_data)
+{
+	struct sk_buff *skb;
+	struct nlmsghdr	*nlh;
+	struct fc_nl_event *event;
+	const char *name;
+	u32 len, skblen;
+	int err;
+
+	if (!scsi_nl_sock) {
+		err = -ENOENT;
+		goto send_fail;
+	}
+
+	len = FC_NL_MSGALIGN(sizeof(*event));
+	skblen = NLMSG_SPACE(len);
+
+	skb = alloc_skb(skblen, GFP_KERNEL);
+	if (!skb) {
+		err = -ENOBUFS;
+		goto send_fail;
+	}
+
+	nlh = nlmsg_put(skb, 0, 0, SCSI_TRANSPORT_MSG,
+				skblen - sizeof(*nlh), 0);
+	if (!nlh) {
+		err = -ENOBUFS;
+		goto send_fail_skb;
+	}
+	event = NLMSG_DATA(nlh);
+
+	INIT_SCSI_NL_HDR(&event->snlh, SCSI_NL_TRANSPORT_FC,
+				FC_NL_ASYNC_EVENT, len);
+	event->seconds = get_seconds();
+	event->vendor_id = 0;
+	event->host_no = shost->host_no;
+	event->event_datalen = sizeof(u32);	/* bytes */
+	event->event_num = event_number;
+	event->event_code = event_code;
+	event->event_data = event_data;
+
+	err = nlmsg_multicast(scsi_nl_sock, skb, 0, SCSI_NL_GRP_FC_EVENTS);
+	if (err && (err != -ESRCH))	/* filter no recipient errors */
+		/* nlmsg_multicast already kfree_skb'd */
+		goto send_fail;
+
+	return;
+
+send_fail_skb:
+	kfree_skb(skb);
+send_fail:
+	name = get_fc_host_event_code_name(event_code);
+	printk(KERN_WARNING
+		"%s: Dropped Event : host %d %s data 0x%08x - err %d\n",
+		__FUNCTION__, shost->host_no,
+		(name) ? name : "<unknown>", event_data, err);
+	return;
+}
+EXPORT_SYMBOL(fc_host_post_event);
+
+
+/**
+ * fc_host_post_vendor_event - called to post a vendor unique event on
+ *                             a fc_host
+ *
+ * @shost:		host the event occurred on
+ * @event_number:	fc event number obtained from get_fc_event_number()
+ * @data_len:		amount, in bytes, of vendor unique data
+ * @data_buf:		pointer to vendor unique data
+ *
+ * Notes:
+ *	This routine assumes no locks are held on entry.
+ **/
+void
+fc_host_post_vendor_event(struct Scsi_Host *shost, u32 event_number,
+		u32 data_len, char * data_buf, u32 vendor_id)
+{
+	struct sk_buff *skb;
+	struct nlmsghdr	*nlh;
+	struct fc_nl_event *event;
+	u32 len, skblen;
+	int err;
+
+	if (!scsi_nl_sock) {
+		err = -ENOENT;
+		goto send_vendor_fail;
+	}
+
+	len = FC_NL_MSGALIGN(sizeof(*event) + data_len);
+	skblen = NLMSG_SPACE(len);
+
+	skb = alloc_skb(skblen, GFP_KERNEL);
+	if (!skb) {
+		err = -ENOBUFS;
+		goto send_vendor_fail;
+	}
+
+	nlh = nlmsg_put(skb, 0, 0, SCSI_TRANSPORT_MSG,
+				skblen - sizeof(*nlh), 0);
+	if (!nlh) {
+		err = -ENOBUFS;
+		goto send_vendor_fail_skb;
+	}
+	event = NLMSG_DATA(nlh);
+
+	INIT_SCSI_NL_HDR(&event->snlh, SCSI_NL_TRANSPORT_FC,
+				FC_NL_ASYNC_EVENT, len);
+	event->seconds = get_seconds();
+	event->vendor_id = vendor_id;
+	event->host_no = shost->host_no;
+	event->event_datalen = data_len;	/* bytes */
+	event->event_num = event_number;
+	event->event_code = FCH_EVT_VENDOR_UNIQUE;
+	memcpy(&event->event_data, data_buf, data_len);
+
+	err = nlmsg_multicast(scsi_nl_sock, skb, 0, SCSI_NL_GRP_FC_EVENTS);
+	if (err && (err != -ESRCH))	/* filter no recipient errors */
+		/* nlmsg_multicast already kfree_skb'd */
+		goto send_vendor_fail;
+
+	return;
+
+send_vendor_fail_skb:
+	kfree_skb(skb);
+send_vendor_fail:
+	printk(KERN_WARNING
+		"%s: Dropped Event : host %d vendor_unique - err %d\n",
+		__FUNCTION__, shost->host_no, err);
+	return;
+}
+EXPORT_SYMBOL(fc_host_post_vendor_event);
+
+
 
 static __init int fc_transport_init(void)
 {
-	int error = transport_class_register(&fc_host_class);
+	int error;
+
+	atomic_set(&fc_event_seq, 0);
+
+	error = transport_class_register(&fc_host_class);
 	if (error)
 		return error;
 	error = transport_class_register(&fc_rport_class);
diff --git a/include/linux/netlink.h b/include/linux/netlink.h
index 855b44668caa..66411622e06e 100644
--- a/include/linux/netlink.h
+++ b/include/linux/netlink.h
@@ -21,6 +21,8 @@
 #define NETLINK_DNRTMSG		14	/* DECnet routing messages */
 #define NETLINK_KOBJECT_UEVENT	15	/* Kernel messages to userspace */
 #define NETLINK_GENERIC		16
+/* leave room for NETLINK_DM (DM Events) */
+#define NETLINK_SCSITRANSPORT	18	/* SCSI Transports */
 
 #define MAX_LINKS 32		
 
diff --git a/include/scsi/scsi_netlink.h b/include/scsi/scsi_netlink.h
new file mode 100644
index 000000000000..7a3a20e640c0
--- /dev/null
+++ b/include/scsi/scsi_netlink.h
@@ -0,0 +1,86 @@
+/*
+ *  SCSI Transport Netlink Interface
+ *    Used for the posting of outbound SCSI transport events
+ *
+ *  Copyright (C) 2006   James Smart, Emulex Corporation
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ *
+ */
+#ifndef SCSI_NETLINK_H
+#define SCSI_NETLINK_H
+
+/*
+ * This file intended to be included by both kernel and user space
+ */
+
+/* Single Netlink Message type to send all SCSI Transport messages */
+#define SCSI_TRANSPORT_MSG		NLMSG_MIN_TYPE + 1
+
+/* SCSI Transport Broadcast Groups */
+	/* leaving groups 0 and 1 unassigned */
+#define SCSI_NL_GRP_FC_EVENTS		(1<<2)		/* Group 2 */
+#define SCSI_NL_GRP_CNT			3
+
+
+/* SCSI_TRANSPORT_MSG event message header */
+struct scsi_nl_hdr {
+	uint8_t version;
+	uint8_t transport;
+	uint16_t magic;
+	uint16_t msgtype;
+	uint16_t msglen;
+} __attribute__((aligned(sizeof(uint64_t))));
+
+/* scsi_nl_hdr->version value */
+#define SCSI_NL_VERSION				1
+
+/* scsi_nl_hdr->magic value */
+#define SCSI_NL_MAGIC				0xA1B2
+
+/* scsi_nl_hdr->transport value */
+#define SCSI_NL_TRANSPORT			0
+#define SCSI_NL_TRANSPORT_FC			1
+#define SCSI_NL_MAX_TRANSPORTS			2
+
+/* scsi_nl_hdr->msgtype values are defined in each transport */
+
+
+/*
+ * Vendor ID:
+ *   If transports post vendor-unique events, they must pass a well-known
+ *   32-bit vendor identifier. This identifier consists of 8 bits indicating
+ *   the "type" of identifier contained, and 24 bits of id data.
+ *
+ *   Identifiers for each type:
+ *    PCI :  ID data is the 16 bit PCI Registered Vendor ID
+ */
+#define SCSI_NL_VID_ID_MASK			0x00FFFFFF
+#define SCSI_NL_VID_TYPE_MASK			0xFF000000
+#define SCSI_NL_VID_TYPE_PCI			0x01000000
+
+
+#define INIT_SCSI_NL_HDR(hdr, t, mtype, mlen)			\
+	{							\
+	(hdr)->version = SCSI_NL_VERSION;			\
+	(hdr)->transport = t;					\
+	(hdr)->magic = SCSI_NL_MAGIC;				\
+	(hdr)->msgtype = mtype;					\
+	(hdr)->msglen = mlen;					\
+	}
+
+
+#endif /* SCSI_NETLINK_H */
+
diff --git a/include/scsi/scsi_netlink_fc.h b/include/scsi/scsi_netlink_fc.h
new file mode 100644
index 000000000000..b213d2909fed
--- /dev/null
+++ b/include/scsi/scsi_netlink_fc.h
@@ -0,0 +1,71 @@
+/*
+ *  FC Transport Netlink Interface
+ *
+ *  Copyright (C) 2006   James Smart, Emulex Corporation
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ *
+ */
+#ifndef SCSI_NETLINK_FC_H
+#define SCSI_NETLINK_FC_H
+
+#include <scsi/scsi_netlink.h>
+
+/*
+ * This file intended to be included by both kernel and user space
+ */
+
+/*
+ * FC Transport Message Types
+ */
+	/* kernel -> user */
+#define FC_NL_ASYNC_EVENT			0x0100
+	/* user -> kernel */
+/* none */
+
+
+/*
+ * Message Structures :
+ */
+
+/* macro to round up message lengths to 8byte boundary */
+#define FC_NL_MSGALIGN(len)		(((len) + 7) & ~7)
+
+
+/*
+ * FC Transport Broadcast Event Message :
+ *   FC_NL_ASYNC_EVENT
+ *
+ * Note: if Vendor Unique message, &event_data will be  start of
+ * 	 vendor unique payload, and the length of the payload is
+ *       per event_datalen
+ *
+ * Note: When specifying vendor_id, be sure to read the Vendor Type and ID
+ *   formatting requirements specified in scsi_netlink.h
+ */
+struct fc_nl_event {
+	struct scsi_nl_hdr snlh;		/* must be 1st element ! */
+	uint64_t seconds;
+	uint32_t vendor_id;
+	uint16_t host_no;
+	uint16_t event_datalen;
+	uint32_t event_num;
+	uint32_t event_code;
+	uint32_t event_data;
+} __attribute__((aligned(sizeof(uint64_t))));
+
+
+#endif /* SCSI_NETLINK_FC_H */
+
diff --git a/include/scsi/scsi_transport_fc.h b/include/scsi/scsi_transport_fc.h
index c74be5dabfeb..f91c5358af3a 100644
--- a/include/scsi/scsi_transport_fc.h
+++ b/include/scsi/scsi_transport_fc.h
@@ -29,6 +29,7 @@
 
 #include <linux/sched.h>
 #include <scsi/scsi.h>
+#include <scsi/scsi_netlink.h>
 
 struct scsi_transport_template;
 
@@ -283,6 +284,30 @@ struct fc_host_statistics {
 };
 
 
+/*
+ * FC Event Codes - Polled and Async, following FC HBAAPI v2.0 guidelines
+ */
+
+/*
+ * fc_host_event_code: If you alter this, you also need to alter
+ * scsi_transport_fc.c (for the ascii descriptions).
+ */
+enum fc_host_event_code  {
+	FCH_EVT_LIP			= 0x1,
+	FCH_EVT_LINKUP			= 0x2,
+	FCH_EVT_LINKDOWN		= 0x3,
+	FCH_EVT_LIPRESET		= 0x4,
+	FCH_EVT_RSCN			= 0x5,
+	FCH_EVT_ADAPTER_CHANGE		= 0x103,
+	FCH_EVT_PORT_UNKNOWN		= 0x200,
+	FCH_EVT_PORT_OFFLINE		= 0x201,
+	FCH_EVT_PORT_ONLINE		= 0x202,
+	FCH_EVT_PORT_FABRIC		= 0x204,
+	FCH_EVT_LINK_UNKNOWN		= 0x500,
+	FCH_EVT_VENDOR_UNIQUE		= 0xffff,
+};
+
+
 /*
  * FC Local Port (Host) Attributes
  *
@@ -526,5 +551,14 @@ struct fc_rport *fc_remote_port_add(struct Scsi_Host *shost,
 void fc_remote_port_delete(struct fc_rport  *rport);
 void fc_remote_port_rolechg(struct fc_rport  *rport, u32 roles);
 int scsi_is_fc_rport(const struct device *);
+u32 fc_get_event_number(void);
+void fc_host_post_event(struct Scsi_Host *shost, u32 event_number,
+		enum fc_host_event_code event_code, u32 event_data);
+void fc_host_post_vendor_event(struct Scsi_Host *shost, u32 event_number,
+		u32 data_len, char * data_buf, u32 vendor_id);
+	/* Note: when specifying vendor_id to fc_host_post_vendor_event()
+	 *   be sure to read the Vendor Type and ID formatting requirements
+	 *   specified in scsi_netlink.h
+	 */
 
 #endif /* SCSI_TRANSPORT_FC_H */
-- 
cgit v1.2.3


From f14e2e29cdd07f80de6dec168dc2bb39de37eec3 Mon Sep 17 00:00:00 2001
From: James Smart <James.Smart@Emulex.Com>
Date: Tue, 22 Aug 2006 09:55:23 -0400
Subject: [SCSI] SCSI & FC transport: extend event vendor id's to 64bits

During discussions with Mike Christie, I became convinced that we needed
a larger vendor id. This patch extends the id from 32 to 64 bits.

This applies on top of the prior patches that add SCSI transport events
via netlink.

Signed-off-by: James Smart <James.Smart@emulex.com>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 drivers/scsi/scsi_transport_fc.c | 2 +-
 include/scsi/scsi_netlink.h      | 7 ++++---
 include/scsi/scsi_netlink_fc.h   | 2 +-
 include/scsi/scsi_transport_fc.h | 2 +-
 4 files changed, 7 insertions(+), 6 deletions(-)

(limited to 'include')

diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c
index 05989f130554..293188cbff8c 100644
--- a/drivers/scsi/scsi_transport_fc.c
+++ b/drivers/scsi/scsi_transport_fc.c
@@ -513,7 +513,7 @@ EXPORT_SYMBOL(fc_host_post_event);
  **/
 void
 fc_host_post_vendor_event(struct Scsi_Host *shost, u32 event_number,
-		u32 data_len, char * data_buf, u32 vendor_id)
+		u32 data_len, char * data_buf, u64 vendor_id)
 {
 	struct sk_buff *skb;
 	struct nlmsghdr	*nlh;
diff --git a/include/scsi/scsi_netlink.h b/include/scsi/scsi_netlink.h
index 7a3a20e640c0..8c1470cc8209 100644
--- a/include/scsi/scsi_netlink.h
+++ b/include/scsi/scsi_netlink.h
@@ -67,9 +67,10 @@ struct scsi_nl_hdr {
  *   Identifiers for each type:
  *    PCI :  ID data is the 16 bit PCI Registered Vendor ID
  */
-#define SCSI_NL_VID_ID_MASK			0x00FFFFFF
-#define SCSI_NL_VID_TYPE_MASK			0xFF000000
-#define SCSI_NL_VID_TYPE_PCI			0x01000000
+#define SCSI_NL_VID_TYPE_SHIFT		56
+#define SCSI_NL_VID_TYPE_MASK		((u64)0xFF << SCSI_NL_VID_TYPE_SHIFT)
+#define SCSI_NL_VID_TYPE_PCI		((u64)0x01 << SCSI_NL_VID_TYPE_SHIFT)
+#define SCSI_NL_VID_ID_MASK		(~ SCSI_NL_VID_TYPE_MASK)
 
 
 #define INIT_SCSI_NL_HDR(hdr, t, mtype, mlen)			\
diff --git a/include/scsi/scsi_netlink_fc.h b/include/scsi/scsi_netlink_fc.h
index b213d2909fed..cbf76e479761 100644
--- a/include/scsi/scsi_netlink_fc.h
+++ b/include/scsi/scsi_netlink_fc.h
@@ -58,7 +58,7 @@
 struct fc_nl_event {
 	struct scsi_nl_hdr snlh;		/* must be 1st element ! */
 	uint64_t seconds;
-	uint32_t vendor_id;
+	uint64_t vendor_id;
 	uint16_t host_no;
 	uint16_t event_datalen;
 	uint32_t event_num;
diff --git a/include/scsi/scsi_transport_fc.h b/include/scsi/scsi_transport_fc.h
index f91c5358af3a..0b11eff989e0 100644
--- a/include/scsi/scsi_transport_fc.h
+++ b/include/scsi/scsi_transport_fc.h
@@ -555,7 +555,7 @@ u32 fc_get_event_number(void);
 void fc_host_post_event(struct Scsi_Host *shost, u32 event_number,
 		enum fc_host_event_code event_code, u32 event_data);
 void fc_host_post_vendor_event(struct Scsi_Host *shost, u32 event_number,
-		u32 data_len, char * data_buf, u32 vendor_id);
+		u32 data_len, char * data_buf, u64 vendor_id);
 	/* Note: when specifying vendor_id to fc_host_post_vendor_event()
 	 *   be sure to read the Vendor Type and ID formatting requirements
 	 *   specified in scsi_netlink.h
-- 
cgit v1.2.3


From 0f29b966d60e9a4f5ecff9f3832257b38aea4f13 Mon Sep 17 00:00:00 2001
From: James Smart <James.Smart@Emulex.Com>
Date: Fri, 18 Aug 2006 17:33:29 -0400
Subject: [SCSI] FC transport: Add dev_loss_tmo callbacks, and new
 fast_io_fail_tmo w/ callback

This patch adds the following functionality to the FC transport:

- dev_loss_tmo LLDD callback :
  Called to essentially confirm the deletion of an rport. Thus, it is
  called whenever the dev_loss_tmo fires, or when the rport is deleted
  due to other circumstances (module unload, etc).  It is expected that
  the callback will initiate the termination of any outstanding i/o on
  the rport.

- fast_io_fail_tmo and LLD callback:
  There are some cases where it may take a long while to truly determine
  device loss, but the system is in a multipathing configuration that if
  the i/o was failed quickly (faster than dev_loss_tmo), it could be
  redirected to a different path and completed sooner.

Many thanks to Mike Reed who cleaned up the initial RFC in support
of this post.

The original RFC is at:
http://marc.theaimsgroup.com/?l=linux-scsi&m=115505981027246&w=2

Signed-off-by: James Smart <James.Smart@emulex.com>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 drivers/scsi/scsi_transport_fc.c | 134 +++++++++++++++++++++++++++++++++++----
 include/scsi/scsi_transport_fc.h |   5 ++
 2 files changed, 128 insertions(+), 11 deletions(-)

(limited to 'include')

diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c
index 293188cbff8c..4ab176ed480d 100644
--- a/drivers/scsi/scsi_transport_fc.c
+++ b/drivers/scsi/scsi_transport_fc.c
@@ -242,6 +242,7 @@ fc_bitfield_name_search(remote_port_roles, fc_remote_port_role_names)
 
 
 static void fc_timeout_deleted_rport(void *data);
+static void fc_timeout_fail_rport_io(void *data);
 static void fc_scsi_scan_rport(void *data);
 
 /*
@@ -249,7 +250,7 @@ static void fc_scsi_scan_rport(void *data);
  * Increase these values if you add attributes
  */
 #define FC_STARGET_NUM_ATTRS 	3
-#define FC_RPORT_NUM_ATTRS	9
+#define FC_RPORT_NUM_ATTRS	10
 #define FC_HOST_NUM_ATTRS	17
 
 struct fc_internal {
@@ -622,11 +623,14 @@ store_fc_rport_##field(struct class_device *cdev, const char *buf,	\
 	struct fc_rport *rport = transport_class_to_rport(cdev);	\
 	struct Scsi_Host *shost = rport_to_shost(rport);		\
 	struct fc_internal *i = to_fc_internal(shost->transportt);	\
+	char *cp;							\
 	if ((rport->port_state == FC_PORTSTATE_BLOCKED) ||		\
 	    (rport->port_state == FC_PORTSTATE_DELETED) ||		\
 	    (rport->port_state == FC_PORTSTATE_NOTPRESENT))		\
 		return -EBUSY;						\
-	val = simple_strtoul(buf, NULL, 0);				\
+	val = simple_strtoul(buf, &cp, 0);				\
+	if (*cp && (*cp != '\n'))					\
+		return -EINVAL;						\
 	i->f->set_rport_##field(rport, val);				\
 	return count;							\
 }
@@ -708,6 +712,13 @@ static FC_CLASS_DEVICE_ATTR(rport, title, S_IRUGO,			\
 	if (i->f->show_rport_##field)					\
 		count++
 
+#define SETUP_PRIVATE_RPORT_ATTRIBUTE_RW(field)				\
+{									\
+	i->private_rport_attrs[count] = class_device_attr_rport_##field; \
+	i->rport_attrs[count] = &i->private_rport_attrs[count];		\
+	count++;							\
+}
+
 
 /* The FC Transport Remote Port Attributes: */
 
@@ -740,12 +751,14 @@ store_fc_rport_dev_loss_tmo(struct class_device *cdev, const char *buf,
 	struct fc_rport *rport = transport_class_to_rport(cdev);
 	struct Scsi_Host *shost = rport_to_shost(rport);
 	struct fc_internal *i = to_fc_internal(shost->transportt);
+	char *cp;
 	if ((rport->port_state == FC_PORTSTATE_BLOCKED) ||
 	    (rport->port_state == FC_PORTSTATE_DELETED) ||
 	    (rport->port_state == FC_PORTSTATE_NOTPRESENT))
 		return -EBUSY;
-	val = simple_strtoul(buf, NULL, 0);
-	if ((val < 0) || (val > SCSI_DEVICE_BLOCK_MAX_TIMEOUT))
+	val = simple_strtoul(buf, &cp, 0);
+	if ((*cp && (*cp != '\n')) ||
+	    (val < 0) || (val > SCSI_DEVICE_BLOCK_MAX_TIMEOUT))
 		return -EINVAL;
 	i->f->set_rport_dev_loss_tmo(rport, val);
 	return count;
@@ -795,6 +808,44 @@ static FC_CLASS_DEVICE_ATTR(rport, roles, S_IRUGO,
 fc_private_rport_rd_enum_attr(port_state, FC_PORTSTATE_MAX_NAMELEN);
 fc_private_rport_rd_attr(scsi_target_id, "%d\n", 20);
 
+/*
+ * fast_io_fail_tmo attribute
+ */
+static ssize_t
+show_fc_rport_fast_io_fail_tmo (struct class_device *cdev, char *buf)
+{
+	struct fc_rport *rport = transport_class_to_rport(cdev);
+
+	if (rport->fast_io_fail_tmo == -1)
+		return snprintf(buf, 5, "off\n");
+	return snprintf(buf, 20, "%d\n", rport->fast_io_fail_tmo);
+}
+
+static ssize_t
+store_fc_rport_fast_io_fail_tmo(struct class_device *cdev, const char *buf,
+			   size_t count)
+{
+	int val;
+	char *cp;
+	struct fc_rport *rport = transport_class_to_rport(cdev);
+
+	if ((rport->port_state == FC_PORTSTATE_BLOCKED) ||
+	    (rport->port_state == FC_PORTSTATE_DELETED) ||
+	    (rport->port_state == FC_PORTSTATE_NOTPRESENT))
+		return -EBUSY;
+	if (strncmp(buf, "off", 3) == 0)
+		rport->fast_io_fail_tmo = -1;
+	else {
+		val = simple_strtoul(buf, &cp, 0);
+		if ((*cp && (*cp != '\n')) ||
+		    (val < 0) || (val >= rport->dev_loss_tmo))
+			return -EINVAL;
+		rport->fast_io_fail_tmo = val;
+	}
+	return count;
+}
+static FC_CLASS_DEVICE_ATTR(rport, fast_io_fail_tmo, S_IRUGO | S_IWUSR,
+	show_fc_rport_fast_io_fail_tmo, store_fc_rport_fast_io_fail_tmo);
 
 
 /*
@@ -880,8 +931,11 @@ store_fc_host_##field(struct class_device *cdev, const char *buf,	\
 	int val;							\
 	struct Scsi_Host *shost = transport_class_to_shost(cdev);	\
 	struct fc_internal *i = to_fc_internal(shost->transportt);	\
+	char *cp;							\
 									\
-	val = simple_strtoul(buf, NULL, 0);				\
+	val = simple_strtoul(buf, &cp, 0);				\
+	if (*cp && (*cp != '\n'))					\
+		return -EINVAL;						\
 	i->f->set_host_##field(shost, val);				\
 	return count;							\
 }
@@ -1481,6 +1535,8 @@ fc_attach_transport(struct fc_function_template *ft)
 	SETUP_PRIVATE_RPORT_ATTRIBUTE_RD(roles);
 	SETUP_PRIVATE_RPORT_ATTRIBUTE_RD(port_state);
 	SETUP_PRIVATE_RPORT_ATTRIBUTE_RD(scsi_target_id);
+	if (ft->terminate_rport_io)
+		SETUP_PRIVATE_RPORT_ATTRIBUTE_RW(fast_io_fail_tmo);
 
 	BUG_ON(count > FC_RPORT_NUM_ATTRS);
 
@@ -1552,7 +1608,7 @@ fc_flush_work(struct Scsi_Host *shost)
  * @delay:	jiffies to delay the work queuing
  *
  * Return value:
- * 	0 on success / != 0 for error
+ * 	1 on success / 0 already queued / < 0 for error
  **/
 static int
 fc_queue_devloss_work(struct Scsi_Host *shost, struct work_struct *work,
@@ -1567,6 +1623,9 @@ fc_queue_devloss_work(struct Scsi_Host *shost, struct work_struct *work,
 		return -EINVAL;
 	}
 
+	if (delay == 0)
+		return queue_work(fc_host_devloss_work_q(shost), work);
+
 	return queue_delayed_work(fc_host_devloss_work_q(shost), work, delay);
 }
 
@@ -1659,10 +1718,23 @@ fc_starget_delete(void *data)
 	struct fc_rport *rport = (struct fc_rport *)data;
 	struct Scsi_Host *shost = rport_to_shost(rport);
 	unsigned long flags;
+	struct fc_internal *i = to_fc_internal(shost->transportt);
+
+	/*
+	 * Involve the LLDD if possible. All io on the rport is to
+	 * be terminated, either as part of the dev_loss_tmo callback
+	 * processing, or via the terminate_rport_io function.
+	 */
+	if (i->f->dev_loss_tmo_callbk)
+		i->f->dev_loss_tmo_callbk(rport);
+	else if (i->f->terminate_rport_io)
+		i->f->terminate_rport_io(rport);
 
 	spin_lock_irqsave(shost->host_lock, flags);
 	if (rport->flags & FC_RPORT_DEVLOSS_PENDING) {
 		spin_unlock_irqrestore(shost->host_lock, flags);
+		if (!cancel_delayed_work(&rport->fail_io_work))
+			fc_flush_devloss(shost);
 		if (!cancel_delayed_work(&rport->dev_loss_work))
 			fc_flush_devloss(shost);
 		spin_lock_irqsave(shost->host_lock, flags);
@@ -1685,10 +1757,7 @@ fc_rport_final_delete(void *data)
 	struct fc_rport *rport = (struct fc_rport *)data;
 	struct device *dev = &rport->dev;
 	struct Scsi_Host *shost = rport_to_shost(rport);
-
-	/* Delete SCSI target and sdevs */
-	if (rport->scsi_target_id != -1)
-		fc_starget_delete(data);
+	struct fc_internal *i = to_fc_internal(shost->transportt);
 
 	/*
 	 * if a scan is pending, flush the SCSI Host work_q so that 
@@ -1697,6 +1766,14 @@ fc_rport_final_delete(void *data)
 	if (rport->flags & FC_RPORT_SCAN_PENDING)
 		scsi_flush_work(shost);
 
+	/* Delete SCSI target and sdevs */
+	if (rport->scsi_target_id != -1)
+		fc_starget_delete(data);
+	else if (i->f->dev_loss_tmo_callbk)
+		i->f->dev_loss_tmo_callbk(rport);
+	else if (i->f->terminate_rport_io)
+		i->f->terminate_rport_io(rport);
+
 	transport_remove_device(dev);
 	device_del(dev);
 	transport_destroy_device(dev);
@@ -1748,8 +1825,10 @@ fc_rport_create(struct Scsi_Host *shost, int channel,
 	if (fci->f->dd_fcrport_size)
 		rport->dd_data = &rport[1];
 	rport->channel = channel;
+	rport->fast_io_fail_tmo = -1;
 
 	INIT_WORK(&rport->dev_loss_work, fc_timeout_deleted_rport, rport);
+	INIT_WORK(&rport->fail_io_work, fc_timeout_fail_rport_io, rport);
 	INIT_WORK(&rport->scan_work, fc_scsi_scan_rport, rport);
 	INIT_WORK(&rport->stgt_delete_work, fc_starget_delete, rport);
 	INIT_WORK(&rport->rport_delete_work, fc_rport_final_delete, rport);
@@ -1913,11 +1992,13 @@ fc_remote_port_add(struct Scsi_Host *shost, int channel,
 				/* restart the target */
 
 				/*
-				 * Stop the target timer first. Take no action
+				 * Stop the target timers first. Take no action
 				 * on the del_timer failure as the state
 				 * machine state change will validate the
 				 * transaction.
 				 */
+				if (!cancel_delayed_work(&rport->fail_io_work))
+					fc_flush_devloss(shost);
 				if (!cancel_delayed_work(work))
 					fc_flush_devloss(shost);
 
@@ -2061,6 +2142,7 @@ void
 fc_remote_port_delete(struct fc_rport  *rport)
 {
 	struct Scsi_Host *shost = rport_to_shost(rport);
+	struct fc_internal *i = to_fc_internal(shost->transportt);
 	int timeout = rport->dev_loss_tmo;
 	unsigned long flags;
 
@@ -2091,6 +2173,12 @@ fc_remote_port_delete(struct fc_rport  *rport)
 
 	scsi_target_block(&rport->dev);
 
+	/* see if we need to kill io faster than waiting for device loss */
+	if ((rport->fast_io_fail_tmo != -1) &&
+	    (rport->fast_io_fail_tmo < timeout) && (i->f->terminate_rport_io))
+		fc_queue_devloss_work(shost, &rport->fail_io_work,
+					rport->fast_io_fail_tmo * HZ);
+
 	/* cap the length the devices can be blocked until they are deleted */
 	fc_queue_devloss_work(shost, &rport->dev_loss_work, timeout * HZ);
 }
@@ -2150,6 +2238,8 @@ fc_remote_port_rolechg(struct fc_rport  *rport, u32 roles)
 		 * machine state change will validate the
 		 * transaction.
 		 */
+		if (!cancel_delayed_work(&rport->fail_io_work))
+			fc_flush_devloss(shost);
 		if (!cancel_delayed_work(&rport->dev_loss_work))
 			fc_flush_devloss(shost);
 
@@ -2270,6 +2360,28 @@ fc_timeout_deleted_rport(void  *data)
 	fc_queue_work(shost, &rport->stgt_delete_work);
 }
 
+/**
+ * fc_timeout_fail_rport_io - Timeout handler for a fast io failing on a
+ *                       disconnected SCSI target.
+ *
+ * @data:	rport to terminate io on.
+ *
+ * Notes: Only requests the failure of the io, not that all are flushed
+ *    prior to returning.
+ **/
+static void
+fc_timeout_fail_rport_io(void  *data)
+{
+	struct fc_rport *rport = (struct fc_rport *)data;
+	struct Scsi_Host *shost = rport_to_shost(rport);
+	struct fc_internal *i = to_fc_internal(shost->transportt);
+
+	if (rport->port_state != FC_PORTSTATE_BLOCKED)
+		return;
+
+	i->f->terminate_rport_io(rport);
+}
+
 /**
  * fc_scsi_scan_rport - called to perform a scsi scan on a remote port.
  *
diff --git a/include/scsi/scsi_transport_fc.h b/include/scsi/scsi_transport_fc.h
index 0b11eff989e0..fd352323378b 100644
--- a/include/scsi/scsi_transport_fc.h
+++ b/include/scsi/scsi_transport_fc.h
@@ -195,6 +195,7 @@ struct fc_rport {	/* aka fc_starget_attrs */
 	u32 roles;
 	enum fc_port_state port_state;	/* Will only be ONLINE or UNKNOWN */
 	u32 scsi_target_id;
+	u32 fast_io_fail_tmo;
 
 	/* exported data */
 	void *dd_data;			/* Used for driver-specific storage */
@@ -207,6 +208,7 @@ struct fc_rport {	/* aka fc_starget_attrs */
 	struct device dev;
  	struct work_struct dev_loss_work;
  	struct work_struct scan_work;
+ 	struct work_struct fail_io_work;
  	struct work_struct stgt_delete_work;
 	struct work_struct rport_delete_work;
 } __attribute__((aligned(sizeof(unsigned long))));
@@ -445,6 +447,9 @@ struct fc_function_template {
 
 	int	(*issue_fc_host_lip)(struct Scsi_Host *);
 
+	void    (*dev_loss_tmo_callbk)(struct fc_rport *);
+	void	(*terminate_rport_io)(struct fc_rport *);
+
 	/* allocation lengths for host-specific data */
 	u32	 			dd_fcrport_size;
 
-- 
cgit v1.2.3


From 88edf74610bd894b93438f389688bc8b4a2d3414 Mon Sep 17 00:00:00 2001
From: James Bottomley <James.Bottomley@steeleye.com>
Date: Wed, 6 Sep 2006 17:36:13 -0500
Subject: [SCSI] SAS: consolidate linkspeed definitions

At the moment we have two separate linkspeed enumerations covering
roughly the same values.  This patch consolidates on a single one enum
sas_linkspeed in scsi_transport_sas.h and uses it everywhere in the
aic94xx driver.  Eventually I'll get around to removing the duplicated
fields in asd_sas_phy and sas_phy ...

Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 drivers/scsi/aic94xx/aic94xx_hwi.c  |  2 +-
 drivers/scsi/aic94xx/aic94xx_init.c | 12 ++++++++----
 drivers/scsi/aic94xx/aic94xx_scb.c  | 26 +++++++++++++-------------
 drivers/scsi/libsas/sas_expander.c  | 27 +++++++--------------------
 drivers/scsi/libsas/sas_internal.h  |  2 +-
 include/scsi/libsas.h               | 14 +++++++-------
 include/scsi/sas.h                  | 14 --------------
 include/scsi/scsi_transport_sas.h   | 26 +++++++++++++++++---------
 8 files changed, 54 insertions(+), 69 deletions(-)

(limited to 'include')

diff --git a/drivers/scsi/aic94xx/aic94xx_hwi.c b/drivers/scsi/aic94xx/aic94xx_hwi.c
index 075cea85b56b..a24201351108 100644
--- a/drivers/scsi/aic94xx/aic94xx_hwi.c
+++ b/drivers/scsi/aic94xx/aic94xx_hwi.c
@@ -96,7 +96,7 @@ static int asd_init_phy(struct asd_phy *phy)
 	sas_phy->type = PHY_TYPE_PHYSICAL;
 	sas_phy->role = PHY_ROLE_INITIATOR;
 	sas_phy->oob_mode = OOB_NOT_CONNECTED;
-	sas_phy->linkrate = PHY_LINKRATE_NONE;
+	sas_phy->linkrate = SAS_LINK_RATE_UNKNOWN;
 
 	phy->id_frm_tok = asd_alloc_coherent(asd_ha,
 					     sizeof(*phy->identify_frame),
diff --git a/drivers/scsi/aic94xx/aic94xx_init.c b/drivers/scsi/aic94xx/aic94xx_init.c
index 69aa70887530..302b54fddf3c 100644
--- a/drivers/scsi/aic94xx/aic94xx_init.c
+++ b/drivers/scsi/aic94xx/aic94xx_init.c
@@ -240,10 +240,14 @@ static int __devinit asd_common_setup(struct asd_ha_struct *asd_ha)
 	/* All phys are enabled, by default. */
 	asd_ha->hw_prof.enabled_phys = 0xFF;
 	for (i = 0; i < ASD_MAX_PHYS; i++) {
-		asd_ha->hw_prof.phy_desc[i].max_sas_lrate = PHY_LINKRATE_3;
-		asd_ha->hw_prof.phy_desc[i].min_sas_lrate = PHY_LINKRATE_1_5;
-		asd_ha->hw_prof.phy_desc[i].max_sata_lrate= PHY_LINKRATE_1_5;
-		asd_ha->hw_prof.phy_desc[i].min_sata_lrate= PHY_LINKRATE_1_5;
+		asd_ha->hw_prof.phy_desc[i].max_sas_lrate =
+			SAS_LINK_RATE_3_0_GBPS;
+		asd_ha->hw_prof.phy_desc[i].min_sas_lrate =
+			SAS_LINK_RATE_1_5_GBPS;
+		asd_ha->hw_prof.phy_desc[i].max_sata_lrate =
+			SAS_LINK_RATE_1_5_GBPS;
+		asd_ha->hw_prof.phy_desc[i].min_sata_lrate =
+			SAS_LINK_RATE_1_5_GBPS;
 	}
 
 	return 0;
diff --git a/drivers/scsi/aic94xx/aic94xx_scb.c b/drivers/scsi/aic94xx/aic94xx_scb.c
index fc1b7438a913..ef8ca08b545f 100644
--- a/drivers/scsi/aic94xx/aic94xx_scb.c
+++ b/drivers/scsi/aic94xx/aic94xx_scb.c
@@ -55,15 +55,15 @@ static inline void get_lrate_mode(struct asd_phy *phy, u8 oob_mode)
 	switch (oob_mode & 7) {
 	case PHY_SPEED_60:
 		/* FIXME: sas transport class doesn't have this */
-		phy->sas_phy.linkrate = PHY_LINKRATE_6;
+		phy->sas_phy.linkrate = SAS_LINK_RATE_6_0_GBPS;
 		phy->sas_phy.phy->negotiated_linkrate = SAS_LINK_RATE_6_0_GBPS;
 		break;
 	case PHY_SPEED_30:
-		phy->sas_phy.linkrate = PHY_LINKRATE_3;
+		phy->sas_phy.linkrate = SAS_LINK_RATE_3_0_GBPS;
 		phy->sas_phy.phy->negotiated_linkrate = SAS_LINK_RATE_3_0_GBPS;
 		break;
 	case PHY_SPEED_15:
-		phy->sas_phy.linkrate = PHY_LINKRATE_1_5;
+		phy->sas_phy.linkrate = SAS_LINK_RATE_1_5_GBPS;
 		phy->sas_phy.phy->negotiated_linkrate = SAS_LINK_RATE_1_5_GBPS;
 		break;
 	}
@@ -540,39 +540,39 @@ static inline void set_speed_mask(u8 *speed_mask, struct asd_phy_desc *pd)
 		| SATA_SPEED_30_DIS | SATA_SPEED_15_DIS;
 
 	switch (pd->max_sas_lrate) {
-	case PHY_LINKRATE_6:
+	case SAS_LINK_RATE_6_0_GBPS:
 		*speed_mask &= ~SAS_SPEED_60_DIS;
 	default:
-	case PHY_LINKRATE_3:
+	case SAS_LINK_RATE_3_0_GBPS:
 		*speed_mask &= ~SAS_SPEED_30_DIS;
-	case PHY_LINKRATE_1_5:
+	case SAS_LINK_RATE_1_5_GBPS:
 		*speed_mask &= ~SAS_SPEED_15_DIS;
 	}
 
 	switch (pd->min_sas_lrate) {
-	case PHY_LINKRATE_6:
+	case SAS_LINK_RATE_6_0_GBPS:
 		*speed_mask |= SAS_SPEED_30_DIS;
-	case PHY_LINKRATE_3:
+	case SAS_LINK_RATE_3_0_GBPS:
 		*speed_mask |= SAS_SPEED_15_DIS;
 	default:
-	case PHY_LINKRATE_1_5:
+	case SAS_LINK_RATE_1_5_GBPS:
 		/* nothing to do */
 		;
 	}
 
 	switch (pd->max_sata_lrate) {
-	case PHY_LINKRATE_3:
+	case SAS_LINK_RATE_3_0_GBPS:
 		*speed_mask &= ~SATA_SPEED_30_DIS;
 	default:
-	case PHY_LINKRATE_1_5:
+	case SAS_LINK_RATE_1_5_GBPS:
 		*speed_mask &= ~SATA_SPEED_15_DIS;
 	}
 
 	switch (pd->min_sata_lrate) {
-	case PHY_LINKRATE_3:
+	case SAS_LINK_RATE_3_0_GBPS:
 		*speed_mask |= SATA_SPEED_15_DIS;
 	default:
-	case PHY_LINKRATE_1_5:
+	case SAS_LINK_RATE_1_5_GBPS:
 		/* nothing to do */
 		;
 	}
diff --git a/drivers/scsi/libsas/sas_expander.c b/drivers/scsi/libsas/sas_expander.c
index b653a263f76a..02e796ee027e 100644
--- a/drivers/scsi/libsas/sas_expander.c
+++ b/drivers/scsi/libsas/sas_expander.c
@@ -191,20 +191,7 @@ static void sas_set_ex_phy(struct domain_device *dev, int phy_id,
 	phy->phy->maximum_linkrate_hw = SAS_LINK_RATE_3_0_GBPS;
 	phy->phy->minimum_linkrate = SAS_LINK_RATE_1_5_GBPS;
 	phy->phy->maximum_linkrate = SAS_LINK_RATE_3_0_GBPS;
-	switch (phy->linkrate) {
-	case PHY_LINKRATE_1_5:
-		phy->phy->negotiated_linkrate = SAS_LINK_RATE_1_5_GBPS;
-		break;
-	case PHY_LINKRATE_3:
-		phy->phy->negotiated_linkrate = SAS_LINK_RATE_3_0_GBPS;
-		break;
-	case PHY_LINKRATE_6:
-		phy->phy->negotiated_linkrate = SAS_LINK_RATE_6_0_GBPS;
-		break;
-	default:
-		phy->phy->negotiated_linkrate = SAS_LINK_RATE_UNKNOWN;
-		break;
-	}
+	phy->phy->negotiated_linkrate = phy->linkrate;
 
 	if (!rediscover)
 		sas_phy_add(phy->phy);
@@ -450,7 +437,7 @@ static void sas_ex_disable_phy(struct domain_device *dev, int phy_id)
 	struct ex_phy *phy = &ex->ex_phy[phy_id];
 
 	sas_smp_phy_control(dev, phy_id, PHY_FUNC_DISABLE);
-	phy->linkrate = PHY_DISABLED;
+	phy->linkrate = SAS_PHY_DISABLED;
 }
 
 static void sas_ex_disable_port(struct domain_device *dev, u8 *sas_addr)
@@ -743,7 +730,7 @@ static int sas_ex_discover_dev(struct domain_device *dev, int phy_id)
 	int res = 0;
 
 	/* Phy state */
-	if (ex_phy->linkrate == PHY_SPINUP_HOLD) {
+	if (ex_phy->linkrate == SAS_SATA_SPINUP_HOLD) {
 		if (!sas_smp_phy_control(dev, phy_id, PHY_FUNC_LINK_RESET))
 			res = sas_ex_phy_discover(dev, phy_id);
 		if (res)
@@ -773,7 +760,7 @@ static int sas_ex_discover_dev(struct domain_device *dev, int phy_id)
 			sas_configure_routing(dev, ex_phy->attached_sas_addr);
 		}
 		return 0;
-	} else if (ex_phy->linkrate == PHY_LINKRATE_UNKNOWN)
+	} else if (ex_phy->linkrate == SAS_LINK_RATE_UNKNOWN)
 		return 0;
 
 	if (ex_phy->attached_dev_type != SAS_END_DEV &&
@@ -922,9 +909,9 @@ static int sas_ex_discover_devices(struct domain_device *dev, int single)
 			continue;
 
 		switch (ex_phy->linkrate) {
-		case PHY_DISABLED:
-		case PHY_RESET_PROBLEM:
-		case PHY_PORT_SELECTOR:
+		case SAS_PHY_DISABLED:
+		case SAS_PHY_RESET_PROBLEM:
+		case SAS_SATA_PORT_SELECTOR:
 			continue;
 		default:
 			res = sas_ex_discover_dev(dev, i);
diff --git a/drivers/scsi/libsas/sas_internal.h b/drivers/scsi/libsas/sas_internal.h
index 89c397680846..0d69ede4b944 100644
--- a/drivers/scsi/libsas/sas_internal.h
+++ b/drivers/scsi/libsas/sas_internal.h
@@ -43,7 +43,7 @@ void sas_scsi_recover_host(struct Scsi_Host *shost);
 
 int sas_show_class(enum sas_class class, char *buf);
 int sas_show_proto(enum sas_proto proto, char *buf);
-int sas_show_linkrate(enum sas_phy_linkrate linkrate, char *buf);
+int sas_show_linkrate(enum sas_linkrate linkrate, char *buf);
 int sas_show_oob_mode(enum sas_oob_mode oob_mode, char *buf);
 
 int  sas_register_phys(struct sas_ha_struct *sas_ha);
diff --git a/include/scsi/libsas.h b/include/scsi/libsas.h
index 72acdabe7f80..8d91313dd888 100644
--- a/include/scsi/libsas.h
+++ b/include/scsi/libsas.h
@@ -114,7 +114,7 @@ struct ex_phy {
 	enum ex_phy_state phy_state;
 
 	enum sas_dev_type attached_dev_type;
-	enum sas_phy_linkrate linkrate;
+	enum sas_linkrate linkrate;
 
 	u8   attached_sata_host:1;
 	u8   attached_sata_dev:1;
@@ -170,9 +170,9 @@ struct sata_device {
 struct domain_device {
         enum sas_dev_type dev_type;
 
-        enum sas_phy_linkrate linkrate;
-        enum sas_phy_linkrate min_linkrate;
-        enum sas_phy_linkrate max_linkrate;
+        enum sas_linkrate linkrate;
+        enum sas_linkrate min_linkrate;
+        enum sas_linkrate max_linkrate;
 
         int  pathways;
 
@@ -220,7 +220,7 @@ struct asd_sas_port {
 	struct domain_device *port_dev;
 	spinlock_t dev_list_lock;
 	struct list_head dev_list;
-	enum   sas_phy_linkrate linkrate;
+	enum   sas_linkrate linkrate;
 
 	struct sas_phy *phy;
 	struct work_struct work;
@@ -276,7 +276,7 @@ struct asd_sas_phy {
 	enum sas_phy_type  type;
 	enum sas_phy_role  role;
 	enum sas_oob_mode  oob_mode;
-	enum sas_phy_linkrate linkrate;
+	enum sas_linkrate linkrate;
 
 	u8   *sas_addr;		  /* must be set */
 	u8   attached_sas_addr[SAS_ADDR_SIZE]; /* class:RO, driver: R/W */
@@ -368,7 +368,7 @@ void sas_hash_addr(u8 *hashed, const u8 *sas_addr);
 static inline void sas_phy_disconnected(struct asd_sas_phy *phy)
 {
 	phy->oob_mode = OOB_NOT_CONNECTED;
-	phy->linkrate = PHY_LINKRATE_NONE;
+	phy->linkrate = SAS_LINK_RATE_UNKNOWN;
 }
 
 /* ---------- Tasks ---------- */
diff --git a/include/scsi/sas.h b/include/scsi/sas.h
index 752853a113dc..9c8a5b91ae64 100644
--- a/include/scsi/sas.h
+++ b/include/scsi/sas.h
@@ -102,20 +102,6 @@ enum sas_dev_type {
 	SATA_PM_PORT= 8,
 };
 
-enum sas_phy_linkrate {
-	PHY_LINKRATE_NONE = 0,
-	PHY_LINKRATE_UNKNOWN = 0,
-	PHY_DISABLED,
-	PHY_RESET_PROBLEM,
-	PHY_SPINUP_HOLD,
-	PHY_PORT_SELECTOR,
-	PHY_LINKRATE_1_5 = 0x08,
-	PHY_LINKRATE_G1  = PHY_LINKRATE_1_5,
-	PHY_LINKRATE_3   = 0x09,
-	PHY_LINKRATE_G2  = PHY_LINKRATE_3,
-	PHY_LINKRATE_6   = 0x0A,
-};
-
 /* Partly from IDENTIFY address frame. */
 enum sas_proto {
 	SATA_PROTO    = 1,
diff --git a/include/scsi/scsi_transport_sas.h b/include/scsi/scsi_transport_sas.h
index eeb2200de855..87de518960c1 100644
--- a/include/scsi/scsi_transport_sas.h
+++ b/include/scsi/scsi_transport_sas.h
@@ -24,15 +24,23 @@ enum sas_protocol {
 };
 
 enum sas_linkrate {
-	SAS_LINK_RATE_UNKNOWN,
-	SAS_PHY_DISABLED,
-	SAS_LINK_RATE_FAILED,
-	SAS_SATA_SPINUP_HOLD,
-	SAS_SATA_PORT_SELECTOR,
-	SAS_LINK_RATE_1_5_GBPS,
-	SAS_LINK_RATE_3_0_GBPS,
-	SAS_LINK_RATE_6_0_GBPS,
-	SAS_LINK_VIRTUAL,
+	/* These Values are defined in the SAS standard */
+	SAS_LINK_RATE_UNKNOWN = 0,
+	SAS_PHY_DISABLED = 1,
+	SAS_PHY_RESET_PROBLEM = 2,
+	SAS_SATA_SPINUP_HOLD = 3,
+	SAS_SATA_PORT_SELECTOR = 4,
+	SAS_PHY_RESET_IN_PROGRESS = 5,
+	SAS_LINK_RATE_1_5_GBPS = 8,
+	SAS_LINK_RATE_G1 = SAS_LINK_RATE_1_5_GBPS,
+	SAS_LINK_RATE_3_0_GBPS = 9,
+	SAS_LINK_RATE_G2 = SAS_LINK_RATE_3_0_GBPS,
+	SAS_LINK_RATE_6_0_GBPS = 10,
+	/* These are virtual to the transport class and may never
+	 * be signalled normally since the standard defined field
+	 * is only 4 bits */
+	SAS_LINK_RATE_FAILED = 0x10,
+	SAS_PHY_VIRTUAL = 0x11,
 };
 
 struct sas_identify {
-- 
cgit v1.2.3


From d24e1eeb3a16e4944288c2f3bf082e1513f4b425 Mon Sep 17 00:00:00 2001
From: James Bottomley <James.Bottomley@steeleye.com>
Date: Wed, 6 Sep 2006 19:25:22 -0500
Subject: [SCSI] scsi_transport_sas: make minimum and maximum linkrate settable
 quantities

According to SPEC, the minimum_linkrate and maximum_linkrate should be
settable by the user.  This patch introduces a callback that allows the
sas class to pass these settings on to the driver.

Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 drivers/scsi/scsi_transport_sas.c | 73 +++++++++++++++++++++++++++++++++++----
 include/scsi/scsi_transport_sas.h |  6 ++++
 2 files changed, 73 insertions(+), 6 deletions(-)

(limited to 'include')

diff --git a/drivers/scsi/scsi_transport_sas.c b/drivers/scsi/scsi_transport_sas.c
index d518c1207fb4..b5b0c2cba96b 100644
--- a/drivers/scsi/scsi_transport_sas.c
+++ b/drivers/scsi/scsi_transport_sas.c
@@ -77,6 +77,24 @@ get_sas_##title##_names(u32 table_key, char *buf)		\
 	return len;						\
 }
 
+#define sas_bitfield_name_set(title, table)			\
+static ssize_t							\
+set_sas_##title##_names(u32 *table_key, const char *buf)	\
+{								\
+	ssize_t len = 0;					\
+	int i;							\
+								\
+	for (i = 0; i < ARRAY_SIZE(table); i++) {		\
+		len = strlen(table[i].name);			\
+		if (strncmp(buf, table[i].name, len) == 0 &&	\
+		    (buf[len] == '\n' || buf[len] == '\0')) {	\
+			*table_key = table[i].value;		\
+			return 0;				\
+		}						\
+	}							\
+	return -EINVAL;						\
+}
+
 #define sas_bitfield_name_search(title, table)			\
 static ssize_t							\
 get_sas_##title##_names(u32 table_key, char *buf)		\
@@ -131,7 +149,7 @@ static struct {
 	{ SAS_LINK_RATE_6_0_GBPS,	"6.0 Gbit" },
 };
 sas_bitfield_name_search(linkspeed, sas_linkspeed_names)
-
+sas_bitfield_name_set(linkspeed, sas_linkspeed_names)
 
 /*
  * SAS host attributes
@@ -253,10 +271,39 @@ show_sas_phy_##field(struct class_device *cdev, char *buf)		\
 	return get_sas_linkspeed_names(phy->field, buf);		\
 }
 
+/* Fudge to tell if we're minimum or maximum */
+#define sas_phy_store_linkspeed(field)					\
+static ssize_t								\
+store_sas_phy_##field(struct class_device *cdev, const char *buf,	\
+		      size_t count)					\
+{									\
+	struct sas_phy *phy = transport_class_to_phy(cdev);		\
+	struct Scsi_Host *shost = dev_to_shost(phy->dev.parent);	\
+	struct sas_internal *i = to_sas_internal(shost->transportt);	\
+	u32 value;							\
+	struct sas_phy_linkrates rates = {0};				\
+	int error;							\
+									\
+	error = set_sas_linkspeed_names(&value, buf);			\
+	if (error)							\
+		return error;						\
+	rates.field = value;						\
+	error = i->f->set_phy_speed(phy, &rates);			\
+									\
+	return error ? error : count;					\
+}
+
+#define sas_phy_linkspeed_rw_attr(field)				\
+	sas_phy_show_linkspeed(field)					\
+	sas_phy_store_linkspeed(field)					\
+static CLASS_DEVICE_ATTR(field, S_IRUGO, show_sas_phy_##field,		\
+	store_sas_phy_##field)
+
 #define sas_phy_linkspeed_attr(field)					\
 	sas_phy_show_linkspeed(field)					\
 static CLASS_DEVICE_ATTR(field, S_IRUGO, show_sas_phy_##field, NULL)
 
+
 #define sas_phy_show_linkerror(field)					\
 static ssize_t								\
 show_sas_phy_##field(struct class_device *cdev, char *buf)		\
@@ -326,9 +373,9 @@ sas_phy_simple_attr(identify.phy_identifier, phy_identifier, "%d\n", u8);
 //sas_phy_simple_attr(port_identifier, port_identifier, "%d\n", int);
 sas_phy_linkspeed_attr(negotiated_linkrate);
 sas_phy_linkspeed_attr(minimum_linkrate_hw);
-sas_phy_linkspeed_attr(minimum_linkrate);
+sas_phy_linkspeed_rw_attr(minimum_linkrate);
 sas_phy_linkspeed_attr(maximum_linkrate_hw);
-sas_phy_linkspeed_attr(maximum_linkrate);
+sas_phy_linkspeed_rw_attr(maximum_linkrate);
 sas_phy_linkerror_attr(invalid_dword_count);
 sas_phy_linkerror_attr(running_disparity_error_count);
 sas_phy_linkerror_attr(loss_of_dword_sync_count);
@@ -1310,13 +1357,23 @@ static int sas_user_scan(struct Scsi_Host *shost, uint channel,
  * Setup / Teardown code
  */
 
-#define SETUP_TEMPLATE(attrb, field, perm, test)				\
+#define SETUP_TEMPLATE(attrb, field, perm, test)			\
 	i->private_##attrb[count] = class_device_attr_##field;		\
 	i->private_##attrb[count].attr.mode = perm;			\
 	i->attrb[count] = &i->private_##attrb[count];			\
 	if (test)							\
 		count++
 
+#define SETUP_TEMPLATE_RW(attrb, field, perm, test, ro_test, ro_perm)	\
+	i->private_##attrb[count] = class_device_attr_##field;		\
+	i->private_##attrb[count].attr.mode = perm;			\
+	if (ro_test) {							\
+		i->private_##attrb[count].attr.mode = ro_perm;		\
+		i->private_##attrb[count].store = NULL;			\
+	}								\
+	i->attrb[count] = &i->private_##attrb[count];			\
+	if (test)							\
+		count++
 
 #define SETUP_RPORT_ATTRIBUTE(field) 					\
 	SETUP_TEMPLATE(rphy_attrs, field, S_IRUGO, 1)
@@ -1327,6 +1384,10 @@ static int sas_user_scan(struct Scsi_Host *shost, uint channel,
 #define SETUP_PHY_ATTRIBUTE(field)					\
 	SETUP_TEMPLATE(phy_attrs, field, S_IRUGO, 1)
 
+#define SETUP_PHY_ATTRIBUTE_RW(field)					\
+	SETUP_TEMPLATE_RW(phy_attrs, field, S_IRUGO | S_IWUSR, 1,	\
+			!i->f->set_phy_speed, S_IRUGO)
+
 #define SETUP_PORT_ATTRIBUTE(field)					\
 	SETUP_TEMPLATE(port_attrs, field, S_IRUGO, 1)
 
@@ -1407,9 +1468,9 @@ sas_attach_transport(struct sas_function_template *ft)
 	//SETUP_PHY_ATTRIBUTE(port_identifier);
 	SETUP_PHY_ATTRIBUTE(negotiated_linkrate);
 	SETUP_PHY_ATTRIBUTE(minimum_linkrate_hw);
-	SETUP_PHY_ATTRIBUTE(minimum_linkrate);
+	SETUP_PHY_ATTRIBUTE_RW(minimum_linkrate);
 	SETUP_PHY_ATTRIBUTE(maximum_linkrate_hw);
-	SETUP_PHY_ATTRIBUTE(maximum_linkrate);
+	SETUP_PHY_ATTRIBUTE_RW(maximum_linkrate);
 
 	SETUP_PHY_ATTRIBUTE(invalid_dword_count);
 	SETUP_PHY_ATTRIBUTE(running_disparity_error_count);
diff --git a/include/scsi/scsi_transport_sas.h b/include/scsi/scsi_transport_sas.h
index 87de518960c1..53024377f3b8 100644
--- a/include/scsi/scsi_transport_sas.h
+++ b/include/scsi/scsi_transport_sas.h
@@ -150,12 +150,18 @@ struct sas_port {
 #define transport_class_to_sas_port(cdev) \
 	dev_to_sas_port((cdev)->dev)
 
+struct sas_phy_linkrates {
+	enum sas_linkrate maximum_linkrate;
+	enum sas_linkrate minimum_linkrate;
+};
+
 /* The functions by which the transport class and the driver communicate */
 struct sas_function_template {
 	int (*get_linkerrors)(struct sas_phy *);
 	int (*get_enclosure_identifier)(struct sas_rphy *, u64 *);
 	int (*get_bay_identifier)(struct sas_rphy *);
 	int (*phy_reset)(struct sas_phy *, int);
+	int (*set_phy_speed)(struct sas_phy *, struct sas_phy_linkrates *);
 };
 
 
-- 
cgit v1.2.3


From a01e70e570a72b8a8c9a58062e4f5bdcd3986222 Mon Sep 17 00:00:00 2001
From: James Bottomley <James.Bottomley@steeleye.com>
Date: Wed, 6 Sep 2006 19:28:07 -0500
Subject: [SCSI] aci94xx: implement link rate setting

This patch implements the ability to set the minimum and maximum
linkrates for both libsas (for expanders) and aic94xx (for the host
phys).  It also tidies up the setting of the hardware min and max to
make sure they're updated when the expander emits a change broadcast.

Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 drivers/scsi/aic94xx/aic94xx.h     |  2 +-
 drivers/scsi/aic94xx/aic94xx_scb.c | 30 ++++++++++++++++++++++++--
 drivers/scsi/libsas/sas_expander.c | 20 +++++++++++------
 drivers/scsi/libsas/sas_init.c     | 44 ++++++++++++++++++++++++++++++++++++--
 drivers/scsi/libsas/sas_internal.h |  2 +-
 drivers/scsi/libsas/sas_phy.c      | 15 +++++++------
 include/scsi/libsas.h              |  2 +-
 include/scsi/sas.h                 |  1 +
 8 files changed, 95 insertions(+), 21 deletions(-)

(limited to 'include')

diff --git a/drivers/scsi/aic94xx/aic94xx.h b/drivers/scsi/aic94xx/aic94xx.h
index cb7caf1c9ce1..1bd5b4ecf3d5 100644
--- a/drivers/scsi/aic94xx/aic94xx.h
+++ b/drivers/scsi/aic94xx/aic94xx.h
@@ -109,6 +109,6 @@ int  asd_clear_nexus_port(struct asd_sas_port *port);
 int  asd_clear_nexus_ha(struct sas_ha_struct *sas_ha);
 
 /* ---------- Phy Management ---------- */
-int  asd_control_phy(struct asd_sas_phy *phy, enum phy_func func);
+int  asd_control_phy(struct asd_sas_phy *phy, enum phy_func func, void *arg);
 
 #endif
diff --git a/drivers/scsi/aic94xx/aic94xx_scb.c b/drivers/scsi/aic94xx/aic94xx_scb.c
index ef8ca08b545f..7ee49b51b724 100644
--- a/drivers/scsi/aic94xx/aic94xx_scb.c
+++ b/drivers/scsi/aic94xx/aic94xx_scb.c
@@ -52,6 +52,8 @@
 
 static inline void get_lrate_mode(struct asd_phy *phy, u8 oob_mode)
 {
+	struct sas_phy *sas_phy = phy->sas_phy.phy;
+
 	switch (oob_mode & 7) {
 	case PHY_SPEED_60:
 		/* FIXME: sas transport class doesn't have this */
@@ -67,6 +69,12 @@ static inline void get_lrate_mode(struct asd_phy *phy, u8 oob_mode)
 		phy->sas_phy.phy->negotiated_linkrate = SAS_LINK_RATE_1_5_GBPS;
 		break;
 	}
+	sas_phy->negotiated_linkrate = phy->sas_phy.linkrate;
+	sas_phy->maximum_linkrate_hw = SAS_LINK_RATE_3_0_GBPS;
+	sas_phy->minimum_linkrate_hw = SAS_LINK_RATE_1_5_GBPS;
+	sas_phy->maximum_linkrate = phy->phy_desc->max_sas_lrate;
+	sas_phy->minimum_linkrate = phy->phy_desc->min_sas_lrate;
+
 	if (oob_mode & SAS_MODE)
 		phy->sas_phy.oob_mode = SAS_OOB_MODE;
 	else if (oob_mode & SATA_MODE)
@@ -710,14 +718,32 @@ static const int phy_func_table[] = {
 	[PHY_FUNC_RELEASE_SPINUP_HOLD] = RELEASE_SPINUP_HOLD,
 };
 
-int asd_control_phy(struct asd_sas_phy *phy, enum phy_func func)
+int asd_control_phy(struct asd_sas_phy *phy, enum phy_func func, void *arg)
 {
 	struct asd_ha_struct *asd_ha = phy->ha->lldd_ha;
+	struct asd_phy_desc *pd = asd_ha->phys[phy->id].phy_desc;
 	struct asd_ascb *ascb;
+	struct sas_phy_linkrates *rates;
 	int res = 1;
 
-	if (func == PHY_FUNC_CLEAR_ERROR_LOG)
+	switch (func) {
+	case PHY_FUNC_CLEAR_ERROR_LOG:
 		return -ENOSYS;
+	case PHY_FUNC_SET_LINK_RATE:
+		rates = arg;
+		if (rates->minimum_linkrate) {
+			pd->min_sas_lrate = rates->minimum_linkrate;
+			pd->min_sata_lrate = rates->minimum_linkrate;
+		}
+		if (rates->maximum_linkrate) {
+			pd->max_sas_lrate = rates->maximum_linkrate;
+			pd->max_sata_lrate = rates->maximum_linkrate;
+		}
+		func = PHY_FUNC_LINK_RESET;
+		break;
+	default:
+		break;
+	}
 
 	ascb = asd_ascb_alloc_list(asd_ha, &res, GFP_KERNEL);
 	if (!ascb)
diff --git a/drivers/scsi/libsas/sas_expander.c b/drivers/scsi/libsas/sas_expander.c
index 02e796ee027e..30b8014bcc7a 100644
--- a/drivers/scsi/libsas/sas_expander.c
+++ b/drivers/scsi/libsas/sas_expander.c
@@ -187,10 +187,10 @@ static void sas_set_ex_phy(struct domain_device *dev, int phy_id,
 	phy->phy->identify.initiator_port_protocols = phy->attached_iproto;
 	phy->phy->identify.target_port_protocols = phy->attached_tproto;
 	phy->phy->identify.phy_identifier = phy_id;
-	phy->phy->minimum_linkrate_hw = SAS_LINK_RATE_1_5_GBPS;
-	phy->phy->maximum_linkrate_hw = SAS_LINK_RATE_3_0_GBPS;
-	phy->phy->minimum_linkrate = SAS_LINK_RATE_1_5_GBPS;
-	phy->phy->maximum_linkrate = SAS_LINK_RATE_3_0_GBPS;
+	phy->phy->minimum_linkrate_hw = dr->hmin_linkrate;
+	phy->phy->maximum_linkrate_hw = dr->hmax_linkrate;
+	phy->phy->minimum_linkrate = dr->pmin_linkrate;
+	phy->phy->maximum_linkrate = dr->pmax_linkrate;
 	phy->phy->negotiated_linkrate = phy->linkrate;
 
 	if (!rediscover)
@@ -404,7 +404,8 @@ out:
 #define PC_RESP_SIZE 8
 
 int sas_smp_phy_control(struct domain_device *dev, int phy_id,
-			enum phy_func phy_func)
+			enum phy_func phy_func,
+			struct sas_phy_linkrates *rates)
 {
 	u8 *pc_req;
 	u8 *pc_resp;
@@ -423,6 +424,10 @@ int sas_smp_phy_control(struct domain_device *dev, int phy_id,
 	pc_req[1] = SMP_PHY_CONTROL;
 	pc_req[9] = phy_id;
 	pc_req[10]= phy_func;
+	if (rates) {
+		pc_req[32] = rates->minimum_linkrate << 4;
+		pc_req[33] = rates->maximum_linkrate << 4;
+	}
 
 	res = smp_execute_task(dev, pc_req, PC_REQ_SIZE, pc_resp,PC_RESP_SIZE);
 
@@ -436,7 +441,7 @@ static void sas_ex_disable_phy(struct domain_device *dev, int phy_id)
 	struct expander_device *ex = &dev->ex_dev;
 	struct ex_phy *phy = &ex->ex_phy[phy_id];
 
-	sas_smp_phy_control(dev, phy_id, PHY_FUNC_DISABLE);
+	sas_smp_phy_control(dev, phy_id, PHY_FUNC_DISABLE, NULL);
 	phy->linkrate = SAS_PHY_DISABLED;
 }
 
@@ -731,7 +736,7 @@ static int sas_ex_discover_dev(struct domain_device *dev, int phy_id)
 
 	/* Phy state */
 	if (ex_phy->linkrate == SAS_SATA_SPINUP_HOLD) {
-		if (!sas_smp_phy_control(dev, phy_id, PHY_FUNC_LINK_RESET))
+		if (!sas_smp_phy_control(dev, phy_id, PHY_FUNC_LINK_RESET, NULL))
 			res = sas_ex_phy_discover(dev, phy_id);
 		if (res)
 			return res;
@@ -1706,6 +1711,7 @@ static int sas_rediscover_dev(struct domain_device *dev, int phy_id)
 		   SAS_ADDR(phy->attached_sas_addr)) {
 		SAS_DPRINTK("ex %016llx phy 0x%x broadcast flutter\n",
 			    SAS_ADDR(dev->sas_addr), phy_id);
+		sas_ex_phy_discover(dev, phy_id);
 	} else
 		res = sas_discover_new(dev, phy_id);
 out:
diff --git a/drivers/scsi/libsas/sas_init.c b/drivers/scsi/libsas/sas_init.c
index b961664b8106..c836a237fb79 100644
--- a/drivers/scsi/libsas/sas_init.c
+++ b/drivers/scsi/libsas/sas_init.c
@@ -159,17 +159,57 @@ static int sas_phy_reset(struct sas_phy *phy, int hard_reset)
 		struct sas_internal *i =
 			to_sas_internal(sas_ha->core.shost->transportt);
 
-		ret = i->dft->lldd_control_phy(asd_phy, reset_type);
+		ret = i->dft->lldd_control_phy(asd_phy, reset_type, NULL);
 	} else {
 		struct sas_rphy *rphy = dev_to_rphy(phy->dev.parent);
 		struct domain_device *ddev = sas_find_dev_by_rphy(rphy);
-		ret = sas_smp_phy_control(ddev, phy->number, reset_type);
+		ret = sas_smp_phy_control(ddev, phy->number, reset_type, NULL);
 	}
 	return ret;
 }
 
+static int sas_set_phy_speed(struct sas_phy *phy,
+			     struct sas_phy_linkrates *rates)
+{
+	int ret;
+
+	if ((rates->minimum_linkrate &&
+	     rates->minimum_linkrate > phy->maximum_linkrate) ||
+	    (rates->maximum_linkrate &&
+	     rates->maximum_linkrate < phy->minimum_linkrate))
+		return -EINVAL;
+
+	if (rates->minimum_linkrate &&
+	    rates->minimum_linkrate < phy->minimum_linkrate_hw)
+		rates->minimum_linkrate = phy->minimum_linkrate_hw;
+
+	if (rates->maximum_linkrate &&
+	    rates->maximum_linkrate > phy->maximum_linkrate_hw)
+		rates->maximum_linkrate = phy->maximum_linkrate_hw;
+
+	if (scsi_is_sas_phy_local(phy)) {
+		struct Scsi_Host *shost = dev_to_shost(phy->dev.parent);
+		struct sas_ha_struct *sas_ha = SHOST_TO_SAS_HA(shost);
+		struct asd_sas_phy *asd_phy = sas_ha->sas_phy[phy->number];
+		struct sas_internal *i =
+			to_sas_internal(sas_ha->core.shost->transportt);
+
+		ret = i->dft->lldd_control_phy(asd_phy, PHY_FUNC_SET_LINK_RATE,
+					       rates);
+	} else {
+		struct sas_rphy *rphy = dev_to_rphy(phy->dev.parent);
+		struct domain_device *ddev = sas_find_dev_by_rphy(rphy);
+		ret = sas_smp_phy_control(ddev, phy->number,
+					  PHY_FUNC_LINK_RESET, rates);
+
+	}
+
+	return ret;
+}
+
 static struct sas_function_template sft = {
 	.phy_reset = sas_phy_reset,
+	.set_phy_speed = sas_set_phy_speed,
 	.get_linkerrors = sas_get_linkerrors,
 };
 
diff --git a/drivers/scsi/libsas/sas_internal.h b/drivers/scsi/libsas/sas_internal.h
index 0d69ede4b944..bffcee474921 100644
--- a/drivers/scsi/libsas/sas_internal.h
+++ b/drivers/scsi/libsas/sas_internal.h
@@ -70,7 +70,7 @@ int sas_notify_lldd_dev_found(struct domain_device *);
 void sas_notify_lldd_dev_gone(struct domain_device *);
 
 int sas_smp_phy_control(struct domain_device *dev, int phy_id,
-			enum phy_func phy_func);
+			enum phy_func phy_func, struct sas_phy_linkrates *);
 int sas_smp_get_phy_events(struct sas_phy *phy);
 
 struct domain_device *sas_find_dev_by_rphy(struct sas_rphy *rphy);
diff --git a/drivers/scsi/libsas/sas_phy.c b/drivers/scsi/libsas/sas_phy.c
index 024ab00e70d2..9340cdbae4a3 100644
--- a/drivers/scsi/libsas/sas_phy.c
+++ b/drivers/scsi/libsas/sas_phy.c
@@ -67,13 +67,14 @@ static void sas_phye_oob_error(void *data)
 		switch (phy->error) {
 		case 1:
 		case 2:
-			i->dft->lldd_control_phy(phy, PHY_FUNC_HARD_RESET);
+			i->dft->lldd_control_phy(phy, PHY_FUNC_HARD_RESET,
+						 NULL);
 			break;
 		case 3:
 		default:
 			phy->error = 0;
 			phy->enabled = 0;
-			i->dft->lldd_control_phy(phy, PHY_FUNC_DISABLE);
+			i->dft->lldd_control_phy(phy, PHY_FUNC_DISABLE, NULL);
 			break;
 		}
 	}
@@ -90,7 +91,7 @@ static void sas_phye_spinup_hold(void *data)
 			&phy->phy_events_pending);
 
 	phy->error = 0;
-	i->dft->lldd_control_phy(phy, PHY_FUNC_RELEASE_SPINUP_HOLD);
+	i->dft->lldd_control_phy(phy, PHY_FUNC_RELEASE_SPINUP_HOLD, NULL);
 }
 
 /* ---------- Phy class registration ---------- */
@@ -144,10 +145,10 @@ int sas_register_phys(struct sas_ha_struct *sas_ha)
 		phy->phy->identify.target_port_protocols = phy->tproto;
 		phy->phy->identify.sas_address = SAS_ADDR(sas_ha->sas_addr);
 		phy->phy->identify.phy_identifier = i;
-		phy->phy->minimum_linkrate_hw = SAS_LINK_RATE_1_5_GBPS;
-		phy->phy->maximum_linkrate_hw = SAS_LINK_RATE_3_0_GBPS;
-		phy->phy->minimum_linkrate = SAS_LINK_RATE_1_5_GBPS;
-		phy->phy->maximum_linkrate = SAS_LINK_RATE_3_0_GBPS;
+		phy->phy->minimum_linkrate_hw = SAS_LINK_RATE_UNKNOWN;
+		phy->phy->maximum_linkrate_hw = SAS_LINK_RATE_UNKNOWN;
+		phy->phy->minimum_linkrate = SAS_LINK_RATE_UNKNOWN;
+		phy->phy->maximum_linkrate = SAS_LINK_RATE_UNKNOWN;
 		phy->phy->negotiated_linkrate = SAS_LINK_RATE_UNKNOWN;
 
 		sas_phy_add(phy->phy);
diff --git a/include/scsi/libsas.h b/include/scsi/libsas.h
index 8d91313dd888..8e39982fc3db 100644
--- a/include/scsi/libsas.h
+++ b/include/scsi/libsas.h
@@ -586,7 +586,7 @@ struct sas_domain_function_template {
 	int (*lldd_clear_nexus_ha)(struct sas_ha_struct *);
 
 	/* Phy management */
-	int (*lldd_control_phy)(struct asd_sas_phy *, enum phy_func);
+	int (*lldd_control_phy)(struct asd_sas_phy *, enum phy_func, void *);
 };
 
 extern int sas_register_ha(struct sas_ha_struct *);
diff --git a/include/scsi/sas.h b/include/scsi/sas.h
index 9c8a5b91ae64..2f4b6afa34fc 100644
--- a/include/scsi/sas.h
+++ b/include/scsi/sas.h
@@ -121,6 +121,7 @@ enum phy_func {
 	PHY_FUNC_CLEAR_AFFIL,
 	PHY_FUNC_TX_SATA_PS_SIGNAL,
 	PHY_FUNC_RELEASE_SPINUP_HOLD = 0x10, /* LOCAL PORT ONLY! */
+	PHY_FUNC_SET_LINK_RATE,
 };
 
 /* SAS LLDD would need to report only _very_few_ of those, like BROADCAST.
-- 
cgit v1.2.3


From 0024300000769eadcb4a4fcdff531d45ee7735d4 Mon Sep 17 00:00:00 2001
From: Olof Johansson <olof@lixom.net>
Date: Wed, 6 Sep 2006 14:35:19 -0500
Subject: [POWERPC] powerpc: Divorce CPU_FTR_CTRL from
 CPU_FTR_PPCAS_ARCH_V2_BASE

The performance monitor implementation (including CTRL register behaviour)
is just included in PPC v2 as an example, it's not truly part of the base.

It's actually a somewhat misleading feature, but I'll leave that be for
now: The presence of the register is not what the feature bit is used
for, but instead it's used to determine if it contains the runlatch
bit for idle reporting of the performance monitor. For alternative
implementations, the register might still exist but the bit might have
different meaning (or no meaning at all).

For now, split it off and don't include it in CPU_FTR_PPCAS_ARCH_V2_BASE.

Signed-off-by: Olof Johansson <olof@lixom.net>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 include/asm-powerpc/cputable.h | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)

(limited to 'include')

diff --git a/include/asm-powerpc/cputable.h b/include/asm-powerpc/cputable.h
index 748bc1805da9..3608259c49cf 100644
--- a/include/asm-powerpc/cputable.h
+++ b/include/asm-powerpc/cputable.h
@@ -148,7 +148,7 @@ extern void do_cpu_ftr_fixups(unsigned long offset);
 
 #define CPU_FTR_PPCAS_ARCH_V2_BASE (CPU_FTR_SLB | \
 					CPU_FTR_TLBIEL | CPU_FTR_NOEXECUTE | \
-					CPU_FTR_NODSISRALIGN | CPU_FTR_CTRL)
+					CPU_FTR_NODSISRALIGN)
 
 /* iSeries doesn't support large pages */
 #ifdef CONFIG_PPC_ISERIES
@@ -313,24 +313,25 @@ extern void do_cpu_ftr_fixups(unsigned long offset);
 	    CPU_FTR_HPTE_TABLE | CPU_FTR_IABR | \
 	    CPU_FTR_MMCRA | CPU_FTR_CTRL)
 #define CPU_FTRS_POWER4	(CPU_FTR_SPLIT_ID_CACHE | CPU_FTR_USE_TB | \
-	    CPU_FTR_HPTE_TABLE | CPU_FTR_PPCAS_ARCH_V2 | CPU_FTR_MMCRA)
+	    CPU_FTR_HPTE_TABLE | CPU_FTR_PPCAS_ARCH_V2 | CPU_FTR_CTRL | \
+	    CPU_FTR_MMCRA)
 #define CPU_FTRS_PPC970	(CPU_FTR_SPLIT_ID_CACHE | CPU_FTR_USE_TB | \
-	    CPU_FTR_HPTE_TABLE | CPU_FTR_PPCAS_ARCH_V2 | \
+	    CPU_FTR_HPTE_TABLE | CPU_FTR_PPCAS_ARCH_V2 | CPU_FTR_CTRL | \
 	    CPU_FTR_ALTIVEC_COMP | CPU_FTR_CAN_NAP | CPU_FTR_MMCRA)
 #define CPU_FTRS_POWER5	(CPU_FTR_SPLIT_ID_CACHE | CPU_FTR_USE_TB | \
-	    CPU_FTR_HPTE_TABLE | CPU_FTR_PPCAS_ARCH_V2 | \
+	    CPU_FTR_HPTE_TABLE | CPU_FTR_PPCAS_ARCH_V2 | CPU_FTR_CTRL | \
 	    CPU_FTR_MMCRA | CPU_FTR_SMT | \
 	    CPU_FTR_COHERENT_ICACHE | CPU_FTR_LOCKLESS_TLBIE | \
 	    CPU_FTR_PURR)
 #define CPU_FTRS_POWER6 (CPU_FTR_SPLIT_ID_CACHE | CPU_FTR_USE_TB | \
-	    CPU_FTR_HPTE_TABLE | CPU_FTR_PPCAS_ARCH_V2 | \
+	    CPU_FTR_HPTE_TABLE | CPU_FTR_PPCAS_ARCH_V2 | CPU_FTR_CTRL | \
 	    CPU_FTR_MMCRA | CPU_FTR_SMT | \
 	    CPU_FTR_COHERENT_ICACHE | CPU_FTR_LOCKLESS_TLBIE | \
 	    CPU_FTR_PURR | CPU_FTR_CI_LARGE_PAGE | CPU_FTR_REAL_LE)
 #define CPU_FTRS_CELL	(CPU_FTR_SPLIT_ID_CACHE | CPU_FTR_USE_TB | \
-	    CPU_FTR_HPTE_TABLE | CPU_FTR_PPCAS_ARCH_V2 | \
+	    CPU_FTR_HPTE_TABLE | CPU_FTR_PPCAS_ARCH_V2 | CPU_FTR_CTRL | \
 	    CPU_FTR_ALTIVEC_COMP | CPU_FTR_MMCRA | CPU_FTR_SMT | \
-	    CPU_FTR_CTRL | CPU_FTR_PAUSE_ZERO | CPU_FTR_CI_LARGE_PAGE)
+	    CPU_FTR_PAUSE_ZERO | CPU_FTR_CI_LARGE_PAGE)
 #define CPU_FTRS_COMPATIBLE	(CPU_FTR_SPLIT_ID_CACHE | CPU_FTR_USE_TB | \
 	    CPU_FTR_HPTE_TABLE | CPU_FTR_PPCAS_ARCH_V2)
 #endif
-- 
cgit v1.2.3


From b3ebd1d862d6c23caa58e40d341eefc426f835e1 Mon Sep 17 00:00:00 2001
From: Olof Johansson <olof@lixom.net>
Date: Wed, 6 Sep 2006 14:35:57 -0500
Subject: [POWERPC] powerpc: PA6T cputable entry, PVR value

Introduce PWRficient PA6T cputable entries and feature bits.

Signed-off-by: Olof Johansson <olof@lixom.net>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 arch/powerpc/kernel/cputable.c | 14 ++++++++++++++
 include/asm-powerpc/cputable.h |  9 +++++++--
 include/asm-powerpc/reg.h      |  1 +
 3 files changed, 22 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/kernel/cputable.c b/arch/powerpc/kernel/cputable.c
index 306da4cd37a0..db65c9f6559a 100644
--- a/arch/powerpc/kernel/cputable.c
+++ b/arch/powerpc/kernel/cputable.c
@@ -58,6 +58,9 @@ extern void __restore_cpu_ppc970(void);
 #define COMMON_USER_POWER6	(COMMON_USER_PPC64 | PPC_FEATURE_ARCH_2_05 |\
 				 PPC_FEATURE_SMT | PPC_FEATURE_ICACHE_SNOOP | \
 				 PPC_FEATURE_TRUE_LE)
+#define COMMON_USER_PA6T	(COMMON_USER_PPC64 | PPC_FEATURE_PA6T |\
+				 PPC_FEATURE_TRUE_LE | \
+				 PPC_FEATURE_HAS_ALTIVEC_COMP)
 #define COMMON_USER_BOOKE	(PPC_FEATURE_32 | PPC_FEATURE_HAS_MMU | \
 				 PPC_FEATURE_BOOKE)
 
@@ -286,6 +289,17 @@ struct cpu_spec	cpu_specs[] = {
 		.dcache_bsize		= 128,
 		.platform		= "ppc-cell-be",
 	},
+	{	/* PA Semi PA6T */
+		.pvr_mask		= 0x7fff0000,
+		.pvr_value		= 0x00900000,
+		.cpu_name		= "PA6T",
+		.cpu_features		= CPU_FTRS_PA6T,
+		.cpu_user_features	= COMMON_USER_PA6T,
+		.icache_bsize		= 64,
+		.dcache_bsize		= 64,
+		.num_pmcs		= 6,
+		.platform		= "pa6t",
+	},
 	{	/* default match */
 		.pvr_mask		= 0x00000000,
 		.pvr_value		= 0x00000000,
diff --git a/include/asm-powerpc/cputable.h b/include/asm-powerpc/cputable.h
index 3608259c49cf..12707ab9dc98 100644
--- a/include/asm-powerpc/cputable.h
+++ b/include/asm-powerpc/cputable.h
@@ -23,6 +23,7 @@
 #define PPC_FEATURE_SMT			0x00004000
 #define PPC_FEATURE_ICACHE_SNOOP	0x00002000
 #define PPC_FEATURE_ARCH_2_05		0x00001000
+#define PPC_FEATURE_PA6T		0x00000800
 
 #define PPC_FEATURE_TRUE_LE		0x00000002
 #define PPC_FEATURE_PPC_LE		0x00000001
@@ -332,6 +333,10 @@ extern void do_cpu_ftr_fixups(unsigned long offset);
 	    CPU_FTR_HPTE_TABLE | CPU_FTR_PPCAS_ARCH_V2 | CPU_FTR_CTRL | \
 	    CPU_FTR_ALTIVEC_COMP | CPU_FTR_MMCRA | CPU_FTR_SMT | \
 	    CPU_FTR_PAUSE_ZERO | CPU_FTR_CI_LARGE_PAGE)
+#define CPU_FTRS_PA6T (CPU_FTR_SPLIT_ID_CACHE | CPU_FTR_USE_TB | \
+	    CPU_FTR_HPTE_TABLE | CPU_FTR_PPCAS_ARCH_V2 | \
+	    CPU_FTR_ALTIVEC_COMP | CPU_FTR_CI_LARGE_PAGE | \
+	    CPU_FTR_PURR | CPU_FTR_REAL_LE)
 #define CPU_FTRS_COMPATIBLE	(CPU_FTR_SPLIT_ID_CACHE | CPU_FTR_USE_TB | \
 	    CPU_FTR_HPTE_TABLE | CPU_FTR_PPCAS_ARCH_V2)
 #endif
@@ -340,7 +345,7 @@ extern void do_cpu_ftr_fixups(unsigned long offset);
 #define CPU_FTRS_POSSIBLE	\
 	    (CPU_FTRS_POWER3 | CPU_FTRS_RS64 | CPU_FTRS_POWER4 |	\
 	    CPU_FTRS_PPC970 | CPU_FTRS_POWER5 | CPU_FTRS_POWER6 |	\
-	    CPU_FTRS_CELL | CPU_FTR_CI_LARGE_PAGE)
+	    CPU_FTRS_CELL | CPU_FTRS_PA6T)
 #else
 enum {
 	CPU_FTRS_POSSIBLE =
@@ -379,7 +384,7 @@ enum {
 #define CPU_FTRS_ALWAYS		\
 	    (CPU_FTRS_POWER3 & CPU_FTRS_RS64 & CPU_FTRS_POWER4 &	\
 	    CPU_FTRS_PPC970 & CPU_FTRS_POWER5 & CPU_FTRS_POWER6 &	\
-	    CPU_FTRS_CELL & CPU_FTRS_POSSIBLE)
+	    CPU_FTRS_CELL & CPU_FTRS_PA6T & CPU_FTRS_POSSIBLE)
 #else
 enum {
 	CPU_FTRS_ALWAYS =
diff --git a/include/asm-powerpc/reg.h b/include/asm-powerpc/reg.h
index cf73475a0c69..3a9fcc15811b 100644
--- a/include/asm-powerpc/reg.h
+++ b/include/asm-powerpc/reg.h
@@ -592,6 +592,7 @@
 #define PV_630p	0x0041
 #define PV_970MP	0x0044
 #define PV_BE		0x0070
+#define PV_PA6T		0x0090
 
 /*
  * Number of entries in the SLB. If this ever changes we should handle
-- 
cgit v1.2.3


From 57852a853b0d6761f270be0961d5d8387e98c8bb Mon Sep 17 00:00:00 2001
From: Mike Kravetz <kravetz@us.ibm.com>
Date: Wed, 6 Sep 2006 16:23:12 -0700
Subject: [POWERPC] powerpc: Instrument Hypervisor Calls

Add instrumentation for hypervisor calls on pseries.  Call statistics
include number of calls, wall time and cpu cycles (if available) and
are made available via debugfs.  Instrumentation code is behind the
HCALL_STATS config option and has no impact if not enabled.

Signed-off-by: Mike Kravetz <kravetz@us.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 arch/powerpc/Kconfig.debug                   |  14 +++
 arch/powerpc/kernel/asm-offsets.c            |   7 ++
 arch/powerpc/platforms/pseries/Makefile      |   1 +
 arch/powerpc/platforms/pseries/hvCall.S      |  72 +++++++++++++++
 arch/powerpc/platforms/pseries/hvCall_inst.c | 129 +++++++++++++++++++++++++++
 include/asm-powerpc/hvcall.h                 |  12 ++-
 6 files changed, 234 insertions(+), 1 deletion(-)
 create mode 100644 arch/powerpc/platforms/pseries/hvCall_inst.c

(limited to 'include')

diff --git a/arch/powerpc/Kconfig.debug b/arch/powerpc/Kconfig.debug
index e29ef77d3b00..d7b2aedd89aa 100644
--- a/arch/powerpc/Kconfig.debug
+++ b/arch/powerpc/Kconfig.debug
@@ -18,6 +18,20 @@ config DEBUG_STACK_USAGE
 
 	  This option will slow down process creation somewhat.
 
+config HCALL_STATS
+	bool "Hypervisor call instrumentation"
+	depends on PPC_PSERIES && DEBUG_FS
+	help
+	  Adds code to keep track of the number of hypervisor calls made and
+	  the amount of time spent in hypervisor callsr.  Wall time spent in
+	  each call is always calculated, and if available CPU cycles spent
+	  are also calculated.  A directory named hcall_inst is added at the
+	  root of the debugfs filesystem.  Within the hcall_inst directory
+	  are files that contain CPU specific call statistics.
+
+	  This option will add a small amount of overhead to all hypervisor
+	  calls.
+
 config DEBUGGER
 	bool "Enable debugger hooks"
 	depends on DEBUG_KERNEL
diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c
index c53acd2a6dfc..c578e7ab8173 100644
--- a/arch/powerpc/kernel/asm-offsets.c
+++ b/arch/powerpc/kernel/asm-offsets.c
@@ -137,6 +137,7 @@ int main(void)
 	DEFINE(PACA_USER_TIME, offsetof(struct paca_struct, user_time));
 	DEFINE(PACA_SYSTEM_TIME, offsetof(struct paca_struct, system_time));
 	DEFINE(PACA_SLBSHADOWPTR, offsetof(struct paca_struct, slb_shadow_ptr));
+	DEFINE(PACA_DATA_OFFSET, offsetof(struct paca_struct, data_offset));
 
 	DEFINE(SLBSHADOW_STACKVSID,
 	       offsetof(struct slb_shadow, save_area[SLB_NUM_BOLTED - 1].vsid));
@@ -165,6 +166,12 @@ int main(void)
 	/* Create extra stack space for SRR0 and SRR1 when calling prom/rtas. */
 	DEFINE(PROM_FRAME_SIZE, STACK_FRAME_OVERHEAD + sizeof(struct pt_regs) + 16);
 	DEFINE(RTAS_FRAME_SIZE, STACK_FRAME_OVERHEAD + sizeof(struct pt_regs) + 16);
+
+	/* hcall statistics */
+	DEFINE(HCALL_STAT_SIZE, sizeof(struct hcall_stats));
+	DEFINE(HCALL_STAT_CALLS, offsetof(struct hcall_stats, num_calls));
+	DEFINE(HCALL_STAT_TB, offsetof(struct hcall_stats, tb_total));
+	DEFINE(HCALL_STAT_PURR, offsetof(struct hcall_stats, purr_total));
 #endif /* CONFIG_PPC64 */
 	DEFINE(GPR0, STACK_FRAME_OVERHEAD+offsetof(struct pt_regs, gpr[0]));
 	DEFINE(GPR1, STACK_FRAME_OVERHEAD+offsetof(struct pt_regs, gpr[1]));
diff --git a/arch/powerpc/platforms/pseries/Makefile b/arch/powerpc/platforms/pseries/Makefile
index e5e0ff466904..997243a91be8 100644
--- a/arch/powerpc/platforms/pseries/Makefile
+++ b/arch/powerpc/platforms/pseries/Makefile
@@ -12,3 +12,4 @@ obj-$(CONFIG_EEH)	+= eeh.o eeh_cache.o eeh_driver.o eeh_event.o
 
 obj-$(CONFIG_HVC_CONSOLE)	+= hvconsole.o
 obj-$(CONFIG_HVCS)		+= hvcserver.o
+obj-$(CONFIG_HCALL_STATS)	+= hvCall_inst.o
diff --git a/arch/powerpc/platforms/pseries/hvCall.S b/arch/powerpc/platforms/pseries/hvCall.S
index 9a99b056bd27..c00cfed7af2c 100644
--- a/arch/powerpc/platforms/pseries/hvCall.S
+++ b/arch/powerpc/platforms/pseries/hvCall.S
@@ -10,9 +10,69 @@
 #include <asm/hvcall.h>
 #include <asm/processor.h>
 #include <asm/ppc_asm.h>
+#include <asm/asm-offsets.h>
 	
 #define STK_PARM(i)     (48 + ((i)-3)*8)
 
+#ifdef CONFIG_HCALL_STATS
+/*
+ * precall must preserve all registers.  use unused STK_PARM()
+ * areas to save snapshots and opcode.
+ */
+#define HCALL_INST_PRECALL					\
+	std	r3,STK_PARM(r3)(r1);	/* save opcode */	\
+	mftb	r0;			/* get timebase and */	\
+	std     r0,STK_PARM(r5)(r1);	/* save for later */	\
+BEGIN_FTR_SECTION;						\
+	mfspr	r0,SPRN_PURR;		/* get PURR and */	\
+	std	r0,STK_PARM(r6)(r1);	/* save for later */	\
+END_FTR_SECTION_IFCLR(CPU_FTR_PURR);
+	
+/*
+ * postcall is performed immediately before function return which
+ * allows liberal use of volatile registers.
+ */
+#define HCALL_INST_POSTCALL					\
+	ld	r4,STK_PARM(r3)(r1);	/* validate opcode */	\
+	cmpldi	cr7,r4,MAX_HCALL_OPCODE;			\
+	bgt-	cr7,1f;						\
+								\
+	/* get time and PURR snapshots after hcall */		\
+	mftb	r7;			/* timebase after */	\
+BEGIN_FTR_SECTION;						\
+	mfspr	r8,SPRN_PURR;		/* PURR after */	\
+	ld	r6,STK_PARM(r6)(r1);	/* PURR before */	\
+	subf	r6,r6,r8;		/* delta */		\
+END_FTR_SECTION_IFCLR(CPU_FTR_PURR);				\
+	ld	r5,STK_PARM(r5)(r1);	/* timebase before */	\
+	subf	r5,r5,r7;		/* time delta */	\
+								\
+	/* calculate address of stat structure r4 = opcode */	\
+	srdi	r4,r4,2;		/* index into array */	\
+	mulli	r4,r4,HCALL_STAT_SIZE;				\
+	LOAD_REG_ADDR(r7, per_cpu__hcall_stats);		\
+	add	r4,r4,r7;					\
+	ld	r7,PACA_DATA_OFFSET(r13); /* per cpu offset */	\
+	add	r4,r4,r7;					\
+								\
+	/* update stats	*/					\
+	ld	r7,HCALL_STAT_CALLS(r4); /* count */		\
+	addi	r7,r7,1;					\
+	std	r7,HCALL_STAT_CALLS(r4);			\
+	ld      r7,HCALL_STAT_TB(r4);	/* timebase */		\
+	add	r7,r7,r5;					\
+	std	r7,HCALL_STAT_TB(r4);				\
+BEGIN_FTR_SECTION;						\
+	ld	r7,HCALL_STAT_PURR(r4);	/* PURR */		\
+	add	r7,r7,r6;					\
+	std	r7,HCALL_STAT_PURR(r4);				\
+END_FTR_SECTION_IFCLR(CPU_FTR_PURR);				\
+1:
+#else
+#define HCALL_INST_PRECALL
+#define HCALL_INST_POSTCALL
+#endif
+
 	.text
 
 _GLOBAL(plpar_hcall_norets)
@@ -21,8 +81,12 @@ _GLOBAL(plpar_hcall_norets)
 	mfcr	r0
 	stw	r0,8(r1)
 
+	HCALL_INST_PRECALL
+
 	HVSC				/* invoke the hypervisor */
 
+	HCALL_INST_POSTCALL
+
 	lwz	r0,8(r1)
 	mtcrf	0xff,r0
 	blr				/* return r3 = status */
@@ -33,6 +97,8 @@ _GLOBAL(plpar_hcall)
 	mfcr	r0
 	stw	r0,8(r1)
 
+	HCALL_INST_PRECALL
+
 	std     r4,STK_PARM(r4)(r1)     /* Save ret buffer */
 
 	mr	r4,r5
@@ -50,6 +116,8 @@ _GLOBAL(plpar_hcall)
 	std	r6, 16(r12)
 	std	r7, 24(r12)
 
+	HCALL_INST_POSTCALL
+
 	lwz	r0,8(r1)
 	mtcrf	0xff,r0
 
@@ -61,6 +129,8 @@ _GLOBAL(plpar_hcall9)
 	mfcr	r0
 	stw	r0,8(r1)
 
+	HCALL_INST_PRECALL
+
 	std     r4,STK_PARM(r4)(r1)     /* Save ret buffer */
 
 	mr	r4,r5
@@ -86,6 +156,8 @@ _GLOBAL(plpar_hcall9)
 	std	r11,56(r12)
 	std	r12,64(r12)
 
+	HCALL_INST_POSTCALL
+
 	lwz	r0,8(r1)
 	mtcrf	0xff,r0
 
diff --git a/arch/powerpc/platforms/pseries/hvCall_inst.c b/arch/powerpc/platforms/pseries/hvCall_inst.c
new file mode 100644
index 000000000000..641e6511cf06
--- /dev/null
+++ b/arch/powerpc/platforms/pseries/hvCall_inst.c
@@ -0,0 +1,129 @@
+/*
+ * Copyright (C) 2006 Mike Kravetz IBM Corporation
+ *
+ * Hypervisor Call Instrumentation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/percpu.h>
+#include <linux/debugfs.h>
+#include <linux/seq_file.h>
+#include <linux/cpumask.h>
+#include <asm/hvcall.h>
+#include <asm/firmware.h>
+#include <asm/cputable.h>
+
+DEFINE_PER_CPU(struct hcall_stats[HCALL_STAT_ARRAY_SIZE], hcall_stats);
+
+/*
+ * Routines for displaying the statistics in debugfs
+ */
+static void *hc_start(struct seq_file *m, loff_t *pos)
+{
+	if ((int)*pos < HCALL_STAT_ARRAY_SIZE)
+		return (void *)(unsigned long)(*pos + 1);
+
+	return NULL;
+}
+
+static void *hc_next(struct seq_file *m, void *p, loff_t * pos)
+{
+	++*pos;
+
+	return hc_start(m, pos);
+}
+
+static void hc_stop(struct seq_file *m, void *p)
+{
+}
+
+static int hc_show(struct seq_file *m, void *p)
+{
+	unsigned long h_num = (unsigned long)p;
+	struct hcall_stats *hs = (struct hcall_stats *)m->private;
+
+	if (hs[h_num].num_calls) {
+		if (!cpu_has_feature(CPU_FTR_PURR))
+			seq_printf(m, "%lu %lu %lu %lu\n", h_num<<2,
+				   hs[h_num].num_calls,
+				   hs[h_num].tb_total,
+				   hs[h_num].purr_total);
+		else
+			seq_printf(m, "%lu %lu %lu\n", h_num<<2,
+				   hs[h_num].num_calls,
+				   hs[h_num].tb_total);
+	}
+
+	return 0;
+}
+
+static struct seq_operations hcall_inst_seq_ops = {
+        .start = hc_start,
+        .next  = hc_next,
+        .stop  = hc_stop,
+        .show  = hc_show
+};
+
+static int hcall_inst_seq_open(struct inode *inode, struct file *file)
+{
+	int rc;
+	struct seq_file *seq;
+
+	rc = seq_open(file, &hcall_inst_seq_ops);
+	seq = file->private_data;
+	seq->private = file->f_dentry->d_inode->u.generic_ip;
+
+	return rc;
+}
+
+static struct file_operations hcall_inst_seq_fops = {
+	.open = hcall_inst_seq_open,
+	.read = seq_read,
+	.llseek = seq_lseek,
+	.release = seq_release,
+};
+
+#define	HCALL_ROOT_DIR		"hcall_inst"
+#define CPU_NAME_BUF_SIZE	32
+
+static int __init hcall_inst_init(void)
+{
+	struct dentry *hcall_root;
+	struct dentry *hcall_file;
+	char cpu_name_buf[CPU_NAME_BUF_SIZE];
+	int cpu;
+
+	if (!firmware_has_feature(FW_FEATURE_LPAR))
+		return 0;
+
+	hcall_root = debugfs_create_dir(HCALL_ROOT_DIR, NULL);
+	if (!hcall_root)
+		return -ENOMEM;
+
+	for_each_possible_cpu(cpu) {
+		snprintf(cpu_name_buf, CPU_NAME_BUF_SIZE, "cpu%d", cpu);
+		hcall_file = debugfs_create_file(cpu_name_buf, S_IRUGO,
+						 hcall_root,
+						 per_cpu(hcall_stats, cpu),
+						 &hcall_inst_seq_fops);
+		if (!hcall_file)
+			return -ENOMEM;
+	}
+
+	return 0;
+}
+__initcall(hcall_inst_init);
diff --git a/include/asm-powerpc/hvcall.h b/include/asm-powerpc/hvcall.h
index 63ce1ac8c1f4..257d1cecb8c9 100644
--- a/include/asm-powerpc/hvcall.h
+++ b/include/asm-powerpc/hvcall.h
@@ -208,7 +208,7 @@
 #define H_JOIN			0x298
 #define H_VASI_STATE            0x2A4
 #define H_ENABLE_CRQ		0x2B0
-#define MAX_HCALL_OPCODES	(H_ENABLE_CRQ >> 2)
+#define MAX_HCALL_OPCODE	H_ENABLE_CRQ
 
 #ifndef __ASSEMBLY__
 
@@ -246,6 +246,16 @@ long plpar_hcall(unsigned long opcode, unsigned long *retbuf, ...);
 #define PLPAR_HCALL9_BUFSIZE 9
 long plpar_hcall9(unsigned long opcode, unsigned long *retbuf, ...);
 
+/* For hcall instrumentation.  One structure per-hcall, per-CPU */
+struct hcall_stats {
+	unsigned long	num_calls;	/* number of calls (on this CPU) */
+	unsigned long	tb_total;	/* total wall time (mftb) of calls. */
+	unsigned long	purr_total;	/* total cpu time (PURR) of calls. */
+};
+void update_hcall_stats(unsigned long opcode, unsigned long tb_delta,
+			unsigned long purr_delta);
+#define HCALL_STAT_ARRAY_SIZE	((MAX_HCALL_OPCODE >> 2) + 1)
+
 #endif /* __ASSEMBLY__ */
 #endif /* __KERNEL__ */
 #endif /* _ASM_POWERPC_HVCALL_H */
-- 
cgit v1.2.3


From 76fd85937097a0c2ec8ab23bf21dc10992d1c398 Mon Sep 17 00:00:00 2001
From: Stephen Hemminger <shemminger@osdl.org>
Date: Fri, 8 Sep 2006 11:16:13 -0700
Subject: [PATCH] ethtool: allow const ethtool_ops

The ethtool_ops structure is immutable, it expected to be setup
by the driver and is never changed. This patch allows drivers to
declare there ethtool_ops structure read-only.

Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 include/linux/netdevice.h |  2 +-
 net/core/ethtool.c        | 14 +++++++-------
 2 files changed, 8 insertions(+), 8 deletions(-)

(limited to 'include')

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 50a4719512ed..a2e747353367 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -342,7 +342,7 @@ struct net_device
 	/* Instance data managed by the core of Wireless Extensions. */
 	struct iw_public_data *	wireless_data;
 
-	struct ethtool_ops *ethtool_ops;
+	const struct ethtool_ops *ethtool_ops;
 
 	/*
 	 * This marks the end of the "visible" part of the structure. All
diff --git a/net/core/ethtool.c b/net/core/ethtool.c
index 2797e2815418..e0ca04f38cef 100644
--- a/net/core/ethtool.c
+++ b/net/core/ethtool.c
@@ -143,7 +143,7 @@ static int ethtool_set_settings(struct net_device *dev, void __user *useraddr)
 static int ethtool_get_drvinfo(struct net_device *dev, void __user *useraddr)
 {
 	struct ethtool_drvinfo info;
-	struct ethtool_ops *ops = dev->ethtool_ops;
+	const struct ethtool_ops *ops = dev->ethtool_ops;
 
 	if (!ops->get_drvinfo)
 		return -EOPNOTSUPP;
@@ -169,7 +169,7 @@ static int ethtool_get_drvinfo(struct net_device *dev, void __user *useraddr)
 static int ethtool_get_regs(struct net_device *dev, char __user *useraddr)
 {
 	struct ethtool_regs regs;
-	struct ethtool_ops *ops = dev->ethtool_ops;
+	const struct ethtool_ops *ops = dev->ethtool_ops;
 	void *regbuf;
 	int reglen, ret;
 
@@ -282,7 +282,7 @@ static int ethtool_get_link(struct net_device *dev, void __user *useraddr)
 static int ethtool_get_eeprom(struct net_device *dev, void __user *useraddr)
 {
 	struct ethtool_eeprom eeprom;
-	struct ethtool_ops *ops = dev->ethtool_ops;
+	const struct ethtool_ops *ops = dev->ethtool_ops;
 	u8 *data;
 	int ret;
 
@@ -327,7 +327,7 @@ static int ethtool_get_eeprom(struct net_device *dev, void __user *useraddr)
 static int ethtool_set_eeprom(struct net_device *dev, void __user *useraddr)
 {
 	struct ethtool_eeprom eeprom;
-	struct ethtool_ops *ops = dev->ethtool_ops;
+	const struct ethtool_ops *ops = dev->ethtool_ops;
 	u8 *data;
 	int ret;
 
@@ -640,7 +640,7 @@ static int ethtool_set_gso(struct net_device *dev, char __user *useraddr)
 static int ethtool_self_test(struct net_device *dev, char __user *useraddr)
 {
 	struct ethtool_test test;
-	struct ethtool_ops *ops = dev->ethtool_ops;
+	const struct ethtool_ops *ops = dev->ethtool_ops;
 	u64 *data;
 	int ret;
 
@@ -673,7 +673,7 @@ static int ethtool_self_test(struct net_device *dev, char __user *useraddr)
 static int ethtool_get_strings(struct net_device *dev, void __user *useraddr)
 {
 	struct ethtool_gstrings gstrings;
-	struct ethtool_ops *ops = dev->ethtool_ops;
+	const struct ethtool_ops *ops = dev->ethtool_ops;
 	u8 *data;
 	int ret;
 
@@ -733,7 +733,7 @@ static int ethtool_phys_id(struct net_device *dev, void __user *useraddr)
 static int ethtool_get_stats(struct net_device *dev, void __user *useraddr)
 {
 	struct ethtool_stats stats;
-	struct ethtool_ops *ops = dev->ethtool_ops;
+	const struct ethtool_ops *ops = dev->ethtool_ops;
 	u64 *data;
 	int ret;
 
-- 
cgit v1.2.3


From fea63e38013ec628ab3f7fddc4c2148064b7910a Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Sat, 16 Sep 2006 03:04:15 +0900
Subject: [PATCH] libata: fix non-uniform ports handling

Non-uniform ports handling got broken while updating libata to handle
those in the same host.  Only separate irq for the non-uniform
secondary port was implemented while all other fields (host flags,
transfer mode...) of the secondary port simply shared those of the
first.

For ata_piix combined mode, which ATM is the only user of non-uniform
ports, this causes the secondary port assume the wrong type.  This can
cause PATA port to use SATA ops, which results in bogus check on PCS
and detection failure.

This patch adds ata_probe_ent->pinfo2 which points to optional
port_info for the secondary port.  For the time being, this seems to
be the simplest solution.  This workaround will be removed together
with ata_probe_ent itself after init model is updated to allow more
flexibility.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Nelson A. de Oliveira <naoliv@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/ata/libata-core.c | 18 +++++++++++++-----
 drivers/ata/libata-sff.c  |  2 ++
 include/linux/libata.h    |  8 ++++++++
 3 files changed, 23 insertions(+), 5 deletions(-)

(limited to 'include')

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 1c9315401f7a..bb66a12c84e5 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -5269,11 +5269,19 @@ void ata_port_init(struct ata_port *ap, struct ata_host *host,
 	ap->host = host;
 	ap->dev = ent->dev;
 	ap->port_no = port_no;
-	ap->pio_mask = ent->pio_mask;
-	ap->mwdma_mask = ent->mwdma_mask;
-	ap->udma_mask = ent->udma_mask;
-	ap->flags |= ent->port_flags;
-	ap->ops = ent->port_ops;
+	if (port_no == 1 && ent->pinfo2) {
+		ap->pio_mask = ent->pinfo2->pio_mask;
+		ap->mwdma_mask = ent->pinfo2->mwdma_mask;
+		ap->udma_mask = ent->pinfo2->udma_mask;
+		ap->flags |= ent->pinfo2->flags;
+		ap->ops = ent->pinfo2->port_ops;
+	} else {
+		ap->pio_mask = ent->pio_mask;
+		ap->mwdma_mask = ent->mwdma_mask;
+		ap->udma_mask = ent->udma_mask;
+		ap->flags |= ent->port_flags;
+		ap->ops = ent->port_ops;
+	}
 	ap->hw_sata_spd_limit = UINT_MAX;
 	ap->active_tag = ATA_TAG_POISON;
 	ap->last_ctl = 0xFF;
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index d51dc41fa195..688bb55e197a 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -858,6 +858,7 @@ ata_pci_init_native_mode(struct pci_dev *pdev, struct ata_port_info **port, int
 			probe_ent->port[p].bmdma_addr = bmdma;
 		}
 		ata_std_ports(&probe_ent->port[p]);
+		probe_ent->pinfo2 = port[1];
 		p++;
 	}
 
@@ -907,6 +908,7 @@ static struct ata_probe_ent *ata_pci_init_legacy_port(struct pci_dev *pdev,
 				probe_ent->_host_flags |= ATA_HOST_SIMPLEX;
 		}
 		ata_std_ports(&probe_ent->port[1]);
+		probe_ent->pinfo2 = port[1];
 	} else
 		probe_ent->dummy_port_mask |= ATA_PORT_SECONDARY;
 
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 8715305f611f..ff67e7524fe9 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -361,6 +361,14 @@ struct ata_probe_ent {
 	unsigned long		_host_flags;
 	void __iomem		*mmio_base;
 	void			*private_data;
+
+	/* port_info for the secondary port.  Together with irq2, it's
+	 * used to implement non-uniform secondary port.  Currently,
+	 * the only user is ata_piix combined mode.  This workaround
+	 * will be removed together with ata_probe_ent when init model
+	 * is updated.
+	 */
+	const struct ata_port_info *pinfo2;
 };
 
 struct ata_host {
-- 
cgit v1.2.3


From 93590859884784520a1850767f86296abc2cdc6d Mon Sep 17 00:00:00 2001
From: Alan Cox <alan@lxorguk.ukuu.org.uk>
Date: Tue, 12 Sep 2006 16:55:12 +0100
Subject: [PATCH] libata: improve handling of diagostic fail (and hardware that
 misreports it)

Our ATA probe code checks that a device is not reporting a diagnostic
failure during start up. Unfortunately at least one device seems to like
doing this - the Gigabyte iRAM.

This is only done for the master right now (which is fine for the iRAM
as it is SATA), as with PATA some combinations of ATAPI device seem to
fool the check into seeing a drive that isn't there if it is applied to
the slave.

Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/ata/libata-core.c | 19 +++++++++++++++++--
 include/linux/libata.h    |  6 ++++++
 2 files changed, 23 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index e85c2f8cf193..753b0152afd1 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -616,8 +616,11 @@ ata_dev_try_classify(struct ata_port *ap, unsigned int device, u8 *r_err)
 	if (r_err)
 		*r_err = err;
 
-	/* see if device passed diags */
-	if (err == 1)
+	/* see if device passed diags: if master then continue and warn later */
+	if (err == 0 && device == 0)
+		/* diagnostic fail : do nothing _YET_ */
+		ap->device[device].horkage |= ATA_HORKAGE_DIAGNOSTIC;
+	else if (err == 1)
 		/* do nothing */ ;
 	else if ((device == 0) && (err == 0x81))
 		/* do nothing */ ;
@@ -1523,6 +1526,18 @@ int ata_dev_configure(struct ata_device *dev, int print_info)
 				       cdb_intr_string);
 	}
 
+	if (dev->horkage & ATA_HORKAGE_DIAGNOSTIC) {
+		/* Let the user know. We don't want to disallow opens for
+		   rescue purposes, or in case the vendor is just a blithering
+		   idiot */
+                if (print_info) {
+			ata_dev_printk(dev, KERN_WARNING,
+"Drive reports diagnostics failure. This may indicate a drive\n");
+			ata_dev_printk(dev, KERN_WARNING,
+"fault or invalid emulation. Contact drive vendor for information.\n");
+		}
+	}
+
 	ata_set_port_max_cmd_len(ap);
 
 	/* limit bridge transfers to udma5, 200 sectors */
diff --git a/include/linux/libata.h b/include/linux/libata.h
index ff67e7524fe9..1ef3d3901b47 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -289,6 +289,11 @@ enum {
 	 * most devices.
 	 */
 	ATA_SPINUP_WAIT		= 8000,
+	
+	/* Horkage types. May be set by libata or controller on drives
+	   (some horkage may be drive/controller pair dependant */
+
+	ATA_HORKAGE_DIAGNOSTIC	= (1 << 0),	/* Failed boot diag */
 };
 
 enum hsm_task_states {
@@ -476,6 +481,7 @@ struct ata_device {
 
 	/* error history */
 	struct ata_ering	ering;
+	unsigned int		horkage;	/* List of broken features */
 };
 
 /* Offset into struct ata_device.  Fields above it are maintained
-- 
cgit v1.2.3


From fa053d2f008cb73fa768b8e171486d8c0b33312b Mon Sep 17 00:00:00 2001
From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Tue, 19 Sep 2006 14:51:40 +1000
Subject: [POWERPC] remove unused io accessors

The io accessors insw_ns, outsw_ns, insl_ns and outsl_ns are unused
(except for one unnecessary use in drivers/net/3c509.c that is addressed
in a previous patch) and are only defined in powerpc/ppc, so remove them.

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 arch/ppc/syslib/m8260_pci_erratum9.c | 16 ----------------
 include/asm-powerpc/io.h             | 13 +------------
 include/asm-ppc/io.h                 | 10 ----------
 include/asm-ppc/mpc8260_pci9.h       |  4 ----
 4 files changed, 1 insertion(+), 42 deletions(-)

(limited to 'include')

diff --git a/arch/ppc/syslib/m8260_pci_erratum9.c b/arch/ppc/syslib/m8260_pci_erratum9.c
index 974581ea4849..5475709ce07b 100644
--- a/arch/ppc/syslib/m8260_pci_erratum9.c
+++ b/arch/ppc/syslib/m8260_pci_erratum9.c
@@ -339,20 +339,6 @@ void insl(unsigned port, void *buf, int nl)
 	idma_pci9_read((u8 *)buf, (u8 *)addr, nl*sizeof(u32), sizeof(u32), 0);
 }
 
-void insw_ns(unsigned port, void *buf, int ns)
-{
-	u8 *addr = (u8 *)(port + _IO_BASE);
-
-	idma_pci9_read((u8 *)buf, (u8 *)addr, ns*sizeof(u16), sizeof(u16), 0);
-}
-
-void insl_ns(unsigned port, void *buf, int nl)
-{
-	u8 *addr = (u8 *)(port + _IO_BASE);
-
-	idma_pci9_read((u8 *)buf, (u8 *)addr, nl*sizeof(u32), sizeof(u32), 0);
-}
-
 void *memcpy_fromio(void *dest, unsigned long src, size_t count)
 {
 	unsigned long pa = iopa((unsigned long) src);
@@ -373,8 +359,6 @@ EXPORT_SYMBOL(inl);
 EXPORT_SYMBOL(insb);
 EXPORT_SYMBOL(insw);
 EXPORT_SYMBOL(insl);
-EXPORT_SYMBOL(insw_ns);
-EXPORT_SYMBOL(insl_ns);
 EXPORT_SYMBOL(memcpy_fromio);
 
 #endif	/* ifdef CONFIG_8260_PCI9 */
diff --git a/include/asm-powerpc/io.h b/include/asm-powerpc/io.h
index 212428db0d8b..9aaced542624 100644
--- a/include/asm-powerpc/io.h
+++ b/include/asm-powerpc/io.h
@@ -76,8 +76,7 @@ extern unsigned long pci_io_base;
 #define insb(port, buf, ns)	_insb((u8 __iomem *)((port)+pci_io_base), (buf), (ns))
 #define insw(port, buf, ns)	_insw_ns((u8 __iomem *)((port)+pci_io_base), (buf), (ns))
 #define insl(port, buf, nl)	_insl_ns((u8 __iomem *)((port)+pci_io_base), (buf), (nl))
-#define insw_ns(port, buf, ns)	_insw_ns((u16 __iomem *)((port)+pci_io_base), (buf), (ns))
-#define insl_ns(port, buf, nl)	_insl_ns((u32 __iomem *)((port)+pci_io_base), (buf), (nl))
+
 #else
 
 static inline unsigned char __raw_readb(const volatile void __iomem *addr)
@@ -138,8 +137,6 @@ static inline void __raw_writeq(unsigned long v, volatile void __iomem *addr)
 #define insb(port, buf, ns)	eeh_insb((port), (buf), (ns))
 #define insw(port, buf, ns)	eeh_insw_ns((port), (buf), (ns))
 #define insl(port, buf, nl)	eeh_insl_ns((port), (buf), (nl))
-#define insw_ns(port, buf, ns)	eeh_insw_ns((port), (buf), (ns))
-#define insl_ns(port, buf, nl)	eeh_insl_ns((port), (buf), (nl))
 
 #endif
 
@@ -180,14 +177,6 @@ static inline void mmiowb(void)
 #define inl_p(port)             inl(port)
 #define outl_p(val, port)       (udelay(1), outl((val), (port)))
 
-/*
- * The *_ns versions below don't do byte-swapping.
- * Neither do the standard versions now, these are just here
- * for older code.
- */
-#define outsw_ns(port, buf, ns)	_outsw_ns((u16 __iomem *)((port)+pci_io_base), (buf), (ns))
-#define outsl_ns(port, buf, nl)	_outsl_ns((u32 __iomem *)((port)+pci_io_base), (buf), (nl))
-
 
 #define IO_SPACE_LIMIT ~(0UL)
 
diff --git a/include/asm-ppc/io.h b/include/asm-ppc/io.h
index 680555be22ec..fb0a8fcc51cf 100644
--- a/include/asm-ppc/io.h
+++ b/include/asm-ppc/io.h
@@ -338,16 +338,6 @@ extern void _outsw_ns(volatile u16 __iomem *port, const void *buf, int ns);
 extern void _insl_ns(volatile u32 __iomem *port, void *buf, int nl);
 extern void _outsl_ns(volatile u32 __iomem *port, const void *buf, int nl);
 
-/*
- * The *_ns versions below don't do byte-swapping.
- * Neither do the standard versions now, these are just here
- * for older code.
- */
-#define insw_ns(port, buf, ns)	_insw_ns((port)+___IO_BASE, (buf), (ns))
-#define outsw_ns(port, buf, ns)	_outsw_ns((port)+___IO_BASE, (buf), (ns))
-#define insl_ns(port, buf, nl)	_insl_ns((port)+___IO_BASE, (buf), (nl))
-#define outsl_ns(port, buf, nl)	_outsl_ns((port)+___IO_BASE, (buf), (nl))
-
 
 #define IO_SPACE_LIMIT ~0
 
diff --git a/include/asm-ppc/mpc8260_pci9.h b/include/asm-ppc/mpc8260_pci9.h
index 26b3f6e787bc..9f7176881c56 100644
--- a/include/asm-ppc/mpc8260_pci9.h
+++ b/include/asm-ppc/mpc8260_pci9.h
@@ -30,8 +30,6 @@
 #undef inb
 #undef inw
 #undef inl
-#undef insw_ns
-#undef insl_ns
 #undef memcpy_fromio
 
 extern int readb(volatile unsigned char *addr);
@@ -43,8 +41,6 @@ extern void insl(unsigned port, void *buf, int nl);
 extern int inb(unsigned port);
 extern int inw(unsigned port);
 extern unsigned inl(unsigned port);
-extern void insw_ns(unsigned port, void *buf, int ns);
-extern void insl_ns(unsigned port, void *buf, int nl);
 extern void *memcpy_fromio(void *dest, unsigned long src, size_t count);
 
 #endif /* !__CONFIG_8260_PCI9_DEFS */
-- 
cgit v1.2.3


From 661f1cdb8b3e3c2c44e97df122c1d5643c054ce8 Mon Sep 17 00:00:00 2001
From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Tue, 19 Sep 2006 16:52:55 +1000
Subject: [POWERPC] remove unused asm routines

_insw, _outsw, _insl amd _outsl are all unused, so remove them.

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 arch/powerpc/kernel/misc.S      | 52 -------------------------
 arch/powerpc/kernel/ppc_ksyms.c |  4 --
 arch/ppc/kernel/misc.S          | 84 -----------------------------------------
 arch/ppc/kernel/ppc_ksyms.c     |  4 --
 include/asm-powerpc/io.h        |  4 --
 include/asm-ppc/io.h            |  4 --
 6 files changed, 152 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/kernel/misc.S b/arch/powerpc/kernel/misc.S
index f770805f1215..dd5f8e429196 100644
--- a/arch/powerpc/kernel/misc.S
+++ b/arch/powerpc/kernel/misc.S
@@ -86,58 +86,6 @@ _GLOBAL(_outsb)
 	sync
 	blr
 
-_GLOBAL(_insw)
-	sync
-	cmpwi	0,r5,0
-	mtctr	r5
-	subi	r4,r4,2
-	blelr-
-00:	lhbrx	r5,0,r3
-	eieio
-	sthu	r5,2(r4)
-	bdnz	00b
-	twi	0,r5,0
-	isync
-	blr
-
-_GLOBAL(_outsw)
-	cmpwi	0,r5,0
-	mtctr	r5
-	subi	r4,r4,2
-	blelr-
-	sync
-00:	lhzu	r5,2(r4)
-	sthbrx	r5,0,r3
-	bdnz	00b
-	sync
-	blr
-
-_GLOBAL(_insl)
-	sync
-	cmpwi	0,r5,0
-	mtctr	r5
-	subi	r4,r4,4
-	blelr-
-00:	lwbrx	r5,0,r3
-	eieio
-	stwu	r5,4(r4)
-	bdnz	00b
-	twi	0,r5,0
-	isync
-	blr
-
-_GLOBAL(_outsl)
-	cmpwi	0,r5,0
-	mtctr	r5
-	subi	r4,r4,4
-	blelr-
-	sync
-00:	lwzu	r5,4(r4)
-	stwbrx	r5,0,r3
-	bdnz	00b
-	sync
-	blr
-
 #ifdef CONFIG_PPC32
 _GLOBAL(__ide_mm_insw)
 #endif
diff --git a/arch/powerpc/kernel/ppc_ksyms.c b/arch/powerpc/kernel/ppc_ksyms.c
index b2edac8ddf0a..314d6114e6ec 100644
--- a/arch/powerpc/kernel/ppc_ksyms.c
+++ b/arch/powerpc/kernel/ppc_ksyms.c
@@ -104,10 +104,6 @@ EXPORT_SYMBOL(__ide_mm_outsl);
 
 EXPORT_SYMBOL(_insb);
 EXPORT_SYMBOL(_outsb);
-EXPORT_SYMBOL(_insw);
-EXPORT_SYMBOL(_outsw);
-EXPORT_SYMBOL(_insl);
-EXPORT_SYMBOL(_outsl);
 EXPORT_SYMBOL(_insw_ns);
 EXPORT_SYMBOL(_outsw_ns);
 EXPORT_SYMBOL(_insl_ns);
diff --git a/arch/ppc/kernel/misc.S b/arch/ppc/kernel/misc.S
index 2fa0075f2b5f..44700bbfe7bc 100644
--- a/arch/ppc/kernel/misc.S
+++ b/arch/ppc/kernel/misc.S
@@ -768,90 +768,6 @@ _GLOBAL(_outsb)
 	bdnz	00b
 	blr
 
-_GLOBAL(_insw)
-	cmpwi	0,r5,0
-	mtctr	r5
-	subi	r4,r4,2
-	blelr-
-00:	lhbrx	r5,0,r3
-01:	eieio
-02:	sthu	r5,2(r4)
-	ISYNC_8xx
-	.section .fixup,"ax"
-03:	blr
-	.text
-	.section __ex_table, "a"
-		.align 2
-		.long 00b, 03b
-		.long 01b, 03b
-		.long 02b, 03b
-	.text
-	bdnz	00b
-	blr
-
-_GLOBAL(_outsw)
-	cmpwi	0,r5,0
-	mtctr	r5
-	subi	r4,r4,2
-	blelr-
-00:	lhzu	r5,2(r4)
-01:	eieio
-02:	sthbrx	r5,0,r3
-	ISYNC_8xx
-	.section .fixup,"ax"
-03:	blr
-	.text
-	.section __ex_table, "a"
-		.align 2
-		.long 00b, 03b
-		.long 01b, 03b
-		.long 02b, 03b
-	.text
-	bdnz	00b
-	blr
-
-_GLOBAL(_insl)
-	cmpwi	0,r5,0
-	mtctr	r5
-	subi	r4,r4,4
-	blelr-
-00:	lwbrx	r5,0,r3
-01:	eieio
-02:	stwu	r5,4(r4)
-	ISYNC_8xx
-	.section .fixup,"ax"
-03:	blr
-	.text
-	.section __ex_table, "a"
-		.align 2
-		.long 00b, 03b
-		.long 01b, 03b
-		.long 02b, 03b
-	.text
-	bdnz	00b
-	blr
-
-_GLOBAL(_outsl)
-	cmpwi	0,r5,0
-	mtctr	r5
-	subi	r4,r4,4
-	blelr-
-00:	lwzu	r5,4(r4)
-01:	stwbrx	r5,0,r3
-02:	eieio
-	ISYNC_8xx
-	.section .fixup,"ax"
-03:	blr
-	.text
-	.section __ex_table, "a"
-		.align 2
-		.long 00b, 03b
-		.long 01b, 03b
-		.long 02b, 03b
-	.text
-	bdnz	00b
-	blr
-
 _GLOBAL(__ide_mm_insw)
 _GLOBAL(_insw_ns)
 	cmpwi	0,r5,0
diff --git a/arch/ppc/kernel/ppc_ksyms.c b/arch/ppc/kernel/ppc_ksyms.c
index d1735401384c..2bd1f7353f56 100644
--- a/arch/ppc/kernel/ppc_ksyms.c
+++ b/arch/ppc/kernel/ppc_ksyms.c
@@ -122,10 +122,6 @@ EXPORT_SYMBOL(__ide_mm_outsl);
 
 EXPORT_SYMBOL(_insb);
 EXPORT_SYMBOL(_outsb);
-EXPORT_SYMBOL(_insw);
-EXPORT_SYMBOL(_outsw);
-EXPORT_SYMBOL(_insl);
-EXPORT_SYMBOL(_outsl);
 EXPORT_SYMBOL(_insw_ns);
 EXPORT_SYMBOL(_outsw_ns);
 EXPORT_SYMBOL(_insl_ns);
diff --git a/include/asm-powerpc/io.h b/include/asm-powerpc/io.h
index 9aaced542624..0ee48436b1e3 100644
--- a/include/asm-powerpc/io.h
+++ b/include/asm-powerpc/io.h
@@ -151,10 +151,6 @@ static inline void __raw_writeq(unsigned long v, volatile void __iomem *addr)
 
 extern void _insb(volatile u8 __iomem *port, void *buf, int ns);
 extern void _outsb(volatile u8 __iomem *port, const void *buf, int ns);
-extern void _insw(volatile u16 __iomem *port, void *buf, int ns);
-extern void _outsw(volatile u16 __iomem *port, const void *buf, int ns);
-extern void _insl(volatile u32 __iomem *port, void *buf, int nl);
-extern void _outsl(volatile u32 __iomem *port, const void *buf, int nl);
 extern void _insw_ns(volatile u16 __iomem *port, void *buf, int ns);
 extern void _outsw_ns(volatile u16 __iomem *port, const void *buf, int ns);
 extern void _insl_ns(volatile u32 __iomem *port, void *buf, int nl);
diff --git a/include/asm-ppc/io.h b/include/asm-ppc/io.h
index fb0a8fcc51cf..9fac420f1648 100644
--- a/include/asm-ppc/io.h
+++ b/include/asm-ppc/io.h
@@ -329,10 +329,6 @@ __do_out_asm(outl, "stwbrx")
 
 extern void _insb(volatile u8 __iomem *port, void *buf, int ns);
 extern void _outsb(volatile u8 __iomem *port, const void *buf, int ns);
-extern void _insw(volatile u16 __iomem *port, void *buf, int ns);
-extern void _outsw(volatile u16 __iomem *port, const void *buf, int ns);
-extern void _insl(volatile u32 __iomem *port, void *buf, int nl);
-extern void _outsl(volatile u32 __iomem *port, const void *buf, int nl);
 extern void _insw_ns(volatile u16 __iomem *port, void *buf, int ns);
 extern void _outsw_ns(volatile u16 __iomem *port, const void *buf, int ns);
 extern void _insl_ns(volatile u32 __iomem *port, void *buf, int nl);
-- 
cgit v1.2.3


From 73ea9e1bcb8eea4f3b2052fe7ccd7ee4b5a271a0 Mon Sep 17 00:00:00 2001
From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Tue, 19 Sep 2006 17:30:20 +1000
Subject: [POWERPC] clean up ide io accessors

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 arch/powerpc/kernel/misc.S      | 12 ------------
 arch/powerpc/kernel/ppc_ksyms.c |  7 -------
 arch/ppc/kernel/misc.S          |  4 ----
 arch/ppc/kernel/ppc_ksyms.c     |  5 -----
 include/asm-powerpc/ide.h       | 12 ++++++------
 include/asm-powerpc/io.h        |  6 ------
 6 files changed, 6 insertions(+), 40 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/kernel/misc.S b/arch/powerpc/kernel/misc.S
index dd5f8e429196..6feb391422ec 100644
--- a/arch/powerpc/kernel/misc.S
+++ b/arch/powerpc/kernel/misc.S
@@ -86,9 +86,6 @@ _GLOBAL(_outsb)
 	sync
 	blr
 
-#ifdef CONFIG_PPC32
-_GLOBAL(__ide_mm_insw)
-#endif
 _GLOBAL(_insw_ns)
 	sync
 	cmpwi	0,r5,0
@@ -103,9 +100,6 @@ _GLOBAL(_insw_ns)
 	isync
 	blr
 
-#ifdef CONFIG_PPC32
-_GLOBAL(__ide_mm_outsw)
-#endif
 _GLOBAL(_outsw_ns)
 	cmpwi	0,r5,0
 	mtctr	r5
@@ -118,9 +112,6 @@ _GLOBAL(_outsw_ns)
 	sync
 	blr
 
-#ifdef CONFIG_PPC32
-_GLOBAL(__ide_mm_insl)
-#endif
 _GLOBAL(_insl_ns)
 	sync
 	cmpwi	0,r5,0
@@ -135,9 +126,6 @@ _GLOBAL(_insl_ns)
 	isync
 	blr
 
-#ifdef CONFIG_PPC32
-_GLOBAL(__ide_mm_outsl)
-#endif
 _GLOBAL(_outsl_ns)
 	cmpwi	0,r5,0
 	mtctr	r5
diff --git a/arch/powerpc/kernel/ppc_ksyms.c b/arch/powerpc/kernel/ppc_ksyms.c
index 314d6114e6ec..75429e580518 100644
--- a/arch/powerpc/kernel/ppc_ksyms.c
+++ b/arch/powerpc/kernel/ppc_ksyms.c
@@ -95,13 +95,6 @@ EXPORT_SYMBOL(__strnlen_user);
 EXPORT_SYMBOL(copy_4K_page);
 #endif
 
-#ifndef  __powerpc64__
-EXPORT_SYMBOL(__ide_mm_insl);
-EXPORT_SYMBOL(__ide_mm_outsw);
-EXPORT_SYMBOL(__ide_mm_insw);
-EXPORT_SYMBOL(__ide_mm_outsl);
-#endif
-
 EXPORT_SYMBOL(_insb);
 EXPORT_SYMBOL(_outsb);
 EXPORT_SYMBOL(_insw_ns);
diff --git a/arch/ppc/kernel/misc.S b/arch/ppc/kernel/misc.S
index 44700bbfe7bc..50b4bbd06804 100644
--- a/arch/ppc/kernel/misc.S
+++ b/arch/ppc/kernel/misc.S
@@ -768,7 +768,6 @@ _GLOBAL(_outsb)
 	bdnz	00b
 	blr
 
-_GLOBAL(__ide_mm_insw)
 _GLOBAL(_insw_ns)
 	cmpwi	0,r5,0
 	mtctr	r5
@@ -790,7 +789,6 @@ _GLOBAL(_insw_ns)
 	bdnz	00b
 	blr
 
-_GLOBAL(__ide_mm_outsw)
 _GLOBAL(_outsw_ns)
 	cmpwi	0,r5,0
 	mtctr	r5
@@ -812,7 +810,6 @@ _GLOBAL(_outsw_ns)
 	bdnz	00b
 	blr
 
-_GLOBAL(__ide_mm_insl)
 _GLOBAL(_insl_ns)
 	cmpwi	0,r5,0
 	mtctr	r5
@@ -834,7 +831,6 @@ _GLOBAL(_insl_ns)
 	bdnz	00b
 	blr
 
-_GLOBAL(__ide_mm_outsl)
 _GLOBAL(_outsl_ns)
 	cmpwi	0,r5,0
 	mtctr	r5
diff --git a/arch/ppc/kernel/ppc_ksyms.c b/arch/ppc/kernel/ppc_ksyms.c
index 2bd1f7353f56..c8b65ca8a350 100644
--- a/arch/ppc/kernel/ppc_ksyms.c
+++ b/arch/ppc/kernel/ppc_ksyms.c
@@ -115,11 +115,6 @@ EXPORT_SYMBOL(outw);
 EXPORT_SYMBOL(outl);
 EXPORT_SYMBOL(outsl);*/
 
-EXPORT_SYMBOL(__ide_mm_insl);
-EXPORT_SYMBOL(__ide_mm_outsw);
-EXPORT_SYMBOL(__ide_mm_insw);
-EXPORT_SYMBOL(__ide_mm_outsl);
-
 EXPORT_SYMBOL(_insb);
 EXPORT_SYMBOL(_outsb);
 EXPORT_SYMBOL(_insw_ns);
diff --git a/include/asm-powerpc/ide.h b/include/asm-powerpc/ide.h
index b09b42af6a1e..c8390f9485de 100644
--- a/include/asm-powerpc/ide.h
+++ b/include/asm-powerpc/ide.h
@@ -12,6 +12,7 @@
 #include <linux/sched.h>
 #include <asm/mpc8xx.h>
 #endif
+#include <asm/io.h>
 
 #ifndef MAX_HWIFS
 #ifdef __powerpc64__
@@ -21,15 +22,14 @@
 #endif
 #endif
 
+#define __ide_mm_insw(p, a, c)	_insw_ns((volatile u16 __iomem *)(p), (a), (c))
+#define __ide_mm_insl(p, a, c)	_insl_ns((volatile u32 __iomem *)(p), (a), (c))
+#define __ide_mm_outsw(p, a, c)	_outsw_ns((volatile u16 __iomem *)(p), (a), (c))
+#define __ide_mm_outsl(p, a, c)	_outsl_ns((volatile u32 __iomem *)(p), (a), (c))
+
 #ifndef  __powerpc64__
 #include <linux/hdreg.h>
 #include <linux/ioport.h>
-#include <asm/io.h>
-
-extern void __ide_mm_insw(void __iomem *port, void *addr, u32 count);
-extern void __ide_mm_outsw(void __iomem *port, void *addr, u32 count);
-extern void __ide_mm_insl(void __iomem *port, void *addr, u32 count);
-extern void __ide_mm_outsl(void __iomem *port, void *addr, u32 count);
 
 struct ide_machdep_calls {
         int         (*default_irq)(unsigned long base);
diff --git a/include/asm-powerpc/io.h b/include/asm-powerpc/io.h
index 0ee48436b1e3..51a598747367 100644
--- a/include/asm-powerpc/io.h
+++ b/include/asm-powerpc/io.h
@@ -28,12 +28,6 @@ extern int check_legacy_ioport(unsigned long base_port);
 
 #include <asm-generic/iomap.h>
 
-#define __ide_mm_insw(p, a, c) _insw_ns((volatile u16 __iomem *)(p), (a), (c))
-#define __ide_mm_insl(p, a, c) _insl_ns((volatile u32 __iomem *)(p), (a), (c))
-#define __ide_mm_outsw(p, a, c) _outsw_ns((volatile u16 __iomem *)(p), (a), (c))
-#define __ide_mm_outsl(p, a, c) _outsl_ns((volatile u32 __iomem *)(p), (a), (c))
-
-
 #define SIO_CONFIG_RA	0x398
 #define SIO_CONFIG_RD	0x399
 
-- 
cgit v1.2.3


From 5adcaf50cf697aa4d0c731107003c1383b59b214 Mon Sep 17 00:00:00 2001
From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Tue, 19 Sep 2006 22:17:49 +1000
Subject: [POWERPC] convert string i/o operations to C

This produces essentially the same code and will make the iSeries i/o
consolidation easier.

The count parameter is changed to long since that will produce the same
(better) code on 32 and 64 bit builds.

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 arch/powerpc/kernel/Makefile    |   2 +-
 arch/powerpc/kernel/io.c        | 117 ++++++++++++++++++++++++++++++++++++++++
 arch/powerpc/kernel/misc.S      |  95 --------------------------------
 arch/powerpc/kernel/ppc_ksyms.c |   7 ---
 include/asm-powerpc/io.h        |  12 ++---
 include/asm-ppc/io.h            |  12 ++---
 6 files changed, 130 insertions(+), 115 deletions(-)
 create mode 100644 arch/powerpc/kernel/io.c

(limited to 'include')

diff --git a/arch/powerpc/kernel/Makefile b/arch/powerpc/kernel/Makefile
index 8b3f4faf5768..8b133afbdc20 100644
--- a/arch/powerpc/kernel/Makefile
+++ b/arch/powerpc/kernel/Makefile
@@ -51,7 +51,7 @@ extra-$(CONFIG_8xx)		:= head_8xx.o
 extra-y				+= vmlinux.lds
 
 obj-y				+= time.o prom.o traps.o setup-common.o \
-				   udbg.o misc.o
+				   udbg.o misc.o io.o
 obj-$(CONFIG_PPC32)		+= entry_32.o setup_32.o misc_32.o
 obj-$(CONFIG_PPC64)		+= misc_64.o dma_64.o iommu.o
 obj-$(CONFIG_PPC_MULTIPLATFORM)	+= prom_init.o
diff --git a/arch/powerpc/kernel/io.c b/arch/powerpc/kernel/io.c
new file mode 100644
index 000000000000..80a3209acef4
--- /dev/null
+++ b/arch/powerpc/kernel/io.c
@@ -0,0 +1,117 @@
+/*
+ * I/O string operations
+ *    Copyright (C) 1995-1996 Gary Thomas (gdt@linuxppc.org)
+ *    Copyright (C) 2006 IBM Corporation
+ *
+ * Largely rewritten by Cort Dougan (cort@cs.nmt.edu)
+ * and Paul Mackerras.
+ *
+ * Adapted for iSeries by Mike Corrigan (mikejc@us.ibm.com)
+ * PPC64 updates by Dave Engebretsen (engebret@us.ibm.com)
+ *
+ * Rewritten in C by Stephen Rothwell.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version
+ * 2 of the License, or (at your option) any later version.
+ */
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/compiler.h>
+#include <linux/module.h>
+
+#include <asm/io.h>
+
+void _insb(volatile u8 __iomem *port, void *buf, long count)
+{
+	u8 *tbuf = buf;
+	u8 tmp;
+
+	if (unlikely(count <= 0))
+		return;
+	asm volatile("sync");
+	do {
+		tmp = *port;
+		asm volatile("eieio");
+		*tbuf++ = tmp;
+	} while (--count != 0);
+	asm volatile("twi 0,%0,0; isync" : : "r" (tmp));
+}
+EXPORT_SYMBOL(_insb);
+
+void _outsb(volatile u8 __iomem *port, const void *buf, long count)
+{
+	const u8 *tbuf = buf;
+
+	if (unlikely(count <= 0))
+		return;
+	asm volatile("sync");
+	do {
+		*port = *tbuf++;
+	} while (--count != 0);
+	asm volatile("sync");
+}
+EXPORT_SYMBOL(_outsb);
+
+void _insw_ns(volatile u16 __iomem *port, void *buf, long count)
+{
+	u16 *tbuf = buf;
+	u16 tmp;
+
+	if (unlikely(count <= 0))
+		return;
+	asm volatile("sync");
+	do {
+		tmp = *port;
+		asm volatile("eieio");
+		*tbuf++ = tmp;
+	} while (--count != 0);
+	asm volatile("twi 0,%0,0; isync" : : "r" (tmp));
+}
+EXPORT_SYMBOL(_insw_ns);
+
+void _outsw_ns(volatile u16 __iomem *port, const void *buf, long count)
+{
+	const u16 *tbuf = buf;
+
+	if (unlikely(count <= 0))
+		return;
+	asm volatile("sync");
+	do {
+		*port = *tbuf++;
+	} while (--count != 0);
+	asm volatile("sync");
+}
+EXPORT_SYMBOL(_outsw_ns);
+
+void _insl_ns(volatile u32 __iomem *port, void *buf, long count)
+{
+	u32 *tbuf = buf;
+	u32 tmp;
+
+	if (unlikely(count <= 0))
+		return;
+	asm volatile("sync");
+	do {
+		tmp = *port;
+		asm volatile("eieio");
+		*tbuf++ = tmp;
+	} while (--count != 0);
+	asm volatile("twi 0,%0,0; isync" : : "r" (tmp));
+}
+EXPORT_SYMBOL(_insl_ns);
+
+void _outsl_ns(volatile u32 __iomem *port, const void *buf, long count)
+{
+	const u32 *tbuf = buf;
+
+	if (unlikely(count <= 0))
+		return;
+	asm volatile("sync");
+	do {
+		*port = *tbuf++;
+	} while (--count != 0);
+	asm volatile("sync");
+}
+EXPORT_SYMBOL(_outsl_ns);
diff --git a/arch/powerpc/kernel/misc.S b/arch/powerpc/kernel/misc.S
index 6feb391422ec..330c9dc7db86 100644
--- a/arch/powerpc/kernel/misc.S
+++ b/arch/powerpc/kernel/misc.S
@@ -43,98 +43,3 @@ _GLOBAL(add_reloc_offset)
 	add	r3,r3,r5
 	mtlr	r0
 	blr
-
-/*
- * I/O string operations
- *
- * insb(port, buf, len)
- * outsb(port, buf, len)
- * insw(port, buf, len)
- * outsw(port, buf, len)
- * insl(port, buf, len)
- * outsl(port, buf, len)
- * insw_ns(port, buf, len)
- * outsw_ns(port, buf, len)
- * insl_ns(port, buf, len)
- * outsl_ns(port, buf, len)
- *
- * The *_ns versions don't do byte-swapping.
- */
-_GLOBAL(_insb)
-	sync
-	cmpwi	0,r5,0
-	mtctr	r5
-	subi	r4,r4,1
-	blelr-
-00:	lbz	r5,0(r3)
-	eieio
-	stbu	r5,1(r4)
-	bdnz	00b
-	twi	0,r5,0
-	isync
-	blr
-
-_GLOBAL(_outsb)
-	cmpwi	0,r5,0
-	mtctr	r5
-	subi	r4,r4,1
-	blelr-
-	sync
-00:	lbzu	r5,1(r4)
-	stb	r5,0(r3)
-	bdnz	00b
-	sync
-	blr
-
-_GLOBAL(_insw_ns)
-	sync
-	cmpwi	0,r5,0
-	mtctr	r5
-	subi	r4,r4,2
-	blelr-
-00:	lhz	r5,0(r3)
-	eieio
-	sthu	r5,2(r4)
-	bdnz	00b
-	twi	0,r5,0
-	isync
-	blr
-
-_GLOBAL(_outsw_ns)
-	cmpwi	0,r5,0
-	mtctr	r5
-	subi	r4,r4,2
-	blelr-
-	sync
-00:	lhzu	r5,2(r4)
-	sth	r5,0(r3)
-	bdnz	00b
-	sync
-	blr
-
-_GLOBAL(_insl_ns)
-	sync
-	cmpwi	0,r5,0
-	mtctr	r5
-	subi	r4,r4,4
-	blelr-
-00:	lwz	r5,0(r3)
-	eieio
-	stwu	r5,4(r4)
-	bdnz	00b
-	twi	0,r5,0
-	isync
-	blr
-
-_GLOBAL(_outsl_ns)
-	cmpwi	0,r5,0
-	mtctr	r5
-	subi	r4,r4,4
-	blelr-
-	sync
-00:	lwzu	r5,4(r4)
-	stw	r5,0(r3)
-	bdnz	00b
-	sync
-	blr
-
diff --git a/arch/powerpc/kernel/ppc_ksyms.c b/arch/powerpc/kernel/ppc_ksyms.c
index 75429e580518..807193a3c784 100644
--- a/arch/powerpc/kernel/ppc_ksyms.c
+++ b/arch/powerpc/kernel/ppc_ksyms.c
@@ -95,13 +95,6 @@ EXPORT_SYMBOL(__strnlen_user);
 EXPORT_SYMBOL(copy_4K_page);
 #endif
 
-EXPORT_SYMBOL(_insb);
-EXPORT_SYMBOL(_outsb);
-EXPORT_SYMBOL(_insw_ns);
-EXPORT_SYMBOL(_outsw_ns);
-EXPORT_SYMBOL(_insl_ns);
-EXPORT_SYMBOL(_outsl_ns);
-
 #if defined(CONFIG_PPC32) && (defined(CONFIG_BLK_DEV_IDE) || defined(CONFIG_BLK_DEV_IDE_MODULE))
 EXPORT_SYMBOL(ppc_ide_md);
 #endif
diff --git a/include/asm-powerpc/io.h b/include/asm-powerpc/io.h
index 51a598747367..57e7d14d6563 100644
--- a/include/asm-powerpc/io.h
+++ b/include/asm-powerpc/io.h
@@ -143,12 +143,12 @@ static inline void __raw_writeq(unsigned long v, volatile void __iomem *addr)
 #define readl_relaxed(addr) readl(addr)
 #define readq_relaxed(addr) readq(addr)
 
-extern void _insb(volatile u8 __iomem *port, void *buf, int ns);
-extern void _outsb(volatile u8 __iomem *port, const void *buf, int ns);
-extern void _insw_ns(volatile u16 __iomem *port, void *buf, int ns);
-extern void _outsw_ns(volatile u16 __iomem *port, const void *buf, int ns);
-extern void _insl_ns(volatile u32 __iomem *port, void *buf, int nl);
-extern void _outsl_ns(volatile u32 __iomem *port, const void *buf, int nl);
+extern void _insb(volatile u8 __iomem *port, void *buf, long count);
+extern void _outsb(volatile u8 __iomem *port, const void *buf, long count);
+extern void _insw_ns(volatile u16 __iomem *port, void *buf, long count);
+extern void _outsw_ns(volatile u16 __iomem *port, const void *buf, long count);
+extern void _insl_ns(volatile u32 __iomem *port, void *buf, long count);
+extern void _outsl_ns(volatile u32 __iomem *port, const void *buf, long count);
 
 static inline void mmiowb(void)
 {
diff --git a/include/asm-ppc/io.h b/include/asm-ppc/io.h
index 9fac420f1648..3d9a9e6f3321 100644
--- a/include/asm-ppc/io.h
+++ b/include/asm-ppc/io.h
@@ -327,12 +327,12 @@ __do_out_asm(outl, "stwbrx")
 #define inl_p(port)		inl((port))
 #define outl_p(val, port)	outl((val), (port))
 
-extern void _insb(volatile u8 __iomem *port, void *buf, int ns);
-extern void _outsb(volatile u8 __iomem *port, const void *buf, int ns);
-extern void _insw_ns(volatile u16 __iomem *port, void *buf, int ns);
-extern void _outsw_ns(volatile u16 __iomem *port, const void *buf, int ns);
-extern void _insl_ns(volatile u32 __iomem *port, void *buf, int nl);
-extern void _outsl_ns(volatile u32 __iomem *port, const void *buf, int nl);
+extern void _insb(volatile u8 __iomem *port, void *buf, long count);
+extern void _outsb(volatile u8 __iomem *port, const void *buf, long count);
+extern void _insw_ns(volatile u16 __iomem *port, void *buf, long count);
+extern void _outsw_ns(volatile u16 __iomem *port, const void *buf, long count);
+extern void _insl_ns(volatile u32 __iomem *port, void *buf, long count);
+extern void _outsl_ns(volatile u32 __iomem *port, const void *buf, long count);
 
 
 #define IO_SPACE_LIMIT ~0
-- 
cgit v1.2.3


From 19e59df4dc2e6f7b46190ee77ce7093769f597a7 Mon Sep 17 00:00:00 2001
From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Thu, 14 Sep 2006 14:55:36 +1000
Subject: [POWERPC] iseries: eliminate a couple of warnings

Copy and paste bug in io.h

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 include/asm-powerpc/io.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/include/asm-powerpc/io.h b/include/asm-powerpc/io.h
index 57e7d14d6563..174fb89d5eda 100644
--- a/include/asm-powerpc/io.h
+++ b/include/asm-powerpc/io.h
@@ -68,8 +68,8 @@ extern unsigned long pci_io_base;
  * for older code.
  */
 #define insb(port, buf, ns)	_insb((u8 __iomem *)((port)+pci_io_base), (buf), (ns))
-#define insw(port, buf, ns)	_insw_ns((u8 __iomem *)((port)+pci_io_base), (buf), (ns))
-#define insl(port, buf, nl)	_insl_ns((u8 __iomem *)((port)+pci_io_base), (buf), (nl))
+#define insw(port, buf, ns)	_insw_ns((u16 __iomem *)((port)+pci_io_base), (buf), (ns))
+#define insl(port, buf, nl)	_insl_ns((u32 __iomem *)((port)+pci_io_base), (buf), (nl))
 
 #else
 
-- 
cgit v1.2.3


From a4dc7ff08915a2035aa6d6decc53fa1deaa410bb Mon Sep 17 00:00:00 2001
From: Paul Mackerras <paulus@samba.org>
Date: Tue, 19 Sep 2006 14:06:27 +1000
Subject: [POWERPC] Define of_read_ulong helper

There are various places where we want to extract an unsigned long
value from a device-tree property that can be 1 or 2 cells in length.
This replaces some open-coded calculations, and one place where we
assumed without checking that properties were the length we wanted,
with a little of_read_ulong() helper.

Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 arch/powerpc/kernel/prom.c         | 19 ++-----------------
 arch/powerpc/kernel/setup-common.c | 13 ++++++++-----
 arch/powerpc/kernel/time.c         |  4 +---
 include/asm-powerpc/prom.h         | 12 +++++++++++-
 4 files changed, 22 insertions(+), 26 deletions(-)

(limited to 'include')

diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c
index bf2005b2feb6..eb913f80bfb1 100644
--- a/arch/powerpc/kernel/prom.c
+++ b/arch/powerpc/kernel/prom.c
@@ -757,24 +757,9 @@ static int __init early_init_dt_scan_root(unsigned long node,
 static unsigned long __init dt_mem_next_cell(int s, cell_t **cellp)
 {
 	cell_t *p = *cellp;
-	unsigned long r;
 
-	/* Ignore more than 2 cells */
-	while (s > sizeof(unsigned long) / 4) {
-		p++;
-		s--;
-	}
-	r = *p++;
-#ifdef CONFIG_PPC64
-	if (s > 1) {
-		r <<= 32;
-		r |= *(p++);
-		s--;
-	}
-#endif
-
-	*cellp = p;
-	return r;
+	*cellp = p + s;
+	return of_read_ulong(p, s);
 }
 
 
diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c
index 465e7435efbc..0af3fc1bdcc9 100644
--- a/arch/powerpc/kernel/setup-common.c
+++ b/arch/powerpc/kernel/setup-common.c
@@ -304,18 +304,21 @@ struct seq_operations cpuinfo_op = {
 void __init check_for_initrd(void)
 {
 #ifdef CONFIG_BLK_DEV_INITRD
-	const unsigned long *prop;
+	const unsigned int *prop;
+	int len;
 
 	DBG(" -> check_for_initrd()\n");
 
 	if (of_chosen) {
-		prop = get_property(of_chosen, "linux,initrd-start", NULL);
+		prop = get_property(of_chosen, "linux,initrd-start", &len);
 		if (prop != NULL) {
-			initrd_start = (unsigned long)__va(*prop);
+			initrd_start = (unsigned long)
+				__va(of_read_ulong(prop, len / 4));
 			prop = get_property(of_chosen,
-					"linux,initrd-end", NULL);
+					"linux,initrd-end", &len);
 			if (prop != NULL) {
-				initrd_end = (unsigned long)__va(*prop);
+				initrd_end = (unsigned long)
+					__va(of_read_ulong(prop, len / 4));
 				initrd_below_start_ok = 1;
 			} else
 				initrd_start = 0;
diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c
index b9a2061cfdb7..7a3c3f791ade 100644
--- a/arch/powerpc/kernel/time.c
+++ b/arch/powerpc/kernel/time.c
@@ -870,9 +870,7 @@ static int __init get_freq(char *name, int cells, unsigned long *val)
 		fp = get_property(cpu, name, NULL);
 		if (fp) {
 			found = 1;
-			*val = 0;
-			while (cells--)
-				*val = (*val << 32) | *fp++;
+			*val = of_read_ulong(fp, cells);
 		}
 
 		of_node_put(cpu);
diff --git a/include/asm-powerpc/prom.h b/include/asm-powerpc/prom.h
index c15e66a2e681..524629769336 100644
--- a/include/asm-powerpc/prom.h
+++ b/include/asm-powerpc/prom.h
@@ -197,7 +197,7 @@ extern int release_OF_resource(struct device_node* node, int index);
  */
 
 
-/* Helper to read a big number */
+/* Helper to read a big number; size is in cells (not bytes) */
 static inline u64 of_read_number(const u32 *cell, int size)
 {
 	u64 r = 0;
@@ -206,6 +206,16 @@ static inline u64 of_read_number(const u32 *cell, int size)
 	return r;
 }
 
+/* Like of_read_number, but we want an unsigned long result */
+#ifdef CONFIG_PPC32
+static inline unsigned long of_read_ulong(const u32 *cell, int size)
+{
+	return cell[size-1];
+}
+#else
+#define of_read_ulong(cell, size)	of_read_number(cell, size)
+#endif
+
 /* Translate an OF address block into a CPU physical address
  */
 #define OF_BAD_ADDR	((u64)-1)
-- 
cgit v1.2.3


From de1a3f1ce6c4c3b2b14cf9157a22d6b4c64f708e Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Wed, 20 Sep 2006 15:58:20 +0200
Subject: [S390] EX_TABLE macro.

Add EX_TABLE helper macro to simplify creation of inline assembly
exception table entries.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 include/asm-s390/processor.h | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

(limited to 'include')

diff --git a/include/asm-s390/processor.h b/include/asm-s390/processor.h
index 5b71d3731723..a3a4e5fd30d7 100644
--- a/include/asm-s390/processor.h
+++ b/include/asm-s390/processor.h
@@ -339,4 +339,21 @@ int unregister_idle_notifier(struct notifier_block *nb);
 
 #endif
 
+/*
+ * Helper macro for exception table entries
+ */
+#ifndef __s390x__
+#define EX_TABLE(_fault,_target)			\
+	".section __ex_table,\"a\"\n"			\
+	"	.align 4\n"				\
+	"	.long  " #_fault "," #_target "\n"	\
+	".previous\n"
+#else
+#define EX_TABLE(_fault,_target)			\
+	".section __ex_table,\"a\"\n"			\
+	"	.align 8\n"				\
+	"	.quad  " #_fault "," #_target "\n"	\
+	".previous\n"
+#endif
+
 #endif                                 /* __ASM_S390_PROCESSOR_H           */
-- 
cgit v1.2.3


From 7561b974e0cbbdca1bb880b55200afd9a1a20737 Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Wed, 20 Sep 2006 15:58:22 +0200
Subject: [S390] remove old z90crypt driver.

The z90crypt driver has served its term. It is replaced by the shiny
new zcrypt device driver.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 drivers/s390/Kconfig              |    9 -
 drivers/s390/crypto/Makefile      |    2 -
 drivers/s390/crypto/z90common.h   |  166 --
 drivers/s390/crypto/z90crypt.h    |   71 -
 drivers/s390/crypto/z90hardware.c | 2531 ---------------------------
 drivers/s390/crypto/z90main.c     | 3379 -------------------------------------
 include/asm-s390/z90crypt.h       |  212 ---
 7 files changed, 6370 deletions(-)
 delete mode 100644 drivers/s390/crypto/z90common.h
 delete mode 100644 drivers/s390/crypto/z90crypt.h
 delete mode 100644 drivers/s390/crypto/z90hardware.c
 delete mode 100644 drivers/s390/crypto/z90main.c
 delete mode 100644 include/asm-s390/z90crypt.h

(limited to 'include')

diff --git a/drivers/s390/Kconfig b/drivers/s390/Kconfig
index 4d36208ff8de..f0ea550d39bc 100644
--- a/drivers/s390/Kconfig
+++ b/drivers/s390/Kconfig
@@ -217,13 +217,4 @@ endmenu
 
 menu "Cryptographic devices"
 
-config Z90CRYPT
-	tristate "Support for PCI-attached cryptographic adapters"
-        default "m"
-        help
-	  Select this option if you want to use a PCI-attached cryptographic
-	  adapter like the PCI Cryptographic Accelerator (PCICA) or the PCI
-	  Cryptographic Coprocessor (PCICC).  This option is also available
-	  as a module called z90crypt.ko.
-
 endmenu
diff --git a/drivers/s390/crypto/Makefile b/drivers/s390/crypto/Makefile
index 15edebbead7f..67e75be8e4e4 100644
--- a/drivers/s390/crypto/Makefile
+++ b/drivers/s390/crypto/Makefile
@@ -2,5 +2,3 @@
 # S/390 crypto devices
 #
 
-z90crypt-objs := z90main.o z90hardware.o
-obj-$(CONFIG_Z90CRYPT) += z90crypt.o
diff --git a/drivers/s390/crypto/z90common.h b/drivers/s390/crypto/z90common.h
deleted file mode 100644
index dbbcda3c846a..000000000000
--- a/drivers/s390/crypto/z90common.h
+++ /dev/null
@@ -1,166 +0,0 @@
-/*
- *  linux/drivers/s390/crypto/z90common.h
- *
- *  z90crypt 1.3.3
- *
- *  Copyright (C)  2001, 2005 IBM Corporation
- *  Author(s): Robert Burroughs (burrough@us.ibm.com)
- *             Eric Rossman (edrossma@us.ibm.com)
- *
- *  Hotplug & misc device support: Jochen Roehrig (roehrig@de.ibm.com)
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2, or (at your option)
- * any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
- */
-
-#ifndef _Z90COMMON_H_
-#define _Z90COMMON_H_
-
-
-#define RESPBUFFSIZE 256
-#define PCI_FUNC_KEY_DECRYPT 0x5044
-#define PCI_FUNC_KEY_ENCRYPT 0x504B
-extern int ext_bitlens;
-
-enum devstat {
-	DEV_GONE,
-	DEV_ONLINE,
-	DEV_QUEUE_FULL,
-	DEV_EMPTY,
-	DEV_NO_WORK,
-	DEV_BAD_MESSAGE,
-	DEV_TSQ_EXCEPTION,
-	DEV_RSQ_EXCEPTION,
-	DEV_SEN_EXCEPTION,
-	DEV_REC_EXCEPTION
-};
-
-enum hdstat {
-	HD_NOT_THERE,
-	HD_BUSY,
-	HD_DECONFIGURED,
-	HD_CHECKSTOPPED,
-	HD_ONLINE,
-	HD_TSQ_EXCEPTION
-};
-
-#define Z90C_NO_DEVICES		1
-#define Z90C_AMBIGUOUS_DOMAIN	2
-#define Z90C_INCORRECT_DOMAIN	3
-#define ENOTINIT		4
-
-#define SEN_BUSY	 7
-#define SEN_USER_ERROR	 8
-#define SEN_QUEUE_FULL	11
-#define SEN_NOT_AVAIL	16
-#define SEN_PAD_ERROR	17
-#define SEN_RETRY	18
-#define SEN_RELEASED	24
-
-#define REC_EMPTY	 4
-#define REC_BUSY	 6
-#define REC_OPERAND_INV	 8
-#define REC_OPERAND_SIZE 9
-#define REC_EVEN_MOD	10
-#define REC_NO_WORK	11
-#define REC_HARDWAR_ERR	12
-#define REC_NO_RESPONSE	13
-#define REC_RETRY_DEV	14
-#define REC_USER_GONE	15
-#define REC_BAD_MESSAGE	16
-#define REC_INVALID_PAD	17
-#define REC_USE_PCICA	18
-
-#define WRONG_DEVICE_TYPE 20
-
-#define REC_FATAL_ERROR 32
-#define SEN_FATAL_ERROR 33
-#define TSQ_FATAL_ERROR 34
-#define RSQ_FATAL_ERROR 35
-
-#define Z90CRYPT_NUM_TYPES	6
-#define PCICA		0
-#define PCICC		1
-#define PCIXCC_MCL2	2
-#define PCIXCC_MCL3	3
-#define CEX2C		4
-#define CEX2A		5
-#define NILDEV		-1
-#define ANYDEV		-1
-#define PCIXCC_UNK	-2
-
-enum hdevice_type {
-	PCICC_HW  = 3,
-	PCICA_HW  = 4,
-	PCIXCC_HW = 5,
-	CEX2A_HW  = 6,
-	CEX2C_HW  = 7
-};
-
-struct CPRBX {
-	unsigned short cprb_len;
-	unsigned char  cprb_ver_id;
-	unsigned char  pad_000[3];
-	unsigned char  func_id[2];
-	unsigned char  cprb_flags[4];
-	unsigned int   req_parml;
-	unsigned int   req_datal;
-	unsigned int   rpl_msgbl;
-	unsigned int   rpld_parml;
-	unsigned int   rpl_datal;
-	unsigned int   rpld_datal;
-	unsigned int   req_extbl;
-	unsigned char  pad_001[4];
-	unsigned int   rpld_extbl;
-	unsigned char  req_parmb[16];
-	unsigned char  req_datab[16];
-	unsigned char  rpl_parmb[16];
-	unsigned char  rpl_datab[16];
-	unsigned char  req_extb[16];
-	unsigned char  rpl_extb[16];
-	unsigned short ccp_rtcode;
-	unsigned short ccp_rscode;
-	unsigned int   mac_data_len;
-	unsigned char  logon_id[8];
-	unsigned char  mac_value[8];
-	unsigned char  mac_content_flgs;
-	unsigned char  pad_002;
-	unsigned short domain;
-	unsigned char  pad_003[12];
-	unsigned char  pad_004[36];
-};
-
-#ifndef DEV_NAME
-#define DEV_NAME	"z90crypt"
-#endif
-#define PRINTK(fmt, args...) \
-	printk(KERN_DEBUG DEV_NAME ": %s -> " fmt, __FUNCTION__ , ## args)
-#define PRINTKN(fmt, args...) \
-	printk(KERN_DEBUG DEV_NAME ": " fmt, ## args)
-#define PRINTKW(fmt, args...) \
-	printk(KERN_WARNING DEV_NAME ": %s -> " fmt, __FUNCTION__ , ## args)
-#define PRINTKC(fmt, args...) \
-	printk(KERN_CRIT DEV_NAME ": %s -> " fmt, __FUNCTION__ , ## args)
-
-#ifdef Z90CRYPT_DEBUG
-#define PDEBUG(fmt, args...) \
-	printk(KERN_DEBUG DEV_NAME ": %s -> " fmt, __FUNCTION__ , ## args)
-#else
-#define PDEBUG(fmt, args...) do {} while (0)
-#endif
-
-#define UMIN(a,b) ((a) < (b) ? (a) : (b))
-#define IS_EVEN(x) ((x) == (2 * ((x) / 2)))
-
-#endif
diff --git a/drivers/s390/crypto/z90crypt.h b/drivers/s390/crypto/z90crypt.h
deleted file mode 100644
index 0ca1d126ccb6..000000000000
--- a/drivers/s390/crypto/z90crypt.h
+++ /dev/null
@@ -1,71 +0,0 @@
-/*
- *  linux/drivers/s390/crypto/z90crypt.h
- *
- *  z90crypt 1.3.3 (kernel-private header)
- *
- *  Copyright (C)  2001, 2005 IBM Corporation
- *  Author(s): Robert Burroughs (burrough@us.ibm.com)
- *             Eric Rossman (edrossma@us.ibm.com)
- *
- *  Hotplug & misc device support: Jochen Roehrig (roehrig@de.ibm.com)
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2, or (at your option)
- * any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
- */
-
-#ifndef _Z90CRYPT_H_
-#define _Z90CRYPT_H_
-
-#include <asm/z90crypt.h>
-
-/**
- * local errno definitions
- */
-#define ENOBUFF	  129	// filp->private_data->...>work_elem_p->buffer is NULL
-#define EWORKPEND 130	// user issues ioctl while another pending
-#define ERELEASED 131	// user released while ioctl pending
-#define EQUIESCE  132	// z90crypt quiescing (no more work allowed)
-#define ETIMEOUT  133	// request timed out
-#define EUNKNOWN  134	// some unrecognized error occured (retry may succeed)
-#define EGETBUFF  135	// Error getting buffer or hardware lacks capability
-			// (retry in software)
-
-/**
- * DEPRECATED STRUCTURES
- */
-
-/**
- * This structure is DEPRECATED and the corresponding ioctl() has been
- * replaced with individual ioctl()s for each piece of data!
- * This structure will NOT survive past version 1.3.1, so switch to the
- * new ioctl()s.
- */
-#define MASK_LENGTH 64 // mask length
-struct ica_z90_status {
-	int totalcount;
-	int leedslitecount; // PCICA
-	int leeds2count;    // PCICC
-	// int PCIXCCCount; is not in struct for backward compatibility
-	int requestqWaitCount;
-	int pendingqWaitCount;
-	int totalOpenCount;
-	int cryptoDomain;
-	// status: 0=not there, 1=PCICA, 2=PCICC, 3=PCIXCC_MCL2, 4=PCIXCC_MCL3,
-	//         5=CEX2C
-	unsigned char status[MASK_LENGTH];
-	// qdepth: # work elements waiting for each device
-	unsigned char qdepth[MASK_LENGTH];
-};
-
-#endif /* _Z90CRYPT_H_ */
diff --git a/drivers/s390/crypto/z90hardware.c b/drivers/s390/crypto/z90hardware.c
deleted file mode 100644
index be60795f4a74..000000000000
--- a/drivers/s390/crypto/z90hardware.c
+++ /dev/null
@@ -1,2531 +0,0 @@
-/*
- *  linux/drivers/s390/crypto/z90hardware.c
- *
- *  z90crypt 1.3.3
- *
- *  Copyright (C)  2001, 2005 IBM Corporation
- *  Author(s): Robert Burroughs (burrough@us.ibm.com)
- *             Eric Rossman (edrossma@us.ibm.com)
- *
- *  Hotplug & misc device support: Jochen Roehrig (roehrig@de.ibm.com)
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2, or (at your option)
- * any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
- */
-
-#include <asm/uaccess.h>
-#include <linux/compiler.h>
-#include <linux/delay.h>
-#include <linux/init.h>
-#include <linux/module.h>
-#include "z90crypt.h"
-#include "z90common.h"
-
-struct cca_token_hdr {
-	unsigned char  token_identifier;
-	unsigned char  version;
-	unsigned short token_length;
-	unsigned char  reserved[4];
-};
-
-#define CCA_TKN_HDR_ID_EXT 0x1E
-
-struct cca_private_ext_ME_sec {
-	unsigned char  section_identifier;
-	unsigned char  version;
-	unsigned short section_length;
-	unsigned char  private_key_hash[20];
-	unsigned char  reserved1[4];
-	unsigned char  key_format;
-	unsigned char  reserved2;
-	unsigned char  key_name_hash[20];
-	unsigned char  key_use_flags[4];
-	unsigned char  reserved3[6];
-	unsigned char  reserved4[24];
-	unsigned char  confounder[24];
-	unsigned char  exponent[128];
-	unsigned char  modulus[128];
-};
-
-#define CCA_PVT_USAGE_ALL 0x80
-
-struct cca_public_sec {
-	unsigned char  section_identifier;
-	unsigned char  version;
-	unsigned short section_length;
-	unsigned char  reserved[2];
-	unsigned short exponent_len;
-	unsigned short modulus_bit_len;
-	unsigned short modulus_byte_len;
-	unsigned char  exponent[3];
-};
-
-struct cca_private_ext_ME {
-	struct cca_token_hdr	      pvtMEHdr;
-	struct cca_private_ext_ME_sec pvtMESec;
-	struct cca_public_sec	      pubMESec;
-};
-
-struct cca_public_key {
-	struct cca_token_hdr  pubHdr;
-	struct cca_public_sec pubSec;
-};
-
-struct cca_pvt_ext_CRT_sec {
-	unsigned char  section_identifier;
-	unsigned char  version;
-	unsigned short section_length;
-	unsigned char  private_key_hash[20];
-	unsigned char  reserved1[4];
-	unsigned char  key_format;
-	unsigned char  reserved2;
-	unsigned char  key_name_hash[20];
-	unsigned char  key_use_flags[4];
-	unsigned short p_len;
-	unsigned short q_len;
-	unsigned short dp_len;
-	unsigned short dq_len;
-	unsigned short u_len;
-	unsigned short mod_len;
-	unsigned char  reserved3[4];
-	unsigned short pad_len;
-	unsigned char  reserved4[52];
-	unsigned char  confounder[8];
-};
-
-#define CCA_PVT_EXT_CRT_SEC_ID_PVT 0x08
-#define CCA_PVT_EXT_CRT_SEC_FMT_CL 0x40
-
-struct cca_private_ext_CRT {
-	struct cca_token_hdr	   pvtCrtHdr;
-	struct cca_pvt_ext_CRT_sec pvtCrtSec;
-	struct cca_public_sec	   pubCrtSec;
-};
-
-struct ap_status_word {
-	unsigned char q_stat_flags;
-	unsigned char response_code;
-	unsigned char reserved[2];
-};
-
-#define AP_Q_STATUS_EMPTY		0x80
-#define AP_Q_STATUS_REPLIES_WAITING	0x40
-#define AP_Q_STATUS_ARRAY_FULL		0x20
-
-#define AP_RESPONSE_NORMAL		0x00
-#define AP_RESPONSE_Q_NOT_AVAIL		0x01
-#define AP_RESPONSE_RESET_IN_PROGRESS	0x02
-#define AP_RESPONSE_DECONFIGURED	0x03
-#define AP_RESPONSE_CHECKSTOPPED	0x04
-#define AP_RESPONSE_BUSY		0x05
-#define AP_RESPONSE_Q_FULL		0x10
-#define AP_RESPONSE_NO_PENDING_REPLY	0x10
-#define AP_RESPONSE_INDEX_TOO_BIG	0x11
-#define AP_RESPONSE_NO_FIRST_PART	0x13
-#define AP_RESPONSE_MESSAGE_TOO_BIG	0x15
-
-#define AP_MAX_CDX_BITL		4
-#define AP_RQID_RESERVED_BITL	4
-#define SKIP_BITL		(AP_MAX_CDX_BITL + AP_RQID_RESERVED_BITL)
-
-struct type4_hdr {
-	unsigned char  reserved1;
-	unsigned char  msg_type_code;
-	unsigned short msg_len;
-	unsigned char  request_code;
-	unsigned char  msg_fmt;
-	unsigned short reserved2;
-};
-
-#define TYPE4_TYPE_CODE 0x04
-#define TYPE4_REQU_CODE 0x40
-
-#define TYPE4_SME_LEN 0x0188
-#define TYPE4_LME_LEN 0x0308
-#define TYPE4_SCR_LEN 0x01E0
-#define TYPE4_LCR_LEN 0x03A0
-
-#define TYPE4_SME_FMT 0x00
-#define TYPE4_LME_FMT 0x10
-#define TYPE4_SCR_FMT 0x40
-#define TYPE4_LCR_FMT 0x50
-
-struct type4_sme {
-	struct type4_hdr header;
-	unsigned char	 message[128];
-	unsigned char	 exponent[128];
-	unsigned char	 modulus[128];
-};
-
-struct type4_lme {
-	struct type4_hdr header;
-	unsigned char	 message[256];
-	unsigned char	 exponent[256];
-	unsigned char	 modulus[256];
-};
-
-struct type4_scr {
-	struct type4_hdr header;
-	unsigned char	 message[128];
-	unsigned char	 dp[72];
-	unsigned char	 dq[64];
-	unsigned char	 p[72];
-	unsigned char	 q[64];
-	unsigned char	 u[72];
-};
-
-struct type4_lcr {
-	struct type4_hdr header;
-	unsigned char	 message[256];
-	unsigned char	 dp[136];
-	unsigned char	 dq[128];
-	unsigned char	 p[136];
-	unsigned char	 q[128];
-	unsigned char	 u[136];
-};
-
-union type4_msg {
-	struct type4_sme sme;
-	struct type4_lme lme;
-	struct type4_scr scr;
-	struct type4_lcr lcr;
-};
-
-struct type84_hdr {
-	unsigned char  reserved1;
-	unsigned char  code;
-	unsigned short len;
-	unsigned char  reserved2[4];
-};
-
-#define TYPE84_RSP_CODE 0x84
-
-struct type6_hdr {
-	unsigned char reserved1;
-	unsigned char type;
-	unsigned char reserved2[2];
-	unsigned char right[4];
-	unsigned char reserved3[2];
-	unsigned char reserved4[2];
-	unsigned char apfs[4];
-	unsigned int  offset1;
-	unsigned int  offset2;
-	unsigned int  offset3;
-	unsigned int  offset4;
-	unsigned char agent_id[16];
-	unsigned char rqid[2];
-	unsigned char reserved5[2];
-	unsigned char function_code[2];
-	unsigned char reserved6[2];
-	unsigned int  ToCardLen1;
-	unsigned int  ToCardLen2;
-	unsigned int  ToCardLen3;
-	unsigned int  ToCardLen4;
-	unsigned int  FromCardLen1;
-	unsigned int  FromCardLen2;
-	unsigned int  FromCardLen3;
-	unsigned int  FromCardLen4;
-};
-
-struct CPRB {
-	unsigned char cprb_len[2];
-	unsigned char cprb_ver_id;
-	unsigned char pad_000;
-	unsigned char srpi_rtcode[4];
-	unsigned char srpi_verb;
-	unsigned char flags;
-	unsigned char func_id[2];
-	unsigned char checkpoint_flag;
-	unsigned char resv2;
-	unsigned char req_parml[2];
-	unsigned char req_parmp[4];
-	unsigned char req_datal[4];
-	unsigned char req_datap[4];
-	unsigned char rpl_parml[2];
-	unsigned char pad_001[2];
-	unsigned char rpl_parmp[4];
-	unsigned char rpl_datal[4];
-	unsigned char rpl_datap[4];
-	unsigned char ccp_rscode[2];
-	unsigned char ccp_rtcode[2];
-	unsigned char repd_parml[2];
-	unsigned char mac_data_len[2];
-	unsigned char repd_datal[4];
-	unsigned char req_pc[2];
-	unsigned char res_origin[8];
-	unsigned char mac_value[8];
-	unsigned char logon_id[8];
-	unsigned char usage_domain[2];
-	unsigned char resv3[18];
-	unsigned char svr_namel[2];
-	unsigned char svr_name[8];
-};
-
-struct type6_msg {
-	struct type6_hdr header;
-	struct CPRB	 CPRB;
-};
-
-struct type86_hdr {
-	unsigned char reserved1;
-	unsigned char type;
-	unsigned char format;
-	unsigned char reserved2;
-	unsigned char reply_code;
-	unsigned char reserved3[3];
-};
-
-#define TYPE86_RSP_CODE 0x86
-#define TYPE86_FMT2	0x02
-
-struct type86_fmt2_msg {
-	struct type86_hdr header;
-	unsigned char	  reserved[4];
-	unsigned char	  apfs[4];
-	unsigned int	  count1;
-	unsigned int	  offset1;
-	unsigned int	  count2;
-	unsigned int	  offset2;
-	unsigned int	  count3;
-	unsigned int	  offset3;
-	unsigned int	  count4;
-	unsigned int	  offset4;
-};
-
-static struct type6_hdr static_type6_hdr = {
-	0x00,
-	0x06,
-	{0x00,0x00},
-	{0x00,0x00,0x00,0x00},
-	{0x00,0x00},
-	{0x00,0x00},
-	{0x00,0x00,0x00,0x00},
-	0x00000058,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	{0x01,0x00,0x43,0x43,0x41,0x2D,0x41,0x50,
-	 0x50,0x4C,0x20,0x20,0x20,0x01,0x01,0x01},
-	{0x00,0x00},
-	{0x00,0x00},
-	{0x50,0x44},
-	{0x00,0x00},
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000
-};
-
-static struct type6_hdr static_type6_hdrX = {
-	0x00,
-	0x06,
-	{0x00,0x00},
-	{0x00,0x00,0x00,0x00},
-	{0x00,0x00},
-	{0x00,0x00},
-	{0x00,0x00,0x00,0x00},
-	0x00000058,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	{0x43,0x41,0x00,0x00,0x00,0x00,0x00,0x00,
-	 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00},
-	{0x00,0x00},
-	{0x00,0x00},
-	{0x50,0x44},
-	{0x00,0x00},
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000
-};
-
-static struct CPRB static_cprb = {
-	{0x70,0x00},
-	0x41,
-	0x00,
-	{0x00,0x00,0x00,0x00},
-	0x00,
-	0x00,
-	{0x54,0x32},
-	0x01,
-	0x00,
-	{0x00,0x00},
-	{0x00,0x00,0x00,0x00},
-	{0x00,0x00,0x00,0x00},
-	{0x00,0x00,0x00,0x00},
-	{0x00,0x00},
-	{0x00,0x00},
-	{0x00,0x00,0x00,0x00},
-	{0x00,0x00,0x00,0x00},
-	{0x00,0x00,0x00,0x00},
-	{0x00,0x00},
-	{0x00,0x00},
-	{0x00,0x00},
-	{0x00,0x00},
-	{0x00,0x00,0x00,0x00},
-	{0x00,0x00},
-	{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00},
-	{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00},
-	{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00},
-	{0x00,0x00},
-	{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-	 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-	 0x00,0x00},
-	{0x08,0x00},
-	{0x49,0x43,0x53,0x46,0x20,0x20,0x20,0x20}
-};
-
-struct function_and_rules_block {
-	unsigned char function_code[2];
-	unsigned char ulen[2];
-	unsigned char only_rule[8];
-};
-
-static struct function_and_rules_block static_pkd_function_and_rules = {
-	{0x50,0x44},
-	{0x0A,0x00},
-	{'P','K','C','S','-','1','.','2'}
-};
-
-static struct function_and_rules_block static_pke_function_and_rules = {
-	{0x50,0x4B},
-	{0x0A,0x00},
-	{'P','K','C','S','-','1','.','2'}
-};
-
-struct T6_keyBlock_hdr {
-	unsigned char blen[2];
-	unsigned char ulen[2];
-	unsigned char flags[2];
-};
-
-static struct T6_keyBlock_hdr static_T6_keyBlock_hdr = {
-	{0x89,0x01},
-	{0x87,0x01},
-	{0x00}
-};
-
-static struct CPRBX static_cprbx = {
-	0x00DC,
-	0x02,
-	{0x00,0x00,0x00},
-	{0x54,0x32},
-	{0x00,0x00,0x00,0x00},
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	{0x00,0x00,0x00,0x00},
-	0x00000000,
-	{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-	 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00},
-	{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-	 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00},
-	{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-	 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00},
-	{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-	 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00},
-	{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-	 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00},
-	{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-	 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00},
-	0x0000,
-	0x0000,
-	0x00000000,
-	{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00},
-	{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00},
-	0x00,
-	0x00,
-	0x0000,
-	{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00},
-	{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-	 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-	 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00}
-};
-
-static struct function_and_rules_block static_pkd_function_and_rulesX_MCL2 = {
-	{0x50,0x44},
-	{0x00,0x0A},
-	{'P','K','C','S','-','1','.','2'}
-};
-
-static struct function_and_rules_block static_pke_function_and_rulesX_MCL2 = {
-	{0x50,0x4B},
-	{0x00,0x0A},
-	{'Z','E','R','O','-','P','A','D'}
-};
-
-static struct function_and_rules_block static_pkd_function_and_rulesX = {
-	{0x50,0x44},
-	{0x00,0x0A},
-	{'Z','E','R','O','-','P','A','D'}
-};
-
-static struct function_and_rules_block static_pke_function_and_rulesX = {
-	{0x50,0x4B},
-	{0x00,0x0A},
-	{'M','R','P',' ',' ',' ',' ',' '}
-};
-
-static unsigned char static_PKE_function_code[2] = {0x50, 0x4B};
-
-struct T6_keyBlock_hdrX {
-	unsigned short blen;
-	unsigned short ulen;
-	unsigned char flags[2];
-};
-
-static unsigned char static_pad[256] = {
-0x1B,0x7B,0x5D,0xB5,0x75,0x01,0x3D,0xFD,0x8D,0xD1,0xC7,0x03,0x2D,0x09,0x23,0x57,
-0x89,0x49,0xB9,0x3F,0xBB,0x99,0x41,0x5B,0x75,0x21,0x7B,0x9D,0x3B,0x6B,0x51,0x39,
-0xBB,0x0D,0x35,0xB9,0x89,0x0F,0x93,0xA5,0x0B,0x47,0xF1,0xD3,0xBB,0xCB,0xF1,0x9D,
-0x23,0x73,0x71,0xFF,0xF3,0xF5,0x45,0xFB,0x61,0x29,0x23,0xFD,0xF1,0x29,0x3F,0x7F,
-0x17,0xB7,0x1B,0xA9,0x19,0xBD,0x57,0xA9,0xD7,0x95,0xA3,0xCB,0xED,0x1D,0xDB,0x45,
-0x7D,0x11,0xD1,0x51,0x1B,0xED,0x71,0xE9,0xB1,0xD1,0xAB,0xAB,0x21,0x2B,0x1B,0x9F,
-0x3B,0x9F,0xF7,0xF7,0xBD,0x63,0xEB,0xAD,0xDF,0xB3,0x6F,0x5B,0xDB,0x8D,0xA9,0x5D,
-0xE3,0x7D,0x77,0x49,0x47,0xF5,0xA7,0xFD,0xAB,0x2F,0x27,0x35,0x77,0xD3,0x49,0xC9,
-0x09,0xEB,0xB1,0xF9,0xBF,0x4B,0xCB,0x2B,0xEB,0xEB,0x05,0xFF,0x7D,0xC7,0x91,0x8B,
-0x09,0x83,0xB9,0xB9,0x69,0x33,0x39,0x6B,0x79,0x75,0x19,0xBF,0xBB,0x07,0x1D,0xBD,
-0x29,0xBF,0x39,0x95,0x93,0x1D,0x35,0xC7,0xC9,0x4D,0xE5,0x97,0x0B,0x43,0x9B,0xF1,
-0x16,0x93,0x03,0x1F,0xA5,0xFB,0xDB,0xF3,0x27,0x4F,0x27,0x61,0x05,0x1F,0xB9,0x23,
-0x2F,0xC3,0x81,0xA9,0x23,0x71,0x55,0x55,0xEB,0xED,0x41,0xE5,0xF3,0x11,0xF1,0x43,
-0x69,0x03,0xBD,0x0B,0x37,0x0F,0x51,0x8F,0x0B,0xB5,0x89,0x5B,0x67,0xA9,0xD9,0x4F,
-0x01,0xF9,0x21,0x77,0x37,0x73,0x79,0xC5,0x7F,0x51,0xC1,0xCF,0x97,0xA1,0x75,0xAD,
-0x35,0x9D,0xD3,0xD3,0xA7,0x9D,0x5D,0x41,0x6F,0x65,0x1B,0xCF,0xA9,0x87,0x91,0x09
-};
-
-static struct cca_private_ext_ME static_pvt_me_key = {
-	{
-		0x1E,
-		0x00,
-		0x0183,
-		{0x00,0x00,0x00,0x00}
-	},
-
-	{
-		0x02,
-		0x00,
-		0x016C,
-		{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00},
-		{0x00,0x00,0x00,0x00},
-		0x00,
-		0x00,
-		{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00},
-		{0x80,0x00,0x00,0x00},
-		{0x00,0x00,0x00,0x00,0x00,0x00},
-		{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00},
-		{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00},
-		{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00},
-		{0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-		 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00}
-	},
-
-	{
-		0x04,
-		0x00,
-		0x000F,
-		{0x00,0x00},
-		0x0003,
-		0x0000,
-		0x0000,
-		{0x01,0x00,0x01}
-	}
-};
-
-static struct cca_public_key static_public_key = {
-	{
-		0x1E,
-		0x00,
-		0x0000,
-		{0x00,0x00,0x00,0x00}
-	},
-
-	{
-		0x04,
-		0x00,
-		0x0000,
-		{0x00,0x00},
-		0x0000,
-		0x0000,
-		0x0000,
-		{0x01,0x00,0x01}
-	}
-};
-
-#define FIXED_TYPE6_ME_LEN 0x0000025F
-
-#define FIXED_TYPE6_ME_EN_LEN 0x000000F0
-
-#define FIXED_TYPE6_ME_LENX 0x000002CB
-
-#define FIXED_TYPE6_ME_EN_LENX 0x0000015C
-
-static struct cca_public_sec static_cca_pub_sec = {
-	0x04,
-	0x00,
-	0x000f,
-	{0x00,0x00},
-	0x0003,
-	0x0000,
-	0x0000,
-	{0x01,0x00,0x01}
-};
-
-#define FIXED_TYPE6_CR_LEN 0x00000177
-
-#define FIXED_TYPE6_CR_LENX 0x000001E3
-
-#define MAX_RESPONSE_SIZE 0x00000710
-
-#define MAX_RESPONSEX_SIZE 0x0000077C
-
-#define RESPONSE_CPRB_SIZE  0x000006B8
-#define RESPONSE_CPRBX_SIZE 0x00000724
-
-struct type50_hdr {
-	u8    reserved1;
-	u8    msg_type_code;
-	u16   msg_len;
-	u8    reserved2;
-	u8    ignored;
-	u16   reserved3;
-};
-
-#define TYPE50_TYPE_CODE 0x50
-
-#define TYPE50_MEB1_LEN (sizeof(struct type50_meb1_msg))
-#define TYPE50_MEB2_LEN (sizeof(struct type50_meb2_msg))
-#define TYPE50_CRB1_LEN (sizeof(struct type50_crb1_msg))
-#define TYPE50_CRB2_LEN (sizeof(struct type50_crb2_msg))
-
-#define TYPE50_MEB1_FMT 0x0001
-#define TYPE50_MEB2_FMT 0x0002
-#define TYPE50_CRB1_FMT 0x0011
-#define TYPE50_CRB2_FMT 0x0012
-
-struct type50_meb1_msg {
-	struct type50_hdr	header;
-	u16			keyblock_type;
-	u8			reserved[6];
-	u8			exponent[128];
-	u8			modulus[128];
-	u8			message[128];
-};
-
-struct type50_meb2_msg {
-	struct type50_hdr	header;
-	u16			keyblock_type;
-	u8			reserved[6];
-	u8			exponent[256];
-	u8			modulus[256];
-	u8			message[256];
-};
-
-struct type50_crb1_msg {
-	struct type50_hdr	header;
-	u16			keyblock_type;
-	u8			reserved[6];
-	u8			p[64];
-	u8			q[64];
-	u8			dp[64];
-	u8			dq[64];
-	u8			u[64];
-	u8			message[128];
-};
-
-struct type50_crb2_msg {
-	struct type50_hdr	header;
-	u16			keyblock_type;
-	u8			reserved[6];
-	u8			p[128];
-	u8			q[128];
-	u8			dp[128];
-	u8			dq[128];
-	u8			u[128];
-	u8			message[256];
-};
-
-union type50_msg {
-	struct type50_meb1_msg meb1;
-	struct type50_meb2_msg meb2;
-	struct type50_crb1_msg crb1;
-	struct type50_crb2_msg crb2;
-};
-
-struct type80_hdr {
-	u8	reserved1;
-	u8	type;
-	u16	len;
-	u8	code;
-	u8	reserved2[3];
-	u8	reserved3[8];
-};
-
-#define TYPE80_RSP_CODE 0x80
-
-struct error_hdr {
-	unsigned char reserved1;
-	unsigned char type;
-	unsigned char reserved2[2];
-	unsigned char reply_code;
-	unsigned char reserved3[3];
-};
-
-#define TYPE82_RSP_CODE 0x82
-#define TYPE88_RSP_CODE 0x88
-
-#define REP82_ERROR_MACHINE_FAILURE  0x10
-#define REP82_ERROR_PREEMPT_FAILURE  0x12
-#define REP82_ERROR_CHECKPT_FAILURE  0x14
-#define REP82_ERROR_MESSAGE_TYPE     0x20
-#define REP82_ERROR_INVALID_COMM_CD  0x21
-#define REP82_ERROR_INVALID_MSG_LEN  0x23
-#define REP82_ERROR_RESERVD_FIELD    0x24
-#define REP82_ERROR_FORMAT_FIELD     0x29
-#define REP82_ERROR_INVALID_COMMAND  0x30
-#define REP82_ERROR_MALFORMED_MSG    0x40
-#define REP82_ERROR_RESERVED_FIELDO  0x50
-#define REP82_ERROR_WORD_ALIGNMENT   0x60
-#define REP82_ERROR_MESSAGE_LENGTH   0x80
-#define REP82_ERROR_OPERAND_INVALID  0x82
-#define REP82_ERROR_OPERAND_SIZE     0x84
-#define REP82_ERROR_EVEN_MOD_IN_OPND 0x85
-#define REP82_ERROR_RESERVED_FIELD   0x88
-#define REP82_ERROR_TRANSPORT_FAIL   0x90
-#define REP82_ERROR_PACKET_TRUNCATED 0xA0
-#define REP82_ERROR_ZERO_BUFFER_LEN  0xB0
-
-#define REP88_ERROR_MODULE_FAILURE   0x10
-#define REP88_ERROR_MODULE_TIMEOUT   0x11
-#define REP88_ERROR_MODULE_NOTINIT   0x13
-#define REP88_ERROR_MODULE_NOTAVAIL  0x14
-#define REP88_ERROR_MODULE_DISABLED  0x15
-#define REP88_ERROR_MODULE_IN_DIAGN  0x17
-#define REP88_ERROR_FASTPATH_DISABLD 0x19
-#define REP88_ERROR_MESSAGE_TYPE     0x20
-#define REP88_ERROR_MESSAGE_MALFORMD 0x22
-#define REP88_ERROR_MESSAGE_LENGTH   0x23
-#define REP88_ERROR_RESERVED_FIELD   0x24
-#define REP88_ERROR_KEY_TYPE         0x34
-#define REP88_ERROR_INVALID_KEY      0x82
-#define REP88_ERROR_OPERAND          0x84
-#define REP88_ERROR_OPERAND_EVEN_MOD 0x85
-
-#define CALLER_HEADER 12
-
-static inline int
-testq(int q_nr, int *q_depth, int *dev_type, struct ap_status_word *stat)
-{
-	int ccode;
-
-	asm volatile
-#ifdef CONFIG_64BIT
-	("	llgfr	0,%4		\n"
-	 "	slgr	1,1		\n"
-	 "	lgr	2,1		\n"
-	 "0:	.long	0xb2af0000	\n"
-	 "1:	ipm	%0		\n"
-	 "	srl	%0,28		\n"
-	 "	iihh	%0,0		\n"
-	 "	iihl	%0,0		\n"
-	 "	lgr	%1,1		\n"
-	 "	lgr	%3,2		\n"
-	 "	srl	%3,24		\n"
-	 "	sll	2,24		\n"
-	 "	srl	2,24		\n"
-	 "	lgr	%2,2		\n"
-	 "2:				\n"
-	 ".section .fixup,\"ax\"	\n"
-	 "3:				\n"
-	 "	lhi	%0,%h5		\n"
-	 "	jg	2b		\n"
-	 ".previous			\n"
-	 ".section __ex_table,\"a\"	\n"
-	 "	.align	8		\n"
-	 "	.quad	0b,3b		\n"
-	 "	.quad	1b,3b		\n"
-	 ".previous"
-	 :"=d" (ccode),"=d" (*stat),"=d" (*q_depth), "=d" (*dev_type)
-	 :"d" (q_nr), "K" (DEV_TSQ_EXCEPTION)
-	 :"cc","0","1","2","memory");
-#else
-	("	lr	0,%4		\n"
-	 "	slr	1,1		\n"
-	 "	lr	2,1		\n"
-	 "0:	.long	0xb2af0000	\n"
-	 "1:	ipm	%0		\n"
-	 "	srl	%0,28		\n"
-	 "	lr	%1,1		\n"
-	 "	lr	%3,2		\n"
-	 "	srl	%3,24		\n"
-	 "	sll	2,24		\n"
-	 "	srl	2,24		\n"
-	 "	lr	%2,2		\n"
-	 "2:				\n"
-	 ".section .fixup,\"ax\"	\n"
-	 "3:				\n"
-	 "	lhi	%0,%h5		\n"
-	 "	bras	1,4f		\n"
-	 "	.long	2b		\n"
-	 "4:				\n"
-	 "	l	1,0(1)		\n"
-	 "	br	1		\n"
-	 ".previous			\n"
-	 ".section __ex_table,\"a\"	\n"
-	 "	.align	4		\n"
-	 "	.long	0b,3b		\n"
-	 "	.long	1b,3b		\n"
-	 ".previous"
-	 :"=d" (ccode),"=d" (*stat),"=d" (*q_depth), "=d" (*dev_type)
-	 :"d" (q_nr), "K" (DEV_TSQ_EXCEPTION)
-	 :"cc","0","1","2","memory");
-#endif
-	return ccode;
-}
-
-static inline int
-resetq(int q_nr, struct ap_status_word *stat_p)
-{
-	int ccode;
-
-	asm volatile
-#ifdef CONFIG_64BIT
-	("	llgfr	0,%2		\n"
-	 "	lghi	1,1		\n"
-	 "	sll	1,24		\n"
-	 "	or	0,1		\n"
-	 "	slgr	1,1		\n"
-	 "	lgr	2,1		\n"
-	 "0:	.long	0xb2af0000	\n"
-	 "1:	ipm	%0		\n"
-	 "	srl	%0,28		\n"
-	 "	iihh	%0,0		\n"
-	 "	iihl	%0,0		\n"
-	 "	lgr	%1,1		\n"
-	 "2:				\n"
-	 ".section .fixup,\"ax\"	\n"
-	 "3:				\n"
-	 "	lhi	%0,%h3		\n"
-	 "	jg	2b		\n"
-	 ".previous			\n"
-	 ".section __ex_table,\"a\"	\n"
-	 "	.align	8		\n"
-	 "	.quad	0b,3b		\n"
-	 "	.quad	1b,3b		\n"
-	 ".previous"
-	 :"=d" (ccode),"=d" (*stat_p)
-	 :"d" (q_nr), "K" (DEV_RSQ_EXCEPTION)
-	 :"cc","0","1","2","memory");
-#else
-	("	lr	0,%2		\n"
-	 "	lhi	1,1		\n"
-	 "	sll	1,24		\n"
-	 "	or	0,1		\n"
-	 "	slr	1,1		\n"
-	 "	lr	2,1		\n"
-	 "0:	.long	0xb2af0000	\n"
-	 "1:	ipm	%0		\n"
-	 "	srl	%0,28		\n"
-	 "	lr	%1,1		\n"
-	 "2:				\n"
-	 ".section .fixup,\"ax\"	\n"
-	 "3:				\n"
-	 "	lhi	%0,%h3		\n"
-	 "	bras	1,4f		\n"
-	 "	.long	2b		\n"
-	 "4:				\n"
-	 "	l	1,0(1)		\n"
-	 "	br	1		\n"
-	 ".previous			\n"
-	 ".section __ex_table,\"a\"	\n"
-	 "	.align	4		\n"
-	 "	.long	0b,3b		\n"
-	 "	.long	1b,3b		\n"
-	 ".previous"
-	 :"=d" (ccode),"=d" (*stat_p)
-	 :"d" (q_nr), "K" (DEV_RSQ_EXCEPTION)
-	 :"cc","0","1","2","memory");
-#endif
-	return ccode;
-}
-
-static inline int
-sen(int msg_len, unsigned char *msg_ext, struct ap_status_word *stat)
-{
-	int ccode;
-
-	asm volatile
-#ifdef CONFIG_64BIT
-	("	lgr	6,%3		\n"
-	 "	llgfr	7,%2		\n"
-	 "	llgt	0,0(6)		\n"
-	 "	lghi	1,64		\n"
-	 "	sll	1,24		\n"
-	 "	or	0,1		\n"
-	 "	la	6,4(6)		\n"
-	 "	llgt	2,0(6)		\n"
-	 "	llgt	3,4(6)		\n"
-	 "	la	6,8(6)		\n"
-	 "	slr	1,1		\n"
-	 "0:	.long	0xb2ad0026	\n"
-	 "1:	brc	2,0b		\n"
-	 "	ipm	%0		\n"
-	 "	srl	%0,28		\n"
-	 "	iihh	%0,0		\n"
-	 "	iihl	%0,0		\n"
-	 "	lgr	%1,1		\n"
-	 "2:				\n"
-	 ".section .fixup,\"ax\"	\n"
-	 "3:				\n"
-	 "	lhi	%0,%h4		\n"
-	 "	jg	2b		\n"
-	 ".previous			\n"
-	 ".section __ex_table,\"a\"	\n"
-	 "	.align	8		\n"
-	 "	.quad	0b,3b		\n"
-	 "	.quad	1b,3b		\n"
-	 ".previous"
-	 :"=d" (ccode),"=d" (*stat)
-	 :"d" (msg_len),"a" (msg_ext), "K" (DEV_SEN_EXCEPTION)
-	 :"cc","0","1","2","3","6","7","memory");
-#else
-	("	lr	6,%3		\n"
-	 "	lr	7,%2		\n"
-	 "	l	0,0(6)		\n"
-	 "	lhi	1,64		\n"
-	 "	sll	1,24		\n"
-	 "	or	0,1		\n"
-	 "	la	6,4(6)		\n"
-	 "	l	2,0(6)		\n"
-	 "	l	3,4(6)		\n"
-	 "	la	6,8(6)		\n"
-	 "	slr	1,1		\n"
-	 "0:	.long	0xb2ad0026	\n"
-	 "1:	brc	2,0b		\n"
-	 "	ipm	%0		\n"
-	 "	srl	%0,28		\n"
-	 "	lr	%1,1		\n"
-	 "2:				\n"
-	 ".section .fixup,\"ax\"	\n"
-	 "3:				\n"
-	 "	lhi	%0,%h4		\n"
-	 "	bras	1,4f		\n"
-	 "	.long	2b		\n"
-	 "4:				\n"
-	 "	l	1,0(1)		\n"
-	 "	br	1		\n"
-	 ".previous			\n"
-	 ".section __ex_table,\"a\"	\n"
-	 "	.align	4		\n"
-	 "	.long	0b,3b		\n"
-	 "	.long	1b,3b		\n"
-	 ".previous"
-	 :"=d" (ccode),"=d" (*stat)
-	 :"d" (msg_len),"a" (msg_ext), "K" (DEV_SEN_EXCEPTION)
-	 :"cc","0","1","2","3","6","7","memory");
-#endif
-	return ccode;
-}
-
-static inline int
-rec(int q_nr, int buff_l, unsigned char *rsp, unsigned char *id,
-    struct ap_status_word *st)
-{
-	int ccode;
-
-	asm volatile
-#ifdef CONFIG_64BIT
-	("	llgfr	0,%2		\n"
-	 "	lgr	3,%4		\n"
-	 "	lgr	6,%3		\n"
-	 "	llgfr	7,%5		\n"
-	 "	lghi	1,128		\n"
-	 "	sll	1,24		\n"
-	 "	or	0,1		\n"
-	 "	slgr	1,1		\n"
-	 "	lgr	2,1		\n"
-	 "	lgr	4,1		\n"
-	 "	lgr	5,1		\n"
-	 "0:	.long	0xb2ae0046	\n"
-	 "1:	brc	2,0b		\n"
-	 "	brc	4,0b		\n"
-	 "	ipm	%0		\n"
-	 "	srl	%0,28		\n"
-	 "	iihh	%0,0		\n"
-	 "	iihl	%0,0		\n"
-	 "	lgr	%1,1		\n"
-	 "	st	4,0(3)		\n"
-	 "	st	5,4(3)		\n"
-	 "2:				\n"
-	 ".section .fixup,\"ax\"	\n"
-	 "3:				\n"
-	 "	lhi   %0,%h6		\n"
-	 "	jg    2b		\n"
-	 ".previous			\n"
-	 ".section __ex_table,\"a\"	\n"
-	 "   .align	8		\n"
-	 "   .quad	0b,3b		\n"
-	 "   .quad	1b,3b		\n"
-	 ".previous"
-	 :"=d"(ccode),"=d"(*st)
-	 :"d" (q_nr), "d" (rsp), "d" (id), "d" (buff_l), "K" (DEV_REC_EXCEPTION)
-	 :"cc","0","1","2","3","4","5","6","7","memory");
-#else
-	("	lr	0,%2		\n"
-	 "	lr	3,%4		\n"
-	 "	lr	6,%3		\n"
-	 "	lr	7,%5		\n"
-	 "	lhi	1,128		\n"
-	 "	sll	1,24		\n"
-	 "	or	0,1		\n"
-	 "	slr	1,1		\n"
-	 "	lr	2,1		\n"
-	 "	lr	4,1		\n"
-	 "	lr	5,1		\n"
-	 "0:	.long	0xb2ae0046	\n"
-	 "1:	brc	2,0b		\n"
-	 "	brc	4,0b		\n"
-	 "	ipm	%0		\n"
-	 "	srl	%0,28		\n"
-	 "	lr	%1,1		\n"
-	 "	st	4,0(3)		\n"
-	 "	st	5,4(3)		\n"
-	 "2:				\n"
-	 ".section .fixup,\"ax\"	\n"
-	 "3:				\n"
-	 "	lhi   %0,%h6		\n"
-	 "	bras  1,4f		\n"
-	 "	.long 2b		\n"
-	 "4:				\n"
-	 "	l     1,0(1)		\n"
-	 "	br    1			\n"
-	 ".previous			\n"
-	 ".section __ex_table,\"a\"	\n"
-	 "   .align	4		\n"
-	 "   .long	0b,3b		\n"
-	 "   .long	1b,3b		\n"
-	 ".previous"
-	 :"=d"(ccode),"=d"(*st)
-	 :"d" (q_nr), "d" (rsp), "d" (id), "d" (buff_l), "K" (DEV_REC_EXCEPTION)
-	 :"cc","0","1","2","3","4","5","6","7","memory");
-#endif
-	return ccode;
-}
-
-static inline void
-itoLe2(int *i_p, unsigned char *lechars)
-{
-	*lechars       = *((unsigned char *) i_p + sizeof(int) - 1);
-	*(lechars + 1) = *((unsigned char *) i_p + sizeof(int) - 2);
-}
-
-static inline void
-le2toI(unsigned char *lechars, int *i_p)
-{
-	unsigned char *ic_p;
-	*i_p = 0;
-	ic_p = (unsigned char *) i_p;
-	*(ic_p + 2) = *(lechars + 1);
-	*(ic_p + 3) = *(lechars);
-}
-
-static inline int
-is_empty(unsigned char *ptr, int len)
-{
-	return !memcmp(ptr, (unsigned char *) &static_pvt_me_key+60, len);
-}
-
-enum hdstat
-query_online(int deviceNr, int cdx, int resetNr, int *q_depth, int *dev_type)
-{
-	int q_nr, i, t_depth, t_dev_type;
-	enum devstat ccode;
-	struct ap_status_word stat_word;
-	enum hdstat stat;
-	int break_out;
-
-	q_nr = (deviceNr << SKIP_BITL) + cdx;
-	stat = HD_BUSY;
-	ccode = testq(q_nr, &t_depth, &t_dev_type, &stat_word);
-	PDEBUG("ccode %d response_code %02X\n", ccode, stat_word.response_code);
-	break_out = 0;
-	for (i = 0; i < resetNr; i++) {
-		if (ccode > 3) {
-			PRINTKC("Exception testing device %d\n", i);
-			return HD_TSQ_EXCEPTION;
-		}
-		switch (ccode) {
-		case 0:
-			PDEBUG("t_dev_type %d\n", t_dev_type);
-			break_out = 1;
-			stat = HD_ONLINE;
-			*q_depth = t_depth + 1;
-			switch (t_dev_type) {
-			case PCICA_HW:
-				*dev_type = PCICA;
-				break;
-			case PCICC_HW:
-				*dev_type = PCICC;
-				break;
-			case PCIXCC_HW:
-				*dev_type = PCIXCC_UNK;
-				break;
-			case CEX2C_HW:
-				*dev_type = CEX2C;
-				break;
-			case CEX2A_HW:
-				*dev_type = CEX2A;
-				break;
-			default:
-				*dev_type = NILDEV;
-				break;
-			}
-			PDEBUG("available device %d: Q depth = %d, dev "
-			       "type = %d, stat = %02X%02X%02X%02X\n",
-			       deviceNr, *q_depth, *dev_type,
-			       stat_word.q_stat_flags,
-			       stat_word.response_code,
-			       stat_word.reserved[0],
-			       stat_word.reserved[1]);
-			break;
-		case 3:
-			switch (stat_word.response_code) {
-			case AP_RESPONSE_NORMAL:
-				stat = HD_ONLINE;
-				break_out = 1;
-				*q_depth = t_depth + 1;
-				*dev_type = t_dev_type;
-				PDEBUG("cc3, available device "
-				       "%d: Q depth = %d, dev "
-				       "type = %d, stat = "
-				       "%02X%02X%02X%02X\n",
-				       deviceNr, *q_depth,
-				       *dev_type,
-				       stat_word.q_stat_flags,
-				       stat_word.response_code,
-				       stat_word.reserved[0],
-				       stat_word.reserved[1]);
-				break;
-			case AP_RESPONSE_Q_NOT_AVAIL:
-				stat = HD_NOT_THERE;
-				break_out = 1;
-				break;
-			case AP_RESPONSE_RESET_IN_PROGRESS:
-				PDEBUG("device %d in reset\n",
-				       deviceNr);
-				break;
-			case AP_RESPONSE_DECONFIGURED:
-				stat = HD_DECONFIGURED;
-				break_out = 1;
-				break;
-			case AP_RESPONSE_CHECKSTOPPED:
-				stat = HD_CHECKSTOPPED;
-				break_out = 1;
-				break;
-			case AP_RESPONSE_BUSY:
-				PDEBUG("device %d busy\n",
-				       deviceNr);
-				break;
-			default:
-				break;
-			}
-			break;
-		default:
-			stat = HD_NOT_THERE;
-			break_out = 1;
-			break;
-		}
-		if (break_out)
-			break;
-
-		udelay(5);
-
-		ccode = testq(q_nr, &t_depth, &t_dev_type, &stat_word);
-	}
-	return stat;
-}
-
-enum devstat
-reset_device(int deviceNr, int cdx, int resetNr)
-{
-	int q_nr, ccode = 0, dummy_qdepth, dummy_devType, i;
-	struct ap_status_word stat_word;
-	enum devstat stat;
-	int break_out;
-
-	q_nr = (deviceNr << SKIP_BITL) + cdx;
-	stat = DEV_GONE;
-	ccode = resetq(q_nr, &stat_word);
-	if (ccode > 3)
-		return DEV_RSQ_EXCEPTION;
-
-	break_out = 0;
-	for (i = 0; i < resetNr; i++) {
-		switch (ccode) {
-		case 0:
-			stat = DEV_ONLINE;
-			if (stat_word.q_stat_flags & AP_Q_STATUS_EMPTY)
-				break_out = 1;
-			break;
-		case 3:
-			switch (stat_word.response_code) {
-			case AP_RESPONSE_NORMAL:
-				stat = DEV_ONLINE;
-				if (stat_word.q_stat_flags & AP_Q_STATUS_EMPTY)
-					break_out = 1;
-				break;
-			case AP_RESPONSE_Q_NOT_AVAIL:
-			case AP_RESPONSE_DECONFIGURED:
-			case AP_RESPONSE_CHECKSTOPPED:
-				stat = DEV_GONE;
-				break_out = 1;
-				break;
-			case AP_RESPONSE_RESET_IN_PROGRESS:
-			case AP_RESPONSE_BUSY:
-			default:
-				break;
-			}
-			break;
-		default:
-			stat = DEV_GONE;
-			break_out = 1;
-			break;
-		}
-		if (break_out == 1)
-			break;
-		udelay(5);
-
-		ccode = testq(q_nr, &dummy_qdepth, &dummy_devType, &stat_word);
-		if (ccode > 3) {
-			stat = DEV_TSQ_EXCEPTION;
-			break;
-		}
-	}
-	PDEBUG("Number of testq's needed for reset: %d\n", i);
-
-	if (i >= resetNr) {
-	  stat = DEV_GONE;
-	}
-
-	return stat;
-}
-
-#ifdef DEBUG_HYDRA_MSGS
-static inline void
-print_buffer(unsigned char *buffer, int bufflen)
-{
-	int i;
-	for (i = 0; i < bufflen; i += 16) {
-		PRINTK("%04X: %02X%02X%02X%02X %02X%02X%02X%02X "
-		       "%02X%02X%02X%02X %02X%02X%02X%02X\n", i,
-		       buffer[i+0], buffer[i+1], buffer[i+2], buffer[i+3],
-		       buffer[i+4], buffer[i+5], buffer[i+6], buffer[i+7],
-		       buffer[i+8], buffer[i+9], buffer[i+10], buffer[i+11],
-		       buffer[i+12], buffer[i+13], buffer[i+14], buffer[i+15]);
-	}
-}
-#endif
-
-enum devstat
-send_to_AP(int dev_nr, int cdx, int msg_len, unsigned char *msg_ext)
-{
-	struct ap_status_word stat_word;
-	enum devstat stat;
-	int ccode;
-	u32 *q_nr_p = (u32 *)msg_ext;
-
-	*q_nr_p = (dev_nr << SKIP_BITL) + cdx;
-	PDEBUG("msg_len passed to sen: %d\n", msg_len);
-	PDEBUG("q number passed to sen: %02x%02x%02x%02x\n",
-	       msg_ext[0], msg_ext[1], msg_ext[2], msg_ext[3]);
-	stat = DEV_GONE;
-
-#ifdef DEBUG_HYDRA_MSGS
-	PRINTK("Request header: %02X%02X%02X%02X %02X%02X%02X%02X "
-	       "%02X%02X%02X%02X\n",
-	       msg_ext[0], msg_ext[1], msg_ext[2], msg_ext[3],
-	       msg_ext[4], msg_ext[5], msg_ext[6], msg_ext[7],
-	       msg_ext[8], msg_ext[9], msg_ext[10], msg_ext[11]);
-	print_buffer(msg_ext+CALLER_HEADER, msg_len);
-#endif
-
-	ccode = sen(msg_len, msg_ext, &stat_word);
-	if (ccode > 3)
-		return DEV_SEN_EXCEPTION;
-
-	PDEBUG("nq cc: %u, st: %02x%02x%02x%02x\n",
-	       ccode, stat_word.q_stat_flags, stat_word.response_code,
-	       stat_word.reserved[0], stat_word.reserved[1]);
-	switch (ccode) {
-	case 0:
-		stat = DEV_ONLINE;
-		break;
-	case 1:
-		stat = DEV_GONE;
-		break;
-	case 3:
-		switch (stat_word.response_code) {
-		case AP_RESPONSE_NORMAL:
-			stat = DEV_ONLINE;
-			break;
-		case AP_RESPONSE_Q_FULL:
-			stat = DEV_QUEUE_FULL;
-			break;
-		default:
-			stat = DEV_GONE;
-			break;
-		}
-		break;
-	default:
-		stat = DEV_GONE;
-		break;
-	}
-
-	return stat;
-}
-
-enum devstat
-receive_from_AP(int dev_nr, int cdx, int resplen, unsigned char *resp,
-		unsigned char *psmid)
-{
-	int ccode;
-	struct ap_status_word stat_word;
-	enum devstat stat;
-
-	memset(resp, 0x00, 8);
-
-	ccode = rec((dev_nr << SKIP_BITL) + cdx, resplen, resp, psmid,
-		    &stat_word);
-	if (ccode > 3)
-		return DEV_REC_EXCEPTION;
-
-	PDEBUG("dq cc: %u, st: %02x%02x%02x%02x\n",
-	       ccode, stat_word.q_stat_flags, stat_word.response_code,
-	       stat_word.reserved[0], stat_word.reserved[1]);
-
-	stat = DEV_GONE;
-	switch (ccode) {
-	case 0:
-		stat = DEV_ONLINE;
-#ifdef DEBUG_HYDRA_MSGS
-		print_buffer(resp, resplen);
-#endif
-		break;
-	case 3:
-		switch (stat_word.response_code) {
-		case AP_RESPONSE_NORMAL:
-			stat = DEV_ONLINE;
-			break;
-		case AP_RESPONSE_NO_PENDING_REPLY:
-			if (stat_word.q_stat_flags & AP_Q_STATUS_EMPTY)
-				stat = DEV_EMPTY;
-			else
-				stat = DEV_NO_WORK;
-			break;
-		case AP_RESPONSE_INDEX_TOO_BIG:
-		case AP_RESPONSE_NO_FIRST_PART:
-		case AP_RESPONSE_MESSAGE_TOO_BIG:
-			stat = DEV_BAD_MESSAGE;
-			break;
-		default:
-			break;
-		}
-		break;
-	default:
-		break;
-	}
-
-	return stat;
-}
-
-static inline int
-pad_msg(unsigned char *buffer, int  totalLength, int msgLength)
-{
-	int pad_len;
-
-	for (pad_len = 0; pad_len < (totalLength - msgLength); pad_len++)
-		if (buffer[pad_len] != 0x00)
-			break;
-	pad_len -= 3;
-	if (pad_len < 8)
-		return SEN_PAD_ERROR;
-
-	buffer[0] = 0x00;
-	buffer[1] = 0x02;
-
-	memcpy(buffer+2, static_pad, pad_len);
-
-	buffer[pad_len + 2] = 0x00;
-
-	return 0;
-}
-
-static inline int
-is_common_public_key(unsigned char *key, int len)
-{
-	int i;
-
-	for (i = 0; i < len; i++)
-		if (key[i])
-			break;
-	key += i;
-	len -= i;
-	if (((len == 1) && (key[0] == 3)) ||
-	    ((len == 3) && (key[0] == 1) && (key[1] == 0) && (key[2] == 1)))
-		return 1;
-
-	return 0;
-}
-
-static int
-ICAMEX_msg_to_type4MEX_msg(struct ica_rsa_modexpo *icaMex_p, int *z90cMsg_l_p,
-			   union type4_msg *z90cMsg_p)
-{
-	int mod_len, msg_size, mod_tgt_len, exp_tgt_len, inp_tgt_len;
-	unsigned char *mod_tgt, *exp_tgt, *inp_tgt;
-	union type4_msg *tmp_type4_msg;
-
-	mod_len = icaMex_p->inputdatalength;
-
-	msg_size = ((mod_len <= 128) ? TYPE4_SME_LEN : TYPE4_LME_LEN) +
-		    CALLER_HEADER;
-
-	memset(z90cMsg_p, 0, msg_size);
-
-	tmp_type4_msg = (union type4_msg *)
-		((unsigned char *) z90cMsg_p + CALLER_HEADER);
-
-	tmp_type4_msg->sme.header.msg_type_code = TYPE4_TYPE_CODE;
-	tmp_type4_msg->sme.header.request_code = TYPE4_REQU_CODE;
-
-	if (mod_len <= 128) {
-		tmp_type4_msg->sme.header.msg_fmt = TYPE4_SME_FMT;
-		tmp_type4_msg->sme.header.msg_len = TYPE4_SME_LEN;
-		mod_tgt = tmp_type4_msg->sme.modulus;
-		mod_tgt_len = sizeof(tmp_type4_msg->sme.modulus);
-		exp_tgt = tmp_type4_msg->sme.exponent;
-		exp_tgt_len = sizeof(tmp_type4_msg->sme.exponent);
-		inp_tgt = tmp_type4_msg->sme.message;
-		inp_tgt_len = sizeof(tmp_type4_msg->sme.message);
-	} else {
-		tmp_type4_msg->lme.header.msg_fmt = TYPE4_LME_FMT;
-		tmp_type4_msg->lme.header.msg_len = TYPE4_LME_LEN;
-		mod_tgt = tmp_type4_msg->lme.modulus;
-		mod_tgt_len = sizeof(tmp_type4_msg->lme.modulus);
-		exp_tgt = tmp_type4_msg->lme.exponent;
-		exp_tgt_len = sizeof(tmp_type4_msg->lme.exponent);
-		inp_tgt = tmp_type4_msg->lme.message;
-		inp_tgt_len = sizeof(tmp_type4_msg->lme.message);
-	}
-
-	mod_tgt += (mod_tgt_len - mod_len);
-	if (copy_from_user(mod_tgt, icaMex_p->n_modulus, mod_len))
-		return SEN_RELEASED;
-	if (is_empty(mod_tgt, mod_len))
-		return SEN_USER_ERROR;
-	exp_tgt += (exp_tgt_len - mod_len);
-	if (copy_from_user(exp_tgt, icaMex_p->b_key, mod_len))
-		return SEN_RELEASED;
-	if (is_empty(exp_tgt, mod_len))
-		return SEN_USER_ERROR;
-	inp_tgt += (inp_tgt_len - mod_len);
-	if (copy_from_user(inp_tgt, icaMex_p->inputdata, mod_len))
-		return SEN_RELEASED;
-	if (is_empty(inp_tgt, mod_len))
-		return SEN_USER_ERROR;
-
-	*z90cMsg_l_p = msg_size - CALLER_HEADER;
-
-	return 0;
-}
-
-static int
-ICACRT_msg_to_type4CRT_msg(struct ica_rsa_modexpo_crt *icaMsg_p,
-			   int *z90cMsg_l_p, union type4_msg *z90cMsg_p)
-{
-	int mod_len, short_len, long_len, tmp_size, p_tgt_len, q_tgt_len,
-	    dp_tgt_len, dq_tgt_len, u_tgt_len, inp_tgt_len;
-	unsigned char *p_tgt, *q_tgt, *dp_tgt, *dq_tgt, *u_tgt, *inp_tgt;
-	union type4_msg *tmp_type4_msg;
-
-	mod_len = icaMsg_p->inputdatalength;
-	short_len = mod_len / 2;
-	long_len = mod_len / 2 + 8;
-
-	tmp_size = ((mod_len <= 128) ? TYPE4_SCR_LEN : TYPE4_LCR_LEN) +
-		    CALLER_HEADER;
-
-	memset(z90cMsg_p, 0, tmp_size);
-
-	tmp_type4_msg = (union type4_msg *)
-		((unsigned char *) z90cMsg_p + CALLER_HEADER);
-
-	tmp_type4_msg->scr.header.msg_type_code = TYPE4_TYPE_CODE;
-	tmp_type4_msg->scr.header.request_code = TYPE4_REQU_CODE;
-	if (mod_len <= 128) {
-		tmp_type4_msg->scr.header.msg_fmt = TYPE4_SCR_FMT;
-		tmp_type4_msg->scr.header.msg_len = TYPE4_SCR_LEN;
-		p_tgt = tmp_type4_msg->scr.p;
-		p_tgt_len = sizeof(tmp_type4_msg->scr.p);
-		q_tgt = tmp_type4_msg->scr.q;
-		q_tgt_len = sizeof(tmp_type4_msg->scr.q);
-		dp_tgt = tmp_type4_msg->scr.dp;
-		dp_tgt_len = sizeof(tmp_type4_msg->scr.dp);
-		dq_tgt = tmp_type4_msg->scr.dq;
-		dq_tgt_len = sizeof(tmp_type4_msg->scr.dq);
-		u_tgt = tmp_type4_msg->scr.u;
-		u_tgt_len = sizeof(tmp_type4_msg->scr.u);
-		inp_tgt = tmp_type4_msg->scr.message;
-		inp_tgt_len = sizeof(tmp_type4_msg->scr.message);
-	} else {
-		tmp_type4_msg->lcr.header.msg_fmt = TYPE4_LCR_FMT;
-		tmp_type4_msg->lcr.header.msg_len = TYPE4_LCR_LEN;
-		p_tgt = tmp_type4_msg->lcr.p;
-		p_tgt_len = sizeof(tmp_type4_msg->lcr.p);
-		q_tgt = tmp_type4_msg->lcr.q;
-		q_tgt_len = sizeof(tmp_type4_msg->lcr.q);
-		dp_tgt = tmp_type4_msg->lcr.dp;
-		dp_tgt_len = sizeof(tmp_type4_msg->lcr.dp);
-		dq_tgt = tmp_type4_msg->lcr.dq;
-		dq_tgt_len = sizeof(tmp_type4_msg->lcr.dq);
-		u_tgt = tmp_type4_msg->lcr.u;
-		u_tgt_len = sizeof(tmp_type4_msg->lcr.u);
-		inp_tgt = tmp_type4_msg->lcr.message;
-		inp_tgt_len = sizeof(tmp_type4_msg->lcr.message);
-	}
-
-	p_tgt += (p_tgt_len - long_len);
-	if (copy_from_user(p_tgt, icaMsg_p->np_prime, long_len))
-		return SEN_RELEASED;
-	if (is_empty(p_tgt, long_len))
-		return SEN_USER_ERROR;
-	q_tgt += (q_tgt_len - short_len);
-	if (copy_from_user(q_tgt, icaMsg_p->nq_prime, short_len))
-		return SEN_RELEASED;
-	if (is_empty(q_tgt, short_len))
-		return SEN_USER_ERROR;
-	dp_tgt += (dp_tgt_len - long_len);
-	if (copy_from_user(dp_tgt, icaMsg_p->bp_key, long_len))
-		return SEN_RELEASED;
-	if (is_empty(dp_tgt, long_len))
-		return SEN_USER_ERROR;
-	dq_tgt += (dq_tgt_len - short_len);
-	if (copy_from_user(dq_tgt, icaMsg_p->bq_key, short_len))
-		return SEN_RELEASED;
-	if (is_empty(dq_tgt, short_len))
-		return SEN_USER_ERROR;
-	u_tgt += (u_tgt_len - long_len);
-	if (copy_from_user(u_tgt, icaMsg_p->u_mult_inv, long_len))
-		return SEN_RELEASED;
-	if (is_empty(u_tgt, long_len))
-		return SEN_USER_ERROR;
-	inp_tgt += (inp_tgt_len - mod_len);
-	if (copy_from_user(inp_tgt, icaMsg_p->inputdata, mod_len))
-		return SEN_RELEASED;
-	if (is_empty(inp_tgt, mod_len))
-		return SEN_USER_ERROR;
-
-	*z90cMsg_l_p = tmp_size - CALLER_HEADER;
-
-	return 0;
-}
-
-static int
-ICAMEX_msg_to_type6MEX_de_msg(struct ica_rsa_modexpo *icaMsg_p, int cdx,
-			      int *z90cMsg_l_p, struct type6_msg *z90cMsg_p)
-{
-	int mod_len, vud_len, tmp_size, total_CPRB_len, parmBlock_l;
-	unsigned char *temp;
-	struct type6_hdr *tp6Hdr_p;
-	struct CPRB *cprb_p;
-	struct cca_private_ext_ME *key_p;
-	static int deprecated_msg_count = 0;
-
-	mod_len = icaMsg_p->inputdatalength;
-	tmp_size = FIXED_TYPE6_ME_LEN + mod_len;
-	total_CPRB_len = tmp_size - sizeof(struct type6_hdr);
-	parmBlock_l = total_CPRB_len - sizeof(struct CPRB);
-	tmp_size = 4*((tmp_size + 3)/4) + CALLER_HEADER;
-
-	memset(z90cMsg_p, 0, tmp_size);
-
-	temp = (unsigned char *)z90cMsg_p + CALLER_HEADER;
-	memcpy(temp, &static_type6_hdr, sizeof(struct type6_hdr));
-	tp6Hdr_p = (struct type6_hdr *)temp;
-	tp6Hdr_p->ToCardLen1 = 4*((total_CPRB_len+3)/4);
-	tp6Hdr_p->FromCardLen1 = RESPONSE_CPRB_SIZE;
-
-	temp += sizeof(struct type6_hdr);
-	memcpy(temp, &static_cprb, sizeof(struct CPRB));
-	cprb_p = (struct CPRB *) temp;
-	cprb_p->usage_domain[0]= (unsigned char)cdx;
-	itoLe2(&parmBlock_l, cprb_p->req_parml);
-	itoLe2((int *)&(tp6Hdr_p->FromCardLen1), cprb_p->rpl_parml);
-
-	temp += sizeof(struct CPRB);
-	memcpy(temp, &static_pkd_function_and_rules,
-	       sizeof(struct function_and_rules_block));
-
-	temp += sizeof(struct function_and_rules_block);
-	vud_len = 2 + icaMsg_p->inputdatalength;
-	itoLe2(&vud_len, temp);
-
-	temp += 2;
-	if (copy_from_user(temp, icaMsg_p->inputdata, mod_len))
-		return SEN_RELEASED;
-	if (is_empty(temp, mod_len))
-		return SEN_USER_ERROR;
-
-	temp += mod_len;
-	memcpy(temp, &static_T6_keyBlock_hdr, sizeof(struct T6_keyBlock_hdr));
-
-	temp += sizeof(struct T6_keyBlock_hdr);
-	memcpy(temp, &static_pvt_me_key, sizeof(struct cca_private_ext_ME));
-	key_p = (struct cca_private_ext_ME *)temp;
-	temp = key_p->pvtMESec.exponent + sizeof(key_p->pvtMESec.exponent)
-	       - mod_len;
-	if (copy_from_user(temp, icaMsg_p->b_key, mod_len))
-		return SEN_RELEASED;
-	if (is_empty(temp, mod_len))
-		return SEN_USER_ERROR;
-
-	if (is_common_public_key(temp, mod_len)) {
-		if (deprecated_msg_count < 20) {
-			PRINTK("Common public key used for modex decrypt\n");
-			deprecated_msg_count++;
-			if (deprecated_msg_count == 20)
-				PRINTK("No longer issuing messages about common"
-				       " public key for modex decrypt.\n");
-		}
-		return SEN_NOT_AVAIL;
-	}
-
-	temp = key_p->pvtMESec.modulus + sizeof(key_p->pvtMESec.modulus)
-	       - mod_len;
-	if (copy_from_user(temp, icaMsg_p->n_modulus, mod_len))
-		return SEN_RELEASED;
-	if (is_empty(temp, mod_len))
-		return SEN_USER_ERROR;
-
-	key_p->pubMESec.modulus_bit_len = 8 * mod_len;
-
-	*z90cMsg_l_p = tmp_size - CALLER_HEADER;
-
-	return 0;
-}
-
-static int
-ICAMEX_msg_to_type6MEX_en_msg(struct ica_rsa_modexpo *icaMsg_p, int cdx,
-			      int *z90cMsg_l_p, struct type6_msg *z90cMsg_p)
-{
-	int mod_len, vud_len, exp_len, key_len;
-	int pad_len, tmp_size, total_CPRB_len, parmBlock_l, i;
-	unsigned char *temp_exp, *exp_p, *temp;
-	struct type6_hdr *tp6Hdr_p;
-	struct CPRB *cprb_p;
-	struct cca_public_key *key_p;
-	struct T6_keyBlock_hdr *keyb_p;
-
-	temp_exp = kmalloc(256, GFP_KERNEL);
-	if (!temp_exp)
-		return EGETBUFF;
-	mod_len = icaMsg_p->inputdatalength;
-	if (copy_from_user(temp_exp, icaMsg_p->b_key, mod_len)) {
-		kfree(temp_exp);
-		return SEN_RELEASED;
-	}
-	if (is_empty(temp_exp, mod_len)) {
-		kfree(temp_exp);
-		return SEN_USER_ERROR;
-	}
-
-	exp_p = temp_exp;
-	for (i = 0; i < mod_len; i++)
-		if (exp_p[i])
-			break;
-	if (i >= mod_len) {
-		kfree(temp_exp);
-		return SEN_USER_ERROR;
-	}
-
-	exp_len = mod_len - i;
-	exp_p += i;
-
-	PDEBUG("exp_len after computation: %08x\n", exp_len);
-	tmp_size = FIXED_TYPE6_ME_EN_LEN + 2 * mod_len + exp_len;
-	total_CPRB_len = tmp_size - sizeof(struct type6_hdr);
-	parmBlock_l = total_CPRB_len - sizeof(struct CPRB);
-	tmp_size = 4*((tmp_size + 3)/4) + CALLER_HEADER;
-
-	vud_len = 2 + mod_len;
-	memset(z90cMsg_p, 0, tmp_size);
-
-	temp = (unsigned char *)z90cMsg_p + CALLER_HEADER;
-	memcpy(temp, &static_type6_hdr, sizeof(struct type6_hdr));
-	tp6Hdr_p = (struct type6_hdr *)temp;
-	tp6Hdr_p->ToCardLen1 = 4*((total_CPRB_len+3)/4);
-	tp6Hdr_p->FromCardLen1 = RESPONSE_CPRB_SIZE;
-	memcpy(tp6Hdr_p->function_code, static_PKE_function_code,
-	       sizeof(static_PKE_function_code));
-	temp += sizeof(struct type6_hdr);
-	memcpy(temp, &static_cprb, sizeof(struct CPRB));
-	cprb_p = (struct CPRB *) temp;
-	cprb_p->usage_domain[0]= (unsigned char)cdx;
-	itoLe2((int *)&(tp6Hdr_p->FromCardLen1), cprb_p->rpl_parml);
-	temp += sizeof(struct CPRB);
-	memcpy(temp, &static_pke_function_and_rules,
-		 sizeof(struct function_and_rules_block));
-	temp += sizeof(struct function_and_rules_block);
-	temp += 2;
-	if (copy_from_user(temp, icaMsg_p->inputdata, mod_len)) {
-		kfree(temp_exp);
-		return SEN_RELEASED;
-	}
-	if (is_empty(temp, mod_len)) {
-		kfree(temp_exp);
-		return SEN_USER_ERROR;
-	}
-	if ((temp[0] != 0x00) || (temp[1] != 0x02)) {
-		kfree(temp_exp);
-		return SEN_NOT_AVAIL;
-	}
-	for (i = 2; i < mod_len; i++)
-		if (temp[i] == 0x00)
-			break;
-	if ((i < 9) || (i > (mod_len - 2))) {
-		kfree(temp_exp);
-		return SEN_NOT_AVAIL;
-	}
-	pad_len = i + 1;
-	vud_len = mod_len - pad_len;
-	memmove(temp, temp+pad_len, vud_len);
-	temp -= 2;
-	vud_len += 2;
-	itoLe2(&vud_len, temp);
-	temp += (vud_len);
-	keyb_p = (struct T6_keyBlock_hdr *)temp;
-	temp += sizeof(struct T6_keyBlock_hdr);
-	memcpy(temp, &static_public_key, sizeof(static_public_key));
-	key_p = (struct cca_public_key *)temp;
-	temp = key_p->pubSec.exponent;
-	memcpy(temp, exp_p, exp_len);
-	kfree(temp_exp);
-	temp += exp_len;
-	if (copy_from_user(temp, icaMsg_p->n_modulus, mod_len))
-		return SEN_RELEASED;
-	if (is_empty(temp, mod_len))
-		return SEN_USER_ERROR;
-	key_p->pubSec.modulus_bit_len = 8 * mod_len;
-	key_p->pubSec.modulus_byte_len = mod_len;
-	key_p->pubSec.exponent_len = exp_len;
-	key_p->pubSec.section_length = CALLER_HEADER + mod_len + exp_len;
-	key_len = key_p->pubSec.section_length + sizeof(struct cca_token_hdr);
-	key_p->pubHdr.token_length = key_len;
-	key_len += 4;
-	itoLe2(&key_len, keyb_p->ulen);
-	key_len += 2;
-	itoLe2(&key_len, keyb_p->blen);
-	parmBlock_l -= pad_len;
-	itoLe2(&parmBlock_l, cprb_p->req_parml);
-	*z90cMsg_l_p = tmp_size - CALLER_HEADER;
-
-	return 0;
-}
-
-static int
-ICACRT_msg_to_type6CRT_msg(struct ica_rsa_modexpo_crt *icaMsg_p, int cdx,
-			   int *z90cMsg_l_p, struct type6_msg *z90cMsg_p)
-{
-	int mod_len, vud_len, tmp_size, total_CPRB_len, parmBlock_l, short_len;
-	int long_len, pad_len, keyPartsLen, tmp_l;
-	unsigned char *tgt_p, *temp;
-	struct type6_hdr *tp6Hdr_p;
-	struct CPRB *cprb_p;
-	struct cca_token_hdr *keyHdr_p;
-	struct cca_pvt_ext_CRT_sec *pvtSec_p;
-	struct cca_public_sec *pubSec_p;
-
-	mod_len = icaMsg_p->inputdatalength;
-	short_len = mod_len / 2;
-	long_len = 8 + short_len;
-	keyPartsLen = 3 * long_len + 2 * short_len;
-	pad_len = (8 - (keyPartsLen % 8)) % 8;
-	keyPartsLen += pad_len + mod_len;
-	tmp_size = FIXED_TYPE6_CR_LEN + keyPartsLen + mod_len;
-	total_CPRB_len = tmp_size -  sizeof(struct type6_hdr);
-	parmBlock_l = total_CPRB_len - sizeof(struct CPRB);
-	vud_len = 2 + mod_len;
-	tmp_size = 4*((tmp_size + 3)/4) + CALLER_HEADER;
-
-	memset(z90cMsg_p, 0, tmp_size);
-	tgt_p = (unsigned char *)z90cMsg_p + CALLER_HEADER;
-	memcpy(tgt_p, &static_type6_hdr, sizeof(struct type6_hdr));
-	tp6Hdr_p = (struct type6_hdr *)tgt_p;
-	tp6Hdr_p->ToCardLen1 = 4*((total_CPRB_len+3)/4);
-	tp6Hdr_p->FromCardLen1 = RESPONSE_CPRB_SIZE;
-	tgt_p += sizeof(struct type6_hdr);
-	cprb_p = (struct CPRB *) tgt_p;
-	memcpy(tgt_p, &static_cprb, sizeof(struct CPRB));
-	cprb_p->usage_domain[0]= *((unsigned char *)(&(cdx))+3);
-	itoLe2(&parmBlock_l, cprb_p->req_parml);
-	memcpy(cprb_p->rpl_parml, cprb_p->req_parml,
-	       sizeof(cprb_p->req_parml));
-	tgt_p += sizeof(struct CPRB);
-	memcpy(tgt_p, &static_pkd_function_and_rules,
-	       sizeof(struct function_and_rules_block));
-	tgt_p += sizeof(struct function_and_rules_block);
-	itoLe2(&vud_len, tgt_p);
-	tgt_p += 2;
-	if (copy_from_user(tgt_p, icaMsg_p->inputdata, mod_len))
-		return SEN_RELEASED;
-	if (is_empty(tgt_p, mod_len))
-		return SEN_USER_ERROR;
-	tgt_p += mod_len;
-	tmp_l = sizeof(struct T6_keyBlock_hdr) + sizeof(struct cca_token_hdr) +
-		sizeof(struct cca_pvt_ext_CRT_sec) + 0x0F + keyPartsLen;
-	itoLe2(&tmp_l, tgt_p);
-	temp = tgt_p + 2;
-	tmp_l -= 2;
-	itoLe2(&tmp_l, temp);
-	tgt_p += sizeof(struct T6_keyBlock_hdr);
-	keyHdr_p = (struct cca_token_hdr *)tgt_p;
-	keyHdr_p->token_identifier = CCA_TKN_HDR_ID_EXT;
-	tmp_l -= 4;
-	keyHdr_p->token_length = tmp_l;
-	tgt_p += sizeof(struct cca_token_hdr);
-	pvtSec_p = (struct cca_pvt_ext_CRT_sec *)tgt_p;
-	pvtSec_p->section_identifier = CCA_PVT_EXT_CRT_SEC_ID_PVT;
-	pvtSec_p->section_length =
-		sizeof(struct cca_pvt_ext_CRT_sec) + keyPartsLen;
-	pvtSec_p->key_format = CCA_PVT_EXT_CRT_SEC_FMT_CL;
-	pvtSec_p->key_use_flags[0] = CCA_PVT_USAGE_ALL;
-	pvtSec_p->p_len = long_len;
-	pvtSec_p->q_len = short_len;
-	pvtSec_p->dp_len = long_len;
-	pvtSec_p->dq_len = short_len;
-	pvtSec_p->u_len = long_len;
-	pvtSec_p->mod_len = mod_len;
-	pvtSec_p->pad_len = pad_len;
-	tgt_p += sizeof(struct cca_pvt_ext_CRT_sec);
-	if (copy_from_user(tgt_p, icaMsg_p->np_prime, long_len))
-		return SEN_RELEASED;
-	if (is_empty(tgt_p, long_len))
-		return SEN_USER_ERROR;
-	tgt_p += long_len;
-	if (copy_from_user(tgt_p, icaMsg_p->nq_prime, short_len))
-		return SEN_RELEASED;
-	if (is_empty(tgt_p, short_len))
-		return SEN_USER_ERROR;
-	tgt_p += short_len;
-	if (copy_from_user(tgt_p, icaMsg_p->bp_key, long_len))
-		return SEN_RELEASED;
-	if (is_empty(tgt_p, long_len))
-		return SEN_USER_ERROR;
-	tgt_p += long_len;
-	if (copy_from_user(tgt_p, icaMsg_p->bq_key, short_len))
-		return SEN_RELEASED;
-	if (is_empty(tgt_p, short_len))
-		return SEN_USER_ERROR;
-	tgt_p += short_len;
-	if (copy_from_user(tgt_p, icaMsg_p->u_mult_inv, long_len))
-		return SEN_RELEASED;
-	if (is_empty(tgt_p, long_len))
-		return SEN_USER_ERROR;
-	tgt_p += long_len;
-	tgt_p += pad_len;
-	memset(tgt_p, 0xFF, mod_len);
-	tgt_p += mod_len;
-	memcpy(tgt_p, &static_cca_pub_sec, sizeof(struct cca_public_sec));
-	pubSec_p = (struct cca_public_sec *) tgt_p;
-	pubSec_p->modulus_bit_len = 8 * mod_len;
-	*z90cMsg_l_p = tmp_size - CALLER_HEADER;
-
-	return 0;
-}
-
-static int
-ICAMEX_msg_to_type6MEX_msgX(struct ica_rsa_modexpo *icaMsg_p, int cdx,
-			    int *z90cMsg_l_p, struct type6_msg *z90cMsg_p,
-			    int dev_type)
-{
-	int mod_len, exp_len, vud_len, tmp_size, total_CPRB_len, parmBlock_l;
-	int key_len, i;
-	unsigned char *temp_exp, *tgt_p, *temp, *exp_p;
-	struct type6_hdr *tp6Hdr_p;
-	struct CPRBX *cprbx_p;
-	struct cca_public_key *key_p;
-	struct T6_keyBlock_hdrX *keyb_p;
-
-	temp_exp = kmalloc(256, GFP_KERNEL);
-	if (!temp_exp)
-		return EGETBUFF;
-	mod_len = icaMsg_p->inputdatalength;
-	if (copy_from_user(temp_exp, icaMsg_p->b_key, mod_len)) {
-		kfree(temp_exp);
-		return SEN_RELEASED;
-	}
-	if (is_empty(temp_exp, mod_len)) {
-		kfree(temp_exp);
-		return SEN_USER_ERROR;
-	}
-	exp_p = temp_exp;
-	for (i = 0; i < mod_len; i++)
-		if (exp_p[i])
-			break;
-	if (i >= mod_len) {
-		kfree(temp_exp);
-		return SEN_USER_ERROR;
-	}
-	exp_len = mod_len - i;
-	exp_p += i;
-	PDEBUG("exp_len after computation: %08x\n", exp_len);
-	tmp_size = FIXED_TYPE6_ME_EN_LENX + 2 * mod_len + exp_len;
-	total_CPRB_len = tmp_size - sizeof(struct type6_hdr);
-	parmBlock_l = total_CPRB_len - sizeof(struct CPRBX);
-	tmp_size = tmp_size + CALLER_HEADER;
-	vud_len = 2 + mod_len;
-	memset(z90cMsg_p, 0, tmp_size);
-	tgt_p = (unsigned char *)z90cMsg_p + CALLER_HEADER;
-	memcpy(tgt_p, &static_type6_hdrX, sizeof(struct type6_hdr));
-	tp6Hdr_p = (struct type6_hdr *)tgt_p;
-	tp6Hdr_p->ToCardLen1 = total_CPRB_len;
-	tp6Hdr_p->FromCardLen1 = RESPONSE_CPRBX_SIZE;
-	memcpy(tp6Hdr_p->function_code, static_PKE_function_code,
-	       sizeof(static_PKE_function_code));
-	tgt_p += sizeof(struct type6_hdr);
-	memcpy(tgt_p, &static_cprbx, sizeof(struct CPRBX));
-	cprbx_p = (struct CPRBX *) tgt_p;
-	cprbx_p->domain = (unsigned short)cdx;
-	cprbx_p->rpl_msgbl = RESPONSE_CPRBX_SIZE;
-	tgt_p += sizeof(struct CPRBX);
-	if (dev_type == PCIXCC_MCL2)
-		memcpy(tgt_p, &static_pke_function_and_rulesX_MCL2,
-		       sizeof(struct function_and_rules_block));
-	else
-		memcpy(tgt_p, &static_pke_function_and_rulesX,
-		       sizeof(struct function_and_rules_block));
-	tgt_p += sizeof(struct function_and_rules_block);
-
-	tgt_p += 2;
-	if (copy_from_user(tgt_p, icaMsg_p->inputdata, mod_len)) {
-		kfree(temp_exp);
-		return SEN_RELEASED;
-	}
-	if (is_empty(tgt_p, mod_len)) {
-		kfree(temp_exp);
-		return SEN_USER_ERROR;
-	}
-	tgt_p -= 2;
-	*((short *)tgt_p) = (short) vud_len;
-	tgt_p += vud_len;
-	keyb_p = (struct T6_keyBlock_hdrX *)tgt_p;
-	tgt_p += sizeof(struct T6_keyBlock_hdrX);
-	memcpy(tgt_p, &static_public_key, sizeof(static_public_key));
-	key_p = (struct cca_public_key *)tgt_p;
-	temp = key_p->pubSec.exponent;
-	memcpy(temp, exp_p, exp_len);
-	kfree(temp_exp);
-	temp += exp_len;
-	if (copy_from_user(temp, icaMsg_p->n_modulus, mod_len))
-		return SEN_RELEASED;
-	if (is_empty(temp, mod_len))
-		return SEN_USER_ERROR;
-	key_p->pubSec.modulus_bit_len = 8 * mod_len;
-	key_p->pubSec.modulus_byte_len = mod_len;
-	key_p->pubSec.exponent_len = exp_len;
-	key_p->pubSec.section_length = CALLER_HEADER + mod_len + exp_len;
-	key_len = key_p->pubSec.section_length + sizeof(struct cca_token_hdr);
-	key_p->pubHdr.token_length = key_len;
-	key_len += 4;
-	keyb_p->ulen = (unsigned short)key_len;
-	key_len += 2;
-	keyb_p->blen = (unsigned short)key_len;
-	cprbx_p->req_parml = parmBlock_l;
-	*z90cMsg_l_p = tmp_size - CALLER_HEADER;
-
-	return 0;
-}
-
-static int
-ICACRT_msg_to_type6CRT_msgX(struct ica_rsa_modexpo_crt *icaMsg_p, int cdx,
-			    int *z90cMsg_l_p, struct type6_msg *z90cMsg_p,
-			    int dev_type)
-{
-	int mod_len, vud_len, tmp_size, total_CPRB_len, parmBlock_l, short_len;
-	int long_len, pad_len, keyPartsLen, tmp_l;
-	unsigned char *tgt_p, *temp;
-	struct type6_hdr *tp6Hdr_p;
-	struct CPRBX *cprbx_p;
-	struct cca_token_hdr *keyHdr_p;
-	struct cca_pvt_ext_CRT_sec *pvtSec_p;
-	struct cca_public_sec *pubSec_p;
-
-	mod_len = icaMsg_p->inputdatalength;
-	short_len = mod_len / 2;
-	long_len = 8 + short_len;
-	keyPartsLen = 3 * long_len + 2 * short_len;
-	pad_len = (8 - (keyPartsLen % 8)) % 8;
-	keyPartsLen += pad_len + mod_len;
-	tmp_size = FIXED_TYPE6_CR_LENX + keyPartsLen + mod_len;
-	total_CPRB_len = tmp_size -  sizeof(struct type6_hdr);
-	parmBlock_l = total_CPRB_len - sizeof(struct CPRBX);
-	vud_len = 2 + mod_len;
-	tmp_size = tmp_size + CALLER_HEADER;
-	memset(z90cMsg_p, 0, tmp_size);
-	tgt_p = (unsigned char *)z90cMsg_p + CALLER_HEADER;
-	memcpy(tgt_p, &static_type6_hdrX, sizeof(struct type6_hdr));
-	tp6Hdr_p = (struct type6_hdr *)tgt_p;
-	tp6Hdr_p->ToCardLen1 = total_CPRB_len;
-	tp6Hdr_p->FromCardLen1 = RESPONSE_CPRBX_SIZE;
-	tgt_p += sizeof(struct type6_hdr);
-	cprbx_p = (struct CPRBX *) tgt_p;
-	memcpy(tgt_p, &static_cprbx, sizeof(struct CPRBX));
-	cprbx_p->domain = (unsigned short)cdx;
-	cprbx_p->req_parml = parmBlock_l;
-	cprbx_p->rpl_msgbl = parmBlock_l;
-	tgt_p += sizeof(struct CPRBX);
-	if (dev_type == PCIXCC_MCL2)
-		memcpy(tgt_p, &static_pkd_function_and_rulesX_MCL2,
-		       sizeof(struct function_and_rules_block));
-	else
-		memcpy(tgt_p, &static_pkd_function_and_rulesX,
-		       sizeof(struct function_and_rules_block));
-	tgt_p += sizeof(struct function_and_rules_block);
-	*((short *)tgt_p) = (short) vud_len;
-	tgt_p += 2;
-	if (copy_from_user(tgt_p, icaMsg_p->inputdata, mod_len))
-		return SEN_RELEASED;
-	if (is_empty(tgt_p, mod_len))
-		return SEN_USER_ERROR;
-	tgt_p += mod_len;
-	tmp_l = sizeof(struct T6_keyBlock_hdr) + sizeof(struct cca_token_hdr) +
-		sizeof(struct cca_pvt_ext_CRT_sec) + 0x0F + keyPartsLen;
-	*((short *)tgt_p) = (short) tmp_l;
-	temp = tgt_p + 2;
-	tmp_l -= 2;
-	*((short *)temp) = (short) tmp_l;
-	tgt_p += sizeof(struct T6_keyBlock_hdr);
-	keyHdr_p = (struct cca_token_hdr *)tgt_p;
-	keyHdr_p->token_identifier = CCA_TKN_HDR_ID_EXT;
-	tmp_l -= 4;
-	keyHdr_p->token_length = tmp_l;
-	tgt_p += sizeof(struct cca_token_hdr);
-	pvtSec_p = (struct cca_pvt_ext_CRT_sec *)tgt_p;
-	pvtSec_p->section_identifier = CCA_PVT_EXT_CRT_SEC_ID_PVT;
-	pvtSec_p->section_length =
-		sizeof(struct cca_pvt_ext_CRT_sec) + keyPartsLen;
-	pvtSec_p->key_format = CCA_PVT_EXT_CRT_SEC_FMT_CL;
-	pvtSec_p->key_use_flags[0] = CCA_PVT_USAGE_ALL;
-	pvtSec_p->p_len = long_len;
-	pvtSec_p->q_len = short_len;
-	pvtSec_p->dp_len = long_len;
-	pvtSec_p->dq_len = short_len;
-	pvtSec_p->u_len = long_len;
-	pvtSec_p->mod_len = mod_len;
-	pvtSec_p->pad_len = pad_len;
-	tgt_p += sizeof(struct cca_pvt_ext_CRT_sec);
-	if (copy_from_user(tgt_p, icaMsg_p->np_prime, long_len))
-		return SEN_RELEASED;
-	if (is_empty(tgt_p, long_len))
-		return SEN_USER_ERROR;
-	tgt_p += long_len;
-	if (copy_from_user(tgt_p, icaMsg_p->nq_prime, short_len))
-		return SEN_RELEASED;
-	if (is_empty(tgt_p, short_len))
-		return SEN_USER_ERROR;
-	tgt_p += short_len;
-	if (copy_from_user(tgt_p, icaMsg_p->bp_key, long_len))
-		return SEN_RELEASED;
-	if (is_empty(tgt_p, long_len))
-		return SEN_USER_ERROR;
-	tgt_p += long_len;
-	if (copy_from_user(tgt_p, icaMsg_p->bq_key, short_len))
-		return SEN_RELEASED;
-	if (is_empty(tgt_p, short_len))
-		return SEN_USER_ERROR;
-	tgt_p += short_len;
-	if (copy_from_user(tgt_p, icaMsg_p->u_mult_inv, long_len))
-		return SEN_RELEASED;
-	if (is_empty(tgt_p, long_len))
-		return SEN_USER_ERROR;
-	tgt_p += long_len;
-	tgt_p += pad_len;
-	memset(tgt_p, 0xFF, mod_len);
-	tgt_p += mod_len;
-	memcpy(tgt_p, &static_cca_pub_sec, sizeof(struct cca_public_sec));
-	pubSec_p = (struct cca_public_sec *) tgt_p;
-	pubSec_p->modulus_bit_len = 8 * mod_len;
-	*z90cMsg_l_p = tmp_size - CALLER_HEADER;
-
-	return 0;
-}
-
-static int
-ICAMEX_msg_to_type50MEX_msg(struct ica_rsa_modexpo *icaMex_p, int *z90cMsg_l_p,
-			    union type50_msg *z90cMsg_p)
-{
-	int mod_len, msg_size, mod_tgt_len, exp_tgt_len, inp_tgt_len;
-	unsigned char *mod_tgt, *exp_tgt, *inp_tgt;
-	union type50_msg *tmp_type50_msg;
-
-	mod_len = icaMex_p->inputdatalength;
-
-	msg_size = ((mod_len <= 128) ? TYPE50_MEB1_LEN : TYPE50_MEB2_LEN) +
-		    CALLER_HEADER;
-
-	memset(z90cMsg_p, 0, msg_size);
-
-	tmp_type50_msg = (union type50_msg *)
-		((unsigned char *) z90cMsg_p + CALLER_HEADER);
-
-	tmp_type50_msg->meb1.header.msg_type_code = TYPE50_TYPE_CODE;
-
-	if (mod_len <= 128) {
-		tmp_type50_msg->meb1.header.msg_len = TYPE50_MEB1_LEN;
-		tmp_type50_msg->meb1.keyblock_type = TYPE50_MEB1_FMT;
-		mod_tgt = tmp_type50_msg->meb1.modulus;
-		mod_tgt_len = sizeof(tmp_type50_msg->meb1.modulus);
-		exp_tgt = tmp_type50_msg->meb1.exponent;
-		exp_tgt_len = sizeof(tmp_type50_msg->meb1.exponent);
-		inp_tgt = tmp_type50_msg->meb1.message;
-		inp_tgt_len = sizeof(tmp_type50_msg->meb1.message);
-	} else {
-		tmp_type50_msg->meb2.header.msg_len = TYPE50_MEB2_LEN;
-		tmp_type50_msg->meb2.keyblock_type = TYPE50_MEB2_FMT;
-		mod_tgt = tmp_type50_msg->meb2.modulus;
-		mod_tgt_len = sizeof(tmp_type50_msg->meb2.modulus);
-		exp_tgt = tmp_type50_msg->meb2.exponent;
-		exp_tgt_len = sizeof(tmp_type50_msg->meb2.exponent);
-		inp_tgt = tmp_type50_msg->meb2.message;
-		inp_tgt_len = sizeof(tmp_type50_msg->meb2.message);
-	}
-
-	mod_tgt += (mod_tgt_len - mod_len);
-	if (copy_from_user(mod_tgt, icaMex_p->n_modulus, mod_len))
-		return SEN_RELEASED;
-	if (is_empty(mod_tgt, mod_len))
-		return SEN_USER_ERROR;
-	exp_tgt += (exp_tgt_len - mod_len);
-	if (copy_from_user(exp_tgt, icaMex_p->b_key, mod_len))
-		return SEN_RELEASED;
-	if (is_empty(exp_tgt, mod_len))
-		return SEN_USER_ERROR;
-	inp_tgt += (inp_tgt_len - mod_len);
-	if (copy_from_user(inp_tgt, icaMex_p->inputdata, mod_len))
-		return SEN_RELEASED;
-	if (is_empty(inp_tgt, mod_len))
-		return SEN_USER_ERROR;
-
-	*z90cMsg_l_p = msg_size - CALLER_HEADER;
-
-	return 0;
-}
-
-static int
-ICACRT_msg_to_type50CRT_msg(struct ica_rsa_modexpo_crt *icaMsg_p,
-			    int *z90cMsg_l_p, union type50_msg *z90cMsg_p)
-{
-	int mod_len, short_len, long_len, tmp_size, p_tgt_len, q_tgt_len,
-	    dp_tgt_len, dq_tgt_len, u_tgt_len, inp_tgt_len, long_offset;
-	unsigned char *p_tgt, *q_tgt, *dp_tgt, *dq_tgt, *u_tgt, *inp_tgt,
-		      temp[8];
-	union type50_msg *tmp_type50_msg;
-
-	mod_len = icaMsg_p->inputdatalength;
-	short_len = mod_len / 2;
-	long_len = mod_len / 2 + 8;
-	long_offset = 0;
-
-	if (long_len > 128) {
-		memset(temp, 0x00, sizeof(temp));
-		if (copy_from_user(temp, icaMsg_p->np_prime, long_len-128))
-			return SEN_RELEASED;
-		if (!is_empty(temp, 8))
-			return SEN_NOT_AVAIL;
-		if (copy_from_user(temp, icaMsg_p->bp_key, long_len-128))
-			return SEN_RELEASED;
-		if (!is_empty(temp, 8))
-			return SEN_NOT_AVAIL;
-		if (copy_from_user(temp, icaMsg_p->u_mult_inv, long_len-128))
-			return SEN_RELEASED;
-		if (!is_empty(temp, 8))
-			return SEN_NOT_AVAIL;
-		long_offset = long_len - 128;
-		long_len = 128;
-	}
-
-	tmp_size = ((long_len <= 64) ? TYPE50_CRB1_LEN : TYPE50_CRB2_LEN) +
-		    CALLER_HEADER;
-
-	memset(z90cMsg_p, 0, tmp_size);
-
-	tmp_type50_msg = (union type50_msg *)
-		((unsigned char *) z90cMsg_p + CALLER_HEADER);
-
-	tmp_type50_msg->crb1.header.msg_type_code = TYPE50_TYPE_CODE;
-	if (long_len <= 64) {
-		tmp_type50_msg->crb1.header.msg_len = TYPE50_CRB1_LEN;
-		tmp_type50_msg->crb1.keyblock_type = TYPE50_CRB1_FMT;
-		p_tgt = tmp_type50_msg->crb1.p;
-		p_tgt_len = sizeof(tmp_type50_msg->crb1.p);
-		q_tgt = tmp_type50_msg->crb1.q;
-		q_tgt_len = sizeof(tmp_type50_msg->crb1.q);
-		dp_tgt = tmp_type50_msg->crb1.dp;
-		dp_tgt_len = sizeof(tmp_type50_msg->crb1.dp);
-		dq_tgt = tmp_type50_msg->crb1.dq;
-		dq_tgt_len = sizeof(tmp_type50_msg->crb1.dq);
-		u_tgt = tmp_type50_msg->crb1.u;
-		u_tgt_len = sizeof(tmp_type50_msg->crb1.u);
-		inp_tgt = tmp_type50_msg->crb1.message;
-		inp_tgt_len = sizeof(tmp_type50_msg->crb1.message);
-	} else {
-		tmp_type50_msg->crb2.header.msg_len = TYPE50_CRB2_LEN;
-		tmp_type50_msg->crb2.keyblock_type = TYPE50_CRB2_FMT;
-		p_tgt = tmp_type50_msg->crb2.p;
-		p_tgt_len = sizeof(tmp_type50_msg->crb2.p);
-		q_tgt = tmp_type50_msg->crb2.q;
-		q_tgt_len = sizeof(tmp_type50_msg->crb2.q);
-		dp_tgt = tmp_type50_msg->crb2.dp;
-		dp_tgt_len = sizeof(tmp_type50_msg->crb2.dp);
-		dq_tgt = tmp_type50_msg->crb2.dq;
-		dq_tgt_len = sizeof(tmp_type50_msg->crb2.dq);
-		u_tgt = tmp_type50_msg->crb2.u;
-		u_tgt_len = sizeof(tmp_type50_msg->crb2.u);
-		inp_tgt = tmp_type50_msg->crb2.message;
-		inp_tgt_len = sizeof(tmp_type50_msg->crb2.message);
-	}
-
-	p_tgt += (p_tgt_len - long_len);
-	if (copy_from_user(p_tgt, icaMsg_p->np_prime + long_offset, long_len))
-		return SEN_RELEASED;
-	if (is_empty(p_tgt, long_len))
-		return SEN_USER_ERROR;
-	q_tgt += (q_tgt_len - short_len);
-	if (copy_from_user(q_tgt, icaMsg_p->nq_prime, short_len))
-		return SEN_RELEASED;
-	if (is_empty(q_tgt, short_len))
-		return SEN_USER_ERROR;
-	dp_tgt += (dp_tgt_len - long_len);
-	if (copy_from_user(dp_tgt, icaMsg_p->bp_key + long_offset, long_len))
-		return SEN_RELEASED;
-	if (is_empty(dp_tgt, long_len))
-		return SEN_USER_ERROR;
-	dq_tgt += (dq_tgt_len - short_len);
-	if (copy_from_user(dq_tgt, icaMsg_p->bq_key, short_len))
-		return SEN_RELEASED;
-	if (is_empty(dq_tgt, short_len))
-		return SEN_USER_ERROR;
-	u_tgt += (u_tgt_len - long_len);
-	if (copy_from_user(u_tgt, icaMsg_p->u_mult_inv + long_offset, long_len))
-		return SEN_RELEASED;
-	if (is_empty(u_tgt, long_len))
-		return SEN_USER_ERROR;
-	inp_tgt += (inp_tgt_len - mod_len);
-	if (copy_from_user(inp_tgt, icaMsg_p->inputdata, mod_len))
-		return SEN_RELEASED;
-	if (is_empty(inp_tgt, mod_len))
-		return SEN_USER_ERROR;
-
-	*z90cMsg_l_p = tmp_size - CALLER_HEADER;
-
-	return 0;
-}
-
-int
-convert_request(unsigned char *buffer, int func, unsigned short function,
-		int cdx, int dev_type, int *msg_l_p, unsigned char *msg_p)
-{
-	if (dev_type == PCICA) {
-		if (func == ICARSACRT)
-			return ICACRT_msg_to_type4CRT_msg(
-				(struct ica_rsa_modexpo_crt *) buffer,
-				msg_l_p, (union type4_msg *) msg_p);
-		else
-			return ICAMEX_msg_to_type4MEX_msg(
-				(struct ica_rsa_modexpo *) buffer,
-				msg_l_p, (union type4_msg *) msg_p);
-	}
-	if (dev_type == PCICC) {
-		if (func == ICARSACRT)
-			return ICACRT_msg_to_type6CRT_msg(
-				(struct ica_rsa_modexpo_crt *) buffer,
-				cdx, msg_l_p, (struct type6_msg *)msg_p);
-		if (function == PCI_FUNC_KEY_ENCRYPT)
-			return ICAMEX_msg_to_type6MEX_en_msg(
-				(struct ica_rsa_modexpo *) buffer,
-				cdx, msg_l_p, (struct type6_msg *) msg_p);
-		else
-			return ICAMEX_msg_to_type6MEX_de_msg(
-				(struct ica_rsa_modexpo *) buffer,
-				cdx, msg_l_p, (struct type6_msg *) msg_p);
-	}
-	if ((dev_type == PCIXCC_MCL2) ||
-	    (dev_type == PCIXCC_MCL3) ||
-	    (dev_type == CEX2C)) {
-		if (func == ICARSACRT)
-			return ICACRT_msg_to_type6CRT_msgX(
-				(struct ica_rsa_modexpo_crt *) buffer,
-				cdx, msg_l_p, (struct type6_msg *) msg_p,
-				dev_type);
-		else
-			return ICAMEX_msg_to_type6MEX_msgX(
-				(struct ica_rsa_modexpo *) buffer,
-				cdx, msg_l_p, (struct type6_msg *) msg_p,
-				dev_type);
-	}
-	if (dev_type == CEX2A) {
-		if (func == ICARSACRT)
-			return ICACRT_msg_to_type50CRT_msg(
-				(struct ica_rsa_modexpo_crt *) buffer,
-				msg_l_p, (union type50_msg *) msg_p);
-		else
-			return ICAMEX_msg_to_type50MEX_msg(
-				(struct ica_rsa_modexpo *) buffer,
-				msg_l_p, (union type50_msg *) msg_p);
-	}
-
-	return 0;
-}
-
-int ext_bitlens_msg_count = 0;
-static inline void
-unset_ext_bitlens(void)
-{
-	if (!ext_bitlens_msg_count) {
-		PRINTK("Unable to use coprocessors for extended bitlengths. "
-		       "Using PCICAs/CEX2As (if present) for extended "
-		       "bitlengths. This is not an error.\n");
-		ext_bitlens_msg_count++;
-	}
-	ext_bitlens = 0;
-}
-
-int
-convert_response(unsigned char *response, unsigned char *buffer,
-		 int *respbufflen_p, unsigned char *resp_buff)
-{
-	struct ica_rsa_modexpo *icaMsg_p = (struct ica_rsa_modexpo *) buffer;
-	struct error_hdr *errh_p = (struct error_hdr *) response;
-	struct type80_hdr *t80h_p = (struct type80_hdr *) response;
-	struct type84_hdr *t84h_p = (struct type84_hdr *) response;
-	struct type86_fmt2_msg *t86m_p =  (struct type86_fmt2_msg *) response;
-	int reply_code, service_rc, service_rs, src_l;
-	unsigned char *src_p, *tgt_p;
-	struct CPRB *cprb_p;
-	struct CPRBX *cprbx_p;
-
-	src_p = 0;
-	reply_code = 0;
-	service_rc = 0;
-	service_rs = 0;
-	src_l = 0;
-	switch (errh_p->type) {
-	case TYPE82_RSP_CODE:
-	case TYPE88_RSP_CODE:
-		reply_code = errh_p->reply_code;
-		src_p = (unsigned char *)errh_p;
-		PRINTK("Hardware error: Type %02X Message Header: "
-		       "%02x%02x%02x%02x%02x%02x%02x%02x\n",
-		       errh_p->type,
-		       src_p[0], src_p[1], src_p[2], src_p[3],
-		       src_p[4], src_p[5], src_p[6], src_p[7]);
-		break;
-	case TYPE80_RSP_CODE:
-		src_l = icaMsg_p->outputdatalength;
-		src_p = response + (int)t80h_p->len - src_l;
-		break;
-	case TYPE84_RSP_CODE:
-		src_l = icaMsg_p->outputdatalength;
-		src_p = response + (int)t84h_p->len - src_l;
-		break;
-	case TYPE86_RSP_CODE:
-		reply_code = t86m_p->header.reply_code;
-		if (reply_code != 0)
-			break;
-		cprb_p = (struct CPRB *)
-			(response + sizeof(struct type86_fmt2_msg));
-		cprbx_p = (struct CPRBX *) cprb_p;
-		if (cprb_p->cprb_ver_id != 0x02) {
-			le2toI(cprb_p->ccp_rtcode, &service_rc);
-			if (service_rc != 0) {
-				le2toI(cprb_p->ccp_rscode, &service_rs);
-				if ((service_rc == 8) && (service_rs == 66))
-					PDEBUG("Bad block format on PCICC\n");
-				else if ((service_rc == 8) && (service_rs == 65))
-					PDEBUG("Probably an even modulus on "
-					       "PCICC\n");
-				else if ((service_rc == 8) && (service_rs == 770)) {
-					PDEBUG("Invalid key length on PCICC\n");
-					unset_ext_bitlens();
-					return REC_USE_PCICA;
-				}
-				else if ((service_rc == 8) && (service_rs == 783)) {
-					PDEBUG("Extended bitlengths not enabled"
-					       "on PCICC\n");
-					unset_ext_bitlens();
-					return REC_USE_PCICA;
-				}
-				else
-					PRINTK("service rc/rs (PCICC): %d/%d\n",
-					       service_rc, service_rs);
-				return REC_OPERAND_INV;
-			}
-			src_p = (unsigned char *)cprb_p + sizeof(struct CPRB);
-			src_p += 4;
-			le2toI(src_p, &src_l);
-			src_l -= 2;
-			src_p += 2;
-		} else {
-			service_rc = (int)cprbx_p->ccp_rtcode;
-			if (service_rc != 0) {
-				service_rs = (int) cprbx_p->ccp_rscode;
-				if ((service_rc == 8) && (service_rs == 66))
-					PDEBUG("Bad block format on PCIXCC\n");
-				else if ((service_rc == 8) && (service_rs == 65))
-					PDEBUG("Probably an even modulus on "
-					       "PCIXCC\n");
-				else if ((service_rc == 8) && (service_rs == 770)) {
-					PDEBUG("Invalid key length on PCIXCC\n");
-					unset_ext_bitlens();
-					return REC_USE_PCICA;
-				}
-				else if ((service_rc == 8) && (service_rs == 783)) {
-					PDEBUG("Extended bitlengths not enabled"
-					       "on PCIXCC\n");
-					unset_ext_bitlens();
-					return REC_USE_PCICA;
-				}
-				else
-					PRINTK("service rc/rs (PCIXCC): %d/%d\n",
-					       service_rc, service_rs);
-				return REC_OPERAND_INV;
-			}
-			src_p = (unsigned char *)
-				cprbx_p + sizeof(struct CPRBX);
-			src_p += 4;
-			src_l = (int)(*((short *) src_p));
-			src_l -= 2;
-			src_p += 2;
-		}
-		break;
-	default:
-		src_p = (unsigned char *)errh_p;
-		PRINTK("Unrecognized Message Header: "
-		       "%02x%02x%02x%02x%02x%02x%02x%02x\n",
-		       src_p[0], src_p[1], src_p[2], src_p[3],
-		       src_p[4], src_p[5], src_p[6], src_p[7]);
-		return REC_BAD_MESSAGE;
-	}
-
-	if (reply_code)
-		switch (reply_code) {
-		case REP82_ERROR_MACHINE_FAILURE:
-			if (errh_p->type == TYPE82_RSP_CODE)
-				PRINTKW("Machine check failure\n");
-			else
-				PRINTKW("Module failure\n");
-			return REC_HARDWAR_ERR;
-		case REP82_ERROR_OPERAND_INVALID:
-			return REC_OPERAND_INV;
-		case REP88_ERROR_MESSAGE_MALFORMD:
-			PRINTKW("Message malformed\n");
-			return REC_OPERAND_INV;
-		case REP82_ERROR_OPERAND_SIZE:
-			return REC_OPERAND_SIZE;
-		case REP82_ERROR_EVEN_MOD_IN_OPND:
-			return REC_EVEN_MOD;
-		case REP82_ERROR_MESSAGE_TYPE:
-			return WRONG_DEVICE_TYPE;
-		case REP82_ERROR_TRANSPORT_FAIL:
-			PRINTKW("Transport failed (APFS = %02X%02X%02X%02X)\n",
-				t86m_p->apfs[0], t86m_p->apfs[1],
-				t86m_p->apfs[2], t86m_p->apfs[3]);
-			return REC_HARDWAR_ERR;
-		default:
-			PRINTKW("reply code = %d\n", reply_code);
-			return REC_HARDWAR_ERR;
-		}
-
-	if (service_rc != 0)
-		return REC_OPERAND_INV;
-
-	if ((src_l > icaMsg_p->outputdatalength) ||
-	    (src_l > RESPBUFFSIZE) ||
-	    (src_l <= 0))
-		return REC_OPERAND_SIZE;
-
-	PDEBUG("Length returned = %d\n", src_l);
-	tgt_p = resp_buff + icaMsg_p->outputdatalength - src_l;
-	memcpy(tgt_p, src_p, src_l);
-	if ((errh_p->type == TYPE86_RSP_CODE) && (resp_buff < tgt_p)) {
-		memset(resp_buff, 0, icaMsg_p->outputdatalength - src_l);
-		if (pad_msg(resp_buff, icaMsg_p->outputdatalength, src_l))
-			return REC_INVALID_PAD;
-	}
-	*respbufflen_p = icaMsg_p->outputdatalength;
-	if (*respbufflen_p == 0)
-		PRINTK("Zero *respbufflen_p\n");
-
-	return 0;
-}
-
diff --git a/drivers/s390/crypto/z90main.c b/drivers/s390/crypto/z90main.c
deleted file mode 100644
index b2f20ab8431a..000000000000
--- a/drivers/s390/crypto/z90main.c
+++ /dev/null
@@ -1,3379 +0,0 @@
-/*
- *  linux/drivers/s390/crypto/z90main.c
- *
- *  z90crypt 1.3.3
- *
- *  Copyright (C)  2001, 2005 IBM Corporation
- *  Author(s): Robert Burroughs (burrough@us.ibm.com)
- *             Eric Rossman (edrossma@us.ibm.com)
- *
- *  Hotplug & misc device support: Jochen Roehrig (roehrig@de.ibm.com)
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2, or (at your option)
- * any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
- */
-
-#include <asm/uaccess.h>       // copy_(from|to)_user
-#include <linux/compat.h>
-#include <linux/compiler.h>
-#include <linux/delay.h>       // mdelay
-#include <linux/init.h>
-#include <linux/interrupt.h>   // for tasklets
-#include <linux/miscdevice.h>
-#include <linux/module.h>
-#include <linux/moduleparam.h>
-#include <linux/proc_fs.h>
-#include <linux/syscalls.h>
-#include "z90crypt.h"
-#include "z90common.h"
-
-/**
- * Defaults that may be modified.
- */
-
-/**
- * You can specify a different minor at compile time.
- */
-#ifndef Z90CRYPT_MINOR
-#define Z90CRYPT_MINOR	MISC_DYNAMIC_MINOR
-#endif
-
-/**
- * You can specify a different domain at compile time or on the insmod
- * command line.
- */
-#ifndef DOMAIN_INDEX
-#define DOMAIN_INDEX	-1
-#endif
-
-/**
- * This is the name under which the device is registered in /proc/modules.
- */
-#define REG_NAME	"z90crypt"
-
-/**
- * Cleanup should run every CLEANUPTIME seconds and should clean up requests
- * older than CLEANUPTIME seconds in the past.
- */
-#ifndef CLEANUPTIME
-#define CLEANUPTIME 15
-#endif
-
-/**
- * Config should run every CONFIGTIME seconds
- */
-#ifndef CONFIGTIME
-#define CONFIGTIME 30
-#endif
-
-/**
- * The first execution of the config task should take place
- * immediately after initialization
- */
-#ifndef INITIAL_CONFIGTIME
-#define INITIAL_CONFIGTIME 1
-#endif
-
-/**
- * Reader should run every READERTIME milliseconds
- * With the 100Hz patch for s390, z90crypt can lock the system solid while
- * under heavy load. We'll try to avoid that.
- */
-#ifndef READERTIME
-#if HZ > 1000
-#define READERTIME 2
-#else
-#define READERTIME 10
-#endif
-#endif
-
-/**
- * turn long device array index into device pointer
- */
-#define LONG2DEVPTR(ndx) (z90crypt.device_p[(ndx)])
-
-/**
- * turn short device array index into long device array index
- */
-#define SHRT2LONG(ndx) (z90crypt.overall_device_x.device_index[(ndx)])
-
-/**
- * turn short device array index into device pointer
- */
-#define SHRT2DEVPTR(ndx) LONG2DEVPTR(SHRT2LONG(ndx))
-
-/**
- * Status for a work-element
- */
-#define STAT_DEFAULT	0x00 // request has not been processed
-
-#define STAT_ROUTED	0x80 // bit 7: requests get routed to specific device
-			     //	       else, device is determined each write
-#define STAT_FAILED	0x40 // bit 6: this bit is set if the request failed
-			     //	       before being sent to the hardware.
-#define STAT_WRITTEN	0x30 // bits 5-4: work to be done, not sent to device
-//			0x20 // UNUSED state
-#define STAT_READPEND	0x10 // bits 5-4: work done, we're returning data now
-#define STAT_NOWORK	0x00 // bits off: no work on any queue
-#define STAT_RDWRMASK	0x30 // mask for bits 5-4
-
-/**
- * Macros to check the status RDWRMASK
- */
-#define CHK_RDWRMASK(statbyte) ((statbyte) & STAT_RDWRMASK)
-#define SET_RDWRMASK(statbyte, newval) \
-	{(statbyte) &= ~STAT_RDWRMASK; (statbyte) |= newval;}
-
-/**
- * Audit Trail.	 Progress of a Work element
- * audit[0]: Unless noted otherwise, these bits are all set by the process
- */
-#define FP_COPYFROM	0x80 // Caller's buffer has been copied to work element
-#define FP_BUFFREQ	0x40 // Low Level buffer requested
-#define FP_BUFFGOT	0x20 // Low Level buffer obtained
-#define FP_SENT		0x10 // Work element sent to a crypto device
-			     // (may be set by process or by reader task)
-#define FP_PENDING	0x08 // Work element placed on pending queue
-			     // (may be set by process or by reader task)
-#define FP_REQUEST	0x04 // Work element placed on request queue
-#define FP_ASLEEP	0x02 // Work element about to sleep
-#define FP_AWAKE	0x01 // Work element has been awakened
-
-/**
- * audit[1]: These bits are set by the reader task and/or the cleanup task
- */
-#define FP_NOTPENDING	  0x80 // Work element removed from pending queue
-#define FP_AWAKENING	  0x40 // Caller about to be awakened
-#define FP_TIMEDOUT	  0x20 // Caller timed out
-#define FP_RESPSIZESET	  0x10 // Response size copied to work element
-#define FP_RESPADDRCOPIED 0x08 // Response address copied to work element
-#define FP_RESPBUFFCOPIED 0x04 // Response buffer copied to work element
-#define FP_REMREQUEST	  0x02 // Work element removed from request queue
-#define FP_SIGNALED	  0x01 // Work element was awakened by a signal
-
-/**
- * audit[2]: unused
- */
-
-/**
- * state of the file handle in private_data.status
- */
-#define STAT_OPEN 0
-#define STAT_CLOSED 1
-
-/**
- * PID() expands to the process ID of the current process
- */
-#define PID() (current->pid)
-
-/**
- * Selected Constants.	The number of APs and the number of devices
- */
-#ifndef Z90CRYPT_NUM_APS
-#define Z90CRYPT_NUM_APS 64
-#endif
-#ifndef Z90CRYPT_NUM_DEVS
-#define Z90CRYPT_NUM_DEVS Z90CRYPT_NUM_APS
-#endif
-
-/**
- * Buffer size for receiving responses. The maximum Response Size
- * is actually the maximum request size, since in an error condition
- * the request itself may be returned unchanged.
- */
-#define MAX_RESPONSE_SIZE 0x0000077C
-
-/**
- * A count and status-byte mask
- */
-struct status {
-	int	      st_count;		    // # of enabled devices
-	int	      disabled_count;	    // # of disabled devices
-	int	      user_disabled_count;  // # of devices disabled via proc fs
-	unsigned char st_mask[Z90CRYPT_NUM_APS]; // current status mask
-};
-
-/**
- * The array of device indexes is a mechanism for fast indexing into
- * a long (and sparse) array.  For instance, if APs 3, 9 and 47 are
- * installed, z90CDeviceIndex[0] is 3, z90CDeviceIndex[1] is 9, and
- * z90CDeviceIndex[2] is 47.
- */
-struct device_x {
-	int device_index[Z90CRYPT_NUM_DEVS];
-};
-
-/**
- * All devices are arranged in a single array: 64 APs
- */
-struct device {
-	int		 dev_type;	    // PCICA, PCICC, PCIXCC_MCL2,
-					    // PCIXCC_MCL3, CEX2C, CEX2A
-	enum devstat	 dev_stat;	    // current device status
-	int		 dev_self_x;	    // Index in array
-	int		 disabled;	    // Set when device is in error
-	int		 user_disabled;	    // Set when device is disabled by user
-	int		 dev_q_depth;	    // q depth
-	unsigned char *	 dev_resp_p;	    // Response buffer address
-	int		 dev_resp_l;	    // Response Buffer length
-	int		 dev_caller_count;  // Number of callers
-	int		 dev_total_req_cnt; // # requests for device since load
-	struct list_head dev_caller_list;   // List of callers
-};
-
-/**
- * There's a struct status and a struct device_x for each device type.
- */
-struct hdware_block {
-	struct status	hdware_mask;
-	struct status	type_mask[Z90CRYPT_NUM_TYPES];
-	struct device_x type_x_addr[Z90CRYPT_NUM_TYPES];
-	unsigned char	device_type_array[Z90CRYPT_NUM_APS];
-};
-
-/**
- * z90crypt is the topmost data structure in the hierarchy.
- */
-struct z90crypt {
-	int		     max_count;		// Nr of possible crypto devices
-	struct status	     mask;
-	int		     q_depth_array[Z90CRYPT_NUM_DEVS];
-	int		     dev_type_array[Z90CRYPT_NUM_DEVS];
-	struct device_x	     overall_device_x;	// array device indexes
-	struct device *	     device_p[Z90CRYPT_NUM_DEVS];
-	int		     terminating;
-	int		     domain_established;// TRUE:  domain has been found
-	int		     cdx;		// Crypto Domain Index
-	int		     len;		// Length of this data structure
-	struct hdware_block *hdware_info;
-};
-
-/**
- * An array of these structures is pointed to from dev_caller
- * The length of the array depends on the device type. For APs,
- * there are 8.
- *
- * The caller buffer is allocated to the user at OPEN. At WRITE,
- * it contains the request; at READ, the response. The function
- * send_to_crypto_device converts the request to device-dependent
- * form and use the caller's OPEN-allocated buffer for the response.
- *
- * For the contents of caller_dev_dep_req and caller_dev_dep_req_p
- * because that points to it, see the discussion in z90hardware.c.
- * Search for "extended request message block".
- */
-struct caller {
-	int		 caller_buf_l;		 // length of original request
-	unsigned char *	 caller_buf_p;		 // Original request on WRITE
-	int		 caller_dev_dep_req_l;	 // len device dependent request
-	unsigned char *	 caller_dev_dep_req_p;	 // Device dependent form
-	unsigned char	 caller_id[8];		 // caller-supplied message id
-	struct list_head caller_liste;
-	unsigned char	 caller_dev_dep_req[MAX_RESPONSE_SIZE];
-};
-
-/**
- * Function prototypes from z90hardware.c
- */
-enum hdstat query_online(int deviceNr, int cdx, int resetNr, int *q_depth,
-			 int *dev_type);
-enum devstat reset_device(int deviceNr, int cdx, int resetNr);
-enum devstat send_to_AP(int dev_nr, int cdx, int msg_len, unsigned char *msg_ext);
-enum devstat receive_from_AP(int dev_nr, int cdx, int resplen,
-			     unsigned char *resp, unsigned char *psmid);
-int convert_request(unsigned char *buffer, int func, unsigned short function,
-		    int cdx, int dev_type, int *msg_l_p, unsigned char *msg_p);
-int convert_response(unsigned char *response, unsigned char *buffer,
-		     int *respbufflen_p, unsigned char *resp_buff);
-
-/**
- * Low level function prototypes
- */
-static int create_z90crypt(int *cdx_p);
-static int refresh_z90crypt(int *cdx_p);
-static int find_crypto_devices(struct status *deviceMask);
-static int create_crypto_device(int index);
-static int destroy_crypto_device(int index);
-static void destroy_z90crypt(void);
-static int refresh_index_array(struct status *status_str,
-			       struct device_x *index_array);
-static int probe_device_type(struct device *devPtr);
-static int probe_PCIXCC_type(struct device *devPtr);
-
-/**
- * proc fs definitions
- */
-static struct proc_dir_entry *z90crypt_entry;
-
-/**
- * data structures
- */
-
-/**
- * work_element.opener points back to this structure
- */
-struct priv_data {
-	pid_t	opener_pid;
-	unsigned char	status;		// 0: open  1: closed
-};
-
-/**
- * A work element is allocated for each request
- */
-struct work_element {
-	struct priv_data *priv_data;
-	pid_t		  pid;
-	int		  devindex;	  // index of device processing this w_e
-					  // (If request did not specify device,
-					  // -1 until placed onto a queue)
-	int		  devtype;
-	struct list_head  liste;	  // used for requestq and pendingq
-	char		  buffer[128];	  // local copy of user request
-	int		  buff_size;	  // size of the buffer for the request
-	char		  resp_buff[RESPBUFFSIZE];
-	int		  resp_buff_size;
-	char __user *	  resp_addr;	  // address of response in user space
-	unsigned int	  funccode;	  // function code of request
-	wait_queue_head_t waitq;
-	unsigned long	  requestsent;	  // time at which the request was sent
-	atomic_t	  alarmrung;	  // wake-up signal
-	unsigned char	  caller_id[8];	  // pid + counter, for this w_e
-	unsigned char	  status[1];	  // bits to mark status of the request
-	unsigned char	  audit[3];	  // record of work element's progress
-	unsigned char *	  requestptr;	  // address of request buffer
-	int		  retcode;	  // return code of request
-};
-
-/**
- * High level function prototypes
- */
-static int z90crypt_open(struct inode *, struct file *);
-static int z90crypt_release(struct inode *, struct file *);
-static ssize_t z90crypt_read(struct file *, char __user *, size_t, loff_t *);
-static ssize_t z90crypt_write(struct file *, const char __user *,
-							size_t, loff_t *);
-static long z90crypt_unlocked_ioctl(struct file *, unsigned int, unsigned long);
-static long z90crypt_compat_ioctl(struct file *, unsigned int, unsigned long);
-
-static void z90crypt_reader_task(unsigned long);
-static void z90crypt_schedule_reader_task(unsigned long);
-static void z90crypt_config_task(unsigned long);
-static void z90crypt_cleanup_task(unsigned long);
-
-static int z90crypt_status(char *, char **, off_t, int, int *, void *);
-static int z90crypt_status_write(struct file *, const char __user *,
-				 unsigned long, void *);
-
-/**
- * Storage allocated at initialization and used throughout the life of
- * this insmod
- */
-static int domain = DOMAIN_INDEX;
-static struct z90crypt z90crypt;
-static int quiesce_z90crypt;
-static spinlock_t queuespinlock;
-static struct list_head request_list;
-static int requestq_count;
-static struct list_head pending_list;
-static int pendingq_count;
-
-static struct tasklet_struct reader_tasklet;
-static struct timer_list reader_timer;
-static struct timer_list config_timer;
-static struct timer_list cleanup_timer;
-static atomic_t total_open;
-static atomic_t z90crypt_step;
-
-static struct file_operations z90crypt_fops = {
-	.owner		= THIS_MODULE,
-	.read		= z90crypt_read,
-	.write		= z90crypt_write,
-	.unlocked_ioctl	= z90crypt_unlocked_ioctl,
-#ifdef CONFIG_COMPAT
-	.compat_ioctl	= z90crypt_compat_ioctl,
-#endif
-	.open		= z90crypt_open,
-	.release	= z90crypt_release
-};
-
-static struct miscdevice z90crypt_misc_device = {
-	.minor	    = Z90CRYPT_MINOR,
-	.name	    = DEV_NAME,
-	.fops	    = &z90crypt_fops,
-};
-
-/**
- * Documentation values.
- */
-MODULE_AUTHOR("zSeries Linux Crypto Team: Robert H. Burroughs, Eric D. Rossman"
-	      "and Jochen Roehrig");
-MODULE_DESCRIPTION("zSeries Linux Cryptographic Coprocessor device driver, "
-		   "Copyright 2001, 2005 IBM Corporation");
-MODULE_LICENSE("GPL");
-module_param(domain, int, 0);
-MODULE_PARM_DESC(domain, "domain index for device");
-
-#ifdef CONFIG_COMPAT
-/**
- * ioctl32 conversion routines
- */
-struct ica_rsa_modexpo_32 { // For 32-bit callers
-	compat_uptr_t	inputdata;
-	unsigned int	inputdatalength;
-	compat_uptr_t	outputdata;
-	unsigned int	outputdatalength;
-	compat_uptr_t	b_key;
-	compat_uptr_t	n_modulus;
-};
-
-static long
-trans_modexpo32(struct file *filp, unsigned int cmd, unsigned long arg)
-{
-	struct ica_rsa_modexpo_32 __user *mex32u = compat_ptr(arg);
-	struct ica_rsa_modexpo_32  mex32k;
-	struct ica_rsa_modexpo __user *mex64;
-	long ret = 0;
-	unsigned int i;
-
-	if (!access_ok(VERIFY_WRITE, mex32u, sizeof(struct ica_rsa_modexpo_32)))
-		return -EFAULT;
-	mex64 = compat_alloc_user_space(sizeof(struct ica_rsa_modexpo));
-	if (!access_ok(VERIFY_WRITE, mex64, sizeof(struct ica_rsa_modexpo)))
-		return -EFAULT;
-	if (copy_from_user(&mex32k, mex32u, sizeof(struct ica_rsa_modexpo_32)))
-		return -EFAULT;
-	if (__put_user(compat_ptr(mex32k.inputdata), &mex64->inputdata)   ||
-	    __put_user(mex32k.inputdatalength, &mex64->inputdatalength)   ||
-	    __put_user(compat_ptr(mex32k.outputdata), &mex64->outputdata) ||
-	    __put_user(mex32k.outputdatalength, &mex64->outputdatalength) ||
-	    __put_user(compat_ptr(mex32k.b_key), &mex64->b_key)           ||
-	    __put_user(compat_ptr(mex32k.n_modulus), &mex64->n_modulus))
-		return -EFAULT;
-	ret = z90crypt_unlocked_ioctl(filp, cmd, (unsigned long)mex64);
-	if (!ret)
-		if (__get_user(i, &mex64->outputdatalength) ||
-		    __put_user(i, &mex32u->outputdatalength))
-			ret = -EFAULT;
-	return ret;
-}
-
-struct ica_rsa_modexpo_crt_32 { // For 32-bit callers
-	compat_uptr_t	inputdata;
-	unsigned int	inputdatalength;
-	compat_uptr_t	outputdata;
-	unsigned int	outputdatalength;
-	compat_uptr_t	bp_key;
-	compat_uptr_t	bq_key;
-	compat_uptr_t	np_prime;
-	compat_uptr_t	nq_prime;
-	compat_uptr_t	u_mult_inv;
-};
-
-static long
-trans_modexpo_crt32(struct file *filp, unsigned int cmd, unsigned long arg)
-{
-	struct ica_rsa_modexpo_crt_32 __user *crt32u = compat_ptr(arg);
-	struct ica_rsa_modexpo_crt_32  crt32k;
-	struct ica_rsa_modexpo_crt __user *crt64;
-	long ret = 0;
-	unsigned int i;
-
-	if (!access_ok(VERIFY_WRITE, crt32u,
-		       sizeof(struct ica_rsa_modexpo_crt_32)))
-		return -EFAULT;
-	crt64 = compat_alloc_user_space(sizeof(struct ica_rsa_modexpo_crt));
-	if (!access_ok(VERIFY_WRITE, crt64, sizeof(struct ica_rsa_modexpo_crt)))
-		return -EFAULT;
-	if (copy_from_user(&crt32k, crt32u,
-			   sizeof(struct ica_rsa_modexpo_crt_32)))
-		return -EFAULT;
-	if (__put_user(compat_ptr(crt32k.inputdata), &crt64->inputdata)   ||
-	    __put_user(crt32k.inputdatalength, &crt64->inputdatalength)   ||
-	    __put_user(compat_ptr(crt32k.outputdata), &crt64->outputdata) ||
-	    __put_user(crt32k.outputdatalength, &crt64->outputdatalength) ||
-	    __put_user(compat_ptr(crt32k.bp_key), &crt64->bp_key)         ||
-	    __put_user(compat_ptr(crt32k.bq_key), &crt64->bq_key)         ||
-	    __put_user(compat_ptr(crt32k.np_prime), &crt64->np_prime)     ||
-	    __put_user(compat_ptr(crt32k.nq_prime), &crt64->nq_prime)     ||
-	    __put_user(compat_ptr(crt32k.u_mult_inv), &crt64->u_mult_inv))
-		return -EFAULT;
-	ret = z90crypt_unlocked_ioctl(filp, cmd, (unsigned long)crt64);
-	if (!ret)
-		if (__get_user(i, &crt64->outputdatalength) ||
-		    __put_user(i, &crt32u->outputdatalength))
-			ret = -EFAULT;
-	return ret;
-}
-
-static long
-z90crypt_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
-{
-	switch (cmd) {
-	case ICAZ90STATUS:
-	case Z90QUIESCE:
-	case Z90STAT_TOTALCOUNT:
-	case Z90STAT_PCICACOUNT:
-	case Z90STAT_PCICCCOUNT:
-	case Z90STAT_PCIXCCCOUNT:
-	case Z90STAT_PCIXCCMCL2COUNT:
-	case Z90STAT_PCIXCCMCL3COUNT:
-	case Z90STAT_CEX2CCOUNT:
-	case Z90STAT_REQUESTQ_COUNT:
-	case Z90STAT_PENDINGQ_COUNT:
-	case Z90STAT_TOTALOPEN_COUNT:
-	case Z90STAT_DOMAIN_INDEX:
-	case Z90STAT_STATUS_MASK:
-	case Z90STAT_QDEPTH_MASK:
-	case Z90STAT_PERDEV_REQCNT:
-		return z90crypt_unlocked_ioctl(filp, cmd, arg);
-	case ICARSAMODEXPO:
-		return trans_modexpo32(filp, cmd, arg);
-	case ICARSACRT:
-		return trans_modexpo_crt32(filp, cmd, arg);
-	default:
-		return -ENOIOCTLCMD;
-  	}
-}
-#endif
-
-/**
- * The module initialization code.
- */
-static int __init
-z90crypt_init_module(void)
-{
-	int result, nresult;
-	struct proc_dir_entry *entry;
-
-	PDEBUG("PID %d\n", PID());
-
-	if ((domain < -1) || (domain > 15)) {
-		PRINTKW("Invalid param: domain = %d.  Not loading.\n", domain);
-		return -EINVAL;
-	}
-
-	/* Register as misc device with given minor (or get a dynamic one). */
-	result = misc_register(&z90crypt_misc_device);
-	if (result < 0) {
-		PRINTKW(KERN_ERR "misc_register (minor %d) failed with %d\n",
-			z90crypt_misc_device.minor, result);
-		return result;
-	}
-
-	PDEBUG("Registered " DEV_NAME " with result %d\n", result);
-
-	result = create_z90crypt(&domain);
-	if (result != 0) {
-		PRINTKW("create_z90crypt (domain index %d) failed with %d.\n",
-			domain, result);
-		result = -ENOMEM;
-		goto init_module_cleanup;
-	}
-
-	if (result == 0) {
-		PRINTKN("Version %d.%d.%d loaded, built on %s %s\n",
-			z90crypt_VERSION, z90crypt_RELEASE, z90crypt_VARIANT,
-			__DATE__, __TIME__);
-		PDEBUG("create_z90crypt (domain index %d) successful.\n",
-		       domain);
-	} else
-		PRINTK("No devices at startup\n");
-
-	/* Initialize globals. */
-	spin_lock_init(&queuespinlock);
-
-	INIT_LIST_HEAD(&pending_list);
-	pendingq_count = 0;
-
-	INIT_LIST_HEAD(&request_list);
-	requestq_count = 0;
-
-	quiesce_z90crypt = 0;
-
-	atomic_set(&total_open, 0);
-	atomic_set(&z90crypt_step, 0);
-
-	/* Set up the cleanup task. */
-	init_timer(&cleanup_timer);
-	cleanup_timer.function = z90crypt_cleanup_task;
-	cleanup_timer.data = 0;
-	cleanup_timer.expires = jiffies + (CLEANUPTIME * HZ);
-	add_timer(&cleanup_timer);
-
-	/* Set up the proc file system */
-	entry = create_proc_entry("driver/z90crypt", 0644, 0);
-	if (entry) {
-		entry->nlink = 1;
-		entry->data = 0;
-		entry->read_proc = z90crypt_status;
-		entry->write_proc = z90crypt_status_write;
-	}
-	else
-		PRINTK("Couldn't create z90crypt proc entry\n");
-	z90crypt_entry = entry;
-
-	/* Set up the configuration task. */
-	init_timer(&config_timer);
-	config_timer.function = z90crypt_config_task;
-	config_timer.data = 0;
-	config_timer.expires = jiffies + (INITIAL_CONFIGTIME * HZ);
-	add_timer(&config_timer);
-
-	/* Set up the reader task */
-	tasklet_init(&reader_tasklet, z90crypt_reader_task, 0);
-	init_timer(&reader_timer);
-	reader_timer.function = z90crypt_schedule_reader_task;
-	reader_timer.data = 0;
-	reader_timer.expires = jiffies + (READERTIME * HZ / 1000);
-	add_timer(&reader_timer);
-
-	return 0; // success
-
-init_module_cleanup:
-	if ((nresult = misc_deregister(&z90crypt_misc_device)))
-		PRINTK("misc_deregister failed with %d.\n", nresult);
-	else
-		PDEBUG("misc_deregister successful.\n");
-
-	return result; // failure
-}
-
-/**
- * The module termination code
- */
-static void __exit
-z90crypt_cleanup_module(void)
-{
-	int nresult;
-
-	PDEBUG("PID %d\n", PID());
-
-	remove_proc_entry("driver/z90crypt", 0);
-
-	if ((nresult = misc_deregister(&z90crypt_misc_device)))
-		PRINTK("misc_deregister failed with %d.\n", nresult);
-	else
-		PDEBUG("misc_deregister successful.\n");
-
-	/* Remove the tasks */
-	tasklet_kill(&reader_tasklet);
-	del_timer(&reader_timer);
-	del_timer(&config_timer);
-	del_timer(&cleanup_timer);
-
-	destroy_z90crypt();
-
-	PRINTKN("Unloaded.\n");
-}
-
-/**
- * Functions running under a process id
- *
- * The I/O functions:
- *     z90crypt_open
- *     z90crypt_release
- *     z90crypt_read
- *     z90crypt_write
- *     z90crypt_unlocked_ioctl
- *     z90crypt_status
- *     z90crypt_status_write
- *	 disable_card
- *	 enable_card
- *
- * Helper functions:
- *     z90crypt_rsa
- *	 z90crypt_prepare
- *	 z90crypt_send
- *	 z90crypt_process_results
- *
- */
-static int
-z90crypt_open(struct inode *inode, struct file *filp)
-{
-	struct priv_data *private_data_p;
-
-	if (quiesce_z90crypt)
-		return -EQUIESCE;
-
-	private_data_p = kzalloc(sizeof(struct priv_data), GFP_KERNEL);
-	if (!private_data_p) {
-		PRINTK("Memory allocate failed\n");
-		return -ENOMEM;
-	}
-
-	private_data_p->status = STAT_OPEN;
-	private_data_p->opener_pid = PID();
-	filp->private_data = private_data_p;
-	atomic_inc(&total_open);
-
-	return 0;
-}
-
-static int
-z90crypt_release(struct inode *inode, struct file *filp)
-{
-	struct priv_data *private_data_p = filp->private_data;
-
-	PDEBUG("PID %d (filp %p)\n", PID(), filp);
-
-	private_data_p->status = STAT_CLOSED;
-	memset(private_data_p, 0, sizeof(struct priv_data));
-	kfree(private_data_p);
-	atomic_dec(&total_open);
-
-	return 0;
-}
-
-/*
- * there are two read functions, of which compile options will choose one
- * without USE_GET_RANDOM_BYTES
- *   => read() always returns -EPERM;
- * otherwise
- *   => read() uses get_random_bytes() kernel function
- */
-#ifndef USE_GET_RANDOM_BYTES
-/**
- * z90crypt_read will not be supported beyond z90crypt 1.3.1
- */
-static ssize_t
-z90crypt_read(struct file *filp, char __user *buf, size_t count, loff_t *f_pos)
-{
-	PDEBUG("filp %p (PID %d)\n", filp, PID());
-	return -EPERM;
-}
-#else // we want to use get_random_bytes
-/**
- * read() just returns a string of random bytes.  Since we have no way
- * to generate these cryptographically, we just execute get_random_bytes
- * for the length specified.
- */
-#include <linux/random.h>
-static ssize_t
-z90crypt_read(struct file *filp, char __user *buf, size_t count, loff_t *f_pos)
-{
-	unsigned char *temp_buff;
-
-	PDEBUG("filp %p (PID %d)\n", filp, PID());
-
-	if (quiesce_z90crypt)
-		return -EQUIESCE;
-	if (count < 0) {
-		PRINTK("Requested random byte count negative: %ld\n", count);
-		return -EINVAL;
-	}
-	if (count > RESPBUFFSIZE) {
-		PDEBUG("count[%d] > RESPBUFFSIZE", count);
-		return -EINVAL;
-	}
-	if (count == 0)
-		return 0;
-	temp_buff = kmalloc(RESPBUFFSIZE, GFP_KERNEL);
-	if (!temp_buff) {
-		PRINTK("Memory allocate failed\n");
-		return -ENOMEM;
-	}
-	get_random_bytes(temp_buff, count);
-
-	if (copy_to_user(buf, temp_buff, count) != 0) {
-		kfree(temp_buff);
-		return -EFAULT;
-	}
-	kfree(temp_buff);
-	return count;
-}
-#endif
-
-/**
- * Write is is not allowed
- */
-static ssize_t
-z90crypt_write(struct file *filp, const char __user *buf, size_t count, loff_t *f_pos)
-{
-	PDEBUG("filp %p (PID %d)\n", filp, PID());
-	return -EPERM;
-}
-
-/**
- * New status functions
- */
-static inline int
-get_status_totalcount(void)
-{
-	return z90crypt.hdware_info->hdware_mask.st_count;
-}
-
-static inline int
-get_status_PCICAcount(void)
-{
-	return z90crypt.hdware_info->type_mask[PCICA].st_count;
-}
-
-static inline int
-get_status_PCICCcount(void)
-{
-	return z90crypt.hdware_info->type_mask[PCICC].st_count;
-}
-
-static inline int
-get_status_PCIXCCcount(void)
-{
-	return z90crypt.hdware_info->type_mask[PCIXCC_MCL2].st_count +
-	       z90crypt.hdware_info->type_mask[PCIXCC_MCL3].st_count;
-}
-
-static inline int
-get_status_PCIXCCMCL2count(void)
-{
-	return z90crypt.hdware_info->type_mask[PCIXCC_MCL2].st_count;
-}
-
-static inline int
-get_status_PCIXCCMCL3count(void)
-{
-	return z90crypt.hdware_info->type_mask[PCIXCC_MCL3].st_count;
-}
-
-static inline int
-get_status_CEX2Ccount(void)
-{
-	return z90crypt.hdware_info->type_mask[CEX2C].st_count;
-}
-
-static inline int
-get_status_CEX2Acount(void)
-{
-	return z90crypt.hdware_info->type_mask[CEX2A].st_count;
-}
-
-static inline int
-get_status_requestq_count(void)
-{
-	return requestq_count;
-}
-
-static inline int
-get_status_pendingq_count(void)
-{
-	return pendingq_count;
-}
-
-static inline int
-get_status_totalopen_count(void)
-{
-	return atomic_read(&total_open);
-}
-
-static inline int
-get_status_domain_index(void)
-{
-	return z90crypt.cdx;
-}
-
-static inline unsigned char *
-get_status_status_mask(unsigned char status[Z90CRYPT_NUM_APS])
-{
-	int i, ix;
-
-	memcpy(status, z90crypt.hdware_info->device_type_array,
-	       Z90CRYPT_NUM_APS);
-
-	for (i = 0; i < get_status_totalcount(); i++) {
-		ix = SHRT2LONG(i);
-		if (LONG2DEVPTR(ix)->user_disabled)
-			status[ix] = 0x0d;
-	}
-
-	return status;
-}
-
-static inline unsigned char *
-get_status_qdepth_mask(unsigned char qdepth[Z90CRYPT_NUM_APS])
-{
-	int i, ix;
-
-	memset(qdepth, 0, Z90CRYPT_NUM_APS);
-
-	for (i = 0; i < get_status_totalcount(); i++) {
-		ix = SHRT2LONG(i);
-		qdepth[ix] = LONG2DEVPTR(ix)->dev_caller_count;
-	}
-
-	return qdepth;
-}
-
-static inline unsigned int *
-get_status_perdevice_reqcnt(unsigned int reqcnt[Z90CRYPT_NUM_APS])
-{
-	int i, ix;
-
-	memset(reqcnt, 0, Z90CRYPT_NUM_APS * sizeof(int));
-
-	for (i = 0; i < get_status_totalcount(); i++) {
-		ix = SHRT2LONG(i);
-		reqcnt[ix] = LONG2DEVPTR(ix)->dev_total_req_cnt;
-	}
-
-	return reqcnt;
-}
-
-static inline void
-init_work_element(struct work_element *we_p,
-		  struct priv_data *priv_data, pid_t pid)
-{
-	int step;
-
-	we_p->requestptr = (unsigned char *)we_p + sizeof(struct work_element);
-	/* Come up with a unique id for this caller. */
-	step = atomic_inc_return(&z90crypt_step);
-	memcpy(we_p->caller_id+0, (void *) &pid, sizeof(pid));
-	memcpy(we_p->caller_id+4, (void *) &step, sizeof(step));
-	we_p->pid = pid;
-	we_p->priv_data = priv_data;
-	we_p->status[0] = STAT_DEFAULT;
-	we_p->audit[0] = 0x00;
-	we_p->audit[1] = 0x00;
-	we_p->audit[2] = 0x00;
-	we_p->resp_buff_size = 0;
-	we_p->retcode = 0;
-	we_p->devindex = -1;
-	we_p->devtype = -1;
-	atomic_set(&we_p->alarmrung, 0);
-	init_waitqueue_head(&we_p->waitq);
-	INIT_LIST_HEAD(&(we_p->liste));
-}
-
-static inline int
-allocate_work_element(struct work_element **we_pp,
-		      struct priv_data *priv_data_p, pid_t pid)
-{
-	struct work_element *we_p;
-
-	we_p = (struct work_element *) get_zeroed_page(GFP_KERNEL);
-	if (!we_p)
-		return -ENOMEM;
-	init_work_element(we_p, priv_data_p, pid);
-	*we_pp = we_p;
-	return 0;
-}
-
-static inline void
-remove_device(struct device *device_p)
-{
-	if (!device_p || (device_p->disabled != 0))
-		return;
-	device_p->disabled = 1;
-	z90crypt.hdware_info->type_mask[device_p->dev_type].disabled_count++;
-	z90crypt.hdware_info->hdware_mask.disabled_count++;
-}
-
-/**
- * Bitlength limits for each card
- *
- * There are new MCLs which allow more bitlengths. See the table for details.
- * The MCL must be applied and the newer bitlengths enabled for these to work.
- *
- * Card Type    Old limit    New limit
- * PCICA          ??-2048     same (the lower limit is less than 128 bit...)
- * PCICC         512-1024     512-2048
- * PCIXCC_MCL2   512-2048     ----- (applying any GA LIC will make an MCL3 card)
- * PCIXCC_MCL3   -----        128-2048
- * CEX2C         512-2048     128-2048
- * CEX2A          ??-2048     same (the lower limit is less than 128 bit...)
- *
- * ext_bitlens (extended bitlengths) is a global, since you should not apply an
- * MCL to just one card in a machine. We assume, at first, that all cards have
- * these capabilities.
- */
-int ext_bitlens = 1; // This is global
-#define PCIXCC_MIN_MOD_SIZE	 16	//  128 bits
-#define OLD_PCIXCC_MIN_MOD_SIZE	 64	//  512 bits
-#define PCICC_MIN_MOD_SIZE	 64	//  512 bits
-#define OLD_PCICC_MAX_MOD_SIZE	128	// 1024 bits
-#define MAX_MOD_SIZE		256	// 2048 bits
-
-static inline int
-select_device_type(int *dev_type_p, int bytelength)
-{
-	static int count = 0;
-	int PCICA_avail, PCIXCC_MCL3_avail, CEX2C_avail, CEX2A_avail,
-	    index_to_use;
-	struct status *stat;
-	if ((*dev_type_p != PCICC) && (*dev_type_p != PCICA) &&
-	    (*dev_type_p != PCIXCC_MCL2) && (*dev_type_p != PCIXCC_MCL3) &&
-	    (*dev_type_p != CEX2C) && (*dev_type_p != CEX2A) &&
-	    (*dev_type_p != ANYDEV))
-		return -1;
-	if (*dev_type_p != ANYDEV) {
-		stat = &z90crypt.hdware_info->type_mask[*dev_type_p];
-		if (stat->st_count >
-		    (stat->disabled_count + stat->user_disabled_count))
-			return 0;
-		return -1;
-	}
-
-	/**
-	 * Assumption: PCICA, PCIXCC_MCL3, CEX2C, and CEX2A are all similar in
-	 * speed.
-	 *
-	 * PCICA and CEX2A do NOT co-exist, so it would be either one or the
-	 * other present.
-	 */
-	stat = &z90crypt.hdware_info->type_mask[PCICA];
-	PCICA_avail = stat->st_count -
-			(stat->disabled_count + stat->user_disabled_count);
-	stat = &z90crypt.hdware_info->type_mask[PCIXCC_MCL3];
-	PCIXCC_MCL3_avail = stat->st_count -
-			(stat->disabled_count + stat->user_disabled_count);
-	stat = &z90crypt.hdware_info->type_mask[CEX2C];
-	CEX2C_avail = stat->st_count -
-			(stat->disabled_count + stat->user_disabled_count);
-	stat = &z90crypt.hdware_info->type_mask[CEX2A];
-	CEX2A_avail = stat->st_count -
-			(stat->disabled_count + stat->user_disabled_count);
-	if (PCICA_avail || PCIXCC_MCL3_avail || CEX2C_avail || CEX2A_avail) {
-		/**
-		 * bitlength is a factor, PCICA or CEX2A are the most capable,
-		 * even with the new MCL for PCIXCC.
-		 */
-		if ((bytelength < PCIXCC_MIN_MOD_SIZE) ||
-		    (!ext_bitlens && (bytelength < OLD_PCIXCC_MIN_MOD_SIZE))) {
-			if (PCICA_avail) {
-				*dev_type_p = PCICA;
-				return 0;
-			}
-			if (CEX2A_avail) {
-				*dev_type_p = CEX2A;
-				return 0;
-			}
-			return -1;
-		}
-
-		index_to_use = count % (PCICA_avail + PCIXCC_MCL3_avail +
-					CEX2C_avail + CEX2A_avail);
-		if (index_to_use < PCICA_avail)
-			*dev_type_p = PCICA;
-		else if (index_to_use < (PCICA_avail + PCIXCC_MCL3_avail))
-			*dev_type_p = PCIXCC_MCL3;
-		else if (index_to_use < (PCICA_avail + PCIXCC_MCL3_avail +
-					 CEX2C_avail))
-			*dev_type_p = CEX2C;
-		else
-			*dev_type_p = CEX2A;
-		count++;
-		return 0;
-	}
-
-	/* Less than OLD_PCIXCC_MIN_MOD_SIZE cannot go to a PCIXCC_MCL2 */
-	if (bytelength < OLD_PCIXCC_MIN_MOD_SIZE)
-		return -1;
-	stat = &z90crypt.hdware_info->type_mask[PCIXCC_MCL2];
-	if (stat->st_count >
-	    (stat->disabled_count + stat->user_disabled_count)) {
-		*dev_type_p = PCIXCC_MCL2;
-		return 0;
-	}
-
-	/**
-	 * Less than PCICC_MIN_MOD_SIZE or more than OLD_PCICC_MAX_MOD_SIZE
-	 * (if we don't have the MCL applied and the newer bitlengths enabled)
-	 * cannot go to a PCICC
-	 */
-	if ((bytelength < PCICC_MIN_MOD_SIZE) ||
-	    (!ext_bitlens && (bytelength > OLD_PCICC_MAX_MOD_SIZE))) {
-		return -1;
-	}
-	stat = &z90crypt.hdware_info->type_mask[PCICC];
-	if (stat->st_count >
-	    (stat->disabled_count + stat->user_disabled_count)) {
-		*dev_type_p = PCICC;
-		return 0;
-	}
-
-	return -1;
-}
-
-/**
- * Try the selected number, then the selected type (can be ANYDEV)
- */
-static inline int
-select_device(int *dev_type_p, int *device_nr_p, int bytelength)
-{
-	int i, indx, devTp, low_count, low_indx;
-	struct device_x *index_p;
-	struct device *dev_ptr;
-
-	PDEBUG("device type = %d, index = %d\n", *dev_type_p, *device_nr_p);
-	if ((*device_nr_p >= 0) && (*device_nr_p < Z90CRYPT_NUM_DEVS)) {
-		PDEBUG("trying index = %d\n", *device_nr_p);
-		dev_ptr = z90crypt.device_p[*device_nr_p];
-
-		if (dev_ptr &&
-		    (dev_ptr->dev_stat != DEV_GONE) &&
-		    (dev_ptr->disabled == 0) &&
-		    (dev_ptr->user_disabled == 0)) {
-			PDEBUG("selected by number, index = %d\n",
-			       *device_nr_p);
-			*dev_type_p = dev_ptr->dev_type;
-			return *device_nr_p;
-		}
-	}
-	*device_nr_p = -1;
-	PDEBUG("trying type = %d\n", *dev_type_p);
-	devTp = *dev_type_p;
-	if (select_device_type(&devTp, bytelength) == -1) {
-		PDEBUG("failed to select by type\n");
-		return -1;
-	}
-	PDEBUG("selected type = %d\n", devTp);
-	index_p = &z90crypt.hdware_info->type_x_addr[devTp];
-	low_count = 0x0000FFFF;
-	low_indx = -1;
-	for (i = 0; i < z90crypt.hdware_info->type_mask[devTp].st_count; i++) {
-		indx = index_p->device_index[i];
-		dev_ptr = z90crypt.device_p[indx];
-		if (dev_ptr &&
-		    (dev_ptr->dev_stat != DEV_GONE) &&
-		    (dev_ptr->disabled == 0) &&
-		    (dev_ptr->user_disabled == 0) &&
-		    (devTp == dev_ptr->dev_type) &&
-		    (low_count > dev_ptr->dev_caller_count)) {
-			low_count = dev_ptr->dev_caller_count;
-			low_indx = indx;
-		}
-	}
-	*device_nr_p = low_indx;
-	return low_indx;
-}
-
-static inline int
-send_to_crypto_device(struct work_element *we_p)
-{
-	struct caller *caller_p;
-	struct device *device_p;
-	int dev_nr;
-	int bytelen = ((struct ica_rsa_modexpo *)we_p->buffer)->inputdatalength;
-
-	if (!we_p->requestptr)
-		return SEN_FATAL_ERROR;
-	caller_p = (struct caller *)we_p->requestptr;
-	dev_nr = we_p->devindex;
-	if (select_device(&we_p->devtype, &dev_nr, bytelen) == -1) {
-		if (z90crypt.hdware_info->hdware_mask.st_count != 0)
-			return SEN_RETRY;
-		else
-			return SEN_NOT_AVAIL;
-	}
-	we_p->devindex = dev_nr;
-	device_p = z90crypt.device_p[dev_nr];
-	if (!device_p)
-		return SEN_NOT_AVAIL;
-	if (device_p->dev_type != we_p->devtype)
-		return SEN_RETRY;
-	if (device_p->dev_caller_count >= device_p->dev_q_depth)
-		return SEN_QUEUE_FULL;
-	PDEBUG("device number prior to send: %d\n", dev_nr);
-	switch (send_to_AP(dev_nr, z90crypt.cdx,
-			   caller_p->caller_dev_dep_req_l,
-			   caller_p->caller_dev_dep_req_p)) {
-	case DEV_SEN_EXCEPTION:
-		PRINTKC("Exception during send to device %d\n", dev_nr);
-		z90crypt.terminating = 1;
-		return SEN_FATAL_ERROR;
-	case DEV_GONE:
-		PRINTK("Device %d not available\n", dev_nr);
-		remove_device(device_p);
-		return SEN_NOT_AVAIL;
-	case DEV_EMPTY:
-		return SEN_NOT_AVAIL;
-	case DEV_NO_WORK:
-		return SEN_FATAL_ERROR;
-	case DEV_BAD_MESSAGE:
-		return SEN_USER_ERROR;
-	case DEV_QUEUE_FULL:
-		return SEN_QUEUE_FULL;
-	default:
-	case DEV_ONLINE:
-		break;
-	}
-	list_add_tail(&(caller_p->caller_liste), &(device_p->dev_caller_list));
-	device_p->dev_caller_count++;
-	return 0;
-}
-
-/**
- * Send puts the user's work on one of two queues:
- *   the pending queue if the send was successful
- *   the request queue if the send failed because device full or busy
- */
-static inline int
-z90crypt_send(struct work_element *we_p, const char *buf)
-{
-	int rv;
-
-	PDEBUG("PID %d\n", PID());
-
-	if (CHK_RDWRMASK(we_p->status[0]) != STAT_NOWORK) {
-		PDEBUG("PID %d tried to send more work but has outstanding "
-		       "work.\n", PID());
-		return -EWORKPEND;
-	}
-	we_p->devindex = -1; // Reset device number
-	spin_lock_irq(&queuespinlock);
-	rv = send_to_crypto_device(we_p);
-	switch (rv) {
-	case 0:
-		we_p->requestsent = jiffies;
-		we_p->audit[0] |= FP_SENT;
-		list_add_tail(&we_p->liste, &pending_list);
-		++pendingq_count;
-		we_p->audit[0] |= FP_PENDING;
-		break;
-	case SEN_BUSY:
-	case SEN_QUEUE_FULL:
-		rv = 0;
-		we_p->devindex = -1; // any device will do
-		we_p->requestsent = jiffies;
-		list_add_tail(&we_p->liste, &request_list);
-		++requestq_count;
-		we_p->audit[0] |= FP_REQUEST;
-		break;
-	case SEN_RETRY:
-		rv = -ERESTARTSYS;
-		break;
-	case SEN_NOT_AVAIL:
-		PRINTK("*** No devices available.\n");
-		rv = we_p->retcode = -ENODEV;
-		we_p->status[0] |= STAT_FAILED;
-		break;
-	case REC_OPERAND_INV:
-	case REC_OPERAND_SIZE:
-	case REC_EVEN_MOD:
-	case REC_INVALID_PAD:
-		rv = we_p->retcode = -EINVAL;
-		we_p->status[0] |= STAT_FAILED;
-		break;
-	default:
-		we_p->retcode = rv;
-		we_p->status[0] |= STAT_FAILED;
-		break;
-	}
-	if (rv != -ERESTARTSYS)
-		SET_RDWRMASK(we_p->status[0], STAT_WRITTEN);
-	spin_unlock_irq(&queuespinlock);
-	if (rv == 0)
-		tasklet_schedule(&reader_tasklet);
-	return rv;
-}
-
-/**
- * process_results copies the user's work from kernel space.
- */
-static inline int
-z90crypt_process_results(struct work_element *we_p, char __user *buf)
-{
-	int rv;
-
-	PDEBUG("we_p %p (PID %d)\n", we_p, PID());
-
-	LONG2DEVPTR(we_p->devindex)->dev_total_req_cnt++;
-	SET_RDWRMASK(we_p->status[0], STAT_READPEND);
-
-	rv = 0;
-	if (!we_p->buffer) {
-		PRINTK("we_p %p PID %d in STAT_READPEND: buffer NULL.\n",
-			we_p, PID());
-		rv = -ENOBUFF;
-	}
-
-	if (!rv)
-		if ((rv = copy_to_user(buf, we_p->buffer, we_p->buff_size))) {
-			PDEBUG("copy_to_user failed: rv = %d\n", rv);
-			rv = -EFAULT;
-		}
-
-	if (!rv)
-		rv = we_p->retcode;
-	if (!rv)
-		if (we_p->resp_buff_size
-		    &&	copy_to_user(we_p->resp_addr, we_p->resp_buff,
-				     we_p->resp_buff_size))
-			rv = -EFAULT;
-
-	SET_RDWRMASK(we_p->status[0], STAT_NOWORK);
-	return rv;
-}
-
-static unsigned char NULL_psmid[8] =
-{0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00};
-
-/**
- * Used in device configuration functions
- */
-#define MAX_RESET 90
-
-/**
- * This is used only for PCICC support
- */
-static inline int
-is_PKCS11_padded(unsigned char *buffer, int length)
-{
-	int i;
-	if ((buffer[0] != 0x00) || (buffer[1] != 0x01))
-		return 0;
-	for (i = 2; i < length; i++)
-		if (buffer[i] != 0xFF)
-			break;
-	if ((i < 10) || (i == length))
-		return 0;
-	if (buffer[i] != 0x00)
-		return 0;
-	return 1;
-}
-
-/**
- * This is used only for PCICC support
- */
-static inline int
-is_PKCS12_padded(unsigned char *buffer, int length)
-{
-	int i;
-	if ((buffer[0] != 0x00) || (buffer[1] != 0x02))
-		return 0;
-	for (i = 2; i < length; i++)
-		if (buffer[i] == 0x00)
-			break;
-	if ((i < 10) || (i == length))
-		return 0;
-	if (buffer[i] != 0x00)
-		return 0;
-	return 1;
-}
-
-/**
- * builds struct caller and converts message from generic format to
- * device-dependent format
- * func is ICARSAMODEXPO or ICARSACRT
- * function is PCI_FUNC_KEY_ENCRYPT or PCI_FUNC_KEY_DECRYPT
- */
-static inline int
-build_caller(struct work_element *we_p, short function)
-{
-	int rv;
-	struct caller *caller_p = (struct caller *)we_p->requestptr;
-
-	if ((we_p->devtype != PCICC) && (we_p->devtype != PCICA) &&
-	    (we_p->devtype != PCIXCC_MCL2) && (we_p->devtype != PCIXCC_MCL3) &&
-	    (we_p->devtype != CEX2C) && (we_p->devtype != CEX2A))
-		return SEN_NOT_AVAIL;
-
-	memcpy(caller_p->caller_id, we_p->caller_id,
-	       sizeof(caller_p->caller_id));
-	caller_p->caller_dev_dep_req_p = caller_p->caller_dev_dep_req;
-	caller_p->caller_dev_dep_req_l = MAX_RESPONSE_SIZE;
-	caller_p->caller_buf_p = we_p->buffer;
-	INIT_LIST_HEAD(&(caller_p->caller_liste));
-
-	rv = convert_request(we_p->buffer, we_p->funccode, function,
-			     z90crypt.cdx, we_p->devtype,
-			     &caller_p->caller_dev_dep_req_l,
-			     caller_p->caller_dev_dep_req_p);
-	if (rv) {
-		if (rv == SEN_NOT_AVAIL)
-			PDEBUG("request can't be processed on hdwr avail\n");
-		else
-			PRINTK("Error from convert_request: %d\n", rv);
-	}
-	else
-		memcpy(&(caller_p->caller_dev_dep_req_p[4]), we_p->caller_id,8);
-	return rv;
-}
-
-static inline void
-unbuild_caller(struct device *device_p, struct caller *caller_p)
-{
-	if (!caller_p)
-		return;
-	if (caller_p->caller_liste.next && caller_p->caller_liste.prev)
-		if (!list_empty(&caller_p->caller_liste)) {
-			list_del_init(&caller_p->caller_liste);
-			device_p->dev_caller_count--;
-		}
-	memset(caller_p->caller_id, 0, sizeof(caller_p->caller_id));
-}
-
-static inline int
-get_crypto_request_buffer(struct work_element *we_p)
-{
-	struct ica_rsa_modexpo *mex_p;
-	struct ica_rsa_modexpo_crt *crt_p;
-	unsigned char *temp_buffer;
-	short function;
-	int rv;
-
-	mex_p =	(struct ica_rsa_modexpo *) we_p->buffer;
-	crt_p = (struct ica_rsa_modexpo_crt *) we_p->buffer;
-
-	PDEBUG("device type input = %d\n", we_p->devtype);
-
-	if (z90crypt.terminating)
-		return REC_NO_RESPONSE;
-	if (memcmp(we_p->caller_id, NULL_psmid, 8) == 0) {
-		PRINTK("psmid zeroes\n");
-		return SEN_FATAL_ERROR;
-	}
-	if (!we_p->buffer) {
-		PRINTK("buffer pointer NULL\n");
-		return SEN_USER_ERROR;
-	}
-	if (!we_p->requestptr) {
-		PRINTK("caller pointer NULL\n");
-		return SEN_USER_ERROR;
-	}
-
-	if ((we_p->devtype != PCICA) && (we_p->devtype != PCICC) &&
-	    (we_p->devtype != PCIXCC_MCL2) && (we_p->devtype != PCIXCC_MCL3) &&
-	    (we_p->devtype != CEX2C) && (we_p->devtype != CEX2A) &&
-	    (we_p->devtype != ANYDEV)) {
-		PRINTK("invalid device type\n");
-		return SEN_USER_ERROR;
-	}
-
-	if ((mex_p->inputdatalength < 1) ||
-	    (mex_p->inputdatalength > MAX_MOD_SIZE)) {
-		PRINTK("inputdatalength[%d] is not valid\n",
-		       mex_p->inputdatalength);
-		return SEN_USER_ERROR;
-	}
-
-	if (mex_p->outputdatalength < mex_p->inputdatalength) {
-		PRINTK("outputdatalength[%d] < inputdatalength[%d]\n",
-		       mex_p->outputdatalength, mex_p->inputdatalength);
-		return SEN_USER_ERROR;
-	}
-
-	if (!mex_p->inputdata || !mex_p->outputdata) {
-		PRINTK("inputdata[%p] or outputdata[%p] is NULL\n",
-		       mex_p->outputdata, mex_p->inputdata);
-		return SEN_USER_ERROR;
-	}
-
-	/**
-	 * As long as outputdatalength is big enough, we can set the
-	 * outputdatalength equal to the inputdatalength, since that is the
-	 * number of bytes we will copy in any case
-	 */
-	mex_p->outputdatalength = mex_p->inputdatalength;
-
-	rv = 0;
-	switch (we_p->funccode) {
-	case ICARSAMODEXPO:
-		if (!mex_p->b_key || !mex_p->n_modulus)
-			rv = SEN_USER_ERROR;
-		break;
-	case ICARSACRT:
-		if (!IS_EVEN(crt_p->inputdatalength)) {
-			PRINTK("inputdatalength[%d] is odd, CRT form\n",
-			       crt_p->inputdatalength);
-			rv = SEN_USER_ERROR;
-			break;
-		}
-		if (!crt_p->bp_key ||
-		    !crt_p->bq_key ||
-		    !crt_p->np_prime ||
-		    !crt_p->nq_prime ||
-		    !crt_p->u_mult_inv) {
-			PRINTK("CRT form, bad data: %p/%p/%p/%p/%p\n",
-			       crt_p->bp_key, crt_p->bq_key,
-			       crt_p->np_prime, crt_p->nq_prime,
-			       crt_p->u_mult_inv);
-			rv = SEN_USER_ERROR;
-		}
-		break;
-	default:
-		PRINTK("bad func = %d\n", we_p->funccode);
-		rv = SEN_USER_ERROR;
-		break;
-	}
-	if (rv != 0)
-		return rv;
-
-	if (select_device_type(&we_p->devtype, mex_p->inputdatalength) < 0)
-		return SEN_NOT_AVAIL;
-
-	temp_buffer = (unsigned char *)we_p + sizeof(struct work_element) +
-		      sizeof(struct caller);
-	if (copy_from_user(temp_buffer, mex_p->inputdata,
-			   mex_p->inputdatalength) != 0)
-		return SEN_RELEASED;
-
-	function = PCI_FUNC_KEY_ENCRYPT;
-	switch (we_p->devtype) {
-	/* PCICA and CEX2A do everything with a simple RSA mod-expo operation */
-	case PCICA:
-	case CEX2A:
-		function = PCI_FUNC_KEY_ENCRYPT;
-		break;
-	/**
-	 * PCIXCC_MCL2 does all Mod-Expo form with a simple RSA mod-expo
-	 * operation, and all CRT forms with a PKCS-1.2 format decrypt.
-	 * PCIXCC_MCL3 and CEX2C do all Mod-Expo and CRT forms with a simple RSA
-	 * mod-expo operation
-	 */
-	case PCIXCC_MCL2:
-		if (we_p->funccode == ICARSAMODEXPO)
-			function = PCI_FUNC_KEY_ENCRYPT;
-		else
-			function = PCI_FUNC_KEY_DECRYPT;
-		break;
-	case PCIXCC_MCL3:
-	case CEX2C:
-		if (we_p->funccode == ICARSAMODEXPO)
-			function = PCI_FUNC_KEY_ENCRYPT;
-		else
-			function = PCI_FUNC_KEY_DECRYPT;
-		break;
-	/**
-	 * PCICC does everything as a PKCS-1.2 format request
-	 */
-	case PCICC:
-		/* PCICC cannot handle input that is is PKCS#1.1 padded */
-		if (is_PKCS11_padded(temp_buffer, mex_p->inputdatalength)) {
-			return SEN_NOT_AVAIL;
-		}
-		if (we_p->funccode == ICARSAMODEXPO) {
-			if (is_PKCS12_padded(temp_buffer,
-					     mex_p->inputdatalength))
-				function = PCI_FUNC_KEY_ENCRYPT;
-			else
-				function = PCI_FUNC_KEY_DECRYPT;
-		} else
-			/* all CRT forms are decrypts */
-			function = PCI_FUNC_KEY_DECRYPT;
-		break;
-	}
-	PDEBUG("function: %04x\n", function);
-	rv = build_caller(we_p, function);
-	PDEBUG("rv from build_caller = %d\n", rv);
-	return rv;
-}
-
-static inline int
-z90crypt_prepare(struct work_element *we_p, unsigned int funccode,
-		 const char __user *buffer)
-{
-	int rv;
-
-	we_p->devindex = -1;
-	if (funccode == ICARSAMODEXPO)
-		we_p->buff_size = sizeof(struct ica_rsa_modexpo);
-	else
-		we_p->buff_size = sizeof(struct ica_rsa_modexpo_crt);
-
-	if (copy_from_user(we_p->buffer, buffer, we_p->buff_size))
-		return -EFAULT;
-
-	we_p->audit[0] |= FP_COPYFROM;
-	SET_RDWRMASK(we_p->status[0], STAT_WRITTEN);
-	we_p->funccode = funccode;
-	we_p->devtype = -1;
-	we_p->audit[0] |= FP_BUFFREQ;
-	rv = get_crypto_request_buffer(we_p);
-	switch (rv) {
-	case 0:
-		we_p->audit[0] |= FP_BUFFGOT;
-		break;
-	case SEN_USER_ERROR:
-		rv = -EINVAL;
-		break;
-	case SEN_QUEUE_FULL:
-		rv = 0;
-		break;
-	case SEN_RELEASED:
-		rv = -EFAULT;
-		break;
-	case REC_NO_RESPONSE:
-		rv = -ENODEV;
-		break;
-	case SEN_NOT_AVAIL:
-	case EGETBUFF:
-		rv = -EGETBUFF;
-		break;
-	default:
-		PRINTK("rv = %d\n", rv);
-		rv = -EGETBUFF;
-		break;
-	}
-	if (CHK_RDWRMASK(we_p->status[0]) == STAT_WRITTEN)
-		SET_RDWRMASK(we_p->status[0], STAT_DEFAULT);
-	return rv;
-}
-
-static inline void
-purge_work_element(struct work_element *we_p)
-{
-	struct list_head *lptr;
-
-	spin_lock_irq(&queuespinlock);
-	list_for_each(lptr, &request_list) {
-		if (lptr == &we_p->liste) {
-			list_del_init(lptr);
-			requestq_count--;
-			break;
-		}
-	}
-	list_for_each(lptr, &pending_list) {
-		if (lptr == &we_p->liste) {
-			list_del_init(lptr);
-			pendingq_count--;
-			break;
-		}
-	}
-	spin_unlock_irq(&queuespinlock);
-}
-
-/**
- * Build the request and send it.
- */
-static inline int
-z90crypt_rsa(struct priv_data *private_data_p, pid_t pid,
-	     unsigned int cmd, unsigned long arg)
-{
-	struct work_element *we_p;
-	int rv;
-
-	if ((rv = allocate_work_element(&we_p, private_data_p, pid))) {
-		PDEBUG("PID %d: allocate_work_element returned ENOMEM\n", pid);
-		return rv;
-	}
-	if ((rv = z90crypt_prepare(we_p, cmd, (const char __user *)arg)))
-		PDEBUG("PID %d: rv = %d from z90crypt_prepare\n", pid, rv);
-	if (!rv)
-		if ((rv = z90crypt_send(we_p, (const char *)arg)))
-			PDEBUG("PID %d: rv %d from z90crypt_send.\n", pid, rv);
-	if (!rv) {
-		we_p->audit[0] |= FP_ASLEEP;
-		wait_event(we_p->waitq, atomic_read(&we_p->alarmrung));
-		we_p->audit[0] |= FP_AWAKE;
-		rv = we_p->retcode;
-	}
-	if (!rv)
-		rv = z90crypt_process_results(we_p, (char __user *)arg);
-
-	if ((we_p->status[0] & STAT_FAILED)) {
-		switch (rv) {
-		/**
-		 * EINVAL *after* receive is almost always a padding error or
-		 * length error issued by a coprocessor (not an accelerator).
-		 * We convert this return value to -EGETBUFF which should
-		 * trigger a fallback to software.
-		 */
-		case -EINVAL:
-			if ((we_p->devtype != PCICA) &&
-			    (we_p->devtype != CEX2A))
-				rv = -EGETBUFF;
-			break;
-		case -ETIMEOUT:
-			if (z90crypt.mask.st_count > 0)
-				rv = -ERESTARTSYS; // retry with another
-			else
-				rv = -ENODEV; // no cards left
-		/* fall through to clean up request queue */
-		case -ERESTARTSYS:
-		case -ERELEASED:
-			switch (CHK_RDWRMASK(we_p->status[0])) {
-			case STAT_WRITTEN:
-				purge_work_element(we_p);
-				break;
-			case STAT_READPEND:
-			case STAT_NOWORK:
-			default:
-				break;
-			}
-			break;
-		default:
-			we_p->status[0] ^= STAT_FAILED;
-			break;
-		}
-	}
-	free_page((long)we_p);
-	return rv;
-}
-
-/**
- * This function is a little long, but it's really just one large switch
- * statement.
- */
-static long
-z90crypt_unlocked_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
-{
-	struct priv_data *private_data_p = filp->private_data;
-	unsigned char *status;
-	unsigned char *qdepth;
-	unsigned int *reqcnt;
-	struct ica_z90_status *pstat;
-	int ret, i, loopLim, tempstat;
-	static int deprecated_msg_count1 = 0;
-	static int deprecated_msg_count2 = 0;
-
-	PDEBUG("filp %p (PID %d), cmd 0x%08X\n", filp, PID(), cmd);
-	PDEBUG("cmd 0x%08X: dir %s, size 0x%04X, type 0x%02X, nr 0x%02X\n",
-		cmd,
-		!_IOC_DIR(cmd) ? "NO"
-		: ((_IOC_DIR(cmd) == (_IOC_READ|_IOC_WRITE)) ? "RW"
-		: ((_IOC_DIR(cmd) == _IOC_READ) ? "RD"
-		: "WR")),
-		_IOC_SIZE(cmd), _IOC_TYPE(cmd), _IOC_NR(cmd));
-
-	if (_IOC_TYPE(cmd) != Z90_IOCTL_MAGIC) {
-		PRINTK("cmd 0x%08X contains bad magic\n", cmd);
-		return -ENOTTY;
-	}
-
-	ret = 0;
-	switch (cmd) {
-	case ICARSAMODEXPO:
-	case ICARSACRT:
-		if (quiesce_z90crypt) {
-			ret = -EQUIESCE;
-			break;
-		}
-		ret = -ENODEV; // Default if no devices
-		loopLim = z90crypt.hdware_info->hdware_mask.st_count -
-			(z90crypt.hdware_info->hdware_mask.disabled_count +
-			 z90crypt.hdware_info->hdware_mask.user_disabled_count);
-		for (i = 0; i < loopLim; i++) {
-			ret = z90crypt_rsa(private_data_p, PID(), cmd, arg);
-			if (ret != -ERESTARTSYS)
-				break;
-		}
-		if (ret == -ERESTARTSYS)
-			ret = -ENODEV;
-		break;
-
-	case Z90STAT_TOTALCOUNT:
-		tempstat = get_status_totalcount();
-		if (copy_to_user((int __user *)arg, &tempstat,sizeof(int)) != 0)
-			ret = -EFAULT;
-		break;
-
-	case Z90STAT_PCICACOUNT:
-		tempstat = get_status_PCICAcount();
-		if (copy_to_user((int __user *)arg, &tempstat, sizeof(int)) != 0)
-			ret = -EFAULT;
-		break;
-
-	case Z90STAT_PCICCCOUNT:
-		tempstat = get_status_PCICCcount();
-		if (copy_to_user((int __user *)arg, &tempstat, sizeof(int)) != 0)
-			ret = -EFAULT;
-		break;
-
-	case Z90STAT_PCIXCCMCL2COUNT:
-		tempstat = get_status_PCIXCCMCL2count();
-		if (copy_to_user((int __user *)arg, &tempstat, sizeof(int)) != 0)
-			ret = -EFAULT;
-		break;
-
-	case Z90STAT_PCIXCCMCL3COUNT:
-		tempstat = get_status_PCIXCCMCL3count();
-		if (copy_to_user((int __user *)arg, &tempstat, sizeof(int)) != 0)
-			ret = -EFAULT;
-		break;
-
-	case Z90STAT_CEX2CCOUNT:
-		tempstat = get_status_CEX2Ccount();
-		if (copy_to_user((int __user *)arg, &tempstat, sizeof(int)) != 0)
-			ret = -EFAULT;
-		break;
-
-	case Z90STAT_CEX2ACOUNT:
-		tempstat = get_status_CEX2Acount();
-		if (copy_to_user((int __user *)arg, &tempstat, sizeof(int)) != 0)
-			ret = -EFAULT;
-		break;
-
-	case Z90STAT_REQUESTQ_COUNT:
-		tempstat = get_status_requestq_count();
-		if (copy_to_user((int __user *)arg, &tempstat, sizeof(int)) != 0)
-			ret = -EFAULT;
-		break;
-
-	case Z90STAT_PENDINGQ_COUNT:
-		tempstat = get_status_pendingq_count();
-		if (copy_to_user((int __user *)arg, &tempstat, sizeof(int)) != 0)
-			ret = -EFAULT;
-		break;
-
-	case Z90STAT_TOTALOPEN_COUNT:
-		tempstat = get_status_totalopen_count();
-		if (copy_to_user((int __user *)arg, &tempstat, sizeof(int)) != 0)
-			ret = -EFAULT;
-		break;
-
-	case Z90STAT_DOMAIN_INDEX:
-		tempstat = get_status_domain_index();
-		if (copy_to_user((int __user *)arg, &tempstat, sizeof(int)) != 0)
-			ret = -EFAULT;
-		break;
-
-	case Z90STAT_STATUS_MASK:
-		status = kmalloc(Z90CRYPT_NUM_APS, GFP_KERNEL);
-		if (!status) {
-			PRINTK("kmalloc for status failed!\n");
-			ret = -ENOMEM;
-			break;
-		}
-		get_status_status_mask(status);
-		if (copy_to_user((char __user *) arg, status, Z90CRYPT_NUM_APS)
-									!= 0)
-			ret = -EFAULT;
-		kfree(status);
-		break;
-
-	case Z90STAT_QDEPTH_MASK:
-		qdepth = kmalloc(Z90CRYPT_NUM_APS, GFP_KERNEL);
-		if (!qdepth) {
-			PRINTK("kmalloc for qdepth failed!\n");
-			ret = -ENOMEM;
-			break;
-		}
-		get_status_qdepth_mask(qdepth);
-		if (copy_to_user((char __user *) arg, qdepth, Z90CRYPT_NUM_APS) != 0)
-			ret = -EFAULT;
-		kfree(qdepth);
-		break;
-
-	case Z90STAT_PERDEV_REQCNT:
-		reqcnt = kmalloc(sizeof(int) * Z90CRYPT_NUM_APS, GFP_KERNEL);
-		if (!reqcnt) {
-			PRINTK("kmalloc for reqcnt failed!\n");
-			ret = -ENOMEM;
-			break;
-		}
-		get_status_perdevice_reqcnt(reqcnt);
-		if (copy_to_user((char __user *) arg, reqcnt,
-				 Z90CRYPT_NUM_APS * sizeof(int)) != 0)
-			ret = -EFAULT;
-		kfree(reqcnt);
-		break;
-
-		/* THIS IS DEPRECATED.	USE THE NEW STATUS CALLS */
-	case ICAZ90STATUS:
-		if (deprecated_msg_count1 < 20) {
-			PRINTK("deprecated call to ioctl (ICAZ90STATUS)!\n");
-			deprecated_msg_count1++;
-			if (deprecated_msg_count1 == 20)
-				PRINTK("No longer issuing messages related to "
-				       "deprecated call to ICAZ90STATUS.\n");
-		}
-
-		pstat = kmalloc(sizeof(struct ica_z90_status), GFP_KERNEL);
-		if (!pstat) {
-			PRINTK("kmalloc for pstat failed!\n");
-			ret = -ENOMEM;
-			break;
-		}
-
-		pstat->totalcount	 = get_status_totalcount();
-		pstat->leedslitecount	 = get_status_PCICAcount();
-		pstat->leeds2count	 = get_status_PCICCcount();
-		pstat->requestqWaitCount = get_status_requestq_count();
-		pstat->pendingqWaitCount = get_status_pendingq_count();
-		pstat->totalOpenCount	 = get_status_totalopen_count();
-		pstat->cryptoDomain	 = get_status_domain_index();
-		get_status_status_mask(pstat->status);
-		get_status_qdepth_mask(pstat->qdepth);
-
-		if (copy_to_user((struct ica_z90_status __user *) arg, pstat,
-				 sizeof(struct ica_z90_status)) != 0)
-			ret = -EFAULT;
-		kfree(pstat);
-		break;
-
-		/* THIS IS DEPRECATED.	USE THE NEW STATUS CALLS */
-	case Z90STAT_PCIXCCCOUNT:
-		if (deprecated_msg_count2 < 20) {
-			PRINTK("deprecated ioctl (Z90STAT_PCIXCCCOUNT)!\n");
-			deprecated_msg_count2++;
-			if (deprecated_msg_count2 == 20)
-				PRINTK("No longer issuing messages about depre"
-				       "cated ioctl Z90STAT_PCIXCCCOUNT.\n");
-		}
-
-		tempstat = get_status_PCIXCCcount();
-		if (copy_to_user((int *)arg, &tempstat, sizeof(int)) != 0)
-			ret = -EFAULT;
-		break;
-
-	case Z90QUIESCE:
-		if (current->euid != 0) {
-			PRINTK("QUIESCE fails: euid %d\n",
-			       current->euid);
-			ret = -EACCES;
-		} else {
-			PRINTK("QUIESCE device from PID %d\n", PID());
-			quiesce_z90crypt = 1;
-		}
-		break;
-
-	default:
-		/* user passed an invalid IOCTL number */
-		PDEBUG("cmd 0x%08X contains invalid ioctl code\n", cmd);
-		ret = -ENOTTY;
-		break;
-	}
-
-	return ret;
-}
-
-static inline int
-sprintcl(unsigned char *outaddr, unsigned char *addr, unsigned int len)
-{
-	int hl, i;
-
-	hl = 0;
-	for (i = 0; i < len; i++)
-		hl += sprintf(outaddr+hl, "%01x", (unsigned int) addr[i]);
-	hl += sprintf(outaddr+hl, " ");
-
-	return hl;
-}
-
-static inline int
-sprintrw(unsigned char *outaddr, unsigned char *addr, unsigned int len)
-{
-	int hl, inl, c, cx;
-
-	hl = sprintf(outaddr, "	   ");
-	inl = 0;
-	for (c = 0; c < (len / 16); c++) {
-		hl += sprintcl(outaddr+hl, addr+inl, 16);
-		inl += 16;
-	}
-
-	cx = len%16;
-	if (cx) {
-		hl += sprintcl(outaddr+hl, addr+inl, cx);
-		inl += cx;
-	}
-
-	hl += sprintf(outaddr+hl, "\n");
-
-	return hl;
-}
-
-static inline int
-sprinthx(unsigned char *title, unsigned char *outaddr,
-	 unsigned char *addr, unsigned int len)
-{
-	int hl, inl, r, rx;
-
-	hl = sprintf(outaddr, "\n%s\n", title);
-	inl = 0;
-	for (r = 0; r < (len / 64); r++) {
-		hl += sprintrw(outaddr+hl, addr+inl, 64);
-		inl += 64;
-	}
-	rx = len % 64;
-	if (rx) {
-		hl += sprintrw(outaddr+hl, addr+inl, rx);
-		inl += rx;
-	}
-
-	hl += sprintf(outaddr+hl, "\n");
-
-	return hl;
-}
-
-static inline int
-sprinthx4(unsigned char *title, unsigned char *outaddr,
-	  unsigned int *array, unsigned int len)
-{
-	int hl, r;
-
-	hl = sprintf(outaddr, "\n%s\n", title);
-
-	for (r = 0; r < len; r++) {
-		if ((r % 8) == 0)
-			hl += sprintf(outaddr+hl, "    ");
-		hl += sprintf(outaddr+hl, "%08X ", array[r]);
-		if ((r % 8) == 7)
-			hl += sprintf(outaddr+hl, "\n");
-	}
-
-	hl += sprintf(outaddr+hl, "\n");
-
-	return hl;
-}
-
-static int
-z90crypt_status(char *resp_buff, char **start, off_t offset,
-		int count, int *eof, void *data)
-{
-	unsigned char *workarea;
-	int len;
-
-	/* resp_buff is a page. Use the right half for a work area */
-	workarea = resp_buff+2000;
-	len = 0;
-	len += sprintf(resp_buff+len, "\nz90crypt version: %d.%d.%d\n",
-		z90crypt_VERSION, z90crypt_RELEASE, z90crypt_VARIANT);
-	len += sprintf(resp_buff+len, "Cryptographic domain: %d\n",
-		get_status_domain_index());
-	len += sprintf(resp_buff+len, "Total device count: %d\n",
-		get_status_totalcount());
-	len += sprintf(resp_buff+len, "PCICA count: %d\n",
-		get_status_PCICAcount());
-	len += sprintf(resp_buff+len, "PCICC count: %d\n",
-		get_status_PCICCcount());
-	len += sprintf(resp_buff+len, "PCIXCC MCL2 count: %d\n",
-		get_status_PCIXCCMCL2count());
-	len += sprintf(resp_buff+len, "PCIXCC MCL3 count: %d\n",
-		get_status_PCIXCCMCL3count());
-	len += sprintf(resp_buff+len, "CEX2C count: %d\n",
-		get_status_CEX2Ccount());
-	len += sprintf(resp_buff+len, "CEX2A count: %d\n",
-		get_status_CEX2Acount());
-	len += sprintf(resp_buff+len, "requestq count: %d\n",
-		get_status_requestq_count());
-	len += sprintf(resp_buff+len, "pendingq count: %d\n",
-		get_status_pendingq_count());
-	len += sprintf(resp_buff+len, "Total open handles: %d\n\n",
-		get_status_totalopen_count());
-	len += sprinthx(
-		"Online devices: 1=PCICA 2=PCICC 3=PCIXCC(MCL2) "
-		"4=PCIXCC(MCL3) 5=CEX2C 6=CEX2A",
-		resp_buff+len,
-		get_status_status_mask(workarea),
-		Z90CRYPT_NUM_APS);
-	len += sprinthx("Waiting work element counts",
-		resp_buff+len,
-		get_status_qdepth_mask(workarea),
-		Z90CRYPT_NUM_APS);
-	len += sprinthx4(
-		"Per-device successfully completed request counts",
-		resp_buff+len,
-		get_status_perdevice_reqcnt((unsigned int *)workarea),
-		Z90CRYPT_NUM_APS);
-	*eof = 1;
-	memset(workarea, 0, Z90CRYPT_NUM_APS * sizeof(unsigned int));
-	return len;
-}
-
-static inline void
-disable_card(int card_index)
-{
-	struct device *devp;
-
-	devp = LONG2DEVPTR(card_index);
-	if (!devp || devp->user_disabled)
-		return;
-	devp->user_disabled = 1;
-	z90crypt.hdware_info->hdware_mask.user_disabled_count++;
-	if (devp->dev_type == -1)
-		return;
-	z90crypt.hdware_info->type_mask[devp->dev_type].user_disabled_count++;
-}
-
-static inline void
-enable_card(int card_index)
-{
-	struct device *devp;
-
-	devp = LONG2DEVPTR(card_index);
-	if (!devp || !devp->user_disabled)
-		return;
-	devp->user_disabled = 0;
-	z90crypt.hdware_info->hdware_mask.user_disabled_count--;
-	if (devp->dev_type == -1)
-		return;
-	z90crypt.hdware_info->type_mask[devp->dev_type].user_disabled_count--;
-}
-
-static int
-z90crypt_status_write(struct file *file, const char __user *buffer,
-		      unsigned long count, void *data)
-{
-	int j, eol;
-	unsigned char *lbuf, *ptr;
-	unsigned int local_count;
-
-#define LBUFSIZE 1200
-	lbuf = kmalloc(LBUFSIZE, GFP_KERNEL);
-	if (!lbuf) {
-		PRINTK("kmalloc failed!\n");
-		return 0;
-	}
-
-	if (count <= 0)
-		return 0;
-
-	local_count = UMIN((unsigned int)count, LBUFSIZE-1);
-
-	if (copy_from_user(lbuf, buffer, local_count) != 0) {
-		kfree(lbuf);
-		return -EFAULT;
-	}
-
-	lbuf[local_count] = '\0';
-
-	ptr = strstr(lbuf, "Online devices");
-	if (ptr == 0) {
-		PRINTK("Unable to parse data (missing \"Online devices\")\n");
-		kfree(lbuf);
-		return count;
-	}
-
-	ptr = strstr(ptr, "\n");
-	if (ptr == 0) {
-		PRINTK("Unable to parse data (missing newline after \"Online devices\")\n");
-		kfree(lbuf);
-		return count;
-	}
-	ptr++;
-
-	if (strstr(ptr, "Waiting work element counts") == NULL) {
-		PRINTK("Unable to parse data (missing \"Waiting work element counts\")\n");
-		kfree(lbuf);
-		return count;
-	}
-
-	j = 0;
-	eol = 0;
-	while ((j < 64) && (*ptr != '\0')) {
-		switch (*ptr) {
-		case '\t':
-		case ' ':
-			break;
-		case '\n':
-		default:
-			eol = 1;
-			break;
-		case '0':	// no device
-		case '1':	// PCICA
-		case '2':	// PCICC
-		case '3':	// PCIXCC_MCL2
-		case '4':	// PCIXCC_MCL3
-		case '5':	// CEX2C
-		case '6':       // CEX2A
-			j++;
-			break;
-		case 'd':
-		case 'D':
-			disable_card(j);
-			j++;
-			break;
-		case 'e':
-		case 'E':
-			enable_card(j);
-			j++;
-			break;
-		}
-		if (eol)
-			break;
-		ptr++;
-	}
-
-	kfree(lbuf);
-	return count;
-}
-
-/**
- * Functions that run under a timer, with no process id
- *
- * The task functions:
- *     z90crypt_reader_task
- *	 helper_send_work
- *	 helper_handle_work_element
- *	 helper_receive_rc
- *     z90crypt_config_task
- *     z90crypt_cleanup_task
- *
- * Helper functions:
- *     z90crypt_schedule_reader_timer
- *     z90crypt_schedule_reader_task
- *     z90crypt_schedule_config_task
- *     z90crypt_schedule_cleanup_task
- */
-static inline int
-receive_from_crypto_device(int index, unsigned char *psmid, int *buff_len_p,
-			   unsigned char *buff, unsigned char __user **dest_p_p)
-{
-	int dv, rv;
-	struct device *dev_ptr;
-	struct caller *caller_p;
-	struct ica_rsa_modexpo *icaMsg_p;
-	struct list_head *ptr, *tptr;
-
-	memcpy(psmid, NULL_psmid, sizeof(NULL_psmid));
-
-	if (z90crypt.terminating)
-		return REC_FATAL_ERROR;
-
-	caller_p = 0;
-	dev_ptr = z90crypt.device_p[index];
-	rv = 0;
-	do {
-		if (!dev_ptr || dev_ptr->disabled) {
-			rv = REC_NO_WORK; // a disabled device can't return work
-			break;
-		}
-		if (dev_ptr->dev_self_x != index) {
-			PRINTKC("Corrupt dev ptr\n");
-			z90crypt.terminating = 1;
-			rv = REC_FATAL_ERROR;
-			break;
-		}
-		if (!dev_ptr->dev_resp_l || !dev_ptr->dev_resp_p) {
-			dv = DEV_REC_EXCEPTION;
-			PRINTK("dev_resp_l = %d, dev_resp_p = %p\n",
-			       dev_ptr->dev_resp_l, dev_ptr->dev_resp_p);
-		} else {
-			PDEBUG("Dequeue called for device %d\n", index);
-			dv = receive_from_AP(index, z90crypt.cdx,
-					     dev_ptr->dev_resp_l,
-					     dev_ptr->dev_resp_p, psmid);
-		}
-		switch (dv) {
-		case DEV_REC_EXCEPTION:
-			rv = REC_FATAL_ERROR;
-			z90crypt.terminating = 1;
-			PRINTKC("Exception in receive from device %d\n",
-				index);
-			break;
-		case DEV_ONLINE:
-			rv = 0;
-			break;
-		case DEV_EMPTY:
-			rv = REC_EMPTY;
-			break;
-		case DEV_NO_WORK:
-			rv = REC_NO_WORK;
-			break;
-		case DEV_BAD_MESSAGE:
-		case DEV_GONE:
-		case REC_HARDWAR_ERR:
-		default:
-			rv = REC_NO_RESPONSE;
-			break;
-		}
-		if (rv)
-			break;
-		if (dev_ptr->dev_caller_count <= 0) {
-			rv = REC_USER_GONE;
-			break;
-	        }
-
-		list_for_each_safe(ptr, tptr, &dev_ptr->dev_caller_list) {
-			caller_p = list_entry(ptr, struct caller, caller_liste);
-			if (!memcmp(caller_p->caller_id, psmid,
-				    sizeof(caller_p->caller_id))) {
-				if (!list_empty(&caller_p->caller_liste)) {
-					list_del_init(ptr);
-					dev_ptr->dev_caller_count--;
-					break;
-				}
-			}
-			caller_p = 0;
-		}
-		if (!caller_p) {
-			PRINTKW("Unable to locate PSMID %02X%02X%02X%02X%02X"
-				"%02X%02X%02X in device list\n",
-				psmid[0], psmid[1], psmid[2], psmid[3],
-				psmid[4], psmid[5], psmid[6], psmid[7]);
-			rv = REC_USER_GONE;
-			break;
-		}
-
-		PDEBUG("caller_p after successful receive: %p\n", caller_p);
-		rv = convert_response(dev_ptr->dev_resp_p,
-				      caller_p->caller_buf_p, buff_len_p, buff);
-		switch (rv) {
-		case REC_USE_PCICA:
-			break;
-		case REC_OPERAND_INV:
-		case REC_OPERAND_SIZE:
-		case REC_EVEN_MOD:
-		case REC_INVALID_PAD:
-			PDEBUG("device %d: 'user error' %d\n", index, rv);
-			break;
-		case WRONG_DEVICE_TYPE:
-		case REC_HARDWAR_ERR:
-		case REC_BAD_MESSAGE:
-			PRINTKW("device %d: hardware error %d\n", index, rv);
-			rv = REC_NO_RESPONSE;
-			break;
-		default:
-			PDEBUG("device %d: rv = %d\n", index, rv);
-			break;
-		}
-	} while (0);
-
-	switch (rv) {
-	case 0:
-		PDEBUG("Successful receive from device %d\n", index);
-		icaMsg_p = (struct ica_rsa_modexpo *)caller_p->caller_buf_p;
-		*dest_p_p = icaMsg_p->outputdata;
-		if (*buff_len_p == 0)
-			PRINTK("Zero *buff_len_p\n");
-		break;
-	case REC_NO_RESPONSE:
-		PRINTKW("Removing device %d from availability\n", index);
-		remove_device(dev_ptr);
-		break;
-	}
-
-	if (caller_p)
-		unbuild_caller(dev_ptr, caller_p);
-
-	return rv;
-}
-
-static inline void
-helper_send_work(int index)
-{
-	struct work_element *rq_p;
-	int rv;
-
-	if (list_empty(&request_list))
-		return;
-	requestq_count--;
-	rq_p = list_entry(request_list.next, struct work_element, liste);
-	list_del_init(&rq_p->liste);
-	rq_p->audit[1] |= FP_REMREQUEST;
-	if (rq_p->devtype == SHRT2DEVPTR(index)->dev_type) {
-		rq_p->devindex = SHRT2LONG(index);
-		rv = send_to_crypto_device(rq_p);
-		if (rv == 0) {
-			rq_p->requestsent = jiffies;
-			rq_p->audit[0] |= FP_SENT;
-			list_add_tail(&rq_p->liste, &pending_list);
-			++pendingq_count;
-			rq_p->audit[0] |= FP_PENDING;
-		} else {
-			switch (rv) {
-			case REC_OPERAND_INV:
-			case REC_OPERAND_SIZE:
-			case REC_EVEN_MOD:
-			case REC_INVALID_PAD:
-				rq_p->retcode = -EINVAL;
-				break;
-			case SEN_NOT_AVAIL:
-			case SEN_RETRY:
-			case REC_NO_RESPONSE:
-			default:
-				if (z90crypt.mask.st_count > 1)
-					rq_p->retcode =
-						-ERESTARTSYS;
-				else
-					rq_p->retcode = -ENODEV;
-				break;
-			}
-			rq_p->status[0] |= STAT_FAILED;
-			rq_p->audit[1] |= FP_AWAKENING;
-			atomic_set(&rq_p->alarmrung, 1);
-			wake_up(&rq_p->waitq);
-		}
-	} else {
-		if (z90crypt.mask.st_count > 1)
-			rq_p->retcode = -ERESTARTSYS;
-		else
-			rq_p->retcode = -ENODEV;
-		rq_p->status[0] |= STAT_FAILED;
-		rq_p->audit[1] |= FP_AWAKENING;
-		atomic_set(&rq_p->alarmrung, 1);
-		wake_up(&rq_p->waitq);
-	}
-}
-
-static inline void
-helper_handle_work_element(int index, unsigned char psmid[8], int rc,
-			   int buff_len, unsigned char *buff,
-			   unsigned char __user *resp_addr)
-{
-	struct work_element *pq_p;
-	struct list_head *lptr, *tptr;
-
-	pq_p = 0;
-	list_for_each_safe(lptr, tptr, &pending_list) {
-		pq_p = list_entry(lptr, struct work_element, liste);
-		if (!memcmp(pq_p->caller_id, psmid, sizeof(pq_p->caller_id))) {
-			list_del_init(lptr);
-			pendingq_count--;
-			pq_p->audit[1] |= FP_NOTPENDING;
-			break;
-		}
-		pq_p = 0;
-	}
-
-	if (!pq_p) {
-		PRINTK("device %d has work but no caller exists on pending Q\n",
-		       SHRT2LONG(index));
-		return;
-	}
-
-	switch (rc) {
-		case 0:
-			pq_p->resp_buff_size = buff_len;
-			pq_p->audit[1] |= FP_RESPSIZESET;
-			if (buff_len) {
-				pq_p->resp_addr = resp_addr;
-				pq_p->audit[1] |= FP_RESPADDRCOPIED;
-				memcpy(pq_p->resp_buff, buff, buff_len);
-				pq_p->audit[1] |= FP_RESPBUFFCOPIED;
-			}
-			break;
-		case REC_OPERAND_INV:
-		case REC_OPERAND_SIZE:
-		case REC_EVEN_MOD:
-		case REC_INVALID_PAD:
-			PDEBUG("-EINVAL after application error %d\n", rc);
-			pq_p->retcode = -EINVAL;
-			pq_p->status[0] |= STAT_FAILED;
-			break;
-		case REC_USE_PCICA:
-			pq_p->retcode = -ERESTARTSYS;
-			pq_p->status[0] |= STAT_FAILED;
-			break;
-		case REC_NO_RESPONSE:
-		default:
-			if (z90crypt.mask.st_count > 1)
-				pq_p->retcode = -ERESTARTSYS;
-			else
-				pq_p->retcode = -ENODEV;
-			pq_p->status[0] |= STAT_FAILED;
-			break;
-	}
-	if ((pq_p->status[0] != STAT_FAILED) || (pq_p->retcode != -ERELEASED)) {
-		pq_p->audit[1] |= FP_AWAKENING;
-		atomic_set(&pq_p->alarmrung, 1);
-		wake_up(&pq_p->waitq);
-	}
-}
-
-/**
- * return TRUE if the work element should be removed from the queue
- */
-static inline int
-helper_receive_rc(int index, int *rc_p)
-{
-	switch (*rc_p) {
-	case 0:
-	case REC_OPERAND_INV:
-	case REC_OPERAND_SIZE:
-	case REC_EVEN_MOD:
-	case REC_INVALID_PAD:
-	case REC_USE_PCICA:
-		break;
-
-	case REC_BUSY:
-	case REC_NO_WORK:
-	case REC_EMPTY:
-	case REC_RETRY_DEV:
-	case REC_FATAL_ERROR:
-		return 0;
-
-	case REC_NO_RESPONSE:
-		break;
-
-	default:
-		PRINTK("rc %d, device %d converted to REC_NO_RESPONSE\n",
-		       *rc_p, SHRT2LONG(index));
-		*rc_p = REC_NO_RESPONSE;
-		break;
-	}
-	return 1;
-}
-
-static inline void
-z90crypt_schedule_reader_timer(void)
-{
-	if (timer_pending(&reader_timer))
-		return;
-	if (mod_timer(&reader_timer, jiffies+(READERTIME*HZ/1000)) != 0)
-		PRINTK("Timer pending while modifying reader timer\n");
-}
-
-static void
-z90crypt_reader_task(unsigned long ptr)
-{
-	int workavail, index, rc, buff_len;
-	unsigned char	psmid[8];
-	unsigned char __user *resp_addr;
-	static unsigned char buff[1024];
-
-	/**
-	 * we use workavail = 2 to ensure 2 passes with nothing dequeued before
-	 * exiting the loop. If (pendingq_count+requestq_count) == 0 after the
-	 * loop, there is no work remaining on the queues.
-	 */
-	resp_addr = 0;
-	workavail = 2;
-	buff_len = 0;
-	while (workavail) {
-		workavail--;
-		rc = 0;
-		spin_lock_irq(&queuespinlock);
-		memset(buff, 0x00, sizeof(buff));
-
-		/* Dequeue once from each device in round robin. */
-		for (index = 0; index < z90crypt.mask.st_count; index++) {
-			PDEBUG("About to receive.\n");
-			rc = receive_from_crypto_device(SHRT2LONG(index),
-							psmid,
-							&buff_len,
-							buff,
-							&resp_addr);
-			PDEBUG("Dequeued: rc = %d.\n", rc);
-
-			if (helper_receive_rc(index, &rc)) {
-				if (rc != REC_NO_RESPONSE) {
-					helper_send_work(index);
-					workavail = 2;
-				}
-
-				helper_handle_work_element(index, psmid, rc,
-							   buff_len, buff,
-							   resp_addr);
-			}
-
-			if (rc == REC_FATAL_ERROR)
-				PRINTKW("REC_FATAL_ERROR from device %d!\n",
-					SHRT2LONG(index));
-		}
-		spin_unlock_irq(&queuespinlock);
-	}
-
-	if (pendingq_count + requestq_count)
-		z90crypt_schedule_reader_timer();
-}
-
-static inline void
-z90crypt_schedule_config_task(unsigned int expiration)
-{
-	if (timer_pending(&config_timer))
-		return;
-	if (mod_timer(&config_timer, jiffies+(expiration*HZ)) != 0)
-		PRINTK("Timer pending while modifying config timer\n");
-}
-
-static void
-z90crypt_config_task(unsigned long ptr)
-{
-	int rc;
-
-	PDEBUG("jiffies %ld\n", jiffies);
-
-	if ((rc = refresh_z90crypt(&z90crypt.cdx)))
-		PRINTK("Error %d detected in refresh_z90crypt.\n", rc);
-	/* If return was fatal, don't bother reconfiguring */
-	if ((rc != TSQ_FATAL_ERROR) && (rc != RSQ_FATAL_ERROR))
-		z90crypt_schedule_config_task(CONFIGTIME);
-}
-
-static inline void
-z90crypt_schedule_cleanup_task(void)
-{
-	if (timer_pending(&cleanup_timer))
-		return;
-	if (mod_timer(&cleanup_timer, jiffies+(CLEANUPTIME*HZ)) != 0)
-		PRINTK("Timer pending while modifying cleanup timer\n");
-}
-
-static inline void
-helper_drain_queues(void)
-{
-	struct work_element *pq_p;
-	struct list_head *lptr, *tptr;
-
-	list_for_each_safe(lptr, tptr, &pending_list) {
-		pq_p = list_entry(lptr, struct work_element, liste);
-		pq_p->retcode = -ENODEV;
-		pq_p->status[0] |= STAT_FAILED;
-		unbuild_caller(LONG2DEVPTR(pq_p->devindex),
-			       (struct caller *)pq_p->requestptr);
-		list_del_init(lptr);
-		pendingq_count--;
-		pq_p->audit[1] |= FP_NOTPENDING;
-		pq_p->audit[1] |= FP_AWAKENING;
-		atomic_set(&pq_p->alarmrung, 1);
-		wake_up(&pq_p->waitq);
-	}
-
-	list_for_each_safe(lptr, tptr, &request_list) {
-		pq_p = list_entry(lptr, struct work_element, liste);
-		pq_p->retcode = -ENODEV;
-		pq_p->status[0] |= STAT_FAILED;
-		list_del_init(lptr);
-		requestq_count--;
-		pq_p->audit[1] |= FP_REMREQUEST;
-		pq_p->audit[1] |= FP_AWAKENING;
-		atomic_set(&pq_p->alarmrung, 1);
-		wake_up(&pq_p->waitq);
-	}
-}
-
-static inline void
-helper_timeout_requests(void)
-{
-	struct work_element *pq_p;
-	struct list_head *lptr, *tptr;
-	long timelimit;
-
-	timelimit = jiffies - (CLEANUPTIME * HZ);
-	/* The list is in strict chronological order */
-	list_for_each_safe(lptr, tptr, &pending_list) {
-		pq_p = list_entry(lptr, struct work_element, liste);
-		if (pq_p->requestsent >= timelimit)
-			break;
-		PRINTKW("Purging(PQ) PSMID %02X%02X%02X%02X%02X%02X%02X%02X\n",
-		       ((struct caller *)pq_p->requestptr)->caller_id[0],
-		       ((struct caller *)pq_p->requestptr)->caller_id[1],
-		       ((struct caller *)pq_p->requestptr)->caller_id[2],
-		       ((struct caller *)pq_p->requestptr)->caller_id[3],
-		       ((struct caller *)pq_p->requestptr)->caller_id[4],
-		       ((struct caller *)pq_p->requestptr)->caller_id[5],
-		       ((struct caller *)pq_p->requestptr)->caller_id[6],
-		       ((struct caller *)pq_p->requestptr)->caller_id[7]);
-		pq_p->retcode = -ETIMEOUT;
-		pq_p->status[0] |= STAT_FAILED;
-		/* get this off any caller queue it may be on */
-		unbuild_caller(LONG2DEVPTR(pq_p->devindex),
-			       (struct caller *) pq_p->requestptr);
-		list_del_init(lptr);
-		pendingq_count--;
-		pq_p->audit[1] |= FP_TIMEDOUT;
-		pq_p->audit[1] |= FP_NOTPENDING;
-		pq_p->audit[1] |= FP_AWAKENING;
-		atomic_set(&pq_p->alarmrung, 1);
-		wake_up(&pq_p->waitq);
-	}
-
-	/**
-	 * If pending count is zero, items left on the request queue may
-	 * never be processed.
-	 */
-	if (pendingq_count <= 0) {
-		list_for_each_safe(lptr, tptr, &request_list) {
-			pq_p = list_entry(lptr, struct work_element, liste);
-			if (pq_p->requestsent >= timelimit)
-				break;
-		PRINTKW("Purging(RQ) PSMID %02X%02X%02X%02X%02X%02X%02X%02X\n",
-		       ((struct caller *)pq_p->requestptr)->caller_id[0],
-		       ((struct caller *)pq_p->requestptr)->caller_id[1],
-		       ((struct caller *)pq_p->requestptr)->caller_id[2],
-		       ((struct caller *)pq_p->requestptr)->caller_id[3],
-		       ((struct caller *)pq_p->requestptr)->caller_id[4],
-		       ((struct caller *)pq_p->requestptr)->caller_id[5],
-		       ((struct caller *)pq_p->requestptr)->caller_id[6],
-		       ((struct caller *)pq_p->requestptr)->caller_id[7]);
-			pq_p->retcode = -ETIMEOUT;
-			pq_p->status[0] |= STAT_FAILED;
-			list_del_init(lptr);
-			requestq_count--;
-			pq_p->audit[1] |= FP_TIMEDOUT;
-			pq_p->audit[1] |= FP_REMREQUEST;
-			pq_p->audit[1] |= FP_AWAKENING;
-			atomic_set(&pq_p->alarmrung, 1);
-			wake_up(&pq_p->waitq);
-		}
-	}
-}
-
-static void
-z90crypt_cleanup_task(unsigned long ptr)
-{
-	PDEBUG("jiffies %ld\n", jiffies);
-	spin_lock_irq(&queuespinlock);
-	if (z90crypt.mask.st_count <= 0) // no devices!
-		helper_drain_queues();
-	else
-		helper_timeout_requests();
-	spin_unlock_irq(&queuespinlock);
-	z90crypt_schedule_cleanup_task();
-}
-
-static void
-z90crypt_schedule_reader_task(unsigned long ptr)
-{
-	tasklet_schedule(&reader_tasklet);
-}
-
-/**
- * Lowlevel Functions:
- *
- *   create_z90crypt:  creates and initializes basic data structures
- *   refresh_z90crypt:	re-initializes basic data structures
- *   find_crypto_devices: returns a count and mask of hardware status
- *   create_crypto_device:  builds the descriptor for a device
- *   destroy_crypto_device:  unallocates the descriptor for a device
- *   destroy_z90crypt:	drains all work, unallocates structs
- */
-
-/**
- * build the z90crypt root structure using the given domain index
- */
-static int
-create_z90crypt(int *cdx_p)
-{
-	struct hdware_block *hdware_blk_p;
-
-	memset(&z90crypt, 0x00, sizeof(struct z90crypt));
-	z90crypt.domain_established = 0;
-	z90crypt.len = sizeof(struct z90crypt);
-	z90crypt.max_count = Z90CRYPT_NUM_DEVS;
-	z90crypt.cdx = *cdx_p;
-
-	hdware_blk_p = kzalloc(sizeof(struct hdware_block), GFP_ATOMIC);
-	if (!hdware_blk_p) {
-		PDEBUG("kmalloc for hardware block failed\n");
-		return ENOMEM;
-	}
-	z90crypt.hdware_info = hdware_blk_p;
-
-	return 0;
-}
-
-static inline int
-helper_scan_devices(int cdx_array[16], int *cdx_p, int *correct_cdx_found)
-{
-	enum hdstat hd_stat;
-	int q_depth, dev_type;
-	int indx, chkdom, numdomains;
-
-	q_depth = dev_type = numdomains = 0;
-	for (chkdom = 0; chkdom <= 15; cdx_array[chkdom++] = -1);
-	for (indx = 0; indx < z90crypt.max_count; indx++) {
-		hd_stat = HD_NOT_THERE;
-		numdomains = 0;
-		for (chkdom = 0; chkdom <= 15; chkdom++) {
-			hd_stat = query_online(indx, chkdom, MAX_RESET,
-					       &q_depth, &dev_type);
-			if (hd_stat == HD_TSQ_EXCEPTION) {
-				z90crypt.terminating = 1;
-				PRINTKC("exception taken!\n");
-				break;
-			}
-			if (hd_stat == HD_ONLINE) {
-				cdx_array[numdomains++] = chkdom;
-				if (*cdx_p == chkdom) {
-					*correct_cdx_found  = 1;
-					break;
-				}
-			}
-		}
-		if ((*correct_cdx_found == 1) || (numdomains != 0))
-			break;
-		if (z90crypt.terminating)
-			break;
-	}
-	return numdomains;
-}
-
-static inline int
-probe_crypto_domain(int *cdx_p)
-{
-	int cdx_array[16];
-	char cdx_array_text[53], temp[5];
-	int correct_cdx_found, numdomains;
-
-	correct_cdx_found = 0;
-	numdomains = helper_scan_devices(cdx_array, cdx_p, &correct_cdx_found);
-
-	if (z90crypt.terminating)
-		return TSQ_FATAL_ERROR;
-
-	if (correct_cdx_found)
-		return 0;
-
-	if (numdomains == 0) {
-		PRINTKW("Unable to find crypto domain: No devices found\n");
-		return Z90C_NO_DEVICES;
-	}
-
-	if (numdomains == 1) {
-		if (*cdx_p == -1) {
-			*cdx_p = cdx_array[0];
-			return 0;
-		}
-		PRINTKW("incorrect domain: specified = %d, found = %d\n",
-		       *cdx_p, cdx_array[0]);
-		return Z90C_INCORRECT_DOMAIN;
-	}
-
-	numdomains--;
-	sprintf(cdx_array_text, "%d", cdx_array[numdomains]);
-	while (numdomains) {
-		numdomains--;
-		sprintf(temp, ", %d", cdx_array[numdomains]);
-		strcat(cdx_array_text, temp);
-	}
-
-	PRINTKW("ambiguous domain detected: specified = %d, found array = %s\n",
-		*cdx_p, cdx_array_text);
-	return Z90C_AMBIGUOUS_DOMAIN;
-}
-
-static int
-refresh_z90crypt(int *cdx_p)
-{
-	int i, j, indx, rv;
-	static struct status local_mask;
-	struct device *devPtr;
-	unsigned char oldStat, newStat;
-	int return_unchanged;
-
-	if (z90crypt.len != sizeof(z90crypt))
-		return ENOTINIT;
-	if (z90crypt.terminating)
-		return TSQ_FATAL_ERROR;
-	rv = 0;
-	if (!z90crypt.hdware_info->hdware_mask.st_count &&
-	    !z90crypt.domain_established) {
-		rv = probe_crypto_domain(cdx_p);
-		if (z90crypt.terminating)
-			return TSQ_FATAL_ERROR;
-		if (rv == Z90C_NO_DEVICES)
-			return 0; // try later
-		if (rv)
-			return rv;
-		z90crypt.cdx = *cdx_p;
-		z90crypt.domain_established = 1;
-	}
-	rv = find_crypto_devices(&local_mask);
-	if (rv) {
-		PRINTK("find crypto devices returned %d\n", rv);
-		return rv;
-	}
-	if (!memcmp(&local_mask, &z90crypt.hdware_info->hdware_mask,
-		    sizeof(struct status))) {
-		return_unchanged = 1;
-		for (i = 0; i < Z90CRYPT_NUM_TYPES; i++) {
-			/**
-			 * Check for disabled cards.  If any device is marked
-			 * disabled, destroy it.
-			 */
-			for (j = 0;
-			     j < z90crypt.hdware_info->type_mask[i].st_count;
-			     j++) {
-				indx = z90crypt.hdware_info->type_x_addr[i].
-								device_index[j];
-				devPtr = z90crypt.device_p[indx];
-				if (devPtr && devPtr->disabled) {
-					local_mask.st_mask[indx] = HD_NOT_THERE;
-					return_unchanged = 0;
-				}
-			}
-		}
-		if (return_unchanged == 1)
-			return 0;
-	}
-
-	spin_lock_irq(&queuespinlock);
-	for (i = 0; i < z90crypt.max_count; i++) {
-		oldStat = z90crypt.hdware_info->hdware_mask.st_mask[i];
-		newStat = local_mask.st_mask[i];
-		if ((oldStat == HD_ONLINE) && (newStat != HD_ONLINE))
-			destroy_crypto_device(i);
-		else if ((oldStat != HD_ONLINE) && (newStat == HD_ONLINE)) {
-			rv = create_crypto_device(i);
-			if (rv >= REC_FATAL_ERROR)
-				return rv;
-			if (rv != 0) {
-				local_mask.st_mask[i] = HD_NOT_THERE;
-				local_mask.st_count--;
-			}
-		}
-	}
-	memcpy(z90crypt.hdware_info->hdware_mask.st_mask, local_mask.st_mask,
-	       sizeof(local_mask.st_mask));
-	z90crypt.hdware_info->hdware_mask.st_count = local_mask.st_count;
-	z90crypt.hdware_info->hdware_mask.disabled_count =
-						      local_mask.disabled_count;
-	refresh_index_array(&z90crypt.mask, &z90crypt.overall_device_x);
-	for (i = 0; i < Z90CRYPT_NUM_TYPES; i++)
-		refresh_index_array(&(z90crypt.hdware_info->type_mask[i]),
-				    &(z90crypt.hdware_info->type_x_addr[i]));
-	spin_unlock_irq(&queuespinlock);
-
-	return rv;
-}
-
-static int
-find_crypto_devices(struct status *deviceMask)
-{
-	int i, q_depth, dev_type;
-	enum hdstat hd_stat;
-
-	deviceMask->st_count = 0;
-	deviceMask->disabled_count = 0;
-	deviceMask->user_disabled_count = 0;
-
-	for (i = 0; i < z90crypt.max_count; i++) {
-		hd_stat = query_online(i, z90crypt.cdx, MAX_RESET, &q_depth,
-				       &dev_type);
-		if (hd_stat == HD_TSQ_EXCEPTION) {
-			z90crypt.terminating = 1;
-			PRINTKC("Exception during probe for crypto devices\n");
-			return TSQ_FATAL_ERROR;
-		}
-		deviceMask->st_mask[i] = hd_stat;
-		if (hd_stat == HD_ONLINE) {
-			PDEBUG("Got an online crypto!: %d\n", i);
-			PDEBUG("Got a queue depth of %d\n", q_depth);
-			PDEBUG("Got a device type of %d\n", dev_type);
-			if (q_depth <= 0)
-				return TSQ_FATAL_ERROR;
-			deviceMask->st_count++;
-			z90crypt.q_depth_array[i] = q_depth;
-			z90crypt.dev_type_array[i] = dev_type;
-		}
-	}
-
-	return 0;
-}
-
-static int
-refresh_index_array(struct status *status_str, struct device_x *index_array)
-{
-	int i, count;
-	enum devstat stat;
-
-	i = -1;
-	count = 0;
-	do {
-		stat = status_str->st_mask[++i];
-		if (stat == DEV_ONLINE)
-			index_array->device_index[count++] = i;
-	} while ((i < Z90CRYPT_NUM_DEVS) && (count < status_str->st_count));
-
-	return count;
-}
-
-static int
-create_crypto_device(int index)
-{
-	int rv, devstat, total_size;
-	struct device *dev_ptr;
-	struct status *type_str_p;
-	int deviceType;
-
-	dev_ptr = z90crypt.device_p[index];
-	if (!dev_ptr) {
-		total_size = sizeof(struct device) +
-			     z90crypt.q_depth_array[index] * sizeof(int);
-
-		dev_ptr = kzalloc(total_size, GFP_ATOMIC);
-		if (!dev_ptr) {
-			PRINTK("kmalloc device %d failed\n", index);
-			return ENOMEM;
-		}
-		dev_ptr->dev_resp_p = kmalloc(MAX_RESPONSE_SIZE, GFP_ATOMIC);
-		if (!dev_ptr->dev_resp_p) {
-			kfree(dev_ptr);
-			PRINTK("kmalloc device %d rec buffer failed\n", index);
-			return ENOMEM;
-		}
-		dev_ptr->dev_resp_l = MAX_RESPONSE_SIZE;
-		INIT_LIST_HEAD(&(dev_ptr->dev_caller_list));
-	}
-
-	devstat = reset_device(index, z90crypt.cdx, MAX_RESET);
-	if (devstat == DEV_RSQ_EXCEPTION) {
-		PRINTK("exception during reset device %d\n", index);
-		kfree(dev_ptr->dev_resp_p);
-		kfree(dev_ptr);
-		return RSQ_FATAL_ERROR;
-	}
-	if (devstat == DEV_ONLINE) {
-		dev_ptr->dev_self_x = index;
-		dev_ptr->dev_type = z90crypt.dev_type_array[index];
-		if (dev_ptr->dev_type == NILDEV) {
-			rv = probe_device_type(dev_ptr);
-			if (rv) {
-				PRINTK("rv = %d from probe_device_type %d\n",
-				       rv, index);
-				kfree(dev_ptr->dev_resp_p);
-				kfree(dev_ptr);
-				return rv;
-			}
-		}
-		if (dev_ptr->dev_type == PCIXCC_UNK) {
-			rv = probe_PCIXCC_type(dev_ptr);
-			if (rv) {
-				PRINTK("rv = %d from probe_PCIXCC_type %d\n",
-				       rv, index);
-				kfree(dev_ptr->dev_resp_p);
-				kfree(dev_ptr);
-				return rv;
-			}
-		}
-		deviceType = dev_ptr->dev_type;
-		z90crypt.dev_type_array[index] = deviceType;
-		if (deviceType == PCICA)
-			z90crypt.hdware_info->device_type_array[index] = 1;
-		else if (deviceType == PCICC)
-			z90crypt.hdware_info->device_type_array[index] = 2;
-		else if (deviceType == PCIXCC_MCL2)
-			z90crypt.hdware_info->device_type_array[index] = 3;
-		else if (deviceType == PCIXCC_MCL3)
-			z90crypt.hdware_info->device_type_array[index] = 4;
-		else if (deviceType == CEX2C)
-			z90crypt.hdware_info->device_type_array[index] = 5;
-		else if (deviceType == CEX2A)
-			z90crypt.hdware_info->device_type_array[index] = 6;
-		else // No idea how this would happen.
-			z90crypt.hdware_info->device_type_array[index] = -1;
-	}
-
-	/**
-	 * 'q_depth' returned by the hardware is one less than
-	 * the actual depth
-	 */
-	dev_ptr->dev_q_depth = z90crypt.q_depth_array[index];
-	dev_ptr->dev_type = z90crypt.dev_type_array[index];
-	dev_ptr->dev_stat = devstat;
-	dev_ptr->disabled = 0;
-	z90crypt.device_p[index] = dev_ptr;
-
-	if (devstat == DEV_ONLINE) {
-		if (z90crypt.mask.st_mask[index] != DEV_ONLINE) {
-			z90crypt.mask.st_mask[index] = DEV_ONLINE;
-			z90crypt.mask.st_count++;
-		}
-		deviceType = dev_ptr->dev_type;
-		type_str_p = &z90crypt.hdware_info->type_mask[deviceType];
-		if (type_str_p->st_mask[index] != DEV_ONLINE) {
-			type_str_p->st_mask[index] = DEV_ONLINE;
-			type_str_p->st_count++;
-		}
-	}
-
-	return 0;
-}
-
-static int
-destroy_crypto_device(int index)
-{
-	struct device *dev_ptr;
-	int t, disabledFlag;
-
-	dev_ptr = z90crypt.device_p[index];
-
-	/* remember device type; get rid of device struct */
-	if (dev_ptr) {
-		disabledFlag = dev_ptr->disabled;
-		t = dev_ptr->dev_type;
-		kfree(dev_ptr->dev_resp_p);
-		kfree(dev_ptr);
-	} else {
-		disabledFlag = 0;
-		t = -1;
-	}
-	z90crypt.device_p[index] = 0;
-
-	/* if the type is valid, remove the device from the type_mask */
-	if ((t != -1) && z90crypt.hdware_info->type_mask[t].st_mask[index]) {
-		  z90crypt.hdware_info->type_mask[t].st_mask[index] = 0x00;
-		  z90crypt.hdware_info->type_mask[t].st_count--;
-		  if (disabledFlag == 1)
-			z90crypt.hdware_info->type_mask[t].disabled_count--;
-	}
-	if (z90crypt.mask.st_mask[index] != DEV_GONE) {
-		z90crypt.mask.st_mask[index] = DEV_GONE;
-		z90crypt.mask.st_count--;
-	}
-	z90crypt.hdware_info->device_type_array[index] = 0;
-
-	return 0;
-}
-
-static void
-destroy_z90crypt(void)
-{
-	int i;
-
-	for (i = 0; i < z90crypt.max_count; i++)
-		if (z90crypt.device_p[i])
-			destroy_crypto_device(i);
-	kfree(z90crypt.hdware_info);
-	memset((void *)&z90crypt, 0, sizeof(z90crypt));
-}
-
-static unsigned char static_testmsg[384] = {
-0x00,0x00,0x00,0x00,0x01,0x02,0x03,0x04,0x05,0x06,0x07,0x08,0x00,0x06,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x58,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x01,0x00,0x43,0x43,
-0x41,0x2d,0x41,0x50,0x50,0x4c,0x20,0x20,0x20,0x01,0x01,0x01,0x00,0x00,0x00,0x00,
-0x50,0x4b,0x00,0x00,0x00,0x00,0x01,0x1c,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x05,0xb8,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x70,0x00,0x41,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x54,0x32,
-0x01,0x00,0xa0,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0xb8,0x05,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x0a,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x08,0x00,0x49,0x43,0x53,0x46,
-0x20,0x20,0x20,0x20,0x50,0x4b,0x0a,0x00,0x50,0x4b,0x43,0x53,0x2d,0x31,0x2e,0x32,
-0x37,0x00,0x11,0x22,0x33,0x44,0x55,0x66,0x77,0x88,0x99,0x00,0x11,0x22,0x33,0x44,
-0x55,0x66,0x77,0x88,0x99,0x00,0x11,0x22,0x33,0x44,0x55,0x66,0x77,0x88,0x99,0x00,
-0x11,0x22,0x33,0x44,0x55,0x66,0x77,0x88,0x99,0x00,0x11,0x22,0x33,0x44,0x55,0x66,
-0x77,0x88,0x99,0x00,0x11,0x22,0x33,0x5d,0x00,0x5b,0x00,0x77,0x88,0x1e,0x00,0x00,
-0x57,0x00,0x00,0x00,0x00,0x04,0x00,0x00,0x4f,0x00,0x00,0x00,0x03,0x02,0x00,0x00,
-0x40,0x01,0x00,0x01,0xce,0x02,0x68,0x2d,0x5f,0xa9,0xde,0x0c,0xf6,0xd2,0x7b,0x58,
-0x4b,0xf9,0x28,0x68,0x3d,0xb4,0xf4,0xef,0x78,0xd5,0xbe,0x66,0x63,0x42,0xef,0xf8,
-0xfd,0xa4,0xf8,0xb0,0x8e,0x29,0xc2,0xc9,0x2e,0xd8,0x45,0xb8,0x53,0x8c,0x6f,0x4e,
-0x72,0x8f,0x6c,0x04,0x9c,0x88,0xfc,0x1e,0xc5,0x83,0x55,0x57,0xf7,0xdd,0xfd,0x4f,
-0x11,0x36,0x95,0x5d,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00
-};
-
-static int
-probe_device_type(struct device *devPtr)
-{
-	int rv, dv, i, index, length;
-	unsigned char psmid[8];
-	static unsigned char loc_testmsg[sizeof(static_testmsg)];
-
-	index = devPtr->dev_self_x;
-	rv = 0;
-	do {
-		memcpy(loc_testmsg, static_testmsg, sizeof(static_testmsg));
-		length = sizeof(static_testmsg) - 24;
-		/* the -24 allows for the header */
-		dv = send_to_AP(index, z90crypt.cdx, length, loc_testmsg);
-		if (dv) {
-			PDEBUG("dv returned by send during probe: %d\n", dv);
-			if (dv == DEV_SEN_EXCEPTION) {
-				rv = SEN_FATAL_ERROR;
-				PRINTKC("exception in send to AP %d\n", index);
-				break;
-			}
-			PDEBUG("return value from send_to_AP: %d\n", rv);
-			switch (dv) {
-			case DEV_GONE:
-				PDEBUG("dev %d not available\n", index);
-				rv = SEN_NOT_AVAIL;
-				break;
-			case DEV_ONLINE:
-				rv = 0;
-				break;
-			case DEV_EMPTY:
-				rv = SEN_NOT_AVAIL;
-				break;
-			case DEV_NO_WORK:
-				rv = SEN_FATAL_ERROR;
-				break;
-			case DEV_BAD_MESSAGE:
-				rv = SEN_USER_ERROR;
-				break;
-			case DEV_QUEUE_FULL:
-				rv = SEN_QUEUE_FULL;
-				break;
-			default:
-				PRINTK("unknown dv=%d for dev %d\n", dv, index);
-				rv = SEN_NOT_AVAIL;
-				break;
-			}
-		}
-
-		if (rv)
-			break;
-
-		for (i = 0; i < 6; i++) {
-			mdelay(300);
-			dv = receive_from_AP(index, z90crypt.cdx,
-					     devPtr->dev_resp_l,
-					     devPtr->dev_resp_p, psmid);
-			PDEBUG("dv returned by DQ = %d\n", dv);
-			if (dv == DEV_REC_EXCEPTION) {
-				rv = REC_FATAL_ERROR;
-				PRINTKC("exception in dequeue %d\n",
-					index);
-				break;
-			}
-			switch (dv) {
-			case DEV_ONLINE:
-				rv = 0;
-				break;
-			case DEV_EMPTY:
-				rv = REC_EMPTY;
-				break;
-			case DEV_NO_WORK:
-				rv = REC_NO_WORK;
-				break;
-			case DEV_BAD_MESSAGE:
-			case DEV_GONE:
-			default:
-				rv = REC_NO_RESPONSE;
-				break;
-			}
-			if ((rv != 0) && (rv != REC_NO_WORK))
-				break;
-			if (rv == 0)
-				break;
-		}
-		if (rv)
-			break;
-		rv = (devPtr->dev_resp_p[0] == 0x00) &&
-		     (devPtr->dev_resp_p[1] == 0x86);
-		if (rv)
-			devPtr->dev_type = PCICC;
-		else
-			devPtr->dev_type = PCICA;
-		rv = 0;
-	} while (0);
-	/* In a general error case, the card is not marked online */
-	return rv;
-}
-
-static unsigned char MCL3_testmsg[] = {
-0x00,0x00,0x00,0x00,0xEE,0xEE,0xEE,0xEE,0xEE,0xEE,0xEE,0xEE,
-0x00,0x06,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x58,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x43,0x41,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x50,0x4B,0x00,0x00,0x00,0x00,0x01,0xC4,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x07,0x24,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0xDC,0x02,0x00,0x00,0x00,0x54,0x32,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0xE8,0x00,0x00,0x00,0x00,0x00,0x00,0x07,0x24,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x04,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x50,0x4B,0x00,0x0A,0x4D,0x52,0x50,0x20,0x20,0x20,0x20,0x20,
-0x00,0x42,0x00,0x01,0x02,0x03,0x04,0x05,0x06,0x07,0x08,0x09,0x0A,0x0B,0x0C,0x0D,
-0x0E,0x0F,0x00,0x11,0x22,0x33,0x44,0x55,0x66,0x77,0x88,0x99,0xAA,0xBB,0xCC,0xDD,
-0xEE,0xFF,0xFF,0xEE,0xDD,0xCC,0xBB,0xAA,0x99,0x88,0x77,0x66,0x55,0x44,0x33,0x22,
-0x11,0x00,0x01,0x23,0x45,0x67,0x89,0xAB,0xCD,0xEF,0xFE,0xDC,0xBA,0x98,0x76,0x54,
-0x32,0x10,0x00,0x9A,0x00,0x98,0x00,0x00,0x1E,0x00,0x00,0x94,0x00,0x00,0x00,0x00,
-0x04,0x00,0x00,0x8C,0x00,0x00,0x00,0x40,0x02,0x00,0x00,0x40,0xBA,0xE8,0x23,0x3C,
-0x75,0xF3,0x91,0x61,0xD6,0x73,0x39,0xCF,0x7B,0x6D,0x8E,0x61,0x97,0x63,0x9E,0xD9,
-0x60,0x55,0xD6,0xC7,0xEF,0xF8,0x1E,0x63,0x95,0x17,0xCC,0x28,0x45,0x60,0x11,0xC5,
-0xC4,0x4E,0x66,0xC6,0xE6,0xC3,0xDE,0x8A,0x19,0x30,0xCF,0x0E,0xD7,0xAA,0xDB,0x01,
-0xD8,0x00,0xBB,0x8F,0x39,0x9F,0x64,0x28,0xF5,0x7A,0x77,0x49,0xCC,0x6B,0xA3,0x91,
-0x97,0x70,0xE7,0x60,0x1E,0x39,0xE1,0xE5,0x33,0xE1,0x15,0x63,0x69,0x08,0x80,0x4C,
-0x67,0xC4,0x41,0x8F,0x48,0xDF,0x26,0x98,0xF1,0xD5,0x8D,0x88,0xD9,0x6A,0xA4,0x96,
-0xC5,0x84,0xD9,0x30,0x49,0x67,0x7D,0x19,0xB1,0xB3,0x45,0x4D,0xB2,0x53,0x9A,0x47,
-0x3C,0x7C,0x55,0xBF,0xCC,0x85,0x00,0x36,0xF1,0x3D,0x93,0x53
-};
-
-static int
-probe_PCIXCC_type(struct device *devPtr)
-{
-	int rv, dv, i, index, length;
-	unsigned char psmid[8];
-	static unsigned char loc_testmsg[548];
-	struct CPRBX *cprbx_p;
-
-	index = devPtr->dev_self_x;
-	rv = 0;
-	do {
-		memcpy(loc_testmsg, MCL3_testmsg, sizeof(MCL3_testmsg));
-		length = sizeof(MCL3_testmsg) - 0x0C;
-		dv = send_to_AP(index, z90crypt.cdx, length, loc_testmsg);
-		if (dv) {
-			PDEBUG("dv returned = %d\n", dv);
-			if (dv == DEV_SEN_EXCEPTION) {
-				rv = SEN_FATAL_ERROR;
-				PRINTKC("exception in send to AP %d\n", index);
-				break;
-			}
-			PDEBUG("return value from send_to_AP: %d\n", rv);
-			switch (dv) {
-			case DEV_GONE:
-				PDEBUG("dev %d not available\n", index);
-				rv = SEN_NOT_AVAIL;
-				break;
-			case DEV_ONLINE:
-				rv = 0;
-				break;
-			case DEV_EMPTY:
-				rv = SEN_NOT_AVAIL;
-				break;
-			case DEV_NO_WORK:
-				rv = SEN_FATAL_ERROR;
-				break;
-			case DEV_BAD_MESSAGE:
-				rv = SEN_USER_ERROR;
-				break;
-			case DEV_QUEUE_FULL:
-				rv = SEN_QUEUE_FULL;
-				break;
-			default:
-				PRINTK("unknown dv=%d for dev %d\n", dv, index);
-				rv = SEN_NOT_AVAIL;
-				break;
-			}
-		}
-
-		if (rv)
-			break;
-
-		for (i = 0; i < 6; i++) {
-			mdelay(300);
-			dv = receive_from_AP(index, z90crypt.cdx,
-					     devPtr->dev_resp_l,
-					     devPtr->dev_resp_p, psmid);
-			PDEBUG("dv returned by DQ = %d\n", dv);
-			if (dv == DEV_REC_EXCEPTION) {
-				rv = REC_FATAL_ERROR;
-				PRINTKC("exception in dequeue %d\n",
-					index);
-				break;
-			}
-			switch (dv) {
-			case DEV_ONLINE:
-				rv = 0;
-				break;
-			case DEV_EMPTY:
-				rv = REC_EMPTY;
-				break;
-			case DEV_NO_WORK:
-				rv = REC_NO_WORK;
-				break;
-			case DEV_BAD_MESSAGE:
-			case DEV_GONE:
-			default:
-				rv = REC_NO_RESPONSE;
-				break;
-			}
-			if ((rv != 0) && (rv != REC_NO_WORK))
-				break;
-			if (rv == 0)
-				break;
-		}
-		if (rv)
-			break;
-		cprbx_p = (struct CPRBX *) (devPtr->dev_resp_p + 48);
-		if ((cprbx_p->ccp_rtcode == 8) && (cprbx_p->ccp_rscode == 33)) {
-			devPtr->dev_type = PCIXCC_MCL2;
-			PDEBUG("device %d is MCL2\n", index);
-		} else {
-			devPtr->dev_type = PCIXCC_MCL3;
-			PDEBUG("device %d is MCL3\n", index);
-		}
-	} while (0);
-	/* In a general error case, the card is not marked online */
-	return rv;
-}
-
-module_init(z90crypt_init_module);
-module_exit(z90crypt_cleanup_module);
diff --git a/include/asm-s390/z90crypt.h b/include/asm-s390/z90crypt.h
deleted file mode 100644
index 31a2439b07bd..000000000000
--- a/include/asm-s390/z90crypt.h
+++ /dev/null
@@ -1,212 +0,0 @@
-/*
- *  include/asm-s390/z90crypt.h
- *
- *  z90crypt 1.3.3 (user-visible header)
- *
- *  Copyright (C)  2001, 2005 IBM Corporation
- *  Author(s): Robert Burroughs
- *             Eric Rossman (edrossma@us.ibm.com)
- *
- *  Hotplug & misc device support: Jochen Roehrig (roehrig@de.ibm.com)
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2, or (at your option)
- * any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
- */
-
-#ifndef __ASM_S390_Z90CRYPT_H
-#define __ASM_S390_Z90CRYPT_H
-#include <linux/ioctl.h>
-
-#define z90crypt_VERSION 1
-#define z90crypt_RELEASE 3	// 2 = PCIXCC, 3 = rewrite for coding standards
-#define z90crypt_VARIANT 3	// 3 = CEX2A support
-
-/**
- * struct ica_rsa_modexpo
- *
- * Requirements:
- * - outputdatalength is at least as large as inputdatalength.
- * - All key parts are right justified in their fields, padded on
- *   the left with zeroes.
- * - length(b_key) = inputdatalength
- * - length(n_modulus) = inputdatalength
- */
-struct ica_rsa_modexpo {
-	char __user *	inputdata;
-	unsigned int	inputdatalength;
-	char __user *	outputdata;
-	unsigned int	outputdatalength;
-	char __user *	b_key;
-	char __user *	n_modulus;
-};
-
-/**
- * struct ica_rsa_modexpo_crt
- *
- * Requirements:
- * - inputdatalength is even.
- * - outputdatalength is at least as large as inputdatalength.
- * - All key parts are right justified in their fields, padded on
- *   the left with zeroes.
- * - length(bp_key)	= inputdatalength/2 + 8
- * - length(bq_key)	= inputdatalength/2
- * - length(np_key)	= inputdatalength/2 + 8
- * - length(nq_key)	= inputdatalength/2
- * - length(u_mult_inv) = inputdatalength/2 + 8
- */
-struct ica_rsa_modexpo_crt {
-	char __user *	inputdata;
-	unsigned int	inputdatalength;
-	char __user *	outputdata;
-	unsigned int	outputdatalength;
-	char __user *	bp_key;
-	char __user *	bq_key;
-	char __user *	np_prime;
-	char __user *	nq_prime;
-	char __user *	u_mult_inv;
-};
-
-#define Z90_IOCTL_MAGIC 'z'  // NOTE:  Need to allocate from linux folks
-
-/**
- * Interface notes:
- *
- * The ioctl()s which are implemented (along with relevant details)
- * are:
- *
- *   ICARSAMODEXPO
- *     Perform an RSA operation using a Modulus-Exponent pair
- *     This takes an ica_rsa_modexpo struct as its arg.
- *
- *     NOTE: please refer to the comments preceding this structure
- *           for the implementation details for the contents of the
- *           block
- *
- *   ICARSACRT
- *     Perform an RSA operation using a Chinese-Remainder Theorem key
- *     This takes an ica_rsa_modexpo_crt struct as its arg.
- *
- *     NOTE: please refer to the comments preceding this structure
- *           for the implementation details for the contents of the
- *           block
- *
- *   Z90STAT_TOTALCOUNT
- *     Return an integer count of all device types together.
- *
- *   Z90STAT_PCICACOUNT
- *     Return an integer count of all PCICAs.
- *
- *   Z90STAT_PCICCCOUNT
- *     Return an integer count of all PCICCs.
- *
- *   Z90STAT_PCIXCCMCL2COUNT
- *     Return an integer count of all MCL2 PCIXCCs.
- *
- *   Z90STAT_PCIXCCMCL3COUNT
- *     Return an integer count of all MCL3 PCIXCCs.
- *
- *   Z90STAT_CEX2CCOUNT
- *     Return an integer count of all CEX2Cs.
- *
- *   Z90STAT_CEX2ACOUNT
- *     Return an integer count of all CEX2As.
- *
- *   Z90STAT_REQUESTQ_COUNT
- *     Return an integer count of the number of entries waiting to be
- *     sent to a device.
- *
- *   Z90STAT_PENDINGQ_COUNT
- *     Return an integer count of the number of entries sent to a
- *     device awaiting the reply.
- *
- *   Z90STAT_TOTALOPEN_COUNT
- *     Return an integer count of the number of open file handles.
- *
- *   Z90STAT_DOMAIN_INDEX
- *     Return the integer value of the Cryptographic Domain.
- *
- *   Z90STAT_STATUS_MASK
- *     Return an 64 element array of unsigned chars for the status of
- *     all devices.
- *       0x01: PCICA
- *       0x02: PCICC
- *       0x03: PCIXCC_MCL2
- *       0x04: PCIXCC_MCL3
- *       0x05: CEX2C
- *       0x06: CEX2A
- *       0x0d: device is disabled via the proc filesystem
- *
- *   Z90STAT_QDEPTH_MASK
- *     Return an 64 element array of unsigned chars for the queue
- *     depth of all devices.
- *
- *   Z90STAT_PERDEV_REQCNT
- *     Return an 64 element array of unsigned integers for the number
- *     of successfully completed requests per device since the device
- *     was detected and made available.
- *
- *   ICAZ90STATUS (deprecated)
- *     Return some device driver status in a ica_z90_status struct
- *     This takes an ica_z90_status struct as its arg.
- *
- *     NOTE: this ioctl() is deprecated, and has been replaced with
- *           single ioctl()s for each type of status being requested
- *
- *   Z90STAT_PCIXCCCOUNT (deprecated)
- *     Return an integer count of all PCIXCCs (MCL2 + MCL3).
- *     This is DEPRECATED now that MCL3 PCIXCCs are treated differently from
- *     MCL2 PCIXCCs.
- *
- *   Z90QUIESCE (not recommended)
- *     Quiesce the driver.  This is intended to stop all new
- *     requests from being processed.  Its use is NOT recommended,
- *     except in circumstances where there is no other way to stop
- *     callers from accessing the driver.  Its original use was to
- *     allow the driver to be "drained" of work in preparation for
- *     a system shutdown.
- *
- *     NOTE: once issued, this ban on new work cannot be undone
- *           except by unloading and reloading the driver.
- */
-
-/**
- * Supported ioctl calls
- */
-#define ICARSAMODEXPO	_IOC(_IOC_READ|_IOC_WRITE, Z90_IOCTL_MAGIC, 0x05, 0)
-#define ICARSACRT	_IOC(_IOC_READ|_IOC_WRITE, Z90_IOCTL_MAGIC, 0x06, 0)
-
-/* DEPRECATED status calls (bound for removal at some point) */
-#define ICAZ90STATUS	_IOR(Z90_IOCTL_MAGIC, 0x10, struct ica_z90_status)
-#define Z90STAT_PCIXCCCOUNT	_IOR(Z90_IOCTL_MAGIC, 0x43, int)
-
-/* unrelated to ICA callers */
-#define Z90QUIESCE	_IO(Z90_IOCTL_MAGIC, 0x11)
-
-/* New status calls */
-#define Z90STAT_TOTALCOUNT	_IOR(Z90_IOCTL_MAGIC, 0x40, int)
-#define Z90STAT_PCICACOUNT	_IOR(Z90_IOCTL_MAGIC, 0x41, int)
-#define Z90STAT_PCICCCOUNT	_IOR(Z90_IOCTL_MAGIC, 0x42, int)
-#define Z90STAT_PCIXCCMCL2COUNT	_IOR(Z90_IOCTL_MAGIC, 0x4b, int)
-#define Z90STAT_PCIXCCMCL3COUNT	_IOR(Z90_IOCTL_MAGIC, 0x4c, int)
-#define Z90STAT_CEX2CCOUNT	_IOR(Z90_IOCTL_MAGIC, 0x4d, int)
-#define Z90STAT_CEX2ACOUNT	_IOR(Z90_IOCTL_MAGIC, 0x4e, int)
-#define Z90STAT_REQUESTQ_COUNT	_IOR(Z90_IOCTL_MAGIC, 0x44, int)
-#define Z90STAT_PENDINGQ_COUNT	_IOR(Z90_IOCTL_MAGIC, 0x45, int)
-#define Z90STAT_TOTALOPEN_COUNT _IOR(Z90_IOCTL_MAGIC, 0x46, int)
-#define Z90STAT_DOMAIN_INDEX	_IOR(Z90_IOCTL_MAGIC, 0x47, int)
-#define Z90STAT_STATUS_MASK	_IOR(Z90_IOCTL_MAGIC, 0x48, char[64])
-#define Z90STAT_QDEPTH_MASK	_IOR(Z90_IOCTL_MAGIC, 0x49, char[64])
-#define Z90STAT_PERDEV_REQCNT	_IOR(Z90_IOCTL_MAGIC, 0x4a, int[64])
-
-#endif /* __ASM_S390_Z90CRYPT_H */
-- 
cgit v1.2.3


From 1534c3820c26aca4e2567f97b8add8bea40e7e2b Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Wed, 20 Sep 2006 15:58:25 +0200
Subject: [S390] zcrypt adjunct processor bus.

Add a bus for the adjunct processor interface. Up to 64 devices can
be connect to the ap bus interface, each device with 16 domains. That
makes 1024 message queues. The interface is asynchronous, the answer
to a message sent to a queue needs to be received at some later point
in time. Unfortunately the interface does not provide interrupts when
a message reply is pending. So the ap bus needs to implement some
fancy polling, each active queue is polled once per 1/HZ second or
continuously if an idle cpus exsists and the poll thread is activ
(see poll_thread parameter).

The ap bus uses the sysfs path /sys/bus/ap and has two bus attributes,
ap_domain and config_time. The ap_domain selects one of the 16 domains
to be used for this system. This limits the maximum number of ap devices
to 64. The config_time attribute contains the number of seconds between
two ap bus scans to find new devices.

The ap bus uses the modalias entries of the form "ap:tN" to autoload
the ap driver for hardware type N. Currently known types are:
3 - PCICC, 4 - PCICA, 5 - PCIXCC, 6 - CEX2A and 7 - CEX2C.

Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Ralph Wuerthner <rwuerthn@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 drivers/s390/crypto/ap_bus.c    | 1221 +++++++++++++++++++++++++++++++++++++++
 drivers/s390/crypto/ap_bus.h    |  158 +++++
 include/linux/mod_devicetable.h |   11 +
 scripts/mod/file2alias.c        |   12 +
 4 files changed, 1402 insertions(+)
 create mode 100644 drivers/s390/crypto/ap_bus.c
 create mode 100644 drivers/s390/crypto/ap_bus.h

(limited to 'include')

diff --git a/drivers/s390/crypto/ap_bus.c b/drivers/s390/crypto/ap_bus.c
new file mode 100644
index 000000000000..6ed0985c0c91
--- /dev/null
+++ b/drivers/s390/crypto/ap_bus.c
@@ -0,0 +1,1221 @@
+/*
+ * linux/drivers/s390/crypto/ap_bus.c
+ *
+ * Copyright (C) 2006 IBM Corporation
+ * Author(s): Cornelia Huck <cornelia.huck@de.ibm.com>
+ *	      Martin Schwidefsky <schwidefsky@de.ibm.com>
+ *	      Ralph Wuerthner <rwuerthn@de.ibm.com>
+ *
+ * Adjunct processor bus.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2, or (at your option)
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/workqueue.h>
+#include <linux/notifier.h>
+#include <linux/kthread.h>
+#include <linux/mutex.h>
+#include <asm/s390_rdev.h>
+
+#include "ap_bus.h"
+
+/* Some prototypes. */
+static void ap_scan_bus(void *);
+static void ap_poll_all(unsigned long);
+static void ap_poll_timeout(unsigned long);
+static int ap_poll_thread_start(void);
+static void ap_poll_thread_stop(void);
+
+/**
+ * Module description.
+ */
+MODULE_AUTHOR("IBM Corporation");
+MODULE_DESCRIPTION("Adjunct Processor Bus driver, "
+		   "Copyright 2006 IBM Corporation");
+MODULE_LICENSE("GPL");
+
+/**
+ * Module parameter
+ */
+int ap_domain_index = -1;	/* Adjunct Processor Domain Index */
+module_param_named(domain, ap_domain_index, int, 0000);
+MODULE_PARM_DESC(domain, "domain index for ap devices");
+EXPORT_SYMBOL(ap_domain_index);
+
+static int ap_thread_flag = 1;
+module_param_named(poll_thread, ap_thread_flag, int, 0000);
+MODULE_PARM_DESC(poll_thread, "Turn on/off poll thread, default is 1 (on).");
+
+static struct device *ap_root_device = NULL;
+
+/**
+ * Workqueue & timer for bus rescan.
+ */
+static struct workqueue_struct *ap_work_queue;
+static struct timer_list ap_config_timer;
+static int ap_config_time = AP_CONFIG_TIME;
+static DECLARE_WORK(ap_config_work, ap_scan_bus, NULL);
+
+/**
+ * Tasklet & timer for AP request polling.
+ */
+static struct timer_list ap_poll_timer = TIMER_INITIALIZER(ap_poll_timeout,0,0);
+static DECLARE_TASKLET(ap_tasklet, ap_poll_all, 0);
+static atomic_t ap_poll_requests = ATOMIC_INIT(0);
+static DECLARE_WAIT_QUEUE_HEAD(ap_poll_wait);
+static struct task_struct *ap_poll_kthread = NULL;
+static DEFINE_MUTEX(ap_poll_thread_mutex);
+
+/**
+ * Test if ap instructions are available.
+ *
+ * Returns 0 if the ap instructions are installed.
+ */
+static inline int ap_instructions_available(void)
+{
+	register unsigned long reg0 asm ("0") = AP_MKQID(0,0);
+	register unsigned long reg1 asm ("1") = -ENODEV;
+	register unsigned long reg2 asm ("2") = 0UL;
+
+	asm volatile(
+		"   .long 0xb2af0000\n"		/* PQAP(TAPQ) */
+		"0: la    %1,0\n"
+		"1:\n"
+		EX_TABLE(0b, 1b)
+		: "+d" (reg0), "+d" (reg1), "+d" (reg2) : : "cc" );
+	return reg1;
+}
+
+/**
+ * Test adjunct processor queue.
+ * @qid: the ap queue number
+ * @queue_depth: pointer to queue depth value
+ * @device_type: pointer to device type value
+ *
+ * Returns ap queue status structure.
+ */
+static inline struct ap_queue_status
+ap_test_queue(ap_qid_t qid, int *queue_depth, int *device_type)
+{
+	register unsigned long reg0 asm ("0") = qid;
+	register struct ap_queue_status reg1 asm ("1");
+	register unsigned long reg2 asm ("2") = 0UL;
+
+	asm volatile(".long 0xb2af0000"		/* PQAP(TAPQ) */
+		     : "+d" (reg0), "=d" (reg1), "+d" (reg2) : : "cc");
+	*device_type = (int) (reg2 >> 24);
+	*queue_depth = (int) (reg2 & 0xff);
+	return reg1;
+}
+
+/**
+ * Reset adjunct processor queue.
+ * @qid: the ap queue number
+ *
+ * Returns ap queue status structure.
+ */
+static inline struct ap_queue_status ap_reset_queue(ap_qid_t qid)
+{
+	register unsigned long reg0 asm ("0") = qid | 0x01000000UL;
+	register struct ap_queue_status reg1 asm ("1");
+	register unsigned long reg2 asm ("2") = 0UL;
+
+	asm volatile(
+		".long 0xb2af0000"		/* PQAP(RAPQ) */
+		: "+d" (reg0), "=d" (reg1), "+d" (reg2) : : "cc");
+	return reg1;
+}
+
+/**
+ * Send message to adjunct processor queue.
+ * @qid: the ap queue number
+ * @psmid: the program supplied message identifier
+ * @msg: the message text
+ * @length: the message length
+ *
+ * Returns ap queue status structure.
+ *
+ * Condition code 1 on NQAP can't happen because the L bit is 1.
+ *
+ * Condition code 2 on NQAP also means the send is incomplete,
+ * because a segment boundary was reached. The NQAP is repeated.
+ */
+static inline struct ap_queue_status
+__ap_send(ap_qid_t qid, unsigned long long psmid, void *msg, size_t length)
+{
+	typedef struct { char _[length]; } msgblock;
+	register unsigned long reg0 asm ("0") = qid | 0x40000000UL;
+	register struct ap_queue_status reg1 asm ("1");
+	register unsigned long reg2 asm ("2") = (unsigned long) msg;
+	register unsigned long reg3 asm ("3") = (unsigned long) length;
+	register unsigned long reg4 asm ("4") = (unsigned int) (psmid >> 32);
+	register unsigned long reg5 asm ("5") = (unsigned int) psmid;
+
+	asm volatile (
+		"0: .long 0xb2ad0042\n"		/* DQAP */
+		"   brc   2,0b"
+		: "+d" (reg0), "=d" (reg1), "+d" (reg2), "+d" (reg3)
+		: "d" (reg4), "d" (reg5), "m" (*(msgblock *) msg)
+		: "cc" );
+	return reg1;
+}
+
+int ap_send(ap_qid_t qid, unsigned long long psmid, void *msg, size_t length)
+{
+	struct ap_queue_status status;
+
+	status = __ap_send(qid, psmid, msg, length);
+	switch (status.response_code) {
+	case AP_RESPONSE_NORMAL:
+		return 0;
+	case AP_RESPONSE_Q_FULL:
+		return -EBUSY;
+	default:	/* Device is gone. */
+		return -ENODEV;
+	}
+}
+EXPORT_SYMBOL(ap_send);
+
+/*
+ * Receive message from adjunct processor queue.
+ * @qid: the ap queue number
+ * @psmid: pointer to program supplied message identifier
+ * @msg: the message text
+ * @length: the message length
+ *
+ * Returns ap queue status structure.
+ *
+ * Condition code 1 on DQAP means the receive has taken place
+ * but only partially.	The response is incomplete, hence the
+ * DQAP is repeated.
+ *
+ * Condition code 2 on DQAP also means the receive is incomplete,
+ * this time because a segment boundary was reached. Again, the
+ * DQAP is repeated.
+ *
+ * Note that gpr2 is used by the DQAP instruction to keep track of
+ * any 'residual' length, in case the instruction gets interrupted.
+ * Hence it gets zeroed before the instruction.
+ */
+static inline struct ap_queue_status
+__ap_recv(ap_qid_t qid, unsigned long long *psmid, void *msg, size_t length)
+{
+	typedef struct { char _[length]; } msgblock;
+	register unsigned long reg0 asm("0") = qid | 0x80000000UL;
+	register struct ap_queue_status reg1 asm ("1");
+	register unsigned long reg2 asm("2") = 0UL;
+	register unsigned long reg4 asm("4") = (unsigned long) msg;
+	register unsigned long reg5 asm("5") = (unsigned long) length;
+	register unsigned long reg6 asm("6") = 0UL;
+	register unsigned long reg7 asm("7") = 0UL;
+
+
+	asm volatile(
+		"0: .long 0xb2ae0064\n"
+		"   brc   6,0b\n"
+		: "+d" (reg0), "=d" (reg1), "+d" (reg2),
+		"+d" (reg4), "+d" (reg5), "+d" (reg6), "+d" (reg7),
+		"=m" (*(msgblock *) msg) : : "cc" );
+	*psmid = (((unsigned long long) reg6) << 32) + reg7;
+	return reg1;
+}
+
+int ap_recv(ap_qid_t qid, unsigned long long *psmid, void *msg, size_t length)
+{
+	struct ap_queue_status status;
+
+	status = __ap_recv(qid, psmid, msg, length);
+	switch (status.response_code) {
+	case AP_RESPONSE_NORMAL:
+		return 0;
+	case AP_RESPONSE_NO_PENDING_REPLY:
+		if (status.queue_empty)
+			return -ENOENT;
+		return -EBUSY;
+	default:
+		return -ENODEV;
+	}
+}
+EXPORT_SYMBOL(ap_recv);
+
+/**
+ * Check if an AP queue is available. The test is repeated for
+ * AP_MAX_RESET times.
+ * @qid: the ap queue number
+ * @queue_depth: pointer to queue depth value
+ * @device_type: pointer to device type value
+ */
+static int ap_query_queue(ap_qid_t qid, int *queue_depth, int *device_type)
+{
+	struct ap_queue_status status;
+	int t_depth, t_device_type, rc, i;
+
+	rc = -EBUSY;
+	for (i = 0; i < AP_MAX_RESET; i++) {
+		status = ap_test_queue(qid, &t_depth, &t_device_type);
+		switch (status.response_code) {
+		case AP_RESPONSE_NORMAL:
+			*queue_depth = t_depth + 1;
+			*device_type = t_device_type;
+			rc = 0;
+			break;
+		case AP_RESPONSE_Q_NOT_AVAIL:
+			rc = -ENODEV;
+			break;
+		case AP_RESPONSE_RESET_IN_PROGRESS:
+			break;
+		case AP_RESPONSE_DECONFIGURED:
+			rc = -ENODEV;
+			break;
+		case AP_RESPONSE_CHECKSTOPPED:
+			rc = -ENODEV;
+			break;
+		case AP_RESPONSE_BUSY:
+			break;
+		default:
+			BUG();
+		}
+		if (rc != -EBUSY)
+			break;
+		if (i < AP_MAX_RESET - 1)
+			udelay(5);
+	}
+	return rc;
+}
+
+/**
+ * Reset an AP queue and wait for it to become available again.
+ * @qid: the ap queue number
+ */
+static int ap_init_queue(ap_qid_t qid)
+{
+	struct ap_queue_status status;
+	int rc, dummy, i;
+
+	rc = -ENODEV;
+	status = ap_reset_queue(qid);
+	for (i = 0; i < AP_MAX_RESET; i++) {
+		switch (status.response_code) {
+		case AP_RESPONSE_NORMAL:
+			if (status.queue_empty)
+				rc = 0;
+			break;
+		case AP_RESPONSE_Q_NOT_AVAIL:
+		case AP_RESPONSE_DECONFIGURED:
+		case AP_RESPONSE_CHECKSTOPPED:
+			i = AP_MAX_RESET;	/* return with -ENODEV */
+			break;
+		case AP_RESPONSE_RESET_IN_PROGRESS:
+		case AP_RESPONSE_BUSY:
+		default:
+			break;
+		}
+		if (rc != -ENODEV)
+			break;
+		if (i < AP_MAX_RESET - 1) {
+			udelay(5);
+			status = ap_test_queue(qid, &dummy, &dummy);
+		}
+	}
+	return rc;
+}
+
+/**
+ * AP device related attributes.
+ */
+static ssize_t ap_hwtype_show(struct device *dev,
+			      struct device_attribute *attr, char *buf)
+{
+	struct ap_device *ap_dev = to_ap_dev(dev);
+	return snprintf(buf, PAGE_SIZE, "%d\n", ap_dev->device_type);
+}
+static DEVICE_ATTR(hwtype, 0444, ap_hwtype_show, NULL);
+
+static ssize_t ap_depth_show(struct device *dev, struct device_attribute *attr,
+			     char *buf)
+{
+	struct ap_device *ap_dev = to_ap_dev(dev);
+	return snprintf(buf, PAGE_SIZE, "%d\n", ap_dev->queue_depth);
+}
+static DEVICE_ATTR(depth, 0444, ap_depth_show, NULL);
+
+static ssize_t ap_request_count_show(struct device *dev,
+				     struct device_attribute *attr,
+				     char *buf)
+{
+	struct ap_device *ap_dev = to_ap_dev(dev);
+	int rc;
+
+	spin_lock_bh(&ap_dev->lock);
+	rc = snprintf(buf, PAGE_SIZE, "%d\n", ap_dev->total_request_count);
+	spin_unlock_bh(&ap_dev->lock);
+	return rc;
+}
+
+static DEVICE_ATTR(request_count, 0444, ap_request_count_show, NULL);
+
+static ssize_t ap_modalias_show(struct device *dev,
+				struct device_attribute *attr, char *buf)
+{
+	return sprintf(buf, "ap:t%02X", to_ap_dev(dev)->device_type);
+}
+
+static DEVICE_ATTR(modalias, 0444, ap_modalias_show, NULL);
+
+static struct attribute *ap_dev_attrs[] = {
+	&dev_attr_hwtype.attr,
+	&dev_attr_depth.attr,
+	&dev_attr_request_count.attr,
+	&dev_attr_modalias.attr,
+	NULL
+};
+static struct attribute_group ap_dev_attr_group = {
+	.attrs = ap_dev_attrs
+};
+
+/**
+ * AP bus driver registration/unregistration.
+ */
+static int ap_bus_match(struct device *dev, struct device_driver *drv)
+{
+	struct ap_device *ap_dev = to_ap_dev(dev);
+	struct ap_driver *ap_drv = to_ap_drv(drv);
+	struct ap_device_id *id;
+
+	/**
+	 * Compare device type of the device with the list of
+	 * supported types of the device_driver.
+	 */
+	for (id = ap_drv->ids; id->match_flags; id++) {
+		if ((id->match_flags & AP_DEVICE_ID_MATCH_DEVICE_TYPE) &&
+		    (id->dev_type != ap_dev->device_type))
+			continue;
+		return 1;
+	}
+	return 0;
+}
+
+/**
+ * uevent function for AP devices. It sets up a single environment
+ * variable DEV_TYPE which contains the hardware device type.
+ */
+static int ap_uevent (struct device *dev, char **envp, int num_envp,
+		       char *buffer, int buffer_size)
+{
+	struct ap_device *ap_dev = to_ap_dev(dev);
+	int length;
+
+	if (!ap_dev)
+		return -ENODEV;
+
+	/* Set up DEV_TYPE environment variable. */
+	envp[0] = buffer;
+	length = scnprintf(buffer, buffer_size, "DEV_TYPE=%04X",
+			   ap_dev->device_type);
+	if (buffer_size - length <= 0)
+		return -ENOMEM;
+	envp[1] = 0;
+	return 0;
+}
+
+static struct bus_type ap_bus_type = {
+	.name = "ap",
+	.match = &ap_bus_match,
+	.uevent = &ap_uevent,
+};
+
+static int ap_device_probe(struct device *dev)
+{
+	struct ap_device *ap_dev = to_ap_dev(dev);
+	struct ap_driver *ap_drv = to_ap_drv(dev->driver);
+	int rc;
+
+	ap_dev->drv = ap_drv;
+	rc = ap_drv->probe ? ap_drv->probe(ap_dev) : -ENODEV;
+	if (rc)
+		ap_dev->unregistered = 1;
+	return rc;
+}
+
+/**
+ * Flush all requests from the request/pending queue of an AP device.
+ * @ap_dev: pointer to the AP device.
+ */
+static inline void __ap_flush_queue(struct ap_device *ap_dev)
+{
+	struct ap_message *ap_msg, *next;
+
+	list_for_each_entry_safe(ap_msg, next, &ap_dev->pendingq, list) {
+		list_del_init(&ap_msg->list);
+		ap_dev->pendingq_count--;
+		ap_dev->drv->receive(ap_dev, ap_msg, ERR_PTR(-ENODEV));
+	}
+	list_for_each_entry_safe(ap_msg, next, &ap_dev->requestq, list) {
+		list_del_init(&ap_msg->list);
+		ap_dev->requestq_count--;
+		ap_dev->drv->receive(ap_dev, ap_msg, ERR_PTR(-ENODEV));
+	}
+}
+
+void ap_flush_queue(struct ap_device *ap_dev)
+{
+	spin_lock_bh(&ap_dev->lock);
+	__ap_flush_queue(ap_dev);
+	spin_unlock_bh(&ap_dev->lock);
+}
+EXPORT_SYMBOL(ap_flush_queue);
+
+static int ap_device_remove(struct device *dev)
+{
+	struct ap_device *ap_dev = to_ap_dev(dev);
+	struct ap_driver *ap_drv = ap_dev->drv;
+
+	spin_lock_bh(&ap_dev->lock);
+	__ap_flush_queue(ap_dev);
+	/**
+	 * set ->unregistered to 1 while holding the lock. This prevents
+	 * new messages to be put on the queue from now on.
+	 */
+	ap_dev->unregistered = 1;
+	spin_unlock_bh(&ap_dev->lock);
+	if (ap_drv->remove)
+		ap_drv->remove(ap_dev);
+	return 0;
+}
+
+int ap_driver_register(struct ap_driver *ap_drv, struct module *owner,
+		       char *name)
+{
+	struct device_driver *drv = &ap_drv->driver;
+
+	drv->bus = &ap_bus_type;
+	drv->probe = ap_device_probe;
+	drv->remove = ap_device_remove;
+	drv->owner = owner;
+	drv->name = name;
+	return driver_register(drv);
+}
+EXPORT_SYMBOL(ap_driver_register);
+
+void ap_driver_unregister(struct ap_driver *ap_drv)
+{
+	driver_unregister(&ap_drv->driver);
+}
+EXPORT_SYMBOL(ap_driver_unregister);
+
+/**
+ * AP bus attributes.
+ */
+static ssize_t ap_domain_show(struct bus_type *bus, char *buf)
+{
+	return snprintf(buf, PAGE_SIZE, "%d\n", ap_domain_index);
+}
+
+static BUS_ATTR(ap_domain, 0444, ap_domain_show, NULL);
+
+static ssize_t ap_config_time_show(struct bus_type *bus, char *buf)
+{
+	return snprintf(buf, PAGE_SIZE, "%d\n", ap_config_time);
+}
+
+static ssize_t ap_config_time_store(struct bus_type *bus,
+				    const char *buf, size_t count)
+{
+	int time;
+
+	if (sscanf(buf, "%d\n", &time) != 1 || time < 5 || time > 120)
+		return -EINVAL;
+	ap_config_time = time;
+	if (!timer_pending(&ap_config_timer) ||
+	    !mod_timer(&ap_config_timer, jiffies + ap_config_time * HZ)) {
+		ap_config_timer.expires = jiffies + ap_config_time * HZ;
+		add_timer(&ap_config_timer);
+	}
+	return count;
+}
+
+static BUS_ATTR(config_time, 0644, ap_config_time_show, ap_config_time_store);
+
+static ssize_t ap_poll_thread_show(struct bus_type *bus, char *buf)
+{
+	return snprintf(buf, PAGE_SIZE, "%d\n", ap_poll_kthread ? 1 : 0);
+}
+
+static ssize_t ap_poll_thread_store(struct bus_type *bus,
+				    const char *buf, size_t count)
+{
+	int flag, rc;
+
+	if (sscanf(buf, "%d\n", &flag) != 1)
+		return -EINVAL;
+	if (flag) {
+		rc = ap_poll_thread_start();
+		if (rc)
+			return rc;
+	}
+	else
+		ap_poll_thread_stop();
+	return count;
+}
+
+static BUS_ATTR(poll_thread, 0644, ap_poll_thread_show, ap_poll_thread_store);
+
+static struct bus_attribute *const ap_bus_attrs[] = {
+	&bus_attr_ap_domain,
+	&bus_attr_config_time,
+	&bus_attr_poll_thread,
+	NULL
+};
+
+/**
+ * Pick one of the 16 ap domains.
+ */
+static inline int ap_select_domain(void)
+{
+	int queue_depth, device_type, count, max_count, best_domain;
+	int rc, i, j;
+
+	/**
+	 * We want to use a single domain. Either the one specified with
+	 * the "domain=" parameter or the domain with the maximum number
+	 * of devices.
+	 */
+	if (ap_domain_index >= 0 && ap_domain_index < AP_DOMAINS)
+		/* Domain has already been selected. */
+		return 0;
+	best_domain = -1;
+	max_count = 0;
+	for (i = 0; i < AP_DOMAINS; i++) {
+		count = 0;
+		for (j = 0; j < AP_DEVICES; j++) {
+			ap_qid_t qid = AP_MKQID(j, i);
+			rc = ap_query_queue(qid, &queue_depth, &device_type);
+			if (rc)
+				continue;
+			count++;
+		}
+		if (count > max_count) {
+			max_count = count;
+			best_domain = i;
+		}
+	}
+	if (best_domain >= 0){
+		ap_domain_index = best_domain;
+		return 0;
+	}
+	return -ENODEV;
+}
+
+/**
+ * Find the device type if query queue returned a device type of 0.
+ * @ap_dev: pointer to the AP device.
+ */
+static int ap_probe_device_type(struct ap_device *ap_dev)
+{
+	static unsigned char msg[] = {
+		0x00,0x06,0x00,0x00,0x00,0x00,0x00,0x00,
+		0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
+		0x00,0x00,0x00,0x58,0x00,0x00,0x00,0x00,
+		0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
+		0x01,0x00,0x43,0x43,0x41,0x2d,0x41,0x50,
+		0x50,0x4c,0x20,0x20,0x20,0x01,0x01,0x01,
+		0x00,0x00,0x00,0x00,0x50,0x4b,0x00,0x00,
+		0x00,0x00,0x01,0x1c,0x00,0x00,0x00,0x00,
+		0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
+		0x00,0x00,0x05,0xb8,0x00,0x00,0x00,0x00,
+		0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
+		0x70,0x00,0x41,0x00,0x00,0x00,0x00,0x00,
+		0x00,0x00,0x54,0x32,0x01,0x00,0xa0,0x00,
+		0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
+		0x00,0x00,0x00,0x00,0xb8,0x05,0x00,0x00,
+		0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
+		0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
+		0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
+		0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
+		0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
+		0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
+		0x00,0x00,0x0a,0x00,0x00,0x00,0x00,0x00,
+		0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
+		0x00,0x00,0x00,0x00,0x00,0x00,0x08,0x00,
+		0x49,0x43,0x53,0x46,0x20,0x20,0x20,0x20,
+		0x50,0x4b,0x0a,0x00,0x50,0x4b,0x43,0x53,
+		0x2d,0x31,0x2e,0x32,0x37,0x00,0x11,0x22,
+		0x33,0x44,0x55,0x66,0x77,0x88,0x99,0x00,
+		0x11,0x22,0x33,0x44,0x55,0x66,0x77,0x88,
+		0x99,0x00,0x11,0x22,0x33,0x44,0x55,0x66,
+		0x77,0x88,0x99,0x00,0x11,0x22,0x33,0x44,
+		0x55,0x66,0x77,0x88,0x99,0x00,0x11,0x22,
+		0x33,0x44,0x55,0x66,0x77,0x88,0x99,0x00,
+		0x11,0x22,0x33,0x5d,0x00,0x5b,0x00,0x77,
+		0x88,0x1e,0x00,0x00,0x57,0x00,0x00,0x00,
+		0x00,0x04,0x00,0x00,0x4f,0x00,0x00,0x00,
+		0x03,0x02,0x00,0x00,0x40,0x01,0x00,0x01,
+		0xce,0x02,0x68,0x2d,0x5f,0xa9,0xde,0x0c,
+		0xf6,0xd2,0x7b,0x58,0x4b,0xf9,0x28,0x68,
+		0x3d,0xb4,0xf4,0xef,0x78,0xd5,0xbe,0x66,
+		0x63,0x42,0xef,0xf8,0xfd,0xa4,0xf8,0xb0,
+		0x8e,0x29,0xc2,0xc9,0x2e,0xd8,0x45,0xb8,
+		0x53,0x8c,0x6f,0x4e,0x72,0x8f,0x6c,0x04,
+		0x9c,0x88,0xfc,0x1e,0xc5,0x83,0x55,0x57,
+		0xf7,0xdd,0xfd,0x4f,0x11,0x36,0x95,0x5d,
+	};
+	struct ap_queue_status status;
+	unsigned long long psmid;
+	char *reply;
+	int rc, i;
+
+	reply = (void *) get_zeroed_page(GFP_KERNEL);
+	if (!reply) {
+		rc = -ENOMEM;
+		goto out;
+	}
+
+	status = __ap_send(ap_dev->qid, 0x0102030405060708ULL,
+			   msg, sizeof(msg));
+	if (status.response_code != AP_RESPONSE_NORMAL) {
+		rc = -ENODEV;
+		goto out_free;
+	}
+
+	/* Wait for the test message to complete. */
+	for (i = 0; i < 6; i++) {
+		mdelay(300);
+		status = __ap_recv(ap_dev->qid, &psmid, reply, 4096);
+		if (status.response_code == AP_RESPONSE_NORMAL &&
+		    psmid == 0x0102030405060708ULL)
+			break;
+	}
+	if (i < 6) {
+		/* Got an answer. */
+		if (reply[0] == 0x00 && reply[1] == 0x86)
+			ap_dev->device_type = AP_DEVICE_TYPE_PCICC;
+		else
+			ap_dev->device_type = AP_DEVICE_TYPE_PCICA;
+		rc = 0;
+	} else
+		rc = -ENODEV;
+
+out_free:
+	free_page((unsigned long) reply);
+out:
+	return rc;
+}
+
+/**
+ * Scan the ap bus for new devices.
+ */
+static int __ap_scan_bus(struct device *dev, void *data)
+{
+	return to_ap_dev(dev)->qid == (ap_qid_t)(unsigned long) data;
+}
+
+static void ap_device_release(struct device *dev)
+{
+	struct ap_device *ap_dev = to_ap_dev(dev);
+
+	kfree(ap_dev);
+}
+
+static void ap_scan_bus(void *data)
+{
+	struct ap_device *ap_dev;
+	struct device *dev;
+	ap_qid_t qid;
+	int queue_depth, device_type;
+	int rc, i;
+
+	if (ap_select_domain() != 0)
+		return;
+	for (i = 0; i < AP_DEVICES; i++) {
+		qid = AP_MKQID(i, ap_domain_index);
+		dev = bus_find_device(&ap_bus_type, NULL,
+				      (void *)(unsigned long)qid,
+				      __ap_scan_bus);
+		if (dev) {
+			put_device(dev);
+			continue;
+		}
+		rc = ap_query_queue(qid, &queue_depth, &device_type);
+		if (rc)
+			continue;
+		rc = ap_init_queue(qid);
+		if (rc)
+			continue;
+		ap_dev = kzalloc(sizeof(*ap_dev), GFP_KERNEL);
+		if (!ap_dev)
+			break;
+		ap_dev->qid = qid;
+		ap_dev->queue_depth = queue_depth;
+		spin_lock_init(&ap_dev->lock);
+		INIT_LIST_HEAD(&ap_dev->pendingq);
+		INIT_LIST_HEAD(&ap_dev->requestq);
+		if (device_type == 0)
+			ap_probe_device_type(ap_dev);
+		else
+			ap_dev->device_type = device_type;
+
+		ap_dev->device.bus = &ap_bus_type;
+		ap_dev->device.parent = ap_root_device;
+		snprintf(ap_dev->device.bus_id, BUS_ID_SIZE, "card%02x",
+			 AP_QID_DEVICE(ap_dev->qid));
+		ap_dev->device.release = ap_device_release;
+		rc = device_register(&ap_dev->device);
+		if (rc) {
+			kfree(ap_dev);
+			continue;
+		}
+		/* Add device attributes. */
+		rc = sysfs_create_group(&ap_dev->device.kobj,
+					&ap_dev_attr_group);
+		if (rc)
+			device_unregister(&ap_dev->device);
+	}
+}
+
+static void
+ap_config_timeout(unsigned long ptr)
+{
+	queue_work(ap_work_queue, &ap_config_work);
+	ap_config_timer.expires = jiffies + ap_config_time * HZ;
+	add_timer(&ap_config_timer);
+}
+
+/**
+ * Set up the timer to run the poll tasklet
+ */
+static inline void ap_schedule_poll_timer(void)
+{
+	if (timer_pending(&ap_poll_timer))
+		return;
+	mod_timer(&ap_poll_timer, jiffies + AP_POLL_TIME);
+}
+
+/**
+ * Receive pending reply messages from an AP device.
+ * @ap_dev: pointer to the AP device
+ * @flags: pointer to control flags, bit 2^0 is set if another poll is
+ *	   required, bit 2^1 is set if the poll timer needs to get armed
+ * Returns 0 if the device is still present, -ENODEV if not.
+ */
+static inline int ap_poll_read(struct ap_device *ap_dev, unsigned long *flags)
+{
+	struct ap_queue_status status;
+	struct ap_message *ap_msg;
+
+	if (ap_dev->queue_count <= 0)
+		return 0;
+	status = __ap_recv(ap_dev->qid, &ap_dev->reply->psmid,
+			   ap_dev->reply->message, ap_dev->reply->length);
+	switch (status.response_code) {
+	case AP_RESPONSE_NORMAL:
+		atomic_dec(&ap_poll_requests);
+		ap_dev->queue_count--;
+		list_for_each_entry(ap_msg, &ap_dev->pendingq, list) {
+			if (ap_msg->psmid != ap_dev->reply->psmid)
+				continue;
+			list_del_init(&ap_msg->list);
+			ap_dev->pendingq_count--;
+			ap_dev->drv->receive(ap_dev, ap_msg, ap_dev->reply);
+			break;
+		}
+		if (ap_dev->queue_count > 0)
+			*flags |= 1;
+		break;
+	case AP_RESPONSE_NO_PENDING_REPLY:
+		if (status.queue_empty) {
+			/* The card shouldn't forget requests but who knows. */
+			ap_dev->queue_count = 0;
+			list_splice_init(&ap_dev->pendingq, &ap_dev->requestq);
+			ap_dev->requestq_count += ap_dev->pendingq_count;
+			ap_dev->pendingq_count = 0;
+		} else
+			*flags |= 2;
+		break;
+	default:
+		return -ENODEV;
+	}
+	return 0;
+}
+
+/**
+ * Send messages from the request queue to an AP device.
+ * @ap_dev: pointer to the AP device
+ * @flags: pointer to control flags, bit 2^0 is set if another poll is
+ *	   required, bit 2^1 is set if the poll timer needs to get armed
+ * Returns 0 if the device is still present, -ENODEV if not.
+ */
+static inline int ap_poll_write(struct ap_device *ap_dev, unsigned long *flags)
+{
+	struct ap_queue_status status;
+	struct ap_message *ap_msg;
+
+	if (ap_dev->requestq_count <= 0 ||
+	    ap_dev->queue_count >= ap_dev->queue_depth)
+		return 0;
+	/* Start the next request on the queue. */
+	ap_msg = list_entry(ap_dev->requestq.next, struct ap_message, list);
+	status = __ap_send(ap_dev->qid, ap_msg->psmid,
+			   ap_msg->message, ap_msg->length);
+	switch (status.response_code) {
+	case AP_RESPONSE_NORMAL:
+		atomic_inc(&ap_poll_requests);
+		ap_dev->queue_count++;
+		list_move_tail(&ap_msg->list, &ap_dev->pendingq);
+		ap_dev->requestq_count--;
+		ap_dev->pendingq_count++;
+		if (ap_dev->queue_count < ap_dev->queue_depth &&
+		    ap_dev->requestq_count > 0)
+			*flags |= 1;
+		*flags |= 2;
+		break;
+	case AP_RESPONSE_Q_FULL:
+		*flags |= 2;
+		break;
+	case AP_RESPONSE_MESSAGE_TOO_BIG:
+		return -EINVAL;
+	default:
+		return -ENODEV;
+	}
+	return 0;
+}
+
+/**
+ * Poll AP device for pending replies and send new messages. If either
+ * ap_poll_read or ap_poll_write returns -ENODEV unregister the device.
+ * @ap_dev: pointer to the bus device
+ * @flags: pointer to control flags, bit 2^0 is set if another poll is
+ *	   required, bit 2^1 is set if the poll timer needs to get armed
+ * Returns 0.
+ */
+static inline int ap_poll_queue(struct ap_device *ap_dev, unsigned long *flags)
+{
+	int rc;
+
+	rc = ap_poll_read(ap_dev, flags);
+	if (rc)
+		return rc;
+	return ap_poll_write(ap_dev, flags);
+}
+
+/**
+ * Queue a message to a device.
+ * @ap_dev: pointer to the AP device
+ * @ap_msg: the message to be queued
+ */
+static int __ap_queue_message(struct ap_device *ap_dev, struct ap_message *ap_msg)
+{
+	struct ap_queue_status status;
+
+	if (list_empty(&ap_dev->requestq) &&
+	    ap_dev->queue_count < ap_dev->queue_depth) {
+		status = __ap_send(ap_dev->qid, ap_msg->psmid,
+				   ap_msg->message, ap_msg->length);
+		switch (status.response_code) {
+		case AP_RESPONSE_NORMAL:
+			list_add_tail(&ap_msg->list, &ap_dev->pendingq);
+			atomic_inc(&ap_poll_requests);
+			ap_dev->pendingq_count++;
+			ap_dev->queue_count++;
+			ap_dev->total_request_count++;
+			break;
+		case AP_RESPONSE_Q_FULL:
+			list_add_tail(&ap_msg->list, &ap_dev->requestq);
+			ap_dev->requestq_count++;
+			ap_dev->total_request_count++;
+			return -EBUSY;
+		case AP_RESPONSE_MESSAGE_TOO_BIG:
+			ap_dev->drv->receive(ap_dev, ap_msg, ERR_PTR(-EINVAL));
+			return -EINVAL;
+		default:	/* Device is gone. */
+			ap_dev->drv->receive(ap_dev, ap_msg, ERR_PTR(-ENODEV));
+			return -ENODEV;
+		}
+	} else {
+		list_add_tail(&ap_msg->list, &ap_dev->requestq);
+		ap_dev->requestq_count++;
+		ap_dev->total_request_count++;
+		return -EBUSY;
+	}
+	ap_schedule_poll_timer();
+	return 0;
+}
+
+void ap_queue_message(struct ap_device *ap_dev, struct ap_message *ap_msg)
+{
+	unsigned long flags;
+	int rc;
+
+	spin_lock_bh(&ap_dev->lock);
+	if (!ap_dev->unregistered) {
+		/* Make room on the queue by polling for finished requests. */
+		rc = ap_poll_queue(ap_dev, &flags);
+		if (!rc)
+			rc = __ap_queue_message(ap_dev, ap_msg);
+		if (!rc)
+			wake_up(&ap_poll_wait);
+	} else {
+		ap_dev->drv->receive(ap_dev, ap_msg, ERR_PTR(-ENODEV));
+		rc = 0;
+	}
+	spin_unlock_bh(&ap_dev->lock);
+	if (rc == -ENODEV)
+		device_unregister(&ap_dev->device);
+}
+EXPORT_SYMBOL(ap_queue_message);
+
+/**
+ * Cancel a crypto request. This is done by removing the request
+ * from the devive pendingq or requestq queue. Note that the
+ * request stays on the AP queue. When it finishes the message
+ * reply will be discarded because the psmid can't be found.
+ * @ap_dev: AP device that has the message queued
+ * @ap_msg: the message that is to be removed
+ */
+void ap_cancel_message(struct ap_device *ap_dev, struct ap_message *ap_msg)
+{
+	struct ap_message *tmp;
+
+	spin_lock_bh(&ap_dev->lock);
+	if (!list_empty(&ap_msg->list)) {
+		list_for_each_entry(tmp, &ap_dev->pendingq, list)
+			if (tmp->psmid == ap_msg->psmid) {
+				ap_dev->pendingq_count--;
+				goto found;
+			}
+		ap_dev->requestq_count--;
+	found:
+		list_del_init(&ap_msg->list);
+	}
+	spin_unlock_bh(&ap_dev->lock);
+}
+EXPORT_SYMBOL(ap_cancel_message);
+
+/**
+ * AP receive polling for finished AP requests
+ */
+static void ap_poll_timeout(unsigned long unused)
+{
+	tasklet_schedule(&ap_tasklet);
+}
+
+/**
+ * Poll all AP devices on the bus in a round robin fashion. Continue
+ * polling until bit 2^0 of the control flags is not set. If bit 2^1
+ * of the control flags has been set arm the poll timer.
+ */
+static int __ap_poll_all(struct device *dev, void *data)
+{
+	struct ap_device *ap_dev = to_ap_dev(dev);
+	int rc;
+
+	spin_lock(&ap_dev->lock);
+	if (!ap_dev->unregistered) {
+		rc = ap_poll_queue(to_ap_dev(dev), (unsigned long *) data);
+	} else
+		rc = 0;
+	spin_unlock(&ap_dev->lock);
+	if (rc)
+		device_unregister(&ap_dev->device);
+	return 0;
+}
+
+static void ap_poll_all(unsigned long dummy)
+{
+	unsigned long flags;
+
+	do {
+		flags = 0;
+		bus_for_each_dev(&ap_bus_type, NULL, &flags, __ap_poll_all);
+	} while (flags & 1);
+	if (flags & 2)
+		ap_schedule_poll_timer();
+}
+
+/**
+ * AP bus poll thread. The purpose of this thread is to poll for
+ * finished requests in a loop if there is a "free" cpu - that is
+ * a cpu that doesn't have anything better to do. The polling stops
+ * as soon as there is another task or if all messages have been
+ * delivered.
+ */
+static int ap_poll_thread(void *data)
+{
+	DECLARE_WAITQUEUE(wait, current);
+	unsigned long flags;
+	int requests;
+
+	set_user_nice(current, -20);
+	while (1) {
+		if (need_resched()) {
+			schedule();
+			continue;
+		}
+		add_wait_queue(&ap_poll_wait, &wait);
+		set_current_state(TASK_INTERRUPTIBLE);
+		if (kthread_should_stop())
+			break;
+		requests = atomic_read(&ap_poll_requests);
+		if (requests <= 0)
+			schedule();
+		set_current_state(TASK_RUNNING);
+		remove_wait_queue(&ap_poll_wait, &wait);
+
+		local_bh_disable();
+		flags = 0;
+		bus_for_each_dev(&ap_bus_type, NULL, &flags, __ap_poll_all);
+		local_bh_enable();
+	}
+	set_current_state(TASK_RUNNING);
+	remove_wait_queue(&ap_poll_wait, &wait);
+	return 0;
+}
+
+static int ap_poll_thread_start(void)
+{
+	int rc;
+
+	mutex_lock(&ap_poll_thread_mutex);
+	if (!ap_poll_kthread) {
+		ap_poll_kthread = kthread_run(ap_poll_thread, NULL, "appoll");
+		rc = IS_ERR(ap_poll_kthread) ? PTR_ERR(ap_poll_kthread) : 0;
+		if (rc)
+			ap_poll_kthread = NULL;
+	}
+	else
+		rc = 0;
+	mutex_unlock(&ap_poll_thread_mutex);
+	return rc;
+}
+
+static void ap_poll_thread_stop(void)
+{
+	mutex_lock(&ap_poll_thread_mutex);
+	if (ap_poll_kthread) {
+		kthread_stop(ap_poll_kthread);
+		ap_poll_kthread = NULL;
+	}
+	mutex_unlock(&ap_poll_thread_mutex);
+}
+
+/**
+ * The module initialization code.
+ */
+int __init ap_module_init(void)
+{
+	int rc, i;
+
+	if (ap_domain_index < -1 || ap_domain_index >= AP_DOMAINS) {
+		printk(KERN_WARNING "Invalid param: domain = %d. "
+		       " Not loading.\n", ap_domain_index);
+		return -EINVAL;
+	}
+	if (ap_instructions_available() != 0) {
+		printk(KERN_WARNING "AP instructions not installed.\n");
+		return -ENODEV;
+	}
+
+	/* Create /sys/bus/ap. */
+	rc = bus_register(&ap_bus_type);
+	if (rc)
+		goto out;
+	for (i = 0; ap_bus_attrs[i]; i++) {
+		rc = bus_create_file(&ap_bus_type, ap_bus_attrs[i]);
+		if (rc)
+			goto out_bus;
+	}
+
+	/* Create /sys/devices/ap. */
+	ap_root_device = s390_root_dev_register("ap");
+	rc = IS_ERR(ap_root_device) ? PTR_ERR(ap_root_device) : 0;
+	if (rc)
+		goto out_bus;
+
+	ap_work_queue = create_singlethread_workqueue("kapwork");
+	if (!ap_work_queue) {
+		rc = -ENOMEM;
+		goto out_root;
+	}
+
+	if (ap_select_domain() == 0)
+		ap_scan_bus(NULL);
+
+	/* Setup the ap bus rescan timer. */
+	init_timer(&ap_config_timer);
+	ap_config_timer.function = ap_config_timeout;
+	ap_config_timer.data = 0;
+	ap_config_timer.expires = jiffies + ap_config_time * HZ;
+	add_timer(&ap_config_timer);
+
+	/* Start the low priority AP bus poll thread. */
+	if (ap_thread_flag) {
+		rc = ap_poll_thread_start();
+		if (rc)
+			goto out_work;
+	}
+
+	return 0;
+
+out_work:
+	del_timer_sync(&ap_config_timer);
+	del_timer_sync(&ap_poll_timer);
+	destroy_workqueue(ap_work_queue);
+out_root:
+	s390_root_dev_unregister(ap_root_device);
+out_bus:
+	while (i--)
+		bus_remove_file(&ap_bus_type, ap_bus_attrs[i]);
+	bus_unregister(&ap_bus_type);
+out:
+	return rc;
+}
+
+static int __ap_match_all(struct device *dev, void *data)
+{
+	return 1;
+}
+
+/**
+ * The module termination code
+ */
+void ap_module_exit(void)
+{
+	int i;
+	struct device *dev;
+
+	ap_poll_thread_stop();
+	del_timer_sync(&ap_config_timer);
+	del_timer_sync(&ap_poll_timer);
+	destroy_workqueue(ap_work_queue);
+	s390_root_dev_unregister(ap_root_device);
+	while ((dev = bus_find_device(&ap_bus_type, NULL, NULL,
+		    __ap_match_all)))
+	{
+		device_unregister(dev);
+		put_device(dev);
+	}
+	for (i = 0; ap_bus_attrs[i]; i++)
+		bus_remove_file(&ap_bus_type, ap_bus_attrs[i]);
+	bus_unregister(&ap_bus_type);
+}
+
+#ifndef CONFIG_ZCRYPT_MONOLITHIC
+module_init(ap_module_init);
+module_exit(ap_module_exit);
+#endif
diff --git a/drivers/s390/crypto/ap_bus.h b/drivers/s390/crypto/ap_bus.h
new file mode 100644
index 000000000000..83b69c01cd6e
--- /dev/null
+++ b/drivers/s390/crypto/ap_bus.h
@@ -0,0 +1,158 @@
+/*
+ * linux/drivers/s390/crypto/ap_bus.h
+ *
+ * Copyright (C) 2006 IBM Corporation
+ * Author(s): Cornelia Huck <cornelia.huck@de.ibm.com>
+ *	      Martin Schwidefsky <schwidefsky@de.ibm.com>
+ *	      Ralph Wuerthner <rwuerthn@de.ibm.com>
+ *
+ * Adjunct processor bus header file.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2, or (at your option)
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#ifndef _AP_BUS_H_
+#define _AP_BUS_H_
+
+#include <linux/device.h>
+#include <linux/mod_devicetable.h>
+#include <linux/types.h>
+
+#define AP_DEVICES 64		/* Number of AP devices. */
+#define AP_DOMAINS 16		/* Number of AP domains. */
+#define AP_MAX_RESET 90		/* Maximum number of resets. */
+#define AP_CONFIG_TIME 30	/* Time in seconds between AP bus rescans. */
+#define AP_POLL_TIME 1		/* Time in ticks between receive polls. */
+
+extern int ap_domain_index;
+
+/**
+ * The ap_qid_t identifier of an ap queue. It contains a
+ * 6 bit device index and a 4 bit queue index (domain).
+ */
+typedef unsigned int ap_qid_t;
+
+#define AP_MKQID(_device,_queue) (((_device) & 63) << 8 | ((_queue) & 15))
+#define AP_QID_DEVICE(_qid) (((_qid) >> 8) & 63)
+#define AP_QID_QUEUE(_qid) ((_qid) & 15)
+
+/**
+ * The ap queue status word is returned by all three AP functions
+ * (PQAP, NQAP and DQAP).  There's a set of flags in the first
+ * byte, followed by a 1 byte response code.
+ */
+struct ap_queue_status {
+	unsigned int queue_empty	: 1;
+	unsigned int replies_waiting	: 1;
+	unsigned int queue_full		: 1;
+	unsigned int pad1		: 5;
+	unsigned int response_code	: 8;
+	unsigned int pad2		: 16;
+};
+
+#define AP_RESPONSE_NORMAL		0x00
+#define AP_RESPONSE_Q_NOT_AVAIL		0x01
+#define AP_RESPONSE_RESET_IN_PROGRESS	0x02
+#define AP_RESPONSE_DECONFIGURED	0x03
+#define AP_RESPONSE_CHECKSTOPPED	0x04
+#define AP_RESPONSE_BUSY		0x05
+#define AP_RESPONSE_Q_FULL		0x10
+#define AP_RESPONSE_NO_PENDING_REPLY	0x10
+#define AP_RESPONSE_INDEX_TOO_BIG	0x11
+#define AP_RESPONSE_NO_FIRST_PART	0x13
+#define AP_RESPONSE_MESSAGE_TOO_BIG	0x15
+
+/**
+ * Known device types
+ */
+#define AP_DEVICE_TYPE_PCICC	3
+#define AP_DEVICE_TYPE_PCICA	4
+#define AP_DEVICE_TYPE_PCIXCC	5
+#define AP_DEVICE_TYPE_CEX2A	6
+#define AP_DEVICE_TYPE_CEX2C	7
+
+struct ap_device;
+struct ap_message;
+
+struct ap_driver {
+	struct device_driver driver;
+	struct ap_device_id *ids;
+
+	int (*probe)(struct ap_device *);
+	void (*remove)(struct ap_device *);
+	/* receive is called from tasklet context */
+	void (*receive)(struct ap_device *, struct ap_message *,
+			struct ap_message *);
+};
+
+#define to_ap_drv(x) container_of((x), struct ap_driver, driver)
+
+int ap_driver_register(struct ap_driver *, struct module *, char *);
+void ap_driver_unregister(struct ap_driver *);
+
+struct ap_device {
+	struct device device;
+	struct ap_driver *drv;		/* Pointer to AP device driver. */
+	spinlock_t lock;		/* Per device lock. */
+
+	ap_qid_t qid;			/* AP queue id. */
+	int queue_depth;		/* AP queue depth.*/
+	int device_type;		/* AP device type. */
+	int unregistered;		/* marks AP device as unregistered */
+
+	int queue_count;		/* # messages currently on AP queue. */
+
+	struct list_head pendingq;	/* List of message sent to AP queue. */
+	int pendingq_count;		/* # requests on pendingq list. */
+	struct list_head requestq;	/* List of message yet to be sent. */
+	int requestq_count;		/* # requests on requestq list. */
+	int total_request_count;	/* # requests ever for this AP device. */
+
+	struct ap_message *reply;	/* Per device reply message. */
+
+	void *private;			/* ap driver private pointer. */
+};
+
+#define to_ap_dev(x) container_of((x), struct ap_device, device)
+
+struct ap_message {
+	struct list_head list;		/* Request queueing. */
+	unsigned long long psmid;	/* Message id. */
+	void *message;			/* Pointer to message buffer. */
+	size_t length;			/* Message length. */
+
+	void *private;			/* ap driver private pointer. */
+};
+
+#define AP_DEVICE(dt)					\
+	.dev_type=(dt),					\
+	.match_flags=AP_DEVICE_ID_MATCH_DEVICE_TYPE,
+
+/**
+ * Note: don't use ap_send/ap_recv after using ap_queue_message
+ * for the first time. Otherwise the ap message queue will get
+ * confused.
+ */
+int ap_send(ap_qid_t, unsigned long long, void *, size_t);
+int ap_recv(ap_qid_t, unsigned long long *, void *, size_t);
+
+void ap_queue_message(struct ap_device *ap_dev, struct ap_message *ap_msg);
+void ap_cancel_message(struct ap_device *ap_dev, struct ap_message *ap_msg);
+void ap_flush_queue(struct ap_device *ap_dev);
+
+int ap_module_init(void);
+void ap_module_exit(void);
+
+#endif /* _AP_BUS_H_ */
diff --git a/include/linux/mod_devicetable.h b/include/linux/mod_devicetable.h
index f6977708585c..f7ca0b09075d 100644
--- a/include/linux/mod_devicetable.h
+++ b/include/linux/mod_devicetable.h
@@ -148,6 +148,17 @@ struct ccw_device_id {
 #define CCW_DEVICE_ID_MATCH_DEVICE_TYPE		0x04
 #define CCW_DEVICE_ID_MATCH_DEVICE_MODEL	0x08
 
+/* s390 AP bus devices */
+struct ap_device_id {
+	__u16 match_flags;	/* which fields to match against */
+	__u8 dev_type;		/* device type */
+	__u8 pad1;
+	__u32 pad2;
+	kernel_ulong_t driver_info;
+};
+
+#define AP_DEVICE_ID_MATCH_DEVICE_TYPE		0x01
+
 
 #define PNP_ID_LEN	8
 #define PNP_MAX_DEVICES	8
diff --git a/scripts/mod/file2alias.c b/scripts/mod/file2alias.c
index e2de650d3dbf..de76da80443f 100644
--- a/scripts/mod/file2alias.c
+++ b/scripts/mod/file2alias.c
@@ -265,6 +265,14 @@ static int do_ccw_entry(const char *filename,
 	return 1;
 }
 
+/* looks like: "ap:tN" */
+static int do_ap_entry(const char *filename,
+		       struct ap_device_id *id, char *alias)
+{
+	sprintf(alias, "ap:t%02X", id->dev_type);
+	return 1;
+}
+
 /* Looks like: "serio:tyNprNidNexN" */
 static int do_serio_entry(const char *filename,
 			  struct serio_device_id *id, char *alias)
@@ -503,6 +511,10 @@ void handle_moddevtable(struct module *mod, struct elf_info *info,
 		do_table(symval, sym->st_size,
 			 sizeof(struct ccw_device_id), "ccw",
 			 do_ccw_entry, mod);
+	else if (sym_is(symname, "__mod_ap_device_table"))
+		do_table(symval, sym->st_size,
+			 sizeof(struct ap_device_id), "ap",
+			 do_ap_entry, mod);
 	else if (sym_is(symname, "__mod_serio_device_table"))
 		do_table(symval, sym->st_size,
 			 sizeof(struct serio_device_id), "serio",
-- 
cgit v1.2.3


From 2dbc2418bac32a18a372ae9aec386f0fe9174389 Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Wed, 20 Sep 2006 15:58:27 +0200
Subject: [S390] zcrypt user space interface.

The user space interface of the zcrypt device driver implements the old
user space interface as defined by the old z90crypt driver. Everything
is there, the /dev/z90crypt misc character device, all the lovely ioctls
and the /proc file. Even writing to the z90crypt proc file to configure
the crypto device still works. It stands to reason to remove the proc
write function someday since a much cleaner configuration via the sysfs
is now available.

The ap bus device drivers register crypto cards to the zcrypt user
space interface. The request router of the user space interface
picks one of the registered cards based on the predicted latency
for the request and calls the driver via a callback found in the
zcrypt_ops of the device. The request router only knows which
operations the card can do and the minimum / maximum number of bits
a request can have.

Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Ralph Wuerthner <rwuerthn@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 drivers/s390/crypto/zcrypt_api.c | 981 +++++++++++++++++++++++++++++++++++++++
 drivers/s390/crypto/zcrypt_api.h | 140 ++++++
 include/asm-s390/zcrypt.h        | 207 +++++++++
 3 files changed, 1328 insertions(+)
 create mode 100644 drivers/s390/crypto/zcrypt_api.c
 create mode 100644 drivers/s390/crypto/zcrypt_api.h
 create mode 100644 include/asm-s390/zcrypt.h

(limited to 'include')

diff --git a/drivers/s390/crypto/zcrypt_api.c b/drivers/s390/crypto/zcrypt_api.c
new file mode 100644
index 000000000000..b3fe003b3d2d
--- /dev/null
+++ b/drivers/s390/crypto/zcrypt_api.c
@@ -0,0 +1,981 @@
+/*
+ *  linux/drivers/s390/crypto/zcrypt_api.c
+ *
+ *  zcrypt 2.0.0
+ *
+ *  Copyright (C)  2001, 2006 IBM Corporation
+ *  Author(s): Robert Burroughs
+ *	       Eric Rossman (edrossma@us.ibm.com)
+ *	       Cornelia Huck <cornelia.huck@de.ibm.com>
+ *
+ *  Hotplug & misc device support: Jochen Roehrig (roehrig@de.ibm.com)
+ *  Major cleanup & driver split: Martin Schwidefsky <schwidefsky@de.ibm.com>
+ *				  Ralph Wuerthner <rwuerthn@de.ibm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2, or (at your option)
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/miscdevice.h>
+#include <linux/fs.h>
+#include <linux/proc_fs.h>
+#include <linux/compat.h>
+#include <asm/atomic.h>
+#include <asm/uaccess.h>
+
+#include "zcrypt_api.h"
+
+/**
+ * Module description.
+ */
+MODULE_AUTHOR("IBM Corporation");
+MODULE_DESCRIPTION("Cryptographic Coprocessor interface, "
+		   "Copyright 2001, 2006 IBM Corporation");
+MODULE_LICENSE("GPL");
+
+static DEFINE_SPINLOCK(zcrypt_device_lock);
+static LIST_HEAD(zcrypt_device_list);
+static int zcrypt_device_count = 0;
+static atomic_t zcrypt_open_count = ATOMIC_INIT(0);
+
+/**
+ * Device attributes common for all crypto devices.
+ */
+static ssize_t zcrypt_type_show(struct device *dev,
+				struct device_attribute *attr, char *buf)
+{
+	struct zcrypt_device *zdev = to_ap_dev(dev)->private;
+	return snprintf(buf, PAGE_SIZE, "%s\n", zdev->type_string);
+}
+
+static DEVICE_ATTR(type, 0444, zcrypt_type_show, NULL);
+
+static ssize_t zcrypt_online_show(struct device *dev,
+				  struct device_attribute *attr, char *buf)
+{
+	struct zcrypt_device *zdev = to_ap_dev(dev)->private;
+	return snprintf(buf, PAGE_SIZE, "%d\n", zdev->online);
+}
+
+static ssize_t zcrypt_online_store(struct device *dev,
+				   struct device_attribute *attr,
+				   const char *buf, size_t count)
+{
+	struct zcrypt_device *zdev = to_ap_dev(dev)->private;
+	int online;
+
+	if (sscanf(buf, "%d\n", &online) != 1 || online < 0 || online > 1)
+		return -EINVAL;
+	zdev->online = online;
+	if (!online)
+		ap_flush_queue(zdev->ap_dev);
+	return count;
+}
+
+static DEVICE_ATTR(online, 0644, zcrypt_online_show, zcrypt_online_store);
+
+static struct attribute * zcrypt_device_attrs[] = {
+	&dev_attr_type.attr,
+	&dev_attr_online.attr,
+	NULL,
+};
+
+static struct attribute_group zcrypt_device_attr_group = {
+	.attrs = zcrypt_device_attrs,
+};
+
+/**
+ * Move the device towards the head of the device list.
+ * Need to be called while holding the zcrypt device list lock.
+ * Note: cards with speed_rating of 0 are kept at the end of the list.
+ */
+static void __zcrypt_increase_preference(struct zcrypt_device *zdev)
+{
+	struct zcrypt_device *tmp;
+	struct list_head *l;
+
+	if (zdev->speed_rating == 0)
+		return;
+	for (l = zdev->list.prev; l != &zcrypt_device_list; l = l->prev) {
+		tmp = list_entry(l, struct zcrypt_device, list);
+		if ((tmp->request_count + 1) * tmp->speed_rating <=
+		    (zdev->request_count + 1) * zdev->speed_rating &&
+		    tmp->speed_rating != 0)
+			break;
+	}
+	if (l == zdev->list.prev)
+		return;
+	/* Move zdev behind l */
+	list_del(&zdev->list);
+	list_add(&zdev->list, l);
+}
+
+/**
+ * Move the device towards the tail of the device list.
+ * Need to be called while holding the zcrypt device list lock.
+ * Note: cards with speed_rating of 0 are kept at the end of the list.
+ */
+static void __zcrypt_decrease_preference(struct zcrypt_device *zdev)
+{
+	struct zcrypt_device *tmp;
+	struct list_head *l;
+
+	if (zdev->speed_rating == 0)
+		return;
+	for (l = zdev->list.next; l != &zcrypt_device_list; l = l->next) {
+		tmp = list_entry(l, struct zcrypt_device, list);
+		if ((tmp->request_count + 1) * tmp->speed_rating >
+		    (zdev->request_count + 1) * zdev->speed_rating ||
+		    tmp->speed_rating == 0)
+			break;
+	}
+	if (l == zdev->list.next)
+		return;
+	/* Move zdev before l */
+	list_del(&zdev->list);
+	list_add_tail(&zdev->list, l);
+}
+
+static void zcrypt_device_release(struct kref *kref)
+{
+	struct zcrypt_device *zdev =
+		container_of(kref, struct zcrypt_device, refcount);
+	zcrypt_device_free(zdev);
+}
+
+void zcrypt_device_get(struct zcrypt_device *zdev)
+{
+	kref_get(&zdev->refcount);
+}
+EXPORT_SYMBOL(zcrypt_device_get);
+
+int zcrypt_device_put(struct zcrypt_device *zdev)
+{
+	return kref_put(&zdev->refcount, zcrypt_device_release);
+}
+EXPORT_SYMBOL(zcrypt_device_put);
+
+struct zcrypt_device *zcrypt_device_alloc(size_t max_response_size)
+{
+	struct zcrypt_device *zdev;
+
+	zdev = kzalloc(sizeof(struct zcrypt_device), GFP_KERNEL);
+	if (!zdev)
+		return NULL;
+	zdev->reply.message = kmalloc(max_response_size, GFP_KERNEL);
+	if (!zdev->reply.message)
+		goto out_free;
+	zdev->reply.length = max_response_size;
+	spin_lock_init(&zdev->lock);
+	INIT_LIST_HEAD(&zdev->list);
+	return zdev;
+
+out_free:
+	kfree(zdev);
+	return NULL;
+}
+EXPORT_SYMBOL(zcrypt_device_alloc);
+
+void zcrypt_device_free(struct zcrypt_device *zdev)
+{
+	kfree(zdev->reply.message);
+	kfree(zdev);
+}
+EXPORT_SYMBOL(zcrypt_device_free);
+
+/**
+ * Register a crypto device.
+ */
+int zcrypt_device_register(struct zcrypt_device *zdev)
+{
+	int rc;
+
+	rc = sysfs_create_group(&zdev->ap_dev->device.kobj,
+				&zcrypt_device_attr_group);
+	if (rc)
+		goto out;
+	get_device(&zdev->ap_dev->device);
+	kref_init(&zdev->refcount);
+	spin_lock_bh(&zcrypt_device_lock);
+	zdev->online = 1;	/* New devices are online by default. */
+	list_add_tail(&zdev->list, &zcrypt_device_list);
+	__zcrypt_increase_preference(zdev);
+	zcrypt_device_count++;
+	spin_unlock_bh(&zcrypt_device_lock);
+out:
+	return rc;
+}
+EXPORT_SYMBOL(zcrypt_device_register);
+
+/**
+ * Unregister a crypto device.
+ */
+void zcrypt_device_unregister(struct zcrypt_device *zdev)
+{
+	spin_lock_bh(&zcrypt_device_lock);
+	zcrypt_device_count--;
+	list_del_init(&zdev->list);
+	spin_unlock_bh(&zcrypt_device_lock);
+	sysfs_remove_group(&zdev->ap_dev->device.kobj,
+			   &zcrypt_device_attr_group);
+	put_device(&zdev->ap_dev->device);
+	zcrypt_device_put(zdev);
+}
+EXPORT_SYMBOL(zcrypt_device_unregister);
+
+/**
+ * zcrypt_read is not be supported beyond zcrypt 1.3.1
+ */
+static ssize_t zcrypt_read(struct file *filp, char __user *buf,
+			   size_t count, loff_t *f_pos)
+{
+	return -EPERM;
+}
+
+/**
+ * Write is is not allowed
+ */
+static ssize_t zcrypt_write(struct file *filp, const char __user *buf,
+			    size_t count, loff_t *f_pos)
+{
+	return -EPERM;
+}
+
+/**
+ * Device open/close functions to count number of users.
+ */
+static int zcrypt_open(struct inode *inode, struct file *filp)
+{
+	atomic_inc(&zcrypt_open_count);
+	return 0;
+}
+
+static int zcrypt_release(struct inode *inode, struct file *filp)
+{
+	atomic_dec(&zcrypt_open_count);
+	return 0;
+}
+
+/**
+ * zcrypt ioctls.
+ */
+static long zcrypt_rsa_modexpo(struct ica_rsa_modexpo *mex)
+{
+	struct zcrypt_device *zdev;
+	int rc;
+
+	if (mex->outputdatalength < mex->inputdatalength)
+		return -EINVAL;
+	/**
+	 * As long as outputdatalength is big enough, we can set the
+	 * outputdatalength equal to the inputdatalength, since that is the
+	 * number of bytes we will copy in any case
+	 */
+	mex->outputdatalength = mex->inputdatalength;
+
+	spin_lock_bh(&zcrypt_device_lock);
+	list_for_each_entry(zdev, &zcrypt_device_list, list) {
+		if (!zdev->online ||
+		    !zdev->ops->rsa_modexpo ||
+		    zdev->min_mod_size > mex->inputdatalength ||
+		    zdev->max_mod_size < mex->inputdatalength)
+			continue;
+		zcrypt_device_get(zdev);
+		get_device(&zdev->ap_dev->device);
+		zdev->request_count++;
+		__zcrypt_decrease_preference(zdev);
+		spin_unlock_bh(&zcrypt_device_lock);
+		if (try_module_get(zdev->ap_dev->drv->driver.owner)) {
+			rc = zdev->ops->rsa_modexpo(zdev, mex);
+			module_put(zdev->ap_dev->drv->driver.owner);
+		}
+		else
+			rc = -EAGAIN;
+		spin_lock_bh(&zcrypt_device_lock);
+		zdev->request_count--;
+		__zcrypt_increase_preference(zdev);
+		put_device(&zdev->ap_dev->device);
+		zcrypt_device_put(zdev);
+		spin_unlock_bh(&zcrypt_device_lock);
+		return rc;
+	}
+	spin_unlock_bh(&zcrypt_device_lock);
+	return -ENODEV;
+}
+
+static long zcrypt_rsa_crt(struct ica_rsa_modexpo_crt *crt)
+{
+	struct zcrypt_device *zdev;
+	unsigned long long z1, z2, z3;
+	int rc, copied;
+
+	if (crt->outputdatalength < crt->inputdatalength ||
+	    (crt->inputdatalength & 1))
+		return -EINVAL;
+	/**
+	 * As long as outputdatalength is big enough, we can set the
+	 * outputdatalength equal to the inputdatalength, since that is the
+	 * number of bytes we will copy in any case
+	 */
+	crt->outputdatalength = crt->inputdatalength;
+
+	copied = 0;
+ restart:
+	spin_lock_bh(&zcrypt_device_lock);
+	list_for_each_entry(zdev, &zcrypt_device_list, list) {
+		if (!zdev->online ||
+		    !zdev->ops->rsa_modexpo_crt ||
+		    zdev->min_mod_size > crt->inputdatalength ||
+		    zdev->max_mod_size < crt->inputdatalength)
+			continue;
+		if (zdev->short_crt && crt->inputdatalength > 240) {
+			/**
+			 * Check inputdata for leading zeros for cards
+			 * that can't handle np_prime, bp_key, or
+			 * u_mult_inv > 128 bytes.
+			 */
+			if (copied == 0) {
+				int len;
+				spin_unlock_bh(&zcrypt_device_lock);
+				/* len is max 256 / 2 - 120 = 8 */
+				len = crt->inputdatalength / 2 - 120;
+				z1 = z2 = z3 = 0;
+				if (copy_from_user(&z1, crt->np_prime, len) ||
+				    copy_from_user(&z2, crt->bp_key, len) ||
+				    copy_from_user(&z3, crt->u_mult_inv, len))
+					return -EFAULT;
+				copied = 1;
+				/**
+				 * We have to restart device lookup -
+				 * the device list may have changed by now.
+				 */
+				goto restart;
+			}
+			if (z1 != 0ULL || z2 != 0ULL || z3 != 0ULL)
+				/* The device can't handle this request. */
+				continue;
+		}
+		zcrypt_device_get(zdev);
+		get_device(&zdev->ap_dev->device);
+		zdev->request_count++;
+		__zcrypt_decrease_preference(zdev);
+		spin_unlock_bh(&zcrypt_device_lock);
+		if (try_module_get(zdev->ap_dev->drv->driver.owner)) {
+			rc = zdev->ops->rsa_modexpo_crt(zdev, crt);
+			module_put(zdev->ap_dev->drv->driver.owner);
+		}
+		else
+			rc = -EAGAIN;
+		spin_lock_bh(&zcrypt_device_lock);
+		zdev->request_count--;
+		__zcrypt_increase_preference(zdev);
+		put_device(&zdev->ap_dev->device);
+		zcrypt_device_put(zdev);
+		spin_unlock_bh(&zcrypt_device_lock);
+		return rc;
+	}
+	spin_unlock_bh(&zcrypt_device_lock);
+	return -ENODEV;
+}
+
+static void zcrypt_status_mask(char status[AP_DEVICES])
+{
+	struct zcrypt_device *zdev;
+
+	memset(status, 0, sizeof(char) * AP_DEVICES);
+	spin_lock_bh(&zcrypt_device_lock);
+	list_for_each_entry(zdev, &zcrypt_device_list, list)
+		status[AP_QID_DEVICE(zdev->ap_dev->qid)] =
+			zdev->online ? zdev->user_space_type : 0x0d;
+	spin_unlock_bh(&zcrypt_device_lock);
+}
+
+static void zcrypt_qdepth_mask(char qdepth[AP_DEVICES])
+{
+	struct zcrypt_device *zdev;
+
+	memset(qdepth, 0, sizeof(char)	* AP_DEVICES);
+	spin_lock_bh(&zcrypt_device_lock);
+	list_for_each_entry(zdev, &zcrypt_device_list, list) {
+		spin_lock(&zdev->ap_dev->lock);
+		qdepth[AP_QID_DEVICE(zdev->ap_dev->qid)] =
+			zdev->ap_dev->pendingq_count +
+			zdev->ap_dev->requestq_count;
+		spin_unlock(&zdev->ap_dev->lock);
+	}
+	spin_unlock_bh(&zcrypt_device_lock);
+}
+
+static void zcrypt_perdev_reqcnt(int reqcnt[AP_DEVICES])
+{
+	struct zcrypt_device *zdev;
+
+	memset(reqcnt, 0, sizeof(int) * AP_DEVICES);
+	spin_lock_bh(&zcrypt_device_lock);
+	list_for_each_entry(zdev, &zcrypt_device_list, list) {
+		spin_lock(&zdev->ap_dev->lock);
+		reqcnt[AP_QID_DEVICE(zdev->ap_dev->qid)] =
+			zdev->ap_dev->total_request_count;
+		spin_unlock(&zdev->ap_dev->lock);
+	}
+	spin_unlock_bh(&zcrypt_device_lock);
+}
+
+static int zcrypt_pendingq_count(void)
+{
+	struct zcrypt_device *zdev;
+	int pendingq_count = 0;
+
+	spin_lock_bh(&zcrypt_device_lock);
+	list_for_each_entry(zdev, &zcrypt_device_list, list) {
+		spin_lock(&zdev->ap_dev->lock);
+		pendingq_count += zdev->ap_dev->pendingq_count;
+		spin_unlock(&zdev->ap_dev->lock);
+	}
+	spin_unlock_bh(&zcrypt_device_lock);
+	return pendingq_count;
+}
+
+static int zcrypt_requestq_count(void)
+{
+	struct zcrypt_device *zdev;
+	int requestq_count = 0;
+
+	spin_lock_bh(&zcrypt_device_lock);
+	list_for_each_entry(zdev, &zcrypt_device_list, list) {
+		spin_lock(&zdev->ap_dev->lock);
+		requestq_count += zdev->ap_dev->requestq_count;
+		spin_unlock(&zdev->ap_dev->lock);
+	}
+	spin_unlock_bh(&zcrypt_device_lock);
+	return requestq_count;
+}
+
+static int zcrypt_count_type(int type)
+{
+	struct zcrypt_device *zdev;
+	int device_count = 0;
+
+	spin_lock_bh(&zcrypt_device_lock);
+	list_for_each_entry(zdev, &zcrypt_device_list, list)
+		if (zdev->user_space_type == type)
+			device_count++;
+	spin_unlock_bh(&zcrypt_device_lock);
+	return device_count;
+}
+
+/**
+ * Old, deprecated combi status call.
+ */
+static long zcrypt_ica_status(struct file *filp, unsigned long arg)
+{
+	struct ica_z90_status *pstat;
+	int ret;
+
+	pstat = kzalloc(sizeof(*pstat), GFP_KERNEL);
+	if (!pstat)
+		return -ENOMEM;
+	pstat->totalcount = zcrypt_device_count;
+	pstat->leedslitecount = zcrypt_count_type(ZCRYPT_PCICA);
+	pstat->leeds2count = zcrypt_count_type(ZCRYPT_PCICC);
+	pstat->requestqWaitCount = zcrypt_requestq_count();
+	pstat->pendingqWaitCount = zcrypt_pendingq_count();
+	pstat->totalOpenCount = atomic_read(&zcrypt_open_count);
+	pstat->cryptoDomain = ap_domain_index;
+	zcrypt_status_mask(pstat->status);
+	zcrypt_qdepth_mask(pstat->qdepth);
+	ret = 0;
+	if (copy_to_user((void __user *) arg, pstat, sizeof(*pstat)))
+		ret = -EFAULT;
+	kfree(pstat);
+	return ret;
+}
+
+static long zcrypt_unlocked_ioctl(struct file *filp, unsigned int cmd,
+				  unsigned long arg)
+{
+	int rc;
+
+	switch (cmd) {
+	case ICARSAMODEXPO: {
+		struct ica_rsa_modexpo __user *umex = (void __user *) arg;
+		struct ica_rsa_modexpo mex;
+		if (copy_from_user(&mex, umex, sizeof(mex)))
+			return -EFAULT;
+		do {
+			rc = zcrypt_rsa_modexpo(&mex);
+		} while (rc == -EAGAIN);
+		if (rc)
+			return rc;
+		return put_user(mex.outputdatalength, &umex->outputdatalength);
+	}
+	case ICARSACRT: {
+		struct ica_rsa_modexpo_crt __user *ucrt = (void __user *) arg;
+		struct ica_rsa_modexpo_crt crt;
+		if (copy_from_user(&crt, ucrt, sizeof(crt)))
+			return -EFAULT;
+		do {
+			rc = zcrypt_rsa_crt(&crt);
+		} while (rc == -EAGAIN);
+		if (rc)
+			return rc;
+		return put_user(crt.outputdatalength, &ucrt->outputdatalength);
+	}
+	case Z90STAT_STATUS_MASK: {
+		char status[AP_DEVICES];
+		zcrypt_status_mask(status);
+		if (copy_to_user((char __user *) arg, status,
+				 sizeof(char) * AP_DEVICES))
+			return -EFAULT;
+		return 0;
+	}
+	case Z90STAT_QDEPTH_MASK: {
+		char qdepth[AP_DEVICES];
+		zcrypt_qdepth_mask(qdepth);
+		if (copy_to_user((char __user *) arg, qdepth,
+				 sizeof(char) * AP_DEVICES))
+			return -EFAULT;
+		return 0;
+	}
+	case Z90STAT_PERDEV_REQCNT: {
+		int reqcnt[AP_DEVICES];
+		zcrypt_perdev_reqcnt(reqcnt);
+		if (copy_to_user((int __user *) arg, reqcnt,
+				 sizeof(int) * AP_DEVICES))
+			return -EFAULT;
+		return 0;
+	}
+	case Z90STAT_REQUESTQ_COUNT:
+		return put_user(zcrypt_requestq_count(), (int __user *) arg);
+	case Z90STAT_PENDINGQ_COUNT:
+		return put_user(zcrypt_pendingq_count(), (int __user *) arg);
+	case Z90STAT_TOTALOPEN_COUNT:
+		return put_user(atomic_read(&zcrypt_open_count),
+				(int __user *) arg);
+	case Z90STAT_DOMAIN_INDEX:
+		return put_user(ap_domain_index, (int __user *) arg);
+	/**
+	 * Deprecated ioctls. Don't add another device count ioctl,
+	 * you can count them yourself in the user space with the
+	 * output of the Z90STAT_STATUS_MASK ioctl.
+	 */
+	case ICAZ90STATUS:
+		return zcrypt_ica_status(filp, arg);
+	case Z90STAT_TOTALCOUNT:
+		return put_user(zcrypt_device_count, (int __user *) arg);
+	case Z90STAT_PCICACOUNT:
+		return put_user(zcrypt_count_type(ZCRYPT_PCICA),
+				(int __user *) arg);
+	case Z90STAT_PCICCCOUNT:
+		return put_user(zcrypt_count_type(ZCRYPT_PCICC),
+				(int __user *) arg);
+	case Z90STAT_PCIXCCMCL2COUNT:
+		return put_user(zcrypt_count_type(ZCRYPT_PCIXCC_MCL2),
+				(int __user *) arg);
+	case Z90STAT_PCIXCCMCL3COUNT:
+		return put_user(zcrypt_count_type(ZCRYPT_PCIXCC_MCL3),
+				(int __user *) arg);
+	case Z90STAT_PCIXCCCOUNT:
+		return put_user(zcrypt_count_type(ZCRYPT_PCIXCC_MCL2) +
+				zcrypt_count_type(ZCRYPT_PCIXCC_MCL3),
+				(int __user *) arg);
+	case Z90STAT_CEX2CCOUNT:
+		return put_user(zcrypt_count_type(ZCRYPT_CEX2C),
+				(int __user *) arg);
+	case Z90STAT_CEX2ACOUNT:
+		return put_user(zcrypt_count_type(ZCRYPT_CEX2A),
+				(int __user *) arg);
+	default:
+		/* unknown ioctl number */
+		return -ENOIOCTLCMD;
+	}
+}
+
+#ifdef CONFIG_COMPAT
+/**
+ * ioctl32 conversion routines
+ */
+struct compat_ica_rsa_modexpo {
+	compat_uptr_t	inputdata;
+	unsigned int	inputdatalength;
+	compat_uptr_t	outputdata;
+	unsigned int	outputdatalength;
+	compat_uptr_t	b_key;
+	compat_uptr_t	n_modulus;
+};
+
+static long trans_modexpo32(struct file *filp, unsigned int cmd,
+			    unsigned long arg)
+{
+	struct compat_ica_rsa_modexpo __user *umex32 = compat_ptr(arg);
+	struct compat_ica_rsa_modexpo mex32;
+	struct ica_rsa_modexpo mex64;
+	long rc;
+
+	if (copy_from_user(&mex32, umex32, sizeof(mex32)))
+		return -EFAULT;
+	mex64.inputdata = compat_ptr(mex32.inputdata);
+	mex64.inputdatalength = mex32.inputdatalength;
+	mex64.outputdata = compat_ptr(mex32.outputdata);
+	mex64.outputdatalength = mex32.outputdatalength;
+	mex64.b_key = compat_ptr(mex32.b_key);
+	mex64.n_modulus = compat_ptr(mex32.n_modulus);
+	do {
+		rc = zcrypt_rsa_modexpo(&mex64);
+	} while (rc == -EAGAIN);
+	if (!rc)
+		rc = put_user(mex64.outputdatalength,
+			      &umex32->outputdatalength);
+	return rc;
+}
+
+struct compat_ica_rsa_modexpo_crt {
+	compat_uptr_t	inputdata;
+	unsigned int	inputdatalength;
+	compat_uptr_t	outputdata;
+	unsigned int	outputdatalength;
+	compat_uptr_t	bp_key;
+	compat_uptr_t	bq_key;
+	compat_uptr_t	np_prime;
+	compat_uptr_t	nq_prime;
+	compat_uptr_t	u_mult_inv;
+};
+
+static long trans_modexpo_crt32(struct file *filp, unsigned int cmd,
+				unsigned long arg)
+{
+	struct compat_ica_rsa_modexpo_crt __user *ucrt32 = compat_ptr(arg);
+	struct compat_ica_rsa_modexpo_crt crt32;
+	struct ica_rsa_modexpo_crt crt64;
+	long rc;
+
+	if (copy_from_user(&crt32, ucrt32, sizeof(crt32)))
+		return -EFAULT;
+	crt64.inputdata = compat_ptr(crt32.inputdata);
+	crt64.inputdatalength = crt32.inputdatalength;
+	crt64.outputdata=  compat_ptr(crt32.outputdata);
+	crt64.outputdatalength = crt32.outputdatalength;
+	crt64.bp_key = compat_ptr(crt32.bp_key);
+	crt64.bq_key = compat_ptr(crt32.bq_key);
+	crt64.np_prime = compat_ptr(crt32.np_prime);
+	crt64.nq_prime = compat_ptr(crt32.nq_prime);
+	crt64.u_mult_inv = compat_ptr(crt32.u_mult_inv);
+	do {
+		rc = zcrypt_rsa_crt(&crt64);
+	} while (rc == -EAGAIN);
+	if (!rc)
+		rc = put_user(crt64.outputdatalength,
+			      &ucrt32->outputdatalength);
+	return rc;
+}
+
+long zcrypt_compat_ioctl(struct file *filp, unsigned int cmd,
+			 unsigned long arg)
+{
+	if (cmd == ICARSAMODEXPO)
+		return trans_modexpo32(filp, cmd, arg);
+	if (cmd == ICARSACRT)
+		return trans_modexpo_crt32(filp, cmd, arg);
+	return zcrypt_unlocked_ioctl(filp, cmd, arg);
+}
+#endif
+
+/**
+ * Misc device file operations.
+ */
+static struct file_operations zcrypt_fops = {
+	.owner		= THIS_MODULE,
+	.read		= zcrypt_read,
+	.write		= zcrypt_write,
+	.unlocked_ioctl	= zcrypt_unlocked_ioctl,
+#ifdef CONFIG_COMPAT
+	.compat_ioctl	= zcrypt_compat_ioctl,
+#endif
+	.open		= zcrypt_open,
+	.release	= zcrypt_release
+};
+
+/**
+ * Misc device.
+ */
+static struct miscdevice zcrypt_misc_device = {
+	.minor	    = MISC_DYNAMIC_MINOR,
+	.name	    = "z90crypt",
+	.fops	    = &zcrypt_fops,
+};
+
+/**
+ * Deprecated /proc entry support.
+ */
+static struct proc_dir_entry *zcrypt_entry;
+
+static inline int sprintcl(unsigned char *outaddr, unsigned char *addr,
+			   unsigned int len)
+{
+	int hl, i;
+
+	hl = 0;
+	for (i = 0; i < len; i++)
+		hl += sprintf(outaddr+hl, "%01x", (unsigned int) addr[i]);
+	hl += sprintf(outaddr+hl, " ");
+	return hl;
+}
+
+static inline int sprintrw(unsigned char *outaddr, unsigned char *addr,
+			   unsigned int len)
+{
+	int hl, inl, c, cx;
+
+	hl = sprintf(outaddr, "	   ");
+	inl = 0;
+	for (c = 0; c < (len / 16); c++) {
+		hl += sprintcl(outaddr+hl, addr+inl, 16);
+		inl += 16;
+	}
+	cx = len%16;
+	if (cx) {
+		hl += sprintcl(outaddr+hl, addr+inl, cx);
+		inl += cx;
+	}
+	hl += sprintf(outaddr+hl, "\n");
+	return hl;
+}
+
+static inline int sprinthx(unsigned char *title, unsigned char *outaddr,
+			   unsigned char *addr, unsigned int len)
+{
+	int hl, inl, r, rx;
+
+	hl = sprintf(outaddr, "\n%s\n", title);
+	inl = 0;
+	for (r = 0; r < (len / 64); r++) {
+		hl += sprintrw(outaddr+hl, addr+inl, 64);
+		inl += 64;
+	}
+	rx = len % 64;
+	if (rx) {
+		hl += sprintrw(outaddr+hl, addr+inl, rx);
+		inl += rx;
+	}
+	hl += sprintf(outaddr+hl, "\n");
+	return hl;
+}
+
+static inline int sprinthx4(unsigned char *title, unsigned char *outaddr,
+			    unsigned int *array, unsigned int len)
+{
+	int hl, r;
+
+	hl = sprintf(outaddr, "\n%s\n", title);
+	for (r = 0; r < len; r++) {
+		if ((r % 8) == 0)
+			hl += sprintf(outaddr+hl, "    ");
+		hl += sprintf(outaddr+hl, "%08X ", array[r]);
+		if ((r % 8) == 7)
+			hl += sprintf(outaddr+hl, "\n");
+	}
+	hl += sprintf(outaddr+hl, "\n");
+	return hl;
+}
+
+static int zcrypt_status_read(char *resp_buff, char **start, off_t offset,
+			      int count, int *eof, void *data)
+{
+	unsigned char *workarea;
+	int len;
+
+	len = 0;
+
+	/* resp_buff is a page. Use the right half for a work area */
+	workarea = resp_buff + 2000;
+	len += sprintf(resp_buff + len, "\nzcrypt version: %d.%d.%d\n",
+		ZCRYPT_VERSION, ZCRYPT_RELEASE, ZCRYPT_VARIANT);
+	len += sprintf(resp_buff + len, "Cryptographic domain: %d\n",
+		       ap_domain_index);
+	len += sprintf(resp_buff + len, "Total device count: %d\n",
+		       zcrypt_device_count);
+	len += sprintf(resp_buff + len, "PCICA count: %d\n",
+		       zcrypt_count_type(ZCRYPT_PCICA));
+	len += sprintf(resp_buff + len, "PCICC count: %d\n",
+		       zcrypt_count_type(ZCRYPT_PCICC));
+	len += sprintf(resp_buff + len, "PCIXCC MCL2 count: %d\n",
+		       zcrypt_count_type(ZCRYPT_PCIXCC_MCL2));
+	len += sprintf(resp_buff + len, "PCIXCC MCL3 count: %d\n",
+		       zcrypt_count_type(ZCRYPT_PCIXCC_MCL3));
+	len += sprintf(resp_buff + len, "CEX2C count: %d\n",
+		       zcrypt_count_type(ZCRYPT_CEX2C));
+	len += sprintf(resp_buff + len, "CEX2A count: %d\n",
+		       zcrypt_count_type(ZCRYPT_CEX2A));
+	len += sprintf(resp_buff + len, "requestq count: %d\n",
+		       zcrypt_requestq_count());
+	len += sprintf(resp_buff + len, "pendingq count: %d\n",
+		       zcrypt_pendingq_count());
+	len += sprintf(resp_buff + len, "Total open handles: %d\n\n",
+		       atomic_read(&zcrypt_open_count));
+	zcrypt_status_mask(workarea);
+	len += sprinthx("Online devices: 1=PCICA 2=PCICC 3=PCIXCC(MCL2) "
+			"4=PCIXCC(MCL3) 5=CEX2C 6=CEX2A",
+			resp_buff+len, workarea, AP_DEVICES);
+	zcrypt_qdepth_mask(workarea);
+	len += sprinthx("Waiting work element counts",
+			resp_buff+len, workarea, AP_DEVICES);
+	zcrypt_perdev_reqcnt((unsigned int *) workarea);
+	len += sprinthx4("Per-device successfully completed request counts",
+			 resp_buff+len,(unsigned int *) workarea, AP_DEVICES);
+	*eof = 1;
+	memset((void *) workarea, 0x00, AP_DEVICES * sizeof(unsigned int));
+	return len;
+}
+
+static void zcrypt_disable_card(int index)
+{
+	struct zcrypt_device *zdev;
+
+	spin_lock_bh(&zcrypt_device_lock);
+	list_for_each_entry(zdev, &zcrypt_device_list, list)
+		if (AP_QID_DEVICE(zdev->ap_dev->qid) == index) {
+			zdev->online = 0;
+			ap_flush_queue(zdev->ap_dev);
+			break;
+		}
+	spin_unlock_bh(&zcrypt_device_lock);
+}
+
+static void zcrypt_enable_card(int index)
+{
+	struct zcrypt_device *zdev;
+
+	spin_lock_bh(&zcrypt_device_lock);
+	list_for_each_entry(zdev, &zcrypt_device_list, list)
+		if (AP_QID_DEVICE(zdev->ap_dev->qid) == index) {
+			zdev->online = 1;
+			break;
+		}
+	spin_unlock_bh(&zcrypt_device_lock);
+}
+
+static int zcrypt_status_write(struct file *file, const char __user *buffer,
+			       unsigned long count, void *data)
+{
+	unsigned char *lbuf, *ptr;
+	unsigned long local_count;
+	int j;
+
+	if (count <= 0)
+		return 0;
+
+#define LBUFSIZE 1200UL
+	lbuf = kmalloc(LBUFSIZE, GFP_KERNEL);
+	if (!lbuf) {
+		PRINTK("kmalloc failed!\n");
+		return 0;
+	}
+
+	local_count = min(LBUFSIZE - 1, count);
+	if (copy_from_user(lbuf, buffer, local_count) != 0) {
+		kfree(lbuf);
+		return -EFAULT;
+	}
+	lbuf[local_count] = '\0';
+
+	ptr = strstr(lbuf, "Online devices");
+	if (!ptr) {
+		PRINTK("Unable to parse data (missing \"Online devices\")\n");
+		goto out;
+	}
+	ptr = strstr(ptr, "\n");
+	if (!ptr) {
+		PRINTK("Unable to parse data (missing newline "
+		       "after \"Online devices\")\n");
+		goto out;
+	}
+	ptr++;
+
+	if (strstr(ptr, "Waiting work element counts") == NULL) {
+		PRINTK("Unable to parse data (missing "
+		       "\"Waiting work element counts\")\n");
+		goto out;
+	}
+
+	for (j = 0; j < 64 && *ptr; ptr++) {
+		/**
+		 * '0' for no device, '1' for PCICA, '2' for PCICC,
+		 * '3' for PCIXCC_MCL2, '4' for PCIXCC_MCL3,
+		 * '5' for CEX2C and '6' for CEX2A'
+		 */
+		if (*ptr >= '0' && *ptr <= '6')
+			j++;
+		else if (*ptr == 'd' || *ptr == 'D')
+			zcrypt_disable_card(j++);
+		else if (*ptr == 'e' || *ptr == 'E')
+			zcrypt_enable_card(j++);
+		else if (*ptr != ' ' && *ptr != '\t')
+			break;
+	}
+out:
+	kfree(lbuf);
+	return count;
+}
+
+/**
+ * The module initialization code.
+ */
+int __init zcrypt_api_init(void)
+{
+	int rc;
+
+	/* Register the request sprayer. */
+	rc = misc_register(&zcrypt_misc_device);
+	if (rc < 0) {
+		PRINTKW(KERN_ERR "misc_register (minor %d) failed with %d\n",
+			zcrypt_misc_device.minor, rc);
+		goto out;
+	}
+
+	/* Set up the proc file system */
+	zcrypt_entry = create_proc_entry("driver/z90crypt", 0644, NULL);
+	if (!zcrypt_entry) {
+		PRINTK("Couldn't create z90crypt proc entry\n");
+		rc = -ENOMEM;
+		goto out_misc;
+	}
+	zcrypt_entry->nlink = 1;
+	zcrypt_entry->data = NULL;
+	zcrypt_entry->read_proc = zcrypt_status_read;
+	zcrypt_entry->write_proc = zcrypt_status_write;
+
+	return 0;
+
+out_misc:
+	misc_deregister(&zcrypt_misc_device);
+out:
+	return rc;
+}
+
+/**
+ * The module termination code.
+ */
+void zcrypt_api_exit(void)
+{
+	remove_proc_entry("driver/z90crypt", NULL);
+	misc_deregister(&zcrypt_misc_device);
+}
+
+#ifndef CONFIG_ZCRYPT_MONOLITHIC
+module_init(zcrypt_api_init);
+module_exit(zcrypt_api_exit);
+#endif
diff --git a/drivers/s390/crypto/zcrypt_api.h b/drivers/s390/crypto/zcrypt_api.h
new file mode 100644
index 000000000000..1f0e61f2e9b4
--- /dev/null
+++ b/drivers/s390/crypto/zcrypt_api.h
@@ -0,0 +1,140 @@
+/*
+ *  linux/drivers/s390/crypto/zcrypt_api.h
+ *
+ *  zcrypt 2.0.0
+ *
+ *  Copyright (C)  2001, 2006 IBM Corporation
+ *  Author(s): Robert Burroughs
+ *	       Eric Rossman (edrossma@us.ibm.com)
+ *	       Cornelia Huck <cornelia.huck@de.ibm.com>
+ *
+ *  Hotplug & misc device support: Jochen Roehrig (roehrig@de.ibm.com)
+ *  Major cleanup & driver split: Martin Schwidefsky <schwidefsky@de.ibm.com>
+ *				  Ralph Wuerthner <rwuerthn@de.ibm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2, or (at your option)
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#ifndef _ZCRYPT_API_H_
+#define _ZCRYPT_API_H_
+
+/**
+ * Macro definitions
+ *
+ * PDEBUG debugs in the form "zcrypt: function_name -> message"
+ *
+ * PRINTK is like PDEBUG, except that it is always enabled
+ * PRINTKN is like PRINTK, except that it does not include the function name
+ * PRINTKW is like PRINTK, except that it uses KERN_WARNING
+ * PRINTKC is like PRINTK, except that it uses KERN_CRIT
+ */
+#define DEV_NAME	"zcrypt"
+
+#define PRINTK(fmt, args...) \
+	printk(KERN_DEBUG DEV_NAME ": %s -> " fmt, __FUNCTION__ , ## args)
+#define PRINTKN(fmt, args...) \
+	printk(KERN_DEBUG DEV_NAME ": " fmt, ## args)
+#define PRINTKW(fmt, args...) \
+	printk(KERN_WARNING DEV_NAME ": %s -> " fmt, __FUNCTION__ , ## args)
+#define PRINTKC(fmt, args...) \
+	printk(KERN_CRIT DEV_NAME ": %s -> " fmt, __FUNCTION__ , ## args)
+
+#ifdef ZCRYPT_DEBUG
+#define PDEBUG(fmt, args...) \
+	printk(KERN_DEBUG DEV_NAME ": %s -> " fmt, __FUNCTION__ , ## args)
+#else
+#define PDEBUG(fmt, args...) do {} while (0)
+#endif
+
+#include "ap_bus.h"
+#include <asm/zcrypt.h>
+
+/* deprecated status calls */
+#define ICAZ90STATUS		_IOR(ZCRYPT_IOCTL_MAGIC, 0x10, struct ica_z90_status)
+#define Z90STAT_PCIXCCCOUNT	_IOR(ZCRYPT_IOCTL_MAGIC, 0x43, int)
+
+/**
+ * This structure is deprecated and the corresponding ioctl() has been
+ * replaced with individual ioctl()s for each piece of data!
+ */
+struct ica_z90_status {
+	int totalcount;
+	int leedslitecount; // PCICA
+	int leeds2count;    // PCICC
+	// int PCIXCCCount; is not in struct for backward compatibility
+	int requestqWaitCount;
+	int pendingqWaitCount;
+	int totalOpenCount;
+	int cryptoDomain;
+	// status: 0=not there, 1=PCICA, 2=PCICC, 3=PCIXCC_MCL2, 4=PCIXCC_MCL3,
+	//	   5=CEX2C
+	unsigned char status[64];
+	// qdepth: # work elements waiting for each device
+	unsigned char qdepth[64];
+};
+
+/**
+ * device type for an actual device is either PCICA, PCICC, PCIXCC_MCL2,
+ * PCIXCC_MCL3, CEX2C, or CEX2A
+ *
+ * NOTE: PCIXCC_MCL3 refers to a PCIXCC with May 2004 version of Licensed
+ *	 Internal Code (LIC) (EC J12220 level 29).
+ *	 PCIXCC_MCL2 refers to any LIC before this level.
+ */
+#define ZCRYPT_PCICA		1
+#define ZCRYPT_PCICC		2
+#define ZCRYPT_PCIXCC_MCL2	3
+#define ZCRYPT_PCIXCC_MCL3	4
+#define ZCRYPT_CEX2C		5
+#define ZCRYPT_CEX2A		6
+
+struct zcrypt_device;
+
+struct zcrypt_ops {
+	long (*rsa_modexpo)(struct zcrypt_device *, struct ica_rsa_modexpo *);
+	long (*rsa_modexpo_crt)(struct zcrypt_device *,
+				struct ica_rsa_modexpo_crt *);
+};
+
+struct zcrypt_device {
+	struct list_head list;		/* Device list. */
+	spinlock_t lock;		/* Per device lock. */
+	struct kref refcount;		/* device refcounting */
+	struct ap_device *ap_dev;	/* The "real" ap device. */
+	struct zcrypt_ops *ops;		/* Crypto operations. */
+	int online;			/* User online/offline */
+
+	int user_space_type;		/* User space device id. */
+	char *type_string;		/* User space device name. */
+	int min_mod_size;		/* Min number of bits. */
+	int max_mod_size;		/* Max number of bits. */
+	int short_crt;			/* Card has crt length restriction. */
+	int speed_rating;		/* Speed of the crypto device. */
+
+	int request_count;		/* # current requests. */
+
+	struct ap_message reply;	/* Per-device reply structure. */
+};
+
+struct zcrypt_device *zcrypt_device_alloc(size_t);
+void zcrypt_device_free(struct zcrypt_device *);
+void zcrypt_device_get(struct zcrypt_device *);
+int zcrypt_device_put(struct zcrypt_device *);
+int zcrypt_device_register(struct zcrypt_device *);
+void zcrypt_device_unregister(struct zcrypt_device *);
+int zcrypt_api_init(void);
+void zcrypt_api_exit(void);
+
+#endif /* _ZCRYPT_API_H_ */
diff --git a/include/asm-s390/zcrypt.h b/include/asm-s390/zcrypt.h
new file mode 100644
index 000000000000..0d6a3e2a3349
--- /dev/null
+++ b/include/asm-s390/zcrypt.h
@@ -0,0 +1,207 @@
+/*
+ *  include/asm-s390/zcrypt.h
+ *
+ *  zcrypt 2.0.0 (user-visible header)
+ *
+ *  Copyright (C)  2001, 2006 IBM Corporation
+ *  Author(s): Robert Burroughs
+ *	       Eric Rossman (edrossma@us.ibm.com)
+ *
+ *  Hotplug & misc device support: Jochen Roehrig (roehrig@de.ibm.com)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2, or (at your option)
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#ifndef __ASM_S390_ZCRYPT_H
+#define __ASM_S390_ZCRYPT_H
+
+#define ZCRYPT_VERSION 2
+#define ZCRYPT_RELEASE 1
+#define ZCRYPT_VARIANT 0
+
+#include <linux/ioctl.h>
+#include <linux/compiler.h>
+
+/**
+ * struct ica_rsa_modexpo
+ *
+ * Requirements:
+ * - outputdatalength is at least as large as inputdatalength.
+ * - All key parts are right justified in their fields, padded on
+ *   the left with zeroes.
+ * - length(b_key) = inputdatalength
+ * - length(n_modulus) = inputdatalength
+ */
+struct ica_rsa_modexpo {
+	char __user *	inputdata;
+	unsigned int	inputdatalength;
+	char __user *	outputdata;
+	unsigned int	outputdatalength;
+	char __user *	b_key;
+	char __user *	n_modulus;
+};
+
+/**
+ * struct ica_rsa_modexpo_crt
+ *
+ * Requirements:
+ * - inputdatalength is even.
+ * - outputdatalength is at least as large as inputdatalength.
+ * - All key parts are right justified in their fields, padded on
+ *   the left with zeroes.
+ * - length(bp_key)	= inputdatalength/2 + 8
+ * - length(bq_key)	= inputdatalength/2
+ * - length(np_key)	= inputdatalength/2 + 8
+ * - length(nq_key)	= inputdatalength/2
+ * - length(u_mult_inv) = inputdatalength/2 + 8
+ */
+struct ica_rsa_modexpo_crt {
+	char __user *	inputdata;
+	unsigned int	inputdatalength;
+	char __user *	outputdata;
+	unsigned int	outputdatalength;
+	char __user *	bp_key;
+	char __user *	bq_key;
+	char __user *	np_prime;
+	char __user *	nq_prime;
+	char __user *	u_mult_inv;
+};
+
+#define ZCRYPT_IOCTL_MAGIC 'z'
+
+/**
+ * Interface notes:
+ *
+ * The ioctl()s which are implemented (along with relevant details)
+ * are:
+ *
+ *   ICARSAMODEXPO
+ *     Perform an RSA operation using a Modulus-Exponent pair
+ *     This takes an ica_rsa_modexpo struct as its arg.
+ *
+ *     NOTE: please refer to the comments preceding this structure
+ *	     for the implementation details for the contents of the
+ *	     block
+ *
+ *   ICARSACRT
+ *     Perform an RSA operation using a Chinese-Remainder Theorem key
+ *     This takes an ica_rsa_modexpo_crt struct as its arg.
+ *
+ *     NOTE: please refer to the comments preceding this structure
+ *	     for the implementation details for the contents of the
+ *	     block
+ *
+ *   Z90STAT_TOTALCOUNT
+ *     Return an integer count of all device types together.
+ *
+ *   Z90STAT_PCICACOUNT
+ *     Return an integer count of all PCICAs.
+ *
+ *   Z90STAT_PCICCCOUNT
+ *     Return an integer count of all PCICCs.
+ *
+ *   Z90STAT_PCIXCCMCL2COUNT
+ *     Return an integer count of all MCL2 PCIXCCs.
+ *
+ *   Z90STAT_PCIXCCMCL3COUNT
+ *     Return an integer count of all MCL3 PCIXCCs.
+ *
+ *   Z90STAT_CEX2CCOUNT
+ *     Return an integer count of all CEX2Cs.
+ *
+ *   Z90STAT_CEX2ACOUNT
+ *     Return an integer count of all CEX2As.
+ *
+ *   Z90STAT_REQUESTQ_COUNT
+ *     Return an integer count of the number of entries waiting to be
+ *     sent to a device.
+ *
+ *   Z90STAT_PENDINGQ_COUNT
+ *     Return an integer count of the number of entries sent to a
+ *     device awaiting the reply.
+ *
+ *   Z90STAT_TOTALOPEN_COUNT
+ *     Return an integer count of the number of open file handles.
+ *
+ *   Z90STAT_DOMAIN_INDEX
+ *     Return the integer value of the Cryptographic Domain.
+ *
+ *   Z90STAT_STATUS_MASK
+ *     Return an 64 element array of unsigned chars for the status of
+ *     all devices.
+ *	 0x01: PCICA
+ *	 0x02: PCICC
+ *	 0x03: PCIXCC_MCL2
+ *	 0x04: PCIXCC_MCL3
+ *	 0x05: CEX2C
+ *	 0x06: CEX2A
+ *	 0x0d: device is disabled via the proc filesystem
+ *
+ *   Z90STAT_QDEPTH_MASK
+ *     Return an 64 element array of unsigned chars for the queue
+ *     depth of all devices.
+ *
+ *   Z90STAT_PERDEV_REQCNT
+ *     Return an 64 element array of unsigned integers for the number
+ *     of successfully completed requests per device since the device
+ *     was detected and made available.
+ *
+ *   ICAZ90STATUS (deprecated)
+ *     Return some device driver status in a ica_z90_status struct
+ *     This takes an ica_z90_status struct as its arg.
+ *
+ *     NOTE: this ioctl() is deprecated, and has been replaced with
+ *	     single ioctl()s for each type of status being requested
+ *
+ *   Z90STAT_PCIXCCCOUNT (deprecated)
+ *     Return an integer count of all PCIXCCs (MCL2 + MCL3).
+ *     This is DEPRECATED now that MCL3 PCIXCCs are treated differently from
+ *     MCL2 PCIXCCs.
+ *
+ *   Z90QUIESCE (not recommended)
+ *     Quiesce the driver.  This is intended to stop all new
+ *     requests from being processed.  Its use is NOT recommended,
+ *     except in circumstances where there is no other way to stop
+ *     callers from accessing the driver.  Its original use was to
+ *     allow the driver to be "drained" of work in preparation for
+ *     a system shutdown.
+ *
+ *     NOTE: once issued, this ban on new work cannot be undone
+ *	     except by unloading and reloading the driver.
+ */
+
+/**
+ * Supported ioctl calls
+ */
+#define ICARSAMODEXPO	_IOC(_IOC_READ|_IOC_WRITE, ZCRYPT_IOCTL_MAGIC, 0x05, 0)
+#define ICARSACRT	_IOC(_IOC_READ|_IOC_WRITE, ZCRYPT_IOCTL_MAGIC, 0x06, 0)
+
+/* New status calls */
+#define Z90STAT_TOTALCOUNT	_IOR(ZCRYPT_IOCTL_MAGIC, 0x40, int)
+#define Z90STAT_PCICACOUNT	_IOR(ZCRYPT_IOCTL_MAGIC, 0x41, int)
+#define Z90STAT_PCICCCOUNT	_IOR(ZCRYPT_IOCTL_MAGIC, 0x42, int)
+#define Z90STAT_PCIXCCMCL2COUNT	_IOR(ZCRYPT_IOCTL_MAGIC, 0x4b, int)
+#define Z90STAT_PCIXCCMCL3COUNT	_IOR(ZCRYPT_IOCTL_MAGIC, 0x4c, int)
+#define Z90STAT_CEX2CCOUNT	_IOR(ZCRYPT_IOCTL_MAGIC, 0x4d, int)
+#define Z90STAT_CEX2ACOUNT	_IOR(ZCRYPT_IOCTL_MAGIC, 0x4e, int)
+#define Z90STAT_REQUESTQ_COUNT	_IOR(ZCRYPT_IOCTL_MAGIC, 0x44, int)
+#define Z90STAT_PENDINGQ_COUNT	_IOR(ZCRYPT_IOCTL_MAGIC, 0x45, int)
+#define Z90STAT_TOTALOPEN_COUNT _IOR(ZCRYPT_IOCTL_MAGIC, 0x46, int)
+#define Z90STAT_DOMAIN_INDEX	_IOR(ZCRYPT_IOCTL_MAGIC, 0x47, int)
+#define Z90STAT_STATUS_MASK	_IOR(ZCRYPT_IOCTL_MAGIC, 0x48, char[64])
+#define Z90STAT_QDEPTH_MASK	_IOR(ZCRYPT_IOCTL_MAGIC, 0x49, char[64])
+#define Z90STAT_PERDEV_REQCNT	_IOR(ZCRYPT_IOCTL_MAGIC, 0x4a, int[64])
+
+#endif /* __ASM_S390_ZCRYPT_H */
-- 
cgit v1.2.3


From 5432114baf0300286a6ca1b0aea549492a379432 Mon Sep 17 00:00:00 2001
From: Ralph Wuerthner <rwuerthn@de.ibm.com>
Date: Wed, 20 Sep 2006 15:58:36 +0200
Subject: [S390] zcrypt secure key cryptography extension.

Allow the user space to send extended cprb messages directly to the
PCIXCC / CEX2C cards. This allows the CCA library to construct special
crypto requests that use "secure" keys that are stored on the card.

Signed-off-by: Ralph Wuerthner <rwuerthn@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 drivers/s390/crypto/zcrypt_api.c     | 112 ++++++++++++++-
 drivers/s390/crypto/zcrypt_api.h     |   3 +-
 drivers/s390/crypto/zcrypt_cca_key.h |   2 +-
 drivers/s390/crypto/zcrypt_cex2a.c   |   2 +-
 drivers/s390/crypto/zcrypt_cex2a.h   |   2 +-
 drivers/s390/crypto/zcrypt_error.h   |   2 +-
 drivers/s390/crypto/zcrypt_mono.c    |   2 +-
 drivers/s390/crypto/zcrypt_pcica.c   |   2 +-
 drivers/s390/crypto/zcrypt_pcica.h   |   2 +-
 drivers/s390/crypto/zcrypt_pcicc.c   |   2 +-
 drivers/s390/crypto/zcrypt_pcicc.h   |   2 +-
 drivers/s390/crypto/zcrypt_pcixcc.c  | 263 +++++++++++++++++++++++++++++++++--
 drivers/s390/crypto/zcrypt_pcixcc.h  |   2 +-
 include/asm-s390/zcrypt.h            |  80 ++++++++++-
 14 files changed, 452 insertions(+), 26 deletions(-)

(limited to 'include')

diff --git a/drivers/s390/crypto/zcrypt_api.c b/drivers/s390/crypto/zcrypt_api.c
index b3fe003b3d2d..1edc10a7a6f2 100644
--- a/drivers/s390/crypto/zcrypt_api.c
+++ b/drivers/s390/crypto/zcrypt_api.c
@@ -1,7 +1,7 @@
 /*
  *  linux/drivers/s390/crypto/zcrypt_api.c
  *
- *  zcrypt 2.0.0
+ *  zcrypt 2.1.0
  *
  *  Copyright (C)  2001, 2006 IBM Corporation
  *  Author(s): Robert Burroughs
@@ -392,6 +392,41 @@ static long zcrypt_rsa_crt(struct ica_rsa_modexpo_crt *crt)
 	return -ENODEV;
 }
 
+static long zcrypt_send_cprb(struct ica_xcRB *xcRB)
+{
+	struct zcrypt_device *zdev;
+	int rc;
+
+	spin_lock_bh(&zcrypt_device_lock);
+	list_for_each_entry(zdev, &zcrypt_device_list, list) {
+		if (!zdev->online || !zdev->ops->send_cprb ||
+		    (xcRB->user_defined != AUTOSELECT &&
+			AP_QID_DEVICE(zdev->ap_dev->qid) != xcRB->user_defined)
+		    )
+			continue;
+		zcrypt_device_get(zdev);
+		get_device(&zdev->ap_dev->device);
+		zdev->request_count++;
+		__zcrypt_decrease_preference(zdev);
+		spin_unlock_bh(&zcrypt_device_lock);
+		if (try_module_get(zdev->ap_dev->drv->driver.owner)) {
+			rc = zdev->ops->send_cprb(zdev, xcRB);
+			module_put(zdev->ap_dev->drv->driver.owner);
+		}
+		else
+			rc = -EAGAIN;
+		spin_lock_bh(&zcrypt_device_lock);
+		zdev->request_count--;
+		__zcrypt_increase_preference(zdev);
+		put_device(&zdev->ap_dev->device);
+		zcrypt_device_put(zdev);
+		spin_unlock_bh(&zcrypt_device_lock);
+		return rc;
+	}
+	spin_unlock_bh(&zcrypt_device_lock);
+	return -ENODEV;
+}
+
 static void zcrypt_status_mask(char status[AP_DEVICES])
 {
 	struct zcrypt_device *zdev;
@@ -535,6 +570,18 @@ static long zcrypt_unlocked_ioctl(struct file *filp, unsigned int cmd,
 			return rc;
 		return put_user(crt.outputdatalength, &ucrt->outputdatalength);
 	}
+	case ZSECSENDCPRB: {
+		struct ica_xcRB __user *uxcRB = (void __user *) arg;
+		struct ica_xcRB xcRB;
+		if (copy_from_user(&xcRB, uxcRB, sizeof(xcRB)))
+			return -EFAULT;
+		do {
+			rc = zcrypt_send_cprb(&xcRB);
+		} while (rc == -EAGAIN);
+		if (copy_to_user(uxcRB, &xcRB, sizeof(xcRB)))
+			return -EFAULT;
+		return rc;
+	}
 	case Z90STAT_STATUS_MASK: {
 		char status[AP_DEVICES];
 		zcrypt_status_mask(status);
@@ -683,6 +730,67 @@ static long trans_modexpo_crt32(struct file *filp, unsigned int cmd,
 	return rc;
 }
 
+struct compat_ica_xcRB {
+	unsigned short	agent_ID;
+	unsigned int	user_defined;
+	unsigned short	request_ID;
+	unsigned int	request_control_blk_length;
+	unsigned char	padding1[16 - sizeof (compat_uptr_t)];
+	compat_uptr_t	request_control_blk_addr;
+	unsigned int	request_data_length;
+	char		padding2[16 - sizeof (compat_uptr_t)];
+	compat_uptr_t	request_data_address;
+	unsigned int	reply_control_blk_length;
+	char		padding3[16 - sizeof (compat_uptr_t)];
+	compat_uptr_t	reply_control_blk_addr;
+	unsigned int	reply_data_length;
+	char		padding4[16 - sizeof (compat_uptr_t)];
+	compat_uptr_t	reply_data_addr;
+	unsigned short	priority_window;
+	unsigned int	status;
+} __attribute__((packed));
+
+static long trans_xcRB32(struct file *filp, unsigned int cmd,
+			 unsigned long arg)
+{
+	struct compat_ica_xcRB __user *uxcRB32 = compat_ptr(arg);
+	struct compat_ica_xcRB xcRB32;
+	struct ica_xcRB xcRB64;
+	long rc;
+
+	if (copy_from_user(&xcRB32, uxcRB32, sizeof(xcRB32)))
+		return -EFAULT;
+	xcRB64.agent_ID = xcRB32.agent_ID;
+	xcRB64.user_defined = xcRB32.user_defined;
+	xcRB64.request_ID = xcRB32.request_ID;
+	xcRB64.request_control_blk_length =
+		xcRB32.request_control_blk_length;
+	xcRB64.request_control_blk_addr =
+		compat_ptr(xcRB32.request_control_blk_addr);
+	xcRB64.request_data_length =
+		xcRB32.request_data_length;
+	xcRB64.request_data_address =
+		compat_ptr(xcRB32.request_data_address);
+	xcRB64.reply_control_blk_length =
+		xcRB32.reply_control_blk_length;
+	xcRB64.reply_control_blk_addr =
+		compat_ptr(xcRB32.reply_control_blk_addr);
+	xcRB64.reply_data_length = xcRB32.reply_data_length;
+	xcRB64.reply_data_addr =
+		compat_ptr(xcRB32.reply_data_addr);
+	xcRB64.priority_window = xcRB32.priority_window;
+	xcRB64.status = xcRB32.status;
+	do {
+		rc = zcrypt_send_cprb(&xcRB64);
+	} while (rc == -EAGAIN);
+	xcRB32.reply_control_blk_length = xcRB64.reply_control_blk_length;
+	xcRB32.reply_data_length = xcRB64.reply_data_length;
+	xcRB32.status = xcRB64.status;
+	if (copy_to_user(uxcRB32, &xcRB32, sizeof(xcRB32)))
+			return -EFAULT;
+	return rc;
+}
+
 long zcrypt_compat_ioctl(struct file *filp, unsigned int cmd,
 			 unsigned long arg)
 {
@@ -690,6 +798,8 @@ long zcrypt_compat_ioctl(struct file *filp, unsigned int cmd,
 		return trans_modexpo32(filp, cmd, arg);
 	if (cmd == ICARSACRT)
 		return trans_modexpo_crt32(filp, cmd, arg);
+	if (cmd == ZSECSENDCPRB)
+		return trans_xcRB32(filp, cmd, arg);
 	return zcrypt_unlocked_ioctl(filp, cmd, arg);
 }
 #endif
diff --git a/drivers/s390/crypto/zcrypt_api.h b/drivers/s390/crypto/zcrypt_api.h
index 1f0e61f2e9b4..de4877ee618f 100644
--- a/drivers/s390/crypto/zcrypt_api.h
+++ b/drivers/s390/crypto/zcrypt_api.h
@@ -1,7 +1,7 @@
 /*
  *  linux/drivers/s390/crypto/zcrypt_api.h
  *
- *  zcrypt 2.0.0
+ *  zcrypt 2.1.0
  *
  *  Copyright (C)  2001, 2006 IBM Corporation
  *  Author(s): Robert Burroughs
@@ -106,6 +106,7 @@ struct zcrypt_ops {
 	long (*rsa_modexpo)(struct zcrypt_device *, struct ica_rsa_modexpo *);
 	long (*rsa_modexpo_crt)(struct zcrypt_device *,
 				struct ica_rsa_modexpo_crt *);
+	long (*send_cprb)(struct zcrypt_device *, struct ica_xcRB *);
 };
 
 struct zcrypt_device {
diff --git a/drivers/s390/crypto/zcrypt_cca_key.h b/drivers/s390/crypto/zcrypt_cca_key.h
index c80f40d44197..8dbcf0eef3e5 100644
--- a/drivers/s390/crypto/zcrypt_cca_key.h
+++ b/drivers/s390/crypto/zcrypt_cca_key.h
@@ -1,7 +1,7 @@
 /*
  *  linux/drivers/s390/crypto/zcrypt_cca_key.h
  *
- *  zcrypt 2.0.0
+ *  zcrypt 2.1.0
  *
  *  Copyright (C)  2001, 2006 IBM Corporation
  *  Author(s): Robert Burroughs
diff --git a/drivers/s390/crypto/zcrypt_cex2a.c b/drivers/s390/crypto/zcrypt_cex2a.c
index 350248e5cd93..a62b00083d0c 100644
--- a/drivers/s390/crypto/zcrypt_cex2a.c
+++ b/drivers/s390/crypto/zcrypt_cex2a.c
@@ -1,7 +1,7 @@
 /*
  *  linux/drivers/s390/crypto/zcrypt_cex2a.c
  *
- *  zcrypt 2.0.0
+ *  zcrypt 2.1.0
  *
  *  Copyright (C)  2001, 2006 IBM Corporation
  *  Author(s): Robert Burroughs
diff --git a/drivers/s390/crypto/zcrypt_cex2a.h b/drivers/s390/crypto/zcrypt_cex2a.h
index 61a78c32dce4..8f69d1dacab8 100644
--- a/drivers/s390/crypto/zcrypt_cex2a.h
+++ b/drivers/s390/crypto/zcrypt_cex2a.h
@@ -1,7 +1,7 @@
 /*
  *  linux/drivers/s390/crypto/zcrypt_cex2a.h
  *
- *  zcrypt 2.0.0
+ *  zcrypt 2.1.0
  *
  *  Copyright (C)  2001, 2006 IBM Corporation
  *  Author(s): Robert Burroughs
diff --git a/drivers/s390/crypto/zcrypt_error.h b/drivers/s390/crypto/zcrypt_error.h
index b22bd055a03b..2cb616ba8bec 100644
--- a/drivers/s390/crypto/zcrypt_error.h
+++ b/drivers/s390/crypto/zcrypt_error.h
@@ -1,7 +1,7 @@
 /*
  *  linux/drivers/s390/crypto/zcrypt_error.h
  *
- *  zcrypt 2.0.0
+ *  zcrypt 2.1.0
  *
  *  Copyright (C)  2001, 2006 IBM Corporation
  *  Author(s): Robert Burroughs
diff --git a/drivers/s390/crypto/zcrypt_mono.c b/drivers/s390/crypto/zcrypt_mono.c
index f48b61a6126c..2a9349ad68b7 100644
--- a/drivers/s390/crypto/zcrypt_mono.c
+++ b/drivers/s390/crypto/zcrypt_mono.c
@@ -1,7 +1,7 @@
 /*
  *  linux/drivers/s390/crypto/zcrypt_mono.c
  *
- *  zcrypt 2.0.0
+ *  zcrypt 2.1.0
  *
  *  Copyright (C)  2001, 2006 IBM Corporation
  *  Author(s): Robert Burroughs
diff --git a/drivers/s390/crypto/zcrypt_pcica.c b/drivers/s390/crypto/zcrypt_pcica.c
index 0ff56e86caae..b6a4ecdc8025 100644
--- a/drivers/s390/crypto/zcrypt_pcica.c
+++ b/drivers/s390/crypto/zcrypt_pcica.c
@@ -1,7 +1,7 @@
 /*
  *  linux/drivers/s390/crypto/zcrypt_pcica.c
  *
- *  zcrypt 2.0.0
+ *  zcrypt 2.1.0
  *
  *  Copyright (C)  2001, 2006 IBM Corporation
  *  Author(s): Robert Burroughs
diff --git a/drivers/s390/crypto/zcrypt_pcica.h b/drivers/s390/crypto/zcrypt_pcica.h
index a08a4f8c33c9..3be11187f6df 100644
--- a/drivers/s390/crypto/zcrypt_pcica.h
+++ b/drivers/s390/crypto/zcrypt_pcica.h
@@ -1,7 +1,7 @@
 /*
  *  linux/drivers/s390/crypto/zcrypt_pcica.h
  *
- *  zcrypt 2.0.0
+ *  zcrypt 2.1.0
  *
  *  Copyright (C)  2001, 2006 IBM Corporation
  *  Author(s): Robert Burroughs
diff --git a/drivers/s390/crypto/zcrypt_pcicc.c b/drivers/s390/crypto/zcrypt_pcicc.c
index 900362983fec..f295a403b29a 100644
--- a/drivers/s390/crypto/zcrypt_pcicc.c
+++ b/drivers/s390/crypto/zcrypt_pcicc.c
@@ -1,7 +1,7 @@
 /*
  *  linux/drivers/s390/crypto/zcrypt_pcicc.c
  *
- *  zcrypt 2.0.0
+ *  zcrypt 2.1.0
  *
  *  Copyright (C)  2001, 2006 IBM Corporation
  *  Author(s): Robert Burroughs
diff --git a/drivers/s390/crypto/zcrypt_pcicc.h b/drivers/s390/crypto/zcrypt_pcicc.h
index 027bafc7312a..6d4454846c8f 100644
--- a/drivers/s390/crypto/zcrypt_pcicc.h
+++ b/drivers/s390/crypto/zcrypt_pcicc.h
@@ -1,7 +1,7 @@
 /*
  *  linux/drivers/s390/crypto/zcrypt_pcicc.h
  *
- *  zcrypt 2.0.0
+ *  zcrypt 2.1.0
  *
  *  Copyright (C)  2001, 2006 IBM Corporation
  *  Author(s): Robert Burroughs
diff --git a/drivers/s390/crypto/zcrypt_pcixcc.c b/drivers/s390/crypto/zcrypt_pcixcc.c
index 6064cf58be43..2da8b9381407 100644
--- a/drivers/s390/crypto/zcrypt_pcixcc.c
+++ b/drivers/s390/crypto/zcrypt_pcixcc.c
@@ -1,7 +1,7 @@
 /*
  *  linux/drivers/s390/crypto/zcrypt_pcixcc.c
  *
- *  zcrypt 2.0.0
+ *  zcrypt 2.1.0
  *
  *  Copyright (C)  2001, 2006 IBM Corporation
  *  Author(s): Robert Burroughs
@@ -60,6 +60,15 @@
 
 #define PCIXCC_CLEANUP_TIME	(15*HZ)
 
+#define CEIL4(x) ((((x)+3)/4)*4)
+
+struct response_type {
+	struct completion work;
+	int type;
+};
+#define PCIXCC_RESPONSE_TYPE_ICA  0
+#define PCIXCC_RESPONSE_TYPE_XCRB 1
+
 static struct ap_device_id zcrypt_pcixcc_ids[] = {
 	{ AP_DEVICE(AP_DEVICE_TYPE_PCIXCC) },
 	{ AP_DEVICE(AP_DEVICE_TYPE_CEX2C) },
@@ -243,6 +252,108 @@ static int ICACRT_msg_to_type6CRT_msgX(struct zcrypt_device *zdev,
 	return 0;
 }
 
+/**
+ * Convert a XCRB message to a type6 CPRB message.
+ *
+ * @zdev: crypto device pointer
+ * @ap_msg: pointer to AP message
+ * @xcRB: pointer to user input data
+ *
+ * Returns 0 on success or -EFAULT.
+ */
+struct type86_fmt2_msg {
+	struct type86_hdr hdr;
+	struct type86_fmt2_ext fmt2;
+} __attribute__((packed));
+
+static int XCRB_msg_to_type6CPRB_msgX(struct zcrypt_device *zdev,
+				       struct ap_message *ap_msg,
+				       struct ica_xcRB *xcRB)
+{
+	static struct type6_hdr static_type6_hdrX = {
+		.type		=  0x06,
+		.offset1	=  0x00000058,
+	};
+	struct {
+		struct type6_hdr hdr;
+		struct ica_CPRBX cprbx;
+	} __attribute__((packed)) *msg = ap_msg->message;
+
+	int rcblen = CEIL4(xcRB->request_control_blk_length);
+	int replylen;
+	char *req_data = ap_msg->message + sizeof(struct type6_hdr) + rcblen;
+	char *function_code;
+
+	/* length checks */
+	ap_msg->length = sizeof(struct type6_hdr) +
+		CEIL4(xcRB->request_control_blk_length) +
+		xcRB->request_data_length;
+	if (ap_msg->length > PCIXCC_MAX_XCRB_MESSAGE_SIZE) {
+		PRINTK("Combined message is too large (%ld/%d/%d).\n",
+		    sizeof(struct type6_hdr),
+		    xcRB->request_control_blk_length,
+		    xcRB->request_data_length);
+		return -EFAULT;
+	}
+	if (CEIL4(xcRB->reply_control_blk_length) >
+	    PCIXCC_MAX_XCRB_REPLY_SIZE) {
+		PDEBUG("Reply CPRB length is too large (%d).\n",
+		    xcRB->request_control_blk_length);
+		return -EFAULT;
+	}
+	if (CEIL4(xcRB->reply_data_length) > PCIXCC_MAX_XCRB_DATA_SIZE) {
+		PDEBUG("Reply data block length is too large (%d).\n",
+		    xcRB->reply_data_length);
+		return -EFAULT;
+	}
+	replylen = CEIL4(xcRB->reply_control_blk_length) +
+		CEIL4(xcRB->reply_data_length) +
+		sizeof(struct type86_fmt2_msg);
+	if (replylen > PCIXCC_MAX_XCRB_RESPONSE_SIZE) {
+		PDEBUG("Reply CPRB + data block > PCIXCC_MAX_XCRB_RESPONSE_SIZE"
+		       " (%d/%d/%d).\n",
+		       sizeof(struct type86_fmt2_msg),
+		       xcRB->reply_control_blk_length,
+		       xcRB->reply_data_length);
+		xcRB->reply_control_blk_length = PCIXCC_MAX_XCRB_RESPONSE_SIZE -
+			(sizeof(struct type86_fmt2_msg) +
+			    CEIL4(xcRB->reply_data_length));
+		PDEBUG("Capping Reply CPRB length at %d\n",
+		       xcRB->reply_control_blk_length);
+	}
+
+	/* prepare type6 header */
+	msg->hdr = static_type6_hdrX;
+	memcpy(msg->hdr.agent_id , &(xcRB->agent_ID), sizeof(xcRB->agent_ID));
+	msg->hdr.ToCardLen1 = xcRB->request_control_blk_length;
+	if (xcRB->request_data_length) {
+		msg->hdr.offset2 = msg->hdr.offset1 + rcblen;
+		msg->hdr.ToCardLen2 = xcRB->request_data_length;
+	}
+	msg->hdr.FromCardLen1 = xcRB->reply_control_blk_length;
+	msg->hdr.FromCardLen2 = xcRB->reply_data_length;
+
+	/* prepare CPRB */
+	if (copy_from_user(&(msg->cprbx), xcRB->request_control_blk_addr,
+		    xcRB->request_control_blk_length))
+		return -EFAULT;
+	if (msg->cprbx.cprb_len + sizeof(msg->hdr.function_code) >
+	    xcRB->request_control_blk_length) {
+		PDEBUG("cprb_len too large (%d/%d)\n", msg->cprbx.cprb_len,
+		    xcRB->request_control_blk_length);
+		return -EFAULT;
+	}
+	function_code = ((unsigned char *)&msg->cprbx) + msg->cprbx.cprb_len;
+	memcpy(msg->hdr.function_code, function_code, sizeof(msg->hdr.function_code));
+
+	/* copy data block */
+	if (xcRB->request_data_length &&
+	    copy_from_user(req_data, xcRB->request_data_address,
+		xcRB->request_data_length))
+		return -EFAULT;
+	return 0;
+}
+
 /**
  * Copy results from a type 86 ICA reply message back to user space.
  *
@@ -363,6 +474,37 @@ static int convert_type86_ica(struct zcrypt_device *zdev,
 	return 0;
 }
 
+/**
+ * Copy results from a type 86 XCRB reply message back to user space.
+ *
+ * @zdev: crypto device pointer
+ * @reply: reply AP message.
+ * @xcRB: pointer to XCRB
+ *
+ * Returns 0 on success or -EINVAL, -EFAULT, -EAGAIN in case of an error.
+ */
+static int convert_type86_xcrb(struct zcrypt_device *zdev,
+			       struct ap_message *reply,
+			       struct ica_xcRB *xcRB)
+{
+	struct type86_fmt2_msg *msg = reply->message;
+	char *data = reply->message;
+
+	/* Copy CPRB to user */
+	if (copy_to_user(xcRB->reply_control_blk_addr,
+		data + msg->fmt2.offset1, msg->fmt2.count1))
+		return -EFAULT;
+	xcRB->reply_control_blk_length = msg->fmt2.count1;
+
+	/* Copy data buffer to user */
+	if (msg->fmt2.count2)
+		if (copy_to_user(xcRB->reply_data_addr,
+			data + msg->fmt2.offset2, msg->fmt2.count2))
+			return -EFAULT;
+	xcRB->reply_data_length = msg->fmt2.count2;
+	return 0;
+}
+
 static int convert_response_ica(struct zcrypt_device *zdev,
 			    struct ap_message *reply,
 			    char __user *outputdata,
@@ -391,6 +533,36 @@ static int convert_response_ica(struct zcrypt_device *zdev,
 	}
 }
 
+static int convert_response_xcrb(struct zcrypt_device *zdev,
+			    struct ap_message *reply,
+			    struct ica_xcRB *xcRB)
+{
+	struct type86x_reply *msg = reply->message;
+
+	/* Response type byte is the second byte in the response. */
+	switch (((unsigned char *) reply->message)[1]) {
+	case TYPE82_RSP_CODE:
+	case TYPE88_RSP_CODE:
+		xcRB->status = 0x0008044DL; /* HDD_InvalidParm */
+		return convert_error(zdev, reply);
+	case TYPE86_RSP_CODE:
+		if (msg->hdr.reply_code) {
+			memcpy(&(xcRB->status), msg->fmt2.apfs, sizeof(u32));
+			return convert_error(zdev, reply);
+		}
+		if (msg->cprbx.cprb_ver_id == 0x02)
+			return convert_type86_xcrb(zdev, reply, xcRB);
+		/* no break, incorrect cprb version is an unknown response */
+	default: /* Unknown response type, this should NEVER EVER happen */
+		PRINTK("Unrecognized Message Header: %08x%08x\n",
+		       *(unsigned int *) reply->message,
+		       *(unsigned int *) (reply->message+4));
+		xcRB->status = 0x0008044DL; /* HDD_InvalidParm */
+		zdev->online = 0;
+		return -EAGAIN;	/* repeat the request on a different device. */
+	}
+}
+
 /**
  * This function is called from the AP bus code after a crypto request
  * "msg" has finished with the reply message "reply".
@@ -407,6 +579,8 @@ static void zcrypt_pcixcc_receive(struct ap_device *ap_dev,
 		.type = TYPE82_RSP_CODE,
 		.reply_code = REP82_ERROR_MACHINE_FAILURE,
 	};
+	struct response_type *resp_type =
+		(struct response_type *) msg->private;
 	struct type86x_reply *t86r = reply->message;
 	int length;
 
@@ -415,12 +589,27 @@ static void zcrypt_pcixcc_receive(struct ap_device *ap_dev,
 		memcpy(msg->message, &error_reply, sizeof(error_reply));
 	else if (t86r->hdr.type == TYPE86_RSP_CODE &&
 		 t86r->cprbx.cprb_ver_id == 0x02) {
-		length = sizeof(struct type86x_reply) + t86r->length - 2;
-		length = min(PCIXCC_MAX_ICA_RESPONSE_SIZE, length);
-		memcpy(msg->message, reply->message, length);
+		switch (resp_type->type) {
+		case PCIXCC_RESPONSE_TYPE_ICA:
+			length = sizeof(struct type86x_reply)
+				+ t86r->length - 2;
+			length = min(PCIXCC_MAX_ICA_RESPONSE_SIZE, length);
+			memcpy(msg->message, reply->message, length);
+			break;
+		case PCIXCC_RESPONSE_TYPE_XCRB:
+			length = t86r->fmt2.offset2 + t86r->fmt2.count2;
+			length = min(PCIXCC_MAX_XCRB_RESPONSE_SIZE, length);
+			memcpy(msg->message, reply->message, length);
+			break;
+		default:
+			PRINTK("Invalid internal response type: %i\n",
+			    resp_type->type);
+			memcpy(msg->message, &error_reply,
+			    sizeof error_reply);
+		}
 	} else
 		memcpy(msg->message, reply->message, sizeof error_reply);
-	complete((struct completion *) msg->private);
+	complete(&(resp_type->work));
 }
 
 static atomic_t zcrypt_step = ATOMIC_INIT(0);
@@ -436,7 +625,9 @@ static long zcrypt_pcixcc_modexpo(struct zcrypt_device *zdev,
 				  struct ica_rsa_modexpo *mex)
 {
 	struct ap_message ap_msg;
-	struct completion work;
+	struct response_type resp_type = {
+		.type = PCIXCC_RESPONSE_TYPE_ICA,
+	};
 	int rc;
 
 	ap_msg.message = (void *) get_zeroed_page(GFP_KERNEL);
@@ -444,14 +635,14 @@ static long zcrypt_pcixcc_modexpo(struct zcrypt_device *zdev,
 		return -ENOMEM;
 	ap_msg.psmid = (((unsigned long long) current->pid) << 32) +
 				atomic_inc_return(&zcrypt_step);
-	ap_msg.private = &work;
+	ap_msg.private = &resp_type;
 	rc = ICAMEX_msg_to_type6MEX_msgX(zdev, &ap_msg, mex);
 	if (rc)
 		goto out_free;
-	init_completion(&work);
+	init_completion(&resp_type.work);
 	ap_queue_message(zdev->ap_dev, &ap_msg);
 	rc = wait_for_completion_interruptible_timeout(
-				&work, PCIXCC_CLEANUP_TIME);
+				&resp_type.work, PCIXCC_CLEANUP_TIME);
 	if (rc > 0)
 		rc = convert_response_ica(zdev, &ap_msg, mex->outputdata,
 					  mex->outputdatalength);
@@ -478,7 +669,9 @@ static long zcrypt_pcixcc_modexpo_crt(struct zcrypt_device *zdev,
 				      struct ica_rsa_modexpo_crt *crt)
 {
 	struct ap_message ap_msg;
-	struct completion work;
+	struct response_type resp_type = {
+		.type = PCIXCC_RESPONSE_TYPE_ICA,
+	};
 	int rc;
 
 	ap_msg.message = (void *) get_zeroed_page(GFP_KERNEL);
@@ -486,14 +679,14 @@ static long zcrypt_pcixcc_modexpo_crt(struct zcrypt_device *zdev,
 		return -ENOMEM;
 	ap_msg.psmid = (((unsigned long long) current->pid) << 32) +
 				atomic_inc_return(&zcrypt_step);
-	ap_msg.private = &work;
+	ap_msg.private = &resp_type;
 	rc = ICACRT_msg_to_type6CRT_msgX(zdev, &ap_msg, crt);
 	if (rc)
 		goto out_free;
-	init_completion(&work);
+	init_completion(&resp_type.work);
 	ap_queue_message(zdev->ap_dev, &ap_msg);
 	rc = wait_for_completion_interruptible_timeout(
-				&work, PCIXCC_CLEANUP_TIME);
+				&resp_type.work, PCIXCC_CLEANUP_TIME);
 	if (rc > 0)
 		rc = convert_response_ica(zdev, &ap_msg, crt->outputdata,
 					  crt->outputdatalength);
@@ -509,12 +702,56 @@ out_free:
 	return rc;
 }
 
+/**
+ * The request distributor calls this function if it picked the PCIXCC/CEX2C
+ * device to handle a send_cprb request.
+ * @zdev: pointer to zcrypt_device structure that identifies the
+ *	  PCIXCC/CEX2C device to the request distributor
+ * @xcRB: pointer to the send_cprb request buffer
+ */
+long zcrypt_pcixcc_send_cprb(struct zcrypt_device *zdev, struct ica_xcRB *xcRB)
+{
+	struct ap_message ap_msg;
+	struct response_type resp_type = {
+		.type = PCIXCC_RESPONSE_TYPE_XCRB,
+	};
+	int rc;
+
+	ap_msg.message = (void *) kmalloc(PCIXCC_MAX_XCRB_MESSAGE_SIZE, GFP_KERNEL);
+	if (!ap_msg.message)
+		return -ENOMEM;
+	ap_msg.psmid = (((unsigned long long) current->pid) << 32) +
+				atomic_inc_return(&zcrypt_step);
+	ap_msg.private = &resp_type;
+	rc = XCRB_msg_to_type6CPRB_msgX(zdev, &ap_msg, xcRB);
+	if (rc)
+		goto out_free;
+	init_completion(&resp_type.work);
+	ap_queue_message(zdev->ap_dev, &ap_msg);
+	rc = wait_for_completion_interruptible_timeout(
+				&resp_type.work, PCIXCC_CLEANUP_TIME);
+	if (rc > 0)
+		rc = convert_response_xcrb(zdev, &ap_msg, xcRB);
+	else {
+		/* Signal pending or message timed out. */
+		ap_cancel_message(zdev->ap_dev, &ap_msg);
+		if (rc == 0)
+			/* Message timed out. */
+			rc = -ETIME;
+	}
+out_free:
+	memset(ap_msg.message, 0x0, ap_msg.length);
+	kfree(ap_msg.message);
+	return rc;
+}
+
 /**
  * The crypto operations for a PCIXCC/CEX2C card.
  */
 static struct zcrypt_ops zcrypt_pcixcc_ops = {
 	.rsa_modexpo = zcrypt_pcixcc_modexpo,
 	.rsa_modexpo_crt = zcrypt_pcixcc_modexpo_crt,
+	.send_cprb = zcrypt_pcixcc_send_cprb,
 };
 
 /**
diff --git a/drivers/s390/crypto/zcrypt_pcixcc.h b/drivers/s390/crypto/zcrypt_pcixcc.h
index d4c44c4d7ad0..a78ff307fd19 100644
--- a/drivers/s390/crypto/zcrypt_pcixcc.h
+++ b/drivers/s390/crypto/zcrypt_pcixcc.h
@@ -1,7 +1,7 @@
 /*
  *  linux/drivers/s390/crypto/zcrypt_pcixcc.h
  *
- *  zcrypt 2.0.0
+ *  zcrypt 2.1.0
  *
  *  Copyright (C)  2001, 2006 IBM Corporation
  *  Author(s): Robert Burroughs
diff --git a/include/asm-s390/zcrypt.h b/include/asm-s390/zcrypt.h
index 0d6a3e2a3349..7244c68464f2 100644
--- a/include/asm-s390/zcrypt.h
+++ b/include/asm-s390/zcrypt.h
@@ -1,7 +1,7 @@
 /*
  *  include/asm-s390/zcrypt.h
  *
- *  zcrypt 2.0.0 (user-visible header)
+ *  zcrypt 2.1.0 (user-visible header)
  *
  *  Copyright (C)  2001, 2006 IBM Corporation
  *  Author(s): Robert Burroughs
@@ -79,6 +79,83 @@ struct ica_rsa_modexpo_crt {
 	char __user *	u_mult_inv;
 };
 
+/**
+ * CPRBX
+ *	  Note that all shorts and ints are big-endian.
+ *	  All pointer fields are 16 bytes long, and mean nothing.
+ *
+ *	  A request CPRB is followed by a request_parameter_block.
+ *
+ *	  The request (or reply) parameter block is organized thus:
+ *	    function code
+ *	    VUD block
+ *	    key block
+ */
+struct ica_CPRBX {
+	unsigned short	cprb_len;	/* CPRB length	      220	 */
+	unsigned char	cprb_ver_id;	/* CPRB version id.   0x02	 */
+	unsigned char	pad_000[3];	/* Alignment pad bytes		 */
+	unsigned char	func_id[2];	/* function id	      0x5432	 */
+	unsigned char	cprb_flags[4];	/* Flags			 */
+	unsigned int	req_parml;	/* request parameter buffer len	 */
+	unsigned int	req_datal;	/* request data buffer		 */
+	unsigned int	rpl_msgbl;	/* reply  message block length	 */
+	unsigned int	rpld_parml;	/* replied parameter block len	 */
+	unsigned int	rpl_datal;	/* reply data block len		 */
+	unsigned int	rpld_datal;	/* replied data block len	 */
+	unsigned int	req_extbl;	/* request extension block len	 */
+	unsigned char	pad_001[4];	/* reserved			 */
+	unsigned int	rpld_extbl;	/* replied extension block len	 */
+	unsigned char	padx000[16 - sizeof (char *)];
+	unsigned char *	req_parmb;	/* request parm block 'address'	 */
+	unsigned char	padx001[16 - sizeof (char *)];
+	unsigned char *	req_datab;	/* request data block 'address'	 */
+	unsigned char	padx002[16 - sizeof (char *)];
+	unsigned char *	rpl_parmb;	/* reply parm block 'address'	 */
+	unsigned char	padx003[16 - sizeof (char *)];
+	unsigned char *	rpl_datab;	/* reply data block 'address'	 */
+	unsigned char	padx004[16 - sizeof (char *)];
+	unsigned char *	req_extb;	/* request extension block 'addr'*/
+	unsigned char	padx005[16 - sizeof (char *)];
+	unsigned char *	rpl_extb;	/* reply extension block 'addres'*/
+	unsigned short	ccp_rtcode;	/* server return code		 */
+	unsigned short	ccp_rscode;	/* server reason code		 */
+	unsigned int	mac_data_len;	/* Mac Data Length		 */
+	unsigned char	logon_id[8];	/* Logon Identifier		 */
+	unsigned char	mac_value[8];	/* Mac Value			 */
+	unsigned char	mac_content_flgs;/* Mac content flag byte	 */
+	unsigned char	pad_002;	/* Alignment			 */
+	unsigned short	domain;		/* Domain			 */
+	unsigned char	usage_domain[4];/* Usage domain			 */
+	unsigned char	cntrl_domain[4];/* Control domain		 */
+	unsigned char	S390enf_mask[4];/* S/390 enforcement mask	 */
+	unsigned char	pad_004[36];	/* reserved			 */
+};
+
+/**
+ * xcRB
+ */
+struct ica_xcRB {
+	unsigned short	agent_ID;
+	unsigned int	user_defined;
+	unsigned short	request_ID;
+	unsigned int	request_control_blk_length;
+	unsigned char	padding1[16 - sizeof (char *)];
+	char __user *	request_control_blk_addr;
+	unsigned int	request_data_length;
+	char		padding2[16 - sizeof (char *)];
+	char __user *	request_data_address;
+	unsigned int	reply_control_blk_length;
+	char		padding3[16 - sizeof (char *)];
+	char __user *	reply_control_blk_addr;
+	unsigned int	reply_data_length;
+	char		padding4[16 - sizeof (char *)];
+	char __user *	reply_data_addr;
+	unsigned short	priority_window;
+	unsigned int	status;
+} __attribute__((packed));
+#define AUTOSELECT ((unsigned int)0xFFFFFFFF)
+
 #define ZCRYPT_IOCTL_MAGIC 'z'
 
 /**
@@ -187,6 +264,7 @@ struct ica_rsa_modexpo_crt {
  */
 #define ICARSAMODEXPO	_IOC(_IOC_READ|_IOC_WRITE, ZCRYPT_IOCTL_MAGIC, 0x05, 0)
 #define ICARSACRT	_IOC(_IOC_READ|_IOC_WRITE, ZCRYPT_IOCTL_MAGIC, 0x06, 0)
+#define ZSECSENDCPRB	_IOC(_IOC_READ|_IOC_WRITE, ZCRYPT_IOCTL_MAGIC, 0x81, 0)
 
 /* New status calls */
 #define Z90STAT_TOTALCOUNT	_IOR(ZCRYPT_IOCTL_MAGIC, 0x40, int)
-- 
cgit v1.2.3


From 4ba069b802c29eee066385f9826e2d83716626b4 Mon Sep 17 00:00:00 2001
From: Michael Grundy <grundym@us.ibm.com>
Date: Wed, 20 Sep 2006 15:58:39 +0200
Subject: [S390] add kprobes support.

Signed-off-by: Michael Grundy <grundym@us.ibm.com>
Signed-off-by: David Wilder <dwilder@us.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/Kconfig              |  14 +
 arch/s390/kernel/Makefile      |   1 +
 arch/s390/kernel/entry.S       |  12 +
 arch/s390/kernel/entry64.S     |  12 +
 arch/s390/kernel/kprobes.c     | 657 +++++++++++++++++++++++++++++++++++++++++
 arch/s390/kernel/traps.c       |  31 +-
 arch/s390/kernel/vmlinux.lds.S |   1 +
 arch/s390/mm/fault.c           |  40 ++-
 include/asm-s390/kdebug.h      |  59 ++++
 include/asm-s390/kprobes.h     | 114 +++++++
 10 files changed, 937 insertions(+), 4 deletions(-)
 create mode 100644 arch/s390/kernel/kprobes.c
 create mode 100644 include/asm-s390/kdebug.h
 create mode 100644 include/asm-s390/kprobes.h

(limited to 'include')

diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig
index 2f4f70c4dbb2..76122ce1e6cb 100644
--- a/arch/s390/Kconfig
+++ b/arch/s390/Kconfig
@@ -487,8 +487,22 @@ source "drivers/net/Kconfig"
 
 source "fs/Kconfig"
 
+menu "Instrumentation Support"
+
 source "arch/s390/oprofile/Kconfig"
 
+config KPROBES
+	bool "Kprobes (EXPERIMENTAL)"
+	depends on EXPERIMENTAL && MODULES
+	help
+	  Kprobes allows you to trap at almost any kernel address and
+	  execute a callback function.	register_kprobe() establishes
+	  a probepoint and specifies the callback.  Kprobes is useful
+	  for kernel debugging, non-intrusive instrumentation and testing.
+	  If in doubt, say "N".
+
+endmenu
+
 source "arch/s390/Kconfig.debug"
 
 source "security/Kconfig"
diff --git a/arch/s390/kernel/Makefile b/arch/s390/kernel/Makefile
index 9a33ed6ca696..33a5069c0e16 100644
--- a/arch/s390/kernel/Makefile
+++ b/arch/s390/kernel/Makefile
@@ -24,6 +24,7 @@ obj-$(CONFIG_COMPAT)		+= compat_linux.o compat_signal.o \
 
 obj-$(CONFIG_VIRT_TIMER)	+= vtime.o
 obj-$(CONFIG_STACKTRACE)	+= stacktrace.o
+obj-$(CONFIG_KPROBES)		+= kprobes.o
 
 # Kexec part
 S390_KEXEC_OBJS := machine_kexec.o crash.o
diff --git a/arch/s390/kernel/entry.S b/arch/s390/kernel/entry.S
index 5b5799ac8f83..0c712b78a7e8 100644
--- a/arch/s390/kernel/entry.S
+++ b/arch/s390/kernel/entry.S
@@ -505,6 +505,8 @@ pgm_no_vtime2:
 	mvc	__THREAD_per+__PER_address(4,%r1),__LC_PER_ADDRESS
 	mvc	__THREAD_per+__PER_access_id(1,%r1),__LC_PER_ACCESS_ID
 	oi	__TI_flags+3(%r9),_TIF_SINGLE_STEP # set TIF_SINGLE_STEP
+	tm	SP_PSW+1(%r15),0x01	# kernel per event ?
+	bz	BASED(kernel_per)
 	l	%r3,__LC_PGM_ILC	 # load program interruption code
 	la	%r8,0x7f
 	nr	%r8,%r3                  # clear per-event-bit and ilc
@@ -536,6 +538,16 @@ pgm_no_vtime3:
 	stosm	__SF_EMPTY(%r15),0x03	# reenable interrupts
 	b	BASED(sysc_do_svc)
 
+#
+# per was called from kernel, must be kprobes
+#
+kernel_per:
+	mvi	SP_TRAP+1(%r15),0x28	# set trap indication to pgm check
+	la	%r2,SP_PTREGS(%r15)	# address of register-save area
+	l	%r1,BASED(.Lhandle_per)	# load adr. of per handler
+	la	%r14,BASED(sysc_leave)	# load adr. of system return
+	br	%r1			# branch to do_single_step
+
 /*
  * IO interrupt handler routine
  */
diff --git a/arch/s390/kernel/entry64.S b/arch/s390/kernel/entry64.S
index 56f5f613b868..8b956d1538f5 100644
--- a/arch/s390/kernel/entry64.S
+++ b/arch/s390/kernel/entry64.S
@@ -518,6 +518,8 @@ pgm_no_vtime2:
 #endif
 	lg	%r9,__LC_THREAD_INFO	# load pointer to thread_info struct
 	lg	%r1,__TI_task(%r9)
+	tm	SP_PSW+1(%r15),0x01	# kernel per event ?
+	jz	kernel_per
 	mvc	__THREAD_per+__PER_atmid(2,%r1),__LC_PER_ATMID
 	mvc	__THREAD_per+__PER_address(8,%r1),__LC_PER_ADDRESS
 	mvc	__THREAD_per+__PER_access_id(1,%r1),__LC_PER_ACCESS_ID
@@ -553,6 +555,16 @@ pgm_no_vtime3:
 	stosm	__SF_EMPTY(%r15),0x03	# reenable interrupts
 	j	sysc_do_svc
 
+#
+# per was called from kernel, must be kprobes
+#
+kernel_per:
+	lhi	%r0,__LC_PGM_OLD_PSW
+	sth	%r0,SP_TRAP(%r15)	# set trap indication to pgm check
+	la	%r2,SP_PTREGS(%r15)	# address of register-save area
+	larl	%r14,sysc_leave		# load adr. of system ret, no work
+	jg	do_single_step		# branch to do_single_step
+
 /*
  * IO interrupt handler routine
  */
diff --git a/arch/s390/kernel/kprobes.c b/arch/s390/kernel/kprobes.c
new file mode 100644
index 000000000000..ca28fb0b3790
--- /dev/null
+++ b/arch/s390/kernel/kprobes.c
@@ -0,0 +1,657 @@
+/*
+ *  Kernel Probes (KProbes)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ * Copyright (C) IBM Corporation, 2002, 2006
+ *
+ * s390 port, used ppc64 as template. Mike Grundy <grundym@us.ibm.com>
+ */
+
+#include <linux/config.h>
+#include <linux/kprobes.h>
+#include <linux/ptrace.h>
+#include <linux/preempt.h>
+#include <linux/stop_machine.h>
+#include <asm/cacheflush.h>
+#include <asm/kdebug.h>
+#include <asm/sections.h>
+#include <asm/uaccess.h>
+#include <linux/module.h>
+
+DEFINE_PER_CPU(struct kprobe *, current_kprobe) = NULL;
+DEFINE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk);
+
+int __kprobes arch_prepare_kprobe(struct kprobe *p)
+{
+	/* Make sure the probe isn't going on a difficult instruction */
+	if (is_prohibited_opcode((kprobe_opcode_t *) p->addr))
+		return -EINVAL;
+
+	if ((unsigned long)p->addr & 0x01) {
+		printk("Attempt to register kprobe at an unaligned address\n");
+		return -EINVAL;
+		}
+
+	/* Use the get_insn_slot() facility for correctness */
+	if (!(p->ainsn.insn = get_insn_slot()))
+		return -ENOMEM;
+
+	memcpy(p->ainsn.insn, p->addr, MAX_INSN_SIZE * sizeof(kprobe_opcode_t));
+
+	get_instruction_type(&p->ainsn);
+	p->opcode = *p->addr;
+	return 0;
+}
+
+int __kprobes is_prohibited_opcode(kprobe_opcode_t *instruction)
+{
+	switch (*(__u8 *) instruction) {
+	case 0x0c:	/* bassm */
+	case 0x0b:	/* bsm	 */
+	case 0x83:	/* diag  */
+	case 0x44:	/* ex	 */
+		return -EINVAL;
+	}
+	switch (*(__u16 *) instruction) {
+	case 0x0101:	/* pr	 */
+	case 0xb25a:	/* bsa	 */
+	case 0xb240:	/* bakr  */
+	case 0xb258:	/* bsg	 */
+	case 0xb218:	/* pc	 */
+	case 0xb228:	/* pt	 */
+		return -EINVAL;
+	}
+	return 0;
+}
+
+void __kprobes get_instruction_type(struct arch_specific_insn *ainsn)
+{
+	/* default fixup method */
+	ainsn->fixup = FIXUP_PSW_NORMAL;
+
+	/* save r1 operand */
+	ainsn->reg = (*ainsn->insn & 0xf0) >> 4;
+
+	/* save the instruction length (pop 5-5) in bytes */
+	switch (*(__u8 *) (ainsn->insn) >> 4) {
+	case 0:
+		ainsn->ilen = 2;
+		break;
+	case 1:
+	case 2:
+		ainsn->ilen = 4;
+		break;
+	case 3:
+		ainsn->ilen = 6;
+		break;
+	}
+
+	switch (*(__u8 *) ainsn->insn) {
+	case 0x05:	/* balr	*/
+	case 0x0d:	/* basr */
+		ainsn->fixup = FIXUP_RETURN_REGISTER;
+		/* if r2 = 0, no branch will be taken */
+		if ((*ainsn->insn & 0x0f) == 0)
+			ainsn->fixup |= FIXUP_BRANCH_NOT_TAKEN;
+		break;
+	case 0x06:	/* bctr	*/
+	case 0x07:	/* bcr	*/
+		ainsn->fixup = FIXUP_BRANCH_NOT_TAKEN;
+		break;
+	case 0x45:	/* bal	*/
+	case 0x4d:	/* bas	*/
+		ainsn->fixup = FIXUP_RETURN_REGISTER;
+		break;
+	case 0x47:	/* bc	*/
+	case 0x46:	/* bct	*/
+	case 0x86:	/* bxh	*/
+	case 0x87:	/* bxle	*/
+		ainsn->fixup = FIXUP_BRANCH_NOT_TAKEN;
+		break;
+	case 0x82:	/* lpsw	*/
+		ainsn->fixup = FIXUP_NOT_REQUIRED;
+		break;
+	case 0xb2:	/* lpswe */
+		if (*(((__u8 *) ainsn->insn) + 1) == 0xb2) {
+			ainsn->fixup = FIXUP_NOT_REQUIRED;
+		}
+		break;
+	case 0xa7:	/* bras	*/
+		if ((*ainsn->insn & 0x0f) == 0x05) {
+			ainsn->fixup |= FIXUP_RETURN_REGISTER;
+		}
+		break;
+	case 0xc0:
+		if ((*ainsn->insn & 0x0f) == 0x00  /* larl  */
+			|| (*ainsn->insn & 0x0f) == 0x05) /* brasl */
+		ainsn->fixup |= FIXUP_RETURN_REGISTER;
+		break;
+	case 0xeb:
+		if (*(((__u8 *) ainsn->insn) + 5 ) == 0x44 ||	/* bxhg  */
+			*(((__u8 *) ainsn->insn) + 5) == 0x45) {/* bxleg */
+			ainsn->fixup = FIXUP_BRANCH_NOT_TAKEN;
+		}
+		break;
+	case 0xe3:	/* bctg	*/
+		if (*(((__u8 *) ainsn->insn) + 5) == 0x46) {
+			ainsn->fixup = FIXUP_BRANCH_NOT_TAKEN;
+		}
+		break;
+	}
+}
+
+static int __kprobes swap_instruction(void *aref)
+{
+	struct ins_replace_args *args = aref;
+	int err = -EFAULT;
+
+	asm volatile(
+		"0: mvc  0(2,%2),0(%3)\n"
+		"1: la   %0,0\n"
+		"2:\n"
+		EX_TABLE(0b,2b)
+		: "+d" (err), "=m" (*args->ptr)
+		: "a" (args->ptr), "a" (&args->new), "m" (args->new));
+	return err;
+}
+
+void __kprobes arch_arm_kprobe(struct kprobe *p)
+{
+	struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+	unsigned long status = kcb->kprobe_status;
+	struct ins_replace_args args;
+
+	args.ptr = p->addr;
+	args.old = p->opcode;
+	args.new = BREAKPOINT_INSTRUCTION;
+
+	kcb->kprobe_status = KPROBE_SWAP_INST;
+	stop_machine_run(swap_instruction, &args, NR_CPUS);
+	kcb->kprobe_status = status;
+}
+
+void __kprobes arch_disarm_kprobe(struct kprobe *p)
+{
+	struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+	unsigned long status = kcb->kprobe_status;
+	struct ins_replace_args args;
+
+	args.ptr = p->addr;
+	args.old = BREAKPOINT_INSTRUCTION;
+	args.new = p->opcode;
+
+	kcb->kprobe_status = KPROBE_SWAP_INST;
+	stop_machine_run(swap_instruction, &args, NR_CPUS);
+	kcb->kprobe_status = status;
+}
+
+void __kprobes arch_remove_kprobe(struct kprobe *p)
+{
+	mutex_lock(&kprobe_mutex);
+	free_insn_slot(p->ainsn.insn);
+	mutex_unlock(&kprobe_mutex);
+}
+
+static void __kprobes prepare_singlestep(struct kprobe *p, struct pt_regs *regs)
+{
+	per_cr_bits kprobe_per_regs[1];
+
+	memset(kprobe_per_regs, 0, sizeof(per_cr_bits));
+	regs->psw.addr = (unsigned long)p->ainsn.insn | PSW_ADDR_AMODE;
+
+	/* Set up the per control reg info, will pass to lctl */
+	kprobe_per_regs[0].em_instruction_fetch = 1;
+	kprobe_per_regs[0].starting_addr = (unsigned long)p->ainsn.insn;
+	kprobe_per_regs[0].ending_addr = (unsigned long)p->ainsn.insn + 1;
+
+	/* Set the PER control regs, turns on single step for this address */
+	__ctl_load(kprobe_per_regs, 9, 11);
+	regs->psw.mask |= PSW_MASK_PER;
+	regs->psw.mask &= ~(PSW_MASK_IO | PSW_MASK_EXT | PSW_MASK_MCHECK);
+}
+
+static void __kprobes save_previous_kprobe(struct kprobe_ctlblk *kcb)
+{
+	kcb->prev_kprobe.kp = kprobe_running();
+	kcb->prev_kprobe.status = kcb->kprobe_status;
+	kcb->prev_kprobe.kprobe_saved_imask = kcb->kprobe_saved_imask;
+	memcpy(kcb->prev_kprobe.kprobe_saved_ctl, kcb->kprobe_saved_ctl,
+					sizeof(kcb->kprobe_saved_ctl));
+}
+
+static void __kprobes restore_previous_kprobe(struct kprobe_ctlblk *kcb)
+{
+	__get_cpu_var(current_kprobe) = kcb->prev_kprobe.kp;
+	kcb->kprobe_status = kcb->prev_kprobe.status;
+	kcb->kprobe_saved_imask = kcb->prev_kprobe.kprobe_saved_imask;
+	memcpy(kcb->kprobe_saved_ctl, kcb->prev_kprobe.kprobe_saved_ctl,
+					sizeof(kcb->kprobe_saved_ctl));
+}
+
+static void __kprobes set_current_kprobe(struct kprobe *p, struct pt_regs *regs,
+						struct kprobe_ctlblk *kcb)
+{
+	__get_cpu_var(current_kprobe) = p;
+	/* Save the interrupt and per flags */
+	kcb->kprobe_saved_imask = regs->psw.mask &
+	    (PSW_MASK_PER | PSW_MASK_IO | PSW_MASK_EXT | PSW_MASK_MCHECK);
+	/* Save the control regs that govern PER */
+	__ctl_store(kcb->kprobe_saved_ctl, 9, 11);
+}
+
+/* Called with kretprobe_lock held */
+void __kprobes arch_prepare_kretprobe(struct kretprobe *rp,
+					struct pt_regs *regs)
+{
+	struct kretprobe_instance *ri;
+
+	if ((ri = get_free_rp_inst(rp)) != NULL) {
+		ri->rp = rp;
+		ri->task = current;
+		ri->ret_addr = (kprobe_opcode_t *) regs->gprs[14];
+
+		/* Replace the return addr with trampoline addr */
+		regs->gprs[14] = (unsigned long)&kretprobe_trampoline;
+
+		add_rp_inst(ri);
+	} else {
+		rp->nmissed++;
+	}
+}
+
+static int __kprobes kprobe_handler(struct pt_regs *regs)
+{
+	struct kprobe *p;
+	int ret = 0;
+	unsigned long *addr = (unsigned long *)
+		((regs->psw.addr & PSW_ADDR_INSN) - 2);
+	struct kprobe_ctlblk *kcb;
+
+	/*
+	 * We don't want to be preempted for the entire
+	 * duration of kprobe processing
+	 */
+	preempt_disable();
+	kcb = get_kprobe_ctlblk();
+
+	/* Check we're not actually recursing */
+	if (kprobe_running()) {
+		p = get_kprobe(addr);
+		if (p) {
+			if (kcb->kprobe_status == KPROBE_HIT_SS &&
+			    *p->ainsn.insn == BREAKPOINT_INSTRUCTION) {
+				regs->psw.mask &= ~PSW_MASK_PER;
+				regs->psw.mask |= kcb->kprobe_saved_imask;
+				goto no_kprobe;
+			}
+			/* We have reentered the kprobe_handler(), since
+			 * another probe was hit while within the handler.
+			 * We here save the original kprobes variables and
+			 * just single step on the instruction of the new probe
+			 * without calling any user handlers.
+			 */
+			save_previous_kprobe(kcb);
+			set_current_kprobe(p, regs, kcb);
+			kprobes_inc_nmissed_count(p);
+			prepare_singlestep(p, regs);
+			kcb->kprobe_status = KPROBE_REENTER;
+			return 1;
+		} else {
+			p = __get_cpu_var(current_kprobe);
+			if (p->break_handler && p->break_handler(p, regs)) {
+				goto ss_probe;
+			}
+		}
+		goto no_kprobe;
+	}
+
+	p = get_kprobe(addr);
+	if (!p) {
+		if (*addr != BREAKPOINT_INSTRUCTION) {
+			/*
+			 * The breakpoint instruction was removed right
+			 * after we hit it.  Another cpu has removed
+			 * either a probepoint or a debugger breakpoint
+			 * at this address.  In either case, no further
+			 * handling of this interrupt is appropriate.
+			 *
+			 */
+			ret = 1;
+		}
+		/* Not one of ours: let kernel handle it */
+		goto no_kprobe;
+	}
+
+	kcb->kprobe_status = KPROBE_HIT_ACTIVE;
+	set_current_kprobe(p, regs, kcb);
+	if (p->pre_handler && p->pre_handler(p, regs))
+		/* handler has already set things up, so skip ss setup */
+		return 1;
+
+ss_probe:
+	prepare_singlestep(p, regs);
+	kcb->kprobe_status = KPROBE_HIT_SS;
+	return 1;
+
+no_kprobe:
+	preempt_enable_no_resched();
+	return ret;
+}
+
+/*
+ * Function return probe trampoline:
+ *	- init_kprobes() establishes a probepoint here
+ *	- When the probed function returns, this probe
+ *		causes the handlers to fire
+ */
+void __kprobes kretprobe_trampoline_holder(void)
+{
+	asm volatile(".global kretprobe_trampoline\n"
+		     "kretprobe_trampoline: bcr 0,0\n");
+}
+
+/*
+ * Called when the probe at kretprobe trampoline is hit
+ */
+int __kprobes trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs)
+{
+	struct kretprobe_instance *ri = NULL;
+	struct hlist_head *head;
+	struct hlist_node *node, *tmp;
+	unsigned long flags, orig_ret_address = 0;
+	unsigned long trampoline_address = (unsigned long)&kretprobe_trampoline;
+
+	spin_lock_irqsave(&kretprobe_lock, flags);
+	head = kretprobe_inst_table_head(current);
+
+	/*
+	 * It is possible to have multiple instances associated with a given
+	 * task either because an multiple functions in the call path
+	 * have a return probe installed on them, and/or more then one return
+	 * return probe was registered for a target function.
+	 *
+	 * We can handle this because:
+	 *     - instances are always inserted at the head of the list
+	 *     - when multiple return probes are registered for the same
+	 *	 function, the first instance's ret_addr will point to the
+	 *	 real return address, and all the rest will point to
+	 *	 kretprobe_trampoline
+	 */
+	hlist_for_each_entry_safe(ri, node, tmp, head, hlist) {
+		if (ri->task != current)
+			/* another task is sharing our hash bucket */
+			continue;
+
+		if (ri->rp && ri->rp->handler)
+			ri->rp->handler(ri, regs);
+
+		orig_ret_address = (unsigned long)ri->ret_addr;
+		recycle_rp_inst(ri);
+
+		if (orig_ret_address != trampoline_address) {
+			/*
+			 * This is the real return address. Any other
+			 * instances associated with this task are for
+			 * other calls deeper on the call stack
+			 */
+			break;
+		}
+	}
+	BUG_ON(!orig_ret_address || (orig_ret_address == trampoline_address));
+	regs->psw.addr = orig_ret_address | PSW_ADDR_AMODE;
+
+	reset_current_kprobe();
+	spin_unlock_irqrestore(&kretprobe_lock, flags);
+	preempt_enable_no_resched();
+
+	/*
+	 * By returning a non-zero value, we are telling
+	 * kprobe_handler() that we don't want the post_handler
+	 * to run (and have re-enabled preemption)
+	 */
+	return 1;
+}
+
+/*
+ * Called after single-stepping.  p->addr is the address of the
+ * instruction whose first byte has been replaced by the "breakpoint"
+ * instruction.  To avoid the SMP problems that can occur when we
+ * temporarily put back the original opcode to single-step, we
+ * single-stepped a copy of the instruction.  The address of this
+ * copy is p->ainsn.insn.
+ */
+static void __kprobes resume_execution(struct kprobe *p, struct pt_regs *regs)
+{
+	struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+
+	regs->psw.addr &= PSW_ADDR_INSN;
+
+	if (p->ainsn.fixup & FIXUP_PSW_NORMAL)
+		regs->psw.addr = (unsigned long)p->addr +
+				((unsigned long)regs->psw.addr -
+				 (unsigned long)p->ainsn.insn);
+
+	if (p->ainsn.fixup & FIXUP_BRANCH_NOT_TAKEN)
+		if ((unsigned long)regs->psw.addr -
+		    (unsigned long)p->ainsn.insn == p->ainsn.ilen)
+			regs->psw.addr = (unsigned long)p->addr + p->ainsn.ilen;
+
+	if (p->ainsn.fixup & FIXUP_RETURN_REGISTER)
+		regs->gprs[p->ainsn.reg] = ((unsigned long)p->addr +
+						(regs->gprs[p->ainsn.reg] -
+						(unsigned long)p->ainsn.insn))
+						| PSW_ADDR_AMODE;
+
+	regs->psw.addr |= PSW_ADDR_AMODE;
+	/* turn off PER mode */
+	regs->psw.mask &= ~PSW_MASK_PER;
+	/* Restore the original per control regs */
+	__ctl_load(kcb->kprobe_saved_ctl, 9, 11);
+	regs->psw.mask |= kcb->kprobe_saved_imask;
+}
+
+static int __kprobes post_kprobe_handler(struct pt_regs *regs)
+{
+	struct kprobe *cur = kprobe_running();
+	struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+
+	if (!cur)
+		return 0;
+
+	if ((kcb->kprobe_status != KPROBE_REENTER) && cur->post_handler) {
+		kcb->kprobe_status = KPROBE_HIT_SSDONE;
+		cur->post_handler(cur, regs, 0);
+	}
+
+	resume_execution(cur, regs);
+
+	/*Restore back the original saved kprobes variables and continue. */
+	if (kcb->kprobe_status == KPROBE_REENTER) {
+		restore_previous_kprobe(kcb);
+		goto out;
+	}
+	reset_current_kprobe();
+out:
+	preempt_enable_no_resched();
+
+	/*
+	 * if somebody else is singlestepping across a probe point, psw mask
+	 * will have PER set, in which case, continue the remaining processing
+	 * of do_single_step, as if this is not a probe hit.
+	 */
+	if (regs->psw.mask & PSW_MASK_PER) {
+		return 0;
+	}
+
+	return 1;
+}
+
+static int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr)
+{
+	struct kprobe *cur = kprobe_running();
+	struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+	const struct exception_table_entry *entry;
+
+	switch(kcb->kprobe_status) {
+	case KPROBE_SWAP_INST:
+		/* We are here because the instruction replacement failed */
+		return 0;
+	case KPROBE_HIT_SS:
+	case KPROBE_REENTER:
+		/*
+		 * We are here because the instruction being single
+		 * stepped caused a page fault. We reset the current
+		 * kprobe and the nip points back to the probe address
+		 * and allow the page fault handler to continue as a
+		 * normal page fault.
+		 */
+		regs->psw.addr = (unsigned long)cur->addr | PSW_ADDR_AMODE;
+		regs->psw.mask &= ~PSW_MASK_PER;
+		regs->psw.mask |= kcb->kprobe_saved_imask;
+		if (kcb->kprobe_status == KPROBE_REENTER)
+			restore_previous_kprobe(kcb);
+		else
+			reset_current_kprobe();
+		preempt_enable_no_resched();
+		break;
+	case KPROBE_HIT_ACTIVE:
+	case KPROBE_HIT_SSDONE:
+		/*
+		 * We increment the nmissed count for accounting,
+		 * we can also use npre/npostfault count for accouting
+		 * these specific fault cases.
+		 */
+		kprobes_inc_nmissed_count(cur);
+
+		/*
+		 * We come here because instructions in the pre/post
+		 * handler caused the page_fault, this could happen
+		 * if handler tries to access user space by
+		 * copy_from_user(), get_user() etc. Let the
+		 * user-specified handler try to fix it first.
+		 */
+		if (cur->fault_handler && cur->fault_handler(cur, regs, trapnr))
+			return 1;
+
+		/*
+		 * In case the user-specified fault handler returned
+		 * zero, try to fix up.
+		 */
+		entry = search_exception_tables(regs->psw.addr & PSW_ADDR_INSN);
+		if (entry) {
+			regs->psw.addr = entry->fixup | PSW_ADDR_AMODE;
+			return 1;
+		}
+
+		/*
+		 * fixup_exception() could not handle it,
+		 * Let do_page_fault() fix it.
+		 */
+		break;
+	default:
+		break;
+	}
+	return 0;
+}
+
+/*
+ * Wrapper routine to for handling exceptions.
+ */
+int __kprobes kprobe_exceptions_notify(struct notifier_block *self,
+				       unsigned long val, void *data)
+{
+	struct die_args *args = (struct die_args *)data;
+	int ret = NOTIFY_DONE;
+
+	switch (val) {
+	case DIE_BPT:
+		if (kprobe_handler(args->regs))
+			ret = NOTIFY_STOP;
+		break;
+	case DIE_SSTEP:
+		if (post_kprobe_handler(args->regs))
+			ret = NOTIFY_STOP;
+		break;
+	case DIE_TRAP:
+	case DIE_PAGE_FAULT:
+		/* kprobe_running() needs smp_processor_id() */
+		preempt_disable();
+		if (kprobe_running() &&
+		    kprobe_fault_handler(args->regs, args->trapnr))
+			ret = NOTIFY_STOP;
+		preempt_enable();
+		break;
+	default:
+		break;
+	}
+	return ret;
+}
+
+int __kprobes setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs)
+{
+	struct jprobe *jp = container_of(p, struct jprobe, kp);
+	unsigned long addr;
+	struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+
+	memcpy(&kcb->jprobe_saved_regs, regs, sizeof(struct pt_regs));
+
+	/* setup return addr to the jprobe handler routine */
+	regs->psw.addr = (unsigned long)(jp->entry) | PSW_ADDR_AMODE;
+
+	/* r14 is the function return address */
+	kcb->jprobe_saved_r14 = (unsigned long)regs->gprs[14];
+	/* r15 is the stack pointer */
+	kcb->jprobe_saved_r15 = (unsigned long)regs->gprs[15];
+	addr = (unsigned long)kcb->jprobe_saved_r15;
+
+	memcpy(kcb->jprobes_stack, (kprobe_opcode_t *) addr,
+	       MIN_STACK_SIZE(addr));
+	return 1;
+}
+
+void __kprobes jprobe_return(void)
+{
+	asm volatile(".word 0x0002");
+}
+
+void __kprobes jprobe_return_end(void)
+{
+	asm volatile("bcr 0,0");
+}
+
+int __kprobes longjmp_break_handler(struct kprobe *p, struct pt_regs *regs)
+{
+	struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+	unsigned long stack_addr = (unsigned long)(kcb->jprobe_saved_r15);
+
+	/* Put the regs back */
+	memcpy(regs, &kcb->jprobe_saved_regs, sizeof(struct pt_regs));
+	/* put the stack back */
+	memcpy((kprobe_opcode_t *) stack_addr, kcb->jprobes_stack,
+	       MIN_STACK_SIZE(stack_addr));
+	preempt_enable_no_resched();
+	return 1;
+}
+
+static struct kprobe trampoline_p = {
+	.addr = (kprobe_opcode_t *) & kretprobe_trampoline,
+	.pre_handler = trampoline_probe_handler
+};
+
+int __init arch_init_kprobes(void)
+{
+	return register_kprobe(&trampoline_p);
+}
diff --git a/arch/s390/kernel/traps.c b/arch/s390/kernel/traps.c
index bde1d1d59858..c4982c963424 100644
--- a/arch/s390/kernel/traps.c
+++ b/arch/s390/kernel/traps.c
@@ -29,6 +29,7 @@
 #include <linux/module.h>
 #include <linux/kallsyms.h>
 #include <linux/reboot.h>
+#include <linux/kprobes.h>
 
 #include <asm/system.h>
 #include <asm/uaccess.h>
@@ -39,6 +40,7 @@
 #include <asm/s390_ext.h>
 #include <asm/lowcore.h>
 #include <asm/debug.h>
+#include <asm/kdebug.h>
 
 /* Called from entry.S only */
 extern void handle_per_exception(struct pt_regs *regs);
@@ -74,6 +76,20 @@ static int kstack_depth_to_print = 12;
 static int kstack_depth_to_print = 20;
 #endif /* CONFIG_64BIT */
 
+ATOMIC_NOTIFIER_HEAD(s390die_chain);
+
+int register_die_notifier(struct notifier_block *nb)
+{
+	return atomic_notifier_chain_register(&s390die_chain, nb);
+}
+EXPORT_SYMBOL(register_die_notifier);
+
+int unregister_die_notifier(struct notifier_block *nb)
+{
+	return atomic_notifier_chain_unregister(&s390die_chain, nb);
+}
+EXPORT_SYMBOL(unregister_die_notifier);
+
 /*
  * For show_trace we have tree different stack to consider:
  *   - the panic stack which is used if the kernel stack has overflown
@@ -305,8 +321,9 @@ report_user_fault(long interruption_code, struct pt_regs *regs)
 #endif
 }
 
-static void inline do_trap(long interruption_code, int signr, char *str,
-                           struct pt_regs *regs, siginfo_t *info)
+static void __kprobes inline do_trap(long interruption_code, int signr,
+					char *str, struct pt_regs *regs,
+					siginfo_t *info)
 {
 	/*
 	 * We got all needed information from the lowcore and can
@@ -315,6 +332,10 @@ static void inline do_trap(long interruption_code, int signr, char *str,
         if (regs->psw.mask & PSW_MASK_PSTATE)
 		local_irq_enable();
 
+	if (notify_die(DIE_TRAP, str, regs, interruption_code,
+				interruption_code, signr) == NOTIFY_STOP)
+		return;
+
         if (regs->psw.mask & PSW_MASK_PSTATE) {
                 struct task_struct *tsk = current;
 
@@ -336,8 +357,12 @@ static inline void __user *get_check_address(struct pt_regs *regs)
 	return (void __user *)((regs->psw.addr-S390_lowcore.pgm_ilc) & PSW_ADDR_INSN);
 }
 
-void do_single_step(struct pt_regs *regs)
+void __kprobes do_single_step(struct pt_regs *regs)
 {
+	if (notify_die(DIE_SSTEP, "sstep", regs, 0, 0,
+					SIGTRAP) == NOTIFY_STOP){
+		return;
+	}
 	if ((current->ptrace & PT_PTRACED) != 0)
 		force_sig(SIGTRAP, current);
 }
diff --git a/arch/s390/kernel/vmlinux.lds.S b/arch/s390/kernel/vmlinux.lds.S
index ff5f7bb34f75..df0c16ab8e92 100644
--- a/arch/s390/kernel/vmlinux.lds.S
+++ b/arch/s390/kernel/vmlinux.lds.S
@@ -24,6 +24,7 @@ SECTIONS
 	*(.text)
 	SCHED_TEXT
 	LOCK_TEXT
+	KPROBES_TEXT
 	*(.fixup)
 	*(.gnu.warning)
 	} = 0x0700
diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c
index 7cd82575813d..44f0cda7e72e 100644
--- a/arch/s390/mm/fault.c
+++ b/arch/s390/mm/fault.c
@@ -25,10 +25,12 @@
 #include <linux/console.h>
 #include <linux/module.h>
 #include <linux/hardirq.h>
+#include <linux/kprobes.h>
 
 #include <asm/system.h>
 #include <asm/uaccess.h>
 #include <asm/pgtable.h>
+#include <asm/kdebug.h>
 
 #ifndef CONFIG_64BIT
 #define __FAIL_ADDR_MASK 0x7ffff000
@@ -48,6 +50,38 @@ extern int sysctl_userprocess_debug;
 
 extern void die(const char *,struct pt_regs *,long);
 
+#ifdef CONFIG_KPROBES
+ATOMIC_NOTIFIER_HEAD(notify_page_fault_chain);
+int register_page_fault_notifier(struct notifier_block *nb)
+{
+	return atomic_notifier_chain_register(&notify_page_fault_chain, nb);
+}
+
+int unregister_page_fault_notifier(struct notifier_block *nb)
+{
+	return atomic_notifier_chain_unregister(&notify_page_fault_chain, nb);
+}
+
+static inline int notify_page_fault(enum die_val val, const char *str,
+			struct pt_regs *regs, long err, int trap, int sig)
+{
+	struct die_args args = {
+		.regs = regs,
+		.str = str,
+		.err = err,
+		.trapnr = trap,
+		.signr = sig
+	};
+	return atomic_notifier_call_chain(&notify_page_fault_chain, val, &args);
+}
+#else
+static inline int notify_page_fault(enum die_val val, const char *str,
+			struct pt_regs *regs, long err, int trap, int sig)
+{
+	return NOTIFY_DONE;
+}
+#endif
+
 extern spinlock_t timerlist_lock;
 
 /*
@@ -159,7 +193,7 @@ static void do_sigsegv(struct pt_regs *regs, unsigned long error_code,
  *   11       Page translation     ->  Not present       (nullification)
  *   3b       Region third trans.  ->  Not present       (nullification)
  */
-static inline void
+static inline void __kprobes
 do_exception(struct pt_regs *regs, unsigned long error_code, int is_protection)
 {
         struct task_struct *tsk;
@@ -173,6 +207,10 @@ do_exception(struct pt_regs *regs, unsigned long error_code, int is_protection)
         tsk = current;
         mm = tsk->mm;
 	
+	if (notify_page_fault(DIE_PAGE_FAULT, "page fault", regs, error_code, 14,
+					SIGSEGV) == NOTIFY_STOP)
+		return;
+
 	/* 
          * Check for low-address protection.  This needs to be treated
 	 * as a special case because the translation exception code 
diff --git a/include/asm-s390/kdebug.h b/include/asm-s390/kdebug.h
new file mode 100644
index 000000000000..40cc68025e01
--- /dev/null
+++ b/include/asm-s390/kdebug.h
@@ -0,0 +1,59 @@
+#ifndef _S390_KDEBUG_H
+#define _S390_KDEBUG_H
+
+/*
+ * Feb 2006 Ported to s390 <grundym@us.ibm.com>
+ */
+#include <linux/notifier.h>
+
+struct pt_regs;
+
+struct die_args {
+	struct pt_regs *regs;
+	const char *str;
+	long err;
+	int trapnr;
+	int signr;
+};
+
+/* Note - you should never unregister because that can race with NMIs.
+ * If you really want to do it first unregister - then synchronize_sched
+ *  - then free.
+ */
+extern int register_die_notifier(struct notifier_block *);
+extern int unregister_die_notifier(struct notifier_block *);
+extern int register_page_fault_notifier(struct notifier_block *);
+extern int unregister_page_fault_notifier(struct notifier_block *);
+extern struct atomic_notifier_head s390die_chain;
+
+
+enum die_val {
+	DIE_OOPS = 1,
+	DIE_BPT,
+	DIE_SSTEP,
+	DIE_PANIC,
+	DIE_NMI,
+	DIE_DIE,
+	DIE_NMIWATCHDOG,
+	DIE_KERNELDEBUG,
+	DIE_TRAP,
+	DIE_GPF,
+	DIE_CALL,
+	DIE_NMI_IPI,
+	DIE_PAGE_FAULT,
+};
+
+static inline int notify_die(enum die_val val, const char *str,
+			struct pt_regs *regs, long err, int trap, int sig)
+{
+	struct die_args args = {
+		.regs = regs,
+		.str = str,
+		.err = err,
+		.trapnr = trap,
+		.signr = sig
+	};
+	return atomic_notifier_call_chain(&s390die_chain, val, &args);
+}
+
+#endif
diff --git a/include/asm-s390/kprobes.h b/include/asm-s390/kprobes.h
new file mode 100644
index 000000000000..b847ff0ec3fa
--- /dev/null
+++ b/include/asm-s390/kprobes.h
@@ -0,0 +1,114 @@
+#ifndef _ASM_S390_KPROBES_H
+#define _ASM_S390_KPROBES_H
+/*
+ *  Kernel Probes (KProbes)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ * Copyright (C) IBM Corporation, 2002, 2006
+ *
+ * 2002-Oct	Created by Vamsi Krishna S <vamsi_krishna@in.ibm.com> Kernel
+ *		Probes initial implementation ( includes suggestions from
+ *		Rusty Russell).
+ * 2004-Nov	Modified for PPC64 by Ananth N Mavinakayanahalli
+ *		<ananth@in.ibm.com>
+ * 2005-Dec	Used as a template for s390 by Mike Grundy
+ *		<grundym@us.ibm.com>
+ */
+#include <linux/types.h>
+#include <linux/ptrace.h>
+#include <linux/percpu.h>
+
+#define  __ARCH_WANT_KPROBES_INSN_SLOT
+struct pt_regs;
+struct kprobe;
+
+typedef u16 kprobe_opcode_t;
+#define BREAKPOINT_INSTRUCTION	0x0002
+
+/* Maximum instruction size is 3 (16bit) halfwords: */
+#define MAX_INSN_SIZE		0x0003
+#define MAX_STACK_SIZE		64
+#define MIN_STACK_SIZE(ADDR) (((MAX_STACK_SIZE) < \
+	(((unsigned long)current_thread_info()) + THREAD_SIZE - (ADDR))) \
+	? (MAX_STACK_SIZE) \
+	: (((unsigned long)current_thread_info()) + THREAD_SIZE - (ADDR)))
+
+#define JPROBE_ENTRY(pentry) (kprobe_opcode_t *)(pentry)
+
+#define ARCH_SUPPORTS_KRETPROBES
+#define ARCH_INACTIVE_KPROBE_COUNT 0
+
+#define KPROBE_SWAP_INST	0x10
+
+#define FIXUP_PSW_NORMAL	0x08
+#define FIXUP_BRANCH_NOT_TAKEN	0x04
+#define FIXUP_RETURN_REGISTER	0x02
+#define FIXUP_NOT_REQUIRED	0x01
+
+/* Architecture specific copy of original instruction */
+struct arch_specific_insn {
+	/* copy of original instruction */
+	kprobe_opcode_t *insn;
+	int fixup;
+	int ilen;
+	int reg;
+};
+
+struct ins_replace_args {
+	kprobe_opcode_t *ptr;
+	kprobe_opcode_t old;
+	kprobe_opcode_t new;
+};
+struct prev_kprobe {
+	struct kprobe *kp;
+	unsigned long status;
+	unsigned long saved_psw;
+	unsigned long kprobe_saved_imask;
+	unsigned long kprobe_saved_ctl[3];
+};
+
+/* per-cpu kprobe control block */
+struct kprobe_ctlblk {
+	unsigned long kprobe_status;
+	unsigned long kprobe_saved_imask;
+	unsigned long kprobe_saved_ctl[3];
+	struct pt_regs jprobe_saved_regs;
+	unsigned long jprobe_saved_r14;
+	unsigned long jprobe_saved_r15;
+	struct prev_kprobe prev_kprobe;
+	kprobe_opcode_t jprobes_stack[MAX_STACK_SIZE];
+};
+
+void arch_remove_kprobe(struct kprobe *p);
+void kretprobe_trampoline(void);
+int  is_prohibited_opcode(kprobe_opcode_t *instruction);
+void get_instruction_type(struct arch_specific_insn *ainsn);
+
+#define flush_insn_slot(p)	do { } while (0)
+
+#endif	/* _ASM_S390_KPROBES_H */
+
+#ifdef CONFIG_KPROBES
+
+extern int kprobe_exceptions_notify(struct notifier_block *self,
+					unsigned long val, void *data);
+#else	/* !CONFIG_KPROBES */
+static inline int kprobe_exceptions_notify(struct notifier_block *self,
+						unsigned long val, void *data)
+{
+	return 0;
+}
+#endif
-- 
cgit v1.2.3


From 65912a84c0f33304fa5ea004c7b6ee58d5f5572e Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Wed, 20 Sep 2006 15:58:41 +0200
Subject: [S390] initrd vs. bootmem bitmap.

Move initrd if the bitmap of the bootmem allocator would overwrite it.
In addition this patch sets the default size and address of the initrd to 0.
Therefore all boot loaders must set the initrd size and address correctly.
This is especially relevant for ftp boot via HMC/SE, where this change
requires a special patch file entry in the .ins file which sets these two
values contained at address 0x10408 and 0x10410.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/head.S   | 10 ++++++----
 arch/s390/kernel/head31.S |  4 ++--
 arch/s390/kernel/head64.S |  4 ++--
 arch/s390/kernel/setup.c  | 41 ++++++++++++++++++++++++++++++++++++++---
 include/asm-s390/setup.h  |  2 --
 5 files changed, 48 insertions(+), 13 deletions(-)

(limited to 'include')

diff --git a/arch/s390/kernel/head.S b/arch/s390/kernel/head.S
index adad8863ee2f..a6e9bdb53591 100644
--- a/arch/s390/kernel/head.S
+++ b/arch/s390/kernel/head.S
@@ -272,7 +272,7 @@ iplstart:
 # load parameter file from ipl device
 #
 .Lagain1:
- 	l     %r2,INITRD_START+ARCH_OFFSET-PARMAREA(%r12) # ramdisk loc. is temp
+	l     %r2,.Linitrd		       # ramdisk loc. is temp
         bas   %r14,.Lloader                    # load parameter file
         ltr   %r2,%r2                          # got anything ?
         bz    .Lnopf
@@ -280,7 +280,7 @@ iplstart:
 	bnh   .Lnotrunc
 	la    %r2,895
 .Lnotrunc:
-	l     %r4,INITRD_START+ARCH_OFFSET-PARMAREA(%r12)
+	l     %r4,.Linitrd
 	clc   0(3,%r4),.L_hdr		       # if it is HDRx
 	bz    .Lagain1			       # skip dataset header
 	clc   0(3,%r4),.L_eof		       # if it is EOFx
@@ -323,14 +323,15 @@ iplstart:
 # load ramdisk from ipl device
 #	
 .Lagain2:
- 	l     %r2,INITRD_START+ARCH_OFFSET-PARMAREA(%r12) # addr of ramdisk
+	l     %r2,.Linitrd		       # addr of ramdisk
+	st    %r2,INITRD_START+ARCH_OFFSET-PARMAREA(%r12)
         bas   %r14,.Lloader                    # load ramdisk
  	st    %r2,INITRD_SIZE+ARCH_OFFSET-PARMAREA(%r12) # store size of ramdisk
         ltr   %r2,%r2
         bnz   .Lrdcont
         st    %r2,INITRD_START+ARCH_OFFSET-PARMAREA(%r12) # no ramdisk found
 .Lrdcont:
-	l     %r2,INITRD_START+ARCH_OFFSET-PARMAREA(%r12)
+	l     %r2,.Linitrd
 
 	clc   0(3,%r2),.L_hdr		       # skip HDRx and EOFx 
 	bz    .Lagain2
@@ -379,6 +380,7 @@ iplstart:
         l     %r1,.Lstartup
         br    %r1
 
+.Linitrd:.long _end + 0x400000		       # default address of initrd
 .Lparm:	.long  PARMAREA
 .Lstartup: .long startup
 .Lcvtab:.long  _ebcasc                         # ebcdic to ascii table
diff --git a/arch/s390/kernel/head31.S b/arch/s390/kernel/head31.S
index a4dc61f3285e..0e46077d7140 100644
--- a/arch/s390/kernel/head31.S
+++ b/arch/s390/kernel/head31.S
@@ -26,8 +26,8 @@ startup:basr	%r13,0			# get base
 #
 	.org	PARMAREA
 	.long	0,0			# IPL_DEVICE
-	.long	0,RAMDISK_ORIGIN	# INITRD_START
-	.long	0,RAMDISK_SIZE		# INITRD_SIZE
+	.long	0,0			# INITRD_START
+	.long	0,0			# INITRD_SIZE
 
 	.org	COMMAND_LINE
 	.byte	"root=/dev/ram0 ro"
diff --git a/arch/s390/kernel/head64.S b/arch/s390/kernel/head64.S
index 9d80c5b1ef95..3e0341acd04e 100644
--- a/arch/s390/kernel/head64.S
+++ b/arch/s390/kernel/head64.S
@@ -26,8 +26,8 @@ startup:basr  %r13,0			 # get base
 #
 	.org   PARMAREA
 	.quad  0			# IPL_DEVICE
-	.quad  RAMDISK_ORIGIN		# INITRD_START
-	.quad  RAMDISK_SIZE		# INITRD_SIZE
+	.quad  0			# INITRD_START
+	.quad  0			# INITRD_SIZE
 
 	.org   COMMAND_LINE
 	.byte  "root=/dev/ram0 ro"
diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c
index c902f059c7aa..89051e8a5d8d 100644
--- a/arch/s390/kernel/setup.c
+++ b/arch/s390/kernel/setup.c
@@ -37,6 +37,7 @@
 #include <linux/kernel_stat.h>
 #include <linux/device.h>
 #include <linux/notifier.h>
+#include <linux/pfn.h>
 
 #include <asm/uaccess.h>
 #include <asm/system.h>
@@ -501,13 +502,47 @@ setup_memory(void)
 	 * partially used pages are not usable - thus
 	 * we are rounding upwards:
 	 */
-	start_pfn = (__pa(&_end) + PAGE_SIZE - 1) >> PAGE_SHIFT;
-	end_pfn = max_pfn = memory_end >> PAGE_SHIFT;
+	start_pfn = PFN_UP(__pa(&_end));
+	end_pfn = max_pfn = PFN_DOWN(memory_end);
 
 	/* Initialize storage key for kernel pages */
 	for (init_pfn = 0 ; init_pfn < start_pfn; init_pfn++)
 		page_set_storage_key(init_pfn << PAGE_SHIFT, PAGE_DEFAULT_KEY);
 
+#ifdef CONFIG_BLK_DEV_INITRD
+	/*
+	 * Move the initrd in case the bitmap of the bootmem allocater
+	 * would overwrite it.
+	 */
+
+	if (INITRD_START && INITRD_SIZE) {
+		unsigned long bmap_size;
+		unsigned long start;
+
+		bmap_size = bootmem_bootmap_pages(end_pfn - start_pfn + 1);
+		bmap_size = PFN_PHYS(bmap_size);
+
+		if (PFN_PHYS(start_pfn) + bmap_size > INITRD_START) {
+			start = PFN_PHYS(start_pfn) + bmap_size + PAGE_SIZE;
+
+			if (start + INITRD_SIZE > memory_end) {
+				printk("initrd extends beyond end of memory "
+				       "(0x%08lx > 0x%08lx)\n"
+				       "disabling initrd\n",
+				       start + INITRD_SIZE, memory_end);
+				INITRD_START = INITRD_SIZE = 0;
+			} else {
+				printk("Moving initrd (0x%08lx -> 0x%08lx, "
+				       "size: %ld)\n",
+				       INITRD_START, start, INITRD_SIZE);
+				memmove((void *) start, (void *) INITRD_START,
+					INITRD_SIZE);
+				INITRD_START = start;
+			}
+		}
+	}
+#endif
+
 	/*
 	 * Initialize the boot-time allocator (with low memory only):
 	 */
@@ -559,7 +594,7 @@ setup_memory(void)
 	reserve_bootmem(start_pfn << PAGE_SHIFT, bootmap_size);
 
 #ifdef CONFIG_BLK_DEV_INITRD
-	if (INITRD_START) {
+	if (INITRD_START && INITRD_SIZE) {
 		if (INITRD_START + INITRD_SIZE <= memory_end) {
 			reserve_bootmem(INITRD_START, INITRD_SIZE);
 			initrd_start = INITRD_START;
diff --git a/include/asm-s390/setup.h b/include/asm-s390/setup.h
index 19e31979309a..02c96d57f0cf 100644
--- a/include/asm-s390/setup.h
+++ b/include/asm-s390/setup.h
@@ -14,8 +14,6 @@
 
 #define PARMAREA		0x10400
 #define COMMAND_LINE_SIZE 	896
-#define RAMDISK_ORIGIN		0x800000
-#define RAMDISK_SIZE		0x800000
 #define MEMORY_CHUNKS		16	/* max 0x7fff */
 #define IPL_PARMBLOCK_ORIGIN	0x2000
 
-- 
cgit v1.2.3


From ff6b8ea68f4b7353f88b97024f28127e2148aa00 Mon Sep 17 00:00:00 2001
From: Michael Holzheu <holzheu@de.ibm.com>
Date: Wed, 20 Sep 2006 15:58:49 +0200
Subject: [S390] ipl/dump on panic.

It is now possible to specify a ccw/fcp dump device which is used to
automatically create a system dump in case of a kernel panic. The dump
device can be configured under /sys/firmware/dump.
In addition it is now possible to specify a ccw/fcp device which is used
for the next reboot of Linux. The reipl device can be configured under
/sys/firmware/reipl.

Signed-off-by: Michael Holzheu <holzheu@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/Makefile     |   2 +-
 arch/s390/kernel/head31.S     |   3 +
 arch/s390/kernel/head64.S     |   3 +
 arch/s390/kernel/ipl.c        | 942 ++++++++++++++++++++++++++++++++++++++++++
 arch/s390/kernel/reipl.S      |  33 +-
 arch/s390/kernel/reipl64.S    |  34 +-
 arch/s390/kernel/reipl_diag.c |  39 --
 arch/s390/kernel/setup.c      | 220 +---------
 arch/s390/kernel/smp.c        |  10 +-
 drivers/s390/cio/cio.c        |  50 ++-
 include/asm-s390/cio.h        |   7 +
 include/asm-s390/lowcore.h    |  13 +-
 include/asm-s390/setup.h      |  54 ++-
 13 files changed, 1099 insertions(+), 311 deletions(-)
 create mode 100644 arch/s390/kernel/ipl.c
 delete mode 100644 arch/s390/kernel/reipl_diag.c

(limited to 'include')

diff --git a/arch/s390/kernel/Makefile b/arch/s390/kernel/Makefile
index 33a5069c0e16..aa978978d3d1 100644
--- a/arch/s390/kernel/Makefile
+++ b/arch/s390/kernel/Makefile
@@ -6,7 +6,7 @@ EXTRA_AFLAGS	:= -traditional
 
 obj-y	:=  bitmap.o traps.o time.o process.o \
             setup.o sys_s390.o ptrace.o signal.o cpcmd.o ebcdic.o \
-            semaphore.o s390_ext.o debug.o profile.o irq.o reipl_diag.o
+	    semaphore.o s390_ext.o debug.o profile.o irq.o ipl.o
 
 obj-y	+= $(if $(CONFIG_64BIT),entry64.o,entry.o)
 obj-y	+= $(if $(CONFIG_64BIT),reipl64.o,reipl.o)
diff --git a/arch/s390/kernel/head31.S b/arch/s390/kernel/head31.S
index 0e46077d7140..d8bb68a72527 100644
--- a/arch/s390/kernel/head31.S
+++ b/arch/s390/kernel/head31.S
@@ -38,6 +38,7 @@ startup:basr	%r13,0			# get base
 startup_continue:
 	basr	%r13,0			# get base
 .LPG1:	GET_IPL_DEVICE
+	mvi	__LC_AR_MODE_ID,0	# set ESA flag (mode 0)
 	lctl	%c0,%c15,.Lctl-.LPG1(%r13) # load control registers
 	l	%r12,.Lparmaddr-.LPG1(%r13) # pointer to parameter area
 					# move IPL device to lowcore
@@ -274,6 +275,8 @@ startup_continue:
 .Lparmaddr: .long PARMAREA
 .Lsccbaddr: .long .Lsccb
 	.org	0x12000
+.globl s390_readinfo_sccb
+s390_readinfo_sccb:
 .Lsccb:
 	.hword	0x1000			# length, one page
 	.byte	0x00,0x00,0x00
diff --git a/arch/s390/kernel/head64.S b/arch/s390/kernel/head64.S
index 3e0341acd04e..c2005101fee1 100644
--- a/arch/s390/kernel/head64.S
+++ b/arch/s390/kernel/head64.S
@@ -41,6 +41,7 @@ startup_continue:
         srl   %r13,1
 	GET_IPL_DEVICE
         lhi   %r1,1                      # mode 1 = esame
+	mvi   __LC_AR_MODE_ID,1		 # set esame flag
         slr   %r0,%r0                    # set cpuid to zero
         sigp  %r1,%r0,0x12               # switch to esame mode
 	sam64				 # switch to 64 bit mode
@@ -269,6 +270,8 @@ startup_continue:
 	.quad	PARMAREA
 
 	.org	0x12000
+.globl s390_readinfo_sccb
+s390_readinfo_sccb:
 .Lsccb:
 	.hword 0x1000			# length, one page
 	.byte 0x00,0x00,0x00
diff --git a/arch/s390/kernel/ipl.c b/arch/s390/kernel/ipl.c
new file mode 100644
index 000000000000..105ee15a2b31
--- /dev/null
+++ b/arch/s390/kernel/ipl.c
@@ -0,0 +1,942 @@
+/*
+ *  arch/s390/kernel/ipl.c
+ *    ipl/reipl/dump support for Linux on s390.
+ *
+ *    Copyright (C) IBM Corp. 2005,2006
+ *    Author(s): Michael Holzheu <holzheu@de.ibm.com>
+ *		 Heiko Carstens <heiko.carstens@de.ibm.com>
+ *		 Volker Sameske <sameske@de.ibm.com>
+ */
+
+#include <linux/types.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/delay.h>
+#include <linux/reboot.h>
+#include <asm/smp.h>
+#include <asm/setup.h>
+#include <asm/cpcmd.h>
+#include <asm/cio.h>
+
+#define IPL_PARM_BLOCK_VERSION 0
+
+enum ipl_type {
+	IPL_TYPE_NONE	 = 1,
+	IPL_TYPE_UNKNOWN = 2,
+	IPL_TYPE_CCW	 = 4,
+	IPL_TYPE_FCP	 = 8,
+};
+
+#define IPL_NONE_STR	 "none"
+#define IPL_UNKNOWN_STR  "unknown"
+#define IPL_CCW_STR	 "ccw"
+#define IPL_FCP_STR	 "fcp"
+
+static char *ipl_type_str(enum ipl_type type)
+{
+	switch (type) {
+	case IPL_TYPE_NONE:
+		return IPL_NONE_STR;
+	case IPL_TYPE_CCW:
+		return IPL_CCW_STR;
+	case IPL_TYPE_FCP:
+		return IPL_FCP_STR;
+	case IPL_TYPE_UNKNOWN:
+	default:
+		return IPL_UNKNOWN_STR;
+	}
+}
+
+enum ipl_method {
+	IPL_METHOD_NONE,
+	IPL_METHOD_CCW_CIO,
+	IPL_METHOD_CCW_DIAG,
+	IPL_METHOD_CCW_VM,
+	IPL_METHOD_FCP_RO_DIAG,
+	IPL_METHOD_FCP_RW_DIAG,
+	IPL_METHOD_FCP_RO_VM,
+};
+
+enum shutdown_action {
+	SHUTDOWN_REIPL,
+	SHUTDOWN_DUMP,
+	SHUTDOWN_STOP,
+};
+
+#define SHUTDOWN_REIPL_STR "reipl"
+#define SHUTDOWN_DUMP_STR  "dump"
+#define SHUTDOWN_STOP_STR  "stop"
+
+static char *shutdown_action_str(enum shutdown_action action)
+{
+	switch (action) {
+	case SHUTDOWN_REIPL:
+		return SHUTDOWN_REIPL_STR;
+	case SHUTDOWN_DUMP:
+		return SHUTDOWN_DUMP_STR;
+	case SHUTDOWN_STOP:
+		return SHUTDOWN_STOP_STR;
+	default:
+		BUG();
+	}
+}
+
+enum diag308_subcode  {
+	DIAG308_IPL   = 3,
+	DIAG308_DUMP  = 4,
+	DIAG308_SET   = 5,
+	DIAG308_STORE = 6,
+};
+
+enum diag308_ipl_type {
+	DIAG308_IPL_TYPE_FCP = 0,
+	DIAG308_IPL_TYPE_CCW = 2,
+};
+
+enum diag308_opt {
+	DIAG308_IPL_OPT_IPL  = 0x10,
+	DIAG308_IPL_OPT_DUMP = 0x20,
+};
+
+enum diag308_rc {
+	DIAG308_RC_OK = 1,
+};
+
+static int diag308_set_works = 0;
+
+static int reipl_capabilities = IPL_TYPE_UNKNOWN;
+static enum ipl_type reipl_type = IPL_TYPE_UNKNOWN;
+static enum ipl_method reipl_method = IPL_METHOD_NONE;
+static struct ipl_parameter_block *reipl_block_fcp;
+static struct ipl_parameter_block *reipl_block_ccw;
+
+static int dump_capabilities = IPL_TYPE_NONE;
+static enum ipl_type dump_type = IPL_TYPE_NONE;
+static enum ipl_method dump_method = IPL_METHOD_NONE;
+static struct ipl_parameter_block *dump_block_fcp;
+static struct ipl_parameter_block *dump_block_ccw;
+
+static enum shutdown_action on_panic_action = SHUTDOWN_STOP;
+
+static int diag308(unsigned long subcode, void *addr)
+{
+	register unsigned long _addr asm("0") = (unsigned long)addr;
+	register unsigned long _rc asm("1") = 0;
+
+	asm volatile (
+		"   diag %0,%2,0x308\n"
+		"0: \n"
+		".section __ex_table,\"a\"\n"
+#ifdef CONFIG_64BIT
+		"   .align 8\n"
+		"   .quad 0b, 0b\n"
+#else
+		"   .align 4\n"
+		"   .long 0b, 0b\n"
+#endif
+		".previous\n"
+		: "+d" (_addr), "+d" (_rc)
+		: "d" (subcode) : "cc", "memory" );
+
+	return _rc;
+}
+
+/* SYSFS */
+
+#define DEFINE_IPL_ATTR_RO(_prefix, _name, _format, _value)		\
+static ssize_t sys_##_prefix##_##_name##_show(struct subsystem *subsys,	\
+		char *page)						\
+{									\
+	return sprintf(page, _format, _value);				\
+}									\
+static struct subsys_attribute sys_##_prefix##_##_name##_attr =		\
+	__ATTR(_name, S_IRUGO, sys_##_prefix##_##_name##_show, NULL);
+
+#define DEFINE_IPL_ATTR_RW(_prefix, _name, _fmt_out, _fmt_in, _value)	\
+static ssize_t sys_##_prefix##_##_name##_show(struct subsystem *subsys,	\
+		char *page)						\
+{									\
+	return sprintf(page, _fmt_out,					\
+			(unsigned long long) _value);			\
+}									\
+static ssize_t sys_##_prefix##_##_name##_store(struct subsystem *subsys,\
+		const char *buf, size_t len)				\
+{									\
+	unsigned long long value;					\
+	if (sscanf(buf, _fmt_in, &value) != 1)				\
+		return -EINVAL;						\
+	_value = value;							\
+	return len;							\
+}									\
+static struct subsys_attribute sys_##_prefix##_##_name##_attr =		\
+	__ATTR(_name,(S_IRUGO | S_IWUSR),				\
+			sys_##_prefix##_##_name##_show,			\
+			sys_##_prefix##_##_name##_store);
+
+static void make_attrs_ro(struct attribute **attrs)
+{
+	while (*attrs) {
+		(*attrs)->mode = S_IRUGO;
+		attrs++;
+	}
+}
+
+/*
+ * ipl section
+ */
+
+static enum ipl_type ipl_get_type(void)
+{
+	struct ipl_parameter_block *ipl = IPL_PARMBLOCK_START;
+
+	if (!IPL_DEVNO_VALID)
+		return IPL_TYPE_UNKNOWN;
+	if (!IPL_PARMBLOCK_VALID)
+		return IPL_TYPE_CCW;
+	if (ipl->hdr.version > IPL_MAX_SUPPORTED_VERSION)
+		return IPL_TYPE_UNKNOWN;
+	if (ipl->hdr.pbt != DIAG308_IPL_TYPE_FCP)
+		return IPL_TYPE_UNKNOWN;
+	return IPL_TYPE_FCP;
+}
+
+static ssize_t ipl_type_show(struct subsystem *subsys, char *page)
+{
+	return sprintf(page, "%s\n", ipl_type_str(ipl_get_type()));
+}
+
+static struct subsys_attribute sys_ipl_type_attr = __ATTR_RO(ipl_type);
+
+static ssize_t sys_ipl_device_show(struct subsystem *subsys, char *page)
+{
+	struct ipl_parameter_block *ipl = IPL_PARMBLOCK_START;
+
+	switch (ipl_get_type()) {
+	case IPL_TYPE_CCW:
+		return sprintf(page, "0.0.%04x\n", ipl_devno);
+	case IPL_TYPE_FCP:
+		return sprintf(page, "0.0.%04x\n", ipl->ipl_info.fcp.devno);
+	default:
+		return 0;
+	}
+}
+
+static struct subsys_attribute sys_ipl_device_attr =
+	__ATTR(device, S_IRUGO, sys_ipl_device_show, NULL);
+
+static ssize_t ipl_parameter_read(struct kobject *kobj, char *buf, loff_t off,
+				  size_t count)
+{
+	unsigned int size = IPL_PARMBLOCK_SIZE;
+
+	if (off > size)
+		return 0;
+	if (off + count > size)
+		count = size - off;
+	memcpy(buf, (void *)IPL_PARMBLOCK_START + off, count);
+	return count;
+}
+
+static struct bin_attribute ipl_parameter_attr = {
+	.attr = {
+		.name = "binary_parameter",
+		.mode = S_IRUGO,
+		.owner = THIS_MODULE,
+	},
+	.size = PAGE_SIZE,
+	.read = &ipl_parameter_read,
+};
+
+static ssize_t ipl_scp_data_read(struct kobject *kobj, char *buf, loff_t off,
+	size_t count)
+{
+	unsigned int size = IPL_PARMBLOCK_START->ipl_info.fcp.scp_data_len;
+	void *scp_data = &IPL_PARMBLOCK_START->ipl_info.fcp.scp_data;
+
+	if (off > size)
+		return 0;
+	if (off + count > size)
+		count = size - off;
+	memcpy(buf, scp_data + off, count);
+	return count;
+}
+
+static struct bin_attribute ipl_scp_data_attr = {
+	.attr = {
+		.name = "scp_data",
+		.mode = S_IRUGO,
+		.owner = THIS_MODULE,
+	},
+	.size = PAGE_SIZE,
+	.read = &ipl_scp_data_read,
+};
+
+/* FCP ipl device attributes */
+
+DEFINE_IPL_ATTR_RO(ipl_fcp, wwpn, "0x%016llx\n", (unsigned long long)
+		   IPL_PARMBLOCK_START->ipl_info.fcp.wwpn);
+DEFINE_IPL_ATTR_RO(ipl_fcp, lun, "0x%016llx\n", (unsigned long long)
+		   IPL_PARMBLOCK_START->ipl_info.fcp.lun);
+DEFINE_IPL_ATTR_RO(ipl_fcp, bootprog, "%lld\n", (unsigned long long)
+		   IPL_PARMBLOCK_START->ipl_info.fcp.bootprog);
+DEFINE_IPL_ATTR_RO(ipl_fcp, br_lba, "%lld\n", (unsigned long long)
+		   IPL_PARMBLOCK_START->ipl_info.fcp.br_lba);
+
+static struct attribute *ipl_fcp_attrs[] = {
+	&sys_ipl_type_attr.attr,
+	&sys_ipl_device_attr.attr,
+	&sys_ipl_fcp_wwpn_attr.attr,
+	&sys_ipl_fcp_lun_attr.attr,
+	&sys_ipl_fcp_bootprog_attr.attr,
+	&sys_ipl_fcp_br_lba_attr.attr,
+	NULL,
+};
+
+static struct attribute_group ipl_fcp_attr_group = {
+	.attrs = ipl_fcp_attrs,
+};
+
+/* CCW ipl device attributes */
+
+static struct attribute *ipl_ccw_attrs[] = {
+	&sys_ipl_type_attr.attr,
+	&sys_ipl_device_attr.attr,
+	NULL,
+};
+
+static struct attribute_group ipl_ccw_attr_group = {
+	.attrs = ipl_ccw_attrs,
+};
+
+/* UNKNOWN ipl device attributes */
+
+static struct attribute *ipl_unknown_attrs[] = {
+	&sys_ipl_type_attr.attr,
+	NULL,
+};
+
+static struct attribute_group ipl_unknown_attr_group = {
+	.attrs = ipl_unknown_attrs,
+};
+
+static decl_subsys(ipl, NULL, NULL);
+
+/*
+ * reipl section
+ */
+
+/* FCP reipl device attributes */
+
+DEFINE_IPL_ATTR_RW(reipl_fcp, wwpn, "0x%016llx\n", "%016llx\n",
+		   reipl_block_fcp->ipl_info.fcp.wwpn);
+DEFINE_IPL_ATTR_RW(reipl_fcp, lun, "0x%016llx\n", "%016llx\n",
+		   reipl_block_fcp->ipl_info.fcp.lun);
+DEFINE_IPL_ATTR_RW(reipl_fcp, bootprog, "%lld\n", "%lld\n",
+		   reipl_block_fcp->ipl_info.fcp.bootprog);
+DEFINE_IPL_ATTR_RW(reipl_fcp, br_lba, "%lld\n", "%lld\n",
+		   reipl_block_fcp->ipl_info.fcp.br_lba);
+DEFINE_IPL_ATTR_RW(reipl_fcp, device, "0.0.%04llx\n", "0.0.%llx\n",
+		   reipl_block_fcp->ipl_info.fcp.devno);
+
+static struct attribute *reipl_fcp_attrs[] = {
+	&sys_reipl_fcp_device_attr.attr,
+	&sys_reipl_fcp_wwpn_attr.attr,
+	&sys_reipl_fcp_lun_attr.attr,
+	&sys_reipl_fcp_bootprog_attr.attr,
+	&sys_reipl_fcp_br_lba_attr.attr,
+	NULL,
+};
+
+static struct attribute_group reipl_fcp_attr_group = {
+	.name  = IPL_FCP_STR,
+	.attrs = reipl_fcp_attrs,
+};
+
+/* CCW reipl device attributes */
+
+DEFINE_IPL_ATTR_RW(reipl_ccw, device, "0.0.%04llx\n", "0.0.%llx\n",
+	reipl_block_ccw->ipl_info.ccw.devno);
+
+static struct attribute *reipl_ccw_attrs[] = {
+	&sys_reipl_ccw_device_attr.attr,
+	NULL,
+};
+
+static struct attribute_group reipl_ccw_attr_group = {
+	.name  = IPL_CCW_STR,
+	.attrs = reipl_ccw_attrs,
+};
+
+/* reipl type */
+
+static int reipl_set_type(enum ipl_type type)
+{
+	if (!(reipl_capabilities & type))
+		return -EINVAL;
+
+	switch(type) {
+	case IPL_TYPE_CCW:
+		if (MACHINE_IS_VM)
+			reipl_method = IPL_METHOD_CCW_VM;
+		else
+			reipl_method = IPL_METHOD_CCW_CIO;
+		break;
+	case IPL_TYPE_FCP:
+		if (diag308_set_works)
+			reipl_method = IPL_METHOD_FCP_RW_DIAG;
+		else if (MACHINE_IS_VM)
+			reipl_method = IPL_METHOD_FCP_RO_VM;
+		else
+			reipl_method = IPL_METHOD_FCP_RO_DIAG;
+		break;
+	default:
+		reipl_method = IPL_METHOD_NONE;
+	}
+	reipl_type = type;
+	return 0;
+}
+
+static ssize_t reipl_type_show(struct subsystem *subsys, char *page)
+{
+	return sprintf(page, "%s\n", ipl_type_str(reipl_type));
+}
+
+static ssize_t reipl_type_store(struct subsystem *subsys, const char *buf,
+				size_t len)
+{
+	int rc = -EINVAL;
+
+	if (strncmp(buf, IPL_CCW_STR, strlen(IPL_CCW_STR)) == 0)
+		rc = reipl_set_type(IPL_TYPE_CCW);
+	else if (strncmp(buf, IPL_FCP_STR, strlen(IPL_FCP_STR)) == 0)
+		rc = reipl_set_type(IPL_TYPE_FCP);
+	return (rc != 0) ? rc : len;
+}
+
+static struct subsys_attribute reipl_type_attr =
+		__ATTR(reipl_type, 0644, reipl_type_show, reipl_type_store);
+
+static decl_subsys(reipl, NULL, NULL);
+
+/*
+ * dump section
+ */
+
+/* FCP dump device attributes */
+
+DEFINE_IPL_ATTR_RW(dump_fcp, wwpn, "0x%016llx\n", "%016llx\n",
+		   dump_block_fcp->ipl_info.fcp.wwpn);
+DEFINE_IPL_ATTR_RW(dump_fcp, lun, "0x%016llx\n", "%016llx\n",
+		   dump_block_fcp->ipl_info.fcp.lun);
+DEFINE_IPL_ATTR_RW(dump_fcp, bootprog, "%lld\n", "%lld\n",
+		   dump_block_fcp->ipl_info.fcp.bootprog);
+DEFINE_IPL_ATTR_RW(dump_fcp, br_lba, "%lld\n", "%lld\n",
+		   dump_block_fcp->ipl_info.fcp.br_lba);
+DEFINE_IPL_ATTR_RW(dump_fcp, device, "0.0.%04llx\n", "0.0.%llx\n",
+		   dump_block_fcp->ipl_info.fcp.devno);
+
+static struct attribute *dump_fcp_attrs[] = {
+	&sys_dump_fcp_device_attr.attr,
+	&sys_dump_fcp_wwpn_attr.attr,
+	&sys_dump_fcp_lun_attr.attr,
+	&sys_dump_fcp_bootprog_attr.attr,
+	&sys_dump_fcp_br_lba_attr.attr,
+	NULL,
+};
+
+static struct attribute_group dump_fcp_attr_group = {
+	.name  = IPL_FCP_STR,
+	.attrs = dump_fcp_attrs,
+};
+
+/* CCW dump device attributes */
+
+DEFINE_IPL_ATTR_RW(dump_ccw, device, "0.0.%04llx\n", "0.0.%llx\n",
+		   dump_block_ccw->ipl_info.ccw.devno);
+
+static struct attribute *dump_ccw_attrs[] = {
+	&sys_dump_ccw_device_attr.attr,
+	NULL,
+};
+
+static struct attribute_group dump_ccw_attr_group = {
+	.name  = IPL_CCW_STR,
+	.attrs = dump_ccw_attrs,
+};
+
+/* dump type */
+
+static int dump_set_type(enum ipl_type type)
+{
+	if (!(dump_capabilities & type))
+		return -EINVAL;
+	switch(type) {
+	case IPL_TYPE_CCW:
+		if (MACHINE_IS_VM)
+			dump_method = IPL_METHOD_CCW_VM;
+		else
+			dump_method = IPL_METHOD_CCW_CIO;
+		break;
+	case IPL_TYPE_FCP:
+		dump_method = IPL_METHOD_FCP_RW_DIAG;
+		break;
+	default:
+		dump_method = IPL_METHOD_NONE;
+	}
+	dump_type = type;
+	return 0;
+}
+
+static ssize_t dump_type_show(struct subsystem *subsys, char *page)
+{
+	return sprintf(page, "%s\n", ipl_type_str(dump_type));
+}
+
+static ssize_t dump_type_store(struct subsystem *subsys, const char *buf,
+			       size_t len)
+{
+	int rc = -EINVAL;
+
+	if (strncmp(buf, IPL_NONE_STR, strlen(IPL_NONE_STR)) == 0)
+		rc = dump_set_type(IPL_TYPE_NONE);
+	else if (strncmp(buf, IPL_CCW_STR, strlen(IPL_CCW_STR)) == 0)
+		rc = dump_set_type(IPL_TYPE_CCW);
+	else if (strncmp(buf, IPL_FCP_STR, strlen(IPL_FCP_STR)) == 0)
+		rc = dump_set_type(IPL_TYPE_FCP);
+	return (rc != 0) ? rc : len;
+}
+
+static struct subsys_attribute dump_type_attr =
+		__ATTR(dump_type, 0644, dump_type_show, dump_type_store);
+
+static decl_subsys(dump, NULL, NULL);
+
+#ifdef CONFIG_SMP
+static void dump_smp_stop_all(void)
+{
+	int cpu;
+	preempt_disable();
+	for_each_online_cpu(cpu) {
+		if (cpu == smp_processor_id())
+			continue;
+		while (signal_processor(cpu, sigp_stop) == sigp_busy)
+			udelay(10);
+	}
+	preempt_enable();
+}
+#else
+#define dump_smp_stop_all() do { } while (0)
+#endif
+
+/*
+ * Shutdown actions section
+ */
+
+static decl_subsys(shutdown_actions, NULL, NULL);
+
+/* on panic */
+
+static ssize_t on_panic_show(struct subsystem *subsys, char *page)
+{
+	return sprintf(page, "%s\n", shutdown_action_str(on_panic_action));
+}
+
+static ssize_t on_panic_store(struct subsystem *subsys, const char *buf,
+			      size_t len)
+{
+	if (strncmp(buf, SHUTDOWN_REIPL_STR, strlen(SHUTDOWN_REIPL_STR)) == 0)
+		on_panic_action = SHUTDOWN_REIPL;
+	else if (strncmp(buf, SHUTDOWN_DUMP_STR,
+			 strlen(SHUTDOWN_DUMP_STR)) == 0)
+		on_panic_action = SHUTDOWN_DUMP;
+	else if (strncmp(buf, SHUTDOWN_STOP_STR,
+			 strlen(SHUTDOWN_STOP_STR)) == 0)
+		on_panic_action = SHUTDOWN_STOP;
+	else
+		return -EINVAL;
+
+	return len;
+}
+
+static struct subsys_attribute on_panic_attr =
+		__ATTR(on_panic, 0644, on_panic_show, on_panic_store);
+
+static void print_fcp_block(struct ipl_parameter_block *fcp_block)
+{
+	printk(KERN_EMERG "wwpn:      %016llx\n",
+		(unsigned long long)fcp_block->ipl_info.fcp.wwpn);
+	printk(KERN_EMERG "lun:       %016llx\n",
+		(unsigned long long)fcp_block->ipl_info.fcp.lun);
+	printk(KERN_EMERG "bootprog:  %lld\n",
+		(unsigned long long)fcp_block->ipl_info.fcp.bootprog);
+	printk(KERN_EMERG "br_lba:    %lld\n",
+		(unsigned long long)fcp_block->ipl_info.fcp.br_lba);
+	printk(KERN_EMERG "device:    %llx\n",
+		(unsigned long long)fcp_block->ipl_info.fcp.devno);
+	printk(KERN_EMERG "opt:       %x\n", fcp_block->ipl_info.fcp.opt);
+}
+
+void do_reipl(void)
+{
+	struct ccw_dev_id devid;
+	static char buf[100];
+
+	switch (reipl_type) {
+	case IPL_TYPE_CCW:
+		printk(KERN_EMERG "reboot on ccw device: 0.0.%04x\n",
+			reipl_block_ccw->ipl_info.ccw.devno);
+		break;
+	case IPL_TYPE_FCP:
+		printk(KERN_EMERG "reboot on fcp device:\n");
+		print_fcp_block(reipl_block_fcp);
+		break;
+	default:
+		break;
+	}
+
+	switch (reipl_method) {
+	case IPL_METHOD_CCW_CIO:
+		devid.devno = reipl_block_ccw->ipl_info.ccw.devno;
+		devid.ssid  = 0;
+		reipl_ccw_dev(&devid);
+		break;
+	case IPL_METHOD_CCW_VM:
+		sprintf(buf, "IPL %X", reipl_block_ccw->ipl_info.ccw.devno);
+		cpcmd(buf, NULL, 0, NULL);
+		break;
+	case IPL_METHOD_CCW_DIAG:
+		diag308(DIAG308_SET, reipl_block_ccw);
+		diag308(DIAG308_IPL, NULL);
+		break;
+	case IPL_METHOD_FCP_RW_DIAG:
+		diag308(DIAG308_SET, reipl_block_fcp);
+		diag308(DIAG308_IPL, NULL);
+		break;
+	case IPL_METHOD_FCP_RO_DIAG:
+		diag308(DIAG308_IPL, NULL);
+		break;
+	case IPL_METHOD_FCP_RO_VM:
+		cpcmd("IPL", NULL, 0, NULL);
+		break;
+	case IPL_METHOD_NONE:
+	default:
+		if (MACHINE_IS_VM)
+			cpcmd("IPL", NULL, 0, NULL);
+		diag308(DIAG308_IPL, NULL);
+		break;
+	}
+	panic("reipl failed!\n");
+}
+
+static void do_dump(void)
+{
+	struct ccw_dev_id devid;
+	static char buf[100];
+
+	switch (dump_type) {
+	case IPL_TYPE_CCW:
+		printk(KERN_EMERG "Automatic dump on ccw device: 0.0.%04x\n",
+		       dump_block_ccw->ipl_info.ccw.devno);
+		break;
+	case IPL_TYPE_FCP:
+		printk(KERN_EMERG "Automatic dump on fcp device:\n");
+		print_fcp_block(dump_block_fcp);
+		break;
+	default:
+		return;
+	}
+
+	switch (dump_method) {
+	case IPL_METHOD_CCW_CIO:
+		dump_smp_stop_all();
+		devid.devno = dump_block_ccw->ipl_info.ccw.devno;
+		devid.ssid  = 0;
+		reipl_ccw_dev(&devid);
+		break;
+	case IPL_METHOD_CCW_VM:
+		dump_smp_stop_all();
+		sprintf(buf, "STORE STATUS");
+		cpcmd(buf, NULL, 0, NULL);
+		sprintf(buf, "IPL %X", dump_block_ccw->ipl_info.ccw.devno);
+		cpcmd(buf, NULL, 0, NULL);
+		break;
+	case IPL_METHOD_CCW_DIAG:
+		diag308(DIAG308_SET, dump_block_ccw);
+		diag308(DIAG308_DUMP, NULL);
+		break;
+	case IPL_METHOD_FCP_RW_DIAG:
+		diag308(DIAG308_SET, dump_block_fcp);
+		diag308(DIAG308_DUMP, NULL);
+		break;
+	case IPL_METHOD_NONE:
+	default:
+		return;
+	}
+	printk(KERN_EMERG "Dump failed!\n");
+}
+
+/* init functions */
+
+static int __init ipl_register_fcp_files(void)
+{
+	int rc;
+
+	rc = sysfs_create_group(&ipl_subsys.kset.kobj,
+				&ipl_fcp_attr_group);
+	if (rc)
+		goto out;
+	rc = sysfs_create_bin_file(&ipl_subsys.kset.kobj,
+				   &ipl_parameter_attr);
+	if (rc)
+		goto out_ipl_parm;
+	rc = sysfs_create_bin_file(&ipl_subsys.kset.kobj,
+				   &ipl_scp_data_attr);
+	if (!rc)
+		goto out;
+
+	sysfs_remove_bin_file(&ipl_subsys.kset.kobj, &ipl_parameter_attr);
+
+out_ipl_parm:
+	sysfs_remove_group(&ipl_subsys.kset.kobj, &ipl_fcp_attr_group);
+out:
+	return rc;
+}
+
+static int __init ipl_init(void)
+{
+	int rc;
+
+	rc = firmware_register(&ipl_subsys);
+	if (rc)
+		return rc;
+	switch (ipl_get_type()) {
+	case IPL_TYPE_CCW:
+		rc = sysfs_create_group(&ipl_subsys.kset.kobj,
+					&ipl_ccw_attr_group);
+		break;
+	case IPL_TYPE_FCP:
+		rc = ipl_register_fcp_files();
+		break;
+	default:
+		rc = sysfs_create_group(&ipl_subsys.kset.kobj,
+					&ipl_unknown_attr_group);
+		break;
+	}
+	if (rc)
+		firmware_unregister(&ipl_subsys);
+	return rc;
+}
+
+static void __init reipl_probe(void)
+{
+	void *buffer;
+
+	buffer = (void *) get_zeroed_page(GFP_KERNEL);
+	if (!buffer)
+		return;
+	if (diag308(DIAG308_STORE, buffer) == DIAG308_RC_OK)
+		diag308_set_works = 1;
+	free_page((unsigned long)buffer);
+}
+
+static int __init reipl_ccw_init(void)
+{
+	int rc;
+
+	reipl_block_ccw = (void *) get_zeroed_page(GFP_KERNEL);
+	if (!reipl_block_ccw)
+		return -ENOMEM;
+	rc = sysfs_create_group(&reipl_subsys.kset.kobj, &reipl_ccw_attr_group);
+	if (rc) {
+		free_page((unsigned long)reipl_block_ccw);
+		return rc;
+	}
+	reipl_block_ccw->hdr.len = IPL_PARM_BLK_CCW_LEN;
+	reipl_block_ccw->hdr.version = IPL_PARM_BLOCK_VERSION;
+	reipl_block_ccw->hdr.blk0_len = sizeof(reipl_block_ccw->ipl_info.ccw);
+	reipl_block_ccw->hdr.pbt = DIAG308_IPL_TYPE_CCW;
+	if (ipl_get_type() == IPL_TYPE_CCW)
+		reipl_block_ccw->ipl_info.ccw.devno = ipl_devno;
+	reipl_capabilities |= IPL_TYPE_CCW;
+	return 0;
+}
+
+static int __init reipl_fcp_init(void)
+{
+	int rc;
+
+	if ((!diag308_set_works) && (ipl_get_type() != IPL_TYPE_FCP))
+		return 0;
+	if ((!diag308_set_works) && (ipl_get_type() == IPL_TYPE_FCP))
+		make_attrs_ro(reipl_fcp_attrs);
+
+	reipl_block_fcp = (void *) get_zeroed_page(GFP_KERNEL);
+	if (!reipl_block_fcp)
+		return -ENOMEM;
+	rc = sysfs_create_group(&reipl_subsys.kset.kobj, &reipl_fcp_attr_group);
+	if (rc) {
+		free_page((unsigned long)reipl_block_fcp);
+		return rc;
+	}
+	if (ipl_get_type() == IPL_TYPE_FCP) {
+		memcpy(reipl_block_fcp, IPL_PARMBLOCK_START, PAGE_SIZE);
+	} else {
+		reipl_block_fcp->hdr.len = IPL_PARM_BLK_FCP_LEN;
+		reipl_block_fcp->hdr.version = IPL_PARM_BLOCK_VERSION;
+		reipl_block_fcp->hdr.blk0_len =
+			sizeof(reipl_block_fcp->ipl_info.fcp);
+		reipl_block_fcp->hdr.pbt = DIAG308_IPL_TYPE_FCP;
+		reipl_block_fcp->ipl_info.fcp.opt = DIAG308_IPL_OPT_IPL;
+	}
+	reipl_capabilities |= IPL_TYPE_FCP;
+	return 0;
+}
+
+static int __init reipl_init(void)
+{
+	int rc;
+
+	rc = firmware_register(&reipl_subsys);
+	if (rc)
+		return rc;
+	rc = subsys_create_file(&reipl_subsys, &reipl_type_attr);
+	if (rc) {
+		firmware_unregister(&reipl_subsys);
+		return rc;
+	}
+	rc = reipl_ccw_init();
+	if (rc)
+		return rc;
+	rc = reipl_fcp_init();
+	if (rc)
+		return rc;
+	rc = reipl_set_type(ipl_get_type());
+	if (rc)
+		return rc;
+	return 0;
+}
+
+static int __init dump_ccw_init(void)
+{
+	int rc;
+
+	dump_block_ccw = (void *) get_zeroed_page(GFP_KERNEL);
+	if (!dump_block_ccw)
+		return -ENOMEM;
+	rc = sysfs_create_group(&dump_subsys.kset.kobj, &dump_ccw_attr_group);
+	if (rc) {
+		free_page((unsigned long)dump_block_ccw);
+		return rc;
+	}
+	dump_block_ccw->hdr.len = IPL_PARM_BLK_CCW_LEN;
+	dump_block_ccw->hdr.version = IPL_PARM_BLOCK_VERSION;
+	dump_block_ccw->hdr.blk0_len = sizeof(reipl_block_ccw->ipl_info.ccw);
+	dump_block_ccw->hdr.pbt = DIAG308_IPL_TYPE_CCW;
+	dump_capabilities |= IPL_TYPE_CCW;
+	return 0;
+}
+
+extern char s390_readinfo_sccb[];
+
+static int __init dump_fcp_init(void)
+{
+	int rc;
+
+	if(!(s390_readinfo_sccb[91] & 0x2))
+		return 0; /* LDIPL DUMP is not installed */
+	if (!diag308_set_works)
+		return 0;
+	dump_block_fcp = (void *) get_zeroed_page(GFP_KERNEL);
+	if (!dump_block_fcp)
+		return -ENOMEM;
+	rc = sysfs_create_group(&dump_subsys.kset.kobj, &dump_fcp_attr_group);
+	if (rc) {
+		free_page((unsigned long)dump_block_fcp);
+		return rc;
+	}
+	dump_block_fcp->hdr.len = IPL_PARM_BLK_FCP_LEN;
+	dump_block_fcp->hdr.version = IPL_PARM_BLOCK_VERSION;
+	dump_block_fcp->hdr.blk0_len = sizeof(dump_block_fcp->ipl_info.fcp);
+	dump_block_fcp->hdr.pbt = DIAG308_IPL_TYPE_FCP;
+	dump_block_fcp->ipl_info.fcp.opt = DIAG308_IPL_OPT_DUMP;
+	dump_capabilities |= IPL_TYPE_FCP;
+	return 0;
+}
+
+#define SHUTDOWN_ON_PANIC_PRIO 0
+
+static int shutdown_on_panic_notify(struct notifier_block *self,
+				    unsigned long event, void *data)
+{
+	if (on_panic_action == SHUTDOWN_DUMP)
+		do_dump();
+	else if (on_panic_action == SHUTDOWN_REIPL)
+		do_reipl();
+	return NOTIFY_OK;
+}
+
+static struct notifier_block shutdown_on_panic_nb = {
+	.notifier_call = shutdown_on_panic_notify,
+	.priority = SHUTDOWN_ON_PANIC_PRIO
+};
+
+static int __init dump_init(void)
+{
+	int rc;
+
+	rc = firmware_register(&dump_subsys);
+	if (rc)
+		return rc;
+	rc = subsys_create_file(&dump_subsys, &dump_type_attr);
+	if (rc) {
+		firmware_unregister(&dump_subsys);
+		return rc;
+	}
+	rc = dump_ccw_init();
+	if (rc)
+		return rc;
+	rc = dump_fcp_init();
+	if (rc)
+		return rc;
+	dump_set_type(IPL_TYPE_NONE);
+	return 0;
+}
+
+static int __init shutdown_actions_init(void)
+{
+	int rc;
+
+	rc = firmware_register(&shutdown_actions_subsys);
+	if (rc)
+		return rc;
+	rc = subsys_create_file(&shutdown_actions_subsys, &on_panic_attr);
+	if (rc) {
+		firmware_unregister(&shutdown_actions_subsys);
+		return rc;
+	}
+	atomic_notifier_chain_register(&panic_notifier_list,
+				       &shutdown_on_panic_nb);
+	return 0;
+}
+
+static int __init s390_ipl_init(void)
+{
+	int rc;
+
+	reipl_probe();
+	rc = ipl_init();
+	if (rc)
+		return rc;
+	rc = reipl_init();
+	if (rc)
+		return rc;
+	rc = dump_init();
+	if (rc)
+		return rc;
+	rc = shutdown_actions_init();
+	if (rc)
+		return rc;
+	return 0;
+}
+
+__initcall(s390_ipl_init);
diff --git a/arch/s390/kernel/reipl.S b/arch/s390/kernel/reipl.S
index 658e5ac484f9..4562cdbce8eb 100644
--- a/arch/s390/kernel/reipl.S
+++ b/arch/s390/kernel/reipl.S
@@ -8,13 +8,30 @@
 
 #include <asm/lowcore.h>
 
-		.globl	do_reipl
-do_reipl:	basr	%r13,0
+		.globl	do_reipl_asm
+do_reipl_asm:	basr	%r13,0
 .Lpg0:		lpsw	.Lnewpsw-.Lpg0(%r13)
-.Lpg1:		lctl	%c6,%c6,.Lall-.Lpg0(%r13)
-                stctl   %c0,%c0,.Lctlsave-.Lpg0(%r13)
-                ni      .Lctlsave-.Lpg0(%r13),0xef
-                lctl    %c0,%c0,.Lctlsave-.Lpg0(%r13)
+
+		# switch off lowcore protection
+
+.Lpg1:		stctl	%c0,%c0,.Lctlsave1-.Lpg0(%r13)
+		stctl	%c0,%c0,.Lctlsave2-.Lpg0(%r13)
+		ni	.Lctlsave1-.Lpg0(%r13),0xef
+		lctl	%c0,%c0,.Lctlsave1-.Lpg0(%r13)
+
+		# do store status of all registers
+
+		stm	%r0,%r15,__LC_GPREGS_SAVE_AREA
+		stctl	%c0,%c15,__LC_CREGS_SAVE_AREA
+		mvc	__LC_CREGS_SAVE_AREA(4),.Lctlsave2-.Lpg0(%r13)
+		stam	%a0,%a15,__LC_AREGS_SAVE_AREA
+		stpx	__LC_PREFIX_SAVE_AREA
+		stckc	.Lclkcmp-.Lpg0(%r13)
+		mvc	__LC_CLOCK_COMP_SAVE_AREA(8),.Lclkcmp-.Lpg0(%r13)
+		stpt	__LC_CPU_TIMER_SAVE_AREA
+		st	%r13, __LC_PSW_SAVE_AREA+4
+
+		lctl	%c6,%c6,.Lall-.Lpg0(%r13)
                 lr      %r1,%r2
         	mvc     __LC_PGM_NEW_PSW(8),.Lpcnew-.Lpg0(%r13)
                 stsch   .Lschib-.Lpg0(%r13)                                    
@@ -46,9 +63,11 @@ do_reipl:	basr	%r13,0
 .Ldisab:	st      %r14,.Ldispsw+4-.Lpg0(%r13)
 		lpsw	.Ldispsw-.Lpg0(%r13)
                 .align 	8
+.Lclkcmp:	.quad	0x0000000000000000
 .Lall:		.long	0xff000000
 .Lnull:		.long   0x00000000
-.Lctlsave:      .long   0x00000000
+.Lctlsave1:	.long	0x00000000
+.Lctlsave2:	.long	0x00000000
                 .align 	8
 .Lnewpsw:	.long   0x00080000,0x80000000+.Lpg1
 .Lpcnew:  	.long   0x00080000,0x80000000+.Lecs
diff --git a/arch/s390/kernel/reipl64.S b/arch/s390/kernel/reipl64.S
index 4d090d60f3ef..95bd1e234f63 100644
--- a/arch/s390/kernel/reipl64.S
+++ b/arch/s390/kernel/reipl64.S
@@ -8,13 +8,30 @@
  */
 
 #include <asm/lowcore.h>
-		.globl	do_reipl
-do_reipl:	basr	%r13,0
-.Lpg0:		lpswe   .Lnewpsw-.Lpg0(%r13)
+		.globl	do_reipl_asm
+do_reipl_asm:	basr	%r13,0
+
+		# do store status of all registers
+
+.Lpg0:		stg	%r1,.Lregsave-.Lpg0(%r13)
+		lghi	%r1,0x1000
+		stmg	%r0,%r15,__LC_GPREGS_SAVE_AREA-0x1000(%r1)
+		lg	%r0,.Lregsave-.Lpg0(%r13)
+		stg	%r0,__LC_GPREGS_SAVE_AREA-0x1000+8(%r1)
+		stctg	%c0,%c15,__LC_CREGS_SAVE_AREA-0x1000(%r1)
+		stam	%a0,%a15,__LC_AREGS_SAVE_AREA-0x1000(%r1)
+		stpx	__LC_PREFIX_SAVE_AREA-0x1000(%r1)
+		stfpc	__LC_FP_CREG_SAVE_AREA-0x1000(%r1)
+		stckc	.Lclkcmp-.Lpg0(%r13)
+		mvc	__LC_CLOCK_COMP_SAVE_AREA-0x1000(8,%r1),.Lclkcmp-.Lpg0(%r13)
+		stpt	__LC_CPU_TIMER_SAVE_AREA-0x1000(%r1)
+		stg	%r13, __LC_PSW_SAVE_AREA-0x1000+8(%r1)
+
+		lpswe	.Lnewpsw-.Lpg0(%r13)
 .Lpg1:		lctlg	%c6,%c6,.Lall-.Lpg0(%r13)
-                stctg   %c0,%c0,.Lctlsave-.Lpg0(%r13)
-                ni      .Lctlsave+4-.Lpg0(%r13),0xef
-                lctlg   %c0,%c0,.Lctlsave-.Lpg0(%r13)
+		stctg	%c0,%c0,.Lregsave-.Lpg0(%r13)
+		ni	.Lregsave+4-.Lpg0(%r13),0xef
+		lctlg	%c0,%c0,.Lregsave-.Lpg0(%r13)
                 lgr     %r1,%r2
         	mvc     __LC_PGM_NEW_PSW(16),.Lpcnew-.Lpg0(%r13)
                 stsch   .Lschib-.Lpg0(%r13)                                    
@@ -50,8 +67,9 @@ do_reipl:	basr	%r13,0
 		st     %r14,.Ldispsw+12-.Lpg0(%r13)
 		lpswe	.Ldispsw-.Lpg0(%r13)
                 .align 	8
+.Lclkcmp:	.quad	0x0000000000000000
 .Lall:		.quad	0x00000000ff000000
-.Lctlsave:      .quad   0x0000000000000000
+.Lregsave:	.quad	0x0000000000000000
 .Lnull:		.long   0x0000000000000000
                 .align 	16
 /*
@@ -92,5 +110,3 @@ do_reipl:	basr	%r13,0
 		.long	0x00000000,0x00000000
 		.long	0x00000000,0x00000000
 	
-
-	
diff --git a/arch/s390/kernel/reipl_diag.c b/arch/s390/kernel/reipl_diag.c
deleted file mode 100644
index 1f33951ba439..000000000000
--- a/arch/s390/kernel/reipl_diag.c
+++ /dev/null
@@ -1,39 +0,0 @@
-/*
- * This file contains the implementation of the
- * Linux re-IPL support
- *
- * (C) Copyright IBM Corp. 2005
- *
- * Author(s): Volker Sameske (sameske@de.ibm.com)
- *
- */
-
-#include <linux/kernel.h>
-
-static unsigned int reipl_diag_rc1;
-static unsigned int reipl_diag_rc2;
-
-/*
- * re-IPL the system using the last used IPL parameters
- */
-void reipl_diag(void)
-{
-        asm volatile (
-		"   la   %%r4,0\n"
-		"   la   %%r5,0\n"
-                "   diag %%r4,%2,0x308\n"
-                "0:\n"
-		"   st   %%r4,%0\n"
-		"   st   %%r5,%1\n"
-                ".section __ex_table,\"a\"\n"
-#ifdef CONFIG_64BIT
-                "   .align 8\n"
-                "   .quad 0b, 0b\n"
-#else
-                "   .align 4\n"
-                "   .long 0b, 0b\n"
-#endif
-                ".previous\n"
-                : "=m" (reipl_diag_rc1), "=m" (reipl_diag_rc2)
-		: "d" (3) : "cc", "4", "5" );
-}
diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c
index 89051e8a5d8d..f2a9165ca4f8 100644
--- a/arch/s390/kernel/setup.c
+++ b/arch/s390/kernel/setup.c
@@ -285,16 +285,9 @@ void (*_machine_power_off)(void) = machine_power_off_smp;
 /*
  * Reboot, halt and power_off routines for non SMP.
  */
-extern void reipl(unsigned long devno);
-extern void reipl_diag(void);
 static void do_machine_restart_nonsmp(char * __unused)
 {
-	reipl_diag();
-
-	if (MACHINE_IS_VM)
-		cpcmd ("IPL", NULL, 0, NULL);
-	else
-		reipl (0x10000 | S390_lowcore.ipl_device);
+	do_reipl();
 }
 
 static void do_machine_halt_nonsmp(void)
@@ -755,214 +748,3 @@ struct seq_operations cpuinfo_op = {
 	.show	= show_cpuinfo,
 };
 
-#define DEFINE_IPL_ATTR(_name, _format, _value)			\
-static ssize_t ipl_##_name##_show(struct subsystem *subsys,	\
-		char *page)					\
-{								\
-	return sprintf(page, _format, _value);			\
-}								\
-static struct subsys_attribute ipl_##_name##_attr =		\
-	__ATTR(_name, S_IRUGO, ipl_##_name##_show, NULL);
-
-DEFINE_IPL_ATTR(wwpn, "0x%016llx\n", (unsigned long long)
-		IPL_PARMBLOCK_START->fcp.wwpn);
-DEFINE_IPL_ATTR(lun, "0x%016llx\n", (unsigned long long)
-		IPL_PARMBLOCK_START->fcp.lun);
-DEFINE_IPL_ATTR(bootprog, "%lld\n", (unsigned long long)
-		IPL_PARMBLOCK_START->fcp.bootprog);
-DEFINE_IPL_ATTR(br_lba, "%lld\n", (unsigned long long)
-		IPL_PARMBLOCK_START->fcp.br_lba);
-
-enum ipl_type_type {
-	ipl_type_unknown,
-	ipl_type_ccw,
-	ipl_type_fcp,
-};
-
-static enum ipl_type_type
-get_ipl_type(void)
-{
-	struct ipl_parameter_block *ipl = IPL_PARMBLOCK_START;
-
-	if (!IPL_DEVNO_VALID)
-		return ipl_type_unknown;
-	if (!IPL_PARMBLOCK_VALID)
-		return ipl_type_ccw;
-	if (ipl->hdr.header.version > IPL_MAX_SUPPORTED_VERSION)
-		return ipl_type_unknown;
-	if (ipl->fcp.pbt != IPL_TYPE_FCP)
-		return ipl_type_unknown;
-	return ipl_type_fcp;
-}
-
-static ssize_t
-ipl_type_show(struct subsystem *subsys, char *page)
-{
-	switch (get_ipl_type()) {
-	case ipl_type_ccw:
-		return sprintf(page, "ccw\n");
-	case ipl_type_fcp:
-		return sprintf(page, "fcp\n");
-	default:
-		return sprintf(page, "unknown\n");
-	}
-}
-
-static struct subsys_attribute ipl_type_attr = __ATTR_RO(ipl_type);
-
-static ssize_t
-ipl_device_show(struct subsystem *subsys, char *page)
-{
-	struct ipl_parameter_block *ipl = IPL_PARMBLOCK_START;
-
-	switch (get_ipl_type()) {
-	case ipl_type_ccw:
-		return sprintf(page, "0.0.%04x\n", ipl_devno);
-	case ipl_type_fcp:
-		return sprintf(page, "0.0.%04x\n", ipl->fcp.devno);
-	default:
-		return 0;
-	}
-}
-
-static struct subsys_attribute ipl_device_attr =
-	__ATTR(device, S_IRUGO, ipl_device_show, NULL);
-
-static struct attribute *ipl_fcp_attrs[] = {
-	&ipl_type_attr.attr,
-	&ipl_device_attr.attr,
-	&ipl_wwpn_attr.attr,
-	&ipl_lun_attr.attr,
-	&ipl_bootprog_attr.attr,
-	&ipl_br_lba_attr.attr,
-	NULL,
-};
-
-static struct attribute_group ipl_fcp_attr_group = {
-	.attrs = ipl_fcp_attrs,
-};
-
-static struct attribute *ipl_ccw_attrs[] = {
-	&ipl_type_attr.attr,
-	&ipl_device_attr.attr,
-	NULL,
-};
-
-static struct attribute_group ipl_ccw_attr_group = {
-	.attrs = ipl_ccw_attrs,
-};
-
-static struct attribute *ipl_unknown_attrs[] = {
-	&ipl_type_attr.attr,
-	NULL,
-};
-
-static struct attribute_group ipl_unknown_attr_group = {
-	.attrs = ipl_unknown_attrs,
-};
-
-static ssize_t
-ipl_parameter_read(struct kobject *kobj, char *buf, loff_t off, size_t count)
-{
-	unsigned int size = IPL_PARMBLOCK_SIZE;
-
-	if (off > size)
-		return 0;
-	if (off + count > size)
-		count = size - off;
-
-	memcpy(buf, (void *) IPL_PARMBLOCK_START + off, count);
-	return count;
-}
-
-static struct bin_attribute ipl_parameter_attr = {
-	.attr = {
-		.name = "binary_parameter",
-		.mode = S_IRUGO,
-		.owner = THIS_MODULE,
-	},
-	.size = PAGE_SIZE,
-	.read = &ipl_parameter_read,
-};
-
-static ssize_t
-ipl_scp_data_read(struct kobject *kobj, char *buf, loff_t off, size_t count)
-{
-	unsigned int size =  IPL_PARMBLOCK_START->fcp.scp_data_len;
-	void *scp_data = &IPL_PARMBLOCK_START->fcp.scp_data;
-
-	if (off > size)
-		return 0;
-	if (off + count > size)
-		count = size - off;
-
-	memcpy(buf, scp_data + off, count);
-	return count;
-}
-
-static struct bin_attribute ipl_scp_data_attr = {
-	.attr = {
-		.name = "scp_data",
-		.mode = S_IRUGO,
-		.owner = THIS_MODULE,
-	},
-	.size = PAGE_SIZE,
-	.read = &ipl_scp_data_read,
-};
-
-static decl_subsys(ipl, NULL, NULL);
-
-static int ipl_register_fcp_files(void)
-{
-	int rc;
-
-	rc = sysfs_create_group(&ipl_subsys.kset.kobj,
-				&ipl_fcp_attr_group);
-	if (rc)
-		goto out;
-	rc = sysfs_create_bin_file(&ipl_subsys.kset.kobj,
-				   &ipl_parameter_attr);
-	if (rc)
-		goto out_ipl_parm;
-	rc = sysfs_create_bin_file(&ipl_subsys.kset.kobj,
-				   &ipl_scp_data_attr);
-	if (!rc)
-		goto out;
-
-	sysfs_remove_bin_file(&ipl_subsys.kset.kobj, &ipl_parameter_attr);
-
-out_ipl_parm:
-	sysfs_remove_group(&ipl_subsys.kset.kobj, &ipl_fcp_attr_group);
-out:
-	return rc;
-}
-
-static int __init
-ipl_device_sysfs_register(void) {
-	int rc;
-
-	rc = firmware_register(&ipl_subsys);
-	if (rc)
-		goto out;
-
-	switch (get_ipl_type()) {
-	case ipl_type_ccw:
-		rc = sysfs_create_group(&ipl_subsys.kset.kobj,
-					&ipl_ccw_attr_group);
-		break;
-	case ipl_type_fcp:
-		rc = ipl_register_fcp_files();
-		break;
-	default:
-		rc = sysfs_create_group(&ipl_subsys.kset.kobj,
-					&ipl_unknown_attr_group);
-		break;
-	}
-
-	if (rc)
-		firmware_unregister(&ipl_subsys);
-out:
-	return rc;
-}
-
-__initcall(ipl_device_sysfs_register);
diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c
index 8e03219eea76..b2e6f4c8d382 100644
--- a/arch/s390/kernel/smp.c
+++ b/arch/s390/kernel/smp.c
@@ -59,9 +59,6 @@ static struct task_struct *current_set[NR_CPUS];
 extern char vmhalt_cmd[];
 extern char vmpoff_cmd[];
 
-extern void reipl(unsigned long devno);
-extern void reipl_diag(void);
-
 static void smp_ext_bitcall(int, ec_bit_sig);
 static void smp_ext_bitcall_others(ec_bit_sig);
 
@@ -279,12 +276,7 @@ static void do_machine_restart(void * __unused)
 	 * interrupted by an external interrupt and s390irq
 	 * locks are always held disabled).
 	 */
-	reipl_diag();
-
-	if (MACHINE_IS_VM)
-		cpcmd ("IPL", NULL, 0, NULL);
-	else
-		reipl (0x10000 | S390_lowcore.ipl_device);
+	do_reipl();
 }
 
 void machine_restart_smp(char * __unused) 
diff --git a/drivers/s390/cio/cio.c b/drivers/s390/cio/cio.c
index 89320c1ad825..050963f15802 100644
--- a/drivers/s390/cio/cio.c
+++ b/drivers/s390/cio/cio.c
@@ -841,14 +841,26 @@ __clear_subchannel_easy(struct subchannel_id schid)
 	return -EBUSY;
 }
 
-extern void do_reipl(unsigned long devno);
-static int
-__shutdown_subchannel_easy(struct subchannel_id schid, void *data)
+struct sch_match_id {
+	struct subchannel_id schid;
+	struct ccw_dev_id devid;
+	int rc;
+};
+
+static int __shutdown_subchannel_easy_and_match(struct subchannel_id schid,
+	void *data)
 {
 	struct schib schib;
+	struct sch_match_id *match_id = data;
 
 	if (stsch_err(schid, &schib))
 		return -ENXIO;
+	if (match_id && schib.pmcw.dnv &&
+		(schib.pmcw.dev == match_id->devid.devno) &&
+		(schid.ssid == match_id->devid.ssid)) {
+		match_id->schid = schid;
+		match_id->rc = 0;
+	}
 	if (!schib.pmcw.ena)
 		return 0;
 	switch(__disable_subchannel_easy(schid, &schib)) {
@@ -864,18 +876,36 @@ __shutdown_subchannel_easy(struct subchannel_id schid, void *data)
 	return 0;
 }
 
-void
-clear_all_subchannels(void)
+static int clear_all_subchannels_and_match(struct ccw_dev_id *devid,
+	struct subchannel_id *schid)
 {
+	struct sch_match_id match_id;
+
+	match_id.devid = *devid;
+	match_id.rc = -ENODEV;
 	local_irq_disable();
-	for_each_subchannel(__shutdown_subchannel_easy, NULL);
+	for_each_subchannel(__shutdown_subchannel_easy_and_match, &match_id);
+	if (match_id.rc == 0)
+		*schid = match_id.schid;
+	return match_id.rc;
 }
 
+
+void clear_all_subchannels(void)
+{
+	local_irq_disable();
+	for_each_subchannel(__shutdown_subchannel_easy_and_match, NULL);
+}
+
+extern void do_reipl_asm(__u32 schid);
+
 /* Make sure all subchannels are quiet before we re-ipl an lpar. */
-void
-reipl(unsigned long devno)
+void reipl_ccw_dev(struct ccw_dev_id *devid)
 {
-	clear_all_subchannels();
+	struct subchannel_id schid;
+
+	if (clear_all_subchannels_and_match(devid, &schid))
+		panic("IPL Device not found\n");
 	cio_reset_channel_paths();
-	do_reipl(devno);
+	do_reipl_asm(*((__u32*)&schid));
 }
diff --git a/include/asm-s390/cio.h b/include/asm-s390/cio.h
index 28fdd6e2b8ba..da063cd5f0a0 100644
--- a/include/asm-s390/cio.h
+++ b/include/asm-s390/cio.h
@@ -270,6 +270,11 @@ struct diag210 {
 	__u32 vrdccrft : 8;    /* real device feature (output) */
 } __attribute__ ((packed,aligned(4)));
 
+struct ccw_dev_id {
+	u8 ssid;
+	u16 devno;
+};
+
 extern int diag210(struct diag210 *addr);
 
 extern void wait_cons_dev(void);
@@ -280,6 +285,8 @@ extern void cio_reset_channel_paths(void);
 
 extern void css_schedule_reprobe(void);
 
+extern void reipl_ccw_dev(struct ccw_dev_id *id);
+
 #endif
 
 #endif
diff --git a/include/asm-s390/lowcore.h b/include/asm-s390/lowcore.h
index 596c8b172104..2e3d4cca5e21 100644
--- a/include/asm-s390/lowcore.h
+++ b/include/asm-s390/lowcore.h
@@ -47,6 +47,7 @@
 #define __LC_PER_ATMID			0x096
 #define __LC_PER_ADDRESS		0x098
 #define __LC_PER_ACCESS_ID		0x0A1
+#define __LC_AR_MODE_ID			0x0A3
 
 #define __LC_SUBCHANNEL_ID              0x0B8
 #define __LC_SUBCHANNEL_NR              0x0BA
@@ -106,18 +107,28 @@
 #define __LC_INT_CLOCK			0xDE8
 #endif /* __s390x__ */
 
-#define __LC_PANIC_MAGIC                0xE00
 
+#define __LC_PANIC_MAGIC		0xE00
 #ifndef __s390x__
 #define __LC_PFAULT_INTPARM             0x080
 #define __LC_CPU_TIMER_SAVE_AREA        0x0D8
+#define __LC_CLOCK_COMP_SAVE_AREA	0x0E0
+#define __LC_PSW_SAVE_AREA		0x100
+#define __LC_PREFIX_SAVE_AREA		0x108
 #define __LC_AREGS_SAVE_AREA            0x120
+#define __LC_FPREGS_SAVE_AREA		0x160
 #define __LC_GPREGS_SAVE_AREA           0x180
 #define __LC_CREGS_SAVE_AREA            0x1C0
 #else /* __s390x__ */
 #define __LC_PFAULT_INTPARM             0x11B8
+#define __LC_FPREGS_SAVE_AREA		0x1200
 #define __LC_GPREGS_SAVE_AREA           0x1280
+#define __LC_PSW_SAVE_AREA		0x1300
+#define __LC_PREFIX_SAVE_AREA		0x1318
+#define __LC_FP_CREG_SAVE_AREA		0x131C
+#define __LC_TODREG_SAVE_AREA		0x1324
 #define __LC_CPU_TIMER_SAVE_AREA        0x1328
+#define __LC_CLOCK_COMP_SAVE_AREA	0x1331
 #define __LC_AREGS_SAVE_AREA            0x1340
 #define __LC_CREGS_SAVE_AREA            0x1380
 #endif /* __s390x__ */
diff --git a/include/asm-s390/setup.h b/include/asm-s390/setup.h
index 02c96d57f0cf..4a1126d8439a 100644
--- a/include/asm-s390/setup.h
+++ b/include/asm-s390/setup.h
@@ -68,39 +68,59 @@ extern unsigned int console_irq;
 #define SET_CONSOLE_3215	do { console_mode = 2; } while (0)
 #define SET_CONSOLE_3270	do { console_mode = 3; } while (0)
 
-struct ipl_list_header {
-	u32 length;
-	u8  reserved[3];
+
+struct ipl_list_hdr {
+	u32 len;
+	u8  reserved1[3];
 	u8  version;
+	u32 blk0_len;
+	u8  pbt;
+	u8  flags;
+	u16 reserved2;
 } __attribute__((packed));
 
 struct ipl_block_fcp {
-	u32 length;
-	u8  pbt;
-	u8  reserved1[322-1];
+	u8  reserved1[313-1];
+	u8  opt;
+	u8  reserved2[3];
+	u16 reserved3;
 	u16 devno;
-	u8  reserved2[4];
+	u8  reserved4[4];
 	u64 wwpn;
 	u64 lun;
 	u32 bootprog;
-	u8  reserved3[12];
+	u8  reserved5[12];
 	u64 br_lba;
 	u32 scp_data_len;
-	u8  reserved4[260];
+	u8  reserved6[260];
 	u8  scp_data[];
 } __attribute__((packed));
 
+struct ipl_block_ccw {
+	u8  load_param[8];
+	u8  reserved1[84];
+	u8  reserved2[2];
+	u16 devno;
+	u8  vm_flags;
+	u8  reserved3[3];
+	u32 vm_parm_len;
+} __attribute__((packed));
+
 struct ipl_parameter_block {
+	struct ipl_list_hdr hdr;
 	union {
-		u32 length;
-		struct ipl_list_header header;
-	} hdr;
-	struct ipl_block_fcp fcp;
+		struct ipl_block_fcp fcp;
+		struct ipl_block_ccw ccw;
+	} ipl_info;
 } __attribute__((packed));
 
-#define IPL_MAX_SUPPORTED_VERSION (0)
+#define IPL_PARM_BLK_FCP_LEN (sizeof(struct ipl_list_hdr) + \
+			      sizeof(struct ipl_block_fcp))
 
-#define IPL_TYPE_FCP (0)
+#define IPL_PARM_BLK_CCW_LEN (sizeof(struct ipl_list_hdr) + \
+			      sizeof(struct ipl_block_ccw))
+
+#define IPL_MAX_SUPPORTED_VERSION (0)
 
 /*
  * IPL validity flags and parameters as detected in head.S
@@ -108,12 +128,14 @@ struct ipl_parameter_block {
 extern u32 ipl_parameter_flags;
 extern u16 ipl_devno;
 
+void do_reipl(void);
+
 #define IPL_DEVNO_VALID		(ipl_parameter_flags & 1)
 #define IPL_PARMBLOCK_VALID	(ipl_parameter_flags & 2)
 
 #define IPL_PARMBLOCK_START	((struct ipl_parameter_block *) \
 				 IPL_PARMBLOCK_ORIGIN)
-#define IPL_PARMBLOCK_SIZE	(IPL_PARMBLOCK_START->hdr.length)
+#define IPL_PARMBLOCK_SIZE	(IPL_PARMBLOCK_START->hdr.len)
 
 #else /* __ASSEMBLY__ */
 
-- 
cgit v1.2.3


From 39b083fe1c3c7b88939f6fa1b0b96e579f12e96f Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Wed, 20 Sep 2006 15:58:51 +0200
Subject: [S390] empty function defines.

Use do { } while (0) constructs instead of empty defines to avoid
subtle compile bugs.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 drivers/s390/cio/qdio.h      | 16 ++++++++--------
 drivers/s390/scsi/zfcp_def.h |  8 ++++----
 include/asm-s390/dma.h       |  2 +-
 include/asm-s390/io.h        |  2 +-
 include/asm-s390/smp.h       |  2 +-
 5 files changed, 15 insertions(+), 15 deletions(-)

(limited to 'include')

diff --git a/drivers/s390/cio/qdio.h b/drivers/s390/cio/qdio.h
index ceb3ab31ee08..124569362f02 100644
--- a/drivers/s390/cio/qdio.h
+++ b/drivers/s390/cio/qdio.h
@@ -191,49 +191,49 @@ enum qdio_irq_states {
 #if QDIO_VERBOSE_LEVEL>8
 #define QDIO_PRINT_STUPID(x...) printk( KERN_DEBUG QDIO_PRINTK_HEADER x)
 #else
-#define QDIO_PRINT_STUPID(x...)
+#define QDIO_PRINT_STUPID(x...) do { } while (0)
 #endif
 
 #if QDIO_VERBOSE_LEVEL>7
 #define QDIO_PRINT_ALL(x...) printk( QDIO_PRINTK_HEADER x)
 #else
-#define QDIO_PRINT_ALL(x...)
+#define QDIO_PRINT_ALL(x...) do { } while (0)
 #endif
 
 #if QDIO_VERBOSE_LEVEL>6
 #define QDIO_PRINT_INFO(x...) printk( QDIO_PRINTK_HEADER x)
 #else
-#define QDIO_PRINT_INFO(x...)
+#define QDIO_PRINT_INFO(x...) do { } while (0)
 #endif
 
 #if QDIO_VERBOSE_LEVEL>5
 #define QDIO_PRINT_WARN(x...) printk( QDIO_PRINTK_HEADER x)
 #else
-#define QDIO_PRINT_WARN(x...)
+#define QDIO_PRINT_WARN(x...) do { } while (0)
 #endif
 
 #if QDIO_VERBOSE_LEVEL>4
 #define QDIO_PRINT_ERR(x...) printk( QDIO_PRINTK_HEADER x)
 #else
-#define QDIO_PRINT_ERR(x...)
+#define QDIO_PRINT_ERR(x...) do { } while (0)
 #endif
 
 #if QDIO_VERBOSE_LEVEL>3
 #define QDIO_PRINT_CRIT(x...) printk( QDIO_PRINTK_HEADER x)
 #else
-#define QDIO_PRINT_CRIT(x...)
+#define QDIO_PRINT_CRIT(x...) do { } while (0)
 #endif
 
 #if QDIO_VERBOSE_LEVEL>2
 #define QDIO_PRINT_ALERT(x...) printk( QDIO_PRINTK_HEADER x)
 #else
-#define QDIO_PRINT_ALERT(x...)
+#define QDIO_PRINT_ALERT(x...) do { } while (0)
 #endif
 
 #if QDIO_VERBOSE_LEVEL>1
 #define QDIO_PRINT_EMERG(x...) printk( QDIO_PRINTK_HEADER x)
 #else
-#define QDIO_PRINT_EMERG(x...)
+#define QDIO_PRINT_EMERG(x...) do { } while (0)
 #endif
 
 #define HEXDUMP16(importance,header,ptr) \
diff --git a/drivers/s390/scsi/zfcp_def.h b/drivers/s390/scsi/zfcp_def.h
index 94d1b74db356..7c84b3d4bd94 100644
--- a/drivers/s390/scsi/zfcp_def.h
+++ b/drivers/s390/scsi/zfcp_def.h
@@ -543,7 +543,7 @@ do { \
 } while (0)
 	
 #if ZFCP_LOG_LEVEL_LIMIT < ZFCP_LOG_LEVEL_NORMAL
-# define ZFCP_LOG_NORMAL(fmt, args...)
+# define ZFCP_LOG_NORMAL(fmt, args...)	do { } while (0)
 #else
 # define ZFCP_LOG_NORMAL(fmt, args...) \
 do { \
@@ -553,7 +553,7 @@ do { \
 #endif
 
 #if ZFCP_LOG_LEVEL_LIMIT < ZFCP_LOG_LEVEL_INFO
-# define ZFCP_LOG_INFO(fmt, args...)
+# define ZFCP_LOG_INFO(fmt, args...)	do { } while (0)
 #else
 # define ZFCP_LOG_INFO(fmt, args...) \
 do { \
@@ -563,14 +563,14 @@ do { \
 #endif
 
 #if ZFCP_LOG_LEVEL_LIMIT < ZFCP_LOG_LEVEL_DEBUG
-# define ZFCP_LOG_DEBUG(fmt, args...)
+# define ZFCP_LOG_DEBUG(fmt, args...)	do { } while (0)
 #else
 # define ZFCP_LOG_DEBUG(fmt, args...) \
 	ZFCP_LOG(ZFCP_LOG_LEVEL_DEBUG, fmt , ##args)
 #endif
 
 #if ZFCP_LOG_LEVEL_LIMIT < ZFCP_LOG_LEVEL_TRACE
-# define ZFCP_LOG_TRACE(fmt, args...)
+# define ZFCP_LOG_TRACE(fmt, args...)	do { } while (0)
 #else
 # define ZFCP_LOG_TRACE(fmt, args...) \
 	ZFCP_LOG(ZFCP_LOG_LEVEL_TRACE, fmt , ##args)
diff --git a/include/asm-s390/dma.h b/include/asm-s390/dma.h
index 02720c449cd8..7425c6af6cd4 100644
--- a/include/asm-s390/dma.h
+++ b/include/asm-s390/dma.h
@@ -11,6 +11,6 @@
 
 #define MAX_DMA_ADDRESS         0x80000000
 
-#define free_dma(x)
+#define free_dma(x)	do { } while (0)
 
 #endif /* _ASM_DMA_H */
diff --git a/include/asm-s390/io.h b/include/asm-s390/io.h
index d4614b35f423..a6cc27e77007 100644
--- a/include/asm-s390/io.h
+++ b/include/asm-s390/io.h
@@ -116,7 +116,7 @@ extern void iounmap(void *addr);
 #define outb(x,addr) ((void) writeb(x,addr))
 #define outb_p(x,addr) outb(x,addr)
 
-#define mmiowb()
+#define mmiowb()	do { } while (0)
 
 /*
  * Convert a physical pointer to a virtual kernel pointer for /dev/mem
diff --git a/include/asm-s390/smp.h b/include/asm-s390/smp.h
index 657646054c5e..9fb02e9779c9 100644
--- a/include/asm-s390/smp.h
+++ b/include/asm-s390/smp.h
@@ -104,7 +104,7 @@ smp_call_function_on(void (*func) (void *info), void *info,
 #define smp_cpu_not_running(cpu)	1
 #define smp_get_cpu(cpu) ({ 0; })
 #define smp_put_cpu(cpu) ({ 0; })
-#define smp_setup_cpu_possible_map()
+#define smp_setup_cpu_possible_map()	do { } while (0)
 #endif
 
 #endif
-- 
cgit v1.2.3


From ba8ce5c6f0a15f08eae39880a0de296007f4a4e7 Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Wed, 20 Sep 2006 15:58:56 +0200
Subject: [S390] #undef in unistd.h

Avoid using #undef in unistd.h.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 include/asm-s390/unistd.h | 170 ++++++++++++++++------------------------------
 1 file changed, 59 insertions(+), 111 deletions(-)

(limited to 'include')

diff --git a/include/asm-s390/unistd.h b/include/asm-s390/unistd.h
index aa7a243862e1..02b942d85c37 100644
--- a/include/asm-s390/unistd.h
+++ b/include/asm-s390/unistd.h
@@ -25,17 +25,12 @@
 #define __NR_unlink              10
 #define __NR_execve              11
 #define __NR_chdir               12
-#define __NR_time                13
 #define __NR_mknod               14
 #define __NR_chmod               15
-#define __NR_lchown              16
 #define __NR_lseek               19
 #define __NR_getpid              20
 #define __NR_mount               21
 #define __NR_umount              22
-#define __NR_setuid              23
-#define __NR_getuid              24
-#define __NR_stime               25
 #define __NR_ptrace              26
 #define __NR_alarm               27
 #define __NR_pause               29
@@ -51,11 +46,7 @@
 #define __NR_pipe                42
 #define __NR_times               43
 #define __NR_brk                 45
-#define __NR_setgid              46
-#define __NR_getgid              47
 #define __NR_signal              48
-#define __NR_geteuid             49
-#define __NR_getegid             50
 #define __NR_acct                51
 #define __NR_umount2             52
 #define __NR_ioctl               54
@@ -69,18 +60,13 @@
 #define __NR_getpgrp             65
 #define __NR_setsid              66
 #define __NR_sigaction           67
-#define __NR_setreuid            70
-#define __NR_setregid            71
 #define __NR_sigsuspend          72
 #define __NR_sigpending          73
 #define __NR_sethostname         74
 #define __NR_setrlimit           75
-#define __NR_getrlimit           76
 #define __NR_getrusage           77
 #define __NR_gettimeofday        78
 #define __NR_settimeofday        79
-#define __NR_getgroups           80
-#define __NR_setgroups           81
 #define __NR_symlink             83
 #define __NR_readlink            85
 #define __NR_uselib              86
@@ -92,12 +78,10 @@
 #define __NR_truncate            92
 #define __NR_ftruncate           93
 #define __NR_fchmod              94
-#define __NR_fchown              95
 #define __NR_getpriority         96
 #define __NR_setpriority         97
 #define __NR_statfs              99
 #define __NR_fstatfs            100
-#define __NR_ioperm             101
 #define __NR_socketcall         102
 #define __NR_syslog             103
 #define __NR_setitimer          104
@@ -131,11 +115,7 @@
 #define __NR_sysfs              135
 #define __NR_personality        136
 #define __NR_afs_syscall        137 /* Syscall for Andrew File System */
-#define __NR_setfsuid           138
-#define __NR_setfsgid           139
-#define __NR__llseek            140
 #define __NR_getdents           141
-#define __NR__newselect         142
 #define __NR_flock              143
 #define __NR_msync              144
 #define __NR_readv              145
@@ -157,13 +137,9 @@
 #define __NR_sched_rr_get_interval      161
 #define __NR_nanosleep          162
 #define __NR_mremap             163
-#define __NR_setresuid          164
-#define __NR_getresuid          165
 #define __NR_query_module       167
 #define __NR_poll               168
 #define __NR_nfsservctl         169
-#define __NR_setresgid          170
-#define __NR_getresgid          171
 #define __NR_prctl              172
 #define __NR_rt_sigreturn       173
 #define __NR_rt_sigaction       174
@@ -174,7 +150,6 @@
 #define __NR_rt_sigsuspend      179
 #define __NR_pread64            180
 #define __NR_pwrite64           181
-#define __NR_chown              182
 #define __NR_getcwd             183
 #define __NR_capget             184
 #define __NR_capset             185
@@ -183,39 +158,11 @@
 #define __NR_getpmsg		188
 #define __NR_putpmsg		189
 #define __NR_vfork		190
-#define __NR_ugetrlimit		191	/* SuS compliant getrlimit */
-#define __NR_mmap2		192
-#define __NR_truncate64		193
-#define __NR_ftruncate64	194
-#define __NR_stat64		195
-#define __NR_lstat64		196
-#define __NR_fstat64		197
-#define __NR_lchown32		198
-#define __NR_getuid32		199
-#define __NR_getgid32		200
-#define __NR_geteuid32		201
-#define __NR_getegid32		202
-#define __NR_setreuid32		203
-#define __NR_setregid32		204
-#define __NR_getgroups32	205
-#define __NR_setgroups32	206
-#define __NR_fchown32		207
-#define __NR_setresuid32	208
-#define __NR_getresuid32	209
-#define __NR_setresgid32	210
-#define __NR_getresgid32	211
-#define __NR_chown32		212
-#define __NR_setuid32		213
-#define __NR_setgid32		214
-#define __NR_setfsuid32		215
-#define __NR_setfsgid32		216
 #define __NR_pivot_root         217
 #define __NR_mincore            218
 #define __NR_madvise            219
 #define __NR_getdents64		220
-#define __NR_fcntl64		221
 #define __NR_readahead		222
-#define __NR_sendfile64		223
 #define __NR_setxattr		224
 #define __NR_lsetxattr		225
 #define __NR_fsetxattr		226
@@ -256,7 +203,6 @@
 #define __NR_clock_getres	(__NR_timer_create+7)
 #define __NR_clock_nanosleep	(__NR_timer_create+8)
 /* Number 263 is reserved for vserver */
-#define __NR_fadvise64_64	264
 #define __NR_statfs64		265
 #define __NR_fstatfs64		266
 #define __NR_remap_file_pages	267
@@ -285,7 +231,6 @@
 #define __NR_mknodat		290
 #define __NR_fchownat		291
 #define __NR_futimesat		292
-#define __NR_fstatat64		293
 #define __NR_unlinkat		294
 #define __NR_renameat		295
 #define __NR_linkat		296
@@ -310,62 +255,65 @@
  * have a different name although they do the same (e.g. __NR_chown32
  * is __NR_chown on 64 bit).
  */
-#ifdef __s390x__
-#undef  __NR_time
-#undef  __NR_lchown
-#undef  __NR_setuid
-#undef  __NR_getuid
-#undef  __NR_stime
-#undef  __NR_setgid
-#undef  __NR_getgid
-#undef  __NR_geteuid
-#undef  __NR_getegid
-#undef  __NR_setreuid
-#undef  __NR_setregid
-#undef  __NR_getrlimit
-#undef  __NR_getgroups
-#undef  __NR_setgroups
-#undef  __NR_fchown
-#undef  __NR_ioperm
-#undef  __NR_setfsuid
-#undef  __NR_setfsgid
-#undef  __NR__llseek
-#undef  __NR__newselect
-#undef  __NR_setresuid
-#undef  __NR_getresuid
-#undef  __NR_setresgid
-#undef  __NR_getresgid
-#undef  __NR_chown
-#undef  __NR_ugetrlimit
-#undef  __NR_mmap2
-#undef  __NR_truncate64
-#undef  __NR_ftruncate64
-#undef  __NR_stat64
-#undef  __NR_lstat64
-#undef  __NR_fstat64
-#undef  __NR_lchown32
-#undef  __NR_getuid32
-#undef  __NR_getgid32
-#undef  __NR_geteuid32
-#undef  __NR_getegid32
-#undef  __NR_setreuid32
-#undef  __NR_setregid32
-#undef  __NR_getgroups32
-#undef  __NR_setgroups32
-#undef  __NR_fchown32
-#undef  __NR_setresuid32
-#undef  __NR_getresuid32
-#undef  __NR_setresgid32
-#undef  __NR_getresgid32
-#undef  __NR_chown32
-#undef  __NR_setuid32
-#undef  __NR_setgid32
-#undef  __NR_setfsuid32
-#undef  __NR_setfsgid32
-#undef  __NR_fcntl64
-#undef  __NR_sendfile64
-#undef  __NR_fadvise64_64
-#undef  __NR_fstatat64
+#ifndef __s390x__
+
+#define __NR_time		 13
+#define __NR_lchown		 16
+#define __NR_setuid		 23
+#define __NR_getuid		 24
+#define __NR_stime		 25
+#define __NR_setgid		 46
+#define __NR_getgid		 47
+#define __NR_geteuid		 49
+#define __NR_getegid		 50
+#define __NR_setreuid		 70
+#define __NR_setregid		 71
+#define __NR_getrlimit		 76
+#define __NR_getgroups		 80
+#define __NR_setgroups		 81
+#define __NR_fchown		 95
+#define __NR_ioperm		101
+#define __NR_setfsuid		138
+#define __NR_setfsgid		139
+#define __NR__llseek		140
+#define __NR__newselect 	142
+#define __NR_setresuid		164
+#define __NR_getresuid		165
+#define __NR_setresgid		170
+#define __NR_getresgid		171
+#define __NR_chown		182
+#define __NR_ugetrlimit		191	/* SuS compliant getrlimit */
+#define __NR_mmap2		192
+#define __NR_truncate64		193
+#define __NR_ftruncate64	194
+#define __NR_stat64		195
+#define __NR_lstat64		196
+#define __NR_fstat64		197
+#define __NR_lchown32		198
+#define __NR_getuid32		199
+#define __NR_getgid32		200
+#define __NR_geteuid32		201
+#define __NR_getegid32		202
+#define __NR_setreuid32		203
+#define __NR_setregid32		204
+#define __NR_getgroups32	205
+#define __NR_setgroups32	206
+#define __NR_fchown32		207
+#define __NR_setresuid32	208
+#define __NR_getresuid32	209
+#define __NR_setresgid32	210
+#define __NR_getresgid32	211
+#define __NR_chown32		212
+#define __NR_setuid32		213
+#define __NR_setgid32		214
+#define __NR_setfsuid32		215
+#define __NR_setfsgid32		216
+#define __NR_fcntl64		221
+#define __NR_sendfile64		223
+#define __NR_fadvise64_64	264
+#define __NR_fstatat64		293
+
+#else
 
 #define __NR_select		142
 #define __NR_getrlimit		191	/* SuS compliant getrlimit */
-- 
cgit v1.2.3


From e87bfe51b5ca2db99dd680bbb1e8fe3c94b607df Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Wed, 20 Sep 2006 15:59:15 +0200
Subject: [S390] convert some assembler to C.

Convert GET_IPL_DEVICE assembler macro to C function.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/head.S    | 59 ----------------------------------------------
 arch/s390/kernel/head31.S  | 43 +++++++++++++++++++++++----------
 arch/s390/kernel/head64.S  | 39 +++++++++++++++++++-----------
 arch/s390/kernel/ipl.c     |  4 ++--
 drivers/s390/cio/cio.c     | 38 +++++++++++++++++++++++++++--
 include/asm-s390/lowcore.h |  1 +
 include/asm-s390/setup.h   |  8 ++++---
 7 files changed, 99 insertions(+), 93 deletions(-)

(limited to 'include')

diff --git a/arch/s390/kernel/head.S b/arch/s390/kernel/head.S
index a6e9bdb53591..0f1db268a8a9 100644
--- a/arch/s390/kernel/head.S
+++ b/arch/s390/kernel/head.S
@@ -481,65 +481,6 @@ start:
 	.byte 0xf0,0xf1,0xf2,0xf3,0xf4,0xf5,0xf6,0xf7 
 	.byte 0xf8,0xf9,0xfa,0xfb,0xfc,0xfd,0xfe,0xff
 
-.macro GET_IPL_DEVICE
-.Lget_ipl_device:
-	l     %r1,0xb8			# get sid
-	sll   %r1,15			# test if subchannel is enabled
-	srl   %r1,31
-	ltr   %r1,%r1
-	bz    2f-.LPG1(%r13)		# subchannel disabled
-	l     %r1,0xb8
-	la    %r5,.Lipl_schib-.LPG1(%r13)
-	stsch 0(%r5)		        # get schib of subchannel
-	bnz   2f-.LPG1(%r13)		# schib not available
-	tm    5(%r5),0x01		# devno valid?
-	bno   2f-.LPG1(%r13)
-	la    %r6,ipl_parameter_flags-.LPG1(%r13)
-	oi    3(%r6),0x01		# set flag
-	la    %r2,ipl_devno-.LPG1(%r13)
-	mvc   0(2,%r2),6(%r5)		# store devno
-	tm    4(%r5),0x80		# qdio capable device?
-	bno   2f-.LPG1(%r13)
-	oi    3(%r6),0x02		# set flag
-
-	# copy ipl parameters
-
-	lhi   %r0,4096
-	l     %r2,20(%r0)		# get address of parameter list
-	lhi   %r3,IPL_PARMBLOCK_ORIGIN
-	st    %r3,20(%r0)
-	lhi   %r4,1
-	cr    %r2,%r3			# start parameters < destination ?
-	jl    0f
-	lhi   %r1,1			# copy direction is upwards
-	j     1f
-0:	lhi   %r1,-1			# copy direction is downwards
-	ar    %r2,%r0
-	ar    %r3,%r0
-	ar    %r2,%r1
-	ar    %r3,%r1
-1:	mvc   0(1,%r3),0(%r2)		# finally copy ipl parameters
-	ar    %r3,%r1
-	ar    %r2,%r1
-	sr    %r0,%r4
-	jne   1b
-	b     2f-.LPG1(%r13)
-
-	.align 4
-.Lipl_schib:
-	.rept 13
-	.long 0
-	.endr
-
-	.globl ipl_parameter_flags
-ipl_parameter_flags:
-	.long 0
-	.globl ipl_devno
-ipl_devno:
-	.word 0
-2:
-.endm
-
 #ifdef CONFIG_64BIT
 #include "head64.S"
 #else
diff --git a/arch/s390/kernel/head31.S b/arch/s390/kernel/head31.S
index d8bb68a72527..1fa9fa1ca740 100644
--- a/arch/s390/kernel/head31.S
+++ b/arch/s390/kernel/head31.S
@@ -37,13 +37,23 @@ startup:basr	%r13,0			# get base
 
 startup_continue:
 	basr	%r13,0			# get base
-.LPG1:	GET_IPL_DEVICE
-	mvi	__LC_AR_MODE_ID,0	# set ESA flag (mode 0)
+.LPG1:	mvi	__LC_AR_MODE_ID,0	# set ESA flag (mode 0)
 	lctl	%c0,%c15,.Lctl-.LPG1(%r13) # load control registers
 	l	%r12,.Lparmaddr-.LPG1(%r13) # pointer to parameter area
 					# move IPL device to lowcore
 	mvc	__LC_IPLDEV(4),IPL_DEVICE-PARMAREA(%r12)
+#
+# Setup stack
+#
+	l	%r15,.Linittu-.LPG1(%r13)
+	mvc	__LC_CURRENT(4),__TI_task(%r15)
+	ahi	%r15,1<<(PAGE_SHIFT+THREAD_ORDER) # init_task_union+THREAD_SIZE
+	st	%r15,__LC_KERNEL_STACK	# set end of kernel stack
+	ahi	%r15,-96
+	xc	__SF_BACKCHAIN(4,%r15),__SF_BACKCHAIN(%r15) # clear backchain
 
+	l	%r14,.Lipl_save_parameters-.LPG1(%r13)
+	basr	%r14,%r14
 #
 # clear bss memory
 #
@@ -115,6 +125,10 @@ startup_continue:
 	b	.Lfchunk-.LPG1(%r13)
 
 	.align 4
+.Lipl_save_parameters:
+	.long	ipl_save_parameters
+.Linittu:
+	.long	init_thread_union
 .Lpmask:
 	.byte	0
 .align 8
@@ -274,6 +288,20 @@ startup_continue:
 .Lbss_end:  .long _end
 .Lparmaddr: .long PARMAREA
 .Lsccbaddr: .long .Lsccb
+
+	.globl ipl_schib
+ipl_schib:
+	.rept 13
+	.long 0
+	.endr
+
+	.globl ipl_flags
+ipl_flags:
+	.long 0
+	.globl ipl_devno
+ipl_devno:
+	.word 0
+
 	.org	0x12000
 .globl s390_readinfo_sccb
 s390_readinfo_sccb:
@@ -305,16 +333,6 @@ s390_readinfo_sccb:
 	.globl	_stext
 _stext:	basr	%r13,0			# get base
 .LPG3:
-#
-# Setup stack
-#
-	l	%r15,.Linittu-.LPG3(%r13)
-	mvc	__LC_CURRENT(4),__TI_task(%r15)
-	ahi	%r15,1<<(PAGE_SHIFT+THREAD_ORDER) # init_task_union+THREAD_SIZE
-	st	%r15,__LC_KERNEL_STACK	# set end of kernel stack
-	ahi	%r15,-96
-	xc	__SF_BACKCHAIN(4,%r15),__SF_BACKCHAIN(%r15) # clear backchain
-
 # check control registers
 	stctl	%c0,%c15,0(%r15)
 	oi	2(%r15),0x40		# enable sigp emergency signal
@@ -333,6 +351,5 @@ _stext:	basr	%r13,0			# get base
 #
 	.align	8
 .Ldw:	.long	0x000a0000,0x00000000
-.Linittu:.long	init_thread_union
 .Lstart:.long	start_kernel
 .Laregs:.long	0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0
diff --git a/arch/s390/kernel/head64.S b/arch/s390/kernel/head64.S
index c2005101fee1..1ebaa338aa7e 100644
--- a/arch/s390/kernel/head64.S
+++ b/arch/s390/kernel/head64.S
@@ -39,7 +39,6 @@ startup_continue:
 	basr  %r13,0			 # get base
 .LPG1:  sll   %r13,1                     # remove high order bit
         srl   %r13,1
-	GET_IPL_DEVICE
         lhi   %r1,1                      # mode 1 = esame
 	mvi   __LC_AR_MODE_ID,1		 # set esame flag
         slr   %r0,%r0                    # set cpuid to zero
@@ -49,7 +48,18 @@ startup_continue:
 	lg    %r12,.Lparmaddr-.LPG1(%r13)# pointer to parameter area
 					 # move IPL device to lowcore
         mvc   __LC_IPLDEV(4),IPL_DEVICE+4-PARMAREA(%r12)
+#
+# Setup stack
+#
+	larl  %r15,init_thread_union
+	lg    %r14,__TI_task(%r15)	# cache current in lowcore
+	stg   %r14,__LC_CURRENT
+	aghi  %r15,1<<(PAGE_SHIFT+THREAD_ORDER) # init_task_union + THREAD_SIZE
+	stg   %r15,__LC_KERNEL_STACK	# set end of kernel stack
+	aghi  %r15,-160
+	xc    __SF_BACKCHAIN(4,%r15),__SF_BACKCHAIN(%r15) # clear backchain
 
+	brasl %r14,ipl_save_parameters
 #
 # clear bss memory
 #
@@ -269,6 +279,19 @@ startup_continue:
 .Lparmaddr:
 	.quad	PARMAREA
 
+	.globl ipl_schib
+ipl_schib:
+	.rept 13
+	.long 0
+	.endr
+
+	.globl ipl_flags
+ipl_flags:
+	.long 0
+	.globl ipl_devno
+ipl_devno:
+	.word 0
+
 	.org	0x12000
 .globl s390_readinfo_sccb
 s390_readinfo_sccb:
@@ -300,24 +323,12 @@ s390_readinfo_sccb:
         .globl _stext
 _stext:	basr  %r13,0                    # get base
 .LPG3:
-#
-# Setup stack
-#
-	larl  %r15,init_thread_union
-	lg    %r14,__TI_task(%r15)      # cache current in lowcore
-	stg   %r14,__LC_CURRENT
-        aghi  %r15,1<<(PAGE_SHIFT+THREAD_ORDER) # init_task_union + THREAD_SIZE
-        stg   %r15,__LC_KERNEL_STACK    # set end of kernel stack
-        aghi  %r15,-160
-        xc    __SF_BACKCHAIN(4,%r15),__SF_BACKCHAIN(%r15) # clear backchain
-
 # check control registers
         stctg  %c0,%c15,0(%r15)
 	oi     6(%r15),0x40             # enable sigp emergency signal
 	oi     4(%r15),0x10             # switch on low address proctection
         lctlg  %c0,%c15,0(%r15)
 
-#
         lam    0,15,.Laregs-.LPG3(%r13) # load access regs needed by uaccess
         brasl  %r14,start_kernel        # go to C code
 #
@@ -325,7 +336,7 @@ _stext:	basr  %r13,0                    # get base
 #
         basr  %r13,0
 	lpswe .Ldw-.(%r13)           # load disabled wait psw
-#
+
             .align 8
 .Ldw:       .quad  0x0002000180000000,0x0000000000000000
 .Laregs:    .long  0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0
diff --git a/arch/s390/kernel/ipl.c b/arch/s390/kernel/ipl.c
index 105ee15a2b31..6555cc48e28f 100644
--- a/arch/s390/kernel/ipl.c
+++ b/arch/s390/kernel/ipl.c
@@ -189,9 +189,9 @@ static enum ipl_type ipl_get_type(void)
 {
 	struct ipl_parameter_block *ipl = IPL_PARMBLOCK_START;
 
-	if (!IPL_DEVNO_VALID)
+	if (!(ipl_flags & IPL_DEVNO_VALID))
 		return IPL_TYPE_UNKNOWN;
-	if (!IPL_PARMBLOCK_VALID)
+	if (!(ipl_flags & IPL_PARMBLOCK_VALID))
 		return IPL_TYPE_CCW;
 	if (ipl->hdr.version > IPL_MAX_SUPPORTED_VERSION)
 		return IPL_TYPE_UNKNOWN;
diff --git a/drivers/s390/cio/cio.c b/drivers/s390/cio/cio.c
index 050963f15802..61eb7caa1567 100644
--- a/drivers/s390/cio/cio.c
+++ b/drivers/s390/cio/cio.c
@@ -16,11 +16,10 @@
 #include <linux/device.h>
 #include <linux/kernel_stat.h>
 #include <linux/interrupt.h>
-
 #include <asm/cio.h>
 #include <asm/delay.h>
 #include <asm/irq.h>
-
+#include <asm/setup.h>
 #include "airq.h"
 #include "cio.h"
 #include "css.h"
@@ -909,3 +908,38 @@ void reipl_ccw_dev(struct ccw_dev_id *devid)
 	cio_reset_channel_paths();
 	do_reipl_asm(*((__u32*)&schid));
 }
+
+extern struct schib ipl_schib;
+
+/*
+ * ipl_save_parameters gets called very early. It is not allowed to access
+ * anything in the bss section at all. The bss section is not cleared yet,
+ * but may contain some ipl parameters written by the firmware.
+ * These parameters (if present) are copied to 0x2000.
+ * To avoid corruption of the ipl parameters, all variables used by this
+ * function must reside on the stack or in the data section.
+ */
+void ipl_save_parameters(void)
+{
+	struct subchannel_id schid;
+	unsigned int *ipl_ptr;
+	void *src, *dst;
+
+	schid = *(struct subchannel_id *)__LC_SUBCHANNEL_ID;
+	if (!schid.one)
+		return;
+	if (stsch(schid, &ipl_schib))
+		return;
+	if (!ipl_schib.pmcw.dnv)
+		return;
+	ipl_devno = ipl_schib.pmcw.dev;
+	ipl_flags |= IPL_DEVNO_VALID;
+	if (!ipl_schib.pmcw.qf)
+		return;
+	ipl_flags |= IPL_PARMBLOCK_VALID;
+	ipl_ptr = (unsigned int *)__LC_IPL_PARMBLOCK_PTR;
+	src = (void *)(unsigned long)*ipl_ptr;
+	dst = (void *)IPL_PARMBLOCK_ORIGIN;
+	memmove(dst, src, PAGE_SIZE);
+	*ipl_ptr = IPL_PARMBLOCK_ORIGIN;
+}
diff --git a/include/asm-s390/lowcore.h b/include/asm-s390/lowcore.h
index 2e3d4cca5e21..18695d10dedf 100644
--- a/include/asm-s390/lowcore.h
+++ b/include/asm-s390/lowcore.h
@@ -35,6 +35,7 @@
 #define __LC_IO_NEW_PSW                 0x01f0
 #endif /* !__s390x__ */
 
+#define __LC_IPL_PARMBLOCK_PTR		0x014
 #define __LC_EXT_PARAMS                 0x080
 #define __LC_CPU_ADDRESS                0x084
 #define __LC_EXT_INT_CODE               0x086
diff --git a/include/asm-s390/setup.h b/include/asm-s390/setup.h
index 4a1126d8439a..00c03e46689b 100644
--- a/include/asm-s390/setup.h
+++ b/include/asm-s390/setup.h
@@ -125,13 +125,15 @@ struct ipl_parameter_block {
 /*
  * IPL validity flags and parameters as detected in head.S
  */
-extern u32 ipl_parameter_flags;
+extern u32 ipl_flags;
 extern u16 ipl_devno;
 
 void do_reipl(void);
 
-#define IPL_DEVNO_VALID		(ipl_parameter_flags & 1)
-#define IPL_PARMBLOCK_VALID	(ipl_parameter_flags & 2)
+enum {
+	IPL_DEVNO_VALID	= 1,
+	IPL_PARMBLOCK_VALID = 2,
+};
 
 #define IPL_PARMBLOCK_START	((struct ipl_parameter_block *) \
 				 IPL_PARMBLOCK_ORIGIN)
-- 
cgit v1.2.3


From 1f38d61347203055b55e34083cce7a9cd8c529a9 Mon Sep 17 00:00:00 2001
From: Gerald Schaefer <geraldsc@de.ibm.com>
Date: Wed, 20 Sep 2006 15:59:26 +0200
Subject: [S390] cleanup appldata.

Introduce asm header that contains the appldata data structures and
the diag inline assembly.

Signed-off-by: Gerald Schaefer <geraldsc@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/appldata/appldata.h      | 16 -------
 arch/s390/appldata/appldata_base.c | 81 +++++-----------------------------
 arch/s390/appldata/appldata_os.c   |  1 +
 include/asm-s390/appldata.h        | 90 ++++++++++++++++++++++++++++++++++++++
 4 files changed, 103 insertions(+), 85 deletions(-)
 create mode 100644 include/asm-s390/appldata.h

(limited to 'include')

diff --git a/arch/s390/appldata/appldata.h b/arch/s390/appldata/appldata.h
index 71d65eb30650..0429481dea63 100644
--- a/arch/s390/appldata/appldata.h
+++ b/arch/s390/appldata/appldata.h
@@ -29,22 +29,6 @@
 #define CTL_APPLDATA_NET_SUM	2125
 #define CTL_APPLDATA_PROC	2126
 
-#ifndef CONFIG_64BIT
-
-#define APPLDATA_START_INTERVAL_REC 0x00	/* Function codes for */
-#define APPLDATA_STOP_REC	    0x01	/* DIAG 0xDC	  */
-#define APPLDATA_GEN_EVENT_RECORD   0x02
-#define APPLDATA_START_CONFIG_REC   0x03
-
-#else
-
-#define APPLDATA_START_INTERVAL_REC 0x80
-#define APPLDATA_STOP_REC	    0x81
-#define APPLDATA_GEN_EVENT_RECORD   0x82
-#define APPLDATA_START_CONFIG_REC   0x83
-
-#endif /* CONFIG_64BIT */
-
 #define P_INFO(x...)	printk(KERN_INFO MY_PRINT_NAME " info: " x)
 #define P_ERROR(x...)	printk(KERN_ERR MY_PRINT_NAME " error: " x)
 #define P_WARNING(x...)	printk(KERN_WARNING MY_PRINT_NAME " status: " x)
diff --git a/arch/s390/appldata/appldata_base.c b/arch/s390/appldata/appldata_base.c
index a0a94e0ef8d1..b69ed742f981 100644
--- a/arch/s390/appldata/appldata_base.c
+++ b/arch/s390/appldata/appldata_base.c
@@ -14,20 +14,20 @@
 #include <linux/init.h>
 #include <linux/slab.h>
 #include <linux/errno.h>
-#include <asm/uaccess.h>
-#include <asm/io.h>
-#include <asm/smp.h>
 #include <linux/interrupt.h>
 #include <linux/proc_fs.h>
 #include <linux/page-flags.h>
 #include <linux/swap.h>
 #include <linux/pagemap.h>
 #include <linux/sysctl.h>
-#include <asm/timer.h>
-//#include <linux/kernel_stat.h>
 #include <linux/notifier.h>
 #include <linux/cpu.h>
 #include <linux/workqueue.h>
+#include <asm/appldata.h>
+#include <asm/timer.h>
+#include <asm/uaccess.h>
+#include <asm/io.h>
+#include <asm/smp.h>
 
 #include "appldata.h"
 
@@ -39,34 +39,6 @@
 
 #define TOD_MICRO	0x01000			/* nr. of TOD clock units
 						   for 1 microsecond */
-
-/*
- * Parameter list for DIAGNOSE X'DC'
- */
-#ifndef CONFIG_64BIT
-struct appldata_parameter_list {
-	u16 diag;		/* The DIAGNOSE code X'00DC'          */
-	u8  function;		/* The function code for the DIAGNOSE */
-	u8  parlist_length;	/* Length of the parameter list       */
-	u32 product_id_addr;	/* Address of the 16-byte product ID  */
-	u16 reserved;
-	u16 buffer_length;	/* Length of the application data buffer  */
-	u32 buffer_addr;	/* Address of the application data buffer */
-};
-#else
-struct appldata_parameter_list {
-	u16 diag;
-	u8  function;
-	u8  parlist_length;
-	u32 unused01;
-	u16 reserved;
-	u16 buffer_length;
-	u32 unused02;
-	u64 product_id_addr;
-	u64 buffer_addr;
-};
-#endif /* CONFIG_64BIT */
-
 /*
  * /proc entries (sysctl)
  */
@@ -181,46 +153,17 @@ static void appldata_work_fn(void *data)
 int appldata_diag(char record_nr, u16 function, unsigned long buffer,
 			u16 length, char *mod_lvl)
 {
-	unsigned long ry;
-	struct appldata_product_id {
-		char prod_nr[7];			/* product nr. */
-		char prod_fn[2];			/* product function */
-		char record_nr;				/* record nr. */
-		char version_nr[2];			/* version */
-		char release_nr[2];			/* release */
-		char mod_lvl[2];			/* modification lvl. */
-	} appldata_product_id = {
-	/* all strings are EBCDIC, record_nr is byte */
+	struct appldata_product_id id = {
 		.prod_nr    = {0xD3, 0xC9, 0xD5, 0xE4,
-				0xE7, 0xD2, 0xD9},	/* "LINUXKR" */
-		.prod_fn    = {0xD5, 0xD3},		/* "NL" */
+			       0xE7, 0xD2, 0xD9},	/* "LINUXKR" */
+		.prod_fn    = 0xD5D3,			/* "NL" */
 		.record_nr  = record_nr,
-		.version_nr = {0xF2, 0xF6},		/* "26" */
-		.release_nr = {0xF0, 0xF1},		/* "01" */
-		.mod_lvl    = {mod_lvl[0], mod_lvl[1]},
-	};
-	struct appldata_parameter_list appldata_parameter_list = {
-				.diag = 0xDC,
-				.function = function,
-				.parlist_length =
-					sizeof(appldata_parameter_list),
-				.buffer_length = length,
-				.product_id_addr =
-					(unsigned long) &appldata_product_id,
-				.buffer_addr = virt_to_phys((void *) buffer)
+		.version_nr = 0xF2F6,			/* "26" */
+		.release_nr = 0xF0F1,			/* "01" */
+		.mod_lvl    = (mod_lvl[0]) << 8 | mod_lvl[1],
 	};
 
-	if (!MACHINE_IS_VM)
-		return -ENOSYS;
-	ry = -1;
-	asm volatile(
-			"diag %1,%0,0xDC\n\t"
-			: "=d" (ry)
-			: "d" (&appldata_parameter_list),
-			  "m" (appldata_parameter_list),
-			  "m" (appldata_product_id)
-			: "cc");
-	return (int) ry;
+	return appldata_asm(&id, function, (void *) buffer, length);
 }
 /************************ timer, work, DIAG <END> ****************************/
 
diff --git a/arch/s390/appldata/appldata_os.c b/arch/s390/appldata/appldata_os.c
index 161acc5c8a1b..76a15523ae9e 100644
--- a/arch/s390/appldata/appldata_os.c
+++ b/arch/s390/appldata/appldata_os.c
@@ -16,6 +16,7 @@
 #include <linux/kernel_stat.h>
 #include <linux/netdevice.h>
 #include <linux/sched.h>
+#include <asm/appldata.h>
 #include <asm/smp.h>
 
 #include "appldata.h"
diff --git a/include/asm-s390/appldata.h b/include/asm-s390/appldata.h
new file mode 100644
index 000000000000..b1770703b706
--- /dev/null
+++ b/include/asm-s390/appldata.h
@@ -0,0 +1,90 @@
+/*
+ * include/asm-s390/appldata.h
+ *
+ * Copyright (C) IBM Corp. 2006
+ *
+ * Author(s): Melissa Howland <melissah@us.ibm.com>
+ */
+
+#ifndef _ASM_S390_APPLDATA_H
+#define _ASM_S390_APPLDATA_H
+
+#include <asm/io.h>
+
+#ifndef CONFIG_64BIT
+
+#define APPLDATA_START_INTERVAL_REC	0x00	/* Function codes for */
+#define APPLDATA_STOP_REC		0x01	/* DIAG 0xDC	      */
+#define APPLDATA_GEN_EVENT_REC		0x02
+#define APPLDATA_START_CONFIG_REC	0x03
+
+/*
+ * Parameter list for DIAGNOSE X'DC'
+ */
+struct appldata_parameter_list {
+	u16 diag;		/* The DIAGNOSE code X'00DC'	      */
+	u8  function;		/* The function code for the DIAGNOSE */
+	u8  parlist_length;	/* Length of the parameter list       */
+	u32 product_id_addr;	/* Address of the 16-byte product ID  */
+	u16 reserved;
+	u16 buffer_length;	/* Length of the application data buffer  */
+	u32 buffer_addr;	/* Address of the application data buffer */
+} __attribute__ ((packed));
+
+#else /* CONFIG_64BIT */
+
+#define APPLDATA_START_INTERVAL_REC	0x80
+#define APPLDATA_STOP_REC		0x81
+#define APPLDATA_GEN_EVENT_REC		0x82
+#define APPLDATA_START_CONFIG_REC	0x83
+
+/*
+ * Parameter list for DIAGNOSE X'DC'
+ */
+struct appldata_parameter_list {
+	u16 diag;
+	u8  function;
+	u8  parlist_length;
+	u32 unused01;
+	u16 reserved;
+	u16 buffer_length;
+	u32 unused02;
+	u64 product_id_addr;
+	u64 buffer_addr;
+} __attribute__ ((packed));
+
+#endif /* CONFIG_64BIT */
+
+struct appldata_product_id {
+	char prod_nr[7];	/* product number */
+	u16  prod_fn;		/* product function */
+	u8   record_nr; 	/* record number */
+	u16  version_nr;	/* version */
+	u16  release_nr;	/* release */
+	u16  mod_lvl;		/* modification level */
+} __attribute__ ((packed));
+
+static inline int appldata_asm(struct appldata_product_id *id,
+			       unsigned short fn, void *buffer,
+			       unsigned short length)
+{
+	struct appldata_parameter_list parm_list;
+	int ry;
+
+	if (!MACHINE_IS_VM)
+		return -ENOSYS;
+	parm_list.diag = 0xdc;
+	parm_list.function = fn;
+	parm_list.parlist_length = sizeof(parm_list);
+	parm_list.buffer_length = length;
+	parm_list.product_id_addr = (unsigned long) id;
+	parm_list.buffer_addr = virt_to_phys(buffer);
+	asm volatile(
+		"diag %1,%0,0xdc"
+		: "=d" (ry)
+		: "d" (&parm_list), "m" (parm_list), "m" (*id)
+		: "cc");
+	return ry;
+}
+
+#endif /* _ASM_S390_APPLDATA_H */
-- 
cgit v1.2.3


From 31b58088292c7f00f0b81088bfb557285b0b6247 Mon Sep 17 00:00:00 2001
From: Melissa Howland <melissah@us.ibm.com>
Date: Wed, 20 Sep 2006 15:59:34 +0200
Subject: [S390] Linux API for writing z/VM APPLDATA Monitor records.

This patch delivers a new Linux API in the form of a misc char
device that is useable from user space and allows write access
to the z/VM APPLDATA Monitor Records collected by the *MONITOR
System Service of z/VM.

Signed-off-by: Melissa Howland <melissah@us.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/defconfig           |   1 +
 drivers/s390/Kconfig          |   6 +
 drivers/s390/char/Makefile    |   1 +
 drivers/s390/char/monwriter.c | 292 ++++++++++++++++++++++++++++++++++++++++++
 include/asm-s390/Kbuild       |   2 +-
 include/asm-s390/monwriter.h  |  33 +++++
 6 files changed, 334 insertions(+), 1 deletion(-)
 create mode 100644 drivers/s390/char/monwriter.c
 create mode 100644 include/asm-s390/monwriter.h

(limited to 'include')

diff --git a/arch/s390/defconfig b/arch/s390/defconfig
index f1d4591eddbb..35da53986b1b 100644
--- a/arch/s390/defconfig
+++ b/arch/s390/defconfig
@@ -428,6 +428,7 @@ CONFIG_S390_TAPE_34XX=m
 # CONFIG_VMLOGRDR is not set
 # CONFIG_VMCP is not set
 # CONFIG_MONREADER is not set
+CONFIG_MONWRITER=m
 
 #
 # Cryptographic devices
diff --git a/drivers/s390/Kconfig b/drivers/s390/Kconfig
index bc4261e8b606..ae89b9b88743 100644
--- a/drivers/s390/Kconfig
+++ b/drivers/s390/Kconfig
@@ -213,6 +213,12 @@ config MONREADER
 	help
 	  Character device driver for reading z/VM monitor service records
 
+config MONWRITER
+	tristate "API for writing z/VM monitor service records"
+	default "m"
+	help
+	  Character device driver for writing z/VM monitor service records
+
 endmenu
 
 menu "Cryptographic devices"
diff --git a/drivers/s390/char/Makefile b/drivers/s390/char/Makefile
index 0c0162ff6c0c..c3e97b4fc186 100644
--- a/drivers/s390/char/Makefile
+++ b/drivers/s390/char/Makefile
@@ -28,3 +28,4 @@ obj-$(CONFIG_S390_TAPE) += tape.o tape_class.o
 obj-$(CONFIG_S390_TAPE_34XX) += tape_34xx.o
 obj-$(CONFIG_S390_TAPE_3590) += tape_3590.o
 obj-$(CONFIG_MONREADER) += monreader.o
+obj-$(CONFIG_MONWRITER) += monwriter.o
diff --git a/drivers/s390/char/monwriter.c b/drivers/s390/char/monwriter.c
new file mode 100644
index 000000000000..1e3939aeb8ab
--- /dev/null
+++ b/drivers/s390/char/monwriter.c
@@ -0,0 +1,292 @@
+/*
+ * drivers/s390/char/monwriter.c
+ *
+ * Character device driver for writing z/VM *MONITOR service records.
+ *
+ * Copyright (C) IBM Corp. 2006
+ *
+ * Author(s): Melissa Howland <Melissa.Howland@us.ibm.com>
+ */
+
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/init.h>
+#include <linux/errno.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/miscdevice.h>
+#include <linux/ctype.h>
+#include <linux/poll.h>
+#include <asm/uaccess.h>
+#include <asm/ebcdic.h>
+#include <asm/io.h>
+#include <asm/appldata.h>
+#include <asm/monwriter.h>
+
+#define MONWRITE_MAX_DATALEN	4024
+
+static int mon_max_bufs = 255;
+
+struct mon_buf {
+	struct list_head list;
+	struct monwrite_hdr hdr;
+	int diag_done;
+	char *data;
+};
+
+struct mon_private {
+	struct list_head list;
+	struct monwrite_hdr hdr;
+	size_t hdr_to_read;
+	size_t data_to_read;
+	struct mon_buf *current_buf;
+	int mon_buf_count;
+};
+
+/*
+ * helper functions
+ */
+
+static int monwrite_diag(struct monwrite_hdr *myhdr, char *buffer, int fcn)
+{
+	struct appldata_product_id id;
+	int rc;
+
+	strcpy(id.prod_nr, "LNXAPPL");
+	id.prod_fn = myhdr->applid;
+	id.record_nr = myhdr->record_num;
+	id.version_nr = myhdr->version;
+	id.release_nr = myhdr->release;
+	id.mod_lvl = myhdr->mod_level;
+	rc = appldata_asm(&id, fcn, (void *) buffer, myhdr->datalen);
+	if (rc <= 0)
+		return rc;
+	if (rc == 5)
+		return -EPERM;
+	printk("DIAG X'DC' error with return code: %i\n", rc);
+	return -EINVAL;
+}
+
+static inline struct mon_buf *monwrite_find_hdr(struct mon_private *monpriv,
+						struct monwrite_hdr *monhdr)
+{
+	struct mon_buf *entry, *next;
+
+	list_for_each_entry_safe(entry, next, &monpriv->list, list)
+		if (entry->hdr.applid == monhdr->applid &&
+		    entry->hdr.record_num == monhdr->record_num &&
+		    entry->hdr.version == monhdr->version &&
+		    entry->hdr.release == monhdr->release &&
+		    entry->hdr.mod_level == monhdr->mod_level)
+			return entry;
+	return NULL;
+}
+
+static int monwrite_new_hdr(struct mon_private *monpriv)
+{
+	struct monwrite_hdr *monhdr = &monpriv->hdr;
+	struct mon_buf *monbuf;
+	int rc;
+
+	if (monhdr->datalen > MONWRITE_MAX_DATALEN ||
+	    monhdr->mon_function > MONWRITE_START_CONFIG ||
+	    monhdr->hdrlen != sizeof(struct monwrite_hdr))
+		return -EINVAL;
+	monbuf = monwrite_find_hdr(monpriv, monhdr);
+	if (monbuf) {
+		if (monhdr->mon_function == MONWRITE_STOP_INTERVAL) {
+			monhdr->datalen = monbuf->hdr.datalen;
+			rc = monwrite_diag(monhdr, monbuf->data,
+					   APPLDATA_STOP_REC);
+			list_del(&monbuf->list);
+			monpriv->mon_buf_count--;
+			kfree(monbuf->data);
+			kfree(monbuf);
+			monbuf = NULL;
+		}
+	} else {
+		if (monpriv->mon_buf_count >= mon_max_bufs)
+			return -ENOSPC;
+		monbuf = kzalloc(sizeof(struct mon_buf), GFP_KERNEL);
+		if (!monbuf)
+			return -ENOMEM;
+		monbuf->data = kzalloc(monbuf->hdr.datalen,
+				       GFP_KERNEL | GFP_DMA);
+		if (!monbuf->data) {
+			kfree(monbuf);
+			return -ENOMEM;
+		}
+		monbuf->hdr = *monhdr;
+		list_add_tail(&monbuf->list, &monpriv->list);
+		monpriv->mon_buf_count++;
+	}
+	monpriv->current_buf = monbuf;
+	return 0;
+}
+
+static int monwrite_new_data(struct mon_private *monpriv)
+{
+	struct monwrite_hdr *monhdr = &monpriv->hdr;
+	struct mon_buf *monbuf = monpriv->current_buf;
+	int rc = 0;
+
+	switch (monhdr->mon_function) {
+	case MONWRITE_START_INTERVAL:
+		if (!monbuf->diag_done) {
+			rc = monwrite_diag(monhdr, monbuf->data,
+					   APPLDATA_START_INTERVAL_REC);
+			monbuf->diag_done = 1;
+		}
+		break;
+	case MONWRITE_START_CONFIG:
+		if (!monbuf->diag_done) {
+			rc = monwrite_diag(monhdr, monbuf->data,
+					   APPLDATA_START_CONFIG_REC);
+			monbuf->diag_done = 1;
+		}
+		break;
+	case MONWRITE_GEN_EVENT:
+		rc = monwrite_diag(monhdr, monbuf->data,
+				   APPLDATA_GEN_EVENT_REC);
+		list_del(&monpriv->current_buf->list);
+		kfree(monpriv->current_buf->data);
+		kfree(monpriv->current_buf);
+		monpriv->current_buf = NULL;
+		break;
+	default:
+		/* monhdr->mon_function is checked in monwrite_new_hdr */
+		BUG();
+	}
+	return rc;
+}
+
+/*
+ * file operations
+ */
+
+static int monwrite_open(struct inode *inode, struct file *filp)
+{
+	struct mon_private *monpriv;
+
+	monpriv = kzalloc(sizeof(struct mon_private), GFP_KERNEL);
+	if (!monpriv)
+		return -ENOMEM;
+	INIT_LIST_HEAD(&monpriv->list);
+	monpriv->hdr_to_read = sizeof(monpriv->hdr);
+	filp->private_data = monpriv;
+	return nonseekable_open(inode, filp);
+}
+
+static int monwrite_close(struct inode *inode, struct file *filp)
+{
+	struct mon_private *monpriv = filp->private_data;
+	struct mon_buf *entry, *next;
+
+	list_for_each_entry_safe(entry, next, &monpriv->list, list) {
+		if (entry->hdr.mon_function != MONWRITE_GEN_EVENT)
+			monwrite_diag(&entry->hdr, entry->data,
+				      APPLDATA_STOP_REC);
+		monpriv->mon_buf_count--;
+		list_del(&entry->list);
+		kfree(entry->data);
+		kfree(entry);
+	}
+	kfree(monpriv);
+	return 0;
+}
+
+static ssize_t monwrite_write(struct file *filp, const char __user *data,
+			      size_t count, loff_t *ppos)
+{
+	struct mon_private *monpriv = filp->private_data;
+	size_t len, written;
+	void *to;
+	int rc;
+
+	for (written = 0; written < count; ) {
+		if (monpriv->hdr_to_read) {
+			len = min(count - written, monpriv->hdr_to_read);
+			to = (char *) &monpriv->hdr +
+				sizeof(monpriv->hdr) - monpriv->hdr_to_read;
+			if (copy_from_user(to, data + written, len)) {
+				rc = -EFAULT;
+				goto out_error;
+			}
+			monpriv->hdr_to_read -= len;
+			written += len;
+			if (monpriv->hdr_to_read > 0)
+				continue;
+			rc = monwrite_new_hdr(monpriv);
+			if (rc)
+				goto out_error;
+			monpriv->data_to_read = monpriv->current_buf ?
+				monpriv->current_buf->hdr.datalen : 0;
+		}
+
+		if (monpriv->data_to_read) {
+			len = min(count - written, monpriv->data_to_read);
+			to = monpriv->current_buf->data +
+				monpriv->hdr.datalen - monpriv->data_to_read;
+			if (copy_from_user(to, data + written, len)) {
+				rc = -EFAULT;
+				goto out_error;
+			}
+			monpriv->data_to_read -= len;
+			written += len;
+			if (monpriv->data_to_read > 0)
+				continue;
+			rc = monwrite_new_data(monpriv);
+			if (rc)
+				goto out_error;
+		}
+		monpriv->hdr_to_read = sizeof(monpriv->hdr);
+	}
+	return written;
+
+out_error:
+	monpriv->data_to_read = 0;
+	monpriv->hdr_to_read = sizeof(struct monwrite_hdr);
+	return rc;
+}
+
+static struct file_operations monwrite_fops = {
+	.owner	 = THIS_MODULE,
+	.open	 = &monwrite_open,
+	.release = &monwrite_close,
+	.write	 = &monwrite_write,
+};
+
+static struct miscdevice mon_dev = {
+	.name	= "monwriter",
+	.fops	= &monwrite_fops,
+	.minor	= MISC_DYNAMIC_MINOR,
+};
+
+/*
+ * module init/exit
+ */
+
+static int __init mon_init(void)
+{
+	if (MACHINE_IS_VM)
+		return misc_register(&mon_dev);
+	else
+		return -ENODEV;
+}
+
+static void __exit mon_exit(void)
+{
+	WARN_ON(misc_deregister(&mon_dev) != 0);
+}
+
+module_init(mon_init);
+module_exit(mon_exit);
+
+module_param_named(max_bufs, mon_max_bufs, int, 0644);
+MODULE_PARM_DESC(max_bufs, "Maximum number of sample monitor data buffers"
+		 "that can be active at one time");
+
+MODULE_AUTHOR("Melissa Howland <Melissa.Howland@us.ibm.com>");
+MODULE_DESCRIPTION("Character device driver for writing z/VM "
+		   "APPLDATA monitor records.");
+MODULE_LICENSE("GPL");
diff --git a/include/asm-s390/Kbuild b/include/asm-s390/Kbuild
index ed8955f49e47..979145026a29 100644
--- a/include/asm-s390/Kbuild
+++ b/include/asm-s390/Kbuild
@@ -1,4 +1,4 @@
 include include/asm-generic/Kbuild.asm
 
 unifdef-y += cmb.h debug.h
-header-y += dasd.h qeth.h tape390.h ucontext.h vtoc.h z90crypt.h
+header-y += dasd.h monwriter.h qeth.h tape390.h ucontext.h vtoc.h z90crypt.h
diff --git a/include/asm-s390/monwriter.h b/include/asm-s390/monwriter.h
new file mode 100644
index 000000000000..f0cbf96c52e6
--- /dev/null
+++ b/include/asm-s390/monwriter.h
@@ -0,0 +1,33 @@
+/*
+ * include/asm-s390/monwriter.h
+ *
+ * Copyright (C) IBM Corp. 2006
+ * Character device driver for writing z/VM APPLDATA monitor records
+ * Version 1.0
+ * Author(s): Melissa Howland <melissah@us.ibm.com>
+ *
+ */
+
+#ifndef _ASM_390_MONWRITER_H
+#define _ASM_390_MONWRITER_H
+
+/* mon_function values */
+#define MONWRITE_START_INTERVAL	0x00 /* start interval recording */
+#define MONWRITE_STOP_INTERVAL	0x01 /* stop interval or config recording */
+#define MONWRITE_GEN_EVENT	0x02 /* generate event record */
+#define MONWRITE_START_CONFIG	0x03 /* start configuration recording */
+
+/* the header the app uses in its write() data */
+struct monwrite_hdr {
+	unsigned char mon_function;
+	unsigned short applid;
+	unsigned char record_num;
+	unsigned short version;
+	unsigned short release;
+	unsigned short mod_level;
+	unsigned short datalen;
+	unsigned char hdrlen;
+
+} __attribute__((packed));
+
+#endif /* _ASM_390_MONWRITER_H */
-- 
cgit v1.2.3


From 9282ed929758b82f448a40d3c17319d794970624 Mon Sep 17 00:00:00 2001
From: Gerald Schaefer <geraldsc@de.ibm.com>
Date: Wed, 20 Sep 2006 15:59:37 +0200
Subject: [S390] Cleanup in page table related code.

Changed and simplified some page table related #defines and code.

Signed-off-by: Gerald Schaefer <geraldsc@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/mm/init.c        |  36 +++++--------
 include/asm-s390/pgalloc.h |  67 ++++++++++++------------
 include/asm-s390/pgtable.h | 124 +++++++++++++++++++++------------------------
 3 files changed, 106 insertions(+), 121 deletions(-)

(limited to 'include')

diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c
index 6e6b6de77770..cfd9b8f7a523 100644
--- a/arch/s390/mm/init.c
+++ b/arch/s390/mm/init.c
@@ -108,16 +108,23 @@ void __init paging_init(void)
         unsigned long pgdir_k = (__pa(swapper_pg_dir) & PAGE_MASK) | _KERNSEG_TABLE;
         static const int ssm_mask = 0x04000000L;
 	unsigned long ro_start_pfn, ro_end_pfn;
+	unsigned long zones_size[MAX_NR_ZONES];
 
 	ro_start_pfn = PFN_DOWN((unsigned long)&__start_rodata);
 	ro_end_pfn = PFN_UP((unsigned long)&__end_rodata);
 
+	memset(zones_size, 0, sizeof(zones_size));
+	zones_size[ZONE_DMA] = max_low_pfn;
+	free_area_init_node(0, &contig_page_data, zones_size,
+			    __pa(PAGE_OFFSET) >> PAGE_SHIFT,
+			    zholes_size);
+
 	/* unmap whole virtual address space */
 	
         pg_dir = swapper_pg_dir;
 
-	for (i=0;i<KERNEL_PGD_PTRS;i++) 
-	        pmd_clear((pmd_t*)pg_dir++);
+	for (i = 0; i < PTRS_PER_PGD; i++)
+		pmd_clear((pmd_t *) pg_dir++);
 		
 	/*
 	 * map whole physical memory to virtual memory (identity mapping) 
@@ -131,10 +138,7 @@ void __init paging_init(void)
                  */
 		pg_table = (pte_t *) alloc_bootmem_pages(PAGE_SIZE);
 
-                pg_dir->pgd0 =  (_PAGE_TABLE | __pa(pg_table));
-                pg_dir->pgd1 =  (_PAGE_TABLE | (__pa(pg_table)+1024));
-                pg_dir->pgd2 =  (_PAGE_TABLE | (__pa(pg_table)+2048));
-                pg_dir->pgd3 =  (_PAGE_TABLE | (__pa(pg_table)+3072));
+		pmd_populate_kernel(&init_mm, (pmd_t *) pg_dir, pg_table);
                 pg_dir++;
 
                 for (tmp = 0 ; tmp < PTRS_PER_PTE ; tmp++,pg_table++) {
@@ -143,8 +147,8 @@ void __init paging_init(void)
 			else
 				pte = pfn_pte(pfn, PAGE_KERNEL);
                         if (pfn >= max_low_pfn)
-                                pte_clear(&init_mm, 0, &pte);
-                        set_pte(pg_table, pte);
+				pte_val(pte) = _PAGE_TYPE_EMPTY;
+			set_pte(pg_table, pte);
                         pfn++;
                 }
         }
@@ -159,16 +163,6 @@ void __init paging_init(void)
 			     : : "m" (pgdir_k), "m" (ssm_mask));
 
         local_flush_tlb();
-
-	{
-		unsigned long zones_size[MAX_NR_ZONES];
-
-		memset(zones_size, 0, sizeof(zones_size));
-		zones_size[ZONE_DMA] = max_low_pfn;
-		free_area_init_node(0, &contig_page_data, zones_size,
-				    __pa(PAGE_OFFSET) >> PAGE_SHIFT,
-				    zholes_size);
-	}
         return;
 }
 
@@ -236,10 +230,8 @@ void __init paging_init(void)
 					pte = pfn_pte(pfn, __pgprot(_PAGE_RO));
 				else
 					pte = pfn_pte(pfn, PAGE_KERNEL);
-                                if (pfn >= max_low_pfn) {
-                                        pte_clear(&init_mm, 0, &pte); 
-                                        continue;
-                                }
+				if (pfn >= max_low_pfn)
+					pte_val(pte) = _PAGE_TYPE_EMPTY;
                                 set_pte(pt_dir, pte);
                                 pfn++;
                         }
diff --git a/include/asm-s390/pgalloc.h b/include/asm-s390/pgalloc.h
index a78e853e0dd5..803bc7064418 100644
--- a/include/asm-s390/pgalloc.h
+++ b/include/asm-s390/pgalloc.h
@@ -21,6 +21,16 @@
 
 extern void diag10(unsigned long addr);
 
+/*
+ * Page allocation orders.
+ */
+#ifndef __s390x__
+# define PGD_ALLOC_ORDER	1
+#else /* __s390x__ */
+# define PMD_ALLOC_ORDER	2
+# define PGD_ALLOC_ORDER	2
+#endif /* __s390x__ */
+
 /*
  * Allocate and free page tables. The xxx_kernel() versions are
  * used to allocate a kernel page table - this turns on ASN bits
@@ -29,30 +39,23 @@ extern void diag10(unsigned long addr);
 
 static inline pgd_t *pgd_alloc(struct mm_struct *mm)
 {
-	pgd_t *pgd;
+	pgd_t *pgd = (pgd_t *) __get_free_pages(GFP_KERNEL, PGD_ALLOC_ORDER);
 	int i;
 
+	if (!pgd)
+		return NULL;
+	for (i = 0; i < PTRS_PER_PGD; i++)
 #ifndef __s390x__
-	pgd = (pgd_t *) __get_free_pages(GFP_KERNEL,1);
-        if (pgd != NULL)
-		for (i = 0; i < USER_PTRS_PER_PGD; i++)
-			pmd_clear(pmd_offset(pgd + i, i*PGDIR_SIZE));
-#else /* __s390x__ */
-	pgd = (pgd_t *) __get_free_pages(GFP_KERNEL,2);
-        if (pgd != NULL)
-		for (i = 0; i < PTRS_PER_PGD; i++)
-			pgd_clear(pgd + i);
-#endif /* __s390x__ */
+		pmd_clear(pmd_offset(pgd + i, i*PGDIR_SIZE));
+#else
+		pgd_clear(pgd + i);
+#endif
 	return pgd;
 }
 
 static inline void pgd_free(pgd_t *pgd)
 {
-#ifndef __s390x__
-        free_pages((unsigned long) pgd, 1);
-#else /* __s390x__ */
-        free_pages((unsigned long) pgd, 2);
-#endif /* __s390x__ */
+	free_pages((unsigned long) pgd, PGD_ALLOC_ORDER);
 }
 
 #ifndef __s390x__
@@ -68,20 +71,19 @@ static inline void pgd_free(pgd_t *pgd)
 #else /* __s390x__ */
 static inline pmd_t * pmd_alloc_one(struct mm_struct *mm, unsigned long vmaddr)
 {
-	pmd_t *pmd;
-        int i;
+	pmd_t *pmd = (pmd_t *) __get_free_pages(GFP_KERNEL, PMD_ALLOC_ORDER);
+	int i;
 
-	pmd = (pmd_t *) __get_free_pages(GFP_KERNEL, 2);
-	if (pmd != NULL) {
-		for (i=0; i < PTRS_PER_PMD; i++)
-			pmd_clear(pmd+i);
-	}
+	if (!pmd)
+		return NULL;
+	for (i=0; i < PTRS_PER_PMD; i++)
+		pmd_clear(pmd + i);
 	return pmd;
 }
 
 static inline void pmd_free (pmd_t *pmd)
 {
-	free_pages((unsigned long) pmd, 2);
+	free_pages((unsigned long) pmd, PMD_ALLOC_ORDER);
 }
 
 #define __pmd_free_tlb(tlb,pmd)			\
@@ -123,15 +125,14 @@ pmd_populate(struct mm_struct *mm, pmd_t *pmd, struct page *page)
 static inline pte_t *
 pte_alloc_one_kernel(struct mm_struct *mm, unsigned long vmaddr)
 {
-	pte_t *pte;
-        int i;
-
-	pte = (pte_t *)__get_free_page(GFP_KERNEL|__GFP_REPEAT);
-	if (pte != NULL) {
-		for (i=0; i < PTRS_PER_PTE; i++) {
-			pte_clear(mm, vmaddr, pte+i);
-			vmaddr += PAGE_SIZE;
-		}
+	pte_t *pte = (pte_t *) __get_free_page(GFP_KERNEL|__GFP_REPEAT);
+	int i;
+
+	if (!pte)
+		return NULL;
+	for (i=0; i < PTRS_PER_PTE; i++) {
+		pte_clear(mm, vmaddr, pte + i);
+		vmaddr += PAGE_SIZE;
 	}
 	return pte;
 }
diff --git a/include/asm-s390/pgtable.h b/include/asm-s390/pgtable.h
index 24312387fa24..1a07028d575e 100644
--- a/include/asm-s390/pgtable.h
+++ b/include/asm-s390/pgtable.h
@@ -89,19 +89,6 @@ extern char empty_zero_page[PAGE_SIZE];
 # define PTRS_PER_PGD    2048
 #endif /* __s390x__ */
 
-/*
- * pgd entries used up by user/kernel:
- */
-#ifndef __s390x__
-# define USER_PTRS_PER_PGD  512
-# define USER_PGD_PTRS      512
-# define KERNEL_PGD_PTRS    512
-#else /* __s390x__ */
-# define USER_PTRS_PER_PGD  2048
-# define USER_PGD_PTRS      2048
-# define KERNEL_PGD_PTRS    2048
-#endif /* __s390x__ */
-
 #define FIRST_USER_ADDRESS  0
 
 #define pte_ERROR(e) \
@@ -216,12 +203,14 @@ extern char empty_zero_page[PAGE_SIZE];
 #define _PAGE_RO        0x200          /* HW read-only                     */
 #define _PAGE_INVALID   0x400          /* HW invalid                       */
 
-/* Mask and four different kinds of invalid pages. */
-#define _PAGE_INVALID_MASK	0x601
-#define _PAGE_INVALID_EMPTY	0x400
-#define _PAGE_INVALID_NONE	0x401
-#define _PAGE_INVALID_SWAP	0x600
-#define _PAGE_INVALID_FILE	0x601
+/* Mask and six different types of pages. */
+#define _PAGE_TYPE_MASK		0x601
+#define _PAGE_TYPE_EMPTY	0x400
+#define _PAGE_TYPE_NONE		0x401
+#define _PAGE_TYPE_SWAP		0x600
+#define _PAGE_TYPE_FILE		0x601
+#define _PAGE_TYPE_RO		0x200
+#define _PAGE_TYPE_RW		0x000
 
 #ifndef __s390x__
 
@@ -280,15 +269,14 @@ extern char empty_zero_page[PAGE_SIZE];
 #endif /* __s390x__ */
 
 /*
- * No mapping available
+ * Page protection definitions.
  */
-#define PAGE_NONE_SHARED  __pgprot(_PAGE_INVALID_NONE)
-#define PAGE_NONE_PRIVATE __pgprot(_PAGE_INVALID_NONE)
-#define PAGE_RO_SHARED	  __pgprot(_PAGE_RO)
-#define PAGE_RO_PRIVATE	  __pgprot(_PAGE_RO)
-#define PAGE_COPY	  __pgprot(_PAGE_RO)
-#define PAGE_SHARED	  __pgprot(0)
-#define PAGE_KERNEL	  __pgprot(0)
+#define PAGE_NONE	__pgprot(_PAGE_TYPE_NONE)
+#define PAGE_RO		__pgprot(_PAGE_TYPE_RO)
+#define PAGE_RW		__pgprot(_PAGE_TYPE_RW)
+
+#define PAGE_KERNEL	PAGE_RW
+#define PAGE_COPY	PAGE_RO
 
 /*
  * The S390 can't do page protection for execute, and considers that the
@@ -296,23 +284,23 @@ extern char empty_zero_page[PAGE_SIZE];
  * the closest we can get..
  */
          /*xwr*/
-#define __P000  PAGE_NONE_PRIVATE
-#define __P001  PAGE_RO_PRIVATE
-#define __P010  PAGE_COPY
-#define __P011  PAGE_COPY
-#define __P100  PAGE_RO_PRIVATE
-#define __P101  PAGE_RO_PRIVATE
-#define __P110  PAGE_COPY
-#define __P111  PAGE_COPY
-
-#define __S000  PAGE_NONE_SHARED
-#define __S001  PAGE_RO_SHARED
-#define __S010  PAGE_SHARED
-#define __S011  PAGE_SHARED
-#define __S100  PAGE_RO_SHARED
-#define __S101  PAGE_RO_SHARED
-#define __S110  PAGE_SHARED
-#define __S111  PAGE_SHARED
+#define __P000	PAGE_NONE
+#define __P001	PAGE_RO
+#define __P010	PAGE_RO
+#define __P011	PAGE_RO
+#define __P100	PAGE_RO
+#define __P101	PAGE_RO
+#define __P110	PAGE_RO
+#define __P111	PAGE_RO
+
+#define __S000	PAGE_NONE
+#define __S001	PAGE_RO
+#define __S010	PAGE_RW
+#define __S011	PAGE_RW
+#define __S100	PAGE_RO
+#define __S101	PAGE_RO
+#define __S110	PAGE_RW
+#define __S111	PAGE_RW
 
 /*
  * Certain architectures need to do special things when PTEs
@@ -377,18 +365,18 @@ static inline int pmd_bad(pmd_t pmd)
 
 static inline int pte_none(pte_t pte)
 {
-	return (pte_val(pte) & _PAGE_INVALID_MASK) == _PAGE_INVALID_EMPTY;
+	return (pte_val(pte) & _PAGE_TYPE_MASK) == _PAGE_TYPE_EMPTY;
 }
 
 static inline int pte_present(pte_t pte)
 {
 	return !(pte_val(pte) & _PAGE_INVALID) ||
-		(pte_val(pte) & _PAGE_INVALID_MASK) == _PAGE_INVALID_NONE;
+		(pte_val(pte) & _PAGE_TYPE_MASK) == _PAGE_TYPE_NONE;
 }
 
 static inline int pte_file(pte_t pte)
 {
-	return (pte_val(pte) & _PAGE_INVALID_MASK) == _PAGE_INVALID_FILE;
+	return (pte_val(pte) & _PAGE_TYPE_MASK) == _PAGE_TYPE_FILE;
 }
 
 #define pte_same(a,b)	(pte_val(a) == pte_val(b))
@@ -461,7 +449,7 @@ static inline void pmd_clear(pmd_t * pmdp)
 
 static inline void pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
 {
-	pte_val(*ptep) = _PAGE_INVALID_EMPTY;
+	pte_val(*ptep) = _PAGE_TYPE_EMPTY;
 }
 
 /*
@@ -477,7 +465,7 @@ static inline pte_t pte_modify(pte_t pte, pgprot_t newprot)
 
 static inline pte_t pte_wrprotect(pte_t pte)
 {
-	/* Do not clobber _PAGE_INVALID_NONE pages!  */
+	/* Do not clobber _PAGE_TYPE_NONE pages!  */
 	if (!(pte_val(pte) & _PAGE_INVALID))
 		pte_val(pte) |= _PAGE_RO;
 	return pte;
@@ -556,26 +544,30 @@ static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr,
 	return pte;
 }
 
-static inline pte_t
-ptep_clear_flush(struct vm_area_struct *vma,
-		 unsigned long address, pte_t *ptep)
+static inline void __ptep_ipte(unsigned long address, pte_t *ptep)
 {
-	pte_t pte = *ptep;
+	if (!(pte_val(*ptep) & _PAGE_INVALID)) {
 #ifndef __s390x__
-	if (!(pte_val(pte) & _PAGE_INVALID)) {
 		/* S390 has 1mb segments, we are emulating 4MB segments */
 		pte_t *pto = (pte_t *) (((unsigned long) ptep) & 0x7ffffc00);
-		__asm__ __volatile__ ("ipte %2,%3"
-				      : "=m" (*ptep) : "m" (*ptep),
-				        "a" (pto), "a" (address) );
+#else
+		/* ipte in zarch mode can do the math */
+		pte_t *pto = ptep;
+#endif
+		asm volatile ("ipte %2,%3"
+			      : "=m" (*ptep) : "m" (*ptep),
+				"a" (pto), "a" (address) );
 	}
-#else /* __s390x__ */
-	if (!(pte_val(pte) & _PAGE_INVALID)) 
-		__asm__ __volatile__ ("ipte %2,%3"
-				      : "=m" (*ptep) : "m" (*ptep),
-				        "a" (ptep), "a" (address) );
-#endif /* __s390x__ */
-	pte_val(*ptep) = _PAGE_INVALID_EMPTY;
+	pte_val(*ptep) = _PAGE_TYPE_EMPTY;
+}
+
+static inline pte_t
+ptep_clear_flush(struct vm_area_struct *vma,
+		 unsigned long address, pte_t *ptep)
+{
+	pte_t pte = *ptep;
+
+	__ptep_ipte(address, ptep);
 	return pte;
 }
 
@@ -755,7 +747,7 @@ static inline pte_t mk_swap_pte(unsigned long type, unsigned long offset)
 {
 	pte_t pte;
 	offset &= __SWP_OFFSET_MASK;
-	pte_val(pte) = _PAGE_INVALID_SWAP | ((type & 0x1f) << 2) |
+	pte_val(pte) = _PAGE_TYPE_SWAP | ((type & 0x1f) << 2) |
 		((offset & 1UL) << 7) | ((offset & ~1UL) << 11);
 	return pte;
 }
@@ -778,7 +770,7 @@ static inline pte_t mk_swap_pte(unsigned long type, unsigned long offset)
 
 #define pgoff_to_pte(__off) \
 	((pte_t) { ((((__off) & 0x7f) << 1) + (((__off) >> 7) << 12)) \
-		   | _PAGE_INVALID_FILE })
+		   | _PAGE_TYPE_FILE })
 
 #endif /* !__ASSEMBLY__ */
 
-- 
cgit v1.2.3


From d02765d1af743567398eb6d523dea0ba5e5e7e8e Mon Sep 17 00:00:00 2001
From: Gerald Schaefer <geraldsc@de.ibm.com>
Date: Wed, 20 Sep 2006 15:59:42 +0200
Subject: [S390] Make user-copy operations run-time configurable.

Introduces a struct uaccess_ops which allows setting user-copy
operations at run-time.

Signed-off-by: Gerald Schaefer <geraldsc@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/s390_ksyms.c |   6 -
 arch/s390/kernel/setup.c      |   7 +
 arch/s390/lib/Makefile        |   3 +-
 arch/s390/lib/uaccess.S       | 211 --------------------------
 arch/s390/lib/uaccess64.S     | 207 -------------------------
 arch/s390/lib/uaccess_std.c   | 340 ++++++++++++++++++++++++++++++++++++++++++
 include/asm-s390/futex.h      |  87 +----------
 include/asm-s390/uaccess.h    | 171 +++++++--------------
 8 files changed, 411 insertions(+), 621 deletions(-)
 delete mode 100644 arch/s390/lib/uaccess.S
 delete mode 100644 arch/s390/lib/uaccess64.S
 create mode 100644 arch/s390/lib/uaccess_std.c

(limited to 'include')

diff --git a/arch/s390/kernel/s390_ksyms.c b/arch/s390/kernel/s390_ksyms.c
index c73a45467fa4..9f19e833a562 100644
--- a/arch/s390/kernel/s390_ksyms.c
+++ b/arch/s390/kernel/s390_ksyms.c
@@ -25,12 +25,6 @@ EXPORT_SYMBOL(_oi_bitmap);
 EXPORT_SYMBOL(_ni_bitmap);
 EXPORT_SYMBOL(_zb_findmap);
 EXPORT_SYMBOL(_sb_findmap);
-EXPORT_SYMBOL(__copy_from_user_asm);
-EXPORT_SYMBOL(__copy_to_user_asm);
-EXPORT_SYMBOL(__copy_in_user_asm);
-EXPORT_SYMBOL(__clear_user_asm);
-EXPORT_SYMBOL(__strncpy_from_user_asm);
-EXPORT_SYMBOL(__strnlen_user_asm);
 EXPORT_SYMBOL(diag10);
 
 /*
diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c
index f2a9165ca4f8..e229af59976c 100644
--- a/arch/s390/kernel/setup.c
+++ b/arch/s390/kernel/setup.c
@@ -50,6 +50,12 @@
 #include <asm/ptrace.h>
 #include <asm/sections.h>
 
+/*
+ * User copy operations.
+ */
+struct uaccess_ops uaccess;
+EXPORT_SYMBOL_GPL(uaccess);
+
 /*
  * Machine setup..
  */
@@ -641,6 +647,7 @@ setup_arch(char **cmdline_p)
 
 	memory_end = memory_size;
 
+	memcpy(&uaccess, &uaccess_std, sizeof(uaccess));
 	parse_early_param();
 
 #ifndef CONFIG_64BIT
diff --git a/arch/s390/lib/Makefile b/arch/s390/lib/Makefile
index e05d087a6eae..96c82424d88b 100644
--- a/arch/s390/lib/Makefile
+++ b/arch/s390/lib/Makefile
@@ -4,6 +4,5 @@
 
 EXTRA_AFLAGS := -traditional
 
-lib-y += delay.o string.o
-lib-y += $(if $(CONFIG_64BIT),uaccess64.o,uaccess.o)
+lib-y += delay.o string.o uaccess_std.o
 lib-$(CONFIG_SMP) += spinlock.o
diff --git a/arch/s390/lib/uaccess.S b/arch/s390/lib/uaccess.S
deleted file mode 100644
index 837275284d9f..000000000000
--- a/arch/s390/lib/uaccess.S
+++ /dev/null
@@ -1,211 +0,0 @@
-/*
- *  arch/s390/lib/uaccess.S
- *    __copy_{from|to}_user functions.
- *
- *  s390
- *    Copyright (C) 2000,2002 IBM Deutschland Entwicklung GmbH, IBM Corporation
- *    Authors(s): Martin Schwidefsky (schwidefsky@de.ibm.com)
- *
- *  These functions have standard call interface
- */
-
-#include <linux/errno.h>
-#include <asm/lowcore.h>
-#include <asm/asm-offsets.h>
-
-        .text
-        .align 4
-        .globl __copy_from_user_asm
-	# %r2 = to, %r3 = n, %r4 = from
-__copy_from_user_asm:
-	slr	%r0,%r0
-0:	mvcp	0(%r3,%r2),0(%r4),%r0
-	jnz	1f
-	slr	%r2,%r2
-	br	%r14
-1:	la	%r2,256(%r2)
-	la	%r4,256(%r4)
-	ahi	%r3,-256
-2:	mvcp	0(%r3,%r2),0(%r4),%r0
-	jnz	1b
-3:	slr	%r2,%r2
-	br	%r14
-4:	lhi	%r0,-4096
-	lr	%r5,%r4
-	slr	%r5,%r0
-	nr	%r5,%r0		# %r5 = (%r4 + 4096) & -4096
-	slr	%r5,%r4		# %r5 = #bytes to next user page boundary
-	clr	%r3,%r5		# copy crosses next page boundary ?
-	jnh	6f		# no, the current page faulted
-	# move with the reduced length which is < 256
-5:	mvcp	0(%r5,%r2),0(%r4),%r0
-	slr	%r3,%r5
-6:	lr	%r2,%r3
-	br	%r14
-        .section __ex_table,"a"
-	.long	0b,4b
-	.long	2b,4b
-	.long	5b,6b
-        .previous
-
-        .align 4
-        .text
-        .globl __copy_to_user_asm
-	# %r2 = from, %r3 = n, %r4 = to
-__copy_to_user_asm:
-	slr	%r0,%r0
-0:	mvcs	0(%r3,%r4),0(%r2),%r0
-	jnz	1f
-	slr	%r2,%r2
-	br	%r14
-1:	la	%r2,256(%r2)
-	la	%r4,256(%r4)
-	ahi	%r3,-256
-2:	mvcs	0(%r3,%r4),0(%r2),%r0
-	jnz	1b
-3:	slr	%r2,%r2
-	br	%r14
-4:	lhi	%r0,-4096
-	lr	%r5,%r4
-	slr	%r5,%r0
-	nr	%r5,%r0		# %r5 = (%r4 + 4096) & -4096
-	slr	%r5,%r4		# %r5 = #bytes to next user page boundary
-	clr	%r3,%r5		# copy crosses next page boundary ?
-	jnh	6f		# no, the current page faulted
-	# move with the reduced length which is < 256
-5:	mvcs	0(%r5,%r4),0(%r2),%r0
-	slr	%r3,%r5
-6:	lr	%r2,%r3
-	br	%r14
-        .section __ex_table,"a"
-	.long	0b,4b
-	.long	2b,4b
-	.long	5b,6b
-        .previous
-
-        .align 4
-        .text
-        .globl __copy_in_user_asm
-	# %r2 = from, %r3 = n, %r4 = to
-__copy_in_user_asm:
-	ahi	%r3,-1
-	jo	6f
-	sacf	256
-	bras	%r1,4f
-0:	ahi	%r3,257
-1:	mvc	0(1,%r4),0(%r2)
-	la	%r2,1(%r2)
-	la	%r4,1(%r4)
-	ahi	%r3,-1
-	jnz	1b
-2:	lr	%r2,%r3
-	br	%r14
-3:	mvc	0(256,%r4),0(%r2)
-	la	%r2,256(%r2)
-	la	%r4,256(%r4)
-4:	ahi	%r3,-256
-	jnm	3b
-5:	ex	%r3,4(%r1)
-	sacf	0
-6:	slr	%r2,%r2
-	br	%r14
-        .section __ex_table,"a"
-	.long	1b,2b
-	.long	3b,0b
-	.long	5b,0b
-        .previous
-
-        .align 4
-        .text
-        .globl __clear_user_asm
-	# %r2 = to, %r3 = n
-__clear_user_asm:
-	bras	%r5,0f
-	.long	empty_zero_page
-0:	l	%r5,0(%r5)
-	slr	%r0,%r0
-1:	mvcs	0(%r3,%r2),0(%r5),%r0
-	jnz	2f
-	slr	%r2,%r2
-	br	%r14
-2:	la	%r2,256(%r2)
-	ahi	%r3,-256
-3:	mvcs	0(%r3,%r2),0(%r5),%r0
-	jnz	2b
-4:	slr	%r2,%r2
-	br	%r14
-5:	lhi	%r0,-4096
-	lr	%r4,%r2
-	slr	%r4,%r0
-	nr	%r4,%r0		# %r4 = (%r2 + 4096) & -4096
-	slr	%r4,%r2		# %r4 = #bytes to next user page boundary
-	clr	%r3,%r4		# clear crosses next page boundary ?
-	jnh	7f		# no, the current page faulted
-	# clear with the reduced length which is < 256
-6:	mvcs	0(%r4,%r2),0(%r5),%r0
-	slr	%r3,%r4
-7:	lr	%r2,%r3
-	br	%r14
-        .section __ex_table,"a"
-	.long	1b,5b
-	.long	3b,5b
-	.long	6b,7b
-        .previous
-
-        .align 4
-        .text
-        .globl __strncpy_from_user_asm
-	# %r2 = count, %r3 = dst, %r4 = src
-__strncpy_from_user_asm:
-	lhi	%r0,0
-	lr	%r1,%r4
-	la	%r4,0(%r4)	# clear high order bit from %r4
-	la	%r2,0(%r2,%r4)	# %r2 points to first byte after string
-	sacf	256
-0:	srst	%r2,%r1
-	jo	0b
-	sacf	0
-	lr	%r1,%r2
-	jh	1f		# \0 found in string ?
-	ahi	%r1,1		# include \0 in copy
-1:	slr	%r1,%r4		# %r1 = copy length (without \0)
-	slr	%r2,%r4		# %r2 = return length (including \0)
-2:	mvcp	0(%r1,%r3),0(%r4),%r0
-	jnz	3f
-	br	%r14
-3:	la	%r3,256(%r3)
-	la	%r4,256(%r4)
-	ahi	%r1,-256
-	mvcp	0(%r1,%r3),0(%r4),%r0
-	jnz	3b
-	br	%r14
-4:	sacf	0
-	lhi	%r2,-EFAULT
-	br	%r14
-	.section __ex_table,"a"
-	.long	0b,4b
-	.previous
-
-        .align 4
-        .text
-        .globl __strnlen_user_asm
-	# %r2 = count, %r3 = src
-__strnlen_user_asm:
-	lhi	%r0,0
-	lr	%r1,%r3
-	la	%r3,0(%r3)	# clear high order bit from %r4
-	la	%r2,0(%r2,%r3)	# %r2 points to first byte after string
-	sacf	256
-0:	srst	%r2,%r1
-	jo	0b
-	sacf	0
-	ahi	%r2,1		# strnlen_user result includes the \0
-				# or return count+1 if \0 not found
-	slr	%r2,%r3
-	br	%r14
-2:	sacf	0
-	slr	%r2,%r2		# return 0 on exception
-	br	%r14
-	.section __ex_table,"a"
-	.long	0b,2b
-	.previous
diff --git a/arch/s390/lib/uaccess64.S b/arch/s390/lib/uaccess64.S
deleted file mode 100644
index 1f755be22f92..000000000000
--- a/arch/s390/lib/uaccess64.S
+++ /dev/null
@@ -1,207 +0,0 @@
-/*
- *  arch/s390x/lib/uaccess.S
- *    __copy_{from|to}_user functions.
- *
- *  s390
- *    Copyright (C) 2000,2002 IBM Deutschland Entwicklung GmbH, IBM Corporation
- *    Authors(s): Martin Schwidefsky (schwidefsky@de.ibm.com)
- *
- *  These functions have standard call interface
- */
-
-#include <linux/errno.h>
-#include <asm/lowcore.h>
-#include <asm/asm-offsets.h>
-
-        .text
-        .align 4
-        .globl __copy_from_user_asm
-	# %r2 = to, %r3 = n, %r4 = from
-__copy_from_user_asm:
-	slgr	%r0,%r0
-0:	mvcp	0(%r3,%r2),0(%r4),%r0
-	jnz	1f
-	slgr	%r2,%r2
-	br	%r14
-1:	la	%r2,256(%r2)
-	la	%r4,256(%r4)
-	aghi	%r3,-256
-2:	mvcp	0(%r3,%r2),0(%r4),%r0
-	jnz	1b
-3:	slgr	%r2,%r2
-	br	%r14
-4:	lghi	%r0,-4096
-	lgr	%r5,%r4
-	slgr	%r5,%r0
-	ngr	%r5,%r0		# %r5 = (%r4 + 4096) & -4096
-	slgr	%r5,%r4		# %r5 = #bytes to next user page boundary
-	clgr	%r3,%r5		# copy crosses next page boundary ?
-	jnh	6f		# no, the current page faulted
-	# move with the reduced length which is < 256
-5:	mvcp	0(%r5,%r2),0(%r4),%r0
-	slgr	%r3,%r5
-6:	lgr	%r2,%r3
-	br	%r14
-        .section __ex_table,"a"
-	.quad	0b,4b
-	.quad	2b,4b
-	.quad	5b,6b
-        .previous
-
-        .align 4
-        .text
-        .globl __copy_to_user_asm
-	# %r2 = from, %r3 = n, %r4 = to
-__copy_to_user_asm:
-	slgr	%r0,%r0
-0:	mvcs	0(%r3,%r4),0(%r2),%r0
-	jnz	1f
-	slgr	%r2,%r2
-	br	%r14
-1:	la	%r2,256(%r2)
-	la	%r4,256(%r4)
-	aghi	%r3,-256
-2:	mvcs	0(%r3,%r4),0(%r2),%r0
-	jnz	1b
-3:	slgr	%r2,%r2
-	br	%r14
-4:	lghi	%r0,-4096
-	lgr	%r5,%r4
-	slgr	%r5,%r0
-	ngr	%r5,%r0		# %r5 = (%r4 + 4096) & -4096
-	slgr	%r5,%r4		# %r5 = #bytes to next user page boundary
-	clgr	%r3,%r5		# copy crosses next page boundary ?
-	jnh	6f		# no, the current page faulted
-	# move with the reduced length which is < 256
-5:	mvcs	0(%r5,%r4),0(%r2),%r0
-	slgr	%r3,%r5
-6:	lgr	%r2,%r3
-	br	%r14
-        .section __ex_table,"a"
-	.quad	0b,4b
-	.quad	2b,4b
-	.quad	5b,6b
-        .previous
-
-        .align 4
-        .text
-        .globl __copy_in_user_asm
-	# %r2 = from, %r3 = n, %r4 = to
-__copy_in_user_asm:
-	aghi	%r3,-1
-	jo	6f
-	sacf	256
-	bras	%r1,4f
-0:	aghi	%r3,257
-1:	mvc	0(1,%r4),0(%r2)
-	la	%r2,1(%r2)
-	la	%r4,1(%r4)
-	aghi	%r3,-1
-	jnz	1b
-2:	lgr	%r2,%r3
-	br	%r14
-3:	mvc	0(256,%r4),0(%r2)
-	la	%r2,256(%r2)
-	la	%r4,256(%r4)
-4:	aghi	%r3,-256
-	jnm	3b
-5:	ex	%r3,4(%r1)
-	sacf	0
-6:	slgr	%r2,%r2
-	br	14
-        .section __ex_table,"a"
-	.quad	1b,2b
-	.quad	3b,0b
-	.quad	5b,0b
-        .previous
-
-        .align 4
-        .text
-        .globl __clear_user_asm
-	# %r2 = to, %r3 = n
-__clear_user_asm:
-	slgr	%r0,%r0
-	larl	%r5,empty_zero_page
-1:	mvcs	0(%r3,%r2),0(%r5),%r0
-	jnz	2f
-	slgr	%r2,%r2
-	br	%r14
-2:	la	%r2,256(%r2)
-	aghi	%r3,-256
-3:	mvcs	0(%r3,%r2),0(%r5),%r0
-	jnz	2b
-4:	slgr	%r2,%r2
-	br	%r14
-5:	lghi	%r0,-4096
-	lgr	%r4,%r2
-	slgr	%r4,%r0
-	ngr	%r4,%r0		# %r4 = (%r2 + 4096) & -4096
-	slgr	%r4,%r2		# %r4 = #bytes to next user page boundary
-	clgr	%r3,%r4		# clear crosses next page boundary ?
-	jnh	7f		# no, the current page faulted
-	# clear with the reduced length which is < 256
-6:	mvcs	0(%r4,%r2),0(%r5),%r0
-	slgr	%r3,%r4
-7:	lgr	%r2,%r3
-	br	%r14
-        .section __ex_table,"a"
-	.quad	1b,5b
-	.quad	3b,5b
-	.quad	6b,7b
-        .previous
-
-        .align 4
-        .text
-        .globl __strncpy_from_user_asm
-	# %r2 = count, %r3 = dst, %r4 = src
-__strncpy_from_user_asm:
-	lghi	%r0,0
-	lgr	%r1,%r4
-	la	%r2,0(%r2,%r4)	# %r2 points to first byte after string
-	sacf	256
-0:	srst	%r2,%r1
-	jo	0b
-	sacf	0
-	lgr	%r1,%r2
-	jh	1f		# \0 found in string ?
-	aghi	%r1,1		# include \0 in copy
-1:	slgr	%r1,%r4		# %r1 = copy length (without \0)
-	slgr	%r2,%r4		# %r2 = return length (including \0)
-2:	mvcp	0(%r1,%r3),0(%r4),%r0
-	jnz	3f
-	br	%r14
-3:	la	%r3,256(%r3)
-	la	%r4,256(%r4)
-	aghi	%r1,-256
-	mvcp	0(%r1,%r3),0(%r4),%r0
-	jnz	3b
-	br	%r14
-4:	sacf	0
-	lghi	%r2,-EFAULT
-	br	%r14
-	.section __ex_table,"a"
-	.quad	0b,4b
-	.previous
-
-        .align 4
-        .text
-        .globl __strnlen_user_asm
-	# %r2 = count, %r3 = src
-__strnlen_user_asm:
-	lghi	%r0,0
-	lgr	%r1,%r3
-	la	%r2,0(%r2,%r3)	# %r2 points to first byte after string
-	sacf	256
-0:	srst	%r2,%r1
-	jo	0b
-	sacf	0
-	aghi	%r2,1		# strnlen_user result includes the \0
-				# or return count+1 if \0 not found
-	slgr	%r2,%r3
-	br	%r14
-2:	sacf	0
-	slgr	%r2,%r2		# return 0 on exception
-	br	%r14
-	.section __ex_table,"a"
-	.quad	0b,2b
-	.previous
diff --git a/arch/s390/lib/uaccess_std.c b/arch/s390/lib/uaccess_std.c
new file mode 100644
index 000000000000..9a4d4a29ea79
--- /dev/null
+++ b/arch/s390/lib/uaccess_std.c
@@ -0,0 +1,340 @@
+/*
+ *  arch/s390/lib/uaccess_std.c
+ *
+ *  Standard user space access functions based on mvcp/mvcs and doing
+ *  interesting things in the secondary space mode.
+ *
+ *    Copyright (C) IBM Corp. 2006
+ *    Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com),
+ *		 Gerald Schaefer (gerald.schaefer@de.ibm.com)
+ */
+
+#include <linux/errno.h>
+#include <linux/mm.h>
+#include <asm/uaccess.h>
+#include <asm/futex.h>
+
+#ifndef __s390x__
+#define AHI	"ahi"
+#define ALR	"alr"
+#define CLR	"clr"
+#define LHI	"lhi"
+#define SLR	"slr"
+#else
+#define AHI	"aghi"
+#define ALR	"algr"
+#define CLR	"clgr"
+#define LHI	"lghi"
+#define SLR	"slgr"
+#endif
+
+size_t copy_from_user_std(size_t size, const void __user *ptr, void *x)
+{
+	unsigned long tmp1, tmp2;
+
+	tmp1 = -256UL;
+	asm volatile(
+		"0: mvcp  0(%0,%2),0(%1),%3\n"
+		"   jz    5f\n"
+		"1:"ALR"  %0,%3\n"
+		"   la    %1,256(%1)\n"
+		"   la    %2,256(%2)\n"
+		"2: mvcp  0(%0,%2),0(%1),%3\n"
+		"   jnz   1b\n"
+		"   j     5f\n"
+		"3: la    %4,255(%1)\n"	/* %4 = ptr + 255 */
+		"  "LHI"  %3,-4096\n"
+		"   nr    %4,%3\n"	/* %4 = (ptr + 255) & -4096 */
+		"  "SLR"  %4,%1\n"
+		"  "CLR"  %0,%4\n"	/* copy crosses next page boundary? */
+		"   jnh   6f\n"
+		"4: mvcp  0(%4,%2),0(%1),%3\n"
+		"  "SLR"  %0,%4\n"
+		"   j     6f\n"
+		"5:"SLR"  %0,%0\n"
+		"6: \n"
+		EX_TABLE(0b,3b) EX_TABLE(2b,3b) EX_TABLE(4b,6b)
+		: "+a" (size), "+a" (ptr), "+a" (x), "+a" (tmp1), "=a" (tmp2)
+		: : "cc", "memory");
+	return size;
+}
+
+size_t copy_from_user_std_small(size_t size, const void __user *ptr, void *x)
+{
+	unsigned long tmp1, tmp2;
+
+	tmp1 = 0UL;
+	asm volatile(
+		"0: mvcp  0(%0,%2),0(%1),%3\n"
+		"  "SLR"  %0,%0\n"
+		"   j     3f\n"
+		"1: la    %4,255(%1)\n" /* %4 = ptr + 255 */
+		"  "LHI"  %3,-4096\n"
+		"   nr    %4,%3\n"	/* %4 = (ptr + 255) & -4096 */
+		"  "SLR"  %4,%1\n"
+		"  "CLR"  %0,%4\n"	/* copy crosses next page boundary? */
+		"   jnh   3f\n"
+		"2: mvcp  0(%4,%2),0(%1),%3\n"
+		"  "SLR"  %0,%4\n"
+		"3:\n"
+		EX_TABLE(0b,1b) EX_TABLE(2b,3b)
+		: "+a" (size), "+a" (ptr), "+a" (x), "+a" (tmp1), "=a" (tmp2)
+		: : "cc", "memory");
+	return size;
+}
+
+size_t copy_to_user_std(size_t size, void __user *ptr, const void *x)
+{
+	unsigned long tmp1, tmp2;
+
+	tmp1 = -256UL;
+	asm volatile(
+		"0: mvcs  0(%0,%1),0(%2),%3\n"
+		"   jz    5f\n"
+		"1:"ALR"  %0,%3\n"
+		"   la    %1,256(%1)\n"
+		"   la    %2,256(%2)\n"
+		"2: mvcs  0(%0,%1),0(%2),%3\n"
+		"   jnz   1b\n"
+		"   j     5f\n"
+		"3: la    %4,255(%1)\n" /* %4 = ptr + 255 */
+		"  "LHI"  %3,-4096\n"
+		"   nr    %4,%3\n"	/* %4 = (ptr + 255) & -4096 */
+		"  "SLR"  %4,%1\n"
+		"  "CLR"  %0,%4\n"	/* copy crosses next page boundary? */
+		"   jnh   6f\n"
+		"4: mvcs  0(%4,%1),0(%2),%3\n"
+		"  "SLR"  %0,%4\n"
+		"   j     6f\n"
+		"5:"SLR"  %0,%0\n"
+		"6: \n"
+		EX_TABLE(0b,3b) EX_TABLE(2b,3b) EX_TABLE(4b,6b)
+		: "+a" (size), "+a" (ptr), "+a" (x), "+a" (tmp1), "=a" (tmp2)
+		: : "cc", "memory");
+	return size;
+}
+
+size_t copy_to_user_std_small(size_t size, void __user *ptr, const void *x)
+{
+	unsigned long tmp1, tmp2;
+
+	tmp1 = 0UL;
+	asm volatile(
+		"0: mvcs  0(%0,%1),0(%2),%3\n"
+		"  "SLR"  %0,%0\n"
+		"   j     3f\n"
+		"1: la    %4,255(%1)\n" /* ptr + 255 */
+		"  "LHI"  %3,-4096\n"
+		"   nr    %4,%3\n"	/* (ptr + 255) & -4096UL */
+		"  "SLR"  %4,%1\n"
+		"  "CLR"  %0,%4\n"	/* copy crosses next page boundary? */
+		"   jnh   3f\n"
+		"2: mvcs  0(%4,%1),0(%2),%3\n"
+		"  "SLR"  %0,%4\n"
+		"3:\n"
+		EX_TABLE(0b,1b) EX_TABLE(2b,3b)
+		: "+a" (size), "+a" (ptr), "+a" (x), "+a" (tmp1), "=a" (tmp2)
+		: : "cc", "memory");
+	return size;
+}
+
+size_t copy_in_user_std(size_t size, void __user *to, const void __user *from)
+{
+	unsigned long tmp1;
+
+	asm volatile(
+		"  "AHI"  %0,-1\n"
+		"   jo    5f\n"
+		"   sacf  256\n"
+		"   bras  %3,3f\n"
+		"0:"AHI"  %0,257\n"
+		"1: mvc   0(1,%1),0(%2)\n"
+		"   la    %1,1(%1)\n"
+		"   la    %2,1(%2)\n"
+		"  "AHI"  %0,-1\n"
+		"   jnz   1b\n"
+		"   j     5f\n"
+		"2: mvc   0(256,%1),0(%2)\n"
+		"   la    %1,256(%1)\n"
+		"   la    %2,256(%2)\n"
+		"3:"AHI"  %0,-256\n"
+		"   jnm   2b\n"
+		"4: ex    %0,1b-0b(%3)\n"
+		"   sacf  0\n"
+		"5: "SLR"  %0,%0\n"
+		"6:\n"
+		EX_TABLE(1b,6b) EX_TABLE(2b,0b) EX_TABLE(4b,0b)
+		: "+a" (size), "+a" (to), "+a" (from), "=a" (tmp1)
+		: : "cc", "memory");
+	return size;
+}
+
+size_t clear_user_std(size_t size, void __user *to)
+{
+	unsigned long tmp1, tmp2;
+
+	asm volatile(
+		"  "AHI"  %0,-1\n"
+		"   jo    5f\n"
+		"   sacf  256\n"
+		"   bras  %3,3f\n"
+		"   xc    0(1,%1),0(%1)\n"
+		"0:"AHI"  %0,257\n"
+		"   la    %2,255(%1)\n" /* %2 = ptr + 255 */
+		"   srl   %2,12\n"
+		"   sll   %2,12\n"	/* %2 = (ptr + 255) & -4096 */
+		"  "SLR"  %2,%1\n"
+		"  "CLR"  %0,%2\n"	/* clear crosses next page boundary? */
+		"   jnh   5f\n"
+		"  "AHI"  %2,-1\n"
+		"1: ex    %2,0(%3)\n"
+		"  "AHI"  %2,1\n"
+		"  "SLR"  %0,%2\n"
+		"   j     5f\n"
+		"2: xc    0(256,%1),0(%1)\n"
+		"   la    %1,256(%1)\n"
+		"3:"AHI"  %0,-256\n"
+		"   jnm   2b\n"
+		"4: ex    %0,0(%3)\n"
+		"   sacf  0\n"
+		"5: "SLR"  %0,%0\n"
+		"6:\n"
+		EX_TABLE(1b,6b) EX_TABLE(2b,0b) EX_TABLE(4b,0b)
+		: "+a" (size), "+a" (to), "=a" (tmp1), "=a" (tmp2)
+		: : "cc", "memory");
+	return size;
+}
+
+size_t strnlen_user_std(size_t size, const char __user *src)
+{
+	register unsigned long reg0 asm("0") = 0UL;
+	unsigned long tmp1, tmp2;
+
+	asm volatile(
+		"   la    %2,0(%1)\n"
+		"   la    %3,0(%0,%1)\n"
+		"  "SLR"  %0,%0\n"
+		"   sacf  256\n"
+		"0: srst  %3,%2\n"
+		"   jo    0b\n"
+		"   la    %0,1(%3)\n"	/* strnlen_user results includes \0 */
+		"  "SLR"  %0,%1\n"
+		"1: sacf  0\n"
+		EX_TABLE(0b,1b)
+		: "+a" (size), "+a" (src), "=a" (tmp1), "=a" (tmp2)
+		: "d" (reg0) : "cc", "memory");
+	return size;
+}
+
+size_t strncpy_from_user_std(size_t size, const char __user *src, char *dst)
+{
+	register unsigned long reg0 asm("0") = 0UL;
+	unsigned long tmp1, tmp2;
+
+	asm volatile(
+		"   la    %3,0(%1)\n"
+		"   la    %4,0(%0,%1)\n"
+		"   sacf  256\n"
+		"0: srst  %4,%3\n"
+		"   jo    0b\n"
+		"   sacf  0\n"
+		"   la    %0,0(%4)\n"
+		"   jh    1f\n"		/* found \0 in string ? */
+		"  "AHI"  %4,1\n"	/* include \0 in copy */
+		"1:"SLR"  %0,%1\n"	/* %0 = return length (without \0) */
+		"  "SLR"  %4,%1\n"	/* %4 = copy length (including \0) */
+		"2: mvcp  0(%4,%2),0(%1),%5\n"
+		"   jz    9f\n"
+		"3:"AHI"  %4,-256\n"
+		"   la    %1,256(%1)\n"
+		"   la    %2,256(%2)\n"
+		"4: mvcp  0(%4,%2),0(%1),%5\n"
+		"   jnz   3b\n"
+		"   j     9f\n"
+		"7: sacf  0\n"
+		"8:"LHI"  %0,%6\n"
+		"9:\n"
+		EX_TABLE(0b,7b) EX_TABLE(2b,8b) EX_TABLE(4b,8b)
+		: "+a" (size), "+a" (src), "+d" (dst), "=a" (tmp1), "=a" (tmp2)
+		: "d" (reg0), "K" (-EFAULT) : "cc", "memory");
+	return size;
+}
+
+#define __futex_atomic_op(insn, ret, oldval, newval, uaddr, oparg)	\
+	asm volatile(							\
+		"   sacf  256\n"					\
+		"0: l     %1,0(%6)\n"					\
+		"1:"insn						\
+		"2: cs    %1,%2,0(%6)\n"				\
+		"3: jl    1b\n"						\
+		"   lhi   %0,0\n"					\
+		"4: sacf  0\n"						\
+		EX_TABLE(0b,4b) EX_TABLE(2b,4b) EX_TABLE(3b,4b)		\
+		: "=d" (ret), "=&d" (oldval), "=&d" (newval),		\
+		  "=m" (*uaddr)						\
+		: "0" (-EFAULT), "d" (oparg), "a" (uaddr),		\
+		  "m" (*uaddr) : "cc");
+
+int futex_atomic_op(int op, int __user *uaddr, int oparg, int *old)
+{
+	int oldval = 0, newval, ret;
+
+	inc_preempt_count();
+
+	switch (op) {
+	case FUTEX_OP_SET:
+		__futex_atomic_op("lr %2,%5\n",
+				  ret, oldval, newval, uaddr, oparg);
+		break;
+	case FUTEX_OP_ADD:
+		__futex_atomic_op("lr %2,%1\nar %2,%5\n",
+				  ret, oldval, newval, uaddr, oparg);
+		break;
+	case FUTEX_OP_OR:
+		__futex_atomic_op("lr %2,%1\nor %2,%5\n",
+				  ret, oldval, newval, uaddr, oparg);
+		break;
+	case FUTEX_OP_ANDN:
+		__futex_atomic_op("lr %2,%1\nnr %2,%5\n",
+				  ret, oldval, newval, uaddr, oparg);
+		break;
+	case FUTEX_OP_XOR:
+		__futex_atomic_op("lr %2,%1\nxr %2,%5\n",
+				  ret, oldval, newval, uaddr, oparg);
+		break;
+	default:
+		ret = -ENOSYS;
+	}
+	dec_preempt_count();
+	*old = oldval;
+	return ret;
+}
+
+int futex_atomic_cmpxchg(int __user *uaddr, int oldval, int newval)
+{
+	int ret;
+
+	asm volatile(
+		"   sacf 256\n"
+		"   cs   %1,%4,0(%5)\n"
+		"0: lr   %0,%1\n"
+		"1: sacf 0\n"
+		EX_TABLE(0b,1b)
+		: "=d" (ret), "+d" (oldval), "=m" (*uaddr)
+		: "0" (-EFAULT), "d" (newval), "a" (uaddr), "m" (*uaddr)
+		: "cc", "memory" );
+	return ret;
+}
+
+struct uaccess_ops uaccess_std = {
+	.copy_from_user = copy_from_user_std,
+	.copy_from_user_small = copy_from_user_std_small,
+	.copy_to_user = copy_to_user_std,
+	.copy_to_user_small = copy_to_user_std_small,
+	.copy_in_user = copy_in_user_std,
+	.clear_user = clear_user_std,
+	.strnlen_user = strnlen_user_std,
+	.strncpy_from_user = strncpy_from_user_std,
+	.futex_atomic_op = futex_atomic_op,
+	.futex_atomic_cmpxchg = futex_atomic_cmpxchg,
+};
diff --git a/include/asm-s390/futex.h b/include/asm-s390/futex.h
index ffedf14f89f6..5e261e1de671 100644
--- a/include/asm-s390/futex.h
+++ b/include/asm-s390/futex.h
@@ -7,75 +7,21 @@
 #include <asm/errno.h>
 #include <asm/uaccess.h>
 
-#ifndef __s390x__
-#define __futex_atomic_fixup \
-		     ".section __ex_table,\"a\"\n"			\
-		     "   .align 4\n"					\
-		     "   .long  0b,4b,2b,4b,3b,4b\n"			\
-		     ".previous"
-#else /* __s390x__ */
-#define __futex_atomic_fixup \
-		     ".section __ex_table,\"a\"\n"			\
-		     "   .align 8\n"					\
-		     "   .quad  0b,4b,2b,4b,3b,4b\n"			\
-		     ".previous"
-#endif /* __s390x__ */
-
-#define __futex_atomic_op(insn, ret, oldval, newval, uaddr, oparg)	\
-	asm volatile("   sacf 256\n"					\
-		     "0: l   %1,0(%6)\n"				\
-		     "1: " insn						\
-		     "2: cs  %1,%2,0(%6)\n"				\
-		     "3: jl  1b\n"					\
-		     "   lhi %0,0\n"					\
-		     "4: sacf 0\n"					\
-		     __futex_atomic_fixup				\
-		     : "=d" (ret), "=&d" (oldval), "=&d" (newval),	\
-		       "=m" (*uaddr)					\
-		     : "0" (-EFAULT), "d" (oparg), "a" (uaddr),		\
-		       "m" (*uaddr) : "cc" );
-
 static inline int futex_atomic_op_inuser (int encoded_op, int __user *uaddr)
 {
 	int op = (encoded_op >> 28) & 7;
 	int cmp = (encoded_op >> 24) & 15;
 	int oparg = (encoded_op << 8) >> 20;
 	int cmparg = (encoded_op << 20) >> 20;
-	int oldval = 0, newval, ret;
+	int oldval, ret;
+
 	if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28))
 		oparg = 1 << oparg;
 
 	if (! access_ok (VERIFY_WRITE, uaddr, sizeof(int)))
 		return -EFAULT;
 
-	inc_preempt_count();
-
-	switch (op) {
-	case FUTEX_OP_SET:
-		__futex_atomic_op("lr %2,%5\n",
-				  ret, oldval, newval, uaddr, oparg);
-		break;
-	case FUTEX_OP_ADD:
-		__futex_atomic_op("lr %2,%1\nar %2,%5\n",
-				  ret, oldval, newval, uaddr, oparg);
-		break;
-	case FUTEX_OP_OR:
-		__futex_atomic_op("lr %2,%1\nor %2,%5\n",
-				  ret, oldval, newval, uaddr, oparg);
-		break;
-	case FUTEX_OP_ANDN:
-		__futex_atomic_op("lr %2,%1\nnr %2,%5\n",
-				  ret, oldval, newval, uaddr, oparg);
-		break;
-	case FUTEX_OP_XOR:
-		__futex_atomic_op("lr %2,%1\nxr %2,%5\n",
-				  ret, oldval, newval, uaddr, oparg);
-		break;
-	default:
-		ret = -ENOSYS;
-	}
-
-	dec_preempt_count();
+	ret = uaccess.futex_atomic_op(op, uaddr, oparg, &oldval);
 
 	if (!ret) {
 		switch (cmp) {
@@ -91,32 +37,13 @@ static inline int futex_atomic_op_inuser (int encoded_op, int __user *uaddr)
 	return ret;
 }
 
-static inline int
-futex_atomic_cmpxchg_inatomic(int __user *uaddr, int oldval, int newval)
+static inline int futex_atomic_cmpxchg_inatomic(int __user *uaddr,
+						int oldval, int newval)
 {
-	int ret;
-
 	if (! access_ok (VERIFY_WRITE, uaddr, sizeof(int)))
 		return -EFAULT;
-	asm volatile("   sacf 256\n"
-		     "   cs   %1,%4,0(%5)\n"
-		     "0: lr   %0,%1\n"
-		     "1: sacf 0\n"
-#ifndef __s390x__
-		     ".section __ex_table,\"a\"\n"
-		     "   .align 4\n"
-		     "   .long  0b,1b\n"
-		     ".previous"
-#else /* __s390x__ */
-		     ".section __ex_table,\"a\"\n"
-		     "   .align 8\n"
-		     "   .quad  0b,1b\n"
-		     ".previous"
-#endif /* __s390x__ */
-		     : "=d" (ret), "+d" (oldval), "=m" (*uaddr)
-		     : "0" (-EFAULT), "d" (newval), "a" (uaddr), "m" (*uaddr)
-		     : "cc", "memory" );
-	return oldval;
+
+	return uaccess.futex_atomic_cmpxchg(uaddr, oldval, newval);
 }
 
 #endif /* __KERNEL__ */
diff --git a/include/asm-s390/uaccess.h b/include/asm-s390/uaccess.h
index 0b7c0ca4c3d7..39a2716ae188 100644
--- a/include/asm-s390/uaccess.h
+++ b/include/asm-s390/uaccess.h
@@ -47,7 +47,7 @@
 		S390_lowcore.user_asce : S390_lowcore.kernel_asce;	\
 	asm volatile ("lctlg 7,7,%0" : : "m" (__pto) );			\
 })
-#else
+#else /* __s390x__ */
 #define set_fs(x) \
 ({									\
 	unsigned long __pto;						\
@@ -56,7 +56,7 @@
 		S390_lowcore.user_asce : S390_lowcore.kernel_asce;	\
 	asm volatile ("lctl  7,7,%0" : : "m" (__pto) );			\
 })
-#endif
+#endif /* __s390x__ */
 
 #define segment_eq(a,b) ((a).ar4 == (b).ar4)
 
@@ -85,76 +85,50 @@ struct exception_table_entry
         unsigned long insn, fixup;
 };
 
-#ifndef __s390x__
-#define __uaccess_fixup \
-	".section .fixup,\"ax\"\n"	\
-	"2: lhi    %0,%4\n"		\
-	"   bras   1,3f\n"		\
-	"   .long  1b\n"		\
-	"3: l      1,0(1)\n"		\
-	"   br     1\n"			\
-	".previous\n"			\
-	".section __ex_table,\"a\"\n"	\
-	"   .align 4\n"			\
-	"   .long  0b,2b\n"		\
-	".previous"
-#define __uaccess_clobber "cc", "1"
-#else /* __s390x__ */
-#define __uaccess_fixup \
-	".section .fixup,\"ax\"\n"	\
-	"2: lghi   %0,%4\n"		\
-	"   jg     1b\n"		\
-	".previous\n"			\
-	".section __ex_table,\"a\"\n"	\
-	"   .align 8\n"			\
-	"   .quad  0b,2b\n"		\
-	".previous"
-#define __uaccess_clobber "cc"
-#endif /* __s390x__ */
+struct uaccess_ops {
+	size_t (*copy_from_user)(size_t, const void __user *, void *);
+	size_t (*copy_from_user_small)(size_t, const void __user *, void *);
+	size_t (*copy_to_user)(size_t, void __user *, const void *);
+	size_t (*copy_to_user_small)(size_t, void __user *, const void *);
+	size_t (*copy_in_user)(size_t, void __user *, const void __user *);
+	size_t (*clear_user)(size_t, void __user *);
+	size_t (*strnlen_user)(size_t, const char __user *);
+	size_t (*strncpy_from_user)(size_t, const char __user *, char *);
+	int (*futex_atomic_op)(int op, int __user *, int oparg, int *old);
+	int (*futex_atomic_cmpxchg)(int __user *, int old, int new);
+};
+
+extern struct uaccess_ops uaccess;
+extern struct uaccess_ops uaccess_std;
+
+static inline int __put_user_fn(size_t size, void __user *ptr, void *x)
+{
+	size = uaccess.copy_to_user_small(size, ptr, x);
+	return size ? -EFAULT : size;
+}
+
+static inline int __get_user_fn(size_t size, const void __user *ptr, void *x)
+{
+	size = uaccess.copy_from_user_small(size, ptr, x);
+	return size ? -EFAULT : size;
+}
 
 /*
  * These are the main single-value transfer routines.  They automatically
  * use the right size if we just have the right pointer type.
  */
-#if __GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ > 2)
-#define __put_user_asm(x, ptr, err) \
-({								\
-	err = 0;						\
-	asm volatile(						\
-		"0: mvcs  0(%1,%2),%3,%0\n"			\
-		"1:\n"						\
-		__uaccess_fixup					\
-		: "+&d" (err)					\
-		: "d" (sizeof(*(ptr))), "a" (ptr), "Q" (x),	\
-		  "K" (-EFAULT)					\
-		: __uaccess_clobber );				\
-})
-#else
-#define __put_user_asm(x, ptr, err) \
-({								\
-	err = 0;						\
-	asm volatile(						\
-		"0: mvcs  0(%1,%2),0(%3),%0\n"			\
-		"1:\n"						\
-		__uaccess_fixup					\
-		: "+&d" (err)					\
-		: "d" (sizeof(*(ptr))), "a" (ptr), "a" (&(x)),	\
-		  "K" (-EFAULT), "m" (x)			\
-		: __uaccess_clobber );				\
-})
-#endif
-
 #define __put_user(x, ptr) \
 ({								\
 	__typeof__(*(ptr)) __x = (x);				\
-	int __pu_err;						\
+	int __pu_err = -EFAULT;					\
         __chk_user_ptr(ptr);                                    \
 	switch (sizeof (*(ptr))) {				\
 	case 1:							\
 	case 2:							\
 	case 4:							\
 	case 8:							\
-		__put_user_asm(__x, ptr, __pu_err);		\
+		__pu_err = __put_user_fn(sizeof (*(ptr)),	\
+					 ptr, &__x);		\
 		break;						\
 	default:						\
 		__put_user_bad();				\
@@ -172,60 +146,36 @@ struct exception_table_entry
 
 extern int __put_user_bad(void) __attribute__((noreturn));
 
-#if __GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ > 2)
-#define __get_user_asm(x, ptr, err) \
-({								\
-	err = 0;						\
-	asm volatile (						\
-		"0: mvcp  %O1(%2,%R1),0(%3),%0\n"		\
-		"1:\n"						\
-		__uaccess_fixup					\
-		: "+&d" (err), "=Q" (x)				\
-		: "d" (sizeof(*(ptr))), "a" (ptr),		\
-		  "K" (-EFAULT)					\
-		: __uaccess_clobber );				\
-})
-#else
-#define __get_user_asm(x, ptr, err) \
-({								\
-	err = 0;						\
-	asm volatile (						\
-		"0: mvcp  0(%2,%5),0(%3),%0\n"			\
-		"1:\n"						\
-		__uaccess_fixup					\
-		: "+&d" (err), "=m" (x)				\
-		: "d" (sizeof(*(ptr))), "a" (ptr),		\
-		  "K" (-EFAULT), "a" (&(x))			\
-		: __uaccess_clobber );				\
-})
-#endif
-
 #define __get_user(x, ptr)					\
 ({								\
-	int __gu_err;						\
-        __chk_user_ptr(ptr);                                    \
+	int __gu_err = -EFAULT;					\
+	__chk_user_ptr(ptr);					\
 	switch (sizeof(*(ptr))) {				\
 	case 1: {						\
 		unsigned char __x;				\
-		__get_user_asm(__x, ptr, __gu_err);		\
+		__gu_err = __get_user_fn(sizeof (*(ptr)),	\
+					 ptr, &__x);		\
 		(x) = *(__force __typeof__(*(ptr)) *) &__x;	\
 		break;						\
 	};							\
 	case 2: {						\
 		unsigned short __x;				\
-		__get_user_asm(__x, ptr, __gu_err);		\
+		__gu_err = __get_user_fn(sizeof (*(ptr)),	\
+					 ptr, &__x);		\
 		(x) = *(__force __typeof__(*(ptr)) *) &__x;	\
 		break;						\
 	};							\
 	case 4: {						\
 		unsigned int __x;				\
-		__get_user_asm(__x, ptr, __gu_err);		\
+		__gu_err = __get_user_fn(sizeof (*(ptr)),	\
+					 ptr, &__x);		\
 		(x) = *(__force __typeof__(*(ptr)) *) &__x;	\
 		break;						\
 	};							\
 	case 8: {						\
 		unsigned long long __x;				\
-		__get_user_asm(__x, ptr, __gu_err);		\
+		__gu_err = __get_user_fn(sizeof (*(ptr)),	\
+					 ptr, &__x);		\
 		(x) = *(__force __typeof__(*(ptr)) *) &__x;	\
 		break;						\
 	};							\
@@ -247,8 +197,6 @@ extern int __get_user_bad(void) __attribute__((noreturn));
 #define __put_user_unaligned __put_user
 #define __get_user_unaligned __get_user
 
-extern long __copy_to_user_asm(const void *from, long n, void __user *to);
-
 /**
  * __copy_to_user: - Copy a block of data into user space, with less checking.
  * @to:   Destination address, in user space.
@@ -266,7 +214,10 @@ extern long __copy_to_user_asm(const void *from, long n, void __user *to);
 static inline unsigned long
 __copy_to_user(void __user *to, const void *from, unsigned long n)
 {
-	return __copy_to_user_asm(from, n, to);
+	if (__builtin_constant_p(n) && (n <= 256))
+		return uaccess.copy_to_user_small(n, to, from);
+	else
+		return uaccess.copy_to_user(n, to, from);
 }
 
 #define __copy_to_user_inatomic __copy_to_user
@@ -294,8 +245,6 @@ copy_to_user(void __user *to, const void *from, unsigned long n)
 	return n;
 }
 
-extern long __copy_from_user_asm(void *to, long n, const void __user *from);
-
 /**
  * __copy_from_user: - Copy a block of data from user space, with less checking.
  * @to:   Destination address, in kernel space.
@@ -316,7 +265,10 @@ extern long __copy_from_user_asm(void *to, long n, const void __user *from);
 static inline unsigned long
 __copy_from_user(void *to, const void __user *from, unsigned long n)
 {
-	return __copy_from_user_asm(to, n, from);
+	if (__builtin_constant_p(n) && (n <= 256))
+		return uaccess.copy_from_user_small(n, from, to);
+	else
+		return uaccess.copy_from_user(n, from, to);
 }
 
 /**
@@ -346,13 +298,10 @@ copy_from_user(void *to, const void __user *from, unsigned long n)
 	return n;
 }
 
-extern unsigned long __copy_in_user_asm(const void __user *from, long n,
-							void __user *to);
-
 static inline unsigned long
 __copy_in_user(void __user *to, const void __user *from, unsigned long n)
 {
-	return __copy_in_user_asm(from, n, to);
+	return uaccess.copy_in_user(n, to, from);
 }
 
 static inline unsigned long
@@ -360,34 +309,28 @@ copy_in_user(void __user *to, const void __user *from, unsigned long n)
 {
 	might_sleep();
 	if (__access_ok(from,n) && __access_ok(to,n))
-		n = __copy_in_user_asm(from, n, to);
+		n = __copy_in_user(to, from, n);
 	return n;
 }
 
 /*
  * Copy a null terminated string from userspace.
  */
-extern long __strncpy_from_user_asm(long count, char *dst,
-					const char __user *src);
-
 static inline long
 strncpy_from_user(char *dst, const char __user *src, long count)
 {
         long res = -EFAULT;
         might_sleep();
         if (access_ok(VERIFY_READ, src, 1))
-                res = __strncpy_from_user_asm(count, dst, src);
+		res = uaccess.strncpy_from_user(count, src, dst);
         return res;
 }
 
-
-extern long __strnlen_user_asm(long count, const char __user *src);
-
 static inline unsigned long
 strnlen_user(const char __user * src, unsigned long n)
 {
 	might_sleep();
-	return __strnlen_user_asm(n, src);
+	return uaccess.strnlen_user(n, src);
 }
 
 /**
@@ -410,12 +353,10 @@ strnlen_user(const char __user * src, unsigned long n)
  * Zero Userspace
  */
 
-extern long __clear_user_asm(void __user *to, long n);
-
 static inline unsigned long
 __clear_user(void __user *to, unsigned long n)
 {
-	return __clear_user_asm(to, n);
+	return uaccess.clear_user(n, to);
 }
 
 static inline unsigned long
@@ -423,7 +364,7 @@ clear_user(void __user *to, unsigned long n)
 {
 	might_sleep();
 	if (access_ok(VERIFY_WRITE, to, n))
-		n = __clear_user_asm(to, n);
+		n = uaccess.clear_user(n, to);
 	return n;
 }
 
-- 
cgit v1.2.3


From 6c2a9e6df60478e712f3c3d98b5047778a82a3d7 Mon Sep 17 00:00:00 2001
From: Gerald Schaefer <geraldsc@de.ibm.com>
Date: Wed, 20 Sep 2006 15:59:44 +0200
Subject: [S390] Use alternative user-copy operations for new hardware.

This introduces new user-copy operations which are optimized for
copying more than 256 Bytes on new hardware.

Signed-off-by: Gerald Schaefer <geraldsc@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/head64.S     |  13 ++++
 arch/s390/kernel/setup.c      |   6 +-
 arch/s390/lib/Makefile        |   1 +
 arch/s390/lib/uaccess_mvcos.c | 156 ++++++++++++++++++++++++++++++++++++++++++
 include/asm-s390/setup.h      |   2 +
 include/asm-s390/uaccess.h    |   1 +
 6 files changed, 178 insertions(+), 1 deletion(-)
 create mode 100644 arch/s390/lib/uaccess_mvcos.c

(limited to 'include')

diff --git a/arch/s390/kernel/head64.S b/arch/s390/kernel/head64.S
index 1ebaa338aa7e..a8bdd96494c7 100644
--- a/arch/s390/kernel/head64.S
+++ b/arch/s390/kernel/head64.S
@@ -250,6 +250,19 @@ startup_continue:
 	oi	7(%r12),0x80		# set IDTE flag
 0:
 
+#
+# find out if we have the MVCOS instruction
+#
+	la	%r1,0f-.LPG1(%r13)	# set program check address
+	stg	%r1,__LC_PGM_NEW_PSW+8
+	.short	0xc800			# mvcos 0(%r0),0(%r0),%r0
+	.short	0x0000
+	.short	0x0000
+0:	tm	0x8f,0x13		# special-operation exception?
+	bno	1f-.LPG1(%r13)		# if yes, MVCOS is present
+	oi	6(%r12),2		# set MVCOS flag
+1:
+
         lpswe .Lentry-.LPG1(13)         # jump to _stext in primary-space,
                                         # virtual and never return ...
         .align 16
diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c
index e229af59976c..e3d9325f6022 100644
--- a/arch/s390/kernel/setup.c
+++ b/arch/s390/kernel/setup.c
@@ -647,7 +647,11 @@ setup_arch(char **cmdline_p)
 
 	memory_end = memory_size;
 
-	memcpy(&uaccess, &uaccess_std, sizeof(uaccess));
+	if (MACHINE_HAS_MVCOS)
+		memcpy(&uaccess, &uaccess_mvcos, sizeof(uaccess));
+	else
+		memcpy(&uaccess, &uaccess_std, sizeof(uaccess));
+
 	parse_early_param();
 
 #ifndef CONFIG_64BIT
diff --git a/arch/s390/lib/Makefile b/arch/s390/lib/Makefile
index 96c82424d88b..c42ffedfdb49 100644
--- a/arch/s390/lib/Makefile
+++ b/arch/s390/lib/Makefile
@@ -5,4 +5,5 @@
 EXTRA_AFLAGS := -traditional
 
 lib-y += delay.o string.o uaccess_std.o
+lib-$(CONFIG_64BIT) += uaccess_mvcos.o
 lib-$(CONFIG_SMP) += spinlock.o
diff --git a/arch/s390/lib/uaccess_mvcos.c b/arch/s390/lib/uaccess_mvcos.c
new file mode 100644
index 000000000000..86c96d6c191a
--- /dev/null
+++ b/arch/s390/lib/uaccess_mvcos.c
@@ -0,0 +1,156 @@
+/*
+ *  arch/s390/lib/uaccess_mvcos.c
+ *
+ *  Optimized user space space access functions based on mvcos.
+ *
+ *    Copyright (C) IBM Corp. 2006
+ *    Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com),
+ *		 Gerald Schaefer (gerald.schaefer@de.ibm.com)
+ */
+
+#include <linux/errno.h>
+#include <linux/mm.h>
+#include <asm/uaccess.h>
+#include <asm/futex.h>
+
+#ifndef __s390x__
+#define AHI	"ahi"
+#define ALR	"alr"
+#define CLR	"clr"
+#define LHI	"lhi"
+#define SLR	"slr"
+#else
+#define AHI	"aghi"
+#define ALR	"algr"
+#define CLR	"clgr"
+#define LHI	"lghi"
+#define SLR	"slgr"
+#endif
+
+size_t copy_from_user_mvcos(size_t size, const void __user *ptr, void *x)
+{
+	register unsigned long reg0 asm("0") = 0x81UL;
+	unsigned long tmp1, tmp2;
+
+	tmp1 = -4096UL;
+	asm volatile(
+		"0: .insn ss,0xc80000000000,0(%0,%2),0(%1),0\n"
+		"   jz    4f\n"
+		"1:"ALR"  %0,%3\n"
+		"  "SLR"  %1,%3\n"
+		"  "SLR"  %2,%3\n"
+		"   j     0b\n"
+		"2: la    %4,4095(%1)\n"/* %4 = ptr + 4095 */
+		"   nr    %4,%3\n"	/* %4 = (ptr + 4095) & -4096 */
+		"  "SLR"  %4,%1\n"
+		"  "CLR"  %0,%4\n"	/* copy crosses next page boundary? */
+		"   jnh   5f\n"
+		"3: .insn ss,0xc80000000000,0(%4,%2),0(%1),0\n"
+		"  "SLR"  %0,%4\n"
+		"   j     5f\n"
+		"4:"SLR"  %0,%0\n"
+		"5: \n"
+		EX_TABLE(0b,2b) EX_TABLE(3b,5b)
+		: "+a" (size), "+a" (ptr), "+a" (x), "+a" (tmp1), "=a" (tmp2)
+		: "d" (reg0) : "cc", "memory");
+	return size;
+}
+
+size_t copy_to_user_mvcos(size_t size, void __user *ptr, const void *x)
+{
+	register unsigned long reg0 asm("0") = 0x810000UL;
+	unsigned long tmp1, tmp2;
+
+	tmp1 = -4096UL;
+	asm volatile(
+		"0: .insn ss,0xc80000000000,0(%0,%1),0(%2),0\n"
+		"   jz    4f\n"
+		"1:"ALR"  %0,%3\n"
+		"  "SLR"  %1,%3\n"
+		"  "SLR"  %2,%3\n"
+		"   j     0b\n"
+		"2: la    %4,4095(%1)\n"/* %4 = ptr + 4095 */
+		"   nr    %4,%3\n"	/* %4 = (ptr + 4095) & -4096 */
+		"  "SLR"  %4,%1\n"
+		"  "CLR"  %0,%4\n"	/* copy crosses next page boundary? */
+		"   jnh   5f\n"
+		"3: .insn ss,0xc80000000000,0(%4,%1),0(%2),0\n"
+		"  "SLR"  %0,%4\n"
+		"   j     5f\n"
+		"4:"SLR"  %0,%0\n"
+		"5: \n"
+		EX_TABLE(0b,2b) EX_TABLE(3b,5b)
+		: "+a" (size), "+a" (ptr), "+a" (x), "+a" (tmp1), "=a" (tmp2)
+		: "d" (reg0) : "cc", "memory");
+	return size;
+}
+
+size_t copy_in_user_mvcos(size_t size, void __user *to, const void __user *from)
+{
+	register unsigned long reg0 asm("0") = 0x810081UL;
+	unsigned long tmp1, tmp2;
+
+	tmp1 = -4096UL;
+	/* FIXME: copy with reduced length. */
+	asm volatile(
+		"0: .insn ss,0xc80000000000,0(%0,%1),0(%2),0\n"
+		"   jz    2f\n"
+		"1:"ALR"  %0,%3\n"
+		"  "SLR"  %1,%3\n"
+		"  "SLR"  %2,%3\n"
+		"   j     0b\n"
+		"2:"SLR"  %0,%0\n"
+		"3: \n"
+		EX_TABLE(0b,3b)
+		: "+a" (size), "+a" (to), "+a" (from), "+a" (tmp1), "=a" (tmp2)
+		: "d" (reg0) : "cc", "memory");
+	return size;
+}
+
+size_t clear_user_mvcos(size_t size, void __user *to)
+{
+	register unsigned long reg0 asm("0") = 0x810000UL;
+	unsigned long tmp1, tmp2;
+
+	tmp1 = -4096UL;
+	asm volatile(
+		"0: .insn ss,0xc80000000000,0(%0,%1),0(%4),0\n"
+		"   jz    4f\n"
+		"1:"ALR"  %0,%2\n"
+		"  "SLR"  %1,%2\n"
+		"   j     0b\n"
+		"2: la    %3,4095(%1)\n"/* %4 = to + 4095 */
+		"   nr    %3,%2\n"	/* %4 = (to + 4095) & -4096 */
+		"  "SLR"  %3,%1\n"
+		"  "CLR"  %0,%3\n"	/* copy crosses next page boundary? */
+		"   jnh   5f\n"
+		"3: .insn ss,0xc80000000000,0(%3,%1),0(%4),0\n"
+		"  "SLR"  %0,%3\n"
+		"   j     5f\n"
+		"4:"SLR"  %0,%0\n"
+		"5: \n"
+		EX_TABLE(0b,2b) EX_TABLE(3b,5b)
+		: "+a" (size), "+a" (to), "+a" (tmp1), "=a" (tmp2)
+		: "a" (empty_zero_page), "d" (reg0) : "cc", "memory");
+	return size;
+}
+
+extern size_t copy_from_user_std_small(size_t, const void __user *, void *);
+extern size_t copy_to_user_std_small(size_t, void __user *, const void *);
+extern size_t strnlen_user_std(size_t, const char __user *);
+extern size_t strncpy_from_user_std(size_t, const char __user *, char *);
+extern int futex_atomic_op(int, int __user *, int, int *);
+extern int futex_atomic_cmpxchg(int __user *, int, int);
+
+struct uaccess_ops uaccess_mvcos = {
+	.copy_from_user = copy_from_user_mvcos,
+	.copy_from_user_small = copy_from_user_std_small,
+	.copy_to_user = copy_to_user_mvcos,
+	.copy_to_user_small = copy_to_user_std_small,
+	.copy_in_user = copy_in_user_mvcos,
+	.clear_user = clear_user_mvcos,
+	.strnlen_user = strnlen_user_std,
+	.strncpy_from_user = strncpy_from_user_std,
+	.futex_atomic_op = futex_atomic_op,
+	.futex_atomic_cmpxchg = futex_atomic_cmpxchg,
+};
diff --git a/include/asm-s390/setup.h b/include/asm-s390/setup.h
index 00c03e46689b..f1959732b6fd 100644
--- a/include/asm-s390/setup.h
+++ b/include/asm-s390/setup.h
@@ -44,10 +44,12 @@ extern unsigned long machine_flags;
 #define MACHINE_HAS_IEEE	(machine_flags & 2)
 #define MACHINE_HAS_CSP		(machine_flags & 8)
 #define MACHINE_HAS_DIAG44	(1)
+#define MACHINE_HAS_MVCOS	(0)
 #else /* __s390x__ */
 #define MACHINE_HAS_IEEE	(1)
 #define MACHINE_HAS_CSP		(1)
 #define MACHINE_HAS_DIAG44	(machine_flags & 32)
+#define MACHINE_HAS_MVCOS	(machine_flags & 512)
 #endif /* __s390x__ */
 
 
diff --git a/include/asm-s390/uaccess.h b/include/asm-s390/uaccess.h
index 39a2716ae188..e2047b0c9092 100644
--- a/include/asm-s390/uaccess.h
+++ b/include/asm-s390/uaccess.h
@@ -100,6 +100,7 @@ struct uaccess_ops {
 
 extern struct uaccess_ops uaccess;
 extern struct uaccess_ops uaccess_std;
+extern struct uaccess_ops uaccess_mvcos;
 
 static inline int __put_user_fn(size_t size, void __user *ptr, void *x)
 {
-- 
cgit v1.2.3


From 799111020c66c41aef621a3b53ad112543754124 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Mon, 21 Aug 2006 21:03:52 +1000
Subject: [CRYPTO] api: Fixed crypto_tfm context alignment

Previously the __aligned__ attribute was added to the crypto_tfm context
member to ensure it is alinged correctly on architectures such as arm.
Unfortunately kmalloc does not use the same minimum alignment rules as
gcc so this is useless.

This patch changes it to use kmalloc's minimum alignment.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 include/linux/crypto.h | 25 +++++++++++++++++++++++--
 1 file changed, 23 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index 7f946241b879..cb1e6631b132 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -21,8 +21,9 @@
 #include <linux/kernel.h>
 #include <linux/types.h>
 #include <linux/list.h>
+#include <linux/slab.h>
 #include <linux/string.h>
-#include <asm/page.h>
+#include <linux/uaccess.h>
 
 /*
  * Algorithm masks and types.
@@ -61,6 +62,26 @@
 #define CRYPTO_DIR_ENCRYPT		1
 #define CRYPTO_DIR_DECRYPT		0
 
+/*
+ * The macro CRYPTO_MINALIGN_ATTR (along with the void * type in the actual
+ * declaration) is used to ensure that the crypto_tfm context structure is
+ * aligned correctly for the given architecture so that there are no alignment
+ * faults for C data types.  In particular, this is required on platforms such
+ * as arm where pointers are 32-bit aligned but there are data types such as
+ * u64 which require 64-bit alignment.
+ */
+#if defined(ARCH_KMALLOC_MINALIGN)
+#define CRYPTO_MINALIGN ARCH_KMALLOC_MINALIGN
+#elif defined(ARCH_SLAB_MINALIGN)
+#define CRYPTO_MINALIGN ARCH_SLAB_MINALIGN
+#endif
+
+#ifdef CRYPTO_MINALIGN
+#define CRYPTO_MINALIGN_ATTR __attribute__ ((__aligned__(CRYPTO_MINALIGN)))
+#else
+#define CRYPTO_MINALIGN_ATTR
+#endif
+
 struct scatterlist;
 struct crypto_tfm;
 
@@ -231,7 +252,7 @@ struct crypto_tfm {
 	
 	struct crypto_alg *__crt_alg;
 
-	char __crt_ctx[] __attribute__ ((__aligned__));
+	void *__crt_ctx[] CRYPTO_MINALIGN_ATTR;
 };
 
 /* 
-- 
cgit v1.2.3


From 2729bb427f686e47970406d6bde6b11892885f29 Mon Sep 17 00:00:00 2001
From: Joachim Fritschi <jfritschi@freenet.de>
Date: Tue, 20 Jun 2006 20:37:23 +1000
Subject: [CRYPTO] twofish: Split out common c code

This patch splits up the twofish crypto routine into a common part ( key
setup  ) which will be uses by all twofish crypto modules ( generic-c , i586
assembler and x86_64 assembler ) and generic-c part. It also creates a new
header file which will be used by all 3 modules.

This eliminates all code duplication.

Correctness was verified with the tcrypt module and automated test scripts.

Signed-off-by: Joachim Fritschi <jfritschi@freenet.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 crypto/Kconfig           |   8 +
 crypto/Makefile          |   1 +
 crypto/twofish.c         | 698 +-------------------------------------------
 crypto/twofish_common.c  | 744 +++++++++++++++++++++++++++++++++++++++++++++++
 include/crypto/twofish.h |  23 ++
 5 files changed, 777 insertions(+), 697 deletions(-)
 create mode 100644 crypto/twofish_common.c
 create mode 100644 include/crypto/twofish.h

(limited to 'include')

diff --git a/crypto/Kconfig b/crypto/Kconfig
index ba133d557045..5472f693e6ec 100644
--- a/crypto/Kconfig
+++ b/crypto/Kconfig
@@ -131,6 +131,7 @@ config CRYPTO_BLOWFISH
 config CRYPTO_TWOFISH
 	tristate "Twofish cipher algorithm"
 	depends on CRYPTO
+	select CRYPTO_TWOFISH_COMMON
 	help
 	  Twofish cipher algorithm.
 	  
@@ -142,6 +143,13 @@ config CRYPTO_TWOFISH
 	  See also:
 	  <http://www.schneier.com/twofish.html>
 
+config CRYPTO_TWOFISH_COMMON
+	tristate
+	depends on CRYPTO
+	help
+	  Common parts of the Twofish cipher algorithm shared by the
+	  generic c and the assembler implementations.
+
 config CRYPTO_SERPENT
 	tristate "Serpent cipher algorithm"
 	depends on CRYPTO
diff --git a/crypto/Makefile b/crypto/Makefile
index d287b9e60c47..fe934f1001c6 100644
--- a/crypto/Makefile
+++ b/crypto/Makefile
@@ -19,6 +19,7 @@ obj-$(CONFIG_CRYPTO_TGR192) += tgr192.o
 obj-$(CONFIG_CRYPTO_DES) += des.o
 obj-$(CONFIG_CRYPTO_BLOWFISH) += blowfish.o
 obj-$(CONFIG_CRYPTO_TWOFISH) += twofish.o
+obj-$(CONFIG_CRYPTO_TWOFISH_COMMON) += twofish_common.o
 obj-$(CONFIG_CRYPTO_SERPENT) += serpent.o
 obj-$(CONFIG_CRYPTO_AES) += aes.o
 obj-$(CONFIG_CRYPTO_CAST5) += cast5.o
diff --git a/crypto/twofish.c b/crypto/twofish.c
index ec2488242e2d..e3b3a0a6cb4d 100644
--- a/crypto/twofish.c
+++ b/crypto/twofish.c
@@ -39,6 +39,7 @@
  */
 
 #include <asm/byteorder.h>
+#include <crypto/twofish.h>
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/types.h>
@@ -46,534 +47,6 @@
 #include <linux/crypto.h>
 #include <linux/bitops.h>
 
-
-/* The large precomputed tables for the Twofish cipher (twofish.c)
- * Taken from the same source as twofish.c
- * Marc Mutz <Marc@Mutz.com>
- */
-
-/* These two tables are the q0 and q1 permutations, exactly as described in
- * the Twofish paper. */
-
-static const u8 q0[256] = {
-   0xA9, 0x67, 0xB3, 0xE8, 0x04, 0xFD, 0xA3, 0x76, 0x9A, 0x92, 0x80, 0x78,
-   0xE4, 0xDD, 0xD1, 0x38, 0x0D, 0xC6, 0x35, 0x98, 0x18, 0xF7, 0xEC, 0x6C,
-   0x43, 0x75, 0x37, 0x26, 0xFA, 0x13, 0x94, 0x48, 0xF2, 0xD0, 0x8B, 0x30,
-   0x84, 0x54, 0xDF, 0x23, 0x19, 0x5B, 0x3D, 0x59, 0xF3, 0xAE, 0xA2, 0x82,
-   0x63, 0x01, 0x83, 0x2E, 0xD9, 0x51, 0x9B, 0x7C, 0xA6, 0xEB, 0xA5, 0xBE,
-   0x16, 0x0C, 0xE3, 0x61, 0xC0, 0x8C, 0x3A, 0xF5, 0x73, 0x2C, 0x25, 0x0B,
-   0xBB, 0x4E, 0x89, 0x6B, 0x53, 0x6A, 0xB4, 0xF1, 0xE1, 0xE6, 0xBD, 0x45,
-   0xE2, 0xF4, 0xB6, 0x66, 0xCC, 0x95, 0x03, 0x56, 0xD4, 0x1C, 0x1E, 0xD7,
-   0xFB, 0xC3, 0x8E, 0xB5, 0xE9, 0xCF, 0xBF, 0xBA, 0xEA, 0x77, 0x39, 0xAF,
-   0x33, 0xC9, 0x62, 0x71, 0x81, 0x79, 0x09, 0xAD, 0x24, 0xCD, 0xF9, 0xD8,
-   0xE5, 0xC5, 0xB9, 0x4D, 0x44, 0x08, 0x86, 0xE7, 0xA1, 0x1D, 0xAA, 0xED,
-   0x06, 0x70, 0xB2, 0xD2, 0x41, 0x7B, 0xA0, 0x11, 0x31, 0xC2, 0x27, 0x90,
-   0x20, 0xF6, 0x60, 0xFF, 0x96, 0x5C, 0xB1, 0xAB, 0x9E, 0x9C, 0x52, 0x1B,
-   0x5F, 0x93, 0x0A, 0xEF, 0x91, 0x85, 0x49, 0xEE, 0x2D, 0x4F, 0x8F, 0x3B,
-   0x47, 0x87, 0x6D, 0x46, 0xD6, 0x3E, 0x69, 0x64, 0x2A, 0xCE, 0xCB, 0x2F,
-   0xFC, 0x97, 0x05, 0x7A, 0xAC, 0x7F, 0xD5, 0x1A, 0x4B, 0x0E, 0xA7, 0x5A,
-   0x28, 0x14, 0x3F, 0x29, 0x88, 0x3C, 0x4C, 0x02, 0xB8, 0xDA, 0xB0, 0x17,
-   0x55, 0x1F, 0x8A, 0x7D, 0x57, 0xC7, 0x8D, 0x74, 0xB7, 0xC4, 0x9F, 0x72,
-   0x7E, 0x15, 0x22, 0x12, 0x58, 0x07, 0x99, 0x34, 0x6E, 0x50, 0xDE, 0x68,
-   0x65, 0xBC, 0xDB, 0xF8, 0xC8, 0xA8, 0x2B, 0x40, 0xDC, 0xFE, 0x32, 0xA4,
-   0xCA, 0x10, 0x21, 0xF0, 0xD3, 0x5D, 0x0F, 0x00, 0x6F, 0x9D, 0x36, 0x42,
-   0x4A, 0x5E, 0xC1, 0xE0
-};
-
-static const u8 q1[256] = {
-   0x75, 0xF3, 0xC6, 0xF4, 0xDB, 0x7B, 0xFB, 0xC8, 0x4A, 0xD3, 0xE6, 0x6B,
-   0x45, 0x7D, 0xE8, 0x4B, 0xD6, 0x32, 0xD8, 0xFD, 0x37, 0x71, 0xF1, 0xE1,
-   0x30, 0x0F, 0xF8, 0x1B, 0x87, 0xFA, 0x06, 0x3F, 0x5E, 0xBA, 0xAE, 0x5B,
-   0x8A, 0x00, 0xBC, 0x9D, 0x6D, 0xC1, 0xB1, 0x0E, 0x80, 0x5D, 0xD2, 0xD5,
-   0xA0, 0x84, 0x07, 0x14, 0xB5, 0x90, 0x2C, 0xA3, 0xB2, 0x73, 0x4C, 0x54,
-   0x92, 0x74, 0x36, 0x51, 0x38, 0xB0, 0xBD, 0x5A, 0xFC, 0x60, 0x62, 0x96,
-   0x6C, 0x42, 0xF7, 0x10, 0x7C, 0x28, 0x27, 0x8C, 0x13, 0x95, 0x9C, 0xC7,
-   0x24, 0x46, 0x3B, 0x70, 0xCA, 0xE3, 0x85, 0xCB, 0x11, 0xD0, 0x93, 0xB8,
-   0xA6, 0x83, 0x20, 0xFF, 0x9F, 0x77, 0xC3, 0xCC, 0x03, 0x6F, 0x08, 0xBF,
-   0x40, 0xE7, 0x2B, 0xE2, 0x79, 0x0C, 0xAA, 0x82, 0x41, 0x3A, 0xEA, 0xB9,
-   0xE4, 0x9A, 0xA4, 0x97, 0x7E, 0xDA, 0x7A, 0x17, 0x66, 0x94, 0xA1, 0x1D,
-   0x3D, 0xF0, 0xDE, 0xB3, 0x0B, 0x72, 0xA7, 0x1C, 0xEF, 0xD1, 0x53, 0x3E,
-   0x8F, 0x33, 0x26, 0x5F, 0xEC, 0x76, 0x2A, 0x49, 0x81, 0x88, 0xEE, 0x21,
-   0xC4, 0x1A, 0xEB, 0xD9, 0xC5, 0x39, 0x99, 0xCD, 0xAD, 0x31, 0x8B, 0x01,
-   0x18, 0x23, 0xDD, 0x1F, 0x4E, 0x2D, 0xF9, 0x48, 0x4F, 0xF2, 0x65, 0x8E,
-   0x78, 0x5C, 0x58, 0x19, 0x8D, 0xE5, 0x98, 0x57, 0x67, 0x7F, 0x05, 0x64,
-   0xAF, 0x63, 0xB6, 0xFE, 0xF5, 0xB7, 0x3C, 0xA5, 0xCE, 0xE9, 0x68, 0x44,
-   0xE0, 0x4D, 0x43, 0x69, 0x29, 0x2E, 0xAC, 0x15, 0x59, 0xA8, 0x0A, 0x9E,
-   0x6E, 0x47, 0xDF, 0x34, 0x35, 0x6A, 0xCF, 0xDC, 0x22, 0xC9, 0xC0, 0x9B,
-   0x89, 0xD4, 0xED, 0xAB, 0x12, 0xA2, 0x0D, 0x52, 0xBB, 0x02, 0x2F, 0xA9,
-   0xD7, 0x61, 0x1E, 0xB4, 0x50, 0x04, 0xF6, 0xC2, 0x16, 0x25, 0x86, 0x56,
-   0x55, 0x09, 0xBE, 0x91
-};
-
-/* These MDS tables are actually tables of MDS composed with q0 and q1,
- * because it is only ever used that way and we can save some time by
- * precomputing.  Of course the main saving comes from precomputing the
- * GF(2^8) multiplication involved in the MDS matrix multiply; by looking
- * things up in these tables we reduce the matrix multiply to four lookups
- * and three XORs.  Semi-formally, the definition of these tables is:
- * mds[0][i] = MDS (q1[i] 0 0 0)^T  mds[1][i] = MDS (0 q0[i] 0 0)^T
- * mds[2][i] = MDS (0 0 q1[i] 0)^T  mds[3][i] = MDS (0 0 0 q0[i])^T
- * where ^T means "transpose", the matrix multiply is performed in GF(2^8)
- * represented as GF(2)[x]/v(x) where v(x)=x^8+x^6+x^5+x^3+1 as described
- * by Schneier et al, and I'm casually glossing over the byte/word
- * conversion issues. */
-
-static const u32 mds[4][256] = {
-   {0xBCBC3275, 0xECEC21F3, 0x202043C6, 0xB3B3C9F4, 0xDADA03DB, 0x02028B7B,
-    0xE2E22BFB, 0x9E9EFAC8, 0xC9C9EC4A, 0xD4D409D3, 0x18186BE6, 0x1E1E9F6B,
-    0x98980E45, 0xB2B2387D, 0xA6A6D2E8, 0x2626B74B, 0x3C3C57D6, 0x93938A32,
-    0x8282EED8, 0x525298FD, 0x7B7BD437, 0xBBBB3771, 0x5B5B97F1, 0x474783E1,
-    0x24243C30, 0x5151E20F, 0xBABAC6F8, 0x4A4AF31B, 0xBFBF4887, 0x0D0D70FA,
-    0xB0B0B306, 0x7575DE3F, 0xD2D2FD5E, 0x7D7D20BA, 0x666631AE, 0x3A3AA35B,
-    0x59591C8A, 0x00000000, 0xCDCD93BC, 0x1A1AE09D, 0xAEAE2C6D, 0x7F7FABC1,
-    0x2B2BC7B1, 0xBEBEB90E, 0xE0E0A080, 0x8A8A105D, 0x3B3B52D2, 0x6464BAD5,
-    0xD8D888A0, 0xE7E7A584, 0x5F5FE807, 0x1B1B1114, 0x2C2CC2B5, 0xFCFCB490,
-    0x3131272C, 0x808065A3, 0x73732AB2, 0x0C0C8173, 0x79795F4C, 0x6B6B4154,
-    0x4B4B0292, 0x53536974, 0x94948F36, 0x83831F51, 0x2A2A3638, 0xC4C49CB0,
-    0x2222C8BD, 0xD5D5F85A, 0xBDBDC3FC, 0x48487860, 0xFFFFCE62, 0x4C4C0796,
-    0x4141776C, 0xC7C7E642, 0xEBEB24F7, 0x1C1C1410, 0x5D5D637C, 0x36362228,
-    0x6767C027, 0xE9E9AF8C, 0x4444F913, 0x1414EA95, 0xF5F5BB9C, 0xCFCF18C7,
-    0x3F3F2D24, 0xC0C0E346, 0x7272DB3B, 0x54546C70, 0x29294CCA, 0xF0F035E3,
-    0x0808FE85, 0xC6C617CB, 0xF3F34F11, 0x8C8CE4D0, 0xA4A45993, 0xCACA96B8,
-    0x68683BA6, 0xB8B84D83, 0x38382820, 0xE5E52EFF, 0xADAD569F, 0x0B0B8477,
-    0xC8C81DC3, 0x9999FFCC, 0x5858ED03, 0x19199A6F, 0x0E0E0A08, 0x95957EBF,
-    0x70705040, 0xF7F730E7, 0x6E6ECF2B, 0x1F1F6EE2, 0xB5B53D79, 0x09090F0C,
-    0x616134AA, 0x57571682, 0x9F9F0B41, 0x9D9D803A, 0x111164EA, 0x2525CDB9,
-    0xAFAFDDE4, 0x4545089A, 0xDFDF8DA4, 0xA3A35C97, 0xEAEAD57E, 0x353558DA,
-    0xEDEDD07A, 0x4343FC17, 0xF8F8CB66, 0xFBFBB194, 0x3737D3A1, 0xFAFA401D,
-    0xC2C2683D, 0xB4B4CCF0, 0x32325DDE, 0x9C9C71B3, 0x5656E70B, 0xE3E3DA72,
-    0x878760A7, 0x15151B1C, 0xF9F93AEF, 0x6363BFD1, 0x3434A953, 0x9A9A853E,
-    0xB1B1428F, 0x7C7CD133, 0x88889B26, 0x3D3DA65F, 0xA1A1D7EC, 0xE4E4DF76,
-    0x8181942A, 0x91910149, 0x0F0FFB81, 0xEEEEAA88, 0x161661EE, 0xD7D77321,
-    0x9797F5C4, 0xA5A5A81A, 0xFEFE3FEB, 0x6D6DB5D9, 0x7878AEC5, 0xC5C56D39,
-    0x1D1DE599, 0x7676A4CD, 0x3E3EDCAD, 0xCBCB6731, 0xB6B6478B, 0xEFEF5B01,
-    0x12121E18, 0x6060C523, 0x6A6AB0DD, 0x4D4DF61F, 0xCECEE94E, 0xDEDE7C2D,
-    0x55559DF9, 0x7E7E5A48, 0x2121B24F, 0x03037AF2, 0xA0A02665, 0x5E5E198E,
-    0x5A5A6678, 0x65654B5C, 0x62624E58, 0xFDFD4519, 0x0606F48D, 0x404086E5,
-    0xF2F2BE98, 0x3333AC57, 0x17179067, 0x05058E7F, 0xE8E85E05, 0x4F4F7D64,
-    0x89896AAF, 0x10109563, 0x74742FB6, 0x0A0A75FE, 0x5C5C92F5, 0x9B9B74B7,
-    0x2D2D333C, 0x3030D6A5, 0x2E2E49CE, 0x494989E9, 0x46467268, 0x77775544,
-    0xA8A8D8E0, 0x9696044D, 0x2828BD43, 0xA9A92969, 0xD9D97929, 0x8686912E,
-    0xD1D187AC, 0xF4F44A15, 0x8D8D1559, 0xD6D682A8, 0xB9B9BC0A, 0x42420D9E,
-    0xF6F6C16E, 0x2F2FB847, 0xDDDD06DF, 0x23233934, 0xCCCC6235, 0xF1F1C46A,
-    0xC1C112CF, 0x8585EBDC, 0x8F8F9E22, 0x7171A1C9, 0x9090F0C0, 0xAAAA539B,
-    0x0101F189, 0x8B8BE1D4, 0x4E4E8CED, 0x8E8E6FAB, 0xABABA212, 0x6F6F3EA2,
-    0xE6E6540D, 0xDBDBF252, 0x92927BBB, 0xB7B7B602, 0x6969CA2F, 0x3939D9A9,
-    0xD3D30CD7, 0xA7A72361, 0xA2A2AD1E, 0xC3C399B4, 0x6C6C4450, 0x07070504,
-    0x04047FF6, 0x272746C2, 0xACACA716, 0xD0D07625, 0x50501386, 0xDCDCF756,
-    0x84841A55, 0xE1E15109, 0x7A7A25BE, 0x1313EF91},
-
-   {0xA9D93939, 0x67901717, 0xB3719C9C, 0xE8D2A6A6, 0x04050707, 0xFD985252,
-    0xA3658080, 0x76DFE4E4, 0x9A084545, 0x92024B4B, 0x80A0E0E0, 0x78665A5A,
-    0xE4DDAFAF, 0xDDB06A6A, 0xD1BF6363, 0x38362A2A, 0x0D54E6E6, 0xC6432020,
-    0x3562CCCC, 0x98BEF2F2, 0x181E1212, 0xF724EBEB, 0xECD7A1A1, 0x6C774141,
-    0x43BD2828, 0x7532BCBC, 0x37D47B7B, 0x269B8888, 0xFA700D0D, 0x13F94444,
-    0x94B1FBFB, 0x485A7E7E, 0xF27A0303, 0xD0E48C8C, 0x8B47B6B6, 0x303C2424,
-    0x84A5E7E7, 0x54416B6B, 0xDF06DDDD, 0x23C56060, 0x1945FDFD, 0x5BA33A3A,
-    0x3D68C2C2, 0x59158D8D, 0xF321ECEC, 0xAE316666, 0xA23E6F6F, 0x82165757,
-    0x63951010, 0x015BEFEF, 0x834DB8B8, 0x2E918686, 0xD9B56D6D, 0x511F8383,
-    0x9B53AAAA, 0x7C635D5D, 0xA63B6868, 0xEB3FFEFE, 0xA5D63030, 0xBE257A7A,
-    0x16A7ACAC, 0x0C0F0909, 0xE335F0F0, 0x6123A7A7, 0xC0F09090, 0x8CAFE9E9,
-    0x3A809D9D, 0xF5925C5C, 0x73810C0C, 0x2C273131, 0x2576D0D0, 0x0BE75656,
-    0xBB7B9292, 0x4EE9CECE, 0x89F10101, 0x6B9F1E1E, 0x53A93434, 0x6AC4F1F1,
-    0xB499C3C3, 0xF1975B5B, 0xE1834747, 0xE66B1818, 0xBDC82222, 0x450E9898,
-    0xE26E1F1F, 0xF4C9B3B3, 0xB62F7474, 0x66CBF8F8, 0xCCFF9999, 0x95EA1414,
-    0x03ED5858, 0x56F7DCDC, 0xD4E18B8B, 0x1C1B1515, 0x1EADA2A2, 0xD70CD3D3,
-    0xFB2BE2E2, 0xC31DC8C8, 0x8E195E5E, 0xB5C22C2C, 0xE9894949, 0xCF12C1C1,
-    0xBF7E9595, 0xBA207D7D, 0xEA641111, 0x77840B0B, 0x396DC5C5, 0xAF6A8989,
-    0x33D17C7C, 0xC9A17171, 0x62CEFFFF, 0x7137BBBB, 0x81FB0F0F, 0x793DB5B5,
-    0x0951E1E1, 0xADDC3E3E, 0x242D3F3F, 0xCDA47676, 0xF99D5555, 0xD8EE8282,
-    0xE5864040, 0xC5AE7878, 0xB9CD2525, 0x4D049696, 0x44557777, 0x080A0E0E,
-    0x86135050, 0xE730F7F7, 0xA1D33737, 0x1D40FAFA, 0xAA346161, 0xED8C4E4E,
-    0x06B3B0B0, 0x706C5454, 0xB22A7373, 0xD2523B3B, 0x410B9F9F, 0x7B8B0202,
-    0xA088D8D8, 0x114FF3F3, 0x3167CBCB, 0xC2462727, 0x27C06767, 0x90B4FCFC,
-    0x20283838, 0xF67F0404, 0x60784848, 0xFF2EE5E5, 0x96074C4C, 0x5C4B6565,
-    0xB1C72B2B, 0xAB6F8E8E, 0x9E0D4242, 0x9CBBF5F5, 0x52F2DBDB, 0x1BF34A4A,
-    0x5FA63D3D, 0x9359A4A4, 0x0ABCB9B9, 0xEF3AF9F9, 0x91EF1313, 0x85FE0808,
-    0x49019191, 0xEE611616, 0x2D7CDEDE, 0x4FB22121, 0x8F42B1B1, 0x3BDB7272,
-    0x47B82F2F, 0x8748BFBF, 0x6D2CAEAE, 0x46E3C0C0, 0xD6573C3C, 0x3E859A9A,
-    0x6929A9A9, 0x647D4F4F, 0x2A948181, 0xCE492E2E, 0xCB17C6C6, 0x2FCA6969,
-    0xFCC3BDBD, 0x975CA3A3, 0x055EE8E8, 0x7AD0EDED, 0xAC87D1D1, 0x7F8E0505,
-    0xD5BA6464, 0x1AA8A5A5, 0x4BB72626, 0x0EB9BEBE, 0xA7608787, 0x5AF8D5D5,
-    0x28223636, 0x14111B1B, 0x3FDE7575, 0x2979D9D9, 0x88AAEEEE, 0x3C332D2D,
-    0x4C5F7979, 0x02B6B7B7, 0xB896CACA, 0xDA583535, 0xB09CC4C4, 0x17FC4343,
-    0x551A8484, 0x1FF64D4D, 0x8A1C5959, 0x7D38B2B2, 0x57AC3333, 0xC718CFCF,
-    0x8DF40606, 0x74695353, 0xB7749B9B, 0xC4F59797, 0x9F56ADAD, 0x72DAE3E3,
-    0x7ED5EAEA, 0x154AF4F4, 0x229E8F8F, 0x12A2ABAB, 0x584E6262, 0x07E85F5F,
-    0x99E51D1D, 0x34392323, 0x6EC1F6F6, 0x50446C6C, 0xDE5D3232, 0x68724646,
-    0x6526A0A0, 0xBC93CDCD, 0xDB03DADA, 0xF8C6BABA, 0xC8FA9E9E, 0xA882D6D6,
-    0x2BCF6E6E, 0x40507070, 0xDCEB8585, 0xFE750A0A, 0x328A9393, 0xA48DDFDF,
-    0xCA4C2929, 0x10141C1C, 0x2173D7D7, 0xF0CCB4B4, 0xD309D4D4, 0x5D108A8A,
-    0x0FE25151, 0x00000000, 0x6F9A1919, 0x9DE01A1A, 0x368F9494, 0x42E6C7C7,
-    0x4AECC9C9, 0x5EFDD2D2, 0xC1AB7F7F, 0xE0D8A8A8},
-
-   {0xBC75BC32, 0xECF3EC21, 0x20C62043, 0xB3F4B3C9, 0xDADBDA03, 0x027B028B,
-    0xE2FBE22B, 0x9EC89EFA, 0xC94AC9EC, 0xD4D3D409, 0x18E6186B, 0x1E6B1E9F,
-    0x9845980E, 0xB27DB238, 0xA6E8A6D2, 0x264B26B7, 0x3CD63C57, 0x9332938A,
-    0x82D882EE, 0x52FD5298, 0x7B377BD4, 0xBB71BB37, 0x5BF15B97, 0x47E14783,
-    0x2430243C, 0x510F51E2, 0xBAF8BAC6, 0x4A1B4AF3, 0xBF87BF48, 0x0DFA0D70,
-    0xB006B0B3, 0x753F75DE, 0xD25ED2FD, 0x7DBA7D20, 0x66AE6631, 0x3A5B3AA3,
-    0x598A591C, 0x00000000, 0xCDBCCD93, 0x1A9D1AE0, 0xAE6DAE2C, 0x7FC17FAB,
-    0x2BB12BC7, 0xBE0EBEB9, 0xE080E0A0, 0x8A5D8A10, 0x3BD23B52, 0x64D564BA,
-    0xD8A0D888, 0xE784E7A5, 0x5F075FE8, 0x1B141B11, 0x2CB52CC2, 0xFC90FCB4,
-    0x312C3127, 0x80A38065, 0x73B2732A, 0x0C730C81, 0x794C795F, 0x6B546B41,
-    0x4B924B02, 0x53745369, 0x9436948F, 0x8351831F, 0x2A382A36, 0xC4B0C49C,
-    0x22BD22C8, 0xD55AD5F8, 0xBDFCBDC3, 0x48604878, 0xFF62FFCE, 0x4C964C07,
-    0x416C4177, 0xC742C7E6, 0xEBF7EB24, 0x1C101C14, 0x5D7C5D63, 0x36283622,
-    0x672767C0, 0xE98CE9AF, 0x441344F9, 0x149514EA, 0xF59CF5BB, 0xCFC7CF18,
-    0x3F243F2D, 0xC046C0E3, 0x723B72DB, 0x5470546C, 0x29CA294C, 0xF0E3F035,
-    0x088508FE, 0xC6CBC617, 0xF311F34F, 0x8CD08CE4, 0xA493A459, 0xCAB8CA96,
-    0x68A6683B, 0xB883B84D, 0x38203828, 0xE5FFE52E, 0xAD9FAD56, 0x0B770B84,
-    0xC8C3C81D, 0x99CC99FF, 0x580358ED, 0x196F199A, 0x0E080E0A, 0x95BF957E,
-    0x70407050, 0xF7E7F730, 0x6E2B6ECF, 0x1FE21F6E, 0xB579B53D, 0x090C090F,
-    0x61AA6134, 0x57825716, 0x9F419F0B, 0x9D3A9D80, 0x11EA1164, 0x25B925CD,
-    0xAFE4AFDD, 0x459A4508, 0xDFA4DF8D, 0xA397A35C, 0xEA7EEAD5, 0x35DA3558,
-    0xED7AEDD0, 0x431743FC, 0xF866F8CB, 0xFB94FBB1, 0x37A137D3, 0xFA1DFA40,
-    0xC23DC268, 0xB4F0B4CC, 0x32DE325D, 0x9CB39C71, 0x560B56E7, 0xE372E3DA,
-    0x87A78760, 0x151C151B, 0xF9EFF93A, 0x63D163BF, 0x345334A9, 0x9A3E9A85,
-    0xB18FB142, 0x7C337CD1, 0x8826889B, 0x3D5F3DA6, 0xA1ECA1D7, 0xE476E4DF,
-    0x812A8194, 0x91499101, 0x0F810FFB, 0xEE88EEAA, 0x16EE1661, 0xD721D773,
-    0x97C497F5, 0xA51AA5A8, 0xFEEBFE3F, 0x6DD96DB5, 0x78C578AE, 0xC539C56D,
-    0x1D991DE5, 0x76CD76A4, 0x3EAD3EDC, 0xCB31CB67, 0xB68BB647, 0xEF01EF5B,
-    0x1218121E, 0x602360C5, 0x6ADD6AB0, 0x4D1F4DF6, 0xCE4ECEE9, 0xDE2DDE7C,
-    0x55F9559D, 0x7E487E5A, 0x214F21B2, 0x03F2037A, 0xA065A026, 0x5E8E5E19,
-    0x5A785A66, 0x655C654B, 0x6258624E, 0xFD19FD45, 0x068D06F4, 0x40E54086,
-    0xF298F2BE, 0x335733AC, 0x17671790, 0x057F058E, 0xE805E85E, 0x4F644F7D,
-    0x89AF896A, 0x10631095, 0x74B6742F, 0x0AFE0A75, 0x5CF55C92, 0x9BB79B74,
-    0x2D3C2D33, 0x30A530D6, 0x2ECE2E49, 0x49E94989, 0x46684672, 0x77447755,
-    0xA8E0A8D8, 0x964D9604, 0x284328BD, 0xA969A929, 0xD929D979, 0x862E8691,
-    0xD1ACD187, 0xF415F44A, 0x8D598D15, 0xD6A8D682, 0xB90AB9BC, 0x429E420D,
-    0xF66EF6C1, 0x2F472FB8, 0xDDDFDD06, 0x23342339, 0xCC35CC62, 0xF16AF1C4,
-    0xC1CFC112, 0x85DC85EB, 0x8F228F9E, 0x71C971A1, 0x90C090F0, 0xAA9BAA53,
-    0x018901F1, 0x8BD48BE1, 0x4EED4E8C, 0x8EAB8E6F, 0xAB12ABA2, 0x6FA26F3E,
-    0xE60DE654, 0xDB52DBF2, 0x92BB927B, 0xB702B7B6, 0x692F69CA, 0x39A939D9,
-    0xD3D7D30C, 0xA761A723, 0xA21EA2AD, 0xC3B4C399, 0x6C506C44, 0x07040705,
-    0x04F6047F, 0x27C22746, 0xAC16ACA7, 0xD025D076, 0x50865013, 0xDC56DCF7,
-    0x8455841A, 0xE109E151, 0x7ABE7A25, 0x139113EF},
-
-   {0xD939A9D9, 0x90176790, 0x719CB371, 0xD2A6E8D2, 0x05070405, 0x9852FD98,
-    0x6580A365, 0xDFE476DF, 0x08459A08, 0x024B9202, 0xA0E080A0, 0x665A7866,
-    0xDDAFE4DD, 0xB06ADDB0, 0xBF63D1BF, 0x362A3836, 0x54E60D54, 0x4320C643,
-    0x62CC3562, 0xBEF298BE, 0x1E12181E, 0x24EBF724, 0xD7A1ECD7, 0x77416C77,
-    0xBD2843BD, 0x32BC7532, 0xD47B37D4, 0x9B88269B, 0x700DFA70, 0xF94413F9,
-    0xB1FB94B1, 0x5A7E485A, 0x7A03F27A, 0xE48CD0E4, 0x47B68B47, 0x3C24303C,
-    0xA5E784A5, 0x416B5441, 0x06DDDF06, 0xC56023C5, 0x45FD1945, 0xA33A5BA3,
-    0x68C23D68, 0x158D5915, 0x21ECF321, 0x3166AE31, 0x3E6FA23E, 0x16578216,
-    0x95106395, 0x5BEF015B, 0x4DB8834D, 0x91862E91, 0xB56DD9B5, 0x1F83511F,
-    0x53AA9B53, 0x635D7C63, 0x3B68A63B, 0x3FFEEB3F, 0xD630A5D6, 0x257ABE25,
-    0xA7AC16A7, 0x0F090C0F, 0x35F0E335, 0x23A76123, 0xF090C0F0, 0xAFE98CAF,
-    0x809D3A80, 0x925CF592, 0x810C7381, 0x27312C27, 0x76D02576, 0xE7560BE7,
-    0x7B92BB7B, 0xE9CE4EE9, 0xF10189F1, 0x9F1E6B9F, 0xA93453A9, 0xC4F16AC4,
-    0x99C3B499, 0x975BF197, 0x8347E183, 0x6B18E66B, 0xC822BDC8, 0x0E98450E,
-    0x6E1FE26E, 0xC9B3F4C9, 0x2F74B62F, 0xCBF866CB, 0xFF99CCFF, 0xEA1495EA,
-    0xED5803ED, 0xF7DC56F7, 0xE18BD4E1, 0x1B151C1B, 0xADA21EAD, 0x0CD3D70C,
-    0x2BE2FB2B, 0x1DC8C31D, 0x195E8E19, 0xC22CB5C2, 0x8949E989, 0x12C1CF12,
-    0x7E95BF7E, 0x207DBA20, 0x6411EA64, 0x840B7784, 0x6DC5396D, 0x6A89AF6A,
-    0xD17C33D1, 0xA171C9A1, 0xCEFF62CE, 0x37BB7137, 0xFB0F81FB, 0x3DB5793D,
-    0x51E10951, 0xDC3EADDC, 0x2D3F242D, 0xA476CDA4, 0x9D55F99D, 0xEE82D8EE,
-    0x8640E586, 0xAE78C5AE, 0xCD25B9CD, 0x04964D04, 0x55774455, 0x0A0E080A,
-    0x13508613, 0x30F7E730, 0xD337A1D3, 0x40FA1D40, 0x3461AA34, 0x8C4EED8C,
-    0xB3B006B3, 0x6C54706C, 0x2A73B22A, 0x523BD252, 0x0B9F410B, 0x8B027B8B,
-    0x88D8A088, 0x4FF3114F, 0x67CB3167, 0x4627C246, 0xC06727C0, 0xB4FC90B4,
-    0x28382028, 0x7F04F67F, 0x78486078, 0x2EE5FF2E, 0x074C9607, 0x4B655C4B,
-    0xC72BB1C7, 0x6F8EAB6F, 0x0D429E0D, 0xBBF59CBB, 0xF2DB52F2, 0xF34A1BF3,
-    0xA63D5FA6, 0x59A49359, 0xBCB90ABC, 0x3AF9EF3A, 0xEF1391EF, 0xFE0885FE,
-    0x01914901, 0x6116EE61, 0x7CDE2D7C, 0xB2214FB2, 0x42B18F42, 0xDB723BDB,
-    0xB82F47B8, 0x48BF8748, 0x2CAE6D2C, 0xE3C046E3, 0x573CD657, 0x859A3E85,
-    0x29A96929, 0x7D4F647D, 0x94812A94, 0x492ECE49, 0x17C6CB17, 0xCA692FCA,
-    0xC3BDFCC3, 0x5CA3975C, 0x5EE8055E, 0xD0ED7AD0, 0x87D1AC87, 0x8E057F8E,
-    0xBA64D5BA, 0xA8A51AA8, 0xB7264BB7, 0xB9BE0EB9, 0x6087A760, 0xF8D55AF8,
-    0x22362822, 0x111B1411, 0xDE753FDE, 0x79D92979, 0xAAEE88AA, 0x332D3C33,
-    0x5F794C5F, 0xB6B702B6, 0x96CAB896, 0x5835DA58, 0x9CC4B09C, 0xFC4317FC,
-    0x1A84551A, 0xF64D1FF6, 0x1C598A1C, 0x38B27D38, 0xAC3357AC, 0x18CFC718,
-    0xF4068DF4, 0x69537469, 0x749BB774, 0xF597C4F5, 0x56AD9F56, 0xDAE372DA,
-    0xD5EA7ED5, 0x4AF4154A, 0x9E8F229E, 0xA2AB12A2, 0x4E62584E, 0xE85F07E8,
-    0xE51D99E5, 0x39233439, 0xC1F66EC1, 0x446C5044, 0x5D32DE5D, 0x72466872,
-    0x26A06526, 0x93CDBC93, 0x03DADB03, 0xC6BAF8C6, 0xFA9EC8FA, 0x82D6A882,
-    0xCF6E2BCF, 0x50704050, 0xEB85DCEB, 0x750AFE75, 0x8A93328A, 0x8DDFA48D,
-    0x4C29CA4C, 0x141C1014, 0x73D72173, 0xCCB4F0CC, 0x09D4D309, 0x108A5D10,
-    0xE2510FE2, 0x00000000, 0x9A196F9A, 0xE01A9DE0, 0x8F94368F, 0xE6C742E6,
-    0xECC94AEC, 0xFDD25EFD, 0xAB7FC1AB, 0xD8A8E0D8}
-};
-
-/* The exp_to_poly and poly_to_exp tables are used to perform efficient
- * operations in GF(2^8) represented as GF(2)[x]/w(x) where
- * w(x)=x^8+x^6+x^3+x^2+1.  We care about doing that because it's part of the
- * definition of the RS matrix in the key schedule.  Elements of that field
- * are polynomials of degree not greater than 7 and all coefficients 0 or 1,
- * which can be represented naturally by bytes (just substitute x=2).  In that
- * form, GF(2^8) addition is the same as bitwise XOR, but GF(2^8)
- * multiplication is inefficient without hardware support.  To multiply
- * faster, I make use of the fact x is a generator for the nonzero elements,
- * so that every element p of GF(2)[x]/w(x) is either 0 or equal to (x)^n for
- * some n in 0..254.  Note that that caret is exponentiation in GF(2^8),
- * *not* polynomial notation.  So if I want to compute pq where p and q are
- * in GF(2^8), I can just say:
- *    1. if p=0 or q=0 then pq=0
- *    2. otherwise, find m and n such that p=x^m and q=x^n
- *    3. pq=(x^m)(x^n)=x^(m+n), so add m and n and find pq
- * The translations in steps 2 and 3 are looked up in the tables
- * poly_to_exp (for step 2) and exp_to_poly (for step 3).  To see this
- * in action, look at the CALC_S macro.  As additional wrinkles, note that
- * one of my operands is always a constant, so the poly_to_exp lookup on it
- * is done in advance; I included the original values in the comments so
- * readers can have some chance of recognizing that this *is* the RS matrix
- * from the Twofish paper.  I've only included the table entries I actually
- * need; I never do a lookup on a variable input of zero and the biggest
- * exponents I'll ever see are 254 (variable) and 237 (constant), so they'll
- * never sum to more than 491.	I'm repeating part of the exp_to_poly table
- * so that I don't have to do mod-255 reduction in the exponent arithmetic.
- * Since I know my constant operands are never zero, I only have to worry
- * about zero values in the variable operand, and I do it with a simple
- * conditional branch.	I know conditionals are expensive, but I couldn't
- * see a non-horrible way of avoiding them, and I did manage to group the
- * statements so that each if covers four group multiplications. */
-
-static const u8 poly_to_exp[255] = {
-   0x00, 0x01, 0x17, 0x02, 0x2E, 0x18, 0x53, 0x03, 0x6A, 0x2F, 0x93, 0x19,
-   0x34, 0x54, 0x45, 0x04, 0x5C, 0x6B, 0xB6, 0x30, 0xA6, 0x94, 0x4B, 0x1A,
-   0x8C, 0x35, 0x81, 0x55, 0xAA, 0x46, 0x0D, 0x05, 0x24, 0x5D, 0x87, 0x6C,
-   0x9B, 0xB7, 0xC1, 0x31, 0x2B, 0xA7, 0xA3, 0x95, 0x98, 0x4C, 0xCA, 0x1B,
-   0xE6, 0x8D, 0x73, 0x36, 0xCD, 0x82, 0x12, 0x56, 0x62, 0xAB, 0xF0, 0x47,
-   0x4F, 0x0E, 0xBD, 0x06, 0xD4, 0x25, 0xD2, 0x5E, 0x27, 0x88, 0x66, 0x6D,
-   0xD6, 0x9C, 0x79, 0xB8, 0x08, 0xC2, 0xDF, 0x32, 0x68, 0x2C, 0xFD, 0xA8,
-   0x8A, 0xA4, 0x5A, 0x96, 0x29, 0x99, 0x22, 0x4D, 0x60, 0xCB, 0xE4, 0x1C,
-   0x7B, 0xE7, 0x3B, 0x8E, 0x9E, 0x74, 0xF4, 0x37, 0xD8, 0xCE, 0xF9, 0x83,
-   0x6F, 0x13, 0xB2, 0x57, 0xE1, 0x63, 0xDC, 0xAC, 0xC4, 0xF1, 0xAF, 0x48,
-   0x0A, 0x50, 0x42, 0x0F, 0xBA, 0xBE, 0xC7, 0x07, 0xDE, 0xD5, 0x78, 0x26,
-   0x65, 0xD3, 0xD1, 0x5F, 0xE3, 0x28, 0x21, 0x89, 0x59, 0x67, 0xFC, 0x6E,
-   0xB1, 0xD7, 0xF8, 0x9D, 0xF3, 0x7A, 0x3A, 0xB9, 0xC6, 0x09, 0x41, 0xC3,
-   0xAE, 0xE0, 0xDB, 0x33, 0x44, 0x69, 0x92, 0x2D, 0x52, 0xFE, 0x16, 0xA9,
-   0x0C, 0x8B, 0x80, 0xA5, 0x4A, 0x5B, 0xB5, 0x97, 0xC9, 0x2A, 0xA2, 0x9A,
-   0xC0, 0x23, 0x86, 0x4E, 0xBC, 0x61, 0xEF, 0xCC, 0x11, 0xE5, 0x72, 0x1D,
-   0x3D, 0x7C, 0xEB, 0xE8, 0xE9, 0x3C, 0xEA, 0x8F, 0x7D, 0x9F, 0xEC, 0x75,
-   0x1E, 0xF5, 0x3E, 0x38, 0xF6, 0xD9, 0x3F, 0xCF, 0x76, 0xFA, 0x1F, 0x84,
-   0xA0, 0x70, 0xED, 0x14, 0x90, 0xB3, 0x7E, 0x58, 0xFB, 0xE2, 0x20, 0x64,
-   0xD0, 0xDD, 0x77, 0xAD, 0xDA, 0xC5, 0x40, 0xF2, 0x39, 0xB0, 0xF7, 0x49,
-   0xB4, 0x0B, 0x7F, 0x51, 0x15, 0x43, 0x91, 0x10, 0x71, 0xBB, 0xEE, 0xBF,
-   0x85, 0xC8, 0xA1
-};
-
-static const u8 exp_to_poly[492] = {
-   0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x4D, 0x9A, 0x79, 0xF2,
-   0xA9, 0x1F, 0x3E, 0x7C, 0xF8, 0xBD, 0x37, 0x6E, 0xDC, 0xF5, 0xA7, 0x03,
-   0x06, 0x0C, 0x18, 0x30, 0x60, 0xC0, 0xCD, 0xD7, 0xE3, 0x8B, 0x5B, 0xB6,
-   0x21, 0x42, 0x84, 0x45, 0x8A, 0x59, 0xB2, 0x29, 0x52, 0xA4, 0x05, 0x0A,
-   0x14, 0x28, 0x50, 0xA0, 0x0D, 0x1A, 0x34, 0x68, 0xD0, 0xED, 0x97, 0x63,
-   0xC6, 0xC1, 0xCF, 0xD3, 0xEB, 0x9B, 0x7B, 0xF6, 0xA1, 0x0F, 0x1E, 0x3C,
-   0x78, 0xF0, 0xAD, 0x17, 0x2E, 0x5C, 0xB8, 0x3D, 0x7A, 0xF4, 0xA5, 0x07,
-   0x0E, 0x1C, 0x38, 0x70, 0xE0, 0x8D, 0x57, 0xAE, 0x11, 0x22, 0x44, 0x88,
-   0x5D, 0xBA, 0x39, 0x72, 0xE4, 0x85, 0x47, 0x8E, 0x51, 0xA2, 0x09, 0x12,
-   0x24, 0x48, 0x90, 0x6D, 0xDA, 0xF9, 0xBF, 0x33, 0x66, 0xCC, 0xD5, 0xE7,
-   0x83, 0x4B, 0x96, 0x61, 0xC2, 0xC9, 0xDF, 0xF3, 0xAB, 0x1B, 0x36, 0x6C,
-   0xD8, 0xFD, 0xB7, 0x23, 0x46, 0x8C, 0x55, 0xAA, 0x19, 0x32, 0x64, 0xC8,
-   0xDD, 0xF7, 0xA3, 0x0B, 0x16, 0x2C, 0x58, 0xB0, 0x2D, 0x5A, 0xB4, 0x25,
-   0x4A, 0x94, 0x65, 0xCA, 0xD9, 0xFF, 0xB3, 0x2B, 0x56, 0xAC, 0x15, 0x2A,
-   0x54, 0xA8, 0x1D, 0x3A, 0x74, 0xE8, 0x9D, 0x77, 0xEE, 0x91, 0x6F, 0xDE,
-   0xF1, 0xAF, 0x13, 0x26, 0x4C, 0x98, 0x7D, 0xFA, 0xB9, 0x3F, 0x7E, 0xFC,
-   0xB5, 0x27, 0x4E, 0x9C, 0x75, 0xEA, 0x99, 0x7F, 0xFE, 0xB1, 0x2F, 0x5E,
-   0xBC, 0x35, 0x6A, 0xD4, 0xE5, 0x87, 0x43, 0x86, 0x41, 0x82, 0x49, 0x92,
-   0x69, 0xD2, 0xE9, 0x9F, 0x73, 0xE6, 0x81, 0x4F, 0x9E, 0x71, 0xE2, 0x89,
-   0x5F, 0xBE, 0x31, 0x62, 0xC4, 0xC5, 0xC7, 0xC3, 0xCB, 0xDB, 0xFB, 0xBB,
-   0x3B, 0x76, 0xEC, 0x95, 0x67, 0xCE, 0xD1, 0xEF, 0x93, 0x6B, 0xD6, 0xE1,
-   0x8F, 0x53, 0xA6, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x4D,
-   0x9A, 0x79, 0xF2, 0xA9, 0x1F, 0x3E, 0x7C, 0xF8, 0xBD, 0x37, 0x6E, 0xDC,
-   0xF5, 0xA7, 0x03, 0x06, 0x0C, 0x18, 0x30, 0x60, 0xC0, 0xCD, 0xD7, 0xE3,
-   0x8B, 0x5B, 0xB6, 0x21, 0x42, 0x84, 0x45, 0x8A, 0x59, 0xB2, 0x29, 0x52,
-   0xA4, 0x05, 0x0A, 0x14, 0x28, 0x50, 0xA0, 0x0D, 0x1A, 0x34, 0x68, 0xD0,
-   0xED, 0x97, 0x63, 0xC6, 0xC1, 0xCF, 0xD3, 0xEB, 0x9B, 0x7B, 0xF6, 0xA1,
-   0x0F, 0x1E, 0x3C, 0x78, 0xF0, 0xAD, 0x17, 0x2E, 0x5C, 0xB8, 0x3D, 0x7A,
-   0xF4, 0xA5, 0x07, 0x0E, 0x1C, 0x38, 0x70, 0xE0, 0x8D, 0x57, 0xAE, 0x11,
-   0x22, 0x44, 0x88, 0x5D, 0xBA, 0x39, 0x72, 0xE4, 0x85, 0x47, 0x8E, 0x51,
-   0xA2, 0x09, 0x12, 0x24, 0x48, 0x90, 0x6D, 0xDA, 0xF9, 0xBF, 0x33, 0x66,
-   0xCC, 0xD5, 0xE7, 0x83, 0x4B, 0x96, 0x61, 0xC2, 0xC9, 0xDF, 0xF3, 0xAB,
-   0x1B, 0x36, 0x6C, 0xD8, 0xFD, 0xB7, 0x23, 0x46, 0x8C, 0x55, 0xAA, 0x19,
-   0x32, 0x64, 0xC8, 0xDD, 0xF7, 0xA3, 0x0B, 0x16, 0x2C, 0x58, 0xB0, 0x2D,
-   0x5A, 0xB4, 0x25, 0x4A, 0x94, 0x65, 0xCA, 0xD9, 0xFF, 0xB3, 0x2B, 0x56,
-   0xAC, 0x15, 0x2A, 0x54, 0xA8, 0x1D, 0x3A, 0x74, 0xE8, 0x9D, 0x77, 0xEE,
-   0x91, 0x6F, 0xDE, 0xF1, 0xAF, 0x13, 0x26, 0x4C, 0x98, 0x7D, 0xFA, 0xB9,
-   0x3F, 0x7E, 0xFC, 0xB5, 0x27, 0x4E, 0x9C, 0x75, 0xEA, 0x99, 0x7F, 0xFE,
-   0xB1, 0x2F, 0x5E, 0xBC, 0x35, 0x6A, 0xD4, 0xE5, 0x87, 0x43, 0x86, 0x41,
-   0x82, 0x49, 0x92, 0x69, 0xD2, 0xE9, 0x9F, 0x73, 0xE6, 0x81, 0x4F, 0x9E,
-   0x71, 0xE2, 0x89, 0x5F, 0xBE, 0x31, 0x62, 0xC4, 0xC5, 0xC7, 0xC3, 0xCB
-};
-
-
-/* The table constants are indices of
- * S-box entries, preprocessed through q0 and q1. */
-static const u8 calc_sb_tbl[512] = {
-    0xA9, 0x75, 0x67, 0xF3, 0xB3, 0xC6, 0xE8, 0xF4,
-    0x04, 0xDB, 0xFD, 0x7B, 0xA3, 0xFB, 0x76, 0xC8,
-    0x9A, 0x4A, 0x92, 0xD3, 0x80, 0xE6, 0x78, 0x6B,
-    0xE4, 0x45, 0xDD, 0x7D, 0xD1, 0xE8, 0x38, 0x4B,
-    0x0D, 0xD6, 0xC6, 0x32, 0x35, 0xD8, 0x98, 0xFD,
-    0x18, 0x37, 0xF7, 0x71, 0xEC, 0xF1, 0x6C, 0xE1,
-    0x43, 0x30, 0x75, 0x0F, 0x37, 0xF8, 0x26, 0x1B,
-    0xFA, 0x87, 0x13, 0xFA, 0x94, 0x06, 0x48, 0x3F,
-    0xF2, 0x5E, 0xD0, 0xBA, 0x8B, 0xAE, 0x30, 0x5B,
-    0x84, 0x8A, 0x54, 0x00, 0xDF, 0xBC, 0x23, 0x9D,
-    0x19, 0x6D, 0x5B, 0xC1, 0x3D, 0xB1, 0x59, 0x0E,
-    0xF3, 0x80, 0xAE, 0x5D, 0xA2, 0xD2, 0x82, 0xD5,
-    0x63, 0xA0, 0x01, 0x84, 0x83, 0x07, 0x2E, 0x14,
-    0xD9, 0xB5, 0x51, 0x90, 0x9B, 0x2C, 0x7C, 0xA3,
-    0xA6, 0xB2, 0xEB, 0x73, 0xA5, 0x4C, 0xBE, 0x54,
-    0x16, 0x92, 0x0C, 0x74, 0xE3, 0x36, 0x61, 0x51,
-    0xC0, 0x38, 0x8C, 0xB0, 0x3A, 0xBD, 0xF5, 0x5A,
-    0x73, 0xFC, 0x2C, 0x60, 0x25, 0x62, 0x0B, 0x96,
-    0xBB, 0x6C, 0x4E, 0x42, 0x89, 0xF7, 0x6B, 0x10,
-    0x53, 0x7C, 0x6A, 0x28, 0xB4, 0x27, 0xF1, 0x8C,
-    0xE1, 0x13, 0xE6, 0x95, 0xBD, 0x9C, 0x45, 0xC7,
-    0xE2, 0x24, 0xF4, 0x46, 0xB6, 0x3B, 0x66, 0x70,
-    0xCC, 0xCA, 0x95, 0xE3, 0x03, 0x85, 0x56, 0xCB,
-    0xD4, 0x11, 0x1C, 0xD0, 0x1E, 0x93, 0xD7, 0xB8,
-    0xFB, 0xA6, 0xC3, 0x83, 0x8E, 0x20, 0xB5, 0xFF,
-    0xE9, 0x9F, 0xCF, 0x77, 0xBF, 0xC3, 0xBA, 0xCC,
-    0xEA, 0x03, 0x77, 0x6F, 0x39, 0x08, 0xAF, 0xBF,
-    0x33, 0x40, 0xC9, 0xE7, 0x62, 0x2B, 0x71, 0xE2,
-    0x81, 0x79, 0x79, 0x0C, 0x09, 0xAA, 0xAD, 0x82,
-    0x24, 0x41, 0xCD, 0x3A, 0xF9, 0xEA, 0xD8, 0xB9,
-    0xE5, 0xE4, 0xC5, 0x9A, 0xB9, 0xA4, 0x4D, 0x97,
-    0x44, 0x7E, 0x08, 0xDA, 0x86, 0x7A, 0xE7, 0x17,
-    0xA1, 0x66, 0x1D, 0x94, 0xAA, 0xA1, 0xED, 0x1D,
-    0x06, 0x3D, 0x70, 0xF0, 0xB2, 0xDE, 0xD2, 0xB3,
-    0x41, 0x0B, 0x7B, 0x72, 0xA0, 0xA7, 0x11, 0x1C,
-    0x31, 0xEF, 0xC2, 0xD1, 0x27, 0x53, 0x90, 0x3E,
-    0x20, 0x8F, 0xF6, 0x33, 0x60, 0x26, 0xFF, 0x5F,
-    0x96, 0xEC, 0x5C, 0x76, 0xB1, 0x2A, 0xAB, 0x49,
-    0x9E, 0x81, 0x9C, 0x88, 0x52, 0xEE, 0x1B, 0x21,
-    0x5F, 0xC4, 0x93, 0x1A, 0x0A, 0xEB, 0xEF, 0xD9,
-    0x91, 0xC5, 0x85, 0x39, 0x49, 0x99, 0xEE, 0xCD,
-    0x2D, 0xAD, 0x4F, 0x31, 0x8F, 0x8B, 0x3B, 0x01,
-    0x47, 0x18, 0x87, 0x23, 0x6D, 0xDD, 0x46, 0x1F,
-    0xD6, 0x4E, 0x3E, 0x2D, 0x69, 0xF9, 0x64, 0x48,
-    0x2A, 0x4F, 0xCE, 0xF2, 0xCB, 0x65, 0x2F, 0x8E,
-    0xFC, 0x78, 0x97, 0x5C, 0x05, 0x58, 0x7A, 0x19,
-    0xAC, 0x8D, 0x7F, 0xE5, 0xD5, 0x98, 0x1A, 0x57,
-    0x4B, 0x67, 0x0E, 0x7F, 0xA7, 0x05, 0x5A, 0x64,
-    0x28, 0xAF, 0x14, 0x63, 0x3F, 0xB6, 0x29, 0xFE,
-    0x88, 0xF5, 0x3C, 0xB7, 0x4C, 0x3C, 0x02, 0xA5,
-    0xB8, 0xCE, 0xDA, 0xE9, 0xB0, 0x68, 0x17, 0x44,
-    0x55, 0xE0, 0x1F, 0x4D, 0x8A, 0x43, 0x7D, 0x69,
-    0x57, 0x29, 0xC7, 0x2E, 0x8D, 0xAC, 0x74, 0x15,
-    0xB7, 0x59, 0xC4, 0xA8, 0x9F, 0x0A, 0x72, 0x9E,
-    0x7E, 0x6E, 0x15, 0x47, 0x22, 0xDF, 0x12, 0x34,
-    0x58, 0x35, 0x07, 0x6A, 0x99, 0xCF, 0x34, 0xDC,
-    0x6E, 0x22, 0x50, 0xC9, 0xDE, 0xC0, 0x68, 0x9B,
-    0x65, 0x89, 0xBC, 0xD4, 0xDB, 0xED, 0xF8, 0xAB,
-    0xC8, 0x12, 0xA8, 0xA2, 0x2B, 0x0D, 0x40, 0x52,
-    0xDC, 0xBB, 0xFE, 0x02, 0x32, 0x2F, 0xA4, 0xA9,
-    0xCA, 0xD7, 0x10, 0x61, 0x21, 0x1E, 0xF0, 0xB4,
-    0xD3, 0x50, 0x5D, 0x04, 0x0F, 0xF6, 0x00, 0xC2,
-    0x6F, 0x16, 0x9D, 0x25, 0x36, 0x86, 0x42, 0x56,
-    0x4A, 0x55, 0x5E, 0x09, 0xC1, 0xBE, 0xE0, 0x91
-};
-
-/* Macro to perform one column of the RS matrix multiplication.  The
- * parameters a, b, c, and d are the four bytes of output; i is the index
- * of the key bytes, and w, x, y, and z, are the column of constants from
- * the RS matrix, preprocessed through the poly_to_exp table. */
-
-#define CALC_S(a, b, c, d, i, w, x, y, z) \
-   if (key[i]) { \
-      tmp = poly_to_exp[key[i] - 1]; \
-      (a) ^= exp_to_poly[tmp + (w)]; \
-      (b) ^= exp_to_poly[tmp + (x)]; \
-      (c) ^= exp_to_poly[tmp + (y)]; \
-      (d) ^= exp_to_poly[tmp + (z)]; \
-   }
-
-/* Macros to calculate the key-dependent S-boxes for a 128-bit key using
- * the S vector from CALC_S.  CALC_SB_2 computes a single entry in all
- * four S-boxes, where i is the index of the entry to compute, and a and b
- * are the index numbers preprocessed through the q0 and q1 tables
- * respectively. */
-
-#define CALC_SB_2(i, a, b) \
-   ctx->s[0][i] = mds[0][q0[(a) ^ sa] ^ se]; \
-   ctx->s[1][i] = mds[1][q0[(b) ^ sb] ^ sf]; \
-   ctx->s[2][i] = mds[2][q1[(a) ^ sc] ^ sg]; \
-   ctx->s[3][i] = mds[3][q1[(b) ^ sd] ^ sh]
-
-/* Macro exactly like CALC_SB_2, but for 192-bit keys. */
-
-#define CALC_SB192_2(i, a, b) \
-   ctx->s[0][i] = mds[0][q0[q0[(b) ^ sa] ^ se] ^ si]; \
-   ctx->s[1][i] = mds[1][q0[q1[(b) ^ sb] ^ sf] ^ sj]; \
-   ctx->s[2][i] = mds[2][q1[q0[(a) ^ sc] ^ sg] ^ sk]; \
-   ctx->s[3][i] = mds[3][q1[q1[(a) ^ sd] ^ sh] ^ sl];
-
-/* Macro exactly like CALC_SB_2, but for 256-bit keys. */
-
-#define CALC_SB256_2(i, a, b) \
-   ctx->s[0][i] = mds[0][q0[q0[q1[(b) ^ sa] ^ se] ^ si] ^ sm]; \
-   ctx->s[1][i] = mds[1][q0[q1[q1[(a) ^ sb] ^ sf] ^ sj] ^ sn]; \
-   ctx->s[2][i] = mds[2][q1[q0[q0[(a) ^ sc] ^ sg] ^ sk] ^ so]; \
-   ctx->s[3][i] = mds[3][q1[q1[q0[(b) ^ sd] ^ sh] ^ sl] ^ sp];
-
-/* Macros to calculate the whitening and round subkeys.  CALC_K_2 computes the
- * last two stages of the h() function for a given index (either 2i or 2i+1).
- * a, b, c, and d are the four bytes going into the last two stages.  For
- * 128-bit keys, this is the entire h() function and a and c are the index
- * preprocessed through q0 and q1 respectively; for longer keys they are the
- * output of previous stages.  j is the index of the first key byte to use.
- * CALC_K computes a pair of subkeys for 128-bit Twofish, by calling CALC_K_2
- * twice, doing the Pseudo-Hadamard Transform, and doing the necessary
- * rotations.  Its parameters are: a, the array to write the results into,
- * j, the index of the first output entry, k and l, the preprocessed indices
- * for index 2i, and m and n, the preprocessed indices for index 2i+1.
- * CALC_K192_2 expands CALC_K_2 to handle 192-bit keys, by doing an
- * additional lookup-and-XOR stage.  The parameters a, b, c and d are the
- * four bytes going into the last three stages.  For 192-bit keys, c = d
- * are the index preprocessed through q0, and a = b are the index
- * preprocessed through q1; j is the index of the first key byte to use.
- * CALC_K192 is identical to CALC_K but for using the CALC_K192_2 macro
- * instead of CALC_K_2.
- * CALC_K256_2 expands CALC_K192_2 to handle 256-bit keys, by doing an
- * additional lookup-and-XOR stage.  The parameters a and b are the index
- * preprocessed through q0 and q1 respectively; j is the index of the first
- * key byte to use.  CALC_K256 is identical to CALC_K but for using the
- * CALC_K256_2 macro instead of CALC_K_2. */
-
-#define CALC_K_2(a, b, c, d, j) \
-     mds[0][q0[a ^ key[(j) + 8]] ^ key[j]] \
-   ^ mds[1][q0[b ^ key[(j) + 9]] ^ key[(j) + 1]] \
-   ^ mds[2][q1[c ^ key[(j) + 10]] ^ key[(j) + 2]] \
-   ^ mds[3][q1[d ^ key[(j) + 11]] ^ key[(j) + 3]]
-
-#define CALC_K(a, j, k, l, m, n) \
-   x = CALC_K_2 (k, l, k, l, 0); \
-   y = CALC_K_2 (m, n, m, n, 4); \
-   y = rol32(y, 8); \
-   x += y; y += x; ctx->a[j] = x; \
-   ctx->a[(j) + 1] = rol32(y, 9)
-
-#define CALC_K192_2(a, b, c, d, j) \
-   CALC_K_2 (q0[a ^ key[(j) + 16]], \
-	     q1[b ^ key[(j) + 17]], \
-	     q0[c ^ key[(j) + 18]], \
-	     q1[d ^ key[(j) + 19]], j)
-
-#define CALC_K192(a, j, k, l, m, n) \
-   x = CALC_K192_2 (l, l, k, k, 0); \
-   y = CALC_K192_2 (n, n, m, m, 4); \
-   y = rol32(y, 8); \
-   x += y; y += x; ctx->a[j] = x; \
-   ctx->a[(j) + 1] = rol32(y, 9)
-
-#define CALC_K256_2(a, b, j) \
-   CALC_K192_2 (q1[b ^ key[(j) + 24]], \
-	        q1[a ^ key[(j) + 25]], \
-	        q0[a ^ key[(j) + 26]], \
-	        q0[b ^ key[(j) + 27]], j)
-
-#define CALC_K256(a, j, k, l, m, n) \
-   x = CALC_K256_2 (k, l, 0); \
-   y = CALC_K256_2 (m, n, 4); \
-   y = rol32(y, 8); \
-   x += y; y += x; ctx->a[j] = x; \
-   ctx->a[(j) + 1] = rol32(y, 9)
-
-
 /* Macros to compute the g() function in the encryption and decryption
  * rounds.  G1 is the straight g() function; G2 includes the 8-bit
  * rotation for the high 32-bit word. */
@@ -630,176 +103,7 @@ static const u8 calc_sb_tbl[512] = {
    x ^= ctx->w[m]; \
    dst[n] = cpu_to_le32(x)
 
-#define TF_MIN_KEY_SIZE 16
-#define TF_MAX_KEY_SIZE 32
-#define TF_BLOCK_SIZE 16
-
-/* Structure for an expanded Twofish key.  s contains the key-dependent
- * S-boxes composed with the MDS matrix; w contains the eight "whitening"
- * subkeys, K[0] through K[7].	k holds the remaining, "round" subkeys.  Note
- * that k[i] corresponds to what the Twofish paper calls K[i+8]. */
-struct twofish_ctx {
-   u32 s[4][256], w[8], k[32];
-};
-
-/* Perform the key setup. */
-static int twofish_setkey(struct crypto_tfm *tfm, const u8 *key,
-			  unsigned int key_len, u32 *flags)
-{
-	
-	struct twofish_ctx *ctx = crypto_tfm_ctx(tfm);
 
-	int i, j, k;
-
-	/* Temporaries for CALC_K. */
-	u32 x, y;
-
-	/* The S vector used to key the S-boxes, split up into individual bytes.
-	 * 128-bit keys use only sa through sh; 256-bit use all of them. */
-	u8 sa = 0, sb = 0, sc = 0, sd = 0, se = 0, sf = 0, sg = 0, sh = 0;
-	u8 si = 0, sj = 0, sk = 0, sl = 0, sm = 0, sn = 0, so = 0, sp = 0;
-
-	/* Temporary for CALC_S. */
-	u8 tmp;
-
-	/* Check key length. */
-	if (key_len != 16 && key_len != 24 && key_len != 32)
-	{
-		*flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
-		return -EINVAL; /* unsupported key length */
-	}
-
-	/* Compute the first two words of the S vector.  The magic numbers are
-	 * the entries of the RS matrix, preprocessed through poly_to_exp. The
-	 * numbers in the comments are the original (polynomial form) matrix
-	 * entries. */
-	CALC_S (sa, sb, sc, sd, 0, 0x00, 0x2D, 0x01, 0x2D); /* 01 A4 02 A4 */
-	CALC_S (sa, sb, sc, sd, 1, 0x2D, 0xA4, 0x44, 0x8A); /* A4 56 A1 55 */
-	CALC_S (sa, sb, sc, sd, 2, 0x8A, 0xD5, 0xBF, 0xD1); /* 55 82 FC 87 */
-	CALC_S (sa, sb, sc, sd, 3, 0xD1, 0x7F, 0x3D, 0x99); /* 87 F3 C1 5A */
-	CALC_S (sa, sb, sc, sd, 4, 0x99, 0x46, 0x66, 0x96); /* 5A 1E 47 58 */
-	CALC_S (sa, sb, sc, sd, 5, 0x96, 0x3C, 0x5B, 0xED); /* 58 C6 AE DB */
-	CALC_S (sa, sb, sc, sd, 6, 0xED, 0x37, 0x4F, 0xE0); /* DB 68 3D 9E */
-	CALC_S (sa, sb, sc, sd, 7, 0xE0, 0xD0, 0x8C, 0x17); /* 9E E5 19 03 */
-	CALC_S (se, sf, sg, sh, 8, 0x00, 0x2D, 0x01, 0x2D); /* 01 A4 02 A4 */
-	CALC_S (se, sf, sg, sh, 9, 0x2D, 0xA4, 0x44, 0x8A); /* A4 56 A1 55 */
-	CALC_S (se, sf, sg, sh, 10, 0x8A, 0xD5, 0xBF, 0xD1); /* 55 82 FC 87 */
-	CALC_S (se, sf, sg, sh, 11, 0xD1, 0x7F, 0x3D, 0x99); /* 87 F3 C1 5A */
-	CALC_S (se, sf, sg, sh, 12, 0x99, 0x46, 0x66, 0x96); /* 5A 1E 47 58 */
-	CALC_S (se, sf, sg, sh, 13, 0x96, 0x3C, 0x5B, 0xED); /* 58 C6 AE DB */
-	CALC_S (se, sf, sg, sh, 14, 0xED, 0x37, 0x4F, 0xE0); /* DB 68 3D 9E */
-	CALC_S (se, sf, sg, sh, 15, 0xE0, 0xD0, 0x8C, 0x17); /* 9E E5 19 03 */
-
-	if (key_len == 24 || key_len == 32) { /* 192- or 256-bit key */
-		/* Calculate the third word of the S vector */
-		CALC_S (si, sj, sk, sl, 16, 0x00, 0x2D, 0x01, 0x2D); /* 01 A4 02 A4 */
-		CALC_S (si, sj, sk, sl, 17, 0x2D, 0xA4, 0x44, 0x8A); /* A4 56 A1 55 */
-		CALC_S (si, sj, sk, sl, 18, 0x8A, 0xD5, 0xBF, 0xD1); /* 55 82 FC 87 */
-		CALC_S (si, sj, sk, sl, 19, 0xD1, 0x7F, 0x3D, 0x99); /* 87 F3 C1 5A */
-		CALC_S (si, sj, sk, sl, 20, 0x99, 0x46, 0x66, 0x96); /* 5A 1E 47 58 */
-		CALC_S (si, sj, sk, sl, 21, 0x96, 0x3C, 0x5B, 0xED); /* 58 C6 AE DB */
-		CALC_S (si, sj, sk, sl, 22, 0xED, 0x37, 0x4F, 0xE0); /* DB 68 3D 9E */
-		CALC_S (si, sj, sk, sl, 23, 0xE0, 0xD0, 0x8C, 0x17); /* 9E E5 19 03 */
-	}
-
-	if (key_len == 32) { /* 256-bit key */
-		/* Calculate the fourth word of the S vector */
-		CALC_S (sm, sn, so, sp, 24, 0x00, 0x2D, 0x01, 0x2D); /* 01 A4 02 A4 */
-		CALC_S (sm, sn, so, sp, 25, 0x2D, 0xA4, 0x44, 0x8A); /* A4 56 A1 55 */
-		CALC_S (sm, sn, so, sp, 26, 0x8A, 0xD5, 0xBF, 0xD1); /* 55 82 FC 87 */
-		CALC_S (sm, sn, so, sp, 27, 0xD1, 0x7F, 0x3D, 0x99); /* 87 F3 C1 5A */
-		CALC_S (sm, sn, so, sp, 28, 0x99, 0x46, 0x66, 0x96); /* 5A 1E 47 58 */
-		CALC_S (sm, sn, so, sp, 29, 0x96, 0x3C, 0x5B, 0xED); /* 58 C6 AE DB */
-		CALC_S (sm, sn, so, sp, 30, 0xED, 0x37, 0x4F, 0xE0); /* DB 68 3D 9E */
-		CALC_S (sm, sn, so, sp, 31, 0xE0, 0xD0, 0x8C, 0x17); /* 9E E5 19 03 */
-
-		/* Compute the S-boxes. */
-		for ( i = j = 0, k = 1; i < 256; i++, j += 2, k += 2 ) {
-			CALC_SB256_2( i, calc_sb_tbl[j], calc_sb_tbl[k] );
-		}
-
-		/* Calculate whitening and round subkeys.  The constants are
-		 * indices of subkeys, preprocessed through q0 and q1. */
-		CALC_K256 (w, 0, 0xA9, 0x75, 0x67, 0xF3);
-		CALC_K256 (w, 2, 0xB3, 0xC6, 0xE8, 0xF4);
-		CALC_K256 (w, 4, 0x04, 0xDB, 0xFD, 0x7B);
-		CALC_K256 (w, 6, 0xA3, 0xFB, 0x76, 0xC8);
-		CALC_K256 (k, 0, 0x9A, 0x4A, 0x92, 0xD3);
-		CALC_K256 (k, 2, 0x80, 0xE6, 0x78, 0x6B);
-		CALC_K256 (k, 4, 0xE4, 0x45, 0xDD, 0x7D);
-		CALC_K256 (k, 6, 0xD1, 0xE8, 0x38, 0x4B);
-		CALC_K256 (k, 8, 0x0D, 0xD6, 0xC6, 0x32);
-		CALC_K256 (k, 10, 0x35, 0xD8, 0x98, 0xFD);
-		CALC_K256 (k, 12, 0x18, 0x37, 0xF7, 0x71);
-		CALC_K256 (k, 14, 0xEC, 0xF1, 0x6C, 0xE1);
-		CALC_K256 (k, 16, 0x43, 0x30, 0x75, 0x0F);
-		CALC_K256 (k, 18, 0x37, 0xF8, 0x26, 0x1B);
-		CALC_K256 (k, 20, 0xFA, 0x87, 0x13, 0xFA);
-		CALC_K256 (k, 22, 0x94, 0x06, 0x48, 0x3F);
-		CALC_K256 (k, 24, 0xF2, 0x5E, 0xD0, 0xBA);
-		CALC_K256 (k, 26, 0x8B, 0xAE, 0x30, 0x5B);
-		CALC_K256 (k, 28, 0x84, 0x8A, 0x54, 0x00);
-		CALC_K256 (k, 30, 0xDF, 0xBC, 0x23, 0x9D);
-	} else if (key_len == 24) { /* 192-bit key */
-		/* Compute the S-boxes. */
-		for ( i = j = 0, k = 1; i < 256; i++, j += 2, k += 2 ) {
-		        CALC_SB192_2( i, calc_sb_tbl[j], calc_sb_tbl[k] );
-		}
-
-		/* Calculate whitening and round subkeys.  The constants are
-		 * indices of subkeys, preprocessed through q0 and q1. */
-		CALC_K192 (w, 0, 0xA9, 0x75, 0x67, 0xF3);
-		CALC_K192 (w, 2, 0xB3, 0xC6, 0xE8, 0xF4);
-		CALC_K192 (w, 4, 0x04, 0xDB, 0xFD, 0x7B);
-		CALC_K192 (w, 6, 0xA3, 0xFB, 0x76, 0xC8);
-		CALC_K192 (k, 0, 0x9A, 0x4A, 0x92, 0xD3);
-		CALC_K192 (k, 2, 0x80, 0xE6, 0x78, 0x6B);
-		CALC_K192 (k, 4, 0xE4, 0x45, 0xDD, 0x7D);
-		CALC_K192 (k, 6, 0xD1, 0xE8, 0x38, 0x4B);
-		CALC_K192 (k, 8, 0x0D, 0xD6, 0xC6, 0x32);
-		CALC_K192 (k, 10, 0x35, 0xD8, 0x98, 0xFD);
-		CALC_K192 (k, 12, 0x18, 0x37, 0xF7, 0x71);
-		CALC_K192 (k, 14, 0xEC, 0xF1, 0x6C, 0xE1);
-		CALC_K192 (k, 16, 0x43, 0x30, 0x75, 0x0F);
-		CALC_K192 (k, 18, 0x37, 0xF8, 0x26, 0x1B);
-		CALC_K192 (k, 20, 0xFA, 0x87, 0x13, 0xFA);
-		CALC_K192 (k, 22, 0x94, 0x06, 0x48, 0x3F);
-		CALC_K192 (k, 24, 0xF2, 0x5E, 0xD0, 0xBA);
-		CALC_K192 (k, 26, 0x8B, 0xAE, 0x30, 0x5B);
-		CALC_K192 (k, 28, 0x84, 0x8A, 0x54, 0x00);
-		CALC_K192 (k, 30, 0xDF, 0xBC, 0x23, 0x9D);
-	} else { /* 128-bit key */
-		/* Compute the S-boxes. */
-		for ( i = j = 0, k = 1; i < 256; i++, j += 2, k += 2 ) {
-			CALC_SB_2( i, calc_sb_tbl[j], calc_sb_tbl[k] );
-		}
-
-		/* Calculate whitening and round subkeys.  The constants are
-		 * indices of subkeys, preprocessed through q0 and q1. */
-		CALC_K (w, 0, 0xA9, 0x75, 0x67, 0xF3);
-		CALC_K (w, 2, 0xB3, 0xC6, 0xE8, 0xF4);
-		CALC_K (w, 4, 0x04, 0xDB, 0xFD, 0x7B);
-		CALC_K (w, 6, 0xA3, 0xFB, 0x76, 0xC8);
-		CALC_K (k, 0, 0x9A, 0x4A, 0x92, 0xD3);
-		CALC_K (k, 2, 0x80, 0xE6, 0x78, 0x6B);
-		CALC_K (k, 4, 0xE4, 0x45, 0xDD, 0x7D);
-		CALC_K (k, 6, 0xD1, 0xE8, 0x38, 0x4B);
-		CALC_K (k, 8, 0x0D, 0xD6, 0xC6, 0x32);
-		CALC_K (k, 10, 0x35, 0xD8, 0x98, 0xFD);
-		CALC_K (k, 12, 0x18, 0x37, 0xF7, 0x71);
-		CALC_K (k, 14, 0xEC, 0xF1, 0x6C, 0xE1);
-		CALC_K (k, 16, 0x43, 0x30, 0x75, 0x0F);
-		CALC_K (k, 18, 0x37, 0xF8, 0x26, 0x1B);
-		CALC_K (k, 20, 0xFA, 0x87, 0x13, 0xFA);
-		CALC_K (k, 22, 0x94, 0x06, 0x48, 0x3F);
-		CALC_K (k, 24, 0xF2, 0x5E, 0xD0, 0xBA);
-		CALC_K (k, 26, 0x8B, 0xAE, 0x30, 0x5B);
-		CALC_K (k, 28, 0x84, 0x8A, 0x54, 0x00);
-		CALC_K (k, 30, 0xDF, 0xBC, 0x23, 0x9D);
-	}
-
-	return 0;
-}
 
 /* Encrypt one block.  in and out may be the same. */
 static void twofish_encrypt(struct crypto_tfm *tfm, u8 *out, const u8 *in)
diff --git a/crypto/twofish_common.c b/crypto/twofish_common.c
new file mode 100644
index 000000000000..1ae0280c2513
--- /dev/null
+++ b/crypto/twofish_common.c
@@ -0,0 +1,744 @@
+/*
+ * Common Twofish algorithm parts shared between the c and assembler
+ * implementations
+ *
+ * Originally Twofish for GPG
+ * By Matthew Skala <mskala@ansuz.sooke.bc.ca>, July 26, 1998
+ * 256-bit key length added March 20, 1999
+ * Some modifications to reduce the text size by Werner Koch, April, 1998
+ * Ported to the kerneli patch by Marc Mutz <Marc@Mutz.com>
+ * Ported to CryptoAPI by Colin Slater <hoho@tacomeat.net>
+ *
+ * The original author has disclaimed all copyright interest in this
+ * code and thus put it in the public domain. The subsequent authors
+ * have put this under the GNU General Public License.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307
+ * USA
+ *
+ * This code is a "clean room" implementation, written from the paper
+ * _Twofish: A 128-Bit Block Cipher_ by Bruce Schneier, John Kelsey,
+ * Doug Whiting, David Wagner, Chris Hall, and Niels Ferguson, available
+ * through http://www.counterpane.com/twofish.html
+ *
+ * For background information on multiplication in finite fields, used for
+ * the matrix operations in the key schedule, see the book _Contemporary
+ * Abstract Algebra_ by Joseph A. Gallian, especially chapter 22 in the
+ * Third Edition.
+ */
+
+#include <crypto/twofish.h>
+#include <linux/bitops.h>
+#include <linux/crypto.h>
+#include <linux/errno.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/types.h>
+
+
+/* The large precomputed tables for the Twofish cipher (twofish.c)
+ * Taken from the same source as twofish.c
+ * Marc Mutz <Marc@Mutz.com>
+ */
+
+/* These two tables are the q0 and q1 permutations, exactly as described in
+ * the Twofish paper. */
+
+static const u8 q0[256] = {
+	0xA9, 0x67, 0xB3, 0xE8, 0x04, 0xFD, 0xA3, 0x76, 0x9A, 0x92, 0x80, 0x78,
+	0xE4, 0xDD, 0xD1, 0x38, 0x0D, 0xC6, 0x35, 0x98, 0x18, 0xF7, 0xEC, 0x6C,
+	0x43, 0x75, 0x37, 0x26, 0xFA, 0x13, 0x94, 0x48, 0xF2, 0xD0, 0x8B, 0x30,
+	0x84, 0x54, 0xDF, 0x23, 0x19, 0x5B, 0x3D, 0x59, 0xF3, 0xAE, 0xA2, 0x82,
+	0x63, 0x01, 0x83, 0x2E, 0xD9, 0x51, 0x9B, 0x7C, 0xA6, 0xEB, 0xA5, 0xBE,
+	0x16, 0x0C, 0xE3, 0x61, 0xC0, 0x8C, 0x3A, 0xF5, 0x73, 0x2C, 0x25, 0x0B,
+	0xBB, 0x4E, 0x89, 0x6B, 0x53, 0x6A, 0xB4, 0xF1, 0xE1, 0xE6, 0xBD, 0x45,
+	0xE2, 0xF4, 0xB6, 0x66, 0xCC, 0x95, 0x03, 0x56, 0xD4, 0x1C, 0x1E, 0xD7,
+	0xFB, 0xC3, 0x8E, 0xB5, 0xE9, 0xCF, 0xBF, 0xBA, 0xEA, 0x77, 0x39, 0xAF,
+	0x33, 0xC9, 0x62, 0x71, 0x81, 0x79, 0x09, 0xAD, 0x24, 0xCD, 0xF9, 0xD8,
+	0xE5, 0xC5, 0xB9, 0x4D, 0x44, 0x08, 0x86, 0xE7, 0xA1, 0x1D, 0xAA, 0xED,
+	0x06, 0x70, 0xB2, 0xD2, 0x41, 0x7B, 0xA0, 0x11, 0x31, 0xC2, 0x27, 0x90,
+	0x20, 0xF6, 0x60, 0xFF, 0x96, 0x5C, 0xB1, 0xAB, 0x9E, 0x9C, 0x52, 0x1B,
+	0x5F, 0x93, 0x0A, 0xEF, 0x91, 0x85, 0x49, 0xEE, 0x2D, 0x4F, 0x8F, 0x3B,
+	0x47, 0x87, 0x6D, 0x46, 0xD6, 0x3E, 0x69, 0x64, 0x2A, 0xCE, 0xCB, 0x2F,
+	0xFC, 0x97, 0x05, 0x7A, 0xAC, 0x7F, 0xD5, 0x1A, 0x4B, 0x0E, 0xA7, 0x5A,
+	0x28, 0x14, 0x3F, 0x29, 0x88, 0x3C, 0x4C, 0x02, 0xB8, 0xDA, 0xB0, 0x17,
+	0x55, 0x1F, 0x8A, 0x7D, 0x57, 0xC7, 0x8D, 0x74, 0xB7, 0xC4, 0x9F, 0x72,
+	0x7E, 0x15, 0x22, 0x12, 0x58, 0x07, 0x99, 0x34, 0x6E, 0x50, 0xDE, 0x68,
+	0x65, 0xBC, 0xDB, 0xF8, 0xC8, 0xA8, 0x2B, 0x40, 0xDC, 0xFE, 0x32, 0xA4,
+	0xCA, 0x10, 0x21, 0xF0, 0xD3, 0x5D, 0x0F, 0x00, 0x6F, 0x9D, 0x36, 0x42,
+	0x4A, 0x5E, 0xC1, 0xE0
+};
+
+static const u8 q1[256] = {
+	0x75, 0xF3, 0xC6, 0xF4, 0xDB, 0x7B, 0xFB, 0xC8, 0x4A, 0xD3, 0xE6, 0x6B,
+	0x45, 0x7D, 0xE8, 0x4B, 0xD6, 0x32, 0xD8, 0xFD, 0x37, 0x71, 0xF1, 0xE1,
+	0x30, 0x0F, 0xF8, 0x1B, 0x87, 0xFA, 0x06, 0x3F, 0x5E, 0xBA, 0xAE, 0x5B,
+	0x8A, 0x00, 0xBC, 0x9D, 0x6D, 0xC1, 0xB1, 0x0E, 0x80, 0x5D, 0xD2, 0xD5,
+	0xA0, 0x84, 0x07, 0x14, 0xB5, 0x90, 0x2C, 0xA3, 0xB2, 0x73, 0x4C, 0x54,
+	0x92, 0x74, 0x36, 0x51, 0x38, 0xB0, 0xBD, 0x5A, 0xFC, 0x60, 0x62, 0x96,
+	0x6C, 0x42, 0xF7, 0x10, 0x7C, 0x28, 0x27, 0x8C, 0x13, 0x95, 0x9C, 0xC7,
+	0x24, 0x46, 0x3B, 0x70, 0xCA, 0xE3, 0x85, 0xCB, 0x11, 0xD0, 0x93, 0xB8,
+	0xA6, 0x83, 0x20, 0xFF, 0x9F, 0x77, 0xC3, 0xCC, 0x03, 0x6F, 0x08, 0xBF,
+	0x40, 0xE7, 0x2B, 0xE2, 0x79, 0x0C, 0xAA, 0x82, 0x41, 0x3A, 0xEA, 0xB9,
+	0xE4, 0x9A, 0xA4, 0x97, 0x7E, 0xDA, 0x7A, 0x17, 0x66, 0x94, 0xA1, 0x1D,
+	0x3D, 0xF0, 0xDE, 0xB3, 0x0B, 0x72, 0xA7, 0x1C, 0xEF, 0xD1, 0x53, 0x3E,
+	0x8F, 0x33, 0x26, 0x5F, 0xEC, 0x76, 0x2A, 0x49, 0x81, 0x88, 0xEE, 0x21,
+	0xC4, 0x1A, 0xEB, 0xD9, 0xC5, 0x39, 0x99, 0xCD, 0xAD, 0x31, 0x8B, 0x01,
+	0x18, 0x23, 0xDD, 0x1F, 0x4E, 0x2D, 0xF9, 0x48, 0x4F, 0xF2, 0x65, 0x8E,
+	0x78, 0x5C, 0x58, 0x19, 0x8D, 0xE5, 0x98, 0x57, 0x67, 0x7F, 0x05, 0x64,
+	0xAF, 0x63, 0xB6, 0xFE, 0xF5, 0xB7, 0x3C, 0xA5, 0xCE, 0xE9, 0x68, 0x44,
+	0xE0, 0x4D, 0x43, 0x69, 0x29, 0x2E, 0xAC, 0x15, 0x59, 0xA8, 0x0A, 0x9E,
+	0x6E, 0x47, 0xDF, 0x34, 0x35, 0x6A, 0xCF, 0xDC, 0x22, 0xC9, 0xC0, 0x9B,
+	0x89, 0xD4, 0xED, 0xAB, 0x12, 0xA2, 0x0D, 0x52, 0xBB, 0x02, 0x2F, 0xA9,
+	0xD7, 0x61, 0x1E, 0xB4, 0x50, 0x04, 0xF6, 0xC2, 0x16, 0x25, 0x86, 0x56,
+	0x55, 0x09, 0xBE, 0x91
+};
+
+/* These MDS tables are actually tables of MDS composed with q0 and q1,
+ * because it is only ever used that way and we can save some time by
+ * precomputing.  Of course the main saving comes from precomputing the
+ * GF(2^8) multiplication involved in the MDS matrix multiply; by looking
+ * things up in these tables we reduce the matrix multiply to four lookups
+ * and three XORs.  Semi-formally, the definition of these tables is:
+ * mds[0][i] = MDS (q1[i] 0 0 0)^T  mds[1][i] = MDS (0 q0[i] 0 0)^T
+ * mds[2][i] = MDS (0 0 q1[i] 0)^T  mds[3][i] = MDS (0 0 0 q0[i])^T
+ * where ^T means "transpose", the matrix multiply is performed in GF(2^8)
+ * represented as GF(2)[x]/v(x) where v(x)=x^8+x^6+x^5+x^3+1 as described
+ * by Schneier et al, and I'm casually glossing over the byte/word
+ * conversion issues. */
+
+static const u32 mds[4][256] = {
+	{
+	0xBCBC3275, 0xECEC21F3, 0x202043C6, 0xB3B3C9F4, 0xDADA03DB, 0x02028B7B,
+	0xE2E22BFB, 0x9E9EFAC8, 0xC9C9EC4A, 0xD4D409D3, 0x18186BE6, 0x1E1E9F6B,
+	0x98980E45, 0xB2B2387D, 0xA6A6D2E8, 0x2626B74B, 0x3C3C57D6, 0x93938A32,
+	0x8282EED8, 0x525298FD, 0x7B7BD437, 0xBBBB3771, 0x5B5B97F1, 0x474783E1,
+	0x24243C30, 0x5151E20F, 0xBABAC6F8, 0x4A4AF31B, 0xBFBF4887, 0x0D0D70FA,
+	0xB0B0B306, 0x7575DE3F, 0xD2D2FD5E, 0x7D7D20BA, 0x666631AE, 0x3A3AA35B,
+	0x59591C8A, 0x00000000, 0xCDCD93BC, 0x1A1AE09D, 0xAEAE2C6D, 0x7F7FABC1,
+	0x2B2BC7B1, 0xBEBEB90E, 0xE0E0A080, 0x8A8A105D, 0x3B3B52D2, 0x6464BAD5,
+	0xD8D888A0, 0xE7E7A584, 0x5F5FE807, 0x1B1B1114, 0x2C2CC2B5, 0xFCFCB490,
+	0x3131272C, 0x808065A3, 0x73732AB2, 0x0C0C8173, 0x79795F4C, 0x6B6B4154,
+	0x4B4B0292, 0x53536974, 0x94948F36, 0x83831F51, 0x2A2A3638, 0xC4C49CB0,
+	0x2222C8BD, 0xD5D5F85A, 0xBDBDC3FC, 0x48487860, 0xFFFFCE62, 0x4C4C0796,
+	0x4141776C, 0xC7C7E642, 0xEBEB24F7, 0x1C1C1410, 0x5D5D637C, 0x36362228,
+	0x6767C027, 0xE9E9AF8C, 0x4444F913, 0x1414EA95, 0xF5F5BB9C, 0xCFCF18C7,
+	0x3F3F2D24, 0xC0C0E346, 0x7272DB3B, 0x54546C70, 0x29294CCA, 0xF0F035E3,
+	0x0808FE85, 0xC6C617CB, 0xF3F34F11, 0x8C8CE4D0, 0xA4A45993, 0xCACA96B8,
+	0x68683BA6, 0xB8B84D83, 0x38382820, 0xE5E52EFF, 0xADAD569F, 0x0B0B8477,
+	0xC8C81DC3, 0x9999FFCC, 0x5858ED03, 0x19199A6F, 0x0E0E0A08, 0x95957EBF,
+	0x70705040, 0xF7F730E7, 0x6E6ECF2B, 0x1F1F6EE2, 0xB5B53D79, 0x09090F0C,
+	0x616134AA, 0x57571682, 0x9F9F0B41, 0x9D9D803A, 0x111164EA, 0x2525CDB9,
+	0xAFAFDDE4, 0x4545089A, 0xDFDF8DA4, 0xA3A35C97, 0xEAEAD57E, 0x353558DA,
+	0xEDEDD07A, 0x4343FC17, 0xF8F8CB66, 0xFBFBB194, 0x3737D3A1, 0xFAFA401D,
+	0xC2C2683D, 0xB4B4CCF0, 0x32325DDE, 0x9C9C71B3, 0x5656E70B, 0xE3E3DA72,
+	0x878760A7, 0x15151B1C, 0xF9F93AEF, 0x6363BFD1, 0x3434A953, 0x9A9A853E,
+	0xB1B1428F, 0x7C7CD133, 0x88889B26, 0x3D3DA65F, 0xA1A1D7EC, 0xE4E4DF76,
+	0x8181942A, 0x91910149, 0x0F0FFB81, 0xEEEEAA88, 0x161661EE, 0xD7D77321,
+	0x9797F5C4, 0xA5A5A81A, 0xFEFE3FEB, 0x6D6DB5D9, 0x7878AEC5, 0xC5C56D39,
+	0x1D1DE599, 0x7676A4CD, 0x3E3EDCAD, 0xCBCB6731, 0xB6B6478B, 0xEFEF5B01,
+	0x12121E18, 0x6060C523, 0x6A6AB0DD, 0x4D4DF61F, 0xCECEE94E, 0xDEDE7C2D,
+	0x55559DF9, 0x7E7E5A48, 0x2121B24F, 0x03037AF2, 0xA0A02665, 0x5E5E198E,
+	0x5A5A6678, 0x65654B5C, 0x62624E58, 0xFDFD4519, 0x0606F48D, 0x404086E5,
+	0xF2F2BE98, 0x3333AC57, 0x17179067, 0x05058E7F, 0xE8E85E05, 0x4F4F7D64,
+	0x89896AAF, 0x10109563, 0x74742FB6, 0x0A0A75FE, 0x5C5C92F5, 0x9B9B74B7,
+	0x2D2D333C, 0x3030D6A5, 0x2E2E49CE, 0x494989E9, 0x46467268, 0x77775544,
+	0xA8A8D8E0, 0x9696044D, 0x2828BD43, 0xA9A92969, 0xD9D97929, 0x8686912E,
+	0xD1D187AC, 0xF4F44A15, 0x8D8D1559, 0xD6D682A8, 0xB9B9BC0A, 0x42420D9E,
+	0xF6F6C16E, 0x2F2FB847, 0xDDDD06DF, 0x23233934, 0xCCCC6235, 0xF1F1C46A,
+	0xC1C112CF, 0x8585EBDC, 0x8F8F9E22, 0x7171A1C9, 0x9090F0C0, 0xAAAA539B,
+	0x0101F189, 0x8B8BE1D4, 0x4E4E8CED, 0x8E8E6FAB, 0xABABA212, 0x6F6F3EA2,
+	0xE6E6540D, 0xDBDBF252, 0x92927BBB, 0xB7B7B602, 0x6969CA2F, 0x3939D9A9,
+	0xD3D30CD7, 0xA7A72361, 0xA2A2AD1E, 0xC3C399B4, 0x6C6C4450, 0x07070504,
+	0x04047FF6, 0x272746C2, 0xACACA716, 0xD0D07625, 0x50501386, 0xDCDCF756,
+	0x84841A55, 0xE1E15109, 0x7A7A25BE, 0x1313EF91},
+
+	{
+	0xA9D93939, 0x67901717, 0xB3719C9C, 0xE8D2A6A6, 0x04050707, 0xFD985252,
+	0xA3658080, 0x76DFE4E4, 0x9A084545, 0x92024B4B, 0x80A0E0E0, 0x78665A5A,
+	0xE4DDAFAF, 0xDDB06A6A, 0xD1BF6363, 0x38362A2A, 0x0D54E6E6, 0xC6432020,
+	0x3562CCCC, 0x98BEF2F2, 0x181E1212, 0xF724EBEB, 0xECD7A1A1, 0x6C774141,
+	0x43BD2828, 0x7532BCBC, 0x37D47B7B, 0x269B8888, 0xFA700D0D, 0x13F94444,
+	0x94B1FBFB, 0x485A7E7E, 0xF27A0303, 0xD0E48C8C, 0x8B47B6B6, 0x303C2424,
+	0x84A5E7E7, 0x54416B6B, 0xDF06DDDD, 0x23C56060, 0x1945FDFD, 0x5BA33A3A,
+	0x3D68C2C2, 0x59158D8D, 0xF321ECEC, 0xAE316666, 0xA23E6F6F, 0x82165757,
+	0x63951010, 0x015BEFEF, 0x834DB8B8, 0x2E918686, 0xD9B56D6D, 0x511F8383,
+	0x9B53AAAA, 0x7C635D5D, 0xA63B6868, 0xEB3FFEFE, 0xA5D63030, 0xBE257A7A,
+	0x16A7ACAC, 0x0C0F0909, 0xE335F0F0, 0x6123A7A7, 0xC0F09090, 0x8CAFE9E9,
+	0x3A809D9D, 0xF5925C5C, 0x73810C0C, 0x2C273131, 0x2576D0D0, 0x0BE75656,
+	0xBB7B9292, 0x4EE9CECE, 0x89F10101, 0x6B9F1E1E, 0x53A93434, 0x6AC4F1F1,
+	0xB499C3C3, 0xF1975B5B, 0xE1834747, 0xE66B1818, 0xBDC82222, 0x450E9898,
+	0xE26E1F1F, 0xF4C9B3B3, 0xB62F7474, 0x66CBF8F8, 0xCCFF9999, 0x95EA1414,
+	0x03ED5858, 0x56F7DCDC, 0xD4E18B8B, 0x1C1B1515, 0x1EADA2A2, 0xD70CD3D3,
+	0xFB2BE2E2, 0xC31DC8C8, 0x8E195E5E, 0xB5C22C2C, 0xE9894949, 0xCF12C1C1,
+	0xBF7E9595, 0xBA207D7D, 0xEA641111, 0x77840B0B, 0x396DC5C5, 0xAF6A8989,
+	0x33D17C7C, 0xC9A17171, 0x62CEFFFF, 0x7137BBBB, 0x81FB0F0F, 0x793DB5B5,
+	0x0951E1E1, 0xADDC3E3E, 0x242D3F3F, 0xCDA47676, 0xF99D5555, 0xD8EE8282,
+	0xE5864040, 0xC5AE7878, 0xB9CD2525, 0x4D049696, 0x44557777, 0x080A0E0E,
+	0x86135050, 0xE730F7F7, 0xA1D33737, 0x1D40FAFA, 0xAA346161, 0xED8C4E4E,
+	0x06B3B0B0, 0x706C5454, 0xB22A7373, 0xD2523B3B, 0x410B9F9F, 0x7B8B0202,
+	0xA088D8D8, 0x114FF3F3, 0x3167CBCB, 0xC2462727, 0x27C06767, 0x90B4FCFC,
+	0x20283838, 0xF67F0404, 0x60784848, 0xFF2EE5E5, 0x96074C4C, 0x5C4B6565,
+	0xB1C72B2B, 0xAB6F8E8E, 0x9E0D4242, 0x9CBBF5F5, 0x52F2DBDB, 0x1BF34A4A,
+	0x5FA63D3D, 0x9359A4A4, 0x0ABCB9B9, 0xEF3AF9F9, 0x91EF1313, 0x85FE0808,
+	0x49019191, 0xEE611616, 0x2D7CDEDE, 0x4FB22121, 0x8F42B1B1, 0x3BDB7272,
+	0x47B82F2F, 0x8748BFBF, 0x6D2CAEAE, 0x46E3C0C0, 0xD6573C3C, 0x3E859A9A,
+	0x6929A9A9, 0x647D4F4F, 0x2A948181, 0xCE492E2E, 0xCB17C6C6, 0x2FCA6969,
+	0xFCC3BDBD, 0x975CA3A3, 0x055EE8E8, 0x7AD0EDED, 0xAC87D1D1, 0x7F8E0505,
+	0xD5BA6464, 0x1AA8A5A5, 0x4BB72626, 0x0EB9BEBE, 0xA7608787, 0x5AF8D5D5,
+	0x28223636, 0x14111B1B, 0x3FDE7575, 0x2979D9D9, 0x88AAEEEE, 0x3C332D2D,
+	0x4C5F7979, 0x02B6B7B7, 0xB896CACA, 0xDA583535, 0xB09CC4C4, 0x17FC4343,
+	0x551A8484, 0x1FF64D4D, 0x8A1C5959, 0x7D38B2B2, 0x57AC3333, 0xC718CFCF,
+	0x8DF40606, 0x74695353, 0xB7749B9B, 0xC4F59797, 0x9F56ADAD, 0x72DAE3E3,
+	0x7ED5EAEA, 0x154AF4F4, 0x229E8F8F, 0x12A2ABAB, 0x584E6262, 0x07E85F5F,
+	0x99E51D1D, 0x34392323, 0x6EC1F6F6, 0x50446C6C, 0xDE5D3232, 0x68724646,
+	0x6526A0A0, 0xBC93CDCD, 0xDB03DADA, 0xF8C6BABA, 0xC8FA9E9E, 0xA882D6D6,
+	0x2BCF6E6E, 0x40507070, 0xDCEB8585, 0xFE750A0A, 0x328A9393, 0xA48DDFDF,
+	0xCA4C2929, 0x10141C1C, 0x2173D7D7, 0xF0CCB4B4, 0xD309D4D4, 0x5D108A8A,
+	0x0FE25151, 0x00000000, 0x6F9A1919, 0x9DE01A1A, 0x368F9494, 0x42E6C7C7,
+	0x4AECC9C9, 0x5EFDD2D2, 0xC1AB7F7F, 0xE0D8A8A8},
+
+	{
+	0xBC75BC32, 0xECF3EC21, 0x20C62043, 0xB3F4B3C9, 0xDADBDA03, 0x027B028B,
+	0xE2FBE22B, 0x9EC89EFA, 0xC94AC9EC, 0xD4D3D409, 0x18E6186B, 0x1E6B1E9F,
+	0x9845980E, 0xB27DB238, 0xA6E8A6D2, 0x264B26B7, 0x3CD63C57, 0x9332938A,
+	0x82D882EE, 0x52FD5298, 0x7B377BD4, 0xBB71BB37, 0x5BF15B97, 0x47E14783,
+	0x2430243C, 0x510F51E2, 0xBAF8BAC6, 0x4A1B4AF3, 0xBF87BF48, 0x0DFA0D70,
+	0xB006B0B3, 0x753F75DE, 0xD25ED2FD, 0x7DBA7D20, 0x66AE6631, 0x3A5B3AA3,
+	0x598A591C, 0x00000000, 0xCDBCCD93, 0x1A9D1AE0, 0xAE6DAE2C, 0x7FC17FAB,
+	0x2BB12BC7, 0xBE0EBEB9, 0xE080E0A0, 0x8A5D8A10, 0x3BD23B52, 0x64D564BA,
+	0xD8A0D888, 0xE784E7A5, 0x5F075FE8, 0x1B141B11, 0x2CB52CC2, 0xFC90FCB4,
+	0x312C3127, 0x80A38065, 0x73B2732A, 0x0C730C81, 0x794C795F, 0x6B546B41,
+	0x4B924B02, 0x53745369, 0x9436948F, 0x8351831F, 0x2A382A36, 0xC4B0C49C,
+	0x22BD22C8, 0xD55AD5F8, 0xBDFCBDC3, 0x48604878, 0xFF62FFCE, 0x4C964C07,
+	0x416C4177, 0xC742C7E6, 0xEBF7EB24, 0x1C101C14, 0x5D7C5D63, 0x36283622,
+	0x672767C0, 0xE98CE9AF, 0x441344F9, 0x149514EA, 0xF59CF5BB, 0xCFC7CF18,
+	0x3F243F2D, 0xC046C0E3, 0x723B72DB, 0x5470546C, 0x29CA294C, 0xF0E3F035,
+	0x088508FE, 0xC6CBC617, 0xF311F34F, 0x8CD08CE4, 0xA493A459, 0xCAB8CA96,
+	0x68A6683B, 0xB883B84D, 0x38203828, 0xE5FFE52E, 0xAD9FAD56, 0x0B770B84,
+	0xC8C3C81D, 0x99CC99FF, 0x580358ED, 0x196F199A, 0x0E080E0A, 0x95BF957E,
+	0x70407050, 0xF7E7F730, 0x6E2B6ECF, 0x1FE21F6E, 0xB579B53D, 0x090C090F,
+	0x61AA6134, 0x57825716, 0x9F419F0B, 0x9D3A9D80, 0x11EA1164, 0x25B925CD,
+	0xAFE4AFDD, 0x459A4508, 0xDFA4DF8D, 0xA397A35C, 0xEA7EEAD5, 0x35DA3558,
+	0xED7AEDD0, 0x431743FC, 0xF866F8CB, 0xFB94FBB1, 0x37A137D3, 0xFA1DFA40,
+	0xC23DC268, 0xB4F0B4CC, 0x32DE325D, 0x9CB39C71, 0x560B56E7, 0xE372E3DA,
+	0x87A78760, 0x151C151B, 0xF9EFF93A, 0x63D163BF, 0x345334A9, 0x9A3E9A85,
+	0xB18FB142, 0x7C337CD1, 0x8826889B, 0x3D5F3DA6, 0xA1ECA1D7, 0xE476E4DF,
+	0x812A8194, 0x91499101, 0x0F810FFB, 0xEE88EEAA, 0x16EE1661, 0xD721D773,
+	0x97C497F5, 0xA51AA5A8, 0xFEEBFE3F, 0x6DD96DB5, 0x78C578AE, 0xC539C56D,
+	0x1D991DE5, 0x76CD76A4, 0x3EAD3EDC, 0xCB31CB67, 0xB68BB647, 0xEF01EF5B,
+	0x1218121E, 0x602360C5, 0x6ADD6AB0, 0x4D1F4DF6, 0xCE4ECEE9, 0xDE2DDE7C,
+	0x55F9559D, 0x7E487E5A, 0x214F21B2, 0x03F2037A, 0xA065A026, 0x5E8E5E19,
+	0x5A785A66, 0x655C654B, 0x6258624E, 0xFD19FD45, 0x068D06F4, 0x40E54086,
+	0xF298F2BE, 0x335733AC, 0x17671790, 0x057F058E, 0xE805E85E, 0x4F644F7D,
+	0x89AF896A, 0x10631095, 0x74B6742F, 0x0AFE0A75, 0x5CF55C92, 0x9BB79B74,
+	0x2D3C2D33, 0x30A530D6, 0x2ECE2E49, 0x49E94989, 0x46684672, 0x77447755,
+	0xA8E0A8D8, 0x964D9604, 0x284328BD, 0xA969A929, 0xD929D979, 0x862E8691,
+	0xD1ACD187, 0xF415F44A, 0x8D598D15, 0xD6A8D682, 0xB90AB9BC, 0x429E420D,
+	0xF66EF6C1, 0x2F472FB8, 0xDDDFDD06, 0x23342339, 0xCC35CC62, 0xF16AF1C4,
+	0xC1CFC112, 0x85DC85EB, 0x8F228F9E, 0x71C971A1, 0x90C090F0, 0xAA9BAA53,
+	0x018901F1, 0x8BD48BE1, 0x4EED4E8C, 0x8EAB8E6F, 0xAB12ABA2, 0x6FA26F3E,
+	0xE60DE654, 0xDB52DBF2, 0x92BB927B, 0xB702B7B6, 0x692F69CA, 0x39A939D9,
+	0xD3D7D30C, 0xA761A723, 0xA21EA2AD, 0xC3B4C399, 0x6C506C44, 0x07040705,
+	0x04F6047F, 0x27C22746, 0xAC16ACA7, 0xD025D076, 0x50865013, 0xDC56DCF7,
+	0x8455841A, 0xE109E151, 0x7ABE7A25, 0x139113EF},
+
+	{
+	0xD939A9D9, 0x90176790, 0x719CB371, 0xD2A6E8D2, 0x05070405, 0x9852FD98,
+	0x6580A365, 0xDFE476DF, 0x08459A08, 0x024B9202, 0xA0E080A0, 0x665A7866,
+	0xDDAFE4DD, 0xB06ADDB0, 0xBF63D1BF, 0x362A3836, 0x54E60D54, 0x4320C643,
+	0x62CC3562, 0xBEF298BE, 0x1E12181E, 0x24EBF724, 0xD7A1ECD7, 0x77416C77,
+	0xBD2843BD, 0x32BC7532, 0xD47B37D4, 0x9B88269B, 0x700DFA70, 0xF94413F9,
+	0xB1FB94B1, 0x5A7E485A, 0x7A03F27A, 0xE48CD0E4, 0x47B68B47, 0x3C24303C,
+	0xA5E784A5, 0x416B5441, 0x06DDDF06, 0xC56023C5, 0x45FD1945, 0xA33A5BA3,
+	0x68C23D68, 0x158D5915, 0x21ECF321, 0x3166AE31, 0x3E6FA23E, 0x16578216,
+	0x95106395, 0x5BEF015B, 0x4DB8834D, 0x91862E91, 0xB56DD9B5, 0x1F83511F,
+	0x53AA9B53, 0x635D7C63, 0x3B68A63B, 0x3FFEEB3F, 0xD630A5D6, 0x257ABE25,
+	0xA7AC16A7, 0x0F090C0F, 0x35F0E335, 0x23A76123, 0xF090C0F0, 0xAFE98CAF,
+	0x809D3A80, 0x925CF592, 0x810C7381, 0x27312C27, 0x76D02576, 0xE7560BE7,
+	0x7B92BB7B, 0xE9CE4EE9, 0xF10189F1, 0x9F1E6B9F, 0xA93453A9, 0xC4F16AC4,
+	0x99C3B499, 0x975BF197, 0x8347E183, 0x6B18E66B, 0xC822BDC8, 0x0E98450E,
+	0x6E1FE26E, 0xC9B3F4C9, 0x2F74B62F, 0xCBF866CB, 0xFF99CCFF, 0xEA1495EA,
+	0xED5803ED, 0xF7DC56F7, 0xE18BD4E1, 0x1B151C1B, 0xADA21EAD, 0x0CD3D70C,
+	0x2BE2FB2B, 0x1DC8C31D, 0x195E8E19, 0xC22CB5C2, 0x8949E989, 0x12C1CF12,
+	0x7E95BF7E, 0x207DBA20, 0x6411EA64, 0x840B7784, 0x6DC5396D, 0x6A89AF6A,
+	0xD17C33D1, 0xA171C9A1, 0xCEFF62CE, 0x37BB7137, 0xFB0F81FB, 0x3DB5793D,
+	0x51E10951, 0xDC3EADDC, 0x2D3F242D, 0xA476CDA4, 0x9D55F99D, 0xEE82D8EE,
+	0x8640E586, 0xAE78C5AE, 0xCD25B9CD, 0x04964D04, 0x55774455, 0x0A0E080A,
+	0x13508613, 0x30F7E730, 0xD337A1D3, 0x40FA1D40, 0x3461AA34, 0x8C4EED8C,
+	0xB3B006B3, 0x6C54706C, 0x2A73B22A, 0x523BD252, 0x0B9F410B, 0x8B027B8B,
+	0x88D8A088, 0x4FF3114F, 0x67CB3167, 0x4627C246, 0xC06727C0, 0xB4FC90B4,
+	0x28382028, 0x7F04F67F, 0x78486078, 0x2EE5FF2E, 0x074C9607, 0x4B655C4B,
+	0xC72BB1C7, 0x6F8EAB6F, 0x0D429E0D, 0xBBF59CBB, 0xF2DB52F2, 0xF34A1BF3,
+	0xA63D5FA6, 0x59A49359, 0xBCB90ABC, 0x3AF9EF3A, 0xEF1391EF, 0xFE0885FE,
+	0x01914901, 0x6116EE61, 0x7CDE2D7C, 0xB2214FB2, 0x42B18F42, 0xDB723BDB,
+	0xB82F47B8, 0x48BF8748, 0x2CAE6D2C, 0xE3C046E3, 0x573CD657, 0x859A3E85,
+	0x29A96929, 0x7D4F647D, 0x94812A94, 0x492ECE49, 0x17C6CB17, 0xCA692FCA,
+	0xC3BDFCC3, 0x5CA3975C, 0x5EE8055E, 0xD0ED7AD0, 0x87D1AC87, 0x8E057F8E,
+	0xBA64D5BA, 0xA8A51AA8, 0xB7264BB7, 0xB9BE0EB9, 0x6087A760, 0xF8D55AF8,
+	0x22362822, 0x111B1411, 0xDE753FDE, 0x79D92979, 0xAAEE88AA, 0x332D3C33,
+	0x5F794C5F, 0xB6B702B6, 0x96CAB896, 0x5835DA58, 0x9CC4B09C, 0xFC4317FC,
+	0x1A84551A, 0xF64D1FF6, 0x1C598A1C, 0x38B27D38, 0xAC3357AC, 0x18CFC718,
+	0xF4068DF4, 0x69537469, 0x749BB774, 0xF597C4F5, 0x56AD9F56, 0xDAE372DA,
+	0xD5EA7ED5, 0x4AF4154A, 0x9E8F229E, 0xA2AB12A2, 0x4E62584E, 0xE85F07E8,
+	0xE51D99E5, 0x39233439, 0xC1F66EC1, 0x446C5044, 0x5D32DE5D, 0x72466872,
+	0x26A06526, 0x93CDBC93, 0x03DADB03, 0xC6BAF8C6, 0xFA9EC8FA, 0x82D6A882,
+	0xCF6E2BCF, 0x50704050, 0xEB85DCEB, 0x750AFE75, 0x8A93328A, 0x8DDFA48D,
+	0x4C29CA4C, 0x141C1014, 0x73D72173, 0xCCB4F0CC, 0x09D4D309, 0x108A5D10,
+	0xE2510FE2, 0x00000000, 0x9A196F9A, 0xE01A9DE0, 0x8F94368F, 0xE6C742E6,
+	0xECC94AEC, 0xFDD25EFD, 0xAB7FC1AB, 0xD8A8E0D8}
+};
+
+/* The exp_to_poly and poly_to_exp tables are used to perform efficient
+ * operations in GF(2^8) represented as GF(2)[x]/w(x) where
+ * w(x)=x^8+x^6+x^3+x^2+1.  We care about doing that because it's part of the
+ * definition of the RS matrix in the key schedule.  Elements of that field
+ * are polynomials of degree not greater than 7 and all coefficients 0 or 1,
+ * which can be represented naturally by bytes (just substitute x=2).  In that
+ * form, GF(2^8) addition is the same as bitwise XOR, but GF(2^8)
+ * multiplication is inefficient without hardware support.  To multiply
+ * faster, I make use of the fact x is a generator for the nonzero elements,
+ * so that every element p of GF(2)[x]/w(x) is either 0 or equal to (x)^n for
+ * some n in 0..254.  Note that that caret is exponentiation in GF(2^8),
+ * *not* polynomial notation.  So if I want to compute pq where p and q are
+ * in GF(2^8), I can just say:
+ *    1. if p=0 or q=0 then pq=0
+ *    2. otherwise, find m and n such that p=x^m and q=x^n
+ *    3. pq=(x^m)(x^n)=x^(m+n), so add m and n and find pq
+ * The translations in steps 2 and 3 are looked up in the tables
+ * poly_to_exp (for step 2) and exp_to_poly (for step 3).  To see this
+ * in action, look at the CALC_S macro.  As additional wrinkles, note that
+ * one of my operands is always a constant, so the poly_to_exp lookup on it
+ * is done in advance; I included the original values in the comments so
+ * readers can have some chance of recognizing that this *is* the RS matrix
+ * from the Twofish paper.  I've only included the table entries I actually
+ * need; I never do a lookup on a variable input of zero and the biggest
+ * exponents I'll ever see are 254 (variable) and 237 (constant), so they'll
+ * never sum to more than 491.	I'm repeating part of the exp_to_poly table
+ * so that I don't have to do mod-255 reduction in the exponent arithmetic.
+ * Since I know my constant operands are never zero, I only have to worry
+ * about zero values in the variable operand, and I do it with a simple
+ * conditional branch.	I know conditionals are expensive, but I couldn't
+ * see a non-horrible way of avoiding them, and I did manage to group the
+ * statements so that each if covers four group multiplications. */
+
+static const u8 poly_to_exp[255] = {
+	0x00, 0x01, 0x17, 0x02, 0x2E, 0x18, 0x53, 0x03, 0x6A, 0x2F, 0x93, 0x19,
+	0x34, 0x54, 0x45, 0x04, 0x5C, 0x6B, 0xB6, 0x30, 0xA6, 0x94, 0x4B, 0x1A,
+	0x8C, 0x35, 0x81, 0x55, 0xAA, 0x46, 0x0D, 0x05, 0x24, 0x5D, 0x87, 0x6C,
+	0x9B, 0xB7, 0xC1, 0x31, 0x2B, 0xA7, 0xA3, 0x95, 0x98, 0x4C, 0xCA, 0x1B,
+	0xE6, 0x8D, 0x73, 0x36, 0xCD, 0x82, 0x12, 0x56, 0x62, 0xAB, 0xF0, 0x47,
+	0x4F, 0x0E, 0xBD, 0x06, 0xD4, 0x25, 0xD2, 0x5E, 0x27, 0x88, 0x66, 0x6D,
+	0xD6, 0x9C, 0x79, 0xB8, 0x08, 0xC2, 0xDF, 0x32, 0x68, 0x2C, 0xFD, 0xA8,
+	0x8A, 0xA4, 0x5A, 0x96, 0x29, 0x99, 0x22, 0x4D, 0x60, 0xCB, 0xE4, 0x1C,
+	0x7B, 0xE7, 0x3B, 0x8E, 0x9E, 0x74, 0xF4, 0x37, 0xD8, 0xCE, 0xF9, 0x83,
+	0x6F, 0x13, 0xB2, 0x57, 0xE1, 0x63, 0xDC, 0xAC, 0xC4, 0xF1, 0xAF, 0x48,
+	0x0A, 0x50, 0x42, 0x0F, 0xBA, 0xBE, 0xC7, 0x07, 0xDE, 0xD5, 0x78, 0x26,
+	0x65, 0xD3, 0xD1, 0x5F, 0xE3, 0x28, 0x21, 0x89, 0x59, 0x67, 0xFC, 0x6E,
+	0xB1, 0xD7, 0xF8, 0x9D, 0xF3, 0x7A, 0x3A, 0xB9, 0xC6, 0x09, 0x41, 0xC3,
+	0xAE, 0xE0, 0xDB, 0x33, 0x44, 0x69, 0x92, 0x2D, 0x52, 0xFE, 0x16, 0xA9,
+	0x0C, 0x8B, 0x80, 0xA5, 0x4A, 0x5B, 0xB5, 0x97, 0xC9, 0x2A, 0xA2, 0x9A,
+	0xC0, 0x23, 0x86, 0x4E, 0xBC, 0x61, 0xEF, 0xCC, 0x11, 0xE5, 0x72, 0x1D,
+	0x3D, 0x7C, 0xEB, 0xE8, 0xE9, 0x3C, 0xEA, 0x8F, 0x7D, 0x9F, 0xEC, 0x75,
+	0x1E, 0xF5, 0x3E, 0x38, 0xF6, 0xD9, 0x3F, 0xCF, 0x76, 0xFA, 0x1F, 0x84,
+	0xA0, 0x70, 0xED, 0x14, 0x90, 0xB3, 0x7E, 0x58, 0xFB, 0xE2, 0x20, 0x64,
+	0xD0, 0xDD, 0x77, 0xAD, 0xDA, 0xC5, 0x40, 0xF2, 0x39, 0xB0, 0xF7, 0x49,
+	0xB4, 0x0B, 0x7F, 0x51, 0x15, 0x43, 0x91, 0x10, 0x71, 0xBB, 0xEE, 0xBF,
+	0x85, 0xC8, 0xA1
+};
+
+static const u8 exp_to_poly[492] = {
+	0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x4D, 0x9A, 0x79, 0xF2,
+	0xA9, 0x1F, 0x3E, 0x7C, 0xF8, 0xBD, 0x37, 0x6E, 0xDC, 0xF5, 0xA7, 0x03,
+	0x06, 0x0C, 0x18, 0x30, 0x60, 0xC0, 0xCD, 0xD7, 0xE3, 0x8B, 0x5B, 0xB6,
+	0x21, 0x42, 0x84, 0x45, 0x8A, 0x59, 0xB2, 0x29, 0x52, 0xA4, 0x05, 0x0A,
+	0x14, 0x28, 0x50, 0xA0, 0x0D, 0x1A, 0x34, 0x68, 0xD0, 0xED, 0x97, 0x63,
+	0xC6, 0xC1, 0xCF, 0xD3, 0xEB, 0x9B, 0x7B, 0xF6, 0xA1, 0x0F, 0x1E, 0x3C,
+	0x78, 0xF0, 0xAD, 0x17, 0x2E, 0x5C, 0xB8, 0x3D, 0x7A, 0xF4, 0xA5, 0x07,
+	0x0E, 0x1C, 0x38, 0x70, 0xE0, 0x8D, 0x57, 0xAE, 0x11, 0x22, 0x44, 0x88,
+	0x5D, 0xBA, 0x39, 0x72, 0xE4, 0x85, 0x47, 0x8E, 0x51, 0xA2, 0x09, 0x12,
+	0x24, 0x48, 0x90, 0x6D, 0xDA, 0xF9, 0xBF, 0x33, 0x66, 0xCC, 0xD5, 0xE7,
+	0x83, 0x4B, 0x96, 0x61, 0xC2, 0xC9, 0xDF, 0xF3, 0xAB, 0x1B, 0x36, 0x6C,
+	0xD8, 0xFD, 0xB7, 0x23, 0x46, 0x8C, 0x55, 0xAA, 0x19, 0x32, 0x64, 0xC8,
+	0xDD, 0xF7, 0xA3, 0x0B, 0x16, 0x2C, 0x58, 0xB0, 0x2D, 0x5A, 0xB4, 0x25,
+	0x4A, 0x94, 0x65, 0xCA, 0xD9, 0xFF, 0xB3, 0x2B, 0x56, 0xAC, 0x15, 0x2A,
+	0x54, 0xA8, 0x1D, 0x3A, 0x74, 0xE8, 0x9D, 0x77, 0xEE, 0x91, 0x6F, 0xDE,
+	0xF1, 0xAF, 0x13, 0x26, 0x4C, 0x98, 0x7D, 0xFA, 0xB9, 0x3F, 0x7E, 0xFC,
+	0xB5, 0x27, 0x4E, 0x9C, 0x75, 0xEA, 0x99, 0x7F, 0xFE, 0xB1, 0x2F, 0x5E,
+	0xBC, 0x35, 0x6A, 0xD4, 0xE5, 0x87, 0x43, 0x86, 0x41, 0x82, 0x49, 0x92,
+	0x69, 0xD2, 0xE9, 0x9F, 0x73, 0xE6, 0x81, 0x4F, 0x9E, 0x71, 0xE2, 0x89,
+	0x5F, 0xBE, 0x31, 0x62, 0xC4, 0xC5, 0xC7, 0xC3, 0xCB, 0xDB, 0xFB, 0xBB,
+	0x3B, 0x76, 0xEC, 0x95, 0x67, 0xCE, 0xD1, 0xEF, 0x93, 0x6B, 0xD6, 0xE1,
+	0x8F, 0x53, 0xA6, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x4D,
+	0x9A, 0x79, 0xF2, 0xA9, 0x1F, 0x3E, 0x7C, 0xF8, 0xBD, 0x37, 0x6E, 0xDC,
+	0xF5, 0xA7, 0x03, 0x06, 0x0C, 0x18, 0x30, 0x60, 0xC0, 0xCD, 0xD7, 0xE3,
+	0x8B, 0x5B, 0xB6, 0x21, 0x42, 0x84, 0x45, 0x8A, 0x59, 0xB2, 0x29, 0x52,
+	0xA4, 0x05, 0x0A, 0x14, 0x28, 0x50, 0xA0, 0x0D, 0x1A, 0x34, 0x68, 0xD0,
+	0xED, 0x97, 0x63, 0xC6, 0xC1, 0xCF, 0xD3, 0xEB, 0x9B, 0x7B, 0xF6, 0xA1,
+	0x0F, 0x1E, 0x3C, 0x78, 0xF0, 0xAD, 0x17, 0x2E, 0x5C, 0xB8, 0x3D, 0x7A,
+	0xF4, 0xA5, 0x07, 0x0E, 0x1C, 0x38, 0x70, 0xE0, 0x8D, 0x57, 0xAE, 0x11,
+	0x22, 0x44, 0x88, 0x5D, 0xBA, 0x39, 0x72, 0xE4, 0x85, 0x47, 0x8E, 0x51,
+	0xA2, 0x09, 0x12, 0x24, 0x48, 0x90, 0x6D, 0xDA, 0xF9, 0xBF, 0x33, 0x66,
+	0xCC, 0xD5, 0xE7, 0x83, 0x4B, 0x96, 0x61, 0xC2, 0xC9, 0xDF, 0xF3, 0xAB,
+	0x1B, 0x36, 0x6C, 0xD8, 0xFD, 0xB7, 0x23, 0x46, 0x8C, 0x55, 0xAA, 0x19,
+	0x32, 0x64, 0xC8, 0xDD, 0xF7, 0xA3, 0x0B, 0x16, 0x2C, 0x58, 0xB0, 0x2D,
+	0x5A, 0xB4, 0x25, 0x4A, 0x94, 0x65, 0xCA, 0xD9, 0xFF, 0xB3, 0x2B, 0x56,
+	0xAC, 0x15, 0x2A, 0x54, 0xA8, 0x1D, 0x3A, 0x74, 0xE8, 0x9D, 0x77, 0xEE,
+	0x91, 0x6F, 0xDE, 0xF1, 0xAF, 0x13, 0x26, 0x4C, 0x98, 0x7D, 0xFA, 0xB9,
+	0x3F, 0x7E, 0xFC, 0xB5, 0x27, 0x4E, 0x9C, 0x75, 0xEA, 0x99, 0x7F, 0xFE,
+	0xB1, 0x2F, 0x5E, 0xBC, 0x35, 0x6A, 0xD4, 0xE5, 0x87, 0x43, 0x86, 0x41,
+	0x82, 0x49, 0x92, 0x69, 0xD2, 0xE9, 0x9F, 0x73, 0xE6, 0x81, 0x4F, 0x9E,
+	0x71, 0xE2, 0x89, 0x5F, 0xBE, 0x31, 0x62, 0xC4, 0xC5, 0xC7, 0xC3, 0xCB
+};
+
+
+/* The table constants are indices of
+ * S-box entries, preprocessed through q0 and q1. */
+static const u8 calc_sb_tbl[512] = {
+	0xA9, 0x75, 0x67, 0xF3, 0xB3, 0xC6, 0xE8, 0xF4,
+	0x04, 0xDB, 0xFD, 0x7B, 0xA3, 0xFB, 0x76, 0xC8,
+	0x9A, 0x4A, 0x92, 0xD3, 0x80, 0xE6, 0x78, 0x6B,
+	0xE4, 0x45, 0xDD, 0x7D, 0xD1, 0xE8, 0x38, 0x4B,
+	0x0D, 0xD6, 0xC6, 0x32, 0x35, 0xD8, 0x98, 0xFD,
+	0x18, 0x37, 0xF7, 0x71, 0xEC, 0xF1, 0x6C, 0xE1,
+	0x43, 0x30, 0x75, 0x0F, 0x37, 0xF8, 0x26, 0x1B,
+	0xFA, 0x87, 0x13, 0xFA, 0x94, 0x06, 0x48, 0x3F,
+	0xF2, 0x5E, 0xD0, 0xBA, 0x8B, 0xAE, 0x30, 0x5B,
+	0x84, 0x8A, 0x54, 0x00, 0xDF, 0xBC, 0x23, 0x9D,
+	0x19, 0x6D, 0x5B, 0xC1, 0x3D, 0xB1, 0x59, 0x0E,
+	0xF3, 0x80, 0xAE, 0x5D, 0xA2, 0xD2, 0x82, 0xD5,
+	0x63, 0xA0, 0x01, 0x84, 0x83, 0x07, 0x2E, 0x14,
+	0xD9, 0xB5, 0x51, 0x90, 0x9B, 0x2C, 0x7C, 0xA3,
+	0xA6, 0xB2, 0xEB, 0x73, 0xA5, 0x4C, 0xBE, 0x54,
+	0x16, 0x92, 0x0C, 0x74, 0xE3, 0x36, 0x61, 0x51,
+	0xC0, 0x38, 0x8C, 0xB0, 0x3A, 0xBD, 0xF5, 0x5A,
+	0x73, 0xFC, 0x2C, 0x60, 0x25, 0x62, 0x0B, 0x96,
+	0xBB, 0x6C, 0x4E, 0x42, 0x89, 0xF7, 0x6B, 0x10,
+	0x53, 0x7C, 0x6A, 0x28, 0xB4, 0x27, 0xF1, 0x8C,
+	0xE1, 0x13, 0xE6, 0x95, 0xBD, 0x9C, 0x45, 0xC7,
+	0xE2, 0x24, 0xF4, 0x46, 0xB6, 0x3B, 0x66, 0x70,
+	0xCC, 0xCA, 0x95, 0xE3, 0x03, 0x85, 0x56, 0xCB,
+	0xD4, 0x11, 0x1C, 0xD0, 0x1E, 0x93, 0xD7, 0xB8,
+	0xFB, 0xA6, 0xC3, 0x83, 0x8E, 0x20, 0xB5, 0xFF,
+	0xE9, 0x9F, 0xCF, 0x77, 0xBF, 0xC3, 0xBA, 0xCC,
+	0xEA, 0x03, 0x77, 0x6F, 0x39, 0x08, 0xAF, 0xBF,
+	0x33, 0x40, 0xC9, 0xE7, 0x62, 0x2B, 0x71, 0xE2,
+	0x81, 0x79, 0x79, 0x0C, 0x09, 0xAA, 0xAD, 0x82,
+	0x24, 0x41, 0xCD, 0x3A, 0xF9, 0xEA, 0xD8, 0xB9,
+	0xE5, 0xE4, 0xC5, 0x9A, 0xB9, 0xA4, 0x4D, 0x97,
+	0x44, 0x7E, 0x08, 0xDA, 0x86, 0x7A, 0xE7, 0x17,
+	0xA1, 0x66, 0x1D, 0x94, 0xAA, 0xA1, 0xED, 0x1D,
+	0x06, 0x3D, 0x70, 0xF0, 0xB2, 0xDE, 0xD2, 0xB3,
+	0x41, 0x0B, 0x7B, 0x72, 0xA0, 0xA7, 0x11, 0x1C,
+	0x31, 0xEF, 0xC2, 0xD1, 0x27, 0x53, 0x90, 0x3E,
+	0x20, 0x8F, 0xF6, 0x33, 0x60, 0x26, 0xFF, 0x5F,
+	0x96, 0xEC, 0x5C, 0x76, 0xB1, 0x2A, 0xAB, 0x49,
+	0x9E, 0x81, 0x9C, 0x88, 0x52, 0xEE, 0x1B, 0x21,
+	0x5F, 0xC4, 0x93, 0x1A, 0x0A, 0xEB, 0xEF, 0xD9,
+	0x91, 0xC5, 0x85, 0x39, 0x49, 0x99, 0xEE, 0xCD,
+	0x2D, 0xAD, 0x4F, 0x31, 0x8F, 0x8B, 0x3B, 0x01,
+	0x47, 0x18, 0x87, 0x23, 0x6D, 0xDD, 0x46, 0x1F,
+	0xD6, 0x4E, 0x3E, 0x2D, 0x69, 0xF9, 0x64, 0x48,
+	0x2A, 0x4F, 0xCE, 0xF2, 0xCB, 0x65, 0x2F, 0x8E,
+	0xFC, 0x78, 0x97, 0x5C, 0x05, 0x58, 0x7A, 0x19,
+	0xAC, 0x8D, 0x7F, 0xE5, 0xD5, 0x98, 0x1A, 0x57,
+	0x4B, 0x67, 0x0E, 0x7F, 0xA7, 0x05, 0x5A, 0x64,
+	0x28, 0xAF, 0x14, 0x63, 0x3F, 0xB6, 0x29, 0xFE,
+	0x88, 0xF5, 0x3C, 0xB7, 0x4C, 0x3C, 0x02, 0xA5,
+	0xB8, 0xCE, 0xDA, 0xE9, 0xB0, 0x68, 0x17, 0x44,
+	0x55, 0xE0, 0x1F, 0x4D, 0x8A, 0x43, 0x7D, 0x69,
+	0x57, 0x29, 0xC7, 0x2E, 0x8D, 0xAC, 0x74, 0x15,
+	0xB7, 0x59, 0xC4, 0xA8, 0x9F, 0x0A, 0x72, 0x9E,
+	0x7E, 0x6E, 0x15, 0x47, 0x22, 0xDF, 0x12, 0x34,
+	0x58, 0x35, 0x07, 0x6A, 0x99, 0xCF, 0x34, 0xDC,
+	0x6E, 0x22, 0x50, 0xC9, 0xDE, 0xC0, 0x68, 0x9B,
+	0x65, 0x89, 0xBC, 0xD4, 0xDB, 0xED, 0xF8, 0xAB,
+	0xC8, 0x12, 0xA8, 0xA2, 0x2B, 0x0D, 0x40, 0x52,
+	0xDC, 0xBB, 0xFE, 0x02, 0x32, 0x2F, 0xA4, 0xA9,
+	0xCA, 0xD7, 0x10, 0x61, 0x21, 0x1E, 0xF0, 0xB4,
+	0xD3, 0x50, 0x5D, 0x04, 0x0F, 0xF6, 0x00, 0xC2,
+	0x6F, 0x16, 0x9D, 0x25, 0x36, 0x86, 0x42, 0x56,
+	0x4A, 0x55, 0x5E, 0x09, 0xC1, 0xBE, 0xE0, 0x91
+};
+
+/* Macro to perform one column of the RS matrix multiplication.  The
+ * parameters a, b, c, and d are the four bytes of output; i is the index
+ * of the key bytes, and w, x, y, and z, are the column of constants from
+ * the RS matrix, preprocessed through the poly_to_exp table. */
+
+#define CALC_S(a, b, c, d, i, w, x, y, z) \
+   if (key[i]) { \
+      tmp = poly_to_exp[key[i] - 1]; \
+      (a) ^= exp_to_poly[tmp + (w)]; \
+      (b) ^= exp_to_poly[tmp + (x)]; \
+      (c) ^= exp_to_poly[tmp + (y)]; \
+      (d) ^= exp_to_poly[tmp + (z)]; \
+   }
+
+/* Macros to calculate the key-dependent S-boxes for a 128-bit key using
+ * the S vector from CALC_S.  CALC_SB_2 computes a single entry in all
+ * four S-boxes, where i is the index of the entry to compute, and a and b
+ * are the index numbers preprocessed through the q0 and q1 tables
+ * respectively. */
+
+#define CALC_SB_2(i, a, b) \
+   ctx->s[0][i] = mds[0][q0[(a) ^ sa] ^ se]; \
+   ctx->s[1][i] = mds[1][q0[(b) ^ sb] ^ sf]; \
+   ctx->s[2][i] = mds[2][q1[(a) ^ sc] ^ sg]; \
+   ctx->s[3][i] = mds[3][q1[(b) ^ sd] ^ sh]
+
+/* Macro exactly like CALC_SB_2, but for 192-bit keys. */
+
+#define CALC_SB192_2(i, a, b) \
+   ctx->s[0][i] = mds[0][q0[q0[(b) ^ sa] ^ se] ^ si]; \
+   ctx->s[1][i] = mds[1][q0[q1[(b) ^ sb] ^ sf] ^ sj]; \
+   ctx->s[2][i] = mds[2][q1[q0[(a) ^ sc] ^ sg] ^ sk]; \
+   ctx->s[3][i] = mds[3][q1[q1[(a) ^ sd] ^ sh] ^ sl];
+
+/* Macro exactly like CALC_SB_2, but for 256-bit keys. */
+
+#define CALC_SB256_2(i, a, b) \
+   ctx->s[0][i] = mds[0][q0[q0[q1[(b) ^ sa] ^ se] ^ si] ^ sm]; \
+   ctx->s[1][i] = mds[1][q0[q1[q1[(a) ^ sb] ^ sf] ^ sj] ^ sn]; \
+   ctx->s[2][i] = mds[2][q1[q0[q0[(a) ^ sc] ^ sg] ^ sk] ^ so]; \
+   ctx->s[3][i] = mds[3][q1[q1[q0[(b) ^ sd] ^ sh] ^ sl] ^ sp];
+
+/* Macros to calculate the whitening and round subkeys.  CALC_K_2 computes the
+ * last two stages of the h() function for a given index (either 2i or 2i+1).
+ * a, b, c, and d are the four bytes going into the last two stages.  For
+ * 128-bit keys, this is the entire h() function and a and c are the index
+ * preprocessed through q0 and q1 respectively; for longer keys they are the
+ * output of previous stages.  j is the index of the first key byte to use.
+ * CALC_K computes a pair of subkeys for 128-bit Twofish, by calling CALC_K_2
+ * twice, doing the Pseudo-Hadamard Transform, and doing the necessary
+ * rotations.  Its parameters are: a, the array to write the results into,
+ * j, the index of the first output entry, k and l, the preprocessed indices
+ * for index 2i, and m and n, the preprocessed indices for index 2i+1.
+ * CALC_K192_2 expands CALC_K_2 to handle 192-bit keys, by doing an
+ * additional lookup-and-XOR stage.  The parameters a, b, c and d are the
+ * four bytes going into the last three stages.  For 192-bit keys, c = d
+ * are the index preprocessed through q0, and a = b are the index
+ * preprocessed through q1; j is the index of the first key byte to use.
+ * CALC_K192 is identical to CALC_K but for using the CALC_K192_2 macro
+ * instead of CALC_K_2.
+ * CALC_K256_2 expands CALC_K192_2 to handle 256-bit keys, by doing an
+ * additional lookup-and-XOR stage.  The parameters a and b are the index
+ * preprocessed through q0 and q1 respectively; j is the index of the first
+ * key byte to use.  CALC_K256 is identical to CALC_K but for using the
+ * CALC_K256_2 macro instead of CALC_K_2. */
+
+#define CALC_K_2(a, b, c, d, j) \
+     mds[0][q0[a ^ key[(j) + 8]] ^ key[j]] \
+   ^ mds[1][q0[b ^ key[(j) + 9]] ^ key[(j) + 1]] \
+   ^ mds[2][q1[c ^ key[(j) + 10]] ^ key[(j) + 2]] \
+   ^ mds[3][q1[d ^ key[(j) + 11]] ^ key[(j) + 3]]
+
+#define CALC_K(a, j, k, l, m, n) \
+   x = CALC_K_2 (k, l, k, l, 0); \
+   y = CALC_K_2 (m, n, m, n, 4); \
+   y = rol32(y, 8); \
+   x += y; y += x; ctx->a[j] = x; \
+   ctx->a[(j) + 1] = rol32(y, 9)
+
+#define CALC_K192_2(a, b, c, d, j) \
+   CALC_K_2 (q0[a ^ key[(j) + 16]], \
+	     q1[b ^ key[(j) + 17]], \
+	     q0[c ^ key[(j) + 18]], \
+	     q1[d ^ key[(j) + 19]], j)
+
+#define CALC_K192(a, j, k, l, m, n) \
+   x = CALC_K192_2 (l, l, k, k, 0); \
+   y = CALC_K192_2 (n, n, m, m, 4); \
+   y = rol32(y, 8); \
+   x += y; y += x; ctx->a[j] = x; \
+   ctx->a[(j) + 1] = rol32(y, 9)
+
+#define CALC_K256_2(a, b, j) \
+   CALC_K192_2 (q1[b ^ key[(j) + 24]], \
+	        q1[a ^ key[(j) + 25]], \
+	        q0[a ^ key[(j) + 26]], \
+	        q0[b ^ key[(j) + 27]], j)
+
+#define CALC_K256(a, j, k, l, m, n) \
+   x = CALC_K256_2 (k, l, 0); \
+   y = CALC_K256_2 (m, n, 4); \
+   y = rol32(y, 8); \
+   x += y; y += x; ctx->a[j] = x; \
+   ctx->a[(j) + 1] = rol32(y, 9)
+
+/* Perform the key setup. */
+int twofish_setkey(struct crypto_tfm *tfm, const u8 *key,
+		   unsigned int key_len, u32 *flags)
+{
+
+	struct twofish_ctx *ctx = crypto_tfm_ctx(tfm);
+
+	int i, j, k;
+
+	/* Temporaries for CALC_K. */
+	u32 x, y;
+
+	/* The S vector used to key the S-boxes, split up into individual bytes.
+	 * 128-bit keys use only sa through sh; 256-bit use all of them. */
+	u8 sa = 0, sb = 0, sc = 0, sd = 0, se = 0, sf = 0, sg = 0, sh = 0;
+	u8 si = 0, sj = 0, sk = 0, sl = 0, sm = 0, sn = 0, so = 0, sp = 0;
+
+	/* Temporary for CALC_S. */
+	u8 tmp;
+
+	/* Check key length. */
+	if (key_len != 16 && key_len != 24 && key_len != 32)
+	{
+		*flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
+		return -EINVAL; /* unsupported key length */
+	}
+
+	/* Compute the first two words of the S vector.  The magic numbers are
+	 * the entries of the RS matrix, preprocessed through poly_to_exp. The
+	 * numbers in the comments are the original (polynomial form) matrix
+	 * entries. */
+	CALC_S (sa, sb, sc, sd, 0, 0x00, 0x2D, 0x01, 0x2D); /* 01 A4 02 A4 */
+	CALC_S (sa, sb, sc, sd, 1, 0x2D, 0xA4, 0x44, 0x8A); /* A4 56 A1 55 */
+	CALC_S (sa, sb, sc, sd, 2, 0x8A, 0xD5, 0xBF, 0xD1); /* 55 82 FC 87 */
+	CALC_S (sa, sb, sc, sd, 3, 0xD1, 0x7F, 0x3D, 0x99); /* 87 F3 C1 5A */
+	CALC_S (sa, sb, sc, sd, 4, 0x99, 0x46, 0x66, 0x96); /* 5A 1E 47 58 */
+	CALC_S (sa, sb, sc, sd, 5, 0x96, 0x3C, 0x5B, 0xED); /* 58 C6 AE DB */
+	CALC_S (sa, sb, sc, sd, 6, 0xED, 0x37, 0x4F, 0xE0); /* DB 68 3D 9E */
+	CALC_S (sa, sb, sc, sd, 7, 0xE0, 0xD0, 0x8C, 0x17); /* 9E E5 19 03 */
+	CALC_S (se, sf, sg, sh, 8, 0x00, 0x2D, 0x01, 0x2D); /* 01 A4 02 A4 */
+	CALC_S (se, sf, sg, sh, 9, 0x2D, 0xA4, 0x44, 0x8A); /* A4 56 A1 55 */
+	CALC_S (se, sf, sg, sh, 10, 0x8A, 0xD5, 0xBF, 0xD1); /* 55 82 FC 87 */
+	CALC_S (se, sf, sg, sh, 11, 0xD1, 0x7F, 0x3D, 0x99); /* 87 F3 C1 5A */
+	CALC_S (se, sf, sg, sh, 12, 0x99, 0x46, 0x66, 0x96); /* 5A 1E 47 58 */
+	CALC_S (se, sf, sg, sh, 13, 0x96, 0x3C, 0x5B, 0xED); /* 58 C6 AE DB */
+	CALC_S (se, sf, sg, sh, 14, 0xED, 0x37, 0x4F, 0xE0); /* DB 68 3D 9E */
+	CALC_S (se, sf, sg, sh, 15, 0xE0, 0xD0, 0x8C, 0x17); /* 9E E5 19 03 */
+
+	if (key_len == 24 || key_len == 32) { /* 192- or 256-bit key */
+		/* Calculate the third word of the S vector */
+		CALC_S (si, sj, sk, sl, 16, 0x00, 0x2D, 0x01, 0x2D); /* 01 A4 02 A4 */
+		CALC_S (si, sj, sk, sl, 17, 0x2D, 0xA4, 0x44, 0x8A); /* A4 56 A1 55 */
+		CALC_S (si, sj, sk, sl, 18, 0x8A, 0xD5, 0xBF, 0xD1); /* 55 82 FC 87 */
+		CALC_S (si, sj, sk, sl, 19, 0xD1, 0x7F, 0x3D, 0x99); /* 87 F3 C1 5A */
+		CALC_S (si, sj, sk, sl, 20, 0x99, 0x46, 0x66, 0x96); /* 5A 1E 47 58 */
+		CALC_S (si, sj, sk, sl, 21, 0x96, 0x3C, 0x5B, 0xED); /* 58 C6 AE DB */
+		CALC_S (si, sj, sk, sl, 22, 0xED, 0x37, 0x4F, 0xE0); /* DB 68 3D 9E */
+		CALC_S (si, sj, sk, sl, 23, 0xE0, 0xD0, 0x8C, 0x17); /* 9E E5 19 03 */
+	}
+
+	if (key_len == 32) { /* 256-bit key */
+		/* Calculate the fourth word of the S vector */
+		CALC_S (sm, sn, so, sp, 24, 0x00, 0x2D, 0x01, 0x2D); /* 01 A4 02 A4 */
+		CALC_S (sm, sn, so, sp, 25, 0x2D, 0xA4, 0x44, 0x8A); /* A4 56 A1 55 */
+		CALC_S (sm, sn, so, sp, 26, 0x8A, 0xD5, 0xBF, 0xD1); /* 55 82 FC 87 */
+		CALC_S (sm, sn, so, sp, 27, 0xD1, 0x7F, 0x3D, 0x99); /* 87 F3 C1 5A */
+		CALC_S (sm, sn, so, sp, 28, 0x99, 0x46, 0x66, 0x96); /* 5A 1E 47 58 */
+		CALC_S (sm, sn, so, sp, 29, 0x96, 0x3C, 0x5B, 0xED); /* 58 C6 AE DB */
+		CALC_S (sm, sn, so, sp, 30, 0xED, 0x37, 0x4F, 0xE0); /* DB 68 3D 9E */
+		CALC_S (sm, sn, so, sp, 31, 0xE0, 0xD0, 0x8C, 0x17); /* 9E E5 19 03 */
+
+		/* Compute the S-boxes. */
+		for ( i = j = 0, k = 1; i < 256; i++, j += 2, k += 2 ) {
+			CALC_SB256_2( i, calc_sb_tbl[j], calc_sb_tbl[k] );
+		}
+
+		/* Calculate whitening and round subkeys.  The constants are
+		 * indices of subkeys, preprocessed through q0 and q1. */
+		CALC_K256 (w, 0, 0xA9, 0x75, 0x67, 0xF3);
+		CALC_K256 (w, 2, 0xB3, 0xC6, 0xE8, 0xF4);
+		CALC_K256 (w, 4, 0x04, 0xDB, 0xFD, 0x7B);
+		CALC_K256 (w, 6, 0xA3, 0xFB, 0x76, 0xC8);
+		CALC_K256 (k, 0, 0x9A, 0x4A, 0x92, 0xD3);
+		CALC_K256 (k, 2, 0x80, 0xE6, 0x78, 0x6B);
+		CALC_K256 (k, 4, 0xE4, 0x45, 0xDD, 0x7D);
+		CALC_K256 (k, 6, 0xD1, 0xE8, 0x38, 0x4B);
+		CALC_K256 (k, 8, 0x0D, 0xD6, 0xC6, 0x32);
+		CALC_K256 (k, 10, 0x35, 0xD8, 0x98, 0xFD);
+		CALC_K256 (k, 12, 0x18, 0x37, 0xF7, 0x71);
+		CALC_K256 (k, 14, 0xEC, 0xF1, 0x6C, 0xE1);
+		CALC_K256 (k, 16, 0x43, 0x30, 0x75, 0x0F);
+		CALC_K256 (k, 18, 0x37, 0xF8, 0x26, 0x1B);
+		CALC_K256 (k, 20, 0xFA, 0x87, 0x13, 0xFA);
+		CALC_K256 (k, 22, 0x94, 0x06, 0x48, 0x3F);
+		CALC_K256 (k, 24, 0xF2, 0x5E, 0xD0, 0xBA);
+		CALC_K256 (k, 26, 0x8B, 0xAE, 0x30, 0x5B);
+		CALC_K256 (k, 28, 0x84, 0x8A, 0x54, 0x00);
+		CALC_K256 (k, 30, 0xDF, 0xBC, 0x23, 0x9D);
+	} else if (key_len == 24) { /* 192-bit key */
+		/* Compute the S-boxes. */
+		for ( i = j = 0, k = 1; i < 256; i++, j += 2, k += 2 ) {
+		        CALC_SB192_2( i, calc_sb_tbl[j], calc_sb_tbl[k] );
+		}
+
+		/* Calculate whitening and round subkeys.  The constants are
+		 * indices of subkeys, preprocessed through q0 and q1. */
+		CALC_K192 (w, 0, 0xA9, 0x75, 0x67, 0xF3);
+		CALC_K192 (w, 2, 0xB3, 0xC6, 0xE8, 0xF4);
+		CALC_K192 (w, 4, 0x04, 0xDB, 0xFD, 0x7B);
+		CALC_K192 (w, 6, 0xA3, 0xFB, 0x76, 0xC8);
+		CALC_K192 (k, 0, 0x9A, 0x4A, 0x92, 0xD3);
+		CALC_K192 (k, 2, 0x80, 0xE6, 0x78, 0x6B);
+		CALC_K192 (k, 4, 0xE4, 0x45, 0xDD, 0x7D);
+		CALC_K192 (k, 6, 0xD1, 0xE8, 0x38, 0x4B);
+		CALC_K192 (k, 8, 0x0D, 0xD6, 0xC6, 0x32);
+		CALC_K192 (k, 10, 0x35, 0xD8, 0x98, 0xFD);
+		CALC_K192 (k, 12, 0x18, 0x37, 0xF7, 0x71);
+		CALC_K192 (k, 14, 0xEC, 0xF1, 0x6C, 0xE1);
+		CALC_K192 (k, 16, 0x43, 0x30, 0x75, 0x0F);
+		CALC_K192 (k, 18, 0x37, 0xF8, 0x26, 0x1B);
+		CALC_K192 (k, 20, 0xFA, 0x87, 0x13, 0xFA);
+		CALC_K192 (k, 22, 0x94, 0x06, 0x48, 0x3F);
+		CALC_K192 (k, 24, 0xF2, 0x5E, 0xD0, 0xBA);
+		CALC_K192 (k, 26, 0x8B, 0xAE, 0x30, 0x5B);
+		CALC_K192 (k, 28, 0x84, 0x8A, 0x54, 0x00);
+		CALC_K192 (k, 30, 0xDF, 0xBC, 0x23, 0x9D);
+	} else { /* 128-bit key */
+		/* Compute the S-boxes. */
+		for ( i = j = 0, k = 1; i < 256; i++, j += 2, k += 2 ) {
+			CALC_SB_2( i, calc_sb_tbl[j], calc_sb_tbl[k] );
+		}
+
+		/* Calculate whitening and round subkeys.  The constants are
+		 * indices of subkeys, preprocessed through q0 and q1. */
+		CALC_K (w, 0, 0xA9, 0x75, 0x67, 0xF3);
+		CALC_K (w, 2, 0xB3, 0xC6, 0xE8, 0xF4);
+		CALC_K (w, 4, 0x04, 0xDB, 0xFD, 0x7B);
+		CALC_K (w, 6, 0xA3, 0xFB, 0x76, 0xC8);
+		CALC_K (k, 0, 0x9A, 0x4A, 0x92, 0xD3);
+		CALC_K (k, 2, 0x80, 0xE6, 0x78, 0x6B);
+		CALC_K (k, 4, 0xE4, 0x45, 0xDD, 0x7D);
+		CALC_K (k, 6, 0xD1, 0xE8, 0x38, 0x4B);
+		CALC_K (k, 8, 0x0D, 0xD6, 0xC6, 0x32);
+		CALC_K (k, 10, 0x35, 0xD8, 0x98, 0xFD);
+		CALC_K (k, 12, 0x18, 0x37, 0xF7, 0x71);
+		CALC_K (k, 14, 0xEC, 0xF1, 0x6C, 0xE1);
+		CALC_K (k, 16, 0x43, 0x30, 0x75, 0x0F);
+		CALC_K (k, 18, 0x37, 0xF8, 0x26, 0x1B);
+		CALC_K (k, 20, 0xFA, 0x87, 0x13, 0xFA);
+		CALC_K (k, 22, 0x94, 0x06, 0x48, 0x3F);
+		CALC_K (k, 24, 0xF2, 0x5E, 0xD0, 0xBA);
+		CALC_K (k, 26, 0x8B, 0xAE, 0x30, 0x5B);
+		CALC_K (k, 28, 0x84, 0x8A, 0x54, 0x00);
+		CALC_K (k, 30, 0xDF, 0xBC, 0x23, 0x9D);
+	}
+
+	return 0;
+}
+
+EXPORT_SYMBOL_GPL(twofish_setkey);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Twofish cipher common functions");
diff --git a/include/crypto/twofish.h b/include/crypto/twofish.h
new file mode 100644
index 000000000000..e4328cfaaf64
--- /dev/null
+++ b/include/crypto/twofish.h
@@ -0,0 +1,23 @@
+#ifndef _CRYPTO_TWOFISH_H
+#define _CRYPTO_TWOFISH_H
+
+#include <linux/types.h>
+
+#define TF_MIN_KEY_SIZE 16
+#define TF_MAX_KEY_SIZE 32
+#define TF_BLOCK_SIZE 16
+
+struct crypto_tfm;
+
+/* Structure for an expanded Twofish key.  s contains the key-dependent
+ * S-boxes composed with the MDS matrix; w contains the eight "whitening"
+ * subkeys, K[0] through K[7].	k holds the remaining, "round" subkeys.  Note
+ * that k[i] corresponds to what the Twofish paper calls K[i+8]. */
+struct twofish_ctx {
+	u32 s[4][256], w[8], k[32];
+};
+
+int twofish_setkey(struct crypto_tfm *tfm, const u8 *key,
+		   unsigned int key_len, u32 *flags);
+
+#endif
-- 
cgit v1.2.3


From 6521f30273fbec65146a0f16de74b7b402b0f7b0 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sun, 6 Aug 2006 20:28:44 +1000
Subject: [CRYPTO] api: Add crypto_alg reference counting

Up until now we've relied on module reference counting to ensure that the
crypto_alg structures don't disappear from under us.  This was good enough
as long as each crypto_alg came from exactly one module.

However, with parameterised crypto algorithms a crypto_alg object may need
two or more modules to operate.  This means that we need to count the
references to the crypto_alg object directly.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 crypto/api.c           | 32 ++++++++++++++++++++++++++------
 crypto/proc.c          |  3 +++
 include/linux/crypto.h |  3 +++
 3 files changed, 32 insertions(+), 6 deletions(-)

(limited to 'include')

diff --git a/crypto/api.c b/crypto/api.c
index 8c2743a05f90..5994a58ef954 100644
--- a/crypto/api.c
+++ b/crypto/api.c
@@ -29,13 +29,26 @@
 LIST_HEAD(crypto_alg_list);
 DECLARE_RWSEM(crypto_alg_sem);
 
-static inline int crypto_mod_get(struct crypto_alg *alg)
+static inline struct crypto_alg *crypto_alg_get(struct crypto_alg *alg)
 {
-	return try_module_get(alg->cra_module);
+	atomic_inc(&alg->cra_refcnt);
+	return alg;
+}
+
+static inline void crypto_alg_put(struct crypto_alg *alg)
+{
+	if (atomic_dec_and_test(&alg->cra_refcnt) && alg->cra_destroy)
+		alg->cra_destroy(alg);
+}
+
+static struct crypto_alg *crypto_mod_get(struct crypto_alg *alg)
+{
+	return try_module_get(alg->cra_module) ? crypto_alg_get(alg) : NULL;
 }
 
-static inline void crypto_mod_put(struct crypto_alg *alg)
+static void crypto_mod_put(struct crypto_alg *alg)
 {
+	crypto_alg_put(alg);
 	module_put(alg->cra_module);
 }
 
@@ -274,6 +287,7 @@ int crypto_register_alg(struct crypto_alg *alg)
 	}
 	
 	list_add(&alg->cra_list, &crypto_alg_list);
+	atomic_set(&alg->cra_refcnt, 1);
 out:	
 	up_write(&crypto_alg_sem);
 	return ret;
@@ -284,8 +298,6 @@ int crypto_unregister_alg(struct crypto_alg *alg)
 	int ret = -ENOENT;
 	struct crypto_alg *q;
 	
-	BUG_ON(!alg->cra_module);
-	
 	down_write(&crypto_alg_sem);
 	list_for_each_entry(q, &crypto_alg_list, cra_list) {
 		if (alg == q) {
@@ -296,7 +308,15 @@ int crypto_unregister_alg(struct crypto_alg *alg)
 	}
 out:	
 	up_write(&crypto_alg_sem);
-	return ret;
+
+	if (ret)
+		return ret;
+
+	BUG_ON(atomic_read(&alg->cra_refcnt) != 1);
+	if (alg->cra_destroy)
+		alg->cra_destroy(alg);
+
+	return 0;
 }
 
 int crypto_alg_available(const char *name, u32 flags)
diff --git a/crypto/proc.c b/crypto/proc.c
index c0a5dd7ce2cc..8543b7a157d6 100644
--- a/crypto/proc.c
+++ b/crypto/proc.c
@@ -12,6 +12,8 @@
  * any later version.
  *
  */
+
+#include <asm/atomic.h>
 #include <linux/init.h>
 #include <linux/crypto.h>
 #include <linux/rwsem.h>
@@ -54,6 +56,7 @@ static int c_show(struct seq_file *m, void *p)
 	seq_printf(m, "driver       : %s\n", alg->cra_driver_name);
 	seq_printf(m, "module       : %s\n", module_name(alg->cra_module));
 	seq_printf(m, "priority     : %d\n", alg->cra_priority);
+	seq_printf(m, "refcnt       : %d\n", atomic_read(&alg->cra_refcnt));
 	
 	switch (alg->cra_flags & CRYPTO_ALG_TYPE_MASK) {
 	case CRYPTO_ALG_TYPE_CIPHER:
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index cb1e6631b132..7f57ff8ec975 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -17,6 +17,7 @@
 #ifndef _LINUX_CRYPTO_H
 #define _LINUX_CRYPTO_H
 
+#include <asm/atomic.h>
 #include <linux/module.h>
 #include <linux/kernel.h>
 #include <linux/types.h>
@@ -148,6 +149,7 @@ struct crypto_alg {
 	unsigned int cra_alignmask;
 
 	int cra_priority;
+	atomic_t cra_refcnt;
 
 	char cra_name[CRYPTO_MAX_ALG_NAME];
 	char cra_driver_name[CRYPTO_MAX_ALG_NAME];
@@ -160,6 +162,7 @@ struct crypto_alg {
 
 	int (*cra_init)(struct crypto_tfm *tfm);
 	void (*cra_exit)(struct crypto_tfm *tfm);
+	void (*cra_destroy)(struct crypto_alg *alg);
 	
 	struct module *cra_module;
 };
-- 
cgit v1.2.3


From 9409f38a0c8773c04bff8dda8c552d7ea013d956 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sun, 6 Aug 2006 19:49:12 +1000
Subject: [IPSEC]: Move linux/crypto.h inclusion out of net/xfrm.h

The header file linux/crypto.h is only needed by a few files so including
it in net/xfrm.h (which is included by half of the networking stack) is a
waste.  This patch moves it out of net/xfrm.h and into the specific header
files that actually need it.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 include/net/ah.h     | 1 +
 include/net/esp.h    | 1 +
 include/net/ipcomp.h | 4 ++++
 include/net/xfrm.h   | 2 +-
 net/xfrm/xfrm_user.c | 1 +
 5 files changed, 8 insertions(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/net/ah.h b/include/net/ah.h
index ceff00afae09..8e27c9ba8b84 100644
--- a/include/net/ah.h
+++ b/include/net/ah.h
@@ -1,6 +1,7 @@
 #ifndef _NET_AH_H
 #define _NET_AH_H
 
+#include <linux/crypto.h>
 #include <net/xfrm.h>
 
 /* This is the maximum truncated ICV length that we know of. */
diff --git a/include/net/esp.h b/include/net/esp.h
index 90cd94fad7d9..6eb837973c84 100644
--- a/include/net/esp.h
+++ b/include/net/esp.h
@@ -1,6 +1,7 @@
 #ifndef _NET_ESP_H
 #define _NET_ESP_H
 
+#include <linux/crypto.h>
 #include <net/xfrm.h>
 #include <asm/scatterlist.h>
 
diff --git a/include/net/ipcomp.h b/include/net/ipcomp.h
index e651a57ecdd5..b94e3047b4d9 100644
--- a/include/net/ipcomp.h
+++ b/include/net/ipcomp.h
@@ -1,8 +1,12 @@
 #ifndef _NET_IPCOMP_H
 #define _NET_IPCOMP_H
 
+#include <linux/types.h>
+
 #define IPCOMP_SCRATCH_SIZE     65400
 
+struct crypto_tfm;
+
 struct ipcomp_data {
 	u16 threshold;
 	struct crypto_tfm **tfms;
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 9c5ee9f20b65..10396b4bde14 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -8,7 +8,6 @@
 #include <linux/list.h>
 #include <linux/skbuff.h>
 #include <linux/socket.h>
-#include <linux/crypto.h>
 #include <linux/pfkeyv2.h>
 #include <linux/in6.h>
 #include <linux/mutex.h>
@@ -985,6 +984,7 @@ extern struct xfrm_algo_desc *xfrm_ealg_get_byname(char *name, int probe);
 extern struct xfrm_algo_desc *xfrm_calg_get_byname(char *name, int probe);
 
 struct crypto_tfm;
+struct scatterlist;
 typedef void (icv_update_fn_t)(struct crypto_tfm *, struct scatterlist *, unsigned int);
 
 extern void skb_icv_walk(const struct sk_buff *skb, struct crypto_tfm *tfm,
diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c
index 3e6a722d072e..7d18ca03c80d 100644
--- a/net/xfrm/xfrm_user.c
+++ b/net/xfrm/xfrm_user.c
@@ -10,6 +10,7 @@
  *
  */
 
+#include <linux/crypto.h>
 #include <linux/module.h>
 #include <linux/kernel.h>
 #include <linux/types.h>
-- 
cgit v1.2.3


From cce9e06d100df19a327b19f23adad76e7bf63edd Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Mon, 21 Aug 2006 21:08:13 +1000
Subject: [CRYPTO] api: Split out low-level API

The crypto API is made up of the part facing users such as IPsec and the
low-level part which is used by cryptographic entities such as algorithms.
This patch splits out the latter so that the two APIs are more clearly
delineated.  As a bonus the low-level API can now be modularised if all
algorithms are built as modules.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 crypto/Kconfig          |  84 ++++++++++++++++++++--------------
 crypto/Makefile         |   7 +--
 crypto/algapi.c         | 118 ++++++++++++++++++++++++++++++++++++++++++++++++
 crypto/api.c            |  97 +--------------------------------------
 crypto/internal.h       |   6 ++-
 crypto/proc.c           |   5 ++
 drivers/crypto/Kconfig  |   3 +-
 include/crypto/algapi.h |  18 ++++++++
 8 files changed, 204 insertions(+), 134 deletions(-)
 create mode 100644 crypto/algapi.c
 create mode 100644 include/crypto/algapi.h

(limited to 'include')

diff --git a/crypto/Kconfig b/crypto/Kconfig
index fa927a287a1d..aabc63195222 100644
--- a/crypto/Kconfig
+++ b/crypto/Kconfig
@@ -9,47 +9,54 @@ config CRYPTO
 	help
 	  This option provides the core Cryptographic API.
 
+if CRYPTO
+
+config CRYPTO_ALGAPI
+	tristate
+	help
+	  This option provides the API for cryptographic algorithms.
+
 config CRYPTO_HMAC
 	bool "HMAC support"
-	depends on CRYPTO
 	help
 	  HMAC: Keyed-Hashing for Message Authentication (RFC2104).
 	  This is required for IPSec.
 
 config CRYPTO_NULL
 	tristate "Null algorithms"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  These are 'Null' algorithms, used by IPsec, which do nothing.
 
 config CRYPTO_MD4
 	tristate "MD4 digest algorithm"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  MD4 message digest algorithm (RFC1320).
 
 config CRYPTO_MD5
 	tristate "MD5 digest algorithm"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  MD5 message digest algorithm (RFC1321).
 
 config CRYPTO_SHA1
 	tristate "SHA1 digest algorithm"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  SHA-1 secure hash standard (FIPS 180-1/DFIPS 180-2).
 
 config CRYPTO_SHA1_S390
 	tristate "SHA1 digest algorithm (s390)"
-	depends on CRYPTO && S390
+	depends on S390
+	select CRYPTO_ALGAPI
 	help
 	  This is the s390 hardware accelerated implementation of the
 	  SHA-1 secure hash standard (FIPS 180-1/DFIPS 180-2).
 
 config CRYPTO_SHA256
 	tristate "SHA256 digest algorithm"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  SHA256 secure hash standard (DFIPS 180-2).
 	  
@@ -58,7 +65,8 @@ config CRYPTO_SHA256
 
 config CRYPTO_SHA256_S390
 	tristate "SHA256 digest algorithm (s390)"
-	depends on CRYPTO && S390
+	depends on S390
+	select CRYPTO_ALGAPI
 	help
 	  This is the s390 hardware accelerated implementation of the
 	  SHA256 secure hash standard (DFIPS 180-2).
@@ -68,7 +76,7 @@ config CRYPTO_SHA256_S390
 
 config CRYPTO_SHA512
 	tristate "SHA384 and SHA512 digest algorithms"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  SHA512 secure hash standard (DFIPS 180-2).
 	  
@@ -80,7 +88,7 @@ config CRYPTO_SHA512
 
 config CRYPTO_WP512
 	tristate "Whirlpool digest algorithms"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  Whirlpool hash algorithm 512, 384 and 256-bit hashes
 
@@ -92,7 +100,7 @@ config CRYPTO_WP512
 
 config CRYPTO_TGR192
 	tristate "Tiger digest algorithms"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  Tiger hash algorithm 192, 160 and 128-bit hashes
 
@@ -105,19 +113,20 @@ config CRYPTO_TGR192
 
 config CRYPTO_DES
 	tristate "DES and Triple DES EDE cipher algorithms"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  DES cipher algorithm (FIPS 46-2), and Triple DES EDE (FIPS 46-3).
 
 config CRYPTO_DES_S390
 	tristate "DES and Triple DES cipher algorithms (s390)"
-	depends on CRYPTO && S390
+	depends on S390
+	select CRYPTO_ALGAPI
 	help
 	  DES cipher algorithm (FIPS 46-2), and Triple DES EDE (FIPS 46-3).
 
 config CRYPTO_BLOWFISH
 	tristate "Blowfish cipher algorithm"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  Blowfish cipher algorithm, by Bruce Schneier.
 	  
@@ -130,7 +139,7 @@ config CRYPTO_BLOWFISH
 
 config CRYPTO_TWOFISH
 	tristate "Twofish cipher algorithm"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	select CRYPTO_TWOFISH_COMMON
 	help
 	  Twofish cipher algorithm.
@@ -145,14 +154,14 @@ config CRYPTO_TWOFISH
 
 config CRYPTO_TWOFISH_COMMON
 	tristate
-	depends on CRYPTO
 	help
 	  Common parts of the Twofish cipher algorithm shared by the
 	  generic c and the assembler implementations.
 
 config CRYPTO_TWOFISH_586
 	tristate "Twofish cipher algorithms (i586)"
-	depends on CRYPTO && ((X86 || UML_X86) && !64BIT)
+	depends on (X86 || UML_X86) && !64BIT
+	select CRYPTO_ALGAPI
 	select CRYPTO_TWOFISH_COMMON
 	help
 	  Twofish cipher algorithm.
@@ -167,7 +176,8 @@ config CRYPTO_TWOFISH_586
 
 config CRYPTO_TWOFISH_X86_64
 	tristate "Twofish cipher algorithm (x86_64)"
-	depends on CRYPTO && ((X86 || UML_X86) && 64BIT)
+	depends on (X86 || UML_X86) && 64BIT
+	select CRYPTO_ALGAPI
 	select CRYPTO_TWOFISH_COMMON
 	help
 	  Twofish cipher algorithm (x86_64).
@@ -182,7 +192,7 @@ config CRYPTO_TWOFISH_X86_64
 
 config CRYPTO_SERPENT
 	tristate "Serpent cipher algorithm"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  Serpent cipher algorithm, by Anderson, Biham & Knudsen.
 
@@ -195,7 +205,7 @@ config CRYPTO_SERPENT
 
 config CRYPTO_AES
 	tristate "AES cipher algorithms"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  AES cipher algorithms (FIPS-197). AES uses the Rijndael 
 	  algorithm.
@@ -215,7 +225,8 @@ config CRYPTO_AES
 
 config CRYPTO_AES_586
 	tristate "AES cipher algorithms (i586)"
-	depends on CRYPTO && ((X86 || UML_X86) && !64BIT)
+	depends on (X86 || UML_X86) && !64BIT
+	select CRYPTO_ALGAPI
 	help
 	  AES cipher algorithms (FIPS-197). AES uses the Rijndael 
 	  algorithm.
@@ -235,7 +246,8 @@ config CRYPTO_AES_586
 
 config CRYPTO_AES_X86_64
 	tristate "AES cipher algorithms (x86_64)"
-	depends on CRYPTO && ((X86 || UML_X86) && 64BIT)
+	depends on (X86 || UML_X86) && 64BIT
+	select CRYPTO_ALGAPI
 	help
 	  AES cipher algorithms (FIPS-197). AES uses the Rijndael 
 	  algorithm.
@@ -255,7 +267,8 @@ config CRYPTO_AES_X86_64
 
 config CRYPTO_AES_S390
 	tristate "AES cipher algorithms (s390)"
-	depends on CRYPTO && S390
+	depends on S390
+	select CRYPTO_ALGAPI
 	help
 	  This is the s390 hardware accelerated implementation of the
 	  AES cipher algorithms (FIPS-197). AES uses the Rijndael
@@ -275,21 +288,21 @@ config CRYPTO_AES_S390
 
 config CRYPTO_CAST5
 	tristate "CAST5 (CAST-128) cipher algorithm"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  The CAST5 encryption algorithm (synonymous with CAST-128) is
 	  described in RFC2144.
 
 config CRYPTO_CAST6
 	tristate "CAST6 (CAST-256) cipher algorithm"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  The CAST6 encryption algorithm (synonymous with CAST-256) is
 	  described in RFC2612.
 
 config CRYPTO_TEA
 	tristate "TEA, XTEA and XETA cipher algorithms"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  TEA cipher algorithm.
 
@@ -306,7 +319,7 @@ config CRYPTO_TEA
 
 config CRYPTO_ARC4
 	tristate "ARC4 cipher algorithm"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  ARC4 cipher algorithm.
 
@@ -317,7 +330,7 @@ config CRYPTO_ARC4
 
 config CRYPTO_KHAZAD
 	tristate "Khazad cipher algorithm"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  Khazad cipher algorithm.
 
@@ -330,7 +343,7 @@ config CRYPTO_KHAZAD
 
 config CRYPTO_ANUBIS
 	tristate "Anubis cipher algorithm"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  Anubis cipher algorithm.
 
@@ -345,7 +358,7 @@ config CRYPTO_ANUBIS
 
 config CRYPTO_DEFLATE
 	tristate "Deflate compression algorithm"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	select ZLIB_INFLATE
 	select ZLIB_DEFLATE
 	help
@@ -356,7 +369,7 @@ config CRYPTO_DEFLATE
 
 config CRYPTO_MICHAEL_MIC
 	tristate "Michael MIC keyed digest algorithm"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	help
 	  Michael MIC is used for message integrity protection in TKIP
 	  (IEEE 802.11i). This algorithm is required for TKIP, but it
@@ -365,7 +378,7 @@ config CRYPTO_MICHAEL_MIC
 
 config CRYPTO_CRC32C
 	tristate "CRC32c CRC algorithm"
-	depends on CRYPTO
+	select CRYPTO_ALGAPI
 	select LIBCRC32C
 	help
 	  Castagnoli, et al Cyclic Redundancy-Check Algorithm.  Used
@@ -375,10 +388,13 @@ config CRYPTO_CRC32C
 
 config CRYPTO_TEST
 	tristate "Testing module"
-	depends on CRYPTO && m
+	depends on m
+	select CRYPTO_ALGAPI
 	help
 	  Quick & dirty crypto test module.
 
 source "drivers/crypto/Kconfig"
-endmenu
 
+endif	# if CRYPTO
+
+endmenu
diff --git a/crypto/Makefile b/crypto/Makefile
index fe934f1001c6..6d51f80753a1 100644
--- a/crypto/Makefile
+++ b/crypto/Makefile
@@ -2,10 +2,11 @@
 # Cryptographic API
 #
 
-proc-crypto-$(CONFIG_PROC_FS) = proc.o
+obj-$(CONFIG_CRYPTO) += api.o scatterwalk.o cipher.o digest.o compress.o
 
-obj-$(CONFIG_CRYPTO) += api.o scatterwalk.o cipher.o digest.o compress.o \
-			$(proc-crypto-y)
+crypto_algapi-$(CONFIG_PROC_FS) += proc.o
+crypto_algapi-objs := algapi.o $(crypto_algapi-y)
+obj-$(CONFIG_CRYPTO_ALGAPI) += crypto_algapi.o
 
 obj-$(CONFIG_CRYPTO_HMAC) += hmac.o
 obj-$(CONFIG_CRYPTO_NULL) += crypto_null.o
diff --git a/crypto/algapi.c b/crypto/algapi.c
new file mode 100644
index 000000000000..a65c6ccfbe17
--- /dev/null
+++ b/crypto/algapi.c
@@ -0,0 +1,118 @@
+/*
+ * Cryptographic API for algorithms (i.e., low-level API).
+ *
+ * Copyright (c) 2006 Herbert Xu <herbert@gondor.apana.org.au>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; either version 2 of the License, or (at your option)
+ * any later version.
+ *
+ */
+
+#include <linux/errno.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/string.h>
+
+#include "internal.h"
+
+static inline int crypto_set_driver_name(struct crypto_alg *alg)
+{
+	static const char suffix[] = "-generic";
+	char *driver_name = alg->cra_driver_name;
+	int len;
+
+	if (*driver_name)
+		return 0;
+
+	len = strlcpy(driver_name, alg->cra_name, CRYPTO_MAX_ALG_NAME);
+	if (len + sizeof(suffix) > CRYPTO_MAX_ALG_NAME)
+		return -ENAMETOOLONG;
+
+	memcpy(driver_name + len, suffix, sizeof(suffix));
+	return 0;
+}
+
+int crypto_register_alg(struct crypto_alg *alg)
+{
+	int ret;
+	struct crypto_alg *q;
+
+	if (alg->cra_alignmask & (alg->cra_alignmask + 1))
+		return -EINVAL;
+
+	if (alg->cra_alignmask & alg->cra_blocksize)
+		return -EINVAL;
+
+	if (alg->cra_blocksize > PAGE_SIZE / 8)
+		return -EINVAL;
+
+	if (alg->cra_priority < 0)
+		return -EINVAL;
+	
+	ret = crypto_set_driver_name(alg);
+	if (unlikely(ret))
+		return ret;
+
+	down_write(&crypto_alg_sem);
+	
+	list_for_each_entry(q, &crypto_alg_list, cra_list) {
+		if (q == alg) {
+			ret = -EEXIST;
+			goto out;
+		}
+	}
+	
+	list_add(&alg->cra_list, &crypto_alg_list);
+	atomic_set(&alg->cra_refcnt, 1);
+out:	
+	up_write(&crypto_alg_sem);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(crypto_register_alg);
+
+int crypto_unregister_alg(struct crypto_alg *alg)
+{
+	int ret = -ENOENT;
+	struct crypto_alg *q;
+	
+	down_write(&crypto_alg_sem);
+	list_for_each_entry(q, &crypto_alg_list, cra_list) {
+		if (alg == q) {
+			list_del(&alg->cra_list);
+			ret = 0;
+			goto out;
+		}
+	}
+out:	
+	up_write(&crypto_alg_sem);
+
+	if (ret)
+		return ret;
+
+	BUG_ON(atomic_read(&alg->cra_refcnt) != 1);
+	if (alg->cra_destroy)
+		alg->cra_destroy(alg);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(crypto_unregister_alg);
+
+static int __init crypto_algapi_init(void)
+{
+	crypto_init_proc();
+	return 0;
+}
+
+static void __exit crypto_algapi_exit(void)
+{
+	crypto_exit_proc();
+}
+
+module_init(crypto_algapi_init);
+module_exit(crypto_algapi_exit);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Cryptographic algorithms API");
diff --git a/crypto/api.c b/crypto/api.c
index 5994a58ef954..c922090b4842 100644
--- a/crypto/api.c
+++ b/crypto/api.c
@@ -15,19 +15,17 @@
  *
  */
 
-#include <linux/compiler.h>
-#include <linux/init.h>
-#include <linux/crypto.h>
 #include <linux/errno.h>
 #include <linux/kernel.h>
 #include <linux/kmod.h>
-#include <linux/rwsem.h>
 #include <linux/slab.h>
 #include <linux/string.h>
 #include "internal.h"
 
 LIST_HEAD(crypto_alg_list);
+EXPORT_SYMBOL_GPL(crypto_alg_list);
 DECLARE_RWSEM(crypto_alg_sem);
+EXPORT_SYMBOL_GPL(crypto_alg_sem);
 
 static inline struct crypto_alg *crypto_alg_get(struct crypto_alg *alg)
 {
@@ -239,86 +237,6 @@ void crypto_free_tfm(struct crypto_tfm *tfm)
 	kfree(tfm);
 }
 
-static inline int crypto_set_driver_name(struct crypto_alg *alg)
-{
-	static const char suffix[] = "-generic";
-	char *driver_name = alg->cra_driver_name;
-	int len;
-
-	if (*driver_name)
-		return 0;
-
-	len = strlcpy(driver_name, alg->cra_name, CRYPTO_MAX_ALG_NAME);
-	if (len + sizeof(suffix) > CRYPTO_MAX_ALG_NAME)
-		return -ENAMETOOLONG;
-
-	memcpy(driver_name + len, suffix, sizeof(suffix));
-	return 0;
-}
-
-int crypto_register_alg(struct crypto_alg *alg)
-{
-	int ret;
-	struct crypto_alg *q;
-
-	if (alg->cra_alignmask & (alg->cra_alignmask + 1))
-		return -EINVAL;
-
-	if (alg->cra_alignmask & alg->cra_blocksize)
-		return -EINVAL;
-
-	if (alg->cra_blocksize > PAGE_SIZE / 8)
-		return -EINVAL;
-
-	if (alg->cra_priority < 0)
-		return -EINVAL;
-	
-	ret = crypto_set_driver_name(alg);
-	if (unlikely(ret))
-		return ret;
-
-	down_write(&crypto_alg_sem);
-	
-	list_for_each_entry(q, &crypto_alg_list, cra_list) {
-		if (q == alg) {
-			ret = -EEXIST;
-			goto out;
-		}
-	}
-	
-	list_add(&alg->cra_list, &crypto_alg_list);
-	atomic_set(&alg->cra_refcnt, 1);
-out:	
-	up_write(&crypto_alg_sem);
-	return ret;
-}
-
-int crypto_unregister_alg(struct crypto_alg *alg)
-{
-	int ret = -ENOENT;
-	struct crypto_alg *q;
-	
-	down_write(&crypto_alg_sem);
-	list_for_each_entry(q, &crypto_alg_list, cra_list) {
-		if (alg == q) {
-			list_del(&alg->cra_list);
-			ret = 0;
-			goto out;
-		}
-	}
-out:	
-	up_write(&crypto_alg_sem);
-
-	if (ret)
-		return ret;
-
-	BUG_ON(atomic_read(&alg->cra_refcnt) != 1);
-	if (alg->cra_destroy)
-		alg->cra_destroy(alg);
-
-	return 0;
-}
-
 int crypto_alg_available(const char *name, u32 flags)
 {
 	int ret = 0;
@@ -332,17 +250,6 @@ int crypto_alg_available(const char *name, u32 flags)
 	return ret;
 }
 
-static int __init init_crypto(void)
-{
-	printk(KERN_INFO "Initializing Cryptographic API\n");
-	crypto_init_proc();
-	return 0;
-}
-
-__initcall(init_crypto);
-
-EXPORT_SYMBOL_GPL(crypto_register_alg);
-EXPORT_SYMBOL_GPL(crypto_unregister_alg);
 EXPORT_SYMBOL_GPL(crypto_alloc_tfm);
 EXPORT_SYMBOL_GPL(crypto_free_tfm);
 EXPORT_SYMBOL_GPL(crypto_alg_available);
diff --git a/crypto/internal.h b/crypto/internal.h
index 959e602909a6..26f47d331551 100644
--- a/crypto/internal.h
+++ b/crypto/internal.h
@@ -12,7 +12,8 @@
  */
 #ifndef _CRYPTO_INTERNAL_H
 #define _CRYPTO_INTERNAL_H
-#include <linux/crypto.h>
+
+#include <crypto/algapi.h>
 #include <linux/mm.h>
 #include <linux/highmem.h>
 #include <linux/interrupt.h>
@@ -64,9 +65,12 @@ static inline void crypto_free_hmac_block(struct crypto_tfm *tfm)
 
 #ifdef CONFIG_PROC_FS
 void __init crypto_init_proc(void);
+void __exit crypto_exit_proc(void);
 #else
 static inline void crypto_init_proc(void)
 { }
+static inline void crypto_exit_proc(void)
+{ }
 #endif
 
 static inline unsigned int crypto_digest_ctxsize(struct crypto_alg *alg,
diff --git a/crypto/proc.c b/crypto/proc.c
index 8543b7a157d6..9e573b17e887 100644
--- a/crypto/proc.c
+++ b/crypto/proc.c
@@ -113,3 +113,8 @@ void __init crypto_init_proc(void)
 	if (proc)
 		proc->proc_fops = &proc_crypto_ops;
 }
+
+void __exit crypto_exit_proc(void)
+{
+	remove_proc_entry("crypto", NULL);
+}
diff --git a/drivers/crypto/Kconfig b/drivers/crypto/Kconfig
index 4263935443cc..ba23683ab8c4 100644
--- a/drivers/crypto/Kconfig
+++ b/drivers/crypto/Kconfig
@@ -2,7 +2,8 @@ menu "Hardware crypto devices"
 
 config CRYPTO_DEV_PADLOCK
 	tristate "Support for VIA PadLock ACE"
-	depends on CRYPTO && X86_32
+	depends on X86_32
+	select CRYPTO_ALGAPI
 	help
 	  Some VIA processors come with an integrated crypto engine
 	  (so called VIA PadLock ACE, Advanced Cryptography Engine)
diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h
new file mode 100644
index 000000000000..ed68d494b364
--- /dev/null
+++ b/include/crypto/algapi.h
@@ -0,0 +1,18 @@
+/*
+ * Cryptographic API for algorithms (i.e., low-level API).
+ *
+ * Copyright (c) 2006 Herbert Xu <herbert@gondor.apana.org.au>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; either version 2 of the License, or (at your option) 
+ * any later version.
+ *
+ */
+#ifndef _CRYPTO_ALGAPI_H
+#define _CRYPTO_ALGAPI_H
+
+#include <linux/crypto.h>
+
+#endif	/* _CRYPTO_ALGAPI_H */
+
-- 
cgit v1.2.3


From 4cc7720cd165273b08a72b4193146dffee58e34b Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sun, 6 Aug 2006 21:16:34 +1000
Subject: [CRYPTO] api: Add template registration

A crypto_template generates a crypto_alg object when given a set of
parameters.  this patch adds the basic data structure fo templates
and code to handle their registration/deregistration.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 crypto/algapi.c         | 156 +++++++++++++++++++++++++++++++++++++++++-------
 crypto/internal.h       |  17 ++++++
 include/crypto/algapi.h |  31 ++++++++++
 3 files changed, 182 insertions(+), 22 deletions(-)

(limited to 'include')

diff --git a/crypto/algapi.c b/crypto/algapi.c
index a65c6ccfbe17..232b37d81613 100644
--- a/crypto/algapi.c
+++ b/crypto/algapi.c
@@ -13,11 +13,14 @@
 #include <linux/errno.h>
 #include <linux/init.h>
 #include <linux/kernel.h>
+#include <linux/list.h>
 #include <linux/module.h>
 #include <linux/string.h>
 
 #include "internal.h"
 
+static LIST_HEAD(crypto_template_list);
+
 static inline int crypto_set_driver_name(struct crypto_alg *alg)
 {
 	static const char suffix[] = "-generic";
@@ -35,11 +38,8 @@ static inline int crypto_set_driver_name(struct crypto_alg *alg)
 	return 0;
 }
 
-int crypto_register_alg(struct crypto_alg *alg)
+static int crypto_check_alg(struct crypto_alg *alg)
 {
-	int ret;
-	struct crypto_alg *q;
-
 	if (alg->cra_alignmask & (alg->cra_alignmask + 1))
 		return -EINVAL;
 
@@ -51,42 +51,52 @@ int crypto_register_alg(struct crypto_alg *alg)
 
 	if (alg->cra_priority < 0)
 		return -EINVAL;
-	
-	ret = crypto_set_driver_name(alg);
-	if (unlikely(ret))
-		return ret;
 
-	down_write(&crypto_alg_sem);
-	
+	return crypto_set_driver_name(alg);
+}
+
+static int __crypto_register_alg(struct crypto_alg *alg)
+{
+	struct crypto_alg *q;
+	int ret = -EEXIST;
+
 	list_for_each_entry(q, &crypto_alg_list, cra_list) {
-		if (q == alg) {
-			ret = -EEXIST;
+		if (q == alg)
 			goto out;
-		}
 	}
 	
 	list_add(&alg->cra_list, &crypto_alg_list);
 	atomic_set(&alg->cra_refcnt, 1);
+	ret = 0;
 out:	
-	up_write(&crypto_alg_sem);
 	return ret;
 }
+
+int crypto_register_alg(struct crypto_alg *alg)
+{
+	int err;
+
+	err = crypto_check_alg(alg);
+	if (err)
+		return err;
+
+	down_write(&crypto_alg_sem);
+	err = __crypto_register_alg(alg);
+	up_write(&crypto_alg_sem);
+
+	return err;
+}
 EXPORT_SYMBOL_GPL(crypto_register_alg);
 
 int crypto_unregister_alg(struct crypto_alg *alg)
 {
 	int ret = -ENOENT;
-	struct crypto_alg *q;
 	
 	down_write(&crypto_alg_sem);
-	list_for_each_entry(q, &crypto_alg_list, cra_list) {
-		if (alg == q) {
-			list_del(&alg->cra_list);
-			ret = 0;
-			goto out;
-		}
+	if (likely(!list_empty(&alg->cra_list))) {
+		list_del_init(&alg->cra_list);
+		ret = 0;
 	}
-out:	
 	up_write(&crypto_alg_sem);
 
 	if (ret)
@@ -100,6 +110,108 @@ out:
 }
 EXPORT_SYMBOL_GPL(crypto_unregister_alg);
 
+int crypto_register_template(struct crypto_template *tmpl)
+{
+	struct crypto_template *q;
+	int err = -EEXIST;
+
+	down_write(&crypto_alg_sem);
+
+	list_for_each_entry(q, &crypto_template_list, list) {
+		if (q == tmpl)
+			goto out;
+	}
+
+	list_add(&tmpl->list, &crypto_template_list);
+	err = 0;
+out:
+	up_write(&crypto_alg_sem);
+	return err;
+}
+EXPORT_SYMBOL_GPL(crypto_register_template);
+
+void crypto_unregister_template(struct crypto_template *tmpl)
+{
+	struct crypto_instance *inst;
+	struct hlist_node *p, *n;
+	struct hlist_head *list;
+
+	down_write(&crypto_alg_sem);
+
+	BUG_ON(list_empty(&tmpl->list));
+	list_del_init(&tmpl->list);
+
+	list = &tmpl->instances;
+	hlist_for_each_entry(inst, p, list, list) {
+		BUG_ON(list_empty(&inst->alg.cra_list));
+		list_del_init(&inst->alg.cra_list);
+	}
+
+	up_write(&crypto_alg_sem);
+
+	hlist_for_each_entry_safe(inst, p, n, list, list) {
+		BUG_ON(atomic_read(&inst->alg.cra_refcnt) != 1);
+		tmpl->free(inst);
+	}
+}
+EXPORT_SYMBOL_GPL(crypto_unregister_template);
+
+static struct crypto_template *__crypto_lookup_template(const char *name)
+{
+	struct crypto_template *q, *tmpl = NULL;
+
+	down_read(&crypto_alg_sem);
+	list_for_each_entry(q, &crypto_template_list, list) {
+		if (strcmp(q->name, name))
+			continue;
+		if (unlikely(!crypto_tmpl_get(q)))
+			continue;
+
+		tmpl = q;
+		break;
+	}
+	up_read(&crypto_alg_sem);
+
+	return tmpl;
+}
+
+struct crypto_template *crypto_lookup_template(const char *name)
+{
+	return try_then_request_module(__crypto_lookup_template(name), name);
+}
+EXPORT_SYMBOL_GPL(crypto_lookup_template);
+
+int crypto_register_instance(struct crypto_template *tmpl,
+			     struct crypto_instance *inst)
+{
+	int err = -EINVAL;
+
+	if (inst->alg.cra_destroy)
+		goto err;
+
+	err = crypto_check_alg(&inst->alg);
+	if (err)
+		goto err;
+
+	inst->alg.cra_module = tmpl->module;
+
+	down_write(&crypto_alg_sem);
+
+	err = __crypto_register_alg(&inst->alg);
+	if (err)
+		goto unlock;
+
+	hlist_add_head(&inst->list, &tmpl->instances);
+	inst->tmpl = tmpl;
+
+unlock:
+	up_write(&crypto_alg_sem);
+
+err:
+	return err;
+}
+EXPORT_SYMBOL_GPL(crypto_register_instance);
+
 static int __init crypto_algapi_init(void)
 {
 	crypto_init_proc();
diff --git a/crypto/internal.h b/crypto/internal.h
index 26f47d331551..c3ab4a950f30 100644
--- a/crypto/internal.h
+++ b/crypto/internal.h
@@ -19,11 +19,15 @@
 #include <linux/interrupt.h>
 #include <linux/init.h>
 #include <linux/list.h>
+#include <linux/module.h>
 #include <linux/kernel.h>
 #include <linux/rwsem.h>
 #include <linux/slab.h>
 #include <asm/kmap_types.h>
 
+struct crypto_instance;
+struct crypto_template;
+
 extern struct list_head crypto_alg_list;
 extern struct rw_semaphore crypto_alg_sem;
 
@@ -112,5 +116,18 @@ void crypto_exit_digest_ops(struct crypto_tfm *tfm);
 void crypto_exit_cipher_ops(struct crypto_tfm *tfm);
 void crypto_exit_compress_ops(struct crypto_tfm *tfm);
 
+int crypto_register_instance(struct crypto_template *tmpl,
+			     struct crypto_instance *inst);
+
+static inline int crypto_tmpl_get(struct crypto_template *tmpl)
+{
+	return try_module_get(tmpl->module);
+}
+
+static inline void crypto_tmpl_put(struct crypto_template *tmpl)
+{
+	module_put(tmpl->module);
+}
+
 #endif	/* _CRYPTO_INTERNAL_H */
 
diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h
index ed68d494b364..ffec530d52fb 100644
--- a/include/crypto/algapi.h
+++ b/include/crypto/algapi.h
@@ -14,5 +14,36 @@
 
 #include <linux/crypto.h>
 
+struct module;
+
+struct crypto_instance {
+	struct crypto_alg alg;
+
+	struct crypto_template *tmpl;
+	struct hlist_node list;
+
+	void *__ctx[] CRYPTO_MINALIGN_ATTR;
+};
+
+struct crypto_template {
+	struct list_head list;
+	struct hlist_head instances;
+	struct module *module;
+
+	struct crypto_instance *(*alloc)(void *param, unsigned int len);
+	void (*free)(struct crypto_instance *inst);
+
+	char name[CRYPTO_MAX_ALG_NAME];
+};
+
+int crypto_register_template(struct crypto_template *tmpl);
+void crypto_unregister_template(struct crypto_template *tmpl);
+struct crypto_template *crypto_lookup_template(const char *name);
+
+static inline void *crypto_instance_ctx(struct crypto_instance *inst)
+{
+	return inst->__ctx;
+}
+
 #endif	/* _CRYPTO_ALGAPI_H */
 
-- 
cgit v1.2.3


From 2825982d9d66ebba4b532a07391dfbb357f71c5f Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sun, 6 Aug 2006 21:23:26 +1000
Subject: [CRYPTO] api: Added event notification

This patch adds a notifier chain for algorithm/template registration events.
This will be used to register compound algorithms such as cbc(aes).  In
future this will also be passed onto user-space through netlink.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 crypto/algapi.c        |  50 +++++++++++++++++++-
 crypto/api.c           | 122 ++++++++++++++++++++++++++++++++++++++++++++-----
 crypto/internal.h      |  37 +++++++++++++++
 include/linux/crypto.h |   4 +-
 4 files changed, 199 insertions(+), 14 deletions(-)

(limited to 'include')

diff --git a/crypto/algapi.c b/crypto/algapi.c
index 232b37d81613..f0df85fc1f50 100644
--- a/crypto/algapi.c
+++ b/crypto/algapi.c
@@ -21,6 +21,24 @@
 
 static LIST_HEAD(crypto_template_list);
 
+void crypto_larval_error(const char *name)
+{
+	struct crypto_alg *alg;
+
+	down_read(&crypto_alg_sem);
+	alg = __crypto_alg_lookup(name);
+	up_read(&crypto_alg_sem);
+
+	if (alg) {
+		if (crypto_is_larval(alg)) {
+			struct crypto_larval *larval = (void *)alg;
+			complete(&larval->completion);
+		}
+		crypto_mod_put(alg);
+	}
+}
+EXPORT_SYMBOL_GPL(crypto_larval_error);
+
 static inline int crypto_set_driver_name(struct crypto_alg *alg)
 {
 	static const char suffix[] = "-generic";
@@ -60,14 +78,27 @@ static int __crypto_register_alg(struct crypto_alg *alg)
 	struct crypto_alg *q;
 	int ret = -EEXIST;
 
+	atomic_set(&alg->cra_refcnt, 1);
 	list_for_each_entry(q, &crypto_alg_list, cra_list) {
 		if (q == alg)
 			goto out;
+		if (crypto_is_larval(q) &&
+		    (!strcmp(alg->cra_name, q->cra_name) ||
+		     !strcmp(alg->cra_driver_name, q->cra_name))) {
+			struct crypto_larval *larval = (void *)q;
+
+			if (!crypto_mod_get(alg))
+				continue;
+			larval->adult = alg;
+			complete(&larval->completion);
+		}
 	}
 	
 	list_add(&alg->cra_list, &crypto_alg_list);
-	atomic_set(&alg->cra_refcnt, 1);
+
+	crypto_notify(CRYPTO_MSG_ALG_REGISTER, alg);
 	ret = 0;
+
 out:	
 	return ret;
 }
@@ -97,6 +128,7 @@ int crypto_unregister_alg(struct crypto_alg *alg)
 		list_del_init(&alg->cra_list);
 		ret = 0;
 	}
+	crypto_notify(CRYPTO_MSG_ALG_UNREGISTER, alg);
 	up_write(&crypto_alg_sem);
 
 	if (ret)
@@ -123,6 +155,7 @@ int crypto_register_template(struct crypto_template *tmpl)
 	}
 
 	list_add(&tmpl->list, &crypto_template_list);
+	crypto_notify(CRYPTO_MSG_TMPL_REGISTER, tmpl);
 	err = 0;
 out:
 	up_write(&crypto_alg_sem);
@@ -145,8 +178,11 @@ void crypto_unregister_template(struct crypto_template *tmpl)
 	hlist_for_each_entry(inst, p, list, list) {
 		BUG_ON(list_empty(&inst->alg.cra_list));
 		list_del_init(&inst->alg.cra_list);
+		crypto_notify(CRYPTO_MSG_ALG_UNREGISTER, &inst->alg);
 	}
 
+	crypto_notify(CRYPTO_MSG_TMPL_UNREGISTER, tmpl);
+
 	up_write(&crypto_alg_sem);
 
 	hlist_for_each_entry_safe(inst, p, n, list, list) {
@@ -212,6 +248,18 @@ err:
 }
 EXPORT_SYMBOL_GPL(crypto_register_instance);
 
+int crypto_register_notifier(struct notifier_block *nb)
+{
+	return blocking_notifier_chain_register(&crypto_chain, nb);
+}
+EXPORT_SYMBOL_GPL(crypto_register_notifier);
+
+int crypto_unregister_notifier(struct notifier_block *nb)
+{
+	return blocking_notifier_chain_unregister(&crypto_chain, nb);
+}
+EXPORT_SYMBOL_GPL(crypto_unregister_notifier);
+
 static int __init crypto_algapi_init(void)
 {
 	crypto_init_proc();
diff --git a/crypto/api.c b/crypto/api.c
index c922090b4842..5a0d6a17cfd7 100644
--- a/crypto/api.c
+++ b/crypto/api.c
@@ -18,6 +18,7 @@
 #include <linux/errno.h>
 #include <linux/kernel.h>
 #include <linux/kmod.h>
+#include <linux/param.h>
 #include <linux/slab.h>
 #include <linux/string.h>
 #include "internal.h"
@@ -27,6 +28,9 @@ EXPORT_SYMBOL_GPL(crypto_alg_list);
 DECLARE_RWSEM(crypto_alg_sem);
 EXPORT_SYMBOL_GPL(crypto_alg_sem);
 
+BLOCKING_NOTIFIER_HEAD(crypto_chain);
+EXPORT_SYMBOL_GPL(crypto_chain);
+
 static inline struct crypto_alg *crypto_alg_get(struct crypto_alg *alg)
 {
 	atomic_inc(&alg->cra_refcnt);
@@ -39,27 +43,24 @@ static inline void crypto_alg_put(struct crypto_alg *alg)
 		alg->cra_destroy(alg);
 }
 
-static struct crypto_alg *crypto_mod_get(struct crypto_alg *alg)
+struct crypto_alg *crypto_mod_get(struct crypto_alg *alg)
 {
 	return try_module_get(alg->cra_module) ? crypto_alg_get(alg) : NULL;
 }
+EXPORT_SYMBOL_GPL(crypto_mod_get);
 
-static void crypto_mod_put(struct crypto_alg *alg)
+void crypto_mod_put(struct crypto_alg *alg)
 {
 	crypto_alg_put(alg);
 	module_put(alg->cra_module);
 }
+EXPORT_SYMBOL_GPL(crypto_mod_put);
 
-static struct crypto_alg *crypto_alg_lookup(const char *name)
+struct crypto_alg *__crypto_alg_lookup(const char *name)
 {
 	struct crypto_alg *q, *alg = NULL;
-	int best = -1;
+	int best = -2;
 
-	if (!name)
-		return NULL;
-	
-	down_read(&crypto_alg_sem);
-	
 	list_for_each_entry(q, &crypto_alg_list, cra_list) {
 		int exact, fuzzy;
 
@@ -79,16 +80,113 @@ static struct crypto_alg *crypto_alg_lookup(const char *name)
 		if (exact)
 			break;
 	}
-	
+
+	return alg;
+}
+EXPORT_SYMBOL_GPL(__crypto_alg_lookup);
+
+static void crypto_larval_destroy(struct crypto_alg *alg)
+{
+	struct crypto_larval *larval = (void *)alg;
+
+	BUG_ON(!crypto_is_larval(alg));
+	if (larval->adult)
+		crypto_mod_put(larval->adult);
+	kfree(larval);
+}
+
+static struct crypto_alg *crypto_larval_alloc(const char *name)
+{
+	struct crypto_alg *alg;
+	struct crypto_larval *larval;
+
+	larval = kzalloc(sizeof(*larval), GFP_KERNEL);
+	if (!larval)
+		return NULL;
+
+	larval->alg.cra_flags = CRYPTO_ALG_LARVAL;
+	larval->alg.cra_priority = -1;
+	larval->alg.cra_destroy = crypto_larval_destroy;
+
+	atomic_set(&larval->alg.cra_refcnt, 2);
+	strlcpy(larval->alg.cra_name, name, CRYPTO_MAX_ALG_NAME);
+	init_completion(&larval->completion);
+
+	down_write(&crypto_alg_sem);
+	alg = __crypto_alg_lookup(name);
+	if (!alg) {
+		alg = &larval->alg;
+		list_add(&alg->cra_list, &crypto_alg_list);
+	}
+	up_write(&crypto_alg_sem);
+
+	if (alg != &larval->alg)
+		kfree(larval);
+
+	return alg;
+}
+
+static void crypto_larval_kill(struct crypto_alg *alg)
+{
+	struct crypto_larval *larval = (void *)alg;
+
+	down_write(&crypto_alg_sem);
+	list_del(&alg->cra_list);
+	up_write(&crypto_alg_sem);
+	complete(&larval->completion);
+	crypto_alg_put(alg);
+}
+
+static struct crypto_alg *crypto_larval_wait(struct crypto_alg *alg)
+{
+	struct crypto_larval *larval = (void *)alg;
+
+	wait_for_completion_interruptible_timeout(&larval->completion, 60 * HZ);
+	alg = larval->adult;
+	if (alg && !crypto_mod_get(alg))
+		alg = NULL;
+	crypto_mod_put(&larval->alg);
+
+	return alg;
+}
+
+static struct crypto_alg *crypto_alg_lookup(const char *name)
+{
+	struct crypto_alg *alg;
+
+	if (!name)
+		return NULL;
+
+	down_read(&crypto_alg_sem);
+	alg = __crypto_alg_lookup(name);
 	up_read(&crypto_alg_sem);
+
 	return alg;
 }
 
 /* A far more intelligent version of this is planned.  For now, just
  * try an exact match on the name of the algorithm. */
-static inline struct crypto_alg *crypto_alg_mod_lookup(const char *name)
+static struct crypto_alg *crypto_alg_mod_lookup(const char *name)
 {
-	return try_then_request_module(crypto_alg_lookup(name), name);
+	struct crypto_alg *alg;
+	struct crypto_alg *larval;
+
+	alg = try_then_request_module(crypto_alg_lookup(name), name);
+	if (alg)
+		return crypto_is_larval(alg) ? crypto_larval_wait(alg) : alg;
+
+	larval = crypto_larval_alloc(name);
+	if (!larval || !crypto_is_larval(larval))
+		return larval;
+
+	if (crypto_notify(CRYPTO_MSG_ALG_REQUEST, larval) == NOTIFY_STOP)
+		alg = crypto_larval_wait(larval);
+	else {
+		crypto_mod_put(larval);
+		alg = NULL;
+	}
+	crypto_larval_kill(larval);
+	return alg;
 }
 
 static int crypto_init_flags(struct crypto_tfm *tfm, u32 flags)
diff --git a/crypto/internal.h b/crypto/internal.h
index c3ab4a950f30..3a08d25fba45 100644
--- a/crypto/internal.h
+++ b/crypto/internal.h
@@ -14,6 +14,7 @@
 #define _CRYPTO_INTERNAL_H
 
 #include <crypto/algapi.h>
+#include <linux/completion.h>
 #include <linux/mm.h>
 #include <linux/highmem.h>
 #include <linux/interrupt.h>
@@ -21,15 +22,32 @@
 #include <linux/list.h>
 #include <linux/module.h>
 #include <linux/kernel.h>
+#include <linux/notifier.h>
 #include <linux/rwsem.h>
 #include <linux/slab.h>
 #include <asm/kmap_types.h>
 
+/* Crypto notification events. */
+enum {
+	CRYPTO_MSG_ALG_REQUEST,
+	CRYPTO_MSG_ALG_REGISTER,
+	CRYPTO_MSG_ALG_UNREGISTER,
+	CRYPTO_MSG_TMPL_REGISTER,
+	CRYPTO_MSG_TMPL_UNREGISTER,
+};
+
 struct crypto_instance;
 struct crypto_template;
 
+struct crypto_larval {
+	struct crypto_alg alg;
+	struct crypto_alg *adult;
+	struct completion completion;
+};
+
 extern struct list_head crypto_alg_list;
 extern struct rw_semaphore crypto_alg_sem;
+extern struct blocking_notifier_head crypto_chain;
 
 extern enum km_type crypto_km_types[];
 
@@ -104,6 +122,10 @@ static inline unsigned int crypto_compress_ctxsize(struct crypto_alg *alg,
 	return alg->cra_ctxsize;
 }
 
+struct crypto_alg *crypto_mod_get(struct crypto_alg *alg);
+void crypto_mod_put(struct crypto_alg *alg);
+struct crypto_alg *__crypto_alg_lookup(const char *name);
+
 int crypto_init_digest_flags(struct crypto_tfm *tfm, u32 flags);
 int crypto_init_cipher_flags(struct crypto_tfm *tfm, u32 flags);
 int crypto_init_compress_flags(struct crypto_tfm *tfm, u32 flags);
@@ -116,9 +138,14 @@ void crypto_exit_digest_ops(struct crypto_tfm *tfm);
 void crypto_exit_cipher_ops(struct crypto_tfm *tfm);
 void crypto_exit_compress_ops(struct crypto_tfm *tfm);
 
+void crypto_larval_error(const char *name);
+
 int crypto_register_instance(struct crypto_template *tmpl,
 			     struct crypto_instance *inst);
 
+int crypto_register_notifier(struct notifier_block *nb);
+int crypto_unregister_notifier(struct notifier_block *nb);
+
 static inline int crypto_tmpl_get(struct crypto_template *tmpl)
 {
 	return try_module_get(tmpl->module);
@@ -129,5 +156,15 @@ static inline void crypto_tmpl_put(struct crypto_template *tmpl)
 	module_put(tmpl->module);
 }
 
+static inline int crypto_is_larval(struct crypto_alg *alg)
+{
+	return alg->cra_flags & CRYPTO_ALG_LARVAL;
+}
+
+static inline int crypto_notify(unsigned long val, void *v)
+{
+	return blocking_notifier_call_chain(&crypto_chain, val, v);
+}
+
 #endif	/* _CRYPTO_INTERNAL_H */
 
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index 7f57ff8ec975..3e3e95aff133 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -29,11 +29,13 @@
 /*
  * Algorithm masks and types.
  */
-#define CRYPTO_ALG_TYPE_MASK		0x000000ff
+#define CRYPTO_ALG_TYPE_MASK		0x0000000f
 #define CRYPTO_ALG_TYPE_CIPHER		0x00000001
 #define CRYPTO_ALG_TYPE_DIGEST		0x00000002
 #define CRYPTO_ALG_TYPE_COMPRESS	0x00000004
 
+#define CRYPTO_ALG_LARVAL		0x00000010
+
 /*
  * Transform masks and values (for crt_flags).
  */
-- 
cgit v1.2.3


From 2b8c19dbdc692e81243a328725a02efb77b144a5 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Thu, 21 Sep 2006 11:31:44 +1000
Subject: [CRYPTO] api: Add cryptomgr

The cryptomgr module is a simple manager of crypto algorithm instances.
It ensures that parameterised algorithms of the type tmpl(alg) (e.g.,
cbc(aes)) are always created.

This is meant to satisfy the needs for most users.  For more complex
cases such as deeper combinations or multiple parameters, a netlink
module will be created which allows arbitrary expressions to be parsed
in user-space.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 crypto/Kconfig         |   8 +++
 crypto/Makefile        |   1 +
 crypto/api.c           |  10 +++-
 crypto/cryptomgr.c     | 146 +++++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/crypto.h |   9 +++
 5 files changed, 173 insertions(+), 1 deletion(-)
 create mode 100644 crypto/cryptomgr.c

(limited to 'include')

diff --git a/crypto/Kconfig b/crypto/Kconfig
index aabc63195222..4ce509dba329 100644
--- a/crypto/Kconfig
+++ b/crypto/Kconfig
@@ -16,6 +16,14 @@ config CRYPTO_ALGAPI
 	help
 	  This option provides the API for cryptographic algorithms.
 
+config CRYPTO_MANAGER
+	tristate "Cryptographic algorithm manager"
+	select CRYPTO_ALGAPI
+	default m
+	help
+	  Create default cryptographic template instantiations such as
+	  cbc(aes).
+
 config CRYPTO_HMAC
 	bool "HMAC support"
 	help
diff --git a/crypto/Makefile b/crypto/Makefile
index 6d51f80753a1..b8745f3d3595 100644
--- a/crypto/Makefile
+++ b/crypto/Makefile
@@ -8,6 +8,7 @@ crypto_algapi-$(CONFIG_PROC_FS) += proc.o
 crypto_algapi-objs := algapi.o $(crypto_algapi-y)
 obj-$(CONFIG_CRYPTO_ALGAPI) += crypto_algapi.o
 
+obj-$(CONFIG_CRYPTO_MANAGER) += cryptomgr.o
 obj-$(CONFIG_CRYPTO_HMAC) += hmac.o
 obj-$(CONFIG_CRYPTO_NULL) += crypto_null.o
 obj-$(CONFIG_CRYPTO_MD4) += md4.o
diff --git a/crypto/api.c b/crypto/api.c
index 5a0d6a17cfd7..67cd6f87b74a 100644
--- a/crypto/api.c
+++ b/crypto/api.c
@@ -18,6 +18,7 @@
 #include <linux/errno.h>
 #include <linux/kernel.h>
 #include <linux/kmod.h>
+#include <linux/module.h>
 #include <linux/param.h>
 #include <linux/slab.h>
 #include <linux/string.h>
@@ -170,6 +171,7 @@ static struct crypto_alg *crypto_alg_mod_lookup(const char *name)
 {
 	struct crypto_alg *alg;
 	struct crypto_alg *larval;
+	int ok;
 
 	alg = try_then_request_module(crypto_alg_lookup(name), name);
 	if (alg)
@@ -179,7 +181,13 @@ static struct crypto_alg *crypto_alg_mod_lookup(const char *name)
 	if (!larval || !crypto_is_larval(larval))
 		return larval;
 
-	if (crypto_notify(CRYPTO_MSG_ALG_REQUEST, larval) == NOTIFY_STOP)
+	ok = crypto_notify(CRYPTO_MSG_ALG_REQUEST, larval);
+	if (ok == NOTIFY_DONE) {
+		request_module("cryptomgr");
+		ok = crypto_notify(CRYPTO_MSG_ALG_REQUEST, larval);
+	}
+
+	if (ok == NOTIFY_STOP)
 		alg = crypto_larval_wait(larval);
 	else {
 		crypto_mod_put(larval);
diff --git a/crypto/cryptomgr.c b/crypto/cryptomgr.c
new file mode 100644
index 000000000000..e0ebe1b44b99
--- /dev/null
+++ b/crypto/cryptomgr.c
@@ -0,0 +1,146 @@
+/*
+ * Create default crypto algorithm instances.
+ *
+ * Copyright (c) 2006 Herbert Xu <herbert@gondor.apana.org.au>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; either version 2 of the License, or (at your option)
+ * any later version.
+ *
+ */
+
+#include <linux/crypto.h>
+#include <linux/ctype.h>
+#include <linux/err.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/notifier.h>
+#include <linux/rtnetlink.h>
+#include <linux/string.h>
+#include <linux/workqueue.h>
+
+#include "internal.h"
+
+struct cryptomgr_param {
+	struct work_struct work;
+
+	struct {
+		struct rtattr attr;
+		struct crypto_attr_alg data;
+	} alg;
+
+	struct {
+		char name[CRYPTO_MAX_ALG_NAME];
+	} larval;
+
+	char template[CRYPTO_MAX_ALG_NAME];
+};
+
+static void cryptomgr_probe(void *data)
+{
+	struct cryptomgr_param *param = data;
+	struct crypto_template *tmpl;
+	struct crypto_instance *inst;
+
+	tmpl = crypto_lookup_template(param->template);
+	if (!tmpl)
+		goto err;
+
+	inst = tmpl->alloc(&param->alg, sizeof(param->alg));
+	if (IS_ERR(inst))
+		goto err;
+	else if ((err = crypto_register_instance(tmpl, inst))) {
+		tmpl->free(inst);
+		goto err;
+	}
+
+	crypto_tmpl_put(tmpl);
+
+out:
+	kfree(param);
+	return;
+
+err:
+	crypto_larval_error(param->larval.name);
+	goto out;
+}
+
+static int cryptomgr_schedule_probe(struct crypto_larval *larval)
+{
+	struct cryptomgr_param *param;
+	const char *name = larval->alg.cra_name;
+	const char *p;
+	unsigned int len;
+
+	param = kmalloc(sizeof(*param), GFP_KERNEL);
+	if (!param)
+		goto err;
+
+	for (p = name; isalnum(*p) || *p == '-' || *p == '_'; p++)
+		;
+
+	len = p - name;
+	if (!len || *p != '(')
+		goto err_free_param;
+
+	memcpy(param->template, name, len);
+	param->template[len] = 0;
+
+	name = p + 1;
+	for (p = name; isalnum(*p) || *p == '-' || *p == '_'; p++)
+		;
+
+	len = p - name;
+	if (!len || *p != ')' || p[1])
+		goto err_free_param;
+
+	param->alg.attr.rta_len = sizeof(param->alg);
+	param->alg.attr.rta_type = CRYPTOA_ALG;
+	memcpy(param->alg.data.name, name, len);
+	param->alg.data.name[len] = 0;
+
+	memcpy(param->larval.name, larval->alg.cra_name, CRYPTO_MAX_ALG_NAME);
+
+	INIT_WORK(&param->work, cryptomgr_probe, param);
+	schedule_work(&param->work);
+
+	return NOTIFY_STOP;
+
+err_free_param:
+	kfree(param);
+err:
+	return NOTIFY_OK;
+}
+
+static int cryptomgr_notify(struct notifier_block *this, unsigned long msg,
+			    void *data)
+{
+	switch (msg) {
+	case CRYPTO_MSG_ALG_REQUEST:
+		return cryptomgr_schedule_probe(data);
+	}
+
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block cryptomgr_notifier = {
+	.notifier_call = cryptomgr_notify,
+};
+
+static int __init cryptomgr_init(void)
+{
+	return crypto_register_notifier(&cryptomgr_notifier);
+}
+
+static void __exit cryptomgr_exit(void)
+{
+	int err = crypto_unregister_notifier(&cryptomgr_notifier);
+	BUG_ON(err);
+}
+
+module_init(cryptomgr_init);
+module_exit(cryptomgr_exit);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Crypto Algorithm Manager");
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index 3e3e95aff133..85f73c381913 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -260,6 +260,15 @@ struct crypto_tfm {
 	void *__crt_ctx[] CRYPTO_MINALIGN_ATTR;
 };
 
+enum {
+	CRYPTOA_UNSPEC,
+	CRYPTOA_ALG,
+};
+
+struct crypto_attr_alg {
+	char name[CRYPTO_MAX_ALG_NAME];
+};
+
 /* 
  * Transform user interface.
  */
-- 
cgit v1.2.3


From 6bfd48096ff8ecabf955958b51ddfa7988eb0a14 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Thu, 21 Sep 2006 11:39:29 +1000
Subject: [CRYPTO] api: Added spawns

Spawns lock a specific crypto algorithm in place.  They can then be used
with crypto_spawn_tfm to allocate a tfm for that algorithm.  When the base
algorithm of a spawn is deregistered, all its spawns will be automatically
removed.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 crypto/algapi.c         | 185 +++++++++++++++++++++++++++++++++++++++++++-----
 crypto/api.c            |  95 +++++++++++++++++--------
 crypto/cryptomgr.c      |  19 +++--
 crypto/internal.h       |  19 +++++
 include/crypto/algapi.h |  11 +++
 include/linux/crypto.h  |   4 ++
 6 files changed, 280 insertions(+), 53 deletions(-)

(limited to 'include')

diff --git a/crypto/algapi.c b/crypto/algapi.c
index acea250677c0..36c4f1bdb521 100644
--- a/crypto/algapi.c
+++ b/crypto/algapi.c
@@ -10,6 +10,7 @@
  *
  */
 
+#include <linux/err.h>
 #include <linux/errno.h>
 #include <linux/init.h>
 #include <linux/kernel.h>
@@ -73,27 +74,96 @@ static int crypto_check_alg(struct crypto_alg *alg)
 	return crypto_set_driver_name(alg);
 }
 
-static int __crypto_register_alg(struct crypto_alg *alg)
+static void crypto_destroy_instance(struct crypto_alg *alg)
+{
+	struct crypto_instance *inst = (void *)alg;
+	struct crypto_template *tmpl = inst->tmpl;
+
+	tmpl->free(inst);
+	crypto_tmpl_put(tmpl);
+}
+
+static void crypto_remove_spawns(struct list_head *spawns,
+				 struct list_head *list)
+{
+	struct crypto_spawn *spawn, *n;
+
+	list_for_each_entry_safe(spawn, n, spawns, list) {
+		struct crypto_instance *inst = spawn->inst;
+		struct crypto_template *tmpl = inst->tmpl;
+
+		list_del_init(&spawn->list);
+		spawn->alg = NULL;
+
+		if (crypto_is_dead(&inst->alg))
+			continue;
+
+		inst->alg.cra_flags |= CRYPTO_ALG_DEAD;
+		if (!tmpl || !crypto_tmpl_get(tmpl))
+			continue;
+
+		crypto_notify(CRYPTO_MSG_ALG_UNREGISTER, &inst->alg);
+		list_move(&inst->alg.cra_list, list);
+		hlist_del(&inst->list);
+		inst->alg.cra_destroy = crypto_destroy_instance;
+
+		if (!list_empty(&inst->alg.cra_users)) {
+			if (&n->list == spawns)
+				n = list_entry(inst->alg.cra_users.next,
+					       typeof(*n), list);
+			__list_splice(&inst->alg.cra_users, spawns->prev);
+		}
+	}
+}
+
+static int __crypto_register_alg(struct crypto_alg *alg,
+				 struct list_head *list)
 {
 	struct crypto_alg *q;
-	int ret = -EEXIST;
+	int ret = -EAGAIN;
+
+	if (crypto_is_dead(alg))
+		goto out;
+
+	INIT_LIST_HEAD(&alg->cra_users);
+
+	ret = -EEXIST;
 
 	atomic_set(&alg->cra_refcnt, 1);
 	list_for_each_entry(q, &crypto_alg_list, cra_list) {
 		if (q == alg)
 			goto out;
-		if (crypto_is_larval(q) &&
-		    (!strcmp(alg->cra_name, q->cra_name) ||
-		     !strcmp(alg->cra_driver_name, q->cra_name))) {
+
+		if (crypto_is_moribund(q))
+			continue;
+
+		if (crypto_is_larval(q)) {
 			struct crypto_larval *larval = (void *)q;
 
+			if (strcmp(alg->cra_name, q->cra_name) &&
+			    strcmp(alg->cra_driver_name, q->cra_name))
+				continue;
+
+			if (larval->adult)
+				continue;
 			if ((q->cra_flags ^ alg->cra_flags) & larval->mask)
 				continue;
 			if (!crypto_mod_get(alg))
 				continue;
+
 			larval->adult = alg;
 			complete(&larval->completion);
+			continue;
 		}
+
+		if (strcmp(alg->cra_name, q->cra_name))
+			continue;
+
+		if (strcmp(alg->cra_driver_name, q->cra_driver_name) &&
+		    q->cra_priority > alg->cra_priority)
+			continue;
+
+		crypto_remove_spawns(&q->cra_users, list);
 	}
 	
 	list_add(&alg->cra_list, &crypto_alg_list);
@@ -105,8 +175,20 @@ out:
 	return ret;
 }
 
+static void crypto_remove_final(struct list_head *list)
+{
+	struct crypto_alg *alg;
+	struct crypto_alg *n;
+
+	list_for_each_entry_safe(alg, n, list, cra_list) {
+		list_del_init(&alg->cra_list);
+		crypto_alg_put(alg);
+	}
+}
+
 int crypto_register_alg(struct crypto_alg *alg)
 {
+	LIST_HEAD(list);
 	int err;
 
 	err = crypto_check_alg(alg);
@@ -114,23 +196,35 @@ int crypto_register_alg(struct crypto_alg *alg)
 		return err;
 
 	down_write(&crypto_alg_sem);
-	err = __crypto_register_alg(alg);
+	err = __crypto_register_alg(alg, &list);
 	up_write(&crypto_alg_sem);
 
+	crypto_remove_final(&list);
 	return err;
 }
 EXPORT_SYMBOL_GPL(crypto_register_alg);
 
+static int crypto_remove_alg(struct crypto_alg *alg, struct list_head *list)
+{
+	if (unlikely(list_empty(&alg->cra_list)))
+		return -ENOENT;
+
+	alg->cra_flags |= CRYPTO_ALG_DEAD;
+
+	crypto_notify(CRYPTO_MSG_ALG_UNREGISTER, alg);
+	list_del_init(&alg->cra_list);
+	crypto_remove_spawns(&alg->cra_users, list);
+
+	return 0;
+}
+
 int crypto_unregister_alg(struct crypto_alg *alg)
 {
-	int ret = -ENOENT;
+	int ret;
+	LIST_HEAD(list);
 	
 	down_write(&crypto_alg_sem);
-	if (likely(!list_empty(&alg->cra_list))) {
-		list_del_init(&alg->cra_list);
-		ret = 0;
-	}
-	crypto_notify(CRYPTO_MSG_ALG_UNREGISTER, alg);
+	ret = crypto_remove_alg(alg, &list);
 	up_write(&crypto_alg_sem);
 
 	if (ret)
@@ -140,6 +234,7 @@ int crypto_unregister_alg(struct crypto_alg *alg)
 	if (alg->cra_destroy)
 		alg->cra_destroy(alg);
 
+	crypto_remove_final(&list);
 	return 0;
 }
 EXPORT_SYMBOL_GPL(crypto_unregister_alg);
@@ -170,6 +265,7 @@ void crypto_unregister_template(struct crypto_template *tmpl)
 	struct crypto_instance *inst;
 	struct hlist_node *p, *n;
 	struct hlist_head *list;
+	LIST_HEAD(users);
 
 	down_write(&crypto_alg_sem);
 
@@ -178,9 +274,8 @@ void crypto_unregister_template(struct crypto_template *tmpl)
 
 	list = &tmpl->instances;
 	hlist_for_each_entry(inst, p, list, list) {
-		BUG_ON(list_empty(&inst->alg.cra_list));
-		list_del_init(&inst->alg.cra_list);
-		crypto_notify(CRYPTO_MSG_ALG_UNREGISTER, &inst->alg);
+		int err = crypto_remove_alg(&inst->alg, &users);
+		BUG_ON(err);
 	}
 
 	crypto_notify(CRYPTO_MSG_TMPL_UNREGISTER, tmpl);
@@ -191,6 +286,7 @@ void crypto_unregister_template(struct crypto_template *tmpl)
 		BUG_ON(atomic_read(&inst->alg.cra_refcnt) != 1);
 		tmpl->free(inst);
 	}
+	crypto_remove_final(&users);
 }
 EXPORT_SYMBOL_GPL(crypto_unregister_template);
 
@@ -222,6 +318,7 @@ EXPORT_SYMBOL_GPL(crypto_lookup_template);
 int crypto_register_instance(struct crypto_template *tmpl,
 			     struct crypto_instance *inst)
 {
+	LIST_HEAD(list);
 	int err = -EINVAL;
 
 	if (inst->alg.cra_destroy)
@@ -235,7 +332,7 @@ int crypto_register_instance(struct crypto_template *tmpl,
 
 	down_write(&crypto_alg_sem);
 
-	err = __crypto_register_alg(&inst->alg);
+	err = __crypto_register_alg(&inst->alg, &list);
 	if (err)
 		goto unlock;
 
@@ -245,11 +342,67 @@ int crypto_register_instance(struct crypto_template *tmpl,
 unlock:
 	up_write(&crypto_alg_sem);
 
+	crypto_remove_final(&list);
+
 err:
 	return err;
 }
 EXPORT_SYMBOL_GPL(crypto_register_instance);
 
+int crypto_init_spawn(struct crypto_spawn *spawn, struct crypto_alg *alg,
+		      struct crypto_instance *inst)
+{
+	int err = -EAGAIN;
+
+	spawn->inst = inst;
+
+	down_write(&crypto_alg_sem);
+	if (!crypto_is_moribund(alg)) {
+		list_add(&spawn->list, &alg->cra_users);
+		spawn->alg = alg;
+		err = 0;
+	}
+	up_write(&crypto_alg_sem);
+
+	return err;
+}
+EXPORT_SYMBOL_GPL(crypto_init_spawn);
+
+void crypto_drop_spawn(struct crypto_spawn *spawn)
+{
+	down_write(&crypto_alg_sem);
+	list_del(&spawn->list);
+	up_write(&crypto_alg_sem);
+}
+EXPORT_SYMBOL_GPL(crypto_drop_spawn);
+
+struct crypto_tfm *crypto_spawn_tfm(struct crypto_spawn *spawn)
+{
+	struct crypto_alg *alg;
+	struct crypto_alg *alg2;
+	struct crypto_tfm *tfm;
+
+	down_read(&crypto_alg_sem);
+	alg = spawn->alg;
+	alg2 = alg;
+	if (alg2)
+		alg2 = crypto_mod_get(alg2);
+	up_read(&crypto_alg_sem);
+
+	if (!alg2) {
+		if (alg)
+			crypto_shoot_alg(alg);
+		return ERR_PTR(-EAGAIN);
+	}
+
+	tfm = __crypto_alloc_tfm(alg, 0);
+	if (IS_ERR(tfm))
+		crypto_mod_put(alg);
+
+	return tfm;
+}
+EXPORT_SYMBOL_GPL(crypto_spawn_tfm);
+
 int crypto_register_notifier(struct notifier_block *nb)
 {
 	return blocking_notifier_chain_register(&crypto_chain, nb);
diff --git a/crypto/api.c b/crypto/api.c
index ddf6a767acdd..7e5522cf856e 100644
--- a/crypto/api.c
+++ b/crypto/api.c
@@ -15,11 +15,13 @@
  *
  */
 
+#include <linux/err.h>
 #include <linux/errno.h>
 #include <linux/kernel.h>
 #include <linux/kmod.h>
 #include <linux/module.h>
 #include <linux/param.h>
+#include <linux/sched.h>
 #include <linux/slab.h>
 #include <linux/string.h>
 #include "internal.h"
@@ -38,12 +40,6 @@ static inline struct crypto_alg *crypto_alg_get(struct crypto_alg *alg)
 	return alg;
 }
 
-static inline void crypto_alg_put(struct crypto_alg *alg)
-{
-	if (atomic_dec_and_test(&alg->cra_refcnt) && alg->cra_destroy)
-		alg->cra_destroy(alg);
-}
-
 struct crypto_alg *crypto_mod_get(struct crypto_alg *alg)
 {
 	return try_module_get(alg->cra_module) ? crypto_alg_get(alg) : NULL;
@@ -65,6 +61,9 @@ struct crypto_alg *__crypto_alg_lookup(const char *name, u32 type, u32 mask)
 	list_for_each_entry(q, &crypto_alg_list, cra_list) {
 		int exact, fuzzy;
 
+		if (crypto_is_moribund(q))
+			continue;
+
 		if ((q->cra_flags ^ type) & mask)
 			continue;
 
@@ -111,7 +110,7 @@ static struct crypto_alg *crypto_larval_alloc(const char *name, u32 type,
 
 	larval = kzalloc(sizeof(*larval), GFP_KERNEL);
 	if (!larval)
-		return NULL;
+		return ERR_PTR(-ENOMEM);
 
 	larval->mask = mask;
 	larval->alg.cra_flags = CRYPTO_ALG_LARVAL | type;
@@ -153,8 +152,11 @@ static struct crypto_alg *crypto_larval_wait(struct crypto_alg *alg)
 
 	wait_for_completion_interruptible_timeout(&larval->completion, 60 * HZ);
 	alg = larval->adult;
-	if (alg && !crypto_mod_get(alg))
-		alg = NULL;
+	if (alg) {
+		if (!crypto_mod_get(alg))
+			alg = ERR_PTR(-EAGAIN);
+	} else
+		alg = ERR_PTR(-ENOENT);
 	crypto_mod_put(&larval->alg);
 
 	return alg;
@@ -165,9 +167,6 @@ static struct crypto_alg *crypto_alg_lookup(const char *name, u32 type,
 {
 	struct crypto_alg *alg;
 
-	if (!name)
-		return NULL;
-
 	down_read(&crypto_alg_sem);
 	alg = __crypto_alg_lookup(name, type, mask);
 	up_read(&crypto_alg_sem);
@@ -181,7 +180,10 @@ struct crypto_alg *crypto_alg_mod_lookup(const char *name, u32 type, u32 mask)
 	struct crypto_alg *larval;
 	int ok;
 
-	mask &= ~CRYPTO_ALG_LARVAL;
+	if (!name)
+		return ERR_PTR(-ENOENT);
+
+	mask &= ~(CRYPTO_ALG_LARVAL | CRYPTO_ALG_DEAD);
 	type &= mask;
 
 	alg = try_then_request_module(crypto_alg_lookup(name, type, mask),
@@ -190,7 +192,7 @@ struct crypto_alg *crypto_alg_mod_lookup(const char *name, u32 type, u32 mask)
 		return crypto_is_larval(alg) ? crypto_larval_wait(alg) : alg;
 
 	larval = crypto_larval_alloc(name, type, mask);
-	if (!larval || !crypto_is_larval(larval))
+	if (IS_ERR(larval) || !crypto_is_larval(larval))
 		return larval;
 
 	ok = crypto_notify(CRYPTO_MSG_ALG_REQUEST, larval);
@@ -203,7 +205,7 @@ struct crypto_alg *crypto_alg_mod_lookup(const char *name, u32 type, u32 mask)
 		alg = crypto_larval_wait(larval);
 	else {
 		crypto_mod_put(larval);
-		alg = NULL;
+		alg = ERR_PTR(-ENOENT);
 	}
 	crypto_larval_kill(larval);
 	return alg;
@@ -298,31 +300,40 @@ static unsigned int crypto_ctxsize(struct crypto_alg *alg, int flags)
 	return len + (alg->cra_alignmask & ~(crypto_tfm_ctx_alignment() - 1));
 }
 
-struct crypto_tfm *crypto_alloc_tfm(const char *name, u32 flags)
+void crypto_shoot_alg(struct crypto_alg *alg)
+{
+	down_write(&crypto_alg_sem);
+	alg->cra_flags |= CRYPTO_ALG_DYING;
+	up_write(&crypto_alg_sem);
+}
+EXPORT_SYMBOL_GPL(crypto_shoot_alg);
+
+struct crypto_tfm *__crypto_alloc_tfm(struct crypto_alg *alg, u32 flags)
 {
 	struct crypto_tfm *tfm = NULL;
-	struct crypto_alg *alg;
 	unsigned int tfm_size;
-
-	alg = crypto_alg_mod_lookup(name, 0, 0);
-	if (alg == NULL)
-		goto out;
+	int err = -ENOMEM;
 
 	tfm_size = sizeof(*tfm) + crypto_ctxsize(alg, flags);
 	tfm = kzalloc(tfm_size, GFP_KERNEL);
 	if (tfm == NULL)
-		goto out_put;
+		goto out;
 
 	tfm->__crt_alg = alg;
-	
-	if (crypto_init_flags(tfm, flags))
+
+	err = crypto_init_flags(tfm, flags);
+	if (err)
 		goto out_free_tfm;
 		
-	if (crypto_init_ops(tfm))
+	err = crypto_init_ops(tfm);
+	if (err)
 		goto out_free_tfm;
 
-	if (alg->cra_init && alg->cra_init(tfm))
+	if (alg->cra_init && (err = alg->cra_init(tfm))) {
+		if (err == -EAGAIN)
+			crypto_shoot_alg(alg);
 		goto cra_init_failed;
+	}
 
 	goto out;
 
@@ -330,12 +341,36 @@ cra_init_failed:
 	crypto_exit_ops(tfm);
 out_free_tfm:
 	kfree(tfm);
-	tfm = NULL;
-out_put:
-	crypto_mod_put(alg);
+	tfm = ERR_PTR(err);
 out:
 	return tfm;
 }
+EXPORT_SYMBOL_GPL(__crypto_alloc_tfm);
+
+struct crypto_tfm *crypto_alloc_tfm(const char *name, u32 flags)
+{
+	struct crypto_tfm *tfm = NULL;
+	int err;
+
+	do {
+		struct crypto_alg *alg;
+
+		alg = crypto_alg_mod_lookup(name, 0, 0);
+		err = PTR_ERR(alg);
+		if (IS_ERR(alg))
+			continue;
+
+		tfm = __crypto_alloc_tfm(alg, flags);
+		err = 0;
+		if (IS_ERR(tfm)) {
+			crypto_mod_put(alg);
+			err = PTR_ERR(tfm);
+			tfm = NULL;
+		}
+	} while (err == -EAGAIN && !signal_pending(current));
+
+	return tfm;
+}
 
 void crypto_free_tfm(struct crypto_tfm *tfm)
 {
@@ -361,7 +396,7 @@ int crypto_alg_available(const char *name, u32 flags)
 	int ret = 0;
 	struct crypto_alg *alg = crypto_alg_mod_lookup(name, 0, 0);
 	
-	if (alg) {
+	if (!IS_ERR(alg)) {
 		crypto_mod_put(alg);
 		ret = 1;
 	}
diff --git a/crypto/cryptomgr.c b/crypto/cryptomgr.c
index ae54942e3b31..9b5b15601068 100644
--- a/crypto/cryptomgr.c
+++ b/crypto/cryptomgr.c
@@ -17,6 +17,7 @@
 #include <linux/module.h>
 #include <linux/notifier.h>
 #include <linux/rtnetlink.h>
+#include <linux/sched.h>
 #include <linux/string.h>
 #include <linux/workqueue.h>
 
@@ -44,21 +45,25 @@ static void cryptomgr_probe(void *data)
 	struct cryptomgr_param *param = data;
 	struct crypto_template *tmpl;
 	struct crypto_instance *inst;
+	int err;
 
 	tmpl = crypto_lookup_template(param->template);
 	if (!tmpl)
 		goto err;
 
-	inst = tmpl->alloc(&param->alg, sizeof(param->alg));
-	if (IS_ERR(inst))
-		goto err;
-	else if ((err = crypto_register_instance(tmpl, inst))) {
-		tmpl->free(inst);
-		goto err;
-	}
+	do {
+		inst = tmpl->alloc(&param->alg, sizeof(param->alg));
+		if (IS_ERR(inst))
+			err = PTR_ERR(inst);
+		else if ((err = crypto_register_instance(tmpl, inst)))
+			tmpl->free(inst);
+	} while (err == -EAGAIN && !signal_pending(current));
 
 	crypto_tmpl_put(tmpl);
 
+	if (err)
+		goto err;
+
 out:
 	kfree(param);
 	return;
diff --git a/crypto/internal.h b/crypto/internal.h
index c08d93bdadc4..03c00b0e6b60 100644
--- a/crypto/internal.h
+++ b/crypto/internal.h
@@ -142,12 +142,21 @@ void crypto_exit_compress_ops(struct crypto_tfm *tfm);
 
 void crypto_larval_error(const char *name, u32 type, u32 mask);
 
+void crypto_shoot_alg(struct crypto_alg *alg);
+struct crypto_tfm *__crypto_alloc_tfm(struct crypto_alg *alg, u32 flags);
+
 int crypto_register_instance(struct crypto_template *tmpl,
 			     struct crypto_instance *inst);
 
 int crypto_register_notifier(struct notifier_block *nb);
 int crypto_unregister_notifier(struct notifier_block *nb);
 
+static inline void crypto_alg_put(struct crypto_alg *alg)
+{
+	if (atomic_dec_and_test(&alg->cra_refcnt) && alg->cra_destroy)
+		alg->cra_destroy(alg);
+}
+
 static inline int crypto_tmpl_get(struct crypto_template *tmpl)
 {
 	return try_module_get(tmpl->module);
@@ -163,6 +172,16 @@ static inline int crypto_is_larval(struct crypto_alg *alg)
 	return alg->cra_flags & CRYPTO_ALG_LARVAL;
 }
 
+static inline int crypto_is_dead(struct crypto_alg *alg)
+{
+	return alg->cra_flags & CRYPTO_ALG_DEAD;
+}
+
+static inline int crypto_is_moribund(struct crypto_alg *alg)
+{
+	return alg->cra_flags & (CRYPTO_ALG_DEAD | CRYPTO_ALG_DYING);
+}
+
 static inline int crypto_notify(unsigned long val, void *v)
 {
 	return blocking_notifier_call_chain(&crypto_chain, val, v);
diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h
index ffec530d52fb..b20f4bdb23ba 100644
--- a/include/crypto/algapi.h
+++ b/include/crypto/algapi.h
@@ -36,10 +36,21 @@ struct crypto_template {
 	char name[CRYPTO_MAX_ALG_NAME];
 };
 
+struct crypto_spawn {
+	struct list_head list;
+	struct crypto_alg *alg;
+	struct crypto_instance *inst;
+};
+
 int crypto_register_template(struct crypto_template *tmpl);
 void crypto_unregister_template(struct crypto_template *tmpl);
 struct crypto_template *crypto_lookup_template(const char *name);
 
+int crypto_init_spawn(struct crypto_spawn *spawn, struct crypto_alg *alg,
+		      struct crypto_instance *inst);
+void crypto_drop_spawn(struct crypto_spawn *spawn);
+struct crypto_tfm *crypto_spawn_tfm(struct crypto_spawn *spawn);
+
 static inline void *crypto_instance_ctx(struct crypto_instance *inst)
 {
 	return inst->__ctx;
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index 85f73c381913..40a6330abc8d 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -35,6 +35,8 @@
 #define CRYPTO_ALG_TYPE_COMPRESS	0x00000004
 
 #define CRYPTO_ALG_LARVAL		0x00000010
+#define CRYPTO_ALG_DEAD			0x00000020
+#define CRYPTO_ALG_DYING		0x00000040
 
 /*
  * Transform masks and values (for crt_flags).
@@ -145,6 +147,8 @@ struct compress_alg {
 
 struct crypto_alg {
 	struct list_head cra_list;
+	struct list_head cra_users;
+
 	u32 cra_flags;
 	unsigned int cra_blocksize;
 	unsigned int cra_ctxsize;
-- 
cgit v1.2.3


From b14cdd6704c96474ba5c74b5959487beaa5ee1cd Mon Sep 17 00:00:00 2001
From: Michal Ludvig <michal@logix.cz>
Date: Sun, 9 Jul 2006 09:02:24 +1000
Subject: [CRYPTO] api: Add missing accessors for new crypto_alg fields

Add missing accessors for cra_driver_name and cra_priority.

Signed-off-by: Michal Ludvig <michal@logix.cz>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 include/linux/crypto.h | 10 ++++++++++
 1 file changed, 10 insertions(+)

(limited to 'include')

diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index 40a6330abc8d..d6e184c876b5 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -297,6 +297,16 @@ static inline const char *crypto_tfm_alg_name(struct crypto_tfm *tfm)
 	return tfm->__crt_alg->cra_name;
 }
 
+static inline const char *crypto_tfm_alg_driver_name(struct crypto_tfm *tfm)
+{
+	return tfm->__crt_alg->cra_driver_name;
+}
+
+static inline int crypto_tfm_alg_priority(struct crypto_tfm *tfm)
+{
+	return tfm->__crt_alg->cra_priority;
+}
+
 static inline const char *crypto_tfm_alg_modname(struct crypto_tfm *tfm)
 {
 	return module_name(tfm->__crt_alg->cra_module);
-- 
cgit v1.2.3


From 560c06ae1ab7c677002ea3b6ac83521bf12ee07d Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sun, 13 Aug 2006 14:16:39 +1000
Subject: [CRYPTO] api: Get rid of flags argument to setkey

Now that the tfm is passed directly to setkey instead of the ctx, we no
longer need to pass the &tfm->crt_flags pointer.

This patch also gets rid of a few unnecessary checks on the key length
for ciphers as the cipher layer guarantees that the key length is within
the bounds specified by the algorithm.

Rather than testing dia_setkey every time, this patch does it only once
during crypto_alloc_tfm.  The redundant check from crypto_digest_setkey
is also removed.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 arch/i386/crypto/aes.c       |  3 ++-
 arch/s390/crypto/aes_s390.c  |  3 ++-
 arch/s390/crypto/des_s390.c  | 13 ++++++++-----
 arch/x86_64/crypto/aes.c     |  5 +++--
 crypto/aes.c                 |  5 +++--
 crypto/anubis.c              |  3 ++-
 crypto/arc4.c                |  2 +-
 crypto/blowfish.c            |  3 +--
 crypto/cast5.c               |  8 +-------
 crypto/cast6.c               |  5 +++--
 crypto/cipher.c              |  4 ++--
 crypto/crc32c.c              |  5 ++---
 crypto/crypto_null.c         |  2 +-
 crypto/des.c                 |  6 ++++--
 crypto/digest.c              | 15 ++++++++++-----
 crypto/khazad.c              |  8 +-------
 crypto/michael_mic.c         |  5 ++---
 crypto/serpent.c             | 19 +++----------------
 crypto/tcrypt.c              |  5 +----
 crypto/tea.c                 | 16 ++--------------
 crypto/twofish_common.c      |  6 +++---
 drivers/crypto/padlock-aes.c |  5 +++--
 include/crypto/twofish.h     |  3 +--
 include/linux/crypto.h       |  6 ++----
 24 files changed, 63 insertions(+), 92 deletions(-)

(limited to 'include')

diff --git a/arch/i386/crypto/aes.c b/arch/i386/crypto/aes.c
index d3806daa3de3..49aad9397f10 100644
--- a/arch/i386/crypto/aes.c
+++ b/arch/i386/crypto/aes.c
@@ -379,12 +379,13 @@ static void gen_tabs(void)
 }
 
 static int aes_set_key(struct crypto_tfm *tfm, const u8 *in_key,
-		       unsigned int key_len, u32 *flags)
+		       unsigned int key_len)
 {
 	int i;
 	u32 ss[8];
 	struct aes_ctx *ctx = crypto_tfm_ctx(tfm);
 	const __le32 *key = (const __le32 *)in_key;
+	u32 *flags = &tfm->crt_flags;
 
 	/* encryption schedule */
 	
diff --git a/arch/s390/crypto/aes_s390.c b/arch/s390/crypto/aes_s390.c
index 5713c7e5bd16..c7c43c9de0d9 100644
--- a/arch/s390/crypto/aes_s390.c
+++ b/arch/s390/crypto/aes_s390.c
@@ -38,9 +38,10 @@ struct s390_aes_ctx {
 };
 
 static int aes_set_key(struct crypto_tfm *tfm, const u8 *in_key,
-		       unsigned int key_len, u32 *flags)
+		       unsigned int key_len)
 {
 	struct s390_aes_ctx *sctx = crypto_tfm_ctx(tfm);
+	u32 *flags = &tfm->crt_flags;
 
 	switch (key_len) {
 	case 16:
diff --git a/arch/s390/crypto/des_s390.c b/arch/s390/crypto/des_s390.c
index b3f7496a79b4..170757b3451d 100644
--- a/arch/s390/crypto/des_s390.c
+++ b/arch/s390/crypto/des_s390.c
@@ -45,9 +45,10 @@ struct crypt_s390_des3_192_ctx {
 };
 
 static int des_setkey(struct crypto_tfm *tfm, const u8 *key,
-		      unsigned int keylen, u32 *flags)
+		      unsigned int keylen)
 {
 	struct crypt_s390_des_ctx *dctx = crypto_tfm_ctx(tfm);
+	u32 *flags = &tfm->crt_flags;
 	int ret;
 
 	/* test if key is valid (not a weak key) */
@@ -167,11 +168,12 @@ static struct crypto_alg des_alg = {
  *
  */
 static int des3_128_setkey(struct crypto_tfm *tfm, const u8 *key,
-			   unsigned int keylen, u32 *flags)
+			   unsigned int keylen)
 {
 	int i, ret;
 	struct crypt_s390_des3_128_ctx *dctx = crypto_tfm_ctx(tfm);
-	const u8* temp_key = key;
+	const u8 *temp_key = key;
+	u32 *flags = &tfm->crt_flags;
 
 	if (!(memcmp(key, &key[DES_KEY_SIZE], DES_KEY_SIZE))) {
 		*flags |= CRYPTO_TFM_RES_BAD_KEY_SCHED;
@@ -303,11 +305,12 @@ static struct crypto_alg des3_128_alg = {
  *
  */
 static int des3_192_setkey(struct crypto_tfm *tfm, const u8 *key,
-			   unsigned int keylen, u32 *flags)
+			   unsigned int keylen)
 {
 	int i, ret;
 	struct crypt_s390_des3_192_ctx *dctx = crypto_tfm_ctx(tfm);
-	const u8* temp_key = key;
+	const u8 *temp_key = key;
+	u32 *flags = &tfm->crt_flags;
 
 	if (!(memcmp(key, &key[DES_KEY_SIZE], DES_KEY_SIZE) &&
 	    memcmp(&key[DES_KEY_SIZE], &key[DES_KEY_SIZE * 2],
diff --git a/arch/x86_64/crypto/aes.c b/arch/x86_64/crypto/aes.c
index 68866fab37aa..5cdb13ea5cc2 100644
--- a/arch/x86_64/crypto/aes.c
+++ b/arch/x86_64/crypto/aes.c
@@ -228,13 +228,14 @@ static void __init gen_tabs(void)
 }
 
 static int aes_set_key(struct crypto_tfm *tfm, const u8 *in_key,
-		       unsigned int key_len, u32 *flags)
+		       unsigned int key_len)
 {
 	struct aes_ctx *ctx = crypto_tfm_ctx(tfm);
 	const __le32 *key = (const __le32 *)in_key;
+	u32 *flags = &tfm->crt_flags;
 	u32 i, j, t, u, v, w;
 
-	if (key_len != 16 && key_len != 24 && key_len != 32) {
+	if (key_len % 8) {
 		*flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
 		return -EINVAL;
 	}
diff --git a/crypto/aes.c b/crypto/aes.c
index a038711831e7..e2440773878c 100644
--- a/crypto/aes.c
+++ b/crypto/aes.c
@@ -249,13 +249,14 @@ gen_tabs (void)
 }
 
 static int aes_set_key(struct crypto_tfm *tfm, const u8 *in_key,
-		       unsigned int key_len, u32 *flags)
+		       unsigned int key_len)
 {
 	struct aes_ctx *ctx = crypto_tfm_ctx(tfm);
 	const __le32 *key = (const __le32 *)in_key;
+	u32 *flags = &tfm->crt_flags;
 	u32 i, t, u, v, w;
 
-	if (key_len != 16 && key_len != 24 && key_len != 32) {
+	if (key_len % 8) {
 		*flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
 		return -EINVAL;
 	}
diff --git a/crypto/anubis.c b/crypto/anubis.c
index 7e2e1a29800e..1c771f7f4dc5 100644
--- a/crypto/anubis.c
+++ b/crypto/anubis.c
@@ -461,10 +461,11 @@ static const u32 rc[] = {
 };
 
 static int anubis_setkey(struct crypto_tfm *tfm, const u8 *in_key,
-			 unsigned int key_len, u32 *flags)
+			 unsigned int key_len)
 {
 	struct anubis_ctx *ctx = crypto_tfm_ctx(tfm);
 	const __be32 *key = (const __be32 *)in_key;
+	u32 *flags = &tfm->crt_flags;
 	int N, R, i, r;
 	u32 kappa[ANUBIS_MAX_N];
 	u32 inter[ANUBIS_MAX_N];
diff --git a/crypto/arc4.c b/crypto/arc4.c
index 5edc6a65b987..8be47e13a9e3 100644
--- a/crypto/arc4.c
+++ b/crypto/arc4.c
@@ -25,7 +25,7 @@ struct arc4_ctx {
 };
 
 static int arc4_set_key(struct crypto_tfm *tfm, const u8 *in_key,
-			unsigned int key_len, u32 *flags)
+			unsigned int key_len)
 {
 	struct arc4_ctx *ctx = crypto_tfm_ctx(tfm);
 	int i, j = 0, k = 0;
diff --git a/crypto/blowfish.c b/crypto/blowfish.c
index 490265f42b3b..55238c4e37f0 100644
--- a/crypto/blowfish.c
+++ b/crypto/blowfish.c
@@ -399,8 +399,7 @@ static void bf_decrypt(struct crypto_tfm *tfm, u8 *dst, const u8 *src)
 /* 
  * Calculates the blowfish S and P boxes for encryption and decryption.
  */
-static int bf_setkey(struct crypto_tfm *tfm, const u8 *key,
-		     unsigned int keylen, u32 *flags)
+static int bf_setkey(struct crypto_tfm *tfm, const u8 *key, unsigned int keylen)
 {
 	struct bf_ctx *ctx = crypto_tfm_ctx(tfm);
 	u32 *P = ctx->p;
diff --git a/crypto/cast5.c b/crypto/cast5.c
index 08eef58c1d3d..13ea60abc19a 100644
--- a/crypto/cast5.c
+++ b/crypto/cast5.c
@@ -769,8 +769,7 @@ static void key_schedule(u32 * x, u32 * z, u32 * k)
 }
 
 
-static int cast5_setkey(struct crypto_tfm *tfm, const u8 *key,
-			unsigned key_len, u32 *flags)
+static int cast5_setkey(struct crypto_tfm *tfm, const u8 *key, unsigned key_len)
 {
 	struct cast5_ctx *c = crypto_tfm_ctx(tfm);
 	int i;
@@ -778,11 +777,6 @@ static int cast5_setkey(struct crypto_tfm *tfm, const u8 *key,
 	u32 z[4];
 	u32 k[16];
 	__be32 p_key[4];
-	
-	if (key_len < 5 || key_len > 16) {
-		*flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
-		return -EINVAL;
-	}
 
 	c->rr = key_len <= 10 ? 1 : 0;
 
diff --git a/crypto/cast6.c b/crypto/cast6.c
index 08e33bfc3ad1..136ab6dfe8c5 100644
--- a/crypto/cast6.c
+++ b/crypto/cast6.c
@@ -382,14 +382,15 @@ static inline void W(u32 *key, unsigned int i) {
 }
 
 static int cast6_setkey(struct crypto_tfm *tfm, const u8 *in_key,
-			unsigned key_len, u32 *flags)
+			unsigned key_len)
 {
 	int i;
 	u32 key[8];
 	__be32 p_key[8]; /* padded key */
 	struct cast6_ctx *c = crypto_tfm_ctx(tfm);
+	u32 *flags = &tfm->crt_flags;
 
-	if (key_len < 16 || key_len > 32 || key_len % 4 != 0) {
+	if (key_len % 4 != 0) {
 		*flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
 		return -EINVAL;
 	}	
diff --git a/crypto/cipher.c b/crypto/cipher.c
index b899eb97abd7..56406a4a88d4 100644
--- a/crypto/cipher.c
+++ b/crypto/cipher.c
@@ -264,12 +264,12 @@ static int setkey(struct crypto_tfm *tfm, const u8 *key, unsigned int keylen)
 {
 	struct cipher_alg *cia = &tfm->__crt_alg->cra_cipher;
 	
+	tfm->crt_flags &= ~CRYPTO_TFM_RES_MASK;
 	if (keylen < cia->cia_min_keysize || keylen > cia->cia_max_keysize) {
 		tfm->crt_flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
 		return -EINVAL;
 	} else
-		return cia->cia_setkey(tfm, key, keylen,
-		                       &tfm->crt_flags);
+		return cia->cia_setkey(tfm, key, keylen);
 }
 
 static int ecb_encrypt(struct crypto_tfm *tfm,
diff --git a/crypto/crc32c.c b/crypto/crc32c.c
index 91ecd895e957..0fa744392a4c 100644
--- a/crypto/crc32c.c
+++ b/crypto/crc32c.c
@@ -44,13 +44,12 @@ static void chksum_init(struct crypto_tfm *tfm)
  * the seed.
  */
 static int chksum_setkey(struct crypto_tfm *tfm, const u8 *key,
-			 unsigned int keylen, u32 *flags)
+			 unsigned int keylen)
 {
 	struct chksum_ctx *mctx = crypto_tfm_ctx(tfm);
 
 	if (keylen != sizeof(mctx->crc)) {
-		if (flags)
-			*flags = CRYPTO_TFM_RES_BAD_KEY_LEN;
+		tfm->crt_flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
 		return -EINVAL;
 	}
 	mctx->key = le32_to_cpu(*(__le32 *)key);
diff --git a/crypto/crypto_null.c b/crypto/crypto_null.c
index a0d956b52949..24dbb5d8617e 100644
--- a/crypto/crypto_null.c
+++ b/crypto/crypto_null.c
@@ -48,7 +48,7 @@ static void null_final(struct crypto_tfm *tfm, u8 *out)
 { }
 
 static int null_setkey(struct crypto_tfm *tfm, const u8 *key,
-		       unsigned int keylen, u32 *flags)
+		       unsigned int keylen)
 { return 0; }
 
 static void null_crypt(struct crypto_tfm *tfm, u8 *dst, const u8 *src)
diff --git a/crypto/des.c b/crypto/des.c
index a9d3c235a6af..1df3a714fa47 100644
--- a/crypto/des.c
+++ b/crypto/des.c
@@ -784,9 +784,10 @@ static void dkey(u32 *pe, const u8 *k)
 }
 
 static int des_setkey(struct crypto_tfm *tfm, const u8 *key,
-		      unsigned int keylen, u32 *flags)
+		      unsigned int keylen)
 {
 	struct des_ctx *dctx = crypto_tfm_ctx(tfm);
+	u32 *flags = &tfm->crt_flags;
 	u32 tmp[DES_EXPKEY_WORDS];
 	int ret;
 
@@ -864,11 +865,12 @@ static void des_decrypt(struct crypto_tfm *tfm, u8 *dst, const u8 *src)
  *
  */
 static int des3_ede_setkey(struct crypto_tfm *tfm, const u8 *key,
-			   unsigned int keylen, u32 *flags)
+			   unsigned int keylen)
 {
 	const u32 *K = (const u32 *)key;
 	struct des3_ede_ctx *dctx = crypto_tfm_ctx(tfm);
 	u32 *expkey = dctx->expkey;
+	u32 *flags = &tfm->crt_flags;
 
 	if (unlikely(!((K[0] ^ K[2]) | (K[1] ^ K[3])) ||
 		     !((K[2] ^ K[4]) | (K[3] ^ K[5]))))
diff --git a/crypto/digest.c b/crypto/digest.c
index 603006a7bef2..0df7f392a56a 100644
--- a/crypto/digest.c
+++ b/crypto/digest.c
@@ -76,12 +76,16 @@ static void final(struct crypto_tfm *tfm, u8 *out)
 		tfm->__crt_alg->cra_digest.dia_final(tfm, out);
 }
 
+static int nosetkey(struct crypto_tfm *tfm, const u8 *key, unsigned int keylen)
+{
+	tfm->crt_flags &= ~CRYPTO_TFM_RES_MASK;
+	return -ENOSYS;
+}
+
 static int setkey(struct crypto_tfm *tfm, const u8 *key, unsigned int keylen)
 {
-	u32 flags;
-	if (tfm->__crt_alg->cra_digest.dia_setkey == NULL)
-		return -ENOSYS;
-	return tfm->__crt_alg->cra_digest.dia_setkey(tfm, key, keylen, &flags);
+	tfm->crt_flags &= ~CRYPTO_TFM_RES_MASK;
+	return tfm->__crt_alg->cra_digest.dia_setkey(tfm, key, keylen);
 }
 
 static void digest(struct crypto_tfm *tfm,
@@ -100,12 +104,13 @@ int crypto_init_digest_flags(struct crypto_tfm *tfm, u32 flags)
 int crypto_init_digest_ops(struct crypto_tfm *tfm)
 {
 	struct digest_tfm *ops = &tfm->crt_digest;
+	struct digest_alg *dalg = &tfm->__crt_alg->cra_digest;
 	
 	ops->dit_init	= init;
 	ops->dit_update	= update;
 	ops->dit_final	= final;
 	ops->dit_digest	= digest;
-	ops->dit_setkey	= setkey;
+	ops->dit_setkey	= dalg->dia_setkey ? setkey : nosetkey;
 	
 	return crypto_alloc_hmac_block(tfm);
 }
diff --git a/crypto/khazad.c b/crypto/khazad.c
index d4c9d3657b36..9fa24a2dd6ff 100644
--- a/crypto/khazad.c
+++ b/crypto/khazad.c
@@ -755,19 +755,13 @@ static const u64 c[KHAZAD_ROUNDS + 1] = {
 };
 
 static int khazad_setkey(struct crypto_tfm *tfm, const u8 *in_key,
-			 unsigned int key_len, u32 *flags)
+			 unsigned int key_len)
 {
 	struct khazad_ctx *ctx = crypto_tfm_ctx(tfm);
 	const __be32 *key = (const __be32 *)in_key;
 	int r;
 	const u64 *S = T7;
 	u64 K2, K1;
-	
-	if (key_len != 16)
-	{
-		*flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
-		return -EINVAL;
-	}
 
 	/* key is supposed to be 32-bit aligned */
 	K2 = ((u64)be32_to_cpu(key[0]) << 32) | be32_to_cpu(key[1]);
diff --git a/crypto/michael_mic.c b/crypto/michael_mic.c
index d061da21cfda..094397b48849 100644
--- a/crypto/michael_mic.c
+++ b/crypto/michael_mic.c
@@ -123,14 +123,13 @@ static void michael_final(struct crypto_tfm *tfm, u8 *out)
 
 
 static int michael_setkey(struct crypto_tfm *tfm, const u8 *key,
-			  unsigned int keylen, u32 *flags)
+			  unsigned int keylen)
 {
 	struct michael_mic_ctx *mctx = crypto_tfm_ctx(tfm);
 	const __le32 *data = (const __le32 *)key;
 
 	if (keylen != 8) {
-		if (flags)
-			*flags = CRYPTO_TFM_RES_BAD_KEY_LEN;
+		tfm->crt_flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
 		return -EINVAL;
 	}
 
diff --git a/crypto/serpent.c b/crypto/serpent.c
index de60cdddbf4a..465d091cd3ec 100644
--- a/crypto/serpent.c
+++ b/crypto/serpent.c
@@ -216,7 +216,7 @@ struct serpent_ctx {
 
 
 static int serpent_setkey(struct crypto_tfm *tfm, const u8 *key,
-			  unsigned int keylen, u32 *flags)
+			  unsigned int keylen)
 {
 	struct serpent_ctx *ctx = crypto_tfm_ctx(tfm);
 	u32 *k = ctx->expkey;
@@ -224,13 +224,6 @@ static int serpent_setkey(struct crypto_tfm *tfm, const u8 *key,
 	u32 r0,r1,r2,r3,r4;
 	int i;
 
-	if ((keylen < SERPENT_MIN_KEY_SIZE)
-			|| (keylen > SERPENT_MAX_KEY_SIZE))
-	{
-		*flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
-		return -EINVAL;
-	}
-
 	/* Copy key, add padding */
 
 	for (i = 0; i < keylen; ++i)
@@ -497,21 +490,15 @@ static struct crypto_alg serpent_alg = {
 };
 
 static int tnepres_setkey(struct crypto_tfm *tfm, const u8 *key,
-			  unsigned int keylen, u32 *flags)
+			  unsigned int keylen)
 {
 	u8 rev_key[SERPENT_MAX_KEY_SIZE];
 	int i;
 
-	if ((keylen < SERPENT_MIN_KEY_SIZE)
-	    || (keylen > SERPENT_MAX_KEY_SIZE)) {
-		*flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
-		return -EINVAL;
-	} 
-
 	for (i = 0; i < keylen; ++i)
 		rev_key[keylen - i - 1] = key[i];
  
-	return serpent_setkey(tfm, rev_key, keylen, flags);
+	return serpent_setkey(tfm, rev_key, keylen);
 }
 
 static void tnepres_encrypt(struct crypto_tfm *tfm, u8 *dst, const u8 *src)
diff --git a/crypto/tcrypt.c b/crypto/tcrypt.c
index bed225e83231..606777074671 100644
--- a/crypto/tcrypt.c
+++ b/crypto/tcrypt.c
@@ -118,10 +118,7 @@ static void test_hash(char *algo, struct hash_testvec *template,
 		sg_set_buf(&sg[0], hash_tv[i].plaintext, hash_tv[i].psize);
 
 		crypto_digest_init(tfm);
-		if (tfm->crt_u.digest.dit_setkey) {
-			crypto_digest_setkey(tfm, hash_tv[i].key,
-					     hash_tv[i].ksize);
-		}
+		crypto_digest_setkey(tfm, hash_tv[i].key, hash_tv[i].ksize);
 		crypto_digest_update(tfm, sg, 1);
 		crypto_digest_final(tfm, result);
 
diff --git a/crypto/tea.c b/crypto/tea.c
index 5367adc82fc9..1c54e26fa529 100644
--- a/crypto/tea.c
+++ b/crypto/tea.c
@@ -46,16 +46,10 @@ struct xtea_ctx {
 };
 
 static int tea_setkey(struct crypto_tfm *tfm, const u8 *in_key,
-		      unsigned int key_len, u32 *flags)
+		      unsigned int key_len)
 {
 	struct tea_ctx *ctx = crypto_tfm_ctx(tfm);
 	const __le32 *key = (const __le32 *)in_key;
-	
-	if (key_len != 16)
-	{
-		*flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
-		return -EINVAL;
-	}
 
 	ctx->KEY[0] = le32_to_cpu(key[0]);
 	ctx->KEY[1] = le32_to_cpu(key[1]);
@@ -125,16 +119,10 @@ static void tea_decrypt(struct crypto_tfm *tfm, u8 *dst, const u8 *src)
 }
 
 static int xtea_setkey(struct crypto_tfm *tfm, const u8 *in_key,
-		       unsigned int key_len, u32 *flags)
+		       unsigned int key_len)
 {
 	struct xtea_ctx *ctx = crypto_tfm_ctx(tfm);
 	const __le32 *key = (const __le32 *)in_key;
-	
-	if (key_len != 16)
-	{
-		*flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
-		return -EINVAL;
-	}
 
 	ctx->KEY[0] = le32_to_cpu(key[0]);
 	ctx->KEY[1] = le32_to_cpu(key[1]);
diff --git a/crypto/twofish_common.c b/crypto/twofish_common.c
index 1ae0280c2513..b4b9c0c3f4ae 100644
--- a/crypto/twofish_common.c
+++ b/crypto/twofish_common.c
@@ -580,11 +580,11 @@ static const u8 calc_sb_tbl[512] = {
    ctx->a[(j) + 1] = rol32(y, 9)
 
 /* Perform the key setup. */
-int twofish_setkey(struct crypto_tfm *tfm, const u8 *key,
-		   unsigned int key_len, u32 *flags)
+int twofish_setkey(struct crypto_tfm *tfm, const u8 *key, unsigned int key_len)
 {
 
 	struct twofish_ctx *ctx = crypto_tfm_ctx(tfm);
+	u32 *flags = &tfm->crt_flags;
 
 	int i, j, k;
 
@@ -600,7 +600,7 @@ int twofish_setkey(struct crypto_tfm *tfm, const u8 *key,
 	u8 tmp;
 
 	/* Check key length. */
-	if (key_len != 16 && key_len != 24 && key_len != 32)
+	if (key_len % 8)
 	{
 		*flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
 		return -EINVAL; /* unsupported key length */
diff --git a/drivers/crypto/padlock-aes.c b/drivers/crypto/padlock-aes.c
index 3a2a71108d35..3e683709243e 100644
--- a/drivers/crypto/padlock-aes.c
+++ b/drivers/crypto/padlock-aes.c
@@ -308,15 +308,16 @@ static inline struct aes_ctx *aes_ctx(struct crypto_tfm *tfm)
 }
 
 static int aes_set_key(struct crypto_tfm *tfm, const u8 *in_key,
-		       unsigned int key_len, u32 *flags)
+		       unsigned int key_len)
 {
 	struct aes_ctx *ctx = aes_ctx(tfm);
 	const __le32 *key = (const __le32 *)in_key;
+	u32 *flags = &tfm->crt_flags;
 	uint32_t i, t, u, v, w;
 	uint32_t P[AES_EXTENDED_KEY_SIZE];
 	uint32_t rounds;
 
-	if (key_len != 16 && key_len != 24 && key_len != 32) {
+	if (key_len % 8) {
 		*flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
 		return -EINVAL;
 	}
diff --git a/include/crypto/twofish.h b/include/crypto/twofish.h
index e4328cfaaf64..c408522595c6 100644
--- a/include/crypto/twofish.h
+++ b/include/crypto/twofish.h
@@ -17,7 +17,6 @@ struct twofish_ctx {
 	u32 s[4][256], w[8], k[32];
 };
 
-int twofish_setkey(struct crypto_tfm *tfm, const u8 *key,
-		   unsigned int key_len, u32 *flags);
+int twofish_setkey(struct crypto_tfm *tfm, const u8 *key, unsigned int key_len);
 
 #endif
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index d6e184c876b5..053bfab43e8d 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -106,7 +106,7 @@ struct cipher_alg {
 	unsigned int cia_min_keysize;
 	unsigned int cia_max_keysize;
 	int (*cia_setkey)(struct crypto_tfm *tfm, const u8 *key,
-	                  unsigned int keylen, u32 *flags);
+	                  unsigned int keylen);
 	void (*cia_encrypt)(struct crypto_tfm *tfm, u8 *dst, const u8 *src);
 	void (*cia_decrypt)(struct crypto_tfm *tfm, u8 *dst, const u8 *src);
 
@@ -131,7 +131,7 @@ struct digest_alg {
 			   unsigned int len);
 	void (*dia_final)(struct crypto_tfm *tfm, u8 *out);
 	int (*dia_setkey)(struct crypto_tfm *tfm, const u8 *key,
-	                  unsigned int keylen, u32 *flags);
+	                  unsigned int keylen);
 };
 
 struct compress_alg {
@@ -397,8 +397,6 @@ static inline int crypto_digest_setkey(struct crypto_tfm *tfm,
                                        const u8 *key, unsigned int keylen)
 {
 	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_DIGEST);
-	if (tfm->crt_digest.dit_setkey == NULL)
-		return -ENOSYS;
 	return tfm->crt_digest.dit_setkey(tfm, key, keylen);
 }
 
-- 
cgit v1.2.3


From df89820ebd5bbf4f3c6b5f8ee7d9e983107f6a91 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Fri, 14 Jul 2006 10:42:27 +1000
Subject: [CRYPTO] cipher: Removed special IV checks for ECB

This patch makes IV operations on ECB fail through nocrypt_iv rather than
calling BUG().  This is needed to generalise CBC/ECB using the template
mechanism.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 crypto/cipher.c        | 2 ++
 include/linux/crypto.h | 2 --
 2 files changed, 2 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/crypto/cipher.c b/crypto/cipher.c
index 56406a4a88d4..aebc4a2adc80 100644
--- a/crypto/cipher.c
+++ b/crypto/cipher.c
@@ -399,6 +399,8 @@ int crypto_init_cipher_ops(struct crypto_tfm *tfm)
 	case CRYPTO_TFM_MODE_ECB:
 		ops->cit_encrypt = ecb_encrypt;
 		ops->cit_decrypt = ecb_decrypt;
+		ops->cit_encrypt_iv = nocrypt_iv;
+		ops->cit_decrypt_iv = nocrypt_iv;
 		break;
 		
 	case CRYPTO_TFM_MODE_CBC:
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index 053bfab43e8d..dbdfc7c79367 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -422,7 +422,6 @@ static inline int crypto_cipher_encrypt_iv(struct crypto_tfm *tfm,
                                            unsigned int nbytes, u8 *iv)
 {
 	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_CIPHER);
-	BUG_ON(tfm->crt_cipher.cit_mode == CRYPTO_TFM_MODE_ECB);
 	return tfm->crt_cipher.cit_encrypt_iv(tfm, dst, src, nbytes, iv);
 }                                        
 
@@ -441,7 +440,6 @@ static inline int crypto_cipher_decrypt_iv(struct crypto_tfm *tfm,
                                            unsigned int nbytes, u8 *iv)
 {
 	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_CIPHER);
-	BUG_ON(tfm->crt_cipher.cit_mode == CRYPTO_TFM_MODE_ECB);
 	return tfm->crt_cipher.cit_decrypt_iv(tfm, dst, src, nbytes, iv);
 }
 
-- 
cgit v1.2.3


From 7fed0bf271b374be4c98a5880faed4b1128e78e9 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sun, 6 Aug 2006 23:10:45 +1000
Subject: [CRYPTO] api: Add common instance initialisation code

This patch adds the helpers crypto_get_attr_alg and crypto_alloc_instance
which can be used by simple one-argument templates like hmac to process
input parameters and allocate instances.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 crypto/algapi.c         | 53 +++++++++++++++++++++++++++++++++++++++++++++++++
 include/crypto/algapi.h |  5 +++++
 2 files changed, 58 insertions(+)

(limited to 'include')

diff --git a/crypto/algapi.c b/crypto/algapi.c
index 36c4f1bdb521..c91530021e9c 100644
--- a/crypto/algapi.c
+++ b/crypto/algapi.c
@@ -16,6 +16,7 @@
 #include <linux/kernel.h>
 #include <linux/list.h>
 #include <linux/module.h>
+#include <linux/rtnetlink.h>
 #include <linux/string.h>
 
 #include "internal.h"
@@ -415,6 +416,58 @@ int crypto_unregister_notifier(struct notifier_block *nb)
 }
 EXPORT_SYMBOL_GPL(crypto_unregister_notifier);
 
+struct crypto_alg *crypto_get_attr_alg(void *param, unsigned int len,
+				       u32 type, u32 mask)
+{
+	struct rtattr *rta = param;
+	struct crypto_attr_alg *alga;
+
+	if (!RTA_OK(rta, len))
+		return ERR_PTR(-EBADR);
+	if (rta->rta_type != CRYPTOA_ALG || RTA_PAYLOAD(rta) < sizeof(*alga))
+		return ERR_PTR(-EINVAL);
+
+	alga = RTA_DATA(rta);
+	alga->name[CRYPTO_MAX_ALG_NAME - 1] = 0;
+
+	return crypto_alg_mod_lookup(alga->name, type, mask);
+}
+EXPORT_SYMBOL_GPL(crypto_get_attr_alg);
+
+struct crypto_instance *crypto_alloc_instance(const char *name,
+					      struct crypto_alg *alg)
+{
+	struct crypto_instance *inst;
+	struct crypto_spawn *spawn;
+	int err;
+
+	inst = kzalloc(sizeof(*inst) + sizeof(*spawn), GFP_KERNEL);
+	if (!inst)
+		return ERR_PTR(-ENOMEM);
+
+	err = -ENAMETOOLONG;
+	if (snprintf(inst->alg.cra_name, CRYPTO_MAX_ALG_NAME, "%s(%s)", name,
+		     alg->cra_name) >= CRYPTO_MAX_ALG_NAME)
+		goto err_free_inst;
+
+	if (snprintf(inst->alg.cra_driver_name, CRYPTO_MAX_ALG_NAME, "%s(%s)",
+		     name, alg->cra_driver_name) >= CRYPTO_MAX_ALG_NAME)
+		goto err_free_inst;
+
+	spawn = crypto_instance_ctx(inst);
+	err = crypto_init_spawn(spawn, alg, inst);
+
+	if (err)
+		goto err_free_inst;
+
+	return inst;
+
+err_free_inst:
+	kfree(inst);
+	return ERR_PTR(err);
+}
+EXPORT_SYMBOL_GPL(crypto_alloc_instance);
+
 static int __init crypto_algapi_init(void)
 {
 	crypto_init_proc();
diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h
index b20f4bdb23ba..1a598f829417 100644
--- a/include/crypto/algapi.h
+++ b/include/crypto/algapi.h
@@ -51,6 +51,11 @@ int crypto_init_spawn(struct crypto_spawn *spawn, struct crypto_alg *alg,
 void crypto_drop_spawn(struct crypto_spawn *spawn);
 struct crypto_tfm *crypto_spawn_tfm(struct crypto_spawn *spawn);
 
+struct crypto_alg *crypto_get_attr_alg(void *param, unsigned int len,
+				       u32 type, u32 mask);
+struct crypto_instance *crypto_alloc_instance(const char *name,
+					      struct crypto_alg *alg);
+
 static inline void *crypto_instance_ctx(struct crypto_instance *inst)
 {
 	return inst->__ctx;
-- 
cgit v1.2.3


From f3f632d61ae9af85d436706ee8e33af1a7fb9c28 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sun, 6 Aug 2006 23:12:59 +1000
Subject: [CRYPTO] api: Added asynchronous flag

This patch adds the asynchronous flag and changes all existing users to
only look up algorithms that are synchronous.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 crypto/api.c           | 5 +++--
 include/linux/crypto.h | 1 +
 2 files changed, 4 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/crypto/api.c b/crypto/api.c
index 7e5522cf856e..1e4692a13474 100644
--- a/crypto/api.c
+++ b/crypto/api.c
@@ -355,7 +355,7 @@ struct crypto_tfm *crypto_alloc_tfm(const char *name, u32 flags)
 	do {
 		struct crypto_alg *alg;
 
-		alg = crypto_alg_mod_lookup(name, 0, 0);
+		alg = crypto_alg_mod_lookup(name, 0, CRYPTO_ALG_ASYNC);
 		err = PTR_ERR(alg);
 		if (IS_ERR(alg))
 			continue;
@@ -394,7 +394,8 @@ void crypto_free_tfm(struct crypto_tfm *tfm)
 int crypto_alg_available(const char *name, u32 flags)
 {
 	int ret = 0;
-	struct crypto_alg *alg = crypto_alg_mod_lookup(name, 0, 0);
+	struct crypto_alg *alg = crypto_alg_mod_lookup(name, 0,
+						       CRYPTO_ALG_ASYNC);
 	
 	if (!IS_ERR(alg)) {
 		crypto_mod_put(alg);
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index dbdfc7c79367..530dc4bf363c 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -37,6 +37,7 @@
 #define CRYPTO_ALG_LARVAL		0x00000010
 #define CRYPTO_ALG_DEAD			0x00000020
 #define CRYPTO_ALG_DYING		0x00000040
+#define CRYPTO_ALG_ASYNC		0x00000080
 
 /*
  * Transform masks and values (for crt_flags).
-- 
cgit v1.2.3


From 6d7d684d635ac5a345f075015f2c84169c111c6a Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sun, 30 Jul 2006 11:53:01 +1000
Subject: [CRYPTO] api: Added crypto_alloc_base

Up until now all crypto transforms have been of the same type, struct
crypto_tfm, regardless of whether they are ciphers, digests, or other
types.  As a result of that, we check the types at run-time before
each crypto operation.

This is rather cumbersome.  We could instead use different C types for
each crypto type to ensure that the correct types are used at compile
time.  That is, we would have crypto_cipher/crypto_digest instead of
just crypto_tfm.  The appropriate type would then be required for the
actual operations such as crypto_digest_digest.

Now that we have the type/mask fields when looking up algorithms, it
is easy to request for an algorithm of the precise type that the user
wants.  However, crypto_alloc_tfm currently does not expose these new
attributes.

This patch introduces the function crypto_alloc_base which will carry
these new parameters.  It will be renamed to crypto_alloc_tfm once
all existing users have been converted.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 crypto/api.c           | 60 ++++++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/crypto.h | 14 +++---------
 2 files changed, 63 insertions(+), 11 deletions(-)

(limited to 'include')

diff --git a/crypto/api.c b/crypto/api.c
index 1e4692a13474..bc4b7901acdf 100644
--- a/crypto/api.c
+++ b/crypto/api.c
@@ -372,6 +372,66 @@ struct crypto_tfm *crypto_alloc_tfm(const char *name, u32 flags)
 	return tfm;
 }
 
+/*
+ *	crypto_alloc_base - Locate algorithm and allocate transform
+ *	@alg_name: Name of algorithm
+ *	@type: Type of algorithm
+ *	@mask: Mask for type comparison
+ *
+ *	crypto_alloc_base() will first attempt to locate an already loaded
+ *	algorithm.  If that fails and the kernel supports dynamically loadable
+ *	modules, it will then attempt to load a module of the same name or
+ *	alias.  If that fails it will send a query to any loaded crypto manager
+ *	to construct an algorithm on the fly.  A refcount is grabbed on the
+ *	algorithm which is then associated with the new transform.
+ *
+ *	The returned transform is of a non-determinate type.  Most people
+ *	should use one of the more specific allocation functions such as
+ *	crypto_alloc_blkcipher.
+ *
+ *	In case of error the return value is an error pointer.
+ */
+struct crypto_tfm *crypto_alloc_base(const char *alg_name, u32 type, u32 mask)
+{
+	struct crypto_tfm *tfm;
+	int err;
+
+	for (;;) {
+		struct crypto_alg *alg;
+
+		alg = crypto_alg_mod_lookup(alg_name, type, mask);
+		err = PTR_ERR(alg);
+		tfm = ERR_PTR(err);
+		if (IS_ERR(alg))
+			goto err;
+
+		tfm = __crypto_alloc_tfm(alg, 0);
+		if (!IS_ERR(tfm))
+			break;
+
+		crypto_mod_put(alg);
+		err = PTR_ERR(tfm);
+
+err:
+		if (err != -EAGAIN)
+			break;
+		if (signal_pending(current)) {
+			err = -EINTR;
+			break;
+		}
+	};
+
+	return tfm;
+}
+EXPORT_SYMBOL_GPL(crypto_alloc_base);
+ 
+/*
+ *	crypto_free_tfm - Free crypto transform
+ *	@tfm: Transform to free
+ *
+ *	crypto_free_tfm() frees up the transform and any associated resources,
+ *	then drops the refcount on the associated algorithm.
+ */
 void crypto_free_tfm(struct crypto_tfm *tfm)
 {
 	struct crypto_alg *alg;
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index 530dc4bf363c..6847ab0ea30e 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -194,8 +194,8 @@ static inline int crypto_alg_available(const char *name, u32 flags)
 
 /*
  * Transforms: user-instantiated objects which encapsulate algorithms
- * and core processing logic.  Managed via crypto_alloc_tfm() and
- * crypto_free_tfm(), as well as the various helpers below.
+ * and core processing logic.  Managed via crypto_alloc_*() and
+ * crypto_free_*(), as well as the various helpers below.
  */
 
 struct cipher_tfm {
@@ -278,16 +278,8 @@ struct crypto_attr_alg {
  * Transform user interface.
  */
  
-/*
- * crypto_alloc_tfm() will first attempt to locate an already loaded algorithm.
- * If that fails and the kernel supports dynamically loadable modules, it
- * will then attempt to load a module of the same name or alias.  A refcount
- * is grabbed on the algorithm which is then associated with the new transform.
- *
- * crypto_free_tfm() frees up the transform and any associated resources,
- * then drops the refcount on the associated algorithm.
- */
 struct crypto_tfm *crypto_alloc_tfm(const char *alg_name, u32 tfm_flags);
+struct crypto_tfm *crypto_alloc_base(const char *alg_name, u32 type, u32 mask);
 void crypto_free_tfm(struct crypto_tfm *tfm);
 
 /*
-- 
cgit v1.2.3


From e853c3cfa8cc24869ecd2526e589bcb176bc12e9 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Tue, 22 Aug 2006 00:06:54 +1000
Subject: [CRYPTO] api: Added crypto_type support

This patch adds the crypto_type structure which will be used for all new
crypto algorithm types, beginning with block ciphers.

The primary purpose of this abstraction is to allow different crypto_type
objects for crypto algorithms of the same type, in particular, there will
be a different crypto_type objects for asynchronous algorithms.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 crypto/api.c            | 32 +++++++++++++++++++++++---------
 crypto/proc.c           |  5 ++++-
 include/crypto/algapi.h |  8 ++++++++
 include/linux/crypto.h  |  3 +++
 4 files changed, 38 insertions(+), 10 deletions(-)

(limited to 'include')

diff --git a/crypto/api.c b/crypto/api.c
index bc4b7901acdf..edaa843d8e83 100644
--- a/crypto/api.c
+++ b/crypto/api.c
@@ -226,17 +226,18 @@ static int crypto_init_flags(struct crypto_tfm *tfm, u32 flags)
 		
 	case CRYPTO_ALG_TYPE_COMPRESS:
 		return crypto_init_compress_flags(tfm, flags);
-	
-	default:
-		break;
 	}
 	
-	BUG();
-	return -EINVAL;
+	return 0;
 }
 
 static int crypto_init_ops(struct crypto_tfm *tfm)
 {
+	const struct crypto_type *type = tfm->__crt_alg->cra_type;
+
+	if (type)
+		return type->init(tfm);
+
 	switch (crypto_tfm_alg_type(tfm)) {
 	case CRYPTO_ALG_TYPE_CIPHER:
 		return crypto_init_cipher_ops(tfm);
@@ -257,6 +258,14 @@ static int crypto_init_ops(struct crypto_tfm *tfm)
 
 static void crypto_exit_ops(struct crypto_tfm *tfm)
 {
+	const struct crypto_type *type = tfm->__crt_alg->cra_type;
+
+	if (type) {
+		if (type->exit)
+			type->exit(tfm);
+		return;
+	}
+
 	switch (crypto_tfm_alg_type(tfm)) {
 	case CRYPTO_ALG_TYPE_CIPHER:
 		crypto_exit_cipher_ops(tfm);
@@ -278,26 +287,31 @@ static void crypto_exit_ops(struct crypto_tfm *tfm)
 
 static unsigned int crypto_ctxsize(struct crypto_alg *alg, int flags)
 {
+	const struct crypto_type *type = alg->cra_type;
 	unsigned int len;
 
+	len = alg->cra_alignmask & ~(crypto_tfm_ctx_alignment() - 1);
+	if (type)
+		return len + type->ctxsize(alg);
+
 	switch (alg->cra_flags & CRYPTO_ALG_TYPE_MASK) {
 	default:
 		BUG();
 
 	case CRYPTO_ALG_TYPE_CIPHER:
-		len = crypto_cipher_ctxsize(alg, flags);
+		len += crypto_cipher_ctxsize(alg, flags);
 		break;
 		
 	case CRYPTO_ALG_TYPE_DIGEST:
-		len = crypto_digest_ctxsize(alg, flags);
+		len += crypto_digest_ctxsize(alg, flags);
 		break;
 		
 	case CRYPTO_ALG_TYPE_COMPRESS:
-		len = crypto_compress_ctxsize(alg, flags);
+		len += crypto_compress_ctxsize(alg, flags);
 		break;
 	}
 
-	return len + (alg->cra_alignmask & ~(crypto_tfm_ctx_alignment() - 1));
+	return len;
 }
 
 void crypto_shoot_alg(struct crypto_alg *alg)
diff --git a/crypto/proc.c b/crypto/proc.c
index 9e573b17e887..dabce0676f63 100644
--- a/crypto/proc.c
+++ b/crypto/proc.c
@@ -78,7 +78,10 @@ static int c_show(struct seq_file *m, void *p)
 		seq_printf(m, "type         : compression\n");
 		break;
 	default:
-		seq_printf(m, "type         : unknown\n");
+		if (alg->cra_type && alg->cra_type->show)
+			alg->cra_type->show(m, alg);
+		else
+			seq_printf(m, "type         : unknown\n");
 		break;
 	}
 
diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h
index 1a598f829417..c533c0a291af 100644
--- a/include/crypto/algapi.h
+++ b/include/crypto/algapi.h
@@ -15,6 +15,14 @@
 #include <linux/crypto.h>
 
 struct module;
+struct seq_file;
+
+struct crypto_type {
+	unsigned int (*ctxsize)(struct crypto_alg *alg);
+	int (*init)(struct crypto_tfm *tfm);
+	void (*exit)(struct crypto_tfm *tfm);
+	void (*show)(struct seq_file *m, struct crypto_alg *alg);
+};
 
 struct crypto_instance {
 	struct crypto_alg alg;
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index 6847ab0ea30e..8e9c407b00d2 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -90,6 +90,7 @@
 
 struct scatterlist;
 struct crypto_tfm;
+struct crypto_type;
 
 struct cipher_desc {
 	struct crypto_tfm *tfm;
@@ -161,6 +162,8 @@ struct crypto_alg {
 	char cra_name[CRYPTO_MAX_ALG_NAME];
 	char cra_driver_name[CRYPTO_MAX_ALG_NAME];
 
+	const struct crypto_type *cra_type;
+
 	union {
 		struct cipher_alg cipher;
 		struct digest_alg digest;
-- 
cgit v1.2.3


From f28776a369b12f9a03a822a8e1090ed670a41f4f Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sun, 13 Aug 2006 20:58:18 +1000
Subject: [CRYPTO] cipher: Added encrypt_one/decrypt_one

This patch adds two new operations for the simple cipher that encrypts or
decrypts a single block at a time.  This will be the main interface after
the existing block operations have moved over to the new block ciphers.

It also adds the crypto_cipher type which is currently only used on the
new operations but will be extended to setkey as well once existing users
have been converted to use block ciphers where applicable.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 crypto/cipher.c         | 48 +++++++++++++++++++++++++
 include/crypto/algapi.h |  5 +++
 include/linux/crypto.h  | 96 +++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 149 insertions(+)

(limited to 'include')

diff --git a/crypto/cipher.c b/crypto/cipher.c
index f573c59ed9dc..d8ca0ec8d0be 100644
--- a/crypto/cipher.c
+++ b/crypto/cipher.c
@@ -388,12 +388,60 @@ int crypto_init_cipher_flags(struct crypto_tfm *tfm, u32 flags)
 	return 0;
 }
 
+static void cipher_crypt_unaligned(void (*fn)(struct crypto_tfm *, u8 *,
+					      const u8 *),
+				   struct crypto_tfm *tfm,
+				   u8 *dst, const u8 *src)
+{
+	unsigned long alignmask = crypto_tfm_alg_alignmask(tfm);
+	unsigned int size = crypto_tfm_alg_blocksize(tfm);
+	u8 buffer[size + alignmask];
+	u8 *tmp = (u8 *)ALIGN((unsigned long)buffer, alignmask + 1);
+
+	memcpy(tmp, src, size);
+	fn(tfm, tmp, tmp);
+	memcpy(dst, tmp, size);
+}
+
+static void cipher_encrypt_unaligned(struct crypto_tfm *tfm,
+				     u8 *dst, const u8 *src)
+{
+	unsigned long alignmask = crypto_tfm_alg_alignmask(tfm);
+	struct cipher_alg *cipher = &tfm->__crt_alg->cra_cipher;
+
+	if (unlikely(((unsigned long)dst | (unsigned long)src) & alignmask)) {
+		cipher_crypt_unaligned(cipher->cia_encrypt, tfm, dst, src);
+		return;
+	}
+
+	cipher->cia_encrypt(tfm, dst, src);
+}
+
+static void cipher_decrypt_unaligned(struct crypto_tfm *tfm,
+				     u8 *dst, const u8 *src)
+{
+	unsigned long alignmask = crypto_tfm_alg_alignmask(tfm);
+	struct cipher_alg *cipher = &tfm->__crt_alg->cra_cipher;
+
+	if (unlikely(((unsigned long)dst | (unsigned long)src) & alignmask)) {
+		cipher_crypt_unaligned(cipher->cia_decrypt, tfm, dst, src);
+		return;
+	}
+
+	cipher->cia_decrypt(tfm, dst, src);
+}
+
 int crypto_init_cipher_ops(struct crypto_tfm *tfm)
 {
 	int ret = 0;
 	struct cipher_tfm *ops = &tfm->crt_cipher;
+	struct cipher_alg *cipher = &tfm->__crt_alg->cra_cipher;
 
 	ops->cit_setkey = setkey;
+	ops->cit_encrypt_one = crypto_tfm_alg_alignmask(tfm) ?
+		cipher_encrypt_unaligned : cipher->cia_encrypt;
+	ops->cit_decrypt_one = crypto_tfm_alg_alignmask(tfm) ?
+		cipher_decrypt_unaligned : cipher->cia_decrypt;
 
 	switch (tfm->crt_cipher.cit_mode) {
 	case CRYPTO_TFM_MODE_ECB:
diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h
index c533c0a291af..6f9fb27b2071 100644
--- a/include/crypto/algapi.h
+++ b/include/crypto/algapi.h
@@ -69,5 +69,10 @@ static inline void *crypto_instance_ctx(struct crypto_instance *inst)
 	return inst->__ctx;
 }
 
+static inline struct cipher_alg *crypto_cipher_alg(struct crypto_cipher *tfm)
+{
+	return &crypto_cipher_tfm(tfm)->__crt_alg->cra_cipher;
+}
+
 #endif	/* _CRYPTO_ALGAPI_H */
 
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index 8e9c407b00d2..fdecee83878c 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -224,6 +224,8 @@ struct cipher_tfm {
 			   struct scatterlist *src,
 			   unsigned int nbytes, u8 *iv);
 	void (*cit_xor_block)(u8 *dst, const u8 *src);
+	void (*cit_encrypt_one)(struct crypto_tfm *tfm, u8 *dst, const u8 *src);
+	void (*cit_decrypt_one)(struct crypto_tfm *tfm, u8 *dst, const u8 *src);
 };
 
 struct digest_tfm {
@@ -268,6 +270,8 @@ struct crypto_tfm {
 	void *__crt_ctx[] CRYPTO_MINALIGN_ATTR;
 };
 
+#define crypto_cipher crypto_tfm
+
 enum {
 	CRYPTOA_UNSPEC,
 	CRYPTOA_ALG,
@@ -347,6 +351,21 @@ static inline unsigned int crypto_tfm_alg_alignmask(struct crypto_tfm *tfm)
 	return tfm->__crt_alg->cra_alignmask;
 }
 
+static inline u32 crypto_tfm_get_flags(struct crypto_tfm *tfm)
+{
+	return tfm->crt_flags;
+}
+
+static inline void crypto_tfm_set_flags(struct crypto_tfm *tfm, u32 flags)
+{
+	tfm->crt_flags |= flags;
+}
+
+static inline void crypto_tfm_clear_flags(struct crypto_tfm *tfm, u32 flags)
+{
+	tfm->crt_flags &= ~flags;
+}
+
 static inline void *crypto_tfm_ctx(struct crypto_tfm *tfm)
 {
 	return tfm->__crt_ctx;
@@ -361,6 +380,83 @@ static inline unsigned int crypto_tfm_ctx_alignment(void)
 /*
  * API wrappers.
  */
+static inline struct crypto_cipher *__crypto_cipher_cast(struct crypto_tfm *tfm)
+{
+	return (struct crypto_cipher *)tfm;
+}
+
+static inline struct crypto_cipher *crypto_cipher_cast(struct crypto_tfm *tfm)
+{
+	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_CIPHER);
+	return __crypto_cipher_cast(tfm);
+}
+
+static inline struct crypto_cipher *crypto_alloc_cipher(const char *alg_name,
+							u32 type, u32 mask)
+{
+	type &= ~CRYPTO_ALG_TYPE_MASK;
+	type |= CRYPTO_ALG_TYPE_CIPHER;
+	mask |= CRYPTO_ALG_TYPE_MASK;
+
+	return __crypto_cipher_cast(crypto_alloc_base(alg_name, type, mask));
+}
+
+static inline struct crypto_tfm *crypto_cipher_tfm(struct crypto_cipher *tfm)
+{
+	return tfm;
+}
+
+static inline void crypto_free_cipher(struct crypto_cipher *tfm)
+{
+	crypto_free_tfm(crypto_cipher_tfm(tfm));
+}
+
+static inline struct cipher_tfm *crypto_cipher_crt(struct crypto_cipher *tfm)
+{
+	return &crypto_cipher_tfm(tfm)->crt_cipher;
+}
+
+static inline unsigned int crypto_cipher_blocksize(struct crypto_cipher *tfm)
+{
+	return crypto_tfm_alg_blocksize(crypto_cipher_tfm(tfm));
+}
+
+static inline unsigned int crypto_cipher_alignmask(struct crypto_cipher *tfm)
+{
+	return crypto_tfm_alg_alignmask(crypto_cipher_tfm(tfm));
+}
+
+static inline u32 crypto_cipher_get_flags(struct crypto_cipher *tfm)
+{
+	return crypto_tfm_get_flags(crypto_cipher_tfm(tfm));
+}
+
+static inline void crypto_cipher_set_flags(struct crypto_cipher *tfm,
+					   u32 flags)
+{
+	crypto_tfm_set_flags(crypto_cipher_tfm(tfm), flags);
+}
+
+static inline void crypto_cipher_clear_flags(struct crypto_cipher *tfm,
+					     u32 flags)
+{
+	crypto_tfm_clear_flags(crypto_cipher_tfm(tfm), flags);
+}
+
+static inline void crypto_cipher_encrypt_one(struct crypto_cipher *tfm,
+					     u8 *dst, const u8 *src)
+{
+	crypto_cipher_crt(tfm)->cit_encrypt_one(crypto_cipher_tfm(tfm),
+						dst, src);
+}
+
+static inline void crypto_cipher_decrypt_one(struct crypto_cipher *tfm,
+					     u8 *dst, const u8 *src)
+{
+	crypto_cipher_crt(tfm)->cit_decrypt_one(crypto_cipher_tfm(tfm),
+						dst, src);
+}
+
 static inline void crypto_digest_init(struct crypto_tfm *tfm)
 {
 	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_DIGEST);
-- 
cgit v1.2.3


From 5c64097aa0f6dc4f27718ef47ca9a12538d62860 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sat, 12 Aug 2006 21:56:17 +1000
Subject: [CRYPTO] scatterwalk: Prepare for block ciphers

This patch prepares the scatterwalk code for use by the new block cipher
type.

Firstly it halves the size of scatter_walk on 32-bit platforms.  This
is important as we allocate at least two of these objects on the stack
for each block cipher operation.

It also exports the symbols since the block cipher code can be built as
a module.

Finally there is a hack in scatterwalk_unmap that relies on progress
being made.  Unfortunately, for hardware crypto we can't guarantee
progress to be made since the hardware can fail.

So this also gets rid of the hack by not advancing the address returned
by scatterwalk_map.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 crypto/cipher.c         | 27 ++++++++-------
 crypto/scatterwalk.c    | 89 ++++++++++++++++++++++---------------------------
 crypto/scatterwalk.h    | 48 +++++++++++++++-----------
 include/crypto/algapi.h |  5 +++
 4 files changed, 87 insertions(+), 82 deletions(-)

(limited to 'include')

diff --git a/crypto/cipher.c b/crypto/cipher.c
index d8ca0ec8d0be..326461780673 100644
--- a/crypto/cipher.c
+++ b/crypto/cipher.c
@@ -45,15 +45,10 @@ static unsigned int crypt_slow(const struct cipher_desc *desc,
 	u8 buffer[bsize * 2 + alignmask];
 	u8 *src = (u8 *)ALIGN((unsigned long)buffer, alignmask + 1);
 	u8 *dst = src + bsize;
-	unsigned int n;
-
-	n = scatterwalk_copychunks(src, in, bsize, 0);
-	scatterwalk_advance(in, n);
 
+	scatterwalk_copychunks(src, in, bsize, 0);
 	desc->prfn(desc, dst, src, bsize);
-
-	n = scatterwalk_copychunks(dst, out, bsize, 1);
-	scatterwalk_advance(out, n);
+	scatterwalk_copychunks(dst, out, bsize, 1);
 
 	return bsize;
 }
@@ -64,12 +59,16 @@ static inline unsigned int crypt_fast(const struct cipher_desc *desc,
 				      unsigned int nbytes, u8 *tmp)
 {
 	u8 *src, *dst;
+	u8 *real_src, *real_dst;
+
+	real_src = scatterwalk_map(in, 0);
+	real_dst = scatterwalk_map(out, 1);
 
-	src = in->data;
-	dst = scatterwalk_samebuf(in, out) ? src : out->data;
+	src = real_src;
+	dst = scatterwalk_samebuf(in, out) ? src : real_dst;
 
 	if (tmp) {
-		memcpy(tmp, in->data, nbytes);
+		memcpy(tmp, src, nbytes);
 		src = tmp;
 		dst = tmp;
 	}
@@ -77,7 +76,10 @@ static inline unsigned int crypt_fast(const struct cipher_desc *desc,
 	nbytes = desc->prfn(desc, dst, src, nbytes);
 
 	if (tmp)
-		memcpy(out->data, tmp, nbytes);
+		memcpy(real_dst, tmp, nbytes);
+
+	scatterwalk_unmap(real_src, 0);
+	scatterwalk_unmap(real_dst, 1);
 
 	scatterwalk_advance(in, nbytes);
 	scatterwalk_advance(out, nbytes);
@@ -126,9 +128,6 @@ static int crypt(const struct cipher_desc *desc,
 			tmp = (u8 *)buffer;
 		}
 
-		scatterwalk_map(&walk_in, 0);
-		scatterwalk_map(&walk_out, 1);
-
 		n = scatterwalk_clamp(&walk_in, n);
 		n = scatterwalk_clamp(&walk_out, n);
 
diff --git a/crypto/scatterwalk.c b/crypto/scatterwalk.c
index 2953e2cc56f0..35172d3f043b 100644
--- a/crypto/scatterwalk.c
+++ b/crypto/scatterwalk.c
@@ -15,9 +15,11 @@
  */
 #include <linux/kernel.h>
 #include <linux/mm.h>
+#include <linux/module.h>
 #include <linux/pagemap.h>
 #include <linux/highmem.h>
-#include <asm/scatterlist.h>
+#include <linux/scatterlist.h>
+
 #include "internal.h"
 #include "scatterwalk.h"
 
@@ -27,88 +29,77 @@ enum km_type crypto_km_types[] = {
 	KM_SOFTIRQ0,
 	KM_SOFTIRQ1,
 };
+EXPORT_SYMBOL_GPL(crypto_km_types);
 
-static void memcpy_dir(void *buf, void *sgdata, size_t nbytes, int out)
+static inline void memcpy_dir(void *buf, void *sgdata, size_t nbytes, int out)
 {
-	if (out)
-		memcpy(sgdata, buf, nbytes);
-	else
-		memcpy(buf, sgdata, nbytes);
+	void *src = out ? buf : sgdata;
+	void *dst = out ? sgdata : buf;
+
+	memcpy(dst, src, nbytes);
 }
 
 void scatterwalk_start(struct scatter_walk *walk, struct scatterlist *sg)
 {
-	unsigned int rest_of_page;
-
 	walk->sg = sg;
 
-	walk->page = sg->page;
-	walk->len_this_segment = sg->length;
-
 	BUG_ON(!sg->length);
 
-	rest_of_page = PAGE_CACHE_SIZE - (sg->offset & (PAGE_CACHE_SIZE - 1));
-	walk->len_this_page = min(sg->length, rest_of_page);
 	walk->offset = sg->offset;
 }
+EXPORT_SYMBOL_GPL(scatterwalk_start);
 
-void scatterwalk_map(struct scatter_walk *walk, int out)
-{
-	walk->data = crypto_kmap(walk->page, out) + walk->offset;
-}
-
-static inline void scatterwalk_unmap(struct scatter_walk *walk, int out)
+void *scatterwalk_map(struct scatter_walk *walk, int out)
 {
-	/* walk->data may be pointing the first byte of the next page;
-	   however, we know we transfered at least one byte.  So,
-	   walk->data - 1 will be a virtual address in the mapped page. */
-	crypto_kunmap(walk->data - 1, out);
+	return crypto_kmap(scatterwalk_page(walk), out) +
+	       offset_in_page(walk->offset);
 }
+EXPORT_SYMBOL_GPL(scatterwalk_map);
 
 static void scatterwalk_pagedone(struct scatter_walk *walk, int out,
 				 unsigned int more)
 {
 	if (out)
-		flush_dcache_page(walk->page);
+		flush_dcache_page(scatterwalk_page(walk));
 
 	if (more) {
-		walk->len_this_segment -= walk->len_this_page;
-
-		if (walk->len_this_segment) {
-			walk->page++;
-			walk->len_this_page = min(walk->len_this_segment,
-						  (unsigned)PAGE_CACHE_SIZE);
-			walk->offset = 0;
-		}
-		else
+		walk->offset += PAGE_SIZE - 1;
+		walk->offset &= PAGE_MASK;
+		if (walk->offset >= walk->sg->offset + walk->sg->length)
 			scatterwalk_start(walk, sg_next(walk->sg));
 	}
 }
 
 void scatterwalk_done(struct scatter_walk *walk, int out, int more)
 {
-	scatterwalk_unmap(walk, out);
-	if (walk->len_this_page == 0 || !more)
+	if (!offset_in_page(walk->offset) || !more)
 		scatterwalk_pagedone(walk, out, more);
 }
+EXPORT_SYMBOL_GPL(scatterwalk_done);
 
-/*
- * Do not call this unless the total length of all of the fragments
- * has been verified as multiple of the block size.
- */
-int scatterwalk_copychunks(void *buf, struct scatter_walk *walk,
-			   size_t nbytes, int out)
+void scatterwalk_copychunks(void *buf, struct scatter_walk *walk,
+			    size_t nbytes, int out)
 {
-	while (nbytes > walk->len_this_page) {
-		memcpy_dir(buf, walk->data, walk->len_this_page, out);
-		buf += walk->len_this_page;
-		nbytes -= walk->len_this_page;
+	for (;;) {
+		unsigned int len_this_page = scatterwalk_pagelen(walk);
+		u8 *vaddr;
+
+		if (len_this_page > nbytes)
+			len_this_page = nbytes;
+
+		vaddr = scatterwalk_map(walk, out);
+		memcpy_dir(buf, vaddr, len_this_page, out);
+		scatterwalk_unmap(vaddr, out);
+
+		if (nbytes == len_this_page)
+			break;
+
+		buf += len_this_page;
+		nbytes -= len_this_page;
 
-		scatterwalk_unmap(walk, out);
 		scatterwalk_pagedone(walk, out, 1);
-		scatterwalk_map(walk, out);
 	}
 
-	memcpy_dir(buf, walk->data, nbytes, out);
-	return nbytes;
+	scatterwalk_advance(walk, nbytes);
 }
+EXPORT_SYMBOL_GPL(scatterwalk_copychunks);
diff --git a/crypto/scatterwalk.h b/crypto/scatterwalk.h
index e79925c474a3..ace595a2e119 100644
--- a/crypto/scatterwalk.h
+++ b/crypto/scatterwalk.h
@@ -14,17 +14,11 @@
 
 #ifndef _CRYPTO_SCATTERWALK_H
 #define _CRYPTO_SCATTERWALK_H
+
 #include <linux/mm.h>
-#include <asm/scatterlist.h>
+#include <linux/scatterlist.h>
 
-struct scatter_walk {
-	struct scatterlist	*sg;
-	struct page		*page;
-	void			*data;
-	unsigned int		len_this_page;
-	unsigned int		len_this_segment;
-	unsigned int		offset;
-};
+#include "internal.h"
 
 /* Define sg_next is an inline routine now in case we want to change
    scatterlist to a linked list later. */
@@ -33,26 +27,31 @@ static inline struct scatterlist *sg_next(struct scatterlist *sg)
 	return sg + 1;
 }
 
-static inline int scatterwalk_samebuf(struct scatter_walk *walk_in,
-				      struct scatter_walk *walk_out)
+static inline unsigned long scatterwalk_samebuf(struct scatter_walk *walk_in,
+						struct scatter_walk *walk_out)
 {
-	return walk_in->page == walk_out->page &&
-	       walk_in->offset == walk_out->offset;
+	return !(((walk_in->sg->page - walk_out->sg->page) << PAGE_SHIFT) +
+		 (int)(walk_in->offset - walk_out->offset));
+}
+
+static inline unsigned int scatterwalk_pagelen(struct scatter_walk *walk)
+{
+	unsigned int len = walk->sg->offset + walk->sg->length - walk->offset;
+	unsigned int len_this_page = offset_in_page(~walk->offset) + 1;
+	return len_this_page > len ? len : len_this_page;
 }
 
 static inline unsigned int scatterwalk_clamp(struct scatter_walk *walk,
 					     unsigned int nbytes)
 {
-	return nbytes > walk->len_this_page ? walk->len_this_page : nbytes;
+	unsigned int len_this_page = scatterwalk_pagelen(walk);
+	return nbytes > len_this_page ? len_this_page : nbytes;
 }
 
 static inline void scatterwalk_advance(struct scatter_walk *walk,
 				       unsigned int nbytes)
 {
-	walk->data += nbytes;
 	walk->offset += nbytes;
-	walk->len_this_page -= nbytes;
-	walk->len_this_segment -= nbytes;
 }
 
 static inline unsigned int scatterwalk_aligned(struct scatter_walk *walk,
@@ -61,9 +60,20 @@ static inline unsigned int scatterwalk_aligned(struct scatter_walk *walk,
 	return !(walk->offset & alignmask);
 }
 
+static inline struct page *scatterwalk_page(struct scatter_walk *walk)
+{
+	return walk->sg->page + (walk->offset >> PAGE_SHIFT);
+}
+
+static inline void scatterwalk_unmap(void *vaddr, int out)
+{
+	crypto_kunmap(vaddr, out);
+}
+
 void scatterwalk_start(struct scatter_walk *walk, struct scatterlist *sg);
-int scatterwalk_copychunks(void *buf, struct scatter_walk *walk, size_t nbytes, int out);
-void scatterwalk_map(struct scatter_walk *walk, int out);
+void scatterwalk_copychunks(void *buf, struct scatter_walk *walk,
+			    size_t nbytes, int out);
+void *scatterwalk_map(struct scatter_walk *walk, int out);
 void scatterwalk_done(struct scatter_walk *walk, int out, int more);
 
 #endif  /* _CRYPTO_SCATTERWALK_H */
diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h
index 6f9fb27b2071..f21ae672e8a8 100644
--- a/include/crypto/algapi.h
+++ b/include/crypto/algapi.h
@@ -50,6 +50,11 @@ struct crypto_spawn {
 	struct crypto_instance *inst;
 };
 
+struct scatter_walk {
+	struct scatterlist *sg;
+	unsigned int offset;
+};
+
 int crypto_register_template(struct crypto_template *tmpl);
 void crypto_unregister_template(struct crypto_template *tmpl);
 struct crypto_template *crypto_lookup_template(const char *name);
-- 
cgit v1.2.3


From 5cde0af2a9825dd1edaca233bd9590566579ef21 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Tue, 22 Aug 2006 00:07:53 +1000
Subject: [CRYPTO] cipher: Added block cipher type

This patch adds the new type of block ciphers.  Unlike current cipher
algorithms which operate on a single block at a time, block ciphers
operate on an arbitrarily long linear area of data.  As it is block-based,
it will skip any data remaining at the end which cannot form a block.

The block cipher has one major difference when compared to the existing
block cipher implementation.  The sg walking is now performed by the
algorithm rather than the cipher mid-layer.  This is needed for drivers
that directly support sg lists.  It also improves performance for all
algorithms as it reduces the total number of indirect calls by one.

In future the existing cipher algorithm will be converted to only have
a single-block interface.  This will be done after all existing users
have switched over to the new block cipher type.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 crypto/Kconfig          |   4 +
 crypto/Makefile         |   2 +
 crypto/blkcipher.c      | 405 ++++++++++++++++++++++++++++++++++++++++++++++++
 include/crypto/algapi.h |  65 ++++++++
 include/linux/crypto.h  | 179 +++++++++++++++++++++
 5 files changed, 655 insertions(+)
 create mode 100644 crypto/blkcipher.c

(limited to 'include')

diff --git a/crypto/Kconfig b/crypto/Kconfig
index 4ce509dba329..68790ad7308d 100644
--- a/crypto/Kconfig
+++ b/crypto/Kconfig
@@ -16,6 +16,10 @@ config CRYPTO_ALGAPI
 	help
 	  This option provides the API for cryptographic algorithms.
 
+config CRYPTO_BLKCIPHER
+	tristate
+	select CRYPTO_ALGAPI
+
 config CRYPTO_MANAGER
 	tristate "Cryptographic algorithm manager"
 	select CRYPTO_ALGAPI
diff --git a/crypto/Makefile b/crypto/Makefile
index b8745f3d3595..b5051951c636 100644
--- a/crypto/Makefile
+++ b/crypto/Makefile
@@ -8,6 +8,8 @@ crypto_algapi-$(CONFIG_PROC_FS) += proc.o
 crypto_algapi-objs := algapi.o $(crypto_algapi-y)
 obj-$(CONFIG_CRYPTO_ALGAPI) += crypto_algapi.o
 
+obj-$(CONFIG_CRYPTO_BLKCIPHER) += blkcipher.o
+
 obj-$(CONFIG_CRYPTO_MANAGER) += cryptomgr.o
 obj-$(CONFIG_CRYPTO_HMAC) += hmac.o
 obj-$(CONFIG_CRYPTO_NULL) += crypto_null.o
diff --git a/crypto/blkcipher.c b/crypto/blkcipher.c
new file mode 100644
index 000000000000..034c939bf91a
--- /dev/null
+++ b/crypto/blkcipher.c
@@ -0,0 +1,405 @@
+/*
+ * Block chaining cipher operations.
+ * 
+ * Generic encrypt/decrypt wrapper for ciphers, handles operations across
+ * multiple page boundaries by using temporary blocks.  In user context,
+ * the kernel is given a chance to schedule us once per page.
+ *
+ * Copyright (c) 2006 Herbert Xu <herbert@gondor.apana.org.au>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; either version 2 of the License, or (at your option) 
+ * any later version.
+ *
+ */
+
+#include <linux/crypto.h>
+#include <linux/errno.h>
+#include <linux/kernel.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/scatterlist.h>
+#include <linux/seq_file.h>
+#include <linux/slab.h>
+#include <linux/string.h>
+
+#include "internal.h"
+#include "scatterwalk.h"
+
+enum {
+	BLKCIPHER_WALK_PHYS = 1 << 0,
+	BLKCIPHER_WALK_SLOW = 1 << 1,
+	BLKCIPHER_WALK_COPY = 1 << 2,
+	BLKCIPHER_WALK_DIFF = 1 << 3,
+};
+
+static int blkcipher_walk_next(struct blkcipher_desc *desc,
+			       struct blkcipher_walk *walk);
+static int blkcipher_walk_first(struct blkcipher_desc *desc,
+				struct blkcipher_walk *walk);
+
+static inline void blkcipher_map_src(struct blkcipher_walk *walk)
+{
+	walk->src.virt.addr = scatterwalk_map(&walk->in, 0);
+}
+
+static inline void blkcipher_map_dst(struct blkcipher_walk *walk)
+{
+	walk->dst.virt.addr = scatterwalk_map(&walk->out, 1);
+}
+
+static inline void blkcipher_unmap_src(struct blkcipher_walk *walk)
+{
+	scatterwalk_unmap(walk->src.virt.addr, 0);
+}
+
+static inline void blkcipher_unmap_dst(struct blkcipher_walk *walk)
+{
+	scatterwalk_unmap(walk->dst.virt.addr, 1);
+}
+
+static inline u8 *blkcipher_get_spot(u8 *start, unsigned int len)
+{
+	if (offset_in_page(start + len) < len)
+		return (u8 *)((unsigned long)(start + len) & PAGE_MASK);
+	return start;
+}
+
+static inline unsigned int blkcipher_done_slow(struct crypto_blkcipher *tfm,
+					       struct blkcipher_walk *walk,
+					       unsigned int bsize)
+{
+	u8 *addr;
+	unsigned int alignmask = crypto_blkcipher_alignmask(tfm);
+
+	addr = (u8 *)ALIGN((unsigned long)walk->buffer, alignmask + 1);
+	addr = blkcipher_get_spot(addr, bsize);
+	scatterwalk_copychunks(addr, &walk->out, bsize, 1);
+	return bsize;
+}
+
+static inline unsigned int blkcipher_done_fast(struct blkcipher_walk *walk,
+					       unsigned int n)
+{
+	n = walk->nbytes - n;
+
+	if (walk->flags & BLKCIPHER_WALK_COPY) {
+		blkcipher_map_dst(walk);
+		memcpy(walk->dst.virt.addr, walk->page, n);
+		blkcipher_unmap_dst(walk);
+	} else if (!(walk->flags & BLKCIPHER_WALK_PHYS)) {
+		blkcipher_unmap_src(walk);
+		if (walk->flags & BLKCIPHER_WALK_DIFF)
+			blkcipher_unmap_dst(walk);
+	}
+
+	scatterwalk_advance(&walk->in, n);
+	scatterwalk_advance(&walk->out, n);
+
+	return n;
+}
+
+int blkcipher_walk_done(struct blkcipher_desc *desc,
+			struct blkcipher_walk *walk, int err)
+{
+	struct crypto_blkcipher *tfm = desc->tfm;
+	unsigned int nbytes = 0;
+
+	if (likely(err >= 0)) {
+		unsigned int bsize = crypto_blkcipher_blocksize(tfm);
+		unsigned int n;
+
+		if (likely(!(walk->flags & BLKCIPHER_WALK_SLOW)))
+			n = blkcipher_done_fast(walk, err);
+		else
+			n = blkcipher_done_slow(tfm, walk, bsize);
+
+		nbytes = walk->total - n;
+		err = 0;
+	}
+
+	scatterwalk_done(&walk->in, 0, nbytes);
+	scatterwalk_done(&walk->out, 1, nbytes);
+
+	walk->total = nbytes;
+	walk->nbytes = nbytes;
+
+	if (nbytes) {
+		crypto_yield(desc->flags);
+		return blkcipher_walk_next(desc, walk);
+	}
+
+	if (walk->iv != desc->info)
+		memcpy(desc->info, walk->iv, crypto_blkcipher_ivsize(tfm));
+	if (walk->buffer != walk->page)
+		kfree(walk->buffer);
+	if (walk->page)
+		free_page((unsigned long)walk->page);
+
+	return err;
+}
+EXPORT_SYMBOL_GPL(blkcipher_walk_done);
+
+static inline int blkcipher_next_slow(struct blkcipher_desc *desc,
+				      struct blkcipher_walk *walk,
+				      unsigned int bsize,
+				      unsigned int alignmask)
+{
+	unsigned int n;
+
+	if (walk->buffer)
+		goto ok;
+
+	walk->buffer = walk->page;
+	if (walk->buffer)
+		goto ok;
+
+	n = bsize * 2 + (alignmask & ~(crypto_tfm_ctx_alignment() - 1));
+	walk->buffer = kmalloc(n, GFP_ATOMIC);
+	if (!walk->buffer)
+		return blkcipher_walk_done(desc, walk, -ENOMEM);
+
+ok:
+	walk->dst.virt.addr = (u8 *)ALIGN((unsigned long)walk->buffer,
+					  alignmask + 1);
+	walk->dst.virt.addr = blkcipher_get_spot(walk->dst.virt.addr, bsize);
+	walk->src.virt.addr = blkcipher_get_spot(walk->dst.virt.addr + bsize,
+						 bsize);
+
+	scatterwalk_copychunks(walk->src.virt.addr, &walk->in, bsize, 0);
+
+	walk->nbytes = bsize;
+	walk->flags |= BLKCIPHER_WALK_SLOW;
+
+	return 0;
+}
+
+static inline int blkcipher_next_copy(struct blkcipher_walk *walk)
+{
+	u8 *tmp = walk->page;
+
+	blkcipher_map_src(walk);
+	memcpy(tmp, walk->src.virt.addr, walk->nbytes);
+	blkcipher_unmap_src(walk);
+
+	walk->src.virt.addr = tmp;
+	walk->dst.virt.addr = tmp;
+
+	return 0;
+}
+
+static inline int blkcipher_next_fast(struct blkcipher_desc *desc,
+				      struct blkcipher_walk *walk)
+{
+	unsigned long diff;
+
+	walk->src.phys.page = scatterwalk_page(&walk->in);
+	walk->src.phys.offset = offset_in_page(walk->in.offset);
+	walk->dst.phys.page = scatterwalk_page(&walk->out);
+	walk->dst.phys.offset = offset_in_page(walk->out.offset);
+
+	if (walk->flags & BLKCIPHER_WALK_PHYS)
+		return 0;
+
+	diff = walk->src.phys.offset - walk->dst.phys.offset;
+	diff |= walk->src.virt.page - walk->dst.virt.page;
+
+	blkcipher_map_src(walk);
+	walk->dst.virt.addr = walk->src.virt.addr;
+
+	if (diff) {
+		walk->flags |= BLKCIPHER_WALK_DIFF;
+		blkcipher_map_dst(walk);
+	}
+
+	return 0;
+}
+
+static int blkcipher_walk_next(struct blkcipher_desc *desc,
+			       struct blkcipher_walk *walk)
+{
+	struct crypto_blkcipher *tfm = desc->tfm;
+	unsigned int alignmask = crypto_blkcipher_alignmask(tfm);
+	unsigned int bsize = crypto_blkcipher_blocksize(tfm);
+	unsigned int n;
+	int err;
+
+	n = walk->total;
+	if (unlikely(n < bsize)) {
+		desc->flags |= CRYPTO_TFM_RES_BAD_BLOCK_LEN;
+		return blkcipher_walk_done(desc, walk, -EINVAL);
+	}
+
+	walk->flags &= ~(BLKCIPHER_WALK_SLOW | BLKCIPHER_WALK_COPY |
+			 BLKCIPHER_WALK_DIFF);
+	if (!scatterwalk_aligned(&walk->in, alignmask) ||
+	    !scatterwalk_aligned(&walk->out, alignmask)) {
+		walk->flags |= BLKCIPHER_WALK_COPY;
+		if (!walk->page) {
+			walk->page = (void *)__get_free_page(GFP_ATOMIC);
+			if (!walk->page)
+				n = 0;
+		}
+	}
+
+	n = scatterwalk_clamp(&walk->in, n);
+	n = scatterwalk_clamp(&walk->out, n);
+
+	if (unlikely(n < bsize)) {
+		err = blkcipher_next_slow(desc, walk, bsize, alignmask);
+		goto set_phys_lowmem;
+	}
+
+	walk->nbytes = n;
+	if (walk->flags & BLKCIPHER_WALK_COPY) {
+		err = blkcipher_next_copy(walk);
+		goto set_phys_lowmem;
+	}
+
+	return blkcipher_next_fast(desc, walk);
+
+set_phys_lowmem:
+	if (walk->flags & BLKCIPHER_WALK_PHYS) {
+		walk->src.phys.page = virt_to_page(walk->src.virt.addr);
+		walk->dst.phys.page = virt_to_page(walk->dst.virt.addr);
+		walk->src.phys.offset &= PAGE_SIZE - 1;
+		walk->dst.phys.offset &= PAGE_SIZE - 1;
+	}
+	return err;
+}
+
+static inline int blkcipher_copy_iv(struct blkcipher_walk *walk,
+				    struct crypto_blkcipher *tfm,
+				    unsigned int alignmask)
+{
+	unsigned bs = crypto_blkcipher_blocksize(tfm);
+	unsigned int ivsize = crypto_blkcipher_ivsize(tfm);
+	unsigned int size = bs * 2 + ivsize + max(bs, ivsize) - (alignmask + 1);
+	u8 *iv;
+
+	size += alignmask & ~(crypto_tfm_ctx_alignment() - 1);
+	walk->buffer = kmalloc(size, GFP_ATOMIC);
+	if (!walk->buffer)
+		return -ENOMEM;
+
+	iv = (u8 *)ALIGN((unsigned long)walk->buffer, alignmask + 1);
+	iv = blkcipher_get_spot(iv, bs) + bs;
+	iv = blkcipher_get_spot(iv, bs) + bs;
+	iv = blkcipher_get_spot(iv, ivsize);
+
+	walk->iv = memcpy(iv, walk->iv, ivsize);
+	return 0;
+}
+
+int blkcipher_walk_virt(struct blkcipher_desc *desc,
+			struct blkcipher_walk *walk)
+{
+	walk->flags &= ~BLKCIPHER_WALK_PHYS;
+	return blkcipher_walk_first(desc, walk);
+}
+EXPORT_SYMBOL_GPL(blkcipher_walk_virt);
+
+int blkcipher_walk_phys(struct blkcipher_desc *desc,
+			struct blkcipher_walk *walk)
+{
+	walk->flags |= BLKCIPHER_WALK_PHYS;
+	return blkcipher_walk_first(desc, walk);
+}
+EXPORT_SYMBOL_GPL(blkcipher_walk_phys);
+
+static int blkcipher_walk_first(struct blkcipher_desc *desc,
+				struct blkcipher_walk *walk)
+{
+	struct crypto_blkcipher *tfm = desc->tfm;
+	unsigned int alignmask = crypto_blkcipher_alignmask(tfm);
+
+	walk->nbytes = walk->total;
+	if (unlikely(!walk->total))
+		return 0;
+
+	walk->buffer = NULL;
+	walk->iv = desc->info;
+	if (unlikely(((unsigned long)walk->iv & alignmask))) {
+		int err = blkcipher_copy_iv(walk, tfm, alignmask);
+		if (err)
+			return err;
+	}
+
+	scatterwalk_start(&walk->in, walk->in.sg);
+	scatterwalk_start(&walk->out, walk->out.sg);
+	walk->page = NULL;
+
+	return blkcipher_walk_next(desc, walk);
+}
+
+static int setkey(struct crypto_tfm *tfm, const u8 *key,
+		  unsigned int keylen)
+{
+	struct blkcipher_alg *cipher = &tfm->__crt_alg->cra_blkcipher;
+
+	if (keylen < cipher->min_keysize || keylen > cipher->max_keysize) {
+		tfm->crt_flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
+		return -EINVAL;
+	}
+
+	return cipher->setkey(tfm, key, keylen);
+}
+
+static unsigned int crypto_blkcipher_ctxsize(struct crypto_alg *alg)
+{
+	struct blkcipher_alg *cipher = &alg->cra_blkcipher;
+	unsigned int len = alg->cra_ctxsize;
+
+	if (cipher->ivsize) {
+		len = ALIGN(len, (unsigned long)alg->cra_alignmask + 1);
+		len += cipher->ivsize;
+	}
+
+	return len;
+}
+
+static int crypto_init_blkcipher_ops(struct crypto_tfm *tfm)
+{
+	struct blkcipher_tfm *crt = &tfm->crt_blkcipher;
+	struct blkcipher_alg *alg = &tfm->__crt_alg->cra_blkcipher;
+	unsigned long align = crypto_tfm_alg_alignmask(tfm) + 1;
+	unsigned long addr;
+
+	if (alg->ivsize > PAGE_SIZE / 8)
+		return -EINVAL;
+
+	crt->setkey = setkey;
+	crt->encrypt = alg->encrypt;
+	crt->decrypt = alg->decrypt;
+
+	addr = (unsigned long)crypto_tfm_ctx(tfm);
+	addr = ALIGN(addr, align);
+	addr += ALIGN(tfm->__crt_alg->cra_ctxsize, align);
+	crt->iv = (void *)addr;
+
+	return 0;
+}
+
+static void crypto_blkcipher_show(struct seq_file *m, struct crypto_alg *alg)
+	__attribute_used__;
+static void crypto_blkcipher_show(struct seq_file *m, struct crypto_alg *alg)
+{
+	seq_printf(m, "type         : blkcipher\n");
+	seq_printf(m, "blocksize    : %u\n", alg->cra_blocksize);
+	seq_printf(m, "min keysize  : %u\n", alg->cra_blkcipher.min_keysize);
+	seq_printf(m, "max keysize  : %u\n", alg->cra_blkcipher.max_keysize);
+	seq_printf(m, "ivsize       : %u\n", alg->cra_blkcipher.ivsize);
+}
+
+const struct crypto_type crypto_blkcipher_type = {
+	.ctxsize = crypto_blkcipher_ctxsize,
+	.init = crypto_init_blkcipher_ops,
+#ifdef CONFIG_PROC_FS
+	.show = crypto_blkcipher_show,
+#endif
+};
+EXPORT_SYMBOL_GPL(crypto_blkcipher_type);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Generic block chaining cipher type");
diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h
index f21ae672e8a8..f3946baf0c07 100644
--- a/include/crypto/algapi.h
+++ b/include/crypto/algapi.h
@@ -55,6 +55,34 @@ struct scatter_walk {
 	unsigned int offset;
 };
 
+struct blkcipher_walk {
+	union {
+		struct {
+			struct page *page;
+			unsigned long offset;
+		} phys;
+
+		struct {
+			u8 *page;
+			u8 *addr;
+		} virt;
+	} src, dst;
+
+	struct scatter_walk in;
+	unsigned int nbytes;
+
+	struct scatter_walk out;
+	unsigned int total;
+
+	void *page;
+	u8 *buffer;
+	u8 *iv;
+
+	int flags;
+};
+
+extern const struct crypto_type crypto_blkcipher_type;
+
 int crypto_register_template(struct crypto_template *tmpl);
 void crypto_unregister_template(struct crypto_template *tmpl);
 struct crypto_template *crypto_lookup_template(const char *name);
@@ -69,15 +97,52 @@ struct crypto_alg *crypto_get_attr_alg(void *param, unsigned int len,
 struct crypto_instance *crypto_alloc_instance(const char *name,
 					      struct crypto_alg *alg);
 
+int blkcipher_walk_done(struct blkcipher_desc *desc,
+			struct blkcipher_walk *walk, int err);
+int blkcipher_walk_virt(struct blkcipher_desc *desc,
+			struct blkcipher_walk *walk);
+int blkcipher_walk_phys(struct blkcipher_desc *desc,
+			struct blkcipher_walk *walk);
+
+static inline void *crypto_tfm_ctx_aligned(struct crypto_tfm *tfm)
+{
+	unsigned long addr = (unsigned long)crypto_tfm_ctx(tfm);
+	unsigned long align = crypto_tfm_alg_alignmask(tfm);
+
+	if (align <= crypto_tfm_ctx_alignment())
+		align = 1;
+	return (void *)ALIGN(addr, align);
+}
+
 static inline void *crypto_instance_ctx(struct crypto_instance *inst)
 {
 	return inst->__ctx;
 }
 
+static inline void *crypto_blkcipher_ctx(struct crypto_blkcipher *tfm)
+{
+	return crypto_tfm_ctx(&tfm->base);
+}
+
+static inline void *crypto_blkcipher_ctx_aligned(struct crypto_blkcipher *tfm)
+{
+	return crypto_tfm_ctx_aligned(&tfm->base);
+}
+
 static inline struct cipher_alg *crypto_cipher_alg(struct crypto_cipher *tfm)
 {
 	return &crypto_cipher_tfm(tfm)->__crt_alg->cra_cipher;
 }
 
+static inline void blkcipher_walk_init(struct blkcipher_walk *walk,
+				       struct scatterlist *dst,
+				       struct scatterlist *src,
+				       unsigned int nbytes)
+{
+	walk->in.sg = src;
+	walk->out.sg = dst;
+	walk->total = nbytes;
+}
+
 #endif	/* _CRYPTO_ALGAPI_H */
 
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index fdecee83878c..5a5466d518e8 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -32,6 +32,7 @@
 #define CRYPTO_ALG_TYPE_MASK		0x0000000f
 #define CRYPTO_ALG_TYPE_CIPHER		0x00000001
 #define CRYPTO_ALG_TYPE_DIGEST		0x00000002
+#define CRYPTO_ALG_TYPE_BLKCIPHER	0x00000003
 #define CRYPTO_ALG_TYPE_COMPRESS	0x00000004
 
 #define CRYPTO_ALG_LARVAL		0x00000010
@@ -89,9 +90,16 @@
 #endif
 
 struct scatterlist;
+struct crypto_blkcipher;
 struct crypto_tfm;
 struct crypto_type;
 
+struct blkcipher_desc {
+	struct crypto_blkcipher *tfm;
+	void *info;
+	u32 flags;
+};
+
 struct cipher_desc {
 	struct crypto_tfm *tfm;
 	void (*crfn)(struct crypto_tfm *tfm, u8 *dst, const u8 *src);
@@ -104,6 +112,21 @@ struct cipher_desc {
  * Algorithms: modular crypto algorithm implementations, managed
  * via crypto_register_alg() and crypto_unregister_alg().
  */
+struct blkcipher_alg {
+	int (*setkey)(struct crypto_tfm *tfm, const u8 *key,
+	              unsigned int keylen);
+	int (*encrypt)(struct blkcipher_desc *desc,
+		       struct scatterlist *dst, struct scatterlist *src,
+		       unsigned int nbytes);
+	int (*decrypt)(struct blkcipher_desc *desc,
+		       struct scatterlist *dst, struct scatterlist *src,
+		       unsigned int nbytes);
+
+	unsigned int min_keysize;
+	unsigned int max_keysize;
+	unsigned int ivsize;
+};
+
 struct cipher_alg {
 	unsigned int cia_min_keysize;
 	unsigned int cia_max_keysize;
@@ -143,6 +166,7 @@ struct compress_alg {
 			      unsigned int slen, u8 *dst, unsigned int *dlen);
 };
 
+#define cra_blkcipher	cra_u.blkcipher
 #define cra_cipher	cra_u.cipher
 #define cra_digest	cra_u.digest
 #define cra_compress	cra_u.compress
@@ -165,6 +189,7 @@ struct crypto_alg {
 	const struct crypto_type *cra_type;
 
 	union {
+		struct blkcipher_alg blkcipher;
 		struct cipher_alg cipher;
 		struct digest_alg digest;
 		struct compress_alg compress;
@@ -201,6 +226,16 @@ static inline int crypto_alg_available(const char *name, u32 flags)
  * crypto_free_*(), as well as the various helpers below.
  */
 
+struct blkcipher_tfm {
+	void *iv;
+	int (*setkey)(struct crypto_tfm *tfm, const u8 *key,
+		      unsigned int keylen);
+	int (*encrypt)(struct blkcipher_desc *desc, struct scatterlist *dst,
+		       struct scatterlist *src, unsigned int nbytes);
+	int (*decrypt)(struct blkcipher_desc *desc, struct scatterlist *dst,
+		       struct scatterlist *src, unsigned int nbytes);
+};
+
 struct cipher_tfm {
 	void *cit_iv;
 	unsigned int cit_ivsize;
@@ -251,6 +286,7 @@ struct compress_tfm {
 	                      u8 *dst, unsigned int *dlen);
 };
 
+#define crt_blkcipher	crt_u.blkcipher
 #define crt_cipher	crt_u.cipher
 #define crt_digest	crt_u.digest
 #define crt_compress	crt_u.compress
@@ -260,6 +296,7 @@ struct crypto_tfm {
 	u32 crt_flags;
 	
 	union {
+		struct blkcipher_tfm blkcipher;
 		struct cipher_tfm cipher;
 		struct digest_tfm digest;
 		struct compress_tfm compress;
@@ -272,6 +309,10 @@ struct crypto_tfm {
 
 #define crypto_cipher crypto_tfm
 
+struct crypto_blkcipher {
+	struct crypto_tfm base;
+};
+
 enum {
 	CRYPTOA_UNSPEC,
 	CRYPTOA_ALG,
@@ -380,6 +421,144 @@ static inline unsigned int crypto_tfm_ctx_alignment(void)
 /*
  * API wrappers.
  */
+static inline struct crypto_blkcipher *__crypto_blkcipher_cast(
+	struct crypto_tfm *tfm)
+{
+	return (struct crypto_blkcipher *)tfm;
+}
+
+static inline struct crypto_blkcipher *crypto_blkcipher_cast(
+	struct crypto_tfm *tfm)
+{
+	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_BLKCIPHER);
+	return __crypto_blkcipher_cast(tfm);
+}
+
+static inline struct crypto_blkcipher *crypto_alloc_blkcipher(
+	const char *alg_name, u32 type, u32 mask)
+{
+	type &= ~CRYPTO_ALG_TYPE_MASK;
+	type |= CRYPTO_ALG_TYPE_BLKCIPHER;
+	mask |= CRYPTO_ALG_TYPE_MASK;
+
+	return __crypto_blkcipher_cast(crypto_alloc_base(alg_name, type, mask));
+}
+
+static inline struct crypto_tfm *crypto_blkcipher_tfm(
+	struct crypto_blkcipher *tfm)
+{
+	return &tfm->base;
+}
+
+static inline void crypto_free_blkcipher(struct crypto_blkcipher *tfm)
+{
+	crypto_free_tfm(crypto_blkcipher_tfm(tfm));
+}
+
+static inline const char *crypto_blkcipher_name(struct crypto_blkcipher *tfm)
+{
+	return crypto_tfm_alg_name(crypto_blkcipher_tfm(tfm));
+}
+
+static inline struct blkcipher_tfm *crypto_blkcipher_crt(
+	struct crypto_blkcipher *tfm)
+{
+	return &crypto_blkcipher_tfm(tfm)->crt_blkcipher;
+}
+
+static inline struct blkcipher_alg *crypto_blkcipher_alg(
+	struct crypto_blkcipher *tfm)
+{
+	return &crypto_blkcipher_tfm(tfm)->__crt_alg->cra_blkcipher;
+}
+
+static inline unsigned int crypto_blkcipher_ivsize(struct crypto_blkcipher *tfm)
+{
+	return crypto_blkcipher_alg(tfm)->ivsize;
+}
+
+static inline unsigned int crypto_blkcipher_blocksize(
+	struct crypto_blkcipher *tfm)
+{
+	return crypto_tfm_alg_blocksize(crypto_blkcipher_tfm(tfm));
+}
+
+static inline unsigned int crypto_blkcipher_alignmask(
+	struct crypto_blkcipher *tfm)
+{
+	return crypto_tfm_alg_alignmask(crypto_blkcipher_tfm(tfm));
+}
+
+static inline u32 crypto_blkcipher_get_flags(struct crypto_blkcipher *tfm)
+{
+	return crypto_tfm_get_flags(crypto_blkcipher_tfm(tfm));
+}
+
+static inline void crypto_blkcipher_set_flags(struct crypto_blkcipher *tfm,
+					      u32 flags)
+{
+	crypto_tfm_set_flags(crypto_blkcipher_tfm(tfm), flags);
+}
+
+static inline void crypto_blkcipher_clear_flags(struct crypto_blkcipher *tfm,
+						u32 flags)
+{
+	crypto_tfm_clear_flags(crypto_blkcipher_tfm(tfm), flags);
+}
+
+static inline int crypto_blkcipher_setkey(struct crypto_blkcipher *tfm,
+					  const u8 *key, unsigned int keylen)
+{
+	return crypto_blkcipher_crt(tfm)->setkey(crypto_blkcipher_tfm(tfm),
+						 key, keylen);
+}
+
+static inline int crypto_blkcipher_encrypt(struct blkcipher_desc *desc,
+					   struct scatterlist *dst,
+					   struct scatterlist *src,
+					   unsigned int nbytes)
+{
+	desc->info = crypto_blkcipher_crt(desc->tfm)->iv;
+	return crypto_blkcipher_crt(desc->tfm)->encrypt(desc, dst, src, nbytes);
+}
+
+static inline int crypto_blkcipher_encrypt_iv(struct blkcipher_desc *desc,
+					      struct scatterlist *dst,
+					      struct scatterlist *src,
+					      unsigned int nbytes)
+{
+	return crypto_blkcipher_crt(desc->tfm)->encrypt(desc, dst, src, nbytes);
+}
+
+static inline int crypto_blkcipher_decrypt(struct blkcipher_desc *desc,
+					   struct scatterlist *dst,
+					   struct scatterlist *src,
+					   unsigned int nbytes)
+{
+	desc->info = crypto_blkcipher_crt(desc->tfm)->iv;
+	return crypto_blkcipher_crt(desc->tfm)->decrypt(desc, dst, src, nbytes);
+}
+
+static inline int crypto_blkcipher_decrypt_iv(struct blkcipher_desc *desc,
+					      struct scatterlist *dst,
+					      struct scatterlist *src,
+					      unsigned int nbytes)
+{
+	return crypto_blkcipher_crt(desc->tfm)->decrypt(desc, dst, src, nbytes);
+}
+
+static inline void crypto_blkcipher_set_iv(struct crypto_blkcipher *tfm,
+					   const u8 *src, unsigned int len)
+{
+	memcpy(crypto_blkcipher_crt(tfm)->iv, src, len);
+}
+
+static inline void crypto_blkcipher_get_iv(struct crypto_blkcipher *tfm,
+					   u8 *dst, unsigned int len)
+{
+	memcpy(dst, crypto_blkcipher_crt(tfm)->iv, len);
+}
+
 static inline struct crypto_cipher *__crypto_cipher_cast(struct crypto_tfm *tfm)
 {
 	return (struct crypto_cipher *)tfm;
-- 
cgit v1.2.3


From db131ef9084110d9e82549c0a627e157e8bb99d7 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Thu, 21 Sep 2006 11:44:08 +1000
Subject: [CRYPTO] cipher: Added block ciphers for CBC/ECB

This patch adds two block cipher algorithms, CBC and ECB.  These
are implemented as templates on top of existing single-block cipher
algorithms.  They invoke the single-block cipher through the new
encrypt_one/decrypt_one interface.

This also optimises the in-place encryption and decryption to remove
the cost of an IV copy each round.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 crypto/Kconfig          |  17 +++
 crypto/Makefile         |   2 +
 crypto/cbc.c            | 344 ++++++++++++++++++++++++++++++++++++++++++++++++
 crypto/ecb.c            | 181 +++++++++++++++++++++++++
 crypto/internal.h       |   1 -
 include/crypto/algapi.h |   2 +
 6 files changed, 546 insertions(+), 1 deletion(-)
 create mode 100644 crypto/cbc.c
 create mode 100644 crypto/ecb.c

(limited to 'include')

diff --git a/crypto/Kconfig b/crypto/Kconfig
index 68790ad7308d..90d467c99c2c 100644
--- a/crypto/Kconfig
+++ b/crypto/Kconfig
@@ -123,6 +123,23 @@ config CRYPTO_TGR192
 	  See also:
 	  <http://www.cs.technion.ac.il/~biham/Reports/Tiger/>.
 
+config CRYPTO_ECB
+	tristate "ECB support"
+	select CRYPTO_BLKCIPHER
+	default m
+	help
+	  ECB: Electronic CodeBook mode
+	  This is the simplest block cipher algorithm.  It simply encrypts
+	  the input block by block.
+
+config CRYPTO_CBC
+	tristate "CBC support"
+	select CRYPTO_BLKCIPHER
+	default m
+	help
+	  CBC: Cipher Block Chaining mode
+	  This block cipher algorithm is required for IPSec.
+
 config CRYPTO_DES
 	tristate "DES and Triple DES EDE cipher algorithms"
 	select CRYPTO_ALGAPI
diff --git a/crypto/Makefile b/crypto/Makefile
index b5051951c636..5e1ff4e0b1fc 100644
--- a/crypto/Makefile
+++ b/crypto/Makefile
@@ -20,6 +20,8 @@ obj-$(CONFIG_CRYPTO_SHA256) += sha256.o
 obj-$(CONFIG_CRYPTO_SHA512) += sha512.o
 obj-$(CONFIG_CRYPTO_WP512) += wp512.o
 obj-$(CONFIG_CRYPTO_TGR192) += tgr192.o
+obj-$(CONFIG_CRYPTO_ECB) += ecb.o
+obj-$(CONFIG_CRYPTO_CBC) += cbc.o
 obj-$(CONFIG_CRYPTO_DES) += des.o
 obj-$(CONFIG_CRYPTO_BLOWFISH) += blowfish.o
 obj-$(CONFIG_CRYPTO_TWOFISH) += twofish.o
diff --git a/crypto/cbc.c b/crypto/cbc.c
new file mode 100644
index 000000000000..f5542b4db387
--- /dev/null
+++ b/crypto/cbc.c
@@ -0,0 +1,344 @@
+/*
+ * CBC: Cipher Block Chaining mode
+ *
+ * Copyright (c) 2006 Herbert Xu <herbert@gondor.apana.org.au>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; either version 2 of the License, or (at your option)
+ * any later version.
+ *
+ */
+
+#include <crypto/algapi.h>
+#include <linux/err.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/scatterlist.h>
+#include <linux/slab.h>
+
+struct crypto_cbc_ctx {
+	struct crypto_cipher *child;
+	void (*xor)(u8 *dst, const u8 *src, unsigned int bs);
+};
+
+static int crypto_cbc_setkey(struct crypto_tfm *parent, const u8 *key,
+			     unsigned int keylen)
+{
+	struct crypto_cbc_ctx *ctx = crypto_tfm_ctx(parent);
+	struct crypto_cipher *child = ctx->child;
+	int err;
+
+	crypto_cipher_clear_flags(child, CRYPTO_TFM_REQ_MASK);
+	crypto_cipher_set_flags(child, crypto_tfm_get_flags(parent) &
+				       CRYPTO_TFM_REQ_MASK);
+	err = crypto_cipher_setkey(child, key, keylen);
+	crypto_tfm_set_flags(parent, crypto_cipher_get_flags(child) &
+				     CRYPTO_TFM_RES_MASK);
+	return err;
+}
+
+static int crypto_cbc_encrypt_segment(struct blkcipher_desc *desc,
+				      struct blkcipher_walk *walk,
+				      struct crypto_cipher *tfm,
+				      void (*xor)(u8 *, const u8 *,
+						  unsigned int))
+{
+	void (*fn)(struct crypto_tfm *, u8 *, const u8 *) =
+		crypto_cipher_alg(tfm)->cia_encrypt;
+	int bsize = crypto_cipher_blocksize(tfm);
+	unsigned int nbytes = walk->nbytes;
+	u8 *src = walk->src.virt.addr;
+	u8 *dst = walk->dst.virt.addr;
+	u8 *iv = walk->iv;
+
+	do {
+		xor(iv, src, bsize);
+		fn(crypto_cipher_tfm(tfm), dst, iv);
+		memcpy(iv, dst, bsize);
+
+		src += bsize;
+		dst += bsize;
+	} while ((nbytes -= bsize) >= bsize);
+
+	return nbytes;
+}
+
+static int crypto_cbc_encrypt_inplace(struct blkcipher_desc *desc,
+				      struct blkcipher_walk *walk,
+				      struct crypto_cipher *tfm,
+				      void (*xor)(u8 *, const u8 *,
+						  unsigned int))
+{
+	void (*fn)(struct crypto_tfm *, u8 *, const u8 *) =
+		crypto_cipher_alg(tfm)->cia_encrypt;
+	int bsize = crypto_cipher_blocksize(tfm);
+	unsigned int nbytes = walk->nbytes;
+	u8 *src = walk->src.virt.addr;
+	u8 *iv = walk->iv;
+
+	do {
+		xor(src, iv, bsize);
+		fn(crypto_cipher_tfm(tfm), src, src);
+		iv = src;
+
+		src += bsize;
+	} while ((nbytes -= bsize) >= bsize);
+
+	memcpy(walk->iv, iv, bsize);
+
+	return nbytes;
+}
+
+static int crypto_cbc_encrypt(struct blkcipher_desc *desc,
+			      struct scatterlist *dst, struct scatterlist *src,
+			      unsigned int nbytes)
+{
+	struct blkcipher_walk walk;
+	struct crypto_blkcipher *tfm = desc->tfm;
+	struct crypto_cbc_ctx *ctx = crypto_blkcipher_ctx(tfm);
+	struct crypto_cipher *child = ctx->child;
+	void (*xor)(u8 *, const u8 *, unsigned int bs) = ctx->xor;
+	int err;
+
+	blkcipher_walk_init(&walk, dst, src, nbytes);
+	err = blkcipher_walk_virt(desc, &walk);
+
+	while ((nbytes = walk.nbytes)) {
+		if (walk.src.virt.addr == walk.dst.virt.addr)
+			nbytes = crypto_cbc_encrypt_inplace(desc, &walk, child,
+							    xor);
+		else
+			nbytes = crypto_cbc_encrypt_segment(desc, &walk, child,
+							    xor);
+		err = blkcipher_walk_done(desc, &walk, nbytes);
+	}
+
+	return err;
+}
+
+static int crypto_cbc_decrypt_segment(struct blkcipher_desc *desc,
+				      struct blkcipher_walk *walk,
+				      struct crypto_cipher *tfm,
+				      void (*xor)(u8 *, const u8 *,
+						  unsigned int))
+{
+	void (*fn)(struct crypto_tfm *, u8 *, const u8 *) =
+		crypto_cipher_alg(tfm)->cia_decrypt;
+	int bsize = crypto_cipher_blocksize(tfm);
+	unsigned int nbytes = walk->nbytes;
+	u8 *src = walk->src.virt.addr;
+	u8 *dst = walk->dst.virt.addr;
+	u8 *iv = walk->iv;
+
+	do {
+		fn(crypto_cipher_tfm(tfm), dst, src);
+		xor(dst, iv, bsize);
+		iv = src;
+
+		src += bsize;
+		dst += bsize;
+	} while ((nbytes -= bsize) >= bsize);
+
+	memcpy(walk->iv, iv, bsize);
+
+	return nbytes;
+}
+
+static int crypto_cbc_decrypt_inplace(struct blkcipher_desc *desc,
+				      struct blkcipher_walk *walk,
+				      struct crypto_cipher *tfm,
+				      void (*xor)(u8 *, const u8 *,
+						  unsigned int))
+{
+	void (*fn)(struct crypto_tfm *, u8 *, const u8 *) =
+		crypto_cipher_alg(tfm)->cia_decrypt;
+	int bsize = crypto_cipher_blocksize(tfm);
+	unsigned long alignmask = crypto_cipher_alignmask(tfm);
+	unsigned int nbytes = walk->nbytes;
+	u8 *src = walk->src.virt.addr;
+	u8 stack[bsize + alignmask];
+	u8 *first_iv = (u8 *)ALIGN((unsigned long)stack, alignmask + 1);
+
+	memcpy(first_iv, walk->iv, bsize);
+
+	/* Start of the last block. */
+	src += nbytes - nbytes % bsize - bsize;
+	memcpy(walk->iv, src, bsize);
+
+	for (;;) {
+		fn(crypto_cipher_tfm(tfm), src, src);
+		if ((nbytes -= bsize) < bsize)
+			break;
+		xor(src, src - bsize, bsize);
+		src -= bsize;
+	}
+
+	xor(src, first_iv, bsize);
+
+	return nbytes;
+}
+
+static int crypto_cbc_decrypt(struct blkcipher_desc *desc,
+			      struct scatterlist *dst, struct scatterlist *src,
+			      unsigned int nbytes)
+{
+	struct blkcipher_walk walk;
+	struct crypto_blkcipher *tfm = desc->tfm;
+	struct crypto_cbc_ctx *ctx = crypto_blkcipher_ctx(tfm);
+	struct crypto_cipher *child = ctx->child;
+	void (*xor)(u8 *, const u8 *, unsigned int bs) = ctx->xor;
+	int err;
+
+	blkcipher_walk_init(&walk, dst, src, nbytes);
+	err = blkcipher_walk_virt(desc, &walk);
+
+	while ((nbytes = walk.nbytes)) {
+		if (walk.src.virt.addr == walk.dst.virt.addr)
+			nbytes = crypto_cbc_decrypt_inplace(desc, &walk, child,
+							    xor);
+		else
+			nbytes = crypto_cbc_decrypt_segment(desc, &walk, child,
+							    xor);
+		err = blkcipher_walk_done(desc, &walk, nbytes);
+	}
+
+	return err;
+}
+
+static void xor_byte(u8 *a, const u8 *b, unsigned int bs)
+{
+	do {
+		*a++ ^= *b++;
+	} while (--bs);
+}
+
+static void xor_quad(u8 *dst, const u8 *src, unsigned int bs)
+{
+	u32 *a = (u32 *)dst;
+	u32 *b = (u32 *)src;
+
+	do {
+		*a++ ^= *b++;
+	} while ((bs -= 4));
+}
+
+static void xor_64(u8 *a, const u8 *b, unsigned int bs)
+{
+	((u32 *)a)[0] ^= ((u32 *)b)[0];
+	((u32 *)a)[1] ^= ((u32 *)b)[1];
+}
+
+static void xor_128(u8 *a, const u8 *b, unsigned int bs)
+{
+	((u32 *)a)[0] ^= ((u32 *)b)[0];
+	((u32 *)a)[1] ^= ((u32 *)b)[1];
+	((u32 *)a)[2] ^= ((u32 *)b)[2];
+	((u32 *)a)[3] ^= ((u32 *)b)[3];
+}
+
+static int crypto_cbc_init_tfm(struct crypto_tfm *tfm)
+{
+	struct crypto_instance *inst = (void *)tfm->__crt_alg;
+	struct crypto_spawn *spawn = crypto_instance_ctx(inst);
+	struct crypto_cbc_ctx *ctx = crypto_tfm_ctx(tfm);
+
+	switch (crypto_tfm_alg_blocksize(tfm)) {
+	case 8:
+		ctx->xor = xor_64;
+		break;
+
+	case 16:
+		ctx->xor = xor_128;
+		break;
+
+	default:
+		if (crypto_tfm_alg_blocksize(tfm) % 4)
+			ctx->xor = xor_byte;
+		else
+			ctx->xor = xor_quad;
+	}
+
+	tfm = crypto_spawn_tfm(spawn);
+	if (IS_ERR(tfm))
+		return PTR_ERR(tfm);
+
+	ctx->child = crypto_cipher_cast(tfm);
+	return 0;
+}
+
+static void crypto_cbc_exit_tfm(struct crypto_tfm *tfm)
+{
+	struct crypto_cbc_ctx *ctx = crypto_tfm_ctx(tfm);
+	crypto_free_cipher(ctx->child);
+}
+
+static struct crypto_instance *crypto_cbc_alloc(void *param, unsigned int len)
+{
+	struct crypto_instance *inst;
+	struct crypto_alg *alg;
+
+	alg = crypto_get_attr_alg(param, len, CRYPTO_ALG_TYPE_CIPHER,
+				  CRYPTO_ALG_TYPE_MASK | CRYPTO_ALG_ASYNC);
+	if (IS_ERR(alg))
+		return ERR_PTR(PTR_ERR(alg));
+
+	inst = crypto_alloc_instance("cbc", alg);
+	if (IS_ERR(inst))
+		goto out_put_alg;
+
+	inst->alg.cra_flags = CRYPTO_ALG_TYPE_BLKCIPHER;
+	inst->alg.cra_priority = alg->cra_priority;
+	inst->alg.cra_blocksize = alg->cra_blocksize;
+	inst->alg.cra_alignmask = alg->cra_alignmask;
+	inst->alg.cra_type = &crypto_blkcipher_type;
+
+	if (!(alg->cra_blocksize % 4))
+		inst->alg.cra_alignmask |= 3;
+	inst->alg.cra_blkcipher.ivsize = alg->cra_blocksize;
+	inst->alg.cra_blkcipher.min_keysize = alg->cra_cipher.cia_min_keysize;
+	inst->alg.cra_blkcipher.max_keysize = alg->cra_cipher.cia_max_keysize;
+
+	inst->alg.cra_ctxsize = sizeof(struct crypto_cbc_ctx);
+
+	inst->alg.cra_init = crypto_cbc_init_tfm;
+	inst->alg.cra_exit = crypto_cbc_exit_tfm;
+
+	inst->alg.cra_blkcipher.setkey = crypto_cbc_setkey;
+	inst->alg.cra_blkcipher.encrypt = crypto_cbc_encrypt;
+	inst->alg.cra_blkcipher.decrypt = crypto_cbc_decrypt;
+
+out_put_alg:
+	crypto_mod_put(alg);
+	return inst;
+}
+
+static void crypto_cbc_free(struct crypto_instance *inst)
+{
+	crypto_drop_spawn(crypto_instance_ctx(inst));
+	kfree(inst);
+}
+
+static struct crypto_template crypto_cbc_tmpl = {
+	.name = "cbc",
+	.alloc = crypto_cbc_alloc,
+	.free = crypto_cbc_free,
+	.module = THIS_MODULE,
+};
+
+static int __init crypto_cbc_module_init(void)
+{
+	return crypto_register_template(&crypto_cbc_tmpl);
+}
+
+static void __exit crypto_cbc_module_exit(void)
+{
+	crypto_unregister_template(&crypto_cbc_tmpl);
+}
+
+module_init(crypto_cbc_module_init);
+module_exit(crypto_cbc_module_exit);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("CBC block cipher algorithm");
diff --git a/crypto/ecb.c b/crypto/ecb.c
new file mode 100644
index 000000000000..f239aa9c4017
--- /dev/null
+++ b/crypto/ecb.c
@@ -0,0 +1,181 @@
+/*
+ * ECB: Electronic CodeBook mode
+ *
+ * Copyright (c) 2006 Herbert Xu <herbert@gondor.apana.org.au>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; either version 2 of the License, or (at your option)
+ * any later version.
+ *
+ */
+
+#include <crypto/algapi.h>
+#include <linux/err.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/scatterlist.h>
+#include <linux/slab.h>
+
+struct crypto_ecb_ctx {
+	struct crypto_cipher *child;
+};
+
+static int crypto_ecb_setkey(struct crypto_tfm *parent, const u8 *key,
+			     unsigned int keylen)
+{
+	struct crypto_ecb_ctx *ctx = crypto_tfm_ctx(parent);
+	struct crypto_cipher *child = ctx->child;
+	int err;
+
+	crypto_cipher_clear_flags(child, CRYPTO_TFM_REQ_MASK);
+	crypto_cipher_set_flags(child, crypto_tfm_get_flags(parent) &
+				       CRYPTO_TFM_REQ_MASK);
+	err = crypto_cipher_setkey(child, key, keylen);
+	crypto_tfm_set_flags(parent, crypto_cipher_get_flags(child) &
+				     CRYPTO_TFM_RES_MASK);
+	return err;
+}
+
+static int crypto_ecb_crypt(struct blkcipher_desc *desc,
+			    struct blkcipher_walk *walk,
+			    struct crypto_cipher *tfm,
+			    void (*fn)(struct crypto_tfm *, u8 *, const u8 *))
+{
+	int bsize = crypto_cipher_blocksize(tfm);
+	unsigned int nbytes;
+	int err;
+
+	err = blkcipher_walk_virt(desc, walk);
+
+	while ((nbytes = walk->nbytes)) {
+		u8 *wsrc = walk->src.virt.addr;
+		u8 *wdst = walk->dst.virt.addr;
+
+		do {
+			fn(crypto_cipher_tfm(tfm), wdst, wsrc);
+	
+			wsrc += bsize;
+			wdst += bsize;
+		} while ((nbytes -= bsize) >= bsize);
+
+		err = blkcipher_walk_done(desc, walk, nbytes);
+	}
+
+	return err;
+}
+
+static int crypto_ecb_encrypt(struct blkcipher_desc *desc,
+			      struct scatterlist *dst, struct scatterlist *src,
+			      unsigned int nbytes)
+{
+	struct blkcipher_walk walk;
+	struct crypto_blkcipher *tfm = desc->tfm;
+	struct crypto_ecb_ctx *ctx = crypto_blkcipher_ctx(tfm);
+	struct crypto_cipher *child = ctx->child;
+
+	blkcipher_walk_init(&walk, dst, src, nbytes);
+	return crypto_ecb_crypt(desc, &walk, child,
+				crypto_cipher_alg(child)->cia_encrypt);
+}
+
+static int crypto_ecb_decrypt(struct blkcipher_desc *desc,
+			      struct scatterlist *dst, struct scatterlist *src,
+			      unsigned int nbytes)
+{
+	struct blkcipher_walk walk;
+	struct crypto_blkcipher *tfm = desc->tfm;
+	struct crypto_ecb_ctx *ctx = crypto_blkcipher_ctx(tfm);
+	struct crypto_cipher *child = ctx->child;
+
+	blkcipher_walk_init(&walk, dst, src, nbytes);
+	return crypto_ecb_crypt(desc, &walk, child,
+				crypto_cipher_alg(child)->cia_decrypt);
+}
+
+static int crypto_ecb_init_tfm(struct crypto_tfm *tfm)
+{
+	struct crypto_instance *inst = (void *)tfm->__crt_alg;
+	struct crypto_spawn *spawn = crypto_instance_ctx(inst);
+	struct crypto_ecb_ctx *ctx = crypto_tfm_ctx(tfm);
+
+	tfm = crypto_spawn_tfm(spawn);
+	if (IS_ERR(tfm))
+		return PTR_ERR(tfm);
+
+	ctx->child = crypto_cipher_cast(tfm);
+	return 0;
+}
+
+static void crypto_ecb_exit_tfm(struct crypto_tfm *tfm)
+{
+	struct crypto_ecb_ctx *ctx = crypto_tfm_ctx(tfm);
+	crypto_free_cipher(ctx->child);
+}
+
+static struct crypto_instance *crypto_ecb_alloc(void *param, unsigned int len)
+{
+	struct crypto_instance *inst;
+	struct crypto_alg *alg;
+
+	alg = crypto_get_attr_alg(param, len, CRYPTO_ALG_TYPE_CIPHER,
+				  CRYPTO_ALG_TYPE_MASK | CRYPTO_ALG_ASYNC);
+	if (IS_ERR(alg))
+		return ERR_PTR(PTR_ERR(alg));
+
+	inst = crypto_alloc_instance("ecb", alg);
+	if (IS_ERR(inst))
+		goto out_put_alg;
+
+	inst->alg.cra_flags = CRYPTO_ALG_TYPE_BLKCIPHER;
+	inst->alg.cra_priority = alg->cra_priority;
+	inst->alg.cra_blocksize = alg->cra_blocksize;
+	inst->alg.cra_alignmask = alg->cra_alignmask;
+	inst->alg.cra_type = &crypto_blkcipher_type;
+
+	inst->alg.cra_blkcipher.min_keysize = alg->cra_cipher.cia_min_keysize;
+	inst->alg.cra_blkcipher.max_keysize = alg->cra_cipher.cia_max_keysize;
+
+	inst->alg.cra_ctxsize = sizeof(struct crypto_ecb_ctx);
+
+	inst->alg.cra_init = crypto_ecb_init_tfm;
+	inst->alg.cra_exit = crypto_ecb_exit_tfm;
+
+	inst->alg.cra_blkcipher.setkey = crypto_ecb_setkey;
+	inst->alg.cra_blkcipher.encrypt = crypto_ecb_encrypt;
+	inst->alg.cra_blkcipher.decrypt = crypto_ecb_decrypt;
+
+out_put_alg:
+	crypto_mod_put(alg);
+	return inst;
+}
+
+static void crypto_ecb_free(struct crypto_instance *inst)
+{
+	crypto_drop_spawn(crypto_instance_ctx(inst));
+	kfree(inst);
+}
+
+static struct crypto_template crypto_ecb_tmpl = {
+	.name = "ecb",
+	.alloc = crypto_ecb_alloc,
+	.free = crypto_ecb_free,
+	.module = THIS_MODULE,
+};
+
+static int __init crypto_ecb_module_init(void)
+{
+	return crypto_register_template(&crypto_ecb_tmpl);
+}
+
+static void __exit crypto_ecb_module_exit(void)
+{
+	crypto_unregister_template(&crypto_ecb_tmpl);
+}
+
+module_init(crypto_ecb_module_init);
+module_exit(crypto_ecb_module_exit);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("ECB block cipher algorithm");
diff --git a/crypto/internal.h b/crypto/internal.h
index 7dc04efb55c6..93d9b10ff914 100644
--- a/crypto/internal.h
+++ b/crypto/internal.h
@@ -131,7 +131,6 @@ static inline unsigned int crypto_compress_ctxsize(struct crypto_alg *alg,
 }
 
 struct crypto_alg *crypto_mod_get(struct crypto_alg *alg);
-void crypto_mod_put(struct crypto_alg *alg);
 struct crypto_alg *__crypto_alg_lookup(const char *name, u32 type, u32 mask);
 struct crypto_alg *crypto_alg_mod_lookup(const char *name, u32 type, u32 mask);
 
diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h
index f3946baf0c07..444f602724db 100644
--- a/include/crypto/algapi.h
+++ b/include/crypto/algapi.h
@@ -83,6 +83,8 @@ struct blkcipher_walk {
 
 extern const struct crypto_type crypto_blkcipher_type;
 
+void crypto_mod_put(struct crypto_alg *alg);
+
 int crypto_register_template(struct crypto_template *tmpl);
 void crypto_unregister_template(struct crypto_template *tmpl);
 struct crypto_template *crypto_lookup_template(const char *name);
-- 
cgit v1.2.3


From 04ff12609445c7b462d7fc7f2d30dad442c922f3 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sun, 13 Aug 2006 08:50:00 +1000
Subject: [IPSEC]: Add compatibility algorithm name support

This patch adds a compatibility name field for each IPsec algorithm.  This
is needed when parameterised algorithms are used.  For example, "md5" will
become "hmac(md5)", and "aes" will become "cbc(aes)".

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 include/net/xfrm.h   | 1 +
 net/xfrm/xfrm_algo.c | 3 ++-
 net/xfrm/xfrm_user.c | 1 +
 3 files changed, 4 insertions(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 10396b4bde14..e9114e41affc 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -854,6 +854,7 @@ struct xfrm_algo_comp_info {
 
 struct xfrm_algo_desc {
 	char *name;
+	char *compat;
 	u8 available:1;
 	union {
 		struct xfrm_algo_auth_info auth;
diff --git a/net/xfrm/xfrm_algo.c b/net/xfrm/xfrm_algo.c
index 04e1aea58bc9..b68974b38741 100644
--- a/net/xfrm/xfrm_algo.c
+++ b/net/xfrm/xfrm_algo.c
@@ -359,7 +359,8 @@ static struct xfrm_algo_desc *xfrm_get_byname(struct xfrm_algo_desc *list,
 		return NULL;
 
 	for (i = 0; i < entries; i++) {
-		if (strcmp(name, list[i].name))
+		if (strcmp(name, list[i].name) &&
+		    (!list[i].compat || strcmp(name, list[i].compat)))
 			continue;
 
 		if (list[i].available)
diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c
index 7d18ca03c80d..fa79ddc4239e 100644
--- a/net/xfrm/xfrm_user.c
+++ b/net/xfrm/xfrm_user.c
@@ -213,6 +213,7 @@ static int attach_one_algo(struct xfrm_algo **algpp, u8 *props,
 		return -ENOMEM;
 
 	memcpy(p, ualg, len);
+	strcpy(p->alg_name, algo->name);
 	*algpp = p;
 	return 0;
 }
-- 
cgit v1.2.3


From 6b7326c8497f954c2cfcb4c49fe42be5b80887bc Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sun, 30 Jul 2006 15:41:01 +1000
Subject: [IPSEC] ESP: Use block ciphers where applicable

This patch converts IPSec/ESP to use the new block cipher type where
applicable.  Similar to the HMAC conversion, existing algorithm names
have been kept for compatibility.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 include/net/esp.h    |  2 +-
 net/ipv4/Kconfig     |  1 +
 net/ipv4/esp4.c      | 49 +++++++++++++++++++++++++++++--------------------
 net/ipv6/Kconfig     |  1 +
 net/ipv6/esp6.c      | 48 ++++++++++++++++++++++++++++--------------------
 net/xfrm/xfrm_algo.c | 24 ++++++++++++++++--------
 6 files changed, 76 insertions(+), 49 deletions(-)

(limited to 'include')

diff --git a/include/net/esp.h b/include/net/esp.h
index 6eb837973c84..af2ff18700c7 100644
--- a/include/net/esp.h
+++ b/include/net/esp.h
@@ -22,7 +22,7 @@ struct esp_data
 		 * >= crypto_tfm_alg_ivsize(tfm). */
 		int			ivlen;
 		int			padlen;		/* 0..255 */
-		struct crypto_tfm	*tfm;		/* crypto handle */
+		struct crypto_blkcipher	*tfm;		/* crypto handle */
 	} conf;
 
 	/* Integrity. It is active when icv_full_len != 0 */
diff --git a/net/ipv4/Kconfig b/net/ipv4/Kconfig
index 8514106761b0..3b5d504a74be 100644
--- a/net/ipv4/Kconfig
+++ b/net/ipv4/Kconfig
@@ -386,6 +386,7 @@ config INET_ESP
 	select CRYPTO
 	select CRYPTO_HMAC
 	select CRYPTO_MD5
+	select CRYPTO_CBC
 	select CRYPTO_SHA1
 	select CRYPTO_DES
 	---help---
diff --git a/net/ipv4/esp4.c b/net/ipv4/esp4.c
index fc2f8ce441de..7c63ae494742 100644
--- a/net/ipv4/esp4.c
+++ b/net/ipv4/esp4.c
@@ -1,3 +1,4 @@
+#include <linux/err.h>
 #include <linux/module.h>
 #include <net/ip.h>
 #include <net/xfrm.h>
@@ -16,7 +17,8 @@ static int esp_output(struct xfrm_state *x, struct sk_buff *skb)
 	int err;
 	struct iphdr *top_iph;
 	struct ip_esp_hdr *esph;
-	struct crypto_tfm *tfm;
+	struct crypto_blkcipher *tfm;
+	struct blkcipher_desc desc;
 	struct esp_data *esp;
 	struct sk_buff *trailer;
 	int blksize;
@@ -36,7 +38,9 @@ static int esp_output(struct xfrm_state *x, struct sk_buff *skb)
 	esp = x->data;
 	alen = esp->auth.icv_trunc_len;
 	tfm = esp->conf.tfm;
-	blksize = ALIGN(crypto_tfm_alg_blocksize(tfm), 4);
+	desc.tfm = tfm;
+	desc.flags = 0;
+	blksize = ALIGN(crypto_blkcipher_blocksize(tfm), 4);
 	clen = ALIGN(clen + 2, blksize);
 	if (esp->conf.padlen)
 		clen = ALIGN(clen, esp->conf.padlen);
@@ -92,7 +96,7 @@ static int esp_output(struct xfrm_state *x, struct sk_buff *skb)
 	xfrm_aevent_doreplay(x);
 
 	if (esp->conf.ivlen)
-		crypto_cipher_set_iv(tfm, esp->conf.ivec, crypto_tfm_alg_ivsize(tfm));
+		crypto_blkcipher_set_iv(tfm, esp->conf.ivec, esp->conf.ivlen);
 
 	do {
 		struct scatterlist *sg = &esp->sgbuf[0];
@@ -103,14 +107,17 @@ static int esp_output(struct xfrm_state *x, struct sk_buff *skb)
 				goto error;
 		}
 		skb_to_sgvec(skb, sg, esph->enc_data+esp->conf.ivlen-skb->data, clen);
-		crypto_cipher_encrypt(tfm, sg, sg, clen);
+		err = crypto_blkcipher_encrypt(&desc, sg, sg, clen);
 		if (unlikely(sg != &esp->sgbuf[0]))
 			kfree(sg);
 	} while (0);
 
+	if (unlikely(err))
+		goto error;
+
 	if (esp->conf.ivlen) {
-		memcpy(esph->enc_data, esp->conf.ivec, crypto_tfm_alg_ivsize(tfm));
-		crypto_cipher_get_iv(tfm, esp->conf.ivec, crypto_tfm_alg_ivsize(tfm));
+		memcpy(esph->enc_data, esp->conf.ivec, esp->conf.ivlen);
+		crypto_blkcipher_get_iv(tfm, esp->conf.ivec, esp->conf.ivlen);
 	}
 
 	if (esp->auth.icv_full_len) {
@@ -121,8 +128,6 @@ static int esp_output(struct xfrm_state *x, struct sk_buff *skb)
 
 	ip_send_check(top_iph);
 
-	err = 0;
-
 error:
 	return err;
 }
@@ -137,8 +142,10 @@ static int esp_input(struct xfrm_state *x, struct sk_buff *skb)
 	struct iphdr *iph;
 	struct ip_esp_hdr *esph;
 	struct esp_data *esp = x->data;
+	struct crypto_blkcipher *tfm = esp->conf.tfm;
+	struct blkcipher_desc desc = { .tfm = tfm };
 	struct sk_buff *trailer;
-	int blksize = ALIGN(crypto_tfm_alg_blocksize(esp->conf.tfm), 4);
+	int blksize = ALIGN(crypto_blkcipher_blocksize(tfm), 4);
 	int alen = esp->auth.icv_trunc_len;
 	int elen = skb->len - sizeof(struct ip_esp_hdr) - esp->conf.ivlen - alen;
 	int nfrags;
@@ -146,6 +153,7 @@ static int esp_input(struct xfrm_state *x, struct sk_buff *skb)
 	u8 nexthdr[2];
 	struct scatterlist *sg;
 	int padlen;
+	int err;
 
 	if (!pskb_may_pull(skb, sizeof(struct ip_esp_hdr)))
 		goto out;
@@ -178,7 +186,7 @@ static int esp_input(struct xfrm_state *x, struct sk_buff *skb)
 
 	/* Get ivec. This can be wrong, check against another impls. */
 	if (esp->conf.ivlen)
-		crypto_cipher_set_iv(esp->conf.tfm, esph->enc_data, crypto_tfm_alg_ivsize(esp->conf.tfm));
+		crypto_blkcipher_set_iv(tfm, esph->enc_data, esp->conf.ivlen);
 
 	sg = &esp->sgbuf[0];
 
@@ -188,9 +196,11 @@ static int esp_input(struct xfrm_state *x, struct sk_buff *skb)
 			goto out;
 	}
 	skb_to_sgvec(skb, sg, sizeof(struct ip_esp_hdr) + esp->conf.ivlen, elen);
-	crypto_cipher_decrypt(esp->conf.tfm, sg, sg, elen);
+	err = crypto_blkcipher_decrypt(&desc, sg, sg, elen);
 	if (unlikely(sg != &esp->sgbuf[0]))
 		kfree(sg);
+	if (unlikely(err))
+		return err;
 
 	if (skb_copy_bits(skb, skb->len-alen-2, nexthdr, 2))
 		BUG();
@@ -254,7 +264,7 @@ out:
 static u32 esp4_get_max_size(struct xfrm_state *x, int mtu)
 {
 	struct esp_data *esp = x->data;
-	u32 blksize = ALIGN(crypto_tfm_alg_blocksize(esp->conf.tfm), 4);
+	u32 blksize = ALIGN(crypto_blkcipher_blocksize(esp->conf.tfm), 4);
 
 	if (x->props.mode) {
 		mtu = ALIGN(mtu + 2, blksize);
@@ -293,7 +303,7 @@ static void esp_destroy(struct xfrm_state *x)
 	if (!esp)
 		return;
 
-	crypto_free_tfm(esp->conf.tfm);
+	crypto_free_blkcipher(esp->conf.tfm);
 	esp->conf.tfm = NULL;
 	kfree(esp->conf.ivec);
 	esp->conf.ivec = NULL;
@@ -307,6 +317,7 @@ static void esp_destroy(struct xfrm_state *x)
 static int esp_init_state(struct xfrm_state *x)
 {
 	struct esp_data *esp = NULL;
+	struct crypto_blkcipher *tfm;
 
 	/* null auth and encryption can have zero length keys */
 	if (x->aalg) {
@@ -351,13 +362,11 @@ static int esp_init_state(struct xfrm_state *x)
 	}
 	esp->conf.key = x->ealg->alg_key;
 	esp->conf.key_len = (x->ealg->alg_key_len+7)/8;
-	if (x->props.ealgo == SADB_EALG_NULL)
-		esp->conf.tfm = crypto_alloc_tfm(x->ealg->alg_name, CRYPTO_TFM_MODE_ECB);
-	else
-		esp->conf.tfm = crypto_alloc_tfm(x->ealg->alg_name, CRYPTO_TFM_MODE_CBC);
-	if (esp->conf.tfm == NULL)
+	tfm = crypto_alloc_blkcipher(x->ealg->alg_name, 0, CRYPTO_ALG_ASYNC);
+	if (IS_ERR(tfm))
 		goto error;
-	esp->conf.ivlen = crypto_tfm_alg_ivsize(esp->conf.tfm);
+	esp->conf.tfm = tfm;
+	esp->conf.ivlen = crypto_blkcipher_ivsize(tfm);
 	esp->conf.padlen = 0;
 	if (esp->conf.ivlen) {
 		esp->conf.ivec = kmalloc(esp->conf.ivlen, GFP_KERNEL);
@@ -365,7 +374,7 @@ static int esp_init_state(struct xfrm_state *x)
 			goto error;
 		get_random_bytes(esp->conf.ivec, esp->conf.ivlen);
 	}
-	if (crypto_cipher_setkey(esp->conf.tfm, esp->conf.key, esp->conf.key_len))
+	if (crypto_blkcipher_setkey(tfm, esp->conf.key, esp->conf.key_len))
 		goto error;
 	x->props.header_len = sizeof(struct ip_esp_hdr) + esp->conf.ivlen;
 	if (x->props.mode)
diff --git a/net/ipv6/Kconfig b/net/ipv6/Kconfig
index e923d4dea418..0ba06c0c5d39 100644
--- a/net/ipv6/Kconfig
+++ b/net/ipv6/Kconfig
@@ -77,6 +77,7 @@ config INET6_ESP
 	select CRYPTO
 	select CRYPTO_HMAC
 	select CRYPTO_MD5
+	select CRYPTO_CBC
 	select CRYPTO_SHA1
 	select CRYPTO_DES
 	---help---
diff --git a/net/ipv6/esp6.c b/net/ipv6/esp6.c
index a278d5e862fe..46a7e687948e 100644
--- a/net/ipv6/esp6.c
+++ b/net/ipv6/esp6.c
@@ -24,6 +24,7 @@
  * 	This file is derived from net/ipv4/esp.c
  */
 
+#include <linux/err.h>
 #include <linux/module.h>
 #include <net/ip.h>
 #include <net/xfrm.h>
@@ -44,7 +45,8 @@ static int esp6_output(struct xfrm_state *x, struct sk_buff *skb)
 	int hdr_len;
 	struct ipv6hdr *top_iph;
 	struct ipv6_esp_hdr *esph;
-	struct crypto_tfm *tfm;
+	struct crypto_blkcipher *tfm;
+	struct blkcipher_desc desc;
 	struct esp_data *esp;
 	struct sk_buff *trailer;
 	int blksize;
@@ -67,7 +69,9 @@ static int esp6_output(struct xfrm_state *x, struct sk_buff *skb)
 
 	alen = esp->auth.icv_trunc_len;
 	tfm = esp->conf.tfm;
-	blksize = ALIGN(crypto_tfm_alg_blocksize(tfm), 4);
+	desc.tfm = tfm;
+	desc.flags = 0;
+	blksize = ALIGN(crypto_blkcipher_blocksize(tfm), 4);
 	clen = ALIGN(clen + 2, blksize);
 	if (esp->conf.padlen)
 		clen = ALIGN(clen, esp->conf.padlen);
@@ -96,7 +100,7 @@ static int esp6_output(struct xfrm_state *x, struct sk_buff *skb)
 	xfrm_aevent_doreplay(x);
 
 	if (esp->conf.ivlen)
-		crypto_cipher_set_iv(tfm, esp->conf.ivec, crypto_tfm_alg_ivsize(tfm));
+		crypto_blkcipher_set_iv(tfm, esp->conf.ivec, esp->conf.ivlen);
 
 	do {
 		struct scatterlist *sg = &esp->sgbuf[0];
@@ -107,14 +111,17 @@ static int esp6_output(struct xfrm_state *x, struct sk_buff *skb)
 				goto error;
 		}
 		skb_to_sgvec(skb, sg, esph->enc_data+esp->conf.ivlen-skb->data, clen);
-		crypto_cipher_encrypt(tfm, sg, sg, clen);
+		err = crypto_blkcipher_encrypt(&desc, sg, sg, clen);
 		if (unlikely(sg != &esp->sgbuf[0]))
 			kfree(sg);
 	} while (0);
 
+	if (unlikely(err))
+		goto error;
+
 	if (esp->conf.ivlen) {
-		memcpy(esph->enc_data, esp->conf.ivec, crypto_tfm_alg_ivsize(tfm));
-		crypto_cipher_get_iv(tfm, esp->conf.ivec, crypto_tfm_alg_ivsize(tfm));
+		memcpy(esph->enc_data, esp->conf.ivec, esp->conf.ivlen);
+		crypto_blkcipher_get_iv(tfm, esp->conf.ivec, esp->conf.ivlen);
 	}
 
 	if (esp->auth.icv_full_len) {
@@ -123,8 +130,6 @@ static int esp6_output(struct xfrm_state *x, struct sk_buff *skb)
 		pskb_put(skb, trailer, alen);
 	}
 
-	err = 0;
-
 error:
 	return err;
 }
@@ -134,8 +139,10 @@ static int esp6_input(struct xfrm_state *x, struct sk_buff *skb)
 	struct ipv6hdr *iph;
 	struct ipv6_esp_hdr *esph;
 	struct esp_data *esp = x->data;
+	struct crypto_blkcipher *tfm = esp->conf.tfm;
+	struct blkcipher_desc desc = { .tfm = tfm };
 	struct sk_buff *trailer;
-	int blksize = ALIGN(crypto_tfm_alg_blocksize(esp->conf.tfm), 4);
+	int blksize = ALIGN(crypto_blkcipher_blocksize(tfm), 4);
 	int alen = esp->auth.icv_trunc_len;
 	int elen = skb->len - sizeof(struct ipv6_esp_hdr) - esp->conf.ivlen - alen;
 
@@ -182,7 +189,7 @@ static int esp6_input(struct xfrm_state *x, struct sk_buff *skb)
 
 	/* Get ivec. This can be wrong, check against another impls. */
 	if (esp->conf.ivlen)
-		crypto_cipher_set_iv(esp->conf.tfm, esph->enc_data, crypto_tfm_alg_ivsize(esp->conf.tfm));
+		crypto_blkcipher_set_iv(tfm, esph->enc_data, esp->conf.ivlen);
 
         {
 		u8 nexthdr[2];
@@ -197,9 +204,11 @@ static int esp6_input(struct xfrm_state *x, struct sk_buff *skb)
 			}
 		}
 		skb_to_sgvec(skb, sg, sizeof(struct ipv6_esp_hdr) + esp->conf.ivlen, elen);
-		crypto_cipher_decrypt(esp->conf.tfm, sg, sg, elen);
+		ret = crypto_blkcipher_decrypt(&desc, sg, sg, elen);
 		if (unlikely(sg != &esp->sgbuf[0]))
 			kfree(sg);
+		if (unlikely(ret))
+			goto out;
 
 		if (skb_copy_bits(skb, skb->len-alen-2, nexthdr, 2))
 			BUG();
@@ -225,7 +234,7 @@ out:
 static u32 esp6_get_max_size(struct xfrm_state *x, int mtu)
 {
 	struct esp_data *esp = x->data;
-	u32 blksize = ALIGN(crypto_tfm_alg_blocksize(esp->conf.tfm), 4);
+	u32 blksize = ALIGN(crypto_blkcipher_blocksize(esp->conf.tfm), 4);
 
 	if (x->props.mode) {
 		mtu = ALIGN(mtu + 2, blksize);
@@ -266,7 +275,7 @@ static void esp6_destroy(struct xfrm_state *x)
 	if (!esp)
 		return;
 
-	crypto_free_tfm(esp->conf.tfm);
+	crypto_free_blkcipher(esp->conf.tfm);
 	esp->conf.tfm = NULL;
 	kfree(esp->conf.ivec);
 	esp->conf.ivec = NULL;
@@ -280,6 +289,7 @@ static void esp6_destroy(struct xfrm_state *x)
 static int esp6_init_state(struct xfrm_state *x)
 {
 	struct esp_data *esp = NULL;
+	struct crypto_blkcipher *tfm;
 
 	/* null auth and encryption can have zero length keys */
 	if (x->aalg) {
@@ -327,13 +337,11 @@ static int esp6_init_state(struct xfrm_state *x)
 	}
 	esp->conf.key = x->ealg->alg_key;
 	esp->conf.key_len = (x->ealg->alg_key_len+7)/8;
-	if (x->props.ealgo == SADB_EALG_NULL)
-		esp->conf.tfm = crypto_alloc_tfm(x->ealg->alg_name, CRYPTO_TFM_MODE_ECB);
-	else
-		esp->conf.tfm = crypto_alloc_tfm(x->ealg->alg_name, CRYPTO_TFM_MODE_CBC);
-	if (esp->conf.tfm == NULL)
+	tfm = crypto_alloc_blkcipher(x->ealg->alg_name, 0, CRYPTO_ALG_ASYNC);
+	if (IS_ERR(tfm))
 		goto error;
-	esp->conf.ivlen = crypto_tfm_alg_ivsize(esp->conf.tfm);
+	esp->conf.tfm = tfm;
+	esp->conf.ivlen = crypto_blkcipher_ivsize(tfm);
 	esp->conf.padlen = 0;
 	if (esp->conf.ivlen) {
 		esp->conf.ivec = kmalloc(esp->conf.ivlen, GFP_KERNEL);
@@ -341,7 +349,7 @@ static int esp6_init_state(struct xfrm_state *x)
 			goto error;
 		get_random_bytes(esp->conf.ivec, esp->conf.ivlen);
 	}
-	if (crypto_cipher_setkey(esp->conf.tfm, esp->conf.key, esp->conf.key_len))
+	if (crypto_blkcipher_setkey(tfm, esp->conf.key, esp->conf.key_len))
 		goto error;
 	x->props.header_len = sizeof(struct ipv6_esp_hdr) + esp->conf.ivlen;
 	if (x->props.mode)
diff --git a/net/xfrm/xfrm_algo.c b/net/xfrm/xfrm_algo.c
index b68974b38741..9b03d8497fba 100644
--- a/net/xfrm/xfrm_algo.c
+++ b/net/xfrm/xfrm_algo.c
@@ -118,7 +118,8 @@ static struct xfrm_algo_desc aalg_list[] = {
 
 static struct xfrm_algo_desc ealg_list[] = {
 {
-	.name = "cipher_null",
+	.name = "ecb(cipher_null)",
+	.compat = "cipher_null",
 	
 	.uinfo = {
 		.encr = {
@@ -135,7 +136,8 @@ static struct xfrm_algo_desc ealg_list[] = {
 	}
 },
 {
-	.name = "des",
+	.name = "cbc(des)",
+	.compat = "des",
 
 	.uinfo = {
 		.encr = {
@@ -152,7 +154,8 @@ static struct xfrm_algo_desc ealg_list[] = {
 	}
 },
 {
-	.name = "des3_ede",
+	.name = "cbc(des3_ede)",
+	.compat = "des3_ede",
 
 	.uinfo = {
 		.encr = {
@@ -169,7 +172,8 @@ static struct xfrm_algo_desc ealg_list[] = {
 	}
 },
 {
-	.name = "cast128",
+	.name = "cbc(cast128)",
+	.compat = "cast128",
 
 	.uinfo = {
 		.encr = {
@@ -186,7 +190,8 @@ static struct xfrm_algo_desc ealg_list[] = {
 	}
 },
 {
-	.name = "blowfish",
+	.name = "cbc(blowfish)",
+	.compat = "blowfish",
 
 	.uinfo = {
 		.encr = {
@@ -203,7 +208,8 @@ static struct xfrm_algo_desc ealg_list[] = {
 	}
 },
 {
-	.name = "aes",
+	.name = "cbc(aes)",
+	.compat = "aes",
 
 	.uinfo = {
 		.encr = {
@@ -220,7 +226,8 @@ static struct xfrm_algo_desc ealg_list[] = {
 	}
 },
 {
-        .name = "serpent",
+        .name = "cbc(serpent)",
+        .compat = "serpent",
 
         .uinfo = {
                 .encr = {
@@ -237,7 +244,8 @@ static struct xfrm_algo_desc ealg_list[] = {
         }
 },
 {
-        .name = "twofish",
+        .name = "cbc(twofish)",
+        .compat = "twofish",
                  
         .uinfo = {
                 .encr = {
-- 
cgit v1.2.3


From 378c6697a282c383d89428380a3405bf95189347 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Tue, 22 Aug 2006 20:33:54 +1000
Subject: [SUNRPC] GSS: Use block ciphers where applicable

This patch converts SUNRPC/GSS to use the new block cipher type where
applicable.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 include/linux/sunrpc/gss_krb5.h       | 19 ++++++------
 include/linux/sunrpc/gss_spkm3.h      |  4 +--
 net/sunrpc/auth_gss/gss_krb5_crypto.c | 57 ++++++++++++++++++++---------------
 net/sunrpc/auth_gss/gss_krb5_mech.c   | 24 ++++++++-------
 net/sunrpc/auth_gss/gss_krb5_seqnum.c |  4 +--
 net/sunrpc/auth_gss/gss_krb5_wrap.c   |  4 +--
 net/sunrpc/auth_gss/gss_spkm3_mech.c  | 29 +++++++++---------
 7 files changed, 76 insertions(+), 65 deletions(-)

(limited to 'include')

diff --git a/include/linux/sunrpc/gss_krb5.h b/include/linux/sunrpc/gss_krb5.h
index 1279280d7196..e30ba201910a 100644
--- a/include/linux/sunrpc/gss_krb5.h
+++ b/include/linux/sunrpc/gss_krb5.h
@@ -46,8 +46,8 @@ struct krb5_ctx {
 	unsigned char		seed[16];
 	int			signalg;
 	int			sealalg;
-	struct crypto_tfm	*enc;
-	struct crypto_tfm	*seq;
+	struct crypto_blkcipher	*enc;
+	struct crypto_blkcipher	*seq;
 	s32			endtime;
 	u32			seq_send;
 	struct xdr_netobj	mech_used;
@@ -136,26 +136,27 @@ gss_unwrap_kerberos(struct gss_ctx *ctx_id, int offset,
 
 
 u32
-krb5_encrypt(struct crypto_tfm * key,
+krb5_encrypt(struct crypto_blkcipher *key,
 	     void *iv, void *in, void *out, int length);
 
 u32
-krb5_decrypt(struct crypto_tfm * key,
+krb5_decrypt(struct crypto_blkcipher *key,
 	     void *iv, void *in, void *out, int length); 
 
 int
-gss_encrypt_xdr_buf(struct crypto_tfm *tfm, struct xdr_buf *outbuf, int offset,
-		struct page **pages);
+gss_encrypt_xdr_buf(struct crypto_blkcipher *tfm, struct xdr_buf *outbuf,
+		    int offset, struct page **pages);
 
 int
-gss_decrypt_xdr_buf(struct crypto_tfm *tfm, struct xdr_buf *inbuf, int offset);
+gss_decrypt_xdr_buf(struct crypto_blkcipher *tfm, struct xdr_buf *inbuf,
+		    int offset);
 
 s32
-krb5_make_seq_num(struct crypto_tfm * key,
+krb5_make_seq_num(struct crypto_blkcipher *key,
 		int direction,
 		s32 seqnum, unsigned char *cksum, unsigned char *buf);
 
 s32
-krb5_get_seq_num(struct crypto_tfm * key,
+krb5_get_seq_num(struct crypto_blkcipher *key,
 	       unsigned char *cksum,
 	       unsigned char *buf, int *direction, s32 * seqnum);
diff --git a/include/linux/sunrpc/gss_spkm3.h b/include/linux/sunrpc/gss_spkm3.h
index 336e218c2782..2cf3fbb40b4f 100644
--- a/include/linux/sunrpc/gss_spkm3.h
+++ b/include/linux/sunrpc/gss_spkm3.h
@@ -19,9 +19,9 @@ struct spkm3_ctx {
 	unsigned int		req_flags ;
 	struct xdr_netobj	share_key;
 	int			conf_alg;
-	struct crypto_tfm*	derived_conf_key;
+	struct crypto_blkcipher	*derived_conf_key;
 	int			intg_alg;
-	struct crypto_tfm*	derived_integ_key;
+	struct crypto_blkcipher	*derived_integ_key;
 	int			keyestb_alg;   /* alg used to get share_key */
 	int			owf_alg;   /* one way function */
 };
diff --git a/net/sunrpc/auth_gss/gss_krb5_crypto.c b/net/sunrpc/auth_gss/gss_krb5_crypto.c
index 76b969e6904f..57192dfe3065 100644
--- a/net/sunrpc/auth_gss/gss_krb5_crypto.c
+++ b/net/sunrpc/auth_gss/gss_krb5_crypto.c
@@ -49,7 +49,7 @@
 
 u32
 krb5_encrypt(
-	struct crypto_tfm *tfm,
+	struct crypto_blkcipher *tfm,
 	void * iv,
 	void * in,
 	void * out,
@@ -58,26 +58,27 @@ krb5_encrypt(
 	u32 ret = -EINVAL;
         struct scatterlist sg[1];
 	u8 local_iv[16] = {0};
+	struct blkcipher_desc desc = { .tfm = tfm, .info = local_iv };
 
 	dprintk("RPC:      krb5_encrypt: input data:\n");
 	print_hexl((u32 *)in, length, 0);
 
-	if (length % crypto_tfm_alg_blocksize(tfm) != 0)
+	if (length % crypto_blkcipher_blocksize(tfm) != 0)
 		goto out;
 
-	if (crypto_tfm_alg_ivsize(tfm) > 16) {
+	if (crypto_blkcipher_ivsize(tfm) > 16) {
 		dprintk("RPC:      gss_k5encrypt: tfm iv size to large %d\n",
-		         crypto_tfm_alg_ivsize(tfm));
+		         crypto_blkcipher_ivsize(tfm));
 		goto out;
 	}
 
 	if (iv)
-		memcpy(local_iv, iv, crypto_tfm_alg_ivsize(tfm));
+		memcpy(local_iv, iv, crypto_blkcipher_ivsize(tfm));
 
 	memcpy(out, in, length);
 	sg_set_buf(sg, out, length);
 
-	ret = crypto_cipher_encrypt_iv(tfm, sg, sg, length, local_iv);
+	ret = crypto_blkcipher_encrypt_iv(&desc, sg, sg, length);
 
 	dprintk("RPC:      krb5_encrypt: output data:\n");
 	print_hexl((u32 *)out, length, 0);
@@ -90,7 +91,7 @@ EXPORT_SYMBOL(krb5_encrypt);
 
 u32
 krb5_decrypt(
-     struct crypto_tfm *tfm,
+     struct crypto_blkcipher *tfm,
      void * iv,
      void * in,
      void * out,
@@ -99,25 +100,26 @@ krb5_decrypt(
 	u32 ret = -EINVAL;
 	struct scatterlist sg[1];
 	u8 local_iv[16] = {0};
+	struct blkcipher_desc desc = { .tfm = tfm, .info = local_iv };
 
 	dprintk("RPC:      krb5_decrypt: input data:\n");
 	print_hexl((u32 *)in, length, 0);
 
-	if (length % crypto_tfm_alg_blocksize(tfm) != 0)
+	if (length % crypto_blkcipher_blocksize(tfm) != 0)
 		goto out;
 
-	if (crypto_tfm_alg_ivsize(tfm) > 16) {
+	if (crypto_blkcipher_ivsize(tfm) > 16) {
 		dprintk("RPC:      gss_k5decrypt: tfm iv size to large %d\n",
-			crypto_tfm_alg_ivsize(tfm));
+			crypto_blkcipher_ivsize(tfm));
 		goto out;
 	}
 	if (iv)
-		memcpy(local_iv,iv, crypto_tfm_alg_ivsize(tfm));
+		memcpy(local_iv,iv, crypto_blkcipher_ivsize(tfm));
 
 	memcpy(out, in, length);
 	sg_set_buf(sg, out, length);
 
-	ret = crypto_cipher_decrypt_iv(tfm, sg, sg, length, local_iv);
+	ret = crypto_blkcipher_decrypt_iv(&desc, sg, sg, length);
 
 	dprintk("RPC:      krb5_decrypt: output_data:\n");
 	print_hexl((u32 *)out, length, 0);
@@ -240,7 +242,7 @@ EXPORT_SYMBOL(make_checksum);
 
 struct encryptor_desc {
 	u8 iv[8]; /* XXX hard-coded blocksize */
-	struct crypto_tfm *tfm;
+	struct blkcipher_desc desc;
 	int pos;
 	struct xdr_buf *outbuf;
 	struct page **pages;
@@ -285,8 +287,8 @@ encryptor(struct scatterlist *sg, void *data)
 	if (thislen == 0)
 		return 0;
 
-	ret = crypto_cipher_encrypt_iv(desc->tfm, desc->outfrags, desc->infrags,
-					thislen, desc->iv);
+	ret = crypto_blkcipher_encrypt_iv(&desc->desc, desc->outfrags,
+					  desc->infrags, thislen);
 	if (ret)
 		return ret;
 	if (fraglen) {
@@ -305,16 +307,18 @@ encryptor(struct scatterlist *sg, void *data)
 }
 
 int
-gss_encrypt_xdr_buf(struct crypto_tfm *tfm, struct xdr_buf *buf, int offset,
-		struct page **pages)
+gss_encrypt_xdr_buf(struct crypto_blkcipher *tfm, struct xdr_buf *buf,
+		    int offset, struct page **pages)
 {
 	int ret;
 	struct encryptor_desc desc;
 
-	BUG_ON((buf->len - offset) % crypto_tfm_alg_blocksize(tfm) != 0);
+	BUG_ON((buf->len - offset) % crypto_blkcipher_blocksize(tfm) != 0);
 
 	memset(desc.iv, 0, sizeof(desc.iv));
-	desc.tfm = tfm;
+	desc.desc.tfm = tfm;
+	desc.desc.info = desc.iv;
+	desc.desc.flags = 0;
 	desc.pos = offset;
 	desc.outbuf = buf;
 	desc.pages = pages;
@@ -329,7 +333,7 @@ EXPORT_SYMBOL(gss_encrypt_xdr_buf);
 
 struct decryptor_desc {
 	u8 iv[8]; /* XXX hard-coded blocksize */
-	struct crypto_tfm *tfm;
+	struct blkcipher_desc desc;
 	struct scatterlist frags[4];
 	int fragno;
 	int fraglen;
@@ -355,8 +359,8 @@ decryptor(struct scatterlist *sg, void *data)
 	if (thislen == 0)
 		return 0;
 
-	ret = crypto_cipher_decrypt_iv(desc->tfm, desc->frags, desc->frags,
-					thislen, desc->iv);
+	ret = crypto_blkcipher_decrypt_iv(&desc->desc, desc->frags,
+					  desc->frags, thislen);
 	if (ret)
 		return ret;
 	if (fraglen) {
@@ -373,15 +377,18 @@ decryptor(struct scatterlist *sg, void *data)
 }
 
 int
-gss_decrypt_xdr_buf(struct crypto_tfm *tfm, struct xdr_buf *buf, int offset)
+gss_decrypt_xdr_buf(struct crypto_blkcipher *tfm, struct xdr_buf *buf,
+		    int offset)
 {
 	struct decryptor_desc desc;
 
 	/* XXXJBF: */
-	BUG_ON((buf->len - offset) % crypto_tfm_alg_blocksize(tfm) != 0);
+	BUG_ON((buf->len - offset) % crypto_blkcipher_blocksize(tfm) != 0);
 
 	memset(desc.iv, 0, sizeof(desc.iv));
-	desc.tfm = tfm;
+	desc.desc.tfm = tfm;
+	desc.desc.info = desc.iv;
+	desc.desc.flags = 0;
 	desc.fragno = 0;
 	desc.fraglen = 0;
 	return process_xdr_buf(buf, offset, buf->len - offset, decryptor, &desc);
diff --git a/net/sunrpc/auth_gss/gss_krb5_mech.c b/net/sunrpc/auth_gss/gss_krb5_mech.c
index 70e1e53a632b..325e72e4fd31 100644
--- a/net/sunrpc/auth_gss/gss_krb5_mech.c
+++ b/net/sunrpc/auth_gss/gss_krb5_mech.c
@@ -34,6 +34,7 @@
  *
  */
 
+#include <linux/err.h>
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/types.h>
@@ -78,10 +79,10 @@ simple_get_netobj(const void *p, const void *end, struct xdr_netobj *res)
 }
 
 static inline const void *
-get_key(const void *p, const void *end, struct crypto_tfm **res)
+get_key(const void *p, const void *end, struct crypto_blkcipher **res)
 {
 	struct xdr_netobj	key;
-	int			alg, alg_mode;
+	int			alg;
 	char			*alg_name;
 
 	p = simple_get_bytes(p, end, &alg, sizeof(alg));
@@ -93,18 +94,19 @@ get_key(const void *p, const void *end, struct crypto_tfm **res)
 
 	switch (alg) {
 		case ENCTYPE_DES_CBC_RAW:
-			alg_name = "des";
-			alg_mode = CRYPTO_TFM_MODE_CBC;
+			alg_name = "cbc(des)";
 			break;
 		default:
 			printk("gss_kerberos_mech: unsupported algorithm %d\n", alg);
 			goto out_err_free_key;
 	}
-	if (!(*res = crypto_alloc_tfm(alg_name, alg_mode))) {
+	*res = crypto_alloc_blkcipher(alg_name, 0, CRYPTO_ALG_ASYNC);
+	if (IS_ERR(*res)) {
 		printk("gss_kerberos_mech: unable to initialize crypto algorithm %s\n", alg_name);
+		*res = NULL;
 		goto out_err_free_key;
 	}
-	if (crypto_cipher_setkey(*res, key.data, key.len)) {
+	if (crypto_blkcipher_setkey(*res, key.data, key.len)) {
 		printk("gss_kerberos_mech: error setting key for crypto algorithm %s\n", alg_name);
 		goto out_err_free_tfm;
 	}
@@ -113,7 +115,7 @@ get_key(const void *p, const void *end, struct crypto_tfm **res)
 	return p;
 
 out_err_free_tfm:
-	crypto_free_tfm(*res);
+	crypto_free_blkcipher(*res);
 out_err_free_key:
 	kfree(key.data);
 	p = ERR_PTR(-EINVAL);
@@ -172,9 +174,9 @@ gss_import_sec_context_kerberos(const void *p,
 	return 0;
 
 out_err_free_key2:
-	crypto_free_tfm(ctx->seq);
+	crypto_free_blkcipher(ctx->seq);
 out_err_free_key1:
-	crypto_free_tfm(ctx->enc);
+	crypto_free_blkcipher(ctx->enc);
 out_err_free_mech:
 	kfree(ctx->mech_used.data);
 out_err_free_ctx:
@@ -187,8 +189,8 @@ static void
 gss_delete_sec_context_kerberos(void *internal_ctx) {
 	struct krb5_ctx *kctx = internal_ctx;
 
-	crypto_free_tfm(kctx->seq);
-	crypto_free_tfm(kctx->enc);
+	crypto_free_blkcipher(kctx->seq);
+	crypto_free_blkcipher(kctx->enc);
 	kfree(kctx->mech_used.data);
 	kfree(kctx);
 }
diff --git a/net/sunrpc/auth_gss/gss_krb5_seqnum.c b/net/sunrpc/auth_gss/gss_krb5_seqnum.c
index c53ead39118d..c604baf3a5f6 100644
--- a/net/sunrpc/auth_gss/gss_krb5_seqnum.c
+++ b/net/sunrpc/auth_gss/gss_krb5_seqnum.c
@@ -41,7 +41,7 @@
 #endif
 
 s32
-krb5_make_seq_num(struct crypto_tfm *key,
+krb5_make_seq_num(struct crypto_blkcipher *key,
 		int direction,
 		s32 seqnum,
 		unsigned char *cksum, unsigned char *buf)
@@ -62,7 +62,7 @@ krb5_make_seq_num(struct crypto_tfm *key,
 }
 
 s32
-krb5_get_seq_num(struct crypto_tfm *key,
+krb5_get_seq_num(struct crypto_blkcipher *key,
 	       unsigned char *cksum,
 	       unsigned char *buf,
 	       int *direction, s32 * seqnum)
diff --git a/net/sunrpc/auth_gss/gss_krb5_wrap.c b/net/sunrpc/auth_gss/gss_krb5_wrap.c
index 89d1f3e14128..f179415d0c38 100644
--- a/net/sunrpc/auth_gss/gss_krb5_wrap.c
+++ b/net/sunrpc/auth_gss/gss_krb5_wrap.c
@@ -149,7 +149,7 @@ gss_wrap_kerberos(struct gss_ctx *ctx, int offset,
 		goto out_err;
 	}
 
-	blocksize = crypto_tfm_alg_blocksize(kctx->enc);
+	blocksize = crypto_blkcipher_blocksize(kctx->enc);
 	gss_krb5_add_padding(buf, offset, blocksize);
 	BUG_ON((buf->len - offset) % blocksize);
 	plainlen = blocksize + buf->len - offset;
@@ -346,7 +346,7 @@ gss_unwrap_kerberos(struct gss_ctx *ctx, int offset, struct xdr_buf *buf)
 	/* Copy the data back to the right position.  XXX: Would probably be
 	 * better to copy and encrypt at the same time. */
 
-	blocksize = crypto_tfm_alg_blocksize(kctx->enc);
+	blocksize = crypto_blkcipher_blocksize(kctx->enc);
 	data_start = ptr + 22 + blocksize;
 	orig_start = buf->head[0].iov_base + offset;
 	data_len = (buf->head[0].iov_base + buf->head[0].iov_len) - data_start;
diff --git a/net/sunrpc/auth_gss/gss_spkm3_mech.c b/net/sunrpc/auth_gss/gss_spkm3_mech.c
index 88dcb52d171b..bdedf456bc17 100644
--- a/net/sunrpc/auth_gss/gss_spkm3_mech.c
+++ b/net/sunrpc/auth_gss/gss_spkm3_mech.c
@@ -34,6 +34,7 @@
  *
  */
 
+#include <linux/err.h>
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/types.h>
@@ -83,10 +84,11 @@ simple_get_netobj(const void *p, const void *end, struct xdr_netobj *res)
 }
 
 static inline const void *
-get_key(const void *p, const void *end, struct crypto_tfm **res, int *resalg)
+get_key(const void *p, const void *end, struct crypto_blkcipher **res,
+	int *resalg)
 {
 	struct xdr_netobj	key = { 0 };
-	int			alg_mode,setkey = 0;
+	int			setkey = 0;
 	char			*alg_name;
 
 	p = simple_get_bytes(p, end, resalg, sizeof(*resalg));
@@ -98,14 +100,12 @@ get_key(const void *p, const void *end, struct crypto_tfm **res, int *resalg)
 
 	switch (*resalg) {
 		case NID_des_cbc:
-			alg_name = "des";
-			alg_mode = CRYPTO_TFM_MODE_CBC;
+			alg_name = "cbc(des)";
 			setkey = 1;
 			break;
 		case NID_cast5_cbc:
 			/* XXXX here in name only, not used */
-			alg_name = "cast5";
-			alg_mode = CRYPTO_TFM_MODE_CBC;
+			alg_name = "cbc(cast5)";
 			setkey = 0; /* XXX will need to set to 1 */
 			break;
 		case NID_md5:
@@ -113,19 +113,20 @@ get_key(const void *p, const void *end, struct crypto_tfm **res, int *resalg)
 				dprintk("RPC: SPKM3 get_key: NID_md5 zero Key length\n");
 			}
 			alg_name = "md5";
-			alg_mode = 0;
 			setkey = 0;
 			break;
 		default:
 			dprintk("gss_spkm3_mech: unsupported algorithm %d\n", *resalg);
 			goto out_err_free_key;
 	}
-	if (!(*res = crypto_alloc_tfm(alg_name, alg_mode))) {
+	*res = crypto_alloc_blkcipher(alg_name, 0, CRYPTO_ALG_ASYNC);
+	if (IS_ERR(*res)) {
 		printk("gss_spkm3_mech: unable to initialize crypto algorthm %s\n", alg_name);
+		*res = NULL;
 		goto out_err_free_key;
 	}
 	if (setkey) {
-		if (crypto_cipher_setkey(*res, key.data, key.len)) {
+		if (crypto_blkcipher_setkey(*res, key.data, key.len)) {
 			printk("gss_spkm3_mech: error setting key for crypto algorthm %s\n", alg_name);
 			goto out_err_free_tfm;
 		}
@@ -136,7 +137,7 @@ get_key(const void *p, const void *end, struct crypto_tfm **res, int *resalg)
 	return p;
 
 out_err_free_tfm:
-	crypto_free_tfm(*res);
+	crypto_free_blkcipher(*res);
 out_err_free_key:
 	if(key.len > 0)
 		kfree(key.data);
@@ -204,9 +205,9 @@ gss_import_sec_context_spkm3(const void *p, size_t len,
 	return 0;
 
 out_err_free_key2:
-	crypto_free_tfm(ctx->derived_integ_key);
+	crypto_free_blkcipher(ctx->derived_integ_key);
 out_err_free_key1:
-	crypto_free_tfm(ctx->derived_conf_key);
+	crypto_free_blkcipher(ctx->derived_conf_key);
 out_err_free_s_key:
 	kfree(ctx->share_key.data);
 out_err_free_mech:
@@ -223,8 +224,8 @@ static void
 gss_delete_sec_context_spkm3(void *internal_ctx) {
 	struct spkm3_ctx *sctx = internal_ctx;
 
-	crypto_free_tfm(sctx->derived_integ_key);
-	crypto_free_tfm(sctx->derived_conf_key);
+	crypto_free_blkcipher(sctx->derived_integ_key);
+	crypto_free_blkcipher(sctx->derived_conf_key);
 	kfree(sctx->share_key.data);
 	kfree(sctx->mech_used.data);
 	kfree(sctx);
-- 
cgit v1.2.3


From 03fd9cee7f46dddcd2562bc175d2c348502ce281 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Mon, 14 Aug 2006 23:11:53 +1000
Subject: [PATCH] scatterlist: Add const to sg_set_buf/sg_init_one pointer
 argument

This patch adds a const modifier to the buf argument of sg_set_buf and
sg_init_one.  This lets people call it with pointers that are const.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 include/linux/scatterlist.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/include/linux/scatterlist.h b/include/linux/scatterlist.h
index 66ff545552f7..4efbd9c445f5 100644
--- a/include/linux/scatterlist.h
+++ b/include/linux/scatterlist.h
@@ -5,7 +5,7 @@
 #include <linux/mm.h>
 #include <linux/string.h>
 
-static inline void sg_set_buf(struct scatterlist *sg, void *buf,
+static inline void sg_set_buf(struct scatterlist *sg, const void *buf,
 			      unsigned int buflen)
 {
 	sg->page = virt_to_page(buf);
@@ -13,7 +13,7 @@ static inline void sg_set_buf(struct scatterlist *sg, void *buf,
 	sg->length = buflen;
 }
 
-static inline void sg_init_one(struct scatterlist *sg, void *buf,
+static inline void sg_init_one(struct scatterlist *sg, const void *buf,
 			       unsigned int buflen)
 {
 	memset(sg, 0, sizeof(*sg));
-- 
cgit v1.2.3


From 7226bc877a22244e8003924031435a4bffd52654 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Mon, 21 Aug 2006 21:40:49 +1000
Subject: [CRYPTO] api: Mark parts of cipher interface as deprecated

Mark the parts of the cipher interface that have been replaced by
block ciphers as deprecated.  Thanks to Andrew Morton for suggesting
doing this before removing them completely.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 crypto/cipher.c        | 34 ++++++++++++++++++++++++++++------
 include/linux/crypto.h | 48 ++++++++++++++++++++++++++++++++++++------------
 2 files changed, 64 insertions(+), 18 deletions(-)

(limited to 'include')

diff --git a/crypto/cipher.c b/crypto/cipher.c
index 326461780673..9e03701cfdcc 100644
--- a/crypto/cipher.c
+++ b/crypto/cipher.c
@@ -23,6 +23,28 @@
 #include "internal.h"
 #include "scatterwalk.h"
 
+struct cipher_alg_compat {
+	unsigned int cia_min_keysize;
+	unsigned int cia_max_keysize;
+	int (*cia_setkey)(struct crypto_tfm *tfm, const u8 *key,
+	                  unsigned int keylen);
+	void (*cia_encrypt)(struct crypto_tfm *tfm, u8 *dst, const u8 *src);
+	void (*cia_decrypt)(struct crypto_tfm *tfm, u8 *dst, const u8 *src);
+
+	unsigned int (*cia_encrypt_ecb)(const struct cipher_desc *desc,
+					u8 *dst, const u8 *src,
+					unsigned int nbytes);
+	unsigned int (*cia_decrypt_ecb)(const struct cipher_desc *desc,
+					u8 *dst, const u8 *src,
+					unsigned int nbytes);
+	unsigned int (*cia_encrypt_cbc)(const struct cipher_desc *desc,
+					u8 *dst, const u8 *src,
+					unsigned int nbytes);
+	unsigned int (*cia_decrypt_cbc)(const struct cipher_desc *desc,
+					u8 *dst, const u8 *src,
+					unsigned int nbytes);
+};
+
 static inline void xor_64(u8 *a, const u8 *b)
 {
 	((u32 *)a)[0] ^= ((u32 *)b)[0];
@@ -276,7 +298,7 @@ static int ecb_encrypt(struct crypto_tfm *tfm,
                        struct scatterlist *src, unsigned int nbytes)
 {
 	struct cipher_desc desc;
-	struct cipher_alg *cipher = &tfm->__crt_alg->cra_cipher;
+	struct cipher_alg_compat *cipher = (void *)&tfm->__crt_alg->cra_cipher;
 
 	desc.tfm = tfm;
 	desc.crfn = cipher->cia_encrypt;
@@ -291,7 +313,7 @@ static int ecb_decrypt(struct crypto_tfm *tfm,
 		       unsigned int nbytes)
 {
 	struct cipher_desc desc;
-	struct cipher_alg *cipher = &tfm->__crt_alg->cra_cipher;
+	struct cipher_alg_compat *cipher = (void *)&tfm->__crt_alg->cra_cipher;
 
 	desc.tfm = tfm;
 	desc.crfn = cipher->cia_decrypt;
@@ -306,7 +328,7 @@ static int cbc_encrypt(struct crypto_tfm *tfm,
 		       unsigned int nbytes)
 {
 	struct cipher_desc desc;
-	struct cipher_alg *cipher = &tfm->__crt_alg->cra_cipher;
+	struct cipher_alg_compat *cipher = (void *)&tfm->__crt_alg->cra_cipher;
 
 	desc.tfm = tfm;
 	desc.crfn = cipher->cia_encrypt;
@@ -322,7 +344,7 @@ static int cbc_encrypt_iv(struct crypto_tfm *tfm,
                           unsigned int nbytes, u8 *iv)
 {
 	struct cipher_desc desc;
-	struct cipher_alg *cipher = &tfm->__crt_alg->cra_cipher;
+	struct cipher_alg_compat *cipher = (void *)&tfm->__crt_alg->cra_cipher;
 
 	desc.tfm = tfm;
 	desc.crfn = cipher->cia_encrypt;
@@ -338,7 +360,7 @@ static int cbc_decrypt(struct crypto_tfm *tfm,
 		       unsigned int nbytes)
 {
 	struct cipher_desc desc;
-	struct cipher_alg *cipher = &tfm->__crt_alg->cra_cipher;
+	struct cipher_alg_compat *cipher = (void *)&tfm->__crt_alg->cra_cipher;
 
 	desc.tfm = tfm;
 	desc.crfn = cipher->cia_decrypt;
@@ -354,7 +376,7 @@ static int cbc_decrypt_iv(struct crypto_tfm *tfm,
                           unsigned int nbytes, u8 *iv)
 {
 	struct cipher_desc desc;
-	struct cipher_alg *cipher = &tfm->__crt_alg->cra_cipher;
+	struct cipher_alg_compat *cipher = (void *)&tfm->__crt_alg->cra_cipher;
 
 	desc.tfm = tfm;
 	desc.crfn = cipher->cia_decrypt;
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index 5a5466d518e8..0be666b50463 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -20,7 +20,6 @@
 #include <asm/atomic.h>
 #include <linux/module.h>
 #include <linux/kernel.h>
-#include <linux/types.h>
 #include <linux/list.h>
 #include <linux/slab.h>
 #include <linux/string.h>
@@ -137,16 +136,16 @@ struct cipher_alg {
 
 	unsigned int (*cia_encrypt_ecb)(const struct cipher_desc *desc,
 					u8 *dst, const u8 *src,
-					unsigned int nbytes);
+					unsigned int nbytes) __deprecated;
 	unsigned int (*cia_decrypt_ecb)(const struct cipher_desc *desc,
 					u8 *dst, const u8 *src,
-					unsigned int nbytes);
+					unsigned int nbytes) __deprecated;
 	unsigned int (*cia_encrypt_cbc)(const struct cipher_desc *desc,
 					u8 *dst, const u8 *src,
-					unsigned int nbytes);
+					unsigned int nbytes) __deprecated;
 	unsigned int (*cia_decrypt_cbc)(const struct cipher_desc *desc,
 					u8 *dst, const u8 *src,
-					unsigned int nbytes);
+					unsigned int nbytes) __deprecated;
 };
 
 struct digest_alg {
@@ -358,18 +357,23 @@ static inline u32 crypto_tfm_alg_type(struct crypto_tfm *tfm)
 	return tfm->__crt_alg->cra_flags & CRYPTO_ALG_TYPE_MASK;
 }
 
+static unsigned int crypto_tfm_alg_min_keysize(struct crypto_tfm *tfm)
+	__deprecated;
 static inline unsigned int crypto_tfm_alg_min_keysize(struct crypto_tfm *tfm)
 {
 	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_CIPHER);
 	return tfm->__crt_alg->cra_cipher.cia_min_keysize;
 }
 
+static unsigned int crypto_tfm_alg_max_keysize(struct crypto_tfm *tfm)
+	__deprecated;
 static inline unsigned int crypto_tfm_alg_max_keysize(struct crypto_tfm *tfm)
 {
 	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_CIPHER);
 	return tfm->__crt_alg->cra_cipher.cia_max_keysize;
 }
 
+static unsigned int crypto_tfm_alg_ivsize(struct crypto_tfm *tfm) __deprecated;
 static inline unsigned int crypto_tfm_alg_ivsize(struct crypto_tfm *tfm)
 {
 	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_CIPHER);
@@ -622,6 +626,13 @@ static inline void crypto_cipher_clear_flags(struct crypto_cipher *tfm,
 	crypto_tfm_clear_flags(crypto_cipher_tfm(tfm), flags);
 }
 
+static inline int crypto_cipher_setkey(struct crypto_cipher *tfm,
+                                       const u8 *key, unsigned int keylen)
+{
+	return crypto_cipher_crt(tfm)->cit_setkey(crypto_cipher_tfm(tfm),
+						  key, keylen);
+}
+
 static inline void crypto_cipher_encrypt_one(struct crypto_cipher *tfm,
 					     u8 *dst, const u8 *src)
 {
@@ -671,13 +682,10 @@ static inline int crypto_digest_setkey(struct crypto_tfm *tfm,
 	return tfm->crt_digest.dit_setkey(tfm, key, keylen);
 }
 
-static inline int crypto_cipher_setkey(struct crypto_tfm *tfm,
-                                       const u8 *key, unsigned int keylen)
-{
-	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_CIPHER);
-	return tfm->crt_cipher.cit_setkey(tfm, key, keylen);
-}
-
+static int crypto_cipher_encrypt(struct crypto_tfm *tfm,
+				 struct scatterlist *dst,
+				 struct scatterlist *src,
+				 unsigned int nbytes) __deprecated;
 static inline int crypto_cipher_encrypt(struct crypto_tfm *tfm,
                                         struct scatterlist *dst,
                                         struct scatterlist *src,
@@ -687,6 +695,10 @@ static inline int crypto_cipher_encrypt(struct crypto_tfm *tfm,
 	return tfm->crt_cipher.cit_encrypt(tfm, dst, src, nbytes);
 }                                        
 
+static int crypto_cipher_encrypt_iv(struct crypto_tfm *tfm,
+				    struct scatterlist *dst,
+				    struct scatterlist *src,
+				    unsigned int nbytes, u8 *iv) __deprecated;
 static inline int crypto_cipher_encrypt_iv(struct crypto_tfm *tfm,
                                            struct scatterlist *dst,
                                            struct scatterlist *src,
@@ -696,6 +708,10 @@ static inline int crypto_cipher_encrypt_iv(struct crypto_tfm *tfm,
 	return tfm->crt_cipher.cit_encrypt_iv(tfm, dst, src, nbytes, iv);
 }                                        
 
+static int crypto_cipher_decrypt(struct crypto_tfm *tfm,
+				 struct scatterlist *dst,
+				 struct scatterlist *src,
+				 unsigned int nbytes) __deprecated;
 static inline int crypto_cipher_decrypt(struct crypto_tfm *tfm,
                                         struct scatterlist *dst,
                                         struct scatterlist *src,
@@ -705,6 +721,10 @@ static inline int crypto_cipher_decrypt(struct crypto_tfm *tfm,
 	return tfm->crt_cipher.cit_decrypt(tfm, dst, src, nbytes);
 }
 
+static int crypto_cipher_decrypt_iv(struct crypto_tfm *tfm,
+				    struct scatterlist *dst,
+				    struct scatterlist *src,
+				    unsigned int nbytes, u8 *iv) __deprecated;
 static inline int crypto_cipher_decrypt_iv(struct crypto_tfm *tfm,
                                            struct scatterlist *dst,
                                            struct scatterlist *src,
@@ -714,6 +734,8 @@ static inline int crypto_cipher_decrypt_iv(struct crypto_tfm *tfm,
 	return tfm->crt_cipher.cit_decrypt_iv(tfm, dst, src, nbytes, iv);
 }
 
+static void crypto_cipher_set_iv(struct crypto_tfm *tfm,
+				 const u8 *src, unsigned int len) __deprecated;
 static inline void crypto_cipher_set_iv(struct crypto_tfm *tfm,
                                         const u8 *src, unsigned int len)
 {
@@ -721,6 +743,8 @@ static inline void crypto_cipher_set_iv(struct crypto_tfm *tfm,
 	memcpy(tfm->crt_cipher.cit_iv, src, len);
 }
 
+static void crypto_cipher_get_iv(struct crypto_tfm *tfm,
+				 u8 *dst, unsigned int len) __deprecated;
 static inline void crypto_cipher_get_iv(struct crypto_tfm *tfm,
                                         u8 *dst, unsigned int len)
 {
-- 
cgit v1.2.3


From 055bcee3102dc35f019b69df9c2618e9d6dd1c09 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sat, 19 Aug 2006 22:24:23 +1000
Subject: [CRYPTO] digest: Added user API for new hash type

The existing digest user interface is inadequate for support asynchronous
operations.  For one it doesn't return a value to indicate success or
failure, nor does it take a per-operation descriptor which is essential
for the issuing of requests while other requests are still outstanding.

This patch is the first in a series of steps to remodel the interface
for asynchronous operations.

For the ease of transition the new interface will be known as "hash"
while the old one will remain as "digest".

This patch also changes sg_next to allow chaining.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 crypto/Kconfig          |   4 ++
 crypto/Makefile         |   3 +
 crypto/digest.c         | 129 ++++++++++++++++++++++++++++--------
 crypto/hash.c           |  61 +++++++++++++++++
 crypto/hmac.c           |  12 ++--
 crypto/scatterwalk.h    |   4 +-
 include/crypto/algapi.h |   6 ++
 include/linux/crypto.h  | 172 ++++++++++++++++++++++++++++++++++++++----------
 8 files changed, 321 insertions(+), 70 deletions(-)
 create mode 100644 crypto/hash.c

(limited to 'include')

diff --git a/crypto/Kconfig b/crypto/Kconfig
index be5eb0cb7c30..69c5f992bcd4 100644
--- a/crypto/Kconfig
+++ b/crypto/Kconfig
@@ -20,6 +20,10 @@ config CRYPTO_BLKCIPHER
 	tristate
 	select CRYPTO_ALGAPI
 
+config CRYPTO_HASH
+	tristate
+	select CRYPTO_ALGAPI
+
 config CRYPTO_MANAGER
 	tristate "Cryptographic algorithm manager"
 	select CRYPTO_ALGAPI
diff --git a/crypto/Makefile b/crypto/Makefile
index 5e1ff4e0b1fc..72366208e291 100644
--- a/crypto/Makefile
+++ b/crypto/Makefile
@@ -10,6 +10,9 @@ obj-$(CONFIG_CRYPTO_ALGAPI) += crypto_algapi.o
 
 obj-$(CONFIG_CRYPTO_BLKCIPHER) += blkcipher.o
 
+crypto_hash-objs := hash.o
+obj-$(CONFIG_CRYPTO_HASH) += crypto_hash.o
+
 obj-$(CONFIG_CRYPTO_MANAGER) += cryptomgr.o
 obj-$(CONFIG_CRYPTO_HMAC) += hmac.o
 obj-$(CONFIG_CRYPTO_NULL) += crypto_null.o
diff --git a/crypto/digest.c b/crypto/digest.c
index 96244a528844..5873063db840 100644
--- a/crypto/digest.c
+++ b/crypto/digest.c
@@ -11,29 +11,89 @@
  * any later version.
  *
  */
-#include <linux/crypto.h>
+
 #include <linux/mm.h>
 #include <linux/errno.h>
 #include <linux/highmem.h>
-#include <asm/scatterlist.h>
+#include <linux/module.h>
+#include <linux/scatterlist.h>
+
 #include "internal.h"
+#include "scatterwalk.h"
 
-static void init(struct crypto_tfm *tfm)
+void crypto_digest_init(struct crypto_tfm *tfm)
 {
-	tfm->__crt_alg->cra_digest.dia_init(tfm);
+	struct crypto_hash *hash = crypto_hash_cast(tfm);
+	struct hash_desc desc = { .tfm = hash, .flags = tfm->crt_flags };
+
+	crypto_hash_init(&desc);
 }
+EXPORT_SYMBOL_GPL(crypto_digest_init);
 
-static void update(struct crypto_tfm *tfm,
-                   struct scatterlist *sg, unsigned int nsg)
+void crypto_digest_update(struct crypto_tfm *tfm,
+			  struct scatterlist *sg, unsigned int nsg)
 {
+	struct crypto_hash *hash = crypto_hash_cast(tfm);
+	struct hash_desc desc = { .tfm = hash, .flags = tfm->crt_flags };
+	unsigned int nbytes = 0;
 	unsigned int i;
+
+	for (i = 0; i < nsg; i++)
+		nbytes += sg[i].length;
+
+	crypto_hash_update(&desc, sg, nbytes);
+}
+EXPORT_SYMBOL_GPL(crypto_digest_update);
+
+void crypto_digest_final(struct crypto_tfm *tfm, u8 *out)
+{
+	struct crypto_hash *hash = crypto_hash_cast(tfm);
+	struct hash_desc desc = { .tfm = hash, .flags = tfm->crt_flags };
+
+	crypto_hash_final(&desc, out);
+}
+EXPORT_SYMBOL_GPL(crypto_digest_final);
+
+void crypto_digest_digest(struct crypto_tfm *tfm,
+			  struct scatterlist *sg, unsigned int nsg, u8 *out)
+{
+	struct crypto_hash *hash = crypto_hash_cast(tfm);
+	struct hash_desc desc = { .tfm = hash, .flags = tfm->crt_flags };
+	unsigned int nbytes = 0;
+	unsigned int i;
+
+	for (i = 0; i < nsg; i++)
+		nbytes += sg[i].length;
+
+	crypto_hash_digest(&desc, sg, nbytes, out);
+}
+EXPORT_SYMBOL_GPL(crypto_digest_digest);
+
+static int init(struct hash_desc *desc)
+{
+	struct crypto_tfm *tfm = crypto_hash_tfm(desc->tfm);
+
+	tfm->__crt_alg->cra_digest.dia_init(tfm);
+	return 0;
+}
+
+static int update(struct hash_desc *desc,
+		  struct scatterlist *sg, unsigned int nbytes)
+{
+	struct crypto_tfm *tfm = crypto_hash_tfm(desc->tfm);
 	unsigned int alignmask = crypto_tfm_alg_alignmask(tfm);
 
-	for (i = 0; i < nsg; i++) {
+	if (!nbytes)
+		return 0;
+
+	for (;;) {
+		struct page *pg = sg->page;
+		unsigned int offset = sg->offset;
+		unsigned int l = sg->length;
 
-		struct page *pg = sg[i].page;
-		unsigned int offset = sg[i].offset;
-		unsigned int l = sg[i].length;
+		if (unlikely(l > nbytes))
+			l = nbytes;
+		nbytes -= l;
 
 		do {
 			unsigned int bytes_from_page = min(l, ((unsigned int)
@@ -55,16 +115,23 @@ static void update(struct crypto_tfm *tfm,
 			tfm->__crt_alg->cra_digest.dia_update(tfm, p,
 							      bytes_from_page);
 			crypto_kunmap(src, 0);
-			crypto_yield(tfm->crt_flags);
+			crypto_yield(desc->flags);
 			offset = 0;
 			pg++;
 			l -= bytes_from_page;
 		} while (l > 0);
+
+		if (!nbytes)
+			break;
+		sg = sg_next(sg);
 	}
+
+	return 0;
 }
 
-static void final(struct crypto_tfm *tfm, u8 *out)
+static int final(struct hash_desc *desc, u8 *out)
 {
+	struct crypto_tfm *tfm = crypto_hash_tfm(desc->tfm);
 	unsigned long alignmask = crypto_tfm_alg_alignmask(tfm);
 	struct digest_alg *digest = &tfm->__crt_alg->cra_digest;
 
@@ -78,26 +145,30 @@ static void final(struct crypto_tfm *tfm, u8 *out)
 		memcpy(out, dst, digest->dia_digestsize);
 	} else
 		digest->dia_final(tfm, out);
+
+	return 0;
 }
 
-static int nosetkey(struct crypto_tfm *tfm, const u8 *key, unsigned int keylen)
+static int nosetkey(struct crypto_hash *tfm, const u8 *key, unsigned int keylen)
 {
-	tfm->crt_flags &= ~CRYPTO_TFM_RES_MASK;
+	crypto_hash_clear_flags(tfm, CRYPTO_TFM_RES_MASK);
 	return -ENOSYS;
 }
 
-static int setkey(struct crypto_tfm *tfm, const u8 *key, unsigned int keylen)
+static int setkey(struct crypto_hash *hash, const u8 *key, unsigned int keylen)
 {
-	tfm->crt_flags &= ~CRYPTO_TFM_RES_MASK;
+	struct crypto_tfm *tfm = crypto_hash_tfm(hash);
+
+	crypto_hash_clear_flags(hash, CRYPTO_TFM_RES_MASK);
 	return tfm->__crt_alg->cra_digest.dia_setkey(tfm, key, keylen);
 }
 
-static void digest(struct crypto_tfm *tfm,
-                   struct scatterlist *sg, unsigned int nsg, u8 *out)
+static int digest(struct hash_desc *desc,
+		  struct scatterlist *sg, unsigned int nbytes, u8 *out)
 {
-	init(tfm);
-	update(tfm, sg, nsg);
-	final(tfm, out);
+	init(desc);
+	update(desc, sg, nbytes);
+	return final(desc, out);
 }
 
 int crypto_init_digest_flags(struct crypto_tfm *tfm, u32 flags)
@@ -107,14 +178,18 @@ int crypto_init_digest_flags(struct crypto_tfm *tfm, u32 flags)
 
 int crypto_init_digest_ops(struct crypto_tfm *tfm)
 {
-	struct digest_tfm *ops = &tfm->crt_digest;
+	struct hash_tfm *ops = &tfm->crt_hash;
 	struct digest_alg *dalg = &tfm->__crt_alg->cra_digest;
+
+	if (dalg->dia_digestsize > crypto_tfm_alg_blocksize(tfm))
+		return -EINVAL;
 	
-	ops->dit_init	= init;
-	ops->dit_update	= update;
-	ops->dit_final	= final;
-	ops->dit_digest	= digest;
-	ops->dit_setkey	= dalg->dia_setkey ? setkey : nosetkey;
+	ops->init	= init;
+	ops->update	= update;
+	ops->final	= final;
+	ops->digest	= digest;
+	ops->setkey	= dalg->dia_setkey ? setkey : nosetkey;
+	ops->digestsize	= dalg->dia_digestsize;
 	
 	return crypto_alloc_hmac_block(tfm);
 }
diff --git a/crypto/hash.c b/crypto/hash.c
new file mode 100644
index 000000000000..cdec23d885fe
--- /dev/null
+++ b/crypto/hash.c
@@ -0,0 +1,61 @@
+/*
+ * Cryptographic Hash operations.
+ * 
+ * Copyright (c) 2006 Herbert Xu <herbert@gondor.apana.org.au>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; either version 2 of the License, or (at your option) 
+ * any later version.
+ */
+
+#include <linux/errno.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/seq_file.h>
+
+#include "internal.h"
+
+static unsigned int crypto_hash_ctxsize(struct crypto_alg *alg)
+{
+	return alg->cra_ctxsize;
+}
+
+static int crypto_init_hash_ops(struct crypto_tfm *tfm)
+{
+	struct hash_tfm *crt = &tfm->crt_hash;
+	struct hash_alg *alg = &tfm->__crt_alg->cra_hash;
+
+	if (alg->digestsize > crypto_tfm_alg_blocksize(tfm))
+		return -EINVAL;
+
+	crt->init = alg->init;
+	crt->update = alg->update;
+	crt->final = alg->final;
+	crt->digest = alg->digest;
+	crt->setkey = alg->setkey;
+	crt->digestsize = alg->digestsize;
+
+	return 0;
+}
+
+static void crypto_hash_show(struct seq_file *m, struct crypto_alg *alg)
+	__attribute_used__;
+static void crypto_hash_show(struct seq_file *m, struct crypto_alg *alg)
+{
+	seq_printf(m, "type         : hash\n");
+	seq_printf(m, "blocksize    : %u\n", alg->cra_blocksize);
+	seq_printf(m, "digestsize   : %u\n", alg->cra_hash.digestsize);
+}
+
+const struct crypto_type crypto_hash_type = {
+	.ctxsize = crypto_hash_ctxsize,
+	.init = crypto_init_hash_ops,
+#ifdef CONFIG_PROC_FS
+	.show = crypto_hash_show,
+#endif
+};
+EXPORT_SYMBOL_GPL(crypto_hash_type);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Generic cryptographic hash type");
diff --git a/crypto/hmac.c b/crypto/hmac.c
index 46120dee5ada..ecf7b0a95b56 100644
--- a/crypto/hmac.c
+++ b/crypto/hmac.c
@@ -35,9 +35,9 @@ int crypto_alloc_hmac_block(struct crypto_tfm *tfm)
 
 	BUG_ON(!crypto_tfm_alg_blocksize(tfm));
 	
-	tfm->crt_digest.dit_hmac_block = kmalloc(crypto_tfm_alg_blocksize(tfm),
-	                                         GFP_KERNEL);
-	if (tfm->crt_digest.dit_hmac_block == NULL)
+	tfm->crt_hash.hmac_block = kmalloc(crypto_tfm_alg_blocksize(tfm),
+					   GFP_KERNEL);
+	if (tfm->crt_hash.hmac_block == NULL)
 		ret = -ENOMEM;
 
 	return ret;
@@ -46,14 +46,14 @@ int crypto_alloc_hmac_block(struct crypto_tfm *tfm)
 
 void crypto_free_hmac_block(struct crypto_tfm *tfm)
 {
-	kfree(tfm->crt_digest.dit_hmac_block);
+	kfree(tfm->crt_hash.hmac_block);
 }
 
 void crypto_hmac_init(struct crypto_tfm *tfm, u8 *key, unsigned int *keylen)
 {
 	unsigned int i;
 	struct scatterlist tmp;
-	char *ipad = tfm->crt_digest.dit_hmac_block;
+	char *ipad = tfm->crt_hash.hmac_block;
 	
 	if (*keylen > crypto_tfm_alg_blocksize(tfm)) {
 		hash_key(tfm, key, *keylen);
@@ -83,7 +83,7 @@ void crypto_hmac_final(struct crypto_tfm *tfm, u8 *key,
 {
 	unsigned int i;
 	struct scatterlist tmp;
-	char *opad = tfm->crt_digest.dit_hmac_block;
+	char *opad = tfm->crt_hash.hmac_block;
 	
 	if (*keylen > crypto_tfm_alg_blocksize(tfm)) {
 		hash_key(tfm, key, *keylen);
diff --git a/crypto/scatterwalk.h b/crypto/scatterwalk.h
index ace595a2e119..f1592cc2d0f4 100644
--- a/crypto/scatterwalk.h
+++ b/crypto/scatterwalk.h
@@ -20,11 +20,9 @@
 
 #include "internal.h"
 
-/* Define sg_next is an inline routine now in case we want to change
-   scatterlist to a linked list later. */
 static inline struct scatterlist *sg_next(struct scatterlist *sg)
 {
-	return sg + 1;
+	return (++sg)->length ? sg : (void *)sg->page;
 }
 
 static inline unsigned long scatterwalk_samebuf(struct scatter_walk *walk_in,
diff --git a/include/crypto/algapi.h b/include/crypto/algapi.h
index 444f602724db..5748aecdb414 100644
--- a/include/crypto/algapi.h
+++ b/include/crypto/algapi.h
@@ -82,6 +82,7 @@ struct blkcipher_walk {
 };
 
 extern const struct crypto_type crypto_blkcipher_type;
+extern const struct crypto_type crypto_hash_type;
 
 void crypto_mod_put(struct crypto_alg *alg);
 
@@ -136,6 +137,11 @@ static inline struct cipher_alg *crypto_cipher_alg(struct crypto_cipher *tfm)
 	return &crypto_cipher_tfm(tfm)->__crt_alg->cra_cipher;
 }
 
+static inline void *crypto_hash_ctx_aligned(struct crypto_hash *tfm)
+{
+	return crypto_tfm_ctx_aligned(&tfm->base);
+}
+
 static inline void blkcipher_walk_init(struct blkcipher_walk *walk,
 				       struct scatterlist *dst,
 				       struct scatterlist *src,
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index 0be666b50463..40c0aab8ad4c 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -31,8 +31,11 @@
 #define CRYPTO_ALG_TYPE_MASK		0x0000000f
 #define CRYPTO_ALG_TYPE_CIPHER		0x00000001
 #define CRYPTO_ALG_TYPE_DIGEST		0x00000002
-#define CRYPTO_ALG_TYPE_BLKCIPHER	0x00000003
-#define CRYPTO_ALG_TYPE_COMPRESS	0x00000004
+#define CRYPTO_ALG_TYPE_HASH		0x00000003
+#define CRYPTO_ALG_TYPE_BLKCIPHER	0x00000004
+#define CRYPTO_ALG_TYPE_COMPRESS	0x00000005
+
+#define CRYPTO_ALG_TYPE_HASH_MASK	0x0000000e
 
 #define CRYPTO_ALG_LARVAL		0x00000010
 #define CRYPTO_ALG_DEAD			0x00000020
@@ -90,6 +93,7 @@
 
 struct scatterlist;
 struct crypto_blkcipher;
+struct crypto_hash;
 struct crypto_tfm;
 struct crypto_type;
 
@@ -107,6 +111,11 @@ struct cipher_desc {
 	void *info;
 };
 
+struct hash_desc {
+	struct crypto_hash *tfm;
+	u32 flags;
+};
+
 /*
  * Algorithms: modular crypto algorithm implementations, managed
  * via crypto_register_alg() and crypto_unregister_alg().
@@ -158,6 +167,19 @@ struct digest_alg {
 	                  unsigned int keylen);
 };
 
+struct hash_alg {
+	int (*init)(struct hash_desc *desc);
+	int (*update)(struct hash_desc *desc, struct scatterlist *sg,
+		      unsigned int nbytes);
+	int (*final)(struct hash_desc *desc, u8 *out);
+	int (*digest)(struct hash_desc *desc, struct scatterlist *sg,
+		      unsigned int nbytes, u8 *out);
+	int (*setkey)(struct crypto_hash *tfm, const u8 *key,
+		      unsigned int keylen);
+
+	unsigned int digestsize;
+};
+
 struct compress_alg {
 	int (*coa_compress)(struct crypto_tfm *tfm, const u8 *src,
 			    unsigned int slen, u8 *dst, unsigned int *dlen);
@@ -168,6 +190,7 @@ struct compress_alg {
 #define cra_blkcipher	cra_u.blkcipher
 #define cra_cipher	cra_u.cipher
 #define cra_digest	cra_u.digest
+#define cra_hash	cra_u.hash
 #define cra_compress	cra_u.compress
 
 struct crypto_alg {
@@ -191,6 +214,7 @@ struct crypto_alg {
 		struct blkcipher_alg blkcipher;
 		struct cipher_alg cipher;
 		struct digest_alg digest;
+		struct hash_alg hash;
 		struct compress_alg compress;
 	} cra_u;
 
@@ -262,18 +286,19 @@ struct cipher_tfm {
 	void (*cit_decrypt_one)(struct crypto_tfm *tfm, u8 *dst, const u8 *src);
 };
 
-struct digest_tfm {
-	void (*dit_init)(struct crypto_tfm *tfm);
-	void (*dit_update)(struct crypto_tfm *tfm,
-	                   struct scatterlist *sg, unsigned int nsg);
-	void (*dit_final)(struct crypto_tfm *tfm, u8 *out);
-	void (*dit_digest)(struct crypto_tfm *tfm, struct scatterlist *sg,
-	                   unsigned int nsg, u8 *out);
-	int (*dit_setkey)(struct crypto_tfm *tfm,
-	                  const u8 *key, unsigned int keylen);
+struct hash_tfm {
+	int (*init)(struct hash_desc *desc);
+	int (*update)(struct hash_desc *desc,
+		      struct scatterlist *sg, unsigned int nsg);
+	int (*final)(struct hash_desc *desc, u8 *out);
+	int (*digest)(struct hash_desc *desc, struct scatterlist *sg,
+		      unsigned int nsg, u8 *out);
+	int (*setkey)(struct crypto_hash *tfm, const u8 *key,
+		      unsigned int keylen);
 #ifdef CONFIG_CRYPTO_HMAC
-	void *dit_hmac_block;
+	void *hmac_block;
 #endif
+	unsigned int digestsize;
 };
 
 struct compress_tfm {
@@ -287,7 +312,7 @@ struct compress_tfm {
 
 #define crt_blkcipher	crt_u.blkcipher
 #define crt_cipher	crt_u.cipher
-#define crt_digest	crt_u.digest
+#define crt_hash	crt_u.hash
 #define crt_compress	crt_u.compress
 
 struct crypto_tfm {
@@ -297,7 +322,7 @@ struct crypto_tfm {
 	union {
 		struct blkcipher_tfm blkcipher;
 		struct cipher_tfm cipher;
-		struct digest_tfm digest;
+		struct hash_tfm hash;
 		struct compress_tfm compress;
 	} crt_u;
 	
@@ -312,6 +337,10 @@ struct crypto_blkcipher {
 	struct crypto_tfm base;
 };
 
+struct crypto_hash {
+	struct crypto_tfm base;
+};
+
 enum {
 	CRYPTOA_UNSPEC,
 	CRYPTOA_ALG,
@@ -647,39 +676,114 @@ static inline void crypto_cipher_decrypt_one(struct crypto_cipher *tfm,
 						dst, src);
 }
 
-static inline void crypto_digest_init(struct crypto_tfm *tfm)
+void crypto_digest_init(struct crypto_tfm *tfm);
+void crypto_digest_update(struct crypto_tfm *tfm,
+			  struct scatterlist *sg, unsigned int nsg);
+void crypto_digest_final(struct crypto_tfm *tfm, u8 *out);
+void crypto_digest_digest(struct crypto_tfm *tfm,
+			  struct scatterlist *sg, unsigned int nsg, u8 *out);
+
+static inline struct crypto_hash *__crypto_hash_cast(struct crypto_tfm *tfm)
 {
-	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_DIGEST);
-	tfm->crt_digest.dit_init(tfm);
+	return (struct crypto_hash *)tfm;
 }
 
-static inline void crypto_digest_update(struct crypto_tfm *tfm,
-                                        struct scatterlist *sg,
-                                        unsigned int nsg)
+static inline struct crypto_hash *crypto_hash_cast(struct crypto_tfm *tfm)
 {
-	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_DIGEST);
-	tfm->crt_digest.dit_update(tfm, sg, nsg);
+	BUG_ON((crypto_tfm_alg_type(tfm) ^ CRYPTO_ALG_TYPE_HASH) &
+	       CRYPTO_ALG_TYPE_HASH_MASK);
+	return __crypto_hash_cast(tfm);
 }
 
-static inline void crypto_digest_final(struct crypto_tfm *tfm, u8 *out)
+static inline int crypto_digest_setkey(struct crypto_tfm *tfm,
+                                       const u8 *key, unsigned int keylen)
 {
-	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_DIGEST);
-	tfm->crt_digest.dit_final(tfm, out);
+	return tfm->crt_hash.setkey(crypto_hash_cast(tfm), key, keylen);
 }
 
-static inline void crypto_digest_digest(struct crypto_tfm *tfm,
-                                        struct scatterlist *sg,
-                                        unsigned int nsg, u8 *out)
+static inline struct crypto_hash *crypto_alloc_hash(const char *alg_name,
+						    u32 type, u32 mask)
 {
-	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_DIGEST);
-	tfm->crt_digest.dit_digest(tfm, sg, nsg, out);
+	type &= ~CRYPTO_ALG_TYPE_MASK;
+	type |= CRYPTO_ALG_TYPE_HASH;
+	mask |= CRYPTO_ALG_TYPE_HASH_MASK;
+
+	return __crypto_hash_cast(crypto_alloc_base(alg_name, type, mask));
 }
 
-static inline int crypto_digest_setkey(struct crypto_tfm *tfm,
-                                       const u8 *key, unsigned int keylen)
+static inline struct crypto_tfm *crypto_hash_tfm(struct crypto_hash *tfm)
 {
-	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_DIGEST);
-	return tfm->crt_digest.dit_setkey(tfm, key, keylen);
+	return &tfm->base;
+}
+
+static inline void crypto_free_hash(struct crypto_hash *tfm)
+{
+	crypto_free_tfm(crypto_hash_tfm(tfm));
+}
+
+static inline struct hash_tfm *crypto_hash_crt(struct crypto_hash *tfm)
+{
+	return &crypto_hash_tfm(tfm)->crt_hash;
+}
+
+static inline unsigned int crypto_hash_blocksize(struct crypto_hash *tfm)
+{
+	return crypto_tfm_alg_blocksize(crypto_hash_tfm(tfm));
+}
+
+static inline unsigned int crypto_hash_alignmask(struct crypto_hash *tfm)
+{
+	return crypto_tfm_alg_alignmask(crypto_hash_tfm(tfm));
+}
+
+static inline unsigned int crypto_hash_digestsize(struct crypto_hash *tfm)
+{
+	return crypto_hash_crt(tfm)->digestsize;
+}
+
+static inline u32 crypto_hash_get_flags(struct crypto_hash *tfm)
+{
+	return crypto_tfm_get_flags(crypto_hash_tfm(tfm));
+}
+
+static inline void crypto_hash_set_flags(struct crypto_hash *tfm, u32 flags)
+{
+	crypto_tfm_set_flags(crypto_hash_tfm(tfm), flags);
+}
+
+static inline void crypto_hash_clear_flags(struct crypto_hash *tfm, u32 flags)
+{
+	crypto_tfm_clear_flags(crypto_hash_tfm(tfm), flags);
+}
+
+static inline int crypto_hash_init(struct hash_desc *desc)
+{
+	return crypto_hash_crt(desc->tfm)->init(desc);
+}
+
+static inline int crypto_hash_update(struct hash_desc *desc,
+				     struct scatterlist *sg,
+				     unsigned int nbytes)
+{
+	return crypto_hash_crt(desc->tfm)->update(desc, sg, nbytes);
+}
+
+static inline int crypto_hash_final(struct hash_desc *desc, u8 *out)
+{
+	return crypto_hash_crt(desc->tfm)->final(desc, out);
+}
+
+static inline int crypto_hash_digest(struct hash_desc *desc,
+				     struct scatterlist *sg,
+				     unsigned int nbytes, u8 *out)
+{
+	return crypto_hash_crt(desc->tfm)->digest(desc, sg, nbytes, out);
+}
+
+static inline int crypto_hash_setkey(struct crypto_hash *hash,
+				     const u8 *key, unsigned int keylen)
+{
+	return crypto_hash_crt(hash)->setkey(hash, key, keylen);
 }
 
 static int crypto_cipher_encrypt(struct crypto_tfm *tfm,
-- 
cgit v1.2.3


From 07d4ee583e21830ec5604d31f65cdc60a6eca19e Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sun, 20 Aug 2006 14:24:50 +1000
Subject: [IPSEC]: Use HMAC template and hash interface

This patch converts IPsec to use the new HMAC template.  The names of
existing simple digest algorithms may still be used to refer to their
HMAC composites.

The same structure can be used by other MACs such as AES-XCBC-MAC.

This patch also switches from the digest interface to hash.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ah.h     | 29 ++++++++++++++++++-----------
 include/net/esp.h    | 28 ++++++++++++++++------------
 include/net/xfrm.h   |  9 +++++----
 net/ipv4/ah4.c       | 36 ++++++++++++++++++++++++------------
 net/ipv4/esp4.c      | 36 +++++++++++++++++++++---------------
 net/ipv6/ah6.c       | 35 +++++++++++++++++++++++------------
 net/ipv6/esp6.c      | 42 ++++++++++++++++++++++++------------------
 net/xfrm/xfrm_algo.c | 40 +++++++++++++++++++++++++++-------------
 8 files changed, 158 insertions(+), 97 deletions(-)

(limited to 'include')

diff --git a/include/net/ah.h b/include/net/ah.h
index 8e27c9ba8b84..8f257c159902 100644
--- a/include/net/ah.h
+++ b/include/net/ah.h
@@ -15,22 +15,29 @@ struct ah_data
 	int			icv_full_len;
 	int			icv_trunc_len;
 
-	void			(*icv)(struct ah_data*,
-	                               struct sk_buff *skb, u8 *icv);
-
-	struct crypto_tfm	*tfm;
+	struct crypto_hash	*tfm;
 };
 
-static inline void
-ah_hmac_digest(struct ah_data *ahp, struct sk_buff *skb, u8 *auth_data)
+static inline int ah_mac_digest(struct ah_data *ahp, struct sk_buff *skb,
+				u8 *auth_data)
 {
-	struct crypto_tfm *tfm = ahp->tfm;
+	struct hash_desc desc;
+	int err;
+
+	desc.tfm = ahp->tfm;
+	desc.flags = 0;
 
 	memset(auth_data, 0, ahp->icv_trunc_len);
-	crypto_hmac_init(tfm, ahp->key, &ahp->key_len);
-	skb_icv_walk(skb, tfm, 0, skb->len, crypto_hmac_update);
-	crypto_hmac_final(tfm, ahp->key, &ahp->key_len, ahp->work_icv);
-	memcpy(auth_data, ahp->work_icv, ahp->icv_trunc_len);
+	err = crypto_hash_init(&desc);
+	if (unlikely(err))
+		goto out;
+	err = skb_icv_walk(skb, &desc, 0, skb->len, crypto_hash_update);
+	if (unlikely(err))
+		goto out;
+	err = crypto_hash_final(&desc, ahp->work_icv);
+
+out:
+	return err;
 }
 
 #endif
diff --git a/include/net/esp.h b/include/net/esp.h
index af2ff18700c7..064366d66eea 100644
--- a/include/net/esp.h
+++ b/include/net/esp.h
@@ -35,7 +35,7 @@ struct esp_data
 		void			(*icv)(struct esp_data*,
 		                               struct sk_buff *skb,
 		                               int offset, int len, u8 *icv);
-		struct crypto_tfm	*tfm;
+		struct crypto_hash	*tfm;
 	} auth;
 };
 
@@ -43,18 +43,22 @@ extern int skb_to_sgvec(struct sk_buff *skb, struct scatterlist *sg, int offset,
 extern int skb_cow_data(struct sk_buff *skb, int tailbits, struct sk_buff **trailer);
 extern void *pskb_put(struct sk_buff *skb, struct sk_buff *tail, int len);
 
-static inline void
-esp_hmac_digest(struct esp_data *esp, struct sk_buff *skb, int offset,
-                int len, u8 *auth_data)
+static inline int esp_mac_digest(struct esp_data *esp, struct sk_buff *skb,
+				 int offset, int len)
 {
-	struct crypto_tfm *tfm = esp->auth.tfm;
-	char *icv = esp->auth.work_icv;
-
-	memset(auth_data, 0, esp->auth.icv_trunc_len);
-	crypto_hmac_init(tfm, esp->auth.key, &esp->auth.key_len);
-	skb_icv_walk(skb, tfm, offset, len, crypto_hmac_update);
-	crypto_hmac_final(tfm, esp->auth.key, &esp->auth.key_len, icv);
-	memcpy(auth_data, icv, esp->auth.icv_trunc_len);
+	struct hash_desc desc;
+	int err;
+
+	desc.tfm = esp->auth.tfm;
+	desc.flags = 0;
+
+	err = crypto_hash_init(&desc);
+	if (unlikely(err))
+		return err;
+	err = skb_icv_walk(skb, &desc, offset, len, crypto_hash_update);
+	if (unlikely(err))
+		return err;
+	return crypto_hash_final(&desc, esp->auth.work_icv);
 }
 
 #endif
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index e9114e41affc..3ecd9fa1ed4b 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -984,12 +984,13 @@ extern struct xfrm_algo_desc *xfrm_aalg_get_byname(char *name, int probe);
 extern struct xfrm_algo_desc *xfrm_ealg_get_byname(char *name, int probe);
 extern struct xfrm_algo_desc *xfrm_calg_get_byname(char *name, int probe);
 
-struct crypto_tfm;
+struct hash_desc;
 struct scatterlist;
-typedef void (icv_update_fn_t)(struct crypto_tfm *, struct scatterlist *, unsigned int);
+typedef int (icv_update_fn_t)(struct hash_desc *, struct scatterlist *,
+			      unsigned int);
 
-extern void skb_icv_walk(const struct sk_buff *skb, struct crypto_tfm *tfm,
-			 int offset, int len, icv_update_fn_t icv_update);
+extern int skb_icv_walk(const struct sk_buff *skb, struct hash_desc *tfm,
+			int offset, int len, icv_update_fn_t icv_update);
 
 static inline int xfrm_addr_cmp(xfrm_address_t *a, xfrm_address_t *b,
 				int family)
diff --git a/net/ipv4/ah4.c b/net/ipv4/ah4.c
index 1366bc6ce6a5..2b98943e6b02 100644
--- a/net/ipv4/ah4.c
+++ b/net/ipv4/ah4.c
@@ -1,3 +1,4 @@
+#include <linux/err.h>
 #include <linux/module.h>
 #include <net/ip.h>
 #include <net/xfrm.h>
@@ -97,7 +98,10 @@ static int ah_output(struct xfrm_state *x, struct sk_buff *skb)
 	ah->spi = x->id.spi;
 	ah->seq_no = htonl(++x->replay.oseq);
 	xfrm_aevent_doreplay(x);
-	ahp->icv(ahp, skb, ah->auth_data);
+	err = ah_mac_digest(ahp, skb, ah->auth_data);
+	if (err)
+		goto error;
+	memcpy(ah->auth_data, ahp->work_icv, ahp->icv_trunc_len);
 
 	top_iph->tos = iph->tos;
 	top_iph->ttl = iph->ttl;
@@ -119,6 +123,7 @@ static int ah_input(struct xfrm_state *x, struct sk_buff *skb)
 {
 	int ah_hlen;
 	int ihl;
+	int err = -EINVAL;
 	struct iphdr *iph;
 	struct ip_auth_hdr *ah;
 	struct ah_data *ahp;
@@ -166,8 +171,11 @@ static int ah_input(struct xfrm_state *x, struct sk_buff *skb)
 		
 		memcpy(auth_data, ah->auth_data, ahp->icv_trunc_len);
 		skb_push(skb, ihl);
-		ahp->icv(ahp, skb, ah->auth_data);
-		if (memcmp(ah->auth_data, auth_data, ahp->icv_trunc_len)) {
+		err = ah_mac_digest(ahp, skb, ah->auth_data);
+		if (err)
+			goto out;
+		err = -EINVAL;
+		if (memcmp(ahp->work_icv, auth_data, ahp->icv_trunc_len)) {
 			x->stats.integrity_failed++;
 			goto out;
 		}
@@ -179,7 +187,7 @@ static int ah_input(struct xfrm_state *x, struct sk_buff *skb)
 	return 0;
 
 out:
-	return -EINVAL;
+	return err;
 }
 
 static void ah4_err(struct sk_buff *skb, u32 info)
@@ -204,6 +212,7 @@ static int ah_init_state(struct xfrm_state *x)
 {
 	struct ah_data *ahp = NULL;
 	struct xfrm_algo_desc *aalg_desc;
+	struct crypto_hash *tfm;
 
 	if (!x->aalg)
 		goto error;
@@ -221,24 +230,27 @@ static int ah_init_state(struct xfrm_state *x)
 
 	ahp->key = x->aalg->alg_key;
 	ahp->key_len = (x->aalg->alg_key_len+7)/8;
-	ahp->tfm = crypto_alloc_tfm(x->aalg->alg_name, 0);
-	if (!ahp->tfm)
+	tfm = crypto_alloc_hash(x->aalg->alg_name, 0, CRYPTO_ALG_ASYNC);
+	if (IS_ERR(tfm))
+		goto error;
+
+	ahp->tfm = tfm;
+	if (crypto_hash_setkey(tfm, ahp->key, ahp->key_len))
 		goto error;
-	ahp->icv = ah_hmac_digest;
 	
 	/*
 	 * Lookup the algorithm description maintained by xfrm_algo,
 	 * verify crypto transform properties, and store information
 	 * we need for AH processing.  This lookup cannot fail here
-	 * after a successful crypto_alloc_tfm().
+	 * after a successful crypto_alloc_hash().
 	 */
 	aalg_desc = xfrm_aalg_get_byname(x->aalg->alg_name, 0);
 	BUG_ON(!aalg_desc);
 
 	if (aalg_desc->uinfo.auth.icv_fullbits/8 !=
-	    crypto_tfm_alg_digestsize(ahp->tfm)) {
+	    crypto_hash_digestsize(tfm)) {
 		printk(KERN_INFO "AH: %s digestsize %u != %hu\n",
-		       x->aalg->alg_name, crypto_tfm_alg_digestsize(ahp->tfm),
+		       x->aalg->alg_name, crypto_hash_digestsize(tfm),
 		       aalg_desc->uinfo.auth.icv_fullbits/8);
 		goto error;
 	}
@@ -262,7 +274,7 @@ static int ah_init_state(struct xfrm_state *x)
 error:
 	if (ahp) {
 		kfree(ahp->work_icv);
-		crypto_free_tfm(ahp->tfm);
+		crypto_free_hash(ahp->tfm);
 		kfree(ahp);
 	}
 	return -EINVAL;
@@ -277,7 +289,7 @@ static void ah_destroy(struct xfrm_state *x)
 
 	kfree(ahp->work_icv);
 	ahp->work_icv = NULL;
-	crypto_free_tfm(ahp->tfm);
+	crypto_free_hash(ahp->tfm);
 	ahp->tfm = NULL;
 	kfree(ahp);
 }
diff --git a/net/ipv4/esp4.c b/net/ipv4/esp4.c
index 7c63ae494742..b428489f6ccd 100644
--- a/net/ipv4/esp4.c
+++ b/net/ipv4/esp4.c
@@ -121,9 +121,9 @@ static int esp_output(struct xfrm_state *x, struct sk_buff *skb)
 	}
 
 	if (esp->auth.icv_full_len) {
-		esp->auth.icv(esp, skb, (u8*)esph-skb->data,
-		              sizeof(struct ip_esp_hdr) + esp->conf.ivlen+clen, trailer->tail);
-		pskb_put(skb, trailer, alen);
+		err = esp_mac_digest(esp, skb, (u8 *)esph - skb->data,
+				     sizeof(*esph) + esp->conf.ivlen + clen);
+		memcpy(pskb_put(skb, trailer, alen), esp->auth.work_icv, alen);
 	}
 
 	ip_send_check(top_iph);
@@ -163,15 +163,16 @@ static int esp_input(struct xfrm_state *x, struct sk_buff *skb)
 
 	/* If integrity check is required, do this. */
 	if (esp->auth.icv_full_len) {
-		u8 sum[esp->auth.icv_full_len];
-		u8 sum1[alen];
-		
-		esp->auth.icv(esp, skb, 0, skb->len-alen, sum);
+		u8 sum[alen];
 
-		if (skb_copy_bits(skb, skb->len-alen, sum1, alen))
+		err = esp_mac_digest(esp, skb, 0, skb->len - alen);
+		if (err)
+			goto out;
+
+		if (skb_copy_bits(skb, skb->len - alen, sum, alen))
 			BUG();
 
-		if (unlikely(memcmp(sum, sum1, alen))) {
+		if (unlikely(memcmp(esp->auth.work_icv, sum, alen))) {
 			x->stats.integrity_failed++;
 			goto out;
 		}
@@ -307,7 +308,7 @@ static void esp_destroy(struct xfrm_state *x)
 	esp->conf.tfm = NULL;
 	kfree(esp->conf.ivec);
 	esp->conf.ivec = NULL;
-	crypto_free_tfm(esp->auth.tfm);
+	crypto_free_hash(esp->auth.tfm);
 	esp->auth.tfm = NULL;
 	kfree(esp->auth.work_icv);
 	esp->auth.work_icv = NULL;
@@ -333,22 +334,27 @@ static int esp_init_state(struct xfrm_state *x)
 
 	if (x->aalg) {
 		struct xfrm_algo_desc *aalg_desc;
+		struct crypto_hash *hash;
 
 		esp->auth.key = x->aalg->alg_key;
 		esp->auth.key_len = (x->aalg->alg_key_len+7)/8;
-		esp->auth.tfm = crypto_alloc_tfm(x->aalg->alg_name, 0);
-		if (esp->auth.tfm == NULL)
+		hash = crypto_alloc_hash(x->aalg->alg_name, 0,
+					 CRYPTO_ALG_ASYNC);
+		if (IS_ERR(hash))
+			goto error;
+
+		esp->auth.tfm = hash;
+		if (crypto_hash_setkey(hash, esp->auth.key, esp->auth.key_len))
 			goto error;
-		esp->auth.icv = esp_hmac_digest;
 
 		aalg_desc = xfrm_aalg_get_byname(x->aalg->alg_name, 0);
 		BUG_ON(!aalg_desc);
 
 		if (aalg_desc->uinfo.auth.icv_fullbits/8 !=
-		    crypto_tfm_alg_digestsize(esp->auth.tfm)) {
+		    crypto_hash_digestsize(hash)) {
 			NETDEBUG(KERN_INFO "ESP: %s digestsize %u != %hu\n",
 				 x->aalg->alg_name,
-				 crypto_tfm_alg_digestsize(esp->auth.tfm),
+				 crypto_hash_digestsize(hash),
 				 aalg_desc->uinfo.auth.icv_fullbits/8);
 			goto error;
 		}
diff --git a/net/ipv6/ah6.c b/net/ipv6/ah6.c
index 9d4831bd4335..00ffa7bc6c9f 100644
--- a/net/ipv6/ah6.c
+++ b/net/ipv6/ah6.c
@@ -213,7 +213,10 @@ static int ah6_output(struct xfrm_state *x, struct sk_buff *skb)
 	ah->spi = x->id.spi;
 	ah->seq_no = htonl(++x->replay.oseq);
 	xfrm_aevent_doreplay(x);
-	ahp->icv(ahp, skb, ah->auth_data);
+	err = ah_mac_digest(ahp, skb, ah->auth_data);
+	if (err)
+		goto error_free_iph;
+	memcpy(ah->auth_data, ahp->work_icv, ahp->icv_trunc_len);
 
 	err = 0;
 
@@ -251,6 +254,7 @@ static int ah6_input(struct xfrm_state *x, struct sk_buff *skb)
 	u16 hdr_len;
 	u16 ah_hlen;
 	int nexthdr;
+	int err = -EINVAL;
 
 	if (!pskb_may_pull(skb, sizeof(struct ip_auth_hdr)))
 		goto out;
@@ -292,8 +296,11 @@ static int ah6_input(struct xfrm_state *x, struct sk_buff *skb)
 		memcpy(auth_data, ah->auth_data, ahp->icv_trunc_len);
 		memset(ah->auth_data, 0, ahp->icv_trunc_len);
 		skb_push(skb, hdr_len);
-		ahp->icv(ahp, skb, ah->auth_data);
-		if (memcmp(ah->auth_data, auth_data, ahp->icv_trunc_len)) {
+		err = ah_mac_digest(ahp, skb, ah->auth_data);
+		if (err)
+			goto free_out;
+		err = -EINVAL;
+		if (memcmp(ahp->work_icv, auth_data, ahp->icv_trunc_len)) {
 			LIMIT_NETDEBUG(KERN_WARNING "ipsec ah authentication error\n");
 			x->stats.integrity_failed++;
 			goto free_out;
@@ -310,7 +317,7 @@ static int ah6_input(struct xfrm_state *x, struct sk_buff *skb)
 free_out:
 	kfree(tmp_hdr);
 out:
-	return -EINVAL;
+	return err;
 }
 
 static void ah6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, 
@@ -338,6 +345,7 @@ static int ah6_init_state(struct xfrm_state *x)
 {
 	struct ah_data *ahp = NULL;
 	struct xfrm_algo_desc *aalg_desc;
+	struct crypto_hash *tfm;
 
 	if (!x->aalg)
 		goto error;
@@ -355,24 +363,27 @@ static int ah6_init_state(struct xfrm_state *x)
 
 	ahp->key = x->aalg->alg_key;
 	ahp->key_len = (x->aalg->alg_key_len+7)/8;
-	ahp->tfm = crypto_alloc_tfm(x->aalg->alg_name, 0);
-	if (!ahp->tfm)
+	tfm = crypto_alloc_hash(x->aalg->alg_name, 0, CRYPTO_ALG_ASYNC);
+	if (IS_ERR(tfm))
+		goto error;
+
+	ahp->tfm = tfm;
+	if (crypto_hash_setkey(tfm, ahp->key, ahp->key_len))
 		goto error;
-	ahp->icv = ah_hmac_digest;
 	
 	/*
 	 * Lookup the algorithm description maintained by xfrm_algo,
 	 * verify crypto transform properties, and store information
 	 * we need for AH processing.  This lookup cannot fail here
-	 * after a successful crypto_alloc_tfm().
+	 * after a successful crypto_alloc_hash().
 	 */
 	aalg_desc = xfrm_aalg_get_byname(x->aalg->alg_name, 0);
 	BUG_ON(!aalg_desc);
 
 	if (aalg_desc->uinfo.auth.icv_fullbits/8 !=
-	    crypto_tfm_alg_digestsize(ahp->tfm)) {
+	    crypto_hash_digestsize(tfm)) {
 		printk(KERN_INFO "AH: %s digestsize %u != %hu\n",
-		       x->aalg->alg_name, crypto_tfm_alg_digestsize(ahp->tfm),
+		       x->aalg->alg_name, crypto_hash_digestsize(tfm),
 		       aalg_desc->uinfo.auth.icv_fullbits/8);
 		goto error;
 	}
@@ -396,7 +407,7 @@ static int ah6_init_state(struct xfrm_state *x)
 error:
 	if (ahp) {
 		kfree(ahp->work_icv);
-		crypto_free_tfm(ahp->tfm);
+		crypto_free_hash(ahp->tfm);
 		kfree(ahp);
 	}
 	return -EINVAL;
@@ -411,7 +422,7 @@ static void ah6_destroy(struct xfrm_state *x)
 
 	kfree(ahp->work_icv);
 	ahp->work_icv = NULL;
-	crypto_free_tfm(ahp->tfm);
+	crypto_free_hash(ahp->tfm);
 	ahp->tfm = NULL;
 	kfree(ahp);
 }
diff --git a/net/ipv6/esp6.c b/net/ipv6/esp6.c
index 46a7e687948e..2ebfd281e721 100644
--- a/net/ipv6/esp6.c
+++ b/net/ipv6/esp6.c
@@ -125,9 +125,9 @@ static int esp6_output(struct xfrm_state *x, struct sk_buff *skb)
 	}
 
 	if (esp->auth.icv_full_len) {
-		esp->auth.icv(esp, skb, (u8*)esph-skb->data,
-			sizeof(struct ipv6_esp_hdr) + esp->conf.ivlen+clen, trailer->tail);
-		pskb_put(skb, trailer, alen);
+		err = esp_mac_digest(esp, skb, (u8 *)esph - skb->data,
+				     sizeof(*esph) + esp->conf.ivlen + clen);
+		memcpy(pskb_put(skb, trailer, alen), esp->auth.work_icv, alen);
 	}
 
 error:
@@ -162,15 +162,16 @@ static int esp6_input(struct xfrm_state *x, struct sk_buff *skb)
 
 	/* If integrity check is required, do this. */
         if (esp->auth.icv_full_len) {
-		u8 sum[esp->auth.icv_full_len];
-		u8 sum1[alen];
+		u8 sum[alen];
 
-		esp->auth.icv(esp, skb, 0, skb->len-alen, sum);
+		ret = esp_mac_digest(esp, skb, 0, skb->len - alen);
+		if (ret)
+			goto out;
 
-		if (skb_copy_bits(skb, skb->len-alen, sum1, alen))
+		if (skb_copy_bits(skb, skb->len - alen, sum, alen))
 			BUG();
 
-		if (unlikely(memcmp(sum, sum1, alen))) {
+		if (unlikely(memcmp(esp->auth.work_icv, sum, alen))) {
 			x->stats.integrity_failed++;
 			ret = -EINVAL;
 			goto out;
@@ -279,7 +280,7 @@ static void esp6_destroy(struct xfrm_state *x)
 	esp->conf.tfm = NULL;
 	kfree(esp->conf.ivec);
 	esp->conf.ivec = NULL;
-	crypto_free_tfm(esp->auth.tfm);
+	crypto_free_hash(esp->auth.tfm);
 	esp->auth.tfm = NULL;
 	kfree(esp->auth.work_icv);
 	esp->auth.work_icv = NULL;
@@ -308,24 +309,29 @@ static int esp6_init_state(struct xfrm_state *x)
 
 	if (x->aalg) {
 		struct xfrm_algo_desc *aalg_desc;
+		struct crypto_hash *hash;
 
 		esp->auth.key = x->aalg->alg_key;
 		esp->auth.key_len = (x->aalg->alg_key_len+7)/8;
-		esp->auth.tfm = crypto_alloc_tfm(x->aalg->alg_name, 0);
-		if (esp->auth.tfm == NULL)
+		hash = crypto_alloc_hash(x->aalg->alg_name, 0,
+					 CRYPTO_ALG_ASYNC);
+		if (IS_ERR(hash))
+			goto error;
+
+		esp->auth.tfm = hash;
+		if (crypto_hash_setkey(hash, esp->auth.key, esp->auth.key_len))
 			goto error;
-		esp->auth.icv = esp_hmac_digest;
  
 		aalg_desc = xfrm_aalg_get_byname(x->aalg->alg_name, 0);
 		BUG_ON(!aalg_desc);
  
 		if (aalg_desc->uinfo.auth.icv_fullbits/8 !=
-			crypto_tfm_alg_digestsize(esp->auth.tfm)) {
-				printk(KERN_INFO "ESP: %s digestsize %u != %hu\n",
-					x->aalg->alg_name,
-					crypto_tfm_alg_digestsize(esp->auth.tfm),
-					aalg_desc->uinfo.auth.icv_fullbits/8);
-				goto error;
+		    crypto_hash_digestsize(hash)) {
+			NETDEBUG(KERN_INFO "ESP: %s digestsize %u != %hu\n",
+				 x->aalg->alg_name,
+				 crypto_hash_digestsize(hash),
+				 aalg_desc->uinfo.auth.icv_fullbits/8);
+			goto error;
 		}
  
 		esp->auth.icv_full_len = aalg_desc->uinfo.auth.icv_fullbits/8;
diff --git a/net/xfrm/xfrm_algo.c b/net/xfrm/xfrm_algo.c
index 9b03d8497fba..87918f281bb4 100644
--- a/net/xfrm/xfrm_algo.c
+++ b/net/xfrm/xfrm_algo.c
@@ -30,7 +30,8 @@
  */
 static struct xfrm_algo_desc aalg_list[] = {
 {
-	.name = "digest_null",
+	.name = "hmac(digest_null)",
+	.compat = "digest_null",
 	
 	.uinfo = {
 		.auth = {
@@ -47,7 +48,8 @@ static struct xfrm_algo_desc aalg_list[] = {
 	}
 },
 {
-	.name = "md5",
+	.name = "hmac(md5)",
+	.compat = "md5",
 
 	.uinfo = {
 		.auth = {
@@ -64,7 +66,8 @@ static struct xfrm_algo_desc aalg_list[] = {
 	}
 },
 {
-	.name = "sha1",
+	.name = "hmac(sha1)",
+	.compat = "sha1",
 
 	.uinfo = {
 		.auth = {
@@ -81,7 +84,8 @@ static struct xfrm_algo_desc aalg_list[] = {
 	}
 },
 {
-	.name = "sha256",
+	.name = "hmac(sha256)",
+	.compat = "sha256",
 
 	.uinfo = {
 		.auth = {
@@ -98,7 +102,8 @@ static struct xfrm_algo_desc aalg_list[] = {
 	}
 },
 {
-	.name = "ripemd160",
+	.name = "hmac(ripemd160)",
+	.compat = "ripemd160",
 
 	.uinfo = {
 		.auth = {
@@ -480,11 +485,12 @@ EXPORT_SYMBOL_GPL(xfrm_count_enc_supported);
 
 /* Move to common area: it is shared with AH. */
 
-void skb_icv_walk(const struct sk_buff *skb, struct crypto_tfm *tfm,
-		  int offset, int len, icv_update_fn_t icv_update)
+int skb_icv_walk(const struct sk_buff *skb, struct hash_desc *desc,
+		 int offset, int len, icv_update_fn_t icv_update)
 {
 	int start = skb_headlen(skb);
 	int i, copy = start - offset;
+	int err;
 	struct scatterlist sg;
 
 	/* Checksum header. */
@@ -496,10 +502,12 @@ void skb_icv_walk(const struct sk_buff *skb, struct crypto_tfm *tfm,
 		sg.offset = (unsigned long)(skb->data + offset) % PAGE_SIZE;
 		sg.length = copy;
 		
-		icv_update(tfm, &sg, 1);
+		err = icv_update(desc, &sg, copy);
+		if (unlikely(err))
+			return err;
 		
 		if ((len -= copy) == 0)
-			return;
+			return 0;
 		offset += copy;
 	}
 
@@ -519,10 +527,12 @@ void skb_icv_walk(const struct sk_buff *skb, struct crypto_tfm *tfm,
 			sg.offset = frag->page_offset + offset-start;
 			sg.length = copy;
 			
-			icv_update(tfm, &sg, 1);
+			err = icv_update(desc, &sg, copy);
+			if (unlikely(err))
+				return err;
 
 			if (!(len -= copy))
-				return;
+				return 0;
 			offset += copy;
 		}
 		start = end;
@@ -540,15 +550,19 @@ void skb_icv_walk(const struct sk_buff *skb, struct crypto_tfm *tfm,
 			if ((copy = end - offset) > 0) {
 				if (copy > len)
 					copy = len;
-				skb_icv_walk(list, tfm, offset-start, copy, icv_update);
+				err = skb_icv_walk(list, desc, offset-start,
+						   copy, icv_update);
+				if (unlikely(err))
+					return err;
 				if ((len -= copy) == 0)
-					return;
+					return 0;
 				offset += copy;
 			}
 			start = end;
 		}
 	}
 	BUG_ON(len);
+	return 0;
 }
 EXPORT_SYMBOL_GPL(skb_icv_walk);
 
-- 
cgit v1.2.3


From 1b489e11d4df82514792f9f981f31976f8a94ddf Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sun, 20 Aug 2006 15:07:14 +1000
Subject: [SCTP]: Use HMAC template and hash interface

This patch converts SCTP to use the new HMAC template and hash interface.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/sctp/constants.h |  4 ++--
 include/net/sctp/sctp.h      | 11 -----------
 include/net/sctp/structs.h   |  3 ++-
 net/sctp/endpointola.c       |  2 +-
 net/sctp/sm_make_chunk.c     | 37 +++++++++++++++++++++++++++----------
 net/sctp/socket.c            |  6 +++---
 6 files changed, 35 insertions(+), 28 deletions(-)

(limited to 'include')

diff --git a/include/net/sctp/constants.h b/include/net/sctp/constants.h
index c51541ee0247..57166bfdf8eb 100644
--- a/include/net/sctp/constants.h
+++ b/include/net/sctp/constants.h
@@ -312,9 +312,9 @@ enum { SCTP_MAX_GABS = 16 };
 				 */
 
 #if defined (CONFIG_SCTP_HMAC_MD5)
-#define SCTP_COOKIE_HMAC_ALG "md5"
+#define SCTP_COOKIE_HMAC_ALG "hmac(md5)"
 #elif defined (CONFIG_SCTP_HMAC_SHA1)
-#define SCTP_COOKIE_HMAC_ALG "sha1"
+#define SCTP_COOKIE_HMAC_ALG "hmac(sha1)"
 #else
 #define SCTP_COOKIE_HMAC_ALG NULL
 #endif
diff --git a/include/net/sctp/sctp.h b/include/net/sctp/sctp.h
index 92eae0e0f3f1..1c1abce5f6b6 100644
--- a/include/net/sctp/sctp.h
+++ b/include/net/sctp/sctp.h
@@ -330,17 +330,6 @@ static inline void sctp_v6_exit(void) { return; }
 
 #endif /* #if defined(CONFIG_IPV6) */
 
-/* Some wrappers, in case crypto not available. */
-#if defined (CONFIG_CRYPTO_HMAC)
-#define sctp_crypto_alloc_tfm crypto_alloc_tfm
-#define sctp_crypto_free_tfm crypto_free_tfm
-#define sctp_crypto_hmac crypto_hmac
-#else
-#define sctp_crypto_alloc_tfm(x...) NULL
-#define sctp_crypto_free_tfm(x...)
-#define sctp_crypto_hmac(x...)
-#endif
-
 
 /* Map an association to an assoc_id. */
 static inline sctp_assoc_t sctp_assoc2id(const struct sctp_association *asoc)
diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h
index e5aa7ff1f5b5..0412e730c765 100644
--- a/include/net/sctp/structs.h
+++ b/include/net/sctp/structs.h
@@ -87,6 +87,7 @@ struct sctp_bind_addr;
 struct sctp_ulpq;
 struct sctp_ep_common;
 struct sctp_ssnmap;
+struct crypto_hash;
 
 
 #include <net/sctp/tsnmap.h>
@@ -264,7 +265,7 @@ struct sctp_sock {
 	struct sctp_pf *pf;
 
 	/* Access to HMAC transform. */
-	struct crypto_tfm *hmac;
+	struct crypto_hash *hmac;
 
 	/* What is our base endpointer? */
 	struct sctp_endpoint *ep;
diff --git a/net/sctp/endpointola.c b/net/sctp/endpointola.c
index ffda1d680529..35c49ff2d062 100644
--- a/net/sctp/endpointola.c
+++ b/net/sctp/endpointola.c
@@ -173,7 +173,7 @@ static void sctp_endpoint_destroy(struct sctp_endpoint *ep)
 	SCTP_ASSERT(ep->base.dead, "Endpoint is not dead", return);
 
 	/* Free up the HMAC transform. */
-	sctp_crypto_free_tfm(sctp_sk(ep->base.sk)->hmac);
+	crypto_free_hash(sctp_sk(ep->base.sk)->hmac);
 
 	/* Cleanup. */
 	sctp_inq_free(&ep->base.inqueue);
diff --git a/net/sctp/sm_make_chunk.c b/net/sctp/sm_make_chunk.c
index 17b509282cf2..7745bdea7817 100644
--- a/net/sctp/sm_make_chunk.c
+++ b/net/sctp/sm_make_chunk.c
@@ -1282,10 +1282,8 @@ static sctp_cookie_param_t *sctp_pack_cookie(const struct sctp_endpoint *ep,
 
 	retval = kmalloc(*cookie_len, GFP_ATOMIC);
 
-	if (!retval) {
-		*cookie_len = 0;
+	if (!retval)
 		goto nodata;
-	}
 
 	/* Clear this memory since we are sending this data structure
 	 * out on the network.
@@ -1321,19 +1319,29 @@ static sctp_cookie_param_t *sctp_pack_cookie(const struct sctp_endpoint *ep,
 	       ntohs(init_chunk->chunk_hdr->length), raw_addrs, addrs_len);
 
   	if (sctp_sk(ep->base.sk)->hmac) {
+		struct hash_desc desc;
+
 		/* Sign the message.  */
 		sg.page = virt_to_page(&cookie->c);
 		sg.offset = (unsigned long)(&cookie->c) % PAGE_SIZE;
 		sg.length = bodysize;
 		keylen = SCTP_SECRET_SIZE;
 		key = (char *)ep->secret_key[ep->current_key];
+  		desc.tfm = sctp_sk(ep->base.sk)->hmac;
+  		desc.flags = 0;
 
-		sctp_crypto_hmac(sctp_sk(ep->base.sk)->hmac, key, &keylen,
-				 &sg, 1, cookie->signature);
+		if (crypto_hash_setkey(desc.tfm, key, keylen) ||
+		    crypto_hash_digest(&desc, &sg, bodysize, cookie->signature))
+			goto free_cookie;
 	}
 
-nodata:
 	return retval;
+
+free_cookie:
+	kfree(retval);
+nodata:
+	*cookie_len = 0;
+	return NULL;
 }
 
 /* Unpack the cookie from COOKIE ECHO chunk, recreating the association.  */
@@ -1354,6 +1362,7 @@ struct sctp_association *sctp_unpack_cookie(
 	sctp_scope_t scope;
 	struct sk_buff *skb = chunk->skb;
 	struct timeval tv;
+	struct hash_desc desc;
 
 	/* Header size is static data prior to the actual cookie, including
 	 * any padding.
@@ -1389,17 +1398,25 @@ struct sctp_association *sctp_unpack_cookie(
 	sg.offset = (unsigned long)(bear_cookie) % PAGE_SIZE;
 	sg.length = bodysize;
 	key = (char *)ep->secret_key[ep->current_key];
+	desc.tfm = sctp_sk(ep->base.sk)->hmac;
+	desc.flags = 0;
 
 	memset(digest, 0x00, SCTP_SIGNATURE_SIZE);
-	sctp_crypto_hmac(sctp_sk(ep->base.sk)->hmac, key, &keylen, &sg,
-			 1, digest);
+	if (crypto_hash_setkey(desc.tfm, key, keylen) ||
+	    crypto_hash_digest(&desc, &sg, bodysize, digest)) {
+		*error = -SCTP_IERROR_NOMEM;
+		goto fail;
+	}
 
 	if (memcmp(digest, cookie->signature, SCTP_SIGNATURE_SIZE)) {
 		/* Try the previous key. */
 		key = (char *)ep->secret_key[ep->last_key];
 		memset(digest, 0x00, SCTP_SIGNATURE_SIZE);
-		sctp_crypto_hmac(sctp_sk(ep->base.sk)->hmac, key, &keylen,
-				 &sg, 1, digest);
+		if (crypto_hash_setkey(desc.tfm, key, keylen) ||
+		    crypto_hash_digest(&desc, &sg, bodysize, digest)) {
+			*error = -SCTP_IERROR_NOMEM;
+			goto fail;
+		}
 
 		if (memcmp(digest, cookie->signature, SCTP_SIGNATURE_SIZE)) {
 			/* Yikes!  Still bad signature! */
diff --git a/net/sctp/socket.c b/net/sctp/socket.c
index dab15949958e..85caf7963886 100644
--- a/net/sctp/socket.c
+++ b/net/sctp/socket.c
@@ -4898,7 +4898,7 @@ SCTP_STATIC int sctp_stream_listen(struct sock *sk, int backlog)
 int sctp_inet_listen(struct socket *sock, int backlog)
 {
 	struct sock *sk = sock->sk;
-	struct crypto_tfm *tfm=NULL;
+	struct crypto_hash *tfm = NULL;
 	int err = -EINVAL;
 
 	if (unlikely(backlog < 0))
@@ -4911,7 +4911,7 @@ int sctp_inet_listen(struct socket *sock, int backlog)
 
 	/* Allocate HMAC for generating cookie. */
 	if (sctp_hmac_alg) {
-		tfm = sctp_crypto_alloc_tfm(sctp_hmac_alg, 0);
+		tfm = crypto_alloc_hash(sctp_hmac_alg, 0, CRYPTO_ALG_ASYNC);
 		if (!tfm) {
 			err = -ENOSYS;
 			goto out;
@@ -4937,7 +4937,7 @@ out:
 	sctp_release_sock(sk);
 	return err;
 cleanup:
-	sctp_crypto_free_tfm(tfm);
+	crypto_free_hash(tfm);
 	goto out;
 }
 
-- 
cgit v1.2.3


From 8425165dfed27945e8509c141cea245d1739e372 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sun, 20 Aug 2006 15:25:22 +1000
Subject: [CRYPTO] digest: Remove old HMAC implementation

This patch removes the old HMAC implementation now that nobody uses it
anymore.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 crypto/Kconfig         |   2 +-
 crypto/digest.c        |   3 +-
 crypto/hmac.c          | 101 -------------------------------------------------
 crypto/internal.h      |  13 -------
 include/linux/crypto.h |  16 --------
 5 files changed, 2 insertions(+), 133 deletions(-)

(limited to 'include')

diff --git a/crypto/Kconfig b/crypto/Kconfig
index f07d9237950f..1e2f39c21180 100644
--- a/crypto/Kconfig
+++ b/crypto/Kconfig
@@ -33,7 +33,7 @@ config CRYPTO_MANAGER
 	  cbc(aes).
 
 config CRYPTO_HMAC
-	bool "HMAC support"
+	tristate "HMAC support"
 	select CRYPTO_HASH
 	help
 	  HMAC: Keyed-Hashing for Message Authentication (RFC2104).
diff --git a/crypto/digest.c b/crypto/digest.c
index 5873063db840..0155a94e4b15 100644
--- a/crypto/digest.c
+++ b/crypto/digest.c
@@ -191,10 +191,9 @@ int crypto_init_digest_ops(struct crypto_tfm *tfm)
 	ops->setkey	= dalg->dia_setkey ? setkey : nosetkey;
 	ops->digestsize	= dalg->dia_digestsize;
 	
-	return crypto_alloc_hmac_block(tfm);
+	return 0;
 }
 
 void crypto_exit_digest_ops(struct crypto_tfm *tfm)
 {
-	crypto_free_hmac_block(tfm);
 }
diff --git a/crypto/hmac.c b/crypto/hmac.c
index eac77e294740..f403b6946047 100644
--- a/crypto/hmac.c
+++ b/crypto/hmac.c
@@ -29,107 +29,6 @@ struct hmac_ctx {
 	struct crypto_hash *child;
 };
 
-static void hash_key(struct crypto_tfm *tfm, u8 *key, unsigned int keylen)
-{
-	struct scatterlist tmp;
-	
-	sg_set_buf(&tmp, key, keylen);
-	crypto_digest_digest(tfm, &tmp, 1, key);
-}
-
-int crypto_alloc_hmac_block(struct crypto_tfm *tfm)
-{
-	int ret = 0;
-
-	BUG_ON(!crypto_tfm_alg_blocksize(tfm));
-	
-	tfm->crt_hash.hmac_block = kmalloc(crypto_tfm_alg_blocksize(tfm),
-					   GFP_KERNEL);
-	if (tfm->crt_hash.hmac_block == NULL)
-		ret = -ENOMEM;
-
-	return ret;
-		
-}
-
-void crypto_free_hmac_block(struct crypto_tfm *tfm)
-{
-	kfree(tfm->crt_hash.hmac_block);
-}
-
-void crypto_hmac_init(struct crypto_tfm *tfm, u8 *key, unsigned int *keylen)
-{
-	unsigned int i;
-	struct scatterlist tmp;
-	char *ipad = tfm->crt_hash.hmac_block;
-	
-	if (*keylen > crypto_tfm_alg_blocksize(tfm)) {
-		hash_key(tfm, key, *keylen);
-		*keylen = crypto_tfm_alg_digestsize(tfm);
-	}
-
-	memset(ipad, 0, crypto_tfm_alg_blocksize(tfm));
-	memcpy(ipad, key, *keylen);
-
-	for (i = 0; i < crypto_tfm_alg_blocksize(tfm); i++)
-		ipad[i] ^= 0x36;
-
-	sg_set_buf(&tmp, ipad, crypto_tfm_alg_blocksize(tfm));
-	
-	crypto_digest_init(tfm);
-	crypto_digest_update(tfm, &tmp, 1);
-}
-
-void crypto_hmac_update(struct crypto_tfm *tfm,
-                        struct scatterlist *sg, unsigned int nsg)
-{
-	crypto_digest_update(tfm, sg, nsg);
-}
-
-void crypto_hmac_final(struct crypto_tfm *tfm, u8 *key,
-                       unsigned int *keylen, u8 *out)
-{
-	unsigned int i;
-	struct scatterlist tmp;
-	char *opad = tfm->crt_hash.hmac_block;
-	
-	if (*keylen > crypto_tfm_alg_blocksize(tfm)) {
-		hash_key(tfm, key, *keylen);
-		*keylen = crypto_tfm_alg_digestsize(tfm);
-	}
-
-	crypto_digest_final(tfm, out);
-
-	memset(opad, 0, crypto_tfm_alg_blocksize(tfm));
-	memcpy(opad, key, *keylen);
-		
-	for (i = 0; i < crypto_tfm_alg_blocksize(tfm); i++)
-		opad[i] ^= 0x5c;
-
-	sg_set_buf(&tmp, opad, crypto_tfm_alg_blocksize(tfm));
-
-	crypto_digest_init(tfm);
-	crypto_digest_update(tfm, &tmp, 1);
-	
-	sg_set_buf(&tmp, out, crypto_tfm_alg_digestsize(tfm));
-	
-	crypto_digest_update(tfm, &tmp, 1);
-	crypto_digest_final(tfm, out);
-}
-
-void crypto_hmac(struct crypto_tfm *tfm, u8 *key, unsigned int *keylen,
-                 struct scatterlist *sg, unsigned int nsg, u8 *out)
-{
-	crypto_hmac_init(tfm, key, keylen);
-	crypto_hmac_update(tfm, sg, nsg);
-	crypto_hmac_final(tfm, key, keylen, out);
-}
-
-EXPORT_SYMBOL_GPL(crypto_hmac_init);
-EXPORT_SYMBOL_GPL(crypto_hmac_update);
-EXPORT_SYMBOL_GPL(crypto_hmac_final);
-EXPORT_SYMBOL_GPL(crypto_hmac);
-
 static inline void *align_ptr(void *p, unsigned int align)
 {
 	return (void *)ALIGN((unsigned long)p, align);
diff --git a/crypto/internal.h b/crypto/internal.h
index 93d9b10ff914..2da6ad4f3593 100644
--- a/crypto/internal.h
+++ b/crypto/internal.h
@@ -73,19 +73,6 @@ static inline void crypto_yield(u32 flags)
 		cond_resched();
 }
 
-#ifdef CONFIG_CRYPTO_HMAC
-int crypto_alloc_hmac_block(struct crypto_tfm *tfm);
-void crypto_free_hmac_block(struct crypto_tfm *tfm);
-#else
-static inline int crypto_alloc_hmac_block(struct crypto_tfm *tfm)
-{
-	return 0;
-}
-
-static inline void crypto_free_hmac_block(struct crypto_tfm *tfm)
-{ }
-#endif
-
 #ifdef CONFIG_PROC_FS
 void __init crypto_init_proc(void);
 void __exit crypto_exit_proc(void);
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index 40c0aab8ad4c..929fb9ad1314 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -295,9 +295,6 @@ struct hash_tfm {
 		      unsigned int nsg, u8 *out);
 	int (*setkey)(struct crypto_hash *tfm, const u8 *key,
 		      unsigned int keylen);
-#ifdef CONFIG_CRYPTO_HMAC
-	void *hmac_block;
-#endif
 	unsigned int digestsize;
 };
 
@@ -872,18 +869,5 @@ static inline int crypto_comp_decompress(struct crypto_tfm *tfm,
 	return tfm->crt_compress.cot_decompress(tfm, src, slen, dst, dlen);
 }
 
-/*
- * HMAC support.
- */
-#ifdef CONFIG_CRYPTO_HMAC
-void crypto_hmac_init(struct crypto_tfm *tfm, u8 *key, unsigned int *keylen);
-void crypto_hmac_update(struct crypto_tfm *tfm,
-                        struct scatterlist *sg, unsigned int nsg);
-void crypto_hmac_final(struct crypto_tfm *tfm, u8 *key,
-                       unsigned int *keylen, u8 *out);
-void crypto_hmac(struct crypto_tfm *tfm, u8 *key, unsigned int *keylen,
-                 struct scatterlist *sg, unsigned int nsg, u8 *out);
-#endif	/* CONFIG_CRYPTO_HMAC */
-
 #endif	/* _LINUX_CRYPTO_H */
 
-- 
cgit v1.2.3


From fce32d70ba834129b164c40c2d4260e5a7a7d850 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sat, 26 Aug 2006 17:35:45 +1000
Subject: [CRYPTO] api: Add crypto_comp and crypto_has_*

This patch adds the crypto_comp type to complete the compile-time checking
conversion.  The functions crypto_has_alg and crypto_has_cipher, etc. are
also added to replace crypto_alg_available.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 crypto/api.c           | 14 ++++++++
 include/linux/crypto.h | 90 ++++++++++++++++++++++++++++++++++++++++++++++----
 2 files changed, 98 insertions(+), 6 deletions(-)

(limited to 'include')

diff --git a/crypto/api.c b/crypto/api.c
index edaa843d8e83..2e84d4b54790 100644
--- a/crypto/api.c
+++ b/crypto/api.c
@@ -482,3 +482,17 @@ int crypto_alg_available(const char *name, u32 flags)
 EXPORT_SYMBOL_GPL(crypto_alloc_tfm);
 EXPORT_SYMBOL_GPL(crypto_free_tfm);
 EXPORT_SYMBOL_GPL(crypto_alg_available);
+
+int crypto_has_alg(const char *name, u32 type, u32 mask)
+{
+	int ret = 0;
+	struct crypto_alg *alg = crypto_alg_mod_lookup(name, type, mask);
+	
+	if (!IS_ERR(alg)) {
+		crypto_mod_put(alg);
+		ret = 1;
+	}
+	
+	return ret;
+}
+EXPORT_SYMBOL_GPL(crypto_has_alg);
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index 929fb9ad1314..cf91c4c0638b 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -236,11 +236,17 @@ int crypto_unregister_alg(struct crypto_alg *alg);
  */
 #ifdef CONFIG_CRYPTO
 int crypto_alg_available(const char *name, u32 flags);
+int crypto_has_alg(const char *name, u32 type, u32 mask);
 #else
 static inline int crypto_alg_available(const char *name, u32 flags)
 {
 	return 0;
 }
+
+static inline int crypto_has_alg(const char *name, u32 type, u32 mask)
+{
+	return 0;
+}
 #endif
 
 /*
@@ -329,6 +335,7 @@ struct crypto_tfm {
 };
 
 #define crypto_cipher crypto_tfm
+#define crypto_comp crypto_tfm
 
 struct crypto_blkcipher {
 	struct crypto_tfm base;
@@ -485,6 +492,15 @@ static inline void crypto_free_blkcipher(struct crypto_blkcipher *tfm)
 	crypto_free_tfm(crypto_blkcipher_tfm(tfm));
 }
 
+static inline int crypto_has_blkcipher(const char *alg_name, u32 type, u32 mask)
+{
+	type &= ~CRYPTO_ALG_TYPE_MASK;
+	type |= CRYPTO_ALG_TYPE_BLKCIPHER;
+	mask |= CRYPTO_ALG_TYPE_MASK;
+
+	return crypto_has_alg(alg_name, type, mask);
+}
+
 static inline const char *crypto_blkcipher_name(struct crypto_blkcipher *tfm)
 {
 	return crypto_tfm_alg_name(crypto_blkcipher_tfm(tfm));
@@ -620,6 +636,15 @@ static inline void crypto_free_cipher(struct crypto_cipher *tfm)
 	crypto_free_tfm(crypto_cipher_tfm(tfm));
 }
 
+static inline int crypto_has_cipher(const char *alg_name, u32 type, u32 mask)
+{
+	type &= ~CRYPTO_ALG_TYPE_MASK;
+	type |= CRYPTO_ALG_TYPE_CIPHER;
+	mask |= CRYPTO_ALG_TYPE_MASK;
+
+	return crypto_has_alg(alg_name, type, mask);
+}
+
 static inline struct cipher_tfm *crypto_cipher_crt(struct crypto_cipher *tfm)
 {
 	return &crypto_cipher_tfm(tfm)->crt_cipher;
@@ -718,6 +743,15 @@ static inline void crypto_free_hash(struct crypto_hash *tfm)
 	crypto_free_tfm(crypto_hash_tfm(tfm));
 }
 
+static inline int crypto_has_hash(const char *alg_name, u32 type, u32 mask)
+{
+	type &= ~CRYPTO_ALG_TYPE_MASK;
+	type |= CRYPTO_ALG_TYPE_HASH;
+	mask |= CRYPTO_ALG_TYPE_HASH_MASK;
+
+	return crypto_has_alg(alg_name, type, mask);
+}
+
 static inline struct hash_tfm *crypto_hash_crt(struct crypto_hash *tfm)
 {
 	return &crypto_hash_tfm(tfm)->crt_hash;
@@ -853,20 +887,64 @@ static inline void crypto_cipher_get_iv(struct crypto_tfm *tfm,
 	memcpy(dst, tfm->crt_cipher.cit_iv, len);
 }
 
-static inline int crypto_comp_compress(struct crypto_tfm *tfm,
+static inline struct crypto_comp *__crypto_comp_cast(struct crypto_tfm *tfm)
+{
+	return (struct crypto_comp *)tfm;
+}
+
+static inline struct crypto_comp *crypto_comp_cast(struct crypto_tfm *tfm)
+{
+	BUG_ON((crypto_tfm_alg_type(tfm) ^ CRYPTO_ALG_TYPE_COMPRESS) &
+	       CRYPTO_ALG_TYPE_MASK);
+	return __crypto_comp_cast(tfm);
+}
+
+static inline struct crypto_comp *crypto_alloc_comp(const char *alg_name,
+						    u32 type, u32 mask)
+{
+	type &= ~CRYPTO_ALG_TYPE_MASK;
+	type |= CRYPTO_ALG_TYPE_COMPRESS;
+	mask |= CRYPTO_ALG_TYPE_MASK;
+
+	return __crypto_comp_cast(crypto_alloc_base(alg_name, type, mask));
+}
+
+static inline struct crypto_tfm *crypto_comp_tfm(struct crypto_comp *tfm)
+{
+	return tfm;
+}
+
+static inline void crypto_free_comp(struct crypto_comp *tfm)
+{
+	crypto_free_tfm(crypto_comp_tfm(tfm));
+}
+
+static inline int crypto_has_comp(const char *alg_name, u32 type, u32 mask)
+{
+	type &= ~CRYPTO_ALG_TYPE_MASK;
+	type |= CRYPTO_ALG_TYPE_COMPRESS;
+	mask |= CRYPTO_ALG_TYPE_MASK;
+
+	return crypto_has_alg(alg_name, type, mask);
+}
+
+static inline struct compress_tfm *crypto_comp_crt(struct crypto_comp *tfm)
+{
+	return &crypto_comp_tfm(tfm)->crt_compress;
+}
+
+static inline int crypto_comp_compress(struct crypto_comp *tfm,
                                        const u8 *src, unsigned int slen,
                                        u8 *dst, unsigned int *dlen)
 {
-	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_COMPRESS);
-	return tfm->crt_compress.cot_compress(tfm, src, slen, dst, dlen);
+	return crypto_comp_crt(tfm)->cot_compress(tfm, src, slen, dst, dlen);
 }
 
-static inline int crypto_comp_decompress(struct crypto_tfm *tfm,
+static inline int crypto_comp_decompress(struct crypto_comp *tfm,
                                          const u8 *src, unsigned int slen,
                                          u8 *dst, unsigned int *dlen)
 {
-	BUG_ON(crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_COMPRESS);
-	return tfm->crt_compress.cot_decompress(tfm, src, slen, dst, dlen);
+	return crypto_comp_crt(tfm)->cot_decompress(tfm, src, slen, dst, dlen);
 }
 
 #endif	/* _LINUX_CRYPTO_H */
-- 
cgit v1.2.3


From e4d5b79c661c7cfca9d8d5afd040a295f128d3cb Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sat, 26 Aug 2006 18:12:40 +1000
Subject: [CRYPTO] users: Use crypto_comp and crypto_has_*

This patch converts all users to use the new crypto_comp type and the
crypto_has_* functions.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 crypto/tcrypt.c          |  8 ++++----
 drivers/crypto/padlock.c |  6 +++---
 drivers/net/ppp_mppe.c   |  4 ++--
 include/linux/crypto.h   |  5 +++++
 include/net/ipcomp.h     |  5 ++---
 net/ipv4/ipcomp.c        | 25 +++++++++++++------------
 net/ipv6/ipcomp6.c       | 25 +++++++++++++------------
 net/xfrm/xfrm_algo.c     | 27 ++++++++++++++++++---------
 8 files changed, 60 insertions(+), 45 deletions(-)

(limited to 'include')

diff --git a/crypto/tcrypt.c b/crypto/tcrypt.c
index 840ab8be0b96..83307420d31c 100644
--- a/crypto/tcrypt.c
+++ b/crypto/tcrypt.c
@@ -749,7 +749,7 @@ static void test_deflate(void)
 {
 	unsigned int i;
 	char result[COMP_BUF_SIZE];
-	struct crypto_tfm *tfm;
+	struct crypto_comp *tfm;
 	struct comp_testvec *tv;
 	unsigned int tsize;
 
@@ -821,7 +821,7 @@ static void test_deflate(void)
 		       ilen, dlen);
 	}
 out:
-	crypto_free_tfm(tfm);
+	crypto_free_comp(tfm);
 }
 
 static void test_available(void)
@@ -830,8 +830,8 @@ static void test_available(void)
 
 	while (*name) {
 		printk("alg %s ", *name);
-		printk((crypto_alg_available(*name, 0)) ?
-			"found\n" : "not found\n");
+		printk(crypto_has_alg(*name, 0, CRYPTO_ALG_ASYNC) ?
+		       "found\n" : "not found\n");
 		name++;
 	}
 }
diff --git a/drivers/crypto/padlock.c b/drivers/crypto/padlock.c
index ce581684f4b4..d6d7dd5bb98c 100644
--- a/drivers/crypto/padlock.c
+++ b/drivers/crypto/padlock.c
@@ -26,13 +26,13 @@ static int __init padlock_init(void)
 {
 	int success = 0;
 
-	if (crypto_alg_available("aes-padlock", 0))
+	if (crypto_has_cipher("aes-padlock", 0, 0))
 		success++;
 
-	if (crypto_alg_available("sha1-padlock", 0))
+	if (crypto_has_hash("sha1-padlock", 0, 0))
 		success++;
 
-	if (crypto_alg_available("sha256-padlock", 0))
+	if (crypto_has_hash("sha256-padlock", 0, 0))
 		success++;
 
 	if (!success) {
diff --git a/drivers/net/ppp_mppe.c b/drivers/net/ppp_mppe.c
index e7a0eb4fca60..f3655fd772f5 100644
--- a/drivers/net/ppp_mppe.c
+++ b/drivers/net/ppp_mppe.c
@@ -710,8 +710,8 @@ static struct compressor ppp_mppe = {
 static int __init ppp_mppe_init(void)
 {
 	int answer;
-	if (!(crypto_alg_available("ecb(arc4)", 0) &&
-	      crypto_alg_available("sha1", 0)))
+	if (!(crypto_has_blkcipher("ecb(arc4)", 0, CRYPTO_ALG_ASYNC) &&
+	      crypto_has_hash("sha1", 0, CRYPTO_ALG_ASYNC)))
 		return -ENODEV;
 
 	sha_pad = kmalloc(sizeof(struct sha_pad), GFP_KERNEL);
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index cf91c4c0638b..d4f9948b64b1 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -928,6 +928,11 @@ static inline int crypto_has_comp(const char *alg_name, u32 type, u32 mask)
 	return crypto_has_alg(alg_name, type, mask);
 }
 
+static inline const char *crypto_comp_name(struct crypto_comp *tfm)
+{
+	return crypto_tfm_alg_name(crypto_comp_tfm(tfm));
+}
+
 static inline struct compress_tfm *crypto_comp_crt(struct crypto_comp *tfm)
 {
 	return &crypto_comp_tfm(tfm)->crt_compress;
diff --git a/include/net/ipcomp.h b/include/net/ipcomp.h
index b94e3047b4d9..87c1af3e5e82 100644
--- a/include/net/ipcomp.h
+++ b/include/net/ipcomp.h
@@ -1,15 +1,14 @@
 #ifndef _NET_IPCOMP_H
 #define _NET_IPCOMP_H
 
+#include <linux/crypto.h>
 #include <linux/types.h>
 
 #define IPCOMP_SCRATCH_SIZE     65400
 
-struct crypto_tfm;
-
 struct ipcomp_data {
 	u16 threshold;
-	struct crypto_tfm **tfms;
+	struct crypto_comp **tfms;
 };
 
 #endif
diff --git a/net/ipv4/ipcomp.c b/net/ipv4/ipcomp.c
index a0c28b2b756e..5bb9c9f03fb6 100644
--- a/net/ipv4/ipcomp.c
+++ b/net/ipv4/ipcomp.c
@@ -32,7 +32,7 @@
 
 struct ipcomp_tfms {
 	struct list_head list;
-	struct crypto_tfm **tfms;
+	struct crypto_comp **tfms;
 	int users;
 };
 
@@ -46,7 +46,7 @@ static int ipcomp_decompress(struct xfrm_state *x, struct sk_buff *skb)
 	int err, plen, dlen;
 	struct ipcomp_data *ipcd = x->data;
 	u8 *start, *scratch;
-	struct crypto_tfm *tfm;
+	struct crypto_comp *tfm;
 	int cpu;
 	
 	plen = skb->len;
@@ -107,7 +107,7 @@ static int ipcomp_compress(struct xfrm_state *x, struct sk_buff *skb)
 	struct iphdr *iph = skb->nh.iph;
 	struct ipcomp_data *ipcd = x->data;
 	u8 *start, *scratch;
-	struct crypto_tfm *tfm;
+	struct crypto_comp *tfm;
 	int cpu;
 	
 	ihlen = iph->ihl * 4;
@@ -302,7 +302,7 @@ static void **ipcomp_alloc_scratches(void)
 	return scratches;
 }
 
-static void ipcomp_free_tfms(struct crypto_tfm **tfms)
+static void ipcomp_free_tfms(struct crypto_comp **tfms)
 {
 	struct ipcomp_tfms *pos;
 	int cpu;
@@ -324,28 +324,28 @@ static void ipcomp_free_tfms(struct crypto_tfm **tfms)
 		return;
 
 	for_each_possible_cpu(cpu) {
-		struct crypto_tfm *tfm = *per_cpu_ptr(tfms, cpu);
-		crypto_free_tfm(tfm);
+		struct crypto_comp *tfm = *per_cpu_ptr(tfms, cpu);
+		crypto_free_comp(tfm);
 	}
 	free_percpu(tfms);
 }
 
-static struct crypto_tfm **ipcomp_alloc_tfms(const char *alg_name)
+static struct crypto_comp **ipcomp_alloc_tfms(const char *alg_name)
 {
 	struct ipcomp_tfms *pos;
-	struct crypto_tfm **tfms;
+	struct crypto_comp **tfms;
 	int cpu;
 
 	/* This can be any valid CPU ID so we don't need locking. */
 	cpu = raw_smp_processor_id();
 
 	list_for_each_entry(pos, &ipcomp_tfms_list, list) {
-		struct crypto_tfm *tfm;
+		struct crypto_comp *tfm;
 
 		tfms = pos->tfms;
 		tfm = *per_cpu_ptr(tfms, cpu);
 
-		if (!strcmp(crypto_tfm_alg_name(tfm), alg_name)) {
+		if (!strcmp(crypto_comp_name(tfm), alg_name)) {
 			pos->users++;
 			return tfms;
 		}
@@ -359,12 +359,13 @@ static struct crypto_tfm **ipcomp_alloc_tfms(const char *alg_name)
 	INIT_LIST_HEAD(&pos->list);
 	list_add(&pos->list, &ipcomp_tfms_list);
 
-	pos->tfms = tfms = alloc_percpu(struct crypto_tfm *);
+	pos->tfms = tfms = alloc_percpu(struct crypto_comp *);
 	if (!tfms)
 		goto error;
 
 	for_each_possible_cpu(cpu) {
-		struct crypto_tfm *tfm = crypto_alloc_tfm(alg_name, 0);
+		struct crypto_comp *tfm = crypto_alloc_comp(alg_name, 0,
+							    CRYPTO_ALG_ASYNC);
 		if (!tfm)
 			goto error;
 		*per_cpu_ptr(tfms, cpu) = tfm;
diff --git a/net/ipv6/ipcomp6.c b/net/ipv6/ipcomp6.c
index 7e4d1c17bfbc..a81e9e9d93bd 100644
--- a/net/ipv6/ipcomp6.c
+++ b/net/ipv6/ipcomp6.c
@@ -53,7 +53,7 @@
 
 struct ipcomp6_tfms {
 	struct list_head list;
-	struct crypto_tfm **tfms;
+	struct crypto_comp **tfms;
 	int users;
 };
 
@@ -70,7 +70,7 @@ static int ipcomp6_input(struct xfrm_state *x, struct sk_buff *skb)
 	int plen, dlen;
 	struct ipcomp_data *ipcd = x->data;
 	u8 *start, *scratch;
-	struct crypto_tfm *tfm;
+	struct crypto_comp *tfm;
 	int cpu;
 
 	if (skb_linearize_cow(skb))
@@ -129,7 +129,7 @@ static int ipcomp6_output(struct xfrm_state *x, struct sk_buff *skb)
 	struct ipcomp_data *ipcd = x->data;
 	int plen, dlen;
 	u8 *start, *scratch;
-	struct crypto_tfm *tfm;
+	struct crypto_comp *tfm;
 	int cpu;
 
 	hdr_len = skb->h.raw - skb->data;
@@ -301,7 +301,7 @@ static void **ipcomp6_alloc_scratches(void)
 	return scratches;
 }
 
-static void ipcomp6_free_tfms(struct crypto_tfm **tfms)
+static void ipcomp6_free_tfms(struct crypto_comp **tfms)
 {
 	struct ipcomp6_tfms *pos;
 	int cpu;
@@ -323,28 +323,28 @@ static void ipcomp6_free_tfms(struct crypto_tfm **tfms)
 		return;
 
 	for_each_possible_cpu(cpu) {
-		struct crypto_tfm *tfm = *per_cpu_ptr(tfms, cpu);
-		crypto_free_tfm(tfm);
+		struct crypto_comp *tfm = *per_cpu_ptr(tfms, cpu);
+		crypto_free_comp(tfm);
 	}
 	free_percpu(tfms);
 }
 
-static struct crypto_tfm **ipcomp6_alloc_tfms(const char *alg_name)
+static struct crypto_comp **ipcomp6_alloc_tfms(const char *alg_name)
 {
 	struct ipcomp6_tfms *pos;
-	struct crypto_tfm **tfms;
+	struct crypto_comp **tfms;
 	int cpu;
 
 	/* This can be any valid CPU ID so we don't need locking. */
 	cpu = raw_smp_processor_id();
 
 	list_for_each_entry(pos, &ipcomp6_tfms_list, list) {
-		struct crypto_tfm *tfm;
+		struct crypto_comp *tfm;
 
 		tfms = pos->tfms;
 		tfm = *per_cpu_ptr(tfms, cpu);
 
-		if (!strcmp(crypto_tfm_alg_name(tfm), alg_name)) {
+		if (!strcmp(crypto_comp_name(tfm), alg_name)) {
 			pos->users++;
 			return tfms;
 		}
@@ -358,12 +358,13 @@ static struct crypto_tfm **ipcomp6_alloc_tfms(const char *alg_name)
 	INIT_LIST_HEAD(&pos->list);
 	list_add(&pos->list, &ipcomp6_tfms_list);
 
-	pos->tfms = tfms = alloc_percpu(struct crypto_tfm *);
+	pos->tfms = tfms = alloc_percpu(struct crypto_comp *);
 	if (!tfms)
 		goto error;
 
 	for_each_possible_cpu(cpu) {
-		struct crypto_tfm *tfm = crypto_alloc_tfm(alg_name, 0);
+		struct crypto_comp *tfm = crypto_alloc_comp(alg_name, 0,
+							    CRYPTO_ALG_ASYNC);
 		if (!tfm)
 			goto error;
 		*per_cpu_ptr(tfms, cpu) = tfm;
diff --git a/net/xfrm/xfrm_algo.c b/net/xfrm/xfrm_algo.c
index 87918f281bb4..5a0dbeb6bbe8 100644
--- a/net/xfrm/xfrm_algo.c
+++ b/net/xfrm/xfrm_algo.c
@@ -363,8 +363,8 @@ struct xfrm_algo_desc *xfrm_calg_get_byid(int alg_id)
 EXPORT_SYMBOL_GPL(xfrm_calg_get_byid);
 
 static struct xfrm_algo_desc *xfrm_get_byname(struct xfrm_algo_desc *list,
-					      int entries, char *name,
-					      int probe)
+					      int entries, u32 type, u32 mask,
+					      char *name, int probe)
 {
 	int i, status;
 
@@ -382,7 +382,7 @@ static struct xfrm_algo_desc *xfrm_get_byname(struct xfrm_algo_desc *list,
 		if (!probe)
 			break;
 
-		status = crypto_alg_available(name, 0);
+		status = crypto_has_alg(name, type, mask | CRYPTO_ALG_ASYNC);
 		if (!status)
 			break;
 
@@ -394,19 +394,25 @@ static struct xfrm_algo_desc *xfrm_get_byname(struct xfrm_algo_desc *list,
 
 struct xfrm_algo_desc *xfrm_aalg_get_byname(char *name, int probe)
 {
-	return xfrm_get_byname(aalg_list, aalg_entries(), name, probe);
+	return xfrm_get_byname(aalg_list, aalg_entries(),
+			       CRYPTO_ALG_TYPE_HASH, CRYPTO_ALG_TYPE_HASH_MASK,
+			       name, probe);
 }
 EXPORT_SYMBOL_GPL(xfrm_aalg_get_byname);
 
 struct xfrm_algo_desc *xfrm_ealg_get_byname(char *name, int probe)
 {
-	return xfrm_get_byname(ealg_list, ealg_entries(), name, probe);
+	return xfrm_get_byname(ealg_list, ealg_entries(),
+			       CRYPTO_ALG_TYPE_BLKCIPHER, CRYPTO_ALG_TYPE_MASK,
+			       name, probe);
 }
 EXPORT_SYMBOL_GPL(xfrm_ealg_get_byname);
 
 struct xfrm_algo_desc *xfrm_calg_get_byname(char *name, int probe)
 {
-	return xfrm_get_byname(calg_list, calg_entries(), name, probe);
+	return xfrm_get_byname(calg_list, calg_entries(),
+			       CRYPTO_ALG_TYPE_COMPRESS, CRYPTO_ALG_TYPE_MASK,
+			       name, probe);
 }
 EXPORT_SYMBOL_GPL(xfrm_calg_get_byname);
 
@@ -441,19 +447,22 @@ void xfrm_probe_algs(void)
 	BUG_ON(in_softirq());
 
 	for (i = 0; i < aalg_entries(); i++) {
-		status = crypto_alg_available(aalg_list[i].name, 0);
+		status = crypto_has_hash(aalg_list[i].name, 0,
+					 CRYPTO_ALG_ASYNC);
 		if (aalg_list[i].available != status)
 			aalg_list[i].available = status;
 	}
 	
 	for (i = 0; i < ealg_entries(); i++) {
-		status = crypto_alg_available(ealg_list[i].name, 0);
+		status = crypto_has_blkcipher(ealg_list[i].name, 0,
+					      CRYPTO_ALG_ASYNC);
 		if (ealg_list[i].available != status)
 			ealg_list[i].available = status;
 	}
 	
 	for (i = 0; i < calg_entries(); i++) {
-		status = crypto_alg_available(calg_list[i].name, 0);
+		status = crypto_has_comp(calg_list[i].name, 0,
+					 CRYPTO_ALG_ASYNC);
 		if (calg_list[i].available != status)
 			calg_list[i].available = status;
 	}
-- 
cgit v1.2.3


From 6010439f47e6b308c031dad7d99686030ef942dd Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sat, 26 Aug 2006 18:34:10 +1000
Subject: [CRYPTO] padlock: Convert padlock-sha to use crypto_hash

This patch converts padlock-sha to use crypto_hash for its fallback.
It also changes the fallback selection to use selection by type instead
of name.  This is done through the new CRYPTO_ALG_NEED_FALLBACK bit,
which is set if and only if an algorithm needs a fallback of the same
type.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 drivers/crypto/padlock-sha.c | 91 +++++++++++++++++---------------------------
 include/linux/crypto.h       |  6 +++
 2 files changed, 41 insertions(+), 56 deletions(-)

(limited to 'include')

diff --git a/drivers/crypto/padlock-sha.c b/drivers/crypto/padlock-sha.c
index b028db61c301..a781fd23b607 100644
--- a/drivers/crypto/padlock-sha.c
+++ b/drivers/crypto/padlock-sha.c
@@ -12,10 +12,11 @@
  *
  */
 
+#include <crypto/algapi.h>
+#include <linux/err.h>
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/errno.h>
-#include <linux/crypto.h>
 #include <linux/cryptohash.h>
 #include <linux/interrupt.h>
 #include <linux/kernel.h>
@@ -30,28 +31,17 @@
 #define SHA256_DIGEST_SIZE      32
 #define SHA256_HMAC_BLOCK_SIZE  64
 
-static char *sha1_fallback = SHA1_DEFAULT_FALLBACK;
-static char *sha256_fallback = SHA256_DEFAULT_FALLBACK;
-
-module_param(sha1_fallback, charp, 0644);
-module_param(sha256_fallback, charp, 0644);
-
-MODULE_PARM_DESC(sha1_fallback, "Fallback driver for SHA1. Default is "
-		 SHA1_DEFAULT_FALLBACK);
-MODULE_PARM_DESC(sha256_fallback, "Fallback driver for SHA256. Default is "
-		 SHA256_DEFAULT_FALLBACK);
-
 struct padlock_sha_ctx {
 	char		*data;
 	size_t		used;
 	int		bypass;
 	void (*f_sha_padlock)(const char *in, char *out, int count);
-	struct crypto_tfm *fallback_tfm;
+	struct hash_desc fallback;
 };
 
 static inline struct padlock_sha_ctx *ctx(struct crypto_tfm *tfm)
 {
-	return (struct padlock_sha_ctx *)(crypto_tfm_ctx(tfm));
+	return crypto_tfm_ctx(tfm);
 }
 
 /* We'll need aligned address on the stack */
@@ -65,14 +55,12 @@ static void padlock_sha_bypass(struct crypto_tfm *tfm)
 	if (ctx(tfm)->bypass)
 		return;
 
-	BUG_ON(!ctx(tfm)->fallback_tfm);
-
-	crypto_digest_init(ctx(tfm)->fallback_tfm);
+	crypto_hash_init(&ctx(tfm)->fallback);
 	if (ctx(tfm)->data && ctx(tfm)->used) {
 		struct scatterlist sg;
 
 		sg_set_buf(&sg, ctx(tfm)->data, ctx(tfm)->used);
-		crypto_digest_update(ctx(tfm)->fallback_tfm, &sg, 1);
+		crypto_hash_update(&ctx(tfm)->fallback, &sg, sg.length);
 	}
 
 	ctx(tfm)->used = 0;
@@ -95,9 +83,8 @@ static void padlock_sha_update(struct crypto_tfm *tfm,
 
 	if (unlikely(ctx(tfm)->bypass)) {
 		struct scatterlist sg;
-		BUG_ON(!ctx(tfm)->fallback_tfm);
 		sg_set_buf(&sg, (uint8_t *)data, length);
-		crypto_digest_update(ctx(tfm)->fallback_tfm, &sg, 1);
+		crypto_hash_update(&ctx(tfm)->fallback, &sg, length);
 		return;
 	}
 
@@ -160,8 +147,7 @@ static void padlock_do_sha256(const char *in, char *out, int count)
 static void padlock_sha_final(struct crypto_tfm *tfm, uint8_t *out)
 {
 	if (unlikely(ctx(tfm)->bypass)) {
-		BUG_ON(!ctx(tfm)->fallback_tfm);
-		crypto_digest_final(ctx(tfm)->fallback_tfm, out);
+		crypto_hash_final(&ctx(tfm)->fallback, out);
 		ctx(tfm)->bypass = 0;
 		return;
 	}
@@ -172,8 +158,11 @@ static void padlock_sha_final(struct crypto_tfm *tfm, uint8_t *out)
 	ctx(tfm)->used = 0;
 }
 
-static int padlock_cra_init(struct crypto_tfm *tfm, const char *fallback_driver_name)
+static int padlock_cra_init(struct crypto_tfm *tfm)
 {
+	const char *fallback_driver_name = tfm->__crt_alg->cra_name;
+	struct crypto_hash *fallback_tfm;
+
 	/* For now we'll allocate one page. This
 	 * could eventually be configurable one day. */
 	ctx(tfm)->data = (char *)__get_free_page(GFP_KERNEL);
@@ -181,14 +170,17 @@ static int padlock_cra_init(struct crypto_tfm *tfm, const char *fallback_driver_
 		return -ENOMEM;
 
 	/* Allocate a fallback and abort if it failed. */
-	ctx(tfm)->fallback_tfm = crypto_alloc_tfm(fallback_driver_name, 0);
-	if (!ctx(tfm)->fallback_tfm) {
+	fallback_tfm = crypto_alloc_hash(fallback_driver_name, 0,
+					 CRYPTO_ALG_ASYNC |
+					 CRYPTO_ALG_NEED_FALLBACK);
+	if (IS_ERR(fallback_tfm)) {
 		printk(KERN_WARNING PFX "Fallback driver '%s' could not be loaded!\n",
 		       fallback_driver_name);
 		free_page((unsigned long)(ctx(tfm)->data));
-		return -ENOENT;
+		return PTR_ERR(fallback_tfm);
 	}
 
+	ctx(tfm)->fallback.tfm = fallback_tfm;
 	return 0;
 }
 
@@ -196,14 +188,14 @@ static int padlock_sha1_cra_init(struct crypto_tfm *tfm)
 {
 	ctx(tfm)->f_sha_padlock = padlock_do_sha1;
 
-	return padlock_cra_init(tfm, sha1_fallback);
+	return padlock_cra_init(tfm);
 }
 
 static int padlock_sha256_cra_init(struct crypto_tfm *tfm)
 {
 	ctx(tfm)->f_sha_padlock = padlock_do_sha256;
 
-	return padlock_cra_init(tfm, sha256_fallback);
+	return padlock_cra_init(tfm);
 }
 
 static void padlock_cra_exit(struct crypto_tfm *tfm)
@@ -213,16 +205,16 @@ static void padlock_cra_exit(struct crypto_tfm *tfm)
 		ctx(tfm)->data = NULL;
 	}
 
-	BUG_ON(!ctx(tfm)->fallback_tfm);
-	crypto_free_tfm(ctx(tfm)->fallback_tfm);
-	ctx(tfm)->fallback_tfm = NULL;
+	crypto_free_hash(ctx(tfm)->fallback.tfm);
+	ctx(tfm)->fallback.tfm = NULL;
 }
 
 static struct crypto_alg sha1_alg = {
 	.cra_name		=	"sha1",
 	.cra_driver_name	=	"sha1-padlock",
 	.cra_priority		=	PADLOCK_CRA_PRIORITY,
-	.cra_flags		=	CRYPTO_ALG_TYPE_DIGEST,
+	.cra_flags		=	CRYPTO_ALG_TYPE_DIGEST |
+					CRYPTO_ALG_NEED_FALLBACK,
 	.cra_blocksize		=	SHA1_HMAC_BLOCK_SIZE,
 	.cra_ctxsize		=	sizeof(struct padlock_sha_ctx),
 	.cra_module		=	THIS_MODULE,
@@ -243,7 +235,8 @@ static struct crypto_alg sha256_alg = {
 	.cra_name		=	"sha256",
 	.cra_driver_name	=	"sha256-padlock",
 	.cra_priority		=	PADLOCK_CRA_PRIORITY,
-	.cra_flags		=	CRYPTO_ALG_TYPE_DIGEST,
+	.cra_flags		=	CRYPTO_ALG_TYPE_DIGEST |
+					CRYPTO_ALG_NEED_FALLBACK,
 	.cra_blocksize		=	SHA256_HMAC_BLOCK_SIZE,
 	.cra_ctxsize		=	sizeof(struct padlock_sha_ctx),
 	.cra_module		=	THIS_MODULE,
@@ -262,29 +255,15 @@ static struct crypto_alg sha256_alg = {
 
 static void __init padlock_sha_check_fallbacks(void)
 {
-	struct crypto_tfm *tfm;
-
-	/* We'll try to allocate one TFM for each fallback
-	 * to test that the modules are available. */
-	tfm = crypto_alloc_tfm(sha1_fallback, 0);
-	if (!tfm) {
-		printk(KERN_WARNING PFX "Couldn't load fallback module for '%s'. Tried '%s'.\n",
-		       sha1_alg.cra_name, sha1_fallback);
-	} else {
-		printk(KERN_NOTICE PFX "Fallback for '%s' is driver '%s' (prio=%d)\n", sha1_alg.cra_name,
-		       crypto_tfm_alg_driver_name(tfm), crypto_tfm_alg_priority(tfm));
-		crypto_free_tfm(tfm);
-	}
-
-	tfm = crypto_alloc_tfm(sha256_fallback, 0);
-	if (!tfm) {
-		printk(KERN_WARNING PFX "Couldn't load fallback module for '%s'. Tried '%s'.\n",
-		       sha256_alg.cra_name, sha256_fallback);
-	} else {
-		printk(KERN_NOTICE PFX "Fallback for '%s' is driver '%s' (prio=%d)\n", sha256_alg.cra_name,
-		       crypto_tfm_alg_driver_name(tfm), crypto_tfm_alg_priority(tfm));
-		crypto_free_tfm(tfm);
-	}
+	if (!crypto_has_hash("sha1", 0, CRYPTO_ALG_ASYNC |
+					CRYPTO_ALG_NEED_FALLBACK))
+		printk(KERN_WARNING PFX
+		       "Couldn't load fallback module for sha1.\n");
+
+	if (!crypto_has_hash("sha256", 0, CRYPTO_ALG_ASYNC |
+					CRYPTO_ALG_NEED_FALLBACK))
+		printk(KERN_WARNING PFX
+		       "Couldn't load fallback module for sha256.\n");
 }
 
 static int __init padlock_init(void)
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index d4f9948b64b1..187c6ea91959 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -42,6 +42,12 @@
 #define CRYPTO_ALG_DYING		0x00000040
 #define CRYPTO_ALG_ASYNC		0x00000080
 
+/*
+ * Set this bit if and only if the algorithm requires another algorithm of
+ * the same type to handle corner cases.
+ */
+#define CRYPTO_ALG_NEED_FALLBACK	0x00000100
+
 /*
  * Transform masks and values (for crt_flags).
  */
-- 
cgit v1.2.3


From 3ad819c61f5f8347f39cdcbe652b3c60ec615888 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Sat, 26 Aug 2006 18:44:31 +1000
Subject: [CRYPTO] api: Deprecate crypto_digest_* and crypto_alg_available

This patch marks the crypto_digest_* functions and crypto_alg_available
as deprecated.  They've been replaced by crypto_hash_* and crypto_has_*
respectively.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 include/linux/crypto.h | 18 +++++++++++++-----
 1 file changed, 13 insertions(+), 5 deletions(-)

(limited to 'include')

diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index 187c6ea91959..8f2ffa4caabf 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -241,9 +241,12 @@ int crypto_unregister_alg(struct crypto_alg *alg);
  * Algorithm query interface.
  */
 #ifdef CONFIG_CRYPTO
-int crypto_alg_available(const char *name, u32 flags);
+int crypto_alg_available(const char *name, u32 flags)
+	__deprecated_for_modules;
 int crypto_has_alg(const char *name, u32 type, u32 mask);
 #else
+static int crypto_alg_available(const char *name, u32 flags);
+	__deprecated_for_modules;
 static inline int crypto_alg_available(const char *name, u32 flags)
 {
 	return 0;
@@ -704,12 +707,15 @@ static inline void crypto_cipher_decrypt_one(struct crypto_cipher *tfm,
 						dst, src);
 }
 
-void crypto_digest_init(struct crypto_tfm *tfm);
+void crypto_digest_init(struct crypto_tfm *tfm) __deprecated_for_modules;
 void crypto_digest_update(struct crypto_tfm *tfm,
-			  struct scatterlist *sg, unsigned int nsg);
-void crypto_digest_final(struct crypto_tfm *tfm, u8 *out);
+			  struct scatterlist *sg, unsigned int nsg)
+	__deprecated_for_modules;
+void crypto_digest_final(struct crypto_tfm *tfm, u8 *out)
+	__deprecated_for_modules;
 void crypto_digest_digest(struct crypto_tfm *tfm,
-			  struct scatterlist *sg, unsigned int nsg, u8 *out);
+			  struct scatterlist *sg, unsigned int nsg, u8 *out)
+	__deprecated_for_modules;
 
 static inline struct crypto_hash *__crypto_hash_cast(struct crypto_tfm *tfm)
 {
@@ -723,6 +729,8 @@ static inline struct crypto_hash *crypto_hash_cast(struct crypto_tfm *tfm)
 	return __crypto_hash_cast(tfm);
 }
 
+static int crypto_digest_setkey(struct crypto_tfm *tfm, const u8 *key,
+				unsigned int keylen) __deprecated;
 static inline int crypto_digest_setkey(struct crypto_tfm *tfm,
                                        const u8 *key, unsigned int keylen)
 {
-- 
cgit v1.2.3


From 47b5c838af92d3504e99633bf568578203b7305f Mon Sep 17 00:00:00 2001
From: Linas Vepstas <linas@austin.ibm.com>
Date: Fri, 15 Sep 2006 18:57:42 -0500
Subject: [POWERPC] EEH: enable MMIO/DMA on frozen slot

Add wrapper around the rtas call to enable MMIO or DMA on a frozen pci
slot.

Signed-off-by: Linas Vepstas <linas@austin.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 arch/powerpc/platforms/pseries/eeh.c | 29 +++++++++++++++++++++++++++++
 include/asm-powerpc/ppc-pci.h        | 11 +++++++++++
 2 files changed, 40 insertions(+)

(limited to 'include')

diff --git a/arch/powerpc/platforms/pseries/eeh.c b/arch/powerpc/platforms/pseries/eeh.c
index fb91842fc819..4534886e3b4e 100644
--- a/arch/powerpc/platforms/pseries/eeh.c
+++ b/arch/powerpc/platforms/pseries/eeh.c
@@ -481,6 +481,35 @@ eeh_slot_availability(struct pci_dn *pdn)
 	return -1;
 }
 
+/**
+ * rtas_pci_enable - enable MMIO or DMA transfers for this slot
+ * @pdn pci device node
+ */
+
+int
+rtas_pci_enable(struct pci_dn *pdn, int function)
+{
+	int config_addr;
+	int rc;
+
+	/* Use PE configuration address, if present */
+	config_addr = pdn->eeh_config_addr;
+	if (pdn->eeh_pe_config_addr)
+		config_addr = pdn->eeh_pe_config_addr;
+
+	rc = rtas_call(ibm_set_eeh_option, 4, 1, NULL,
+	               config_addr,
+	               BUID_HI(pdn->phb->buid),
+	               BUID_LO(pdn->phb->buid),
+		            function);
+
+	if (rc)
+		printk(KERN_WARNING "EEH: Cannot enable function %d, err=%d dn=%s\n",
+		        function, rc, pdn->node->full_name);
+
+	return rc;
+}
+
 /**
  * rtas_pci_slot_reset - raises/lowers the pci #RST line
  * @pdn pci device node
diff --git a/include/asm-powerpc/ppc-pci.h b/include/asm-powerpc/ppc-pci.h
index cf79bc7ebb55..1115756c79f9 100644
--- a/include/asm-powerpc/ppc-pci.h
+++ b/include/asm-powerpc/ppc-pci.h
@@ -68,6 +68,17 @@ struct pci_dev *pci_get_device_by_addr(unsigned long addr);
  */
 void eeh_slot_error_detail (struct pci_dn *pdn, int severity);
 
+/**
+ * rtas_pci_enableo - enable IO transfers for this slot
+ * @pdn:       pci device node
+ * @function:  either EEH_THAW_MMIO or EEH_THAW_DMA 
+ *
+ * Enable I/O transfers to this slot 
+ */
+#define EEH_THAW_MMIO 2
+#define EEH_THAW_DMA  3
+int rtas_pci_enable(struct pci_dn *pdn, int function);
+
 /**
  * rtas_set_slot_reset -- unfreeze a frozen slot
  *
-- 
cgit v1.2.3


From 7da8a2e5c1fd2ee513fdeac8d13c4f3623838fd0 Mon Sep 17 00:00:00 2001
From: Josh Boyer <jdub@us.ibm.com>
Date: Wed, 20 Sep 2006 09:11:59 -0500
Subject: [POWERPC] 40x: Fix debug status register defines

This fixes some debug register defines on PPC 40x that were incorrect.

Signed-off-by: Josh Boyer <jdub@us.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 include/asm-ppc/reg_booke.h | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

(limited to 'include')

diff --git a/include/asm-ppc/reg_booke.h b/include/asm-ppc/reg_booke.h
index 4944c0fb8bea..602fbadeaf48 100644
--- a/include/asm-ppc/reg_booke.h
+++ b/include/asm-ppc/reg_booke.h
@@ -300,14 +300,14 @@ do {						\
 #define DBSR_IC		0x80000000	/* Instruction Completion */
 #define DBSR_BT		0x40000000	/* Branch taken */
 #define DBSR_TIE	0x10000000	/* Trap Instruction debug Event */
-#define DBSR_IAC1	0x00800000	/* Instruction Address Compare 1 Event */
-#define DBSR_IAC2	0x00400000	/* Instruction Address Compare 2 Event */
-#define DBSR_IAC3	0x00200000	/* Instruction Address Compare 3 Event */
-#define DBSR_IAC4	0x00100000	/* Instruction Address Compare 4 Event */
-#define DBSR_DAC1R	0x00080000	/* Data Address Compare 1 Read Event */
-#define DBSR_DAC1W	0x00040000	/* Data Address Compare 1 Write Event */
-#define DBSR_DAC2R	0x00020000	/* Data Address Compare 2 Read Event */
-#define DBSR_DAC2W	0x00010000	/* Data Address Compare 2 Write Event */
+#define DBSR_IAC1	0x04000000	/* Instruction Address Compare 1 Event */
+#define DBSR_IAC2	0x02000000	/* Instruction Address Compare 2 Event */
+#define DBSR_IAC3	0x00080000	/* Instruction Address Compare 3 Event */
+#define DBSR_IAC4	0x00040000	/* Instruction Address Compare 4 Event */
+#define DBSR_DAC1R	0x01000000	/* Data Address Compare 1 Read Event */
+#define DBSR_DAC1W	0x00800000	/* Data Address Compare 1 Write Event */
+#define DBSR_DAC2R	0x00400000	/* Data Address Compare 2 Read Event */
+#define DBSR_DAC2W	0x00200000	/* Data Address Compare 2 Write Event */
 #endif
 
 /* Bit definitions related to the ESR. */
-- 
cgit v1.2.3


From caf81329c39b5c48f6cc0d78fa159b5a587e37f9 Mon Sep 17 00:00:00 2001
From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Thu, 21 Sep 2006 18:00:00 +1000
Subject: [POWERPC] Merge iSeries i/o operations with the rest

This patch changes the io operations so that they are out of line if
CONFIG_PPC_ISERIES is set and includes a firmware feature check in
that case.

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 arch/powerpc/kernel/io.c                 |  14 ++
 arch/powerpc/platforms/iseries/pci.c     | 280 +++++++++++++++++++++++++------
 include/asm-powerpc/io.h                 | 148 ++++++++--------
 include/asm-powerpc/iseries/iseries_io.h |  60 -------
 4 files changed, 319 insertions(+), 183 deletions(-)
 delete mode 100644 include/asm-powerpc/iseries/iseries_io.h

(limited to 'include')

diff --git a/arch/powerpc/kernel/io.c b/arch/powerpc/kernel/io.c
index 80a3209acef4..e98180686b35 100644
--- a/arch/powerpc/kernel/io.c
+++ b/arch/powerpc/kernel/io.c
@@ -22,12 +22,16 @@
 #include <linux/module.h>
 
 #include <asm/io.h>
+#include <asm/firmware.h>
+#include <asm/bug.h>
 
 void _insb(volatile u8 __iomem *port, void *buf, long count)
 {
 	u8 *tbuf = buf;
 	u8 tmp;
 
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
 	if (unlikely(count <= 0))
 		return;
 	asm volatile("sync");
@@ -44,6 +48,8 @@ void _outsb(volatile u8 __iomem *port, const void *buf, long count)
 {
 	const u8 *tbuf = buf;
 
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
 	if (unlikely(count <= 0))
 		return;
 	asm volatile("sync");
@@ -59,6 +65,8 @@ void _insw_ns(volatile u16 __iomem *port, void *buf, long count)
 	u16 *tbuf = buf;
 	u16 tmp;
 
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
 	if (unlikely(count <= 0))
 		return;
 	asm volatile("sync");
@@ -75,6 +83,8 @@ void _outsw_ns(volatile u16 __iomem *port, const void *buf, long count)
 {
 	const u16 *tbuf = buf;
 
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
 	if (unlikely(count <= 0))
 		return;
 	asm volatile("sync");
@@ -90,6 +100,8 @@ void _insl_ns(volatile u32 __iomem *port, void *buf, long count)
 	u32 *tbuf = buf;
 	u32 tmp;
 
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
 	if (unlikely(count <= 0))
 		return;
 	asm volatile("sync");
@@ -106,6 +118,8 @@ void _outsl_ns(volatile u32 __iomem *port, const void *buf, long count)
 {
 	const u32 *tbuf = buf;
 
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
 	if (unlikely(count <= 0))
 		return;
 	asm volatile("sync");
diff --git a/arch/powerpc/platforms/iseries/pci.c b/arch/powerpc/platforms/iseries/pci.c
index f4d427a7bb2d..3eb12065df23 100644
--- a/arch/powerpc/platforms/iseries/pci.c
+++ b/arch/powerpc/platforms/iseries/pci.c
@@ -34,6 +34,7 @@
 #include <asm/pci-bridge.h>
 #include <asm/iommu.h>
 #include <asm/abs_addr.h>
+#include <asm/firmware.h>
 
 #include <asm/iseries/hv_call_xm.h>
 #include <asm/iseries/mf.h>
@@ -269,46 +270,6 @@ void pcibios_fixup_resources(struct pci_dev *pdev)
 {
 }
 
-/*
- * I/0 Memory copy MUST use mmio commands on iSeries
- * To do; For performance, include the hv call directly
- */
-void iSeries_memset_io(volatile void __iomem *dest, char c, size_t Count)
-{
-	u8 ByteValue = c;
-	long NumberOfBytes = Count;
-
-	while (NumberOfBytes > 0) {
-		iSeries_Write_Byte(ByteValue, dest++);
-		-- NumberOfBytes;
-	}
-}
-EXPORT_SYMBOL(iSeries_memset_io);
-
-void iSeries_memcpy_toio(volatile void __iomem *dest, void *source, size_t count)
-{
-	char *src = source;
-	long NumberOfBytes = count;
-
-	while (NumberOfBytes > 0) {
-		iSeries_Write_Byte(*src++, dest++);
-		-- NumberOfBytes;
-	}
-}
-EXPORT_SYMBOL(iSeries_memcpy_toio);
-
-void iSeries_memcpy_fromio(void *dest, const volatile void __iomem *src, size_t count)
-{
-	char *dst = dest;
-	long NumberOfBytes = count;
-
-	while (NumberOfBytes > 0) {
-		*dst++ = iSeries_Read_Byte(src++);
-		-- NumberOfBytes;
-	}
-}
-EXPORT_SYMBOL(iSeries_memcpy_fromio);
-
 /*
  * Look down the chain to find the matching Device Device
  */
@@ -491,7 +452,7 @@ static inline struct device_node *xlate_iomm_address(
  * iSeries_Read_Word = Read Word  (16 bit)
  * iSeries_Read_Long = Read Long  (32 bit)
  */
-u8 iSeries_Read_Byte(const volatile void __iomem *IoAddress)
+static u8 iSeries_Read_Byte(const volatile void __iomem *IoAddress)
 {
 	u64 BarOffset;
 	u64 dsa;
@@ -518,9 +479,8 @@ u8 iSeries_Read_Byte(const volatile void __iomem *IoAddress)
 
 	return (u8)ret.value;
 }
-EXPORT_SYMBOL(iSeries_Read_Byte);
 
-u16 iSeries_Read_Word(const volatile void __iomem *IoAddress)
+static u16 iSeries_Read_Word(const volatile void __iomem *IoAddress)
 {
 	u64 BarOffset;
 	u64 dsa;
@@ -548,9 +508,8 @@ u16 iSeries_Read_Word(const volatile void __iomem *IoAddress)
 
 	return swab16((u16)ret.value);
 }
-EXPORT_SYMBOL(iSeries_Read_Word);
 
-u32 iSeries_Read_Long(const volatile void __iomem *IoAddress)
+static u32 iSeries_Read_Long(const volatile void __iomem *IoAddress)
 {
 	u64 BarOffset;
 	u64 dsa;
@@ -578,7 +537,6 @@ u32 iSeries_Read_Long(const volatile void __iomem *IoAddress)
 
 	return swab32((u32)ret.value);
 }
-EXPORT_SYMBOL(iSeries_Read_Long);
 
 /*
  * Write MM I/O Instructions for the iSeries
@@ -587,7 +545,7 @@ EXPORT_SYMBOL(iSeries_Read_Long);
  * iSeries_Write_Word = Write Word(16 bit)
  * iSeries_Write_Long = Write Long(32 bit)
  */
-void iSeries_Write_Byte(u8 data, volatile void __iomem *IoAddress)
+static void iSeries_Write_Byte(u8 data, volatile void __iomem *IoAddress)
 {
 	u64 BarOffset;
 	u64 dsa;
@@ -612,9 +570,8 @@ void iSeries_Write_Byte(u8 data, volatile void __iomem *IoAddress)
 		rc = HvCall4(HvCallPciBarStore8, dsa, BarOffset, data, 0);
 	} while (CheckReturnCode("WWB", DevNode, &retry, rc) != 0);
 }
-EXPORT_SYMBOL(iSeries_Write_Byte);
 
-void iSeries_Write_Word(u16 data, volatile void __iomem *IoAddress)
+static void iSeries_Write_Word(u16 data, volatile void __iomem *IoAddress)
 {
 	u64 BarOffset;
 	u64 dsa;
@@ -639,9 +596,8 @@ void iSeries_Write_Word(u16 data, volatile void __iomem *IoAddress)
 		rc = HvCall4(HvCallPciBarStore16, dsa, BarOffset, swab16(data), 0);
 	} while (CheckReturnCode("WWW", DevNode, &retry, rc) != 0);
 }
-EXPORT_SYMBOL(iSeries_Write_Word);
 
-void iSeries_Write_Long(u32 data, volatile void __iomem *IoAddress)
+static void iSeries_Write_Long(u32 data, volatile void __iomem *IoAddress)
 {
 	u64 BarOffset;
 	u64 dsa;
@@ -666,4 +622,224 @@ void iSeries_Write_Long(u32 data, volatile void __iomem *IoAddress)
 		rc = HvCall4(HvCallPciBarStore32, dsa, BarOffset, swab32(data), 0);
 	} while (CheckReturnCode("WWL", DevNode, &retry, rc) != 0);
 }
-EXPORT_SYMBOL(iSeries_Write_Long);
+
+extern unsigned char __raw_readb(const volatile void __iomem *addr)
+{
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
+	return *(volatile unsigned char __force *)addr;
+}
+EXPORT_SYMBOL(__raw_readb);
+
+extern unsigned short __raw_readw(const volatile void __iomem *addr)
+{
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
+	return *(volatile unsigned short __force *)addr;
+}
+EXPORT_SYMBOL(__raw_readw);
+
+extern unsigned int __raw_readl(const volatile void __iomem *addr)
+{
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
+	return *(volatile unsigned int __force *)addr;
+}
+EXPORT_SYMBOL(__raw_readl);
+
+extern unsigned long __raw_readq(const volatile void __iomem *addr)
+{
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
+	return *(volatile unsigned long __force *)addr;
+}
+EXPORT_SYMBOL(__raw_readq);
+
+extern void __raw_writeb(unsigned char v, volatile void __iomem *addr)
+{
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
+	*(volatile unsigned char __force *)addr = v;
+}
+EXPORT_SYMBOL(__raw_writeb);
+
+extern void __raw_writew(unsigned short v, volatile void __iomem *addr)
+{
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
+	*(volatile unsigned short __force *)addr = v;
+}
+EXPORT_SYMBOL(__raw_writew);
+
+extern void __raw_writel(unsigned int v, volatile void __iomem *addr)
+{
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
+	*(volatile unsigned int __force *)addr = v;
+}
+EXPORT_SYMBOL(__raw_writel);
+
+extern void __raw_writeq(unsigned long v, volatile void __iomem *addr)
+{
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
+	*(volatile unsigned long __force *)addr = v;
+}
+EXPORT_SYMBOL(__raw_writeq);
+
+int in_8(const volatile unsigned char __iomem *addr)
+{
+	if (firmware_has_feature(FW_FEATURE_ISERIES))
+		return iSeries_Read_Byte(addr);
+	return __in_8(addr);
+}
+EXPORT_SYMBOL(in_8);
+
+void out_8(volatile unsigned char __iomem *addr, int val)
+{
+	if (firmware_has_feature(FW_FEATURE_ISERIES))
+		iSeries_Write_Byte(val, addr);
+	else
+		__out_8(addr, val);
+}
+EXPORT_SYMBOL(out_8);
+
+int in_le16(const volatile unsigned short __iomem *addr)
+{
+	if (firmware_has_feature(FW_FEATURE_ISERIES))
+		return iSeries_Read_Word(addr);
+	return __in_le16(addr);
+}
+EXPORT_SYMBOL(in_le16);
+
+int in_be16(const volatile unsigned short __iomem *addr)
+{
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
+	return __in_be16(addr);
+}
+EXPORT_SYMBOL(in_be16);
+
+void out_le16(volatile unsigned short __iomem *addr, int val)
+{
+	if (firmware_has_feature(FW_FEATURE_ISERIES))
+		iSeries_Write_Word(val, addr);
+	else
+		__out_le16(addr, val);
+}
+EXPORT_SYMBOL(out_le16);
+
+void out_be16(volatile unsigned short __iomem *addr, int val)
+{
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
+	__out_be16(addr, val);
+}
+EXPORT_SYMBOL(out_be16);
+
+unsigned in_le32(const volatile unsigned __iomem *addr)
+{
+	if (firmware_has_feature(FW_FEATURE_ISERIES))
+		return iSeries_Read_Long(addr);
+	return __in_le32(addr);
+}
+EXPORT_SYMBOL(in_le32);
+
+unsigned in_be32(const volatile unsigned __iomem *addr)
+{
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
+	return __in_be32(addr);
+}
+EXPORT_SYMBOL(in_be32);
+
+void out_le32(volatile unsigned __iomem *addr, int val)
+{
+	if (firmware_has_feature(FW_FEATURE_ISERIES))
+		iSeries_Write_Long(val, addr);
+	else
+		__out_le32(addr, val);
+}
+EXPORT_SYMBOL(out_le32);
+
+void out_be32(volatile unsigned __iomem *addr, int val)
+{
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
+	__out_be32(addr, val);
+}
+EXPORT_SYMBOL(out_be32);
+
+unsigned long in_le64(const volatile unsigned long __iomem *addr)
+{
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
+	return __in_le64(addr);
+}
+EXPORT_SYMBOL(in_le64);
+
+unsigned long in_be64(const volatile unsigned long __iomem *addr)
+{
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
+	return __in_be64(addr);
+}
+EXPORT_SYMBOL(in_be64);
+
+void out_le64(volatile unsigned long __iomem *addr, unsigned long val)
+{
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
+	__out_le64(addr, val);
+}
+EXPORT_SYMBOL(out_le64);
+
+void out_be64(volatile unsigned long __iomem *addr, unsigned long val)
+{
+	BUG_ON(firmware_has_feature(FW_FEATURE_ISERIES));
+
+	__out_be64(addr, val);
+}
+EXPORT_SYMBOL(out_be64);
+
+void memset_io(volatile void __iomem *addr, int c, unsigned long n)
+{
+	if (firmware_has_feature(FW_FEATURE_ISERIES)) {
+		volatile char __iomem *d = addr;
+
+		while (n-- > 0) {
+			iSeries_Write_Byte(c, d++);
+		}
+	} else
+		eeh_memset_io(addr, c, n);
+}
+EXPORT_SYMBOL(memset_io);
+
+void memcpy_fromio(void *dest, const volatile void __iomem *src,
+                                 unsigned long n)
+{
+	if (firmware_has_feature(FW_FEATURE_ISERIES)) {
+		char *d = dest;
+		const volatile char __iomem *s = src;
+
+		while (n-- > 0) {
+			*d++ = iSeries_Read_Byte(s++);
+		}
+	} else
+		eeh_memcpy_fromio(dest, src, n);
+}
+EXPORT_SYMBOL(memcpy_fromio);
+
+void memcpy_toio(volatile void __iomem *dest, const void *src, unsigned long n)
+{
+	if (firmware_has_feature(FW_FEATURE_ISERIES)) {
+		const char *s = src;
+		volatile char __iomem *d = dest;
+
+		while (n-- > 0) {
+			iSeries_Write_Byte(*s++, d++);
+		}
+	} else
+		eeh_memcpy_toio(dest, src, n);
+}
+EXPORT_SYMBOL(memcpy_toio);
diff --git a/include/asm-powerpc/io.h b/include/asm-powerpc/io.h
index 174fb89d5eda..46bae1cf385b 100644
--- a/include/asm-powerpc/io.h
+++ b/include/asm-powerpc/io.h
@@ -20,9 +20,6 @@ extern int check_legacy_ioport(unsigned long base_port);
 #include <asm/page.h>
 #include <asm/byteorder.h>
 #include <asm/paca.h>
-#ifdef CONFIG_PPC_ISERIES 
-#include <asm/iseries/iseries_io.h>
-#endif  
 #include <asm/synch.h>
 #include <asm/delay.h>
 
@@ -37,41 +34,53 @@ extern unsigned long isa_io_base;
 extern unsigned long pci_io_base;
 
 #ifdef CONFIG_PPC_ISERIES
-/* __raw_* accessors aren't supported on iSeries */
-#define __raw_readb(addr)	{ BUG(); 0; }
-#define __raw_readw(addr)       { BUG(); 0; }
-#define __raw_readl(addr)       { BUG(); 0; }
-#define __raw_readq(addr)       { BUG(); 0; }
-#define __raw_writeb(v, addr)   { BUG(); 0; }
-#define __raw_writew(v, addr)   { BUG(); 0; }
-#define __raw_writel(v, addr)   { BUG(); 0; }
-#define __raw_writeq(v, addr)   { BUG(); 0; }
-#define readb(addr)		iSeries_Read_Byte(addr)
-#define readw(addr)		iSeries_Read_Word(addr)
-#define readl(addr)		iSeries_Read_Long(addr)
-#define writeb(data, addr)	iSeries_Write_Byte((data),(addr))
-#define writew(data, addr)	iSeries_Write_Word((data),(addr))
-#define writel(data, addr)	iSeries_Write_Long((data),(addr))
-#define memset_io(a,b,c)	iSeries_memset_io((a),(b),(c))
-#define memcpy_fromio(a,b,c)	iSeries_memcpy_fromio((a), (b), (c))
-#define memcpy_toio(a,b,c)	iSeries_memcpy_toio((a), (b), (c))
-
-#define inb(addr)		readb(((void __iomem *)(long)(addr)))
-#define inw(addr)		readw(((void __iomem *)(long)(addr)))
-#define inl(addr)		readl(((void __iomem *)(long)(addr)))
-#define outb(data,addr)		writeb(data,((void __iomem *)(long)(addr)))
-#define outw(data,addr)		writew(data,((void __iomem *)(long)(addr)))
-#define outl(data,addr)		writel(data,((void __iomem *)(long)(addr)))
-/*
- * The *_ns versions below don't do byte-swapping.
- * Neither do the standard versions now, these are just here
- * for older code.
- */
-#define insb(port, buf, ns)	_insb((u8 __iomem *)((port)+pci_io_base), (buf), (ns))
-#define insw(port, buf, ns)	_insw_ns((u16 __iomem *)((port)+pci_io_base), (buf), (ns))
-#define insl(port, buf, nl)	_insl_ns((u32 __iomem *)((port)+pci_io_base), (buf), (nl))
 
-#else
+extern int in_8(const volatile unsigned char __iomem *addr);
+extern void out_8(volatile unsigned char __iomem *addr, int val);
+extern int in_le16(const volatile unsigned short __iomem *addr);
+extern int in_be16(const volatile unsigned short __iomem *addr);
+extern void out_le16(volatile unsigned short __iomem *addr, int val);
+extern void out_be16(volatile unsigned short __iomem *addr, int val);
+extern unsigned in_le32(const volatile unsigned __iomem *addr);
+extern unsigned in_be32(const volatile unsigned __iomem *addr);
+extern void out_le32(volatile unsigned __iomem *addr, int val);
+extern void out_be32(volatile unsigned __iomem *addr, int val);
+extern unsigned long in_le64(const volatile unsigned long __iomem *addr);
+extern unsigned long in_be64(const volatile unsigned long __iomem *addr);
+extern void out_le64(volatile unsigned long __iomem *addr, unsigned long val);
+extern void out_be64(volatile unsigned long __iomem *addr, unsigned long val);
+
+extern unsigned char __raw_readb(const volatile void __iomem *addr);
+extern unsigned short __raw_readw(const volatile void __iomem *addr);
+extern unsigned int __raw_readl(const volatile void __iomem *addr);
+extern unsigned long __raw_readq(const volatile void __iomem *addr);
+extern void __raw_writeb(unsigned char v, volatile void __iomem *addr);
+extern void __raw_writew(unsigned short v, volatile void __iomem *addr);
+extern void __raw_writel(unsigned int v, volatile void __iomem *addr);
+extern void __raw_writeq(unsigned long v, volatile void __iomem *addr);
+
+extern void memset_io(volatile void __iomem *addr, int c, unsigned long n);
+extern void memcpy_fromio(void *dest, const volatile void __iomem *src,
+                                 unsigned long n);
+extern void memcpy_toio(volatile void __iomem *dest, const void *src,
+                                 unsigned long n);
+
+#else /* CONFIG_PPC_ISERIES */
+
+#define in_8(addr)		__in_8((addr))
+#define out_8(addr, val)	__out_8((addr), (val))
+#define in_le16(addr)		__in_le16((addr))
+#define in_be16(addr)		__in_be16((addr))
+#define out_le16(addr, val)	__out_le16((addr), (val))
+#define out_be16(addr, val)	__out_be16((addr), (val))
+#define in_le32(addr)		__in_le32((addr))
+#define in_be32(addr)		__in_be32((addr))
+#define out_le32(addr, val)	__out_le32((addr), (val))
+#define out_be32(addr, val)	__out_be32((addr), (val))
+#define in_le64(addr)		__in_le64((addr))
+#define in_be64(addr)		__in_be64((addr))
+#define out_le64(addr, val)	__out_le64((addr), (val))
+#define out_be64(addr, val)	__out_be64((addr), (val))
 
 static inline unsigned char __raw_readb(const volatile void __iomem *addr)
 {
@@ -105,23 +114,11 @@ static inline void __raw_writeq(unsigned long v, volatile void __iomem *addr)
 {
 	*(volatile unsigned long __force *)addr = v;
 }
-#define readb(addr)		eeh_readb(addr)
-#define readw(addr)		eeh_readw(addr)
-#define readl(addr)		eeh_readl(addr)
-#define readq(addr)		eeh_readq(addr)
-#define writeb(data, addr)	eeh_writeb((data), (addr))
-#define writew(data, addr)	eeh_writew((data), (addr))
-#define writel(data, addr)	eeh_writel((data), (addr))
-#define writeq(data, addr)	eeh_writeq((data), (addr))
 #define memset_io(a,b,c)	eeh_memset_io((a),(b),(c))
 #define memcpy_fromio(a,b,c)	eeh_memcpy_fromio((a),(b),(c))
 #define memcpy_toio(a,b,c)	eeh_memcpy_toio((a),(b),(c))
-#define inb(port)		eeh_inb((unsigned long)port)
-#define outb(val, port)		eeh_outb(val, (unsigned long)port)
-#define inw(port)		eeh_inw((unsigned long)port)
-#define outw(val, port)		eeh_outw(val, (unsigned long)port)
-#define inl(port)		eeh_inl((unsigned long)port)
-#define outl(val, port)		eeh_outl(val, (unsigned long)port)
+
+#endif /* CONFIG_PPC_ISERIES */
 
 /*
  * The insw/outsw/insl/outsl macros don't do byte-swapping.
@@ -132,12 +129,25 @@ static inline void __raw_writeq(unsigned long v, volatile void __iomem *addr)
 #define insw(port, buf, ns)	eeh_insw_ns((port), (buf), (ns))
 #define insl(port, buf, nl)	eeh_insl_ns((port), (buf), (nl))
 
-#endif
-
 #define outsb(port, buf, ns)  _outsb((u8 __iomem *)((port)+pci_io_base), (buf), (ns))
 #define outsw(port, buf, ns)  _outsw_ns((u16 __iomem *)((port)+pci_io_base), (buf), (ns))
 #define outsl(port, buf, nl)  _outsl_ns((u32 __iomem *)((port)+pci_io_base), (buf), (nl))
 
+#define readb(addr)		eeh_readb(addr)
+#define readw(addr)		eeh_readw(addr)
+#define readl(addr)		eeh_readl(addr)
+#define readq(addr)		eeh_readq(addr)
+#define writeb(data, addr)	eeh_writeb((data), (addr))
+#define writew(data, addr)	eeh_writew((data), (addr))
+#define writel(data, addr)	eeh_writel((data), (addr))
+#define writeq(data, addr)	eeh_writeq((data), (addr))
+#define inb(port)		eeh_inb((unsigned long)port)
+#define outb(val, port)		eeh_outb(val, (unsigned long)port)
+#define inw(port)		eeh_inw((unsigned long)port)
+#define outw(val, port)		eeh_outw(val, (unsigned long)port)
+#define inl(port)		eeh_inl((unsigned long)port)
+#define outl(val, port)		eeh_outl(val, (unsigned long)port)
+
 #define readb_relaxed(addr) readb(addr)
 #define readw_relaxed(addr) readw(addr)
 #define readl_relaxed(addr) readl(addr)
@@ -258,7 +268,7 @@ static inline void iosync(void)
  * and should not be used directly by device drivers.  Use inb/readb
  * instead.
  */
-static inline int in_8(const volatile unsigned char __iomem *addr)
+static inline int __in_8(const volatile unsigned char __iomem *addr)
 {
 	int ret;
 
@@ -267,14 +277,14 @@ static inline int in_8(const volatile unsigned char __iomem *addr)
 	return ret;
 }
 
-static inline void out_8(volatile unsigned char __iomem *addr, int val)
+static inline void __out_8(volatile unsigned char __iomem *addr, int val)
 {
 	__asm__ __volatile__("sync; stb%U0%X0 %1,%0"
 			     : "=m" (*addr) : "r" (val));
 	get_paca()->io_sync = 1;
 }
 
-static inline int in_le16(const volatile unsigned short __iomem *addr)
+static inline int __in_le16(const volatile unsigned short __iomem *addr)
 {
 	int ret;
 
@@ -283,7 +293,7 @@ static inline int in_le16(const volatile unsigned short __iomem *addr)
 	return ret;
 }
 
-static inline int in_be16(const volatile unsigned short __iomem *addr)
+static inline int __in_be16(const volatile unsigned short __iomem *addr)
 {
 	int ret;
 
@@ -292,21 +302,21 @@ static inline int in_be16(const volatile unsigned short __iomem *addr)
 	return ret;
 }
 
-static inline void out_le16(volatile unsigned short __iomem *addr, int val)
+static inline void __out_le16(volatile unsigned short __iomem *addr, int val)
 {
 	__asm__ __volatile__("sync; sthbrx %1,0,%2"
 			     : "=m" (*addr) : "r" (val), "r" (addr));
 	get_paca()->io_sync = 1;
 }
 
-static inline void out_be16(volatile unsigned short __iomem *addr, int val)
+static inline void __out_be16(volatile unsigned short __iomem *addr, int val)
 {
 	__asm__ __volatile__("sync; sth%U0%X0 %1,%0"
 			     : "=m" (*addr) : "r" (val));
 	get_paca()->io_sync = 1;
 }
 
-static inline unsigned in_le32(const volatile unsigned __iomem *addr)
+static inline unsigned __in_le32(const volatile unsigned __iomem *addr)
 {
 	unsigned ret;
 
@@ -315,7 +325,7 @@ static inline unsigned in_le32(const volatile unsigned __iomem *addr)
 	return ret;
 }
 
-static inline unsigned in_be32(const volatile unsigned __iomem *addr)
+static inline unsigned __in_be32(const volatile unsigned __iomem *addr)
 {
 	unsigned ret;
 
@@ -324,21 +334,21 @@ static inline unsigned in_be32(const volatile unsigned __iomem *addr)
 	return ret;
 }
 
-static inline void out_le32(volatile unsigned __iomem *addr, int val)
+static inline void __out_le32(volatile unsigned __iomem *addr, int val)
 {
 	__asm__ __volatile__("sync; stwbrx %1,0,%2" : "=m" (*addr)
 			     : "r" (val), "r" (addr));
 	get_paca()->io_sync = 1;
 }
 
-static inline void out_be32(volatile unsigned __iomem *addr, int val)
+static inline void __out_be32(volatile unsigned __iomem *addr, int val)
 {
 	__asm__ __volatile__("sync; stw%U0%X0 %1,%0"
 			     : "=m" (*addr) : "r" (val));
 	get_paca()->io_sync = 1;
 }
 
-static inline unsigned long in_le64(const volatile unsigned long __iomem *addr)
+static inline unsigned long __in_le64(const volatile unsigned long __iomem *addr)
 {
 	unsigned long tmp, ret;
 
@@ -358,7 +368,7 @@ static inline unsigned long in_le64(const volatile unsigned long __iomem *addr)
 	return ret;
 }
 
-static inline unsigned long in_be64(const volatile unsigned long __iomem *addr)
+static inline unsigned long __in_be64(const volatile unsigned long __iomem *addr)
 {
 	unsigned long ret;
 
@@ -367,7 +377,7 @@ static inline unsigned long in_be64(const volatile unsigned long __iomem *addr)
 	return ret;
 }
 
-static inline void out_le64(volatile unsigned long __iomem *addr, unsigned long val)
+static inline void __out_le64(volatile unsigned long __iomem *addr, unsigned long val)
 {
 	unsigned long tmp;
 
@@ -385,15 +395,13 @@ static inline void out_le64(volatile unsigned long __iomem *addr, unsigned long
 	get_paca()->io_sync = 1;
 }
 
-static inline void out_be64(volatile unsigned long __iomem *addr, unsigned long val)
+static inline void __out_be64(volatile unsigned long __iomem *addr, unsigned long val)
 {
 	__asm__ __volatile__("sync; std%U0%X0 %1,%0" : "=m" (*addr) : "r" (val));
 	get_paca()->io_sync = 1;
 }
 
-#ifndef CONFIG_PPC_ISERIES 
 #include <asm/eeh.h>
-#endif
 
 /**
  *	check_signature		-	find BIOS signatures
@@ -409,7 +417,6 @@ static inline int check_signature(const volatile void __iomem * io_addr,
 	const unsigned char *signature, int length)
 {
 	int retval = 0;
-#ifndef CONFIG_PPC_ISERIES 
 	do {
 		if (readb(io_addr) != *signature)
 			goto out;
@@ -419,7 +426,6 @@ static inline int check_signature(const volatile void __iomem * io_addr,
 	} while (length);
 	retval = 1;
 out:
-#endif
 	return retval;
 }
 
diff --git a/include/asm-powerpc/iseries/iseries_io.h b/include/asm-powerpc/iseries/iseries_io.h
deleted file mode 100644
index f29009bd63c9..000000000000
--- a/include/asm-powerpc/iseries/iseries_io.h
+++ /dev/null
@@ -1,60 +0,0 @@
-#ifndef _ASM_POWERPC_ISERIES_ISERIES_IO_H
-#define _ASM_POWERPC_ISERIES_ISERIES_IO_H
-
-
-#ifdef CONFIG_PPC_ISERIES
-#include <linux/types.h>
-/*
- * Created by Allan Trautman on Thu Dec 28 2000.
- *
- * Remaps the io.h for the iSeries Io
- * Copyright (C) 2000  Allan H Trautman, IBM Corporation
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the:
- * Free Software Foundation, Inc.,
- * 59 Temple Place, Suite 330,
- * Boston, MA  02111-1307  USA
- *
- * Change Activity:
- *   Created December 28, 2000
- * End Change Activity
- */
-
-#ifdef CONFIG_PCI
-extern u8   iSeries_Read_Byte(const volatile void __iomem * IoAddress);
-extern u16  iSeries_Read_Word(const volatile void __iomem * IoAddress);
-extern u32  iSeries_Read_Long(const volatile void __iomem * IoAddress);
-extern void iSeries_Write_Byte(u8  IoData, volatile void __iomem * IoAddress);
-extern void iSeries_Write_Word(u16 IoData, volatile void __iomem * IoAddress);
-extern void iSeries_Write_Long(u32 IoData, volatile void __iomem * IoAddress);
-
-extern void iSeries_memset_io(volatile void __iomem *dest, char x, size_t n);
-extern void iSeries_memcpy_toio(volatile void __iomem *dest, void *source,
-		size_t n);
-extern void iSeries_memcpy_fromio(void *dest,
-		const volatile void __iomem *source, size_t n);
-#else
-static inline u8 iSeries_Read_Byte(const volatile void __iomem *IoAddress)
-{
-	return 0xff;
-}
-
-static inline void iSeries_Write_Byte(u8 IoData,
-		volatile void __iomem *IoAddress)
-{
-}
-#endif	/* CONFIG_PCI */
-
-#endif /* CONFIG_PPC_ISERIES */
-#endif /* _ASM_POWERPC_ISERIES_ISERIES_IO_H */
-- 
cgit v1.2.3


From 2954da897c40de0f3abdd6a100f2978f30d04068 Mon Sep 17 00:00:00 2001
From: Michael Ellerman <michael@ellerman.id.au>
Date: Thu, 21 Sep 2006 18:21:35 +1000
Subject: [POWERPC] Remove DISCONTIGMEM cruft from page.h

This looks like cruft to me, these functions don't exist AFAICT,
and I can't see that it's possible to even enable DISCONTIGMEM on
powerpc anymore. CC'ing some folks who might know better, based on
the who-touched-it-last principle.

Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 include/asm-powerpc/page.h | 6 ------
 1 file changed, 6 deletions(-)

(limited to 'include')

diff --git a/include/asm-powerpc/page.h b/include/asm-powerpc/page.h
index fb597b37c2a2..b4d38b0b15f8 100644
--- a/include/asm-powerpc/page.h
+++ b/include/asm-powerpc/page.h
@@ -55,12 +55,6 @@
 #define PAGE_OFFSET     ASM_CONST(CONFIG_KERNEL_START)
 #define KERNELBASE      (PAGE_OFFSET + PHYSICAL_START)
 
-#ifdef CONFIG_DISCONTIGMEM
-#define page_to_pfn(page)	discontigmem_page_to_pfn(page)
-#define pfn_to_page(pfn)	discontigmem_pfn_to_page(pfn)
-#define pfn_valid(pfn)		discontigmem_pfn_valid(pfn)
-#endif
-
 #ifdef CONFIG_FLATMEM
 #define pfn_valid(pfn)		((pfn) < max_mapnr)
 #endif
-- 
cgit v1.2.3


From 7d452c326c2ac879aced884411a0fe3ba75d9c87 Mon Sep 17 00:00:00 2001
From: Arnd Bergmann <arnd.bergmann@de.ibm.com>
Date: Thu, 21 Sep 2006 12:29:51 +0200
Subject: [POWERPC] powerpc: fix building gdb against asm/ptrace.h

Ulrich Weigand found a bug with the current version of the
asm-powerpc/ptrace.h that prevents building at least the
SPU target version of gdb, since some ptrace opcodes are
not defined.

The problem seems to have originated in the merging of 32 and
64 bit versions of that file, the problem is that some opcodes
are only valid on 64 bit kernels, but are also used by 32 bit
programs, so they can't depends on the __powerpc64__ symbol.

Signed-off-by: Arnd Bergmann <arnd.bergmann@de.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 include/asm-powerpc/ptrace.h | 4 ----
 1 file changed, 4 deletions(-)

(limited to 'include')

diff --git a/include/asm-powerpc/ptrace.h b/include/asm-powerpc/ptrace.h
index dc4cb9cc73a1..4435efe85d0e 100644
--- a/include/asm-powerpc/ptrace.h
+++ b/include/asm-powerpc/ptrace.h
@@ -215,12 +215,10 @@ do {									      \
 #define PTRACE_GETVRREGS	18
 #define PTRACE_SETVRREGS	19
 
-#ifndef __powerpc64__
 /* Get/set all the upper 32-bits of the SPE registers, accumulator, and
  * spefscr, in one go */
 #define PTRACE_GETEVRREGS	20
 #define PTRACE_SETEVRREGS	21
-#endif /* __powerpc64__ */
 
 /*
  * Get or set a debug register. The first 16 are DABR registers and the
@@ -235,7 +233,6 @@ do {									      \
 #define PPC_PTRACE_GETFPREGS	0x97	/* Get FPRs 0 - 31 */
 #define PPC_PTRACE_SETFPREGS	0x96	/* Set FPRs 0 - 31 */
 
-#ifdef __powerpc64__
 /* Calls to trace a 64bit program from a 32bit program */
 #define PPC_PTRACE_PEEKTEXT_3264 0x95
 #define PPC_PTRACE_PEEKDATA_3264 0x94
@@ -243,6 +240,5 @@ do {									      \
 #define PPC_PTRACE_POKEDATA_3264 0x92
 #define PPC_PTRACE_PEEKUSR_3264  0x91
 #define PPC_PTRACE_POKEUSR_3264  0x90
-#endif /* __powerpc64__ */
 
 #endif /* _ASM_POWERPC_PTRACE_H */
-- 
cgit v1.2.3


From 187ef15268e638603dea55a91fdfa29feaed6d13 Mon Sep 17 00:00:00 2001
From: Håvard Skinnemoen <hskinnemoen@atmel.com>
Date: Fri, 22 Sep 2006 10:07:08 +0100
Subject: [MTD] Unlock NOR flash automatically where necessary
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Introduce the MTD_STUPID_LOCK flag which indicates that the flash chip is
always locked after power-up, so all sectors need to be unlocked before it
is usable.

If this flag is set, and the chip provides an unlock() operation,
mtd_add_device will unlock the whole MTD device if it's writeable.  This
means that non-writeable partitions will stay locked.

Set MTD_STUPID_LOCK in fixup_use_atmel_lock() so that these chips will work
as expected.

Signed-off-by: Håvard Skinnemoen <hskinnemoen@atmel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
---
 drivers/mtd/chips/cfi_cmdset_0002.c |  1 +
 drivers/mtd/mtdcore.c               | 10 ++++++++++
 include/mtd/mtd-abi.h               |  1 +
 3 files changed, 12 insertions(+)

(limited to 'include')

diff --git a/drivers/mtd/chips/cfi_cmdset_0002.c b/drivers/mtd/chips/cfi_cmdset_0002.c
index a482e8922de1..702ae4cd8691 100644
--- a/drivers/mtd/chips/cfi_cmdset_0002.c
+++ b/drivers/mtd/chips/cfi_cmdset_0002.c
@@ -212,6 +212,7 @@ static void fixup_use_atmel_lock(struct mtd_info *mtd, void *param)
 {
 	mtd->lock = cfi_atmel_lock;
 	mtd->unlock = cfi_atmel_unlock;
+	mtd->flags |= MTD_STUPID_LOCK;
 }
 
 static struct cfi_fixup cfi_fixup_table[] = {
diff --git a/drivers/mtd/mtdcore.c b/drivers/mtd/mtdcore.c
index 168d3ba063c3..c4d26de74349 100644
--- a/drivers/mtd/mtdcore.c
+++ b/drivers/mtd/mtdcore.c
@@ -57,6 +57,16 @@ int add_mtd_device(struct mtd_info *mtd)
 			mtd->index = i;
 			mtd->usecount = 0;
 
+			/* Some chips always power up locked. Unlock them now */
+			if ((mtd->flags & MTD_WRITEABLE)
+			    && (mtd->flags & MTD_STUPID_LOCK) && mtd->unlock) {
+				if (mtd->unlock(mtd, 0, mtd->size))
+					printk(KERN_WARNING
+					       "%s: unlock failed, "
+					       "writes may not work\n",
+					       mtd->name);
+			}
+
 			DEBUG(0, "mtd: Giving out device %d to %s\n",i, mtd->name);
 			/* No need to get a refcount on the module containing
 			   the notifier, since we hold the mtd_table_mutex */
diff --git a/include/mtd/mtd-abi.h b/include/mtd/mtd-abi.h
index 1da3f7fa7993..b0a67b7ffdcd 100644
--- a/include/mtd/mtd-abi.h
+++ b/include/mtd/mtd-abi.h
@@ -34,6 +34,7 @@ struct mtd_oob_buf {
 #define MTD_WRITEABLE		0x400	/* Device is writeable */
 #define MTD_BIT_WRITEABLE	0x800	/* Single bits can be flipped */
 #define MTD_NO_ERASE		0x1000	/* No erase necessary */
+#define MTD_STUPID_LOCK		0x2000	/* Always locked after reset */
 
 // Some common devices / combinations of capabilities
 #define MTD_CAP_ROM		0
-- 
cgit v1.2.3


From 25f0c659fe64832d8ee06aa623fffaad708dcf8b Mon Sep 17 00:00:00 2001
From: Amol Lad <amol@verismonetworks.com>
Date: Thu, 21 Sep 2006 18:12:43 +0530
Subject: ioremap balanced with iounmap for drivers/mtd subsystem

ioremap must be balanced by an iounmap and failing to do so can result
in a memory leak.

Tested (compilation only) with:
- allmodconfig
- Modifying drivers/mtd/maps/Kconfig and drivers/mtd/nand/Kconfig to
make sure that the changed file is compiling without warning

Signed-off-by: Amol Lad <amol@verismonetworks.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
---
 drivers/mtd/maps/arctic-mtd.c     | 14 ++++++++++++--
 drivers/mtd/maps/beech-mtd.c      | 14 ++++++++++++--
 drivers/mtd/maps/cstm_mips_ixx.c  | 18 ++++++++++++++++--
 drivers/mtd/maps/ebony.c          |  4 ++++
 drivers/mtd/maps/fortunet.c       |  3 +++
 drivers/mtd/maps/lasat.c          |  2 ++
 drivers/mtd/maps/nettel.c         | 34 +++++++++++++++++++++++++++-------
 drivers/mtd/maps/ocotea.c         |  4 ++++
 drivers/mtd/maps/pcmciamtd.c      |  4 ++++
 drivers/mtd/maps/redwood.c        | 11 ++++++++++-
 drivers/mtd/maps/sbc8240.c        | 11 ++++++++++-
 drivers/mtd/maps/walnut.c         |  4 ++++
 drivers/mtd/nand/edb7312.c        |  3 +++
 drivers/mtd/nand/ppchameleonevb.c |  7 +++++++
 include/linux/utsrelease.h        |  1 +
 15 files changed, 119 insertions(+), 15 deletions(-)
 create mode 100644 include/linux/utsrelease.h

(limited to 'include')

diff --git a/drivers/mtd/maps/arctic-mtd.c b/drivers/mtd/maps/arctic-mtd.c
index d95ae582fbe9..642d96bc8919 100644
--- a/drivers/mtd/maps/arctic-mtd.c
+++ b/drivers/mtd/maps/arctic-mtd.c
@@ -96,6 +96,8 @@ static struct mtd_partition arctic_partitions[PARTITIONS] = {
 static int __init
 init_arctic_mtd(void)
 {
+	int err = 0;
+
 	printk("%s: 0x%08x at 0x%08x\n", NAME, SIZE, PADDR);
 
 	arctic_mtd_map.virt = ioremap(PADDR, SIZE);
@@ -109,12 +111,20 @@ init_arctic_mtd(void)
 	printk("%s: probing %d-bit flash bus\n", NAME, BUSWIDTH * 8);
 	arctic_mtd = do_map_probe("cfi_probe", &arctic_mtd_map);
 
-	if (!arctic_mtd)
+	if (!arctic_mtd) {
+		iounmap((void *) arctic_mtd_map.virt);
 		return -ENXIO;
+	}
 
 	arctic_mtd->owner = THIS_MODULE;
 
-	return add_mtd_partitions(arctic_mtd, arctic_partitions, PARTITIONS);
+	err = add_mtd_partitions(arctic_mtd, arctic_partitions, PARTITIONS);
+	if (err) {
+		printk("%s: add_mtd_partitions failed\n", NAME);
+		iounmap((void *) arctic_mtd_map.virt);
+	}
+
+	return err;
 }
 
 static void __exit
diff --git a/drivers/mtd/maps/beech-mtd.c b/drivers/mtd/maps/beech-mtd.c
index 5df7361d1407..a64b1a5ab316 100644
--- a/drivers/mtd/maps/beech-mtd.c
+++ b/drivers/mtd/maps/beech-mtd.c
@@ -72,6 +72,8 @@ static struct mtd_partition beech_partitions[2] = {
 static int __init
 init_beech_mtd(void)
 {
+	int err = 0;
+
 	printk("%s: 0x%08x at 0x%08x\n", NAME, SIZE, PADDR);
 
 	beech_mtd_map.virt = ioremap(PADDR, SIZE);
@@ -86,12 +88,20 @@ init_beech_mtd(void)
 	printk("%s: probing %d-bit flash bus\n", NAME, BUSWIDTH * 8);
 	beech_mtd = do_map_probe("cfi_probe", &beech_mtd_map);
 
-	if (!beech_mtd)
+	if (!beech_mtd) {
+		iounmap((void *) beech_mtd_map.virt);
 		return -ENXIO;
+	}
 
 	beech_mtd->owner = THIS_MODULE;
 
-	return add_mtd_partitions(beech_mtd, beech_partitions, 2);
+	err = add_mtd_partitions(beech_mtd, beech_partitions, 2);
+	if (err) {
+		printk("%s: add_mtd_partitions failed\n", NAME);
+		iounmap((void *) beech_mtd_map.virt);
+	}
+
+	return err;
 }
 
 static void __exit
diff --git a/drivers/mtd/maps/cstm_mips_ixx.c b/drivers/mtd/maps/cstm_mips_ixx.c
index aa56defb94c8..d6bef100d69a 100644
--- a/drivers/mtd/maps/cstm_mips_ixx.c
+++ b/drivers/mtd/maps/cstm_mips_ixx.c
@@ -171,7 +171,14 @@ int __init init_cstm_mips_ixx(void)
 		cstm_mips_ixx_map[i].phys = cstm_mips_ixx_board_desc[i].window_addr;
 		cstm_mips_ixx_map[i].virt = ioremap(cstm_mips_ixx_board_desc[i].window_addr, cstm_mips_ixx_board_desc[i].window_size);
 		if (!cstm_mips_ixx_map[i].virt) {
+			int j = 0;
 			printk(KERN_WARNING "Failed to ioremap\n");
+			for (j = 0; j < i; j++) {
+				if (cstm_mips_ixx_map[j].virt) {
+					iounmap((void *)cstm_mips_ixx_map[j].virt);
+					cstm_mips_ixx_map[j].virt = 0;
+				}
+			}
 			return -EIO;
 	        }
 		cstm_mips_ixx_map[i].name = cstm_mips_ixx_board_desc[i].name;
@@ -204,8 +211,15 @@ int __init init_cstm_mips_ixx(void)
 	                cstm_mips_ixx_map[i].map_priv_2 = (unsigned long)mymtd;
 		        add_mtd_partitions(mymtd, parts, cstm_mips_ixx_board_desc[i].num_partitions);
 		}
-		else
-	           return -ENXIO;
+		else {
+			for (i = 0; i < PHYSMAP_NUMBER; i++) {
+				if (cstm_mips_ixx_map[i].virt) {
+					iounmap((void *)cstm_mips_ixx_map[i].virt);
+					cstm_mips_ixx_map[i].virt = 0;
+				}
+			}
+			return -ENXIO;
+		}
 	}
 	return 0;
 }
diff --git a/drivers/mtd/maps/ebony.c b/drivers/mtd/maps/ebony.c
index 641e1dd8479e..1488bb92f26f 100644
--- a/drivers/mtd/maps/ebony.c
+++ b/drivers/mtd/maps/ebony.c
@@ -108,6 +108,7 @@ int __init init_ebony(void)
 					ARRAY_SIZE(ebony_small_partitions));
 	} else {
 		printk("map probe failed for flash\n");
+		iounmap(ebony_small_map.virt);
 		return -ENXIO;
 	}
 
@@ -117,6 +118,7 @@ int __init init_ebony(void)
 
 	if (!ebony_large_map.virt) {
 		printk("Failed to ioremap flash\n");
+		iounmap(ebony_small_map.virt);
 		return -EIO;
 	}
 
@@ -129,6 +131,8 @@ int __init init_ebony(void)
 					ARRAY_SIZE(ebony_large_partitions));
 	} else {
 		printk("map probe failed for flash\n");
+		iounmap(ebony_small_map.virt);
+		iounmap(ebony_large_map.virt);
 		return -ENXIO;
 	}
 
diff --git a/drivers/mtd/maps/fortunet.c b/drivers/mtd/maps/fortunet.c
index c6bf4e1219ef..7c50c271651c 100644
--- a/drivers/mtd/maps/fortunet.c
+++ b/drivers/mtd/maps/fortunet.c
@@ -218,8 +218,11 @@ int __init init_fortunet(void)
 				map_regions[ix].map_info.size);
 			if(!map_regions[ix].map_info.virt)
 			{
+				int j = 0;
 				printk(MTD_FORTUNET_PK "%s flash failed to ioremap!\n",
 					map_regions[ix].map_info.name);
+				for (j = 0 ; j < ix; j++)
+					iounmap(map_regions[j].map_info.virt);
 				return -ENXIO;
 			}
 			simple_map_init(&map_regions[ix].map_info);
diff --git a/drivers/mtd/maps/lasat.c b/drivers/mtd/maps/lasat.c
index 1c13d2dc0cdf..e34376321050 100644
--- a/drivers/mtd/maps/lasat.c
+++ b/drivers/mtd/maps/lasat.c
@@ -79,6 +79,7 @@ static int __init init_lasat(void)
 		return 0;
 	}
 
+	iounmap(lasat_map.virt);
 	return -ENXIO;
 }
 
@@ -89,6 +90,7 @@ static void __exit cleanup_lasat(void)
 		map_destroy(lasat_mtd);
 	}
 	if (lasat_map.virt) {
+		iounmap(lasat_map.virt);
 		lasat_map.virt = 0;
 	}
 }
diff --git a/drivers/mtd/maps/nettel.c b/drivers/mtd/maps/nettel.c
index 0994b5b2e331..198e840ff6db 100644
--- a/drivers/mtd/maps/nettel.c
+++ b/drivers/mtd/maps/nettel.c
@@ -277,6 +277,7 @@ int __init nettel_init(void)
 	nettel_amd_map.virt = ioremap_nocache(amdaddr, maxsize);
 	if (!nettel_amd_map.virt) {
 		printk("SNAPGEAR: failed to ioremap() BOOTCS\n");
+		iounmap(nettel_mmcrp);
 		return(-EIO);
 	}
 	simple_map_init(&nettel_amd_map);
@@ -337,7 +338,8 @@ int __init nettel_init(void)
 		nettel_amd_map.virt = NULL;
 #else
 		/* Only AMD flash supported */
-		return(-ENXIO);
+		rc = -ENXIO;
+		goto out_unmap2;
 #endif
 	}
 
@@ -361,14 +363,15 @@ int __init nettel_init(void)
 	nettel_intel_map.virt = ioremap_nocache(intel0addr, maxsize);
 	if (!nettel_intel_map.virt) {
 		printk("SNAPGEAR: failed to ioremap() ROMCS1\n");
-		return(-EIO);
+		rc = -EIO;
+		goto out_unmap2;
 	}
 	simple_map_init(&nettel_intel_map);
 
 	intel_mtd = do_map_probe("cfi_probe", &nettel_intel_map);
 	if (!intel_mtd) {
-		iounmap(nettel_intel_map.virt);
-		return(-ENXIO);
+		rc = -ENXIO;
+		goto out_unmap1;
 	}
 
 	/* Set PAR to the detected size */
@@ -394,13 +397,14 @@ int __init nettel_init(void)
 	nettel_intel_map.virt = ioremap_nocache(intel0addr, maxsize);
 	if (!nettel_intel_map.virt) {
 		printk("SNAPGEAR: failed to ioremap() ROMCS1/2\n");
-		return(-EIO);
+		rc = -EIO;
+		goto out_unmap2;
 	}
 
 	intel_mtd = do_map_probe("cfi_probe", &nettel_intel_map);
 	if (! intel_mtd) {
-		iounmap((void *) nettel_intel_map.virt);
-		return(-ENXIO);
+		rc = -ENXIO;
+		goto out_unmap1;
 	}
 
 	intel1size = intel_mtd->size - intel0size;
@@ -456,6 +460,18 @@ int __init nettel_init(void)
 #endif
 
 	return(rc);
+
+#ifdef CONFIG_MTD_CFI_INTELEXT
+out_unmap1:
+	iounmap((void *) nettel_intel_map.virt);
+#endif
+
+out_unmap2:
+	iounmap(nettel_mmcrp);
+	iounmap(nettel_amd_map.virt);
+
+	return(rc);
+		
 }
 
 /****************************************************************************/
@@ -469,6 +485,10 @@ void __exit nettel_cleanup(void)
 		del_mtd_partitions(amd_mtd);
 		map_destroy(amd_mtd);
 	}
+	if (nettel_mmcrp) {
+		iounmap(nettel_mmcrp);
+		nettel_mmcrp = NULL;
+	}
 	if (nettel_amd_map.virt) {
 		iounmap(nettel_amd_map.virt);
 		nettel_amd_map.virt = NULL;
diff --git a/drivers/mtd/maps/ocotea.c b/drivers/mtd/maps/ocotea.c
index 2f07602ba940..5522eac8c980 100644
--- a/drivers/mtd/maps/ocotea.c
+++ b/drivers/mtd/maps/ocotea.c
@@ -97,6 +97,7 @@ int __init init_ocotea(void)
 					ARRAY_SIZE(ocotea_small_partitions));
 	} else {
 		printk("map probe failed for flash\n");
+		iounmap(ocotea_small_map.virt);
 		return -ENXIO;
 	}
 
@@ -106,6 +107,7 @@ int __init init_ocotea(void)
 
 	if (!ocotea_large_map.virt) {
 		printk("Failed to ioremap flash\n");
+		iounmap(ocotea_small_map.virt);
 		return -EIO;
 	}
 
@@ -118,6 +120,8 @@ int __init init_ocotea(void)
 					ARRAY_SIZE(ocotea_large_partitions));
 	} else {
 		printk("map probe failed for flash\n");
+		iounmap(ocotea_small_map.virt);
+		iounmap(ocotea_large_map.virt);
 		return -ENXIO;
 	}
 
diff --git a/drivers/mtd/maps/pcmciamtd.c b/drivers/mtd/maps/pcmciamtd.c
index c861134cbc48..995347b1beba 100644
--- a/drivers/mtd/maps/pcmciamtd.c
+++ b/drivers/mtd/maps/pcmciamtd.c
@@ -602,6 +602,10 @@ static int pcmciamtd_config(struct pcmcia_device *link)
 	ret = pcmcia_request_configuration(link, &link->conf);
 	if(ret != CS_SUCCESS) {
 		cs_error(link, RequestConfiguration, ret);
+		if (dev->win_base) {
+			iounmap(dev->win_base);
+			dev->win_base = NULL;
+		}
 		return -ENODEV;
 	}
 
diff --git a/drivers/mtd/maps/redwood.c b/drivers/mtd/maps/redwood.c
index ec8fdae1dd99..2257d2b500c0 100644
--- a/drivers/mtd/maps/redwood.c
+++ b/drivers/mtd/maps/redwood.c
@@ -126,6 +126,8 @@ static struct mtd_info *redwood_mtd;
 
 int __init init_redwood_flash(void)
 {
+	int err = 0;
+
 	printk(KERN_NOTICE "redwood: flash mapping: %x at %x\n",
 			WINDOW_SIZE, WINDOW_ADDR);
 
@@ -141,11 +143,18 @@ int __init init_redwood_flash(void)
 
 	if (redwood_mtd) {
 		redwood_mtd->owner = THIS_MODULE;
-		return add_mtd_partitions(redwood_mtd,
+		err = add_mtd_partitions(redwood_mtd,
 				redwood_flash_partitions,
 				NUM_REDWOOD_FLASH_PARTITIONS);
+		if (err) {
+			printk("init_redwood_flash: add_mtd_partitions failed\n");
+			iounmap(redwood_flash_map.virt);
+		}
+		return err;
+
 	}
 
+	iounmap(redwood_flash_map.virt);
 	return -ENXIO;
 }
 
diff --git a/drivers/mtd/maps/sbc8240.c b/drivers/mtd/maps/sbc8240.c
index 7d0fcf8f4f33..b8c1331b7a04 100644
--- a/drivers/mtd/maps/sbc8240.c
+++ b/drivers/mtd/maps/sbc8240.c
@@ -156,7 +156,7 @@ int __init init_sbc8240_mtd (void)
 	};
 
 	int devicesfound = 0;
-	int i;
+	int i,j;
 
 	for (i = 0; i < NUM_FLASH_BANKS; i++) {
 		printk (KERN_NOTICE MSG_PREFIX
@@ -166,6 +166,10 @@ int __init init_sbc8240_mtd (void)
 			(unsigned long) ioremap (pt[i].addr, pt[i].size);
 		if (!sbc8240_map[i].map_priv_1) {
 			printk (MSG_PREFIX "failed to ioremap\n");
+			for (j = 0; j < i; j++) {
+				iounmap((void *) sbc8240_map[j].map_priv_1);
+				sbc8240_map[j].map_priv_1 = 0;
+			}
 			return -EIO;
 		}
 		simple_map_init(&sbc8240_mtd[i]);
@@ -175,6 +179,11 @@ int __init init_sbc8240_mtd (void)
 		if (sbc8240_mtd[i]) {
 			sbc8240_mtd[i]->module = THIS_MODULE;
 			devicesfound++;
+		} else {
+			if (sbc8240_map[i].map_priv_1) {
+				iounmap((void *) sbc8240_map[i].map_priv_1);
+				sbc8240_map[i].map_priv_1 = 0;
+			}
 		}
 	}
 
diff --git a/drivers/mtd/maps/walnut.c b/drivers/mtd/maps/walnut.c
index ec80eec376bf..ca932122fb64 100644
--- a/drivers/mtd/maps/walnut.c
+++ b/drivers/mtd/maps/walnut.c
@@ -68,6 +68,7 @@ int __init init_walnut(void)
 
 	if (WALNUT_FLASH_ONBD_N(fpga_brds1)) {
 		printk("The on-board flash is disabled (U79 sw 5)!");
+		iounmap(fpga_status_adr);
 		return -EIO;
 	}
 	if (WALNUT_FLASH_SRAM_SEL(fpga_brds1))
@@ -81,6 +82,7 @@ int __init init_walnut(void)
 
 	if (!walnut_map.virt) {
 		printk("Failed to ioremap flash.\n");
+		iounmap(fpga_status_adr);
 		return -EIO;
 	}
 
@@ -93,9 +95,11 @@ int __init init_walnut(void)
 					ARRAY_SIZE(walnut_partitions));
 	} else {
 		printk("map probe failed for flash\n");
+		iounmap(fpga_status_adr);
 		return -ENXIO;
 	}
 
+	iounmap(fpga_status_adr);
 	return 0;
 }
 
diff --git a/drivers/mtd/nand/edb7312.c b/drivers/mtd/nand/edb7312.c
index 516c0e5e564c..12017f3c6bd6 100644
--- a/drivers/mtd/nand/edb7312.c
+++ b/drivers/mtd/nand/edb7312.c
@@ -198,6 +198,9 @@ static void __exit ep7312_cleanup(void)
 	/* Release resources, unregister device */
 	nand_release(ap7312_mtd);
 
+	/* Release io resource */
+	iounmap((void *)this->IO_ADDR_R);
+
 	/* Free the MTD device structure */
 	kfree(ep7312_mtd);
 }
diff --git a/drivers/mtd/nand/ppchameleonevb.c b/drivers/mtd/nand/ppchameleonevb.c
index 22fa65c12ab9..eb7d4d443deb 100644
--- a/drivers/mtd/nand/ppchameleonevb.c
+++ b/drivers/mtd/nand/ppchameleonevb.c
@@ -276,6 +276,7 @@ static int __init ppchameleonevb_init(void)
 	/* Scan to find existence of the device (it could not be mounted) */
 	if (nand_scan(ppchameleon_mtd, 1)) {
 		iounmap((void *)ppchameleon_fio_base);
+		ppchameleon_fio_base = NULL;
 		kfree(ppchameleon_mtd);
 		goto nand_evb_init;
 	}
@@ -314,6 +315,8 @@ static int __init ppchameleonevb_init(void)
 	ppchameleonevb_mtd = kmalloc(sizeof(struct mtd_info) + sizeof(struct nand_chip), GFP_KERNEL);
 	if (!ppchameleonevb_mtd) {
 		printk("Unable to allocate PPChameleonEVB NAND MTD device structure.\n");
+		if (ppchameleon_fio_base)
+			iounmap(ppchameleon_fio_base);
 		return -ENOMEM;
 	}
 
@@ -322,6 +325,8 @@ static int __init ppchameleonevb_init(void)
 	if (!ppchameleonevb_fio_base) {
 		printk("ioremap PPChameleonEVB NAND flash failed\n");
 		kfree(ppchameleonevb_mtd);
+		if (ppchameleon_fio_base)
+			iounmap(ppchameleon_fio_base);
 		return -EIO;
 	}
 
@@ -378,6 +383,8 @@ static int __init ppchameleonevb_init(void)
 	if (nand_scan(ppchameleonevb_mtd, 1)) {
 		iounmap((void *)ppchameleonevb_fio_base);
 		kfree(ppchameleonevb_mtd);
+		if (ppchameleon_fio_base)
+			iounmap(ppchameleon_fio_base);
 		return -ENXIO;
 	}
 #ifdef CONFIG_MTD_PARTITIONS
diff --git a/include/linux/utsrelease.h b/include/linux/utsrelease.h
new file mode 100644
index 000000000000..7fe1de0ad4c1
--- /dev/null
+++ b/include/linux/utsrelease.h
@@ -0,0 +1 @@
+#define UTS_RELEASE "2.6.18"
-- 
cgit v1.2.3


From 734a56285dbeedc6cc10aef6f700eeab7c65ea9f Mon Sep 17 00:00:00 2001
From: David Woodhouse <dwmw2@infradead.org>
Date: Fri, 22 Sep 2006 18:41:20 +0100
Subject: Remove accidentally-added include/linux/utsrelease.h

Signed-off-by: David Woodhouse <dwmw2@infradead.org>
---
 include/linux/utsrelease.h | 1 -
 1 file changed, 1 deletion(-)
 delete mode 100644 include/linux/utsrelease.h

(limited to 'include')

diff --git a/include/linux/utsrelease.h b/include/linux/utsrelease.h
deleted file mode 100644
index 7fe1de0ad4c1..000000000000
--- a/include/linux/utsrelease.h
+++ /dev/null
@@ -1 +0,0 @@
-#define UTS_RELEASE "2.6.18"
-- 
cgit v1.2.3


From 892c141e62982272b9c738b5520ad0e5e1ad7b42 Mon Sep 17 00:00:00 2001
From: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Date: Fri, 4 Aug 2006 23:08:56 -0700
Subject: [MLSXFRM]: Add security sid to sock

This adds security for IP sockets at the sock level. Security at the
sock level is needed to enforce the SELinux security policy for
security associations even when a sock is orphaned (such as in the TCP
LAST_ACK state).

This will also be used to enforce SELinux controls over data arriving
at or leaving a child socket while it's still waiting to be accepted.

Signed-off-by: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/security.h          | 12 ++++++++++++
 include/net/sock.h                | 13 +++++++++++++
 net/core/sock.c                   |  2 +-
 security/dummy.c                  |  5 +++++
 security/selinux/hooks.c          | 38 +++++++++++++++++++++-----------------
 security/selinux/include/objsec.h |  1 +
 6 files changed, 53 insertions(+), 18 deletions(-)

(limited to 'include')

diff --git a/include/linux/security.h b/include/linux/security.h
index 6bc2aad494ff..4d7fb59996b0 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -812,6 +812,8 @@ struct swap_info_struct;
  *      which is used to copy security attributes between local stream sockets.
  * @sk_free_security:
  *	Deallocate security structure.
+ * @sk_clone_security:
+ *	Clone/copy security structure.
  * @sk_getsid:
  *	Retrieve the LSM-specific sid for the sock to enable caching of network
  *	authorizations.
@@ -1332,6 +1334,7 @@ struct security_operations {
 	int (*socket_getpeersec_dgram) (struct socket *sock, struct sk_buff *skb, u32 *secid);
 	int (*sk_alloc_security) (struct sock *sk, int family, gfp_t priority);
 	void (*sk_free_security) (struct sock *sk);
+	void (*sk_clone_security) (const struct sock *sk, struct sock *newsk);
 	unsigned int (*sk_getsid) (struct sock *sk, struct flowi *fl, u8 dir);
 #endif	/* CONFIG_SECURITY_NETWORK */
 
@@ -2885,6 +2888,11 @@ static inline void security_sk_free(struct sock *sk)
 	return security_ops->sk_free_security(sk);
 }
 
+static inline void security_sk_clone(const struct sock *sk, struct sock *newsk)
+{
+	return security_ops->sk_clone_security(sk, newsk);
+}
+
 static inline unsigned int security_sk_sid(struct sock *sk, struct flowi *fl, u8 dir)
 {
 	return security_ops->sk_getsid(sk, fl, dir);
@@ -3011,6 +3019,10 @@ static inline void security_sk_free(struct sock *sk)
 {
 }
 
+static inline void security_sk_clone(const struct sock *sk, struct sock *newsk)
+{
+}
+
 static inline unsigned int security_sk_sid(struct sock *sk, struct flowi *fl, u8 dir)
 {
 	return 0;
diff --git a/include/net/sock.h b/include/net/sock.h
index 324b3ea233d6..91cdceb3c028 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -972,6 +972,19 @@ static inline void sock_graft(struct sock *sk, struct socket *parent)
 	write_unlock_bh(&sk->sk_callback_lock);
 }
 
+static inline void sock_copy(struct sock *nsk, const struct sock *osk)
+{
+#ifdef CONFIG_SECURITY_NETWORK
+	void *sptr = nsk->sk_security;
+#endif
+
+	memcpy(nsk, osk, osk->sk_prot->obj_size);
+#ifdef CONFIG_SECURITY_NETWORK
+	nsk->sk_security = sptr;
+	security_sk_clone(osk, nsk);
+#endif
+}
+
 extern int sock_i_uid(struct sock *sk);
 extern unsigned long sock_i_ino(struct sock *sk);
 
diff --git a/net/core/sock.c b/net/core/sock.c
index 51fcfbc041a7..b67d868649cd 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -911,7 +911,7 @@ struct sock *sk_clone(const struct sock *sk, const gfp_t priority)
 	if (newsk != NULL) {
 		struct sk_filter *filter;
 
-		memcpy(newsk, sk, sk->sk_prot->obj_size);
+		sock_copy(newsk, sk);
 
 		/* SANITY */
 		sk_node_init(&newsk->sk_node);
diff --git a/security/dummy.c b/security/dummy.c
index 58c6d399c844..bd3bc5faa9a8 100644
--- a/security/dummy.c
+++ b/security/dummy.c
@@ -805,6 +805,10 @@ static inline void dummy_sk_free_security (struct sock *sk)
 {
 }
 
+static inline void dummy_sk_clone_security (const struct sock *sk, struct sock *newsk)
+{
+}
+
 static unsigned int dummy_sk_getsid(struct sock *sk, struct flowi *fl, u8 dir)
 {
 	return 0;
@@ -1060,6 +1064,7 @@ void security_fixup_ops (struct security_operations *ops)
 	set_to_dummy_if_null(ops, socket_getpeersec_dgram);
 	set_to_dummy_if_null(ops, sk_alloc_security);
 	set_to_dummy_if_null(ops, sk_free_security);
+	set_to_dummy_if_null(ops, sk_clone_security);
 	set_to_dummy_if_null(ops, sk_getsid);
  #endif	/* CONFIG_SECURITY_NETWORK */
 #ifdef  CONFIG_SECURITY_NETWORK_XFRM
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index 5d1b8c733199..d67abf77584a 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -269,15 +269,13 @@ static int sk_alloc_security(struct sock *sk, int family, gfp_t priority)
 {
 	struct sk_security_struct *ssec;
 
-	if (family != PF_UNIX)
-		return 0;
-
 	ssec = kzalloc(sizeof(*ssec), priority);
 	if (!ssec)
 		return -ENOMEM;
 
 	ssec->sk = sk;
 	ssec->peer_sid = SECINITSID_UNLABELED;
+	ssec->sid = SECINITSID_UNLABELED;
 	sk->sk_security = ssec;
 
 	return 0;
@@ -287,9 +285,6 @@ static void sk_free_security(struct sock *sk)
 {
 	struct sk_security_struct *ssec = sk->sk_security;
 
-	if (sk->sk_family != PF_UNIX)
-		return;
-
 	sk->sk_security = NULL;
 	kfree(ssec);
 }
@@ -3068,6 +3063,7 @@ static void selinux_socket_post_create(struct socket *sock, int family,
 {
 	struct inode_security_struct *isec;
 	struct task_security_struct *tsec;
+	struct sk_security_struct *sksec;
 	u32 newsid;
 
 	isec = SOCK_INODE(sock)->i_security;
@@ -3078,6 +3074,11 @@ static void selinux_socket_post_create(struct socket *sock, int family,
 	isec->sid = kern ? SECINITSID_KERNEL : newsid;
 	isec->initialized = 1;
 
+	if (sock->sk) {
+		sksec = sock->sk->sk_security;
+		sksec->sid = isec->sid;
+	}
+
 	return;
 }
 
@@ -3551,22 +3552,24 @@ static void selinux_sk_free_security(struct sock *sk)
 	sk_free_security(sk);
 }
 
-static unsigned int selinux_sk_getsid_security(struct sock *sk, struct flowi *fl, u8 dir)
+static void selinux_sk_clone_security(const struct sock *sk, struct sock *newsk)
 {
-	struct inode_security_struct *isec;
-	u32 sock_sid = SECINITSID_ANY_SOCKET;
+	struct sk_security_struct *ssec = sk->sk_security;
+	struct sk_security_struct *newssec = newsk->sk_security;
 
+	newssec->sid = ssec->sid;
+	newssec->peer_sid = ssec->peer_sid;
+}
+
+static unsigned int selinux_sk_getsid_security(struct sock *sk, struct flowi *fl, u8 dir)
+{
 	if (!sk)
 		return selinux_no_sk_sid(fl);
+	else {
+		struct sk_security_struct *sksec = sk->sk_security;
 
-	read_lock_bh(&sk->sk_callback_lock);
-	isec = get_sock_isec(sk);
-
-	if (isec)
-		sock_sid = isec->sid;
-
-	read_unlock_bh(&sk->sk_callback_lock);
-	return sock_sid;
+		return sksec->sid;
+	}
 }
 
 static int selinux_nlmsg_perm(struct sock *sk, struct sk_buff *skb)
@@ -4618,6 +4621,7 @@ static struct security_operations selinux_ops = {
 	.socket_getpeersec_dgram =	selinux_socket_getpeersec_dgram,
 	.sk_alloc_security =		selinux_sk_alloc_security,
 	.sk_free_security =		selinux_sk_free_security,
+	.sk_clone_security =		selinux_sk_clone_security,
 	.sk_getsid = 			selinux_sk_getsid_security,
 
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
diff --git a/security/selinux/include/objsec.h b/security/selinux/include/objsec.h
index 940178865fc7..79b9e0af19a0 100644
--- a/security/selinux/include/objsec.h
+++ b/security/selinux/include/objsec.h
@@ -99,6 +99,7 @@ struct netif_security_struct {
 
 struct sk_security_struct {
 	struct sock *sk;		/* back pointer to sk object */
+	u32 sid;			/* SID of this object */
 	u32 peer_sid;			/* SID of peer */
 };
 
-- 
cgit v1.2.3


From b6340fcd761acf9249b3acbc95c4dc555d9beb07 Mon Sep 17 00:00:00 2001
From: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Date: Mon, 24 Jul 2006 23:28:37 -0700
Subject: [MLSXFRM]: Add security sid to flowi

This adds security to flow key for labeling of flows as also to allow
for making flow cache lookups based on the security label seemless.

Signed-off-by: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 Documentation/networking/secid.txt | 14 ++++++++++++++
 include/net/flow.h                 |  1 +
 2 files changed, 15 insertions(+)
 create mode 100644 Documentation/networking/secid.txt

(limited to 'include')

diff --git a/Documentation/networking/secid.txt b/Documentation/networking/secid.txt
new file mode 100644
index 000000000000..95ea06784333
--- /dev/null
+++ b/Documentation/networking/secid.txt
@@ -0,0 +1,14 @@
+flowi structure:
+
+The secid member in the flow structure is used in LSMs (e.g. SELinux) to indicate
+the label of the flow. This label of the flow is currently used in selecting
+matching labeled xfrm(s).
+
+If this is an outbound flow, the label is derived from the socket, if any, or
+the incoming packet this flow is being generated as a response to (e.g. tcp
+resets, timewait ack, etc.). It is also conceivable that the label could be
+derived from other sources such as process context, device, etc., in special
+cases, as may be appropriate.
+
+If this is an inbound flow, the label is derived from the IPSec security
+associations, if any, used by the packet.
diff --git a/include/net/flow.h b/include/net/flow.h
index 04d89f763451..1cee5a83433a 100644
--- a/include/net/flow.h
+++ b/include/net/flow.h
@@ -78,6 +78,7 @@ struct flowi {
 #define fl_icmp_type	uli_u.icmpt.type
 #define fl_icmp_code	uli_u.icmpt.code
 #define fl_ipsec_spi	uli_u.spi
+	__u32           secid;	/* used by xfrm; see secid.txt */
 } __attribute__((__aligned__(BITS_PER_LONG/8)));
 
 #define FLOW_DIR_IN	0
-- 
cgit v1.2.3


From e0d1caa7b0d5f02e4f34aa09c695d04251310c6c Mon Sep 17 00:00:00 2001
From: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Date: Mon, 24 Jul 2006 23:29:07 -0700
Subject: [MLSXFRM]: Flow based matching of xfrm policy and state

This implements a seemless mechanism for xfrm policy selection and
state matching based on the flow sid. This also includes the necessary
SELinux enforcement pieces.

Signed-off-by: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/security.h        | 106 +++++++++++++++++----
 include/net/flow.h              |   4 +-
 net/core/flow.c                 |   7 +-
 net/xfrm/xfrm_policy.c          |  28 +++---
 net/xfrm/xfrm_state.c           |  12 ++-
 security/dummy.c                |  23 ++++-
 security/selinux/hooks.c        |   7 +-
 security/selinux/include/xfrm.h |  23 +++--
 security/selinux/xfrm.c         | 199 +++++++++++++++++++++++++++++++++-------
 9 files changed, 329 insertions(+), 80 deletions(-)

(limited to 'include')

diff --git a/include/linux/security.h b/include/linux/security.h
index 4d7fb59996b0..2c4921d79d19 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -31,6 +31,7 @@
 #include <linux/msg.h>
 #include <linux/sched.h>
 #include <linux/key.h>
+#include <linux/xfrm.h>
 
 struct ctl_table;
 
@@ -825,9 +826,8 @@ struct swap_info_struct;
  *	used by the XFRM system.
  *	@sec_ctx contains the security context information being provided by
  *	the user-level policy update program (e.g., setkey).
- *	Allocate a security structure to the xp->security field.
- *	The security field is initialized to NULL when the xfrm_policy is
- *	allocated.
+ *	Allocate a security structure to the xp->security field; the security
+ *	field is initialized to NULL when the xfrm_policy is allocated.
  *	Return 0 if operation was successful (memory to allocate, legal context)
  * @xfrm_policy_clone_security:
  *	@old contains an existing xfrm_policy in the SPD.
@@ -846,9 +846,14 @@ struct swap_info_struct;
  *	Database by the XFRM system.
  *	@sec_ctx contains the security context information being provided by
  *	the user-level SA generation program (e.g., setkey or racoon).
- *	Allocate a security structure to the x->security field.  The
- *	security field is initialized to NULL when the xfrm_state is
- *	allocated.
+ *	@polsec contains the security context information associated with a xfrm
+ *	policy rule from which to take the base context. polsec must be NULL
+ *	when sec_ctx is specified.
+ *	@secid contains the secid from which to take the mls portion of the context.
+ *	Allocate a security structure to the x->security field; the security
+ *	field is initialized to NULL when the xfrm_state is allocated. Set the
+ *	context to correspond to either sec_ctx or polsec, with the mls portion
+ *	taken from secid in the latter case.
  *	Return 0 if operation was successful (memory to allocate, legal context).
  * @xfrm_state_free_security:
  *	@x contains the xfrm_state.
@@ -859,13 +864,26 @@ struct swap_info_struct;
  * @xfrm_policy_lookup:
  *	@xp contains the xfrm_policy for which the access control is being
  *	checked.
- *	@sk_sid contains the sock security label that is used to authorize
+ *	@fl_secid contains the flow security label that is used to authorize
  *	access to the policy xp.
  *	@dir contains the direction of the flow (input or output).
- *	Check permission when a sock selects a xfrm_policy for processing
+ *	Check permission when a flow selects a xfrm_policy for processing
  *	XFRMs on a packet.  The hook is called when selecting either a
  *	per-socket policy or a generic xfrm policy.
  *	Return 0 if permission is granted.
+ * @xfrm_state_pol_flow_match:
+ *	@x contains the state to match.
+ *	@xp contains the policy to check for a match.
+ *	@fl contains the flow to check for a match.
+ *	Return 1 if there is a match.
+ * @xfrm_flow_state_match:
+ *	@fl contains the flow key to match.
+ *	@xfrm points to the xfrm_state to match.
+ *	Return 1 if there is a match.
+ * @xfrm_decode_session:
+ *	@skb points to skb to decode.
+ *	@fl points to the flow key to set.
+ *	Return 0 if successful decoding.
  *
  * Security hooks affecting all Key Management operations
  *
@@ -1343,10 +1361,16 @@ struct security_operations {
 	int (*xfrm_policy_clone_security) (struct xfrm_policy *old, struct xfrm_policy *new);
 	void (*xfrm_policy_free_security) (struct xfrm_policy *xp);
 	int (*xfrm_policy_delete_security) (struct xfrm_policy *xp);
-	int (*xfrm_state_alloc_security) (struct xfrm_state *x, struct xfrm_user_sec_ctx *sec_ctx);
+	int (*xfrm_state_alloc_security) (struct xfrm_state *x,
+		struct xfrm_user_sec_ctx *sec_ctx, struct xfrm_sec_ctx *polsec,
+		u32 secid);
 	void (*xfrm_state_free_security) (struct xfrm_state *x);
 	int (*xfrm_state_delete_security) (struct xfrm_state *x);
-	int (*xfrm_policy_lookup)(struct xfrm_policy *xp, u32 sk_sid, u8 dir);
+	int (*xfrm_policy_lookup)(struct xfrm_policy *xp, u32 fl_secid, u8 dir);
+	int (*xfrm_state_pol_flow_match)(struct xfrm_state *x,
+			struct xfrm_policy *xp, struct flowi *fl);
+	int (*xfrm_flow_state_match)(struct flowi *fl, struct xfrm_state *xfrm);
+	int (*xfrm_decode_session)(struct sk_buff *skb, struct flowi *fl);
 #endif	/* CONFIG_SECURITY_NETWORK_XFRM */
 
 	/* key management security hooks */
@@ -3050,9 +3074,18 @@ static inline int security_xfrm_policy_delete(struct xfrm_policy *xp)
 	return security_ops->xfrm_policy_delete_security(xp);
 }
 
-static inline int security_xfrm_state_alloc(struct xfrm_state *x, struct xfrm_user_sec_ctx *sec_ctx)
+static inline int security_xfrm_state_alloc(struct xfrm_state *x,
+			struct xfrm_user_sec_ctx *sec_ctx)
+{
+	return security_ops->xfrm_state_alloc_security(x, sec_ctx, NULL, 0);
+}
+
+static inline int security_xfrm_state_alloc_acquire(struct xfrm_state *x,
+				struct xfrm_sec_ctx *polsec, u32 secid)
 {
-	return security_ops->xfrm_state_alloc_security(x, sec_ctx);
+	if (!polsec)
+		return 0;
+	return security_ops->xfrm_state_alloc_security(x, NULL, polsec, secid);
 }
 
 static inline int security_xfrm_state_delete(struct xfrm_state *x)
@@ -3065,9 +3098,25 @@ static inline void security_xfrm_state_free(struct xfrm_state *x)
 	security_ops->xfrm_state_free_security(x);
 }
 
-static inline int security_xfrm_policy_lookup(struct xfrm_policy *xp, u32 sk_sid, u8 dir)
+static inline int security_xfrm_policy_lookup(struct xfrm_policy *xp, u32 fl_secid, u8 dir)
+{
+	return security_ops->xfrm_policy_lookup(xp, fl_secid, dir);
+}
+
+static inline int security_xfrm_state_pol_flow_match(struct xfrm_state *x,
+			struct xfrm_policy *xp, struct flowi *fl)
+{
+	return security_ops->xfrm_state_pol_flow_match(x, xp, fl);
+}
+
+static inline int security_xfrm_flow_state_match(struct flowi *fl, struct xfrm_state *xfrm)
+{
+	return security_ops->xfrm_flow_state_match(fl, xfrm);
+}
+
+static inline int security_xfrm_decode_session(struct sk_buff *skb, struct flowi *fl)
 {
-	return security_ops->xfrm_policy_lookup(xp, sk_sid, dir);
+	return security_ops->xfrm_decode_session(skb, fl);
 }
 #else	/* CONFIG_SECURITY_NETWORK_XFRM */
 static inline int security_xfrm_policy_alloc(struct xfrm_policy *xp, struct xfrm_user_sec_ctx *sec_ctx)
@@ -3089,7 +3138,14 @@ static inline int security_xfrm_policy_delete(struct xfrm_policy *xp)
 	return 0;
 }
 
-static inline int security_xfrm_state_alloc(struct xfrm_state *x, struct xfrm_user_sec_ctx *sec_ctx)
+static inline int security_xfrm_state_alloc(struct xfrm_state *x,
+					struct xfrm_user_sec_ctx *sec_ctx)
+{
+	return 0;
+}
+
+static inline int security_xfrm_state_alloc_acquire(struct xfrm_state *x,
+					struct xfrm_sec_ctx *polsec, u32 secid)
 {
 	return 0;
 }
@@ -3103,10 +3159,28 @@ static inline int security_xfrm_state_delete(struct xfrm_state *x)
 	return 0;
 }
 
-static inline int security_xfrm_policy_lookup(struct xfrm_policy *xp, u32 sk_sid, u8 dir)
+static inline int security_xfrm_policy_lookup(struct xfrm_policy *xp, u32 fl_secid, u8 dir)
 {
 	return 0;
 }
+
+static inline int security_xfrm_state_pol_flow_match(struct xfrm_state *x,
+			struct xfrm_policy *xp, struct flowi *fl)
+{
+	return 1;
+}
+
+static inline int security_xfrm_flow_state_match(struct flowi *fl,
+                                struct xfrm_state *xfrm)
+{
+	return 1;
+}
+
+static inline int security_xfrm_decode_session(struct sk_buff *skb, struct flowi *fl)
+{
+	return 0;
+}
+
 #endif	/* CONFIG_SECURITY_NETWORK_XFRM */
 
 #ifdef CONFIG_KEYS
diff --git a/include/net/flow.h b/include/net/flow.h
index 1cee5a83433a..21d988b2058a 100644
--- a/include/net/flow.h
+++ b/include/net/flow.h
@@ -86,10 +86,10 @@ struct flowi {
 #define FLOW_DIR_FWD	2
 
 struct sock;
-typedef void (*flow_resolve_t)(struct flowi *key, u32 sk_sid, u16 family, u8 dir,
+typedef void (*flow_resolve_t)(struct flowi *key, u16 family, u8 dir,
 			       void **objp, atomic_t **obj_refp);
 
-extern void *flow_cache_lookup(struct flowi *key, u32 sk_sid, u16 family, u8 dir,
+extern void *flow_cache_lookup(struct flowi *key, u16 family, u8 dir,
 	 		       flow_resolve_t resolver);
 extern void flow_cache_flush(void);
 extern atomic_t flow_cache_genid;
diff --git a/net/core/flow.c b/net/core/flow.c
index 2191af5f26ac..645241165e6c 100644
--- a/net/core/flow.c
+++ b/net/core/flow.c
@@ -32,7 +32,6 @@ struct flow_cache_entry {
 	u8			dir;
 	struct flowi		key;
 	u32			genid;
-	u32			sk_sid;
 	void			*object;
 	atomic_t		*object_ref;
 };
@@ -165,7 +164,7 @@ static int flow_key_compare(struct flowi *key1, struct flowi *key2)
 	return 0;
 }
 
-void *flow_cache_lookup(struct flowi *key, u32 sk_sid, u16 family, u8 dir,
+void *flow_cache_lookup(struct flowi *key, u16 family, u8 dir,
 			flow_resolve_t resolver)
 {
 	struct flow_cache_entry *fle, **head;
@@ -189,7 +188,6 @@ void *flow_cache_lookup(struct flowi *key, u32 sk_sid, u16 family, u8 dir,
 	for (fle = *head; fle; fle = fle->next) {
 		if (fle->family == family &&
 		    fle->dir == dir &&
-		    fle->sk_sid == sk_sid &&
 		    flow_key_compare(key, &fle->key) == 0) {
 			if (fle->genid == atomic_read(&flow_cache_genid)) {
 				void *ret = fle->object;
@@ -214,7 +212,6 @@ void *flow_cache_lookup(struct flowi *key, u32 sk_sid, u16 family, u8 dir,
 			*head = fle;
 			fle->family = family;
 			fle->dir = dir;
-			fle->sk_sid = sk_sid;
 			memcpy(&fle->key, key, sizeof(*key));
 			fle->object = NULL;
 			flow_count(cpu)++;
@@ -226,7 +223,7 @@ nocache:
 		void *obj;
 		atomic_t *obj_ref;
 
-		resolver(key, sk_sid, family, dir, &obj, &obj_ref);
+		resolver(key, family, dir, &obj, &obj_ref);
 
 		if (fle) {
 			fle->genid = atomic_read(&flow_cache_genid);
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 3da67ca2c3ce..79405daadc52 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -597,7 +597,7 @@ EXPORT_SYMBOL(xfrm_policy_walk);
 
 /* Find policy to apply to this flow. */
 
-static void xfrm_policy_lookup(struct flowi *fl, u32 sk_sid, u16 family, u8 dir,
+static void xfrm_policy_lookup(struct flowi *fl, u16 family, u8 dir,
 			       void **objp, atomic_t **obj_refp)
 {
 	struct xfrm_policy *pol;
@@ -613,7 +613,7 @@ static void xfrm_policy_lookup(struct flowi *fl, u32 sk_sid, u16 family, u8 dir,
 		match = xfrm_selector_match(sel, fl, family);
 
 		if (match) {
- 			if (!security_xfrm_policy_lookup(pol, sk_sid, dir)) {
+ 			if (!security_xfrm_policy_lookup(pol, fl->secid, dir)) {
 				xfrm_pol_hold(pol);
 				break;
 			}
@@ -641,7 +641,7 @@ static inline int policy_to_flow_dir(int dir)
 	};
 }
 
-static struct xfrm_policy *xfrm_sk_policy_lookup(struct sock *sk, int dir, struct flowi *fl, u32 sk_sid)
+static struct xfrm_policy *xfrm_sk_policy_lookup(struct sock *sk, int dir, struct flowi *fl)
 {
 	struct xfrm_policy *pol;
 
@@ -652,7 +652,7 @@ static struct xfrm_policy *xfrm_sk_policy_lookup(struct sock *sk, int dir, struc
  		int err = 0;
 
 		if (match)
-		  err = security_xfrm_policy_lookup(pol, sk_sid, policy_to_flow_dir(dir));
+		  err = security_xfrm_policy_lookup(pol, fl->secid, policy_to_flow_dir(dir));
 
  		if (match && !err)
 			xfrm_pol_hold(pol);
@@ -862,19 +862,20 @@ int xfrm_lookup(struct dst_entry **dst_p, struct flowi *fl,
 	u32 genid;
 	u16 family;
 	u8 dir = policy_to_flow_dir(XFRM_POLICY_OUT);
-	u32 sk_sid = security_sk_sid(sk, fl, dir);
+
+	fl->secid = security_sk_sid(sk, fl, dir);
 restart:
 	genid = atomic_read(&flow_cache_genid);
 	policy = NULL;
 	if (sk && sk->sk_policy[1])
-		policy = xfrm_sk_policy_lookup(sk, XFRM_POLICY_OUT, fl, sk_sid);
+		policy = xfrm_sk_policy_lookup(sk, XFRM_POLICY_OUT, fl);
 
 	if (!policy) {
 		/* To accelerate a bit...  */
 		if ((dst_orig->flags & DST_NOXFRM) || !xfrm_policy_list[XFRM_POLICY_OUT])
 			return 0;
 
-		policy = flow_cache_lookup(fl, sk_sid, dst_orig->ops->family,
+		policy = flow_cache_lookup(fl, dst_orig->ops->family,
 					   dir, xfrm_policy_lookup);
 	}
 
@@ -1032,13 +1033,15 @@ int
 xfrm_decode_session(struct sk_buff *skb, struct flowi *fl, unsigned short family)
 {
 	struct xfrm_policy_afinfo *afinfo = xfrm_policy_get_afinfo(family);
+	int err;
 
 	if (unlikely(afinfo == NULL))
 		return -EAFNOSUPPORT;
 
 	afinfo->decode_session(skb, fl);
+	err = security_xfrm_decode_session(skb, fl);
 	xfrm_policy_put_afinfo(afinfo);
-	return 0;
+	return err;
 }
 EXPORT_SYMBOL(xfrm_decode_session);
 
@@ -1058,14 +1061,11 @@ int __xfrm_policy_check(struct sock *sk, int dir, struct sk_buff *skb,
 	struct xfrm_policy *pol;
 	struct flowi fl;
 	u8 fl_dir = policy_to_flow_dir(dir);
-	u32 sk_sid;
 
 	if (xfrm_decode_session(skb, &fl, family) < 0)
 		return 0;
 	nf_nat_decode_session(skb, &fl, family);
 
-	sk_sid = security_sk_sid(sk, &fl, fl_dir);
-
 	/* First, check used SA against their selectors. */
 	if (skb->sp) {
 		int i;
@@ -1079,10 +1079,10 @@ int __xfrm_policy_check(struct sock *sk, int dir, struct sk_buff *skb,
 
 	pol = NULL;
 	if (sk && sk->sk_policy[dir])
-		pol = xfrm_sk_policy_lookup(sk, dir, &fl, sk_sid);
+		pol = xfrm_sk_policy_lookup(sk, dir, &fl);
 
 	if (!pol)
-		pol = flow_cache_lookup(&fl, sk_sid, family, fl_dir,
+		pol = flow_cache_lookup(&fl, family, fl_dir,
 					xfrm_policy_lookup);
 
 	if (!pol)
@@ -1298,6 +1298,8 @@ int xfrm_bundle_ok(struct xfrm_dst *first, struct flowi *fl, int family)
 
 		if (fl && !xfrm_selector_match(&dst->xfrm->sel, fl, family))
 			return 0;
+		if (fl && !security_xfrm_flow_state_match(fl, dst->xfrm))
+			return 0;
 		if (dst->xfrm->km.state != XFRM_STATE_VALID)
 			return 0;
 
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 0021aad5db43..be02bd981d12 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -367,7 +367,7 @@ xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr,
 			 */
 			if (x->km.state == XFRM_STATE_VALID) {
 				if (!xfrm_selector_match(&x->sel, fl, family) ||
-				    !xfrm_sec_ctx_match(pol->security, x->security))
+				    !security_xfrm_state_pol_flow_match(x, pol, fl))
 					continue;
 				if (!best ||
 				    best->km.dying > x->km.dying ||
@@ -379,7 +379,7 @@ xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr,
 			} else if (x->km.state == XFRM_STATE_ERROR ||
 				   x->km.state == XFRM_STATE_EXPIRED) {
  				if (xfrm_selector_match(&x->sel, fl, family) &&
-				    xfrm_sec_ctx_match(pol->security, x->security))
+				    security_xfrm_state_pol_flow_match(x, pol, fl))
 					error = -ESRCH;
 			}
 		}
@@ -403,6 +403,14 @@ xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr,
 		 * to current session. */
 		xfrm_init_tempsel(x, fl, tmpl, daddr, saddr, family);
 
+		error = security_xfrm_state_alloc_acquire(x, pol->security, fl->secid);
+		if (error) {
+			x->km.state = XFRM_STATE_DEAD;
+			xfrm_state_put(x);
+			x = NULL;
+			goto out;
+		}
+
 		if (km_query(x, tmpl, pol) == 0) {
 			x->km.state = XFRM_STATE_ACQ;
 			list_add_tail(&x->bydst, xfrm_state_bydst+h);
diff --git a/security/dummy.c b/security/dummy.c
index bd3bc5faa9a8..c1f10654871e 100644
--- a/security/dummy.c
+++ b/security/dummy.c
@@ -835,7 +835,8 @@ static int dummy_xfrm_policy_delete_security(struct xfrm_policy *xp)
 	return 0;
 }
 
-static int dummy_xfrm_state_alloc_security(struct xfrm_state *x, struct xfrm_user_sec_ctx *sec_ctx)
+static int dummy_xfrm_state_alloc_security(struct xfrm_state *x,
+	struct xfrm_user_sec_ctx *sec_ctx, struct xfrm_sec_ctx *pol, u32 secid)
 {
 	return 0;
 }
@@ -853,6 +854,23 @@ static int dummy_xfrm_policy_lookup(struct xfrm_policy *xp, u32 sk_sid, u8 dir)
 {
 	return 0;
 }
+
+static int dummy_xfrm_state_pol_flow_match(struct xfrm_state *x,
+				struct xfrm_policy *xp, struct flowi *fl)
+{
+	return 1;
+}
+
+static int dummy_xfrm_flow_state_match(struct flowi *fl, struct xfrm_state *xfrm)
+{
+	return 1;
+}
+
+static int dummy_xfrm_decode_session(struct sk_buff *skb, struct flowi *fl)
+{
+	return 0;
+}
+
 #endif /* CONFIG_SECURITY_NETWORK_XFRM */
 static int dummy_register_security (const char *name, struct security_operations *ops)
 {
@@ -1076,6 +1094,9 @@ void security_fixup_ops (struct security_operations *ops)
 	set_to_dummy_if_null(ops, xfrm_state_free_security);
 	set_to_dummy_if_null(ops, xfrm_state_delete_security);
 	set_to_dummy_if_null(ops, xfrm_policy_lookup);
+	set_to_dummy_if_null(ops, xfrm_state_pol_flow_match);
+	set_to_dummy_if_null(ops, xfrm_flow_state_match);
+	set_to_dummy_if_null(ops, xfrm_decode_session);
 #endif	/* CONFIG_SECURITY_NETWORK_XFRM */
 #ifdef CONFIG_KEYS
 	set_to_dummy_if_null(ops, key_alloc);
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index d67abf77584a..5c189da07bc9 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -3468,7 +3468,7 @@ static int selinux_socket_sock_rcv_skb(struct sock *sk, struct sk_buff *skb)
 	if (err)
 		goto out;
 
-	err = selinux_xfrm_sock_rcv_skb(sock_sid, skb);
+	err = selinux_xfrm_sock_rcv_skb(sock_sid, skb, &ad);
 out:	
 	return err;
 }
@@ -3720,7 +3720,7 @@ static unsigned int selinux_ip_postroute_last(unsigned int hooknum,
 	if (err)
 		goto out;
 
-	err = selinux_xfrm_postroute_last(isec->sid, skb);
+	err = selinux_xfrm_postroute_last(isec->sid, skb, &ad);
 out:
 	return err ? NF_DROP : NF_ACCEPT;
 }
@@ -4633,6 +4633,9 @@ static struct security_operations selinux_ops = {
 	.xfrm_state_free_security =	selinux_xfrm_state_free,
 	.xfrm_state_delete_security =	selinux_xfrm_state_delete,
 	.xfrm_policy_lookup = 		selinux_xfrm_policy_lookup,
+	.xfrm_state_pol_flow_match =	selinux_xfrm_state_pol_flow_match,
+	.xfrm_flow_state_match =	selinux_xfrm_flow_state_match,
+	.xfrm_decode_session =		selinux_xfrm_decode_session,
 #endif
 
 #ifdef CONFIG_KEYS
diff --git a/security/selinux/include/xfrm.h b/security/selinux/include/xfrm.h
index c96498a10eb8..f51a3e84bd9b 100644
--- a/security/selinux/include/xfrm.h
+++ b/security/selinux/include/xfrm.h
@@ -2,6 +2,7 @@
  * SELinux support for the XFRM LSM hooks
  *
  * Author : Trent Jaeger, <jaegert@us.ibm.com>
+ * Updated : Venkat Yekkirala, <vyekkirala@TrustedCS.com>
  */
 #ifndef _SELINUX_XFRM_H_
 #define _SELINUX_XFRM_H_
@@ -10,10 +11,16 @@ int selinux_xfrm_policy_alloc(struct xfrm_policy *xp, struct xfrm_user_sec_ctx *
 int selinux_xfrm_policy_clone(struct xfrm_policy *old, struct xfrm_policy *new);
 void selinux_xfrm_policy_free(struct xfrm_policy *xp);
 int selinux_xfrm_policy_delete(struct xfrm_policy *xp);
-int selinux_xfrm_state_alloc(struct xfrm_state *x, struct xfrm_user_sec_ctx *sec_ctx);
+int selinux_xfrm_state_alloc(struct xfrm_state *x,
+	struct xfrm_user_sec_ctx *sec_ctx, struct xfrm_sec_ctx *pol, u32 secid);
 void selinux_xfrm_state_free(struct xfrm_state *x);
 int selinux_xfrm_state_delete(struct xfrm_state *x);
-int selinux_xfrm_policy_lookup(struct xfrm_policy *xp, u32 sk_sid, u8 dir);
+int selinux_xfrm_policy_lookup(struct xfrm_policy *xp, u32 fl_secid, u8 dir);
+int selinux_xfrm_state_pol_flow_match(struct xfrm_state *x,
+			struct xfrm_policy *xp, struct flowi *fl);
+int selinux_xfrm_flow_state_match(struct flowi *fl, struct xfrm_state *xfrm);
+int selinux_xfrm_decode_session(struct sk_buff *skb, struct flowi *fl);
+
 
 /*
  * Extract the security blob from the sock (it's actually on the socket)
@@ -39,17 +46,21 @@ static inline u32 selinux_no_sk_sid(struct flowi *fl)
 }
 
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
-int selinux_xfrm_sock_rcv_skb(u32 sid, struct sk_buff *skb);
-int selinux_xfrm_postroute_last(u32 isec_sid, struct sk_buff *skb);
+int selinux_xfrm_sock_rcv_skb(u32 sid, struct sk_buff *skb,
+			struct avc_audit_data *ad);
+int selinux_xfrm_postroute_last(u32 isec_sid, struct sk_buff *skb,
+			struct avc_audit_data *ad);
 u32 selinux_socket_getpeer_stream(struct sock *sk);
 u32 selinux_socket_getpeer_dgram(struct sk_buff *skb);
 #else
-static inline int selinux_xfrm_sock_rcv_skb(u32 isec_sid, struct sk_buff *skb)
+static inline int selinux_xfrm_sock_rcv_skb(u32 isec_sid, struct sk_buff *skb,
+			struct avc_audit_data *ad)
 {
 	return 0;
 }
 
-static inline int selinux_xfrm_postroute_last(u32 isec_sid, struct sk_buff *skb)
+static inline int selinux_xfrm_postroute_last(u32 isec_sid, struct sk_buff *skb,
+			struct avc_audit_data *ad)
 {
 	return 0;
 }
diff --git a/security/selinux/xfrm.c b/security/selinux/xfrm.c
index 6c985ced8102..a502b0540e3d 100644
--- a/security/selinux/xfrm.c
+++ b/security/selinux/xfrm.c
@@ -6,7 +6,12 @@
  *  Authors:  Serge Hallyn <sergeh@us.ibm.com>
  *	      Trent Jaeger <jaegert@us.ibm.com>
  *
+ *  Updated: Venkat Yekkirala <vyekkirala@TrustedCS.com>
+ *
+ *           Granular IPSec Associations for use in MLS environments.
+ *
  *  Copyright (C) 2005 International Business Machines Corporation
+ *  Copyright (C) 2006 Trusted Computer Solutions, Inc.
  *
  *	This program is free software; you can redistribute it and/or modify
  *	it under the terms of the GNU General Public License version 2,
@@ -67,10 +72,10 @@ static inline int selinux_authorizable_xfrm(struct xfrm_state *x)
 }
 
 /*
- * LSM hook implementation that authorizes that a socket can be used
- * with the corresponding xfrm_sec_ctx and direction.
+ * LSM hook implementation that authorizes that a flow can use
+ * a xfrm policy rule.
  */
-int selinux_xfrm_policy_lookup(struct xfrm_policy *xp, u32 sk_sid, u8 dir)
+int selinux_xfrm_policy_lookup(struct xfrm_policy *xp, u32 fl_secid, u8 dir)
 {
 	int rc = 0;
 	u32 sel_sid = SECINITSID_UNLABELED;
@@ -84,27 +89,129 @@ int selinux_xfrm_policy_lookup(struct xfrm_policy *xp, u32 sk_sid, u8 dir)
 		sel_sid = ctx->ctx_sid;
 	}
 
-	rc = avc_has_perm(sk_sid, sel_sid, SECCLASS_ASSOCIATION,
-			  ((dir == FLOW_DIR_IN) ? ASSOCIATION__RECVFROM :
-			   ((dir == FLOW_DIR_OUT) ?  ASSOCIATION__SENDTO :
-			    (ASSOCIATION__SENDTO | ASSOCIATION__RECVFROM))),
+	rc = avc_has_perm(fl_secid, sel_sid, SECCLASS_ASSOCIATION,
+			  ASSOCIATION__POLMATCH,
 			  NULL);
 
 	return rc;
 }
 
+/*
+ * LSM hook implementation that authorizes that a state matches
+ * the given policy, flow combo.
+ */
+
+int selinux_xfrm_state_pol_flow_match(struct xfrm_state *x, struct xfrm_policy *xp,
+			struct flowi *fl)
+{
+	u32 state_sid;
+	u32 pol_sid;
+	int err;
+
+	if (x->security)
+		state_sid = x->security->ctx_sid;
+	else
+		state_sid = SECINITSID_UNLABELED;
+
+	if (xp->security)
+		pol_sid = xp->security->ctx_sid;
+	else
+		pol_sid = SECINITSID_UNLABELED;
+
+	err = avc_has_perm(state_sid, pol_sid, SECCLASS_ASSOCIATION,
+			  ASSOCIATION__POLMATCH,
+			  NULL);
+
+	if (err)
+		return 0;
+
+	return selinux_xfrm_flow_state_match(fl, x);
+}
+
+/*
+ * LSM hook implementation that authorizes that a particular outgoing flow
+ * can use a given security association.
+ */
+
+int selinux_xfrm_flow_state_match(struct flowi *fl, struct xfrm_state *xfrm)
+{
+	int rc = 0;
+	u32 sel_sid = SECINITSID_UNLABELED;
+	struct xfrm_sec_ctx *ctx;
+
+	/* Context sid is either set to label or ANY_ASSOC */
+	if ((ctx = xfrm->security)) {
+		if (!selinux_authorizable_ctx(ctx))
+			return 0;
+
+		sel_sid = ctx->ctx_sid;
+	}
+
+	rc = avc_has_perm(fl->secid, sel_sid, SECCLASS_ASSOCIATION,
+			  ASSOCIATION__SENDTO,
+			  NULL)? 0:1;
+
+	return rc;
+}
+
+/*
+ * LSM hook implementation that determines the sid for the session.
+ */
+
+int selinux_xfrm_decode_session(struct sk_buff *skb, struct flowi *fl)
+{
+	struct sec_path *sp;
+
+	fl->secid = SECSID_NULL;
+
+	if (skb == NULL)
+		return 0;
+
+	sp = skb->sp;
+	if (sp) {
+		int i, sid_set = 0;
+
+		for (i = sp->len-1; i >= 0; i--) {
+			struct xfrm_state *x = sp->xvec[i];
+			if (selinux_authorizable_xfrm(x)) {
+				struct xfrm_sec_ctx *ctx = x->security;
+
+				if (!sid_set) {
+					fl->secid = ctx->ctx_sid;
+					sid_set = 1;
+				}
+				else if (fl->secid != ctx->ctx_sid)
+					return -EINVAL;
+			}
+		}
+	}
+
+	return 0;
+}
+
 /*
  * Security blob allocation for xfrm_policy and xfrm_state
  * CTX does not have a meaningful value on input
  */
-static int selinux_xfrm_sec_ctx_alloc(struct xfrm_sec_ctx **ctxp, struct xfrm_user_sec_ctx *uctx)
+static int selinux_xfrm_sec_ctx_alloc(struct xfrm_sec_ctx **ctxp,
+	struct xfrm_user_sec_ctx *uctx, struct xfrm_sec_ctx *pol, u32 sid)
 {
 	int rc = 0;
 	struct task_security_struct *tsec = current->security;
-	struct xfrm_sec_ctx *ctx;
+	struct xfrm_sec_ctx *ctx = NULL;
+	char *ctx_str = NULL;
+	u32 str_len;
+	u32 ctx_sid;
+
+	BUG_ON(uctx && pol);
+
+	if (pol)
+		goto from_policy;
 
 	BUG_ON(!uctx);
-	BUG_ON(uctx->ctx_doi != XFRM_SC_ALG_SELINUX);
+
+	if (uctx->ctx_doi != XFRM_SC_ALG_SELINUX)
+		return -EINVAL;
 
 	if (uctx->ctx_len >= PAGE_SIZE)
 		return -ENOMEM;
@@ -141,9 +248,41 @@ static int selinux_xfrm_sec_ctx_alloc(struct xfrm_sec_ctx **ctxp, struct xfrm_us
 
 	return rc;
 
+from_policy:
+	BUG_ON(!pol);
+	rc = security_sid_mls_copy(pol->ctx_sid, sid, &ctx_sid);
+	if (rc)
+		goto out;
+
+	rc = security_sid_to_context(ctx_sid, &ctx_str, &str_len);
+	if (rc)
+		goto out;
+
+	*ctxp = ctx = kmalloc(sizeof(*ctx) +
+			      str_len,
+			      GFP_ATOMIC);
+
+	if (!ctx) {
+		rc = -ENOMEM;
+		goto out;
+	}
+
+
+	ctx->ctx_doi = XFRM_SC_DOI_LSM;
+	ctx->ctx_alg = XFRM_SC_ALG_SELINUX;
+	ctx->ctx_sid = ctx_sid;
+	ctx->ctx_len = str_len;
+	memcpy(ctx->ctx_str,
+	       ctx_str,
+	       str_len);
+
+	goto out2;
+
 out:
 	*ctxp = NULL;
 	kfree(ctx);
+out2:
+	kfree(ctx_str);
 	return rc;
 }
 
@@ -157,7 +296,7 @@ int selinux_xfrm_policy_alloc(struct xfrm_policy *xp, struct xfrm_user_sec_ctx *
 
 	BUG_ON(!xp);
 
-	err = selinux_xfrm_sec_ctx_alloc(&xp->security, uctx);
+	err = selinux_xfrm_sec_ctx_alloc(&xp->security, uctx, NULL, 0);
 	return err;
 }
 
@@ -217,13 +356,14 @@ int selinux_xfrm_policy_delete(struct xfrm_policy *xp)
  * LSM hook implementation that allocs and transfers sec_ctx spec to
  * xfrm_state.
  */
-int selinux_xfrm_state_alloc(struct xfrm_state *x, struct xfrm_user_sec_ctx *uctx)
+int selinux_xfrm_state_alloc(struct xfrm_state *x, struct xfrm_user_sec_ctx *uctx,
+		struct xfrm_sec_ctx *pol, u32 secid)
 {
 	int err;
 
 	BUG_ON(!x);
 
-	err = selinux_xfrm_sec_ctx_alloc(&x->security, uctx);
+	err = selinux_xfrm_sec_ctx_alloc(&x->security, uctx, pol, secid);
 	return err;
 }
 
@@ -329,38 +469,30 @@ int selinux_xfrm_state_delete(struct xfrm_state *x)
  * we need to check for unlabelled access since this may not have
  * gone thru the IPSec process.
  */
-int selinux_xfrm_sock_rcv_skb(u32 isec_sid, struct sk_buff *skb)
+int selinux_xfrm_sock_rcv_skb(u32 isec_sid, struct sk_buff *skb,
+				struct avc_audit_data *ad)
 {
 	int i, rc = 0;
 	struct sec_path *sp;
+	u32 sel_sid = SECINITSID_UNLABELED;
 
 	sp = skb->sp;
 
 	if (sp) {
-		/*
-		 * __xfrm_policy_check does not approve unless xfrm_policy_ok
-		 * says that spi's match for policy and the socket.
-		 *
-		 *  Only need to verify the existence of an authorizable sp.
-		 */
 		for (i = 0; i < sp->len; i++) {
 			struct xfrm_state *x = sp->xvec[i];
 
-			if (x && selinux_authorizable_xfrm(x))
-				goto accept;
+			if (x && selinux_authorizable_xfrm(x)) {
+				struct xfrm_sec_ctx *ctx = x->security;
+				sel_sid = ctx->ctx_sid;
+				break;
+			}
 		}
 	}
 
-	/* check SELinux sock for unlabelled access */
-	rc = avc_has_perm(isec_sid, SECINITSID_UNLABELED, SECCLASS_ASSOCIATION,
-			  ASSOCIATION__RECVFROM, NULL);
-	if (rc)
-		goto drop;
-
-accept:
-	return 0;
+	rc = avc_has_perm(isec_sid, sel_sid, SECCLASS_ASSOCIATION,
+			  ASSOCIATION__RECVFROM, ad);
 
-drop:
 	return rc;
 }
 
@@ -371,7 +503,8 @@ drop:
  * If we do have a authorizable security association, then it has already been
  * checked in xfrm_policy_lookup hook.
  */
-int selinux_xfrm_postroute_last(u32 isec_sid, struct sk_buff *skb)
+int selinux_xfrm_postroute_last(u32 isec_sid, struct sk_buff *skb,
+					struct avc_audit_data *ad)
 {
 	struct dst_entry *dst;
 	int rc = 0;
@@ -391,7 +524,7 @@ int selinux_xfrm_postroute_last(u32 isec_sid, struct sk_buff *skb)
 	}
 
 	rc = avc_has_perm(isec_sid, SECINITSID_UNLABELED, SECCLASS_ASSOCIATION,
-			  ASSOCIATION__SENDTO, NULL);
+			  ASSOCIATION__SENDTO, ad);
 out:
 	return rc;
 }
-- 
cgit v1.2.3


From beb8d13bed80f8388f1a9a107d07ddd342e627e8 Mon Sep 17 00:00:00 2001
From: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Date: Fri, 4 Aug 2006 23:12:42 -0700
Subject: [MLSXFRM]: Add flow labeling

This labels the flows that could utilize IPSec xfrms at the points the
flows are defined so that IPSec policy and SAs at the right label can
be used.

The following protos are currently not handled, but they should
continue to be able to use single-labeled IPSec like they currently
do.

ipmr
ip_gre
ipip
igmp
sit
sctp
ip6_tunnel (IPv6 over IPv6 tunnel device)
decnet

Signed-off-by: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/security.h         | 38 +++++++++++++++++++++++++-------------
 include/net/route.h              |  3 +++
 net/dccp/ipv4.c                  |  1 +
 net/dccp/ipv6.c                  |  6 ++++++
 net/ipv4/af_inet.c               |  1 +
 net/ipv4/icmp.c                  |  2 ++
 net/ipv4/inet_connection_sock.c  |  1 +
 net/ipv4/ip_output.c             |  2 ++
 net/ipv4/netfilter/ipt_REJECT.c  |  1 +
 net/ipv4/raw.c                   |  1 +
 net/ipv4/syncookies.c            |  1 +
 net/ipv4/udp.c                   |  1 +
 net/ipv6/af_inet6.c              |  1 +
 net/ipv6/datagram.c              |  2 ++
 net/ipv6/icmp.c                  |  2 ++
 net/ipv6/inet6_connection_sock.c |  1 +
 net/ipv6/ndisc.c                 |  1 +
 net/ipv6/netfilter/ip6t_REJECT.c |  1 +
 net/ipv6/raw.c                   |  1 +
 net/ipv6/tcp_ipv6.c              |  7 +++++++
 net/ipv6/udp.c                   |  2 ++
 net/xfrm/xfrm_policy.c           |  3 +--
 security/dummy.c                 |  7 +++----
 security/selinux/hooks.c         |  8 ++++----
 security/selinux/include/xfrm.h  | 14 +-------------
 security/selinux/xfrm.c          | 11 +++++++----
 26 files changed, 79 insertions(+), 40 deletions(-)

(limited to 'include')

diff --git a/include/linux/security.h b/include/linux/security.h
index 2c4921d79d19..f3909d189fe0 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -32,6 +32,7 @@
 #include <linux/sched.h>
 #include <linux/key.h>
 #include <linux/xfrm.h>
+#include <net/flow.h>
 
 struct ctl_table;
 
@@ -815,8 +816,8 @@ struct swap_info_struct;
  *	Deallocate security structure.
  * @sk_clone_security:
  *	Clone/copy security structure.
- * @sk_getsid:
- *	Retrieve the LSM-specific sid for the sock to enable caching of network
+ * @sk_getsecid:
+ *	Retrieve the LSM-specific secid for the sock to enable caching of network
  *	authorizations.
  *
  * Security hooks for XFRM operations.
@@ -882,8 +883,9 @@ struct swap_info_struct;
  *	Return 1 if there is a match.
  * @xfrm_decode_session:
  *	@skb points to skb to decode.
- *	@fl points to the flow key to set.
- *	Return 0 if successful decoding.
+ *	@secid points to the flow key secid to set.
+ *	@ckall says if all xfrms used should be checked for same secid.
+ *	Return 0 if ckall is zero or all xfrms used have the same secid.
  *
  * Security hooks affecting all Key Management operations
  *
@@ -1353,7 +1355,7 @@ struct security_operations {
 	int (*sk_alloc_security) (struct sock *sk, int family, gfp_t priority);
 	void (*sk_free_security) (struct sock *sk);
 	void (*sk_clone_security) (const struct sock *sk, struct sock *newsk);
-	unsigned int (*sk_getsid) (struct sock *sk, struct flowi *fl, u8 dir);
+	void (*sk_getsecid) (struct sock *sk, u32 *secid);
 #endif	/* CONFIG_SECURITY_NETWORK */
 
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
@@ -1370,7 +1372,7 @@ struct security_operations {
 	int (*xfrm_state_pol_flow_match)(struct xfrm_state *x,
 			struct xfrm_policy *xp, struct flowi *fl);
 	int (*xfrm_flow_state_match)(struct flowi *fl, struct xfrm_state *xfrm);
-	int (*xfrm_decode_session)(struct sk_buff *skb, struct flowi *fl);
+	int (*xfrm_decode_session)(struct sk_buff *skb, u32 *secid, int ckall);
 #endif	/* CONFIG_SECURITY_NETWORK_XFRM */
 
 	/* key management security hooks */
@@ -2917,9 +2919,9 @@ static inline void security_sk_clone(const struct sock *sk, struct sock *newsk)
 	return security_ops->sk_clone_security(sk, newsk);
 }
 
-static inline unsigned int security_sk_sid(struct sock *sk, struct flowi *fl, u8 dir)
+static inline void security_sk_classify_flow(struct sock *sk, struct flowi *fl)
 {
-	return security_ops->sk_getsid(sk, fl, dir);
+	security_ops->sk_getsecid(sk, &fl->secid);
 }
 #else	/* CONFIG_SECURITY_NETWORK */
 static inline int security_unix_stream_connect(struct socket * sock,
@@ -3047,9 +3049,8 @@ static inline void security_sk_clone(const struct sock *sk, struct sock *newsk)
 {
 }
 
-static inline unsigned int security_sk_sid(struct sock *sk, struct flowi *fl, u8 dir)
+static inline void security_sk_classify_flow(struct sock *sk, struct flowi *fl)
 {
-	return 0;
 }
 #endif	/* CONFIG_SECURITY_NETWORK */
 
@@ -3114,9 +3115,16 @@ static inline int security_xfrm_flow_state_match(struct flowi *fl, struct xfrm_s
 	return security_ops->xfrm_flow_state_match(fl, xfrm);
 }
 
-static inline int security_xfrm_decode_session(struct sk_buff *skb, struct flowi *fl)
+static inline int security_xfrm_decode_session(struct sk_buff *skb, u32 *secid)
+{
+	return security_ops->xfrm_decode_session(skb, secid, 1);
+}
+
+static inline void security_skb_classify_flow(struct sk_buff *skb, struct flowi *fl)
 {
-	return security_ops->xfrm_decode_session(skb, fl);
+	int rc = security_ops->xfrm_decode_session(skb, &fl->secid, 0);
+
+	BUG_ON(rc);
 }
 #else	/* CONFIG_SECURITY_NETWORK_XFRM */
 static inline int security_xfrm_policy_alloc(struct xfrm_policy *xp, struct xfrm_user_sec_ctx *sec_ctx)
@@ -3176,11 +3184,15 @@ static inline int security_xfrm_flow_state_match(struct flowi *fl,
 	return 1;
 }
 
-static inline int security_xfrm_decode_session(struct sk_buff *skb, struct flowi *fl)
+static inline int security_xfrm_decode_session(struct sk_buff *skb, u32 *secid)
 {
 	return 0;
 }
 
+static inline void security_skb_classify_flow(struct sk_buff *skb, struct flowi *fl)
+{
+}
+
 #endif	/* CONFIG_SECURITY_NETWORK_XFRM */
 
 #ifdef CONFIG_KEYS
diff --git a/include/net/route.h b/include/net/route.h
index c4a068692dcc..7f93ac0e0899 100644
--- a/include/net/route.h
+++ b/include/net/route.h
@@ -32,6 +32,7 @@
 #include <linux/route.h>
 #include <linux/ip.h>
 #include <linux/cache.h>
+#include <linux/security.h>
 
 #ifndef __KERNEL__
 #warning This file is not supposed to be used outside of kernel.
@@ -166,6 +167,7 @@ static inline int ip_route_connect(struct rtable **rp, u32 dst,
 		ip_rt_put(*rp);
 		*rp = NULL;
 	}
+	security_sk_classify_flow(sk, &fl);
 	return ip_route_output_flow(rp, &fl, sk, 0);
 }
 
@@ -182,6 +184,7 @@ static inline int ip_route_newports(struct rtable **rp, u8 protocol,
 		fl.proto = protocol;
 		ip_rt_put(*rp);
 		*rp = NULL;
+		security_sk_classify_flow(sk, &fl);
 		return ip_route_output_flow(rp, &fl, sk, 0);
 	}
 	return 0;
diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c
index 7f56f7e8f571..386498053b1c 100644
--- a/net/dccp/ipv4.c
+++ b/net/dccp/ipv4.c
@@ -678,6 +678,7 @@ static struct dst_entry* dccp_v4_route_skb(struct sock *sk,
 			   	     }
 			  };
 
+	security_skb_classify_flow(skb, &fl);
 	if (ip_route_output_flow(&rt, &fl, sk, 0)) {
 		IP_INC_STATS_BH(IPSTATS_MIB_OUTNOROUTES);
 		return NULL;
diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c
index 610c722ac27f..53d255c01431 100644
--- a/net/dccp/ipv6.c
+++ b/net/dccp/ipv6.c
@@ -201,6 +201,7 @@ static int dccp_v6_connect(struct sock *sk, struct sockaddr *uaddr,
 	fl.oif = sk->sk_bound_dev_if;
 	fl.fl_ip_dport = usin->sin6_port;
 	fl.fl_ip_sport = inet->sport;
+	security_sk_classify_flow(sk, &fl);
 
 	if (np->opt != NULL && np->opt->srcrt != NULL) {
 		const struct rt0_hdr *rt0 = (struct rt0_hdr *)np->opt->srcrt;
@@ -322,6 +323,7 @@ static void dccp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
 			fl.oif = sk->sk_bound_dev_if;
 			fl.fl_ip_dport = inet->dport;
 			fl.fl_ip_sport = inet->sport;
+			security_sk_classify_flow(sk, &fl);
 
 			err = ip6_dst_lookup(sk, &dst, &fl);
 			if (err) {
@@ -422,6 +424,7 @@ static int dccp_v6_send_response(struct sock *sk, struct request_sock *req,
 	fl.oif = ireq6->iif;
 	fl.fl_ip_dport = inet_rsk(req)->rmt_port;
 	fl.fl_ip_sport = inet_sk(sk)->sport;
+	security_sk_classify_flow(sk, &fl);
 
 	if (dst == NULL) {
 		opt = np->opt;
@@ -566,6 +569,7 @@ static void dccp_v6_ctl_send_reset(struct sk_buff *rxskb)
 	fl.oif = inet6_iif(rxskb);
 	fl.fl_ip_dport = dh->dccph_dport;
 	fl.fl_ip_sport = dh->dccph_sport;
+	security_skb_classify_flow(rxskb, &fl);
 
 	/* sk = NULL, but it is safe for now. RST socket required. */
 	if (!ip6_dst_lookup(NULL, &skb->dst, &fl)) {
@@ -622,6 +626,7 @@ static void dccp_v6_reqsk_send_ack(struct sk_buff *rxskb,
 	fl.oif = inet6_iif(rxskb);
 	fl.fl_ip_dport = dh->dccph_dport;
 	fl.fl_ip_sport = dh->dccph_sport;
+	security_skb_classify_flow(rxskb, &fl);
 
 	if (!ip6_dst_lookup(NULL, &skb->dst, &fl)) {
 		if (xfrm_lookup(&skb->dst, &fl, NULL, 0) >= 0) {
@@ -842,6 +847,7 @@ static struct sock *dccp_v6_request_recv_sock(struct sock *sk,
 		fl.oif = sk->sk_bound_dev_if;
 		fl.fl_ip_dport = inet_rsk(req)->rmt_port;
 		fl.fl_ip_sport = inet_sk(sk)->sport;
+		security_sk_classify_flow(sk, &fl);
 
 		if (ip6_dst_lookup(sk, &dst, &fl))
 			goto out;
diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index c84a32070f8d..fc40da3b6d39 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -1074,6 +1074,7 @@ int inet_sk_rebuild_header(struct sock *sk)
 		},
 	};
 						
+	security_sk_classify_flow(sk, &fl);
 	err = ip_route_output_flow(&rt, &fl, sk, 0);
 }
 	if (!err)
diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c
index 4c86ac3d882d..6ad797c14163 100644
--- a/net/ipv4/icmp.c
+++ b/net/ipv4/icmp.c
@@ -406,6 +406,7 @@ static void icmp_reply(struct icmp_bxm *icmp_param, struct sk_buff *skb)
 						.saddr = rt->rt_spec_dst,
 						.tos = RT_TOS(skb->nh.iph->tos) } },
 				    .proto = IPPROTO_ICMP };
+		security_skb_classify_flow(skb, &fl);
 		if (ip_route_output_key(&rt, &fl))
 			goto out_unlock;
 	}
@@ -560,6 +561,7 @@ void icmp_send(struct sk_buff *skb_in, int type, int code, u32 info)
 				}
 			}
 		};
+		security_skb_classify_flow(skb_in, &fl);
 		if (ip_route_output_key(&rt, &fl))
 			goto out_unlock;
 	}
diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c
index e50a1bfd7ccc..772b4eac78bc 100644
--- a/net/ipv4/inet_connection_sock.c
+++ b/net/ipv4/inet_connection_sock.c
@@ -327,6 +327,7 @@ struct dst_entry* inet_csk_route_req(struct sock *sk,
 				       { .sport = inet_sk(sk)->sport,
 					 .dport = ireq->rmt_port } } };
 
+	security_sk_classify_flow(sk, &fl);
 	if (ip_route_output_flow(&rt, &fl, sk, 0)) {
 		IP_INC_STATS_BH(IPSTATS_MIB_OUTNOROUTES);
 		return NULL;
diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index a2ede167e045..308bdeac3455 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -328,6 +328,7 @@ int ip_queue_xmit(struct sk_buff *skb, int ipfragok)
 			 * keep trying until route appears or the connection times
 			 * itself out.
 			 */
+			security_sk_classify_flow(sk, &fl);
 			if (ip_route_output_flow(&rt, &fl, sk, 0))
 				goto no_route;
 		}
@@ -1366,6 +1367,7 @@ void ip_send_reply(struct sock *sk, struct sk_buff *skb, struct ip_reply_arg *ar
 					       { .sport = skb->h.th->dest,
 					         .dport = skb->h.th->source } },
 				    .proto = sk->sk_protocol };
+		security_skb_classify_flow(skb, &fl);
 		if (ip_route_output_key(&rt, &fl))
 			return;
 	}
diff --git a/net/ipv4/netfilter/ipt_REJECT.c b/net/ipv4/netfilter/ipt_REJECT.c
index 269bc2067cb8..7f905bf2bde5 100644
--- a/net/ipv4/netfilter/ipt_REJECT.c
+++ b/net/ipv4/netfilter/ipt_REJECT.c
@@ -90,6 +90,7 @@ static inline struct rtable *route_reverse(struct sk_buff *skb,
 	fl.proto = IPPROTO_TCP;
 	fl.fl_ip_sport = tcph->dest;
 	fl.fl_ip_dport = tcph->source;
+	security_skb_classify_flow(skb, &fl);
 
 	xfrm_lookup((struct dst_entry **)&rt, &fl, NULL, 0);
 
diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c
index 62b2762a2420..fe44cb50a1c5 100644
--- a/net/ipv4/raw.c
+++ b/net/ipv4/raw.c
@@ -484,6 +484,7 @@ static int raw_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
 		if (!inet->hdrincl)
 			raw_probe_proto_opt(&fl, msg);
 
+		security_sk_classify_flow(sk, &fl);
 		err = ip_route_output_flow(&rt, &fl, sk, !(msg->msg_flags&MSG_DONTWAIT));
 	}
 	if (err)
diff --git a/net/ipv4/syncookies.c b/net/ipv4/syncookies.c
index e20be3331f67..307dc3c0d635 100644
--- a/net/ipv4/syncookies.c
+++ b/net/ipv4/syncookies.c
@@ -259,6 +259,7 @@ struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb,
 				    .uli_u = { .ports =
 					       { .sport = skb->h.th->dest,
 						 .dport = skb->h.th->source } } };
+		security_sk_classify_flow(sk, &fl);
 		if (ip_route_output_key(&rt, &fl)) {
 			reqsk_free(req);
 			goto out; 
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index f136cec96d95..a4d005eccc7f 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -603,6 +603,7 @@ int udp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
 				    .uli_u = { .ports =
 					       { .sport = inet->sport,
 						 .dport = dport } } };
+		security_sk_classify_flow(sk, &fl);
 		err = ip_route_output_flow(&rt, &fl, sk, !(msg->msg_flags&MSG_DONTWAIT));
 		if (err)
 			goto out;
diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c
index ac85e9c532c2..82a1b1a328db 100644
--- a/net/ipv6/af_inet6.c
+++ b/net/ipv6/af_inet6.c
@@ -637,6 +637,7 @@ int inet6_sk_rebuild_header(struct sock *sk)
 		fl.oif = sk->sk_bound_dev_if;
 		fl.fl_ip_dport = inet->dport;
 		fl.fl_ip_sport = inet->sport;
+		security_sk_classify_flow(sk, &fl);
 
 		if (np->opt && np->opt->srcrt) {
 			struct rt0_hdr *rt0 = (struct rt0_hdr *) np->opt->srcrt;
diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c
index 3b55b4c8e2d1..c73508e090a6 100644
--- a/net/ipv6/datagram.c
+++ b/net/ipv6/datagram.c
@@ -156,6 +156,8 @@ ipv4_connected:
 	if (!fl.oif && (addr_type&IPV6_ADDR_MULTICAST))
 		fl.oif = np->mcast_oif;
 
+	security_sk_classify_flow(sk, &fl);
+
 	if (flowlabel) {
 		if (flowlabel->opt && flowlabel->opt->srcrt) {
 			struct rt0_hdr *rt0 = (struct rt0_hdr *) flowlabel->opt->srcrt;
diff --git a/net/ipv6/icmp.c b/net/ipv6/icmp.c
index 356a8a7ef22a..dbfce089e916 100644
--- a/net/ipv6/icmp.c
+++ b/net/ipv6/icmp.c
@@ -358,6 +358,7 @@ void icmpv6_send(struct sk_buff *skb, int type, int code, __u32 info,
 	fl.oif = iif;
 	fl.fl_icmp_type = type;
 	fl.fl_icmp_code = code;
+	security_skb_classify_flow(skb, &fl);
 
 	if (icmpv6_xmit_lock())
 		return;
@@ -472,6 +473,7 @@ static void icmpv6_echo_reply(struct sk_buff *skb)
 		ipv6_addr_copy(&fl.fl6_src, saddr);
 	fl.oif = skb->dev->ifindex;
 	fl.fl_icmp_type = ICMPV6_ECHO_REPLY;
+	security_skb_classify_flow(skb, &fl);
 
 	if (icmpv6_xmit_lock())
 		return;
diff --git a/net/ipv6/inet6_connection_sock.c b/net/ipv6/inet6_connection_sock.c
index bf491077b822..7a51a258615d 100644
--- a/net/ipv6/inet6_connection_sock.c
+++ b/net/ipv6/inet6_connection_sock.c
@@ -157,6 +157,7 @@ int inet6_csk_xmit(struct sk_buff *skb, int ipfragok)
 	fl.oif = sk->sk_bound_dev_if;
 	fl.fl_ip_sport = inet->sport;
 	fl.fl_ip_dport = inet->dport;
+	security_sk_classify_flow(sk, &fl);
 
 	if (np->opt && np->opt->srcrt) {
 		struct rt0_hdr *rt0 = (struct rt0_hdr *)np->opt->srcrt;
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index b50055b9278d..67cfc3813c32 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -419,6 +419,7 @@ static inline void ndisc_flow_init(struct flowi *fl, u8 type,
 	fl->proto	 	= IPPROTO_ICMPV6;
 	fl->fl_icmp_type	= type;
 	fl->fl_icmp_code	= 0;
+	security_sk_classify_flow(ndisc_socket->sk, fl);
 }
 
 static void ndisc_send_na(struct net_device *dev, struct neighbour *neigh,
diff --git a/net/ipv6/netfilter/ip6t_REJECT.c b/net/ipv6/netfilter/ip6t_REJECT.c
index 8629ba195d2d..c4eba1aeb323 100644
--- a/net/ipv6/netfilter/ip6t_REJECT.c
+++ b/net/ipv6/netfilter/ip6t_REJECT.c
@@ -96,6 +96,7 @@ static void send_reset(struct sk_buff *oldskb)
 	ipv6_addr_copy(&fl.fl6_dst, &oip6h->saddr);
 	fl.fl_ip_sport = otcph.dest;
 	fl.fl_ip_dport = otcph.source;
+	security_skb_classify_flow(oldskb, &fl);
 	dst = ip6_route_output(NULL, &fl);
 	if (dst == NULL)
 		return;
diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c
index 15b862d8acab..d5040e172292 100644
--- a/net/ipv6/raw.c
+++ b/net/ipv6/raw.c
@@ -759,6 +759,7 @@ static int rawv6_sendmsg(struct kiocb *iocb, struct sock *sk,
 
 	if (!fl.oif && ipv6_addr_is_multicast(&fl.fl6_dst))
 		fl.oif = np->mcast_oif;
+	security_sk_classify_flow(sk, &fl);
 
 	err = ip6_dst_lookup(sk, &dst, &fl);
 	if (err)
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 802a1a6b1037..46922e57e311 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -251,6 +251,8 @@ static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr,
 		final_p = &final;
 	}
 
+	security_sk_classify_flow(sk, &fl);
+
 	err = ip6_dst_lookup(sk, &dst, &fl);
 	if (err)
 		goto failure;
@@ -374,6 +376,7 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
 			fl.oif = sk->sk_bound_dev_if;
 			fl.fl_ip_dport = inet->dport;
 			fl.fl_ip_sport = inet->sport;
+			security_skb_classify_flow(skb, &fl);
 
 			if ((err = ip6_dst_lookup(sk, &dst, &fl))) {
 				sk->sk_err_soft = -err;
@@ -467,6 +470,7 @@ static int tcp_v6_send_synack(struct sock *sk, struct request_sock *req,
 	fl.oif = treq->iif;
 	fl.fl_ip_dport = inet_rsk(req)->rmt_port;
 	fl.fl_ip_sport = inet_sk(sk)->sport;
+	security_sk_classify_flow(sk, &fl);
 
 	if (dst == NULL) {
 		opt = np->opt;
@@ -625,6 +629,7 @@ static void tcp_v6_send_reset(struct sk_buff *skb)
 	fl.oif = inet6_iif(skb);
 	fl.fl_ip_dport = t1->dest;
 	fl.fl_ip_sport = t1->source;
+	security_skb_classify_flow(skb, &fl);
 
 	/* sk = NULL, but it is safe for now. RST socket required. */
 	if (!ip6_dst_lookup(NULL, &buff->dst, &fl)) {
@@ -691,6 +696,7 @@ static void tcp_v6_send_ack(struct sk_buff *skb, u32 seq, u32 ack, u32 win, u32
 	fl.oif = inet6_iif(skb);
 	fl.fl_ip_dport = t1->dest;
 	fl.fl_ip_sport = t1->source;
+	security_skb_classify_flow(skb, &fl);
 
 	if (!ip6_dst_lookup(NULL, &buff->dst, &fl)) {
 		if (xfrm_lookup(&buff->dst, &fl, NULL, 0) >= 0) {
@@ -923,6 +929,7 @@ static struct sock * tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb,
 		fl.oif = sk->sk_bound_dev_if;
 		fl.fl_ip_dport = inet_rsk(req)->rmt_port;
 		fl.fl_ip_sport = inet_sk(sk)->sport;
+		security_sk_classify_flow(sk, &fl);
 
 		if (ip6_dst_lookup(sk, &dst, &fl))
 			goto out;
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index 3d54f246411e..82c7c9cde2a8 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -782,6 +782,8 @@ do_udp_sendmsg:
 		connected = 0;
 	}
 
+	security_sk_classify_flow(sk, fl);
+
 	err = ip6_sk_dst_lookup(sk, &dst, fl);
 	if (err)
 		goto out;
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 79405daadc52..32c963c90573 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -863,7 +863,6 @@ int xfrm_lookup(struct dst_entry **dst_p, struct flowi *fl,
 	u16 family;
 	u8 dir = policy_to_flow_dir(XFRM_POLICY_OUT);
 
-	fl->secid = security_sk_sid(sk, fl, dir);
 restart:
 	genid = atomic_read(&flow_cache_genid);
 	policy = NULL;
@@ -1039,7 +1038,7 @@ xfrm_decode_session(struct sk_buff *skb, struct flowi *fl, unsigned short family
 		return -EAFNOSUPPORT;
 
 	afinfo->decode_session(skb, fl);
-	err = security_xfrm_decode_session(skb, fl);
+	err = security_xfrm_decode_session(skb, &fl->secid);
 	xfrm_policy_put_afinfo(afinfo);
 	return err;
 }
diff --git a/security/dummy.c b/security/dummy.c
index c1f10654871e..c0ff6b9bfd7d 100644
--- a/security/dummy.c
+++ b/security/dummy.c
@@ -809,9 +809,8 @@ static inline void dummy_sk_clone_security (const struct sock *sk, struct sock *
 {
 }
 
-static unsigned int dummy_sk_getsid(struct sock *sk, struct flowi *fl, u8 dir)
+static inline void dummy_sk_getsecid(struct sock *sk, u32 *secid)
 {
-	return 0;
 }
 #endif	/* CONFIG_SECURITY_NETWORK */
 
@@ -866,7 +865,7 @@ static int dummy_xfrm_flow_state_match(struct flowi *fl, struct xfrm_state *xfrm
 	return 1;
 }
 
-static int dummy_xfrm_decode_session(struct sk_buff *skb, struct flowi *fl)
+static int dummy_xfrm_decode_session(struct sk_buff *skb, u32 *fl, int ckall)
 {
 	return 0;
 }
@@ -1083,7 +1082,7 @@ void security_fixup_ops (struct security_operations *ops)
 	set_to_dummy_if_null(ops, sk_alloc_security);
 	set_to_dummy_if_null(ops, sk_free_security);
 	set_to_dummy_if_null(ops, sk_clone_security);
-	set_to_dummy_if_null(ops, sk_getsid);
+	set_to_dummy_if_null(ops, sk_getsecid);
  #endif	/* CONFIG_SECURITY_NETWORK */
 #ifdef  CONFIG_SECURITY_NETWORK_XFRM
 	set_to_dummy_if_null(ops, xfrm_policy_alloc_security);
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index 5c189da07bc9..4e5989d584ce 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -3561,14 +3561,14 @@ static void selinux_sk_clone_security(const struct sock *sk, struct sock *newsk)
 	newssec->peer_sid = ssec->peer_sid;
 }
 
-static unsigned int selinux_sk_getsid_security(struct sock *sk, struct flowi *fl, u8 dir)
+static void selinux_sk_getsecid(struct sock *sk, u32 *secid)
 {
 	if (!sk)
-		return selinux_no_sk_sid(fl);
+		*secid = SECINITSID_ANY_SOCKET;
 	else {
 		struct sk_security_struct *sksec = sk->sk_security;
 
-		return sksec->sid;
+		*secid = sksec->sid;
 	}
 }
 
@@ -4622,7 +4622,7 @@ static struct security_operations selinux_ops = {
 	.sk_alloc_security =		selinux_sk_alloc_security,
 	.sk_free_security =		selinux_sk_free_security,
 	.sk_clone_security =		selinux_sk_clone_security,
-	.sk_getsid = 			selinux_sk_getsid_security,
+	.sk_getsecid = 			selinux_sk_getsecid,
 
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
 	.xfrm_policy_alloc_security =	selinux_xfrm_policy_alloc,
diff --git a/security/selinux/include/xfrm.h b/security/selinux/include/xfrm.h
index f51a3e84bd9b..8e45c1d588a8 100644
--- a/security/selinux/include/xfrm.h
+++ b/security/selinux/include/xfrm.h
@@ -19,7 +19,7 @@ int selinux_xfrm_policy_lookup(struct xfrm_policy *xp, u32 fl_secid, u8 dir);
 int selinux_xfrm_state_pol_flow_match(struct xfrm_state *x,
 			struct xfrm_policy *xp, struct flowi *fl);
 int selinux_xfrm_flow_state_match(struct flowi *fl, struct xfrm_state *xfrm);
-int selinux_xfrm_decode_session(struct sk_buff *skb, struct flowi *fl);
+int selinux_xfrm_decode_session(struct sk_buff *skb, u32 *fl, int ckall);
 
 
 /*
@@ -33,18 +33,6 @@ static inline struct inode_security_struct *get_sock_isec(struct sock *sk)
 	return SOCK_INODE(sk->sk_socket)->i_security;
 }
 
-
-static inline u32 selinux_no_sk_sid(struct flowi *fl)
-{
-	/* NOTE: no sock occurs on ICMP reply, forwards, ... */
-	/* icmp_reply: authorize as kernel packet */
-	if (fl && fl->proto == IPPROTO_ICMP) {
-		return SECINITSID_KERNEL;
-	}
-
-	return SECINITSID_ANY_SOCKET;
-}
-
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
 int selinux_xfrm_sock_rcv_skb(u32 sid, struct sk_buff *skb,
 			struct avc_audit_data *ad);
diff --git a/security/selinux/xfrm.c b/security/selinux/xfrm.c
index a502b0540e3d..c750ef7af66f 100644
--- a/security/selinux/xfrm.c
+++ b/security/selinux/xfrm.c
@@ -158,11 +158,11 @@ int selinux_xfrm_flow_state_match(struct flowi *fl, struct xfrm_state *xfrm)
  * LSM hook implementation that determines the sid for the session.
  */
 
-int selinux_xfrm_decode_session(struct sk_buff *skb, struct flowi *fl)
+int selinux_xfrm_decode_session(struct sk_buff *skb, u32 *sid, int ckall)
 {
 	struct sec_path *sp;
 
-	fl->secid = SECSID_NULL;
+	*sid = SECSID_NULL;
 
 	if (skb == NULL)
 		return 0;
@@ -177,10 +177,13 @@ int selinux_xfrm_decode_session(struct sk_buff *skb, struct flowi *fl)
 				struct xfrm_sec_ctx *ctx = x->security;
 
 				if (!sid_set) {
-					fl->secid = ctx->ctx_sid;
+					*sid = ctx->ctx_sid;
 					sid_set = 1;
+
+					if (!ckall)
+						break;
 				}
-				else if (fl->secid != ctx->ctx_sid)
+				else if (*sid != ctx->ctx_sid)
 					return -EINVAL;
 			}
 		}
-- 
cgit v1.2.3


From cb969f072b6d67770b559617f14e767f47e77ece Mon Sep 17 00:00:00 2001
From: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Date: Mon, 24 Jul 2006 23:32:20 -0700
Subject: [MLSXFRM]: Default labeling of socket specific IPSec policies

This defaults the label of socket-specific IPSec policies to be the
same as the socket they are set on.

Signed-off-by: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/security.h        | 19 ++++++++++++++++---
 include/net/xfrm.h              |  2 +-
 net/key/af_key.c                | 15 +++++++++++----
 net/xfrm/xfrm_state.c           |  2 +-
 net/xfrm/xfrm_user.c            | 13 +++++++++++--
 security/dummy.c                |  3 ++-
 security/selinux/include/xfrm.h |  3 ++-
 security/selinux/xfrm.c         | 33 ++++++++++++++++++++++-----------
 8 files changed, 66 insertions(+), 24 deletions(-)

(limited to 'include')

diff --git a/include/linux/security.h b/include/linux/security.h
index f3909d189fe0..8e3dc6c51a6d 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -827,8 +827,10 @@ struct swap_info_struct;
  *	used by the XFRM system.
  *	@sec_ctx contains the security context information being provided by
  *	the user-level policy update program (e.g., setkey).
+ *	@sk refers to the sock from which to derive the security context.
  *	Allocate a security structure to the xp->security field; the security
- *	field is initialized to NULL when the xfrm_policy is allocated.
+ *	field is initialized to NULL when the xfrm_policy is allocated. Only
+ *	one of sec_ctx or sock can be specified.
  *	Return 0 if operation was successful (memory to allocate, legal context)
  * @xfrm_policy_clone_security:
  *	@old contains an existing xfrm_policy in the SPD.
@@ -1359,7 +1361,8 @@ struct security_operations {
 #endif	/* CONFIG_SECURITY_NETWORK */
 
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
-	int (*xfrm_policy_alloc_security) (struct xfrm_policy *xp, struct xfrm_user_sec_ctx *sec_ctx);
+	int (*xfrm_policy_alloc_security) (struct xfrm_policy *xp,
+			struct xfrm_user_sec_ctx *sec_ctx, struct sock *sk);
 	int (*xfrm_policy_clone_security) (struct xfrm_policy *old, struct xfrm_policy *new);
 	void (*xfrm_policy_free_security) (struct xfrm_policy *xp);
 	int (*xfrm_policy_delete_security) (struct xfrm_policy *xp);
@@ -3057,7 +3060,12 @@ static inline void security_sk_classify_flow(struct sock *sk, struct flowi *fl)
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
 static inline int security_xfrm_policy_alloc(struct xfrm_policy *xp, struct xfrm_user_sec_ctx *sec_ctx)
 {
-	return security_ops->xfrm_policy_alloc_security(xp, sec_ctx);
+	return security_ops->xfrm_policy_alloc_security(xp, sec_ctx, NULL);
+}
+
+static inline int security_xfrm_sock_policy_alloc(struct xfrm_policy *xp, struct sock *sk)
+{
+	return security_ops->xfrm_policy_alloc_security(xp, NULL, sk);
 }
 
 static inline int security_xfrm_policy_clone(struct xfrm_policy *old, struct xfrm_policy *new)
@@ -3132,6 +3140,11 @@ static inline int security_xfrm_policy_alloc(struct xfrm_policy *xp, struct xfrm
 	return 0;
 }
 
+static inline int security_xfrm_sock_policy_alloc(struct xfrm_policy *xp, struct sock *sk)
+{
+	return 0;
+}
+
 static inline int security_xfrm_policy_clone(struct xfrm_policy *old, struct xfrm_policy *new)
 {
 	return 0;
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 3ecd9fa1ed4b..00bf86e6e82b 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -362,7 +362,7 @@ struct xfrm_mgr
 	char			*id;
 	int			(*notify)(struct xfrm_state *x, struct km_event *c);
 	int			(*acquire)(struct xfrm_state *x, struct xfrm_tmpl *, struct xfrm_policy *xp, int dir);
-	struct xfrm_policy	*(*compile_policy)(u16 family, int opt, u8 *data, int len, int *dir);
+	struct xfrm_policy	*(*compile_policy)(struct sock *sk, int opt, u8 *data, int len, int *dir);
 	int			(*new_mapping)(struct xfrm_state *x, xfrm_address_t *ipaddr, u16 sport);
 	int			(*notify_policy)(struct xfrm_policy *x, int dir, struct km_event *c);
 };
diff --git a/net/key/af_key.c b/net/key/af_key.c
index a065e1a67773..797c744a8438 100644
--- a/net/key/af_key.c
+++ b/net/key/af_key.c
@@ -2843,14 +2843,14 @@ static int pfkey_send_acquire(struct xfrm_state *x, struct xfrm_tmpl *t, struct
 	return pfkey_broadcast(skb, GFP_ATOMIC, BROADCAST_REGISTERED, NULL);
 }
 
-static struct xfrm_policy *pfkey_compile_policy(u16 family, int opt,
+static struct xfrm_policy *pfkey_compile_policy(struct sock *sk, int opt,
                                                 u8 *data, int len, int *dir)
 {
 	struct xfrm_policy *xp;
 	struct sadb_x_policy *pol = (struct sadb_x_policy*)data;
 	struct sadb_x_sec_ctx *sec_ctx;
 
-	switch (family) {
+	switch (sk->sk_family) {
 	case AF_INET:
 		if (opt != IP_IPSEC_POLICY) {
 			*dir = -EOPNOTSUPP;
@@ -2891,7 +2891,7 @@ static struct xfrm_policy *pfkey_compile_policy(u16 family, int opt,
 	xp->lft.hard_byte_limit = XFRM_INF;
 	xp->lft.soft_packet_limit = XFRM_INF;
 	xp->lft.hard_packet_limit = XFRM_INF;
-	xp->family = family;
+	xp->family = sk->sk_family;
 
 	xp->xfrm_nr = 0;
 	if (pol->sadb_x_policy_type == IPSEC_POLICY_IPSEC &&
@@ -2907,8 +2907,10 @@ static struct xfrm_policy *pfkey_compile_policy(u16 family, int opt,
 		p += pol->sadb_x_policy_len*8;
 		sec_ctx = (struct sadb_x_sec_ctx *)p;
 		if (len < pol->sadb_x_policy_len*8 +
-		    sec_ctx->sadb_x_sec_len)
+		    sec_ctx->sadb_x_sec_len) {
+			*dir = -EINVAL;
 			goto out;
+		}
 		if ((*dir = verify_sec_ctx_len(p)))
 			goto out;
 		uctx = pfkey_sadb2xfrm_user_sec_ctx(sec_ctx);
@@ -2918,6 +2920,11 @@ static struct xfrm_policy *pfkey_compile_policy(u16 family, int opt,
 		if (*dir)
 			goto out;
 	}
+	else {
+		*dir = security_xfrm_sock_policy_alloc(xp, sk);
+		if (*dir)
+			goto out;
+	}
 
 	*dir = pol->sadb_x_policy_dir-1;
 	return xp;
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index be02bd981d12..1c796087ee78 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -1026,7 +1026,7 @@ int xfrm_user_policy(struct sock *sk, int optname, u8 __user *optval, int optlen
 	err = -EINVAL;
 	read_lock(&xfrm_km_lock);
 	list_for_each_entry(km, &xfrm_km_list, list) {
-		pol = km->compile_policy(sk->sk_family, optname, data,
+		pol = km->compile_policy(sk, optname, data,
 					 optlen, &err);
 		if (err >= 0)
 			break;
diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c
index dac8db1088bc..f70e158874d2 100644
--- a/net/xfrm/xfrm_user.c
+++ b/net/xfrm/xfrm_user.c
@@ -1757,7 +1757,7 @@ static int xfrm_send_acquire(struct xfrm_state *x, struct xfrm_tmpl *xt,
 /* User gives us xfrm_user_policy_info followed by an array of 0
  * or more templates.
  */
-static struct xfrm_policy *xfrm_compile_policy(u16 family, int opt,
+static struct xfrm_policy *xfrm_compile_policy(struct sock *sk, int opt,
 					       u8 *data, int len, int *dir)
 {
 	struct xfrm_userpolicy_info *p = (struct xfrm_userpolicy_info *)data;
@@ -1765,7 +1765,7 @@ static struct xfrm_policy *xfrm_compile_policy(u16 family, int opt,
 	struct xfrm_policy *xp;
 	int nr;
 
-	switch (family) {
+	switch (sk->sk_family) {
 	case AF_INET:
 		if (opt != IP_XFRM_POLICY) {
 			*dir = -EOPNOTSUPP;
@@ -1807,6 +1807,15 @@ static struct xfrm_policy *xfrm_compile_policy(u16 family, int opt,
 	copy_from_user_policy(xp, p);
 	copy_templates(xp, ut, nr);
 
+	if (!xp->security) {
+		int err = security_xfrm_sock_policy_alloc(xp, sk);
+		if (err) {
+			kfree(xp);
+			*dir = err;
+			return NULL;
+		}
+	}
+
 	*dir = p->dir;
 
 	return xp;
diff --git a/security/dummy.c b/security/dummy.c
index c0ff6b9bfd7d..66cc06404930 100644
--- a/security/dummy.c
+++ b/security/dummy.c
@@ -815,7 +815,8 @@ static inline void dummy_sk_getsecid(struct sock *sk, u32 *secid)
 #endif	/* CONFIG_SECURITY_NETWORK */
 
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
-static int dummy_xfrm_policy_alloc_security(struct xfrm_policy *xp, struct xfrm_user_sec_ctx *sec_ctx)
+static int dummy_xfrm_policy_alloc_security(struct xfrm_policy *xp,
+		struct xfrm_user_sec_ctx *sec_ctx, struct sock *sk)
 {
 	return 0;
 }
diff --git a/security/selinux/include/xfrm.h b/security/selinux/include/xfrm.h
index 8e45c1d588a8..1822c73e5085 100644
--- a/security/selinux/include/xfrm.h
+++ b/security/selinux/include/xfrm.h
@@ -7,7 +7,8 @@
 #ifndef _SELINUX_XFRM_H_
 #define _SELINUX_XFRM_H_
 
-int selinux_xfrm_policy_alloc(struct xfrm_policy *xp, struct xfrm_user_sec_ctx *sec_ctx);
+int selinux_xfrm_policy_alloc(struct xfrm_policy *xp,
+		struct xfrm_user_sec_ctx *sec_ctx, struct sock *sk);
 int selinux_xfrm_policy_clone(struct xfrm_policy *old, struct xfrm_policy *new);
 void selinux_xfrm_policy_free(struct xfrm_policy *xp);
 int selinux_xfrm_policy_delete(struct xfrm_policy *xp);
diff --git a/security/selinux/xfrm.c b/security/selinux/xfrm.c
index c750ef7af66f..d3690f985135 100644
--- a/security/selinux/xfrm.c
+++ b/security/selinux/xfrm.c
@@ -208,10 +208,8 @@ static int selinux_xfrm_sec_ctx_alloc(struct xfrm_sec_ctx **ctxp,
 
 	BUG_ON(uctx && pol);
 
-	if (pol)
-		goto from_policy;
-
-	BUG_ON(!uctx);
+	if (!uctx)
+		goto not_from_user;
 
 	if (uctx->ctx_doi != XFRM_SC_ALG_SELINUX)
 		return -EINVAL;
@@ -251,11 +249,14 @@ static int selinux_xfrm_sec_ctx_alloc(struct xfrm_sec_ctx **ctxp,
 
 	return rc;
 
-from_policy:
-	BUG_ON(!pol);
-	rc = security_sid_mls_copy(pol->ctx_sid, sid, &ctx_sid);
-	if (rc)
-		goto out;
+not_from_user:
+	if (pol) {
+		rc = security_sid_mls_copy(pol->ctx_sid, sid, &ctx_sid);
+		if (rc)
+			goto out;
+	}
+	else
+		ctx_sid = sid;
 
 	rc = security_sid_to_context(ctx_sid, &ctx_str, &str_len);
 	if (rc)
@@ -293,13 +294,23 @@ out2:
  * LSM hook implementation that allocs and transfers uctx spec to
  * xfrm_policy.
  */
-int selinux_xfrm_policy_alloc(struct xfrm_policy *xp, struct xfrm_user_sec_ctx *uctx)
+int selinux_xfrm_policy_alloc(struct xfrm_policy *xp,
+		struct xfrm_user_sec_ctx *uctx, struct sock *sk)
 {
 	int err;
+	u32 sid;
 
 	BUG_ON(!xp);
+	BUG_ON(uctx && sk);
+
+	if (sk) {
+		struct sk_security_struct *ssec = sk->sk_security;
+		sid = ssec->sid;
+	}
+	else
+		sid = SECSID_NULL;
 
-	err = selinux_xfrm_sec_ctx_alloc(&xp->security, uctx, NULL, 0);
+	err = selinux_xfrm_sec_ctx_alloc(&xp->security, uctx, NULL, sid);
 	return err;
 }
 
-- 
cgit v1.2.3


From 4237c75c0a35535d7f9f2bfeeb4b4df1e068a0bf Mon Sep 17 00:00:00 2001
From: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Date: Mon, 24 Jul 2006 23:32:50 -0700
Subject: [MLSXFRM]: Auto-labeling of child sockets

This automatically labels the TCP, Unix stream, and dccp child sockets
as well as openreqs to be at the same MLS level as the peer. This will
result in the selection of appropriately labeled IPSec Security
Associations.

This also uses the sock's sid (as opposed to the isec sid) in SELinux
enforcement of secmark in rcv_skb and postroute_last hooks.

Signed-off-by: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/security.h        |  55 ++++++++++++++++
 include/net/request_sock.h      |   1 +
 include/net/sock.h              |   1 +
 net/dccp/ipv4.c                 |   3 +
 net/dccp/ipv6.c                 |   7 +-
 net/ipv4/inet_connection_sock.c |   4 +-
 net/ipv4/syncookies.c           |   6 +-
 net/ipv4/tcp_ipv4.c             |   3 +
 net/ipv6/tcp_ipv6.c             |   6 +-
 security/dummy.c                |  24 +++++++
 security/selinux/hooks.c        | 137 +++++++++++++++++++++++++++-------------
 security/selinux/xfrm.c         |   1 -
 12 files changed, 197 insertions(+), 51 deletions(-)

(limited to 'include')

diff --git a/include/linux/security.h b/include/linux/security.h
index 8e3dc6c51a6d..bb4c80fdfe7a 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -90,6 +90,7 @@ extern int cap_netlink_recv(struct sk_buff *skb, int cap);
 struct nfsctl_arg;
 struct sched_param;
 struct swap_info_struct;
+struct request_sock;
 
 /* bprm_apply_creds unsafe reasons */
 #define LSM_UNSAFE_SHARE	1
@@ -819,6 +820,14 @@ struct swap_info_struct;
  * @sk_getsecid:
  *	Retrieve the LSM-specific secid for the sock to enable caching of network
  *	authorizations.
+ * @sock_graft:
+ *	Sets the socket's isec sid to the sock's sid.
+ * @inet_conn_request:
+ *	Sets the openreq's sid to socket's sid with MLS portion taken from peer sid.
+ * @inet_csk_clone:
+ *	Sets the new child socket's sid to the openreq sid.
+ * @req_classify_flow:
+ *	Sets the flow's sid to the openreq sid.
  *
  * Security hooks for XFRM operations.
  *
@@ -1358,6 +1367,11 @@ struct security_operations {
 	void (*sk_free_security) (struct sock *sk);
 	void (*sk_clone_security) (const struct sock *sk, struct sock *newsk);
 	void (*sk_getsecid) (struct sock *sk, u32 *secid);
+	void (*sock_graft)(struct sock* sk, struct socket *parent);
+	int (*inet_conn_request)(struct sock *sk, struct sk_buff *skb,
+					struct request_sock *req);
+	void (*inet_csk_clone)(struct sock *newsk, const struct request_sock *req);
+	void (*req_classify_flow)(const struct request_sock *req, struct flowi *fl);
 #endif	/* CONFIG_SECURITY_NETWORK */
 
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
@@ -2926,6 +2940,28 @@ static inline void security_sk_classify_flow(struct sock *sk, struct flowi *fl)
 {
 	security_ops->sk_getsecid(sk, &fl->secid);
 }
+
+static inline void security_req_classify_flow(const struct request_sock *req, struct flowi *fl)
+{
+	security_ops->req_classify_flow(req, fl);
+}
+
+static inline void security_sock_graft(struct sock* sk, struct socket *parent)
+{
+	security_ops->sock_graft(sk, parent);
+}
+
+static inline int security_inet_conn_request(struct sock *sk,
+			struct sk_buff *skb, struct request_sock *req)
+{
+	return security_ops->inet_conn_request(sk, skb, req);
+}
+
+static inline void security_inet_csk_clone(struct sock *newsk,
+			const struct request_sock *req)
+{
+	security_ops->inet_csk_clone(newsk, req);
+}
 #else	/* CONFIG_SECURITY_NETWORK */
 static inline int security_unix_stream_connect(struct socket * sock,
 					       struct socket * other, 
@@ -3055,6 +3091,25 @@ static inline void security_sk_clone(const struct sock *sk, struct sock *newsk)
 static inline void security_sk_classify_flow(struct sock *sk, struct flowi *fl)
 {
 }
+
+static inline void security_req_classify_flow(const struct request_sock *req, struct flowi *fl)
+{
+}
+
+static inline void security_sock_graft(struct sock* sk, struct socket *parent)
+{
+}
+
+static inline int security_inet_conn_request(struct sock *sk,
+			struct sk_buff *skb, struct request_sock *req)
+{
+	return 0;
+}
+
+static inline void security_inet_csk_clone(struct sock *newsk,
+			const struct request_sock *req)
+{
+}
 #endif	/* CONFIG_SECURITY_NETWORK */
 
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
diff --git a/include/net/request_sock.h b/include/net/request_sock.h
index c5d7f920c352..8e165ca16bd8 100644
--- a/include/net/request_sock.h
+++ b/include/net/request_sock.h
@@ -53,6 +53,7 @@ struct request_sock {
 	unsigned long			expires;
 	struct request_sock_ops		*rsk_ops;
 	struct sock			*sk;
+	u32				secid;
 };
 
 static inline struct request_sock *reqsk_alloc(struct request_sock_ops *ops)
diff --git a/include/net/sock.h b/include/net/sock.h
index 91cdceb3c028..337ebec84c70 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -969,6 +969,7 @@ static inline void sock_graft(struct sock *sk, struct socket *parent)
 	sk->sk_sleep = &parent->wait;
 	parent->sk = sk;
 	sk->sk_socket = parent;
+	security_sock_graft(sk, parent);
 	write_unlock_bh(&sk->sk_callback_lock);
 }
 
diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c
index 386498053b1c..171d363876ee 100644
--- a/net/dccp/ipv4.c
+++ b/net/dccp/ipv4.c
@@ -501,6 +501,9 @@ int dccp_v4_conn_request(struct sock *sk, struct sk_buff *skb)
 
 	dccp_openreq_init(req, &dp, skb);
 
+	if (security_inet_conn_request(sk, skb, req))
+		goto drop_and_free;
+
 	ireq = inet_rsk(req);
 	ireq->loc_addr = daddr;
 	ireq->rmt_addr = saddr;
diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c
index 53d255c01431..231bc7c7e749 100644
--- a/net/dccp/ipv6.c
+++ b/net/dccp/ipv6.c
@@ -424,7 +424,7 @@ static int dccp_v6_send_response(struct sock *sk, struct request_sock *req,
 	fl.oif = ireq6->iif;
 	fl.fl_ip_dport = inet_rsk(req)->rmt_port;
 	fl.fl_ip_sport = inet_sk(sk)->sport;
-	security_sk_classify_flow(sk, &fl);
+	security_req_classify_flow(req, &fl);
 
 	if (dst == NULL) {
 		opt = np->opt;
@@ -626,7 +626,7 @@ static void dccp_v6_reqsk_send_ack(struct sk_buff *rxskb,
 	fl.oif = inet6_iif(rxskb);
 	fl.fl_ip_dport = dh->dccph_dport;
 	fl.fl_ip_sport = dh->dccph_sport;
-	security_skb_classify_flow(rxskb, &fl);
+	security_req_classify_flow(req, &fl);
 
 	if (!ip6_dst_lookup(NULL, &skb->dst, &fl)) {
 		if (xfrm_lookup(&skb->dst, &fl, NULL, 0) >= 0) {
@@ -709,6 +709,9 @@ static int dccp_v6_conn_request(struct sock *sk, struct sk_buff *skb)
 
 	dccp_openreq_init(req, &dp, skb);
 
+	if (security_inet_conn_request(sk, skb, req))
+		goto drop_and_free;
+
 	ireq6 = inet6_rsk(req);
 	ireq = inet_rsk(req);
 	ipv6_addr_copy(&ireq6->rmt_addr, &skb->nh.ipv6h->saddr);
diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c
index 772b4eac78bc..07204391d083 100644
--- a/net/ipv4/inet_connection_sock.c
+++ b/net/ipv4/inet_connection_sock.c
@@ -327,7 +327,7 @@ struct dst_entry* inet_csk_route_req(struct sock *sk,
 				       { .sport = inet_sk(sk)->sport,
 					 .dport = ireq->rmt_port } } };
 
-	security_sk_classify_flow(sk, &fl);
+	security_req_classify_flow(req, &fl);
 	if (ip_route_output_flow(&rt, &fl, sk, 0)) {
 		IP_INC_STATS_BH(IPSTATS_MIB_OUTNOROUTES);
 		return NULL;
@@ -510,6 +510,8 @@ struct sock *inet_csk_clone(struct sock *sk, const struct request_sock *req,
 
 		/* Deinitialize accept_queue to trap illegal accesses. */
 		memset(&newicsk->icsk_accept_queue, 0, sizeof(newicsk->icsk_accept_queue));
+
+		security_inet_csk_clone(newsk, req);
 	}
 	return newsk;
 }
diff --git a/net/ipv4/syncookies.c b/net/ipv4/syncookies.c
index 307dc3c0d635..661e0a4bca72 100644
--- a/net/ipv4/syncookies.c
+++ b/net/ipv4/syncookies.c
@@ -214,6 +214,10 @@ struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb,
 	if (!req)
 		goto out;
 
+	if (security_inet_conn_request(sk, skb, req)) {
+		reqsk_free(req);
+		goto out;
+	}
 	ireq = inet_rsk(req);
 	treq = tcp_rsk(req);
 	treq->rcv_isn		= htonl(skb->h.th->seq) - 1;
@@ -259,7 +263,7 @@ struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb,
 				    .uli_u = { .ports =
 					       { .sport = skb->h.th->dest,
 						 .dport = skb->h.th->source } } };
-		security_sk_classify_flow(sk, &fl);
+		security_req_classify_flow(req, &fl);
 		if (ip_route_output_key(&rt, &fl)) {
 			reqsk_free(req);
 			goto out; 
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 4b04c3edd4a9..43f6740244f8 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -798,6 +798,9 @@ int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb)
 
 	tcp_openreq_init(req, &tmp_opt, skb);
 
+	if (security_inet_conn_request(sk, skb, req))
+		goto drop_and_free;
+
 	ireq = inet_rsk(req);
 	ireq->loc_addr = daddr;
 	ireq->rmt_addr = saddr;
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 46922e57e311..302786a11cd6 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -470,7 +470,7 @@ static int tcp_v6_send_synack(struct sock *sk, struct request_sock *req,
 	fl.oif = treq->iif;
 	fl.fl_ip_dport = inet_rsk(req)->rmt_port;
 	fl.fl_ip_sport = inet_sk(sk)->sport;
-	security_sk_classify_flow(sk, &fl);
+	security_req_classify_flow(req, &fl);
 
 	if (dst == NULL) {
 		opt = np->opt;
@@ -826,6 +826,8 @@ static int tcp_v6_conn_request(struct sock *sk, struct sk_buff *skb)
 
 	tcp_rsk(req)->snt_isn = isn;
 
+	security_inet_conn_request(sk, skb, req);
+
 	if (tcp_v6_send_synack(sk, req, NULL))
 		goto drop;
 
@@ -929,7 +931,7 @@ static struct sock * tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb,
 		fl.oif = sk->sk_bound_dev_if;
 		fl.fl_ip_dport = inet_rsk(req)->rmt_port;
 		fl.fl_ip_sport = inet_sk(sk)->sport;
-		security_sk_classify_flow(sk, &fl);
+		security_req_classify_flow(req, &fl);
 
 		if (ip6_dst_lookup(sk, &dst, &fl))
 			goto out;
diff --git a/security/dummy.c b/security/dummy.c
index 66cc06404930..1c45f8e4aad1 100644
--- a/security/dummy.c
+++ b/security/dummy.c
@@ -812,6 +812,26 @@ static inline void dummy_sk_clone_security (const struct sock *sk, struct sock *
 static inline void dummy_sk_getsecid(struct sock *sk, u32 *secid)
 {
 }
+
+static inline void dummy_sock_graft(struct sock* sk, struct socket *parent)
+{
+}
+
+static inline int dummy_inet_conn_request(struct sock *sk,
+			struct sk_buff *skb, struct request_sock *req)
+{
+	return 0;
+}
+
+static inline void dummy_inet_csk_clone(struct sock *newsk,
+			const struct request_sock *req)
+{
+}
+
+static inline void dummy_req_classify_flow(const struct request_sock *req,
+			struct flowi *fl)
+{
+}
 #endif	/* CONFIG_SECURITY_NETWORK */
 
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
@@ -1084,6 +1104,10 @@ void security_fixup_ops (struct security_operations *ops)
 	set_to_dummy_if_null(ops, sk_free_security);
 	set_to_dummy_if_null(ops, sk_clone_security);
 	set_to_dummy_if_null(ops, sk_getsecid);
+	set_to_dummy_if_null(ops, sock_graft);
+	set_to_dummy_if_null(ops, inet_conn_request);
+	set_to_dummy_if_null(ops, inet_csk_clone);
+	set_to_dummy_if_null(ops, req_classify_flow);
  #endif	/* CONFIG_SECURITY_NETWORK */
 #ifdef  CONFIG_SECURITY_NETWORK_XFRM
 	set_to_dummy_if_null(ops, xfrm_policy_alloc_security);
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index 4e5989d584ce..1dc935f7b919 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -3328,8 +3328,9 @@ static int selinux_socket_unix_stream_connect(struct socket *sock,
 	/* server child socket */
 	ssec = newsk->sk_security;
 	ssec->peer_sid = isec->sid;
-	
-	return 0;
+	err = security_sid_mls_copy(other_isec->sid, ssec->peer_sid, &ssec->sid);
+
+	return err;
 }
 
 static int selinux_socket_unix_may_send(struct socket *sock,
@@ -3355,11 +3356,29 @@ static int selinux_socket_unix_may_send(struct socket *sock,
 }
 
 static int selinux_sock_rcv_skb_compat(struct sock *sk, struct sk_buff *skb,
-		struct avc_audit_data *ad, u32 sock_sid, u16 sock_class,
-		u16 family, char *addrp, int len)
+		struct avc_audit_data *ad, u16 family, char *addrp, int len)
 {
 	int err = 0;
 	u32 netif_perm, node_perm, node_sid, if_sid, recv_perm = 0;
+	struct socket *sock;
+	u16 sock_class = 0;
+	u32 sock_sid = 0;
+
+ 	read_lock_bh(&sk->sk_callback_lock);
+ 	sock = sk->sk_socket;
+ 	if (sock) {
+ 		struct inode *inode;
+ 		inode = SOCK_INODE(sock);
+ 		if (inode) {
+ 			struct inode_security_struct *isec;
+ 			isec = inode->i_security;
+ 			sock_sid = isec->sid;
+ 			sock_class = isec->sclass;
+ 		}
+ 	}
+ 	read_unlock_bh(&sk->sk_callback_lock);
+ 	if (!sock_sid)
+  		goto out;
 
 	if (!skb->dev)
 		goto out;
@@ -3419,12 +3438,10 @@ out:
 static int selinux_socket_sock_rcv_skb(struct sock *sk, struct sk_buff *skb)
 {
 	u16 family;
-	u16 sock_class = 0;
 	char *addrp;
 	int len, err = 0;
-	u32 sock_sid = 0;
-	struct socket *sock;
 	struct avc_audit_data ad;
+	struct sk_security_struct *sksec = sk->sk_security;
 
 	family = sk->sk_family;
 	if (family != PF_INET && family != PF_INET6)
@@ -3434,22 +3451,6 @@ static int selinux_socket_sock_rcv_skb(struct sock *sk, struct sk_buff *skb)
 	if (family == PF_INET6 && skb->protocol == ntohs(ETH_P_IP))
 		family = PF_INET;
 
- 	read_lock_bh(&sk->sk_callback_lock);
- 	sock = sk->sk_socket;
- 	if (sock) {
- 		struct inode *inode;
- 		inode = SOCK_INODE(sock);
- 		if (inode) {
- 			struct inode_security_struct *isec;
- 			isec = inode->i_security;
- 			sock_sid = isec->sid;
- 			sock_class = isec->sclass;
- 		}
- 	}
- 	read_unlock_bh(&sk->sk_callback_lock);
- 	if (!sock_sid)
-  		goto out;
-
 	AVC_AUDIT_DATA_INIT(&ad, NET);
 	ad.u.net.netif = skb->dev ? skb->dev->name : "[unknown]";
 	ad.u.net.family = family;
@@ -3459,16 +3460,15 @@ static int selinux_socket_sock_rcv_skb(struct sock *sk, struct sk_buff *skb)
 		goto out;
 
 	if (selinux_compat_net)
-		err = selinux_sock_rcv_skb_compat(sk, skb, &ad, sock_sid,
-						  sock_class, family,
+		err = selinux_sock_rcv_skb_compat(sk, skb, &ad, family,
 						  addrp, len);
 	else
-		err = avc_has_perm(sock_sid, skb->secmark, SECCLASS_PACKET,
+		err = avc_has_perm(sksec->sid, skb->secmark, SECCLASS_PACKET,
 				   PACKET__RECV, &ad);
 	if (err)
 		goto out;
 
-	err = selinux_xfrm_sock_rcv_skb(sock_sid, skb, &ad);
+	err = selinux_xfrm_sock_rcv_skb(sksec->sid, skb, &ad);
 out:	
 	return err;
 }
@@ -3572,6 +3572,49 @@ static void selinux_sk_getsecid(struct sock *sk, u32 *secid)
 	}
 }
 
+void selinux_sock_graft(struct sock* sk, struct socket *parent)
+{
+	struct inode_security_struct *isec = SOCK_INODE(parent)->i_security;
+	struct sk_security_struct *sksec = sk->sk_security;
+
+	isec->sid = sksec->sid;
+}
+
+int selinux_inet_conn_request(struct sock *sk, struct sk_buff *skb,
+					   struct request_sock *req)
+{
+	struct sk_security_struct *sksec = sk->sk_security;
+	int err;
+	u32 newsid = 0;
+	u32 peersid;
+
+	err = selinux_xfrm_decode_session(skb, &peersid, 0);
+	BUG_ON(err);
+
+	err = security_sid_mls_copy(sksec->sid, peersid, &newsid);
+	if (err)
+		return err;
+
+	req->secid = newsid;
+	return 0;
+}
+
+void selinux_inet_csk_clone(struct sock *newsk, const struct request_sock *req)
+{
+	struct sk_security_struct *newsksec = newsk->sk_security;
+
+	newsksec->sid = req->secid;
+	/* NOTE: Ideally, we should also get the isec->sid for the
+	   new socket in sync, but we don't have the isec available yet.
+	   So we will wait until sock_graft to do it, by which
+	   time it will have been created and available. */
+}
+
+void selinux_req_classify_flow(const struct request_sock *req, struct flowi *fl)
+{
+	fl->secid = req->secid;
+}
+
 static int selinux_nlmsg_perm(struct sock *sk, struct sk_buff *skb)
 {
 	int err = 0;
@@ -3611,12 +3654,24 @@ out:
 #ifdef CONFIG_NETFILTER
 
 static int selinux_ip_postroute_last_compat(struct sock *sk, struct net_device *dev,
-					    struct inode_security_struct *isec,
 					    struct avc_audit_data *ad,
 					    u16 family, char *addrp, int len)
 {
-	int err;
+	int err = 0;
 	u32 netif_perm, node_perm, node_sid, if_sid, send_perm = 0;
+	struct socket *sock;
+	struct inode *inode;
+	struct inode_security_struct *isec;
+
+	sock = sk->sk_socket;
+	if (!sock)
+		goto out;
+
+	inode = SOCK_INODE(sock);
+	if (!inode)
+		goto out;
+
+	isec = inode->i_security;
 	
 	err = sel_netif_sids(dev, &if_sid, NULL);
 	if (err)
@@ -3681,26 +3736,16 @@ static unsigned int selinux_ip_postroute_last(unsigned int hooknum,
 	char *addrp;
 	int len, err = 0;
 	struct sock *sk;
-	struct socket *sock;
-	struct inode *inode;
 	struct sk_buff *skb = *pskb;
-	struct inode_security_struct *isec;
 	struct avc_audit_data ad;
 	struct net_device *dev = (struct net_device *)out;
+	struct sk_security_struct *sksec;
 
 	sk = skb->sk;
 	if (!sk)
 		goto out;
 
-	sock = sk->sk_socket;
-	if (!sock)
-		goto out;
-
-	inode = SOCK_INODE(sock);
-	if (!inode)
-		goto out;
-
-	isec = inode->i_security;
+	sksec = sk->sk_security;
 
 	AVC_AUDIT_DATA_INIT(&ad, NET);
 	ad.u.net.netif = dev->name;
@@ -3711,16 +3756,16 @@ static unsigned int selinux_ip_postroute_last(unsigned int hooknum,
 		goto out;
 
 	if (selinux_compat_net)
-		err = selinux_ip_postroute_last_compat(sk, dev, isec, &ad,
+		err = selinux_ip_postroute_last_compat(sk, dev, &ad,
 						       family, addrp, len);
 	else
-		err = avc_has_perm(isec->sid, skb->secmark, SECCLASS_PACKET,
+		err = avc_has_perm(sksec->sid, skb->secmark, SECCLASS_PACKET,
 				   PACKET__SEND, &ad);
 
 	if (err)
 		goto out;
 
-	err = selinux_xfrm_postroute_last(isec->sid, skb, &ad);
+	err = selinux_xfrm_postroute_last(sksec->sid, skb, &ad);
 out:
 	return err ? NF_DROP : NF_ACCEPT;
 }
@@ -4623,6 +4668,10 @@ static struct security_operations selinux_ops = {
 	.sk_free_security =		selinux_sk_free_security,
 	.sk_clone_security =		selinux_sk_clone_security,
 	.sk_getsecid = 			selinux_sk_getsecid,
+	.sock_graft =			selinux_sock_graft,
+	.inet_conn_request =		selinux_inet_conn_request,
+	.inet_csk_clone =		selinux_inet_csk_clone,
+	.req_classify_flow =		selinux_req_classify_flow,
 
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
 	.xfrm_policy_alloc_security =	selinux_xfrm_policy_alloc,
diff --git a/security/selinux/xfrm.c b/security/selinux/xfrm.c
index d3690f985135..3e742b850af6 100644
--- a/security/selinux/xfrm.c
+++ b/security/selinux/xfrm.c
@@ -271,7 +271,6 @@ not_from_user:
 		goto out;
 	}
 
-
 	ctx->ctx_doi = XFRM_SC_DOI_LSM;
 	ctx->ctx_alg = XFRM_SC_ALG_SELINUX;
 	ctx->ctx_sid = ctx_sid;
-- 
cgit v1.2.3


From 11a03f78fbf15a866ba3bf6359a75cdfd1ced703 Mon Sep 17 00:00:00 2001
From: Paul Moore <paul.moore@hp.com>
Date: Thu, 3 Aug 2006 16:46:20 -0700
Subject: [NetLabel]: core network changes

Changes to the core network stack to support the NetLabel subsystem.  This
includes changes to the IPv4 option handling to support CIPSO labels.

Signed-off-by: Paul Moore <paul.moore@hp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/ip.h       |   1 +
 include/net/cipso_ipv4.h | 250 ++++++++++++++++++++++++++++++++++++++++
 include/net/inet_sock.h  |   2 +-
 include/net/netlabel.h   | 291 +++++++++++++++++++++++++++++++++++++++++++++++
 net/ipv4/ah4.c           |   2 +-
 net/ipv4/ip_options.c    |  19 ++++
 6 files changed, 563 insertions(+), 2 deletions(-)
 create mode 100644 include/net/cipso_ipv4.h
 create mode 100644 include/net/netlabel.h

(limited to 'include')

diff --git a/include/linux/ip.h b/include/linux/ip.h
index 4b55cf1df732..2f4600146f83 100644
--- a/include/linux/ip.h
+++ b/include/linux/ip.h
@@ -57,6 +57,7 @@
 #define IPOPT_SEC	(2 |IPOPT_CONTROL|IPOPT_COPY)
 #define IPOPT_LSRR	(3 |IPOPT_CONTROL|IPOPT_COPY)
 #define IPOPT_TIMESTAMP	(4 |IPOPT_MEASUREMENT)
+#define IPOPT_CIPSO	(6 |IPOPT_CONTROL|IPOPT_COPY)
 #define IPOPT_RR	(7 |IPOPT_CONTROL)
 #define IPOPT_SID	(8 |IPOPT_CONTROL|IPOPT_COPY)
 #define IPOPT_SSRR	(9 |IPOPT_CONTROL|IPOPT_COPY)
diff --git a/include/net/cipso_ipv4.h b/include/net/cipso_ipv4.h
new file mode 100644
index 000000000000..c7175e725804
--- /dev/null
+++ b/include/net/cipso_ipv4.h
@@ -0,0 +1,250 @@
+/*
+ * CIPSO - Commercial IP Security Option
+ *
+ * This is an implementation of the CIPSO 2.2 protocol as specified in
+ * draft-ietf-cipso-ipsecurity-01.txt with additional tag types as found in
+ * FIPS-188, copies of both documents can be found in the Documentation
+ * directory.  While CIPSO never became a full IETF RFC standard many vendors
+ * have chosen to adopt the protocol and over the years it has become a
+ * de-facto standard for labeled networking.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software;  you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY;  without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program;  if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _CIPSO_IPV4_H
+#define _CIPSO_IPV4_H
+
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/list.h>
+#include <net/netlabel.h>
+
+/* known doi values */
+#define CIPSO_V4_DOI_UNKNOWN          0x00000000
+
+/* tag types */
+#define CIPSO_V4_TAG_INVALID          0
+#define CIPSO_V4_TAG_RBITMAP          1
+#define CIPSO_V4_TAG_ENUM             2
+#define CIPSO_V4_TAG_RANGE            5
+#define CIPSO_V4_TAG_PBITMAP          6
+#define CIPSO_V4_TAG_FREEFORM         7
+
+/* doi mapping types */
+#define CIPSO_V4_MAP_UNKNOWN          0
+#define CIPSO_V4_MAP_STD              1
+#define CIPSO_V4_MAP_PASS             2
+
+/* limits */
+#define CIPSO_V4_MAX_REM_LVLS         256
+#define CIPSO_V4_INV_LVL              0x80000000
+#define CIPSO_V4_MAX_LOC_LVLS         (CIPSO_V4_INV_LVL - 1)
+#define CIPSO_V4_MAX_REM_CATS         65536
+#define CIPSO_V4_INV_CAT              0x80000000
+#define CIPSO_V4_MAX_LOC_CATS         (CIPSO_V4_INV_CAT - 1)
+
+/*
+ * CIPSO DOI definitions
+ */
+
+/* DOI definition struct */
+#define CIPSO_V4_TAG_MAXCNT           5
+struct cipso_v4_doi {
+	u32 doi;
+	u32 type;
+	union {
+		struct cipso_v4_std_map_tbl *std;
+	} map;
+	u8 tags[CIPSO_V4_TAG_MAXCNT];
+
+	u32 valid;
+	struct list_head list;
+	struct rcu_head rcu;
+	struct list_head dom_list;
+};
+
+/* Standard CIPSO mapping table */
+/* NOTE: the highest order bit (i.e. 0x80000000) is an 'invalid' flag, if the
+ *       bit is set then consider that value as unspecified, meaning the
+ *       mapping for that particular level/category is invalid */
+struct cipso_v4_std_map_tbl {
+	struct {
+		u32 *cipso;
+		u32 *local;
+		u32 cipso_size;
+		u32 local_size;
+	} lvl;
+	struct {
+		u32 *cipso;
+		u32 *local;
+		u32 cipso_size;
+		u32 local_size;
+	} cat;
+};
+
+/*
+ * Sysctl Variables
+ */
+
+#ifdef CONFIG_NETLABEL
+extern int cipso_v4_cache_enabled;
+extern int cipso_v4_cache_bucketsize;
+extern int cipso_v4_rbm_optfmt;
+extern int cipso_v4_rbm_strictvalid;
+#endif
+
+/*
+ * Helper Functions
+ */
+
+#define CIPSO_V4_OPTEXIST(x) (IPCB(x)->opt.cipso != 0)
+#define CIPSO_V4_OPTPTR(x) ((x)->nh.raw + IPCB(x)->opt.cipso)
+
+/*
+ * DOI List Functions
+ */
+
+#ifdef CONFIG_NETLABEL
+int cipso_v4_doi_add(struct cipso_v4_doi *doi_def);
+int cipso_v4_doi_remove(u32 doi, void (*callback) (struct rcu_head * head));
+struct cipso_v4_doi *cipso_v4_doi_getdef(u32 doi);
+struct sk_buff *cipso_v4_doi_dump_all(size_t headroom);
+struct sk_buff *cipso_v4_doi_dump(u32 doi, size_t headroom);
+int cipso_v4_doi_domhsh_add(struct cipso_v4_doi *doi_def, const char *domain);
+int cipso_v4_doi_domhsh_remove(struct cipso_v4_doi *doi_def,
+			       const char *domain);
+#else
+static inline int cipso_v4_doi_add(struct cipso_v4_doi *doi_def)
+{
+	return -ENOSYS;
+}
+
+static inline int cipso_v4_doi_remove(u32 doi,
+				    void (*callback) (struct rcu_head * head))
+{
+	return 0;
+}
+
+static inline struct cipso_v4_doi *cipso_v4_doi_getdef(u32 doi)
+{
+	return NULL;
+}
+
+static inline struct sk_buff *cipso_v4_doi_dump_all(size_t headroom)
+{
+	return NULL;
+}
+
+static inline struct sk_buff *cipso_v4_doi_dump(u32 doi, size_t headroom)
+{
+	return NULL;
+}
+
+static inline int cipso_v4_doi_domhsh_add(struct cipso_v4_doi *doi_def,
+					  const char *domain)
+{
+	return -ENOSYS;
+}
+
+static inline int cipso_v4_doi_domhsh_remove(struct cipso_v4_doi *doi_def,
+					     const char *domain)
+{
+	return 0;
+}
+#endif /* CONFIG_NETLABEL */
+
+/*
+ * Label Mapping Cache Functions
+ */
+
+#ifdef CONFIG_NETLABEL
+void cipso_v4_cache_invalidate(void);
+int cipso_v4_cache_add(const struct sk_buff *skb,
+		       const struct netlbl_lsm_secattr *secattr);
+#else
+static inline void cipso_v4_cache_invalidate(void)
+{
+	return;
+}
+
+static inline int cipso_v4_cache_add(const struct sk_buff *skb,
+				     const struct netlbl_lsm_secattr *secattr)
+{
+	return 0;
+}
+#endif /* CONFIG_NETLABEL */
+
+/*
+ * Protocol Handling Functions
+ */
+
+#ifdef CONFIG_NETLABEL
+void cipso_v4_error(struct sk_buff *skb, int error, u32 gateway);
+int cipso_v4_socket_setopt(struct socket *sock,
+			   unsigned char *opt,
+			   u32 opt_len);
+int cipso_v4_socket_setattr(const struct socket *sock,
+			    const struct cipso_v4_doi *doi_def,
+			    const struct netlbl_lsm_secattr *secattr);
+int cipso_v4_socket_getopt(const struct socket *sock,
+			   unsigned char **opt,
+			   u32 *opt_len);
+int cipso_v4_socket_getattr(const struct socket *sock,
+			    struct netlbl_lsm_secattr *secattr);
+int cipso_v4_skbuff_getattr(const struct sk_buff *skb,
+			    struct netlbl_lsm_secattr *secattr);
+int cipso_v4_validate(unsigned char **option);
+#else
+static inline void cipso_v4_error(struct sk_buff *skb,
+				  int error,
+				  u32 gateway)
+{
+	return;
+}
+
+static inline int cipso_v4_socket_setattr(const struct socket *sock,
+				  const struct cipso_v4_doi *doi_def,
+				  const struct netlbl_lsm_secattr *secattr)
+{
+	return -ENOSYS;
+}
+
+static inline int cipso_v4_socket_getattr(const struct socket *sock,
+					  struct netlbl_lsm_secattr *secattr)
+{
+	return -ENOSYS;
+}
+
+static inline int cipso_v4_skbuff_getattr(const struct sk_buff *skb,
+					  struct netlbl_lsm_secattr *secattr)
+{
+	return -ENOSYS;
+}
+
+static inline int cipso_v4_validate(unsigned char **option)
+{
+	return -ENOSYS;
+}
+#endif /* CONFIG_NETLABEL */
+
+#endif /* _CIPSO_IPV4_H */
diff --git a/include/net/inet_sock.h b/include/net/inet_sock.h
index 1f4a9a60d4cc..f4caad56cd03 100644
--- a/include/net/inet_sock.h
+++ b/include/net/inet_sock.h
@@ -51,7 +51,7 @@ struct ip_options {
 			ts_needtime:1,
 			ts_needaddr:1;
 	unsigned char	router_alert;
-	unsigned char	__pad1;
+	unsigned char	cipso;
 	unsigned char	__pad2;
 	unsigned char	__data[0];
 };
diff --git a/include/net/netlabel.h b/include/net/netlabel.h
new file mode 100644
index 000000000000..7cae730832c7
--- /dev/null
+++ b/include/net/netlabel.h
@@ -0,0 +1,291 @@
+/*
+ * NetLabel System
+ *
+ * The NetLabel system manages static and dynamic label mappings for network
+ * protocols such as CIPSO and RIPSO.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software;  you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY;  without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program;  if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _NETLABEL_H
+#define _NETLABEL_H
+
+#include <linux/types.h>
+#include <linux/skbuff.h>
+#include <net/netlink.h>
+
+/*
+ * NetLabel - A management interface for maintaining network packet label
+ *            mapping tables for explicit packet labling protocols.
+ *
+ * Network protocols such as CIPSO and RIPSO require a label translation layer
+ * to convert the label on the packet into something meaningful on the host
+ * machine.  In the current Linux implementation these mapping tables live
+ * inside the kernel; NetLabel provides a mechanism for user space applications
+ * to manage these mapping tables.
+ *
+ * NetLabel makes use of the Generic NETLINK mechanism as a transport layer to
+ * send messages between kernel and user space.  The general format of a
+ * NetLabel message is shown below:
+ *
+ *  +-----------------+-------------------+--------- --- -- -
+ *  | struct nlmsghdr | struct genlmsghdr | payload
+ *  +-----------------+-------------------+--------- --- -- -
+ *
+ * The 'nlmsghdr' and 'genlmsghdr' structs should be dealt with like normal.
+ * The payload is dependent on the subsystem specified in the
+ * 'nlmsghdr->nlmsg_type' and should be defined below, supporting functions
+ * should be defined in the corresponding net/netlabel/netlabel_<subsys>.h|c
+ * file.  All of the fields in the NetLabel payload are NETLINK attributes, the
+ * length of each field is the length of the NETLINK attribute payload, see
+ * include/net/netlink.h for more information on NETLINK attributes.
+ *
+ */
+
+/*
+ * NetLabel NETLINK protocol
+ */
+
+#define NETLBL_PROTO_VERSION            1
+
+/* NetLabel NETLINK types/families */
+#define NETLBL_NLTYPE_NONE              0
+#define NETLBL_NLTYPE_MGMT              1
+#define NETLBL_NLTYPE_MGMT_NAME         "NLBL_MGMT"
+#define NETLBL_NLTYPE_RIPSO             2
+#define NETLBL_NLTYPE_RIPSO_NAME        "NLBL_RIPSO"
+#define NETLBL_NLTYPE_CIPSOV4           3
+#define NETLBL_NLTYPE_CIPSOV4_NAME      "NLBL_CIPSOv4"
+#define NETLBL_NLTYPE_CIPSOV6           4
+#define NETLBL_NLTYPE_CIPSOV6_NAME      "NLBL_CIPSOv6"
+#define NETLBL_NLTYPE_UNLABELED         5
+#define NETLBL_NLTYPE_UNLABELED_NAME    "NLBL_UNLBL"
+
+/* NetLabel return codes */
+#define NETLBL_E_OK                     0
+
+/*
+ * Helper functions
+ */
+
+#define NETLBL_LEN_U8                   nla_total_size(sizeof(u8))
+#define NETLBL_LEN_U16                  nla_total_size(sizeof(u16))
+#define NETLBL_LEN_U32                  nla_total_size(sizeof(u32))
+
+/**
+ * netlbl_netlink_alloc_skb - Allocate a NETLINK message buffer
+ * @head: the amount of headroom in bytes
+ * @body: the desired size (minus headroom) in bytes
+ * @gfp_flags: the alloc flags to pass to alloc_skb()
+ *
+ * Description:
+ * Allocate a NETLINK message buffer based on the sizes given in @head and
+ * @body.  If @head is greater than zero skb_reserve() is called to reserve
+ * @head bytes at the start of the buffer.  Returns a valid sk_buff pointer on
+ * success, NULL on failure.
+ *
+ */
+static inline struct sk_buff *netlbl_netlink_alloc_skb(size_t head,
+						       size_t body,
+						       int gfp_flags)
+{
+	struct sk_buff *skb;
+
+	skb = alloc_skb(NLMSG_ALIGN(head + body), gfp_flags);
+	if (skb == NULL)
+		return NULL;
+	if (head > 0) {
+		skb_reserve(skb, head);
+		if (skb_tailroom(skb) < body) {
+			kfree_skb(skb);
+			return NULL;
+		}
+	}
+
+	return skb;
+}
+
+/*
+ * NetLabel - Kernel API for accessing the network packet label mappings.
+ *
+ * The following functions are provided for use by other kernel modules,
+ * specifically kernel LSM modules, to provide a consistent, transparent API
+ * for dealing with explicit packet labeling protocols such as CIPSO and
+ * RIPSO.  The functions defined here are implemented in the
+ * net/netlabel/netlabel_kapi.c file.
+ *
+ */
+
+/* Domain mapping definition struct */
+struct netlbl_dom_map;
+
+/* Domain mapping operations */
+int netlbl_domhsh_remove(const char *domain);
+
+/* LSM security attributes */
+struct netlbl_lsm_cache {
+	void (*free) (const void *data);
+	void *data;
+};
+struct netlbl_lsm_secattr {
+	char *domain;
+
+	u32 mls_lvl;
+	u32 mls_lvl_vld;
+	unsigned char *mls_cat;
+	size_t mls_cat_len;
+
+	struct netlbl_lsm_cache cache;
+};
+
+/*
+ * LSM security attribute operations
+ */
+
+
+/**
+ * netlbl_secattr_init - Initialize a netlbl_lsm_secattr struct
+ * @secattr: the struct to initialize
+ *
+ * Description:
+ * Initialize an already allocated netlbl_lsm_secattr struct.  Returns zero on
+ * success, negative values on error.
+ *
+ */
+static inline int netlbl_secattr_init(struct netlbl_lsm_secattr *secattr)
+{
+	memset(secattr, 0, sizeof(*secattr));
+	return 0;
+}
+
+/**
+ * netlbl_secattr_destroy - Clears a netlbl_lsm_secattr struct
+ * @secattr: the struct to clear
+ * @clear_cache: cache clear flag
+ *
+ * Description:
+ * Destroys the @secattr struct, including freeing all of the internal buffers.
+ * If @clear_cache is true then free the cache fields, otherwise leave them
+ * intact.  The struct must be reset with a call to netlbl_secattr_init()
+ * before reuse.
+ *
+ */
+static inline void netlbl_secattr_destroy(struct netlbl_lsm_secattr *secattr,
+					  u32 clear_cache)
+{
+	if (clear_cache && secattr->cache.data != NULL && secattr->cache.free)
+		secattr->cache.free(secattr->cache.data);
+	kfree(secattr->domain);
+	kfree(secattr->mls_cat);
+}
+
+/**
+ * netlbl_secattr_alloc - Allocate and initialize a netlbl_lsm_secattr struct
+ * @flags: the memory allocation flags
+ *
+ * Description:
+ * Allocate and initialize a netlbl_lsm_secattr struct.  Returns a valid
+ * pointer on success, or NULL on failure.
+ *
+ */
+static inline struct netlbl_lsm_secattr *netlbl_secattr_alloc(int flags)
+{
+	return kzalloc(sizeof(struct netlbl_lsm_secattr), flags);
+}
+
+/**
+ * netlbl_secattr_free - Frees a netlbl_lsm_secattr struct
+ * @secattr: the struct to free
+ * @clear_cache: cache clear flag
+ *
+ * Description:
+ * Frees @secattr including all of the internal buffers.  If @clear_cache is
+ * true then free the cache fields, otherwise leave them intact.
+ *
+ */
+static inline void netlbl_secattr_free(struct netlbl_lsm_secattr *secattr,
+				       u32 clear_cache)
+{
+	netlbl_secattr_destroy(secattr, clear_cache);
+	kfree(secattr);
+}
+
+/*
+ * LSM protocol operations
+ */
+
+#ifdef CONFIG_NETLABEL
+int netlbl_socket_setattr(const struct socket *sock,
+			  const struct netlbl_lsm_secattr *secattr);
+int netlbl_socket_getattr(const struct socket *sock,
+			  struct netlbl_lsm_secattr *secattr);
+int netlbl_skbuff_getattr(const struct sk_buff *skb,
+			  struct netlbl_lsm_secattr *secattr);
+void netlbl_skbuff_err(struct sk_buff *skb, int error);
+#else
+static inline int netlbl_socket_setattr(const struct socket *sock,
+				     const struct netlbl_lsm_secattr *secattr)
+{
+	return -ENOSYS;
+}
+
+static inline int netlbl_socket_getattr(const struct socket *sock,
+					struct netlbl_lsm_secattr *secattr)
+{
+	return -ENOSYS;
+}
+
+static inline int netlbl_skbuff_getattr(const struct sk_buff *skb,
+					struct netlbl_lsm_secattr *secattr)
+{
+	return -ENOSYS;
+}
+
+static inline void netlbl_skbuff_err(struct sk_buff *skb, int error)
+{
+	return;
+}
+#endif /* CONFIG_NETLABEL */
+
+/*
+ * LSM label mapping cache operations
+ */
+
+#ifdef CONFIG_NETLABEL
+void netlbl_cache_invalidate(void);
+int netlbl_cache_add(const struct sk_buff *skb,
+		     const struct netlbl_lsm_secattr *secattr);
+#else
+static inline void netlbl_cache_invalidate(void)
+{
+	return;
+}
+
+static inline int netlbl_cache_add(const struct sk_buff *skb,
+				   const struct netlbl_lsm_secattr *secattr)
+{
+	return 0;
+}
+#endif /* CONFIG_NETLABEL */
+
+#endif /* _NETLABEL_H */
diff --git a/net/ipv4/ah4.c b/net/ipv4/ah4.c
index 2b98943e6b02..008e69d2e423 100644
--- a/net/ipv4/ah4.c
+++ b/net/ipv4/ah4.c
@@ -35,7 +35,7 @@ static int ip_clear_mutable_options(struct iphdr *iph, u32 *daddr)
 		switch (*optptr) {
 		case IPOPT_SEC:
 		case 0x85:	/* Some "Extended Security" crap. */
-		case 0x86:	/* Another "Commercial Security" crap. */
+		case IPOPT_CIPSO:
 		case IPOPT_RA:
 		case 0x80|21:	/* RFC1770 */
 			break;
diff --git a/net/ipv4/ip_options.c b/net/ipv4/ip_options.c
index 406056edc02b..e0a93b4fa8cc 100644
--- a/net/ipv4/ip_options.c
+++ b/net/ipv4/ip_options.c
@@ -24,6 +24,7 @@
 #include <net/ip.h>
 #include <net/icmp.h>
 #include <net/route.h>
+#include <net/cipso_ipv4.h>
 
 /* 
  * Write options to IP header, record destination address to
@@ -194,6 +195,13 @@ int ip_options_echo(struct ip_options * dopt, struct sk_buff * skb)
 			dopt->is_strictroute = sopt->is_strictroute;
 		}
 	}
+	if (sopt->cipso) {
+		optlen  = sptr[sopt->cipso+1];
+		dopt->cipso = dopt->optlen+sizeof(struct iphdr);
+		memcpy(dptr, sptr+sopt->cipso, optlen);
+		dptr += optlen;
+		dopt->optlen += optlen;
+	}
 	while (dopt->optlen & 3) {
 		*dptr++ = IPOPT_END;
 		dopt->optlen++;
@@ -434,6 +442,17 @@ int ip_options_compile(struct ip_options * opt, struct sk_buff * skb)
 			if (optptr[2] == 0 && optptr[3] == 0)
 				opt->router_alert = optptr - iph;
 			break;
+		      case IPOPT_CIPSO:
+		        if (opt->cipso) {
+				pp_ptr = optptr;
+				goto error;
+			}
+			opt->cipso = optptr - iph;
+		        if (cipso_v4_validate(&optptr)) {
+				pp_ptr = optptr;
+				goto error;
+			}
+			break;
 		      case IPOPT_SEC:
 		      case IPOPT_SID:
 		      default:
-- 
cgit v1.2.3


From 446fda4f26822b2d42ab3396aafcedf38a9ff2b6 Mon Sep 17 00:00:00 2001
From: Paul Moore <paul.moore@hp.com>
Date: Thu, 3 Aug 2006 16:48:06 -0700
Subject: [NetLabel]: CIPSOv4 engine

Add support for the Commercial IP Security Option (CIPSO) to the IPv4
network stack.  CIPSO has become a de-facto standard for
trusted/labeled networking amongst existing Trusted Operating Systems
such as Trusted Solaris, HP-UX CMW, etc.  This implementation is
designed to be used with the NetLabel subsystem to provide explicit
packet labeling to LSM developers.

The CIPSO/IPv4 packet labeling works by the LSM calling a NetLabel API
function which attaches a CIPSO label (IPv4 option) to a given socket;
this in turn attaches the CIPSO label to every packet leaving the
socket without any extra processing on the outbound side.  On the
inbound side the individual packet's sk_buff is examined through a
call to a NetLabel API function to determine if a CIPSO/IPv4 label is
present and if so the security attributes of the CIPSO label are
returned to the caller of the NetLabel API function.

Signed-off-by: Paul Moore <paul.moore@hp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/sysctl.h     |    4 +
 net/ipv4/Makefile          |    1 +
 net/ipv4/cipso_ipv4.c      | 1607 ++++++++++++++++++++++++++++++++++++++++++++
 net/ipv4/sysctl_net_ipv4.c |   35 +
 4 files changed, 1647 insertions(+)
 create mode 100644 net/ipv4/cipso_ipv4.c

(limited to 'include')

diff --git a/include/linux/sysctl.h b/include/linux/sysctl.h
index e4b1a4d4dcf3..af61d9235409 100644
--- a/include/linux/sysctl.h
+++ b/include/linux/sysctl.h
@@ -411,6 +411,10 @@ enum
 	NET_IPV4_TCP_WORKAROUND_SIGNED_WINDOWS=115,
 	NET_TCP_DMA_COPYBREAK=116,
 	NET_TCP_SLOW_START_AFTER_IDLE=117,
+	NET_CIPSOV4_CACHE_ENABLE=118,
+	NET_CIPSOV4_CACHE_BUCKET_SIZE=119,
+	NET_CIPSOV4_RBM_OPTFMT=120,
+	NET_CIPSOV4_RBM_STRICTVALID=121,
 };
 
 enum {
diff --git a/net/ipv4/Makefile b/net/ipv4/Makefile
index 4878fc5be85f..f66049e28aeb 100644
--- a/net/ipv4/Makefile
+++ b/net/ipv4/Makefile
@@ -47,6 +47,7 @@ obj-$(CONFIG_TCP_CONG_VEGAS) += tcp_vegas.o
 obj-$(CONFIG_TCP_CONG_VENO) += tcp_veno.o
 obj-$(CONFIG_TCP_CONG_SCALABLE) += tcp_scalable.o
 obj-$(CONFIG_TCP_CONG_LP) += tcp_lp.o
+obj-$(CONFIG_NETLABEL) += cipso_ipv4.o
 
 obj-$(CONFIG_XFRM) += xfrm4_policy.o xfrm4_state.o xfrm4_input.o \
 		      xfrm4_output.o
diff --git a/net/ipv4/cipso_ipv4.c b/net/ipv4/cipso_ipv4.c
new file mode 100644
index 000000000000..b82a101c95c5
--- /dev/null
+++ b/net/ipv4/cipso_ipv4.c
@@ -0,0 +1,1607 @@
+/*
+ * CIPSO - Commercial IP Security Option
+ *
+ * This is an implementation of the CIPSO 2.2 protocol as specified in
+ * draft-ietf-cipso-ipsecurity-01.txt with additional tag types as found in
+ * FIPS-188, copies of both documents can be found in the Documentation
+ * directory.  While CIPSO never became a full IETF RFC standard many vendors
+ * have chosen to adopt the protocol and over the years it has become a
+ * de-facto standard for labeled networking.
+ *
+ * Author: Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software;  you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY;  without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program;  if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/string.h>
+#include <linux/jhash.h>
+#include <net/ip.h>
+#include <net/icmp.h>
+#include <net/tcp.h>
+#include <net/netlabel.h>
+#include <net/cipso_ipv4.h>
+#include <asm/bug.h>
+
+struct cipso_v4_domhsh_entry {
+	char *domain;
+	u32 valid;
+	struct list_head list;
+	struct rcu_head rcu;
+};
+
+/* List of available DOI definitions */
+/* XXX - Updates should be minimal so having a single lock for the
+ * cipso_v4_doi_list and the cipso_v4_doi_list->dom_list should be
+ * okay. */
+/* XXX - This currently assumes a minimal number of different DOIs in use,
+ * if in practice there are a lot of different DOIs this list should
+ * probably be turned into a hash table or something similar so we
+ * can do quick lookups. */
+DEFINE_SPINLOCK(cipso_v4_doi_list_lock);
+static struct list_head cipso_v4_doi_list = LIST_HEAD_INIT(cipso_v4_doi_list);
+
+/* Label mapping cache */
+int cipso_v4_cache_enabled = 1;
+int cipso_v4_cache_bucketsize = 10;
+#define CIPSO_V4_CACHE_BUCKETBITS     7
+#define CIPSO_V4_CACHE_BUCKETS        (1 << CIPSO_V4_CACHE_BUCKETBITS)
+#define CIPSO_V4_CACHE_REORDERLIMIT   10
+struct cipso_v4_map_cache_bkt {
+	spinlock_t lock;
+	u32 size;
+	struct list_head list;
+};
+struct cipso_v4_map_cache_entry {
+	u32 hash;
+	unsigned char *key;
+	size_t key_len;
+
+	struct netlbl_lsm_cache lsm_data;
+
+	u32 activity;
+	struct list_head list;
+};
+static struct cipso_v4_map_cache_bkt *cipso_v4_cache = NULL;
+
+/* Restricted bitmap (tag #1) flags */
+int cipso_v4_rbm_optfmt = 0;
+int cipso_v4_rbm_strictvalid = 1;
+
+/*
+ * Helper Functions
+ */
+
+/**
+ * cipso_v4_bitmap_walk - Walk a bitmap looking for a bit
+ * @bitmap: the bitmap
+ * @bitmap_len: length in bits
+ * @offset: starting offset
+ * @state: if non-zero, look for a set (1) bit else look for a cleared (0) bit
+ *
+ * Description:
+ * Starting at @offset, walk the bitmap from left to right until either the
+ * desired bit is found or we reach the end.  Return the bit offset, -1 if
+ * not found, or -2 if error.
+ */
+static int cipso_v4_bitmap_walk(const unsigned char *bitmap,
+				u32 bitmap_len,
+				u32 offset,
+				u8 state)
+{
+	u32 bit_spot;
+	u32 byte_offset;
+	unsigned char bitmask;
+	unsigned char byte;
+
+	/* gcc always rounds to zero when doing integer division */
+	byte_offset = offset / 8;
+	byte = bitmap[byte_offset];
+	bit_spot = offset;
+	bitmask = 0x80 >> (offset % 8);
+
+	while (bit_spot < bitmap_len) {
+		if ((state && (byte & bitmask) == bitmask) ||
+		    (state == 0 && (byte & bitmask) == 0))
+			return bit_spot;
+
+		bit_spot++;
+		bitmask >>= 1;
+		if (bitmask == 0) {
+			byte = bitmap[++byte_offset];
+			bitmask = 0x80;
+		}
+	}
+
+	return -1;
+}
+
+/**
+ * cipso_v4_bitmap_setbit - Sets a single bit in a bitmap
+ * @bitmap: the bitmap
+ * @bit: the bit
+ * @state: if non-zero, set the bit (1) else clear the bit (0)
+ *
+ * Description:
+ * Set a single bit in the bitmask.  Returns zero on success, negative values
+ * on error.
+ */
+static void cipso_v4_bitmap_setbit(unsigned char *bitmap,
+				   u32 bit,
+				   u8 state)
+{
+	u32 byte_spot;
+	u8 bitmask;
+
+	/* gcc always rounds to zero when doing integer division */
+	byte_spot = bit / 8;
+	bitmask = 0x80 >> (bit % 8);
+	if (state)
+		bitmap[byte_spot] |= bitmask;
+	else
+		bitmap[byte_spot] &= ~bitmask;
+}
+
+/**
+ * cipso_v4_doi_domhsh_free - Frees a domain list entry
+ * @entry: the entry's RCU field
+ *
+ * Description:
+ * This function is designed to be used as a callback to the call_rcu()
+ * function so that the memory allocated to a domain list entry can be released
+ * safely.
+ *
+ */
+static void cipso_v4_doi_domhsh_free(struct rcu_head *entry)
+{
+	struct cipso_v4_domhsh_entry *ptr;
+
+	ptr = container_of(entry, struct cipso_v4_domhsh_entry, rcu);
+	kfree(ptr->domain);
+	kfree(ptr);
+}
+
+/**
+ * cipso_v4_cache_entry_free - Frees a cache entry
+ * @entry: the entry to free
+ *
+ * Description:
+ * This function frees the memory associated with a cache entry.
+ *
+ */
+static void cipso_v4_cache_entry_free(struct cipso_v4_map_cache_entry *entry)
+{
+	if (entry->lsm_data.free)
+		entry->lsm_data.free(entry->lsm_data.data);
+	kfree(entry->key);
+	kfree(entry);
+}
+
+/**
+ * cipso_v4_map_cache_hash - Hashing function for the CIPSO cache
+ * @key: the hash key
+ * @key_len: the length of the key in bytes
+ *
+ * Description:
+ * The CIPSO tag hashing function.  Returns a 32-bit hash value.
+ *
+ */
+static u32 cipso_v4_map_cache_hash(const unsigned char *key, u32 key_len)
+{
+	return jhash(key, key_len, 0);
+}
+
+/*
+ * Label Mapping Cache Functions
+ */
+
+/**
+ * cipso_v4_cache_init - Initialize the CIPSO cache
+ *
+ * Description:
+ * Initializes the CIPSO label mapping cache, this function should be called
+ * before any of the other functions defined in this file.  Returns zero on
+ * success, negative values on error.
+ *
+ */
+static int cipso_v4_cache_init(void)
+{
+	u32 iter;
+
+	cipso_v4_cache = kcalloc(CIPSO_V4_CACHE_BUCKETS,
+				 sizeof(struct cipso_v4_map_cache_bkt),
+				 GFP_KERNEL);
+	if (cipso_v4_cache == NULL)
+		return -ENOMEM;
+
+	for (iter = 0; iter < CIPSO_V4_CACHE_BUCKETS; iter++) {
+		spin_lock_init(&cipso_v4_cache[iter].lock);
+		cipso_v4_cache[iter].size = 0;
+		INIT_LIST_HEAD(&cipso_v4_cache[iter].list);
+	}
+
+	return 0;
+}
+
+/**
+ * cipso_v4_cache_invalidate - Invalidates the current CIPSO cache
+ *
+ * Description:
+ * Invalidates and frees any entries in the CIPSO cache.  Returns zero on
+ * success and negative values on failure.
+ *
+ */
+void cipso_v4_cache_invalidate(void)
+{
+	struct cipso_v4_map_cache_entry *entry, *tmp_entry;
+	u32 iter;
+
+	for (iter = 0; iter < CIPSO_V4_CACHE_BUCKETS; iter++) {
+		spin_lock(&cipso_v4_cache[iter].lock);
+		list_for_each_entry_safe(entry,
+					 tmp_entry,
+					 &cipso_v4_cache[iter].list, list) {
+			list_del(&entry->list);
+			cipso_v4_cache_entry_free(entry);
+		}
+		cipso_v4_cache[iter].size = 0;
+		spin_unlock(&cipso_v4_cache[iter].lock);
+	}
+
+	return;
+}
+
+/**
+ * cipso_v4_cache_check - Check the CIPSO cache for a label mapping
+ * @key: the buffer to check
+ * @key_len: buffer length in bytes
+ * @secattr: the security attribute struct to use
+ *
+ * Description:
+ * This function checks the cache to see if a label mapping already exists for
+ * the given key.  If there is a match then the cache is adjusted and the
+ * @secattr struct is populated with the correct LSM security attributes.  The
+ * cache is adjusted in the following manner if the entry is not already the
+ * first in the cache bucket:
+ *
+ *  1. The cache entry's activity counter is incremented
+ *  2. The previous (higher ranking) entry's activity counter is decremented
+ *  3. If the difference between the two activity counters is geater than
+ *     CIPSO_V4_CACHE_REORDERLIMIT the two entries are swapped
+ *
+ * Returns zero on success, -ENOENT for a cache miss, and other negative values
+ * on error.
+ *
+ */
+static int cipso_v4_cache_check(const unsigned char *key,
+				u32 key_len,
+				struct netlbl_lsm_secattr *secattr)
+{
+	u32 bkt;
+	struct cipso_v4_map_cache_entry *entry;
+	struct cipso_v4_map_cache_entry *prev_entry = NULL;
+	u32 hash;
+
+	if (!cipso_v4_cache_enabled)
+		return -ENOENT;
+
+	hash = cipso_v4_map_cache_hash(key, key_len);
+	bkt = hash & (CIPSO_V4_CACHE_BUCKETBITS - 1);
+	spin_lock(&cipso_v4_cache[bkt].lock);
+	list_for_each_entry(entry, &cipso_v4_cache[bkt].list, list) {
+		if (entry->hash == hash &&
+		    entry->key_len == key_len &&
+		    memcmp(entry->key, key, key_len) == 0) {
+			entry->activity += 1;
+			secattr->cache.free = entry->lsm_data.free;
+			secattr->cache.data = entry->lsm_data.data;
+			if (prev_entry == NULL) {
+				spin_unlock(&cipso_v4_cache[bkt].lock);
+				return 0;
+			}
+
+			if (prev_entry->activity > 0)
+				prev_entry->activity -= 1;
+			if (entry->activity > prev_entry->activity &&
+			    entry->activity - prev_entry->activity >
+			    CIPSO_V4_CACHE_REORDERLIMIT) {
+				__list_del(entry->list.prev, entry->list.next);
+				__list_add(&entry->list,
+					   prev_entry->list.prev,
+					   &prev_entry->list);
+			}
+
+			spin_unlock(&cipso_v4_cache[bkt].lock);
+			return 0;
+		}
+		prev_entry = entry;
+	}
+	spin_unlock(&cipso_v4_cache[bkt].lock);
+
+	return -ENOENT;
+}
+
+/**
+ * cipso_v4_cache_add - Add an entry to the CIPSO cache
+ * @skb: the packet
+ * @secattr: the packet's security attributes
+ *
+ * Description:
+ * Add a new entry into the CIPSO label mapping cache.  Add the new entry to
+ * head of the cache bucket's list, if the cache bucket is out of room remove
+ * the last entry in the list first.  It is important to note that there is
+ * currently no checking for duplicate keys.  Returns zero on success,
+ * negative values on failure.
+ *
+ */
+int cipso_v4_cache_add(const struct sk_buff *skb,
+		       const struct netlbl_lsm_secattr *secattr)
+{
+	int ret_val = -EPERM;
+	u32 bkt;
+	struct cipso_v4_map_cache_entry *entry = NULL;
+	struct cipso_v4_map_cache_entry *old_entry = NULL;
+	unsigned char *cipso_ptr;
+	u32 cipso_ptr_len;
+
+	if (!cipso_v4_cache_enabled || cipso_v4_cache_bucketsize <= 0)
+		return 0;
+
+	cipso_ptr = CIPSO_V4_OPTPTR(skb);
+	cipso_ptr_len = cipso_ptr[1];
+
+	entry = kzalloc(sizeof(*entry), GFP_ATOMIC);
+	if (entry == NULL)
+		return -ENOMEM;
+	entry->key = kmalloc(cipso_ptr_len, GFP_ATOMIC);
+	if (entry->key == NULL) {
+		ret_val = -ENOMEM;
+		goto cache_add_failure;
+	}
+	memcpy(entry->key, cipso_ptr, cipso_ptr_len);
+	entry->key_len = cipso_ptr_len;
+	entry->hash = cipso_v4_map_cache_hash(cipso_ptr, cipso_ptr_len);
+	entry->lsm_data.free = secattr->cache.free;
+	entry->lsm_data.data = secattr->cache.data;
+
+	bkt = entry->hash & (CIPSO_V4_CACHE_BUCKETBITS - 1);
+	spin_lock(&cipso_v4_cache[bkt].lock);
+	if (cipso_v4_cache[bkt].size < cipso_v4_cache_bucketsize) {
+		list_add(&entry->list, &cipso_v4_cache[bkt].list);
+		cipso_v4_cache[bkt].size += 1;
+	} else {
+		old_entry = list_entry(cipso_v4_cache[bkt].list.prev,
+				       struct cipso_v4_map_cache_entry, list);
+		list_del(&old_entry->list);
+		list_add(&entry->list, &cipso_v4_cache[bkt].list);
+		cipso_v4_cache_entry_free(old_entry);
+	}
+	spin_unlock(&cipso_v4_cache[bkt].lock);
+
+	return 0;
+
+cache_add_failure:
+	if (entry)
+		cipso_v4_cache_entry_free(entry);
+	return ret_val;
+}
+
+/*
+ * DOI List Functions
+ */
+
+/**
+ * cipso_v4_doi_search - Searches for a DOI definition
+ * @doi: the DOI to search for
+ *
+ * Description:
+ * Search the DOI definition list for a DOI definition with a DOI value that
+ * matches @doi.  The caller is responsibile for calling rcu_read_[un]lock().
+ * Returns a pointer to the DOI definition on success and NULL on failure.
+ */
+static struct cipso_v4_doi *cipso_v4_doi_search(u32 doi)
+{
+	struct cipso_v4_doi *iter;
+
+	list_for_each_entry_rcu(iter, &cipso_v4_doi_list, list)
+		if (iter->doi == doi && iter->valid)
+			return iter;
+	return NULL;
+}
+
+/**
+ * cipso_v4_doi_add - Add a new DOI to the CIPSO protocol engine
+ * @doi_def: the DOI structure
+ *
+ * Description:
+ * The caller defines a new DOI for use by the CIPSO engine and calls this
+ * function to add it to the list of acceptable domains.  The caller must
+ * ensure that the mapping table specified in @doi_def->map meets all of the
+ * requirements of the mapping type (see cipso_ipv4.h for details).  Returns
+ * zero on success and non-zero on failure.
+ *
+ */
+int cipso_v4_doi_add(struct cipso_v4_doi *doi_def)
+{
+	if (doi_def == NULL || doi_def->doi == CIPSO_V4_DOI_UNKNOWN)
+		return -EINVAL;
+
+	doi_def->valid = 1;
+	INIT_RCU_HEAD(&doi_def->rcu);
+	INIT_LIST_HEAD(&doi_def->dom_list);
+
+	rcu_read_lock();
+	if (cipso_v4_doi_search(doi_def->doi) != NULL)
+		goto doi_add_failure_rlock;
+	spin_lock(&cipso_v4_doi_list_lock);
+	if (cipso_v4_doi_search(doi_def->doi) != NULL)
+		goto doi_add_failure_slock;
+	list_add_tail_rcu(&doi_def->list, &cipso_v4_doi_list);
+	spin_unlock(&cipso_v4_doi_list_lock);
+	rcu_read_unlock();
+
+	return 0;
+
+doi_add_failure_slock:
+	spin_unlock(&cipso_v4_doi_list_lock);
+doi_add_failure_rlock:
+	rcu_read_unlock();
+	return -EEXIST;
+}
+
+/**
+ * cipso_v4_doi_remove - Remove an existing DOI from the CIPSO protocol engine
+ * @doi: the DOI value
+ * @callback: the DOI cleanup/free callback
+ *
+ * Description:
+ * Removes a DOI definition from the CIPSO engine, @callback is called to
+ * free any memory.  The NetLabel routines will be called to release their own
+ * LSM domain mappings as well as our own domain list.  Returns zero on
+ * success and negative values on failure.
+ *
+ */
+int cipso_v4_doi_remove(u32 doi, void (*callback) (struct rcu_head * head))
+{
+	struct cipso_v4_doi *doi_def;
+	struct cipso_v4_domhsh_entry *dom_iter;
+
+	rcu_read_lock();
+	if (cipso_v4_doi_search(doi) != NULL) {
+		spin_lock(&cipso_v4_doi_list_lock);
+		doi_def = cipso_v4_doi_search(doi);
+		if (doi_def == NULL) {
+			spin_unlock(&cipso_v4_doi_list_lock);
+			rcu_read_unlock();
+			return -ENOENT;
+		}
+		doi_def->valid = 0;
+		list_del_rcu(&doi_def->list);
+		spin_unlock(&cipso_v4_doi_list_lock);
+		list_for_each_entry_rcu(dom_iter, &doi_def->dom_list, list)
+			if (dom_iter->valid)
+				netlbl_domhsh_remove(dom_iter->domain);
+		cipso_v4_cache_invalidate();
+		rcu_read_unlock();
+
+		call_rcu(&doi_def->rcu, callback);
+		return 0;
+	}
+	rcu_read_unlock();
+
+	return -ENOENT;
+}
+
+/**
+ * cipso_v4_doi_getdef - Returns a pointer to a valid DOI definition
+ * @doi: the DOI value
+ *
+ * Description:
+ * Searches for a valid DOI definition and if one is found it is returned to
+ * the caller.  Otherwise NULL is returned.  The caller must ensure that
+ * rcu_read_lock() is held while accessing the returned definition.
+ *
+ */
+struct cipso_v4_doi *cipso_v4_doi_getdef(u32 doi)
+{
+	return cipso_v4_doi_search(doi);
+}
+
+/**
+ * cipso_v4_doi_dump_all - Dump all the CIPSO DOI definitions into a sk_buff
+ * @headroom: the amount of headroom to allocate for the sk_buff
+ *
+ * Description:
+ * Dump a list of all the configured DOI values into a sk_buff.  The returned
+ * sk_buff has room at the front of the sk_buff for @headroom bytes.  See
+ * net/netlabel/netlabel_cipso_v4.h for the LISTALL message format.  This
+ * function may fail if another process is changing the DOI list at the same
+ * time.  Returns a pointer to a sk_buff on success, NULL on error.
+ *
+ */
+struct sk_buff *cipso_v4_doi_dump_all(size_t headroom)
+{
+	struct sk_buff *skb = NULL;
+	struct cipso_v4_doi *iter;
+	u32 doi_cnt = 0;
+	ssize_t buf_len;
+
+	buf_len = NETLBL_LEN_U32;
+	rcu_read_lock();
+	list_for_each_entry_rcu(iter, &cipso_v4_doi_list, list)
+		if (iter->valid) {
+			doi_cnt += 1;
+			buf_len += 2 * NETLBL_LEN_U32;
+		}
+
+	skb = netlbl_netlink_alloc_skb(headroom, buf_len, GFP_ATOMIC);
+	if (skb == NULL)
+		goto doi_dump_all_failure;
+
+	if (nla_put_u32(skb, NLA_U32, doi_cnt) != 0)
+		goto doi_dump_all_failure;
+	buf_len -= NETLBL_LEN_U32;
+	list_for_each_entry_rcu(iter, &cipso_v4_doi_list, list)
+		if (iter->valid) {
+			if (buf_len < 2 * NETLBL_LEN_U32)
+				goto doi_dump_all_failure;
+			if (nla_put_u32(skb, NLA_U32, iter->doi) != 0)
+				goto doi_dump_all_failure;
+			if (nla_put_u32(skb, NLA_U32, iter->type) != 0)
+				goto doi_dump_all_failure;
+			buf_len -= 2 * NETLBL_LEN_U32;
+		}
+	rcu_read_unlock();
+
+	return skb;
+
+doi_dump_all_failure:
+	rcu_read_unlock();
+	kfree(skb);
+	return NULL;
+}
+
+/**
+ * cipso_v4_doi_dump - Dump a CIPSO DOI definition into a sk_buff
+ * @doi: the DOI value
+ * @headroom: the amount of headroom to allocate for the sk_buff
+ *
+ * Description:
+ * Lookup the DOI definition matching @doi and dump it's contents into a
+ * sk_buff.  The returned sk_buff has room at the front of the sk_buff for
+ * @headroom bytes.  See net/netlabel/netlabel_cipso_v4.h for the LIST message
+ * format.  This function may fail if another process is changing the DOI list
+ * at the same time.  Returns a pointer to a sk_buff on success, NULL on error.
+ *
+ */
+struct sk_buff *cipso_v4_doi_dump(u32 doi, size_t headroom)
+{
+	struct sk_buff *skb = NULL;
+	struct cipso_v4_doi *iter;
+	u32 tag_cnt = 0;
+	u32 lvl_cnt = 0;
+	u32 cat_cnt = 0;
+	ssize_t buf_len;
+	ssize_t tmp;
+
+	rcu_read_lock();
+	iter = cipso_v4_doi_getdef(doi);
+	if (iter == NULL)
+		goto doi_dump_failure;
+	buf_len = NETLBL_LEN_U32;
+	switch (iter->type) {
+	case CIPSO_V4_MAP_PASS:
+		buf_len += NETLBL_LEN_U32;
+		while(tag_cnt < CIPSO_V4_TAG_MAXCNT &&
+		      iter->tags[tag_cnt] != CIPSO_V4_TAG_INVALID) {
+			tag_cnt += 1;
+			buf_len += NETLBL_LEN_U8;
+		}
+		break;
+	case CIPSO_V4_MAP_STD:
+		buf_len += 3 * NETLBL_LEN_U32;
+		while (tag_cnt < CIPSO_V4_TAG_MAXCNT &&
+		       iter->tags[tag_cnt] != CIPSO_V4_TAG_INVALID) {
+			tag_cnt += 1;
+			buf_len += NETLBL_LEN_U8;
+		}
+		for (tmp = 0; tmp < iter->map.std->lvl.local_size; tmp++)
+			if (iter->map.std->lvl.local[tmp] !=
+			    CIPSO_V4_INV_LVL) {
+				lvl_cnt += 1;
+				buf_len += NETLBL_LEN_U32 + NETLBL_LEN_U8;
+			}
+		for (tmp = 0; tmp < iter->map.std->cat.local_size; tmp++)
+			if (iter->map.std->cat.local[tmp] !=
+			    CIPSO_V4_INV_CAT) {
+				cat_cnt += 1;
+				buf_len += NETLBL_LEN_U32 + NETLBL_LEN_U16;
+			}
+		break;
+	}
+
+	skb = netlbl_netlink_alloc_skb(headroom, buf_len, GFP_ATOMIC);
+	if (skb == NULL)
+		goto doi_dump_failure;
+
+	if (nla_put_u32(skb, NLA_U32, iter->type) != 0)
+		goto doi_dump_failure;
+	buf_len -= NETLBL_LEN_U32;
+	if (iter != cipso_v4_doi_getdef(doi))
+		goto doi_dump_failure;
+	switch (iter->type) {
+	case CIPSO_V4_MAP_PASS:
+		if (nla_put_u32(skb, NLA_U32, tag_cnt) != 0)
+			goto doi_dump_failure;
+		buf_len -= NETLBL_LEN_U32;
+		for (tmp = 0;
+		     tmp < CIPSO_V4_TAG_MAXCNT &&
+			     iter->tags[tmp] != CIPSO_V4_TAG_INVALID;
+		     tmp++) {
+			if (buf_len < NETLBL_LEN_U8)
+				goto doi_dump_failure;
+			if (nla_put_u8(skb, NLA_U8, iter->tags[tmp]) != 0)
+				goto doi_dump_failure;
+			buf_len -= NETLBL_LEN_U8;
+		}
+		break;
+	case CIPSO_V4_MAP_STD:
+		if (nla_put_u32(skb, NLA_U32, tag_cnt) != 0)
+			goto doi_dump_failure;
+		if (nla_put_u32(skb, NLA_U32, lvl_cnt) != 0)
+			goto doi_dump_failure;
+		if (nla_put_u32(skb, NLA_U32, cat_cnt) != 0)
+			goto doi_dump_failure;
+		buf_len -= 3 * NETLBL_LEN_U32;
+		for (tmp = 0;
+		     tmp < CIPSO_V4_TAG_MAXCNT &&
+			     iter->tags[tmp] != CIPSO_V4_TAG_INVALID;
+		     tmp++) {
+			if (buf_len < NETLBL_LEN_U8)
+				goto doi_dump_failure;
+			if (nla_put_u8(skb, NLA_U8, iter->tags[tmp]) != 0)
+				goto doi_dump_failure;
+			buf_len -= NETLBL_LEN_U8;
+		}
+		for (tmp = 0; tmp < iter->map.std->lvl.local_size; tmp++)
+			if (iter->map.std->lvl.local[tmp] !=
+			    CIPSO_V4_INV_LVL) {
+				if (buf_len < NETLBL_LEN_U32 + NETLBL_LEN_U8)
+					goto doi_dump_failure;
+				if (nla_put_u32(skb, NLA_U32, tmp) != 0)
+					goto doi_dump_failure;
+				if (nla_put_u8(skb,
+					   NLA_U8,
+					   iter->map.std->lvl.local[tmp]) != 0)
+					goto doi_dump_failure;
+				buf_len -= NETLBL_LEN_U32 + NETLBL_LEN_U8;
+			}
+		for (tmp = 0; tmp < iter->map.std->cat.local_size; tmp++)
+			if (iter->map.std->cat.local[tmp] !=
+			    CIPSO_V4_INV_CAT) {
+				if (buf_len < NETLBL_LEN_U32 + NETLBL_LEN_U16)
+					goto doi_dump_failure;
+				if (nla_put_u32(skb, NLA_U32, tmp) != 0)
+					goto doi_dump_failure;
+				if (nla_put_u16(skb,
+					   NLA_U16,
+					   iter->map.std->cat.local[tmp]) != 0)
+					goto doi_dump_failure;
+				buf_len -= NETLBL_LEN_U32 + NETLBL_LEN_U16;
+			}
+		break;
+	}
+	rcu_read_unlock();
+
+	return skb;
+
+doi_dump_failure:
+	rcu_read_unlock();
+	kfree(skb);
+	return NULL;
+}
+
+/**
+ * cipso_v4_doi_domhsh_add - Adds a domain entry to a DOI definition
+ * @doi_def: the DOI definition
+ * @domain: the domain to add
+ *
+ * Description:
+ * Adds the @domain to the the DOI specified by @doi_def, this function
+ * should only be called by external functions (i.e. NetLabel).  This function
+ * does allocate memory.  Returns zero on success, negative values on failure.
+ *
+ */
+int cipso_v4_doi_domhsh_add(struct cipso_v4_doi *doi_def, const char *domain)
+{
+	struct cipso_v4_domhsh_entry *iter;
+	struct cipso_v4_domhsh_entry *new_dom;
+
+	new_dom = kzalloc(sizeof(*new_dom), GFP_KERNEL);
+	if (new_dom == NULL)
+		return -ENOMEM;
+	if (domain) {
+		new_dom->domain = kstrdup(domain, GFP_KERNEL);
+		if (new_dom->domain == NULL) {
+			kfree(new_dom);
+			return -ENOMEM;
+		}
+	}
+	new_dom->valid = 1;
+	INIT_RCU_HEAD(&new_dom->rcu);
+
+	rcu_read_lock();
+	spin_lock(&cipso_v4_doi_list_lock);
+	list_for_each_entry_rcu(iter, &doi_def->dom_list, list)
+		if (iter->valid &&
+		    ((domain != NULL && iter->domain != NULL &&
+		      strcmp(iter->domain, domain) == 0) ||
+		     (domain == NULL && iter->domain == NULL))) {
+			spin_unlock(&cipso_v4_doi_list_lock);
+			rcu_read_unlock();
+			kfree(new_dom->domain);
+			kfree(new_dom);
+			return -EEXIST;
+		}
+	list_add_tail_rcu(&new_dom->list, &doi_def->dom_list);
+	spin_unlock(&cipso_v4_doi_list_lock);
+	rcu_read_unlock();
+
+	return 0;
+}
+
+/**
+ * cipso_v4_doi_domhsh_remove - Removes a domain entry from a DOI definition
+ * @doi_def: the DOI definition
+ * @domain: the domain to remove
+ *
+ * Description:
+ * Removes the @domain from the DOI specified by @doi_def, this function
+ * should only be called by external functions (i.e. NetLabel).   Returns zero
+ * on success and negative values on error.
+ *
+ */
+int cipso_v4_doi_domhsh_remove(struct cipso_v4_doi *doi_def,
+			       const char *domain)
+{
+	struct cipso_v4_domhsh_entry *iter;
+
+	rcu_read_lock();
+	spin_lock(&cipso_v4_doi_list_lock);
+	list_for_each_entry_rcu(iter, &doi_def->dom_list, list)
+		if (iter->valid &&
+		    ((domain != NULL && iter->domain != NULL &&
+		      strcmp(iter->domain, domain) == 0) ||
+		     (domain == NULL && iter->domain == NULL))) {
+			iter->valid = 0;
+			list_del_rcu(&iter->list);
+			spin_unlock(&cipso_v4_doi_list_lock);
+			rcu_read_unlock();
+			call_rcu(&iter->rcu, cipso_v4_doi_domhsh_free);
+
+			return 0;
+		}
+	spin_unlock(&cipso_v4_doi_list_lock);
+	rcu_read_unlock();
+
+	return -ENOENT;
+}
+
+/*
+ * Label Mapping Functions
+ */
+
+/**
+ * cipso_v4_map_lvl_valid - Checks to see if the given level is understood
+ * @doi_def: the DOI definition
+ * @level: the level to check
+ *
+ * Description:
+ * Checks the given level against the given DOI definition and returns a
+ * negative value if the level does not have a valid mapping and a zero value
+ * if the level is defined by the DOI.
+ *
+ */
+static int cipso_v4_map_lvl_valid(const struct cipso_v4_doi *doi_def, u8 level)
+{
+	switch (doi_def->type) {
+	case CIPSO_V4_MAP_PASS:
+		return 0;
+	case CIPSO_V4_MAP_STD:
+		if (doi_def->map.std->lvl.cipso[level] < CIPSO_V4_INV_LVL)
+			return 0;
+		break;
+	}
+
+	return -EFAULT;
+}
+
+/**
+ * cipso_v4_map_lvl_hton - Perform a level mapping from the host to the network
+ * @doi_def: the DOI definition
+ * @host_lvl: the host MLS level
+ * @net_lvl: the network/CIPSO MLS level
+ *
+ * Description:
+ * Perform a label mapping to translate a local MLS level to the correct
+ * CIPSO level using the given DOI definition.  Returns zero on success,
+ * negative values otherwise.
+ *
+ */
+static int cipso_v4_map_lvl_hton(const struct cipso_v4_doi *doi_def,
+				 u32 host_lvl,
+				 u32 *net_lvl)
+{
+	switch (doi_def->type) {
+	case CIPSO_V4_MAP_PASS:
+		*net_lvl = host_lvl;
+		return 0;
+	case CIPSO_V4_MAP_STD:
+		if (host_lvl < doi_def->map.std->lvl.local_size) {
+			*net_lvl = doi_def->map.std->lvl.local[host_lvl];
+			return 0;
+		}
+		break;
+	}
+
+	return -EINVAL;
+}
+
+/**
+ * cipso_v4_map_lvl_ntoh - Perform a level mapping from the network to the host
+ * @doi_def: the DOI definition
+ * @net_lvl: the network/CIPSO MLS level
+ * @host_lvl: the host MLS level
+ *
+ * Description:
+ * Perform a label mapping to translate a CIPSO level to the correct local MLS
+ * level using the given DOI definition.  Returns zero on success, negative
+ * values otherwise.
+ *
+ */
+static int cipso_v4_map_lvl_ntoh(const struct cipso_v4_doi *doi_def,
+				 u32 net_lvl,
+				 u32 *host_lvl)
+{
+	struct cipso_v4_std_map_tbl *map_tbl;
+
+	switch (doi_def->type) {
+	case CIPSO_V4_MAP_PASS:
+		*host_lvl = net_lvl;
+		return 0;
+	case CIPSO_V4_MAP_STD:
+		map_tbl = doi_def->map.std;
+		if (net_lvl < map_tbl->lvl.cipso_size &&
+		    map_tbl->lvl.cipso[net_lvl] < CIPSO_V4_INV_LVL) {
+			*host_lvl = doi_def->map.std->lvl.cipso[net_lvl];
+			return 0;
+		}
+		break;
+	}
+
+	return -EINVAL;
+}
+
+/**
+ * cipso_v4_map_cat_rbm_valid - Checks to see if the category bitmap is valid
+ * @doi_def: the DOI definition
+ * @bitmap: category bitmap
+ * @bitmap_len: bitmap length in bytes
+ *
+ * Description:
+ * Checks the given category bitmap against the given DOI definition and
+ * returns a negative value if any of the categories in the bitmap do not have
+ * a valid mapping and a zero value if all of the categories are valid.
+ *
+ */
+static int cipso_v4_map_cat_rbm_valid(const struct cipso_v4_doi *doi_def,
+				      const unsigned char *bitmap,
+				      u32 bitmap_len)
+{
+	int cat = -1;
+	u32 bitmap_len_bits = bitmap_len * 8;
+	u32 cipso_cat_size = doi_def->map.std->cat.cipso_size;
+	u32 *cipso_array = doi_def->map.std->cat.cipso;
+
+	switch (doi_def->type) {
+	case CIPSO_V4_MAP_PASS:
+		return 0;
+	case CIPSO_V4_MAP_STD:
+		for (;;) {
+			cat = cipso_v4_bitmap_walk(bitmap,
+						   bitmap_len_bits,
+						   cat + 1,
+						   1);
+			if (cat < 0)
+				break;
+			if (cat >= cipso_cat_size ||
+			    cipso_array[cat] >= CIPSO_V4_INV_CAT)
+				return -EFAULT;
+		}
+
+		if (cat == -1)
+			return 0;
+		break;
+	}
+
+	return -EFAULT;
+}
+
+/**
+ * cipso_v4_map_cat_rbm_hton - Perform a category mapping from host to network
+ * @doi_def: the DOI definition
+ * @host_cat: the category bitmap in host format
+ * @host_cat_len: the length of the host's category bitmap in bytes
+ * @net_cat: the zero'd out category bitmap in network/CIPSO format
+ * @net_cat_len: the length of the CIPSO bitmap in bytes
+ *
+ * Description:
+ * Perform a label mapping to translate a local MLS category bitmap to the
+ * correct CIPSO bitmap using the given DOI definition.  Returns the minimum
+ * size in bytes of the network bitmap on success, negative values otherwise.
+ *
+ */
+static int cipso_v4_map_cat_rbm_hton(const struct cipso_v4_doi *doi_def,
+				     const unsigned char *host_cat,
+				     u32 host_cat_len,
+				     unsigned char *net_cat,
+				     u32 net_cat_len)
+{
+	int host_spot = -1;
+	u32 net_spot;
+	u32 net_spot_max = 0;
+	u32 host_clen_bits = host_cat_len * 8;
+	u32 net_clen_bits = net_cat_len * 8;
+	u32 host_cat_size = doi_def->map.std->cat.local_size;
+	u32 *host_cat_array = doi_def->map.std->cat.local;
+
+	switch (doi_def->type) {
+	case CIPSO_V4_MAP_PASS:
+		net_spot_max = host_cat_len - 1;
+		while (net_spot_max > 0 && host_cat[net_spot_max] == 0)
+			net_spot_max--;
+		if (net_spot_max > net_cat_len)
+			return -EINVAL;
+		memcpy(net_cat, host_cat, net_spot_max);
+		return net_spot_max;
+	case CIPSO_V4_MAP_STD:
+		for (;;) {
+			host_spot = cipso_v4_bitmap_walk(host_cat,
+							 host_clen_bits,
+							 host_spot + 1,
+							 1);
+			if (host_spot < 0)
+				break;
+			if (host_spot >= host_cat_size)
+				return -EPERM;
+
+			net_spot = host_cat_array[host_spot];
+			if (net_spot >= net_clen_bits)
+				return -ENOSPC;
+			cipso_v4_bitmap_setbit(net_cat, net_spot, 1);
+
+			if (net_spot > net_spot_max)
+				net_spot_max = net_spot;
+		}
+
+		if (host_spot == -2)
+			return -EFAULT;
+
+		if (++net_spot_max % 8)
+			return net_spot_max / 8 + 1;
+		return net_spot_max / 8;
+	}
+
+	return -EINVAL;
+}
+
+/**
+ * cipso_v4_map_cat_rbm_ntoh - Perform a category mapping from network to host
+ * @doi_def: the DOI definition
+ * @net_cat: the category bitmap in network/CIPSO format
+ * @net_cat_len: the length of the CIPSO bitmap in bytes
+ * @host_cat: the zero'd out category bitmap in host format
+ * @host_cat_len: the length of the host's category bitmap in bytes
+ *
+ * Description:
+ * Perform a label mapping to translate a CIPSO bitmap to the correct local
+ * MLS category bitmap using the given DOI definition.  Returns the minimum
+ * size in bytes of the host bitmap on success, negative values otherwise.
+ *
+ */
+static int cipso_v4_map_cat_rbm_ntoh(const struct cipso_v4_doi *doi_def,
+				     const unsigned char *net_cat,
+				     u32 net_cat_len,
+				     unsigned char *host_cat,
+				     u32 host_cat_len)
+{
+	u32 host_spot;
+	u32 host_spot_max = 0;
+	int net_spot = -1;
+	u32 net_clen_bits = net_cat_len * 8;
+	u32 host_clen_bits = host_cat_len * 8;
+	u32 net_cat_size = doi_def->map.std->cat.cipso_size;
+	u32 *net_cat_array = doi_def->map.std->cat.cipso;
+
+	switch (doi_def->type) {
+	case CIPSO_V4_MAP_PASS:
+		if (net_cat_len > host_cat_len)
+			return -EINVAL;
+		memcpy(host_cat, net_cat, net_cat_len);
+		return net_cat_len;
+	case CIPSO_V4_MAP_STD:
+		for (;;) {
+			net_spot = cipso_v4_bitmap_walk(net_cat,
+							net_clen_bits,
+							net_spot + 1,
+							1);
+			if (net_spot < 0)
+				break;
+			if (net_spot >= net_cat_size ||
+			    net_cat_array[net_spot] >= CIPSO_V4_INV_CAT)
+				return -EPERM;
+
+			host_spot = net_cat_array[net_spot];
+			if (host_spot >= host_clen_bits)
+				return -ENOSPC;
+			cipso_v4_bitmap_setbit(host_cat, host_spot, 1);
+
+			if (host_spot > host_spot_max)
+				host_spot_max = host_spot;
+		}
+
+		if (net_spot == -2)
+			return -EFAULT;
+
+		if (++host_spot_max % 8)
+			return host_spot_max / 8 + 1;
+		return host_spot_max / 8;
+	}
+
+	return -EINVAL;
+}
+
+/*
+ * Protocol Handling Functions
+ */
+
+#define CIPSO_V4_HDR_LEN              6
+
+/**
+ * cipso_v4_gentag_hdr - Generate a CIPSO option header
+ * @doi_def: the DOI definition
+ * @len: the total tag length in bytes
+ * @buf: the CIPSO option buffer
+ *
+ * Description:
+ * Write a CIPSO header into the beginning of @buffer.  Return zero on success,
+ * negative values on failure.
+ *
+ */
+static int cipso_v4_gentag_hdr(const struct cipso_v4_doi *doi_def,
+			       u32 len,
+			       unsigned char *buf)
+{
+	if (CIPSO_V4_HDR_LEN + len > 40)
+		return -ENOSPC;
+
+	buf[0] = IPOPT_CIPSO;
+	buf[1] = CIPSO_V4_HDR_LEN + len;
+	*(u32 *)&buf[2] = htonl(doi_def->doi);
+
+	return 0;
+}
+
+#define CIPSO_V4_TAG1_CAT_LEN         30
+
+/**
+ * cipso_v4_gentag_rbm - Generate a CIPSO restricted bitmap tag (type #1)
+ * @doi_def: the DOI definition
+ * @secattr: the security attributes
+ * @buffer: the option buffer
+ * @buffer_len: length of buffer in bytes
+ *
+ * Description:
+ * Generate a CIPSO option using the restricted bitmap tag, tag type #1.  The
+ * actual buffer length may be larger than the indicated size due to
+ * translation between host and network category bitmaps.  Returns zero on
+ * success, negative values on failure.
+ *
+ */
+static int cipso_v4_gentag_rbm(const struct cipso_v4_doi *doi_def,
+			       const struct netlbl_lsm_secattr *secattr,
+			       unsigned char **buffer,
+			       u32 *buffer_len)
+{
+	int ret_val = -EPERM;
+	unsigned char *buf = NULL;
+	u32 buf_len;
+	u32 level;
+
+	if (secattr->mls_cat) {
+		buf = kzalloc(CIPSO_V4_HDR_LEN + 4 + CIPSO_V4_TAG1_CAT_LEN,
+			      GFP_ATOMIC);
+		if (buf == NULL)
+			return -ENOMEM;
+
+		ret_val = cipso_v4_map_cat_rbm_hton(doi_def,
+						    secattr->mls_cat,
+						    secattr->mls_cat_len,
+						    &buf[CIPSO_V4_HDR_LEN + 4],
+						    CIPSO_V4_TAG1_CAT_LEN);
+		if (ret_val < 0)
+			goto gentag_failure;
+
+		/* This will send packets using the "optimized" format when
+		 * possibile as specified in  section 3.4.2.6 of the
+		 * CIPSO draft. */
+		if (cipso_v4_rbm_optfmt && (ret_val > 0 && ret_val < 10))
+			ret_val = 10;
+
+		buf_len = 4 + ret_val;
+	} else {
+		buf = kzalloc(CIPSO_V4_HDR_LEN + 4, GFP_ATOMIC);
+		if (buf == NULL)
+			return -ENOMEM;
+		buf_len = 4;
+	}
+
+	ret_val = cipso_v4_map_lvl_hton(doi_def, secattr->mls_lvl, &level);
+	if (ret_val != 0)
+		goto gentag_failure;
+
+	ret_val = cipso_v4_gentag_hdr(doi_def, buf_len, buf);
+	if (ret_val != 0)
+		goto gentag_failure;
+
+	buf[CIPSO_V4_HDR_LEN] = 0x01;
+	buf[CIPSO_V4_HDR_LEN + 1] = buf_len;
+	buf[CIPSO_V4_HDR_LEN + 3] = level;
+
+	*buffer = buf;
+	*buffer_len = CIPSO_V4_HDR_LEN + buf_len;
+
+	return 0;
+
+gentag_failure:
+	kfree(buf);
+	return ret_val;
+}
+
+/**
+ * cipso_v4_parsetag_rbm - Parse a CIPSO restricted bitmap tag
+ * @doi_def: the DOI definition
+ * @tag: the CIPSO tag
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Parse a CIPSO restricted bitmap tag (tag type #1) and return the security
+ * attributes in @secattr.  Return zero on success, negatives values on
+ * failure.
+ *
+ */
+static int cipso_v4_parsetag_rbm(const struct cipso_v4_doi *doi_def,
+				 const unsigned char *tag,
+				 struct netlbl_lsm_secattr *secattr)
+{
+	int ret_val;
+	u8 tag_len = tag[1];
+	u32 level;
+
+	ret_val = cipso_v4_map_lvl_ntoh(doi_def, tag[3], &level);
+	if (ret_val != 0)
+		return ret_val;
+	secattr->mls_lvl = level;
+	secattr->mls_lvl_vld = 1;
+
+	if (tag_len > 4) {
+		switch (doi_def->type) {
+		case CIPSO_V4_MAP_PASS:
+			secattr->mls_cat_len = tag_len - 4;
+			break;
+		case CIPSO_V4_MAP_STD:
+			secattr->mls_cat_len =
+				doi_def->map.std->cat.local_size;
+			break;
+		}
+		secattr->mls_cat = kzalloc(secattr->mls_cat_len, GFP_ATOMIC);
+		if (secattr->mls_cat == NULL)
+			return -ENOMEM;
+
+		ret_val = cipso_v4_map_cat_rbm_ntoh(doi_def,
+						    &tag[4],
+						    tag_len - 4,
+						    secattr->mls_cat,
+						    secattr->mls_cat_len);
+		if (ret_val < 0) {
+			kfree(secattr->mls_cat);
+			return ret_val;
+		}
+		secattr->mls_cat_len = ret_val;
+	}
+
+	return 0;
+}
+
+/**
+ * cipso_v4_validate - Validate a CIPSO option
+ * @option: the start of the option, on error it is set to point to the error
+ *
+ * Description:
+ * This routine is called to validate a CIPSO option, it checks all of the
+ * fields to ensure that they are at least valid, see the draft snippet below
+ * for details.  If the option is valid then a zero value is returned and
+ * the value of @option is unchanged.  If the option is invalid then a
+ * non-zero value is returned and @option is adjusted to point to the
+ * offending portion of the option.  From the IETF draft ...
+ *
+ *  "If any field within the CIPSO options, such as the DOI identifier, is not
+ *   recognized the IP datagram is discarded and an ICMP 'parameter problem'
+ *   (type 12) is generated and returned.  The ICMP code field is set to 'bad
+ *   parameter' (code 0) and the pointer is set to the start of the CIPSO field
+ *   that is unrecognized."
+ *
+ */
+int cipso_v4_validate(unsigned char **option)
+{
+	unsigned char *opt = *option;
+	unsigned char *tag;
+	unsigned char opt_iter;
+	unsigned char err_offset = 0;
+	u8 opt_len;
+	u8 tag_len;
+	struct cipso_v4_doi *doi_def = NULL;
+	u32 tag_iter;
+
+	/* caller already checks for length values that are too large */
+	opt_len = opt[1];
+	if (opt_len < 8) {
+		err_offset = 1;
+		goto validate_return;
+	}
+
+	rcu_read_lock();
+	doi_def = cipso_v4_doi_getdef(ntohl(*((u32 *)&opt[2])));
+	if (doi_def == NULL) {
+		err_offset = 2;
+		goto validate_return_locked;
+	}
+
+	opt_iter = 6;
+	tag = opt + opt_iter;
+	while (opt_iter < opt_len) {
+		for (tag_iter = 0; doi_def->tags[tag_iter] != tag[0];)
+			if (doi_def->tags[tag_iter] == CIPSO_V4_TAG_INVALID ||
+			    ++tag_iter == CIPSO_V4_TAG_MAXCNT) {
+				err_offset = opt_iter;
+				goto validate_return_locked;
+			}
+
+		tag_len = tag[1];
+		if (tag_len > (opt_len - opt_iter)) {
+			err_offset = opt_iter + 1;
+			goto validate_return_locked;
+		}
+
+		switch (tag[0]) {
+		case CIPSO_V4_TAG_RBITMAP:
+			if (tag_len < 4) {
+				err_offset = opt_iter + 1;
+				goto validate_return_locked;
+			}
+
+			/* We are already going to do all the verification
+			 * necessary at the socket layer so from our point of
+			 * view it is safe to turn these checks off (and less
+			 * work), however, the CIPSO draft says we should do
+			 * all the CIPSO validations here but it doesn't
+			 * really specify _exactly_ what we need to validate
+			 * ... so, just make it a sysctl tunable. */
+			if (cipso_v4_rbm_strictvalid) {
+				if (cipso_v4_map_lvl_valid(doi_def,
+							   tag[3]) < 0) {
+					err_offset = opt_iter + 3;
+					goto validate_return_locked;
+				}
+				if (tag_len > 4 &&
+				    cipso_v4_map_cat_rbm_valid(doi_def,
+							    &tag[4],
+							    tag_len - 4) < 0) {
+					err_offset = opt_iter + 4;
+					goto validate_return_locked;
+				}
+			}
+			break;
+		default:
+			err_offset = opt_iter;
+			goto validate_return_locked;
+		}
+
+		tag += tag_len;
+		opt_iter += tag_len;
+	}
+
+validate_return_locked:
+	rcu_read_unlock();
+validate_return:
+	*option = opt + err_offset;
+	return err_offset;
+}
+
+/**
+ * cipso_v4_error - Send the correct reponse for a bad packet
+ * @skb: the packet
+ * @error: the error code
+ * @gateway: CIPSO gateway flag
+ *
+ * Description:
+ * Based on the error code given in @error, send an ICMP error message back to
+ * the originating host.  From the IETF draft ...
+ *
+ *  "If the contents of the CIPSO [option] are valid but the security label is
+ *   outside of the configured host or port label range, the datagram is
+ *   discarded and an ICMP 'destination unreachable' (type 3) is generated and
+ *   returned.  The code field of the ICMP is set to 'communication with
+ *   destination network administratively prohibited' (code 9) or to
+ *   'communication with destination host administratively prohibited'
+ *   (code 10).  The value of the code is dependent on whether the originator
+ *   of the ICMP message is acting as a CIPSO host or a CIPSO gateway.  The
+ *   recipient of the ICMP message MUST be able to handle either value.  The
+ *   same procedure is performed if a CIPSO [option] can not be added to an
+ *   IP packet because it is too large to fit in the IP options area."
+ *
+ *  "If the error is triggered by receipt of an ICMP message, the message is
+ *   discarded and no response is permitted (consistent with general ICMP
+ *   processing rules)."
+ *
+ */
+void cipso_v4_error(struct sk_buff *skb, int error, u32 gateway)
+{
+	if (skb->nh.iph->protocol == IPPROTO_ICMP || error != -EACCES)
+		return;
+
+	if (gateway)
+		icmp_send(skb, ICMP_DEST_UNREACH, ICMP_NET_ANO, 0);
+	else
+		icmp_send(skb, ICMP_DEST_UNREACH, ICMP_HOST_ANO, 0);
+}
+
+/**
+ * cipso_v4_socket_setattr - Add a CIPSO option to a socket
+ * @sock: the socket
+ * @doi_def: the CIPSO DOI to use
+ * @secattr: the specific security attributes of the socket
+ *
+ * Description:
+ * Set the CIPSO option on the given socket using the DOI definition and
+ * security attributes passed to the function.  This function requires
+ * exclusive access to @sock->sk, which means it either needs to be in the
+ * process of being created or locked via lock_sock(sock->sk).  Returns zero on
+ * success and negative values on failure.
+ *
+ */
+int cipso_v4_socket_setattr(const struct socket *sock,
+			    const struct cipso_v4_doi *doi_def,
+			    const struct netlbl_lsm_secattr *secattr)
+{
+	int ret_val = -EPERM;
+	u32 iter;
+	unsigned char *buf = NULL;
+	u32 buf_len = 0;
+	u32 opt_len;
+	struct ip_options *opt = NULL;
+	struct sock *sk;
+	struct inet_sock *sk_inet;
+	struct inet_connection_sock *sk_conn;
+
+	/* In the case of sock_create_lite(), the sock->sk field is not
+	 * defined yet but it is not a problem as the only users of these
+	 * "lite" PF_INET sockets are functions which do an accept() call
+	 * afterwards so we will label the socket as part of the accept(). */
+	sk = sock->sk;
+	if (sk == NULL)
+		return 0;
+
+	/* XXX - This code assumes only one tag per CIPSO option which isn't
+	 * really a good assumption to make but since we only support the MAC
+	 * tags right now it is a safe assumption. */
+	iter = 0;
+	do {
+		switch (doi_def->tags[iter]) {
+		case CIPSO_V4_TAG_RBITMAP:
+			ret_val = cipso_v4_gentag_rbm(doi_def,
+						      secattr,
+						      &buf,
+						      &buf_len);
+			break;
+		default:
+			ret_val = -EPERM;
+			goto socket_setattr_failure;
+		}
+
+		iter++;
+	} while (ret_val != 0 &&
+		 iter < CIPSO_V4_TAG_MAXCNT &&
+		 doi_def->tags[iter] != CIPSO_V4_TAG_INVALID);
+	if (ret_val != 0)
+		goto socket_setattr_failure;
+
+	/* We can't use ip_options_get() directly because it makes a call to
+	 * ip_options_get_alloc() which allocates memory with GFP_KERNEL and
+	 * we can't block here. */
+	opt_len = (buf_len + 3) & ~3;
+	opt = kzalloc(sizeof(*opt) + opt_len, GFP_ATOMIC);
+	if (opt == NULL) {
+		ret_val = -ENOMEM;
+		goto socket_setattr_failure;
+	}
+	memcpy(opt->__data, buf, buf_len);
+	opt->optlen = opt_len;
+	opt->is_data = 1;
+	kfree(buf);
+	buf = NULL;
+	ret_val = ip_options_compile(opt, NULL);
+	if (ret_val != 0)
+		goto socket_setattr_failure;
+
+	sk_inet = inet_sk(sk);
+	if (sk_inet->is_icsk) {
+		sk_conn = inet_csk(sk);
+		if (sk_inet->opt)
+			sk_conn->icsk_ext_hdr_len -= sk_inet->opt->optlen;
+		sk_conn->icsk_ext_hdr_len += opt->optlen;
+		sk_conn->icsk_sync_mss(sk, sk_conn->icsk_pmtu_cookie);
+	}
+	opt = xchg(&sk_inet->opt, opt);
+	kfree(opt);
+
+	return 0;
+
+socket_setattr_failure:
+	kfree(buf);
+	kfree(opt);
+	return ret_val;
+}
+
+/**
+ * cipso_v4_socket_getattr - Get the security attributes from a socket
+ * @sock: the socket
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Query @sock to see if there is a CIPSO option attached to the socket and if
+ * there is return the CIPSO security attributes in @secattr.  Returns zero on
+ * success and negative values on failure.
+ *
+ */
+int cipso_v4_socket_getattr(const struct socket *sock,
+			    struct netlbl_lsm_secattr *secattr)
+{
+	int ret_val = -ENOMSG;
+	struct sock *sk;
+	struct inet_sock *sk_inet;
+	unsigned char *cipso_ptr;
+	u32 doi;
+	struct cipso_v4_doi *doi_def;
+
+	sk = sock->sk;
+	lock_sock(sk);
+	sk_inet = inet_sk(sk);
+	if (sk_inet->opt == NULL || sk_inet->opt->cipso == 0)
+		goto socket_getattr_return;
+	cipso_ptr = sk_inet->opt->__data + sk_inet->opt->cipso -
+		sizeof(struct iphdr);
+	ret_val = cipso_v4_cache_check(cipso_ptr, cipso_ptr[1], secattr);
+	if (ret_val == 0)
+		goto socket_getattr_return;
+
+	doi = ntohl(*(u32 *)&cipso_ptr[2]);
+	rcu_read_lock();
+	doi_def = cipso_v4_doi_getdef(doi);
+	if (doi_def == NULL) {
+		rcu_read_unlock();
+		goto socket_getattr_return;
+	}
+	switch (cipso_ptr[6]) {
+	case CIPSO_V4_TAG_RBITMAP:
+		ret_val = cipso_v4_parsetag_rbm(doi_def,
+						&cipso_ptr[6],
+						secattr);
+		break;
+	}
+	rcu_read_unlock();
+
+socket_getattr_return:
+	release_sock(sk);
+	return ret_val;
+}
+
+/**
+ * cipso_v4_skbuff_getattr - Get the security attributes from the CIPSO option
+ * @skb: the packet
+ * @secattr: the security attributes
+ *
+ * Description:
+ * Parse the given packet's CIPSO option and return the security attributes.
+ * Returns zero on success and negative values on failure.
+ *
+ */
+int cipso_v4_skbuff_getattr(const struct sk_buff *skb,
+			    struct netlbl_lsm_secattr *secattr)
+{
+	int ret_val = -ENOMSG;
+	unsigned char *cipso_ptr;
+	u32 doi;
+	struct cipso_v4_doi *doi_def;
+
+	if (!CIPSO_V4_OPTEXIST(skb))
+		return -ENOMSG;
+	cipso_ptr = CIPSO_V4_OPTPTR(skb);
+	if (cipso_v4_cache_check(cipso_ptr, cipso_ptr[1], secattr) == 0)
+		return 0;
+
+	doi = ntohl(*(u32 *)&cipso_ptr[2]);
+	rcu_read_lock();
+	doi_def = cipso_v4_doi_getdef(doi);
+	if (doi_def == NULL)
+		goto skbuff_getattr_return;
+	switch (cipso_ptr[6]) {
+	case CIPSO_V4_TAG_RBITMAP:
+		ret_val = cipso_v4_parsetag_rbm(doi_def,
+						&cipso_ptr[6],
+						secattr);
+		break;
+	}
+
+skbuff_getattr_return:
+	rcu_read_unlock();
+	return ret_val;
+}
+
+/*
+ * Setup Functions
+ */
+
+/**
+ * cipso_v4_init - Initialize the CIPSO module
+ *
+ * Description:
+ * Initialize the CIPSO module and prepare it for use.  Returns zero on success
+ * and negative values on failure.
+ *
+ */
+static int __init cipso_v4_init(void)
+{
+	int ret_val;
+
+	ret_val = cipso_v4_cache_init();
+	if (ret_val != 0)
+		panic("Failed to initialize the CIPSO/IPv4 cache (%d)\n",
+		      ret_val);
+
+	return 0;
+}
+
+subsys_initcall(cipso_v4_init);
diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c
index 70cea9d08a38..19b2071ff319 100644
--- a/net/ipv4/sysctl_net_ipv4.c
+++ b/net/ipv4/sysctl_net_ipv4.c
@@ -17,6 +17,7 @@
 #include <net/ip.h>
 #include <net/route.h>
 #include <net/tcp.h>
+#include <net/cipso_ipv4.h>
 
 /* From af_inet.c */
 extern int sysctl_ip_nonlocal_bind;
@@ -697,6 +698,40 @@ ctl_table ipv4_table[] = {
 		.mode		= 0644,
 		.proc_handler	= &proc_dointvec
 	},
+#ifdef CONFIG_NETLABEL
+	{
+		.ctl_name	= NET_CIPSOV4_CACHE_ENABLE,
+		.procname	= "cipso_cache_enable",
+		.data		= &cipso_v4_cache_enabled,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= &proc_dointvec,
+	},
+	{
+		.ctl_name	= NET_CIPSOV4_CACHE_BUCKET_SIZE,
+		.procname	= "cipso_cache_bucket_size",
+		.data		= &cipso_v4_cache_bucketsize,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= &proc_dointvec,
+	},
+	{
+		.ctl_name	= NET_CIPSOV4_RBM_OPTFMT,
+		.procname	= "cipso_rbm_optfmt",
+		.data		= &cipso_v4_rbm_optfmt,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= &proc_dointvec,
+	},
+	{
+		.ctl_name	= NET_CIPSOV4_RBM_STRICTVALID,
+		.procname	= "cipso_rbm_strictvalid",
+		.data		= &cipso_v4_rbm_strictvalid,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= &proc_dointvec,
+	},
+#endif /* CONFIG_NETLABEL */
 	{ .ctl_name = 0 }
 };
 
-- 
cgit v1.2.3


From 7420ed23a4f77480b5b7b3245e5da30dd24b7575 Mon Sep 17 00:00:00 2001
From: Venkat Yekkirala <vyekkirala@TrustedCS.com>
Date: Fri, 4 Aug 2006 23:17:57 -0700
Subject: [NetLabel]: SELinux support

Add NetLabel support to the SELinux LSM and modify the
socket_post_create() LSM hook to return an error code.  The most
significant part of this patch is the addition of NetLabel hooks into
the following SELinux LSM hooks:

 * selinux_file_permission()
 * selinux_socket_sendmsg()
 * selinux_socket_post_create()
 * selinux_socket_sock_rcv_skb()
 * selinux_socket_getpeersec_stream()
 * selinux_socket_getpeersec_dgram()
 * selinux_sock_graft()
 * selinux_inet_conn_request()

The basic reasoning behind this patch is that outgoing packets are
"NetLabel'd" by labeling their socket and the NetLabel security
attributes are checked via the additional hook in
selinux_socket_sock_rcv_skb().  NetLabel itself is only a labeling
mechanism, similar to filesystem extended attributes, it is up to the
SELinux enforcement mechanism to perform the actual access checks.

In addition to the changes outlined above this patch also includes
some changes to the extended bitmap (ebitmap) and multi-level security
(mls) code to import and export SELinux TE/MLS attributes into and out
of NetLabel.

Signed-off-by: Paul Moore <paul.moore@hp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/security.h                    |  25 +-
 net/socket.c                                |  13 +-
 security/dummy.c                            |   6 +-
 security/selinux/hooks.c                    |  56 +++-
 security/selinux/include/objsec.h           |   8 +
 security/selinux/include/selinux_netlabel.h | 125 +++++++
 security/selinux/ss/ebitmap.c               | 144 ++++++++
 security/selinux/ss/ebitmap.h               |   6 +
 security/selinux/ss/mls.c                   | 156 +++++++++
 security/selinux/ss/mls.h                   |  21 ++
 security/selinux/ss/services.c              | 488 ++++++++++++++++++++++++++++
 11 files changed, 1020 insertions(+), 28 deletions(-)
 create mode 100644 security/selinux/include/selinux_netlabel.h

(limited to 'include')

diff --git a/include/linux/security.h b/include/linux/security.h
index bb4c80fdfe7a..9f56fb8a4a6c 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -1341,8 +1341,8 @@ struct security_operations {
 	int (*unix_may_send) (struct socket * sock, struct socket * other);
 
 	int (*socket_create) (int family, int type, int protocol, int kern);
-	void (*socket_post_create) (struct socket * sock, int family,
-				    int type, int protocol, int kern);
+	int (*socket_post_create) (struct socket * sock, int family,
+				   int type, int protocol, int kern);
 	int (*socket_bind) (struct socket * sock,
 			    struct sockaddr * address, int addrlen);
 	int (*socket_connect) (struct socket * sock,
@@ -2824,13 +2824,13 @@ static inline int security_socket_create (int family, int type,
 	return security_ops->socket_create(family, type, protocol, kern);
 }
 
-static inline void security_socket_post_create(struct socket * sock, 
-					       int family,
-					       int type, 
-					       int protocol, int kern)
+static inline int security_socket_post_create(struct socket * sock,
+					      int family,
+					      int type,
+					      int protocol, int kern)
 {
-	security_ops->socket_post_create(sock, family, type,
-					 protocol, kern);
+	return security_ops->socket_post_create(sock, family, type,
+						protocol, kern);
 }
 
 static inline int security_socket_bind(struct socket * sock, 
@@ -2982,11 +2982,12 @@ static inline int security_socket_create (int family, int type,
 	return 0;
 }
 
-static inline void security_socket_post_create(struct socket * sock, 
-					       int family,
-					       int type, 
-					       int protocol, int kern)
+static inline int security_socket_post_create(struct socket * sock,
+					      int family,
+					      int type,
+					      int protocol, int kern)
 {
+	return 0;
 }
 
 static inline int security_socket_bind(struct socket * sock, 
diff --git a/net/socket.c b/net/socket.c
index 6d261bf206fc..6756e57e1ff0 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -973,11 +973,18 @@ int sock_create_lite(int family, int type, int protocol, struct socket **res)
 		goto out;
 	}
 
-	security_socket_post_create(sock, family, type, protocol, 1);
 	sock->type = type;
+	err = security_socket_post_create(sock, family, type, protocol, 1);
+	if (err)
+		goto out_release;
+
 out:
 	*res = sock;
 	return err;
+out_release:
+	sock_release(sock);
+	sock = NULL;
+	goto out;
 }
 
 /* No kernel lock held - perfect */
@@ -1214,7 +1221,9 @@ static int __sock_create(int family, int type, int protocol, struct socket **res
 	 */
 	module_put(net_families[family]->owner);
 	*res = sock;
-	security_socket_post_create(sock, family, type, protocol, kern);
+	err = security_socket_post_create(sock, family, type, protocol, kern);
+	if (err)
+		goto out_release;
 
 out:
 	net_family_read_unlock();
diff --git a/security/dummy.c b/security/dummy.c
index 1c45f8e4aad1..aeee70565509 100644
--- a/security/dummy.c
+++ b/security/dummy.c
@@ -709,10 +709,10 @@ static int dummy_socket_create (int family, int type,
 	return 0;
 }
 
-static void dummy_socket_post_create (struct socket *sock, int family, int type,
-				      int protocol, int kern)
+static int dummy_socket_post_create (struct socket *sock, int family, int type,
+				     int protocol, int kern)
 {
-	return;
+	return 0;
 }
 
 static int dummy_socket_bind (struct socket *sock, struct sockaddr *address,
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index 33028b3b19ce..2a6bbb921e1e 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -12,6 +12,8 @@
  *  Copyright (C) 2003 Red Hat, Inc., James Morris <jmorris@redhat.com>
  *  Copyright (C) 2004-2005 Trusted Computer Solutions, Inc.
  *                          <dgoeddel@trustedcs.com>
+ *  Copyright (C) 2006 Hewlett-Packard Development Company, L.P.
+ *                     Paul Moore, <paul.moore@hp.com>
  *
  *	This program is free software; you can redistribute it and/or modify
  *	it under the terms of the GNU General Public License version 2,
@@ -74,6 +76,7 @@
 #include "objsec.h"
 #include "netif.h"
 #include "xfrm.h"
+#include "selinux_netlabel.h"
 
 #define XATTR_SELINUX_SUFFIX "selinux"
 #define XATTR_NAME_SELINUX XATTR_SECURITY_PREFIX XATTR_SELINUX_SUFFIX
@@ -2395,6 +2398,7 @@ static int selinux_inode_listsecurity(struct inode *inode, char *buffer, size_t
 
 static int selinux_file_permission(struct file *file, int mask)
 {
+	int rc;
 	struct inode *inode = file->f_dentry->d_inode;
 
 	if (!mask) {
@@ -2406,8 +2410,12 @@ static int selinux_file_permission(struct file *file, int mask)
 	if ((file->f_flags & O_APPEND) && (mask & MAY_WRITE))
 		mask |= MAY_APPEND;
 
-	return file_has_perm(current, file,
-			     file_mask_to_av(inode->i_mode, mask));
+	rc = file_has_perm(current, file,
+			   file_mask_to_av(inode->i_mode, mask));
+	if (rc)
+		return rc;
+
+	return selinux_netlbl_inode_permission(inode, mask);
 }
 
 static int selinux_file_alloc_security(struct file *file)
@@ -3058,9 +3066,10 @@ out:
 	return err;
 }
 
-static void selinux_socket_post_create(struct socket *sock, int family,
-				       int type, int protocol, int kern)
+static int selinux_socket_post_create(struct socket *sock, int family,
+				      int type, int protocol, int kern)
 {
+	int err = 0;
 	struct inode_security_struct *isec;
 	struct task_security_struct *tsec;
 	struct sk_security_struct *sksec;
@@ -3077,9 +3086,12 @@ static void selinux_socket_post_create(struct socket *sock, int family,
 	if (sock->sk) {
 		sksec = sock->sk->sk_security;
 		sksec->sid = isec->sid;
+		err = selinux_netlbl_socket_post_create(sock,
+							family,
+							isec->sid);
 	}
 
-	return;
+	return err;
 }
 
 /* Range of port numbers used to automatically bind.
@@ -3260,7 +3272,13 @@ static int selinux_socket_accept(struct socket *sock, struct socket *newsock)
 static int selinux_socket_sendmsg(struct socket *sock, struct msghdr *msg,
  				  int size)
 {
-	return socket_has_perm(current, sock, SOCKET__WRITE);
+	int rc;
+
+	rc = socket_has_perm(current, sock, SOCKET__WRITE);
+	if (rc)
+		return rc;
+
+	return selinux_netlbl_inode_permission(SOCK_INODE(sock), MAY_WRITE);
 }
 
 static int selinux_socket_recvmsg(struct socket *sock, struct msghdr *msg,
@@ -3468,6 +3486,10 @@ static int selinux_socket_sock_rcv_skb(struct sock *sk, struct sk_buff *skb)
 	if (err)
 		goto out;
 
+	err = selinux_netlbl_sock_rcv_skb(sksec, skb, &ad);
+	if (err)
+		goto out;
+
 	err = selinux_xfrm_sock_rcv_skb(sksec->sid, skb, &ad);
 out:	
 	return err;
@@ -3491,8 +3513,9 @@ static int selinux_socket_getpeersec_stream(struct socket *sock, char __user *op
 		peer_sid = ssec->peer_sid;
 	}
 	else if (isec->sclass == SECCLASS_TCP_SOCKET) {
-		peer_sid = selinux_socket_getpeer_stream(sock->sk);
-
+		peer_sid = selinux_netlbl_socket_getpeersec_stream(sock);
+		if (peer_sid == SECSID_NULL)
+			peer_sid = selinux_socket_getpeer_stream(sock->sk);
 		if (peer_sid == SECSID_NULL) {
 			err = -ENOPROTOOPT;
 			goto out;
@@ -3532,8 +3555,11 @@ static int selinux_socket_getpeersec_dgram(struct socket *sock, struct sk_buff *
 
 	if (sock && (sock->sk->sk_family == PF_UNIX))
 		selinux_get_inode_sid(SOCK_INODE(sock), &peer_secid);
-	else if (skb)
-		peer_secid = selinux_socket_getpeer_dgram(skb);
+	else if (skb) {
+		peer_secid = selinux_netlbl_socket_getpeersec_dgram(skb);
+		if (peer_secid == SECSID_NULL)
+			peer_secid = selinux_socket_getpeer_dgram(skb);
+	}
 
 	if (peer_secid == SECSID_NULL)
 		err = -EINVAL;
@@ -3578,6 +3604,8 @@ void selinux_sock_graft(struct sock* sk, struct socket *parent)
 	struct sk_security_struct *sksec = sk->sk_security;
 
 	isec->sid = sksec->sid;
+
+	selinux_netlbl_sock_graft(sk, parent);
 }
 
 int selinux_inet_conn_request(struct sock *sk, struct sk_buff *skb,
@@ -3585,9 +3613,15 @@ int selinux_inet_conn_request(struct sock *sk, struct sk_buff *skb,
 {
 	struct sk_security_struct *sksec = sk->sk_security;
 	int err;
-	u32 newsid = 0;
+	u32 newsid;
 	u32 peersid;
 
+	newsid = selinux_netlbl_inet_conn_request(skb, sksec->sid);
+	if (newsid != SECSID_NULL) {
+		req->secid = newsid;
+		return 0;
+	}
+
 	err = selinux_xfrm_decode_session(skb, &peersid, 0);
 	BUG_ON(err);
 
diff --git a/security/selinux/include/objsec.h b/security/selinux/include/objsec.h
index 79b9e0af19a0..0a39bfd1319f 100644
--- a/security/selinux/include/objsec.h
+++ b/security/selinux/include/objsec.h
@@ -101,6 +101,14 @@ struct sk_security_struct {
 	struct sock *sk;		/* back pointer to sk object */
 	u32 sid;			/* SID of this object */
 	u32 peer_sid;			/* SID of peer */
+#ifdef CONFIG_NETLABEL
+	u16 sclass;			/* sock security class */
+	enum {				/* NetLabel state */
+		NLBL_UNSET = 0,
+		NLBL_REQUIRE,
+		NLBL_LABELED,
+	} nlbl_state;
+#endif
 };
 
 struct key_security_struct {
diff --git a/security/selinux/include/selinux_netlabel.h b/security/selinux/include/selinux_netlabel.h
new file mode 100644
index 000000000000..88c463eef1e1
--- /dev/null
+++ b/security/selinux/include/selinux_netlabel.h
@@ -0,0 +1,125 @@
+/*
+ * SELinux interface to the NetLabel subsystem
+ *
+ * Author : Paul Moore <paul.moore@hp.com>
+ *
+ */
+
+/*
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ *
+ * This program is free software;  you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY;  without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program;  if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ */
+
+#ifndef _SELINUX_NETLABEL_H_
+#define _SELINUX_NETLABEL_H_
+
+#ifdef CONFIG_NETLABEL
+void selinux_netlbl_cache_invalidate(void);
+int selinux_netlbl_socket_post_create(struct socket *sock,
+				      int sock_family,
+				      u32 sid);
+void selinux_netlbl_sock_graft(struct sock *sk, struct socket *sock);
+u32 selinux_netlbl_inet_conn_request(struct sk_buff *skb, u32 sock_sid);
+int selinux_netlbl_sock_rcv_skb(struct sk_security_struct *sksec,
+				struct sk_buff *skb,
+				struct avc_audit_data *ad);
+u32 selinux_netlbl_socket_getpeersec_stream(struct socket *sock);
+u32 selinux_netlbl_socket_getpeersec_dgram(struct sk_buff *skb);
+
+int __selinux_netlbl_inode_permission(struct inode *inode, int mask);
+/**
+ * selinux_netlbl_inode_permission - Verify the socket is NetLabel labeled
+ * @inode: the file descriptor's inode
+ * @mask: the permission mask
+ *
+ * Description:
+ * Looks at a file's inode and if it is marked as a socket protected by
+ * NetLabel then verify that the socket has been labeled, if not try to label
+ * the socket now with the inode's SID.  Returns zero on success, negative
+ * values on failure.
+ *
+ */
+static inline int selinux_netlbl_inode_permission(struct inode *inode,
+						  int mask)
+{
+	int rc = 0;
+	struct inode_security_struct *isec;
+	struct sk_security_struct *sksec;
+
+	if (!S_ISSOCK(inode->i_mode))
+		return 0;
+
+	isec = inode->i_security;
+	sksec = SOCKET_I(inode)->sk->sk_security;
+	down(&isec->sem);
+	if (unlikely(sksec->nlbl_state == NLBL_REQUIRE &&
+		     (mask & (MAY_WRITE | MAY_APPEND))))
+		rc = __selinux_netlbl_inode_permission(inode, mask);
+	up(&isec->sem);
+
+	return rc;
+}
+#else
+static inline void selinux_netlbl_cache_invalidate(void)
+{
+	return;
+}
+
+static inline int selinux_netlbl_socket_post_create(struct socket *sock,
+						    int sock_family,
+						    u32 sid)
+{
+	return 0;
+}
+
+static inline void selinux_netlbl_sock_graft(struct sock *sk,
+					     struct socket *sock)
+{
+	return;
+}
+
+static inline u32 selinux_netlbl_inet_conn_request(struct sk_buff *skb,
+						   u32 sock_sid)
+{
+	return SECSID_NULL;
+}
+
+static inline int selinux_netlbl_sock_rcv_skb(struct sk_security_struct *sksec,
+					      struct sk_buff *skb,
+					      struct avc_audit_data *ad)
+{
+	return 0;
+}
+
+static inline u32 selinux_netlbl_socket_getpeersec_stream(struct socket *sock)
+{
+	return SECSID_NULL;
+}
+
+static inline u32 selinux_netlbl_socket_getpeersec_dgram(struct sk_buff *skb)
+{
+	return SECSID_NULL;
+}
+
+static inline int selinux_netlbl_inode_permission(struct inode *inode,
+						  int mask)
+{
+	return 0;
+}
+#endif /* CONFIG_NETLABEL */
+
+#endif
diff --git a/security/selinux/ss/ebitmap.c b/security/selinux/ss/ebitmap.c
index 47024a6e1844..4b915eb60c45 100644
--- a/security/selinux/ss/ebitmap.c
+++ b/security/selinux/ss/ebitmap.c
@@ -3,6 +3,14 @@
  *
  * Author : Stephen Smalley, <sds@epoch.ncsc.mil>
  */
+/*
+ * Updated: Hewlett-Packard <paul.moore@hp.com>
+ *
+ *      Added ebitmap_export() and ebitmap_import()
+ *
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ */
+
 #include <linux/kernel.h>
 #include <linux/slab.h>
 #include <linux/errno.h>
@@ -59,6 +67,142 @@ int ebitmap_cpy(struct ebitmap *dst, struct ebitmap *src)
 	return 0;
 }
 
+/**
+ * ebitmap_export - Export an ebitmap to a unsigned char bitmap string
+ * @src: the ebitmap to export
+ * @dst: the resulting bitmap string
+ * @dst_len: length of dst in bytes
+ *
+ * Description:
+ * Allocate a buffer at least src->highbit bits long and export the extensible
+ * bitmap into the buffer.  The bitmap string will be in little endian format,
+ * i.e. LSB first.  The value returned in dst_len may not the true size of the
+ * buffer as the length of the buffer is rounded up to a multiple of MAPTYPE.
+ * The caller must free the buffer when finished. Returns zero on success,
+ * negative values on failure.
+ *
+ */
+int ebitmap_export(const struct ebitmap *src,
+		   unsigned char **dst,
+		   size_t *dst_len)
+{
+	size_t bitmap_len;
+	unsigned char *bitmap;
+	struct ebitmap_node *iter_node;
+	MAPTYPE node_val;
+	size_t bitmap_byte;
+	unsigned char bitmask;
+
+	bitmap_len = src->highbit / 8;
+	if (src->highbit % 7)
+		bitmap_len += 1;
+	if (bitmap_len == 0)
+		return -EINVAL;
+
+	bitmap = kzalloc((bitmap_len & ~(sizeof(MAPTYPE) - 1)) +
+			 sizeof(MAPTYPE),
+			 GFP_ATOMIC);
+	if (bitmap == NULL)
+		return -ENOMEM;
+
+	iter_node = src->node;
+	do {
+		bitmap_byte = iter_node->startbit / 8;
+		bitmask = 0x80;
+		node_val = iter_node->map;
+		do {
+			if (bitmask == 0) {
+				bitmap_byte++;
+				bitmask = 0x80;
+			}
+			if (node_val & (MAPTYPE)0x01)
+				bitmap[bitmap_byte] |= bitmask;
+			node_val >>= 1;
+			bitmask >>= 1;
+		} while (node_val > 0);
+		iter_node = iter_node->next;
+	} while (iter_node);
+
+	*dst = bitmap;
+	*dst_len = bitmap_len;
+	return 0;
+}
+
+/**
+ * ebitmap_import - Import an unsigned char bitmap string into an ebitmap
+ * @src: the bitmap string
+ * @src_len: the bitmap length in bytes
+ * @dst: the empty ebitmap
+ *
+ * Description:
+ * This function takes a little endian bitmap string in src and imports it into
+ * the ebitmap pointed to by dst.  Returns zero on success, negative values on
+ * failure.
+ *
+ */
+int ebitmap_import(const unsigned char *src,
+		   size_t src_len,
+		   struct ebitmap *dst)
+{
+	size_t src_off = 0;
+	struct ebitmap_node *node_new;
+	struct ebitmap_node *node_last = NULL;
+	size_t iter;
+	size_t iter_bit;
+	size_t iter_limit;
+	unsigned char src_byte;
+
+	do {
+		iter_limit = src_len - src_off;
+		if (iter_limit >= sizeof(MAPTYPE)) {
+			if (*(MAPTYPE *)&src[src_off] == 0) {
+				src_off += sizeof(MAPTYPE);
+				continue;
+			}
+			iter_limit = sizeof(MAPTYPE);
+		} else {
+			iter = src_off;
+			src_byte = 0;
+			do {
+				src_byte |= src[iter++];
+			} while (iter < src_len && src_byte == 0);
+			if (src_byte == 0)
+				break;
+		}
+
+		node_new = kzalloc(sizeof(*node_new), GFP_ATOMIC);
+		if (unlikely(node_new == NULL)) {
+			ebitmap_destroy(dst);
+			return -ENOMEM;
+		}
+		node_new->startbit = src_off * 8;
+		iter = 0;
+		do {
+			src_byte = src[src_off++];
+			iter_bit = iter++ * 8;
+			while (src_byte != 0) {
+				if (src_byte & 0x80)
+					node_new->map |= MAPBIT << iter_bit;
+				iter_bit++;
+				src_byte <<= 1;
+			}
+		} while (iter < iter_limit);
+
+		if (node_last != NULL)
+			node_last->next = node_new;
+		else
+			dst->node = node_new;
+		node_last = node_new;
+	} while (src_off < src_len);
+
+	if (likely(node_last != NULL))
+		dst->highbit = node_last->startbit + MAPSIZE;
+	else
+		ebitmap_init(dst);
+
+	return 0;
+}
+
 int ebitmap_contains(struct ebitmap *e1, struct ebitmap *e2)
 {
 	struct ebitmap_node *n1, *n2;
diff --git a/security/selinux/ss/ebitmap.h b/security/selinux/ss/ebitmap.h
index 8bf41055a6cb..da2d4651b10d 100644
--- a/security/selinux/ss/ebitmap.h
+++ b/security/selinux/ss/ebitmap.h
@@ -69,6 +69,12 @@ static inline int ebitmap_node_get_bit(struct ebitmap_node * n,
 
 int ebitmap_cmp(struct ebitmap *e1, struct ebitmap *e2);
 int ebitmap_cpy(struct ebitmap *dst, struct ebitmap *src);
+int ebitmap_export(const struct ebitmap *src,
+		   unsigned char **dst,
+		   size_t *dst_len);
+int ebitmap_import(const unsigned char *src,
+		   size_t src_len,
+		   struct ebitmap *dst);
 int ebitmap_contains(struct ebitmap *e1, struct ebitmap *e2);
 int ebitmap_get_bit(struct ebitmap *e, unsigned long bit);
 int ebitmap_set_bit(struct ebitmap *e, unsigned long bit, int value);
diff --git a/security/selinux/ss/mls.c b/security/selinux/ss/mls.c
index e15f7e0399b8..119bd6078ba1 100644
--- a/security/selinux/ss/mls.c
+++ b/security/selinux/ss/mls.c
@@ -10,6 +10,13 @@
  *
  * Copyright (C) 2004-2006 Trusted Computer Solutions, Inc.
  */
+/*
+ * Updated: Hewlett-Packard <paul.moore@hp.com>
+ *
+ *      Added support to import/export the MLS label
+ *
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ */
 
 #include <linux/kernel.h>
 #include <linux/slab.h>
@@ -565,3 +572,152 @@ int mls_compute_sid(struct context *scontext,
 	return -EINVAL;
 }
 
+/**
+ * mls_export_lvl - Export the MLS sensitivity levels
+ * @context: the security context
+ * @low: the low sensitivity level
+ * @high: the high sensitivity level
+ *
+ * Description:
+ * Given the security context copy the low MLS sensitivity level into lvl_low
+ * and the high sensitivity level in lvl_high.  The MLS levels are only
+ * exported if the pointers are not NULL, if they are NULL then that level is
+ * not exported.
+ *
+ */
+void mls_export_lvl(const struct context *context, u32 *low, u32 *high)
+{
+	if (!selinux_mls_enabled)
+		return;
+
+	if (low != NULL)
+		*low = context->range.level[0].sens - 1;
+	if (high != NULL)
+		*high = context->range.level[1].sens - 1;
+}
+
+/**
+ * mls_import_lvl - Import the MLS sensitivity levels
+ * @context: the security context
+ * @low: the low sensitivity level
+ * @high: the high sensitivity level
+ *
+ * Description:
+ * Given the security context and the two sensitivty levels, set the MLS levels
+ * in the context according the two given as parameters.  Returns zero on
+ * success, negative values on failure.
+ *
+ */
+void mls_import_lvl(struct context *context, u32 low, u32 high)
+{
+	if (!selinux_mls_enabled)
+		return;
+
+	context->range.level[0].sens = low + 1;
+	context->range.level[1].sens = high + 1;
+}
+
+/**
+ * mls_export_cat - Export the MLS categories
+ * @context: the security context
+ * @low: the low category
+ * @low_len: length of the cat_low bitmap in bytes
+ * @high: the high category
+ * @high_len: length of the cat_high bitmap in bytes
+ *
+ * Description:
+ * Given the security context export the low MLS category bitmap into cat_low
+ * and the high category bitmap into cat_high.  The MLS categories are only
+ * exported if the pointers are not NULL, if they are NULL then that level is
+ * not exported.  The caller is responsibile for freeing the memory when
+ * finished.  Returns zero on success, negative values on failure.
+ *
+ */
+int mls_export_cat(const struct context *context,
+		   unsigned char **low,
+		   size_t *low_len,
+		   unsigned char **high,
+		   size_t *high_len)
+{
+	int rc = -EPERM;
+
+	if (!selinux_mls_enabled)
+		return 0;
+
+	if (low != NULL) {
+		rc = ebitmap_export(&context->range.level[0].cat,
+				    low,
+				    low_len);
+		if (rc != 0)
+			goto export_cat_failure;
+	}
+	if (high != NULL) {
+		rc = ebitmap_export(&context->range.level[1].cat,
+				    high,
+				    high_len);
+		if (rc != 0)
+			goto export_cat_failure;
+	}
+
+	return 0;
+
+export_cat_failure:
+	if (low != NULL)
+		kfree(*low);
+	if (high != NULL)
+		kfree(*high);
+	return rc;
+}
+
+/**
+ * mls_import_cat - Import the MLS categories
+ * @context: the security context
+ * @low: the low category
+ * @low_len: length of the cat_low bitmap in bytes
+ * @high: the high category
+ * @high_len: length of the cat_high bitmap in bytes
+ *
+ * Description:
+ * Given the security context and the two category bitmap strings import the
+ * categories into the security context.  The MLS categories are only imported
+ * if the pointers are not NULL, if they are NULL they are skipped.  Returns
+ * zero on success, negative values on failure.
+ *
+ */
+int mls_import_cat(struct context *context,
+		   const unsigned char *low,
+		   size_t low_len,
+		   const unsigned char *high,
+		   size_t high_len)
+{
+	int rc = -EPERM;
+
+	if (!selinux_mls_enabled)
+		return 0;
+
+	if (low != NULL) {
+		rc = ebitmap_import(low,
+				    low_len,
+				    &context->range.level[0].cat);
+		if (rc != 0)
+			goto import_cat_failure;
+	}
+	if (high != NULL) {
+		if (high == low)
+			rc = ebitmap_cpy(&context->range.level[1].cat,
+					 &context->range.level[0].cat);
+		else
+			rc = ebitmap_import(high,
+					    high_len,
+					    &context->range.level[1].cat);
+		if (rc != 0)
+			goto import_cat_failure;
+	}
+
+	return 0;
+
+import_cat_failure:
+	ebitmap_destroy(&context->range.level[0].cat);
+	ebitmap_destroy(&context->range.level[1].cat);
+	return rc;
+}
diff --git a/security/selinux/ss/mls.h b/security/selinux/ss/mls.h
index 90c5e88987fa..df6032c6d492 100644
--- a/security/selinux/ss/mls.h
+++ b/security/selinux/ss/mls.h
@@ -10,6 +10,13 @@
  *
  * Copyright (C) 2004-2006 Trusted Computer Solutions, Inc.
  */
+/*
+ * Updated: Hewlett-Packard <paul.moore@hp.com>
+ *
+ *      Added support to import/export the MLS label
+ *
+ * (c) Copyright Hewlett-Packard Development Company, L.P., 2006
+ */
 
 #ifndef _SS_MLS_H_
 #define _SS_MLS_H_
@@ -62,5 +69,19 @@ int mls_compute_sid(struct context *scontext,
 int mls_setup_user_range(struct context *fromcon, struct user_datum *user,
                          struct context *usercon);
 
+void mls_export_lvl(const struct context *context, u32 *low, u32 *high);
+void mls_import_lvl(struct context *context, u32 low, u32 high);
+
+int mls_export_cat(const struct context *context,
+		   unsigned char **low,
+		   size_t *low_len,
+		   unsigned char **high,
+		   size_t *high_len);
+int mls_import_cat(struct context *context,
+		   const unsigned char *low,
+		   size_t low_len,
+		   const unsigned char *high,
+		   size_t high_len);
+
 #endif	/* _SS_MLS_H */
 
diff --git a/security/selinux/ss/services.c b/security/selinux/ss/services.c
index b00ec69f0ffd..910afa1ffc31 100644
--- a/security/selinux/ss/services.c
+++ b/security/selinux/ss/services.c
@@ -13,6 +13,11 @@
  *
  * 	Added conditional policy language extensions
  *
+ * Updated: Hewlett-Packard <paul.moore@hp.com>
+ *
+ *      Added support for NetLabel
+ *
+ * Copyright (C) 2006 Hewlett-Packard Development Company, L.P.
  * Copyright (C) 2004-2006 Trusted Computer Solutions, Inc.
  * Copyright (C) 2003 - 2004 Tresys Technology, LLC
  * Copyright (C) 2003 Red Hat, Inc., James Morris <jmorris@redhat.com>
@@ -29,6 +34,8 @@
 #include <linux/sched.h>
 #include <linux/audit.h>
 #include <linux/mutex.h>
+#include <net/sock.h>
+#include <net/netlabel.h>
 
 #include "flask.h"
 #include "avc.h"
@@ -40,6 +47,8 @@
 #include "services.h"
 #include "conditional.h"
 #include "mls.h"
+#include "objsec.h"
+#include "selinux_netlabel.h"
 
 extern void selnl_notify_policyload(u32 seqno);
 unsigned int policydb_loaded_version;
@@ -1241,6 +1250,7 @@ int security_load_policy(void *data, size_t len)
 		selinux_complete_init();
 		avc_ss_reset(seqno);
 		selnl_notify_policyload(seqno);
+		selinux_netlbl_cache_invalidate();
 		return 0;
 	}
 
@@ -1295,6 +1305,7 @@ int security_load_policy(void *data, size_t len)
 
 	avc_ss_reset(seqno);
 	selnl_notify_policyload(seqno);
+	selinux_netlbl_cache_invalidate();
 
 	return 0;
 
@@ -2133,3 +2144,480 @@ void selinux_audit_set_callback(int (*callback)(void))
 {
 	aurule_callback = callback;
 }
+
+#ifdef CONFIG_NETLABEL
+/*
+ * This is the structure we store inside the NetLabel cache block.
+ */
+#define NETLBL_CACHE(x)           ((struct netlbl_cache *)(x))
+#define NETLBL_CACHE_T_NONE       0
+#define NETLBL_CACHE_T_SID        1
+#define NETLBL_CACHE_T_MLS        2
+struct netlbl_cache {
+	u32 type;
+	union {
+		u32 sid;
+		struct mls_range mls_label;
+	} data;
+};
+
+/**
+ * selinux_netlbl_cache_free - Free the NetLabel cached data
+ * @data: the data to free
+ *
+ * Description:
+ * This function is intended to be used as the free() callback inside the
+ * netlbl_lsm_cache structure.
+ *
+ */
+static void selinux_netlbl_cache_free(const void *data)
+{
+	struct netlbl_cache *cache = NETLBL_CACHE(data);
+	switch (cache->type) {
+	case NETLBL_CACHE_T_MLS:
+		ebitmap_destroy(&cache->data.mls_label.level[0].cat);
+		break;
+	}
+	kfree(data);
+}
+
+/**
+ * selinux_netlbl_cache_add - Add an entry to the NetLabel cache
+ * @skb: the packet
+ * @ctx: the SELinux context
+ *
+ * Description:
+ * Attempt to cache the context in @ctx, which was derived from the packet in
+ * @skb, in the NetLabel subsystem cache.
+ *
+ */
+static void selinux_netlbl_cache_add(struct sk_buff *skb, struct context *ctx)
+{
+	struct netlbl_cache *cache = NULL;
+	struct netlbl_lsm_secattr secattr;
+
+	netlbl_secattr_init(&secattr);
+
+	cache = kzalloc(sizeof(*cache),	GFP_ATOMIC);
+	if (cache == NULL)
+		goto netlbl_cache_add_failure;
+	secattr.cache.free = selinux_netlbl_cache_free;
+	secattr.cache.data = (void *)cache;
+
+	cache->type = NETLBL_CACHE_T_MLS;
+	if (ebitmap_cpy(&cache->data.mls_label.level[0].cat,
+			&ctx->range.level[0].cat) != 0)
+		goto netlbl_cache_add_failure;
+	cache->data.mls_label.level[1].cat.highbit =
+		cache->data.mls_label.level[0].cat.highbit;
+	cache->data.mls_label.level[1].cat.node =
+		cache->data.mls_label.level[0].cat.node;
+	cache->data.mls_label.level[0].sens = ctx->range.level[0].sens;
+	cache->data.mls_label.level[1].sens = ctx->range.level[0].sens;
+
+	if (netlbl_cache_add(skb, &secattr) != 0)
+		goto netlbl_cache_add_failure;
+
+	return;
+
+netlbl_cache_add_failure:
+	netlbl_secattr_destroy(&secattr, 1);
+}
+
+/**
+ * selinux_netlbl_cache_invalidate - Invalidate the NetLabel cache
+ *
+ * Description:
+ * Invalidate the NetLabel security attribute mapping cache.
+ *
+ */
+void selinux_netlbl_cache_invalidate(void)
+{
+	netlbl_cache_invalidate();
+}
+
+/**
+ * selinux_netlbl_secattr_to_sid - Convert a NetLabel secattr to a SELinux SID
+ * @skb: the network packet
+ * @secattr: the NetLabel packet security attributes
+ * @base_sid: the SELinux SID to use as a context for MLS only attributes
+ * @sid: the SELinux SID
+ *
+ * Description:
+ * Convert the given NetLabel packet security attributes in @secattr into a
+ * SELinux SID.  If the @secattr field does not contain a full SELinux
+ * SID/context then use the context in @base_sid as the foundation.  If @skb
+ * is not NULL attempt to cache as much data as possibile.  Returns zero on
+ * success, negative values on failure.
+ *
+ */
+static int selinux_netlbl_secattr_to_sid(struct sk_buff *skb,
+					 struct netlbl_lsm_secattr *secattr,
+					 u32 base_sid,
+					 u32 *sid)
+{
+	int rc = -EIDRM;
+	struct context *ctx;
+	struct context ctx_new;
+	struct netlbl_cache *cache;
+
+	POLICY_RDLOCK;
+
+	if (secattr->cache.data) {
+		cache = NETLBL_CACHE(secattr->cache.data);
+		switch (cache->type) {
+		case NETLBL_CACHE_T_SID:
+			*sid = cache->data.sid;
+			rc = 0;
+			break;
+		case NETLBL_CACHE_T_MLS:
+			ctx = sidtab_search(&sidtab, base_sid);
+			if (ctx == NULL)
+				goto netlbl_secattr_to_sid_return;
+
+			ctx_new.user = ctx->user;
+			ctx_new.role = ctx->role;
+			ctx_new.type = ctx->type;
+			ctx_new.range.level[0].sens =
+				cache->data.mls_label.level[0].sens;
+			ctx_new.range.level[0].cat.highbit =
+				cache->data.mls_label.level[0].cat.highbit;
+			ctx_new.range.level[0].cat.node =
+				cache->data.mls_label.level[0].cat.node;
+			ctx_new.range.level[1].sens =
+				cache->data.mls_label.level[1].sens;
+			ctx_new.range.level[1].cat.highbit =
+				cache->data.mls_label.level[1].cat.highbit;
+			ctx_new.range.level[1].cat.node =
+				cache->data.mls_label.level[1].cat.node;
+
+			rc = sidtab_context_to_sid(&sidtab, &ctx_new, sid);
+			break;
+		default:
+			goto netlbl_secattr_to_sid_return;
+		}
+	} else if (secattr->mls_lvl_vld) {
+		ctx = sidtab_search(&sidtab, base_sid);
+		if (ctx == NULL)
+			goto netlbl_secattr_to_sid_return;
+
+		ctx_new.user = ctx->user;
+		ctx_new.role = ctx->role;
+		ctx_new.type = ctx->type;
+		mls_import_lvl(&ctx_new, secattr->mls_lvl, secattr->mls_lvl);
+		if (secattr->mls_cat) {
+			if (mls_import_cat(&ctx_new,
+					   secattr->mls_cat,
+					   secattr->mls_cat_len,
+					   NULL,
+					   0) != 0)
+				goto netlbl_secattr_to_sid_return;
+			ctx_new.range.level[1].cat.highbit =
+				ctx_new.range.level[0].cat.highbit;
+			ctx_new.range.level[1].cat.node =
+				ctx_new.range.level[0].cat.node;
+		} else {
+			ebitmap_init(&ctx_new.range.level[0].cat);
+			ebitmap_init(&ctx_new.range.level[1].cat);
+		}
+		if (mls_context_isvalid(&policydb, &ctx_new) != 1)
+			goto netlbl_secattr_to_sid_return_cleanup;
+
+		rc = sidtab_context_to_sid(&sidtab, &ctx_new, sid);
+		if (rc != 0)
+			goto netlbl_secattr_to_sid_return_cleanup;
+
+		if (skb != NULL)
+			selinux_netlbl_cache_add(skb, &ctx_new);
+		ebitmap_destroy(&ctx_new.range.level[0].cat);
+	} else {
+		*sid = SECINITSID_UNLABELED;
+		rc = 0;
+	}
+
+netlbl_secattr_to_sid_return:
+	POLICY_RDUNLOCK;
+	return rc;
+netlbl_secattr_to_sid_return_cleanup:
+	ebitmap_destroy(&ctx_new.range.level[0].cat);
+	goto netlbl_secattr_to_sid_return;
+}
+
+/**
+ * selinux_netlbl_skbuff_getsid - Get the sid of a packet using NetLabel
+ * @skb: the packet
+ * @base_sid: the SELinux SID to use as a context for MLS only attributes
+ * @sid: the SID
+ *
+ * Description:
+ * Call the NetLabel mechanism to get the security attributes of the given
+ * packet and use those attributes to determine the correct context/SID to
+ * assign to the packet.  Returns zero on success, negative values on failure.
+ *
+ */
+static int selinux_netlbl_skbuff_getsid(struct sk_buff *skb,
+					u32 base_sid,
+					u32 *sid)
+{
+	int rc;
+	struct netlbl_lsm_secattr secattr;
+
+	netlbl_secattr_init(&secattr);
+	rc = netlbl_skbuff_getattr(skb, &secattr);
+	if (rc == 0)
+		rc = selinux_netlbl_secattr_to_sid(skb,
+						   &secattr,
+						   base_sid,
+						   sid);
+	netlbl_secattr_destroy(&secattr, 0);
+
+	return rc;
+}
+
+/**
+ * selinux_netlbl_socket_setsid - Label a socket using the NetLabel mechanism
+ * @sock: the socket to label
+ * @sid: the SID to use
+ *
+ * Description:
+ * Attempt to label a socket using the NetLabel mechanism using the given
+ * SID.  Returns zero values on success, negative values on failure.
+ *
+ */
+static int selinux_netlbl_socket_setsid(struct socket *sock, u32 sid)
+{
+	int rc = -ENOENT;
+	struct sk_security_struct *sksec = sock->sk->sk_security;
+	struct netlbl_lsm_secattr secattr;
+	struct context *ctx;
+
+	if (!ss_initialized)
+		return 0;
+
+	POLICY_RDLOCK;
+
+	ctx = sidtab_search(&sidtab, sid);
+	if (ctx == NULL)
+		goto netlbl_socket_setsid_return;
+
+	netlbl_secattr_init(&secattr);
+	secattr.domain = kstrdup(policydb.p_type_val_to_name[ctx->type - 1],
+				 GFP_ATOMIC);
+	mls_export_lvl(ctx, &secattr.mls_lvl, NULL);
+	secattr.mls_lvl_vld = 1;
+	mls_export_cat(ctx,
+		       &secattr.mls_cat,
+		       &secattr.mls_cat_len,
+		       NULL,
+		       NULL);
+
+	rc = netlbl_socket_setattr(sock, &secattr);
+	if (rc == 0)
+		sksec->nlbl_state = NLBL_LABELED;
+
+	netlbl_secattr_destroy(&secattr, 0);
+
+netlbl_socket_setsid_return:
+	POLICY_RDUNLOCK;
+	return rc;
+}
+
+/**
+ * selinux_netlbl_socket_post_create - Label a socket using NetLabel
+ * @sock: the socket to label
+ * @sock_family: the socket family
+ * @sid: the SID to use
+ *
+ * Description:
+ * Attempt to label a socket using the NetLabel mechanism using the given
+ * SID.  Returns zero values on success, negative values on failure.
+ *
+ */
+int selinux_netlbl_socket_post_create(struct socket *sock,
+				      int sock_family,
+				      u32 sid)
+{
+	struct inode_security_struct *isec = SOCK_INODE(sock)->i_security;
+	struct sk_security_struct *sksec = sock->sk->sk_security;
+
+	if (sock_family != PF_INET)
+		return 0;
+
+	sksec->sclass = isec->sclass;
+	sksec->nlbl_state = NLBL_REQUIRE;
+	return selinux_netlbl_socket_setsid(sock, sid);
+}
+
+/**
+ * selinux_netlbl_sock_graft - Netlabel the new socket
+ * @sk: the new connection
+ * @sock: the new socket
+ *
+ * Description:
+ * The connection represented by @sk is being grafted onto @sock so set the
+ * socket's NetLabel to match the SID of @sk.
+ *
+ */
+void selinux_netlbl_sock_graft(struct sock *sk, struct socket *sock)
+{
+	struct inode_security_struct *isec = SOCK_INODE(sock)->i_security;
+	struct sk_security_struct *sksec = sk->sk_security;
+
+	if (sk->sk_family != PF_INET)
+		return;
+
+	sksec->nlbl_state = NLBL_REQUIRE;
+	sksec->peer_sid = sksec->sid;
+	sksec->sclass = isec->sclass;
+
+	/* Try to set the NetLabel on the socket to save time later, if we fail
+	 * here we will pick up the pieces in later calls to
+	 * selinux_netlbl_inode_permission(). */
+	selinux_netlbl_socket_setsid(sock, sksec->sid);
+}
+
+/**
+ * selinux_netlbl_inet_conn_request - Handle a new connection request
+ * @skb: the packet
+ * @sock_sid: the SID of the parent socket
+ *
+ * Description:
+ * If present, use the security attributes of the packet in @skb and the
+ * parent sock's SID to arrive at a SID for the new child sock.  Returns the
+ * SID of the connection or SECSID_NULL on failure.
+ *
+ */
+u32 selinux_netlbl_inet_conn_request(struct sk_buff *skb, u32 sock_sid)
+{
+	int rc;
+	u32 peer_sid;
+
+	rc = selinux_netlbl_skbuff_getsid(skb, sock_sid, &peer_sid);
+	if (rc != 0)
+		return SECSID_NULL;
+
+	if (peer_sid == SECINITSID_UNLABELED)
+		return SECSID_NULL;
+
+	return peer_sid;
+}
+
+/**
+ * __selinux_netlbl_inode_permission - Label a socket using NetLabel
+ * @inode: the file descriptor's inode
+ * @mask: the permission mask
+ *
+ * Description:
+ * Try to label a socket with the inode's SID using NetLabel.  Returns zero on
+ * success, negative values on failure.
+ *
+ */
+int __selinux_netlbl_inode_permission(struct inode *inode, int mask)
+{
+	int rc;
+	struct socket *sock = SOCKET_I(inode);
+	struct sk_security_struct *sksec = sock->sk->sk_security;
+
+	lock_sock(sock->sk);
+	rc = selinux_netlbl_socket_setsid(sock, sksec->sid);
+	release_sock(sock->sk);
+
+	return rc;
+}
+
+/**
+ * selinux_netlbl_sock_rcv_skb - Do an inbound access check using NetLabel
+ * @sksec: the sock's sk_security_struct
+ * @skb: the packet
+ * @ad: the audit data
+ *
+ * Description:
+ * Fetch the NetLabel security attributes from @skb and perform an access check
+ * against the receiving socket.  Returns zero on success, negative values on
+ * error.
+ *
+ */
+int selinux_netlbl_sock_rcv_skb(struct sk_security_struct *sksec,
+				struct sk_buff *skb,
+				struct avc_audit_data *ad)
+{
+	int rc;
+	u32 netlbl_sid;
+	u32 recv_perm;
+
+	rc = selinux_netlbl_skbuff_getsid(skb, sksec->sid, &netlbl_sid);
+	if (rc != 0)
+		return rc;
+
+	if (netlbl_sid == SECINITSID_UNLABELED)
+		return 0;
+
+	switch (sksec->sclass) {
+	case SECCLASS_UDP_SOCKET:
+		recv_perm = UDP_SOCKET__RECV_MSG;
+		break;
+	case SECCLASS_TCP_SOCKET:
+		recv_perm = TCP_SOCKET__RECV_MSG;
+		break;
+	default:
+		recv_perm = RAWIP_SOCKET__RECV_MSG;
+	}
+
+	rc = avc_has_perm(sksec->sid,
+			  netlbl_sid,
+			  sksec->sclass,
+			  recv_perm,
+			  ad);
+	if (rc == 0)
+		return 0;
+
+	netlbl_skbuff_err(skb, rc);
+	return rc;
+}
+
+/**
+ * selinux_netlbl_socket_peersid - Return the peer SID of a connected socket
+ * @sock: the socket
+ *
+ * Description:
+ * Examine @sock to find the connected peer's SID.  Returns the SID on success
+ * or SECSID_NULL on error.
+ *
+ */
+u32 selinux_netlbl_socket_getpeersec_stream(struct socket *sock)
+{
+	struct sk_security_struct *sksec = sock->sk->sk_security;
+
+	if (sksec->peer_sid == SECINITSID_UNLABELED)
+		return SECSID_NULL;
+
+	return sksec->peer_sid;
+}
+
+/**
+ * selinux_netlbl_socket_getpeersec_dgram - Return the SID of a NetLabel packet
+ * @skb: the packet
+ *
+ * Description:
+ * Examine @skb to find the SID assigned to it by NetLabel.  Returns the SID on
+ * success, SECSID_NULL on error.
+ *
+ */
+u32 selinux_netlbl_socket_getpeersec_dgram(struct sk_buff *skb)
+{
+	int peer_sid;
+	struct sock *sk = skb->sk;
+	struct inode_security_struct *isec;
+
+	if (sk == NULL || sk->sk_socket == NULL)
+		return SECSID_NULL;
+
+	isec = SOCK_INODE(sk->sk_socket)->i_security;
+	if (selinux_netlbl_skbuff_getsid(skb, isec->sid, &peer_sid) != 0)
+		return SECSID_NULL;
+	if (peer_sid == SECINITSID_UNLABELED)
+		return SECSID_NULL;
+
+	return peer_sid;
+}
+#endif /* CONFIG_NETLABEL */
-- 
cgit v1.2.3


From c71099acce933455123ee505cc75964610a209ad Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Fri, 4 Aug 2006 23:20:06 -0700
Subject: [IPV6]: Multiple Routing Tables

Adds the framework to support multiple IPv6 routing tables.
Currently all automatically generated routes are put into the
same table. This could be changed at a later point after
considering the produced locking overhead.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip6_fib.h   |  39 ++++-
 include/net/ip6_route.h |   3 +-
 net/ipv6/Kconfig        |   6 +
 net/ipv6/addrconf.c     |   6 +-
 net/ipv6/ip6_fib.c      | 144 +++++++++++++++++-
 net/ipv6/route.c        | 380 ++++++++++++++++++++++++++++++++----------------
 6 files changed, 441 insertions(+), 137 deletions(-)

(limited to 'include')

diff --git a/include/net/ip6_fib.h b/include/net/ip6_fib.h
index a66e9de16a6c..818411519c89 100644
--- a/include/net/ip6_fib.h
+++ b/include/net/ip6_fib.h
@@ -51,6 +51,8 @@ struct rt6key
 	int		plen;
 };
 
+struct fib6_table;
+
 struct rt6_info
 {
 	union {
@@ -71,6 +73,7 @@ struct rt6_info
 	u32				rt6i_flags;
 	u32				rt6i_metric;
 	atomic_t			rt6i_ref;
+	struct fib6_table		*rt6i_table;
 
 	struct rt6key			rt6i_dst;
 	struct rt6key			rt6i_src;
@@ -143,12 +146,43 @@ struct rt6_statistics {
 
 typedef void			(*f_pnode)(struct fib6_node *fn, void *);
 
-extern struct fib6_node		ip6_routing_table;
+struct fib6_table {
+	struct hlist_node	tb6_hlist;
+	u32			tb6_id;
+	rwlock_t		tb6_lock;
+	struct fib6_node	tb6_root;
+};
+
+#define RT6_TABLE_UNSPEC	RT_TABLE_UNSPEC
+#define RT6_TABLE_MAIN		RT_TABLE_MAIN
+#define RT6_TABLE_LOCAL		RT6_TABLE_MAIN
+#define RT6_TABLE_DFLT		RT6_TABLE_MAIN
+#define RT6_TABLE_INFO		RT6_TABLE_MAIN
+#define RT6_TABLE_PREFIX	RT6_TABLE_MAIN
+
+#ifdef CONFIG_IPV6_MULTIPLE_TABLES
+#define FIB6_TABLE_MIN		1
+#define FIB6_TABLE_MAX		RT_TABLE_MAX
+#else
+#define FIB6_TABLE_MIN		RT_TABLE_MAIN
+#define FIB6_TABLE_MAX		FIB6_TABLE_MIN
+#endif
+
+#define RT6_F_STRICT		1
+#define RT6_F_HAS_SADDR		2
+
+typedef struct rt6_info *(*pol_lookup_t)(struct fib6_table *,
+					 struct flowi *, int);
 
 /*
  *	exported functions
  */
 
+extern struct fib6_table *	fib6_get_table(u32 id);
+extern struct fib6_table *	fib6_new_table(u32 id);
+extern struct dst_entry *	fib6_rule_lookup(struct flowi *fl, int flags,
+						 pol_lookup_t lookup);
+
 extern struct fib6_node		*fib6_lookup(struct fib6_node *root,
 					     struct in6_addr *daddr,
 					     struct in6_addr *saddr);
@@ -161,6 +195,9 @@ extern void			fib6_clean_tree(struct fib6_node *root,
 						int (*func)(struct rt6_info *, void *arg),
 						int prune, void *arg);
 
+extern void			fib6_clean_all(int (*func)(struct rt6_info *, void *arg),
+					       int prune, void *arg);
+
 extern int			fib6_walk(struct fib6_walker_t *w);
 extern int			fib6_walk_continue(struct fib6_walker_t *w);
 
diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index 96b0e66406ec..d49c8c90eb68 100644
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -58,7 +58,8 @@ extern int			ipv6_route_ioctl(unsigned int cmd, void __user *arg);
 extern int			ip6_route_add(struct in6_rtmsg *rtmsg,
 					      struct nlmsghdr *,
 					      void *rtattr,
-					      struct netlink_skb_parms *req);
+					      struct netlink_skb_parms *req,
+					      u32 table_id);
 extern int			ip6_ins_rt(struct rt6_info *,
 					   struct nlmsghdr *,
 					   void *rtattr,
diff --git a/net/ipv6/Kconfig b/net/ipv6/Kconfig
index 0ba06c0c5d39..159c63d99c81 100644
--- a/net/ipv6/Kconfig
+++ b/net/ipv6/Kconfig
@@ -136,3 +136,9 @@ config IPV6_TUNNEL
 
 	  If unsure, say N.
 
+config IPV6_MULTIPLE_TABLES
+	bool "IPv6: Multiple Routing Tables"
+	depends on IPV6 && EXPERIMENTAL
+	---help---
+	  Support multiple routing tables.
+
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index c7852b38e03e..318767fcefdc 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -1525,7 +1525,7 @@ addrconf_prefix_route(struct in6_addr *pfx, int plen, struct net_device *dev,
 	if (dev->type == ARPHRD_SIT && (dev->flags&IFF_POINTOPOINT))
 		rtmsg.rtmsg_flags |= RTF_NONEXTHOP;
 
-	ip6_route_add(&rtmsg, NULL, NULL, NULL);
+	ip6_route_add(&rtmsg, NULL, NULL, NULL, RT6_TABLE_PREFIX);
 }
 
 /* Create "default" multicast route to the interface */
@@ -1542,7 +1542,7 @@ static void addrconf_add_mroute(struct net_device *dev)
 	rtmsg.rtmsg_ifindex = dev->ifindex;
 	rtmsg.rtmsg_flags = RTF_UP;
 	rtmsg.rtmsg_type = RTMSG_NEWROUTE;
-	ip6_route_add(&rtmsg, NULL, NULL, NULL);
+	ip6_route_add(&rtmsg, NULL, NULL, NULL, RT6_TABLE_LOCAL);
 }
 
 static void sit_route_add(struct net_device *dev)
@@ -1559,7 +1559,7 @@ static void sit_route_add(struct net_device *dev)
 	rtmsg.rtmsg_flags	= RTF_UP|RTF_NONEXTHOP;
 	rtmsg.rtmsg_ifindex	= dev->ifindex;
 
-	ip6_route_add(&rtmsg, NULL, NULL, NULL);
+	ip6_route_add(&rtmsg, NULL, NULL, NULL, RT6_TABLE_MAIN);
 }
 
 static void addrconf_add_lroute(struct net_device *dev)
diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
index 764221220afd..fcd7da830aca 100644
--- a/net/ipv6/ip6_fib.c
+++ b/net/ipv6/ip6_fib.c
@@ -26,6 +26,7 @@
 #include <linux/netdevice.h>
 #include <linux/in6.h>
 #include <linux/init.h>
+#include <linux/list.h>
 
 #ifdef 	CONFIG_PROC_FS
 #include <linux/proc_fs.h>
@@ -147,6 +148,126 @@ static __inline__ void rt6_release(struct rt6_info *rt)
 		dst_free(&rt->u.dst);
 }
 
+static struct fib6_table fib6_main_tbl = {
+	.tb6_id		= RT6_TABLE_MAIN,
+	.tb6_lock	= RW_LOCK_UNLOCKED,
+	.tb6_root	= {
+		.leaf		= &ip6_null_entry,
+		.fn_flags	= RTN_ROOT | RTN_TL_ROOT | RTN_RTINFO,
+	},
+};
+
+#ifdef CONFIG_IPV6_MULTIPLE_TABLES
+
+#define FIB_TABLE_HASHSZ 256
+static struct hlist_head fib_table_hash[FIB_TABLE_HASHSZ];
+
+static struct fib6_table *fib6_alloc_table(u32 id)
+{
+	struct fib6_table *table;
+
+	table = kzalloc(sizeof(*table), GFP_ATOMIC);
+	if (table != NULL) {
+		table->tb6_id = id;
+		table->tb6_lock = RW_LOCK_UNLOCKED;
+		table->tb6_root.leaf = &ip6_null_entry;
+		table->tb6_root.fn_flags = RTN_ROOT | RTN_TL_ROOT | RTN_RTINFO;
+	}
+
+	return table;
+}
+
+static void fib6_link_table(struct fib6_table *tb)
+{
+	unsigned int h;
+
+	h = tb->tb6_id & (FIB_TABLE_HASHSZ - 1);
+
+	/*
+	 * No protection necessary, this is the only list mutatation
+	 * operation, tables never disappear once they exist.
+	 */
+	hlist_add_head_rcu(&tb->tb6_hlist, &fib_table_hash[h]);
+}
+
+struct fib6_table *fib6_new_table(u32 id)
+{
+	struct fib6_table *tb;
+
+	if (id == 0)
+		id = RT6_TABLE_MAIN;
+	tb = fib6_get_table(id);
+	if (tb)
+		return tb;
+
+	tb = fib6_alloc_table(id);
+	if (tb != NULL)
+		fib6_link_table(tb);
+
+	return tb;
+}
+
+struct fib6_table *fib6_get_table(u32 id)
+{
+	struct fib6_table *tb;
+	struct hlist_node *node;
+	unsigned int h;
+
+	if (id == 0)
+		id = RT6_TABLE_MAIN;
+	h = id & (FIB_TABLE_HASHSZ - 1);
+	rcu_read_lock();
+	hlist_for_each_entry_rcu(tb, node, &fib_table_hash[h], tb6_hlist) {
+		if (tb->tb6_id == id) {
+			rcu_read_unlock();
+			return tb;
+		}
+	}
+	rcu_read_unlock();
+
+	return NULL;
+}
+
+struct dst_entry *fib6_rule_lookup(struct flowi *fl, int flags,
+				   pol_lookup_t lookup)
+{
+	/*
+	 * TODO: Add rule lookup
+	 */
+	struct fib6_table *table = fib6_get_table(RT6_TABLE_MAIN);
+
+	return (struct dst_entry *) lookup(table, fl, flags);
+}
+
+static void __init fib6_tables_init(void)
+{
+	fib6_link_table(&fib6_main_tbl);
+}
+
+#else
+
+struct fib6_table *fib6_new_table(u32 id)
+{
+	return fib6_get_table(id);
+}
+
+struct fib6_table *fib6_get_table(u32 id)
+{
+	return &fib6_main_tbl;
+}
+
+struct dst_entry *fib6_rule_lookup(struct flowi *fl, int flags,
+				   pol_lookup_t lookup)
+{
+	return (struct dst_entry *) lookup(&fib6_main_tbl, fl, flags);
+}
+
+static void __init fib6_tables_init(void)
+{
+}
+
+#endif
+
 
 /*
  *	Routing Table
@@ -1064,6 +1185,22 @@ void fib6_clean_tree(struct fib6_node *root,
 	fib6_walk(&c.w);
 }
 
+void fib6_clean_all(int (*func)(struct rt6_info *, void *arg),
+		    int prune, void *arg)
+{
+	int i;
+	struct fib6_table *table;
+
+	for (i = FIB6_TABLE_MIN; i <= FIB6_TABLE_MAX; i++) {
+		table = fib6_get_table(i);
+		if (table != NULL) {
+			write_lock_bh(&table->tb6_lock);
+			fib6_clean_tree(&table->tb6_root, func, prune, arg);
+			write_unlock_bh(&table->tb6_lock);
+		}
+	}
+}
+
 static int fib6_prune_clone(struct rt6_info *rt, void *arg)
 {
 	if (rt->rt6i_flags & RTF_CACHE) {
@@ -1142,11 +1279,8 @@ void fib6_run_gc(unsigned long dummy)
 	}
 	gc_args.more = 0;
 
-
-	write_lock_bh(&rt6_lock);
 	ndisc_dst_gc(&gc_args.more);
-	fib6_clean_tree(&ip6_routing_table, fib6_age, 0, NULL);
-	write_unlock_bh(&rt6_lock);
+	fib6_clean_all(fib6_age, 0, NULL);
 
 	if (gc_args.more)
 		mod_timer(&ip6_fib_timer, jiffies + ip6_rt_gc_interval);
@@ -1165,6 +1299,8 @@ void __init fib6_init(void)
 					   NULL, NULL);
 	if (!fib6_node_kmem)
 		panic("cannot create fib6_nodes cache");
+
+	fib6_tables_init();
 }
 
 void fib6_gc_cleanup(void)
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index ce1f49b595b0..73efdadb9ab8 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -140,16 +140,6 @@ struct rt6_info ip6_null_entry = {
 	.rt6i_ref	= ATOMIC_INIT(1),
 };
 
-struct fib6_node ip6_routing_table = {
-	.leaf		= &ip6_null_entry,
-	.fn_flags	= RTN_ROOT | RTN_TL_ROOT | RTN_RTINFO,
-};
-
-/* Protects all the ip6 fib */
-
-DEFINE_RWLOCK(rt6_lock);
-
-
 /* allocate dst with ip6_dst_ops */
 static __inline__ struct rt6_info *ip6_dst_alloc(void)
 {
@@ -188,8 +178,14 @@ static __inline__ int rt6_check_expired(const struct rt6_info *rt)
 		time_after(jiffies, rt->rt6i_expires));
 }
 
+static inline int rt6_need_strict(struct in6_addr *daddr)
+{
+	return (ipv6_addr_type(daddr) &
+		(IPV6_ADDR_MULTICAST | IPV6_ADDR_LINKLOCAL));
+}
+
 /*
- *	Route lookup. Any rt6_lock is implied.
+ *	Route lookup. Any table->tb6_lock is implied.
  */
 
 static __inline__ struct rt6_info *rt6_device_match(struct rt6_info *rt,
@@ -441,27 +437,66 @@ int rt6_route_rcv(struct net_device *dev, u8 *opt, int len,
 }
 #endif
 
-struct rt6_info *rt6_lookup(struct in6_addr *daddr, struct in6_addr *saddr,
-			    int oif, int strict)
+#define BACKTRACK() \
+if (rt == &ip6_null_entry && flags & RT6_F_STRICT) { \
+	while ((fn = fn->parent) != NULL) { \
+		if (fn->fn_flags & RTN_TL_ROOT) { \
+			dst_hold(&rt->u.dst); \
+			goto out; \
+		} \
+		if (fn->fn_flags & RTN_RTINFO) \
+			goto restart; \
+	} \
+}
+
+static struct rt6_info *ip6_pol_route_lookup(struct fib6_table *table,
+					     struct flowi *fl, int flags)
 {
 	struct fib6_node *fn;
 	struct rt6_info *rt;
 
-	read_lock_bh(&rt6_lock);
-	fn = fib6_lookup(&ip6_routing_table, daddr, saddr);
-	rt = rt6_device_match(fn->leaf, oif, strict);
+	read_lock_bh(&table->tb6_lock);
+	fn = fib6_lookup(&table->tb6_root, &fl->fl6_dst, &fl->fl6_src);
+restart:
+	rt = fn->leaf;
+	rt = rt6_device_match(rt, fl->oif, flags & RT6_F_STRICT);
+	BACKTRACK();
 	dst_hold(&rt->u.dst);
-	rt->u.dst.__use++;
-	read_unlock_bh(&rt6_lock);
+out:
+	read_unlock_bh(&table->tb6_lock);
 
 	rt->u.dst.lastuse = jiffies;
-	if (rt->u.dst.error == 0)
-		return rt;
-	dst_release(&rt->u.dst);
+	rt->u.dst.__use++;
+
+	return rt;
+
+}
+
+struct rt6_info *rt6_lookup(struct in6_addr *daddr, struct in6_addr *saddr,
+			    int oif, int strict)
+{
+	struct flowi fl = {
+		.oif = oif,
+		.nl_u = {
+			.ip6_u = {
+				.daddr = *daddr,
+				/* TODO: saddr */
+			},
+		},
+	};
+	struct dst_entry *dst;
+	int flags = strict ? RT6_F_STRICT : 0;
+
+	dst = fib6_rule_lookup(&fl, flags, ip6_pol_route_lookup);
+	if (dst->error == 0)
+		return (struct rt6_info *) dst;
+
+	dst_release(dst);
+
 	return NULL;
 }
 
-/* ip6_ins_rt is called with FREE rt6_lock.
+/* ip6_ins_rt is called with FREE table->tb6_lock.
    It takes new route entry, the addition fails by any reason the
    route is freed. In any case, if caller does not hold it, it may
    be destroyed.
@@ -471,10 +506,12 @@ int ip6_ins_rt(struct rt6_info *rt, struct nlmsghdr *nlh,
 		void *_rtattr, struct netlink_skb_parms *req)
 {
 	int err;
+	struct fib6_table *table;
 
-	write_lock_bh(&rt6_lock);
-	err = fib6_add(&ip6_routing_table, rt, nlh, _rtattr, req);
-	write_unlock_bh(&rt6_lock);
+	table = rt->rt6i_table;
+	write_lock_bh(&table->tb6_lock);
+	err = fib6_add(&table->tb6_root, rt, nlh, _rtattr, req);
+	write_unlock_bh(&table->tb6_lock);
 
 	return err;
 }
@@ -532,51 +569,40 @@ static struct rt6_info *rt6_alloc_clone(struct rt6_info *ort, struct in6_addr *d
 	return rt;
 }
 
-#define BACKTRACK() \
-if (rt == &ip6_null_entry) { \
-       while ((fn = fn->parent) != NULL) { \
-		if (fn->fn_flags & RTN_ROOT) { \
-			goto out; \
-		} \
-		if (fn->fn_flags & RTN_RTINFO) \
-			goto restart; \
-	} \
-}
-
-
-void ip6_route_input(struct sk_buff *skb)
+struct rt6_info *ip6_pol_route_input(struct fib6_table *table, struct flowi *fl,
+				     int flags)
 {
 	struct fib6_node *fn;
 	struct rt6_info *rt, *nrt;
-	int strict;
+	int strict = 0;
 	int attempts = 3;
 	int err;
 	int reachable = RT6_SELECT_F_REACHABLE;
 
-	strict = ipv6_addr_type(&skb->nh.ipv6h->daddr) & (IPV6_ADDR_MULTICAST|IPV6_ADDR_LINKLOCAL) ? RT6_SELECT_F_IFACE : 0;
+	if (flags & RT6_F_STRICT)
+		strict = RT6_SELECT_F_IFACE;
 
 relookup:
-	read_lock_bh(&rt6_lock);
+	read_lock_bh(&table->tb6_lock);
 
 restart_2:
-	fn = fib6_lookup(&ip6_routing_table, &skb->nh.ipv6h->daddr,
-			 &skb->nh.ipv6h->saddr);
+	fn = fib6_lookup(&table->tb6_root, &fl->fl6_dst, &fl->fl6_src);
 
 restart:
-	rt = rt6_select(&fn->leaf, skb->dev->ifindex, strict | reachable);
+	rt = rt6_select(&fn->leaf, fl->iif, strict | reachable);
 	BACKTRACK();
 	if (rt == &ip6_null_entry ||
 	    rt->rt6i_flags & RTF_CACHE)
 		goto out;
 
 	dst_hold(&rt->u.dst);
-	read_unlock_bh(&rt6_lock);
+	read_unlock_bh(&table->tb6_lock);
 
 	if (!rt->rt6i_nexthop && !(rt->rt6i_flags & RTF_NONEXTHOP))
-		nrt = rt6_alloc_cow(rt, &skb->nh.ipv6h->daddr, &skb->nh.ipv6h->saddr);
+		nrt = rt6_alloc_cow(rt, &fl->fl6_dst, &fl->fl6_src);
 	else {
 #if CLONE_OFFLINK_ROUTE
-		nrt = rt6_alloc_clone(rt, &skb->nh.ipv6h->daddr);
+		nrt = rt6_alloc_clone(rt, &fl->fl6_dst);
 #else
 		goto out2;
 #endif
@@ -587,7 +613,7 @@ restart:
 
 	dst_hold(&rt->u.dst);
 	if (nrt) {
-		err = ip6_ins_rt(nrt, NULL, NULL, &NETLINK_CB(skb));
+		err = ip6_ins_rt(nrt, NULL, NULL, NULL);
 		if (!err)
 			goto out2;
 	}
@@ -596,7 +622,7 @@ restart:
 		goto out2;
 
 	/*
-	 * Race condition! In the gap, when rt6_lock was
+	 * Race condition! In the gap, when table->tb6_lock was
 	 * released someone could insert this route.  Relookup.
 	 */
 	dst_release(&rt->u.dst);
@@ -608,30 +634,54 @@ out:
 		goto restart_2;
 	}
 	dst_hold(&rt->u.dst);
-	read_unlock_bh(&rt6_lock);
+	read_unlock_bh(&table->tb6_lock);
 out2:
 	rt->u.dst.lastuse = jiffies;
 	rt->u.dst.__use++;
-	skb->dst = (struct dst_entry *) rt;
-	return;
+
+	return rt;
 }
 
-struct dst_entry * ip6_route_output(struct sock *sk, struct flowi *fl)
+void ip6_route_input(struct sk_buff *skb)
+{
+	struct ipv6hdr *iph = skb->nh.ipv6h;
+	struct flowi fl = {
+		.iif = skb->dev->ifindex,
+		.nl_u = {
+			.ip6_u = {
+				.daddr = iph->daddr,
+				.saddr = iph->saddr,
+				.flowlabel = (* (u32 *) iph)&IPV6_FLOWINFO_MASK,
+			},
+		},
+		.proto = iph->nexthdr,
+	};
+	int flags = 0;
+
+	if (rt6_need_strict(&iph->daddr))
+		flags |= RT6_F_STRICT;
+
+	skb->dst = fib6_rule_lookup(&fl, flags, ip6_pol_route_input);
+}
+
+static struct rt6_info *ip6_pol_route_output(struct fib6_table *table,
+					     struct flowi *fl, int flags)
 {
 	struct fib6_node *fn;
 	struct rt6_info *rt, *nrt;
-	int strict;
+	int strict = 0;
 	int attempts = 3;
 	int err;
 	int reachable = RT6_SELECT_F_REACHABLE;
 
-	strict = ipv6_addr_type(&fl->fl6_dst) & (IPV6_ADDR_MULTICAST|IPV6_ADDR_LINKLOCAL) ? RT6_SELECT_F_IFACE : 0;
+	if (flags & RT6_F_STRICT)
+		strict = RT6_SELECT_F_IFACE;
 
 relookup:
-	read_lock_bh(&rt6_lock);
+	read_lock_bh(&table->tb6_lock);
 
 restart_2:
-	fn = fib6_lookup(&ip6_routing_table, &fl->fl6_dst, &fl->fl6_src);
+	fn = fib6_lookup(&table->tb6_root, &fl->fl6_dst, &fl->fl6_src);
 
 restart:
 	rt = rt6_select(&fn->leaf, fl->oif, strict | reachable);
@@ -641,7 +691,7 @@ restart:
 		goto out;
 
 	dst_hold(&rt->u.dst);
-	read_unlock_bh(&rt6_lock);
+	read_unlock_bh(&table->tb6_lock);
 
 	if (!rt->rt6i_nexthop && !(rt->rt6i_flags & RTF_NONEXTHOP))
 		nrt = rt6_alloc_cow(rt, &fl->fl6_dst, &fl->fl6_src);
@@ -667,7 +717,7 @@ restart:
 		goto out2;
 
 	/*
-	 * Race condition! In the gap, when rt6_lock was
+	 * Race condition! In the gap, when table->tb6_lock was
 	 * released someone could insert this route.  Relookup.
 	 */
 	dst_release(&rt->u.dst);
@@ -679,11 +729,21 @@ out:
 		goto restart_2;
 	}
 	dst_hold(&rt->u.dst);
-	read_unlock_bh(&rt6_lock);
+	read_unlock_bh(&table->tb6_lock);
 out2:
 	rt->u.dst.lastuse = jiffies;
 	rt->u.dst.__use++;
-	return &rt->u.dst;
+	return rt;
+}
+
+struct dst_entry * ip6_route_output(struct sock *sk, struct flowi *fl)
+{
+	int flags = 0;
+
+	if (rt6_need_strict(&fl->fl6_dst))
+		flags |= RT6_F_STRICT;
+
+	return fib6_rule_lookup(fl, flags, ip6_pol_route_output);
 }
 
 
@@ -906,7 +966,8 @@ int ipv6_get_hoplimit(struct net_device *dev)
  */
 
 int ip6_route_add(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh, 
-		void *_rtattr, struct netlink_skb_parms *req)
+		  void *_rtattr, struct netlink_skb_parms *req,
+		  u32 table_id)
 {
 	int err;
 	struct rtmsg *r;
@@ -914,6 +975,7 @@ int ip6_route_add(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh,
 	struct rt6_info *rt = NULL;
 	struct net_device *dev = NULL;
 	struct inet6_dev *idev = NULL;
+	struct fib6_table *table;
 	int addr_type;
 
 	rta = (struct rtattr **) _rtattr;
@@ -937,6 +999,12 @@ int ip6_route_add(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh,
 	if (rtmsg->rtmsg_metric == 0)
 		rtmsg->rtmsg_metric = IP6_RT_PRIO_USER;
 
+	table = fib6_new_table(table_id);
+	if (table == NULL) {
+		err = -ENOBUFS;
+		goto out;
+	}
+
 	rt = ip6_dst_alloc();
 
 	if (rt == NULL) {
@@ -1093,6 +1161,7 @@ install_route:
 		rt->u.dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(dst_mtu(&rt->u.dst));
 	rt->u.dst.dev = dev;
 	rt->rt6i_idev = idev;
+	rt->rt6i_table = table;
 	return ip6_ins_rt(rt, nlh, _rtattr, req);
 
 out:
@@ -1108,26 +1177,35 @@ out:
 int ip6_del_rt(struct rt6_info *rt, struct nlmsghdr *nlh, void *_rtattr, struct netlink_skb_parms *req)
 {
 	int err;
+	struct fib6_table *table;
 
-	write_lock_bh(&rt6_lock);
+	table = rt->rt6i_table;
+	write_lock_bh(&table->tb6_lock);
 
 	err = fib6_del(rt, nlh, _rtattr, req);
 	dst_release(&rt->u.dst);
 
-	write_unlock_bh(&rt6_lock);
+	write_unlock_bh(&table->tb6_lock);
 
 	return err;
 }
 
-static int ip6_route_del(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh, void *_rtattr, struct netlink_skb_parms *req)
+static int ip6_route_del(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh,
+			 void *_rtattr, struct netlink_skb_parms *req,
+			 u32 table_id)
 {
+	struct fib6_table *table;
 	struct fib6_node *fn;
 	struct rt6_info *rt;
 	int err = -ESRCH;
 
-	read_lock_bh(&rt6_lock);
+	table = fib6_get_table(table_id);
+	if (table == NULL)
+		return err;
+
+	read_lock_bh(&table->tb6_lock);
 
-	fn = fib6_locate(&ip6_routing_table,
+	fn = fib6_locate(&table->tb6_root,
 			 &rtmsg->rtmsg_dst, rtmsg->rtmsg_dst_len,
 			 &rtmsg->rtmsg_src, rtmsg->rtmsg_src_len);
 	
@@ -1144,12 +1222,12 @@ static int ip6_route_del(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh, void *_r
 			    rtmsg->rtmsg_metric != rt->rt6i_metric)
 				continue;
 			dst_hold(&rt->u.dst);
-			read_unlock_bh(&rt6_lock);
+			read_unlock_bh(&table->tb6_lock);
 
 			return ip6_del_rt(rt, nlh, _rtattr, req);
 		}
 	}
-	read_unlock_bh(&rt6_lock);
+	read_unlock_bh(&table->tb6_lock);
 
 	return err;
 }
@@ -1161,10 +1239,15 @@ void rt6_redirect(struct in6_addr *dest, struct in6_addr *saddr,
 		  struct neighbour *neigh, u8 *lladdr, int on_link)
 {
 	struct rt6_info *rt, *nrt = NULL;
-	int strict;
 	struct fib6_node *fn;
+	struct fib6_table *table;
 	struct netevent_redirect netevent;
 
+	/* TODO: Very lazy, might need to check all tables */
+	table = fib6_get_table(RT6_TABLE_MAIN);
+	if (table == NULL)
+		return;
+
 	/*
 	 * Get the "current" route for this destination and
 	 * check if the redirect has come from approriate router.
@@ -1175,10 +1258,9 @@ void rt6_redirect(struct in6_addr *dest, struct in6_addr *saddr,
 	 * is a bit fuzzy and one might need to check all possible
 	 * routes.
 	 */
-	strict = ipv6_addr_type(dest) & (IPV6_ADDR_MULTICAST | IPV6_ADDR_LINKLOCAL);
 
-	read_lock_bh(&rt6_lock);
-	fn = fib6_lookup(&ip6_routing_table, dest, NULL);
+	read_lock_bh(&table->tb6_lock);
+	fn = fib6_lookup(&table->tb6_root, dest, NULL);
 restart:
 	for (rt = fn->leaf; rt; rt = rt->u.next) {
 		/*
@@ -1201,7 +1283,7 @@ restart:
 	}
 	if (rt)
 		dst_hold(&rt->u.dst);
-	else if (strict) {
+	else if (rt6_need_strict(dest)) {
 		while ((fn = fn->parent) != NULL) {
 			if (fn->fn_flags & RTN_ROOT)
 				break;
@@ -1209,7 +1291,7 @@ restart:
 				goto restart;
 		}
 	}
-	read_unlock_bh(&rt6_lock);
+	read_unlock_bh(&table->tb6_lock);
 
 	if (!rt) {
 		if (net_ratelimit())
@@ -1384,6 +1466,7 @@ static struct rt6_info * ip6_rt_copy(struct rt6_info *ort)
 #ifdef CONFIG_IPV6_SUBTREES
 		memcpy(&rt->rt6i_src, &ort->rt6i_src, sizeof(struct rt6key));
 #endif
+		rt->rt6i_table = ort->rt6i_table;
 	}
 	return rt;
 }
@@ -1394,9 +1477,14 @@ static struct rt6_info *rt6_get_route_info(struct in6_addr *prefix, int prefixle
 {
 	struct fib6_node *fn;
 	struct rt6_info *rt = NULL;
+	struct fib6_table *table;
+
+	table = fib6_get_table(RT6_TABLE_INFO);
+	if (table == NULL)
+		return NULL;
 
-	write_lock_bh(&rt6_lock);
-	fn = fib6_locate(&ip6_routing_table, prefix ,prefixlen, NULL, 0);
+	write_lock_bh(&table->tb6_lock);
+	fn = fib6_locate(&table->tb6_root, prefix ,prefixlen, NULL, 0);
 	if (!fn)
 		goto out;
 
@@ -1411,7 +1499,7 @@ static struct rt6_info *rt6_get_route_info(struct in6_addr *prefix, int prefixle
 		break;
 	}
 out:
-	write_unlock_bh(&rt6_lock);
+	write_unlock_bh(&table->tb6_lock);
 	return rt;
 }
 
@@ -1433,7 +1521,7 @@ static struct rt6_info *rt6_add_route_info(struct in6_addr *prefix, int prefixle
 		rtmsg.rtmsg_flags |= RTF_DEFAULT;
 	rtmsg.rtmsg_ifindex = ifindex;
 
-	ip6_route_add(&rtmsg, NULL, NULL, NULL);
+	ip6_route_add(&rtmsg, NULL, NULL, NULL, RT6_TABLE_INFO);
 
 	return rt6_get_route_info(prefix, prefixlen, gwaddr, ifindex);
 }
@@ -1442,12 +1530,14 @@ static struct rt6_info *rt6_add_route_info(struct in6_addr *prefix, int prefixle
 struct rt6_info *rt6_get_dflt_router(struct in6_addr *addr, struct net_device *dev)
 {	
 	struct rt6_info *rt;
-	struct fib6_node *fn;
+	struct fib6_table *table;
 
-	fn = &ip6_routing_table;
+	table = fib6_get_table(RT6_TABLE_DFLT);
+	if (table == NULL)
+		return NULL;
 
-	write_lock_bh(&rt6_lock);
-	for (rt = fn->leaf; rt; rt=rt->u.next) {
+	write_lock_bh(&table->tb6_lock);
+	for (rt = table->tb6_root.leaf; rt; rt=rt->u.next) {
 		if (dev == rt->rt6i_dev &&
 		    ((rt->rt6i_flags & (RTF_ADDRCONF | RTF_DEFAULT)) == (RTF_ADDRCONF | RTF_DEFAULT)) &&
 		    ipv6_addr_equal(&rt->rt6i_gateway, addr))
@@ -1455,7 +1545,7 @@ struct rt6_info *rt6_get_dflt_router(struct in6_addr *addr, struct net_device *d
 	}
 	if (rt)
 		dst_hold(&rt->u.dst);
-	write_unlock_bh(&rt6_lock);
+	write_unlock_bh(&table->tb6_lock);
 	return rt;
 }
 
@@ -1474,28 +1564,31 @@ struct rt6_info *rt6_add_dflt_router(struct in6_addr *gwaddr,
 
 	rtmsg.rtmsg_ifindex = dev->ifindex;
 
-	ip6_route_add(&rtmsg, NULL, NULL, NULL);
+	ip6_route_add(&rtmsg, NULL, NULL, NULL, RT6_TABLE_DFLT);
 	return rt6_get_dflt_router(gwaddr, dev);
 }
 
 void rt6_purge_dflt_routers(void)
 {
 	struct rt6_info *rt;
+	struct fib6_table *table;
+
+	/* NOTE: Keep consistent with rt6_get_dflt_router */
+	table = fib6_get_table(RT6_TABLE_DFLT);
+	if (table == NULL)
+		return;
 
 restart:
-	read_lock_bh(&rt6_lock);
-	for (rt = ip6_routing_table.leaf; rt; rt = rt->u.next) {
+	read_lock_bh(&table->tb6_lock);
+	for (rt = table->tb6_root.leaf; rt; rt = rt->u.next) {
 		if (rt->rt6i_flags & (RTF_DEFAULT | RTF_ADDRCONF)) {
 			dst_hold(&rt->u.dst);
-
-			read_unlock_bh(&rt6_lock);
-
+			read_unlock_bh(&table->tb6_lock);
 			ip6_del_rt(rt, NULL, NULL, NULL);
-
 			goto restart;
 		}
 	}
-	read_unlock_bh(&rt6_lock);
+	read_unlock_bh(&table->tb6_lock);
 }
 
 int ipv6_route_ioctl(unsigned int cmd, void __user *arg)
@@ -1516,10 +1609,12 @@ int ipv6_route_ioctl(unsigned int cmd, void __user *arg)
 		rtnl_lock();
 		switch (cmd) {
 		case SIOCADDRT:
-			err = ip6_route_add(&rtmsg, NULL, NULL, NULL);
+			err = ip6_route_add(&rtmsg, NULL, NULL, NULL,
+					    RT6_TABLE_MAIN);
 			break;
 		case SIOCDELRT:
-			err = ip6_route_del(&rtmsg, NULL, NULL, NULL);
+			err = ip6_route_del(&rtmsg, NULL, NULL, NULL,
+					    RT6_TABLE_MAIN);
 			break;
 		default:
 			err = -EINVAL;
@@ -1593,6 +1688,7 @@ struct rt6_info *addrconf_dst_alloc(struct inet6_dev *idev,
 
 	ipv6_addr_copy(&rt->rt6i_dst.addr, addr);
 	rt->rt6i_dst.plen = 128;
+	rt->rt6i_table = fib6_get_table(RT6_TABLE_LOCAL);
 
 	atomic_set(&rt->u.dst.__refcnt, 1);
 
@@ -1611,9 +1707,7 @@ static int fib6_ifdown(struct rt6_info *rt, void *arg)
 
 void rt6_ifdown(struct net_device *dev)
 {
-	write_lock_bh(&rt6_lock);
-	fib6_clean_tree(&ip6_routing_table, fib6_ifdown, 0, dev);
-	write_unlock_bh(&rt6_lock);
+	fib6_clean_all(fib6_ifdown, 0, dev);
 }
 
 struct rt6_mtu_change_arg
@@ -1663,13 +1757,12 @@ static int rt6_mtu_change_route(struct rt6_info *rt, void *p_arg)
 
 void rt6_mtu_change(struct net_device *dev, unsigned mtu)
 {
-	struct rt6_mtu_change_arg arg;
+	struct rt6_mtu_change_arg arg = {
+		.dev = dev,
+		.mtu = mtu,
+	};
 
-	arg.dev = dev;
-	arg.mtu = mtu;
-	read_lock_bh(&rt6_lock);
-	fib6_clean_tree(&ip6_routing_table, rt6_mtu_change_route, 0, &arg);
-	read_unlock_bh(&rt6_lock);
+	fib6_clean_all(rt6_mtu_change_route, 0, &arg);
 }
 
 static int inet6_rtm_to_rtmsg(struct rtmsg *r, struct rtattr **rta,
@@ -1719,7 +1812,7 @@ int inet6_rtm_delroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 
 	if (inet6_rtm_to_rtmsg(r, arg, &rtmsg))
 		return -EINVAL;
-	return ip6_route_del(&rtmsg, nlh, arg, &NETLINK_CB(skb));
+	return ip6_route_del(&rtmsg, nlh, arg, &NETLINK_CB(skb), r->rtm_table);
 }
 
 int inet6_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
@@ -1729,7 +1822,7 @@ int inet6_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 
 	if (inet6_rtm_to_rtmsg(r, arg, &rtmsg))
 		return -EINVAL;
-	return ip6_route_add(&rtmsg, nlh, arg, &NETLINK_CB(skb));
+	return ip6_route_add(&rtmsg, nlh, arg, &NETLINK_CB(skb), r->rtm_table);
 }
 
 struct rt6_rtnl_dump_arg
@@ -1761,6 +1854,10 @@ static int rt6_fill_node(struct sk_buff *skb, struct rt6_info *rt,
 	rtm->rtm_dst_len = rt->rt6i_dst.plen;
 	rtm->rtm_src_len = rt->rt6i_src.plen;
 	rtm->rtm_tos = 0;
+	if (rt->rt6i_table)
+		rtm->rtm_table = rt->rt6i_table->tb6_id;
+	else
+		rtm->rtm_table = RT6_TABLE_UNSPEC;
 	rtm->rtm_table = RT_TABLE_MAIN;
 	if (rt->rt6i_flags&RTF_REJECT)
 		rtm->rtm_type = RTN_UNREACHABLE;
@@ -1868,7 +1965,6 @@ static void fib6_dump_end(struct netlink_callback *cb)
 
 	if (w) {
 		cb->args[0] = 0;
-		fib6_walker_unlink(w);
 		kfree(w);
 	}
 	cb->done = (void*)cb->args[1];
@@ -1883,13 +1979,20 @@ static int fib6_dump_done(struct netlink_callback *cb)
 
 int inet6_dump_fib(struct sk_buff *skb, struct netlink_callback *cb)
 {
+	struct fib6_table *table;
 	struct rt6_rtnl_dump_arg arg;
 	struct fib6_walker_t *w;
-	int res;
+	int i, res = 0;
 
 	arg.skb = skb;
 	arg.cb = cb;
 
+	/*
+	 * cb->args[0] = pointer to walker structure
+	 * cb->args[1] = saved cb->done() pointer
+	 * cb->args[2] = current table being dumped
+	 */
+
 	w = (void*)cb->args[0];
 	if (w == NULL) {
 		/* New dump:
@@ -1905,24 +2008,48 @@ int inet6_dump_fib(struct sk_buff *skb, struct netlink_callback *cb)
 		w = kzalloc(sizeof(*w), GFP_ATOMIC);
 		if (w == NULL)
 			return -ENOMEM;
-		RT6_TRACE("dump<%p", w);
-		w->root = &ip6_routing_table;
 		w->func = fib6_dump_node;
 		w->args = &arg;
 		cb->args[0] = (long)w;
-		read_lock_bh(&rt6_lock);
-		res = fib6_walk(w);
-		read_unlock_bh(&rt6_lock);
+		cb->args[2] = FIB6_TABLE_MIN;
 	} else {
 		w->args = &arg;
-		read_lock_bh(&rt6_lock);
-		res = fib6_walk_continue(w);
-		read_unlock_bh(&rt6_lock);
+		i = cb->args[2];
+		if (i > FIB6_TABLE_MAX)
+			goto end;
+
+		table = fib6_get_table(i);
+		if (table != NULL) {
+			read_lock_bh(&table->tb6_lock);
+			w->root = &table->tb6_root;
+			res = fib6_walk_continue(w);
+			read_unlock_bh(&table->tb6_lock);
+			if (res != 0) {
+				if (res < 0)
+					fib6_walker_unlink(w);
+				goto end;
+			}
+		}
+
+		fib6_walker_unlink(w);
+		cb->args[2] = ++i;
 	}
-#if RT6_DEBUG >= 3
-	if (res <= 0 && skb->len == 0)
-		RT6_TRACE("%p>dump end\n", w);
-#endif
+
+	for (i = cb->args[2]; i <= FIB6_TABLE_MAX; i++) {
+		table = fib6_get_table(i);
+		if (table == NULL)
+			continue;
+
+		read_lock_bh(&table->tb6_lock);
+		w->root = &table->tb6_root;
+		res = fib6_walk(w);
+		read_unlock_bh(&table->tb6_lock);
+		if (res)
+			break;
+	}
+end:
+	cb->args[2] = i;
+
 	res = res < 0 ? res : skb->len;
 	/* res < 0 is an error. (really, impossible)
 	   res == 0 means that dump is complete, but skb still can contain data.
@@ -2102,16 +2229,13 @@ static int rt6_info_route(struct rt6_info *rt, void *p_arg)
 
 static int rt6_proc_info(char *buffer, char **start, off_t offset, int length)
 {
-	struct rt6_proc_arg arg;
-	arg.buffer = buffer;
-	arg.offset = offset;
-	arg.length = length;
-	arg.skip = 0;
-	arg.len = 0;
+	struct rt6_proc_arg arg = {
+		.buffer = buffer,
+		.offset = offset,
+		.length = length,
+	};
 
-	read_lock_bh(&rt6_lock);
-	fib6_clean_tree(&ip6_routing_table, rt6_info_route, 0, &arg);
-	read_unlock_bh(&rt6_lock);
+	fib6_clean_all(rt6_info_route, 0, &arg);
 
 	*start = buffer;
 	if (offset)
-- 
cgit v1.2.3


From 14c0b97ddfc2944982d078b8e33b088840068976 Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Fri, 4 Aug 2006 03:38:38 -0700
Subject: [NET]: Protocol Independant Policy Routing Rules Framework

Derived from net/ipv/fib_rules.c

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/fib_rules.h |  60 +++++++
 include/net/fib_rules.h   |  90 ++++++++++
 net/Kconfig               |   3 +
 net/core/Makefile         |   1 +
 net/core/fib_rules.c      | 416 ++++++++++++++++++++++++++++++++++++++++++++++
 net/core/rtnetlink.c      |   9 +-
 6 files changed, 577 insertions(+), 2 deletions(-)
 create mode 100644 include/linux/fib_rules.h
 create mode 100644 include/net/fib_rules.h
 create mode 100644 net/core/fib_rules.c

(limited to 'include')

diff --git a/include/linux/fib_rules.h b/include/linux/fib_rules.h
new file mode 100644
index 000000000000..5e503f0ca6e4
--- /dev/null
+++ b/include/linux/fib_rules.h
@@ -0,0 +1,60 @@
+#ifndef __LINUX_FIB_RULES_H
+#define __LINUX_FIB_RULES_H
+
+#include <linux/types.h>
+#include <linux/rtnetlink.h>
+
+/* rule is permanent, and cannot be deleted */
+#define FIB_RULE_PERMANENT	1
+
+struct fib_rule_hdr
+{
+	__u8		family;
+	__u8		dst_len;
+	__u8		src_len;
+	__u8		tos;
+
+	__u8		table;
+	__u8		res1;	/* reserved */
+	__u8		res2;	/* reserved */
+	__u8		action;
+
+	__u32		flags;
+};
+
+enum
+{
+	FRA_UNSPEC,
+	FRA_DST,	/* destination address */
+	FRA_SRC,	/* source address */
+	FRA_IFNAME,	/* interface name */
+	FRA_UNUSED1,
+	FRA_UNUSED2,
+	FRA_PRIORITY,	/* priority/preference */
+	FRA_UNUSED3,
+	FRA_UNUSED4,
+	FRA_UNUSED5,
+	FRA_FWMARK,	/* netfilter mark (IPv4) */
+	FRA_FLOW,	/* flow/class id */
+	__FRA_MAX
+};
+
+#define FRA_MAX (__FRA_MAX - 1)
+
+enum
+{
+	FR_ACT_UNSPEC,
+	FR_ACT_TO_TBL,		/* Pass to fixed table */
+	FR_ACT_RES1,
+	FR_ACT_RES2,
+	FR_ACT_RES3,
+	FR_ACT_RES4,
+	FR_ACT_BLACKHOLE,	/* Drop without notification */
+	FR_ACT_UNREACHABLE,	/* Drop with ENETUNREACH */
+	FR_ACT_PROHIBIT,	/* Drop with EACCES */
+	__FR_ACT_MAX,
+};
+
+#define FR_ACT_MAX (__FR_ACT_MAX - 1)
+
+#endif
diff --git a/include/net/fib_rules.h b/include/net/fib_rules.h
new file mode 100644
index 000000000000..61375d9e53f8
--- /dev/null
+++ b/include/net/fib_rules.h
@@ -0,0 +1,90 @@
+#ifndef __NET_FIB_RULES_H
+#define __NET_FIB_RULES_H
+
+#include <linux/types.h>
+#include <linux/netdevice.h>
+#include <linux/fib_rules.h>
+#include <net/flow.h>
+#include <net/netlink.h>
+
+struct fib_rule
+{
+	struct list_head	list;
+	atomic_t		refcnt;
+	int			ifindex;
+	char			ifname[IFNAMSIZ];
+	u32			pref;
+	u32			flags;
+	u32			table;
+	u8			action;
+	struct rcu_head		rcu;
+};
+
+struct fib_lookup_arg
+{
+	void			*lookup_ptr;
+	void			*result;
+	struct fib_rule		*rule;
+};
+
+struct fib_rules_ops
+{
+	int			family;
+	struct list_head	list;
+	int			rule_size;
+
+	int			(*action)(struct fib_rule *,
+					  struct flowi *, int,
+					  struct fib_lookup_arg *);
+	int			(*match)(struct fib_rule *,
+					 struct flowi *, int);
+	int			(*configure)(struct fib_rule *,
+					     struct sk_buff *,
+					     struct nlmsghdr *,
+					     struct fib_rule_hdr *,
+					     struct nlattr **);
+	int			(*compare)(struct fib_rule *,
+					   struct fib_rule_hdr *,
+					   struct nlattr **);
+	int			(*fill)(struct fib_rule *, struct sk_buff *,
+					struct nlmsghdr *,
+					struct fib_rule_hdr *);
+	u32			(*default_pref)(void);
+
+	int			nlgroup;
+	struct nla_policy	*policy;
+	struct list_head	*rules_list;
+	struct module		*owner;
+};
+
+static inline void fib_rule_get(struct fib_rule *rule)
+{
+	atomic_inc(&rule->refcnt);
+}
+
+static inline void fib_rule_put_rcu(struct rcu_head *head)
+{
+	struct fib_rule *rule = container_of(head, struct fib_rule, rcu);
+	kfree(rule);
+}
+
+static inline void fib_rule_put(struct fib_rule *rule)
+{
+	if (atomic_dec_and_test(&rule->refcnt))
+		call_rcu(&rule->rcu, fib_rule_put_rcu);
+}
+
+extern int			fib_rules_register(struct fib_rules_ops *);
+extern int			fib_rules_unregister(struct fib_rules_ops *);
+
+extern int			fib_rules_lookup(struct fib_rules_ops *,
+						 struct flowi *, int flags,
+						 struct fib_lookup_arg *);
+
+extern int			fib_nl_newrule(struct sk_buff *,
+					       struct nlmsghdr *, void *);
+extern int			fib_nl_delrule(struct sk_buff *,
+					       struct nlmsghdr *, void *);
+extern int			fib_rules_dump(struct sk_buff *,
+					       struct netlink_callback *, int);
+#endif
diff --git a/net/Kconfig b/net/Kconfig
index eb855b7fa642..6528a935622c 100644
--- a/net/Kconfig
+++ b/net/Kconfig
@@ -251,6 +251,9 @@ config WIRELESS_EXT
 
 source "net/netlabel/Kconfig"
 
+config FIB_RULES
+	bool
+
 endif   # if NET
 endmenu # Networking
 
diff --git a/net/core/Makefile b/net/core/Makefile
index 2645ba428d48..119568077dab 100644
--- a/net/core/Makefile
+++ b/net/core/Makefile
@@ -17,3 +17,4 @@ obj-$(CONFIG_NET_PKTGEN) += pktgen.o
 obj-$(CONFIG_WIRELESS_EXT) += wireless.o
 obj-$(CONFIG_NETPOLL) += netpoll.o
 obj-$(CONFIG_NET_DMA) += user_dma.o
+obj-$(CONFIG_FIB_RULES) += fib_rules.o
diff --git a/net/core/fib_rules.c b/net/core/fib_rules.c
new file mode 100644
index 000000000000..6cdad24038e2
--- /dev/null
+++ b/net/core/fib_rules.c
@@ -0,0 +1,416 @@
+/*
+ * net/core/fib_rules.c		Generic Routing Rules
+ *
+ *	This program is free software; you can redistribute it and/or
+ *	modify it under the terms of the GNU General Public License as
+ *	published by the Free Software Foundation, version 2.
+ *
+ * Authors:	Thomas Graf <tgraf@suug.ch>
+ */
+
+#include <linux/config.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/list.h>
+#include <net/fib_rules.h>
+
+static LIST_HEAD(rules_ops);
+static DEFINE_SPINLOCK(rules_mod_lock);
+
+static void notify_rule_change(int event, struct fib_rule *rule,
+			       struct fib_rules_ops *ops);
+
+static struct fib_rules_ops *lookup_rules_ops(int family)
+{
+	struct fib_rules_ops *ops;
+
+	rcu_read_lock();
+	list_for_each_entry_rcu(ops, &rules_ops, list) {
+		if (ops->family == family) {
+			if (!try_module_get(ops->owner))
+				ops = NULL;
+			rcu_read_unlock();
+			return ops;
+		}
+	}
+	rcu_read_unlock();
+
+	return NULL;
+}
+
+static void rules_ops_put(struct fib_rules_ops *ops)
+{
+	if (ops)
+		module_put(ops->owner);
+}
+
+int fib_rules_register(struct fib_rules_ops *ops)
+{
+	int err = -EEXIST;
+	struct fib_rules_ops *o;
+
+	if (ops->rule_size < sizeof(struct fib_rule))
+		return -EINVAL;
+
+	if (ops->match == NULL || ops->configure == NULL ||
+	    ops->compare == NULL || ops->fill == NULL ||
+	    ops->action == NULL)
+		return -EINVAL;
+
+	spin_lock(&rules_mod_lock);
+	list_for_each_entry(o, &rules_ops, list)
+		if (ops->family == o->family)
+			goto errout;
+
+	list_add_tail_rcu(&ops->list, &rules_ops);
+	err = 0;
+errout:
+	spin_unlock(&rules_mod_lock);
+
+	return err;
+}
+
+EXPORT_SYMBOL_GPL(fib_rules_register);
+
+static void cleanup_ops(struct fib_rules_ops *ops)
+{
+	struct fib_rule *rule, *tmp;
+
+	list_for_each_entry_safe(rule, tmp, ops->rules_list, list) {
+		list_del_rcu(&rule->list);
+		fib_rule_put(rule);
+	}
+}
+
+int fib_rules_unregister(struct fib_rules_ops *ops)
+{
+	int err = 0;
+	struct fib_rules_ops *o;
+
+	spin_lock(&rules_mod_lock);
+	list_for_each_entry(o, &rules_ops, list) {
+		if (o == ops) {
+			list_del_rcu(&o->list);
+			cleanup_ops(ops);
+			goto out;
+		}
+	}
+
+	err = -ENOENT;
+out:
+	spin_unlock(&rules_mod_lock);
+
+	synchronize_rcu();
+
+	return err;
+}
+
+EXPORT_SYMBOL_GPL(fib_rules_unregister);
+
+int fib_rules_lookup(struct fib_rules_ops *ops, struct flowi *fl,
+		     int flags, struct fib_lookup_arg *arg)
+{
+	struct fib_rule *rule;
+	int err;
+
+	rcu_read_lock();
+
+	list_for_each_entry_rcu(rule, ops->rules_list, list) {
+		if (rule->ifindex && (rule->ifindex != fl->iif))
+			continue;
+
+		if (!ops->match(rule, fl, flags))
+			continue;
+
+		err = ops->action(rule, fl, flags, arg);
+		if (err != -EAGAIN) {
+			fib_rule_get(rule);
+			arg->rule = rule;
+			goto out;
+		}
+	}
+
+	err = -ENETUNREACH;
+out:
+	rcu_read_unlock();
+
+	return err;
+}
+
+EXPORT_SYMBOL_GPL(fib_rules_lookup);
+
+int fib_nl_newrule(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
+{
+	struct fib_rule_hdr *frh = nlmsg_data(nlh);
+	struct fib_rules_ops *ops = NULL;
+	struct fib_rule *rule, *r, *last = NULL;
+	struct nlattr *tb[FRA_MAX+1];
+	int err = -EINVAL;
+
+	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*frh)))
+		goto errout;
+
+	ops = lookup_rules_ops(frh->family);
+	if (ops == NULL) {
+		err = EAFNOSUPPORT;
+		goto errout;
+	}
+
+	err = nlmsg_parse(nlh, sizeof(*frh), tb, FRA_MAX, ops->policy);
+	if (err < 0)
+		goto errout;
+
+	if (tb[FRA_IFNAME] && nla_len(tb[FRA_IFNAME]) > IFNAMSIZ)
+		goto errout;
+
+	rule = kzalloc(ops->rule_size, GFP_KERNEL);
+	if (rule == NULL) {
+		err = -ENOMEM;
+		goto errout;
+	}
+
+	if (tb[FRA_PRIORITY])
+		rule->pref = nla_get_u32(tb[FRA_PRIORITY]);
+
+	if (tb[FRA_IFNAME]) {
+		struct net_device *dev;
+
+		rule->ifindex = -1;
+		if (nla_strlcpy(rule->ifname, tb[FRA_IFNAME],
+				IFNAMSIZ) >= IFNAMSIZ)
+			goto errout_free;
+
+		dev = __dev_get_by_name(rule->ifname);
+		if (dev)
+			rule->ifindex = dev->ifindex;
+	}
+
+	rule->action = frh->action;
+	rule->flags = frh->flags;
+	rule->table = frh->table;
+
+	if (!rule->pref && ops->default_pref)
+		rule->pref = ops->default_pref();
+
+	err = ops->configure(rule, skb, nlh, frh, tb);
+	if (err < 0)
+		goto errout_free;
+
+	list_for_each_entry(r, ops->rules_list, list) {
+		if (r->pref > rule->pref)
+			break;
+		last = r;
+	}
+
+	fib_rule_get(rule);
+
+	if (last)
+		list_add_rcu(&rule->list, &last->list);
+	else
+		list_add_rcu(&rule->list, ops->rules_list);
+
+	notify_rule_change(RTM_NEWRULE, rule, ops);
+	rules_ops_put(ops);
+	return 0;
+
+errout_free:
+	kfree(rule);
+errout:
+	rules_ops_put(ops);
+	return err;
+}
+
+int fib_nl_delrule(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
+{
+	struct fib_rule_hdr *frh = nlmsg_data(nlh);
+	struct fib_rules_ops *ops = NULL;
+	struct fib_rule *rule;
+	struct nlattr *tb[FRA_MAX+1];
+	int err = -EINVAL;
+
+	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*frh)))
+		goto errout;
+
+	ops = lookup_rules_ops(frh->family);
+	if (ops == NULL) {
+		err = EAFNOSUPPORT;
+		goto errout;
+	}
+
+	err = nlmsg_parse(nlh, sizeof(*frh), tb, FRA_MAX, ops->policy);
+	if (err < 0)
+		goto errout;
+
+	list_for_each_entry(rule, ops->rules_list, list) {
+		if (frh->action && (frh->action != rule->action))
+			continue;
+
+		if (frh->table && (frh->table != rule->table))
+			continue;
+
+		if (tb[FRA_PRIORITY] &&
+		    (rule->pref != nla_get_u32(tb[FRA_PRIORITY])))
+			continue;
+
+		if (tb[FRA_IFNAME] &&
+		    nla_strcmp(tb[FRA_IFNAME], rule->ifname))
+			continue;
+
+		if (!ops->compare(rule, frh, tb))
+			continue;
+
+		if (rule->flags & FIB_RULE_PERMANENT) {
+			err = -EPERM;
+			goto errout;
+		}
+
+		list_del_rcu(&rule->list);
+		synchronize_rcu();
+		notify_rule_change(RTM_DELRULE, rule, ops);
+		fib_rule_put(rule);
+		rules_ops_put(ops);
+		return 0;
+	}
+
+	err = -ENOENT;
+errout:
+	rules_ops_put(ops);
+	return err;
+}
+
+static int fib_nl_fill_rule(struct sk_buff *skb, struct fib_rule *rule,
+			    u32 pid, u32 seq, int type, int flags,
+			    struct fib_rules_ops *ops)
+{
+	struct nlmsghdr *nlh;
+	struct fib_rule_hdr *frh;
+
+	nlh = nlmsg_put(skb, pid, seq, type, sizeof(*frh), flags);
+	if (nlh == NULL)
+		return -1;
+
+	frh = nlmsg_data(nlh);
+	frh->table = rule->table;
+	frh->res1 = 0;
+	frh->res2 = 0;
+	frh->action = rule->action;
+	frh->flags = rule->flags;
+
+	if (rule->ifname[0])
+		NLA_PUT_STRING(skb, FRA_IFNAME, rule->ifname);
+
+	if (rule->pref)
+		NLA_PUT_U32(skb, FRA_PRIORITY, rule->pref);
+
+	if (ops->fill(rule, skb, nlh, frh) < 0)
+		goto nla_put_failure;
+
+	return nlmsg_end(skb, nlh);
+
+nla_put_failure:
+	return nlmsg_cancel(skb, nlh);
+}
+
+int fib_rules_dump(struct sk_buff *skb, struct netlink_callback *cb, int family)
+{
+	int idx = 0;
+	struct fib_rule *rule;
+	struct fib_rules_ops *ops;
+
+	ops = lookup_rules_ops(family);
+	if (ops == NULL)
+		return -EAFNOSUPPORT;
+
+	rcu_read_lock();
+	list_for_each_entry(rule, ops->rules_list, list) {
+		if (idx < cb->args[0])
+			goto skip;
+
+		if (fib_nl_fill_rule(skb, rule, NETLINK_CB(cb->skb).pid,
+				     cb->nlh->nlmsg_seq, RTM_NEWRULE,
+				     NLM_F_MULTI, ops) < 0)
+			break;
+skip:
+		idx++;
+	}
+	rcu_read_unlock();
+	cb->args[0] = idx;
+	rules_ops_put(ops);
+
+	return skb->len;
+}
+
+EXPORT_SYMBOL_GPL(fib_rules_dump);
+
+static void notify_rule_change(int event, struct fib_rule *rule,
+			       struct fib_rules_ops *ops)
+{
+	int size = nlmsg_total_size(sizeof(struct fib_rule_hdr) + 128);
+	struct sk_buff *skb = alloc_skb(size, GFP_KERNEL);
+
+	if (skb == NULL)
+		netlink_set_err(rtnl, 0, ops->nlgroup, ENOBUFS);
+	else if (fib_nl_fill_rule(skb, rule, 0, 0, event, 0, ops) < 0) {
+		kfree_skb(skb);
+		netlink_set_err(rtnl, 0, ops->nlgroup, EINVAL);
+	} else
+		netlink_broadcast(rtnl, skb, 0, ops->nlgroup, GFP_KERNEL);
+}
+
+static void attach_rules(struct list_head *rules, struct net_device *dev)
+{
+	struct fib_rule *rule;
+
+	list_for_each_entry(rule, rules, list) {
+		if (rule->ifindex == -1 &&
+		    strcmp(dev->name, rule->ifname) == 0)
+			rule->ifindex = dev->ifindex;
+	}
+}
+
+static void detach_rules(struct list_head *rules, struct net_device *dev)
+{
+	struct fib_rule *rule;
+
+	list_for_each_entry(rule, rules, list)
+		if (rule->ifindex == dev->ifindex)
+			rule->ifindex = -1;
+}
+
+
+static int fib_rules_event(struct notifier_block *this, unsigned long event,
+			    void *ptr)
+{
+	struct net_device *dev = ptr;
+	struct fib_rules_ops *ops;
+
+	ASSERT_RTNL();
+	rcu_read_lock();
+
+	switch (event) {
+	case NETDEV_REGISTER:
+		list_for_each_entry(ops, &rules_ops, list)
+			attach_rules(ops->rules_list, dev);
+		break;
+
+	case NETDEV_UNREGISTER:
+		list_for_each_entry(ops, &rules_ops, list)
+			detach_rules(ops->rules_list, dev);
+		break;
+	}
+
+	rcu_read_unlock();
+
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block fib_rules_notifier = {
+	.notifier_call = fib_rules_event,
+};
+
+static int __init fib_rules_init(void)
+{
+	return register_netdevice_notifier(&fib_rules_notifier);
+}
+
+subsys_initcall(fib_rules_init);
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index 30cc1ba6ed5c..aa7cff2257b1 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -49,6 +49,7 @@
 #include <net/udp.h>
 #include <net/sock.h>
 #include <net/pkt_sched.h>
+#include <net/fib_rules.h>
 #include <net/netlink.h>
 #ifdef CONFIG_NET_WIRELESS_RTNETLINK
 #include <linux/wireless.h>
@@ -103,7 +104,7 @@ static const int rtm_min[RTM_NR_FAMILIES] =
 	[RTM_FAM(RTM_NEWADDR)]      = NLMSG_LENGTH(sizeof(struct ifaddrmsg)),
 	[RTM_FAM(RTM_NEWROUTE)]     = NLMSG_LENGTH(sizeof(struct rtmsg)),
 	[RTM_FAM(RTM_NEWNEIGH)]     = NLMSG_LENGTH(sizeof(struct ndmsg)),
-	[RTM_FAM(RTM_NEWRULE)]      = NLMSG_LENGTH(sizeof(struct rtmsg)),
+	[RTM_FAM(RTM_NEWRULE)]      = NLMSG_LENGTH(sizeof(struct fib_rule_hdr)),
 	[RTM_FAM(RTM_NEWQDISC)]     = NLMSG_LENGTH(sizeof(struct tcmsg)),
 	[RTM_FAM(RTM_NEWTCLASS)]    = NLMSG_LENGTH(sizeof(struct tcmsg)),
 	[RTM_FAM(RTM_NEWTFILTER)]   = NLMSG_LENGTH(sizeof(struct tcmsg)),
@@ -120,7 +121,7 @@ static const int rta_max[RTM_NR_FAMILIES] =
 	[RTM_FAM(RTM_NEWADDR)]      = IFA_MAX,
 	[RTM_FAM(RTM_NEWROUTE)]     = RTA_MAX,
 	[RTM_FAM(RTM_NEWNEIGH)]     = NDA_MAX,
-	[RTM_FAM(RTM_NEWRULE)]      = RTA_MAX,
+	[RTM_FAM(RTM_NEWRULE)]      = FRA_MAX,
 	[RTM_FAM(RTM_NEWQDISC)]     = TCA_MAX,
 	[RTM_FAM(RTM_NEWTCLASS)]    = TCA_MAX,
 	[RTM_FAM(RTM_NEWTFILTER)]   = TCA_MAX,
@@ -757,6 +758,10 @@ static struct rtnetlink_link link_rtnetlink_table[RTM_NR_MSGTYPES] =
 	[RTM_NEWNEIGH    - RTM_BASE] = { .doit   = neigh_add		 },
 	[RTM_DELNEIGH    - RTM_BASE] = { .doit   = neigh_delete		 },
 	[RTM_GETNEIGH    - RTM_BASE] = { .dumpit = neigh_dump_info	 },
+#ifdef CONFIG_FIB_RULES
+	[RTM_NEWRULE     - RTM_BASE] = { .doit   = fib_nl_newrule	 },
+	[RTM_DELRULE     - RTM_BASE] = { .doit   = fib_nl_delrule	 },
+#endif
 	[RTM_GETRULE     - RTM_BASE] = { .dumpit = rtnetlink_dump_all	 },
 	[RTM_GETNEIGHTBL - RTM_BASE] = { .dumpit = neightbl_dump_info	 },
 	[RTM_SETNEIGHTBL - RTM_BASE] = { .doit   = neightbl_set		 },
-- 
cgit v1.2.3


From 101367c2f8c464ea96643192673aa18d88e6336d Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Fri, 4 Aug 2006 03:39:02 -0700
Subject: [IPV6]: Policy Routing Rules

Adds support for policy routing rules including a new
local table for routes with a local destination.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/rtnetlink.h |   2 +
 include/net/ip6_fib.h     |   9 +-
 include/net/ip6_route.h   |   5 +
 net/ipv6/Kconfig          |   1 +
 net/ipv6/Makefile         |   1 +
 net/ipv6/addrconf.c       |   1 +
 net/ipv6/fib6_rules.c     | 251 ++++++++++++++++++++++++++++++++++++++++++++++
 net/ipv6/ip6_fib.c        |  21 ++--
 net/ipv6/route.c          |  50 +++++++++
 9 files changed, 329 insertions(+), 12 deletions(-)
 create mode 100644 net/ipv6/fib6_rules.c

(limited to 'include')

diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index facd9ee37b76..bf353538ae93 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -889,6 +889,8 @@ enum rtnetlink_groups {
 	RTNLGRP_NOP4,
 	RTNLGRP_IPV6_PREFIX,
 #define RTNLGRP_IPV6_PREFIX	RTNLGRP_IPV6_PREFIX
+	RTNLGRP_IPV6_RULE,
+#define RTNLGRP_IPV6_RULE	RTNLGRP_IPV6_RULE
 	__RTNLGRP_MAX
 };
 #define RTNLGRP_MAX	(__RTNLGRP_MAX - 1)
diff --git a/include/net/ip6_fib.h b/include/net/ip6_fib.h
index 818411519c89..7b47e8d5a765 100644
--- a/include/net/ip6_fib.h
+++ b/include/net/ip6_fib.h
@@ -155,7 +155,6 @@ struct fib6_table {
 
 #define RT6_TABLE_UNSPEC	RT_TABLE_UNSPEC
 #define RT6_TABLE_MAIN		RT_TABLE_MAIN
-#define RT6_TABLE_LOCAL		RT6_TABLE_MAIN
 #define RT6_TABLE_DFLT		RT6_TABLE_MAIN
 #define RT6_TABLE_INFO		RT6_TABLE_MAIN
 #define RT6_TABLE_PREFIX	RT6_TABLE_MAIN
@@ -163,9 +162,11 @@ struct fib6_table {
 #ifdef CONFIG_IPV6_MULTIPLE_TABLES
 #define FIB6_TABLE_MIN		1
 #define FIB6_TABLE_MAX		RT_TABLE_MAX
+#define RT6_TABLE_LOCAL		RT_TABLE_LOCAL
 #else
 #define FIB6_TABLE_MIN		RT_TABLE_MAIN
 #define FIB6_TABLE_MAX		FIB6_TABLE_MIN
+#define RT6_TABLE_LOCAL		RT6_TABLE_MAIN
 #endif
 
 #define RT6_F_STRICT		1
@@ -221,5 +222,11 @@ extern void			fib6_run_gc(unsigned long dummy);
 extern void			fib6_gc_cleanup(void);
 
 extern void			fib6_init(void);
+
+extern void			fib6_rules_init(void);
+extern void			fib6_rules_cleanup(void);
+extern int			fib6_rules_dump(struct sk_buff *,
+						struct netlink_callback *);
+
 #endif
 #endif
diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index d49c8c90eb68..9bfa3cc6cedb 100644
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -41,6 +41,11 @@ struct pol_chain {
 
 extern struct rt6_info	ip6_null_entry;
 
+#ifdef CONFIG_IPV6_MULTIPLE_TABLES
+extern struct rt6_info	ip6_prohibit_entry;
+extern struct rt6_info	ip6_blk_hole_entry;
+#endif
+
 extern int ip6_rt_gc_interval;
 
 extern void			ip6_route_input(struct sk_buff *skb);
diff --git a/net/ipv6/Kconfig b/net/ipv6/Kconfig
index 159c63d99c81..36a6c2b79889 100644
--- a/net/ipv6/Kconfig
+++ b/net/ipv6/Kconfig
@@ -139,6 +139,7 @@ config IPV6_TUNNEL
 config IPV6_MULTIPLE_TABLES
 	bool "IPv6: Multiple Routing Tables"
 	depends on IPV6 && EXPERIMENTAL
+	select FIB_RULES
 	---help---
 	  Support multiple routing tables.
 
diff --git a/net/ipv6/Makefile b/net/ipv6/Makefile
index 386e0a626948..9eebf6091279 100644
--- a/net/ipv6/Makefile
+++ b/net/ipv6/Makefile
@@ -13,6 +13,7 @@ ipv6-objs :=	af_inet6.o anycast.o ip6_output.o ip6_input.o addrconf.o sit.o \
 ipv6-$(CONFIG_XFRM) += xfrm6_policy.o xfrm6_state.o xfrm6_input.o \
 	xfrm6_output.o
 ipv6-$(CONFIG_NETFILTER) += netfilter.o
+ipv6-$(CONFIG_IPV6_MULTIPLE_TABLES) += fib6_rules.o
 ipv6-objs += $(ipv6-y)
 
 obj-$(CONFIG_INET6_AH) += ah6.o
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 318767fcefdc..ed766eebc022 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -3528,6 +3528,7 @@ static struct rtnetlink_link inet6_rtnetlink_table[RTM_NR_MSGTYPES] = {
 	[RTM_DELROUTE - RTM_BASE] = { .doit	= inet6_rtm_delroute, },
 	[RTM_GETROUTE - RTM_BASE] = { .doit	= inet6_rtm_getroute,
 				      .dumpit	= inet6_dump_fib, },
+	[RTM_GETRULE  - RTM_BASE] = { .dumpit   = fib6_rules_dump,   },
 };
 
 static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
diff --git a/net/ipv6/fib6_rules.c b/net/ipv6/fib6_rules.c
new file mode 100644
index 000000000000..c3c8195744ee
--- /dev/null
+++ b/net/ipv6/fib6_rules.c
@@ -0,0 +1,251 @@
+/*
+ * net/ipv6/fib6_rules.c	IPv6 Routing Policy Rules
+ *
+ * Copyright (C)2003-2006 Helsinki University of Technology
+ * Copyright (C)2003-2006 USAGI/WIDE Project
+ *
+ *	This program is free software; you can redistribute it and/or
+ *	modify it under the terms of the GNU General Public License as
+ *	published by the Free Software Foundation, version 2.
+ *
+ * Authors
+ *	Thomas Graf		<tgraf@suug.ch>
+ *	Ville Nuorvala		<vnuorval@tcs.hut.fi>
+ */
+
+#include <linux/config.h>
+#include <linux/netdevice.h>
+
+#include <net/fib_rules.h>
+#include <net/ipv6.h>
+#include <net/ip6_route.h>
+#include <net/netlink.h>
+
+struct fib6_rule
+{
+	struct fib_rule		common;
+	struct rt6key		src;
+	struct rt6key		dst;
+	u8			tclass;
+};
+
+static struct fib_rules_ops fib6_rules_ops;
+
+static struct fib6_rule main_rule = {
+	.common = {
+		.refcnt =	ATOMIC_INIT(2),
+		.pref =		0x7FFE,
+		.action =	FR_ACT_TO_TBL,
+		.table =	RT6_TABLE_MAIN,
+	},
+};
+
+static struct fib6_rule local_rule = {
+	.common = {
+		.refcnt =	ATOMIC_INIT(2),
+		.pref =		0,
+		.action =	FR_ACT_TO_TBL,
+		.table =	RT6_TABLE_LOCAL,
+		.flags =	FIB_RULE_PERMANENT,
+	},
+};
+
+static LIST_HEAD(fib6_rules);
+
+struct dst_entry *fib6_rule_lookup(struct flowi *fl, int flags,
+				   pol_lookup_t lookup)
+{
+	struct fib_lookup_arg arg = {
+		.lookup_ptr = lookup,
+	};
+
+	fib_rules_lookup(&fib6_rules_ops, fl, flags, &arg);
+	if (arg.rule)
+		fib_rule_put(arg.rule);
+
+	return (struct dst_entry *) arg.result;
+}
+
+int fib6_rule_action(struct fib_rule *rule, struct flowi *flp,
+		     int flags, struct fib_lookup_arg *arg)
+{
+	struct rt6_info *rt = NULL;
+	struct fib6_table *table;
+	pol_lookup_t lookup = arg->lookup_ptr;
+
+	switch (rule->action) {
+	case FR_ACT_TO_TBL:
+		break;
+	case FR_ACT_UNREACHABLE:
+		rt = &ip6_null_entry;
+		goto discard_pkt;
+	default:
+	case FR_ACT_BLACKHOLE:
+		rt = &ip6_blk_hole_entry;
+		goto discard_pkt;
+	case FR_ACT_PROHIBIT:
+		rt = &ip6_prohibit_entry;
+		goto discard_pkt;
+	}
+
+	table = fib6_get_table(rule->table);
+	if (table)
+		rt = lookup(table, flp, flags);
+
+	if (rt != &ip6_null_entry)
+		goto out;
+
+	dst_release(&rt->u.dst);
+discard_pkt:
+	dst_hold(&rt->u.dst);
+out:
+	arg->result = rt;
+	return rt == NULL ? -EAGAIN : 0;
+}
+
+
+static int fib6_rule_match(struct fib_rule *rule, struct flowi *fl, int flags)
+{
+	struct fib6_rule *r = (struct fib6_rule *) rule;
+
+	if (!ipv6_prefix_equal(&fl->fl6_dst, &r->dst.addr, r->dst.plen))
+		return 0;
+
+	if ((flags & RT6_F_HAS_SADDR) &&
+	    !ipv6_prefix_equal(&fl->fl6_src, &r->src.addr, r->src.plen))
+		return 0;
+
+	return 1;
+}
+
+static struct nla_policy fib6_rule_policy[RTA_MAX+1] __read_mostly = {
+	[FRA_IFNAME]	= { .type = NLA_STRING },
+	[FRA_PRIORITY]	= { .type = NLA_U32 },
+	[FRA_SRC]	= { .minlen = sizeof(struct in6_addr) },
+	[FRA_DST]	= { .minlen = sizeof(struct in6_addr) },
+};
+
+static int fib6_rule_configure(struct fib_rule *rule, struct sk_buff *skb,
+			       struct nlmsghdr *nlh, struct fib_rule_hdr *frh,
+			       struct nlattr **tb)
+{
+	int err = -EINVAL;
+	struct fib6_rule *rule6 = (struct fib6_rule *) rule;
+
+	if (frh->src_len > 128 || frh->dst_len > 128 ||
+	    (frh->tos & ~IPV6_FLOWINFO_MASK))
+		goto errout;
+
+	if (rule->action == FR_ACT_TO_TBL) {
+		if (rule->table == RT6_TABLE_UNSPEC)
+			goto errout;
+
+		if (fib6_new_table(rule->table) == NULL) {
+			err = -ENOBUFS;
+			goto errout;
+		}
+	}
+
+	if (tb[FRA_SRC])
+		nla_memcpy(&rule6->src.addr, tb[FRA_SRC],
+			   sizeof(struct in6_addr));
+
+	if (tb[FRA_DST])
+		nla_memcpy(&rule6->dst.addr, tb[FRA_DST],
+			   sizeof(struct in6_addr));
+
+	rule6->src.plen = frh->src_len;
+	rule6->dst.plen = frh->dst_len;
+	rule6->tclass = frh->tos;
+
+	err = 0;
+errout:
+	return err;
+}
+
+static int fib6_rule_compare(struct fib_rule *rule, struct fib_rule_hdr *frh,
+			     struct nlattr **tb)
+{
+	struct fib6_rule *rule6 = (struct fib6_rule *) rule;
+
+	if (frh->src_len && (rule6->src.plen != frh->src_len))
+		return 0;
+
+	if (frh->dst_len && (rule6->dst.plen != frh->dst_len))
+		return 0;
+
+	if (frh->tos && (rule6->tclass != frh->tos))
+		return 0;
+
+	if (tb[FRA_SRC] &&
+	    nla_memcmp(tb[FRA_SRC], &rule6->src.addr, sizeof(struct in6_addr)))
+		return 0;
+
+	if (tb[FRA_DST] &&
+	    nla_memcmp(tb[FRA_DST], &rule6->dst.addr, sizeof(struct in6_addr)))
+		return 0;
+
+	return 1;
+}
+
+static int fib6_rule_fill(struct fib_rule *rule, struct sk_buff *skb,
+			  struct nlmsghdr *nlh, struct fib_rule_hdr *frh)
+{
+	struct fib6_rule *rule6 = (struct fib6_rule *) rule;
+
+	frh->family = AF_INET6;
+	frh->dst_len = rule6->dst.plen;
+	frh->src_len = rule6->src.plen;
+	frh->tos = rule6->tclass;
+
+	if (rule6->dst.plen)
+		NLA_PUT(skb, FRA_DST, sizeof(struct in6_addr),
+			&rule6->dst.addr);
+
+	if (rule6->src.plen)
+		NLA_PUT(skb, FRA_SRC, sizeof(struct in6_addr),
+			&rule6->src.addr);
+
+	return 0;
+
+nla_put_failure:
+	return -ENOBUFS;
+}
+
+int fib6_rules_dump(struct sk_buff *skb, struct netlink_callback *cb)
+{
+	return fib_rules_dump(skb, cb, AF_INET6);
+}
+
+static u32 fib6_rule_default_pref(void)
+{
+	return 0x3FFF;
+}
+
+static struct fib_rules_ops fib6_rules_ops = {
+	.family			= AF_INET6,
+	.rule_size		= sizeof(struct fib6_rule),
+	.action			= fib6_rule_action,
+	.match			= fib6_rule_match,
+	.configure		= fib6_rule_configure,
+	.compare		= fib6_rule_compare,
+	.fill			= fib6_rule_fill,
+	.default_pref		= fib6_rule_default_pref,
+	.nlgroup		= RTNLGRP_IPV6_RULE,
+	.policy			= fib6_rule_policy,
+	.rules_list		= &fib6_rules,
+	.owner			= THIS_MODULE,
+};
+
+void __init fib6_rules_init(void)
+{
+	list_add_tail(&local_rule.common.list, &fib6_rules);
+	list_add_tail(&main_rule.common.list, &fib6_rules);
+
+	fib_rules_register(&fib6_rules_ops);
+}
+
+void fib6_rules_cleanup(void)
+{
+	fib_rules_unregister(&fib6_rules_ops);
+}
diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
index fcd7da830aca..ce226c14bef5 100644
--- a/net/ipv6/ip6_fib.c
+++ b/net/ipv6/ip6_fib.c
@@ -159,6 +159,15 @@ static struct fib6_table fib6_main_tbl = {
 
 #ifdef CONFIG_IPV6_MULTIPLE_TABLES
 
+static struct fib6_table fib6_local_tbl = {
+	.tb6_id		= RT6_TABLE_LOCAL,
+	.tb6_lock	= RW_LOCK_UNLOCKED,
+	.tb6_root 	= {
+		.leaf		= &ip6_null_entry,
+		.fn_flags	= RTN_ROOT | RTN_TL_ROOT | RTN_RTINFO,
+	},
+};
+
 #define FIB_TABLE_HASHSZ 256
 static struct hlist_head fib_table_hash[FIB_TABLE_HASHSZ];
 
@@ -228,20 +237,10 @@ struct fib6_table *fib6_get_table(u32 id)
 	return NULL;
 }
 
-struct dst_entry *fib6_rule_lookup(struct flowi *fl, int flags,
-				   pol_lookup_t lookup)
-{
-	/*
-	 * TODO: Add rule lookup
-	 */
-	struct fib6_table *table = fib6_get_table(RT6_TABLE_MAIN);
-
-	return (struct dst_entry *) lookup(table, fl, flags);
-}
-
 static void __init fib6_tables_init(void)
 {
 	fib6_link_table(&fib6_main_tbl);
+	fib6_link_table(&fib6_local_tbl);
 }
 
 #else
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 73efdadb9ab8..438977e2085d 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -140,6 +140,50 @@ struct rt6_info ip6_null_entry = {
 	.rt6i_ref	= ATOMIC_INIT(1),
 };
 
+#ifdef CONFIG_IPV6_MULTIPLE_TABLES
+
+struct rt6_info ip6_prohibit_entry = {
+	.u = {
+		.dst = {
+			.__refcnt	= ATOMIC_INIT(1),
+			.__use		= 1,
+			.dev		= &loopback_dev,
+			.obsolete	= -1,
+			.error		= -EACCES,
+			.metrics	= { [RTAX_HOPLIMIT - 1] = 255, },
+			.input		= ip6_pkt_discard,
+			.output		= ip6_pkt_discard_out,
+			.ops		= &ip6_dst_ops,
+			.path		= (struct dst_entry*)&ip6_prohibit_entry,
+		}
+	},
+	.rt6i_flags	= (RTF_REJECT | RTF_NONEXTHOP),
+	.rt6i_metric	= ~(u32) 0,
+	.rt6i_ref	= ATOMIC_INIT(1),
+};
+
+struct rt6_info ip6_blk_hole_entry = {
+	.u = {
+		.dst = {
+			.__refcnt	= ATOMIC_INIT(1),
+			.__use		= 1,
+			.dev		= &loopback_dev,
+			.obsolete	= -1,
+			.error		= -EINVAL,
+			.metrics	= { [RTAX_HOPLIMIT - 1] = 255, },
+			.input		= ip6_pkt_discard,
+			.output		= ip6_pkt_discard_out,
+			.ops		= &ip6_dst_ops,
+			.path		= (struct dst_entry*)&ip6_blk_hole_entry,
+		}
+	},
+	.rt6i_flags	= (RTF_REJECT | RTF_NONEXTHOP),
+	.rt6i_metric	= ~(u32) 0,
+	.rt6i_ref	= ATOMIC_INIT(1),
+};
+
+#endif
+
 /* allocate dst with ip6_dst_ops */
 static __inline__ struct rt6_info *ip6_dst_alloc(void)
 {
@@ -2408,10 +2452,16 @@ void __init ip6_route_init(void)
 #ifdef CONFIG_XFRM
 	xfrm6_init();
 #endif
+#ifdef CONFIG_IPV6_MULTIPLE_TABLES
+	fib6_rules_init();
+#endif
 }
 
 void ip6_route_cleanup(void)
 {
+#ifdef CONFIG_IPV6_MULTIPLE_TABLES
+	fib6_rules_cleanup();
+#endif
 #ifdef CONFIG_PROC_FS
 	proc_net_remove("ipv6_route");
 	proc_net_remove("rt6_stats");
-- 
cgit v1.2.3


From e1ef4bf23b1ced0bf78a1c98289f746486e5c912 Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Fri, 4 Aug 2006 03:39:22 -0700
Subject: [IPV4]: Use Protocol Independant Policy Routing Rules Framework

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip_fib.h    |  14 +-
 net/ipv4/Kconfig        |   1 +
 net/ipv4/devinet.c      |   4 +-
 net/ipv4/fib_frontend.c |   2 +-
 net/ipv4/fib_rules.c    | 605 +++++++++++++++++++-----------------------------
 5 files changed, 249 insertions(+), 377 deletions(-)

(limited to 'include')

diff --git a/include/net/ip_fib.h b/include/net/ip_fib.h
index a095d1dec7a4..14c82e611c95 100644
--- a/include/net/ip_fib.h
+++ b/include/net/ip_fib.h
@@ -18,6 +18,7 @@
 
 #include <net/flow.h>
 #include <linux/seq_file.h>
+#include <net/fib_rules.h>
 
 /* WARNING: The ordering of these elements must match ordering
  *          of RTA_* rtnetlink attribute numbers.
@@ -203,9 +204,8 @@ static inline void fib_select_default(const struct flowi *flp, struct fib_result
 #define ip_fib_main_table (fib_tables[RT_TABLE_MAIN])
 
 extern struct fib_table * fib_tables[RT_TABLE_MAX+1];
-extern int fib_lookup(const struct flowi *flp, struct fib_result *res);
+extern int fib_lookup(struct flowi *flp, struct fib_result *res);
 extern struct fib_table *__fib_new_table(int id);
-extern void fib_rule_put(struct fib_rule *r);
 
 static inline struct fib_table *fib_get_table(int id)
 {
@@ -251,15 +251,15 @@ extern u32  __fib_res_prefsrc(struct fib_result *res);
 extern struct fib_table *fib_hash_init(int id);
 
 #ifdef CONFIG_IP_MULTIPLE_TABLES
-/* Exported by fib_rules.c */
+extern int fib4_rules_dump(struct sk_buff *skb, struct netlink_callback *cb);
+
+extern void __init fib4_rules_init(void);
+extern void __exit fib4_rules_cleanup(void);
 
-extern int inet_rtm_delrule(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg);
-extern int inet_rtm_newrule(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg);
-extern int inet_dump_rules(struct sk_buff *skb, struct netlink_callback *cb);
 #ifdef CONFIG_NET_CLS_ROUTE
 extern u32 fib_rules_tclass(struct fib_result *res);
 #endif
-extern void fib_rules_init(void);
+
 #endif
 
 static inline void fib_combine_itag(u32 *itag, struct fib_result *res)
diff --git a/net/ipv4/Kconfig b/net/ipv4/Kconfig
index 3b5d504a74be..1650b64415aa 100644
--- a/net/ipv4/Kconfig
+++ b/net/ipv4/Kconfig
@@ -88,6 +88,7 @@ config IP_FIB_HASH
 config IP_MULTIPLE_TABLES
 	bool "IP: policy routing"
 	depends on IP_ADVANCED_ROUTER
+	select FIB_RULES
 	---help---
 	  Normally, a router decides what to do with a received packet based
 	  solely on the packet's final destination address. If you say Y here,
diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c
index a6cc31d911eb..9f3ffbec3296 100644
--- a/net/ipv4/devinet.c
+++ b/net/ipv4/devinet.c
@@ -1151,9 +1151,7 @@ static struct rtnetlink_link inet_rtnetlink_table[RTM_NR_MSGTYPES] = {
 	[RTM_GETROUTE - RTM_BASE] = { .doit	= inet_rtm_getroute,
 				      .dumpit	= inet_dump_fib,	},
 #ifdef CONFIG_IP_MULTIPLE_TABLES
-	[RTM_NEWRULE  - RTM_BASE] = { .doit	= inet_rtm_newrule,	},
-	[RTM_DELRULE  - RTM_BASE] = { .doit	= inet_rtm_delrule,	},
-	[RTM_GETRULE  - RTM_BASE] = { .dumpit	= inet_dump_rules,	},
+	[RTM_GETRULE  - RTM_BASE] = { .dumpit	= fib4_rules_dump,	},
 #endif
 };
 
diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c
index ba2a70745a63..fe4a53d4d10d 100644
--- a/net/ipv4/fib_frontend.c
+++ b/net/ipv4/fib_frontend.c
@@ -656,7 +656,7 @@ void __init ip_fib_init(void)
 	ip_fib_local_table = fib_hash_init(RT_TABLE_LOCAL);
 	ip_fib_main_table  = fib_hash_init(RT_TABLE_MAIN);
 #else
-	fib_rules_init();
+	fib4_rules_init();
 #endif
 
 	register_netdevice_notifier(&fib_netdev_notifier);
diff --git a/net/ipv4/fib_rules.c b/net/ipv4/fib_rules.c
index 79b04718bdfd..23ec6ae1a0f6 100644
--- a/net/ipv4/fib_rules.c
+++ b/net/ipv4/fib_rules.c
@@ -5,9 +5,8 @@
  *
  *		IPv4 Forwarding Information Base: policy rules.
  *
- * Version:	$Id: fib_rules.c,v 1.17 2001/10/31 21:55:54 davem Exp $
- *
  * Authors:	Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
+ * 		Thomas Graf <tgraf@suug.ch>
  *
  *		This program is free software; you can redistribute it and/or
  *		modify it under the terms of the GNU General Public License
@@ -19,129 +18,154 @@
  *		Marc Boucher	:	routing by fwmark
  */
 
-#include <asm/uaccess.h>
-#include <asm/system.h>
-#include <linux/bitops.h>
 #include <linux/types.h>
 #include <linux/kernel.h>
-#include <linux/sched.h>
-#include <linux/mm.h>
-#include <linux/string.h>
-#include <linux/socket.h>
-#include <linux/sockios.h>
-#include <linux/errno.h>
-#include <linux/in.h>
-#include <linux/inet.h>
-#include <linux/inetdevice.h>
 #include <linux/netdevice.h>
-#include <linux/if_arp.h>
-#include <linux/proc_fs.h>
-#include <linux/skbuff.h>
 #include <linux/netlink.h>
+#include <linux/inetdevice.h>
 #include <linux/init.h>
 #include <linux/list.h>
 #include <linux/rcupdate.h>
-
 #include <net/ip.h>
-#include <net/protocol.h>
 #include <net/route.h>
 #include <net/tcp.h>
-#include <net/sock.h>
 #include <net/ip_fib.h>
+#include <net/fib_rules.h>
 
-#define FRprintk(a...)
+static struct fib_rules_ops fib4_rules_ops;
 
-struct fib_rule
+struct fib4_rule
 {
-	struct hlist_node hlist;
-	atomic_t	r_clntref;
-	u32		r_preference;
-	unsigned char	r_table;
-	unsigned char	r_action;
-	unsigned char	r_dst_len;
-	unsigned char	r_src_len;
-	u32		r_src;
-	u32		r_srcmask;
-	u32		r_dst;
-	u32		r_dstmask;
-	u32		r_srcmap;
-	u8		r_flags;
-	u8		r_tos;
+	struct fib_rule		common;
+	u8			dst_len;
+	u8			src_len;
+	u8			tos;
+	u32			src;
+	u32			srcmask;
+	u32			dst;
+	u32			dstmask;
 #ifdef CONFIG_IP_ROUTE_FWMARK
-	u32		r_fwmark;
+	u32			fwmark;
 #endif
-	int		r_ifindex;
 #ifdef CONFIG_NET_CLS_ROUTE
-	__u32		r_tclassid;
+	u32			tclassid;
 #endif
-	char		r_ifname[IFNAMSIZ];
-	int		r_dead;
-	struct		rcu_head rcu;
 };
 
-static struct fib_rule default_rule = {
-	.r_clntref =	ATOMIC_INIT(2),
-	.r_preference =	0x7FFF,
-	.r_table =	RT_TABLE_DEFAULT,
-	.r_action =	RTN_UNICAST,
+static struct fib4_rule default_rule = {
+	.common = {
+		.refcnt =	ATOMIC_INIT(2),
+		.pref =		0x7FFF,
+		.table =	RT_TABLE_DEFAULT,
+		.action =	FR_ACT_TO_TBL,
+	},
 };
 
-static struct fib_rule main_rule = {
-	.r_clntref =	ATOMIC_INIT(2),
-	.r_preference =	0x7FFE,
-	.r_table =	RT_TABLE_MAIN,
-	.r_action =	RTN_UNICAST,
+static struct fib4_rule main_rule = {
+	.common = {
+		.refcnt =	ATOMIC_INIT(2),
+		.pref =		0x7FFE,
+		.table =	RT_TABLE_MAIN,
+		.action =	FR_ACT_TO_TBL,
+	},
 };
 
-static struct fib_rule local_rule = {
-	.r_clntref =	ATOMIC_INIT(2),
-	.r_table =	RT_TABLE_LOCAL,
-	.r_action =	RTN_UNICAST,
+static struct fib4_rule local_rule = {
+	.common = {
+		.refcnt =	ATOMIC_INIT(2),
+		.table =	RT_TABLE_LOCAL,
+		.action =	FR_ACT_TO_TBL,
+		.flags =	FIB_RULE_PERMANENT,
+	},
 };
 
-static struct hlist_head fib_rules;
+static LIST_HEAD(fib4_rules);
+
+#ifdef CONFIG_NET_CLS_ROUTE
+u32 fib_rules_tclass(struct fib_result *res)
+{
+	return res->r ? ((struct fib4_rule *) res->r)->tclassid : 0;
+}
+#endif
 
-/* writer func called from netlink -- rtnl_sem hold*/
+int fib_lookup(struct flowi *flp, struct fib_result *res)
+{
+	struct fib_lookup_arg arg = {
+		.result = res,
+	};
+	int err;
 
-static void rtmsg_rule(int, struct fib_rule *);
+	err = fib_rules_lookup(&fib4_rules_ops, flp, 0, &arg);
+	res->r = arg.rule;
 
-int inet_rtm_delrule(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
+	return err;
+}
+
+int fib4_rule_action(struct fib_rule *rule, struct flowi *flp, int flags,
+		     struct fib_lookup_arg *arg)
 {
-	struct rtattr **rta = arg;
-	struct rtmsg *rtm = NLMSG_DATA(nlh);
-	struct fib_rule *r;
-	struct hlist_node *node;
-	int err = -ESRCH;
-
-	hlist_for_each_entry(r, node, &fib_rules, hlist) {
-		if ((!rta[RTA_SRC-1] || memcmp(RTA_DATA(rta[RTA_SRC-1]), &r->r_src, 4) == 0) &&
-		    rtm->rtm_src_len == r->r_src_len &&
-		    rtm->rtm_dst_len == r->r_dst_len &&
-		    (!rta[RTA_DST-1] || memcmp(RTA_DATA(rta[RTA_DST-1]), &r->r_dst, 4) == 0) &&
-		    rtm->rtm_tos == r->r_tos &&
-#ifdef CONFIG_IP_ROUTE_FWMARK
-		    (!rta[RTA_PROTOINFO-1] || memcmp(RTA_DATA(rta[RTA_PROTOINFO-1]), &r->r_fwmark, 4) == 0) &&
-#endif
-		    (!rtm->rtm_type || rtm->rtm_type == r->r_action) &&
-		    (!rta[RTA_PRIORITY-1] || memcmp(RTA_DATA(rta[RTA_PRIORITY-1]), &r->r_preference, 4) == 0) &&
-		    (!rta[RTA_IIF-1] || rtattr_strcmp(rta[RTA_IIF-1], r->r_ifname) == 0) &&
-		    (!rtm->rtm_table || (r && rtm->rtm_table == r->r_table))) {
-			err = -EPERM;
-			if (r == &local_rule)
-				break;
-
-			hlist_del_rcu(&r->hlist);
-			r->r_dead = 1;
-			rtmsg_rule(RTM_DELRULE, r);
-			fib_rule_put(r);
-			err = 0;
-			break;
-		}
+	int err = -EAGAIN;
+	struct fib_table *tbl;
+
+	switch (rule->action) {
+	case FR_ACT_TO_TBL:
+		break;
+
+	case FR_ACT_UNREACHABLE:
+		err = -ENETUNREACH;
+		goto errout;
+
+	case FR_ACT_PROHIBIT:
+		err = -EACCES;
+		goto errout;
+
+	case FR_ACT_BLACKHOLE:
+	default:
+		err = -EINVAL;
+		goto errout;
 	}
+
+	if ((tbl = fib_get_table(rule->table)) == NULL)
+		goto errout;
+
+	err = tbl->tb_lookup(tbl, flp, (struct fib_result *) arg->result);
+	if (err > 0)
+		err = -EAGAIN;
+errout:
 	return err;
 }
 
-/* Allocate new unique table id */
+
+void fib_select_default(const struct flowi *flp, struct fib_result *res)
+{
+	if (res->r && res->r->action == FR_ACT_TO_TBL &&
+	    FIB_RES_GW(*res) && FIB_RES_NH(*res).nh_scope == RT_SCOPE_LINK) {
+		struct fib_table *tb;
+		if ((tb = fib_get_table(res->r->table)) != NULL)
+			tb->tb_select_default(tb, flp, res);
+	}
+}
+
+static int fib4_rule_match(struct fib_rule *rule, struct flowi *fl, int flags)
+{
+	struct fib4_rule *r = (struct fib4_rule *) rule;
+	u32 daddr = fl->fl4_dst;
+	u32 saddr = fl->fl4_src;
+
+	if (((saddr ^ r->src) & r->srcmask) ||
+	    ((daddr ^ r->dst) & r->dstmask))
+		return 0;
+
+	if (r->tos && (r->tos != fl->fl4_tos))
+		return 0;
+
+#ifdef CONFIG_IP_ROUTE_FWMARK
+	if (r->fwmark && (r->fwmark != fl->fl4_fwmark))
+		return 0;
+#endif
+
+	return 1;
+}
 
 static struct fib_table *fib_empty_table(void)
 {
@@ -153,329 +177,178 @@ static struct fib_table *fib_empty_table(void)
 	return NULL;
 }
 
-static inline void fib_rule_put_rcu(struct rcu_head *head)
-{
-	struct fib_rule *r = container_of(head, struct fib_rule, rcu);
-	kfree(r);
-}
+static struct nla_policy fib4_rule_policy[FRA_MAX+1] __read_mostly = {
+	[FRA_IFNAME]	= { .type = NLA_STRING },
+	[FRA_PRIORITY]	= { .type = NLA_U32 },
+	[FRA_SRC]	= { .type = NLA_U32 },
+	[FRA_DST]	= { .type = NLA_U32 },
+	[FRA_FWMARK]	= { .type = NLA_U32 },
+	[FRA_FLOW]	= { .type = NLA_U32 },
+};
 
-void fib_rule_put(struct fib_rule *r)
+static int fib4_rule_configure(struct fib_rule *rule, struct sk_buff *skb,
+			       struct nlmsghdr *nlh, struct fib_rule_hdr *frh,
+			       struct nlattr **tb)
 {
-	if (atomic_dec_and_test(&r->r_clntref)) {
-		if (r->r_dead)
-			call_rcu(&r->rcu, fib_rule_put_rcu);
-		else
-			printk("Freeing alive rule %p\n", r);
-	}
-}
+	int err = -EINVAL;
+	struct fib4_rule *rule4 = (struct fib4_rule *) rule;
 
-/* writer func called from netlink -- rtnl_sem hold*/
+	if (frh->src_len > 32 || frh->dst_len > 32 ||
+	    (frh->tos & ~IPTOS_TOS_MASK))
+		goto errout;
 
-int inet_rtm_newrule(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
-{
-	struct rtattr **rta = arg;
-	struct rtmsg *rtm = NLMSG_DATA(nlh);
-	struct fib_rule *r, *new_r, *last = NULL;
-	struct hlist_node *node = NULL;
-	unsigned char table_id;
-
-	if (rtm->rtm_src_len > 32 || rtm->rtm_dst_len > 32 ||
-	    (rtm->rtm_tos & ~IPTOS_TOS_MASK))
-		return -EINVAL;
-
-	if (rta[RTA_IIF-1] && RTA_PAYLOAD(rta[RTA_IIF-1]) > IFNAMSIZ)
-		return -EINVAL;
-
-	table_id = rtm->rtm_table;
-	if (table_id == RT_TABLE_UNSPEC) {
-		struct fib_table *table;
-		if (rtm->rtm_type == RTN_UNICAST) {
-			if ((table = fib_empty_table()) == NULL)
-				return -ENOBUFS;
-			table_id = table->tb_id;
-		}
-	}
+	if (rule->table == RT_TABLE_UNSPEC) {
+		if (rule->action == FR_ACT_TO_TBL) {
+			struct fib_table *table;
 
-	new_r = kzalloc(sizeof(*new_r), GFP_KERNEL);
-	if (!new_r)
-		return -ENOMEM;
-
-	if (rta[RTA_SRC-1])
-		memcpy(&new_r->r_src, RTA_DATA(rta[RTA_SRC-1]), 4);
-	if (rta[RTA_DST-1])
-		memcpy(&new_r->r_dst, RTA_DATA(rta[RTA_DST-1]), 4);
-	if (rta[RTA_GATEWAY-1])
-		memcpy(&new_r->r_srcmap, RTA_DATA(rta[RTA_GATEWAY-1]), 4);
-	new_r->r_src_len = rtm->rtm_src_len;
-	new_r->r_dst_len = rtm->rtm_dst_len;
-	new_r->r_srcmask = inet_make_mask(rtm->rtm_src_len);
-	new_r->r_dstmask = inet_make_mask(rtm->rtm_dst_len);
-	new_r->r_tos = rtm->rtm_tos;
-#ifdef CONFIG_IP_ROUTE_FWMARK
-	if (rta[RTA_PROTOINFO-1])
-		memcpy(&new_r->r_fwmark, RTA_DATA(rta[RTA_PROTOINFO-1]), 4);
-#endif
-	new_r->r_action = rtm->rtm_type;
-	new_r->r_flags = rtm->rtm_flags;
-	if (rta[RTA_PRIORITY-1])
-		memcpy(&new_r->r_preference, RTA_DATA(rta[RTA_PRIORITY-1]), 4);
-	new_r->r_table = table_id;
-	if (rta[RTA_IIF-1]) {
-		struct net_device *dev;
-		rtattr_strlcpy(new_r->r_ifname, rta[RTA_IIF-1], IFNAMSIZ);
-		new_r->r_ifindex = -1;
-		dev = __dev_get_by_name(new_r->r_ifname);
-		if (dev)
-			new_r->r_ifindex = dev->ifindex;
-	}
-#ifdef CONFIG_NET_CLS_ROUTE
-	if (rta[RTA_FLOW-1])
-		memcpy(&new_r->r_tclassid, RTA_DATA(rta[RTA_FLOW-1]), 4);
-#endif
-	r = container_of(fib_rules.first, struct fib_rule, hlist);
+			table = fib_empty_table();
+			if (table == NULL) {
+				err = -ENOBUFS;
+				goto errout;
+			}
 
-	if (!new_r->r_preference) {
-		if (r && r->hlist.next != NULL) {
-			r = container_of(r->hlist.next, struct fib_rule, hlist);
-			if (r->r_preference)
-				new_r->r_preference = r->r_preference - 1;
+			rule->table = table->tb_id;
 		}
 	}
 
-	hlist_for_each_entry(r, node, &fib_rules, hlist) {
-		if (r->r_preference > new_r->r_preference)
-			break;
-		last = r;
-	}
-	atomic_inc(&new_r->r_clntref);
+	if (tb[FRA_SRC])
+		rule4->src = nla_get_u32(tb[FRA_SRC]);
 
-	if (last)
-		hlist_add_after_rcu(&last->hlist, &new_r->hlist);
-	else
-		hlist_add_before_rcu(&new_r->hlist, &r->hlist);
+	if (tb[FRA_DST])
+		rule4->dst = nla_get_u32(tb[FRA_DST]);
 
-	rtmsg_rule(RTM_NEWRULE, new_r);
-	return 0;
-}
+#ifdef CONFIG_IP_ROUTE_FWMARK
+	if (tb[FRA_FWMARK])
+		rule4->fwmark = nla_get_u32(tb[FRA_FWMARK]);
+#endif
 
 #ifdef CONFIG_NET_CLS_ROUTE
-u32 fib_rules_tclass(struct fib_result *res)
-{
-	if (res->r)
-		return res->r->r_tclassid;
-	return 0;
-}
+	if (tb[FRA_FLOW])
+		rule4->tclassid = nla_get_u32(tb[FRA_FLOW]);
 #endif
 
-/* callers should hold rtnl semaphore */
-
-static void fib_rules_detach(struct net_device *dev)
-{
-	struct hlist_node *node;
-	struct fib_rule *r;
-
-	hlist_for_each_entry(r, node, &fib_rules, hlist) {
-		if (r->r_ifindex == dev->ifindex)
-			r->r_ifindex = -1;
+	rule4->src_len = frh->src_len;
+	rule4->srcmask = inet_make_mask(rule4->src_len);
+	rule4->dst_len = frh->dst_len;
+	rule4->dstmask = inet_make_mask(rule4->dst_len);
+	rule4->tos = frh->tos;
 
-	}
-}
-
-/* callers should hold rtnl semaphore */
-
-static void fib_rules_attach(struct net_device *dev)
-{
-	struct hlist_node *node;
-	struct fib_rule *r;
-
-	hlist_for_each_entry(r, node, &fib_rules, hlist) {
-		if (r->r_ifindex == -1 && strcmp(dev->name, r->r_ifname) == 0)
-			r->r_ifindex = dev->ifindex;
-	}
+	err = 0;
+errout:
+	return err;
 }
 
-int fib_lookup(const struct flowi *flp, struct fib_result *res)
+static int fib4_rule_compare(struct fib_rule *rule, struct fib_rule_hdr *frh,
+			     struct nlattr **tb)
 {
-	int err;
-	struct fib_rule *r, *policy;
-	struct fib_table *tb;
-	struct hlist_node *node;
+	struct fib4_rule *rule4 = (struct fib4_rule *) rule;
 
-	u32 daddr = flp->fl4_dst;
-	u32 saddr = flp->fl4_src;
+	if (frh->src_len && (rule4->src_len != frh->src_len))
+		return 0;
 
-FRprintk("Lookup: %u.%u.%u.%u <- %u.%u.%u.%u ",
-	NIPQUAD(flp->fl4_dst), NIPQUAD(flp->fl4_src));
+	if (frh->dst_len && (rule4->dst_len != frh->dst_len))
+		return 0;
 
-	rcu_read_lock();
+	if (frh->tos && (rule4->tos != frh->tos))
+		return 0;
 
-	hlist_for_each_entry_rcu(r, node, &fib_rules, hlist) {
-		if (((saddr^r->r_src) & r->r_srcmask) ||
-		    ((daddr^r->r_dst) & r->r_dstmask) ||
-		    (r->r_tos && r->r_tos != flp->fl4_tos) ||
 #ifdef CONFIG_IP_ROUTE_FWMARK
-		    (r->r_fwmark && r->r_fwmark != flp->fl4_fwmark) ||
+	if (tb[FRA_FWMARK] && (rule4->fwmark != nla_get_u32(tb[FRA_FWMARK])))
+		return 0;
 #endif
-		    (r->r_ifindex && r->r_ifindex != flp->iif))
-			continue;
-
-FRprintk("tb %d r %d ", r->r_table, r->r_action);
-		switch (r->r_action) {
-		case RTN_UNICAST:
-			policy = r;
-			break;
-		case RTN_UNREACHABLE:
-			rcu_read_unlock();
-			return -ENETUNREACH;
-		default:
-		case RTN_BLACKHOLE:
-			rcu_read_unlock();
-			return -EINVAL;
-		case RTN_PROHIBIT:
-			rcu_read_unlock();
-			return -EACCES;
-		}
 
-		if ((tb = fib_get_table(r->r_table)) == NULL)
-			continue;
-		err = tb->tb_lookup(tb, flp, res);
-		if (err == 0) {
-			res->r = policy;
-			if (policy)
-				atomic_inc(&policy->r_clntref);
-			rcu_read_unlock();
-			return 0;
-		}
-		if (err < 0 && err != -EAGAIN) {
-			rcu_read_unlock();
-			return err;
-		}
-	}
-FRprintk("FAILURE\n");
-	rcu_read_unlock();
-	return -ENETUNREACH;
-}
+#ifdef CONFIG_NET_CLS_ROUTE
+	if (tb[FRA_FLOW] && (rule4->tclassid != nla_get_u32(tb[FRA_FLOW])))
+		return 0;
+#endif
 
-void fib_select_default(const struct flowi *flp, struct fib_result *res)
-{
-	if (res->r && res->r->r_action == RTN_UNICAST &&
-	    FIB_RES_GW(*res) && FIB_RES_NH(*res).nh_scope == RT_SCOPE_LINK) {
-		struct fib_table *tb;
-		if ((tb = fib_get_table(res->r->r_table)) != NULL)
-			tb->tb_select_default(tb, flp, res);
-	}
-}
+	if (tb[FRA_SRC] && (rule4->src != nla_get_u32(tb[FRA_SRC])))
+		return 0;
 
-static int fib_rules_event(struct notifier_block *this, unsigned long event, void *ptr)
-{
-	struct net_device *dev = ptr;
+	if (tb[FRA_DST] && (rule4->dst != nla_get_u32(tb[FRA_DST])))
+		return 0;
 
-	if (event == NETDEV_UNREGISTER)
-		fib_rules_detach(dev);
-	else if (event == NETDEV_REGISTER)
-		fib_rules_attach(dev);
-	return NOTIFY_DONE;
+	return 1;
 }
 
+static int fib4_rule_fill(struct fib_rule *rule, struct sk_buff *skb,
+			  struct nlmsghdr *nlh, struct fib_rule_hdr *frh)
+{
+	struct fib4_rule *rule4 = (struct fib4_rule *) rule;
 
-static struct notifier_block fib_rules_notifier = {
-	.notifier_call =fib_rules_event,
-};
+	frh->family = AF_INET;
+	frh->dst_len = rule4->dst_len;
+	frh->src_len = rule4->src_len;
+	frh->tos = rule4->tos;
 
-static __inline__ int inet_fill_rule(struct sk_buff *skb,
-				     struct fib_rule *r,
-				     u32 pid, u32 seq, int event,
-				     unsigned int flags)
-{
-	struct rtmsg *rtm;
-	struct nlmsghdr  *nlh;
-	unsigned char	 *b = skb->tail;
-
-	nlh = NLMSG_NEW(skb, pid, seq, event, sizeof(*rtm), flags);
-	rtm = NLMSG_DATA(nlh);
-	rtm->rtm_family = AF_INET;
-	rtm->rtm_dst_len = r->r_dst_len;
-	rtm->rtm_src_len = r->r_src_len;
-	rtm->rtm_tos = r->r_tos;
 #ifdef CONFIG_IP_ROUTE_FWMARK
-	if (r->r_fwmark)
-		RTA_PUT(skb, RTA_PROTOINFO, 4, &r->r_fwmark);
+	if (rule4->fwmark)
+		NLA_PUT_U32(skb, FRA_FWMARK, rule4->fwmark);
 #endif
-	rtm->rtm_table = r->r_table;
-	rtm->rtm_protocol = 0;
-	rtm->rtm_scope = 0;
-	rtm->rtm_type = r->r_action;
-	rtm->rtm_flags = r->r_flags;
-
-	if (r->r_dst_len)
-		RTA_PUT(skb, RTA_DST, 4, &r->r_dst);
-	if (r->r_src_len)
-		RTA_PUT(skb, RTA_SRC, 4, &r->r_src);
-	if (r->r_ifname[0])
-		RTA_PUT(skb, RTA_IIF, IFNAMSIZ, &r->r_ifname);
-	if (r->r_preference)
-		RTA_PUT(skb, RTA_PRIORITY, 4, &r->r_preference);
-	if (r->r_srcmap)
-		RTA_PUT(skb, RTA_GATEWAY, 4, &r->r_srcmap);
+
+	if (rule4->dst_len)
+		NLA_PUT_U32(skb, FRA_DST, rule4->dst);
+
+	if (rule4->src_len)
+		NLA_PUT_U32(skb, FRA_SRC, rule4->src);
+
 #ifdef CONFIG_NET_CLS_ROUTE
-	if (r->r_tclassid)
-		RTA_PUT(skb, RTA_FLOW, 4, &r->r_tclassid);
+	if (rule4->tclassid)
+		NLA_PUT_U32(skb, FRA_FLOW, rule4->tclassid);
 #endif
-	nlh->nlmsg_len = skb->tail - b;
-	return skb->len;
+	return 0;
 
-nlmsg_failure:
-rtattr_failure:
-	skb_trim(skb, b - skb->data);
-	return -1;
+nla_put_failure:
+	return -ENOBUFS;
 }
 
-/* callers should hold rtnl semaphore */
+int fib4_rules_dump(struct sk_buff *skb, struct netlink_callback *cb)
+{
+	return fib_rules_dump(skb, cb, AF_INET);
+}
 
-static void rtmsg_rule(int event, struct fib_rule *r)
+static u32 fib4_rule_default_pref(void)
 {
-	int size = NLMSG_SPACE(sizeof(struct rtmsg) + 128);
-	struct sk_buff *skb = alloc_skb(size, GFP_KERNEL);
-
-	if (!skb)
-		netlink_set_err(rtnl, 0, RTNLGRP_IPV4_RULE, ENOBUFS);
-	else if (inet_fill_rule(skb, r, 0, 0, event, 0) < 0) {
-		kfree_skb(skb);
-		netlink_set_err(rtnl, 0, RTNLGRP_IPV4_RULE, EINVAL);
-	} else {
-		netlink_broadcast(rtnl, skb, 0, RTNLGRP_IPV4_RULE, GFP_KERNEL);
+	struct list_head *pos;
+	struct fib_rule *rule;
+
+	if (!list_empty(&fib4_rules)) {
+		pos = fib4_rules.next;
+		if (pos->next != &fib4_rules) {
+			rule = list_entry(pos->next, struct fib_rule, list);
+			if (rule->pref)
+				return rule->pref - 1;
+		}
 	}
+
+	return 0;
 }
 
-int inet_dump_rules(struct sk_buff *skb, struct netlink_callback *cb)
+static struct fib_rules_ops fib4_rules_ops = {
+	.family		= AF_INET,
+	.rule_size	= sizeof(struct fib4_rule),
+	.action		= fib4_rule_action,
+	.match		= fib4_rule_match,
+	.configure	= fib4_rule_configure,
+	.compare	= fib4_rule_compare,
+	.fill		= fib4_rule_fill,
+	.default_pref	= fib4_rule_default_pref,
+	.nlgroup	= RTNLGRP_IPV4_RULE,
+	.policy		= fib4_rule_policy,
+	.rules_list	= &fib4_rules,
+	.owner		= THIS_MODULE,
+};
+
+void __init fib4_rules_init(void)
 {
-	int idx = 0;
-	int s_idx = cb->args[0];
-	struct fib_rule *r;
-	struct hlist_node *node;
-
-	rcu_read_lock();
-	hlist_for_each_entry(r, node, &fib_rules, hlist) {
-		if (idx < s_idx)
-			goto next;
-		if (inet_fill_rule(skb, r, NETLINK_CB(cb->skb).pid,
-				   cb->nlh->nlmsg_seq,
-				   RTM_NEWRULE, NLM_F_MULTI) < 0)
-			break;
-next:
-		idx++;
-	}
-	rcu_read_unlock();
-	cb->args[0] = idx;
+	list_add_tail(&local_rule.common.list, &fib4_rules);
+	list_add_tail(&main_rule.common.list, &fib4_rules);
+	list_add_tail(&default_rule.common.list, &fib4_rules);
 
-	return skb->len;
+	fib_rules_register(&fib4_rules_ops);
 }
 
-void __init fib_rules_init(void)
+void __exit fib4_rules_cleanup(void)
 {
-	INIT_HLIST_HEAD(&fib_rules);
-	hlist_add_head(&local_rule.hlist, &fib_rules);
-	hlist_add_after(&local_rule.hlist, &main_rule.hlist);
-	hlist_add_after(&main_rule.hlist, &default_rule.hlist);
-	register_netdevice_notifier(&fib_rules_notifier);
+	fib_rules_unregister(&fib4_rules_ops);
 }
-- 
cgit v1.2.3


From fe4944e59c357f945f81bc67edb7ed1392e875ad Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Fri, 4 Aug 2006 23:03:05 -0700
Subject: [NETLINK]: Extend netlink messaging interface

Adds:
 nlmsg_get_pos()                 return current position in message
 nlmsg_trim()                    trim part of message
 nla_reserve_nohdr(skb, len)     reserve room for an attribute w/o hdr
 nla_put_nohdr(skb, len, data)   add attribute w/o hdr
 nla_find_nested()               find attribute in nested attributes

Fixes nlmsg_new() to take allocation flags and consider size.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/netlink.h   | 74 +++++++++++++++++++++++++++++++++++++++++-------
 kernel/taskstats.c      |  2 +-
 net/netlink/attr.c      | 75 +++++++++++++++++++++++++++++++++++++++++++++++++
 net/netlink/genetlink.c |  2 +-
 4 files changed, 141 insertions(+), 12 deletions(-)

(limited to 'include')

diff --git a/include/net/netlink.h b/include/net/netlink.h
index 640c26a90cf1..3a5e40b1e045 100644
--- a/include/net/netlink.h
+++ b/include/net/netlink.h
@@ -35,6 +35,8 @@
  *   nlmsg_put()			add a netlink message to an skb
  *   nlmsg_put_answer()			callback based nlmsg_put()
  *   nlmsg_end()			finanlize netlink message
+ *   nlmsg_get_pos()			return current position in message
+ *   nlmsg_trim()			trim part of message
  *   nlmsg_cancel()			cancel message construction
  *   nlmsg_free()			free a netlink message
  *
@@ -80,8 +82,10 @@
  *   struct nlattr			netlink attribtue header
  *
  * Attribute Construction:
- *   nla_reserve(skb, type, len)	reserve skb tailroom for an attribute
+ *   nla_reserve(skb, type, len)	reserve room for an attribute
+ *   nla_reserve_nohdr(skb, len)	reserve room for an attribute w/o hdr
  *   nla_put(skb, type, len, data)	add attribute to skb
+ *   nla_put_nohdr(skb, len, data)	add attribute w/o hdr
  *
  * Attribute Construction for Basic Types:
  *   nla_put_u8(skb, type, value)	add u8 attribute to skb
@@ -139,6 +143,7 @@
  *   nla_next(nla, remaining)		get next netlink attribute
  *   nla_validate()			validate a stream of attributes
  *   nla_find()				find attribute in stream of attributes
+ *   nla_find_nested()			find attribute in nested attributes
  *   nla_parse()			parse and validate stream of attrs
  *   nla_parse_nested()			parse nested attribuets
  *   nla_for_each_attr()		loop over all attributes
@@ -203,12 +208,18 @@ extern int		nla_memcmp(const struct nlattr *nla, const void *data,
 extern int		nla_strcmp(const struct nlattr *nla, const char *str);
 extern struct nlattr *	__nla_reserve(struct sk_buff *skb, int attrtype,
 				      int attrlen);
+extern void *		__nla_reserve_nohdr(struct sk_buff *skb, int attrlen);
 extern struct nlattr *	nla_reserve(struct sk_buff *skb, int attrtype,
 				    int attrlen);
+extern void *		nla_reserve_nohdr(struct sk_buff *skb, int attrlen);
 extern void		__nla_put(struct sk_buff *skb, int attrtype,
 				  int attrlen, const void *data);
+extern void		__nla_put_nohdr(struct sk_buff *skb, int attrlen,
+					const void *data);
 extern int		nla_put(struct sk_buff *skb, int attrtype,
 				int attrlen, const void *data);
+extern int		nla_put_nohdr(struct sk_buff *skb, int attrlen,
+				      const void *data);
 
 /**************************************************************************
  * Netlink Messages
@@ -453,12 +464,13 @@ static inline struct nlmsghdr *nlmsg_put_answer(struct sk_buff *skb,
 /**
  * nlmsg_new - Allocate a new netlink message
  * @size: maximum size of message
+ * @flags: the type of memory to allocate.
  *
  * Use NLMSG_GOODSIZE if size isn't know and you need a good default size.
  */
-static inline struct sk_buff *nlmsg_new(int size)
+static inline struct sk_buff *nlmsg_new(int size, gfp_t flags)
 {
-	return alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL);
+	return alloc_skb(size, flags);
 }
 
 /**
@@ -479,6 +491,32 @@ static inline int nlmsg_end(struct sk_buff *skb, struct nlmsghdr *nlh)
 	return skb->len;
 }
 
+/**
+ * nlmsg_get_pos - return current position in netlink message
+ * @skb: socket buffer the message is stored in
+ *
+ * Returns a pointer to the current tail of the message.
+ */
+static inline void *nlmsg_get_pos(struct sk_buff *skb)
+{
+	return skb->tail;
+}
+
+/**
+ * nlmsg_trim - Trim message to a mark
+ * @skb: socket buffer the message is stored in
+ * @mark: mark to trim to
+ *
+ * Trims the message to the provided mark. Returns -1.
+ */
+static inline int nlmsg_trim(struct sk_buff *skb, void *mark)
+{
+	if (mark)
+		skb_trim(skb, (unsigned char *) mark - skb->data);
+
+	return -1;
+}
+
 /**
  * nlmsg_cancel - Cancel construction of a netlink message
  * @skb: socket buffer the message is stored in
@@ -489,9 +527,7 @@ static inline int nlmsg_end(struct sk_buff *skb, struct nlmsghdr *nlh)
  */
 static inline int nlmsg_cancel(struct sk_buff *skb, struct nlmsghdr *nlh)
 {
-	skb_trim(skb, (unsigned char *) nlh - skb->data);
-
-	return -1;
+	return nlmsg_trim(skb, nlh);
 }
 
 /**
@@ -630,6 +666,18 @@ static inline struct nlattr *nla_next(const struct nlattr *nla, int *remaining)
 	return (struct nlattr *) ((char *) nla + totlen);
 }
 
+/**
+ * nla_find_nested - find attribute in a set of nested attributes
+ * @nla: attribute containing the nested attributes
+ * @attrtype: type of attribute to look for
+ *
+ * Returns the first attribute which matches the specified type.
+ */
+static inline struct nlattr *nla_find_nested(struct nlattr *nla, int attrtype)
+{
+	return nla_find(nla_data(nla), nla_len(nla), attrtype);
+}
+
 /**
  * nla_parse_nested - parse nested attributes
  * @tb: destination array with maxtype+1 elements
@@ -862,10 +910,7 @@ static inline int nla_nest_end(struct sk_buff *skb, struct nlattr *start)
  */
 static inline int nla_nest_cancel(struct sk_buff *skb, struct nlattr *start)
 {
-	if (start)
-		skb_trim(skb, (unsigned char *) start - skb->data);
-
-	return -1;
+	return nlmsg_trim(skb, start);
 }
 
 /**
@@ -880,4 +925,13 @@ static inline int nla_nest_cancel(struct sk_buff *skb, struct nlattr *start)
 	     nla_ok(pos, rem); \
 	     pos = nla_next(pos, &(rem)))
 
+/**
+ * nla_for_each_nested - iterate over nested attributes
+ * @pos: loop counter, set to current attribute
+ * @nla: attribute containing the nested attributes
+ * @rem: initialized to len, holds bytes currently remaining in stream
+ */
+#define nla_for_each_nested(pos, nla, rem) \
+	nla_for_each_attr(pos, nla_data(nla), nla_len(nla), rem)
+
 #endif
diff --git a/kernel/taskstats.c b/kernel/taskstats.c
index e78187657330..2ed4040d0dc5 100644
--- a/kernel/taskstats.c
+++ b/kernel/taskstats.c
@@ -75,7 +75,7 @@ static int prepare_reply(struct genl_info *info, u8 cmd, struct sk_buff **skbp,
 	/*
 	 * If new attributes are added, please revisit this allocation
 	 */
-	skb = nlmsg_new(size);
+	skb = nlmsg_new(size, GFP_KERNEL);
 	if (!skb)
 		return -ENOMEM;
 
diff --git a/net/netlink/attr.c b/net/netlink/attr.c
index dddbd15135a8..136e529e5780 100644
--- a/net/netlink/attr.c
+++ b/net/netlink/attr.c
@@ -254,6 +254,26 @@ struct nlattr *__nla_reserve(struct sk_buff *skb, int attrtype, int attrlen)
 	return nla;
 }
 
+/**
+ * __nla_reserve_nohdr - reserve room for attribute without header
+ * @skb: socket buffer to reserve room on
+ * @attrlen: length of attribute payload
+ *
+ * Reserves room for attribute payload without a header.
+ *
+ * The caller is responsible to ensure that the skb provides enough
+ * tailroom for the payload.
+ */
+void *__nla_reserve_nohdr(struct sk_buff *skb, int attrlen)
+{
+	void *start;
+
+	start = skb_put(skb, NLA_ALIGN(attrlen));
+	memset(start, 0, NLA_ALIGN(attrlen));
+
+	return start;
+}
+
 /**
  * nla_reserve - reserve room for attribute on the skb
  * @skb: socket buffer to reserve room on
@@ -274,6 +294,24 @@ struct nlattr *nla_reserve(struct sk_buff *skb, int attrtype, int attrlen)
 	return __nla_reserve(skb, attrtype, attrlen);
 }
 
+/**
+ * nla_reserve - reserve room for attribute without header
+ * @skb: socket buffer to reserve room on
+ * @len: length of attribute payload
+ *
+ * Reserves room for attribute payload without a header.
+ *
+ * Returns NULL if the tailroom of the skb is insufficient to store
+ * the attribute payload.
+ */
+void *nla_reserve_nohdr(struct sk_buff *skb, int attrlen)
+{
+	if (unlikely(skb_tailroom(skb) < NLA_ALIGN(attrlen)))
+		return NULL;
+
+	return __nla_reserve_nohdr(skb, attrlen);
+}
+
 /**
  * __nla_put - Add a netlink attribute to a socket buffer
  * @skb: socket buffer to add attribute to
@@ -293,6 +331,22 @@ void __nla_put(struct sk_buff *skb, int attrtype, int attrlen,
 	memcpy(nla_data(nla), data, attrlen);
 }
 
+/**
+ * __nla_put_nohdr - Add a netlink attribute without header
+ * @skb: socket buffer to add attribute to
+ * @attrlen: length of attribute payload
+ * @data: head of attribute payload
+ *
+ * The caller is responsible to ensure that the skb provides enough
+ * tailroom for the attribute payload.
+ */
+void __nla_put_nohdr(struct sk_buff *skb, int attrlen, const void *data)
+{
+	void *start;
+
+	start = __nla_reserve_nohdr(skb, attrlen);
+	memcpy(start, data, attrlen);
+}
 
 /**
  * nla_put - Add a netlink attribute to a socket buffer
@@ -313,15 +367,36 @@ int nla_put(struct sk_buff *skb, int attrtype, int attrlen, const void *data)
 	return 0;
 }
 
+/**
+ * nla_put_nohdr - Add a netlink attribute without header
+ * @skb: socket buffer to add attribute to
+ * @attrlen: length of attribute payload
+ * @data: head of attribute payload
+ *
+ * Returns -1 if the tailroom of the skb is insufficient to store
+ * the attribute payload.
+ */
+int nla_put_nohdr(struct sk_buff *skb, int attrlen, const void *data)
+{
+	if (unlikely(skb_tailroom(skb) < NLA_ALIGN(attrlen)))
+		return -1;
+
+	__nla_put_nohdr(skb, attrlen, data);
+	return 0;
+}
 
 EXPORT_SYMBOL(nla_validate);
 EXPORT_SYMBOL(nla_parse);
 EXPORT_SYMBOL(nla_find);
 EXPORT_SYMBOL(nla_strlcpy);
 EXPORT_SYMBOL(__nla_reserve);
+EXPORT_SYMBOL(__nla_reserve_nohdr);
 EXPORT_SYMBOL(nla_reserve);
+EXPORT_SYMBOL(nla_reserve_nohdr);
 EXPORT_SYMBOL(__nla_put);
+EXPORT_SYMBOL(__nla_put_nohdr);
 EXPORT_SYMBOL(nla_put);
+EXPORT_SYMBOL(nla_put_nohdr);
 EXPORT_SYMBOL(nla_memcpy);
 EXPORT_SYMBOL(nla_memcmp);
 EXPORT_SYMBOL(nla_strcmp);
diff --git a/net/netlink/genetlink.c b/net/netlink/genetlink.c
index a298f77cc3e3..75bb47a898dd 100644
--- a/net/netlink/genetlink.c
+++ b/net/netlink/genetlink.c
@@ -440,7 +440,7 @@ static struct sk_buff *ctrl_build_msg(struct genl_family *family, u32 pid,
 	struct sk_buff *skb;
 	int err;
 
-	skb = nlmsg_new(NLMSG_GOODSIZE);
+	skb = nlmsg_new(NLMSG_GOODSIZE, GFP_KERNEL);
 	if (skb == NULL)
 		return ERR_PTR(-ENOBUFS);
 
-- 
cgit v1.2.3


From 1823730fbc89fadde72a7bb3b7bdf03cc7b8835c Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Fri, 4 Aug 2006 23:04:54 -0700
Subject: [IPv4]: Move interface address bits to linux/if_addr.h

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/if_addr.h   | 53 ++++++++++++++++++++++++++++++++++++++++++++
 include/linux/rtnetlink.h | 56 -----------------------------------------------
 net/core/rtnetlink.c      |  1 +
 net/decnet/dn_dev.c       |  1 +
 net/ipv4/devinet.c        |  1 +
 net/ipv4/fib_frontend.c   |  1 +
 net/ipv6/addrconf.c       |  1 +
 net/ipv6/ndisc.c          |  1 +
 8 files changed, 59 insertions(+), 56 deletions(-)
 create mode 100644 include/linux/if_addr.h

(limited to 'include')

diff --git a/include/linux/if_addr.h b/include/linux/if_addr.h
new file mode 100644
index 000000000000..e1590454db59
--- /dev/null
+++ b/include/linux/if_addr.h
@@ -0,0 +1,53 @@
+#ifndef __LINUX_IF_ADDR_H
+#define __LINUX_IF_ADDR_H
+
+#include <linux/netlink.h>
+
+struct ifaddrmsg
+{
+	__u8		ifa_family;
+	__u8		ifa_prefixlen;	/* The prefix length		*/
+	__u8		ifa_flags;	/* Flags			*/
+	__u8		ifa_scope;	/* Address scope		*/
+	__u32		ifa_index;	/* Link index			*/
+};
+
+/*
+ * Important comment:
+ * IFA_ADDRESS is prefix address, rather than local interface address.
+ * It makes no difference for normally configured broadcast interfaces,
+ * but for point-to-point IFA_ADDRESS is DESTINATION address,
+ * local address is supplied in IFA_LOCAL attribute.
+ */
+enum
+{
+	IFA_UNSPEC,
+	IFA_ADDRESS,
+	IFA_LOCAL,
+	IFA_LABEL,
+	IFA_BROADCAST,
+	IFA_ANYCAST,
+	IFA_CACHEINFO,
+	IFA_MULTICAST,
+	__IFA_MAX,
+};
+
+#define IFA_MAX (__IFA_MAX - 1)
+
+/* ifa_flags */
+#define IFA_F_SECONDARY		0x01
+#define IFA_F_TEMPORARY		IFA_F_SECONDARY
+
+#define IFA_F_DEPRECATED	0x20
+#define IFA_F_TENTATIVE		0x40
+#define IFA_F_PERMANENT		0x80
+
+struct ifa_cacheinfo
+{
+	__u32	ifa_prefered;
+	__u32	ifa_valid;
+	__u32	cstamp; /* created timestamp, hundredths of seconds */
+	__u32	tstamp; /* updated timestamp, hundredths of seconds */
+};
+
+#endif
diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index bf353538ae93..890c4d4038b6 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -384,62 +384,6 @@ struct rta_session
 };
 
 
-/*********************************************************
- *		Interface address.
- ****/
-
-struct ifaddrmsg
-{
-	unsigned char	ifa_family;
-	unsigned char	ifa_prefixlen;	/* The prefix length		*/
-	unsigned char	ifa_flags;	/* Flags			*/
-	unsigned char	ifa_scope;	/* See above			*/
-	int		ifa_index;	/* Link index			*/
-};
-
-enum
-{
-	IFA_UNSPEC,
-	IFA_ADDRESS,
-	IFA_LOCAL,
-	IFA_LABEL,
-	IFA_BROADCAST,
-	IFA_ANYCAST,
-	IFA_CACHEINFO,
-	IFA_MULTICAST,
-	__IFA_MAX
-};
-
-#define IFA_MAX (__IFA_MAX - 1)
-
-/* ifa_flags */
-
-#define IFA_F_SECONDARY		0x01
-#define IFA_F_TEMPORARY		IFA_F_SECONDARY
-
-#define IFA_F_DEPRECATED	0x20
-#define IFA_F_TENTATIVE		0x40
-#define IFA_F_PERMANENT		0x80
-
-struct ifa_cacheinfo
-{
-	__u32	ifa_prefered;
-	__u32	ifa_valid;
-	__u32	cstamp; /* created timestamp, hundredths of seconds */
-	__u32	tstamp; /* updated timestamp, hundredths of seconds */
-};
-
-
-#define IFA_RTA(r)  ((struct rtattr*)(((char*)(r)) + NLMSG_ALIGN(sizeof(struct ifaddrmsg))))
-#define IFA_PAYLOAD(n) NLMSG_PAYLOAD(n,sizeof(struct ifaddrmsg))
-
-/*
-   Important comment:
-   IFA_ADDRESS is prefix address, rather than local interface address.
-   It makes no difference for normally configured broadcast interfaces,
-   but for point-to-point IFA_ADDRESS is DESTINATION address,
-   local address is supplied in IFA_LOCAL attribute.
- */
 
 /**************************************************************
  *		Neighbour discovery.
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index aa7cff2257b1..35712031e2c3 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -35,6 +35,7 @@
 #include <linux/init.h>
 #include <linux/security.h>
 #include <linux/mutex.h>
+#include <linux/if_addr.h>
 
 #include <asm/uaccess.h>
 #include <asm/system.h>
diff --git a/net/decnet/dn_dev.c b/net/decnet/dn_dev.c
index 476455fbdb03..632c5a90b589 100644
--- a/net/decnet/dn_dev.c
+++ b/net/decnet/dn_dev.c
@@ -34,6 +34,7 @@
 #include <linux/seq_file.h>
 #include <linux/timer.h>
 #include <linux/string.h>
+#include <linux/if_addr.h>
 #include <linux/if_arp.h>
 #include <linux/if_ether.h>
 #include <linux/skbuff.h>
diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c
index 80bf5b2ea2e6..398e7b9ca66b 100644
--- a/net/ipv4/devinet.c
+++ b/net/ipv4/devinet.c
@@ -43,6 +43,7 @@
 #include <linux/in.h>
 #include <linux/errno.h>
 #include <linux/interrupt.h>
+#include <linux/if_addr.h>
 #include <linux/if_ether.h>
 #include <linux/inet.h>
 #include <linux/netdevice.h>
diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c
index fe4a53d4d10d..a83f1aa8034e 100644
--- a/net/ipv4/fib_frontend.c
+++ b/net/ipv4/fib_frontend.c
@@ -32,6 +32,7 @@
 #include <linux/inet.h>
 #include <linux/inetdevice.h>
 #include <linux/netdevice.h>
+#include <linux/if_addr.h>
 #include <linux/if_arp.h>
 #include <linux/skbuff.h>
 #include <linux/netlink.h>
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index ed766eebc022..c2a4db843e51 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -48,6 +48,7 @@
 #include <linux/net.h>
 #include <linux/in6.h>
 #include <linux/netdevice.h>
+#include <linux/if_addr.h>
 #include <linux/if_arp.h>
 #include <linux/if_arcnet.h>
 #include <linux/if_infiniband.h>
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index 67cfc3813c32..5743e8bffefd 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -62,6 +62,7 @@
 #include <linux/sysctl.h>
 #endif
 
+#include <linux/if_addr.h>
 #include <linux/if_arp.h>
 #include <linux/ipv6.h>
 #include <linux/icmpv6.h>
-- 
cgit v1.2.3


From 0844565fb8a9418f5a860aa480c1aef70319c9a2 Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Fri, 4 Aug 2006 23:05:56 -0700
Subject: [NET]: Move netlink interface bits to linux/if.h

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/if.h        | 129 ++++++++++++++++++++++++++++++++++++++++++++
 include/linux/rtnetlink.h | 133 +---------------------------------------------
 2 files changed, 130 insertions(+), 132 deletions(-)

(limited to 'include')

diff --git a/include/linux/if.h b/include/linux/if.h
index 374e20ad8b0d..cd080d765324 100644
--- a/include/linux/if.h
+++ b/include/linux/if.h
@@ -212,5 +212,134 @@ struct ifconf
 #define	ifc_buf	ifc_ifcu.ifcu_buf		/* buffer address	*/
 #define	ifc_req	ifc_ifcu.ifcu_req		/* array of structures	*/
 
+/* The struct should be in sync with struct net_device_stats */
+struct rtnl_link_stats
+{
+	__u32	rx_packets;		/* total packets received	*/
+	__u32	tx_packets;		/* total packets transmitted	*/
+	__u32	rx_bytes;		/* total bytes received 	*/
+	__u32	tx_bytes;		/* total bytes transmitted	*/
+	__u32	rx_errors;		/* bad packets received		*/
+	__u32	tx_errors;		/* packet transmit problems	*/
+	__u32	rx_dropped;		/* no space in linux buffers	*/
+	__u32	tx_dropped;		/* no space available in linux	*/
+	__u32	multicast;		/* multicast packets received	*/
+	__u32	collisions;
+
+	/* detailed rx_errors: */
+	__u32	rx_length_errors;
+	__u32	rx_over_errors;		/* receiver ring buff overflow	*/
+	__u32	rx_crc_errors;		/* recved pkt with crc error	*/
+	__u32	rx_frame_errors;	/* recv'd frame alignment error */
+	__u32	rx_fifo_errors;		/* recv'r fifo overrun		*/
+	__u32	rx_missed_errors;	/* receiver missed packet	*/
+
+	/* detailed tx_errors */
+	__u32	tx_aborted_errors;
+	__u32	tx_carrier_errors;
+	__u32	tx_fifo_errors;
+	__u32	tx_heartbeat_errors;
+	__u32	tx_window_errors;
+
+	/* for cslip etc */
+	__u32	rx_compressed;
+	__u32	tx_compressed;
+};
+
+/* The struct should be in sync with struct ifmap */
+struct rtnl_link_ifmap
+{
+	__u64	mem_start;
+	__u64	mem_end;
+	__u64	base_addr;
+	__u16	irq;
+	__u8	dma;
+	__u8	port;
+};
+
+enum
+{
+	IFLA_UNSPEC,
+	IFLA_ADDRESS,
+	IFLA_BROADCAST,
+	IFLA_IFNAME,
+	IFLA_MTU,
+	IFLA_LINK,
+	IFLA_QDISC,
+	IFLA_STATS,
+	IFLA_COST,
+#define IFLA_COST IFLA_COST
+	IFLA_PRIORITY,
+#define IFLA_PRIORITY IFLA_PRIORITY
+	IFLA_MASTER,
+#define IFLA_MASTER IFLA_MASTER
+	IFLA_WIRELESS,		/* Wireless Extension event - see wireless.h */
+#define IFLA_WIRELESS IFLA_WIRELESS
+	IFLA_PROTINFO,		/* Protocol specific information for a link */
+#define IFLA_PROTINFO IFLA_PROTINFO
+	IFLA_TXQLEN,
+#define IFLA_TXQLEN IFLA_TXQLEN
+	IFLA_MAP,
+#define IFLA_MAP IFLA_MAP
+	IFLA_WEIGHT,
+#define IFLA_WEIGHT IFLA_WEIGHT
+	IFLA_OPERSTATE,
+	IFLA_LINKMODE,
+	__IFLA_MAX
+};
+
+
+#define IFLA_MAX (__IFLA_MAX - 1)
+
+/* ifi_flags.
+
+   IFF_* flags.
+
+   The only change is:
+   IFF_LOOPBACK, IFF_BROADCAST and IFF_POINTOPOINT are
+   more not changeable by user. They describe link media
+   characteristics and set by device driver.
+
+   Comments:
+   - Combination IFF_BROADCAST|IFF_POINTOPOINT is invalid
+   - If neither of these three flags are set;
+     the interface is NBMA.
+
+   - IFF_MULTICAST does not mean anything special:
+   multicasts can be used on all not-NBMA links.
+   IFF_MULTICAST means that this media uses special encapsulation
+   for multicast frames. Apparently, all IFF_POINTOPOINT and
+   IFF_BROADCAST devices are able to use multicasts too.
+ */
+
+/* IFLA_LINK.
+   For usual devices it is equal ifi_index.
+   If it is a "virtual interface" (f.e. tunnel), ifi_link
+   can point to real physical interface (f.e. for bandwidth calculations),
+   or maybe 0, what means, that real media is unknown (usual
+   for IPIP tunnels, when route to endpoint is allowed to change)
+ */
+
+/* Subtype attributes for IFLA_PROTINFO */
+enum
+{
+	IFLA_INET6_UNSPEC,
+	IFLA_INET6_FLAGS,	/* link flags			*/
+	IFLA_INET6_CONF,	/* sysctl parameters		*/
+	IFLA_INET6_STATS,	/* statistics			*/
+	IFLA_INET6_MCAST,	/* MC things. What of them?	*/
+	IFLA_INET6_CACHEINFO,	/* time values and max reasm size */
+	__IFLA_INET6_MAX
+};
+
+#define IFLA_INET6_MAX	(__IFLA_INET6_MAX - 1)
+
+struct ifla_cacheinfo
+{
+	__u32	max_reasm_len;
+	__u32	tstamp;		/* ipv6InterfaceTable updated timestamp */
+	__u32	reachable_time;
+	__u32	retrans_time;
+};
 
 #endif /* _LINUX_IF_H */
diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index 890c4d4038b6..84f3eb426da2 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -2,6 +2,7 @@
 #define __LINUX_RTNETLINK_H
 
 #include <linux/netlink.h>
+#include <linux/if.h>
 
 /****
  *		Routing/neighbour discovery messages.
@@ -607,138 +608,6 @@ struct prefix_cacheinfo
 	__u32	valid_time;
 };
 
-/* The struct should be in sync with struct net_device_stats */
-struct rtnl_link_stats
-{
-	__u32	rx_packets;		/* total packets received	*/
-	__u32	tx_packets;		/* total packets transmitted	*/
-	__u32	rx_bytes;		/* total bytes received 	*/
-	__u32	tx_bytes;		/* total bytes transmitted	*/
-	__u32	rx_errors;		/* bad packets received		*/
-	__u32	tx_errors;		/* packet transmit problems	*/
-	__u32	rx_dropped;		/* no space in linux buffers	*/
-	__u32	tx_dropped;		/* no space available in linux	*/
-	__u32	multicast;		/* multicast packets received	*/
-	__u32	collisions;
-
-	/* detailed rx_errors: */
-	__u32	rx_length_errors;
-	__u32	rx_over_errors;		/* receiver ring buff overflow	*/
-	__u32	rx_crc_errors;		/* recved pkt with crc error	*/
-	__u32	rx_frame_errors;	/* recv'd frame alignment error */
-	__u32	rx_fifo_errors;		/* recv'r fifo overrun		*/
-	__u32	rx_missed_errors;	/* receiver missed packet	*/
-
-	/* detailed tx_errors */
-	__u32	tx_aborted_errors;
-	__u32	tx_carrier_errors;
-	__u32	tx_fifo_errors;
-	__u32	tx_heartbeat_errors;
-	__u32	tx_window_errors;
-	
-	/* for cslip etc */
-	__u32	rx_compressed;
-	__u32	tx_compressed;
-};
-
-/* The struct should be in sync with struct ifmap */
-struct rtnl_link_ifmap
-{
-	__u64	mem_start;
-	__u64	mem_end;
-	__u64	base_addr;
-	__u16	irq;
-	__u8	dma;
-	__u8	port;
-};
-
-enum
-{
-	IFLA_UNSPEC,
-	IFLA_ADDRESS,
-	IFLA_BROADCAST,
-	IFLA_IFNAME,
-	IFLA_MTU,
-	IFLA_LINK,
-	IFLA_QDISC,
-	IFLA_STATS,
-	IFLA_COST,
-#define IFLA_COST IFLA_COST
-	IFLA_PRIORITY,
-#define IFLA_PRIORITY IFLA_PRIORITY
-	IFLA_MASTER,
-#define IFLA_MASTER IFLA_MASTER
-	IFLA_WIRELESS,		/* Wireless Extension event - see wireless.h */
-#define IFLA_WIRELESS IFLA_WIRELESS
-	IFLA_PROTINFO,		/* Protocol specific information for a link */
-#define IFLA_PROTINFO IFLA_PROTINFO
-	IFLA_TXQLEN,
-#define IFLA_TXQLEN IFLA_TXQLEN
-	IFLA_MAP,
-#define IFLA_MAP IFLA_MAP
-	IFLA_WEIGHT,
-#define IFLA_WEIGHT IFLA_WEIGHT
-	IFLA_OPERSTATE,
-	IFLA_LINKMODE,
-	__IFLA_MAX
-};
-
-
-#define IFLA_MAX (__IFLA_MAX - 1)
-
-#define IFLA_RTA(r)  ((struct rtattr*)(((char*)(r)) + NLMSG_ALIGN(sizeof(struct ifinfomsg))))
-#define IFLA_PAYLOAD(n) NLMSG_PAYLOAD(n,sizeof(struct ifinfomsg))
-
-/* ifi_flags.
-
-   IFF_* flags.
-
-   The only change is:
-   IFF_LOOPBACK, IFF_BROADCAST and IFF_POINTOPOINT are
-   more not changeable by user. They describe link media
-   characteristics and set by device driver.
-
-   Comments:
-   - Combination IFF_BROADCAST|IFF_POINTOPOINT is invalid
-   - If neither of these three flags are set;
-     the interface is NBMA.
-
-   - IFF_MULTICAST does not mean anything special:
-   multicasts can be used on all not-NBMA links.
-   IFF_MULTICAST means that this media uses special encapsulation
-   for multicast frames. Apparently, all IFF_POINTOPOINT and
-   IFF_BROADCAST devices are able to use multicasts too.
- */
-
-/* IFLA_LINK.
-   For usual devices it is equal ifi_index.
-   If it is a "virtual interface" (f.e. tunnel), ifi_link
-   can point to real physical interface (f.e. for bandwidth calculations),
-   or maybe 0, what means, that real media is unknown (usual
-   for IPIP tunnels, when route to endpoint is allowed to change)
- */
-
-/* Subtype attributes for IFLA_PROTINFO */
-enum
-{
-	IFLA_INET6_UNSPEC,
-	IFLA_INET6_FLAGS,	/* link flags			*/
-	IFLA_INET6_CONF,	/* sysctl parameters		*/
-	IFLA_INET6_STATS,	/* statistics			*/
-	IFLA_INET6_MCAST,	/* MC things. What of them?	*/
-	IFLA_INET6_CACHEINFO,	/* time values and max reasm size */
-	__IFLA_INET6_MAX
-};
-
-#define IFLA_INET6_MAX	(__IFLA_INET6_MAX - 1)
-
-struct ifla_cacheinfo
-{
-	__u32	max_reasm_len;
-	__u32	tstamp;		/* ipv6InterfaceTable updated timestamp */
-	__u32	reachable_time;
-	__u32	retrans_time;
-};
 
 /*****************************************************************
  *		Traffic control messages.
-- 
cgit v1.2.3


From 84fa7933a33f806bbbaae6775e87459b1ec584c0 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 29 Aug 2006 16:44:56 -0700
Subject: [NET]: Replace CHECKSUM_HW by CHECKSUM_PARTIAL/CHECKSUM_COMPLETE

Replace CHECKSUM_HW by CHECKSUM_PARTIAL (for outgoing packets, whose
checksum still needs to be completed) and CHECKSUM_COMPLETE (for
incoming packets, device supplied full checksum).

Patch originally from Herbert Xu, updated by myself for 2.6.18-rc3.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/atm/he.c                            |  2 +-
 drivers/net/3c59x.c                         |  2 +-
 drivers/net/8139cp.c                        |  6 +++---
 drivers/net/acenic.c                        |  8 ++++----
 drivers/net/bnx2.c                          |  2 +-
 drivers/net/cassini.c                       |  4 ++--
 drivers/net/chelsio/sge.c                   | 10 +++++-----
 drivers/net/dl2k.c                          |  2 +-
 drivers/net/e1000/e1000_main.c              |  8 ++++----
 drivers/net/forcedeth.c                     |  3 ++-
 drivers/net/gianfar.c                       |  2 +-
 drivers/net/hamachi.c                       |  2 +-
 drivers/net/ibm_emac/ibm_emac_core.c        |  2 +-
 drivers/net/ioc3-eth.c                      |  2 +-
 drivers/net/ixgb/ixgb_main.c                |  2 +-
 drivers/net/mv643xx_eth.c                   |  2 +-
 drivers/net/myri10ge/myri10ge.c             |  8 ++++----
 drivers/net/ns83820.c                       |  2 +-
 drivers/net/r8169.c                         |  2 +-
 drivers/net/s2io.c                          |  2 +-
 drivers/net/sk98lin/skge.c                  |  6 +++---
 drivers/net/skge.c                          |  4 ++--
 drivers/net/sky2.c                          |  6 +++---
 drivers/net/starfire.c                      |  6 +++---
 drivers/net/sungem.c                        |  4 ++--
 drivers/net/sunhme.c                        |  6 +++---
 drivers/net/tg3.c                           |  6 +++---
 drivers/net/typhoon.c                       |  2 +-
 drivers/net/via-rhine.c                     |  2 +-
 drivers/net/via-velocity.c                  |  2 +-
 include/linux/netdevice.h                   |  4 ++--
 include/linux/skbuff.h                      | 17 +++++++++--------
 net/core/datagram.c                         |  4 ++--
 net/core/dev.c                              | 12 ++++++------
 net/core/netpoll.c                          |  2 +-
 net/core/skbuff.c                           | 14 +++++++-------
 net/ipv4/icmp.c                             |  2 +-
 net/ipv4/igmp.c                             |  2 +-
 net/ipv4/ip_fragment.c                      |  2 +-
 net/ipv4/ip_gre.c                           |  4 ++--
 net/ipv4/ip_output.c                        |  6 +++---
 net/ipv4/ipvs/ip_vs_proto_tcp.c             |  8 ++++----
 net/ipv4/ipvs/ip_vs_proto_udp.c             |  8 ++++----
 net/ipv4/netfilter.c                        |  2 +-
 net/ipv4/netfilter/ip_conntrack_proto_tcp.c |  3 +--
 net/ipv4/netfilter/ip_conntrack_proto_udp.c |  3 +--
 net/ipv4/netfilter/ip_nat_standalone.c      |  5 +++--
 net/ipv4/netfilter/ip_queue.c               |  6 +++---
 net/ipv4/netfilter/ipt_ECN.c                |  9 +++++----
 net/ipv4/netfilter/ipt_TCPMSS.c             |  5 +++--
 net/ipv4/tcp.c                              |  8 ++++----
 net/ipv4/tcp_ipv4.c                         |  6 +++---
 net/ipv4/tcp_output.c                       | 18 ++++++++----------
 net/ipv4/udp.c                              |  6 +++---
 net/ipv4/xfrm4_output.c                     |  4 ++--
 net/ipv6/exthdrs.c                          |  2 +-
 net/ipv6/icmp.c                             |  2 +-
 net/ipv6/ip6_output.c                       |  2 +-
 net/ipv6/netfilter.c                        |  2 +-
 net/ipv6/netfilter/ip6_queue.c              |  6 +++---
 net/ipv6/netfilter/nf_conntrack_reasm.c     |  6 +++---
 net/ipv6/raw.c                              |  2 +-
 net/ipv6/reassembly.c                       |  6 +++---
 net/ipv6/tcp_ipv6.c                         |  6 +++---
 net/ipv6/udp.c                              |  2 +-
 net/ipv6/xfrm6_output.c                     |  4 ++--
 net/netfilter/nf_conntrack_proto_tcp.c      |  3 +--
 net/netfilter/nf_conntrack_proto_udp.c      |  3 +--
 net/netfilter/nfnetlink_queue.c             |  6 +++---
 net/packet/af_packet.c                      |  2 +-
 net/sched/sch_netem.c                       |  4 ++--
 net/sunrpc/socklib.c                        |  2 +-
 72 files changed, 168 insertions(+), 169 deletions(-)

(limited to 'include')

diff --git a/drivers/atm/he.c b/drivers/atm/he.c
index ffcb9fd31c38..41e052fecd7f 100644
--- a/drivers/atm/he.c
+++ b/drivers/atm/he.c
@@ -1912,7 +1912,7 @@ he_service_rbrq(struct he_dev *he_dev, int group)
 				skb->tail = skb->data + skb->len;
 #ifdef USE_CHECKSUM_HW
 				if (vcc->vpi == 0 && vcc->vci >= ATM_NOT_RSV_VCI) {
-					skb->ip_summed = CHECKSUM_HW;
+					skb->ip_summed = CHECKSUM_COMPLETE;
 					skb->csum = TCP_CKSUM(skb->data,
 							he_vcc->pdu_len);
 				}
diff --git a/drivers/net/3c59x.c b/drivers/net/3c59x.c
index 80e8ca013e44..29dede2eaa85 100644
--- a/drivers/net/3c59x.c
+++ b/drivers/net/3c59x.c
@@ -2077,7 +2077,7 @@ boomerang_start_xmit(struct sk_buff *skb, struct net_device *dev)
 
 	vp->tx_ring[entry].next = 0;
 #if DO_ZEROCOPY
-	if (skb->ip_summed != CHECKSUM_HW)
+	if (skb->ip_summed != CHECKSUM_PARTIAL)
 			vp->tx_ring[entry].status = cpu_to_le32(skb->len | TxIntrUploaded);
 	else
 			vp->tx_ring[entry].status = cpu_to_le32(skb->len | TxIntrUploaded | AddTCPChksum | AddUDPChksum);
diff --git a/drivers/net/8139cp.c b/drivers/net/8139cp.c
index 1428bb7715af..a48b211c489d 100644
--- a/drivers/net/8139cp.c
+++ b/drivers/net/8139cp.c
@@ -813,7 +813,7 @@ static int cp_start_xmit (struct sk_buff *skb, struct net_device *dev)
 
 		if (mss)
 			flags |= LargeSend | ((mss & MSSMask) << MSSShift);
-		else if (skb->ip_summed == CHECKSUM_HW) {
+		else if (skb->ip_summed == CHECKSUM_PARTIAL) {
 			const struct iphdr *ip = skb->nh.iph;
 			if (ip->protocol == IPPROTO_TCP)
 				flags |= IPCS | TCPCS;
@@ -867,7 +867,7 @@ static int cp_start_xmit (struct sk_buff *skb, struct net_device *dev)
 			if (mss)
 				ctrl |= LargeSend |
 					((mss & MSSMask) << MSSShift);
-			else if (skb->ip_summed == CHECKSUM_HW) {
+			else if (skb->ip_summed == CHECKSUM_PARTIAL) {
 				if (ip->protocol == IPPROTO_TCP)
 					ctrl |= IPCS | TCPCS;
 				else if (ip->protocol == IPPROTO_UDP)
@@ -898,7 +898,7 @@ static int cp_start_xmit (struct sk_buff *skb, struct net_device *dev)
 		txd->addr = cpu_to_le64(first_mapping);
 		wmb();
 
-		if (skb->ip_summed == CHECKSUM_HW) {
+		if (skb->ip_summed == CHECKSUM_PARTIAL) {
 			if (ip->protocol == IPPROTO_TCP)
 				txd->opts1 = cpu_to_le32(first_eor | first_len |
 							 FirstFrag | DescOwn |
diff --git a/drivers/net/acenic.c b/drivers/net/acenic.c
index 1c01e9b3d07c..826548644d7b 100644
--- a/drivers/net/acenic.c
+++ b/drivers/net/acenic.c
@@ -2040,7 +2040,7 @@ static void ace_rx_int(struct net_device *dev, u32 rxretprd, u32 rxretcsm)
 		 */
 		if (bd_flags & BD_FLG_TCP_UDP_SUM) {
 			skb->csum = htons(csum);
-			skb->ip_summed = CHECKSUM_HW;
+			skb->ip_summed = CHECKSUM_COMPLETE;
 		} else {
 			skb->ip_summed = CHECKSUM_NONE;
 		}
@@ -2511,7 +2511,7 @@ restart:
 
 		mapping = ace_map_tx_skb(ap, skb, skb, idx);
 		flagsize = (skb->len << 16) | (BD_FLG_END);
-		if (skb->ip_summed == CHECKSUM_HW)
+		if (skb->ip_summed == CHECKSUM_PARTIAL)
 			flagsize |= BD_FLG_TCP_UDP_SUM;
 #if ACENIC_DO_VLAN
 		if (vlan_tx_tag_present(skb)) {
@@ -2534,7 +2534,7 @@ restart:
 
 		mapping = ace_map_tx_skb(ap, skb, NULL, idx);
 		flagsize = (skb_headlen(skb) << 16);
-		if (skb->ip_summed == CHECKSUM_HW)
+		if (skb->ip_summed == CHECKSUM_PARTIAL)
 			flagsize |= BD_FLG_TCP_UDP_SUM;
 #if ACENIC_DO_VLAN
 		if (vlan_tx_tag_present(skb)) {
@@ -2560,7 +2560,7 @@ restart:
 					       PCI_DMA_TODEVICE);
 
 			flagsize = (frag->size << 16);
-			if (skb->ip_summed == CHECKSUM_HW)
+			if (skb->ip_summed == CHECKSUM_PARTIAL)
 				flagsize |= BD_FLG_TCP_UDP_SUM;
 			idx = (idx + 1) % ACE_TX_RING_ENTRIES(ap);
 
diff --git a/drivers/net/bnx2.c b/drivers/net/bnx2.c
index 652eb05a6c2d..7857b4630124 100644
--- a/drivers/net/bnx2.c
+++ b/drivers/net/bnx2.c
@@ -4423,7 +4423,7 @@ bnx2_start_xmit(struct sk_buff *skb, struct net_device *dev)
 	ring_prod = TX_RING_IDX(prod);
 
 	vlan_tag_flags = 0;
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		vlan_tag_flags |= TX_BD_FLAGS_TCP_UDP_CKSUM;
 	}
 
diff --git a/drivers/net/cassini.c b/drivers/net/cassini.c
index a31544ccb3c4..558fdb8ad2dc 100644
--- a/drivers/net/cassini.c
+++ b/drivers/net/cassini.c
@@ -2167,7 +2167,7 @@ end_copy_pkt:
 			cas_page_unmap(addr);
 	}
 	skb->csum = ntohs(i ^ 0xffff);
-	skb->ip_summed = CHECKSUM_HW;
+	skb->ip_summed = CHECKSUM_COMPLETE;
 	skb->protocol = eth_type_trans(skb, cp->dev);
 	return len;
 }
@@ -2821,7 +2821,7 @@ static inline int cas_xmit_tx_ringN(struct cas *cp, int ring,
 	}
 
 	ctrl = 0;
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		u64 csum_start_off, csum_stuff_off;
 
 		csum_start_off = (u64) (skb->h.raw - skb->data);
diff --git a/drivers/net/chelsio/sge.c b/drivers/net/chelsio/sge.c
index 61b3754f50ff..ddd0bdb498f4 100644
--- a/drivers/net/chelsio/sge.c
+++ b/drivers/net/chelsio/sge.c
@@ -1470,9 +1470,9 @@ int t1_start_xmit(struct sk_buff *skb, struct net_device *dev)
 		}
 
 		if (!(adapter->flags & UDP_CSUM_CAPABLE) &&
-		    skb->ip_summed == CHECKSUM_HW &&
+		    skb->ip_summed == CHECKSUM_PARTIAL &&
 		    skb->nh.iph->protocol == IPPROTO_UDP)
-			if (unlikely(skb_checksum_help(skb, 0))) {
+			if (unlikely(skb_checksum_help(skb))) {
 				dev_kfree_skb_any(skb);
 				return NETDEV_TX_OK;
 			}
@@ -1495,11 +1495,11 @@ int t1_start_xmit(struct sk_buff *skb, struct net_device *dev)
 		cpl = (struct cpl_tx_pkt *)__skb_push(skb, sizeof(*cpl));
 		cpl->opcode = CPL_TX_PKT;
 		cpl->ip_csum_dis = 1;    /* SW calculates IP csum */
-		cpl->l4_csum_dis = skb->ip_summed == CHECKSUM_HW ? 0 : 1;
+		cpl->l4_csum_dis = skb->ip_summed == CHECKSUM_PARTIAL ? 0 : 1;
 		/* the length field isn't used so don't bother setting it */
 
-		st->tx_cso += (skb->ip_summed == CHECKSUM_HW);
-		sge->stats.tx_do_cksum += (skb->ip_summed == CHECKSUM_HW);
+		st->tx_cso += (skb->ip_summed == CHECKSUM_PARTIAL);
+		sge->stats.tx_do_cksum += (skb->ip_summed == CHECKSUM_PARTIAL);
 		sge->stats.tx_reg_pkts++;
 	}
 	cpl->iff = dev->if_port;
diff --git a/drivers/net/dl2k.c b/drivers/net/dl2k.c
index 402961e68c89..b74e67654764 100644
--- a/drivers/net/dl2k.c
+++ b/drivers/net/dl2k.c
@@ -611,7 +611,7 @@ start_xmit (struct sk_buff *skb, struct net_device *dev)
 	txdesc = &np->tx_ring[entry];
 
 #if 0
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		txdesc->status |=
 		    cpu_to_le64 (TCPChecksumEnable | UDPChecksumEnable |
 				 IPChecksumEnable);
diff --git a/drivers/net/e1000/e1000_main.c b/drivers/net/e1000/e1000_main.c
index 98ef9f85482f..2ab9f96f5dab 100644
--- a/drivers/net/e1000/e1000_main.c
+++ b/drivers/net/e1000/e1000_main.c
@@ -2600,7 +2600,7 @@ e1000_tx_csum(struct e1000_adapter *adapter, struct e1000_tx_ring *tx_ring,
 	unsigned int i;
 	uint8_t css;
 
-	if (likely(skb->ip_summed == CHECKSUM_HW)) {
+	if (likely(skb->ip_summed == CHECKSUM_PARTIAL)) {
 		css = skb->h.raw - skb->data;
 
 		i = tx_ring->next_to_use;
@@ -2927,11 +2927,11 @@ e1000_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
 	}
 
 	/* reserve a descriptor for the offload context */
-	if ((mss) || (skb->ip_summed == CHECKSUM_HW))
+	if ((mss) || (skb->ip_summed == CHECKSUM_PARTIAL))
 		count++;
 	count++;
 #else
-	if (skb->ip_summed == CHECKSUM_HW)
+	if (skb->ip_summed == CHECKSUM_PARTIAL)
 		count++;
 #endif
 
@@ -3608,7 +3608,7 @@ e1000_rx_checksum(struct e1000_adapter *adapter,
 		 */
 		csum = ntohl(csum ^ 0xFFFF);
 		skb->csum = csum;
-		skb->ip_summed = CHECKSUM_HW;
+		skb->ip_summed = CHECKSUM_COMPLETE;
 	}
 	adapter->hw_csum_good++;
 }
diff --git a/drivers/net/forcedeth.c b/drivers/net/forcedeth.c
index 11b8f1b43dd5..32cacf115f75 100644
--- a/drivers/net/forcedeth.c
+++ b/drivers/net/forcedeth.c
@@ -1503,7 +1503,8 @@ static int nv_start_xmit(struct sk_buff *skb, struct net_device *dev)
 		tx_flags_extra = NV_TX2_TSO | (skb_shinfo(skb)->gso_size << NV_TX2_TSO_SHIFT);
 	else
 #endif
-	tx_flags_extra = (skb->ip_summed == CHECKSUM_HW ? (NV_TX2_CHECKSUM_L3|NV_TX2_CHECKSUM_L4) : 0);
+	tx_flags_extra = skb->ip_summed == CHECKSUM_PARTIAL ?
+			 NV_TX2_CHECKSUM_L3 | NV_TX2_CHECKSUM_L4 : 0;
 
 	/* vlan tag */
 	if (np->vlangrp && vlan_tx_tag_present(skb)) {
diff --git a/drivers/net/gianfar.c b/drivers/net/gianfar.c
index ebbbd6ca6204..ba960913c034 100644
--- a/drivers/net/gianfar.c
+++ b/drivers/net/gianfar.c
@@ -947,7 +947,7 @@ static int gfar_start_xmit(struct sk_buff *skb, struct net_device *dev)
 
 	/* Set up checksumming */
 	if (likely((dev->features & NETIF_F_IP_CSUM)
-			&& (CHECKSUM_HW == skb->ip_summed))) {
+			&& (CHECKSUM_PARTIAL == skb->ip_summed))) {
 		fcb = gfar_add_fcb(skb, txbdp);
 		status |= TXBD_TOE;
 		gfar_tx_checksum(skb, fcb);
diff --git a/drivers/net/hamachi.c b/drivers/net/hamachi.c
index 409c6aab0411..763373ae9666 100644
--- a/drivers/net/hamachi.c
+++ b/drivers/net/hamachi.c
@@ -1648,7 +1648,7 @@ static int hamachi_rx(struct net_device *dev)
 						* could do the pseudo myself and return
 						* CHECKSUM_UNNECESSARY
 						*/
-						skb->ip_summed = CHECKSUM_HW;
+						skb->ip_summed = CHECKSUM_COMPLETE;
 					}
 				}	
 			}
diff --git a/drivers/net/ibm_emac/ibm_emac_core.c b/drivers/net/ibm_emac/ibm_emac_core.c
index 82468e2dc799..57e214d85e9a 100644
--- a/drivers/net/ibm_emac/ibm_emac_core.c
+++ b/drivers/net/ibm_emac/ibm_emac_core.c
@@ -1036,7 +1036,7 @@ static inline u16 emac_tx_csum(struct ocp_enet_private *dev,
 			       struct sk_buff *skb)
 {
 #if defined(CONFIG_IBM_EMAC_TAH)
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		++dev->stats.tx_packets_csum;
 		return EMAC_TX_CTRL_TAH_CSUM;
 	}
diff --git a/drivers/net/ioc3-eth.c b/drivers/net/ioc3-eth.c
index 68d8af7df08e..65f897ddb920 100644
--- a/drivers/net/ioc3-eth.c
+++ b/drivers/net/ioc3-eth.c
@@ -1387,7 +1387,7 @@ static int ioc3_start_xmit(struct sk_buff *skb, struct net_device *dev)
 	 * MAC header which should not be summed and the TCP/UDP pseudo headers
 	 * manually.
 	 */
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		int proto = ntohs(skb->nh.iph->protocol);
 		unsigned int csoff;
 		struct iphdr *ih = skb->nh.iph;
diff --git a/drivers/net/ixgb/ixgb_main.c b/drivers/net/ixgb/ixgb_main.c
index 7bbd447289b5..9405b44f3214 100644
--- a/drivers/net/ixgb/ixgb_main.c
+++ b/drivers/net/ixgb/ixgb_main.c
@@ -1232,7 +1232,7 @@ ixgb_tx_csum(struct ixgb_adapter *adapter, struct sk_buff *skb)
 	unsigned int i;
 	uint8_t css, cso;
 
-	if(likely(skb->ip_summed == CHECKSUM_HW)) {
+	if(likely(skb->ip_summed == CHECKSUM_PARTIAL)) {
 		css = skb->h.raw - skb->data;
 		cso = (skb->h.raw + skb->csum) - skb->data;
 
diff --git a/drivers/net/mv643xx_eth.c b/drivers/net/mv643xx_eth.c
index eeab1df5bef3..38df58fdb358 100644
--- a/drivers/net/mv643xx_eth.c
+++ b/drivers/net/mv643xx_eth.c
@@ -1147,7 +1147,7 @@ static void eth_tx_submit_descs_for_skb(struct mv643xx_private *mp,
 	desc->byte_cnt = length;
 	desc->buf_ptr = dma_map_single(NULL, skb->data, length, DMA_TO_DEVICE);
 
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		BUG_ON(skb->protocol != ETH_P_IP);
 
 		cmd_sts |= ETH_GEN_TCP_UDP_CHECKSUM |
diff --git a/drivers/net/myri10ge/myri10ge.c b/drivers/net/myri10ge/myri10ge.c
index 9bdd43ab3573..9f16681d0e7e 100644
--- a/drivers/net/myri10ge/myri10ge.c
+++ b/drivers/net/myri10ge/myri10ge.c
@@ -930,7 +930,7 @@ static inline void myri10ge_vlan_ip_csum(struct sk_buff *skb, u16 hw_csum)
 	    (vh->h_vlan_encapsulated_proto == htons(ETH_P_IP) ||
 	     vh->h_vlan_encapsulated_proto == htons(ETH_P_IPV6))) {
 		skb->csum = hw_csum;
-		skb->ip_summed = CHECKSUM_HW;
+		skb->ip_summed = CHECKSUM_COMPLETE;
 	}
 }
 
@@ -973,7 +973,7 @@ myri10ge_rx_done(struct myri10ge_priv *mgp, struct myri10ge_rx_buf *rx,
 		if ((skb->protocol == ntohs(ETH_P_IP)) ||
 		    (skb->protocol == ntohs(ETH_P_IPV6))) {
 			skb->csum = ntohs((u16) csum);
-			skb->ip_summed = CHECKSUM_HW;
+			skb->ip_summed = CHECKSUM_COMPLETE;
 		} else
 			myri10ge_vlan_ip_csum(skb, ntohs((u16) csum));
 	}
@@ -1897,13 +1897,13 @@ again:
 	pseudo_hdr_offset = 0;
 	odd_flag = 0;
 	flags = (MXGEFW_FLAGS_NO_TSO | MXGEFW_FLAGS_FIRST);
-	if (likely(skb->ip_summed == CHECKSUM_HW)) {
+	if (likely(skb->ip_summed == CHECKSUM_PARTIAL)) {
 		cksum_offset = (skb->h.raw - skb->data);
 		pseudo_hdr_offset = (skb->h.raw + skb->csum) - skb->data;
 		/* If the headers are excessively large, then we must
 		 * fall back to a software checksum */
 		if (unlikely(cksum_offset > 255 || pseudo_hdr_offset > 127)) {
-			if (skb_checksum_help(skb, 0))
+			if (skb_checksum_help(skb))
 				goto drop;
 			cksum_offset = 0;
 			pseudo_hdr_offset = 0;
diff --git a/drivers/net/ns83820.c b/drivers/net/ns83820.c
index 0e76859c90a2..5143f5dbb2e5 100644
--- a/drivers/net/ns83820.c
+++ b/drivers/net/ns83820.c
@@ -1153,7 +1153,7 @@ again:
 	if (!nr_frags)
 		frag = NULL;
 	extsts = 0;
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		extsts |= EXTSTS_IPPKT;
 		if (IPPROTO_TCP == skb->nh.iph->protocol)
 			extsts |= EXTSTS_TCPPKT;
diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c
index 4c2f575faad7..d9b960aa9b0d 100644
--- a/drivers/net/r8169.c
+++ b/drivers/net/r8169.c
@@ -2169,7 +2169,7 @@ static inline u32 rtl8169_tso_csum(struct sk_buff *skb, struct net_device *dev)
 		if (mss)
 			return LargeSend | ((mss & MSSMask) << MSSShift);
 	}
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		const struct iphdr *ip = skb->nh.iph;
 
 		if (ip->protocol == IPPROTO_TCP)
diff --git a/drivers/net/s2io.c b/drivers/net/s2io.c
index e72e0e099060..5b3713f622d7 100644
--- a/drivers/net/s2io.c
+++ b/drivers/net/s2io.c
@@ -3893,7 +3893,7 @@ static int s2io_xmit(struct sk_buff *skb, struct net_device *dev)
 		txdp->Control_1 |= TXD_TCP_LSO_MSS(s2io_tcp_mss(skb));
 	}
 #endif
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		txdp->Control_2 |=
 		    (TXD_TX_CKO_IPV4_EN | TXD_TX_CKO_TCP_EN |
 		     TXD_TX_CKO_UDP_EN);
diff --git a/drivers/net/sk98lin/skge.c b/drivers/net/sk98lin/skge.c
index ee62845d3ac9..eb3b35180c2f 100644
--- a/drivers/net/sk98lin/skge.c
+++ b/drivers/net/sk98lin/skge.c
@@ -1559,7 +1559,7 @@ struct sk_buff	*pMessage)	/* pointer to send-message              */
 	pTxd->VDataHigh = (SK_U32) (PhysAddr >> 32);
 	pTxd->pMBuf     = pMessage;
 
-	if (pMessage->ip_summed == CHECKSUM_HW) {
+	if (pMessage->ip_summed == CHECKSUM_PARTIAL) {
 		u16 hdrlen = pMessage->h.raw - pMessage->data;
 		u16 offset = hdrlen + pMessage->csum;
 
@@ -1678,7 +1678,7 @@ struct sk_buff	*pMessage)	/* pointer to send-message              */
 	/* 
 	** Does the HW need to evaluate checksum for TCP or UDP packets? 
 	*/
-	if (pMessage->ip_summed == CHECKSUM_HW) {
+	if (pMessage->ip_summed == CHECKSUM_PARTIAL) {
 		u16 hdrlen = pMessage->h.raw - pMessage->data;
 		u16 offset = hdrlen + pMessage->csum;
 
@@ -2158,7 +2158,7 @@ rx_start:
 
 #ifdef USE_SK_RX_CHECKSUM
 		pMsg->csum = pRxd->TcpSums & 0xffff;
-		pMsg->ip_summed = CHECKSUM_HW;
+		pMsg->ip_summed = CHECKSUM_COMPLETE;
 #else
 		pMsg->ip_summed = CHECKSUM_NONE;
 #endif
diff --git a/drivers/net/skge.c b/drivers/net/skge.c
index ad878dfddef4..b3d6fa3d6df4 100644
--- a/drivers/net/skge.c
+++ b/drivers/net/skge.c
@@ -2338,7 +2338,7 @@ static int skge_xmit_frame(struct sk_buff *skb, struct net_device *dev)
 	td->dma_lo = map;
 	td->dma_hi = map >> 32;
 
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		int offset = skb->h.raw - skb->data;
 
 		/* This seems backwards, but it is what the sk98lin
@@ -2642,7 +2642,7 @@ static inline struct sk_buff *skge_rx_get(struct skge_port *skge,
 	skb->dev = skge->netdev;
 	if (skge->rx_csum) {
 		skb->csum = csum;
-		skb->ip_summed = CHECKSUM_HW;
+		skb->ip_summed = CHECKSUM_COMPLETE;
 	}
 
 	skb->protocol = eth_type_trans(skb, skge->netdev);
diff --git a/drivers/net/sky2.c b/drivers/net/sky2.c
index 933e87f1cc68..8e92566b587e 100644
--- a/drivers/net/sky2.c
+++ b/drivers/net/sky2.c
@@ -1163,7 +1163,7 @@ static unsigned tx_le_req(const struct sk_buff *skb)
 	if (skb_is_gso(skb))
 		++count;
 
-	if (skb->ip_summed == CHECKSUM_HW)
+	if (skb->ip_summed == CHECKSUM_PARTIAL)
 		++count;
 
 	return count;
@@ -1272,7 +1272,7 @@ static int sky2_xmit_frame(struct sk_buff *skb, struct net_device *dev)
 #endif
 
 	/* Handle TCP checksum offload */
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		u16 hdr = skb->h.raw - skb->data;
 		u16 offset = hdr + skb->csum;
 
@@ -2000,7 +2000,7 @@ static int sky2_status_intr(struct sky2_hw *hw, int to_do)
 #endif
 		case OP_RXCHKS:
 			skb = sky2->rx_ring[sky2->rx_next].skb;
-			skb->ip_summed = CHECKSUM_HW;
+			skb->ip_summed = CHECKSUM_COMPLETE;
 			skb->csum = le16_to_cpu(status);
 			break;
 
diff --git a/drivers/net/starfire.c b/drivers/net/starfire.c
index c0a62b00ffc8..2607aa51d8e0 100644
--- a/drivers/net/starfire.c
+++ b/drivers/net/starfire.c
@@ -1230,7 +1230,7 @@ static int start_tx(struct sk_buff *skb, struct net_device *dev)
 	}
 
 #if defined(ZEROCOPY) && defined(HAS_BROKEN_FIRMWARE)
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		if (skb_padto(skb, (skb->len + PADDING_MASK) & ~PADDING_MASK))
 			return NETDEV_TX_OK;
 	}
@@ -1252,7 +1252,7 @@ static int start_tx(struct sk_buff *skb, struct net_device *dev)
 				status |= TxDescIntr;
 				np->reap_tx = 0;
 			}
-			if (skb->ip_summed == CHECKSUM_HW) {
+			if (skb->ip_summed == CHECKSUM_PARTIAL) {
 				status |= TxCalTCP;
 				np->stats.tx_compressed++;
 			}
@@ -1499,7 +1499,7 @@ static int __netdev_rx(struct net_device *dev, int *quota)
 		 * Until then, the printk stays. :-) -Ion
 		 */
 		else if (le16_to_cpu(desc->status2) & 0x0040) {
-			skb->ip_summed = CHECKSUM_HW;
+			skb->ip_summed = CHECKSUM_COMPLETE;
 			skb->csum = le16_to_cpu(desc->csum);
 			printk(KERN_DEBUG "%s: checksum_hw, status2 = %#x\n", dev->name, le16_to_cpu(desc->status2));
 		}
diff --git a/drivers/net/sungem.c b/drivers/net/sungem.c
index d7b1d1882cab..b388651b7836 100644
--- a/drivers/net/sungem.c
+++ b/drivers/net/sungem.c
@@ -855,7 +855,7 @@ static int gem_rx(struct gem *gp, int work_to_do)
 		}
 
 		skb->csum = ntohs((status & RXDCTRL_TCPCSUM) ^ 0xffff);
-		skb->ip_summed = CHECKSUM_HW;
+		skb->ip_summed = CHECKSUM_COMPLETE;
 		skb->protocol = eth_type_trans(skb, gp->dev);
 
 		netif_receive_skb(skb);
@@ -1026,7 +1026,7 @@ static int gem_start_xmit(struct sk_buff *skb, struct net_device *dev)
 	unsigned long flags;
 
 	ctrl = 0;
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		u64 csum_start_off, csum_stuff_off;
 
 		csum_start_off = (u64) (skb->h.raw - skb->data);
diff --git a/drivers/net/sunhme.c b/drivers/net/sunhme.c
index c6f5bc3c042f..17981da22730 100644
--- a/drivers/net/sunhme.c
+++ b/drivers/net/sunhme.c
@@ -1207,7 +1207,7 @@ static void happy_meal_transceiver_check(struct happy_meal *hp, void __iomem *tr
  * flags, thus:
  *
  * 	skb->csum = rxd->rx_flags & 0xffff;
- * 	skb->ip_summed = CHECKSUM_HW;
+ * 	skb->ip_summed = CHECKSUM_COMPLETE;
  *
  * before sending off the skb to the protocols, and we are good as gold.
  */
@@ -2074,7 +2074,7 @@ static void happy_meal_rx(struct happy_meal *hp, struct net_device *dev)
 
 		/* This card is _fucking_ hot... */
 		skb->csum = ntohs(csum ^ 0xffff);
-		skb->ip_summed = CHECKSUM_HW;
+		skb->ip_summed = CHECKSUM_COMPLETE;
 
 		RXD(("len=%d csum=%4x]", len, csum));
 		skb->protocol = eth_type_trans(skb, dev);
@@ -2268,7 +2268,7 @@ static int happy_meal_start_xmit(struct sk_buff *skb, struct net_device *dev)
  	u32 tx_flags;
 
 	tx_flags = TXFLAG_OWN;
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		u32 csum_start_off, csum_stuff_off;
 
 		csum_start_off = (u32) (skb->h.raw - skb->data);
diff --git a/drivers/net/tg3.c b/drivers/net/tg3.c
index eafabb253f08..6f5d3a38c582 100644
--- a/drivers/net/tg3.c
+++ b/drivers/net/tg3.c
@@ -3851,11 +3851,11 @@ static int tg3_start_xmit(struct sk_buff *skb, struct net_device *dev)
 		skb->h.th->check = 0;
 
 	}
-	else if (skb->ip_summed == CHECKSUM_HW)
+	else if (skb->ip_summed == CHECKSUM_PARTIAL)
 		base_flags |= TXD_FLAG_TCPUDP_CSUM;
 #else
 	mss = 0;
-	if (skb->ip_summed == CHECKSUM_HW)
+	if (skb->ip_summed == CHECKSUM_PARTIAL)
 		base_flags |= TXD_FLAG_TCPUDP_CSUM;
 #endif
 #if TG3_VLAN_TAG_USED
@@ -3981,7 +3981,7 @@ static int tg3_start_xmit_dma_bug(struct sk_buff *skb, struct net_device *dev)
 
 	entry = tp->tx_prod;
 	base_flags = 0;
-	if (skb->ip_summed == CHECKSUM_HW)
+	if (skb->ip_summed == CHECKSUM_PARTIAL)
 		base_flags |= TXD_FLAG_TCPUDP_CSUM;
 #if TG3_TSO_SUPPORT != 0
 	mss = 0;
diff --git a/drivers/net/typhoon.c b/drivers/net/typhoon.c
index 4103c37172f9..c6e601dc6bbc 100644
--- a/drivers/net/typhoon.c
+++ b/drivers/net/typhoon.c
@@ -830,7 +830,7 @@ typhoon_start_tx(struct sk_buff *skb, struct net_device *dev)
 	first_txd->addrHi = (u64)((unsigned long) skb) >> 32;
 	first_txd->processFlags = 0;
 
-	if(skb->ip_summed == CHECKSUM_HW) {
+	if(skb->ip_summed == CHECKSUM_PARTIAL) {
 		/* The 3XP will figure out if this is UDP/TCP */
 		first_txd->processFlags |= TYPHOON_TX_PF_TCP_CHKSUM;
 		first_txd->processFlags |= TYPHOON_TX_PF_UDP_CHKSUM;
diff --git a/drivers/net/via-rhine.c b/drivers/net/via-rhine.c
index ae971080e2e4..66547159bfd9 100644
--- a/drivers/net/via-rhine.c
+++ b/drivers/net/via-rhine.c
@@ -1230,7 +1230,7 @@ static int rhine_start_tx(struct sk_buff *skb, struct net_device *dev)
 	rp->tx_skbuff[entry] = skb;
 
 	if ((rp->quirks & rqRhineI) &&
-	    (((unsigned long)skb->data & 3) || skb_shinfo(skb)->nr_frags != 0 || skb->ip_summed == CHECKSUM_HW)) {
+	    (((unsigned long)skb->data & 3) || skb_shinfo(skb)->nr_frags != 0 || skb->ip_summed == CHECKSUM_PARTIAL)) {
 		/* Must use alignment buffer. */
 		if (skb->len > PKT_BUF_SZ) {
 			/* packet too long, drop it */
diff --git a/drivers/net/via-velocity.c b/drivers/net/via-velocity.c
index aa9cd92f46b2..f1e0c746a388 100644
--- a/drivers/net/via-velocity.c
+++ b/drivers/net/via-velocity.c
@@ -2002,7 +2002,7 @@ static int velocity_xmit(struct sk_buff *skb, struct net_device *dev)
 	 *	Handle hardware checksum
 	 */
 	if ((vptr->flags & VELOCITY_FLAGS_TX_CSUM)
-				 && (skb->ip_summed == CHECKSUM_HW)) {
+				 && (skb->ip_summed == CHECKSUM_PARTIAL)) {
 		struct iphdr *ip = skb->nh.iph;
 		if (ip->protocol == IPPROTO_TCP)
 			td_ptr->tdesc1.TCR |= TCR0_TCPCK;
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 50a4719512ed..4f2c2b6beb5e 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -976,7 +976,7 @@ extern void		dev_mcast_init(void);
 extern int		netdev_max_backlog;
 extern int		weight_p;
 extern int		netdev_set_master(struct net_device *dev, struct net_device *master);
-extern int skb_checksum_help(struct sk_buff *skb, int inward);
+extern int skb_checksum_help(struct sk_buff *skb);
 extern struct sk_buff *skb_gso_segment(struct sk_buff *skb, int features);
 #ifdef CONFIG_BUG
 extern void netdev_rx_csum_fault(struct net_device *dev);
@@ -1012,7 +1012,7 @@ static inline int netif_needs_gso(struct net_device *dev, struct sk_buff *skb)
 {
 	return skb_is_gso(skb) &&
 	       (!skb_gso_ok(skb, dev->features) ||
-		unlikely(skb->ip_summed != CHECKSUM_HW));
+		unlikely(skb->ip_summed != CHECKSUM_PARTIAL));
 }
 
 /* On bonding slaves other than the currently active slave, suppress
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 755e9cddac47..85577a4ffa61 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -34,8 +34,9 @@
 #define HAVE_ALIGNABLE_SKB	/* Ditto 8)		   */
 
 #define CHECKSUM_NONE 0
-#define CHECKSUM_HW 1
+#define CHECKSUM_PARTIAL 1
 #define CHECKSUM_UNNECESSARY 2
+#define CHECKSUM_COMPLETE 3
 
 #define SKB_DATA_ALIGN(X)	(((X) + (SMP_CACHE_BYTES - 1)) & \
 				 ~(SMP_CACHE_BYTES - 1))
@@ -56,17 +57,17 @@
  *	      Apparently with secret goal to sell you new device, when you
  *	      will add new protocol to your host. F.e. IPv6. 8)
  *
- *	HW: the most generic way. Device supplied checksum of _all_
+ *	COMPLETE: the most generic way. Device supplied checksum of _all_
  *	    the packet as seen by netif_rx in skb->csum.
  *	    NOTE: Even if device supports only some protocols, but
- *	    is able to produce some skb->csum, it MUST use HW,
+ *	    is able to produce some skb->csum, it MUST use COMPLETE,
  *	    not UNNECESSARY.
  *
  * B. Checksumming on output.
  *
  *	NONE: skb is checksummed by protocol or csum is not required.
  *
- *	HW: device is required to csum packet as seen by hard_start_xmit
+ *	PARTIAL: device is required to csum packet as seen by hard_start_xmit
  *	from skb->h.raw to the end and to record the checksum
  *	at skb->h.raw+skb->csum.
  *
@@ -1261,14 +1262,14 @@ static inline int skb_linearize_cow(struct sk_buff *skb)
  *	@len: length of data pulled
  *
  *	After doing a pull on a received packet, you need to call this to
- *	update the CHECKSUM_HW checksum, or set ip_summed to CHECKSUM_NONE
- *	so that it can be recomputed from scratch.
+ *	update the CHECKSUM_COMPLETE checksum, or set ip_summed to
+ *	CHECKSUM_NONE so that it can be recomputed from scratch.
  */
 
 static inline void skb_postpull_rcsum(struct sk_buff *skb,
 				      const void *start, unsigned int len)
 {
-	if (skb->ip_summed == CHECKSUM_HW)
+	if (skb->ip_summed == CHECKSUM_COMPLETE)
 		skb->csum = csum_sub(skb->csum, csum_partial(start, len, 0));
 }
 
@@ -1287,7 +1288,7 @@ static inline int pskb_trim_rcsum(struct sk_buff *skb, unsigned int len)
 {
 	if (likely(len >= skb->len))
 		return 0;
-	if (skb->ip_summed == CHECKSUM_HW)
+	if (skb->ip_summed == CHECKSUM_COMPLETE)
 		skb->ip_summed = CHECKSUM_NONE;
 	return __pskb_trim(skb, len);
 }
diff --git a/net/core/datagram.c b/net/core/datagram.c
index aecddcc30401..f558c61aecc7 100644
--- a/net/core/datagram.c
+++ b/net/core/datagram.c
@@ -417,7 +417,7 @@ unsigned int __skb_checksum_complete(struct sk_buff *skb)
 
 	sum = (u16)csum_fold(skb_checksum(skb, 0, skb->len, skb->csum));
 	if (likely(!sum)) {
-		if (unlikely(skb->ip_summed == CHECKSUM_HW))
+		if (unlikely(skb->ip_summed == CHECKSUM_COMPLETE))
 			netdev_rx_csum_fault(skb->dev);
 		skb->ip_summed = CHECKSUM_UNNECESSARY;
 	}
@@ -462,7 +462,7 @@ int skb_copy_and_csum_datagram_iovec(struct sk_buff *skb,
 			goto fault;
 		if ((unsigned short)csum_fold(csum))
 			goto csum_error;
-		if (unlikely(skb->ip_summed == CHECKSUM_HW))
+		if (unlikely(skb->ip_summed == CHECKSUM_COMPLETE))
 			netdev_rx_csum_fault(skb->dev);
 		iov->iov_len -= chunk;
 		iov->iov_base += chunk;
diff --git a/net/core/dev.c b/net/core/dev.c
index d4a1ec3bded5..fc82f6f6e1c1 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -1166,12 +1166,12 @@ EXPORT_SYMBOL(netif_device_attach);
  * Invalidate hardware checksum when packet is to be mangled, and
  * complete checksum manually on outgoing path.
  */
-int skb_checksum_help(struct sk_buff *skb, int inward)
+int skb_checksum_help(struct sk_buff *skb)
 {
 	unsigned int csum;
 	int ret = 0, offset = skb->h.raw - skb->data;
 
-	if (inward)
+	if (skb->ip_summed == CHECKSUM_COMPLETE)
 		goto out_set_summed;
 
 	if (unlikely(skb_shinfo(skb)->gso_size)) {
@@ -1223,7 +1223,7 @@ struct sk_buff *skb_gso_segment(struct sk_buff *skb, int features)
 	skb->mac_len = skb->nh.raw - skb->data;
 	__skb_pull(skb, skb->mac_len);
 
-	if (unlikely(skb->ip_summed != CHECKSUM_HW)) {
+	if (unlikely(skb->ip_summed != CHECKSUM_PARTIAL)) {
 		if (skb_header_cloned(skb) &&
 		    (err = pskb_expand_head(skb, 0, 0, GFP_ATOMIC)))
 			return ERR_PTR(err);
@@ -1232,7 +1232,7 @@ struct sk_buff *skb_gso_segment(struct sk_buff *skb, int features)
 	rcu_read_lock();
 	list_for_each_entry_rcu(ptype, &ptype_base[ntohs(type) & 15], list) {
 		if (ptype->type == type && !ptype->dev && ptype->gso_segment) {
-			if (unlikely(skb->ip_summed != CHECKSUM_HW)) {
+			if (unlikely(skb->ip_summed != CHECKSUM_PARTIAL)) {
 				err = ptype->gso_send_check(skb);
 				segs = ERR_PTR(err);
 				if (err || skb_gso_ok(skb, features))
@@ -1444,11 +1444,11 @@ int dev_queue_xmit(struct sk_buff *skb)
 	/* If packet is not checksummed and device does not support
 	 * checksumming for this protocol, complete checksumming here.
 	 */
-	if (skb->ip_summed == CHECKSUM_HW &&
+	if (skb->ip_summed == CHECKSUM_PARTIAL &&
 	    (!(dev->features & NETIF_F_GEN_CSUM) &&
 	     (!(dev->features & NETIF_F_IP_CSUM) ||
 	      skb->protocol != htons(ETH_P_IP))))
-	      	if (skb_checksum_help(skb, 0))
+	      	if (skb_checksum_help(skb))
 	      		goto out_kfree_skb;
 
 gso:
diff --git a/net/core/netpoll.c b/net/core/netpoll.c
index 471da451cd48..ead5920c26d6 100644
--- a/net/core/netpoll.c
+++ b/net/core/netpoll.c
@@ -110,7 +110,7 @@ static int checksum_udp(struct sk_buff *skb, struct udphdr *uh,
 
 	psum = csum_tcpudp_nofold(saddr, daddr, ulen, IPPROTO_UDP, 0);
 
-	if (skb->ip_summed == CHECKSUM_HW &&
+	if (skb->ip_summed == CHECKSUM_COMPLETE &&
 	    !(u16)csum_fold(csum_add(psum, skb->csum)))
 		return 0;
 
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index c54f3664bce5..8a476f1956e5 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -1397,7 +1397,7 @@ void skb_copy_and_csum_dev(const struct sk_buff *skb, u8 *to)
 	unsigned int csum;
 	long csstart;
 
-	if (skb->ip_summed == CHECKSUM_HW)
+	if (skb->ip_summed == CHECKSUM_PARTIAL)
 		csstart = skb->h.raw - skb->data;
 	else
 		csstart = skb_headlen(skb);
@@ -1411,7 +1411,7 @@ void skb_copy_and_csum_dev(const struct sk_buff *skb, u8 *to)
 		csum = skb_copy_and_csum_bits(skb, csstart, to + csstart,
 					      skb->len - csstart, 0);
 
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		long csstuff = csstart + skb->csum;
 
 		*((unsigned short *)(to + csstuff)) = csum_fold(csum);
@@ -1898,10 +1898,10 @@ int skb_append_datato_frags(struct sock *sk, struct sk_buff *skb,
  *	@len: length of data pulled
  *
  *	This function performs an skb_pull on the packet and updates
- *	update the CHECKSUM_HW checksum.  It should be used on receive
- *	path processing instead of skb_pull unless you know that the
- *	checksum difference is zero (e.g., a valid IP header) or you
- *	are setting ip_summed to CHECKSUM_NONE.
+ *	update the CHECKSUM_COMPLETE checksum.  It should be used on
+ *	receive path processing instead of skb_pull unless you know
+ *	that the checksum difference is zero (e.g., a valid IP header)
+ *	or you are setting ip_summed to CHECKSUM_NONE.
  */
 unsigned char *skb_pull_rcsum(struct sk_buff *skb, unsigned int len)
 {
@@ -1994,7 +1994,7 @@ struct sk_buff *skb_segment(struct sk_buff *skb, int features)
 		frag = skb_shinfo(nskb)->frags;
 		k = 0;
 
-		nskb->ip_summed = CHECKSUM_HW;
+		nskb->ip_summed = CHECKSUM_PARTIAL;
 		nskb->csum = skb->csum;
 		memcpy(skb_put(nskb, hsize), skb->data + offset, hsize);
 
diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c
index 6ad797c14163..6d223e5c6741 100644
--- a/net/ipv4/icmp.c
+++ b/net/ipv4/icmp.c
@@ -930,7 +930,7 @@ int icmp_rcv(struct sk_buff *skb)
 	ICMP_INC_STATS_BH(ICMP_MIB_INMSGS);
 
 	switch (skb->ip_summed) {
-	case CHECKSUM_HW:
+	case CHECKSUM_COMPLETE:
 		if (!(u16)csum_fold(skb->csum))
 			break;
 		/* fall through */
diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c
index 8e8117c19e4d..7003e763d970 100644
--- a/net/ipv4/igmp.c
+++ b/net/ipv4/igmp.c
@@ -931,7 +931,7 @@ int igmp_rcv(struct sk_buff *skb)
 		goto drop;
 
 	switch (skb->ip_summed) {
-	case CHECKSUM_HW:
+	case CHECKSUM_COMPLETE:
 		if (!(u16)csum_fold(skb->csum))
 			break;
 		/* fall through */
diff --git a/net/ipv4/ip_fragment.c b/net/ipv4/ip_fragment.c
index b84b53a47526..8d7f107c2eef 100644
--- a/net/ipv4/ip_fragment.c
+++ b/net/ipv4/ip_fragment.c
@@ -665,7 +665,7 @@ static struct sk_buff *ip_frag_reasm(struct ipq *qp, struct net_device *dev)
 		head->len += fp->len;
 		if (head->ip_summed != fp->ip_summed)
 			head->ip_summed = CHECKSUM_NONE;
-		else if (head->ip_summed == CHECKSUM_HW)
+		else if (head->ip_summed == CHECKSUM_COMPLETE)
 			head->csum = csum_add(head->csum, fp->csum);
 		head->truesize += fp->truesize;
 		atomic_sub(fp->truesize, &ip_frag_mem);
diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
index 0f9b3a31997b..e66f6ff2e198 100644
--- a/net/ipv4/ip_gre.c
+++ b/net/ipv4/ip_gre.c
@@ -576,7 +576,7 @@ static int ipgre_rcv(struct sk_buff *skb)
 
 		if (flags&GRE_CSUM) {
 			switch (skb->ip_summed) {
-			case CHECKSUM_HW:
+			case CHECKSUM_COMPLETE:
 				csum = (u16)csum_fold(skb->csum);
 				if (!csum)
 					break;
@@ -584,7 +584,7 @@ static int ipgre_rcv(struct sk_buff *skb)
 			case CHECKSUM_NONE:
 				skb->csum = 0;
 				csum = __skb_checksum_complete(skb);
-				skb->ip_summed = CHECKSUM_HW;
+				skb->ip_summed = CHECKSUM_COMPLETE;
 			}
 			offset += 4;
 		}
diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index 308bdeac3455..1b9b6742ef77 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -680,7 +680,7 @@ ip_generic_getfrag(void *from, char *to, int offset, int len, int odd, struct sk
 {
 	struct iovec *iov = from;
 
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		if (memcpy_fromiovecend(to, iov, offset, len) < 0)
 			return -EFAULT;
 	} else {
@@ -736,7 +736,7 @@ static inline int ip_ufo_append_data(struct sock *sk,
 		/* initialize protocol header pointer */
 		skb->h.raw = skb->data + fragheaderlen;
 
-		skb->ip_summed = CHECKSUM_HW;
+		skb->ip_summed = CHECKSUM_PARTIAL;
 		skb->csum = 0;
 		sk->sk_sndmsg_off = 0;
 	}
@@ -844,7 +844,7 @@ int ip_append_data(struct sock *sk,
 	    length + fragheaderlen <= mtu &&
 	    rt->u.dst.dev->features & NETIF_F_ALL_CSUM &&
 	    !exthdrlen)
-		csummode = CHECKSUM_HW;
+		csummode = CHECKSUM_PARTIAL;
 
 	inet->cork.length += length;
 	if (((length > mtu) && (sk->sk_protocol == IPPROTO_UDP)) &&
diff --git a/net/ipv4/ipvs/ip_vs_proto_tcp.c b/net/ipv4/ipvs/ip_vs_proto_tcp.c
index bc28b1160a3a..820e8318d10d 100644
--- a/net/ipv4/ipvs/ip_vs_proto_tcp.c
+++ b/net/ipv4/ipvs/ip_vs_proto_tcp.c
@@ -151,7 +151,7 @@ tcp_snat_handler(struct sk_buff **pskb,
 		/* Only port and addr are changed, do fast csum update */
 		tcp_fast_csum_update(tcph, cp->daddr, cp->vaddr,
 				     cp->dport, cp->vport);
-		if ((*pskb)->ip_summed == CHECKSUM_HW)
+		if ((*pskb)->ip_summed == CHECKSUM_COMPLETE)
 			(*pskb)->ip_summed = CHECKSUM_NONE;
 	} else {
 		/* full checksum calculation */
@@ -204,7 +204,7 @@ tcp_dnat_handler(struct sk_buff **pskb,
 		/* Only port and addr are changed, do fast csum update */
 		tcp_fast_csum_update(tcph, cp->vaddr, cp->daddr,
 				     cp->vport, cp->dport);
-		if ((*pskb)->ip_summed == CHECKSUM_HW)
+		if ((*pskb)->ip_summed == CHECKSUM_COMPLETE)
 			(*pskb)->ip_summed = CHECKSUM_NONE;
 	} else {
 		/* full checksum calculation */
@@ -229,7 +229,7 @@ tcp_csum_check(struct sk_buff *skb, struct ip_vs_protocol *pp)
 	switch (skb->ip_summed) {
 	case CHECKSUM_NONE:
 		skb->csum = skb_checksum(skb, tcphoff, skb->len - tcphoff, 0);
-	case CHECKSUM_HW:
+	case CHECKSUM_COMPLETE:
 		if (csum_tcpudp_magic(skb->nh.iph->saddr, skb->nh.iph->daddr,
 				      skb->len - tcphoff,
 				      skb->nh.iph->protocol, skb->csum)) {
@@ -239,7 +239,7 @@ tcp_csum_check(struct sk_buff *skb, struct ip_vs_protocol *pp)
 		}
 		break;
 	default:
-		/* CHECKSUM_UNNECESSARY */
+		/* No need to checksum. */
 		break;
 	}
 
diff --git a/net/ipv4/ipvs/ip_vs_proto_udp.c b/net/ipv4/ipvs/ip_vs_proto_udp.c
index 89d9175d8f28..90c8166c0ec1 100644
--- a/net/ipv4/ipvs/ip_vs_proto_udp.c
+++ b/net/ipv4/ipvs/ip_vs_proto_udp.c
@@ -161,7 +161,7 @@ udp_snat_handler(struct sk_buff **pskb,
 		/* Only port and addr are changed, do fast csum update */
 		udp_fast_csum_update(udph, cp->daddr, cp->vaddr,
 				     cp->dport, cp->vport);
-		if ((*pskb)->ip_summed == CHECKSUM_HW)
+		if ((*pskb)->ip_summed == CHECKSUM_COMPLETE)
 			(*pskb)->ip_summed = CHECKSUM_NONE;
 	} else {
 		/* full checksum calculation */
@@ -216,7 +216,7 @@ udp_dnat_handler(struct sk_buff **pskb,
 		/* Only port and addr are changed, do fast csum update */
 		udp_fast_csum_update(udph, cp->vaddr, cp->daddr,
 				     cp->vport, cp->dport);
-		if ((*pskb)->ip_summed == CHECKSUM_HW)
+		if ((*pskb)->ip_summed == CHECKSUM_COMPLETE)
 			(*pskb)->ip_summed = CHECKSUM_NONE;
 	} else {
 		/* full checksum calculation */
@@ -250,7 +250,7 @@ udp_csum_check(struct sk_buff *skb, struct ip_vs_protocol *pp)
 		case CHECKSUM_NONE:
 			skb->csum = skb_checksum(skb, udphoff,
 						 skb->len - udphoff, 0);
-		case CHECKSUM_HW:
+		case CHECKSUM_COMPLETE:
 			if (csum_tcpudp_magic(skb->nh.iph->saddr,
 					      skb->nh.iph->daddr,
 					      skb->len - udphoff,
@@ -262,7 +262,7 @@ udp_csum_check(struct sk_buff *skb, struct ip_vs_protocol *pp)
 			}
 			break;
 		default:
-			/* CHECKSUM_UNNECESSARY */
+			/* No need to checksum. */
 			break;
 		}
 	}
diff --git a/net/ipv4/netfilter.c b/net/ipv4/netfilter.c
index 6a9e34b794bc..f88347de21a9 100644
--- a/net/ipv4/netfilter.c
+++ b/net/ipv4/netfilter.c
@@ -168,7 +168,7 @@ unsigned int nf_ip_checksum(struct sk_buff *skb, unsigned int hook,
 	unsigned int csum = 0;
 
 	switch (skb->ip_summed) {
-	case CHECKSUM_HW:
+	case CHECKSUM_COMPLETE:
 		if (hook != NF_IP_PRE_ROUTING && hook != NF_IP_LOCAL_IN)
 			break;
 		if ((protocol == 0 && !(u16)csum_fold(skb->csum)) ||
diff --git a/net/ipv4/netfilter/ip_conntrack_proto_tcp.c b/net/ipv4/netfilter/ip_conntrack_proto_tcp.c
index fb920e76ec10..9de81ff645d5 100644
--- a/net/ipv4/netfilter/ip_conntrack_proto_tcp.c
+++ b/net/ipv4/netfilter/ip_conntrack_proto_tcp.c
@@ -865,8 +865,7 @@ static int tcp_error(struct sk_buff *skb,
   
 	/* Checksum invalid? Ignore.
 	 * We skip checking packets on the outgoing path
-	 * because the semantic of CHECKSUM_HW is different there 
-	 * and moreover root might send raw packets.
+	 * because it is assumed to be correct.
 	 */
 	/* FIXME: Source route IP option packets --RR */
 	if (ip_conntrack_checksum && hooknum == NF_IP_PRE_ROUTING &&
diff --git a/net/ipv4/netfilter/ip_conntrack_proto_udp.c b/net/ipv4/netfilter/ip_conntrack_proto_udp.c
index 9b2c16b4d2ff..e58e52f14553 100644
--- a/net/ipv4/netfilter/ip_conntrack_proto_udp.c
+++ b/net/ipv4/netfilter/ip_conntrack_proto_udp.c
@@ -117,8 +117,7 @@ static int udp_error(struct sk_buff *skb, enum ip_conntrack_info *ctinfo,
 
 	/* Checksum invalid? Ignore.
 	 * We skip checking packets on the outgoing path
-	 * because the semantic of CHECKSUM_HW is different there 
-	 * and moreover root might send raw packets.
+	 * because the checksum is assumed to be correct.
 	 * FIXME: Source route IP option packets --RR */
 	if (ip_conntrack_checksum && hooknum == NF_IP_PRE_ROUTING &&
 	    nf_ip_checksum(skb, hooknum, iph->ihl * 4, IPPROTO_UDP)) {
diff --git a/net/ipv4/netfilter/ip_nat_standalone.c b/net/ipv4/netfilter/ip_nat_standalone.c
index 17de077a7901..f4f00c816d87 100644
--- a/net/ipv4/netfilter/ip_nat_standalone.c
+++ b/net/ipv4/netfilter/ip_nat_standalone.c
@@ -111,8 +111,9 @@ ip_nat_fn(unsigned int hooknum,
 		       & htons(IP_MF|IP_OFFSET)));
 
 	/* If we had a hardware checksum before, it's now invalid */
-	if ((*pskb)->ip_summed == CHECKSUM_HW)
-		if (skb_checksum_help(*pskb, (out == NULL)))
+	if ((*pskb)->ip_summed == CHECKSUM_PARTIAL ||
+	    (*pskb)->ip_summed == CHECKSUM_COMPLETE)
+		if (skb_checksum_help(*pskb))
 			return NF_DROP;
 
 	ct = ip_conntrack_get(*pskb, &ctinfo);
diff --git a/net/ipv4/netfilter/ip_queue.c b/net/ipv4/netfilter/ip_queue.c
index 198ac36db861..276a964ee6cf 100644
--- a/net/ipv4/netfilter/ip_queue.c
+++ b/net/ipv4/netfilter/ip_queue.c
@@ -208,9 +208,9 @@ ipq_build_packet_message(struct ipq_queue_entry *entry, int *errp)
 		break;
 	
 	case IPQ_COPY_PACKET:
-		if (entry->skb->ip_summed == CHECKSUM_HW &&
-		    (*errp = skb_checksum_help(entry->skb,
-		                               entry->info->outdev == NULL))) {
+		if ((entry->skb->ip_summed == CHECKSUM_PARTIAL ||
+		     entry->skb->ip_summed == CHECKSUM_COMPLETE) &&
+		    (*errp = skb_checksum_help(entry->skb))) {
 			read_unlock_bh(&queue_lock);
 			return NULL;
 		}
diff --git a/net/ipv4/netfilter/ipt_ECN.c b/net/ipv4/netfilter/ipt_ECN.c
index 4adf5c9d34f5..4ec43f98fe49 100644
--- a/net/ipv4/netfilter/ipt_ECN.c
+++ b/net/ipv4/netfilter/ipt_ECN.c
@@ -49,7 +49,7 @@ set_ect_ip(struct sk_buff **pskb, const struct ipt_ECN_info *einfo)
 
 /* Return 0 if there was an error. */
 static inline int
-set_ect_tcp(struct sk_buff **pskb, const struct ipt_ECN_info *einfo, int inward)
+set_ect_tcp(struct sk_buff **pskb, const struct ipt_ECN_info *einfo)
 {
 	struct tcphdr _tcph, *tcph;
 	u_int16_t diffs[2];
@@ -70,8 +70,9 @@ set_ect_tcp(struct sk_buff **pskb, const struct ipt_ECN_info *einfo, int inward)
 		return 0;
 	tcph = (void *)(*pskb)->nh.iph + (*pskb)->nh.iph->ihl*4;
 
-	if ((*pskb)->ip_summed == CHECKSUM_HW &&
-	    skb_checksum_help(*pskb, inward))
+	if (((*pskb)->ip_summed == CHECKSUM_PARTIAL ||
+	     (*pskb)->ip_summed == CHECKSUM_COMPLETE) &&
+	    skb_checksum_help(*pskb))
 		return 0;
 
 	diffs[0] = ((u_int16_t *)tcph)[6];
@@ -106,7 +107,7 @@ target(struct sk_buff **pskb,
 
 	if (einfo->operation & (IPT_ECN_OP_SET_ECE | IPT_ECN_OP_SET_CWR)
 	    && (*pskb)->nh.iph->protocol == IPPROTO_TCP)
-		if (!set_ect_tcp(pskb, einfo, (out == NULL)))
+		if (!set_ect_tcp(pskb, einfo))
 			return NF_DROP;
 
 	return IPT_CONTINUE;
diff --git a/net/ipv4/netfilter/ipt_TCPMSS.c b/net/ipv4/netfilter/ipt_TCPMSS.c
index ef2fe5b3f0d8..c998dc0fcd15 100644
--- a/net/ipv4/netfilter/ipt_TCPMSS.c
+++ b/net/ipv4/netfilter/ipt_TCPMSS.c
@@ -62,8 +62,9 @@ ipt_tcpmss_target(struct sk_buff **pskb,
 	if (!skb_make_writable(pskb, (*pskb)->len))
 		return NF_DROP;
 
-	if ((*pskb)->ip_summed == CHECKSUM_HW &&
-	    skb_checksum_help(*pskb, out == NULL))
+	if (((*pskb)->ip_summed == CHECKSUM_PARTIAL ||
+	     (*pskb)->ip_summed == CHECKSUM_COMPLETE) &&
+	    skb_checksum_help(*pskb))
 		return NF_DROP;
 
 	iph = (*pskb)->nh.iph;
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 934396bb1376..b0124e69ab38 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -568,7 +568,7 @@ new_segment:
 		skb->truesize += copy;
 		sk->sk_wmem_queued += copy;
 		sk->sk_forward_alloc -= copy;
-		skb->ip_summed = CHECKSUM_HW;
+		skb->ip_summed = CHECKSUM_PARTIAL;
 		tp->write_seq += copy;
 		TCP_SKB_CB(skb)->end_seq += copy;
 		skb_shinfo(skb)->gso_segs = 0;
@@ -723,7 +723,7 @@ new_segment:
 				 * Check whether we can use HW checksum.
 				 */
 				if (sk->sk_route_caps & NETIF_F_ALL_CSUM)
-					skb->ip_summed = CHECKSUM_HW;
+					skb->ip_summed = CHECKSUM_PARTIAL;
 
 				skb_entail(sk, tp, skb);
 				copy = size_goal;
@@ -2205,7 +2205,7 @@ struct sk_buff *tcp_tso_segment(struct sk_buff *skb, int features)
 		th->fin = th->psh = 0;
 
 		th->check = ~csum_fold(th->check + delta);
-		if (skb->ip_summed != CHECKSUM_HW)
+		if (skb->ip_summed != CHECKSUM_PARTIAL)
 			th->check = csum_fold(csum_partial(skb->h.raw, thlen,
 							   skb->csum));
 
@@ -2219,7 +2219,7 @@ struct sk_buff *tcp_tso_segment(struct sk_buff *skb, int features)
 
 	delta = htonl(oldlen + (skb->tail - skb->h.raw) + skb->data_len);
 	th->check = ~csum_fold(th->check + delta);
-	if (skb->ip_summed != CHECKSUM_HW)
+	if (skb->ip_summed != CHECKSUM_PARTIAL)
 		th->check = csum_fold(csum_partial(skb->h.raw, thlen,
 						   skb->csum));
 
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 43f6740244f8..b2aa512a30e9 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -484,7 +484,7 @@ void tcp_v4_send_check(struct sock *sk, int len, struct sk_buff *skb)
 	struct inet_sock *inet = inet_sk(sk);
 	struct tcphdr *th = skb->h.th;
 
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		th->check = ~tcp_v4_check(th, len, inet->saddr, inet->daddr, 0);
 		skb->csum = offsetof(struct tcphdr, check);
 	} else {
@@ -509,7 +509,7 @@ int tcp_v4_gso_send_check(struct sk_buff *skb)
 	th->check = 0;
 	th->check = ~tcp_v4_check(th, skb->len, iph->saddr, iph->daddr, 0);
 	skb->csum = offsetof(struct tcphdr, check);
-	skb->ip_summed = CHECKSUM_HW;
+	skb->ip_summed = CHECKSUM_PARTIAL;
 	return 0;
 }
 
@@ -973,7 +973,7 @@ static struct sock *tcp_v4_hnd_req(struct sock *sk, struct sk_buff *skb)
 
 static int tcp_v4_checksum_init(struct sk_buff *skb)
 {
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_COMPLETE) {
 		if (!tcp_v4_check(skb->h.th, skb->len, skb->nh.iph->saddr,
 				  skb->nh.iph->daddr, skb->csum)) {
 			skb->ip_summed = CHECKSUM_UNNECESSARY;
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index b4f3ffe1b3b4..9252a50c4b49 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -577,7 +577,7 @@ int tcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, unsigned int mss
 	TCP_SKB_CB(buff)->sacked = TCP_SKB_CB(skb)->sacked;
 	TCP_SKB_CB(skb)->sacked &= ~TCPCB_AT_TAIL;
 
-	if (!skb_shinfo(skb)->nr_frags && skb->ip_summed != CHECKSUM_HW) {
+	if (!skb_shinfo(skb)->nr_frags && skb->ip_summed != CHECKSUM_PARTIAL) {
 		/* Copy and checksum data tail into the new buffer. */
 		buff->csum = csum_partial_copy_nocheck(skb->data + len, skb_put(buff, nsize),
 						       nsize, 0);
@@ -586,7 +586,7 @@ int tcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, unsigned int mss
 
 		skb->csum = csum_block_sub(skb->csum, buff->csum, len);
 	} else {
-		skb->ip_summed = CHECKSUM_HW;
+		skb->ip_summed = CHECKSUM_PARTIAL;
 		skb_split(skb, buff, len);
 	}
 
@@ -689,7 +689,7 @@ int tcp_trim_head(struct sock *sk, struct sk_buff *skb, u32 len)
 		__pskb_trim_head(skb, len - skb_headlen(skb));
 
 	TCP_SKB_CB(skb)->seq += len;
-	skb->ip_summed = CHECKSUM_HW;
+	skb->ip_summed = CHECKSUM_PARTIAL;
 
 	skb->truesize	     -= len;
 	sk->sk_wmem_queued   -= len;
@@ -1062,7 +1062,7 @@ static int tso_fragment(struct sock *sk, struct sk_buff *skb, unsigned int len,
 	/* This packet was never sent out yet, so no SACK bits. */
 	TCP_SKB_CB(buff)->sacked = 0;
 
-	buff->ip_summed = skb->ip_summed = CHECKSUM_HW;
+	buff->ip_summed = skb->ip_summed = CHECKSUM_PARTIAL;
 	skb_split(skb, buff, len);
 
 	/* Fix up tso_factor for both original and new SKB.  */
@@ -1206,8 +1206,7 @@ static int tcp_mtu_probe(struct sock *sk)
 	TCP_SKB_CB(nskb)->flags = TCPCB_FLAG_ACK;
 	TCP_SKB_CB(nskb)->sacked = 0;
 	nskb->csum = 0;
-	if (skb->ip_summed == CHECKSUM_HW)
-		nskb->ip_summed = CHECKSUM_HW;
+	nskb->ip_summed = skb->ip_summed;
 
 	len = 0;
 	while (len < probe_size) {
@@ -1231,7 +1230,7 @@ static int tcp_mtu_probe(struct sock *sk)
 			                           ~(TCPCB_FLAG_FIN|TCPCB_FLAG_PSH);
 			if (!skb_shinfo(skb)->nr_frags) {
 				skb_pull(skb, copy);
-				if (skb->ip_summed != CHECKSUM_HW)
+				if (skb->ip_summed != CHECKSUM_PARTIAL)
 					skb->csum = csum_partial(skb->data, skb->len, 0);
 			} else {
 				__pskb_trim_head(skb, copy);
@@ -1572,10 +1571,9 @@ static void tcp_retrans_try_collapse(struct sock *sk, struct sk_buff *skb, int m
 
 		memcpy(skb_put(skb, next_skb_size), next_skb->data, next_skb_size);
 
-		if (next_skb->ip_summed == CHECKSUM_HW)
-			skb->ip_summed = CHECKSUM_HW;
+		skb->ip_summed = next_skb->ip_summed;
 
-		if (skb->ip_summed != CHECKSUM_HW)
+		if (skb->ip_summed != CHECKSUM_PARTIAL)
 			skb->csum = csum_block_add(skb->csum, next_skb->csum, skb_size);
 
 		/* Update sequence range on original skb. */
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index a4d005eccc7f..87152510980c 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -429,7 +429,7 @@ static int udp_push_pending_frames(struct sock *sk, struct udp_sock *up)
 		/*
 		 * Only one fragment on the socket.
 		 */
-		if (skb->ip_summed == CHECKSUM_HW) {
+		if (skb->ip_summed == CHECKSUM_PARTIAL) {
 			skb->csum = offsetof(struct udphdr, check);
 			uh->check = ~csum_tcpudp_magic(fl->fl4_src, fl->fl4_dst,
 					up->len, IPPROTO_UDP, 0);
@@ -448,7 +448,7 @@ static int udp_push_pending_frames(struct sock *sk, struct udp_sock *up)
 		 * fragments on the socket so that all csums of sk_buffs
 		 * should be together.
 		 */
-		if (skb->ip_summed == CHECKSUM_HW) {
+		if (skb->ip_summed == CHECKSUM_PARTIAL) {
 			int offset = (unsigned char *)uh - skb->data;
 			skb->csum = skb_checksum(skb, offset, skb->len - offset, 0);
 
@@ -1088,7 +1088,7 @@ static void udp_checksum_init(struct sk_buff *skb, struct udphdr *uh,
 {
 	if (uh->check == 0) {
 		skb->ip_summed = CHECKSUM_UNNECESSARY;
-	} else if (skb->ip_summed == CHECKSUM_HW) {
+	} else if (skb->ip_summed == CHECKSUM_COMPLETE) {
 		if (!udp_check(uh, ulen, saddr, daddr, skb->csum))
 			skb->ip_summed = CHECKSUM_UNNECESSARY;
 	}
diff --git a/net/ipv4/xfrm4_output.c b/net/ipv4/xfrm4_output.c
index d16f863cf687..4a96a9e3ef3b 100644
--- a/net/ipv4/xfrm4_output.c
+++ b/net/ipv4/xfrm4_output.c
@@ -48,8 +48,8 @@ static int xfrm4_output_one(struct sk_buff *skb)
 	struct xfrm_state *x = dst->xfrm;
 	int err;
 	
-	if (skb->ip_summed == CHECKSUM_HW) {
-		err = skb_checksum_help(skb, 0);
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
+		err = skb_checksum_help(skb);
 		if (err)
 			goto error_nolock;
 	}
diff --git a/net/ipv6/exthdrs.c b/net/ipv6/exthdrs.c
index 86dac106873b..05afa6b1912b 100644
--- a/net/ipv6/exthdrs.c
+++ b/net/ipv6/exthdrs.c
@@ -294,7 +294,7 @@ looped_back:
 		hdr = (struct ipv6_rt_hdr *) skb2->h.raw;
 	}
 
-	if (skb->ip_summed == CHECKSUM_HW)
+	if (skb->ip_summed == CHECKSUM_COMPLETE)
 		skb->ip_summed = CHECKSUM_NONE;
 
 	i = n - --hdr->segments_left;
diff --git a/net/ipv6/icmp.c b/net/ipv6/icmp.c
index dbfce089e916..103055107674 100644
--- a/net/ipv6/icmp.c
+++ b/net/ipv6/icmp.c
@@ -606,7 +606,7 @@ static int icmpv6_rcv(struct sk_buff **pskb)
 
 	/* Perform checksum. */
 	switch (skb->ip_summed) {
-	case CHECKSUM_HW:
+	case CHECKSUM_COMPLETE:
 		if (!csum_ipv6_magic(saddr, daddr, skb->len, IPPROTO_ICMPV6,
 				     skb->csum))
 			break;
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index 4fb47a252913..65514f21c186 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -866,7 +866,7 @@ static inline int ip6_ufo_append_data(struct sock *sk,
 		/* initialize protocol header pointer */
 		skb->h.raw = skb->data + fragheaderlen;
 
-		skb->ip_summed = CHECKSUM_HW;
+		skb->ip_summed = CHECKSUM_PARTIAL;
 		skb->csum = 0;
 		sk->sk_sndmsg_off = 0;
 	}
diff --git a/net/ipv6/netfilter.c b/net/ipv6/netfilter.c
index 395a417ba955..580b1aba6722 100644
--- a/net/ipv6/netfilter.c
+++ b/net/ipv6/netfilter.c
@@ -87,7 +87,7 @@ unsigned int nf_ip6_checksum(struct sk_buff *skb, unsigned int hook,
 	unsigned int csum = 0;
 
 	switch (skb->ip_summed) {
-	case CHECKSUM_HW:
+	case CHECKSUM_COMPLETE:
 		if (hook != NF_IP6_PRE_ROUTING && hook != NF_IP6_LOCAL_IN)
 			break;
 		if (!csum_ipv6_magic(&ip6h->saddr, &ip6h->daddr,
diff --git a/net/ipv6/netfilter/ip6_queue.c b/net/ipv6/netfilter/ip6_queue.c
index 968a14be0d05..c01c126224e2 100644
--- a/net/ipv6/netfilter/ip6_queue.c
+++ b/net/ipv6/netfilter/ip6_queue.c
@@ -206,9 +206,9 @@ ipq_build_packet_message(struct ipq_queue_entry *entry, int *errp)
 		break;
 	
 	case IPQ_COPY_PACKET:
-		if (entry->skb->ip_summed == CHECKSUM_HW &&
-		    (*errp = skb_checksum_help(entry->skb,
-		                               entry->info->outdev == NULL))) {
+		if ((entry->skb->ip_summed == CHECKSUM_PARTIAL ||
+		     entry->skb->ip_summed == CHECKSUM_COMPLETE) &&
+		    (*errp = skb_checksum_help(entry->skb))) {
 			read_unlock_bh(&queue_lock);
 			return NULL;
 		}
diff --git a/net/ipv6/netfilter/nf_conntrack_reasm.c b/net/ipv6/netfilter/nf_conntrack_reasm.c
index 00d5583807f7..7a4e4c2e3197 100644
--- a/net/ipv6/netfilter/nf_conntrack_reasm.c
+++ b/net/ipv6/netfilter/nf_conntrack_reasm.c
@@ -408,7 +408,7 @@ static int nf_ct_frag6_queue(struct nf_ct_frag6_queue *fq, struct sk_buff *skb,
  		return -1;
 	}
 
- 	if (skb->ip_summed == CHECKSUM_HW)
+ 	if (skb->ip_summed == CHECKSUM_COMPLETE)
  		skb->csum = csum_sub(skb->csum,
  				     csum_partial(skb->nh.raw,
 						  (u8*)(fhdr + 1) - skb->nh.raw,
@@ -640,7 +640,7 @@ nf_ct_frag6_reasm(struct nf_ct_frag6_queue *fq, struct net_device *dev)
 		head->len += fp->len;
 		if (head->ip_summed != fp->ip_summed)
 			head->ip_summed = CHECKSUM_NONE;
-		else if (head->ip_summed == CHECKSUM_HW)
+		else if (head->ip_summed == CHECKSUM_COMPLETE)
 			head->csum = csum_add(head->csum, fp->csum);
 		head->truesize += fp->truesize;
 		atomic_sub(fp->truesize, &nf_ct_frag6_mem);
@@ -652,7 +652,7 @@ nf_ct_frag6_reasm(struct nf_ct_frag6_queue *fq, struct net_device *dev)
 	head->nh.ipv6h->payload_len = htons(payload_len);
 
 	/* Yes, and fold redundant checksum back. 8) */
-	if (head->ip_summed == CHECKSUM_HW)
+	if (head->ip_summed == CHECKSUM_COMPLETE)
 		head->csum = csum_partial(head->nh.raw, head->h.raw-head->nh.raw, head->csum);
 
 	fq->fragments = NULL;
diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c
index d5040e172292..d4af1cb5e19f 100644
--- a/net/ipv6/raw.c
+++ b/net/ipv6/raw.c
@@ -334,7 +334,7 @@ int rawv6_rcv(struct sock *sk, struct sk_buff *skb)
 	if (!rp->checksum)
 		skb->ip_summed = CHECKSUM_UNNECESSARY;
 
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_COMPLETE) {
 		skb_postpull_rcsum(skb, skb->nh.raw,
 		                   skb->h.raw - skb->nh.raw);
 		if (!csum_ipv6_magic(&skb->nh.ipv6h->saddr,
diff --git a/net/ipv6/reassembly.c b/net/ipv6/reassembly.c
index 4e299c69e1c6..a8623d2b0879 100644
--- a/net/ipv6/reassembly.c
+++ b/net/ipv6/reassembly.c
@@ -433,7 +433,7 @@ static void ip6_frag_queue(struct frag_queue *fq, struct sk_buff *skb,
  		return;
 	}
 
- 	if (skb->ip_summed == CHECKSUM_HW)
+ 	if (skb->ip_summed == CHECKSUM_COMPLETE)
  		skb->csum = csum_sub(skb->csum,
  				     csum_partial(skb->nh.raw, (u8*)(fhdr+1)-skb->nh.raw, 0));
 
@@ -647,7 +647,7 @@ static int ip6_frag_reasm(struct frag_queue *fq, struct sk_buff **skb_in,
 		head->len += fp->len;
 		if (head->ip_summed != fp->ip_summed)
 			head->ip_summed = CHECKSUM_NONE;
-		else if (head->ip_summed == CHECKSUM_HW)
+		else if (head->ip_summed == CHECKSUM_COMPLETE)
 			head->csum = csum_add(head->csum, fp->csum);
 		head->truesize += fp->truesize;
 		atomic_sub(fp->truesize, &ip6_frag_mem);
@@ -662,7 +662,7 @@ static int ip6_frag_reasm(struct frag_queue *fq, struct sk_buff **skb_in,
 	*skb_in = head;
 
 	/* Yes, and fold redundant checksum back. 8) */
-	if (head->ip_summed == CHECKSUM_HW)
+	if (head->ip_summed == CHECKSUM_COMPLETE)
 		head->csum = csum_partial(head->nh.raw, head->h.raw-head->nh.raw, head->csum);
 
 	IP6_INC_STATS_BH(IPSTATS_MIB_REASMOKS);
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 302786a11cd6..7f1b660493b7 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -545,7 +545,7 @@ static void tcp_v6_send_check(struct sock *sk, int len, struct sk_buff *skb)
 	struct ipv6_pinfo *np = inet6_sk(sk);
 	struct tcphdr *th = skb->h.th;
 
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		th->check = ~csum_ipv6_magic(&np->saddr, &np->daddr, len, IPPROTO_TCP,  0);
 		skb->csum = offsetof(struct tcphdr, check);
 	} else {
@@ -570,7 +570,7 @@ static int tcp_v6_gso_send_check(struct sk_buff *skb)
 	th->check = ~csum_ipv6_magic(&ipv6h->saddr, &ipv6h->daddr, skb->len,
 				     IPPROTO_TCP, 0);
 	skb->csum = offsetof(struct tcphdr, check);
-	skb->ip_summed = CHECKSUM_HW;
+	skb->ip_summed = CHECKSUM_PARTIAL;
 	return 0;
 }
 
@@ -1033,7 +1033,7 @@ out:
 
 static int tcp_v6_checksum_init(struct sk_buff *skb)
 {
-	if (skb->ip_summed == CHECKSUM_HW) {
+	if (skb->ip_summed == CHECKSUM_COMPLETE) {
 		if (!tcp_v6_check(skb->h.th,skb->len,&skb->nh.ipv6h->saddr,
 				  &skb->nh.ipv6h->daddr,skb->csum)) {
 			skb->ip_summed = CHECKSUM_UNNECESSARY;
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index 82c7c9cde2a8..780b89f6dfcc 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -475,7 +475,7 @@ static int udpv6_rcv(struct sk_buff **pskb)
 		uh = skb->h.uh;
 	}
 
-	if (skb->ip_summed == CHECKSUM_HW &&
+	if (skb->ip_summed == CHECKSUM_COMPLETE &&
 	    !csum_ipv6_magic(saddr, daddr, ulen, IPPROTO_UDP, skb->csum))
 		skb->ip_summed = CHECKSUM_UNNECESSARY;
 
diff --git a/net/ipv6/xfrm6_output.c b/net/ipv6/xfrm6_output.c
index c8c8b44a0f58..6d111743e508 100644
--- a/net/ipv6/xfrm6_output.c
+++ b/net/ipv6/xfrm6_output.c
@@ -41,8 +41,8 @@ static int xfrm6_output_one(struct sk_buff *skb)
 	struct xfrm_state *x = dst->xfrm;
 	int err;
 	
-	if (skb->ip_summed == CHECKSUM_HW) {
-		err = skb_checksum_help(skb, 0);
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
+		err = skb_checksum_help(skb);
 		if (err)
 			goto error_nolock;
 	}
diff --git a/net/netfilter/nf_conntrack_proto_tcp.c b/net/netfilter/nf_conntrack_proto_tcp.c
index af8adcba23a7..308d2abd7ee5 100644
--- a/net/netfilter/nf_conntrack_proto_tcp.c
+++ b/net/netfilter/nf_conntrack_proto_tcp.c
@@ -823,8 +823,7 @@ static int tcp_error(struct sk_buff *skb,
   
 	/* Checksum invalid? Ignore.
 	 * We skip checking packets on the outgoing path
-	 * because the semantic of CHECKSUM_HW is different there 
-	 * and moreover root might send raw packets.
+	 * because the checksum is assumed to be correct.
 	 */
 	/* FIXME: Source route IP option packets --RR */
 	if (nf_conntrack_checksum &&
diff --git a/net/netfilter/nf_conntrack_proto_udp.c b/net/netfilter/nf_conntrack_proto_udp.c
index ae07ebe3ab37..d36e03139e8b 100644
--- a/net/netfilter/nf_conntrack_proto_udp.c
+++ b/net/netfilter/nf_conntrack_proto_udp.c
@@ -131,8 +131,7 @@ static int udp_error(struct sk_buff *skb, unsigned int dataoff,
 
 	/* Checksum invalid? Ignore.
 	 * We skip checking packets on the outgoing path
-	 * because the semantic of CHECKSUM_HW is different there
-	 * and moreover root might send raw packets.
+	 * because the checksum is assumed to be correct.
 	 * FIXME: Source route IP option packets --RR */
 	if (nf_conntrack_checksum &&
 	    ((pf == PF_INET && hooknum == NF_IP_PRE_ROUTING) ||
diff --git a/net/netfilter/nfnetlink_queue.c b/net/netfilter/nfnetlink_queue.c
index 49ef41e34c48..eddfbe4441a2 100644
--- a/net/netfilter/nfnetlink_queue.c
+++ b/net/netfilter/nfnetlink_queue.c
@@ -377,9 +377,9 @@ nfqnl_build_packet_message(struct nfqnl_instance *queue,
 		break;
 	
 	case NFQNL_COPY_PACKET:
-		if (entskb->ip_summed == CHECKSUM_HW &&
-		    (*errp = skb_checksum_help(entskb,
-		                               outdev == NULL))) {
+		if ((entskb->ip_summed == CHECKSUM_PARTIAL ||
+		     entskb->ip_summed == CHECKSUM_COMPLETE) &&
+		    (*errp = skb_checksum_help(entskb))) {
 			spin_unlock_bh(&queue->lock);
 			return NULL;
 		}
diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
index 4172a5235916..300215bdbf46 100644
--- a/net/packet/af_packet.c
+++ b/net/packet/af_packet.c
@@ -586,7 +586,7 @@ static int tpacket_rcv(struct sk_buff *skb, struct net_device *dev, struct packe
 		else if (skb->pkt_type == PACKET_OUTGOING) {
 			/* Special case: outgoing packets have ll header at head */
 			skb_pull(skb, skb->nh.raw - skb->data);
-			if (skb->ip_summed == CHECKSUM_HW)
+			if (skb->ip_summed == CHECKSUM_PARTIAL)
 				status |= TP_STATUS_CSUMNOTREADY;
 		}
 	}
diff --git a/net/sched/sch_netem.c b/net/sched/sch_netem.c
index a08ec4c7c55d..45939bafbdf8 100644
--- a/net/sched/sch_netem.c
+++ b/net/sched/sch_netem.c
@@ -192,8 +192,8 @@ static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch)
 	 */
 	if (q->corrupt && q->corrupt >= get_crandom(&q->corrupt_cor)) {
 		if (!(skb = skb_unshare(skb, GFP_ATOMIC))
-		    || (skb->ip_summed == CHECKSUM_HW
-			&& skb_checksum_help(skb, 0))) {
+		    || (skb->ip_summed == CHECKSUM_PARTIAL
+			&& skb_checksum_help(skb))) {
 			sch->qstats.drops++;
 			return NET_XMIT_DROP;
 		}
diff --git a/net/sunrpc/socklib.c b/net/sunrpc/socklib.c
index eb330d4f66d6..6f17527b9e69 100644
--- a/net/sunrpc/socklib.c
+++ b/net/sunrpc/socklib.c
@@ -168,7 +168,7 @@ int csum_partial_copy_to_xdr(struct xdr_buf *xdr, struct sk_buff *skb)
 		return -1;
 	if ((unsigned short)csum_fold(desc.csum))
 		return -1;
-	if (unlikely(skb->ip_summed == CHECKSUM_HW))
+	if (unlikely(skb->ip_summed == CHECKSUM_COMPLETE))
 		netdev_rx_csum_fault(skb->dev);
 	return 0;
 no_checksum:
-- 
cgit v1.2.3


From 4cf411de49c65140b3c259748629b561c0d3340f Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Sat, 5 Aug 2006 00:58:33 -0700
Subject: [NETFILTER]: Get rid of HW checksum invalidation

Update hardware checksums incrementally to avoid breaking GSO.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter.h                  |  6 +++
 include/linux/netfilter_ipv4/ip_nat.h      |  4 --
 include/linux/netfilter_ipv4/ip_nat_core.h |  8 ++--
 net/ipv4/netfilter/ip_nat_core.c           | 52 ++++++++++----------------
 net/ipv4/netfilter/ip_nat_helper.c         | 59 +++++++++++++++++++++---------
 net/ipv4/netfilter/ip_nat_proto_gre.c      |  5 ++-
 net/ipv4/netfilter/ip_nat_proto_icmp.c     |  8 ++--
 net/ipv4/netfilter/ip_nat_proto_tcp.c      |  7 ++--
 net/ipv4/netfilter/ip_nat_proto_udp.c      | 15 +++++---
 net/ipv4/netfilter/ip_nat_standalone.c     | 10 +----
 net/ipv4/netfilter/ipt_ECN.c               | 19 +++-------
 net/ipv4/netfilter/ipt_REJECT.c            |  1 +
 net/ipv4/netfilter/ipt_TCPMSS.c            | 39 ++++++++------------
 net/netfilter/core.c                       | 22 +++++++++++
 14 files changed, 138 insertions(+), 117 deletions(-)

(limited to 'include')

diff --git a/include/linux/netfilter.h b/include/linux/netfilter.h
index 10168e26a846..b7e67d1d4382 100644
--- a/include/linux/netfilter.h
+++ b/include/linux/netfilter.h
@@ -282,6 +282,12 @@ extern void nf_invalidate_cache(int pf);
    Returns true or false. */
 extern int skb_make_writable(struct sk_buff **pskb, unsigned int writable_len);
 
+extern u_int16_t nf_csum_update(u_int32_t oldval, u_int32_t newval,
+				u_int32_t csum);
+extern u_int16_t nf_proto_csum_update(struct sk_buff *skb,
+				      u_int32_t oldval, u_int32_t newval,
+				      u_int16_t csum, int pseudohdr);
+
 struct nf_afinfo {
 	unsigned short	family;
 	unsigned int	(*checksum)(struct sk_buff *skb, unsigned int hook,
diff --git a/include/linux/netfilter_ipv4/ip_nat.h b/include/linux/netfilter_ipv4/ip_nat.h
index e9f5ed1d9f68..98f8407e4cb5 100644
--- a/include/linux/netfilter_ipv4/ip_nat.h
+++ b/include/linux/netfilter_ipv4/ip_nat.h
@@ -72,10 +72,6 @@ extern unsigned int ip_nat_setup_info(struct ip_conntrack *conntrack,
 extern int ip_nat_used_tuple(const struct ip_conntrack_tuple *tuple,
 			     const struct ip_conntrack *ignored_conntrack);
 
-/* Calculate relative checksum. */
-extern u_int16_t ip_nat_cheat_check(u_int32_t oldvalinv,
-				    u_int32_t newval,
-				    u_int16_t oldcheck);
 #else  /* !__KERNEL__: iptables wants this to compile. */
 #define ip_nat_multi_range ip_nat_multi_range_compat
 #endif /*__KERNEL__*/
diff --git a/include/linux/netfilter_ipv4/ip_nat_core.h b/include/linux/netfilter_ipv4/ip_nat_core.h
index 30db23f06b03..60566f9fd7b3 100644
--- a/include/linux/netfilter_ipv4/ip_nat_core.h
+++ b/include/linux/netfilter_ipv4/ip_nat_core.h
@@ -11,8 +11,8 @@ extern unsigned int ip_nat_packet(struct ip_conntrack *ct,
 			       unsigned int hooknum,
 			       struct sk_buff **pskb);
 
-extern int ip_nat_icmp_reply_translation(struct sk_buff **pskb,
-					 struct ip_conntrack *ct,
-					 enum ip_nat_manip_type manip,
-					 enum ip_conntrack_dir dir);
+extern int ip_nat_icmp_reply_translation(struct ip_conntrack *ct,
+					 enum ip_conntrack_info ctinfo,
+					 unsigned int hooknum,
+					 struct sk_buff **pskb);
 #endif /* _IP_NAT_CORE_H */
diff --git a/net/ipv4/netfilter/ip_nat_core.c b/net/ipv4/netfilter/ip_nat_core.c
index 1741d555ad0d..4c540d03d48e 100644
--- a/net/ipv4/netfilter/ip_nat_core.c
+++ b/net/ipv4/netfilter/ip_nat_core.c
@@ -101,18 +101,6 @@ static void ip_nat_cleanup_conntrack(struct ip_conntrack *conn)
 	write_unlock_bh(&ip_nat_lock);
 }
 
-/* We do checksum mangling, so if they were wrong before they're still
- * wrong.  Also works for incomplete packets (eg. ICMP dest
- * unreachables.) */
-u_int16_t
-ip_nat_cheat_check(u_int32_t oldvalinv, u_int32_t newval, u_int16_t oldcheck)
-{
-	u_int32_t diffs[] = { oldvalinv, newval };
-	return csum_fold(csum_partial((char *)diffs, sizeof(diffs),
-				      oldcheck^0xFFFF));
-}
-EXPORT_SYMBOL(ip_nat_cheat_check);
-
 /* Is this tuple already taken? (not by us) */
 int
 ip_nat_used_tuple(const struct ip_conntrack_tuple *tuple,
@@ -378,12 +366,12 @@ manip_pkt(u_int16_t proto,
 	iph = (void *)(*pskb)->data + iphdroff;
 
 	if (maniptype == IP_NAT_MANIP_SRC) {
-		iph->check = ip_nat_cheat_check(~iph->saddr, target->src.ip,
-						iph->check);
+		iph->check = nf_csum_update(~iph->saddr, target->src.ip,
+					    iph->check);
 		iph->saddr = target->src.ip;
 	} else {
-		iph->check = ip_nat_cheat_check(~iph->daddr, target->dst.ip,
-						iph->check);
+		iph->check = nf_csum_update(~iph->daddr, target->dst.ip,
+					    iph->check);
 		iph->daddr = target->dst.ip;
 	}
 	return 1;
@@ -423,10 +411,10 @@ unsigned int ip_nat_packet(struct ip_conntrack *ct,
 EXPORT_SYMBOL_GPL(ip_nat_packet);
 
 /* Dir is direction ICMP is coming from (opposite to packet it contains) */
-int ip_nat_icmp_reply_translation(struct sk_buff **pskb,
-				  struct ip_conntrack *ct,
-				  enum ip_nat_manip_type manip,
-				  enum ip_conntrack_dir dir)
+int ip_nat_icmp_reply_translation(struct ip_conntrack *ct,
+				  enum ip_conntrack_info ctinfo,
+				  unsigned int hooknum,
+				  struct sk_buff **pskb)
 {
 	struct {
 		struct icmphdr icmp;
@@ -434,7 +422,9 @@ int ip_nat_icmp_reply_translation(struct sk_buff **pskb,
 	} *inside;
 	struct ip_conntrack_tuple inner, target;
 	int hdrlen = (*pskb)->nh.iph->ihl * 4;
+	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
 	unsigned long statusbit;
+	enum ip_nat_manip_type manip = HOOK2MANIP(hooknum);
 
 	if (!skb_make_writable(pskb, hdrlen + sizeof(*inside)))
 		return 0;
@@ -443,12 +433,8 @@ int ip_nat_icmp_reply_translation(struct sk_buff **pskb,
 
 	/* We're actually going to mangle it beyond trivial checksum
 	   adjustment, so make sure the current checksum is correct. */
-	if ((*pskb)->ip_summed != CHECKSUM_UNNECESSARY) {
-		hdrlen = (*pskb)->nh.iph->ihl * 4;
-		if ((u16)csum_fold(skb_checksum(*pskb, hdrlen,
-						(*pskb)->len - hdrlen, 0)))
-			return 0;
-	}
+	if (nf_ip_checksum(*pskb, hooknum, hdrlen, 0))
+		return 0;
 
 	/* Must be RELATED */
 	IP_NF_ASSERT((*pskb)->nfctinfo == IP_CT_RELATED ||
@@ -487,12 +473,14 @@ int ip_nat_icmp_reply_translation(struct sk_buff **pskb,
 		       !manip))
 		return 0;
 
-	/* Reloading "inside" here since manip_pkt inner. */
-	inside = (void *)(*pskb)->data + (*pskb)->nh.iph->ihl*4;
-	inside->icmp.checksum = 0;
-	inside->icmp.checksum = csum_fold(skb_checksum(*pskb, hdrlen,
-						       (*pskb)->len - hdrlen,
-						       0));
+	if ((*pskb)->ip_summed != CHECKSUM_PARTIAL) {
+		/* Reloading "inside" here since manip_pkt inner. */
+		inside = (void *)(*pskb)->data + (*pskb)->nh.iph->ihl*4;
+		inside->icmp.checksum = 0;
+		inside->icmp.checksum = csum_fold(skb_checksum(*pskb, hdrlen,
+							       (*pskb)->len - hdrlen,
+							       0));
+	}
 
 	/* Change outer to look the reply to an incoming packet
 	 * (proto 0 means don't invert per-proto part). */
diff --git a/net/ipv4/netfilter/ip_nat_helper.c b/net/ipv4/netfilter/ip_nat_helper.c
index cbcaa45370ae..021c3daae3ed 100644
--- a/net/ipv4/netfilter/ip_nat_helper.c
+++ b/net/ipv4/netfilter/ip_nat_helper.c
@@ -165,7 +165,7 @@ ip_nat_mangle_tcp_packet(struct sk_buff **pskb,
 {
 	struct iphdr *iph;
 	struct tcphdr *tcph;
-	int datalen;
+	int oldlen, datalen;
 
 	if (!skb_make_writable(pskb, (*pskb)->len))
 		return 0;
@@ -180,13 +180,22 @@ ip_nat_mangle_tcp_packet(struct sk_buff **pskb,
 	iph = (*pskb)->nh.iph;
 	tcph = (void *)iph + iph->ihl*4;
 
+	oldlen = (*pskb)->len - iph->ihl*4;
 	mangle_contents(*pskb, iph->ihl*4 + tcph->doff*4,
 			match_offset, match_len, rep_buffer, rep_len);
 
 	datalen = (*pskb)->len - iph->ihl*4;
-	tcph->check = 0;
-	tcph->check = tcp_v4_check(tcph, datalen, iph->saddr, iph->daddr,
-				   csum_partial((char *)tcph, datalen, 0));
+	if ((*pskb)->ip_summed != CHECKSUM_PARTIAL) {
+		tcph->check = 0;
+		tcph->check = tcp_v4_check(tcph, datalen,
+					   iph->saddr, iph->daddr,
+					   csum_partial((char *)tcph,
+					   		datalen, 0));
+	} else
+		tcph->check = nf_proto_csum_update(*pskb,
+						   htons(oldlen) ^ 0xFFFF,
+						   htons(datalen),
+						   tcph->check, 1);
 
 	if (rep_len != match_len) {
 		set_bit(IPS_SEQ_ADJUST_BIT, &ct->status);
@@ -221,6 +230,7 @@ ip_nat_mangle_udp_packet(struct sk_buff **pskb,
 {
 	struct iphdr *iph;
 	struct udphdr *udph;
+	int datalen, oldlen;
 
 	/* UDP helpers might accidentally mangle the wrong packet */
 	iph = (*pskb)->nh.iph;
@@ -238,22 +248,32 @@ ip_nat_mangle_udp_packet(struct sk_buff **pskb,
 
 	iph = (*pskb)->nh.iph;
 	udph = (void *)iph + iph->ihl*4;
+
+	oldlen = (*pskb)->len - iph->ihl*4;
 	mangle_contents(*pskb, iph->ihl*4 + sizeof(*udph),
 			match_offset, match_len, rep_buffer, rep_len);
 
 	/* update the length of the UDP packet */
-	udph->len = htons((*pskb)->len - iph->ihl*4);
+	datalen = (*pskb)->len - iph->ihl*4;
+	udph->len = htons(datalen);
 
 	/* fix udp checksum if udp checksum was previously calculated */
-	if (udph->check) {
-		int datalen = (*pskb)->len - iph->ihl * 4;
+	if (!udph->check && (*pskb)->ip_summed != CHECKSUM_PARTIAL)
+		return 1;
+
+	if ((*pskb)->ip_summed != CHECKSUM_PARTIAL) {
 		udph->check = 0;
 		udph->check = csum_tcpudp_magic(iph->saddr, iph->daddr,
 		                                datalen, IPPROTO_UDP,
 		                                csum_partial((char *)udph,
 		                                             datalen, 0));
-	}
-
+		if (!udph->check)
+			udph->check = -1;
+	} else
+		udph->check = nf_proto_csum_update(*pskb,
+						   htons(oldlen) ^ 0xFFFF,
+						   htons(datalen),
+						   udph->check, 1);
 	return 1;
 }
 EXPORT_SYMBOL(ip_nat_mangle_udp_packet);
@@ -293,11 +313,14 @@ sack_adjust(struct sk_buff *skb,
 			ntohl(sack->start_seq), new_start_seq,
 			ntohl(sack->end_seq), new_end_seq);
 
-		tcph->check = 
-			ip_nat_cheat_check(~sack->start_seq, new_start_seq,
-					   ip_nat_cheat_check(~sack->end_seq, 
-						   	      new_end_seq,
-							      tcph->check));
+		tcph->check = nf_proto_csum_update(skb,
+						   ~sack->start_seq,
+						   new_start_seq,
+						   tcph->check, 0);
+		tcph->check = nf_proto_csum_update(skb,
+						   ~sack->end_seq,
+						   new_end_seq,
+						   tcph->check, 0);
 		sack->start_seq = new_start_seq;
 		sack->end_seq = new_end_seq;
 		sackoff += sizeof(*sack);
@@ -381,10 +404,10 @@ ip_nat_seq_adjust(struct sk_buff **pskb,
 		newack = ntohl(tcph->ack_seq) - other_way->offset_before;
 	newack = htonl(newack);
 
-	tcph->check = ip_nat_cheat_check(~tcph->seq, newseq,
-					 ip_nat_cheat_check(~tcph->ack_seq, 
-					 		    newack, 
-							    tcph->check));
+	tcph->check = nf_proto_csum_update(*pskb, ~tcph->seq, newseq,
+					   tcph->check, 0);
+	tcph->check = nf_proto_csum_update(*pskb, ~tcph->ack_seq, newack,
+					   tcph->check, 0);
 
 	DEBUGP("Adjusting sequence number from %u->%u, ack from %u->%u\n",
 		ntohl(tcph->seq), ntohl(newseq), ntohl(tcph->ack_seq),
diff --git a/net/ipv4/netfilter/ip_nat_proto_gre.c b/net/ipv4/netfilter/ip_nat_proto_gre.c
index 38acfdf540eb..70a65372225a 100644
--- a/net/ipv4/netfilter/ip_nat_proto_gre.c
+++ b/net/ipv4/netfilter/ip_nat_proto_gre.c
@@ -130,9 +130,10 @@ gre_manip_pkt(struct sk_buff **pskb,
 			if (greh->csum) {
 				/* FIXME: Never tested this code... */
 				*(gre_csum(greh)) = 
-					ip_nat_cheat_check(~*(gre_key(greh)),
+					nf_proto_csum_update(*pskb,
+							~*(gre_key(greh)),
 							tuple->dst.u.gre.key,
-							*(gre_csum(greh)));
+							*(gre_csum(greh)), 0);
 			}
 			*(gre_key(greh)) = tuple->dst.u.gre.key;
 			break;
diff --git a/net/ipv4/netfilter/ip_nat_proto_icmp.c b/net/ipv4/netfilter/ip_nat_proto_icmp.c
index 31a3f4ccb99c..ec50cc295317 100644
--- a/net/ipv4/netfilter/ip_nat_proto_icmp.c
+++ b/net/ipv4/netfilter/ip_nat_proto_icmp.c
@@ -66,10 +66,10 @@ icmp_manip_pkt(struct sk_buff **pskb,
 		return 0;
 
 	hdr = (struct icmphdr *)((*pskb)->data + hdroff);
-
-	hdr->checksum = ip_nat_cheat_check(hdr->un.echo.id ^ 0xFFFF,
-					    tuple->src.u.icmp.id,
-					    hdr->checksum);
+	hdr->checksum = nf_proto_csum_update(*pskb,
+					     hdr->un.echo.id ^ 0xFFFF,
+					     tuple->src.u.icmp.id,
+					     hdr->checksum, 0);
 	hdr->un.echo.id = tuple->src.u.icmp.id;
 	return 1;
 }
diff --git a/net/ipv4/netfilter/ip_nat_proto_tcp.c b/net/ipv4/netfilter/ip_nat_proto_tcp.c
index a3d14079eba6..72a6307bd2db 100644
--- a/net/ipv4/netfilter/ip_nat_proto_tcp.c
+++ b/net/ipv4/netfilter/ip_nat_proto_tcp.c
@@ -129,10 +129,9 @@ tcp_manip_pkt(struct sk_buff **pskb,
 	if (hdrsize < sizeof(*hdr))
 		return 1;
 
-	hdr->check = ip_nat_cheat_check(~oldip, newip,
-					ip_nat_cheat_check(oldport ^ 0xFFFF,
-							   newport,
-							   hdr->check));
+	hdr->check = nf_proto_csum_update(*pskb, ~oldip, newip, hdr->check, 1);
+	hdr->check = nf_proto_csum_update(*pskb, oldport ^ 0xFFFF, newport,
+					  hdr->check, 0);
 	return 1;
 }
 
diff --git a/net/ipv4/netfilter/ip_nat_proto_udp.c b/net/ipv4/netfilter/ip_nat_proto_udp.c
index ec6053fdc867..5da196ae758c 100644
--- a/net/ipv4/netfilter/ip_nat_proto_udp.c
+++ b/net/ipv4/netfilter/ip_nat_proto_udp.c
@@ -113,11 +113,16 @@ udp_manip_pkt(struct sk_buff **pskb,
 		newport = tuple->dst.u.udp.port;
 		portptr = &hdr->dest;
 	}
-	if (hdr->check) /* 0 is a special case meaning no checksum */
-		hdr->check = ip_nat_cheat_check(~oldip, newip,
-					ip_nat_cheat_check(*portptr ^ 0xFFFF,
-							   newport,
-							   hdr->check));
+
+	if (hdr->check || (*pskb)->ip_summed == CHECKSUM_PARTIAL) {
+		hdr->check = nf_proto_csum_update(*pskb, ~oldip, newip,
+						  hdr->check, 1);
+		hdr->check = nf_proto_csum_update(*pskb,
+						  *portptr ^ 0xFFFF, newport,
+						  hdr->check, 0);
+		if (!hdr->check)
+			hdr->check = -1;
+	}
 	*portptr = newport;
 	return 1;
 }
diff --git a/net/ipv4/netfilter/ip_nat_standalone.c b/net/ipv4/netfilter/ip_nat_standalone.c
index f4f00c816d87..f3b778355432 100644
--- a/net/ipv4/netfilter/ip_nat_standalone.c
+++ b/net/ipv4/netfilter/ip_nat_standalone.c
@@ -110,12 +110,6 @@ ip_nat_fn(unsigned int hooknum,
 	IP_NF_ASSERT(!((*pskb)->nh.iph->frag_off
 		       & htons(IP_MF|IP_OFFSET)));
 
-	/* If we had a hardware checksum before, it's now invalid */
-	if ((*pskb)->ip_summed == CHECKSUM_PARTIAL ||
-	    (*pskb)->ip_summed == CHECKSUM_COMPLETE)
-		if (skb_checksum_help(*pskb))
-			return NF_DROP;
-
 	ct = ip_conntrack_get(*pskb, &ctinfo);
 	/* Can't track?  It's not due to stress, or conntrack would
 	   have dropped it.  Hence it's the user's responsibilty to
@@ -146,8 +140,8 @@ ip_nat_fn(unsigned int hooknum,
 	case IP_CT_RELATED:
 	case IP_CT_RELATED+IP_CT_IS_REPLY:
 		if ((*pskb)->nh.iph->protocol == IPPROTO_ICMP) {
-			if (!ip_nat_icmp_reply_translation(pskb, ct, maniptype,
-							   CTINFO2DIR(ctinfo)))
+			if (!ip_nat_icmp_reply_translation(ct, ctinfo,
+							   hooknum, pskb))
 				return NF_DROP;
 			else
 				return NF_ACCEPT;
diff --git a/net/ipv4/netfilter/ipt_ECN.c b/net/ipv4/netfilter/ipt_ECN.c
index 4ec43f98fe49..35916c74fe4e 100644
--- a/net/ipv4/netfilter/ipt_ECN.c
+++ b/net/ipv4/netfilter/ipt_ECN.c
@@ -52,7 +52,7 @@ static inline int
 set_ect_tcp(struct sk_buff **pskb, const struct ipt_ECN_info *einfo)
 {
 	struct tcphdr _tcph, *tcph;
-	u_int16_t diffs[2];
+	u_int16_t oldval;
 
 	/* Not enought header? */
 	tcph = skb_header_pointer(*pskb, (*pskb)->nh.iph->ihl*4,
@@ -70,23 +70,16 @@ set_ect_tcp(struct sk_buff **pskb, const struct ipt_ECN_info *einfo)
 		return 0;
 	tcph = (void *)(*pskb)->nh.iph + (*pskb)->nh.iph->ihl*4;
 
-	if (((*pskb)->ip_summed == CHECKSUM_PARTIAL ||
-	     (*pskb)->ip_summed == CHECKSUM_COMPLETE) &&
-	    skb_checksum_help(*pskb))
-		return 0;
-
-	diffs[0] = ((u_int16_t *)tcph)[6];
+	oldval = ((u_int16_t *)tcph)[6];
 	if (einfo->operation & IPT_ECN_OP_SET_ECE)
 		tcph->ece = einfo->proto.tcp.ece;
 	if (einfo->operation & IPT_ECN_OP_SET_CWR)
 		tcph->cwr = einfo->proto.tcp.cwr;
-	diffs[1] = ((u_int16_t *)tcph)[6];
-	diffs[0] = diffs[0] ^ 0xFFFF;
 
-	if ((*pskb)->ip_summed != CHECKSUM_UNNECESSARY)
-		tcph->check = csum_fold(csum_partial((char *)diffs,
-						     sizeof(diffs),
-						     tcph->check^0xFFFF));
+	tcph->check = nf_proto_csum_update((*pskb),
+					   oldval ^ 0xFFFF,
+					   ((u_int16_t *)tcph)[6],
+					   tcph->check, 0);
 	return 1;
 }
 
diff --git a/net/ipv4/netfilter/ipt_REJECT.c b/net/ipv4/netfilter/ipt_REJECT.c
index 7f905bf2bde5..95c6662b663c 100644
--- a/net/ipv4/netfilter/ipt_REJECT.c
+++ b/net/ipv4/netfilter/ipt_REJECT.c
@@ -185,6 +185,7 @@ static void send_reset(struct sk_buff *oldskb, int hook)
 	tcph->urg_ptr = 0;
 
 	/* Adjust TCP checksum */
+	nskb->ip_summed = CHECKSUM_NONE;
 	tcph->check = 0;
 	tcph->check = tcp_v4_check(tcph, sizeof(struct tcphdr),
 				   nskb->nh.iph->saddr,
diff --git a/net/ipv4/netfilter/ipt_TCPMSS.c b/net/ipv4/netfilter/ipt_TCPMSS.c
index c998dc0fcd15..0fce85e05507 100644
--- a/net/ipv4/netfilter/ipt_TCPMSS.c
+++ b/net/ipv4/netfilter/ipt_TCPMSS.c
@@ -27,14 +27,6 @@ MODULE_DESCRIPTION("iptables TCP MSS modification module");
 #define DEBUGP(format, args...)
 #endif
 
-static u_int16_t
-cheat_check(u_int32_t oldvalinv, u_int32_t newval, u_int16_t oldcheck)
-{
-	u_int32_t diffs[] = { oldvalinv, newval };
-	return csum_fold(csum_partial((char *)diffs, sizeof(diffs),
-                                      oldcheck^0xFFFF));
-}
-
 static inline unsigned int
 optlen(const u_int8_t *opt, unsigned int offset)
 {
@@ -62,11 +54,6 @@ ipt_tcpmss_target(struct sk_buff **pskb,
 	if (!skb_make_writable(pskb, (*pskb)->len))
 		return NF_DROP;
 
-	if (((*pskb)->ip_summed == CHECKSUM_PARTIAL ||
-	     (*pskb)->ip_summed == CHECKSUM_COMPLETE) &&
-	    skb_checksum_help(*pskb))
-		return NF_DROP;
-
 	iph = (*pskb)->nh.iph;
 	tcplen = (*pskb)->len - iph->ihl*4;
 
@@ -120,9 +107,10 @@ ipt_tcpmss_target(struct sk_buff **pskb,
 			opt[i+2] = (newmss & 0xff00) >> 8;
 			opt[i+3] = (newmss & 0x00ff);
 
-			tcph->check = cheat_check(htons(oldmss)^0xFFFF,
-						  htons(newmss),
-						  tcph->check);
+			tcph->check = nf_proto_csum_update(*pskb,
+							   htons(oldmss)^0xFFFF,
+							   htons(newmss),
+							   tcph->check, 0);
 
 			DEBUGP(KERN_INFO "ipt_tcpmss_target: %u.%u.%u.%u:%hu"
 			       "->%u.%u.%u.%u:%hu changed TCP MSS option"
@@ -162,8 +150,10 @@ ipt_tcpmss_target(struct sk_buff **pskb,
  	opt = (u_int8_t *)tcph + sizeof(struct tcphdr);
 	memmove(opt + TCPOLEN_MSS, opt, tcplen - sizeof(struct tcphdr));
 
-	tcph->check = cheat_check(htons(tcplen) ^ 0xFFFF,
-				  htons(tcplen + TCPOLEN_MSS), tcph->check);
+	tcph->check = nf_proto_csum_update(*pskb,
+					   htons(tcplen) ^ 0xFFFF,
+				           htons(tcplen + TCPOLEN_MSS),
+					   tcph->check, 1);
 	tcplen += TCPOLEN_MSS;
 
 	opt[0] = TCPOPT_MSS;
@@ -171,16 +161,19 @@ ipt_tcpmss_target(struct sk_buff **pskb,
 	opt[2] = (newmss & 0xff00) >> 8;
 	opt[3] = (newmss & 0x00ff);
 
-	tcph->check = cheat_check(~0, *((u_int32_t *)opt), tcph->check);
+	tcph->check = nf_proto_csum_update(*pskb, ~0, *((u_int32_t *)opt),
+					   tcph->check, 0);
 
 	oldval = ((u_int16_t *)tcph)[6];
 	tcph->doff += TCPOLEN_MSS/4;
-	tcph->check = cheat_check(oldval ^ 0xFFFF,
-				  ((u_int16_t *)tcph)[6], tcph->check);
+	tcph->check = nf_proto_csum_update(*pskb,
+					   oldval ^ 0xFFFF,
+					   ((u_int16_t *)tcph)[6],
+					   tcph->check, 0);
 
 	newtotlen = htons(ntohs(iph->tot_len) + TCPOLEN_MSS);
-	iph->check = cheat_check(iph->tot_len ^ 0xFFFF,
-				 newtotlen, iph->check);
+	iph->check = nf_csum_update(iph->tot_len ^ 0xFFFF,
+				    newtotlen, iph->check);
 	iph->tot_len = newtotlen;
 
 	DEBUGP(KERN_INFO "ipt_tcpmss_target: %u.%u.%u.%u:%hu"
diff --git a/net/netfilter/core.c b/net/netfilter/core.c
index 5d29d5e23624..27f639f3ac2a 100644
--- a/net/netfilter/core.c
+++ b/net/netfilter/core.c
@@ -222,6 +222,28 @@ copy_skb:
 }
 EXPORT_SYMBOL(skb_make_writable);
 
+u_int16_t nf_csum_update(u_int32_t oldval, u_int32_t newval, u_int32_t csum)
+{
+	u_int32_t diff[] = { oldval, newval };
+
+	return csum_fold(csum_partial((char *)diff, sizeof(diff), ~csum));
+}
+EXPORT_SYMBOL(nf_csum_update);
+
+u_int16_t nf_proto_csum_update(struct sk_buff *skb,
+			       u_int32_t oldval, u_int32_t newval,
+			       u_int16_t csum, int pseudohdr)
+{
+	if (skb->ip_summed != CHECKSUM_PARTIAL) {
+		csum = nf_csum_update(oldval, newval, csum);
+		if (skb->ip_summed == CHECKSUM_COMPLETE && pseudohdr)
+			skb->csum = nf_csum_update(oldval, newval, skb->csum);
+	} else if (pseudohdr)
+		csum = ~nf_csum_update(oldval, newval, ~csum);
+
+	return csum;
+}
+EXPORT_SYMBOL(nf_proto_csum_update);
 
 /* This does not belong here, but locally generated errors need it if connection
    tracking in use: without this, connection may not be in hash table, and hence
-- 
cgit v1.2.3


From 9067c722cf6930adf1df2d169de9094dd90b0c33 Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Mon, 7 Aug 2006 17:57:44 -0700
Subject: [NEIGH]: Move netlink neighbour bits to linux/neighbour.h

Moves netlink neighbour bits to linux/neighbour.h. Also
moves bits to be exported to userspace from net/neighbour.h
to linux/neighbour.h and removes __KERNEL__ guards, userspace
is not supposed to be using it.

rtnetlink_rcv_msg() is not longer required to parse attributes
for the neighbour layer, remove dependency on obsolete and
buggy rta_buf.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/neighbour.h | 65 +++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/rtnetlink.h | 63 ---------------------------------------------
 include/net/neighbour.h   | 39 ++--------------------------
 net/core/rtnetlink.c      |  2 --
 4 files changed, 67 insertions(+), 102 deletions(-)
 create mode 100644 include/linux/neighbour.h

(limited to 'include')

diff --git a/include/linux/neighbour.h b/include/linux/neighbour.h
new file mode 100644
index 000000000000..8e8293d86fbb
--- /dev/null
+++ b/include/linux/neighbour.h
@@ -0,0 +1,65 @@
+#ifndef __LINUX_NEIGHBOUR_H
+#define __LINUX_NEIGHBOUR_H
+
+#include <linux/netlink.h>
+
+struct ndmsg
+{
+	__u8		ndm_family;
+	__u8		ndm_pad1;
+	__u16		ndm_pad2;
+	__s32		ndm_ifindex;
+	__u16		ndm_state;
+	__u8		ndm_flags;
+	__u8		ndm_type;
+};
+
+enum
+{
+	NDA_UNSPEC,
+	NDA_DST,
+	NDA_LLADDR,
+	NDA_CACHEINFO,
+	NDA_PROBES,
+	__NDA_MAX
+};
+
+#define NDA_MAX (__NDA_MAX - 1)
+
+/*
+ *	Neighbor Cache Entry Flags
+ */
+
+#define NTF_PROXY	0x08	/* == ATF_PUBL */
+#define NTF_ROUTER	0x80
+
+/*
+ *	Neighbor Cache Entry States.
+ */
+
+#define NUD_INCOMPLETE	0x01
+#define NUD_REACHABLE	0x02
+#define NUD_STALE	0x04
+#define NUD_DELAY	0x08
+#define NUD_PROBE	0x10
+#define NUD_FAILED	0x20
+
+/* Dummy states */
+#define NUD_NOARP	0x40
+#define NUD_PERMANENT	0x80
+#define NUD_NONE	0x00
+
+/* NUD_NOARP & NUD_PERMANENT are pseudostates, they never change
+   and make no address resolution or NUD.
+   NUD_PERMANENT is also cannot be deleted by garbage collectors.
+ */
+
+struct nda_cacheinfo
+{
+	__u32		ndm_confirmed;
+	__u32		ndm_used;
+	__u32		ndm_updated;
+	__u32		ndm_refcnt;
+};
+
+#endif
diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index 84f3eb426da2..9750f0214c22 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -386,69 +386,6 @@ struct rta_session
 
 
 
-/**************************************************************
- *		Neighbour discovery.
- ****/
-
-struct ndmsg
-{
-	unsigned char	ndm_family;
-	unsigned char	ndm_pad1;
-	unsigned short	ndm_pad2;
-	int		ndm_ifindex;	/* Link index			*/
-	__u16		ndm_state;
-	__u8		ndm_flags;
-	__u8		ndm_type;
-};
-
-enum
-{
-	NDA_UNSPEC,
-	NDA_DST,
-	NDA_LLADDR,
-	NDA_CACHEINFO,
-	NDA_PROBES,
-	__NDA_MAX
-};
-
-#define NDA_MAX (__NDA_MAX - 1)
-
-#define NDA_RTA(r)  ((struct rtattr*)(((char*)(r)) + NLMSG_ALIGN(sizeof(struct ndmsg))))
-#define NDA_PAYLOAD(n) NLMSG_PAYLOAD(n,sizeof(struct ndmsg))
-
-/*
- *	Neighbor Cache Entry Flags
- */
-
-#define NTF_PROXY	0x08	/* == ATF_PUBL */
-#define NTF_ROUTER	0x80
-
-/*
- *	Neighbor Cache Entry States.
- */
-
-#define NUD_INCOMPLETE	0x01
-#define NUD_REACHABLE	0x02
-#define NUD_STALE	0x04
-#define NUD_DELAY	0x08
-#define NUD_PROBE	0x10
-#define NUD_FAILED	0x20
-
-/* Dummy states */
-#define NUD_NOARP	0x40
-#define NUD_PERMANENT	0x80
-#define NUD_NONE	0x00
-
-
-struct nda_cacheinfo
-{
-	__u32		ndm_confirmed;
-	__u32		ndm_used;
-	__u32		ndm_updated;
-	__u32		ndm_refcnt;
-};
-
-
 /*****************************************************************
  *		Neighbour tables specific messages.
  *
diff --git a/include/net/neighbour.h b/include/net/neighbour.h
index 4901ee446879..74c4b6ff8a5c 100644
--- a/include/net/neighbour.h
+++ b/include/net/neighbour.h
@@ -1,6 +1,8 @@
 #ifndef _NET_NEIGHBOUR_H
 #define _NET_NEIGHBOUR_H
 
+#include <linux/neighbour.h>
+
 /*
  *	Generic neighbour manipulation
  *
@@ -14,40 +16,6 @@
  *		- Add neighbour cache statistics like rtstat
  */
 
-/* The following flags & states are exported to user space,
-   so that they should be moved to include/linux/ directory.
- */
-
-/*
- *	Neighbor Cache Entry Flags
- */
-
-#define NTF_PROXY	0x08	/* == ATF_PUBL */
-#define NTF_ROUTER	0x80
-
-/*
- *	Neighbor Cache Entry States.
- */
-
-#define NUD_INCOMPLETE	0x01
-#define NUD_REACHABLE	0x02
-#define NUD_STALE	0x04
-#define NUD_DELAY	0x08
-#define NUD_PROBE	0x10
-#define NUD_FAILED	0x20
-
-/* Dummy states */
-#define NUD_NOARP	0x40
-#define NUD_PERMANENT	0x80
-#define NUD_NONE	0x00
-
-/* NUD_NOARP & NUD_PERMANENT are pseudostates, they never change
-   and make no address resolution or NUD.
-   NUD_PERMANENT is also cannot be deleted by garbage collectors.
- */
-
-#ifdef __KERNEL__
-
 #include <asm/atomic.h>
 #include <linux/netdevice.h>
 #include <linux/skbuff.h>
@@ -374,6 +342,3 @@ struct neighbour_cb {
 #define NEIGH_CB(skb)	((struct neighbour_cb *)(skb)->cb)
 
 #endif
-#endif
-
-
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index 93ba04fb8444..78ccbd4c4e37 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -104,7 +104,6 @@ static const int rtm_min[RTM_NR_FAMILIES] =
 	[RTM_FAM(RTM_NEWLINK)]      = NLMSG_LENGTH(sizeof(struct ifinfomsg)),
 	[RTM_FAM(RTM_NEWADDR)]      = NLMSG_LENGTH(sizeof(struct ifaddrmsg)),
 	[RTM_FAM(RTM_NEWROUTE)]     = NLMSG_LENGTH(sizeof(struct rtmsg)),
-	[RTM_FAM(RTM_NEWNEIGH)]     = NLMSG_LENGTH(sizeof(struct ndmsg)),
 	[RTM_FAM(RTM_NEWRULE)]      = NLMSG_LENGTH(sizeof(struct fib_rule_hdr)),
 	[RTM_FAM(RTM_NEWQDISC)]     = NLMSG_LENGTH(sizeof(struct tcmsg)),
 	[RTM_FAM(RTM_NEWTCLASS)]    = NLMSG_LENGTH(sizeof(struct tcmsg)),
@@ -121,7 +120,6 @@ static const int rta_max[RTM_NR_FAMILIES] =
 	[RTM_FAM(RTM_NEWLINK)]      = IFLA_MAX,
 	[RTM_FAM(RTM_NEWADDR)]      = IFA_MAX,
 	[RTM_FAM(RTM_NEWROUTE)]     = RTA_MAX,
-	[RTM_FAM(RTM_NEWNEIGH)]     = NDA_MAX,
 	[RTM_FAM(RTM_NEWRULE)]      = FRA_MAX,
 	[RTM_FAM(RTM_NEWQDISC)]     = TCA_MAX,
 	[RTM_FAM(RTM_NEWTCLASS)]    = TCA_MAX,
-- 
cgit v1.2.3


From b63bbc5006a0a62fabc81c4f77e95f16ff16f340 Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Mon, 7 Aug 2006 18:00:57 -0700
Subject: [NEIGH]: Move netlink neighbour table bits to linux/neighbour.h

rtnetlink_rcv_msg() is not longer required to parse attributes
for the neighbour tables layer, remove dependency on obsolete and
buggy rta_buf.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/neighbour.h |  94 ++++++++++++++++++++++++++++++++++++++++++
 include/linux/rtnetlink.h | 101 ----------------------------------------------
 net/core/rtnetlink.c      |   2 -
 3 files changed, 94 insertions(+), 103 deletions(-)

(limited to 'include')

diff --git a/include/linux/neighbour.h b/include/linux/neighbour.h
index 8e8293d86fbb..bd3bbf668cdb 100644
--- a/include/linux/neighbour.h
+++ b/include/linux/neighbour.h
@@ -62,4 +62,98 @@ struct nda_cacheinfo
 	__u32		ndm_refcnt;
 };
 
+/*****************************************************************
+ *		Neighbour tables specific messages.
+ *
+ * To retrieve the neighbour tables send RTM_GETNEIGHTBL with the
+ * NLM_F_DUMP flag set. Every neighbour table configuration is
+ * spread over multiple messages to avoid running into message
+ * size limits on systems with many interfaces. The first message
+ * in the sequence transports all not device specific data such as
+ * statistics, configuration, and the default parameter set.
+ * This message is followed by 0..n messages carrying device
+ * specific parameter sets.
+ * Although the ordering should be sufficient, NDTA_NAME can be
+ * used to identify sequences. The initial message can be identified
+ * by checking for NDTA_CONFIG. The device specific messages do
+ * not contain this TLV but have NDTPA_IFINDEX set to the
+ * corresponding interface index.
+ *
+ * To change neighbour table attributes, send RTM_SETNEIGHTBL
+ * with NDTA_NAME set. Changeable attribute include NDTA_THRESH[1-3],
+ * NDTA_GC_INTERVAL, and all TLVs in NDTA_PARMS unless marked
+ * otherwise. Device specific parameter sets can be changed by
+ * setting NDTPA_IFINDEX to the interface index of the corresponding
+ * device.
+ ****/
+
+struct ndt_stats
+{
+	__u64		ndts_allocs;
+	__u64		ndts_destroys;
+	__u64		ndts_hash_grows;
+	__u64		ndts_res_failed;
+	__u64		ndts_lookups;
+	__u64		ndts_hits;
+	__u64		ndts_rcv_probes_mcast;
+	__u64		ndts_rcv_probes_ucast;
+	__u64		ndts_periodic_gc_runs;
+	__u64		ndts_forced_gc_runs;
+};
+
+enum {
+	NDTPA_UNSPEC,
+	NDTPA_IFINDEX,			/* u32, unchangeable */
+	NDTPA_REFCNT,			/* u32, read-only */
+	NDTPA_REACHABLE_TIME,		/* u64, read-only, msecs */
+	NDTPA_BASE_REACHABLE_TIME,	/* u64, msecs */
+	NDTPA_RETRANS_TIME,		/* u64, msecs */
+	NDTPA_GC_STALETIME,		/* u64, msecs */
+	NDTPA_DELAY_PROBE_TIME,		/* u64, msecs */
+	NDTPA_QUEUE_LEN,		/* u32 */
+	NDTPA_APP_PROBES,		/* u32 */
+	NDTPA_UCAST_PROBES,		/* u32 */
+	NDTPA_MCAST_PROBES,		/* u32 */
+	NDTPA_ANYCAST_DELAY,		/* u64, msecs */
+	NDTPA_PROXY_DELAY,		/* u64, msecs */
+	NDTPA_PROXY_QLEN,		/* u32 */
+	NDTPA_LOCKTIME,			/* u64, msecs */
+	__NDTPA_MAX
+};
+#define NDTPA_MAX (__NDTPA_MAX - 1)
+
+struct ndtmsg
+{
+	__u8		ndtm_family;
+	__u8		ndtm_pad1;
+	__u16		ndtm_pad2;
+};
+
+struct ndt_config
+{
+	__u16		ndtc_key_len;
+	__u16		ndtc_entry_size;
+	__u32		ndtc_entries;
+	__u32		ndtc_last_flush;	/* delta to now in msecs */
+	__u32		ndtc_last_rand;		/* delta to now in msecs */
+	__u32		ndtc_hash_rnd;
+	__u32		ndtc_hash_mask;
+	__u32		ndtc_hash_chain_gc;
+	__u32		ndtc_proxy_qlen;
+};
+
+enum {
+	NDTA_UNSPEC,
+	NDTA_NAME,			/* char *, unchangeable */
+	NDTA_THRESH1,			/* u32 */
+	NDTA_THRESH2,			/* u32 */
+	NDTA_THRESH3,			/* u32 */
+	NDTA_CONFIG,			/* struct ndt_config, read-only */
+	NDTA_PARMS,			/* nested TLV NDTPA_* */
+	NDTA_STATS,			/* struct ndt_stats, read-only */
+	NDTA_GC_INTERVAL,		/* u64, msecs */
+	__NDTA_MAX
+};
+#define NDTA_MAX (__NDTA_MAX - 1)
+
 #endif
diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index 9750f0214c22..784a1a29490e 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -384,107 +384,6 @@ struct rta_session
 	} u;
 };
 
-
-
-/*****************************************************************
- *		Neighbour tables specific messages.
- *
- * To retrieve the neighbour tables send RTM_GETNEIGHTBL with the
- * NLM_F_DUMP flag set. Every neighbour table configuration is
- * spread over multiple messages to avoid running into message
- * size limits on systems with many interfaces. The first message
- * in the sequence transports all not device specific data such as
- * statistics, configuration, and the default parameter set.
- * This message is followed by 0..n messages carrying device
- * specific parameter sets.
- * Although the ordering should be sufficient, NDTA_NAME can be
- * used to identify sequences. The initial message can be identified
- * by checking for NDTA_CONFIG. The device specific messages do
- * not contain this TLV but have NDTPA_IFINDEX set to the
- * corresponding interface index.
- *
- * To change neighbour table attributes, send RTM_SETNEIGHTBL
- * with NDTA_NAME set. Changeable attribute include NDTA_THRESH[1-3],
- * NDTA_GC_INTERVAL, and all TLVs in NDTA_PARMS unless marked
- * otherwise. Device specific parameter sets can be changed by
- * setting NDTPA_IFINDEX to the interface index of the corresponding
- * device.
- ****/
-
-struct ndt_stats
-{
-	__u64		ndts_allocs;
-	__u64		ndts_destroys;
-	__u64		ndts_hash_grows;
-	__u64		ndts_res_failed;
-	__u64		ndts_lookups;
-	__u64		ndts_hits;
-	__u64		ndts_rcv_probes_mcast;
-	__u64		ndts_rcv_probes_ucast;
-	__u64		ndts_periodic_gc_runs;
-	__u64		ndts_forced_gc_runs;
-};
-
-enum {
-	NDTPA_UNSPEC,
-	NDTPA_IFINDEX,			/* u32, unchangeable */
-	NDTPA_REFCNT,			/* u32, read-only */
-	NDTPA_REACHABLE_TIME,		/* u64, read-only, msecs */
-	NDTPA_BASE_REACHABLE_TIME,	/* u64, msecs */
-	NDTPA_RETRANS_TIME,		/* u64, msecs */
-	NDTPA_GC_STALETIME,		/* u64, msecs */
-	NDTPA_DELAY_PROBE_TIME,		/* u64, msecs */
-	NDTPA_QUEUE_LEN,		/* u32 */
-	NDTPA_APP_PROBES,		/* u32 */
-	NDTPA_UCAST_PROBES,		/* u32 */
-	NDTPA_MCAST_PROBES,		/* u32 */
-	NDTPA_ANYCAST_DELAY,		/* u64, msecs */
-	NDTPA_PROXY_DELAY,		/* u64, msecs */
-	NDTPA_PROXY_QLEN,		/* u32 */
-	NDTPA_LOCKTIME,			/* u64, msecs */
-	__NDTPA_MAX
-};
-#define NDTPA_MAX (__NDTPA_MAX - 1)
-
-struct ndtmsg
-{
-	__u8		ndtm_family;
-	__u8		ndtm_pad1;
-	__u16		ndtm_pad2;
-};
-
-struct ndt_config
-{
-	__u16		ndtc_key_len;
-	__u16		ndtc_entry_size;
-	__u32		ndtc_entries;
-	__u32		ndtc_last_flush;	/* delta to now in msecs */
-	__u32		ndtc_last_rand;		/* delta to now in msecs */
-	__u32		ndtc_hash_rnd;
-	__u32		ndtc_hash_mask;
-	__u32		ndtc_hash_chain_gc;
-	__u32		ndtc_proxy_qlen;
-};
-
-enum {
-	NDTA_UNSPEC,
-	NDTA_NAME,			/* char *, unchangeable */
-	NDTA_THRESH1,			/* u32 */
-	NDTA_THRESH2,			/* u32 */
-	NDTA_THRESH3,			/* u32 */
-	NDTA_CONFIG,			/* struct ndt_config, read-only */
-	NDTA_PARMS,			/* nested TLV NDTPA_* */
-	NDTA_STATS,			/* struct ndt_stats, read-only */
-	NDTA_GC_INTERVAL,		/* u64, msecs */
-	__NDTA_MAX
-};
-#define NDTA_MAX (__NDTA_MAX - 1)
-
-#define NDTA_RTA(r) ((struct rtattr*)(((char*)(r)) + \
-		     NLMSG_ALIGN(sizeof(struct ndtmsg))))
-#define NDTA_PAYLOAD(n) NLMSG_PAYLOAD(n,sizeof(struct ndtmsg))
-
-
 /****
  *		General form of address family dependent message.
  ****/
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index 78ccbd4c4e37..a1b783a6afc6 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -112,7 +112,6 @@ static const int rtm_min[RTM_NR_FAMILIES] =
 	[RTM_FAM(RTM_NEWPREFIX)]    = NLMSG_LENGTH(sizeof(struct rtgenmsg)),
 	[RTM_FAM(RTM_GETMULTICAST)] = NLMSG_LENGTH(sizeof(struct rtgenmsg)),
 	[RTM_FAM(RTM_GETANYCAST)]   = NLMSG_LENGTH(sizeof(struct rtgenmsg)),
-	[RTM_FAM(RTM_NEWNEIGHTBL)]  = NLMSG_LENGTH(sizeof(struct ndtmsg)),
 };
 
 static const int rta_max[RTM_NR_FAMILIES] =
@@ -125,7 +124,6 @@ static const int rta_max[RTM_NR_FAMILIES] =
 	[RTM_FAM(RTM_NEWTCLASS)]    = TCA_MAX,
 	[RTM_FAM(RTM_NEWTFILTER)]   = TCA_MAX,
 	[RTM_FAM(RTM_NEWACTION)]    = TCAA_MAX,
-	[RTM_FAM(RTM_NEWNEIGHTBL)]  = NDTA_MAX,
 };
 
 void __rta_fill(struct sk_buff *skb, int attrtype, int attrlen, const void *data)
-- 
cgit v1.2.3


From ac5a488ef252ed673cb067843e411f8cc43f7ab9 Mon Sep 17 00:00:00 2001
From: Sridhar Samudrala <sri@us.ibm.com>
Date: Mon, 7 Aug 2006 20:57:31 -0700
Subject: [NET]: Round out in-kernel sockets API

This patch implements wrapper functions that provide a convenient way
to access the sockets API for in-kernel users like sunrpc, cifs &
ocfs2 etc and any future users.

Signed-off-by: Sridhar Samudrala <sri@us.ibm.com>
Acked-by: James Morris <jmorris@namei.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/net.h |  19 +++++++++
 net/socket.c        | 113 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 132 insertions(+)

(limited to 'include')

diff --git a/include/linux/net.h b/include/linux/net.h
index b20c53c74413..19da2c08d7b6 100644
--- a/include/linux/net.h
+++ b/include/linux/net.h
@@ -208,6 +208,25 @@ extern int   	     kernel_recvmsg(struct socket *sock, struct msghdr *msg,
 				    struct kvec *vec, size_t num,
 				    size_t len, int flags);
 
+extern int kernel_bind(struct socket *sock, struct sockaddr *addr,
+		       int addrlen);
+extern int kernel_listen(struct socket *sock, int backlog);
+extern int kernel_accept(struct socket *sock, struct socket **newsock,
+			 int flags);
+extern int kernel_connect(struct socket *sock, struct sockaddr *addr,
+			  int addrlen, int flags);
+extern int kernel_getsockname(struct socket *sock, struct sockaddr *addr,
+			      int *addrlen);
+extern int kernel_getpeername(struct socket *sock, struct sockaddr *addr,
+			      int *addrlen);
+extern int kernel_getsockopt(struct socket *sock, int level, int optname,
+			     char *optval, int *optlen);
+extern int kernel_setsockopt(struct socket *sock, int level, int optname,
+			     char *optval, int optlen);
+extern int kernel_sendpage(struct socket *sock, struct page *page, int offset,
+			   size_t size, int flags);
+extern int kernel_sock_ioctl(struct socket *sock, int cmd, unsigned long arg);
+
 #ifndef CONFIG_SMP
 #define SOCKOPS_WRAPPED(name) name
 #define SOCKOPS_WRAP(name, fam)
diff --git a/net/socket.c b/net/socket.c
index 6756e57e1ff0..2eaebf934a1a 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -2170,6 +2170,109 @@ static long compat_sock_ioctl(struct file *file, unsigned cmd,
 }
 #endif
 
+int kernel_bind(struct socket *sock, struct sockaddr *addr, int addrlen)
+{
+	return sock->ops->bind(sock, addr, addrlen);
+}
+
+int kernel_listen(struct socket *sock, int backlog)
+{
+	return sock->ops->listen(sock, backlog);
+}
+
+int kernel_accept(struct socket *sock, struct socket **newsock, int flags)
+{
+	struct sock *sk = sock->sk;
+	int err;
+
+	err = sock_create_lite(sk->sk_family, sk->sk_type, sk->sk_protocol,
+			       newsock);
+	if (err < 0)
+		goto done;
+
+	err = sock->ops->accept(sock, *newsock, flags);
+	if (err < 0) {
+		sock_release(*newsock);
+		goto done;
+	}
+
+	(*newsock)->ops = sock->ops;
+
+done:
+	return err;
+}
+
+int kernel_connect(struct socket *sock, struct sockaddr *addr, int addrlen,
+                   int flags)
+{
+	return sock->ops->connect(sock, addr, addrlen, flags);
+}
+
+int kernel_getsockname(struct socket *sock, struct sockaddr *addr,
+			 int *addrlen)
+{
+	return sock->ops->getname(sock, addr, addrlen, 0);
+}
+
+int kernel_getpeername(struct socket *sock, struct sockaddr *addr,
+			 int *addrlen)
+{
+	return sock->ops->getname(sock, addr, addrlen, 1);
+}
+
+int kernel_getsockopt(struct socket *sock, int level, int optname,
+			char *optval, int *optlen)
+{
+	mm_segment_t oldfs = get_fs();
+	int err;
+
+	set_fs(KERNEL_DS);
+	if (level == SOL_SOCKET)
+		err = sock_getsockopt(sock, level, optname, optval, optlen);
+	else
+		err = sock->ops->getsockopt(sock, level, optname, optval,
+					    optlen);
+	set_fs(oldfs);
+	return err;
+}
+
+int kernel_setsockopt(struct socket *sock, int level, int optname,
+			char *optval, int optlen)
+{
+	mm_segment_t oldfs = get_fs();
+	int err;
+
+	set_fs(KERNEL_DS);
+	if (level == SOL_SOCKET)
+		err = sock_setsockopt(sock, level, optname, optval, optlen);
+	else
+		err = sock->ops->setsockopt(sock, level, optname, optval,
+					    optlen);
+	set_fs(oldfs);
+	return err;
+}
+
+int kernel_sendpage(struct socket *sock, struct page *page, int offset,
+		    size_t size, int flags)
+{
+	if (sock->ops->sendpage)
+		return sock->ops->sendpage(sock, page, offset, size, flags);
+
+	return sock_no_sendpage(sock, page, offset, size, flags);
+}
+
+int kernel_sock_ioctl(struct socket *sock, int cmd, unsigned long arg)
+{
+	mm_segment_t oldfs = get_fs();
+	int err;
+
+	set_fs(KERNEL_DS);
+	err = sock->ops->ioctl(sock, cmd, arg);
+	set_fs(oldfs);
+
+	return err;
+}
+
 /* ABI emulation layers need these two */
 EXPORT_SYMBOL(move_addr_to_kernel);
 EXPORT_SYMBOL(move_addr_to_user);
@@ -2186,3 +2289,13 @@ EXPORT_SYMBOL(sock_wake_async);
 EXPORT_SYMBOL(sockfd_lookup);
 EXPORT_SYMBOL(kernel_sendmsg);
 EXPORT_SYMBOL(kernel_recvmsg);
+EXPORT_SYMBOL(kernel_bind);
+EXPORT_SYMBOL(kernel_listen);
+EXPORT_SYMBOL(kernel_accept);
+EXPORT_SYMBOL(kernel_connect);
+EXPORT_SYMBOL(kernel_getsockname);
+EXPORT_SYMBOL(kernel_getpeername);
+EXPORT_SYMBOL(kernel_getsockopt);
+EXPORT_SYMBOL(kernel_setsockopt);
+EXPORT_SYMBOL(kernel_sendpage);
+EXPORT_SYMBOL(kernel_sock_ioctl);
-- 
cgit v1.2.3


From 8ce11e6a9faf1f1c849b77104adc1642c46aee95 Mon Sep 17 00:00:00 2001
From: Adrian Bunk <bunk@stusta.de>
Date: Mon, 7 Aug 2006 21:50:48 -0700
Subject: [NET]: Make code static.

This patch makes needlessly global code static.

Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip6_fib.h              | 4 ----
 net/ipv4/cipso_ipv4.c              | 2 +-
 net/ipv4/fib_rules.c               | 4 ++--
 net/ipv6/fib6_rules.c              | 4 ++--
 net/ipv6/ip6_fib.c                 | 6 +++---
 net/ipv6/route.c                   | 6 +++---
 net/netlabel/netlabel_domainhash.c | 4 ++--
 7 files changed, 13 insertions(+), 17 deletions(-)

(limited to 'include')

diff --git a/include/net/ip6_fib.h b/include/net/ip6_fib.h
index 7b47e8d5a765..c0660cea9a2f 100644
--- a/include/net/ip6_fib.h
+++ b/include/net/ip6_fib.h
@@ -192,10 +192,6 @@ struct fib6_node		*fib6_locate(struct fib6_node *root,
 					     struct in6_addr *daddr, int dst_len,
 					     struct in6_addr *saddr, int src_len);
 
-extern void			fib6_clean_tree(struct fib6_node *root,
-						int (*func)(struct rt6_info *, void *arg),
-						int prune, void *arg);
-
 extern void			fib6_clean_all(int (*func)(struct rt6_info *, void *arg),
 					       int prune, void *arg);
 
diff --git a/net/ipv4/cipso_ipv4.c b/net/ipv4/cipso_ipv4.c
index b82a101c95c5..80a2a0911b49 100644
--- a/net/ipv4/cipso_ipv4.c
+++ b/net/ipv4/cipso_ipv4.c
@@ -60,7 +60,7 @@ struct cipso_v4_domhsh_entry {
  * if in practice there are a lot of different DOIs this list should
  * probably be turned into a hash table or something similar so we
  * can do quick lookups. */
-DEFINE_SPINLOCK(cipso_v4_doi_list_lock);
+static DEFINE_SPINLOCK(cipso_v4_doi_list_lock);
 static struct list_head cipso_v4_doi_list = LIST_HEAD_INIT(cipso_v4_doi_list);
 
 /* Label mapping cache */
diff --git a/net/ipv4/fib_rules.c b/net/ipv4/fib_rules.c
index 23ec6ae1a0f6..03d1e8a43a48 100644
--- a/net/ipv4/fib_rules.c
+++ b/net/ipv4/fib_rules.c
@@ -101,8 +101,8 @@ int fib_lookup(struct flowi *flp, struct fib_result *res)
 	return err;
 }
 
-int fib4_rule_action(struct fib_rule *rule, struct flowi *flp, int flags,
-		     struct fib_lookup_arg *arg)
+static int fib4_rule_action(struct fib_rule *rule, struct flowi *flp,
+			    int flags, struct fib_lookup_arg *arg)
 {
 	int err = -EAGAIN;
 	struct fib_table *tbl;
diff --git a/net/ipv6/fib6_rules.c b/net/ipv6/fib6_rules.c
index 94a46ec967a4..bf9bba83b852 100644
--- a/net/ipv6/fib6_rules.c
+++ b/net/ipv6/fib6_rules.c
@@ -66,8 +66,8 @@ struct dst_entry *fib6_rule_lookup(struct flowi *fl, int flags,
 	return (struct dst_entry *) arg.result;
 }
 
-int fib6_rule_action(struct fib_rule *rule, struct flowi *flp,
-		     int flags, struct fib_lookup_arg *arg)
+static int fib6_rule_action(struct fib_rule *rule, struct flowi *flp,
+			    int flags, struct fib_lookup_arg *arg)
 {
 	struct rt6_info *rt = NULL;
 	struct fib6_table *table;
diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
index ce226c14bef5..1f2316187ca4 100644
--- a/net/ipv6/ip6_fib.c
+++ b/net/ipv6/ip6_fib.c
@@ -1169,9 +1169,9 @@ static int fib6_clean_node(struct fib6_walker_t *w)
  *	ignoring pure split nodes) will be scanned.
  */
 
-void fib6_clean_tree(struct fib6_node *root,
-		     int (*func)(struct rt6_info *, void *arg),
-		     int prune, void *arg)
+static void fib6_clean_tree(struct fib6_node *root,
+			    int (*func)(struct rt6_info *, void *arg),
+			    int prune, void *arg)
 {
 	struct fib6_cleaner_t c;
 
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 41c5905d3191..e08d84063c1f 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -613,8 +613,8 @@ static struct rt6_info *rt6_alloc_clone(struct rt6_info *ort, struct in6_addr *d
 	return rt;
 }
 
-struct rt6_info *ip6_pol_route_input(struct fib6_table *table, struct flowi *fl,
-				     int flags)
+static struct rt6_info *ip6_pol_route_input(struct fib6_table *table,
+					    struct flowi *fl, int flags)
 {
 	struct fib6_node *fn;
 	struct rt6_info *rt, *nrt;
@@ -872,7 +872,7 @@ static inline unsigned int ipv6_advmss(unsigned int mtu)
 }
 
 static struct dst_entry *ndisc_dst_gc_list;
-DEFINE_SPINLOCK(ndisc_lock);
+static DEFINE_SPINLOCK(ndisc_lock);
 
 struct dst_entry *ndisc_dst_alloc(struct net_device *dev, 
 				  struct neighbour *neigh,
diff --git a/net/netlabel/netlabel_domainhash.c b/net/netlabel/netlabel_domainhash.c
index 5bb3fad4a115..0489a1378101 100644
--- a/net/netlabel/netlabel_domainhash.c
+++ b/net/netlabel/netlabel_domainhash.c
@@ -50,11 +50,11 @@ struct netlbl_domhsh_tbl {
 /* Domain hash table */
 /* XXX - updates should be so rare that having one spinlock for the entire
  * hash table should be okay */
-DEFINE_SPINLOCK(netlbl_domhsh_lock);
+static DEFINE_SPINLOCK(netlbl_domhsh_lock);
 static struct netlbl_domhsh_tbl *netlbl_domhsh = NULL;
 
 /* Default domain mapping */
-DEFINE_SPINLOCK(netlbl_domhsh_def_lock);
+static DEFINE_SPINLOCK(netlbl_domhsh_def_lock);
 static struct netlbl_dom_map *netlbl_domhsh_def = NULL;
 
 /*
-- 
cgit v1.2.3


From 0298f36a579b5bd7f10f6f6d57e5929977a865a1 Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@sunset.davemloft.net>
Date: Mon, 7 Aug 2006 21:56:52 -0700
Subject: [IPV4]: Kill fib4_rules_clean().

As noted by Adrian Bunk this function is totally unused.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip_fib.h | 1 -
 net/ipv4/fib_rules.c | 5 -----
 2 files changed, 6 deletions(-)

(limited to 'include')

diff --git a/include/net/ip_fib.h b/include/net/ip_fib.h
index 14c82e611c95..adf73586bc05 100644
--- a/include/net/ip_fib.h
+++ b/include/net/ip_fib.h
@@ -254,7 +254,6 @@ extern struct fib_table *fib_hash_init(int id);
 extern int fib4_rules_dump(struct sk_buff *skb, struct netlink_callback *cb);
 
 extern void __init fib4_rules_init(void);
-extern void __exit fib4_rules_cleanup(void);
 
 #ifdef CONFIG_NET_CLS_ROUTE
 extern u32 fib_rules_tclass(struct fib_result *res);
diff --git a/net/ipv4/fib_rules.c b/net/ipv4/fib_rules.c
index 03d1e8a43a48..d242e5291fcc 100644
--- a/net/ipv4/fib_rules.c
+++ b/net/ipv4/fib_rules.c
@@ -347,8 +347,3 @@ void __init fib4_rules_init(void)
 
 	fib_rules_register(&fib4_rules_ops);
 }
-
-void __exit fib4_rules_cleanup(void)
-{
-	fib_rules_unregister(&fib4_rules_ops);
-}
-- 
cgit v1.2.3


From 1a01912ae0a5666c4c24eaae2b4821711e2ad79a Mon Sep 17 00:00:00 2001
From: Louis Nyffenegger <louis.nyffenegger@gmail.com>
Date: Tue, 8 Aug 2006 00:56:11 -0700
Subject: [INET]: Remove is_setbyuser patch

The value is_setbyuser from struct ip_options is never used and set
only one time (http://linux-net.osdl.org/index.php/TODO#IPV4).
This little patch removes it from the kernel source.

Signed-off-by: Louis Nyffenegger <louis.nyffenegger@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/inet_sock.h | 4 +---
 net/ipv4/ip_options.c   | 1 -
 2 files changed, 1 insertion(+), 4 deletions(-)

(limited to 'include')

diff --git a/include/net/inet_sock.h b/include/net/inet_sock.h
index f4caad56cd03..f6242710f2ff 100644
--- a/include/net/inet_sock.h
+++ b/include/net/inet_sock.h
@@ -27,7 +27,6 @@
 /** struct ip_options - IP Options
  *
  * @faddr - Saved first hop address
- * @is_setbyuser - Set by setsockopt?
  * @is_data - Options in __data, rather than skb
  * @is_strictroute - Strict source route
  * @srr_is_hit - Packet destination addr was our one
@@ -42,8 +41,7 @@ struct ip_options {
 	unsigned char	srr;
 	unsigned char	rr;
 	unsigned char	ts;
-	unsigned char	is_setbyuser:1,
-			is_data:1,
+	unsigned char	is_data:1,
 			is_strictroute:1,
 			srr_is_hit:1,
 			is_changed:1,
diff --git a/net/ipv4/ip_options.c b/net/ipv4/ip_options.c
index e0a93b4fa8cc..e7437c091326 100644
--- a/net/ipv4/ip_options.c
+++ b/net/ipv4/ip_options.c
@@ -525,7 +525,6 @@ static int ip_options_get_finish(struct ip_options **optp,
 		opt->__data[optlen++] = IPOPT_END;
 	opt->optlen = optlen;
 	opt->is_data = 1;
-	opt->is_setbyuser = 1;
 	if (optlen && ip_options_compile(opt, NULL)) {
 		kfree(opt);
 		return -EINVAL;
-- 
cgit v1.2.3


From 99a92ff50424146ba01a222248fd47a1cd55b78f Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Tue, 8 Aug 2006 02:18:10 -0700
Subject: [IPV4]: Uninline inet_lookup_listener

By modern standards this function is way too big to be inlined.  It's
even bigger than __inet_lookup_listener :)

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/inet_hashtables.h | 33 ++-------------------------------
 net/ipv4/inet_hashtables.c    | 35 ++++++++++++++++++++++++++++++++---
 2 files changed, 34 insertions(+), 34 deletions(-)

(limited to 'include')

diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h
index 98e0bb3014fe..bd513f3b9c7e 100644
--- a/include/net/inet_hashtables.h
+++ b/include/net/inet_hashtables.h
@@ -271,39 +271,10 @@ static inline int inet_iif(const struct sk_buff *skb)
 	return ((struct rtable *)skb->dst)->rt_iif;
 }
 
-extern struct sock *__inet_lookup_listener(const struct hlist_head *head,
-					   const u32 daddr,
-					   const unsigned short hnum,
-					   const int dif);
-
-/* Optimize the common listener case. */
-static inline struct sock *
+extern struct sock *
 		inet_lookup_listener(struct inet_hashinfo *hashinfo,
 				     const u32 daddr,
-				     const unsigned short hnum, const int dif)
-{
-	struct sock *sk = NULL;
-	const struct hlist_head *head;
-
-	read_lock(&hashinfo->lhash_lock);
-	head = &hashinfo->listening_hash[inet_lhashfn(hnum)];
-	if (!hlist_empty(head)) {
-		const struct inet_sock *inet = inet_sk((sk = __sk_head(head)));
-
-		if (inet->num == hnum && !sk->sk_node.next &&
-		    (!inet->rcv_saddr || inet->rcv_saddr == daddr) &&
-		    (sk->sk_family == PF_INET || !ipv6_only_sock(sk)) &&
-		    !sk->sk_bound_dev_if)
-			goto sherry_cache;
-		sk = __inet_lookup_listener(head, daddr, hnum, dif);
-	}
-	if (sk) {
-sherry_cache:
-		sock_hold(sk);
-	}
-	read_unlock(&hashinfo->lhash_lock);
-	return sk;
-}
+				     const unsigned short hnum, const int dif);
 
 /* Socket demux engine toys. */
 #ifdef __BIG_ENDIAN
diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c
index 95fac5532994..bfc39066e730 100644
--- a/net/ipv4/inet_hashtables.c
+++ b/net/ipv4/inet_hashtables.c
@@ -124,8 +124,10 @@ EXPORT_SYMBOL(inet_listen_wlock);
  * remote address for the connection. So always assume those are both
  * wildcarded during the search since they can never be otherwise.
  */
-struct sock *__inet_lookup_listener(const struct hlist_head *head, const u32 daddr,
-				    const unsigned short hnum, const int dif)
+static struct sock *__inet_lookup_listener(const struct hlist_head *head,
+					   const u32 daddr,
+					   const unsigned short hnum,
+					   const int dif)
 {
 	struct sock *result = NULL, *sk;
 	const struct hlist_node *node;
@@ -159,7 +161,34 @@ struct sock *__inet_lookup_listener(const struct hlist_head *head, const u32 dad
 	return result;
 }
 
-EXPORT_SYMBOL_GPL(__inet_lookup_listener);
+/* Optimize the common listener case. */
+struct sock *inet_lookup_listener(struct inet_hashinfo *hashinfo,
+				  const u32 daddr, const unsigned short hnum,
+				  const int dif)
+{
+	struct sock *sk = NULL;
+	const struct hlist_head *head;
+
+	read_lock(&hashinfo->lhash_lock);
+	head = &hashinfo->listening_hash[inet_lhashfn(hnum)];
+	if (!hlist_empty(head)) {
+		const struct inet_sock *inet = inet_sk((sk = __sk_head(head)));
+
+		if (inet->num == hnum && !sk->sk_node.next &&
+		    (!inet->rcv_saddr || inet->rcv_saddr == daddr) &&
+		    (sk->sk_family == PF_INET || !ipv6_only_sock(sk)) &&
+		    !sk->sk_bound_dev_if)
+			goto sherry_cache;
+		sk = __inet_lookup_listener(head, daddr, hnum, dif);
+	}
+	if (sk) {
+sherry_cache:
+		sock_hold(sk);
+	}
+	read_unlock(&hashinfo->lhash_lock);
+	return sk;
+}
+EXPORT_SYMBOL_GPL(inet_lookup_listener);
 
 /* called with local bh disabled */
 static int __inet_check_established(struct inet_timewait_death_row *death_row,
-- 
cgit v1.2.3


From 8f491069b40be5d627007a343f99759e9da6a178 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Wed, 9 Aug 2006 15:47:12 -0700
Subject: [IPV4]: Use network-order dport for all visible inet_lookup_*

Right now most inet_lookup_* functions take a host-order hnum instead
of a network-order dport because that's how it is represented
internally.

This means that users of these functions have to be careful about
using the right byte-order.  To add more confusion, inet_lookup takes
a network-order dport unlike all other functions.

So this patch changes all visible inet_lookup functions to take a
dport and move all dport->hnum conversion inside them.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/inet_hashtables.h | 31 ++++++++++++++++++++++++-------
 net/dccp/ipv4.c               | 10 +++++-----
 net/ipv4/inet_hashtables.c    | 18 +++++++++---------
 net/ipv4/tcp_ipv4.c           | 10 +++++-----
 4 files changed, 43 insertions(+), 26 deletions(-)

(limited to 'include')

diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h
index bd513f3b9c7e..b4491c9e2a5a 100644
--- a/include/net/inet_hashtables.h
+++ b/include/net/inet_hashtables.h
@@ -271,10 +271,16 @@ static inline int inet_iif(const struct sk_buff *skb)
 	return ((struct rtable *)skb->dst)->rt_iif;
 }
 
-extern struct sock *
-		inet_lookup_listener(struct inet_hashinfo *hashinfo,
-				     const u32 daddr,
-				     const unsigned short hnum, const int dif);
+extern struct sock *__inet_lookup_listener(struct inet_hashinfo *hashinfo,
+					   const u32 daddr,
+					   const unsigned short hnum,
+					   const int dif);
+
+static inline struct sock *inet_lookup_listener(struct inet_hashinfo *hashinfo,
+						u32 daddr, u16 dport, int dif)
+{
+	return __inet_lookup_listener(hashinfo, daddr, ntohs(dport), dif);
+}
 
 /* Socket demux engine toys. */
 #ifdef __BIG_ENDIAN
@@ -362,14 +368,25 @@ hit:
 	goto out;
 }
 
+static inline struct sock *
+	inet_lookup_established(struct inet_hashinfo *hashinfo,
+				const u32 saddr, const u16 sport,
+				const u32 daddr, const u16 dport,
+				const int dif)
+{
+	return __inet_lookup_established(hashinfo, saddr, sport, daddr,
+					 ntohs(dport), dif);
+}
+
 static inline struct sock *__inet_lookup(struct inet_hashinfo *hashinfo,
 					 const u32 saddr, const u16 sport,
-					 const u32 daddr, const u16 hnum,
+					 const u32 daddr, const u16 dport,
 					 const int dif)
 {
+	u16 hnum = ntohs(dport);
 	struct sock *sk = __inet_lookup_established(hashinfo, saddr, sport, daddr,
 						    hnum, dif);
-	return sk ? : inet_lookup_listener(hashinfo, daddr, hnum, dif);
+	return sk ? : __inet_lookup_listener(hashinfo, daddr, hnum, dif);
 }
 
 static inline struct sock *inet_lookup(struct inet_hashinfo *hashinfo,
@@ -380,7 +397,7 @@ static inline struct sock *inet_lookup(struct inet_hashinfo *hashinfo,
 	struct sock *sk;
 
 	local_bh_disable();
-	sk = __inet_lookup(hashinfo, saddr, sport, daddr, ntohs(dport), dif);
+	sk = __inet_lookup(hashinfo, saddr, sport, daddr, dport, dif);
 	local_bh_enable();
 
 	return sk;
diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c
index 171d363876ee..9a1a76a7dc41 100644
--- a/net/dccp/ipv4.c
+++ b/net/dccp/ipv4.c
@@ -608,10 +608,10 @@ static struct sock *dccp_v4_hnd_req(struct sock *sk, struct sk_buff *skb)
 	if (req != NULL)
 		return dccp_check_req(sk, skb, req, prev);
 
-	nsk = __inet_lookup_established(&dccp_hashinfo,
-					iph->saddr, dh->dccph_sport,
-					iph->daddr, ntohs(dh->dccph_dport),
-					inet_iif(skb));
+	nsk = inet_lookup_established(&dccp_hashinfo,
+				      iph->saddr, dh->dccph_sport,
+				      iph->daddr, dh->dccph_dport,
+				      inet_iif(skb));
 	if (nsk != NULL) {
 		if (nsk->sk_state != DCCP_TIME_WAIT) {
 			bh_lock_sock(nsk);
@@ -925,7 +925,7 @@ static int dccp_v4_rcv(struct sk_buff *skb)
 	 * 	Look up flow ID in table and get corresponding socket */
 	sk = __inet_lookup(&dccp_hashinfo,
 			   skb->nh.iph->saddr, dh->dccph_sport,
-			   skb->nh.iph->daddr, ntohs(dh->dccph_dport),
+			   skb->nh.iph->daddr, dh->dccph_dport,
 			   inet_iif(skb));
 
 	/* 
diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c
index bfc39066e730..fb296c9a7f3f 100644
--- a/net/ipv4/inet_hashtables.c
+++ b/net/ipv4/inet_hashtables.c
@@ -124,10 +124,10 @@ EXPORT_SYMBOL(inet_listen_wlock);
  * remote address for the connection. So always assume those are both
  * wildcarded during the search since they can never be otherwise.
  */
-static struct sock *__inet_lookup_listener(const struct hlist_head *head,
-					   const u32 daddr,
-					   const unsigned short hnum,
-					   const int dif)
+static struct sock *inet_lookup_listener_slow(const struct hlist_head *head,
+					      const u32 daddr,
+					      const unsigned short hnum,
+					      const int dif)
 {
 	struct sock *result = NULL, *sk;
 	const struct hlist_node *node;
@@ -162,9 +162,9 @@ static struct sock *__inet_lookup_listener(const struct hlist_head *head,
 }
 
 /* Optimize the common listener case. */
-struct sock *inet_lookup_listener(struct inet_hashinfo *hashinfo,
-				  const u32 daddr, const unsigned short hnum,
-				  const int dif)
+struct sock *__inet_lookup_listener(struct inet_hashinfo *hashinfo,
+				    const u32 daddr, const unsigned short hnum,
+				    const int dif)
 {
 	struct sock *sk = NULL;
 	const struct hlist_head *head;
@@ -179,7 +179,7 @@ struct sock *inet_lookup_listener(struct inet_hashinfo *hashinfo,
 		    (sk->sk_family == PF_INET || !ipv6_only_sock(sk)) &&
 		    !sk->sk_bound_dev_if)
 			goto sherry_cache;
-		sk = __inet_lookup_listener(head, daddr, hnum, dif);
+		sk = inet_lookup_listener_slow(head, daddr, hnum, dif);
 	}
 	if (sk) {
 sherry_cache:
@@ -188,7 +188,7 @@ sherry_cache:
 	read_unlock(&hashinfo->lhash_lock);
 	return sk;
 }
-EXPORT_SYMBOL_GPL(inet_lookup_listener);
+EXPORT_SYMBOL_GPL(__inet_lookup_listener);
 
 /* called with local bh disabled */
 static int __inet_check_established(struct inet_timewait_death_row *death_row,
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index b2aa512a30e9..2973dee0a489 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -951,9 +951,9 @@ static struct sock *tcp_v4_hnd_req(struct sock *sk, struct sk_buff *skb)
 	if (req)
 		return tcp_check_req(sk, skb, req, prev);
 
-	nsk = __inet_lookup_established(&tcp_hashinfo, skb->nh.iph->saddr,
-					th->source, skb->nh.iph->daddr,
-					ntohs(th->dest), inet_iif(skb));
+	nsk = inet_lookup_established(&tcp_hashinfo, skb->nh.iph->saddr,
+				      th->source, skb->nh.iph->daddr,
+				      th->dest, inet_iif(skb));
 
 	if (nsk) {
 		if (nsk->sk_state != TCP_TIME_WAIT) {
@@ -1090,7 +1090,7 @@ int tcp_v4_rcv(struct sk_buff *skb)
 	TCP_SKB_CB(skb)->sacked	 = 0;
 
 	sk = __inet_lookup(&tcp_hashinfo, skb->nh.iph->saddr, th->source,
-			   skb->nh.iph->daddr, ntohs(th->dest),
+			   skb->nh.iph->daddr, th->dest,
 			   inet_iif(skb));
 
 	if (!sk)
@@ -1168,7 +1168,7 @@ do_time_wait:
 	case TCP_TW_SYN: {
 		struct sock *sk2 = inet_lookup_listener(&tcp_hashinfo,
 							skb->nh.iph->daddr,
-							ntohs(th->dest),
+							th->dest,
 							inet_iif(skb));
 		if (sk2) {
 			inet_twsk_deschedule((struct inet_timewait_sock *)sk,
-- 
cgit v1.2.3


From a8731cbf61c8768ea129780b70dc7dfc6795aad4 Mon Sep 17 00:00:00 2001
From: Steven Whitehouse <steve@chygwyn.com>
Date: Wed, 9 Aug 2006 15:56:46 -0700
Subject: [DECNET]: Covert rules to use generic code

This patch converts the DECnet rules code to use the generic
rules system created by Thomas Graf <tgraf@suug.ch>.

Signed-off-by: Steven Whitehouse <steve@chygwyn.com>
Acked-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/rtnetlink.h |   3 +-
 include/net/dn_fib.h      |   8 +-
 net/decnet/Kconfig        |   1 +
 net/decnet/af_decnet.c    |   1 +
 net/decnet/dn_dev.c       |   3 +-
 net/decnet/dn_fib.c       |   1 +
 net/decnet/dn_route.c     |   3 +-
 net/decnet/dn_rules.c     | 494 +++++++++++++++++-----------------------------
 net/decnet/dn_table.c     |   1 +
 9 files changed, 196 insertions(+), 319 deletions(-)

(limited to 'include')

diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index 784a1a29490e..0aaffa2ae666 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -534,7 +534,8 @@ enum rtnetlink_groups {
 	RTNLGRP_NOP2,
 	RTNLGRP_DECnet_ROUTE,
 #define RTNLGRP_DECnet_ROUTE	RTNLGRP_DECnet_ROUTE
-	RTNLGRP_NOP3,
+	RTNLGRP_DECnet_RULE,
+#define RTNLGRP_DECnet_RULE	RTNLGRP_DECnet_RULE
 	RTNLGRP_NOP4,
 	RTNLGRP_IPV6_PREFIX,
 #define RTNLGRP_IPV6_PREFIX	RTNLGRP_IPV6_PREFIX
diff --git a/include/net/dn_fib.h b/include/net/dn_fib.h
index a15dcf0d5c1e..32bc8ce5c5ce 100644
--- a/include/net/dn_fib.h
+++ b/include/net/dn_fib.h
@@ -22,7 +22,7 @@ struct dn_kern_rta
 };
 
 struct dn_fib_res {
-	struct dn_fib_rule *r;
+	struct fib_rule *r;
 	struct dn_fib_info *fi;
 	unsigned char prefixlen;
 	unsigned char nh_sel;
@@ -147,10 +147,8 @@ extern void dn_fib_table_cleanup(void);
  */
 extern void dn_fib_rules_init(void);
 extern void dn_fib_rules_cleanup(void);
-extern void dn_fib_rule_put(struct dn_fib_rule *);
-extern __le16 dn_fib_rules_policy(__le16 saddr, struct dn_fib_res *res, unsigned *flags);
 extern unsigned dnet_addr_type(__le16 addr);
-extern int dn_fib_lookup(const struct flowi *fl, struct dn_fib_res *res);
+extern int dn_fib_lookup(struct flowi *fl, struct dn_fib_res *res);
 
 /*
  * rtnetlink interface
@@ -176,7 +174,7 @@ static inline void dn_fib_res_put(struct dn_fib_res *res)
 	if (res->fi)
 		dn_fib_info_put(res->fi);
 	if (res->r)
-		dn_fib_rule_put(res->r);
+		fib_rule_put(res->r);
 }
 
 extern struct dn_fib_table *dn_fib_tables[];
diff --git a/net/decnet/Kconfig b/net/decnet/Kconfig
index 92f2ec46fd22..36e72cb145b0 100644
--- a/net/decnet/Kconfig
+++ b/net/decnet/Kconfig
@@ -27,6 +27,7 @@ config DECNET
 config DECNET_ROUTER
 	bool "DECnet: router support (EXPERIMENTAL)"
 	depends on DECNET && EXPERIMENTAL
+	select FIB_RULES
 	---help---
 	  Add support for turning your DECnet Endnode into a level 1 or 2
 	  router.  This is an experimental, but functional option.  If you
diff --git a/net/decnet/af_decnet.c b/net/decnet/af_decnet.c
index 5486247735f6..70e027375682 100644
--- a/net/decnet/af_decnet.c
+++ b/net/decnet/af_decnet.c
@@ -130,6 +130,7 @@ Version 0.0.6    2.1.110   07-aug-98   Eduardo Marcelo Serrat
 #include <linux/poll.h>
 #include <net/neighbour.h>
 #include <net/dst.h>
+#include <net/fib_rules.h>
 #include <net/dn.h>
 #include <net/dn_nsp.h>
 #include <net/dn_dev.h>
diff --git a/net/decnet/dn_dev.c b/net/decnet/dn_dev.c
index 632c5a90b589..88ea7a13bb24 100644
--- a/net/decnet/dn_dev.c
+++ b/net/decnet/dn_dev.c
@@ -46,6 +46,7 @@
 #include <net/neighbour.h>
 #include <net/dst.h>
 #include <net/flow.h>
+#include <net/fib_rules.h>
 #include <net/dn.h>
 #include <net/dn_dev.h>
 #include <net/dn_route.h>
@@ -1418,8 +1419,6 @@ static struct rtnetlink_link dnet_rtnetlink_table[RTM_NR_MSGTYPES] =
 	[RTM_DELROUTE - RTM_BASE] = { .doit	= dn_fib_rtm_delroute,	},
 	[RTM_GETROUTE - RTM_BASE] = { .doit	= dn_cache_getroute,
 				      .dumpit	= dn_fib_dump,		},
-	[RTM_NEWRULE  - RTM_BASE] = { .doit	= dn_fib_rtm_newrule,	},
-	[RTM_DELRULE  - RTM_BASE] = { .doit	= dn_fib_rtm_delrule,	},
 	[RTM_GETRULE  - RTM_BASE] = { .dumpit	= dn_fib_dump_rules,	},
 #else
 	[RTM_GETROUTE - RTM_BASE] = { .doit	= dn_cache_getroute,
diff --git a/net/decnet/dn_fib.c b/net/decnet/dn_fib.c
index fa20e2efcfc1..846df3954a63 100644
--- a/net/decnet/dn_fib.c
+++ b/net/decnet/dn_fib.c
@@ -34,6 +34,7 @@
 #include <net/neighbour.h>
 #include <net/dst.h>
 #include <net/flow.h>
+#include <net/fib_rules.h>
 #include <net/dn.h>
 #include <net/dn_route.h>
 #include <net/dn_fib.h>
diff --git a/net/decnet/dn_route.c b/net/decnet/dn_route.c
index 743e9fcf7c5a..5e6f4616ca10 100644
--- a/net/decnet/dn_route.c
+++ b/net/decnet/dn_route.c
@@ -80,6 +80,7 @@
 #include <net/neighbour.h>
 #include <net/dst.h>
 #include <net/flow.h>
+#include <net/fib_rules.h>
 #include <net/dn.h>
 #include <net/dn_dev.h>
 #include <net/dn_nsp.h>
@@ -1284,7 +1285,7 @@ static int dn_route_input_slow(struct sk_buff *skb)
 		dev_hold(out_dev);
 
 		if (res.r)
-			src_map = dn_fib_rules_policy(fl.fld_src, &res, &flags);
+			src_map = fl.fld_src; /* no NAT support for now */
 
 		gateway = DN_FIB_RES_GW(res);
 		if (res.type == RTN_NAT) {
diff --git a/net/decnet/dn_rules.c b/net/decnet/dn_rules.c
index 6986be754ef2..096f1273e714 100644
--- a/net/decnet/dn_rules.c
+++ b/net/decnet/dn_rules.c
@@ -11,259 +11,198 @@
  *
  *
  * Changes:
+ *              Steve Whitehouse <steve@chygwyn.com>
+ *              Updated for Thomas Graf's generic rules
  *
  */
-#include <linux/string.h>
 #include <linux/net.h>
-#include <linux/socket.h>
-#include <linux/sockios.h>
 #include <linux/init.h>
-#include <linux/skbuff.h>
 #include <linux/netlink.h>
 #include <linux/rtnetlink.h>
-#include <linux/proc_fs.h>
 #include <linux/netdevice.h>
-#include <linux/timer.h>
 #include <linux/spinlock.h>
-#include <linux/in_route.h>
 #include <linux/list.h>
 #include <linux/rcupdate.h>
-#include <asm/atomic.h>
-#include <asm/uaccess.h>
 #include <net/neighbour.h>
 #include <net/dst.h>
 #include <net/flow.h>
+#include <net/fib_rules.h>
 #include <net/dn.h>
 #include <net/dn_fib.h>
 #include <net/dn_neigh.h>
 #include <net/dn_dev.h>
 
+static struct fib_rules_ops dn_fib_rules_ops;
+
 struct dn_fib_rule
 {
-	struct hlist_node	r_hlist;
-	atomic_t		r_clntref;
-	u32			r_preference;
-	unsigned char		r_table;
-	unsigned char		r_action;
-	unsigned char		r_dst_len;
-	unsigned char		r_src_len;
-	__le16			r_src;
-	__le16			r_srcmask;
-	__le16			r_dst;
-	__le16			r_dstmask;
-	__le16			r_srcmap;
-	u8			r_flags;
+	struct fib_rule		common;
+	unsigned char		dst_len;
+	unsigned char		src_len;
+	__le16			src;
+	__le16			srcmask;
+	__le16			dst;
+	__le16			dstmask;
+	__le16			srcmap;
+	u8			flags;
 #ifdef CONFIG_DECNET_ROUTE_FWMARK
-	u32			r_fwmark;
+	u32			fwmark;
 #endif
-	int			r_ifindex;
-	char			r_ifname[IFNAMSIZ];
-	int			r_dead;
-	struct rcu_head		rcu;
 };
 
 static struct dn_fib_rule default_rule = {
-	.r_clntref =		ATOMIC_INIT(2),
-	.r_preference =		0x7fff,
-	.r_table =		RT_TABLE_MAIN,
-	.r_action =		RTN_UNICAST
+	.common = {
+		.refcnt =		ATOMIC_INIT(2),
+		.pref =			0x7fff,
+		.table =		RT_TABLE_MAIN,
+		.action =		FR_ACT_TO_TBL,
+	},
 };
 
-static struct hlist_head dn_fib_rules;
+static LIST_HEAD(dn_fib_rules);
+
 
-int dn_fib_rtm_delrule(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
+int dn_fib_lookup(struct flowi *flp, struct dn_fib_res *res)
 {
-	struct rtattr **rta = arg;
-	struct rtmsg *rtm = NLMSG_DATA(nlh);
-	struct dn_fib_rule *r;
-	struct hlist_node *node;
-	int err = -ESRCH;
-
-	hlist_for_each_entry(r, node, &dn_fib_rules, r_hlist) {
-		if ((!rta[RTA_SRC-1] || memcmp(RTA_DATA(rta[RTA_SRC-1]), &r->r_src, 2) == 0) &&
-			rtm->rtm_src_len == r->r_src_len &&
-			rtm->rtm_dst_len == r->r_dst_len &&
-			(!rta[RTA_DST-1] || memcmp(RTA_DATA(rta[RTA_DST-1]), &r->r_dst, 2) == 0) &&
-#ifdef CONFIG_DECNET_ROUTE_FWMARK
-			(!rta[RTA_PROTOINFO-1] || memcmp(RTA_DATA(rta[RTA_PROTOINFO-1]), &r->r_fwmark, 4) == 0) &&
-#endif
-			(!rtm->rtm_type || rtm->rtm_type == r->r_action) &&
-			(!rta[RTA_PRIORITY-1] || memcmp(RTA_DATA(rta[RTA_PRIORITY-1]), &r->r_preference, 4) == 0) &&
-			(!rta[RTA_IIF-1] || rtattr_strcmp(rta[RTA_IIF-1], r->r_ifname) == 0) &&
-			(!rtm->rtm_table || (r && rtm->rtm_table == r->r_table))) {
-
-			err = -EPERM;
-			if (r == &default_rule)
-				break;
-
-			hlist_del_rcu(&r->r_hlist);
-			r->r_dead = 1;
-			dn_fib_rule_put(r);
-			err = 0;
-			break;
-		}
-	}
+	struct fib_lookup_arg arg = {
+		.result = res,
+	};
+	int err;
+
+	err = fib_rules_lookup(&dn_fib_rules_ops, flp, 0, &arg);
+	res->r = arg.rule;
 
 	return err;
 }
 
-static inline void dn_fib_rule_put_rcu(struct rcu_head *head)
+int dn_fib_rule_action(struct fib_rule *rule, struct flowi *flp, int flags,
+		       struct fib_lookup_arg *arg)
 {
-	struct dn_fib_rule *r = container_of(head, struct dn_fib_rule, rcu);
-	kfree(r);
-}
+	int err = -EAGAIN;
+	struct dn_fib_table *tbl;
 
-void dn_fib_rule_put(struct dn_fib_rule *r)
-{
-	if (atomic_dec_and_test(&r->r_clntref)) {
-		if (r->r_dead)
-			call_rcu(&r->rcu, dn_fib_rule_put_rcu);
-		else
-			printk(KERN_DEBUG "Attempt to free alive dn_fib_rule\n");
+	switch(rule->action) {
+	case FR_ACT_TO_TBL:
+		break;
+
+	case FR_ACT_UNREACHABLE:
+		err = -ENETUNREACH;
+		goto errout;
+
+	case FR_ACT_PROHIBIT:
+		err = -EACCES;
+		goto errout;
+
+	case FR_ACT_BLACKHOLE:
+	default:
+		err = -EINVAL;
+		goto errout;
 	}
+
+	tbl = dn_fib_get_table(rule->table, 0);
+	if (tbl == NULL)
+		goto errout;
+
+	err = tbl->lookup(tbl, flp, (struct dn_fib_res *)arg->result);
+	if (err > 0)
+		err = -EAGAIN;
+errout:
+	return err;
 }
 
+static struct nla_policy dn_fib_rule_policy[FRA_MAX+1] __read_mostly = {
+	[FRA_IFNAME]	= { .type = NLA_STRING },
+	[FRA_PRIORITY]	= { .type = NLA_U32 },
+	[FRA_SRC]	= { .type = NLA_U16 },
+	[FRA_DST]	= { .type = NLA_U16 },
+	[FRA_FWMARK]	= { .type = NLA_U32 },
+};
 
-int dn_fib_rtm_newrule(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
+static int dn_fib_rule_match(struct fib_rule *rule, struct flowi *fl, int flags)
 {
-	struct rtattr **rta = arg;
-	struct rtmsg *rtm = NLMSG_DATA(nlh);
-	struct dn_fib_rule *r, *new_r, *last = NULL;
-	struct hlist_node *node = NULL;
-	unsigned char table_id;
-
-	if (rtm->rtm_src_len > 16 || rtm->rtm_dst_len > 16)
-		return -EINVAL;
-
-	if (rta[RTA_IIF-1] && RTA_PAYLOAD(rta[RTA_IIF-1]) > IFNAMSIZ)
-		return -EINVAL;
-
-	if (rtm->rtm_type == RTN_NAT)
-		return -EINVAL;
-
-	table_id = rtm->rtm_table;
-	if (table_id == RT_TABLE_UNSPEC) {
-		struct dn_fib_table *tb;
-		if (rtm->rtm_type == RTN_UNICAST) {
-			if ((tb = dn_fib_empty_table()) == NULL)
-				return -ENOBUFS;
-			table_id = tb->n;
-		}
-	}
+	struct dn_fib_rule *r = (struct dn_fib_rule *)rule;
+	u16 daddr = fl->fld_dst;
+	u16 saddr = fl->fld_src;
+
+	if (((saddr ^ r->src) & r->srcmask) ||
+	    ((daddr ^ r->dst) & r->dstmask))
+		return 0;
 
-	new_r = kzalloc(sizeof(*new_r), GFP_KERNEL);
-	if (!new_r)
-		return -ENOMEM;
-
-	if (rta[RTA_SRC-1])
-		memcpy(&new_r->r_src, RTA_DATA(rta[RTA_SRC-1]), 2);
-	if (rta[RTA_DST-1])
-		memcpy(&new_r->r_dst, RTA_DATA(rta[RTA_DST-1]), 2);
-	if (rta[RTA_GATEWAY-1])
-		memcpy(&new_r->r_srcmap, RTA_DATA(rta[RTA_GATEWAY-1]), 2);
-	new_r->r_src_len = rtm->rtm_src_len;
-	new_r->r_dst_len = rtm->rtm_dst_len;
-	new_r->r_srcmask = dnet_make_mask(rtm->rtm_src_len);
-	new_r->r_dstmask = dnet_make_mask(rtm->rtm_dst_len);
 #ifdef CONFIG_DECNET_ROUTE_FWMARK
-	if (rta[RTA_PROTOINFO-1])
-		memcpy(&new_r->r_fwmark, RTA_DATA(rta[RTA_PROTOINFO-1]), 4);
+	if (r->fwmark && (r->fwmark != fl->fld_fwmark))
+		return 0;
 #endif
-	new_r->r_action = rtm->rtm_type;
-	new_r->r_flags = rtm->rtm_flags;
-	if (rta[RTA_PRIORITY-1])
-		memcpy(&new_r->r_preference, RTA_DATA(rta[RTA_PRIORITY-1]), 4);
-	new_r->r_table = table_id;
-	if (rta[RTA_IIF-1]) {
-		struct net_device *dev;
-		rtattr_strlcpy(new_r->r_ifname, rta[RTA_IIF-1], IFNAMSIZ);
-		new_r->r_ifindex = -1;
-		dev = dev_get_by_name(new_r->r_ifname);
-		if (dev) {
-			new_r->r_ifindex = dev->ifindex;
-			dev_put(dev);
-		}
-	}
 
-	r = container_of(dn_fib_rules.first, struct dn_fib_rule, r_hlist);
-	if (!new_r->r_preference) {
-		if (r && r->r_hlist.next != NULL) {
-			r = container_of(r->r_hlist.next, struct dn_fib_rule, r_hlist);
-			if (r->r_preference)
-				new_r->r_preference = r->r_preference - 1;
+	return 1;
+}
+
+static int dn_fib_rule_configure(struct fib_rule *rule, struct sk_buff *skb,
+				 struct nlmsghdr *nlh, struct fib_rule_hdr *frh,
+				 struct nlattr **tb)
+{
+	int err = -EINVAL;
+	struct dn_fib_rule *r = (struct dn_fib_rule *)rule;
+
+	if (frh->src_len > 16 || frh->dst_len > 16 || frh->tos)
+		goto  errout;
+
+	if (rule->table == RT_TABLE_UNSPEC) {
+		if (rule->action == FR_ACT_TO_TBL) {
+			struct dn_fib_table *table;
+
+			table = dn_fib_empty_table();
+			if (table == NULL) {
+				err = -ENOBUFS;
+				goto errout;
+			}
+
+			rule->table = table->n;
 		}
 	}
 
-	hlist_for_each_entry(r, node, &dn_fib_rules, r_hlist) {
-		if (r->r_preference > new_r->r_preference)
-			break;
-		last = r;
-	}
-	atomic_inc(&new_r->r_clntref);
+	if (tb[FRA_SRC])
+		r->src = nla_get_u16(tb[FRA_SRC]);
 
-	if (last)
-		hlist_add_after_rcu(&last->r_hlist, &new_r->r_hlist);
-	else
-		hlist_add_before_rcu(&new_r->r_hlist, &r->r_hlist);
-	return 0;
-}
+	if (tb[FRA_DST])
+		r->dst = nla_get_u16(tb[FRA_DST]);
 
+#ifdef CONFIG_DECNET_ROUTE_FWMARK
+	if (tb[FRA_FWMARK])
+		r->fwmark = nla_get_u32(tb[FRA_FWMARK]);
+#endif
+
+	r->src_len = frh->src_len;
+	r->srcmask = dnet_make_mask(r->src_len);
+	r->dst_len = frh->dst_len;
+	r->dstmask = dnet_make_mask(r->dst_len);
+	err = 0;
+errout:
+	return err;
+}
 
-int dn_fib_lookup(const struct flowi *flp, struct dn_fib_res *res)
+static int dn_fib_rule_compare(struct fib_rule *rule, struct fib_rule_hdr *frh,
+			       struct nlattr **tb)
 {
-	struct dn_fib_rule *r, *policy;
-	struct dn_fib_table *tb;
-	__le16 saddr = flp->fld_src;
-	__le16 daddr = flp->fld_dst;
-	struct hlist_node *node;
-	int err;
+	struct dn_fib_rule *r = (struct dn_fib_rule *)rule;
+
+	if (frh->src_len && (r->src_len != frh->src_len))
+		return 0;
 
-	rcu_read_lock();
+	if (frh->dst_len && (r->dst_len != frh->dst_len))
+		return 0;
 
-	hlist_for_each_entry_rcu(r, node, &dn_fib_rules, r_hlist) {
-		if (((saddr^r->r_src) & r->r_srcmask) ||
-		    ((daddr^r->r_dst) & r->r_dstmask) ||
 #ifdef CONFIG_DECNET_ROUTE_FWMARK
-		    (r->r_fwmark && r->r_fwmark != flp->fld_fwmark) ||
+	if (tb[FRA_FWMARK] && (r->fwmark != nla_get_u32(tb[FRA_FWMARK])))
+		return 0;
 #endif
-		    (r->r_ifindex && r->r_ifindex != flp->iif))
-			continue;
-
-		switch(r->r_action) {
-			case RTN_UNICAST:
-			case RTN_NAT:
-				policy = r;
-				break;
-			case RTN_UNREACHABLE:
-				rcu_read_unlock();
-				return -ENETUNREACH;
-			default:
-			case RTN_BLACKHOLE:
-				rcu_read_unlock();
-				return -EINVAL;
-			case RTN_PROHIBIT:
-				rcu_read_unlock();
-				return -EACCES;
-		}
 
-		if ((tb = dn_fib_get_table(r->r_table, 0)) == NULL)
-			continue;
-		err = tb->lookup(tb, flp, res);
-		if (err == 0) {
-			res->r = policy;
-			if (policy)
-				atomic_inc(&policy->r_clntref);
-			rcu_read_unlock();
-			return 0;
-		}
-		if (err < 0 && err != -EAGAIN) {
-			rcu_read_unlock();
-			return err;
-		}
-	}
+	if (tb[FRA_SRC] && (r->src != nla_get_u32(tb[FRA_SRC])))
+		return 0;
+
+	if (tb[FRA_DST] && (r->dst != nla_get_u32(tb[FRA_DST])))
+		return 0;
 
-	rcu_read_unlock();
-	return -ESRCH;
+	return 1;
 }
 
 unsigned dnet_addr_type(__le16 addr)
@@ -284,142 +223,77 @@ unsigned dnet_addr_type(__le16 addr)
 	return ret;
 }
 
-__le16 dn_fib_rules_policy(__le16 saddr, struct dn_fib_res *res, unsigned *flags)
+static int dn_fib_rule_fill(struct fib_rule *rule, struct sk_buff *skb,
+			    struct nlmsghdr *nlh, struct fib_rule_hdr *frh)
 {
-	struct dn_fib_rule *r = res->r;
-
-	if (r->r_action == RTN_NAT) {
-		int addrtype = dnet_addr_type(r->r_srcmap);
-
-		if (addrtype == RTN_NAT) {
-			saddr = (saddr&~r->r_srcmask)|r->r_srcmap;
-			*flags |= RTCF_SNAT;
-		} else if (addrtype == RTN_LOCAL || r->r_srcmap == 0) {
-			saddr = r->r_srcmap;
-			*flags |= RTCF_MASQ;
-		}
-	}
-	return saddr;
-}
+	struct dn_fib_rule *r = (struct dn_fib_rule *)rule;
 
-static void dn_fib_rules_detach(struct net_device *dev)
-{
-	struct hlist_node *node;
-	struct dn_fib_rule *r;
+	frh->family = AF_DECnet;
+	frh->dst_len = r->dst_len;
+	frh->src_len = r->src_len;
+	frh->tos = 0;
 
-	hlist_for_each_entry(r, node, &dn_fib_rules, r_hlist) {
-		if (r->r_ifindex == dev->ifindex)
-			r->r_ifindex = -1;
-	}
-}
+#ifdef CONFIG_DECNET_ROUTE_FWMARK
+	if (r->fwmark)
+		NLA_PUT_U32(skb, FRA_FWMARK, r->fwmark);
+#endif
+	if (r->dst_len)
+		NLA_PUT_U16(skb, FRA_DST, r->dst);
+	if (r->src_len)
+		NLA_PUT_U16(skb, FRA_SRC, r->src);
 
-static void dn_fib_rules_attach(struct net_device *dev)
-{
-	struct hlist_node *node;
-	struct dn_fib_rule *r;
+	return 0;
 
-	hlist_for_each_entry(r, node, &dn_fib_rules, r_hlist) {
-		if (r->r_ifindex == -1 && strcmp(dev->name, r->r_ifname) == 0)
-			r->r_ifindex = dev->ifindex;
-	}
+nla_put_failure:
+	return -ENOBUFS;
 }
 
-static int dn_fib_rules_event(struct notifier_block *this, unsigned long event, void *ptr)
+static u32 dn_fib_rule_default_pref(void)
 {
-	struct net_device *dev = ptr;
-
-	switch(event) {
-		case NETDEV_UNREGISTER:
-			dn_fib_rules_detach(dev);
-			dn_fib_sync_down(0, dev, 1);
-		case NETDEV_REGISTER:
-			dn_fib_rules_attach(dev);
-			dn_fib_sync_up(dev);
+	struct list_head *pos;
+	struct fib_rule *rule;
+
+	if (!list_empty(&dn_fib_rules)) {
+		pos = dn_fib_rules.next;
+		if (pos->next != &dn_fib_rules) {
+			rule = list_entry(pos->next, struct fib_rule, list);
+			if (rule->pref)
+				return rule->pref - 1;
+		}
 	}
 
-	return NOTIFY_DONE;
-}
-
-
-static struct notifier_block dn_fib_rules_notifier = {
-	.notifier_call =	dn_fib_rules_event,
-};
-
-static int dn_fib_fill_rule(struct sk_buff *skb, struct dn_fib_rule *r,
-			    struct netlink_callback *cb, unsigned int flags)
-{
-	struct rtmsg *rtm;
-	struct nlmsghdr *nlh;
-	unsigned char *b = skb->tail;
-
-
-	nlh = NLMSG_NEW_ANSWER(skb, cb, RTM_NEWRULE, sizeof(*rtm), flags);
-	rtm = NLMSG_DATA(nlh);
-	rtm->rtm_family = AF_DECnet;
-	rtm->rtm_dst_len = r->r_dst_len;
-	rtm->rtm_src_len = r->r_src_len;
-	rtm->rtm_tos = 0;
-#ifdef CONFIG_DECNET_ROUTE_FWMARK
-	if (r->r_fwmark)
-		RTA_PUT(skb, RTA_PROTOINFO, 4, &r->r_fwmark);
-#endif
-	rtm->rtm_table = r->r_table;
-	rtm->rtm_protocol = 0;
-	rtm->rtm_scope = 0;
-	rtm->rtm_type = r->r_action;
-	rtm->rtm_flags = r->r_flags;
-
-	if (r->r_dst_len)
-		RTA_PUT(skb, RTA_DST, 2, &r->r_dst);
-	if (r->r_src_len)
-		RTA_PUT(skb, RTA_SRC, 2, &r->r_src);
-	if (r->r_ifname[0])
-		RTA_PUT(skb, RTA_IIF, IFNAMSIZ, &r->r_ifname);
-	if (r->r_preference)
-		RTA_PUT(skb, RTA_PRIORITY, 4, &r->r_preference);
-	if (r->r_srcmap)
-		RTA_PUT(skb, RTA_GATEWAY, 2, &r->r_srcmap);
-	nlh->nlmsg_len = skb->tail - b;
-	return skb->len;
-
-nlmsg_failure:
-rtattr_failure:
-	skb_trim(skb, b - skb->data);
-	return -1;
+	return 0;
 }
 
 int dn_fib_dump_rules(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	int idx = 0;
-	int s_idx = cb->args[0];
-	struct dn_fib_rule *r;
-	struct hlist_node *node;
-
-	rcu_read_lock();
-	hlist_for_each_entry(r, node, &dn_fib_rules, r_hlist) {
-		if (idx < s_idx)
-			goto next;
-		if (dn_fib_fill_rule(skb, r, cb, NLM_F_MULTI) < 0)
-			break;
-next:
-		idx++;
-	}
-	rcu_read_unlock();
-	cb->args[0] = idx;
-
-	return skb->len;
+	return fib_rules_dump(skb, cb, AF_DECnet);
 }
 
+static struct fib_rules_ops dn_fib_rules_ops = {
+	.family		= AF_DECnet,
+	.rule_size	= sizeof(struct dn_fib_rule),
+	.action		= dn_fib_rule_action,
+	.match		= dn_fib_rule_match,
+	.configure	= dn_fib_rule_configure,
+	.compare	= dn_fib_rule_compare,
+	.fill		= dn_fib_rule_fill,
+	.default_pref	= dn_fib_rule_default_pref,
+	.nlgroup	= RTNLGRP_DECnet_RULE,
+	.policy		= dn_fib_rule_policy,
+	.rules_list	= &dn_fib_rules,
+	.owner		= THIS_MODULE,
+};
+
 void __init dn_fib_rules_init(void)
 {
-	INIT_HLIST_HEAD(&dn_fib_rules);
-	hlist_add_head(&default_rule.r_hlist, &dn_fib_rules);
-	register_netdevice_notifier(&dn_fib_rules_notifier);
+	list_add_tail(&default_rule.common.list, &dn_fib_rules);
+	fib_rules_register(&dn_fib_rules_ops);
 }
 
 void __exit dn_fib_rules_cleanup(void)
 {
-	unregister_netdevice_notifier(&dn_fib_rules_notifier);
+	fib_rules_unregister(&dn_fib_rules_ops);
 }
 
 
diff --git a/net/decnet/dn_table.c b/net/decnet/dn_table.c
index e926c952e363..2e01b67398c8 100644
--- a/net/decnet/dn_table.c
+++ b/net/decnet/dn_table.c
@@ -30,6 +30,7 @@
 #include <net/neighbour.h>
 #include <net/dst.h>
 #include <net/flow.h>
+#include <net/fib_rules.h>
 #include <net/dn.h>
 #include <net/dn_route.h>
 #include <net/dn_fib.h>
-- 
cgit v1.2.3


From 757dbb494be3309fe41ce4c62f8057d8b41d8897 Mon Sep 17 00:00:00 2001
From: Stephen Hemminger <shemminger@osdl.org>
Date: Wed, 9 Aug 2006 20:50:00 -0700
Subject: [NET]: drop unused elements from net_proto_family

Three values in net_proto_family are defined but never used.

Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/net.h | 5 -----
 1 file changed, 5 deletions(-)

(limited to 'include')

diff --git a/include/linux/net.h b/include/linux/net.h
index 19da2c08d7b6..1bd76327ee2b 100644
--- a/include/linux/net.h
+++ b/include/linux/net.h
@@ -169,11 +169,6 @@ struct proto_ops {
 struct net_proto_family {
 	int		family;
 	int		(*create)(struct socket *sock, int protocol);
-	/* These are counters for the number of different methods of
-	   each we support */
-	short		authentication;
-	short		encryption;
-	short		encrypt_net;
 	struct module	*owner;
 };
 
-- 
cgit v1.2.3


From f0fd27d42e39b91f85e1840ec49b072fd6c545b8 Mon Sep 17 00:00:00 2001
From: Stephen Hemminger <shemminger@osdl.org>
Date: Wed, 9 Aug 2006 21:03:17 -0700
Subject: [NET]: sock_register interface changes

The sock_register() doesn't change the family, so the protocols can
define it read-only.  No caller ever checks return value from
sock_unregister()

Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/net.h |  4 ++--
 net/socket.c        | 10 ++++------
 2 files changed, 6 insertions(+), 8 deletions(-)

(limited to 'include')

diff --git a/include/linux/net.h b/include/linux/net.h
index 1bd76327ee2b..c257f716e00f 100644
--- a/include/linux/net.h
+++ b/include/linux/net.h
@@ -176,8 +176,8 @@ struct iovec;
 struct kvec;
 
 extern int	     sock_wake_async(struct socket *sk, int how, int band);
-extern int	     sock_register(struct net_proto_family *fam);
-extern int	     sock_unregister(int family);
+extern int	     sock_register(const struct net_proto_family *fam);
+extern void	     sock_unregister(int family);
 extern int	     sock_create(int family, int type, int proto,
 				 struct socket **res);
 extern int	     sock_create_kern(int family, int type, int proto,
diff --git a/net/socket.c b/net/socket.c
index b5a3fcb9ed6d..4147fe4bf41d 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -147,7 +147,7 @@ static struct file_operations socket_file_ops = {
  */
 
 static DEFINE_SPINLOCK(net_family_lock);
-static const struct net_proto_family *net_families[NPROTO];
+static const struct net_proto_family *net_families[NPROTO] __read_mostly;
 
 /*
  *	Statistics counters of the socket lists
@@ -2080,7 +2080,7 @@ asmlinkage long sys_socketcall(int call, unsigned long __user *args)
  *	socket interface. The value ops->family coresponds to the
  *	socket system call protocol family.
  */
-int sock_register(struct net_proto_family *ops)
+int sock_register(const struct net_proto_family *ops)
 {
 	int err;
 
@@ -2116,10 +2116,9 @@ int sock_register(struct net_proto_family *ops)
  *	a module then it needs to provide its own protection in
  *	the ops->create routine.
  */
-int sock_unregister(int family)
+void sock_unregister(int family)
 {
-	if (family < 0 || family >= NPROTO)
-		return -EINVAL;
+	BUG_ON(family < 0 || family >= NPROTO);
 
 	spin_lock(&net_family_lock);
 	net_families[family] = NULL;
@@ -2128,7 +2127,6 @@ int sock_unregister(int family)
 	synchronize_rcu();
 
 	printk(KERN_INFO "NET: Unregistered protocol family %d\n", family);
-	return 0;
 }
 
 static int __init sock_init(void)
-- 
cgit v1.2.3


From d924424aaed116b362c6d0e667d912b77e655085 Mon Sep 17 00:00:00 2001
From: Stephen Hemminger <shemminger@osdl.org>
Date: Thu, 10 Aug 2006 23:03:23 -0700
Subject: [NEIGHBOUR]: Use ALIGN() macro.

Rather than opencoding the mask, it looks better to use ALIGN()
macro from kernel.h.

Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/neighbour.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/net/neighbour.h b/include/net/neighbour.h
index 74c4b6ff8a5c..bd187daffdb9 100644
--- a/include/net/neighbour.h
+++ b/include/net/neighbour.h
@@ -101,7 +101,7 @@ struct neighbour
 	__u8			dead;
 	atomic_t		probes;
 	rwlock_t		lock;
-	unsigned char		ha[(MAX_ADDR_LEN+sizeof(unsigned long)-1)&~(sizeof(unsigned long)-1)];
+	unsigned char		ha[ALIGN(MAX_ADDR_LEN, sizeof(unsigned long))];
 	struct hh_cache		*hh;
 	atomic_t		refcnt;
 	int			(*output)(struct sk_buff *skb);
-- 
cgit v1.2.3


From 2dfe55b47e3d66ded5a84caf71e0da5710edf48b Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Thu, 10 Aug 2006 23:08:33 -0700
Subject: [NET]: Use u32 for routing table IDs

Use u32 for routing table IDs in net/ipv4 and net/decnet in preparation of
support for a larger number of routing tables. net/ipv6 already uses u32
everywhere and needs no further changes. No functional changes are made by
this patch.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/dn_fib.h     |  4 ++--
 include/net/ip_fib.h     | 14 +++++++-------
 net/decnet/dn_fib.c      |  6 +++---
 net/decnet/dn_table.c    | 10 +++++-----
 net/ipv4/fib_frontend.c  |  8 ++++----
 net/ipv4/fib_hash.c      |  4 ++--
 net/ipv4/fib_lookup.h    |  4 ++--
 net/ipv4/fib_rules.c     |  2 +-
 net/ipv4/fib_semantics.c |  4 ++--
 net/ipv4/fib_trie.c      |  6 +++---
 10 files changed, 31 insertions(+), 31 deletions(-)

(limited to 'include')

diff --git a/include/net/dn_fib.h b/include/net/dn_fib.h
index 32bc8ce5c5ce..cd9c3782f838 100644
--- a/include/net/dn_fib.h
+++ b/include/net/dn_fib.h
@@ -94,7 +94,7 @@ struct dn_fib_node {
 
 
 struct dn_fib_table {
-	int n;
+	u32 n;
 
 	int (*insert)(struct dn_fib_table *t, struct rtmsg *r, 
 			struct dn_kern_rta *rta, struct nlmsghdr *n, 
@@ -137,7 +137,7 @@ extern int dn_fib_sync_up(struct net_device *dev);
 /*
  * dn_tables.c
  */
-extern struct dn_fib_table *dn_fib_get_table(int n, int creat);
+extern struct dn_fib_table *dn_fib_get_table(u32 n, int creat);
 extern struct dn_fib_table *dn_fib_empty_table(void);
 extern void dn_fib_table_init(void);
 extern void dn_fib_table_cleanup(void);
diff --git a/include/net/ip_fib.h b/include/net/ip_fib.h
index adf73586bc05..0dcbf166eb94 100644
--- a/include/net/ip_fib.h
+++ b/include/net/ip_fib.h
@@ -150,7 +150,7 @@ struct fib_result_nl {
 #endif /* CONFIG_IP_ROUTE_MULTIPATH_WRANDOM */
 
 struct fib_table {
-	unsigned char	tb_id;
+	u32		tb_id;
 	unsigned	tb_stamp;
 	int		(*tb_lookup)(struct fib_table *tb, const struct flowi *flp, struct fib_result *res);
 	int		(*tb_insert)(struct fib_table *table, struct rtmsg *r,
@@ -173,14 +173,14 @@ struct fib_table {
 extern struct fib_table *ip_fib_local_table;
 extern struct fib_table *ip_fib_main_table;
 
-static inline struct fib_table *fib_get_table(int id)
+static inline struct fib_table *fib_get_table(u32 id)
 {
 	if (id != RT_TABLE_LOCAL)
 		return ip_fib_main_table;
 	return ip_fib_local_table;
 }
 
-static inline struct fib_table *fib_new_table(int id)
+static inline struct fib_table *fib_new_table(u32 id)
 {
 	return fib_get_table(id);
 }
@@ -205,9 +205,9 @@ static inline void fib_select_default(const struct flowi *flp, struct fib_result
 
 extern struct fib_table * fib_tables[RT_TABLE_MAX+1];
 extern int fib_lookup(struct flowi *flp, struct fib_result *res);
-extern struct fib_table *__fib_new_table(int id);
+extern struct fib_table *__fib_new_table(u32 id);
 
-static inline struct fib_table *fib_get_table(int id)
+static inline struct fib_table *fib_get_table(u32 id)
 {
 	if (id == 0)
 		id = RT_TABLE_MAIN;
@@ -215,7 +215,7 @@ static inline struct fib_table *fib_get_table(int id)
 	return fib_tables[id];
 }
 
-static inline struct fib_table *fib_new_table(int id)
+static inline struct fib_table *fib_new_table(u32 id)
 {
 	if (id == 0)
 		id = RT_TABLE_MAIN;
@@ -248,7 +248,7 @@ extern int fib_convert_rtentry(int cmd, struct nlmsghdr *nl, struct rtmsg *rtm,
 extern u32  __fib_res_prefsrc(struct fib_result *res);
 
 /* Exported by fib_hash.c */
-extern struct fib_table *fib_hash_init(int id);
+extern struct fib_table *fib_hash_init(u32 id);
 
 #ifdef CONFIG_IP_MULTIPLE_TABLES
 extern int fib4_rules_dump(struct sk_buff *skb, struct netlink_callback *cb);
diff --git a/net/decnet/dn_fib.c b/net/decnet/dn_fib.c
index ed5fb5c3eab5..7b3bf5c3d720 100644
--- a/net/decnet/dn_fib.c
+++ b/net/decnet/dn_fib.c
@@ -534,8 +534,8 @@ int dn_fib_rtm_newroute(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 
 int dn_fib_dump(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	int t;
-	int s_t;
+	u32 t;
+	u32 s_t;
 	struct dn_fib_table *tb;
 
 	if (NLMSG_PAYLOAD(cb->nlh, 0) >= sizeof(struct rtmsg) &&
@@ -765,7 +765,7 @@ void dn_fib_flush(void)
 {
         int flushed = 0;
         struct dn_fib_table *tb;
-        int id;
+        u32 id;
 
         for(id = RT_TABLE_MAX; id > 0; id--) {
                 if ((tb = dn_fib_get_table(id, 0)) == NULL)
diff --git a/net/decnet/dn_table.c b/net/decnet/dn_table.c
index 2e01b67398c8..1601ee5406a8 100644
--- a/net/decnet/dn_table.c
+++ b/net/decnet/dn_table.c
@@ -264,7 +264,7 @@ static int dn_fib_nh_match(struct rtmsg *r, struct nlmsghdr *nlh, struct dn_kern
 }
 
 static int dn_fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event,
-                        u8 tb_id, u8 type, u8 scope, void *dst, int dst_len,
+                        u32 tb_id, u8 type, u8 scope, void *dst, int dst_len,
                         struct dn_fib_info *fi, unsigned int flags)
 {
         struct rtmsg *rtm;
@@ -327,7 +327,7 @@ rtattr_failure:
 }
 
 
-static void dn_rtmsg_fib(int event, struct dn_fib_node *f, int z, int tb_id,
+static void dn_rtmsg_fib(int event, struct dn_fib_node *f, int z, u32 tb_id,
                         struct nlmsghdr *nlh, struct netlink_skb_parms *req)
 {
         struct sk_buff *skb;
@@ -740,7 +740,7 @@ out:
 }
 
 
-struct dn_fib_table *dn_fib_get_table(int n, int create)
+struct dn_fib_table *dn_fib_get_table(u32 n, int create)
 {
         struct dn_fib_table *t;
 
@@ -777,7 +777,7 @@ struct dn_fib_table *dn_fib_get_table(int n, int create)
         return t;
 }
 
-static void dn_fib_del_tree(int n)
+static void dn_fib_del_tree(u32 n)
 {
 	struct dn_fib_table *t;
 
@@ -791,7 +791,7 @@ static void dn_fib_del_tree(int n)
 
 struct dn_fib_table *dn_fib_empty_table(void)
 {
-        int id;
+        u32 id;
 
         for(id = RT_TABLE_MIN; id <= RT_TABLE_MAX; id++)
                 if (dn_fib_tables[id] == NULL)
diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c
index a83f1aa8034e..06f4b23f6f57 100644
--- a/net/ipv4/fib_frontend.c
+++ b/net/ipv4/fib_frontend.c
@@ -62,7 +62,7 @@ struct fib_table *ip_fib_main_table;
 
 struct fib_table *fib_tables[RT_TABLE_MAX+1];
 
-struct fib_table *__fib_new_table(int id)
+struct fib_table *__fib_new_table(u32 id)
 {
 	struct fib_table *tb;
 
@@ -82,7 +82,7 @@ static void fib_flush(void)
 	int flushed = 0;
 #ifdef CONFIG_IP_MULTIPLE_TABLES
 	struct fib_table *tb;
-	int id;
+	u32 id;
 
 	for (id = RT_TABLE_MAX; id>0; id--) {
 		if ((tb = fib_get_table(id))==NULL)
@@ -333,8 +333,8 @@ int inet_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 
 int inet_dump_fib(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	int t;
-	int s_t;
+	u32 t;
+	u32 s_t;
 	struct fib_table *tb;
 
 	if (NLMSG_PAYLOAD(cb->nlh, 0) >= sizeof(struct rtmsg) &&
diff --git a/net/ipv4/fib_hash.c b/net/ipv4/fib_hash.c
index 72c633b357cf..f8d5c8024ccb 100644
--- a/net/ipv4/fib_hash.c
+++ b/net/ipv4/fib_hash.c
@@ -765,9 +765,9 @@ static int fn_hash_dump(struct fib_table *tb, struct sk_buff *skb, struct netlin
 }
 
 #ifdef CONFIG_IP_MULTIPLE_TABLES
-struct fib_table * fib_hash_init(int id)
+struct fib_table * fib_hash_init(u32 id)
 #else
-struct fib_table * __init fib_hash_init(int id)
+struct fib_table * __init fib_hash_init(u32 id)
 #endif
 {
 	struct fib_table *tb;
diff --git a/net/ipv4/fib_lookup.h b/net/ipv4/fib_lookup.h
index ef6609ea0eb7..ddd52496b451 100644
--- a/net/ipv4/fib_lookup.h
+++ b/net/ipv4/fib_lookup.h
@@ -30,11 +30,11 @@ extern struct fib_info *fib_create_info(const struct rtmsg *r,
 extern int fib_nh_match(struct rtmsg *r, struct nlmsghdr *,
 			struct kern_rta *rta, struct fib_info *fi);
 extern int fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event,
-			 u8 tb_id, u8 type, u8 scope, void *dst,
+			 u32 tb_id, u8 type, u8 scope, void *dst,
 			 int dst_len, u8 tos, struct fib_info *fi,
 			 unsigned int);
 extern void rtmsg_fib(int event, u32 key, struct fib_alias *fa,
-		      int z, int tb_id,
+		      int z, u32 tb_id,
 		      struct nlmsghdr *n, struct netlink_skb_parms *req);
 extern struct fib_alias *fib_find_alias(struct list_head *fah,
 					u8 tos, u32 prio);
diff --git a/net/ipv4/fib_rules.c b/net/ipv4/fib_rules.c
index d242e5291fcc..58fb91b00fdf 100644
--- a/net/ipv4/fib_rules.c
+++ b/net/ipv4/fib_rules.c
@@ -169,7 +169,7 @@ static int fib4_rule_match(struct fib_rule *rule, struct flowi *fl, int flags)
 
 static struct fib_table *fib_empty_table(void)
 {
-	int id;
+	u32 id;
 
 	for (id = 1; id <= RT_TABLE_MAX; id++)
 		if (fib_tables[id] == NULL)
diff --git a/net/ipv4/fib_semantics.c b/net/ipv4/fib_semantics.c
index 38bca473c7e2..c7a112b5a185 100644
--- a/net/ipv4/fib_semantics.c
+++ b/net/ipv4/fib_semantics.c
@@ -273,7 +273,7 @@ int ip_fib_check_default(u32 gw, struct net_device *dev)
 }
 
 void rtmsg_fib(int event, u32 key, struct fib_alias *fa,
-	       int z, int tb_id,
+	       int z, u32 tb_id,
 	       struct nlmsghdr *n, struct netlink_skb_parms *req)
 {
 	struct sk_buff *skb;
@@ -939,7 +939,7 @@ u32 __fib_res_prefsrc(struct fib_result *res)
 
 int
 fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event,
-	      u8 tb_id, u8 type, u8 scope, void *dst, int dst_len, u8 tos,
+	      u32 tb_id, u8 type, u8 scope, void *dst, int dst_len, u8 tos,
 	      struct fib_info *fi, unsigned int flags)
 {
 	struct rtmsg *rtm;
diff --git a/net/ipv4/fib_trie.c b/net/ipv4/fib_trie.c
index 01801c0f885d..4a27b2d573a3 100644
--- a/net/ipv4/fib_trie.c
+++ b/net/ipv4/fib_trie.c
@@ -1148,7 +1148,7 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 
 	key = ntohl(key);
 
-	pr_debug("Insert table=%d %08x/%d\n", tb->tb_id, key, plen);
+	pr_debug("Insert table=%u %08x/%d\n", tb->tb_id, key, plen);
 
 	mask = ntohl(inet_make_mask(plen));
 
@@ -1943,9 +1943,9 @@ out:
 /* Fix more generic FIB names for init later */
 
 #ifdef CONFIG_IP_MULTIPLE_TABLES
-struct fib_table * fib_hash_init(int id)
+struct fib_table * fib_hash_init(u32 id)
 #else
-struct fib_table * __init fib_hash_init(int id)
+struct fib_table * __init fib_hash_init(u32 id)
 #endif
 {
 	struct fib_table *tb;
-- 
cgit v1.2.3


From 9e762a4a89b302cb3b26a1f9bb33eff459eaeca9 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Thu, 10 Aug 2006 23:09:48 -0700
Subject: [NET]: Introduce RTA_TABLE/FRA_TABLE attributes

Introduce RTA_TABLE route attribute and FRA_TABLE routing rule attribute
to hold 32 bit routing table IDs. Usespace compatibility is provided by
continuing to accept and send the rtm_table field, but because of its
limited size it can only carry the low 8 bits of the table ID. This
implies that if larger IDs are used, _all_ userspace programs using them
need to use RTA_TABLE.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/fib_rules.h |  4 ++++
 include/linux/rtnetlink.h |  8 ++++++++
 include/net/fib_rules.h   |  7 +++++++
 net/core/fib_rules.c      |  5 +++--
 net/decnet/dn_fib.c       |  7 ++++---
 net/decnet/dn_route.c     |  1 +
 net/decnet/dn_table.c     |  1 +
 net/ipv4/fib_frontend.c   |  7 ++++---
 net/ipv4/fib_rules.c      |  1 +
 net/ipv4/fib_semantics.c  |  1 +
 net/ipv4/route.c          |  1 +
 net/ipv6/fib6_rules.c     |  1 +
 net/ipv6/route.c          | 13 +++++++++----
 13 files changed, 45 insertions(+), 12 deletions(-)

(limited to 'include')

diff --git a/include/linux/fib_rules.h b/include/linux/fib_rules.h
index 5e503f0ca6e4..19a82b6c1c1f 100644
--- a/include/linux/fib_rules.h
+++ b/include/linux/fib_rules.h
@@ -36,6 +36,10 @@ enum
 	FRA_UNUSED5,
 	FRA_FWMARK,	/* netfilter mark (IPv4) */
 	FRA_FLOW,	/* flow/class id */
+	FRA_UNUSED6,
+	FRA_UNUSED7,
+	FRA_UNUSED8,
+	FRA_TABLE,	/* Extended table id */
 	__FRA_MAX
 };
 
diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index 0aaffa2ae666..ea422a539a03 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -264,6 +264,7 @@ enum rtattr_type_t
 	RTA_CACHEINFO,
 	RTA_SESSION,
 	RTA_MP_ALGO,
+	RTA_TABLE,
 	__RTA_MAX
 };
 
@@ -717,6 +718,13 @@ extern void __rtnl_unlock(void);
 	} \
 } while(0)
 
+static inline u32 rtm_get_table(struct rtattr **rta, u8 table)
+{
+	return RTA_GET_U32(rta[RTA_TABLE-1]);
+rtattr_failure:
+	return table;
+}
+
 #endif /* __KERNEL__ */
 
 
diff --git a/include/net/fib_rules.h b/include/net/fib_rules.h
index 61375d9e53f8..8e2f473d3e82 100644
--- a/include/net/fib_rules.h
+++ b/include/net/fib_rules.h
@@ -74,6 +74,13 @@ static inline void fib_rule_put(struct fib_rule *rule)
 		call_rcu(&rule->rcu, fib_rule_put_rcu);
 }
 
+static inline u32 frh_get_table(struct fib_rule_hdr *frh, struct nlattr **nla)
+{
+	if (nla[FRA_TABLE])
+		return nla_get_u32(nla[FRA_TABLE]);
+	return frh->table;
+}
+
 extern int			fib_rules_register(struct fib_rules_ops *);
 extern int			fib_rules_unregister(struct fib_rules_ops *);
 
diff --git a/net/core/fib_rules.c b/net/core/fib_rules.c
index 6cdad24038e2..873b04d5df81 100644
--- a/net/core/fib_rules.c
+++ b/net/core/fib_rules.c
@@ -187,7 +187,7 @@ int fib_nl_newrule(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 
 	rule->action = frh->action;
 	rule->flags = frh->flags;
-	rule->table = frh->table;
+	rule->table = frh_get_table(frh, tb);
 
 	if (!rule->pref && ops->default_pref)
 		rule->pref = ops->default_pref();
@@ -245,7 +245,7 @@ int fib_nl_delrule(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 		if (frh->action && (frh->action != rule->action))
 			continue;
 
-		if (frh->table && (frh->table != rule->table))
+		if (frh->table && (frh_get_table(frh, tb) != rule->table))
 			continue;
 
 		if (tb[FRA_PRIORITY] &&
@@ -291,6 +291,7 @@ static int fib_nl_fill_rule(struct sk_buff *skb, struct fib_rule *rule,
 
 	frh = nlmsg_data(nlh);
 	frh->table = rule->table;
+	NLA_PUT_U32(skb, FRA_TABLE, rule->table);
 	frh->res1 = 0;
 	frh->res2 = 0;
 	frh->action = rule->action;
diff --git a/net/decnet/dn_fib.c b/net/decnet/dn_fib.c
index 7b3bf5c3d720..fb596373daa8 100644
--- a/net/decnet/dn_fib.c
+++ b/net/decnet/dn_fib.c
@@ -491,7 +491,8 @@ static int dn_fib_check_attr(struct rtmsg *r, struct rtattr **rta)
 		if (attr) {
 			if (RTA_PAYLOAD(attr) < 4 && RTA_PAYLOAD(attr) != 2)
 				return -EINVAL;
-			if (i != RTA_MULTIPATH && i != RTA_METRICS)
+			if (i != RTA_MULTIPATH && i != RTA_METRICS &&
+			    i != RTA_TABLE)
 				rta[i-1] = (struct rtattr *)RTA_DATA(attr);
 		}
 	}
@@ -508,7 +509,7 @@ int dn_fib_rtm_delroute(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 	if (dn_fib_check_attr(r, rta))
 		return -EINVAL;
 
-	tb = dn_fib_get_table(r->rtm_table, 0);
+	tb = dn_fib_get_table(rtm_get_table(rta, r->rtm_table), 0);
 	if (tb)
 		return tb->delete(tb, r, (struct dn_kern_rta *)rta, nlh, &NETLINK_CB(skb));
 
@@ -524,7 +525,7 @@ int dn_fib_rtm_newroute(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 	if (dn_fib_check_attr(r, rta))
 		return -EINVAL;
 
-	tb = dn_fib_get_table(r->rtm_table, 1);
+	tb = dn_fib_get_table(rtm_get_table(rta, r->rtm_table), 1);
 	if (tb) 
 		return tb->insert(tb, r, (struct dn_kern_rta *)rta, nlh, &NETLINK_CB(skb));
 
diff --git a/net/decnet/dn_route.c b/net/decnet/dn_route.c
index 5e6f4616ca10..4c963213fba5 100644
--- a/net/decnet/dn_route.c
+++ b/net/decnet/dn_route.c
@@ -1486,6 +1486,7 @@ static int dn_rt_fill_info(struct sk_buff *skb, u32 pid, u32 seq,
 	r->rtm_src_len = 0;
 	r->rtm_tos = 0;
 	r->rtm_table = RT_TABLE_MAIN;
+	RTA_PUT_U32(skb, RTA_TABLE, RT_TABLE_MAIN);
 	r->rtm_type = rt->rt_type;
 	r->rtm_flags = (rt->rt_flags & ~0xFFFF) | RTM_F_CLONED;
 	r->rtm_scope = RT_SCOPE_UNIVERSE;
diff --git a/net/decnet/dn_table.c b/net/decnet/dn_table.c
index 1601ee5406a8..eca7c1e10c80 100644
--- a/net/decnet/dn_table.c
+++ b/net/decnet/dn_table.c
@@ -278,6 +278,7 @@ static int dn_fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event,
         rtm->rtm_src_len = 0;
         rtm->rtm_tos = 0;
         rtm->rtm_table = tb_id;
+	RTA_PUT_U32(skb, RTA_TABLE, tb_id);
         rtm->rtm_flags = fi->fib_flags;
         rtm->rtm_scope = scope;
 	rtm->rtm_type  = type;
diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c
index 06f4b23f6f57..2696ede52de2 100644
--- a/net/ipv4/fib_frontend.c
+++ b/net/ipv4/fib_frontend.c
@@ -294,7 +294,8 @@ static int inet_check_attr(struct rtmsg *r, struct rtattr **rta)
 		if (attr) {
 			if (RTA_PAYLOAD(attr) < 4)
 				return -EINVAL;
-			if (i != RTA_MULTIPATH && i != RTA_METRICS)
+			if (i != RTA_MULTIPATH && i != RTA_METRICS &&
+			    i != RTA_TABLE)
 				*rta = (struct rtattr*)RTA_DATA(attr);
 		}
 	}
@@ -310,7 +311,7 @@ int inet_rtm_delroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 	if (inet_check_attr(r, rta))
 		return -EINVAL;
 
-	tb = fib_get_table(r->rtm_table);
+	tb = fib_get_table(rtm_get_table(rta, r->rtm_table));
 	if (tb)
 		return tb->tb_delete(tb, r, (struct kern_rta*)rta, nlh, &NETLINK_CB(skb));
 	return -ESRCH;
@@ -325,7 +326,7 @@ int inet_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 	if (inet_check_attr(r, rta))
 		return -EINVAL;
 
-	tb = fib_new_table(r->rtm_table);
+	tb = fib_new_table(rtm_get_table(rta, r->rtm_table));
 	if (tb)
 		return tb->tb_insert(tb, r, (struct kern_rta*)rta, nlh, &NETLINK_CB(skb));
 	return -ENOBUFS;
diff --git a/net/ipv4/fib_rules.c b/net/ipv4/fib_rules.c
index 58fb91b00fdf..0330b9cc4b58 100644
--- a/net/ipv4/fib_rules.c
+++ b/net/ipv4/fib_rules.c
@@ -184,6 +184,7 @@ static struct nla_policy fib4_rule_policy[FRA_MAX+1] __read_mostly = {
 	[FRA_DST]	= { .type = NLA_U32 },
 	[FRA_FWMARK]	= { .type = NLA_U32 },
 	[FRA_FLOW]	= { .type = NLA_U32 },
+	[FRA_TABLE]	= { .type = NLA_U32 },
 };
 
 static int fib4_rule_configure(struct fib_rule *rule, struct sk_buff *skb,
diff --git a/net/ipv4/fib_semantics.c b/net/ipv4/fib_semantics.c
index c7a112b5a185..ab753df20a39 100644
--- a/net/ipv4/fib_semantics.c
+++ b/net/ipv4/fib_semantics.c
@@ -953,6 +953,7 @@ fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event,
 	rtm->rtm_src_len = 0;
 	rtm->rtm_tos = tos;
 	rtm->rtm_table = tb_id;
+	RTA_PUT_U32(skb, RTA_TABLE, tb_id);
 	rtm->rtm_type = type;
 	rtm->rtm_flags = fi->fib_flags;
 	rtm->rtm_scope = scope;
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index b873cbcdd0b8..12128b82c9dc 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -2652,6 +2652,7 @@ static int rt_fill_info(struct sk_buff *skb, u32 pid, u32 seq, int event,
 	r->rtm_src_len	= 0;
 	r->rtm_tos	= rt->fl.fl4_tos;
 	r->rtm_table	= RT_TABLE_MAIN;
+	RTA_PUT_U32(skb, RTA_TABLE, RT_TABLE_MAIN);
 	r->rtm_type	= rt->rt_type;
 	r->rtm_scope	= RT_SCOPE_UNIVERSE;
 	r->rtm_protocol = RTPROT_UNSPEC;
diff --git a/net/ipv6/fib6_rules.c b/net/ipv6/fib6_rules.c
index 22a2fdb09831..2c4fbc855e6c 100644
--- a/net/ipv6/fib6_rules.c
+++ b/net/ipv6/fib6_rules.c
@@ -129,6 +129,7 @@ static struct nla_policy fib6_rule_policy[RTA_MAX+1] __read_mostly = {
 	[FRA_PRIORITY]	= { .type = NLA_U32 },
 	[FRA_SRC]	= { .minlen = sizeof(struct in6_addr) },
 	[FRA_DST]	= { .minlen = sizeof(struct in6_addr) },
+	[FRA_TABLE]	= { .type = NLA_U32 },
 };
 
 static int fib6_rule_configure(struct fib_rule *rule, struct sk_buff *skb,
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index e08d84063c1f..843c5509fced 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -1859,7 +1859,8 @@ int inet6_rtm_delroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 
 	if (inet6_rtm_to_rtmsg(r, arg, &rtmsg))
 		return -EINVAL;
-	return ip6_route_del(&rtmsg, nlh, arg, &NETLINK_CB(skb), r->rtm_table);
+	return ip6_route_del(&rtmsg, nlh, arg, &NETLINK_CB(skb),
+			     rtm_get_table(arg, r->rtm_table));
 }
 
 int inet6_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
@@ -1869,7 +1870,8 @@ int inet6_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 
 	if (inet6_rtm_to_rtmsg(r, arg, &rtmsg))
 		return -EINVAL;
-	return ip6_route_add(&rtmsg, nlh, arg, &NETLINK_CB(skb), r->rtm_table);
+	return ip6_route_add(&rtmsg, nlh, arg, &NETLINK_CB(skb),
+			     rtm_get_table(arg, r->rtm_table));
 }
 
 struct rt6_rtnl_dump_arg
@@ -1887,6 +1889,7 @@ static int rt6_fill_node(struct sk_buff *skb, struct rt6_info *rt,
 	struct nlmsghdr  *nlh;
 	unsigned char	 *b = skb->tail;
 	struct rta_cacheinfo ci;
+	u32 table;
 
 	if (prefix) {	/* user wants prefix routes only */
 		if (!(rt->rt6i_flags & RTF_PREFIX_RT)) {
@@ -1902,9 +1905,11 @@ static int rt6_fill_node(struct sk_buff *skb, struct rt6_info *rt,
 	rtm->rtm_src_len = rt->rt6i_src.plen;
 	rtm->rtm_tos = 0;
 	if (rt->rt6i_table)
-		rtm->rtm_table = rt->rt6i_table->tb6_id;
+		table = rt->rt6i_table->tb6_id;
 	else
-		rtm->rtm_table = RT6_TABLE_UNSPEC;
+		table = RT6_TABLE_UNSPEC;
+	rtm->rtm_table = table;
+	RTA_PUT_U32(skb, RTA_TABLE, table);
 	if (rt->rt6i_flags&RTF_REJECT)
 		rtm->rtm_type = RTN_UNREACHABLE;
 	else if (rt->rt6i_dev && (rt->rt6i_dev->flags&IFF_LOOPBACK))
-- 
cgit v1.2.3


From 1af5a8c4a11cfed0c9a7f30fcfb689981750599c Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Thu, 10 Aug 2006 23:10:46 -0700
Subject: [IPV4]: Increase number of possible routing tables to 2^32

Increase the number of possible routing tables to 2^32 by replacing the
fixed sized array of pointers by a hash table and replacing iterations
over all possible table IDs by hash table walking.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip_fib.h    |  25 +++---------
 net/ipv4/fib_frontend.c | 102 ++++++++++++++++++++++++++++++++----------------
 net/ipv4/fib_hash.c     |  26 ++++++------
 net/ipv4/fib_rules.c    |   4 +-
 net/ipv4/fib_trie.c     |  26 ++++++------
 5 files changed, 101 insertions(+), 82 deletions(-)

(limited to 'include')

diff --git a/include/net/ip_fib.h b/include/net/ip_fib.h
index 0dcbf166eb94..8e9ba563d342 100644
--- a/include/net/ip_fib.h
+++ b/include/net/ip_fib.h
@@ -150,6 +150,7 @@ struct fib_result_nl {
 #endif /* CONFIG_IP_ROUTE_MULTIPATH_WRANDOM */
 
 struct fib_table {
+	struct hlist_node tb_hlist;
 	u32		tb_id;
 	unsigned	tb_stamp;
 	int		(*tb_lookup)(struct fib_table *tb, const struct flowi *flp, struct fib_result *res);
@@ -200,29 +201,13 @@ static inline void fib_select_default(const struct flowi *flp, struct fib_result
 }
 
 #else /* CONFIG_IP_MULTIPLE_TABLES */
-#define ip_fib_local_table (fib_tables[RT_TABLE_LOCAL])
-#define ip_fib_main_table (fib_tables[RT_TABLE_MAIN])
+#define ip_fib_local_table fib_get_table(RT_TABLE_LOCAL)
+#define ip_fib_main_table fib_get_table(RT_TABLE_MAIN)
 
-extern struct fib_table * fib_tables[RT_TABLE_MAX+1];
 extern int fib_lookup(struct flowi *flp, struct fib_result *res);
-extern struct fib_table *__fib_new_table(u32 id);
-
-static inline struct fib_table *fib_get_table(u32 id)
-{
-	if (id == 0)
-		id = RT_TABLE_MAIN;
-
-	return fib_tables[id];
-}
-
-static inline struct fib_table *fib_new_table(u32 id)
-{
-	if (id == 0)
-		id = RT_TABLE_MAIN;
-
-	return fib_tables[id] ? : __fib_new_table(id);
-}
 
+extern struct fib_table *fib_new_table(u32 id);
+extern struct fib_table *fib_get_table(u32 id);
 extern void fib_select_default(const struct flowi *flp, struct fib_result *res);
 
 #endif /* CONFIG_IP_MULTIPLE_TABLES */
diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c
index 2696ede52de2..ad4c14f968a1 100644
--- a/net/ipv4/fib_frontend.c
+++ b/net/ipv4/fib_frontend.c
@@ -37,6 +37,7 @@
 #include <linux/skbuff.h>
 #include <linux/netlink.h>
 #include <linux/init.h>
+#include <linux/list.h>
 
 #include <net/ip.h>
 #include <net/protocol.h>
@@ -51,48 +52,67 @@
 
 #ifndef CONFIG_IP_MULTIPLE_TABLES
 
-#define RT_TABLE_MIN RT_TABLE_MAIN
-
 struct fib_table *ip_fib_local_table;
 struct fib_table *ip_fib_main_table;
 
-#else
+#define FIB_TABLE_HASHSZ 1
+static struct hlist_head fib_table_hash[FIB_TABLE_HASHSZ];
 
-#define RT_TABLE_MIN 1
+#else
 
-struct fib_table *fib_tables[RT_TABLE_MAX+1];
+#define FIB_TABLE_HASHSZ 256
+static struct hlist_head fib_table_hash[FIB_TABLE_HASHSZ];
 
-struct fib_table *__fib_new_table(u32 id)
+struct fib_table *fib_new_table(u32 id)
 {
 	struct fib_table *tb;
+	unsigned int h;
 
+	if (id == 0)
+		id = RT_TABLE_MAIN;
+	tb = fib_get_table(id);
+	if (tb)
+		return tb;
 	tb = fib_hash_init(id);
 	if (!tb)
 		return NULL;
-	fib_tables[id] = tb;
+	h = id & (FIB_TABLE_HASHSZ - 1);
+	hlist_add_head_rcu(&tb->tb_hlist, &fib_table_hash[h]);
 	return tb;
 }
 
+struct fib_table *fib_get_table(u32 id)
+{
+	struct fib_table *tb;
+	struct hlist_node *node;
+	unsigned int h;
 
+	if (id == 0)
+		id = RT_TABLE_MAIN;
+	h = id & (FIB_TABLE_HASHSZ - 1);
+	rcu_read_lock();
+	hlist_for_each_entry_rcu(tb, node, &fib_table_hash[h], tb_hlist) {
+		if (tb->tb_id == id) {
+			rcu_read_unlock();
+			return tb;
+		}
+	}
+	rcu_read_unlock();
+	return NULL;
+}
 #endif /* CONFIG_IP_MULTIPLE_TABLES */
 
-
 static void fib_flush(void)
 {
 	int flushed = 0;
-#ifdef CONFIG_IP_MULTIPLE_TABLES
 	struct fib_table *tb;
-	u32 id;
+	struct hlist_node *node;
+	unsigned int h;
 
-	for (id = RT_TABLE_MAX; id>0; id--) {
-		if ((tb = fib_get_table(id))==NULL)
-			continue;
-		flushed += tb->tb_flush(tb);
+	for (h = 0; h < FIB_TABLE_HASHSZ; h++) {
+		hlist_for_each_entry(tb, node, &fib_table_hash[h], tb_hlist)
+			flushed += tb->tb_flush(tb);
 	}
-#else /* CONFIG_IP_MULTIPLE_TABLES */
-	flushed += ip_fib_main_table->tb_flush(ip_fib_main_table);
-	flushed += ip_fib_local_table->tb_flush(ip_fib_local_table);
-#endif /* CONFIG_IP_MULTIPLE_TABLES */
 
 	if (flushed)
 		rt_cache_flush(-1);
@@ -334,29 +354,37 @@ int inet_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 
 int inet_dump_fib(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	u32 t;
-	u32 s_t;
+	unsigned int h, s_h;
+	unsigned int e = 0, s_e;
 	struct fib_table *tb;
+	struct hlist_node *node;
+	int dumped = 0;
 
 	if (NLMSG_PAYLOAD(cb->nlh, 0) >= sizeof(struct rtmsg) &&
 	    ((struct rtmsg*)NLMSG_DATA(cb->nlh))->rtm_flags&RTM_F_CLONED)
 		return ip_rt_dump(skb, cb);
 
-	s_t = cb->args[0];
-	if (s_t == 0)
-		s_t = cb->args[0] = RT_TABLE_MIN;
-
-	for (t=s_t; t<=RT_TABLE_MAX; t++) {
-		if (t < s_t) continue;
-		if (t > s_t)
-			memset(&cb->args[1], 0, sizeof(cb->args)-sizeof(cb->args[0]));
-		if ((tb = fib_get_table(t))==NULL)
-			continue;
-		if (tb->tb_dump(tb, skb, cb) < 0) 
-			break;
+	s_h = cb->args[0];
+	s_e = cb->args[1];
+
+	for (h = s_h; h < FIB_TABLE_HASHSZ; h++, s_e = 0) {
+		e = 0;
+		hlist_for_each_entry(tb, node, &fib_table_hash[h], tb_hlist) {
+			if (e < s_e)
+				goto next;
+			if (dumped)
+				memset(&cb->args[2], 0, sizeof(cb->args) -
+				                 2 * sizeof(cb->args[0]));
+			if (tb->tb_dump(tb, skb, cb) < 0)
+				goto out;
+			dumped = 1;
+next:
+			e++;
+		}
 	}
-
-	cb->args[0] = t;
+out:
+	cb->args[1] = e;
+	cb->args[0] = h;
 
 	return skb->len;
 }
@@ -654,9 +682,15 @@ static struct notifier_block fib_netdev_notifier = {
 
 void __init ip_fib_init(void)
 {
+	unsigned int i;
+
+	for (i = 0; i < FIB_TABLE_HASHSZ; i++)
+		INIT_HLIST_HEAD(&fib_table_hash[i]);
 #ifndef CONFIG_IP_MULTIPLE_TABLES
 	ip_fib_local_table = fib_hash_init(RT_TABLE_LOCAL);
+	hlist_add_head_rcu(&ip_fib_local_table->tb_hlist, &fib_table_hash[0]);
 	ip_fib_main_table  = fib_hash_init(RT_TABLE_MAIN);
+	hlist_add_head_rcu(&ip_fib_main_table->tb_hlist, &fib_table_hash[0]);
 #else
 	fib4_rules_init();
 #endif
diff --git a/net/ipv4/fib_hash.c b/net/ipv4/fib_hash.c
index f8d5c8024ccb..b5bee1a71e5c 100644
--- a/net/ipv4/fib_hash.c
+++ b/net/ipv4/fib_hash.c
@@ -684,7 +684,7 @@ fn_hash_dump_bucket(struct sk_buff *skb, struct netlink_callback *cb,
 	struct fib_node *f;
 	int i, s_i;
 
-	s_i = cb->args[3];
+	s_i = cb->args[4];
 	i = 0;
 	hlist_for_each_entry(f, node, head, fn_hash) {
 		struct fib_alias *fa;
@@ -704,14 +704,14 @@ fn_hash_dump_bucket(struct sk_buff *skb, struct netlink_callback *cb,
 					  fa->fa_tos,
 					  fa->fa_info,
 					  NLM_F_MULTI) < 0) {
-				cb->args[3] = i;
+				cb->args[4] = i;
 				return -1;
 			}
 		next:
 			i++;
 		}
 	}
-	cb->args[3] = i;
+	cb->args[4] = i;
 	return skb->len;
 }
 
@@ -722,21 +722,21 @@ fn_hash_dump_zone(struct sk_buff *skb, struct netlink_callback *cb,
 {
 	int h, s_h;
 
-	s_h = cb->args[2];
+	s_h = cb->args[3];
 	for (h=0; h < fz->fz_divisor; h++) {
 		if (h < s_h) continue;
 		if (h > s_h)
-			memset(&cb->args[3], 0,
-			       sizeof(cb->args) - 3*sizeof(cb->args[0]));
+			memset(&cb->args[4], 0,
+			       sizeof(cb->args) - 4*sizeof(cb->args[0]));
 		if (fz->fz_hash == NULL ||
 		    hlist_empty(&fz->fz_hash[h]))
 			continue;
 		if (fn_hash_dump_bucket(skb, cb, tb, fz, &fz->fz_hash[h])<0) {
-			cb->args[2] = h;
+			cb->args[3] = h;
 			return -1;
 		}
 	}
-	cb->args[2] = h;
+	cb->args[3] = h;
 	return skb->len;
 }
 
@@ -746,21 +746,21 @@ static int fn_hash_dump(struct fib_table *tb, struct sk_buff *skb, struct netlin
 	struct fn_zone *fz;
 	struct fn_hash *table = (struct fn_hash*)tb->tb_data;
 
-	s_m = cb->args[1];
+	s_m = cb->args[2];
 	read_lock(&fib_hash_lock);
 	for (fz = table->fn_zone_list, m=0; fz; fz = fz->fz_next, m++) {
 		if (m < s_m) continue;
 		if (m > s_m)
-			memset(&cb->args[2], 0,
-			       sizeof(cb->args) - 2*sizeof(cb->args[0]));
+			memset(&cb->args[3], 0,
+			       sizeof(cb->args) - 3*sizeof(cb->args[0]));
 		if (fn_hash_dump_zone(skb, cb, tb, fz) < 0) {
-			cb->args[1] = m;
+			cb->args[2] = m;
 			read_unlock(&fib_hash_lock);
 			return -1;
 		}
 	}
 	read_unlock(&fib_hash_lock);
-	cb->args[1] = m;
+	cb->args[2] = m;
 	return skb->len;
 }
 
diff --git a/net/ipv4/fib_rules.c b/net/ipv4/fib_rules.c
index 0330b9cc4b58..ce185ac6f260 100644
--- a/net/ipv4/fib_rules.c
+++ b/net/ipv4/fib_rules.c
@@ -172,8 +172,8 @@ static struct fib_table *fib_empty_table(void)
 	u32 id;
 
 	for (id = 1; id <= RT_TABLE_MAX; id++)
-		if (fib_tables[id] == NULL)
-			return __fib_new_table(id);
+		if (fib_get_table(id) == NULL)
+			return fib_new_table(id);
 	return NULL;
 }
 
diff --git a/net/ipv4/fib_trie.c b/net/ipv4/fib_trie.c
index 4a27b2d573a3..2a580eb2579b 100644
--- a/net/ipv4/fib_trie.c
+++ b/net/ipv4/fib_trie.c
@@ -1848,7 +1848,7 @@ static int fn_trie_dump_fa(t_key key, int plen, struct list_head *fah, struct fi
 
 	u32 xkey = htonl(key);
 
-	s_i = cb->args[3];
+	s_i = cb->args[4];
 	i = 0;
 
 	/* rcu_read_lock is hold by caller */
@@ -1870,12 +1870,12 @@ static int fn_trie_dump_fa(t_key key, int plen, struct list_head *fah, struct fi
 				  plen,
 				  fa->fa_tos,
 				  fa->fa_info, 0) < 0) {
-			cb->args[3] = i;
+			cb->args[4] = i;
 			return -1;
 		}
 		i++;
 	}
-	cb->args[3] = i;
+	cb->args[4] = i;
 	return skb->len;
 }
 
@@ -1886,14 +1886,14 @@ static int fn_trie_dump_plen(struct trie *t, int plen, struct fib_table *tb, str
 	struct list_head *fa_head;
 	struct leaf *l = NULL;
 
-	s_h = cb->args[2];
+	s_h = cb->args[3];
 
 	for (h = 0; (l = nextleaf(t, l)) != NULL; h++) {
 		if (h < s_h)
 			continue;
 		if (h > s_h)
-			memset(&cb->args[3], 0,
-			       sizeof(cb->args) - 3*sizeof(cb->args[0]));
+			memset(&cb->args[4], 0,
+			       sizeof(cb->args) - 4*sizeof(cb->args[0]));
 
 		fa_head = get_fa_head(l, plen);
 
@@ -1904,11 +1904,11 @@ static int fn_trie_dump_plen(struct trie *t, int plen, struct fib_table *tb, str
 			continue;
 
 		if (fn_trie_dump_fa(l->key, plen, fa_head, tb, skb, cb)<0) {
-			cb->args[2] = h;
+			cb->args[3] = h;
 			return -1;
 		}
 	}
-	cb->args[2] = h;
+	cb->args[3] = h;
 	return skb->len;
 }
 
@@ -1917,23 +1917,23 @@ static int fn_trie_dump(struct fib_table *tb, struct sk_buff *skb, struct netlin
 	int m, s_m;
 	struct trie *t = (struct trie *) tb->tb_data;
 
-	s_m = cb->args[1];
+	s_m = cb->args[2];
 
 	rcu_read_lock();
 	for (m = 0; m <= 32; m++) {
 		if (m < s_m)
 			continue;
 		if (m > s_m)
-			memset(&cb->args[2], 0,
-				sizeof(cb->args) - 2*sizeof(cb->args[0]));
+			memset(&cb->args[3], 0,
+				sizeof(cb->args) - 3*sizeof(cb->args[0]));
 
 		if (fn_trie_dump_plen(t, 32-m, tb, skb, cb)<0) {
-			cb->args[1] = m;
+			cb->args[2] = m;
 			goto out;
 		}
 	}
 	rcu_read_unlock();
-	cb->args[1] = m;
+	cb->args[2] = m;
 	return skb->len;
 out:
 	rcu_read_unlock();
-- 
cgit v1.2.3


From 1b43af5480c351dbcb2eef478bafe179cbeb6e83 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Thu, 10 Aug 2006 23:11:17 -0700
Subject: [IPV6]: Increase number of possible routing tables to 2^32

Increase number of possible routing tables to 2^32 by replacing iterations
over all possible table IDs by hash table walking.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip6_route.h |   7 ++
 net/ipv6/ip6_fib.c      | 171 ++++++++++++++++++++++++++++++++++++++++++------
 net/ipv6/route.c        | 128 +-----------------------------------
 3 files changed, 159 insertions(+), 147 deletions(-)

(limited to 'include')

diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index 9bfa3cc6cedb..01bfe404784f 100644
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -137,6 +137,13 @@ extern int inet6_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *a
 extern int inet6_rtm_delroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg);
 extern int inet6_rtm_getroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg);
 
+struct rt6_rtnl_dump_arg
+{
+	struct sk_buff *skb;
+	struct netlink_callback *cb;
+};
+
+extern int rt6_dump_route(struct rt6_info *rt, void *p_arg);
 extern void rt6_ifdown(struct net_device *dev);
 extern void rt6_mtu_change(struct net_device *dev, unsigned mtu);
 
diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
index 1f2316187ca4..bececbe9dd2c 100644
--- a/net/ipv6/ip6_fib.c
+++ b/net/ipv6/ip6_fib.c
@@ -158,7 +158,26 @@ static struct fib6_table fib6_main_tbl = {
 };
 
 #ifdef CONFIG_IPV6_MULTIPLE_TABLES
+#define FIB_TABLE_HASHSZ 256
+#else
+#define FIB_TABLE_HASHSZ 1
+#endif
+static struct hlist_head fib_table_hash[FIB_TABLE_HASHSZ];
+
+static void fib6_link_table(struct fib6_table *tb)
+{
+	unsigned int h;
+
+	h = tb->tb6_id & (FIB_TABLE_HASHSZ - 1);
 
+	/*
+	 * No protection necessary, this is the only list mutatation
+	 * operation, tables never disappear once they exist.
+	 */
+	hlist_add_head_rcu(&tb->tb6_hlist, &fib_table_hash[h]);
+}
+
+#ifdef CONFIG_IPV6_MULTIPLE_TABLES
 static struct fib6_table fib6_local_tbl = {
 	.tb6_id		= RT6_TABLE_LOCAL,
 	.tb6_lock	= RW_LOCK_UNLOCKED,
@@ -168,9 +187,6 @@ static struct fib6_table fib6_local_tbl = {
 	},
 };
 
-#define FIB_TABLE_HASHSZ 256
-static struct hlist_head fib_table_hash[FIB_TABLE_HASHSZ];
-
 static struct fib6_table *fib6_alloc_table(u32 id)
 {
 	struct fib6_table *table;
@@ -186,19 +202,6 @@ static struct fib6_table *fib6_alloc_table(u32 id)
 	return table;
 }
 
-static void fib6_link_table(struct fib6_table *tb)
-{
-	unsigned int h;
-
-	h = tb->tb6_id & (FIB_TABLE_HASHSZ - 1);
-
-	/*
-	 * No protection necessary, this is the only list mutatation
-	 * operation, tables never disappear once they exist.
-	 */
-	hlist_add_head_rcu(&tb->tb6_hlist, &fib_table_hash[h]);
-}
-
 struct fib6_table *fib6_new_table(u32 id)
 {
 	struct fib6_table *tb;
@@ -263,10 +266,135 @@ struct dst_entry *fib6_rule_lookup(struct flowi *fl, int flags,
 
 static void __init fib6_tables_init(void)
 {
+	fib6_link_table(&fib6_main_tbl);
 }
 
 #endif
 
+static int fib6_dump_node(struct fib6_walker_t *w)
+{
+	int res;
+	struct rt6_info *rt;
+
+	for (rt = w->leaf; rt; rt = rt->u.next) {
+		res = rt6_dump_route(rt, w->args);
+		if (res < 0) {
+			/* Frame is full, suspend walking */
+			w->leaf = rt;
+			return 1;
+		}
+		BUG_TRAP(res!=0);
+	}
+	w->leaf = NULL;
+	return 0;
+}
+
+static void fib6_dump_end(struct netlink_callback *cb)
+{
+	struct fib6_walker_t *w = (void*)cb->args[2];
+
+	if (w) {
+		cb->args[2] = 0;
+		kfree(w);
+	}
+	cb->done = (void*)cb->args[3];
+	cb->args[1] = 3;
+}
+
+static int fib6_dump_done(struct netlink_callback *cb)
+{
+	fib6_dump_end(cb);
+	return cb->done ? cb->done(cb) : 0;
+}
+
+static int fib6_dump_table(struct fib6_table *table, struct sk_buff *skb,
+			   struct netlink_callback *cb)
+{
+	struct fib6_walker_t *w;
+	int res;
+
+	w = (void *)cb->args[2];
+	w->root = &table->tb6_root;
+
+	if (cb->args[4] == 0) {
+		read_lock_bh(&table->tb6_lock);
+		res = fib6_walk(w);
+		read_unlock_bh(&table->tb6_lock);
+		if (res > 0)
+			cb->args[4] = 1;
+	} else {
+		read_lock_bh(&table->tb6_lock);
+		res = fib6_walk_continue(w);
+		read_unlock_bh(&table->tb6_lock);
+		if (res != 0) {
+			if (res < 0)
+				fib6_walker_unlink(w);
+			goto end;
+		}
+		fib6_walker_unlink(w);
+		cb->args[4] = 0;
+	}
+end:
+	return res;
+}
+
+int inet6_dump_fib(struct sk_buff *skb, struct netlink_callback *cb)
+{
+	unsigned int h, s_h;
+	unsigned int e = 0, s_e;
+	struct rt6_rtnl_dump_arg arg;
+	struct fib6_walker_t *w;
+	struct fib6_table *tb;
+	struct hlist_node *node;
+	int res = 0;
+
+	s_h = cb->args[0];
+	s_e = cb->args[1];
+
+	w = (void *)cb->args[2];
+	if (w == NULL) {
+		/* New dump:
+		 *
+		 * 1. hook callback destructor.
+		 */
+		cb->args[3] = (long)cb->done;
+		cb->done = fib6_dump_done;
+
+		/*
+		 * 2. allocate and initialize walker.
+		 */
+		w = kzalloc(sizeof(*w), GFP_ATOMIC);
+		if (w == NULL)
+			return -ENOMEM;
+		w->func = fib6_dump_node;
+		cb->args[2] = (long)w;
+	}
+
+	arg.skb = skb;
+	arg.cb = cb;
+	w->args = &arg;
+
+	for (h = s_h; h < FIB_TABLE_HASHSZ; h++, s_e = 0) {
+		e = 0;
+		hlist_for_each_entry(tb, node, &fib_table_hash[h], tb6_hlist) {
+			if (e < s_e)
+				goto next;
+			res = fib6_dump_table(tb, skb, cb);
+			if (res != 0)
+				goto out;
+next:
+			e++;
+		}
+	}
+out:
+	cb->args[1] = e;
+	cb->args[0] = h;
+
+	res = res < 0 ? res : skb->len;
+	if (res <= 0)
+		fib6_dump_end(cb);
+	return res;
+}
 
 /*
  *	Routing Table
@@ -1187,17 +1315,20 @@ static void fib6_clean_tree(struct fib6_node *root,
 void fib6_clean_all(int (*func)(struct rt6_info *, void *arg),
 		    int prune, void *arg)
 {
-	int i;
 	struct fib6_table *table;
+	struct hlist_node *node;
+	unsigned int h;
 
-	for (i = FIB6_TABLE_MIN; i <= FIB6_TABLE_MAX; i++) {
-		table = fib6_get_table(i);
-		if (table != NULL) {
+	rcu_read_lock();
+	for (h = 0; h < FIB_TABLE_HASHSZ; h++) {
+		hlist_for_each_entry_rcu(table, node, &fib_table_hash[h],
+					 tb6_hlist) {
 			write_lock_bh(&table->tb6_lock);
 			fib6_clean_tree(&table->tb6_root, func, prune, arg);
 			write_unlock_bh(&table->tb6_lock);
 		}
 	}
+	rcu_read_unlock();
 }
 
 static int fib6_prune_clone(struct rt6_info *rt, void *arg)
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 843c5509fced..9ce28277f47f 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -1874,12 +1874,6 @@ int inet6_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 			     rtm_get_table(arg, r->rtm_table));
 }
 
-struct rt6_rtnl_dump_arg
-{
-	struct sk_buff *skb;
-	struct netlink_callback *cb;
-};
-
 static int rt6_fill_node(struct sk_buff *skb, struct rt6_info *rt,
 			 struct in6_addr *dst, struct in6_addr *src,
 			 int iif, int type, u32 pid, u32 seq,
@@ -1976,7 +1970,7 @@ rtattr_failure:
 	return -1;
 }
 
-static int rt6_dump_route(struct rt6_info *rt, void *p_arg)
+int rt6_dump_route(struct rt6_info *rt, void *p_arg)
 {
 	struct rt6_rtnl_dump_arg *arg = (struct rt6_rtnl_dump_arg *) p_arg;
 	int prefix;
@@ -1992,126 +1986,6 @@ static int rt6_dump_route(struct rt6_info *rt, void *p_arg)
 		     prefix, NLM_F_MULTI);
 }
 
-static int fib6_dump_node(struct fib6_walker_t *w)
-{
-	int res;
-	struct rt6_info *rt;
-
-	for (rt = w->leaf; rt; rt = rt->u.next) {
-		res = rt6_dump_route(rt, w->args);
-		if (res < 0) {
-			/* Frame is full, suspend walking */
-			w->leaf = rt;
-			return 1;
-		}
-		BUG_TRAP(res!=0);
-	}
-	w->leaf = NULL;
-	return 0;
-}
-
-static void fib6_dump_end(struct netlink_callback *cb)
-{
-	struct fib6_walker_t *w = (void*)cb->args[0];
-
-	if (w) {
-		cb->args[0] = 0;
-		kfree(w);
-	}
-	cb->done = (void*)cb->args[1];
-	cb->args[1] = 0;
-}
-
-static int fib6_dump_done(struct netlink_callback *cb)
-{
-	fib6_dump_end(cb);
-	return cb->done ? cb->done(cb) : 0;
-}
-
-int inet6_dump_fib(struct sk_buff *skb, struct netlink_callback *cb)
-{
-	struct fib6_table *table;
-	struct rt6_rtnl_dump_arg arg;
-	struct fib6_walker_t *w;
-	int i, res = 0;
-
-	arg.skb = skb;
-	arg.cb = cb;
-
-	/*
-	 * cb->args[0] = pointer to walker structure
-	 * cb->args[1] = saved cb->done() pointer
-	 * cb->args[2] = current table being dumped
-	 */
-
-	w = (void*)cb->args[0];
-	if (w == NULL) {
-		/* New dump:
-		 * 
-		 * 1. hook callback destructor.
-		 */
-		cb->args[1] = (long)cb->done;
-		cb->done = fib6_dump_done;
-
-		/*
-		 * 2. allocate and initialize walker.
-		 */
-		w = kzalloc(sizeof(*w), GFP_ATOMIC);
-		if (w == NULL)
-			return -ENOMEM;
-		w->func = fib6_dump_node;
-		w->args = &arg;
-		cb->args[0] = (long)w;
-		cb->args[2] = FIB6_TABLE_MIN;
-	} else {
-		w->args = &arg;
-		i = cb->args[2];
-		if (i > FIB6_TABLE_MAX)
-			goto end;
-
-		table = fib6_get_table(i);
-		if (table != NULL) {
-			read_lock_bh(&table->tb6_lock);
-			w->root = &table->tb6_root;
-			res = fib6_walk_continue(w);
-			read_unlock_bh(&table->tb6_lock);
-			if (res != 0) {
-				if (res < 0)
-					fib6_walker_unlink(w);
-				goto end;
-			}
-		}
-
-		fib6_walker_unlink(w);
-		cb->args[2] = ++i;
-	}
-
-	for (i = cb->args[2]; i <= FIB6_TABLE_MAX; i++) {
-		table = fib6_get_table(i);
-		if (table == NULL)
-			continue;
-
-		read_lock_bh(&table->tb6_lock);
-		w->root = &table->tb6_root;
-		res = fib6_walk(w);
-		read_unlock_bh(&table->tb6_lock);
-		if (res)
-			break;
-	}
-end:
-	cb->args[2] = i;
-
-	res = res < 0 ? res : skb->len;
-	/* res < 0 is an error. (really, impossible)
-	   res == 0 means that dump is complete, but skb still can contain data.
-	   res > 0 dump is not complete, but frame is full.
-	 */
-	/* Destroy walker, if dump of this table is complete. */
-	if (res <= 0)
-		fib6_dump_end(cb);
-	return res;
-}
-
 int inet6_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void *arg)
 {
 	struct rtattr **rta = arg;
-- 
cgit v1.2.3


From abcab268303c22d24fc89fedd35d82271d20f5da Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Thu, 10 Aug 2006 23:11:47 -0700
Subject: [DECNET]: Increase number of possible routing tables to 2^32

Increase the number of possible routing tables to 2^32 by replacing the
fixed sized array of pointers by a hash table and replacing iterations
over all possible table IDs by hash table walking.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/dn_fib.h  |   3 +-
 net/decnet/dn_fib.c   |  49 --------------------
 net/decnet/dn_rules.c |   2 +-
 net/decnet/dn_table.c | 125 ++++++++++++++++++++++++++++++++++++--------------
 4 files changed, 93 insertions(+), 86 deletions(-)

(limited to 'include')

diff --git a/include/net/dn_fib.h b/include/net/dn_fib.h
index cd9c3782f838..d97aa10c463f 100644
--- a/include/net/dn_fib.h
+++ b/include/net/dn_fib.h
@@ -94,6 +94,7 @@ struct dn_fib_node {
 
 
 struct dn_fib_table {
+	struct hlist_node hlist;
 	u32 n;
 
 	int (*insert)(struct dn_fib_table *t, struct rtmsg *r, 
@@ -177,8 +178,6 @@ static inline void dn_fib_res_put(struct dn_fib_res *res)
 		fib_rule_put(res->r);
 }
 
-extern struct dn_fib_table *dn_fib_tables[];
-
 #else /* Endnode */
 
 #define dn_fib_init()  do { } while(0)
diff --git a/net/decnet/dn_fib.c b/net/decnet/dn_fib.c
index fb596373daa8..5ccca3ed53bd 100644
--- a/net/decnet/dn_fib.c
+++ b/net/decnet/dn_fib.c
@@ -532,39 +532,6 @@ int dn_fib_rtm_newroute(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 	return -ENOBUFS;
 }
 
-
-int dn_fib_dump(struct sk_buff *skb, struct netlink_callback *cb)
-{
-	u32 t;
-	u32 s_t;
-	struct dn_fib_table *tb;
-
-	if (NLMSG_PAYLOAD(cb->nlh, 0) >= sizeof(struct rtmsg) &&
-		((struct rtmsg *)NLMSG_DATA(cb->nlh))->rtm_flags&RTM_F_CLONED)
-			return dn_cache_dump(skb, cb);
-
-	s_t = cb->args[0];
-	if (s_t == 0)
-		s_t = cb->args[0] = RT_MIN_TABLE;
-
-	for(t = s_t; t <= RT_TABLE_MAX; t++) {
-		if (t < s_t)
-			continue;
-		if (t > s_t)
-			memset(&cb->args[1], 0,
-			       sizeof(cb->args) - sizeof(cb->args[0]));
-		tb = dn_fib_get_table(t, 0);
-		if (tb == NULL)
-			continue;
-		if (tb->dump(tb, skb, cb) < 0)
-			break;
-	}
-
-	cb->args[0] = t;
-
-	return skb->len;
-}
-
 static void fib_magic(int cmd, int type, __le16 dst, int dst_len, struct dn_ifaddr *ifa)
 {
 	struct dn_fib_table *tb;
@@ -762,22 +729,6 @@ int dn_fib_sync_up(struct net_device *dev)
         return ret;
 }
 
-void dn_fib_flush(void)
-{
-        int flushed = 0;
-        struct dn_fib_table *tb;
-        u32 id;
-
-        for(id = RT_TABLE_MAX; id > 0; id--) {
-                if ((tb = dn_fib_get_table(id, 0)) == NULL)
-                        continue;
-                flushed += tb->flush(tb);
-        }
-
-        if (flushed)
-                dn_rt_cache_flush(-1);
-}
-
 static struct notifier_block dn_fib_dnaddr_notifier = {
 	.notifier_call = dn_fib_dnaddr_event,
 };
diff --git a/net/decnet/dn_rules.c b/net/decnet/dn_rules.c
index 096f1273e714..878312ff34ec 100644
--- a/net/decnet/dn_rules.c
+++ b/net/decnet/dn_rules.c
@@ -210,7 +210,7 @@ unsigned dnet_addr_type(__le16 addr)
 	struct flowi fl = { .nl_u = { .dn_u = { .daddr = addr } } };
 	struct dn_fib_res res;
 	unsigned ret = RTN_UNICAST;
-	struct dn_fib_table *tb = dn_fib_tables[RT_TABLE_LOCAL];
+	struct dn_fib_table *tb = dn_fib_get_table(RT_TABLE_LOCAL, 0);
 
 	res.r = NULL;
 
diff --git a/net/decnet/dn_table.c b/net/decnet/dn_table.c
index eca7c1e10c80..10e87262b6fb 100644
--- a/net/decnet/dn_table.c
+++ b/net/decnet/dn_table.c
@@ -75,9 +75,9 @@ for( ; ((f) = *(fp)) != NULL; (fp) = &(f)->fn_next)
 for( ; ((f) = *(fp)) != NULL && dn_key_eq((f)->fn_key, (key)); (fp) = &(f)->fn_next)
 
 #define RT_TABLE_MIN 1
-
+#define DN_FIB_TABLE_HASHSZ 256
+static struct hlist_head dn_fib_table_hash[DN_FIB_TABLE_HASHSZ];
 static DEFINE_RWLOCK(dn_fib_tables_lock);
-struct dn_fib_table *dn_fib_tables[RT_TABLE_MAX + 1];
 
 static kmem_cache_t *dn_hash_kmem __read_mostly;
 static int dn_fib_hash_zombies;
@@ -361,7 +361,7 @@ static __inline__ int dn_hash_dump_bucket(struct sk_buff *skb,
 {
 	int i, s_i;
 
-	s_i = cb->args[3];
+	s_i = cb->args[4];
 	for(i = 0; f; i++, f = f->fn_next) {
 		if (i < s_i)
 			continue;
@@ -374,11 +374,11 @@ static __inline__ int dn_hash_dump_bucket(struct sk_buff *skb,
 				(f->fn_state & DN_S_ZOMBIE) ? 0 : f->fn_type,
 				f->fn_scope, &f->fn_key, dz->dz_order, 
 				f->fn_info, NLM_F_MULTI) < 0) {
-			cb->args[3] = i;
+			cb->args[4] = i;
 			return -1;
 		}
 	}
-	cb->args[3] = i;
+	cb->args[4] = i;
 	return skb->len;
 }
 
@@ -389,20 +389,20 @@ static __inline__ int dn_hash_dump_zone(struct sk_buff *skb,
 {
 	int h, s_h;
 
-	s_h = cb->args[2];
+	s_h = cb->args[3];
 	for(h = 0; h < dz->dz_divisor; h++) {
 		if (h < s_h)
 			continue;
 		if (h > s_h)
-			memset(&cb->args[3], 0, sizeof(cb->args) - 3*sizeof(cb->args[0]));
+			memset(&cb->args[4], 0, sizeof(cb->args) - 4*sizeof(cb->args[0]));
 		if (dz->dz_hash == NULL || dz->dz_hash[h] == NULL)
 			continue;
 		if (dn_hash_dump_bucket(skb, cb, tb, dz, dz->dz_hash[h]) < 0) {
-			cb->args[2] = h;
+			cb->args[3] = h;
 			return -1;
 		}
 	}
-	cb->args[2] = h;
+	cb->args[3] = h;
 	return skb->len;
 }
 
@@ -413,26 +413,63 @@ static int dn_fib_table_dump(struct dn_fib_table *tb, struct sk_buff *skb,
 	struct dn_zone *dz;
 	struct dn_hash *table = (struct dn_hash *)tb->data;
 
-	s_m = cb->args[1];
+	s_m = cb->args[2];
 	read_lock(&dn_fib_tables_lock);
 	for(dz = table->dh_zone_list, m = 0; dz; dz = dz->dz_next, m++) {
 		if (m < s_m)
 			continue;
 		if (m > s_m)
-			memset(&cb->args[2], 0, sizeof(cb->args) - 2*sizeof(cb->args[0]));
+			memset(&cb->args[3], 0, sizeof(cb->args) - 3*sizeof(cb->args[0]));
 
 		if (dn_hash_dump_zone(skb, cb, tb, dz) < 0) {
-			cb->args[1] = m;
+			cb->args[2] = m;
 			read_unlock(&dn_fib_tables_lock);
 			return -1;
 		}
 	}
 	read_unlock(&dn_fib_tables_lock);
-	cb->args[1] = m;
+	cb->args[2] = m;
 
         return skb->len;
 }
 
+int dn_fib_dump(struct sk_buff *skb, struct netlink_callback *cb)
+{
+	unsigned int h, s_h;
+	unsigned int e = 0, s_e;
+	struct dn_fib_table *tb;
+	struct hlist_node *node;
+	int dumped = 0;
+
+	if (NLMSG_PAYLOAD(cb->nlh, 0) >= sizeof(struct rtmsg) &&
+		((struct rtmsg *)NLMSG_DATA(cb->nlh))->rtm_flags&RTM_F_CLONED)
+			return dn_cache_dump(skb, cb);
+
+	s_h = cb->args[0];
+	s_e = cb->args[1];
+
+	for (h = s_h; h < DN_FIB_TABLE_HASHSZ; h++, s_h = 0) {
+		e = 0;
+		hlist_for_each_entry(tb, node, &dn_fib_table_hash[h], hlist) {
+			if (e < s_e)
+				goto next;
+			if (dumped)
+				memset(&cb->args[2], 0, sizeof(cb->args) -
+				                 2 * sizeof(cb->args[0]));
+			if (tb->dump(tb, skb, cb) < 0)
+				goto out;
+			dumped = 1;
+next:
+			e++;
+		}
+	}
+out:
+	cb->args[1] = e;
+	cb->args[0] = h;
+
+	return skb->len;
+}
+
 static int dn_fib_table_insert(struct dn_fib_table *tb, struct rtmsg *r, struct dn_kern_rta *rta, struct nlmsghdr *n, struct netlink_skb_parms *req)
 {
 	struct dn_hash *table = (struct dn_hash *)tb->data;
@@ -744,6 +781,8 @@ out:
 struct dn_fib_table *dn_fib_get_table(u32 n, int create)
 {
         struct dn_fib_table *t;
+	struct hlist_node *node;
+	unsigned int h;
 
         if (n < RT_TABLE_MIN)
                 return NULL;
@@ -751,8 +790,15 @@ struct dn_fib_table *dn_fib_get_table(u32 n, int create)
         if (n > RT_TABLE_MAX)
                 return NULL;
 
-        if (dn_fib_tables[n]) 
-                return dn_fib_tables[n];
+	h = n & (DN_FIB_TABLE_HASHSZ - 1);
+	rcu_read_lock();
+	hlist_for_each_entry_rcu(t, node, &dn_fib_table_hash[h], hlist) {
+		if (t->n == n) {
+			rcu_read_unlock();
+			return t;
+		}
+	}
+	rcu_read_unlock();
 
         if (!create)
                 return NULL;
@@ -773,33 +819,37 @@ struct dn_fib_table *dn_fib_get_table(u32 n, int create)
         t->flush  = dn_fib_table_flush;
         t->dump = dn_fib_table_dump;
 	memset(t->data, 0, sizeof(struct dn_hash));
-        dn_fib_tables[n] = t;
+	hlist_add_head_rcu(&t->hlist, &dn_fib_table_hash[h]);
 
         return t;
 }
 
-static void dn_fib_del_tree(u32 n)
-{
-	struct dn_fib_table *t;
-
-	write_lock(&dn_fib_tables_lock);
-	t = dn_fib_tables[n];
-	dn_fib_tables[n] = NULL;
-	write_unlock(&dn_fib_tables_lock);
-
-	kfree(t);
-}
-
 struct dn_fib_table *dn_fib_empty_table(void)
 {
         u32 id;
 
         for(id = RT_TABLE_MIN; id <= RT_TABLE_MAX; id++)
-                if (dn_fib_tables[id] == NULL)
+		if (dn_fib_get_table(id, 0) == NULL)
                         return dn_fib_get_table(id, 1);
         return NULL;
 }
 
+void dn_fib_flush(void)
+{
+        int flushed = 0;
+        struct dn_fib_table *tb;
+	struct hlist_node *node;
+	unsigned int h;
+
+	for (h = 0; h < DN_FIB_TABLE_HASHSZ; h++) {
+		hlist_for_each_entry(tb, node, &dn_fib_table_hash[h], hlist)
+	                flushed += tb->flush(tb);
+        }
+
+        if (flushed)
+                dn_rt_cache_flush(-1);
+}
+
 void __init dn_fib_table_init(void)
 {
 	dn_hash_kmem = kmem_cache_create("dn_fib_info_cache",
@@ -810,10 +860,17 @@ void __init dn_fib_table_init(void)
 
 void __exit dn_fib_table_cleanup(void)
 {
-	int i;
-
-	for (i = RT_TABLE_MIN; i <= RT_TABLE_MAX; ++i)
-		dn_fib_del_tree(i);
+	struct dn_fib_table *t;
+	struct hlist_node *node, *next;
+	unsigned int h;
 
-	return;
+	write_lock(&dn_fib_tables_lock);
+	for (h = 0; h < DN_FIB_TABLE_HASHSZ; h++) {
+		hlist_for_each_entry_safe(t, node, next, &dn_fib_table_hash[h],
+		                          hlist) {
+			hlist_del(&t->hlist);
+			kfree(t);
+		}
+	}
+	write_unlock(&dn_fib_tables_lock);
 }
-- 
cgit v1.2.3


From b801f54917b7c6e8540f877ee562cd0725e62ebd Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Thu, 10 Aug 2006 23:12:34 -0700
Subject: [NET]: Increate RT_TABLE_MAX to 2^32

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/rtnetlink.h | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

(limited to 'include')

diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index ea422a539a03..7e4aa48680a7 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -239,10 +239,8 @@ enum rt_class_t
 	RT_TABLE_DEFAULT=253,
 	RT_TABLE_MAIN=254,
 	RT_TABLE_LOCAL=255,
-	__RT_TABLE_MAX
+	RT_TABLE_MAX=0xFFFFFFFF
 };
-#define RT_TABLE_MAX (__RT_TABLE_MAX - 1)
-
 
 
 /* Routing message attributes */
-- 
cgit v1.2.3


From 90d41122f79c8c3687d965dde4c6d30a6e0cac4c Mon Sep 17 00:00:00 2001
From: Adrian Bunk <bunk@stusta.de>
Date: Mon, 14 Aug 2006 23:49:16 -0700
Subject: [IPV6] ip6_fib.c: make code static

Make the following needlessly global code static:
- fib6_walker_lock
- struct fib6_walker_list
- fib6_walk_continue()
- fib6_walk()

Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip6_fib.h | 25 -------------------------
 net/ipv6/ip6_fib.c    | 29 ++++++++++++++++++++++++-----
 2 files changed, 24 insertions(+), 30 deletions(-)

(limited to 'include')

diff --git a/include/net/ip6_fib.h b/include/net/ip6_fib.h
index c0660cea9a2f..69c444209781 100644
--- a/include/net/ip6_fib.h
+++ b/include/net/ip6_fib.h
@@ -92,28 +92,6 @@ struct fib6_walker_t
 	void *args;
 };
 
-extern struct fib6_walker_t fib6_walker_list;
-extern rwlock_t fib6_walker_lock;
-
-static inline void fib6_walker_link(struct fib6_walker_t *w)
-{
-	write_lock_bh(&fib6_walker_lock);
-	w->next = fib6_walker_list.next;
-	w->prev = &fib6_walker_list;
-	w->next->prev = w;
-	w->prev->next = w;
-	write_unlock_bh(&fib6_walker_lock);
-}
-
-static inline void fib6_walker_unlink(struct fib6_walker_t *w)
-{
-	write_lock_bh(&fib6_walker_lock);
-	w->next->prev = w->prev;
-	w->prev->next = w->next;
-	w->prev = w->next = w;
-	write_unlock_bh(&fib6_walker_lock);
-}
-
 struct rt6_statistics {
 	__u32		fib_nodes;
 	__u32		fib_route_nodes;
@@ -195,9 +173,6 @@ struct fib6_node		*fib6_locate(struct fib6_node *root,
 extern void			fib6_clean_all(int (*func)(struct rt6_info *, void *arg),
 					       int prune, void *arg);
 
-extern int			fib6_walk(struct fib6_walker_t *w);
-extern int			fib6_walk_continue(struct fib6_walker_t *w);
-
 extern int			fib6_add(struct fib6_node *root,
 					 struct rt6_info *rt,
 					 struct nlmsghdr *nlh,
diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
index bececbe9dd2c..be36f4acda94 100644
--- a/net/ipv6/ip6_fib.c
+++ b/net/ipv6/ip6_fib.c
@@ -69,8 +69,7 @@ struct fib6_cleaner_t
 	void *arg;
 };
 
-DEFINE_RWLOCK(fib6_walker_lock);
-
+static DEFINE_RWLOCK(fib6_walker_lock);
 
 #ifdef CONFIG_IPV6_SUBTREES
 #define FWS_INIT FWS_S
@@ -82,6 +81,8 @@ DEFINE_RWLOCK(fib6_walker_lock);
 
 static void fib6_prune_clones(struct fib6_node *fn, struct rt6_info *rt);
 static struct fib6_node * fib6_repair_tree(struct fib6_node *fn);
+static int fib6_walk(struct fib6_walker_t *w);
+static int fib6_walk_continue(struct fib6_walker_t *w);
 
 /*
  *	A routing update causes an increase of the serial number on the
@@ -94,13 +95,31 @@ static __u32 rt_sernum;
 
 static DEFINE_TIMER(ip6_fib_timer, fib6_run_gc, 0, 0);
 
-struct fib6_walker_t fib6_walker_list = {
+static struct fib6_walker_t fib6_walker_list = {
 	.prev	= &fib6_walker_list,
 	.next	= &fib6_walker_list, 
 };
 
 #define FOR_WALKERS(w) for ((w)=fib6_walker_list.next; (w) != &fib6_walker_list; (w)=(w)->next)
 
+static inline void fib6_walker_link(struct fib6_walker_t *w)
+{
+	write_lock_bh(&fib6_walker_lock);
+	w->next = fib6_walker_list.next;
+	w->prev = &fib6_walker_list;
+	w->next->prev = w;
+	w->prev->next = w;
+	write_unlock_bh(&fib6_walker_lock);
+}
+
+static inline void fib6_walker_unlink(struct fib6_walker_t *w)
+{
+	write_lock_bh(&fib6_walker_lock);
+	w->next->prev = w->prev;
+	w->prev->next = w->next;
+	w->prev = w->next = w;
+	write_unlock_bh(&fib6_walker_lock);
+}
 static __inline__ u32 fib6_new_sernum(void)
 {
 	u32 n = ++rt_sernum;
@@ -1173,7 +1192,7 @@ int fib6_del(struct rt6_info *rt, struct nlmsghdr *nlh, void *_rtattr, struct ne
  *	<0  -> walk is terminated by an error.
  */
 
-int fib6_walk_continue(struct fib6_walker_t *w)
+static int fib6_walk_continue(struct fib6_walker_t *w)
 {
 	struct fib6_node *fn, *pn;
 
@@ -1247,7 +1266,7 @@ int fib6_walk_continue(struct fib6_walker_t *w)
 	}
 }
 
-int fib6_walk(struct fib6_walker_t *w)
+static int fib6_walk(struct fib6_walker_t *w)
 {
 	int res;
 
-- 
cgit v1.2.3


From 2aa7f36cdb332a32849afbf25fcbf35dce5b1940 Mon Sep 17 00:00:00 2001
From: Adrian Bunk <bunk@stusta.de>
Date: Mon, 14 Aug 2006 23:55:20 -0700
Subject: [DECNET]: cleanups

- make the following needlessly global functions static:
  - dn_fib.c: dn_fib_sync_down()
  - dn_fib.c: dn_fib_sync_up()
  - dn_rules.c: dn_fib_rule_action()
- remove the following unneeded prototype:
  - dn_fib.c: dn_cache_dump()

Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/dn_fib.h  | 3 ---
 net/decnet/dn_fib.c   | 9 +++++----
 net/decnet/dn_rules.c | 4 ++--
 3 files changed, 7 insertions(+), 9 deletions(-)

(limited to 'include')

diff --git a/include/net/dn_fib.h b/include/net/dn_fib.h
index d97aa10c463f..f01626cbbed6 100644
--- a/include/net/dn_fib.h
+++ b/include/net/dn_fib.h
@@ -131,9 +131,6 @@ extern __le16 dn_fib_get_attr16(struct rtattr *attr, int attrlen, int type);
 extern void dn_fib_flush(void);
 extern void dn_fib_select_multipath(const struct flowi *fl,
 					struct dn_fib_res *res);
-extern int dn_fib_sync_down(__le16 local, struct net_device *dev,
-				int force);
-extern int dn_fib_sync_up(struct net_device *dev);
 
 /*
  * dn_tables.c
diff --git a/net/decnet/dn_fib.c b/net/decnet/dn_fib.c
index 5ccca3ed53bd..1cf010124ec5 100644
--- a/net/decnet/dn_fib.c
+++ b/net/decnet/dn_fib.c
@@ -55,8 +55,6 @@
 
 #define endfor_nexthops(fi) }
 
-extern int dn_cache_dump(struct sk_buff *skb, struct netlink_callback *cb);
-
 static DEFINE_SPINLOCK(dn_fib_multipath_lock);
 static struct dn_fib_info *dn_fib_info_list;
 static DEFINE_SPINLOCK(dn_fib_info_lock);
@@ -80,6 +78,9 @@ static struct
 	[RTN_XRESOLVE] =    { .error = -EINVAL, .scope = RT_SCOPE_NOWHERE },
 };
 
+static int dn_fib_sync_down(__le16 local, struct net_device *dev, int force);
+static int dn_fib_sync_up(struct net_device *dev);
+
 void dn_fib_free_info(struct dn_fib_info *fi)
 {
 	if (fi->fib_dead == 0) {
@@ -651,7 +652,7 @@ static int dn_fib_dnaddr_event(struct notifier_block *this, unsigned long event,
 	return NOTIFY_DONE;
 }
 
-int dn_fib_sync_down(__le16 local, struct net_device *dev, int force)
+static int dn_fib_sync_down(__le16 local, struct net_device *dev, int force)
 {
         int ret = 0;
         int scope = RT_SCOPE_NOWHERE;
@@ -695,7 +696,7 @@ int dn_fib_sync_down(__le16 local, struct net_device *dev, int force)
 }
 
 
-int dn_fib_sync_up(struct net_device *dev)
+static int dn_fib_sync_up(struct net_device *dev)
 {
         int ret = 0;
 
diff --git a/net/decnet/dn_rules.c b/net/decnet/dn_rules.c
index 977bb56c3ce4..50e819edf8c7 100644
--- a/net/decnet/dn_rules.c
+++ b/net/decnet/dn_rules.c
@@ -75,8 +75,8 @@ int dn_fib_lookup(struct flowi *flp, struct dn_fib_res *res)
 	return err;
 }
 
-int dn_fib_rule_action(struct fib_rule *rule, struct flowi *flp, int flags,
-		       struct fib_lookup_arg *arg)
+static int dn_fib_rule_action(struct fib_rule *rule, struct flowi *flp,
+			      int flags, struct fib_lookup_arg *arg)
 {
 	int err = -EAGAIN;
 	struct dn_fib_table *tbl;
-- 
cgit v1.2.3


From 81aa646cc4df3779bcbf9d18cc2c0813ee9b3262 Mon Sep 17 00:00:00 2001
From: Martin Bligh <mbligh@google.com>
Date: Mon, 14 Aug 2006 23:57:10 -0700
Subject: [IPV4]: add the UdpSndbufErrors and UdpRcvbufErrors MIBs

Signed-off-by: Martin Bligh <mbligh@google.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
---
 include/linux/snmp.h |  2 ++
 net/ipv4/proc.c      |  2 ++
 net/ipv4/udp.c       | 16 +++++++++++++++-
 3 files changed, 19 insertions(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/linux/snmp.h b/include/linux/snmp.h
index 4db25d5c7cd1..30156556f78d 100644
--- a/include/linux/snmp.h
+++ b/include/linux/snmp.h
@@ -155,6 +155,8 @@ enum
 	UDP_MIB_NOPORTS,			/* NoPorts */
 	UDP_MIB_INERRORS,			/* InErrors */
 	UDP_MIB_OUTDATAGRAMS,			/* OutDatagrams */
+	UDP_MIB_RCVBUFERRORS,			/* RcvbufErrors */
+	UDP_MIB_SNDBUFERRORS,			/* SndbufErrors */
 	__UDP_MIB_MAX
 };
 
diff --git a/net/ipv4/proc.c b/net/ipv4/proc.c
index d61e2a9d394d..9c6cbe3d9fb8 100644
--- a/net/ipv4/proc.c
+++ b/net/ipv4/proc.c
@@ -173,6 +173,8 @@ static const struct snmp_mib snmp4_udp_list[] = {
 	SNMP_MIB_ITEM("NoPorts", UDP_MIB_NOPORTS),
 	SNMP_MIB_ITEM("InErrors", UDP_MIB_INERRORS),
 	SNMP_MIB_ITEM("OutDatagrams", UDP_MIB_OUTDATAGRAMS),
+	SNMP_MIB_ITEM("RcvbufErrors", UDP_MIB_RCVBUFERRORS),
+	SNMP_MIB_ITEM("SndbufErrors", UDP_MIB_SNDBUFERRORS),
 	SNMP_MIB_SENTINEL
 };
 
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index 87152510980c..514c1e9ae810 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -662,6 +662,16 @@ out:
 		UDP_INC_STATS_USER(UDP_MIB_OUTDATAGRAMS);
 		return len;
 	}
+	/*
+	 * ENOBUFS = no kernel mem, SOCK_NOSPACE = no sndbuf space.  Reporting
+	 * ENOBUFS might not be good (it's not tunable per se), but otherwise
+	 * we don't have a good statistic (IpOutDiscards but it can be too many
+	 * things).  We could add another new stat but at least for now that
+	 * seems like overkill.
+	 */
+	if (err == -ENOBUFS || test_bit(SOCK_NOSPACE, &sk->sk_socket->flags)) {
+		UDP_INC_STATS_USER(UDP_MIB_SNDBUFERRORS);
+	}
 	return err;
 
 do_confirm:
@@ -981,6 +991,7 @@ static int udp_encap_rcv(struct sock * sk, struct sk_buff *skb)
 static int udp_queue_rcv_skb(struct sock * sk, struct sk_buff *skb)
 {
 	struct udp_sock *up = udp_sk(sk);
+	int rc;
 
 	/*
 	 *	Charge it to the socket, dropping if the queue is full.
@@ -1027,7 +1038,10 @@ static int udp_queue_rcv_skb(struct sock * sk, struct sk_buff *skb)
 		skb->ip_summed = CHECKSUM_UNNECESSARY;
 	}
 
-	if (sock_queue_rcv_skb(sk,skb)<0) {
+	if ((rc = sock_queue_rcv_skb(sk,skb)) < 0) {
+		/* Note that an ENOMEM error is charged twice */
+		if (rc == -ENOMEM)
+			UDP_INC_STATS_BH(UDP_MIB_RCVBUFERRORS);
 		UDP_INC_STATS_BH(UDP_MIB_INERRORS);
 		kfree_skb(skb);
 		return -1;
-- 
cgit v1.2.3


From 2942e90050569525628a9f34e0daaa9b661b49cc Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Tue, 15 Aug 2006 00:30:25 -0700
Subject: [RTNETLINK]: Use rtnl_unicast() for rtnetlink unicasts

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/rtnetlink.h |  1 +
 net/core/rtnetlink.c      | 10 +++++++---
 net/decnet/dn_route.c     |  4 +---
 net/ipv4/ipmr.c           |  7 ++++---
 net/ipv4/route.c          |  7 +++----
 net/ipv6/addrconf.c       |  4 +---
 net/ipv6/route.c          |  4 +---
 net/sched/act_api.c       |  7 ++-----
 8 files changed, 20 insertions(+), 24 deletions(-)

(limited to 'include')

diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index 7e4aa48680a7..0e4f478e2cb5 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -584,6 +584,7 @@ struct rtnetlink_link
 
 extern struct rtnetlink_link * rtnetlink_links[NPROTO];
 extern int rtnetlink_send(struct sk_buff *skb, u32 pid, u32 group, int echo);
+extern int rtnl_unicast(struct sk_buff *skb, u32 pid);
 extern int rtnetlink_put_metrics(struct sk_buff *skb, u32 *metrics);
 
 extern void __rta_fill(struct sk_buff *skb, int attrtype, int attrlen, const void *data);
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index a1b783a6afc6..e02fa6a33f42 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -166,6 +166,11 @@ int rtnetlink_send(struct sk_buff *skb, u32 pid, unsigned group, int echo)
 	return err;
 }
 
+int rtnl_unicast(struct sk_buff *skb, u32 pid)
+{
+	return nlmsg_unicast(rtnl, skb, pid);
+}
+
 int rtnetlink_put_metrics(struct sk_buff *skb, u32 *metrics)
 {
 	struct rtattr *mx = (struct rtattr*)skb->tail;
@@ -574,9 +579,7 @@ static int rtnl_getlink(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 		goto errout;
 	}
 
-	err = netlink_unicast(rtnl, skb, NETLINK_CB(skb).pid, MSG_DONTWAIT);
-	if (err > 0)
-		err = 0;
+	err = rtnl_unicast(skb, NETLINK_CB(skb).pid);
 errout:
 	kfree(iw_buf);
 	dev_put(dev);
@@ -825,3 +828,4 @@ EXPORT_SYMBOL(rtnl);
 EXPORT_SYMBOL(rtnl_lock);
 EXPORT_SYMBOL(rtnl_trylock);
 EXPORT_SYMBOL(rtnl_unlock);
+EXPORT_SYMBOL(rtnl_unicast);
diff --git a/net/decnet/dn_route.c b/net/decnet/dn_route.c
index 4c963213fba5..c5daf3557c1f 100644
--- a/net/decnet/dn_route.c
+++ b/net/decnet/dn_route.c
@@ -1611,9 +1611,7 @@ int dn_cache_getroute(struct sk_buff *in_skb, struct nlmsghdr *nlh, void *arg)
 		goto out_free;
 	}
 
-	err = netlink_unicast(rtnl, skb, NETLINK_CB(in_skb).pid, MSG_DONTWAIT);
-
-	return err;
+	return rtnl_unicast(skb, NETLINK_CB(in_skb).pid);
 
 out_free:
 	kfree_skb(skb);
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index 85893eef6b16..98f0aa0d4216 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -312,7 +312,8 @@ static void ipmr_destroy_unres(struct mfc_cache *c)
 			e = NLMSG_DATA(nlh);
 			e->error = -ETIMEDOUT;
 			memset(&e->msg, 0, sizeof(e->msg));
-			netlink_unicast(rtnl, skb, NETLINK_CB(skb).dst_pid, MSG_DONTWAIT);
+
+			rtnl_unicast(skb, NETLINK_CB(skb).pid);
 		} else
 			kfree_skb(skb);
 	}
@@ -512,7 +513,6 @@ static void ipmr_cache_resolve(struct mfc_cache *uc, struct mfc_cache *c)
 
 	while((skb=__skb_dequeue(&uc->mfc_un.unres.unresolved))) {
 		if (skb->nh.iph->version == 0) {
-			int err;
 			struct nlmsghdr *nlh = (struct nlmsghdr *)skb_pull(skb, sizeof(struct iphdr));
 
 			if (ipmr_fill_mroute(skb, c, NLMSG_DATA(nlh)) > 0) {
@@ -525,7 +525,8 @@ static void ipmr_cache_resolve(struct mfc_cache *uc, struct mfc_cache *c)
 				e->error = -EMSGSIZE;
 				memset(&e->msg, 0, sizeof(e->msg));
 			}
-			err = netlink_unicast(rtnl, skb, NETLINK_CB(skb).dst_pid, MSG_DONTWAIT);
+
+			rtnl_unicast(skb, NETLINK_CB(skb).pid);
 		} else
 			ip_mr_forward(skb, c, 0);
 	}
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 12128b82c9dc..b8f6cadc5b3a 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -2809,10 +2809,9 @@ int inet_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void *arg)
 		goto out_free;
 	}
 
-	err = netlink_unicast(rtnl, skb, NETLINK_CB(in_skb).pid, MSG_DONTWAIT);
-	if (err > 0)
-		err = 0;
-out:	return err;
+	err = rtnl_unicast(skb, NETLINK_CB(in_skb).pid);
+out:
+	return err;
 
 out_free:
 	kfree_skb(skb);
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 9ba1e811ba50..4f991a2234d0 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -3268,9 +3268,7 @@ static int inet6_rtm_getaddr(struct sk_buff *in_skb,
 		goto out_free;
 	}
 
-	err = netlink_unicast(rtnl, skb, NETLINK_CB(in_skb).pid, MSG_DONTWAIT);
-	if (err > 0)
-		err = 0;
+	err = rtnl_unicast(skb, NETLINK_CB(in_skb).pid);
 out:
 	in6_ifa_put(ifa);
 	return err;
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 9ce28277f47f..024c8e26c2ec 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -2044,9 +2044,7 @@ int inet6_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void *arg)
 		goto out_free;
 	}
 
-	err = netlink_unicast(rtnl, skb, NETLINK_CB(in_skb).pid, MSG_DONTWAIT);
-	if (err > 0)
-		err = 0;
+	err = rtnl_unicast(skb, NETLINK_CB(in_skb).pid);
 out:
 	return err;
 out_free:
diff --git a/net/sched/act_api.c b/net/sched/act_api.c
index a2587b52e531..6990747d6d5a 100644
--- a/net/sched/act_api.c
+++ b/net/sched/act_api.c
@@ -459,7 +459,6 @@ static int
 act_get_notify(u32 pid, struct nlmsghdr *n, struct tc_action *a, int event)
 {
 	struct sk_buff *skb;
-	int err = 0;
 
 	skb = alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL);
 	if (!skb)
@@ -468,10 +467,8 @@ act_get_notify(u32 pid, struct nlmsghdr *n, struct tc_action *a, int event)
 		kfree_skb(skb);
 		return -EINVAL;
 	}
-	err = netlink_unicast(rtnl, skb, pid, MSG_DONTWAIT);
-	if (err > 0)
-		err = 0;
-	return err;
+
+	return rtnl_unicast(skb, pid);
 }
 
 static struct tc_action *
-- 
cgit v1.2.3


From d387f6ad10764fc2174373b4a1cca443adee36e3 Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Tue, 15 Aug 2006 00:31:06 -0700
Subject: [NETLINK]: Add notification message sending interface

Adds nlmsg_notify() implementing proper notification logic. The
message is multicasted to all listeners in the group. The
applications the requests orignates from can request a unicast
back report in which case said socket will be excluded from the
multicast to avoid duplicated notifications.

nlmsg_multicast() is extended to take allocation flags to
allow notification in atomic contexts.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/genetlink.h      |  5 +++--
 include/net/netlink.h        |  6 ++++--
 net/netlabel/netlabel_user.c |  2 +-
 net/netlink/af_netlink.c     | 34 +++++++++++++++++++++++++++++++++-
 net/netlink/genetlink.c      |  2 +-
 5 files changed, 42 insertions(+), 7 deletions(-)

(limited to 'include')

diff --git a/include/net/genetlink.h b/include/net/genetlink.h
index 8c2287264266..97d6d3aba9d2 100644
--- a/include/net/genetlink.h
+++ b/include/net/genetlink.h
@@ -133,11 +133,12 @@ static inline int genlmsg_cancel(struct sk_buff *skb, void *hdr)
  * @skb: netlink message as socket buffer
  * @pid: own netlink pid to avoid sending to yourself
  * @group: multicast group id
+ * @flags: allocation flags
  */
 static inline int genlmsg_multicast(struct sk_buff *skb, u32 pid,
-				    unsigned int group)
+				    unsigned int group, gfp_t flags)
 {
-	return nlmsg_multicast(genl_sock, skb, pid, group);
+	return nlmsg_multicast(genl_sock, skb, pid, group, flags);
 }
 
 /**
diff --git a/include/net/netlink.h b/include/net/netlink.h
index 3a5e40b1e045..b154b81d9a7a 100644
--- a/include/net/netlink.h
+++ b/include/net/netlink.h
@@ -43,6 +43,7 @@
  * Message Sending:
  *   nlmsg_multicast()			multicast message to several groups
  *   nlmsg_unicast()			unicast a message to a single socket
+ *   nlmsg_notify()			send notification message
  *
  * Message Length Calculations:
  *   nlmsg_msg_size(payload)		length of message w/o padding
@@ -545,15 +546,16 @@ static inline void nlmsg_free(struct sk_buff *skb)
  * @skb: netlink message as socket buffer
  * @pid: own netlink pid to avoid sending to yourself
  * @group: multicast group id
+ * @flags: allocation flags
  */
 static inline int nlmsg_multicast(struct sock *sk, struct sk_buff *skb,
-				  u32 pid, unsigned int group)
+				  u32 pid, unsigned int group, gfp_t flags)
 {
 	int err;
 
 	NETLINK_CB(skb).dst_group = group;
 
-	err = netlink_broadcast(sk, skb, pid, group, GFP_KERNEL);
+	err = netlink_broadcast(sk, skb, pid, group, flags);
 	if (err > 0)
 		err = 0;
 
diff --git a/net/netlabel/netlabel_user.c b/net/netlabel/netlabel_user.c
index 80022221b0a7..73cbe66e42ff 100644
--- a/net/netlabel/netlabel_user.c
+++ b/net/netlabel/netlabel_user.c
@@ -154,5 +154,5 @@ int netlbl_netlink_snd(struct sk_buff *skb, u32 pid)
  */
 int netlbl_netlink_snd_multicast(struct sk_buff *skb, u32 pid, u32 group)
 {
-	return genlmsg_multicast(skb, pid, group);
+	return genlmsg_multicast(skb, pid, group, GFP_KERNEL);
 }
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 0f36ddc0b72d..a80e4456e204 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -1549,6 +1549,38 @@ void netlink_queue_skip(struct nlmsghdr *nlh, struct sk_buff *skb)
 	skb_pull(skb, msglen);
 }
 
+/**
+ * nlmsg_notify - send a notification netlink message
+ * @sk: netlink socket to use
+ * @skb: notification message
+ * @pid: destination netlink pid for reports or 0
+ * @group: destination multicast group or 0
+ * @report: 1 to report back, 0 to disable
+ * @flags: allocation flags
+ */
+int nlmsg_notify(struct sock *sk, struct sk_buff *skb, u32 pid,
+		 unsigned int group, int report, gfp_t flags)
+{
+	int err = 0;
+
+	if (group) {
+		int exclude_pid = 0;
+
+		if (report) {
+			atomic_inc(&skb->users);
+			exclude_pid = pid;
+		}
+
+		/* errors reported via destination sk->sk_err */
+		nlmsg_multicast(sk, skb, exclude_pid, group, flags);
+	}
+
+	if (report)
+		err = nlmsg_unicast(sk, skb, pid);
+
+	return err;
+}
+
 #ifdef CONFIG_PROC_FS
 struct nl_seq_iter {
 	int link;
@@ -1802,4 +1834,4 @@ EXPORT_SYMBOL(netlink_set_err);
 EXPORT_SYMBOL(netlink_set_nonroot);
 EXPORT_SYMBOL(netlink_unicast);
 EXPORT_SYMBOL(netlink_unregister_notifier);
-
+EXPORT_SYMBOL(nlmsg_notify);
diff --git a/net/netlink/genetlink.c b/net/netlink/genetlink.c
index 75bb47a898dd..d32599116c56 100644
--- a/net/netlink/genetlink.c
+++ b/net/netlink/genetlink.c
@@ -510,7 +510,7 @@ static int genl_ctrl_event(int event, void *data)
 		if (IS_ERR(msg))
 			return PTR_ERR(msg);
 
-		genlmsg_multicast(msg, 0, GENL_ID_CTRL);
+		genlmsg_multicast(msg, 0, GENL_ID_CTRL, GFP_KERNEL);
 		break;
 	}
 
-- 
cgit v1.2.3


From 97676b6b5538b3e059d33b8338e7d5cc41c5f1f1 Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Tue, 15 Aug 2006 00:31:41 -0700
Subject: [RTNETLINK]: Add rtnetlink notification interface

Adds rtnl_notify() to send rtnetlink notification messages and
rtnl_set_sk_err() to report notification errors as socket
errors in order to indicate the need of a resync due to loss
of events.

nlmsg_report() is added to properly document the meaning of
NLM_F_ECHO.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/rtnetlink.h |  3 +++
 include/net/netlink.h     | 17 +++++++++++++++++
 net/core/rtnetlink.c      | 18 ++++++++++++++++++
 3 files changed, 38 insertions(+)

(limited to 'include')

diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index 0e4f478e2cb5..ecbe0349060c 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -585,6 +585,9 @@ struct rtnetlink_link
 extern struct rtnetlink_link * rtnetlink_links[NPROTO];
 extern int rtnetlink_send(struct sk_buff *skb, u32 pid, u32 group, int echo);
 extern int rtnl_unicast(struct sk_buff *skb, u32 pid);
+extern int rtnl_notify(struct sk_buff *skb, u32 pid, u32 group,
+		       struct nlmsghdr *nlh, gfp_t flags);
+extern void rtnl_set_sk_err(u32 group, int error);
 extern int rtnetlink_put_metrics(struct sk_buff *skb, u32 *metrics);
 
 extern void __rta_fill(struct sk_buff *skb, int attrtype, int attrlen, const void *data);
diff --git a/include/net/netlink.h b/include/net/netlink.h
index b154b81d9a7a..bf593eb59e1b 100644
--- a/include/net/netlink.h
+++ b/include/net/netlink.h
@@ -65,6 +65,9 @@
  *   nlmsg_validate()			validate netlink message incl. attrs
  *   nlmsg_for_each_attr()		loop over all attributes
  *
+ * Misc:
+ *   nlmsg_report()			report back to application?
+ *
  * ------------------------------------------------------------------------
  *                          Attributes Interface
  * ------------------------------------------------------------------------
@@ -194,6 +197,9 @@ extern void		netlink_run_queue(struct sock *sk, unsigned int *qlen,
 						    struct nlmsghdr *, int *));
 extern void		netlink_queue_skip(struct nlmsghdr *nlh,
 					   struct sk_buff *skb);
+extern int		nlmsg_notify(struct sock *sk, struct sk_buff *skb,
+				     u32 pid, unsigned int group, int report,
+				     gfp_t flags);
 
 extern int		nla_validate(struct nlattr *head, int len, int maxtype,
 				     struct nla_policy *policy);
@@ -375,6 +381,17 @@ static inline int nlmsg_validate(struct nlmsghdr *nlh, int hdrlen, int maxtype,
 			    nlmsg_attrlen(nlh, hdrlen), maxtype, policy);
 }
 
+/**
+ * nlmsg_report - need to report back to application?
+ * @nlh: netlink message header
+ *
+ * Returns 1 if a report back to the application is requested.
+ */
+static inline int nlmsg_report(struct nlmsghdr *nlh)
+{
+	return !!(nlh->nlmsg_flags & NLM_F_ECHO);
+}
+
 /**
  * nlmsg_for_each_attr - iterate over a stream of attributes
  * @pos: loop counter, set to current attribute
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index e02fa6a33f42..2b1af17e6389 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -171,6 +171,22 @@ int rtnl_unicast(struct sk_buff *skb, u32 pid)
 	return nlmsg_unicast(rtnl, skb, pid);
 }
 
+int rtnl_notify(struct sk_buff *skb, u32 pid, u32 group,
+		struct nlmsghdr *nlh, gfp_t flags)
+{
+	int report = 0;
+
+	if (nlh)
+		report = nlmsg_report(nlh);
+
+	return nlmsg_notify(rtnl, skb, pid, group, report, flags);
+}
+
+void rtnl_set_sk_err(u32 group, int error)
+{
+	netlink_set_err(rtnl, 0, group, error);
+}
+
 int rtnetlink_put_metrics(struct sk_buff *skb, u32 *metrics)
 {
 	struct rtattr *mx = (struct rtattr*)skb->tail;
@@ -829,3 +845,5 @@ EXPORT_SYMBOL(rtnl_lock);
 EXPORT_SYMBOL(rtnl_trylock);
 EXPORT_SYMBOL(rtnl_unlock);
 EXPORT_SYMBOL(rtnl_unicast);
+EXPORT_SYMBOL(rtnl_notify);
+EXPORT_SYMBOL(rtnl_set_sk_err);
-- 
cgit v1.2.3


From 56fc85ac961e2c20dcb5ef07e2628b3f93de2e49 Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Tue, 15 Aug 2006 00:37:29 -0700
Subject: [RTNETLINK]: Unexport rtnl socket

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/rtnetlink.h | 2 --
 net/core/rtnetlink.c      | 4 +---
 2 files changed, 1 insertion(+), 5 deletions(-)

(limited to 'include')

diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index ecbe0349060c..9c92dc8b9a08 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -574,8 +574,6 @@ extern int rtattr_parse(struct rtattr *tb[], int maxattr, struct rtattr *rta, in
 #define rtattr_parse_nested(tb, max, rta) \
 	rtattr_parse((tb), (max), RTA_DATA((rta)), RTA_PAYLOAD((rta)))
 
-extern struct sock *rtnl;
-
 struct rtnetlink_link
 {
 	int (*doit)(struct sk_buff *, struct nlmsghdr*, void *attr);
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index f5300b5dd0fd..dfc58269240a 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -58,6 +58,7 @@
 #endif	/* CONFIG_NET_WIRELESS_RTNETLINK */
 
 static DEFINE_MUTEX(rtnl_mutex);
+static struct sock *rtnl;
 
 void rtnl_lock(void)
 {
@@ -95,8 +96,6 @@ int rtattr_parse(struct rtattr *tb[], int maxattr, struct rtattr *rta, int len)
 	return 0;
 }
 
-struct sock *rtnl;
-
 struct rtnetlink_link * rtnetlink_links[NPROTO];
 
 static const int rtm_min[RTM_NR_FAMILIES] =
@@ -842,7 +841,6 @@ EXPORT_SYMBOL(rtattr_strlcpy);
 EXPORT_SYMBOL(rtattr_parse);
 EXPORT_SYMBOL(rtnetlink_links);
 EXPORT_SYMBOL(rtnetlink_put_metrics);
-EXPORT_SYMBOL(rtnl);
 EXPORT_SYMBOL(rtnl_lock);
 EXPORT_SYMBOL(rtnl_trylock);
 EXPORT_SYMBOL(rtnl_unlock);
-- 
cgit v1.2.3


From 4e902c57417c4c285b98ba2722468d1c3ed83d1b Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Thu, 17 Aug 2006 18:14:52 -0700
Subject: [IPv4]: FIB configuration using struct fib_config

Introduces struct fib_config replacing the ugly struct kern_rta
prone to ordering issues. Avoids creating faked netlink messages
for auto generated routes or requests via ioctl.

A new interface net/nexthop.h is added to help navigate through
nexthop configuration arrays.

A new struct nl_info will be used to carry the necessary netlink
information to be used for notifications later on.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip_fib.h     |  55 +++----
 include/net/netlink.h    |  10 ++
 include/net/nexthop.h    |  33 ++++
 net/ipv4/fib_frontend.c  | 364 +++++++++++++++++++++++++++++++++++---------
 net/ipv4/fib_hash.c      |  94 ++++++------
 net/ipv4/fib_lookup.h    |  11 +-
 net/ipv4/fib_semantics.c | 385 ++++++++++++++---------------------------------
 net/ipv4/fib_trie.c      |  76 ++++------
 8 files changed, 560 insertions(+), 468 deletions(-)
 create mode 100644 include/net/nexthop.h

(limited to 'include')

diff --git a/include/net/ip_fib.h b/include/net/ip_fib.h
index 8e9ba563d342..42ed96fab3f5 100644
--- a/include/net/ip_fib.h
+++ b/include/net/ip_fib.h
@@ -20,25 +20,32 @@
 #include <linux/seq_file.h>
 #include <net/fib_rules.h>
 
-/* WARNING: The ordering of these elements must match ordering
- *          of RTA_* rtnetlink attribute numbers.
- */
-struct kern_rta {
-	void		*rta_dst;
-	void		*rta_src;
-	int		*rta_iif;
-	int		*rta_oif;
-	void		*rta_gw;
-	u32		*rta_priority;
-	void		*rta_prefsrc;
-	struct rtattr	*rta_mx;
-	struct rtattr	*rta_mp;
-	unsigned char	*rta_protoinfo;
-	u32		*rta_flow;
-	struct rta_cacheinfo *rta_ci;
-	struct rta_session *rta_sess;
-	u32		*rta_mp_alg;
-};
+struct fib_config {
+	u8			fc_family;
+	u8			fc_dst_len;
+	u8			fc_src_len;
+	u8			fc_tos;
+	u8			fc_protocol;
+	u8			fc_scope;
+	u8			fc_type;
+	/* 1 byte unused */
+	u32			fc_table;
+	u32			fc_dst;
+	u32			fc_src;
+	u32			fc_gw;
+	int			fc_oif;
+	u32			fc_flags;
+	u32			fc_priority;
+	u32			fc_prefsrc;
+	struct nlattr		*fc_mx;
+	struct rtnexthop	*fc_mp;
+	int			fc_mx_len;
+	int			fc_mp_len;
+	u32			fc_flow;
+	u32			fc_mp_alg;
+	u32			fc_nlflags;
+	struct nl_info		fc_nlinfo;
+ };
 
 struct fib_info;
 
@@ -154,12 +161,8 @@ struct fib_table {
 	u32		tb_id;
 	unsigned	tb_stamp;
 	int		(*tb_lookup)(struct fib_table *tb, const struct flowi *flp, struct fib_result *res);
-	int		(*tb_insert)(struct fib_table *table, struct rtmsg *r,
-				     struct kern_rta *rta, struct nlmsghdr *n,
-				     struct netlink_skb_parms *req);
-	int		(*tb_delete)(struct fib_table *table, struct rtmsg *r,
-				     struct kern_rta *rta, struct nlmsghdr *n,
-				     struct netlink_skb_parms *req);
+	int		(*tb_insert)(struct fib_table *, struct fib_config *);
+	int		(*tb_delete)(struct fib_table *, struct fib_config *);
 	int		(*tb_dump)(struct fib_table *table, struct sk_buff *skb,
 				     struct netlink_callback *cb);
 	int		(*tb_flush)(struct fib_table *table);
@@ -228,8 +231,6 @@ struct rtentry;
 extern int ip_fib_check_default(u32 gw, struct net_device *dev);
 extern int fib_sync_down(u32 local, struct net_device *dev, int force);
 extern int fib_sync_up(struct net_device *dev);
-extern int fib_convert_rtentry(int cmd, struct nlmsghdr *nl, struct rtmsg *rtm,
-			       struct kern_rta *rta, struct rtentry *r);
 extern u32  __fib_res_prefsrc(struct fib_result *res);
 
 /* Exported by fib_hash.c */
diff --git a/include/net/netlink.h b/include/net/netlink.h
index bf593eb59e1b..47044da167c5 100644
--- a/include/net/netlink.h
+++ b/include/net/netlink.h
@@ -192,6 +192,16 @@ struct nla_policy {
 	u16		minlen;
 };
 
+/**
+ * struct nl_info - netlink source information
+ * @nlh: Netlink message header of original request
+ * @pid: Netlink PID of requesting application
+ */
+struct nl_info {
+	struct nlmsghdr		*nlh;
+	u32			pid;
+};
+
 extern void		netlink_run_queue(struct sock *sk, unsigned int *qlen,
 					  int (*cb)(struct sk_buff *,
 						    struct nlmsghdr *, int *));
diff --git a/include/net/nexthop.h b/include/net/nexthop.h
new file mode 100644
index 000000000000..3334dbfa5aa4
--- /dev/null
+++ b/include/net/nexthop.h
@@ -0,0 +1,33 @@
+#ifndef __NET_NEXTHOP_H
+#define __NET_NEXTHOP_H
+
+#include <linux/rtnetlink.h>
+#include <net/netlink.h>
+
+static inline int rtnh_ok(const struct rtnexthop *rtnh, int remaining)
+{
+	return remaining >= sizeof(*rtnh) &&
+	       rtnh->rtnh_len >= sizeof(*rtnh) &&
+	       rtnh->rtnh_len <= remaining;
+}
+
+static inline struct rtnexthop *rtnh_next(const struct rtnexthop *rtnh,
+                                         int *remaining)
+{
+	int totlen = NLA_ALIGN(rtnh->rtnh_len);
+
+	*remaining -= totlen;
+	return (struct rtnexthop *) ((char *) rtnh + totlen);
+}
+
+static inline struct nlattr *rtnh_attrs(const struct rtnexthop *rtnh)
+{
+	return (struct nlattr *) ((char *) rtnh + NLA_ALIGN(sizeof(*rtnh)));
+}
+
+static inline int rtnh_attrlen(const struct rtnexthop *rtnh)
+{
+	return rtnh->rtnh_len - NLA_ALIGN(sizeof(*rtnh));
+}
+
+#endif
diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c
index ad4c14f968a1..acc18bdf2dee 100644
--- a/net/ipv4/fib_frontend.c
+++ b/net/ipv4/fib_frontend.c
@@ -253,42 +253,190 @@ e_inval:
 
 #ifndef CONFIG_IP_NOSIOCRT
 
+static inline u32 sk_extract_addr(struct sockaddr *addr)
+{
+	return ((struct sockaddr_in *) addr)->sin_addr.s_addr;
+}
+
+static int put_rtax(struct nlattr *mx, int len, int type, u32 value)
+{
+	struct nlattr *nla;
+
+	nla = (struct nlattr *) ((char *) mx + len);
+	nla->nla_type = type;
+	nla->nla_len = nla_attr_size(4);
+	*(u32 *) nla_data(nla) = value;
+
+	return len + nla_total_size(4);
+}
+
+static int rtentry_to_fib_config(int cmd, struct rtentry *rt,
+				 struct fib_config *cfg)
+{
+	u32 addr;
+	int plen;
+
+	memset(cfg, 0, sizeof(*cfg));
+
+	if (rt->rt_dst.sa_family != AF_INET)
+		return -EAFNOSUPPORT;
+
+	/*
+	 * Check mask for validity:
+	 * a) it must be contiguous.
+	 * b) destination must have all host bits clear.
+	 * c) if application forgot to set correct family (AF_INET),
+	 *    reject request unless it is absolutely clear i.e.
+	 *    both family and mask are zero.
+	 */
+	plen = 32;
+	addr = sk_extract_addr(&rt->rt_dst);
+	if (!(rt->rt_flags & RTF_HOST)) {
+		u32 mask = sk_extract_addr(&rt->rt_genmask);
+
+		if (rt->rt_genmask.sa_family != AF_INET) {
+			if (mask || rt->rt_genmask.sa_family)
+				return -EAFNOSUPPORT;
+		}
+
+		if (bad_mask(mask, addr))
+			return -EINVAL;
+
+		plen = inet_mask_len(mask);
+	}
+
+	cfg->fc_dst_len = plen;
+	cfg->fc_dst = addr;
+
+	if (cmd != SIOCDELRT) {
+		cfg->fc_nlflags = NLM_F_CREATE;
+		cfg->fc_protocol = RTPROT_BOOT;
+	}
+
+	if (rt->rt_metric)
+		cfg->fc_priority = rt->rt_metric - 1;
+
+	if (rt->rt_flags & RTF_REJECT) {
+		cfg->fc_scope = RT_SCOPE_HOST;
+		cfg->fc_type = RTN_UNREACHABLE;
+		return 0;
+	}
+
+	cfg->fc_scope = RT_SCOPE_NOWHERE;
+	cfg->fc_type = RTN_UNICAST;
+
+	if (rt->rt_dev) {
+		char *colon;
+		struct net_device *dev;
+		char devname[IFNAMSIZ];
+
+		if (copy_from_user(devname, rt->rt_dev, IFNAMSIZ-1))
+			return -EFAULT;
+
+		devname[IFNAMSIZ-1] = 0;
+		colon = strchr(devname, ':');
+		if (colon)
+			*colon = 0;
+		dev = __dev_get_by_name(devname);
+		if (!dev)
+			return -ENODEV;
+		cfg->fc_oif = dev->ifindex;
+		if (colon) {
+			struct in_ifaddr *ifa;
+			struct in_device *in_dev = __in_dev_get_rtnl(dev);
+			if (!in_dev)
+				return -ENODEV;
+			*colon = ':';
+			for (ifa = in_dev->ifa_list; ifa; ifa = ifa->ifa_next)
+				if (strcmp(ifa->ifa_label, devname) == 0)
+					break;
+			if (ifa == NULL)
+				return -ENODEV;
+			cfg->fc_prefsrc = ifa->ifa_local;
+		}
+	}
+
+	addr = sk_extract_addr(&rt->rt_gateway);
+	if (rt->rt_gateway.sa_family == AF_INET && addr) {
+		cfg->fc_gw = addr;
+		if (rt->rt_flags & RTF_GATEWAY &&
+		    inet_addr_type(addr) == RTN_UNICAST)
+			cfg->fc_scope = RT_SCOPE_UNIVERSE;
+	}
+
+	if (cmd == SIOCDELRT)
+		return 0;
+
+	if (rt->rt_flags & RTF_GATEWAY && !cfg->fc_gw)
+		return -EINVAL;
+
+	if (cfg->fc_scope == RT_SCOPE_NOWHERE)
+		cfg->fc_scope = RT_SCOPE_LINK;
+
+	if (rt->rt_flags & (RTF_MTU | RTF_WINDOW | RTF_IRTT)) {
+		struct nlattr *mx;
+		int len = 0;
+
+		mx = kzalloc(3 * nla_total_size(4), GFP_KERNEL);
+ 		if (mx == NULL)
+			return -ENOMEM;
+
+		if (rt->rt_flags & RTF_MTU)
+			len = put_rtax(mx, len, RTAX_ADVMSS, rt->rt_mtu - 40);
+
+		if (rt->rt_flags & RTF_WINDOW)
+			len = put_rtax(mx, len, RTAX_WINDOW, rt->rt_window);
+
+		if (rt->rt_flags & RTF_IRTT)
+			len = put_rtax(mx, len, RTAX_RTT, rt->rt_irtt << 3);
+
+		cfg->fc_mx = mx;
+		cfg->fc_mx_len = len;
+	}
+
+	return 0;
+}
+
 /*
  *	Handle IP routing ioctl calls. These are used to manipulate the routing tables
  */
  
 int ip_rt_ioctl(unsigned int cmd, void __user *arg)
 {
+	struct fib_config cfg;
+	struct rtentry rt;
 	int err;
-	struct kern_rta rta;
-	struct rtentry  r;
-	struct {
-		struct nlmsghdr nlh;
-		struct rtmsg	rtm;
-	} req;
 
 	switch (cmd) {
 	case SIOCADDRT:		/* Add a route */
 	case SIOCDELRT:		/* Delete a route */
 		if (!capable(CAP_NET_ADMIN))
 			return -EPERM;
-		if (copy_from_user(&r, arg, sizeof(struct rtentry)))
+
+		if (copy_from_user(&rt, arg, sizeof(rt)))
 			return -EFAULT;
+
 		rtnl_lock();
-		err = fib_convert_rtentry(cmd, &req.nlh, &req.rtm, &rta, &r);
+		err = rtentry_to_fib_config(cmd, &rt, &cfg);
 		if (err == 0) {
+			struct fib_table *tb;
+
 			if (cmd == SIOCDELRT) {
-				struct fib_table *tb = fib_get_table(req.rtm.rtm_table);
-				err = -ESRCH;
+				tb = fib_get_table(cfg.fc_table);
 				if (tb)
-					err = tb->tb_delete(tb, &req.rtm, &rta, &req.nlh, NULL);
+					err = tb->tb_delete(tb, &cfg);
+				else
+					err = -ESRCH;
 			} else {
-				struct fib_table *tb = fib_new_table(req.rtm.rtm_table);
-				err = -ENOBUFS;
+				tb = fib_new_table(cfg.fc_table);
 				if (tb)
-					err = tb->tb_insert(tb, &req.rtm, &rta, &req.nlh, NULL);
+					err = tb->tb_insert(tb, &cfg);
+				else
+					err = -ENOBUFS;
 			}
-			kfree(rta.rta_mx);
+
+			/* allocated by rtentry_to_fib_config() */
+			kfree(cfg.fc_mx);
 		}
 		rtnl_unlock();
 		return err;
@@ -305,51 +453,134 @@ int ip_rt_ioctl(unsigned int cmd, void *arg)
 
 #endif
 
-static int inet_check_attr(struct rtmsg *r, struct rtattr **rta)
+static struct nla_policy rtm_ipv4_policy[RTA_MAX+1] __read_mostly = {
+	[RTA_DST]		= { .type = NLA_U32 },
+	[RTA_SRC]		= { .type = NLA_U32 },
+	[RTA_IIF]		= { .type = NLA_U32 },
+	[RTA_OIF]		= { .type = NLA_U32 },
+	[RTA_GATEWAY]		= { .type = NLA_U32 },
+	[RTA_PRIORITY]		= { .type = NLA_U32 },
+	[RTA_PREFSRC]		= { .type = NLA_U32 },
+	[RTA_METRICS]		= { .type = NLA_NESTED },
+	[RTA_MULTIPATH]		= { .minlen = sizeof(struct rtnexthop) },
+	[RTA_PROTOINFO]		= { .type = NLA_U32 },
+	[RTA_FLOW]		= { .type = NLA_U32 },
+	[RTA_MP_ALGO]		= { .type = NLA_U32 },
+};
+
+static int rtm_to_fib_config(struct sk_buff *skb, struct nlmsghdr *nlh,
+			     struct fib_config *cfg)
 {
-	int i;
-
-	for (i=1; i<=RTA_MAX; i++, rta++) {
-		struct rtattr *attr = *rta;
-		if (attr) {
-			if (RTA_PAYLOAD(attr) < 4)
-				return -EINVAL;
-			if (i != RTA_MULTIPATH && i != RTA_METRICS &&
-			    i != RTA_TABLE)
-				*rta = (struct rtattr*)RTA_DATA(attr);
+	struct nlattr *attr;
+	int err, remaining;
+	struct rtmsg *rtm;
+
+	err = nlmsg_validate(nlh, sizeof(*rtm), RTA_MAX, rtm_ipv4_policy);
+	if (err < 0)
+		goto errout;
+
+	memset(cfg, 0, sizeof(*cfg));
+
+	rtm = nlmsg_data(nlh);
+	cfg->fc_family = rtm->rtm_family;
+	cfg->fc_dst_len = rtm->rtm_dst_len;
+	cfg->fc_src_len = rtm->rtm_src_len;
+	cfg->fc_tos = rtm->rtm_tos;
+	cfg->fc_table = rtm->rtm_table;
+	cfg->fc_protocol = rtm->rtm_protocol;
+	cfg->fc_scope = rtm->rtm_scope;
+	cfg->fc_type = rtm->rtm_type;
+	cfg->fc_flags = rtm->rtm_flags;
+	cfg->fc_nlflags = nlh->nlmsg_flags;
+
+	cfg->fc_nlinfo.pid = NETLINK_CB(skb).pid;
+	cfg->fc_nlinfo.nlh = nlh;
+
+	nlmsg_for_each_attr(attr, nlh, sizeof(struct rtmsg), remaining) {
+		switch (attr->nla_type) {
+		case RTA_DST:
+			cfg->fc_dst = nla_get_u32(attr);
+			break;
+		case RTA_SRC:
+			cfg->fc_src = nla_get_u32(attr);
+			break;
+		case RTA_OIF:
+			cfg->fc_oif = nla_get_u32(attr);
+			break;
+		case RTA_GATEWAY:
+			cfg->fc_gw = nla_get_u32(attr);
+			break;
+		case RTA_PRIORITY:
+			cfg->fc_priority = nla_get_u32(attr);
+			break;
+		case RTA_PREFSRC:
+			cfg->fc_prefsrc = nla_get_u32(attr);
+			break;
+		case RTA_METRICS:
+			cfg->fc_mx = nla_data(attr);
+			cfg->fc_mx_len = nla_len(attr);
+			break;
+		case RTA_MULTIPATH:
+			cfg->fc_mp = nla_data(attr);
+			cfg->fc_mp_len = nla_len(attr);
+			break;
+		case RTA_FLOW:
+			cfg->fc_flow = nla_get_u32(attr);
+			break;
+		case RTA_MP_ALGO:
+			cfg->fc_mp_alg = nla_get_u32(attr);
+			break;
+		case RTA_TABLE:
+			cfg->fc_table = nla_get_u32(attr);
+			break;
 		}
 	}
+
 	return 0;
+errout:
+	return err;
 }
 
 int inet_rtm_delroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 {
-	struct fib_table * tb;
-	struct rtattr **rta = arg;
-	struct rtmsg *r = NLMSG_DATA(nlh);
+	struct fib_config cfg;
+	struct fib_table *tb;
+	int err;
 
-	if (inet_check_attr(r, rta))
-		return -EINVAL;
+	err = rtm_to_fib_config(skb, nlh, &cfg);
+	if (err < 0)
+		goto errout;
 
-	tb = fib_get_table(rtm_get_table(rta, r->rtm_table));
-	if (tb)
-		return tb->tb_delete(tb, r, (struct kern_rta*)rta, nlh, &NETLINK_CB(skb));
-	return -ESRCH;
+	tb = fib_get_table(cfg.fc_table);
+	if (tb == NULL) {
+		err = -ESRCH;
+		goto errout;
+	}
+
+	err = tb->tb_delete(tb, &cfg);
+errout:
+	return err;
 }
 
 int inet_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 {
-	struct fib_table * tb;
-	struct rtattr **rta = arg;
-	struct rtmsg *r = NLMSG_DATA(nlh);
+	struct fib_config cfg;
+	struct fib_table *tb;
+	int err;
 
-	if (inet_check_attr(r, rta))
-		return -EINVAL;
+	err = rtm_to_fib_config(skb, nlh, &cfg);
+	if (err < 0)
+		goto errout;
 
-	tb = fib_new_table(rtm_get_table(rta, r->rtm_table));
-	if (tb)
-		return tb->tb_insert(tb, r, (struct kern_rta*)rta, nlh, &NETLINK_CB(skb));
-	return -ENOBUFS;
+	tb = fib_new_table(cfg.fc_table);
+	if (tb == NULL) {
+		err = -ENOBUFS;
+		goto errout;
+	}
+
+	err = tb->tb_insert(tb, &cfg);
+errout:
+	return err;
 }
 
 int inet_dump_fib(struct sk_buff *skb, struct netlink_callback *cb)
@@ -396,17 +627,19 @@ out:
    only when netlink is already locked.
  */
 
-static void fib_magic(int cmd, int type, u32 dst, int dst_len, struct in_ifaddr *ifa)
+static void fib_magic(int cmd, int type, u32 dst, int dst_len,
+		      struct in_ifaddr *ifa)
 {
-	struct fib_table * tb;
-	struct {
-		struct nlmsghdr	nlh;
-		struct rtmsg	rtm;
-	} req;
-	struct kern_rta rta;
-
-	memset(&req.rtm, 0, sizeof(req.rtm));
-	memset(&rta, 0, sizeof(rta));
+	struct fib_table *tb;
+	struct fib_config cfg = {
+		.fc_protocol = RTPROT_KERNEL,
+		.fc_type = type,
+		.fc_dst = dst,
+		.fc_dst_len = dst_len,
+		.fc_prefsrc = ifa->ifa_local,
+		.fc_oif = ifa->ifa_dev->dev->ifindex,
+		.fc_nlflags = NLM_F_CREATE | NLM_F_APPEND,
+	};
 
 	if (type == RTN_UNICAST)
 		tb = fib_new_table(RT_TABLE_MAIN);
@@ -416,26 +649,17 @@ static void fib_magic(int cmd, int type, u32 dst, int dst_len, struct in_ifaddr
 	if (tb == NULL)
 		return;
 
-	req.nlh.nlmsg_len = sizeof(req);
-	req.nlh.nlmsg_type = cmd;
-	req.nlh.nlmsg_flags = NLM_F_REQUEST|NLM_F_CREATE|NLM_F_APPEND;
-	req.nlh.nlmsg_pid = 0;
-	req.nlh.nlmsg_seq = 0;
+	cfg.fc_table = tb->tb_id;
 
-	req.rtm.rtm_dst_len = dst_len;
-	req.rtm.rtm_table = tb->tb_id;
-	req.rtm.rtm_protocol = RTPROT_KERNEL;
-	req.rtm.rtm_scope = (type != RTN_LOCAL ? RT_SCOPE_LINK : RT_SCOPE_HOST);
-	req.rtm.rtm_type = type;
-
-	rta.rta_dst = &dst;
-	rta.rta_prefsrc = &ifa->ifa_local;
-	rta.rta_oif = &ifa->ifa_dev->dev->ifindex;
+	if (type != RTN_LOCAL)
+		cfg.fc_scope = RT_SCOPE_LINK;
+	else
+		cfg.fc_scope = RT_SCOPE_HOST;
 
 	if (cmd == RTM_NEWROUTE)
-		tb->tb_insert(tb, &req.rtm, &rta, &req.nlh, NULL);
+		tb->tb_insert(tb, &cfg);
 	else
-		tb->tb_delete(tb, &req.rtm, &rta, &req.nlh, NULL);
+		tb->tb_delete(tb, &cfg);
 }
 
 void fib_add_ifaddr(struct in_ifaddr *ifa)
diff --git a/net/ipv4/fib_hash.c b/net/ipv4/fib_hash.c
index b5bee1a71e5c..357557549ce5 100644
--- a/net/ipv4/fib_hash.c
+++ b/net/ipv4/fib_hash.c
@@ -379,42 +379,39 @@ static struct fib_node *fib_find_node(struct fn_zone *fz, u32 key)
 	return NULL;
 }
 
-static int
-fn_hash_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
-	       struct nlmsghdr *n, struct netlink_skb_parms *req)
+static int fn_hash_insert(struct fib_table *tb, struct fib_config *cfg)
 {
 	struct fn_hash *table = (struct fn_hash *) tb->tb_data;
 	struct fib_node *new_f, *f;
 	struct fib_alias *fa, *new_fa;
 	struct fn_zone *fz;
 	struct fib_info *fi;
-	int z = r->rtm_dst_len;
-	int type = r->rtm_type;
-	u8 tos = r->rtm_tos;
+	u8 tos = cfg->fc_tos;
 	u32 key;
 	int err;
 
-	if (z > 32)
+	if (cfg->fc_dst_len > 32)
 		return -EINVAL;
-	fz = table->fn_zones[z];
-	if (!fz && !(fz = fn_new_zone(table, z)))
+
+	fz = table->fn_zones[cfg->fc_dst_len];
+	if (!fz && !(fz = fn_new_zone(table, cfg->fc_dst_len)))
 		return -ENOBUFS;
 
 	key = 0;
-	if (rta->rta_dst) {
-		u32 dst;
-		memcpy(&dst, rta->rta_dst, 4);
-		if (dst & ~FZ_MASK(fz))
+	if (cfg->fc_dst) {
+		if (cfg->fc_dst & ~FZ_MASK(fz))
 			return -EINVAL;
-		key = fz_key(dst, fz);
+		key = fz_key(cfg->fc_dst, fz);
 	}
 
-	if  ((fi = fib_create_info(r, rta, n, &err)) == NULL)
-		return err;
+	fi = fib_create_info(cfg);
+	if (IS_ERR(fi))
+		return PTR_ERR(fi);
 
 	if (fz->fz_nent > (fz->fz_divisor<<1) &&
 	    fz->fz_divisor < FZ_MAX_DIVISOR &&
-	    (z==32 || (1<<z) > fz->fz_divisor))
+	    (cfg->fc_dst_len == 32 ||
+	     (1 << cfg->fc_dst_len) > fz->fz_divisor))
 		fn_rehash_zone(fz);
 
 	f = fib_find_node(fz, key);
@@ -440,18 +437,18 @@ fn_hash_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 		struct fib_alias *fa_orig;
 
 		err = -EEXIST;
-		if (n->nlmsg_flags & NLM_F_EXCL)
+		if (cfg->fc_nlflags & NLM_F_EXCL)
 			goto out;
 
-		if (n->nlmsg_flags & NLM_F_REPLACE) {
+		if (cfg->fc_nlflags & NLM_F_REPLACE) {
 			struct fib_info *fi_drop;
 			u8 state;
 
 			write_lock_bh(&fib_hash_lock);
 			fi_drop = fa->fa_info;
 			fa->fa_info = fi;
-			fa->fa_type = type;
-			fa->fa_scope = r->rtm_scope;
+			fa->fa_type = cfg->fc_type;
+			fa->fa_scope = cfg->fc_scope;
 			state = fa->fa_state;
 			fa->fa_state &= ~FA_S_ACCESSED;
 			fib_hash_genid++;
@@ -474,17 +471,17 @@ fn_hash_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 				break;
 			if (fa->fa_info->fib_priority != fi->fib_priority)
 				break;
-			if (fa->fa_type == type &&
-			    fa->fa_scope == r->rtm_scope &&
+			if (fa->fa_type == cfg->fc_type &&
+			    fa->fa_scope == cfg->fc_scope &&
 			    fa->fa_info == fi)
 				goto out;
 		}
-		if (!(n->nlmsg_flags & NLM_F_APPEND))
+		if (!(cfg->fc_nlflags & NLM_F_APPEND))
 			fa = fa_orig;
 	}
 
 	err = -ENOENT;
-	if (!(n->nlmsg_flags&NLM_F_CREATE))
+	if (!(cfg->fc_nlflags & NLM_F_CREATE))
 		goto out;
 
 	err = -ENOBUFS;
@@ -506,8 +503,8 @@ fn_hash_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 
 	new_fa->fa_info = fi;
 	new_fa->fa_tos = tos;
-	new_fa->fa_type = type;
-	new_fa->fa_scope = r->rtm_scope;
+	new_fa->fa_type = cfg->fc_type;
+	new_fa->fa_scope = cfg->fc_scope;
 	new_fa->fa_state = 0;
 
 	/*
@@ -526,7 +523,8 @@ fn_hash_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 		fz->fz_nent++;
 	rt_cache_flush(-1);
 
-	rtmsg_fib(RTM_NEWROUTE, key, new_fa, z, tb->tb_id, n, req);
+	rtmsg_fib(RTM_NEWROUTE, key, new_fa, cfg->fc_dst_len, tb->tb_id,
+		  &cfg->fc_nlinfo);
 	return 0;
 
 out_free_new_fa:
@@ -537,30 +535,25 @@ out:
 }
 
 
-static int
-fn_hash_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
-	       struct nlmsghdr *n, struct netlink_skb_parms *req)
+static int fn_hash_delete(struct fib_table *tb, struct fib_config *cfg)
 {
 	struct fn_hash *table = (struct fn_hash*)tb->tb_data;
 	struct fib_node *f;
 	struct fib_alias *fa, *fa_to_delete;
-	int z = r->rtm_dst_len;
 	struct fn_zone *fz;
 	u32 key;
-	u8 tos = r->rtm_tos;
 
-	if (z > 32)
+	if (cfg->fc_dst_len > 32)
 		return -EINVAL;
-	if ((fz  = table->fn_zones[z]) == NULL)
+
+	if ((fz  = table->fn_zones[cfg->fc_dst_len]) == NULL)
 		return -ESRCH;
 
 	key = 0;
-	if (rta->rta_dst) {
-		u32 dst;
-		memcpy(&dst, rta->rta_dst, 4);
-		if (dst & ~FZ_MASK(fz))
+	if (cfg->fc_dst) {
+		if (cfg->fc_dst & ~FZ_MASK(fz))
 			return -EINVAL;
-		key = fz_key(dst, fz);
+		key = fz_key(cfg->fc_dst, fz);
 	}
 
 	f = fib_find_node(fz, key);
@@ -568,7 +561,7 @@ fn_hash_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 	if (!f)
 		fa = NULL;
 	else
-		fa = fib_find_alias(&f->fn_alias, tos, 0);
+		fa = fib_find_alias(&f->fn_alias, cfg->fc_tos, 0);
 	if (!fa)
 		return -ESRCH;
 
@@ -577,16 +570,16 @@ fn_hash_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 	list_for_each_entry_continue(fa, &f->fn_alias, fa_list) {
 		struct fib_info *fi = fa->fa_info;
 
-		if (fa->fa_tos != tos)
+		if (fa->fa_tos != cfg->fc_tos)
 			break;
 
-		if ((!r->rtm_type ||
-		     fa->fa_type == r->rtm_type) &&
-		    (r->rtm_scope == RT_SCOPE_NOWHERE ||
-		     fa->fa_scope == r->rtm_scope) &&
-		    (!r->rtm_protocol ||
-		     fi->fib_protocol == r->rtm_protocol) &&
-		    fib_nh_match(r, n, rta, fi) == 0) {
+		if ((!cfg->fc_type ||
+		     fa->fa_type == cfg->fc_type) &&
+		    (cfg->fc_scope == RT_SCOPE_NOWHERE ||
+		     fa->fa_scope == cfg->fc_scope) &&
+		    (!cfg->fc_protocol ||
+		     fi->fib_protocol == cfg->fc_protocol) &&
+		    fib_nh_match(cfg, fi) == 0) {
 			fa_to_delete = fa;
 			break;
 		}
@@ -596,7 +589,8 @@ fn_hash_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 		int kill_fn;
 
 		fa = fa_to_delete;
-		rtmsg_fib(RTM_DELROUTE, key, fa, z, tb->tb_id, n, req);
+		rtmsg_fib(RTM_DELROUTE, key, fa, cfg->fc_dst_len,
+			  tb->tb_id, &cfg->fc_nlinfo);
 
 		kill_fn = 0;
 		write_lock_bh(&fib_hash_lock);
diff --git a/net/ipv4/fib_lookup.h b/net/ipv4/fib_lookup.h
index ddd52496b451..d6d1a89e4003 100644
--- a/net/ipv4/fib_lookup.h
+++ b/net/ipv4/fib_lookup.h
@@ -23,19 +23,14 @@ extern int fib_semantic_match(struct list_head *head,
 			      struct fib_result *res, __u32 zone, __u32 mask,
 				int prefixlen);
 extern void fib_release_info(struct fib_info *);
-extern struct fib_info *fib_create_info(const struct rtmsg *r,
-					struct kern_rta *rta,
-					const struct nlmsghdr *,
-					int *err);
-extern int fib_nh_match(struct rtmsg *r, struct nlmsghdr *,
-			struct kern_rta *rta, struct fib_info *fi);
+extern struct fib_info *fib_create_info(struct fib_config *cfg);
+extern int fib_nh_match(struct fib_config *cfg, struct fib_info *fi);
 extern int fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event,
 			 u32 tb_id, u8 type, u8 scope, void *dst,
 			 int dst_len, u8 tos, struct fib_info *fi,
 			 unsigned int);
 extern void rtmsg_fib(int event, u32 key, struct fib_alias *fa,
-		      int z, u32 tb_id,
-		      struct nlmsghdr *n, struct netlink_skb_parms *req);
+		      int dst_len, u32 tb_id, struct nl_info *info);
 extern struct fib_alias *fib_find_alias(struct list_head *fah,
 					u8 tos, u32 prio);
 extern int fib_detect_death(struct fib_info *fi, int order,
diff --git a/net/ipv4/fib_semantics.c b/net/ipv4/fib_semantics.c
index 5dfdad5cbcd4..340f9db389e5 100644
--- a/net/ipv4/fib_semantics.c
+++ b/net/ipv4/fib_semantics.c
@@ -44,6 +44,7 @@
 #include <net/ip_fib.h>
 #include <net/ip_mp_alg.h>
 #include <net/netlink.h>
+#include <net/nexthop.h>
 
 #include "fib_lookup.h"
 
@@ -273,27 +274,27 @@ int ip_fib_check_default(u32 gw, struct net_device *dev)
 }
 
 void rtmsg_fib(int event, u32 key, struct fib_alias *fa,
-	       int z, u32 tb_id,
-	       struct nlmsghdr *n, struct netlink_skb_parms *req)
+	       int dst_len, u32 tb_id, struct nl_info *info)
 {
 	struct sk_buff *skb;
-	u32 pid = req ? req->pid : n->nlmsg_pid;
 	int payload = sizeof(struct rtmsg) + 256;
+	u32 seq = info->nlh ? info->nlh->nlmsg_seq : 0;
 	int err = -ENOBUFS;
 
 	skb = nlmsg_new(nlmsg_total_size(payload), GFP_KERNEL);
 	if (skb == NULL)
 		goto errout;
 
-	err = fib_dump_info(skb, pid, n->nlmsg_seq, event, tb_id,
-			    fa->fa_type, fa->fa_scope, &key, z, fa->fa_tos,
-			    fa->fa_info, 0);
+	err = fib_dump_info(skb, info->pid, seq, event, tb_id,
+			    fa->fa_type, fa->fa_scope, &key, dst_len,
+			    fa->fa_tos, fa->fa_info, 0);
 	if (err < 0) {
 		kfree_skb(skb);
 		goto errout;
 	}
 
-	err = rtnl_notify(skb, pid, RTNLGRP_IPV4_ROUTE, n, GFP_KERNEL);
+	err = rtnl_notify(skb, info->pid, RTNLGRP_IPV4_ROUTE,
+			  info->nlh, GFP_KERNEL);
 errout:
 	if (err < 0)
 		rtnl_set_sk_err(RTNLGRP_IPV4_ROUTE, err);
@@ -342,102 +343,100 @@ int fib_detect_death(struct fib_info *fi, int order,
 
 #ifdef CONFIG_IP_ROUTE_MULTIPATH
 
-static u32 fib_get_attr32(struct rtattr *attr, int attrlen, int type)
-{
-	while (RTA_OK(attr,attrlen)) {
-		if (attr->rta_type == type)
-			return *(u32*)RTA_DATA(attr);
-		attr = RTA_NEXT(attr, attrlen);
-	}
-	return 0;
-}
-
-static int
-fib_count_nexthops(struct rtattr *rta)
+static int fib_count_nexthops(struct rtnexthop *rtnh, int remaining)
 {
 	int nhs = 0;
-	struct rtnexthop *nhp = RTA_DATA(rta);
-	int nhlen = RTA_PAYLOAD(rta);
 
-	while (nhlen >= (int)sizeof(struct rtnexthop)) {
-		if ((nhlen -= nhp->rtnh_len) < 0)
-			return 0;
+	while (rtnh_ok(rtnh, remaining)) {
 		nhs++;
-		nhp = RTNH_NEXT(nhp);
-	};
-	return nhs;
+		rtnh = rtnh_next(rtnh, &remaining);
+	}
+
+	/* leftover implies invalid nexthop configuration, discard it */
+	return remaining > 0 ? 0 : nhs;
 }
 
-static int
-fib_get_nhs(struct fib_info *fi, const struct rtattr *rta, const struct rtmsg *r)
+static int fib_get_nhs(struct fib_info *fi, struct rtnexthop *rtnh,
+		       int remaining, struct fib_config *cfg)
 {
-	struct rtnexthop *nhp = RTA_DATA(rta);
-	int nhlen = RTA_PAYLOAD(rta);
-
 	change_nexthops(fi) {
-		int attrlen = nhlen - sizeof(struct rtnexthop);
-		if (attrlen < 0 || (nhlen -= nhp->rtnh_len) < 0)
+		int attrlen;
+
+		if (!rtnh_ok(rtnh, remaining))
 			return -EINVAL;
-		nh->nh_flags = (r->rtm_flags&~0xFF) | nhp->rtnh_flags;
-		nh->nh_oif = nhp->rtnh_ifindex;
-		nh->nh_weight = nhp->rtnh_hops + 1;
-		if (attrlen) {
-			nh->nh_gw = fib_get_attr32(RTNH_DATA(nhp), attrlen, RTA_GATEWAY);
+
+		nh->nh_flags = (cfg->fc_flags & ~0xFF) | rtnh->rtnh_flags;
+		nh->nh_oif = rtnh->rtnh_ifindex;
+		nh->nh_weight = rtnh->rtnh_hops + 1;
+
+		attrlen = rtnh_attrlen(rtnh);
+		if (attrlen > 0) {
+			struct nlattr *nla, *attrs = rtnh_attrs(rtnh);
+
+			nla = nla_find(attrs, attrlen, RTA_GATEWAY);
+			nh->nh_gw = nla ? nla_get_u32(nla) : 0;
 #ifdef CONFIG_NET_CLS_ROUTE
-			nh->nh_tclassid = fib_get_attr32(RTNH_DATA(nhp), attrlen, RTA_FLOW);
+			nla = nla_find(attrs, attrlen, RTA_FLOW);
+			nh->nh_tclassid = nla ? nla_get_u32(nla) : 0;
 #endif
 		}
-		nhp = RTNH_NEXT(nhp);
+
+		rtnh = rtnh_next(rtnh, &remaining);
 	} endfor_nexthops(fi);
+
 	return 0;
 }
 
 #endif
 
-int fib_nh_match(struct rtmsg *r, struct nlmsghdr *nlh, struct kern_rta *rta,
-		 struct fib_info *fi)
+int fib_nh_match(struct fib_config *cfg, struct fib_info *fi)
 {
 #ifdef CONFIG_IP_ROUTE_MULTIPATH
-	struct rtnexthop *nhp;
-	int nhlen;
+	struct rtnexthop *rtnh;
+	int remaining;
 #endif
 
-	if (rta->rta_priority &&
-	    *rta->rta_priority != fi->fib_priority)
+	if (cfg->fc_priority && cfg->fc_priority != fi->fib_priority)
 		return 1;
 
-	if (rta->rta_oif || rta->rta_gw) {
-		if ((!rta->rta_oif || *rta->rta_oif == fi->fib_nh->nh_oif) &&
-		    (!rta->rta_gw  || memcmp(rta->rta_gw, &fi->fib_nh->nh_gw, 4) == 0))
+	if (cfg->fc_oif || cfg->fc_gw) {
+		if ((!cfg->fc_oif || cfg->fc_oif == fi->fib_nh->nh_oif) &&
+		    (!cfg->fc_gw  || cfg->fc_gw == fi->fib_nh->nh_gw))
 			return 0;
 		return 1;
 	}
 
 #ifdef CONFIG_IP_ROUTE_MULTIPATH
-	if (rta->rta_mp == NULL)
+	if (cfg->fc_mp == NULL)
 		return 0;
-	nhp = RTA_DATA(rta->rta_mp);
-	nhlen = RTA_PAYLOAD(rta->rta_mp);
+
+	rtnh = cfg->fc_mp;
+	remaining = cfg->fc_mp_len;
 	
 	for_nexthops(fi) {
-		int attrlen = nhlen - sizeof(struct rtnexthop);
-		u32 gw;
+		int attrlen;
 
-		if (attrlen < 0 || (nhlen -= nhp->rtnh_len) < 0)
+		if (!rtnh_ok(rtnh, remaining))
 			return -EINVAL;
-		if (nhp->rtnh_ifindex && nhp->rtnh_ifindex != nh->nh_oif)
+
+		if (rtnh->rtnh_ifindex && rtnh->rtnh_ifindex != nh->nh_oif)
 			return 1;
-		if (attrlen) {
-			gw = fib_get_attr32(RTNH_DATA(nhp), attrlen, RTA_GATEWAY);
-			if (gw && gw != nh->nh_gw)
+
+		attrlen = rtnh_attrlen(rtnh);
+		if (attrlen < 0) {
+			struct nlattr *nla, *attrs = rtnh_attrs(rtnh);
+
+			nla = nla_find(attrs, attrlen, RTA_GATEWAY);
+			if (nla && nla_get_u32(nla) != nh->nh_gw)
 				return 1;
 #ifdef CONFIG_NET_CLS_ROUTE
-			gw = fib_get_attr32(RTNH_DATA(nhp), attrlen, RTA_FLOW);
-			if (gw && gw != nh->nh_tclassid)
+			nla = nla_find(attrs, attrlen, RTA_FLOW);
+			if (nla && nla_get_u32(nla) != nh->nh_tclassid)
 				return 1;
 #endif
 		}
-		nhp = RTNH_NEXT(nhp);
+
+		rtnh = rtnh_next(rtnh, &remaining);
 	} endfor_nexthops(fi);
 #endif
 	return 0;
@@ -488,7 +487,8 @@ int fib_nh_match(struct rtmsg *r, struct nlmsghdr *nlh, struct kern_rta *rta,
 						|-> {local prefix} (terminal node)
  */
 
-static int fib_check_nh(const struct rtmsg *r, struct fib_info *fi, struct fib_nh *nh)
+static int fib_check_nh(struct fib_config *cfg, struct fib_info *fi,
+			struct fib_nh *nh)
 {
 	int err;
 
@@ -502,7 +502,7 @@ static int fib_check_nh(const struct rtmsg *r, struct fib_info *fi, struct fib_n
 		if (nh->nh_flags&RTNH_F_ONLINK) {
 			struct net_device *dev;
 
-			if (r->rtm_scope >= RT_SCOPE_LINK)
+			if (cfg->fc_scope >= RT_SCOPE_LINK)
 				return -EINVAL;
 			if (inet_addr_type(nh->nh_gw) != RTN_UNICAST)
 				return -EINVAL;
@@ -516,10 +516,15 @@ static int fib_check_nh(const struct rtmsg *r, struct fib_info *fi, struct fib_n
 			return 0;
 		}
 		{
-			struct flowi fl = { .nl_u = { .ip4_u =
-						      { .daddr = nh->nh_gw,
-							.scope = r->rtm_scope + 1 } },
-					    .oif = nh->nh_oif };
+			struct flowi fl = {
+				.nl_u = {
+					.ip4_u = {
+						.daddr = nh->nh_gw,
+						.scope = cfg->fc_scope + 1,
+					},
+				},
+				.oif = nh->nh_oif,
+			};
 
 			/* It is not necessary, but requires a bit of thinking */
 			if (fl.fl4_scope < RT_SCOPE_LINK)
@@ -646,39 +651,28 @@ static void fib_hash_move(struct hlist_head *new_info_hash,
 	fib_hash_free(old_laddrhash, bytes);
 }
 
-struct fib_info *
-fib_create_info(const struct rtmsg *r, struct kern_rta *rta,
-		const struct nlmsghdr *nlh, int *errp)
+struct fib_info *fib_create_info(struct fib_config *cfg)
 {
 	int err;
 	struct fib_info *fi = NULL;
 	struct fib_info *ofi;
-#ifdef CONFIG_IP_ROUTE_MULTIPATH
 	int nhs = 1;
-#else
-	const int nhs = 1;
-#endif
-#ifdef CONFIG_IP_ROUTE_MULTIPATH_CACHED
-	u32 mp_alg = IP_MP_ALG_NONE;
-#endif
 
 	/* Fast check to catch the most weird cases */
-	if (fib_props[r->rtm_type].scope > r->rtm_scope)
+	if (fib_props[cfg->fc_type].scope > cfg->fc_scope)
 		goto err_inval;
 
 #ifdef CONFIG_IP_ROUTE_MULTIPATH
-	if (rta->rta_mp) {
-		nhs = fib_count_nexthops(rta->rta_mp);
+	if (cfg->fc_mp) {
+		nhs = fib_count_nexthops(cfg->fc_mp, cfg->fc_mp_len);
 		if (nhs == 0)
 			goto err_inval;
 	}
 #endif
 #ifdef CONFIG_IP_ROUTE_MULTIPATH_CACHED
-	if (rta->rta_mp_alg) {
-		mp_alg = *rta->rta_mp_alg;
-
-		if (mp_alg < IP_MP_ALG_NONE ||
-		    mp_alg > IP_MP_ALG_MAX)
+	if (cfg->fc_mp_alg) {
+		if (cfg->fc_mp_alg < IP_MP_ALG_NONE ||
+		    cfg->fc_mp_alg > IP_MP_ALG_MAX)
 			goto err_inval;
 	}
 #endif
@@ -714,43 +708,42 @@ fib_create_info(const struct rtmsg *r, struct kern_rta *rta,
 		goto failure;
 	fib_info_cnt++;
 
-	fi->fib_protocol = r->rtm_protocol;
+	fi->fib_protocol = cfg->fc_protocol;
+	fi->fib_flags = cfg->fc_flags;
+	fi->fib_priority = cfg->fc_priority;
+	fi->fib_prefsrc = cfg->fc_prefsrc;
 
 	fi->fib_nhs = nhs;
 	change_nexthops(fi) {
 		nh->nh_parent = fi;
 	} endfor_nexthops(fi)
 
-	fi->fib_flags = r->rtm_flags;
-	if (rta->rta_priority)
-		fi->fib_priority = *rta->rta_priority;
-	if (rta->rta_mx) {
-		int attrlen = RTA_PAYLOAD(rta->rta_mx);
-		struct rtattr *attr = RTA_DATA(rta->rta_mx);
-
-		while (RTA_OK(attr, attrlen)) {
-			unsigned flavor = attr->rta_type;
-			if (flavor) {
-				if (flavor > RTAX_MAX)
+	if (cfg->fc_mx) {
+		struct nlattr *nla;
+		int remaining;
+
+		nla_for_each_attr(nla, cfg->fc_mx, cfg->fc_mx_len, remaining) {
+			int type = nla->nla_type;
+
+			if (type) {
+				if (type > RTAX_MAX)
 					goto err_inval;
-				fi->fib_metrics[flavor-1] = *(unsigned*)RTA_DATA(attr);
+				fi->fib_metrics[type - 1] = nla_get_u32(nla);
 			}
-			attr = RTA_NEXT(attr, attrlen);
 		}
 	}
-	if (rta->rta_prefsrc)
-		memcpy(&fi->fib_prefsrc, rta->rta_prefsrc, 4);
 
-	if (rta->rta_mp) {
+	if (cfg->fc_mp) {
 #ifdef CONFIG_IP_ROUTE_MULTIPATH
-		if ((err = fib_get_nhs(fi, rta->rta_mp, r)) != 0)
+		err = fib_get_nhs(fi, cfg->fc_mp, cfg->fc_mp_len, cfg);
+		if (err != 0)
 			goto failure;
-		if (rta->rta_oif && fi->fib_nh->nh_oif != *rta->rta_oif)
+		if (cfg->fc_oif && fi->fib_nh->nh_oif != cfg->fc_oif)
 			goto err_inval;
-		if (rta->rta_gw && memcmp(&fi->fib_nh->nh_gw, rta->rta_gw, 4))
+		if (cfg->fc_gw && fi->fib_nh->nh_gw != cfg->fc_gw)
 			goto err_inval;
 #ifdef CONFIG_NET_CLS_ROUTE
-		if (rta->rta_flow && memcmp(&fi->fib_nh->nh_tclassid, rta->rta_flow, 4))
+		if (cfg->fc_flow && fi->fib_nh->nh_tclassid != cfg->fc_flow)
 			goto err_inval;
 #endif
 #else
@@ -758,34 +751,32 @@ fib_create_info(const struct rtmsg *r, struct kern_rta *rta,
 #endif
 	} else {
 		struct fib_nh *nh = fi->fib_nh;
-		if (rta->rta_oif)
-			nh->nh_oif = *rta->rta_oif;
-		if (rta->rta_gw)
-			memcpy(&nh->nh_gw, rta->rta_gw, 4);
+
+		nh->nh_oif = cfg->fc_oif;
+		nh->nh_gw = cfg->fc_gw;
+		nh->nh_flags = cfg->fc_flags;
 #ifdef CONFIG_NET_CLS_ROUTE
-		if (rta->rta_flow)
-			memcpy(&nh->nh_tclassid, rta->rta_flow, 4);
+		nh->nh_tclassid = cfg->fc_flow;
 #endif
-		nh->nh_flags = r->rtm_flags;
 #ifdef CONFIG_IP_ROUTE_MULTIPATH
 		nh->nh_weight = 1;
 #endif
 	}
 
 #ifdef CONFIG_IP_ROUTE_MULTIPATH_CACHED
-	fi->fib_mp_alg = mp_alg;
+	fi->fib_mp_alg = cfg->fc_mp_alg;
 #endif
 
-	if (fib_props[r->rtm_type].error) {
-		if (rta->rta_gw || rta->rta_oif || rta->rta_mp)
+	if (fib_props[cfg->fc_type].error) {
+		if (cfg->fc_gw || cfg->fc_oif || cfg->fc_mp)
 			goto err_inval;
 		goto link_it;
 	}
 
-	if (r->rtm_scope > RT_SCOPE_HOST)
+	if (cfg->fc_scope > RT_SCOPE_HOST)
 		goto err_inval;
 
-	if (r->rtm_scope == RT_SCOPE_HOST) {
+	if (cfg->fc_scope == RT_SCOPE_HOST) {
 		struct fib_nh *nh = fi->fib_nh;
 
 		/* Local address is added. */
@@ -798,14 +789,14 @@ fib_create_info(const struct rtmsg *r, struct kern_rta *rta,
 			goto failure;
 	} else {
 		change_nexthops(fi) {
-			if ((err = fib_check_nh(r, fi, nh)) != 0)
+			if ((err = fib_check_nh(cfg, fi, nh)) != 0)
 				goto failure;
 		} endfor_nexthops(fi)
 	}
 
 	if (fi->fib_prefsrc) {
-		if (r->rtm_type != RTN_LOCAL || rta->rta_dst == NULL ||
-		    memcmp(&fi->fib_prefsrc, rta->rta_dst, 4))
+		if (cfg->fc_type != RTN_LOCAL || !cfg->fc_dst ||
+		    fi->fib_prefsrc != cfg->fc_dst)
 			if (inet_addr_type(fi->fib_prefsrc) != RTN_LOCAL)
 				goto err_inval;
 	}
@@ -846,12 +837,12 @@ err_inval:
 	err = -EINVAL;
 
 failure:
-        *errp = err;
         if (fi) {
 		fi->fib_dead = 1;
 		free_fib_info(fi);
 	}
-	return NULL;
+
+	return ERR_PTR(err);
 }
 
 /* Note! fib_semantic_match intentionally uses  RCU list functions. */
@@ -1012,150 +1003,6 @@ rtattr_failure:
 	return -1;
 }
 
-#ifndef CONFIG_IP_NOSIOCRT
-
-int
-fib_convert_rtentry(int cmd, struct nlmsghdr *nl, struct rtmsg *rtm,
-		    struct kern_rta *rta, struct rtentry *r)
-{
-	int    plen;
-	u32    *ptr;
-
-	memset(rtm, 0, sizeof(*rtm));
-	memset(rta, 0, sizeof(*rta));
-
-	if (r->rt_dst.sa_family != AF_INET)
-		return -EAFNOSUPPORT;
-
-	/* Check mask for validity:
-	   a) it must be contiguous.
-	   b) destination must have all host bits clear.
-	   c) if application forgot to set correct family (AF_INET),
-	      reject request unless it is absolutely clear i.e.
-	      both family and mask are zero.
-	 */
-	plen = 32;
-	ptr = &((struct sockaddr_in*)&r->rt_dst)->sin_addr.s_addr;
-	if (!(r->rt_flags&RTF_HOST)) {
-		u32 mask = ((struct sockaddr_in*)&r->rt_genmask)->sin_addr.s_addr;
-		if (r->rt_genmask.sa_family != AF_INET) {
-			if (mask || r->rt_genmask.sa_family)
-				return -EAFNOSUPPORT;
-		}
-		if (bad_mask(mask, *ptr))
-			return -EINVAL;
-		plen = inet_mask_len(mask);
-	}
-
-	nl->nlmsg_flags = NLM_F_REQUEST;
-	nl->nlmsg_pid = 0;
-	nl->nlmsg_seq = 0;
-	nl->nlmsg_len = NLMSG_LENGTH(sizeof(*rtm));
-	if (cmd == SIOCDELRT) {
-		nl->nlmsg_type = RTM_DELROUTE;
-		nl->nlmsg_flags = 0;
-	} else {
-		nl->nlmsg_type = RTM_NEWROUTE;
-		nl->nlmsg_flags = NLM_F_REQUEST|NLM_F_CREATE;
-		rtm->rtm_protocol = RTPROT_BOOT;
-	}
-
-	rtm->rtm_dst_len = plen;
-	rta->rta_dst = ptr;
-
-	if (r->rt_metric) {
-		*(u32*)&r->rt_pad3 = r->rt_metric - 1;
-		rta->rta_priority = (u32*)&r->rt_pad3;
-	}
-	if (r->rt_flags&RTF_REJECT) {
-		rtm->rtm_scope = RT_SCOPE_HOST;
-		rtm->rtm_type = RTN_UNREACHABLE;
-		return 0;
-	}
-	rtm->rtm_scope = RT_SCOPE_NOWHERE;
-	rtm->rtm_type = RTN_UNICAST;
-
-	if (r->rt_dev) {
-		char *colon;
-		struct net_device *dev;
-		char   devname[IFNAMSIZ];
-
-		if (copy_from_user(devname, r->rt_dev, IFNAMSIZ-1))
-			return -EFAULT;
-		devname[IFNAMSIZ-1] = 0;
-		colon = strchr(devname, ':');
-		if (colon)
-			*colon = 0;
-		dev = __dev_get_by_name(devname);
-		if (!dev)
-			return -ENODEV;
-		rta->rta_oif = &dev->ifindex;
-		if (colon) {
-			struct in_ifaddr *ifa;
-			struct in_device *in_dev = __in_dev_get_rtnl(dev);
-			if (!in_dev)
-				return -ENODEV;
-			*colon = ':';
-			for (ifa = in_dev->ifa_list; ifa; ifa = ifa->ifa_next)
-				if (strcmp(ifa->ifa_label, devname) == 0)
-					break;
-			if (ifa == NULL)
-				return -ENODEV;
-			rta->rta_prefsrc = &ifa->ifa_local;
-		}
-	}
-
-	ptr = &((struct sockaddr_in*)&r->rt_gateway)->sin_addr.s_addr;
-	if (r->rt_gateway.sa_family == AF_INET && *ptr) {
-		rta->rta_gw = ptr;
-		if (r->rt_flags&RTF_GATEWAY && inet_addr_type(*ptr) == RTN_UNICAST)
-			rtm->rtm_scope = RT_SCOPE_UNIVERSE;
-	}
-
-	if (cmd == SIOCDELRT)
-		return 0;
-
-	if (r->rt_flags&RTF_GATEWAY && rta->rta_gw == NULL)
-		return -EINVAL;
-
-	if (rtm->rtm_scope == RT_SCOPE_NOWHERE)
-		rtm->rtm_scope = RT_SCOPE_LINK;
-
-	if (r->rt_flags&(RTF_MTU|RTF_WINDOW|RTF_IRTT)) {
-		struct rtattr *rec;
-		struct rtattr *mx = kmalloc(RTA_LENGTH(3*RTA_LENGTH(4)), GFP_KERNEL);
-		if (mx == NULL)
-			return -ENOMEM;
-		rta->rta_mx = mx;
-		mx->rta_type = RTA_METRICS;
-		mx->rta_len  = RTA_LENGTH(0);
-		if (r->rt_flags&RTF_MTU) {
-			rec = (void*)((char*)mx + RTA_ALIGN(mx->rta_len));
-			rec->rta_type = RTAX_ADVMSS;
-			rec->rta_len = RTA_LENGTH(4);
-			mx->rta_len += RTA_LENGTH(4);
-			*(u32*)RTA_DATA(rec) = r->rt_mtu - 40;
-		}
-		if (r->rt_flags&RTF_WINDOW) {
-			rec = (void*)((char*)mx + RTA_ALIGN(mx->rta_len));
-			rec->rta_type = RTAX_WINDOW;
-			rec->rta_len = RTA_LENGTH(4);
-			mx->rta_len += RTA_LENGTH(4);
-			*(u32*)RTA_DATA(rec) = r->rt_window;
-		}
-		if (r->rt_flags&RTF_IRTT) {
-			rec = (void*)((char*)mx + RTA_ALIGN(mx->rta_len));
-			rec->rta_type = RTAX_RTT;
-			rec->rta_len = RTA_LENGTH(4);
-			mx->rta_len += RTA_LENGTH(4);
-			*(u32*)RTA_DATA(rec) = r->rt_irtt<<3;
-		}
-	}
-	return 0;
-}
-
-#endif
-
 /*
    Update FIB if:
    - local address disappeared -> we must delete all the entries
diff --git a/net/ipv4/fib_trie.c b/net/ipv4/fib_trie.c
index 2a580eb2579b..41bef0a88ab6 100644
--- a/net/ipv4/fib_trie.c
+++ b/net/ipv4/fib_trie.c
@@ -1124,17 +1124,14 @@ err:
 	return fa_head;
 }
 
-static int
-fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
-	       struct nlmsghdr *nlhdr, struct netlink_skb_parms *req)
+static int fn_trie_insert(struct fib_table *tb, struct fib_config *cfg)
 {
 	struct trie *t = (struct trie *) tb->tb_data;
 	struct fib_alias *fa, *new_fa;
 	struct list_head *fa_head = NULL;
 	struct fib_info *fi;
-	int plen = r->rtm_dst_len;
-	int type = r->rtm_type;
-	u8 tos = r->rtm_tos;
+	int plen = cfg->fc_dst_len;
+	u8 tos = cfg->fc_tos;
 	u32 key, mask;
 	int err;
 	struct leaf *l;
@@ -1142,11 +1139,7 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 	if (plen > 32)
 		return -EINVAL;
 
-	key = 0;
-	if (rta->rta_dst)
-		memcpy(&key, rta->rta_dst, 4);
-
-	key = ntohl(key);
+	key = ntohl(cfg->fc_dst);
 
 	pr_debug("Insert table=%u %08x/%d\n", tb->tb_id, key, plen);
 
@@ -1157,10 +1150,11 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 
 	key = key & mask;
 
-	fi = fib_create_info(r, rta, nlhdr, &err);
-
-	if (!fi)
+	fi = fib_create_info(cfg);
+	if (IS_ERR(fi)) {
+		err = PTR_ERR(fi);
 		goto err;
+	}
 
 	l = fib_find_node(t, key);
 	fa = NULL;
@@ -1185,10 +1179,10 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 		struct fib_alias *fa_orig;
 
 		err = -EEXIST;
-		if (nlhdr->nlmsg_flags & NLM_F_EXCL)
+		if (cfg->fc_nlflags & NLM_F_EXCL)
 			goto out;
 
-		if (nlhdr->nlmsg_flags & NLM_F_REPLACE) {
+		if (cfg->fc_nlflags & NLM_F_REPLACE) {
 			struct fib_info *fi_drop;
 			u8 state;
 
@@ -1200,8 +1194,8 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 			fi_drop = fa->fa_info;
 			new_fa->fa_tos = fa->fa_tos;
 			new_fa->fa_info = fi;
-			new_fa->fa_type = type;
-			new_fa->fa_scope = r->rtm_scope;
+			new_fa->fa_type = cfg->fc_type;
+			new_fa->fa_scope = cfg->fc_scope;
 			state = fa->fa_state;
 			new_fa->fa_state &= ~FA_S_ACCESSED;
 
@@ -1224,17 +1218,17 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 				break;
 			if (fa->fa_info->fib_priority != fi->fib_priority)
 				break;
-			if (fa->fa_type == type &&
-			    fa->fa_scope == r->rtm_scope &&
+			if (fa->fa_type == cfg->fc_type &&
+			    fa->fa_scope == cfg->fc_scope &&
 			    fa->fa_info == fi) {
 				goto out;
 			}
 		}
-		if (!(nlhdr->nlmsg_flags & NLM_F_APPEND))
+		if (!(cfg->fc_nlflags & NLM_F_APPEND))
 			fa = fa_orig;
 	}
 	err = -ENOENT;
-	if (!(nlhdr->nlmsg_flags & NLM_F_CREATE))
+	if (!(cfg->fc_nlflags & NLM_F_CREATE))
 		goto out;
 
 	err = -ENOBUFS;
@@ -1244,8 +1238,8 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 
 	new_fa->fa_info = fi;
 	new_fa->fa_tos = tos;
-	new_fa->fa_type = type;
-	new_fa->fa_scope = r->rtm_scope;
+	new_fa->fa_type = cfg->fc_type;
+	new_fa->fa_scope = cfg->fc_scope;
 	new_fa->fa_state = 0;
 	/*
 	 * Insert new entry to the list.
@@ -1262,7 +1256,8 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 			  (fa ? &fa->fa_list : fa_head));
 
 	rt_cache_flush(-1);
-	rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen, tb->tb_id, nlhdr, req);
+	rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen, tb->tb_id,
+		  &cfg->fc_nlinfo);
 succeeded:
 	return 0;
 
@@ -1548,28 +1543,21 @@ static int trie_leaf_remove(struct trie *t, t_key key)
 	return 1;
 }
 
-static int
-fn_trie_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
-		struct nlmsghdr *nlhdr, struct netlink_skb_parms *req)
+static int fn_trie_delete(struct fib_table *tb, struct fib_config *cfg)
 {
 	struct trie *t = (struct trie *) tb->tb_data;
 	u32 key, mask;
-	int plen = r->rtm_dst_len;
-	u8 tos = r->rtm_tos;
+	int plen = cfg->fc_dst_len;
+	u8 tos = cfg->fc_tos;
 	struct fib_alias *fa, *fa_to_delete;
 	struct list_head *fa_head;
 	struct leaf *l;
 	struct leaf_info *li;
 
-
 	if (plen > 32)
 		return -EINVAL;
 
-	key = 0;
-	if (rta->rta_dst)
-		memcpy(&key, rta->rta_dst, 4);
-
-	key = ntohl(key);
+	key = ntohl(cfg->fc_dst);
 	mask = ntohl(inet_make_mask(plen));
 
 	if (key & ~mask)
@@ -1598,13 +1586,12 @@ fn_trie_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 		if (fa->fa_tos != tos)
 			break;
 
-		if ((!r->rtm_type ||
-		     fa->fa_type == r->rtm_type) &&
-		    (r->rtm_scope == RT_SCOPE_NOWHERE ||
-		     fa->fa_scope == r->rtm_scope) &&
-		    (!r->rtm_protocol ||
-		     fi->fib_protocol == r->rtm_protocol) &&
-		    fib_nh_match(r, nlhdr, rta, fi) == 0) {
+		if ((!cfg->fc_type || fa->fa_type == cfg->fc_type) &&
+		    (cfg->fc_scope == RT_SCOPE_NOWHERE ||
+		     fa->fa_scope == cfg->fc_scope) &&
+		    (!cfg->fc_protocol ||
+		     fi->fib_protocol == cfg->fc_protocol) &&
+		    fib_nh_match(cfg, fi) == 0) {
 			fa_to_delete = fa;
 			break;
 		}
@@ -1614,7 +1601,8 @@ fn_trie_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 		return -ESRCH;
 
 	fa = fa_to_delete;
-	rtmsg_fib(RTM_DELROUTE, htonl(key), fa, plen, tb->tb_id, nlhdr, req);
+	rtmsg_fib(RTM_DELROUTE, htonl(key), fa, plen, tb->tb_id,
+		  &cfg->fc_nlinfo);
 
 	l = fib_find_node(t, key);
 	li = find_leaf_info(l, plen);
-- 
cgit v1.2.3


From d889ce3b29e55b91257964b4c9aac70b91fedd91 Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Thu, 17 Aug 2006 18:15:44 -0700
Subject: [IPv4]: Convert route get to new netlink api

Fixes various unvalidated netlink attributes causing memory
corruptions when left empty by userspace applications.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip_fib.h    |  1 +
 net/ipv4/fib_frontend.c |  2 +-
 net/ipv4/route.c        | 84 ++++++++++++++++++++++++++-----------------------
 3 files changed, 47 insertions(+), 40 deletions(-)

(limited to 'include')

diff --git a/include/net/ip_fib.h b/include/net/ip_fib.h
index 42ed96fab3f5..fcc159a4ac17 100644
--- a/include/net/ip_fib.h
+++ b/include/net/ip_fib.h
@@ -216,6 +216,7 @@ extern void fib_select_default(const struct flowi *flp, struct fib_result *res);
 #endif /* CONFIG_IP_MULTIPLE_TABLES */
 
 /* Exported by fib_frontend.c */
+extern struct nla_policy rtm_ipv4_policy[];
 extern void		ip_fib_init(void);
 extern int inet_rtm_delroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg);
 extern int inet_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg);
diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c
index d537c933abe3..d0abeab16e66 100644
--- a/net/ipv4/fib_frontend.c
+++ b/net/ipv4/fib_frontend.c
@@ -453,7 +453,7 @@ int ip_rt_ioctl(unsigned int cmd, void *arg)
 
 #endif
 
-static struct nla_policy rtm_ipv4_policy[RTA_MAX+1] __read_mostly = {
+struct nla_policy rtm_ipv4_policy[RTA_MAX+1] __read_mostly = {
 	[RTA_DST]		= { .type = NLA_U32 },
 	[RTA_SRC]		= { .type = NLA_U32 },
 	[RTA_IIF]		= { .type = NLA_U32 },
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 31b67059ac29..a4d4cb85a16c 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -2737,18 +2737,24 @@ nla_put_failure:
 
 int inet_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void *arg)
 {
-	struct rtattr **rta = arg;
-	struct rtmsg *rtm = NLMSG_DATA(nlh);
+	struct rtmsg *rtm;
+	struct nlattr *tb[RTA_MAX+1];
 	struct rtable *rt = NULL;
-	u32 dst = 0;
-	u32 src = 0;
-	int iif = 0;
-	int err = -ENOBUFS;
+	u32 dst, src, iif;
+	int err;
 	struct sk_buff *skb;
 
+	err = nlmsg_parse(nlh, sizeof(*rtm), tb, RTA_MAX, rtm_ipv4_policy);
+	if (err < 0)
+		goto errout;
+
+	rtm = nlmsg_data(nlh);
+
 	skb = alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL);
-	if (!skb)
-		goto out;
+	if (skb == NULL) {
+		err = -ENOBUFS;
+		goto errout;
+	}
 
 	/* Reserve room for dummy headers, this skb can pass
 	   through good chunk of routing engine.
@@ -2759,61 +2765,61 @@ int inet_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void *arg)
 	skb->nh.iph->protocol = IPPROTO_ICMP;
 	skb_reserve(skb, MAX_HEADER + sizeof(struct iphdr));
 
-	if (rta[RTA_SRC - 1])
-		memcpy(&src, RTA_DATA(rta[RTA_SRC - 1]), 4);
-	if (rta[RTA_DST - 1])
-		memcpy(&dst, RTA_DATA(rta[RTA_DST - 1]), 4);
-	if (rta[RTA_IIF - 1])
-		memcpy(&iif, RTA_DATA(rta[RTA_IIF - 1]), sizeof(int));
+	src = tb[RTA_SRC] ? nla_get_u32(tb[RTA_SRC]) : 0;
+	dst = tb[RTA_DST] ? nla_get_u32(tb[RTA_DST]) : 0;
+	iif = tb[RTA_IIF] ? nla_get_u32(tb[RTA_IIF]) : 0;
 
 	if (iif) {
-		struct net_device *dev = __dev_get_by_index(iif);
-		err = -ENODEV;
-		if (!dev)
-			goto out_free;
+		struct net_device *dev;
+
+		dev = __dev_get_by_index(iif);
+		if (dev == NULL) {
+			err = -ENODEV;
+			goto errout_free;
+		}
+
 		skb->protocol	= htons(ETH_P_IP);
 		skb->dev	= dev;
 		local_bh_disable();
 		err = ip_route_input(skb, dst, src, rtm->rtm_tos, dev);
 		local_bh_enable();
-		rt = (struct rtable*)skb->dst;
-		if (!err && rt->u.dst.error)
+
+		rt = (struct rtable*) skb->dst;
+		if (err == 0 && rt->u.dst.error)
 			err = -rt->u.dst.error;
 	} else {
-		struct flowi fl = { .nl_u = { .ip4_u = { .daddr = dst,
-							 .saddr = src,
-							 .tos = rtm->rtm_tos } } };
-		int oif = 0;
-		if (rta[RTA_OIF - 1])
-			memcpy(&oif, RTA_DATA(rta[RTA_OIF - 1]), sizeof(int));
-		fl.oif = oif;
+		struct flowi fl = {
+			.nl_u = {
+				.ip4_u = {
+					.daddr = dst,
+					.saddr = src,
+					.tos = rtm->rtm_tos,
+				},
+			},
+			.oif = tb[RTA_OIF] ? nla_get_u32(tb[RTA_OIF]) : 0,
+		};
 		err = ip_route_output_key(&rt, &fl);
 	}
+
 	if (err)
-		goto out_free;
+		goto errout_free;
 
 	skb->dst = &rt->u.dst;
 	if (rtm->rtm_flags & RTM_F_NOTIFY)
 		rt->rt_flags |= RTCF_NOTIFY;
 
-	NETLINK_CB(skb).dst_pid = NETLINK_CB(in_skb).pid;
-
 	err = rt_fill_info(skb, NETLINK_CB(in_skb).pid, nlh->nlmsg_seq,
 				RTM_NEWROUTE, 0, 0);
-	if (!err)
-		goto out_free;
-	if (err < 0) {
-		err = -EMSGSIZE;
-		goto out_free;
-	}
+	if (err <= 0)
+		goto errout_free;
 
 	err = rtnl_unicast(skb, NETLINK_CB(in_skb).pid);
-out:
+errout:
 	return err;
 
-out_free:
+errout_free:
 	kfree_skb(skb);
-	goto out;
+	goto errout;
 }
 
 int ip_rt_dump(struct sk_buff *skb,  struct netlink_callback *cb)
-- 
cgit v1.2.3


From e9ce1cd3cf6cf35b21d0ce990f2e738f35907386 Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@sunset.davemloft.net>
Date: Mon, 21 Aug 2006 23:54:55 -0700
Subject: [PKT_SCHED]: Kill pkt_act.h inlining.

This was simply making templates of functions and mostly causing a lot
of code duplication in the classifier action modules.

We solve this more cleanly by having a common "struct tcf_common" that
hash worker functions contained once in act_api.c can work with.

Callers work with real action objects that have the common struct
plus their module specific struct members.  You go from a common
object to the higher level one using a "to_foo()" macro which makes
use of container_of() to do the dirty work.

This also kills off act_generic.h which was only used by act_simple.c
and keeping it around was more work than the it's value.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/act_api.h          | 136 +++++++----
 include/net/act_generic.h      | 142 ------------
 include/net/pkt_act.h          | 273 ----------------------
 include/net/tc_act/tc_defact.h |  13 +-
 include/net/tc_act/tc_gact.h   |  18 +-
 include/net/tc_act/tc_ipt.h    |  15 +-
 include/net/tc_act/tc_mirred.h |  17 +-
 include/net/tc_act/tc_pedit.h  |  15 +-
 net/sched/act_api.c            | 246 ++++++++++++++++++--
 net/sched/act_gact.c           | 142 ++++++------
 net/sched/act_ipt.c            | 175 +++++++-------
 net/sched/act_mirred.c         | 159 ++++++-------
 net/sched/act_pedit.c          | 166 ++++++--------
 net/sched/act_police.c         | 508 +++++++++++++++++++++--------------------
 net/sched/act_simple.c         | 183 ++++++++++++---
 15 files changed, 1061 insertions(+), 1147 deletions(-)
 delete mode 100644 include/net/act_generic.h
 delete mode 100644 include/net/pkt_act.h

(limited to 'include')

diff --git a/include/net/act_api.h b/include/net/act_api.h
index 11e9eaf79f5a..8b06c2f3657f 100644
--- a/include/net/act_api.h
+++ b/include/net/act_api.h
@@ -8,70 +8,110 @@
 #include <net/sch_generic.h>
 #include <net/pkt_sched.h>
 
-#define tca_gen(name) \
-struct tcf_##name *next; \
-	u32 index; \
-	int refcnt; \
-	int bindcnt; \
-	u32 capab; \
-	int action; \
-	struct tcf_t tm; \
-	struct gnet_stats_basic bstats; \
-	struct gnet_stats_queue qstats; \
-	struct gnet_stats_rate_est rate_est; \
-	spinlock_t *stats_lock; \
-	spinlock_t lock
-
-struct tcf_police
-{
-	tca_gen(police);
-	int		result;
-	u32		ewma_rate;
-	u32		burst;
-	u32		mtu;
-	u32		toks;
-	u32		ptoks;
-	psched_time_t	t_c;
-	struct qdisc_rate_table *R_tab;
-	struct qdisc_rate_table *P_tab;
+struct tcf_common {
+	struct tcf_common		*tcfc_next;
+	u32				tcfc_index;
+	int				tcfc_refcnt;
+	int				tcfc_bindcnt;
+	u32				tcfc_capab;
+	int				tcfc_action;
+	struct tcf_t			tcfc_tm;
+	struct gnet_stats_basic		tcfc_bstats;
+	struct gnet_stats_queue		tcfc_qstats;
+	struct gnet_stats_rate_est	tcfc_rate_est;
+	spinlock_t			*tcfc_stats_lock;
+	spinlock_t			tcfc_lock;
+};
+#define tcf_next	common.tcfc_next
+#define tcf_index	common.tcfc_index
+#define tcf_refcnt	common.tcfc_refcnt
+#define tcf_bindcnt	common.tcfc_bindcnt
+#define tcf_capab	common.tcfc_capab
+#define tcf_action	common.tcfc_action
+#define tcf_tm		common.tcfc_tm
+#define tcf_bstats	common.tcfc_bstats
+#define tcf_qstats	common.tcfc_qstats
+#define tcf_rate_est	common.tcfc_rate_est
+#define tcf_stats_lock	common.tcfc_stats_lock
+#define tcf_lock	common.tcfc_lock
+
+struct tcf_police {
+	struct tcf_common	common;
+	int			tcfp_result;
+	u32			tcfp_ewma_rate;
+	u32			tcfp_burst;
+	u32			tcfp_mtu;
+	u32			tcfp_toks;
+	u32			tcfp_ptoks;
+	psched_time_t		tcfp_t_c;
+	struct qdisc_rate_table	*tcfp_R_tab;
+	struct qdisc_rate_table	*tcfp_P_tab;
 };
+#define to_police(pc)	\
+	container_of(pc, struct tcf_police, common)
+
+struct tcf_hashinfo {
+	struct tcf_common	**htab;
+	unsigned int		hmask;
+	rwlock_t		*lock;
+};
+
+static inline unsigned int tcf_hash(u32 index, unsigned int hmask)
+{
+	return index & hmask;
+}
 
 #ifdef CONFIG_NET_CLS_ACT
 
 #define ACT_P_CREATED 1
 #define ACT_P_DELETED 1
 
-struct tcf_act_hdr
-{
-	tca_gen(act_hdr);
+struct tcf_act_hdr {
+	struct tcf_common	common;
 };
 
-struct tc_action
-{
-	void *priv;
-	struct tc_action_ops *ops;
-	__u32   type;   /* for backward compat(TCA_OLD_COMPAT) */
-	__u32   order; 
-	struct tc_action *next;
+struct tc_action {
+	void			*priv;
+	struct tc_action_ops	*ops;
+	__u32			type; /* for backward compat(TCA_OLD_COMPAT) */
+	__u32			order;
+	struct tc_action	*next;
 };
 
 #define TCA_CAP_NONE 0
-struct tc_action_ops
-{
+struct tc_action_ops {
 	struct tc_action_ops *next;
+	struct tcf_hashinfo *hinfo;
 	char    kind[IFNAMSIZ];
 	__u32   type; /* TBD to match kind */
 	__u32 	capab;  /* capabilities includes 4 bit version */
 	struct module		*owner;
 	int     (*act)(struct sk_buff *, struct tc_action *, struct tcf_result *);
 	int     (*get_stats)(struct sk_buff *, struct tc_action *);
-	int     (*dump)(struct sk_buff *, struct tc_action *,int , int);
+	int     (*dump)(struct sk_buff *, struct tc_action *, int, int);
 	int     (*cleanup)(struct tc_action *, int bind);
-	int     (*lookup)(struct tc_action *, u32 );
-	int     (*init)(struct rtattr *,struct rtattr *,struct tc_action *, int , int );
-	int     (*walk)(struct sk_buff *, struct netlink_callback *, int , struct tc_action *);
+	int     (*lookup)(struct tc_action *, u32);
+	int     (*init)(struct rtattr *, struct rtattr *, struct tc_action *, int , int);
+	int     (*walk)(struct sk_buff *, struct netlink_callback *, int, struct tc_action *);
 };
 
+extern struct tcf_common *tcf_hash_lookup(u32 index,
+					  struct tcf_hashinfo *hinfo);
+extern void tcf_hash_destroy(struct tcf_common *p, struct tcf_hashinfo *hinfo);
+extern int tcf_hash_release(struct tcf_common *p, int bind,
+			    struct tcf_hashinfo *hinfo);
+extern int tcf_generic_walker(struct sk_buff *skb, struct netlink_callback *cb,
+			      int type, struct tc_action *a);
+extern u32 tcf_hash_new_index(u32 *idx_gen, struct tcf_hashinfo *hinfo);
+extern int tcf_hash_search(struct tc_action *a, u32 index);
+extern struct tcf_common *tcf_hash_check(u32 index, struct tc_action *a,
+					 int bind, struct tcf_hashinfo *hinfo);
+extern struct tcf_common *tcf_hash_create(u32 index, struct rtattr *est,
+					  struct tc_action *a, int size,
+					  int bind, u32 *idx_gen,
+					  struct tcf_hashinfo *hinfo);
+extern void tcf_hash_insert(struct tcf_common *p, struct tcf_hashinfo *hinfo);
+
 extern int tcf_register_action(struct tc_action_ops *a);
 extern int tcf_unregister_action(struct tc_action_ops *a);
 extern void tcf_action_destroy(struct tc_action *a, int bind);
@@ -96,17 +136,17 @@ tcf_police_release(struct tcf_police *p, int bind)
 	int ret = 0;
 #ifdef CONFIG_NET_CLS_ACT
 	if (p) {
-		if (bind) {
-			 p->bindcnt--;
-		}
-		p->refcnt--;
-		if (p->refcnt <= 0 && !p->bindcnt) {
+		if (bind)
+			p->tcf_bindcnt--;
+
+		p->tcf_refcnt--;
+		if (p->tcf_refcnt <= 0 && !p->tcf_bindcnt) {
 			tcf_police_destroy(p);
 			ret = 1;
 		}
 	}
 #else
-	if (p && --p->refcnt == 0)
+	if (p && --p->tcf_refcnt == 0)
 		tcf_police_destroy(p);
 
 #endif /* CONFIG_NET_CLS_ACT */
diff --git a/include/net/act_generic.h b/include/net/act_generic.h
deleted file mode 100644
index c9daa7e52300..000000000000
--- a/include/net/act_generic.h
+++ /dev/null
@@ -1,142 +0,0 @@
-/*
- * include/net/act_generic.h
- *
-*/
-#ifndef _NET_ACT_GENERIC_H
-#define _NET_ACT_GENERIC_H
-static inline int tcf_defact_release(struct tcf_defact *p, int bind)
-{
-	int ret = 0;
-	if (p) {
-		if (bind) {
-			p->bindcnt--;
-		}
-		p->refcnt--;
-		if (p->bindcnt <= 0 && p->refcnt <= 0) {
-			kfree(p->defdata);
-			tcf_hash_destroy(p);
-			ret = 1;
-		}
-	}
-	return ret;
-}
-
-static inline int
-alloc_defdata(struct tcf_defact *p, u32 datalen, void *defdata)
-{
-	p->defdata = kmalloc(datalen, GFP_KERNEL);
-	if (p->defdata == NULL)
-		return -ENOMEM;
-	p->datalen = datalen;
-	memcpy(p->defdata, defdata, datalen);
-	return 0;
-}
-
-static inline int
-realloc_defdata(struct tcf_defact *p, u32 datalen, void *defdata)
-{
-	/* safer to be just brute force for now */
-	kfree(p->defdata);
-	return alloc_defdata(p, datalen, defdata);
-}
-
-static inline int
-tcf_defact_init(struct rtattr *rta, struct rtattr *est,
-		struct tc_action *a, int ovr, int bind)
-{
-	struct rtattr *tb[TCA_DEF_MAX];
-	struct tc_defact *parm;
-	struct tcf_defact *p;
-	void *defdata;
-	u32 datalen = 0;
-	int ret = 0;
-
-	if (rta == NULL || rtattr_parse_nested(tb, TCA_DEF_MAX, rta) < 0)
-		return -EINVAL;
-
-	if (tb[TCA_DEF_PARMS - 1] == NULL || 
-	    RTA_PAYLOAD(tb[TCA_DEF_PARMS - 1]) < sizeof(*parm))
-		return -EINVAL;
-
-	parm = RTA_DATA(tb[TCA_DEF_PARMS - 1]);
-	defdata = RTA_DATA(tb[TCA_DEF_DATA - 1]);
-	if (defdata == NULL)
-		return -EINVAL;
-
-	datalen = RTA_PAYLOAD(tb[TCA_DEF_DATA - 1]);
-	if (datalen <= 0)
-		return -EINVAL;
-
-	p = tcf_hash_check(parm->index, a, ovr, bind);
-	if (p == NULL) {
-		p = tcf_hash_create(parm->index, est, a, sizeof(*p), ovr, bind);
-		if (p == NULL)
-			return -ENOMEM;
-
-		ret = alloc_defdata(p, datalen, defdata);
-		if (ret < 0) {
-			kfree(p);
-			return ret;
-		}
-		ret = ACT_P_CREATED;
-	} else {
-		if (!ovr) {
-			tcf_defact_release(p, bind);
-			return -EEXIST;
-		}
-		realloc_defdata(p, datalen, defdata);
-	}
-
-	spin_lock_bh(&p->lock);
-	p->action = parm->action;
-	spin_unlock_bh(&p->lock);
-	if (ret == ACT_P_CREATED)
-		tcf_hash_insert(p);
-	return ret;
-}
-
-static inline int tcf_defact_cleanup(struct tc_action *a, int bind)
-{
-	struct tcf_defact *p = PRIV(a, defact);
-
-	if (p != NULL)
-		return tcf_defact_release(p, bind);
-	return 0;
-}
-
-static inline int
-tcf_defact_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref)
-{
-	unsigned char *b = skb->tail;
-	struct tc_defact opt;
-	struct tcf_defact *p = PRIV(a, defact);
-	struct tcf_t t;
-
-	opt.index = p->index;
-	opt.refcnt = p->refcnt - ref;
-	opt.bindcnt = p->bindcnt - bind;
-	opt.action = p->action;
-	RTA_PUT(skb, TCA_DEF_PARMS, sizeof(opt), &opt);
-	RTA_PUT(skb, TCA_DEF_DATA, p->datalen, p->defdata);
-	t.install = jiffies_to_clock_t(jiffies - p->tm.install);
-	t.lastuse = jiffies_to_clock_t(jiffies - p->tm.lastuse);
-	t.expires = jiffies_to_clock_t(p->tm.expires);
-	RTA_PUT(skb, TCA_DEF_TM, sizeof(t), &t);
-	return skb->len;
-
-rtattr_failure:
-	skb_trim(skb, b - skb->data);
-	return -1;
-}
-
-#define tca_use_default_ops \
-	.dump           =       tcf_defact_dump, \
-	.cleanup        =       tcf_defact_cleanup, \
-	.init           =       tcf_defact_init, \
-	.walk           =       tcf_generic_walker, \
-
-#define tca_use_default_defines(name) \
-	static u32 idx_gen; \
-	static struct tcf_defact *tcf_##name_ht[MY_TAB_SIZE]; \
-	static DEFINE_RWLOCK(##name_lock);
-#endif /* _NET_ACT_GENERIC_H */
diff --git a/include/net/pkt_act.h b/include/net/pkt_act.h
deleted file mode 100644
index cf5e4d2e4c21..000000000000
--- a/include/net/pkt_act.h
+++ /dev/null
@@ -1,273 +0,0 @@
-#ifndef __NET_PKT_ACT_H
-#define __NET_PKT_ACT_H
-
-#include <asm/uaccess.h>
-#include <asm/system.h>
-#include <linux/bitops.h>
-#include <linux/types.h>
-#include <linux/kernel.h>
-#include <linux/sched.h>
-#include <linux/string.h>
-#include <linux/mm.h>
-#include <linux/socket.h>
-#include <linux/sockios.h>
-#include <linux/in.h>
-#include <linux/errno.h>
-#include <linux/interrupt.h>
-#include <linux/skbuff.h>
-#include <linux/rtnetlink.h>
-#include <linux/module.h>
-#include <linux/init.h>
-#include <linux/proc_fs.h>
-#include <net/sock.h>
-#include <net/pkt_sched.h>
-
-#define tca_st(val) (struct tcf_##val *)
-#define PRIV(a,name) ( tca_st(name) (a)->priv)
-
-#if 0 /* control */
-#define DPRINTK(format,args...) printk(KERN_DEBUG format,##args)
-#else
-#define DPRINTK(format,args...)
-#endif
-
-#if 0 /* data */
-#define D2PRINTK(format,args...) printk(KERN_DEBUG format,##args)
-#else
-#define D2PRINTK(format,args...)
-#endif
-
-static __inline__ unsigned
-tcf_hash(u32 index)
-{
-	return index & MY_TAB_MASK;
-}
-
-/* probably move this from being inline
- * and put into act_generic
-*/
-static inline void
-tcf_hash_destroy(struct tcf_st *p)
-{
-	unsigned h = tcf_hash(p->index);
-	struct tcf_st **p1p;
-
-	for (p1p = &tcf_ht[h]; *p1p; p1p = &(*p1p)->next) {
-		if (*p1p == p) {
-			write_lock_bh(&tcf_t_lock);
-			*p1p = p->next;
-			write_unlock_bh(&tcf_t_lock);
-#ifdef CONFIG_NET_ESTIMATOR
-			gen_kill_estimator(&p->bstats, &p->rate_est);
-#endif
-			kfree(p);
-			return;
-		}
-	}
-	BUG_TRAP(0);
-}
-
-static inline int
-tcf_hash_release(struct tcf_st *p, int bind )
-{
-	int ret = 0;
-	if (p) {
-		if (bind) {
-			p->bindcnt--;
-		}
-		p->refcnt--;
-	       	if(p->bindcnt <=0 && p->refcnt <= 0) {
-			tcf_hash_destroy(p);
-			ret = 1;
-		}
-	}
-	return ret;
-}
-
-static __inline__ int
-tcf_dump_walker(struct sk_buff *skb, struct netlink_callback *cb,
-		struct tc_action *a)
-{
-	struct tcf_st *p;
-	int err =0, index =  -1,i= 0, s_i = 0, n_i = 0;
-	struct rtattr *r ;
-
-	read_lock(&tcf_t_lock);
-
-	s_i = cb->args[0];
-
-	for (i = 0; i < MY_TAB_SIZE; i++) {
-		p = tcf_ht[tcf_hash(i)];
-
-		for (; p; p = p->next) {
-			index++;
-			if (index < s_i)
-				continue;
-			a->priv = p;
-			a->order = n_i;
-			r = (struct rtattr*) skb->tail;
-			RTA_PUT(skb, a->order, 0, NULL);
-			err = tcf_action_dump_1(skb, a, 0, 0);
-			if (0 > err) {
-				index--;
-				skb_trim(skb, (u8*)r - skb->data);
-				goto done;
-			}
-			r->rta_len = skb->tail - (u8*)r;
-			n_i++;
-			if (n_i >= TCA_ACT_MAX_PRIO) {
-				goto done;
-			}
-		}
-	}
-done:
-	read_unlock(&tcf_t_lock);
-	if (n_i)
-		cb->args[0] += n_i;
-	return n_i;
-
-rtattr_failure:
-	skb_trim(skb, (u8*)r - skb->data);
-	goto done;
-}
-
-static __inline__ int
-tcf_del_walker(struct sk_buff *skb, struct tc_action *a)
-{
-	struct tcf_st *p, *s_p;
-	struct rtattr *r ;
-	int i= 0, n_i = 0;
-
-	r = (struct rtattr*) skb->tail;
-	RTA_PUT(skb, a->order, 0, NULL);
-	RTA_PUT(skb, TCA_KIND, IFNAMSIZ, a->ops->kind);
-	for (i = 0; i < MY_TAB_SIZE; i++) {
-		p = tcf_ht[tcf_hash(i)];
-
-		while (p != NULL) {
-			s_p = p->next;
-			if (ACT_P_DELETED == tcf_hash_release(p, 0)) {
-				 module_put(a->ops->owner);
-			}
-			n_i++;
-			p = s_p;
-		}
-	}
-	RTA_PUT(skb, TCA_FCNT, 4, &n_i);
-	r->rta_len = skb->tail - (u8*)r;
-
-	return n_i;
-rtattr_failure:
-	skb_trim(skb, (u8*)r - skb->data);
-	return -EINVAL;
-}
-
-static __inline__ int
-tcf_generic_walker(struct sk_buff *skb, struct netlink_callback *cb, int type,
-		struct tc_action *a)
-{
-		if (type == RTM_DELACTION) {
-			return tcf_del_walker(skb,a);
-		} else if (type == RTM_GETACTION) {
-			return tcf_dump_walker(skb,cb,a);
-		} else {
-			printk("tcf_generic_walker: unknown action %d\n",type);
-			return -EINVAL;
-		}
-}
-
-static __inline__ struct tcf_st *
-tcf_hash_lookup(u32 index)
-{
-	struct tcf_st *p;
-
-	read_lock(&tcf_t_lock);
-	for (p = tcf_ht[tcf_hash(index)]; p; p = p->next) {
-		if (p->index == index)
-			break;
-	}
-	read_unlock(&tcf_t_lock);
-	return p;
-}
-
-static __inline__ u32
-tcf_hash_new_index(void)
-{
-	do {
-		if (++idx_gen == 0)
-			idx_gen = 1;
-	} while (tcf_hash_lookup(idx_gen));
-
-	return idx_gen;
-}
-
-
-static inline int
-tcf_hash_search(struct tc_action *a, u32 index)
-{
-	struct tcf_st *p = tcf_hash_lookup(index);
-
-	if (p != NULL) {
-		a->priv = p;
-		return 1;
-	}
-	return 0;
-}
-
-#ifdef CONFIG_NET_ACT_INIT
-static inline struct tcf_st *
-tcf_hash_check(u32 index, struct tc_action *a, int ovr, int bind)
-{
-	struct tcf_st *p = NULL;
-	if (index && (p = tcf_hash_lookup(index)) != NULL) {
-		if (bind) {
-			p->bindcnt++;
-			p->refcnt++;
-		}
-		a->priv = p;
-	}
-	return p;
-}
-
-static inline struct tcf_st *
-tcf_hash_create(u32 index, struct rtattr *est, struct tc_action *a, int size, int ovr, int bind)
-{
-	struct tcf_st *p = NULL;
-
-	p = kmalloc(size, GFP_KERNEL);
-	if (p == NULL)
-		return p;
-
-	memset(p, 0, size);
-	p->refcnt = 1;
-
-	if (bind) {
-		p->bindcnt = 1;
-	}
-
-	spin_lock_init(&p->lock);
-	p->stats_lock = &p->lock;
-	p->index = index ? : tcf_hash_new_index();
-	p->tm.install = jiffies;
-	p->tm.lastuse = jiffies;
-#ifdef CONFIG_NET_ESTIMATOR
-	if (est)
-		gen_new_estimator(&p->bstats, &p->rate_est, p->stats_lock, est);
-#endif
-	a->priv = (void *) p;
-	return p;
-}
-
-static inline void tcf_hash_insert(struct tcf_st *p)
-{
-	unsigned h = tcf_hash(p->index);
-
-	write_lock_bh(&tcf_t_lock);
-	p->next = tcf_ht[h];
-	tcf_ht[h] = p;
-	write_unlock_bh(&tcf_t_lock);
-}
-
-#endif
-
-#endif
diff --git a/include/net/tc_act/tc_defact.h b/include/net/tc_act/tc_defact.h
index 463aa671f95d..65f024b80958 100644
--- a/include/net/tc_act/tc_defact.h
+++ b/include/net/tc_act/tc_defact.h
@@ -3,11 +3,12 @@
 
 #include <net/act_api.h>
 
-struct tcf_defact
-{
-	tca_gen(defact);
-	u32     datalen;
-	void    *defdata;
+struct tcf_defact {
+	struct tcf_common	common;
+	u32     		tcfd_datalen;
+	void    		*tcfd_defdata;
 };
+#define to_defact(pc) \
+	container_of(pc, struct tcf_defact, common)
 
-#endif
+#endif /* __NET_TC_DEF_H */
diff --git a/include/net/tc_act/tc_gact.h b/include/net/tc_act/tc_gact.h
index 59f0d9628ad1..9e3f6767b80e 100644
--- a/include/net/tc_act/tc_gact.h
+++ b/include/net/tc_act/tc_gact.h
@@ -3,15 +3,15 @@
 
 #include <net/act_api.h>
 
-struct tcf_gact
-{
-        tca_gen(gact);
+struct tcf_gact {
+	struct tcf_common	common;
 #ifdef CONFIG_GACT_PROB
-        u16                 ptype;
-        u16                 pval;
-        int                 paction;
+        u16			tcfg_ptype;
+        u16			tcfg_pval;
+        int			tcfg_paction;
 #endif
-                                                                                
 };
-                                                                                
-#endif
+#define to_gact(pc) \
+	container_of(pc, struct tcf_gact, common)
+
+#endif /* __NET_TC_GACT_H */
diff --git a/include/net/tc_act/tc_ipt.h b/include/net/tc_act/tc_ipt.h
index cb37ad08427f..f7d25dfcc4b7 100644
--- a/include/net/tc_act/tc_ipt.h
+++ b/include/net/tc_act/tc_ipt.h
@@ -5,12 +5,13 @@
 
 struct xt_entry_target;
 
-struct tcf_ipt
-{
-	tca_gen(ipt);
-	u32 hook;
-	char *tname;
-	struct xt_entry_target *t;
+struct tcf_ipt {
+	struct tcf_common	common;
+	u32			tcfi_hook;
+	char			*tcfi_tname;
+	struct xt_entry_target	*tcfi_t;
 };
+#define to_ipt(pc) \
+	container_of(pc, struct tcf_ipt, common)
 
-#endif
+#endif /* __NET_TC_IPT_H */
diff --git a/include/net/tc_act/tc_mirred.h b/include/net/tc_act/tc_mirred.h
index b5c32f65c12c..ceac661cdfd5 100644
--- a/include/net/tc_act/tc_mirred.h
+++ b/include/net/tc_act/tc_mirred.h
@@ -3,13 +3,14 @@
 
 #include <net/act_api.h>
 
-struct tcf_mirred
-{
-	tca_gen(mirred);
-	int eaction;
-	int ifindex;
-	int ok_push;
-	struct net_device *dev;
+struct tcf_mirred {
+	struct tcf_common	common;
+	int			tcfm_eaction;
+	int			tcfm_ifindex;
+	int			tcfm_ok_push;
+	struct net_device	*tcfm_dev;
 };
+#define to_mirred(pc) \
+	container_of(pc, struct tcf_mirred, common)
 
-#endif
+#endif /* __NET_TC_MIR_H */
diff --git a/include/net/tc_act/tc_pedit.h b/include/net/tc_act/tc_pedit.h
index eb21689d759d..e6f6e15956f5 100644
--- a/include/net/tc_act/tc_pedit.h
+++ b/include/net/tc_act/tc_pedit.h
@@ -3,12 +3,13 @@
 
 #include <net/act_api.h>
 
-struct tcf_pedit
-{
-	tca_gen(pedit);
-	unsigned char           nkeys;
-	unsigned char           flags;
-	struct tc_pedit_key     *keys;
+struct tcf_pedit {
+	struct tcf_common	common;
+	unsigned char		tcfp_nkeys;
+	unsigned char		tcfp_flags;
+	struct tc_pedit_key	*tcfp_keys;
 };
+#define to_pedit(pc) \
+	container_of(pc, struct tcf_pedit, common)
 
-#endif
+#endif /* __NET_TC_PED_H */
diff --git a/net/sched/act_api.c b/net/sched/act_api.c
index 6990747d6d5a..835070e9169c 100644
--- a/net/sched/act_api.c
+++ b/net/sched/act_api.c
@@ -33,16 +33,230 @@
 #include <net/sch_generic.h>
 #include <net/act_api.h>
 
-#if 0 /* control */
-#define DPRINTK(format, args...) printk(KERN_DEBUG format, ##args)
-#else
-#define DPRINTK(format, args...)
+void tcf_hash_destroy(struct tcf_common *p, struct tcf_hashinfo *hinfo)
+{
+	unsigned int h = tcf_hash(p->tcfc_index, hinfo->hmask);
+	struct tcf_common **p1p;
+
+	for (p1p = &hinfo->htab[h]; *p1p; p1p = &(*p1p)->tcfc_next) {
+		if (*p1p == p) {
+			write_lock_bh(hinfo->lock);
+			*p1p = p->tcfc_next;
+			write_unlock_bh(hinfo->lock);
+#ifdef CONFIG_NET_ESTIMATOR
+			gen_kill_estimator(&p->tcfc_bstats,
+					   &p->tcfc_rate_est);
 #endif
-#if 0 /* data */
-#define D2PRINTK(format, args...) printk(KERN_DEBUG format, ##args)
-#else
-#define D2PRINTK(format, args...)
+			kfree(p);
+			return;
+		}
+	}
+	BUG_TRAP(0);
+}
+EXPORT_SYMBOL(tcf_hash_destroy);
+
+int tcf_hash_release(struct tcf_common *p, int bind,
+		     struct tcf_hashinfo *hinfo)
+{
+	int ret = 0;
+
+	if (p) {
+		if (bind)
+			p->tcfc_bindcnt--;
+
+		p->tcfc_refcnt--;
+	       	if (p->tcfc_bindcnt <= 0 && p->tcfc_refcnt <= 0) {
+			tcf_hash_destroy(p, hinfo);
+			ret = 1;
+		}
+	}
+	return ret;
+}
+EXPORT_SYMBOL(tcf_hash_release);
+
+static int tcf_dump_walker(struct sk_buff *skb, struct netlink_callback *cb,
+			   struct tc_action *a, struct tcf_hashinfo *hinfo)
+{
+	struct tcf_common *p;
+	int err = 0, index = -1,i = 0, s_i = 0, n_i = 0;
+	struct rtattr *r ;
+
+	read_lock(hinfo->lock);
+
+	s_i = cb->args[0];
+
+	for (i = 0; i < (hinfo->hmask + 1); i++) {
+		p = hinfo->htab[tcf_hash(i, hinfo->hmask)];
+
+		for (; p; p = p->tcfc_next) {
+			index++;
+			if (index < s_i)
+				continue;
+			a->priv = p;
+			a->order = n_i;
+			r = (struct rtattr*) skb->tail;
+			RTA_PUT(skb, a->order, 0, NULL);
+			err = tcf_action_dump_1(skb, a, 0, 0);
+			if (err < 0) {
+				index--;
+				skb_trim(skb, (u8*)r - skb->data);
+				goto done;
+			}
+			r->rta_len = skb->tail - (u8*)r;
+			n_i++;
+			if (n_i >= TCA_ACT_MAX_PRIO)
+				goto done;
+		}
+	}
+done:
+	read_unlock(hinfo->lock);
+	if (n_i)
+		cb->args[0] += n_i;
+	return n_i;
+
+rtattr_failure:
+	skb_trim(skb, (u8*)r - skb->data);
+	goto done;
+}
+
+static int tcf_del_walker(struct sk_buff *skb, struct tc_action *a,
+			  struct tcf_hashinfo *hinfo)
+{
+	struct tcf_common *p, *s_p;
+	struct rtattr *r ;
+	int i= 0, n_i = 0;
+
+	r = (struct rtattr*) skb->tail;
+	RTA_PUT(skb, a->order, 0, NULL);
+	RTA_PUT(skb, TCA_KIND, IFNAMSIZ, a->ops->kind);
+	for (i = 0; i < (hinfo->hmask + 1); i++) {
+		p = hinfo->htab[tcf_hash(i, hinfo->hmask)];
+
+		while (p != NULL) {
+			s_p = p->tcfc_next;
+			if (ACT_P_DELETED == tcf_hash_release(p, 0, hinfo))
+				 module_put(a->ops->owner);
+			n_i++;
+			p = s_p;
+		}
+	}
+	RTA_PUT(skb, TCA_FCNT, 4, &n_i);
+	r->rta_len = skb->tail - (u8*)r;
+
+	return n_i;
+rtattr_failure:
+	skb_trim(skb, (u8*)r - skb->data);
+	return -EINVAL;
+}
+
+int tcf_generic_walker(struct sk_buff *skb, struct netlink_callback *cb,
+		       int type, struct tc_action *a)
+{
+	struct tcf_hashinfo *hinfo = a->ops->hinfo;
+
+	if (type == RTM_DELACTION) {
+		return tcf_del_walker(skb, a, hinfo);
+	} else if (type == RTM_GETACTION) {
+		return tcf_dump_walker(skb, cb, a, hinfo);
+	} else {
+		printk("tcf_generic_walker: unknown action %d\n", type);
+		return -EINVAL;
+	}
+}
+EXPORT_SYMBOL(tcf_generic_walker);
+
+struct tcf_common *tcf_hash_lookup(u32 index, struct tcf_hashinfo *hinfo)
+{
+	struct tcf_common *p;
+
+	read_lock(hinfo->lock);
+	for (p = hinfo->htab[tcf_hash(index, hinfo->hmask)]; p;
+	     p = p->tcfc_next) {
+		if (p->tcfc_index == index)
+			break;
+	}
+	read_unlock(hinfo->lock);
+
+	return p;
+}
+EXPORT_SYMBOL(tcf_hash_lookup);
+
+u32 tcf_hash_new_index(u32 *idx_gen, struct tcf_hashinfo *hinfo)
+{
+	u32 val = *idx_gen;
+
+	do {
+		if (++val == 0)
+			val = 1;
+	} while (tcf_hash_lookup(val, hinfo));
+
+	return (*idx_gen = val);
+}
+EXPORT_SYMBOL(tcf_hash_new_index);
+
+int tcf_hash_search(struct tc_action *a, u32 index)
+{
+	struct tcf_hashinfo *hinfo = a->ops->hinfo;
+	struct tcf_common *p = tcf_hash_lookup(index, hinfo);
+
+	if (p) {
+		a->priv = p;
+		return 1;
+	}
+	return 0;
+}
+EXPORT_SYMBOL(tcf_hash_search);
+
+struct tcf_common *tcf_hash_check(u32 index, struct tc_action *a, int bind,
+				  struct tcf_hashinfo *hinfo)
+{
+	struct tcf_common *p = NULL;
+	if (index && (p = tcf_hash_lookup(index, hinfo)) != NULL) {
+		if (bind) {
+			p->tcfc_bindcnt++;
+			p->tcfc_refcnt++;
+		}
+		a->priv = p;
+	}
+	return p;
+}
+EXPORT_SYMBOL(tcf_hash_check);
+
+struct tcf_common *tcf_hash_create(u32 index, struct rtattr *est, struct tc_action *a, int size, int bind, u32 *idx_gen, struct tcf_hashinfo *hinfo)
+{
+	struct tcf_common *p = kzalloc(size, GFP_KERNEL);
+
+	if (unlikely(!p))
+		return p;
+	p->tcfc_refcnt = 1;
+	if (bind)
+		p->tcfc_bindcnt = 1;
+
+	spin_lock_init(&p->tcfc_lock);
+	p->tcfc_stats_lock = &p->tcfc_lock;
+	p->tcfc_index = index ? index : tcf_hash_new_index(idx_gen, hinfo);
+	p->tcfc_tm.install = jiffies;
+	p->tcfc_tm.lastuse = jiffies;
+#ifdef CONFIG_NET_ESTIMATOR
+	if (est)
+		gen_new_estimator(&p->tcfc_bstats, &p->tcfc_rate_est,
+				  p->tcfc_stats_lock, est);
 #endif
+	a->priv = (void *) p;
+	return p;
+}
+EXPORT_SYMBOL(tcf_hash_create);
+
+void tcf_hash_insert(struct tcf_common *p, struct tcf_hashinfo *hinfo)
+{
+	unsigned int h = tcf_hash(p->tcfc_index, hinfo->hmask);
+
+	write_lock_bh(hinfo->lock);
+	p->tcfc_next = hinfo->htab[h];
+	hinfo->htab[h] = p;
+	write_unlock_bh(hinfo->lock);
+}
+EXPORT_SYMBOL(tcf_hash_insert);
 
 static struct tc_action_ops *act_base = NULL;
 static DEFINE_RWLOCK(act_mod_lock);
@@ -155,9 +369,6 @@ int tcf_action_exec(struct sk_buff *skb, struct tc_action *act,
 
 	if (skb->tc_verd & TC_NCLS) {
 		skb->tc_verd = CLR_TC_NCLS(skb->tc_verd);
-		D2PRINTK("(%p)tcf_action_exec: cleared TC_NCLS in %s out %s\n",
-		         skb, skb->input_dev ? skb->input_dev->name : "xxx",
-		         skb->dev->name);
 		ret = TC_ACT_OK;
 		goto exec_done;
 	}
@@ -187,8 +398,6 @@ void tcf_action_destroy(struct tc_action *act, int bind)
 
 	for (a = act; a; a = act) {
 		if (a->ops && a->ops->cleanup) {
-			DPRINTK("tcf_action_destroy destroying %p next %p\n",
-			        a, a->next);
 			if (a->ops->cleanup(a, bind) == ACT_P_DELETED)
 				module_put(a->ops->owner);
 			act = act->next;
@@ -331,7 +540,6 @@ struct tc_action *tcf_action_init_1(struct rtattr *rta, struct rtattr *est,
 	if (*err != ACT_P_CREATED)
 		module_put(a_o->owner);
 	a->ops = a_o;
-	DPRINTK("tcf_action_init_1: successfull %s\n", act_name);
 
 	*err = 0;
 	return a;
@@ -392,12 +600,12 @@ int tcf_action_copy_stats(struct sk_buff *skb, struct tc_action *a,
 	if (compat_mode) {
 		if (a->type == TCA_OLD_COMPAT)
 			err = gnet_stats_start_copy_compat(skb, 0,
-				TCA_STATS, TCA_XSTATS, h->stats_lock, &d);
+				TCA_STATS, TCA_XSTATS, h->tcf_stats_lock, &d);
 		else
 			return 0;
 	} else
 		err = gnet_stats_start_copy(skb, TCA_ACT_STATS,
-			h->stats_lock, &d);
+			h->tcf_stats_lock, &d);
 
 	if (err < 0)
 		goto errout;
@@ -406,11 +614,11 @@ int tcf_action_copy_stats(struct sk_buff *skb, struct tc_action *a,
 		if (a->ops->get_stats(skb, a) < 0)
 			goto errout;
 
-	if (gnet_stats_copy_basic(&d, &h->bstats) < 0 ||
+	if (gnet_stats_copy_basic(&d, &h->tcf_bstats) < 0 ||
 #ifdef CONFIG_NET_ESTIMATOR
-	    gnet_stats_copy_rate_est(&d, &h->rate_est) < 0 ||
+	    gnet_stats_copy_rate_est(&d, &h->tcf_rate_est) < 0 ||
 #endif
-	    gnet_stats_copy_queue(&d, &h->qstats) < 0)
+	    gnet_stats_copy_queue(&d, &h->tcf_qstats) < 0)
 		goto errout;
 
 	if (gnet_stats_finish_copy(&d) < 0)
diff --git a/net/sched/act_gact.c b/net/sched/act_gact.c
index e75a147ad60f..6cff56696a81 100644
--- a/net/sched/act_gact.c
+++ b/net/sched/act_gact.c
@@ -34,48 +34,43 @@
 #include <linux/tc_act/tc_gact.h>
 #include <net/tc_act/tc_gact.h>
 
-/* use generic hash table */
-#define MY_TAB_SIZE	16
-#define MY_TAB_MASK	15
-
-static u32 idx_gen;
-static struct tcf_gact *tcf_gact_ht[MY_TAB_SIZE];
+#define GACT_TAB_MASK	15
+static struct tcf_common *tcf_gact_ht[GACT_TAB_MASK + 1];
+static u32 gact_idx_gen;
 static DEFINE_RWLOCK(gact_lock);
 
-/* ovewrride the defaults */
-#define tcf_st		tcf_gact
-#define tc_st		tc_gact
-#define tcf_t_lock	gact_lock
-#define tcf_ht		tcf_gact_ht
-
-#define CONFIG_NET_ACT_INIT 1
-#include <net/pkt_act.h>
+static struct tcf_hashinfo gact_hash_info = {
+	.htab	=	tcf_gact_ht,
+	.hmask	=	GACT_TAB_MASK,
+	.lock	=	&gact_lock,
+};
 
 #ifdef CONFIG_GACT_PROB
-static int gact_net_rand(struct tcf_gact *p)
+static int gact_net_rand(struct tcf_gact *gact)
 {
-	if (net_random()%p->pval)
-		return p->action;
-	return p->paction;
+	if (net_random() % gact->tcfg_pval)
+		return gact->tcf_action;
+	return gact->tcfg_paction;
 }
 
-static int gact_determ(struct tcf_gact *p)
+static int gact_determ(struct tcf_gact *gact)
 {
-	if (p->bstats.packets%p->pval)
-		return p->action;
-	return p->paction;
+	if (gact->tcf_bstats.packets % gact->tcfg_pval)
+		return gact->tcf_action;
+	return gact->tcfg_paction;
 }
 
-typedef int (*g_rand)(struct tcf_gact *p);
+typedef int (*g_rand)(struct tcf_gact *gact);
 static g_rand gact_rand[MAX_RAND]= { NULL, gact_net_rand, gact_determ };
-#endif
+#endif /* CONFIG_GACT_PROB */
 
 static int tcf_gact_init(struct rtattr *rta, struct rtattr *est,
                          struct tc_action *a, int ovr, int bind)
 {
 	struct rtattr *tb[TCA_GACT_MAX];
 	struct tc_gact *parm;
-	struct tcf_gact *p;
+	struct tcf_gact *gact;
+	struct tcf_common *pc;
 	int ret = 0;
 
 	if (rta == NULL || rtattr_parse_nested(tb, TCA_GACT_MAX, rta) < 0)
@@ -94,105 +89,106 @@ static int tcf_gact_init(struct rtattr *rta, struct rtattr *est,
 		return -EOPNOTSUPP;
 #endif
 
-	p = tcf_hash_check(parm->index, a, ovr, bind);
-	if (p == NULL) {
-		p = tcf_hash_create(parm->index, est, a, sizeof(*p), ovr, bind);
-		if (p == NULL)
+	pc = tcf_hash_check(parm->index, a, bind, &gact_hash_info);
+	if (!pc) {
+		pc = tcf_hash_create(parm->index, est, a, sizeof(*gact),
+				     bind, &gact_idx_gen, &gact_hash_info);
+		if (unlikely(!pc))
 			return -ENOMEM;
 		ret = ACT_P_CREATED;
 	} else {
 		if (!ovr) {
-			tcf_hash_release(p, bind);
+			tcf_hash_release(pc, bind, &gact_hash_info);
 			return -EEXIST;
 		}
 	}
 
-	spin_lock_bh(&p->lock);
-	p->action = parm->action;
+	gact = to_gact(pc);
+
+	spin_lock_bh(&gact->tcf_lock);
+	gact->tcf_action = parm->action;
 #ifdef CONFIG_GACT_PROB
 	if (tb[TCA_GACT_PROB-1] != NULL) {
 		struct tc_gact_p *p_parm = RTA_DATA(tb[TCA_GACT_PROB-1]);
-		p->paction = p_parm->paction;
-		p->pval    = p_parm->pval;
-		p->ptype   = p_parm->ptype;
+		gact->tcfg_paction = p_parm->paction;
+		gact->tcfg_pval    = p_parm->pval;
+		gact->tcfg_ptype   = p_parm->ptype;
 	}
 #endif
-	spin_unlock_bh(&p->lock);
+	spin_unlock_bh(&gact->tcf_lock);
 	if (ret == ACT_P_CREATED)
-		tcf_hash_insert(p);
+		tcf_hash_insert(pc, &gact_hash_info);
 	return ret;
 }
 
-static int
-tcf_gact_cleanup(struct tc_action *a, int bind)
+static int tcf_gact_cleanup(struct tc_action *a, int bind)
 {
-	struct tcf_gact *p = PRIV(a, gact);
+	struct tcf_gact *gact = a->priv;
 
-	if (p != NULL)
-		return tcf_hash_release(p, bind);
+	if (gact)
+		return tcf_hash_release(&gact->common, bind, &gact_hash_info);
 	return 0;
 }
 
-static int
-tcf_gact(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res)
+static int tcf_gact(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res)
 {
-	struct tcf_gact *p = PRIV(a, gact);
+	struct tcf_gact *gact = a->priv;
 	int action = TC_ACT_SHOT;
 
-	spin_lock(&p->lock);
+	spin_lock(&gact->tcf_lock);
 #ifdef CONFIG_GACT_PROB
-	if (p->ptype && gact_rand[p->ptype] != NULL)
-		action = gact_rand[p->ptype](p);
+	if (gact->tcfg_ptype && gact_rand[gact->tcfg_ptype] != NULL)
+		action = gact_rand[gact->tcfg_ptype](gact);
 	else
-		action = p->action;
+		action = gact->tcf_action;
 #else
-	action = p->action;
+	action = gact->tcf_action;
 #endif
-	p->bstats.bytes += skb->len;
-	p->bstats.packets++;
+	gact->tcf_bstats.bytes += skb->len;
+	gact->tcf_bstats.packets++;
 	if (action == TC_ACT_SHOT)
-		p->qstats.drops++;
-	p->tm.lastuse = jiffies;
-	spin_unlock(&p->lock);
+		gact->tcf_qstats.drops++;
+	gact->tcf_tm.lastuse = jiffies;
+	spin_unlock(&gact->tcf_lock);
 
 	return action;
 }
 
-static int
-tcf_gact_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref)
+static int tcf_gact_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref)
 {
 	unsigned char *b = skb->tail;
 	struct tc_gact opt;
-	struct tcf_gact *p = PRIV(a, gact);
+	struct tcf_gact *gact = a->priv;
 	struct tcf_t t;
 
-	opt.index = p->index;
-	opt.refcnt = p->refcnt - ref;
-	opt.bindcnt = p->bindcnt - bind;
-	opt.action = p->action;
+	opt.index = gact->tcf_index;
+	opt.refcnt = gact->tcf_refcnt - ref;
+	opt.bindcnt = gact->tcf_bindcnt - bind;
+	opt.action = gact->tcf_action;
 	RTA_PUT(skb, TCA_GACT_PARMS, sizeof(opt), &opt);
 #ifdef CONFIG_GACT_PROB
-	if (p->ptype) {
+	if (gact->tcfg_ptype) {
 		struct tc_gact_p p_opt;
-		p_opt.paction = p->paction;
-		p_opt.pval = p->pval;
-		p_opt.ptype = p->ptype;
+		p_opt.paction = gact->tcfg_paction;
+		p_opt.pval = gact->tcfg_pval;
+		p_opt.ptype = gact->tcfg_ptype;
 		RTA_PUT(skb, TCA_GACT_PROB, sizeof(p_opt), &p_opt);
 	}
 #endif
-	t.install = jiffies_to_clock_t(jiffies - p->tm.install);
-	t.lastuse = jiffies_to_clock_t(jiffies - p->tm.lastuse);
-	t.expires = jiffies_to_clock_t(p->tm.expires);
+	t.install = jiffies_to_clock_t(jiffies - gact->tcf_tm.install);
+	t.lastuse = jiffies_to_clock_t(jiffies - gact->tcf_tm.lastuse);
+	t.expires = jiffies_to_clock_t(gact->tcf_tm.expires);
 	RTA_PUT(skb, TCA_GACT_TM, sizeof(t), &t);
 	return skb->len;
 
-      rtattr_failure:
+rtattr_failure:
 	skb_trim(skb, b - skb->data);
 	return -1;
 }
 
 static struct tc_action_ops act_gact_ops = {
 	.kind		=	"gact",
+	.hinfo		=	&gact_hash_info,
 	.type		=	TCA_ACT_GACT,
 	.capab		=	TCA_CAP_NONE,
 	.owner		=	THIS_MODULE,
@@ -208,8 +204,7 @@ MODULE_AUTHOR("Jamal Hadi Salim(2002-4)");
 MODULE_DESCRIPTION("Generic Classifier actions");
 MODULE_LICENSE("GPL");
 
-static int __init
-gact_init_module(void)
+static int __init gact_init_module(void)
 {
 #ifdef CONFIG_GACT_PROB
 	printk("GACT probability on\n");
@@ -219,8 +214,7 @@ gact_init_module(void)
 	return tcf_register_action(&act_gact_ops);
 }
 
-static void __exit
-gact_cleanup_module(void)
+static void __exit gact_cleanup_module(void)
 {
 	tcf_unregister_action(&act_gact_ops);
 }
diff --git a/net/sched/act_ipt.c b/net/sched/act_ipt.c
index d799e01248c4..224c078a398e 100644
--- a/net/sched/act_ipt.c
+++ b/net/sched/act_ipt.c
@@ -38,25 +38,19 @@
 
 #include <linux/netfilter_ipv4/ip_tables.h>
 
-/* use generic hash table */
-#define MY_TAB_SIZE     16
-#define MY_TAB_MASK     15
 
-static u32 idx_gen;
-static struct tcf_ipt *tcf_ipt_ht[MY_TAB_SIZE];
-/* ipt hash table lock */
+#define IPT_TAB_MASK     15
+static struct tcf_common *tcf_ipt_ht[IPT_TAB_MASK + 1];
+static u32 ipt_idx_gen;
 static DEFINE_RWLOCK(ipt_lock);
 
-/* ovewrride the defaults */
-#define tcf_st		tcf_ipt
-#define tcf_t_lock	ipt_lock
-#define tcf_ht		tcf_ipt_ht
-
-#define CONFIG_NET_ACT_INIT
-#include <net/pkt_act.h>
+static struct tcf_hashinfo ipt_hash_info = {
+	.htab	=	tcf_ipt_ht,
+	.hmask	=	IPT_TAB_MASK,
+	.lock	=	&ipt_lock,
+};
 
-static int
-ipt_init_target(struct ipt_entry_target *t, char *table, unsigned int hook)
+static int ipt_init_target(struct ipt_entry_target *t, char *table, unsigned int hook)
 {
 	struct ipt_target *target;
 	int ret = 0;
@@ -65,7 +59,6 @@ ipt_init_target(struct ipt_entry_target *t, char *table, unsigned int hook)
 	if (!target)
 		return -ENOENT;
 
-	DPRINTK("ipt_init_target: found %s\n", target->name);
 	t->u.kernel.target = target;
 
 	ret = xt_check_target(target, AF_INET, t->u.target_size - sizeof(*t),
@@ -78,8 +71,6 @@ ipt_init_target(struct ipt_entry_target *t, char *table, unsigned int hook)
 		    			       t->u.kernel.target, t->data,
 					       t->u.target_size - sizeof(*t),
 					       hook)) {
-		DPRINTK("ipt_init_target: check failed for `%s'.\n",
-			t->u.kernel.target->name);
 		module_put(t->u.kernel.target->me);
 		ret = -EINVAL;
 	}
@@ -87,8 +78,7 @@ ipt_init_target(struct ipt_entry_target *t, char *table, unsigned int hook)
 	return ret;
 }
 
-static void
-ipt_destroy_target(struct ipt_entry_target *t)
+static void ipt_destroy_target(struct ipt_entry_target *t)
 {
 	if (t->u.kernel.target->destroy)
 		t->u.kernel.target->destroy(t->u.kernel.target, t->data,
@@ -96,31 +86,30 @@ ipt_destroy_target(struct ipt_entry_target *t)
         module_put(t->u.kernel.target->me);
 }
 
-static int
-tcf_ipt_release(struct tcf_ipt *p, int bind)
+static int tcf_ipt_release(struct tcf_ipt *ipt, int bind)
 {
 	int ret = 0;
-	if (p) {
+	if (ipt) {
 		if (bind)
-			p->bindcnt--;
-		p->refcnt--;
-		if (p->bindcnt <= 0 && p->refcnt <= 0) {
-			ipt_destroy_target(p->t);
-			kfree(p->tname);
-			kfree(p->t);
-			tcf_hash_destroy(p);
+			ipt->tcf_bindcnt--;
+		ipt->tcf_refcnt--;
+		if (ipt->tcf_bindcnt <= 0 && ipt->tcf_refcnt <= 0) {
+			ipt_destroy_target(ipt->tcfi_t);
+			kfree(ipt->tcfi_tname);
+			kfree(ipt->tcfi_t);
+			tcf_hash_destroy(&ipt->common, &ipt_hash_info);
 			ret = ACT_P_DELETED;
 		}
 	}
 	return ret;
 }
 
-static int
-tcf_ipt_init(struct rtattr *rta, struct rtattr *est, struct tc_action *a,
-             int ovr, int bind)
+static int tcf_ipt_init(struct rtattr *rta, struct rtattr *est,
+			struct tc_action *a, int ovr, int bind)
 {
 	struct rtattr *tb[TCA_IPT_MAX];
-	struct tcf_ipt *p;
+	struct tcf_ipt *ipt;
+	struct tcf_common *pc;
 	struct ipt_entry_target *td, *t;
 	char *tname;
 	int ret = 0, err;
@@ -144,49 +133,51 @@ tcf_ipt_init(struct rtattr *rta, struct rtattr *est, struct tc_action *a,
 	    RTA_PAYLOAD(tb[TCA_IPT_INDEX-1]) >= sizeof(u32))
 		index = *(u32 *)RTA_DATA(tb[TCA_IPT_INDEX-1]);
 
-	p = tcf_hash_check(index, a, ovr, bind);
-	if (p == NULL) {
-		p = tcf_hash_create(index, est, a, sizeof(*p), ovr, bind);
-		if (p == NULL)
+	pc = tcf_hash_check(index, a, bind, &ipt_hash_info);
+	if (!pc) {
+		pc = tcf_hash_create(index, est, a, sizeof(*ipt), bind,
+				     &ipt_idx_gen, &ipt_hash_info);
+		if (unlikely(!pc))
 			return -ENOMEM;
 		ret = ACT_P_CREATED;
 	} else {
 		if (!ovr) {
-			tcf_ipt_release(p, bind);
+			tcf_ipt_release(to_ipt(pc), bind);
 			return -EEXIST;
 		}
 	}
+	ipt = to_ipt(pc);
 
 	hook = *(u32 *)RTA_DATA(tb[TCA_IPT_HOOK-1]);
 
 	err = -ENOMEM;
 	tname = kmalloc(IFNAMSIZ, GFP_KERNEL);
-	if (tname == NULL)
+	if (unlikely(!tname))
 		goto err1;
 	if (tb[TCA_IPT_TABLE - 1] == NULL ||
 	    rtattr_strlcpy(tname, tb[TCA_IPT_TABLE-1], IFNAMSIZ) >= IFNAMSIZ)
 		strcpy(tname, "mangle");
 
 	t = kmalloc(td->u.target_size, GFP_KERNEL);
-	if (t == NULL)
+	if (unlikely(!t))
 		goto err2;
 	memcpy(t, td, td->u.target_size);
 
 	if ((err = ipt_init_target(t, tname, hook)) < 0)
 		goto err3;
 
-	spin_lock_bh(&p->lock);
+	spin_lock_bh(&ipt->tcf_lock);
 	if (ret != ACT_P_CREATED) {
-		ipt_destroy_target(p->t);
-		kfree(p->tname);
-		kfree(p->t);
+		ipt_destroy_target(ipt->tcfi_t);
+		kfree(ipt->tcfi_tname);
+		kfree(ipt->tcfi_t);
 	}
-	p->tname = tname;
-	p->t     = t;
-	p->hook  = hook;
-	spin_unlock_bh(&p->lock);
+	ipt->tcfi_tname = tname;
+	ipt->tcfi_t     = t;
+	ipt->tcfi_hook  = hook;
+	spin_unlock_bh(&ipt->tcf_lock);
 	if (ret == ACT_P_CREATED)
-		tcf_hash_insert(p);
+		tcf_hash_insert(pc, &ipt_hash_info);
 	return ret;
 
 err3:
@@ -194,33 +185,32 @@ err3:
 err2:
 	kfree(tname);
 err1:
-	kfree(p);
+	kfree(pc);
 	return err;
 }
 
-static int
-tcf_ipt_cleanup(struct tc_action *a, int bind)
+static int tcf_ipt_cleanup(struct tc_action *a, int bind)
 {
-	struct tcf_ipt *p = PRIV(a, ipt);
-	return tcf_ipt_release(p, bind);
+	struct tcf_ipt *ipt = a->priv;
+	return tcf_ipt_release(ipt, bind);
 }
 
-static int
-tcf_ipt(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res)
+static int tcf_ipt(struct sk_buff *skb, struct tc_action *a,
+		   struct tcf_result *res)
 {
 	int ret = 0, result = 0;
-	struct tcf_ipt *p = PRIV(a, ipt);
+	struct tcf_ipt *ipt = a->priv;
 
 	if (skb_cloned(skb)) {
 		if (pskb_expand_head(skb, 0, 0, GFP_ATOMIC))
 			return TC_ACT_UNSPEC;
 	}
 
-	spin_lock(&p->lock);
+	spin_lock(&ipt->tcf_lock);
 
-	p->tm.lastuse = jiffies;
-	p->bstats.bytes += skb->len;
-	p->bstats.packets++;
+	ipt->tcf_tm.lastuse = jiffies;
+	ipt->tcf_bstats.bytes += skb->len;
+	ipt->tcf_bstats.packets++;
 
 	/* yes, we have to worry about both in and out dev
 	 worry later - danger - this API seems to have changed
@@ -229,16 +219,17 @@ tcf_ipt(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res)
 	/* iptables targets take a double skb pointer in case the skb
 	 * needs to be replaced. We don't own the skb, so this must not
 	 * happen. The pskb_expand_head above should make sure of this */
-	ret = p->t->u.kernel.target->target(&skb, skb->dev, NULL, p->hook,
-					    p->t->u.kernel.target, p->t->data,
-					    NULL);
+	ret = ipt->tcfi_t->u.kernel.target->target(&skb, skb->dev, NULL,
+						   ipt->tcfi_hook,
+						   ipt->tcfi_t->u.kernel.target,
+						   ipt->tcfi_t->data, NULL);
 	switch (ret) {
 	case NF_ACCEPT:
 		result = TC_ACT_OK;
 		break;
 	case NF_DROP:
 		result = TC_ACT_SHOT;
-		p->qstats.drops++;
+		ipt->tcf_qstats.drops++;
 		break;
 	case IPT_CONTINUE:
 		result = TC_ACT_PIPE;
@@ -249,53 +240,46 @@ tcf_ipt(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res)
 		result = TC_POLICE_OK;
 		break;
 	}
-	spin_unlock(&p->lock);
+	spin_unlock(&ipt->tcf_lock);
 	return result;
 
 }
 
-static int
-tcf_ipt_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref)
+static int tcf_ipt_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref)
 {
+	unsigned char *b = skb->tail;
+	struct tcf_ipt *ipt = a->priv;
 	struct ipt_entry_target *t;
 	struct tcf_t tm;
 	struct tc_cnt c;
-	unsigned char *b = skb->tail;
-	struct tcf_ipt *p = PRIV(a, ipt);
 
 	/* for simple targets kernel size == user size
 	** user name = target name
 	** for foolproof you need to not assume this
 	*/
 
-	t = kmalloc(p->t->u.user.target_size, GFP_ATOMIC);
-	if (t == NULL)
+	t = kmalloc(ipt->tcfi_t->u.user.target_size, GFP_ATOMIC);
+	if (unlikely(!t))
 		goto rtattr_failure;
 
-	c.bindcnt = p->bindcnt - bind;
-	c.refcnt = p->refcnt - ref;
-	memcpy(t, p->t, p->t->u.user.target_size);
-	strcpy(t->u.user.name, p->t->u.kernel.target->name);
-
-	DPRINTK("\ttcf_ipt_dump tablename %s length %d\n", p->tname,
-		strlen(p->tname));
-	DPRINTK("\tdump target name %s size %d size user %d "
-	        "data[0] %x data[1] %x\n", p->t->u.kernel.target->name,
-	        p->t->u.target_size, p->t->u.user.target_size,
-	        p->t->data[0], p->t->data[1]);
-	RTA_PUT(skb, TCA_IPT_TARG, p->t->u.user.target_size, t);
-	RTA_PUT(skb, TCA_IPT_INDEX, 4, &p->index);
-	RTA_PUT(skb, TCA_IPT_HOOK, 4, &p->hook);
+	c.bindcnt = ipt->tcf_bindcnt - bind;
+	c.refcnt = ipt->tcf_refcnt - ref;
+	memcpy(t, ipt->tcfi_t, ipt->tcfi_t->u.user.target_size);
+	strcpy(t->u.user.name, ipt->tcfi_t->u.kernel.target->name);
+
+	RTA_PUT(skb, TCA_IPT_TARG, ipt->tcfi_t->u.user.target_size, t);
+	RTA_PUT(skb, TCA_IPT_INDEX, 4, &ipt->tcf_index);
+	RTA_PUT(skb, TCA_IPT_HOOK, 4, &ipt->tcfi_hook);
 	RTA_PUT(skb, TCA_IPT_CNT, sizeof(struct tc_cnt), &c);
-	RTA_PUT(skb, TCA_IPT_TABLE, IFNAMSIZ, p->tname);
-	tm.install = jiffies_to_clock_t(jiffies - p->tm.install);
-	tm.lastuse = jiffies_to_clock_t(jiffies - p->tm.lastuse);
-	tm.expires = jiffies_to_clock_t(p->tm.expires);
+	RTA_PUT(skb, TCA_IPT_TABLE, IFNAMSIZ, ipt->tcfi_tname);
+	tm.install = jiffies_to_clock_t(jiffies - ipt->tcf_tm.install);
+	tm.lastuse = jiffies_to_clock_t(jiffies - ipt->tcf_tm.lastuse);
+	tm.expires = jiffies_to_clock_t(ipt->tcf_tm.expires);
 	RTA_PUT(skb, TCA_IPT_TM, sizeof (tm), &tm);
 	kfree(t);
 	return skb->len;
 
-      rtattr_failure:
+rtattr_failure:
 	skb_trim(skb, b - skb->data);
 	kfree(t);
 	return -1;
@@ -303,6 +287,7 @@ tcf_ipt_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref)
 
 static struct tc_action_ops act_ipt_ops = {
 	.kind		=	"ipt",
+	.hinfo		=	&ipt_hash_info,
 	.type		=	TCA_ACT_IPT,
 	.capab		=	TCA_CAP_NONE,
 	.owner		=	THIS_MODULE,
@@ -318,14 +303,12 @@ MODULE_AUTHOR("Jamal Hadi Salim(2002-4)");
 MODULE_DESCRIPTION("Iptables target actions");
 MODULE_LICENSE("GPL");
 
-static int __init
-ipt_init_module(void)
+static int __init ipt_init_module(void)
 {
 	return tcf_register_action(&act_ipt_ops);
 }
 
-static void __exit
-ipt_cleanup_module(void)
+static void __exit ipt_cleanup_module(void)
 {
 	tcf_unregister_action(&act_ipt_ops);
 }
diff --git a/net/sched/act_mirred.c b/net/sched/act_mirred.c
index fc562047ecc5..483897271f15 100644
--- a/net/sched/act_mirred.c
+++ b/net/sched/act_mirred.c
@@ -39,46 +39,39 @@
 #include <linux/etherdevice.h>
 #include <linux/if_arp.h>
 
-
-/* use generic hash table */
-#define MY_TAB_SIZE     8
-#define MY_TAB_MASK     (MY_TAB_SIZE - 1)
-static u32 idx_gen;
-static struct tcf_mirred *tcf_mirred_ht[MY_TAB_SIZE];
+#define MIRRED_TAB_MASK     7
+static struct tcf_common *tcf_mirred_ht[MIRRED_TAB_MASK + 1];
+static u32 mirred_idx_gen;
 static DEFINE_RWLOCK(mirred_lock);
 
-/* ovewrride the defaults */
-#define tcf_st		tcf_mirred
-#define tc_st		tc_mirred
-#define tcf_t_lock	mirred_lock
-#define tcf_ht		tcf_mirred_ht
-
-#define CONFIG_NET_ACT_INIT 1
-#include <net/pkt_act.h>
+static struct tcf_hashinfo mirred_hash_info = {
+	.htab	=	tcf_mirred_ht,
+	.hmask	=	MIRRED_TAB_MASK,
+	.lock	=	&mirred_lock,
+};
 
-static inline int
-tcf_mirred_release(struct tcf_mirred *p, int bind)
+static inline int tcf_mirred_release(struct tcf_mirred *m, int bind)
 {
-	if (p) {
+	if (m) {
 		if (bind)
-			p->bindcnt--;
-		p->refcnt--;
-		if(!p->bindcnt && p->refcnt <= 0) {
-			dev_put(p->dev);
-			tcf_hash_destroy(p);
+			m->tcf_bindcnt--;
+		m->tcf_refcnt--;
+		if(!m->tcf_bindcnt && m->tcf_refcnt <= 0) {
+			dev_put(m->tcfm_dev);
+			tcf_hash_destroy(&m->common, &mirred_hash_info);
 			return 1;
 		}
 	}
 	return 0;
 }
 
-static int
-tcf_mirred_init(struct rtattr *rta, struct rtattr *est, struct tc_action *a,
-                int ovr, int bind)
+static int tcf_mirred_init(struct rtattr *rta, struct rtattr *est,
+			   struct tc_action *a, int ovr, int bind)
 {
 	struct rtattr *tb[TCA_MIRRED_MAX];
 	struct tc_mirred *parm;
-	struct tcf_mirred *p;
+	struct tcf_mirred *m;
+	struct tcf_common *pc;
 	struct net_device *dev = NULL;
 	int ret = 0;
 	int ok_push = 0;
@@ -110,64 +103,62 @@ tcf_mirred_init(struct rtattr *rta, struct rtattr *est, struct tc_action *a,
 		}
 	}
 
-	p = tcf_hash_check(parm->index, a, ovr, bind);
-	if (p == NULL) {
+	pc = tcf_hash_check(parm->index, a, bind, &mirred_hash_info);
+	if (!pc) {
 		if (!parm->ifindex)
 			return -EINVAL;
-		p = tcf_hash_create(parm->index, est, a, sizeof(*p), ovr, bind);
-		if (p == NULL)
+		pc = tcf_hash_create(parm->index, est, a, sizeof(*m), bind,
+				     &mirred_idx_gen, &mirred_hash_info);
+		if (unlikely(!pc))
 			return -ENOMEM;
 		ret = ACT_P_CREATED;
 	} else {
 		if (!ovr) {
-			tcf_mirred_release(p, bind);
+			tcf_mirred_release(to_mirred(pc), bind);
 			return -EEXIST;
 		}
 	}
+	m = to_mirred(pc);
 
-	spin_lock_bh(&p->lock);
-	p->action = parm->action;
-	p->eaction = parm->eaction;
+	spin_lock_bh(&m->tcf_lock);
+	m->tcf_action = parm->action;
+	m->tcfm_eaction = parm->eaction;
 	if (parm->ifindex) {
-		p->ifindex = parm->ifindex;
+		m->tcfm_ifindex = parm->ifindex;
 		if (ret != ACT_P_CREATED)
-			dev_put(p->dev);
-		p->dev = dev;
+			dev_put(m->tcfm_dev);
+		m->tcfm_dev = dev;
 		dev_hold(dev);
-		p->ok_push = ok_push;
+		m->tcfm_ok_push = ok_push;
 	}
-	spin_unlock_bh(&p->lock);
+	spin_unlock_bh(&m->tcf_lock);
 	if (ret == ACT_P_CREATED)
-		tcf_hash_insert(p);
+		tcf_hash_insert(pc, &mirred_hash_info);
 
-	DPRINTK("tcf_mirred_init index %d action %d eaction %d device %s "
-	        "ifindex %d\n", parm->index, parm->action, parm->eaction,
-	        dev->name, parm->ifindex);
 	return ret;
 }
 
-static int
-tcf_mirred_cleanup(struct tc_action *a, int bind)
+static int tcf_mirred_cleanup(struct tc_action *a, int bind)
 {
-	struct tcf_mirred *p = PRIV(a, mirred);
+	struct tcf_mirred *m = a->priv;
 
-	if (p != NULL)
-		return tcf_mirred_release(p, bind);
+	if (m)
+		return tcf_mirred_release(m, bind);
 	return 0;
 }
 
-static int
-tcf_mirred(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res)
+static int tcf_mirred(struct sk_buff *skb, struct tc_action *a,
+		      struct tcf_result *res)
 {
-	struct tcf_mirred *p = PRIV(a, mirred);
+	struct tcf_mirred *m = a->priv;
 	struct net_device *dev;
 	struct sk_buff *skb2 = NULL;
 	u32 at = G_TC_AT(skb->tc_verd);
 
-	spin_lock(&p->lock);
+	spin_lock(&m->tcf_lock);
 
-	dev = p->dev;
-	p->tm.lastuse = jiffies;
+	dev = m->tcfm_dev;
+	m->tcf_tm.lastuse = jiffies;
 
 	if (!(dev->flags&IFF_UP) ) {
 		if (net_ratelimit())
@@ -176,10 +167,10 @@ tcf_mirred(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res)
 bad_mirred:
 		if (skb2 != NULL)
 			kfree_skb(skb2);
-		p->qstats.overlimits++;
-		p->bstats.bytes += skb->len;
-		p->bstats.packets++;
-		spin_unlock(&p->lock);
+		m->tcf_qstats.overlimits++;
+		m->tcf_bstats.bytes += skb->len;
+		m->tcf_bstats.packets++;
+		spin_unlock(&m->tcf_lock);
 		/* should we be asking for packet to be dropped?
 		 * may make sense for redirect case only
 		*/
@@ -189,59 +180,59 @@ bad_mirred:
 	skb2 = skb_clone(skb, GFP_ATOMIC);
 	if (skb2 == NULL)
 		goto bad_mirred;
-	if (p->eaction != TCA_EGRESS_MIRROR && p->eaction != TCA_EGRESS_REDIR) {
+	if (m->tcfm_eaction != TCA_EGRESS_MIRROR &&
+	    m->tcfm_eaction != TCA_EGRESS_REDIR) {
 		if (net_ratelimit())
-			printk("tcf_mirred unknown action %d\n", p->eaction);
+			printk("tcf_mirred unknown action %d\n",
+			       m->tcfm_eaction);
 		goto bad_mirred;
 	}
 
-	p->bstats.bytes += skb2->len;
-	p->bstats.packets++;
+	m->tcf_bstats.bytes += skb2->len;
+	m->tcf_bstats.packets++;
 	if (!(at & AT_EGRESS))
-		if (p->ok_push)
+		if (m->tcfm_ok_push)
 			skb_push(skb2, skb2->dev->hard_header_len);
 
 	/* mirror is always swallowed */
-	if (p->eaction != TCA_EGRESS_MIRROR)
+	if (m->tcfm_eaction != TCA_EGRESS_MIRROR)
 		skb2->tc_verd = SET_TC_FROM(skb2->tc_verd, at);
 
 	skb2->dev = dev;
 	skb2->input_dev = skb->dev;
 	dev_queue_xmit(skb2);
-	spin_unlock(&p->lock);
-	return p->action;
+	spin_unlock(&m->tcf_lock);
+	return m->tcf_action;
 }
 
-static int
-tcf_mirred_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref)
+static int tcf_mirred_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref)
 {
 	unsigned char *b = skb->tail;
+	struct tcf_mirred *m = a->priv;
 	struct tc_mirred opt;
-	struct tcf_mirred *p = PRIV(a, mirred);
 	struct tcf_t t;
 
-	opt.index = p->index;
-	opt.action = p->action;
-	opt.refcnt = p->refcnt - ref;
-	opt.bindcnt = p->bindcnt - bind;
-	opt.eaction = p->eaction;
-	opt.ifindex = p->ifindex;
-	DPRINTK("tcf_mirred_dump index %d action %d eaction %d ifindex %d\n",
-	         p->index, p->action, p->eaction, p->ifindex);
+	opt.index = m->tcf_index;
+	opt.action = m->tcf_action;
+	opt.refcnt = m->tcf_refcnt - ref;
+	opt.bindcnt = m->tcf_bindcnt - bind;
+	opt.eaction = m->tcfm_eaction;
+	opt.ifindex = m->tcfm_ifindex;
 	RTA_PUT(skb, TCA_MIRRED_PARMS, sizeof(opt), &opt);
-	t.install = jiffies_to_clock_t(jiffies - p->tm.install);
-	t.lastuse = jiffies_to_clock_t(jiffies - p->tm.lastuse);
-	t.expires = jiffies_to_clock_t(p->tm.expires);
+	t.install = jiffies_to_clock_t(jiffies - m->tcf_tm.install);
+	t.lastuse = jiffies_to_clock_t(jiffies - m->tcf_tm.lastuse);
+	t.expires = jiffies_to_clock_t(m->tcf_tm.expires);
 	RTA_PUT(skb, TCA_MIRRED_TM, sizeof(t), &t);
 	return skb->len;
 
-      rtattr_failure:
+rtattr_failure:
 	skb_trim(skb, b - skb->data);
 	return -1;
 }
 
 static struct tc_action_ops act_mirred_ops = {
 	.kind		=	"mirred",
+	.hinfo		=	&mirred_hash_info,
 	.type		=	TCA_ACT_MIRRED,
 	.capab		=	TCA_CAP_NONE,
 	.owner		=	THIS_MODULE,
@@ -257,15 +248,13 @@ MODULE_AUTHOR("Jamal Hadi Salim(2002)");
 MODULE_DESCRIPTION("Device Mirror/redirect actions");
 MODULE_LICENSE("GPL");
 
-static int __init
-mirred_init_module(void)
+static int __init mirred_init_module(void)
 {
 	printk("Mirror/redirect action on\n");
 	return tcf_register_action(&act_mirred_ops);
 }
 
-static void __exit
-mirred_cleanup_module(void)
+static void __exit mirred_cleanup_module(void)
 {
 	tcf_unregister_action(&act_mirred_ops);
 }
diff --git a/net/sched/act_pedit.c b/net/sched/act_pedit.c
index f257475e0e0c..8ac65c219b98 100644
--- a/net/sched/act_pedit.c
+++ b/net/sched/act_pedit.c
@@ -33,32 +33,25 @@
 #include <linux/tc_act/tc_pedit.h>
 #include <net/tc_act/tc_pedit.h>
 
-
-#define PEDIT_DEB 1
-
-/* use generic hash table */
-#define MY_TAB_SIZE     16
-#define MY_TAB_MASK     15
-static u32 idx_gen;
-static struct tcf_pedit *tcf_pedit_ht[MY_TAB_SIZE];
+#define PEDIT_TAB_MASK	15
+static struct tcf_common *tcf_pedit_ht[PEDIT_TAB_MASK + 1];
+static u32 pedit_idx_gen;
 static DEFINE_RWLOCK(pedit_lock);
 
-#define tcf_st		tcf_pedit
-#define tc_st		tc_pedit
-#define tcf_t_lock	pedit_lock
-#define tcf_ht		tcf_pedit_ht
-
-#define CONFIG_NET_ACT_INIT 1
-#include <net/pkt_act.h>
+static struct tcf_hashinfo pedit_hash_info = {
+	.htab	=	tcf_pedit_ht,
+	.hmask	=	PEDIT_TAB_MASK,
+	.lock	=	&pedit_lock,
+};
 
-static int
-tcf_pedit_init(struct rtattr *rta, struct rtattr *est, struct tc_action *a,
-               int ovr, int bind)
+static int tcf_pedit_init(struct rtattr *rta, struct rtattr *est,
+			  struct tc_action *a, int ovr, int bind)
 {
 	struct rtattr *tb[TCA_PEDIT_MAX];
 	struct tc_pedit *parm;
 	int ret = 0;
 	struct tcf_pedit *p;
+	struct tcf_common *pc;
 	struct tc_pedit_key *keys = NULL;
 	int ksize;
 
@@ -73,54 +66,56 @@ tcf_pedit_init(struct rtattr *rta, struct rtattr *est, struct tc_action *a,
 	if (RTA_PAYLOAD(tb[TCA_PEDIT_PARMS-1]) < sizeof(*parm) + ksize)
 		return -EINVAL;
 
-	p = tcf_hash_check(parm->index, a, ovr, bind);
-	if (p == NULL) {
+	pc = tcf_hash_check(parm->index, a, bind, &pedit_hash_info);
+	if (!pc) {
 		if (!parm->nkeys)
 			return -EINVAL;
-		p = tcf_hash_create(parm->index, est, a, sizeof(*p), ovr, bind);
-		if (p == NULL)
+		pc = tcf_hash_create(parm->index, est, a, sizeof(*p), bind,
+				     &pedit_idx_gen, &pedit_hash_info);
+		if (unlikely(!pc))
 			return -ENOMEM;
+		p = to_pedit(pc);
 		keys = kmalloc(ksize, GFP_KERNEL);
 		if (keys == NULL) {
-			kfree(p);
+			kfree(pc);
 			return -ENOMEM;
 		}
 		ret = ACT_P_CREATED;
 	} else {
+		p = to_pedit(pc);
 		if (!ovr) {
-			tcf_hash_release(p, bind);
+			tcf_hash_release(pc, bind, &pedit_hash_info);
 			return -EEXIST;
 		}
-		if (p->nkeys && p->nkeys != parm->nkeys) {
+		if (p->tcfp_nkeys && p->tcfp_nkeys != parm->nkeys) {
 			keys = kmalloc(ksize, GFP_KERNEL);
 			if (keys == NULL)
 				return -ENOMEM;
 		}
 	}
 
-	spin_lock_bh(&p->lock);
-	p->flags = parm->flags;
-	p->action = parm->action;
+	spin_lock_bh(&p->tcf_lock);
+	p->tcfp_flags = parm->flags;
+	p->tcf_action = parm->action;
 	if (keys) {
-		kfree(p->keys);
-		p->keys = keys;
-		p->nkeys = parm->nkeys;
+		kfree(p->tcfp_keys);
+		p->tcfp_keys = keys;
+		p->tcfp_nkeys = parm->nkeys;
 	}
-	memcpy(p->keys, parm->keys, ksize);
-	spin_unlock_bh(&p->lock);
+	memcpy(p->tcfp_keys, parm->keys, ksize);
+	spin_unlock_bh(&p->tcf_lock);
 	if (ret == ACT_P_CREATED)
-		tcf_hash_insert(p);
+		tcf_hash_insert(pc, &pedit_hash_info);
 	return ret;
 }
 
-static int
-tcf_pedit_cleanup(struct tc_action *a, int bind)
+static int tcf_pedit_cleanup(struct tc_action *a, int bind)
 {
-	struct tcf_pedit *p = PRIV(a, pedit);
+	struct tcf_pedit *p = a->priv;
 
-	if (p != NULL) {
-		struct tc_pedit_key *keys = p->keys;
-		if (tcf_hash_release(p, bind)) {
+	if (p) {
+		struct tc_pedit_key *keys = p->tcfp_keys;
+		if (tcf_hash_release(&p->common, bind, &pedit_hash_info)) {
 			kfree(keys);
 			return 1;
 		}
@@ -128,30 +123,30 @@ tcf_pedit_cleanup(struct tc_action *a, int bind)
 	return 0;
 }
 
-static int
-tcf_pedit(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res)
+static int tcf_pedit(struct sk_buff *skb, struct tc_action *a,
+		     struct tcf_result *res)
 {
-	struct tcf_pedit *p = PRIV(a, pedit);
+	struct tcf_pedit *p = a->priv;
 	int i, munged = 0;
 	u8 *pptr;
 
 	if (!(skb->tc_verd & TC_OK2MUNGE)) {
 		/* should we set skb->cloned? */
 		if (pskb_expand_head(skb, 0, 0, GFP_ATOMIC)) {
-			return p->action;
+			return p->tcf_action;
 		}
 	}
 
 	pptr = skb->nh.raw;
 
-	spin_lock(&p->lock);
+	spin_lock(&p->tcf_lock);
 
-	p->tm.lastuse = jiffies;
+	p->tcf_tm.lastuse = jiffies;
 
-	if (p->nkeys > 0) {
-		struct tc_pedit_key *tkey = p->keys;
+	if (p->tcfp_nkeys > 0) {
+		struct tc_pedit_key *tkey = p->tcfp_keys;
 
-		for (i = p->nkeys; i > 0; i--, tkey++) {
+		for (i = p->tcfp_nkeys; i > 0; i--, tkey++) {
 			u32 *ptr;
 			int offset = tkey->off;
 
@@ -169,7 +164,8 @@ tcf_pedit(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res)
 				printk("offset must be on 32 bit boundaries\n");
 				goto bad;
 			}
-			if (skb->len < 0 || (offset > 0 && offset > skb->len)) {
+			if (skb->len < 0 ||
+			    (offset > 0 && offset > skb->len)) {
 				printk("offset %d cant exceed pkt length %d\n",
 				       offset, skb->len);
 				goto bad;
@@ -185,63 +181,47 @@ tcf_pedit(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res)
 			skb->tc_verd = SET_TC_MUNGED(skb->tc_verd);
 		goto done;
 	} else {
-		printk("pedit BUG: index %d\n",p->index);
+		printk("pedit BUG: index %d\n", p->tcf_index);
 	}
 
 bad:
-	p->qstats.overlimits++;
+	p->tcf_qstats.overlimits++;
 done:
-	p->bstats.bytes += skb->len;
-	p->bstats.packets++;
-	spin_unlock(&p->lock);
-	return p->action;
+	p->tcf_bstats.bytes += skb->len;
+	p->tcf_bstats.packets++;
+	spin_unlock(&p->tcf_lock);
+	return p->tcf_action;
 }
 
-static int
-tcf_pedit_dump(struct sk_buff *skb, struct tc_action *a,int bind, int ref)
+static int tcf_pedit_dump(struct sk_buff *skb, struct tc_action *a,
+			  int bind, int ref)
 {
 	unsigned char *b = skb->tail;
+	struct tcf_pedit *p = a->priv;
 	struct tc_pedit *opt;
-	struct tcf_pedit *p = PRIV(a, pedit);
 	struct tcf_t t;
 	int s; 
 		
-	s = sizeof(*opt) + p->nkeys * sizeof(struct tc_pedit_key);
+	s = sizeof(*opt) + p->tcfp_nkeys * sizeof(struct tc_pedit_key);
 
 	/* netlink spinlocks held above us - must use ATOMIC */
 	opt = kzalloc(s, GFP_ATOMIC);
-	if (opt == NULL)
+	if (unlikely(!opt))
 		return -ENOBUFS;
 
-	memcpy(opt->keys, p->keys, p->nkeys * sizeof(struct tc_pedit_key));
-	opt->index = p->index;
-	opt->nkeys = p->nkeys;
-	opt->flags = p->flags;
-	opt->action = p->action;
-	opt->refcnt = p->refcnt - ref;
-	opt->bindcnt = p->bindcnt - bind;
-
-
-#ifdef PEDIT_DEB
-	{                
-		/* Debug - get rid of later */
-		int i;
-		struct tc_pedit_key *key = opt->keys;
-
-		for (i=0; i<opt->nkeys; i++, key++) {
-			printk( "\n key #%d",i);
-			printk( "  at %d: val %08x mask %08x",
-			(unsigned int)key->off,
-			(unsigned int)key->val,
-			(unsigned int)key->mask);
-		}
-	}
-#endif
+	memcpy(opt->keys, p->tcfp_keys,
+	       p->tcfp_nkeys * sizeof(struct tc_pedit_key));
+	opt->index = p->tcf_index;
+	opt->nkeys = p->tcfp_nkeys;
+	opt->flags = p->tcfp_flags;
+	opt->action = p->tcf_action;
+	opt->refcnt = p->tcf_refcnt - ref;
+	opt->bindcnt = p->tcf_bindcnt - bind;
 
 	RTA_PUT(skb, TCA_PEDIT_PARMS, s, opt);
-	t.install = jiffies_to_clock_t(jiffies - p->tm.install);
-	t.lastuse = jiffies_to_clock_t(jiffies - p->tm.lastuse);
-	t.expires = jiffies_to_clock_t(p->tm.expires);
+	t.install = jiffies_to_clock_t(jiffies - p->tcf_tm.install);
+	t.lastuse = jiffies_to_clock_t(jiffies - p->tcf_tm.lastuse);
+	t.expires = jiffies_to_clock_t(p->tcf_tm.expires);
 	RTA_PUT(skb, TCA_PEDIT_TM, sizeof(t), &t);
 	kfree(opt);
 	return skb->len;
@@ -252,9 +232,9 @@ rtattr_failure:
 	return -1;
 }
 
-static
-struct tc_action_ops act_pedit_ops = {
+static struct tc_action_ops act_pedit_ops = {
 	.kind		=	"pedit",
+	.hinfo		=	&pedit_hash_info,
 	.type		=	TCA_ACT_PEDIT,
 	.capab		=	TCA_CAP_NONE,
 	.owner		=	THIS_MODULE,
@@ -270,14 +250,12 @@ MODULE_AUTHOR("Jamal Hadi Salim(2002-4)");
 MODULE_DESCRIPTION("Generic Packet Editor actions");
 MODULE_LICENSE("GPL");
 
-static int __init
-pedit_init_module(void)
+static int __init pedit_init_module(void)
 {
 	return tcf_register_action(&act_pedit_ops);
 }
 
-static void __exit
-pedit_cleanup_module(void)
+static void __exit pedit_cleanup_module(void)
 {
 	tcf_unregister_action(&act_pedit_ops);
 }
diff --git a/net/sched/act_police.c b/net/sched/act_police.c
index da905d7b4b40..fed47b658837 100644
--- a/net/sched/act_police.c
+++ b/net/sched/act_police.c
@@ -32,43 +32,27 @@
 #include <net/sock.h>
 #include <net/act_api.h>
 
-#define L2T(p,L)   ((p)->R_tab->data[(L)>>(p)->R_tab->rate.cell_log])
-#define L2T_P(p,L) ((p)->P_tab->data[(L)>>(p)->P_tab->rate.cell_log])
-#define PRIV(a) ((struct tcf_police *) (a)->priv)
-
-/* use generic hash table */
-#define MY_TAB_SIZE     16
-#define MY_TAB_MASK     15
-static u32 idx_gen;
-static struct tcf_police *tcf_police_ht[MY_TAB_SIZE];
-/* Policer hash table lock */
-static DEFINE_RWLOCK(police_lock);
-
-/* Each policer is serialized by its individual spinlock */
+#define L2T(p,L)   ((p)->tcfp_R_tab->data[(L)>>(p)->tcfp_R_tab->rate.cell_log])
+#define L2T_P(p,L) ((p)->tcfp_P_tab->data[(L)>>(p)->tcfp_P_tab->rate.cell_log])
 
-static __inline__ unsigned tcf_police_hash(u32 index)
-{
-	return index&0xF;
-}
+#define POL_TAB_MASK     15
+static struct tcf_common *tcf_police_ht[POL_TAB_MASK + 1];
+static u32 police_idx_gen;
+static DEFINE_RWLOCK(police_lock);
 
-static __inline__ struct tcf_police * tcf_police_lookup(u32 index)
-{
-	struct tcf_police *p;
+static struct tcf_hashinfo police_hash_info = {
+	.htab	=	tcf_police_ht,
+	.hmask	=	POL_TAB_MASK,
+	.lock	=	&police_lock,
+};
 
-	read_lock(&police_lock);
-	for (p = tcf_police_ht[tcf_police_hash(index)]; p; p = p->next) {
-		if (p->index == index)
-			break;
-	}
-	read_unlock(&police_lock);
-	return p;
-}
+/* Each policer is serialized by its individual spinlock */
 
 #ifdef CONFIG_NET_CLS_ACT
 static int tcf_act_police_walker(struct sk_buff *skb, struct netlink_callback *cb,
                               int type, struct tc_action *a)
 {
-	struct tcf_police *p;
+	struct tcf_common *p;
 	int err = 0, index = -1, i = 0, s_i = 0, n_i = 0;
 	struct rtattr *r;
 
@@ -76,10 +60,10 @@ static int tcf_act_police_walker(struct sk_buff *skb, struct netlink_callback *c
 
 	s_i = cb->args[0];
 
-	for (i = 0; i < MY_TAB_SIZE; i++) {
-		p = tcf_police_ht[tcf_police_hash(i)];
+	for (i = 0; i < (POL_TAB_MASK + 1); i++) {
+		p = tcf_police_ht[tcf_hash(i, POL_TAB_MASK)];
 
-		for (; p; p = p->next) {
+		for (; p; p = p->tcfc_next) {
 			index++;
 			if (index < s_i)
 				continue;
@@ -110,48 +94,26 @@ rtattr_failure:
 	skb_trim(skb, (u8*)r - skb->data);
 	goto done;
 }
-
-static inline int
-tcf_act_police_hash_search(struct tc_action *a, u32 index)
-{
-	struct tcf_police *p = tcf_police_lookup(index);
-
-	if (p != NULL) {
-		a->priv = p;
-		return 1;
-	} else {
-		return 0;
-	}
-}
 #endif
 
-static inline u32 tcf_police_new_index(void)
-{
-	do {
-		if (++idx_gen == 0)
-			idx_gen = 1;
-	} while (tcf_police_lookup(idx_gen));
-
-	return idx_gen;
-}
-
 void tcf_police_destroy(struct tcf_police *p)
 {
-	unsigned h = tcf_police_hash(p->index);
-	struct tcf_police **p1p;
+	unsigned int h = tcf_hash(p->tcf_index, POL_TAB_MASK);
+	struct tcf_common **p1p;
 	
-	for (p1p = &tcf_police_ht[h]; *p1p; p1p = &(*p1p)->next) {
-		if (*p1p == p) {
+	for (p1p = &tcf_police_ht[h]; *p1p; p1p = &(*p1p)->tcfc_next) {
+		if (*p1p == &p->common) {
 			write_lock_bh(&police_lock);
-			*p1p = p->next;
+			*p1p = p->tcf_next;
 			write_unlock_bh(&police_lock);
 #ifdef CONFIG_NET_ESTIMATOR
-			gen_kill_estimator(&p->bstats, &p->rate_est);
+			gen_kill_estimator(&p->tcf_bstats,
+					   &p->tcf_rate_est);
 #endif
-			if (p->R_tab)
-				qdisc_put_rtab(p->R_tab);
-			if (p->P_tab)
-				qdisc_put_rtab(p->P_tab);
+			if (p->tcfp_R_tab)
+				qdisc_put_rtab(p->tcfp_R_tab);
+			if (p->tcfp_P_tab)
+				qdisc_put_rtab(p->tcfp_P_tab);
 			kfree(p);
 			return;
 		}
@@ -167,7 +129,7 @@ static int tcf_act_police_locate(struct rtattr *rta, struct rtattr *est,
 	int ret = 0, err;
 	struct rtattr *tb[TCA_POLICE_MAX];
 	struct tc_police *parm;
-	struct tcf_police *p;
+	struct tcf_police *police;
 	struct qdisc_rate_table *R_tab = NULL, *P_tab = NULL;
 
 	if (rta == NULL || rtattr_parse_nested(tb, TCA_POLICE_MAX, rta) < 0)
@@ -185,27 +147,32 @@ static int tcf_act_police_locate(struct rtattr *rta, struct rtattr *est,
 	    RTA_PAYLOAD(tb[TCA_POLICE_RESULT-1]) != sizeof(u32))
 		return -EINVAL;
 
-	if (parm->index && (p = tcf_police_lookup(parm->index)) != NULL) {
-		a->priv = p;
-		if (bind) {
-			p->bindcnt += 1;
-			p->refcnt += 1;
+	if (parm->index) {
+		struct tcf_common *pc;
+
+		pc = tcf_hash_lookup(parm->index, &police_hash_info);
+		if (pc != NULL) {
+			a->priv = pc;
+			police = to_police(pc);
+			if (bind) {
+				police->tcf_bindcnt += 1;
+				police->tcf_refcnt += 1;
+			}
+			if (ovr)
+				goto override;
+			return ret;
 		}
-		if (ovr)
-			goto override;
-		return ret;
 	}
 
-	p = kzalloc(sizeof(*p), GFP_KERNEL);
-	if (p == NULL)
+	police = kzalloc(sizeof(*police), GFP_KERNEL);
+	if (police == NULL)
 		return -ENOMEM;
-
 	ret = ACT_P_CREATED;
-	p->refcnt = 1;
-	spin_lock_init(&p->lock);
-	p->stats_lock = &p->lock;
+	police->tcf_refcnt = 1;
+	spin_lock_init(&police->tcf_lock);
+	police->tcf_stats_lock = &police->tcf_lock;
 	if (bind)
-		p->bindcnt = 1;
+		police->tcf_bindcnt = 1;
 override:
 	if (parm->rate.rate) {
 		err = -ENOMEM;
@@ -215,67 +182,71 @@ override:
 		if (parm->peakrate.rate) {
 			P_tab = qdisc_get_rtab(&parm->peakrate,
 					       tb[TCA_POLICE_PEAKRATE-1]);
-			if (p->P_tab == NULL) {
+			if (P_tab == NULL) {
 				qdisc_put_rtab(R_tab);
 				goto failure;
 			}
 		}
 	}
 	/* No failure allowed after this point */
-	spin_lock_bh(&p->lock);
+	spin_lock_bh(&police->tcf_lock);
 	if (R_tab != NULL) {
-		qdisc_put_rtab(p->R_tab);
-		p->R_tab = R_tab;
+		qdisc_put_rtab(police->tcfp_R_tab);
+		police->tcfp_R_tab = R_tab;
 	}
 	if (P_tab != NULL) {
-		qdisc_put_rtab(p->P_tab);
-		p->P_tab = P_tab;
+		qdisc_put_rtab(police->tcfp_P_tab);
+		police->tcfp_P_tab = P_tab;
 	}
 
 	if (tb[TCA_POLICE_RESULT-1])
-		p->result = *(u32*)RTA_DATA(tb[TCA_POLICE_RESULT-1]);
-	p->toks = p->burst = parm->burst;
-	p->mtu = parm->mtu;
-	if (p->mtu == 0) {
-		p->mtu = ~0;
-		if (p->R_tab)
-			p->mtu = 255<<p->R_tab->rate.cell_log;
+		police->tcfp_result = *(u32*)RTA_DATA(tb[TCA_POLICE_RESULT-1]);
+	police->tcfp_toks = police->tcfp_burst = parm->burst;
+	police->tcfp_mtu = parm->mtu;
+	if (police->tcfp_mtu == 0) {
+		police->tcfp_mtu = ~0;
+		if (police->tcfp_R_tab)
+			police->tcfp_mtu = 255<<police->tcfp_R_tab->rate.cell_log;
 	}
-	if (p->P_tab)
-		p->ptoks = L2T_P(p, p->mtu);
-	p->action = parm->action;
+	if (police->tcfp_P_tab)
+		police->tcfp_ptoks = L2T_P(police, police->tcfp_mtu);
+	police->tcf_action = parm->action;
 
 #ifdef CONFIG_NET_ESTIMATOR
 	if (tb[TCA_POLICE_AVRATE-1])
-		p->ewma_rate = *(u32*)RTA_DATA(tb[TCA_POLICE_AVRATE-1]);
+		police->tcfp_ewma_rate =
+			*(u32*)RTA_DATA(tb[TCA_POLICE_AVRATE-1]);
 	if (est)
-		gen_replace_estimator(&p->bstats, &p->rate_est, p->stats_lock, est);
+		gen_replace_estimator(&police->tcf_bstats,
+				      &police->tcf_rate_est,
+				      police->tcf_stats_lock, est);
 #endif
 
-	spin_unlock_bh(&p->lock);
+	spin_unlock_bh(&police->tcf_lock);
 	if (ret != ACT_P_CREATED)
 		return ret;
 
-	PSCHED_GET_TIME(p->t_c);
-	p->index = parm->index ? : tcf_police_new_index();
-	h = tcf_police_hash(p->index);
+	PSCHED_GET_TIME(police->tcfp_t_c);
+	police->tcf_index = parm->index ? parm->index :
+		tcf_hash_new_index(&police_idx_gen, &police_hash_info);
+	h = tcf_hash(police->tcf_index, POL_TAB_MASK);
 	write_lock_bh(&police_lock);
-	p->next = tcf_police_ht[h];
-	tcf_police_ht[h] = p;
+	police->tcf_next = tcf_police_ht[h];
+	tcf_police_ht[h] = &police->common;
 	write_unlock_bh(&police_lock);
 
-	a->priv = p;
+	a->priv = police;
 	return ret;
 
 failure:
 	if (ret == ACT_P_CREATED)
-		kfree(p);
+		kfree(police);
 	return err;
 }
 
 static int tcf_act_police_cleanup(struct tc_action *a, int bind)
 {
-	struct tcf_police *p = PRIV(a);
+	struct tcf_police *p = a->priv;
 
 	if (p != NULL)
 		return tcf_police_release(p, bind);
@@ -285,86 +256,87 @@ static int tcf_act_police_cleanup(struct tc_action *a, int bind)
 static int tcf_act_police(struct sk_buff *skb, struct tc_action *a,
                           struct tcf_result *res)
 {
+	struct tcf_police *police = a->priv;
 	psched_time_t now;
-	struct tcf_police *p = PRIV(a);
 	long toks;
 	long ptoks = 0;
 
-	spin_lock(&p->lock);
+	spin_lock(&police->tcf_lock);
 
-	p->bstats.bytes += skb->len;
-	p->bstats.packets++;
+	police->tcf_bstats.bytes += skb->len;
+	police->tcf_bstats.packets++;
 
 #ifdef CONFIG_NET_ESTIMATOR
-	if (p->ewma_rate && p->rate_est.bps >= p->ewma_rate) {
-		p->qstats.overlimits++;
-		spin_unlock(&p->lock);
-		return p->action;
+	if (police->tcfp_ewma_rate &&
+	    police->tcf_rate_est.bps >= police->tcfp_ewma_rate) {
+		police->tcf_qstats.overlimits++;
+		spin_unlock(&police->tcf_lock);
+		return police->tcf_action;
 	}
 #endif
 
-	if (skb->len <= p->mtu) {
-		if (p->R_tab == NULL) {
-			spin_unlock(&p->lock);
-			return p->result;
+	if (skb->len <= police->tcfp_mtu) {
+		if (police->tcfp_R_tab == NULL) {
+			spin_unlock(&police->tcf_lock);
+			return police->tcfp_result;
 		}
 
 		PSCHED_GET_TIME(now);
 
-		toks = PSCHED_TDIFF_SAFE(now, p->t_c, p->burst);
-
-		if (p->P_tab) {
-			ptoks = toks + p->ptoks;
-			if (ptoks > (long)L2T_P(p, p->mtu))
-				ptoks = (long)L2T_P(p, p->mtu);
-			ptoks -= L2T_P(p, skb->len);
+		toks = PSCHED_TDIFF_SAFE(now, police->tcfp_t_c,
+					 police->tcfp_burst);
+		if (police->tcfp_P_tab) {
+			ptoks = toks + police->tcfp_ptoks;
+			if (ptoks > (long)L2T_P(police, police->tcfp_mtu))
+				ptoks = (long)L2T_P(police, police->tcfp_mtu);
+			ptoks -= L2T_P(police, skb->len);
 		}
-		toks += p->toks;
-		if (toks > (long)p->burst)
-			toks = p->burst;
-		toks -= L2T(p, skb->len);
-
+		toks += police->tcfp_toks;
+		if (toks > (long)police->tcfp_burst)
+			toks = police->tcfp_burst;
+		toks -= L2T(police, skb->len);
 		if ((toks|ptoks) >= 0) {
-			p->t_c = now;
-			p->toks = toks;
-			p->ptoks = ptoks;
-			spin_unlock(&p->lock);
-			return p->result;
+			police->tcfp_t_c = now;
+			police->tcfp_toks = toks;
+			police->tcfp_ptoks = ptoks;
+			spin_unlock(&police->tcf_lock);
+			return police->tcfp_result;
 		}
 	}
 
-	p->qstats.overlimits++;
-	spin_unlock(&p->lock);
-	return p->action;
+	police->tcf_qstats.overlimits++;
+	spin_unlock(&police->tcf_lock);
+	return police->tcf_action;
 }
 
 static int
 tcf_act_police_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref)
 {
 	unsigned char	 *b = skb->tail;
+	struct tcf_police *police = a->priv;
 	struct tc_police opt;
-	struct tcf_police *p = PRIV(a);
-
-	opt.index = p->index;
-	opt.action = p->action;
-	opt.mtu = p->mtu;
-	opt.burst = p->burst;
-	opt.refcnt = p->refcnt - ref;
-	opt.bindcnt = p->bindcnt - bind;
-	if (p->R_tab)
-		opt.rate = p->R_tab->rate;
+
+	opt.index = police->tcf_index;
+	opt.action = police->tcf_action;
+	opt.mtu = police->tcfp_mtu;
+	opt.burst = police->tcfp_burst;
+	opt.refcnt = police->tcf_refcnt - ref;
+	opt.bindcnt = police->tcf_bindcnt - bind;
+	if (police->tcfp_R_tab)
+		opt.rate = police->tcfp_R_tab->rate;
 	else
 		memset(&opt.rate, 0, sizeof(opt.rate));
-	if (p->P_tab)
-		opt.peakrate = p->P_tab->rate;
+	if (police->tcfp_P_tab)
+		opt.peakrate = police->tcfp_P_tab->rate;
 	else
 		memset(&opt.peakrate, 0, sizeof(opt.peakrate));
 	RTA_PUT(skb, TCA_POLICE_TBF, sizeof(opt), &opt);
-	if (p->result)
-		RTA_PUT(skb, TCA_POLICE_RESULT, sizeof(int), &p->result);
+	if (police->tcfp_result)
+		RTA_PUT(skb, TCA_POLICE_RESULT, sizeof(int),
+			&police->tcfp_result);
 #ifdef CONFIG_NET_ESTIMATOR
-	if (p->ewma_rate)
-		RTA_PUT(skb, TCA_POLICE_AVRATE, 4, &p->ewma_rate);
+	if (police->tcfp_ewma_rate)
+		RTA_PUT(skb, TCA_POLICE_AVRATE, 4, &police->tcfp_ewma_rate);
 #endif
 	return skb->len;
 
@@ -379,13 +351,14 @@ MODULE_LICENSE("GPL");
 
 static struct tc_action_ops act_police_ops = {
 	.kind		=	"police",
+	.hinfo		=	&police_hash_info,
 	.type		=	TCA_ID_POLICE,
 	.capab		=	TCA_CAP_NONE,
 	.owner		=	THIS_MODULE,
 	.act		=	tcf_act_police,
 	.dump		=	tcf_act_police_dump,
 	.cleanup	=	tcf_act_police_cleanup,
-	.lookup		=	tcf_act_police_hash_search,
+	.lookup		=	tcf_hash_search,
 	.init		=	tcf_act_police_locate,
 	.walk		=	tcf_act_police_walker
 };
@@ -407,10 +380,39 @@ module_exit(police_cleanup_module);
 
 #else /* CONFIG_NET_CLS_ACT */
 
-struct tcf_police * tcf_police_locate(struct rtattr *rta, struct rtattr *est)
+static struct tcf_common *tcf_police_lookup(u32 index)
 {
-	unsigned h;
-	struct tcf_police *p;
+	struct tcf_hashinfo *hinfo = &police_hash_info;
+	struct tcf_common *p;
+
+	read_lock(hinfo->lock);
+	for (p = hinfo->htab[tcf_hash(index, hinfo->hmask)]; p;
+	     p = p->tcfc_next) {
+		if (p->tcfc_index == index)
+			break;
+	}
+	read_unlock(hinfo->lock);
+
+	return p;
+}
+
+static u32 tcf_police_new_index(void)
+{
+	u32 *idx_gen = &police_idx_gen;
+	u32 val = *idx_gen;
+
+	do {
+		if (++val == 0)
+			val = 1;
+	} while (tcf_police_lookup(val));
+
+	return (*idx_gen = val);
+}
+
+struct tcf_police *tcf_police_locate(struct rtattr *rta, struct rtattr *est)
+{
+	unsigned int h;
+	struct tcf_police *police;
 	struct rtattr *tb[TCA_POLICE_MAX];
 	struct tc_police *parm;
 
@@ -423,149 +425,158 @@ struct tcf_police * tcf_police_locate(struct rtattr *rta, struct rtattr *est)
 
 	parm = RTA_DATA(tb[TCA_POLICE_TBF-1]);
 
-	if (parm->index && (p = tcf_police_lookup(parm->index)) != NULL) {
-		p->refcnt++;
-		return p;
-	}
+	if (parm->index) {
+		struct tcf_common *pc;
 
-	p = kzalloc(sizeof(*p), GFP_KERNEL);
-	if (p == NULL)
+		pc = tcf_police_lookup(parm->index);
+		if (pc) {
+			police = to_police(pc);
+			police->tcf_refcnt++;
+			return police;
+		}
+	}
+	police = kzalloc(sizeof(*police), GFP_KERNEL);
+	if (unlikely(!police))
 		return NULL;
 
-	p->refcnt = 1;
-	spin_lock_init(&p->lock);
-	p->stats_lock = &p->lock;
+	police->tcf_refcnt = 1;
+	spin_lock_init(&police->tcf_lock);
+	police->tcf_stats_lock = &police->tcf_lock;
 	if (parm->rate.rate) {
-		p->R_tab = qdisc_get_rtab(&parm->rate, tb[TCA_POLICE_RATE-1]);
-		if (p->R_tab == NULL)
+		police->tcfp_R_tab =
+			qdisc_get_rtab(&parm->rate, tb[TCA_POLICE_RATE-1]);
+		if (police->tcfp_R_tab == NULL)
 			goto failure;
 		if (parm->peakrate.rate) {
-			p->P_tab = qdisc_get_rtab(&parm->peakrate,
-			                          tb[TCA_POLICE_PEAKRATE-1]);
-			if (p->P_tab == NULL)
+			police->tcfp_P_tab =
+				qdisc_get_rtab(&parm->peakrate,
+					       tb[TCA_POLICE_PEAKRATE-1]);
+			if (police->tcfp_P_tab == NULL)
 				goto failure;
 		}
 	}
 	if (tb[TCA_POLICE_RESULT-1]) {
 		if (RTA_PAYLOAD(tb[TCA_POLICE_RESULT-1]) != sizeof(u32))
 			goto failure;
-		p->result = *(u32*)RTA_DATA(tb[TCA_POLICE_RESULT-1]);
+		police->tcfp_result = *(u32*)RTA_DATA(tb[TCA_POLICE_RESULT-1]);
 	}
 #ifdef CONFIG_NET_ESTIMATOR
 	if (tb[TCA_POLICE_AVRATE-1]) {
 		if (RTA_PAYLOAD(tb[TCA_POLICE_AVRATE-1]) != sizeof(u32))
 			goto failure;
-		p->ewma_rate = *(u32*)RTA_DATA(tb[TCA_POLICE_AVRATE-1]);
+		police->tcfp_ewma_rate =
+			*(u32*)RTA_DATA(tb[TCA_POLICE_AVRATE-1]);
 	}
 #endif
-	p->toks = p->burst = parm->burst;
-	p->mtu = parm->mtu;
-	if (p->mtu == 0) {
-		p->mtu = ~0;
-		if (p->R_tab)
-			p->mtu = 255<<p->R_tab->rate.cell_log;
+	police->tcfp_toks = police->tcfp_burst = parm->burst;
+	police->tcfp_mtu = parm->mtu;
+	if (police->tcfp_mtu == 0) {
+		police->tcfp_mtu = ~0;
+		if (police->tcfp_R_tab)
+			police->tcfp_mtu = 255<<police->tcfp_R_tab->rate.cell_log;
 	}
-	if (p->P_tab)
-		p->ptoks = L2T_P(p, p->mtu);
-	PSCHED_GET_TIME(p->t_c);
-	p->index = parm->index ? : tcf_police_new_index();
-	p->action = parm->action;
+	if (police->tcfp_P_tab)
+		police->tcfp_ptoks = L2T_P(police, police->tcfp_mtu);
+	PSCHED_GET_TIME(police->tcfp_t_c);
+	police->tcf_index = parm->index ? parm->index :
+		tcf_police_new_index();
+	police->tcf_action = parm->action;
 #ifdef CONFIG_NET_ESTIMATOR
 	if (est)
-		gen_new_estimator(&p->bstats, &p->rate_est, p->stats_lock, est);
+		gen_new_estimator(&police->tcf_bstats, &police->tcf_rate_est,
+				  police->tcf_stats_lock, est);
 #endif
-	h = tcf_police_hash(p->index);
+	h = tcf_hash(police->tcf_index, POL_TAB_MASK);
 	write_lock_bh(&police_lock);
-	p->next = tcf_police_ht[h];
-	tcf_police_ht[h] = p;
+	police->tcf_next = tcf_police_ht[h];
+	tcf_police_ht[h] = &police->common;
 	write_unlock_bh(&police_lock);
-	return p;
+	return police;
 
 failure:
-	if (p->R_tab)
-		qdisc_put_rtab(p->R_tab);
-	kfree(p);
+	if (police->tcfp_R_tab)
+		qdisc_put_rtab(police->tcfp_R_tab);
+	kfree(police);
 	return NULL;
 }
 
-int tcf_police(struct sk_buff *skb, struct tcf_police *p)
+int tcf_police(struct sk_buff *skb, struct tcf_police *police)
 {
 	psched_time_t now;
 	long toks;
 	long ptoks = 0;
 
-	spin_lock(&p->lock);
+	spin_lock(&police->tcf_lock);
 
-	p->bstats.bytes += skb->len;
-	p->bstats.packets++;
+	police->tcf_bstats.bytes += skb->len;
+	police->tcf_bstats.packets++;
 
 #ifdef CONFIG_NET_ESTIMATOR
-	if (p->ewma_rate && p->rate_est.bps >= p->ewma_rate) {
-		p->qstats.overlimits++;
-		spin_unlock(&p->lock);
-		return p->action;
+	if (police->tcfp_ewma_rate &&
+	    police->tcf_rate_est.bps >= police->tcfp_ewma_rate) {
+		police->tcf_qstats.overlimits++;
+		spin_unlock(&police->tcf_lock);
+		return police->tcf_action;
 	}
 #endif
-
-	if (skb->len <= p->mtu) {
-		if (p->R_tab == NULL) {
-			spin_unlock(&p->lock);
-			return p->result;
+	if (skb->len <= police->tcfp_mtu) {
+		if (police->tcfp_R_tab == NULL) {
+			spin_unlock(&police->tcf_lock);
+			return police->tcfp_result;
 		}
 
 		PSCHED_GET_TIME(now);
-
-		toks = PSCHED_TDIFF_SAFE(now, p->t_c, p->burst);
-
-		if (p->P_tab) {
-			ptoks = toks + p->ptoks;
-			if (ptoks > (long)L2T_P(p, p->mtu))
-				ptoks = (long)L2T_P(p, p->mtu);
-			ptoks -= L2T_P(p, skb->len);
+		toks = PSCHED_TDIFF_SAFE(now, police->tcfp_t_c,
+					 police->tcfp_burst);
+		if (police->tcfp_P_tab) {
+			ptoks = toks + police->tcfp_ptoks;
+			if (ptoks > (long)L2T_P(police, police->tcfp_mtu))
+				ptoks = (long)L2T_P(police, police->tcfp_mtu);
+			ptoks -= L2T_P(police, skb->len);
 		}
-		toks += p->toks;
-		if (toks > (long)p->burst)
-			toks = p->burst;
-		toks -= L2T(p, skb->len);
-
+		toks += police->tcfp_toks;
+		if (toks > (long)police->tcfp_burst)
+			toks = police->tcfp_burst;
+		toks -= L2T(police, skb->len);
 		if ((toks|ptoks) >= 0) {
-			p->t_c = now;
-			p->toks = toks;
-			p->ptoks = ptoks;
-			spin_unlock(&p->lock);
-			return p->result;
+			police->tcfp_t_c = now;
+			police->tcfp_toks = toks;
+			police->tcfp_ptoks = ptoks;
+			spin_unlock(&police->tcf_lock);
+			return police->tcfp_result;
 		}
 	}
 
-	p->qstats.overlimits++;
-	spin_unlock(&p->lock);
-	return p->action;
+	police->tcf_qstats.overlimits++;
+	spin_unlock(&police->tcf_lock);
+	return police->tcf_action;
 }
 EXPORT_SYMBOL(tcf_police);
 
-int tcf_police_dump(struct sk_buff *skb, struct tcf_police *p)
+int tcf_police_dump(struct sk_buff *skb, struct tcf_police *police)
 {
-	unsigned char	 *b = skb->tail;
+	unsigned char *b = skb->tail;
 	struct tc_police opt;
 
-	opt.index = p->index;
-	opt.action = p->action;
-	opt.mtu = p->mtu;
-	opt.burst = p->burst;
-	if (p->R_tab)
-		opt.rate = p->R_tab->rate;
+	opt.index = police->tcf_index;
+	opt.action = police->tcf_action;
+	opt.mtu = police->tcfp_mtu;
+	opt.burst = police->tcfp_burst;
+	if (police->tcfp_R_tab)
+		opt.rate = police->tcfp_R_tab->rate;
 	else
 		memset(&opt.rate, 0, sizeof(opt.rate));
-	if (p->P_tab)
-		opt.peakrate = p->P_tab->rate;
+	if (police->tcfp_P_tab)
+		opt.peakrate = police->tcfp_P_tab->rate;
 	else
 		memset(&opt.peakrate, 0, sizeof(opt.peakrate));
 	RTA_PUT(skb, TCA_POLICE_TBF, sizeof(opt), &opt);
-	if (p->result)
-		RTA_PUT(skb, TCA_POLICE_RESULT, sizeof(int), &p->result);
+	if (police->tcfp_result)
+		RTA_PUT(skb, TCA_POLICE_RESULT, sizeof(int),
+			&police->tcfp_result);
 #ifdef CONFIG_NET_ESTIMATOR
-	if (p->ewma_rate)
-		RTA_PUT(skb, TCA_POLICE_AVRATE, 4, &p->ewma_rate);
+	if (police->tcfp_ewma_rate)
+		RTA_PUT(skb, TCA_POLICE_AVRATE, 4, &police->tcfp_ewma_rate);
 #endif
 	return skb->len;
 
@@ -574,19 +585,20 @@ rtattr_failure:
 	return -1;
 }
 
-int tcf_police_dump_stats(struct sk_buff *skb, struct tcf_police *p)
+int tcf_police_dump_stats(struct sk_buff *skb, struct tcf_police *police)
 {
 	struct gnet_dump d;
 	
 	if (gnet_stats_start_copy_compat(skb, TCA_STATS2, TCA_STATS,
-			TCA_XSTATS, p->stats_lock, &d) < 0)
+					 TCA_XSTATS, police->tcf_stats_lock,
+					 &d) < 0)
 		goto errout;
 	
-	if (gnet_stats_copy_basic(&d, &p->bstats) < 0 ||
+	if (gnet_stats_copy_basic(&d, &police->tcf_bstats) < 0 ||
 #ifdef CONFIG_NET_ESTIMATOR
-	    gnet_stats_copy_rate_est(&d, &p->rate_est) < 0 ||
+	    gnet_stats_copy_rate_est(&d, &police->tcf_rate_est) < 0 ||
 #endif
-	    gnet_stats_copy_queue(&d, &p->qstats) < 0)
+	    gnet_stats_copy_queue(&d, &police->tcf_qstats) < 0)
 		goto errout;
 
 	if (gnet_stats_finish_copy(&d) < 0)
diff --git a/net/sched/act_simple.c b/net/sched/act_simple.c
index 17105c82537f..8c1ab8ad8fa6 100644
--- a/net/sched/act_simple.c
+++ b/net/sched/act_simple.c
@@ -20,54 +20,175 @@
 
 #define TCA_ACT_SIMP 22
 
-/* XXX: Hide all these common elements under some macro 
- * probably
-*/
 #include <linux/tc_act/tc_defact.h>
 #include <net/tc_act/tc_defact.h>
 
-/* use generic hash table with 8 buckets */
-#define MY_TAB_SIZE     8
-#define MY_TAB_MASK     (MY_TAB_SIZE - 1)
-static u32 idx_gen;
-static struct tcf_defact *tcf_simp_ht[MY_TAB_SIZE];
+#define SIMP_TAB_MASK     7
+static struct tcf_common *tcf_simp_ht[SIMP_TAB_MASK + 1];
+static u32 simp_idx_gen;
 static DEFINE_RWLOCK(simp_lock);
 
-/* override the defaults */
-#define tcf_st		tcf_defact
-#define tc_st		tc_defact
-#define tcf_t_lock	simp_lock
-#define tcf_ht		tcf_simp_ht
-
-#define CONFIG_NET_ACT_INIT 1
-#include <net/pkt_act.h>
-#include <net/act_generic.h>
+struct tcf_hashinfo simp_hash_info = {
+	.htab	=	tcf_simp_ht,
+	.hmask	=	SIMP_TAB_MASK,
+	.lock	=	&simp_lock,
+};
 
 static int tcf_simp(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res)
 {
-	struct tcf_defact *p = PRIV(a, defact);
+	struct tcf_defact *d = a->priv;
 
-	spin_lock(&p->lock);
-	p->tm.lastuse = jiffies;
-	p->bstats.bytes += skb->len;
-	p->bstats.packets++;
+	spin_lock(&d->tcf_lock);
+	d->tcf_tm.lastuse = jiffies;
+	d->tcf_bstats.bytes += skb->len;
+	d->tcf_bstats.packets++;
 
 	/* print policy string followed by _ then packet count 
 	 * Example if this was the 3rd packet and the string was "hello" 
 	 * then it would look like "hello_3" (without quotes) 
 	 **/
-	printk("simple: %s_%d\n", (char *)p->defdata, p->bstats.packets);
-	spin_unlock(&p->lock);
-	return p->action;
+	printk("simple: %s_%d\n",
+	       (char *)d->tcfd_defdata, d->tcf_bstats.packets);
+	spin_unlock(&d->tcf_lock);
+	return d->tcf_action;
+}
+
+static int tcf_simp_release(struct tcf_defact *d, int bind)
+{
+	int ret = 0;
+	if (d) {
+		if (bind)
+			d->tcf_bindcnt--;
+		d->tcf_refcnt--;
+		if (d->tcf_bindcnt <= 0 && d->tcf_refcnt <= 0) {
+			kfree(d->tcfd_defdata);
+			tcf_hash_destroy(&d->common, &simp_hash_info);
+			ret = 1;
+		}
+	}
+	return ret;
+}
+
+static int alloc_defdata(struct tcf_defact *d, u32 datalen, void *defdata)
+{
+	d->tcfd_defdata = kmalloc(datalen, GFP_KERNEL);
+	if (unlikely(!d->tcfd_defdata))
+		return -ENOMEM;
+	d->tcfd_datalen = datalen;
+	memcpy(d->tcfd_defdata, defdata, datalen);
+	return 0;
+}
+
+static int realloc_defdata(struct tcf_defact *d, u32 datalen, void *defdata)
+{
+	kfree(d->tcfd_defdata);
+	return alloc_defdata(d, datalen, defdata);
+}
+
+static int tcf_simp_init(struct rtattr *rta, struct rtattr *est,
+			 struct tc_action *a, int ovr, int bind)
+{
+	struct rtattr *tb[TCA_DEF_MAX];
+	struct tc_defact *parm;
+	struct tcf_defact *d;
+	struct tcf_common *pc;
+	void *defdata;
+	u32 datalen = 0;
+	int ret = 0;
+
+	if (rta == NULL || rtattr_parse_nested(tb, TCA_DEF_MAX, rta) < 0)
+		return -EINVAL;
+
+	if (tb[TCA_DEF_PARMS - 1] == NULL ||
+	    RTA_PAYLOAD(tb[TCA_DEF_PARMS - 1]) < sizeof(*parm))
+		return -EINVAL;
+
+	parm = RTA_DATA(tb[TCA_DEF_PARMS - 1]);
+	defdata = RTA_DATA(tb[TCA_DEF_DATA - 1]);
+	if (defdata == NULL)
+		return -EINVAL;
+
+	datalen = RTA_PAYLOAD(tb[TCA_DEF_DATA - 1]);
+	if (datalen <= 0)
+		return -EINVAL;
+
+	pc = tcf_hash_check(parm->index, a, bind, &simp_hash_info);
+	if (!pc) {
+		pc = tcf_hash_create(parm->index, est, a, sizeof(*d), bind,
+				     &simp_idx_gen, &simp_hash_info);
+		if (unlikely(!pc))
+			return -ENOMEM;
+
+		d = to_defact(pc);
+		ret = alloc_defdata(d, datalen, defdata);
+		if (ret < 0) {
+			kfree(pc);
+			return ret;
+		}
+		ret = ACT_P_CREATED;
+	} else {
+		d = to_defact(pc);
+		if (!ovr) {
+			tcf_simp_release(d, bind);
+			return -EEXIST;
+		}
+		realloc_defdata(d, datalen, defdata);
+	}
+
+	spin_lock_bh(&d->tcf_lock);
+	d->tcf_action = parm->action;
+	spin_unlock_bh(&d->tcf_lock);
+
+	if (ret == ACT_P_CREATED)
+		tcf_hash_insert(pc, &simp_hash_info);
+	return ret;
+}
+
+static inline int tcf_simp_cleanup(struct tc_action *a, int bind)
+{
+	struct tcf_defact *d = a->priv;
+
+	if (d)
+		return tcf_simp_release(d, bind);
+	return 0;
+}
+
+static inline int tcf_simp_dump(struct sk_buff *skb, struct tc_action *a,
+				int bind, int ref)
+{
+	unsigned char *b = skb->tail;
+	struct tcf_defact *d = a->priv;
+	struct tc_defact opt;
+	struct tcf_t t;
+
+	opt.index = d->tcf_index;
+	opt.refcnt = d->tcf_refcnt - ref;
+	opt.bindcnt = d->tcf_bindcnt - bind;
+	opt.action = d->tcf_action;
+	RTA_PUT(skb, TCA_DEF_PARMS, sizeof(opt), &opt);
+	RTA_PUT(skb, TCA_DEF_DATA, d->tcfd_datalen, d->tcfd_defdata);
+	t.install = jiffies_to_clock_t(jiffies - d->tcf_tm.install);
+	t.lastuse = jiffies_to_clock_t(jiffies - d->tcf_tm.lastuse);
+	t.expires = jiffies_to_clock_t(d->tcf_tm.expires);
+	RTA_PUT(skb, TCA_DEF_TM, sizeof(t), &t);
+	return skb->len;
+
+rtattr_failure:
+	skb_trim(skb, b - skb->data);
+	return -1;
 }
 
 static struct tc_action_ops act_simp_ops = {
-	.kind = "simple",
-	.type = TCA_ACT_SIMP,
-	.capab = TCA_CAP_NONE,
-	.owner = THIS_MODULE,
-	.act = tcf_simp,
-	tca_use_default_ops
+	.kind		=	"simple",
+	.hinfo		=	&simp_hash_info,
+	.type		=	TCA_ACT_SIMP,
+	.capab		=	TCA_CAP_NONE,
+	.owner		=	THIS_MODULE,
+	.act		=	tcf_simp,
+	.dump		=	tcf_simp_dump,
+	.cleanup	=	tcf_simp_cleanup,
+	.init		=	tcf_simp_init,
+	.walk		=	tcf_generic_walker,
 };
 
 MODULE_AUTHOR("Jamal Hadi Salim(2005)");
-- 
cgit v1.2.3


From e0a1ad73d34fd6dfdb630479400511e9879069c0 Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Tue, 22 Aug 2006 00:00:21 -0700
Subject: [IPv6] route: Simplify ip6_del_rt()

Provide a simple ip6_del_rt() for the majority of users and
an alternative for the exception via netlink. Avoids code
obfuscation.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip6_route.h |  5 +----
 net/ipv6/addrconf.c     |  6 +++---
 net/ipv6/ndisc.c        |  4 ++--
 net/ipv6/route.c        | 18 ++++++++++++------
 4 files changed, 18 insertions(+), 15 deletions(-)

(limited to 'include')

diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index 01bfe404784f..a7e6086a2bd4 100644
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -69,10 +69,7 @@ extern int			ip6_ins_rt(struct rt6_info *,
 					   struct nlmsghdr *,
 					   void *rtattr,
 					   struct netlink_skb_parms *req);
-extern int			ip6_del_rt(struct rt6_info *,
-					   struct nlmsghdr *,
-					   void *rtattr,
-					   struct netlink_skb_parms *req);
+extern int			ip6_del_rt(struct rt6_info *);
 
 extern int			ip6_rt_addr_add(struct in6_addr *addr,
 						struct net_device *dev,
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index f1ede9004887..27f2e3309598 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -736,7 +736,7 @@ static void ipv6_del_addr(struct inet6_ifaddr *ifp)
 
 		if (rt && ((rt->rt6i_flags & (RTF_GATEWAY | RTF_DEFAULT)) == 0)) {
 			if (onlink == 0) {
-				ip6_del_rt(rt, NULL, NULL, NULL);
+				ip6_del_rt(rt);
 				rt = NULL;
 			} else if (!(rt->rt6i_flags & RTF_EXPIRES)) {
 				rt->rt6i_expires = expires;
@@ -1662,7 +1662,7 @@ void addrconf_prefix_rcv(struct net_device *dev, u8 *opt, int len)
 		if (rt && ((rt->rt6i_flags & (RTF_GATEWAY | RTF_DEFAULT)) == 0)) {
 			if (rt->rt6i_flags&RTF_EXPIRES) {
 				if (valid_lft == 0) {
-					ip6_del_rt(rt, NULL, NULL, NULL);
+					ip6_del_rt(rt);
 					rt = NULL;
 				} else {
 					rt->rt6i_expires = jiffies + rt_expires;
@@ -3557,7 +3557,7 @@ static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
 			addrconf_leave_anycast(ifp);
 		addrconf_leave_solict(ifp->idev, &ifp->addr);
 		dst_hold(&ifp->rt->u.dst);
-		if (ip6_del_rt(ifp->rt, NULL, NULL, NULL))
+		if (ip6_del_rt(ifp->rt))
 			dst_free(&ifp->rt->u.dst);
 		break;
 	}
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index 5743e8bffefd..419d65163819 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -961,7 +961,7 @@ static void ndisc_recv_na(struct sk_buff *skb)
 			struct rt6_info *rt;
 			rt = rt6_get_dflt_router(saddr, dev);
 			if (rt)
-				ip6_del_rt(rt, NULL, NULL, NULL);
+				ip6_del_rt(rt);
 		}
 
 out:
@@ -1114,7 +1114,7 @@ static void ndisc_router_discovery(struct sk_buff *skb)
 
 	if (rt && lifetime == 0) {
 		neigh_clone(neigh);
-		ip6_del_rt(rt, NULL, NULL, NULL);
+		ip6_del_rt(rt);
 		rt = NULL;
 	}
 
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 1aca787ead85..8d511de0db1b 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -457,7 +457,7 @@ int rt6_route_rcv(struct net_device *dev, u8 *opt, int len,
 	rt = rt6_get_route_info(prefix, rinfo->prefix_len, gwaddr, dev->ifindex);
 
 	if (rt && !lifetime) {
-		ip6_del_rt(rt, NULL, NULL, NULL);
+		ip6_del_rt(rt);
 		rt = NULL;
 	}
 
@@ -813,7 +813,7 @@ static struct dst_entry *ip6_negative_advice(struct dst_entry *dst)
 
 	if (rt) {
 		if (rt->rt6i_flags & RTF_CACHE)
-			ip6_del_rt(rt, NULL, NULL, NULL);
+			ip6_del_rt(rt);
 		else
 			dst_release(dst);
 	}
@@ -1218,7 +1218,8 @@ out:
 	return err;
 }
 
-int ip6_del_rt(struct rt6_info *rt, struct nlmsghdr *nlh, void *_rtattr, struct netlink_skb_parms *req)
+static int __ip6_del_rt(struct rt6_info *rt, struct nlmsghdr *nlh,
+			void *_rtattr, struct netlink_skb_parms *req)
 {
 	int err;
 	struct fib6_table *table;
@@ -1237,6 +1238,11 @@ int ip6_del_rt(struct rt6_info *rt, struct nlmsghdr *nlh, void *_rtattr, struct
 	return err;
 }
 
+int ip6_del_rt(struct rt6_info *rt)
+{
+	return __ip6_del_rt(rt, NULL, NULL, NULL);
+}
+
 static int ip6_route_del(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh,
 			 void *_rtattr, struct netlink_skb_parms *req,
 			 u32 table_id)
@@ -1271,7 +1277,7 @@ static int ip6_route_del(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh,
 			dst_hold(&rt->u.dst);
 			read_unlock_bh(&table->tb6_lock);
 
-			return ip6_del_rt(rt, nlh, _rtattr, req);
+			return __ip6_del_rt(rt, nlh, _rtattr, req);
 		}
 	}
 	read_unlock_bh(&table->tb6_lock);
@@ -1395,7 +1401,7 @@ restart:
 	call_netevent_notifiers(NETEVENT_REDIRECT, &netevent);
 
 	if (rt->rt6i_flags&RTF_CACHE) {
-		ip6_del_rt(rt, NULL, NULL, NULL);
+		ip6_del_rt(rt);
 		return;
 	}
 
@@ -1631,7 +1637,7 @@ restart:
 		if (rt->rt6i_flags & (RTF_DEFAULT | RTF_ADDRCONF)) {
 			dst_hold(&rt->u.dst);
 			read_unlock_bh(&table->tb6_lock);
-			ip6_del_rt(rt, NULL, NULL, NULL);
+			ip6_del_rt(rt);
 			goto restart;
 		}
 	}
-- 
cgit v1.2.3


From 40e22e8f3d4d4f1ff68fb03683f007c53ee8b348 Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Tue, 22 Aug 2006 00:00:45 -0700
Subject: [IPv6] route: Simplify ip6_ins_rt()

Provide a simple ip6_ins_rt() for the majority of users and
an alternative for the exception via netlink. Avoids code
obfuscation.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip6_route.h |  5 +----
 net/ipv6/addrconf.c     |  2 +-
 net/ipv6/anycast.c      |  2 +-
 net/ipv6/route.c        | 19 ++++++++++++-------
 4 files changed, 15 insertions(+), 13 deletions(-)

(limited to 'include')

diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index a7e6086a2bd4..172c4761e2bf 100644
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -65,10 +65,7 @@ extern int			ip6_route_add(struct in6_rtmsg *rtmsg,
 					      void *rtattr,
 					      struct netlink_skb_parms *req,
 					      u32 table_id);
-extern int			ip6_ins_rt(struct rt6_info *,
-					   struct nlmsghdr *,
-					   void *rtattr,
-					   struct netlink_skb_parms *req);
+extern int			ip6_ins_rt(struct rt6_info *);
 extern int			ip6_del_rt(struct rt6_info *);
 
 extern int			ip6_rt_addr_add(struct in6_addr *addr,
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 27f2e3309598..aafba9ea9cb6 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -3548,7 +3548,7 @@ static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
 
 	switch (event) {
 	case RTM_NEWADDR:
-		ip6_ins_rt(ifp->rt, NULL, NULL, NULL);
+		ip6_ins_rt(ifp->rt);
 		if (ifp->idev->cnf.forwarding)
 			addrconf_join_anycast(ifp);
 		break;
diff --git a/net/ipv6/anycast.c b/net/ipv6/anycast.c
index f6881d7a0385..abbc35a13e08 100644
--- a/net/ipv6/anycast.c
+++ b/net/ipv6/anycast.c
@@ -335,7 +335,7 @@ int ipv6_dev_ac_inc(struct net_device *dev, struct in6_addr *addr)
 	write_unlock_bh(&idev->lock);
 
 	dst_hold(&rt->u.dst);
-	if (ip6_ins_rt(rt, NULL, NULL, NULL))
+	if (ip6_ins_rt(rt))
 		dst_release(&rt->u.dst);
 
 	addrconf_join_solict(dev, &aca->aca_addr);
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 8d511de0db1b..9ec348a72a95 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -546,8 +546,8 @@ struct rt6_info *rt6_lookup(struct in6_addr *daddr, struct in6_addr *saddr,
    be destroyed.
  */
 
-int ip6_ins_rt(struct rt6_info *rt, struct nlmsghdr *nlh,
-		void *_rtattr, struct netlink_skb_parms *req)
+static int __ip6_ins_rt(struct rt6_info *rt, struct nlmsghdr *nlh,
+			void *_rtattr, struct netlink_skb_parms *req)
 {
 	int err;
 	struct fib6_table *table;
@@ -560,6 +560,11 @@ int ip6_ins_rt(struct rt6_info *rt, struct nlmsghdr *nlh,
 	return err;
 }
 
+int ip6_ins_rt(struct rt6_info *rt)
+{
+	return __ip6_ins_rt(rt, NULL, NULL, NULL);
+}
+
 static struct rt6_info *rt6_alloc_cow(struct rt6_info *ort, struct in6_addr *daddr,
 				      struct in6_addr *saddr)
 {
@@ -657,7 +662,7 @@ restart:
 
 	dst_hold(&rt->u.dst);
 	if (nrt) {
-		err = ip6_ins_rt(nrt, NULL, NULL, NULL);
+		err = ip6_ins_rt(nrt);
 		if (!err)
 			goto out2;
 	}
@@ -752,7 +757,7 @@ restart:
 
 	dst_hold(&rt->u.dst);
 	if (nrt) {
-		err = ip6_ins_rt(nrt, NULL, NULL, NULL);
+		err = ip6_ins_rt(nrt);
 		if (!err)
 			goto out2;
 	}
@@ -1206,7 +1211,7 @@ install_route:
 	rt->u.dst.dev = dev;
 	rt->rt6i_idev = idev;
 	rt->rt6i_table = table;
-	return ip6_ins_rt(rt, nlh, _rtattr, req);
+	return __ip6_ins_rt(rt, nlh, _rtattr, req);
 
 out:
 	if (dev)
@@ -1393,7 +1398,7 @@ restart:
 	nrt->u.dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(neigh->dev);
 	nrt->u.dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(dst_mtu(&nrt->u.dst));
 
-	if (ip6_ins_rt(nrt, NULL, NULL, NULL))
+	if (ip6_ins_rt(nrt))
 		goto out;
 
 	netevent.old = &rt->u.dst;
@@ -1483,7 +1488,7 @@ void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr,
 		dst_set_expires(&nrt->u.dst, ip6_rt_mtu_expires);
 		nrt->rt6i_flags |= RTF_DYNAMIC|RTF_EXPIRES;
 
-		ip6_ins_rt(nrt, NULL, NULL, NULL);
+		ip6_ins_rt(nrt);
 	}
 out:
 	dst_release(&rt->u.dst);
-- 
cgit v1.2.3


From 86872cb57925c46a6499887d77afb880a892c0ec Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Tue, 22 Aug 2006 00:01:08 -0700
Subject: [IPv6] route: FIB6 configuration using struct fib6_config

Replaces the struct in6_rtmsg based interface orignating from
the ioctl interface with a struct fib6_config based on. Allows
changing the interface without breaking the ioctl interface
and avoids passing on tons of parameters.

The recently introduced struct nl_info is used to pass on
netlink authorship information for notifications.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip6_fib.h   |  38 ++++--
 include/net/ip6_route.h |   6 +-
 net/ipv6/addrconf.c     |  65 +++++-----
 net/ipv6/ip6_fib.c      |  19 ++-
 net/ipv6/route.c        | 331 ++++++++++++++++++++++++++++--------------------
 5 files changed, 259 insertions(+), 200 deletions(-)

(limited to 'include')

diff --git a/include/net/ip6_fib.h b/include/net/ip6_fib.h
index 69c444209781..9610b887ffb5 100644
--- a/include/net/ip6_fib.h
+++ b/include/net/ip6_fib.h
@@ -16,14 +16,35 @@
 #ifdef __KERNEL__
 
 #include <linux/ipv6_route.h>
-
-#include <net/dst.h>
-#include <net/flow.h>
 #include <linux/rtnetlink.h>
 #include <linux/spinlock.h>
+#include <net/dst.h>
+#include <net/flow.h>
+#include <net/netlink.h>
 
 struct rt6_info;
 
+struct fib6_config
+{
+	u32		fc_table;
+	u32		fc_metric;
+	int		fc_dst_len;
+	int		fc_src_len;
+	int		fc_ifindex;
+	u32		fc_flags;
+	u32		fc_protocol;
+
+	struct in6_addr	fc_dst;
+	struct in6_addr	fc_src;
+	struct in6_addr	fc_gateway;
+
+	unsigned long	fc_expires;
+	struct nlattr	*fc_mx;
+	int		fc_mx_len;
+
+	struct nl_info	fc_nlinfo;
+};
+
 struct fib6_node
 {
 	struct fib6_node	*parent;
@@ -175,18 +196,13 @@ extern void			fib6_clean_all(int (*func)(struct rt6_info *, void *arg),
 
 extern int			fib6_add(struct fib6_node *root,
 					 struct rt6_info *rt,
-					 struct nlmsghdr *nlh,
-					 void *rtattr,
-					 struct netlink_skb_parms *req);
+					 struct nl_info *info);
 
 extern int			fib6_del(struct rt6_info *rt,
-					 struct nlmsghdr *nlh,
-					 void *rtattr,
-					 struct netlink_skb_parms *req);
+					 struct nl_info *info);
 
 extern void			inet6_rt_notify(int event, struct rt6_info *rt,
-						struct nlmsghdr *nlh,
-						struct netlink_skb_parms *req);
+						struct nl_info *info);
 
 extern void			fib6_run_gc(unsigned long dummy);
 
diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index 172c4761e2bf..3f170f667c7b 100644
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -60,11 +60,7 @@ extern void			ip6_route_cleanup(void);
 
 extern int			ipv6_route_ioctl(unsigned int cmd, void __user *arg);
 
-extern int			ip6_route_add(struct in6_rtmsg *rtmsg,
-					      struct nlmsghdr *,
-					      void *rtattr,
-					      struct netlink_skb_parms *req,
-					      u32 table_id);
+extern int			ip6_route_add(struct fib6_config *cfg);
 extern int			ip6_ins_rt(struct rt6_info *);
 extern int			ip6_del_rt(struct rt6_info *);
 
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index aafba9ea9cb6..fc9cff3426c4 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -1509,59 +1509,56 @@ static void
 addrconf_prefix_route(struct in6_addr *pfx, int plen, struct net_device *dev,
 		      unsigned long expires, u32 flags)
 {
-	struct in6_rtmsg rtmsg;
+	struct fib6_config cfg = {
+		.fc_table = RT6_TABLE_PREFIX,
+		.fc_metric = IP6_RT_PRIO_ADDRCONF,
+		.fc_ifindex = dev->ifindex,
+		.fc_expires = expires,
+		.fc_dst_len = plen,
+		.fc_flags = RTF_UP | flags,
+	};
 
-	memset(&rtmsg, 0, sizeof(rtmsg));
-	ipv6_addr_copy(&rtmsg.rtmsg_dst, pfx);
-	rtmsg.rtmsg_dst_len = plen;
-	rtmsg.rtmsg_metric = IP6_RT_PRIO_ADDRCONF;
-	rtmsg.rtmsg_ifindex = dev->ifindex;
-	rtmsg.rtmsg_info = expires;
-	rtmsg.rtmsg_flags = RTF_UP|flags;
-	rtmsg.rtmsg_type = RTMSG_NEWROUTE;
+	ipv6_addr_copy(&cfg.fc_dst, pfx);
 
 	/* Prevent useless cloning on PtP SIT.
 	   This thing is done here expecting that the whole
 	   class of non-broadcast devices need not cloning.
 	 */
-	if (dev->type == ARPHRD_SIT && (dev->flags&IFF_POINTOPOINT))
-		rtmsg.rtmsg_flags |= RTF_NONEXTHOP;
+	if (dev->type == ARPHRD_SIT && (dev->flags & IFF_POINTOPOINT))
+		cfg.fc_flags |= RTF_NONEXTHOP;
 
-	ip6_route_add(&rtmsg, NULL, NULL, NULL, RT6_TABLE_PREFIX);
+	ip6_route_add(&cfg);
 }
 
 /* Create "default" multicast route to the interface */
 
 static void addrconf_add_mroute(struct net_device *dev)
 {
-	struct in6_rtmsg rtmsg;
-
-	memset(&rtmsg, 0, sizeof(rtmsg));
-	ipv6_addr_set(&rtmsg.rtmsg_dst,
-		      htonl(0xFF000000), 0, 0, 0);
-	rtmsg.rtmsg_dst_len = 8;
-	rtmsg.rtmsg_metric = IP6_RT_PRIO_ADDRCONF;
-	rtmsg.rtmsg_ifindex = dev->ifindex;
-	rtmsg.rtmsg_flags = RTF_UP;
-	rtmsg.rtmsg_type = RTMSG_NEWROUTE;
-	ip6_route_add(&rtmsg, NULL, NULL, NULL, RT6_TABLE_LOCAL);
+	struct fib6_config cfg = {
+		.fc_table = RT6_TABLE_LOCAL,
+		.fc_metric = IP6_RT_PRIO_ADDRCONF,
+		.fc_ifindex = dev->ifindex,
+		.fc_dst_len = 8,
+		.fc_flags = RTF_UP,
+	};
+
+	ipv6_addr_set(&cfg.fc_dst, htonl(0xFF000000), 0, 0, 0);
+
+	ip6_route_add(&cfg);
 }
 
 static void sit_route_add(struct net_device *dev)
 {
-	struct in6_rtmsg rtmsg;
-
-	memset(&rtmsg, 0, sizeof(rtmsg));
-
-	rtmsg.rtmsg_type	= RTMSG_NEWROUTE;
-	rtmsg.rtmsg_metric	= IP6_RT_PRIO_ADDRCONF;
+	struct fib6_config cfg = {
+		.fc_table = RT6_TABLE_MAIN,
+		.fc_metric = IP6_RT_PRIO_ADDRCONF,
+		.fc_ifindex = dev->ifindex,
+		.fc_dst_len = 96,
+		.fc_flags = RTF_UP | RTF_NONEXTHOP,
+	};
 
 	/* prefix length - 96 bits "::d.d.d.d" */
-	rtmsg.rtmsg_dst_len	= 96;
-	rtmsg.rtmsg_flags	= RTF_UP|RTF_NONEXTHOP;
-	rtmsg.rtmsg_ifindex	= dev->ifindex;
-
-	ip6_route_add(&rtmsg, NULL, NULL, NULL, RT6_TABLE_MAIN);
+	ip6_route_add(&cfg);
 }
 
 static void addrconf_add_lroute(struct net_device *dev)
diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
index be36f4acda94..667b1b1ea25d 100644
--- a/net/ipv6/ip6_fib.c
+++ b/net/ipv6/ip6_fib.c
@@ -610,7 +610,7 @@ insert_above:
  */
 
 static int fib6_add_rt2node(struct fib6_node *fn, struct rt6_info *rt,
-		struct nlmsghdr *nlh,  struct netlink_skb_parms *req)
+			    struct nl_info *info)
 {
 	struct rt6_info *iter = NULL;
 	struct rt6_info **ins;
@@ -665,7 +665,7 @@ out:
 	*ins = rt;
 	rt->rt6i_node = fn;
 	atomic_inc(&rt->rt6i_ref);
-	inet6_rt_notify(RTM_NEWROUTE, rt, nlh, req);
+	inet6_rt_notify(RTM_NEWROUTE, rt, info);
 	rt6_stats.fib_rt_entries++;
 
 	if ((fn->fn_flags & RTN_RTINFO) == 0) {
@@ -695,8 +695,7 @@ void fib6_force_start_gc(void)
  *	with source addr info in sub-trees
  */
 
-int fib6_add(struct fib6_node *root, struct rt6_info *rt, 
-		struct nlmsghdr *nlh, void *_rtattr, struct netlink_skb_parms *req)
+int fib6_add(struct fib6_node *root, struct rt6_info *rt, struct nl_info *info)
 {
 	struct fib6_node *fn;
 	int err = -ENOMEM;
@@ -769,7 +768,7 @@ int fib6_add(struct fib6_node *root, struct rt6_info *rt,
 	}
 #endif
 
-	err = fib6_add_rt2node(fn, rt, nlh, req);
+	err = fib6_add_rt2node(fn, rt, info);
 
 	if (err == 0) {
 		fib6_start_gc(rt);
@@ -1076,7 +1075,7 @@ static struct fib6_node * fib6_repair_tree(struct fib6_node *fn)
 }
 
 static void fib6_del_route(struct fib6_node *fn, struct rt6_info **rtp,
-    struct nlmsghdr *nlh, void *_rtattr, struct netlink_skb_parms *req)
+			   struct nl_info *info)
 {
 	struct fib6_walker_t *w;
 	struct rt6_info *rt = *rtp;
@@ -1132,11 +1131,11 @@ static void fib6_del_route(struct fib6_node *fn, struct rt6_info **rtp,
 		if (atomic_read(&rt->rt6i_ref) != 1) BUG();
 	}
 
-	inet6_rt_notify(RTM_DELROUTE, rt, nlh, req);
+	inet6_rt_notify(RTM_DELROUTE, rt, info);
 	rt6_release(rt);
 }
 
-int fib6_del(struct rt6_info *rt, struct nlmsghdr *nlh, void *_rtattr, struct netlink_skb_parms *req)
+int fib6_del(struct rt6_info *rt, struct nl_info *info)
 {
 	struct fib6_node *fn = rt->rt6i_node;
 	struct rt6_info **rtp;
@@ -1161,7 +1160,7 @@ int fib6_del(struct rt6_info *rt, struct nlmsghdr *nlh, void *_rtattr, struct ne
 
 	for (rtp = &fn->leaf; *rtp; rtp = &(*rtp)->u.next) {
 		if (*rtp == rt) {
-			fib6_del_route(fn, rtp, nlh, _rtattr, req);
+			fib6_del_route(fn, rtp, info);
 			return 0;
 		}
 	}
@@ -1290,7 +1289,7 @@ static int fib6_clean_node(struct fib6_walker_t *w)
 		res = c->func(rt, c->arg);
 		if (res < 0) {
 			w->leaf = rt;
-			res = fib6_del(rt, NULL, NULL, NULL);
+			res = fib6_del(rt, NULL);
 			if (res) {
 #if RT6_DEBUG >= 2
 				printk(KERN_DEBUG "fib6_clean_node: del failed: rt=%p@%p err=%d\n", rt, rt->rt6i_node, res);
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 9ec348a72a95..7bcffa6ddba3 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -546,15 +546,14 @@ struct rt6_info *rt6_lookup(struct in6_addr *daddr, struct in6_addr *saddr,
    be destroyed.
  */
 
-static int __ip6_ins_rt(struct rt6_info *rt, struct nlmsghdr *nlh,
-			void *_rtattr, struct netlink_skb_parms *req)
+static int __ip6_ins_rt(struct rt6_info *rt, struct nl_info *info)
 {
 	int err;
 	struct fib6_table *table;
 
 	table = rt->rt6i_table;
 	write_lock_bh(&table->tb6_lock);
-	err = fib6_add(&table->tb6_root, rt, nlh, _rtattr, req);
+	err = fib6_add(&table->tb6_root, rt, info);
 	write_unlock_bh(&table->tb6_lock);
 
 	return err;
@@ -562,7 +561,7 @@ static int __ip6_ins_rt(struct rt6_info *rt, struct nlmsghdr *nlh,
 
 int ip6_ins_rt(struct rt6_info *rt)
 {
-	return __ip6_ins_rt(rt, NULL, NULL, NULL);
+	return __ip6_ins_rt(rt, NULL);
 }
 
 static struct rt6_info *rt6_alloc_cow(struct rt6_info *ort, struct in6_addr *daddr,
@@ -1014,30 +1013,24 @@ int ipv6_get_hoplimit(struct net_device *dev)
  *
  */
 
-int ip6_route_add(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh, 
-		  void *_rtattr, struct netlink_skb_parms *req,
-		  u32 table_id)
+int ip6_route_add(struct fib6_config *cfg)
 {
 	int err;
-	struct rtmsg *r;
-	struct rtattr **rta;
 	struct rt6_info *rt = NULL;
 	struct net_device *dev = NULL;
 	struct inet6_dev *idev = NULL;
 	struct fib6_table *table;
 	int addr_type;
 
-	rta = (struct rtattr **) _rtattr;
-
-	if (rtmsg->rtmsg_dst_len > 128 || rtmsg->rtmsg_src_len > 128)
+	if (cfg->fc_dst_len > 128 || cfg->fc_src_len > 128)
 		return -EINVAL;
 #ifndef CONFIG_IPV6_SUBTREES
-	if (rtmsg->rtmsg_src_len)
+	if (cfg->fc_src_len)
 		return -EINVAL;
 #endif
-	if (rtmsg->rtmsg_ifindex) {
+	if (cfg->fc_ifindex) {
 		err = -ENODEV;
-		dev = dev_get_by_index(rtmsg->rtmsg_ifindex);
+		dev = dev_get_by_index(cfg->fc_ifindex);
 		if (!dev)
 			goto out;
 		idev = in6_dev_get(dev);
@@ -1045,10 +1038,10 @@ int ip6_route_add(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh,
 			goto out;
 	}
 
-	if (rtmsg->rtmsg_metric == 0)
-		rtmsg->rtmsg_metric = IP6_RT_PRIO_USER;
+	if (cfg->fc_metric == 0)
+		cfg->fc_metric = IP6_RT_PRIO_USER;
 
-	table = fib6_new_table(table_id);
+	table = fib6_new_table(cfg->fc_table);
 	if (table == NULL) {
 		err = -ENOBUFS;
 		goto out;
@@ -1062,14 +1055,13 @@ int ip6_route_add(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh,
 	}
 
 	rt->u.dst.obsolete = -1;
-	rt->rt6i_expires = jiffies + clock_t_to_jiffies(rtmsg->rtmsg_info);
-	if (nlh && (r = NLMSG_DATA(nlh))) {
-		rt->rt6i_protocol = r->rtm_protocol;
-	} else {
-		rt->rt6i_protocol = RTPROT_BOOT;
-	}
+	rt->rt6i_expires = jiffies + clock_t_to_jiffies(cfg->fc_expires);
+
+	if (cfg->fc_protocol == RTPROT_UNSPEC)
+		cfg->fc_protocol = RTPROT_BOOT;
+	rt->rt6i_protocol = cfg->fc_protocol;
 
-	addr_type = ipv6_addr_type(&rtmsg->rtmsg_dst);
+	addr_type = ipv6_addr_type(&cfg->fc_dst);
 
 	if (addr_type & IPV6_ADDR_MULTICAST)
 		rt->u.dst.input = ip6_mc_input;
@@ -1078,24 +1070,22 @@ int ip6_route_add(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh,
 
 	rt->u.dst.output = ip6_output;
 
-	ipv6_addr_prefix(&rt->rt6i_dst.addr, 
-			 &rtmsg->rtmsg_dst, rtmsg->rtmsg_dst_len);
-	rt->rt6i_dst.plen = rtmsg->rtmsg_dst_len;
+	ipv6_addr_prefix(&rt->rt6i_dst.addr, &cfg->fc_dst, cfg->fc_dst_len);
+	rt->rt6i_dst.plen = cfg->fc_dst_len;
 	if (rt->rt6i_dst.plen == 128)
 	       rt->u.dst.flags = DST_HOST;
 
 #ifdef CONFIG_IPV6_SUBTREES
-	ipv6_addr_prefix(&rt->rt6i_src.addr, 
-			 &rtmsg->rtmsg_src, rtmsg->rtmsg_src_len);
-	rt->rt6i_src.plen = rtmsg->rtmsg_src_len;
+	ipv6_addr_prefix(&rt->rt6i_src.addr, &cfg->fc_src, cfg->fc_src_len);
+	rt->rt6i_src.plen = cfg->fc_src_len;
 #endif
 
-	rt->rt6i_metric = rtmsg->rtmsg_metric;
+	rt->rt6i_metric = cfg->fc_metric;
 
 	/* We cannot add true routes via loopback here,
 	   they would result in kernel looping; promote them to reject routes
 	 */
-	if ((rtmsg->rtmsg_flags&RTF_REJECT) ||
+	if ((cfg->fc_flags & RTF_REJECT) ||
 	    (dev && (dev->flags&IFF_LOOPBACK) && !(addr_type&IPV6_ADDR_LOOPBACK))) {
 		/* hold loopback dev/idev if we haven't done so. */
 		if (dev != &loopback_dev) {
@@ -1118,12 +1108,12 @@ int ip6_route_add(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh,
 		goto install_route;
 	}
 
-	if (rtmsg->rtmsg_flags & RTF_GATEWAY) {
+	if (cfg->fc_flags & RTF_GATEWAY) {
 		struct in6_addr *gw_addr;
 		int gwa_type;
 
-		gw_addr = &rtmsg->rtmsg_gateway;
-		ipv6_addr_copy(&rt->rt6i_gateway, &rtmsg->rtmsg_gateway);
+		gw_addr = &cfg->fc_gateway;
+		ipv6_addr_copy(&rt->rt6i_gateway, gw_addr);
 		gwa_type = ipv6_addr_type(gw_addr);
 
 		if (gwa_type != (IPV6_ADDR_LINKLOCAL|IPV6_ADDR_UNICAST)) {
@@ -1140,7 +1130,7 @@ int ip6_route_add(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh,
 			if (!(gwa_type&IPV6_ADDR_UNICAST))
 				goto out;
 
-			grt = rt6_lookup(gw_addr, NULL, rtmsg->rtmsg_ifindex, 1);
+			grt = rt6_lookup(gw_addr, NULL, cfg->fc_ifindex, 1);
 
 			err = -EHOSTUNREACH;
 			if (grt == NULL)
@@ -1172,7 +1162,7 @@ int ip6_route_add(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh,
 	if (dev == NULL)
 		goto out;
 
-	if (rtmsg->rtmsg_flags & (RTF_GATEWAY|RTF_NONEXTHOP)) {
+	if (cfg->fc_flags & (RTF_GATEWAY | RTF_NONEXTHOP)) {
 		rt->rt6i_nexthop = __neigh_lookup_errno(&nd_tbl, &rt->rt6i_gateway, dev);
 		if (IS_ERR(rt->rt6i_nexthop)) {
 			err = PTR_ERR(rt->rt6i_nexthop);
@@ -1181,24 +1171,24 @@ int ip6_route_add(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh,
 		}
 	}
 
-	rt->rt6i_flags = rtmsg->rtmsg_flags;
+	rt->rt6i_flags = cfg->fc_flags;
 
 install_route:
-	if (rta && rta[RTA_METRICS-1]) {
-		int attrlen = RTA_PAYLOAD(rta[RTA_METRICS-1]);
-		struct rtattr *attr = RTA_DATA(rta[RTA_METRICS-1]);
-
-		while (RTA_OK(attr, attrlen)) {
-			unsigned flavor = attr->rta_type;
-			if (flavor) {
-				if (flavor > RTAX_MAX) {
+	if (cfg->fc_mx) {
+		struct nlattr *nla;
+		int remaining;
+
+		nla_for_each_attr(nla, cfg->fc_mx, cfg->fc_mx_len, remaining) {
+			int type = nla->nla_type;
+
+			if (type) {
+				if (type > RTAX_MAX) {
 					err = -EINVAL;
 					goto out;
 				}
-				rt->u.dst.metrics[flavor-1] =
-					*(u32 *)RTA_DATA(attr);
+
+				rt->u.dst.metrics[type - 1] = nla_get_u32(nla);
 			}
-			attr = RTA_NEXT(attr, attrlen);
 		}
 	}
 
@@ -1211,7 +1201,7 @@ install_route:
 	rt->u.dst.dev = dev;
 	rt->rt6i_idev = idev;
 	rt->rt6i_table = table;
-	return __ip6_ins_rt(rt, nlh, _rtattr, req);
+	return __ip6_ins_rt(rt, &cfg->fc_nlinfo);
 
 out:
 	if (dev)
@@ -1223,8 +1213,7 @@ out:
 	return err;
 }
 
-static int __ip6_del_rt(struct rt6_info *rt, struct nlmsghdr *nlh,
-			void *_rtattr, struct netlink_skb_parms *req)
+static int __ip6_del_rt(struct rt6_info *rt, struct nl_info *info)
 {
 	int err;
 	struct fib6_table *table;
@@ -1235,7 +1224,7 @@ static int __ip6_del_rt(struct rt6_info *rt, struct nlmsghdr *nlh,
 	table = rt->rt6i_table;
 	write_lock_bh(&table->tb6_lock);
 
-	err = fib6_del(rt, nlh, _rtattr, req);
+	err = fib6_del(rt, info);
 	dst_release(&rt->u.dst);
 
 	write_unlock_bh(&table->tb6_lock);
@@ -1245,44 +1234,41 @@ static int __ip6_del_rt(struct rt6_info *rt, struct nlmsghdr *nlh,
 
 int ip6_del_rt(struct rt6_info *rt)
 {
-	return __ip6_del_rt(rt, NULL, NULL, NULL);
+	return __ip6_del_rt(rt, NULL);
 }
 
-static int ip6_route_del(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh,
-			 void *_rtattr, struct netlink_skb_parms *req,
-			 u32 table_id)
+static int ip6_route_del(struct fib6_config *cfg)
 {
 	struct fib6_table *table;
 	struct fib6_node *fn;
 	struct rt6_info *rt;
 	int err = -ESRCH;
 
-	table = fib6_get_table(table_id);
+	table = fib6_get_table(cfg->fc_table);
 	if (table == NULL)
 		return err;
 
 	read_lock_bh(&table->tb6_lock);
 
 	fn = fib6_locate(&table->tb6_root,
-			 &rtmsg->rtmsg_dst, rtmsg->rtmsg_dst_len,
-			 &rtmsg->rtmsg_src, rtmsg->rtmsg_src_len);
+			 &cfg->fc_dst, cfg->fc_dst_len,
+			 &cfg->fc_src, cfg->fc_src_len);
 	
 	if (fn) {
 		for (rt = fn->leaf; rt; rt = rt->u.next) {
-			if (rtmsg->rtmsg_ifindex &&
+			if (cfg->fc_ifindex &&
 			    (rt->rt6i_dev == NULL ||
-			     rt->rt6i_dev->ifindex != rtmsg->rtmsg_ifindex))
+			     rt->rt6i_dev->ifindex != cfg->fc_ifindex))
 				continue;
-			if (rtmsg->rtmsg_flags&RTF_GATEWAY &&
-			    !ipv6_addr_equal(&rtmsg->rtmsg_gateway, &rt->rt6i_gateway))
+			if (cfg->fc_flags & RTF_GATEWAY &&
+			    !ipv6_addr_equal(&cfg->fc_gateway, &rt->rt6i_gateway))
 				continue;
-			if (rtmsg->rtmsg_metric &&
-			    rtmsg->rtmsg_metric != rt->rt6i_metric)
+			if (cfg->fc_metric && cfg->fc_metric != rt->rt6i_metric)
 				continue;
 			dst_hold(&rt->u.dst);
 			read_unlock_bh(&table->tb6_lock);
 
-			return __ip6_del_rt(rt, nlh, _rtattr, req);
+			return __ip6_del_rt(rt, &cfg->fc_nlinfo);
 		}
 	}
 	read_unlock_bh(&table->tb6_lock);
@@ -1565,21 +1551,23 @@ static struct rt6_info *rt6_add_route_info(struct in6_addr *prefix, int prefixle
 					   struct in6_addr *gwaddr, int ifindex,
 					   unsigned pref)
 {
-	struct in6_rtmsg rtmsg;
+	struct fib6_config cfg = {
+		.fc_table	= RT6_TABLE_INFO,
+		.fc_metric	= 1024,
+		.fc_ifindex	= ifindex,
+		.fc_dst_len	= prefixlen,
+		.fc_flags	= RTF_GATEWAY | RTF_ADDRCONF | RTF_ROUTEINFO |
+				  RTF_UP | RTF_PREF(pref),
+	};
+
+	ipv6_addr_copy(&cfg.fc_dst, prefix);
+	ipv6_addr_copy(&cfg.fc_gateway, gwaddr);
 
-	memset(&rtmsg, 0, sizeof(rtmsg));
-	rtmsg.rtmsg_type = RTMSG_NEWROUTE;
-	ipv6_addr_copy(&rtmsg.rtmsg_dst, prefix);
-	rtmsg.rtmsg_dst_len = prefixlen;
-	ipv6_addr_copy(&rtmsg.rtmsg_gateway, gwaddr);
-	rtmsg.rtmsg_metric = 1024;
-	rtmsg.rtmsg_flags = RTF_GATEWAY | RTF_ADDRCONF | RTF_ROUTEINFO | RTF_UP | RTF_PREF(pref);
 	/* We should treat it as a default route if prefix length is 0. */
 	if (!prefixlen)
-		rtmsg.rtmsg_flags |= RTF_DEFAULT;
-	rtmsg.rtmsg_ifindex = ifindex;
+		cfg.fc_flags |= RTF_DEFAULT;
 
-	ip6_route_add(&rtmsg, NULL, NULL, NULL, RT6_TABLE_INFO);
+	ip6_route_add(&cfg);
 
 	return rt6_get_route_info(prefix, prefixlen, gwaddr, ifindex);
 }
@@ -1611,18 +1599,18 @@ struct rt6_info *rt6_add_dflt_router(struct in6_addr *gwaddr,
 				     struct net_device *dev,
 				     unsigned int pref)
 {
-	struct in6_rtmsg rtmsg;
+	struct fib6_config cfg = {
+		.fc_table	= RT6_TABLE_DFLT,
+		.fc_metric	= 1024,
+		.fc_ifindex	= dev->ifindex,
+		.fc_flags	= RTF_GATEWAY | RTF_ADDRCONF | RTF_DEFAULT |
+				  RTF_UP | RTF_EXPIRES | RTF_PREF(pref),
+	};
 
-	memset(&rtmsg, 0, sizeof(struct in6_rtmsg));
-	rtmsg.rtmsg_type = RTMSG_NEWROUTE;
-	ipv6_addr_copy(&rtmsg.rtmsg_gateway, gwaddr);
-	rtmsg.rtmsg_metric = 1024;
-	rtmsg.rtmsg_flags = RTF_GATEWAY | RTF_ADDRCONF | RTF_DEFAULT | RTF_UP | RTF_EXPIRES |
-			    RTF_PREF(pref);
+	ipv6_addr_copy(&cfg.fc_gateway, gwaddr);
 
-	rtmsg.rtmsg_ifindex = dev->ifindex;
+	ip6_route_add(&cfg);
 
-	ip6_route_add(&rtmsg, NULL, NULL, NULL, RT6_TABLE_DFLT);
 	return rt6_get_dflt_router(gwaddr, dev);
 }
 
@@ -1649,8 +1637,27 @@ restart:
 	read_unlock_bh(&table->tb6_lock);
 }
 
+static void rtmsg_to_fib6_config(struct in6_rtmsg *rtmsg,
+				 struct fib6_config *cfg)
+{
+	memset(cfg, 0, sizeof(*cfg));
+
+	cfg->fc_table = RT6_TABLE_MAIN;
+	cfg->fc_ifindex = rtmsg->rtmsg_ifindex;
+	cfg->fc_metric = rtmsg->rtmsg_metric;
+	cfg->fc_expires = rtmsg->rtmsg_info;
+	cfg->fc_dst_len = rtmsg->rtmsg_dst_len;
+	cfg->fc_src_len = rtmsg->rtmsg_src_len;
+	cfg->fc_flags = rtmsg->rtmsg_flags;
+
+	ipv6_addr_copy(&cfg->fc_dst, &rtmsg->rtmsg_dst);
+	ipv6_addr_copy(&cfg->fc_src, &rtmsg->rtmsg_src);
+	ipv6_addr_copy(&cfg->fc_gateway, &rtmsg->rtmsg_gateway);
+}
+
 int ipv6_route_ioctl(unsigned int cmd, void __user *arg)
 {
+	struct fib6_config cfg;
 	struct in6_rtmsg rtmsg;
 	int err;
 
@@ -1663,16 +1670,16 @@ int ipv6_route_ioctl(unsigned int cmd, void __user *arg)
 				     sizeof(struct in6_rtmsg));
 		if (err)
 			return -EFAULT;
-			
+
+		rtmsg_to_fib6_config(&rtmsg, &cfg);
+
 		rtnl_lock();
 		switch (cmd) {
 		case SIOCADDRT:
-			err = ip6_route_add(&rtmsg, NULL, NULL, NULL,
-					    RT6_TABLE_MAIN);
+			err = ip6_route_add(&cfg);
 			break;
 		case SIOCDELRT:
-			err = ip6_route_del(&rtmsg, NULL, NULL, NULL,
-					    RT6_TABLE_MAIN);
+			err = ip6_route_del(&cfg);
 			break;
 		default:
 			err = -EINVAL;
@@ -1823,66 +1830,104 @@ void rt6_mtu_change(struct net_device *dev, unsigned mtu)
 	fib6_clean_all(rt6_mtu_change_route, 0, &arg);
 }
 
-static int inet6_rtm_to_rtmsg(struct rtmsg *r, struct rtattr **rta,
-			      struct in6_rtmsg *rtmsg)
+static struct nla_policy rtm_ipv6_policy[RTA_MAX+1] __read_mostly = {
+	[RTA_GATEWAY]           = { .minlen = sizeof(struct in6_addr) },
+	[RTA_OIF]               = { .type = NLA_U32 },
+	[RTA_PRIORITY]          = { .type = NLA_U32 },
+	[RTA_METRICS]           = { .type = NLA_NESTED },
+};
+
+static int rtm_to_fib6_config(struct sk_buff *skb, struct nlmsghdr *nlh,
+			      struct fib6_config *cfg)
 {
-	memset(rtmsg, 0, sizeof(*rtmsg));
+	struct rtmsg *rtm;
+	struct nlattr *tb[RTA_MAX+1];
+	int err;
+
+	err = nlmsg_parse(nlh, sizeof(*rtm), tb, RTA_MAX, rtm_ipv6_policy);
+	if (err < 0)
+		goto errout;
 
-	rtmsg->rtmsg_dst_len = r->rtm_dst_len;
-	rtmsg->rtmsg_src_len = r->rtm_src_len;
-	rtmsg->rtmsg_flags = RTF_UP;
-	if (r->rtm_type == RTN_UNREACHABLE)
-		rtmsg->rtmsg_flags |= RTF_REJECT;
+	err = -EINVAL;
+	rtm = nlmsg_data(nlh);
+	memset(cfg, 0, sizeof(*cfg));
 
-	if (rta[RTA_GATEWAY-1]) {
-		if (rta[RTA_GATEWAY-1]->rta_len != RTA_LENGTH(16))
-			return -EINVAL;
-		memcpy(&rtmsg->rtmsg_gateway, RTA_DATA(rta[RTA_GATEWAY-1]), 16);
-		rtmsg->rtmsg_flags |= RTF_GATEWAY;
-	}
-	if (rta[RTA_DST-1]) {
-		if (RTA_PAYLOAD(rta[RTA_DST-1]) < ((r->rtm_dst_len+7)>>3))
-			return -EINVAL;
-		memcpy(&rtmsg->rtmsg_dst, RTA_DATA(rta[RTA_DST-1]), ((r->rtm_dst_len+7)>>3));
+	cfg->fc_table = rtm->rtm_table;
+	cfg->fc_dst_len = rtm->rtm_dst_len;
+	cfg->fc_src_len = rtm->rtm_src_len;
+	cfg->fc_flags = RTF_UP;
+	cfg->fc_protocol = rtm->rtm_protocol;
+
+	if (rtm->rtm_type == RTN_UNREACHABLE)
+		cfg->fc_flags |= RTF_REJECT;
+
+	cfg->fc_nlinfo.pid = NETLINK_CB(skb).pid;
+	cfg->fc_nlinfo.nlh = nlh;
+
+	if (tb[RTA_GATEWAY]) {
+		nla_memcpy(&cfg->fc_gateway, tb[RTA_GATEWAY], 16);
+		cfg->fc_flags |= RTF_GATEWAY;
 	}
-	if (rta[RTA_SRC-1]) {
-		if (RTA_PAYLOAD(rta[RTA_SRC-1]) < ((r->rtm_src_len+7)>>3))
-			return -EINVAL;
-		memcpy(&rtmsg->rtmsg_src, RTA_DATA(rta[RTA_SRC-1]), ((r->rtm_src_len+7)>>3));
+
+	if (tb[RTA_DST]) {
+		int plen = (rtm->rtm_dst_len + 7) >> 3;
+
+		if (nla_len(tb[RTA_DST]) < plen)
+			goto errout;
+
+		nla_memcpy(&cfg->fc_dst, tb[RTA_DST], plen);
 	}
-	if (rta[RTA_OIF-1]) {
-		if (rta[RTA_OIF-1]->rta_len != RTA_LENGTH(sizeof(int)))
-			return -EINVAL;
-		memcpy(&rtmsg->rtmsg_ifindex, RTA_DATA(rta[RTA_OIF-1]), sizeof(int));
+
+	if (tb[RTA_SRC]) {
+		int plen = (rtm->rtm_src_len + 7) >> 3;
+
+		if (nla_len(tb[RTA_SRC]) < plen)
+			goto errout;
+
+		nla_memcpy(&cfg->fc_src, tb[RTA_SRC], plen);
 	}
-	if (rta[RTA_PRIORITY-1]) {
-		if (rta[RTA_PRIORITY-1]->rta_len != RTA_LENGTH(4))
-			return -EINVAL;
-		memcpy(&rtmsg->rtmsg_metric, RTA_DATA(rta[RTA_PRIORITY-1]), 4);
+
+	if (tb[RTA_OIF])
+		cfg->fc_ifindex = nla_get_u32(tb[RTA_OIF]);
+
+	if (tb[RTA_PRIORITY])
+		cfg->fc_metric = nla_get_u32(tb[RTA_PRIORITY]);
+
+	if (tb[RTA_METRICS]) {
+		cfg->fc_mx = nla_data(tb[RTA_METRICS]);
+		cfg->fc_mx_len = nla_len(tb[RTA_METRICS]);
 	}
-	return 0;
+
+	if (tb[RTA_TABLE])
+		cfg->fc_table = nla_get_u32(tb[RTA_TABLE]);
+
+	err = 0;
+errout:
+	return err;
 }
 
 int inet6_rtm_delroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 {
-	struct rtmsg *r = NLMSG_DATA(nlh);
-	struct in6_rtmsg rtmsg;
+	struct fib6_config cfg;
+	int err;
 
-	if (inet6_rtm_to_rtmsg(r, arg, &rtmsg))
-		return -EINVAL;
-	return ip6_route_del(&rtmsg, nlh, arg, &NETLINK_CB(skb),
-			     rtm_get_table(arg, r->rtm_table));
+	err = rtm_to_fib6_config(skb, nlh, &cfg);
+	if (err < 0)
+		return err;
+
+	return ip6_route_del(&cfg);
 }
 
 int inet6_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 {
-	struct rtmsg *r = NLMSG_DATA(nlh);
-	struct in6_rtmsg rtmsg;
+	struct fib6_config cfg;
+	int err;
 
-	if (inet6_rtm_to_rtmsg(r, arg, &rtmsg))
-		return -EINVAL;
-	return ip6_route_add(&rtmsg, nlh, arg, &NETLINK_CB(skb),
-			     rtm_get_table(arg, r->rtm_table));
+	err = rtm_to_fib6_config(skb, nlh, &cfg);
+	if (err < 0)
+		return err;
+
+	return ip6_route_add(&cfg);
 }
 
 static int rt6_fill_node(struct sk_buff *skb, struct rt6_info *rt,
@@ -2063,15 +2108,21 @@ out_free:
 	goto out;	
 }
 
-void inet6_rt_notify(int event, struct rt6_info *rt, struct nlmsghdr *nlh, 
-			struct netlink_skb_parms *req)
+void inet6_rt_notify(int event, struct rt6_info *rt, struct nl_info *info)
 {
 	struct sk_buff *skb;
-	u32 pid = req ? req->pid : 0;
-	u32 seq = nlh ? nlh->nlmsg_seq : 0;
+	u32 pid = 0, seq = 0;
+	struct nlmsghdr *nlh = NULL;
 	int payload = sizeof(struct rtmsg) + 256;
 	int err = -ENOBUFS;
 
+	if (info) {
+		pid = info->pid;
+		nlh = info->nlh;
+		if (nlh)
+			seq = nlh->nlmsg_seq;
+	}
+
 	skb = nlmsg_new(nlmsg_total_size(payload), gfp_any());
 	if (skb == NULL)
 		goto errout;
-- 
cgit v1.2.3


From ac0b04627269ff16c3c7ab854a65fe6780c6e3e5 Mon Sep 17 00:00:00 2001
From: Sridhar Samudrala <sri@us.ibm.com>
Date: Tue, 22 Aug 2006 00:15:33 -0700
Subject: [SCTP]: Extend /proc/net/sctp/snmp to provide more statistics.

This patch adds more statistics info under /proc/net/sctp/snmp
that should be useful for debugging. The additional events that
are counted now include timer expirations, retransmits, packet
and data chunk discards.

The Data chunk discards include all the cases where a data chunk
is discarded including high tsn, bad stream, dup tsn and the most
useful one(out of receive buffer/rwnd).

Also moved the SCTP MIB data structures from the generic include
directories to include/sctp/sctp.h.

Signed-off-by: Sridhar Samudrala <sri@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/snmp.h    | 33 ---------------------------------
 include/net/sctp/sctp.h | 44 ++++++++++++++++++++++++++++++++++++++++++++
 include/net/snmp.h      |  6 ------
 net/sctp/input.c        |  8 ++++++--
 net/sctp/inqueue.c      |  4 ++--
 net/sctp/outqueue.c     |  6 +++++-
 net/sctp/proc.c         | 17 ++++++++++++++++-
 net/sctp/sm_statefuns.c | 15 +++++++++++++++
 8 files changed, 88 insertions(+), 45 deletions(-)

(limited to 'include')

diff --git a/include/linux/snmp.h b/include/linux/snmp.h
index 30156556f78d..854aa6b543f1 100644
--- a/include/linux/snmp.h
+++ b/include/linux/snmp.h
@@ -160,39 +160,6 @@ enum
 	__UDP_MIB_MAX
 };
 
-/* sctp mib definitions */
-/*
- * draft-ietf-sigtran-sctp-mib-07.txt
- */
-enum
-{
-	SCTP_MIB_NUM = 0,
-	SCTP_MIB_CURRESTAB,			/* CurrEstab */
-	SCTP_MIB_ACTIVEESTABS,			/* ActiveEstabs */
-	SCTP_MIB_PASSIVEESTABS,			/* PassiveEstabs */
-	SCTP_MIB_ABORTEDS,			/* Aborteds */
-	SCTP_MIB_SHUTDOWNS,			/* Shutdowns */
-	SCTP_MIB_OUTOFBLUES,			/* OutOfBlues */
-	SCTP_MIB_CHECKSUMERRORS,		/* ChecksumErrors */
-	SCTP_MIB_OUTCTRLCHUNKS,			/* OutCtrlChunks */
-	SCTP_MIB_OUTORDERCHUNKS,		/* OutOrderChunks */
-	SCTP_MIB_OUTUNORDERCHUNKS,		/* OutUnorderChunks */
-	SCTP_MIB_INCTRLCHUNKS,			/* InCtrlChunks */
-	SCTP_MIB_INORDERCHUNKS,			/* InOrderChunks */
-	SCTP_MIB_INUNORDERCHUNKS,		/* InUnorderChunks */
-	SCTP_MIB_FRAGUSRMSGS,			/* FragUsrMsgs */
-	SCTP_MIB_REASMUSRMSGS,			/* ReasmUsrMsgs */
-	SCTP_MIB_OUTSCTPPACKS,			/* OutSCTPPacks */
-	SCTP_MIB_INSCTPPACKS,			/* InSCTPPacks */
-	SCTP_MIB_RTOALGORITHM,			/* RtoAlgorithm */
-	SCTP_MIB_RTOMIN,			/* RtoMin */
-	SCTP_MIB_RTOMAX,			/* RtoMax */
-	SCTP_MIB_RTOINITIAL,			/* RtoInitial */
-	SCTP_MIB_VALCOOKIELIFE,			/* ValCookieLife */
-	SCTP_MIB_MAXINITRETR,			/* MaxInitRetr */
-	__SCTP_MIB_MAX
-};
-
 /* linux mib definitions */
 enum
 {
diff --git a/include/net/sctp/sctp.h b/include/net/sctp/sctp.h
index 1c1abce5f6b6..e274fd479990 100644
--- a/include/net/sctp/sctp.h
+++ b/include/net/sctp/sctp.h
@@ -216,6 +216,50 @@ DECLARE_SNMP_STAT(struct sctp_mib, sctp_statistics);
 
 #endif /* !TEST_FRAME */
 
+/* sctp mib definitions */
+enum
+{
+	SCTP_MIB_NUM = 0,
+	SCTP_MIB_CURRESTAB,			/* CurrEstab */
+	SCTP_MIB_ACTIVEESTABS,			/* ActiveEstabs */
+	SCTP_MIB_PASSIVEESTABS,			/* PassiveEstabs */
+	SCTP_MIB_ABORTEDS,			/* Aborteds */
+	SCTP_MIB_SHUTDOWNS,			/* Shutdowns */
+	SCTP_MIB_OUTOFBLUES,			/* OutOfBlues */
+	SCTP_MIB_CHECKSUMERRORS,		/* ChecksumErrors */
+	SCTP_MIB_OUTCTRLCHUNKS,			/* OutCtrlChunks */
+	SCTP_MIB_OUTORDERCHUNKS,		/* OutOrderChunks */
+	SCTP_MIB_OUTUNORDERCHUNKS,		/* OutUnorderChunks */
+	SCTP_MIB_INCTRLCHUNKS,			/* InCtrlChunks */
+	SCTP_MIB_INORDERCHUNKS,			/* InOrderChunks */
+	SCTP_MIB_INUNORDERCHUNKS,		/* InUnorderChunks */
+	SCTP_MIB_FRAGUSRMSGS,			/* FragUsrMsgs */
+	SCTP_MIB_REASMUSRMSGS,			/* ReasmUsrMsgs */
+	SCTP_MIB_OUTSCTPPACKS,			/* OutSCTPPacks */
+	SCTP_MIB_INSCTPPACKS,			/* InSCTPPacks */
+	SCTP_MIB_T1_INIT_EXPIREDS,
+	SCTP_MIB_T1_COOKIE_EXPIREDS,
+	SCTP_MIB_T2_SHUTDOWN_EXPIREDS,
+	SCTP_MIB_T3_RTX_EXPIREDS,
+	SCTP_MIB_T4_RTO_EXPIREDS,
+	SCTP_MIB_T5_SHUTDOWN_GUARD_EXPIREDS,
+	SCTP_MIB_DELAY_SACK_EXPIREDS,
+	SCTP_MIB_AUTOCLOSE_EXPIREDS,
+	SCTP_MIB_T3_RETRANSMITS,
+	SCTP_MIB_PMTUD_RETRANSMITS,
+	SCTP_MIB_FAST_RETRANSMITS,
+	SCTP_MIB_IN_PKT_SOFTIRQ,
+	SCTP_MIB_IN_PKT_BACKLOG,
+	SCTP_MIB_IN_PKT_DISCARDS,
+	SCTP_MIB_IN_DATA_CHUNK_DISCARDS,
+	__SCTP_MIB_MAX
+};
+
+#define SCTP_MIB_MAX    __SCTP_MIB_MAX
+struct sctp_mib {
+        unsigned long   mibs[SCTP_MIB_MAX];
+} __SNMP_MIB_ALIGN__;
+
 
 /* Print debugging messages.  */
 #if SCTP_DEBUG
diff --git a/include/net/snmp.h b/include/net/snmp.h
index a36bed8ea210..464970e39ec0 100644
--- a/include/net/snmp.h
+++ b/include/net/snmp.h
@@ -100,12 +100,6 @@ struct udp_mib {
 	unsigned long	mibs[UDP_MIB_MAX];
 } __SNMP_MIB_ALIGN__;
 
-/* SCTP */
-#define SCTP_MIB_MAX	__SCTP_MIB_MAX
-struct sctp_mib {
-	unsigned long	mibs[SCTP_MIB_MAX];
-} __SNMP_MIB_ALIGN__;
-
 /* Linux */
 #define LINUX_MIB_MAX	__LINUX_MIB_MAX
 struct linux_mib {
diff --git a/net/sctp/input.c b/net/sctp/input.c
index 42b66e74bbb5..8a34d95602ce 100644
--- a/net/sctp/input.c
+++ b/net/sctp/input.c
@@ -255,10 +255,13 @@ int sctp_rcv(struct sk_buff *skb)
 	 */
 	sctp_bh_lock_sock(sk);
 
-	if (sock_owned_by_user(sk))
+	if (sock_owned_by_user(sk)) {
+		SCTP_INC_STATS_BH(SCTP_MIB_IN_PKT_BACKLOG);
 		sctp_add_backlog(sk, skb);
-	else
+	} else {
+		SCTP_INC_STATS_BH(SCTP_MIB_IN_PKT_SOFTIRQ);
 		sctp_inq_push(&chunk->rcvr->inqueue, chunk);
+	}
 
 	sctp_bh_unlock_sock(sk);
 
@@ -271,6 +274,7 @@ int sctp_rcv(struct sk_buff *skb)
 	return 0;
 
 discard_it:
+	SCTP_INC_STATS_BH(SCTP_MIB_IN_PKT_DISCARDS);
 	kfree_skb(skb);
 	return 0;
 
diff --git a/net/sctp/inqueue.c b/net/sctp/inqueue.c
index cf0c767d43ae..cf6deed7e849 100644
--- a/net/sctp/inqueue.c
+++ b/net/sctp/inqueue.c
@@ -87,7 +87,7 @@ void sctp_inq_free(struct sctp_inq *queue)
 /* Put a new packet in an SCTP inqueue.
  * We assume that packet->sctp_hdr is set and in host byte order.
  */
-void sctp_inq_push(struct sctp_inq *q, struct sctp_chunk *packet)
+void sctp_inq_push(struct sctp_inq *q, struct sctp_chunk *chunk)
 {
 	/* Directly call the packet handling routine. */
 
@@ -96,7 +96,7 @@ void sctp_inq_push(struct sctp_inq *q, struct sctp_chunk *packet)
 	 * Eventually, we should clean up inqueue to not rely
 	 * on the BH related data structures.
 	 */
-	list_add_tail(&packet->list, &q->in_chunk_list);
+	list_add_tail(&chunk->list, &q->in_chunk_list);
 	q->immediate.func(q->immediate.data);
 }
 
diff --git a/net/sctp/outqueue.c b/net/sctp/outqueue.c
index 30b710c54e64..37074a39ecbb 100644
--- a/net/sctp/outqueue.c
+++ b/net/sctp/outqueue.c
@@ -467,6 +467,7 @@ void sctp_retransmit(struct sctp_outq *q, struct sctp_transport *transport,
 
 	switch(reason) {
 	case SCTP_RTXR_T3_RTX:
+		SCTP_INC_STATS(SCTP_MIB_T3_RETRANSMITS);
 		sctp_transport_lower_cwnd(transport, SCTP_LOWER_CWND_T3_RTX);
 		/* Update the retran path if the T3-rtx timer has expired for
 		 * the current retran path.
@@ -475,12 +476,15 @@ void sctp_retransmit(struct sctp_outq *q, struct sctp_transport *transport,
 			sctp_assoc_update_retran_path(transport->asoc);
 		break;
 	case SCTP_RTXR_FAST_RTX:
+		SCTP_INC_STATS(SCTP_MIB_FAST_RETRANSMITS);
 		sctp_transport_lower_cwnd(transport, SCTP_LOWER_CWND_FAST_RTX);
 		fast_retransmit = 1;
 		break;
 	case SCTP_RTXR_PMTUD:
-	default:
+		SCTP_INC_STATS(SCTP_MIB_PMTUD_RETRANSMITS);
 		break;
+	default:
+		BUG();
 	}
 
 	sctp_retransmit_mark(q, transport, fast_retransmit);
diff --git a/net/sctp/proc.c b/net/sctp/proc.c
index 5b3b0e0ae7e5..a356d8d310a9 100644
--- a/net/sctp/proc.c
+++ b/net/sctp/proc.c
@@ -57,6 +57,21 @@ static struct snmp_mib sctp_snmp_list[] = {
 	SNMP_MIB_ITEM("SctpReasmUsrMsgs", SCTP_MIB_REASMUSRMSGS),
 	SNMP_MIB_ITEM("SctpOutSCTPPacks", SCTP_MIB_OUTSCTPPACKS),
 	SNMP_MIB_ITEM("SctpInSCTPPacks", SCTP_MIB_INSCTPPACKS),
+	SNMP_MIB_ITEM("SctpT1InitExpireds", SCTP_MIB_T1_INIT_EXPIREDS),
+	SNMP_MIB_ITEM("SctpT1CookieExpireds", SCTP_MIB_T1_COOKIE_EXPIREDS),
+	SNMP_MIB_ITEM("SctpT2ShutdownExpireds", SCTP_MIB_T2_SHUTDOWN_EXPIREDS),
+	SNMP_MIB_ITEM("SctpT3RtxExpireds", SCTP_MIB_T3_RTX_EXPIREDS),
+	SNMP_MIB_ITEM("SctpT4RtoExpireds", SCTP_MIB_T4_RTO_EXPIREDS),
+	SNMP_MIB_ITEM("SctpT5ShutdownGuardExpireds", SCTP_MIB_T5_SHUTDOWN_GUARD_EXPIREDS),
+	SNMP_MIB_ITEM("SctpDelaySackExpireds", SCTP_MIB_DELAY_SACK_EXPIREDS),
+	SNMP_MIB_ITEM("SctpAutocloseExpireds", SCTP_MIB_AUTOCLOSE_EXPIREDS),
+	SNMP_MIB_ITEM("SctpT3Retransmits", SCTP_MIB_T3_RETRANSMITS),
+	SNMP_MIB_ITEM("SctpPmtudRetransmits", SCTP_MIB_PMTUD_RETRANSMITS),
+	SNMP_MIB_ITEM("SctpFastRetransmits", SCTP_MIB_FAST_RETRANSMITS),
+	SNMP_MIB_ITEM("SctpInPktSoftirq", SCTP_MIB_IN_PKT_SOFTIRQ),
+	SNMP_MIB_ITEM("SctpInPktBacklog", SCTP_MIB_IN_PKT_BACKLOG),
+	SNMP_MIB_ITEM("SctpInPktDiscards", SCTP_MIB_IN_PKT_DISCARDS),
+	SNMP_MIB_ITEM("SctpInDataChunkDiscards", SCTP_MIB_IN_DATA_CHUNK_DISCARDS),
 	SNMP_MIB_SENTINEL
 };
 
@@ -328,8 +343,8 @@ static int sctp_assocs_seq_show(struct seq_file *seq, void *v)
 			   "%8p %8p %-3d %-3d %-2d %-4d %4d %8d %8d %7d %5lu %-5d %5d ",
 			   assoc, sk, sctp_sk(sk)->type, sk->sk_state,
 			   assoc->state, hash, assoc->assoc_id,
-			   (sk->sk_rcvbuf - assoc->rwnd),
 			   assoc->sndbuf_used,
+			   (sk->sk_rcvbuf - assoc->rwnd),
 			   sock_i_uid(sk), sock_i_ino(sk),
 			   epb->bind_addr.port,
 			   assoc->peer.port);
diff --git a/net/sctp/sm_statefuns.c b/net/sctp/sm_statefuns.c
index 5b5ae7958322..32f57f42af9e 100644
--- a/net/sctp/sm_statefuns.c
+++ b/net/sctp/sm_statefuns.c
@@ -2663,9 +2663,11 @@ sctp_disposition_t sctp_sf_eat_data_6_2(const struct sctp_endpoint *ep,
 		break;
 	case SCTP_IERROR_HIGH_TSN:
 	case SCTP_IERROR_BAD_STREAM:
+		SCTP_INC_STATS(SCTP_MIB_IN_DATA_CHUNK_DISCARDS);
 		goto discard_noforce;
 	case SCTP_IERROR_DUP_TSN:
 	case SCTP_IERROR_IGNORE_TSN:
+		SCTP_INC_STATS(SCTP_MIB_IN_DATA_CHUNK_DISCARDS);
 		goto discard_force;
 	case SCTP_IERROR_NO_DATA:
 		goto consume;
@@ -3652,6 +3654,7 @@ sctp_disposition_t sctp_sf_pdiscard(const struct sctp_endpoint *ep,
 				    void *arg,
 				    sctp_cmd_seq_t *commands)
 {
+	SCTP_INC_STATS(SCTP_MIB_IN_PKT_DISCARDS);
 	sctp_add_cmd_sf(commands, SCTP_CMD_DISCARD_PACKET, SCTP_NULL());
 
 	return SCTP_DISPOSITION_CONSUME;
@@ -4548,6 +4551,8 @@ sctp_disposition_t sctp_sf_do_6_3_3_rtx(const struct sctp_endpoint *ep,
 {
 	struct sctp_transport *transport = arg;
 
+	SCTP_INC_STATS(SCTP_MIB_T3_RTX_EXPIREDS);
+
 	if (asoc->overall_error_count >= asoc->max_retrans) {
 		sctp_add_cmd_sf(commands, SCTP_CMD_SET_SK_ERR,
 				SCTP_ERROR(ETIMEDOUT));
@@ -4616,6 +4621,7 @@ sctp_disposition_t sctp_sf_do_6_2_sack(const struct sctp_endpoint *ep,
 				       void *arg,
 				       sctp_cmd_seq_t *commands)
 {
+	SCTP_INC_STATS(SCTP_MIB_DELAY_SACK_EXPIREDS);
 	sctp_add_cmd_sf(commands, SCTP_CMD_GEN_SACK, SCTP_FORCE());
 	return SCTP_DISPOSITION_CONSUME;
 }
@@ -4650,6 +4656,7 @@ sctp_disposition_t sctp_sf_t1_init_timer_expire(const struct sctp_endpoint *ep,
 	int attempts = asoc->init_err_counter + 1;
 
 	SCTP_DEBUG_PRINTK("Timer T1 expired (INIT).\n");
+	SCTP_INC_STATS(SCTP_MIB_T1_INIT_EXPIREDS);
 
 	if (attempts <= asoc->max_init_attempts) {
 		bp = (struct sctp_bind_addr *) &asoc->base.bind_addr;
@@ -4709,6 +4716,7 @@ sctp_disposition_t sctp_sf_t1_cookie_timer_expire(const struct sctp_endpoint *ep
 	int attempts = asoc->init_err_counter + 1;
 
 	SCTP_DEBUG_PRINTK("Timer T1 expired (COOKIE-ECHO).\n");
+	SCTP_INC_STATS(SCTP_MIB_T1_COOKIE_EXPIREDS);
 
 	if (attempts <= asoc->max_init_attempts) {
 		repl = sctp_make_cookie_echo(asoc, NULL);
@@ -4753,6 +4761,8 @@ sctp_disposition_t sctp_sf_t2_timer_expire(const struct sctp_endpoint *ep,
 	struct sctp_chunk *reply = NULL;
 
 	SCTP_DEBUG_PRINTK("Timer T2 expired.\n");
+	SCTP_INC_STATS(SCTP_MIB_T2_SHUTDOWN_EXPIREDS);
+
 	if (asoc->overall_error_count >= asoc->max_retrans) {
 		sctp_add_cmd_sf(commands, SCTP_CMD_SET_SK_ERR,
 				SCTP_ERROR(ETIMEDOUT));
@@ -4814,6 +4824,8 @@ sctp_disposition_t sctp_sf_t4_timer_expire(
 	struct sctp_chunk *chunk = asoc->addip_last_asconf;
 	struct sctp_transport *transport = chunk->transport;
 
+	SCTP_INC_STATS(SCTP_MIB_T4_RTO_EXPIREDS);
+
 	/* ADDIP 4.1 B1) Increment the error counters and perform path failure
 	 * detection on the appropriate destination address as defined in
 	 * RFC2960 [5] section 8.1 and 8.2.
@@ -4880,6 +4892,7 @@ sctp_disposition_t sctp_sf_t5_timer_expire(const struct sctp_endpoint *ep,
 	struct sctp_chunk *reply = NULL;
 
 	SCTP_DEBUG_PRINTK("Timer T5 expired.\n");
+	SCTP_INC_STATS(SCTP_MIB_T5_SHUTDOWN_GUARD_EXPIREDS);
 
 	reply = sctp_make_abort(asoc, NULL, 0);
 	if (!reply)
@@ -4910,6 +4923,8 @@ sctp_disposition_t sctp_sf_autoclose_timer_expire(
 {
 	int disposition;
 
+	SCTP_INC_STATS(SCTP_MIB_AUTOCLOSE_EXPIREDS);
+
 	/* From 9.2 Shutdown of an Association
 	 * Upon receipt of the SHUTDOWN primitive from its upper
 	 * layer, the endpoint enters SHUTDOWN-PENDING state and
-- 
cgit v1.2.3


From 9ba1627617d396135a4d679542a3623d5819e628 Mon Sep 17 00:00:00 2001
From: Yasuyuki Kozakai <yasuyuki.kozakai@toshiba.co.jp>
Date: Tue, 22 Aug 2006 00:29:37 -0700
Subject: [NETFILTER]: x_tables: replace IPv4 dscp match by address family
 independent version

This replaces IPv4 dscp match by address family independent version.
This also
	- utilizes dsfield.h to get the DS field in IPv4/IPv6 header, and
	- checks for the DSCP value from user space.
	- fixes Kconfig help text.

Signed-off-by: Yasuyuki Kozakai <yasuyuki.kozakai@toshiba.co.jp>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/xt_dscp.h       |  23 +++++++
 include/linux/netfilter_ipv4/ipt_dscp.h |  14 ++--
 net/ipv4/netfilter/Kconfig              |  11 ----
 net/ipv4/netfilter/Makefile             |   1 -
 net/ipv4/netfilter/ipt_dscp.c           |  54 ---------------
 net/netfilter/Kconfig                   |  11 ++++
 net/netfilter/Makefile                  |   1 +
 net/netfilter/xt_dscp.c                 | 113 ++++++++++++++++++++++++++++++++
 8 files changed, 154 insertions(+), 74 deletions(-)
 create mode 100644 include/linux/netfilter/xt_dscp.h
 delete mode 100644 net/ipv4/netfilter/ipt_dscp.c
 create mode 100644 net/netfilter/xt_dscp.c

(limited to 'include')

diff --git a/include/linux/netfilter/xt_dscp.h b/include/linux/netfilter/xt_dscp.h
new file mode 100644
index 000000000000..1da61e6acaf7
--- /dev/null
+++ b/include/linux/netfilter/xt_dscp.h
@@ -0,0 +1,23 @@
+/* x_tables module for matching the IPv4/IPv6 DSCP field
+ *
+ * (C) 2002 Harald Welte <laforge@gnumonks.org>
+ * This software is distributed under GNU GPL v2, 1991
+ *
+ * See RFC2474 for a description of the DSCP field within the IP Header.
+ *
+ * xt_dscp.h,v 1.3 2002/08/05 19:00:21 laforge Exp
+*/
+#ifndef _XT_DSCP_H
+#define _XT_DSCP_H
+
+#define XT_DSCP_MASK	0xfc	/* 11111100 */
+#define XT_DSCP_SHIFT	2
+#define XT_DSCP_MAX	0x3f	/* 00111111 */
+
+/* match info */
+struct xt_dscp_info {
+	u_int8_t dscp;
+	u_int8_t invert;
+};
+
+#endif /* _XT_DSCP_H */
diff --git a/include/linux/netfilter_ipv4/ipt_dscp.h b/include/linux/netfilter_ipv4/ipt_dscp.h
index 2fa6dfe92894..4b82ca912b0e 100644
--- a/include/linux/netfilter_ipv4/ipt_dscp.h
+++ b/include/linux/netfilter_ipv4/ipt_dscp.h
@@ -10,14 +10,12 @@
 #ifndef _IPT_DSCP_H
 #define _IPT_DSCP_H
 
-#define IPT_DSCP_MASK	0xfc	/* 11111100 */
-#define IPT_DSCP_SHIFT	2
-#define IPT_DSCP_MAX	0x3f	/* 00111111 */
+#include <linux/netfilter/xt_dscp.h>
 
-/* match info */
-struct ipt_dscp_info {
-	u_int8_t dscp;
-	u_int8_t invert;
-};
+#define IPT_DSCP_MASK	XT_DSCP_MASK
+#define IPT_DSCP_SHIFT	XT_DSCP_SHIFT
+#define IPT_DSCP_MAX	XT_DSCP_MAX
+
+#define ipt_dscp_info	xt_dscp_info
 
 #endif /* _IPT_DSCP_H */
diff --git a/net/ipv4/netfilter/Kconfig b/net/ipv4/netfilter/Kconfig
index ef0b5aac5838..d88d71d1ce0d 100644
--- a/net/ipv4/netfilter/Kconfig
+++ b/net/ipv4/netfilter/Kconfig
@@ -278,17 +278,6 @@ config IP_NF_MATCH_ECN
 
 	  To compile it as a module, choose M here.  If unsure, say N.
 
-config IP_NF_MATCH_DSCP
-	tristate "DSCP match support"
-	depends on IP_NF_IPTABLES
-	help
-	  This option adds a `DSCP' match, which allows you to match against
-	  the IPv4 header DSCP field (DSCP codepoint).
-
-	  The DSCP codepoint can have any value between 0x0 and 0x4f.
-
-	  To compile it as a module, choose M here.  If unsure, say N.
-
 config IP_NF_MATCH_AH
 	tristate "AH match support"
 	depends on IP_NF_IPTABLES
diff --git a/net/ipv4/netfilter/Makefile b/net/ipv4/netfilter/Makefile
index 3ded4a3af59c..b946b0f3ea9d 100644
--- a/net/ipv4/netfilter/Makefile
+++ b/net/ipv4/netfilter/Makefile
@@ -59,7 +59,6 @@ obj-$(CONFIG_IP_NF_MATCH_OWNER) += ipt_owner.o
 obj-$(CONFIG_IP_NF_MATCH_TOS) += ipt_tos.o
 obj-$(CONFIG_IP_NF_MATCH_RECENT) += ipt_recent.o
 obj-$(CONFIG_IP_NF_MATCH_ECN) += ipt_ecn.o
-obj-$(CONFIG_IP_NF_MATCH_DSCP) += ipt_dscp.o
 obj-$(CONFIG_IP_NF_MATCH_AH) += ipt_ah.o
 obj-$(CONFIG_IP_NF_MATCH_TTL) += ipt_ttl.o
 obj-$(CONFIG_IP_NF_MATCH_ADDRTYPE) += ipt_addrtype.o
diff --git a/net/ipv4/netfilter/ipt_dscp.c b/net/ipv4/netfilter/ipt_dscp.c
deleted file mode 100644
index 47177591aeb6..000000000000
--- a/net/ipv4/netfilter/ipt_dscp.c
+++ /dev/null
@@ -1,54 +0,0 @@
-/* IP tables module for matching the value of the IPv4 DSCP field
- *
- * ipt_dscp.c,v 1.3 2002/08/05 19:00:21 laforge Exp
- *
- * (C) 2002 by Harald Welte <laforge@netfilter.org>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- */
-
-#include <linux/module.h>
-#include <linux/skbuff.h>
-
-#include <linux/netfilter_ipv4/ipt_dscp.h>
-#include <linux/netfilter_ipv4/ip_tables.h>
-
-MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>");
-MODULE_DESCRIPTION("iptables DSCP matching module");
-MODULE_LICENSE("GPL");
-
-static int match(const struct sk_buff *skb,
-		 const struct net_device *in, const struct net_device *out,
-		 const struct xt_match *match, const void *matchinfo,
-		 int offset, unsigned int protoff, int *hotdrop)
-{
-	const struct ipt_dscp_info *info = matchinfo;
-	const struct iphdr *iph = skb->nh.iph;
-
-	u_int8_t sh_dscp = ((info->dscp << IPT_DSCP_SHIFT) & IPT_DSCP_MASK);
-
-	return ((iph->tos&IPT_DSCP_MASK) == sh_dscp) ^ info->invert;
-}
-
-static struct ipt_match dscp_match = {
-	.name		= "dscp",
-	.match		= match,
-	.matchsize	= sizeof(struct ipt_dscp_info),
-	.me		= THIS_MODULE,
-};
-
-static int __init ipt_dscp_init(void)
-{
-	return ipt_register_match(&dscp_match);
-}
-
-static void __exit ipt_dscp_fini(void)
-{
-	ipt_unregister_match(&dscp_match);
-
-}
-
-module_init(ipt_dscp_init);
-module_exit(ipt_dscp_fini);
diff --git a/net/netfilter/Kconfig b/net/netfilter/Kconfig
index a9894ddfd72a..f781405f5d65 100644
--- a/net/netfilter/Kconfig
+++ b/net/netfilter/Kconfig
@@ -263,6 +263,17 @@ config NETFILTER_XT_MATCH_DCCP
 	  If you want to compile it as a module, say M here and read
 	  <file:Documentation/modules.txt>.  If unsure, say `N'.
 
+config NETFILTER_XT_MATCH_DSCP
+	tristate '"DSCP" match support'
+	depends on NETFILTER_XTABLES
+	help
+	  This option adds a `DSCP' match, which allows you to match against
+	  the IPv4/IPv6 header DSCP field (differentiated services codepoint).
+
+	  The DSCP field can have any value between 0x0 and 0x3f inclusive.
+
+	  To compile it as a module, choose M here.  If unsure, say N.
+
 config NETFILTER_XT_MATCH_ESP
 	tristate '"ESP" match support'
 	depends on NETFILTER_XTABLES
diff --git a/net/netfilter/Makefile b/net/netfilter/Makefile
index 6fa4b7580458..0b8a70c1df46 100644
--- a/net/netfilter/Makefile
+++ b/net/netfilter/Makefile
@@ -37,6 +37,7 @@ obj-$(CONFIG_NETFILTER_XT_MATCH_CONNBYTES) += xt_connbytes.o
 obj-$(CONFIG_NETFILTER_XT_MATCH_CONNMARK) += xt_connmark.o
 obj-$(CONFIG_NETFILTER_XT_MATCH_CONNTRACK) += xt_conntrack.o
 obj-$(CONFIG_NETFILTER_XT_MATCH_DCCP) += xt_dccp.o
+obj-$(CONFIG_NETFILTER_XT_MATCH_DSCP) += xt_dscp.o
 obj-$(CONFIG_NETFILTER_XT_MATCH_ESP) += xt_esp.o
 obj-$(CONFIG_NETFILTER_XT_MATCH_HELPER) += xt_helper.o
 obj-$(CONFIG_NETFILTER_XT_MATCH_LENGTH) += xt_length.o
diff --git a/net/netfilter/xt_dscp.c b/net/netfilter/xt_dscp.c
new file mode 100644
index 000000000000..82e250d1f007
--- /dev/null
+++ b/net/netfilter/xt_dscp.c
@@ -0,0 +1,113 @@
+/* IP tables module for matching the value of the IPv4/IPv6 DSCP field
+ *
+ * xt_dscp.c,v 1.3 2002/08/05 19:00:21 laforge Exp
+ *
+ * (C) 2002 by Harald Welte <laforge@netfilter.org>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/module.h>
+#include <linux/skbuff.h>
+#include <linux/ip.h>
+#include <linux/ipv6.h>
+#include <net/dsfield.h>
+
+#include <linux/netfilter/xt_dscp.h>
+#include <linux/netfilter/x_tables.h>
+
+MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>");
+MODULE_DESCRIPTION("x_tables DSCP matching module");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("ipt_dscp");
+MODULE_ALIAS("ip6t_dscp");
+
+static int match(const struct sk_buff *skb,
+		 const struct net_device *in,
+		 const struct net_device *out,
+		 const struct xt_match *match,
+		 const void *matchinfo,
+		 int offset,
+		 unsigned int protoff,
+		 int *hotdrop)
+{
+	const struct xt_dscp_info *info = matchinfo;
+	u_int8_t dscp = ipv4_get_dsfield(skb->nh.iph) >> XT_DSCP_SHIFT;
+
+	return (dscp == info->dscp) ^ !!info->invert;
+}
+
+static int match6(const struct sk_buff *skb,
+		  const struct net_device *in,
+		  const struct net_device *out,
+		  const struct xt_match *match,
+		  const void *matchinfo,
+		  int offset,
+		  unsigned int protoff,
+		  int *hotdrop)
+{
+	const struct xt_dscp_info *info = matchinfo;
+	u_int8_t dscp = ipv6_get_dsfield(skb->nh.ipv6h) >> XT_DSCP_SHIFT;
+
+	return (dscp == info->dscp) ^ !!info->invert;
+}
+
+static int checkentry(const char *tablename,
+		      const void *info,
+		      const struct xt_match *match,
+		      void *matchinfo,
+		      unsigned int matchsize,
+		      unsigned int hook_mask)
+{
+	const u_int8_t dscp = ((struct xt_dscp_info *)matchinfo)->dscp;
+
+	if (dscp > XT_DSCP_MAX) {
+		printk(KERN_ERR "xt_dscp: dscp %x out of range\n", dscp);
+		return 0;
+	}
+
+	return 1;
+}
+
+static struct xt_match dscp_match = {
+	.name		= "dscp",
+	.match		= match,
+	.checkentry	= checkentry,
+	.matchsize	= sizeof(struct xt_dscp_info),
+	.family		= AF_INET,
+	.me		= THIS_MODULE,
+};
+
+static struct xt_match dscp6_match = {
+	.name		= "dscp",
+	.match		= match6,
+	.checkentry	= checkentry,
+	.matchsize	= sizeof(struct xt_dscp_info),
+	.family		= AF_INET6,
+	.me		= THIS_MODULE,
+};
+
+static int __init xt_dscp_match_init(void)
+{
+	int ret;
+	ret = xt_register_match(&dscp_match);
+	if (ret)
+		return ret;
+
+	ret = xt_register_match(&dscp6_match);
+	if (ret)
+		xt_unregister_match(&dscp_match);
+
+	return ret;
+}
+
+static void __exit xt_dscp_match_fini(void)
+{
+	xt_unregister_match(&dscp_match);
+	xt_unregister_match(&dscp6_match);
+}
+
+module_init(xt_dscp_match_init);
+module_exit(xt_dscp_match_fini);
-- 
cgit v1.2.3


From a468701db58a8b3e08e3f55fa6ac66db42014922 Mon Sep 17 00:00:00 2001
From: Yasuyuki Kozakai <yasuyuki.kozakai@toshiba.co.jp>
Date: Tue, 22 Aug 2006 00:30:26 -0700
Subject: [NETFILTER]: x_tables: replace IPv4 DSCP target by address family
 independent version

This replaces IPv4 DSCP target by address family independent version.
This also
	- utilizes dsfield.h to get/mangle DS field in IPv4/IPv6 header
	- fixes Kconfig help text.

Signed-off-by: Yasuyuki Kozakai <yasuyuki.kozakai@toshiba.co.jp>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/xt_DSCP.h       |  20 +++++
 include/linux/netfilter_ipv4/ipt_DSCP.h |   6 +-
 net/ipv4/netfilter/Kconfig              |  11 ---
 net/ipv4/netfilter/Makefile             |   1 -
 net/ipv4/netfilter/ipt_DSCP.c           |  96 -----------------------
 net/netfilter/Kconfig                   |  12 +++
 net/netfilter/Makefile                  |   1 +
 net/netfilter/xt_DSCP.c                 | 130 ++++++++++++++++++++++++++++++++
 8 files changed, 165 insertions(+), 112 deletions(-)
 create mode 100644 include/linux/netfilter/xt_DSCP.h
 delete mode 100644 net/ipv4/netfilter/ipt_DSCP.c
 create mode 100644 net/netfilter/xt_DSCP.c

(limited to 'include')

diff --git a/include/linux/netfilter/xt_DSCP.h b/include/linux/netfilter/xt_DSCP.h
new file mode 100644
index 000000000000..3c7c963997bd
--- /dev/null
+++ b/include/linux/netfilter/xt_DSCP.h
@@ -0,0 +1,20 @@
+/* x_tables module for setting the IPv4/IPv6 DSCP field
+ *
+ * (C) 2002 Harald Welte <laforge@gnumonks.org>
+ * based on ipt_FTOS.c (C) 2000 by Matthew G. Marsh <mgm@paktronix.com>
+ * This software is distributed under GNU GPL v2, 1991
+ *
+ * See RFC2474 for a description of the DSCP field within the IP Header.
+ *
+ * xt_DSCP.h,v 1.7 2002/03/14 12:03:13 laforge Exp
+*/
+#ifndef _XT_DSCP_TARGET_H
+#define _XT_DSCP_TARGET_H
+#include <linux/netfilter/xt_dscp.h>
+
+/* target info */
+struct xt_DSCP_info {
+	u_int8_t dscp;
+};
+
+#endif /* _XT_DSCP_TARGET_H */
diff --git a/include/linux/netfilter_ipv4/ipt_DSCP.h b/include/linux/netfilter_ipv4/ipt_DSCP.h
index b30f510b5bef..3491e524d5ea 100644
--- a/include/linux/netfilter_ipv4/ipt_DSCP.h
+++ b/include/linux/netfilter_ipv4/ipt_DSCP.h
@@ -11,10 +11,8 @@
 #ifndef _IPT_DSCP_TARGET_H
 #define _IPT_DSCP_TARGET_H
 #include <linux/netfilter_ipv4/ipt_dscp.h>
+#include <linux/netfilter/xt_DSCP.h>
 
-/* target info */
-struct ipt_DSCP_info {
-	u_int8_t dscp;
-};
+#define ipt_DSCP_info xt_DSCP_info
 
 #endif /* _IPT_DSCP_TARGET_H */
diff --git a/net/ipv4/netfilter/Kconfig b/net/ipv4/netfilter/Kconfig
index d88d71d1ce0d..a55b8ff70ded 100644
--- a/net/ipv4/netfilter/Kconfig
+++ b/net/ipv4/netfilter/Kconfig
@@ -557,17 +557,6 @@ config IP_NF_TARGET_ECN
 
 	  To compile it as a module, choose M here.  If unsure, say N.
 
-config IP_NF_TARGET_DSCP
-	tristate "DSCP target support"
-	depends on IP_NF_MANGLE
-	help
-	  This option adds a `DSCP' match, which allows you to match against
-	  the IPv4 header DSCP field (DSCP codepoint).
-
-	  The DSCP codepoint can have any value between 0x0 and 0x4f.
-
-	  To compile it as a module, choose M here.  If unsure, say N.
-
 config IP_NF_TARGET_TTL
 	tristate  'TTL target support'
 	depends on IP_NF_MANGLE
diff --git a/net/ipv4/netfilter/Makefile b/net/ipv4/netfilter/Makefile
index b946b0f3ea9d..09aaed1a8063 100644
--- a/net/ipv4/netfilter/Makefile
+++ b/net/ipv4/netfilter/Makefile
@@ -67,7 +67,6 @@ obj-$(CONFIG_IP_NF_MATCH_ADDRTYPE) += ipt_addrtype.o
 obj-$(CONFIG_IP_NF_TARGET_REJECT) += ipt_REJECT.o
 obj-$(CONFIG_IP_NF_TARGET_TOS) += ipt_TOS.o
 obj-$(CONFIG_IP_NF_TARGET_ECN) += ipt_ECN.o
-obj-$(CONFIG_IP_NF_TARGET_DSCP) += ipt_DSCP.o
 obj-$(CONFIG_IP_NF_TARGET_MASQUERADE) += ipt_MASQUERADE.o
 obj-$(CONFIG_IP_NF_TARGET_REDIRECT) += ipt_REDIRECT.o
 obj-$(CONFIG_IP_NF_TARGET_NETMAP) += ipt_NETMAP.o
diff --git a/net/ipv4/netfilter/ipt_DSCP.c b/net/ipv4/netfilter/ipt_DSCP.c
deleted file mode 100644
index c8e971288dfe..000000000000
--- a/net/ipv4/netfilter/ipt_DSCP.c
+++ /dev/null
@@ -1,96 +0,0 @@
-/* iptables module for setting the IPv4 DSCP field, Version 1.8
- *
- * (C) 2002 by Harald Welte <laforge@netfilter.org>
- * based on ipt_FTOS.c (C) 2000 by Matthew G. Marsh <mgm@paktronix.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as 
- * published by the Free Software Foundation.
- * 
- * See RFC2474 for a description of the DSCP field within the IP Header.
- *
- * ipt_DSCP.c,v 1.8 2002/08/06 18:41:57 laforge Exp
-*/
-
-#include <linux/module.h>
-#include <linux/skbuff.h>
-#include <linux/ip.h>
-#include <net/checksum.h>
-
-#include <linux/netfilter_ipv4/ip_tables.h>
-#include <linux/netfilter_ipv4/ipt_DSCP.h>
-
-MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>");
-MODULE_DESCRIPTION("iptables DSCP modification module");
-MODULE_LICENSE("GPL");
-
-static unsigned int
-target(struct sk_buff **pskb,
-       const struct net_device *in,
-       const struct net_device *out,
-       unsigned int hooknum,
-       const struct xt_target *target,
-       const void *targinfo,
-       void *userinfo)
-{
-	const struct ipt_DSCP_info *dinfo = targinfo;
-	u_int8_t sh_dscp = ((dinfo->dscp << IPT_DSCP_SHIFT) & IPT_DSCP_MASK);
-
-
-	if (((*pskb)->nh.iph->tos & IPT_DSCP_MASK) != sh_dscp) {
-		u_int16_t diffs[2];
-
-		if (!skb_make_writable(pskb, sizeof(struct iphdr)))
-			return NF_DROP;
-
-		diffs[0] = htons((*pskb)->nh.iph->tos) ^ 0xFFFF;
-		(*pskb)->nh.iph->tos = ((*pskb)->nh.iph->tos & ~IPT_DSCP_MASK)
-			| sh_dscp;
-		diffs[1] = htons((*pskb)->nh.iph->tos);
-		(*pskb)->nh.iph->check
-			= csum_fold(csum_partial((char *)diffs,
-						 sizeof(diffs),
-						 (*pskb)->nh.iph->check
-						 ^ 0xFFFF));
-	}
-	return IPT_CONTINUE;
-}
-
-static int
-checkentry(const char *tablename,
-	   const void *e_void,
-	   const struct xt_target *target,
-           void *targinfo,
-           unsigned int targinfosize,
-           unsigned int hook_mask)
-{
-	const u_int8_t dscp = ((struct ipt_DSCP_info *)targinfo)->dscp;
-
-	if ((dscp > IPT_DSCP_MAX)) {
-		printk(KERN_WARNING "DSCP: dscp %x out of range\n", dscp);
-		return 0;
-	}
-	return 1;
-}
-
-static struct ipt_target ipt_dscp_reg = {
-	.name		= "DSCP",
-	.target		= target,
-	.targetsize	= sizeof(struct ipt_DSCP_info),
-	.table		= "mangle",
-	.checkentry	= checkentry,
-	.me		= THIS_MODULE,
-};
-
-static int __init ipt_dscp_init(void)
-{
-	return ipt_register_target(&ipt_dscp_reg);
-}
-
-static void __exit ipt_dscp_fini(void)
-{
-	ipt_unregister_target(&ipt_dscp_reg);
-}
-
-module_init(ipt_dscp_init);
-module_exit(ipt_dscp_fini);
diff --git a/net/netfilter/Kconfig b/net/netfilter/Kconfig
index f781405f5d65..0a28d2c5c44f 100644
--- a/net/netfilter/Kconfig
+++ b/net/netfilter/Kconfig
@@ -148,6 +148,18 @@ config NETFILTER_XT_TARGET_CONNMARK
 	  <file:Documentation/modules.txt>.  The module will be called
 	  ipt_CONNMARK.o.  If unsure, say `N'.
 
+config NETFILTER_XT_TARGET_DSCP
+	tristate '"DSCP" target support'
+	depends on NETFILTER_XTABLES
+	depends on IP_NF_MANGLE || IP6_NF_MANGLE
+	help
+	  This option adds a `DSCP' target, which allows you to manipulate
+	  the IPv4/IPv6 header DSCP field (differentiated services codepoint).
+
+	  The DSCP field can have any value between 0x0 and 0x3f inclusive.
+
+	  To compile it as a module, choose M here.  If unsure, say N.
+
 config NETFILTER_XT_TARGET_MARK
 	tristate '"MARK" target support'
 	depends on NETFILTER_XTABLES
diff --git a/net/netfilter/Makefile b/net/netfilter/Makefile
index 0b8a70c1df46..a74be492fd0a 100644
--- a/net/netfilter/Makefile
+++ b/net/netfilter/Makefile
@@ -25,6 +25,7 @@ obj-$(CONFIG_NETFILTER_XTABLES) += x_tables.o xt_tcpudp.o
 # targets
 obj-$(CONFIG_NETFILTER_XT_TARGET_CLASSIFY) += xt_CLASSIFY.o
 obj-$(CONFIG_NETFILTER_XT_TARGET_CONNMARK) += xt_CONNMARK.o
+obj-$(CONFIG_NETFILTER_XT_TARGET_DSCP) += xt_DSCP.o
 obj-$(CONFIG_NETFILTER_XT_TARGET_MARK) += xt_MARK.o
 obj-$(CONFIG_NETFILTER_XT_TARGET_NFQUEUE) += xt_NFQUEUE.o
 obj-$(CONFIG_NETFILTER_XT_TARGET_NOTRACK) += xt_NOTRACK.o
diff --git a/net/netfilter/xt_DSCP.c b/net/netfilter/xt_DSCP.c
new file mode 100644
index 000000000000..79df8165cd79
--- /dev/null
+++ b/net/netfilter/xt_DSCP.c
@@ -0,0 +1,130 @@
+/* x_tables module for setting the IPv4/IPv6 DSCP field, Version 1.8
+ *
+ * (C) 2002 by Harald Welte <laforge@netfilter.org>
+ * based on ipt_FTOS.c (C) 2000 by Matthew G. Marsh <mgm@paktronix.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * See RFC2474 for a description of the DSCP field within the IP Header.
+ *
+ * xt_DSCP.c,v 1.8 2002/08/06 18:41:57 laforge Exp
+*/
+
+#include <linux/module.h>
+#include <linux/skbuff.h>
+#include <linux/ip.h>
+#include <linux/ipv6.h>
+#include <net/dsfield.h>
+
+#include <linux/netfilter/x_tables.h>
+#include <linux/netfilter/xt_DSCP.h>
+
+MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>");
+MODULE_DESCRIPTION("x_tables DSCP modification module");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("ipt_DSCP");
+MODULE_ALIAS("ip6t_DSCP");
+
+static unsigned int target(struct sk_buff **pskb,
+			   const struct net_device *in,
+			   const struct net_device *out,
+			   unsigned int hooknum,
+			   const struct xt_target *target,
+			   const void *targinfo,
+			   void *userinfo)
+{
+	const struct xt_DSCP_info *dinfo = targinfo;
+	u_int8_t dscp = ipv4_get_dsfield((*pskb)->nh.iph) >> XT_DSCP_SHIFT;
+
+	if (dscp != dinfo->dscp) {
+		if (!skb_make_writable(pskb, sizeof(struct iphdr)))
+			return NF_DROP;
+
+		ipv4_change_dsfield((*pskb)->nh.iph, (__u8)(~XT_DSCP_MASK),
+				    dinfo->dscp << XT_DSCP_SHIFT);
+
+	}
+	return XT_CONTINUE;
+}
+
+static unsigned int target6(struct sk_buff **pskb,
+			    const struct net_device *in,
+			    const struct net_device *out,
+			    unsigned int hooknum,
+			    const struct xt_target *target,
+			    const void *targinfo,
+			    void *userinfo)
+{
+	const struct xt_DSCP_info *dinfo = targinfo;
+	u_int8_t dscp = ipv6_get_dsfield((*pskb)->nh.ipv6h) >> XT_DSCP_SHIFT;
+
+	if (dscp != dinfo->dscp) {
+		if (!skb_make_writable(pskb, sizeof(struct ipv6hdr)))
+			return NF_DROP;
+
+		ipv6_change_dsfield((*pskb)->nh.ipv6h, (__u8)(~XT_DSCP_MASK),
+				    dinfo->dscp << XT_DSCP_SHIFT);
+	}
+	return XT_CONTINUE;
+}
+
+static int checkentry(const char *tablename,
+		      const void *e_void,
+		      const struct xt_target *target,
+		      void *targinfo,
+		      unsigned int targinfosize,
+		      unsigned int hook_mask)
+{
+	const u_int8_t dscp = ((struct xt_DSCP_info *)targinfo)->dscp;
+
+	if ((dscp > XT_DSCP_MAX)) {
+		printk(KERN_WARNING "DSCP: dscp %x out of range\n", dscp);
+		return 0;
+	}
+	return 1;
+}
+
+static struct xt_target xt_dscp_reg = {
+	.name		= "DSCP",
+	.target		= target,
+	.targetsize	= sizeof(struct xt_DSCP_info),
+	.table		= "mangle",
+	.checkentry	= checkentry,
+	.family		= AF_INET,
+	.me		= THIS_MODULE,
+};
+
+static struct xt_target xt_dscp6_reg = {
+	.name		= "DSCP",
+	.target		= target6,
+	.targetsize	= sizeof(struct xt_DSCP_info),
+	.table		= "mangle",
+	.checkentry	= checkentry,
+	.family		= AF_INET6,
+	.me		= THIS_MODULE,
+};
+
+static int __init xt_dscp_target_init(void)
+{
+	int ret;
+	ret = xt_register_target(&xt_dscp_reg);
+	if (ret)
+		return ret;
+
+	ret = xt_register_target(&xt_dscp6_reg);
+	if (ret)
+		xt_unregister_target(&xt_dscp_reg);
+
+	return ret;
+}
+
+static void __exit xt_dscp_target_fini(void)
+{
+	xt_unregister_target(&xt_dscp_reg);
+	xt_unregister_target(&xt_dscp6_reg);
+}
+
+module_init(xt_dscp_target_init);
+module_exit(xt_dscp_target_fini);
-- 
cgit v1.2.3


From 2521c12cf1a29f6c380b13ca32a38175f6beed08 Mon Sep 17 00:00:00 2001
From: Pablo Neira Ayuso <pablo@netfilter.org>
Date: Tue, 22 Aug 2006 00:31:24 -0700
Subject: [NETFILTER]: conntrack: introduce connection mark event

This patch introduces the mark event. ctnetlink can use this to know if
the mark needs to be dumped.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nf_conntrack_common.h |  4 ++++
 net/netfilter/xt_CONNMARK.c                   | 16 ++++++++++++++--
 2 files changed, 18 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/include/linux/netfilter/nf_conntrack_common.h b/include/linux/netfilter/nf_conntrack_common.h
index d2e4bd7a7a14..9e0dae07861e 100644
--- a/include/linux/netfilter/nf_conntrack_common.h
+++ b/include/linux/netfilter/nf_conntrack_common.h
@@ -125,6 +125,10 @@ enum ip_conntrack_events
 	/* Counter highest bit has been set */
 	IPCT_COUNTER_FILLING_BIT = 11,
 	IPCT_COUNTER_FILLING = (1 << IPCT_COUNTER_FILLING_BIT),
+
+	/* Mark is set */
+	IPCT_MARK_BIT = 12,
+	IPCT_MARK = (1 << IPCT_MARK_BIT),
 };
 
 enum ip_conntrack_expect_events {
diff --git a/net/netfilter/xt_CONNMARK.c b/net/netfilter/xt_CONNMARK.c
index 60c375d36f01..784482b74e58 100644
--- a/net/netfilter/xt_CONNMARK.c
+++ b/net/netfilter/xt_CONNMARK.c
@@ -52,13 +52,25 @@ target(struct sk_buff **pskb,
 	    switch(markinfo->mode) {
 	    case XT_CONNMARK_SET:
 		newmark = (*ctmark & ~markinfo->mask) | markinfo->mark;
-		if (newmark != *ctmark)
+		if (newmark != *ctmark) {
 		    *ctmark = newmark;
+#ifdef CONFIG_IP_NF_CONNTRACK_EVENTS
+		    ip_conntrack_event_cache(IPCT_MARK, *pskb);
+#else
+		    nf_conntrack_event_cache(IPCT_MARK, *pskb);
+#endif
+		}
 		break;
 	    case XT_CONNMARK_SAVE:
 		newmark = (*ctmark & ~markinfo->mask) | ((*pskb)->nfmark & markinfo->mask);
-		if (*ctmark != newmark)
+		if (*ctmark != newmark) {
 		    *ctmark = newmark;
+#ifdef CONFIG_IP_NF_CONNTRACK_EVENTS
+		    ip_conntrack_event_cache(IPCT_MARK, *pskb);
+#else
+		    nf_conntrack_event_cache(IPCT_MARK, *pskb);
+#endif
+		}
 		break;
 	    case XT_CONNMARK_RESTORE:
 		nfmark = (*pskb)->nfmark;
-- 
cgit v1.2.3


From 52d9c42ef2563d2c420eb23b96bf5a4cae9e167b Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 22 Aug 2006 00:33:45 -0700
Subject: [NETFILTER]: x_tables: add helpers for mass match/target registration

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/x_tables.h |  5 ++++
 net/netfilter/x_tables.c           | 60 ++++++++++++++++++++++++++++++++++++++
 2 files changed, 65 insertions(+)

(limited to 'include')

diff --git a/include/linux/netfilter/x_tables.h b/include/linux/netfilter/x_tables.h
index 48cc32d83f77..9a9912430e3a 100644
--- a/include/linux/netfilter/x_tables.h
+++ b/include/linux/netfilter/x_tables.h
@@ -290,8 +290,13 @@ struct xt_table_info
 
 extern int xt_register_target(struct xt_target *target);
 extern void xt_unregister_target(struct xt_target *target);
+extern int xt_register_targets(struct xt_target *target, unsigned int n);
+extern void xt_unregister_targets(struct xt_target *target, unsigned int n);
+
 extern int xt_register_match(struct xt_match *target);
 extern void xt_unregister_match(struct xt_match *target);
+extern int xt_register_matches(struct xt_match *match, unsigned int n);
+extern void xt_unregister_matches(struct xt_match *match, unsigned int n);
 
 extern int xt_check_match(const struct xt_match *match, unsigned short family,
 			  unsigned int size, const char *table, unsigned int hook,
diff --git a/net/netfilter/x_tables.c b/net/netfilter/x_tables.c
index 174e8f970095..8037ba63d587 100644
--- a/net/netfilter/x_tables.c
+++ b/net/netfilter/x_tables.c
@@ -86,6 +86,36 @@ xt_unregister_target(struct xt_target *target)
 }
 EXPORT_SYMBOL(xt_unregister_target);
 
+int
+xt_register_targets(struct xt_target *target, unsigned int n)
+{
+	unsigned int i;
+	int err = 0;
+
+	for (i = 0; i < n; i++) {
+		err = xt_register_target(&target[i]);
+		if (err)
+			goto err;
+	}
+	return err;
+
+err:
+	if (i > 0)
+		xt_unregister_targets(target, i);
+	return err;
+}
+EXPORT_SYMBOL(xt_register_targets);
+
+void
+xt_unregister_targets(struct xt_target *target, unsigned int n)
+{
+	unsigned int i;
+
+	for (i = 0; i < n; i++)
+		xt_unregister_target(&target[i]);
+}
+EXPORT_SYMBOL(xt_unregister_targets);
+
 int
 xt_register_match(struct xt_match *match)
 {
@@ -113,6 +143,36 @@ xt_unregister_match(struct xt_match *match)
 }
 EXPORT_SYMBOL(xt_unregister_match);
 
+int
+xt_register_matches(struct xt_match *match, unsigned int n)
+{
+	unsigned int i;
+	int err = 0;
+
+	for (i = 0; i < n; i++) {
+		err = xt_register_match(&match[i]);
+		if (err)
+			goto err;
+	}
+	return err;
+
+err:
+	if (i > 0)
+		xt_unregister_matches(match, i);
+	return err;
+}
+EXPORT_SYMBOL(xt_register_matches);
+
+void
+xt_unregister_matches(struct xt_match *match, unsigned int n)
+{
+	unsigned int i;
+
+	for (i = 0; i < n; i++)
+		xt_unregister_match(&match[i]);
+}
+EXPORT_SYMBOL(xt_unregister_matches);
+
 
 /*
  * These are weird, but module loading must not be done with mutex
-- 
cgit v1.2.3


From fe1cb10873b44cf89082465823ee6d4d4ac63ad7 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 22 Aug 2006 00:35:47 -0700
Subject: [NETFILTER]: x_tables: remove unused argument to target functions

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/x_tables.h        | 3 +--
 include/linux/netfilter_arp/arp_tables.h  | 3 +--
 include/linux/netfilter_ipv4/ip_tables.h  | 3 +--
 include/linux/netfilter_ipv6/ip6_tables.h | 3 +--
 net/ipv4/netfilter/arp_tables.c           | 9 +++------
 net/ipv4/netfilter/arpt_mangle.c          | 2 +-
 net/ipv4/netfilter/arptable_filter.c      | 2 +-
 net/ipv4/netfilter/ip_nat_rule.c          | 8 +++-----
 net/ipv4/netfilter/ip_tables.c            | 9 +++------
 net/ipv4/netfilter/ipt_CLUSTERIP.c        | 3 +--
 net/ipv4/netfilter/ipt_ECN.c              | 3 +--
 net/ipv4/netfilter/ipt_LOG.c              | 3 +--
 net/ipv4/netfilter/ipt_MASQUERADE.c       | 3 +--
 net/ipv4/netfilter/ipt_NETMAP.c           | 3 +--
 net/ipv4/netfilter/ipt_REDIRECT.c         | 3 +--
 net/ipv4/netfilter/ipt_REJECT.c           | 3 +--
 net/ipv4/netfilter/ipt_SAME.c             | 3 +--
 net/ipv4/netfilter/ipt_TCPMSS.c           | 3 +--
 net/ipv4/netfilter/ipt_TOS.c              | 3 +--
 net/ipv4/netfilter/ipt_TTL.c              | 2 +-
 net/ipv4/netfilter/ipt_ULOG.c             | 2 +-
 net/ipv4/netfilter/iptable_filter.c       | 4 ++--
 net/ipv4/netfilter/iptable_mangle.c       | 4 ++--
 net/ipv4/netfilter/iptable_raw.c          | 2 +-
 net/ipv6/netfilter/ip6_tables.c           | 9 +++------
 net/ipv6/netfilter/ip6t_HL.c              | 2 +-
 net/ipv6/netfilter/ip6t_LOG.c             | 3 +--
 net/ipv6/netfilter/ip6t_REJECT.c          | 3 +--
 net/ipv6/netfilter/ip6table_filter.c      | 4 ++--
 net/ipv6/netfilter/ip6table_mangle.c      | 4 ++--
 net/ipv6/netfilter/ip6table_raw.c         | 2 +-
 net/netfilter/xt_CLASSIFY.c               | 3 +--
 net/netfilter/xt_CONNMARK.c               | 3 +--
 net/netfilter/xt_CONNSECMARK.c            | 2 +-
 net/netfilter/xt_DSCP.c                   | 6 ++----
 net/netfilter/xt_MARK.c                   | 6 ++----
 net/netfilter/xt_NFQUEUE.c                | 3 +--
 net/netfilter/xt_NOTRACK.c                | 3 +--
 net/netfilter/xt_SECMARK.c                | 2 +-
 net/netfilter/xt_connbytes.c              | 2 +-
 net/sched/act_ipt.c                       | 2 +-
 41 files changed, 55 insertions(+), 90 deletions(-)

(limited to 'include')

diff --git a/include/linux/netfilter/x_tables.h b/include/linux/netfilter/x_tables.h
index 9a9912430e3a..9cef0e91542b 100644
--- a/include/linux/netfilter/x_tables.h
+++ b/include/linux/netfilter/x_tables.h
@@ -211,8 +211,7 @@ struct xt_target
 			       const struct net_device *out,
 			       unsigned int hooknum,
 			       const struct xt_target *target,
-			       const void *targinfo,
-			       void *userdata);
+			       const void *targinfo);
 
 	/* Called when user tries to insert an entry of this type:
            hook_mask is a bitmask of hooks from which it can be
diff --git a/include/linux/netfilter_arp/arp_tables.h b/include/linux/netfilter_arp/arp_tables.h
index 62cc27daca4e..149e87c9ab13 100644
--- a/include/linux/netfilter_arp/arp_tables.h
+++ b/include/linux/netfilter_arp/arp_tables.h
@@ -248,8 +248,7 @@ extern unsigned int arpt_do_table(struct sk_buff **pskb,
 				  unsigned int hook,
 				  const struct net_device *in,
 				  const struct net_device *out,
-				  struct arpt_table *table,
-				  void *userdata);
+				  struct arpt_table *table);
 
 #define ARPT_ALIGN(s) (((s) + (__alignof__(struct arpt_entry)-1)) & ~(__alignof__(struct arpt_entry)-1))
 #endif /*__KERNEL__*/
diff --git a/include/linux/netfilter_ipv4/ip_tables.h b/include/linux/netfilter_ipv4/ip_tables.h
index c0dac16e1902..a536bbdef145 100644
--- a/include/linux/netfilter_ipv4/ip_tables.h
+++ b/include/linux/netfilter_ipv4/ip_tables.h
@@ -312,8 +312,7 @@ extern unsigned int ipt_do_table(struct sk_buff **pskb,
 				 unsigned int hook,
 				 const struct net_device *in,
 				 const struct net_device *out,
-				 struct ipt_table *table,
-				 void *userdata);
+				 struct ipt_table *table);
 
 #define IPT_ALIGN(s) XT_ALIGN(s)
 
diff --git a/include/linux/netfilter_ipv6/ip6_tables.h b/include/linux/netfilter_ipv6/ip6_tables.h
index d0d5d1ee4be3..d7a8e9c0dad0 100644
--- a/include/linux/netfilter_ipv6/ip6_tables.h
+++ b/include/linux/netfilter_ipv6/ip6_tables.h
@@ -300,8 +300,7 @@ extern unsigned int ip6t_do_table(struct sk_buff **pskb,
 				  unsigned int hook,
 				  const struct net_device *in,
 				  const struct net_device *out,
-				  struct ip6t_table *table,
-				  void *userdata);
+				  struct ip6t_table *table);
 
 /* Check for an extension */
 extern int ip6t_ext_hdr(u8 nexthdr);
diff --git a/net/ipv4/netfilter/arp_tables.c b/net/ipv4/netfilter/arp_tables.c
index 8d1d7a6e72a5..c6bd270bf46a 100644
--- a/net/ipv4/netfilter/arp_tables.c
+++ b/net/ipv4/netfilter/arp_tables.c
@@ -208,8 +208,7 @@ static unsigned int arpt_error(struct sk_buff **pskb,
 			       const struct net_device *out,
 			       unsigned int hooknum,
 			       const struct xt_target *target,
-			       const void *targinfo,
-			       void *userinfo)
+			       const void *targinfo)
 {
 	if (net_ratelimit())
 		printk("arp_tables: error: '%s'\n", (char *)targinfo);
@@ -226,8 +225,7 @@ unsigned int arpt_do_table(struct sk_buff **pskb,
 			   unsigned int hook,
 			   const struct net_device *in,
 			   const struct net_device *out,
-			   struct arpt_table *table,
-			   void *userdata)
+			   struct arpt_table *table)
 {
 	static const char nulldevname[IFNAMSIZ];
 	unsigned int verdict = NF_DROP;
@@ -302,8 +300,7 @@ unsigned int arpt_do_table(struct sk_buff **pskb,
 								     in, out,
 								     hook,
 								     t->u.kernel.target,
-								     t->data,
-								     userdata);
+								     t->data);
 
 				/* Target might have changed stuff. */
 				arp = (*pskb)->nh.arph;
diff --git a/net/ipv4/netfilter/arpt_mangle.c b/net/ipv4/netfilter/arpt_mangle.c
index a58325c1ceb9..05fb2421bb26 100644
--- a/net/ipv4/netfilter/arpt_mangle.c
+++ b/net/ipv4/netfilter/arpt_mangle.c
@@ -11,7 +11,7 @@ static unsigned int
 target(struct sk_buff **pskb,
        const struct net_device *in, const struct net_device *out,
        unsigned int hooknum, const struct xt_target *target,
-       const void *targinfo, void *userinfo)
+       const void *targinfo)
 {
 	const struct arpt_mangle *mangle = targinfo;
 	struct arphdr *arp;
diff --git a/net/ipv4/netfilter/arptable_filter.c b/net/ipv4/netfilter/arptable_filter.c
index d7c472faa53b..7edea2a1696c 100644
--- a/net/ipv4/netfilter/arptable_filter.c
+++ b/net/ipv4/netfilter/arptable_filter.c
@@ -155,7 +155,7 @@ static unsigned int arpt_hook(unsigned int hook,
 			      const struct net_device *out,
 			      int (*okfn)(struct sk_buff *))
 {
-	return arpt_do_table(pskb, hook, in, out, &packet_filter, NULL);
+	return arpt_do_table(pskb, hook, in, out, &packet_filter);
 }
 
 static struct nf_hook_ops arpt_ops[] = {
diff --git a/net/ipv4/netfilter/ip_nat_rule.c b/net/ipv4/netfilter/ip_nat_rule.c
index 1aba926c1cb0..1aa0e4f462a5 100644
--- a/net/ipv4/netfilter/ip_nat_rule.c
+++ b/net/ipv4/netfilter/ip_nat_rule.c
@@ -104,8 +104,7 @@ static unsigned int ipt_snat_target(struct sk_buff **pskb,
 				    const struct net_device *out,
 				    unsigned int hooknum,
 				    const struct ipt_target *target,
-				    const void *targinfo,
-				    void *userinfo)
+				    const void *targinfo)
 {
 	struct ip_conntrack *ct;
 	enum ip_conntrack_info ctinfo;
@@ -147,8 +146,7 @@ static unsigned int ipt_dnat_target(struct sk_buff **pskb,
 				    const struct net_device *out,
 				    unsigned int hooknum,
 				    const struct ipt_target *target,
-				    const void *targinfo,
-				    void *userinfo)
+				    const void *targinfo)
 {
 	struct ip_conntrack *ct;
 	enum ip_conntrack_info ctinfo;
@@ -255,7 +253,7 @@ int ip_nat_rule_find(struct sk_buff **pskb,
 {
 	int ret;
 
-	ret = ipt_do_table(pskb, hooknum, in, out, &nat_table, NULL);
+	ret = ipt_do_table(pskb, hooknum, in, out, &nat_table);
 
 	if (ret == NF_ACCEPT) {
 		if (!ip_nat_initialized(ct, HOOK2MANIP(hooknum)))
diff --git a/net/ipv4/netfilter/ip_tables.c b/net/ipv4/netfilter/ip_tables.c
index 048514f15f2f..8ce5b6f76447 100644
--- a/net/ipv4/netfilter/ip_tables.c
+++ b/net/ipv4/netfilter/ip_tables.c
@@ -180,8 +180,7 @@ ipt_error(struct sk_buff **pskb,
 	  const struct net_device *out,
 	  unsigned int hooknum,
 	  const struct xt_target *target,
-	  const void *targinfo,
-	  void *userinfo)
+	  const void *targinfo)
 {
 	if (net_ratelimit())
 		printk("ip_tables: error: `%s'\n", (char *)targinfo);
@@ -217,8 +216,7 @@ ipt_do_table(struct sk_buff **pskb,
 	     unsigned int hook,
 	     const struct net_device *in,
 	     const struct net_device *out,
-	     struct ipt_table *table,
-	     void *userdata)
+	     struct ipt_table *table)
 {
 	static const char nulldevname[IFNAMSIZ] __attribute__((aligned(sizeof(long))));
 	u_int16_t offset;
@@ -308,8 +306,7 @@ ipt_do_table(struct sk_buff **pskb,
 								     in, out,
 								     hook,
 								     t->u.kernel.target,
-								     t->data,
-								     userdata);
+								     t->data);
 
 #ifdef CONFIG_NETFILTER_DEBUG
 				if (((struct ipt_entry *)table_base)->comefrom
diff --git a/net/ipv4/netfilter/ipt_CLUSTERIP.c b/net/ipv4/netfilter/ipt_CLUSTERIP.c
index d994c5f5744c..a08383cf9e7a 100644
--- a/net/ipv4/netfilter/ipt_CLUSTERIP.c
+++ b/net/ipv4/netfilter/ipt_CLUSTERIP.c
@@ -302,8 +302,7 @@ target(struct sk_buff **pskb,
        const struct net_device *out,
        unsigned int hooknum,
        const struct xt_target *target,
-       const void *targinfo,
-       void *userinfo)
+       const void *targinfo)
 {
 	const struct ipt_clusterip_tgt_info *cipinfo = targinfo;
 	enum ip_conntrack_info ctinfo;
diff --git a/net/ipv4/netfilter/ipt_ECN.c b/net/ipv4/netfilter/ipt_ECN.c
index 7e30e6d2b5da..1c3da4a48e5f 100644
--- a/net/ipv4/netfilter/ipt_ECN.c
+++ b/net/ipv4/netfilter/ipt_ECN.c
@@ -85,8 +85,7 @@ target(struct sk_buff **pskb,
        const struct net_device *out,
        unsigned int hooknum,
        const struct xt_target *target,
-       const void *targinfo,
-       void *userinfo)
+       const void *targinfo)
 {
 	const struct ipt_ECN_info *einfo = targinfo;
 
diff --git a/net/ipv4/netfilter/ipt_LOG.c b/net/ipv4/netfilter/ipt_LOG.c
index b98f7b08b084..a8d356c6191f 100644
--- a/net/ipv4/netfilter/ipt_LOG.c
+++ b/net/ipv4/netfilter/ipt_LOG.c
@@ -416,8 +416,7 @@ ipt_log_target(struct sk_buff **pskb,
 	       const struct net_device *out,
 	       unsigned int hooknum,
 	       const struct xt_target *target,
-	       const void *targinfo,
-	       void *userinfo)
+	       const void *targinfo)
 {
 	const struct ipt_log_info *loginfo = targinfo;
 	struct nf_loginfo li;
diff --git a/net/ipv4/netfilter/ipt_MASQUERADE.c b/net/ipv4/netfilter/ipt_MASQUERADE.c
index ebd94f2abf0d..9659793c66c0 100644
--- a/net/ipv4/netfilter/ipt_MASQUERADE.c
+++ b/net/ipv4/netfilter/ipt_MASQUERADE.c
@@ -64,8 +64,7 @@ masquerade_target(struct sk_buff **pskb,
 		  const struct net_device *out,
 		  unsigned int hooknum,
 		  const struct xt_target *target,
-		  const void *targinfo,
-		  void *userinfo)
+		  const void *targinfo)
 {
 	struct ip_conntrack *ct;
 	enum ip_conntrack_info ctinfo;
diff --git a/net/ipv4/netfilter/ipt_NETMAP.c b/net/ipv4/netfilter/ipt_NETMAP.c
index 736c4b5a86a7..fd5e74a19fb5 100644
--- a/net/ipv4/netfilter/ipt_NETMAP.c
+++ b/net/ipv4/netfilter/ipt_NETMAP.c
@@ -55,8 +55,7 @@ target(struct sk_buff **pskb,
        const struct net_device *out,
        unsigned int hooknum,
        const struct xt_target *target,
-       const void *targinfo,
-       void *userinfo)
+       const void *targinfo)
 {
 	struct ip_conntrack *ct;
 	enum ip_conntrack_info ctinfo;
diff --git a/net/ipv4/netfilter/ipt_REDIRECT.c b/net/ipv4/netfilter/ipt_REDIRECT.c
index f290463232de..839fe99f71d4 100644
--- a/net/ipv4/netfilter/ipt_REDIRECT.c
+++ b/net/ipv4/netfilter/ipt_REDIRECT.c
@@ -58,8 +58,7 @@ redirect_target(struct sk_buff **pskb,
 		const struct net_device *out,
 		unsigned int hooknum,
 		const struct xt_target *target,
-		const void *targinfo,
-		void *userinfo)
+		const void *targinfo)
 {
 	struct ip_conntrack *ct;
 	enum ip_conntrack_info ctinfo;
diff --git a/net/ipv4/netfilter/ipt_REJECT.c b/net/ipv4/netfilter/ipt_REJECT.c
index 95c6662b663c..1dfd8e56be8b 100644
--- a/net/ipv4/netfilter/ipt_REJECT.c
+++ b/net/ipv4/netfilter/ipt_REJECT.c
@@ -228,8 +228,7 @@ static unsigned int reject(struct sk_buff **pskb,
 			   const struct net_device *out,
 			   unsigned int hooknum,
 			   const struct xt_target *target,
-			   const void *targinfo,
-			   void *userinfo)
+			   const void *targinfo)
 {
 	const struct ipt_reject_info *reject = targinfo;
 
diff --git a/net/ipv4/netfilter/ipt_SAME.c b/net/ipv4/netfilter/ipt_SAME.c
index 7169b09b5a67..cf801749490f 100644
--- a/net/ipv4/netfilter/ipt_SAME.c
+++ b/net/ipv4/netfilter/ipt_SAME.c
@@ -133,8 +133,7 @@ same_target(struct sk_buff **pskb,
 		const struct net_device *out,
 		unsigned int hooknum,
 		const struct xt_target *target,
-		const void *targinfo,
-		void *userinfo)
+		const void *targinfo)
 {
 	struct ip_conntrack *ct;
 	enum ip_conntrack_info ctinfo;
diff --git a/net/ipv4/netfilter/ipt_TCPMSS.c b/net/ipv4/netfilter/ipt_TCPMSS.c
index 0fce85e05507..6d668dcfc22a 100644
--- a/net/ipv4/netfilter/ipt_TCPMSS.c
+++ b/net/ipv4/netfilter/ipt_TCPMSS.c
@@ -41,8 +41,7 @@ ipt_tcpmss_target(struct sk_buff **pskb,
 		  const struct net_device *out,
 		  unsigned int hooknum,
 		  const struct xt_target *target,
-		  const void *targinfo,
-		  void *userinfo)
+		  const void *targinfo)
 {
 	const struct ipt_tcpmss_info *tcpmssinfo = targinfo;
 	struct tcphdr *tcph;
diff --git a/net/ipv4/netfilter/ipt_TOS.c b/net/ipv4/netfilter/ipt_TOS.c
index 52e9d705d48e..043df0137084 100644
--- a/net/ipv4/netfilter/ipt_TOS.c
+++ b/net/ipv4/netfilter/ipt_TOS.c
@@ -26,8 +26,7 @@ target(struct sk_buff **pskb,
        const struct net_device *out,
        unsigned int hooknum,
        const struct xt_target *target,
-       const void *targinfo,
-       void *userinfo)
+       const void *targinfo)
 {
 	const struct ipt_tos_target_info *tosinfo = targinfo;
 	struct iphdr *iph = (*pskb)->nh.iph;
diff --git a/net/ipv4/netfilter/ipt_TTL.c b/net/ipv4/netfilter/ipt_TTL.c
index 2afb2a8aa8c5..164007107b5e 100644
--- a/net/ipv4/netfilter/ipt_TTL.c
+++ b/net/ipv4/netfilter/ipt_TTL.c
@@ -23,7 +23,7 @@ static unsigned int
 ipt_ttl_target(struct sk_buff **pskb,
 	       const struct net_device *in, const struct net_device *out,
 	       unsigned int hooknum, const struct xt_target *target,
-	       const void *targinfo, void *userinfo)
+	       const void *targinfo)
 {
 	struct iphdr *iph;
 	const struct ipt_TTL_info *info = targinfo;
diff --git a/net/ipv4/netfilter/ipt_ULOG.c b/net/ipv4/netfilter/ipt_ULOG.c
index d46fd677fa11..4c5f0a117862 100644
--- a/net/ipv4/netfilter/ipt_ULOG.c
+++ b/net/ipv4/netfilter/ipt_ULOG.c
@@ -308,7 +308,7 @@ static unsigned int ipt_ulog_target(struct sk_buff **pskb,
 				    const struct net_device *out,
 				    unsigned int hooknum,
 				    const struct xt_target *target,
-				    const void *targinfo, void *userinfo)
+				    const void *targinfo)
 {
 	struct ipt_ulog_info *loginfo = (struct ipt_ulog_info *) targinfo;
 
diff --git a/net/ipv4/netfilter/iptable_filter.c b/net/ipv4/netfilter/iptable_filter.c
index 7f417484bfbf..e2e7dd8d7903 100644
--- a/net/ipv4/netfilter/iptable_filter.c
+++ b/net/ipv4/netfilter/iptable_filter.c
@@ -90,7 +90,7 @@ ipt_hook(unsigned int hook,
 	 const struct net_device *out,
 	 int (*okfn)(struct sk_buff *))
 {
-	return ipt_do_table(pskb, hook, in, out, &packet_filter, NULL);
+	return ipt_do_table(pskb, hook, in, out, &packet_filter);
 }
 
 static unsigned int
@@ -108,7 +108,7 @@ ipt_local_out_hook(unsigned int hook,
 		return NF_ACCEPT;
 	}
 
-	return ipt_do_table(pskb, hook, in, out, &packet_filter, NULL);
+	return ipt_do_table(pskb, hook, in, out, &packet_filter);
 }
 
 static struct nf_hook_ops ipt_ops[] = {
diff --git a/net/ipv4/netfilter/iptable_mangle.c b/net/ipv4/netfilter/iptable_mangle.c
index 4e7998beda63..79336cb42527 100644
--- a/net/ipv4/netfilter/iptable_mangle.c
+++ b/net/ipv4/netfilter/iptable_mangle.c
@@ -119,7 +119,7 @@ ipt_route_hook(unsigned int hook,
 	 const struct net_device *out,
 	 int (*okfn)(struct sk_buff *))
 {
-	return ipt_do_table(pskb, hook, in, out, &packet_mangler, NULL);
+	return ipt_do_table(pskb, hook, in, out, &packet_mangler);
 }
 
 static unsigned int
@@ -148,7 +148,7 @@ ipt_local_hook(unsigned int hook,
 	daddr = (*pskb)->nh.iph->daddr;
 	tos = (*pskb)->nh.iph->tos;
 
-	ret = ipt_do_table(pskb, hook, in, out, &packet_mangler, NULL);
+	ret = ipt_do_table(pskb, hook, in, out, &packet_mangler);
 	/* Reroute for ANY change. */
 	if (ret != NF_DROP && ret != NF_STOLEN && ret != NF_QUEUE
 	    && ((*pskb)->nh.iph->saddr != saddr
diff --git a/net/ipv4/netfilter/iptable_raw.c b/net/ipv4/netfilter/iptable_raw.c
index 7912cce1e1b8..bcbeb4aeacd9 100644
--- a/net/ipv4/netfilter/iptable_raw.c
+++ b/net/ipv4/netfilter/iptable_raw.c
@@ -95,7 +95,7 @@ ipt_hook(unsigned int hook,
 	 const struct net_device *out,
 	 int (*okfn)(struct sk_buff *))
 {
-	return ipt_do_table(pskb, hook, in, out, &packet_raw, NULL);
+	return ipt_do_table(pskb, hook, in, out, &packet_raw);
 }
 
 /* 'raw' is the very first table. */
diff --git a/net/ipv6/netfilter/ip6_tables.c b/net/ipv6/netfilter/ip6_tables.c
index c9d6b23cd3f7..38cd7ffda9a0 100644
--- a/net/ipv6/netfilter/ip6_tables.c
+++ b/net/ipv6/netfilter/ip6_tables.c
@@ -220,8 +220,7 @@ ip6t_error(struct sk_buff **pskb,
 	  const struct net_device *out,
 	  unsigned int hooknum,
 	  const struct xt_target *target,
-	  const void *targinfo,
-	  void *userinfo)
+	  const void *targinfo)
 {
 	if (net_ratelimit())
 		printk("ip6_tables: error: `%s'\n", (char *)targinfo);
@@ -258,8 +257,7 @@ ip6t_do_table(struct sk_buff **pskb,
 	      unsigned int hook,
 	      const struct net_device *in,
 	      const struct net_device *out,
-	      struct xt_table *table,
-	      void *userdata)
+	      struct xt_table *table)
 {
 	static const char nulldevname[IFNAMSIZ] __attribute__((aligned(sizeof(long))));
 	int offset = 0;
@@ -349,8 +347,7 @@ ip6t_do_table(struct sk_buff **pskb,
 								     in, out,
 								     hook,
 								     t->u.kernel.target,
-								     t->data,
-								     userdata);
+								     t->data);
 
 #ifdef CONFIG_NETFILTER_DEBUG
 				if (((struct ip6t_entry *)table_base)->comefrom
diff --git a/net/ipv6/netfilter/ip6t_HL.c b/net/ipv6/netfilter/ip6t_HL.c
index b8eff8ee69b1..c85d124f9a3d 100644
--- a/net/ipv6/netfilter/ip6t_HL.c
+++ b/net/ipv6/netfilter/ip6t_HL.c
@@ -22,7 +22,7 @@ static unsigned int ip6t_hl_target(struct sk_buff **pskb,
 				   const struct net_device *out,
 				   unsigned int hooknum,
 				   const struct xt_target *target,
-				   const void *targinfo, void *userinfo)
+				   const void *targinfo)
 {
 	struct ipv6hdr *ip6h;
 	const struct ip6t_HL_info *info = targinfo;
diff --git a/net/ipv6/netfilter/ip6t_LOG.c b/net/ipv6/netfilter/ip6t_LOG.c
index 73c6300109d6..acb91733e1fd 100644
--- a/net/ipv6/netfilter/ip6t_LOG.c
+++ b/net/ipv6/netfilter/ip6t_LOG.c
@@ -427,8 +427,7 @@ ip6t_log_target(struct sk_buff **pskb,
 		const struct net_device *out,
 		unsigned int hooknum,
 		const struct xt_target *target,
-		const void *targinfo,
-		void *userinfo)
+		const void *targinfo)
 {
 	const struct ip6t_log_info *loginfo = targinfo;
 	struct nf_loginfo li;
diff --git a/net/ipv6/netfilter/ip6t_REJECT.c b/net/ipv6/netfilter/ip6t_REJECT.c
index 7929ff402166..343acd3cbf5e 100644
--- a/net/ipv6/netfilter/ip6t_REJECT.c
+++ b/net/ipv6/netfilter/ip6t_REJECT.c
@@ -180,8 +180,7 @@ static unsigned int reject6_target(struct sk_buff **pskb,
 			   const struct net_device *out,
 			   unsigned int hooknum,
 			   const struct xt_target *target,
-			   const void *targinfo,
-			   void *userinfo)
+			   const void *targinfo)
 {
 	const struct ip6t_reject_info *reject = targinfo;
 
diff --git a/net/ipv6/netfilter/ip6table_filter.c b/net/ipv6/netfilter/ip6table_filter.c
index 60976c0c58e8..2fc07c74decf 100644
--- a/net/ipv6/netfilter/ip6table_filter.c
+++ b/net/ipv6/netfilter/ip6table_filter.c
@@ -108,7 +108,7 @@ ip6t_hook(unsigned int hook,
 	 const struct net_device *out,
 	 int (*okfn)(struct sk_buff *))
 {
-	return ip6t_do_table(pskb, hook, in, out, &packet_filter, NULL);
+	return ip6t_do_table(pskb, hook, in, out, &packet_filter);
 }
 
 static unsigned int
@@ -128,7 +128,7 @@ ip6t_local_out_hook(unsigned int hook,
 	}
 #endif
 
-	return ip6t_do_table(pskb, hook, in, out, &packet_filter, NULL);
+	return ip6t_do_table(pskb, hook, in, out, &packet_filter);
 }
 
 static struct nf_hook_ops ip6t_ops[] = {
diff --git a/net/ipv6/netfilter/ip6table_mangle.c b/net/ipv6/netfilter/ip6table_mangle.c
index 03a13eab1dae..32db04fd8310 100644
--- a/net/ipv6/netfilter/ip6table_mangle.c
+++ b/net/ipv6/netfilter/ip6table_mangle.c
@@ -138,7 +138,7 @@ ip6t_route_hook(unsigned int hook,
 	 const struct net_device *out,
 	 int (*okfn)(struct sk_buff *))
 {
-	return ip6t_do_table(pskb, hook, in, out, &packet_mangler, NULL);
+	return ip6t_do_table(pskb, hook, in, out, &packet_mangler);
 }
 
 static unsigned int
@@ -174,7 +174,7 @@ ip6t_local_hook(unsigned int hook,
 	/* flowlabel and prio (includes version, which shouldn't change either */
 	flowlabel = *((u_int32_t *) (*pskb)->nh.ipv6h);
 
-	ret = ip6t_do_table(pskb, hook, in, out, &packet_mangler, NULL);
+	ret = ip6t_do_table(pskb, hook, in, out, &packet_mangler);
 
 	if (ret != NF_DROP && ret != NF_STOLEN 
 		&& (memcmp(&(*pskb)->nh.ipv6h->saddr, &saddr, sizeof(saddr))
diff --git a/net/ipv6/netfilter/ip6table_raw.c b/net/ipv6/netfilter/ip6table_raw.c
index 61a7c58e99f8..b4154da575c0 100644
--- a/net/ipv6/netfilter/ip6table_raw.c
+++ b/net/ipv6/netfilter/ip6table_raw.c
@@ -122,7 +122,7 @@ ip6t_hook(unsigned int hook,
 	 const struct net_device *out,
 	 int (*okfn)(struct sk_buff *))
 {
-	return ip6t_do_table(pskb, hook, in, out, &packet_raw, NULL);
+	return ip6t_do_table(pskb, hook, in, out, &packet_raw);
 }
 
 static struct nf_hook_ops ip6t_ops[] = { 
diff --git a/net/netfilter/xt_CLASSIFY.c b/net/netfilter/xt_CLASSIFY.c
index 1f92edd05933..50de965bb104 100644
--- a/net/netfilter/xt_CLASSIFY.c
+++ b/net/netfilter/xt_CLASSIFY.c
@@ -29,8 +29,7 @@ target(struct sk_buff **pskb,
        const struct net_device *out,
        unsigned int hooknum,
        const struct xt_target *target,
-       const void *targinfo,
-       void *userinfo)
+       const void *targinfo)
 {
 	const struct xt_classify_target_info *clinfo = targinfo;
 
diff --git a/net/netfilter/xt_CONNMARK.c b/net/netfilter/xt_CONNMARK.c
index e577356b5c71..c2125f6ee128 100644
--- a/net/netfilter/xt_CONNMARK.c
+++ b/net/netfilter/xt_CONNMARK.c
@@ -38,8 +38,7 @@ target(struct sk_buff **pskb,
        const struct net_device *out,
        unsigned int hooknum,
        const struct xt_target *target,
-       const void *targinfo,
-       void *userinfo)
+       const void *targinfo)
 {
 	const struct xt_connmark_target_info *markinfo = targinfo;
 	u_int32_t diff;
diff --git a/net/netfilter/xt_CONNSECMARK.c b/net/netfilter/xt_CONNSECMARK.c
index 48f7fc3c85cd..4b9cc65bb82b 100644
--- a/net/netfilter/xt_CONNSECMARK.c
+++ b/net/netfilter/xt_CONNSECMARK.c
@@ -66,7 +66,7 @@ static void secmark_restore(struct sk_buff *skb)
 static unsigned int target(struct sk_buff **pskb, const struct net_device *in,
 			   const struct net_device *out, unsigned int hooknum,
 			   const struct xt_target *target,
-			   const void *targinfo, void *userinfo)
+			   const void *targinfo)
 {
 	struct sk_buff *skb = *pskb;
 	const struct xt_connsecmark_target_info *info = targinfo;
diff --git a/net/netfilter/xt_DSCP.c b/net/netfilter/xt_DSCP.c
index a1cd9723644f..9d23c9580d80 100644
--- a/net/netfilter/xt_DSCP.c
+++ b/net/netfilter/xt_DSCP.c
@@ -32,8 +32,7 @@ static unsigned int target(struct sk_buff **pskb,
 			   const struct net_device *out,
 			   unsigned int hooknum,
 			   const struct xt_target *target,
-			   const void *targinfo,
-			   void *userinfo)
+			   const void *targinfo)
 {
 	const struct xt_DSCP_info *dinfo = targinfo;
 	u_int8_t dscp = ipv4_get_dsfield((*pskb)->nh.iph) >> XT_DSCP_SHIFT;
@@ -54,8 +53,7 @@ static unsigned int target6(struct sk_buff **pskb,
 			    const struct net_device *out,
 			    unsigned int hooknum,
 			    const struct xt_target *target,
-			    const void *targinfo,
-			    void *userinfo)
+			    const void *targinfo)
 {
 	const struct xt_DSCP_info *dinfo = targinfo;
 	u_int8_t dscp = ipv6_get_dsfield((*pskb)->nh.ipv6h) >> XT_DSCP_SHIFT;
diff --git a/net/netfilter/xt_MARK.c b/net/netfilter/xt_MARK.c
index 0a6127219467..95a171c87994 100644
--- a/net/netfilter/xt_MARK.c
+++ b/net/netfilter/xt_MARK.c
@@ -27,8 +27,7 @@ target_v0(struct sk_buff **pskb,
 	  const struct net_device *out,
 	  unsigned int hooknum,
 	  const struct xt_target *target,
-	  const void *targinfo,
-	  void *userinfo)
+	  const void *targinfo)
 {
 	const struct xt_mark_target_info *markinfo = targinfo;
 
@@ -44,8 +43,7 @@ target_v1(struct sk_buff **pskb,
 	  const struct net_device *out,
 	  unsigned int hooknum,
 	  const struct xt_target *target,
-	  const void *targinfo,
-	  void *userinfo)
+	  const void *targinfo)
 {
 	const struct xt_mark_target_info_v1 *markinfo = targinfo;
 	int mark = 0;
diff --git a/net/netfilter/xt_NFQUEUE.c b/net/netfilter/xt_NFQUEUE.c
index 7b982283abdb..db9b896e57c8 100644
--- a/net/netfilter/xt_NFQUEUE.c
+++ b/net/netfilter/xt_NFQUEUE.c
@@ -29,8 +29,7 @@ target(struct sk_buff **pskb,
        const struct net_device *out,
        unsigned int hooknum,
        const struct xt_target *target,
-       const void *targinfo,
-       void *userinfo)
+       const void *targinfo)
 {
 	const struct xt_NFQ_info *tinfo = targinfo;
 
diff --git a/net/netfilter/xt_NOTRACK.c b/net/netfilter/xt_NOTRACK.c
index cab881d4424c..6d00dcaed238 100644
--- a/net/netfilter/xt_NOTRACK.c
+++ b/net/netfilter/xt_NOTRACK.c
@@ -16,8 +16,7 @@ target(struct sk_buff **pskb,
        const struct net_device *out,
        unsigned int hooknum,
        const struct xt_target *target,
-       const void *targinfo,
-       void *userinfo)
+       const void *targinfo)
 {
 	/* Previously seen (loopback)? Ignore. */
 	if ((*pskb)->nfct != NULL)
diff --git a/net/netfilter/xt_SECMARK.c b/net/netfilter/xt_SECMARK.c
index 4300988786c9..8a04dcf2611e 100644
--- a/net/netfilter/xt_SECMARK.c
+++ b/net/netfilter/xt_SECMARK.c
@@ -31,7 +31,7 @@ static u8 mode;
 static unsigned int target(struct sk_buff **pskb, const struct net_device *in,
 			   const struct net_device *out, unsigned int hooknum,
 			   const struct xt_target *target,
-			   const void *targinfo, void *userinfo)
+			   const void *targinfo)
 {
 	u32 secmark = 0;
 	const struct xt_secmark_target_info *info = targinfo;
diff --git a/net/netfilter/xt_connbytes.c b/net/netfilter/xt_connbytes.c
index 2d49948d3c38..d725e8b84503 100644
--- a/net/netfilter/xt_connbytes.c
+++ b/net/netfilter/xt_connbytes.c
@@ -143,7 +143,7 @@ static int check(const char *tablename,
 	return 1;
 }
 
-static struct xt_match xt_connbytes_match = {
+static struct xt_match xt_connbytes_match[] = {
 	{
 		.name		= "connbytes",
 		.family		= AF_INET,
diff --git a/net/sched/act_ipt.c b/net/sched/act_ipt.c
index 224c078a398e..45a3143b8629 100644
--- a/net/sched/act_ipt.c
+++ b/net/sched/act_ipt.c
@@ -222,7 +222,7 @@ static int tcf_ipt(struct sk_buff *skb, struct tc_action *a,
 	ret = ipt->tcfi_t->u.kernel.target->target(&skb, skb->dev, NULL,
 						   ipt->tcfi_hook,
 						   ipt->tcfi_t->u.kernel.target,
-						   ipt->tcfi_t->data, NULL);
+						   ipt->tcfi_t->data);
 	switch (ret) {
 	case NF_ACCEPT:
 		result = TC_ACT_OK;
-- 
cgit v1.2.3


From efa741656e9ebf5fd6e0432b0d1b3c7f156392d3 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 22 Aug 2006 00:36:37 -0700
Subject: [NETFILTER]: x_tables: remove unused size argument to check/destroy
 functions

The size is verified by x_tables and isn't needed by the modules anymore.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/x_tables.h   |  8 ++------
 net/ipv4/netfilter/arp_tables.c      |  5 +----
 net/ipv4/netfilter/arpt_mangle.c     |  2 +-
 net/ipv4/netfilter/ip_nat_rule.c     |  2 --
 net/ipv4/netfilter/ip_tables.c       | 14 +++-----------
 net/ipv4/netfilter/ipt_CLUSTERIP.c   |  4 +---
 net/ipv4/netfilter/ipt_ECN.c         |  1 -
 net/ipv4/netfilter/ipt_LOG.c         |  1 -
 net/ipv4/netfilter/ipt_MASQUERADE.c  |  1 -
 net/ipv4/netfilter/ipt_NETMAP.c      |  1 -
 net/ipv4/netfilter/ipt_REDIRECT.c    |  1 -
 net/ipv4/netfilter/ipt_REJECT.c      |  1 -
 net/ipv4/netfilter/ipt_SAME.c        |  4 +---
 net/ipv4/netfilter/ipt_TCPMSS.c      |  1 -
 net/ipv4/netfilter/ipt_TOS.c         |  1 -
 net/ipv4/netfilter/ipt_TTL.c         |  1 -
 net/ipv4/netfilter/ipt_ULOG.c        |  1 -
 net/ipv4/netfilter/ipt_ah.c          |  1 -
 net/ipv4/netfilter/ipt_ecn.c         |  3 +--
 net/ipv4/netfilter/ipt_hashlimit.c   |  4 +---
 net/ipv4/netfilter/ipt_owner.c       |  1 -
 net/ipv4/netfilter/ipt_recent.c      |  5 ++---
 net/ipv6/netfilter/ip6_tables.c      | 10 ++--------
 net/ipv6/netfilter/ip6t_HL.c         |  1 -
 net/ipv6/netfilter/ip6t_LOG.c        |  1 -
 net/ipv6/netfilter/ip6t_REJECT.c     |  1 -
 net/ipv6/netfilter/ip6t_ah.c         |  1 -
 net/ipv6/netfilter/ip6t_dst.c        |  1 -
 net/ipv6/netfilter/ip6t_frag.c       |  1 -
 net/ipv6/netfilter/ip6t_hbh.c        |  1 -
 net/ipv6/netfilter/ip6t_ipv6header.c |  1 -
 net/ipv6/netfilter/ip6t_owner.c      |  1 -
 net/ipv6/netfilter/ip6t_rt.c         |  1 -
 net/netfilter/xt_CONNMARK.c          |  1 -
 net/netfilter/xt_CONNSECMARK.c       |  2 +-
 net/netfilter/xt_DSCP.c              |  1 -
 net/netfilter/xt_MARK.c              |  2 --
 net/netfilter/xt_SECMARK.c           |  2 +-
 net/netfilter/xt_connbytes.c         |  1 -
 net/netfilter/xt_connmark.c          |  3 +--
 net/netfilter/xt_conntrack.c         |  3 +--
 net/netfilter/xt_dccp.c              |  1 -
 net/netfilter/xt_dscp.c              |  1 -
 net/netfilter/xt_esp.c               |  1 -
 net/netfilter/xt_helper.c            |  3 +--
 net/netfilter/xt_limit.c             |  1 -
 net/netfilter/xt_mark.c              |  1 -
 net/netfilter/xt_multiport.c         |  4 ----
 net/netfilter/xt_physdev.c           |  1 -
 net/netfilter/xt_policy.c            |  3 +--
 net/netfilter/xt_quota.c             |  2 +-
 net/netfilter/xt_sctp.c              |  1 -
 net/netfilter/xt_state.c             |  3 +--
 net/netfilter/xt_statistic.c         |  2 +-
 net/netfilter/xt_string.c            |  4 +---
 net/netfilter/xt_tcpudp.c            |  2 --
 net/sched/act_ipt.c                  |  4 +---
 57 files changed, 26 insertions(+), 106 deletions(-)

(limited to 'include')

diff --git a/include/linux/netfilter/x_tables.h b/include/linux/netfilter/x_tables.h
index 9cef0e91542b..9d97102a9347 100644
--- a/include/linux/netfilter/x_tables.h
+++ b/include/linux/netfilter/x_tables.h
@@ -174,12 +174,10 @@ struct xt_match
 			  const void *ip,
 			  const struct xt_match *match,
 			  void *matchinfo,
-			  unsigned int matchinfosize,
 			  unsigned int hook_mask);
 
 	/* Called when entry of this type deleted. */
-	void (*destroy)(const struct xt_match *match, void *matchinfo,
-			unsigned int matchinfosize);
+	void (*destroy)(const struct xt_match *match, void *matchinfo);
 
 	/* Called when userspace align differs from kernel space one */
 	int (*compat)(void *match, void **dstptr, int *size, int convert);
@@ -221,12 +219,10 @@ struct xt_target
 			  const void *entry,
 			  const struct xt_target *target,
 			  void *targinfo,
-			  unsigned int targinfosize,
 			  unsigned int hook_mask);
 
 	/* Called when entry of this type deleted. */
-	void (*destroy)(const struct xt_target *target, void *targinfo,
-			unsigned int targinfosize);
+	void (*destroy)(const struct xt_target *target, void *targinfo);
 
 	/* Called when userspace align differs from kernel space one */
 	int (*compat)(void *target, void **dstptr, int *size, int convert);
diff --git a/net/ipv4/netfilter/arp_tables.c b/net/ipv4/netfilter/arp_tables.c
index c6bd270bf46a..4f10b06413a1 100644
--- a/net/ipv4/netfilter/arp_tables.c
+++ b/net/ipv4/netfilter/arp_tables.c
@@ -491,8 +491,6 @@ static inline int check_entry(struct arpt_entry *e, const char *name, unsigned i
 		}
 	} else if (t->u.kernel.target->checkentry
 		   && !t->u.kernel.target->checkentry(name, e, target, t->data,
-						      t->u.target_size
-						      - sizeof(*t),
 						      e->comefrom)) {
 		duprintf("arp_tables: check failed for `%s'.\n",
 			 t->u.kernel.target->name);
@@ -559,8 +557,7 @@ static inline int cleanup_entry(struct arpt_entry *e, unsigned int *i)
 
 	t = arpt_get_target(e);
 	if (t->u.kernel.target->destroy)
-		t->u.kernel.target->destroy(t->u.kernel.target, t->data,
-					    t->u.target_size - sizeof(*t));
+		t->u.kernel.target->destroy(t->u.kernel.target, t->data);
 	module_put(t->u.kernel.target->me);
 	return 0;
 }
diff --git a/net/ipv4/netfilter/arpt_mangle.c b/net/ipv4/netfilter/arpt_mangle.c
index 05fb2421bb26..d12b1df252a1 100644
--- a/net/ipv4/netfilter/arpt_mangle.c
+++ b/net/ipv4/netfilter/arpt_mangle.c
@@ -67,7 +67,7 @@ target(struct sk_buff **pskb,
 
 static int
 checkentry(const char *tablename, const void *e, const struct xt_target *target,
-           void *targinfo, unsigned int targinfosize, unsigned int hook_mask)
+           void *targinfo, unsigned int hook_mask)
 {
 	const struct arpt_mangle *mangle = targinfo;
 
diff --git a/net/ipv4/netfilter/ip_nat_rule.c b/net/ipv4/netfilter/ip_nat_rule.c
index 1aa0e4f462a5..e59f5a8ecb6b 100644
--- a/net/ipv4/netfilter/ip_nat_rule.c
+++ b/net/ipv4/netfilter/ip_nat_rule.c
@@ -172,7 +172,6 @@ static int ipt_snat_checkentry(const char *tablename,
 			       const void *entry,
 			       const struct ipt_target *target,
 			       void *targinfo,
-			       unsigned int targinfosize,
 			       unsigned int hook_mask)
 {
 	struct ip_nat_multi_range_compat *mr = targinfo;
@@ -189,7 +188,6 @@ static int ipt_dnat_checkentry(const char *tablename,
 			       const void *entry,
 			       const struct ipt_target *target,
 			       void *targinfo,
-			       unsigned int targinfosize,
 			       unsigned int hook_mask)
 {
 	struct ip_nat_multi_range_compat *mr = targinfo;
diff --git a/net/ipv4/netfilter/ip_tables.c b/net/ipv4/netfilter/ip_tables.c
index 8ce5b6f76447..a0f36806998c 100644
--- a/net/ipv4/netfilter/ip_tables.c
+++ b/net/ipv4/netfilter/ip_tables.c
@@ -464,8 +464,7 @@ cleanup_match(struct ipt_entry_match *m, unsigned int *i)
 		return 1;
 
 	if (m->u.kernel.match->destroy)
-		m->u.kernel.match->destroy(m->u.kernel.match, m->data,
-					   m->u.match_size - sizeof(*m));
+		m->u.kernel.match->destroy(m->u.kernel.match, m->data);
 	module_put(m->u.kernel.match->me);
 	return 0;
 }
@@ -518,7 +517,6 @@ check_match(struct ipt_entry_match *m,
 
 	if (m->u.kernel.match->checkentry
 	    && !m->u.kernel.match->checkentry(name, ip, match, m->data,
-					      m->u.match_size - sizeof(*m),
 					      hookmask)) {
 		duprintf("ip_tables: check failed for `%s'.\n",
 			 m->u.kernel.match->name);
@@ -579,8 +577,6 @@ check_entry(struct ipt_entry *e, const char *name, unsigned int size,
 		}
 	} else if (t->u.kernel.target->checkentry
 		   && !t->u.kernel.target->checkentry(name, e, target, t->data,
-						      t->u.target_size
-						      - sizeof(*t),
 						      e->comefrom)) {
 		duprintf("ip_tables: check failed for `%s'.\n",
 			 t->u.kernel.target->name);
@@ -652,8 +648,7 @@ cleanup_entry(struct ipt_entry *e, unsigned int *i)
 	IPT_MATCH_ITERATE(e, cleanup_match, NULL);
 	t = ipt_get_target(e);
 	if (t->u.kernel.target->destroy)
-		t->u.kernel.target->destroy(t->u.kernel.target, t->data,
-					    t->u.target_size - sizeof(*t));
+		t->u.kernel.target->destroy(t->u.kernel.target, t->data);
 	module_put(t->u.kernel.target->me);
 	return 0;
 }
@@ -1599,7 +1594,6 @@ static inline int compat_copy_match_from_user(struct ipt_entry_match *m,
 
 	if (m->u.kernel.match->checkentry
 	    && !m->u.kernel.match->checkentry(name, ip, match, dm->data,
-					      dm->u.match_size - sizeof(*dm),
 					      hookmask)) {
 		duprintf("ip_tables: check failed for `%s'.\n",
 			 m->u.kernel.match->name);
@@ -1658,8 +1652,7 @@ static int compat_copy_entry_from_user(struct ipt_entry *e, void **dstptr,
 			goto out;
 	} else if (t->u.kernel.target->checkentry
 		   && !t->u.kernel.target->checkentry(name, de, target,
-				t->data, t->u.target_size - sizeof(*t),
-				de->comefrom)) {
+						      t->data, de->comefrom)) {
 		duprintf("ip_tables: compat: check failed for `%s'.\n",
 			 t->u.kernel.target->name);
 		goto out;
@@ -2182,7 +2175,6 @@ icmp_checkentry(const char *tablename,
 	   const void *info,
 	   const struct xt_match *match,
 	   void *matchinfo,
-	   unsigned int matchsize,
 	   unsigned int hook_mask)
 {
 	const struct ipt_icmp *icmpinfo = matchinfo;
diff --git a/net/ipv4/netfilter/ipt_CLUSTERIP.c b/net/ipv4/netfilter/ipt_CLUSTERIP.c
index a08383cf9e7a..41589665fc5d 100644
--- a/net/ipv4/netfilter/ipt_CLUSTERIP.c
+++ b/net/ipv4/netfilter/ipt_CLUSTERIP.c
@@ -372,7 +372,6 @@ checkentry(const char *tablename,
 	   const void *e_void,
 	   const struct xt_target *target,
            void *targinfo,
-           unsigned int targinfosize,
            unsigned int hook_mask)
 {
 	struct ipt_clusterip_tgt_info *cipinfo = targinfo;
@@ -449,8 +448,7 @@ checkentry(const char *tablename,
 }
 
 /* drop reference count of cluster config when rule is deleted */
-static void destroy(const struct xt_target *target, void *targinfo,
-		    unsigned int targinfosize)
+static void destroy(const struct xt_target *target, void *targinfo)
 {
 	struct ipt_clusterip_tgt_info *cipinfo = targinfo;
 
diff --git a/net/ipv4/netfilter/ipt_ECN.c b/net/ipv4/netfilter/ipt_ECN.c
index 1c3da4a48e5f..23f9c7ebe7eb 100644
--- a/net/ipv4/netfilter/ipt_ECN.c
+++ b/net/ipv4/netfilter/ipt_ECN.c
@@ -106,7 +106,6 @@ checkentry(const char *tablename,
 	   const void *e_void,
 	   const struct xt_target *target,
            void *targinfo,
-           unsigned int targinfosize,
            unsigned int hook_mask)
 {
 	const struct ipt_ECN_info *einfo = (struct ipt_ECN_info *)targinfo;
diff --git a/net/ipv4/netfilter/ipt_LOG.c b/net/ipv4/netfilter/ipt_LOG.c
index a8d356c6191f..7dc820df8bc5 100644
--- a/net/ipv4/netfilter/ipt_LOG.c
+++ b/net/ipv4/netfilter/ipt_LOG.c
@@ -439,7 +439,6 @@ static int ipt_log_checkentry(const char *tablename,
 			      const void *e,
 			      const struct xt_target *target,
 			      void *targinfo,
-			      unsigned int targinfosize,
 			      unsigned int hook_mask)
 {
 	const struct ipt_log_info *loginfo = targinfo;
diff --git a/net/ipv4/netfilter/ipt_MASQUERADE.c b/net/ipv4/netfilter/ipt_MASQUERADE.c
index 9659793c66c0..bc65168a3437 100644
--- a/net/ipv4/netfilter/ipt_MASQUERADE.c
+++ b/net/ipv4/netfilter/ipt_MASQUERADE.c
@@ -42,7 +42,6 @@ masquerade_check(const char *tablename,
 		 const void *e,
 		 const struct xt_target *target,
 		 void *targinfo,
-		 unsigned int targinfosize,
 		 unsigned int hook_mask)
 {
 	const struct ip_nat_multi_range_compat *mr = targinfo;
diff --git a/net/ipv4/netfilter/ipt_NETMAP.c b/net/ipv4/netfilter/ipt_NETMAP.c
index fd5e74a19fb5..beb2914225ff 100644
--- a/net/ipv4/netfilter/ipt_NETMAP.c
+++ b/net/ipv4/netfilter/ipt_NETMAP.c
@@ -33,7 +33,6 @@ check(const char *tablename,
       const void *e,
       const struct xt_target *target,
       void *targinfo,
-      unsigned int targinfosize,
       unsigned int hook_mask)
 {
 	const struct ip_nat_multi_range_compat *mr = targinfo;
diff --git a/net/ipv4/netfilter/ipt_REDIRECT.c b/net/ipv4/netfilter/ipt_REDIRECT.c
index 839fe99f71d4..f03d43671c6d 100644
--- a/net/ipv4/netfilter/ipt_REDIRECT.c
+++ b/net/ipv4/netfilter/ipt_REDIRECT.c
@@ -36,7 +36,6 @@ redirect_check(const char *tablename,
 	       const void *e,
 	       const struct xt_target *target,
 	       void *targinfo,
-	       unsigned int targinfosize,
 	       unsigned int hook_mask)
 {
 	const struct ip_nat_multi_range_compat *mr = targinfo;
diff --git a/net/ipv4/netfilter/ipt_REJECT.c b/net/ipv4/netfilter/ipt_REJECT.c
index 1dfd8e56be8b..b81821edd893 100644
--- a/net/ipv4/netfilter/ipt_REJECT.c
+++ b/net/ipv4/netfilter/ipt_REJECT.c
@@ -276,7 +276,6 @@ static int check(const char *tablename,
 		 const void *e_void,
 		 const struct xt_target *target,
 		 void *targinfo,
-		 unsigned int targinfosize,
 		 unsigned int hook_mask)
 {
  	const struct ipt_reject_info *rejinfo = targinfo;
diff --git a/net/ipv4/netfilter/ipt_SAME.c b/net/ipv4/netfilter/ipt_SAME.c
index cf801749490f..efbcb1198832 100644
--- a/net/ipv4/netfilter/ipt_SAME.c
+++ b/net/ipv4/netfilter/ipt_SAME.c
@@ -52,7 +52,6 @@ same_check(const char *tablename,
 	      const void *e,
 	      const struct xt_target *target,
 	      void *targinfo,
-	      unsigned int targinfosize,
 	      unsigned int hook_mask)
 {
 	unsigned int count, countess, rangeip, index = 0;
@@ -116,8 +115,7 @@ same_check(const char *tablename,
 }
 
 static void 
-same_destroy(const struct xt_target *target, void *targinfo,
-		unsigned int targinfosize)
+same_destroy(const struct xt_target *target, void *targinfo)
 {
 	struct ipt_same_info *mr = targinfo;
 
diff --git a/net/ipv4/netfilter/ipt_TCPMSS.c b/net/ipv4/netfilter/ipt_TCPMSS.c
index 6d668dcfc22a..ac8a35eeea3f 100644
--- a/net/ipv4/netfilter/ipt_TCPMSS.c
+++ b/net/ipv4/netfilter/ipt_TCPMSS.c
@@ -207,7 +207,6 @@ ipt_tcpmss_checkentry(const char *tablename,
 		      const void *e_void,
 		      const struct xt_target *target,
 		      void *targinfo,
-		      unsigned int targinfosize,
 		      unsigned int hook_mask)
 {
 	const struct ipt_tcpmss_info *tcpmssinfo = targinfo;
diff --git a/net/ipv4/netfilter/ipt_TOS.c b/net/ipv4/netfilter/ipt_TOS.c
index 043df0137084..471a4c438b0a 100644
--- a/net/ipv4/netfilter/ipt_TOS.c
+++ b/net/ipv4/netfilter/ipt_TOS.c
@@ -49,7 +49,6 @@ checkentry(const char *tablename,
 	   const void *e_void,
 	   const struct xt_target *target,
            void *targinfo,
-           unsigned int targinfosize,
            unsigned int hook_mask)
 {
 	const u_int8_t tos = ((struct ipt_tos_target_info *)targinfo)->tos;
diff --git a/net/ipv4/netfilter/ipt_TTL.c b/net/ipv4/netfilter/ipt_TTL.c
index 164007107b5e..214d9d9c428f 100644
--- a/net/ipv4/netfilter/ipt_TTL.c
+++ b/net/ipv4/netfilter/ipt_TTL.c
@@ -67,7 +67,6 @@ static int ipt_ttl_checkentry(const char *tablename,
 		const void *e,
 		const struct xt_target *target,
 		void *targinfo,
-		unsigned int targinfosize,
 		unsigned int hook_mask)
 {
 	struct ipt_TTL_info *info = targinfo;
diff --git a/net/ipv4/netfilter/ipt_ULOG.c b/net/ipv4/netfilter/ipt_ULOG.c
index 4c5f0a117862..2b104ea54f48 100644
--- a/net/ipv4/netfilter/ipt_ULOG.c
+++ b/net/ipv4/netfilter/ipt_ULOG.c
@@ -346,7 +346,6 @@ static int ipt_ulog_checkentry(const char *tablename,
 			       const void *e,
 			       const struct xt_target *target,
 			       void *targinfo,
-			       unsigned int targinfosize,
 			       unsigned int hookmask)
 {
 	struct ipt_ulog_info *loginfo = (struct ipt_ulog_info *) targinfo;
diff --git a/net/ipv4/netfilter/ipt_ah.c b/net/ipv4/netfilter/ipt_ah.c
index 2927135873d7..1798f86bc534 100644
--- a/net/ipv4/netfilter/ipt_ah.c
+++ b/net/ipv4/netfilter/ipt_ah.c
@@ -74,7 +74,6 @@ checkentry(const char *tablename,
 	   const void *ip_void,
 	   const struct xt_match *match,
 	   void *matchinfo,
-	   unsigned int matchinfosize,
 	   unsigned int hook_mask)
 {
 	const struct ipt_ah *ahinfo = matchinfo;
diff --git a/net/ipv4/netfilter/ipt_ecn.c b/net/ipv4/netfilter/ipt_ecn.c
index b28250414933..dafbdec0efc0 100644
--- a/net/ipv4/netfilter/ipt_ecn.c
+++ b/net/ipv4/netfilter/ipt_ecn.c
@@ -88,8 +88,7 @@ static int match(const struct sk_buff *skb,
 
 static int checkentry(const char *tablename, const void *ip_void,
 		      const struct xt_match *match,
-		      void *matchinfo, unsigned int matchsize,
-		      unsigned int hook_mask)
+		      void *matchinfo, unsigned int hook_mask)
 {
 	const struct ipt_ecn_info *info = matchinfo;
 	const struct ipt_ip *ip = ip_void;
diff --git a/net/ipv4/netfilter/ipt_hashlimit.c b/net/ipv4/netfilter/ipt_hashlimit.c
index 3bd2368e1fc9..b5b74b07370c 100644
--- a/net/ipv4/netfilter/ipt_hashlimit.c
+++ b/net/ipv4/netfilter/ipt_hashlimit.c
@@ -478,7 +478,6 @@ hashlimit_checkentry(const char *tablename,
 		     const void *inf,
 		     const struct xt_match *match,
 		     void *matchinfo,
-		     unsigned int matchsize,
 		     unsigned int hook_mask)
 {
 	struct ipt_hashlimit_info *r = matchinfo;
@@ -529,8 +528,7 @@ hashlimit_checkentry(const char *tablename,
 }
 
 static void
-hashlimit_destroy(const struct xt_match *match, void *matchinfo,
-		  unsigned int matchsize)
+hashlimit_destroy(const struct xt_match *match, void *matchinfo)
 {
 	struct ipt_hashlimit_info *r = matchinfo;
 
diff --git a/net/ipv4/netfilter/ipt_owner.c b/net/ipv4/netfilter/ipt_owner.c
index 5ac6ac023b5e..78c336f12a9e 100644
--- a/net/ipv4/netfilter/ipt_owner.c
+++ b/net/ipv4/netfilter/ipt_owner.c
@@ -56,7 +56,6 @@ checkentry(const char *tablename,
            const void *ip,
 	   const struct xt_match *match,
            void *matchinfo,
-           unsigned int matchsize,
            unsigned int hook_mask)
 {
 	const struct ipt_owner_info *info = matchinfo;
diff --git a/net/ipv4/netfilter/ipt_recent.c b/net/ipv4/netfilter/ipt_recent.c
index 682c0946201e..32ae8d7ac506 100644
--- a/net/ipv4/netfilter/ipt_recent.c
+++ b/net/ipv4/netfilter/ipt_recent.c
@@ -238,7 +238,7 @@ out:
 static int
 ipt_recent_checkentry(const char *tablename, const void *ip,
 		      const struct xt_match *match, void *matchinfo,
-		      unsigned int matchsize, unsigned int hook_mask)
+		      unsigned int hook_mask)
 {
 	const struct ipt_recent_info *info = matchinfo;
 	struct recent_table *t;
@@ -294,8 +294,7 @@ out:
 }
 
 static void
-ipt_recent_destroy(const struct xt_match *match, void *matchinfo,
-		   unsigned int matchsize)
+ipt_recent_destroy(const struct xt_match *match, void *matchinfo)
 {
 	const struct ipt_recent_info *info = matchinfo;
 	struct recent_table *t;
diff --git a/net/ipv6/netfilter/ip6_tables.c b/net/ipv6/netfilter/ip6_tables.c
index 38cd7ffda9a0..d1c315364ee7 100644
--- a/net/ipv6/netfilter/ip6_tables.c
+++ b/net/ipv6/netfilter/ip6_tables.c
@@ -504,8 +504,7 @@ cleanup_match(struct ip6t_entry_match *m, unsigned int *i)
 		return 1;
 
 	if (m->u.kernel.match->destroy)
-		m->u.kernel.match->destroy(m->u.kernel.match, m->data,
-					   m->u.match_size - sizeof(*m));
+		m->u.kernel.match->destroy(m->u.kernel.match, m->data);
 	module_put(m->u.kernel.match->me);
 	return 0;
 }
@@ -558,7 +557,6 @@ check_match(struct ip6t_entry_match *m,
 
 	if (m->u.kernel.match->checkentry
 	    && !m->u.kernel.match->checkentry(name, ipv6, match,  m->data,
-					      m->u.match_size - sizeof(*m),
 					      hookmask)) {
 		duprintf("ip_tables: check failed for `%s'.\n",
 			 m->u.kernel.match->name);
@@ -619,8 +617,6 @@ check_entry(struct ip6t_entry *e, const char *name, unsigned int size,
 		}
 	} else if (t->u.kernel.target->checkentry
 		   && !t->u.kernel.target->checkentry(name, e, target, t->data,
-						      t->u.target_size
-						      - sizeof(*t),
 						      e->comefrom)) {
 		duprintf("ip_tables: check failed for `%s'.\n",
 			 t->u.kernel.target->name);
@@ -692,8 +688,7 @@ cleanup_entry(struct ip6t_entry *e, unsigned int *i)
 	IP6T_MATCH_ITERATE(e, cleanup_match, NULL);
 	t = ip6t_get_target(e);
 	if (t->u.kernel.target->destroy)
-		t->u.kernel.target->destroy(t->u.kernel.target, t->data,
-					    t->u.target_size - sizeof(*t));
+		t->u.kernel.target->destroy(t->u.kernel.target, t->data);
 	module_put(t->u.kernel.target->me);
 	return 0;
 }
@@ -1349,7 +1344,6 @@ icmp6_checkentry(const char *tablename,
 	   const void *entry,
 	   const struct xt_match *match,
 	   void *matchinfo,
-	   unsigned int matchsize,
 	   unsigned int hook_mask)
 {
 	const struct ip6t_icmp *icmpinfo = matchinfo;
diff --git a/net/ipv6/netfilter/ip6t_HL.c b/net/ipv6/netfilter/ip6t_HL.c
index c85d124f9a3d..e54ea92d107b 100644
--- a/net/ipv6/netfilter/ip6t_HL.c
+++ b/net/ipv6/netfilter/ip6t_HL.c
@@ -66,7 +66,6 @@ static int ip6t_hl_checkentry(const char *tablename,
 		const void *entry,
 		const struct xt_target *target,
 		void *targinfo,
-		unsigned int targinfosize,
 		unsigned int hook_mask)
 {
 	struct ip6t_HL_info *info = targinfo;
diff --git a/net/ipv6/netfilter/ip6t_LOG.c b/net/ipv6/netfilter/ip6t_LOG.c
index acb91733e1fd..0cf537d30185 100644
--- a/net/ipv6/netfilter/ip6t_LOG.c
+++ b/net/ipv6/netfilter/ip6t_LOG.c
@@ -451,7 +451,6 @@ static int ip6t_log_checkentry(const char *tablename,
 			       const void *entry,
 			       const struct xt_target *target,
 			       void *targinfo,
-			       unsigned int targinfosize,
 			       unsigned int hook_mask)
 {
 	const struct ip6t_log_info *loginfo = targinfo;
diff --git a/net/ipv6/netfilter/ip6t_REJECT.c b/net/ipv6/netfilter/ip6t_REJECT.c
index 343acd3cbf5e..311eae82feb3 100644
--- a/net/ipv6/netfilter/ip6t_REJECT.c
+++ b/net/ipv6/netfilter/ip6t_REJECT.c
@@ -223,7 +223,6 @@ static int check(const char *tablename,
 		 const void *entry,
 		 const struct xt_target *target,
 		 void *targinfo,
-		 unsigned int targinfosize,
 		 unsigned int hook_mask)
 {
  	const struct ip6t_reject_info *rejinfo = targinfo;
diff --git a/net/ipv6/netfilter/ip6t_ah.c b/net/ipv6/netfilter/ip6t_ah.c
index 2f7bb20c758b..ec1b1608156c 100644
--- a/net/ipv6/netfilter/ip6t_ah.c
+++ b/net/ipv6/netfilter/ip6t_ah.c
@@ -102,7 +102,6 @@ checkentry(const char *tablename,
           const void *entry,
 	  const struct xt_match *match,
           void *matchinfo,
-          unsigned int matchinfosize,
           unsigned int hook_mask)
 {
 	const struct ip6t_ah *ahinfo = matchinfo;
diff --git a/net/ipv6/netfilter/ip6t_dst.c b/net/ipv6/netfilter/ip6t_dst.c
index 9422413d0571..223c335467cc 100644
--- a/net/ipv6/netfilter/ip6t_dst.c
+++ b/net/ipv6/netfilter/ip6t_dst.c
@@ -182,7 +182,6 @@ checkentry(const char *tablename,
 	   const void *info,
 	   const struct xt_match *match,
 	   void *matchinfo,
-	   unsigned int matchinfosize,
 	   unsigned int hook_mask)
 {
 	const struct ip6t_opts *optsinfo = matchinfo;
diff --git a/net/ipv6/netfilter/ip6t_frag.c b/net/ipv6/netfilter/ip6t_frag.c
index 06768c84bd31..78d9c8b9e28a 100644
--- a/net/ipv6/netfilter/ip6t_frag.c
+++ b/net/ipv6/netfilter/ip6t_frag.c
@@ -119,7 +119,6 @@ checkentry(const char *tablename,
 	   const void *ip,
 	   const struct xt_match *match,
 	   void *matchinfo,
-	   unsigned int matchinfosize,
 	   unsigned int hook_mask)
 {
 	const struct ip6t_frag *fraginfo = matchinfo;
diff --git a/net/ipv6/netfilter/ip6t_hbh.c b/net/ipv6/netfilter/ip6t_hbh.c
index 374f1be85c0d..72defc816563 100644
--- a/net/ipv6/netfilter/ip6t_hbh.c
+++ b/net/ipv6/netfilter/ip6t_hbh.c
@@ -182,7 +182,6 @@ checkentry(const char *tablename,
 	   const void *entry,
 	   const struct xt_match *match,
 	   void *matchinfo,
-	   unsigned int matchinfosize,
 	   unsigned int hook_mask)
 {
 	const struct ip6t_opts *optsinfo = matchinfo;
diff --git a/net/ipv6/netfilter/ip6t_ipv6header.c b/net/ipv6/netfilter/ip6t_ipv6header.c
index 9375eeb1369f..3093c398002f 100644
--- a/net/ipv6/netfilter/ip6t_ipv6header.c
+++ b/net/ipv6/netfilter/ip6t_ipv6header.c
@@ -128,7 +128,6 @@ ipv6header_checkentry(const char *tablename,
 		      const void *ip,
 		      const struct xt_match *match,
 		      void *matchinfo,
-		      unsigned int matchsize,
 		      unsigned int hook_mask)
 {
 	const struct ip6t_ipv6header_info *info = matchinfo;
diff --git a/net/ipv6/netfilter/ip6t_owner.c b/net/ipv6/netfilter/ip6t_owner.c
index 5d047990cd44..4eb9bbc4ebc3 100644
--- a/net/ipv6/netfilter/ip6t_owner.c
+++ b/net/ipv6/netfilter/ip6t_owner.c
@@ -57,7 +57,6 @@ checkentry(const char *tablename,
 	   const void *ip,
 	   const struct xt_match *match,
 	   void *matchinfo,
-	   unsigned int matchsize,
 	   unsigned int hook_mask)
 {
 	const struct ip6t_owner_info *info = matchinfo;
diff --git a/net/ipv6/netfilter/ip6t_rt.c b/net/ipv6/netfilter/ip6t_rt.c
index fbb0184a41d8..bcb2e168a5bc 100644
--- a/net/ipv6/netfilter/ip6t_rt.c
+++ b/net/ipv6/netfilter/ip6t_rt.c
@@ -197,7 +197,6 @@ checkentry(const char *tablename,
 	   const void *entry,
 	   const struct xt_match *match,
 	   void *matchinfo,
-	   unsigned int matchinfosize,
 	   unsigned int hook_mask)
 {
 	const struct ip6t_rt *rtinfo = matchinfo;
diff --git a/net/netfilter/xt_CONNMARK.c b/net/netfilter/xt_CONNMARK.c
index c2125f6ee128..0e4249ddc17b 100644
--- a/net/netfilter/xt_CONNMARK.c
+++ b/net/netfilter/xt_CONNMARK.c
@@ -89,7 +89,6 @@ checkentry(const char *tablename,
 	   const void *entry,
 	   const struct xt_target *target,
 	   void *targinfo,
-	   unsigned int targinfosize,
 	   unsigned int hook_mask)
 {
 	struct xt_connmark_target_info *matchinfo = targinfo;
diff --git a/net/netfilter/xt_CONNSECMARK.c b/net/netfilter/xt_CONNSECMARK.c
index 4b9cc65bb82b..4b0e14bb1726 100644
--- a/net/netfilter/xt_CONNSECMARK.c
+++ b/net/netfilter/xt_CONNSECMARK.c
@@ -89,7 +89,7 @@ static unsigned int target(struct sk_buff **pskb, const struct net_device *in,
 
 static int checkentry(const char *tablename, const void *entry,
 		      const struct xt_target *target, void *targinfo,
-		      unsigned int targinfosize, unsigned int hook_mask)
+		      unsigned int hook_mask)
 {
 	struct xt_connsecmark_target_info *info = targinfo;
 
diff --git a/net/netfilter/xt_DSCP.c b/net/netfilter/xt_DSCP.c
index 9d23c9580d80..a7cc75aeb38d 100644
--- a/net/netfilter/xt_DSCP.c
+++ b/net/netfilter/xt_DSCP.c
@@ -72,7 +72,6 @@ static int checkentry(const char *tablename,
 		      const void *e_void,
 		      const struct xt_target *target,
 		      void *targinfo,
-		      unsigned int targinfosize,
 		      unsigned int hook_mask)
 {
 	const u_int8_t dscp = ((struct xt_DSCP_info *)targinfo)->dscp;
diff --git a/net/netfilter/xt_MARK.c b/net/netfilter/xt_MARK.c
index 95a171c87994..782f8d8c3edf 100644
--- a/net/netfilter/xt_MARK.c
+++ b/net/netfilter/xt_MARK.c
@@ -74,7 +74,6 @@ checkentry_v0(const char *tablename,
 	      const void *entry,
 	      const struct xt_target *target,
 	      void *targinfo,
-	      unsigned int targinfosize,
 	      unsigned int hook_mask)
 {
 	struct xt_mark_target_info *markinfo = targinfo;
@@ -91,7 +90,6 @@ checkentry_v1(const char *tablename,
 	      const void *entry,
 	      const struct xt_target *target,
 	      void *targinfo,
-	      unsigned int targinfosize,
 	      unsigned int hook_mask)
 {
 	struct xt_mark_target_info_v1 *markinfo = targinfo;
diff --git a/net/netfilter/xt_SECMARK.c b/net/netfilter/xt_SECMARK.c
index 8a04dcf2611e..451b67c4bb53 100644
--- a/net/netfilter/xt_SECMARK.c
+++ b/net/netfilter/xt_SECMARK.c
@@ -85,7 +85,7 @@ static int checkentry_selinux(struct xt_secmark_target_info *info)
 
 static int checkentry(const char *tablename, const void *entry,
 		      const struct xt_target *target, void *targinfo,
-		      unsigned int targinfosize, unsigned int hook_mask)
+		      unsigned int hook_mask)
 {
 	struct xt_secmark_target_info *info = targinfo;
 
diff --git a/net/netfilter/xt_connbytes.c b/net/netfilter/xt_connbytes.c
index d725e8b84503..dcc497ea8183 100644
--- a/net/netfilter/xt_connbytes.c
+++ b/net/netfilter/xt_connbytes.c
@@ -125,7 +125,6 @@ static int check(const char *tablename,
 		 const void *ip,
 		 const struct xt_match *match,
 		 void *matchinfo,
-		 unsigned int matchsize,
 		 unsigned int hook_mask)
 {
 	const struct xt_connbytes_info *sinfo = matchinfo;
diff --git a/net/netfilter/xt_connmark.c b/net/netfilter/xt_connmark.c
index a97b2d455b79..c9104d05a19c 100644
--- a/net/netfilter/xt_connmark.c
+++ b/net/netfilter/xt_connmark.c
@@ -55,7 +55,6 @@ checkentry(const char *tablename,
 	   const void *ip,
 	   const struct xt_match *match,
 	   void *matchinfo,
-	   unsigned int matchsize,
 	   unsigned int hook_mask)
 {
 	struct xt_connmark_info *cm = matchinfo;
@@ -75,7 +74,7 @@ checkentry(const char *tablename,
 }
 
 static void
-destroy(const struct xt_match *match, void *matchinfo, unsigned int matchsize)
+destroy(const struct xt_match *match, void *matchinfo)
 {
 #if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
 	nf_ct_l3proto_module_put(match->family);
diff --git a/net/netfilter/xt_conntrack.c b/net/netfilter/xt_conntrack.c
index 1540885174ee..39c57e9f7563 100644
--- a/net/netfilter/xt_conntrack.c
+++ b/net/netfilter/xt_conntrack.c
@@ -208,7 +208,6 @@ checkentry(const char *tablename,
 	   const void *ip,
 	   const struct xt_match *match,
 	   void *matchinfo,
-	   unsigned int matchsize,
 	   unsigned int hook_mask)
 {
 #if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
@@ -222,7 +221,7 @@ checkentry(const char *tablename,
 }
 
 static void
-destroy(const struct xt_match *match, void *matchinfo, unsigned int matchsize)
+destroy(const struct xt_match *match, void *matchinfo)
 {
 #if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
 	nf_ct_l3proto_module_put(match->family);
diff --git a/net/netfilter/xt_dccp.c b/net/netfilter/xt_dccp.c
index 5ca6f5288f46..3e6cf430e518 100644
--- a/net/netfilter/xt_dccp.c
+++ b/net/netfilter/xt_dccp.c
@@ -131,7 +131,6 @@ checkentry(const char *tablename,
 	   const void *inf,
 	   const struct xt_match *match,
 	   void *matchinfo,
-	   unsigned int matchsize,
 	   unsigned int hook_mask)
 {
 	const struct xt_dccp_info *info = matchinfo;
diff --git a/net/netfilter/xt_dscp.c b/net/netfilter/xt_dscp.c
index d84075c30159..26c7f4ad102a 100644
--- a/net/netfilter/xt_dscp.c
+++ b/net/netfilter/xt_dscp.c
@@ -58,7 +58,6 @@ static int checkentry(const char *tablename,
 		      const void *info,
 		      const struct xt_match *match,
 		      void *matchinfo,
-		      unsigned int matchsize,
 		      unsigned int hook_mask)
 {
 	const u_int8_t dscp = ((struct xt_dscp_info *)matchinfo)->dscp;
diff --git a/net/netfilter/xt_esp.c b/net/netfilter/xt_esp.c
index 7b19bc9ea205..7c95f149d942 100644
--- a/net/netfilter/xt_esp.c
+++ b/net/netfilter/xt_esp.c
@@ -79,7 +79,6 @@ checkentry(const char *tablename,
 	   const void *ip_void,
 	   const struct xt_match *match,
 	   void *matchinfo,
-	   unsigned int matchinfosize,
 	   unsigned int hook_mask)
 {
 	const struct xt_esp *espinfo = matchinfo;
diff --git a/net/netfilter/xt_helper.c b/net/netfilter/xt_helper.c
index db453a7a154e..5d7818b73e3a 100644
--- a/net/netfilter/xt_helper.c
+++ b/net/netfilter/xt_helper.c
@@ -139,7 +139,6 @@ static int check(const char *tablename,
 		 const void *inf,
 		 const struct xt_match *match,
 		 void *matchinfo,
-		 unsigned int matchsize,
 		 unsigned int hook_mask)
 {
 	struct xt_helper_info *info = matchinfo;
@@ -156,7 +155,7 @@ static int check(const char *tablename,
 }
 
 static void
-destroy(const struct xt_match *match, void *matchinfo, unsigned int matchsize)
+destroy(const struct xt_match *match, void *matchinfo)
 {
 #if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
 	nf_ct_l3proto_module_put(match->family);
diff --git a/net/netfilter/xt_limit.c b/net/netfilter/xt_limit.c
index e8d5e7ac695a..b9c9ff3a06ea 100644
--- a/net/netfilter/xt_limit.c
+++ b/net/netfilter/xt_limit.c
@@ -110,7 +110,6 @@ ipt_limit_checkentry(const char *tablename,
 		     const void *inf,
 		     const struct xt_match *match,
 		     void *matchinfo,
-		     unsigned int matchsize,
 		     unsigned int hook_mask)
 {
 	struct xt_rateinfo *r = matchinfo;
diff --git a/net/netfilter/xt_mark.c b/net/netfilter/xt_mark.c
index 39f9b079f5d4..e8059cd17275 100644
--- a/net/netfilter/xt_mark.c
+++ b/net/netfilter/xt_mark.c
@@ -39,7 +39,6 @@ checkentry(const char *tablename,
            const void *entry,
 	   const struct xt_match *match,
            void *matchinfo,
-           unsigned int matchsize,
            unsigned int hook_mask)
 {
 	const struct xt_mark_info *minfo = matchinfo;
diff --git a/net/netfilter/xt_multiport.c b/net/netfilter/xt_multiport.c
index e74f9bb98b3c..d3aefd380930 100644
--- a/net/netfilter/xt_multiport.c
+++ b/net/netfilter/xt_multiport.c
@@ -176,7 +176,6 @@ checkentry(const char *tablename,
 	   const void *info,
 	   const struct xt_match *match,
 	   void *matchinfo,
-	   unsigned int matchsize,
 	   unsigned int hook_mask)
 {
 	const struct ipt_ip *ip = info;
@@ -191,7 +190,6 @@ checkentry_v1(const char *tablename,
 	      const void *info,
 	      const struct xt_match *match,
 	      void *matchinfo,
-	      unsigned int matchsize,
 	      unsigned int hook_mask)
 {
 	const struct ipt_ip *ip = info;
@@ -206,7 +204,6 @@ checkentry6(const char *tablename,
 	    const void *info,
 	    const struct xt_match *match,
 	    void *matchinfo,
-	    unsigned int matchsize,
 	    unsigned int hook_mask)
 {
 	const struct ip6t_ip6 *ip = info;
@@ -221,7 +218,6 @@ checkentry6_v1(const char *tablename,
 	       const void *info,
 	       const struct xt_match *match,
 	       void *matchinfo,
-	       unsigned int matchsize,
 	       unsigned int hook_mask)
 {
 	const struct ip6t_ip6 *ip = info;
diff --git a/net/netfilter/xt_physdev.c b/net/netfilter/xt_physdev.c
index af3d70f96ecd..fd8f954cded5 100644
--- a/net/netfilter/xt_physdev.c
+++ b/net/netfilter/xt_physdev.c
@@ -106,7 +106,6 @@ checkentry(const char *tablename,
 		       const void *ip,
 		       const struct xt_match *match,
 		       void *matchinfo,
-		       unsigned int matchsize,
 		       unsigned int hook_mask)
 {
 	const struct xt_physdev_info *info = matchinfo;
diff --git a/net/netfilter/xt_policy.c b/net/netfilter/xt_policy.c
index f5639c451112..e9d81378d653 100644
--- a/net/netfilter/xt_policy.c
+++ b/net/netfilter/xt_policy.c
@@ -135,8 +135,7 @@ static int match(const struct sk_buff *skb,
 
 static int checkentry(const char *tablename, const void *ip_void,
                       const struct xt_match *match,
-                      void *matchinfo, unsigned int matchsize,
-                      unsigned int hook_mask)
+                      void *matchinfo, unsigned int hook_mask)
 {
 	struct xt_policy_info *info = matchinfo;
 
diff --git a/net/netfilter/xt_quota.c b/net/netfilter/xt_quota.c
index cc44f87cb8e6..b75fa2c70e66 100644
--- a/net/netfilter/xt_quota.c
+++ b/net/netfilter/xt_quota.c
@@ -41,7 +41,7 @@ match(const struct sk_buff *skb,
 static int
 checkentry(const char *tablename, const void *entry,
 	   const struct xt_match *match, void *matchinfo,
-	   unsigned int matchsize, unsigned int hook_mask)
+	   unsigned int hook_mask)
 {
 	struct xt_quota_info *q = (struct xt_quota_info *)matchinfo;
 
diff --git a/net/netfilter/xt_sctp.c b/net/netfilter/xt_sctp.c
index 5628621170e6..7956acaaa24b 100644
--- a/net/netfilter/xt_sctp.c
+++ b/net/netfilter/xt_sctp.c
@@ -163,7 +163,6 @@ checkentry(const char *tablename,
 	   const void *inf,
 	   const struct xt_match *match,
 	   void *matchinfo,
-	   unsigned int matchsize,
 	   unsigned int hook_mask)
 {
 	const struct xt_sctp_info *info = matchinfo;
diff --git a/net/netfilter/xt_state.c b/net/netfilter/xt_state.c
index 5f9492e3b2b1..d9010b16a1f9 100644
--- a/net/netfilter/xt_state.c
+++ b/net/netfilter/xt_state.c
@@ -48,7 +48,6 @@ static int check(const char *tablename,
 		 const void *inf,
 		 const struct xt_match *match,
 		 void *matchinfo,
-		 unsigned int matchsize,
 		 unsigned int hook_mask)
 {
 #if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
@@ -62,7 +61,7 @@ static int check(const char *tablename,
 }
 
 static void
-destroy(const struct xt_match *match, void *matchinfo, unsigned int matchsize)
+destroy(const struct xt_match *match, void *matchinfo)
 {
 #if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
 	nf_ct_l3proto_module_put(match->family);
diff --git a/net/netfilter/xt_statistic.c b/net/netfilter/xt_statistic.c
index 5181630a87fc..091a9f89f5d5 100644
--- a/net/netfilter/xt_statistic.c
+++ b/net/netfilter/xt_statistic.c
@@ -55,7 +55,7 @@ match(const struct sk_buff *skb,
 static int
 checkentry(const char *tablename, const void *entry,
 	   const struct xt_match *match, void *matchinfo,
-	   unsigned int matchsize, unsigned int hook_mask)
+	   unsigned int hook_mask)
 {
 	struct xt_statistic_info *info = (struct xt_statistic_info *)matchinfo;
 
diff --git a/net/netfilter/xt_string.c b/net/netfilter/xt_string.c
index 1a1c1d17d85e..4453252400aa 100644
--- a/net/netfilter/xt_string.c
+++ b/net/netfilter/xt_string.c
@@ -46,7 +46,6 @@ static int checkentry(const char *tablename,
 		      const void *ip,
 		      const struct xt_match *match,
 		      void *matchinfo,
-		      unsigned int matchsize,
 		      unsigned int hook_mask)
 {
 	struct xt_string_info *conf = matchinfo;
@@ -69,8 +68,7 @@ static int checkentry(const char *tablename,
 	return 1;
 }
 
-static void destroy(const struct xt_match *match, void *matchinfo,
-		    unsigned int matchsize)
+static void destroy(const struct xt_match *match, void *matchinfo)
 {
 	textsearch_destroy(STRING_TEXT_PRIV(matchinfo)->config);
 }
diff --git a/net/netfilter/xt_tcpudp.c b/net/netfilter/xt_tcpudp.c
index 54aab051af86..e76a68e0bc66 100644
--- a/net/netfilter/xt_tcpudp.c
+++ b/net/netfilter/xt_tcpudp.c
@@ -141,7 +141,6 @@ tcp_checkentry(const char *tablename,
 	       const void *info,
 	       const struct xt_match *match,
 	       void *matchinfo,
-	       unsigned int matchsize,
 	       unsigned int hook_mask)
 {
 	const struct xt_tcp *tcpinfo = matchinfo;
@@ -190,7 +189,6 @@ udp_checkentry(const char *tablename,
 	       const void *info,
 	       const struct xt_match *match,
 	       void *matchinfo,
-	       unsigned int matchsize,
 	       unsigned int hook_mask)
 {
 	const struct xt_tcp *udpinfo = matchinfo;
diff --git a/net/sched/act_ipt.c b/net/sched/act_ipt.c
index 45a3143b8629..d8c9310da6e5 100644
--- a/net/sched/act_ipt.c
+++ b/net/sched/act_ipt.c
@@ -69,7 +69,6 @@ static int ipt_init_target(struct ipt_entry_target *t, char *table, unsigned int
 	if (t->u.kernel.target->checkentry
 	    && !t->u.kernel.target->checkentry(table, NULL,
 		    			       t->u.kernel.target, t->data,
-					       t->u.target_size - sizeof(*t),
 					       hook)) {
 		module_put(t->u.kernel.target->me);
 		ret = -EINVAL;
@@ -81,8 +80,7 @@ static int ipt_init_target(struct ipt_entry_target *t, char *table, unsigned int
 static void ipt_destroy_target(struct ipt_entry_target *t)
 {
 	if (t->u.kernel.target->destroy)
-		t->u.kernel.target->destroy(t->u.kernel.target, t->data,
-		                            t->u.target_size - sizeof(*t));
+		t->u.kernel.target->destroy(t->u.kernel.target, t->data);
         module_put(t->u.kernel.target->me);
 }
 
-- 
cgit v1.2.3


From 53e26658282373b84ba85a0c9807cb762f7738a6 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 22 Aug 2006 00:43:20 -0700
Subject: [NETFILTER]: nfnetlink: remove unnecessary packed attributes

Remove unnecessary packed attributes in nfnetlink structures. Unfortunately
in a few cases they have to stay to avoid changing structure sizes.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nfnetlink.h       | 4 ++--
 include/linux/netfilter/nfnetlink_log.h   | 6 +++---
 include/linux/netfilter/nfnetlink_queue.h | 8 ++++----
 3 files changed, 9 insertions(+), 9 deletions(-)

(limited to 'include')

diff --git a/include/linux/netfilter/nfnetlink.h b/include/linux/netfilter/nfnetlink.h
index 9f5b12cf489b..6d8e3e5a80e9 100644
--- a/include/linux/netfilter/nfnetlink.h
+++ b/include/linux/netfilter/nfnetlink.h
@@ -43,7 +43,7 @@ struct nfattr
 	u_int16_t nfa_len;
 	u_int16_t nfa_type;	/* we use 15 bits for the type, and the highest
 				 * bit to indicate whether the payload is nested */
-} __attribute__ ((packed));
+};
 
 /* FIXME: Apart from NFNL_NFA_NESTED shamelessly copy and pasted from
  * rtnetlink.h, it's time to put this in a generic file */
@@ -79,7 +79,7 @@ struct nfgenmsg {
 	u_int8_t  nfgen_family;		/* AF_xxx */
 	u_int8_t  version;		/* nfnetlink version */
 	u_int16_t res_id;		/* resource id */
-} __attribute__ ((packed));
+};
 
 #define NFNETLINK_V0	0
 
diff --git a/include/linux/netfilter/nfnetlink_log.h b/include/linux/netfilter/nfnetlink_log.h
index a7497c7436df..87b92f8b988f 100644
--- a/include/linux/netfilter/nfnetlink_log.h
+++ b/include/linux/netfilter/nfnetlink_log.h
@@ -19,18 +19,18 @@ struct nfulnl_msg_packet_hdr {
 	u_int16_t	hw_protocol;	/* hw protocol (network order) */
 	u_int8_t	hook;		/* netfilter hook */
 	u_int8_t	_pad;
-} __attribute__ ((packed));
+};
 
 struct nfulnl_msg_packet_hw {
 	u_int16_t	hw_addrlen;
 	u_int16_t	_pad;
 	u_int8_t	hw_addr[8];
-} __attribute__ ((packed));
+};
 
 struct nfulnl_msg_packet_timestamp {
 	aligned_u64	sec;
 	aligned_u64	usec;
-} __attribute__ ((packed));
+};
 
 #define NFULNL_PREFIXLEN	30	/* just like old log target */
 
diff --git a/include/linux/netfilter/nfnetlink_queue.h b/include/linux/netfilter/nfnetlink_queue.h
index 9e774373244c..36af0360b56d 100644
--- a/include/linux/netfilter/nfnetlink_queue.h
+++ b/include/linux/netfilter/nfnetlink_queue.h
@@ -22,12 +22,12 @@ struct nfqnl_msg_packet_hw {
 	u_int16_t	hw_addrlen;
 	u_int16_t	_pad;
 	u_int8_t	hw_addr[8];
-} __attribute__ ((packed));
+};
 
 struct nfqnl_msg_packet_timestamp {
 	aligned_u64	sec;
 	aligned_u64	usec;
-} __attribute__ ((packed));
+};
 
 enum nfqnl_attr_type {
 	NFQA_UNSPEC,
@@ -49,7 +49,7 @@ enum nfqnl_attr_type {
 struct nfqnl_msg_verdict_hdr {
 	u_int32_t verdict;
 	u_int32_t id;
-} __attribute__ ((packed));
+};
 
 
 enum nfqnl_msg_config_cmds {
@@ -64,7 +64,7 @@ struct nfqnl_msg_config_cmd {
 	u_int8_t	command;	/* nfqnl_msg_config_cmds */
 	u_int8_t	_pad;
 	u_int16_t	pf;		/* AF_xxx for PF_[UN]BIND */
-} __attribute__ ((packed));
+};
 
 enum nfqnl_config_mode {
 	NFQNL_COPY_NONE,
-- 
cgit v1.2.3


From 91270cf81765152f6e77953440beb4d3b34a71b5 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 22 Aug 2006 00:43:38 -0700
Subject: [NETFILTER]: x_tables: add data member to struct xt_match

Shared match functions can use this to make runtime decisions basen on the
used match.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/x_tables.h | 3 +++
 1 file changed, 3 insertions(+)

(limited to 'include')

diff --git a/include/linux/netfilter/x_tables.h b/include/linux/netfilter/x_tables.h
index 9d97102a9347..03d1027fb0e8 100644
--- a/include/linux/netfilter/x_tables.h
+++ b/include/linux/netfilter/x_tables.h
@@ -185,6 +185,9 @@ struct xt_match
 	/* Set this to THIS_MODULE if you are a module, otherwise NULL */
 	struct module *me;
 
+	/* Free to use by each match */
+	unsigned long data;
+
 	char *table;
 	unsigned int matchsize;
 	unsigned int hooks;
-- 
cgit v1.2.3


From 3fd091e73b81f131e1567c4d4a1ec042940bf2f7 Mon Sep 17 00:00:00 2001
From: Vladislav Yasevich <vladislav.yasevich@hp.com>
Date: Tue, 22 Aug 2006 13:29:17 -0700
Subject: [SCTP]: Remove multiple levels of msecs to jiffies conversions.

The SCTP sysctl entries are displayed in milliseconds, but stored
internally in jiffies. This results in multiple levels of msecs to
jiffies conversion and as a result produces a truncation error. This
patch makes things consistent in that we store and display defaults
in milliseconds and only convert once for use by association.
This patch also adds some sane min/max values so that we don't go off
the deep end.

Signed-off-by: Vladislav Yasevich <vladislav.yasevich@hp.com>
Signed-off-by: Sridhar Samudrala <sri@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/sctp/constants.h |  13 ++--
 include/net/sctp/structs.h   |  12 ++--
 net/sctp/protocol.c          |   2 +-
 net/sctp/socket.c            |  15 +++--
 net/sctp/sysctl.c            | 140 +++++++++++++++++++------------------------
 net/sctp/transport.c         |   2 +-
 6 files changed, 84 insertions(+), 100 deletions(-)

(limited to 'include')

diff --git a/include/net/sctp/constants.h b/include/net/sctp/constants.h
index 57166bfdf8eb..6c632e26f72d 100644
--- a/include/net/sctp/constants.h
+++ b/include/net/sctp/constants.h
@@ -264,10 +264,10 @@ enum { SCTP_MAX_DUP_TSNS = 16 };
 enum { SCTP_MAX_GABS = 16 };
 
 /* Heartbeat interval - 30 secs */
-#define SCTP_DEFAULT_TIMEOUT_HEARTBEAT	(30 * HZ)
+#define SCTP_DEFAULT_TIMEOUT_HEARTBEAT	(30*1000)
 
 /* Delayed sack timer - 200ms */
-#define SCTP_DEFAULT_TIMEOUT_SACK	((200 * HZ) / 1000)
+#define SCTP_DEFAULT_TIMEOUT_SACK	(200)
 
 /* RTO.Initial              - 3  seconds
  * RTO.Min                  - 1  second
@@ -275,9 +275,9 @@ enum { SCTP_MAX_GABS = 16 };
  * RTO.Alpha                - 1/8
  * RTO.Beta                 - 1/4
  */
-#define SCTP_RTO_INITIAL	(3 * HZ)
-#define SCTP_RTO_MIN		(1 * HZ)
-#define SCTP_RTO_MAX		(60 * HZ)
+#define SCTP_RTO_INITIAL	(3 * 1000)
+#define SCTP_RTO_MIN		(1 * 1000)
+#define SCTP_RTO_MAX		(60 * 1000)
 
 #define SCTP_RTO_ALPHA          3   /* 1/8 when converted to right shifts. */
 #define SCTP_RTO_BETA           2   /* 1/4 when converted to right shifts. */
@@ -290,8 +290,7 @@ enum { SCTP_MAX_GABS = 16 };
 #define SCTP_DEF_MAX_INIT 6
 #define SCTP_DEF_MAX_SEND 10
 
-#define SCTP_DEFAULT_COOKIE_LIFE_SEC	60 /* seconds */
-#define SCTP_DEFAULT_COOKIE_LIFE_USEC	0  /* microseconds */
+#define SCTP_DEFAULT_COOKIE_LIFE	(60 * 1000) /* 60 seconds */
 
 #define SCTP_DEFAULT_MINWINDOW	1500	/* default minimum rwnd size */
 #define SCTP_DEFAULT_MAXWINDOW	65535	/* default rwnd size */
diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h
index 0412e730c765..c6d93bb0dcd2 100644
--- a/include/net/sctp/structs.h
+++ b/include/net/sctp/structs.h
@@ -128,9 +128,9 @@ extern struct sctp_globals {
 	 * RTO.Alpha		    - 1/8  (3 when converted to right shifts.)
 	 * RTO.Beta		    - 1/4  (2 when converted to right shifts.)
 	 */
-	unsigned long rto_initial;
-	unsigned long rto_min;
-	unsigned long rto_max;
+	unsigned int rto_initial;
+	unsigned int rto_min;
+	unsigned int rto_max;
 
 	/* Note: rto_alpha and rto_beta are really defined as inverse
 	 * powers of two to facilitate integer operations.
@@ -145,13 +145,13 @@ extern struct sctp_globals {
 	int cookie_preserve_enable;
 
 	/* Valid.Cookie.Life	    - 60  seconds  */
-	unsigned long valid_cookie_life;
+	unsigned int valid_cookie_life;
 
 	/* Delayed SACK timeout  200ms default*/
-	unsigned long sack_timeout;
+	unsigned int sack_timeout;
 
 	/* HB.interval		    - 30 seconds  */
-	unsigned long hb_interval;
+	unsigned int hb_interval;
 
 	/* Association.Max.Retrans  - 10 attempts
 	 * Path.Max.Retrans	    - 5	 attempts (per destination address)
diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c
index 1ab03a27a76e..5692ef5485d3 100644
--- a/net/sctp/protocol.c
+++ b/net/sctp/protocol.c
@@ -1049,7 +1049,7 @@ SCTP_STATIC __init int sctp_init(void)
 	sctp_rto_beta			= SCTP_RTO_BETA;
 
 	/* Valid.Cookie.Life        - 60  seconds */
-	sctp_valid_cookie_life		= 60 * HZ;
+	sctp_valid_cookie_life		= SCTP_DEFAULT_COOKIE_LIFE;
 
 	/* Whether Cookie Preservative is enabled(1) or not(0) */
 	sctp_cookie_preserve_enable 	= 1;
diff --git a/net/sctp/socket.c b/net/sctp/socket.c
index 3b6e82cb372f..7c1dbb1d10df 100644
--- a/net/sctp/socket.c
+++ b/net/sctp/socket.c
@@ -3045,14 +3045,14 @@ SCTP_STATIC int sctp_init_sock(struct sock *sk)
 	sp->initmsg.sinit_num_ostreams   = sctp_max_outstreams;
 	sp->initmsg.sinit_max_instreams  = sctp_max_instreams;
 	sp->initmsg.sinit_max_attempts   = sctp_max_retrans_init;
-	sp->initmsg.sinit_max_init_timeo = jiffies_to_msecs(sctp_rto_max);
+	sp->initmsg.sinit_max_init_timeo = sctp_rto_max;
 
 	/* Initialize default RTO related parameters.  These parameters can
 	 * be modified for with the SCTP_RTOINFO socket option.
 	 */
-	sp->rtoinfo.srto_initial = jiffies_to_msecs(sctp_rto_initial);
-	sp->rtoinfo.srto_max     = jiffies_to_msecs(sctp_rto_max);
-	sp->rtoinfo.srto_min     = jiffies_to_msecs(sctp_rto_min);
+	sp->rtoinfo.srto_initial = sctp_rto_initial;
+	sp->rtoinfo.srto_max     = sctp_rto_max;
+	sp->rtoinfo.srto_min     = sctp_rto_min;
 
 	/* Initialize default association related parameters. These parameters
 	 * can be modified with the SCTP_ASSOCINFO socket option.
@@ -3061,8 +3061,7 @@ SCTP_STATIC int sctp_init_sock(struct sock *sk)
 	sp->assocparams.sasoc_number_peer_destinations = 0;
 	sp->assocparams.sasoc_peer_rwnd = 0;
 	sp->assocparams.sasoc_local_rwnd = 0;
-	sp->assocparams.sasoc_cookie_life = 
-		jiffies_to_msecs(sctp_valid_cookie_life);
+	sp->assocparams.sasoc_cookie_life = sctp_valid_cookie_life;
 
 	/* Initialize default event subscriptions. By default, all the
 	 * options are off. 
@@ -3072,10 +3071,10 @@ SCTP_STATIC int sctp_init_sock(struct sock *sk)
 	/* Default Peer Address Parameters.  These defaults can
 	 * be modified via SCTP_PEER_ADDR_PARAMS
 	 */
-	sp->hbinterval  = jiffies_to_msecs(sctp_hb_interval);
+	sp->hbinterval  = sctp_hb_interval;
 	sp->pathmaxrxt  = sctp_max_retrans_path;
 	sp->pathmtu     = 0; // allow default discovery
-	sp->sackdelay   = jiffies_to_msecs(sctp_sack_timeout);
+	sp->sackdelay   = sctp_sack_timeout;
 	sp->param_flags = SPP_HB_ENABLE |
 	                  SPP_PMTUD_ENABLE |
 	                  SPP_SACKDELAY_ENABLE;
diff --git a/net/sctp/sysctl.c b/net/sctp/sysctl.c
index dc6f3ff32358..633cd178654b 100644
--- a/net/sctp/sysctl.c
+++ b/net/sctp/sysctl.c
@@ -45,9 +45,10 @@
 #include <net/sctp/sctp.h>
 #include <linux/sysctl.h>
 
-static ctl_handler sctp_sysctl_jiffies_ms;
-static long rto_timer_min = 1;
-static long rto_timer_max = 86400000; /* One day */
+static int zero = 0;
+static int one = 1;
+static int timer_max = 86400000; /* ms in one day */
+static int int_max = INT_MAX;
 static long sack_timer_min = 1;
 static long sack_timer_max = 500;
 
@@ -56,45 +57,45 @@ static ctl_table sctp_table[] = {
 		.ctl_name	= NET_SCTP_RTO_INITIAL,
 		.procname	= "rto_initial",
 		.data		= &sctp_rto_initial,
-		.maxlen		= sizeof(long),
+		.maxlen		= sizeof(unsigned int),
 		.mode		= 0644,
-		.proc_handler	= &proc_doulongvec_ms_jiffies_minmax,
-		.strategy	= &sctp_sysctl_jiffies_ms,
-		.extra1         = &rto_timer_min,
-		.extra2         = &rto_timer_max
+		.proc_handler	= &proc_dointvec_minmax,
+		.strategy	= &sysctl_intvec,
+		.extra1         = &one,
+		.extra2         = &timer_max
 	},
 	{
 		.ctl_name	= NET_SCTP_RTO_MIN,
 		.procname	= "rto_min",
 		.data		= &sctp_rto_min,
-		.maxlen		= sizeof(long),
+		.maxlen		= sizeof(unsigned int),
 		.mode		= 0644,
-		.proc_handler	= &proc_doulongvec_ms_jiffies_minmax,
-		.strategy	= &sctp_sysctl_jiffies_ms,
-		.extra1         = &rto_timer_min,
-		.extra2         = &rto_timer_max
+		.proc_handler	= &proc_dointvec_minmax,
+		.strategy	= &sysctl_intvec,
+		.extra1         = &one,
+		.extra2         = &timer_max
 	},
 	{
 		.ctl_name	= NET_SCTP_RTO_MAX,
 		.procname	= "rto_max",
 		.data		= &sctp_rto_max,
-		.maxlen		= sizeof(long),
+		.maxlen		= sizeof(unsigned int),
 		.mode		= 0644,
-		.proc_handler	= &proc_doulongvec_ms_jiffies_minmax,
-		.strategy	= &sctp_sysctl_jiffies_ms,
-		.extra1         = &rto_timer_min,
-		.extra2         = &rto_timer_max
+		.proc_handler	= &proc_dointvec_minmax,
+		.strategy	= &sysctl_intvec,
+		.extra1         = &one,
+		.extra2         = &timer_max
 	},
 	{
 		.ctl_name	= NET_SCTP_VALID_COOKIE_LIFE,
 		.procname	= "valid_cookie_life",
 		.data		= &sctp_valid_cookie_life,
-		.maxlen		= sizeof(long),
+		.maxlen		= sizeof(unsigned int),
 		.mode		= 0644,
-		.proc_handler	= &proc_doulongvec_ms_jiffies_minmax,
-		.strategy	= &sctp_sysctl_jiffies_ms,
-		.extra1         = &rto_timer_min,
-		.extra2         = &rto_timer_max
+		.proc_handler	= &proc_dointvec_minmax,
+		.strategy	= &sysctl_intvec,
+		.extra1         = &one,
+		.extra2         = &timer_max
 	},
 	{
 		.ctl_name	= NET_SCTP_MAX_BURST,
@@ -102,7 +103,10 @@ static ctl_table sctp_table[] = {
 		.data		= &sctp_max_burst,
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
-		.proc_handler	= &proc_dointvec
+		.proc_handler	= &proc_dointvec_minmax,
+		.strategy	= &sysctl_intvec,
+		.extra1		= &zero,
+		.extra2		= &int_max
 	},
 	{
 		.ctl_name	= NET_SCTP_ASSOCIATION_MAX_RETRANS,
@@ -110,7 +114,10 @@ static ctl_table sctp_table[] = {
 		.data		= &sctp_max_retrans_association,
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
-		.proc_handler	= &proc_dointvec
+		.proc_handler	= &proc_dointvec_minmax,
+		.strategy	= &sysctl_intvec,
+		.extra1		= &one,
+		.extra2		= &int_max
 	},
 	{
 		.ctl_name	= NET_SCTP_SNDBUF_POLICY,
@@ -118,7 +125,8 @@ static ctl_table sctp_table[] = {
 		.data		= &sctp_sndbuf_policy,
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
-		.proc_handler	= &proc_dointvec
+		.proc_handler	= &proc_dointvec,
+		.strategy	= &sysctl_intvec
 	},
 	{
 		.ctl_name	= NET_SCTP_RCVBUF_POLICY,
@@ -126,7 +134,8 @@ static ctl_table sctp_table[] = {
 		.data		= &sctp_rcvbuf_policy,
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
-		.proc_handler	= &proc_dointvec
+		.proc_handler	= &proc_dointvec,
+		.strategy	= &sysctl_intvec
 	},
 	{
 		.ctl_name	= NET_SCTP_PATH_MAX_RETRANS,
@@ -134,7 +143,10 @@ static ctl_table sctp_table[] = {
 		.data		= &sctp_max_retrans_path,
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
-		.proc_handler	= &proc_dointvec
+		.proc_handler	= &proc_dointvec_minmax,
+		.strategy	= &sysctl_intvec,
+		.extra1		= &one,
+		.extra2		= &int_max
 	},
 	{
 		.ctl_name	= NET_SCTP_MAX_INIT_RETRANSMITS,
@@ -142,18 +154,21 @@ static ctl_table sctp_table[] = {
 		.data		= &sctp_max_retrans_init,
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
-		.proc_handler	= &proc_dointvec
+		.proc_handler	= &proc_dointvec_minmax,
+		.strategy	= &sysctl_intvec,
+		.extra1		= &one,
+		.extra2		= &int_max
 	},
 	{
 		.ctl_name	= NET_SCTP_HB_INTERVAL,
 		.procname	= "hb_interval",
 		.data		= &sctp_hb_interval,
-		.maxlen		= sizeof(long),
+		.maxlen		= sizeof(unsigned int),
 		.mode		= 0644,
-		.proc_handler	= &proc_doulongvec_ms_jiffies_minmax,
-		.strategy	= &sctp_sysctl_jiffies_ms,
-		.extra1         = &rto_timer_min,
-		.extra2         = &rto_timer_max
+		.proc_handler	= &proc_dointvec_minmax,
+		.strategy	= &sysctl_intvec,
+		.extra1         = &one,
+		.extra2         = &timer_max
 	},
 	{
 		.ctl_name	= NET_SCTP_PRESERVE_ENABLE,
@@ -161,23 +176,26 @@ static ctl_table sctp_table[] = {
 		.data		= &sctp_cookie_preserve_enable,
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
-		.proc_handler	= &proc_dointvec
+		.proc_handler	= &proc_dointvec,
+		.strategy	= &sysctl_intvec
 	},
 	{
 		.ctl_name	= NET_SCTP_RTO_ALPHA,
 		.procname	= "rto_alpha_exp_divisor",
 		.data		= &sctp_rto_alpha,
 		.maxlen		= sizeof(int),
-		.mode		= 0644,
-		.proc_handler	= &proc_dointvec
+		.mode		= 0444,
+		.proc_handler	= &proc_dointvec,
+		.strategy	= &sysctl_intvec
 	},
 	{
 		.ctl_name	= NET_SCTP_RTO_BETA,
 		.procname	= "rto_beta_exp_divisor",
 		.data		= &sctp_rto_beta,
 		.maxlen		= sizeof(int),
-		.mode		= 0644,
-		.proc_handler	= &proc_dointvec
+		.mode		= 0444,
+		.proc_handler	= &proc_dointvec,
+		.strategy	= &sysctl_intvec
 	},
 	{
 		.ctl_name	= NET_SCTP_ADDIP_ENABLE,
@@ -185,7 +203,8 @@ static ctl_table sctp_table[] = {
 		.data		= &sctp_addip_enable,
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
-		.proc_handler	= &proc_dointvec
+		.proc_handler	= &proc_dointvec,
+		.strategy	= &sysctl_intvec
 	},
 	{
 		.ctl_name	= NET_SCTP_PRSCTP_ENABLE,
@@ -193,7 +212,8 @@ static ctl_table sctp_table[] = {
 		.data		= &sctp_prsctp_enable,
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
-		.proc_handler	= &proc_dointvec
+		.proc_handler	= &proc_dointvec,
+		.strategy	= &sysctl_intvec
 	},
 	{
 		.ctl_name	= NET_SCTP_SACK_TIMEOUT,
@@ -201,8 +221,8 @@ static ctl_table sctp_table[] = {
 		.data		= &sctp_sack_timeout,
 		.maxlen		= sizeof(long),
 		.mode		= 0644,
-		.proc_handler	= &proc_doulongvec_ms_jiffies_minmax,
-		.strategy	= &sctp_sysctl_jiffies_ms,
+		.proc_handler	= &proc_dointvec_minmax,
+		.strategy	= &sysctl_intvec,
 		.extra1         = &sack_timer_min,
 		.extra2         = &sack_timer_max,
 	},
@@ -242,37 +262,3 @@ void sctp_sysctl_unregister(void)
 {
 	unregister_sysctl_table(sctp_sysctl_header);
 }
-
-/* Strategy function to convert jiffies to milliseconds.  */
-static int sctp_sysctl_jiffies_ms(ctl_table *table, int __user *name, int nlen,
-		void __user *oldval, size_t __user *oldlenp,
-		void __user *newval, size_t newlen, void **context) {
-
-	if (oldval) {
-		size_t olen;
-
-		if (oldlenp) {
-			if (get_user(olen, oldlenp))
-				return -EFAULT;
-
-			if (olen != sizeof (int))
-				return -EINVAL;
-		}
-		if (put_user((*(int *)(table->data) * 1000) / HZ,
-			(int __user *)oldval) ||
-		    (oldlenp && put_user(sizeof (int), oldlenp)))
-			return -EFAULT;
-	}
-	if (newval && newlen) {
-		int new;
-
-		if (newlen != sizeof (int))
-			return -EINVAL;
-
-		if (get_user(new, (int __user *)newval))
-			return -EFAULT;
-
-		*(int *)(table->data) = (new * HZ) / 1000;
-	}
-	return 1;
-}
diff --git a/net/sctp/transport.c b/net/sctp/transport.c
index 2763aa93de1a..3e5936a5f671 100644
--- a/net/sctp/transport.c
+++ b/net/sctp/transport.c
@@ -75,7 +75,7 @@ static struct sctp_transport *sctp_transport_init(struct sctp_transport *peer,
 	 * parameter 'RTO.Initial'.
 	 */
 	peer->rtt = 0;
-	peer->rto = sctp_rto_initial;
+	peer->rto = msecs_to_jiffies(sctp_rto_initial);
 	peer->rttvar = 0;
 	peer->srtt = 0;
 	peer->rto_pending = 0;
-- 
cgit v1.2.3


From 5e032e32ecc2e6cb0385dc115ca9bfe5e19a9539 Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Wed, 23 Aug 2006 17:12:24 -0700
Subject: [IPV6] NDISC: Take source address into account for redirects.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: Ville Nuorvala <vnuorval@tcs.hut.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip6_route.h | 1 +
 net/ipv6/ndisc.c        | 3 ++-
 net/ipv6/route.c        | 5 +++--
 3 files changed, 6 insertions(+), 3 deletions(-)

(limited to 'include')

diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index 3f170f667c7b..249ce4545ef0 100644
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -110,6 +110,7 @@ extern int			rt6_route_rcv(struct net_device *dev,
 					      struct in6_addr *gwaddr);
 
 extern void			rt6_redirect(struct in6_addr *dest,
+					     struct in6_addr *src,
 					     struct in6_addr *saddr,
 					     struct neighbour *neigh,
 					     u8 *lladdr,
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index 419d65163819..32f28dec399e 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -1346,7 +1346,8 @@ static void ndisc_redirect_rcv(struct sk_buff *skb)
 
 	neigh = __neigh_lookup(&nd_tbl, target, skb->dev, 1);
 	if (neigh) {
-		rt6_redirect(dest, &skb->nh.ipv6h->saddr, neigh, lladdr, 
+		rt6_redirect(dest, &skb->nh.ipv6h->daddr,
+			     &skb->nh.ipv6h->saddr, neigh, lladdr,
 			     on_link);
 		neigh_release(neigh);
 	}
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 5d6e9083ca2c..a9b08a2422e0 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -1279,7 +1279,8 @@ static int ip6_route_del(struct fib6_config *cfg)
 /*
  *	Handle redirects
  */
-void rt6_redirect(struct in6_addr *dest, struct in6_addr *saddr,
+void rt6_redirect(struct in6_addr *dest, struct in6_addr *src,
+		  struct in6_addr *saddr,
 		  struct neighbour *neigh, u8 *lladdr, int on_link)
 {
 	struct rt6_info *rt, *nrt = NULL;
@@ -1304,7 +1305,7 @@ void rt6_redirect(struct in6_addr *dest, struct in6_addr *saddr,
 	 */
 
 	read_lock_bh(&table->tb6_lock);
-	fn = fib6_lookup(&table->tb6_root, dest, NULL);
+	fn = fib6_lookup(&table->tb6_root, dest, src);
 restart:
 	for (rt = fn->leaf; rt; rt = rt->u.next) {
 		/*
-- 
cgit v1.2.3


From 8e1ef0a95b87e8b4292b2ba733e8cb854ea2d2fe Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Tue, 29 Aug 2006 17:15:09 -0700
Subject: [IPV6]: Cache source address as well in ipv6_pinfo{}.

Based on MIPL2 kernel patch.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: Ville Nuorvala <vnuorval@tcs.hut.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/ipv6.h             | 3 +++
 include/net/ip6_route.h          | 9 ++++++---
 net/dccp/ipv6.c                  | 4 ++--
 net/ipv6/af_inet6.c              | 2 +-
 net/ipv6/datagram.c              | 7 ++++++-
 net/ipv6/inet6_connection_sock.c | 2 +-
 net/ipv6/ip6_output.c            | 3 +++
 net/ipv6/tcp_ipv6.c              | 4 ++--
 net/ipv6/udp.c                   | 7 ++++++-
 9 files changed, 30 insertions(+), 11 deletions(-)

(limited to 'include')

diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index 297853c841b4..02d14a3ff2af 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -242,6 +242,9 @@ struct ipv6_pinfo {
 	struct in6_addr 	rcv_saddr;
 	struct in6_addr		daddr;
 	struct in6_addr		*daddr_cache;
+#ifdef CONFIG_IPV6_SUBTREES
+	struct in6_addr		*saddr_cache;
+#endif
 
 	__u32			flow_label;
 	__u32			frag_size;
diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index 249ce4545ef0..0d40f84df21b 100644
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -144,21 +144,24 @@ extern rwlock_t rt6_lock;
  *	Store a destination cache entry in a socket
  */
 static inline void __ip6_dst_store(struct sock *sk, struct dst_entry *dst,
-				   struct in6_addr *daddr)
+				   struct in6_addr *daddr, struct in6_addr *saddr)
 {
 	struct ipv6_pinfo *np = inet6_sk(sk);
 	struct rt6_info *rt = (struct rt6_info *) dst;
 
 	sk_setup_caps(sk, dst);
 	np->daddr_cache = daddr;
+#ifdef CONFIG_IPV6_SUBTREES
+	np->saddr_cache = saddr;
+#endif
 	np->dst_cookie = rt->rt6i_node ? rt->rt6i_node->fn_sernum : 0;
 }
 
 static inline void ip6_dst_store(struct sock *sk, struct dst_entry *dst,
-				 struct in6_addr *daddr)
+				 struct in6_addr *daddr, struct in6_addr *saddr)
 {
 	write_lock(&sk->sk_dst_lock);
-	__ip6_dst_store(sk, dst, daddr);
+	__ip6_dst_store(sk, dst, daddr, saddr);
 	write_unlock(&sk->sk_dst_lock);
 }
 
diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c
index 231bc7c7e749..f9c5e12d7038 100644
--- a/net/dccp/ipv6.c
+++ b/net/dccp/ipv6.c
@@ -231,7 +231,7 @@ static int dccp_v6_connect(struct sock *sk, struct sockaddr *uaddr,
 	ipv6_addr_copy(&np->saddr, saddr);
 	inet->rcv_saddr = LOOPBACK4_IPV6;
 
-	__ip6_dst_store(sk, dst, NULL);
+	__ip6_dst_store(sk, dst, NULL, NULL);
 
 	icsk->icsk_ext_hdr_len = 0;
 	if (np->opt != NULL)
@@ -872,7 +872,7 @@ static struct sock *dccp_v6_request_recv_sock(struct sock *sk,
 	 * comment in that function for the gory details. -acme
 	 */
 
-	__ip6_dst_store(newsk, dst, NULL);
+	__ip6_dst_store(newsk, dst, NULL, NULL);
 	newsk->sk_route_caps = dst->dev->features & ~(NETIF_F_IP_CSUM |
 						      NETIF_F_TSO);
 	newdp6 = (struct dccp6_sock *)newsk;
diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c
index 2ff600cfe3a4..57ee5ddea96f 100644
--- a/net/ipv6/af_inet6.c
+++ b/net/ipv6/af_inet6.c
@@ -659,7 +659,7 @@ int inet6_sk_rebuild_header(struct sock *sk)
 			return err;
 		}
 
-		__ip6_dst_store(sk, dst, NULL);
+		__ip6_dst_store(sk, dst, NULL, NULL);
 	}
 
 	return 0;
diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c
index c73508e090a6..8561b9da6db6 100644
--- a/net/ipv6/datagram.c
+++ b/net/ipv6/datagram.c
@@ -193,7 +193,12 @@ ipv4_connected:
 
 	ip6_dst_store(sk, dst,
 		      ipv6_addr_equal(&fl.fl6_dst, &np->daddr) ?
-		      &np->daddr : NULL);
+		      &np->daddr : NULL,
+#ifdef CONFIG_IPV6_SUBTREES
+		      ipv6_addr_equal(&fl.fl6_src, &np->saddr) ?
+		      &np->saddr :
+#endif
+		      NULL);
 
 	sk->sk_state = TCP_ESTABLISHED;
 out:
diff --git a/net/ipv6/inet6_connection_sock.c b/net/ipv6/inet6_connection_sock.c
index 7a51a258615d..827f41d1478b 100644
--- a/net/ipv6/inet6_connection_sock.c
+++ b/net/ipv6/inet6_connection_sock.c
@@ -186,7 +186,7 @@ int inet6_csk_xmit(struct sk_buff *skb, int ipfragok)
 			return err;
 		}
 
-		__ip6_dst_store(sk, dst, NULL);
+		__ip6_dst_store(sk, dst, NULL, NULL);
 	}
 
 	skb->dst = dst_clone(dst);
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index 0a18cb6b1cbb..2a376b7d91b4 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -762,6 +762,9 @@ static struct dst_entry *ip6_sk_dst_check(struct sock *sk,
 	 * 2. oif also should be the same.
 	 */
 	if (ip6_rt_check(&rt->rt6i_dst, &fl->fl6_dst, np->daddr_cache) ||
+#ifdef CONFIG_IPV6_SUBTREES
+	    ip6_rt_check(&rt->rt6i_src, &fl->fl6_src, np->saddr_cache) ||
+#endif
 	    (fl->oif && fl->oif != dst->dev->ifindex)) {
 		dst_release(dst);
 		dst = NULL;
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 7f1b660493b7..2b18918f3011 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -272,7 +272,7 @@ static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr,
 	inet->rcv_saddr = LOOPBACK4_IPV6;
 
 	sk->sk_gso_type = SKB_GSO_TCPV6;
-	__ip6_dst_store(sk, dst, NULL);
+	__ip6_dst_store(sk, dst, NULL, NULL);
 
 	icsk->icsk_ext_hdr_len = 0;
 	if (np->opt)
@@ -954,7 +954,7 @@ static struct sock * tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb,
 	 */
 
 	newsk->sk_gso_type = SKB_GSO_TCPV6;
-	__ip6_dst_store(newsk, dst, NULL);
+	__ip6_dst_store(newsk, dst, NULL, NULL);
 
 	newtcp6sk = (struct tcp6_sock *)newsk;
 	inet_sk(newsk)->pinet6 = &newtcp6sk->inet6;
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index eb9e1b39c8f8..b9cc55ccb000 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -847,7 +847,12 @@ do_append_data:
 		if (connected) {
 			ip6_dst_store(sk, dst,
 				      ipv6_addr_equal(&fl->fl6_dst, &np->daddr) ?
-				      &np->daddr : NULL);
+				      &np->daddr : NULL,
+#ifdef CONFIG_IPV6_SUBTREES
+				      ipv6_addr_equal(&fl->fl6_src, &np->saddr) ?
+				      &np->saddr :
+#endif
+				      NULL);
 		} else {
 			dst_release(dst);
 		}
-- 
cgit v1.2.3


From 7fc33165a74301b2c5c90b2f2a1f6907cbd5c6f1 Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Wed, 23 Aug 2006 17:22:24 -0700
Subject: [IPV6] ROUTE: Put SUBTREE() as FIB6_SUBTREE() into ip6_fib.h for
 future use.

Based on MIPL2 kernel patch.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: Ville Nuorvala <vnuorval@tcs.hut.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip6_fib.h |  5 +++++
 net/ipv6/ip6_fib.c    | 20 +++++++++-----------
 2 files changed, 14 insertions(+), 11 deletions(-)

(limited to 'include')

diff --git a/include/net/ip6_fib.h b/include/net/ip6_fib.h
index 9610b887ffb5..6a3f26a04509 100644
--- a/include/net/ip6_fib.h
+++ b/include/net/ip6_fib.h
@@ -60,6 +60,11 @@ struct fib6_node
 	__u32			fn_sernum;
 };
 
+#ifndef CONFIG_IPV6_SUBTREES
+#define FIB6_SUBTREE(fn)	NULL
+#else
+#define FIB6_SUBTREE(fn)	((fn)->subtree)
+#endif
 
 /*
  *	routing information
diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
index b706424e70b8..6536e33d8353 100644
--- a/net/ipv6/ip6_fib.c
+++ b/net/ipv6/ip6_fib.c
@@ -73,10 +73,8 @@ static DEFINE_RWLOCK(fib6_walker_lock);
 
 #ifdef CONFIG_IPV6_SUBTREES
 #define FWS_INIT FWS_S
-#define SUBTREE(fn) ((fn)->subtree)
 #else
 #define FWS_INIT FWS_L
-#define SUBTREE(fn) NULL
 #endif
 
 static void fib6_prune_clones(struct fib6_node *fn, struct rt6_info *rt);
@@ -854,7 +852,7 @@ static struct fib6_node * fib6_lookup_1(struct fib6_node *root,
 	}
 
 	while(fn) {
-		if (SUBTREE(fn) || fn->fn_flags & RTN_RTINFO) {
+		if (FIB6_SUBTREE(fn) || fn->fn_flags & RTN_RTINFO) {
 			struct rt6key *key;
 
 			key = (struct rt6key *) ((u8 *) fn->leaf +
@@ -985,7 +983,7 @@ static struct rt6_info * fib6_find_prefix(struct fib6_node *fn)
 		if(fn->right)
 			return fn->right->leaf;
 
-		fn = SUBTREE(fn);
+		fn = FIB6_SUBTREE(fn);
 	}
 	return NULL;
 }
@@ -1016,7 +1014,7 @@ static struct fib6_node * fib6_repair_tree(struct fib6_node *fn)
 		if (fn->right) child = fn->right, children |= 1;
 		if (fn->left) child = fn->left, children |= 2;
 
-		if (children == 3 || SUBTREE(fn) 
+		if (children == 3 || FIB6_SUBTREE(fn)
 #ifdef CONFIG_IPV6_SUBTREES
 		    /* Subtree root (i.e. fn) may have one child */
 		    || (children && fn->fn_flags&RTN_ROOT)
@@ -1035,9 +1033,9 @@ static struct fib6_node * fib6_repair_tree(struct fib6_node *fn)
 
 		pn = fn->parent;
 #ifdef CONFIG_IPV6_SUBTREES
-		if (SUBTREE(pn) == fn) {
+		if (FIB6_SUBTREE(pn) == fn) {
 			BUG_TRAP(fn->fn_flags&RTN_ROOT);
-			SUBTREE(pn) = NULL;
+			FIB6_SUBTREE(pn) = NULL;
 			nstate = FWS_L;
 		} else {
 			BUG_TRAP(!(fn->fn_flags&RTN_ROOT));
@@ -1085,7 +1083,7 @@ static struct fib6_node * fib6_repair_tree(struct fib6_node *fn)
 		read_unlock(&fib6_walker_lock);
 
 		node_free(fn);
-		if (pn->fn_flags&RTN_RTINFO || SUBTREE(pn))
+		if (pn->fn_flags&RTN_RTINFO || FIB6_SUBTREE(pn))
 			return pn;
 
 		rt6_release(pn->leaf);
@@ -1228,8 +1226,8 @@ static int fib6_walk_continue(struct fib6_walker_t *w)
 		switch (w->state) {
 #ifdef CONFIG_IPV6_SUBTREES
 		case FWS_S:
-			if (SUBTREE(fn)) {
-				w->node = SUBTREE(fn);
+			if (FIB6_SUBTREE(fn)) {
+				w->node = FIB6_SUBTREE(fn);
 				continue;
 			}
 			w->state = FWS_L;
@@ -1263,7 +1261,7 @@ static int fib6_walk_continue(struct fib6_walker_t *w)
 			pn = fn->parent;
 			w->node = pn;
 #ifdef CONFIG_IPV6_SUBTREES
-			if (SUBTREE(pn) == fn) {
+			if (FIB6_SUBTREE(pn) == fn) {
 				BUG_TRAP(fn->fn_flags&RTN_ROOT);
 				w->state = FWS_L;
 				continue;
-- 
cgit v1.2.3


From 77d16f450ae0452d7d4b009f78debb1294fb435c Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Wed, 23 Aug 2006 17:25:05 -0700
Subject: [IPV6] ROUTE: Unify RT6_F_xxx and RT6_SELECT_F_xxx flags

Unify RT6_F_xxx and RT6_SELECT_F_xxx flags into
RT6_LOOKUP_F_xxx flags, and put them into ip6_route.h

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Acked-by: Ville Nuorvala <vnuorval@tcs.hut.fi
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip6_fib.h   |  3 ---
 include/net/ip6_route.h |  4 ++++
 net/ipv6/fib6_rules.c   |  2 +-
 net/ipv6/route.c        | 32 ++++++++++++--------------------
 4 files changed, 17 insertions(+), 24 deletions(-)

(limited to 'include')

diff --git a/include/net/ip6_fib.h b/include/net/ip6_fib.h
index 6a3f26a04509..e4438de3bd6b 100644
--- a/include/net/ip6_fib.h
+++ b/include/net/ip6_fib.h
@@ -173,9 +173,6 @@ struct fib6_table {
 #define RT6_TABLE_LOCAL		RT6_TABLE_MAIN
 #endif
 
-#define RT6_F_STRICT		1
-#define RT6_F_HAS_SADDR		2
-
 typedef struct rt6_info *(*pol_lookup_t)(struct fib6_table *,
 					 struct flowi *, int);
 
diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index 0d40f84df21b..297909570041 100644
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -32,6 +32,10 @@ struct route_info {
 #include <linux/ip.h>
 #include <linux/ipv6.h>
 
+#define RT6_LOOKUP_F_IFACE	0x1
+#define RT6_LOOKUP_F_REACHABLE	0x2
+#define RT6_LOOKUP_F_HAS_SADDR	0x4
+
 struct pol_chain {
 	int			type;
 	int			priority;
diff --git a/net/ipv6/fib6_rules.c b/net/ipv6/fib6_rules.c
index 2c4fbc855e6c..7b4908cc52b3 100644
--- a/net/ipv6/fib6_rules.c
+++ b/net/ipv6/fib6_rules.c
@@ -117,7 +117,7 @@ static int fib6_rule_match(struct fib_rule *rule, struct flowi *fl, int flags)
 	if (!ipv6_prefix_equal(&fl->fl6_dst, &r->dst.addr, r->dst.plen))
 		return 0;
 
-	if ((flags & RT6_F_HAS_SADDR) &&
+	if ((flags & RT6_LOOKUP_F_HAS_SADDR) &&
 	    !ipv6_prefix_equal(&fl->fl6_src, &r->src.addr, r->src.plen))
 		return 0;
 
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index fd6f2ec4fa09..20691285aee5 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -76,9 +76,6 @@
 
 #define CLONE_OFFLINK_ROUTE 0
 
-#define RT6_SELECT_F_IFACE	0x1
-#define RT6_SELECT_F_REACHABLE	0x2
-
 static int ip6_rt_max_size = 4096;
 static int ip6_rt_gc_min_interval = HZ / 2;
 static int ip6_rt_gc_timeout = 60*HZ;
@@ -340,7 +337,7 @@ static int rt6_score_route(struct rt6_info *rt, int oif,
 	int m, n;
 		
 	m = rt6_check_dev(rt, oif);
-	if (!m && (strict & RT6_SELECT_F_IFACE))
+	if (!m && (strict & RT6_LOOKUP_F_IFACE))
 		return -1;
 #ifdef CONFIG_IPV6_ROUTER_PREF
 	m |= IPV6_DECODE_PREF(IPV6_EXTRACT_PREF(rt->rt6i_flags)) << 2;
@@ -348,7 +345,7 @@ static int rt6_score_route(struct rt6_info *rt, int oif,
 	n = rt6_check_neigh(rt);
 	if (n > 1)
 		m |= 16;
-	else if (!n && strict & RT6_SELECT_F_REACHABLE)
+	else if (!n && strict & RT6_LOOKUP_F_REACHABLE)
 		return -1;
 	return m;
 }
@@ -388,7 +385,7 @@ static struct rt6_info *rt6_select(struct rt6_info **head, int oif,
 	}
 
 	if (!match &&
-	    (strict & RT6_SELECT_F_REACHABLE) &&
+	    (strict & RT6_LOOKUP_F_REACHABLE) &&
 	    last && last != rt0) {
 		/* no entries matched; do round-robin */
 		static DEFINE_SPINLOCK(lock);
@@ -511,7 +508,7 @@ static struct rt6_info *ip6_pol_route_lookup(struct fib6_table *table,
 	fn = fib6_lookup(&table->tb6_root, &fl->fl6_dst, &fl->fl6_src);
 restart:
 	rt = fn->leaf;
-	rt = rt6_device_match(rt, fl->oif, flags & RT6_F_STRICT);
+	rt = rt6_device_match(rt, fl->oif, flags);
 	BACKTRACK(&fl->fl6_src);
 	dst_hold(&rt->u.dst);
 out:
@@ -537,7 +534,7 @@ struct rt6_info *rt6_lookup(struct in6_addr *daddr, struct in6_addr *saddr,
 		},
 	};
 	struct dst_entry *dst;
-	int flags = strict ? RT6_F_STRICT : 0;
+	int flags = strict ? RT6_LOOKUP_F_IFACE : 0;
 
 	dst = fib6_rule_lookup(&fl, flags, ip6_pol_route_lookup);
 	if (dst->error == 0)
@@ -633,10 +630,9 @@ static struct rt6_info *ip6_pol_route_input(struct fib6_table *table,
 	int strict = 0;
 	int attempts = 3;
 	int err;
-	int reachable = RT6_SELECT_F_REACHABLE;
+	int reachable = RT6_LOOKUP_F_REACHABLE;
 
-	if (flags & RT6_F_STRICT)
-		strict = RT6_SELECT_F_IFACE;
+	strict |= flags & RT6_LOOKUP_F_IFACE;
 
 relookup:
 	read_lock_bh(&table->tb6_lock);
@@ -712,10 +708,7 @@ void ip6_route_input(struct sk_buff *skb)
 		},
 		.proto = iph->nexthdr,
 	};
-	int flags = 0;
-
-	if (rt6_need_strict(&iph->daddr))
-		flags |= RT6_F_STRICT;
+	int flags = rt6_need_strict(&iph->daddr) ? RT6_LOOKUP_F_IFACE : 0;
 
 	skb->dst = fib6_rule_lookup(&fl, flags, ip6_pol_route_input);
 }
@@ -728,10 +721,9 @@ static struct rt6_info *ip6_pol_route_output(struct fib6_table *table,
 	int strict = 0;
 	int attempts = 3;
 	int err;
-	int reachable = RT6_SELECT_F_REACHABLE;
+	int reachable = RT6_LOOKUP_F_REACHABLE;
 
-	if (flags & RT6_F_STRICT)
-		strict = RT6_SELECT_F_IFACE;
+	strict |= flags & RT6_LOOKUP_F_IFACE;
 
 relookup:
 	read_lock_bh(&table->tb6_lock);
@@ -797,7 +789,7 @@ struct dst_entry * ip6_route_output(struct sock *sk, struct flowi *fl)
 	int flags = 0;
 
 	if (rt6_need_strict(&fl->fl6_dst))
-		flags |= RT6_F_STRICT;
+		flags |= RT6_LOOKUP_F_IFACE;
 
 	return fib6_rule_lookup(fl, flags, ip6_pol_route_output);
 }
@@ -1362,7 +1354,7 @@ static struct rt6_info *ip6_route_redirect(struct in6_addr *dest,
 		},
 		.gateway = *gateway,
 	};
-	int flags = rt6_need_strict(dest) ? RT6_F_STRICT : 0;
+	int flags = rt6_need_strict(dest) ? RT6_LOOKUP_F_IFACE : 0;
 
 	return (struct rt6_info *)fib6_rule_lookup((struct flowi *)&rdfl, flags, __ip6_route_redirect);
 }
-- 
cgit v1.2.3


From 7e49e6de30efa716614e280d97963c570f3acf29 Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Fri, 22 Sep 2006 15:05:15 -0700
Subject: [XFRM]: Add XFRM_MODE_xxx for future use.

Transformation mode is used as either IPsec transport or tunnel.
It is required to add two more items, route optimization and inbound trigger
for Mobile IPv6.
Based on MIPL2 kernel patch.

This patch was also written by: Ville Nuorvala <vnuorval@tcs.hut.fi>

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/xfrm.h    |  6 ++++--
 include/net/xfrm.h      |  2 +-
 net/ipv4/ah4.c          |  2 +-
 net/ipv4/esp4.c         |  6 +++---
 net/ipv4/ipcomp.c       |  8 ++++----
 net/ipv4/xfrm4_input.c  |  2 +-
 net/ipv4/xfrm4_output.c |  4 ++--
 net/ipv4/xfrm4_policy.c |  2 +-
 net/ipv4/xfrm4_state.c  |  2 +-
 net/ipv4/xfrm4_tunnel.c |  2 +-
 net/ipv6/ah6.c          |  2 +-
 net/ipv6/esp6.c         |  4 ++--
 net/ipv6/ipcomp6.c      |  6 +++---
 net/ipv6/xfrm6_input.c  |  2 +-
 net/ipv6/xfrm6_output.c |  4 ++--
 net/ipv6/xfrm6_policy.c |  2 +-
 net/ipv6/xfrm6_state.c  |  2 +-
 net/ipv6/xfrm6_tunnel.c |  2 +-
 net/key/af_key.c        |  6 +++---
 net/xfrm/xfrm_policy.c  | 11 ++++++-----
 net/xfrm/xfrm_user.c    |  4 ++--
 21 files changed, 42 insertions(+), 39 deletions(-)

(limited to 'include')

diff --git a/include/linux/xfrm.h b/include/linux/xfrm.h
index 46a15c7a1a13..5154064b6d95 100644
--- a/include/linux/xfrm.h
+++ b/include/linux/xfrm.h
@@ -120,7 +120,9 @@ enum
 
 #define XFRM_MODE_TRANSPORT 0
 #define XFRM_MODE_TUNNEL 1
-#define XFRM_MODE_MAX 2
+#define XFRM_MODE_ROUTEOPTIMIZATION 2
+#define XFRM_MODE_IN_TRIGGER 3
+#define XFRM_MODE_MAX 4
 
 /* Netlink configuration messages.  */
 enum {
@@ -247,7 +249,7 @@ struct xfrm_usersa_info {
 	__u32				seq;
 	__u32				reqid;
 	__u16				family;
-	__u8				mode; /* 0=transport,1=tunnel */
+	__u8				mode;		/* XFRM_MODE_xxx */
 	__u8				replay_window;
 	__u8				flags;
 #define XFRM_STATE_NOECN	1
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 00bf86e6e82b..762795624b10 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -298,7 +298,7 @@ struct xfrm_tmpl
 
 	__u32			reqid;
 
-/* Mode: transport/tunnel */
+/* Mode: transport, tunnel etc. */
 	__u8			mode;
 
 /* Sharing mode: unique, this session only, this user only etc. */
diff --git a/net/ipv4/ah4.c b/net/ipv4/ah4.c
index 008e69d2e423..99542977e47e 100644
--- a/net/ipv4/ah4.c
+++ b/net/ipv4/ah4.c
@@ -265,7 +265,7 @@ static int ah_init_state(struct xfrm_state *x)
 		goto error;
 	
 	x->props.header_len = XFRM_ALIGN8(sizeof(struct ip_auth_hdr) + ahp->icv_trunc_len);
-	if (x->props.mode)
+	if (x->props.mode == XFRM_MODE_TUNNEL)
 		x->props.header_len += sizeof(struct iphdr);
 	x->data = ahp;
 
diff --git a/net/ipv4/esp4.c b/net/ipv4/esp4.c
index b428489f6ccd..e87377e1d6b6 100644
--- a/net/ipv4/esp4.c
+++ b/net/ipv4/esp4.c
@@ -248,7 +248,7 @@ static int esp_input(struct xfrm_state *x, struct sk_buff *skb)
 		 *    as per draft-ietf-ipsec-udp-encaps-06,
 		 *    section 3.1.2
 		 */
-		if (!x->props.mode)
+		if (x->props.mode == XFRM_MODE_TRANSPORT)
 			skb->ip_summed = CHECKSUM_UNNECESSARY;
 	}
 
@@ -267,7 +267,7 @@ static u32 esp4_get_max_size(struct xfrm_state *x, int mtu)
 	struct esp_data *esp = x->data;
 	u32 blksize = ALIGN(crypto_blkcipher_blocksize(esp->conf.tfm), 4);
 
-	if (x->props.mode) {
+	if (x->props.mode == XFRM_MODE_TUNNEL) {
 		mtu = ALIGN(mtu + 2, blksize);
 	} else {
 		/* The worst case. */
@@ -383,7 +383,7 @@ static int esp_init_state(struct xfrm_state *x)
 	if (crypto_blkcipher_setkey(tfm, esp->conf.key, esp->conf.key_len))
 		goto error;
 	x->props.header_len = sizeof(struct ip_esp_hdr) + esp->conf.ivlen;
-	if (x->props.mode)
+	if (x->props.mode == XFRM_MODE_TUNNEL)
 		x->props.header_len += sizeof(struct iphdr);
 	if (x->encap) {
 		struct xfrm_encap_tmpl *encap = x->encap;
diff --git a/net/ipv4/ipcomp.c b/net/ipv4/ipcomp.c
index 5bb9c9f03fb6..17342430a843 100644
--- a/net/ipv4/ipcomp.c
+++ b/net/ipv4/ipcomp.c
@@ -176,7 +176,7 @@ static int ipcomp_output(struct xfrm_state *x, struct sk_buff *skb)
 	return 0;
 
 out_ok:
-	if (x->props.mode)
+	if (x->props.mode == XFRM_MODE_TUNNEL)
 		ip_send_check(iph);
 	return 0;
 }
@@ -216,7 +216,7 @@ static struct xfrm_state *ipcomp_tunnel_create(struct xfrm_state *x)
 	t->id.daddr.a4 = x->id.daddr.a4;
 	memcpy(&t->sel, &x->sel, sizeof(t->sel));
 	t->props.family = AF_INET;
-	t->props.mode = 1;
+	t->props.mode = XFRM_MODE_TUNNEL;
 	t->props.saddr.a4 = x->props.saddr.a4;
 	t->props.flags = x->props.flags;
 
@@ -416,7 +416,7 @@ static int ipcomp_init_state(struct xfrm_state *x)
 		goto out;
 
 	x->props.header_len = 0;
-	if (x->props.mode)
+	if (x->props.mode == XFRM_MODE_TUNNEL)
 		x->props.header_len += sizeof(struct iphdr);
 
 	mutex_lock(&ipcomp_resource_mutex);
@@ -428,7 +428,7 @@ static int ipcomp_init_state(struct xfrm_state *x)
 		goto error;
 	mutex_unlock(&ipcomp_resource_mutex);
 
-	if (x->props.mode) {
+	if (x->props.mode == XFRM_MODE_TUNNEL) {
 		err = ipcomp_tunnel_attach(x);
 		if (err)
 			goto error_tunnel;
diff --git a/net/ipv4/xfrm4_input.c b/net/ipv4/xfrm4_input.c
index 817ed84511a6..040e8475f295 100644
--- a/net/ipv4/xfrm4_input.c
+++ b/net/ipv4/xfrm4_input.c
@@ -106,7 +106,7 @@ int xfrm4_rcv_encap(struct sk_buff *skb, __u16 encap_type)
 		if (x->mode->input(x, skb))
 			goto drop;
 
-		if (x->props.mode) {
+		if (x->props.mode == XFRM_MODE_TUNNEL) {
 			decaps = 1;
 			break;
 		}
diff --git a/net/ipv4/xfrm4_output.c b/net/ipv4/xfrm4_output.c
index 4a96a9e3ef3b..5fd115f0c547 100644
--- a/net/ipv4/xfrm4_output.c
+++ b/net/ipv4/xfrm4_output.c
@@ -54,7 +54,7 @@ static int xfrm4_output_one(struct sk_buff *skb)
 			goto error_nolock;
 	}
 
-	if (x->props.mode) {
+	if (x->props.mode == XFRM_MODE_TUNNEL) {
 		err = xfrm4_tunnel_check_size(skb);
 		if (err)
 			goto error_nolock;
@@ -85,7 +85,7 @@ static int xfrm4_output_one(struct sk_buff *skb)
 		}
 		dst = skb->dst;
 		x = dst->xfrm;
-	} while (x && !x->props.mode);
+	} while (x && (x->props.mode != XFRM_MODE_TUNNEL));
 
 	IPCB(skb)->flags |= IPSKB_XFRM_TRANSFORMED;
 	err = 0;
diff --git a/net/ipv4/xfrm4_policy.c b/net/ipv4/xfrm4_policy.c
index 8f50eae47d03..a5bed741de2c 100644
--- a/net/ipv4/xfrm4_policy.c
+++ b/net/ipv4/xfrm4_policy.c
@@ -96,7 +96,7 @@ __xfrm4_bundle_create(struct xfrm_policy *policy, struct xfrm_state **xfrm, int
 
 		dst1->next = dst_prev;
 		dst_prev = dst1;
-		if (xfrm[i]->props.mode) {
+		if (xfrm[i]->props.mode != XFRM_MODE_TRANSPORT) {
 			remote = xfrm[i]->id.daddr.a4;
 			local  = xfrm[i]->props.saddr.a4;
 			tunnel = 1;
diff --git a/net/ipv4/xfrm4_state.c b/net/ipv4/xfrm4_state.c
index 81e1751c966e..97b0c7589711 100644
--- a/net/ipv4/xfrm4_state.c
+++ b/net/ipv4/xfrm4_state.c
@@ -42,7 +42,7 @@ __xfrm4_init_tempsel(struct xfrm_state *x, struct flowi *fl,
 	x->props.saddr = tmpl->saddr;
 	if (x->props.saddr.a4 == 0)
 		x->props.saddr.a4 = saddr->a4;
-	if (tmpl->mode && x->props.saddr.a4 == 0) {
+	if (tmpl->mode == XFRM_MODE_TUNNEL && x->props.saddr.a4 == 0) {
 		struct rtable *rt;
 	        struct flowi fl_tunnel = {
         	        .nl_u = {
diff --git a/net/ipv4/xfrm4_tunnel.c b/net/ipv4/xfrm4_tunnel.c
index f8ceaa127c83..f110af5b1319 100644
--- a/net/ipv4/xfrm4_tunnel.c
+++ b/net/ipv4/xfrm4_tunnel.c
@@ -28,7 +28,7 @@ static int ipip_xfrm_rcv(struct xfrm_state *x, struct sk_buff *skb)
 
 static int ipip_init_state(struct xfrm_state *x)
 {
-	if (!x->props.mode)
+	if (x->props.mode != XFRM_MODE_TUNNEL)
 		return -EINVAL;
 
 	if (x->encap)
diff --git a/net/ipv6/ah6.c b/net/ipv6/ah6.c
index 00ffa7bc6c9f..60954fc7eb36 100644
--- a/net/ipv6/ah6.c
+++ b/net/ipv6/ah6.c
@@ -398,7 +398,7 @@ static int ah6_init_state(struct xfrm_state *x)
 		goto error;
 	
 	x->props.header_len = XFRM_ALIGN8(sizeof(struct ipv6_auth_hdr) + ahp->icv_trunc_len);
-	if (x->props.mode)
+	if (x->props.mode == XFRM_MODE_TUNNEL)
 		x->props.header_len += sizeof(struct ipv6hdr);
 	x->data = ahp;
 
diff --git a/net/ipv6/esp6.c b/net/ipv6/esp6.c
index 2ebfd281e721..2b8e52e1d0ab 100644
--- a/net/ipv6/esp6.c
+++ b/net/ipv6/esp6.c
@@ -237,7 +237,7 @@ static u32 esp6_get_max_size(struct xfrm_state *x, int mtu)
 	struct esp_data *esp = x->data;
 	u32 blksize = ALIGN(crypto_blkcipher_blocksize(esp->conf.tfm), 4);
 
-	if (x->props.mode) {
+	if (x->props.mode == XFRM_MODE_TUNNEL) {
 		mtu = ALIGN(mtu + 2, blksize);
 	} else {
 		/* The worst case. */
@@ -358,7 +358,7 @@ static int esp6_init_state(struct xfrm_state *x)
 	if (crypto_blkcipher_setkey(tfm, esp->conf.key, esp->conf.key_len))
 		goto error;
 	x->props.header_len = sizeof(struct ipv6_esp_hdr) + esp->conf.ivlen;
-	if (x->props.mode)
+	if (x->props.mode == XFRM_MODE_TUNNEL)
 		x->props.header_len += sizeof(struct ipv6hdr);
 	x->data = esp;
 	return 0;
diff --git a/net/ipv6/ipcomp6.c b/net/ipv6/ipcomp6.c
index a81e9e9d93bd..19eba8d9f851 100644
--- a/net/ipv6/ipcomp6.c
+++ b/net/ipv6/ipcomp6.c
@@ -212,7 +212,7 @@ static struct xfrm_state *ipcomp6_tunnel_create(struct xfrm_state *x)
 	memcpy(t->id.daddr.a6, x->id.daddr.a6, sizeof(struct in6_addr));
 	memcpy(&t->sel, &x->sel, sizeof(t->sel));
 	t->props.family = AF_INET6;
-	t->props.mode = 1;
+	t->props.mode = XFRM_MODE_TUNNEL;
 	memcpy(t->props.saddr.a6, x->props.saddr.a6, sizeof(struct in6_addr));
 
 	if (xfrm_init_state(t))
@@ -417,7 +417,7 @@ static int ipcomp6_init_state(struct xfrm_state *x)
 		goto out;
 
 	x->props.header_len = 0;
-	if (x->props.mode)
+	if (x->props.mode == XFRM_MODE_TUNNEL)
 		x->props.header_len += sizeof(struct ipv6hdr);
 	
 	mutex_lock(&ipcomp6_resource_mutex);
@@ -429,7 +429,7 @@ static int ipcomp6_init_state(struct xfrm_state *x)
 		goto error;
 	mutex_unlock(&ipcomp6_resource_mutex);
 
-	if (x->props.mode) {
+	if (x->props.mode == XFRM_MODE_TUNNEL) {
 		err = ipcomp6_tunnel_attach(x);
 		if (err)
 			goto error_tunnel;
diff --git a/net/ipv6/xfrm6_input.c b/net/ipv6/xfrm6_input.c
index 0405d74ff910..ee2f6b3908b6 100644
--- a/net/ipv6/xfrm6_input.c
+++ b/net/ipv6/xfrm6_input.c
@@ -72,7 +72,7 @@ int xfrm6_rcv_spi(struct sk_buff *skb, u32 spi)
 		if (x->mode->input(x, skb))
 			goto drop;
 
-		if (x->props.mode) { /* XXX */
+		if (x->props.mode == XFRM_MODE_TUNNEL) { /* XXX */
 			decaps = 1;
 			break;
 		}
diff --git a/net/ipv6/xfrm6_output.c b/net/ipv6/xfrm6_output.c
index 6d111743e508..26f18869f77b 100644
--- a/net/ipv6/xfrm6_output.c
+++ b/net/ipv6/xfrm6_output.c
@@ -47,7 +47,7 @@ static int xfrm6_output_one(struct sk_buff *skb)
 			goto error_nolock;
 	}
 
-	if (x->props.mode) {
+	if (x->props.mode == XFRM_MODE_TUNNEL) {
 		err = xfrm6_tunnel_check_size(skb);
 		if (err)
 			goto error_nolock;
@@ -80,7 +80,7 @@ static int xfrm6_output_one(struct sk_buff *skb)
 		}
 		dst = skb->dst;
 		x = dst->xfrm;
-	} while (x && !x->props.mode);
+	} while (x && (x->props.mode != XFRM_MODE_TUNNEL));
 
 	IP6CB(skb)->flags |= IP6SKB_XFRM_TRANSFORMED;
 	err = 0;
diff --git a/net/ipv6/xfrm6_policy.c b/net/ipv6/xfrm6_policy.c
index 73cd250aecbb..81355bb50328 100644
--- a/net/ipv6/xfrm6_policy.c
+++ b/net/ipv6/xfrm6_policy.c
@@ -114,7 +114,7 @@ __xfrm6_bundle_create(struct xfrm_policy *policy, struct xfrm_state **xfrm, int
 
 		dst1->next = dst_prev;
 		dst_prev = dst1;
-		if (xfrm[i]->props.mode) {
+		if (xfrm[i]->props.mode != XFRM_MODE_TRANSPORT) {
 			remote = (struct in6_addr*)&xfrm[i]->id.daddr;
 			local  = (struct in6_addr*)&xfrm[i]->props.saddr;
 			tunnel = 1;
diff --git a/net/ipv6/xfrm6_state.c b/net/ipv6/xfrm6_state.c
index b33296b3f6de..a1a1f5476442 100644
--- a/net/ipv6/xfrm6_state.c
+++ b/net/ipv6/xfrm6_state.c
@@ -42,7 +42,7 @@ __xfrm6_init_tempsel(struct xfrm_state *x, struct flowi *fl,
 	memcpy(&x->props.saddr, &tmpl->saddr, sizeof(x->props.saddr));
 	if (ipv6_addr_any((struct in6_addr*)&x->props.saddr))
 		memcpy(&x->props.saddr, saddr, sizeof(x->props.saddr));
-	if (tmpl->mode && ipv6_addr_any((struct in6_addr*)&x->props.saddr)) {
+	if (tmpl->mode == XFRM_MODE_TUNNEL && ipv6_addr_any((struct in6_addr*)&x->props.saddr)) {
 		struct rt6_info *rt;
 		struct flowi fl_tunnel = {
 			.nl_u = {
diff --git a/net/ipv6/xfrm6_tunnel.c b/net/ipv6/xfrm6_tunnel.c
index c8f9369c2a87..59685ee8f700 100644
--- a/net/ipv6/xfrm6_tunnel.c
+++ b/net/ipv6/xfrm6_tunnel.c
@@ -307,7 +307,7 @@ static int xfrm6_tunnel_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
 
 static int xfrm6_tunnel_init_state(struct xfrm_state *x)
 {
-	if (!x->props.mode)
+	if (x->props.mode != XFRM_MODE_TUNNEL)
 		return -EINVAL;
 
 	if (x->encap)
diff --git a/net/key/af_key.c b/net/key/af_key.c
index 797c744a8438..19e047b0e678 100644
--- a/net/key/af_key.c
+++ b/net/key/af_key.c
@@ -1765,7 +1765,7 @@ parse_ipsecrequest(struct xfrm_policy *xp, struct sadb_x_ipsecrequest *rq)
 	}
 
 	/* addresses present only in tunnel mode */
-	if (t->mode) {
+	if (t->mode == XFRM_MODE_TUNNEL) {
 		switch (xp->family) {
 		case AF_INET:
 			sin = (void*)(rq+1);
@@ -1997,7 +1997,7 @@ static void pfkey_xfrm_policy2msg(struct sk_buff *skb, struct xfrm_policy *xp, i
 		int req_size;
 
 		req_size = sizeof(struct sadb_x_ipsecrequest);
-		if (t->mode)
+		if (t->mode == XFRM_MODE_TUNNEL)
 			req_size += 2*socklen;
 		else
 			size -= 2*socklen;
@@ -2013,7 +2013,7 @@ static void pfkey_xfrm_policy2msg(struct sk_buff *skb, struct xfrm_policy *xp, i
 		if (t->optional)
 			rq->sadb_x_ipsecrequest_level = IPSEC_LEVEL_USE;
 		rq->sadb_x_ipsecrequest_reqid = t->reqid;
-		if (t->mode) {
+		if (t->mode == XFRM_MODE_TUNNEL) {
 			switch (xp->family) {
 			case AF_INET:
 				sin = (void*)(rq+1);
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 32c963c90573..a0d58971391d 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -779,7 +779,7 @@ xfrm_tmpl_resolve(struct xfrm_policy *policy, struct flowi *fl,
 		xfrm_address_t *local  = saddr;
 		struct xfrm_tmpl *tmpl = &policy->xfrm_vec[i];
 
-		if (tmpl->mode) {
+		if (tmpl->mode == XFRM_MODE_TUNNEL) {
 			remote = &tmpl->id.daddr;
 			local = &tmpl->saddr;
 		}
@@ -1005,7 +1005,8 @@ xfrm_state_ok(struct xfrm_tmpl *tmpl, struct xfrm_state *x,
 		(x->props.reqid == tmpl->reqid || !tmpl->reqid) &&
 		x->props.mode == tmpl->mode &&
 		(tmpl->aalgos & (1<<x->props.aalgo)) &&
-		!(x->props.mode && xfrm_state_addr_cmp(tmpl, x, family));
+		!(x->props.mode != XFRM_MODE_TRANSPORT &&
+		  xfrm_state_addr_cmp(tmpl, x, family));
 }
 
 static inline int
@@ -1015,14 +1016,14 @@ xfrm_policy_ok(struct xfrm_tmpl *tmpl, struct sec_path *sp, int start,
 	int idx = start;
 
 	if (tmpl->optional) {
-		if (!tmpl->mode)
+		if (tmpl->mode == XFRM_MODE_TRANSPORT)
 			return start;
 	} else
 		start = -1;
 	for (; idx < sp->len; idx++) {
 		if (xfrm_state_ok(tmpl, sp->xvec[idx], family))
 			return ++idx;
-		if (sp->xvec[idx]->props.mode)
+		if (sp->xvec[idx]->props.mode != XFRM_MODE_TRANSPORT)
 			break;
 	}
 	return start;
@@ -1047,7 +1048,7 @@ EXPORT_SYMBOL(xfrm_decode_session);
 static inline int secpath_has_tunnel(struct sec_path *sp, int k)
 {
 	for (; k < sp->len; k++) {
-		if (sp->xvec[k]->props.mode)
+		if (sp->xvec[k]->props.mode != XFRM_MODE_TRANSPORT)
 			return 1;
 	}
 
diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c
index f70e158874d2..0d580ac19771 100644
--- a/net/xfrm/xfrm_user.c
+++ b/net/xfrm/xfrm_user.c
@@ -174,8 +174,8 @@ static int verify_newsa_info(struct xfrm_usersa_info *p,
 
 	err = -EINVAL;
 	switch (p->mode) {
-	case 0:
-	case 1:
+	case XFRM_MODE_TRANSPORT:
+	case XFRM_MODE_TUNNEL:
 		break;
 
 	default:
-- 
cgit v1.2.3


From 5794708f11551b6d19b10673abf4b0202f66b44d Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Fri, 22 Sep 2006 15:06:24 -0700
Subject: [XFRM]: Introduce a helper to compare id protocol.

Put the helper to header for future use.
Based on MIPL2 kernel patch.

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h    | 6 ++++++
 net/xfrm/xfrm_state.c | 6 +++---
 2 files changed, 9 insertions(+), 3 deletions(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 762795624b10..5b364b0a6a28 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -9,6 +9,7 @@
 #include <linux/skbuff.h>
 #include <linux/socket.h>
 #include <linux/pfkeyv2.h>
+#include <linux/ipsec.h>
 #include <linux/in6.h>
 #include <linux/mutex.h>
 
@@ -835,6 +836,11 @@ static inline int xfrm_state_kern(struct xfrm_state *x)
 	return atomic_read(&x->tunnel_users);
 }
 
+static inline int xfrm_id_proto_match(u8 proto, u8 userproto)
+{
+	return (userproto == IPSEC_PROTO_ANY || proto == userproto);
+}
+
 /*
  * xfrm algorithm information
  */
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 1c796087ee78..34c038cbdf46 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -294,7 +294,7 @@ void xfrm_state_flush(u8 proto)
 restart:
 		list_for_each_entry(x, xfrm_state_bydst+i, bydst) {
 			if (!xfrm_state_kern(x) &&
-			    (proto == IPSEC_PROTO_ANY || x->id.proto == proto)) {
+			    xfrm_id_proto_match(x->id.proto, proto)) {
 				xfrm_state_hold(x);
 				spin_unlock_bh(&xfrm_state_lock);
 
@@ -772,7 +772,7 @@ int xfrm_state_walk(u8 proto, int (*func)(struct xfrm_state *, int, void*),
 	spin_lock_bh(&xfrm_state_lock);
 	for (i = 0; i < XFRM_DST_HSIZE; i++) {
 		list_for_each_entry(x, xfrm_state_bydst+i, bydst) {
-			if (proto == IPSEC_PROTO_ANY || x->id.proto == proto)
+			if (xfrm_id_proto_match(x->id.proto, proto))
 				count++;
 		}
 	}
@@ -783,7 +783,7 @@ int xfrm_state_walk(u8 proto, int (*func)(struct xfrm_state *, int, void*),
 
 	for (i = 0; i < XFRM_DST_HSIZE; i++) {
 		list_for_each_entry(x, xfrm_state_bydst+i, bydst) {
-			if (proto != IPSEC_PROTO_ANY && x->id.proto != proto)
+			if (!xfrm_id_proto_match(x->id.proto, proto))
 				continue;
 			err = func(x, --count, data);
 			if (err)
-- 
cgit v1.2.3


From dc00a525603650a1471c823a1e48c6505c2f9765 Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 17:49:52 -0700
Subject: [XFRM] STATE: Allow non IPsec protocol.

It will be added two more transformation protocols (routing header
and destination options header) for Mobile IPv6.
xfrm_id_proto_match() can be handle zero as all, IPSEC_PROTO_ANY as
all IPsec and otherwise as exact one.
Based on MIPL2 kernel patch.

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h   | 5 ++++-
 net/xfrm/xfrm_user.c | 2 +-
 2 files changed, 5 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 5b364b0a6a28..2a7d2132a1ae 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -838,7 +838,10 @@ static inline int xfrm_state_kern(struct xfrm_state *x)
 
 static inline int xfrm_id_proto_match(u8 proto, u8 userproto)
 {
-	return (userproto == IPSEC_PROTO_ANY || proto == userproto);
+	return (!userproto || proto == userproto ||
+		(userproto == IPSEC_PROTO_ANY && (proto == IPPROTO_AH ||
+						  proto == IPPROTO_ESP ||
+						  proto == IPPROTO_COMP)));
 }
 
 /*
diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c
index 0d580ac19771..41f3d51ffc33 100644
--- a/net/xfrm/xfrm_user.c
+++ b/net/xfrm/xfrm_user.c
@@ -542,7 +542,7 @@ static int xfrm_dump_sa(struct sk_buff *skb, struct netlink_callback *cb)
 	info.nlmsg_flags = NLM_F_MULTI;
 	info.this_idx = 0;
 	info.start_idx = cb->args[0];
-	(void) xfrm_state_walk(IPSEC_PROTO_ANY, dump_one_state, &info);
+	(void) xfrm_state_walk(0, dump_one_state, &info);
 	cb->args[0] = info.this_idx;
 
 	return skb->len;
-- 
cgit v1.2.3


From 622dc8281a80374873686514e46f852093d91106 Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 17:52:01 -0700
Subject: [XFRM]: Expand XFRM_MAX_DEPTH for route optimization.

XFRM_MAX_DEPTH is a limit of transformation states to be applied to the same
flow. Two more extension headers are used by Mobile IPv6 transformation.
Based on MIPL2 kernel patch.

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 2a7d2132a1ae..aa3be68041be 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -314,7 +314,7 @@ struct xfrm_tmpl
 	__u32			calgos;
 };
 
-#define XFRM_MAX_DEPTH		4
+#define XFRM_MAX_DEPTH		6
 
 struct xfrm_policy
 {
-- 
cgit v1.2.3


From 6c44e6b7ab500d7e3e3f406c83325671be51a752 Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 17:53:57 -0700
Subject: [XFRM] STATE: Add source address list.

Support source address based searching.
Mobile IPv6 will use it.
Based on MIPL2 kernel patch.

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h     | 26 ++++++++++++++++++++++++++
 net/ipv4/xfrm4_state.c |  3 +++
 net/ipv6/xfrm6_state.c |  3 +++
 net/xfrm/xfrm_state.c  | 21 +++++++++++++++++++--
 4 files changed, 51 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index aa3be68041be..88145e3348d0 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -95,6 +95,7 @@ struct xfrm_state
 {
 	/* Note: bydst is re-used during gc */
 	struct list_head	bydst;
+	struct list_head	bysrc;
 	struct list_head	byspi;
 
 	atomic_t		refcnt;
@@ -236,6 +237,7 @@ extern int __xfrm_state_delete(struct xfrm_state *x);
 struct xfrm_state_afinfo {
 	unsigned short		family;
 	struct list_head	*state_bydst;
+	struct list_head	*state_bysrc;
 	struct list_head	*state_byspi;
 	int			(*init_flags)(struct xfrm_state *x);
 	void			(*init_tempsel)(struct xfrm_state *x, struct flowi *fl,
@@ -420,6 +422,30 @@ unsigned xfrm_dst_hash(xfrm_address_t *addr, unsigned short family)
 	return 0;
 }
 
+static __inline__
+unsigned __xfrm4_src_hash(xfrm_address_t *addr)
+{
+	return __xfrm4_dst_hash(addr);
+}
+
+static __inline__
+unsigned __xfrm6_src_hash(xfrm_address_t *addr)
+{
+	return __xfrm6_dst_hash(addr);
+}
+
+static __inline__
+unsigned xfrm_src_hash(xfrm_address_t *addr, unsigned short family)
+{
+	switch (family) {
+	case AF_INET:
+		return __xfrm4_src_hash(addr);
+	case AF_INET6:
+		return __xfrm6_src_hash(addr);
+	}
+	return 0;
+}
+
 static __inline__
 unsigned __xfrm4_spi_hash(xfrm_address_t *addr, u32 spi, u8 proto)
 {
diff --git a/net/ipv4/xfrm4_state.c b/net/ipv4/xfrm4_state.c
index 97b0c7589711..c56b258fad73 100644
--- a/net/ipv4/xfrm4_state.c
+++ b/net/ipv4/xfrm4_state.c
@@ -122,6 +122,9 @@ __xfrm4_find_acq(u8 mode, u32 reqid, u8 proto,
 		add_timer(&x0->timer);
 		xfrm_state_hold(x0);
 		list_add_tail(&x0->bydst, xfrm4_state_afinfo.state_bydst+h);
+		h = __xfrm4_src_hash(saddr);
+		xfrm_state_hold(x0);
+		list_add_tail(&x0->bysrc, xfrm4_state_afinfo.state_bysrc+h);
 		wake_up(&km_waitq);
 	}
 	if (x0)
diff --git a/net/ipv6/xfrm6_state.c b/net/ipv6/xfrm6_state.c
index a1a1f5476442..2fb07850449f 100644
--- a/net/ipv6/xfrm6_state.c
+++ b/net/ipv6/xfrm6_state.c
@@ -126,6 +126,9 @@ __xfrm6_find_acq(u8 mode, u32 reqid, u8 proto,
 		add_timer(&x0->timer);
 		xfrm_state_hold(x0);
 		list_add_tail(&x0->bydst, xfrm6_state_afinfo.state_bydst+h);
+		h = __xfrm6_src_hash(saddr);
+		xfrm_state_hold(x0);
+		list_add_tail(&x0->bysrc, xfrm6_state_afinfo.state_bysrc+h);
 		wake_up(&km_waitq);
 	}
 	if (x0)
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 34c038cbdf46..2a9992894e69 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -45,6 +45,7 @@ static DEFINE_SPINLOCK(xfrm_state_lock);
  * Also, it can be used by ah/esp icmp error handler to find offending SA.
  */
 static struct list_head xfrm_state_bydst[XFRM_DST_HSIZE];
+static struct list_head xfrm_state_bysrc[XFRM_DST_HSIZE];
 static struct list_head xfrm_state_byspi[XFRM_DST_HSIZE];
 
 DECLARE_WAIT_QUEUE_HEAD(km_waitq);
@@ -200,6 +201,7 @@ struct xfrm_state *xfrm_state_alloc(void)
 		atomic_set(&x->refcnt, 1);
 		atomic_set(&x->tunnel_users, 0);
 		INIT_LIST_HEAD(&x->bydst);
+		INIT_LIST_HEAD(&x->bysrc);
 		INIT_LIST_HEAD(&x->byspi);
 		init_timer(&x->timer);
 		x->timer.function = xfrm_timer_handler;
@@ -240,6 +242,8 @@ int __xfrm_state_delete(struct xfrm_state *x)
 		spin_lock(&xfrm_state_lock);
 		list_del(&x->bydst);
 		__xfrm_state_put(x);
+		list_del(&x->bysrc);
+		__xfrm_state_put(x);
 		if (x->id.spi) {
 			list_del(&x->byspi);
 			__xfrm_state_put(x);
@@ -415,6 +419,8 @@ xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr,
 			x->km.state = XFRM_STATE_ACQ;
 			list_add_tail(&x->bydst, xfrm_state_bydst+h);
 			xfrm_state_hold(x);
+			list_add_tail(&x->bysrc, xfrm_state_bysrc+h);
+			xfrm_state_hold(x);
 			if (x->id.spi) {
 				h = xfrm_spi_hash(&x->id.daddr, x->id.spi, x->id.proto, family);
 				list_add(&x->byspi, xfrm_state_byspi+h);
@@ -448,11 +454,19 @@ static void __xfrm_state_insert(struct xfrm_state *x)
 	list_add(&x->bydst, xfrm_state_bydst+h);
 	xfrm_state_hold(x);
 
-	h = xfrm_spi_hash(&x->id.daddr, x->id.spi, x->id.proto, x->props.family);
+	h = xfrm_src_hash(&x->props.saddr, x->props.family);
 
-	list_add(&x->byspi, xfrm_state_byspi+h);
+	list_add(&x->bysrc, xfrm_state_bysrc+h);
 	xfrm_state_hold(x);
 
+	if (xfrm_id_proto_match(x->id.proto, IPSEC_PROTO_ANY)) {
+		h = xfrm_spi_hash(&x->id.daddr, x->id.spi, x->id.proto,
+				  x->props.family);
+
+		list_add(&x->byspi, xfrm_state_byspi+h);
+		xfrm_state_hold(x);
+	}
+
 	if (!mod_timer(&x->timer, jiffies + HZ))
 		xfrm_state_hold(x);
 
@@ -1075,6 +1089,7 @@ int xfrm_state_register_afinfo(struct xfrm_state_afinfo *afinfo)
 		err = -ENOBUFS;
 	else {
 		afinfo->state_bydst = xfrm_state_bydst;
+		afinfo->state_bysrc = xfrm_state_bysrc;
 		afinfo->state_byspi = xfrm_state_byspi;
 		xfrm_state_afinfo[afinfo->family] = afinfo;
 	}
@@ -1097,6 +1112,7 @@ int xfrm_state_unregister_afinfo(struct xfrm_state_afinfo *afinfo)
 		else {
 			xfrm_state_afinfo[afinfo->family] = NULL;
 			afinfo->state_byspi = NULL;
+			afinfo->state_bysrc = NULL;
 			afinfo->state_bydst = NULL;
 		}
 	}
@@ -1218,6 +1234,7 @@ void __init xfrm_state_init(void)
 
 	for (i=0; i<XFRM_DST_HSIZE; i++) {
 		INIT_LIST_HEAD(&xfrm_state_bydst[i]);
+		INIT_LIST_HEAD(&xfrm_state_bysrc[i]);
 		INIT_LIST_HEAD(&xfrm_state_byspi[i]);
 	}
 	INIT_WORK(&xfrm_state_gc_work, xfrm_state_gc_task, NULL);
-- 
cgit v1.2.3


From eb2971b68a7d17a7d0fa2c7fc6fbc4bfe41cd694 Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 17:56:04 -0700
Subject: [XFRM] STATE: Search by address using source address list.

This is a support to search transformation states by its addresses
by using source address list for Mobile IPv6 usage.
To use it from user-space, it is also added a message type for
source address as a xfrm state option.
Based on MIPL2 kernel patch.

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/xfrm.h   |  1 +
 include/net/xfrm.h     |  2 ++
 net/ipv4/xfrm4_state.c |  9 ++++++++
 net/ipv6/xfrm6_state.c | 21 ++++++++++++++++++
 net/xfrm/xfrm_state.c  | 37 +++++++++++++++++++++++++++----
 net/xfrm/xfrm_user.c   | 59 +++++++++++++++++++++++++++++++++++++++++++++-----
 6 files changed, 119 insertions(+), 10 deletions(-)

(limited to 'include')

diff --git a/include/linux/xfrm.h b/include/linux/xfrm.h
index 5154064b6d95..66343d3d4b91 100644
--- a/include/linux/xfrm.h
+++ b/include/linux/xfrm.h
@@ -234,6 +234,7 @@ enum xfrm_attr_type_t {
 	XFRMA_REPLAY_VAL,
 	XFRMA_REPLAY_THRESH,
 	XFRMA_ETIMER_THRESH,
+	XFRMA_SRCADDR,		/* xfrm_address_t */
 	__XFRMA_MAX
 
 #define XFRMA_MAX (__XFRMA_MAX - 1)
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 88145e3348d0..d9c40e713184 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -244,6 +244,7 @@ struct xfrm_state_afinfo {
 						struct xfrm_tmpl *tmpl,
 						xfrm_address_t *daddr, xfrm_address_t *saddr);
 	struct xfrm_state	*(*state_lookup)(xfrm_address_t *daddr, u32 spi, u8 proto);
+	struct xfrm_state	*(*state_lookup_byaddr)(xfrm_address_t *daddr, xfrm_address_t *saddr, u8 proto);
 	struct xfrm_state	*(*find_acq)(u8 mode, u32 reqid, u8 proto, 
 					     xfrm_address_t *daddr, xfrm_address_t *saddr, 
 					     int create);
@@ -937,6 +938,7 @@ extern void xfrm_state_insert(struct xfrm_state *x);
 extern int xfrm_state_add(struct xfrm_state *x);
 extern int xfrm_state_update(struct xfrm_state *x);
 extern struct xfrm_state *xfrm_state_lookup(xfrm_address_t *daddr, u32 spi, u8 proto, unsigned short family);
+extern struct xfrm_state *xfrm_state_lookup_byaddr(xfrm_address_t *daddr, xfrm_address_t *saddr, u8 proto, unsigned short family);
 extern struct xfrm_state *xfrm_find_acq_byseq(u32 seq);
 extern int xfrm_state_delete(struct xfrm_state *x);
 extern void xfrm_state_flush(u8 proto);
diff --git a/net/ipv4/xfrm4_state.c b/net/ipv4/xfrm4_state.c
index c56b258fad73..616be131b4e3 100644
--- a/net/ipv4/xfrm4_state.c
+++ b/net/ipv4/xfrm4_state.c
@@ -80,6 +80,14 @@ __xfrm4_state_lookup(xfrm_address_t *daddr, u32 spi, u8 proto)
 	return NULL;
 }
 
+/* placeholder until ipv4's code is written */
+static struct xfrm_state *
+__xfrm4_state_lookup_byaddr(xfrm_address_t *daddr, xfrm_address_t *saddr,
+			    u8 proto)
+{
+	return NULL;
+}
+
 static struct xfrm_state *
 __xfrm4_find_acq(u8 mode, u32 reqid, u8 proto, 
 		 xfrm_address_t *daddr, xfrm_address_t *saddr, 
@@ -137,6 +145,7 @@ static struct xfrm_state_afinfo xfrm4_state_afinfo = {
 	.init_flags		= xfrm4_init_flags,
 	.init_tempsel		= __xfrm4_init_tempsel,
 	.state_lookup		= __xfrm4_state_lookup,
+	.state_lookup_byaddr	= __xfrm4_state_lookup_byaddr,
 	.find_acq		= __xfrm4_find_acq,
 };
 
diff --git a/net/ipv6/xfrm6_state.c b/net/ipv6/xfrm6_state.c
index 2fb07850449f..9c95b9d3e110 100644
--- a/net/ipv6/xfrm6_state.c
+++ b/net/ipv6/xfrm6_state.c
@@ -63,6 +63,26 @@ __xfrm6_init_tempsel(struct xfrm_state *x, struct flowi *fl,
 	x->props.family = AF_INET6;
 }
 
+static struct xfrm_state *
+__xfrm6_state_lookup_byaddr(xfrm_address_t *daddr, xfrm_address_t *saddr,
+			    u8 proto)
+{
+	struct xfrm_state *x = NULL;
+	unsigned h;
+
+	h = __xfrm6_src_hash(saddr);
+	list_for_each_entry(x, xfrm6_state_afinfo.state_bysrc+h, bysrc) {
+		if (x->props.family == AF_INET6 &&
+		    ipv6_addr_equal((struct in6_addr *)daddr, (struct in6_addr *)x->id.daddr.a6) &&
+		    ipv6_addr_equal((struct in6_addr *)saddr, (struct in6_addr *)x->props.saddr.a6) &&
+		    proto == x->id.proto) {
+			xfrm_state_hold(x);
+			return x;
+		}
+	}
+	return NULL;
+}
+
 static struct xfrm_state *
 __xfrm6_state_lookup(xfrm_address_t *daddr, u32 spi, u8 proto)
 {
@@ -140,6 +160,7 @@ static struct xfrm_state_afinfo xfrm6_state_afinfo = {
 	.family			= AF_INET6,
 	.init_tempsel		= __xfrm6_init_tempsel,
 	.state_lookup		= __xfrm6_state_lookup,
+	.state_lookup_byaddr	= __xfrm6_state_lookup_byaddr,
 	.find_acq		= __xfrm6_find_acq,
 };
 
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 2a9992894e69..11f480b12952 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -487,6 +487,16 @@ void xfrm_state_insert(struct xfrm_state *x)
 }
 EXPORT_SYMBOL(xfrm_state_insert);
 
+static inline struct xfrm_state *
+__xfrm_state_locate(struct xfrm_state_afinfo *afinfo, struct xfrm_state *x,
+		    int use_spi)
+{
+	if (use_spi)
+		return afinfo->state_lookup(&x->id.daddr, x->id.spi, x->id.proto);
+	else
+		return afinfo->state_lookup_byaddr(&x->id.daddr, &x->props.saddr, x->id.proto);
+}
+
 static struct xfrm_state *__xfrm_find_acq_byseq(u32 seq);
 
 int xfrm_state_add(struct xfrm_state *x)
@@ -495,6 +505,7 @@ int xfrm_state_add(struct xfrm_state *x)
 	struct xfrm_state *x1;
 	int family;
 	int err;
+	int use_spi = xfrm_id_proto_match(x->id.proto, IPSEC_PROTO_ANY);
 
 	family = x->props.family;
 	afinfo = xfrm_state_get_afinfo(family);
@@ -503,7 +514,7 @@ int xfrm_state_add(struct xfrm_state *x)
 
 	spin_lock_bh(&xfrm_state_lock);
 
-	x1 = afinfo->state_lookup(&x->id.daddr, x->id.spi, x->id.proto);
+	x1 = __xfrm_state_locate(afinfo, x, use_spi);
 	if (x1) {
 		xfrm_state_put(x1);
 		x1 = NULL;
@@ -511,7 +522,7 @@ int xfrm_state_add(struct xfrm_state *x)
 		goto out;
 	}
 
-	if (x->km.seq) {
+	if (use_spi && x->km.seq) {
 		x1 = __xfrm_find_acq_byseq(x->km.seq);
 		if (x1 && xfrm_addr_cmp(&x1->id.daddr, &x->id.daddr, family)) {
 			xfrm_state_put(x1);
@@ -519,7 +530,7 @@ int xfrm_state_add(struct xfrm_state *x)
 		}
 	}
 
-	if (!x1)
+	if (use_spi && !x1)
 		x1 = afinfo->find_acq(
 			x->props.mode, x->props.reqid, x->id.proto,
 			&x->id.daddr, &x->props.saddr, 0);
@@ -548,13 +559,14 @@ int xfrm_state_update(struct xfrm_state *x)
 	struct xfrm_state_afinfo *afinfo;
 	struct xfrm_state *x1;
 	int err;
+	int use_spi = xfrm_id_proto_match(x->id.proto, IPSEC_PROTO_ANY);
 
 	afinfo = xfrm_state_get_afinfo(x->props.family);
 	if (unlikely(afinfo == NULL))
 		return -EAFNOSUPPORT;
 
 	spin_lock_bh(&xfrm_state_lock);
-	x1 = afinfo->state_lookup(&x->id.daddr, x->id.spi, x->id.proto);
+	x1 = __xfrm_state_locate(afinfo, x, use_spi);
 
 	err = -ESRCH;
 	if (!x1)
@@ -674,6 +686,23 @@ xfrm_state_lookup(xfrm_address_t *daddr, u32 spi, u8 proto,
 }
 EXPORT_SYMBOL(xfrm_state_lookup);
 
+struct xfrm_state *
+xfrm_state_lookup_byaddr(xfrm_address_t *daddr, xfrm_address_t *saddr,
+			 u8 proto, unsigned short family)
+{
+	struct xfrm_state *x;
+	struct xfrm_state_afinfo *afinfo = xfrm_state_get_afinfo(family);
+	if (!afinfo)
+		return NULL;
+
+	spin_lock_bh(&xfrm_state_lock);
+	x = afinfo->state_lookup_byaddr(daddr, saddr, proto);
+	spin_unlock_bh(&xfrm_state_lock);
+	xfrm_state_put_afinfo(afinfo);
+	return x;
+}
+EXPORT_SYMBOL(xfrm_state_lookup_byaddr);
+
 struct xfrm_state *
 xfrm_find_acq(u8 mode, u32 reqid, u8 proto, 
 	      xfrm_address_t *daddr, xfrm_address_t *saddr, 
diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c
index 41f3d51ffc33..b5f8ab71aa54 100644
--- a/net/xfrm/xfrm_user.c
+++ b/net/xfrm/xfrm_user.c
@@ -87,6 +87,22 @@ static int verify_encap_tmpl(struct rtattr **xfrma)
 	return 0;
 }
 
+static int verify_one_addr(struct rtattr **xfrma, enum xfrm_attr_type_t type,
+			   xfrm_address_t **addrp)
+{
+	struct rtattr *rt = xfrma[type - 1];
+
+	if (!rt)
+		return 0;
+
+	if ((rt->rta_len - sizeof(*rt)) < sizeof(**addrp))
+		return -EINVAL;
+
+	if (addrp)
+		*addrp = RTA_DATA(rt);
+
+	return 0;
+}
 
 static inline int verify_sec_ctx_len(struct rtattr **xfrma)
 {
@@ -418,16 +434,48 @@ out:
 	return err;
 }
 
+static struct xfrm_state *xfrm_user_state_lookup(struct xfrm_usersa_id *p,
+						 struct rtattr **xfrma,
+						 int *errp)
+{
+	struct xfrm_state *x = NULL;
+	int err;
+
+	if (xfrm_id_proto_match(p->proto, IPSEC_PROTO_ANY)) {
+		err = -ESRCH;
+		x = xfrm_state_lookup(&p->daddr, p->spi, p->proto, p->family);
+	} else {
+		xfrm_address_t *saddr = NULL;
+
+		err = verify_one_addr(xfrma, XFRMA_SRCADDR, &saddr);
+		if (err)
+			goto out;
+
+		if (!saddr) {
+			err = -EINVAL;
+			goto out;
+		}
+
+		x = xfrm_state_lookup_byaddr(&p->daddr, saddr, p->proto,
+					     p->family);
+	}
+
+ out:
+	if (!x && errp)
+		*errp = err;
+	return x;
+}
+
 static int xfrm_del_sa(struct sk_buff *skb, struct nlmsghdr *nlh, void **xfrma)
 {
 	struct xfrm_state *x;
-	int err;
+	int err = -ESRCH;
 	struct km_event c;
 	struct xfrm_usersa_id *p = NLMSG_DATA(nlh);
 
-	x = xfrm_state_lookup(&p->daddr, p->spi, p->proto, p->family);
+	x = xfrm_user_state_lookup(p, (struct rtattr **)xfrma, &err);
 	if (x == NULL)
-		return -ESRCH;
+		return err;
 
 	if ((err = security_xfrm_state_delete(x)) != 0)
 		goto out;
@@ -578,10 +626,9 @@ static int xfrm_get_sa(struct sk_buff *skb, struct nlmsghdr *nlh, void **xfrma)
 	struct xfrm_usersa_id *p = NLMSG_DATA(nlh);
 	struct xfrm_state *x;
 	struct sk_buff *resp_skb;
-	int err;
+	int err = -ESRCH;
 
-	x = xfrm_state_lookup(&p->daddr, p->spi, p->proto, p->family);
-	err = -ESRCH;
+	x = xfrm_user_state_lookup(p, (struct rtattr **)xfrma, &err);
 	if (x == NULL)
 		goto out_noput;
 
-- 
cgit v1.2.3


From aee5adb4307c4c63a4dc5f3b49984d76f8a71b5b Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 17:57:28 -0700
Subject: [XFRM] STATE: Add a hook to find offset to be inserted header in
 outbound.

On current kernel, ip6_find_1stfragopt() is used by IPv6 IPsec to find
offset to be inserted header in outbound for transport mode. (BTW, no
usage may be needed for IPv4 case.)  Mobile IPv6 requires another
logic for routing header and destination options header
respectively. This patch is common platform for the offset and adopts
it to IPsec.

Based on MIPL2 kernel patch.

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h              | 3 +++
 net/ipv6/ah6.c                  | 3 ++-
 net/ipv6/esp6.c                 | 3 ++-
 net/ipv6/ipcomp6.c              | 1 +
 net/ipv6/ipv6_syms.c            | 1 +
 net/ipv6/xfrm6_mode_transport.c | 2 +-
 net/ipv6/xfrm6_output.c         | 6 ++++++
 7 files changed, 16 insertions(+), 3 deletions(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index d9c40e713184..eed48f832ce1 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -265,6 +265,7 @@ struct xfrm_type
 	void			(*destructor)(struct xfrm_state *);
 	int			(*input)(struct xfrm_state *, struct sk_buff *skb);
 	int			(*output)(struct xfrm_state *, struct sk_buff *pskb);
+	int			(*hdr_offset)(struct xfrm_state *, struct sk_buff *, u8 **);
 	/* Estimate maximal size of result of transformation of a dgram */
 	u32			(*get_max_size)(struct xfrm_state *, int size);
 };
@@ -960,6 +961,8 @@ extern u32 xfrm6_tunnel_alloc_spi(xfrm_address_t *saddr);
 extern void xfrm6_tunnel_free_spi(xfrm_address_t *saddr);
 extern u32 xfrm6_tunnel_spi_lookup(xfrm_address_t *saddr);
 extern int xfrm6_output(struct sk_buff *skb);
+extern int xfrm6_find_1stfragopt(struct xfrm_state *x, struct sk_buff *skb,
+				 u8 **prevhdr);
 
 #ifdef CONFIG_XFRM
 extern int xfrm4_rcv_encap(struct sk_buff *skb, __u16 encap_type);
diff --git a/net/ipv6/ah6.c b/net/ipv6/ah6.c
index 60954fc7eb36..6c0aa51319a5 100644
--- a/net/ipv6/ah6.c
+++ b/net/ipv6/ah6.c
@@ -435,7 +435,8 @@ static struct xfrm_type ah6_type =
 	.init_state	= ah6_init_state,
 	.destructor	= ah6_destroy,
 	.input		= ah6_input,
-	.output		= ah6_output
+	.output		= ah6_output,
+	.hdr_offset	= xfrm6_find_1stfragopt,
 };
 
 static struct inet6_protocol ah6_protocol = {
diff --git a/net/ipv6/esp6.c b/net/ipv6/esp6.c
index 2b8e52e1d0ab..ae50b9511151 100644
--- a/net/ipv6/esp6.c
+++ b/net/ipv6/esp6.c
@@ -379,7 +379,8 @@ static struct xfrm_type esp6_type =
 	.destructor	= esp6_destroy,
 	.get_max_size	= esp6_get_max_size,
 	.input		= esp6_input,
-	.output		= esp6_output
+	.output		= esp6_output,
+	.hdr_offset	= xfrm6_find_1stfragopt,
 };
 
 static struct inet6_protocol esp6_protocol = {
diff --git a/net/ipv6/ipcomp6.c b/net/ipv6/ipcomp6.c
index 19eba8d9f851..ad9c6e824e62 100644
--- a/net/ipv6/ipcomp6.c
+++ b/net/ipv6/ipcomp6.c
@@ -461,6 +461,7 @@ static struct xfrm_type ipcomp6_type =
 	.destructor	= ipcomp6_destroy,
 	.input		= ipcomp6_input,
 	.output		= ipcomp6_output,
+	.hdr_offset	= xfrm6_find_1stfragopt,
 };
 
 static struct inet6_protocol ipcomp6_protocol = 
diff --git a/net/ipv6/ipv6_syms.c b/net/ipv6/ipv6_syms.c
index dd4d1ce77769..e1a741612888 100644
--- a/net/ipv6/ipv6_syms.c
+++ b/net/ipv6/ipv6_syms.c
@@ -31,6 +31,7 @@ EXPORT_SYMBOL(ipv6_chk_addr);
 EXPORT_SYMBOL(in6_dev_finish_destroy);
 #ifdef CONFIG_XFRM
 EXPORT_SYMBOL(xfrm6_rcv);
+EXPORT_SYMBOL(xfrm6_find_1stfragopt);
 #endif
 EXPORT_SYMBOL(rt6_lookup);
 EXPORT_SYMBOL(ipv6_push_nfrag_opts);
diff --git a/net/ipv6/xfrm6_mode_transport.c b/net/ipv6/xfrm6_mode_transport.c
index 711d713e36d8..a5dce216024d 100644
--- a/net/ipv6/xfrm6_mode_transport.c
+++ b/net/ipv6/xfrm6_mode_transport.c
@@ -35,7 +35,7 @@ static int xfrm6_transport_output(struct sk_buff *skb)
 	skb_push(skb, x->props.header_len);
 	iph = skb->nh.ipv6h;
 
-	hdr_len = ip6_find_1stfragopt(skb, &prevhdr);
+	hdr_len = x->type->hdr_offset(x, skb, &prevhdr);
 	skb->nh.raw = prevhdr - x->props.header_len;
 	skb->h.raw = skb->data + hdr_len;
 	memmove(skb->data, iph, hdr_len);
diff --git a/net/ipv6/xfrm6_output.c b/net/ipv6/xfrm6_output.c
index 26f18869f77b..b4628fbf8ff5 100644
--- a/net/ipv6/xfrm6_output.c
+++ b/net/ipv6/xfrm6_output.c
@@ -17,6 +17,12 @@
 #include <net/ipv6.h>
 #include <net/xfrm.h>
 
+int xfrm6_find_1stfragopt(struct xfrm_state *x, struct sk_buff *skb,
+			  u8 **prevhdr)
+{
+	return ip6_find_1stfragopt(skb, prevhdr);
+}
+
 static int xfrm6_tunnel_check_size(struct sk_buff *skb)
 {
 	int mtu, ret = 0;
-- 
cgit v1.2.3


From fbd9a5b47ee9c319ff0cae584391241ce78ffd6b Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 18:08:21 -0700
Subject: [XFRM] STATE: Common receive function for route optimization
 extension headers.

XFRM_STATE_WILDRECV flag is introduced; the last resort state is set
it and receives packet which is not route optimized but uses such
extension headers i.e. Mobile IPv6 signaling (binding update and
acknowledgement).  A node enabled Mobile IPv6 adds the state.

Based on MIPL2 kernel patch.

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/xfrm.h   |   1 +
 include/net/xfrm.h     |   2 +
 net/ipv6/ipv6_syms.c   |   1 +
 net/ipv6/xfrm6_input.c | 108 +++++++++++++++++++++++++++++++++++++++++++++++++
 net/xfrm/xfrm_state.c  |   1 +
 5 files changed, 113 insertions(+)

(limited to 'include')

diff --git a/include/linux/xfrm.h b/include/linux/xfrm.h
index 66343d3d4b91..a7c9e4cfb15b 100644
--- a/include/linux/xfrm.h
+++ b/include/linux/xfrm.h
@@ -256,6 +256,7 @@ struct xfrm_usersa_info {
 #define XFRM_STATE_NOECN	1
 #define XFRM_STATE_DECAP_DSCP	2
 #define XFRM_STATE_NOPMTUDISC	4
+#define XFRM_STATE_WILDRECV	8
 };
 
 struct xfrm_usersa_id {
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index eed48f832ce1..0d735a5aba61 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -955,6 +955,8 @@ extern int xfrm4_tunnel_register(struct xfrm_tunnel *handler);
 extern int xfrm4_tunnel_deregister(struct xfrm_tunnel *handler);
 extern int xfrm6_rcv_spi(struct sk_buff *skb, u32 spi);
 extern int xfrm6_rcv(struct sk_buff **pskb);
+extern int xfrm6_input_addr(struct sk_buff *skb, xfrm_address_t *daddr,
+			    xfrm_address_t *saddr, u8 proto);
 extern int xfrm6_tunnel_register(struct xfrm6_tunnel *handler);
 extern int xfrm6_tunnel_deregister(struct xfrm6_tunnel *handler);
 extern u32 xfrm6_tunnel_alloc_spi(xfrm_address_t *saddr);
diff --git a/net/ipv6/ipv6_syms.c b/net/ipv6/ipv6_syms.c
index e1a741612888..7b7b90d9c3d0 100644
--- a/net/ipv6/ipv6_syms.c
+++ b/net/ipv6/ipv6_syms.c
@@ -31,6 +31,7 @@ EXPORT_SYMBOL(ipv6_chk_addr);
 EXPORT_SYMBOL(in6_dev_finish_destroy);
 #ifdef CONFIG_XFRM
 EXPORT_SYMBOL(xfrm6_rcv);
+EXPORT_SYMBOL(xfrm6_input_addr);
 EXPORT_SYMBOL(xfrm6_find_1stfragopt);
 #endif
 EXPORT_SYMBOL(rt6_lookup);
diff --git a/net/ipv6/xfrm6_input.c b/net/ipv6/xfrm6_input.c
index ee2f6b3908b6..a40a05789013 100644
--- a/net/ipv6/xfrm6_input.c
+++ b/net/ipv6/xfrm6_input.c
@@ -138,3 +138,111 @@ int xfrm6_rcv(struct sk_buff **pskb)
 {
 	return xfrm6_rcv_spi(*pskb, 0);
 }
+
+int xfrm6_input_addr(struct sk_buff *skb, xfrm_address_t *daddr,
+		     xfrm_address_t *saddr, u8 proto)
+{
+ 	struct xfrm_state *x = NULL;
+ 	int wildcard = 0;
+	struct in6_addr any;
+	xfrm_address_t *xany;
+	struct xfrm_state *xfrm_vec_one = NULL;
+ 	int nh = 0;
+	int i = 0;
+
+	ipv6_addr_set(&any, 0, 0, 0, 0);
+	xany = (xfrm_address_t *)&any;
+
+	for (i = 0; i < 3; i++) {
+		xfrm_address_t *dst, *src;
+		switch (i) {
+		case 0:
+			dst = daddr;
+			src = saddr;
+			break;
+		case 1:
+			/* lookup state with wild-card source address */
+			wildcard = 1;
+			dst = daddr;
+			src = xany;
+			break;
+		case 2:
+		default:
+ 			/* lookup state with wild-card addresses */
+			wildcard = 1; /* XXX */
+			dst = xany;
+			src = xany;
+			break;
+ 		}
+
+		x = xfrm_state_lookup_byaddr(dst, src, proto, AF_INET6);
+		if (!x)
+			continue;
+
+		spin_lock(&x->lock);
+
+		if (wildcard) {
+			if ((x->props.flags & XFRM_STATE_WILDRECV) == 0) {
+				spin_unlock(&x->lock);
+				xfrm_state_put(x);
+				x = NULL;
+				continue;
+			}
+		}
+
+		if (unlikely(x->km.state != XFRM_STATE_VALID)) {
+			spin_unlock(&x->lock);
+			xfrm_state_put(x);
+ 			x = NULL;
+ 			continue;
+		}
+		if (xfrm_state_check_expire(x)) {
+			spin_unlock(&x->lock);
+			xfrm_state_put(x);
+			x = NULL;
+			continue;
+		}
+
+		nh = x->type->input(x, skb);
+		if (nh <= 0) {
+			spin_unlock(&x->lock);
+			xfrm_state_put(x);
+			x = NULL;
+			continue;
+		}
+
+		x->curlft.bytes += skb->len;
+		x->curlft.packets++;
+
+		spin_unlock(&x->lock);
+
+		xfrm_vec_one = x;
+		break;
+	}
+
+	if (!xfrm_vec_one)
+		goto drop;
+
+	/* Allocate new secpath or COW existing one. */
+	if (!skb->sp || atomic_read(&skb->sp->refcnt) != 1) {
+		struct sec_path *sp;
+		sp = secpath_dup(skb->sp);
+		if (!sp)
+			goto drop;
+		if (skb->sp)
+			secpath_put(skb->sp);
+		skb->sp = sp;
+	}
+
+	if (1 + skb->sp->len > XFRM_MAX_DEPTH)
+		goto drop;
+
+	skb->sp->xvec[skb->sp->len] = xfrm_vec_one;
+	skb->sp->len ++;
+
+	return 1;
+drop:
+	if (xfrm_vec_one)
+		xfrm_state_put(xfrm_vec_one);
+	return -1;
+}
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 11f480b12952..f05371556cce 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -352,6 +352,7 @@ xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr,
 	list_for_each_entry(x, xfrm_state_bydst+h, bydst) {
 		if (x->props.family == family &&
 		    x->props.reqid == tmpl->reqid &&
+		    !(x->props.flags & XFRM_STATE_WILDRECV) &&
 		    xfrm_state_addr_check(x, daddr, saddr, family) &&
 		    tmpl->mode == x->props.mode &&
 		    tmpl->id.proto == x->id.proto &&
-- 
cgit v1.2.3


From 99505a843673faeae962a8cde128c7c034ba6b5e Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 18:10:33 -0700
Subject: [XFRM] STATE: Add a hook to obtain local/remote outbound address.

Outbound transformation replaces both source and destination address with
state's end-point addresses at the same time when IPsec tunnel mode.
It is also required to change them for Mobile IPv6 route optimization, but we
should care about the following differences:
 - changing result is not end-point but care-of address
 - either source or destination is replaced for each state
This hook is a common platform to change outbound address.
Based on MIPL2 kernel patch.

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h      |  2 ++
 net/ipv6/xfrm6_policy.c | 20 ++++++++++++++++++--
 2 files changed, 20 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 0d735a5aba61..aa3ac994477b 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -266,6 +266,8 @@ struct xfrm_type
 	int			(*input)(struct xfrm_state *, struct sk_buff *skb);
 	int			(*output)(struct xfrm_state *, struct sk_buff *pskb);
 	int			(*hdr_offset)(struct xfrm_state *, struct sk_buff *, u8 **);
+	xfrm_address_t		*(*local_addr)(struct xfrm_state *, xfrm_address_t *);
+	xfrm_address_t		*(*remote_addr)(struct xfrm_state *, xfrm_address_t *);
 	/* Estimate maximal size of result of transformation of a dgram */
 	u32			(*get_max_size)(struct xfrm_state *, int size);
 };
diff --git a/net/ipv6/xfrm6_policy.c b/net/ipv6/xfrm6_policy.c
index 81355bb50328..9328fc88708a 100644
--- a/net/ipv6/xfrm6_policy.c
+++ b/net/ipv6/xfrm6_policy.c
@@ -59,6 +59,22 @@ __xfrm6_find_bundle(struct flowi *fl, struct xfrm_policy *policy)
 	return dst;
 }
 
+static inline struct in6_addr*
+__xfrm6_bundle_addr_remote(struct xfrm_state *x, struct in6_addr *addr)
+{
+	return (x->type->remote_addr) ?
+		(struct in6_addr*)x->type->remote_addr(x, (xfrm_address_t *)addr) :
+		(struct in6_addr*)&x->id.daddr;
+}
+
+static inline struct in6_addr*
+__xfrm6_bundle_addr_local(struct xfrm_state *x, struct in6_addr *addr)
+{
+	return (x->type->local_addr) ?
+		(struct in6_addr*)x->type->local_addr(x, (xfrm_address_t *)addr) :
+		(struct in6_addr*)&x->props.saddr;
+}
+
 /* Allocate chain of dst_entry's, attach known xfrm's, calculate
  * all the metrics... Shortly, bundle a bundle.
  */
@@ -115,8 +131,8 @@ __xfrm6_bundle_create(struct xfrm_policy *policy, struct xfrm_state **xfrm, int
 		dst1->next = dst_prev;
 		dst_prev = dst1;
 		if (xfrm[i]->props.mode != XFRM_MODE_TRANSPORT) {
-			remote = (struct in6_addr*)&xfrm[i]->id.daddr;
-			local  = (struct in6_addr*)&xfrm[i]->props.saddr;
+			remote = __xfrm6_bundle_addr_remote(xfrm[i], remote);
+			local  = __xfrm6_bundle_addr_local(xfrm[i], local);
 			tunnel = 1;
 		}
 		header_len += xfrm[i]->props.header_len;
-- 
cgit v1.2.3


From 1b5c229987dc4d0c92a38fac0cde2aeec08cd775 Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 18:11:50 -0700
Subject: [XFRM] STATE: Support non-fragment outbound transformation headers.

For originated outbound IPv6 packets which will fragment, ip6_append_data()
should know length of extension headers before sending them and
the length is carried by dst_entry.
IPv6 IPsec headers fragment then transformation was
designed to place all headers after fragment header.
OTOH Mobile IPv6 extension headers do not fragment then
it is a good idea to make dst_entry have non-fragment length to tell it
to ip6_append_data().

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/dst.h       |  1 +
 include/net/xfrm.h      |  2 ++
 net/ipv4/xfrm4_policy.c |  1 +
 net/ipv6/ip6_output.c   |  2 +-
 net/ipv6/xfrm6_policy.c | 24 ++++++++++++++++++++++--
 5 files changed, 27 insertions(+), 3 deletions(-)

(limited to 'include')

diff --git a/include/net/dst.h b/include/net/dst.h
index 36d54fc248b0..a8d825f90305 100644
--- a/include/net/dst.h
+++ b/include/net/dst.h
@@ -54,6 +54,7 @@ struct dst_entry
 	unsigned long		expires;
 
 	unsigned short		header_len;	/* more space at head required */
+	unsigned short		nfheader_len;	/* more non-fragment space at head required */
 	unsigned short		trailer_len;	/* space to reserve at tail */
 
 	u32			metrics[RTAX_MAX];
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index aa3ac994477b..aa93cc1f6299 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -260,6 +260,8 @@ struct xfrm_type
 	char			*description;
 	struct module		*owner;
 	__u8			proto;
+	__u8			flags;
+#define XFRM_TYPE_NON_FRAGMENT	1
 
 	int			(*init_state)(struct xfrm_state *x);
 	void			(*destructor)(struct xfrm_state *);
diff --git a/net/ipv4/xfrm4_policy.c b/net/ipv4/xfrm4_policy.c
index a5bed741de2c..e517981ceadd 100644
--- a/net/ipv4/xfrm4_policy.c
+++ b/net/ipv4/xfrm4_policy.c
@@ -135,6 +135,7 @@ __xfrm4_bundle_create(struct xfrm_policy *policy, struct xfrm_state **xfrm, int
 		dst_prev->flags	       |= DST_HOST;
 		dst_prev->lastuse	= jiffies;
 		dst_prev->header_len	= header_len;
+		dst_prev->nfheader_len	= 0;
 		dst_prev->trailer_len	= trailer_len;
 		memcpy(&dst_prev->metrics, &x->route->metrics, sizeof(dst_prev->metrics));
 
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index 2a376b7d91b4..258e3e45f5e0 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -971,7 +971,7 @@ int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to,
 
 	hh_len = LL_RESERVED_SPACE(rt->u.dst.dev);
 
-	fragheaderlen = sizeof(struct ipv6hdr) + (opt ? opt->opt_nflen : 0);
+	fragheaderlen = sizeof(struct ipv6hdr) + rt->u.dst.nfheader_len + (opt ? opt->opt_nflen : 0);
 	maxfraglen = ((mtu - fragheaderlen) & ~7) + fragheaderlen - sizeof(struct frag_hdr);
 
 	if (mtu <= sizeof(struct ipv6hdr) + IPV6_MAXPLEN) {
diff --git a/net/ipv6/xfrm6_policy.c b/net/ipv6/xfrm6_policy.c
index 9328fc88708a..a3f68c8b737e 100644
--- a/net/ipv6/xfrm6_policy.c
+++ b/net/ipv6/xfrm6_policy.c
@@ -75,6 +75,24 @@ __xfrm6_bundle_addr_local(struct xfrm_state *x, struct in6_addr *addr)
 		(struct in6_addr*)&x->props.saddr;
 }
 
+static inline void
+__xfrm6_bundle_len_inc(int *len, int *nflen, struct xfrm_state *x)
+{
+	if (x->type->flags & XFRM_TYPE_NON_FRAGMENT)
+		*nflen += x->props.header_len;
+	else
+		*len += x->props.header_len;
+}
+
+static inline void
+__xfrm6_bundle_len_dec(int *len, int *nflen, struct xfrm_state *x)
+{
+	if (x->type->flags & XFRM_TYPE_NON_FRAGMENT)
+		*nflen -= x->props.header_len;
+	else
+		*len -= x->props.header_len;
+}
+
 /* Allocate chain of dst_entry's, attach known xfrm's, calculate
  * all the metrics... Shortly, bundle a bundle.
  */
@@ -99,6 +117,7 @@ __xfrm6_bundle_create(struct xfrm_policy *policy, struct xfrm_state **xfrm, int
 	int i;
 	int err = 0;
 	int header_len = 0;
+	int nfheader_len = 0;
 	int trailer_len = 0;
 
 	dst = dst_prev = NULL;
@@ -135,7 +154,7 @@ __xfrm6_bundle_create(struct xfrm_policy *policy, struct xfrm_state **xfrm, int
 			local  = __xfrm6_bundle_addr_local(xfrm[i], local);
 			tunnel = 1;
 		}
-		header_len += xfrm[i]->props.header_len;
+		__xfrm6_bundle_len_inc(&header_len, &nfheader_len, xfrm[i]);
 		trailer_len += xfrm[i]->props.trailer_len;
 
 		if (tunnel) {
@@ -170,6 +189,7 @@ __xfrm6_bundle_create(struct xfrm_policy *policy, struct xfrm_state **xfrm, int
 		dst_prev->flags	       |= DST_HOST;
 		dst_prev->lastuse	= jiffies;
 		dst_prev->header_len	= header_len;
+		dst_prev->nfheader_len	= nfheader_len;
 		dst_prev->trailer_len	= trailer_len;
 		memcpy(&dst_prev->metrics, &x->route->metrics, sizeof(dst_prev->metrics));
 
@@ -188,7 +208,7 @@ __xfrm6_bundle_create(struct xfrm_policy *policy, struct xfrm_state **xfrm, int
 		x->u.rt6.rt6i_src      = rt0->rt6i_src;	
 		x->u.rt6.rt6i_idev     = rt0->rt6i_idev;
 		in6_dev_hold(rt0->rt6i_idev);
-		header_len -= x->u.dst.xfrm->props.header_len;
+		__xfrm6_bundle_len_dec(&header_len, &nfheader_len, x->u.dst.xfrm);
 		trailer_len -= x->u.dst.xfrm->props.trailer_len;
 	}
 
-- 
cgit v1.2.3


From 060f02a3bdd4d9ba8aa3c48e9b470672b1f3a585 Mon Sep 17 00:00:00 2001
From: Noriaki TAKAMIYA <takamiya@po.ntts.co.jp>
Date: Wed, 23 Aug 2006 18:18:55 -0700
Subject: [XFRM] STATE: Introduce care-of address.

Care-of address is carried by state as a transformation option like
IPsec encryption/authentication algorithm.

Based on MIPL2 kernel patch.

Signed-off-by: Noriaki TAKAMIYA <takamiya@po.ntts.co.jp>
Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/xfrm.h  |  1 +
 include/net/xfrm.h    |  3 +++
 net/xfrm/xfrm_state.c |  6 ++++++
 net/xfrm/xfrm_user.c  | 28 +++++++++++++++++++++++++++-
 4 files changed, 37 insertions(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/linux/xfrm.h b/include/linux/xfrm.h
index a7c9e4cfb15b..b53f799189af 100644
--- a/include/linux/xfrm.h
+++ b/include/linux/xfrm.h
@@ -235,6 +235,7 @@ enum xfrm_attr_type_t {
 	XFRMA_REPLAY_THRESH,
 	XFRMA_ETIMER_THRESH,
 	XFRMA_SRCADDR,		/* xfrm_address_t */
+	XFRMA_COADDR,		/* xfrm_address_t */
 	__XFRMA_MAX
 
 #define XFRMA_MAX (__XFRMA_MAX - 1)
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index aa93cc1f6299..872a2a4022b2 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -134,6 +134,9 @@ struct xfrm_state
 	/* Data for encapsulator */
 	struct xfrm_encap_tmpl	*encap;
 
+	/* Data for care-of address */
+	xfrm_address_t	*coaddr;
+
 	/* IPComp needs an IPIP tunnel for handling uncompressed packets */
 	struct xfrm_state	*tunnel;
 
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index f05371556cce..3da89c01ea71 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -78,6 +78,7 @@ static void xfrm_state_gc_destroy(struct xfrm_state *x)
 	kfree(x->ealg);
 	kfree(x->calg);
 	kfree(x->encap);
+	kfree(x->coaddr);
 	if (x->mode)
 		xfrm_put_mode(x->mode);
 	if (x->type) {
@@ -603,6 +604,11 @@ out:
 	if (likely(x1->km.state == XFRM_STATE_VALID)) {
 		if (x->encap && x1->encap)
 			memcpy(x1->encap, x->encap, sizeof(*x1->encap));
+		if (x->coaddr && x1->coaddr) {
+			memcpy(x1->coaddr, x->coaddr, sizeof(*x1->coaddr));
+		}
+		if (!use_spi && memcmp(&x1->sel, &x->sel, sizeof(x1->sel)))
+			memcpy(&x1->sel, &x->sel, sizeof(x1->sel));
 		memcpy(&x1->lft, &x->lft, sizeof(x1->lft));
 		x1->km.dying = 0;
 
diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c
index b5f8ab71aa54..939808de9e20 100644
--- a/net/xfrm/xfrm_user.c
+++ b/net/xfrm/xfrm_user.c
@@ -187,11 +187,14 @@ static int verify_newsa_info(struct xfrm_usersa_info *p,
 		goto out;
 	if ((err = verify_sec_ctx_len(xfrma)))
 		goto out;
+	if ((err = verify_one_addr(xfrma, XFRMA_COADDR, NULL)))
+		goto out;
 
 	err = -EINVAL;
 	switch (p->mode) {
 	case XFRM_MODE_TRANSPORT:
 	case XFRM_MODE_TUNNEL:
+	case XFRM_MODE_ROUTEOPTIMIZATION:
 		break;
 
 	default:
@@ -276,6 +279,24 @@ static int attach_sec_ctx(struct xfrm_state *x, struct rtattr *u_arg)
 	return security_xfrm_state_alloc(x, uctx);
 }
 
+static int attach_one_addr(xfrm_address_t **addrpp, struct rtattr *u_arg)
+{
+	struct rtattr *rta = u_arg;
+	xfrm_address_t *p, *uaddrp;
+
+	if (!rta)
+		return 0;
+
+	uaddrp = RTA_DATA(rta);
+	p = kmalloc(sizeof(*p), GFP_KERNEL);
+	if (!p)
+		return -ENOMEM;
+
+	memcpy(p, uaddrp, sizeof(*p));
+	*addrpp = p;
+	return 0;
+}
+
 static void copy_from_user_state(struct xfrm_state *x, struct xfrm_usersa_info *p)
 {
 	memcpy(&x->id, &p->id, sizeof(x->id));
@@ -365,7 +386,8 @@ static struct xfrm_state *xfrm_state_construct(struct xfrm_usersa_info *p,
 		goto error;
 	if ((err = attach_encap_tmpl(&x->encap, xfrma[XFRMA_ENCAP-1])))
 		goto error;
-
+	if ((err = attach_one_addr(&x->coaddr, xfrma[XFRMA_COADDR-1])))
+		goto error;
 	err = xfrm_init_state(x);
 	if (err)
 		goto error;
@@ -569,6 +591,10 @@ static int dump_one_state(struct xfrm_state *x, int count, void *ptr)
 		uctx->ctx_len = x->security->ctx_len;
 		memcpy(uctx + 1, x->security->ctx_str, x->security->ctx_len);
 	}
+
+	if (x->coaddr)
+		RTA_PUT(skb, XFRMA_COADDR, sizeof(*x->coaddr), x->coaddr);
+
 	nlh->nlmsg_len = skb->tail - b;
 out:
 	sp->this_idx++;
-- 
cgit v1.2.3


From 9afaca057980c02771f4657c455cc7592fcd7373 Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 18:20:16 -0700
Subject: [XFRM] IPV6: Update outbound state timestamp for each sending.

With this patch transformation state is updated last used time
for each sending. Xtime is used for it like other state lifetime
expiration.
Mobile IPv6 enabled nodes will want to know traffic status of each
binding (e.g. judgement to request binding refresh by correspondent node,
or to keep home/care-of nonce alive by mobile node).
The last used timestamp is an important hint about it.
Based on MIPL2 kernel patch.

This patch was also written by: Henrik Petander <petander@tcs.hut.fi>

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/xfrm.h    | 1 +
 include/net/xfrm.h      | 3 +++
 net/ipv6/xfrm6_output.c | 2 ++
 net/xfrm/xfrm_user.c    | 3 +++
 4 files changed, 9 insertions(+)

(limited to 'include')

diff --git a/include/linux/xfrm.h b/include/linux/xfrm.h
index b53f799189af..1d8c1f22c12d 100644
--- a/include/linux/xfrm.h
+++ b/include/linux/xfrm.h
@@ -236,6 +236,7 @@ enum xfrm_attr_type_t {
 	XFRMA_ETIMER_THRESH,
 	XFRMA_SRCADDR,		/* xfrm_address_t */
 	XFRMA_COADDR,		/* xfrm_address_t */
+	XFRMA_LASTUSED,
 	__XFRMA_MAX
 
 #define XFRMA_MAX (__XFRMA_MAX - 1)
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 872a2a4022b2..248874ecf8df 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -167,6 +167,9 @@ struct xfrm_state
 	struct xfrm_lifetime_cur curlft;
 	struct timer_list	timer;
 
+	/* Last used time */
+	u64			lastused;
+
 	/* Reference to data common to all the instances of this
 	 * transformer. */
 	struct xfrm_type	*type;
diff --git a/net/ipv6/xfrm6_output.c b/net/ipv6/xfrm6_output.c
index b4628fbf8ff5..db58104e710b 100644
--- a/net/ipv6/xfrm6_output.c
+++ b/net/ipv6/xfrm6_output.c
@@ -75,6 +75,8 @@ static int xfrm6_output_one(struct sk_buff *skb)
 
 		x->curlft.bytes += skb->len;
 		x->curlft.packets++;
+		if (x->props.mode == XFRM_MODE_ROUTEOPTIMIZATION)
+			x->lastused = (u64)xtime.tv_sec;
 
 		spin_unlock_bh(&x->lock);
 
diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c
index 939808de9e20..f643063a1cbd 100644
--- a/net/xfrm/xfrm_user.c
+++ b/net/xfrm/xfrm_user.c
@@ -595,6 +595,9 @@ static int dump_one_state(struct xfrm_state *x, int count, void *ptr)
 	if (x->coaddr)
 		RTA_PUT(skb, XFRMA_COADDR, sizeof(*x->coaddr), x->coaddr);
 
+	if (x->lastused)
+		RTA_PUT(skb, XFRMA_LASTUSED, sizeof(x->lastused), &x->lastused);
+
 	nlh->nlmsg_len = skb->tail - b;
 out:
 	sp->this_idx++;
-- 
cgit v1.2.3


From e53820de0f81da1429048634cadc6ef5f50c2f8b Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 19:12:01 -0700
Subject: [XFRM] IPV6: Restrict bundle reusing

For outbound transformation, bundle is checked whether it is
suitable for current flow to be reused or not. In such IPv6 case
as below, transformation may apply incorrect bundle for the flow instead
of creating another bundle:

- The policy selector has destination prefix length < 128
  (Two or more addresses can be matched it)
- Its bundle holds dst entry of default route whose prefix length < 128
  (Previous traffic was used such route as next hop)
- The policy and the bundle were used a transport mode state and
  this time flow address is not matched the bundled state.

This issue is found by Mobile IPv6 usage to protect mobility signaling
by IPsec, but it is not a Mobile IPv6 specific.
This patch adds strict check to xfrm_bundle_ok() for each
state mode and address when prefix length is less than 128.

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h      | 19 ++++++++++++++++++-
 net/ipv4/xfrm4_policy.c |  2 +-
 net/ipv6/xfrm6_policy.c |  4 +++-
 net/xfrm/xfrm_policy.c  |  8 ++++++--
 4 files changed, 28 insertions(+), 5 deletions(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 248874ecf8df..7f1630630dcf 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -869,6 +869,23 @@ xfrm_state_addr_check(struct xfrm_state *x,
 	return 0;
 }
 
+static __inline__ int
+xfrm_state_addr_flow_check(struct xfrm_state *x, struct flowi *fl,
+			   unsigned short family)
+{
+	switch (family) {
+	case AF_INET:
+		return __xfrm4_state_addr_check(x,
+						(xfrm_address_t *)&fl->fl4_dst,
+						(xfrm_address_t *)&fl->fl4_src);
+	case AF_INET6:
+		return __xfrm6_state_addr_check(x,
+						(xfrm_address_t *)&fl->fl6_dst,
+						(xfrm_address_t *)&fl->fl6_src);
+	}
+	return 0;
+}
+
 static inline int xfrm_state_kern(struct xfrm_state *x)
 {
 	return atomic_read(&x->tunnel_users);
@@ -1014,7 +1031,7 @@ extern void xfrm_policy_flush(void);
 extern int xfrm_sk_policy_insert(struct sock *sk, int dir, struct xfrm_policy *pol);
 extern int xfrm_flush_bundles(void);
 extern void xfrm_flush_all_bundles(void);
-extern int xfrm_bundle_ok(struct xfrm_dst *xdst, struct flowi *fl, int family);
+extern int xfrm_bundle_ok(struct xfrm_dst *xdst, struct flowi *fl, int family, int strict);
 extern void xfrm_init_pmtu(struct dst_entry *dst);
 
 extern wait_queue_head_t km_waitq;
diff --git a/net/ipv4/xfrm4_policy.c b/net/ipv4/xfrm4_policy.c
index e517981ceadd..42d8ded0f96a 100644
--- a/net/ipv4/xfrm4_policy.c
+++ b/net/ipv4/xfrm4_policy.c
@@ -33,7 +33,7 @@ __xfrm4_find_bundle(struct flowi *fl, struct xfrm_policy *policy)
 		    xdst->u.rt.fl.fl4_dst == fl->fl4_dst &&
 	    	    xdst->u.rt.fl.fl4_src == fl->fl4_src &&
 	    	    xdst->u.rt.fl.fl4_tos == fl->fl4_tos &&
-		    xfrm_bundle_ok(xdst, fl, AF_INET)) {
+		    xfrm_bundle_ok(xdst, fl, AF_INET, 0)) {
 			dst_clone(dst);
 			break;
 		}
diff --git a/net/ipv6/xfrm6_policy.c b/net/ipv6/xfrm6_policy.c
index a3f68c8b737e..729b4748d6d3 100644
--- a/net/ipv6/xfrm6_policy.c
+++ b/net/ipv6/xfrm6_policy.c
@@ -50,7 +50,9 @@ __xfrm6_find_bundle(struct flowi *fl, struct xfrm_policy *policy)
 				 xdst->u.rt6.rt6i_src.plen);
 		if (ipv6_addr_equal(&xdst->u.rt6.rt6i_dst.addr, &fl_dst_prefix) &&
 		    ipv6_addr_equal(&xdst->u.rt6.rt6i_src.addr, &fl_src_prefix) &&
-		    xfrm_bundle_ok(xdst, fl, AF_INET6)) {
+		    xfrm_bundle_ok(xdst, fl, AF_INET6,
+				   (xdst->u.rt6.rt6i_dst.plen != 128 ||
+				    xdst->u.rt6.rt6i_src.plen != 128))) {
 			dst_clone(dst);
 			break;
 		}
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 56abb5c057d4..ad2a5cba1f5b 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -1167,7 +1167,7 @@ static struct dst_entry *xfrm_dst_check(struct dst_entry *dst, u32 cookie)
 
 static int stale_bundle(struct dst_entry *dst)
 {
-	return !xfrm_bundle_ok((struct xfrm_dst *)dst, NULL, AF_UNSPEC);
+	return !xfrm_bundle_ok((struct xfrm_dst *)dst, NULL, AF_UNSPEC, 0);
 }
 
 void xfrm_dst_ifdown(struct dst_entry *dst, struct net_device *dev)
@@ -1282,7 +1282,7 @@ EXPORT_SYMBOL(xfrm_init_pmtu);
  * still valid.
  */
 
-int xfrm_bundle_ok(struct xfrm_dst *first, struct flowi *fl, int family)
+int xfrm_bundle_ok(struct xfrm_dst *first, struct flowi *fl, int family, int strict)
 {
 	struct dst_entry *dst = &first->u.dst;
 	struct xfrm_dst *last;
@@ -1304,6 +1304,10 @@ int xfrm_bundle_ok(struct xfrm_dst *first, struct flowi *fl, int family)
 		if (dst->xfrm->km.state != XFRM_STATE_VALID)
 			return 0;
 
+		if (strict && fl && dst->xfrm->props.mode != XFRM_MODE_TUNNEL &&
+		    !xfrm_state_addr_flow_check(dst->xfrm, fl, family))
+			return 0;
+
 		mtu = dst_mtu(dst->child);
 		if (xdst->child_mtu_cached != mtu) {
 			last = xdst;
-- 
cgit v1.2.3


From 642ec62eee5bdc158e01029220c8a23c685778fb Mon Sep 17 00:00:00 2001
From: Noriaki TAKAMIYA <takamiya@po.ntts.co.jp>
Date: Wed, 23 Aug 2006 19:15:07 -0700
Subject: [IPV6] MIP6: Add routing header type 2 definition.

Add routing header type 2 definition for Mobile IPv6.
Based on MIPL2 kernel patch.

Signed-off-by: Noriaki TAKAMIYA <takamiya@po.ntts.co.jp>
Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/ipv6.h | 13 +++++++++++++
 1 file changed, 13 insertions(+)

(limited to 'include')

diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index 02d14a3ff2af..d995662e94c4 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -29,6 +29,7 @@ struct in6_ifreq {
 
 #define IPV6_SRCRT_STRICT	0x01	/* this hop must be a neighbor	*/
 #define IPV6_SRCRT_TYPE_0	0	/* IPv6 type 0 Routing Header	*/
+#define IPV6_SRCRT_TYPE_2	2	/* IPv6 type 2 Routing Header	*/
 
 /*
  *	routing header
@@ -73,6 +74,18 @@ struct rt0_hdr {
 #define rt0_type		rt_hdr.type
 };
 
+/*
+ *	routing header type 2
+ */
+
+struct rt2_hdr {
+	struct ipv6_rt_hdr	rt_hdr;
+	__u32			reserved;
+	struct in6_addr		addr;
+
+#define rt2_type		rt_hdr.type
+};
+
 struct ipv6_auth_hdr {
 	__u8  nexthdr;
 	__u8  hdrlen;           /* This one is measured in 32 bit units! */
-- 
cgit v1.2.3


From 65d4ed92219b28875efb52de5700da8c3dfa83e1 Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 19:16:22 -0700
Subject: [IPV6] MIP6: Add inbound interface of routing header type 2.

Add inbound interface of routing header type 2 for Mobile IPv6.
Based on MIPL2 kernel patch.

This patch was also written by: Ville Nuorvala <vnuorval@tcs.hut.fi>

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/addrconf.h |  7 +++++
 net/ipv6/exthdrs.c     | 69 ++++++++++++++++++++++++++++++++++++++++++++------
 2 files changed, 68 insertions(+), 8 deletions(-)

(limited to 'include')

diff --git a/include/net/addrconf.h b/include/net/addrconf.h
index 3d71251b3eca..5fc8627435eb 100644
--- a/include/net/addrconf.h
+++ b/include/net/addrconf.h
@@ -61,6 +61,13 @@ extern int			addrconf_set_dstaddr(void __user *arg);
 extern int			ipv6_chk_addr(struct in6_addr *addr,
 					      struct net_device *dev,
 					      int strict);
+/* XXX: this is a placeholder till addrconf supports */
+#ifdef CONFIG_IPV6_MIP6
+static inline int ipv6_chk_home_addr(struct in6_addr *addr)
+{
+	return 0;
+}
+#endif
 extern struct inet6_ifaddr *	ipv6_get_ifaddr(struct in6_addr *addr,
 						struct net_device *dev,
 						int strict);
diff --git a/net/ipv6/exthdrs.c b/net/ipv6/exthdrs.c
index 05afa6b1912b..8d3a0e17314d 100644
--- a/net/ipv6/exthdrs.c
+++ b/net/ipv6/exthdrs.c
@@ -43,6 +43,9 @@
 #include <net/ndisc.h>
 #include <net/ip6_route.h>
 #include <net/addrconf.h>
+#ifdef CONFIG_IPV6_MIP6
+#include <net/xfrm.h>
+#endif
 
 #include <asm/uaccess.h>
 
@@ -219,7 +222,7 @@ static int ipv6_rthdr_rcv(struct sk_buff **skbp)
 {
 	struct sk_buff *skb = *skbp;
 	struct inet6_skb_parm *opt = IP6CB(skb);
-	struct in6_addr *addr;
+	struct in6_addr *addr = NULL;
 	struct in6_addr daddr;
 	int n, i;
 
@@ -244,6 +247,23 @@ static int ipv6_rthdr_rcv(struct sk_buff **skbp)
 
 looped_back:
 	if (hdr->segments_left == 0) {
+		switch (hdr->type) {
+#ifdef CONFIG_IPV6_MIP6
+		case IPV6_SRCRT_TYPE_2:
+			/* Silently discard type 2 header unless it was
+			 * processed by own
+			 */
+			if (!addr) {
+				IP6_INC_STATS_BH(IPSTATS_MIB_INADDRERRORS);
+				kfree_skb(skb);
+				return -1;
+			}
+			break;
+#endif
+		default:
+			break;
+		}
+
 		opt->lastopt = skb->h.raw - skb->nh.raw;
 		opt->srcrt = skb->h.raw - skb->nh.raw;
 		skb->h.raw += (hdr->hdrlen + 1) << 3;
@@ -253,17 +273,29 @@ looped_back:
 		return 1;
 	}
 
-	if (hdr->type != IPV6_SRCRT_TYPE_0) {
+	switch (hdr->type) {
+	case IPV6_SRCRT_TYPE_0:
+		if (hdr->hdrlen & 0x01) {
+			IP6_INC_STATS_BH(IPSTATS_MIB_INHDRERRORS);
+			icmpv6_param_prob(skb, ICMPV6_HDR_FIELD, (&hdr->hdrlen) - skb->nh.raw);
+			return -1;
+		}
+		break;
+#ifdef CONFIG_IPV6_MIP6
+	case IPV6_SRCRT_TYPE_2:
+		/* Silently discard invalid RTH type 2 */
+		if (hdr->hdrlen != 2 || hdr->segments_left != 1) {
+			IP6_INC_STATS_BH(IPSTATS_MIB_INHDRERRORS);
+			kfree_skb(skb);
+			return -1;
+		}
+		break;
+#endif
+	default:
 		IP6_INC_STATS_BH(IPSTATS_MIB_INHDRERRORS);
 		icmpv6_param_prob(skb, ICMPV6_HDR_FIELD, (&hdr->type) - skb->nh.raw);
 		return -1;
 	}
-	
-	if (hdr->hdrlen & 0x01) {
-		IP6_INC_STATS_BH(IPSTATS_MIB_INHDRERRORS);
-		icmpv6_param_prob(skb, ICMPV6_HDR_FIELD, (&hdr->hdrlen) - skb->nh.raw);
-		return -1;
-	}
 
 	/*
 	 *	This is the routing header forwarding algorithm from
@@ -303,6 +335,27 @@ looped_back:
 	addr = rthdr->addr;
 	addr += i - 1;
 
+	switch (hdr->type) {
+#ifdef CONFIG_IPV6_MIP6
+	case IPV6_SRCRT_TYPE_2:
+		if (xfrm6_input_addr(skb, (xfrm_address_t *)addr,
+				     (xfrm_address_t *)&skb->nh.ipv6h->saddr,
+				     IPPROTO_ROUTING) < 0) {
+			IP6_INC_STATS_BH(IPSTATS_MIB_INADDRERRORS);
+			kfree_skb(skb);
+			return -1;
+		}
+		if (!ipv6_chk_home_addr(addr)) {
+			IP6_INC_STATS_BH(IPSTATS_MIB_INADDRERRORS);
+			kfree_skb(skb);
+			return -1;
+		}
+		break;
+#endif
+	default:
+		break;
+	}
+
 	if (ipv6_addr_is_multicast(addr)) {
 		IP6_INC_STATS_BH(IPSTATS_MIB_INADDRERRORS);
 		kfree_skb(skb);
-- 
cgit v1.2.3


From c61a404325093250b676f40ad8f4dd00f3bcab5f Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 19:18:35 -0700
Subject: [IPV6]: Find option offset by type.

This is a helper to search option offset from extension header which
can carry TLV option like destination options header.

Mobile IPv6 home address option will use it.

Based on MIPL2 kernel patch.

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ipv6.h |  2 ++
 net/ipv6/exthdrs.c | 43 +++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 45 insertions(+)

(limited to 'include')

diff --git a/include/net/ipv6.h b/include/net/ipv6.h
index ece7e8a84ffd..c4ea12710576 100644
--- a/include/net/ipv6.h
+++ b/include/net/ipv6.h
@@ -506,6 +506,8 @@ extern int			ipv6_skip_exthdr(const struct sk_buff *, int start,
 
 extern int 			ipv6_ext_hdr(u8 nexthdr);
 
+extern int ipv6_find_tlv(struct sk_buff *skb, int offset, int type);
+
 extern struct ipv6_txoptions *	ipv6_invert_rthdr(struct sock *sk,
 						  struct ipv6_rt_hdr *hdr);
 
diff --git a/net/ipv6/exthdrs.c b/net/ipv6/exthdrs.c
index 8d3a0e17314d..50ff49e518bc 100644
--- a/net/ipv6/exthdrs.c
+++ b/net/ipv6/exthdrs.c
@@ -49,6 +49,49 @@
 
 #include <asm/uaccess.h>
 
+int ipv6_find_tlv(struct sk_buff *skb, int offset, int type)
+{
+	int packet_len = skb->tail - skb->nh.raw;
+	struct ipv6_opt_hdr *hdr;
+	int len;
+
+	if (offset + 2 > packet_len)
+		goto bad;
+	hdr = (struct ipv6_opt_hdr*)(skb->nh.raw + offset);
+	len = ((hdr->hdrlen + 1) << 3);
+
+	if (offset + len > packet_len)
+		goto bad;
+
+	offset += 2;
+	len -= 2;
+
+	while (len > 0) {
+		int opttype = skb->nh.raw[offset];
+		int optlen;
+
+		if (opttype == type)
+			return offset;
+
+		switch (opttype) {
+		case IPV6_TLV_PAD0:
+			optlen = 1;
+			break;
+		default:
+			optlen = skb->nh.raw[offset + 1] + 2;
+			if (optlen > len)
+				goto bad;
+			break;
+		}
+		offset += optlen;
+		len -= optlen;
+	}
+	/* not_found */
+	return -1;
+ bad:
+	return -1;
+}
+
 /*
  *	Parsing tlv encoded headers.
  *
-- 
cgit v1.2.3


From a80ff03e05e4343d647780c116b02ec86078fd24 Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 19:19:50 -0700
Subject: [IPV6]: Allow to replace skbuff by TLV parser.

In receiving Mobile IPv6 home address option which is a TLV carried by
destination options header, kernel will try to mangle source adderss
of packet. Think of cloned skbuff it is required to replace it by the
parser just like routing header case.

This is a framework to achieve that to allow TLV parser to replace
inbound skbuff pointer.

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ipv6.h   |  2 +-
 net/ipv6/exthdrs.c   | 29 +++++++++++++++++++----------
 net/ipv6/ip6_input.c |  2 +-
 3 files changed, 21 insertions(+), 12 deletions(-)

(limited to 'include')

diff --git a/include/net/ipv6.h b/include/net/ipv6.h
index c4ea12710576..8e6ec6063f8c 100644
--- a/include/net/ipv6.h
+++ b/include/net/ipv6.h
@@ -229,7 +229,7 @@ extern int 			ip6_ra_control(struct sock *sk, int sel,
 					       void (*destructor)(struct sock *));
 
 
-extern int			ipv6_parse_hopopts(struct sk_buff *skb);
+extern int			ipv6_parse_hopopts(struct sk_buff **skbp);
 
 extern struct ipv6_txoptions *  ipv6_dup_options(struct sock *sk, struct ipv6_txoptions *opt);
 extern struct ipv6_txoptions *	ipv6_renew_options(struct sock *sk, struct ipv6_txoptions *opt,
diff --git a/net/ipv6/exthdrs.c b/net/ipv6/exthdrs.c
index 50ff49e518bc..1cdd0f0b5d34 100644
--- a/net/ipv6/exthdrs.c
+++ b/net/ipv6/exthdrs.c
@@ -102,7 +102,7 @@ int ipv6_find_tlv(struct sk_buff *skb, int offset, int type)
 
 struct tlvtype_proc {
 	int	type;
-	int	(*func)(struct sk_buff *skb, int offset);
+	int	(*func)(struct sk_buff **skbp, int offset);
 };
 
 /*********************
@@ -111,8 +111,10 @@ struct tlvtype_proc {
 
 /* An unknown option is detected, decide what to do */
 
-static int ip6_tlvopt_unknown(struct sk_buff *skb, int optoff)
+static int ip6_tlvopt_unknown(struct sk_buff **skbp, int optoff)
 {
+	struct sk_buff *skb = *skbp;
+
 	switch ((skb->nh.raw[optoff] & 0xC0) >> 6) {
 	case 0: /* ignore */
 		return 1;
@@ -137,8 +139,9 @@ static int ip6_tlvopt_unknown(struct sk_buff *skb, int optoff)
 
 /* Parse tlv encoded option header (hop-by-hop or destination) */
 
-static int ip6_parse_tlv(struct tlvtype_proc *procs, struct sk_buff *skb)
+static int ip6_parse_tlv(struct tlvtype_proc *procs, struct sk_buff **skbp)
 {
+	struct sk_buff *skb = *skbp;
 	struct tlvtype_proc *curr;
 	int off = skb->h.raw - skb->nh.raw;
 	int len = ((skb->h.raw[1]+1)<<3);
@@ -168,13 +171,13 @@ static int ip6_parse_tlv(struct tlvtype_proc *procs, struct sk_buff *skb)
 					/* type specific length/alignment 
 					   checks will be performed in the 
 					   func(). */
-					if (curr->func(skb, off) == 0)
+					if (curr->func(skbp, off) == 0)
 						return 0;
 					break;
 				}
 			}
 			if (curr->type < 0) {
-				if (ip6_tlvopt_unknown(skb, off) == 0)
+				if (ip6_tlvopt_unknown(skbp, off) == 0)
 					return 0;
 			}
 			break;
@@ -213,7 +216,8 @@ static int ipv6_destopt_rcv(struct sk_buff **skbp)
 	opt->lastopt = skb->h.raw - skb->nh.raw;
 	opt->dst1 = skb->h.raw - skb->nh.raw;
 
-	if (ip6_parse_tlv(tlvprocdestopt_lst, skb)) {
+	if (ip6_parse_tlv(tlvprocdestopt_lst, skbp)) {
+		skb = *skbp;
 		skb->h.raw += ((skb->h.raw[1]+1)<<3);
 		opt->nhoff = opt->dst1;
 		return 1;
@@ -517,8 +521,10 @@ EXPORT_SYMBOL_GPL(ipv6_invert_rthdr);
 
 /* Router Alert as of RFC 2711 */
 
-static int ipv6_hop_ra(struct sk_buff *skb, int optoff)
+static int ipv6_hop_ra(struct sk_buff **skbp, int optoff)
 {
+	struct sk_buff *skb = *skbp;
+
 	if (skb->nh.raw[optoff+1] == 2) {
 		IP6CB(skb)->ra = optoff;
 		return 1;
@@ -531,8 +537,9 @@ static int ipv6_hop_ra(struct sk_buff *skb, int optoff)
 
 /* Jumbo payload */
 
-static int ipv6_hop_jumbo(struct sk_buff *skb, int optoff)
+static int ipv6_hop_jumbo(struct sk_buff **skbp, int optoff)
 {
+	struct sk_buff *skb = *skbp;
 	u32 pkt_len;
 
 	if (skb->nh.raw[optoff+1] != 4 || (optoff&3) != 2) {
@@ -581,8 +588,9 @@ static struct tlvtype_proc tlvprochopopt_lst[] = {
 	{ -1, }
 };
 
-int ipv6_parse_hopopts(struct sk_buff *skb)
+int ipv6_parse_hopopts(struct sk_buff **skbp)
 {
+	struct sk_buff *skb = *skbp;
 	struct inet6_skb_parm *opt = IP6CB(skb);
 
 	/*
@@ -598,7 +606,8 @@ int ipv6_parse_hopopts(struct sk_buff *skb)
 	}
 
 	opt->hop = sizeof(struct ipv6hdr);
-	if (ip6_parse_tlv(tlvprochopopt_lst, skb)) {
+	if (ip6_parse_tlv(tlvprochopopt_lst, skbp)) {
+		skb = *skbp;
 		skb->h.raw += (skb->h.raw[1]+1)<<3;
 		opt->nhoff = sizeof(struct ipv6hdr);
 		return 1;
diff --git a/net/ipv6/ip6_input.c b/net/ipv6/ip6_input.c
index 25c2a9e03895..6b8e6d76a58b 100644
--- a/net/ipv6/ip6_input.c
+++ b/net/ipv6/ip6_input.c
@@ -111,7 +111,7 @@ int ipv6_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt
 	}
 
 	if (hdr->nexthdr == NEXTHDR_HOP) {
-		if (ipv6_parse_hopopts(skb) < 0) {
+		if (ipv6_parse_hopopts(&skb) < 0) {
 			IP6_INC_STATS_BH(IPSTATS_MIB_INHDRERRORS);
 			return 0;
 		}
-- 
cgit v1.2.3


From 842426e719f86cd5709617208efae93ff1a1e2d8 Mon Sep 17 00:00:00 2001
From: Noriaki TAKAMIYA <takamiya@po.ntts.co.jp>
Date: Wed, 23 Aug 2006 19:21:34 -0700
Subject: [IPV6] MIP6: Add home address option definition.

Add home address option definition for Mobile IPv6.
Based on MIPL2 kernel patch.

Signed-off-by: Noriaki TAKAMIYA <takamiya@po.ntts.co.jp>
Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/in6.h  |  1 +
 include/linux/ipv6.h | 10 ++++++++++
 2 files changed, 11 insertions(+)

(limited to 'include')

diff --git a/include/linux/in6.h b/include/linux/in6.h
index 304aaedea305..086ec2ac8c5f 100644
--- a/include/linux/in6.h
+++ b/include/linux/in6.h
@@ -142,6 +142,7 @@ struct in6_flowlabel_req
 #define IPV6_TLV_PADN		1
 #define IPV6_TLV_ROUTERALERT	5
 #define IPV6_TLV_JUMBO		194
+#define IPV6_TLV_HAO		201	/* home address option */
 
 /*
  *	IPV6 socket options
diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index d995662e94c4..5bf4406e26d4 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -86,6 +86,16 @@ struct rt2_hdr {
 #define rt2_type		rt_hdr.type
 };
 
+/*
+ *	home address option in destination options header
+ */
+
+struct ipv6_destopt_hao {
+	__u8			type;
+	__u8			length;
+	struct in6_addr		addr;
+} __attribute__ ((__packed__));
+
 struct ipv6_auth_hdr {
 	__u8  nexthdr;
 	__u8  hdrlen;           /* This one is measured in 32 bit units! */
-- 
cgit v1.2.3


From a831f5bbc89a9978795504be9e1ff412043f8f77 Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 19:24:48 -0700
Subject: [IPV6] MIP6: Add inbound interface of home address option.

Add inbound function of home address option by registering it to TLV
table for destination options header.

Based on MIPL2 kernel patch.

This patch was also written by: Ville Nuorvala <vnuorval@tcs.hut.fi>

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/ipv6.h |  3 ++
 net/ipv6/exthdrs.c   | 84 +++++++++++++++++++++++++++++++++++++++++++++++++++-
 2 files changed, 86 insertions(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index 5bf4406e26d4..db3b2ba0f4f8 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -226,6 +226,9 @@ struct inet6_skb_parm {
 	__u16			dst0;
 	__u16			srcrt;
 	__u16			dst1;
+#ifdef CONFIG_IPV6_MIP6
+	__u16			dsthao;
+#endif
 	__u16			lastopt;
 	__u32			nhoff;
 	__u16			flags;
diff --git a/net/ipv6/exthdrs.c b/net/ipv6/exthdrs.c
index 1cdd0f0b5d34..6a6466bb5f26 100644
--- a/net/ipv6/exthdrs.c
+++ b/net/ipv6/exthdrs.c
@@ -196,8 +196,80 @@ bad:
   Destination options header.
  *****************************/
 
+#ifdef CONFIG_IPV6_MIP6
+static int ipv6_dest_hao(struct sk_buff **skbp, int optoff)
+{
+	struct sk_buff *skb = *skbp;
+	struct ipv6_destopt_hao *hao;
+	struct inet6_skb_parm *opt = IP6CB(skb);
+	struct ipv6hdr *ipv6h = (struct ipv6hdr *)skb->nh.raw;
+	struct in6_addr tmp_addr;
+	int ret;
+
+	if (opt->dsthao) {
+		LIMIT_NETDEBUG(KERN_DEBUG "hao duplicated\n");
+		goto discard;
+	}
+	opt->dsthao = opt->dst1;
+	opt->dst1 = 0;
+
+	hao = (struct ipv6_destopt_hao *)(skb->nh.raw + optoff);
+
+	if (hao->length != 16) {
+		LIMIT_NETDEBUG(
+			KERN_DEBUG "hao invalid option length = %d\n", hao->length);
+		goto discard;
+	}
+
+	if (!(ipv6_addr_type(&hao->addr) & IPV6_ADDR_UNICAST)) {
+		LIMIT_NETDEBUG(
+			KERN_DEBUG "hao is not an unicast addr: " NIP6_FMT "\n", NIP6(hao->addr));
+		goto discard;
+	}
+
+	ret = xfrm6_input_addr(skb, (xfrm_address_t *)&ipv6h->daddr,
+			       (xfrm_address_t *)&hao->addr, IPPROTO_DSTOPTS);
+	if (unlikely(ret < 0))
+		goto discard;
+
+	if (skb_cloned(skb)) {
+		struct sk_buff *skb2 = skb_copy(skb, GFP_ATOMIC);
+		if (skb2 == NULL)
+			goto discard;
+
+		kfree_skb(skb);
+
+		/* update all variable using below by copied skbuff */
+		*skbp = skb = skb2;
+		hao = (struct ipv6_destopt_hao *)(skb2->nh.raw + optoff);
+		ipv6h = (struct ipv6hdr *)skb2->nh.raw;
+	}
+
+	if (skb->ip_summed == CHECKSUM_COMPLETE)
+		skb->ip_summed = CHECKSUM_NONE;
+
+	ipv6_addr_copy(&tmp_addr, &ipv6h->saddr);
+	ipv6_addr_copy(&ipv6h->saddr, &hao->addr);
+	ipv6_addr_copy(&hao->addr, &tmp_addr);
+
+	if (skb->tstamp.off_sec == 0)
+		__net_timestamp(skb);
+
+	return 1;
+
+ discard:
+	kfree_skb(skb);
+	return 0;
+}
+#endif
+
 static struct tlvtype_proc tlvprocdestopt_lst[] = {
-	/* No destination options are defined now */
+#ifdef CONFIG_IPV6_MIP6
+	{
+		.type	= IPV6_TLV_HAO,
+		.func	= ipv6_dest_hao,
+	},
+#endif
 	{-1,			NULL}
 };
 
@@ -205,6 +277,9 @@ static int ipv6_destopt_rcv(struct sk_buff **skbp)
 {
 	struct sk_buff *skb = *skbp;
 	struct inet6_skb_parm *opt = IP6CB(skb);
+#ifdef CONFIG_IPV6_MIP6
+	__u16 dstbuf;
+#endif
 
 	if (!pskb_may_pull(skb, (skb->h.raw-skb->data)+8) ||
 	    !pskb_may_pull(skb, (skb->h.raw-skb->data)+((skb->h.raw[1]+1)<<3))) {
@@ -215,11 +290,18 @@ static int ipv6_destopt_rcv(struct sk_buff **skbp)
 
 	opt->lastopt = skb->h.raw - skb->nh.raw;
 	opt->dst1 = skb->h.raw - skb->nh.raw;
+#ifdef CONFIG_IPV6_MIP6
+	dstbuf = opt->dst1;
+#endif
 
 	if (ip6_parse_tlv(tlvprocdestopt_lst, skbp)) {
 		skb = *skbp;
 		skb->h.raw += ((skb->h.raw[1]+1)<<3);
+#ifdef CONFIG_IPV6_MIP6
+		opt->nhoff = dstbuf;
+#else
 		opt->nhoff = opt->dst1;
+#endif
 		return 1;
 	}
 
-- 
cgit v1.2.3


From 8dd7368dd97def967bbb3aec67b882e8dfd1a528 Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@sunset.davemloft.net>
Date: Wed, 23 Aug 2006 19:25:55 -0700
Subject: [IPV6]: Put dsthao after flags in order to pack inet6_skb_parm
 better.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/ipv6.h | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

(limited to 'include')

diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index db3b2ba0f4f8..1d6d3ccc9413 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -226,12 +226,12 @@ struct inet6_skb_parm {
 	__u16			dst0;
 	__u16			srcrt;
 	__u16			dst1;
-#ifdef CONFIG_IPV6_MIP6
-	__u16			dsthao;
-#endif
 	__u16			lastopt;
 	__u32			nhoff;
 	__u16			flags;
+#ifdef CONFIG_IPV6_MIP6
+	__u16			dsthao;
+#endif
 
 #define IP6SKB_XFRM_TRANSFORMED	1
 };
-- 
cgit v1.2.3


From 2c8d7ca0f76103855ad1f2a930e05683b64a00eb Mon Sep 17 00:00:00 2001
From: Noriaki TAKAMIYA <takamiya@po.ntts.co.jp>
Date: Wed, 23 Aug 2006 20:31:11 -0700
Subject: [IPV6] MIP6: Add routing header type 2 transformation.

Add routing header type 2 transformation for Mobile IPv6.
Based on MIPL2 kernel patch.

Signed-off-by: Noriaki TAKAMIYA <takamiya@po.ntts.co.jp>
Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/mip6.h  |  31 +++++++++
 net/ipv6/Makefile   |   2 +
 net/ipv6/af_inet6.c |   9 +++
 net/ipv6/mip6.c     | 181 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 4 files changed, 223 insertions(+)
 create mode 100644 include/net/mip6.h
 create mode 100644 net/ipv6/mip6.c

(limited to 'include')

diff --git a/include/net/mip6.h b/include/net/mip6.h
new file mode 100644
index 000000000000..644b8b673048
--- /dev/null
+++ b/include/net/mip6.h
@@ -0,0 +1,31 @@
+/*
+ * Copyright (C)2003-2006 Helsinki University of Technology
+ * Copyright (C)2003-2006 USAGI/WIDE Project
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+/*
+ * Authors:
+ *	Noriaki TAKAMIYA @USAGI
+ *	Masahide NAKAMURA @USAGI
+ *	YOSHIFUJI Hideaki @USAGI
+ */
+#ifndef _NET_MIP6_H
+#define _NET_MIP6_H
+
+extern int mip6_init(void);
+extern void mip6_fini(void);
+
+#endif
diff --git a/net/ipv6/Makefile b/net/ipv6/Makefile
index 87e912e31922..0213c6612b58 100644
--- a/net/ipv6/Makefile
+++ b/net/ipv6/Makefile
@@ -14,6 +14,8 @@ ipv6-$(CONFIG_XFRM) += xfrm6_policy.o xfrm6_state.o xfrm6_input.o \
 	xfrm6_output.o
 ipv6-$(CONFIG_NETFILTER) += netfilter.o
 ipv6-$(CONFIG_IPV6_MULTIPLE_TABLES) += fib6_rules.o
+ipv6-$(CONFIG_IPV6_MIP6) += mip6.o
+
 ipv6-objs += $(ipv6-y)
 
 obj-$(CONFIG_INET6_AH) += ah6.o
diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c
index 57ee5ddea96f..fc9c8a99bea6 100644
--- a/net/ipv6/af_inet6.c
+++ b/net/ipv6/af_inet6.c
@@ -59,6 +59,9 @@
 #ifdef CONFIG_IPV6_TUNNEL
 #include <net/ip6_tunnel.h>
 #endif
+#ifdef CONFIG_IPV6_MIP6
+#include <net/mip6.h>
+#endif
 
 #include <asm/uaccess.h>
 #include <asm/system.h>
@@ -857,6 +860,9 @@ static int __init inet6_init(void)
 	ipv6_frag_init();
 	ipv6_nodata_init();
 	ipv6_destopt_init();
+#ifdef CONFIG_IPV6_MIP6
+	mip6_init();
+#endif
 
 	/* Init v6 transport protocols. */
 	udpv6_init();
@@ -919,6 +925,9 @@ static void __exit inet6_exit(void)
  	udp6_proc_exit();
  	tcp6_proc_exit();
  	raw6_proc_exit();
+#endif
+#ifdef CONFIG_IPV6_MIP6
+	mip6_fini();
 #endif
 	/* Cleanup code parts. */
 	sit_cleanup();
diff --git a/net/ipv6/mip6.c b/net/ipv6/mip6.c
new file mode 100644
index 000000000000..63e548b6f81e
--- /dev/null
+++ b/net/ipv6/mip6.c
@@ -0,0 +1,181 @@
+/*
+ * Copyright (C)2003-2006 Helsinki University of Technology
+ * Copyright (C)2003-2006 USAGI/WIDE Project
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+/*
+ * Authors:
+ *	Noriaki TAKAMIYA @USAGI
+ *	Masahide NAKAMURA @USAGI
+ */
+
+#include <linux/config.h>
+#include <linux/module.h>
+#include <linux/skbuff.h>
+#include <linux/ipv6.h>
+#include <net/ipv6.h>
+#include <net/xfrm.h>
+#include <net/mip6.h>
+
+static xfrm_address_t *mip6_xfrm_addr(struct xfrm_state *x, xfrm_address_t *addr)
+{
+	return x->coaddr;
+}
+
+static int mip6_rthdr_input(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct rt2_hdr *rt2 = (struct rt2_hdr *)skb->data;
+
+	if (!ipv6_addr_equal(&rt2->addr, (struct in6_addr *)x->coaddr) &&
+	    !ipv6_addr_any((struct in6_addr *)x->coaddr))
+		return -ENOENT;
+
+	return rt2->rt_hdr.nexthdr;
+}
+
+/* Routing Header type 2 is inserted.
+ * IP Header's dst address is replaced with Routing Header's Home Address.
+ */
+static int mip6_rthdr_output(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct ipv6hdr *iph;
+	struct rt2_hdr *rt2;
+	u8 nexthdr;
+
+	iph = (struct ipv6hdr *)skb->data;
+	iph->payload_len = htons(skb->len - sizeof(*iph));
+
+	nexthdr = *skb->nh.raw;
+	*skb->nh.raw = IPPROTO_ROUTING;
+
+	rt2 = (struct rt2_hdr *)skb->h.raw;
+	rt2->rt_hdr.nexthdr = nexthdr;
+	rt2->rt_hdr.hdrlen = (x->props.header_len >> 3) - 1;
+	rt2->rt_hdr.type = IPV6_SRCRT_TYPE_2;
+	rt2->rt_hdr.segments_left = 1;
+	memset(&rt2->reserved, 0, sizeof(rt2->reserved));
+
+	BUG_TRAP(rt2->rt_hdr.hdrlen == 2);
+
+	memcpy(&rt2->addr, &iph->daddr, sizeof(rt2->addr));
+	memcpy(&iph->daddr, x->coaddr, sizeof(iph->daddr));
+
+	return 0;
+}
+
+static int mip6_rthdr_offset(struct xfrm_state *x, struct sk_buff *skb,
+			     u8 **nexthdr)
+{
+	u16 offset = sizeof(struct ipv6hdr);
+	struct ipv6_opt_hdr *exthdr = (struct ipv6_opt_hdr*)(skb->nh.ipv6h + 1);
+	unsigned int packet_len = skb->tail - skb->nh.raw;
+	int found_rhdr = 0;
+
+	*nexthdr = &skb->nh.ipv6h->nexthdr;
+
+	while (offset + 1 <= packet_len) {
+
+		switch (**nexthdr) {
+		case NEXTHDR_HOP:
+			break;
+		case NEXTHDR_ROUTING:
+			if (offset + 3 <= packet_len) {
+				struct ipv6_rt_hdr *rt;
+				rt = (struct ipv6_rt_hdr *)(skb->nh.raw + offset);
+				if (rt->type != 0)
+					return offset;
+			}
+			found_rhdr = 1;
+			break;
+		case NEXTHDR_DEST:
+			if (ipv6_find_tlv(skb, offset, IPV6_TLV_HAO) >= 0)
+				return offset;
+
+			if (found_rhdr)
+				return offset;
+
+			break;
+		default:
+			return offset;
+		}
+
+		offset += ipv6_optlen(exthdr);
+		*nexthdr = &exthdr->nexthdr;
+		exthdr = (struct ipv6_opt_hdr*)(skb->nh.raw + offset);
+	}
+
+	return offset;
+}
+
+static int mip6_rthdr_init_state(struct xfrm_state *x)
+{
+	if (x->id.spi) {
+		printk(KERN_INFO "%s: spi is not 0: %u\n", __FUNCTION__,
+		       x->id.spi);
+		return -EINVAL;
+	}
+	if (x->props.mode != XFRM_MODE_ROUTEOPTIMIZATION) {
+		printk(KERN_INFO "%s: state's mode is not %u: %u\n",
+		       __FUNCTION__, XFRM_MODE_ROUTEOPTIMIZATION, x->props.mode);
+		return -EINVAL;
+	}
+
+	x->props.header_len = sizeof(struct rt2_hdr);
+
+	return 0;
+}
+
+/*
+ * Do nothing about destroying since it has no specific operation for routing
+ * header type 2 unlike IPsec protocols.
+ */
+static void mip6_rthdr_destroy(struct xfrm_state *x)
+{
+}
+
+static struct xfrm_type mip6_rthdr_type =
+{
+	.description	= "MIP6RT",
+	.owner		= THIS_MODULE,
+	.proto	     	= IPPROTO_ROUTING,
+	.flags		= XFRM_TYPE_NON_FRAGMENT,
+	.init_state	= mip6_rthdr_init_state,
+	.destructor	= mip6_rthdr_destroy,
+	.input		= mip6_rthdr_input,
+	.output		= mip6_rthdr_output,
+	.hdr_offset	= mip6_rthdr_offset,
+	.remote_addr	= mip6_xfrm_addr,
+};
+
+int __init mip6_init(void)
+{
+	printk(KERN_INFO "Mobile IPv6\n");
+
+	if (xfrm_register_type(&mip6_rthdr_type, AF_INET6) < 0) {
+		printk(KERN_INFO "%s: can't add xfrm type(rthdr)\n", __FUNCTION__);
+		goto mip6_rthdr_xfrm_fail;
+	}
+	return 0;
+
+ mip6_rthdr_xfrm_fail:
+	return -EAGAIN;
+}
+
+void __exit mip6_fini(void)
+{
+	if (xfrm_unregister_type(&mip6_rthdr_type, AF_INET6) < 0)
+		printk(KERN_INFO "%s: can't remove xfrm type(rthdr)\n", __FUNCTION__);
+}
-- 
cgit v1.2.3


From 3d126890dd67beffec27c1b6f51c040fc8d0b526 Mon Sep 17 00:00:00 2001
From: Noriaki TAKAMIYA <takamiya@po.ntts.co.jp>
Date: Wed, 23 Aug 2006 20:32:34 -0700
Subject: [IPV6] MIP6: Add destination options header transformation.

Add destination options header transformation for Mobile IPv6.
Based on MIPL2 kernel patch.

This patch was also written by: Ville Nuorvala <vnuorval@tcs.hut.fi>

Signed-off-by: Noriaki TAKAMIYA <takamiya@po.ntts.co.jp>
Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/mip6.h |   3 +
 net/ipv6/mip6.c    | 167 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 170 insertions(+)

(limited to 'include')

diff --git a/include/net/mip6.h b/include/net/mip6.h
index 644b8b673048..42b65bace122 100644
--- a/include/net/mip6.h
+++ b/include/net/mip6.h
@@ -25,6 +25,9 @@
 #ifndef _NET_MIP6_H
 #define _NET_MIP6_H
 
+#define MIP6_OPT_PAD_1	0
+#define MIP6_OPT_PAD_N	1
+
 extern int mip6_init(void);
 extern void mip6_fini(void);
 
diff --git a/net/ipv6/mip6.c b/net/ipv6/mip6.c
index 63e548b6f81e..a8adf891fe0e 100644
--- a/net/ipv6/mip6.c
+++ b/net/ipv6/mip6.c
@@ -35,6 +35,165 @@ static xfrm_address_t *mip6_xfrm_addr(struct xfrm_state *x, xfrm_address_t *addr
 	return x->coaddr;
 }
 
+static inline unsigned int calc_padlen(unsigned int len, unsigned int n)
+{
+	return (n - len + 16) & 0x7;
+}
+
+static inline void *mip6_padn(__u8 *data, __u8 padlen)
+{
+	if (!data)
+		return NULL;
+	if (padlen == 1) {
+		data[0] = MIP6_OPT_PAD_1;
+	} else if (padlen > 1) {
+		data[0] = MIP6_OPT_PAD_N;
+		data[1] = padlen - 2;
+		if (padlen > 2)
+			memset(data+2, 0, data[1]);
+	}
+	return data + padlen;
+}
+
+static int mip6_destopt_input(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct ipv6hdr *iph = skb->nh.ipv6h;
+	struct ipv6_destopt_hdr *destopt = (struct ipv6_destopt_hdr *)skb->data;
+
+	if (!ipv6_addr_equal(&iph->saddr, (struct in6_addr *)x->coaddr) &&
+	    !ipv6_addr_any((struct in6_addr *)x->coaddr))
+		return -ENOENT;
+
+	return destopt->nexthdr;
+}
+
+/* Destination Option Header is inserted.
+ * IP Header's src address is replaced with Home Address Option in
+ * Destination Option Header.
+ */
+static int mip6_destopt_output(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct ipv6hdr *iph;
+	struct ipv6_destopt_hdr *dstopt;
+	struct ipv6_destopt_hao *hao;
+	u8 nexthdr;
+	int len;
+
+	iph = (struct ipv6hdr *)skb->data;
+	iph->payload_len = htons(skb->len - sizeof(*iph));
+
+	nexthdr = *skb->nh.raw;
+	*skb->nh.raw = IPPROTO_DSTOPTS;
+
+	dstopt = (struct ipv6_destopt_hdr *)skb->h.raw;
+	dstopt->nexthdr = nexthdr;
+
+	hao = mip6_padn((char *)(dstopt + 1),
+			calc_padlen(sizeof(*dstopt), 6));
+
+	hao->type = IPV6_TLV_HAO;
+	hao->length = sizeof(*hao) - 2;
+	BUG_TRAP(hao->length == 16);
+
+	len = ((char *)hao - (char *)dstopt) + sizeof(*hao);
+
+	memcpy(&hao->addr, &iph->saddr, sizeof(hao->addr));
+	memcpy(&iph->saddr, x->coaddr, sizeof(iph->saddr));
+
+	BUG_TRAP(len == x->props.header_len);
+	dstopt->hdrlen = (x->props.header_len >> 3) - 1;
+
+	return 0;
+}
+
+static int mip6_destopt_offset(struct xfrm_state *x, struct sk_buff *skb,
+			       u8 **nexthdr)
+{
+	u16 offset = sizeof(struct ipv6hdr);
+	struct ipv6_opt_hdr *exthdr = (struct ipv6_opt_hdr*)(skb->nh.ipv6h + 1);
+	unsigned int packet_len = skb->tail - skb->nh.raw;
+	int found_rhdr = 0;
+
+	*nexthdr = &skb->nh.ipv6h->nexthdr;
+
+	while (offset + 1 <= packet_len) {
+
+		switch (**nexthdr) {
+		case NEXTHDR_HOP:
+			break;
+		case NEXTHDR_ROUTING:
+			found_rhdr = 1;
+			break;
+		case NEXTHDR_DEST:
+			/*
+			 * HAO MUST NOT appear more than once.
+			 * XXX: It is better to try to find by the end of
+			 * XXX: packet if HAO exists.
+			 */
+			if (ipv6_find_tlv(skb, offset, IPV6_TLV_HAO) >= 0) {
+				LIMIT_NETDEBUG(KERN_WARNING "mip6: hao exists already, override\n");
+				return offset;
+			}
+
+			if (found_rhdr)
+				return offset;
+
+			break;
+		default:
+			return offset;
+		}
+
+		offset += ipv6_optlen(exthdr);
+		*nexthdr = &exthdr->nexthdr;
+		exthdr = (struct ipv6_opt_hdr*)(skb->nh.raw + offset);
+	}
+
+	return offset;
+}
+
+static int mip6_destopt_init_state(struct xfrm_state *x)
+{
+	if (x->id.spi) {
+		printk(KERN_INFO "%s: spi is not 0: %u\n", __FUNCTION__,
+		       x->id.spi);
+		return -EINVAL;
+	}
+	if (x->props.mode != XFRM_MODE_ROUTEOPTIMIZATION) {
+		printk(KERN_INFO "%s: state's mode is not %u: %u\n",
+		       __FUNCTION__, XFRM_MODE_ROUTEOPTIMIZATION, x->props.mode);
+		return -EINVAL;
+	}
+
+	x->props.header_len = sizeof(struct ipv6_destopt_hdr) +
+		calc_padlen(sizeof(struct ipv6_destopt_hdr), 6) +
+		sizeof(struct ipv6_destopt_hao);
+	BUG_TRAP(x->props.header_len == 24);
+
+	return 0;
+}
+
+/*
+ * Do nothing about destroying since it has no specific operation for
+ * destination options header unlike IPsec protocols.
+ */
+static void mip6_destopt_destroy(struct xfrm_state *x)
+{
+}
+
+static struct xfrm_type mip6_destopt_type =
+{
+	.description	= "MIP6DESTOPT",
+	.owner		= THIS_MODULE,
+	.proto	     	= IPPROTO_DSTOPTS,
+	.flags		= XFRM_TYPE_NON_FRAGMENT,
+	.init_state	= mip6_destopt_init_state,
+	.destructor	= mip6_destopt_destroy,
+	.input		= mip6_destopt_input,
+	.output		= mip6_destopt_output,
+	.hdr_offset	= mip6_destopt_offset,
+	.local_addr	= mip6_xfrm_addr,
+};
+
 static int mip6_rthdr_input(struct xfrm_state *x, struct sk_buff *skb)
 {
 	struct rt2_hdr *rt2 = (struct rt2_hdr *)skb->data;
@@ -164,6 +323,10 @@ int __init mip6_init(void)
 {
 	printk(KERN_INFO "Mobile IPv6\n");
 
+	if (xfrm_register_type(&mip6_destopt_type, AF_INET6) < 0) {
+		printk(KERN_INFO "%s: can't add xfrm type(destopt)\n", __FUNCTION__);
+		goto mip6_destopt_xfrm_fail;
+	}
 	if (xfrm_register_type(&mip6_rthdr_type, AF_INET6) < 0) {
 		printk(KERN_INFO "%s: can't add xfrm type(rthdr)\n", __FUNCTION__);
 		goto mip6_rthdr_xfrm_fail;
@@ -171,6 +334,8 @@ int __init mip6_init(void)
 	return 0;
 
  mip6_rthdr_xfrm_fail:
+	xfrm_unregister_type(&mip6_destopt_type, AF_INET6);
+ mip6_destopt_xfrm_fail:
 	return -EAGAIN;
 }
 
@@ -178,4 +343,6 @@ void __exit mip6_fini(void)
 {
 	if (xfrm_unregister_type(&mip6_rthdr_type, AF_INET6) < 0)
 		printk(KERN_INFO "%s: can't remove xfrm type(rthdr)\n", __FUNCTION__);
+	if (xfrm_unregister_type(&mip6_destopt_type, AF_INET6) < 0)
+		printk(KERN_INFO "%s: can't remove xfrm type(destopt)\n", __FUNCTION__);
 }
-- 
cgit v1.2.3


From 2b741653b6c824fe7520ee92b6795f11c5f24b24 Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 20:34:26 -0700
Subject: [IPV6] MIP6: Add Mobility header definition.

Add Mobility header definition for Mobile IPv6.
Based on MIPL2 kernel patch.

This patch was also written by: Antti Tuominen <anttit@tcs.hut.fi>

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/in6.h |  1 +
 include/net/flow.h  |  9 +++++++++
 include/net/ipv6.h  |  1 +
 include/net/mip6.h  | 23 +++++++++++++++++++++++
 4 files changed, 34 insertions(+)

(limited to 'include')

diff --git a/include/linux/in6.h b/include/linux/in6.h
index 086ec2ac8c5f..d776829b443f 100644
--- a/include/linux/in6.h
+++ b/include/linux/in6.h
@@ -134,6 +134,7 @@ struct in6_flowlabel_req
 #define IPPROTO_ICMPV6		58	/* ICMPv6			*/
 #define IPPROTO_NONE		59	/* IPv6 no next header		*/
 #define IPPROTO_DSTOPTS		60	/* IPv6 destination options	*/
+#define IPPROTO_MH		135	/* IPv6 mobility header		*/
 
 /*
  *	IPv6 TLV options.
diff --git a/include/net/flow.h b/include/net/flow.h
index 21d988b2058a..e0522914316e 100644
--- a/include/net/flow.h
+++ b/include/net/flow.h
@@ -72,12 +72,21 @@ struct flowi {
 		} dnports;
 
 		__u32		spi;
+
+#ifdef CONFIG_IPV6_MIP6
+		struct {
+			__u8	type;
+		} mht;
+#endif
 	} uli_u;
 #define fl_ip_sport	uli_u.ports.sport
 #define fl_ip_dport	uli_u.ports.dport
 #define fl_icmp_type	uli_u.icmpt.type
 #define fl_icmp_code	uli_u.icmpt.code
 #define fl_ipsec_spi	uli_u.spi
+#ifdef CONFIG_IPV6_MIP6
+#define fl_mh_type	uli_u.mht.type
+#endif
 	__u32           secid;	/* used by xfrm; see secid.txt */
 } __attribute__((__aligned__(BITS_PER_LONG/8)));
 
diff --git a/include/net/ipv6.h b/include/net/ipv6.h
index 8e6ec6063f8c..72bf47b2a4e0 100644
--- a/include/net/ipv6.h
+++ b/include/net/ipv6.h
@@ -40,6 +40,7 @@
 #define NEXTHDR_ICMP		58	/* ICMP for IPv6. */
 #define NEXTHDR_NONE		59	/* No next header */
 #define NEXTHDR_DEST		60	/* Destination options header. */
+#define NEXTHDR_MOBILITY	135	/* Mobility header. */
 
 #define NEXTHDR_MAX		255
 
diff --git a/include/net/mip6.h b/include/net/mip6.h
index 42b65bace122..fd43178faace 100644
--- a/include/net/mip6.h
+++ b/include/net/mip6.h
@@ -28,6 +28,29 @@
 #define MIP6_OPT_PAD_1	0
 #define MIP6_OPT_PAD_N	1
 
+/*
+ * Mobility Header
+ */
+struct ip6_mh {
+	__u8	ip6mh_proto;
+	__u8	ip6mh_hdrlen;
+	__u8	ip6mh_type;
+	__u8	ip6mh_reserved;
+	__u16	ip6mh_cksum;
+	/* Followed by type specific messages */
+	__u8	data[0];
+} __attribute__ ((__packed__));
+
+#define IP6_MH_TYPE_BRR		0   /* Binding Refresh Request */
+#define IP6_MH_TYPE_HOTI	1   /* HOTI Message   */
+#define IP6_MH_TYPE_COTI	2   /* COTI Message  */
+#define IP6_MH_TYPE_HOT		3   /* HOT Message   */
+#define IP6_MH_TYPE_COT		4   /* COT Message  */
+#define IP6_MH_TYPE_BU		5   /* Binding Update */
+#define IP6_MH_TYPE_BACK	6   /* Binding ACK */
+#define IP6_MH_TYPE_BERROR	7   /* Binding Error */
+#define IP6_MH_TYPE_MAX		IP6_MH_TYPE_BERROR
+
 extern int mip6_init(void);
 extern void mip6_fini(void);
 
-- 
cgit v1.2.3


From 7be96f7628469e56f91d51f13b03e9bcff113c7f Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 20:35:31 -0700
Subject: [IPV6] MIP6: Add receiving mobility header functions through raw
 socket.

Like ICMPv6, mobility header is handled through raw socket.
In inbound case, check only whether ICMPv6 error should be sent as a reply
or not by kernel.
Based on MIPL2 kernel patch.

This patch was also written by: Ville Nuorvala <vnuorval@tcs.hut.fi>
This patch was also written by: Antti Tuominen <anttit@tcs.hut.fi>

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/mip6.h |  4 +++
 net/ipv6/mip6.c    | 83 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
 net/ipv6/raw.c     | 29 ++++++++++++++++++-
 3 files changed, 115 insertions(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/net/mip6.h b/include/net/mip6.h
index fd43178faace..68263c6d9996 100644
--- a/include/net/mip6.h
+++ b/include/net/mip6.h
@@ -25,6 +25,9 @@
 #ifndef _NET_MIP6_H
 #define _NET_MIP6_H
 
+#include <linux/skbuff.h>
+#include <net/sock.h>
+
 #define MIP6_OPT_PAD_1	0
 #define MIP6_OPT_PAD_N	1
 
@@ -53,5 +56,6 @@ struct ip6_mh {
 
 extern int mip6_init(void);
 extern void mip6_fini(void);
+extern int mip6_mh_filter(struct sock *sk, struct sk_buff *skb);
 
 #endif
diff --git a/net/ipv6/mip6.c b/net/ipv6/mip6.c
index a8adf891fe0e..7b5f89321482 100644
--- a/net/ipv6/mip6.c
+++ b/net/ipv6/mip6.c
@@ -26,7 +26,10 @@
 #include <linux/module.h>
 #include <linux/skbuff.h>
 #include <linux/ipv6.h>
+#include <linux/icmpv6.h>
+#include <net/sock.h>
 #include <net/ipv6.h>
+#include <net/ip6_checksum.h>
 #include <net/xfrm.h>
 #include <net/mip6.h>
 
@@ -55,6 +58,86 @@ static inline void *mip6_padn(__u8 *data, __u8 padlen)
 	return data + padlen;
 }
 
+static inline void mip6_param_prob(struct sk_buff *skb, int code, int pos)
+{
+	icmpv6_send(skb, ICMPV6_PARAMPROB, code, pos, skb->dev);
+}
+
+static int mip6_mh_len(int type)
+{
+	int len = 0;
+
+	switch (type) {
+	case IP6_MH_TYPE_BRR:
+		len = 0;
+		break;
+	case IP6_MH_TYPE_HOTI:
+	case IP6_MH_TYPE_COTI:
+	case IP6_MH_TYPE_BU:
+	case IP6_MH_TYPE_BACK:
+		len = 1;
+		break;
+	case IP6_MH_TYPE_HOT:
+	case IP6_MH_TYPE_COT:
+	case IP6_MH_TYPE_BERROR:
+		len = 2;
+		break;
+	}
+	return len;
+}
+
+int mip6_mh_filter(struct sock *sk, struct sk_buff *skb)
+{
+	struct ip6_mh *mh;
+	int mhlen;
+
+	if (!pskb_may_pull(skb, (skb->h.raw - skb->data) + 8) ||
+	    !pskb_may_pull(skb, (skb->h.raw - skb->data) + ((skb->h.raw[1] + 1) << 3)))
+		return -1;
+
+	mh = (struct ip6_mh *)skb->h.raw;
+
+	if (mh->ip6mh_hdrlen < mip6_mh_len(mh->ip6mh_type)) {
+		LIMIT_NETDEBUG(KERN_DEBUG "mip6: MH message too short: %d vs >=%d\n",
+			       mh->ip6mh_hdrlen, mip6_mh_len(mh->ip6mh_type));
+		mip6_param_prob(skb, 0, (&mh->ip6mh_hdrlen) - skb->nh.raw);
+		return -1;
+	}
+	mhlen = (mh->ip6mh_hdrlen + 1) << 3;
+
+	if (skb->ip_summed == CHECKSUM_COMPLETE) {
+		skb->ip_summed = CHECKSUM_UNNECESSARY;
+		if (csum_ipv6_magic(&skb->nh.ipv6h->saddr,
+				    &skb->nh.ipv6h->daddr,
+				    mhlen, IPPROTO_MH,
+				    skb->csum)) {
+			LIMIT_NETDEBUG(KERN_DEBUG "mip6: MH hw checksum failed\n");
+			skb->ip_summed = CHECKSUM_NONE;
+		}
+	}
+	if (skb->ip_summed == CHECKSUM_NONE) {
+		if (csum_ipv6_magic(&skb->nh.ipv6h->saddr,
+				    &skb->nh.ipv6h->daddr,
+				    mhlen, IPPROTO_MH,
+				    skb_checksum(skb, 0, mhlen, 0))) {
+			LIMIT_NETDEBUG(KERN_DEBUG "mip6: MH checksum failed [%04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x > %04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x]\n",
+				       NIP6(skb->nh.ipv6h->saddr),
+				       NIP6(skb->nh.ipv6h->daddr));
+			return -1;
+		}
+		skb->ip_summed = CHECKSUM_UNNECESSARY;
+	}
+
+	if (mh->ip6mh_proto != IPPROTO_NONE) {
+		LIMIT_NETDEBUG(KERN_DEBUG "mip6: MH invalid payload proto = %d\n",
+			       mh->ip6mh_proto);
+		mip6_param_prob(skb, 0, (&mh->ip6mh_proto) - skb->nh.raw);
+		return -1;
+	}
+
+	return 0;
+}
+
 static int mip6_destopt_input(struct xfrm_state *x, struct sk_buff *skb)
 {
 	struct ipv6hdr *iph = skb->nh.ipv6h;
diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c
index d4af1cb5e19f..ecca8aae3c4b 100644
--- a/net/ipv6/raw.c
+++ b/net/ipv6/raw.c
@@ -50,6 +50,9 @@
 #include <net/udp.h>
 #include <net/inet_common.h>
 #include <net/tcp_states.h>
+#ifdef CONFIG_IPV6_MIP6
+#include <net/mip6.h>
+#endif
 
 #include <net/rawv6.h>
 #include <net/xfrm.h>
@@ -169,8 +172,32 @@ int ipv6_raw_deliver(struct sk_buff *skb, int nexthdr)
 	sk = __raw_v6_lookup(sk, nexthdr, daddr, saddr, IP6CB(skb)->iif);
 
 	while (sk) {
+		int filtered;
+
 		delivered = 1;
-		if (nexthdr != IPPROTO_ICMPV6 || !icmpv6_filter(sk, skb)) {
+		switch (nexthdr) {
+		case IPPROTO_ICMPV6:
+			filtered = icmpv6_filter(sk, skb);
+			break;
+#ifdef CONFIG_IPV6_MIP6
+		case IPPROTO_MH:
+			/* XXX: To validate MH only once for each packet,
+			 * this is placed here. It should be after checking
+			 * xfrm policy, however it doesn't. The checking xfrm
+			 * policy is placed in rawv6_rcv() because it is
+			 * required for each socket.
+			 */
+			filtered = mip6_mh_filter(sk, skb);
+			break;
+#endif
+		default:
+			filtered = 0;
+			break;
+		}
+
+		if (filtered < 0)
+			break;
+		if (filtered == 0) {
 			struct sk_buff *clone = skb_clone(skb, GFP_ATOMIC);
 
 			/* Not releasing hash table! */
-- 
cgit v1.2.3


From 2ce4272a699c731b9736d76126dc742353e381db Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 20:39:03 -0700
Subject: [IPV6] MIP6: Transformation support mobility header.

Transformation support mobility header.
Based on MIPL2 kernel patch.

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h      |  5 +++++
 net/ipv6/xfrm6_policy.c | 15 +++++++++++++++
 2 files changed, 20 insertions(+)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 7f1630630dcf..13488e7ba68c 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -546,6 +546,11 @@ u16 xfrm_flowi_sport(struct flowi *fl)
 	case IPPROTO_ICMPV6:
 		port = htons(fl->fl_icmp_type);
 		break;
+#ifdef CONFIG_IPV6_MIP6
+	case IPPROTO_MH:
+		port = htons(fl->fl_mh_type);
+		break;
+#endif
 	default:
 		port = 0;	/*XXX*/
 	}
diff --git a/net/ipv6/xfrm6_policy.c b/net/ipv6/xfrm6_policy.c
index 729b4748d6d3..98c2fe449b3f 100644
--- a/net/ipv6/xfrm6_policy.c
+++ b/net/ipv6/xfrm6_policy.c
@@ -18,6 +18,9 @@
 #include <net/ip.h>
 #include <net/ipv6.h>
 #include <net/ip6_route.h>
+#ifdef CONFIG_IPV6_MIP6
+#include <net/mip6.h>
+#endif
 
 static struct dst_ops xfrm6_dst_ops;
 static struct xfrm_policy_afinfo xfrm6_policy_afinfo;
@@ -270,6 +273,18 @@ _decode_session6(struct sk_buff *skb, struct flowi *fl)
 			fl->proto = nexthdr;
 			return;
 
+#ifdef CONFIG_IPV6_MIP6
+		case IPPROTO_MH:
+			if (pskb_may_pull(skb, skb->nh.raw + offset + 3 - skb->data)) {
+				struct ip6_mh *mh;
+				mh = (struct ip6_mh *)exthdr;
+
+				fl->fl_mh_type = mh->ip6mh_type;
+			}
+			fl->proto = nexthdr;
+			return;
+#endif
+
 		/* XXX Why are there these headers? */
 		case IPPROTO_AH:
 		case IPPROTO_ESP:
-- 
cgit v1.2.3


From df0ba92a99ca757039dfa84a929281ea3f7a50e8 Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 20:41:00 -0700
Subject: [XFRM]: Trace which secpath state is reject factor.

For Mobile IPv6 usage, it is required to trace which secpath state is
reject factor in order to notify it to user space (to know the address
which cannot be used route optimized communication).

Based on MIPL2 kernel patch.

This patch was also written by: Henrik Petander <petander@tcs.hut.fi>

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h     |  1 +
 net/xfrm/xfrm_policy.c | 55 +++++++++++++++++++++++++++++++++++++++++++-------
 2 files changed, 49 insertions(+), 7 deletions(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 13488e7ba68c..9ebbdc1dd471 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -273,6 +273,7 @@ struct xfrm_type
 	void			(*destructor)(struct xfrm_state *);
 	int			(*input)(struct xfrm_state *, struct sk_buff *skb);
 	int			(*output)(struct xfrm_state *, struct sk_buff *pskb);
+	int			(*reject)(struct xfrm_state *, struct sk_buff *, struct flowi *);
 	int			(*hdr_offset)(struct xfrm_state *, struct sk_buff *, u8 **);
 	xfrm_address_t		*(*local_addr)(struct xfrm_state *, xfrm_address_t *);
 	xfrm_address_t		*(*remote_addr)(struct xfrm_state *, xfrm_address_t *);
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index ad2a5cba1f5b..d125a2649037 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -988,6 +988,23 @@ error:
 }
 EXPORT_SYMBOL(xfrm_lookup);
 
+static inline int
+xfrm_secpath_reject(int idx, struct sk_buff *skb, struct flowi *fl)
+{
+	struct xfrm_state *x;
+	int err;
+
+	if (!skb->sp || idx < 0 || idx >= skb->sp->len)
+		return 0;
+	x = skb->sp->xvec[idx];
+	if (!x->type->reject)
+		return 0;
+	xfrm_state_hold(x);
+	err = x->type->reject(x, skb, fl);
+	xfrm_state_put(x);
+	return err;
+}
+
 /* When skb is transformed back to its "native" form, we have to
  * check policy restrictions. At the moment we make this in maximally
  * stupid way. Shame on me. :-) Of course, connected sockets must
@@ -1010,6 +1027,13 @@ xfrm_state_ok(struct xfrm_tmpl *tmpl, struct xfrm_state *x,
 		  xfrm_state_addr_cmp(tmpl, x, family));
 }
 
+/*
+ * 0 or more than 0 is returned when validation is succeeded (either bypass
+ * because of optional transport mode, or next index of the mathced secpath
+ * state with the template.
+ * -1 is returned when no matching template is found.
+ * Otherwise "-2 - errored_index" is returned.
+ */
 static inline int
 xfrm_policy_ok(struct xfrm_tmpl *tmpl, struct sec_path *sp, int start,
 	       unsigned short family)
@@ -1024,8 +1048,11 @@ xfrm_policy_ok(struct xfrm_tmpl *tmpl, struct sec_path *sp, int start,
 	for (; idx < sp->len; idx++) {
 		if (xfrm_state_ok(tmpl, sp->xvec[idx], family))
 			return ++idx;
-		if (sp->xvec[idx]->props.mode != XFRM_MODE_TRANSPORT)
+		if (sp->xvec[idx]->props.mode != XFRM_MODE_TRANSPORT) {
+			if (start == -1)
+				start = -2-idx;
 			break;
+		}
 	}
 	return start;
 }
@@ -1046,11 +1073,14 @@ xfrm_decode_session(struct sk_buff *skb, struct flowi *fl, unsigned short family
 }
 EXPORT_SYMBOL(xfrm_decode_session);
 
-static inline int secpath_has_nontransport(struct sec_path *sp, int k)
+static inline int secpath_has_nontransport(struct sec_path *sp, int k, int *idxp)
 {
 	for (; k < sp->len; k++) {
-		if (sp->xvec[k]->props.mode != XFRM_MODE_TRANSPORT)
+		if (sp->xvec[k]->props.mode != XFRM_MODE_TRANSPORT) {
+			if (idxp)
+				*idxp = k;
 			return 1;
+		}
 	}
 
 	return 0;
@@ -1062,6 +1092,8 @@ int __xfrm_policy_check(struct sock *sk, int dir, struct sk_buff *skb,
 	struct xfrm_policy *pol;
 	struct flowi fl;
 	u8 fl_dir = policy_to_flow_dir(dir);
+	int xerr_idx = -1;
+	int *xerr_idxp = &xerr_idx;
 
 	if (xfrm_decode_session(skb, &fl, family) < 0)
 		return 0;
@@ -1086,8 +1118,13 @@ int __xfrm_policy_check(struct sock *sk, int dir, struct sk_buff *skb,
 		pol = flow_cache_lookup(&fl, family, fl_dir,
 					xfrm_policy_lookup);
 
-	if (!pol)
-		return !skb->sp || !secpath_has_nontransport(skb->sp, 0);
+	if (!pol) {
+		if (skb->sp && secpath_has_nontransport(skb->sp, 0, xerr_idxp)) {
+			xfrm_secpath_reject(xerr_idx, skb, &fl);
+			return 0;
+		}
+		return 1;
+	}
 
 	pol->curlft.use_time = (unsigned long)xtime.tv_sec;
 
@@ -1107,11 +1144,14 @@ int __xfrm_policy_check(struct sock *sk, int dir, struct sk_buff *skb,
 		 */
 		for (i = pol->xfrm_nr-1, k = 0; i >= 0; i--) {
 			k = xfrm_policy_ok(pol->xfrm_vec+i, sp, k, family);
-			if (k < 0)
+			if (k < 0) {
+				if (k < -1 && xerr_idxp)
+					*xerr_idxp = -(2+k);
 				goto reject;
+			}
 		}
 
-		if (secpath_has_nontransport(sp, k))
+		if (secpath_has_nontransport(sp, k, xerr_idxp))
 			goto reject;
 
 		xfrm_pol_put(pol);
@@ -1119,6 +1159,7 @@ int __xfrm_policy_check(struct sock *sk, int dir, struct sk_buff *skb,
 	}
 
 reject:
+	xfrm_secpath_reject(xerr_idx, skb, &fl);
 	xfrm_pol_put(pol);
 	return 0;
 }
-- 
cgit v1.2.3


From 97a64b4577ae2bc5599dbd008a3cd9e25de9b9f5 Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 20:44:06 -0700
Subject: [XFRM]: Introduce XFRM_MSG_REPORT.

XFRM_MSG_REPORT is a message as notification of state protocol and
selector from kernel to user-space.

Mobile IPv6 will use it when inbound reject is occurred at route
optimization to make user-space know a binding error requirement.

Based on MIPL2 kernel patch.

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/xfrm.h  | 12 ++++++++++++
 include/net/xfrm.h    |  2 ++
 net/xfrm/xfrm_state.c | 19 +++++++++++++++++++
 net/xfrm/xfrm_user.c  | 46 ++++++++++++++++++++++++++++++++++++++++++++++
 4 files changed, 79 insertions(+)

(limited to 'include')

diff --git a/include/linux/xfrm.h b/include/linux/xfrm.h
index 1d8c1f22c12d..4009f4445fa9 100644
--- a/include/linux/xfrm.h
+++ b/include/linux/xfrm.h
@@ -166,6 +166,10 @@ enum {
 #define XFRM_MSG_NEWAE XFRM_MSG_NEWAE
 	XFRM_MSG_GETAE,
 #define XFRM_MSG_GETAE XFRM_MSG_GETAE
+
+	XFRM_MSG_REPORT,
+#define XFRM_MSG_REPORT XFRM_MSG_REPORT
+
 	__XFRM_MSG_MAX
 };
 #define XFRM_MSG_MAX (__XFRM_MSG_MAX - 1)
@@ -325,12 +329,18 @@ struct xfrm_usersa_flush {
 	__u8				proto;
 };
 
+struct xfrm_user_report {
+	__u8				proto;
+	struct xfrm_selector		sel;
+};
+
 #ifndef __KERNEL__
 /* backwards compatibility for userspace */
 #define XFRMGRP_ACQUIRE		1
 #define XFRMGRP_EXPIRE		2
 #define XFRMGRP_SA		4
 #define XFRMGRP_POLICY		8
+#define XFRMGRP_REPORT		0x10
 #endif
 
 enum xfrm_nlgroups {
@@ -346,6 +356,8 @@ enum xfrm_nlgroups {
 #define XFRMNLGRP_POLICY	XFRMNLGRP_POLICY
 	XFRMNLGRP_AEVENTS,
 #define XFRMNLGRP_AEVENTS	XFRMNLGRP_AEVENTS
+	XFRMNLGRP_REPORT,
+#define XFRMNLGRP_REPORT	XFRMNLGRP_REPORT
 	__XFRMNLGRP_MAX
 };
 #define XFRMNLGRP_MAX	(__XFRMNLGRP_MAX - 1)
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 9ebbdc1dd471..0b223eed4c9b 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -381,6 +381,7 @@ struct xfrm_mgr
 	struct xfrm_policy	*(*compile_policy)(struct sock *sk, int opt, u8 *data, int len, int *dir);
 	int			(*new_mapping)(struct xfrm_state *x, xfrm_address_t *ipaddr, u16 sport);
 	int			(*notify_policy)(struct xfrm_policy *x, int dir, struct km_event *c);
+	int			(*report)(u8 proto, struct xfrm_selector *sel, xfrm_address_t *addr);
 };
 
 extern int xfrm_register_km(struct xfrm_mgr *km);
@@ -1043,6 +1044,7 @@ extern void xfrm_init_pmtu(struct dst_entry *dst);
 extern wait_queue_head_t km_waitq;
 extern int km_new_mapping(struct xfrm_state *x, xfrm_address_t *ipaddr, u16 sport);
 extern void km_policy_expired(struct xfrm_policy *pol, int dir, int hard, u32 pid);
+extern int km_report(u8 proto, struct xfrm_selector *sel, xfrm_address_t *addr);
 
 extern void xfrm_input_init(void);
 extern int xfrm_parse_spi(struct sk_buff *skb, u8 nexthdr, u32 *spi, u32 *seq);
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 3da89c01ea71..a26ef6952c30 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -1055,6 +1055,25 @@ void km_policy_expired(struct xfrm_policy *pol, int dir, int hard, u32 pid)
 }
 EXPORT_SYMBOL(km_policy_expired);
 
+int km_report(u8 proto, struct xfrm_selector *sel, xfrm_address_t *addr)
+{
+	int err = -EINVAL;
+	int ret;
+	struct xfrm_mgr *km;
+
+	read_lock(&xfrm_km_lock);
+	list_for_each_entry(km, &xfrm_km_list, list) {
+		if (km->report) {
+			ret = km->report(proto, sel, addr);
+			if (!ret)
+				err = ret;
+		}
+	}
+	read_unlock(&xfrm_km_lock);
+	return err;
+}
+EXPORT_SYMBOL(km_report);
+
 int xfrm_user_policy(struct sock *sk, int optname, u8 __user *optval, int optlen)
 {
 	int err;
diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c
index 770bd2410749..7303b820bea4 100644
--- a/net/xfrm/xfrm_user.c
+++ b/net/xfrm/xfrm_user.c
@@ -1491,6 +1491,7 @@ static const int xfrm_msg_min[XFRM_NR_MSGTYPES] = {
 	[XFRM_MSG_FLUSHPOLICY - XFRM_MSG_BASE] = NLMSG_LENGTH(0),
 	[XFRM_MSG_NEWAE       - XFRM_MSG_BASE] = XMSGSIZE(xfrm_aevent_id),
 	[XFRM_MSG_GETAE       - XFRM_MSG_BASE] = XMSGSIZE(xfrm_aevent_id),
+	[XFRM_MSG_REPORT      - XFRM_MSG_BASE] = XMSGSIZE(xfrm_user_report),
 };
 
 #undef XMSGSIZE
@@ -2058,12 +2059,57 @@ static int xfrm_send_policy_notify(struct xfrm_policy *xp, int dir, struct km_ev
 
 }
 
+static int build_report(struct sk_buff *skb, u8 proto,
+			struct xfrm_selector *sel, xfrm_address_t *addr)
+{
+	struct xfrm_user_report *ur;
+	struct nlmsghdr *nlh;
+	unsigned char *b = skb->tail;
+
+	nlh = NLMSG_PUT(skb, 0, 0, XFRM_MSG_REPORT, sizeof(*ur));
+	ur = NLMSG_DATA(nlh);
+	nlh->nlmsg_flags = 0;
+
+	ur->proto = proto;
+	memcpy(&ur->sel, sel, sizeof(ur->sel));
+
+	if (addr)
+		RTA_PUT(skb, XFRMA_COADDR, sizeof(*addr), addr);
+
+	nlh->nlmsg_len = skb->tail - b;
+	return skb->len;
+
+nlmsg_failure:
+rtattr_failure:
+	skb_trim(skb, b - skb->data);
+	return -1;
+}
+
+static int xfrm_send_report(u8 proto, struct xfrm_selector *sel,
+			    xfrm_address_t *addr)
+{
+	struct sk_buff *skb;
+	size_t len;
+
+	len = NLMSG_ALIGN(NLMSG_LENGTH(sizeof(struct xfrm_user_report)));
+	skb = alloc_skb(len, GFP_ATOMIC);
+	if (skb == NULL)
+		return -ENOMEM;
+
+	if (build_report(skb, proto, sel, addr) < 0)
+		BUG();
+
+	NETLINK_CB(skb).dst_group = XFRMNLGRP_REPORT;
+	return netlink_broadcast(xfrm_nl, skb, 0, XFRMNLGRP_REPORT, GFP_ATOMIC);
+}
+
 static struct xfrm_mgr netlink_mgr = {
 	.id		= "netlink",
 	.notify		= xfrm_send_state_notify,
 	.acquire	= xfrm_send_acquire,
 	.compile_policy	= xfrm_compile_policy,
 	.notify_policy	= xfrm_send_policy_notify,
+	.report		= xfrm_send_report,
 };
 
 static int __init xfrm_user_init(void)
-- 
cgit v1.2.3


From 4e81bb8336a0ac50289d4d4c7a55e559b994ee8f Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 22:43:30 -0700
Subject: [XFRM] POLICY: sub policy support.

Sub policy is introduced. Main and sub policy are applied the same flow.
(Policy that current kernel uses is named as main.)
It is required another transformation policy management to keep IPsec
and Mobile IPv6 lives separate.
Policy which lives shorter time in kernel should be a sub i.e. normally
main is for IPsec and sub is for Mobile IPv6.
(Such usage as two IPsec policies on different database can be used, too.)

Limitation or TODOs:
 - Sub policy is not supported for per socket one (it is always inserted as main).
 - Current kernel makes cached outbound with flowi to skip searching database.
   However this patch makes it disabled only when "two policies are used and
   the first matched one is bypass case" because neither flowi nor bundle
   information knows about transformation template size.

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/xfrm.h   |   7 ++
 include/net/xfrm.h     |  45 +++++++--
 net/xfrm/xfrm_policy.c | 252 ++++++++++++++++++++++++++++++++++++++++++-------
 3 files changed, 260 insertions(+), 44 deletions(-)

(limited to 'include')

diff --git a/include/linux/xfrm.h b/include/linux/xfrm.h
index 4009f4445fa9..492fb9818747 100644
--- a/include/linux/xfrm.h
+++ b/include/linux/xfrm.h
@@ -102,6 +102,13 @@ struct xfrm_stats {
 	__u32	integrity_failed;
 };
 
+enum
+{
+	XFRM_POLICY_TYPE_MAIN	= 0,
+	XFRM_POLICY_TYPE_SUB	= 1,
+	XFRM_POLICY_TYPE_MAX	= 2
+};
+
 enum
 {
 	XFRM_POLICY_IN	= 0,
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 0b223eed4c9b..4655ca25f808 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -341,6 +341,7 @@ struct xfrm_policy
 	atomic_t		refcnt;
 	struct timer_list	timer;
 
+	u8			type;
 	u32			priority;
 	u32			index;
 	struct xfrm_selector	selector;
@@ -389,6 +390,19 @@ extern int xfrm_unregister_km(struct xfrm_mgr *km);
 
 
 extern struct xfrm_policy *xfrm_policy_list[XFRM_POLICY_MAX*2];
+#ifdef CONFIG_XFRM_SUB_POLICY
+extern struct xfrm_policy *xfrm_policy_list_sub[XFRM_POLICY_MAX*2];
+
+static inline int xfrm_policy_lists_empty(int dir)
+{
+	return (!xfrm_policy_list[dir] && !xfrm_policy_list_sub[dir]);
+}
+#else
+static inline int xfrm_policy_lists_empty(int dir)
+{
+	return (!xfrm_policy_list[dir]);
+}
+#endif
 
 static inline void xfrm_pol_hold(struct xfrm_policy *policy)
 {
@@ -404,6 +418,20 @@ static inline void xfrm_pol_put(struct xfrm_policy *policy)
 		__xfrm_policy_destroy(policy);
 }
 
+#ifdef CONFIG_XFRM_SUB_POLICY
+static inline void xfrm_pols_put(struct xfrm_policy **pols, int npols)
+{
+	int i;
+	for (i = npols - 1; i >= 0; --i)
+		xfrm_pol_put(pols[i]);
+}
+#else
+static inline void xfrm_pols_put(struct xfrm_policy **pols, int npols)
+{
+	xfrm_pol_put(pols[0]);
+}
+#endif
+
 #define XFRM_DST_HSIZE		1024
 
 static __inline__
@@ -737,8 +765,8 @@ static inline int xfrm_policy_check(struct sock *sk, int dir, struct sk_buff *sk
 {
 	if (sk && sk->sk_policy[XFRM_POLICY_IN])
 		return __xfrm_policy_check(sk, dir, skb, family);
-		
-	return	(!xfrm_policy_list[dir] && !skb->sp) ||
+
+	return	(xfrm_policy_lists_empty(dir) && !skb->sp) ||
 		(skb->dst->flags & DST_NOPOLICY) ||
 		__xfrm_policy_check(sk, dir, skb, family);
 }
@@ -758,7 +786,7 @@ extern int __xfrm_route_forward(struct sk_buff *skb, unsigned short family);
 
 static inline int xfrm_route_forward(struct sk_buff *skb, unsigned short family)
 {
-	return	!xfrm_policy_list[XFRM_POLICY_OUT] ||
+	return	xfrm_policy_lists_empty(XFRM_POLICY_OUT) ||
 		(skb->dst->flags & DST_NOXFRM) ||
 		__xfrm_route_forward(skb, family);
 }
@@ -1023,18 +1051,19 @@ static inline int xfrm_dst_lookup(struct xfrm_dst **dst, struct flowi *fl, unsig
 #endif
 
 struct xfrm_policy *xfrm_policy_alloc(gfp_t gfp);
-extern int xfrm_policy_walk(int (*func)(struct xfrm_policy *, int, int, void*), void *);
+extern int xfrm_policy_walk(u8 type, int (*func)(struct xfrm_policy *, int, int, void*), void *);
 int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl);
-struct xfrm_policy *xfrm_policy_bysel_ctx(int dir, struct xfrm_selector *sel,
+struct xfrm_policy *xfrm_policy_bysel_ctx(u8 type, int dir,
+					  struct xfrm_selector *sel,
 					  struct xfrm_sec_ctx *ctx, int delete);
-struct xfrm_policy *xfrm_policy_byid(int dir, u32 id, int delete);
-void xfrm_policy_flush(void);
+struct xfrm_policy *xfrm_policy_byid(u8, int dir, u32 id, int delete);
+void xfrm_policy_flush(u8 type);
 u32 xfrm_get_acqseq(void);
 void xfrm_alloc_spi(struct xfrm_state *x, u32 minspi, u32 maxspi);
 struct xfrm_state * xfrm_find_acq(u8 mode, u32 reqid, u8 proto, 
 				  xfrm_address_t *daddr, xfrm_address_t *saddr, 
 				  int create, unsigned short family);
-extern void xfrm_policy_flush(void);
+extern void xfrm_policy_flush(u8 type);
 extern int xfrm_sk_policy_insert(struct sock *sk, int dir, struct xfrm_policy *pol);
 extern int xfrm_flush_bundles(void);
 extern void xfrm_flush_all_bundles(void);
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index d125a2649037..96de6c76ed57 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -32,6 +32,24 @@ static DEFINE_RWLOCK(xfrm_policy_lock);
 
 struct xfrm_policy *xfrm_policy_list[XFRM_POLICY_MAX*2];
 EXPORT_SYMBOL(xfrm_policy_list);
+#ifdef CONFIG_XFRM_SUB_POLICY
+struct xfrm_policy *xfrm_policy_list_sub[XFRM_POLICY_MAX*2];
+EXPORT_SYMBOL(xfrm_policy_list_sub);
+
+#define XFRM_POLICY_LISTS(type) \
+	((type == XFRM_POLICY_TYPE_SUB) ? xfrm_policy_list_sub : \
+	 xfrm_policy_list)
+#define XFRM_POLICY_LISTHEAD(type, dir) \
+	((type == XFRM_POLICY_TYPE_SUB) ? xfrm_policy_list_sub[dir] : \
+	 xfrm_policy_list[dir])
+#define XFRM_POLICY_LISTHEADP(type, dir) \
+	((type == XFRM_POLICY_TYPE_SUB) ? &xfrm_policy_list_sub[dir] : \
+	 &xfrm_policy_list[dir])
+#else
+#define XFRM_POLICY_LISTS(type)              xfrm_policy_list
+#define XFRM_POLICY_LISTHEAD(type, dif)      xfrm_policy_list[dir]
+#define XFRM_POLICY_LISTHEADP(type, dif)     &xfrm_policy_list[dir]
+#endif
 
 static DEFINE_RWLOCK(xfrm_policy_afinfo_lock);
 static struct xfrm_policy_afinfo *xfrm_policy_afinfo[NPROTO];
@@ -397,7 +415,7 @@ static void xfrm_policy_kill(struct xfrm_policy *policy)
 
 /* Generate new index... KAME seems to generate them ordered by cost
  * of an absolute inpredictability of ordering of rules. This will not pass. */
-static u32 xfrm_gen_index(int dir)
+static u32 xfrm_gen_index(u8 type, int dir)
 {
 	u32 idx;
 	struct xfrm_policy *p;
@@ -408,7 +426,7 @@ static u32 xfrm_gen_index(int dir)
 		idx_generator += 8;
 		if (idx == 0)
 			idx = 8;
-		for (p = xfrm_policy_list[dir]; p; p = p->next) {
+		for (p = XFRM_POLICY_LISTHEAD(type, dir); p; p = p->next) {
 			if (p->index == idx)
 				break;
 		}
@@ -425,7 +443,7 @@ int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl)
 	struct dst_entry *gc_list;
 
 	write_lock_bh(&xfrm_policy_lock);
-	for (p = &xfrm_policy_list[dir]; (pol=*p)!=NULL;) {
+	for (p = XFRM_POLICY_LISTHEADP(policy->type, dir); (pol=*p)!=NULL;) {
 		if (!delpol && memcmp(&policy->selector, &pol->selector, sizeof(pol->selector)) == 0 &&
 		    xfrm_sec_ctx_match(pol->security, policy->security)) {
 			if (excl) {
@@ -452,7 +470,7 @@ int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl)
 	policy->next = *p;
 	*p = policy;
 	atomic_inc(&flow_cache_genid);
-	policy->index = delpol ? delpol->index : xfrm_gen_index(dir);
+	policy->index = delpol ? delpol->index : xfrm_gen_index(policy->type, dir);
 	policy->curlft.add_time = (unsigned long)xtime.tv_sec;
 	policy->curlft.use_time = 0;
 	if (!mod_timer(&policy->timer, jiffies + HZ))
@@ -493,13 +511,14 @@ int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl)
 }
 EXPORT_SYMBOL(xfrm_policy_insert);
 
-struct xfrm_policy *xfrm_policy_bysel_ctx(int dir, struct xfrm_selector *sel,
+struct xfrm_policy *xfrm_policy_bysel_ctx(u8 type, int dir,
+					  struct xfrm_selector *sel,
 					  struct xfrm_sec_ctx *ctx, int delete)
 {
 	struct xfrm_policy *pol, **p;
 
 	write_lock_bh(&xfrm_policy_lock);
-	for (p = &xfrm_policy_list[dir]; (pol=*p)!=NULL; p = &pol->next) {
+	for (p = XFRM_POLICY_LISTHEADP(type, dir); (pol=*p)!=NULL; p = &pol->next) {
 		if ((memcmp(sel, &pol->selector, sizeof(*sel)) == 0) &&
 		    (xfrm_sec_ctx_match(ctx, pol->security))) {
 			xfrm_pol_hold(pol);
@@ -518,12 +537,12 @@ struct xfrm_policy *xfrm_policy_bysel_ctx(int dir, struct xfrm_selector *sel,
 }
 EXPORT_SYMBOL(xfrm_policy_bysel_ctx);
 
-struct xfrm_policy *xfrm_policy_byid(int dir, u32 id, int delete)
+struct xfrm_policy *xfrm_policy_byid(u8 type, int dir, u32 id, int delete)
 {
 	struct xfrm_policy *pol, **p;
 
 	write_lock_bh(&xfrm_policy_lock);
-	for (p = &xfrm_policy_list[dir]; (pol=*p)!=NULL; p = &pol->next) {
+	for (p = XFRM_POLICY_LISTHEADP(type, dir); (pol=*p)!=NULL; p = &pol->next) {
 		if (pol->index == id) {
 			xfrm_pol_hold(pol);
 			if (delete)
@@ -541,15 +560,16 @@ struct xfrm_policy *xfrm_policy_byid(int dir, u32 id, int delete)
 }
 EXPORT_SYMBOL(xfrm_policy_byid);
 
-void xfrm_policy_flush(void)
+void xfrm_policy_flush(u8 type)
 {
 	struct xfrm_policy *xp;
+	struct xfrm_policy **p_list = XFRM_POLICY_LISTS(type);
 	int dir;
 
 	write_lock_bh(&xfrm_policy_lock);
 	for (dir = 0; dir < XFRM_POLICY_MAX; dir++) {
-		while ((xp = xfrm_policy_list[dir]) != NULL) {
-			xfrm_policy_list[dir] = xp->next;
+		while ((xp = p_list[dir]) != NULL) {
+			p_list[dir] = xp->next;
 			write_unlock_bh(&xfrm_policy_lock);
 
 			xfrm_policy_kill(xp);
@@ -562,7 +582,7 @@ void xfrm_policy_flush(void)
 }
 EXPORT_SYMBOL(xfrm_policy_flush);
 
-int xfrm_policy_walk(int (*func)(struct xfrm_policy *, int, int, void*),
+int xfrm_policy_walk(u8 type, int (*func)(struct xfrm_policy *, int, int, void*),
 		     void *data)
 {
 	struct xfrm_policy *xp;
@@ -572,7 +592,7 @@ int xfrm_policy_walk(int (*func)(struct xfrm_policy *, int, int, void*),
 
 	read_lock_bh(&xfrm_policy_lock);
 	for (dir = 0; dir < 2*XFRM_POLICY_MAX; dir++) {
-		for (xp = xfrm_policy_list[dir]; xp; xp = xp->next)
+		for (xp = XFRM_POLICY_LISTHEAD(type, dir); xp; xp = xp->next)
 			count++;
 	}
 
@@ -582,7 +602,7 @@ int xfrm_policy_walk(int (*func)(struct xfrm_policy *, int, int, void*),
 	}
 
 	for (dir = 0; dir < 2*XFRM_POLICY_MAX; dir++) {
-		for (xp = xfrm_policy_list[dir]; xp; xp = xp->next) {
+		for (xp = XFRM_POLICY_LISTHEAD(type, dir); xp; xp = xp->next) {
 			error = func(xp, dir%XFRM_POLICY_MAX, --count, data);
 			if (error)
 				goto out;
@@ -597,13 +617,13 @@ EXPORT_SYMBOL(xfrm_policy_walk);
 
 /* Find policy to apply to this flow. */
 
-static void xfrm_policy_lookup(struct flowi *fl, u16 family, u8 dir,
-			       void **objp, atomic_t **obj_refp)
+static struct xfrm_policy *xfrm_policy_lookup_bytype(u8 type, struct flowi *fl,
+						     u16 family, u8 dir)
 {
 	struct xfrm_policy *pol;
 
 	read_lock_bh(&xfrm_policy_lock);
-	for (pol = xfrm_policy_list[dir]; pol; pol = pol->next) {
+	for (pol = XFRM_POLICY_LISTHEAD(type, dir); pol; pol = pol->next) {
 		struct xfrm_selector *sel = &pol->selector;
 		int match;
 
@@ -620,6 +640,25 @@ static void xfrm_policy_lookup(struct flowi *fl, u16 family, u8 dir,
 		}
 	}
 	read_unlock_bh(&xfrm_policy_lock);
+
+	return pol;
+}
+
+static void xfrm_policy_lookup(struct flowi *fl, u16 family, u8 dir,
+			       void **objp, atomic_t **obj_refp)
+{
+	struct xfrm_policy *pol;
+
+#ifdef CONFIG_XFRM_SUB_POLICY
+	pol = xfrm_policy_lookup_bytype(XFRM_POLICY_TYPE_SUB, fl, family, dir);
+	if (pol)
+		goto end;
+#endif
+	pol = xfrm_policy_lookup_bytype(XFRM_POLICY_TYPE_MAIN, fl, family, dir);
+
+#ifdef CONFIG_XFRM_SUB_POLICY
+ end:
+#endif
 	if ((*objp = (void *) pol) != NULL)
 		*obj_refp = &pol->refcnt;
 }
@@ -665,8 +704,10 @@ static struct xfrm_policy *xfrm_sk_policy_lookup(struct sock *sk, int dir, struc
 
 static void __xfrm_policy_link(struct xfrm_policy *pol, int dir)
 {
-	pol->next = xfrm_policy_list[dir];
-	xfrm_policy_list[dir] = pol;
+	struct xfrm_policy **p_list = XFRM_POLICY_LISTS(pol->type);
+
+	pol->next = p_list[dir];
+	p_list[dir] = pol;
 	xfrm_pol_hold(pol);
 }
 
@@ -675,7 +716,7 @@ static struct xfrm_policy *__xfrm_policy_unlink(struct xfrm_policy *pol,
 {
 	struct xfrm_policy **polp;
 
-	for (polp = &xfrm_policy_list[dir];
+	for (polp = XFRM_POLICY_LISTHEADP(pol->type, dir);
 	     *polp != NULL; polp = &(*polp)->next) {
 		if (*polp == pol) {
 			*polp = pol->next;
@@ -704,12 +745,17 @@ int xfrm_sk_policy_insert(struct sock *sk, int dir, struct xfrm_policy *pol)
 {
 	struct xfrm_policy *old_pol;
 
+#ifdef CONFIG_XFRM_SUB_POLICY
+	if (pol && pol->type != XFRM_POLICY_TYPE_MAIN)
+		return -EINVAL;
+#endif
+
 	write_lock_bh(&xfrm_policy_lock);
 	old_pol = sk->sk_policy[dir];
 	sk->sk_policy[dir] = pol;
 	if (pol) {
 		pol->curlft.add_time = (unsigned long)xtime.tv_sec;
-		pol->index = xfrm_gen_index(XFRM_POLICY_MAX+dir);
+		pol->index = xfrm_gen_index(pol->type, XFRM_POLICY_MAX+dir);
 		__xfrm_policy_link(pol, XFRM_POLICY_MAX+dir);
 	}
 	if (old_pol)
@@ -738,6 +784,7 @@ static struct xfrm_policy *clone_policy(struct xfrm_policy *old, int dir)
 		newp->flags = old->flags;
 		newp->xfrm_nr = old->xfrm_nr;
 		newp->index = old->index;
+		newp->type = old->type;
 		memcpy(newp->xfrm_vec, old->xfrm_vec,
 		       newp->xfrm_nr*sizeof(struct xfrm_tmpl));
 		write_lock_bh(&xfrm_policy_lock);
@@ -764,9 +811,9 @@ int __xfrm_sk_clone_policy(struct sock *sk)
 /* Resolve list of templates for the flow, given policy. */
 
 static int
-xfrm_tmpl_resolve(struct xfrm_policy *policy, struct flowi *fl,
-		  struct xfrm_state **xfrm,
-		  unsigned short family)
+xfrm_tmpl_resolve_one(struct xfrm_policy *policy, struct flowi *fl,
+		      struct xfrm_state **xfrm,
+		      unsigned short family)
 {
 	int nx;
 	int i, error;
@@ -809,6 +856,38 @@ fail:
 	return error;
 }
 
+static int
+xfrm_tmpl_resolve(struct xfrm_policy **pols, int npols, struct flowi *fl,
+		  struct xfrm_state **xfrm,
+		  unsigned short family)
+{
+	int cnx = 0;
+	int error;
+	int ret;
+	int i;
+
+	for (i = 0; i < npols; i++) {
+		if (cnx + pols[i]->xfrm_nr >= XFRM_MAX_DEPTH) {
+			error = -ENOBUFS;
+			goto fail;
+		}
+		ret = xfrm_tmpl_resolve_one(pols[i], fl, &xfrm[cnx], family);
+		if (ret < 0) {
+			error = ret;
+			goto fail;
+		} else
+			cnx += ret;
+	}
+
+	return cnx;
+
+ fail:
+	for (cnx--; cnx>=0; cnx--)
+		xfrm_state_put(xfrm[cnx]);
+	return error;
+
+}
+
 /* Check that the bundle accepts the flow and its components are
  * still valid.
  */
@@ -855,6 +934,11 @@ int xfrm_lookup(struct dst_entry **dst_p, struct flowi *fl,
 		struct sock *sk, int flags)
 {
 	struct xfrm_policy *policy;
+	struct xfrm_policy *pols[XFRM_POLICY_TYPE_MAX];
+	int npols;
+	int pol_dead;
+	int xfrm_nr;
+	int pi;
 	struct xfrm_state *xfrm[XFRM_MAX_DEPTH];
 	struct dst_entry *dst, *dst_orig = *dst_p;
 	int nx = 0;
@@ -866,12 +950,18 @@ int xfrm_lookup(struct dst_entry **dst_p, struct flowi *fl,
 restart:
 	genid = atomic_read(&flow_cache_genid);
 	policy = NULL;
+	for (pi = 0; pi < ARRAY_SIZE(pols); pi++)
+		pols[pi] = NULL;
+	npols = 0;
+	pol_dead = 0;
+	xfrm_nr = 0;
+
 	if (sk && sk->sk_policy[1])
 		policy = xfrm_sk_policy_lookup(sk, XFRM_POLICY_OUT, fl);
 
 	if (!policy) {
 		/* To accelerate a bit...  */
-		if ((dst_orig->flags & DST_NOXFRM) || !xfrm_policy_list[XFRM_POLICY_OUT])
+		if ((dst_orig->flags & DST_NOXFRM) || xfrm_policy_lists_empty(XFRM_POLICY_OUT))
 			return 0;
 
 		policy = flow_cache_lookup(fl, dst_orig->ops->family,
@@ -883,6 +973,9 @@ restart:
 
 	family = dst_orig->ops->family;
 	policy->curlft.use_time = (unsigned long)xtime.tv_sec;
+	pols[0] = policy;
+	npols ++;
+	xfrm_nr += pols[0]->xfrm_nr;
 
 	switch (policy->action) {
 	case XFRM_POLICY_BLOCK:
@@ -891,11 +984,13 @@ restart:
 		goto error;
 
 	case XFRM_POLICY_ALLOW:
+#ifndef CONFIG_XFRM_SUB_POLICY
 		if (policy->xfrm_nr == 0) {
 			/* Flow passes not transformed. */
 			xfrm_pol_put(policy);
 			return 0;
 		}
+#endif
 
 		/* Try to find matching bundle.
 		 *
@@ -911,7 +1006,36 @@ restart:
 		if (dst)
 			break;
 
-		nx = xfrm_tmpl_resolve(policy, fl, xfrm, family);
+#ifdef CONFIG_XFRM_SUB_POLICY
+		if (pols[0]->type != XFRM_POLICY_TYPE_MAIN) {
+			pols[1] = xfrm_policy_lookup_bytype(XFRM_POLICY_TYPE_MAIN,
+							    fl, family,
+							    XFRM_POLICY_OUT);
+			if (pols[1]) {
+				if (pols[1]->action == XFRM_POLICY_BLOCK) {
+					err = -EPERM;
+					goto error;
+				}
+				npols ++;
+				xfrm_nr += pols[1]->xfrm_nr;
+			}
+		}
+
+		/*
+		 * Because neither flowi nor bundle information knows about
+		 * transformation template size. On more than one policy usage
+		 * we can realize whether all of them is bypass or not after
+		 * they are searched. See above not-transformed bypass
+		 * is surrounded by non-sub policy configuration, too.
+		 */
+		if (xfrm_nr == 0) {
+			/* Flow passes not transformed. */
+			xfrm_pols_put(pols, npols);
+			return 0;
+		}
+
+#endif
+		nx = xfrm_tmpl_resolve(pols, npols, fl, xfrm, family);
 
 		if (unlikely(nx<0)) {
 			err = nx;
@@ -924,7 +1048,7 @@ restart:
 				set_current_state(TASK_RUNNING);
 				remove_wait_queue(&km_waitq, &wait);
 
-				nx = xfrm_tmpl_resolve(policy, fl, xfrm, family);
+				nx = xfrm_tmpl_resolve(pols, npols, fl, xfrm, family);
 
 				if (nx == -EAGAIN && signal_pending(current)) {
 					err = -ERESTART;
@@ -932,7 +1056,7 @@ restart:
 				}
 				if (nx == -EAGAIN ||
 				    genid != atomic_read(&flow_cache_genid)) {
-					xfrm_pol_put(policy);
+					xfrm_pols_put(pols, npols);
 					goto restart;
 				}
 				err = nx;
@@ -942,7 +1066,7 @@ restart:
 		}
 		if (nx == 0) {
 			/* Flow passes not transformed. */
-			xfrm_pol_put(policy);
+			xfrm_pols_put(pols, npols);
 			return 0;
 		}
 
@@ -956,8 +1080,14 @@ restart:
 			goto error;
 		}
 
+		for (pi = 0; pi < npols; pi++) {
+			read_lock_bh(&pols[pi]->lock);
+			pol_dead |= pols[pi]->dead;
+			read_unlock_bh(&pols[pi]->lock);
+		}
+
 		write_lock_bh(&policy->lock);
-		if (unlikely(policy->dead || stale_bundle(dst))) {
+		if (unlikely(pol_dead || stale_bundle(dst))) {
 			/* Wow! While we worked on resolving, this
 			 * policy has gone. Retry. It is not paranoia,
 			 * we just cannot enlist new bundle to dead object.
@@ -977,12 +1107,12 @@ restart:
 	}
 	*dst_p = dst;
 	dst_release(dst_orig);
-	xfrm_pol_put(policy);
+ 	xfrm_pols_put(pols, npols);
 	return 0;
 
 error:
 	dst_release(dst_orig);
-	xfrm_pol_put(policy);
+	xfrm_pols_put(pols, npols);
 	*dst_p = NULL;
 	return err;
 }
@@ -1090,6 +1220,10 @@ int __xfrm_policy_check(struct sock *sk, int dir, struct sk_buff *skb,
 			unsigned short family)
 {
 	struct xfrm_policy *pol;
+	struct xfrm_policy *pols[XFRM_POLICY_TYPE_MAX];
+	int npols = 0;
+	int xfrm_nr;
+	int pi;
 	struct flowi fl;
 	u8 fl_dir = policy_to_flow_dir(dir);
 	int xerr_idx = -1;
@@ -1128,22 +1262,50 @@ int __xfrm_policy_check(struct sock *sk, int dir, struct sk_buff *skb,
 
 	pol->curlft.use_time = (unsigned long)xtime.tv_sec;
 
+	pols[0] = pol;
+	npols ++;
+#ifdef CONFIG_XFRM_SUB_POLICY
+	if (pols[0]->type != XFRM_POLICY_TYPE_MAIN) {
+		pols[1] = xfrm_policy_lookup_bytype(XFRM_POLICY_TYPE_MAIN,
+						    &fl, family,
+						    XFRM_POLICY_IN);
+		if (pols[1]) {
+			pols[1]->curlft.use_time = (unsigned long)xtime.tv_sec;
+			npols ++;
+		}
+	}
+#endif
+
 	if (pol->action == XFRM_POLICY_ALLOW) {
 		struct sec_path *sp;
 		static struct sec_path dummy;
+		struct xfrm_tmpl *tp[XFRM_MAX_DEPTH];
+		struct xfrm_tmpl **tpp = tp;
+		int ti = 0;
 		int i, k;
 
 		if ((sp = skb->sp) == NULL)
 			sp = &dummy;
 
+		for (pi = 0; pi < npols; pi++) {
+			if (pols[pi] != pol &&
+			    pols[pi]->action != XFRM_POLICY_ALLOW)
+				goto reject;
+			if (ti + pols[pi]->xfrm_nr >= XFRM_MAX_DEPTH)
+				goto reject_error;
+			for (i = 0; i < pols[pi]->xfrm_nr; i++)
+				tpp[ti++] = &pols[pi]->xfrm_vec[i];
+		}
+		xfrm_nr = ti;
+
 		/* For each tunnel xfrm, find the first matching tmpl.
 		 * For each tmpl before that, find corresponding xfrm.
 		 * Order is _important_. Later we will implement
 		 * some barriers, but at the moment barriers
 		 * are implied between each two transformations.
 		 */
-		for (i = pol->xfrm_nr-1, k = 0; i >= 0; i--) {
-			k = xfrm_policy_ok(pol->xfrm_vec+i, sp, k, family);
+		for (i = xfrm_nr-1, k = 0; i >= 0; i--) {
+			k = xfrm_policy_ok(tpp[i], sp, k, family);
 			if (k < 0) {
 				if (k < -1 && xerr_idxp)
 					*xerr_idxp = -(2+k);
@@ -1154,13 +1316,14 @@ int __xfrm_policy_check(struct sock *sk, int dir, struct sk_buff *skb,
 		if (secpath_has_nontransport(sp, k, xerr_idxp))
 			goto reject;
 
-		xfrm_pol_put(pol);
+		xfrm_pols_put(pols, npols);
 		return 1;
 	}
 
 reject:
 	xfrm_secpath_reject(xerr_idx, skb, &fl);
-	xfrm_pol_put(pol);
+reject_error:
+	xfrm_pols_put(pols, npols);
 	return 0;
 }
 EXPORT_SYMBOL(__xfrm_policy_check);
@@ -1246,6 +1409,23 @@ static void xfrm_prune_bundles(int (*func)(struct dst_entry *))
 
 	read_lock_bh(&xfrm_policy_lock);
 	for (i=0; i<2*XFRM_POLICY_MAX; i++) {
+#ifdef CONFIG_XFRM_SUB_POLICY
+		for (pol = xfrm_policy_list_sub[i]; pol; pol = pol->next) {
+			write_lock(&pol->lock);
+			dstp = &pol->bundles;
+			while ((dst=*dstp) != NULL) {
+				if (func(dst)) {
+					*dstp = dst->next;
+					dst->next = gc_list;
+					gc_list = dst;
+				} else {
+					dstp = &dst->next;
+				}
+			}
+			write_unlock(&pol->lock);
+		}
+
+#endif
 		for (pol = xfrm_policy_list[i]; pol; pol = pol->next) {
 			write_lock(&pol->lock);
 			dstp = &pol->bundles;
-- 
cgit v1.2.3


From 41a49cc3c02ace59d4dddae91ea211c330970ee3 Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 22:48:31 -0700
Subject: [XFRM]: Add sorting interface for state and template.

Under two transformation policies it is required to merge them.
This is a platform to sort state for outbound and templates
for inbound respectively.
It will be used when Mobile IPv6 and IPsec are used at the same time.

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h     | 20 ++++++++++++++++++++
 net/xfrm/xfrm_policy.c | 16 ++++++++++++++--
 net/xfrm/xfrm_state.c  | 38 ++++++++++++++++++++++++++++++++++++++
 3 files changed, 72 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 4655ca25f808..d341603e4ba8 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -254,6 +254,8 @@ struct xfrm_state_afinfo {
 	struct xfrm_state	*(*find_acq)(u8 mode, u32 reqid, u8 proto, 
 					     xfrm_address_t *daddr, xfrm_address_t *saddr, 
 					     int create);
+	int			(*tmpl_sort)(struct xfrm_tmpl **dst, struct xfrm_tmpl **src, int n);
+	int			(*state_sort)(struct xfrm_state **dst, struct xfrm_state **src, int n);
 };
 
 extern int xfrm_state_register_afinfo(struct xfrm_state_afinfo *afinfo);
@@ -1002,6 +1004,24 @@ extern int xfrm_state_add(struct xfrm_state *x);
 extern int xfrm_state_update(struct xfrm_state *x);
 extern struct xfrm_state *xfrm_state_lookup(xfrm_address_t *daddr, u32 spi, u8 proto, unsigned short family);
 extern struct xfrm_state *xfrm_state_lookup_byaddr(xfrm_address_t *daddr, xfrm_address_t *saddr, u8 proto, unsigned short family);
+#ifdef CONFIG_XFRM_SUB_POLICY
+extern int xfrm_tmpl_sort(struct xfrm_tmpl **dst, struct xfrm_tmpl **src,
+			  int n, unsigned short family);
+extern int xfrm_state_sort(struct xfrm_state **dst, struct xfrm_state **src,
+			   int n, unsigned short family);
+#else
+static inline int xfrm_tmpl_sort(struct xfrm_tmpl **dst, struct xfrm_tmpl **src,
+				 int n, unsigned short family)
+{
+	return -ENOSYS;
+}
+
+static inline int xfrm_state_sort(struct xfrm_state **dst, struct xfrm_state **src,
+				  int n, unsigned short family)
+{
+	return -ENOSYS;
+}
+#endif
 extern struct xfrm_state *xfrm_find_acq_byseq(u32 seq);
 extern int xfrm_state_delete(struct xfrm_state *x);
 extern void xfrm_state_flush(u8 proto);
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 96de6c76ed57..1732159ffd01 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -861,6 +861,8 @@ xfrm_tmpl_resolve(struct xfrm_policy **pols, int npols, struct flowi *fl,
 		  struct xfrm_state **xfrm,
 		  unsigned short family)
 {
+	struct xfrm_state *tp[XFRM_MAX_DEPTH];
+	struct xfrm_state **tpp = (npols > 1) ? tp : xfrm;
 	int cnx = 0;
 	int error;
 	int ret;
@@ -871,7 +873,8 @@ xfrm_tmpl_resolve(struct xfrm_policy **pols, int npols, struct flowi *fl,
 			error = -ENOBUFS;
 			goto fail;
 		}
-		ret = xfrm_tmpl_resolve_one(pols[i], fl, &xfrm[cnx], family);
+
+		ret = xfrm_tmpl_resolve_one(pols[i], fl, &tpp[cnx], family);
 		if (ret < 0) {
 			error = ret;
 			goto fail;
@@ -879,11 +882,15 @@ xfrm_tmpl_resolve(struct xfrm_policy **pols, int npols, struct flowi *fl,
 			cnx += ret;
 	}
 
+	/* found states are sorted for outbound processing */
+	if (npols > 1)
+		xfrm_state_sort(xfrm, tpp, cnx, family);
+
 	return cnx;
 
  fail:
 	for (cnx--; cnx>=0; cnx--)
-		xfrm_state_put(xfrm[cnx]);
+		xfrm_state_put(tpp[cnx]);
 	return error;
 
 }
@@ -1280,6 +1287,7 @@ int __xfrm_policy_check(struct sock *sk, int dir, struct sk_buff *skb,
 		struct sec_path *sp;
 		static struct sec_path dummy;
 		struct xfrm_tmpl *tp[XFRM_MAX_DEPTH];
+		struct xfrm_tmpl *stp[XFRM_MAX_DEPTH];
 		struct xfrm_tmpl **tpp = tp;
 		int ti = 0;
 		int i, k;
@@ -1297,6 +1305,10 @@ int __xfrm_policy_check(struct sock *sk, int dir, struct sk_buff *skb,
 				tpp[ti++] = &pols[pi]->xfrm_vec[i];
 		}
 		xfrm_nr = ti;
+		if (npols > 1) {
+			xfrm_tmpl_sort(stp, tpp, xfrm_nr, family);
+			tpp = stp;
+		}
 
 		/* For each tunnel xfrm, find the first matching tmpl.
 		 * For each tmpl before that, find corresponding xfrm.
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index a26ef6952c30..622e92a08d0b 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -728,6 +728,44 @@ xfrm_find_acq(u8 mode, u32 reqid, u8 proto,
 }
 EXPORT_SYMBOL(xfrm_find_acq);
 
+#ifdef CONFIG_XFRM_SUB_POLICY
+int
+xfrm_tmpl_sort(struct xfrm_tmpl **dst, struct xfrm_tmpl **src, int n,
+	       unsigned short family)
+{
+	int err = 0;
+	struct xfrm_state_afinfo *afinfo = xfrm_state_get_afinfo(family);
+	if (!afinfo)
+		return -EAFNOSUPPORT;
+
+	spin_lock_bh(&xfrm_state_lock);
+	if (afinfo->tmpl_sort)
+		err = afinfo->tmpl_sort(dst, src, n);
+	spin_unlock_bh(&xfrm_state_lock);
+	xfrm_state_put_afinfo(afinfo);
+	return err;
+}
+EXPORT_SYMBOL(xfrm_tmpl_sort);
+
+int
+xfrm_state_sort(struct xfrm_state **dst, struct xfrm_state **src, int n,
+		unsigned short family)
+{
+	int err = 0;
+	struct xfrm_state_afinfo *afinfo = xfrm_state_get_afinfo(family);
+	if (!afinfo)
+		return -EAFNOSUPPORT;
+
+	spin_lock_bh(&xfrm_state_lock);
+	if (afinfo->state_sort)
+		err = afinfo->state_sort(dst, src, n);
+	spin_unlock_bh(&xfrm_state_lock);
+	xfrm_state_put_afinfo(afinfo);
+	return err;
+}
+EXPORT_SYMBOL(xfrm_state_sort);
+#endif
+
 /* Silly enough, but I'm lazy to build resolution list */
 
 static struct xfrm_state *__xfrm_find_acq_byseq(u32 seq)
-- 
cgit v1.2.3


From f7b6983f0feeefcd2a594138adcffe640593d8de Mon Sep 17 00:00:00 2001
From: Masahide NAKAMURA <nakam@linux-ipv6.org>
Date: Wed, 23 Aug 2006 22:49:28 -0700
Subject: [XFRM] POLICY: Support netlink socket interface for sub policy.

Sub policy can be used through netlink socket.
PF_KEY uses main only and it is TODO to support sub.

Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/xfrm.h |   7 +++
 include/net/xfrm.h   |   1 +
 net/key/af_key.c     |  18 +++++--
 net/xfrm/xfrm_user.c | 134 ++++++++++++++++++++++++++++++++++++++++++++++-----
 4 files changed, 142 insertions(+), 18 deletions(-)

(limited to 'include')

diff --git a/include/linux/xfrm.h b/include/linux/xfrm.h
index 492fb9818747..14ecd19f4cdc 100644
--- a/include/linux/xfrm.h
+++ b/include/linux/xfrm.h
@@ -230,6 +230,12 @@ enum xfrm_ae_ftype_t {
 #define XFRM_AE_MAX (__XFRM_AE_MAX - 1)
 };
 
+struct xfrm_userpolicy_type {
+	__u8		type;
+	__u16		reserved1;
+	__u8		reserved2;
+};
+
 /* Netlink message attributes.  */
 enum xfrm_attr_type_t {
 	XFRMA_UNSPEC,
@@ -248,6 +254,7 @@ enum xfrm_attr_type_t {
 	XFRMA_SRCADDR,		/* xfrm_address_t */
 	XFRMA_COADDR,		/* xfrm_address_t */
 	XFRMA_LASTUSED,
+	XFRMA_POLICY_TYPE,	/* struct xfrm_userpolicy_type */
 	__XFRMA_MAX
 
 #define XFRMA_MAX (__XFRMA_MAX - 1)
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index d341603e4ba8..c75b3287d8f8 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -203,6 +203,7 @@ struct km_event
 		u32 proto;
 		u32 byid;
 		u32 aevent;
+		u32 type;
 	} data;
 
 	u32	seq;
diff --git a/net/key/af_key.c b/net/key/af_key.c
index 19e047b0e678..83b443ddc72f 100644
--- a/net/key/af_key.c
+++ b/net/key/af_key.c
@@ -1731,7 +1731,8 @@ static u32 gen_reqid(void)
 		++reqid;
 		if (reqid == 0)
 			reqid = IPSEC_MANUAL_REQID_MAX+1;
-		if (xfrm_policy_walk(check_reqid, (void*)&reqid) != -EEXIST)
+		if (xfrm_policy_walk(XFRM_POLICY_TYPE_MAIN, check_reqid,
+				     (void*)&reqid) != -EEXIST)
 			return reqid;
 	} while (reqid != start);
 	return 0;
@@ -2268,7 +2269,8 @@ static int pfkey_spddelete(struct sock *sk, struct sk_buff *skb, struct sadb_msg
 			return err;
 	}
 
-	xp = xfrm_policy_bysel_ctx(pol->sadb_x_policy_dir-1, &sel, tmp.security, 1);
+	xp = xfrm_policy_bysel_ctx(XFRM_POLICY_TYPE_MAIN, pol->sadb_x_policy_dir-1,
+				   &sel, tmp.security, 1);
 	security_xfrm_policy_free(&tmp);
 	if (xp == NULL)
 		return -ENOENT;
@@ -2330,7 +2332,7 @@ static int pfkey_spdget(struct sock *sk, struct sk_buff *skb, struct sadb_msg *h
 	if (dir >= XFRM_POLICY_MAX)
 		return -EINVAL;
 
-	xp = xfrm_policy_byid(dir, pol->sadb_x_policy_id,
+	xp = xfrm_policy_byid(XFRM_POLICY_TYPE_MAIN, dir, pol->sadb_x_policy_id,
 			      hdr->sadb_msg_type == SADB_X_SPDDELETE2);
 	if (xp == NULL)
 		return -ENOENT;
@@ -2378,7 +2380,7 @@ static int pfkey_spddump(struct sock *sk, struct sk_buff *skb, struct sadb_msg *
 {
 	struct pfkey_dump_data data = { .skb = skb, .hdr = hdr, .sk = sk };
 
-	return xfrm_policy_walk(dump_sp, &data);
+	return xfrm_policy_walk(XFRM_POLICY_TYPE_MAIN, dump_sp, &data);
 }
 
 static int key_notify_policy_flush(struct km_event *c)
@@ -2405,7 +2407,8 @@ static int pfkey_spdflush(struct sock *sk, struct sk_buff *skb, struct sadb_msg
 {
 	struct km_event c;
 
-	xfrm_policy_flush();
+	xfrm_policy_flush(XFRM_POLICY_TYPE_MAIN);
+	c.data.type = XFRM_POLICY_TYPE_MAIN;
 	c.event = XFRM_MSG_FLUSHPOLICY;
 	c.pid = hdr->sadb_msg_pid;
 	c.seq = hdr->sadb_msg_seq;
@@ -2667,6 +2670,9 @@ static int pfkey_send_notify(struct xfrm_state *x, struct km_event *c)
 
 static int pfkey_send_policy_notify(struct xfrm_policy *xp, int dir, struct km_event *c)
 {
+	if (xp && xp->type != XFRM_POLICY_TYPE_MAIN)
+		return 0;
+
 	switch (c->event) {
 	case XFRM_MSG_POLEXPIRE:
 		return key_notify_policy_expire(xp, c);
@@ -2675,6 +2681,8 @@ static int pfkey_send_policy_notify(struct xfrm_policy *xp, int dir, struct km_e
 	case XFRM_MSG_UPDPOLICY:
 		return key_notify_policy(xp, dir, c);
 	case XFRM_MSG_FLUSHPOLICY:
+		if (c->data.type != XFRM_POLICY_TYPE_MAIN)
+			break;
 		return key_notify_policy_flush(c);
 	default:
 		printk("pfkey: Unknown policy event %d\n", c->event);
diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c
index 7303b820bea4..c59a78d2923a 100644
--- a/net/xfrm/xfrm_user.c
+++ b/net/xfrm/xfrm_user.c
@@ -786,6 +786,22 @@ static int verify_policy_dir(__u8 dir)
 	return 0;
 }
 
+static int verify_policy_type(__u8 type)
+{
+	switch (type) {
+	case XFRM_POLICY_TYPE_MAIN:
+#ifdef CONFIG_XFRM_SUB_POLICY
+	case XFRM_POLICY_TYPE_SUB:
+#endif
+		break;
+
+	default:
+		return -EINVAL;
+	};
+
+	return 0;
+}
+
 static int verify_newpolicy_info(struct xfrm_userpolicy_info *p)
 {
 	switch (p->share) {
@@ -879,6 +895,29 @@ static int copy_from_user_tmpl(struct xfrm_policy *pol, struct rtattr **xfrma)
 	return 0;
 }
 
+static int copy_from_user_policy_type(u8 *tp, struct rtattr **xfrma)
+{
+	struct rtattr *rt = xfrma[XFRMA_POLICY_TYPE-1];
+	struct xfrm_userpolicy_type *upt;
+	__u8 type = XFRM_POLICY_TYPE_MAIN;
+	int err;
+
+	if (rt) {
+		if (rt->rta_len < sizeof(*upt))
+			return -EINVAL;
+
+		upt = RTA_DATA(rt);
+		type = upt->type;
+	}
+
+	err = verify_policy_type(type);
+	if (err)
+		return err;
+
+	*tp = type;
+	return 0;
+}
+
 static void copy_from_user_policy(struct xfrm_policy *xp, struct xfrm_userpolicy_info *p)
 {
 	xp->priority = p->priority;
@@ -917,16 +956,20 @@ static struct xfrm_policy *xfrm_policy_construct(struct xfrm_userpolicy_info *p,
 
 	copy_from_user_policy(xp, p);
 
+	err = copy_from_user_policy_type(&xp->type, xfrma);
+	if (err)
+		goto error;
+
 	if (!(err = copy_from_user_tmpl(xp, xfrma)))
 		err = copy_from_user_sec_ctx(xp, xfrma);
-
-	if (err) {
-		*errp = err;
-		kfree(xp);
-		xp = NULL;
-	}
+	if (err)
+		goto error;
 
 	return xp;
+ error:
+	*errp = err;
+	kfree(xp);
+	return NULL;
 }
 
 static int xfrm_add_policy(struct sk_buff *skb, struct nlmsghdr *nlh, void **xfrma)
@@ -1037,6 +1080,29 @@ static inline int copy_to_user_sec_ctx(struct xfrm_policy *xp, struct sk_buff *s
 	return 0;
 }
 
+#ifdef CONFIG_XFRM_SUB_POLICY
+static int copy_to_user_policy_type(struct xfrm_policy *xp, struct sk_buff *skb)
+{
+	struct xfrm_userpolicy_type upt;
+
+	memset(&upt, 0, sizeof(upt));
+	upt.type = xp->type;
+
+	RTA_PUT(skb, XFRMA_POLICY_TYPE, sizeof(upt), &upt);
+
+	return 0;
+
+rtattr_failure:
+	return -1;
+}
+
+#else
+static inline int copy_to_user_policy_type(struct xfrm_policy *xp, struct sk_buff *skb)
+{
+	return 0;
+}
+#endif
+
 static int dump_one_policy(struct xfrm_policy *xp, int dir, int count, void *ptr)
 {
 	struct xfrm_dump_info *sp = ptr;
@@ -1060,6 +1126,8 @@ static int dump_one_policy(struct xfrm_policy *xp, int dir, int count, void *ptr
 		goto nlmsg_failure;
 	if (copy_to_user_sec_ctx(xp, skb))
 		goto nlmsg_failure;
+	if (copy_to_user_policy_type(xp, skb) < 0)
+		goto nlmsg_failure;
 
 	nlh->nlmsg_len = skb->tail - b;
 out:
@@ -1081,7 +1149,10 @@ static int xfrm_dump_policy(struct sk_buff *skb, struct netlink_callback *cb)
 	info.nlmsg_flags = NLM_F_MULTI;
 	info.this_idx = 0;
 	info.start_idx = cb->args[0];
-	(void) xfrm_policy_walk(dump_one_policy, &info);
+	(void) xfrm_policy_walk(XFRM_POLICY_TYPE_MAIN, dump_one_policy, &info);
+#ifdef CONFIG_XFRM_SUB_POLICY
+	(void) xfrm_policy_walk(XFRM_POLICY_TYPE_SUB, dump_one_policy, &info);
+#endif
 	cb->args[0] = info.this_idx;
 
 	return skb->len;
@@ -1117,6 +1188,7 @@ static int xfrm_get_policy(struct sk_buff *skb, struct nlmsghdr *nlh, void **xfr
 {
 	struct xfrm_policy *xp;
 	struct xfrm_userpolicy_id *p;
+	__u8 type = XFRM_POLICY_TYPE_MAIN;
 	int err;
 	struct km_event c;
 	int delete;
@@ -1124,12 +1196,16 @@ static int xfrm_get_policy(struct sk_buff *skb, struct nlmsghdr *nlh, void **xfr
 	p = NLMSG_DATA(nlh);
 	delete = nlh->nlmsg_type == XFRM_MSG_DELPOLICY;
 
+	err = copy_from_user_policy_type(&type, (struct rtattr **)xfrma);
+	if (err)
+		return err;
+
 	err = verify_policy_dir(p->dir);
 	if (err)
 		return err;
 
 	if (p->index)
-		xp = xfrm_policy_byid(p->dir, p->index, delete);
+		xp = xfrm_policy_byid(type, p->dir, p->index, delete);
 	else {
 		struct rtattr **rtattrs = (struct rtattr **)xfrma;
 		struct rtattr *rt = rtattrs[XFRMA_SEC_CTX-1];
@@ -1146,7 +1222,7 @@ static int xfrm_get_policy(struct sk_buff *skb, struct nlmsghdr *nlh, void **xfr
 			if ((err = security_xfrm_policy_alloc(&tmp, uctx)))
 				return err;
 		}
-		xp = xfrm_policy_bysel_ctx(p->dir, &p->sel, tmp.security, delete);
+		xp = xfrm_policy_bysel_ctx(type, p->dir, &p->sel, tmp.security, delete);
 		security_xfrm_policy_free(&tmp);
 	}
 	if (xp == NULL)
@@ -1329,9 +1405,16 @@ out:
 
 static int xfrm_flush_policy(struct sk_buff *skb, struct nlmsghdr *nlh, void **xfrma)
 {
-struct km_event c;
+	struct km_event c;
+	__u8 type = XFRM_POLICY_TYPE_MAIN;
+	int err;
+
+	err = copy_from_user_policy_type(&type, (struct rtattr **)xfrma);
+	if (err)
+		return err;
 
-	xfrm_policy_flush();
+	xfrm_policy_flush(type);
+	c.data.type = type;
 	c.event = nlh->nlmsg_type;
 	c.seq = nlh->nlmsg_seq;
 	c.pid = nlh->nlmsg_pid;
@@ -1344,10 +1427,15 @@ static int xfrm_add_pol_expire(struct sk_buff *skb, struct nlmsghdr *nlh, void *
 	struct xfrm_policy *xp;
 	struct xfrm_user_polexpire *up = NLMSG_DATA(nlh);
 	struct xfrm_userpolicy_info *p = &up->pol;
+	__u8 type = XFRM_POLICY_TYPE_MAIN;
 	int err = -ENOENT;
 
+	err = copy_from_user_policy_type(&type, (struct rtattr **)xfrma);
+	if (err)
+		return err;
+
 	if (p->index)
-		xp = xfrm_policy_byid(p->dir, p->index, 0);
+		xp = xfrm_policy_byid(type, p->dir, p->index, 0);
 	else {
 		struct rtattr **rtattrs = (struct rtattr **)xfrma;
 		struct rtattr *rt = rtattrs[XFRMA_SEC_CTX-1];
@@ -1364,7 +1452,7 @@ static int xfrm_add_pol_expire(struct sk_buff *skb, struct nlmsghdr *nlh, void *
 			if ((err = security_xfrm_policy_alloc(&tmp, uctx)))
 				return err;
 		}
-		xp = xfrm_policy_bysel_ctx(p->dir, &p->sel, tmp.security, 0);
+		xp = xfrm_policy_bysel_ctx(type, p->dir, &p->sel, tmp.security, 0);
 		security_xfrm_policy_free(&tmp);
 	}
 
@@ -1818,6 +1906,8 @@ static int build_acquire(struct sk_buff *skb, struct xfrm_state *x,
 		goto nlmsg_failure;
 	if (copy_to_user_state_sec_ctx(x, skb))
 		goto nlmsg_failure;
+	if (copy_to_user_policy_type(xp, skb) < 0)
+		goto nlmsg_failure;
 
 	nlh->nlmsg_len = skb->tail - b;
 	return skb->len;
@@ -1898,6 +1988,7 @@ static struct xfrm_policy *xfrm_compile_policy(struct sock *sk, int opt,
 	}
 
 	copy_from_user_policy(xp, p);
+	xp->type = XFRM_POLICY_TYPE_MAIN;
 	copy_templates(xp, ut, nr);
 
 	if (!xp->security) {
@@ -1931,6 +2022,8 @@ static int build_polexpire(struct sk_buff *skb, struct xfrm_policy *xp,
 		goto nlmsg_failure;
 	if (copy_to_user_sec_ctx(xp, skb))
 		goto nlmsg_failure;
+	if (copy_to_user_policy_type(xp, skb) < 0)
+		goto nlmsg_failure;
 	upe->hard = !!hard;
 
 	nlh->nlmsg_len = skb->tail - b;
@@ -2002,6 +2095,8 @@ static int xfrm_notify_policy(struct xfrm_policy *xp, int dir, struct km_event *
 	copy_to_user_policy(xp, p, dir);
 	if (copy_to_user_tmpl(xp, skb) < 0)
 		goto nlmsg_failure;
+	if (copy_to_user_policy_type(xp, skb) < 0)
+		goto nlmsg_failure;
 
 	nlh->nlmsg_len = skb->tail - b;
 
@@ -2019,6 +2114,9 @@ static int xfrm_notify_policy_flush(struct km_event *c)
 	struct nlmsghdr *nlh;
 	struct sk_buff *skb;
 	unsigned char *b;
+#ifdef CONFIG_XFRM_SUB_POLICY
+	struct xfrm_userpolicy_type upt;
+#endif
 	int len = NLMSG_LENGTH(0);
 
 	skb = alloc_skb(len, GFP_ATOMIC);
@@ -2028,6 +2126,13 @@ static int xfrm_notify_policy_flush(struct km_event *c)
 
 
 	nlh = NLMSG_PUT(skb, c->pid, c->seq, XFRM_MSG_FLUSHPOLICY, 0);
+	nlh->nlmsg_flags = 0;
+
+#ifdef CONFIG_XFRM_SUB_POLICY
+	memset(&upt, 0, sizeof(upt));
+	upt.type = c->data.type;
+	RTA_PUT(skb, XFRMA_POLICY_TYPE, sizeof(upt), &upt);
+#endif
 
 	nlh->nlmsg_len = skb->tail - b;
 
@@ -2035,6 +2140,9 @@ static int xfrm_notify_policy_flush(struct km_event *c)
 	return netlink_broadcast(xfrm_nl, skb, 0, XFRMNLGRP_POLICY, GFP_ATOMIC);
 
 nlmsg_failure:
+#ifdef CONFIG_XFRM_SUB_POLICY
+rtattr_failure:
+#endif
 	kfree_skb(skb);
 	return -1;
 }
-- 
cgit v1.2.3


From 2770834c9f44afd1bfa13914c7285470775af657 Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@sunset.davemloft.net>
Date: Thu, 24 Aug 2006 00:13:10 -0700
Subject: [XFRM]: Pull xfrm_state_bydst hash table knowledge out of afinfo.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h     |  16 -------
 net/ipv4/xfrm4_state.c |  53 ------------------------
 net/ipv6/xfrm6_state.c |  56 -------------------------
 net/xfrm/xfrm_state.c  | 110 ++++++++++++++++++++++++++++++++++++++++++++-----
 4 files changed, 100 insertions(+), 135 deletions(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index c75b3287d8f8..cc83443f301e 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -243,7 +243,6 @@ extern int __xfrm_state_delete(struct xfrm_state *x);
 
 struct xfrm_state_afinfo {
 	unsigned short		family;
-	struct list_head	*state_bydst;
 	struct list_head	*state_bysrc;
 	struct list_head	*state_byspi;
 	int			(*init_flags)(struct xfrm_state *x);
@@ -252,9 +251,6 @@ struct xfrm_state_afinfo {
 						xfrm_address_t *daddr, xfrm_address_t *saddr);
 	struct xfrm_state	*(*state_lookup)(xfrm_address_t *daddr, u32 spi, u8 proto);
 	struct xfrm_state	*(*state_lookup_byaddr)(xfrm_address_t *daddr, xfrm_address_t *saddr, u8 proto);
-	struct xfrm_state	*(*find_acq)(u8 mode, u32 reqid, u8 proto, 
-					     xfrm_address_t *daddr, xfrm_address_t *saddr, 
-					     int create);
 	int			(*tmpl_sort)(struct xfrm_tmpl **dst, struct xfrm_tmpl **src, int n);
 	int			(*state_sort)(struct xfrm_state **dst, struct xfrm_state **src, int n);
 };
@@ -455,18 +451,6 @@ unsigned __xfrm6_dst_hash(xfrm_address_t *addr)
 	return h;
 }
 
-static __inline__
-unsigned xfrm_dst_hash(xfrm_address_t *addr, unsigned short family)
-{
-	switch (family) {
-	case AF_INET:
-		return __xfrm4_dst_hash(addr);
-	case AF_INET6:
-		return __xfrm6_dst_hash(addr);
-	}
-	return 0;
-}
-
 static __inline__
 unsigned __xfrm4_src_hash(xfrm_address_t *addr)
 {
diff --git a/net/ipv4/xfrm4_state.c b/net/ipv4/xfrm4_state.c
index 616be131b4e3..9dc1afc17b6d 100644
--- a/net/ipv4/xfrm4_state.c
+++ b/net/ipv4/xfrm4_state.c
@@ -88,65 +88,12 @@ __xfrm4_state_lookup_byaddr(xfrm_address_t *daddr, xfrm_address_t *saddr,
 	return NULL;
 }
 
-static struct xfrm_state *
-__xfrm4_find_acq(u8 mode, u32 reqid, u8 proto, 
-		 xfrm_address_t *daddr, xfrm_address_t *saddr, 
-		 int create)
-{
-	struct xfrm_state *x, *x0;
-	unsigned h = __xfrm4_dst_hash(daddr);
-
-	x0 = NULL;
-
-	list_for_each_entry(x, xfrm4_state_afinfo.state_bydst+h, bydst) {
-		if (x->props.family == AF_INET &&
-		    daddr->a4 == x->id.daddr.a4 &&
-		    mode == x->props.mode &&
-		    proto == x->id.proto &&
-		    saddr->a4 == x->props.saddr.a4 &&
-		    reqid == x->props.reqid &&
-		    x->km.state == XFRM_STATE_ACQ &&
-		    !x->id.spi) {
-			    x0 = x;
-			    break;
-		    }
-	}
-	if (!x0 && create && (x0 = xfrm_state_alloc()) != NULL) {
-		x0->sel.daddr.a4 = daddr->a4;
-		x0->sel.saddr.a4 = saddr->a4;
-		x0->sel.prefixlen_d = 32;
-		x0->sel.prefixlen_s = 32;
-		x0->props.saddr.a4 = saddr->a4;
-		x0->km.state = XFRM_STATE_ACQ;
-		x0->id.daddr.a4 = daddr->a4;
-		x0->id.proto = proto;
-		x0->props.family = AF_INET;
-		x0->props.mode = mode;
-		x0->props.reqid = reqid;
-		x0->props.family = AF_INET;
-		x0->lft.hard_add_expires_seconds = XFRM_ACQ_EXPIRES;
-		xfrm_state_hold(x0);
-		x0->timer.expires = jiffies + XFRM_ACQ_EXPIRES*HZ;
-		add_timer(&x0->timer);
-		xfrm_state_hold(x0);
-		list_add_tail(&x0->bydst, xfrm4_state_afinfo.state_bydst+h);
-		h = __xfrm4_src_hash(saddr);
-		xfrm_state_hold(x0);
-		list_add_tail(&x0->bysrc, xfrm4_state_afinfo.state_bysrc+h);
-		wake_up(&km_waitq);
-	}
-	if (x0)
-		xfrm_state_hold(x0);
-	return x0;
-}
-
 static struct xfrm_state_afinfo xfrm4_state_afinfo = {
 	.family			= AF_INET,
 	.init_flags		= xfrm4_init_flags,
 	.init_tempsel		= __xfrm4_init_tempsel,
 	.state_lookup		= __xfrm4_state_lookup,
 	.state_lookup_byaddr	= __xfrm4_state_lookup_byaddr,
-	.find_acq		= __xfrm4_find_acq,
 };
 
 void __init xfrm4_state_init(void)
diff --git a/net/ipv6/xfrm6_state.c b/net/ipv6/xfrm6_state.c
index 6269584e610e..40fcaab7e028 100644
--- a/net/ipv6/xfrm6_state.c
+++ b/net/ipv6/xfrm6_state.c
@@ -101,61 +101,6 @@ __xfrm6_state_lookup(xfrm_address_t *daddr, u32 spi, u8 proto)
 	return NULL;
 }
 
-static struct xfrm_state *
-__xfrm6_find_acq(u8 mode, u32 reqid, u8 proto, 
-		 xfrm_address_t *daddr, xfrm_address_t *saddr, 
-		 int create)
-{
-	struct xfrm_state *x, *x0;
-	unsigned h = __xfrm6_dst_hash(daddr);
-
-	x0 = NULL;
-
-	list_for_each_entry(x, xfrm6_state_afinfo.state_bydst+h, bydst) {
-		if (x->props.family == AF_INET6 &&
-		    ipv6_addr_equal((struct in6_addr *)daddr, (struct in6_addr *)x->id.daddr.a6) &&
-		    mode == x->props.mode &&
-		    proto == x->id.proto &&
-		    ipv6_addr_equal((struct in6_addr *)saddr, (struct in6_addr *)x->props.saddr.a6) &&
-		    reqid == x->props.reqid &&
-		    x->km.state == XFRM_STATE_ACQ &&
-		    !x->id.spi) {
-			    x0 = x;
-			    break;
-		    }
-	}
-	if (!x0 && create && (x0 = xfrm_state_alloc()) != NULL) {
-		ipv6_addr_copy((struct in6_addr *)x0->sel.daddr.a6,
-			       (struct in6_addr *)daddr);
-		ipv6_addr_copy((struct in6_addr *)x0->sel.saddr.a6,
-			       (struct in6_addr *)saddr);
-		x0->sel.prefixlen_d = 128;
-		x0->sel.prefixlen_s = 128;
-		ipv6_addr_copy((struct in6_addr *)x0->props.saddr.a6,
-			       (struct in6_addr *)saddr);
-		x0->km.state = XFRM_STATE_ACQ;
-		ipv6_addr_copy((struct in6_addr *)x0->id.daddr.a6,
-			       (struct in6_addr *)daddr);
-		x0->id.proto = proto;
-		x0->props.family = AF_INET6;
-		x0->props.mode = mode;
-		x0->props.reqid = reqid;
-		x0->lft.hard_add_expires_seconds = XFRM_ACQ_EXPIRES;
-		xfrm_state_hold(x0);
-		x0->timer.expires = jiffies + XFRM_ACQ_EXPIRES*HZ;
-		add_timer(&x0->timer);
-		xfrm_state_hold(x0);
-		list_add_tail(&x0->bydst, xfrm6_state_afinfo.state_bydst+h);
-		h = __xfrm6_src_hash(saddr);
-		xfrm_state_hold(x0);
-		list_add_tail(&x0->bysrc, xfrm6_state_afinfo.state_bysrc+h);
-		wake_up(&km_waitq);
-	}
-	if (x0)
-		xfrm_state_hold(x0);
-	return x0;
-}
-
 static int
 __xfrm6_state_sort(struct xfrm_state **dst, struct xfrm_state **src, int n)
 {
@@ -280,7 +225,6 @@ static struct xfrm_state_afinfo xfrm6_state_afinfo = {
 	.init_tempsel		= __xfrm6_init_tempsel,
 	.state_lookup		= __xfrm6_state_lookup,
 	.state_lookup_byaddr	= __xfrm6_state_lookup_byaddr,
-	.find_acq		= __xfrm6_find_acq,
 	.tmpl_sort		= __xfrm6_tmpl_sort,
 	.state_sort		= __xfrm6_state_sort,
 };
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 622e92a08d0b..80f5f9dc2b9e 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -48,6 +48,18 @@ static struct list_head xfrm_state_bydst[XFRM_DST_HSIZE];
 static struct list_head xfrm_state_bysrc[XFRM_DST_HSIZE];
 static struct list_head xfrm_state_byspi[XFRM_DST_HSIZE];
 
+static __inline__
+unsigned xfrm_dst_hash(xfrm_address_t *addr, unsigned short family)
+{
+	switch (family) {
+	case AF_INET:
+		return __xfrm4_dst_hash(addr);
+	case AF_INET6:
+		return __xfrm6_dst_hash(addr);
+	}
+	return 0;
+}
+
 DECLARE_WAIT_QUEUE_HEAD(km_waitq);
 EXPORT_SYMBOL(km_waitq);
 
@@ -489,6 +501,89 @@ void xfrm_state_insert(struct xfrm_state *x)
 }
 EXPORT_SYMBOL(xfrm_state_insert);
 
+/* xfrm_state_lock is held */
+static struct xfrm_state *__find_acq_core(unsigned short family, u8 mode, u32 reqid, u8 proto, xfrm_address_t *daddr, xfrm_address_t *saddr, int create)
+{
+	unsigned int h = xfrm_dst_hash(daddr, family);
+	struct xfrm_state *x;
+
+	list_for_each_entry(x, xfrm_state_bydst+h, bydst) {
+		if (x->props.reqid  != reqid ||
+		    x->props.mode   != mode ||
+		    x->props.family != family ||
+		    x->km.state     != XFRM_STATE_ACQ ||
+		    x->id.spi       != 0)
+			continue;
+
+		switch (family) {
+		case AF_INET:
+			if (x->id.daddr.a4    != daddr->a4 ||
+			    x->props.saddr.a4 != saddr->a4)
+				continue;
+			break;
+		case AF_INET6:
+			if (!ipv6_addr_equal((struct in6_addr *)x->id.daddr.a6,
+					     (struct in6_addr *)daddr) ||
+			    !ipv6_addr_equal((struct in6_addr *)
+					     x->props.saddr.a6,
+					     (struct in6_addr *)saddr))
+				continue;
+			break;
+		};
+
+		xfrm_state_hold(x);
+		return x;
+	}
+
+	if (!create)
+		return NULL;
+
+	x = xfrm_state_alloc();
+	if (likely(x)) {
+		switch (family) {
+		case AF_INET:
+			x->sel.daddr.a4 = daddr->a4;
+			x->sel.saddr.a4 = saddr->a4;
+			x->sel.prefixlen_d = 32;
+			x->sel.prefixlen_s = 32;
+			x->props.saddr.a4 = saddr->a4;
+			x->id.daddr.a4 = daddr->a4;
+			break;
+
+		case AF_INET6:
+			ipv6_addr_copy((struct in6_addr *)x->sel.daddr.a6,
+				       (struct in6_addr *)daddr);
+			ipv6_addr_copy((struct in6_addr *)x->sel.saddr.a6,
+				       (struct in6_addr *)saddr);
+			x->sel.prefixlen_d = 128;
+			x->sel.prefixlen_s = 128;
+			ipv6_addr_copy((struct in6_addr *)x->props.saddr.a6,
+				       (struct in6_addr *)saddr);
+			ipv6_addr_copy((struct in6_addr *)x->id.daddr.a6,
+				       (struct in6_addr *)daddr);
+			break;
+		};
+
+		x->km.state = XFRM_STATE_ACQ;
+		x->id.proto = proto;
+		x->props.family = family;
+		x->props.mode = mode;
+		x->props.reqid = reqid;
+		x->lft.hard_add_expires_seconds = XFRM_ACQ_EXPIRES;
+		xfrm_state_hold(x);
+		x->timer.expires = jiffies + XFRM_ACQ_EXPIRES*HZ;
+		add_timer(&x->timer);
+		xfrm_state_hold(x);
+		list_add_tail(&x->bydst, xfrm_state_bydst+h);
+		h = xfrm_src_hash(saddr, family);
+		xfrm_state_hold(x);
+		list_add_tail(&x->bysrc, xfrm_state_bysrc+h);
+		wake_up(&km_waitq);
+	}
+
+	return x;
+}
+
 static inline struct xfrm_state *
 __xfrm_state_locate(struct xfrm_state_afinfo *afinfo, struct xfrm_state *x,
 		    int use_spi)
@@ -533,9 +628,9 @@ int xfrm_state_add(struct xfrm_state *x)
 	}
 
 	if (use_spi && !x1)
-		x1 = afinfo->find_acq(
-			x->props.mode, x->props.reqid, x->id.proto,
-			&x->id.daddr, &x->props.saddr, 0);
+		x1 = __find_acq_core(family, x->props.mode, x->props.reqid,
+				     x->id.proto,
+				     &x->id.daddr, &x->props.saddr, 0);
 
 	__xfrm_state_insert(x);
 	err = 0;
@@ -716,14 +811,11 @@ xfrm_find_acq(u8 mode, u32 reqid, u8 proto,
 	      int create, unsigned short family)
 {
 	struct xfrm_state *x;
-	struct xfrm_state_afinfo *afinfo = xfrm_state_get_afinfo(family);
-	if (!afinfo)
-		return NULL;
 
 	spin_lock_bh(&xfrm_state_lock);
-	x = afinfo->find_acq(mode, reqid, proto, daddr, saddr, create);
+	x = __find_acq_core(family, mode, reqid, proto, daddr, saddr, create);
 	spin_unlock_bh(&xfrm_state_lock);
-	xfrm_state_put_afinfo(afinfo);
+
 	return x;
 }
 EXPORT_SYMBOL(xfrm_find_acq);
@@ -1181,7 +1273,6 @@ int xfrm_state_register_afinfo(struct xfrm_state_afinfo *afinfo)
 	if (unlikely(xfrm_state_afinfo[afinfo->family] != NULL))
 		err = -ENOBUFS;
 	else {
-		afinfo->state_bydst = xfrm_state_bydst;
 		afinfo->state_bysrc = xfrm_state_bysrc;
 		afinfo->state_byspi = xfrm_state_byspi;
 		xfrm_state_afinfo[afinfo->family] = afinfo;
@@ -1206,7 +1297,6 @@ int xfrm_state_unregister_afinfo(struct xfrm_state_afinfo *afinfo)
 			xfrm_state_afinfo[afinfo->family] = NULL;
 			afinfo->state_byspi = NULL;
 			afinfo->state_bysrc = NULL;
-			afinfo->state_bydst = NULL;
 		}
 	}
 	write_unlock_bh(&xfrm_state_afinfo_lock);
-- 
cgit v1.2.3


From edcd582152090bfb0ccb4ad444c151798a73eda8 Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@sunset.davemloft.net>
Date: Thu, 24 Aug 2006 00:42:45 -0700
Subject: [XFRM]: Pull xfrm_state_by{spi,src} hash table knowledge out of
 afinfo.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h     |  78 ------------------
 net/ipv4/xfrm4_state.c |  28 -------
 net/ipv6/xfrm6_state.c |  40 ----------
 net/xfrm/xfrm_state.c  | 210 +++++++++++++++++++++++++++++++++++++------------
 4 files changed, 159 insertions(+), 197 deletions(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index cc83443f301e..dd3b84b9c04e 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -243,14 +243,10 @@ extern int __xfrm_state_delete(struct xfrm_state *x);
 
 struct xfrm_state_afinfo {
 	unsigned short		family;
-	struct list_head	*state_bysrc;
-	struct list_head	*state_byspi;
 	int			(*init_flags)(struct xfrm_state *x);
 	void			(*init_tempsel)(struct xfrm_state *x, struct flowi *fl,
 						struct xfrm_tmpl *tmpl,
 						xfrm_address_t *daddr, xfrm_address_t *saddr);
-	struct xfrm_state	*(*state_lookup)(xfrm_address_t *daddr, u32 spi, u8 proto);
-	struct xfrm_state	*(*state_lookup_byaddr)(xfrm_address_t *daddr, xfrm_address_t *saddr, u8 proto);
 	int			(*tmpl_sort)(struct xfrm_tmpl **dst, struct xfrm_tmpl **src, int n);
 	int			(*state_sort)(struct xfrm_state **dst, struct xfrm_state **src, int n);
 };
@@ -431,80 +427,6 @@ static inline void xfrm_pols_put(struct xfrm_policy **pols, int npols)
 }
 #endif
 
-#define XFRM_DST_HSIZE		1024
-
-static __inline__
-unsigned __xfrm4_dst_hash(xfrm_address_t *addr)
-{
-	unsigned h;
-	h = ntohl(addr->a4);
-	h = (h ^ (h>>16)) % XFRM_DST_HSIZE;
-	return h;
-}
-
-static __inline__
-unsigned __xfrm6_dst_hash(xfrm_address_t *addr)
-{
-	unsigned h;
-	h = ntohl(addr->a6[2]^addr->a6[3]);
-	h = (h ^ (h>>16)) % XFRM_DST_HSIZE;
-	return h;
-}
-
-static __inline__
-unsigned __xfrm4_src_hash(xfrm_address_t *addr)
-{
-	return __xfrm4_dst_hash(addr);
-}
-
-static __inline__
-unsigned __xfrm6_src_hash(xfrm_address_t *addr)
-{
-	return __xfrm6_dst_hash(addr);
-}
-
-static __inline__
-unsigned xfrm_src_hash(xfrm_address_t *addr, unsigned short family)
-{
-	switch (family) {
-	case AF_INET:
-		return __xfrm4_src_hash(addr);
-	case AF_INET6:
-		return __xfrm6_src_hash(addr);
-	}
-	return 0;
-}
-
-static __inline__
-unsigned __xfrm4_spi_hash(xfrm_address_t *addr, u32 spi, u8 proto)
-{
-	unsigned h;
-	h = ntohl(addr->a4^spi^proto);
-	h = (h ^ (h>>10) ^ (h>>20)) % XFRM_DST_HSIZE;
-	return h;
-}
-
-static __inline__
-unsigned __xfrm6_spi_hash(xfrm_address_t *addr, u32 spi, u8 proto)
-{
-	unsigned h;
-	h = ntohl(addr->a6[2]^addr->a6[3]^spi^proto);
-	h = (h ^ (h>>10) ^ (h>>20)) % XFRM_DST_HSIZE;
-	return h;
-}
-
-static __inline__
-unsigned xfrm_spi_hash(xfrm_address_t *addr, u32 spi, u8 proto, unsigned short family)
-{
-	switch (family) {
-	case AF_INET:
-		return __xfrm4_spi_hash(addr, spi, proto);
-	case AF_INET6:
-		return __xfrm6_spi_hash(addr, spi, proto);
-	}
-	return 0;	/*XXX*/
-}
-
 extern void __xfrm_state_destroy(struct xfrm_state *);
 
 static inline void __xfrm_state_put(struct xfrm_state *x)
diff --git a/net/ipv4/xfrm4_state.c b/net/ipv4/xfrm4_state.c
index 9dc1afc17b6d..6a2a4ab42772 100644
--- a/net/ipv4/xfrm4_state.c
+++ b/net/ipv4/xfrm4_state.c
@@ -62,38 +62,10 @@ __xfrm4_init_tempsel(struct xfrm_state *x, struct flowi *fl,
 	x->props.family = AF_INET;
 }
 
-static struct xfrm_state *
-__xfrm4_state_lookup(xfrm_address_t *daddr, u32 spi, u8 proto)
-{
-	unsigned h = __xfrm4_spi_hash(daddr, spi, proto);
-	struct xfrm_state *x;
-
-	list_for_each_entry(x, xfrm4_state_afinfo.state_byspi+h, byspi) {
-		if (x->props.family == AF_INET &&
-		    spi == x->id.spi &&
-		    daddr->a4 == x->id.daddr.a4 &&
-		    proto == x->id.proto) {
-			xfrm_state_hold(x);
-			return x;
-		}
-	}
-	return NULL;
-}
-
-/* placeholder until ipv4's code is written */
-static struct xfrm_state *
-__xfrm4_state_lookup_byaddr(xfrm_address_t *daddr, xfrm_address_t *saddr,
-			    u8 proto)
-{
-	return NULL;
-}
-
 static struct xfrm_state_afinfo xfrm4_state_afinfo = {
 	.family			= AF_INET,
 	.init_flags		= xfrm4_init_flags,
 	.init_tempsel		= __xfrm4_init_tempsel,
-	.state_lookup		= __xfrm4_state_lookup,
-	.state_lookup_byaddr	= __xfrm4_state_lookup_byaddr,
 };
 
 void __init xfrm4_state_init(void)
diff --git a/net/ipv6/xfrm6_state.c b/net/ipv6/xfrm6_state.c
index 40fcaab7e028..d88cd92c864e 100644
--- a/net/ipv6/xfrm6_state.c
+++ b/net/ipv6/xfrm6_state.c
@@ -63,44 +63,6 @@ __xfrm6_init_tempsel(struct xfrm_state *x, struct flowi *fl,
 	x->props.family = AF_INET6;
 }
 
-static struct xfrm_state *
-__xfrm6_state_lookup_byaddr(xfrm_address_t *daddr, xfrm_address_t *saddr,
-			    u8 proto)
-{
-	struct xfrm_state *x = NULL;
-	unsigned h;
-
-	h = __xfrm6_src_hash(saddr);
-	list_for_each_entry(x, xfrm6_state_afinfo.state_bysrc+h, bysrc) {
-		if (x->props.family == AF_INET6 &&
-		    ipv6_addr_equal((struct in6_addr *)daddr, (struct in6_addr *)x->id.daddr.a6) &&
-		    ipv6_addr_equal((struct in6_addr *)saddr, (struct in6_addr *)x->props.saddr.a6) &&
-		    proto == x->id.proto) {
-			xfrm_state_hold(x);
-			return x;
-		}
-	}
-	return NULL;
-}
-
-static struct xfrm_state *
-__xfrm6_state_lookup(xfrm_address_t *daddr, u32 spi, u8 proto)
-{
-	unsigned h = __xfrm6_spi_hash(daddr, spi, proto);
-	struct xfrm_state *x;
-
-	list_for_each_entry(x, xfrm6_state_afinfo.state_byspi+h, byspi) {
-		if (x->props.family == AF_INET6 &&
-		    spi == x->id.spi &&
-		    ipv6_addr_equal((struct in6_addr *)daddr, (struct in6_addr *)x->id.daddr.a6) &&
-		    proto == x->id.proto) {
-			xfrm_state_hold(x);
-			return x;
-		}
-	}
-	return NULL;
-}
-
 static int
 __xfrm6_state_sort(struct xfrm_state **dst, struct xfrm_state **src, int n)
 {
@@ -223,8 +185,6 @@ __xfrm6_tmpl_sort(struct xfrm_tmpl **dst, struct xfrm_tmpl **src, int n)
 static struct xfrm_state_afinfo xfrm6_state_afinfo = {
 	.family			= AF_INET6,
 	.init_tempsel		= __xfrm6_init_tempsel,
-	.state_lookup		= __xfrm6_state_lookup,
-	.state_lookup_byaddr	= __xfrm6_state_lookup_byaddr,
 	.tmpl_sort		= __xfrm6_tmpl_sort,
 	.state_sort		= __xfrm6_state_sort,
 };
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 80f5f9dc2b9e..4a3832f81c37 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -38,6 +38,8 @@ EXPORT_SYMBOL(sysctl_xfrm_aevent_rseqth);
 
 static DEFINE_SPINLOCK(xfrm_state_lock);
 
+#define XFRM_DST_HSIZE		1024
+
 /* Hash table to find appropriate SA towards given target (endpoint
  * of tunnel or destination of transport mode) allowed by selector.
  *
@@ -48,6 +50,48 @@ static struct list_head xfrm_state_bydst[XFRM_DST_HSIZE];
 static struct list_head xfrm_state_bysrc[XFRM_DST_HSIZE];
 static struct list_head xfrm_state_byspi[XFRM_DST_HSIZE];
 
+static __inline__
+unsigned __xfrm4_dst_hash(xfrm_address_t *addr)
+{
+	unsigned h;
+	h = ntohl(addr->a4);
+	h = (h ^ (h>>16)) % XFRM_DST_HSIZE;
+	return h;
+}
+
+static __inline__
+unsigned __xfrm6_dst_hash(xfrm_address_t *addr)
+{
+	unsigned h;
+	h = ntohl(addr->a6[2]^addr->a6[3]);
+	h = (h ^ (h>>16)) % XFRM_DST_HSIZE;
+	return h;
+}
+
+static __inline__
+unsigned __xfrm4_src_hash(xfrm_address_t *addr)
+{
+	return __xfrm4_dst_hash(addr);
+}
+
+static __inline__
+unsigned __xfrm6_src_hash(xfrm_address_t *addr)
+{
+	return __xfrm6_dst_hash(addr);
+}
+
+static __inline__
+unsigned xfrm_src_hash(xfrm_address_t *addr, unsigned short family)
+{
+	switch (family) {
+	case AF_INET:
+		return __xfrm4_src_hash(addr);
+	case AF_INET6:
+		return __xfrm6_src_hash(addr);
+	}
+	return 0;
+}
+
 static __inline__
 unsigned xfrm_dst_hash(xfrm_address_t *addr, unsigned short family)
 {
@@ -60,6 +104,36 @@ unsigned xfrm_dst_hash(xfrm_address_t *addr, unsigned short family)
 	return 0;
 }
 
+static __inline__
+unsigned __xfrm4_spi_hash(xfrm_address_t *addr, u32 spi, u8 proto)
+{
+	unsigned h;
+	h = ntohl(addr->a4^spi^proto);
+	h = (h ^ (h>>10) ^ (h>>20)) % XFRM_DST_HSIZE;
+	return h;
+}
+
+static __inline__
+unsigned __xfrm6_spi_hash(xfrm_address_t *addr, u32 spi, u8 proto)
+{
+	unsigned h;
+	h = ntohl(addr->a6[2]^addr->a6[3]^spi^proto);
+	h = (h ^ (h>>10) ^ (h>>20)) % XFRM_DST_HSIZE;
+	return h;
+}
+
+static __inline__
+unsigned xfrm_spi_hash(xfrm_address_t *addr, u32 spi, u8 proto, unsigned short family)
+{
+	switch (family) {
+	case AF_INET:
+		return __xfrm4_spi_hash(addr, spi, proto);
+	case AF_INET6:
+		return __xfrm6_spi_hash(addr, spi, proto);
+	}
+	return 0;	/*XXX*/
+}
+
 DECLARE_WAIT_QUEUE_HEAD(km_waitq);
 EXPORT_SYMBOL(km_waitq);
 
@@ -342,6 +416,83 @@ xfrm_init_tempsel(struct xfrm_state *x, struct flowi *fl,
 	return 0;
 }
 
+static struct xfrm_state *__xfrm_state_lookup(xfrm_address_t *daddr, u32 spi, u8 proto, unsigned short family)
+{
+	unsigned int h = xfrm_spi_hash(daddr, spi, proto, family);
+	struct xfrm_state *x;
+
+	list_for_each_entry(x, xfrm_state_byspi+h, byspi) {
+		if (x->props.family != family ||
+		    x->id.spi       != spi ||
+		    x->id.proto     != proto)
+			continue;
+
+		switch (family) {
+		case AF_INET:
+			if (x->id.daddr.a4 != daddr->a4)
+				continue;
+			break;
+		case AF_INET6:
+			if (!ipv6_addr_equal((struct in6_addr *)daddr,
+					     (struct in6_addr *)
+					     x->id.daddr.a6))
+				continue;
+			break;
+		};
+
+		xfrm_state_hold(x);
+		return x;
+	}
+
+	return NULL;
+}
+
+static struct xfrm_state *__xfrm_state_lookup_byaddr(xfrm_address_t *daddr, xfrm_address_t *saddr, u8 proto, unsigned short family)
+{
+	unsigned int h = xfrm_src_hash(saddr, family);
+	struct xfrm_state *x;
+
+	list_for_each_entry(x, xfrm_state_bysrc+h, bysrc) {
+		if (x->props.family != family ||
+		    x->id.proto     != proto)
+			continue;
+
+		switch (family) {
+		case AF_INET:
+			if (x->id.daddr.a4 != daddr->a4 ||
+			    x->props.saddr.a4 != saddr->a4)
+				continue;
+			break;
+		case AF_INET6:
+			if (!ipv6_addr_equal((struct in6_addr *)daddr,
+					     (struct in6_addr *)
+					     x->id.daddr.a6) ||
+			    !ipv6_addr_equal((struct in6_addr *)saddr,
+					     (struct in6_addr *)
+					     x->props.saddr.a6))
+				continue;
+			break;
+		};
+
+		xfrm_state_hold(x);
+		return x;
+	}
+
+	return NULL;
+}
+
+static inline struct xfrm_state *
+__xfrm_state_locate(struct xfrm_state *x, int use_spi, int family)
+{
+	if (use_spi)
+		return __xfrm_state_lookup(&x->id.daddr, x->id.spi,
+					   x->id.proto, family);
+	else
+		return __xfrm_state_lookup_byaddr(&x->id.daddr,
+						  &x->props.saddr,
+						  x->id.proto, family);
+}
+
 struct xfrm_state *
 xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr, 
 		struct flowi *fl, struct xfrm_tmpl *tmpl,
@@ -353,14 +504,7 @@ xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr,
 	int acquire_in_progress = 0;
 	int error = 0;
 	struct xfrm_state *best = NULL;
-	struct xfrm_state_afinfo *afinfo;
 	
-	afinfo = xfrm_state_get_afinfo(family);
-	if (afinfo == NULL) {
-		*err = -EAFNOSUPPORT;
-		return NULL;
-	}
-
 	spin_lock_bh(&xfrm_state_lock);
 	list_for_each_entry(x, xfrm_state_bydst+h, bydst) {
 		if (x->props.family == family &&
@@ -406,8 +550,8 @@ xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr,
 	x = best;
 	if (!x && !error && !acquire_in_progress) {
 		if (tmpl->id.spi &&
-		    (x0 = afinfo->state_lookup(daddr, tmpl->id.spi,
-		                               tmpl->id.proto)) != NULL) {
+		    (x0 = __xfrm_state_lookup(daddr, tmpl->id.spi,
+					      tmpl->id.proto, family)) != NULL) {
 			xfrm_state_put(x0);
 			error = -EEXIST;
 			goto out;
@@ -457,7 +601,6 @@ out:
 	else
 		*err = acquire_in_progress ? -EAGAIN : error;
 	spin_unlock_bh(&xfrm_state_lock);
-	xfrm_state_put_afinfo(afinfo);
 	return x;
 }
 
@@ -584,34 +727,20 @@ static struct xfrm_state *__find_acq_core(unsigned short family, u8 mode, u32 re
 	return x;
 }
 
-static inline struct xfrm_state *
-__xfrm_state_locate(struct xfrm_state_afinfo *afinfo, struct xfrm_state *x,
-		    int use_spi)
-{
-	if (use_spi)
-		return afinfo->state_lookup(&x->id.daddr, x->id.spi, x->id.proto);
-	else
-		return afinfo->state_lookup_byaddr(&x->id.daddr, &x->props.saddr, x->id.proto);
-}
-
 static struct xfrm_state *__xfrm_find_acq_byseq(u32 seq);
 
 int xfrm_state_add(struct xfrm_state *x)
 {
-	struct xfrm_state_afinfo *afinfo;
 	struct xfrm_state *x1;
 	int family;
 	int err;
 	int use_spi = xfrm_id_proto_match(x->id.proto, IPSEC_PROTO_ANY);
 
 	family = x->props.family;
-	afinfo = xfrm_state_get_afinfo(family);
-	if (unlikely(afinfo == NULL))
-		return -EAFNOSUPPORT;
 
 	spin_lock_bh(&xfrm_state_lock);
 
-	x1 = __xfrm_state_locate(afinfo, x, use_spi);
+	x1 = __xfrm_state_locate(x, use_spi, family);
 	if (x1) {
 		xfrm_state_put(x1);
 		x1 = NULL;
@@ -637,7 +766,6 @@ int xfrm_state_add(struct xfrm_state *x)
 
 out:
 	spin_unlock_bh(&xfrm_state_lock);
-	xfrm_state_put_afinfo(afinfo);
 
 	if (!err)
 		xfrm_flush_all_bundles();
@@ -653,17 +781,12 @@ EXPORT_SYMBOL(xfrm_state_add);
 
 int xfrm_state_update(struct xfrm_state *x)
 {
-	struct xfrm_state_afinfo *afinfo;
 	struct xfrm_state *x1;
 	int err;
 	int use_spi = xfrm_id_proto_match(x->id.proto, IPSEC_PROTO_ANY);
 
-	afinfo = xfrm_state_get_afinfo(x->props.family);
-	if (unlikely(afinfo == NULL))
-		return -EAFNOSUPPORT;
-
 	spin_lock_bh(&xfrm_state_lock);
-	x1 = __xfrm_state_locate(afinfo, x, use_spi);
+	x1 = __xfrm_state_locate(x, use_spi, x->props.family);
 
 	err = -ESRCH;
 	if (!x1)
@@ -683,7 +806,6 @@ int xfrm_state_update(struct xfrm_state *x)
 
 out:
 	spin_unlock_bh(&xfrm_state_lock);
-	xfrm_state_put_afinfo(afinfo);
 
 	if (err)
 		return err;
@@ -776,14 +898,10 @@ xfrm_state_lookup(xfrm_address_t *daddr, u32 spi, u8 proto,
 		  unsigned short family)
 {
 	struct xfrm_state *x;
-	struct xfrm_state_afinfo *afinfo = xfrm_state_get_afinfo(family);
-	if (!afinfo)
-		return NULL;
 
 	spin_lock_bh(&xfrm_state_lock);
-	x = afinfo->state_lookup(daddr, spi, proto);
+	x = __xfrm_state_lookup(daddr, spi, proto, family);
 	spin_unlock_bh(&xfrm_state_lock);
-	xfrm_state_put_afinfo(afinfo);
 	return x;
 }
 EXPORT_SYMBOL(xfrm_state_lookup);
@@ -793,14 +911,10 @@ xfrm_state_lookup_byaddr(xfrm_address_t *daddr, xfrm_address_t *saddr,
 			 u8 proto, unsigned short family)
 {
 	struct xfrm_state *x;
-	struct xfrm_state_afinfo *afinfo = xfrm_state_get_afinfo(family);
-	if (!afinfo)
-		return NULL;
 
 	spin_lock_bh(&xfrm_state_lock);
-	x = afinfo->state_lookup_byaddr(daddr, saddr, proto);
+	x = __xfrm_state_lookup_byaddr(daddr, saddr, proto, family);
 	spin_unlock_bh(&xfrm_state_lock);
-	xfrm_state_put_afinfo(afinfo);
 	return x;
 }
 EXPORT_SYMBOL(xfrm_state_lookup_byaddr);
@@ -1272,11 +1386,8 @@ int xfrm_state_register_afinfo(struct xfrm_state_afinfo *afinfo)
 	write_lock_bh(&xfrm_state_afinfo_lock);
 	if (unlikely(xfrm_state_afinfo[afinfo->family] != NULL))
 		err = -ENOBUFS;
-	else {
-		afinfo->state_bysrc = xfrm_state_bysrc;
-		afinfo->state_byspi = xfrm_state_byspi;
+	else
 		xfrm_state_afinfo[afinfo->family] = afinfo;
-	}
 	write_unlock_bh(&xfrm_state_afinfo_lock);
 	return err;
 }
@@ -1293,11 +1404,8 @@ int xfrm_state_unregister_afinfo(struct xfrm_state_afinfo *afinfo)
 	if (likely(xfrm_state_afinfo[afinfo->family] != NULL)) {
 		if (unlikely(xfrm_state_afinfo[afinfo->family] != afinfo))
 			err = -EINVAL;
-		else {
+		else
 			xfrm_state_afinfo[afinfo->family] = NULL;
-			afinfo->state_byspi = NULL;
-			afinfo->state_bysrc = NULL;
-		}
 	}
 	write_unlock_bh(&xfrm_state_afinfo_lock);
 	return err;
-- 
cgit v1.2.3


From 8f126e37c0b250310a48a609bedf92a19a5559ec Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@sunset.davemloft.net>
Date: Thu, 24 Aug 2006 02:45:07 -0700
Subject: [XFRM]: Convert xfrm_state hash linkage to hlists.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h    |  6 ++--
 net/xfrm/xfrm_state.c | 92 ++++++++++++++++++++++++++++-----------------------
 2 files changed, 54 insertions(+), 44 deletions(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index dd3b84b9c04e..3405e5d9d51c 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -94,9 +94,9 @@ extern struct mutex xfrm_cfg_mutex;
 struct xfrm_state
 {
 	/* Note: bydst is re-used during gc */
-	struct list_head	bydst;
-	struct list_head	bysrc;
-	struct list_head	byspi;
+	struct hlist_node	bydst;
+	struct hlist_node	bysrc;
+	struct hlist_node	byspi;
 
 	atomic_t		refcnt;
 	spinlock_t		lock;
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 4a3832f81c37..fe3c8c38d5e1 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -46,9 +46,9 @@ static DEFINE_SPINLOCK(xfrm_state_lock);
  * Main use is finding SA after policy selected tunnel or transport mode.
  * Also, it can be used by ah/esp icmp error handler to find offending SA.
  */
-static struct list_head xfrm_state_bydst[XFRM_DST_HSIZE];
-static struct list_head xfrm_state_bysrc[XFRM_DST_HSIZE];
-static struct list_head xfrm_state_byspi[XFRM_DST_HSIZE];
+static struct hlist_head xfrm_state_bydst[XFRM_DST_HSIZE];
+static struct hlist_head xfrm_state_bysrc[XFRM_DST_HSIZE];
+static struct hlist_head xfrm_state_byspi[XFRM_DST_HSIZE];
 
 static __inline__
 unsigned __xfrm4_dst_hash(xfrm_address_t *addr)
@@ -141,7 +141,7 @@ static DEFINE_RWLOCK(xfrm_state_afinfo_lock);
 static struct xfrm_state_afinfo *xfrm_state_afinfo[NPROTO];
 
 static struct work_struct xfrm_state_gc_work;
-static struct list_head xfrm_state_gc_list = LIST_HEAD_INIT(xfrm_state_gc_list);
+static HLIST_HEAD(xfrm_state_gc_list);
 static DEFINE_SPINLOCK(xfrm_state_gc_lock);
 
 static int xfrm_state_gc_flush_bundles;
@@ -178,8 +178,8 @@ static void xfrm_state_gc_destroy(struct xfrm_state *x)
 static void xfrm_state_gc_task(void *data)
 {
 	struct xfrm_state *x;
-	struct list_head *entry, *tmp;
-	struct list_head gc_list = LIST_HEAD_INIT(gc_list);
+	struct hlist_node *entry, *tmp;
+	struct hlist_head gc_list;
 
 	if (xfrm_state_gc_flush_bundles) {
 		xfrm_state_gc_flush_bundles = 0;
@@ -187,13 +187,13 @@ static void xfrm_state_gc_task(void *data)
 	}
 
 	spin_lock_bh(&xfrm_state_gc_lock);
-	list_splice_init(&xfrm_state_gc_list, &gc_list);
+	gc_list.first = xfrm_state_gc_list.first;
+	INIT_HLIST_HEAD(&xfrm_state_gc_list);
 	spin_unlock_bh(&xfrm_state_gc_lock);
 
-	list_for_each_safe(entry, tmp, &gc_list) {
-		x = list_entry(entry, struct xfrm_state, bydst);
+	hlist_for_each_entry_safe(x, entry, tmp, &gc_list, bydst)
 		xfrm_state_gc_destroy(x);
-	}
+
 	wake_up(&km_waitq);
 }
 
@@ -287,9 +287,9 @@ struct xfrm_state *xfrm_state_alloc(void)
 	if (x) {
 		atomic_set(&x->refcnt, 1);
 		atomic_set(&x->tunnel_users, 0);
-		INIT_LIST_HEAD(&x->bydst);
-		INIT_LIST_HEAD(&x->bysrc);
-		INIT_LIST_HEAD(&x->byspi);
+		INIT_HLIST_NODE(&x->bydst);
+		INIT_HLIST_NODE(&x->bysrc);
+		INIT_HLIST_NODE(&x->byspi);
 		init_timer(&x->timer);
 		x->timer.function = xfrm_timer_handler;
 		x->timer.data	  = (unsigned long)x;
@@ -314,7 +314,7 @@ void __xfrm_state_destroy(struct xfrm_state *x)
 	BUG_TRAP(x->km.state == XFRM_STATE_DEAD);
 
 	spin_lock_bh(&xfrm_state_gc_lock);
-	list_add(&x->bydst, &xfrm_state_gc_list);
+	hlist_add_head(&x->bydst, &xfrm_state_gc_list);
 	spin_unlock_bh(&xfrm_state_gc_lock);
 	schedule_work(&xfrm_state_gc_work);
 }
@@ -327,12 +327,12 @@ int __xfrm_state_delete(struct xfrm_state *x)
 	if (x->km.state != XFRM_STATE_DEAD) {
 		x->km.state = XFRM_STATE_DEAD;
 		spin_lock(&xfrm_state_lock);
-		list_del(&x->bydst);
+		hlist_del(&x->bydst);
 		__xfrm_state_put(x);
-		list_del(&x->bysrc);
+		hlist_del(&x->bysrc);
 		__xfrm_state_put(x);
 		if (x->id.spi) {
-			list_del(&x->byspi);
+			hlist_del(&x->byspi);
 			__xfrm_state_put(x);
 		}
 		spin_unlock(&xfrm_state_lock);
@@ -378,12 +378,13 @@ EXPORT_SYMBOL(xfrm_state_delete);
 void xfrm_state_flush(u8 proto)
 {
 	int i;
-	struct xfrm_state *x;
 
 	spin_lock_bh(&xfrm_state_lock);
 	for (i = 0; i < XFRM_DST_HSIZE; i++) {
+		struct hlist_node *entry;
+		struct xfrm_state *x;
 restart:
-		list_for_each_entry(x, xfrm_state_bydst+i, bydst) {
+		hlist_for_each_entry(x, entry, xfrm_state_bydst+i, bydst) {
 			if (!xfrm_state_kern(x) &&
 			    xfrm_id_proto_match(x->id.proto, proto)) {
 				xfrm_state_hold(x);
@@ -420,8 +421,9 @@ static struct xfrm_state *__xfrm_state_lookup(xfrm_address_t *daddr, u32 spi, u8
 {
 	unsigned int h = xfrm_spi_hash(daddr, spi, proto, family);
 	struct xfrm_state *x;
+	struct hlist_node *entry;
 
-	list_for_each_entry(x, xfrm_state_byspi+h, byspi) {
+	hlist_for_each_entry(x, entry, xfrm_state_byspi+h, byspi) {
 		if (x->props.family != family ||
 		    x->id.spi       != spi ||
 		    x->id.proto     != proto)
@@ -451,8 +453,9 @@ static struct xfrm_state *__xfrm_state_lookup_byaddr(xfrm_address_t *daddr, xfrm
 {
 	unsigned int h = xfrm_src_hash(saddr, family);
 	struct xfrm_state *x;
+	struct hlist_node *entry;
 
-	list_for_each_entry(x, xfrm_state_bysrc+h, bysrc) {
+	hlist_for_each_entry(x, entry, xfrm_state_bysrc+h, bysrc) {
 		if (x->props.family != family ||
 		    x->id.proto     != proto)
 			continue;
@@ -499,14 +502,15 @@ xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr,
 		struct xfrm_policy *pol, int *err,
 		unsigned short family)
 {
-	unsigned h = xfrm_dst_hash(daddr, family);
+	unsigned int h = xfrm_dst_hash(daddr, family);
+	struct hlist_node *entry;
 	struct xfrm_state *x, *x0;
 	int acquire_in_progress = 0;
 	int error = 0;
 	struct xfrm_state *best = NULL;
 	
 	spin_lock_bh(&xfrm_state_lock);
-	list_for_each_entry(x, xfrm_state_bydst+h, bydst) {
+	hlist_for_each_entry(x, entry, xfrm_state_bydst+h, bydst) {
 		if (x->props.family == family &&
 		    x->props.reqid == tmpl->reqid &&
 		    !(x->props.flags & XFRM_STATE_WILDRECV) &&
@@ -575,13 +579,14 @@ xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr,
 
 		if (km_query(x, tmpl, pol) == 0) {
 			x->km.state = XFRM_STATE_ACQ;
-			list_add_tail(&x->bydst, xfrm_state_bydst+h);
+			hlist_add_head(&x->bydst, xfrm_state_bydst+h);
 			xfrm_state_hold(x);
-			list_add_tail(&x->bysrc, xfrm_state_bysrc+h);
+			h = xfrm_src_hash(saddr, family);
+			hlist_add_head(&x->bysrc, xfrm_state_bysrc+h);
 			xfrm_state_hold(x);
 			if (x->id.spi) {
 				h = xfrm_spi_hash(&x->id.daddr, x->id.spi, x->id.proto, family);
-				list_add(&x->byspi, xfrm_state_byspi+h);
+				hlist_add_head(&x->byspi, xfrm_state_byspi+h);
 				xfrm_state_hold(x);
 			}
 			x->lft.hard_add_expires_seconds = XFRM_ACQ_EXPIRES;
@@ -608,19 +613,19 @@ static void __xfrm_state_insert(struct xfrm_state *x)
 {
 	unsigned h = xfrm_dst_hash(&x->id.daddr, x->props.family);
 
-	list_add(&x->bydst, xfrm_state_bydst+h);
+	hlist_add_head(&x->bydst, xfrm_state_bydst+h);
 	xfrm_state_hold(x);
 
 	h = xfrm_src_hash(&x->props.saddr, x->props.family);
 
-	list_add(&x->bysrc, xfrm_state_bysrc+h);
+	hlist_add_head(&x->bysrc, xfrm_state_bysrc+h);
 	xfrm_state_hold(x);
 
 	if (xfrm_id_proto_match(x->id.proto, IPSEC_PROTO_ANY)) {
 		h = xfrm_spi_hash(&x->id.daddr, x->id.spi, x->id.proto,
 				  x->props.family);
 
-		list_add(&x->byspi, xfrm_state_byspi+h);
+		hlist_add_head(&x->byspi, xfrm_state_byspi+h);
 		xfrm_state_hold(x);
 	}
 
@@ -648,9 +653,10 @@ EXPORT_SYMBOL(xfrm_state_insert);
 static struct xfrm_state *__find_acq_core(unsigned short family, u8 mode, u32 reqid, u8 proto, xfrm_address_t *daddr, xfrm_address_t *saddr, int create)
 {
 	unsigned int h = xfrm_dst_hash(daddr, family);
+	struct hlist_node *entry;
 	struct xfrm_state *x;
 
-	list_for_each_entry(x, xfrm_state_bydst+h, bydst) {
+	hlist_for_each_entry(x, entry, xfrm_state_bydst+h, bydst) {
 		if (x->props.reqid  != reqid ||
 		    x->props.mode   != mode ||
 		    x->props.family != family ||
@@ -717,10 +723,10 @@ static struct xfrm_state *__find_acq_core(unsigned short family, u8 mode, u32 re
 		x->timer.expires = jiffies + XFRM_ACQ_EXPIRES*HZ;
 		add_timer(&x->timer);
 		xfrm_state_hold(x);
-		list_add_tail(&x->bydst, xfrm_state_bydst+h);
+		hlist_add_head(&x->bydst, xfrm_state_bydst+h);
 		h = xfrm_src_hash(saddr, family);
 		xfrm_state_hold(x);
-		list_add_tail(&x->bysrc, xfrm_state_bysrc+h);
+		hlist_add_head(&x->bysrc, xfrm_state_bysrc+h);
 		wake_up(&km_waitq);
 	}
 
@@ -977,11 +983,14 @@ EXPORT_SYMBOL(xfrm_state_sort);
 static struct xfrm_state *__xfrm_find_acq_byseq(u32 seq)
 {
 	int i;
-	struct xfrm_state *x;
 
 	for (i = 0; i < XFRM_DST_HSIZE; i++) {
-		list_for_each_entry(x, xfrm_state_bydst+i, bydst) {
-			if (x->km.seq == seq && x->km.state == XFRM_STATE_ACQ) {
+		struct hlist_node *entry;
+		struct xfrm_state *x;
+
+		hlist_for_each_entry(x, entry, xfrm_state_bydst+i, bydst) {
+			if (x->km.seq == seq &&
+			    x->km.state == XFRM_STATE_ACQ) {
 				xfrm_state_hold(x);
 				return x;
 			}
@@ -1047,7 +1056,7 @@ xfrm_alloc_spi(struct xfrm_state *x, u32 minspi, u32 maxspi)
 	if (x->id.spi) {
 		spin_lock_bh(&xfrm_state_lock);
 		h = xfrm_spi_hash(&x->id.daddr, x->id.spi, x->id.proto, x->props.family);
-		list_add(&x->byspi, xfrm_state_byspi+h);
+		hlist_add_head(&x->byspi, xfrm_state_byspi+h);
 		xfrm_state_hold(x);
 		spin_unlock_bh(&xfrm_state_lock);
 		wake_up(&km_waitq);
@@ -1060,12 +1069,13 @@ int xfrm_state_walk(u8 proto, int (*func)(struct xfrm_state *, int, void*),
 {
 	int i;
 	struct xfrm_state *x;
+	struct hlist_node *entry;
 	int count = 0;
 	int err = 0;
 
 	spin_lock_bh(&xfrm_state_lock);
 	for (i = 0; i < XFRM_DST_HSIZE; i++) {
-		list_for_each_entry(x, xfrm_state_bydst+i, bydst) {
+		hlist_for_each_entry(x, entry, xfrm_state_bydst+i, bydst) {
 			if (xfrm_id_proto_match(x->id.proto, proto))
 				count++;
 		}
@@ -1076,7 +1086,7 @@ int xfrm_state_walk(u8 proto, int (*func)(struct xfrm_state *, int, void*),
 	}
 
 	for (i = 0; i < XFRM_DST_HSIZE; i++) {
-		list_for_each_entry(x, xfrm_state_bydst+i, bydst) {
+		hlist_for_each_entry(x, entry, xfrm_state_bydst+i, bydst) {
 			if (!xfrm_id_proto_match(x->id.proto, proto))
 				continue;
 			err = func(x, --count, data);
@@ -1524,9 +1534,9 @@ void __init xfrm_state_init(void)
 	int i;
 
 	for (i=0; i<XFRM_DST_HSIZE; i++) {
-		INIT_LIST_HEAD(&xfrm_state_bydst[i]);
-		INIT_LIST_HEAD(&xfrm_state_bysrc[i]);
-		INIT_LIST_HEAD(&xfrm_state_byspi[i]);
+		INIT_HLIST_HEAD(&xfrm_state_bydst[i]);
+		INIT_HLIST_HEAD(&xfrm_state_bysrc[i]);
+		INIT_HLIST_HEAD(&xfrm_state_byspi[i]);
 	}
 	INIT_WORK(&xfrm_state_gc_work, xfrm_state_gc_task, NULL);
 }
-- 
cgit v1.2.3


From f034b5d4efdfe0fb9e2a1ce1d95fa7914f24de49 Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@sunset.davemloft.net>
Date: Thu, 24 Aug 2006 03:08:07 -0700
Subject: [XFRM]: Dynamic xfrm_state hash table sizing.

The grow algorithm is simple, we grow if:

1) we see a hash chain collision at insert, and
2) we haven't hit the hash size limit (currently 1*1024*1024 slots), and
3) the number of xfrm_state objects is > the current hash mask

All of this needs some tweaking.

Remove __initdata from "hashdist" so we can use it safely at run time.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/bootmem.h |   2 +-
 mm/page_alloc.c         |   2 +-
 net/xfrm/xfrm_state.c   | 247 ++++++++++++++++++++++++++++++++++++++----------
 3 files changed, 197 insertions(+), 54 deletions(-)

(limited to 'include')

diff --git a/include/linux/bootmem.h b/include/linux/bootmem.h
index 1021f508d82c..e319c649e4fd 100644
--- a/include/linux/bootmem.h
+++ b/include/linux/bootmem.h
@@ -114,7 +114,7 @@ extern void *__init alloc_large_system_hash(const char *tablename,
 #else
 #define HASHDIST_DEFAULT 0
 #endif
-extern int __initdata hashdist;		/* Distribute hashes across NUMA nodes? */
+extern int hashdist;		/* Distribute hashes across NUMA nodes? */
 
 
 #endif /* _LINUX_BOOTMEM_H */
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 54a4f5375bba..3b5358a0561f 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -2363,7 +2363,7 @@ int percpu_pagelist_fraction_sysctl_handler(ctl_table *table, int write,
 	return 0;
 }
 
-__initdata int hashdist = HASHDIST_DEFAULT;
+int hashdist = HASHDIST_DEFAULT;
 
 #ifdef CONFIG_NUMA
 static int __init set_hashdist(char *str)
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index fe3c8c38d5e1..445263c54c94 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -18,6 +18,9 @@
 #include <linux/pfkeyv2.h>
 #include <linux/ipsec.h>
 #include <linux/module.h>
+#include <linux/bootmem.h>
+#include <linux/vmalloc.h>
+#include <linux/cache.h>
 #include <asm/uaccess.h>
 
 struct sock *xfrm_nl;
@@ -38,102 +41,230 @@ EXPORT_SYMBOL(sysctl_xfrm_aevent_rseqth);
 
 static DEFINE_SPINLOCK(xfrm_state_lock);
 
-#define XFRM_DST_HSIZE		1024
-
 /* Hash table to find appropriate SA towards given target (endpoint
  * of tunnel or destination of transport mode) allowed by selector.
  *
  * Main use is finding SA after policy selected tunnel or transport mode.
  * Also, it can be used by ah/esp icmp error handler to find offending SA.
  */
-static struct hlist_head xfrm_state_bydst[XFRM_DST_HSIZE];
-static struct hlist_head xfrm_state_bysrc[XFRM_DST_HSIZE];
-static struct hlist_head xfrm_state_byspi[XFRM_DST_HSIZE];
-
-static __inline__
-unsigned __xfrm4_dst_hash(xfrm_address_t *addr)
+static struct hlist_head *xfrm_state_bydst __read_mostly;
+static struct hlist_head *xfrm_state_bysrc __read_mostly;
+static struct hlist_head *xfrm_state_byspi __read_mostly;
+static unsigned int xfrm_state_hmask __read_mostly;
+static unsigned int xfrm_state_hashmax __read_mostly = 1 * 1024 * 1024;
+static unsigned int xfrm_state_num;
+
+static inline unsigned int __xfrm4_dst_hash(xfrm_address_t *addr, unsigned int hmask)
 {
-	unsigned h;
+	unsigned int h;
 	h = ntohl(addr->a4);
-	h = (h ^ (h>>16)) % XFRM_DST_HSIZE;
+	h = (h ^ (h>>16)) & hmask;
 	return h;
 }
 
-static __inline__
-unsigned __xfrm6_dst_hash(xfrm_address_t *addr)
+static inline unsigned int __xfrm6_dst_hash(xfrm_address_t *addr, unsigned int hmask)
 {
-	unsigned h;
+	unsigned int h;
 	h = ntohl(addr->a6[2]^addr->a6[3]);
-	h = (h ^ (h>>16)) % XFRM_DST_HSIZE;
+	h = (h ^ (h>>16)) & hmask;
 	return h;
 }
 
-static __inline__
-unsigned __xfrm4_src_hash(xfrm_address_t *addr)
+static inline unsigned int __xfrm4_src_hash(xfrm_address_t *addr, unsigned int hmask)
 {
-	return __xfrm4_dst_hash(addr);
+	return __xfrm4_dst_hash(addr, hmask);
 }
 
-static __inline__
-unsigned __xfrm6_src_hash(xfrm_address_t *addr)
+static inline unsigned int __xfrm6_src_hash(xfrm_address_t *addr, unsigned int hmask)
 {
-	return __xfrm6_dst_hash(addr);
+	return __xfrm6_dst_hash(addr, hmask);
 }
 
-static __inline__
-unsigned xfrm_src_hash(xfrm_address_t *addr, unsigned short family)
+static inline unsigned __xfrm_src_hash(xfrm_address_t *addr, unsigned short family,  unsigned int hmask)
 {
 	switch (family) {
 	case AF_INET:
-		return __xfrm4_src_hash(addr);
+		return __xfrm4_src_hash(addr, hmask);
 	case AF_INET6:
-		return __xfrm6_src_hash(addr);
+		return __xfrm6_src_hash(addr, hmask);
 	}
 	return 0;
 }
 
-static __inline__
-unsigned xfrm_dst_hash(xfrm_address_t *addr, unsigned short family)
+static inline unsigned xfrm_src_hash(xfrm_address_t *addr, unsigned short family)
+{
+	return __xfrm_src_hash(addr, family, xfrm_state_hmask);
+}
+
+static inline unsigned int __xfrm_dst_hash(xfrm_address_t *addr, unsigned short family, unsigned int hmask)
 {
 	switch (family) {
 	case AF_INET:
-		return __xfrm4_dst_hash(addr);
+		return __xfrm4_dst_hash(addr, hmask);
 	case AF_INET6:
-		return __xfrm6_dst_hash(addr);
+		return __xfrm6_dst_hash(addr, hmask);
 	}
 	return 0;
 }
 
-static __inline__
-unsigned __xfrm4_spi_hash(xfrm_address_t *addr, u32 spi, u8 proto)
+static inline unsigned int xfrm_dst_hash(xfrm_address_t *addr, unsigned short family)
+{
+	return __xfrm_dst_hash(addr, family, xfrm_state_hmask);
+}
+
+static inline unsigned int __xfrm4_spi_hash(xfrm_address_t *addr, u32 spi, u8 proto,
+					unsigned int hmask)
 {
-	unsigned h;
+	unsigned int h;
 	h = ntohl(addr->a4^spi^proto);
-	h = (h ^ (h>>10) ^ (h>>20)) % XFRM_DST_HSIZE;
+	h = (h ^ (h>>10) ^ (h>>20)) & hmask;
 	return h;
 }
 
-static __inline__
-unsigned __xfrm6_spi_hash(xfrm_address_t *addr, u32 spi, u8 proto)
+static inline unsigned int __xfrm6_spi_hash(xfrm_address_t *addr, u32 spi, u8 proto,
+					    unsigned int hmask)
 {
-	unsigned h;
+	unsigned int h;
 	h = ntohl(addr->a6[2]^addr->a6[3]^spi^proto);
-	h = (h ^ (h>>10) ^ (h>>20)) % XFRM_DST_HSIZE;
+	h = (h ^ (h>>10) ^ (h>>20)) & hmask;
 	return h;
 }
 
-static __inline__
-unsigned xfrm_spi_hash(xfrm_address_t *addr, u32 spi, u8 proto, unsigned short family)
+static inline
+unsigned __xfrm_spi_hash(xfrm_address_t *addr, u32 spi, u8 proto, unsigned short family,
+			 unsigned int hmask)
 {
 	switch (family) {
 	case AF_INET:
-		return __xfrm4_spi_hash(addr, spi, proto);
+		return __xfrm4_spi_hash(addr, spi, proto, hmask);
 	case AF_INET6:
-		return __xfrm6_spi_hash(addr, spi, proto);
+		return __xfrm6_spi_hash(addr, spi, proto, hmask);
 	}
 	return 0;	/*XXX*/
 }
 
+static inline unsigned int
+xfrm_spi_hash(xfrm_address_t *addr, u32 spi, u8 proto, unsigned short family)
+{
+	return __xfrm_spi_hash(addr, spi, proto, family, xfrm_state_hmask);
+}
+
+static struct hlist_head *xfrm_state_hash_alloc(unsigned int sz)
+{
+	struct hlist_head *n;
+
+	if (sz <= PAGE_SIZE)
+		n = kmalloc(sz, GFP_KERNEL);
+	else if (hashdist)
+		n = __vmalloc(sz, GFP_KERNEL, PAGE_KERNEL);
+	else
+		n = (struct hlist_head *)
+			__get_free_pages(GFP_KERNEL, get_order(sz));
+
+	if (n)
+		memset(n, 0, sz);
+
+	return n;
+}
+
+static void xfrm_state_hash_free(struct hlist_head *n, unsigned int sz)
+{
+	if (sz <= PAGE_SIZE)
+		kfree(n);
+	else if (hashdist)
+		vfree(n);
+	else
+		free_pages((unsigned long)n, get_order(sz));
+}
+
+static void xfrm_hash_transfer(struct hlist_head *list,
+			       struct hlist_head *ndsttable,
+			       struct hlist_head *nsrctable,
+			       struct hlist_head *nspitable,
+			       unsigned int nhashmask)
+{
+	struct hlist_node *entry, *tmp;
+	struct xfrm_state *x;
+
+	hlist_for_each_entry_safe(x, entry, tmp, list, bydst) {
+		unsigned int h;
+
+		h = __xfrm_dst_hash(&x->id.daddr, x->props.family, nhashmask);
+		hlist_add_head(&x->bydst, ndsttable+h);
+
+		h = __xfrm_src_hash(&x->props.saddr, x->props.family,
+				    nhashmask);
+		hlist_add_head(&x->bysrc, nsrctable+h);
+
+		h = __xfrm_spi_hash(&x->id.daddr, x->id.spi, x->id.proto,
+				    x->props.family, nhashmask);
+		hlist_add_head(&x->byspi, nspitable+h);
+	}
+}
+
+static unsigned long xfrm_hash_new_size(void)
+{
+	return ((xfrm_state_hmask + 1) << 1) *
+		sizeof(struct hlist_head);
+}
+
+static DEFINE_MUTEX(hash_resize_mutex);
+
+static void xfrm_hash_resize(void *__unused)
+{
+	struct hlist_head *ndst, *nsrc, *nspi, *odst, *osrc, *ospi;
+	unsigned long nsize, osize;
+	unsigned int nhashmask, ohashmask;
+	int i;
+
+	mutex_lock(&hash_resize_mutex);
+
+	nsize = xfrm_hash_new_size();
+	ndst = xfrm_state_hash_alloc(nsize);
+	if (!ndst)
+		goto out_unlock;
+	nsrc = xfrm_state_hash_alloc(nsize);
+	if (!nsrc) {
+		xfrm_state_hash_free(ndst, nsize);
+		goto out_unlock;
+	}
+	nspi = xfrm_state_hash_alloc(nsize);
+	if (!nspi) {
+		xfrm_state_hash_free(ndst, nsize);
+		xfrm_state_hash_free(nsrc, nsize);
+		goto out_unlock;
+	}
+
+	spin_lock_bh(&xfrm_state_lock);
+
+	nhashmask = (nsize / sizeof(struct hlist_head)) - 1U;
+	for (i = xfrm_state_hmask; i >= 0; i--)
+		xfrm_hash_transfer(xfrm_state_bydst+i, ndst, nsrc, nspi,
+				   nhashmask);
+
+	odst = xfrm_state_bydst;
+	osrc = xfrm_state_bysrc;
+	ospi = xfrm_state_byspi;
+	ohashmask = xfrm_state_hmask;
+
+	xfrm_state_bydst = ndst;
+	xfrm_state_bysrc = nsrc;
+	xfrm_state_byspi = nspi;
+	xfrm_state_hmask = nhashmask;
+
+	spin_unlock_bh(&xfrm_state_lock);
+
+	osize = (ohashmask + 1) * sizeof(struct hlist_head);
+	xfrm_state_hash_free(odst, osize);
+	xfrm_state_hash_free(osrc, osize);
+	xfrm_state_hash_free(ospi, osize);
+
+out_unlock:
+	mutex_unlock(&hash_resize_mutex);
+}
+
+static DECLARE_WORK(xfrm_hash_work, xfrm_hash_resize, NULL);
+
 DECLARE_WAIT_QUEUE_HEAD(km_waitq);
 EXPORT_SYMBOL(km_waitq);
 
@@ -335,6 +466,7 @@ int __xfrm_state_delete(struct xfrm_state *x)
 			hlist_del(&x->byspi);
 			__xfrm_state_put(x);
 		}
+		xfrm_state_num--;
 		spin_unlock(&xfrm_state_lock);
 		if (del_timer(&x->timer))
 			__xfrm_state_put(x);
@@ -380,7 +512,7 @@ void xfrm_state_flush(u8 proto)
 	int i;
 
 	spin_lock_bh(&xfrm_state_lock);
-	for (i = 0; i < XFRM_DST_HSIZE; i++) {
+	for (i = 0; i < xfrm_state_hmask; i++) {
 		struct hlist_node *entry;
 		struct xfrm_state *x;
 restart:
@@ -611,7 +743,7 @@ out:
 
 static void __xfrm_state_insert(struct xfrm_state *x)
 {
-	unsigned h = xfrm_dst_hash(&x->id.daddr, x->props.family);
+	unsigned int h = xfrm_dst_hash(&x->id.daddr, x->props.family);
 
 	hlist_add_head(&x->bydst, xfrm_state_bydst+h);
 	xfrm_state_hold(x);
@@ -637,6 +769,13 @@ static void __xfrm_state_insert(struct xfrm_state *x)
 		xfrm_state_hold(x);
 
 	wake_up(&km_waitq);
+
+	xfrm_state_num++;
+
+	if (x->bydst.next != NULL &&
+	    (xfrm_state_hmask + 1) < xfrm_state_hashmax &&
+	    xfrm_state_num > xfrm_state_hmask)
+		schedule_work(&xfrm_hash_work);
 }
 
 void xfrm_state_insert(struct xfrm_state *x)
@@ -984,7 +1123,7 @@ static struct xfrm_state *__xfrm_find_acq_byseq(u32 seq)
 {
 	int i;
 
-	for (i = 0; i < XFRM_DST_HSIZE; i++) {
+	for (i = 0; i <= xfrm_state_hmask; i++) {
 		struct hlist_node *entry;
 		struct xfrm_state *x;
 
@@ -1026,7 +1165,7 @@ EXPORT_SYMBOL(xfrm_get_acqseq);
 void
 xfrm_alloc_spi(struct xfrm_state *x, u32 minspi, u32 maxspi)
 {
-	u32 h;
+	unsigned int h;
 	struct xfrm_state *x0;
 
 	if (x->id.spi)
@@ -1074,7 +1213,7 @@ int xfrm_state_walk(u8 proto, int (*func)(struct xfrm_state *, int, void*),
 	int err = 0;
 
 	spin_lock_bh(&xfrm_state_lock);
-	for (i = 0; i < XFRM_DST_HSIZE; i++) {
+	for (i = 0; i <= xfrm_state_hmask; i++) {
 		hlist_for_each_entry(x, entry, xfrm_state_bydst+i, bydst) {
 			if (xfrm_id_proto_match(x->id.proto, proto))
 				count++;
@@ -1085,7 +1224,7 @@ int xfrm_state_walk(u8 proto, int (*func)(struct xfrm_state *, int, void*),
 		goto out;
 	}
 
-	for (i = 0; i < XFRM_DST_HSIZE; i++) {
+	for (i = 0; i <= xfrm_state_hmask; i++) {
 		hlist_for_each_entry(x, entry, xfrm_state_bydst+i, bydst) {
 			if (!xfrm_id_proto_match(x->id.proto, proto))
 				continue;
@@ -1531,13 +1670,17 @@ EXPORT_SYMBOL(xfrm_init_state);
  
 void __init xfrm_state_init(void)
 {
-	int i;
+	unsigned int sz;
+
+	sz = sizeof(struct hlist_head) * 8;
+
+	xfrm_state_bydst = xfrm_state_hash_alloc(sz);
+	xfrm_state_bysrc = xfrm_state_hash_alloc(sz);
+	xfrm_state_byspi = xfrm_state_hash_alloc(sz);
+	if (!xfrm_state_bydst || !xfrm_state_bysrc || !xfrm_state_byspi)
+		panic("XFRM: Cannot allocate bydst/bysrc/byspi hashes.");
+	xfrm_state_hmask = ((sz / sizeof(struct hlist_head)) - 1);
 
-	for (i=0; i<XFRM_DST_HSIZE; i++) {
-		INIT_HLIST_HEAD(&xfrm_state_bydst[i]);
-		INIT_HLIST_HEAD(&xfrm_state_bysrc[i]);
-		INIT_HLIST_HEAD(&xfrm_state_byspi[i]);
-	}
 	INIT_WORK(&xfrm_state_gc_work, xfrm_state_gc_task, NULL);
 }
 
-- 
cgit v1.2.3


From 9d4a706d852411154d0c91b9ffb3bec68b94b25c Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@sunset.davemloft.net>
Date: Thu, 24 Aug 2006 03:18:09 -0700
Subject: [XFRM]: Add generation count to xfrm_state and xfrm_dst.

Each xfrm_state inserted gets a new generation counter
value.  When a bundle is created, the xfrm_dst objects
get the current generation counter of the xfrm_state
they will attach to at dst->xfrm.

xfrm_bundle_ok() will return false if it sees an
xfrm_dst with a generation count different from the
generation count of the xfrm_state that dst points to.

This provides a facility by which to passively and
cheaply invalidate cached IPSEC routes during SA
database changes.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h      | 3 +++
 net/ipv4/xfrm4_policy.c | 1 +
 net/ipv6/xfrm6_policy.c | 1 +
 net/xfrm/xfrm_policy.c  | 2 ++
 net/xfrm/xfrm_state.c   | 3 +++
 5 files changed, 10 insertions(+)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 3405e5d9d51c..fd4a300b5baf 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -104,6 +104,8 @@ struct xfrm_state
 	struct xfrm_id		id;
 	struct xfrm_selector	sel;
 
+	u32			genid;
+
 	/* Key manger bits */
 	struct {
 		u8		state;
@@ -590,6 +592,7 @@ struct xfrm_dst
 		struct rt6_info		rt6;
 	} u;
 	struct dst_entry *route;
+	u32 genid;
 	u32 route_mtu_cached;
 	u32 child_mtu_cached;
 	u32 route_cookie;
diff --git a/net/ipv4/xfrm4_policy.c b/net/ipv4/xfrm4_policy.c
index 42d8ded0f96a..479598566f1d 100644
--- a/net/ipv4/xfrm4_policy.c
+++ b/net/ipv4/xfrm4_policy.c
@@ -93,6 +93,7 @@ __xfrm4_bundle_create(struct xfrm_policy *policy, struct xfrm_state **xfrm, int
 
 		xdst = (struct xfrm_dst *)dst1;
 		xdst->route = &rt->u.dst;
+		xdst->genid = xfrm[i]->genid;
 
 		dst1->next = dst_prev;
 		dst_prev = dst1;
diff --git a/net/ipv6/xfrm6_policy.c b/net/ipv6/xfrm6_policy.c
index 98c2fe449b3f..9391c4c94feb 100644
--- a/net/ipv6/xfrm6_policy.c
+++ b/net/ipv6/xfrm6_policy.c
@@ -149,6 +149,7 @@ __xfrm6_bundle_create(struct xfrm_policy *policy, struct xfrm_state **xfrm, int
 
 		xdst = (struct xfrm_dst *)dst1;
 		xdst->route = &rt->u.dst;
+		xdst->genid = xfrm[i]->genid;
 		if (rt->rt6i_node)
 			xdst->route_cookie = rt->rt6i_node->fn_sernum;
 
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 1732159ffd01..7fc6944ee36f 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -1536,6 +1536,8 @@ int xfrm_bundle_ok(struct xfrm_dst *first, struct flowi *fl, int family, int str
 			return 0;
 		if (dst->xfrm->km.state != XFRM_STATE_VALID)
 			return 0;
+		if (xdst->genid != dst->xfrm->genid)
+			return 0;
 
 		if (strict && fl && dst->xfrm->props.mode != XFRM_MODE_TUNNEL &&
 		    !xfrm_state_addr_flow_check(dst->xfrm, fl, family))
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 445263c54c94..535d43c14720 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -53,6 +53,7 @@ static struct hlist_head *xfrm_state_byspi __read_mostly;
 static unsigned int xfrm_state_hmask __read_mostly;
 static unsigned int xfrm_state_hashmax __read_mostly = 1 * 1024 * 1024;
 static unsigned int xfrm_state_num;
+static unsigned int xfrm_state_genid;
 
 static inline unsigned int __xfrm4_dst_hash(xfrm_address_t *addr, unsigned int hmask)
 {
@@ -745,6 +746,8 @@ static void __xfrm_state_insert(struct xfrm_state *x)
 {
 	unsigned int h = xfrm_dst_hash(&x->id.daddr, x->props.family);
 
+	x->genid = ++xfrm_state_genid;
+
 	hlist_add_head(&x->bydst, xfrm_state_bydst+h);
 	xfrm_state_hold(x);
 
-- 
cgit v1.2.3


From c7f5ea3a4d1ae6b3b426e113358fdc57494bc754 Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@sunset.davemloft.net>
Date: Thu, 24 Aug 2006 03:29:04 -0700
Subject: [XFRM]: Do not flush all bundles on SA insert.

Instead, simply set all potentially aliasing existing xfrm_state
objects to have the current generation counter value.

This will make routes get relooked up the next time an existing
route mentioning these aliased xfrm_state objects gets used,
via xfrm_dst_check().

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h     |  1 -
 net/xfrm/xfrm_policy.c | 10 ----------
 net/xfrm/xfrm_state.c  | 25 ++++++++++++++++++++-----
 3 files changed, 20 insertions(+), 16 deletions(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index fd4a300b5baf..a620a43c9eeb 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -996,7 +996,6 @@ struct xfrm_state * xfrm_find_acq(u8 mode, u32 reqid, u8 proto,
 extern void xfrm_policy_flush(u8 type);
 extern int xfrm_sk_policy_insert(struct sock *sk, int dir, struct xfrm_policy *pol);
 extern int xfrm_flush_bundles(void);
-extern void xfrm_flush_all_bundles(void);
 extern int xfrm_bundle_ok(struct xfrm_dst *xdst, struct flowi *fl, int family, int strict);
 extern void xfrm_init_pmtu(struct dst_entry *dst);
 
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 7fc6944ee36f..cfa5c692f2e8 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -1478,16 +1478,6 @@ int xfrm_flush_bundles(void)
 	return 0;
 }
 
-static int always_true(struct dst_entry *dst)
-{
-	return 1;
-}
-
-void xfrm_flush_all_bundles(void)
-{
-	xfrm_prune_bundles(always_true);
-}
-
 void xfrm_init_pmtu(struct dst_entry *dst)
 {
 	do {
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 98200397e098..77ef796c9d0d 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -761,13 +761,30 @@ static void __xfrm_state_insert(struct xfrm_state *x)
 		schedule_work(&xfrm_hash_work);
 }
 
+/* xfrm_state_lock is held */
+static void __xfrm_state_bump_genids(struct xfrm_state *xnew)
+{
+	unsigned short family = xnew->props.family;
+	u32 reqid = xnew->props.reqid;
+	struct xfrm_state *x;
+	struct hlist_node *entry;
+	unsigned int h;
+
+	h = xfrm_dst_hash(&xnew->id.daddr, reqid, family);
+	hlist_for_each_entry(x, entry, xfrm_state_bydst+h, bydst) {
+		if (x->props.family	== family &&
+		    x->props.reqid	== reqid &&
+		    !xfrm_addr_cmp(&x->id.daddr, &xnew->id.daddr, family))
+			x->genid = xfrm_state_genid;
+	}
+}
+
 void xfrm_state_insert(struct xfrm_state *x)
 {
 	spin_lock_bh(&xfrm_state_lock);
+	__xfrm_state_bump_genids(x);
 	__xfrm_state_insert(x);
 	spin_unlock_bh(&xfrm_state_lock);
-
-	xfrm_flush_all_bundles();
 }
 EXPORT_SYMBOL(xfrm_state_insert);
 
@@ -889,15 +906,13 @@ int xfrm_state_add(struct xfrm_state *x)
 				     x->id.proto,
 				     &x->id.daddr, &x->props.saddr, 0);
 
+	__xfrm_state_bump_genids(x);
 	__xfrm_state_insert(x);
 	err = 0;
 
 out:
 	spin_unlock_bh(&xfrm_state_lock);
 
-	if (!err)
-		xfrm_flush_all_bundles();
-
 	if (x1) {
 		xfrm_state_delete(x1);
 		xfrm_state_put(x1);
-- 
cgit v1.2.3


From 1c0953997567b22e32fdf85d3b4bc0f2461fd161 Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@sunset.davemloft.net>
Date: Thu, 24 Aug 2006 03:30:28 -0700
Subject: [XFRM]: Purge dst references to deleted SAs passively.

Just let GC and other normal mechanisms take care of getting
rid of DST cache references to deleted xfrm_state objects
instead of walking all the policy bundles.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h     |  1 -
 net/xfrm/xfrm_policy.c |  2 +-
 net/xfrm/xfrm_state.c  | 17 -----------------
 3 files changed, 1 insertion(+), 19 deletions(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index a620a43c9eeb..c7870b6eae01 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -995,7 +995,6 @@ struct xfrm_state * xfrm_find_acq(u8 mode, u32 reqid, u8 proto,
 				  int create, unsigned short family);
 extern void xfrm_policy_flush(u8 type);
 extern int xfrm_sk_policy_insert(struct sock *sk, int dir, struct xfrm_policy *pol);
-extern int xfrm_flush_bundles(void);
 extern int xfrm_bundle_ok(struct xfrm_dst *xdst, struct flowi *fl, int family, int strict);
 extern void xfrm_init_pmtu(struct dst_entry *dst);
 
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index cfa5c692f2e8..1bcaae4adf3a 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -1472,7 +1472,7 @@ static void __xfrm_garbage_collect(void)
 	xfrm_prune_bundles(unused_bundle);
 }
 
-int xfrm_flush_bundles(void)
+static int xfrm_flush_bundles(void)
 {
 	xfrm_prune_bundles(stale_bundle);
 	return 0;
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 77ef796c9d0d..9ff00b7d6ad3 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -256,8 +256,6 @@ static struct work_struct xfrm_state_gc_work;
 static HLIST_HEAD(xfrm_state_gc_list);
 static DEFINE_SPINLOCK(xfrm_state_gc_lock);
 
-static int xfrm_state_gc_flush_bundles;
-
 int __xfrm_state_delete(struct xfrm_state *x);
 
 static struct xfrm_state_afinfo *xfrm_state_get_afinfo(unsigned short family);
@@ -293,11 +291,6 @@ static void xfrm_state_gc_task(void *data)
 	struct hlist_node *entry, *tmp;
 	struct hlist_head gc_list;
 
-	if (xfrm_state_gc_flush_bundles) {
-		xfrm_state_gc_flush_bundles = 0;
-		xfrm_flush_bundles();
-	}
-
 	spin_lock_bh(&xfrm_state_gc_lock);
 	gc_list.first = xfrm_state_gc_list.first;
 	INIT_HLIST_HEAD(&xfrm_state_gc_list);
@@ -454,16 +447,6 @@ int __xfrm_state_delete(struct xfrm_state *x)
 		if (del_timer(&x->rtimer))
 			__xfrm_state_put(x);
 
-		/* The number two in this test is the reference
-		 * mentioned in the comment below plus the reference
-		 * our caller holds.  A larger value means that
-		 * there are DSTs attached to this xfrm_state.
-		 */
-		if (atomic_read(&x->refcnt) > 2) {
-			xfrm_state_gc_flush_bundles = 1;
-			schedule_work(&xfrm_state_gc_work);
-		}
-
 		/* All xfrm_state objects are created by xfrm_state_alloc.
 		 * The xfrm_state_alloc call gives a reference, and that
 		 * is what we are dropping here.
-- 
cgit v1.2.3


From 2518c7c2b3d7f0a6b302b4efe17c911f8dd4049f Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@sunset.davemloft.net>
Date: Thu, 24 Aug 2006 04:45:07 -0700
Subject: [XFRM]: Hash policies when non-prefixed.

This idea is from Alexey Kuznetsov.

It is common for policies to be non-prefixed.  And for
that case we can optimize lookups, insert, etc. quite
a bit.

For each direction, we have a dynamically sized policy
hash table for non-prefixed policies.  We also have a
hash table on policy->index.

For prefixed policies, we have a list per-direction which
we will consult on lookups when a non-prefix hashtable
lookup fails.

This still isn't as efficient as I would like it.  There
are four immediate problems:

1) Lots of excessive refcounting, which can be fixed just
   like xfrm_state was
2) We do 2 hash probes on insert, one to look for dups and
   one to allocate a unique policy->index.  Althought I wonder
   how much this matters since xfrm_state inserts do up to
   3 hash probes and that seems to perform fine.
3) xfrm_policy_insert() is very complex because of the priority
   ordering and entry replacement logic.
4) Lots of counter bumping, in addition to policy refcounts,
   in the form of xfrm_policy_count[].  This is merely used
   to let code path(s) know that some IPSEC rules exist.  So
   this count is indexed per-direction, maybe that is overkill.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h     |  23 +-
 net/xfrm/xfrm_policy.c | 681 +++++++++++++++++++++++++++++++++++++++----------
 2 files changed, 546 insertions(+), 158 deletions(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index c7870b6eae01..0acabf2a0a8f 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -331,7 +331,8 @@ struct xfrm_tmpl
 struct xfrm_policy
 {
 	struct xfrm_policy	*next;
-	struct list_head	list;
+	struct hlist_node	bydst;
+	struct hlist_node	byidx;
 
 	/* This lock only affects elements except for entry. */
 	rwlock_t		lock;
@@ -385,21 +386,7 @@ struct xfrm_mgr
 extern int xfrm_register_km(struct xfrm_mgr *km);
 extern int xfrm_unregister_km(struct xfrm_mgr *km);
 
-
-extern struct xfrm_policy *xfrm_policy_list[XFRM_POLICY_MAX*2];
-#ifdef CONFIG_XFRM_SUB_POLICY
-extern struct xfrm_policy *xfrm_policy_list_sub[XFRM_POLICY_MAX*2];
-
-static inline int xfrm_policy_lists_empty(int dir)
-{
-	return (!xfrm_policy_list[dir] && !xfrm_policy_list_sub[dir]);
-}
-#else
-static inline int xfrm_policy_lists_empty(int dir)
-{
-	return (!xfrm_policy_list[dir]);
-}
-#endif
+extern unsigned int xfrm_policy_count[XFRM_POLICY_MAX*2];
 
 static inline void xfrm_pol_hold(struct xfrm_policy *policy)
 {
@@ -678,7 +665,7 @@ static inline int xfrm_policy_check(struct sock *sk, int dir, struct sk_buff *sk
 	if (sk && sk->sk_policy[XFRM_POLICY_IN])
 		return __xfrm_policy_check(sk, dir, skb, family);
 
-	return	(xfrm_policy_lists_empty(dir) && !skb->sp) ||
+	return	(!xfrm_policy_count[dir] && !skb->sp) ||
 		(skb->dst->flags & DST_NOPOLICY) ||
 		__xfrm_policy_check(sk, dir, skb, family);
 }
@@ -698,7 +685,7 @@ extern int __xfrm_route_forward(struct sk_buff *skb, unsigned short family);
 
 static inline int xfrm_route_forward(struct sk_buff *skb, unsigned short family)
 {
-	return	xfrm_policy_lists_empty(XFRM_POLICY_OUT) ||
+	return	!xfrm_policy_count[XFRM_POLICY_OUT] ||
 		(skb->dst->flags & DST_NOXFRM) ||
 		__xfrm_route_forward(skb, family);
 }
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 1bcaae4adf3a..087a5443b051 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -22,6 +22,9 @@
 #include <linux/netdevice.h>
 #include <linux/netfilter.h>
 #include <linux/module.h>
+#include <linux/bootmem.h>
+#include <linux/vmalloc.h>
+#include <linux/cache.h>
 #include <net/xfrm.h>
 #include <net/ip.h>
 
@@ -30,26 +33,8 @@ EXPORT_SYMBOL(xfrm_cfg_mutex);
 
 static DEFINE_RWLOCK(xfrm_policy_lock);
 
-struct xfrm_policy *xfrm_policy_list[XFRM_POLICY_MAX*2];
-EXPORT_SYMBOL(xfrm_policy_list);
-#ifdef CONFIG_XFRM_SUB_POLICY
-struct xfrm_policy *xfrm_policy_list_sub[XFRM_POLICY_MAX*2];
-EXPORT_SYMBOL(xfrm_policy_list_sub);
-
-#define XFRM_POLICY_LISTS(type) \
-	((type == XFRM_POLICY_TYPE_SUB) ? xfrm_policy_list_sub : \
-	 xfrm_policy_list)
-#define XFRM_POLICY_LISTHEAD(type, dir) \
-	((type == XFRM_POLICY_TYPE_SUB) ? xfrm_policy_list_sub[dir] : \
-	 xfrm_policy_list[dir])
-#define XFRM_POLICY_LISTHEADP(type, dir) \
-	((type == XFRM_POLICY_TYPE_SUB) ? &xfrm_policy_list_sub[dir] : \
-	 &xfrm_policy_list[dir])
-#else
-#define XFRM_POLICY_LISTS(type)              xfrm_policy_list
-#define XFRM_POLICY_LISTHEAD(type, dif)      xfrm_policy_list[dir]
-#define XFRM_POLICY_LISTHEADP(type, dif)     &xfrm_policy_list[dir]
-#endif
+unsigned int xfrm_policy_count[XFRM_POLICY_MAX*2];
+EXPORT_SYMBOL(xfrm_policy_count);
 
 static DEFINE_RWLOCK(xfrm_policy_afinfo_lock);
 static struct xfrm_policy_afinfo *xfrm_policy_afinfo[NPROTO];
@@ -57,8 +42,7 @@ static struct xfrm_policy_afinfo *xfrm_policy_afinfo[NPROTO];
 static kmem_cache_t *xfrm_dst_cache __read_mostly;
 
 static struct work_struct xfrm_policy_gc_work;
-static struct list_head xfrm_policy_gc_list =
-	LIST_HEAD_INIT(xfrm_policy_gc_list);
+static HLIST_HEAD(xfrm_policy_gc_list);
 static DEFINE_SPINLOCK(xfrm_policy_gc_lock);
 
 static struct xfrm_policy_afinfo *xfrm_policy_get_afinfo(unsigned short family);
@@ -328,8 +312,10 @@ struct xfrm_policy *xfrm_policy_alloc(gfp_t gfp)
 	policy = kzalloc(sizeof(struct xfrm_policy), gfp);
 
 	if (policy) {
-		atomic_set(&policy->refcnt, 1);
+		INIT_HLIST_NODE(&policy->bydst);
+		INIT_HLIST_NODE(&policy->byidx);
 		rwlock_init(&policy->lock);
+		atomic_set(&policy->refcnt, 1);
 		init_timer(&policy->timer);
 		policy->timer.data = (unsigned long)policy;
 		policy->timer.function = xfrm_policy_timer;
@@ -375,17 +361,16 @@ static void xfrm_policy_gc_kill(struct xfrm_policy *policy)
 static void xfrm_policy_gc_task(void *data)
 {
 	struct xfrm_policy *policy;
-	struct list_head *entry, *tmp;
-	struct list_head gc_list = LIST_HEAD_INIT(gc_list);
+	struct hlist_node *entry, *tmp;
+	struct hlist_head gc_list;
 
 	spin_lock_bh(&xfrm_policy_gc_lock);
-	list_splice_init(&xfrm_policy_gc_list, &gc_list);
+	gc_list.first = xfrm_policy_gc_list.first;
+	INIT_HLIST_HEAD(&xfrm_policy_gc_list);
 	spin_unlock_bh(&xfrm_policy_gc_lock);
 
-	list_for_each_safe(entry, tmp, &gc_list) {
-		policy = list_entry(entry, struct xfrm_policy, list);
+	hlist_for_each_entry_safe(policy, entry, tmp, &gc_list, bydst)
 		xfrm_policy_gc_kill(policy);
-	}
 }
 
 /* Rule must be locked. Release descentant resources, announce
@@ -407,70 +392,354 @@ static void xfrm_policy_kill(struct xfrm_policy *policy)
 	}
 
 	spin_lock(&xfrm_policy_gc_lock);
-	list_add(&policy->list, &xfrm_policy_gc_list);
+	hlist_add_head(&policy->bydst, &xfrm_policy_gc_list);
 	spin_unlock(&xfrm_policy_gc_lock);
 
 	schedule_work(&xfrm_policy_gc_work);
 }
 
+struct xfrm_policy_hash {
+	struct hlist_head	*table;
+	unsigned int		hmask;
+};
+
+static struct hlist_head xfrm_policy_inexact[XFRM_POLICY_MAX*2];
+static struct xfrm_policy_hash xfrm_policy_bydst[XFRM_POLICY_MAX*2] __read_mostly;
+static struct hlist_head *xfrm_policy_byidx __read_mostly;
+static unsigned int xfrm_idx_hmask __read_mostly;
+static unsigned int xfrm_policy_hashmax __read_mostly = 1 * 1024 * 1024;
+
+static inline unsigned int __idx_hash(u32 index, unsigned int hmask)
+{
+	return (index ^ (index >> 8)) & hmask;
+}
+
+static inline unsigned int idx_hash(u32 index)
+{
+	return __idx_hash(index, xfrm_idx_hmask);
+}
+
+static inline unsigned int __sel_hash(struct xfrm_selector *sel, unsigned short family, unsigned int hmask)
+{
+	xfrm_address_t *daddr = &sel->daddr;
+	xfrm_address_t *saddr = &sel->saddr;
+	unsigned int h = 0;
+
+	switch (family) {
+	case AF_INET:
+		if (sel->prefixlen_d != 32 ||
+		    sel->prefixlen_s != 32)
+			return hmask + 1;
+
+		h = ntohl(daddr->a4 ^ saddr->a4);
+		break;
+
+	case AF_INET6:
+		if (sel->prefixlen_d != 128 ||
+		    sel->prefixlen_s != 128)
+			return hmask + 1;
+
+		h = ntohl(daddr->a6[2] ^ daddr->a6[3] ^
+			  saddr->a6[2] ^ saddr->a6[3]);
+		break;
+	};
+	h ^= (h >> 16);
+	return h & hmask;
+}
+
+static inline unsigned int __addr_hash(xfrm_address_t *daddr, xfrm_address_t *saddr, unsigned short family, unsigned int hmask)
+{
+	unsigned int h = 0;
+
+	switch (family) {
+	case AF_INET:
+		h = ntohl(daddr->a4 ^ saddr->a4);
+		break;
+
+	case AF_INET6:
+		h = ntohl(daddr->a6[2] ^ daddr->a6[3] ^
+			  saddr->a6[2] ^ saddr->a6[3]);
+		break;
+	};
+	h ^= (h >> 16);
+	return h & hmask;
+}
+
+static struct hlist_head *policy_hash_bysel(struct xfrm_selector *sel, unsigned short family, int dir)
+{
+	unsigned int hmask = xfrm_policy_bydst[dir].hmask;
+	unsigned int hash = __sel_hash(sel, family, hmask);
+
+	return (hash == hmask + 1 ?
+		&xfrm_policy_inexact[dir] :
+		xfrm_policy_bydst[dir].table + hash);
+}
+
+static struct hlist_head *policy_hash_direct(xfrm_address_t *daddr, xfrm_address_t *saddr, unsigned short family, int dir)
+{
+	unsigned int hmask = xfrm_policy_bydst[dir].hmask;
+	unsigned int hash = __addr_hash(daddr, saddr, family, hmask);
+
+	return xfrm_policy_bydst[dir].table + hash;
+}
+
+static struct hlist_head *xfrm_policy_hash_alloc(unsigned int sz)
+{
+	struct hlist_head *n;
+
+	if (sz <= PAGE_SIZE)
+		n = kmalloc(sz, GFP_KERNEL);
+	else if (hashdist)
+		n = __vmalloc(sz, GFP_KERNEL, PAGE_KERNEL);
+	else
+		n = (struct hlist_head *)
+			__get_free_pages(GFP_KERNEL, get_order(sz));
+
+	if (n)
+		memset(n, 0, sz);
+
+	return n;
+}
+
+static void xfrm_policy_hash_free(struct hlist_head *n, unsigned int sz)
+{
+	if (sz <= PAGE_SIZE)
+		kfree(n);
+	else if (hashdist)
+		vfree(n);
+	else
+		free_pages((unsigned long)n, get_order(sz));
+}
+
+static void xfrm_dst_hash_transfer(struct hlist_head *list,
+				   struct hlist_head *ndsttable,
+				   unsigned int nhashmask)
+{
+	struct hlist_node *entry, *tmp;
+	struct xfrm_policy *pol;
+
+	hlist_for_each_entry_safe(pol, entry, tmp, list, bydst) {
+		unsigned int h;
+
+		h = __addr_hash(&pol->selector.daddr, &pol->selector.saddr,
+				pol->family, nhashmask);
+		hlist_add_head(&pol->bydst, ndsttable+h);
+	}
+}
+
+static void xfrm_idx_hash_transfer(struct hlist_head *list,
+				   struct hlist_head *nidxtable,
+				   unsigned int nhashmask)
+{
+	struct hlist_node *entry, *tmp;
+	struct xfrm_policy *pol;
+
+	hlist_for_each_entry_safe(pol, entry, tmp, list, byidx) {
+		unsigned int h;
+
+		h = __idx_hash(pol->index, nhashmask);
+		hlist_add_head(&pol->byidx, nidxtable+h);
+	}
+}
+
+static unsigned long xfrm_new_hash_mask(unsigned int old_hmask)
+{
+	return ((old_hmask + 1) << 1) - 1;
+}
+
+static void xfrm_bydst_resize(int dir)
+{
+	unsigned int hmask = xfrm_policy_bydst[dir].hmask;
+	unsigned int nhashmask = xfrm_new_hash_mask(hmask);
+	unsigned int nsize = (nhashmask + 1) * sizeof(struct hlist_head);
+	struct hlist_head *odst = xfrm_policy_bydst[dir].table;
+	struct hlist_head *ndst = xfrm_policy_hash_alloc(nsize);
+	int i;
+
+	if (!ndst)
+		return;
+
+	write_lock_bh(&xfrm_policy_lock);
+
+	for (i = hmask; i >= 0; i--)
+		xfrm_dst_hash_transfer(odst + i, ndst, nhashmask);
+
+	xfrm_policy_bydst[dir].table = ndst;
+	xfrm_policy_bydst[dir].hmask = nhashmask;
+
+	write_unlock_bh(&xfrm_policy_lock);
+
+	xfrm_policy_hash_free(odst, (hmask + 1) * sizeof(struct hlist_head));
+}
+
+static void xfrm_byidx_resize(int total)
+{
+	unsigned int hmask = xfrm_idx_hmask;
+	unsigned int nhashmask = xfrm_new_hash_mask(hmask);
+	unsigned int nsize = (nhashmask + 1) * sizeof(struct hlist_head);
+	struct hlist_head *oidx = xfrm_policy_byidx;
+	struct hlist_head *nidx = xfrm_policy_hash_alloc(nsize);
+	int i;
+
+	if (!nidx)
+		return;
+
+	write_lock_bh(&xfrm_policy_lock);
+
+	for (i = hmask; i >= 0; i--)
+		xfrm_idx_hash_transfer(oidx + i, nidx, nhashmask);
+
+	xfrm_policy_byidx = nidx;
+	xfrm_idx_hmask = nhashmask;
+
+	write_unlock_bh(&xfrm_policy_lock);
+
+	xfrm_policy_hash_free(oidx, (hmask + 1) * sizeof(struct hlist_head));
+}
+
+static inline int xfrm_bydst_should_resize(int dir, int *total)
+{
+	unsigned int cnt = xfrm_policy_count[dir];
+	unsigned int hmask = xfrm_policy_bydst[dir].hmask;
+
+	if (total)
+		*total += cnt;
+
+	if ((hmask + 1) < xfrm_policy_hashmax &&
+	    cnt > hmask)
+		return 1;
+
+	return 0;
+}
+
+static inline int xfrm_byidx_should_resize(int total)
+{
+	unsigned int hmask = xfrm_idx_hmask;
+
+	if ((hmask + 1) < xfrm_policy_hashmax &&
+	    total > hmask)
+		return 1;
+
+	return 0;
+}
+
+static DEFINE_MUTEX(hash_resize_mutex);
+
+static void xfrm_hash_resize(void *__unused)
+{
+	int dir, total;
+
+	mutex_lock(&hash_resize_mutex);
+
+	total = 0;
+	for (dir = 0; dir < XFRM_POLICY_MAX * 2; dir++) {
+		if (xfrm_bydst_should_resize(dir, &total))
+			xfrm_bydst_resize(dir);
+	}
+	if (xfrm_byidx_should_resize(total))
+		xfrm_byidx_resize(total);
+
+	mutex_unlock(&hash_resize_mutex);
+}
+
+static DECLARE_WORK(xfrm_hash_work, xfrm_hash_resize, NULL);
+
 /* Generate new index... KAME seems to generate them ordered by cost
  * of an absolute inpredictability of ordering of rules. This will not pass. */
 static u32 xfrm_gen_index(u8 type, int dir)
 {
-	u32 idx;
-	struct xfrm_policy *p;
 	static u32 idx_generator;
 
 	for (;;) {
+		struct hlist_node *entry;
+		struct hlist_head *list;
+		struct xfrm_policy *p;
+		u32 idx;
+		int found;
+
 		idx = (idx_generator | dir);
 		idx_generator += 8;
 		if (idx == 0)
 			idx = 8;
-		for (p = XFRM_POLICY_LISTHEAD(type, dir); p; p = p->next) {
-			if (p->index == idx)
+		list = xfrm_policy_byidx + idx_hash(idx);
+		found = 0;
+		hlist_for_each_entry(p, entry, list, byidx) {
+			if (p->index == idx) {
+				found = 1;
 				break;
+			}
 		}
-		if (!p)
+		if (!found)
 			return idx;
 	}
 }
 
+static inline int selector_cmp(struct xfrm_selector *s1, struct xfrm_selector *s2)
+{
+	u32 *p1 = (u32 *) s1;
+	u32 *p2 = (u32 *) s2;
+	int len = sizeof(struct xfrm_selector) / sizeof(u32);
+	int i;
+
+	for (i = 0; i < len; i++) {
+		if (p1[i] != p2[i])
+			return 1;
+	}
+
+	return 0;
+}
+
 int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl)
 {
-	struct xfrm_policy *pol, **p;
-	struct xfrm_policy *delpol = NULL;
-	struct xfrm_policy **newpos = NULL;
+	struct xfrm_policy *pol;
+	struct xfrm_policy *delpol;
+	struct hlist_head *chain;
+	struct hlist_node *entry, *newpos, *last;
 	struct dst_entry *gc_list;
 
 	write_lock_bh(&xfrm_policy_lock);
-	for (p = XFRM_POLICY_LISTHEADP(policy->type, dir); (pol=*p)!=NULL;) {
-		if (!delpol && memcmp(&policy->selector, &pol->selector, sizeof(pol->selector)) == 0 &&
+	chain = policy_hash_bysel(&policy->selector, policy->family, dir);
+	delpol = NULL;
+	newpos = NULL;
+	last = NULL;
+	hlist_for_each_entry(pol, entry, chain, bydst) {
+		if (!delpol &&
+		    pol->type == policy->type &&
+		    !selector_cmp(&pol->selector, &policy->selector) &&
 		    xfrm_sec_ctx_match(pol->security, policy->security)) {
 			if (excl) {
 				write_unlock_bh(&xfrm_policy_lock);
 				return -EEXIST;
 			}
-			*p = pol->next;
 			delpol = pol;
 			if (policy->priority > pol->priority)
 				continue;
 		} else if (policy->priority >= pol->priority) {
-			p = &pol->next;
+			last = &pol->bydst;
 			continue;
 		}
 		if (!newpos)
-			newpos = p;
+			newpos = &pol->bydst;
 		if (delpol)
 			break;
-		p = &pol->next;
+		last = &pol->bydst;
 	}
+	if (!newpos)
+		newpos = last;
 	if (newpos)
-		p = newpos;
+		hlist_add_after(newpos, &policy->bydst);
+	else
+		hlist_add_head(&policy->bydst, chain);
 	xfrm_pol_hold(policy);
-	policy->next = *p;
-	*p = policy;
+	xfrm_policy_count[dir]++;
 	atomic_inc(&flow_cache_genid);
+	if (delpol) {
+		hlist_del(&delpol->bydst);
+		hlist_del(&delpol->byidx);
+		xfrm_policy_count[dir]--;
+	}
 	policy->index = delpol ? delpol->index : xfrm_gen_index(policy->type, dir);
+	hlist_add_head(&policy->byidx, xfrm_policy_byidx+idx_hash(policy->index));
 	policy->curlft.add_time = (unsigned long)xtime.tv_sec;
 	policy->curlft.use_time = 0;
 	if (!mod_timer(&policy->timer, jiffies + HZ))
@@ -479,10 +748,13 @@ int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl)
 
 	if (delpol)
 		xfrm_policy_kill(delpol);
+	else if (xfrm_bydst_should_resize(dir, NULL))
+		schedule_work(&xfrm_hash_work);
 
 	read_lock_bh(&xfrm_policy_lock);
 	gc_list = NULL;
-	for (policy = policy->next; policy; policy = policy->next) {
+	entry = &policy->bydst;
+	hlist_for_each_entry_continue(policy, entry, bydst) {
 		struct dst_entry *dst;
 
 		write_lock(&policy->lock);
@@ -515,67 +787,112 @@ struct xfrm_policy *xfrm_policy_bysel_ctx(u8 type, int dir,
 					  struct xfrm_selector *sel,
 					  struct xfrm_sec_ctx *ctx, int delete)
 {
-	struct xfrm_policy *pol, **p;
+	struct xfrm_policy *pol, *ret;
+	struct hlist_head *chain;
+	struct hlist_node *entry;
 
 	write_lock_bh(&xfrm_policy_lock);
-	for (p = XFRM_POLICY_LISTHEADP(type, dir); (pol=*p)!=NULL; p = &pol->next) {
-		if ((memcmp(sel, &pol->selector, sizeof(*sel)) == 0) &&
-		    (xfrm_sec_ctx_match(ctx, pol->security))) {
+	chain = policy_hash_bysel(sel, sel->family, dir);
+	ret = NULL;
+	hlist_for_each_entry(pol, entry, chain, bydst) {
+		if (pol->type == type &&
+		    !selector_cmp(sel, &pol->selector) &&
+		    xfrm_sec_ctx_match(ctx, pol->security)) {
 			xfrm_pol_hold(pol);
-			if (delete)
-				*p = pol->next;
+			if (delete) {
+				hlist_del(&pol->bydst);
+				hlist_del(&pol->byidx);
+				xfrm_policy_count[dir]--;
+			}
+			ret = pol;
 			break;
 		}
 	}
 	write_unlock_bh(&xfrm_policy_lock);
 
-	if (pol && delete) {
+	if (ret && delete) {
 		atomic_inc(&flow_cache_genid);
-		xfrm_policy_kill(pol);
+		xfrm_policy_kill(ret);
 	}
-	return pol;
+	return ret;
 }
 EXPORT_SYMBOL(xfrm_policy_bysel_ctx);
 
 struct xfrm_policy *xfrm_policy_byid(u8 type, int dir, u32 id, int delete)
 {
-	struct xfrm_policy *pol, **p;
+	struct xfrm_policy *pol, *ret;
+	struct hlist_head *chain;
+	struct hlist_node *entry;
 
 	write_lock_bh(&xfrm_policy_lock);
-	for (p = XFRM_POLICY_LISTHEADP(type, dir); (pol=*p)!=NULL; p = &pol->next) {
-		if (pol->index == id) {
+	chain = xfrm_policy_byidx + idx_hash(id);
+	ret = NULL;
+	hlist_for_each_entry(pol, entry, chain, byidx) {
+		if (pol->type == type && pol->index == id) {
 			xfrm_pol_hold(pol);
-			if (delete)
-				*p = pol->next;
+			if (delete) {
+				hlist_del(&pol->bydst);
+				hlist_del(&pol->byidx);
+				xfrm_policy_count[dir]--;
+			}
+			ret = pol;
 			break;
 		}
 	}
 	write_unlock_bh(&xfrm_policy_lock);
 
-	if (pol && delete) {
+	if (ret && delete) {
 		atomic_inc(&flow_cache_genid);
-		xfrm_policy_kill(pol);
+		xfrm_policy_kill(ret);
 	}
-	return pol;
+	return ret;
 }
 EXPORT_SYMBOL(xfrm_policy_byid);
 
 void xfrm_policy_flush(u8 type)
 {
-	struct xfrm_policy *xp;
-	struct xfrm_policy **p_list = XFRM_POLICY_LISTS(type);
 	int dir;
 
 	write_lock_bh(&xfrm_policy_lock);
 	for (dir = 0; dir < XFRM_POLICY_MAX; dir++) {
-		while ((xp = p_list[dir]) != NULL) {
-			p_list[dir] = xp->next;
+		struct xfrm_policy *pol;
+		struct hlist_node *entry;
+		int i;
+
+	again1:
+		hlist_for_each_entry(pol, entry,
+				     &xfrm_policy_inexact[dir], bydst) {
+			if (pol->type != type)
+				continue;
+			hlist_del(&pol->bydst);
+			hlist_del(&pol->byidx);
 			write_unlock_bh(&xfrm_policy_lock);
 
-			xfrm_policy_kill(xp);
+			xfrm_policy_kill(pol);
 
 			write_lock_bh(&xfrm_policy_lock);
+			goto again1;
+		}
+
+		for (i = xfrm_policy_bydst[dir].hmask; i >= 0; i--) {
+	again2:
+			hlist_for_each_entry(pol, entry,
+					     xfrm_policy_bydst[dir].table + i,
+					     bydst) {
+				if (pol->type != type)
+					continue;
+				hlist_del(&pol->bydst);
+				hlist_del(&pol->byidx);
+				write_unlock_bh(&xfrm_policy_lock);
+
+				xfrm_policy_kill(pol);
+
+				write_lock_bh(&xfrm_policy_lock);
+				goto again2;
+			}
 		}
+
+		xfrm_policy_count[dir] = 0;
 	}
 	atomic_inc(&flow_cache_genid);
 	write_unlock_bh(&xfrm_policy_lock);
@@ -585,15 +902,27 @@ EXPORT_SYMBOL(xfrm_policy_flush);
 int xfrm_policy_walk(u8 type, int (*func)(struct xfrm_policy *, int, int, void*),
 		     void *data)
 {
-	struct xfrm_policy *xp;
-	int dir;
-	int count = 0;
-	int error = 0;
+	struct xfrm_policy *pol;
+	struct hlist_node *entry;
+	int dir, count, error;
 
 	read_lock_bh(&xfrm_policy_lock);
+	count = 0;
 	for (dir = 0; dir < 2*XFRM_POLICY_MAX; dir++) {
-		for (xp = XFRM_POLICY_LISTHEAD(type, dir); xp; xp = xp->next)
-			count++;
+		struct hlist_head *table = xfrm_policy_bydst[dir].table;
+		int i;
+
+		hlist_for_each_entry(pol, entry,
+				     &xfrm_policy_inexact[dir], bydst) {
+			if (pol->type == type)
+				count++;
+		}
+		for (i = xfrm_policy_bydst[dir].hmask; i >= 0; i--) {
+			hlist_for_each_entry(pol, entry, table + i, bydst) {
+				if (pol->type == type)
+					count++;
+			}
+		}
 	}
 
 	if (count == 0) {
@@ -602,13 +931,28 @@ int xfrm_policy_walk(u8 type, int (*func)(struct xfrm_policy *, int, int, void*)
 	}
 
 	for (dir = 0; dir < 2*XFRM_POLICY_MAX; dir++) {
-		for (xp = XFRM_POLICY_LISTHEAD(type, dir); xp; xp = xp->next) {
-			error = func(xp, dir%XFRM_POLICY_MAX, --count, data);
+		struct hlist_head *table = xfrm_policy_bydst[dir].table;
+		int i;
+
+		hlist_for_each_entry(pol, entry,
+				     &xfrm_policy_inexact[dir], bydst) {
+			if (pol->type != type)
+				continue;
+			error = func(pol, dir % XFRM_POLICY_MAX, --count, data);
 			if (error)
 				goto out;
 		}
+		for (i = xfrm_policy_bydst[dir].hmask; i >= 0; i--) {
+			hlist_for_each_entry(pol, entry, table + i, bydst) {
+				if (pol->type != type)
+					continue;
+				error = func(pol, dir % XFRM_POLICY_MAX, --count, data);
+				if (error)
+					goto out;
+			}
+		}
 	}
-
+	error = 0;
 out:
 	read_unlock_bh(&xfrm_policy_lock);
 	return error;
@@ -617,31 +961,61 @@ EXPORT_SYMBOL(xfrm_policy_walk);
 
 /* Find policy to apply to this flow. */
 
-static struct xfrm_policy *xfrm_policy_lookup_bytype(u8 type, struct flowi *fl,
-						     u16 family, u8 dir)
+static int xfrm_policy_match(struct xfrm_policy *pol, struct flowi *fl,
+			     u8 type, u16 family, int dir)
 {
-	struct xfrm_policy *pol;
+	struct xfrm_selector *sel = &pol->selector;
+	int match;
 
-	read_lock_bh(&xfrm_policy_lock);
-	for (pol = XFRM_POLICY_LISTHEAD(type, dir); pol; pol = pol->next) {
-		struct xfrm_selector *sel = &pol->selector;
-		int match;
+	if (pol->family != family ||
+	    pol->type != type)
+		return 0;
 
-		if (pol->family != family)
-			continue;
+	match = xfrm_selector_match(sel, fl, family);
+	if (match) {
+		if (!security_xfrm_policy_lookup(pol, fl->secid, dir))
+			return 1;
+	}
+
+	return 0;
+}
 
-		match = xfrm_selector_match(sel, fl, family);
+static struct xfrm_policy *xfrm_policy_lookup_bytype(u8 type, struct flowi *fl,
+						     u16 family, u8 dir)
+{
+	struct xfrm_policy *pol, *ret;
+	xfrm_address_t *daddr, *saddr;
+	struct hlist_node *entry;
+	struct hlist_head *chain;
 
-		if (match) {
- 			if (!security_xfrm_policy_lookup(pol, fl->secid, dir)) {
+	daddr = xfrm_flowi_daddr(fl, family);
+	saddr = xfrm_flowi_saddr(fl, family);
+	if (unlikely(!daddr || !saddr))
+		return NULL;
+
+	read_lock_bh(&xfrm_policy_lock);
+	chain = policy_hash_direct(daddr, saddr, family, dir);
+	ret = NULL;
+	hlist_for_each_entry(pol, entry, chain, bydst) {
+		if (xfrm_policy_match(pol, fl, type, family, dir)) {
+			xfrm_pol_hold(pol);
+			ret = pol;
+			break;
+		}
+	}
+	if (!ret) {
+		chain = &xfrm_policy_inexact[dir];
+		hlist_for_each_entry(pol, entry, chain, bydst) {
+			if (xfrm_policy_match(pol, fl, type, family, dir)) {
 				xfrm_pol_hold(pol);
+				ret = pol;
 				break;
 			}
 		}
 	}
 	read_unlock_bh(&xfrm_policy_lock);
 
-	return pol;
+	return ret;
 }
 
 static void xfrm_policy_lookup(struct flowi *fl, u16 family, u8 dir,
@@ -657,7 +1031,7 @@ static void xfrm_policy_lookup(struct flowi *fl, u16 family, u8 dir,
 	pol = xfrm_policy_lookup_bytype(XFRM_POLICY_TYPE_MAIN, fl, family, dir);
 
 #ifdef CONFIG_XFRM_SUB_POLICY
- end:
+end:
 #endif
 	if ((*objp = (void *) pol) != NULL)
 		*obj_refp = &pol->refcnt;
@@ -704,26 +1078,29 @@ static struct xfrm_policy *xfrm_sk_policy_lookup(struct sock *sk, int dir, struc
 
 static void __xfrm_policy_link(struct xfrm_policy *pol, int dir)
 {
-	struct xfrm_policy **p_list = XFRM_POLICY_LISTS(pol->type);
+	struct hlist_head *chain = policy_hash_bysel(&pol->selector,
+						     pol->family, dir);
 
-	pol->next = p_list[dir];
-	p_list[dir] = pol;
+	hlist_add_head(&pol->bydst, chain);
+	hlist_add_head(&pol->byidx, xfrm_policy_byidx+idx_hash(pol->index));
+	xfrm_policy_count[dir]++;
 	xfrm_pol_hold(pol);
+
+	if (xfrm_bydst_should_resize(dir, NULL))
+		schedule_work(&xfrm_hash_work);
 }
 
 static struct xfrm_policy *__xfrm_policy_unlink(struct xfrm_policy *pol,
 						int dir)
 {
-	struct xfrm_policy **polp;
+	if (hlist_unhashed(&pol->bydst))
+		return NULL;
 
-	for (polp = XFRM_POLICY_LISTHEADP(pol->type, dir);
-	     *polp != NULL; polp = &(*polp)->next) {
-		if (*polp == pol) {
-			*polp = pol->next;
-			return pol;
-		}
-	}
-	return NULL;
+	hlist_del(&pol->bydst);
+	hlist_del(&pol->byidx);
+	xfrm_policy_count[dir]--;
+
+	return pol;
 }
 
 int xfrm_policy_delete(struct xfrm_policy *pol, int dir)
@@ -968,7 +1345,8 @@ restart:
 
 	if (!policy) {
 		/* To accelerate a bit...  */
-		if ((dst_orig->flags & DST_NOXFRM) || xfrm_policy_lists_empty(XFRM_POLICY_OUT))
+		if ((dst_orig->flags & DST_NOXFRM) ||
+		    !xfrm_policy_count[XFRM_POLICY_OUT])
 			return 0;
 
 		policy = flow_cache_lookup(fl, dst_orig->ops->family,
@@ -1413,50 +1791,50 @@ static struct dst_entry *xfrm_negative_advice(struct dst_entry *dst)
 	return dst;
 }
 
+static void prune_one_bundle(struct xfrm_policy *pol, int (*func)(struct dst_entry *), struct dst_entry **gc_list_p)
+{
+	struct dst_entry *dst, **dstp;
+
+	write_lock(&pol->lock);
+	dstp = &pol->bundles;
+	while ((dst=*dstp) != NULL) {
+		if (func(dst)) {
+			*dstp = dst->next;
+			dst->next = *gc_list_p;
+			*gc_list_p = dst;
+		} else {
+			dstp = &dst->next;
+		}
+	}
+	write_unlock(&pol->lock);
+}
+
 static void xfrm_prune_bundles(int (*func)(struct dst_entry *))
 {
-	int i;
-	struct xfrm_policy *pol;
-	struct dst_entry *dst, **dstp, *gc_list = NULL;
+	struct dst_entry *gc_list = NULL;
+	int dir;
 
 	read_lock_bh(&xfrm_policy_lock);
-	for (i=0; i<2*XFRM_POLICY_MAX; i++) {
-#ifdef CONFIG_XFRM_SUB_POLICY
-		for (pol = xfrm_policy_list_sub[i]; pol; pol = pol->next) {
-			write_lock(&pol->lock);
-			dstp = &pol->bundles;
-			while ((dst=*dstp) != NULL) {
-				if (func(dst)) {
-					*dstp = dst->next;
-					dst->next = gc_list;
-					gc_list = dst;
-				} else {
-					dstp = &dst->next;
-				}
-			}
-			write_unlock(&pol->lock);
-		}
+	for (dir = 0; dir < XFRM_POLICY_MAX * 2; dir++) {
+		struct xfrm_policy *pol;
+		struct hlist_node *entry;
+		struct hlist_head *table;
+		int i;
 
-#endif
-		for (pol = xfrm_policy_list[i]; pol; pol = pol->next) {
-			write_lock(&pol->lock);
-			dstp = &pol->bundles;
-			while ((dst=*dstp) != NULL) {
-				if (func(dst)) {
-					*dstp = dst->next;
-					dst->next = gc_list;
-					gc_list = dst;
-				} else {
-					dstp = &dst->next;
-				}
-			}
-			write_unlock(&pol->lock);
+		hlist_for_each_entry(pol, entry,
+				     &xfrm_policy_inexact[dir], bydst)
+			prune_one_bundle(pol, func, &gc_list);
+
+		table = xfrm_policy_bydst[dir].table;
+		for (i = xfrm_policy_bydst[dir].hmask; i >= 0; i--) {
+			hlist_for_each_entry(pol, entry, table + i, bydst)
+				prune_one_bundle(pol, func, &gc_list);
 		}
 	}
 	read_unlock_bh(&xfrm_policy_lock);
 
 	while (gc_list) {
-		dst = gc_list;
+		struct dst_entry *dst = gc_list;
 		gc_list = dst->next;
 		dst_free(dst);
 	}
@@ -1680,6 +2058,9 @@ static struct notifier_block xfrm_dev_notifier = {
 
 static void __init xfrm_policy_init(void)
 {
+	unsigned int hmask, sz;
+	int dir;
+
 	xfrm_dst_cache = kmem_cache_create("xfrm_dst_cache",
 					   sizeof(struct xfrm_dst),
 					   0, SLAB_HWCACHE_ALIGN,
@@ -1687,6 +2068,26 @@ static void __init xfrm_policy_init(void)
 	if (!xfrm_dst_cache)
 		panic("XFRM: failed to allocate xfrm_dst_cache\n");
 
+	hmask = 8 - 1;
+	sz = (hmask+1) * sizeof(struct hlist_head);
+
+	xfrm_policy_byidx = xfrm_policy_hash_alloc(sz);
+	xfrm_idx_hmask = hmask;
+	if (!xfrm_policy_byidx)
+		panic("XFRM: failed to allocate byidx hash\n");
+
+	for (dir = 0; dir < XFRM_POLICY_MAX * 2; dir++) {
+		struct xfrm_policy_hash *htab;
+
+		INIT_HLIST_HEAD(&xfrm_policy_inexact[dir]);
+
+		htab = &xfrm_policy_bydst[dir];
+		htab->table = xfrm_policy_hash_alloc(sz);
+		htab->hmask = hmask;
+		if (!htab->table)
+			panic("XFRM: failed to allocate bydst hash\n");
+	}
+
 	INIT_WORK(&xfrm_policy_gc_work, xfrm_policy_gc_task, NULL);
 	register_netdevice_notifier(&xfrm_dev_notifier);
 }
-- 
cgit v1.2.3


From e4bec827feda76d5e7417a2696a75424834d564f Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@davemloft.net>
Date: Fri, 22 Sep 2006 15:17:35 -0700
Subject: [IPSEC] esp: Defer output IV initialization to first use.

First of all, if the xfrm_state only gets used for input
packets this entropy is a complete waste.

Secondly, it is often the case that a configuration loads
many rules (perhaps even dynamically) and they don't all
necessarily ever get used.

This get_random_bytes() call was showing up in the profiles
for xfrm_state inserts which is how I noticed this.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/esp.h | 5 +++--
 net/ipv4/esp4.c   | 9 +++++++--
 net/ipv6/esp6.c   | 9 +++++++--
 3 files changed, 17 insertions(+), 6 deletions(-)

(limited to 'include')

diff --git a/include/net/esp.h b/include/net/esp.h
index 064366d66eea..713d039f4af7 100644
--- a/include/net/esp.h
+++ b/include/net/esp.h
@@ -15,13 +15,14 @@ struct esp_data
 	struct {
 		u8			*key;		/* Key */
 		int			key_len;	/* Key length */
-		u8			*ivec;		/* ivec buffer */
+		int			padlen;		/* 0..255 */
 		/* ivlen is offset from enc_data, where encrypted data start.
 		 * It is logically different of crypto_tfm_alg_ivsize(tfm).
 		 * We assume that it is either zero (no ivec), or
 		 * >= crypto_tfm_alg_ivsize(tfm). */
 		int			ivlen;
-		int			padlen;		/* 0..255 */
+		int			ivinitted;
+		u8			*ivec;		/* ivec buffer */
 		struct crypto_blkcipher	*tfm;		/* crypto handle */
 	} conf;
 
diff --git a/net/ipv4/esp4.c b/net/ipv4/esp4.c
index e87377e1d6b6..13b29360d102 100644
--- a/net/ipv4/esp4.c
+++ b/net/ipv4/esp4.c
@@ -95,8 +95,13 @@ static int esp_output(struct xfrm_state *x, struct sk_buff *skb)
 	esph->seq_no = htonl(++x->replay.oseq);
 	xfrm_aevent_doreplay(x);
 
-	if (esp->conf.ivlen)
+	if (esp->conf.ivlen) {
+		if (unlikely(!esp->conf.ivinitted)) {
+			get_random_bytes(esp->conf.ivec, esp->conf.ivlen);
+			esp->conf.ivinitted = 1;
+		}
 		crypto_blkcipher_set_iv(tfm, esp->conf.ivec, esp->conf.ivlen);
+	}
 
 	do {
 		struct scatterlist *sg = &esp->sgbuf[0];
@@ -378,7 +383,7 @@ static int esp_init_state(struct xfrm_state *x)
 		esp->conf.ivec = kmalloc(esp->conf.ivlen, GFP_KERNEL);
 		if (unlikely(esp->conf.ivec == NULL))
 			goto error;
-		get_random_bytes(esp->conf.ivec, esp->conf.ivlen);
+		esp->conf.ivinitted = 0;
 	}
 	if (crypto_blkcipher_setkey(tfm, esp->conf.key, esp->conf.key_len))
 		goto error;
diff --git a/net/ipv6/esp6.c b/net/ipv6/esp6.c
index ae50b9511151..e78680a9985b 100644
--- a/net/ipv6/esp6.c
+++ b/net/ipv6/esp6.c
@@ -99,8 +99,13 @@ static int esp6_output(struct xfrm_state *x, struct sk_buff *skb)
 	esph->seq_no = htonl(++x->replay.oseq);
 	xfrm_aevent_doreplay(x);
 
-	if (esp->conf.ivlen)
+	if (esp->conf.ivlen) {
+		if (unlikely(!esp->conf.ivinitted)) {
+			get_random_bytes(esp->conf.ivec, esp->conf.ivlen);
+			esp->conf.ivinitted = 1;
+		}
 		crypto_blkcipher_set_iv(tfm, esp->conf.ivec, esp->conf.ivlen);
+	}
 
 	do {
 		struct scatterlist *sg = &esp->sgbuf[0];
@@ -353,7 +358,7 @@ static int esp6_init_state(struct xfrm_state *x)
 		esp->conf.ivec = kmalloc(esp->conf.ivlen, GFP_KERNEL);
 		if (unlikely(esp->conf.ivec == NULL))
 			goto error;
-		get_random_bytes(esp->conf.ivec, esp->conf.ivlen);
+		esp->conf.ivinitted = 0;
 	}
 	if (crypto_blkcipher_setkey(tfm, esp->conf.key, esp->conf.key_len))
 		goto error;
-- 
cgit v1.2.3


From 75bff8f023e02b045a8f68f36fa7da98dca124b8 Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Mon, 21 Aug 2006 19:22:01 +0900
Subject: [IPV6] ROUTE: Routing by FWMARK.

Based on patch by Jean Lorchat <lorchat@sfc.wide.ad.jp>.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/fib_rules.h |  2 +-
 include/net/flow.h        |  2 ++
 net/ipv6/Kconfig          |  7 +++++++
 net/ipv6/fib6_rules.c     | 23 +++++++++++++++++++++++
 net/ipv6/route.c          |  1 +
 5 files changed, 34 insertions(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/linux/fib_rules.h b/include/linux/fib_rules.h
index 19a82b6c1c1f..2987549d6044 100644
--- a/include/linux/fib_rules.h
+++ b/include/linux/fib_rules.h
@@ -34,7 +34,7 @@ enum
 	FRA_UNUSED3,
 	FRA_UNUSED4,
 	FRA_UNUSED5,
-	FRA_FWMARK,	/* netfilter mark (IPv4) */
+	FRA_FWMARK,	/* netfilter mark (IPv4/IPv6) */
 	FRA_FLOW,	/* flow/class id */
 	FRA_UNUSED6,
 	FRA_UNUSED7,
diff --git a/include/net/flow.h b/include/net/flow.h
index e0522914316e..3ca210ec1379 100644
--- a/include/net/flow.h
+++ b/include/net/flow.h
@@ -26,6 +26,7 @@ struct flowi {
 		struct {
 			struct in6_addr		daddr;
 			struct in6_addr		saddr;
+			__u32			fwmark;
 			__u32			flowlabel;
 		} ip6_u;
 
@@ -42,6 +43,7 @@ struct flowi {
 #define fld_scope	nl_u.dn_u.scope
 #define fl6_dst		nl_u.ip6_u.daddr
 #define fl6_src		nl_u.ip6_u.saddr
+#define fl6_fwmark	nl_u.ip6_u.fwmark
 #define fl6_flowlabel	nl_u.ip6_u.flowlabel
 #define fl4_dst		nl_u.ip4_u.daddr
 #define fl4_src		nl_u.ip4_u.saddr
diff --git a/net/ipv6/Kconfig b/net/ipv6/Kconfig
index 21e0cc808f44..a2d211da2aba 100644
--- a/net/ipv6/Kconfig
+++ b/net/ipv6/Kconfig
@@ -173,3 +173,10 @@ config IPV6_MULTIPLE_TABLES
 	---help---
 	  Support multiple routing tables.
 
+config IPV6_ROUTE_FWMARK
+	bool "IPv6: use netfilter MARK value as routing key"
+	depends on IPV6_MULTIPLE_TABLES && NETFILTER
+	---help---
+	  If you say Y here, you will be able to specify different routes for
+	  packets with different mark values (see iptables(8), MARK target).
+
diff --git a/net/ipv6/fib6_rules.c b/net/ipv6/fib6_rules.c
index 91f6233d8efd..aebd9e2b85a8 100644
--- a/net/ipv6/fib6_rules.c
+++ b/net/ipv6/fib6_rules.c
@@ -26,6 +26,9 @@ struct fib6_rule
 	struct fib_rule		common;
 	struct rt6key		src;
 	struct rt6key		dst;
+#ifdef CONFIG_IPV6_ROUTE_FWMARK
+	u8			fwmark;
+#endif
 	u8			tclass;
 };
 
@@ -124,6 +127,11 @@ static int fib6_rule_match(struct fib_rule *rule, struct flowi *fl, int flags)
 	if (r->tclass && r->tclass != ((ntohl(fl->fl6_flowlabel) >> 20) & 0xff))
 		return 0;
 
+#ifdef CONFIG_IPV6_ROUTE_FWMARK
+	if (r->fwmark && (r->fwmark != fl->fl6_fwmark))
+		return 0;
+#endif
+
 	return 1;
 }
 
@@ -164,6 +172,11 @@ static int fib6_rule_configure(struct fib_rule *rule, struct sk_buff *skb,
 		nla_memcpy(&rule6->dst.addr, tb[FRA_DST],
 			   sizeof(struct in6_addr));
 
+#ifdef CONFIG_IPV6_ROUTE_FWMARK
+	if (tb[FRA_FWMARK])
+		rule6->fwmark = nla_get_u32(tb[FRA_FWMARK]);
+#endif
+
 	rule6->src.plen = frh->src_len;
 	rule6->dst.plen = frh->dst_len;
 	rule6->tclass = frh->tos;
@@ -195,6 +208,11 @@ static int fib6_rule_compare(struct fib_rule *rule, struct fib_rule_hdr *frh,
 	    nla_memcmp(tb[FRA_DST], &rule6->dst.addr, sizeof(struct in6_addr)))
 		return 0;
 
+#ifdef CONFIG_IPV6_ROUTE_FWMARK
+	if (tb[FRA_FWMARK] && (rule6->fwmark != nla_get_u32(tb[FRA_FWMARK])))
+		return 0;
+#endif
+
 	return 1;
 }
 
@@ -216,6 +234,11 @@ static int fib6_rule_fill(struct fib_rule *rule, struct sk_buff *skb,
 		NLA_PUT(skb, FRA_SRC, sizeof(struct in6_addr),
 			&rule6->src.addr);
 
+#ifdef CONFIG_IPV6_ROUTE_FWMARK
+	if (rule6->fwmark)
+		NLA_PUT_U32(skb, FRA_FWMARK, rule6->fwmark);
+#endif
+
 	return 0;
 
 nla_put_failure:
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 20691285aee5..649350bd9299 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -703,6 +703,7 @@ void ip6_route_input(struct sk_buff *skb)
 			.ip6_u = {
 				.daddr = iph->daddr,
 				.saddr = iph->saddr,
+				.fwmark = skb->nfmark,
 				.flowlabel = (* (u32 *) iph)&IPV6_FLOWINFO_MASK,
 			},
 		},
-- 
cgit v1.2.3


From 1aaec67f9335a17856dfacdd3e5cc6f4c18faeec Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Sun, 25 Jun 2006 23:54:55 +0900
Subject: [NET]: Add common helper functions to convert IPv6/IPv4 address
 string to network address structure.

These helpers can be used in netfilter, cifs etc.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/inet.h |   2 +
 net/core/utils.c     | 215 +++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 217 insertions(+)

(limited to 'include')

diff --git a/include/linux/inet.h b/include/linux/inet.h
index 6c5587af118d..b7c6da7d6d32 100644
--- a/include/linux/inet.h
+++ b/include/linux/inet.h
@@ -46,5 +46,7 @@
 #include <linux/types.h>
 
 extern __be32 in_aton(const char *str);
+extern int in4_pton(const char *src, int srclen, u8 *dst, char delim, const char **end);
+extern int in6_pton(const char *src, int srclen, u8 *dst, char delim, const char **end);
 #endif
 #endif	/* _LINUX_INET_H */
diff --git a/net/core/utils.c b/net/core/utils.c
index e31c90e05594..5a06e8a72c17 100644
--- a/net/core/utils.c
+++ b/net/core/utils.c
@@ -4,6 +4,7 @@
  *	Authors:
  *	net_random Alan Cox
  *	net_ratelimit Andy Kleen
+ *	in{4,6}_pton YOSHIFUJI Hideaki, Copyright (C)2006 USAGI/WIDE Project
  *
  *	Created by Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
  *
@@ -191,3 +192,217 @@ __be32 in_aton(const char *str)
 }
 
 EXPORT_SYMBOL(in_aton);
+
+#define IN6PTON_XDIGIT		0x00010000
+#define IN6PTON_DIGIT		0x00020000
+#define IN6PTON_COLON_MASK	0x00700000
+#define IN6PTON_COLON_1		0x00100000	/* single : requested */
+#define IN6PTON_COLON_2		0x00200000	/* second : requested */
+#define IN6PTON_COLON_1_2	0x00400000	/* :: requested */
+#define IN6PTON_DOT		0x00800000	/* . */
+#define IN6PTON_DELIM		0x10000000
+#define IN6PTON_NULL		0x20000000	/* first/tail */
+#define IN6PTON_UNKNOWN		0x40000000
+
+static inline int digit2bin(char c, char delim)
+{
+	if (c == delim || c == '\0')
+		return IN6PTON_DELIM;
+	if (c == '.')
+		return IN6PTON_DOT;
+	if (c >= '0' && c <= '9')
+		return (IN6PTON_DIGIT | (c - '0'));
+	return IN6PTON_UNKNOWN;
+}
+
+static inline int xdigit2bin(char c, char delim)
+{
+	if (c == delim || c == '\0')
+		return IN6PTON_DELIM;
+	if (c == ':')
+		return IN6PTON_COLON_MASK;
+	if (c == '.')
+		return IN6PTON_DOT;
+	if (c >= '0' && c <= '9')
+		return (IN6PTON_XDIGIT | IN6PTON_DIGIT| (c - '0'));
+	if (c >= 'a' && c <= 'f')
+		return (IN6PTON_XDIGIT | (c - 'a' + 10));
+	if (c >= 'A' && c <= 'F')
+		return (IN6PTON_XDIGIT | (c - 'A' + 10));
+	return IN6PTON_UNKNOWN;
+}
+
+int in4_pton(const char *src, int srclen,
+	     u8 *dst,
+	     char delim, const char **end)
+{
+	const char *s;
+	u8 *d;
+	u8 dbuf[4];
+	int ret = 0;
+	int i;
+	int w = 0;
+
+	if (srclen < 0)
+		srclen = strlen(src);
+	s = src;
+	d = dbuf;
+	i = 0;
+	while(1) {
+		int c;
+		c = xdigit2bin(srclen > 0 ? *s : '\0', delim);
+		if (!(c & (IN6PTON_DIGIT | IN6PTON_DOT | IN6PTON_DELIM))) {
+			goto out;
+		}
+		if (c & (IN6PTON_DOT | IN6PTON_DELIM)) {
+			if (w == 0)
+				goto out;
+			*d++ = w & 0xff;
+			w = 0;
+			i++;
+			if (c & IN6PTON_DELIM) {
+				if (i != 4)
+					goto out;
+				break;
+			}
+			goto cont;
+		}
+		w = (w * 10) + c;
+		if ((w & 0xffff) > 255) {
+			goto out;
+		}
+cont:
+		if (i >= 4)
+			goto out;
+		s++;
+		srclen--;
+	}
+	ret = 1;
+	memcpy(dst, dbuf, sizeof(dbuf));
+out:
+	if (end)
+		*end = s;
+	return ret;
+}
+
+EXPORT_SYMBOL(in4_pton);
+
+int in6_pton(const char *src, int srclen,
+	     u8 *dst,
+	     char delim, const char **end)
+{
+	const char *s, *tok = NULL;
+	u8 *d, *dc = NULL;
+	u8 dbuf[16];
+	int ret = 0;
+	int i;
+	int state = IN6PTON_COLON_1_2 | IN6PTON_XDIGIT | IN6PTON_NULL;
+	int w = 0;
+
+	memset(dbuf, 0, sizeof(dbuf));
+
+	s = src;
+	d = dbuf;
+	if (srclen < 0)
+		srclen = strlen(src);
+
+	printf("srclen=%d\n", srclen);
+
+	while (1) {
+		int c;
+
+		c = xdigit2bin(srclen > 0 ? *s : '\0', delim);
+		if (!(c & state))
+			goto out;
+		if (c & (IN6PTON_DELIM | IN6PTON_COLON_MASK)) {
+			/* process one 16-bit word */
+			if (!(state & IN6PTON_NULL)) {
+				*d++ = (w >> 8) & 0xff;
+				*d++ = w & 0xff;
+			}
+			w = 0;
+			if (c & IN6PTON_DELIM) {
+				/* We've processed last word */
+				break;
+			}
+			/*
+			 * COLON_1 => XDIGIT
+			 * COLON_2 => XDIGIT|DELIM
+			 * COLON_1_2 => COLON_2
+			 */
+			switch (state & IN6PTON_COLON_MASK) {
+			case IN6PTON_COLON_2:
+				dc = d;
+				state = IN6PTON_XDIGIT | IN6PTON_DELIM;
+				if (dc - dbuf >= sizeof(dbuf))
+					state |= IN6PTON_NULL;
+				break;
+			case IN6PTON_COLON_1|IN6PTON_COLON_1_2:
+				state = IN6PTON_XDIGIT | IN6PTON_COLON_2;
+				break;
+			case IN6PTON_COLON_1:
+				state = IN6PTON_XDIGIT;
+				break;
+			case IN6PTON_COLON_1_2:
+				state = IN6PTON_COLON_2;
+				break;
+			default:
+				state = 0;
+			}
+			tok = s + 1;
+			goto cont;
+		}
+
+		if (c & IN6PTON_DOT) {
+			ret = in4_pton(tok ? tok : s, srclen + (int)(s - tok), d, delim, &s);
+			if (ret > 0) {
+				d += 4;
+				break;
+			}
+			goto out;
+		}
+
+		w = (w << 4) | (0xff & c);
+		state = IN6PTON_COLON_1 | IN6PTON_DELIM;
+		if (!(w & 0xf000)) {
+			state |= IN6PTON_XDIGIT;
+		}
+		if (!dc && d + 2 < dbuf + sizeof(dbuf)) {
+			state |= IN6PTON_COLON_1_2;
+			state &= ~IN6PTON_DELIM;
+		}
+		if (d + 2 >= dbuf + sizeof(dbuf)) {
+			state &= ~(IN6PTON_COLON_1|IN6PTON_COLON_1_2);
+		}
+cont:
+		if ((dc && d + 4 < dbuf + sizeof(dbuf)) ||
+		    d + 4 == dbuf + sizeof(dbuf)) {
+			state |= IN6PTON_DOT;
+		}
+		if (d >= dbuf + sizeof(dbuf)) {
+			state &= ~(IN6PTON_XDIGIT|IN6PTON_COLON_MASK);
+		}
+		s++;
+		srclen--;
+	}
+
+	i = 15; d--;
+
+	if (dc) {
+		while(d >= dc)
+			dst[i--] = *d--;
+		while(i >= dc - dbuf)
+			dst[i--] = 0;
+		while(i >= 0)
+			dst[i--] = *d--;
+	} else
+		memcpy(dst, dbuf, sizeof(dbuf));
+
+	ret = 1;
+out:
+	if (end)
+		*end = s;
+	return ret;
+}
+
+EXPORT_SYMBOL(in6_pton);
-- 
cgit v1.2.3


From bbfb39cbf63829d1db607aa90cbdca557a3a131d Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Fri, 25 Aug 2006 16:10:14 -0700
Subject: [IPV4]: Add support for fwmark masks in routing rules

Add a FRA_FWMASK attributes for fwmark masks. For compatibility a mask of
0xFFFFFFFF is used when a mark value != 0 is sent without a mask.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/fib_rules.h |  3 ++-
 net/ipv4/fib_rules.c      | 21 +++++++++++++++++++--
 2 files changed, 21 insertions(+), 3 deletions(-)

(limited to 'include')

diff --git a/include/linux/fib_rules.h b/include/linux/fib_rules.h
index 2987549d6044..4418c8d9d479 100644
--- a/include/linux/fib_rules.h
+++ b/include/linux/fib_rules.h
@@ -34,12 +34,13 @@ enum
 	FRA_UNUSED3,
 	FRA_UNUSED4,
 	FRA_UNUSED5,
-	FRA_FWMARK,	/* netfilter mark (IPv4/IPv6) */
+	FRA_FWMARK,	/* netfilter mark */
 	FRA_FLOW,	/* flow/class id */
 	FRA_UNUSED6,
 	FRA_UNUSED7,
 	FRA_UNUSED8,
 	FRA_TABLE,	/* Extended table id */
+	FRA_FWMASK,	/* mask for netfilter mark */
 	__FRA_MAX
 };
 
diff --git a/net/ipv4/fib_rules.c b/net/ipv4/fib_rules.c
index ce185ac6f260..280f424ca9c9 100644
--- a/net/ipv4/fib_rules.c
+++ b/net/ipv4/fib_rules.c
@@ -46,6 +46,7 @@ struct fib4_rule
 	u32			dstmask;
 #ifdef CONFIG_IP_ROUTE_FWMARK
 	u32			fwmark;
+	u32			fwmask;
 #endif
 #ifdef CONFIG_NET_CLS_ROUTE
 	u32			tclassid;
@@ -160,7 +161,7 @@ static int fib4_rule_match(struct fib_rule *rule, struct flowi *fl, int flags)
 		return 0;
 
 #ifdef CONFIG_IP_ROUTE_FWMARK
-	if (r->fwmark && (r->fwmark != fl->fl4_fwmark))
+	if ((r->fwmark ^ fl->fl4_fwmark) & r->fwmask)
 		return 0;
 #endif
 
@@ -183,6 +184,7 @@ static struct nla_policy fib4_rule_policy[FRA_MAX+1] __read_mostly = {
 	[FRA_SRC]	= { .type = NLA_U32 },
 	[FRA_DST]	= { .type = NLA_U32 },
 	[FRA_FWMARK]	= { .type = NLA_U32 },
+	[FRA_FWMASK]	= { .type = NLA_U32 },
 	[FRA_FLOW]	= { .type = NLA_U32 },
 	[FRA_TABLE]	= { .type = NLA_U32 },
 };
@@ -219,8 +221,17 @@ static int fib4_rule_configure(struct fib_rule *rule, struct sk_buff *skb,
 		rule4->dst = nla_get_u32(tb[FRA_DST]);
 
 #ifdef CONFIG_IP_ROUTE_FWMARK
-	if (tb[FRA_FWMARK])
+	if (tb[FRA_FWMARK]) {
 		rule4->fwmark = nla_get_u32(tb[FRA_FWMARK]);
+		if (rule4->fwmark)
+			/* compatibility: if the mark value is non-zero all bits
+			 * are compared unless a mask is explicitly specified.
+			 */
+			rule4->fwmask = 0xFFFFFFFF;
+	}
+
+	if (tb[FRA_FWMASK])
+		rule4->fwmask = nla_get_u32(tb[FRA_FWMASK]);
 #endif
 
 #ifdef CONFIG_NET_CLS_ROUTE
@@ -256,6 +267,9 @@ static int fib4_rule_compare(struct fib_rule *rule, struct fib_rule_hdr *frh,
 #ifdef CONFIG_IP_ROUTE_FWMARK
 	if (tb[FRA_FWMARK] && (rule4->fwmark != nla_get_u32(tb[FRA_FWMARK])))
 		return 0;
+
+	if (tb[FRA_FWMASK] && (rule4->fwmask != nla_get_u32(tb[FRA_FWMASK])))
+		return 0;
 #endif
 
 #ifdef CONFIG_NET_CLS_ROUTE
@@ -285,6 +299,9 @@ static int fib4_rule_fill(struct fib_rule *rule, struct sk_buff *skb,
 #ifdef CONFIG_IP_ROUTE_FWMARK
 	if (rule4->fwmark)
 		NLA_PUT_U32(skb, FRA_FWMARK, rule4->fwmark);
+
+	if (rule4->fwmask || rule4->fwmark)
+		NLA_PUT_U32(skb, FRA_FWMASK, rule4->fwmask);
 #endif
 
 	if (rule4->dst_len)
-- 
cgit v1.2.3


From b4e9b520ca5d07a37ea59648e7f50f478e7487a3 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Fri, 25 Aug 2006 16:11:42 -0700
Subject: [NET_SCHED]: Add mask support to fwmark classifier

Support masking the nfmark value before the search. The mask value is
global for all filters contained in one instance. It can only be set
when a new instance is created, all filters must specify the same mask.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/pkt_cls.h |  1 +
 net/sched/cls_fw.c      | 25 ++++++++++++++++++++++++-
 2 files changed, 25 insertions(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/linux/pkt_cls.h b/include/linux/pkt_cls.h
index bd2c5a2bbbf5..c3f01b3085a4 100644
--- a/include/linux/pkt_cls.h
+++ b/include/linux/pkt_cls.h
@@ -305,6 +305,7 @@ enum
 	TCA_FW_POLICE,
 	TCA_FW_INDEV, /*  used by CONFIG_NET_CLS_IND */
 	TCA_FW_ACT, /* used by CONFIG_NET_CLS_ACT */
+	TCA_FW_MASK,
 	__TCA_FW_MAX
 };
 
diff --git a/net/sched/cls_fw.c b/net/sched/cls_fw.c
index e6973d9b686d..e54acc6bcccd 100644
--- a/net/sched/cls_fw.c
+++ b/net/sched/cls_fw.c
@@ -50,6 +50,7 @@
 struct fw_head
 {
 	struct fw_filter *ht[HTSIZE];
+	u32 mask;
 };
 
 struct fw_filter
@@ -101,7 +102,7 @@ static int fw_classify(struct sk_buff *skb, struct tcf_proto *tp,
 	struct fw_filter *f;
 	int r;
 #ifdef CONFIG_NETFILTER
-	u32 id = skb->nfmark;
+	u32 id = skb->nfmark & head->mask;
 #else
 	u32 id = 0;
 #endif
@@ -209,7 +210,9 @@ static int
 fw_change_attrs(struct tcf_proto *tp, struct fw_filter *f,
 	struct rtattr **tb, struct rtattr **tca, unsigned long base)
 {
+	struct fw_head *head = (struct fw_head *)tp->root;
 	struct tcf_exts e;
+	u32 mask;
 	int err;
 
 	err = tcf_exts_validate(tp, tb, tca[TCA_RATE-1], &e, &fw_ext_map);
@@ -232,6 +235,15 @@ fw_change_attrs(struct tcf_proto *tp, struct fw_filter *f,
 	}
 #endif /* CONFIG_NET_CLS_IND */
 
+	if (tb[TCA_FW_MASK-1]) {
+		if (RTA_PAYLOAD(tb[TCA_FW_MASK-1]) != sizeof(u32))
+			goto errout;
+		mask = *(u32*)RTA_DATA(tb[TCA_FW_MASK-1]);
+		if (mask != head->mask)
+			goto errout;
+	} else if (head->mask != 0xFFFFFFFF)
+		goto errout;
+
 	tcf_exts_change(tp, &f->exts, &e);
 
 	return 0;
@@ -267,9 +279,17 @@ static int fw_change(struct tcf_proto *tp, unsigned long base,
 		return -EINVAL;
 
 	if (head == NULL) {
+		u32 mask = 0xFFFFFFFF;
+		if (tb[TCA_FW_MASK-1]) {
+			if (RTA_PAYLOAD(tb[TCA_FW_MASK-1]) != sizeof(u32))
+				return -EINVAL;
+			mask = *(u32*)RTA_DATA(tb[TCA_FW_MASK-1]);
+		}
+
 		head = kzalloc(sizeof(struct fw_head), GFP_KERNEL);
 		if (head == NULL)
 			return -ENOBUFS;
+		head->mask = mask;
 
 		tcf_tree_lock(tp);
 		tp->root = head;
@@ -330,6 +350,7 @@ static void fw_walk(struct tcf_proto *tp, struct tcf_walker *arg)
 static int fw_dump(struct tcf_proto *tp, unsigned long fh,
 		   struct sk_buff *skb, struct tcmsg *t)
 {
+	struct fw_head *head = (struct fw_head *)tp->root;
 	struct fw_filter *f = (struct fw_filter*)fh;
 	unsigned char	 *b = skb->tail;
 	struct rtattr *rta;
@@ -351,6 +372,8 @@ static int fw_dump(struct tcf_proto *tp, unsigned long fh,
 	if (strlen(f->indev))
 		RTA_PUT(skb, TCA_FW_INDEV, IFNAMSIZ, f->indev);
 #endif /* CONFIG_NET_CLS_IND */
+	if (head->mask != 0xFFFFFFFF)
+		RTA_PUT(skb, TCA_FW_MASK, 4, &head->mask);
 
 	if (tcf_exts_dump(skb, &f->exts, &fw_ext_map) < 0)
 		goto rtattr_failure;
-- 
cgit v1.2.3


From 97e5848dd39e7e76bd6077735ebb5473763ab9c5 Mon Sep 17 00:00:00 2001
From: Ian McDonald <ian.mcdonald@jandi.co.nz>
Date: Sat, 26 Aug 2006 19:16:45 -0700
Subject: [DCCP]: Introduce tx buffering

This adds transmit buffering to DCCP.

I have tested with CCID2/3 and with loss and rate limiting.

Signed off by: Ian McDonald <ian.mcdonald@jandi.co.nz>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/dccp.h |  2 ++
 net/dccp/dccp.h      |  2 +-
 net/dccp/output.c    | 90 +++++++++++++++++++++++++++++++++++++---------------
 net/dccp/proto.c     | 16 +++-------
 4 files changed, 73 insertions(+), 37 deletions(-)

(limited to 'include')

diff --git a/include/linux/dccp.h b/include/linux/dccp.h
index 676333b9fad0..2d7671c92c0b 100644
--- a/include/linux/dccp.h
+++ b/include/linux/dccp.h
@@ -438,6 +438,7 @@ struct dccp_ackvec;
  * @dccps_role - Role of this sock, one of %dccp_role
  * @dccps_ndp_count - number of Non Data Packets since last data packet
  * @dccps_hc_rx_ackvec - rx half connection ack vector
+ * @dccps_xmit_timer - timer for when CCID is not ready to send
  */
 struct dccp_sock {
 	/* inet_connection_sock has to be the first member of dccp_sock */
@@ -470,6 +471,7 @@ struct dccp_sock {
 	enum dccp_role			dccps_role:2;
 	__u8				dccps_hc_rx_insert_options:1;
 	__u8				dccps_hc_tx_insert_options:1;
+	struct timer_list		dccps_xmit_timer;
 };
  
 static inline struct dccp_sock *dccp_sk(const struct sock *sk)
diff --git a/net/dccp/dccp.h b/net/dccp/dccp.h
index a5c5475724c0..0a21be437ed3 100644
--- a/net/dccp/dccp.h
+++ b/net/dccp/dccp.h
@@ -130,7 +130,7 @@ extern void dccp_send_delayed_ack(struct sock *sk);
 extern void dccp_send_sync(struct sock *sk, const u64 seq,
 			   const enum dccp_pkt_type pkt_type);
 
-extern int dccp_write_xmit(struct sock *sk, struct sk_buff *skb, long *timeo);
+extern void dccp_write_xmit(struct sock *sk, int block);
 extern void dccp_write_space(struct sock *sk);
 
 extern void dccp_init_xmit_timers(struct sock *sk);
diff --git a/net/dccp/output.c b/net/dccp/output.c
index 58669beee132..7102e3aed4ca 100644
--- a/net/dccp/output.c
+++ b/net/dccp/output.c
@@ -198,7 +198,7 @@ static int dccp_wait_for_ccid(struct sock *sk, struct sk_buff *skb,
 	while (1) {
 		prepare_to_wait(sk->sk_sleep, &wait, TASK_INTERRUPTIBLE);
 
-		if (sk->sk_err || (sk->sk_shutdown & SEND_SHUTDOWN))
+		if (sk->sk_err)
 			goto do_error;
 		if (!*timeo)
 			goto do_nonblock;
@@ -234,37 +234,72 @@ do_interrupted:
 	goto out;
 }
 
-int dccp_write_xmit(struct sock *sk, struct sk_buff *skb, long *timeo)
+static void dccp_write_xmit_timer(unsigned long data) {
+	struct sock *sk = (struct sock *)data;
+	struct dccp_sock *dp = dccp_sk(sk);
+
+	bh_lock_sock(sk);
+	if (sock_owned_by_user(sk))
+		sk_reset_timer(sk, &dp->dccps_xmit_timer, jiffies+1);
+	else
+		dccp_write_xmit(sk, 0);
+	bh_unlock_sock(sk);
+	sock_put(sk);
+}
+
+void dccp_write_xmit(struct sock *sk, int block)
 {
-	const struct dccp_sock *dp = dccp_sk(sk);
-	int err = ccid_hc_tx_send_packet(dp->dccps_hc_tx_ccid, sk, skb,
+	struct dccp_sock *dp = dccp_sk(sk);
+	struct sk_buff *skb;
+	long timeo = 30000; 	/* If a packet is taking longer than 2 secs
+				   we have other issues */
+
+	while ((skb = skb_peek(&sk->sk_write_queue))) {
+		int err = ccid_hc_tx_send_packet(dp->dccps_hc_tx_ccid, sk, skb,
 					 skb->len);
 
-	if (err > 0)
-		err = dccp_wait_for_ccid(sk, skb, timeo);
+		if (err > 0) {
+			if (!block) {
+				sk_reset_timer(sk, &dp->dccps_xmit_timer,
+						msecs_to_jiffies(err)+jiffies);
+				break;
+			} else
+				err = dccp_wait_for_ccid(sk, skb, &timeo);
+			if (err) {
+				printk(KERN_CRIT "%s:err at dccp_wait_for_ccid"
+						 " %d\n", __FUNCTION__, err);
+				dump_stack();
+			}
+		}
 
-	if (err == 0) {
-		struct dccp_skb_cb *dcb = DCCP_SKB_CB(skb);
-		const int len = skb->len;
+		skb_dequeue(&sk->sk_write_queue);
+		if (err == 0) {
+			struct dccp_skb_cb *dcb = DCCP_SKB_CB(skb);
+			const int len = skb->len;
 
-		if (sk->sk_state == DCCP_PARTOPEN) {
-			/* See 8.1.5.  Handshake Completion */
-			inet_csk_schedule_ack(sk);
-			inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK,
+			if (sk->sk_state == DCCP_PARTOPEN) {
+				/* See 8.1.5.  Handshake Completion */
+				inet_csk_schedule_ack(sk);
+				inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK,
 						  inet_csk(sk)->icsk_rto,
 						  DCCP_RTO_MAX);
-			dcb->dccpd_type = DCCP_PKT_DATAACK;
-		} else if (dccp_ack_pending(sk))
-			dcb->dccpd_type = DCCP_PKT_DATAACK;
-		else
-			dcb->dccpd_type = DCCP_PKT_DATA;
-
-		err = dccp_transmit_skb(sk, skb);
-		ccid_hc_tx_packet_sent(dp->dccps_hc_tx_ccid, sk, 0, len);
-	} else
-		kfree_skb(skb);
-
-	return err;
+				dcb->dccpd_type = DCCP_PKT_DATAACK;
+			} else if (dccp_ack_pending(sk))
+				dcb->dccpd_type = DCCP_PKT_DATAACK;
+			else
+				dcb->dccpd_type = DCCP_PKT_DATA;
+
+			err = dccp_transmit_skb(sk, skb);
+			ccid_hc_tx_packet_sent(dp->dccps_hc_tx_ccid, sk, 0, len);
+			if (err) {
+				printk(KERN_CRIT "%s:err from "
+					         "ccid_hc_tx_packet_sent %d\n",
+					         __FUNCTION__, err);
+				dump_stack();
+			}
+		} else
+			kfree(skb);
+	}
 }
 
 int dccp_retransmit_skb(struct sock *sk, struct sk_buff *skb)
@@ -426,6 +461,9 @@ static inline void dccp_connect_init(struct sock *sk)
 	dccp_set_seqno(&dp->dccps_awl, max48(dp->dccps_awl, dp->dccps_iss));
 
 	icsk->icsk_retransmits = 0;
+	init_timer(&dp->dccps_xmit_timer);
+	dp->dccps_xmit_timer.data = (unsigned long)sk;
+	dp->dccps_xmit_timer.function = dccp_write_xmit_timer;
 }
 
 int dccp_connect(struct sock *sk)
@@ -560,8 +598,10 @@ void dccp_send_close(struct sock *sk, const int active)
 					DCCP_PKT_CLOSE : DCCP_PKT_CLOSEREQ;
 
 	if (active) {
+		dccp_write_xmit(sk, 1);
 		dccp_skb_entail(sk, skb);
 		dccp_transmit_skb(sk, skb_clone(skb, prio));
+		/* FIXME do we need a retransmit timer here? */
 	} else
 		dccp_transmit_skb(sk, skb);
 }
diff --git a/net/dccp/proto.c b/net/dccp/proto.c
index 6f14bb5a28d4..962df0ea31aa 100644
--- a/net/dccp/proto.c
+++ b/net/dccp/proto.c
@@ -662,17 +662,8 @@ int dccp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
 	if (rc != 0)
 		goto out_discard;
 
-	rc = dccp_write_xmit(sk, skb, &timeo);
-	/*
-	 * XXX we don't use sk_write_queue, so just discard the packet.
-	 *     Current plan however is to _use_ sk_write_queue with
-	 *     an algorith similar to tcp_sendmsg, where the main difference
-	 *     is that in DCCP we have to respect packet boundaries, so
-	 *     no coalescing of skbs.
-	 *
-	 *     This bug was _quickly_ found & fixed by just looking at an OSTRA
-	 *     generated callgraph 8) -acme
-	 */
+	skb_queue_tail(&sk->sk_write_queue, skb);
+	dccp_write_xmit(sk,0);
 out_release:
 	release_sock(sk);
 	return rc ? : len;
@@ -846,6 +837,7 @@ static int dccp_close_state(struct sock *sk)
 
 void dccp_close(struct sock *sk, long timeout)
 {
+	struct dccp_sock *dp = dccp_sk(sk);
 	struct sk_buff *skb;
 	int state;
 
@@ -862,6 +854,8 @@ void dccp_close(struct sock *sk, long timeout)
 		goto adjudge_to_death;
 	}
 
+	sk_stop_timer(sk, &dp->dccps_xmit_timer);
+
 	/*
 	 * We need to flush the recv. buffs.  We do this only on the
 	 * descriptor close, not protocol-sourced closes, because the
-- 
cgit v1.2.3


From ff5dfe736dd9f6c74b206aa77c0465dfd503bdb9 Mon Sep 17 00:00:00 2001
From: Johannes Berg <johannes@sipsolutions.net>
Date: Sat, 26 Aug 2006 19:17:53 -0700
Subject: [NETLINK]: remove third bogus argument from NLA_PUT_FLAG

This patch removes the 'value' argument from NLA_PUT_FLAG which is
unused anyway. The documentation comment was already correct so it
doesn't need an update :)

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/netlink.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/net/netlink.h b/include/net/netlink.h
index 47044da167c5..bcb27e3a312e 100644
--- a/include/net/netlink.h
+++ b/include/net/netlink.h
@@ -828,7 +828,7 @@ static inline int nla_put_msecs(struct sk_buff *skb, int attrtype,
 #define NLA_PUT_STRING(skb, attrtype, value) \
 	NLA_PUT(skb, attrtype, strlen(value) + 1, value)
 
-#define NLA_PUT_FLAG(skb, attrtype, value) \
+#define NLA_PUT_FLAG(skb, attrtype) \
 	NLA_PUT(skb, attrtype, 0, NULL)
 
 #define NLA_PUT_MSECS(skb, attrtype, jiffies) \
-- 
cgit v1.2.3


From 25030a7f9eeab2dcefff036469e0e2b4f956198f Mon Sep 17 00:00:00 2001
From: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Date: Sat, 26 Aug 2006 20:06:05 -0700
Subject: [UDP]: Unify UDPv4 and UDPv6 ->get_port()

This patch creates one common function which is called by
udp_v4_get_port() and udp_v6_get_port(). As a result,
  * duplicated code is removed
  * udp_port_rover and local port lookup can now be removed from udp.h
  * further savings follow since the same function will be used by UDP-Litev4
    and UDP-Litev6

In contrast to the patch sent in response to Yoshifujis comments
(fixed by this variant), the code below also removes the
EXPORT_SYMBOL(udp_port_rover), since udp_port_rover can now remain
local to net/ipv4/udp.c.

Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/udp.h | 18 ++---------
 net/ipv4/udp.c    | 96 ++++++++++++++++++++++++++++++++++---------------------
 net/ipv6/udp.c    | 76 ++-----------------------------------------
 3 files changed, 64 insertions(+), 126 deletions(-)

(limited to 'include')

diff --git a/include/net/udp.h b/include/net/udp.h
index 766fba1369ce..c490a0f662ac 100644
--- a/include/net/udp.h
+++ b/include/net/udp.h
@@ -30,25 +30,9 @@
 
 #define UDP_HTABLE_SIZE		128
 
-/* udp.c: This needs to be shared by v4 and v6 because the lookup
- *        and hashing code needs to work with different AF's yet
- *        the port space is shared.
- */
 extern struct hlist_head udp_hash[UDP_HTABLE_SIZE];
 extern rwlock_t udp_hash_lock;
 
-extern int udp_port_rover;
-
-static inline int udp_lport_inuse(u16 num)
-{
-	struct sock *sk;
-	struct hlist_node *node;
-
-	sk_for_each(sk, node, &udp_hash[num & (UDP_HTABLE_SIZE - 1)])
-		if (inet_sk(sk)->num == num)
-			return 1;
-	return 0;
-}
 
 /* Note: this must match 'valbool' in sock_setsockopt */
 #define UDP_CSUM_NOXMIT		1
@@ -63,6 +47,8 @@ extern struct proto udp_prot;
 
 struct sk_buff;
 
+extern int	udp_get_port(struct sock *sk, unsigned short snum,
+			     int (*saddr_cmp)(struct sock *, struct sock *));
 extern void	udp_err(struct sk_buff *, u32);
 
 extern int	udp_sendmsg(struct kiocb *iocb, struct sock *sk,
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index 514c1e9ae810..7552b50bcd84 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -118,14 +118,34 @@ DEFINE_SNMP_STAT(struct udp_mib, udp_statistics) __read_mostly;
 struct hlist_head udp_hash[UDP_HTABLE_SIZE];
 DEFINE_RWLOCK(udp_hash_lock);
 
-/* Shared by v4/v6 udp. */
+/* Shared by v4/v6 udp_get_port */
 int udp_port_rover;
 
-static int udp_v4_get_port(struct sock *sk, unsigned short snum)
+static inline int udp_lport_inuse(u16 num)
 {
+	struct sock *sk;
 	struct hlist_node *node;
+
+	sk_for_each(sk, node, &udp_hash[num & (UDP_HTABLE_SIZE - 1)])
+		if (inet_sk(sk)->num == num)
+			return 1;
+	return 0;
+}
+
+/**
+ *  udp_get_port  -  common port lookup for IPv4 and IPv6
+ *
+ *  @sk:          socket struct in question
+ *  @snum:        port number to look up
+ *  @saddr_comp:  AF-dependent comparison of bound local IP addresses
+ */
+int udp_get_port(struct sock *sk, unsigned short snum,
+		 int (*saddr_cmp)(struct sock *sk1, struct sock *sk2))
+{
+	struct hlist_node *node;
+	struct hlist_head *head;
 	struct sock *sk2;
-	struct inet_sock *inet = inet_sk(sk);
+	int    error = 1;
 
 	write_lock_bh(&udp_hash_lock);
 	if (snum == 0) {
@@ -137,11 +157,10 @@ static int udp_v4_get_port(struct sock *sk, unsigned short snum)
 		best_size_so_far = 32767;
 		best = result = udp_port_rover;
 		for (i = 0; i < UDP_HTABLE_SIZE; i++, result++) {
-			struct hlist_head *list;
 			int size;
 
-			list = &udp_hash[result & (UDP_HTABLE_SIZE - 1)];
-			if (hlist_empty(list)) {
+			head = &udp_hash[result & (UDP_HTABLE_SIZE - 1)];
+			if (hlist_empty(head)) {
 				if (result > sysctl_local_port_range[1])
 					result = sysctl_local_port_range[0] +
 						((result - sysctl_local_port_range[0]) &
@@ -149,12 +168,11 @@ static int udp_v4_get_port(struct sock *sk, unsigned short snum)
 				goto gotit;
 			}
 			size = 0;
-			sk_for_each(sk2, node, list)
-				if (++size >= best_size_so_far)
-					goto next;
-			best_size_so_far = size;
-			best = result;
-		next:;
+			sk_for_each(sk2, node, head)
+				if (++size < best_size_so_far) {
+					best_size_so_far = size;
+					best = result;
+				}
 		}
 		result = best;
 		for(i = 0; i < (1 << 16) / UDP_HTABLE_SIZE; i++, result += UDP_HTABLE_SIZE) {
@@ -170,38 +188,44 @@ static int udp_v4_get_port(struct sock *sk, unsigned short snum)
 gotit:
 		udp_port_rover = snum = result;
 	} else {
-		sk_for_each(sk2, node,
-			    &udp_hash[snum & (UDP_HTABLE_SIZE - 1)]) {
-			struct inet_sock *inet2 = inet_sk(sk2);
-
-			if (inet2->num == snum &&
-			    sk2 != sk &&
-			    !ipv6_only_sock(sk2) &&
-			    (!sk2->sk_bound_dev_if ||
-			     !sk->sk_bound_dev_if ||
-			     sk2->sk_bound_dev_if == sk->sk_bound_dev_if) &&
-			    (!inet2->rcv_saddr ||
-			     !inet->rcv_saddr ||
-			     inet2->rcv_saddr == inet->rcv_saddr) &&
-			    (!sk2->sk_reuse || !sk->sk_reuse))
+		head = &udp_hash[snum & (UDP_HTABLE_SIZE - 1)];
+
+		sk_for_each(sk2, node, head)
+			if (inet_sk(sk2)->num == snum                        &&
+			    sk2 != sk                                        &&
+			    (!sk2->sk_reuse        || !sk->sk_reuse)         &&
+			    (!sk2->sk_bound_dev_if || !sk->sk_bound_dev_if
+			     || sk2->sk_bound_dev_if == sk->sk_bound_dev_if) &&
+			    (*saddr_cmp)(sk, sk2)                              )
 				goto fail;
-		}
 	}
-	inet->num = snum;
+	inet_sk(sk)->num = snum;
 	if (sk_unhashed(sk)) {
-		struct hlist_head *h = &udp_hash[snum & (UDP_HTABLE_SIZE - 1)];
-
-		sk_add_node(sk, h);
+		head = &udp_hash[snum & (UDP_HTABLE_SIZE - 1)];
+		sk_add_node(sk, head);
 		sock_prot_inc_use(sk->sk_prot);
 	}
-	write_unlock_bh(&udp_hash_lock);
-	return 0;
-
+	error = 0;
 fail:
 	write_unlock_bh(&udp_hash_lock);
-	return 1;
+	return error;
+}
+
+static inline int  ipv4_rcv_saddr_equal(struct sock *sk1, struct sock *sk2)
+{
+	struct inet_sock *inet1 = inet_sk(sk1), *inet2 = inet_sk(sk2);
+
+	return 	( !ipv6_only_sock(sk2)  &&
+		  (!inet1->rcv_saddr || !inet2->rcv_saddr ||
+		   inet1->rcv_saddr == inet2->rcv_saddr      ));
+}
+
+static inline int udp_v4_get_port(struct sock *sk, unsigned short snum)
+{
+	return udp_get_port(sk, snum, ipv4_rcv_saddr_equal);
 }
 
+
 static void udp_v4_hash(struct sock *sk)
 {
 	BUG();
@@ -1596,7 +1620,7 @@ EXPORT_SYMBOL(udp_disconnect);
 EXPORT_SYMBOL(udp_hash);
 EXPORT_SYMBOL(udp_hash_lock);
 EXPORT_SYMBOL(udp_ioctl);
-EXPORT_SYMBOL(udp_port_rover);
+EXPORT_SYMBOL(udp_get_port);
 EXPORT_SYMBOL(udp_prot);
 EXPORT_SYMBOL(udp_sendmsg);
 EXPORT_SYMBOL(udp_poll);
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index b9cc55ccb000..9662561701d1 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -61,81 +61,9 @@
 
 DEFINE_SNMP_STAT(struct udp_mib, udp_stats_in6) __read_mostly;
 
-/* Grrr, addr_type already calculated by caller, but I don't want
- * to add some silly "cookie" argument to this method just for that.
- */
-static int udp_v6_get_port(struct sock *sk, unsigned short snum)
+static inline int udp_v6_get_port(struct sock *sk, unsigned short snum)
 {
-	struct sock *sk2;
-	struct hlist_node *node;
-
-	write_lock_bh(&udp_hash_lock);
-	if (snum == 0) {
-		int best_size_so_far, best, result, i;
-
-		if (udp_port_rover > sysctl_local_port_range[1] ||
-		    udp_port_rover < sysctl_local_port_range[0])
-			udp_port_rover = sysctl_local_port_range[0];
-		best_size_so_far = 32767;
-		best = result = udp_port_rover;
-		for (i = 0; i < UDP_HTABLE_SIZE; i++, result++) {
-			int size;
-			struct hlist_head *list;
-
-			list = &udp_hash[result & (UDP_HTABLE_SIZE - 1)];
-			if (hlist_empty(list)) {
-				if (result > sysctl_local_port_range[1])
-					result = sysctl_local_port_range[0] +
-						((result - sysctl_local_port_range[0]) &
-						 (UDP_HTABLE_SIZE - 1));
-				goto gotit;
-			}
-			size = 0;
-			sk_for_each(sk2, node, list)
-				if (++size >= best_size_so_far)
-					goto next;
-			best_size_so_far = size;
-			best = result;
-		next:;
-		}
-		result = best;
-		for(i = 0; i < (1 << 16) / UDP_HTABLE_SIZE; i++, result += UDP_HTABLE_SIZE) {
-			if (result > sysctl_local_port_range[1])
-				result = sysctl_local_port_range[0]
-					+ ((result - sysctl_local_port_range[0]) &
-					   (UDP_HTABLE_SIZE - 1));
-			if (!udp_lport_inuse(result))
-				break;
-		}
-		if (i >= (1 << 16) / UDP_HTABLE_SIZE)
-			goto fail;
-gotit:
-		udp_port_rover = snum = result;
-	} else {
-		sk_for_each(sk2, node,
-			    &udp_hash[snum & (UDP_HTABLE_SIZE - 1)]) {
-			if (inet_sk(sk2)->num == snum &&
-			    sk2 != sk &&
-			    (!sk2->sk_bound_dev_if ||
-			     !sk->sk_bound_dev_if ||
-			     sk2->sk_bound_dev_if == sk->sk_bound_dev_if) &&
-			    (!sk2->sk_reuse || !sk->sk_reuse) &&
-			    ipv6_rcv_saddr_equal(sk, sk2))
-				goto fail;
-		}
-	}
-
-	inet_sk(sk)->num = snum;
-	if (sk_unhashed(sk)) {
-		sk_add_node(sk, &udp_hash[snum & (UDP_HTABLE_SIZE - 1)]);
-		sock_prot_inc_use(sk->sk_prot);
-	}
-	write_unlock_bh(&udp_hash_lock);
-	return 0;
-
-fail:
-	write_unlock_bh(&udp_hash_lock);
-	return 1;
+	return udp_get_port(sk, snum, ipv6_rcv_saddr_equal);
 }
 
 static void udp_v6_hash(struct sock *sk)
-- 
cgit v1.2.3


From e3b4eadbea77ecb3c3a74d1bc81b392f454c7f2e Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@sunset.davemloft.net>
Date: Sat, 26 Aug 2006 20:10:15 -0700
Subject: [UDP]: saddr_cmp function should take const socket pointers
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This also kills a warning while building ipv6:

net/ipv6/udp.c: In function ‘udp_v6_get_port’:
net/ipv6/udp.c:66: warning: passing argument 3 of ‘udp_get_port’ from incompatible pointer type

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/udp.h | 2 +-
 net/ipv4/udp.c    | 4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)

(limited to 'include')

diff --git a/include/net/udp.h b/include/net/udp.h
index c490a0f662ac..db0c05f67546 100644
--- a/include/net/udp.h
+++ b/include/net/udp.h
@@ -48,7 +48,7 @@ extern struct proto udp_prot;
 struct sk_buff;
 
 extern int	udp_get_port(struct sock *sk, unsigned short snum,
-			     int (*saddr_cmp)(struct sock *, struct sock *));
+			     int (*saddr_cmp)(const struct sock *, const struct sock *));
 extern void	udp_err(struct sk_buff *, u32);
 
 extern int	udp_sendmsg(struct kiocb *iocb, struct sock *sk,
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index aa1823050b00..77e265d7bb8f 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -139,7 +139,7 @@ static inline int udp_lport_inuse(u16 num)
  *  @saddr_comp:  AF-dependent comparison of bound local IP addresses
  */
 int udp_get_port(struct sock *sk, unsigned short snum,
-		 int (*saddr_cmp)(struct sock *sk1, struct sock *sk2))
+		 int (*saddr_cmp)(const struct sock *sk1, const struct sock *sk2))
 {
 	struct hlist_node *node;
 	struct hlist_head *head;
@@ -210,7 +210,7 @@ fail:
 	return error;
 }
 
-static inline int  ipv4_rcv_saddr_equal(struct sock *sk1, struct sock *sk2)
+static inline int ipv4_rcv_saddr_equal(const struct sock *sk1, const struct sock *sk2)
 {
 	struct inet_sock *inet1 = inet_sk(sk1), *inet2 = inet_sk(sk2);
 
-- 
cgit v1.2.3


From a5531a5d852008be40811496029012f4ad3093d1 Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Sat, 26 Aug 2006 20:11:47 -0700
Subject: [NETLINK]: Improve string attribute validation

Introduces a new attribute type NLA_NUL_STRING to support NUL
terminated strings. Attributes of this kind require to carry
a terminating NUL within the maximum specified in the policy.

The `old' NLA_STRING which is not required to be NUL terminated
is extended to provide means to specify a maximum length of the
string.

Aims at easing the pain with using nla_strlcpy() on temporary
buffers.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/netlink.h | 15 +++++++++++----
 net/netlink/attr.c    | 49 +++++++++++++++++++++++++++++++++++++++----------
 2 files changed, 50 insertions(+), 14 deletions(-)

(limited to 'include')

diff --git a/include/net/netlink.h b/include/net/netlink.h
index bcb27e3a312e..11dc2e7f679a 100644
--- a/include/net/netlink.h
+++ b/include/net/netlink.h
@@ -167,6 +167,7 @@ enum {
 	NLA_FLAG,
 	NLA_MSECS,
 	NLA_NESTED,
+	NLA_NUL_STRING,
 	__NLA_TYPE_MAX,
 };
 
@@ -175,21 +176,27 @@ enum {
 /**
  * struct nla_policy - attribute validation policy
  * @type: Type of attribute or NLA_UNSPEC
- * @minlen: Minimal length of payload required to be available
+ * @len: Type specific length of payload
  *
  * Policies are defined as arrays of this struct, the array must be
  * accessible by attribute type up to the highest identifier to be expected.
  *
+ * Meaning of `len' field:
+ *    NLA_STRING           Maximum length of string
+ *    NLA_NUL_STRING       Maximum length of string (excluding NUL)
+ *    NLA_FLAG             Unused
+ *    All other            Exact length of attribute payload
+ *
  * Example:
  * static struct nla_policy my_policy[ATTR_MAX+1] __read_mostly = {
  * 	[ATTR_FOO] = { .type = NLA_U16 },
- *	[ATTR_BAR] = { .type = NLA_STRING },
- *	[ATTR_BAZ] = { .minlen = sizeof(struct mystruct) },
+ *	[ATTR_BAR] = { .type = NLA_STRING, len = BARSIZ },
+ *	[ATTR_BAZ] = { .len = sizeof(struct mystruct) },
  * };
  */
 struct nla_policy {
 	u16		type;
-	u16		minlen;
+	u16		len;
 };
 
 /**
diff --git a/net/netlink/attr.c b/net/netlink/attr.c
index 136e529e5780..004139557e09 100644
--- a/net/netlink/attr.c
+++ b/net/netlink/attr.c
@@ -20,7 +20,6 @@ static u16 nla_attr_minlen[NLA_TYPE_MAX+1] __read_mostly = {
 	[NLA_U16]	= sizeof(u16),
 	[NLA_U32]	= sizeof(u32),
 	[NLA_U64]	= sizeof(u64),
-	[NLA_STRING]	= 1,
 	[NLA_NESTED]	= NLA_HDRLEN,
 };
 
@@ -28,7 +27,7 @@ static int validate_nla(struct nlattr *nla, int maxtype,
 			struct nla_policy *policy)
 {
 	struct nla_policy *pt;
-	int minlen = 0;
+	int minlen = 0, attrlen = nla_len(nla);
 
 	if (nla->nla_type <= 0 || nla->nla_type > maxtype)
 		return 0;
@@ -37,16 +36,46 @@ static int validate_nla(struct nlattr *nla, int maxtype,
 
 	BUG_ON(pt->type > NLA_TYPE_MAX);
 
-	if (pt->minlen)
-		minlen = pt->minlen;
-	else if (pt->type != NLA_UNSPEC)
-		minlen = nla_attr_minlen[pt->type];
+	switch (pt->type) {
+	case NLA_FLAG:
+		if (attrlen > 0)
+			return -ERANGE;
+		break;
 
-	if (pt->type == NLA_FLAG && nla_len(nla) > 0)
-		return -ERANGE;
+	case NLA_NUL_STRING:
+		if (pt->len)
+			minlen = min_t(int, attrlen, pt->len + 1);
+		else
+			minlen = attrlen;
 
-	if (nla_len(nla) < minlen)
-		return -ERANGE;
+		if (!minlen || memchr(nla_data(nla), '\0', minlen) == NULL)
+			return -EINVAL;
+		/* fall through */
+
+	case NLA_STRING:
+		if (attrlen < 1)
+			return -ERANGE;
+
+		if (pt->len) {
+			char *buf = nla_data(nla);
+
+			if (buf[attrlen - 1] == '\0')
+				attrlen--;
+
+			if (attrlen > pt->len)
+				return -ERANGE;
+		}
+		break;
+
+	default:
+		if (pt->len)
+			minlen = pt->len;
+		else if (pt->type != NLA_UNSPEC)
+			minlen = nla_attr_minlen[pt->type];
+
+		if (attrlen < minlen)
+			return -ERANGE;
+	}
 
 	return 0;
 }
-- 
cgit v1.2.3


From def42ff4dd6f54ebcf78192579a8ff1f81d8e2e8 Mon Sep 17 00:00:00 2001
From: Alexey Dobriyan <adobriyan@gmail.com>
Date: Mon, 28 Aug 2006 23:57:56 -0700
Subject: [IPV4]: Make struct in_addr::s_addr __be32

There will be relatively small increase in sparse endian warnings, but
this (and sin_port) patch is a first step to make networking code
endian clean.

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/in.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/linux/in.h b/include/linux/in.h
index 94f557fa4636..9a9d5dd32e73 100644
--- a/include/linux/in.h
+++ b/include/linux/in.h
@@ -52,7 +52,7 @@ enum {
 
 /* Internet address. */
 struct in_addr {
-	__u32	s_addr;
+	__be32	s_addr;
 };
 
 #define IP_TOS		1
-- 
cgit v1.2.3


From cd360007a0eb8cbf17c006cca42aa884d33f96be Mon Sep 17 00:00:00 2001
From: Alexey Dobriyan <adobriyan@gmail.com>
Date: Mon, 28 Aug 2006 23:58:32 -0700
Subject: [IPV4]: Make struct sockaddr_in::sin_port __be16

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/in.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/linux/in.h b/include/linux/in.h
index 9a9d5dd32e73..bcaca8399aed 100644
--- a/include/linux/in.h
+++ b/include/linux/in.h
@@ -177,7 +177,7 @@ struct in_pktinfo
 #define __SOCK_SIZE__	16		/* sizeof(struct sockaddr)	*/
 struct sockaddr_in {
   sa_family_t		sin_family;	/* Address family		*/
-  unsigned short int	sin_port;	/* Port number			*/
+  __be16		sin_port;	/* Port number			*/
   struct in_addr	sin_addr;	/* Internet address		*/
 
   /* Pad to size of `struct sockaddr'. */
-- 
cgit v1.2.3


From 07317621d004e8e6967f2dac8562825267e56135 Mon Sep 17 00:00:00 2001
From: Stephen Hemminger <shemminger@osdl.org>
Date: Tue, 29 Aug 2006 17:48:17 -0700
Subject: [NETFILTER] bridge: code rearrangement for clarity

Cleanup and rearrangement for better style and clarity:
	Split the function nf_bridge_maybe_copy_header into two pieces
	Move copy portion out of line.
	Use Ethernet header size macros.
	Use header file to handle CONFIG_NETFILTER_BRIDGE differences

Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter_bridge.h | 26 +++++++-------------------
 net/bridge/br_forward.c          |  5 +----
 net/bridge/br_netfilter.c        | 27 +++++++++++++++++++++++++--
 3 files changed, 33 insertions(+), 25 deletions(-)

(limited to 'include')

diff --git a/include/linux/netfilter_bridge.h b/include/linux/netfilter_bridge.h
index 427c67ff89e9..274fe4b33155 100644
--- a/include/linux/netfilter_bridge.h
+++ b/include/linux/netfilter_bridge.h
@@ -47,26 +47,12 @@ enum nf_br_hook_priorities {
 
 
 /* Only used in br_forward.c */
-static inline
-int nf_bridge_maybe_copy_header(struct sk_buff *skb)
+extern int nf_bridge_copy_header(struct sk_buff *skb);
+static inline int nf_bridge_maybe_copy_header(struct sk_buff *skb)
 {
-	int err;
-
-	if (skb->nf_bridge) {
-		if (skb->protocol == __constant_htons(ETH_P_8021Q)) {
-			err = skb_cow(skb, 18);
-			if (err)
-				return err;
-			memcpy(skb->data - 18, skb->nf_bridge->data, 18);
-			skb_push(skb, 4);
-		} else {
-			err = skb_cow(skb, 16);
-			if (err)
-				return err;
-			memcpy(skb->data - 16, skb->nf_bridge->data, 16);
-		}
-	}
-	return 0;
+	if (skb->nf_bridge)
+		return nf_bridge_copy_header(skb);
+  	return 0;
 }
 
 /* This is called by the IP fragmenting code and it ensures there is
@@ -90,6 +76,8 @@ struct bridge_skb_cb {
 };
 
 extern int brnf_deferred_hooks;
+#else
+#define nf_bridge_maybe_copy_header(skb)	(0)
 #endif /* CONFIG_BRIDGE_NETFILTER */
 
 #endif /* __KERNEL__ */
diff --git a/net/bridge/br_forward.c b/net/bridge/br_forward.c
index 864fbbc7b24d..191b861e5e53 100644
--- a/net/bridge/br_forward.c
+++ b/net/bridge/br_forward.c
@@ -38,13 +38,10 @@ int br_dev_queue_push_xmit(struct sk_buff *skb)
 	if (packet_length(skb) > skb->dev->mtu && !skb_is_gso(skb))
 		kfree_skb(skb);
 	else {
-#ifdef CONFIG_BRIDGE_NETFILTER
 		/* ip_refrag calls ip_fragment, doesn't copy the MAC header. */
 		if (nf_bridge_maybe_copy_header(skb))
 			kfree_skb(skb);
-		else
-#endif
-		{
+		else {
 			skb_push(skb, ETH_HLEN);
 
 			dev_queue_xmit(skb);
diff --git a/net/bridge/br_netfilter.c b/net/bridge/br_netfilter.c
index 05b3de888243..b498efcfe451 100644
--- a/net/bridge/br_netfilter.c
+++ b/net/bridge/br_netfilter.c
@@ -127,14 +127,37 @@ static inline struct nf_bridge_info *nf_bridge_alloc(struct sk_buff *skb)
 
 static inline void nf_bridge_save_header(struct sk_buff *skb)
 {
-        int header_size = 16;
+        int header_size = ETH_HLEN;
 
 	if (skb->protocol == htons(ETH_P_8021Q))
-		header_size = 18;
+		header_size += VLAN_HLEN;
 
 	memcpy(skb->nf_bridge->data, skb->data - header_size, header_size);
 }
 
+/*
+ * When forwarding bridge frames, we save a copy of the original
+ * header before processing.
+ */
+int nf_bridge_copy_header(struct sk_buff *skb)
+{
+	int err;
+        int header_size = ETH_HLEN;
+
+	if (skb->protocol == htons(ETH_P_8021Q))
+		header_size += VLAN_HLEN;
+
+	err = skb_cow(skb, header_size);
+	if (err)
+		return err;
+
+	memcpy(skb->data - header_size, skb->nf_bridge->data, header_size);
+
+	if (skb->protocol == htons(ETH_P_8021Q))
+		__skb_push(skb, VLAN_HLEN);
+	return 0;
+}
+
 /* PF_BRIDGE/PRE_ROUTING *********************************************/
 /* Undo the changes made for ip6tables PREROUTING and continue the
  * bridge PRE_ROUTING hook. */
-- 
cgit v1.2.3


From 9bcfcaf5e9cc887eb39236e43bdbe4b4b2572229 Mon Sep 17 00:00:00 2001
From: Stephen Hemminger <shemminger@osdl.org>
Date: Tue, 29 Aug 2006 17:48:57 -0700
Subject: [NETFILTER] bridge: simplify nf_bridge_pad

Do some simple optimization on the nf_bridge_pad() function
and don't use magic constants. Eliminate a double call and
the #ifdef'd code for CONFIG_BRIDGE_NETFILTER.

Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter_bridge.h | 16 +++++-----------
 net/ipv4/ip_output.c             | 15 +++++++--------
 2 files changed, 12 insertions(+), 19 deletions(-)

(limited to 'include')

diff --git a/include/linux/netfilter_bridge.h b/include/linux/netfilter_bridge.h
index 274fe4b33155..9a4dd11af86e 100644
--- a/include/linux/netfilter_bridge.h
+++ b/include/linux/netfilter_bridge.h
@@ -5,9 +5,8 @@
  */
 
 #include <linux/netfilter.h>
-#if defined(__KERNEL__) && defined(CONFIG_BRIDGE_NETFILTER)
 #include <linux/if_ether.h>
-#endif
+#include <linux/if_vlan.h>
 
 /* Bridge Hooks */
 /* After promisc drops, checksum checks. */
@@ -57,16 +56,10 @@ static inline int nf_bridge_maybe_copy_header(struct sk_buff *skb)
 
 /* This is called by the IP fragmenting code and it ensures there is
  * enough room for the encapsulating header (if there is one). */
-static inline
-int nf_bridge_pad(struct sk_buff *skb)
+static inline int nf_bridge_pad(const struct sk_buff *skb)
 {
-	if (skb->protocol == __constant_htons(ETH_P_IP))
-		return 0;
-	if (skb->nf_bridge) {
-		if (skb->protocol == __constant_htons(ETH_P_8021Q))
-			return 4;
-	}
-	return 0;
+ 	return (skb->nf_bridge && skb->protocol == htons(ETH_P_8021Q))
+		? VLAN_HLEN : 0;
 }
 
 struct bridge_skb_cb {
@@ -78,6 +71,7 @@ struct bridge_skb_cb {
 extern int brnf_deferred_hooks;
 #else
 #define nf_bridge_maybe_copy_header(skb)	(0)
+#define nf_bridge_pad(skb)			(0)
 #endif /* CONFIG_BRIDGE_NETFILTER */
 
 #endif /* __KERNEL__ */
diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index 81b2795a4c20..97aee76fb746 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -426,7 +426,7 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff*))
 	int ptr;
 	struct net_device *dev;
 	struct sk_buff *skb2;
-	unsigned int mtu, hlen, left, len, ll_rs;
+	unsigned int mtu, hlen, left, len, ll_rs, pad;
 	int offset;
 	__be16 not_last_frag;
 	struct rtable *rt = (struct rtable*)skb->dst;
@@ -556,14 +556,13 @@ slow_path:
 	left = skb->len - hlen;		/* Space per frame */
 	ptr = raw + hlen;		/* Where to start from */
 
-#ifdef CONFIG_BRIDGE_NETFILTER
 	/* for bridged IP traffic encapsulated inside f.e. a vlan header,
-	 * we need to make room for the encapsulating header */
-	ll_rs = LL_RESERVED_SPACE_EXTRA(rt->u.dst.dev, nf_bridge_pad(skb));
-	mtu -= nf_bridge_pad(skb);
-#else
-	ll_rs = LL_RESERVED_SPACE(rt->u.dst.dev);
-#endif
+	 * we need to make room for the encapsulating header
+	 */
+	pad = nf_bridge_pad(skb);
+	ll_rs = LL_RESERVED_SPACE_EXTRA(rt->u.dst.dev, pad);
+	mtu -= pad;
+
 	/*
 	 *	Fragment the datagram.
 	 */
-- 
cgit v1.2.3


From 1b7f775209bbee6b993587bae69acb9fc12ceb17 Mon Sep 17 00:00:00 2001
From: Paul Moore <paul.moore@hp.com>
Date: Tue, 29 Aug 2006 17:54:17 -0700
Subject: [NetLabel]: remove unused function prototypes

Removed some older function prototypes for functions that no longer exist.

Signed-off-by: Paul Moore <paul.moore@hp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/cipso_ipv4.h | 6 ------
 1 file changed, 6 deletions(-)

(limited to 'include')

diff --git a/include/net/cipso_ipv4.h b/include/net/cipso_ipv4.h
index c7175e725804..5aed72ab652b 100644
--- a/include/net/cipso_ipv4.h
+++ b/include/net/cipso_ipv4.h
@@ -200,15 +200,9 @@ static inline int cipso_v4_cache_add(const struct sk_buff *skb,
 
 #ifdef CONFIG_NETLABEL
 void cipso_v4_error(struct sk_buff *skb, int error, u32 gateway);
-int cipso_v4_socket_setopt(struct socket *sock,
-			   unsigned char *opt,
-			   u32 opt_len);
 int cipso_v4_socket_setattr(const struct socket *sock,
 			    const struct cipso_v4_doi *doi_def,
 			    const struct netlbl_lsm_secattr *secattr);
-int cipso_v4_socket_getopt(const struct socket *sock,
-			   unsigned char **opt,
-			   u32 *opt_len);
 int cipso_v4_socket_getattr(const struct socket *sock,
 			    struct netlbl_lsm_secattr *secattr);
 int cipso_v4_skbuff_getattr(const struct sk_buff *skb,
-- 
cgit v1.2.3


From 7a0e1d602288370801c353221c6a938eab925053 Mon Sep 17 00:00:00 2001
From: Paul Moore <paul.moore@hp.com>
Date: Tue, 29 Aug 2006 17:56:04 -0700
Subject: [NetLabel]: add some missing #includes to various header files

Add some missing include files to the NetLabel related header files.

Signed-off-by: Paul Moore <paul.moore@hp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/cipso_ipv4.h                    | 2 ++
 include/net/netlabel.h                      | 1 +
 net/netlabel/netlabel_domainhash.h          | 4 ++++
 net/netlabel/netlabel_user.h                | 5 +++--
 security/selinux/include/selinux_netlabel.h | 9 +++++++++
 5 files changed, 19 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/include/net/cipso_ipv4.h b/include/net/cipso_ipv4.h
index 5aed72ab652b..59406e0dc5b2 100644
--- a/include/net/cipso_ipv4.h
+++ b/include/net/cipso_ipv4.h
@@ -37,6 +37,8 @@
 #include <linux/types.h>
 #include <linux/rcupdate.h>
 #include <linux/list.h>
+#include <linux/net.h>
+#include <linux/skbuff.h>
 #include <net/netlabel.h>
 
 /* known doi values */
diff --git a/include/net/netlabel.h b/include/net/netlabel.h
index 7cae730832c7..fc2b72fc7e07 100644
--- a/include/net/netlabel.h
+++ b/include/net/netlabel.h
@@ -31,6 +31,7 @@
 #define _NETLABEL_H
 
 #include <linux/types.h>
+#include <linux/net.h>
 #include <linux/skbuff.h>
 #include <net/netlink.h>
 
diff --git a/net/netlabel/netlabel_domainhash.h b/net/netlabel/netlabel_domainhash.h
index 9217863ce0d3..99a2287de246 100644
--- a/net/netlabel/netlabel_domainhash.h
+++ b/net/netlabel/netlabel_domainhash.h
@@ -32,6 +32,10 @@
 #ifndef _NETLABEL_DOMAINHASH_H
 #define _NETLABEL_DOMAINHASH_H
 
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/list.h>
+
 /* Domain hash table size */
 /* XXX - currently this number is an uneducated guess */
 #define NETLBL_DOMHSH_BITSIZE       7
diff --git a/net/netlabel/netlabel_user.h b/net/netlabel/netlabel_user.h
index ccf237b3a128..385a6c7488c6 100644
--- a/net/netlabel/netlabel_user.h
+++ b/net/netlabel/netlabel_user.h
@@ -31,11 +31,12 @@
 #ifndef _NETLABEL_USER_H
 #define _NETLABEL_USER_H
 
+#include <linux/types.h>
 #include <linux/skbuff.h>
 #include <linux/capability.h>
-#include <linux/genetlink.h>
-#include <net/netlabel.h>
+#include <net/netlink.h>
 #include <net/genetlink.h>
+#include <net/netlabel.h>
 
 /* NetLabel NETLINK helper functions */
 
diff --git a/security/selinux/include/selinux_netlabel.h b/security/selinux/include/selinux_netlabel.h
index d69ec650cdbe..ecab4bddaaf4 100644
--- a/security/selinux/include/selinux_netlabel.h
+++ b/security/selinux/include/selinux_netlabel.h
@@ -27,6 +27,15 @@
 #ifndef _SELINUX_NETLABEL_H_
 #define _SELINUX_NETLABEL_H_
 
+#include <linux/types.h>
+#include <linux/fs.h>
+#include <linux/net.h>
+#include <linux/skbuff.h>
+#include <net/sock.h>
+
+#include "avc.h"
+#include "objsec.h"
+
 #ifdef CONFIG_NETLABEL
 void selinux_netlbl_cache_invalidate(void);
 int selinux_netlbl_socket_post_create(struct socket *sock,
-- 
cgit v1.2.3


From fda9ef5d679b07c9d9097aaf6ef7f069d794a8f9 Mon Sep 17 00:00:00 2001
From: Dmitry Mishin <dim@openvz.org>
Date: Thu, 31 Aug 2006 15:28:39 -0700
Subject: [NET]: Fix sk->sk_filter field access

Function sk_filter() is called from tcp_v{4,6}_rcv() functions with arg
needlock = 0, while socket is not locked at that moment. In order to avoid
this and similar issues in the future, use rcu for sk->sk_filter field read
protection.

Signed-off-by: Dmitry Mishin <dim@openvz.org>
Signed-off-by: Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
Signed-off-by: Kirill Korotaev <dev@openvz.org>
---
 include/linux/filter.h | 13 +++++++------
 include/net/sock.h     | 34 +++++++++++++++++-----------------
 net/core/filter.c      |  8 ++++----
 net/core/sock.c        | 22 +++++++++-------------
 net/dccp/ipv6.c        |  2 +-
 net/decnet/dn_nsp_in.c |  2 +-
 net/ipv4/tcp_ipv4.c    |  2 +-
 net/ipv6/tcp_ipv6.c    |  4 ++--
 net/packet/af_packet.c | 43 ++++++++++++++++++-------------------------
 net/sctp/input.c       |  2 +-
 10 files changed, 61 insertions(+), 71 deletions(-)

(limited to 'include')

diff --git a/include/linux/filter.h b/include/linux/filter.h
index c6cb8f095088..91b2e3b9251e 100644
--- a/include/linux/filter.h
+++ b/include/linux/filter.h
@@ -25,10 +25,10 @@
  
 struct sock_filter	/* Filter block */
 {
-        __u16	code;   /* Actual filter code */
-        __u8	jt;	/* Jump true */
-        __u8	jf;	/* Jump false */
-        __u32	k;      /* Generic multiuse field */
+	__u16	code;   /* Actual filter code */
+	__u8	jt;	/* Jump true */
+	__u8	jf;	/* Jump false */
+	__u32	k;      /* Generic multiuse field */
 };
 
 struct sock_fprog	/* Required for SO_ATTACH_FILTER. */
@@ -41,8 +41,9 @@ struct sock_fprog	/* Required for SO_ATTACH_FILTER. */
 struct sk_filter
 {
 	atomic_t		refcnt;
-        unsigned int         	len;	/* Number of filter blocks */
-        struct sock_filter     	insns[0];
+	unsigned int         	len;	/* Number of filter blocks */
+	struct rcu_head		rcu;
+	struct sock_filter     	insns[0];
 };
 
 static inline unsigned int sk_filter_len(struct sk_filter *fp)
diff --git a/include/net/sock.h b/include/net/sock.h
index 337ebec84c70..edd4d73ce7f5 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -862,30 +862,24 @@ extern void sock_init_data(struct socket *sock, struct sock *sk);
  *
  */
 
-static inline int sk_filter(struct sock *sk, struct sk_buff *skb, int needlock)
+static inline int sk_filter(struct sock *sk, struct sk_buff *skb)
 {
 	int err;
+	struct sk_filter *filter;
 	
 	err = security_sock_rcv_skb(sk, skb);
 	if (err)
 		return err;
 	
-	if (sk->sk_filter) {
-		struct sk_filter *filter;
-		
-		if (needlock)
-			bh_lock_sock(sk);
-		
-		filter = sk->sk_filter;
-		if (filter) {
-			unsigned int pkt_len = sk_run_filter(skb, filter->insns,
-							     filter->len);
-			err = pkt_len ? pskb_trim(skb, pkt_len) : -EPERM;
-		}
-
-		if (needlock)
-			bh_unlock_sock(sk);
+	rcu_read_lock_bh();
+	filter = sk->sk_filter;
+	if (filter) {
+		unsigned int pkt_len = sk_run_filter(skb, filter->insns,
+				filter->len);
+		err = pkt_len ? pskb_trim(skb, pkt_len) : -EPERM;
 	}
+ 	rcu_read_unlock_bh();
+
 	return err;
 }
 
@@ -897,6 +891,12 @@ static inline int sk_filter(struct sock *sk, struct sk_buff *skb, int needlock)
  *	Remove a filter from a socket and release its resources.
  */
  
+static inline void sk_filter_rcu_free(struct rcu_head *rcu)
+{
+	struct sk_filter *fp = container_of(rcu, struct sk_filter, rcu);
+	kfree(fp);
+}
+
 static inline void sk_filter_release(struct sock *sk, struct sk_filter *fp)
 {
 	unsigned int size = sk_filter_len(fp);
@@ -904,7 +904,7 @@ static inline void sk_filter_release(struct sock *sk, struct sk_filter *fp)
 	atomic_sub(size, &sk->sk_omem_alloc);
 
 	if (atomic_dec_and_test(&fp->refcnt))
-		kfree(fp);
+		call_rcu_bh(&fp->rcu, sk_filter_rcu_free);
 }
 
 static inline void sk_filter_charge(struct sock *sk, struct sk_filter *fp)
diff --git a/net/core/filter.c b/net/core/filter.c
index 5b4486a60cf6..6732782a5a40 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -422,10 +422,10 @@ int sk_attach_filter(struct sock_fprog *fprog, struct sock *sk)
 	if (!err) {
 		struct sk_filter *old_fp;
 
-		spin_lock_bh(&sk->sk_lock.slock);
-		old_fp = sk->sk_filter;
-		sk->sk_filter = fp;
-		spin_unlock_bh(&sk->sk_lock.slock);
+		rcu_read_lock_bh();
+		old_fp = rcu_dereference(sk->sk_filter);
+		rcu_assign_pointer(sk->sk_filter, fp);
+		rcu_read_unlock_bh();
 		fp = old_fp;
 	}
 
diff --git a/net/core/sock.c b/net/core/sock.c
index cfaf09039b02..b77e155cbe6c 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -247,11 +247,7 @@ int sock_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
 		goto out;
 	}
 
-	/* It would be deadlock, if sock_queue_rcv_skb is used
-	   with socket lock! We assume that users of this
-	   function are lock free.
-	*/
-	err = sk_filter(sk, skb, 1);
+	err = sk_filter(sk, skb);
 	if (err)
 		goto out;
 
@@ -278,7 +274,7 @@ int sk_receive_skb(struct sock *sk, struct sk_buff *skb)
 {
 	int rc = NET_RX_SUCCESS;
 
-	if (sk_filter(sk, skb, 0))
+	if (sk_filter(sk, skb))
 		goto discard_and_relse;
 
 	skb->dev = NULL;
@@ -606,15 +602,15 @@ set_rcvbuf:
 			break;
 
 		case SO_DETACH_FILTER:
-			spin_lock_bh(&sk->sk_lock.slock);
-			filter = sk->sk_filter;
+			rcu_read_lock_bh();
+			filter = rcu_dereference(sk->sk_filter);
                         if (filter) {
-				sk->sk_filter = NULL;
-				spin_unlock_bh(&sk->sk_lock.slock);
+				rcu_assign_pointer(sk->sk_filter, NULL);
 				sk_filter_release(sk, filter);
+				rcu_read_unlock_bh();
 				break;
 			}
-			spin_unlock_bh(&sk->sk_lock.slock);
+			rcu_read_unlock_bh();
 			ret = -ENONET;
 			break;
 
@@ -884,10 +880,10 @@ void sk_free(struct sock *sk)
 	if (sk->sk_destruct)
 		sk->sk_destruct(sk);
 
-	filter = sk->sk_filter;
+	filter = rcu_dereference(sk->sk_filter);
 	if (filter) {
 		sk_filter_release(sk, filter);
-		sk->sk_filter = NULL;
+		rcu_assign_pointer(sk->sk_filter, NULL);
 	}
 
 	sock_disable_timestamp(sk);
diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c
index f9c5e12d7038..7a47399cf31f 100644
--- a/net/dccp/ipv6.c
+++ b/net/dccp/ipv6.c
@@ -970,7 +970,7 @@ static int dccp_v6_do_rcv(struct sock *sk, struct sk_buff *skb)
 	if (skb->protocol == htons(ETH_P_IP))
 		return dccp_v4_do_rcv(sk, skb);
 
-	if (sk_filter(sk, skb, 0))
+	if (sk_filter(sk, skb))
 		goto discard;
 
 	/*
diff --git a/net/decnet/dn_nsp_in.c b/net/decnet/dn_nsp_in.c
index 86f7f3b28e70..72ecc6e62ec4 100644
--- a/net/decnet/dn_nsp_in.c
+++ b/net/decnet/dn_nsp_in.c
@@ -586,7 +586,7 @@ static __inline__ int dn_queue_skb(struct sock *sk, struct sk_buff *skb, int sig
         	goto out;
         }
 
-	err = sk_filter(sk, skb, 0);
+	err = sk_filter(sk, skb);
 	if (err)
 		goto out;
 
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 23b46e36b147..39b179856082 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1104,7 +1104,7 @@ process:
 		goto discard_and_relse;
 	nf_reset(skb);
 
-	if (sk_filter(sk, skb, 0))
+	if (sk_filter(sk, skb))
 		goto discard_and_relse;
 
 	skb->dev = NULL;
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 2b18918f3011..2546fc9f0a78 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -1075,7 +1075,7 @@ static int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb)
 	if (skb->protocol == htons(ETH_P_IP))
 		return tcp_v4_do_rcv(sk, skb);
 
-	if (sk_filter(sk, skb, 0))
+	if (sk_filter(sk, skb))
 		goto discard;
 
 	/*
@@ -1232,7 +1232,7 @@ process:
 	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
 		goto discard_and_relse;
 
-	if (sk_filter(sk, skb, 0))
+	if (sk_filter(sk, skb))
 		goto discard_and_relse;
 
 	skb->dev = NULL;
diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
index 300215bdbf46..f4ccb90e6739 100644
--- a/net/packet/af_packet.c
+++ b/net/packet/af_packet.c
@@ -427,21 +427,24 @@ out_unlock:
 }
 #endif
 
-static inline unsigned run_filter(struct sk_buff *skb, struct sock *sk, unsigned res)
+static inline int run_filter(struct sk_buff *skb, struct sock *sk,
+							unsigned *snaplen)
 {
 	struct sk_filter *filter;
+	int err = 0;
 
-	bh_lock_sock(sk);
-	filter = sk->sk_filter;
-	/*
-	 * Our caller already checked that filter != NULL but we need to
-	 * verify that under bh_lock_sock() to be safe
-	 */
-	if (likely(filter != NULL))
-		res = sk_run_filter(skb, filter->insns, filter->len);
-	bh_unlock_sock(sk);
+	rcu_read_lock_bh();
+	filter = rcu_dereference(sk->sk_filter);
+	if (filter != NULL) {
+		err = sk_run_filter(skb, filter->insns, filter->len);
+		if (!err)
+			err = -EPERM;
+		else if (*snaplen > err)
+			*snaplen = err;
+	}
+	rcu_read_unlock_bh();
 
-	return res;
+	return err;
 }
 
 /*
@@ -491,13 +494,8 @@ static int packet_rcv(struct sk_buff *skb, struct net_device *dev, struct packet
 
 	snaplen = skb->len;
 
-	if (sk->sk_filter) {
-		unsigned res = run_filter(skb, sk, snaplen);
-		if (res == 0)
-			goto drop_n_restore;
-		if (snaplen > res)
-			snaplen = res;
-	}
+	if (run_filter(skb, sk, &snaplen) < 0)
+		goto drop_n_restore;
 
 	if (atomic_read(&sk->sk_rmem_alloc) + skb->truesize >=
 	    (unsigned)sk->sk_rcvbuf)
@@ -593,13 +591,8 @@ static int tpacket_rcv(struct sk_buff *skb, struct net_device *dev, struct packe
 
 	snaplen = skb->len;
 
-	if (sk->sk_filter) {
-		unsigned res = run_filter(skb, sk, snaplen);
-		if (res == 0)
-			goto drop_n_restore;
-		if (snaplen > res)
-			snaplen = res;
-	}
+	if (run_filter(skb, sk, &snaplen) < 0)
+		goto drop_n_restore;
 
 	if (sk->sk_type == SOCK_DGRAM) {
 		macoff = netoff = TPACKET_ALIGN(TPACKET_HDRLEN) + 16;
diff --git a/net/sctp/input.c b/net/sctp/input.c
index 8a34d95602ce..03f65de75d88 100644
--- a/net/sctp/input.c
+++ b/net/sctp/input.c
@@ -228,7 +228,7 @@ int sctp_rcv(struct sk_buff *skb)
 		goto discard_release;
 	nf_reset(skb);
 
-	if (sk_filter(sk, skb, 1))
+	if (sk_filter(sk, skb))
                 goto discard_release;
 
 	/* Create an SCTP packet structure. */
-- 
cgit v1.2.3


From eb878e84575fbce21d2edb079eada78bfa27023d Mon Sep 17 00:00:00 2001
From: Jamal Hadi Salim <hadi@cyberus.ca>
Date: Thu, 31 Aug 2006 17:42:59 -0700
Subject: [IPSEC]: output mode to take an xfrm state as input param

Expose IPSEC modes output path to take an xfrm state as input param.
This makes it consistent with the input mode processing (which already
takes the xfrm state as a param).

Signed-off-by: Jamal Hadi Salim <hadi@cyberus.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h              | 2 +-
 net/ipv4/xfrm4_mode_transport.c | 4 +---
 net/ipv4/xfrm4_mode_tunnel.c    | 3 +--
 net/ipv4/xfrm4_output.c         | 2 +-
 net/ipv6/xfrm6_mode_ro.c        | 3 +--
 net/ipv6/xfrm6_mode_transport.c | 3 +--
 net/ipv6/xfrm6_mode_tunnel.c    | 3 +--
 net/ipv6/xfrm6_output.c         | 2 +-
 8 files changed, 8 insertions(+), 14 deletions(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 0acabf2a0a8f..4d6dc627df9b 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -285,7 +285,7 @@ extern void xfrm_put_type(struct xfrm_type *type);
 
 struct xfrm_mode {
 	int (*input)(struct xfrm_state *x, struct sk_buff *skb);
-	int (*output)(struct sk_buff *skb);
+	int (*output)(struct xfrm_state *x,struct sk_buff *skb);
 
 	struct module *owner;
 	unsigned int encap;
diff --git a/net/ipv4/xfrm4_mode_transport.c b/net/ipv4/xfrm4_mode_transport.c
index a9e6b3dd19c9..92676b7e4034 100644
--- a/net/ipv4/xfrm4_mode_transport.c
+++ b/net/ipv4/xfrm4_mode_transport.c
@@ -21,9 +21,8 @@
  * On exit, skb->h will be set to the start of the payload to be processed
  * by x->type->output and skb->nh will be set to the top IP header.
  */
-static int xfrm4_transport_output(struct sk_buff *skb)
+static int xfrm4_transport_output(struct xfrm_state *x, struct sk_buff *skb)
 {
-	struct xfrm_state *x;
 	struct iphdr *iph;
 	int ihl;
 
@@ -33,7 +32,6 @@ static int xfrm4_transport_output(struct sk_buff *skb)
 	ihl = iph->ihl * 4;
 	skb->h.raw += ihl;
 
-	x = skb->dst->xfrm;
 	skb->nh.raw = memmove(skb_push(skb, x->props.header_len), iph, ihl);
 	return 0;
 }
diff --git a/net/ipv4/xfrm4_mode_tunnel.c b/net/ipv4/xfrm4_mode_tunnel.c
index 13cafbe56ce3..e23c21d31a53 100644
--- a/net/ipv4/xfrm4_mode_tunnel.c
+++ b/net/ipv4/xfrm4_mode_tunnel.c
@@ -33,10 +33,9 @@ static inline void ipip_ecn_decapsulate(struct sk_buff *skb)
  * On exit, skb->h will be set to the start of the payload to be processed
  * by x->type->output and skb->nh will be set to the top IP header.
  */
-static int xfrm4_tunnel_output(struct sk_buff *skb)
+static int xfrm4_tunnel_output(struct xfrm_state *x, struct sk_buff *skb)
 {
 	struct dst_entry *dst = skb->dst;
-	struct xfrm_state *x = dst->xfrm;
 	struct iphdr *iph, *top_iph;
 	int flags;
 
diff --git a/net/ipv4/xfrm4_output.c b/net/ipv4/xfrm4_output.c
index 5fd115f0c547..04403fb01a58 100644
--- a/net/ipv4/xfrm4_output.c
+++ b/net/ipv4/xfrm4_output.c
@@ -66,7 +66,7 @@ static int xfrm4_output_one(struct sk_buff *skb)
 		if (err)
 			goto error;
 
-		err = x->mode->output(skb);
+		err = x->mode->output(x, skb);
 		if (err)
 			goto error;
 
diff --git a/net/ipv6/xfrm6_mode_ro.c b/net/ipv6/xfrm6_mode_ro.c
index c11c335312f9..6031c16d46ca 100644
--- a/net/ipv6/xfrm6_mode_ro.c
+++ b/net/ipv6/xfrm6_mode_ro.c
@@ -43,9 +43,8 @@
  * its absence, that of the top IP header.  The value of skb->data will always
  * point to the top IP header.
  */
-static int xfrm6_ro_output(struct sk_buff *skb)
+static int xfrm6_ro_output(struct xfrm_state *x, struct sk_buff *skb)
 {
-	struct xfrm_state *x = skb->dst->xfrm;
 	struct ipv6hdr *iph;
 	u8 *prevhdr;
 	int hdr_len;
diff --git a/net/ipv6/xfrm6_mode_transport.c b/net/ipv6/xfrm6_mode_transport.c
index a5dce216024d..3a4b39b12bad 100644
--- a/net/ipv6/xfrm6_mode_transport.c
+++ b/net/ipv6/xfrm6_mode_transport.c
@@ -25,9 +25,8 @@
  * its absence, that of the top IP header.  The value of skb->data will always
  * point to the top IP header.
  */
-static int xfrm6_transport_output(struct sk_buff *skb)
+static int xfrm6_transport_output(struct xfrm_state *x, struct sk_buff *skb)
 {
-	struct xfrm_state *x = skb->dst->xfrm;
 	struct ipv6hdr *iph;
 	u8 *prevhdr;
 	int hdr_len;
diff --git a/net/ipv6/xfrm6_mode_tunnel.c b/net/ipv6/xfrm6_mode_tunnel.c
index 8af79be2edca..5e7d8a7d6414 100644
--- a/net/ipv6/xfrm6_mode_tunnel.c
+++ b/net/ipv6/xfrm6_mode_tunnel.c
@@ -37,10 +37,9 @@ static inline void ipip6_ecn_decapsulate(struct sk_buff *skb)
  * its absence, that of the top IP header.  The value of skb->data will always
  * point to the top IP header.
  */
-static int xfrm6_tunnel_output(struct sk_buff *skb)
+static int xfrm6_tunnel_output(struct xfrm_state *x, struct sk_buff *skb)
 {
 	struct dst_entry *dst = skb->dst;
-	struct xfrm_state *x = dst->xfrm;
 	struct ipv6hdr *iph, *top_iph;
 	int dsfield;
 
diff --git a/net/ipv6/xfrm6_output.c b/net/ipv6/xfrm6_output.c
index db58104e710b..c260ea104c52 100644
--- a/net/ipv6/xfrm6_output.c
+++ b/net/ipv6/xfrm6_output.c
@@ -65,7 +65,7 @@ static int xfrm6_output_one(struct sk_buff *skb)
 		if (err)
 			goto error;
 
-		err = x->mode->output(skb);
+		err = x->mode->output(x, skb);
 		if (err)
 			goto error;
 
-- 
cgit v1.2.3


From eb328111efde7bca782f340fe805756039ec6a0c Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Mon, 18 Sep 2006 00:01:59 -0700
Subject: [GENL]: Provide more information to userspace about registered genl
 families

Additionaly exports the following information when providing
the list of registered generic netlink families:
  - protocol version
  - header size
  - maximum number of attributes
  - list of available operations including
      - id
      - flags
      - avaiability of policy and doit/dumpit function

libnl HEAD provides a utility to read this new information:

	0x0010 nlctrl version 1
	    hdrsize 0 maxattr 6
	      op GETFAMILY (0x03) [POLICY,DOIT,DUMPIT]
	0x0011 NLBL_MGMT version 1
	    hdrsize 0 maxattr 0
	      op unknown (0x02) [DOIT]
	      op unknown (0x03) [DOIT]
	      ....

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/genetlink.h | 18 ++++++++++++++++++
 include/net/genetlink.h   |  2 --
 net/netlink/genetlink.c   | 40 ++++++++++++++++++++++++++++++++++++++++
 3 files changed, 58 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/include/linux/genetlink.h b/include/linux/genetlink.h
index 84f12a41dc01..9049dc65ae51 100644
--- a/include/linux/genetlink.h
+++ b/include/linux/genetlink.h
@@ -16,6 +16,8 @@ struct genlmsghdr {
 
 #define GENL_HDRLEN	NLMSG_ALIGN(sizeof(struct genlmsghdr))
 
+#define GENL_ADMIN_PERM		0x01
+
 /*
  * List of reserved static generic netlink identifiers:
  */
@@ -43,9 +45,25 @@ enum {
 	CTRL_ATTR_UNSPEC,
 	CTRL_ATTR_FAMILY_ID,
 	CTRL_ATTR_FAMILY_NAME,
+	CTRL_ATTR_VERSION,
+	CTRL_ATTR_HDRSIZE,
+	CTRL_ATTR_MAXATTR,
+	CTRL_ATTR_OPS,
 	__CTRL_ATTR_MAX,
 };
 
 #define CTRL_ATTR_MAX (__CTRL_ATTR_MAX - 1)
 
+enum {
+	CTRL_ATTR_OP_UNSPEC,
+	CTRL_ATTR_OP_ID,
+	CTRL_ATTR_OP_FLAGS,
+	CTRL_ATTR_OP_POLICY,
+	CTRL_ATTR_OP_DOIT,
+	CTRL_ATTR_OP_DUMPIT,
+	__CTRL_ATTR_OP_MAX,
+};
+
+#define CTRL_ATTR_OP_MAX (__CTRL_ATTR_OP_MAX - 1)
+
 #endif	/* __LINUX_GENERIC_NETLINK_H */
diff --git a/include/net/genetlink.h b/include/net/genetlink.h
index 97d6d3aba9d2..4a38d85e4e25 100644
--- a/include/net/genetlink.h
+++ b/include/net/genetlink.h
@@ -27,8 +27,6 @@ struct genl_family
 	struct list_head	family_list;	/* private */
 };
 
-#define GENL_ADMIN_PERM		0x01
-
 /**
  * struct genl_info - receiving information
  * @snd_seq: sending sequence number
diff --git a/net/netlink/genetlink.c b/net/netlink/genetlink.c
index 3ac942cdb677..49bc2db7982b 100644
--- a/net/netlink/genetlink.c
+++ b/net/netlink/genetlink.c
@@ -387,7 +387,10 @@ static void genl_rcv(struct sock *sk, int len)
 static int ctrl_fill_info(struct genl_family *family, u32 pid, u32 seq,
 			  u32 flags, struct sk_buff *skb, u8 cmd)
 {
+	struct nlattr *nla_ops;
+	struct genl_ops *ops;
 	void *hdr;
+	int idx = 1;
 
 	hdr = genlmsg_put(skb, pid, seq, GENL_ID_CTRL, 0, flags, cmd,
 			  family->version);
@@ -396,6 +399,37 @@ static int ctrl_fill_info(struct genl_family *family, u32 pid, u32 seq,
 
 	NLA_PUT_STRING(skb, CTRL_ATTR_FAMILY_NAME, family->name);
 	NLA_PUT_U16(skb, CTRL_ATTR_FAMILY_ID, family->id);
+	NLA_PUT_U32(skb, CTRL_ATTR_VERSION, family->version);
+	NLA_PUT_U32(skb, CTRL_ATTR_HDRSIZE, family->hdrsize);
+	NLA_PUT_U32(skb, CTRL_ATTR_MAXATTR, family->maxattr);
+
+	nla_ops = nla_nest_start(skb, CTRL_ATTR_OPS);
+	if (nla_ops == NULL)
+		goto nla_put_failure;
+
+	list_for_each_entry(ops, &family->ops_list, ops_list) {
+		struct nlattr *nest;
+
+		nest = nla_nest_start(skb, idx++);
+		if (nest == NULL)
+			goto nla_put_failure;
+
+		NLA_PUT_U32(skb, CTRL_ATTR_OP_ID, ops->cmd);
+		NLA_PUT_U32(skb, CTRL_ATTR_OP_FLAGS, ops->flags);
+
+		if (ops->policy)
+			NLA_PUT_FLAG(skb, CTRL_ATTR_OP_POLICY);
+
+		if (ops->doit)
+			NLA_PUT_FLAG(skb, CTRL_ATTR_OP_DOIT);
+
+		if (ops->dumpit)
+			NLA_PUT_FLAG(skb, CTRL_ATTR_OP_DUMPIT);
+
+		nla_nest_end(skb, nest);
+	}
+
+	nla_nest_end(skb, nla_ops);
 
 	return genlmsg_end(skb, hdr);
 
@@ -411,6 +445,9 @@ static int ctrl_dumpfamily(struct sk_buff *skb, struct netlink_callback *cb)
 	int chains_to_skip = cb->args[0];
 	int fams_to_skip = cb->args[1];
 
+	if (chains_to_skip != 0)
+		genl_lock();
+
 	for (i = 0; i < GENL_FAM_TAB_SIZE; i++) {
 		if (i < chains_to_skip)
 			continue;
@@ -428,6 +465,9 @@ static int ctrl_dumpfamily(struct sk_buff *skb, struct netlink_callback *cb)
 	}
 
 errout:
+	if (chains_to_skip != 0)
+		genl_unlock();
+
 	cb->args[0] = i;
 	cb->args[1] = n;
 
-- 
cgit v1.2.3


From 161643660129dd7d98f0b12418c0a2710ffa7db6 Mon Sep 17 00:00:00 2001
From: Adrian Bunk <bunk@stusta.de>
Date: Mon, 18 Sep 2006 00:40:38 -0700
Subject: [SCTP]: Cleanups

This patch contains the following cleanups:
- make the following needlessly global function static:
  - socket.c: sctp_apply_peer_addr_params()
- add proper prototypes for the several global functions in
  include/net/sctp/sctp.h

Note that this fixes wrong prototypes for the following functions:
- sctp_snmp_proc_exit()
- sctp_eps_proc_exit()
- sctp_assocs_proc_exit()

The latter was spotted by the GNU C compiler and reported
by David Woodhouse.

Signed-off-by: Adrian Bunk <bunk@stusta.de>
Acked-by: Sridhar Samudrala <sri@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/sctp/sctp.h | 13 +++++++++++++
 net/sctp/ipv6.c         |  1 -
 net/sctp/protocol.c     |  7 -------
 net/sctp/socket.c       | 14 +++++++-------
 4 files changed, 20 insertions(+), 15 deletions(-)

(limited to 'include')

diff --git a/include/net/sctp/sctp.h b/include/net/sctp/sctp.h
index e274fd479990..ee68a3124076 100644
--- a/include/net/sctp/sctp.h
+++ b/include/net/sctp/sctp.h
@@ -128,6 +128,8 @@ extern int sctp_copy_local_addr_list(struct sctp_bind_addr *,
 				     int flags);
 extern struct sctp_pf *sctp_get_pf_specific(sa_family_t family);
 extern int sctp_register_pf(struct sctp_pf *, sa_family_t);
+int sctp_inetaddr_event(struct notifier_block *this, unsigned long ev,
+                        void *ptr);
 
 /*
  * sctp/socket.c
@@ -177,6 +179,17 @@ void sctp_icmp_proto_unreachable(struct sock *sk,
 void sctp_backlog_migrate(struct sctp_association *assoc,
 			  struct sock *oldsk, struct sock *newsk);
 
+/*
+ * sctp/proc.c
+ */
+int sctp_snmp_proc_init(void);
+void sctp_snmp_proc_exit(void);
+int sctp_eps_proc_init(void);
+void sctp_eps_proc_exit(void);
+int sctp_assocs_proc_init(void);
+void sctp_assocs_proc_exit(void);
+
+
 /*
  *  Section:  Macros, externs, and inlines
  */
diff --git a/net/sctp/ipv6.c b/net/sctp/ipv6.c
index 99c0cefc04e0..fd87e3ceb56e 100644
--- a/net/sctp/ipv6.c
+++ b/net/sctp/ipv6.c
@@ -78,7 +78,6 @@
 
 #include <asm/uaccess.h>
 
-extern int sctp_inetaddr_event(struct notifier_block *, unsigned long, void *);
 static struct notifier_block sctp_inet6addr_notifier = {
 	.notifier_call = sctp_inetaddr_event,
 };
diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c
index d9dd4c47bc29..fac7674438a4 100644
--- a/net/sctp/protocol.c
+++ b/net/sctp/protocol.c
@@ -82,13 +82,6 @@ static struct sctp_af *sctp_af_v6_specific;
 kmem_cache_t *sctp_chunk_cachep __read_mostly;
 kmem_cache_t *sctp_bucket_cachep __read_mostly;
 
-extern int sctp_snmp_proc_init(void);
-extern int sctp_snmp_proc_exit(void);
-extern int sctp_eps_proc_init(void);
-extern int sctp_eps_proc_exit(void);
-extern int sctp_assocs_proc_init(void);
-extern int sctp_assocs_proc_exit(void);
-
 /* Return the address of the control sock. */
 struct sock *sctp_get_ctl_sock(void)
 {
diff --git a/net/sctp/socket.c b/net/sctp/socket.c
index 7c1dbb1d10df..79c3e072cf28 100644
--- a/net/sctp/socket.c
+++ b/net/sctp/socket.c
@@ -2081,13 +2081,13 @@ static int sctp_setsockopt_autoclose(struct sock *sk, char __user *optval,
  *                     SPP_SACKDELAY_ENABLE, setting both will have undefined
  *                     results.
  */
-int sctp_apply_peer_addr_params(struct sctp_paddrparams *params,
-				struct sctp_transport   *trans,
-				struct sctp_association *asoc,
-				struct sctp_sock        *sp,
-				int                      hb_change,
-				int                      pmtud_change,
-				int                      sackdelay_change)
+static int sctp_apply_peer_addr_params(struct sctp_paddrparams *params,
+				       struct sctp_transport   *trans,
+				       struct sctp_association *asoc,
+				       struct sctp_sock        *sp,
+				       int                      hb_change,
+				       int                      pmtud_change,
+				       int                      sackdelay_change)
 {
 	int error;
 
-- 
cgit v1.2.3


From 1ef9696c909060ccdae3ade245ca88692b49285b Mon Sep 17 00:00:00 2001
From: Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
Date: Tue, 19 Sep 2006 12:52:50 -0700
Subject: [TCP]: Send ACKs each 2nd received segment.

It does not affect either mss-sized connections (obviously) or
connections controlled by Nagle (because there is only one small
segment in flight).

The idea is to record the fact that a small segment arrives on a
connection, where one small segment has already been received and
still not-ACKed. In this case ACK is forced after tcp_recvmsg() drains
receive buffer.

In other words, it is a "soft" each-2nd-segment ACK, which is enough
to preserve ACK clock even when ABC is enabled.

Signed-off-by: Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/inet_connection_sock.h | 3 ++-
 net/ipv4/tcp.c                     | 7 +++++--
 net/ipv4/tcp_input.c               | 2 ++
 3 files changed, 9 insertions(+), 3 deletions(-)

(limited to 'include')

diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h
index 9bf73fe50948..de4e83b6da4b 100644
--- a/include/net/inet_connection_sock.h
+++ b/include/net/inet_connection_sock.h
@@ -147,7 +147,8 @@ extern struct sock *inet_csk_clone(struct sock *sk,
 enum inet_csk_ack_state_t {
 	ICSK_ACK_SCHED	= 1,
 	ICSK_ACK_TIMER  = 2,
-	ICSK_ACK_PUSHED = 4
+	ICSK_ACK_PUSHED = 4,
+	ICSK_ACK_PUSHED2 = 8
 };
 
 extern void inet_csk_init_xmit_timers(struct sock *sk,
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 29e3d606db78..66e9a729f6df 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -955,8 +955,11 @@ void tcp_cleanup_rbuf(struct sock *sk, int copied)
 		     * receive buffer and there was a small segment
 		     * in queue.
 		     */
-		    (copied > 0 && (icsk->icsk_ack.pending & ICSK_ACK_PUSHED) &&
-		     !icsk->icsk_ack.pingpong && !atomic_read(&sk->sk_rmem_alloc)))
+		    (copied > 0 &&
+		     ((icsk->icsk_ack.pending & ICSK_ACK_PUSHED2) ||
+		      ((icsk->icsk_ack.pending & ICSK_ACK_PUSHED) &&
+		       !icsk->icsk_ack.pingpong)) &&
+		      !atomic_read(&sk->sk_rmem_alloc)))
 			time_to_ack = 1;
 	}
 
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 511b738f118a..b3def0df14fb 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -156,6 +156,8 @@ static void tcp_measure_rcv_mss(struct sock *sk,
 				return;
 			}
 		}
+		if (icsk->icsk_ack.pending & ICSK_ACK_PUSHED)
+			icsk->icsk_ack.pending |= ICSK_ACK_PUSHED2;
 		icsk->icsk_ack.pending |= ICSK_ACK_PUSHED;
 	}
 }
-- 
cgit v1.2.3


From a1e59abf824969554b90facd44a4ab16e265afa4 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 19 Sep 2006 12:57:34 -0700
Subject: [XFRM]: Fix wildcard as tunnel source

Hashing SAs by source address breaks templates with wildcards as tunnel
source since the source address used for hashing/lookup is still 0/0.
Move source address lookup to xfrm_tmpl_resolve_one() so we can use the
real address in the lookup.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/xfrm.h      | 13 +++++++++++++
 net/ipv4/xfrm4_policy.c | 20 ++++++++++++++++++++
 net/ipv4/xfrm4_state.c  | 15 ---------------
 net/ipv6/xfrm6_policy.c | 21 +++++++++++++++++++++
 net/ipv6/xfrm6_state.c  | 16 ----------------
 net/xfrm/xfrm_policy.c  | 21 +++++++++++++++++++++
 6 files changed, 75 insertions(+), 31 deletions(-)

(limited to 'include')

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 4d6dc627df9b..11e0b1d6bd47 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -222,6 +222,7 @@ struct xfrm_policy_afinfo {
 	struct dst_ops		*dst_ops;
 	void			(*garbage_collect)(void);
 	int			(*dst_lookup)(struct xfrm_dst **dst, struct flowi *fl);
+	int			(*get_saddr)(xfrm_address_t *saddr, xfrm_address_t *daddr);
 	struct dst_entry	*(*find_bundle)(struct flowi *fl, struct xfrm_policy *policy);
 	int			(*bundle_create)(struct xfrm_policy *policy, 
 						 struct xfrm_state **xfrm, 
@@ -630,6 +631,18 @@ secpath_reset(struct sk_buff *skb)
 #endif
 }
 
+static inline int
+xfrm_addr_any(xfrm_address_t *addr, unsigned short family)
+{
+	switch (family) {
+	case AF_INET:
+		return addr->a4 == 0;
+	case AF_INET6:
+		return ipv6_addr_any((struct in6_addr *)&addr->a6);
+	}
+	return 0;
+}
+
 static inline int
 __xfrm4_state_addr_cmp(struct xfrm_tmpl *tmpl, struct xfrm_state *x)
 {
diff --git a/net/ipv4/xfrm4_policy.c b/net/ipv4/xfrm4_policy.c
index 479598566f1d..eabcd27b1767 100644
--- a/net/ipv4/xfrm4_policy.c
+++ b/net/ipv4/xfrm4_policy.c
@@ -21,6 +21,25 @@ static int xfrm4_dst_lookup(struct xfrm_dst **dst, struct flowi *fl)
 	return __ip_route_output_key((struct rtable**)dst, fl);
 }
 
+static int xfrm4_get_saddr(xfrm_address_t *saddr, xfrm_address_t *daddr)
+{
+	struct rtable *rt;
+	struct flowi fl_tunnel = {
+		.nl_u = {
+			.ip4_u = {
+				.daddr = daddr->a4,
+			},
+		},
+	};
+
+	if (!xfrm4_dst_lookup((struct xfrm_dst **)&rt, &fl_tunnel)) {
+		saddr->a4 = rt->rt_src;
+		dst_release(&rt->u.dst);
+		return 0;
+	}
+	return -EHOSTUNREACH;
+}
+
 static struct dst_entry *
 __xfrm4_find_bundle(struct flowi *fl, struct xfrm_policy *policy)
 {
@@ -298,6 +317,7 @@ static struct xfrm_policy_afinfo xfrm4_policy_afinfo = {
 	.family = 		AF_INET,
 	.dst_ops =		&xfrm4_dst_ops,
 	.dst_lookup =		xfrm4_dst_lookup,
+	.get_saddr =		xfrm4_get_saddr,
 	.find_bundle = 		__xfrm4_find_bundle,
 	.bundle_create =	__xfrm4_bundle_create,
 	.decode_session =	_decode_session4,
diff --git a/net/ipv4/xfrm4_state.c b/net/ipv4/xfrm4_state.c
index 6a2a4ab42772..fe2034494d08 100644
--- a/net/ipv4/xfrm4_state.c
+++ b/net/ipv4/xfrm4_state.c
@@ -42,21 +42,6 @@ __xfrm4_init_tempsel(struct xfrm_state *x, struct flowi *fl,
 	x->props.saddr = tmpl->saddr;
 	if (x->props.saddr.a4 == 0)
 		x->props.saddr.a4 = saddr->a4;
-	if (tmpl->mode == XFRM_MODE_TUNNEL && x->props.saddr.a4 == 0) {
-		struct rtable *rt;
-	        struct flowi fl_tunnel = {
-        	        .nl_u = {
-        			.ip4_u = {
-					.daddr = x->id.daddr.a4,
-				}
-			}
-		};
-		if (!xfrm_dst_lookup((struct xfrm_dst **)&rt,
-		                     &fl_tunnel, AF_INET)) {
-			x->props.saddr.a4 = rt->rt_src;
-			dst_release(&rt->u.dst);
-		}
-	}
 	x->props.mode = tmpl->mode;
 	x->props.reqid = tmpl->reqid;
 	x->props.family = AF_INET;
diff --git a/net/ipv6/xfrm6_policy.c b/net/ipv6/xfrm6_policy.c
index 9391c4c94feb..6a252e2134d1 100644
--- a/net/ipv6/xfrm6_policy.c
+++ b/net/ipv6/xfrm6_policy.c
@@ -34,6 +34,26 @@ static int xfrm6_dst_lookup(struct xfrm_dst **dst, struct flowi *fl)
 	return err;
 }
 
+static int xfrm6_get_saddr(xfrm_address_t *saddr, xfrm_address_t *daddr)
+{
+	struct rt6_info *rt;
+	struct flowi fl_tunnel = {
+		.nl_u = {
+			.ip6_u = {
+				.daddr = *(struct in6_addr *)&daddr->a6,
+			},
+		},
+	};
+
+	if (!xfrm6_dst_lookup((struct xfrm_dst **)&rt, &fl_tunnel)) {
+		ipv6_get_saddr(&rt->u.dst, (struct in6_addr *)&daddr->a6,
+			       (struct in6_addr *)&saddr->a6);
+		dst_release(&rt->u.dst);
+		return 0;
+	}
+	return -EHOSTUNREACH;
+}
+
 static struct dst_entry *
 __xfrm6_find_bundle(struct flowi *fl, struct xfrm_policy *policy)
 {
@@ -362,6 +382,7 @@ static struct xfrm_policy_afinfo xfrm6_policy_afinfo = {
 	.family =		AF_INET6,
 	.dst_ops =		&xfrm6_dst_ops,
 	.dst_lookup =		xfrm6_dst_lookup,
+	.get_saddr = 		xfrm6_get_saddr,
 	.find_bundle =		__xfrm6_find_bundle,
 	.bundle_create =	__xfrm6_bundle_create,
 	.decode_session =	_decode_session6,
diff --git a/net/ipv6/xfrm6_state.c b/net/ipv6/xfrm6_state.c
index d88cd92c864e..711bfafb2472 100644
--- a/net/ipv6/xfrm6_state.c
+++ b/net/ipv6/xfrm6_state.c
@@ -42,22 +42,6 @@ __xfrm6_init_tempsel(struct xfrm_state *x, struct flowi *fl,
 	memcpy(&x->props.saddr, &tmpl->saddr, sizeof(x->props.saddr));
 	if (ipv6_addr_any((struct in6_addr*)&x->props.saddr))
 		memcpy(&x->props.saddr, saddr, sizeof(x->props.saddr));
-	if (tmpl->mode == XFRM_MODE_TUNNEL && ipv6_addr_any((struct in6_addr*)&x->props.saddr)) {
-		struct rt6_info *rt;
-		struct flowi fl_tunnel = {
-			.nl_u = {
-				.ip6_u = {
-					.daddr = *(struct in6_addr *)daddr,
-				}
-			}
-		};
-		if (!xfrm_dst_lookup((struct xfrm_dst **)&rt,
-		                     &fl_tunnel, AF_INET6)) {
-			ipv6_get_saddr(&rt->u.dst, (struct in6_addr *)daddr,
-			               (struct in6_addr *)&x->props.saddr);
-			dst_release(&rt->u.dst);
-		}
-	}
 	x->props.mode = tmpl->mode;
 	x->props.reqid = tmpl->reqid;
 	x->props.family = AF_INET6;
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 537854fe47ca..b6e2e79d7261 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -1107,6 +1107,20 @@ int __xfrm_sk_clone_policy(struct sock *sk)
 	return 0;
 }
 
+static int
+xfrm_get_saddr(xfrm_address_t *local, xfrm_address_t *remote,
+	       unsigned short family)
+{
+	int err;
+	struct xfrm_policy_afinfo *afinfo = xfrm_policy_get_afinfo(family);
+
+	if (unlikely(afinfo == NULL))
+		return -EINVAL;
+	err = afinfo->get_saddr(local, remote);
+	xfrm_policy_put_afinfo(afinfo);
+	return err;
+}
+
 /* Resolve list of templates for the flow, given policy. */
 
 static int
@@ -1118,6 +1132,7 @@ xfrm_tmpl_resolve_one(struct xfrm_policy *policy, struct flowi *fl,
 	int i, error;
 	xfrm_address_t *daddr = xfrm_flowi_daddr(fl, family);
 	xfrm_address_t *saddr = xfrm_flowi_saddr(fl, family);
+	xfrm_address_t tmp;
 
 	for (nx=0, i = 0; i < policy->xfrm_nr; i++) {
 		struct xfrm_state *x;
@@ -1128,6 +1143,12 @@ xfrm_tmpl_resolve_one(struct xfrm_policy *policy, struct flowi *fl,
 		if (tmpl->mode == XFRM_MODE_TUNNEL) {
 			remote = &tmpl->id.daddr;
 			local = &tmpl->saddr;
+			if (xfrm_addr_any(local, family)) {
+				error = xfrm_get_saddr(&tmp, remote, family);
+				if (error)
+					goto fail;
+				local = &tmp;
+			}
 		}
 
 		x = xfrm_state_find(remote, local, fl, tmpl, policy, &error, family);
-- 
cgit v1.2.3


From 1bf38a36b6a0e810dafae048fdbb999e587f0f2f Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Wed, 20 Sep 2006 11:57:09 -0700
Subject: [NETFILTER]: remove unused include file

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter_logging.h | 33 ---------------------------------
 1 file changed, 33 deletions(-)
 delete mode 100644 include/linux/netfilter_logging.h

(limited to 'include')

diff --git a/include/linux/netfilter_logging.h b/include/linux/netfilter_logging.h
deleted file mode 100644
index 562bb6aad4e1..000000000000
--- a/include/linux/netfilter_logging.h
+++ /dev/null
@@ -1,33 +0,0 @@
-/* Internal logging interface, which relies on the real 
-   LOG target modules */
-#ifndef __LINUX_NETFILTER_LOGGING_H
-#define __LINUX_NETFILTER_LOGGING_H
-
-#ifdef __KERNEL__
-#include <asm/atomic.h>
-
-struct nf_logging_t {
-	void (*nf_log_packet)(struct sk_buff **pskb,
-			      unsigned int hooknum,
-			      const struct net_device *in,
-			      const struct net_device *out,
-			      const char *prefix);
-	void (*nf_log)(char *pfh, size_t len,
-		       const char *prefix);
-};
-
-extern void nf_log_register(int pf, const struct nf_logging_t *logging);
-extern void nf_log_unregister(int pf, const struct nf_logging_t *logging);
-
-extern void nf_log_packet(int pf,
-			  struct sk_buff **pskb,
-			  unsigned int hooknum,
-			  const struct net_device *in,
-			  const struct net_device *out,
-			  const char *fmt, ...);
-extern void nf_log(int pf,
-		   char *pfh, size_t len,
-		   const char *fmt, ...);
-#endif /*__KERNEL__*/
-
-#endif /*__LINUX_NETFILTER_LOGGING_H*/
-- 
cgit v1.2.3


From df0933dcb027e156cb5253570ad694b81bd52b69 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Wed, 20 Sep 2006 11:57:53 -0700
Subject: [NETFILTER]: kill listhelp.h

Kill listhelp.h and use the list.h functions instead.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/x_tables.h           |   4 -
 include/linux/netfilter_ipv4/listhelp.h      | 123 -----------------
 net/bridge/netfilter/ebtables.c              |  76 ++++++-----
 net/ipv4/netfilter/arp_tables.c              |   2 -
 net/ipv4/netfilter/ip_conntrack_core.c       | 189 ++++++++++++---------------
 net/ipv4/netfilter/ip_conntrack_proto_gre.c  |  24 ++--
 net/ipv4/netfilter/ip_conntrack_standalone.c |   1 -
 net/ipv4/netfilter/ip_nat_core.c             |   4 -
 net/ipv4/netfilter/ip_nat_helper.c           |   4 -
 net/ipv4/netfilter/ip_nat_rule.c             |   4 -
 net/ipv4/netfilter/ip_nat_standalone.c       |   4 -
 net/ipv6/netfilter/ip6_tables.c              |   3 -
 net/netfilter/nf_conntrack_core.c            | 185 ++++++++++++--------------
 net/netfilter/nf_conntrack_standalone.c      |   1 -
 net/netfilter/x_tables.c                     |  17 ++-
 15 files changed, 237 insertions(+), 404 deletions(-)
 delete mode 100644 include/linux/netfilter_ipv4/listhelp.h

(limited to 'include')

diff --git a/include/linux/netfilter/x_tables.h b/include/linux/netfilter/x_tables.h
index 03d1027fb0e8..c832295dbf61 100644
--- a/include/linux/netfilter/x_tables.h
+++ b/include/linux/netfilter/x_tables.h
@@ -138,10 +138,6 @@ struct xt_counters_info
 
 #include <linux/netdevice.h>
 
-#define ASSERT_READ_LOCK(x)
-#define ASSERT_WRITE_LOCK(x)
-#include <linux/netfilter_ipv4/listhelp.h>
-
 #ifdef CONFIG_COMPAT
 #define COMPAT_TO_USER		1
 #define COMPAT_FROM_USER	-1
diff --git a/include/linux/netfilter_ipv4/listhelp.h b/include/linux/netfilter_ipv4/listhelp.h
deleted file mode 100644
index 5d92cf044d91..000000000000
--- a/include/linux/netfilter_ipv4/listhelp.h
+++ /dev/null
@@ -1,123 +0,0 @@
-#ifndef _LISTHELP_H
-#define _LISTHELP_H
-#include <linux/list.h>
-
-/* Header to do more comprehensive job than linux/list.h; assume list
-   is first entry in structure. */
-
-/* Return pointer to first true entry, if any, or NULL.  A macro
-   required to allow inlining of cmpfn. */
-#define LIST_FIND(head, cmpfn, type, args...)		\
-({							\
-	const struct list_head *__i, *__j = NULL;	\
-							\
-	ASSERT_READ_LOCK(head);				\
-	list_for_each(__i, (head))			\
-		if (cmpfn((const type)__i , ## args)) {	\
-			__j = __i;			\
-			break;				\
-		}					\
-	(type)__j;					\
-})
-
-#define LIST_FIND_W(head, cmpfn, type, args...)		\
-({							\
-	const struct list_head *__i, *__j = NULL;	\
-							\
-	ASSERT_WRITE_LOCK(head);			\
-	list_for_each(__i, (head))			\
-		if (cmpfn((type)__i , ## args)) {	\
-			__j = __i;			\
-			break;				\
-		}					\
-	(type)__j;					\
-})
-
-/* Just like LIST_FIND but we search backwards */
-#define LIST_FIND_B(head, cmpfn, type, args...)		\
-({							\
-	const struct list_head *__i, *__j = NULL;	\
-							\
-	ASSERT_READ_LOCK(head);				\
-	list_for_each_prev(__i, (head))			\
-		if (cmpfn((const type)__i , ## args)) {	\
-			__j = __i;			\
-			break;				\
-		}					\
-	(type)__j;					\
-})
-
-static inline int
-__list_cmp_same(const void *p1, const void *p2) { return p1 == p2; }
-
-/* Is this entry in the list? */
-static inline int
-list_inlist(struct list_head *head, const void *entry)
-{
-	return LIST_FIND(head, __list_cmp_same, void *, entry) != NULL;
-}
-
-/* Delete from list. */
-#ifdef CONFIG_NETFILTER_DEBUG
-#define LIST_DELETE(head, oldentry)					\
-do {									\
-	ASSERT_WRITE_LOCK(head);					\
-	if (!list_inlist(head, oldentry))				\
-		printk("LIST_DELETE: %s:%u `%s'(%p) not in %s.\n",	\
-		       __FILE__, __LINE__, #oldentry, oldentry, #head);	\
-        else list_del((struct list_head *)oldentry);			\
-} while(0)
-#else
-#define LIST_DELETE(head, oldentry) list_del((struct list_head *)oldentry)
-#endif
-
-/* Append. */
-static inline void
-list_append(struct list_head *head, void *new)
-{
-	ASSERT_WRITE_LOCK(head);
-	list_add((new), (head)->prev);
-}
-
-/* Prepend. */
-static inline void
-list_prepend(struct list_head *head, void *new)
-{
-	ASSERT_WRITE_LOCK(head);
-	list_add(new, head);
-}
-
-/* Insert according to ordering function; insert before first true. */
-#define LIST_INSERT(head, new, cmpfn)				\
-do {								\
-	struct list_head *__i;					\
-	ASSERT_WRITE_LOCK(head);				\
-	list_for_each(__i, (head))				\
-		if ((new), (typeof (new))__i)			\
-			break;					\
-	list_add((struct list_head *)(new), __i->prev);		\
-} while(0)
-
-/* If the field after the list_head is a nul-terminated string, you
-   can use these functions. */
-static inline int __list_cmp_name(const void *i, const char *name)
-{
-	return strcmp(name, i+sizeof(struct list_head)) == 0;
-}
-
-/* Returns false if same name already in list, otherwise does insert. */
-static inline int
-list_named_insert(struct list_head *head, void *new)
-{
-	if (LIST_FIND(head, __list_cmp_name, void *,
-		      new + sizeof(struct list_head)))
-		return 0;
-	list_prepend(head, new);
-	return 1;
-}
-
-/* Find this named element in the list. */
-#define list_named_find(head, name)			\
-LIST_FIND(head, __list_cmp_name, void *, name)
-
-#endif /*_LISTHELP_H*/
diff --git a/net/bridge/netfilter/ebtables.c b/net/bridge/netfilter/ebtables.c
index d06a5075b5f6..3df55b2bd91d 100644
--- a/net/bridge/netfilter/ebtables.c
+++ b/net/bridge/netfilter/ebtables.c
@@ -24,6 +24,7 @@
 #include <linux/vmalloc.h>
 #include <linux/netfilter_bridge/ebtables.h>
 #include <linux/spinlock.h>
+#include <linux/mutex.h>
 #include <asm/uaccess.h>
 #include <linux/smp.h>
 #include <linux/cpumask.h>
@@ -31,12 +32,6 @@
 /* needed for logical [in,out]-dev filtering */
 #include "../br_private.h"
 
-/* list_named_find */
-#define ASSERT_READ_LOCK(x)
-#define ASSERT_WRITE_LOCK(x)
-#include <linux/netfilter_ipv4/listhelp.h>
-#include <linux/mutex.h>
-
 #define BUGPRINT(format, args...) printk("kernel msg: ebtables bug: please "\
                                          "report to author: "format, ## args)
 /* #define BUGPRINT(format, args...) */
@@ -278,18 +273,22 @@ static inline void *
 find_inlist_lock_noload(struct list_head *head, const char *name, int *error,
    struct mutex *mutex)
 {
-	void *ret;
+	struct {
+		struct list_head list;
+		char name[EBT_FUNCTION_MAXNAMELEN];
+	} *e;
 
 	*error = mutex_lock_interruptible(mutex);
 	if (*error != 0)
 		return NULL;
 
-	ret = list_named_find(head, name);
-	if (!ret) {
-		*error = -ENOENT;
-		mutex_unlock(mutex);
+	list_for_each_entry(e, head, list) {
+		if (strcmp(e->name, name) == 0)
+			return e;
 	}
-	return ret;
+	*error = -ENOENT;
+	mutex_unlock(mutex);
+	return NULL;
 }
 
 #ifndef CONFIG_KMOD
@@ -1043,15 +1042,19 @@ free_newinfo:
 
 int ebt_register_target(struct ebt_target *target)
 {
+	struct ebt_target *t;
 	int ret;
 
 	ret = mutex_lock_interruptible(&ebt_mutex);
 	if (ret != 0)
 		return ret;
-	if (!list_named_insert(&ebt_targets, target)) {
-		mutex_unlock(&ebt_mutex);
-		return -EEXIST;
+	list_for_each_entry(t, &ebt_targets, list) {
+		if (strcmp(t->name, target->name) == 0) {
+			mutex_unlock(&ebt_mutex);
+			return -EEXIST;
+		}
 	}
+	list_add(&target->list, &ebt_targets);
 	mutex_unlock(&ebt_mutex);
 
 	return 0;
@@ -1060,21 +1063,25 @@ int ebt_register_target(struct ebt_target *target)
 void ebt_unregister_target(struct ebt_target *target)
 {
 	mutex_lock(&ebt_mutex);
-	LIST_DELETE(&ebt_targets, target);
+	list_del(&target->list);
 	mutex_unlock(&ebt_mutex);
 }
 
 int ebt_register_match(struct ebt_match *match)
 {
+	struct ebt_match *m;
 	int ret;
 
 	ret = mutex_lock_interruptible(&ebt_mutex);
 	if (ret != 0)
 		return ret;
-	if (!list_named_insert(&ebt_matches, match)) {
-		mutex_unlock(&ebt_mutex);
-		return -EEXIST;
+	list_for_each_entry(m, &ebt_matches, list) {
+		if (strcmp(m->name, match->name) == 0) {
+			mutex_unlock(&ebt_mutex);
+			return -EEXIST;
+		}
 	}
+	list_add(&match->list, &ebt_matches);
 	mutex_unlock(&ebt_mutex);
 
 	return 0;
@@ -1083,21 +1090,25 @@ int ebt_register_match(struct ebt_match *match)
 void ebt_unregister_match(struct ebt_match *match)
 {
 	mutex_lock(&ebt_mutex);
-	LIST_DELETE(&ebt_matches, match);
+	list_del(&match->list);
 	mutex_unlock(&ebt_mutex);
 }
 
 int ebt_register_watcher(struct ebt_watcher *watcher)
 {
+	struct ebt_watcher *w;
 	int ret;
 
 	ret = mutex_lock_interruptible(&ebt_mutex);
 	if (ret != 0)
 		return ret;
-	if (!list_named_insert(&ebt_watchers, watcher)) {
-		mutex_unlock(&ebt_mutex);
-		return -EEXIST;
+	list_for_each_entry(w, &ebt_watchers, list) {
+		if (strcmp(w->name, watcher->name) == 0) {
+			mutex_unlock(&ebt_mutex);
+			return -EEXIST;
+		}
 	}
+	list_add(&watcher->list, &ebt_watchers);
 	mutex_unlock(&ebt_mutex);
 
 	return 0;
@@ -1106,13 +1117,14 @@ int ebt_register_watcher(struct ebt_watcher *watcher)
 void ebt_unregister_watcher(struct ebt_watcher *watcher)
 {
 	mutex_lock(&ebt_mutex);
-	LIST_DELETE(&ebt_watchers, watcher);
+	list_del(&watcher->list);
 	mutex_unlock(&ebt_mutex);
 }
 
 int ebt_register_table(struct ebt_table *table)
 {
 	struct ebt_table_info *newinfo;
+	struct ebt_table *t;
 	int ret, i, countersize;
 
 	if (!table || !table->table ||!table->table->entries ||
@@ -1158,10 +1170,12 @@ int ebt_register_table(struct ebt_table *table)
 	if (ret != 0)
 		goto free_chainstack;
 
-	if (list_named_find(&ebt_tables, table->name)) {
-		ret = -EEXIST;
-		BUGPRINT("Table name already exists\n");
-		goto free_unlock;
+	list_for_each_entry(t, &ebt_tables, list) {
+		if (strcmp(t->name, table->name) == 0) {
+			ret = -EEXIST;
+			BUGPRINT("Table name already exists\n");
+			goto free_unlock;
+		}
 	}
 
 	/* Hold a reference count if the chains aren't empty */
@@ -1169,7 +1183,7 @@ int ebt_register_table(struct ebt_table *table)
 		ret = -ENOENT;
 		goto free_unlock;
 	}
-	list_prepend(&ebt_tables, table);
+	list_add(&table->list, &ebt_tables);
 	mutex_unlock(&ebt_mutex);
 	return 0;
 free_unlock:
@@ -1195,7 +1209,7 @@ void ebt_unregister_table(struct ebt_table *table)
 		return;
 	}
 	mutex_lock(&ebt_mutex);
-	LIST_DELETE(&ebt_tables, table);
+	list_del(&table->list);
 	mutex_unlock(&ebt_mutex);
 	vfree(table->private->entries);
 	if (table->private->chainstack) {
@@ -1465,7 +1479,7 @@ static int __init ebtables_init(void)
 	int ret;
 
 	mutex_lock(&ebt_mutex);
-	list_named_insert(&ebt_targets, &ebt_standard_target);
+	list_add(&ebt_standard_target.list, &ebt_targets);
 	mutex_unlock(&ebt_mutex);
 	if ((ret = nf_register_sockopt(&ebt_sockopts)) < 0)
 		return ret;
diff --git a/net/ipv4/netfilter/arp_tables.c b/net/ipv4/netfilter/arp_tables.c
index 4f10b06413a1..aaeaa9ce0f28 100644
--- a/net/ipv4/netfilter/arp_tables.c
+++ b/net/ipv4/netfilter/arp_tables.c
@@ -56,8 +56,6 @@ do {								\
 #define ARP_NF_ASSERT(x)
 #endif
 
-#include <linux/netfilter_ipv4/listhelp.h>
-
 static inline int arp_devaddr_compare(const struct arpt_devaddr_info *ap,
 				      char *hdr_addr, int len)
 {
diff --git a/net/ipv4/netfilter/ip_conntrack_core.c b/net/ipv4/netfilter/ip_conntrack_core.c
index 5da25ad50309..2568d480e9a9 100644
--- a/net/ipv4/netfilter/ip_conntrack_core.c
+++ b/net/ipv4/netfilter/ip_conntrack_core.c
@@ -47,7 +47,6 @@
 #include <linux/netfilter_ipv4/ip_conntrack_protocol.h>
 #include <linux/netfilter_ipv4/ip_conntrack_helper.h>
 #include <linux/netfilter_ipv4/ip_conntrack_core.h>
-#include <linux/netfilter_ipv4/listhelp.h>
 
 #define IP_CONNTRACK_VERSION	"2.4"
 
@@ -294,15 +293,10 @@ void ip_ct_remove_expectations(struct ip_conntrack *ct)
 static void
 clean_from_lists(struct ip_conntrack *ct)
 {
-	unsigned int ho, hr;
-	
 	DEBUGP("clean_from_lists(%p)\n", ct);
 	ASSERT_WRITE_LOCK(&ip_conntrack_lock);
-
-	ho = hash_conntrack(&ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple);
-	hr = hash_conntrack(&ct->tuplehash[IP_CT_DIR_REPLY].tuple);
-	LIST_DELETE(&ip_conntrack_hash[ho], &ct->tuplehash[IP_CT_DIR_ORIGINAL]);
-	LIST_DELETE(&ip_conntrack_hash[hr], &ct->tuplehash[IP_CT_DIR_REPLY]);
+	list_del(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list);
+	list_del(&ct->tuplehash[IP_CT_DIR_REPLY].list);
 
 	/* Destroy all pending expectations */
 	ip_ct_remove_expectations(ct);
@@ -367,16 +361,6 @@ static void death_by_timeout(unsigned long ul_conntrack)
 	ip_conntrack_put(ct);
 }
 
-static inline int
-conntrack_tuple_cmp(const struct ip_conntrack_tuple_hash *i,
-		    const struct ip_conntrack_tuple *tuple,
-		    const struct ip_conntrack *ignored_conntrack)
-{
-	ASSERT_READ_LOCK(&ip_conntrack_lock);
-	return tuplehash_to_ctrack(i) != ignored_conntrack
-		&& ip_ct_tuple_equal(tuple, &i->tuple);
-}
-
 struct ip_conntrack_tuple_hash *
 __ip_conntrack_find(const struct ip_conntrack_tuple *tuple,
 		    const struct ip_conntrack *ignored_conntrack)
@@ -386,7 +370,8 @@ __ip_conntrack_find(const struct ip_conntrack_tuple *tuple,
 
 	ASSERT_READ_LOCK(&ip_conntrack_lock);
 	list_for_each_entry(h, &ip_conntrack_hash[hash], list) {
-		if (conntrack_tuple_cmp(h, tuple, ignored_conntrack)) {
+		if (tuplehash_to_ctrack(h) != ignored_conntrack &&
+		    ip_ct_tuple_equal(tuple, &h->tuple)) {
 			CONNTRACK_STAT_INC(found);
 			return h;
 		}
@@ -417,10 +402,10 @@ static void __ip_conntrack_hash_insert(struct ip_conntrack *ct,
 					unsigned int repl_hash) 
 {
 	ct->id = ++ip_conntrack_next_id;
-	list_prepend(&ip_conntrack_hash[hash],
-		     &ct->tuplehash[IP_CT_DIR_ORIGINAL].list);
-	list_prepend(&ip_conntrack_hash[repl_hash],
-		     &ct->tuplehash[IP_CT_DIR_REPLY].list);
+	list_add(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list,
+		 &ip_conntrack_hash[hash]);
+	list_add(&ct->tuplehash[IP_CT_DIR_REPLY].list,
+		 &ip_conntrack_hash[repl_hash]);
 }
 
 void ip_conntrack_hash_insert(struct ip_conntrack *ct)
@@ -440,6 +425,7 @@ int
 __ip_conntrack_confirm(struct sk_buff **pskb)
 {
 	unsigned int hash, repl_hash;
+	struct ip_conntrack_tuple_hash *h;
 	struct ip_conntrack *ct;
 	enum ip_conntrack_info ctinfo;
 
@@ -470,43 +456,43 @@ __ip_conntrack_confirm(struct sk_buff **pskb)
 	/* See if there's one in the list already, including reverse:
            NAT could have grabbed it without realizing, since we're
            not in the hash.  If there is, we lost race. */
-	if (!LIST_FIND(&ip_conntrack_hash[hash],
-		       conntrack_tuple_cmp,
-		       struct ip_conntrack_tuple_hash *,
-		       &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple, NULL)
-	    && !LIST_FIND(&ip_conntrack_hash[repl_hash],
-			  conntrack_tuple_cmp,
-			  struct ip_conntrack_tuple_hash *,
-			  &ct->tuplehash[IP_CT_DIR_REPLY].tuple, NULL)) {
-		/* Remove from unconfirmed list */
-		list_del(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list);
+	list_for_each_entry(h, &ip_conntrack_hash[hash], list)
+		if (ip_ct_tuple_equal(&ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple,
+				      &h->tuple))
+			goto out;
+	list_for_each_entry(h, &ip_conntrack_hash[repl_hash], list)
+		if (ip_ct_tuple_equal(&ct->tuplehash[IP_CT_DIR_REPLY].tuple,
+				      &h->tuple))
+			goto out;
 
-		__ip_conntrack_hash_insert(ct, hash, repl_hash);
-		/* Timer relative to confirmation time, not original
-		   setting time, otherwise we'd get timer wrap in
-		   weird delay cases. */
-		ct->timeout.expires += jiffies;
-		add_timer(&ct->timeout);
-		atomic_inc(&ct->ct_general.use);
-		set_bit(IPS_CONFIRMED_BIT, &ct->status);
-		CONNTRACK_STAT_INC(insert);
-		write_unlock_bh(&ip_conntrack_lock);
-		if (ct->helper)
-			ip_conntrack_event_cache(IPCT_HELPER, *pskb);
+	/* Remove from unconfirmed list */
+	list_del(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list);
+
+	__ip_conntrack_hash_insert(ct, hash, repl_hash);
+	/* Timer relative to confirmation time, not original
+	   setting time, otherwise we'd get timer wrap in
+	   weird delay cases. */
+	ct->timeout.expires += jiffies;
+	add_timer(&ct->timeout);
+	atomic_inc(&ct->ct_general.use);
+	set_bit(IPS_CONFIRMED_BIT, &ct->status);
+	CONNTRACK_STAT_INC(insert);
+	write_unlock_bh(&ip_conntrack_lock);
+	if (ct->helper)
+		ip_conntrack_event_cache(IPCT_HELPER, *pskb);
 #ifdef CONFIG_IP_NF_NAT_NEEDED
-		if (test_bit(IPS_SRC_NAT_DONE_BIT, &ct->status) ||
-		    test_bit(IPS_DST_NAT_DONE_BIT, &ct->status))
-			ip_conntrack_event_cache(IPCT_NATINFO, *pskb);
+	if (test_bit(IPS_SRC_NAT_DONE_BIT, &ct->status) ||
+	    test_bit(IPS_DST_NAT_DONE_BIT, &ct->status))
+		ip_conntrack_event_cache(IPCT_NATINFO, *pskb);
 #endif
-		ip_conntrack_event_cache(master_ct(ct) ?
-					 IPCT_RELATED : IPCT_NEW, *pskb);
+	ip_conntrack_event_cache(master_ct(ct) ?
+				 IPCT_RELATED : IPCT_NEW, *pskb);
 
-		return NF_ACCEPT;
-	}
+	return NF_ACCEPT;
 
+out:
 	CONNTRACK_STAT_INC(insert_failed);
 	write_unlock_bh(&ip_conntrack_lock);
-
 	return NF_DROP;
 }
 
@@ -527,23 +513,21 @@ ip_conntrack_tuple_taken(const struct ip_conntrack_tuple *tuple,
 
 /* There's a small race here where we may free a just-assured
    connection.  Too bad: we're in trouble anyway. */
-static inline int unreplied(const struct ip_conntrack_tuple_hash *i)
-{
-	return !(test_bit(IPS_ASSURED_BIT, &tuplehash_to_ctrack(i)->status));
-}
-
 static int early_drop(struct list_head *chain)
 {
 	/* Traverse backwards: gives us oldest, which is roughly LRU */
 	struct ip_conntrack_tuple_hash *h;
-	struct ip_conntrack *ct = NULL;
+	struct ip_conntrack *ct = NULL, *tmp;
 	int dropped = 0;
 
 	read_lock_bh(&ip_conntrack_lock);
-	h = LIST_FIND_B(chain, unreplied, struct ip_conntrack_tuple_hash *);
-	if (h) {
-		ct = tuplehash_to_ctrack(h);
-		atomic_inc(&ct->ct_general.use);
+	list_for_each_entry_reverse(h, chain, list) {
+		tmp = tuplehash_to_ctrack(h);
+		if (!test_bit(IPS_ASSURED_BIT, &tmp->status)) {
+			ct = tmp;
+			atomic_inc(&ct->ct_general.use);
+			break;
+		}
 	}
 	read_unlock_bh(&ip_conntrack_lock);
 
@@ -559,18 +543,16 @@ static int early_drop(struct list_head *chain)
 	return dropped;
 }
 
-static inline int helper_cmp(const struct ip_conntrack_helper *i,
-			     const struct ip_conntrack_tuple *rtuple)
-{
-	return ip_ct_tuple_mask_cmp(rtuple, &i->tuple, &i->mask);
-}
-
 static struct ip_conntrack_helper *
 __ip_conntrack_helper_find( const struct ip_conntrack_tuple *tuple)
 {
-	return LIST_FIND(&helpers, helper_cmp,
-			 struct ip_conntrack_helper *,
-			 tuple);
+	struct ip_conntrack_helper *h;
+
+	list_for_each_entry(h, &helpers, list) {
+		if (ip_ct_tuple_mask_cmp(tuple, &h->tuple, &h->mask))
+			return h;
+	}
+	return NULL;
 }
 
 struct ip_conntrack_helper *
@@ -1062,7 +1044,7 @@ int ip_conntrack_helper_register(struct ip_conntrack_helper *me)
 {
 	BUG_ON(me->timeout == 0);
 	write_lock_bh(&ip_conntrack_lock);
-	list_prepend(&helpers, me);
+	list_add(&me->list, &helpers);
 	write_unlock_bh(&ip_conntrack_lock);
 
 	return 0;
@@ -1081,24 +1063,24 @@ __ip_conntrack_helper_find_byname(const char *name)
 	return NULL;
 }
 
-static inline int unhelp(struct ip_conntrack_tuple_hash *i,
-			 const struct ip_conntrack_helper *me)
+static inline void unhelp(struct ip_conntrack_tuple_hash *i,
+			  const struct ip_conntrack_helper *me)
 {
 	if (tuplehash_to_ctrack(i)->helper == me) {
  		ip_conntrack_event(IPCT_HELPER, tuplehash_to_ctrack(i));
 		tuplehash_to_ctrack(i)->helper = NULL;
 	}
-	return 0;
 }
 
 void ip_conntrack_helper_unregister(struct ip_conntrack_helper *me)
 {
 	unsigned int i;
+	struct ip_conntrack_tuple_hash *h;
 	struct ip_conntrack_expect *exp, *tmp;
 
 	/* Need write lock here, to delete helper. */
 	write_lock_bh(&ip_conntrack_lock);
-	LIST_DELETE(&helpers, me);
+	list_del(&me->list);
 
 	/* Get rid of expectations */
 	list_for_each_entry_safe(exp, tmp, &ip_conntrack_expect_list, list) {
@@ -1108,10 +1090,12 @@ void ip_conntrack_helper_unregister(struct ip_conntrack_helper *me)
 		}
 	}
 	/* Get rid of expecteds, set helpers to NULL. */
-	LIST_FIND_W(&unconfirmed, unhelp, struct ip_conntrack_tuple_hash*, me);
-	for (i = 0; i < ip_conntrack_htable_size; i++)
-		LIST_FIND_W(&ip_conntrack_hash[i], unhelp,
-			    struct ip_conntrack_tuple_hash *, me);
+	list_for_each_entry(h, &unconfirmed, list)
+		unhelp(h, me);
+	for (i = 0; i < ip_conntrack_htable_size; i++) {
+		list_for_each_entry(h, &ip_conntrack_hash[i], list)
+			unhelp(h, me);
+	}
 	write_unlock_bh(&ip_conntrack_lock);
 
 	/* Someone could be still looking at the helper in a bh. */
@@ -1237,46 +1221,43 @@ static void ip_conntrack_attach(struct sk_buff *nskb, struct sk_buff *skb)
 	nf_conntrack_get(nskb->nfct);
 }
 
-static inline int
-do_iter(const struct ip_conntrack_tuple_hash *i,
-	int (*iter)(struct ip_conntrack *i, void *data),
-	void *data)
-{
-	return iter(tuplehash_to_ctrack(i), data);
-}
-
 /* Bring out ya dead! */
-static struct ip_conntrack_tuple_hash *
+static struct ip_conntrack *
 get_next_corpse(int (*iter)(struct ip_conntrack *i, void *data),
 		void *data, unsigned int *bucket)
 {
-	struct ip_conntrack_tuple_hash *h = NULL;
+	struct ip_conntrack_tuple_hash *h;
+	struct ip_conntrack *ct;
 
 	write_lock_bh(&ip_conntrack_lock);
 	for (; *bucket < ip_conntrack_htable_size; (*bucket)++) {
-		h = LIST_FIND_W(&ip_conntrack_hash[*bucket], do_iter,
-				struct ip_conntrack_tuple_hash *, iter, data);
-		if (h)
-			break;
+		list_for_each_entry(h, &ip_conntrack_hash[*bucket], list) {
+			ct = tuplehash_to_ctrack(h);
+			if (iter(ct, data))
+				goto found;
+		}
+	}
+	list_for_each_entry(h, &unconfirmed, list) {
+		ct = tuplehash_to_ctrack(h);
+		if (iter(ct, data))
+			goto found;
 	}
-	if (!h)
-		h = LIST_FIND_W(&unconfirmed, do_iter,
-				struct ip_conntrack_tuple_hash *, iter, data);
-	if (h)
-		atomic_inc(&tuplehash_to_ctrack(h)->ct_general.use);
 	write_unlock_bh(&ip_conntrack_lock);
+	return NULL;
 
-	return h;
+found:
+	atomic_inc(&ct->ct_general.use);
+	write_unlock_bh(&ip_conntrack_lock);
+	return ct;
 }
 
 void
 ip_ct_iterate_cleanup(int (*iter)(struct ip_conntrack *i, void *), void *data)
 {
-	struct ip_conntrack_tuple_hash *h;
+	struct ip_conntrack *ct;
 	unsigned int bucket = 0;
 
-	while ((h = get_next_corpse(iter, data, &bucket)) != NULL) {
-		struct ip_conntrack *ct = tuplehash_to_ctrack(h);
+	while ((ct = get_next_corpse(iter, data, &bucket)) != NULL) {
 		/* Time to push up daises... */
 		if (del_timer(&ct->timeout))
 			death_by_timeout((unsigned long)ct);
diff --git a/net/ipv4/netfilter/ip_conntrack_proto_gre.c b/net/ipv4/netfilter/ip_conntrack_proto_gre.c
index 4ee016c427b4..92c6d8b178c9 100644
--- a/net/ipv4/netfilter/ip_conntrack_proto_gre.c
+++ b/net/ipv4/netfilter/ip_conntrack_proto_gre.c
@@ -37,7 +37,6 @@ static DEFINE_RWLOCK(ip_ct_gre_lock);
 #define ASSERT_READ_LOCK(x)
 #define ASSERT_WRITE_LOCK(x)
 
-#include <linux/netfilter_ipv4/listhelp.h>
 #include <linux/netfilter_ipv4/ip_conntrack_protocol.h>
 #include <linux/netfilter_ipv4/ip_conntrack_helper.h>
 #include <linux/netfilter_ipv4/ip_conntrack_core.h>
@@ -82,10 +81,12 @@ static __be16 gre_keymap_lookup(struct ip_conntrack_tuple *t)
 	__be16 key = 0;
 
 	read_lock_bh(&ip_ct_gre_lock);
-	km = LIST_FIND(&gre_keymap_list, gre_key_cmpfn,
-			struct ip_ct_gre_keymap *, t);
-	if (km)
-		key = km->tuple.src.u.gre.key;
+	list_for_each_entry(km, &gre_keymap_list, list) {
+		if (gre_key_cmpfn(km, t)) {
+			key = km->tuple.src.u.gre.key;
+			break;
+		}
+	}
 	read_unlock_bh(&ip_ct_gre_lock);
 	
 	DEBUGP("lookup src key 0x%x up key for ", key);
@@ -99,7 +100,7 @@ int
 ip_ct_gre_keymap_add(struct ip_conntrack *ct,
 		     struct ip_conntrack_tuple *t, int reply)
 {
-	struct ip_ct_gre_keymap **exist_km, *km, *old;
+	struct ip_ct_gre_keymap **exist_km, *km;
 
 	if (!ct->helper || strcmp(ct->helper->name, "pptp")) {
 		DEBUGP("refusing to add GRE keymap to non-pptp session\n");
@@ -113,13 +114,10 @@ ip_ct_gre_keymap_add(struct ip_conntrack *ct,
 
 	if (*exist_km) {
 		/* check whether it's a retransmission */
-		old = LIST_FIND(&gre_keymap_list, gre_key_cmpfn,
-				struct ip_ct_gre_keymap *, t);
-		if (old == *exist_km) {
-			DEBUGP("retransmission\n");
-			return 0;
+		list_for_each_entry(km, &gre_keymap_list, list) {
+			if (gre_key_cmpfn(km, t) && km == *exist_km)
+				return 0;
 		}
-
 		DEBUGP("trying to override keymap_%s for ct %p\n", 
 			reply? "reply":"orig", ct);
 		return -EEXIST;
@@ -136,7 +134,7 @@ ip_ct_gre_keymap_add(struct ip_conntrack *ct,
 	DUMP_TUPLE_GRE(&km->tuple);
 
 	write_lock_bh(&ip_ct_gre_lock);
-	list_append(&gre_keymap_list, km);
+	list_add_tail(&km->list, &gre_keymap_list);
 	write_unlock_bh(&ip_ct_gre_lock);
 
 	return 0;
diff --git a/net/ipv4/netfilter/ip_conntrack_standalone.c b/net/ipv4/netfilter/ip_conntrack_standalone.c
index 3f5d495b853b..02135756562e 100644
--- a/net/ipv4/netfilter/ip_conntrack_standalone.c
+++ b/net/ipv4/netfilter/ip_conntrack_standalone.c
@@ -35,7 +35,6 @@
 #include <linux/netfilter_ipv4/ip_conntrack_protocol.h>
 #include <linux/netfilter_ipv4/ip_conntrack_core.h>
 #include <linux/netfilter_ipv4/ip_conntrack_helper.h>
-#include <linux/netfilter_ipv4/listhelp.h>
 
 #if 0
 #define DEBUGP printk
diff --git a/net/ipv4/netfilter/ip_nat_core.c b/net/ipv4/netfilter/ip_nat_core.c
index 4c540d03d48e..71f3e09cbc84 100644
--- a/net/ipv4/netfilter/ip_nat_core.c
+++ b/net/ipv4/netfilter/ip_nat_core.c
@@ -22,9 +22,6 @@
 #include <linux/udp.h>
 #include <linux/jhash.h>
 
-#define ASSERT_READ_LOCK(x)
-#define ASSERT_WRITE_LOCK(x)
-
 #include <linux/netfilter_ipv4/ip_conntrack.h>
 #include <linux/netfilter_ipv4/ip_conntrack_core.h>
 #include <linux/netfilter_ipv4/ip_conntrack_protocol.h>
@@ -33,7 +30,6 @@
 #include <linux/netfilter_ipv4/ip_nat_core.h>
 #include <linux/netfilter_ipv4/ip_nat_helper.h>
 #include <linux/netfilter_ipv4/ip_conntrack_helper.h>
-#include <linux/netfilter_ipv4/listhelp.h>
 
 #if 0
 #define DEBUGP printk
diff --git a/net/ipv4/netfilter/ip_nat_helper.c b/net/ipv4/netfilter/ip_nat_helper.c
index 021c3daae3ed..7f6a75984f6c 100644
--- a/net/ipv4/netfilter/ip_nat_helper.c
+++ b/net/ipv4/netfilter/ip_nat_helper.c
@@ -27,16 +27,12 @@
 #include <net/tcp.h>
 #include <net/udp.h>
 
-#define ASSERT_READ_LOCK(x)
-#define ASSERT_WRITE_LOCK(x)
-
 #include <linux/netfilter_ipv4/ip_conntrack.h>
 #include <linux/netfilter_ipv4/ip_conntrack_helper.h>
 #include <linux/netfilter_ipv4/ip_nat.h>
 #include <linux/netfilter_ipv4/ip_nat_protocol.h>
 #include <linux/netfilter_ipv4/ip_nat_core.h>
 #include <linux/netfilter_ipv4/ip_nat_helper.h>
-#include <linux/netfilter_ipv4/listhelp.h>
 
 #if 0
 #define DEBUGP printk
diff --git a/net/ipv4/netfilter/ip_nat_rule.c b/net/ipv4/netfilter/ip_nat_rule.c
index e59f5a8ecb6b..7b703839aa58 100644
--- a/net/ipv4/netfilter/ip_nat_rule.c
+++ b/net/ipv4/netfilter/ip_nat_rule.c
@@ -19,14 +19,10 @@
 #include <net/route.h>
 #include <linux/bitops.h>
 
-#define ASSERT_READ_LOCK(x)
-#define ASSERT_WRITE_LOCK(x)
-
 #include <linux/netfilter_ipv4/ip_tables.h>
 #include <linux/netfilter_ipv4/ip_nat.h>
 #include <linux/netfilter_ipv4/ip_nat_core.h>
 #include <linux/netfilter_ipv4/ip_nat_rule.h>
-#include <linux/netfilter_ipv4/listhelp.h>
 
 #if 0
 #define DEBUGP printk
diff --git a/net/ipv4/netfilter/ip_nat_standalone.c b/net/ipv4/netfilter/ip_nat_standalone.c
index f3b778355432..9c577db62047 100644
--- a/net/ipv4/netfilter/ip_nat_standalone.c
+++ b/net/ipv4/netfilter/ip_nat_standalone.c
@@ -30,9 +30,6 @@
 #include <net/checksum.h>
 #include <linux/spinlock.h>
 
-#define ASSERT_READ_LOCK(x)
-#define ASSERT_WRITE_LOCK(x)
-
 #include <linux/netfilter_ipv4/ip_nat.h>
 #include <linux/netfilter_ipv4/ip_nat_rule.h>
 #include <linux/netfilter_ipv4/ip_nat_protocol.h>
@@ -40,7 +37,6 @@
 #include <linux/netfilter_ipv4/ip_nat_helper.h>
 #include <linux/netfilter_ipv4/ip_tables.h>
 #include <linux/netfilter_ipv4/ip_conntrack_core.h>
-#include <linux/netfilter_ipv4/listhelp.h>
 
 #if 0
 #define DEBUGP printk
diff --git a/net/ipv6/netfilter/ip6_tables.c b/net/ipv6/netfilter/ip6_tables.c
index d1c315364ee7..73d477ce216b 100644
--- a/net/ipv6/netfilter/ip6_tables.c
+++ b/net/ipv6/netfilter/ip6_tables.c
@@ -70,9 +70,6 @@ do {								\
 #define IP_NF_ASSERT(x)
 #endif
 
-
-#include <linux/netfilter_ipv4/listhelp.h>
-
 #if 0
 /* All the better to debug you with... */
 #define static
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index 3b64dbee6620..927137b8b3b5 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -57,7 +57,6 @@
 #include <net/netfilter/nf_conntrack_protocol.h>
 #include <net/netfilter/nf_conntrack_helper.h>
 #include <net/netfilter/nf_conntrack_core.h>
-#include <linux/netfilter_ipv4/listhelp.h>
 
 #define NF_CONNTRACK_VERSION	"0.5.0"
 
@@ -539,15 +538,10 @@ void nf_ct_remove_expectations(struct nf_conn *ct)
 static void
 clean_from_lists(struct nf_conn *ct)
 {
-	unsigned int ho, hr;
-	
 	DEBUGP("clean_from_lists(%p)\n", ct);
 	ASSERT_WRITE_LOCK(&nf_conntrack_lock);
-
-	ho = hash_conntrack(&ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple);
-	hr = hash_conntrack(&ct->tuplehash[IP_CT_DIR_REPLY].tuple);
-	LIST_DELETE(&nf_conntrack_hash[ho], &ct->tuplehash[IP_CT_DIR_ORIGINAL]);
-	LIST_DELETE(&nf_conntrack_hash[hr], &ct->tuplehash[IP_CT_DIR_REPLY]);
+	list_del(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list);
+	list_del(&ct->tuplehash[IP_CT_DIR_REPLY].list);
 
 	/* Destroy all pending expectations */
 	nf_ct_remove_expectations(ct);
@@ -617,16 +611,6 @@ static void death_by_timeout(unsigned long ul_conntrack)
 	nf_ct_put(ct);
 }
 
-static inline int
-conntrack_tuple_cmp(const struct nf_conntrack_tuple_hash *i,
-		    const struct nf_conntrack_tuple *tuple,
-		    const struct nf_conn *ignored_conntrack)
-{
-	ASSERT_READ_LOCK(&nf_conntrack_lock);
-	return nf_ct_tuplehash_to_ctrack(i) != ignored_conntrack
-		&& nf_ct_tuple_equal(tuple, &i->tuple);
-}
-
 struct nf_conntrack_tuple_hash *
 __nf_conntrack_find(const struct nf_conntrack_tuple *tuple,
 		    const struct nf_conn *ignored_conntrack)
@@ -636,7 +620,8 @@ __nf_conntrack_find(const struct nf_conntrack_tuple *tuple,
 
 	ASSERT_READ_LOCK(&nf_conntrack_lock);
 	list_for_each_entry(h, &nf_conntrack_hash[hash], list) {
-		if (conntrack_tuple_cmp(h, tuple, ignored_conntrack)) {
+		if (nf_ct_tuplehash_to_ctrack(h) != ignored_conntrack &&
+		    nf_ct_tuple_equal(tuple, &h->tuple)) {
 			NF_CT_STAT_INC(found);
 			return h;
 		}
@@ -667,10 +652,10 @@ static void __nf_conntrack_hash_insert(struct nf_conn *ct,
 				       unsigned int repl_hash) 
 {
 	ct->id = ++nf_conntrack_next_id;
-	list_prepend(&nf_conntrack_hash[hash],
-		     &ct->tuplehash[IP_CT_DIR_ORIGINAL].list);
-	list_prepend(&nf_conntrack_hash[repl_hash],
-		     &ct->tuplehash[IP_CT_DIR_REPLY].list);
+	list_add(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list,
+		 &nf_conntrack_hash[hash]);
+	list_add(&ct->tuplehash[IP_CT_DIR_REPLY].list,
+		 &nf_conntrack_hash[repl_hash]);
 }
 
 void nf_conntrack_hash_insert(struct nf_conn *ct)
@@ -690,7 +675,9 @@ int
 __nf_conntrack_confirm(struct sk_buff **pskb)
 {
 	unsigned int hash, repl_hash;
+	struct nf_conntrack_tuple_hash *h;
 	struct nf_conn *ct;
+	struct nf_conn_help *help;
 	enum ip_conntrack_info ctinfo;
 
 	ct = nf_ct_get(*pskb, &ctinfo);
@@ -720,41 +707,41 @@ __nf_conntrack_confirm(struct sk_buff **pskb)
 	/* See if there's one in the list already, including reverse:
 	   NAT could have grabbed it without realizing, since we're
 	   not in the hash.  If there is, we lost race. */
-	if (!LIST_FIND(&nf_conntrack_hash[hash],
-		       conntrack_tuple_cmp,
-		       struct nf_conntrack_tuple_hash *,
-		       &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple, NULL)
-	    && !LIST_FIND(&nf_conntrack_hash[repl_hash],
-			  conntrack_tuple_cmp,
-			  struct nf_conntrack_tuple_hash *,
-			  &ct->tuplehash[IP_CT_DIR_REPLY].tuple, NULL)) {
-		struct nf_conn_help *help;
-		/* Remove from unconfirmed list */
-		list_del(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list);
+	list_for_each_entry(h, &nf_conntrack_hash[hash], list)
+		if (nf_ct_tuple_equal(&ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple,
+				      &h->tuple))
+			goto out;
+	list_for_each_entry(h, &nf_conntrack_hash[repl_hash], list)
+		if (nf_ct_tuple_equal(&ct->tuplehash[IP_CT_DIR_REPLY].tuple,
+				      &h->tuple))
+			goto out;
 
-		__nf_conntrack_hash_insert(ct, hash, repl_hash);
-		/* Timer relative to confirmation time, not original
-		   setting time, otherwise we'd get timer wrap in
-		   weird delay cases. */
-		ct->timeout.expires += jiffies;
-		add_timer(&ct->timeout);
-		atomic_inc(&ct->ct_general.use);
-		set_bit(IPS_CONFIRMED_BIT, &ct->status);
-		NF_CT_STAT_INC(insert);
-		write_unlock_bh(&nf_conntrack_lock);
-		help = nfct_help(ct);
-		if (help && help->helper)
-			nf_conntrack_event_cache(IPCT_HELPER, *pskb);
+	/* Remove from unconfirmed list */
+	list_del(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list);
+
+	__nf_conntrack_hash_insert(ct, hash, repl_hash);
+	/* Timer relative to confirmation time, not original
+	   setting time, otherwise we'd get timer wrap in
+	   weird delay cases. */
+	ct->timeout.expires += jiffies;
+	add_timer(&ct->timeout);
+	atomic_inc(&ct->ct_general.use);
+	set_bit(IPS_CONFIRMED_BIT, &ct->status);
+	NF_CT_STAT_INC(insert);
+	write_unlock_bh(&nf_conntrack_lock);
+	help = nfct_help(ct);
+	if (help && help->helper)
+		nf_conntrack_event_cache(IPCT_HELPER, *pskb);
 #ifdef CONFIG_NF_NAT_NEEDED
-		if (test_bit(IPS_SRC_NAT_DONE_BIT, &ct->status) ||
-		    test_bit(IPS_DST_NAT_DONE_BIT, &ct->status))
-			nf_conntrack_event_cache(IPCT_NATINFO, *pskb);
+	if (test_bit(IPS_SRC_NAT_DONE_BIT, &ct->status) ||
+	    test_bit(IPS_DST_NAT_DONE_BIT, &ct->status))
+		nf_conntrack_event_cache(IPCT_NATINFO, *pskb);
 #endif
-		nf_conntrack_event_cache(master_ct(ct) ?
-					 IPCT_RELATED : IPCT_NEW, *pskb);
-		return NF_ACCEPT;
-	}
+	nf_conntrack_event_cache(master_ct(ct) ?
+				 IPCT_RELATED : IPCT_NEW, *pskb);
+	return NF_ACCEPT;
 
+out:
 	NF_CT_STAT_INC(insert_failed);
 	write_unlock_bh(&nf_conntrack_lock);
 	return NF_DROP;
@@ -777,24 +764,21 @@ nf_conntrack_tuple_taken(const struct nf_conntrack_tuple *tuple,
 
 /* There's a small race here where we may free a just-assured
    connection.  Too bad: we're in trouble anyway. */
-static inline int unreplied(const struct nf_conntrack_tuple_hash *i)
-{
-	return !(test_bit(IPS_ASSURED_BIT,
-			  &nf_ct_tuplehash_to_ctrack(i)->status));
-}
-
 static int early_drop(struct list_head *chain)
 {
 	/* Traverse backwards: gives us oldest, which is roughly LRU */
 	struct nf_conntrack_tuple_hash *h;
-	struct nf_conn *ct = NULL;
+	struct nf_conn *ct = NULL, *tmp;
 	int dropped = 0;
 
 	read_lock_bh(&nf_conntrack_lock);
-	h = LIST_FIND_B(chain, unreplied, struct nf_conntrack_tuple_hash *);
-	if (h) {
-		ct = nf_ct_tuplehash_to_ctrack(h);
-		atomic_inc(&ct->ct_general.use);
+	list_for_each_entry_reverse(h, chain, list) {
+		tmp = nf_ct_tuplehash_to_ctrack(h);
+		if (!test_bit(IPS_ASSURED_BIT, &tmp->status)) {
+			ct = tmp;
+			atomic_inc(&ct->ct_general.use);
+			break;
+		}
 	}
 	read_unlock_bh(&nf_conntrack_lock);
 
@@ -810,18 +794,16 @@ static int early_drop(struct list_head *chain)
 	return dropped;
 }
 
-static inline int helper_cmp(const struct nf_conntrack_helper *i,
-			     const struct nf_conntrack_tuple *rtuple)
-{
-	return nf_ct_tuple_mask_cmp(rtuple, &i->tuple, &i->mask);
-}
-
 static struct nf_conntrack_helper *
 __nf_ct_helper_find(const struct nf_conntrack_tuple *tuple)
 {
-	return LIST_FIND(&helpers, helper_cmp,
-			 struct nf_conntrack_helper *,
-			 tuple);
+	struct nf_conntrack_helper *h;
+
+	list_for_each_entry(h, &helpers, list) {
+		if (nf_ct_tuple_mask_cmp(tuple, &h->tuple, &h->mask))
+			return h;
+	}
+	return NULL;
 }
 
 struct nf_conntrack_helper *
@@ -1323,7 +1305,7 @@ int nf_conntrack_helper_register(struct nf_conntrack_helper *me)
 		return ret;
 	}
 	write_lock_bh(&nf_conntrack_lock);
-	list_prepend(&helpers, me);
+	list_add(&me->list, &helpers);
 	write_unlock_bh(&nf_conntrack_lock);
 
 	return 0;
@@ -1342,8 +1324,8 @@ __nf_conntrack_helper_find_byname(const char *name)
 	return NULL;
 }
 
-static inline int unhelp(struct nf_conntrack_tuple_hash *i,
-			 const struct nf_conntrack_helper *me)
+static inline void unhelp(struct nf_conntrack_tuple_hash *i,
+			  const struct nf_conntrack_helper *me)
 {
 	struct nf_conn *ct = nf_ct_tuplehash_to_ctrack(i);
 	struct nf_conn_help *help = nfct_help(ct);
@@ -1352,17 +1334,17 @@ static inline int unhelp(struct nf_conntrack_tuple_hash *i,
 		nf_conntrack_event(IPCT_HELPER, ct);
 		help->helper = NULL;
 	}
-	return 0;
 }
 
 void nf_conntrack_helper_unregister(struct nf_conntrack_helper *me)
 {
 	unsigned int i;
+	struct nf_conntrack_tuple_hash *h;
 	struct nf_conntrack_expect *exp, *tmp;
 
 	/* Need write lock here, to delete helper. */
 	write_lock_bh(&nf_conntrack_lock);
-	LIST_DELETE(&helpers, me);
+	list_del(&me->list);
 
 	/* Get rid of expectations */
 	list_for_each_entry_safe(exp, tmp, &nf_conntrack_expect_list, list) {
@@ -1374,10 +1356,12 @@ void nf_conntrack_helper_unregister(struct nf_conntrack_helper *me)
 	}
 
 	/* Get rid of expecteds, set helpers to NULL. */
-	LIST_FIND_W(&unconfirmed, unhelp, struct nf_conntrack_tuple_hash*, me);
-	for (i = 0; i < nf_conntrack_htable_size; i++)
-		LIST_FIND_W(&nf_conntrack_hash[i], unhelp,
-			    struct nf_conntrack_tuple_hash *, me);
+	list_for_each_entry(h, &unconfirmed, list)
+		unhelp(h, me);
+	for (i = 0; i < nf_conntrack_htable_size; i++) {
+		list_for_each_entry(h, &nf_conntrack_hash[i], list)
+			unhelp(h, me);
+	}
 	write_unlock_bh(&nf_conntrack_lock);
 
 	/* Someone could be still looking at the helper in a bh. */
@@ -1510,37 +1494,40 @@ do_iter(const struct nf_conntrack_tuple_hash *i,
 }
 
 /* Bring out ya dead! */
-static struct nf_conntrack_tuple_hash *
+static struct nf_conn *
 get_next_corpse(int (*iter)(struct nf_conn *i, void *data),
 		void *data, unsigned int *bucket)
 {
-	struct nf_conntrack_tuple_hash *h = NULL;
+	struct nf_conntrack_tuple_hash *h;
+	struct nf_conn *ct;
 
 	write_lock_bh(&nf_conntrack_lock);
 	for (; *bucket < nf_conntrack_htable_size; (*bucket)++) {
-		h = LIST_FIND_W(&nf_conntrack_hash[*bucket], do_iter,
-				struct nf_conntrack_tuple_hash *, iter, data);
-		if (h)
-			break;
+		list_for_each_entry(h, &nf_conntrack_hash[*bucket], list) {
+			ct = nf_ct_tuplehash_to_ctrack(h);
+			if (iter(ct, data))
+				goto found;
+		}
  	}
-	if (!h)
-		h = LIST_FIND_W(&unconfirmed, do_iter,
-				struct nf_conntrack_tuple_hash *, iter, data);
-	if (h)
-		atomic_inc(&nf_ct_tuplehash_to_ctrack(h)->ct_general.use);
+	list_for_each_entry(h, &unconfirmed, list) {
+		ct = nf_ct_tuplehash_to_ctrack(h);
+		if (iter(ct, data))
+			goto found;
+	}
+	return NULL;
+found:
+	atomic_inc(&nf_ct_tuplehash_to_ctrack(h)->ct_general.use);
 	write_unlock_bh(&nf_conntrack_lock);
-
-	return h;
+	return ct;
 }
 
 void
 nf_ct_iterate_cleanup(int (*iter)(struct nf_conn *i, void *data), void *data)
 {
-	struct nf_conntrack_tuple_hash *h;
+	struct nf_conn *ct;
 	unsigned int bucket = 0;
 
-	while ((h = get_next_corpse(iter, data, &bucket)) != NULL) {
-		struct nf_conn *ct = nf_ct_tuplehash_to_ctrack(h);
+	while ((ct = get_next_corpse(iter, data, &bucket)) != NULL) {
 		/* Time to push up daises... */
 		if (del_timer(&ct->timeout))
 			death_by_timeout((unsigned long)ct);
diff --git a/net/netfilter/nf_conntrack_standalone.c b/net/netfilter/nf_conntrack_standalone.c
index 9a1de0ca475b..5954f6773810 100644
--- a/net/netfilter/nf_conntrack_standalone.c
+++ b/net/netfilter/nf_conntrack_standalone.c
@@ -37,7 +37,6 @@
 #include <net/netfilter/nf_conntrack_protocol.h>
 #include <net/netfilter/nf_conntrack_core.h>
 #include <net/netfilter/nf_conntrack_helper.h>
-#include <linux/netfilter_ipv4/listhelp.h>
 
 #if 0
 #define DEBUGP printk
diff --git a/net/netfilter/x_tables.c b/net/netfilter/x_tables.c
index 8037ba63d587..be7baf4f6846 100644
--- a/net/netfilter/x_tables.c
+++ b/net/netfilter/x_tables.c
@@ -81,7 +81,7 @@ xt_unregister_target(struct xt_target *target)
 	int af = target->family;
 
 	mutex_lock(&xt[af].mutex);
-	LIST_DELETE(&xt[af].target, target);
+	list_del(&target->list);
 	mutex_unlock(&xt[af].mutex);
 }
 EXPORT_SYMBOL(xt_unregister_target);
@@ -138,7 +138,7 @@ xt_unregister_match(struct xt_match *match)
 	int af =  match->family;
 
 	mutex_lock(&xt[af].mutex);
-	LIST_DELETE(&xt[af].match, match);
+	list_del(&match->list);
 	mutex_unlock(&xt[af].mutex);
 }
 EXPORT_SYMBOL(xt_unregister_match);
@@ -575,15 +575,18 @@ int xt_register_table(struct xt_table *table,
 {
 	int ret;
 	struct xt_table_info *private;
+	struct xt_table *t;
 
 	ret = mutex_lock_interruptible(&xt[table->af].mutex);
 	if (ret != 0)
 		return ret;
 
 	/* Don't autoload: we'd eat our tail... */
-	if (list_named_find(&xt[table->af].tables, table->name)) {
-		ret = -EEXIST;
-		goto unlock;
+	list_for_each_entry(t, &xt[table->af].tables, list) {
+		if (strcmp(t->name, table->name) == 0) {
+			ret = -EEXIST;
+			goto unlock;
+		}
 	}
 
 	/* Simplifies replace_table code. */
@@ -598,7 +601,7 @@ int xt_register_table(struct xt_table *table,
 	/* save number of initial entries */
 	private->initial_entries = private->number;
 
-	list_prepend(&xt[table->af].tables, table);
+	list_add(&table->list, &xt[table->af].tables);
 
 	ret = 0;
  unlock:
@@ -613,7 +616,7 @@ void *xt_unregister_table(struct xt_table *table)
 
 	mutex_lock(&xt[table->af].mutex);
 	private = table->private;
-	LIST_DELETE(&xt[table->af].tables, table);
+	list_del(&table->list);
 	mutex_unlock(&xt[table->af].mutex);
 
 	return private;
-- 
cgit v1.2.3


From 9123de2c043996050bacf77031cad845f5976f5d Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Wed, 20 Sep 2006 11:59:42 -0700
Subject: [NETFILTER]: ip6table_mangle: reroute when nfmark changes in
 NF_IP6_LOCAL_OUT

Now that IPv6 supports policy routing we need to reroute in NF_IP6_LOCAL_OUT
when the mark value changes.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter_ipv6.h       | 1 +
 include/net/ip6_route.h              | 2 --
 net/ipv6/netfilter/ip6table_mangle.c | 8 ++------
 3 files changed, 3 insertions(+), 8 deletions(-)

(limited to 'include')

diff --git a/include/linux/netfilter_ipv6.h b/include/linux/netfilter_ipv6.h
index 52a7b9e76428..d97e268cdfe5 100644
--- a/include/linux/netfilter_ipv6.h
+++ b/include/linux/netfilter_ipv6.h
@@ -73,6 +73,7 @@ enum nf_ip6_hook_priorities {
 };
 
 #ifdef CONFIG_NETFILTER
+extern int ip6_route_me_harder(struct sk_buff *skb);
 extern unsigned int nf_ip6_checksum(struct sk_buff *skb, unsigned int hook,
 				    unsigned int dataoff, u_int8_t protocol);
 
diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index 297909570041..6ca6b71dfe0f 100644
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -57,8 +57,6 @@ extern void			ip6_route_input(struct sk_buff *skb);
 extern struct dst_entry *	ip6_route_output(struct sock *sk,
 						 struct flowi *fl);
 
-extern int			ip6_route_me_harder(struct sk_buff *skb);
-
 extern void			ip6_route_init(void);
 extern void			ip6_route_cleanup(void);
 
diff --git a/net/ipv6/netfilter/ip6table_mangle.c b/net/ipv6/netfilter/ip6table_mangle.c
index 32db04fd8310..386ea260e767 100644
--- a/net/ipv6/netfilter/ip6table_mangle.c
+++ b/net/ipv6/netfilter/ip6table_mangle.c
@@ -180,12 +180,8 @@ ip6t_local_hook(unsigned int hook,
 		&& (memcmp(&(*pskb)->nh.ipv6h->saddr, &saddr, sizeof(saddr))
 		    || memcmp(&(*pskb)->nh.ipv6h->daddr, &daddr, sizeof(daddr))
 		    || (*pskb)->nfmark != nfmark
-		    || (*pskb)->nh.ipv6h->hop_limit != hop_limit)) {
-
-		/* something which could affect routing has changed */
-
-		DEBUGP("ip6table_mangle: we'd need to re-route a packet\n");
-	}
+		    || (*pskb)->nh.ipv6h->hop_limit != hop_limit))
+		return ip6_route_me_harder(*pskb) == 0 ? ret : NF_DROP;
 
 	return ret;
 }
-- 
cgit v1.2.3


From c1fe3ca5106d9568791433fa6c7f27e71ac69e1b Mon Sep 17 00:00:00 2001
From: George Hansper <georgeh@anstat.com.au>
Date: Wed, 20 Sep 2006 12:03:23 -0700
Subject: [NETFILTER]: TCP conntrack: improve dead connection detection

Don't count window updates as retransmissions.

Signed-off-by: George Hansper <georgeh@anstat.com.au>
Signed-off-by: Patrick McHardy <kaber@trash.net>
---
 include/linux/netfilter/nf_conntrack_tcp.h  | 1 +
 net/ipv4/netfilter/ip_conntrack_proto_tcp.c | 4 +++-
 net/netfilter/nf_conntrack_proto_tcp.c      | 4 +++-
 3 files changed, 7 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/include/linux/netfilter/nf_conntrack_tcp.h b/include/linux/netfilter/nf_conntrack_tcp.h
index b2feeffde384..6b01ba297727 100644
--- a/include/linux/netfilter/nf_conntrack_tcp.h
+++ b/include/linux/netfilter/nf_conntrack_tcp.h
@@ -49,6 +49,7 @@ struct ip_ct_tcp
 	u_int32_t	last_seq;	/* Last sequence number seen in dir */
 	u_int32_t	last_ack;	/* Last sequence number seen in opposite dir */
 	u_int32_t	last_end;	/* Last seq + len */
+	u_int16_t	last_win;	/* Last window advertisement seen in dir */
 };
 
 #endif /* __KERNEL__ */
diff --git a/net/ipv4/netfilter/ip_conntrack_proto_tcp.c b/net/ipv4/netfilter/ip_conntrack_proto_tcp.c
index 75a7237eb8c1..03ae9a04cb37 100644
--- a/net/ipv4/netfilter/ip_conntrack_proto_tcp.c
+++ b/net/ipv4/netfilter/ip_conntrack_proto_tcp.c
@@ -731,13 +731,15 @@ static int tcp_in_window(struct ip_ct_tcp *state,
 			if (state->last_dir == dir
 			    && state->last_seq == seq
 			    && state->last_ack == ack
-			    && state->last_end == end)
+			    && state->last_end == end
+			    && state->last_win == win)
 				state->retrans++;
 			else {
 				state->last_dir = dir;
 				state->last_seq = seq;
 				state->last_ack = ack;
 				state->last_end = end;
+				state->last_win = win;
 				state->retrans = 0;
 			}
 		}
diff --git a/net/netfilter/nf_conntrack_proto_tcp.c b/net/netfilter/nf_conntrack_proto_tcp.c
index 9fc0ee61f92a..238bbb5b72ef 100644
--- a/net/netfilter/nf_conntrack_proto_tcp.c
+++ b/net/netfilter/nf_conntrack_proto_tcp.c
@@ -688,13 +688,15 @@ static int tcp_in_window(struct ip_ct_tcp *state,
 			if (state->last_dir == dir
 			    && state->last_seq == seq
 			    && state->last_ack == ack
-			    && state->last_end == end)
+			    && state->last_end == end
+			    && state->last_win == win)
 				state->retrans++;
 			else {
 				state->last_dir = dir;
 				state->last_seq = seq;
 				state->last_ack = ack;
 				state->last_end = end;
+				state->last_win = win;
 				state->retrans = 0;
 			}
 		}
-- 
cgit v1.2.3


From 9fa492cdc160cd27ce1046cb36f47d3b2b1efa21 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Wed, 20 Sep 2006 12:05:37 -0700
Subject: [NETFILTER]: x_tables: simplify compat API

Split the xt_compat_match/xt_compat_target into smaller type-safe functions
performing just one operation. Handle all alignment and size-related
conversions centrally in these function instead of requiring each module to
implement a full-blown conversion function. Replace ->compat callback by
->compat_from_user and ->compat_to_user callbacks, responsible for
converting just a single private structure.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/x_tables.h |  29 +++---
 net/ipv4/netfilter/ip_tables.c     | 115 +++++-----------------
 net/netfilter/x_tables.c           | 192 +++++++++++++++++++++----------------
 3 files changed, 151 insertions(+), 185 deletions(-)

(limited to 'include')

diff --git a/include/linux/netfilter/x_tables.h b/include/linux/netfilter/x_tables.h
index c832295dbf61..739a98eebe2c 100644
--- a/include/linux/netfilter/x_tables.h
+++ b/include/linux/netfilter/x_tables.h
@@ -138,12 +138,6 @@ struct xt_counters_info
 
 #include <linux/netdevice.h>
 
-#ifdef CONFIG_COMPAT
-#define COMPAT_TO_USER		1
-#define COMPAT_FROM_USER	-1
-#define COMPAT_CALC_SIZE	0
-#endif
-
 struct xt_match
 {
 	struct list_head list;
@@ -176,7 +170,8 @@ struct xt_match
 	void (*destroy)(const struct xt_match *match, void *matchinfo);
 
 	/* Called when userspace align differs from kernel space one */
-	int (*compat)(void *match, void **dstptr, int *size, int convert);
+	void (*compat_from_user)(void *dst, void *src);
+	int (*compat_to_user)(void __user *dst, void *src);
 
 	/* Set this to THIS_MODULE if you are a module, otherwise NULL */
 	struct module *me;
@@ -186,6 +181,7 @@ struct xt_match
 
 	char *table;
 	unsigned int matchsize;
+	unsigned int compatsize;
 	unsigned int hooks;
 	unsigned short proto;
 
@@ -224,13 +220,15 @@ struct xt_target
 	void (*destroy)(const struct xt_target *target, void *targinfo);
 
 	/* Called when userspace align differs from kernel space one */
-	int (*compat)(void *target, void **dstptr, int *size, int convert);
+	void (*compat_from_user)(void *dst, void *src);
+	int (*compat_to_user)(void __user *dst, void *src);
 
 	/* Set this to THIS_MODULE if you are a module, otherwise NULL */
 	struct module *me;
 
 	char *table;
 	unsigned int targetsize;
+	unsigned int compatsize;
 	unsigned int hooks;
 	unsigned short proto;
 
@@ -387,9 +385,18 @@ struct compat_xt_counters_info
 
 extern void xt_compat_lock(int af);
 extern void xt_compat_unlock(int af);
-extern int xt_compat_match(void *match, void **dstptr, int *size, int convert);
-extern int xt_compat_target(void *target, void **dstptr, int *size,
-		int convert);
+
+extern int xt_compat_match_offset(struct xt_match *match);
+extern void xt_compat_match_from_user(struct xt_entry_match *m,
+				      void **dstptr, int *size);
+extern int xt_compat_match_to_user(struct xt_entry_match *m,
+				   void * __user *dstptr, int *size);
+
+extern int xt_compat_target_offset(struct xt_target *target);
+extern void xt_compat_target_from_user(struct xt_entry_target *t,
+				       void **dstptr, int *size);
+extern int xt_compat_target_to_user(struct xt_entry_target *t,
+				    void * __user *dstptr, int *size);
 
 #endif /* CONFIG_COMPAT */
 #endif /* __KERNEL__ */
diff --git a/net/ipv4/netfilter/ip_tables.c b/net/ipv4/netfilter/ip_tables.c
index 673581db986e..800067d69a9a 100644
--- a/net/ipv4/netfilter/ip_tables.c
+++ b/net/ipv4/netfilter/ip_tables.c
@@ -942,73 +942,28 @@ static short compat_calc_jump(u_int16_t offset)
 	return delta;
 }
 
-struct compat_ipt_standard_target
+static void compat_standard_from_user(void *dst, void *src)
 {
-	struct compat_xt_entry_target target;
-	compat_int_t verdict;
-};
-
-struct compat_ipt_standard
-{
-	struct compat_ipt_entry entry;
-	struct compat_ipt_standard_target target;
-};
+	int v = *(compat_int_t *)src;
 
-#define IPT_ST_LEN		XT_ALIGN(sizeof(struct ipt_standard_target))
-#define IPT_ST_COMPAT_LEN	COMPAT_XT_ALIGN(sizeof(struct compat_ipt_standard_target))
-#define IPT_ST_OFFSET		(IPT_ST_LEN - IPT_ST_COMPAT_LEN)
+	if (v > 0)
+		v += compat_calc_jump(v);
+	memcpy(dst, &v, sizeof(v));
+}
 
-static int compat_ipt_standard_fn(void *target,
-		void **dstptr, int *size, int convert)
+static int compat_standard_to_user(void __user *dst, void *src)
 {
-	struct compat_ipt_standard_target compat_st, *pcompat_st;
-	struct ipt_standard_target st, *pst;
-	int ret;
+	compat_int_t cv = *(int *)src;
 
-	ret = 0;
-	switch (convert) {
-		case COMPAT_TO_USER:
-			pst = target;
-			memcpy(&compat_st.target, &pst->target,
-				sizeof(compat_st.target));
-			compat_st.verdict = pst->verdict;
-			if (compat_st.verdict > 0)
-				compat_st.verdict -=
-					compat_calc_jump(compat_st.verdict);
-			compat_st.target.u.user.target_size = IPT_ST_COMPAT_LEN;
-			if (copy_to_user(*dstptr, &compat_st, IPT_ST_COMPAT_LEN))
-				ret = -EFAULT;
-			*size -= IPT_ST_OFFSET;
-			*dstptr += IPT_ST_COMPAT_LEN;
-			break;
-		case COMPAT_FROM_USER:
-			pcompat_st = target;
-			memcpy(&st.target, &pcompat_st->target, IPT_ST_COMPAT_LEN);
-			st.verdict = pcompat_st->verdict;
-			if (st.verdict > 0)
-				st.verdict += compat_calc_jump(st.verdict);
-			st.target.u.user.target_size = IPT_ST_LEN;
-			memcpy(*dstptr, &st, IPT_ST_LEN);
-			*size += IPT_ST_OFFSET;
-			*dstptr += IPT_ST_LEN;
-			break;
-		case COMPAT_CALC_SIZE:
-			*size += IPT_ST_OFFSET;
-			break;
-		default:
-			ret = -ENOPROTOOPT;
-			break;
-	}
-	return ret;
+	if (cv > 0)
+		cv -= compat_calc_jump(cv);
+	return copy_to_user(dst, &cv, sizeof(cv)) ? -EFAULT : 0;
 }
 
 static inline int
 compat_calc_match(struct ipt_entry_match *m, int * size)
 {
-	if (m->u.kernel.match->compat)
-		m->u.kernel.match->compat(m, NULL, size, COMPAT_CALC_SIZE);
-	else
-		xt_compat_match(m, NULL, size, COMPAT_CALC_SIZE);
+	*size += xt_compat_match_offset(m->u.kernel.match);
 	return 0;
 }
 
@@ -1023,10 +978,7 @@ static int compat_calc_entry(struct ipt_entry *e, struct xt_table_info *info,
 	entry_offset = (void *)e - base;
 	IPT_MATCH_ITERATE(e, compat_calc_match, &off);
 	t = ipt_get_target(e);
-	if (t->u.kernel.target->compat)
-		t->u.kernel.target->compat(t, NULL, &off, COMPAT_CALC_SIZE);
-	else
-		xt_compat_target(t, NULL, &off, COMPAT_CALC_SIZE);
+	off += xt_compat_target_offset(t->u.kernel.target);
 	newinfo->size -= off;
 	ret = compat_add_offset(entry_offset, off);
 	if (ret)
@@ -1412,17 +1364,13 @@ struct compat_ipt_replace {
 };
 
 static inline int compat_copy_match_to_user(struct ipt_entry_match *m,
-		void __user **dstptr, compat_uint_t *size)
+		void * __user *dstptr, compat_uint_t *size)
 {
-	if (m->u.kernel.match->compat)
-		return m->u.kernel.match->compat(m, dstptr, size,
-				COMPAT_TO_USER);
-	else
-		return xt_compat_match(m, dstptr, size, COMPAT_TO_USER);
+	return xt_compat_match_to_user(m, dstptr, size);
 }
 
 static int compat_copy_entry_to_user(struct ipt_entry *e,
-		void __user **dstptr, compat_uint_t *size)
+		void * __user *dstptr, compat_uint_t *size)
 {
 	struct ipt_entry_target __user *t;
 	struct compat_ipt_entry __user *ce;
@@ -1442,11 +1390,7 @@ static int compat_copy_entry_to_user(struct ipt_entry *e,
 	if (ret)
 		goto out;
 	t = ipt_get_target(e);
-	if (t->u.kernel.target->compat)
-		ret = t->u.kernel.target->compat(t, dstptr, size,
-				COMPAT_TO_USER);
-	else
-		ret = xt_compat_target(t, dstptr, size, COMPAT_TO_USER);
+	ret = xt_compat_target_to_user(t, dstptr, size);
 	if (ret)
 		goto out;
 	ret = -EFAULT;
@@ -1478,11 +1422,7 @@ compat_check_calc_match(struct ipt_entry_match *m,
 		return match ? PTR_ERR(match) : -ENOENT;
 	}
 	m->u.kernel.match = match;
-
-	if (m->u.kernel.match->compat)
-		m->u.kernel.match->compat(m, NULL, size, COMPAT_CALC_SIZE);
-	else
-		xt_compat_match(m, NULL, size, COMPAT_CALC_SIZE);
+	*size += xt_compat_match_offset(match);
 
 	(*i)++;
 	return 0;
@@ -1543,10 +1483,7 @@ check_compat_entry_size_and_hooks(struct ipt_entry *e,
 	}
 	t->u.kernel.target = target;
 
-	if (t->u.kernel.target->compat)
-		t->u.kernel.target->compat(t, NULL, &off, COMPAT_CALC_SIZE);
-	else
-		xt_compat_target(t, NULL, &off, COMPAT_CALC_SIZE);
+	off += xt_compat_target_offset(target);
 	*size += off;
 	ret = compat_add_offset(entry_offset, off);
 	if (ret)
@@ -1584,10 +1521,7 @@ static inline int compat_copy_match_from_user(struct ipt_entry_match *m,
 
 	dm = (struct ipt_entry_match *)*dstptr;
 	match = m->u.kernel.match;
-	if (match->compat)
-		match->compat(m, dstptr, size, COMPAT_FROM_USER);
-	else
-		xt_compat_match(m, dstptr, size, COMPAT_FROM_USER);
+	xt_compat_match_from_user(m, dstptr, size);
 
 	ret = xt_check_match(match, AF_INET, dm->u.match_size - sizeof(*dm),
 			     name, hookmask, ip->proto,
@@ -1635,10 +1569,7 @@ static int compat_copy_entry_from_user(struct ipt_entry *e, void **dstptr,
 	de->target_offset = e->target_offset - (origsize - *size);
 	t = ipt_get_target(e);
 	target = t->u.kernel.target;
-	if (target->compat)
-		target->compat(t, dstptr, size, COMPAT_FROM_USER);
-	else
-		xt_compat_target(t, dstptr, size, COMPAT_FROM_USER);
+	xt_compat_target_from_user(t, dstptr, size);
 
 	de->next_offset = e->next_offset - (origsize - *size);
 	for (h = 0; h < NF_IP_NUMHOOKS; h++) {
@@ -2205,7 +2136,9 @@ static struct ipt_target ipt_standard_target = {
 	.targetsize	= sizeof(int),
 	.family		= AF_INET,
 #ifdef CONFIG_COMPAT
-	.compat		= &compat_ipt_standard_fn,
+	.compatsize	= sizeof(compat_int_t),
+	.compat_from_user = compat_standard_from_user,
+	.compat_to_user	= compat_standard_to_user,
 #endif
 };
 
diff --git a/net/netfilter/x_tables.c b/net/netfilter/x_tables.c
index be7baf4f6846..58522fc65d33 100644
--- a/net/netfilter/x_tables.c
+++ b/net/netfilter/x_tables.c
@@ -333,52 +333,65 @@ int xt_check_match(const struct xt_match *match, unsigned short family,
 EXPORT_SYMBOL_GPL(xt_check_match);
 
 #ifdef CONFIG_COMPAT
-int xt_compat_match(void *match, void **dstptr, int *size, int convert)
+int xt_compat_match_offset(struct xt_match *match)
 {
-	struct xt_match *m;
-	struct compat_xt_entry_match *pcompat_m;
-	struct xt_entry_match *pm;
-	u_int16_t msize;
-	int off, ret;
+	u_int16_t csize = match->compatsize ? : match->matchsize;
+	return XT_ALIGN(match->matchsize) - COMPAT_XT_ALIGN(csize);
+}
+EXPORT_SYMBOL_GPL(xt_compat_match_offset);
 
-	ret = 0;
-	m = ((struct xt_entry_match *)match)->u.kernel.match;
-	off = XT_ALIGN(m->matchsize) - COMPAT_XT_ALIGN(m->matchsize);
-	switch (convert) {
-		case COMPAT_TO_USER:
-			pm = (struct xt_entry_match *)match;
-			msize = pm->u.user.match_size;
-			if (copy_to_user(*dstptr, pm, msize)) {
-				ret = -EFAULT;
-				break;
-			}
-			msize -= off;
-			if (put_user(msize, (u_int16_t *)*dstptr))
-				ret = -EFAULT;
-			*size -= off;
-			*dstptr += msize;
-			break;
-		case COMPAT_FROM_USER:
-			pcompat_m = (struct compat_xt_entry_match *)match;
-			pm = (struct xt_entry_match *)*dstptr;
-			msize = pcompat_m->u.user.match_size;
-			memcpy(pm, pcompat_m, msize);
-			msize += off;
-			pm->u.user.match_size = msize;
-			*size += off;
-			*dstptr += msize;
-			break;
-		case COMPAT_CALC_SIZE:
-			*size += off;
-			break;
-		default:
-			ret = -ENOPROTOOPT;
-			break;
+void xt_compat_match_from_user(struct xt_entry_match *m, void **dstptr,
+			       int *size)
+{
+	struct xt_match *match = m->u.kernel.match;
+	struct compat_xt_entry_match *cm = (struct compat_xt_entry_match *)m;
+	int pad, off = xt_compat_match_offset(match);
+	u_int16_t msize = cm->u.user.match_size;
+
+	m = *dstptr;
+	memcpy(m, cm, sizeof(*cm));
+	if (match->compat_from_user)
+		match->compat_from_user(m->data, cm->data);
+	else
+		memcpy(m->data, cm->data, msize - sizeof(*cm));
+	pad = XT_ALIGN(match->matchsize) - match->matchsize;
+	if (pad > 0)
+		memset(m->data + match->matchsize, 0, pad);
+
+	msize += off;
+	m->u.user.match_size = msize;
+
+	*size += off;
+	*dstptr += msize;
+}
+EXPORT_SYMBOL_GPL(xt_compat_match_from_user);
+
+int xt_compat_match_to_user(struct xt_entry_match *m, void __user **dstptr,
+			    int *size)
+{
+	struct xt_match *match = m->u.kernel.match;
+	struct compat_xt_entry_match __user *cm = *dstptr;
+	int off = xt_compat_match_offset(match);
+	u_int16_t msize = m->u.user.match_size - off;
+
+	if (copy_to_user(cm, m, sizeof(*cm)) ||
+	    put_user(msize, &cm->u.user.match_size))
+	    	return -EFAULT;
+
+	if (match->compat_to_user) {
+		if (match->compat_to_user((void __user *)cm->data, m->data))
+			return -EFAULT;
+	} else {
+		if (copy_to_user(cm->data, m->data, msize - sizeof(*cm)))
+			return -EFAULT;
 	}
-	return ret;
+
+	*size -= off;
+	*dstptr += msize;
+	return 0;
 }
-EXPORT_SYMBOL_GPL(xt_compat_match);
-#endif
+EXPORT_SYMBOL_GPL(xt_compat_match_to_user);
+#endif /* CONFIG_COMPAT */
 
 int xt_check_target(const struct xt_target *target, unsigned short family,
 		    unsigned int size, const char *table, unsigned int hook_mask,
@@ -410,51 +423,64 @@ int xt_check_target(const struct xt_target *target, unsigned short family,
 EXPORT_SYMBOL_GPL(xt_check_target);
 
 #ifdef CONFIG_COMPAT
-int xt_compat_target(void *target, void **dstptr, int *size, int convert)
+int xt_compat_target_offset(struct xt_target *target)
 {
-	struct xt_target *t;
-	struct compat_xt_entry_target *pcompat;
-	struct xt_entry_target *pt;
-	u_int16_t tsize;
-	int off, ret;
+	u_int16_t csize = target->compatsize ? : target->targetsize;
+	return XT_ALIGN(target->targetsize) - COMPAT_XT_ALIGN(csize);
+}
+EXPORT_SYMBOL_GPL(xt_compat_target_offset);
 
-	ret = 0;
-	t = ((struct xt_entry_target *)target)->u.kernel.target;
-	off = XT_ALIGN(t->targetsize) - COMPAT_XT_ALIGN(t->targetsize);
-	switch (convert) {
-		case COMPAT_TO_USER:
-			pt = (struct xt_entry_target *)target;
-			tsize = pt->u.user.target_size;
-			if (copy_to_user(*dstptr, pt, tsize)) {
-				ret = -EFAULT;
-				break;
-			}
-			tsize -= off;
-			if (put_user(tsize, (u_int16_t *)*dstptr))
-				ret = -EFAULT;
-			*size -= off;
-			*dstptr += tsize;
-			break;
-		case COMPAT_FROM_USER:
-			pcompat = (struct compat_xt_entry_target *)target;
-			pt = (struct xt_entry_target *)*dstptr;
-			tsize = pcompat->u.user.target_size;
-			memcpy(pt, pcompat, tsize);
-			tsize += off;
-			pt->u.user.target_size = tsize;
-			*size += off;
-			*dstptr += tsize;
-			break;
-		case COMPAT_CALC_SIZE:
-			*size += off;
-			break;
-		default:
-			ret = -ENOPROTOOPT;
-			break;
+void xt_compat_target_from_user(struct xt_entry_target *t, void **dstptr,
+			        int *size)
+{
+	struct xt_target *target = t->u.kernel.target;
+	struct compat_xt_entry_target *ct = (struct compat_xt_entry_target *)t;
+	int pad, off = xt_compat_target_offset(target);
+	u_int16_t tsize = ct->u.user.target_size;
+
+	t = *dstptr;
+	memcpy(t, ct, sizeof(*ct));
+	if (target->compat_from_user)
+		target->compat_from_user(t->data, ct->data);
+	else
+		memcpy(t->data, ct->data, tsize - sizeof(*ct));
+	pad = XT_ALIGN(target->targetsize) - target->targetsize;
+	if (pad > 0)
+		memset(t->data + target->targetsize, 0, pad);
+
+	tsize += off;
+	t->u.user.target_size = tsize;
+
+	*size += off;
+	*dstptr += tsize;
+}
+EXPORT_SYMBOL_GPL(xt_compat_target_from_user);
+
+int xt_compat_target_to_user(struct xt_entry_target *t, void __user **dstptr,
+			     int *size)
+{
+	struct xt_target *target = t->u.kernel.target;
+	struct compat_xt_entry_target __user *ct = *dstptr;
+	int off = xt_compat_target_offset(target);
+	u_int16_t tsize = t->u.user.target_size - off;
+
+	if (copy_to_user(ct, t, sizeof(*ct)) ||
+	    put_user(tsize, &ct->u.user.target_size))
+	    	return -EFAULT;
+
+	if (target->compat_to_user) {
+		if (target->compat_to_user((void __user *)ct->data, t->data))
+			return -EFAULT;
+	} else {
+		if (copy_to_user(ct->data, t->data, tsize - sizeof(*ct)))
+			return -EFAULT;
 	}
-	return ret;
+
+	*size -= off;
+	*dstptr += tsize;
+	return 0;
 }
-EXPORT_SYMBOL_GPL(xt_compat_target);
+EXPORT_SYMBOL_GPL(xt_compat_target_to_user);
 #endif
 
 struct xt_table_info *xt_alloc_table_info(unsigned int size)
-- 
cgit v1.2.3


From edd5a329cf69c112882e03c8ab55e985062a5d2a Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Wed, 20 Sep 2006 12:07:39 -0700
Subject: [NETFILTER]: PPTP conntrack: fix whitespace errors

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter_ipv4/ip_conntrack_pptp.h | 26 +++----
 net/ipv4/netfilter/ip_conntrack_helper_pptp.c    | 76 ++++++++++----------
 net/ipv4/netfilter/ip_conntrack_proto_gre.c      | 28 ++++----
 net/ipv4/netfilter/ip_nat_helper_pptp.c          | 92 ++++++++++++------------
 net/ipv4/netfilter/ip_nat_proto_gre.c            | 20 +++---
 5 files changed, 121 insertions(+), 121 deletions(-)

(limited to 'include')

diff --git a/include/linux/netfilter_ipv4/ip_conntrack_pptp.h b/include/linux/netfilter_ipv4/ip_conntrack_pptp.h
index 816144c75de0..88f66d3c8765 100644
--- a/include/linux/netfilter_ipv4/ip_conntrack_pptp.h
+++ b/include/linux/netfilter_ipv4/ip_conntrack_pptp.h
@@ -285,19 +285,19 @@ struct PptpSetLinkInfo {
 };
 
 union pptp_ctrl_union {
-		struct PptpStartSessionRequest	sreq;
-		struct PptpStartSessionReply	srep;
-		struct PptpStopSessionRequest	streq;
-		struct PptpStopSessionReply	strep;
-                struct PptpOutCallRequest       ocreq;
-                struct PptpOutCallReply         ocack;
-                struct PptpInCallRequest        icreq;
-                struct PptpInCallReply          icack;
-                struct PptpInCallConnected      iccon;
-		struct PptpClearCallRequest	clrreq;
-                struct PptpCallDisconnectNotify disc;
-                struct PptpWanErrorNotify       wanerr;
-                struct PptpSetLinkInfo          setlink;
+	struct PptpStartSessionRequest	sreq;
+	struct PptpStartSessionReply	srep;
+	struct PptpStopSessionRequest	streq;
+	struct PptpStopSessionReply	strep;
+	struct PptpOutCallRequest	ocreq;
+	struct PptpOutCallReply		ocack;
+	struct PptpInCallRequest	icreq;
+	struct PptpInCallReply		icack;
+	struct PptpInCallConnected	iccon;
+	struct PptpClearCallRequest	clrreq;
+	struct PptpCallDisconnectNotify disc;
+	struct PptpWanErrorNotify	wanerr;
+	struct PptpSetLinkInfo		setlink;
 };
 
 extern int
diff --git a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c
index b020a33e65e9..6c94dd5d476c 100644
--- a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c
+++ b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c
@@ -20,11 +20,11 @@
  * 	 - We can only support one single call within each session
  *
  * TODO:
- *	 - testing of incoming PPTP calls 
+ *	 - testing of incoming PPTP calls
  *
- * Changes: 
+ * Changes:
  * 	2002-02-05 - Version 1.3
- * 	  - Call ip_conntrack_unexpect_related() from 
+ * 	  - Call ip_conntrack_unexpect_related() from
  * 	    pptp_destroy_siblings() to destroy expectations in case
  * 	    CALL_DISCONNECT_NOTIFY or tcp fin packet was seen
  * 	    (Philip Craig <philipc@snapgear.com>)
@@ -141,7 +141,7 @@ static void pptp_expectfn(struct ip_conntrack *ct,
 		invert_tuplepr(&inv_t, &exp->tuple);
 		DEBUGP("trying to unexpect other dir: ");
 		DUMP_TUPLE(&inv_t);
-	
+
 		exp_other = ip_conntrack_expect_find(&inv_t);
 		if (exp_other) {
 			/* delete other expectation.  */
@@ -194,7 +194,7 @@ static void pptp_destroy_siblings(struct ip_conntrack *ct)
 {
 	struct ip_conntrack_tuple t;
 
-	/* Since ct->sibling_list has literally rusted away in 2.6.11, 
+	/* Since ct->sibling_list has literally rusted away in 2.6.11,
 	 * we now need another way to find out about our sibling
 	 * contrack and expects... -HW */
 
@@ -264,7 +264,7 @@ exp_gre(struct ip_conntrack *master,
 	exp_orig->mask.dst.u.gre.key = htons(0xffff);
 	exp_orig->mask.dst.ip = 0xffffffff;
 	exp_orig->mask.dst.protonum = 0xff;
-		
+
 	exp_orig->master = master;
 	exp_orig->expectfn = pptp_expectfn;
 	exp_orig->flags = 0;
@@ -322,7 +322,7 @@ out_unexpect_orig:
 	goto out_put_both;
 }
 
-static inline int 
+static inline int
 pptp_inbound_pkt(struct sk_buff **pskb,
 		 struct tcphdr *tcph,
 		 unsigned int nexthdr_off,
@@ -336,7 +336,7 @@ pptp_inbound_pkt(struct sk_buff **pskb,
 	struct ip_ct_pptp_master *info = &ct->help.ct_pptp_info;
 	u_int16_t msg;
 	__be16 *cid, *pcid;
-	u_int32_t seq;	
+	u_int32_t seq;
 
 	ctlh = skb_header_pointer(*pskb, nexthdr_off, sizeof(_ctlh), &_ctlh);
 	if (!ctlh) {
@@ -373,7 +373,7 @@ pptp_inbound_pkt(struct sk_buff **pskb,
 		}
 		if (pptpReq->srep.resultCode == PPTP_START_OK)
 			info->sstate = PPTP_SESSION_CONFIRMED;
-		else 
+		else
 			info->sstate = PPTP_SESSION_ERROR;
 		break;
 
@@ -420,22 +420,22 @@ pptp_inbound_pkt(struct sk_buff **pskb,
 		pcid = &pptpReq->ocack.peersCallID;
 
 		info->pac_call_id = ntohs(*cid);
-		
+
 		if (htons(info->pns_call_id) != *pcid) {
 			DEBUGP("%s for unknown callid %u\n",
 				pptp_msg_name[msg], ntohs(*pcid));
 			break;
 		}
 
-		DEBUGP("%s, CID=%X, PCID=%X\n", pptp_msg_name[msg], 
+		DEBUGP("%s, CID=%X, PCID=%X\n", pptp_msg_name[msg],
 			ntohs(*cid), ntohs(*pcid));
-		
+
 		info->cstate = PPTP_CALL_OUT_CONF;
 
 		seq = ntohl(tcph->seq) + sizeof(struct pptp_pkt_hdr)
 				       + sizeof(struct PptpControlHeader)
 				       + ((void *)pcid - (void *)pptpReq);
-			
+
 		if (exp_gre(ct, seq, *cid, *pcid) != 0)
 			printk("ip_conntrack_pptp: error during exp_gre\n");
 		break;
@@ -479,7 +479,7 @@ pptp_inbound_pkt(struct sk_buff **pskb,
 		cid = &info->pac_call_id;
 
 		if (info->pns_call_id != ntohs(*pcid)) {
-			DEBUGP("%s for unknown CallID %u\n", 
+			DEBUGP("%s for unknown CallID %u\n",
 				pptp_msg_name[msg], ntohs(*pcid));
 			break;
 		}
@@ -491,7 +491,7 @@ pptp_inbound_pkt(struct sk_buff **pskb,
 		seq = ntohl(tcph->seq) + sizeof(struct pptp_pkt_hdr)
 				       + sizeof(struct PptpControlHeader)
 				       + ((void *)pcid - (void *)pptpReq);
-			
+
 		if (exp_gre(ct, seq, *cid, *pcid) != 0)
 			printk("ip_conntrack_pptp: error during exp_gre\n");
 
@@ -554,7 +554,7 @@ pptp_outbound_pkt(struct sk_buff **pskb,
 		return NF_ACCEPT;
 	nexthdr_off += sizeof(_ctlh);
 	datalen -= sizeof(_ctlh);
-	
+
 	reqlen = datalen;
 	if (reqlen > sizeof(*pptpReq))
 		reqlen = sizeof(*pptpReq);
@@ -606,7 +606,7 @@ pptp_outbound_pkt(struct sk_buff **pskb,
 		/* client answers incoming call */
 		if (info->cstate != PPTP_CALL_IN_REQ
 		    && info->cstate != PPTP_CALL_IN_REP) {
-			DEBUGP("%s without incall_req\n", 
+			DEBUGP("%s without incall_req\n",
 				pptp_msg_name[msg]);
 			break;
 		}
@@ -616,7 +616,7 @@ pptp_outbound_pkt(struct sk_buff **pskb,
 		}
 		pcid = &pptpReq->icack.peersCallID;
 		if (info->pac_call_id != ntohs(*pcid)) {
-			DEBUGP("%s for unknown call %u\n", 
+			DEBUGP("%s for unknown call %u\n",
 				pptp_msg_name[msg], ntohs(*pcid));
 			break;
 		}
@@ -644,12 +644,12 @@ pptp_outbound_pkt(struct sk_buff **pskb,
 		/* I don't have to explain these ;) */
 		break;
 	default:
-		DEBUGP("invalid %s (TY=%d)\n", (msg <= PPTP_MSG_MAX)? 
+		DEBUGP("invalid %s (TY=%d)\n", (msg <= PPTP_MSG_MAX)?
 			pptp_msg_name[msg]:pptp_msg_name[0], msg);
 		/* unknown: no need to create GRE masq table entry */
 		break;
 	}
-	
+
 	if (ip_nat_pptp_hook_outbound)
 		return ip_nat_pptp_hook_outbound(pskb, ct, ctinfo, ctlh,
 						 pptpReq);
@@ -659,7 +659,7 @@ pptp_outbound_pkt(struct sk_buff **pskb,
 
 
 /* track caller id inside control connection, call expect_related */
-static int 
+static int
 conntrack_pptp_help(struct sk_buff **pskb,
 		    struct ip_conntrack *ct, enum ip_conntrack_info ctinfo)
 
@@ -676,12 +676,12 @@ conntrack_pptp_help(struct sk_buff **pskb,
 	int ret;
 
 	/* don't do any tracking before tcp handshake complete */
-	if (ctinfo != IP_CT_ESTABLISHED 
+	if (ctinfo != IP_CT_ESTABLISHED
 	    && ctinfo != IP_CT_ESTABLISHED+IP_CT_IS_REPLY) {
 		DEBUGP("ctinfo = %u, skipping\n", ctinfo);
 		return NF_ACCEPT;
 	}
-	
+
 	nexthdr_off = (*pskb)->nh.iph->ihl*4;
 	tcph = skb_header_pointer(*pskb, nexthdr_off, sizeof(_tcph), &_tcph);
 	BUG_ON(!tcph);
@@ -735,28 +735,28 @@ conntrack_pptp_help(struct sk_buff **pskb,
 }
 
 /* control protocol helper */
-static struct ip_conntrack_helper pptp = { 
+static struct ip_conntrack_helper pptp = {
 	.list = { NULL, NULL },
-	.name = "pptp", 
+	.name = "pptp",
 	.me = THIS_MODULE,
 	.max_expected = 2,
 	.timeout = 5 * 60,
-	.tuple = { .src = { .ip = 0, 
-		 	    .u = { .tcp = { .port =  
-				    __constant_htons(PPTP_CONTROL_PORT) } } 
-			  }, 
-		   .dst = { .ip = 0, 
+	.tuple = { .src = { .ip = 0,
+		 	    .u = { .tcp = { .port =
+				    __constant_htons(PPTP_CONTROL_PORT) } }
+			  },
+		   .dst = { .ip = 0,
 			    .u = { .all = 0 },
 			    .protonum = IPPROTO_TCP
-			  } 
+			  }
 		 },
-	.mask = { .src = { .ip = 0, 
-			   .u = { .tcp = { .port = __constant_htons(0xffff) } } 
-			 }, 
-		  .dst = { .ip = 0, 
+	.mask = { .src = { .ip = 0,
+			   .u = { .tcp = { .port = __constant_htons(0xffff) } }
+			 },
+		  .dst = { .ip = 0,
 			   .u = { .all = 0 },
-			   .protonum = 0xff 
-		 	 } 
+			   .protonum = 0xff
+		 	 }
 		},
 	.help = conntrack_pptp_help
 };
@@ -768,7 +768,7 @@ extern int __init ip_ct_proto_gre_init(void);
 static int __init ip_conntrack_helper_pptp_init(void)
 {
 	int retcode;
- 
+
 	retcode = ip_ct_proto_gre_init();
 	if (retcode < 0)
 		return retcode;
diff --git a/net/ipv4/netfilter/ip_conntrack_proto_gre.c b/net/ipv4/netfilter/ip_conntrack_proto_gre.c
index 92c6d8b178c9..5fe026f467d3 100644
--- a/net/ipv4/netfilter/ip_conntrack_proto_gre.c
+++ b/net/ipv4/netfilter/ip_conntrack_proto_gre.c
@@ -1,15 +1,15 @@
 /*
- * ip_conntrack_proto_gre.c - Version 3.0 
+ * ip_conntrack_proto_gre.c - Version 3.0
  *
  * Connection tracking protocol helper module for GRE.
  *
  * GRE is a generic encapsulation protocol, which is generally not very
  * suited for NAT, as it has no protocol-specific part as port numbers.
  *
- * It has an optional key field, which may help us distinguishing two 
+ * It has an optional key field, which may help us distinguishing two
  * connections between the same two hosts.
  *
- * GRE is defined in RFC 1701 and RFC 1702, as well as RFC 2784 
+ * GRE is defined in RFC 1701 and RFC 1702, as well as RFC 2784
  *
  * PPTP is built on top of a modified version of GRE, and has a mandatory
  * field called "CallID", which serves us for the same purpose as the key
@@ -61,7 +61,7 @@ MODULE_DESCRIPTION("netfilter connection tracking protocol helper for GRE");
 #define DEBUGP(x, args...)
 #define DUMP_TUPLE_GRE(x)
 #endif
-				
+
 /* GRE KEYMAP HANDLING FUNCTIONS */
 static LIST_HEAD(gre_keymap_list);
 
@@ -88,7 +88,7 @@ static __be16 gre_keymap_lookup(struct ip_conntrack_tuple *t)
 		}
 	}
 	read_unlock_bh(&ip_ct_gre_lock);
-	
+
 	DEBUGP("lookup src key 0x%x up key for ", key);
 	DUMP_TUPLE_GRE(t);
 
@@ -107,7 +107,7 @@ ip_ct_gre_keymap_add(struct ip_conntrack *ct,
 		return -1;
 	}
 
-	if (!reply) 
+	if (!reply)
 		exist_km = &ct->help.ct_pptp_info.keymap_orig;
 	else
 		exist_km = &ct->help.ct_pptp_info.keymap_reply;
@@ -118,7 +118,7 @@ ip_ct_gre_keymap_add(struct ip_conntrack *ct,
 			if (gre_key_cmpfn(km, t) && km == *exist_km)
 				return 0;
 		}
-		DEBUGP("trying to override keymap_%s for ct %p\n", 
+		DEBUGP("trying to override keymap_%s for ct %p\n",
 			reply? "reply":"orig", ct);
 		return -EEXIST;
 	}
@@ -152,7 +152,7 @@ void ip_ct_gre_keymap_destroy(struct ip_conntrack *ct)
 
 	write_lock_bh(&ip_ct_gre_lock);
 	if (ct->help.ct_pptp_info.keymap_orig) {
-		DEBUGP("removing %p from list\n", 
+		DEBUGP("removing %p from list\n",
 			ct->help.ct_pptp_info.keymap_orig);
 		list_del(&ct->help.ct_pptp_info.keymap_orig->list);
 		kfree(ct->help.ct_pptp_info.keymap_orig);
@@ -220,7 +220,7 @@ static int gre_pkt_to_tuple(const struct sk_buff *skb,
 static int gre_print_tuple(struct seq_file *s,
 			   const struct ip_conntrack_tuple *tuple)
 {
-	return seq_printf(s, "srckey=0x%x dstkey=0x%x ", 
+	return seq_printf(s, "srckey=0x%x dstkey=0x%x ",
 			  ntohs(tuple->src.u.gre.key),
 			  ntohs(tuple->dst.u.gre.key));
 }
@@ -250,14 +250,14 @@ static int gre_packet(struct ip_conntrack *ct,
 	} else
 		ip_ct_refresh_acct(ct, conntrackinfo, skb,
 				   ct->proto.gre.timeout);
-	
+
 	return NF_ACCEPT;
 }
 
 /* Called when a new connection for this protocol found. */
 static int gre_new(struct ip_conntrack *ct,
 		   const struct sk_buff *skb)
-{ 
+{
 	DEBUGP(": ");
 	DUMP_TUPLE_GRE(&ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple);
 
@@ -283,9 +283,9 @@ static void gre_destroy(struct ip_conntrack *ct)
 }
 
 /* protocol helper struct */
-static struct ip_conntrack_protocol gre = { 
+static struct ip_conntrack_protocol gre = {
 	.proto		 = IPPROTO_GRE,
-	.name		 = "gre", 
+	.name		 = "gre",
 	.pkt_to_tuple	 = gre_pkt_to_tuple,
 	.invert_tuple	 = gre_invert_tuple,
 	.print_tuple	 = gre_print_tuple,
@@ -323,7 +323,7 @@ void ip_ct_proto_gre_fini(void)
 	}
 	write_unlock_bh(&ip_ct_gre_lock);
 
-	ip_conntrack_protocol_unregister(&gre); 
+	ip_conntrack_protocol_unregister(&gre);
 }
 
 EXPORT_SYMBOL(ip_ct_gre_keymap_add);
diff --git a/net/ipv4/netfilter/ip_nat_helper_pptp.c b/net/ipv4/netfilter/ip_nat_helper_pptp.c
index 1d149964dc38..5dde1da1c300 100644
--- a/net/ipv4/netfilter/ip_nat_helper_pptp.c
+++ b/net/ipv4/netfilter/ip_nat_helper_pptp.c
@@ -32,7 +32,7 @@
  *     2005-06-10 - Version 3.0
  *       - kernel >= 2.6.11 version,
  *	   funded by Oxcoda NetBox Blue (http://www.netboxblue.com/)
- * 
+ *
  */
 
 #include <linux/module.h>
@@ -93,10 +93,10 @@ static void pptp_nat_expected(struct ip_conntrack *ct,
 		DEBUGP("we are PAC->PNS\n");
 		/* build tuple for PNS->PAC */
 		t.src.ip = master->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.ip;
-		t.src.u.gre.key = 
+		t.src.u.gre.key =
 			htons(master->nat.help.nat_pptp_info.pns_call_id);
 		t.dst.ip = master->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.ip;
-		t.dst.u.gre.key = 
+		t.dst.u.gre.key =
 			htons(master->nat.help.nat_pptp_info.pac_call_id);
 		t.dst.protonum = IPPROTO_GRE;
 	}
@@ -153,47 +153,47 @@ pptp_outbound_pkt(struct sk_buff **pskb,
 	unsigned int cid_off;
 
 	new_callid = htons(ct_pptp_info->pns_call_id);
-	
+
 	switch (msg = ntohs(ctlh->messageType)) {
-		case PPTP_OUT_CALL_REQUEST:
-			cid_off = offsetof(union pptp_ctrl_union, ocreq.callID);
-			/* FIXME: ideally we would want to reserve a call ID
-			 * here.  current netfilter NAT core is not able to do
-			 * this :( For now we use TCP source port. This breaks
-			 * multiple calls within one control session */
-
-			/* save original call ID in nat_info */
-			nat_pptp_info->pns_call_id = ct_pptp_info->pns_call_id;
-
-			/* don't use tcph->source since we are at a DSTmanip
-			 * hook (e.g. PREROUTING) and pkt is not mangled yet */
-			new_callid = ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.u.tcp.port;
-
-			/* save new call ID in ct info */
-			ct_pptp_info->pns_call_id = ntohs(new_callid);
-			break;
-		case PPTP_IN_CALL_REPLY:
-			cid_off = offsetof(union pptp_ctrl_union, icreq.callID);
-			break;
-		case PPTP_CALL_CLEAR_REQUEST:
-			cid_off = offsetof(union pptp_ctrl_union, clrreq.callID);
-			break;
-		default:
-			DEBUGP("unknown outbound packet 0x%04x:%s\n", msg,
-			      (msg <= PPTP_MSG_MAX)? 
-			      pptp_msg_name[msg]:pptp_msg_name[0]);
-			/* fall through */
-
-		case PPTP_SET_LINK_INFO:
-			/* only need to NAT in case PAC is behind NAT box */
-		case PPTP_START_SESSION_REQUEST:
-		case PPTP_START_SESSION_REPLY:
-		case PPTP_STOP_SESSION_REQUEST:
-		case PPTP_STOP_SESSION_REPLY:
-		case PPTP_ECHO_REQUEST:
-		case PPTP_ECHO_REPLY:
-			/* no need to alter packet */
-			return NF_ACCEPT;
+	case PPTP_OUT_CALL_REQUEST:
+		cid_off = offsetof(union pptp_ctrl_union, ocreq.callID);
+		/* FIXME: ideally we would want to reserve a call ID
+		 * here.  current netfilter NAT core is not able to do
+		 * this :( For now we use TCP source port. This breaks
+		 * multiple calls within one control session */
+
+		/* save original call ID in nat_info */
+		nat_pptp_info->pns_call_id = ct_pptp_info->pns_call_id;
+
+		/* don't use tcph->source since we are at a DSTmanip
+		 * hook (e.g. PREROUTING) and pkt is not mangled yet */
+		new_callid = ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.u.tcp.port;
+
+		/* save new call ID in ct info */
+		ct_pptp_info->pns_call_id = ntohs(new_callid);
+		break;
+	case PPTP_IN_CALL_REPLY:
+		cid_off = offsetof(union pptp_ctrl_union, icreq.callID);
+		break;
+	case PPTP_CALL_CLEAR_REQUEST:
+		cid_off = offsetof(union pptp_ctrl_union, clrreq.callID);
+		break;
+	default:
+		DEBUGP("unknown outbound packet 0x%04x:%s\n", msg,
+		      (msg <= PPTP_MSG_MAX)?
+		      pptp_msg_name[msg]:pptp_msg_name[0]);
+		/* fall through */
+
+	case PPTP_SET_LINK_INFO:
+		/* only need to NAT in case PAC is behind NAT box */
+	case PPTP_START_SESSION_REQUEST:
+	case PPTP_START_SESSION_REPLY:
+	case PPTP_STOP_SESSION_REQUEST:
+	case PPTP_STOP_SESSION_REPLY:
+	case PPTP_ECHO_REQUEST:
+	case PPTP_ECHO_REPLY:
+		/* no need to alter packet */
+		return NF_ACCEPT;
 	}
 
 	/* only OUT_CALL_REQUEST, IN_CALL_REPLY, CALL_CLEAR_REQUEST pass
@@ -216,9 +216,9 @@ static int
 pptp_exp_gre(struct ip_conntrack_expect *expect_orig,
 	     struct ip_conntrack_expect *expect_reply)
 {
-	struct ip_ct_pptp_master *ct_pptp_info = 
+	struct ip_ct_pptp_master *ct_pptp_info =
 				&expect_orig->master->help.ct_pptp_info;
-	struct ip_nat_pptp *nat_pptp_info = 
+	struct ip_nat_pptp *nat_pptp_info =
 				&expect_orig->master->nat.help.nat_pptp_info;
 
 	struct ip_conntrack *ct = expect_orig->master;
@@ -324,7 +324,7 @@ pptp_inbound_pkt(struct sk_buff **pskb,
 		break;
 
 	default:
-		DEBUGP("unknown inbound packet %s\n", (msg <= PPTP_MSG_MAX)? 
+		DEBUGP("unknown inbound packet %s\n", (msg <= PPTP_MSG_MAX)?
 			pptp_msg_name[msg]:pptp_msg_name[0]);
 		/* fall through */
 
diff --git a/net/ipv4/netfilter/ip_nat_proto_gre.c b/net/ipv4/netfilter/ip_nat_proto_gre.c
index 70a65372225a..a5226691f02c 100644
--- a/net/ipv4/netfilter/ip_nat_proto_gre.c
+++ b/net/ipv4/netfilter/ip_nat_proto_gre.c
@@ -6,10 +6,10 @@
  * GRE is a generic encapsulation protocol, which is generally not very
  * suited for NAT, as it has no protocol-specific part as port numbers.
  *
- * It has an optional key field, which may help us distinguishing two 
+ * It has an optional key field, which may help us distinguishing two
  * connections between the same two hosts.
  *
- * GRE is defined in RFC 1701 and RFC 1702, as well as RFC 2784 
+ * GRE is defined in RFC 1701 and RFC 1702, as well as RFC 2784
  *
  * PPTP is built on top of a modified version of GRE, and has a mandatory
  * field called "CallID", which serves us for the same purpose as the key
@@ -60,7 +60,7 @@ gre_in_range(const struct ip_conntrack_tuple *tuple,
 }
 
 /* generate unique tuple ... */
-static int 
+static int
 gre_unique_tuple(struct ip_conntrack_tuple *tuple,
 		 const struct ip_nat_range *range,
 		 enum ip_nat_manip_type maniptype,
@@ -84,7 +84,7 @@ gre_unique_tuple(struct ip_conntrack_tuple *tuple,
 		range_size = ntohs(range->max.gre.key) - min + 1;
 	}
 
-	DEBUGP("min = %u, range_size = %u\n", min, range_size); 
+	DEBUGP("min = %u, range_size = %u\n", min, range_size);
 
 	for (i = 0; i < range_size; i++, key++) {
 		*keyptr = htons(min + key % range_size);
@@ -117,7 +117,7 @@ gre_manip_pkt(struct sk_buff **pskb,
 	greh = (void *)(*pskb)->data + hdroff;
 	pgreh = (struct gre_hdr_pptp *) greh;
 
-	/* we only have destination manip of a packet, since 'source key' 
+	/* we only have destination manip of a packet, since 'source key'
 	 * is not present in the packet itself */
 	if (maniptype == IP_NAT_MANIP_DST) {
 		/* key manipulation is always dest */
@@ -129,7 +129,7 @@ gre_manip_pkt(struct sk_buff **pskb,
 			}
 			if (greh->csum) {
 				/* FIXME: Never tested this code... */
-				*(gre_csum(greh)) = 
+				*(gre_csum(greh)) =
 					nf_proto_csum_update(*pskb,
 							~*(gre_key(greh)),
 							tuple->dst.u.gre.key,
@@ -138,7 +138,7 @@ gre_manip_pkt(struct sk_buff **pskb,
 			*(gre_key(greh)) = tuple->dst.u.gre.key;
 			break;
 		case GRE_VERSION_PPTP:
-			DEBUGP("call_id -> 0x%04x\n", 
+			DEBUGP("call_id -> 0x%04x\n",
 				ntohs(tuple->dst.u.gre.key));
 			pgreh->call_id = tuple->dst.u.gre.key;
 			break;
@@ -152,8 +152,8 @@ gre_manip_pkt(struct sk_buff **pskb,
 }
 
 /* nat helper struct */
-static struct ip_nat_protocol gre = { 
-	.name		= "GRE", 
+static struct ip_nat_protocol gre = {
+	.name		= "GRE",
 	.protonum	= IPPROTO_GRE,
 	.manip_pkt	= gre_manip_pkt,
 	.in_range	= gre_in_range,
@@ -164,7 +164,7 @@ static struct ip_nat_protocol gre = {
 	.nfattr_to_range	= ip_nat_port_nfattr_to_range,
 #endif
 };
-				  
+
 int __init ip_nat_proto_gre_init(void)
 {
 	return ip_nat_protocol_register(&gre);
-- 
cgit v1.2.3


From 955b944293dd4c931ec866ebe19a6b2463b8f9a0 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Wed, 20 Sep 2006 12:08:03 -0700
Subject: [NETFILTER]: PPTP conntrack: get rid of unnecessary byte order
 conversions

The conntrack structure contains the call ID in host byte order for no
reason, get rid of back and forth conversions.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter_ipv4/ip_conntrack_pptp.h   |  8 ++---
 .../linux/netfilter_ipv4/ip_conntrack_proto_gre.h  | 22 ++++++------
 include/linux/netfilter_ipv4/ip_nat_pptp.h         |  4 +--
 net/ipv4/netfilter/ip_conntrack_helper_pptp.c      | 22 ++++++------
 net/ipv4/netfilter/ip_nat_helper_pptp.c            | 42 +++++++++++-----------
 net/ipv4/netfilter/ip_nat_proto_gre.c              |  2 +-
 6 files changed, 50 insertions(+), 50 deletions(-)

(limited to 'include')

diff --git a/include/linux/netfilter_ipv4/ip_conntrack_pptp.h b/include/linux/netfilter_ipv4/ip_conntrack_pptp.h
index 88f66d3c8765..0d35623f9453 100644
--- a/include/linux/netfilter_ipv4/ip_conntrack_pptp.h
+++ b/include/linux/netfilter_ipv4/ip_conntrack_pptp.h
@@ -31,8 +31,8 @@ struct ip_ct_pptp_master {
 	/* everything below is going to be per-expectation in newnat,
 	 * since there could be more than one call within one session */
 	enum pptp_ctrlcall_state cstate;	/* call state */
-	u_int16_t pac_call_id;			/* call id of PAC, host byte order */
-	u_int16_t pns_call_id;			/* call id of PNS, host byte order */
+	__be16 pac_call_id;			/* call id of PAC, host byte order */
+	__be16 pns_call_id;			/* call id of PNS, host byte order */
 
 	/* in pre-2.6.11 this used to be per-expect. Now it is per-conntrack
 	 * and therefore imposes a fixed limit on the number of maps */
@@ -42,8 +42,8 @@ struct ip_ct_pptp_master {
 /* conntrack_expect private member */
 struct ip_ct_pptp_expect {
 	enum pptp_ctrlcall_state cstate; 	/* call state */
-	u_int16_t pac_call_id;			/* call id of PAC */
-	u_int16_t pns_call_id;			/* call id of PNS */
+	__be16 pac_call_id;			/* call id of PAC */
+	__be16 pns_call_id;			/* call id of PNS */
 };
 
 
diff --git a/include/linux/netfilter_ipv4/ip_conntrack_proto_gre.h b/include/linux/netfilter_ipv4/ip_conntrack_proto_gre.h
index 8d090ef82f5f..1d853aa873eb 100644
--- a/include/linux/netfilter_ipv4/ip_conntrack_proto_gre.h
+++ b/include/linux/netfilter_ipv4/ip_conntrack_proto_gre.h
@@ -49,18 +49,18 @@ struct gre_hdr {
 #else
 #error "Adjust your <asm/byteorder.h> defines"
 #endif
-	__u16	protocol;
+	__be16	protocol;
 };
 
 /* modified GRE header for PPTP */
 struct gre_hdr_pptp {
-	__u8  flags;		/* bitfield */
-	__u8  version;		/* should be GRE_VERSION_PPTP */
-	__u16 protocol;		/* should be GRE_PROTOCOL_PPTP */
-	__u16 payload_len;	/* size of ppp payload, not inc. gre header */
-	__u16 call_id;		/* peer's call_id for this session */
-	__u32 seq;		/* sequence number.  Present if S==1 */
-	__u32 ack;		/* seq number of highest packet recieved by */
+	__u8   flags;		/* bitfield */
+	__u8   version;		/* should be GRE_VERSION_PPTP */
+	__be16 protocol;	/* should be GRE_PROTOCOL_PPTP */
+	__be16 payload_len;	/* size of ppp payload, not inc. gre header */
+	__be16 call_id;		/* peer's call_id for this session */
+	__be32 seq;		/* sequence number.  Present if S==1 */
+	__be32 ack;		/* seq number of highest packet recieved by */
 				/*  sender in this session */
 };
 
@@ -92,13 +92,13 @@ void ip_ct_gre_keymap_destroy(struct ip_conntrack *ct);
 
 
 /* get pointer to gre key, if present */
-static inline u_int32_t *gre_key(struct gre_hdr *greh)
+static inline __be32 *gre_key(struct gre_hdr *greh)
 {
 	if (!greh->key)
 		return NULL;
 	if (greh->csum || greh->routing)
-		return (u_int32_t *) (greh+sizeof(*greh)+4);
-	return (u_int32_t *) (greh+sizeof(*greh));
+		return (__be32 *) (greh+sizeof(*greh)+4);
+	return (__be32 *) (greh+sizeof(*greh));
 }
 
 /* get pointer ot gre csum, if present */
diff --git a/include/linux/netfilter_ipv4/ip_nat_pptp.h b/include/linux/netfilter_ipv4/ip_nat_pptp.h
index eaf66c2e8f93..36668bf0f373 100644
--- a/include/linux/netfilter_ipv4/ip_nat_pptp.h
+++ b/include/linux/netfilter_ipv4/ip_nat_pptp.h
@@ -4,8 +4,8 @@
 
 /* conntrack private data */
 struct ip_nat_pptp {
-	u_int16_t pns_call_id;		/* NAT'ed PNS call id */
-	u_int16_t pac_call_id;		/* NAT'ed PAC call id */
+	__be16 pns_call_id;		/* NAT'ed PNS call id */
+	__be16 pac_call_id;		/* NAT'ed PAC call id */
 };
 
 #endif /* _NAT_PPTP_H */
diff --git a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c
index 6c94dd5d476c..57637ca2b82c 100644
--- a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c
+++ b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c
@@ -201,8 +201,8 @@ static void pptp_destroy_siblings(struct ip_conntrack *ct)
 	/* try original (pns->pac) tuple */
 	memcpy(&t, &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple, sizeof(t));
 	t.dst.protonum = IPPROTO_GRE;
-	t.src.u.gre.key = htons(ct->help.ct_pptp_info.pns_call_id);
-	t.dst.u.gre.key = htons(ct->help.ct_pptp_info.pac_call_id);
+	t.src.u.gre.key = ct->help.ct_pptp_info.pns_call_id;
+	t.dst.u.gre.key = ct->help.ct_pptp_info.pac_call_id;
 
 	if (!destroy_sibling_or_exp(&t))
 		DEBUGP("failed to timeout original pns->pac ct/exp\n");
@@ -210,8 +210,8 @@ static void pptp_destroy_siblings(struct ip_conntrack *ct)
 	/* try reply (pac->pns) tuple */
 	memcpy(&t, &ct->tuplehash[IP_CT_DIR_REPLY].tuple, sizeof(t));
 	t.dst.protonum = IPPROTO_GRE;
-	t.src.u.gre.key = htons(ct->help.ct_pptp_info.pac_call_id);
-	t.dst.u.gre.key = htons(ct->help.ct_pptp_info.pns_call_id);
+	t.src.u.gre.key = ct->help.ct_pptp_info.pac_call_id;
+	t.dst.u.gre.key = ct->help.ct_pptp_info.pns_call_id;
 
 	if (!destroy_sibling_or_exp(&t))
 		DEBUGP("failed to timeout reply pac->pns ct/exp\n");
@@ -419,9 +419,9 @@ pptp_inbound_pkt(struct sk_buff **pskb,
 		cid = &pptpReq->ocack.callID;
 		pcid = &pptpReq->ocack.peersCallID;
 
-		info->pac_call_id = ntohs(*cid);
+		info->pac_call_id = *cid;
 
-		if (htons(info->pns_call_id) != *pcid) {
+		if (info->pns_call_id != *pcid) {
 			DEBUGP("%s for unknown callid %u\n",
 				pptp_msg_name[msg], ntohs(*pcid));
 			break;
@@ -454,7 +454,7 @@ pptp_inbound_pkt(struct sk_buff **pskb,
 		pcid = &pptpReq->icack.peersCallID;
 		DEBUGP("%s, PCID=%X\n", pptp_msg_name[msg], ntohs(*pcid));
 		info->cstate = PPTP_CALL_IN_REQ;
-		info->pac_call_id = ntohs(*pcid);
+		info->pac_call_id = *pcid;
 		break;
 
 	case PPTP_IN_CALL_CONNECT:
@@ -478,7 +478,7 @@ pptp_inbound_pkt(struct sk_buff **pskb,
 		pcid = &pptpReq->iccon.peersCallID;
 		cid = &info->pac_call_id;
 
-		if (info->pns_call_id != ntohs(*pcid)) {
+		if (info->pns_call_id != *pcid) {
 			DEBUGP("%s for unknown CallID %u\n",
 				pptp_msg_name[msg], ntohs(*pcid));
 			break;
@@ -595,7 +595,7 @@ pptp_outbound_pkt(struct sk_buff **pskb,
 		/* track PNS call id */
 		cid = &pptpReq->ocreq.callID;
 		DEBUGP("%s, CID=%X\n", pptp_msg_name[msg], ntohs(*cid));
-		info->pns_call_id = ntohs(*cid);
+		info->pns_call_id = *cid;
 		break;
 	case PPTP_IN_CALL_REPLY:
 		if (reqlen < sizeof(_pptpReq.icack)) {
@@ -615,7 +615,7 @@ pptp_outbound_pkt(struct sk_buff **pskb,
 			break;
 		}
 		pcid = &pptpReq->icack.peersCallID;
-		if (info->pac_call_id != ntohs(*pcid)) {
+		if (info->pac_call_id != *pcid) {
 			DEBUGP("%s for unknown call %u\n",
 				pptp_msg_name[msg], ntohs(*pcid));
 			break;
@@ -623,7 +623,7 @@ pptp_outbound_pkt(struct sk_buff **pskb,
 		DEBUGP("%s, CID=%X\n", pptp_msg_name[msg], ntohs(*pcid));
 		/* part two of the three-way handshake */
 		info->cstate = PPTP_CALL_IN_REP;
-		info->pns_call_id = ntohs(pptpReq->icack.callID);
+		info->pns_call_id = pptpReq->icack.callID;
 		break;
 
 	case PPTP_CALL_CLEAR_REQUEST:
diff --git a/net/ipv4/netfilter/ip_nat_helper_pptp.c b/net/ipv4/netfilter/ip_nat_helper_pptp.c
index 5dde1da1c300..6e8bd6b3431f 100644
--- a/net/ipv4/netfilter/ip_nat_helper_pptp.c
+++ b/net/ipv4/netfilter/ip_nat_helper_pptp.c
@@ -85,19 +85,17 @@ static void pptp_nat_expected(struct ip_conntrack *ct,
 		DEBUGP("we are PNS->PAC\n");
 		/* therefore, build tuple for PAC->PNS */
 		t.src.ip = master->tuplehash[IP_CT_DIR_REPLY].tuple.src.ip;
-		t.src.u.gre.key = htons(master->help.ct_pptp_info.pac_call_id);
+		t.src.u.gre.key = master->help.ct_pptp_info.pac_call_id;
 		t.dst.ip = master->tuplehash[IP_CT_DIR_REPLY].tuple.dst.ip;
-		t.dst.u.gre.key = htons(master->help.ct_pptp_info.pns_call_id);
+		t.dst.u.gre.key = master->help.ct_pptp_info.pns_call_id;
 		t.dst.protonum = IPPROTO_GRE;
 	} else {
 		DEBUGP("we are PAC->PNS\n");
 		/* build tuple for PNS->PAC */
 		t.src.ip = master->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.ip;
-		t.src.u.gre.key =
-			htons(master->nat.help.nat_pptp_info.pns_call_id);
+		t.src.u.gre.key = master->nat.help.nat_pptp_info.pns_call_id;
 		t.dst.ip = master->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.ip;
-		t.dst.u.gre.key =
-			htons(master->nat.help.nat_pptp_info.pac_call_id);
+		t.dst.u.gre.key = master->nat.help.nat_pptp_info.pac_call_id;
 		t.dst.protonum = IPPROTO_GRE;
 	}
 
@@ -149,10 +147,11 @@ pptp_outbound_pkt(struct sk_buff **pskb,
 {
 	struct ip_ct_pptp_master *ct_pptp_info = &ct->help.ct_pptp_info;
 	struct ip_nat_pptp *nat_pptp_info = &ct->nat.help.nat_pptp_info;
-	u_int16_t msg, new_callid;
+	u_int16_t msg;
+	__be16 new_callid;
 	unsigned int cid_off;
 
-	new_callid = htons(ct_pptp_info->pns_call_id);
+	new_callid = ct_pptp_info->pns_call_id;
 
 	switch (msg = ntohs(ctlh->messageType)) {
 	case PPTP_OUT_CALL_REQUEST:
@@ -170,7 +169,7 @@ pptp_outbound_pkt(struct sk_buff **pskb,
 		new_callid = ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.u.tcp.port;
 
 		/* save new call ID in ct info */
-		ct_pptp_info->pns_call_id = ntohs(new_callid);
+		ct_pptp_info->pns_call_id = new_callid;
 		break;
 	case PPTP_IN_CALL_REPLY:
 		cid_off = offsetof(union pptp_ctrl_union, icreq.callID);
@@ -235,14 +234,14 @@ pptp_exp_gre(struct ip_conntrack_expect *expect_orig,
 
 	/* alter expectation for PNS->PAC direction */
 	invert_tuplepr(&inv_t, &expect_orig->tuple);
-	expect_orig->saved_proto.gre.key = htons(ct_pptp_info->pns_call_id);
-	expect_orig->tuple.src.u.gre.key = htons(nat_pptp_info->pns_call_id);
-	expect_orig->tuple.dst.u.gre.key = htons(ct_pptp_info->pac_call_id);
+	expect_orig->saved_proto.gre.key = ct_pptp_info->pns_call_id;
+	expect_orig->tuple.src.u.gre.key = nat_pptp_info->pns_call_id;
+	expect_orig->tuple.dst.u.gre.key = ct_pptp_info->pac_call_id;
 	expect_orig->dir = IP_CT_DIR_ORIGINAL;
 	inv_t.src.ip = reply_t->src.ip;
 	inv_t.dst.ip = reply_t->dst.ip;
-	inv_t.src.u.gre.key = htons(nat_pptp_info->pac_call_id);
-	inv_t.dst.u.gre.key = htons(ct_pptp_info->pns_call_id);
+	inv_t.src.u.gre.key = nat_pptp_info->pac_call_id;
+	inv_t.dst.u.gre.key = ct_pptp_info->pns_call_id;
 
 	if (!ip_conntrack_expect_related(expect_orig)) {
 		DEBUGP("successfully registered expect\n");
@@ -253,14 +252,14 @@ pptp_exp_gre(struct ip_conntrack_expect *expect_orig,
 
 	/* alter expectation for PAC->PNS direction */
 	invert_tuplepr(&inv_t, &expect_reply->tuple);
-	expect_reply->saved_proto.gre.key = htons(nat_pptp_info->pns_call_id);
-	expect_reply->tuple.src.u.gre.key = htons(nat_pptp_info->pac_call_id);
-	expect_reply->tuple.dst.u.gre.key = htons(ct_pptp_info->pns_call_id);
+	expect_reply->saved_proto.gre.key = nat_pptp_info->pns_call_id;
+	expect_reply->tuple.src.u.gre.key = nat_pptp_info->pac_call_id;
+	expect_reply->tuple.dst.u.gre.key = ct_pptp_info->pns_call_id;
 	expect_reply->dir = IP_CT_DIR_REPLY;
 	inv_t.src.ip = orig_t->src.ip;
 	inv_t.dst.ip = orig_t->dst.ip;
-	inv_t.src.u.gre.key = htons(nat_pptp_info->pns_call_id);
-	inv_t.dst.u.gre.key = htons(ct_pptp_info->pac_call_id);
+	inv_t.src.u.gre.key = nat_pptp_info->pns_call_id;
+	inv_t.dst.u.gre.key = ct_pptp_info->pac_call_id;
 
 	if (!ip_conntrack_expect_related(expect_reply)) {
 		DEBUGP("successfully registered expect\n");
@@ -297,10 +296,11 @@ pptp_inbound_pkt(struct sk_buff **pskb,
 		 union pptp_ctrl_union *pptpReq)
 {
 	struct ip_nat_pptp *nat_pptp_info = &ct->nat.help.nat_pptp_info;
-	u_int16_t msg, new_cid = 0, new_pcid;
+	u_int16_t msg, new_cid = 0;
+	__be16 new_pcid;
 	unsigned int pcid_off, cid_off = 0;
 
-	new_pcid = htons(nat_pptp_info->pns_call_id);
+	new_pcid = nat_pptp_info->pns_call_id;
 
 	switch (msg = ntohs(ctlh->messageType)) {
 	case PPTP_OUT_CALL_REPLY:
diff --git a/net/ipv4/netfilter/ip_nat_proto_gre.c b/net/ipv4/netfilter/ip_nat_proto_gre.c
index a5226691f02c..bf91f9312b3c 100644
--- a/net/ipv4/netfilter/ip_nat_proto_gre.c
+++ b/net/ipv4/netfilter/ip_nat_proto_gre.c
@@ -67,7 +67,7 @@ gre_unique_tuple(struct ip_conntrack_tuple *tuple,
 		 const struct ip_conntrack *conntrack)
 {
 	static u_int16_t key;
-	u_int16_t *keyptr;
+	__be16 *keyptr;
 	unsigned int min, i, range_size;
 
 	if (maniptype == IP_NAT_MANIP_SRC)
-- 
cgit v1.2.3


From 6013c0a13e335674a783215e182c367406294392 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Wed, 20 Sep 2006 12:08:56 -0700
Subject: [NETFILTER]: PPTP conntrack: fix header definitions

Fix a few header definitions to match RFC2637. Most importantly the
PptpOutCallRequest header included an invalid padding field and a
size check was disabled because of this.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter_ipv4/ip_conntrack_pptp.h | 9 +++++----
 net/ipv4/netfilter/ip_conntrack_helper_pptp.c    | 2 +-
 2 files changed, 6 insertions(+), 5 deletions(-)

(limited to 'include')

diff --git a/include/linux/netfilter_ipv4/ip_conntrack_pptp.h b/include/linux/netfilter_ipv4/ip_conntrack_pptp.h
index 0d35623f9453..620bf06fabc2 100644
--- a/include/linux/netfilter_ipv4/ip_conntrack_pptp.h
+++ b/include/linux/netfilter_ipv4/ip_conntrack_pptp.h
@@ -107,8 +107,7 @@ struct PptpControlHeader {
 
 struct PptpStartSessionRequest {
 	__be16	protocolVersion;
-	__u8	reserved1;
-	__u8	reserved2;
+	__u16	reserved1;
 	__be32	framingCapability;
 	__be32	bearerCapability;
 	__be16	maxChannels;
@@ -143,6 +142,8 @@ struct PptpStartSessionReply {
 
 struct PptpStopSessionRequest {
 	__u8	reason;
+	__u8	reserved1;
+	__u16	reserved2;
 };
 
 /* PptpStopSessionResultCode */
@@ -152,6 +153,7 @@ struct PptpStopSessionRequest {
 struct PptpStopSessionReply {
 	__u8	resultCode;
 	__u8	generalErrorCode;
+	__u16	reserved1;
 };
 
 struct PptpEchoRequest {
@@ -188,9 +190,8 @@ struct PptpOutCallRequest {
 	__be32	framingType;
 	__be16	packetWindow;
 	__be16	packetProcDelay;
-	__u16	reserved1;
 	__be16	phoneNumberLength;
-	__u16	reserved2;
+	__u16	reserved1;
 	__u8	phoneNumber[64];
 	__u8	subAddress[64];
 };
diff --git a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c
index 0510ee50dc65..1a8da9015d8c 100644
--- a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c
+++ b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c
@@ -569,7 +569,7 @@ pptp_outbound_pkt(struct sk_buff **pskb,
 	case PPTP_OUT_CALL_REQUEST:
 		if (reqlen < sizeof(_pptpReq.ocreq)) {
 			DEBUGP("%s: short packet\n", pptp_msg_name[msg]);
-			/* FIXME: break; */
+			break;
 		}
 
 		/* client initiating connection to server */
-- 
cgit v1.2.3


From cf9f81523ef3e95d9f222c896d266e4562999150 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Wed, 20 Sep 2006 12:09:34 -0700
Subject: [NETFILTER]: PPTP conntrack: simplify expectation handling

Remove duplicated expectation handling in the NAT helper and simplify
the remains in the conntrack helper.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter_ipv4/ip_conntrack_pptp.h |  2 +-
 net/ipv4/netfilter/ip_conntrack_helper_pptp.c    | 92 ++++++++----------------
 net/ipv4/netfilter/ip_nat_helper_pptp.c          | 58 +--------------
 3 files changed, 35 insertions(+), 117 deletions(-)

(limited to 'include')

diff --git a/include/linux/netfilter_ipv4/ip_conntrack_pptp.h b/include/linux/netfilter_ipv4/ip_conntrack_pptp.h
index 620bf06fabc2..2644b1faddd6 100644
--- a/include/linux/netfilter_ipv4/ip_conntrack_pptp.h
+++ b/include/linux/netfilter_ipv4/ip_conntrack_pptp.h
@@ -315,7 +315,7 @@ extern int
 			  struct PptpControlHeader *ctlh,
 			  union pptp_ctrl_union *pptpReq);
 
-extern int
+extern void
 (*ip_nat_pptp_hook_exp_gre)(struct ip_conntrack_expect *exp_orig,
 			    struct ip_conntrack_expect *exp_reply);
 
diff --git a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c
index 5f7af6ef3881..57eac6e3871a 100644
--- a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c
+++ b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c
@@ -80,7 +80,7 @@ int
 			  struct PptpControlHeader *ctlh,
 			  union pptp_ctrl_union *pptpReq);
 
-int
+void
 (*ip_nat_pptp_hook_exp_gre)(struct ip_conntrack_expect *expect_orig,
 			    struct ip_conntrack_expect *expect_reply);
 
@@ -219,93 +219,63 @@ static void pptp_destroy_siblings(struct ip_conntrack *ct)
 
 /* expect GRE connections (PNS->PAC and PAC->PNS direction) */
 static inline int
-exp_gre(struct ip_conntrack *master,
+exp_gre(struct ip_conntrack *ct,
 	__be16 callid,
 	__be16 peer_callid)
 {
-	struct ip_conntrack_tuple inv_tuple;
-	struct ip_conntrack_tuple exp_tuples[] = {
-		/* tuple in original direction, PNS->PAC */
-		{ .src = { .ip = master->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.ip,
-			   .u = { .gre = { .key = peer_callid } }
-			 },
-		  .dst = { .ip = master->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.ip,
-			   .u = { .gre = { .key = callid } },
-			   .protonum = IPPROTO_GRE
-			 },
-		 },
-		/* tuple in reply direction, PAC->PNS */
-		{ .src = { .ip = master->tuplehash[IP_CT_DIR_REPLY].tuple.src.ip,
-			   .u = { .gre = { .key = callid } }
-			 },
-		  .dst = { .ip = master->tuplehash[IP_CT_DIR_REPLY].tuple.dst.ip,
-			   .u = { .gre = { .key = peer_callid } },
-			   .protonum = IPPROTO_GRE
-			 },
-		 }
-	};
 	struct ip_conntrack_expect *exp_orig, *exp_reply;
 	int ret = 1;
 
-	exp_orig = ip_conntrack_expect_alloc(master);
+	exp_orig = ip_conntrack_expect_alloc(ct);
 	if (exp_orig == NULL)
 		goto out;
 
-	exp_reply = ip_conntrack_expect_alloc(master);
+	exp_reply = ip_conntrack_expect_alloc(ct);
 	if (exp_reply == NULL)
 		goto out_put_orig;
 
-	memcpy(&exp_orig->tuple, &exp_tuples[0], sizeof(exp_orig->tuple));
+	/* original direction, PNS->PAC */
+	exp_orig->tuple.src.ip = ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.ip;
+	exp_orig->tuple.src.u.gre.key = peer_callid;
+	exp_orig->tuple.dst.ip = ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.ip;
+	exp_orig->tuple.dst.u.gre.key = callid;
+	exp_orig->tuple.dst.protonum = IPPROTO_GRE;
 
 	exp_orig->mask.src.ip = 0xffffffff;
 	exp_orig->mask.src.u.all = 0;
-	exp_orig->mask.dst.u.all = 0;
 	exp_orig->mask.dst.u.gre.key = htons(0xffff);
 	exp_orig->mask.dst.ip = 0xffffffff;
 	exp_orig->mask.dst.protonum = 0xff;
 
-	exp_orig->master = master;
+	exp_orig->master = ct;
 	exp_orig->expectfn = pptp_expectfn;
 	exp_orig->flags = 0;
 
 	/* both expectations are identical apart from tuple */
 	memcpy(exp_reply, exp_orig, sizeof(*exp_reply));
-	memcpy(&exp_reply->tuple, &exp_tuples[1], sizeof(exp_reply->tuple));
-
-	if (ip_nat_pptp_hook_exp_gre)
-		ret = ip_nat_pptp_hook_exp_gre(exp_orig, exp_reply);
-	else {
 
-		DEBUGP("calling expect_related PNS->PAC");
-		DUMP_TUPLE(&exp_orig->tuple);
+	/* reply direction, PAC->PNS */
+	exp_reply->tuple.src.ip = ct->tuplehash[IP_CT_DIR_REPLY].tuple.src.ip;
+	exp_reply->tuple.src.u.gre.key = callid;
+	exp_reply->tuple.dst.ip = ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.ip;
+	exp_reply->tuple.dst.u.gre.key = peer_callid;
+	exp_reply->tuple.dst.protonum = IPPROTO_GRE;
 
-		if (ip_conntrack_expect_related(exp_orig) != 0) {
-			DEBUGP("cannot expect_related()\n");
-			goto out_put_both;
-		}
-
-		DEBUGP("calling expect_related PAC->PNS");
-		DUMP_TUPLE(&exp_reply->tuple);
-
-		if (ip_conntrack_expect_related(exp_reply) != 0) {
-			DEBUGP("cannot expect_related()\n");
-			goto out_unexpect_orig;
-		}
-
-		/* Add GRE keymap entries */
-		if (ip_ct_gre_keymap_add(master, &exp_reply->tuple, 0) != 0) {
-			DEBUGP("cannot keymap_add() exp\n");
-			goto out_unexpect_both;
-		}
-
-		invert_tuplepr(&inv_tuple, &exp_reply->tuple);
-		if (ip_ct_gre_keymap_add(master, &inv_tuple, 1) != 0) {
-			ip_ct_gre_keymap_destroy(master);
-			DEBUGP("cannot keymap_add() exp_inv\n");
-			goto out_unexpect_both;
-		}
-		ret = 0;
+	if (ip_nat_pptp_hook_exp_gre)
+		ip_nat_pptp_hook_exp_gre(exp_orig, exp_reply);
+	if (ip_conntrack_expect_related(exp_orig) != 0)
+		goto out_put_both;
+	if (ip_conntrack_expect_related(exp_reply) != 0)
+		goto out_unexpect_orig;
+
+	/* Add GRE keymap entries */
+	if (ip_ct_gre_keymap_add(ct, &exp_orig->tuple, 0) != 0)
+		goto out_unexpect_both;
+	if (ip_ct_gre_keymap_add(ct, &exp_reply->tuple, 1) != 0) {
+		ip_ct_gre_keymap_destroy(ct);
+		goto out_unexpect_both;
 	}
+	ret = 0;
 
 out_put_both:
 	ip_conntrack_expect_put(exp_reply);
diff --git a/net/ipv4/netfilter/ip_nat_helper_pptp.c b/net/ipv4/netfilter/ip_nat_helper_pptp.c
index 0f5e753b481d..84f6bd09fcd4 100644
--- a/net/ipv4/netfilter/ip_nat_helper_pptp.c
+++ b/net/ipv4/netfilter/ip_nat_helper_pptp.c
@@ -211,80 +211,28 @@ pptp_outbound_pkt(struct sk_buff **pskb,
 	return NF_ACCEPT;
 }
 
-static int
+static void
 pptp_exp_gre(struct ip_conntrack_expect *expect_orig,
 	     struct ip_conntrack_expect *expect_reply)
 {
-	struct ip_ct_pptp_master *ct_pptp_info =
-				&expect_orig->master->help.ct_pptp_info;
-	struct ip_nat_pptp *nat_pptp_info =
-				&expect_orig->master->nat.help.nat_pptp_info;
-
 	struct ip_conntrack *ct = expect_orig->master;
-
-	struct ip_conntrack_tuple inv_t;
-	struct ip_conntrack_tuple *orig_t, *reply_t;
+	struct ip_ct_pptp_master *ct_pptp_info = &ct->help.ct_pptp_info;
+	struct ip_nat_pptp *nat_pptp_info = &ct->nat.help.nat_pptp_info;
 
 	/* save original PAC call ID in nat_info */
 	nat_pptp_info->pac_call_id = ct_pptp_info->pac_call_id;
 
-	/* alter expectation */
-	orig_t = &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple;
-	reply_t = &ct->tuplehash[IP_CT_DIR_REPLY].tuple;
-
 	/* alter expectation for PNS->PAC direction */
-	invert_tuplepr(&inv_t, &expect_orig->tuple);
 	expect_orig->saved_proto.gre.key = ct_pptp_info->pns_call_id;
 	expect_orig->tuple.src.u.gre.key = nat_pptp_info->pns_call_id;
 	expect_orig->tuple.dst.u.gre.key = ct_pptp_info->pac_call_id;
 	expect_orig->dir = IP_CT_DIR_ORIGINAL;
-	inv_t.src.ip = reply_t->src.ip;
-	inv_t.dst.ip = reply_t->dst.ip;
-	inv_t.src.u.gre.key = nat_pptp_info->pac_call_id;
-	inv_t.dst.u.gre.key = ct_pptp_info->pns_call_id;
-
-	if (!ip_conntrack_expect_related(expect_orig)) {
-		DEBUGP("successfully registered expect\n");
-	} else {
-		DEBUGP("can't expect_related(expect_orig)\n");
-		return 1;
-	}
 
 	/* alter expectation for PAC->PNS direction */
-	invert_tuplepr(&inv_t, &expect_reply->tuple);
 	expect_reply->saved_proto.gre.key = nat_pptp_info->pns_call_id;
 	expect_reply->tuple.src.u.gre.key = nat_pptp_info->pac_call_id;
 	expect_reply->tuple.dst.u.gre.key = ct_pptp_info->pns_call_id;
 	expect_reply->dir = IP_CT_DIR_REPLY;
-	inv_t.src.ip = orig_t->src.ip;
-	inv_t.dst.ip = orig_t->dst.ip;
-	inv_t.src.u.gre.key = nat_pptp_info->pns_call_id;
-	inv_t.dst.u.gre.key = ct_pptp_info->pac_call_id;
-
-	if (!ip_conntrack_expect_related(expect_reply)) {
-		DEBUGP("successfully registered expect\n");
-	} else {
-		DEBUGP("can't expect_related(expect_reply)\n");
-		ip_conntrack_unexpect_related(expect_orig);
-		return 1;
-	}
-
-	if (ip_ct_gre_keymap_add(ct, &expect_reply->tuple, 0) < 0) {
-		DEBUGP("can't register original keymap\n");
-		ip_conntrack_unexpect_related(expect_orig);
-		ip_conntrack_unexpect_related(expect_reply);
-		return 1;
-	}
-
-	if (ip_ct_gre_keymap_add(ct, &inv_t, 1) < 0) {
-		DEBUGP("can't register reply keymap\n");
-		ip_conntrack_unexpect_related(expect_orig);
-		ip_conntrack_unexpect_related(expect_reply);
-		ip_ct_gre_keymap_destroy(ct);
-		return 1;
-	}
-
-	return 0;
 }
 
 /* inbound packets == from PAC to PNS */
-- 
cgit v1.2.3


From 4c5de695cf7f71c85ad8cfff509f6475b8bd4d27 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Wed, 20 Sep 2006 12:11:30 -0700
Subject: [NETFILTER]: PPTP conntrack: fix another GRE keymap leak

When the master PPTP connection times out while still having unfullfilled
expectations (and a GRE keymap entry) associated with it, the keymap entry
is not destroyed.

Add a destroy callback to struct ip_conntrack_helper and use it to destroy
PPTP siblings when the master is destroyed.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter_ipv4/ip_conntrack_helper.h |  2 ++
 net/ipv4/netfilter/ip_conntrack_core.c             |  5 +++++
 net/ipv4/netfilter/ip_conntrack_helper_pptp.c      | 12 ++----------
 3 files changed, 9 insertions(+), 10 deletions(-)

(limited to 'include')

diff --git a/include/linux/netfilter_ipv4/ip_conntrack_helper.h b/include/linux/netfilter_ipv4/ip_conntrack_helper.h
index 8d69279ccfe4..77fe868d36ff 100644
--- a/include/linux/netfilter_ipv4/ip_conntrack_helper.h
+++ b/include/linux/netfilter_ipv4/ip_conntrack_helper.h
@@ -25,6 +25,8 @@ struct ip_conntrack_helper
 		    struct ip_conntrack *ct,
 		    enum ip_conntrack_info conntrackinfo);
 
+	void (*destroy)(struct ip_conntrack *ct);
+
 	int (*to_nfattr)(struct sk_buff *skb, const struct ip_conntrack *ct);
 };
 
diff --git a/net/ipv4/netfilter/ip_conntrack_core.c b/net/ipv4/netfilter/ip_conntrack_core.c
index 2b6f24fc727e..c432b3163609 100644
--- a/net/ipv4/netfilter/ip_conntrack_core.c
+++ b/net/ipv4/netfilter/ip_conntrack_core.c
@@ -307,6 +307,7 @@ destroy_conntrack(struct nf_conntrack *nfct)
 {
 	struct ip_conntrack *ct = (struct ip_conntrack *)nfct;
 	struct ip_conntrack_protocol *proto;
+	struct ip_conntrack_helper *helper;
 
 	DEBUGP("destroy_conntrack(%p)\n", ct);
 	IP_NF_ASSERT(atomic_read(&nfct->use) == 0);
@@ -315,6 +316,10 @@ destroy_conntrack(struct nf_conntrack *nfct)
 	ip_conntrack_event(IPCT_DESTROY, ct);
 	set_bit(IPS_DYING_BIT, &ct->status);
 
+	helper = ct->helper;
+	if (helper && helper->destroy)
+		helper->destroy(ct);
+
 	/* To make sure we don't get any weird locking issues here:
 	 * destroy_conntrack() MUST NOT be called with a write lock
 	 * to ip_conntrack_lock!!! -HW */
diff --git a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c
index 98267b0d2a47..fb0aee691721 100644
--- a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c
+++ b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c
@@ -553,15 +553,6 @@ conntrack_pptp_help(struct sk_buff **pskb,
 	nexthdr_off += tcph->doff * 4;
  	datalen = tcplen - tcph->doff * 4;
 
-	if (tcph->fin || tcph->rst) {
-		DEBUGP("RST/FIN received, timeouting GRE\n");
-		/* can't do this after real newnat */
-		info->cstate = PPTP_CALL_NONE;
-
-		/* untrack this call id, unexpect GRE packets */
-		pptp_destroy_siblings(ct);
-	}
-
 	pptph = skb_header_pointer(*pskb, nexthdr_off, sizeof(_pptph), &_pptph);
 	if (!pptph) {
 		DEBUGP("no full PPTP header, can't track\n");
@@ -640,7 +631,8 @@ static struct ip_conntrack_helper pptp = {
 			   .protonum = 0xff
 		 	 }
 		},
-	.help = conntrack_pptp_help
+	.help = conntrack_pptp_help,
+	.destroy = pptp_destroy_siblings,
 };
 
 extern void ip_ct_proto_gre_fini(void);
-- 
cgit v1.2.3


From 62dd93181aaa1d5a501a9cebcb254f44b8a48af7 Mon Sep 17 00:00:00 2001
From: Ville Nuorvala <vnuorval@tcs.hut.fi>
Date: Fri, 22 Sep 2006 14:43:19 -0700
Subject: [IPV6] NDISC: Set per-entry is_router flag in Proxy NA.

We have sent NA with router flag from the node-wide forwarding
configuration.  This is not appropriate for proxy NA, and it should be
set according to each proxy entry's configuration.

This is used by Mobile IPv6 home agent to support physical home link
in acting as a proxy router for mobile node which is not a router,
for example.

Based on MIPL2 kernel patch.

Signed-off-by: Ville Nuorvala <vnuorval@tcs.hut.fi>
Signed-off-by: Masahide NAKAMURA <nakam@linux-ipv6.org>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/net/neighbour.h |  1 +
 net/core/neighbour.c    | 11 ++++++++---
 net/ipv6/ndisc.c        | 14 +++++++++++---
 3 files changed, 20 insertions(+), 6 deletions(-)

(limited to 'include')

diff --git a/include/net/neighbour.h b/include/net/neighbour.h
index bd187daffdb9..c8aacbd2e333 100644
--- a/include/net/neighbour.h
+++ b/include/net/neighbour.h
@@ -126,6 +126,7 @@ struct pneigh_entry
 {
 	struct pneigh_entry	*next;
 	struct net_device		*dev;
+	u8			flags;
 	u8			key[0];
 };
 
diff --git a/net/core/neighbour.c b/net/core/neighbour.c
index a45bd2124d6b..b6c69e1463e8 100644
--- a/net/core/neighbour.c
+++ b/net/core/neighbour.c
@@ -1544,9 +1544,14 @@ int neigh_add(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 		lladdr = tb[NDA_LLADDR] ? nla_data(tb[NDA_LLADDR]) : NULL;
 
 		if (ndm->ndm_flags & NTF_PROXY) {
-			err = 0;
-			if (pneigh_lookup(tbl, dst, dev, 1) == NULL)
-				err = -ENOBUFS;
+			struct pneigh_entry *pn;
+
+			err = -ENOBUFS;
+			pn = pneigh_lookup(tbl, dst, dev, 1);
+			if (pn) {
+				pn->flags = ndm->ndm_flags;
+				err = 0;
+			}
 			goto out_dev_put;
 		}
 
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index 0e0d6ce69021..ddf038636f01 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -736,8 +736,10 @@ static void ndisc_recv_ns(struct sk_buff *skb)
 	struct inet6_ifaddr *ifp;
 	struct inet6_dev *idev = NULL;
 	struct neighbour *neigh;
+	struct pneigh_entry *pneigh = NULL;
 	int dad = ipv6_addr_any(saddr);
 	int inc;
+	int is_router;
 
 	if (ipv6_addr_is_multicast(&msg->target)) {
 		ND_PRINTK2(KERN_WARNING 
@@ -822,7 +824,8 @@ static void ndisc_recv_ns(struct sk_buff *skb)
 
 		if (ipv6_chk_acast_addr(dev, &msg->target) ||
 		    (idev->cnf.forwarding && 
-		     pneigh_lookup(&nd_tbl, &msg->target, dev, 0))) {
+		     (pneigh = pneigh_lookup(&nd_tbl,
+					     &msg->target, dev, 0)) != NULL)) {
 			if (!(NEIGH_CB(skb)->flags & LOCALLY_ENQUEUED) &&
 			    skb->pkt_type != PACKET_HOST &&
 			    inc != 0 &&
@@ -843,12 +846,17 @@ static void ndisc_recv_ns(struct sk_buff *skb)
 			goto out;
 	}
 
+	if (pneigh)
+		is_router = pneigh->flags & NTF_ROUTER;
+	else
+		is_router = idev->cnf.forwarding;
+
 	if (dad) {
 		struct in6_addr maddr;
 
 		ipv6_addr_all_nodes(&maddr);
 		ndisc_send_na(dev, NULL, &maddr, &msg->target,
-			      idev->cnf.forwarding, 0, (ifp != NULL), 1);
+			      is_router, 0, (ifp != NULL), 1);
 		goto out;
 	}
 
@@ -869,7 +877,7 @@ static void ndisc_recv_ns(struct sk_buff *skb)
 			     NEIGH_UPDATE_F_OVERRIDE);
 	if (neigh || !dev->hard_header) {
 		ndisc_send_na(dev, neigh, saddr, &msg->target,
-			      idev->cnf.forwarding, 
+			      is_router,
 			      1, (ifp != NULL && inc), inc);
 		if (neigh)
 			neigh_release(neigh);
-- 
cgit v1.2.3


From fbea49e1e2404baa2d88ab47e2db89e49551b53b Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Fri, 22 Sep 2006 14:43:49 -0700
Subject: [IPV6] NDISC: Add proxy_ndp sysctl.

We do not always need proxy NDP functionality even we
enable forwarding.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 Documentation/networking/ip-sysctl.txt |  3 +++
 include/linux/ipv6.h                   |  2 ++
 include/linux/sysctl.h                 |  1 +
 net/ipv6/addrconf.c                    | 11 +++++++++++
 net/ipv6/ip6_output.c                  |  4 +++-
 net/ipv6/ndisc.c                       |  8 +++++++-
 6 files changed, 27 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/Documentation/networking/ip-sysctl.txt b/Documentation/networking/ip-sysctl.txt
index 307cd4ec8edd..935e298f674a 100644
--- a/Documentation/networking/ip-sysctl.txt
+++ b/Documentation/networking/ip-sysctl.txt
@@ -765,6 +765,9 @@ conf/all/forwarding - BOOLEAN
 
 	This referred to as global forwarding.
 
+proxy_ndp - BOOLEAN
+	Do proxy ndp.
+
 conf/interface/*:
 	Change special settings per interface.
 
diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index 1d6d3ccc9413..caca57df0d7d 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -176,6 +176,7 @@ struct ipv6_devconf {
 	__s32		accept_ra_rt_info_max_plen;
 #endif
 #endif
+	__s32		proxy_ndp;
 	void		*sysctl;
 };
 
@@ -203,6 +204,7 @@ enum {
 	DEVCONF_ACCEPT_RA_RTR_PREF,
 	DEVCONF_RTR_PROBE_INTERVAL,
 	DEVCONF_ACCEPT_RA_RT_INFO_MAX_PLEN,
+	DEVCONF_PROXY_NDP,
 	DEVCONF_MAX
 };
 
diff --git a/include/linux/sysctl.h b/include/linux/sysctl.h
index af61d9235409..736ed917a4f8 100644
--- a/include/linux/sysctl.h
+++ b/include/linux/sysctl.h
@@ -556,6 +556,7 @@ enum {
 	NET_IPV6_ACCEPT_RA_RTR_PREF=20,
 	NET_IPV6_RTR_PROBE_INTERVAL=21,
 	NET_IPV6_ACCEPT_RA_RT_INFO_MAX_PLEN=22,
+	NET_IPV6_PROXY_NDP=23,
 	__NET_IPV6_MAX
 };
 
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 1e5a296d0a82..825a291d5aa5 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -175,6 +175,7 @@ struct ipv6_devconf ipv6_devconf __read_mostly = {
 	.accept_ra_rt_info_max_plen = 0,
 #endif
 #endif
+	.proxy_ndp		= 0,
 };
 
 static struct ipv6_devconf ipv6_devconf_dflt __read_mostly = {
@@ -205,6 +206,7 @@ static struct ipv6_devconf ipv6_devconf_dflt __read_mostly = {
 	.accept_ra_rt_info_max_plen = 0,
 #endif
 #endif
+	.proxy_ndp		= 0,
 };
 
 /* IPv6 Wildcard Address and Loopback Address defined by RFC2553 */
@@ -3337,6 +3339,7 @@ static void inline ipv6_store_devconf(struct ipv6_devconf *cnf,
 	array[DEVCONF_ACCEPT_RA_RT_INFO_MAX_PLEN] = cnf->accept_ra_rt_info_max_plen;
 #endif
 #endif
+	array[DEVCONF_PROXY_NDP] = cnf->proxy_ndp;
 }
 
 /* Maximum length of ifinfomsg attributes */
@@ -3859,6 +3862,14 @@ static struct addrconf_sysctl_table
 		},
 #endif
 #endif
+		{
+			.ctl_name	=	NET_IPV6_PROXY_NDP,
+			.procname	=	"proxy_ndp",
+			.data		=	&ipv6_devconf.proxy_ndp,
+			.maxlen		=	sizeof(int),
+			.mode		=	0644,
+			.proc_handler	=	&proc_dointvec,
+		},
 		{
 			.ctl_name	=	0,	/* sentinel */
 		}
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index b2be749d2217..66716911962e 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -412,7 +412,9 @@ int ip6_forward(struct sk_buff *skb)
 		return -ETIMEDOUT;
 	}
 
-	if (pneigh_lookup(&nd_tbl, &hdr->daddr, skb->dev, 0)) {
+	/* XXX: idev->cnf.proxy_ndp? */
+	if (ipv6_devconf.proxy_ndp &&
+	    pneigh_lookup(&nd_tbl, &hdr->daddr, skb->dev, 0)) {
 		int proxied = ip6_forward_proxy_check(skb);
 		if (proxied > 0)
 			return ip6_input(skb);
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index ddf038636f01..76517a5f6576 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -824,6 +824,7 @@ static void ndisc_recv_ns(struct sk_buff *skb)
 
 		if (ipv6_chk_acast_addr(dev, &msg->target) ||
 		    (idev->cnf.forwarding && 
+		     (ipv6_devconf.proxy_ndp || idev->cnf.proxy_ndp) &&
 		     (pneigh = pneigh_lookup(&nd_tbl,
 					     &msg->target, dev, 0)) != NULL)) {
 			if (!(NEIGH_CB(skb)->flags & LOCALLY_ENQUEUED) &&
@@ -966,8 +967,13 @@ static void ndisc_recv_na(struct sk_buff *skb)
 		 * has already sent a NA to us.
 		 */
 		if (lladdr && !memcmp(lladdr, dev->dev_addr, dev->addr_len) &&
-		    pneigh_lookup(&nd_tbl, &msg->target, dev, 0))
+		    ipv6_devconf.forwarding && ipv6_devconf.proxy_ndp &&
+		    pneigh_lookup(&nd_tbl, &msg->target, dev, 0)) {
+			/* XXX: idev->cnf.prixy_ndp */
+			WARN_ON(skb->dst != NULL &&
+				((struct rt6_info *)skb->dst)->rt6i_idev);
 			goto out;
+		}
 
 		neigh_update(neigh, lladdr,
 			     msg->icmph.icmp6_solicited ? NUD_REACHABLE : NUD_STALE,
-- 
cgit v1.2.3


From 8814c4b533817df825485ff32ce6ac406c3a54d1 Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Fri, 22 Sep 2006 14:44:24 -0700
Subject: [IPV6] ADDRCONF: Convert addrconf_lock to RCU.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/addrconf.h | 10 ++++------
 include/net/if_inet6.h |  1 +
 net/core/pktgen.c      |  4 ++--
 net/ipv6/addrconf.c    | 46 ++++++++++++++++++++++++----------------------
 net/ipv6/anycast.c     |  4 ++--
 net/ipv6/ipv6_syms.c   |  1 -
 net/sctp/ipv6.c        |  6 +++---
 7 files changed, 36 insertions(+), 36 deletions(-)

(limited to 'include')

diff --git a/include/net/addrconf.h b/include/net/addrconf.h
index 5fc8627435eb..aa2ed8f0a9dd 100644
--- a/include/net/addrconf.h
+++ b/include/net/addrconf.h
@@ -133,20 +133,18 @@ extern int unregister_inet6addr_notifier(struct notifier_block *nb);
 static inline struct inet6_dev *
 __in6_dev_get(struct net_device *dev)
 {
-	return (struct inet6_dev *)dev->ip6_ptr;
+	return rcu_dereference(dev->ip6_ptr);
 }
 
-extern rwlock_t addrconf_lock;
-
 static inline struct inet6_dev *
 in6_dev_get(struct net_device *dev)
 {
 	struct inet6_dev *idev = NULL;
-	read_lock(&addrconf_lock);
-	idev = dev->ip6_ptr;
+	rcu_read_lock();
+	idev = __in6_dev_get(dev);
 	if (idev)
 		atomic_inc(&idev->refcnt);
-	read_unlock(&addrconf_lock);
+	rcu_read_unlock();
 	return idev;
 }
 
diff --git a/include/net/if_inet6.h b/include/net/if_inet6.h
index e459e1a0ae4a..34489c13c119 100644
--- a/include/net/if_inet6.h
+++ b/include/net/if_inet6.h
@@ -189,6 +189,7 @@ struct inet6_dev
 	struct ipv6_devconf	cnf;
 	struct ipv6_devstat	stats;
 	unsigned long		tstamp; /* ipv6InterfaceTable update timestamp */
+	struct rcu_head		rcu;
 };
 
 extern struct ipv6_devconf ipv6_devconf;
diff --git a/net/core/pktgen.c b/net/core/pktgen.c
index 6a7320b39ed0..72145d4a2600 100644
--- a/net/core/pktgen.c
+++ b/net/core/pktgen.c
@@ -1786,7 +1786,7 @@ static void pktgen_setup_inject(struct pktgen_dev *pkt_dev)
 			 * use ipv6_get_lladdr if/when it's get exported
 			 */
 
-			read_lock(&addrconf_lock);
+			rcu_read_lock();
 			if ((idev = __in6_dev_get(pkt_dev->odev)) != NULL) {
 				struct inet6_ifaddr *ifp;
 
@@ -1805,7 +1805,7 @@ static void pktgen_setup_inject(struct pktgen_dev *pkt_dev)
 				}
 				read_unlock_bh(&idev->lock);
 			}
-			read_unlock(&addrconf_lock);
+			rcu_read_unlock();
 			if (err)
 				printk("pktgen: ERROR: IPv6 link address not availble.\n");
 		}
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 825a291d5aa5..c09ebb7bb98a 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -119,9 +119,6 @@ static int ipv6_count_addresses(struct inet6_dev *idev);
 static struct inet6_ifaddr		*inet6_addr_lst[IN6_ADDR_HSIZE];
 static DEFINE_RWLOCK(addrconf_hash_lock);
 
-/* Protects inet6 devices */
-DEFINE_RWLOCK(addrconf_lock);
-
 static void addrconf_verify(unsigned long);
 
 static DEFINE_TIMER(addr_chk_timer, addrconf_verify, 0, 0);
@@ -318,6 +315,12 @@ static void addrconf_mod_timer(struct inet6_ifaddr *ifp,
 
 /* Nobody refers to this device, we may destroy it. */
 
+static void in6_dev_finish_destroy_rcu(struct rcu_head *head)
+{
+	struct inet6_dev *idev = container_of(head, struct inet6_dev, rcu);
+	kfree(idev);
+}
+
 void in6_dev_finish_destroy(struct inet6_dev *idev)
 {
 	struct net_device *dev = idev->dev;
@@ -332,7 +335,7 @@ void in6_dev_finish_destroy(struct inet6_dev *idev)
 		return;
 	}
 	snmp6_free_dev(idev);
-	kfree(idev);
+	call_rcu(&idev->rcu, in6_dev_finish_destroy_rcu);
 }
 
 static struct inet6_dev * ipv6_add_dev(struct net_device *dev)
@@ -408,9 +411,8 @@ static struct inet6_dev * ipv6_add_dev(struct net_device *dev)
 	if (netif_carrier_ok(dev))
 		ndev->if_flags |= IF_READY;
 
-	write_lock_bh(&addrconf_lock);
-	dev->ip6_ptr = ndev;
-	write_unlock_bh(&addrconf_lock);
+	/* protected by rtnl_lock */
+	rcu_assign_pointer(dev->ip6_ptr, ndev);
 
 	ipv6_mc_init_dev(ndev);
 	ndev->tstamp = jiffies;
@@ -474,7 +476,7 @@ static void addrconf_forward_change(void)
 
 	read_lock(&dev_base_lock);
 	for (dev=dev_base; dev; dev=dev->next) {
-		read_lock(&addrconf_lock);
+		rcu_read_lock();
 		idev = __in6_dev_get(dev);
 		if (idev) {
 			int changed = (!idev->cnf.forwarding) ^ (!ipv6_devconf.forwarding);
@@ -482,7 +484,7 @@ static void addrconf_forward_change(void)
 			if (changed)
 				dev_forward_change(idev);
 		}
-		read_unlock(&addrconf_lock);
+		rcu_read_unlock();
 	}
 	read_unlock(&dev_base_lock);
 }
@@ -543,7 +545,7 @@ ipv6_add_addr(struct inet6_dev *idev, const struct in6_addr *addr, int pfxlen,
 	int hash;
 	int err = 0;
 
-	read_lock_bh(&addrconf_lock);
+	rcu_read_lock_bh();
 	if (idev->dead) {
 		err = -ENODEV;			/*XXX*/
 		goto out2;
@@ -612,7 +614,7 @@ ipv6_add_addr(struct inet6_dev *idev, const struct in6_addr *addr, int pfxlen,
 	in6_ifa_hold(ifa);
 	write_unlock(&idev->lock);
 out2:
-	read_unlock_bh(&addrconf_lock);
+	rcu_read_unlock_bh();
 
 	if (likely(err == 0))
 		atomic_notifier_call_chain(&inet6addr_chain, NETDEV_UP, ifa);
@@ -915,7 +917,7 @@ int ipv6_dev_get_saddr(struct net_device *daddr_dev,
 	memset(&hiscore, 0, sizeof(hiscore));
 
 	read_lock(&dev_base_lock);
-	read_lock(&addrconf_lock);
+	rcu_read_lock();
 
 	for (dev = dev_base; dev; dev=dev->next) {
 		struct inet6_dev *idev;
@@ -1127,7 +1129,7 @@ record_it:
 		}
 		read_unlock_bh(&idev->lock);
 	}
-	read_unlock(&addrconf_lock);
+	rcu_read_unlock();
 	read_unlock(&dev_base_lock);
 
 	if (!ifa_result)
@@ -1151,7 +1153,7 @@ int ipv6_get_lladdr(struct net_device *dev, struct in6_addr *addr)
 	struct inet6_dev *idev;
 	int err = -EADDRNOTAVAIL;
 
-	read_lock(&addrconf_lock);
+	rcu_read_lock();
 	if ((idev = __in6_dev_get(dev)) != NULL) {
 		struct inet6_ifaddr *ifp;
 
@@ -1165,7 +1167,7 @@ int ipv6_get_lladdr(struct net_device *dev, struct in6_addr *addr)
 		}
 		read_unlock_bh(&idev->lock);
 	}
-	read_unlock(&addrconf_lock);
+	rcu_read_unlock();
 	return err;
 }
 
@@ -1466,7 +1468,7 @@ static void ipv6_regen_rndid(unsigned long data)
 	struct inet6_dev *idev = (struct inet6_dev *) data;
 	unsigned long expires;
 
-	read_lock_bh(&addrconf_lock);
+	rcu_read_lock_bh();
 	write_lock_bh(&idev->lock);
 
 	if (idev->dead)
@@ -1490,7 +1492,7 @@ static void ipv6_regen_rndid(unsigned long data)
 
 out:
 	write_unlock_bh(&idev->lock);
-	read_unlock_bh(&addrconf_lock);
+	rcu_read_unlock_bh();
 	in6_dev_put(idev);
 }
 
@@ -2342,10 +2344,10 @@ static int addrconf_ifdown(struct net_device *dev, int how)
 	           Do not dev_put!
 	 */
 	if (how == 1) {
-		write_lock_bh(&addrconf_lock);
-		dev->ip6_ptr = NULL;
 		idev->dead = 1;
-		write_unlock_bh(&addrconf_lock);
+
+		/* protected by rtnl_lock */
+		rcu_assign_pointer(dev->ip6_ptr, NULL);
 
 		/* Step 1.5: remove snmp6 entry */
 		snmp6_unregister_dev(idev);
@@ -3573,10 +3575,10 @@ static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
 
 static void ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
 {
-	read_lock_bh(&addrconf_lock);
+	rcu_read_lock_bh();
 	if (likely(ifp->idev->dead == 0))
 		__ipv6_ifa_notify(event, ifp);
-	read_unlock_bh(&addrconf_lock);
+	rcu_read_unlock_bh();
 }
 
 #ifdef CONFIG_SYSCTL
diff --git a/net/ipv6/anycast.c b/net/ipv6/anycast.c
index b80fc502ca03..a9604764e015 100644
--- a/net/ipv6/anycast.c
+++ b/net/ipv6/anycast.c
@@ -56,7 +56,7 @@ ip6_onlink(struct in6_addr *addr, struct net_device *dev)
 	int	onlink;
 
 	onlink = 0;
-	read_lock(&addrconf_lock);
+	rcu_read_lock();
 	idev = __in6_dev_get(dev);
 	if (idev) {
 		read_lock_bh(&idev->lock);
@@ -68,7 +68,7 @@ ip6_onlink(struct in6_addr *addr, struct net_device *dev)
 		}
 		read_unlock_bh(&idev->lock);
 	}
-	read_unlock(&addrconf_lock);
+	rcu_read_unlock();
 	return onlink;
 }
 
diff --git a/net/ipv6/ipv6_syms.c b/net/ipv6/ipv6_syms.c
index 7b7b90d9c3d0..0e8e0676a033 100644
--- a/net/ipv6/ipv6_syms.c
+++ b/net/ipv6/ipv6_syms.c
@@ -14,7 +14,6 @@ EXPORT_SYMBOL(ndisc_mc_map);
 EXPORT_SYMBOL(register_inet6addr_notifier);
 EXPORT_SYMBOL(unregister_inet6addr_notifier);
 EXPORT_SYMBOL(ip6_route_output);
-EXPORT_SYMBOL(addrconf_lock);
 EXPORT_SYMBOL(ipv6_setsockopt);
 EXPORT_SYMBOL(ipv6_getsockopt);
 EXPORT_SYMBOL(inet6_register_protosw);
diff --git a/net/sctp/ipv6.c b/net/sctp/ipv6.c
index fd87e3ceb56e..249e5033c1a8 100644
--- a/net/sctp/ipv6.c
+++ b/net/sctp/ipv6.c
@@ -321,9 +321,9 @@ static void sctp_v6_copy_addrlist(struct list_head *addrlist,
 	struct inet6_ifaddr *ifp;
 	struct sctp_sockaddr_entry *addr;
 
-	read_lock(&addrconf_lock);
+	rcu_read_lock();
 	if ((in6_dev = __in6_dev_get(dev)) == NULL) {
-		read_unlock(&addrconf_lock);
+		rcu_read_unlock();
 		return;
 	}
 
@@ -342,7 +342,7 @@ static void sctp_v6_copy_addrlist(struct list_head *addrlist,
 	}
 
 	read_unlock(&in6_dev->lock);
-	read_unlock(&addrconf_lock);
+	rcu_read_unlock();
 }
 
 /* Initialize a sockaddr_storage from in incoming skb. */
-- 
cgit v1.2.3


From 55ebaef1d5db9c1c76ba01a87fd986db5dee550d Mon Sep 17 00:00:00 2001
From: Noriaki TAKAMIYA <takamiya@po.ntts.co.jp>
Date: Fri, 22 Sep 2006 14:45:27 -0700
Subject: [IPV6] ADDRCONF: Allow non-DAD'able addresses.

IFA_F_NODAD flag, similar to IN6_IFF_NODAD in BSDs, is introduced
to skip DAD.

This flag should be set to Mobile IPv6 Home Address(es) on Mobile
Node because DAD would fail if we should perform DAD; our Home Agent
protects our Home Address(es).

Signed-off-by: Noriaki TAKAMIYA <takamiya@po.ntts.co.jp>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/if_addr.h |  1 +
 net/ipv6/addrconf.c     | 31 ++++++++++++++++---------------
 2 files changed, 17 insertions(+), 15 deletions(-)

(limited to 'include')

diff --git a/include/linux/if_addr.h b/include/linux/if_addr.h
index e1590454db59..ca24b9de13fb 100644
--- a/include/linux/if_addr.h
+++ b/include/linux/if_addr.h
@@ -38,6 +38,7 @@ enum
 #define IFA_F_SECONDARY		0x01
 #define IFA_F_TEMPORARY		IFA_F_SECONDARY
 
+#define	IFA_F_NODAD		0x02
 #define IFA_F_DEPRECATED	0x20
 #define IFA_F_TENTATIVE		0x40
 #define IFA_F_PERMANENT		0x80
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index c09ebb7bb98a..adb583a26151 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -1873,12 +1873,11 @@ err_exit:
  *	Manual configuration of address on an interface
  */
 static int inet6_addr_add(int ifindex, struct in6_addr *pfx, int plen,
-			  __u32 prefered_lft, __u32 valid_lft)
+			  __u8 ifa_flags, __u32 prefered_lft, __u32 valid_lft)
 {
 	struct inet6_ifaddr *ifp;
 	struct inet6_dev *idev;
 	struct net_device *dev;
-	__u8 ifa_flags = 0;
 	int scope;
 
 	ASSERT_RTNL();
@@ -1971,7 +1970,7 @@ int addrconf_add_ifaddr(void __user *arg)
 
 	rtnl_lock();
 	err = inet6_addr_add(ireq.ifr6_ifindex, &ireq.ifr6_addr, ireq.ifr6_prefixlen,
-			     INFINITY_LIFE_TIME, INFINITY_LIFE_TIME);
+			     IFA_F_PERMANENT, INFINITY_LIFE_TIME, INFINITY_LIFE_TIME);
 	rtnl_unlock();
 	return err;
 }
@@ -2514,7 +2513,8 @@ static void addrconf_dad_start(struct inet6_ifaddr *ifp, u32 flags)
 	spin_lock_bh(&ifp->lock);
 
 	if (dev->flags&(IFF_NOARP|IFF_LOOPBACK) ||
-	    !(ifp->flags&IFA_F_TENTATIVE)) {
+	    !(ifp->flags&IFA_F_TENTATIVE) ||
+	    ifp->flags & IFA_F_NODAD) {
 		ifp->flags &= ~IFA_F_TENTATIVE;
 		spin_unlock_bh(&ifp->lock);
 		read_unlock_bh(&idev->lock);
@@ -2912,28 +2912,25 @@ inet6_rtm_deladdr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 	return inet6_addr_del(ifm->ifa_index, pfx, ifm->ifa_prefixlen);
 }
 
-static int inet6_addr_modify(struct inet6_ifaddr *ifp, u32 prefered_lft,
-			     u32 valid_lft)
+static int inet6_addr_modify(struct inet6_ifaddr *ifp, u8 ifa_flags,
+			     u32 prefered_lft, u32 valid_lft)
 {
-	int ifa_flags = 0;
-
 	if (!valid_lft || (prefered_lft > valid_lft))
 		return -EINVAL;
 
 	if (valid_lft == INFINITY_LIFE_TIME)
-		ifa_flags = IFA_F_PERMANENT;
+		ifa_flags |= IFA_F_PERMANENT;
 	else if (valid_lft >= 0x7FFFFFFF/HZ)
 		valid_lft = 0x7FFFFFFF/HZ;
 
 	if (prefered_lft == 0)
-		ifa_flags = IFA_F_DEPRECATED;
+		ifa_flags |= IFA_F_DEPRECATED;
 	else if ((prefered_lft >= 0x7FFFFFFF/HZ) &&
 		 (prefered_lft != INFINITY_LIFE_TIME))
 		prefered_lft = 0x7FFFFFFF/HZ;
 
 	spin_lock_bh(&ifp->lock);
-	ifp->flags = (ifp->flags & ~(IFA_F_DEPRECATED|IFA_F_PERMANENT)) | ifa_flags;
-
+	ifp->flags = (ifp->flags & ~(IFA_F_DEPRECATED | IFA_F_PERMANENT | IFA_F_NODAD)) | ifa_flags;
 	ifp->tstamp = jiffies;
 	ifp->valid_lft = valid_lft;
 	ifp->prefered_lft = prefered_lft;
@@ -2955,7 +2952,8 @@ inet6_rtm_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 	struct in6_addr *pfx;
 	struct inet6_ifaddr *ifa;
 	struct net_device *dev;
-	u32 valid_lft, preferred_lft;
+	u32 valid_lft = INFINITY_LIFE_TIME, preferred_lft = INFINITY_LIFE_TIME;
+	u8 ifa_flags;
 	int err;
 
 	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv6_policy);
@@ -2982,6 +2980,9 @@ inet6_rtm_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 	if (dev == NULL)
 		return -ENODEV;
 
+	/* We ignore other flags so far. */
+	ifa_flags = ifm->ifa_flags & IFA_F_NODAD;
+
 	ifa = ipv6_get_ifaddr(pfx, dev, 1);
 	if (ifa == NULL) {
 		/*
@@ -2989,14 +2990,14 @@ inet6_rtm_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 		 * userspace alreay relies on not having to provide this.
 		 */
 		return inet6_addr_add(ifm->ifa_index, pfx, ifm->ifa_prefixlen,
-				      preferred_lft, valid_lft);
+				      ifa_flags, preferred_lft, valid_lft);
 	}
 
 	if (nlh->nlmsg_flags & NLM_F_EXCL ||
 	    !(nlh->nlmsg_flags & NLM_F_REPLACE))
 		err = -EEXIST;
 	else
-		err = inet6_addr_modify(ifa, preferred_lft, valid_lft);
+		err = inet6_addr_modify(ifa, ifa_flags, preferred_lft, valid_lft);
 
 	in6_ifa_put(ifa);
 
-- 
cgit v1.2.3


From 3b9f9a1c3903b64c38505f9fed3bb11e48dbc931 Mon Sep 17 00:00:00 2001
From: Noriaki TAKAMIYA <takamiya@po.ntts.co.jp>
Date: Fri, 22 Sep 2006 14:45:56 -0700
Subject: [IPV6] ADDRCONF: Mobile IPv6 Home Address support.

IFA_F_HOMEADDRESS is introduced for Mobile IPv6 Home Addresses on
Mobile Node.

The IFA_F_HOMEADDRESS flag should be set for Mobile IPv6 Home
Addresses for 2 purposes. 1) We need to check this on receipt of
Type 2 Routing Header (RFC3775 Secion 6.4), 2) We prefer Home
Address(es) in source address selection (RFC3484 Section 5 Rule 4).

Signed-off-by: Noriaki TAKAMIYA <takamiya@po.ntts.co.jp>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/if_addr.h |  1 +
 include/net/addrconf.h  |  6 +-----
 net/ipv6/addrconf.c     | 44 +++++++++++++++++++++++++++++++++++++++++---
 3 files changed, 43 insertions(+), 8 deletions(-)

(limited to 'include')

diff --git a/include/linux/if_addr.h b/include/linux/if_addr.h
index ca24b9de13fb..dbe8f6120a40 100644
--- a/include/linux/if_addr.h
+++ b/include/linux/if_addr.h
@@ -39,6 +39,7 @@ enum
 #define IFA_F_TEMPORARY		IFA_F_SECONDARY
 
 #define	IFA_F_NODAD		0x02
+#define	IFA_F_HOMEADDRESS	0x10
 #define IFA_F_DEPRECATED	0x20
 #define IFA_F_TENTATIVE		0x40
 #define IFA_F_PERMANENT		0x80
diff --git a/include/net/addrconf.h b/include/net/addrconf.h
index aa2ed8f0a9dd..44f1b673f916 100644
--- a/include/net/addrconf.h
+++ b/include/net/addrconf.h
@@ -61,12 +61,8 @@ extern int			addrconf_set_dstaddr(void __user *arg);
 extern int			ipv6_chk_addr(struct in6_addr *addr,
 					      struct net_device *dev,
 					      int strict);
-/* XXX: this is a placeholder till addrconf supports */
 #ifdef CONFIG_IPV6_MIP6
-static inline int ipv6_chk_home_addr(struct in6_addr *addr)
-{
-	return 0;
-}
+extern int			ipv6_chk_home_addr(struct in6_addr *addr);
 #endif
 extern struct inet6_ifaddr *	ipv6_get_ifaddr(struct in6_addr *addr,
 						struct net_device *dev,
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index adb583a26151..c18676352397 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -1038,9 +1038,27 @@ int ipv6_dev_get_saddr(struct net_device *daddr_dev,
 					continue;
 			}
 
-			/* Rule 4: Prefer home address -- not implemented yet */
+			/* Rule 4: Prefer home address */
+#ifdef CONFIG_IPV6_MIP6
+			if (hiscore.rule < 4) {
+				if (ifa_result->flags & IFA_F_HOMEADDRESS)
+					hiscore.attrs |= IPV6_SADDR_SCORE_HOA;
+				hiscore.rule++;
+			}
+			if (ifa->flags & IFA_F_HOMEADDRESS) {
+				score.attrs |= IPV6_SADDR_SCORE_HOA;
+				if (!(ifa_result->flags & IFA_F_HOMEADDRESS)) {
+					score.rule = 4;
+					goto record_it;
+				}
+			} else {
+				if (hiscore.attrs & IPV6_SADDR_SCORE_HOA)
+					continue;
+			}
+#else
 			if (hiscore.rule < 4)
 				hiscore.rule++;
+#endif
 
 			/* Rule 5: Prefer outgoing interface */
 			if (hiscore.rule < 5) {
@@ -2759,6 +2777,26 @@ void if6_proc_exit(void)
 }
 #endif	/* CONFIG_PROC_FS */
 
+#ifdef CONFIG_IPV6_MIP6
+/* Check if address is a home address configured on any interface. */
+int ipv6_chk_home_addr(struct in6_addr *addr)
+{
+	int ret = 0;
+	struct inet6_ifaddr * ifp;
+	u8 hash = ipv6_addr_hash(addr);
+	read_lock_bh(&addrconf_hash_lock);
+	for (ifp = inet6_addr_lst[hash]; ifp; ifp = ifp->lst_next) {
+		if (ipv6_addr_cmp(&ifp->addr, addr) == 0 &&
+		    (ifp->flags & IFA_F_HOMEADDRESS)) {
+			ret = 1;
+			break;
+		}
+	}
+	read_unlock_bh(&addrconf_hash_lock);
+	return ret;
+}
+#endif
+
 /*
  *	Periodic address status verification
  */
@@ -2930,7 +2968,7 @@ static int inet6_addr_modify(struct inet6_ifaddr *ifp, u8 ifa_flags,
 		prefered_lft = 0x7FFFFFFF/HZ;
 
 	spin_lock_bh(&ifp->lock);
-	ifp->flags = (ifp->flags & ~(IFA_F_DEPRECATED | IFA_F_PERMANENT | IFA_F_NODAD)) | ifa_flags;
+	ifp->flags = (ifp->flags & ~(IFA_F_DEPRECATED | IFA_F_PERMANENT | IFA_F_NODAD | IFA_F_HOMEADDRESS)) | ifa_flags;
 	ifp->tstamp = jiffies;
 	ifp->valid_lft = valid_lft;
 	ifp->prefered_lft = prefered_lft;
@@ -2981,7 +3019,7 @@ inet6_rtm_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 		return -ENODEV;
 
 	/* We ignore other flags so far. */
-	ifa_flags = ifm->ifa_flags & IFA_F_NODAD;
+	ifa_flags = ifm->ifa_flags & (IFA_F_NODAD | IFA_F_HOMEADDRESS);
 
 	ifa = ipv6_get_ifaddr(pfx, dev, 1);
 	if (ifa == NULL) {
-- 
cgit v1.2.3


From 64f817ba98095156149ba5991592d5d039f6da74 Mon Sep 17 00:00:00 2001
From: Ralph Campbell <ralph.campbell@qlogic.com>
Date: Fri, 22 Sep 2006 15:22:24 -0700
Subject: IB/uverbs: Allow resize CQ operation to return driver-specific data

Add a ib_uverbs_resize_cq_resp.driver_data field so that low-level
drivers can return data from a resize CQ operation to userspace.  Have
ib_uverbs_resize_cq() only copy the cqe field, to avoid having to bump
the userspace ABI.

Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
---
 drivers/infiniband/core/uverbs_cmd.c | 3 +--
 include/rdma/ib_user_verbs.h         | 2 ++
 2 files changed, 3 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index 8b6df7cec0bf..deae43f31e79 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -879,11 +879,10 @@ ssize_t ib_uverbs_resize_cq(struct ib_uverbs_file *file,
 	if (ret)
 		goto out;
 
-	memset(&resp, 0, sizeof resp);
 	resp.cqe = cq->cqe;
 
 	if (copy_to_user((void __user *) (unsigned long) cmd.response,
-			 &resp, sizeof resp))
+			 &resp, sizeof resp.cqe))
 		ret = -EFAULT;
 
 out:
diff --git a/include/rdma/ib_user_verbs.h b/include/rdma/ib_user_verbs.h
index 7b5372010f4b..db1b814b62cc 100644
--- a/include/rdma/ib_user_verbs.h
+++ b/include/rdma/ib_user_verbs.h
@@ -275,6 +275,8 @@ struct ib_uverbs_resize_cq {
 
 struct ib_uverbs_resize_cq_resp {
 	__u32 cqe;
+	__u32 reserved;
+	__u64 driver_data[0];
 };
 
 struct ib_uverbs_poll_cq {
-- 
cgit v1.2.3


From 9bc57e2d19db4da81c1150120658cc3658a99ed4 Mon Sep 17 00:00:00 2001
From: Ralph Campbell <ralphc@pathscale.com>
Date: Fri, 11 Aug 2006 14:58:09 -0700
Subject: IB/uverbs: Pass userspace data to modify_srq and modify_qp methods

Pass a struct ib_udata to the low-level driver's ->modify_srq() and
->modify_qp() methods, so that it can get to the device-specific data
passed in by the userspace driver.

Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
---
 drivers/infiniband/core/uverbs_cmd.c      | 13 ++++++++++---
 drivers/infiniband/core/verbs.c           |  4 ++--
 drivers/infiniband/hw/ehca/ehca_iverbs.h  |  3 ++-
 drivers/infiniband/hw/ehca/ehca_qp.c      |  3 ++-
 drivers/infiniband/hw/ipath/ipath_qp.c    |  3 ++-
 drivers/infiniband/hw/ipath/ipath_srq.c   |  4 +++-
 drivers/infiniband/hw/ipath/ipath_verbs.h |  5 +++--
 drivers/infiniband/hw/mthca/mthca_dev.h   |  5 +++--
 drivers/infiniband/hw/mthca/mthca_qp.c    |  3 ++-
 drivers/infiniband/hw/mthca/mthca_srq.c   |  2 +-
 include/rdma/ib_verbs.h                   |  6 ++++--
 11 files changed, 34 insertions(+), 17 deletions(-)

(limited to 'include')

diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index deae43f31e79..3fcb5d189a23 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -841,7 +841,6 @@ ssize_t ib_uverbs_create_cq(struct ib_uverbs_file *file,
 err_copy:
 	idr_remove_uobj(&ib_uverbs_cq_idr, &obj->uobject);
 
-
 err_free:
 	ib_destroy_cq(cq);
 
@@ -1273,6 +1272,7 @@ ssize_t ib_uverbs_modify_qp(struct ib_uverbs_file *file,
 			    int out_len)
 {
 	struct ib_uverbs_modify_qp cmd;
+	struct ib_udata            udata;
 	struct ib_qp              *qp;
 	struct ib_qp_attr         *attr;
 	int                        ret;
@@ -1280,6 +1280,9 @@ ssize_t ib_uverbs_modify_qp(struct ib_uverbs_file *file,
 	if (copy_from_user(&cmd, buf, sizeof cmd))
 		return -EFAULT;
 
+	INIT_UDATA(&udata, buf + sizeof cmd, NULL, in_len - sizeof cmd,
+		   out_len);
+
 	attr = kmalloc(sizeof *attr, GFP_KERNEL);
 	if (!attr)
 		return -ENOMEM;
@@ -1336,7 +1339,7 @@ ssize_t ib_uverbs_modify_qp(struct ib_uverbs_file *file,
 	attr->alt_ah_attr.ah_flags 	    = cmd.alt_dest.is_global ? IB_AH_GRH : 0;
 	attr->alt_ah_attr.port_num 	    = cmd.alt_dest.port_num;
 
-	ret = ib_modify_qp(qp, attr, cmd.attr_mask);
+	ret = qp->device->modify_qp(qp, attr, cmd.attr_mask, &udata);
 
 	put_qp_read(qp);
 
@@ -2054,6 +2057,7 @@ ssize_t ib_uverbs_modify_srq(struct ib_uverbs_file *file,
 			     int out_len)
 {
 	struct ib_uverbs_modify_srq cmd;
+	struct ib_udata             udata;
 	struct ib_srq              *srq;
 	struct ib_srq_attr          attr;
 	int                         ret;
@@ -2061,6 +2065,9 @@ ssize_t ib_uverbs_modify_srq(struct ib_uverbs_file *file,
 	if (copy_from_user(&cmd, buf, sizeof cmd))
 		return -EFAULT;
 
+	INIT_UDATA(&udata, buf + sizeof cmd, NULL, in_len - sizeof cmd,
+		   out_len);
+
 	srq = idr_read_srq(cmd.srq_handle, file->ucontext);
 	if (!srq)
 		return -EINVAL;
@@ -2068,7 +2075,7 @@ ssize_t ib_uverbs_modify_srq(struct ib_uverbs_file *file,
 	attr.max_wr    = cmd.max_wr;
 	attr.srq_limit = cmd.srq_limit;
 
-	ret = ib_modify_srq(srq, &attr, cmd.attr_mask);
+	ret = srq->device->modify_srq(srq, &attr, cmd.attr_mask, &udata);
 
 	put_srq_read(srq);
 
diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index 468999c38803..06f98e9e14f9 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -231,7 +231,7 @@ int ib_modify_srq(struct ib_srq *srq,
 		  struct ib_srq_attr *srq_attr,
 		  enum ib_srq_attr_mask srq_attr_mask)
 {
-	return srq->device->modify_srq(srq, srq_attr, srq_attr_mask);
+	return srq->device->modify_srq(srq, srq_attr, srq_attr_mask, NULL);
 }
 EXPORT_SYMBOL(ib_modify_srq);
 
@@ -547,7 +547,7 @@ int ib_modify_qp(struct ib_qp *qp,
 		 struct ib_qp_attr *qp_attr,
 		 int qp_attr_mask)
 {
-	return qp->device->modify_qp(qp, qp_attr, qp_attr_mask);
+	return qp->device->modify_qp(qp, qp_attr, qp_attr_mask, NULL);
 }
 EXPORT_SYMBOL(ib_modify_qp);
 
diff --git a/drivers/infiniband/hw/ehca/ehca_iverbs.h b/drivers/infiniband/hw/ehca/ehca_iverbs.h
index bbdc437f5167..319c39d47f3a 100644
--- a/drivers/infiniband/hw/ehca/ehca_iverbs.h
+++ b/drivers/infiniband/hw/ehca/ehca_iverbs.h
@@ -143,7 +143,8 @@ struct ib_qp *ehca_create_qp(struct ib_pd *pd,
 
 int ehca_destroy_qp(struct ib_qp *qp);
 
-int ehca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask);
+int ehca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask,
+		   struct ib_udata *udata);
 
 int ehca_query_qp(struct ib_qp *qp, struct ib_qp_attr *qp_attr,
 		  int qp_attr_mask, struct ib_qp_init_attr *qp_init_attr);
diff --git a/drivers/infiniband/hw/ehca/ehca_qp.c b/drivers/infiniband/hw/ehca/ehca_qp.c
index 4b27bedc6c24..4394123cdbd7 100644
--- a/drivers/infiniband/hw/ehca/ehca_qp.c
+++ b/drivers/infiniband/hw/ehca/ehca_qp.c
@@ -1230,7 +1230,8 @@ modify_qp_exit1:
 	return ret;
 }
 
-int ehca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask)
+int ehca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask,
+		   struct ib_udata *udata)
 {
 	struct ehca_qp *my_qp = container_of(ibqp, struct ehca_qp, ib_qp);
 	struct ehca_pd *my_pd = container_of(my_qp->ib_qp.pd, struct ehca_pd,
diff --git a/drivers/infiniband/hw/ipath/ipath_qp.c b/drivers/infiniband/hw/ipath/ipath_qp.c
index 83e557be591e..44c32d2db990 100644
--- a/drivers/infiniband/hw/ipath/ipath_qp.c
+++ b/drivers/infiniband/hw/ipath/ipath_qp.c
@@ -426,11 +426,12 @@ void ipath_error_qp(struct ipath_qp *qp)
  * @ibqp: the queue pair who's attributes we're modifying
  * @attr: the new attributes
  * @attr_mask: the mask of attributes to modify
+ * @udata: user data for ipathverbs.so
  *
  * Returns 0 on success, otherwise returns an errno.
  */
 int ipath_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
-		    int attr_mask)
+		    int attr_mask, struct ib_udata *udata)
 {
 	struct ipath_ibdev *dev = to_idev(ibqp->device);
 	struct ipath_qp *qp = to_iqp(ibqp);
diff --git a/drivers/infiniband/hw/ipath/ipath_srq.c b/drivers/infiniband/hw/ipath/ipath_srq.c
index f760434660bd..fa77da6667ed 100644
--- a/drivers/infiniband/hw/ipath/ipath_srq.c
+++ b/drivers/infiniband/hw/ipath/ipath_srq.c
@@ -188,9 +188,11 @@ bail:
  * @ibsrq: the SRQ to modify
  * @attr: the new attributes of the SRQ
  * @attr_mask: indicates which attributes to modify
+ * @udata: user data for ipathverbs.so
  */
 int ipath_modify_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr,
-		     enum ib_srq_attr_mask attr_mask)
+		     enum ib_srq_attr_mask attr_mask,
+		     struct ib_udata *udata)
 {
 	struct ipath_srq *srq = to_isrq(ibsrq);
 	unsigned long flags;
diff --git a/drivers/infiniband/hw/ipath/ipath_verbs.h b/drivers/infiniband/hw/ipath/ipath_verbs.h
index 2df684727dc1..698396778f00 100644
--- a/drivers/infiniband/hw/ipath/ipath_verbs.h
+++ b/drivers/infiniband/hw/ipath/ipath_verbs.h
@@ -579,7 +579,7 @@ struct ib_qp *ipath_create_qp(struct ib_pd *ibpd,
 int ipath_destroy_qp(struct ib_qp *ibqp);
 
 int ipath_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
-		    int attr_mask);
+		    int attr_mask, struct ib_udata *udata);
 
 int ipath_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 		   int attr_mask, struct ib_qp_init_attr *init_attr);
@@ -638,7 +638,8 @@ struct ib_srq *ipath_create_srq(struct ib_pd *ibpd,
 				struct ib_udata *udata);
 
 int ipath_modify_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr,
-		     enum ib_srq_attr_mask attr_mask);
+		     enum ib_srq_attr_mask attr_mask,
+		     struct ib_udata *udata);
 
 int ipath_query_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr);
 
diff --git a/drivers/infiniband/hw/mthca/mthca_dev.h b/drivers/infiniband/hw/mthca/mthca_dev.h
index f8160b8de090..33bd0b8bfd13 100644
--- a/drivers/infiniband/hw/mthca/mthca_dev.h
+++ b/drivers/infiniband/hw/mthca/mthca_dev.h
@@ -506,7 +506,7 @@ int mthca_alloc_srq(struct mthca_dev *dev, struct mthca_pd *pd,
 		    struct ib_srq_attr *attr, struct mthca_srq *srq);
 void mthca_free_srq(struct mthca_dev *dev, struct mthca_srq *srq);
 int mthca_modify_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr,
-		     enum ib_srq_attr_mask attr_mask);
+		     enum ib_srq_attr_mask attr_mask, struct ib_udata *udata);
 int mthca_query_srq(struct ib_srq *srq, struct ib_srq_attr *srq_attr);
 int mthca_max_srq_sge(struct mthca_dev *dev);
 void mthca_srq_event(struct mthca_dev *dev, u32 srqn,
@@ -521,7 +521,8 @@ void mthca_qp_event(struct mthca_dev *dev, u32 qpn,
 		    enum ib_event_type event_type);
 int mthca_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, int qp_attr_mask,
 		   struct ib_qp_init_attr *qp_init_attr);
-int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask);
+int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask,
+		    struct ib_udata *udata);
 int mthca_tavor_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 			  struct ib_send_wr **bad_wr);
 int mthca_tavor_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr,
diff --git a/drivers/infiniband/hw/mthca/mthca_qp.c b/drivers/infiniband/hw/mthca/mthca_qp.c
index 2e8f6f36e0a5..6d6ba4180a39 100644
--- a/drivers/infiniband/hw/mthca/mthca_qp.c
+++ b/drivers/infiniband/hw/mthca/mthca_qp.c
@@ -527,7 +527,8 @@ static int mthca_path_set(struct mthca_dev *dev, struct ib_ah_attr *ah,
 	return 0;
 }
 
-int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask)
+int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask,
+		    struct ib_udata *udata)
 {
 	struct mthca_dev *dev = to_mdev(ibqp->device);
 	struct mthca_qp *qp = to_mqp(ibqp);
diff --git a/drivers/infiniband/hw/mthca/mthca_srq.c b/drivers/infiniband/hw/mthca/mthca_srq.c
index b60a9d79ae54..0f316c87bf64 100644
--- a/drivers/infiniband/hw/mthca/mthca_srq.c
+++ b/drivers/infiniband/hw/mthca/mthca_srq.c
@@ -358,7 +358,7 @@ void mthca_free_srq(struct mthca_dev *dev, struct mthca_srq *srq)
 }
 
 int mthca_modify_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr,
-		     enum ib_srq_attr_mask attr_mask)
+		     enum ib_srq_attr_mask attr_mask, struct ib_udata *udata)
 {
 	struct mthca_dev *dev = to_mdev(ibsrq->device);
 	struct mthca_srq *srq = to_msrq(ibsrq);
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index ee1f3a355666..61eed3996117 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -888,7 +888,8 @@ struct ib_device {
 						 struct ib_udata *udata);
 	int                        (*modify_srq)(struct ib_srq *srq,
 						 struct ib_srq_attr *srq_attr,
-						 enum ib_srq_attr_mask srq_attr_mask);
+						 enum ib_srq_attr_mask srq_attr_mask,
+						 struct ib_udata *udata);
 	int                        (*query_srq)(struct ib_srq *srq,
 						struct ib_srq_attr *srq_attr);
 	int                        (*destroy_srq)(struct ib_srq *srq);
@@ -900,7 +901,8 @@ struct ib_device {
 						struct ib_udata *udata);
 	int                        (*modify_qp)(struct ib_qp *qp,
 						struct ib_qp_attr *qp_attr,
-						int qp_attr_mask);
+						int qp_attr_mask,
+						struct ib_udata *udata);
 	int                        (*query_qp)(struct ib_qp *qp,
 					       struct ib_qp_attr *qp_attr,
 					       int qp_attr_mask,
-- 
cgit v1.2.3


From 922a8e9fb2e0711212badce47a41137e2ca04cb3 Mon Sep 17 00:00:00 2001
From: Tom Tucker <tom@opengridcomputing.com>
Date: Thu, 3 Aug 2006 16:02:40 -0500
Subject: RDMA: iWARP Connection Manager.

Add an iWARP Connection Manager (CM), which abstracts connection
management for iWARP devices (RNICs).  It is a logical instance of the
xx_cm where xx is the transport type (ib or iw).  The symbols exported
are used by the transport independent rdma_cm module, and are
available also for transport dependent ULPs.

Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
---
 drivers/infiniband/core/iwcm.c | 1019 ++++++++++++++++++++++++++++++++++++++++
 drivers/infiniband/core/iwcm.h |   62 +++
 include/rdma/iw_cm.h           |  258 ++++++++++
 3 files changed, 1339 insertions(+)
 create mode 100644 drivers/infiniband/core/iwcm.c
 create mode 100644 drivers/infiniband/core/iwcm.h
 create mode 100644 include/rdma/iw_cm.h

(limited to 'include')

diff --git a/drivers/infiniband/core/iwcm.c b/drivers/infiniband/core/iwcm.c
new file mode 100644
index 000000000000..c3fb304a4e86
--- /dev/null
+++ b/drivers/infiniband/core/iwcm.c
@@ -0,0 +1,1019 @@
+/*
+ * Copyright (c) 2004, 2005 Intel Corporation.  All rights reserved.
+ * Copyright (c) 2004 Topspin Corporation.  All rights reserved.
+ * Copyright (c) 2004, 2005 Voltaire Corporation.  All rights reserved.
+ * Copyright (c) 2005 Sun Microsystems, Inc. All rights reserved.
+ * Copyright (c) 2005 Open Grid Computing, Inc. All rights reserved.
+ * Copyright (c) 2005 Network Appliance, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+#include <linux/dma-mapping.h>
+#include <linux/err.h>
+#include <linux/idr.h>
+#include <linux/interrupt.h>
+#include <linux/pci.h>
+#include <linux/rbtree.h>
+#include <linux/spinlock.h>
+#include <linux/workqueue.h>
+#include <linux/completion.h>
+
+#include <rdma/iw_cm.h>
+#include <rdma/ib_addr.h>
+
+#include "iwcm.h"
+
+MODULE_AUTHOR("Tom Tucker");
+MODULE_DESCRIPTION("iWARP CM");
+MODULE_LICENSE("Dual BSD/GPL");
+
+static struct workqueue_struct *iwcm_wq;
+struct iwcm_work {
+	struct work_struct work;
+	struct iwcm_id_private *cm_id;
+	struct list_head list;
+	struct iw_cm_event event;
+	struct list_head free_list;
+};
+
+/*
+ * The following services provide a mechanism for pre-allocating iwcm_work
+ * elements.  The design pre-allocates them  based on the cm_id type:
+ *	LISTENING IDS: 	Get enough elements preallocated to handle the
+ *			listen backlog.
+ *	ACTIVE IDS:	4: CONNECT_REPLY, ESTABLISHED, DISCONNECT, CLOSE
+ *	PASSIVE IDS:	3: ESTABLISHED, DISCONNECT, CLOSE
+ *
+ * Allocating them in connect and listen avoids having to deal
+ * with allocation failures on the event upcall from the provider (which
+ * is called in the interrupt context).
+ *
+ * One exception is when creating the cm_id for incoming connection requests.
+ * There are two cases:
+ * 1) in the event upcall, cm_event_handler(), for a listening cm_id.  If
+ *    the backlog is exceeded, then no more connection request events will
+ *    be processed.  cm_event_handler() returns -ENOMEM in this case.  Its up
+ *    to the provider to reject the connectino request.
+ * 2) in the connection request workqueue handler, cm_conn_req_handler().
+ *    If work elements cannot be allocated for the new connect request cm_id,
+ *    then IWCM will call the provider reject method.  This is ok since
+ *    cm_conn_req_handler() runs in the workqueue thread context.
+ */
+
+static struct iwcm_work *get_work(struct iwcm_id_private *cm_id_priv)
+{
+	struct iwcm_work *work;
+
+	if (list_empty(&cm_id_priv->work_free_list))
+		return NULL;
+	work = list_entry(cm_id_priv->work_free_list.next, struct iwcm_work,
+			  free_list);
+	list_del_init(&work->free_list);
+	return work;
+}
+
+static void put_work(struct iwcm_work *work)
+{
+	list_add(&work->free_list, &work->cm_id->work_free_list);
+}
+
+static void dealloc_work_entries(struct iwcm_id_private *cm_id_priv)
+{
+	struct list_head *e, *tmp;
+
+	list_for_each_safe(e, tmp, &cm_id_priv->work_free_list)
+		kfree(list_entry(e, struct iwcm_work, free_list));
+}
+
+static int alloc_work_entries(struct iwcm_id_private *cm_id_priv, int count)
+{
+	struct iwcm_work *work;
+
+	BUG_ON(!list_empty(&cm_id_priv->work_free_list));
+	while (count--) {
+		work = kmalloc(sizeof(struct iwcm_work), GFP_KERNEL);
+		if (!work) {
+			dealloc_work_entries(cm_id_priv);
+			return -ENOMEM;
+		}
+		work->cm_id = cm_id_priv;
+		INIT_LIST_HEAD(&work->list);
+		put_work(work);
+	}
+	return 0;
+}
+
+/*
+ * Save private data from incoming connection requests in the
+ * cm_id_priv so the low level driver doesn't have to.  Adjust
+ * the event ptr to point to the local copy.
+ */
+static int copy_private_data(struct iwcm_id_private *cm_id_priv,
+		       struct iw_cm_event *event)
+{
+	void *p;
+
+	p = kmalloc(event->private_data_len, GFP_ATOMIC);
+	if (!p)
+		return -ENOMEM;
+	memcpy(p, event->private_data, event->private_data_len);
+	event->private_data = p;
+	return 0;
+}
+
+/*
+ * Release a reference on cm_id. If the last reference is being removed
+ * and iw_destroy_cm_id is waiting, wake up the waiting thread.
+ */
+static int iwcm_deref_id(struct iwcm_id_private *cm_id_priv)
+{
+	int ret = 0;
+
+	BUG_ON(atomic_read(&cm_id_priv->refcount)==0);
+	if (atomic_dec_and_test(&cm_id_priv->refcount)) {
+		BUG_ON(!list_empty(&cm_id_priv->work_list));
+		if (waitqueue_active(&cm_id_priv->destroy_comp.wait)) {
+			BUG_ON(cm_id_priv->state != IW_CM_STATE_DESTROYING);
+			BUG_ON(test_bit(IWCM_F_CALLBACK_DESTROY,
+					&cm_id_priv->flags));
+			ret = 1;
+		}
+		complete(&cm_id_priv->destroy_comp);
+	}
+
+	return ret;
+}
+
+static void add_ref(struct iw_cm_id *cm_id)
+{
+	struct iwcm_id_private *cm_id_priv;
+	cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
+	atomic_inc(&cm_id_priv->refcount);
+}
+
+static void rem_ref(struct iw_cm_id *cm_id)
+{
+	struct iwcm_id_private *cm_id_priv;
+	cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
+	iwcm_deref_id(cm_id_priv);
+}
+
+static int cm_event_handler(struct iw_cm_id *cm_id, struct iw_cm_event *event);
+
+struct iw_cm_id *iw_create_cm_id(struct ib_device *device,
+				 iw_cm_handler cm_handler,
+				 void *context)
+{
+	struct iwcm_id_private *cm_id_priv;
+
+	cm_id_priv = kzalloc(sizeof(*cm_id_priv), GFP_KERNEL);
+	if (!cm_id_priv)
+		return ERR_PTR(-ENOMEM);
+
+	cm_id_priv->state = IW_CM_STATE_IDLE;
+	cm_id_priv->id.device = device;
+	cm_id_priv->id.cm_handler = cm_handler;
+	cm_id_priv->id.context = context;
+	cm_id_priv->id.event_handler = cm_event_handler;
+	cm_id_priv->id.add_ref = add_ref;
+	cm_id_priv->id.rem_ref = rem_ref;
+	spin_lock_init(&cm_id_priv->lock);
+	atomic_set(&cm_id_priv->refcount, 1);
+	init_waitqueue_head(&cm_id_priv->connect_wait);
+	init_completion(&cm_id_priv->destroy_comp);
+	INIT_LIST_HEAD(&cm_id_priv->work_list);
+	INIT_LIST_HEAD(&cm_id_priv->work_free_list);
+
+	return &cm_id_priv->id;
+}
+EXPORT_SYMBOL(iw_create_cm_id);
+
+
+static int iwcm_modify_qp_err(struct ib_qp *qp)
+{
+	struct ib_qp_attr qp_attr;
+
+	if (!qp)
+		return -EINVAL;
+
+	qp_attr.qp_state = IB_QPS_ERR;
+	return ib_modify_qp(qp, &qp_attr, IB_QP_STATE);
+}
+
+/*
+ * This is really the RDMAC CLOSING state. It is most similar to the
+ * IB SQD QP state.
+ */
+static int iwcm_modify_qp_sqd(struct ib_qp *qp)
+{
+	struct ib_qp_attr qp_attr;
+
+	BUG_ON(qp == NULL);
+	qp_attr.qp_state = IB_QPS_SQD;
+	return ib_modify_qp(qp, &qp_attr, IB_QP_STATE);
+}
+
+/*
+ * CM_ID <-- CLOSING
+ *
+ * Block if a passive or active connection is currenlty being processed. Then
+ * process the event as follows:
+ * - If we are ESTABLISHED, move to CLOSING and modify the QP state
+ *   based on the abrupt flag
+ * - If the connection is already in the CLOSING or IDLE state, the peer is
+ *   disconnecting concurrently with us and we've already seen the
+ *   DISCONNECT event -- ignore the request and return 0
+ * - Disconnect on a listening endpoint returns -EINVAL
+ */
+int iw_cm_disconnect(struct iw_cm_id *cm_id, int abrupt)
+{
+	struct iwcm_id_private *cm_id_priv;
+	unsigned long flags;
+	int ret = 0;
+	struct ib_qp *qp = NULL;
+
+	cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
+	/* Wait if we're currently in a connect or accept downcall */
+	wait_event(cm_id_priv->connect_wait,
+		   !test_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags));
+
+	spin_lock_irqsave(&cm_id_priv->lock, flags);
+	switch (cm_id_priv->state) {
+	case IW_CM_STATE_ESTABLISHED:
+		cm_id_priv->state = IW_CM_STATE_CLOSING;
+
+		/* QP could be <nul> for user-mode client */
+		if (cm_id_priv->qp)
+			qp = cm_id_priv->qp;
+		else
+			ret = -EINVAL;
+		break;
+	case IW_CM_STATE_LISTEN:
+		ret = -EINVAL;
+		break;
+	case IW_CM_STATE_CLOSING:
+		/* remote peer closed first */
+	case IW_CM_STATE_IDLE:
+		/* accept or connect returned !0 */
+		break;
+	case IW_CM_STATE_CONN_RECV:
+		/*
+		 * App called disconnect before/without calling accept after
+		 * connect_request event delivered.
+		 */
+		break;
+	case IW_CM_STATE_CONN_SENT:
+		/* Can only get here if wait above fails */
+	default:
+		BUG();
+	}
+	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+
+	if (qp) {
+		if (abrupt)
+			ret = iwcm_modify_qp_err(qp);
+		else
+			ret = iwcm_modify_qp_sqd(qp);
+
+		/*
+		 * If both sides are disconnecting the QP could
+		 * already be in ERR or SQD states
+		 */
+		ret = 0;
+	}
+
+	return ret;
+}
+EXPORT_SYMBOL(iw_cm_disconnect);
+
+/*
+ * CM_ID <-- DESTROYING
+ *
+ * Clean up all resources associated with the connection and release
+ * the initial reference taken by iw_create_cm_id.
+ */
+static void destroy_cm_id(struct iw_cm_id *cm_id)
+{
+	struct iwcm_id_private *cm_id_priv;
+	unsigned long flags;
+	int ret;
+
+	cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
+	/*
+	 * Wait if we're currently in a connect or accept downcall. A
+	 * listening endpoint should never block here.
+	 */
+	wait_event(cm_id_priv->connect_wait,
+		   !test_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags));
+
+	spin_lock_irqsave(&cm_id_priv->lock, flags);
+	switch (cm_id_priv->state) {
+	case IW_CM_STATE_LISTEN:
+		cm_id_priv->state = IW_CM_STATE_DESTROYING;
+		spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+		/* destroy the listening endpoint */
+		ret = cm_id->device->iwcm->destroy_listen(cm_id);
+		spin_lock_irqsave(&cm_id_priv->lock, flags);
+		break;
+	case IW_CM_STATE_ESTABLISHED:
+		cm_id_priv->state = IW_CM_STATE_DESTROYING;
+		spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+		/* Abrupt close of the connection */
+		(void)iwcm_modify_qp_err(cm_id_priv->qp);
+		spin_lock_irqsave(&cm_id_priv->lock, flags);
+		break;
+	case IW_CM_STATE_IDLE:
+	case IW_CM_STATE_CLOSING:
+		cm_id_priv->state = IW_CM_STATE_DESTROYING;
+		break;
+	case IW_CM_STATE_CONN_RECV:
+		/*
+		 * App called destroy before/without calling accept after
+		 * receiving connection request event notification.
+		 */
+		cm_id_priv->state = IW_CM_STATE_DESTROYING;
+		break;
+	case IW_CM_STATE_CONN_SENT:
+	case IW_CM_STATE_DESTROYING:
+	default:
+		BUG();
+		break;
+	}
+	if (cm_id_priv->qp) {
+		cm_id_priv->id.device->iwcm->rem_ref(cm_id_priv->qp);
+		cm_id_priv->qp = NULL;
+	}
+	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+
+	(void)iwcm_deref_id(cm_id_priv);
+}
+
+/*
+ * This function is only called by the application thread and cannot
+ * be called by the event thread. The function will wait for all
+ * references to be released on the cm_id and then kfree the cm_id
+ * object.
+ */
+void iw_destroy_cm_id(struct iw_cm_id *cm_id)
+{
+	struct iwcm_id_private *cm_id_priv;
+
+	cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
+	BUG_ON(test_bit(IWCM_F_CALLBACK_DESTROY, &cm_id_priv->flags));
+
+	destroy_cm_id(cm_id);
+
+	wait_for_completion(&cm_id_priv->destroy_comp);
+
+	dealloc_work_entries(cm_id_priv);
+
+	kfree(cm_id_priv);
+}
+EXPORT_SYMBOL(iw_destroy_cm_id);
+
+/*
+ * CM_ID <-- LISTEN
+ *
+ * Start listening for connect requests. Generates one CONNECT_REQUEST
+ * event for each inbound connect request.
+ */
+int iw_cm_listen(struct iw_cm_id *cm_id, int backlog)
+{
+	struct iwcm_id_private *cm_id_priv;
+	unsigned long flags;
+	int ret = 0;
+
+	cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
+
+	ret = alloc_work_entries(cm_id_priv, backlog);
+	if (ret)
+		return ret;
+
+	spin_lock_irqsave(&cm_id_priv->lock, flags);
+	switch (cm_id_priv->state) {
+	case IW_CM_STATE_IDLE:
+		cm_id_priv->state = IW_CM_STATE_LISTEN;
+		spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+		ret = cm_id->device->iwcm->create_listen(cm_id, backlog);
+		if (ret)
+			cm_id_priv->state = IW_CM_STATE_IDLE;
+		spin_lock_irqsave(&cm_id_priv->lock, flags);
+		break;
+	default:
+		ret = -EINVAL;
+	}
+	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+
+	return ret;
+}
+EXPORT_SYMBOL(iw_cm_listen);
+
+/*
+ * CM_ID <-- IDLE
+ *
+ * Rejects an inbound connection request. No events are generated.
+ */
+int iw_cm_reject(struct iw_cm_id *cm_id,
+		 const void *private_data,
+		 u8 private_data_len)
+{
+	struct iwcm_id_private *cm_id_priv;
+	unsigned long flags;
+	int ret;
+
+	cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
+	set_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
+
+	spin_lock_irqsave(&cm_id_priv->lock, flags);
+	if (cm_id_priv->state != IW_CM_STATE_CONN_RECV) {
+		spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+		clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
+		wake_up_all(&cm_id_priv->connect_wait);
+		return -EINVAL;
+	}
+	cm_id_priv->state = IW_CM_STATE_IDLE;
+	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+
+	ret = cm_id->device->iwcm->reject(cm_id, private_data,
+					  private_data_len);
+
+	clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
+	wake_up_all(&cm_id_priv->connect_wait);
+
+	return ret;
+}
+EXPORT_SYMBOL(iw_cm_reject);
+
+/*
+ * CM_ID <-- ESTABLISHED
+ *
+ * Accepts an inbound connection request and generates an ESTABLISHED
+ * event. Callers of iw_cm_disconnect and iw_destroy_cm_id will block
+ * until the ESTABLISHED event is received from the provider.
+ */
+int iw_cm_accept(struct iw_cm_id *cm_id,
+		 struct iw_cm_conn_param *iw_param)
+{
+	struct iwcm_id_private *cm_id_priv;
+	struct ib_qp *qp;
+	unsigned long flags;
+	int ret;
+
+	cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
+	set_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
+
+	spin_lock_irqsave(&cm_id_priv->lock, flags);
+	if (cm_id_priv->state != IW_CM_STATE_CONN_RECV) {
+		spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+		clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
+		wake_up_all(&cm_id_priv->connect_wait);
+		return -EINVAL;
+	}
+	/* Get the ib_qp given the QPN */
+	qp = cm_id->device->iwcm->get_qp(cm_id->device, iw_param->qpn);
+	if (!qp) {
+		spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+		return -EINVAL;
+	}
+	cm_id->device->iwcm->add_ref(qp);
+	cm_id_priv->qp = qp;
+	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+
+	ret = cm_id->device->iwcm->accept(cm_id, iw_param);
+	if (ret) {
+		/* An error on accept precludes provider events */
+		BUG_ON(cm_id_priv->state != IW_CM_STATE_CONN_RECV);
+		cm_id_priv->state = IW_CM_STATE_IDLE;
+		spin_lock_irqsave(&cm_id_priv->lock, flags);
+		if (cm_id_priv->qp) {
+			cm_id->device->iwcm->rem_ref(qp);
+			cm_id_priv->qp = NULL;
+		}
+		spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+		clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
+		wake_up_all(&cm_id_priv->connect_wait);
+	}
+
+	return ret;
+}
+EXPORT_SYMBOL(iw_cm_accept);
+
+/*
+ * Active Side: CM_ID <-- CONN_SENT
+ *
+ * If successful, results in the generation of a CONNECT_REPLY
+ * event. iw_cm_disconnect and iw_cm_destroy will block until the
+ * CONNECT_REPLY event is received from the provider.
+ */
+int iw_cm_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *iw_param)
+{
+	struct iwcm_id_private *cm_id_priv;
+	int ret = 0;
+	unsigned long flags;
+	struct ib_qp *qp;
+
+	cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
+
+	ret = alloc_work_entries(cm_id_priv, 4);
+	if (ret)
+		return ret;
+
+	set_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
+	spin_lock_irqsave(&cm_id_priv->lock, flags);
+
+	if (cm_id_priv->state != IW_CM_STATE_IDLE) {
+		spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+		clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
+		wake_up_all(&cm_id_priv->connect_wait);
+		return -EINVAL;
+	}
+
+	/* Get the ib_qp given the QPN */
+	qp = cm_id->device->iwcm->get_qp(cm_id->device, iw_param->qpn);
+	if (!qp) {
+		spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+		return -EINVAL;
+	}
+	cm_id->device->iwcm->add_ref(qp);
+	cm_id_priv->qp = qp;
+	cm_id_priv->state = IW_CM_STATE_CONN_SENT;
+	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+
+	ret = cm_id->device->iwcm->connect(cm_id, iw_param);
+	if (ret) {
+		spin_lock_irqsave(&cm_id_priv->lock, flags);
+		if (cm_id_priv->qp) {
+			cm_id->device->iwcm->rem_ref(qp);
+			cm_id_priv->qp = NULL;
+		}
+		spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+		BUG_ON(cm_id_priv->state != IW_CM_STATE_CONN_SENT);
+		cm_id_priv->state = IW_CM_STATE_IDLE;
+		clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
+		wake_up_all(&cm_id_priv->connect_wait);
+	}
+
+	return ret;
+}
+EXPORT_SYMBOL(iw_cm_connect);
+
+/*
+ * Passive Side: new CM_ID <-- CONN_RECV
+ *
+ * Handles an inbound connect request. The function creates a new
+ * iw_cm_id to represent the new connection and inherits the client
+ * callback function and other attributes from the listening parent.
+ *
+ * The work item contains a pointer to the listen_cm_id and the event. The
+ * listen_cm_id contains the client cm_handler, context and
+ * device. These are copied when the device is cloned. The event
+ * contains the new four tuple.
+ *
+ * An error on the child should not affect the parent, so this
+ * function does not return a value.
+ */
+static void cm_conn_req_handler(struct iwcm_id_private *listen_id_priv,
+				struct iw_cm_event *iw_event)
+{
+	unsigned long flags;
+	struct iw_cm_id *cm_id;
+	struct iwcm_id_private *cm_id_priv;
+	int ret;
+
+	/*
+	 * The provider should never generate a connection request
+	 * event with a bad status.
+	 */
+	BUG_ON(iw_event->status);
+
+	/*
+	 * We could be destroying the listening id. If so, ignore this
+	 * upcall.
+	 */
+	spin_lock_irqsave(&listen_id_priv->lock, flags);
+	if (listen_id_priv->state != IW_CM_STATE_LISTEN) {
+		spin_unlock_irqrestore(&listen_id_priv->lock, flags);
+		return;
+	}
+	spin_unlock_irqrestore(&listen_id_priv->lock, flags);
+
+	cm_id = iw_create_cm_id(listen_id_priv->id.device,
+				listen_id_priv->id.cm_handler,
+				listen_id_priv->id.context);
+	/* If the cm_id could not be created, ignore the request */
+	if (IS_ERR(cm_id))
+		return;
+
+	cm_id->provider_data = iw_event->provider_data;
+	cm_id->local_addr = iw_event->local_addr;
+	cm_id->remote_addr = iw_event->remote_addr;
+
+	cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
+	cm_id_priv->state = IW_CM_STATE_CONN_RECV;
+
+	ret = alloc_work_entries(cm_id_priv, 3);
+	if (ret) {
+		iw_cm_reject(cm_id, NULL, 0);
+		iw_destroy_cm_id(cm_id);
+		return;
+	}
+
+	/* Call the client CM handler */
+	ret = cm_id->cm_handler(cm_id, iw_event);
+	if (ret) {
+		set_bit(IWCM_F_CALLBACK_DESTROY, &cm_id_priv->flags);
+		destroy_cm_id(cm_id);
+		if (atomic_read(&cm_id_priv->refcount)==0)
+			kfree(cm_id);
+	}
+
+	if (iw_event->private_data_len)
+		kfree(iw_event->private_data);
+}
+
+/*
+ * Passive Side: CM_ID <-- ESTABLISHED
+ *
+ * The provider generated an ESTABLISHED event which means that
+ * the MPA negotion has completed successfully and we are now in MPA
+ * FPDU mode.
+ *
+ * This event can only be received in the CONN_RECV state. If the
+ * remote peer closed, the ESTABLISHED event would be received followed
+ * by the CLOSE event. If the app closes, it will block until we wake
+ * it up after processing this event.
+ */
+static int cm_conn_est_handler(struct iwcm_id_private *cm_id_priv,
+			       struct iw_cm_event *iw_event)
+{
+	unsigned long flags;
+	int ret = 0;
+
+	spin_lock_irqsave(&cm_id_priv->lock, flags);
+
+	/*
+	 * We clear the CONNECT_WAIT bit here to allow the callback
+	 * function to call iw_cm_disconnect. Calling iw_destroy_cm_id
+	 * from a callback handler is not allowed.
+	 */
+	clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
+	BUG_ON(cm_id_priv->state != IW_CM_STATE_CONN_RECV);
+	cm_id_priv->state = IW_CM_STATE_ESTABLISHED;
+	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+	ret = cm_id_priv->id.cm_handler(&cm_id_priv->id, iw_event);
+	wake_up_all(&cm_id_priv->connect_wait);
+
+	return ret;
+}
+
+/*
+ * Active Side: CM_ID <-- ESTABLISHED
+ *
+ * The app has called connect and is waiting for the established event to
+ * post it's requests to the server. This event will wake up anyone
+ * blocked in iw_cm_disconnect or iw_destroy_id.
+ */
+static int cm_conn_rep_handler(struct iwcm_id_private *cm_id_priv,
+			       struct iw_cm_event *iw_event)
+{
+	unsigned long flags;
+	int ret = 0;
+
+	spin_lock_irqsave(&cm_id_priv->lock, flags);
+	/*
+	 * Clear the connect wait bit so a callback function calling
+	 * iw_cm_disconnect will not wait and deadlock this thread
+	 */
+	clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
+	BUG_ON(cm_id_priv->state != IW_CM_STATE_CONN_SENT);
+	if (iw_event->status == IW_CM_EVENT_STATUS_ACCEPTED) {
+		cm_id_priv->id.local_addr = iw_event->local_addr;
+		cm_id_priv->id.remote_addr = iw_event->remote_addr;
+		cm_id_priv->state = IW_CM_STATE_ESTABLISHED;
+	} else {
+		/* REJECTED or RESET */
+		cm_id_priv->id.device->iwcm->rem_ref(cm_id_priv->qp);
+		cm_id_priv->qp = NULL;
+		cm_id_priv->state = IW_CM_STATE_IDLE;
+	}
+	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+	ret = cm_id_priv->id.cm_handler(&cm_id_priv->id, iw_event);
+
+	if (iw_event->private_data_len)
+		kfree(iw_event->private_data);
+
+	/* Wake up waiters on connect complete */
+	wake_up_all(&cm_id_priv->connect_wait);
+
+	return ret;
+}
+
+/*
+ * CM_ID <-- CLOSING
+ *
+ * If in the ESTABLISHED state, move to CLOSING.
+ */
+static void cm_disconnect_handler(struct iwcm_id_private *cm_id_priv,
+				  struct iw_cm_event *iw_event)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&cm_id_priv->lock, flags);
+	if (cm_id_priv->state == IW_CM_STATE_ESTABLISHED)
+		cm_id_priv->state = IW_CM_STATE_CLOSING;
+	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+}
+
+/*
+ * CM_ID <-- IDLE
+ *
+ * If in the ESTBLISHED or CLOSING states, the QP will have have been
+ * moved by the provider to the ERR state. Disassociate the CM_ID from
+ * the QP,  move to IDLE, and remove the 'connected' reference.
+ *
+ * If in some other state, the cm_id was destroyed asynchronously.
+ * This is the last reference that will result in waking up
+ * the app thread blocked in iw_destroy_cm_id.
+ */
+static int cm_close_handler(struct iwcm_id_private *cm_id_priv,
+				  struct iw_cm_event *iw_event)
+{
+	unsigned long flags;
+	int ret = 0;
+	spin_lock_irqsave(&cm_id_priv->lock, flags);
+
+	if (cm_id_priv->qp) {
+		cm_id_priv->id.device->iwcm->rem_ref(cm_id_priv->qp);
+		cm_id_priv->qp = NULL;
+	}
+	switch (cm_id_priv->state) {
+	case IW_CM_STATE_ESTABLISHED:
+	case IW_CM_STATE_CLOSING:
+		cm_id_priv->state = IW_CM_STATE_IDLE;
+		spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+		ret = cm_id_priv->id.cm_handler(&cm_id_priv->id, iw_event);
+		spin_lock_irqsave(&cm_id_priv->lock, flags);
+		break;
+	case IW_CM_STATE_DESTROYING:
+		break;
+	default:
+		BUG();
+	}
+	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+
+	return ret;
+}
+
+static int process_event(struct iwcm_id_private *cm_id_priv,
+			 struct iw_cm_event *iw_event)
+{
+	int ret = 0;
+
+	switch (iw_event->event) {
+	case IW_CM_EVENT_CONNECT_REQUEST:
+		cm_conn_req_handler(cm_id_priv, iw_event);
+		break;
+	case IW_CM_EVENT_CONNECT_REPLY:
+		ret = cm_conn_rep_handler(cm_id_priv, iw_event);
+		break;
+	case IW_CM_EVENT_ESTABLISHED:
+		ret = cm_conn_est_handler(cm_id_priv, iw_event);
+		break;
+	case IW_CM_EVENT_DISCONNECT:
+		cm_disconnect_handler(cm_id_priv, iw_event);
+		break;
+	case IW_CM_EVENT_CLOSE:
+		ret = cm_close_handler(cm_id_priv, iw_event);
+		break;
+	default:
+		BUG();
+	}
+
+	return ret;
+}
+
+/*
+ * Process events on the work_list for the cm_id. If the callback
+ * function requests that the cm_id be deleted, a flag is set in the
+ * cm_id flags to indicate that when the last reference is
+ * removed, the cm_id is to be destroyed. This is necessary to
+ * distinguish between an object that will be destroyed by the app
+ * thread asleep on the destroy_comp list vs. an object destroyed
+ * here synchronously when the last reference is removed.
+ */
+static void cm_work_handler(void *arg)
+{
+	struct iwcm_work *work = arg, lwork;
+	struct iwcm_id_private *cm_id_priv = work->cm_id;
+	unsigned long flags;
+	int empty;
+	int ret = 0;
+
+	spin_lock_irqsave(&cm_id_priv->lock, flags);
+	empty = list_empty(&cm_id_priv->work_list);
+	while (!empty) {
+		work = list_entry(cm_id_priv->work_list.next,
+				  struct iwcm_work, list);
+		list_del_init(&work->list);
+		empty = list_empty(&cm_id_priv->work_list);
+		lwork = *work;
+		put_work(work);
+		spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+
+		ret = process_event(cm_id_priv, &work->event);
+		if (ret) {
+			set_bit(IWCM_F_CALLBACK_DESTROY, &cm_id_priv->flags);
+			destroy_cm_id(&cm_id_priv->id);
+		}
+		BUG_ON(atomic_read(&cm_id_priv->refcount)==0);
+		if (iwcm_deref_id(cm_id_priv))
+			return;
+
+		if (atomic_read(&cm_id_priv->refcount)==0 &&
+		    test_bit(IWCM_F_CALLBACK_DESTROY, &cm_id_priv->flags)) {
+			dealloc_work_entries(cm_id_priv);
+			kfree(cm_id_priv);
+			return;
+		}
+		spin_lock_irqsave(&cm_id_priv->lock, flags);
+	}
+	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+}
+
+/*
+ * This function is called on interrupt context. Schedule events on
+ * the iwcm_wq thread to allow callback functions to downcall into
+ * the CM and/or block.  Events are queued to a per-CM_ID
+ * work_list. If this is the first event on the work_list, the work
+ * element is also queued on the iwcm_wq thread.
+ *
+ * Each event holds a reference on the cm_id. Until the last posted
+ * event has been delivered and processed, the cm_id cannot be
+ * deleted.
+ *
+ * Returns:
+ * 	      0	- the event was handled.
+ *	-ENOMEM	- the event was not handled due to lack of resources.
+ */
+static int cm_event_handler(struct iw_cm_id *cm_id,
+			     struct iw_cm_event *iw_event)
+{
+	struct iwcm_work *work;
+	struct iwcm_id_private *cm_id_priv;
+	unsigned long flags;
+	int ret = 0;
+
+	cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
+
+	spin_lock_irqsave(&cm_id_priv->lock, flags);
+	work = get_work(cm_id_priv);
+	if (!work) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	INIT_WORK(&work->work, cm_work_handler, work);
+	work->cm_id = cm_id_priv;
+	work->event = *iw_event;
+
+	if ((work->event.event == IW_CM_EVENT_CONNECT_REQUEST ||
+	     work->event.event == IW_CM_EVENT_CONNECT_REPLY) &&
+	    work->event.private_data_len) {
+		ret = copy_private_data(cm_id_priv, &work->event);
+		if (ret) {
+			put_work(work);
+			goto out;
+		}
+	}
+
+	atomic_inc(&cm_id_priv->refcount);
+	if (list_empty(&cm_id_priv->work_list)) {
+		list_add_tail(&work->list, &cm_id_priv->work_list);
+		queue_work(iwcm_wq, &work->work);
+	} else
+		list_add_tail(&work->list, &cm_id_priv->work_list);
+out:
+	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+	return ret;
+}
+
+static int iwcm_init_qp_init_attr(struct iwcm_id_private *cm_id_priv,
+				  struct ib_qp_attr *qp_attr,
+				  int *qp_attr_mask)
+{
+	unsigned long flags;
+	int ret;
+
+	spin_lock_irqsave(&cm_id_priv->lock, flags);
+	switch (cm_id_priv->state) {
+	case IW_CM_STATE_IDLE:
+	case IW_CM_STATE_CONN_SENT:
+	case IW_CM_STATE_CONN_RECV:
+	case IW_CM_STATE_ESTABLISHED:
+		*qp_attr_mask = IB_QP_STATE | IB_QP_ACCESS_FLAGS;
+		qp_attr->qp_access_flags = IB_ACCESS_LOCAL_WRITE |
+					   IB_ACCESS_REMOTE_WRITE|
+					   IB_ACCESS_REMOTE_READ;
+		ret = 0;
+		break;
+	default:
+		ret = -EINVAL;
+		break;
+	}
+	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+	return ret;
+}
+
+static int iwcm_init_qp_rts_attr(struct iwcm_id_private *cm_id_priv,
+				  struct ib_qp_attr *qp_attr,
+				  int *qp_attr_mask)
+{
+	unsigned long flags;
+	int ret;
+
+	spin_lock_irqsave(&cm_id_priv->lock, flags);
+	switch (cm_id_priv->state) {
+	case IW_CM_STATE_IDLE:
+	case IW_CM_STATE_CONN_SENT:
+	case IW_CM_STATE_CONN_RECV:
+	case IW_CM_STATE_ESTABLISHED:
+		*qp_attr_mask = 0;
+		ret = 0;
+		break;
+	default:
+		ret = -EINVAL;
+		break;
+	}
+	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+	return ret;
+}
+
+int iw_cm_init_qp_attr(struct iw_cm_id *cm_id,
+		       struct ib_qp_attr *qp_attr,
+		       int *qp_attr_mask)
+{
+	struct iwcm_id_private *cm_id_priv;
+	int ret;
+
+	cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
+	switch (qp_attr->qp_state) {
+	case IB_QPS_INIT:
+	case IB_QPS_RTR:
+		ret = iwcm_init_qp_init_attr(cm_id_priv,
+					     qp_attr, qp_attr_mask);
+		break;
+	case IB_QPS_RTS:
+		ret = iwcm_init_qp_rts_attr(cm_id_priv,
+					    qp_attr, qp_attr_mask);
+		break;
+	default:
+		ret = -EINVAL;
+		break;
+	}
+	return ret;
+}
+EXPORT_SYMBOL(iw_cm_init_qp_attr);
+
+static int __init iw_cm_init(void)
+{
+	iwcm_wq = create_singlethread_workqueue("iw_cm_wq");
+	if (!iwcm_wq)
+		return -ENOMEM;
+
+	return 0;
+}
+
+static void __exit iw_cm_cleanup(void)
+{
+	destroy_workqueue(iwcm_wq);
+}
+
+module_init(iw_cm_init);
+module_exit(iw_cm_cleanup);
diff --git a/drivers/infiniband/core/iwcm.h b/drivers/infiniband/core/iwcm.h
new file mode 100644
index 000000000000..3f6cc82564c8
--- /dev/null
+++ b/drivers/infiniband/core/iwcm.h
@@ -0,0 +1,62 @@
+/*
+ * Copyright (c) 2005 Network Appliance, Inc. All rights reserved.
+ * Copyright (c) 2005 Open Grid Computing, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#ifndef IWCM_H
+#define IWCM_H
+
+enum iw_cm_state {
+	IW_CM_STATE_IDLE,             /* unbound, inactive */
+	IW_CM_STATE_LISTEN,           /* listen waiting for connect */
+	IW_CM_STATE_CONN_RECV,        /* inbound waiting for user accept */
+	IW_CM_STATE_CONN_SENT,        /* outbound waiting for peer accept */
+	IW_CM_STATE_ESTABLISHED,      /* established */
+	IW_CM_STATE_CLOSING,	      /* disconnect */
+	IW_CM_STATE_DESTROYING        /* object being deleted */
+};
+
+struct iwcm_id_private {
+	struct iw_cm_id	id;
+	enum iw_cm_state state;
+	unsigned long flags;
+	struct ib_qp *qp;
+	struct completion destroy_comp;
+	wait_queue_head_t connect_wait;
+	struct list_head work_list;
+	spinlock_t lock;
+	atomic_t refcount;
+	struct list_head work_free_list;
+};
+
+#define IWCM_F_CALLBACK_DESTROY   1
+#define IWCM_F_CONNECT_WAIT       2
+
+#endif /* IWCM_H */
diff --git a/include/rdma/iw_cm.h b/include/rdma/iw_cm.h
new file mode 100644
index 000000000000..aeefa9b740dc
--- /dev/null
+++ b/include/rdma/iw_cm.h
@@ -0,0 +1,258 @@
+/*
+ * Copyright (c) 2005 Network Appliance, Inc. All rights reserved.
+ * Copyright (c) 2005 Open Grid Computing, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#ifndef IW_CM_H
+#define IW_CM_H
+
+#include <linux/in.h>
+#include <rdma/ib_cm.h>
+
+struct iw_cm_id;
+
+enum iw_cm_event_type {
+	IW_CM_EVENT_CONNECT_REQUEST = 1, /* connect request received */
+	IW_CM_EVENT_CONNECT_REPLY,	 /* reply from active connect request */
+	IW_CM_EVENT_ESTABLISHED,	 /* passive side accept successful */
+	IW_CM_EVENT_DISCONNECT,		 /* orderly shutdown */
+	IW_CM_EVENT_CLOSE		 /* close complete */
+};
+
+enum iw_cm_event_status {
+	IW_CM_EVENT_STATUS_OK = 0,	 /* request successful */
+	IW_CM_EVENT_STATUS_ACCEPTED = 0, /* connect request accepted */
+	IW_CM_EVENT_STATUS_REJECTED,	 /* connect request rejected */
+	IW_CM_EVENT_STATUS_TIMEOUT,	 /* the operation timed out */
+	IW_CM_EVENT_STATUS_RESET,	 /* reset from remote peer */
+	IW_CM_EVENT_STATUS_EINVAL,	 /* asynchronous failure for bad parm */
+};
+
+struct iw_cm_event {
+	enum iw_cm_event_type event;
+	enum iw_cm_event_status status;
+	struct sockaddr_in local_addr;
+	struct sockaddr_in remote_addr;
+	void *private_data;
+	u8 private_data_len;
+	void* provider_data;
+};
+
+/**
+ * iw_cm_handler - Function to be called by the IW CM when delivering events
+ * to the client.
+ *
+ * @cm_id: The IW CM identifier associated with the event.
+ * @event: Pointer to the event structure.
+ */
+typedef int (*iw_cm_handler)(struct iw_cm_id *cm_id,
+			     struct iw_cm_event *event);
+
+/**
+ * iw_event_handler - Function called by the provider when delivering provider
+ * events to the IW CM.  Returns either 0 indicating the event was processed
+ * or -errno if the event could not be processed.
+ *
+ * @cm_id: The IW CM identifier associated with the event.
+ * @event: Pointer to the event structure.
+ */
+typedef int (*iw_event_handler)(struct iw_cm_id *cm_id,
+				 struct iw_cm_event *event);
+
+struct iw_cm_id {
+	iw_cm_handler		cm_handler;      /* client callback function */
+	void		        *context;	 /* client cb context */
+	struct ib_device	*device;
+	struct sockaddr_in      local_addr;
+	struct sockaddr_in	remote_addr;
+	void			*provider_data;	 /* provider private data */
+	iw_event_handler        event_handler;   /* cb for provider
+						    events */
+	/* Used by provider to add and remove refs on IW cm_id */
+	void (*add_ref)(struct iw_cm_id *);
+	void (*rem_ref)(struct iw_cm_id *);
+};
+
+struct iw_cm_conn_param {
+	const void *private_data;
+	u16 private_data_len;
+	u32 ord;
+	u32 ird;
+	u32 qpn;
+};
+
+struct iw_cm_verbs {
+	void		(*add_ref)(struct ib_qp *qp);
+
+	void		(*rem_ref)(struct ib_qp *qp);
+
+	struct ib_qp *	(*get_qp)(struct ib_device *device,
+				  int qpn);
+
+	int		(*connect)(struct iw_cm_id *cm_id,
+				   struct iw_cm_conn_param *conn_param);
+
+	int		(*accept)(struct iw_cm_id *cm_id,
+				  struct iw_cm_conn_param *conn_param);
+
+	int		(*reject)(struct iw_cm_id *cm_id,
+				  const void *pdata, u8 pdata_len);
+
+	int		(*create_listen)(struct iw_cm_id *cm_id,
+					 int backlog);
+
+	int		(*destroy_listen)(struct iw_cm_id *cm_id);
+};
+
+/**
+ * iw_create_cm_id - Create an IW CM identifier.
+ *
+ * @device: The IB device on which to create the IW CM identier.
+ * @event_handler: User callback invoked to report events associated with the
+ *   returned IW CM identifier.
+ * @context: User specified context associated with the id.
+ */
+struct iw_cm_id *iw_create_cm_id(struct ib_device *device,
+				 iw_cm_handler cm_handler, void *context);
+
+/**
+ * iw_destroy_cm_id - Destroy an IW CM identifier.
+ *
+ * @cm_id: The previously created IW CM identifier to destroy.
+ *
+ * The client can assume that no events will be delivered for the CM ID after
+ * this function returns.
+ */
+void iw_destroy_cm_id(struct iw_cm_id *cm_id);
+
+/**
+ * iw_cm_bind_qp - Unbind the specified IW CM identifier and QP
+ *
+ * @cm_id: The IW CM idenfier to unbind from the QP.
+ * @qp: The QP
+ *
+ * This is called by the provider when destroying the QP to ensure
+ * that any references held by the IWCM are released. It may also
+ * be called by the IWCM when destroying a CM_ID to that any
+ * references held by the provider are released.
+ */
+void iw_cm_unbind_qp(struct iw_cm_id *cm_id, struct ib_qp *qp);
+
+/**
+ * iw_cm_get_qp - Return the ib_qp associated with a QPN
+ *
+ * @ib_device: The IB device
+ * @qpn: The queue pair number
+ */
+struct ib_qp *iw_cm_get_qp(struct ib_device *device, int qpn);
+
+/**
+ * iw_cm_listen - Listen for incoming connection requests on the
+ * specified IW CM id.
+ *
+ * @cm_id: The IW CM identifier.
+ * @backlog: The maximum number of outstanding un-accepted inbound listen
+ *   requests to queue.
+ *
+ * The source address and port number are specified in the IW CM identifier
+ * structure.
+ */
+int iw_cm_listen(struct iw_cm_id *cm_id, int backlog);
+
+/**
+ * iw_cm_accept - Called to accept an incoming connect request.
+ *
+ * @cm_id: The IW CM identifier associated with the connection request.
+ * @iw_param: Pointer to a structure containing connection establishment
+ *   parameters.
+ *
+ * The specified cm_id will have been provided in the event data for a
+ * CONNECT_REQUEST event. Subsequent events related to this connection will be
+ * delivered to the specified IW CM identifier prior and may occur prior to
+ * the return of this function. If this function returns a non-zero value, the
+ * client can assume that no events will be delivered to the specified IW CM
+ * identifier.
+ */
+int iw_cm_accept(struct iw_cm_id *cm_id, struct iw_cm_conn_param *iw_param);
+
+/**
+ * iw_cm_reject - Reject an incoming connection request.
+ *
+ * @cm_id: Connection identifier associated with the request.
+ * @private_daa: Pointer to data to deliver to the remote peer as part of the
+ *   reject message.
+ * @private_data_len: The number of bytes in the private_data parameter.
+ *
+ * The client can assume that no events will be delivered to the specified IW
+ * CM identifier following the return of this function. The private_data
+ * buffer is available for reuse when this function returns.
+ */
+int iw_cm_reject(struct iw_cm_id *cm_id, const void *private_data,
+		 u8 private_data_len);
+
+/**
+ * iw_cm_connect - Called to request a connection to a remote peer.
+ *
+ * @cm_id: The IW CM identifier for the connection.
+ * @iw_param: Pointer to a structure containing connection  establishment
+ *   parameters.
+ *
+ * Events may be delivered to the specified IW CM identifier prior to the
+ * return of this function. If this function returns a non-zero value, the
+ * client can assume that no events will be delivered to the specified IW CM
+ * identifier.
+ */
+int iw_cm_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *iw_param);
+
+/**
+ * iw_cm_disconnect - Close the specified connection.
+ *
+ * @cm_id: The IW CM identifier to close.
+ * @abrupt: If 0, the connection will be closed gracefully, otherwise, the
+ *   connection will be reset.
+ *
+ * The IW CM identifier is still active until the IW_CM_EVENT_CLOSE event is
+ * delivered.
+ */
+int iw_cm_disconnect(struct iw_cm_id *cm_id, int abrupt);
+
+/**
+ * iw_cm_init_qp_attr - Called to initialize the attributes of the QP
+ * associated with a IW CM identifier.
+ *
+ * @cm_id: The IW CM identifier associated with the QP
+ * @qp_attr: Pointer to the QP attributes structure.
+ * @qp_attr_mask: Pointer to a bit vector specifying which QP attributes are
+ *   valid.
+ */
+int iw_cm_init_qp_attr(struct iw_cm_id *cm_id, struct ib_qp_attr *qp_attr,
+		       int *qp_attr_mask);
+
+#endif /* IW_CM_H */
-- 
cgit v1.2.3


From 07ebafbaaa72aa6a35472879008f5a1d1d469a0c Mon Sep 17 00:00:00 2001
From: Tom Tucker <tom@opengridcomputing.com>
Date: Thu, 3 Aug 2006 16:02:42 -0500
Subject: RDMA: iWARP Core Changes.

Modifications to the existing rdma header files, core files, drivers,
and ulp files to support iWARP, including:
 - Hook iWARP CM into the build system and use it in rdma_cm.
 - Convert enum ib_node_type to enum rdma_node_type, which includes
   the possibility of RDMA_NODE_RNIC, and update everything for this.

Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
---
 drivers/infiniband/core/Makefile             |   4 +-
 drivers/infiniband/core/addr.c               |  18 +-
 drivers/infiniband/core/cache.c              |   5 +-
 drivers/infiniband/core/cm.c                 |   3 +
 drivers/infiniband/core/cma.c                | 356 ++++++++++++++++++++++++---
 drivers/infiniband/core/device.c             |   4 +-
 drivers/infiniband/core/mad.c                |   7 +-
 drivers/infiniband/core/sa_query.c           |   5 +-
 drivers/infiniband/core/smi.c                |  16 +-
 drivers/infiniband/core/sysfs.c              |  11 +-
 drivers/infiniband/core/ucm.c                |   3 +-
 drivers/infiniband/core/user_mad.c           |   5 +-
 drivers/infiniband/core/verbs.c              |  17 ++
 drivers/infiniband/hw/ehca/ehca_main.c       |   2 +-
 drivers/infiniband/hw/ipath/ipath_verbs.c    |   2 +-
 drivers/infiniband/hw/mthca/mthca_provider.c |   2 +-
 drivers/infiniband/ulp/ipoib/ipoib_main.c    |   8 +-
 drivers/infiniband/ulp/srp/ib_srp.c          |   2 +-
 include/rdma/ib_addr.h                       |  17 +-
 include/rdma/ib_verbs.h                      |  25 +-
 20 files changed, 431 insertions(+), 81 deletions(-)

(limited to 'include')

diff --git a/drivers/infiniband/core/Makefile b/drivers/infiniband/core/Makefile
index 68e73ec2d1f8..163d991eb8c9 100644
--- a/drivers/infiniband/core/Makefile
+++ b/drivers/infiniband/core/Makefile
@@ -1,7 +1,7 @@
 infiniband-$(CONFIG_INFINIBAND_ADDR_TRANS)	:= ib_addr.o rdma_cm.o
 
 obj-$(CONFIG_INFINIBAND) +=		ib_core.o ib_mad.o ib_sa.o \
-					ib_cm.o $(infiniband-y)
+					ib_cm.o iw_cm.o $(infiniband-y)
 obj-$(CONFIG_INFINIBAND_USER_MAD) +=	ib_umad.o
 obj-$(CONFIG_INFINIBAND_USER_ACCESS) +=	ib_uverbs.o ib_ucm.o
 
@@ -14,6 +14,8 @@ ib_sa-y :=			sa_query.o
 
 ib_cm-y :=			cm.o
 
+iw_cm-y :=			iwcm.o
+
 rdma_cm-y :=			cma.o
 
 ib_addr-y :=			addr.o
diff --git a/drivers/infiniband/core/addr.c b/drivers/infiniband/core/addr.c
index d8e54e002ce3..9cbf09e2052f 100644
--- a/drivers/infiniband/core/addr.c
+++ b/drivers/infiniband/core/addr.c
@@ -61,12 +61,15 @@ static LIST_HEAD(req_list);
 static DECLARE_WORK(work, process_req, NULL);
 static struct workqueue_struct *addr_wq;
 
-static int copy_addr(struct rdma_dev_addr *dev_addr, struct net_device *dev,
-		     unsigned char *dst_dev_addr)
+int rdma_copy_addr(struct rdma_dev_addr *dev_addr, struct net_device *dev,
+		     const unsigned char *dst_dev_addr)
 {
 	switch (dev->type) {
 	case ARPHRD_INFINIBAND:
-		dev_addr->dev_type = IB_NODE_CA;
+		dev_addr->dev_type = RDMA_NODE_IB_CA;
+		break;
+	case ARPHRD_ETHER:
+		dev_addr->dev_type = RDMA_NODE_RNIC;
 		break;
 	default:
 		return -EADDRNOTAVAIL;
@@ -78,6 +81,7 @@ static int copy_addr(struct rdma_dev_addr *dev_addr, struct net_device *dev,
 		memcpy(dev_addr->dst_dev_addr, dst_dev_addr, MAX_ADDR_LEN);
 	return 0;
 }
+EXPORT_SYMBOL(rdma_copy_addr);
 
 int rdma_translate_ip(struct sockaddr *addr, struct rdma_dev_addr *dev_addr)
 {
@@ -89,7 +93,7 @@ int rdma_translate_ip(struct sockaddr *addr, struct rdma_dev_addr *dev_addr)
 	if (!dev)
 		return -EADDRNOTAVAIL;
 
-	ret = copy_addr(dev_addr, dev, NULL);
+	ret = rdma_copy_addr(dev_addr, dev, NULL);
 	dev_put(dev);
 	return ret;
 }
@@ -161,7 +165,7 @@ static int addr_resolve_remote(struct sockaddr_in *src_in,
 
 	/* If the device does ARP internally, return 'done' */
 	if (rt->idev->dev->flags & IFF_NOARP) {
-		copy_addr(addr, rt->idev->dev, NULL);
+		rdma_copy_addr(addr, rt->idev->dev, NULL);
 		goto put;
 	}
 
@@ -181,7 +185,7 @@ static int addr_resolve_remote(struct sockaddr_in *src_in,
 		src_in->sin_addr.s_addr = rt->rt_src;
 	}
 
-	ret = copy_addr(addr, neigh->dev, neigh->ha);
+	ret = rdma_copy_addr(addr, neigh->dev, neigh->ha);
 release:
 	neigh_release(neigh);
 put:
@@ -245,7 +249,7 @@ static int addr_resolve_local(struct sockaddr_in *src_in,
 	if (ZERONET(src_ip)) {
 		src_in->sin_family = dst_in->sin_family;
 		src_in->sin_addr.s_addr = dst_ip;
-		ret = copy_addr(addr, dev, dev->dev_addr);
+		ret = rdma_copy_addr(addr, dev, dev->dev_addr);
 	} else if (LOOPBACK(src_ip)) {
 		ret = rdma_translate_ip((struct sockaddr *)dst_in, addr);
 		if (!ret)
diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c
index 75313ade2e0d..20e9f64e67a6 100644
--- a/drivers/infiniband/core/cache.c
+++ b/drivers/infiniband/core/cache.c
@@ -62,12 +62,13 @@ struct ib_update_work {
 
 static inline int start_port(struct ib_device *device)
 {
-	return device->node_type == IB_NODE_SWITCH ? 0 : 1;
+	return (device->node_type == RDMA_NODE_IB_SWITCH) ? 0 : 1;
 }
 
 static inline int end_port(struct ib_device *device)
 {
-	return device->node_type == IB_NODE_SWITCH ? 0 : device->phys_port_cnt;
+	return (device->node_type == RDMA_NODE_IB_SWITCH) ?
+		0 : device->phys_port_cnt;
 }
 
 int ib_get_cached_gid(struct ib_device *device,
diff --git a/drivers/infiniband/core/cm.c b/drivers/infiniband/core/cm.c
index 1c145fe92a54..e130d2e89515 100644
--- a/drivers/infiniband/core/cm.c
+++ b/drivers/infiniband/core/cm.c
@@ -3280,6 +3280,9 @@ static void cm_add_one(struct ib_device *device)
 	int ret;
 	u8 i;
 
+	if (rdma_node_get_transport(device->node_type) != RDMA_TRANSPORT_IB)
+		return;
+
 	cm_dev = kmalloc(sizeof(*cm_dev) + sizeof(*port) *
 			 device->phys_port_cnt, GFP_KERNEL);
 	if (!cm_dev)
diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c
index 9d58bb59cd45..e88a7c652ca0 100644
--- a/drivers/infiniband/core/cma.c
+++ b/drivers/infiniband/core/cma.c
@@ -35,6 +35,7 @@
 #include <linux/mutex.h>
 #include <linux/random.h>
 #include <linux/idr.h>
+#include <linux/inetdevice.h>
 
 #include <net/tcp.h>
 
@@ -43,6 +44,7 @@
 #include <rdma/ib_cache.h>
 #include <rdma/ib_cm.h>
 #include <rdma/ib_sa.h>
+#include <rdma/iw_cm.h>
 
 MODULE_AUTHOR("Sean Hefty");
 MODULE_DESCRIPTION("Generic RDMA CM Agent");
@@ -124,6 +126,7 @@ struct rdma_id_private {
 	int			query_id;
 	union {
 		struct ib_cm_id	*ib;
+		struct iw_cm_id	*iw;
 	} cm_id;
 
 	u32			seq_num;
@@ -259,14 +262,23 @@ static void cma_detach_from_dev(struct rdma_id_private *id_priv)
 	id_priv->cma_dev = NULL;
 }
 
-static int cma_acquire_ib_dev(struct rdma_id_private *id_priv)
+static int cma_acquire_dev(struct rdma_id_private *id_priv)
 {
+	enum rdma_node_type dev_type = id_priv->id.route.addr.dev_addr.dev_type;
 	struct cma_device *cma_dev;
 	union ib_gid gid;
 	int ret = -ENODEV;
 
-	ib_addr_get_sgid(&id_priv->id.route.addr.dev_addr, &gid),
-
+	switch (rdma_node_get_transport(dev_type)) {
+	case RDMA_TRANSPORT_IB:
+		ib_addr_get_sgid(&id_priv->id.route.addr.dev_addr, &gid);
+		break;
+	case RDMA_TRANSPORT_IWARP:
+		iw_addr_get_sgid(&id_priv->id.route.addr.dev_addr, &gid);
+		break;
+	default:
+		return -ENODEV;
+	}
 	mutex_lock(&lock);
 	list_for_each_entry(cma_dev, &dev_list, list) {
 		ret = ib_find_cached_gid(cma_dev->device, &gid,
@@ -280,16 +292,6 @@ static int cma_acquire_ib_dev(struct rdma_id_private *id_priv)
 	return ret;
 }
 
-static int cma_acquire_dev(struct rdma_id_private *id_priv)
-{
-	switch (id_priv->id.route.addr.dev_addr.dev_type) {
-	case IB_NODE_CA:
-		return cma_acquire_ib_dev(id_priv);
-	default:
-		return -ENODEV;
-	}
-}
-
 static void cma_deref_id(struct rdma_id_private *id_priv)
 {
 	if (atomic_dec_and_test(&id_priv->refcount))
@@ -347,6 +349,16 @@ static int cma_init_ib_qp(struct rdma_id_private *id_priv, struct ib_qp *qp)
 					  IB_QP_PKEY_INDEX | IB_QP_PORT);
 }
 
+static int cma_init_iw_qp(struct rdma_id_private *id_priv, struct ib_qp *qp)
+{
+	struct ib_qp_attr qp_attr;
+
+	qp_attr.qp_state = IB_QPS_INIT;
+	qp_attr.qp_access_flags = IB_ACCESS_LOCAL_WRITE;
+
+	return ib_modify_qp(qp, &qp_attr, IB_QP_STATE | IB_QP_ACCESS_FLAGS);
+}
+
 int rdma_create_qp(struct rdma_cm_id *id, struct ib_pd *pd,
 		   struct ib_qp_init_attr *qp_init_attr)
 {
@@ -362,10 +374,13 @@ int rdma_create_qp(struct rdma_cm_id *id, struct ib_pd *pd,
 	if (IS_ERR(qp))
 		return PTR_ERR(qp);
 
-	switch (id->device->node_type) {
-	case IB_NODE_CA:
+	switch (rdma_node_get_transport(id->device->node_type)) {
+	case RDMA_TRANSPORT_IB:
 		ret = cma_init_ib_qp(id_priv, qp);
 		break;
+	case RDMA_TRANSPORT_IWARP:
+		ret = cma_init_iw_qp(id_priv, qp);
+		break;
 	default:
 		ret = -ENOSYS;
 		break;
@@ -451,13 +466,17 @@ int rdma_init_qp_attr(struct rdma_cm_id *id, struct ib_qp_attr *qp_attr,
 	int ret;
 
 	id_priv = container_of(id, struct rdma_id_private, id);
-	switch (id_priv->id.device->node_type) {
-	case IB_NODE_CA:
+	switch (rdma_node_get_transport(id_priv->id.device->node_type)) {
+	case RDMA_TRANSPORT_IB:
 		ret = ib_cm_init_qp_attr(id_priv->cm_id.ib, qp_attr,
 					 qp_attr_mask);
 		if (qp_attr->qp_state == IB_QPS_RTR)
 			qp_attr->rq_psn = id_priv->seq_num;
 		break;
+	case RDMA_TRANSPORT_IWARP:
+		ret = iw_cm_init_qp_attr(id_priv->cm_id.iw, qp_attr,
+					qp_attr_mask);
+		break;
 	default:
 		ret = -ENOSYS;
 		break;
@@ -590,8 +609,8 @@ static int cma_notify_user(struct rdma_id_private *id_priv,
 
 static void cma_cancel_route(struct rdma_id_private *id_priv)
 {
-	switch (id_priv->id.device->node_type) {
-	case IB_NODE_CA:
+	switch (rdma_node_get_transport(id_priv->id.device->node_type)) {
+	case RDMA_TRANSPORT_IB:
 		if (id_priv->query)
 			ib_sa_cancel_query(id_priv->query_id, id_priv->query);
 		break;
@@ -611,11 +630,15 @@ static void cma_destroy_listen(struct rdma_id_private *id_priv)
 	cma_exch(id_priv, CMA_DESTROYING);
 
 	if (id_priv->cma_dev) {
-		switch (id_priv->id.device->node_type) {
-		case IB_NODE_CA:
+		switch (rdma_node_get_transport(id_priv->id.device->node_type)) {
+		case RDMA_TRANSPORT_IB:
 			if (id_priv->cm_id.ib && !IS_ERR(id_priv->cm_id.ib))
 				ib_destroy_cm_id(id_priv->cm_id.ib);
 			break;
+		case RDMA_TRANSPORT_IWARP:
+			if (id_priv->cm_id.iw && !IS_ERR(id_priv->cm_id.iw))
+				iw_destroy_cm_id(id_priv->cm_id.iw);
+			break;
 		default:
 			break;
 		}
@@ -690,11 +713,15 @@ void rdma_destroy_id(struct rdma_cm_id *id)
 	cma_cancel_operation(id_priv, state);
 
 	if (id_priv->cma_dev) {
-		switch (id->device->node_type) {
-		case IB_NODE_CA:
+		switch (rdma_node_get_transport(id->device->node_type)) {
+		case RDMA_TRANSPORT_IB:
 			if (id_priv->cm_id.ib && !IS_ERR(id_priv->cm_id.ib))
 				ib_destroy_cm_id(id_priv->cm_id.ib);
 			break;
+		case RDMA_TRANSPORT_IWARP:
+			if (id_priv->cm_id.iw && !IS_ERR(id_priv->cm_id.iw))
+				iw_destroy_cm_id(id_priv->cm_id.iw);
+			break;
 		default:
 			break;
 		}
@@ -869,7 +896,7 @@ static struct rdma_id_private *cma_new_id(struct rdma_cm_id *listen_id,
 	ib_addr_set_sgid(&rt->addr.dev_addr, &rt->path_rec[0].sgid);
 	ib_addr_set_dgid(&rt->addr.dev_addr, &rt->path_rec[0].dgid);
 	ib_addr_set_pkey(&rt->addr.dev_addr, be16_to_cpu(rt->path_rec[0].pkey));
-	rt->addr.dev_addr.dev_type = IB_NODE_CA;
+	rt->addr.dev_addr.dev_type = RDMA_NODE_IB_CA;
 
 	id_priv = container_of(id, struct rdma_id_private, id);
 	id_priv->state = CMA_CONNECT;
@@ -898,7 +925,7 @@ static int cma_req_handler(struct ib_cm_id *cm_id, struct ib_cm_event *ib_event)
 	}
 
 	atomic_inc(&conn_id->dev_remove);
-	ret = cma_acquire_ib_dev(conn_id);
+	ret = cma_acquire_dev(conn_id);
 	if (ret) {
 		ret = -ENODEV;
 		cma_release_remove(conn_id);
@@ -982,6 +1009,128 @@ static void cma_set_compare_data(enum rdma_port_space ps, struct sockaddr *addr,
 	}
 }
 
+static int cma_iw_handler(struct iw_cm_id *iw_id, struct iw_cm_event *iw_event)
+{
+	struct rdma_id_private *id_priv = iw_id->context;
+	enum rdma_cm_event_type event = 0;
+	struct sockaddr_in *sin;
+	int ret = 0;
+
+	atomic_inc(&id_priv->dev_remove);
+
+	switch (iw_event->event) {
+	case IW_CM_EVENT_CLOSE:
+		event = RDMA_CM_EVENT_DISCONNECTED;
+		break;
+	case IW_CM_EVENT_CONNECT_REPLY:
+		sin = (struct sockaddr_in *) &id_priv->id.route.addr.src_addr;
+		*sin = iw_event->local_addr;
+		sin = (struct sockaddr_in *) &id_priv->id.route.addr.dst_addr;
+		*sin = iw_event->remote_addr;
+		if (iw_event->status)
+			event = RDMA_CM_EVENT_REJECTED;
+		else
+			event = RDMA_CM_EVENT_ESTABLISHED;
+		break;
+	case IW_CM_EVENT_ESTABLISHED:
+		event = RDMA_CM_EVENT_ESTABLISHED;
+		break;
+	default:
+		BUG_ON(1);
+	}
+
+	ret = cma_notify_user(id_priv, event, iw_event->status,
+			      iw_event->private_data,
+			      iw_event->private_data_len);
+	if (ret) {
+		/* Destroy the CM ID by returning a non-zero value. */
+		id_priv->cm_id.iw = NULL;
+		cma_exch(id_priv, CMA_DESTROYING);
+		cma_release_remove(id_priv);
+		rdma_destroy_id(&id_priv->id);
+		return ret;
+	}
+
+	cma_release_remove(id_priv);
+	return ret;
+}
+
+static int iw_conn_req_handler(struct iw_cm_id *cm_id,
+			       struct iw_cm_event *iw_event)
+{
+	struct rdma_cm_id *new_cm_id;
+	struct rdma_id_private *listen_id, *conn_id;
+	struct sockaddr_in *sin;
+	struct net_device *dev = NULL;
+	int ret;
+
+	listen_id = cm_id->context;
+	atomic_inc(&listen_id->dev_remove);
+	if (!cma_comp(listen_id, CMA_LISTEN)) {
+		ret = -ECONNABORTED;
+		goto out;
+	}
+
+	/* Create a new RDMA id for the new IW CM ID */
+	new_cm_id = rdma_create_id(listen_id->id.event_handler,
+				   listen_id->id.context,
+				   RDMA_PS_TCP);
+	if (!new_cm_id) {
+		ret = -ENOMEM;
+		goto out;
+	}
+	conn_id = container_of(new_cm_id, struct rdma_id_private, id);
+	atomic_inc(&conn_id->dev_remove);
+	conn_id->state = CMA_CONNECT;
+
+	dev = ip_dev_find(iw_event->local_addr.sin_addr.s_addr);
+	if (!dev) {
+		ret = -EADDRNOTAVAIL;
+		cma_release_remove(conn_id);
+		rdma_destroy_id(new_cm_id);
+		goto out;
+	}
+	ret = rdma_copy_addr(&conn_id->id.route.addr.dev_addr, dev, NULL);
+	if (ret) {
+		cma_release_remove(conn_id);
+		rdma_destroy_id(new_cm_id);
+		goto out;
+	}
+
+	ret = cma_acquire_dev(conn_id);
+	if (ret) {
+		cma_release_remove(conn_id);
+		rdma_destroy_id(new_cm_id);
+		goto out;
+	}
+
+	conn_id->cm_id.iw = cm_id;
+	cm_id->context = conn_id;
+	cm_id->cm_handler = cma_iw_handler;
+
+	sin = (struct sockaddr_in *) &new_cm_id->route.addr.src_addr;
+	*sin = iw_event->local_addr;
+	sin = (struct sockaddr_in *) &new_cm_id->route.addr.dst_addr;
+	*sin = iw_event->remote_addr;
+
+	ret = cma_notify_user(conn_id, RDMA_CM_EVENT_CONNECT_REQUEST, 0,
+			      iw_event->private_data,
+			      iw_event->private_data_len);
+	if (ret) {
+		/* User wants to destroy the CM ID */
+		conn_id->cm_id.iw = NULL;
+		cma_exch(conn_id, CMA_DESTROYING);
+		cma_release_remove(conn_id);
+		rdma_destroy_id(&conn_id->id);
+	}
+
+out:
+	if (dev)
+		dev_put(dev);
+	cma_release_remove(listen_id);
+	return ret;
+}
+
 static int cma_ib_listen(struct rdma_id_private *id_priv)
 {
 	struct ib_cm_compare_data compare_data;
@@ -1011,6 +1160,30 @@ static int cma_ib_listen(struct rdma_id_private *id_priv)
 	return ret;
 }
 
+static int cma_iw_listen(struct rdma_id_private *id_priv, int backlog)
+{
+	int ret;
+	struct sockaddr_in *sin;
+
+	id_priv->cm_id.iw = iw_create_cm_id(id_priv->id.device,
+					    iw_conn_req_handler,
+					    id_priv);
+	if (IS_ERR(id_priv->cm_id.iw))
+		return PTR_ERR(id_priv->cm_id.iw);
+
+	sin = (struct sockaddr_in *) &id_priv->id.route.addr.src_addr;
+	id_priv->cm_id.iw->local_addr = *sin;
+
+	ret = iw_cm_listen(id_priv->cm_id.iw, backlog);
+
+	if (ret) {
+		iw_destroy_cm_id(id_priv->cm_id.iw);
+		id_priv->cm_id.iw = NULL;
+	}
+
+	return ret;
+}
+
 static int cma_listen_handler(struct rdma_cm_id *id,
 			      struct rdma_cm_event *event)
 {
@@ -1087,12 +1260,17 @@ int rdma_listen(struct rdma_cm_id *id, int backlog)
 
 	id_priv->backlog = backlog;
 	if (id->device) {
-		switch (id->device->node_type) {
-		case IB_NODE_CA:
+		switch (rdma_node_get_transport(id->device->node_type)) {
+		case RDMA_TRANSPORT_IB:
 			ret = cma_ib_listen(id_priv);
 			if (ret)
 				goto err;
 			break;
+		case RDMA_TRANSPORT_IWARP:
+			ret = cma_iw_listen(id_priv, backlog);
+			if (ret)
+				goto err;
+			break;
 		default:
 			ret = -ENOSYS;
 			goto err;
@@ -1231,6 +1409,23 @@ err:
 }
 EXPORT_SYMBOL(rdma_set_ib_paths);
 
+static int cma_resolve_iw_route(struct rdma_id_private *id_priv, int timeout_ms)
+{
+	struct cma_work *work;
+
+	work = kzalloc(sizeof *work, GFP_KERNEL);
+	if (!work)
+		return -ENOMEM;
+
+	work->id = id_priv;
+	INIT_WORK(&work->work, cma_work_handler, work);
+	work->old_state = CMA_ROUTE_QUERY;
+	work->new_state = CMA_ROUTE_RESOLVED;
+	work->event.event = RDMA_CM_EVENT_ROUTE_RESOLVED;
+	queue_work(cma_wq, &work->work);
+	return 0;
+}
+
 int rdma_resolve_route(struct rdma_cm_id *id, int timeout_ms)
 {
 	struct rdma_id_private *id_priv;
@@ -1241,10 +1436,13 @@ int rdma_resolve_route(struct rdma_cm_id *id, int timeout_ms)
 		return -EINVAL;
 
 	atomic_inc(&id_priv->refcount);
-	switch (id->device->node_type) {
-	case IB_NODE_CA:
+	switch (rdma_node_get_transport(id->device->node_type)) {
+	case RDMA_TRANSPORT_IB:
 		ret = cma_resolve_ib_route(id_priv, timeout_ms);
 		break;
+	case RDMA_TRANSPORT_IWARP:
+		ret = cma_resolve_iw_route(id_priv, timeout_ms);
+		break;
 	default:
 		ret = -ENOSYS;
 		break;
@@ -1649,6 +1847,47 @@ out:
 	return ret;
 }
 
+static int cma_connect_iw(struct rdma_id_private *id_priv,
+			  struct rdma_conn_param *conn_param)
+{
+	struct iw_cm_id *cm_id;
+	struct sockaddr_in* sin;
+	int ret;
+	struct iw_cm_conn_param iw_param;
+
+	cm_id = iw_create_cm_id(id_priv->id.device, cma_iw_handler, id_priv);
+	if (IS_ERR(cm_id)) {
+		ret = PTR_ERR(cm_id);
+		goto out;
+	}
+
+	id_priv->cm_id.iw = cm_id;
+
+	sin = (struct sockaddr_in*) &id_priv->id.route.addr.src_addr;
+	cm_id->local_addr = *sin;
+
+	sin = (struct sockaddr_in*) &id_priv->id.route.addr.dst_addr;
+	cm_id->remote_addr = *sin;
+
+	ret = cma_modify_qp_rtr(&id_priv->id);
+	if (ret) {
+		iw_destroy_cm_id(cm_id);
+		return ret;
+	}
+
+	iw_param.ord = conn_param->initiator_depth;
+	iw_param.ird = conn_param->responder_resources;
+	iw_param.private_data = conn_param->private_data;
+	iw_param.private_data_len = conn_param->private_data_len;
+	if (id_priv->id.qp)
+		iw_param.qpn = id_priv->qp_num;
+	else
+		iw_param.qpn = conn_param->qp_num;
+	ret = iw_cm_connect(cm_id, &iw_param);
+out:
+	return ret;
+}
+
 int rdma_connect(struct rdma_cm_id *id, struct rdma_conn_param *conn_param)
 {
 	struct rdma_id_private *id_priv;
@@ -1664,10 +1903,13 @@ int rdma_connect(struct rdma_cm_id *id, struct rdma_conn_param *conn_param)
 		id_priv->srq = conn_param->srq;
 	}
 
-	switch (id->device->node_type) {
-	case IB_NODE_CA:
+	switch (rdma_node_get_transport(id->device->node_type)) {
+	case RDMA_TRANSPORT_IB:
 		ret = cma_connect_ib(id_priv, conn_param);
 		break;
+	case RDMA_TRANSPORT_IWARP:
+		ret = cma_connect_iw(id_priv, conn_param);
+		break;
 	default:
 		ret = -ENOSYS;
 		break;
@@ -1708,6 +1950,28 @@ static int cma_accept_ib(struct rdma_id_private *id_priv,
 	return ib_send_cm_rep(id_priv->cm_id.ib, &rep);
 }
 
+static int cma_accept_iw(struct rdma_id_private *id_priv,
+		  struct rdma_conn_param *conn_param)
+{
+	struct iw_cm_conn_param iw_param;
+	int ret;
+
+	ret = cma_modify_qp_rtr(&id_priv->id);
+	if (ret)
+		return ret;
+
+	iw_param.ord = conn_param->initiator_depth;
+	iw_param.ird = conn_param->responder_resources;
+	iw_param.private_data = conn_param->private_data;
+	iw_param.private_data_len = conn_param->private_data_len;
+	if (id_priv->id.qp) {
+		iw_param.qpn = id_priv->qp_num;
+	} else
+		iw_param.qpn = conn_param->qp_num;
+
+	return iw_cm_accept(id_priv->cm_id.iw, &iw_param);
+}
+
 int rdma_accept(struct rdma_cm_id *id, struct rdma_conn_param *conn_param)
 {
 	struct rdma_id_private *id_priv;
@@ -1723,13 +1987,16 @@ int rdma_accept(struct rdma_cm_id *id, struct rdma_conn_param *conn_param)
 		id_priv->srq = conn_param->srq;
 	}
 
-	switch (id->device->node_type) {
-	case IB_NODE_CA:
+	switch (rdma_node_get_transport(id->device->node_type)) {
+	case RDMA_TRANSPORT_IB:
 		if (conn_param)
 			ret = cma_accept_ib(id_priv, conn_param);
 		else
 			ret = cma_rep_recv(id_priv);
 		break;
+	case RDMA_TRANSPORT_IWARP:
+		ret = cma_accept_iw(id_priv, conn_param);
+		break;
 	default:
 		ret = -ENOSYS;
 		break;
@@ -1756,12 +2023,16 @@ int rdma_reject(struct rdma_cm_id *id, const void *private_data,
 	if (!cma_comp(id_priv, CMA_CONNECT))
 		return -EINVAL;
 
-	switch (id->device->node_type) {
-	case IB_NODE_CA:
+	switch (rdma_node_get_transport(id->device->node_type)) {
+	case RDMA_TRANSPORT_IB:
 		ret = ib_send_cm_rej(id_priv->cm_id.ib,
 				     IB_CM_REJ_CONSUMER_DEFINED, NULL, 0,
 				     private_data, private_data_len);
 		break;
+	case RDMA_TRANSPORT_IWARP:
+		ret = iw_cm_reject(id_priv->cm_id.iw,
+				   private_data, private_data_len);
+		break;
 	default:
 		ret = -ENOSYS;
 		break;
@@ -1780,17 +2051,20 @@ int rdma_disconnect(struct rdma_cm_id *id)
 	    !cma_comp(id_priv, CMA_DISCONNECT))
 		return -EINVAL;
 
-	ret = cma_modify_qp_err(id);
-	if (ret)
-		goto out;
-
-	switch (id->device->node_type) {
-	case IB_NODE_CA:
+	switch (rdma_node_get_transport(id->device->node_type)) {
+	case RDMA_TRANSPORT_IB:
+		ret = cma_modify_qp_err(id);
+		if (ret)
+			goto out;
 		/* Initiate or respond to a disconnect. */
 		if (ib_send_cm_dreq(id_priv->cm_id.ib, NULL, 0))
 			ib_send_cm_drep(id_priv->cm_id.ib, NULL, 0);
 		break;
+	case RDMA_TRANSPORT_IWARP:
+		ret = iw_cm_disconnect(id_priv->cm_id.iw, 0);
+		break;
 	default:
+		ret = -EINVAL;
 		break;
 	}
 out:
diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c
index b2f3cb91d9bc..d978fbe97535 100644
--- a/drivers/infiniband/core/device.c
+++ b/drivers/infiniband/core/device.c
@@ -505,7 +505,7 @@ int ib_query_port(struct ib_device *device,
 		  u8 port_num,
 		  struct ib_port_attr *port_attr)
 {
-	if (device->node_type == IB_NODE_SWITCH) {
+	if (device->node_type == RDMA_NODE_IB_SWITCH) {
 		if (port_num)
 			return -EINVAL;
 	} else if (port_num < 1 || port_num > device->phys_port_cnt)
@@ -580,7 +580,7 @@ int ib_modify_port(struct ib_device *device,
 		   u8 port_num, int port_modify_mask,
 		   struct ib_port_modify *port_modify)
 {
-	if (device->node_type == IB_NODE_SWITCH) {
+	if (device->node_type == RDMA_NODE_IB_SWITCH) {
 		if (port_num)
 			return -EINVAL;
 	} else if (port_num < 1 || port_num > device->phys_port_cnt)
diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c
index 32d3028b274b..082f03c158f0 100644
--- a/drivers/infiniband/core/mad.c
+++ b/drivers/infiniband/core/mad.c
@@ -2876,7 +2876,10 @@ static void ib_mad_init_device(struct ib_device *device)
 {
 	int start, end, i;
 
-	if (device->node_type == IB_NODE_SWITCH) {
+	if (rdma_node_get_transport(device->node_type) != RDMA_TRANSPORT_IB)
+		return;
+
+	if (device->node_type == RDMA_NODE_IB_SWITCH) {
 		start = 0;
 		end   = 0;
 	} else {
@@ -2923,7 +2926,7 @@ static void ib_mad_remove_device(struct ib_device *device)
 {
 	int i, num_ports, cur_port;
 
-	if (device->node_type == IB_NODE_SWITCH) {
+	if (device->node_type == RDMA_NODE_IB_SWITCH) {
 		num_ports = 1;
 		cur_port = 0;
 	} else {
diff --git a/drivers/infiniband/core/sa_query.c b/drivers/infiniband/core/sa_query.c
index df762ba4868f..ca8760a7d88c 100644
--- a/drivers/infiniband/core/sa_query.c
+++ b/drivers/infiniband/core/sa_query.c
@@ -919,7 +919,10 @@ static void ib_sa_add_one(struct ib_device *device)
 	struct ib_sa_device *sa_dev;
 	int s, e, i;
 
-	if (device->node_type == IB_NODE_SWITCH)
+	if (rdma_node_get_transport(device->node_type) != RDMA_TRANSPORT_IB)
+		return;
+
+	if (device->node_type == RDMA_NODE_IB_SWITCH)
 		s = e = 0;
 	else {
 		s = 1;
diff --git a/drivers/infiniband/core/smi.c b/drivers/infiniband/core/smi.c
index 35852e794e26..54b81e17ad50 100644
--- a/drivers/infiniband/core/smi.c
+++ b/drivers/infiniband/core/smi.c
@@ -64,7 +64,7 @@ int smi_handle_dr_smp_send(struct ib_smp *smp,
 
 		/* C14-9:2 */
 		if (hop_ptr && hop_ptr < hop_cnt) {
-			if (node_type != IB_NODE_SWITCH)
+			if (node_type != RDMA_NODE_IB_SWITCH)
 				return 0;
 
 			/* smp->return_path set when received */
@@ -77,7 +77,7 @@ int smi_handle_dr_smp_send(struct ib_smp *smp,
 		if (hop_ptr == hop_cnt) {
 			/* smp->return_path set when received */
 			smp->hop_ptr++;
-			return (node_type == IB_NODE_SWITCH ||
+			return (node_type == RDMA_NODE_IB_SWITCH ||
 				smp->dr_dlid == IB_LID_PERMISSIVE);
 		}
 
@@ -95,7 +95,7 @@ int smi_handle_dr_smp_send(struct ib_smp *smp,
 
 		/* C14-13:2 */
 		if (2 <= hop_ptr && hop_ptr <= hop_cnt) {
-			if (node_type != IB_NODE_SWITCH)
+			if (node_type != RDMA_NODE_IB_SWITCH)
 				return 0;
 
 			smp->hop_ptr--;
@@ -107,7 +107,7 @@ int smi_handle_dr_smp_send(struct ib_smp *smp,
 		if (hop_ptr == 1) {
 			smp->hop_ptr--;
 			/* C14-13:3 -- SMPs destined for SM shouldn't be here */
-			return (node_type == IB_NODE_SWITCH ||
+			return (node_type == RDMA_NODE_IB_SWITCH ||
 				smp->dr_slid == IB_LID_PERMISSIVE);
 		}
 
@@ -142,7 +142,7 @@ int smi_handle_dr_smp_recv(struct ib_smp *smp,
 
 		/* C14-9:2 -- intermediate hop */
 		if (hop_ptr && hop_ptr < hop_cnt) {
-			if (node_type != IB_NODE_SWITCH)
+			if (node_type != RDMA_NODE_IB_SWITCH)
 				return 0;
 
 			smp->return_path[hop_ptr] = port_num;
@@ -156,7 +156,7 @@ int smi_handle_dr_smp_recv(struct ib_smp *smp,
 				smp->return_path[hop_ptr] = port_num;
 			/* smp->hop_ptr updated when sending */
 
-			return (node_type == IB_NODE_SWITCH ||
+			return (node_type == RDMA_NODE_IB_SWITCH ||
 				smp->dr_dlid == IB_LID_PERMISSIVE);
 		}
 
@@ -175,7 +175,7 @@ int smi_handle_dr_smp_recv(struct ib_smp *smp,
 
 		/* C14-13:2 */
 		if (2 <= hop_ptr && hop_ptr <= hop_cnt) {
-			if (node_type != IB_NODE_SWITCH)
+			if (node_type != RDMA_NODE_IB_SWITCH)
 				return 0;
 
 			/* smp->hop_ptr updated when sending */
@@ -190,7 +190,7 @@ int smi_handle_dr_smp_recv(struct ib_smp *smp,
 				return 1;
 			}
 			/* smp->hop_ptr updated when sending */
-			return (node_type == IB_NODE_SWITCH);
+			return (node_type == RDMA_NODE_IB_SWITCH);
 		}
 
 		/* C14-13:4 -- hop_ptr = 0 -> give to SM */
diff --git a/drivers/infiniband/core/sysfs.c b/drivers/infiniband/core/sysfs.c
index fb6660564a30..709323c14c5d 100644
--- a/drivers/infiniband/core/sysfs.c
+++ b/drivers/infiniband/core/sysfs.c
@@ -589,10 +589,11 @@ static ssize_t show_node_type(struct class_device *cdev, char *buf)
 		return -ENODEV;
 
 	switch (dev->node_type) {
-	case IB_NODE_CA:     return sprintf(buf, "%d: CA\n", dev->node_type);
-	case IB_NODE_SWITCH: return sprintf(buf, "%d: switch\n", dev->node_type);
-	case IB_NODE_ROUTER: return sprintf(buf, "%d: router\n", dev->node_type);
-	default:             return sprintf(buf, "%d: <unknown>\n", dev->node_type);
+	case RDMA_NODE_IB_CA:	  return sprintf(buf, "%d: CA\n", dev->node_type);
+	case RDMA_NODE_RNIC:	  return sprintf(buf, "%d: RNIC\n", dev->node_type);
+	case RDMA_NODE_IB_SWITCH: return sprintf(buf, "%d: switch\n", dev->node_type);
+	case RDMA_NODE_IB_ROUTER: return sprintf(buf, "%d: router\n", dev->node_type);
+	default:		  return sprintf(buf, "%d: <unknown>\n", dev->node_type);
 	}
 }
 
@@ -708,7 +709,7 @@ int ib_device_register_sysfs(struct ib_device *device)
 	if (ret)
 		goto err_put;
 
-	if (device->node_type == IB_NODE_SWITCH) {
+	if (device->node_type == RDMA_NODE_IB_SWITCH) {
 		ret = add_port(device, 0);
 		if (ret)
 			goto err_put;
diff --git a/drivers/infiniband/core/ucm.c b/drivers/infiniband/core/ucm.c
index e74c964af7fa..ad4f4d5c2924 100644
--- a/drivers/infiniband/core/ucm.c
+++ b/drivers/infiniband/core/ucm.c
@@ -1247,7 +1247,8 @@ static void ib_ucm_add_one(struct ib_device *device)
 {
 	struct ib_ucm_device *ucm_dev;
 
-	if (!device->alloc_ucontext)
+	if (!device->alloc_ucontext ||
+	    rdma_node_get_transport(device->node_type) != RDMA_TRANSPORT_IB)
 		return;
 
 	ucm_dev = kzalloc(sizeof *ucm_dev, GFP_KERNEL);
diff --git a/drivers/infiniband/core/user_mad.c b/drivers/infiniband/core/user_mad.c
index 8a455aec758f..807fbd6b8414 100644
--- a/drivers/infiniband/core/user_mad.c
+++ b/drivers/infiniband/core/user_mad.c
@@ -1032,7 +1032,10 @@ static void ib_umad_add_one(struct ib_device *device)
 	struct ib_umad_device *umad_dev;
 	int s, e, i;
 
-	if (device->node_type == IB_NODE_SWITCH)
+	if (rdma_node_get_transport(device->node_type) != RDMA_TRANSPORT_IB)
+		return;
+
+	if (device->node_type == RDMA_NODE_IB_SWITCH)
 		s = e = 0;
 	else {
 		s = 1;
diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index 06f98e9e14f9..8b5dd3649bbf 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -79,6 +79,23 @@ enum ib_rate mult_to_ib_rate(int mult)
 }
 EXPORT_SYMBOL(mult_to_ib_rate);
 
+enum rdma_transport_type
+rdma_node_get_transport(enum rdma_node_type node_type)
+{
+	switch (node_type) {
+	case RDMA_NODE_IB_CA:
+	case RDMA_NODE_IB_SWITCH:
+	case RDMA_NODE_IB_ROUTER:
+		return RDMA_TRANSPORT_IB;
+	case RDMA_NODE_RNIC:
+		return RDMA_TRANSPORT_IWARP;
+	default:
+		BUG();
+		return 0;
+	}
+}
+EXPORT_SYMBOL(rdma_node_get_transport);
+
 /* Protection domains */
 
 struct ib_pd *ib_alloc_pd(struct ib_device *device)
diff --git a/drivers/infiniband/hw/ehca/ehca_main.c b/drivers/infiniband/hw/ehca/ehca_main.c
index 2a99f2d13cdb..2380994418a5 100644
--- a/drivers/infiniband/hw/ehca/ehca_main.c
+++ b/drivers/infiniband/hw/ehca/ehca_main.c
@@ -269,7 +269,7 @@ int ehca_register_device(struct ehca_shca *shca)
 		(1ull << IB_USER_VERBS_CMD_ATTACH_MCAST)	|
 		(1ull << IB_USER_VERBS_CMD_DETACH_MCAST);
 
-	shca->ib_device.node_type           = IB_NODE_CA;
+	shca->ib_device.node_type           = RDMA_NODE_IB_CA;
 	shca->ib_device.phys_port_cnt       = shca->num_ports;
 	shca->ib_device.dma_device          = &shca->ibmebus_dev->ofdev.dev;
 	shca->ib_device.query_device        = ehca_query_device;
diff --git a/drivers/infiniband/hw/ipath/ipath_verbs.c b/drivers/infiniband/hw/ipath/ipath_verbs.c
index fbda7739715f..b8381c5e72bd 100644
--- a/drivers/infiniband/hw/ipath/ipath_verbs.c
+++ b/drivers/infiniband/hw/ipath/ipath_verbs.c
@@ -1538,7 +1538,7 @@ int ipath_register_ib_device(struct ipath_devdata *dd)
 		(1ull << IB_USER_VERBS_CMD_QUERY_SRQ)		|
 		(1ull << IB_USER_VERBS_CMD_DESTROY_SRQ)		|
 		(1ull << IB_USER_VERBS_CMD_POST_SRQ_RECV);
-	dev->node_type = IB_NODE_CA;
+	dev->node_type = RDMA_NODE_IB_CA;
 	dev->phys_port_cnt = 1;
 	dev->dma_device = &dd->pcidev->dev;
 	dev->class_dev.dev = dev->dma_device;
diff --git a/drivers/infiniband/hw/mthca/mthca_provider.c b/drivers/infiniband/hw/mthca/mthca_provider.c
index 265b1d1c4a62..981fe2eebdfa 100644
--- a/drivers/infiniband/hw/mthca/mthca_provider.c
+++ b/drivers/infiniband/hw/mthca/mthca_provider.c
@@ -1288,7 +1288,7 @@ int mthca_register_device(struct mthca_dev *dev)
 		(1ull << IB_USER_VERBS_CMD_DESTROY_QP)		|
 		(1ull << IB_USER_VERBS_CMD_ATTACH_MCAST)	|
 		(1ull << IB_USER_VERBS_CMD_DETACH_MCAST);
-	dev->ib_dev.node_type            = IB_NODE_CA;
+	dev->ib_dev.node_type            = RDMA_NODE_IB_CA;
 	dev->ib_dev.phys_port_cnt        = dev->limits.num_ports;
 	dev->ib_dev.dma_device           = &dev->pdev->dev;
 	dev->ib_dev.class_dev.dev        = &dev->pdev->dev;
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c
index 36d76987a481..e9a7659eb1d7 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_main.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c
@@ -1111,13 +1111,16 @@ static void ipoib_add_one(struct ib_device *device)
 	struct ipoib_dev_priv *priv;
 	int s, e, p;
 
+	if (rdma_node_get_transport(device->node_type) != RDMA_TRANSPORT_IB)
+		return;
+
 	dev_list = kmalloc(sizeof *dev_list, GFP_KERNEL);
 	if (!dev_list)
 		return;
 
 	INIT_LIST_HEAD(dev_list);
 
-	if (device->node_type == IB_NODE_SWITCH) {
+	if (device->node_type == RDMA_NODE_IB_SWITCH) {
 		s = 0;
 		e = 0;
 	} else {
@@ -1141,6 +1144,9 @@ static void ipoib_remove_one(struct ib_device *device)
 	struct ipoib_dev_priv *priv, *tmp;
 	struct list_head *dev_list;
 
+	if (rdma_node_get_transport(device->node_type) != RDMA_TRANSPORT_IB)
+		return;
+
 	dev_list = ib_get_client_data(device, &ipoib_client);
 
 	list_for_each_entry_safe(priv, tmp, dev_list, list) {
diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
index 61c13d1e0506..feb1fcd0f2fb 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.c
+++ b/drivers/infiniband/ulp/srp/ib_srp.c
@@ -1899,7 +1899,7 @@ static void srp_add_one(struct ib_device *device)
 	if (IS_ERR(srp_dev->fmr_pool))
 		srp_dev->fmr_pool = NULL;
 
-	if (device->node_type == IB_NODE_SWITCH) {
+	if (device->node_type == RDMA_NODE_IB_SWITCH) {
 		s = 0;
 		e = 0;
 	} else {
diff --git a/include/rdma/ib_addr.h b/include/rdma/ib_addr.h
index 0ff67398928d..81b62307621d 100644
--- a/include/rdma/ib_addr.h
+++ b/include/rdma/ib_addr.h
@@ -40,7 +40,7 @@ struct rdma_dev_addr {
 	unsigned char src_dev_addr[MAX_ADDR_LEN];
 	unsigned char dst_dev_addr[MAX_ADDR_LEN];
 	unsigned char broadcast[MAX_ADDR_LEN];
-	enum ib_node_type dev_type;
+	enum rdma_node_type dev_type;
 };
 
 /**
@@ -72,6 +72,9 @@ int rdma_resolve_ip(struct sockaddr *src_addr, struct sockaddr *dst_addr,
 
 void rdma_addr_cancel(struct rdma_dev_addr *addr);
 
+int rdma_copy_addr(struct rdma_dev_addr *dev_addr, struct net_device *dev,
+	      const unsigned char *dst_dev_addr);
+
 static inline int ip_addr_size(struct sockaddr *addr)
 {
 	return addr->sa_family == AF_INET6 ?
@@ -113,4 +116,16 @@ static inline void ib_addr_set_dgid(struct rdma_dev_addr *dev_addr,
 	memcpy(dev_addr->dst_dev_addr + 4, gid, sizeof *gid);
 }
 
+static inline void iw_addr_get_sgid(struct rdma_dev_addr *dev_addr,
+				    union ib_gid *gid)
+{
+	memcpy(gid, dev_addr->src_dev_addr, sizeof *gid);
+}
+
+static inline void iw_addr_get_dgid(struct rdma_dev_addr *dev_addr,
+				    union ib_gid *gid)
+{
+	memcpy(gid, dev_addr->dst_dev_addr, sizeof *gid);
+}
+
 #endif /* IB_ADDR_H */
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index 61eed3996117..8eacc3510993 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -56,12 +56,22 @@ union ib_gid {
 	} global;
 };
 
-enum ib_node_type {
-	IB_NODE_CA 	= 1,
-	IB_NODE_SWITCH,
-	IB_NODE_ROUTER
+enum rdma_node_type {
+	/* IB values map to NodeInfo:NodeType. */
+	RDMA_NODE_IB_CA 	= 1,
+	RDMA_NODE_IB_SWITCH,
+	RDMA_NODE_IB_ROUTER,
+	RDMA_NODE_RNIC
 };
 
+enum rdma_transport_type {
+	RDMA_TRANSPORT_IB,
+	RDMA_TRANSPORT_IWARP
+};
+
+enum rdma_transport_type
+rdma_node_get_transport(enum rdma_node_type node_type) __attribute_const__;
+
 enum ib_device_cap_flags {
 	IB_DEVICE_RESIZE_MAX_WR		= 1,
 	IB_DEVICE_BAD_PKEY_CNTR		= (1<<1),
@@ -78,6 +88,9 @@ enum ib_device_cap_flags {
 	IB_DEVICE_RC_RNR_NAK_GEN	= (1<<12),
 	IB_DEVICE_SRQ_RESIZE		= (1<<13),
 	IB_DEVICE_N_NOTIFY_CQ		= (1<<14),
+	IB_DEVICE_ZERO_STAG		= (1<<15),
+	IB_DEVICE_SEND_W_INV		= (1<<16),
+	IB_DEVICE_MEM_WINDOW		= (1<<17)
 };
 
 enum ib_atomic_cap {
@@ -835,6 +848,8 @@ struct ib_cache {
 	u8                     *lmc_cache;
 };
 
+struct iw_cm_verbs;
+
 struct ib_device {
 	struct device                *dma_device;
 
@@ -851,6 +866,8 @@ struct ib_device {
 
 	u32                           flags;
 
+	struct iw_cm_verbs	     *iwcm;
+
 	int		           (*query_device)(struct ib_device *device,
 						   struct ib_device_attr *device_attr);
 	int		           (*query_port)(struct ib_device *device,
-- 
cgit v1.2.3


From c1a0b23bf477c2e1068905f4e2b5c3cee139e853 Mon Sep 17 00:00:00 2001
From: "Michael S. Tsirkin" <mst@mellanox.co.il>
Date: Mon, 21 Aug 2006 16:40:12 -0700
Subject: IB/sa: Require SA registration

Require users to register with SA module, to prevent the sa_query
module text from going away while an SA query callback is still
running.  Update all in-tree users for the new interface.

Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
---
 drivers/infiniband/core/cma.c                  |  7 ++-
 drivers/infiniband/core/sa_query.c             | 60 ++++++++++++++++++++++----
 drivers/infiniband/ulp/ipoib/ipoib.h           |  2 +
 drivers/infiniband/ulp/ipoib/ipoib_main.c      | 12 ++++--
 drivers/infiniband/ulp/ipoib/ipoib_multicast.c | 12 +++---
 drivers/infiniband/ulp/srp/ib_srp.c            |  9 +++-
 include/rdma/ib_sa.h                           | 41 +++++++++++++++---
 7 files changed, 116 insertions(+), 27 deletions(-)

(limited to 'include')

diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c
index 488fa1d642a7..1178bd434d1b 100644
--- a/drivers/infiniband/core/cma.c
+++ b/drivers/infiniband/core/cma.c
@@ -62,6 +62,7 @@ static struct ib_client cma_client = {
 	.remove = cma_remove_one
 };
 
+static struct ib_sa_client sa_client;
 static LIST_HEAD(dev_list);
 static LIST_HEAD(listen_any_list);
 static DEFINE_MUTEX(lock);
@@ -1323,7 +1324,7 @@ static int cma_query_ib_route(struct rdma_id_private *id_priv, int timeout_ms,
 	path_rec.pkey = cpu_to_be16(ib_addr_get_pkey(addr));
 	path_rec.numb_path = 1;
 
-	id_priv->query_id = ib_sa_path_rec_get(id_priv->id.device,
+	id_priv->query_id = ib_sa_path_rec_get(&sa_client, id_priv->id.device,
 				id_priv->id.port_num, &path_rec,
 				IB_SA_PATH_REC_DGID | IB_SA_PATH_REC_SGID |
 				IB_SA_PATH_REC_PKEY | IB_SA_PATH_REC_NUMB_PATH,
@@ -2199,12 +2200,15 @@ static int cma_init(void)
 	if (!cma_wq)
 		return -ENOMEM;
 
+	ib_sa_register_client(&sa_client);
+
 	ret = ib_register_client(&cma_client);
 	if (ret)
 		goto err;
 	return 0;
 
 err:
+	ib_sa_unregister_client(&sa_client);
 	destroy_workqueue(cma_wq);
 	return ret;
 }
@@ -2212,6 +2216,7 @@ err:
 static void cma_cleanup(void)
 {
 	ib_unregister_client(&cma_client);
+	ib_sa_unregister_client(&sa_client);
 	destroy_workqueue(cma_wq);
 	idr_destroy(&sdp_ps);
 	idr_destroy(&tcp_ps);
diff --git a/drivers/infiniband/core/sa_query.c b/drivers/infiniband/core/sa_query.c
index ca8760a7d88c..1706d3c7e95e 100644
--- a/drivers/infiniband/core/sa_query.c
+++ b/drivers/infiniband/core/sa_query.c
@@ -1,6 +1,7 @@
 /*
  * Copyright (c) 2004 Topspin Communications.  All rights reserved.
  * Copyright (c) 2005 Voltaire, Inc.  All rights reserved.
+ * Copyright (c) 2006 Intel Corporation.  All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU
@@ -75,6 +76,7 @@ struct ib_sa_device {
 struct ib_sa_query {
 	void (*callback)(struct ib_sa_query *, int, struct ib_sa_mad *);
 	void (*release)(struct ib_sa_query *);
+	struct ib_sa_client    *client;
 	struct ib_sa_port      *port;
 	struct ib_mad_send_buf *mad_buf;
 	struct ib_sa_sm_ah     *sm_ah;
@@ -415,6 +417,31 @@ static void ib_sa_event(struct ib_event_handler *handler, struct ib_event *event
 	}
 }
 
+void ib_sa_register_client(struct ib_sa_client *client)
+{
+	atomic_set(&client->users, 1);
+	init_completion(&client->comp);
+}
+EXPORT_SYMBOL(ib_sa_register_client);
+
+static inline void ib_sa_client_get(struct ib_sa_client *client)
+{
+	atomic_inc(&client->users);
+}
+
+static inline void ib_sa_client_put(struct ib_sa_client *client)
+{
+	if (atomic_dec_and_test(&client->users))
+		complete(&client->comp);
+}
+
+void ib_sa_unregister_client(struct ib_sa_client *client)
+{
+	ib_sa_client_put(client);
+	wait_for_completion(&client->comp);
+}
+EXPORT_SYMBOL(ib_sa_unregister_client);
+
 /**
  * ib_sa_cancel_query - try to cancel an SA query
  * @id:ID of query to cancel
@@ -557,6 +584,7 @@ static void ib_sa_path_rec_release(struct ib_sa_query *sa_query)
 
 /**
  * ib_sa_path_rec_get - Start a Path get query
+ * @client:SA client
  * @device:device to send query on
  * @port_num: port number to send query on
  * @rec:Path Record to send in query
@@ -579,7 +607,8 @@ static void ib_sa_path_rec_release(struct ib_sa_query *sa_query)
  * error code.  Otherwise it is a query ID that can be used to cancel
  * the query.
  */
-int ib_sa_path_rec_get(struct ib_device *device, u8 port_num,
+int ib_sa_path_rec_get(struct ib_sa_client *client,
+		       struct ib_device *device, u8 port_num,
 		       struct ib_sa_path_rec *rec,
 		       ib_sa_comp_mask comp_mask,
 		       int timeout_ms, gfp_t gfp_mask,
@@ -614,8 +643,10 @@ int ib_sa_path_rec_get(struct ib_device *device, u8 port_num,
 		goto err1;
 	}
 
-	query->callback = callback;
-	query->context  = context;
+	ib_sa_client_get(client);
+	query->sa_query.client = client;
+	query->callback        = callback;
+	query->context         = context;
 
 	mad = query->sa_query.mad_buf->mad;
 	init_mad(mad, agent);
@@ -639,6 +670,7 @@ int ib_sa_path_rec_get(struct ib_device *device, u8 port_num,
 
 err2:
 	*sa_query = NULL;
+	ib_sa_client_put(query->sa_query.client);
 	ib_free_send_mad(query->sa_query.mad_buf);
 
 err1:
@@ -671,6 +703,7 @@ static void ib_sa_service_rec_release(struct ib_sa_query *sa_query)
 
 /**
  * ib_sa_service_rec_query - Start Service Record operation
+ * @client:SA client
  * @device:device to send request on
  * @port_num: port number to send request on
  * @method:SA method - should be get, set, or delete
@@ -695,7 +728,8 @@ static void ib_sa_service_rec_release(struct ib_sa_query *sa_query)
  * error code.  Otherwise it is a request ID that can be used to cancel
  * the query.
  */
-int ib_sa_service_rec_query(struct ib_device *device, u8 port_num, u8 method,
+int ib_sa_service_rec_query(struct ib_sa_client *client,
+			    struct ib_device *device, u8 port_num, u8 method,
 			    struct ib_sa_service_rec *rec,
 			    ib_sa_comp_mask comp_mask,
 			    int timeout_ms, gfp_t gfp_mask,
@@ -735,8 +769,10 @@ int ib_sa_service_rec_query(struct ib_device *device, u8 port_num, u8 method,
 		goto err1;
 	}
 
-	query->callback = callback;
-	query->context  = context;
+	ib_sa_client_get(client);
+	query->sa_query.client = client;
+	query->callback        = callback;
+	query->context         = context;
 
 	mad = query->sa_query.mad_buf->mad;
 	init_mad(mad, agent);
@@ -761,6 +797,7 @@ int ib_sa_service_rec_query(struct ib_device *device, u8 port_num, u8 method,
 
 err2:
 	*sa_query = NULL;
+	ib_sa_client_put(query->sa_query.client);
 	ib_free_send_mad(query->sa_query.mad_buf);
 
 err1:
@@ -791,7 +828,8 @@ static void ib_sa_mcmember_rec_release(struct ib_sa_query *sa_query)
 	kfree(container_of(sa_query, struct ib_sa_mcmember_query, sa_query));
 }
 
-int ib_sa_mcmember_rec_query(struct ib_device *device, u8 port_num,
+int ib_sa_mcmember_rec_query(struct ib_sa_client *client,
+			     struct ib_device *device, u8 port_num,
 			     u8 method,
 			     struct ib_sa_mcmember_rec *rec,
 			     ib_sa_comp_mask comp_mask,
@@ -827,8 +865,10 @@ int ib_sa_mcmember_rec_query(struct ib_device *device, u8 port_num,
 		goto err1;
 	}
 
-	query->callback = callback;
-	query->context  = context;
+	ib_sa_client_get(client);
+	query->sa_query.client = client;
+	query->callback        = callback;
+	query->context         = context;
 
 	mad = query->sa_query.mad_buf->mad;
 	init_mad(mad, agent);
@@ -853,6 +893,7 @@ int ib_sa_mcmember_rec_query(struct ib_device *device, u8 port_num,
 
 err2:
 	*sa_query = NULL;
+	ib_sa_client_put(query->sa_query.client);
 	ib_free_send_mad(query->sa_query.mad_buf);
 
 err1:
@@ -889,6 +930,7 @@ static void send_handler(struct ib_mad_agent *agent,
 
 	ib_free_send_mad(mad_send_wc->send_buf);
 	kref_put(&query->sm_ah->ref, free_sm_ah);
+	ib_sa_client_put(query->client);
 	query->release(query);
 }
 
diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h b/drivers/infiniband/ulp/ipoib/ipoib.h
index 474aa214ab57..0b8a79d53a00 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib.h
+++ b/drivers/infiniband/ulp/ipoib/ipoib.h
@@ -336,6 +336,8 @@ static inline void ipoib_unregister_debugfs(void) { }
 extern int ipoib_sendq_size;
 extern int ipoib_recvq_size;
 
+extern struct ib_sa_client ipoib_sa_client;
+
 #ifdef CONFIG_INFINIBAND_IPOIB_DEBUG
 extern int ipoib_debug_level;
 
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c
index e9a7659eb1d7..ae3a4982cddb 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_main.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c
@@ -82,6 +82,8 @@ static const u8 ipv4_bcast_addr[] = {
 
 struct workqueue_struct *ipoib_workqueue;
 
+struct ib_sa_client ipoib_sa_client;
+
 static void ipoib_add_one(struct ib_device *device);
 static void ipoib_remove_one(struct ib_device *device);
 
@@ -463,7 +465,7 @@ static int path_rec_start(struct net_device *dev,
 	init_completion(&path->done);
 
 	path->query_id =
-		ib_sa_path_rec_get(priv->ca, priv->port,
+		ib_sa_path_rec_get(&ipoib_sa_client, priv->ca, priv->port,
 				   &path->pathrec,
 				   IB_SA_PATH_REC_DGID		|
 				   IB_SA_PATH_REC_SGID		|
@@ -1191,13 +1193,16 @@ static int __init ipoib_init_module(void)
 		goto err_fs;
 	}
 
+	ib_sa_register_client(&ipoib_sa_client);
+
 	ret = ib_register_client(&ipoib_client);
 	if (ret)
-		goto err_wq;
+		goto err_sa;
 
 	return 0;
 
-err_wq:
+err_sa:
+	ib_sa_unregister_client(&ipoib_sa_client);
 	destroy_workqueue(ipoib_workqueue);
 
 err_fs:
@@ -1209,6 +1214,7 @@ err_fs:
 static void __exit ipoib_cleanup_module(void)
 {
 	ib_unregister_client(&ipoib_client);
+	ib_sa_unregister_client(&ipoib_sa_client);
 	ipoib_unregister_debugfs();
 	destroy_workqueue(ipoib_workqueue);
 }
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
index 60b09f5cb347..fb3e4875a46d 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
@@ -361,7 +361,7 @@ static int ipoib_mcast_sendonly_join(struct ipoib_mcast *mcast)
 
 	init_completion(&mcast->done);
 
-	ret = ib_sa_mcmember_rec_set(priv->ca, priv->port, &rec,
+	ret = ib_sa_mcmember_rec_set(&ipoib_sa_client, priv->ca, priv->port, &rec,
 				     IB_SA_MCMEMBER_REC_MGID		|
 				     IB_SA_MCMEMBER_REC_PORT_GID	|
 				     IB_SA_MCMEMBER_REC_PKEY		|
@@ -485,9 +485,9 @@ static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast,
 
 	init_completion(&mcast->done);
 
-	ret = ib_sa_mcmember_rec_set(priv->ca, priv->port, &rec, comp_mask,
-				     mcast->backoff * 1000, GFP_ATOMIC,
-				     ipoib_mcast_join_complete,
+	ret = ib_sa_mcmember_rec_set(&ipoib_sa_client, priv->ca, priv->port,
+				     &rec, comp_mask, mcast->backoff * 1000,
+				     GFP_ATOMIC, ipoib_mcast_join_complete,
 				     mcast, &mcast->query);
 
 	if (ret < 0) {
@@ -528,7 +528,7 @@ void ipoib_mcast_join_task(void *dev_ptr)
 			priv->local_rate = attr.active_speed *
 				ib_width_enum_to_int(attr.active_width);
 		} else
-			ipoib_warn(priv, "ib_query_port failed\n");
+		ipoib_warn(priv, "ib_query_port failed\n");
 	}
 
 	if (!priv->broadcast) {
@@ -681,7 +681,7 @@ static int ipoib_mcast_leave(struct net_device *dev, struct ipoib_mcast *mcast)
 	 * Just make one shot at leaving and don't wait for a reply;
 	 * if we fail, too bad.
 	 */
-	ret = ib_sa_mcmember_rec_delete(priv->ca, priv->port, &rec,
+	ret = ib_sa_mcmember_rec_delete(&ipoib_sa_client, priv->ca, priv->port, &rec,
 					IB_SA_MCMEMBER_REC_MGID		|
 					IB_SA_MCMEMBER_REC_PORT_GID	|
 					IB_SA_MCMEMBER_REC_PKEY		|
diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
index feb1fcd0f2fb..44b9e5be6687 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.c
+++ b/drivers/infiniband/ulp/srp/ib_srp.c
@@ -96,6 +96,8 @@ static struct ib_client srp_client = {
 	.remove = srp_remove_one
 };
 
+static struct ib_sa_client srp_sa_client;
+
 static inline struct srp_target_port *host_to_target(struct Scsi_Host *host)
 {
 	return (struct srp_target_port *) host->hostdata;
@@ -267,7 +269,8 @@ static int srp_lookup_path(struct srp_target_port *target)
 
 	init_completion(&target->done);
 
-	target->path_query_id = ib_sa_path_rec_get(target->srp_host->dev->dev,
+	target->path_query_id = ib_sa_path_rec_get(&srp_sa_client,
+						   target->srp_host->dev->dev,
 						   target->srp_host->port,
 						   &target->path,
 						   IB_SA_PATH_REC_DGID		|
@@ -1998,9 +2001,12 @@ static int __init srp_init_module(void)
 		return ret;
 	}
 
+	ib_sa_register_client(&srp_sa_client);
+
 	ret = ib_register_client(&srp_client);
 	if (ret) {
 		printk(KERN_ERR PFX "couldn't register IB client\n");
+		ib_sa_unregister_client(&srp_sa_client);
 		class_unregister(&srp_class);
 		return ret;
 	}
@@ -2011,6 +2017,7 @@ static int __init srp_init_module(void)
 static void __exit srp_cleanup_module(void)
 {
 	ib_unregister_client(&srp_client);
+	ib_sa_unregister_client(&srp_sa_client);
 	class_unregister(&srp_class);
 }
 
diff --git a/include/rdma/ib_sa.h b/include/rdma/ib_sa.h
index c99e4420fd7e..58bb5f716fe3 100644
--- a/include/rdma/ib_sa.h
+++ b/include/rdma/ib_sa.h
@@ -1,6 +1,7 @@
 /*
  * Copyright (c) 2004 Topspin Communications.  All rights reserved.
  * Copyright (c) 2005 Voltaire, Inc.  All rights reserved.
+ * Copyright (c) 2006 Intel Corporation.  All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU
@@ -36,8 +37,11 @@
 #ifndef IB_SA_H
 #define IB_SA_H
 
+#include <linux/completion.h>
 #include <linux/compiler.h>
 
+#include <asm/atomic.h>
+
 #include <rdma/ib_verbs.h>
 #include <rdma/ib_mad.h>
 
@@ -250,11 +254,28 @@ struct ib_sa_service_rec {
 	u64		data64[2];
 };
 
+struct ib_sa_client {
+	atomic_t users;
+	struct completion comp;
+};
+
+/**
+ * ib_sa_register_client - Register an SA client.
+ */
+void ib_sa_register_client(struct ib_sa_client *client);
+
+/**
+ * ib_sa_unregister_client - Deregister an SA client.
+ * @client: Client object to deregister.
+ */
+void ib_sa_unregister_client(struct ib_sa_client *client);
+
 struct ib_sa_query;
 
 void ib_sa_cancel_query(int id, struct ib_sa_query *query);
 
-int ib_sa_path_rec_get(struct ib_device *device, u8 port_num,
+int ib_sa_path_rec_get(struct ib_sa_client *client,
+		       struct ib_device *device, u8 port_num,
 		       struct ib_sa_path_rec *rec,
 		       ib_sa_comp_mask comp_mask,
 		       int timeout_ms, gfp_t gfp_mask,
@@ -264,7 +285,8 @@ int ib_sa_path_rec_get(struct ib_device *device, u8 port_num,
 		       void *context,
 		       struct ib_sa_query **query);
 
-int ib_sa_mcmember_rec_query(struct ib_device *device, u8 port_num,
+int ib_sa_mcmember_rec_query(struct ib_sa_client *client,
+			     struct ib_device *device, u8 port_num,
 			     u8 method,
 			     struct ib_sa_mcmember_rec *rec,
 			     ib_sa_comp_mask comp_mask,
@@ -275,7 +297,8 @@ int ib_sa_mcmember_rec_query(struct ib_device *device, u8 port_num,
 			     void *context,
 			     struct ib_sa_query **query);
 
-int ib_sa_service_rec_query(struct ib_device *device, u8 port_num,
+int ib_sa_service_rec_query(struct ib_sa_client *client,
+			 struct ib_device *device, u8 port_num,
 			 u8 method,
 			 struct ib_sa_service_rec *rec,
 			 ib_sa_comp_mask comp_mask,
@@ -288,6 +311,7 @@ int ib_sa_service_rec_query(struct ib_device *device, u8 port_num,
 
 /**
  * ib_sa_mcmember_rec_set - Start an MCMember set query
+ * @client:SA client
  * @device:device to send query on
  * @port_num: port number to send query on
  * @rec:MCMember Record to send in query
@@ -311,7 +335,8 @@ int ib_sa_service_rec_query(struct ib_device *device, u8 port_num,
  * cancel the query.
  */
 static inline int
-ib_sa_mcmember_rec_set(struct ib_device *device, u8 port_num,
+ib_sa_mcmember_rec_set(struct ib_sa_client *client,
+		       struct ib_device *device, u8 port_num,
 		       struct ib_sa_mcmember_rec *rec,
 		       ib_sa_comp_mask comp_mask,
 		       int timeout_ms, gfp_t gfp_mask,
@@ -321,7 +346,7 @@ ib_sa_mcmember_rec_set(struct ib_device *device, u8 port_num,
 		       void *context,
 		       struct ib_sa_query **query)
 {
-	return ib_sa_mcmember_rec_query(device, port_num,
+	return ib_sa_mcmember_rec_query(client, device, port_num,
 					IB_MGMT_METHOD_SET,
 					rec, comp_mask,
 					timeout_ms, gfp_mask, callback,
@@ -330,6 +355,7 @@ ib_sa_mcmember_rec_set(struct ib_device *device, u8 port_num,
 
 /**
  * ib_sa_mcmember_rec_delete - Start an MCMember delete query
+ * @client:SA client
  * @device:device to send query on
  * @port_num: port number to send query on
  * @rec:MCMember Record to send in query
@@ -353,7 +379,8 @@ ib_sa_mcmember_rec_set(struct ib_device *device, u8 port_num,
  * cancel the query.
  */
 static inline int
-ib_sa_mcmember_rec_delete(struct ib_device *device, u8 port_num,
+ib_sa_mcmember_rec_delete(struct ib_sa_client *client,
+			  struct ib_device *device, u8 port_num,
 			  struct ib_sa_mcmember_rec *rec,
 			  ib_sa_comp_mask comp_mask,
 			  int timeout_ms, gfp_t gfp_mask,
@@ -363,7 +390,7 @@ ib_sa_mcmember_rec_delete(struct ib_device *device, u8 port_num,
 			  void *context,
 			  struct ib_sa_query **query)
 {
-	return ib_sa_mcmember_rec_query(device, port_num,
+	return ib_sa_mcmember_rec_query(client, device, port_num,
 					IB_SA_METHOD_DELETE,
 					rec, comp_mask,
 					timeout_ms, gfp_mask, callback,
-- 
cgit v1.2.3


From 07eeec0627e93a1a753c4df004a97a4d0a7b9ceb Mon Sep 17 00:00:00 2001
From: Or Gerlitz <ogerlitz@voltaire.com>
Date: Tue, 12 Sep 2006 09:03:33 -0700
Subject: RDMA/cma: Document rdma_destroy_id() function

Clarify that rdma_destroy_id cancels outstanding asynchronous operations on the
Associated id.

Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
---
 include/rdma/rdma_cm.h | 8 ++++++++
 1 file changed, 8 insertions(+)

(limited to 'include')

diff --git a/include/rdma/rdma_cm.h b/include/rdma/rdma_cm.h
index 402c63d7226b..1566be568ab6 100644
--- a/include/rdma/rdma_cm.h
+++ b/include/rdma/rdma_cm.h
@@ -117,6 +117,14 @@ struct rdma_cm_id {
 struct rdma_cm_id *rdma_create_id(rdma_cm_event_handler event_handler,
 				  void *context, enum rdma_port_space ps);
 
+/**
+  * rdma_destroy_id - Destroys an RDMA identifier.
+  *
+  * @id: RDMA identifier.
+  *
+  * Note: calling this function has the effect of canceling in-flight
+  * asynchronous operations associated with the id.
+  */
 void rdma_destroy_id(struct rdma_cm_id *id);
 
 /**
-- 
cgit v1.2.3


From 951f7fc1372da3d826b1d975b3cc5e3db92af5d0 Mon Sep 17 00:00:00 2001
From: Or Gerlitz <ogerlitz@voltaire.com>
Date: Fri, 22 Sep 2006 15:22:54 -0700
Subject: RDMA/cma: Document rdma_accept() error handling

Document the reject sending and modifying QP to error done in rdma_accept().

Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
---
 include/rdma/rdma_cm.h | 4 ++++
 1 file changed, 4 insertions(+)

(limited to 'include')

diff --git a/include/rdma/rdma_cm.h b/include/rdma/rdma_cm.h
index 1566be568ab6..deb5a0a4cee5 100644
--- a/include/rdma/rdma_cm.h
+++ b/include/rdma/rdma_cm.h
@@ -245,6 +245,10 @@ int rdma_listen(struct rdma_cm_id *id, int backlog);
  * Typically, this routine is only called by the listener to accept a connection
  * request.  It must also be called on the active side of a connection if the
  * user is performing their own QP transitions.
+ *
+ * In the case of error, a reject message is sent to the remote side and the
+ * state of the qp associated with the id is modified to error, such that any
+ * previously posted receive buffers would be flushed.
  */
 int rdma_accept(struct rdma_cm_id *id, struct rdma_conn_param *conn_param);
 
-- 
cgit v1.2.3


From aec79fcc3ea3b536a2788b4e22b7ebabbb176485 Mon Sep 17 00:00:00 2001
From: "Michael S. Tsirkin" <mst@mellanox.co.il>
Date: Mon, 18 Sep 2006 22:17:08 +0300
Subject: IB/sa: fix ib_sa_selector names

Relevant SA queries are actually "greater than" / "less than", not
"greater than or equal" / "less than or equal" as the names imply.
(See IB spec 1.2 Vol 1, 15.2.5.16 PATHRECORD/Table 205 PathRecord)

Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
---
 include/rdma/ib_sa.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/include/rdma/ib_sa.h b/include/rdma/ib_sa.h
index 58bb5f716fe3..97715b0c20b6 100644
--- a/include/rdma/ib_sa.h
+++ b/include/rdma/ib_sa.h
@@ -83,8 +83,8 @@ enum {
 };
 
 enum ib_sa_selector {
-	IB_SA_GTE  = 0,
-	IB_SA_LTE  = 1,
+	IB_SA_GT   = 0,
+	IB_SA_LT   = 1,
 	IB_SA_EQ   = 2,
 	/*
 	 * The meaning of "best" depends on the attribute: for
-- 
cgit v1.2.3


From 1d22e05df8183b36b3cc0760344774040abc74d5 Mon Sep 17 00:00:00 2001
From: Lennert Buytenhek <buytenh@wantstofly.org>
Date: Fri, 22 Sep 2006 02:28:13 +0200
Subject: [PATCH] Cirrus Logic ep93xx ethernet driver

The Cirrus Logic ep93xx is an ARM SoC that includes an ethernet MAC
-- this patch adds a driver for that ethernet MAC.

Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/net/arm/Kconfig                   |   7 +
 drivers/net/arm/Makefile                  |   1 +
 drivers/net/arm/ep93xx_eth.c              | 944 ++++++++++++++++++++++++++++++
 include/asm-arm/arch-ep93xx/ep93xx-regs.h |   1 +
 include/asm-arm/arch-ep93xx/platform.h    |   6 +
 5 files changed, 959 insertions(+)
 create mode 100644 drivers/net/arm/ep93xx_eth.c

(limited to 'include')

diff --git a/drivers/net/arm/Kconfig b/drivers/net/arm/Kconfig
index 77fe20dbea35..678e4f48d36b 100644
--- a/drivers/net/arm/Kconfig
+++ b/drivers/net/arm/Kconfig
@@ -39,3 +39,10 @@ config ARM_AT91_ETHER
 	help
 	  If you wish to compile a kernel for the AT91RM9200 and enable
 	  ethernet support, then you should always answer Y to this.
+
+config EP93XX_ETH
+	tristate "EP93xx Ethernet support"
+	depends on NET_ETHERNET && ARM && ARCH_EP93XX
+	help
+	  This is a driver for the ethernet hardware included in EP93xx CPUs.
+	  Say Y if you are building a kernel for EP93xx based devices.
diff --git a/drivers/net/arm/Makefile b/drivers/net/arm/Makefile
index 42c95b79c261..a4c868278e11 100644
--- a/drivers/net/arm/Makefile
+++ b/drivers/net/arm/Makefile
@@ -8,3 +8,4 @@ obj-$(CONFIG_ARM_ETHERH)	+= etherh.o
 obj-$(CONFIG_ARM_ETHER3)	+= ether3.o
 obj-$(CONFIG_ARM_ETHER1)	+= ether1.o
 obj-$(CONFIG_ARM_AT91_ETHER)	+= at91_ether.o
+obj-$(CONFIG_EP93XX_ETH)	+= ep93xx_eth.o
diff --git a/drivers/net/arm/ep93xx_eth.c b/drivers/net/arm/ep93xx_eth.c
new file mode 100644
index 000000000000..cef00744a9dc
--- /dev/null
+++ b/drivers/net/arm/ep93xx_eth.c
@@ -0,0 +1,944 @@
+/*
+ * EP93xx ethernet network device driver
+ * Copyright (C) 2006 Lennert Buytenhek <buytenh@wantstofly.org>
+ * Dedicated to Marija Kulikova.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/config.h>
+#include <linux/dma-mapping.h>
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/netdevice.h>
+#include <linux/mii.h>
+#include <linux/etherdevice.h>
+#include <linux/ethtool.h>
+#include <linux/init.h>
+#include <linux/moduleparam.h>
+#include <linux/platform_device.h>
+#include <linux/delay.h>
+#include <asm/arch/ep93xx-regs.h>
+#include <asm/arch/platform.h>
+#include <asm/io.h>
+
+#define DRV_MODULE_NAME		"ep93xx-eth"
+#define DRV_MODULE_VERSION	"0.1"
+
+#define RX_QUEUE_ENTRIES	64
+#define TX_QUEUE_ENTRIES	8
+
+#define MAX_PKT_SIZE		2044
+#define PKT_BUF_SIZE		2048
+
+#define REG_RXCTL		0x0000
+#define  REG_RXCTL_DEFAULT	0x00073800
+#define REG_TXCTL		0x0004
+#define  REG_TXCTL_ENABLE	0x00000001
+#define REG_MIICMD		0x0010
+#define  REG_MIICMD_READ	0x00008000
+#define  REG_MIICMD_WRITE	0x00004000
+#define REG_MIIDATA		0x0014
+#define REG_MIISTS		0x0018
+#define  REG_MIISTS_BUSY	0x00000001
+#define REG_SELFCTL		0x0020
+#define  REG_SELFCTL_RESET	0x00000001
+#define REG_INTEN		0x0024
+#define  REG_INTEN_TX		0x00000008
+#define  REG_INTEN_RX		0x00000007
+#define REG_INTSTSP		0x0028
+#define  REG_INTSTS_TX		0x00000008
+#define  REG_INTSTS_RX		0x00000004
+#define REG_INTSTSC		0x002c
+#define REG_AFP			0x004c
+#define REG_INDAD0		0x0050
+#define REG_INDAD1		0x0051
+#define REG_INDAD2		0x0052
+#define REG_INDAD3		0x0053
+#define REG_INDAD4		0x0054
+#define REG_INDAD5		0x0055
+#define REG_GIINTMSK		0x0064
+#define  REG_GIINTMSK_ENABLE	0x00008000
+#define REG_BMCTL		0x0080
+#define  REG_BMCTL_ENABLE_TX	0x00000100
+#define  REG_BMCTL_ENABLE_RX	0x00000001
+#define REG_BMSTS		0x0084
+#define  REG_BMSTS_RX_ACTIVE	0x00000008
+#define REG_RXDQBADD		0x0090
+#define REG_RXDQBLEN		0x0094
+#define REG_RXDCURADD		0x0098
+#define REG_RXDENQ		0x009c
+#define REG_RXSTSQBADD		0x00a0
+#define REG_RXSTSQBLEN		0x00a4
+#define REG_RXSTSQCURADD	0x00a8
+#define REG_RXSTSENQ		0x00ac
+#define REG_TXDQBADD		0x00b0
+#define REG_TXDQBLEN		0x00b4
+#define REG_TXDQCURADD		0x00b8
+#define REG_TXDENQ		0x00bc
+#define REG_TXSTSQBADD		0x00c0
+#define REG_TXSTSQBLEN		0x00c4
+#define REG_TXSTSQCURADD	0x00c8
+#define REG_MAXFRMLEN		0x00e8
+
+struct ep93xx_rdesc
+{
+	u32	buf_addr;
+	u32	rdesc1;
+};
+
+#define RDESC1_NSOF		0x80000000
+#define RDESC1_BUFFER_INDEX	0x7fff0000
+#define RDESC1_BUFFER_LENGTH	0x0000ffff
+
+struct ep93xx_rstat
+{
+	u32	rstat0;
+	u32	rstat1;
+};
+
+#define RSTAT0_RFP		0x80000000
+#define RSTAT0_RWE		0x40000000
+#define RSTAT0_EOF		0x20000000
+#define RSTAT0_EOB		0x10000000
+#define RSTAT0_AM		0x00c00000
+#define RSTAT0_RX_ERR		0x00200000
+#define RSTAT0_OE		0x00100000
+#define RSTAT0_FE		0x00080000
+#define RSTAT0_RUNT		0x00040000
+#define RSTAT0_EDATA		0x00020000
+#define RSTAT0_CRCE		0x00010000
+#define RSTAT0_CRCI		0x00008000
+#define RSTAT0_HTI		0x00003f00
+#define RSTAT1_RFP		0x80000000
+#define RSTAT1_BUFFER_INDEX	0x7fff0000
+#define RSTAT1_FRAME_LENGTH	0x0000ffff
+
+struct ep93xx_tdesc
+{
+	u32	buf_addr;
+	u32	tdesc1;
+};
+
+#define TDESC1_EOF		0x80000000
+#define TDESC1_BUFFER_INDEX	0x7fff0000
+#define TDESC1_BUFFER_ABORT	0x00008000
+#define TDESC1_BUFFER_LENGTH	0x00000fff
+
+struct ep93xx_tstat
+{
+	u32	tstat0;
+};
+
+#define TSTAT0_TXFP		0x80000000
+#define TSTAT0_TXWE		0x40000000
+#define TSTAT0_FA		0x20000000
+#define TSTAT0_LCRS		0x10000000
+#define TSTAT0_OW		0x04000000
+#define TSTAT0_TXU		0x02000000
+#define TSTAT0_ECOLL		0x01000000
+#define TSTAT0_NCOLL		0x001f0000
+#define TSTAT0_BUFFER_INDEX	0x00007fff
+
+struct ep93xx_descs
+{
+	struct ep93xx_rdesc	rdesc[RX_QUEUE_ENTRIES];
+	struct ep93xx_tdesc	tdesc[TX_QUEUE_ENTRIES];
+	struct ep93xx_rstat	rstat[RX_QUEUE_ENTRIES];
+	struct ep93xx_tstat	tstat[TX_QUEUE_ENTRIES];
+};
+
+struct ep93xx_priv
+{
+	struct resource		*res;
+	void			*base_addr;
+	int			irq;
+
+	struct ep93xx_descs	*descs;
+	dma_addr_t		descs_dma_addr;
+
+	void			*rx_buf[RX_QUEUE_ENTRIES];
+	void			*tx_buf[TX_QUEUE_ENTRIES];
+
+	spinlock_t		rx_lock;
+	unsigned int		rx_pointer;
+	unsigned int		tx_clean_pointer;
+	unsigned int		tx_pointer;
+	spinlock_t		tx_pending_lock;
+	unsigned int		tx_pending;
+
+	struct net_device_stats	stats;
+
+	struct mii_if_info	mii;
+	u8			mdc_divisor;
+};
+
+#define rdb(ep, off)		__raw_readb((ep)->base_addr + (off))
+#define rdw(ep, off)		__raw_readw((ep)->base_addr + (off))
+#define rdl(ep, off)		__raw_readl((ep)->base_addr + (off))
+#define wrb(ep, off, val)	__raw_writeb((val), (ep)->base_addr + (off))
+#define wrw(ep, off, val)	__raw_writew((val), (ep)->base_addr + (off))
+#define wrl(ep, off, val)	__raw_writel((val), (ep)->base_addr + (off))
+
+static int ep93xx_mdio_read(struct net_device *dev, int phy_id, int reg);
+
+static struct net_device_stats *ep93xx_get_stats(struct net_device *dev)
+{
+	struct ep93xx_priv *ep = netdev_priv(dev);
+	return &(ep->stats);
+}
+
+static int ep93xx_rx(struct net_device *dev, int *budget)
+{
+	struct ep93xx_priv *ep = netdev_priv(dev);
+	int tail_offset;
+	int rx_done;
+	int processed;
+
+	tail_offset = rdl(ep, REG_RXSTSQCURADD) - ep->descs_dma_addr;
+
+	rx_done = 0;
+	processed = 0;
+	while (*budget > 0) {
+		int entry;
+		struct ep93xx_rstat *rstat;
+		u32 rstat0;
+		u32 rstat1;
+		int length;
+		struct sk_buff *skb;
+
+		entry = ep->rx_pointer;
+		rstat = ep->descs->rstat + entry;
+		if ((void *)rstat - (void *)ep->descs == tail_offset) {
+			rx_done = 1;
+			break;
+		}
+
+		rstat0 = rstat->rstat0;
+		rstat1 = rstat->rstat1;
+		rstat->rstat0 = 0;
+		rstat->rstat1 = 0;
+
+		if (!(rstat0 & RSTAT0_RFP))
+			printk(KERN_CRIT "ep93xx_rx: buffer not done "
+					 " %.8x %.8x\n", rstat0, rstat1);
+		if (!(rstat0 & RSTAT0_EOF))
+			printk(KERN_CRIT "ep93xx_rx: not end-of-frame "
+					 " %.8x %.8x\n", rstat0, rstat1);
+		if (!(rstat0 & RSTAT0_EOB))
+			printk(KERN_CRIT "ep93xx_rx: not end-of-buffer "
+					 " %.8x %.8x\n", rstat0, rstat1);
+		if (!(rstat1 & RSTAT1_RFP))
+			printk(KERN_CRIT "ep93xx_rx: buffer1 not done "
+					 " %.8x %.8x\n", rstat0, rstat1);
+		if ((rstat1 & RSTAT1_BUFFER_INDEX) >> 16 != entry)
+			printk(KERN_CRIT "ep93xx_rx: entry mismatch "
+					 " %.8x %.8x\n", rstat0, rstat1);
+
+		if (!(rstat0 & RSTAT0_RWE)) {
+			printk(KERN_NOTICE "ep93xx_rx: receive error "
+					 " %.8x %.8x\n", rstat0, rstat1);
+
+			ep->stats.rx_errors++;
+			if (rstat0 & RSTAT0_OE)
+				ep->stats.rx_fifo_errors++;
+			if (rstat0 & RSTAT0_FE)
+				ep->stats.rx_frame_errors++;
+			if (rstat0 & (RSTAT0_RUNT | RSTAT0_EDATA))
+				ep->stats.rx_length_errors++;
+			if (rstat0 & RSTAT0_CRCE)
+				ep->stats.rx_crc_errors++;
+			goto err;
+		}
+
+		length = rstat1 & RSTAT1_FRAME_LENGTH;
+		if (length > MAX_PKT_SIZE) {
+			printk(KERN_NOTICE "ep93xx_rx: invalid length "
+					 " %.8x %.8x\n", rstat0, rstat1);
+			goto err;
+		}
+
+		/* Strip FCS.  */
+		if (rstat0 & RSTAT0_CRCI)
+			length -= 4;
+
+		skb = dev_alloc_skb(length + 2);
+		if (likely(skb != NULL)) {
+			skb->dev = dev;
+			skb_reserve(skb, 2);
+			dma_sync_single(NULL, ep->descs->rdesc[entry].buf_addr,
+						length, DMA_FROM_DEVICE);
+			eth_copy_and_sum(skb, ep->rx_buf[entry], length, 0);
+			skb_put(skb, length);
+			skb->protocol = eth_type_trans(skb, dev);
+
+			dev->last_rx = jiffies;
+
+			netif_receive_skb(skb);
+
+			ep->stats.rx_packets++;
+			ep->stats.rx_bytes += length;
+		} else {
+			ep->stats.rx_dropped++;
+		}
+
+err:
+		ep->rx_pointer = (entry + 1) & (RX_QUEUE_ENTRIES - 1);
+		processed++;
+		dev->quota--;
+		(*budget)--;
+	}
+
+	if (processed) {
+		wrw(ep, REG_RXDENQ, processed);
+		wrw(ep, REG_RXSTSENQ, processed);
+	}
+
+	return !rx_done;
+}
+
+static int ep93xx_have_more_rx(struct ep93xx_priv *ep)
+{
+	struct ep93xx_rstat *rstat;
+	int tail_offset;
+
+	rstat = ep->descs->rstat + ep->rx_pointer;
+	tail_offset = rdl(ep, REG_RXSTSQCURADD) - ep->descs_dma_addr;
+
+	return !((void *)rstat - (void *)ep->descs == tail_offset);
+}
+
+static int ep93xx_poll(struct net_device *dev, int *budget)
+{
+	struct ep93xx_priv *ep = netdev_priv(dev);
+
+	/*
+	 * @@@ Have to stop polling if device is downed while we
+	 * are polling.
+	 */
+
+poll_some_more:
+	if (ep93xx_rx(dev, budget))
+		return 1;
+
+	netif_rx_complete(dev);
+
+	spin_lock_irq(&ep->rx_lock);
+	wrl(ep, REG_INTEN, REG_INTEN_TX | REG_INTEN_RX);
+	if (ep93xx_have_more_rx(ep)) {
+		wrl(ep, REG_INTEN, REG_INTEN_TX);
+		wrl(ep, REG_INTSTSP, REG_INTSTS_RX);
+		spin_unlock_irq(&ep->rx_lock);
+
+		if (netif_rx_reschedule(dev, 0))
+			goto poll_some_more;
+
+		return 0;
+	}
+	spin_unlock_irq(&ep->rx_lock);
+
+	return 0;
+}
+
+static int ep93xx_xmit(struct sk_buff *skb, struct net_device *dev)
+{
+	struct ep93xx_priv *ep = netdev_priv(dev);
+	int entry;
+
+	if (unlikely(skb->len) > MAX_PKT_SIZE) {
+		ep->stats.tx_dropped++;
+		dev_kfree_skb(skb);
+		return NETDEV_TX_OK;
+	}
+
+	entry = ep->tx_pointer;
+	ep->tx_pointer = (ep->tx_pointer + 1) & (TX_QUEUE_ENTRIES - 1);
+
+	ep->descs->tdesc[entry].tdesc1 =
+		TDESC1_EOF | (entry << 16) | (skb->len & 0xfff);
+	skb_copy_and_csum_dev(skb, ep->tx_buf[entry]);
+	dma_sync_single(NULL, ep->descs->tdesc[entry].buf_addr,
+				skb->len, DMA_TO_DEVICE);
+	dev_kfree_skb(skb);
+
+	dev->trans_start = jiffies;
+
+	spin_lock_irq(&ep->tx_pending_lock);
+	ep->tx_pending++;
+	if (ep->tx_pending == TX_QUEUE_ENTRIES)
+		netif_stop_queue(dev);
+	spin_unlock_irq(&ep->tx_pending_lock);
+
+	wrl(ep, REG_TXDENQ, 1);
+
+	return NETDEV_TX_OK;
+}
+
+static void ep93xx_tx_complete(struct net_device *dev)
+{
+	struct ep93xx_priv *ep = netdev_priv(dev);
+	int tail_offset;
+	int wake;
+
+	tail_offset = rdl(ep, REG_TXSTSQCURADD) - ep->descs_dma_addr;
+	wake = 0;
+
+	spin_lock(&ep->tx_pending_lock);
+	while (1) {
+		int entry;
+		struct ep93xx_tstat *tstat;
+		u32 tstat0;
+
+		entry = ep->tx_clean_pointer;
+		tstat = ep->descs->tstat + entry;
+		if ((void *)tstat - (void *)ep->descs == tail_offset)
+			break;
+
+		tstat0 = tstat->tstat0;
+		tstat->tstat0 = 0;
+
+		if (!(tstat0 & TSTAT0_TXFP))
+			printk(KERN_CRIT "ep93xx_tx_complete: buffer not done "
+					 " %.8x\n", tstat0);
+		if (tstat0 & TSTAT0_FA)
+			printk(KERN_CRIT "ep93xx_tx_complete: frame aborted "
+					 " %.8x\n", tstat0);
+		if ((tstat0 & TSTAT0_BUFFER_INDEX) != entry)
+			printk(KERN_CRIT "ep93xx_tx_complete: entry mismatch "
+					 " %.8x\n", tstat0);
+
+		if (tstat0 & TSTAT0_TXWE) {
+			int length = ep->descs->tdesc[entry].tdesc1 & 0xfff;
+
+			ep->stats.tx_packets++;
+			ep->stats.tx_bytes += length;
+		} else {
+			ep->stats.tx_errors++;
+		}
+
+		if (tstat0 & TSTAT0_OW)
+			ep->stats.tx_window_errors++;
+		if (tstat0 & TSTAT0_TXU)
+			ep->stats.tx_fifo_errors++;
+		ep->stats.collisions += (tstat0 >> 16) & 0x1f;
+
+		ep->tx_clean_pointer = (entry + 1) & (TX_QUEUE_ENTRIES - 1);
+		if (ep->tx_pending == TX_QUEUE_ENTRIES)
+			wake = 1;
+		ep->tx_pending--;
+	}
+	spin_unlock(&ep->tx_pending_lock);
+
+	if (wake)
+		netif_wake_queue(dev);
+}
+
+static irqreturn_t ep93xx_irq(int irq, void *dev_id, struct pt_regs *regs)
+{
+	struct net_device *dev = dev_id;
+	struct ep93xx_priv *ep = netdev_priv(dev);
+	u32 status;
+
+	status = rdl(ep, REG_INTSTSC);
+	if (status == 0)
+		return IRQ_NONE;
+
+	if (status & REG_INTSTS_RX) {
+		spin_lock(&ep->rx_lock);
+		if (likely(__netif_rx_schedule_prep(dev))) {
+			wrl(ep, REG_INTEN, REG_INTEN_TX);
+			__netif_rx_schedule(dev);
+		}
+		spin_unlock(&ep->rx_lock);
+	}
+
+	if (status & REG_INTSTS_TX)
+		ep93xx_tx_complete(dev);
+
+	return IRQ_HANDLED;
+}
+
+static void ep93xx_free_buffers(struct ep93xx_priv *ep)
+{
+	int i;
+
+	for (i = 0; i < RX_QUEUE_ENTRIES; i += 2) {
+		dma_addr_t d;
+
+		d = ep->descs->rdesc[i].buf_addr;
+		if (d)
+			dma_unmap_single(NULL, d, PAGE_SIZE, DMA_FROM_DEVICE);
+
+		if (ep->rx_buf[i] != NULL)
+			free_page((unsigned long)ep->rx_buf[i]);
+	}
+
+	for (i = 0; i < TX_QUEUE_ENTRIES; i += 2) {
+		dma_addr_t d;
+
+		d = ep->descs->tdesc[i].buf_addr;
+		if (d)
+			dma_unmap_single(NULL, d, PAGE_SIZE, DMA_TO_DEVICE);
+
+		if (ep->tx_buf[i] != NULL)
+			free_page((unsigned long)ep->tx_buf[i]);
+	}
+
+	dma_free_coherent(NULL, sizeof(struct ep93xx_descs), ep->descs,
+							ep->descs_dma_addr);
+}
+
+/*
+ * The hardware enforces a sub-2K maximum packet size, so we put
+ * two buffers on every hardware page.
+ */
+static int ep93xx_alloc_buffers(struct ep93xx_priv *ep)
+{
+	int i;
+
+	ep->descs = dma_alloc_coherent(NULL, sizeof(struct ep93xx_descs),
+				&ep->descs_dma_addr, GFP_KERNEL | GFP_DMA);
+	if (ep->descs == NULL)
+		return 1;
+
+	for (i = 0; i < RX_QUEUE_ENTRIES; i += 2) {
+		void *page;
+		dma_addr_t d;
+
+		page = (void *)__get_free_page(GFP_KERNEL | GFP_DMA);
+		if (page == NULL)
+			goto err;
+
+		d = dma_map_single(NULL, page, PAGE_SIZE, DMA_FROM_DEVICE);
+		if (dma_mapping_error(d)) {
+			free_page((unsigned long)page);
+			goto err;
+		}
+
+		ep->rx_buf[i] = page;
+		ep->descs->rdesc[i].buf_addr = d;
+		ep->descs->rdesc[i].rdesc1 = (i << 16) | PKT_BUF_SIZE;
+
+		ep->rx_buf[i + 1] = page + PKT_BUF_SIZE;
+		ep->descs->rdesc[i + 1].buf_addr = d + PKT_BUF_SIZE;
+		ep->descs->rdesc[i + 1].rdesc1 = ((i + 1) << 16) | PKT_BUF_SIZE;
+	}
+
+	for (i = 0; i < TX_QUEUE_ENTRIES; i += 2) {
+		void *page;
+		dma_addr_t d;
+
+		page = (void *)__get_free_page(GFP_KERNEL | GFP_DMA);
+		if (page == NULL)
+			goto err;
+
+		d = dma_map_single(NULL, page, PAGE_SIZE, DMA_TO_DEVICE);
+		if (dma_mapping_error(d)) {
+			free_page((unsigned long)page);
+			goto err;
+		}
+
+		ep->tx_buf[i] = page;
+		ep->descs->tdesc[i].buf_addr = d;
+
+		ep->tx_buf[i + 1] = page + PKT_BUF_SIZE;
+		ep->descs->tdesc[i + 1].buf_addr = d + PKT_BUF_SIZE;
+	}
+
+	return 0;
+
+err:
+	ep93xx_free_buffers(ep);
+	return 1;
+}
+
+static int ep93xx_start_hw(struct net_device *dev)
+{
+	struct ep93xx_priv *ep = netdev_priv(dev);
+	unsigned long addr;
+	int i;
+
+	wrl(ep, REG_SELFCTL, REG_SELFCTL_RESET);
+	for (i = 0; i < 10; i++) {
+		if ((rdl(ep, REG_SELFCTL) & REG_SELFCTL_RESET) == 0)
+			break;
+		msleep(1);
+	}
+
+	if (i == 10) {
+		printk(KERN_CRIT DRV_MODULE_NAME ": hw failed to reset\n");
+		return 1;
+	}
+
+	wrl(ep, REG_SELFCTL, ((ep->mdc_divisor - 1) << 9));
+
+	/* Does the PHY support preamble suppress?  */
+	if ((ep93xx_mdio_read(dev, ep->mii.phy_id, MII_BMSR) & 0x0040) != 0)
+		wrl(ep, REG_SELFCTL, ((ep->mdc_divisor - 1) << 9) | (1 << 8));
+
+	/* Receive descriptor ring.  */
+	addr = ep->descs_dma_addr + offsetof(struct ep93xx_descs, rdesc);
+	wrl(ep, REG_RXDQBADD, addr);
+	wrl(ep, REG_RXDCURADD, addr);
+	wrw(ep, REG_RXDQBLEN, RX_QUEUE_ENTRIES * sizeof(struct ep93xx_rdesc));
+
+	/* Receive status ring.  */
+	addr = ep->descs_dma_addr + offsetof(struct ep93xx_descs, rstat);
+	wrl(ep, REG_RXSTSQBADD, addr);
+	wrl(ep, REG_RXSTSQCURADD, addr);
+	wrw(ep, REG_RXSTSQBLEN, RX_QUEUE_ENTRIES * sizeof(struct ep93xx_rstat));
+
+	/* Transmit descriptor ring.  */
+	addr = ep->descs_dma_addr + offsetof(struct ep93xx_descs, tdesc);
+	wrl(ep, REG_TXDQBADD, addr);
+	wrl(ep, REG_TXDQCURADD, addr);
+	wrw(ep, REG_TXDQBLEN, TX_QUEUE_ENTRIES * sizeof(struct ep93xx_tdesc));
+
+	/* Transmit status ring.  */
+	addr = ep->descs_dma_addr + offsetof(struct ep93xx_descs, tstat);
+	wrl(ep, REG_TXSTSQBADD, addr);
+	wrl(ep, REG_TXSTSQCURADD, addr);
+	wrw(ep, REG_TXSTSQBLEN, TX_QUEUE_ENTRIES * sizeof(struct ep93xx_tstat));
+
+	wrl(ep, REG_BMCTL, REG_BMCTL_ENABLE_TX | REG_BMCTL_ENABLE_RX);
+	wrl(ep, REG_INTEN, REG_INTEN_TX | REG_INTEN_RX);
+	wrl(ep, REG_GIINTMSK, 0);
+
+	for (i = 0; i < 10; i++) {
+		if ((rdl(ep, REG_BMSTS) & REG_BMSTS_RX_ACTIVE) != 0)
+			break;
+		msleep(1);
+	}
+
+	if (i == 10) {
+		printk(KERN_CRIT DRV_MODULE_NAME ": hw failed to start\n");
+		return 1;
+	}
+
+	wrl(ep, REG_RXDENQ, RX_QUEUE_ENTRIES);
+	wrl(ep, REG_RXSTSENQ, RX_QUEUE_ENTRIES);
+
+	wrb(ep, REG_INDAD0, dev->dev_addr[0]);
+	wrb(ep, REG_INDAD1, dev->dev_addr[1]);
+	wrb(ep, REG_INDAD2, dev->dev_addr[2]);
+	wrb(ep, REG_INDAD3, dev->dev_addr[3]);
+	wrb(ep, REG_INDAD4, dev->dev_addr[4]);
+	wrb(ep, REG_INDAD5, dev->dev_addr[5]);
+	wrl(ep, REG_AFP, 0);
+
+	wrl(ep, REG_MAXFRMLEN, (MAX_PKT_SIZE << 16) | MAX_PKT_SIZE);
+
+	wrl(ep, REG_RXCTL, REG_RXCTL_DEFAULT);
+	wrl(ep, REG_TXCTL, REG_TXCTL_ENABLE);
+
+	return 0;
+}
+
+static void ep93xx_stop_hw(struct net_device *dev)
+{
+	struct ep93xx_priv *ep = netdev_priv(dev);
+	int i;
+
+	wrl(ep, REG_SELFCTL, REG_SELFCTL_RESET);
+	for (i = 0; i < 10; i++) {
+		if ((rdl(ep, REG_SELFCTL) & REG_SELFCTL_RESET) == 0)
+			break;
+		msleep(1);
+	}
+
+	if (i == 10)
+		printk(KERN_CRIT DRV_MODULE_NAME ": hw failed to reset\n");
+}
+
+static int ep93xx_open(struct net_device *dev)
+{
+	struct ep93xx_priv *ep = netdev_priv(dev);
+	int err;
+
+	if (ep93xx_alloc_buffers(ep))
+		return -ENOMEM;
+
+	if (is_zero_ether_addr(dev->dev_addr)) {
+		random_ether_addr(dev->dev_addr);
+		printk(KERN_INFO "%s: generated random MAC address "
+			"%.2x:%.2x:%.2x:%.2x:%.2x:%.2x.\n", dev->name,
+			dev->dev_addr[0], dev->dev_addr[1],
+			dev->dev_addr[2], dev->dev_addr[3],
+			dev->dev_addr[4], dev->dev_addr[5]);
+	}
+
+	if (ep93xx_start_hw(dev)) {
+		ep93xx_free_buffers(ep);
+		return -EIO;
+	}
+
+	spin_lock_init(&ep->rx_lock);
+	ep->rx_pointer = 0;
+	ep->tx_clean_pointer = 0;
+	ep->tx_pointer = 0;
+	spin_lock_init(&ep->tx_pending_lock);
+	ep->tx_pending = 0;
+
+	err = request_irq(ep->irq, ep93xx_irq, IRQF_SHARED, dev->name, dev);
+	if (err) {
+		ep93xx_stop_hw(dev);
+		ep93xx_free_buffers(ep);
+		return err;
+	}
+
+	wrl(ep, REG_GIINTMSK, REG_GIINTMSK_ENABLE);
+
+	netif_start_queue(dev);
+
+	return 0;
+}
+
+static int ep93xx_close(struct net_device *dev)
+{
+	struct ep93xx_priv *ep = netdev_priv(dev);
+
+	netif_stop_queue(dev);
+
+	wrl(ep, REG_GIINTMSK, 0);
+	free_irq(ep->irq, dev);
+	ep93xx_stop_hw(dev);
+	ep93xx_free_buffers(ep);
+
+	return 0;
+}
+
+static int ep93xx_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd)
+{
+	struct ep93xx_priv *ep = netdev_priv(dev);
+	struct mii_ioctl_data *data = if_mii(ifr);
+
+	return generic_mii_ioctl(&ep->mii, data, cmd, NULL);
+}
+
+static int ep93xx_mdio_read(struct net_device *dev, int phy_id, int reg)
+{
+	struct ep93xx_priv *ep = netdev_priv(dev);
+	int data;
+	int i;
+
+	wrl(ep, REG_MIICMD, REG_MIICMD_READ | (phy_id << 5) | reg);
+
+	for (i = 0; i < 10; i++) {
+		if ((rdl(ep, REG_MIISTS) & REG_MIISTS_BUSY) == 0)
+			break;
+		msleep(1);
+	}
+
+	if (i == 10) {
+		printk(KERN_INFO DRV_MODULE_NAME ": mdio read timed out\n");
+		data = 0xffff;
+	} else {
+		data = rdl(ep, REG_MIIDATA);
+	}
+
+	return data;
+}
+
+static void ep93xx_mdio_write(struct net_device *dev, int phy_id, int reg, int data)
+{
+	struct ep93xx_priv *ep = netdev_priv(dev);
+	int i;
+
+	wrl(ep, REG_MIIDATA, data);
+	wrl(ep, REG_MIICMD, REG_MIICMD_WRITE | (phy_id << 5) | reg);
+
+	for (i = 0; i < 10; i++) {
+		if ((rdl(ep, REG_MIISTS) & REG_MIISTS_BUSY) == 0)
+			break;
+		msleep(1);
+	}
+
+	if (i == 10)
+		printk(KERN_INFO DRV_MODULE_NAME ": mdio write timed out\n");
+}
+
+static void ep93xx_get_drvinfo(struct net_device *dev, struct ethtool_drvinfo *info)
+{
+	strcpy(info->driver, DRV_MODULE_NAME);
+	strcpy(info->version, DRV_MODULE_VERSION);
+}
+
+static int ep93xx_get_settings(struct net_device *dev, struct ethtool_cmd *cmd)
+{
+	struct ep93xx_priv *ep = netdev_priv(dev);
+	return mii_ethtool_gset(&ep->mii, cmd);
+}
+
+static int ep93xx_set_settings(struct net_device *dev, struct ethtool_cmd *cmd)
+{
+	struct ep93xx_priv *ep = netdev_priv(dev);
+	return mii_ethtool_sset(&ep->mii, cmd);
+}
+
+static int ep93xx_nway_reset(struct net_device *dev)
+{
+	struct ep93xx_priv *ep = netdev_priv(dev);
+	return mii_nway_restart(&ep->mii);
+}
+
+static u32 ep93xx_get_link(struct net_device *dev)
+{
+	struct ep93xx_priv *ep = netdev_priv(dev);
+	return mii_link_ok(&ep->mii);
+}
+
+static struct ethtool_ops ep93xx_ethtool_ops = {
+	.get_drvinfo		= ep93xx_get_drvinfo,
+	.get_settings		= ep93xx_get_settings,
+	.set_settings		= ep93xx_set_settings,
+	.nway_reset		= ep93xx_nway_reset,
+	.get_link		= ep93xx_get_link,
+};
+
+struct net_device *ep93xx_dev_alloc(struct ep93xx_eth_data *data)
+{
+	struct net_device *dev;
+	struct ep93xx_priv *ep;
+
+	dev = alloc_etherdev(sizeof(struct ep93xx_priv));
+	if (dev == NULL)
+		return NULL;
+	ep = netdev_priv(dev);
+
+	memcpy(dev->dev_addr, data->dev_addr, ETH_ALEN);
+
+	dev->get_stats = ep93xx_get_stats;
+	dev->ethtool_ops = &ep93xx_ethtool_ops;
+	dev->poll = ep93xx_poll;
+	dev->hard_start_xmit = ep93xx_xmit;
+	dev->open = ep93xx_open;
+	dev->stop = ep93xx_close;
+	dev->do_ioctl = ep93xx_ioctl;
+
+	dev->features |= NETIF_F_SG | NETIF_F_HW_CSUM;
+	dev->weight = 64;
+
+	return dev;
+}
+
+
+static int ep93xx_eth_remove(struct platform_device *pdev)
+{
+	struct net_device *dev;
+	struct ep93xx_priv *ep;
+
+	dev = platform_get_drvdata(pdev);
+	if (dev == NULL)
+		return 0;
+	platform_set_drvdata(pdev, NULL);
+
+	ep = netdev_priv(dev);
+
+	/* @@@ Force down.  */
+	unregister_netdev(dev);
+	ep93xx_free_buffers(ep);
+
+	if (ep->base_addr != NULL)
+		iounmap(ep->base_addr);
+
+	if (ep->res != NULL) {
+		release_resource(ep->res);
+		kfree(ep->res);
+	}
+
+	free_netdev(dev);
+
+	return 0;
+}
+
+static int ep93xx_eth_probe(struct platform_device *pdev)
+{
+	struct ep93xx_eth_data *data;
+	struct net_device *dev;
+	struct ep93xx_priv *ep;
+	int err;
+
+	data = pdev->dev.platform_data;
+	if (pdev == NULL)
+		return -ENODEV;
+
+	dev = ep93xx_dev_alloc(data);
+	if (dev == NULL) {
+		err = -ENOMEM;
+		goto err_out;
+	}
+	ep = netdev_priv(dev);
+
+	platform_set_drvdata(pdev, dev);
+
+	ep->res = request_mem_region(pdev->resource[0].start,
+			pdev->resource[0].end - pdev->resource[0].start + 1,
+			pdev->dev.bus_id);
+	if (ep->res == NULL) {
+		dev_err(&pdev->dev, "Could not reserve memory region\n");
+		err = -ENOMEM;
+		goto err_out;
+	}
+
+	ep->base_addr = ioremap(pdev->resource[0].start,
+			pdev->resource[0].end - pdev->resource[0].start);
+	if (ep->base_addr == NULL) {
+		dev_err(&pdev->dev, "Failed to ioremap ethernet registers\n");
+		err = -EIO;
+		goto err_out;
+	}
+	ep->irq = pdev->resource[1].start;
+
+	ep->mii.phy_id = data->phy_id;
+	ep->mii.phy_id_mask = 0x1f;
+	ep->mii.reg_num_mask = 0x1f;
+	ep->mii.dev = dev;
+	ep->mii.mdio_read = ep93xx_mdio_read;
+	ep->mii.mdio_write = ep93xx_mdio_write;
+	ep->mdc_divisor = 40;	/* Max HCLK 100 MHz, min MDIO clk 2.5 MHz.  */
+
+	err = register_netdev(dev);
+	if (err) {
+		dev_err(&pdev->dev, "Failed to register netdev\n");
+		goto err_out;
+	}
+
+	printk(KERN_INFO "%s: ep93xx on-chip ethernet, IRQ %d, "
+			 "%.2x:%.2x:%.2x:%.2x:%.2x:%.2x.\n", dev->name,
+			ep->irq, data->dev_addr[0], data->dev_addr[1],
+			data->dev_addr[2], data->dev_addr[3],
+			data->dev_addr[4], data->dev_addr[5]);
+
+	return 0;
+
+err_out:
+	ep93xx_eth_remove(pdev);
+	return err;
+}
+
+
+static struct platform_driver ep93xx_eth_driver = {
+	.probe		= ep93xx_eth_probe,
+	.remove		= ep93xx_eth_remove,
+	.driver		= {
+		.name	= "ep93xx-eth",
+	},
+};
+
+static int __init ep93xx_eth_init_module(void)
+{
+	printk(KERN_INFO DRV_MODULE_NAME " version " DRV_MODULE_VERSION " loading\n");
+	return platform_driver_register(&ep93xx_eth_driver);
+}
+
+static void __exit ep93xx_eth_cleanup_module(void)
+{
+	platform_driver_unregister(&ep93xx_eth_driver);
+}
+
+module_init(ep93xx_eth_init_module);
+module_exit(ep93xx_eth_cleanup_module);
+MODULE_LICENSE("GPL");
diff --git a/include/asm-arm/arch-ep93xx/ep93xx-regs.h b/include/asm-arm/arch-ep93xx/ep93xx-regs.h
index 8c322975f96e..593f562f85c3 100644
--- a/include/asm-arm/arch-ep93xx/ep93xx-regs.h
+++ b/include/asm-arm/arch-ep93xx/ep93xx-regs.h
@@ -27,6 +27,7 @@
 #define EP93XX_DMA_BASE			(EP93XX_AHB_VIRT_BASE + 0x00000000)
 
 #define EP93XX_ETHERNET_BASE		(EP93XX_AHB_VIRT_BASE + 0x00010000)
+#define EP93XX_ETHERNET_PHYS_BASE	(EP93XX_AHB_PHYS_BASE + 0x00010000)
 
 #define EP93XX_USB_BASE			(EP93XX_AHB_VIRT_BASE + 0x00020000)
 #define EP93XX_USB_PHYS_BASE		(EP93XX_AHB_PHYS_BASE + 0x00020000)
diff --git a/include/asm-arm/arch-ep93xx/platform.h b/include/asm-arm/arch-ep93xx/platform.h
index d7a34ce20293..b4a8deb8bdef 100644
--- a/include/asm-arm/arch-ep93xx/platform.h
+++ b/include/asm-arm/arch-ep93xx/platform.h
@@ -11,5 +11,11 @@ void ep93xx_init_devices(void);
 void ep93xx_clock_init(void);
 extern struct sys_timer ep93xx_timer;
 
+struct ep93xx_eth_data
+{
+	unsigned char	dev_addr[6];
+	unsigned char	phy_id;
+};
+
 
 #endif
-- 
cgit v1.2.3


From a83fbf635992442edf6aa3252e4008d4a08edf12 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@ftp.linux.org.uk>
Date: Sat, 23 Sep 2006 00:10:18 +0100
Subject: [PATCH] fix missing ifdefs in syscall classes hookup for generic
 targets

several targets have no ....at() family and m32r calls its only chown variant
chown32(), with __NR_chown being undefined.  creat(2) is also absent in some
targets.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
---
 include/asm-generic/audit_change_attr.h | 4 ++++
 include/asm-generic/audit_dir_write.h   | 4 ++++
 lib/audit.c                             | 2 ++
 3 files changed, 10 insertions(+)

(limited to 'include')

diff --git a/include/asm-generic/audit_change_attr.h b/include/asm-generic/audit_change_attr.h
index cb05bf69745a..50764550a60c 100644
--- a/include/asm-generic/audit_change_attr.h
+++ b/include/asm-generic/audit_change_attr.h
@@ -1,16 +1,20 @@
 __NR_chmod,
 __NR_fchmod,
+#ifdef __NR_chown
 __NR_chown,
 __NR_fchown,
 __NR_lchown,
+#endif
 __NR_setxattr,
 __NR_lsetxattr,
 __NR_fsetxattr,
 __NR_removexattr,
 __NR_lremovexattr,
 __NR_fremovexattr,
+#ifdef __NR_fchownat
 __NR_fchownat,
 __NR_fchmodat,
+#endif
 #ifdef __NR_chown32
 __NR_chown32,
 __NR_fchown32,
diff --git a/include/asm-generic/audit_dir_write.h b/include/asm-generic/audit_dir_write.h
index 161a7a58fbab..6621bd82cbe8 100644
--- a/include/asm-generic/audit_dir_write.h
+++ b/include/asm-generic/audit_dir_write.h
@@ -1,14 +1,18 @@
 __NR_rename,
 __NR_mkdir,
 __NR_rmdir,
+#ifdef __NR_creat
 __NR_creat,
+#endif
 __NR_link,
 __NR_unlink,
 __NR_symlink,
 __NR_mknod,
+#ifdef __NR_mkdirat
 __NR_mkdirat,
 __NR_mknodat,
 __NR_unlinkat,
 __NR_renameat,
 __NR_linkat,
 __NR_symlinkat,
+#endif
diff --git a/lib/audit.c b/lib/audit.c
index 8c21625ef938..3b1289fadf06 100644
--- a/lib/audit.c
+++ b/lib/audit.c
@@ -28,8 +28,10 @@ int audit_classify_syscall(int abi, unsigned syscall)
 	switch(syscall) {
 	case __NR_open:
 		return 2;
+#ifdef __NR_openat
 	case __NR_openat:
 		return 3;
+#endif
 #ifdef __NR_socketcall
 	case __NR_socketcall:
 		return 4;
-- 
cgit v1.2.3


From 1c3c07e9f6cc50dab2aeb8051325e317d4f6c70e Mon Sep 17 00:00:00 2001
From: Trond Myklebust <Trond.Myklebust@netapp.com>
Date: Tue, 25 Jul 2006 11:28:18 -0400
Subject: NFS: Add a new ACCESS rpc call cache to the linux nfs client

The current access cache only allows one entry at a time to be cached for each
inode. Add a per-inode red-black tree in order to allow more than one to
be cached at a time.

Should significantly cut down the time spent in path traversal for shared
directories such as ${PATH}, /usr/share, etc.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/dir.c           | 133 ++++++++++++++++++++++++++++++++++++++++++-------
 fs/nfs/inode.c         |  13 ++---
 include/linux/nfs_fs.h |   5 +-
 3 files changed, 124 insertions(+), 27 deletions(-)

(limited to 'include')

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index e7ffb4deb3e5..094afded2b11 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -1638,35 +1638,134 @@ out:
 	return error;
 }
 
-int nfs_access_get_cached(struct inode *inode, struct rpc_cred *cred, struct nfs_access_entry *res)
+static void nfs_access_free_entry(struct nfs_access_entry *entry)
+{
+	put_rpccred(entry->cred);
+	kfree(entry);
+}
+
+static void __nfs_access_zap_cache(struct inode *inode)
 {
 	struct nfs_inode *nfsi = NFS_I(inode);
-	struct nfs_access_entry *cache = &nfsi->cache_access;
+	struct rb_root *root_node = &nfsi->access_cache;
+	struct rb_node *n, *dispose = NULL;
+	struct nfs_access_entry *entry;
+
+	/* Unhook entries from the cache */
+	while ((n = rb_first(root_node)) != NULL) {
+		entry = rb_entry(n, struct nfs_access_entry, rb_node);
+		rb_erase(n, root_node);
+		n->rb_left = dispose;
+		dispose = n;
+	}
+	nfsi->cache_validity &= ~NFS_INO_INVALID_ACCESS;
+	spin_unlock(&inode->i_lock);
 
-	if (cache->cred != cred
-			|| time_after(jiffies, cache->jiffies + NFS_ATTRTIMEO(inode))
-			|| (nfsi->cache_validity & NFS_INO_INVALID_ACCESS))
-		return -ENOENT;
-	memcpy(res, cache, sizeof(*res));
-	return 0;
+	/* Now kill them all! */
+	while (dispose != NULL) {
+		n = dispose;
+		dispose = n->rb_left;
+		nfs_access_free_entry(rb_entry(n, struct nfs_access_entry, rb_node));
+	}
 }
 
-void nfs_access_add_cache(struct inode *inode, struct nfs_access_entry *set)
+void nfs_access_zap_cache(struct inode *inode)
 {
-	struct nfs_inode *nfsi = NFS_I(inode);
-	struct nfs_access_entry *cache = &nfsi->cache_access;
+	spin_lock(&inode->i_lock);
+	/* This will release the spinlock */
+	__nfs_access_zap_cache(inode);
+}
 
-	if (cache->cred != set->cred) {
-		if (cache->cred)
-			put_rpccred(cache->cred);
-		cache->cred = get_rpccred(set->cred);
+static struct nfs_access_entry *nfs_access_search_rbtree(struct inode *inode, struct rpc_cred *cred)
+{
+	struct rb_node *n = NFS_I(inode)->access_cache.rb_node;
+	struct nfs_access_entry *entry;
+
+	while (n != NULL) {
+		entry = rb_entry(n, struct nfs_access_entry, rb_node);
+
+		if (cred < entry->cred)
+			n = n->rb_left;
+		else if (cred > entry->cred)
+			n = n->rb_right;
+		else
+			return entry;
 	}
-	/* FIXME: replace current access_cache BKL reliance with inode->i_lock */
+	return NULL;
+}
+
+int nfs_access_get_cached(struct inode *inode, struct rpc_cred *cred, struct nfs_access_entry *res)
+{
+	struct nfs_inode *nfsi = NFS_I(inode);
+	struct nfs_access_entry *cache;
+	int err = -ENOENT;
+
 	spin_lock(&inode->i_lock);
-	nfsi->cache_validity &= ~NFS_INO_INVALID_ACCESS;
+	if (nfsi->cache_validity & NFS_INO_INVALID_ACCESS)
+		goto out_zap;
+	cache = nfs_access_search_rbtree(inode, cred);
+	if (cache == NULL)
+		goto out;
+	if (time_after(jiffies, cache->jiffies + NFS_ATTRTIMEO(inode)))
+		goto out_stale;
+	res->jiffies = cache->jiffies;
+	res->cred = cache->cred;
+	res->mask = cache->mask;
+	err = 0;
+out:
+	spin_unlock(&inode->i_lock);
+	return err;
+out_stale:
+	rb_erase(&cache->rb_node, &nfsi->access_cache);
+	spin_unlock(&inode->i_lock);
+	nfs_access_free_entry(cache);
+	return -ENOENT;
+out_zap:
+	/* This will release the spinlock */
+	__nfs_access_zap_cache(inode);
+	return -ENOENT;
+}
+
+static void nfs_access_add_rbtree(struct inode *inode, struct nfs_access_entry *set)
+{
+	struct rb_root *root_node = &NFS_I(inode)->access_cache;
+	struct rb_node **p = &root_node->rb_node;
+	struct rb_node *parent = NULL;
+	struct nfs_access_entry *entry;
+
+	spin_lock(&inode->i_lock);
+	while (*p != NULL) {
+		parent = *p;
+		entry = rb_entry(parent, struct nfs_access_entry, rb_node);
+
+		if (set->cred < entry->cred)
+			p = &parent->rb_left;
+		else if (set->cred > entry->cred)
+			p = &parent->rb_right;
+		else
+			goto found;
+	}
+	rb_link_node(&set->rb_node, parent, p);
+	rb_insert_color(&set->rb_node, root_node);
 	spin_unlock(&inode->i_lock);
+	return;
+found:
+	rb_replace_node(parent, &set->rb_node, root_node);
+	spin_unlock(&inode->i_lock);
+	nfs_access_free_entry(entry);
+}
+
+void nfs_access_add_cache(struct inode *inode, struct nfs_access_entry *set)
+{
+	struct nfs_access_entry *cache = kmalloc(sizeof(*cache), GFP_KERNEL);
+	if (cache == NULL)
+		return;
+	RB_CLEAR_NODE(&cache->rb_node);
 	cache->jiffies = set->jiffies;
+	cache->cred = get_rpccred(set->cred);
 	cache->mask = set->mask;
+
+	nfs_access_add_rbtree(inode, cache);
 }
 
 static int nfs_do_access(struct inode *inode, struct rpc_cred *cred, int mask)
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index d349fb2245da..b94ab060bb1e 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -76,19 +76,14 @@ int nfs_write_inode(struct inode *inode, int sync)
 
 void nfs_clear_inode(struct inode *inode)
 {
-	struct nfs_inode *nfsi = NFS_I(inode);
-	struct rpc_cred *cred;
-
 	/*
 	 * The following should never happen...
 	 */
 	BUG_ON(nfs_have_writebacks(inode));
-	BUG_ON (!list_empty(&nfsi->open_files));
+	BUG_ON(!list_empty(&NFS_I(inode)->open_files));
+	BUG_ON(atomic_read(&NFS_I(inode)->data_updates) != 0);
 	nfs_zap_acl_cache(inode);
-	cred = nfsi->cache_access.cred;
-	if (cred)
-		put_rpccred(cred);
-	BUG_ON(atomic_read(&nfsi->data_updates) != 0);
+	nfs_access_zap_cache(inode);
 }
 
 /**
@@ -290,7 +285,7 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr)
 		nfsi->attrtimeo = NFS_MINATTRTIMEO(inode);
 		nfsi->attrtimeo_timestamp = jiffies;
 		memset(nfsi->cookieverf, 0, sizeof(nfsi->cookieverf));
-		nfsi->cache_access.cred = NULL;
+		nfsi->access_cache = RB_ROOT;
 
 		unlock_new_inode(inode);
 	} else
diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h
index 6c2066caeaab..cc013ed2e52e 100644
--- a/include/linux/nfs_fs.h
+++ b/include/linux/nfs_fs.h
@@ -42,6 +42,7 @@
 #include <linux/in.h>
 #include <linux/mm.h>
 #include <linux/pagemap.h>
+#include <linux/rbtree.h>
 #include <linux/rwsem.h>
 #include <linux/wait.h>
 
@@ -69,6 +70,7 @@
  * NFSv3/v4 Access mode cache entry
  */
 struct nfs_access_entry {
+	struct rb_node		rb_node;
 	unsigned long		jiffies;
 	struct rpc_cred *	cred;
 	int			mask;
@@ -145,7 +147,7 @@ struct nfs_inode {
 	 */
 	atomic_t		data_updates;
 
-	struct nfs_access_entry	cache_access;
+	struct rb_root		access_cache;
 #ifdef CONFIG_NFS_V3_ACL
 	struct posix_acl	*acl_access;
 	struct posix_acl	*acl_default;
@@ -297,6 +299,7 @@ extern int nfs_getattr(struct vfsmount *, struct dentry *, struct kstat *);
 extern int nfs_permission(struct inode *, int, struct nameidata *);
 extern int nfs_access_get_cached(struct inode *, struct rpc_cred *, struct nfs_access_entry *);
 extern void nfs_access_add_cache(struct inode *, struct nfs_access_entry *);
+extern void nfs_access_zap_cache(struct inode *inode);
 extern int nfs_open(struct inode *, struct file *);
 extern int nfs_release(struct inode *, struct file *);
 extern int nfs_attribute_timeout(struct inode *inode);
-- 
cgit v1.2.3


From cfcea3e8c66c2dcde98d5c2693d4bff50b5cac97 Mon Sep 17 00:00:00 2001
From: Trond Myklebust <Trond.Myklebust@netapp.com>
Date: Tue, 25 Jul 2006 11:28:18 -0400
Subject: NFS: Add a global LRU list for the ACCESS cache

...in order to allow the addition of a memory shrinker.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/dir.c           | 35 ++++++++++++++++++++++++++++++++++-
 fs/nfs/inode.c         |  2 ++
 include/linux/nfs_fs.h |  4 ++++
 3 files changed, 40 insertions(+), 1 deletion(-)

(limited to 'include')

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 094afded2b11..bf4f5ffda703 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -1638,10 +1638,17 @@ out:
 	return error;
 }
 
+static DEFINE_SPINLOCK(nfs_access_lru_lock);
+static LIST_HEAD(nfs_access_lru_list);
+static atomic_long_t nfs_access_nr_entries;
+
 static void nfs_access_free_entry(struct nfs_access_entry *entry)
 {
 	put_rpccred(entry->cred);
 	kfree(entry);
+	smp_mb__before_atomic_dec();
+	atomic_long_dec(&nfs_access_nr_entries);
+	smp_mb__after_atomic_dec();
 }
 
 static void __nfs_access_zap_cache(struct inode *inode)
@@ -1655,6 +1662,7 @@ static void __nfs_access_zap_cache(struct inode *inode)
 	while ((n = rb_first(root_node)) != NULL) {
 		entry = rb_entry(n, struct nfs_access_entry, rb_node);
 		rb_erase(n, root_node);
+		list_del(&entry->lru);
 		n->rb_left = dispose;
 		dispose = n;
 	}
@@ -1671,6 +1679,13 @@ static void __nfs_access_zap_cache(struct inode *inode)
 
 void nfs_access_zap_cache(struct inode *inode)
 {
+	/* Remove from global LRU init */
+	if (test_and_clear_bit(NFS_INO_ACL_LRU_SET, &NFS_FLAGS(inode))) {
+		spin_lock(&nfs_access_lru_lock);
+		list_del_init(&NFS_I(inode)->access_cache_inode_lru);
+		spin_unlock(&nfs_access_lru_lock);
+	}
+
 	spin_lock(&inode->i_lock);
 	/* This will release the spinlock */
 	__nfs_access_zap_cache(inode);
@@ -1711,12 +1726,14 @@ int nfs_access_get_cached(struct inode *inode, struct rpc_cred *cred, struct nfs
 	res->jiffies = cache->jiffies;
 	res->cred = cache->cred;
 	res->mask = cache->mask;
+	list_move_tail(&cache->lru, &nfsi->access_cache_entry_lru);
 	err = 0;
 out:
 	spin_unlock(&inode->i_lock);
 	return err;
 out_stale:
 	rb_erase(&cache->rb_node, &nfsi->access_cache);
+	list_del(&cache->lru);
 	spin_unlock(&inode->i_lock);
 	nfs_access_free_entry(cache);
 	return -ENOENT;
@@ -1728,7 +1745,8 @@ out_zap:
 
 static void nfs_access_add_rbtree(struct inode *inode, struct nfs_access_entry *set)
 {
-	struct rb_root *root_node = &NFS_I(inode)->access_cache;
+	struct nfs_inode *nfsi = NFS_I(inode);
+	struct rb_root *root_node = &nfsi->access_cache;
 	struct rb_node **p = &root_node->rb_node;
 	struct rb_node *parent = NULL;
 	struct nfs_access_entry *entry;
@@ -1747,10 +1765,13 @@ static void nfs_access_add_rbtree(struct inode *inode, struct nfs_access_entry *
 	}
 	rb_link_node(&set->rb_node, parent, p);
 	rb_insert_color(&set->rb_node, root_node);
+	list_add_tail(&set->lru, &nfsi->access_cache_entry_lru);
 	spin_unlock(&inode->i_lock);
 	return;
 found:
 	rb_replace_node(parent, &set->rb_node, root_node);
+	list_add_tail(&set->lru, &nfsi->access_cache_entry_lru);
+	list_del(&entry->lru);
 	spin_unlock(&inode->i_lock);
 	nfs_access_free_entry(entry);
 }
@@ -1766,6 +1787,18 @@ void nfs_access_add_cache(struct inode *inode, struct nfs_access_entry *set)
 	cache->mask = set->mask;
 
 	nfs_access_add_rbtree(inode, cache);
+
+	/* Update accounting */
+	smp_mb__before_atomic_inc();
+	atomic_long_inc(&nfs_access_nr_entries);
+	smp_mb__after_atomic_inc();
+
+	/* Add inode to global LRU list */
+	if (!test_and_set_bit(NFS_INO_ACL_LRU_SET, &NFS_FLAGS(inode))) {
+		spin_lock(&nfs_access_lru_lock);
+		list_add_tail(&NFS_I(inode)->access_cache_inode_lru, &nfs_access_lru_list);
+		spin_unlock(&nfs_access_lru_lock);
+	}
 }
 
 static int nfs_do_access(struct inode *inode, struct rpc_cred *cred, int mask)
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index b94ab060bb1e..6ed018c9aad2 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -1104,6 +1104,8 @@ static void init_once(void * foo, kmem_cache_t * cachep, unsigned long flags)
 		INIT_LIST_HEAD(&nfsi->dirty);
 		INIT_LIST_HEAD(&nfsi->commit);
 		INIT_LIST_HEAD(&nfsi->open_files);
+		INIT_LIST_HEAD(&nfsi->access_cache_entry_lru);
+		INIT_LIST_HEAD(&nfsi->access_cache_inode_lru);
 		INIT_RADIX_TREE(&nfsi->nfs_page_tree, GFP_ATOMIC);
 		atomic_set(&nfsi->data_updates, 0);
 		nfsi->ndirty = 0;
diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h
index cc013ed2e52e..a36e01cd6321 100644
--- a/include/linux/nfs_fs.h
+++ b/include/linux/nfs_fs.h
@@ -71,6 +71,7 @@
  */
 struct nfs_access_entry {
 	struct rb_node		rb_node;
+	struct list_head	lru;
 	unsigned long		jiffies;
 	struct rpc_cred *	cred;
 	int			mask;
@@ -148,6 +149,8 @@ struct nfs_inode {
 	atomic_t		data_updates;
 
 	struct rb_root		access_cache;
+	struct list_head	access_cache_entry_lru;
+	struct list_head	access_cache_inode_lru;
 #ifdef CONFIG_NFS_V3_ACL
 	struct posix_acl	*acl_access;
 	struct posix_acl	*acl_default;
@@ -201,6 +204,7 @@ struct nfs_inode {
 #define NFS_INO_REVALIDATING	(0)		/* revalidating attrs */
 #define NFS_INO_ADVISE_RDPLUS	(1)		/* advise readdirplus */
 #define NFS_INO_STALE		(2)		/* possible stale inode */
+#define NFS_INO_ACL_LRU_SET	(3)		/* Inode is on the LRU list */
 
 static inline struct nfs_inode *NFS_I(struct inode *inode)
 {
-- 
cgit v1.2.3


From 770bfad846ab6628444428467b11fa6773ae9ea1 Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Tue, 22 Aug 2006 20:06:07 -0400
Subject: NFS: Add dentry materialisation op

The attached patch adds a new directory cache management function that prepares
a disconnected anonymous function to be connected into the dentry tree. The
anonymous dentry is transferred the name and parentage from another dentry.

The following changes were made in [try #2]:

 (*) d_materialise_dentry() now switches the parentage of the two nodes around
     correctly when one or other of them is self-referential.

The following changes were made in [try #7]:

 (*) d_instantiate_unique() has had the interior part split out as function
     __d_instantiate_unique(). Callers of this latter function must be holding
     the appropriate locks.

 (*) _d_rehash() has been added as a wrapper around __d_rehash() to call it
     with the most obvious hash list (the one from the name). d_rehash() now
     calls _d_rehash().

 (*) d_materialise_dentry() is now __d_materialise_dentry() and is static.

 (*) d_materialise_unique() added to perform the combination of d_find_alias(),
     d_materialise_dentry() and d_add_unique() that the NFS client was doing
     twice, all within a single dcache_lock critical section. This reduces the
     number of times two different spinlocks were being accessed.

The following further changes were made:

 (*) Add the dentries onto their parents d_subdirs lists.

Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/dcache.c            | 164 ++++++++++++++++++++++++++++++++++++++++++++-----
 include/linux/dcache.h |   1 +
 2 files changed, 151 insertions(+), 14 deletions(-)

(limited to 'include')

diff --git a/fs/dcache.c b/fs/dcache.c
index 1b4a3a34ec57..17b392a2049e 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -828,17 +828,19 @@ void d_instantiate(struct dentry *entry, struct inode * inode)
  * (or otherwise set) by the caller to indicate that it is now
  * in use by the dcache.
  */
-struct dentry *d_instantiate_unique(struct dentry *entry, struct inode *inode)
+static struct dentry *__d_instantiate_unique(struct dentry *entry,
+					     struct inode *inode)
 {
 	struct dentry *alias;
 	int len = entry->d_name.len;
 	const char *name = entry->d_name.name;
 	unsigned int hash = entry->d_name.hash;
 
-	BUG_ON(!list_empty(&entry->d_alias));
-	spin_lock(&dcache_lock);
-	if (!inode)
-		goto do_negative;
+	if (!inode) {
+		entry->d_inode = NULL;
+		return NULL;
+	}
+
 	list_for_each_entry(alias, &inode->i_dentry, d_alias) {
 		struct qstr *qstr = &alias->d_name;
 
@@ -851,19 +853,35 @@ struct dentry *d_instantiate_unique(struct dentry *entry, struct inode *inode)
 		if (memcmp(qstr->name, name, len))
 			continue;
 		dget_locked(alias);
-		spin_unlock(&dcache_lock);
-		BUG_ON(!d_unhashed(alias));
-		iput(inode);
 		return alias;
 	}
+
 	list_add(&entry->d_alias, &inode->i_dentry);
-do_negative:
 	entry->d_inode = inode;
 	fsnotify_d_instantiate(entry, inode);
-	spin_unlock(&dcache_lock);
-	security_d_instantiate(entry, inode);
 	return NULL;
 }
+
+struct dentry *d_instantiate_unique(struct dentry *entry, struct inode *inode)
+{
+	struct dentry *result;
+
+	BUG_ON(!list_empty(&entry->d_alias));
+
+	spin_lock(&dcache_lock);
+	result = __d_instantiate_unique(entry, inode);
+	spin_unlock(&dcache_lock);
+
+	if (!result) {
+		security_d_instantiate(entry, inode);
+		return NULL;
+	}
+
+	BUG_ON(!d_unhashed(result));
+	iput(inode);
+	return result;
+}
+
 EXPORT_SYMBOL(d_instantiate_unique);
 
 /**
@@ -1235,6 +1253,11 @@ static void __d_rehash(struct dentry * entry, struct hlist_head *list)
  	hlist_add_head_rcu(&entry->d_hash, list);
 }
 
+static void _d_rehash(struct dentry * entry)
+{
+	__d_rehash(entry, d_hash(entry->d_parent, entry->d_name.hash));
+}
+
 /**
  * d_rehash	- add an entry back to the hash
  * @entry: dentry to add to the hash
@@ -1244,11 +1267,9 @@ static void __d_rehash(struct dentry * entry, struct hlist_head *list)
  
 void d_rehash(struct dentry * entry)
 {
-	struct hlist_head *list = d_hash(entry->d_parent, entry->d_name.hash);
-
 	spin_lock(&dcache_lock);
 	spin_lock(&entry->d_lock);
-	__d_rehash(entry, list);
+	_d_rehash(entry);
 	spin_unlock(&entry->d_lock);
 	spin_unlock(&dcache_lock);
 }
@@ -1386,6 +1407,120 @@ already_unhashed:
 	spin_unlock(&dcache_lock);
 }
 
+/*
+ * Prepare an anonymous dentry for life in the superblock's dentry tree as a
+ * named dentry in place of the dentry to be replaced.
+ */
+static void __d_materialise_dentry(struct dentry *dentry, struct dentry *anon)
+{
+	struct dentry *dparent, *aparent;
+
+	switch_names(dentry, anon);
+	do_switch(dentry->d_name.len, anon->d_name.len);
+	do_switch(dentry->d_name.hash, anon->d_name.hash);
+
+	dparent = dentry->d_parent;
+	aparent = anon->d_parent;
+
+	dentry->d_parent = (aparent == anon) ? dentry : aparent;
+	list_del(&dentry->d_u.d_child);
+	if (!IS_ROOT(dentry))
+		list_add(&dentry->d_u.d_child, &dentry->d_parent->d_subdirs);
+	else
+		INIT_LIST_HEAD(&dentry->d_u.d_child);
+
+	anon->d_parent = (dparent == dentry) ? anon : dparent;
+	list_del(&anon->d_u.d_child);
+	if (!IS_ROOT(anon))
+		list_add(&anon->d_u.d_child, &anon->d_parent->d_subdirs);
+	else
+		INIT_LIST_HEAD(&anon->d_u.d_child);
+
+	anon->d_flags &= ~DCACHE_DISCONNECTED;
+}
+
+/**
+ * d_materialise_unique - introduce an inode into the tree
+ * @dentry: candidate dentry
+ * @inode: inode to bind to the dentry, to which aliases may be attached
+ *
+ * Introduces an dentry into the tree, substituting an extant disconnected
+ * root directory alias in its place if there is one
+ */
+struct dentry *d_materialise_unique(struct dentry *dentry, struct inode *inode)
+{
+	struct dentry *alias, *actual;
+
+	BUG_ON(!d_unhashed(dentry));
+
+	spin_lock(&dcache_lock);
+
+	if (!inode) {
+		actual = dentry;
+		dentry->d_inode = NULL;
+		goto found_lock;
+	}
+
+	/* See if a disconnected directory already exists as an anonymous root
+	 * that we should splice into the tree instead */
+	if (S_ISDIR(inode->i_mode) && (alias = __d_find_alias(inode, 1))) {
+		spin_lock(&alias->d_lock);
+
+		/* Is this a mountpoint that we could splice into our tree? */
+		if (IS_ROOT(alias))
+			goto connect_mountpoint;
+
+		if (alias->d_name.len == dentry->d_name.len &&
+		    alias->d_parent == dentry->d_parent &&
+		    memcmp(alias->d_name.name,
+			   dentry->d_name.name,
+			   dentry->d_name.len) == 0)
+			goto replace_with_alias;
+
+		spin_unlock(&alias->d_lock);
+
+		/* Doh! Seem to be aliasing directories for some reason... */
+		dput(alias);
+	}
+
+	/* Add a unique reference */
+	actual = __d_instantiate_unique(dentry, inode);
+	if (!actual)
+		actual = dentry;
+	else if (unlikely(!d_unhashed(actual)))
+		goto shouldnt_be_hashed;
+
+found_lock:
+	spin_lock(&actual->d_lock);
+found:
+	_d_rehash(actual);
+	spin_unlock(&actual->d_lock);
+	spin_unlock(&dcache_lock);
+
+	if (actual == dentry) {
+		security_d_instantiate(dentry, inode);
+		return NULL;
+	}
+
+	iput(inode);
+	return actual;
+
+	/* Convert the anonymous/root alias into an ordinary dentry */
+connect_mountpoint:
+	__d_materialise_dentry(dentry, alias);
+
+	/* Replace the candidate dentry with the alias in the tree */
+replace_with_alias:
+	__d_drop(alias);
+	actual = alias;
+	goto found;
+
+shouldnt_be_hashed:
+	spin_unlock(&dcache_lock);
+	BUG();
+	goto shouldnt_be_hashed;
+}
+
 /**
  * d_path - return the path of a dentry
  * @dentry: dentry to report
@@ -1784,6 +1919,7 @@ EXPORT_SYMBOL(d_instantiate);
 EXPORT_SYMBOL(d_invalidate);
 EXPORT_SYMBOL(d_lookup);
 EXPORT_SYMBOL(d_move);
+EXPORT_SYMBOL_GPL(d_materialise_unique);
 EXPORT_SYMBOL(d_path);
 EXPORT_SYMBOL(d_prune_aliases);
 EXPORT_SYMBOL(d_rehash);
diff --git a/include/linux/dcache.h b/include/linux/dcache.h
index 471781ffeab1..44605be59409 100644
--- a/include/linux/dcache.h
+++ b/include/linux/dcache.h
@@ -221,6 +221,7 @@ static inline int dname_external(struct dentry *dentry)
  */
 extern void d_instantiate(struct dentry *, struct inode *);
 extern struct dentry * d_instantiate_unique(struct dentry *, struct inode *);
+extern struct dentry * d_materialise_unique(struct dentry *, struct inode *);
 extern void d_delete(struct dentry *);
 
 /* allocate/de-allocate */
-- 
cgit v1.2.3


From adfa6f980bd46974e6b32b22dd0c45e3f52063f4 Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Tue, 22 Aug 2006 20:06:08 -0400
Subject: NFS: Rename struct nfs4_client to struct nfs_client

Rename struct nfs4_client to struct nfs_client so that it can become the basis
for a general client record for NFS2 and NFS3 in addition to NFS4.

Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/callback.c         |  2 +-
 fs/nfs/callback_proc.c    |  4 ++--
 fs/nfs/delegation.c       | 24 +++++++++++------------
 fs/nfs/delegation.h       | 10 +++++-----
 fs/nfs/idmap.c            | 12 ++++++------
 fs/nfs/nfs4_fs.h          | 30 ++++++++++++++--------------
 fs/nfs/nfs4proc.c         | 32 +++++++++++++++---------------
 fs/nfs/nfs4renewd.c       |  8 ++++----
 fs/nfs/nfs4state.c        | 50 +++++++++++++++++++++++------------------------
 fs/nfs/nfs4xdr.c          | 18 ++++++++---------
 fs/nfs/super.c            |  4 ++--
 include/linux/nfs_fs_sb.h |  2 +-
 include/linux/nfs_idmap.h | 14 ++++++-------
 13 files changed, 105 insertions(+), 105 deletions(-)

(limited to 'include')

diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c
index b1f7dc415392..1b596b6d9dc2 100644
--- a/fs/nfs/callback.c
+++ b/fs/nfs/callback.c
@@ -167,7 +167,7 @@ void nfs_callback_down(void)
 static int nfs_callback_authenticate(struct svc_rqst *rqstp)
 {
 	struct in_addr *addr = &rqstp->rq_addr.sin_addr;
-	struct nfs4_client *clp;
+	struct nfs_client *clp;
 
 	/* Don't talk to strangers */
 	clp = nfs4_find_client(addr);
diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c
index 7719483ecdfc..55d6e2ec157f 100644
--- a/fs/nfs/callback_proc.c
+++ b/fs/nfs/callback_proc.c
@@ -15,7 +15,7 @@
  
 unsigned nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres *res)
 {
-	struct nfs4_client *clp;
+	struct nfs_client *clp;
 	struct nfs_delegation *delegation;
 	struct nfs_inode *nfsi;
 	struct inode *inode;
@@ -56,7 +56,7 @@ out:
 
 unsigned nfs4_callback_recall(struct cb_recallargs *args, void *dummy)
 {
-	struct nfs4_client *clp;
+	struct nfs_client *clp;
 	struct inode *inode;
 	unsigned res;
 	
diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c
index 9540a316c05e..5a1105c258bd 100644
--- a/fs/nfs/delegation.c
+++ b/fs/nfs/delegation.c
@@ -114,7 +114,7 @@ void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred, st
  */
 int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res)
 {
-	struct nfs4_client *clp = NFS_SERVER(inode)->nfs4_state;
+	struct nfs_client *clp = NFS_SERVER(inode)->nfs4_state;
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_delegation *delegation;
 	int status = 0;
@@ -176,7 +176,7 @@ static void nfs_msync_inode(struct inode *inode)
  */
 int __nfs_inode_return_delegation(struct inode *inode)
 {
-	struct nfs4_client *clp = NFS_SERVER(inode)->nfs4_state;
+	struct nfs_client *clp = NFS_SERVER(inode)->nfs4_state;
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_delegation *delegation;
 	int res = 0;
@@ -208,7 +208,7 @@ int __nfs_inode_return_delegation(struct inode *inode)
  */
 void nfs_return_all_delegations(struct super_block *sb)
 {
-	struct nfs4_client *clp = NFS_SB(sb)->nfs4_state;
+	struct nfs_client *clp = NFS_SB(sb)->nfs4_state;
 	struct nfs_delegation *delegation;
 	struct inode *inode;
 
@@ -232,7 +232,7 @@ restart:
 
 int nfs_do_expire_all_delegations(void *ptr)
 {
-	struct nfs4_client *clp = ptr;
+	struct nfs_client *clp = ptr;
 	struct nfs_delegation *delegation;
 	struct inode *inode;
 
@@ -258,7 +258,7 @@ out:
 	module_put_and_exit(0);
 }
 
-void nfs_expire_all_delegations(struct nfs4_client *clp)
+void nfs_expire_all_delegations(struct nfs_client *clp)
 {
 	struct task_struct *task;
 
@@ -276,7 +276,7 @@ void nfs_expire_all_delegations(struct nfs4_client *clp)
 /*
  * Return all delegations following an NFS4ERR_CB_PATH_DOWN error.
  */
-void nfs_handle_cb_pathdown(struct nfs4_client *clp)
+void nfs_handle_cb_pathdown(struct nfs_client *clp)
 {
 	struct nfs_delegation *delegation;
 	struct inode *inode;
@@ -299,7 +299,7 @@ restart:
 
 struct recall_threadargs {
 	struct inode *inode;
-	struct nfs4_client *clp;
+	struct nfs_client *clp;
 	const nfs4_stateid *stateid;
 
 	struct completion started;
@@ -310,7 +310,7 @@ static int recall_thread(void *data)
 {
 	struct recall_threadargs *args = (struct recall_threadargs *)data;
 	struct inode *inode = igrab(args->inode);
-	struct nfs4_client *clp = NFS_SERVER(inode)->nfs4_state;
+	struct nfs_client *clp = NFS_SERVER(inode)->nfs4_state;
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_delegation *delegation;
 
@@ -371,7 +371,7 @@ out_module_put:
 /*
  * Retrieve the inode associated with a delegation
  */
-struct inode *nfs_delegation_find_inode(struct nfs4_client *clp, const struct nfs_fh *fhandle)
+struct inode *nfs_delegation_find_inode(struct nfs_client *clp, const struct nfs_fh *fhandle)
 {
 	struct nfs_delegation *delegation;
 	struct inode *res = NULL;
@@ -389,7 +389,7 @@ struct inode *nfs_delegation_find_inode(struct nfs4_client *clp, const struct nf
 /*
  * Mark all delegations as needing to be reclaimed
  */
-void nfs_delegation_mark_reclaim(struct nfs4_client *clp)
+void nfs_delegation_mark_reclaim(struct nfs_client *clp)
 {
 	struct nfs_delegation *delegation;
 	spin_lock(&clp->cl_lock);
@@ -401,7 +401,7 @@ void nfs_delegation_mark_reclaim(struct nfs4_client *clp)
 /*
  * Reap all unclaimed delegations after reboot recovery is done
  */
-void nfs_delegation_reap_unclaimed(struct nfs4_client *clp)
+void nfs_delegation_reap_unclaimed(struct nfs_client *clp)
 {
 	struct nfs_delegation *delegation, *n;
 	LIST_HEAD(head);
@@ -423,7 +423,7 @@ void nfs_delegation_reap_unclaimed(struct nfs4_client *clp)
 
 int nfs4_copy_delegation_stateid(nfs4_stateid *dst, struct inode *inode)
 {
-	struct nfs4_client *clp = NFS_SERVER(inode)->nfs4_state;
+	struct nfs_client *clp = NFS_SERVER(inode)->nfs4_state;
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_delegation *delegation;
 	int res = 0;
diff --git a/fs/nfs/delegation.h b/fs/nfs/delegation.h
index 3858694652fa..2cfd4b24c7fe 100644
--- a/fs/nfs/delegation.h
+++ b/fs/nfs/delegation.h
@@ -29,13 +29,13 @@ void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred, st
 int __nfs_inode_return_delegation(struct inode *inode);
 int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *stateid);
 
-struct inode *nfs_delegation_find_inode(struct nfs4_client *clp, const struct nfs_fh *fhandle);
+struct inode *nfs_delegation_find_inode(struct nfs_client *clp, const struct nfs_fh *fhandle);
 void nfs_return_all_delegations(struct super_block *sb);
-void nfs_expire_all_delegations(struct nfs4_client *clp);
-void nfs_handle_cb_pathdown(struct nfs4_client *clp);
+void nfs_expire_all_delegations(struct nfs_client *clp);
+void nfs_handle_cb_pathdown(struct nfs_client *clp);
 
-void nfs_delegation_mark_reclaim(struct nfs4_client *clp);
-void nfs_delegation_reap_unclaimed(struct nfs4_client *clp);
+void nfs_delegation_mark_reclaim(struct nfs_client *clp);
+void nfs_delegation_reap_unclaimed(struct nfs_client *clp);
 
 /* NFSv4 delegation-related procedures */
 int nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, const nfs4_stateid *stateid);
diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c
index 873deb96a6cd..d05148ec9414 100644
--- a/fs/nfs/idmap.c
+++ b/fs/nfs/idmap.c
@@ -109,7 +109,7 @@ static struct rpc_pipe_ops idmap_upcall_ops = {
 };
 
 void
-nfs_idmap_new(struct nfs4_client *clp)
+nfs_idmap_new(struct nfs_client *clp)
 {
 	struct idmap *idmap;
 
@@ -138,7 +138,7 @@ nfs_idmap_new(struct nfs4_client *clp)
 }
 
 void
-nfs_idmap_delete(struct nfs4_client *clp)
+nfs_idmap_delete(struct nfs_client *clp)
 {
 	struct idmap *idmap = clp->cl_idmap;
 
@@ -491,27 +491,27 @@ static unsigned int fnvhash32(const void *buf, size_t buflen)
 	return (hash);
 }
 
-int nfs_map_name_to_uid(struct nfs4_client *clp, const char *name, size_t namelen, __u32 *uid)
+int nfs_map_name_to_uid(struct nfs_client *clp, const char *name, size_t namelen, __u32 *uid)
 {
 	struct idmap *idmap = clp->cl_idmap;
 
 	return nfs_idmap_id(idmap, &idmap->idmap_user_hash, name, namelen, uid);
 }
 
-int nfs_map_group_to_gid(struct nfs4_client *clp, const char *name, size_t namelen, __u32 *uid)
+int nfs_map_group_to_gid(struct nfs_client *clp, const char *name, size_t namelen, __u32 *uid)
 {
 	struct idmap *idmap = clp->cl_idmap;
 
 	return nfs_idmap_id(idmap, &idmap->idmap_group_hash, name, namelen, uid);
 }
 
-int nfs_map_uid_to_name(struct nfs4_client *clp, __u32 uid, char *buf)
+int nfs_map_uid_to_name(struct nfs_client *clp, __u32 uid, char *buf)
 {
 	struct idmap *idmap = clp->cl_idmap;
 
 	return nfs_idmap_name(idmap, &idmap->idmap_user_hash, uid, buf);
 }
-int nfs_map_gid_to_group(struct nfs4_client *clp, __u32 uid, char *buf)
+int nfs_map_gid_to_group(struct nfs_client *clp, __u32 uid, char *buf)
 {
 	struct idmap *idmap = clp->cl_idmap;
 
diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h
index 9a102860df37..4e334cb48498 100644
--- a/fs/nfs/nfs4_fs.h
+++ b/fs/nfs/nfs4_fs.h
@@ -43,9 +43,9 @@ enum nfs4_client_state {
 };
 
 /*
- * The nfs4_client identifies our client state to the server.
+ * The nfs_client identifies our client state to the server.
  */
-struct nfs4_client {
+struct nfs_client {
 	struct list_head	cl_servers;	/* Global list of servers */
 	struct in_addr		cl_addr;	/* Server identifier */
 	u64			cl_clientid;	/* constant */
@@ -127,7 +127,7 @@ static inline void nfs_confirm_seqid(struct nfs_seqid_counter *seqid, int status
 struct nfs4_state_owner {
 	spinlock_t	     so_lock;
 	struct list_head     so_list;	 /* per-clientid list of state_owners */
-	struct nfs4_client   *so_client;
+	struct nfs_client    *so_client;
 	u32                  so_id;      /* 32-bit identifier, unique */
 	atomic_t	     so_count;
 
@@ -210,10 +210,10 @@ extern ssize_t nfs4_listxattr(struct dentry *, char *, size_t);
 
 /* nfs4proc.c */
 extern int nfs4_map_errors(int err);
-extern int nfs4_proc_setclientid(struct nfs4_client *, u32, unsigned short, struct rpc_cred *);
-extern int nfs4_proc_setclientid_confirm(struct nfs4_client *, struct rpc_cred *);
-extern int nfs4_proc_async_renew(struct nfs4_client *, struct rpc_cred *);
-extern int nfs4_proc_renew(struct nfs4_client *, struct rpc_cred *);
+extern int nfs4_proc_setclientid(struct nfs_client *, u32, unsigned short, struct rpc_cred *);
+extern int nfs4_proc_setclientid_confirm(struct nfs_client *, struct rpc_cred *);
+extern int nfs4_proc_async_renew(struct nfs_client *, struct rpc_cred *);
+extern int nfs4_proc_renew(struct nfs_client *, struct rpc_cred *);
 extern int nfs4_do_close(struct inode *inode, struct nfs4_state *state);
 extern struct dentry *nfs4_atomic_open(struct inode *, struct dentry *, struct nameidata *);
 extern int nfs4_open_revalidate(struct inode *, struct dentry *, int, struct nameidata *);
@@ -231,19 +231,19 @@ extern const u32 nfs4_fsinfo_bitmap[2];
 extern const u32 nfs4_fs_locations_bitmap[2];
 
 /* nfs4renewd.c */
-extern void nfs4_schedule_state_renewal(struct nfs4_client *);
+extern void nfs4_schedule_state_renewal(struct nfs_client *);
 extern void nfs4_renewd_prepare_shutdown(struct nfs_server *);
-extern void nfs4_kill_renewd(struct nfs4_client *);
+extern void nfs4_kill_renewd(struct nfs_client *);
 extern void nfs4_renew_state(void *);
 
 /* nfs4state.c */
 extern void init_nfsv4_state(struct nfs_server *);
 extern void destroy_nfsv4_state(struct nfs_server *);
-extern struct nfs4_client *nfs4_get_client(struct in_addr *);
-extern void nfs4_put_client(struct nfs4_client *clp);
-extern struct nfs4_client *nfs4_find_client(struct in_addr *);
-struct rpc_cred *nfs4_get_renew_cred(struct nfs4_client *clp);
-extern u32 nfs4_alloc_lockowner_id(struct nfs4_client *);
+extern struct nfs_client *nfs4_get_client(struct in_addr *);
+extern void nfs4_put_client(struct nfs_client *clp);
+extern struct nfs_client *nfs4_find_client(struct in_addr *);
+struct rpc_cred *nfs4_get_renew_cred(struct nfs_client *clp);
+extern u32 nfs4_alloc_lockowner_id(struct nfs_client *);
 
 extern struct nfs4_state_owner * nfs4_get_state_owner(struct nfs_server *, struct rpc_cred *);
 extern void nfs4_put_state_owner(struct nfs4_state_owner *);
@@ -252,7 +252,7 @@ extern struct nfs4_state * nfs4_get_open_state(struct inode *, struct nfs4_state
 extern void nfs4_put_open_state(struct nfs4_state *);
 extern void nfs4_close_state(struct nfs4_state *, mode_t);
 extern void nfs4_state_set_mode_locked(struct nfs4_state *, mode_t);
-extern void nfs4_schedule_state_recovery(struct nfs4_client *);
+extern void nfs4_schedule_state_recovery(struct nfs_client *);
 extern void nfs4_put_lock_state(struct nfs4_lock_state *lsp);
 extern int nfs4_set_lock_state(struct nfs4_state *state, struct file_lock *fl);
 extern void nfs4_copy_stateid(nfs4_stateid *, struct nfs4_state *, fl_owner_t);
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index b14145b7b87f..168f3ffb059f 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -64,7 +64,7 @@ static int nfs4_do_fsinfo(struct nfs_server *, struct nfs_fh *, struct nfs_fsinf
 static int nfs4_async_handle_error(struct rpc_task *, const struct nfs_server *);
 static int _nfs4_proc_access(struct inode *inode, struct nfs_access_entry *entry);
 static int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct nfs4_exception *exception);
-static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs4_client *clp);
+static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs_client *clp);
 
 /* Prevent leaks of NFSv4 errors into userland */
 int nfs4_map_errors(int err)
@@ -195,7 +195,7 @@ static void nfs4_setup_readdir(u64 cookie, u32 *verifier, struct dentry *dentry,
 
 static void renew_lease(const struct nfs_server *server, unsigned long timestamp)
 {
-	struct nfs4_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs4_state;
 	spin_lock(&clp->cl_lock);
 	if (time_before(clp->cl_last_renewal,timestamp))
 		clp->cl_last_renewal = timestamp;
@@ -792,7 +792,7 @@ out:
 
 int nfs4_recover_expired_lease(struct nfs_server *server)
 {
-	struct nfs4_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs4_state;
 
 	if (test_and_clear_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state))
 		nfs4_schedule_state_recovery(clp);
@@ -867,7 +867,7 @@ static int _nfs4_open_delegated(struct inode *inode, int flags, struct rpc_cred
 {
 	struct nfs_delegation *delegation;
 	struct nfs_server *server = NFS_SERVER(inode);
-	struct nfs4_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs4_state;
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs4_state_owner *sp = NULL;
 	struct nfs4_state *state = NULL;
@@ -953,7 +953,7 @@ static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, st
 	struct nfs4_state_owner  *sp;
 	struct nfs4_state     *state = NULL;
 	struct nfs_server       *server = NFS_SERVER(dir);
-	struct nfs4_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs4_state;
 	struct nfs4_opendata *opendata;
 	int                     status;
 
@@ -2521,7 +2521,7 @@ static void nfs4_proc_commit_setup(struct nfs_write_data *data, int how)
  */
 static void nfs4_renew_done(struct rpc_task *task, void *data)
 {
-	struct nfs4_client *clp = (struct nfs4_client *)task->tk_msg.rpc_argp;
+	struct nfs_client *clp = (struct nfs_client *)task->tk_msg.rpc_argp;
 	unsigned long timestamp = (unsigned long)data;
 
 	if (task->tk_status < 0) {
@@ -2543,7 +2543,7 @@ static const struct rpc_call_ops nfs4_renew_ops = {
 	.rpc_call_done = nfs4_renew_done,
 };
 
-int nfs4_proc_async_renew(struct nfs4_client *clp, struct rpc_cred *cred)
+int nfs4_proc_async_renew(struct nfs_client *clp, struct rpc_cred *cred)
 {
 	struct rpc_message msg = {
 		.rpc_proc	= &nfs4_procedures[NFSPROC4_CLNT_RENEW],
@@ -2555,7 +2555,7 @@ int nfs4_proc_async_renew(struct nfs4_client *clp, struct rpc_cred *cred)
 			&nfs4_renew_ops, (void *)jiffies);
 }
 
-int nfs4_proc_renew(struct nfs4_client *clp, struct rpc_cred *cred)
+int nfs4_proc_renew(struct nfs_client *clp, struct rpc_cred *cred)
 {
 	struct rpc_message msg = {
 		.rpc_proc	= &nfs4_procedures[NFSPROC4_CLNT_RENEW],
@@ -2791,7 +2791,7 @@ static int nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t buflen
 static int
 nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server)
 {
-	struct nfs4_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs4_state;
 
 	if (!clp || task->tk_status >= 0)
 		return 0;
@@ -2828,7 +2828,7 @@ static int nfs4_wait_bit_interruptible(void *word)
 	return 0;
 }
 
-static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs4_client *clp)
+static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs_client *clp)
 {
 	sigset_t oldset;
 	int res;
@@ -2871,7 +2871,7 @@ static int nfs4_delay(struct rpc_clnt *clnt, long *timeout)
  */
 int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct nfs4_exception *exception)
 {
-	struct nfs4_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs4_state;
 	int ret = errorcode;
 
 	exception->retry = 0;
@@ -2898,7 +2898,7 @@ int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct
 	return nfs4_map_errors(ret);
 }
 
-int nfs4_proc_setclientid(struct nfs4_client *clp, u32 program, unsigned short port, struct rpc_cred *cred)
+int nfs4_proc_setclientid(struct nfs_client *clp, u32 program, unsigned short port, struct rpc_cred *cred)
 {
 	nfs4_verifier sc_verifier;
 	struct nfs4_setclientid setclientid = {
@@ -2945,7 +2945,7 @@ int nfs4_proc_setclientid(struct nfs4_client *clp, u32 program, unsigned short p
 	return status;
 }
 
-static int _nfs4_proc_setclientid_confirm(struct nfs4_client *clp, struct rpc_cred *cred)
+static int _nfs4_proc_setclientid_confirm(struct nfs_client *clp, struct rpc_cred *cred)
 {
 	struct nfs_fsinfo fsinfo;
 	struct rpc_message msg = {
@@ -2969,7 +2969,7 @@ static int _nfs4_proc_setclientid_confirm(struct nfs4_client *clp, struct rpc_cr
 	return status;
 }
 
-int nfs4_proc_setclientid_confirm(struct nfs4_client *clp, struct rpc_cred *cred)
+int nfs4_proc_setclientid_confirm(struct nfs_client *clp, struct rpc_cred *cred)
 {
 	long timeout;
 	int err;
@@ -3106,7 +3106,7 @@ static int _nfs4_proc_getlk(struct nfs4_state *state, int cmd, struct file_lock
 {
 	struct inode *inode = state->inode;
 	struct nfs_server *server = NFS_SERVER(inode);
-	struct nfs4_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs4_state;
 	struct nfs_lockt_args arg = {
 		.fh = NFS_FH(inode),
 		.fl = request,
@@ -3513,7 +3513,7 @@ static int nfs4_lock_expired(struct nfs4_state *state, struct file_lock *request
 
 static int _nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock *request)
 {
-	struct nfs4_client *clp = state->owner->so_client;
+	struct nfs_client *clp = state->owner->so_client;
 	unsigned char fl_flags = request->fl_flags;
 	int status;
 
diff --git a/fs/nfs/nfs4renewd.c b/fs/nfs/nfs4renewd.c
index 5d764d8e6d8a..208764069f61 100644
--- a/fs/nfs/nfs4renewd.c
+++ b/fs/nfs/nfs4renewd.c
@@ -61,7 +61,7 @@
 void
 nfs4_renew_state(void *data)
 {
-	struct nfs4_client *clp = (struct nfs4_client *)data;
+	struct nfs_client *clp = (struct nfs_client *)data;
 	struct rpc_cred *cred;
 	long lease, timeout;
 	unsigned long last, now;
@@ -108,7 +108,7 @@ out:
 
 /* Must be called with clp->cl_sem locked for writes */
 void
-nfs4_schedule_state_renewal(struct nfs4_client *clp)
+nfs4_schedule_state_renewal(struct nfs_client *clp)
 {
 	long timeout;
 
@@ -127,7 +127,7 @@ nfs4_schedule_state_renewal(struct nfs4_client *clp)
 void
 nfs4_renewd_prepare_shutdown(struct nfs_server *server)
 {
-	struct nfs4_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs4_state;
 
 	if (!clp)
 		return;
@@ -140,7 +140,7 @@ nfs4_renewd_prepare_shutdown(struct nfs_server *server)
 
 /* Must be called with clp->cl_sem locked for writes */
 void
-nfs4_kill_renewd(struct nfs4_client *clp)
+nfs4_kill_renewd(struct nfs_client *clp)
 {
 	down_read(&clp->cl_sem);
 	if (!list_empty(&clp->cl_superblocks)) {
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index 090a36b07a22..c0b6439f1f71 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -83,10 +83,10 @@ destroy_nfsv4_state(struct nfs_server *server)
  * Since these are allocated/deallocated very rarely, we don't
  * bother putting them in a slab cache...
  */
-static struct nfs4_client *
+static struct nfs_client *
 nfs4_alloc_client(struct in_addr *addr)
 {
-	struct nfs4_client *clp;
+	struct nfs_client *clp;
 
 	if (nfs_callback_up() < 0)
 		return NULL;
@@ -111,7 +111,7 @@ nfs4_alloc_client(struct in_addr *addr)
 }
 
 static void
-nfs4_free_client(struct nfs4_client *clp)
+nfs4_free_client(struct nfs_client *clp)
 {
 	struct nfs4_state_owner *sp;
 
@@ -130,9 +130,9 @@ nfs4_free_client(struct nfs4_client *clp)
 	nfs_callback_down();
 }
 
-static struct nfs4_client *__nfs4_find_client(struct in_addr *addr)
+static struct nfs_client *__nfs4_find_client(struct in_addr *addr)
 {
-	struct nfs4_client *clp;
+	struct nfs_client *clp;
 	list_for_each_entry(clp, &nfs4_clientid_list, cl_servers) {
 		if (memcmp(&clp->cl_addr, addr, sizeof(clp->cl_addr)) == 0) {
 			atomic_inc(&clp->cl_count);
@@ -142,19 +142,19 @@ static struct nfs4_client *__nfs4_find_client(struct in_addr *addr)
 	return NULL;
 }
 
-struct nfs4_client *nfs4_find_client(struct in_addr *addr)
+struct nfs_client *nfs4_find_client(struct in_addr *addr)
 {
-	struct nfs4_client *clp;
+	struct nfs_client *clp;
 	spin_lock(&state_spinlock);
 	clp = __nfs4_find_client(addr);
 	spin_unlock(&state_spinlock);
 	return clp;
 }
 
-struct nfs4_client *
+struct nfs_client *
 nfs4_get_client(struct in_addr *addr)
 {
-	struct nfs4_client *clp, *new = NULL;
+	struct nfs_client *clp, *new = NULL;
 
 	spin_lock(&state_spinlock);
 	for (;;) {
@@ -180,7 +180,7 @@ nfs4_get_client(struct in_addr *addr)
 }
 
 void
-nfs4_put_client(struct nfs4_client *clp)
+nfs4_put_client(struct nfs_client *clp)
 {
 	if (!atomic_dec_and_lock(&clp->cl_count, &state_spinlock))
 		return;
@@ -192,7 +192,7 @@ nfs4_put_client(struct nfs4_client *clp)
 	nfs4_free_client(clp);
 }
 
-static int nfs4_init_client(struct nfs4_client *clp, struct rpc_cred *cred)
+static int nfs4_init_client(struct nfs_client *clp, struct rpc_cred *cred)
 {
 	int status = nfs4_proc_setclientid(clp, NFS4_CALLBACK,
 			nfs_callback_tcpport, cred);
@@ -204,13 +204,13 @@ static int nfs4_init_client(struct nfs4_client *clp, struct rpc_cred *cred)
 }
 
 u32
-nfs4_alloc_lockowner_id(struct nfs4_client *clp)
+nfs4_alloc_lockowner_id(struct nfs_client *clp)
 {
 	return clp->cl_lockowner_id ++;
 }
 
 static struct nfs4_state_owner *
-nfs4_client_grab_unused(struct nfs4_client *clp, struct rpc_cred *cred)
+nfs4_client_grab_unused(struct nfs_client *clp, struct rpc_cred *cred)
 {
 	struct nfs4_state_owner *sp = NULL;
 
@@ -224,7 +224,7 @@ nfs4_client_grab_unused(struct nfs4_client *clp, struct rpc_cred *cred)
 	return sp;
 }
 
-struct rpc_cred *nfs4_get_renew_cred(struct nfs4_client *clp)
+struct rpc_cred *nfs4_get_renew_cred(struct nfs_client *clp)
 {
 	struct nfs4_state_owner *sp;
 	struct rpc_cred *cred = NULL;
@@ -238,7 +238,7 @@ struct rpc_cred *nfs4_get_renew_cred(struct nfs4_client *clp)
 	return cred;
 }
 
-struct rpc_cred *nfs4_get_setclientid_cred(struct nfs4_client *clp)
+struct rpc_cred *nfs4_get_setclientid_cred(struct nfs_client *clp)
 {
 	struct nfs4_state_owner *sp;
 
@@ -251,7 +251,7 @@ struct rpc_cred *nfs4_get_setclientid_cred(struct nfs4_client *clp)
 }
 
 static struct nfs4_state_owner *
-nfs4_find_state_owner(struct nfs4_client *clp, struct rpc_cred *cred)
+nfs4_find_state_owner(struct nfs_client *clp, struct rpc_cred *cred)
 {
 	struct nfs4_state_owner *sp, *res = NULL;
 
@@ -294,7 +294,7 @@ nfs4_alloc_state_owner(void)
 void
 nfs4_drop_state_owner(struct nfs4_state_owner *sp)
 {
-	struct nfs4_client *clp = sp->so_client;
+	struct nfs_client *clp = sp->so_client;
 	spin_lock(&clp->cl_lock);
 	list_del_init(&sp->so_list);
 	spin_unlock(&clp->cl_lock);
@@ -306,7 +306,7 @@ nfs4_drop_state_owner(struct nfs4_state_owner *sp)
  */
 struct nfs4_state_owner *nfs4_get_state_owner(struct nfs_server *server, struct rpc_cred *cred)
 {
-	struct nfs4_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs4_state;
 	struct nfs4_state_owner *sp, *new;
 
 	get_rpccred(cred);
@@ -337,7 +337,7 @@ struct nfs4_state_owner *nfs4_get_state_owner(struct nfs_server *server, struct
  */
 void nfs4_put_state_owner(struct nfs4_state_owner *sp)
 {
-	struct nfs4_client *clp = sp->so_client;
+	struct nfs_client *clp = sp->so_client;
 	struct rpc_cred *cred = sp->so_cred;
 
 	if (!atomic_dec_and_lock(&sp->so_count, &clp->cl_lock))
@@ -540,7 +540,7 @@ __nfs4_find_lock_state(struct nfs4_state *state, fl_owner_t fl_owner)
 static struct nfs4_lock_state *nfs4_alloc_lock_state(struct nfs4_state *state, fl_owner_t fl_owner)
 {
 	struct nfs4_lock_state *lsp;
-	struct nfs4_client *clp = state->owner->so_client;
+	struct nfs_client *clp = state->owner->so_client;
 
 	lsp = kzalloc(sizeof(*lsp), GFP_KERNEL);
 	if (lsp == NULL)
@@ -752,7 +752,7 @@ out:
 
 static int reclaimer(void *);
 
-static inline void nfs4_clear_recover_bit(struct nfs4_client *clp)
+static inline void nfs4_clear_recover_bit(struct nfs_client *clp)
 {
 	smp_mb__before_clear_bit();
 	clear_bit(NFS4CLNT_STATE_RECOVER, &clp->cl_state);
@@ -764,7 +764,7 @@ static inline void nfs4_clear_recover_bit(struct nfs4_client *clp)
 /*
  * State recovery routine
  */
-static void nfs4_recover_state(struct nfs4_client *clp)
+static void nfs4_recover_state(struct nfs_client *clp)
 {
 	struct task_struct *task;
 
@@ -782,7 +782,7 @@ static void nfs4_recover_state(struct nfs4_client *clp)
 /*
  * Schedule a state recovery attempt
  */
-void nfs4_schedule_state_recovery(struct nfs4_client *clp)
+void nfs4_schedule_state_recovery(struct nfs_client *clp)
 {
 	if (!clp)
 		return;
@@ -879,7 +879,7 @@ out_err:
 	return status;
 }
 
-static void nfs4_state_mark_reclaim(struct nfs4_client *clp)
+static void nfs4_state_mark_reclaim(struct nfs_client *clp)
 {
 	struct nfs4_state_owner *sp;
 	struct nfs4_state *state;
@@ -903,7 +903,7 @@ static void nfs4_state_mark_reclaim(struct nfs4_client *clp)
 
 static int reclaimer(void *ptr)
 {
-	struct nfs4_client *clp = ptr;
+	struct nfs_client *clp = ptr;
 	struct nfs4_state_owner *sp;
 	struct nfs4_state_recovery_ops *ops;
 	struct rpc_cred *cred;
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index 1dee6ef7e5a9..04748ab9ed55 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -1160,7 +1160,7 @@ static int encode_rename(struct xdr_stream *xdr, const struct qstr *oldname, con
 	return 0;
 }
 
-static int encode_renew(struct xdr_stream *xdr, const struct nfs4_client *client_stateid)
+static int encode_renew(struct xdr_stream *xdr, const struct nfs_client *client_stateid)
 {
 	uint32_t *p;
 
@@ -1246,7 +1246,7 @@ static int encode_setclientid(struct xdr_stream *xdr, const struct nfs4_setclien
 	return 0;
 }
 
-static int encode_setclientid_confirm(struct xdr_stream *xdr, const struct nfs4_client *client_state)
+static int encode_setclientid_confirm(struct xdr_stream *xdr, const struct nfs_client *client_state)
 {
         uint32_t *p;
 
@@ -1945,7 +1945,7 @@ static int nfs4_xdr_enc_server_caps(struct rpc_rqst *req, uint32_t *p, const str
 /*
  * a RENEW request
  */
-static int nfs4_xdr_enc_renew(struct rpc_rqst *req, uint32_t *p, struct nfs4_client *clp)
+static int nfs4_xdr_enc_renew(struct rpc_rqst *req, uint32_t *p, struct nfs_client *clp)
 {
 	struct xdr_stream xdr;
 	struct compound_hdr hdr = {
@@ -1975,7 +1975,7 @@ static int nfs4_xdr_enc_setclientid(struct rpc_rqst *req, uint32_t *p, struct nf
 /*
  * a SETCLIENTID_CONFIRM request
  */
-static int nfs4_xdr_enc_setclientid_confirm(struct rpc_rqst *req, uint32_t *p, struct nfs4_client *clp)
+static int nfs4_xdr_enc_setclientid_confirm(struct rpc_rqst *req, uint32_t *p, struct nfs_client *clp)
 {
 	struct xdr_stream xdr;
 	struct compound_hdr hdr = {
@@ -2132,7 +2132,7 @@ static int decode_op_hdr(struct xdr_stream *xdr, enum nfs_opnum4 expected)
 }
 
 /* Dummy routine */
-static int decode_ace(struct xdr_stream *xdr, void *ace, struct nfs4_client *clp)
+static int decode_ace(struct xdr_stream *xdr, void *ace, struct nfs_client *clp)
 {
 	uint32_t *p;
 	unsigned int strlen;
@@ -2636,7 +2636,7 @@ static int decode_attr_nlink(struct xdr_stream *xdr, uint32_t *bitmap, uint32_t
 	return 0;
 }
 
-static int decode_attr_owner(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs4_client *clp, int32_t *uid)
+static int decode_attr_owner(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs_client *clp, int32_t *uid)
 {
 	uint32_t len, *p;
 
@@ -2660,7 +2660,7 @@ static int decode_attr_owner(struct xdr_stream *xdr, uint32_t *bitmap, struct nf
 	return 0;
 }
 
-static int decode_attr_group(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs4_client *clp, int32_t *gid)
+static int decode_attr_group(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs_client *clp, int32_t *gid)
 {
 	uint32_t len, *p;
 
@@ -3565,7 +3565,7 @@ static int decode_setattr(struct xdr_stream *xdr, struct nfs_setattrres *res)
 	return 0;
 }
 
-static int decode_setclientid(struct xdr_stream *xdr, struct nfs4_client *clp)
+static int decode_setclientid(struct xdr_stream *xdr, struct nfs_client *clp)
 {
 	uint32_t *p;
 	uint32_t opnum;
@@ -4335,7 +4335,7 @@ static int nfs4_xdr_dec_renew(struct rpc_rqst *rqstp, uint32_t *p, void *dummy)
  * a SETCLIENTID request
  */
 static int nfs4_xdr_dec_setclientid(struct rpc_rqst *req, uint32_t *p,
-		struct nfs4_client *clp)
+		struct nfs_client *clp)
 {
 	struct xdr_stream xdr;
 	struct compound_hdr hdr;
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 63497345806b..d03ede5b1aca 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -1099,7 +1099,7 @@ static int nfs_clone_nfs_sb(struct file_system_type *fs_type,
 static struct rpc_clnt *nfs4_create_client(struct nfs_server *server,
 	struct rpc_timeout *timeparms, int proto, rpc_authflavor_t flavor)
 {
-	struct nfs4_client *clp;
+	struct nfs_client *clp;
 	struct rpc_xprt *xprt = NULL;
 	struct rpc_clnt *clnt = NULL;
 	int err = -EIO;
@@ -1416,7 +1416,7 @@ static inline char *nfs4_dup_path(const struct dentry *dentry)
 static struct super_block *nfs4_clone_sb(struct nfs_server *server, struct nfs_clone_mount *data)
 {
 	const struct dentry *dentry = data->dentry;
-	struct nfs4_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs4_state;
 	struct super_block *sb;
 
 	server->fsid = data->fattr->fsid;
diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index 6b4a13c79474..4db90df2aed0 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -43,7 +43,7 @@ struct nfs_server {
 	 */
 	char			ip_addr[16];
 	char *			mnt_path;
-	struct nfs4_client *	nfs4_state;	/* all NFSv4 state starts here */
+	struct nfs_client *	nfs4_state;	/* all NFSv4 state starts here */
 	struct list_head	nfs4_siblings;	/* List of other nfs_server structs
 						 * that share the same clientid
 						 */
diff --git a/include/linux/nfs_idmap.h b/include/linux/nfs_idmap.h
index 102e56094296..678fe68982ef 100644
--- a/include/linux/nfs_idmap.h
+++ b/include/linux/nfs_idmap.h
@@ -62,15 +62,15 @@ struct idmap_msg {
 #ifdef __KERNEL__
 
 /* Forward declaration to make this header independent of others */
-struct nfs4_client;
+struct nfs_client;
 
-void nfs_idmap_new(struct nfs4_client *);
-void nfs_idmap_delete(struct nfs4_client *);
+void nfs_idmap_new(struct nfs_client *);
+void nfs_idmap_delete(struct nfs_client *);
 
-int nfs_map_name_to_uid(struct nfs4_client *, const char *, size_t, __u32 *);
-int nfs_map_group_to_gid(struct nfs4_client *, const char *, size_t, __u32 *);
-int nfs_map_uid_to_name(struct nfs4_client *, __u32, char *);
-int nfs_map_gid_to_group(struct nfs4_client *, __u32, char *);
+int nfs_map_name_to_uid(struct nfs_client *, const char *, size_t, __u32 *);
+int nfs_map_group_to_gid(struct nfs_client *, const char *, size_t, __u32 *);
+int nfs_map_uid_to_name(struct nfs_client *, __u32, char *);
+int nfs_map_gid_to_group(struct nfs_client *, __u32, char *);
 
 extern unsigned int nfs_idmap_cache_timeout;
 #endif /* __KERNEL__ */
-- 
cgit v1.2.3


From 7539bbab8062aadc1db95a22b377146843cfa88f Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Tue, 22 Aug 2006 20:06:09 -0400
Subject: NFS: Rename nfs_server::nfs4_state

Rename nfs_server::nfs4_state to nfs_client as it will be used to represent the
client state for NFS2 and NFS3 also.

Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/delegation.c       | 12 ++++++------
 fs/nfs/nfs4proc.c         | 26 +++++++++++++-------------
 fs/nfs/nfs4renewd.c       |  2 +-
 fs/nfs/nfs4state.c        | 10 +++++-----
 fs/nfs/nfs4xdr.c          | 10 +++++-----
 fs/nfs/super.c            |  6 +++---
 include/linux/nfs_fs_sb.h |  2 +-
 7 files changed, 34 insertions(+), 34 deletions(-)

(limited to 'include')

diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c
index 5a1105c258bd..cfe239736ac0 100644
--- a/fs/nfs/delegation.c
+++ b/fs/nfs/delegation.c
@@ -52,7 +52,7 @@ static int nfs_delegation_claim_locks(struct nfs_open_context *ctx, struct nfs4_
 			case -NFS4ERR_EXPIRED:
 				/* kill_proc(fl->fl_pid, SIGLOST, 1); */
 			case -NFS4ERR_STALE_CLIENTID:
-				nfs4_schedule_state_recovery(NFS_SERVER(inode)->nfs4_state);
+				nfs4_schedule_state_recovery(NFS_SERVER(inode)->nfs_client);
 				goto out_err;
 		}
 	}
@@ -114,7 +114,7 @@ void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred, st
  */
 int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res)
 {
-	struct nfs_client *clp = NFS_SERVER(inode)->nfs4_state;
+	struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_delegation *delegation;
 	int status = 0;
@@ -176,7 +176,7 @@ static void nfs_msync_inode(struct inode *inode)
  */
 int __nfs_inode_return_delegation(struct inode *inode)
 {
-	struct nfs_client *clp = NFS_SERVER(inode)->nfs4_state;
+	struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_delegation *delegation;
 	int res = 0;
@@ -208,7 +208,7 @@ int __nfs_inode_return_delegation(struct inode *inode)
  */
 void nfs_return_all_delegations(struct super_block *sb)
 {
-	struct nfs_client *clp = NFS_SB(sb)->nfs4_state;
+	struct nfs_client *clp = NFS_SB(sb)->nfs_client;
 	struct nfs_delegation *delegation;
 	struct inode *inode;
 
@@ -310,7 +310,7 @@ static int recall_thread(void *data)
 {
 	struct recall_threadargs *args = (struct recall_threadargs *)data;
 	struct inode *inode = igrab(args->inode);
-	struct nfs_client *clp = NFS_SERVER(inode)->nfs4_state;
+	struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_delegation *delegation;
 
@@ -423,7 +423,7 @@ void nfs_delegation_reap_unclaimed(struct nfs_client *clp)
 
 int nfs4_copy_delegation_stateid(nfs4_stateid *dst, struct inode *inode)
 {
-	struct nfs_client *clp = NFS_SERVER(inode)->nfs4_state;
+	struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_delegation *delegation;
 	int res = 0;
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 168f3ffb059f..b46597fc81e1 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -195,7 +195,7 @@ static void nfs4_setup_readdir(u64 cookie, u32 *verifier, struct dentry *dentry,
 
 static void renew_lease(const struct nfs_server *server, unsigned long timestamp)
 {
-	struct nfs_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs_client;
 	spin_lock(&clp->cl_lock);
 	if (time_before(clp->cl_last_renewal,timestamp))
 		clp->cl_last_renewal = timestamp;
@@ -252,7 +252,7 @@ static struct nfs4_opendata *nfs4_opendata_alloc(struct dentry *dentry,
 	atomic_inc(&sp->so_count);
 	p->o_arg.fh = NFS_FH(dir);
 	p->o_arg.open_flags = flags,
-	p->o_arg.clientid = server->nfs4_state->cl_clientid;
+	p->o_arg.clientid = server->nfs_client->cl_clientid;
 	p->o_arg.id = sp->so_id;
 	p->o_arg.name = &dentry->d_name;
 	p->o_arg.server = server;
@@ -550,7 +550,7 @@ int nfs4_open_delegation_recall(struct dentry *dentry, struct nfs4_state *state)
 			case -NFS4ERR_STALE_STATEID:
 			case -NFS4ERR_EXPIRED:
 				/* Don't recall a delegation if it was lost */
-				nfs4_schedule_state_recovery(server->nfs4_state);
+				nfs4_schedule_state_recovery(server->nfs_client);
 				return err;
 		}
 		err = nfs4_handle_exception(server, err, &exception);
@@ -792,7 +792,7 @@ out:
 
 int nfs4_recover_expired_lease(struct nfs_server *server)
 {
-	struct nfs_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs_client;
 
 	if (test_and_clear_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state))
 		nfs4_schedule_state_recovery(clp);
@@ -867,7 +867,7 @@ static int _nfs4_open_delegated(struct inode *inode, int flags, struct rpc_cred
 {
 	struct nfs_delegation *delegation;
 	struct nfs_server *server = NFS_SERVER(inode);
-	struct nfs_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs_client;
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs4_state_owner *sp = NULL;
 	struct nfs4_state *state = NULL;
@@ -953,7 +953,7 @@ static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, st
 	struct nfs4_state_owner  *sp;
 	struct nfs4_state     *state = NULL;
 	struct nfs_server       *server = NFS_SERVER(dir);
-	struct nfs_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs_client;
 	struct nfs4_opendata *opendata;
 	int                     status;
 
@@ -1133,7 +1133,7 @@ static void nfs4_close_done(struct rpc_task *task, void *data)
 			break;
 		case -NFS4ERR_STALE_STATEID:
 		case -NFS4ERR_EXPIRED:
-			nfs4_schedule_state_recovery(server->nfs4_state);
+			nfs4_schedule_state_recovery(server->nfs_client);
 			break;
 		default:
 			if (nfs4_async_handle_error(task, server) == -EAGAIN) {
@@ -2791,7 +2791,7 @@ static int nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t buflen
 static int
 nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server)
 {
-	struct nfs_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs_client;
 
 	if (!clp || task->tk_status >= 0)
 		return 0;
@@ -2871,7 +2871,7 @@ static int nfs4_delay(struct rpc_clnt *clnt, long *timeout)
  */
 int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct nfs4_exception *exception)
 {
-	struct nfs_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs_client;
 	int ret = errorcode;
 
 	exception->retry = 0;
@@ -3077,7 +3077,7 @@ int nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, const nfs4
 		switch (err) {
 			case -NFS4ERR_STALE_STATEID:
 			case -NFS4ERR_EXPIRED:
-				nfs4_schedule_state_recovery(server->nfs4_state);
+				nfs4_schedule_state_recovery(server->nfs_client);
 			case 0:
 				return 0;
 		}
@@ -3106,7 +3106,7 @@ static int _nfs4_proc_getlk(struct nfs4_state *state, int cmd, struct file_lock
 {
 	struct inode *inode = state->inode;
 	struct nfs_server *server = NFS_SERVER(inode);
-	struct nfs_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs_client;
 	struct nfs_lockt_args arg = {
 		.fh = NFS_FH(inode),
 		.fl = request,
@@ -3231,7 +3231,7 @@ static void nfs4_locku_done(struct rpc_task *task, void *data)
 			break;
 		case -NFS4ERR_STALE_STATEID:
 		case -NFS4ERR_EXPIRED:
-			nfs4_schedule_state_recovery(calldata->server->nfs4_state);
+			nfs4_schedule_state_recovery(calldata->server->nfs_client);
 			break;
 		default:
 			if (nfs4_async_handle_error(task, calldata->server) == -EAGAIN) {
@@ -3343,7 +3343,7 @@ static struct nfs4_lockdata *nfs4_alloc_lockdata(struct file_lock *fl,
 	if (p->arg.lock_seqid == NULL)
 		goto out_free;
 	p->arg.lock_stateid = &lsp->ls_stateid;
-	p->arg.lock_owner.clientid = server->nfs4_state->cl_clientid;
+	p->arg.lock_owner.clientid = server->nfs_client->cl_clientid;
 	p->arg.lock_owner.id = lsp->ls_id;
 	p->lsp = lsp;
 	atomic_inc(&lsp->ls_count);
diff --git a/fs/nfs/nfs4renewd.c b/fs/nfs/nfs4renewd.c
index 208764069f61..ff947ecb8b81 100644
--- a/fs/nfs/nfs4renewd.c
+++ b/fs/nfs/nfs4renewd.c
@@ -127,7 +127,7 @@ nfs4_schedule_state_renewal(struct nfs_client *clp)
 void
 nfs4_renewd_prepare_shutdown(struct nfs_server *server)
 {
-	struct nfs_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs_client;
 
 	if (!clp)
 		return;
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index c0b6439f1f71..fa51a7d4c022 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -61,7 +61,7 @@ static LIST_HEAD(nfs4_clientid_list);
 void
 init_nfsv4_state(struct nfs_server *server)
 {
-	server->nfs4_state = NULL;
+	server->nfs_client = NULL;
 	INIT_LIST_HEAD(&server->nfs4_siblings);
 }
 
@@ -70,9 +70,9 @@ destroy_nfsv4_state(struct nfs_server *server)
 {
 	kfree(server->mnt_path);
 	server->mnt_path = NULL;
-	if (server->nfs4_state) {
-		nfs4_put_client(server->nfs4_state);
-		server->nfs4_state = NULL;
+	if (server->nfs_client) {
+		nfs4_put_client(server->nfs_client);
+		server->nfs_client = NULL;
 	}
 }
 
@@ -306,7 +306,7 @@ nfs4_drop_state_owner(struct nfs4_state_owner *sp)
  */
 struct nfs4_state_owner *nfs4_get_state_owner(struct nfs_server *server, struct rpc_cred *cred)
 {
-	struct nfs_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs_client;
 	struct nfs4_state_owner *sp, *new;
 
 	get_rpccred(cred);
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index 04748ab9ed55..99926067eca4 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -529,7 +529,7 @@ static int encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const s
 	if (iap->ia_valid & ATTR_MODE)
 		len += 4;
 	if (iap->ia_valid & ATTR_UID) {
-		owner_namelen = nfs_map_uid_to_name(server->nfs4_state, iap->ia_uid, owner_name);
+		owner_namelen = nfs_map_uid_to_name(server->nfs_client, iap->ia_uid, owner_name);
 		if (owner_namelen < 0) {
 			printk(KERN_WARNING "nfs: couldn't resolve uid %d to string\n",
 			       iap->ia_uid);
@@ -541,7 +541,7 @@ static int encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const s
 		len += 4 + (XDR_QUADLEN(owner_namelen) << 2);
 	}
 	if (iap->ia_valid & ATTR_GID) {
-		owner_grouplen = nfs_map_gid_to_group(server->nfs4_state, iap->ia_gid, owner_group);
+		owner_grouplen = nfs_map_gid_to_group(server->nfs_client, iap->ia_gid, owner_group);
 		if (owner_grouplen < 0) {
 			printk(KERN_WARNING "nfs4: couldn't resolve gid %d to string\n",
 			       iap->ia_gid);
@@ -3051,9 +3051,9 @@ static int decode_getfattr(struct xdr_stream *xdr, struct nfs_fattr *fattr, cons
 	fattr->mode |= fmode;
 	if ((status = decode_attr_nlink(xdr, bitmap, &fattr->nlink)) != 0)
 		goto xdr_error;
-	if ((status = decode_attr_owner(xdr, bitmap, server->nfs4_state, &fattr->uid)) != 0)
+	if ((status = decode_attr_owner(xdr, bitmap, server->nfs_client, &fattr->uid)) != 0)
 		goto xdr_error;
-	if ((status = decode_attr_group(xdr, bitmap, server->nfs4_state, &fattr->gid)) != 0)
+	if ((status = decode_attr_group(xdr, bitmap, server->nfs_client, &fattr->gid)) != 0)
 		goto xdr_error;
 	if ((status = decode_attr_rdev(xdr, bitmap, &fattr->rdev)) != 0)
 		goto xdr_error;
@@ -3254,7 +3254,7 @@ static int decode_delegation(struct xdr_stream *xdr, struct nfs_openres *res)
 			if (decode_space_limit(xdr, &res->maxsize) < 0)
 				return -EIO;
 	}
-	return decode_ace(xdr, NULL, res->server->nfs4_state);
+	return decode_ace(xdr, NULL, res->server->nfs_client);
 }
 
 static int decode_open(struct xdr_stream *xdr, struct nfs_openres *res)
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index d03ede5b1aca..ab4c78ee840c 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -1141,7 +1141,7 @@ static struct rpc_clnt *nfs4_create_client(struct nfs_server *server,
 	list_add_tail(&server->nfs4_siblings, &clp->cl_superblocks);
 	clnt = rpc_clone_client(clp->cl_rpcclient);
 	if (!IS_ERR(clnt))
-		server->nfs4_state = clp;
+		server->nfs_client = clp;
 	up_write(&clp->cl_sem);
 	clp = NULL;
 
@@ -1151,7 +1151,7 @@ static struct rpc_clnt *nfs4_create_client(struct nfs_server *server,
 		return clnt;
 	}
 
-	if (server->nfs4_state->cl_idmap == NULL) {
+	if (server->nfs_client->cl_idmap == NULL) {
 		dprintk("%s: failed to create idmapper.\n", __FUNCTION__);
 		return ERR_PTR(-ENOMEM);
 	}
@@ -1416,7 +1416,7 @@ static inline char *nfs4_dup_path(const struct dentry *dentry)
 static struct super_block *nfs4_clone_sb(struct nfs_server *server, struct nfs_clone_mount *data)
 {
 	const struct dentry *dentry = data->dentry;
-	struct nfs_client *clp = server->nfs4_state;
+	struct nfs_client *clp = server->nfs_client;
 	struct super_block *sb;
 
 	server->fsid = data->fattr->fsid;
diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index 4db90df2aed0..fc20d6b934fb 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -43,7 +43,7 @@ struct nfs_server {
 	 */
 	char			ip_addr[16];
 	char *			mnt_path;
-	struct nfs_client *	nfs4_state;	/* all NFSv4 state starts here */
+	struct nfs_client *	nfs_client;	/* all NFSv4 state starts here */
 	struct list_head	nfs4_siblings;	/* List of other nfs_server structs
 						 * that share the same clientid
 						 */
-- 
cgit v1.2.3


From b7162792b5c0e0f6e91b8997f8e6bbc76ec5420a Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Tue, 22 Aug 2006 20:06:09 -0400
Subject: NFS: Return an error when starting the idmapping pipe

Return an error when starting the idmapping pipe so that we can detect it
failing.

Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/idmap.c            | 12 ++++++++----
 fs/nfs/super.c            |  3 ++-
 include/linux/nfs_idmap.h |  2 +-
 3 files changed, 11 insertions(+), 6 deletions(-)

(limited to 'include')

diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c
index d05148ec9414..231c20ffc0ff 100644
--- a/fs/nfs/idmap.c
+++ b/fs/nfs/idmap.c
@@ -108,15 +108,17 @@ static struct rpc_pipe_ops idmap_upcall_ops = {
         .destroy_msg    = idmap_pipe_destroy_msg,
 };
 
-void
+int
 nfs_idmap_new(struct nfs_client *clp)
 {
 	struct idmap *idmap;
+	int error;
 
 	if (clp->cl_idmap != NULL)
-		return;
+		return 0;
+
         if ((idmap = kzalloc(sizeof(*idmap), GFP_KERNEL)) == NULL)
-                return;
+                return -ENOMEM;
 
 	snprintf(idmap->idmap_path, sizeof(idmap->idmap_path),
 	    "%s/idmap", clp->cl_rpcclient->cl_pathname);
@@ -124,8 +126,9 @@ nfs_idmap_new(struct nfs_client *clp)
         idmap->idmap_dentry = rpc_mkpipe(idmap->idmap_path,
 	    idmap, &idmap_upcall_ops, 0);
         if (IS_ERR(idmap->idmap_dentry)) {
+		error = PTR_ERR(idmap->idmap_dentry);
 		kfree(idmap);
-		return;
+		return error;
 	}
 
         mutex_init(&idmap->idmap_lock);
@@ -135,6 +138,7 @@ nfs_idmap_new(struct nfs_client *clp)
 	idmap->idmap_group_hash.h_type = IDMAP_TYPE_GROUP;
 
 	clp->cl_idmap = idmap;
+	return 0;
 }
 
 void
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index ab4c78ee840c..3ee85c4e65d8 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -1136,7 +1136,8 @@ static struct rpc_clnt *nfs4_create_client(struct nfs_server *server,
 		clnt->cl_softrtry = 1;
 		clp->cl_rpcclient = clnt;
 		memcpy(clp->cl_ipaddr, server->ip_addr, sizeof(clp->cl_ipaddr));
-		nfs_idmap_new(clp);
+		if (nfs_idmap_new(clp) < 0)
+			goto out_fail;
 	}
 	list_add_tail(&server->nfs4_siblings, &clp->cl_superblocks);
 	clnt = rpc_clone_client(clp->cl_rpcclient);
diff --git a/include/linux/nfs_idmap.h b/include/linux/nfs_idmap.h
index 678fe68982ef..15a9f3b7289a 100644
--- a/include/linux/nfs_idmap.h
+++ b/include/linux/nfs_idmap.h
@@ -64,7 +64,7 @@ struct idmap_msg {
 /* Forward declaration to make this header independent of others */
 struct nfs_client;
 
-void nfs_idmap_new(struct nfs_client *);
+int nfs_idmap_new(struct nfs_client *);
 void nfs_idmap_delete(struct nfs_client *);
 
 int nfs_map_name_to_uid(struct nfs_client *, const char *, size_t, __u32 *);
-- 
cgit v1.2.3


From 2b3de4411b3ccaeb00018c99d1bbe7203554cf7f Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Tue, 22 Aug 2006 20:06:09 -0400
Subject: NFS: Add a lookupfh NFS RPC op

Add a lookup filehandle NFS RPC op so that a file handle can be looked up
without requiring dentries and inodes and other VFS stuff when doing an NFS4
pathwalk during mounting.

Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/nfs4proc.c       | 47 +++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/nfs_xdr.h |  3 +++
 2 files changed, 50 insertions(+)

(limited to 'include')

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index b46597fc81e1..de2006f754ef 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -1583,6 +1583,52 @@ nfs4_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr,
 	return status;
 }
 
+static int _nfs4_proc_lookupfh(struct nfs_server *server, struct nfs_fh *dirfh,
+		struct qstr *name, struct nfs_fh *fhandle,
+		struct nfs_fattr *fattr)
+{
+	int		       status;
+	struct nfs4_lookup_arg args = {
+		.bitmask = server->attr_bitmask,
+		.dir_fh = dirfh,
+		.name = name,
+	};
+	struct nfs4_lookup_res res = {
+		.server = server,
+		.fattr = fattr,
+		.fh = fhandle,
+	};
+	struct rpc_message msg = {
+		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LOOKUP],
+		.rpc_argp = &args,
+		.rpc_resp = &res,
+	};
+
+	nfs_fattr_init(fattr);
+
+	dprintk("NFS call  lookupfh %s\n", name->name);
+	status = rpc_call_sync(server->client, &msg, 0);
+	dprintk("NFS reply lookupfh: %d\n", status);
+	if (status == -NFS4ERR_MOVED)
+		status = -EREMOTE;
+	return status;
+}
+
+static int nfs4_proc_lookupfh(struct nfs_server *server, struct nfs_fh *dirfh,
+			      struct qstr *name, struct nfs_fh *fhandle,
+			      struct nfs_fattr *fattr)
+{
+	struct nfs4_exception exception = { };
+	int err;
+	do {
+		err = nfs4_handle_exception(server,
+				_nfs4_proc_lookupfh(server, dirfh, name,
+						    fhandle, fattr),
+				&exception);
+	} while (exception.retry);
+	return err;
+}
+
 static int _nfs4_proc_lookup(struct inode *dir, struct qstr *name,
 		struct nfs_fh *fhandle, struct nfs_fattr *fattr)
 {
@@ -3723,6 +3769,7 @@ struct nfs_rpc_ops	nfs_v4_clientops = {
 	.getroot	= nfs4_proc_get_root,
 	.getattr	= nfs4_proc_getattr,
 	.setattr	= nfs4_proc_setattr,
+	.lookupfh	= nfs4_proc_lookupfh,
 	.lookup		= nfs4_proc_lookup,
 	.access		= nfs4_proc_access,
 	.readlink	= nfs4_proc_readlink,
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index 41e5a19199e9..26879771831d 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -770,6 +770,9 @@ struct nfs_rpc_ops {
 
 	int	(*getroot) (struct nfs_server *, struct nfs_fh *,
 			    struct nfs_fsinfo *);
+	int	(*lookupfh)(struct nfs_server *, struct nfs_fh *,
+			    struct qstr *, struct nfs_fh *,
+			    struct nfs_fattr *);
 	int	(*getattr) (struct nfs_server *, struct nfs_fh *,
 			    struct nfs_fattr *);
 	int	(*setattr) (struct dentry *, struct nfs_fattr *,
-- 
cgit v1.2.3


From e9326dcab413848e70ab746c7c5363da13e5f801 Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Tue, 22 Aug 2006 20:06:10 -0400
Subject: NFS: Add a server capabilities NFS RPC op

Add a set_capabilities NFS RPC op so that the server capabilities can be set.

Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/nfs4proc.c       | 1 +
 include/linux/nfs_xdr.h | 1 +
 2 files changed, 2 insertions(+)

(limited to 'include')

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index de2006f754ef..850f0851023a 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -3790,6 +3790,7 @@ struct nfs_rpc_ops	nfs_v4_clientops = {
 	.statfs		= nfs4_proc_statfs,
 	.fsinfo		= nfs4_proc_fsinfo,
 	.pathconf	= nfs4_proc_pathconf,
+	.set_capabilities = nfs4_server_capabilities,
 	.decode_dirent	= nfs4_decode_dirent,
 	.read_setup	= nfs4_proc_read_setup,
 	.read_done	= nfs4_read_done,
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index 26879771831d..dd9ae6761f71 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -809,6 +809,7 @@ struct nfs_rpc_ops {
 			    struct nfs_fsinfo *);
 	int	(*pathconf) (struct nfs_server *, struct nfs_fh *,
 			     struct nfs_pathconf *);
+	int	(*set_capabilities)(struct nfs_server *, struct nfs_fh *);
 	u32 *	(*decode_dirent)(u32 *, struct nfs_entry *, int plus);
 	void	(*read_setup)   (struct nfs_read_data *);
 	int	(*read_done)  (struct rpc_task *, struct nfs_read_data *);
-- 
cgit v1.2.3


From 24c8dbbb5f777187d660393599641ab3307b4b97 Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Tue, 22 Aug 2006 20:06:10 -0400
Subject: NFS: Generalise the nfs_client structure

Generalise the nfs_client structure by:

 (1) Moving nfs_client to a more general place (nfs_fs_sb.h).

 (2) Renaming its maintenance routines to be non-NFS4 specific.

 (3) Move those maintenance routines to a new non-NFS4 specific file (client.c)
     and move the declarations to internal.h.

 (4) Make nfs_find/get_client() take a full sockaddr_in to include the port
     number (will be required for NFS2/3).

 (5) Make nfs_find/get_client() take the NFS protocol version (again will be
     required to differentiate NFS2, 3 & 4 client records).

Also:

 (6) Make nfs_client construction proceed akin to inodes, marking them as under
     construction and providing a function to indicate completion.

 (7) Make nfs_get_client() wait interruptibly if it finds a client that it can
     share, but that client is currently being constructed.

 (8) Make nfs4_create_client() use (6) and (7) instead of locking cl_sem.

Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/Makefile           |   6 +-
 fs/nfs/callback.c         |   9 +-
 fs/nfs/callback_proc.c    |   9 +-
 fs/nfs/client.c           | 312 ++++++++++++++++++++++++++++++++++++++++++++++
 fs/nfs/delegation.c       |   9 +-
 fs/nfs/internal.h         |   6 +
 fs/nfs/nfs4_fs.h          |  52 --------
 fs/nfs/nfs4proc.c         |   2 +-
 fs/nfs/nfs4state.c        | 128 +------------------
 fs/nfs/super.c            |  53 ++++----
 include/linux/nfs_fs.h    |   1 +
 include/linux/nfs_fs_sb.h |  60 +++++++++
 12 files changed, 425 insertions(+), 222 deletions(-)
 create mode 100644 fs/nfs/client.c

(limited to 'include')

diff --git a/fs/nfs/Makefile b/fs/nfs/Makefile
index 0b572a0c1967..3b993a6f8163 100644
--- a/fs/nfs/Makefile
+++ b/fs/nfs/Makefile
@@ -4,9 +4,9 @@
 
 obj-$(CONFIG_NFS_FS) += nfs.o
 
-nfs-y 			:= dir.o file.o inode.o super.o nfs2xdr.o pagelist.o \
-			   proc.o read.o symlink.o unlink.o write.o \
-			   namespace.o
+nfs-y 			:= client.o dir.o file.o inode.o super.o nfs2xdr.o \
+			   pagelist.o proc.o read.o symlink.o unlink.o \
+			   write.o namespace.o
 nfs-$(CONFIG_ROOT_NFS)	+= nfsroot.o mount_clnt.o      
 nfs-$(CONFIG_NFS_V3)	+= nfs3proc.o nfs3xdr.o
 nfs-$(CONFIG_NFS_V3_ACL)	+= nfs3acl.o
diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c
index 1b596b6d9dc2..a3ee11364db0 100644
--- a/fs/nfs/callback.c
+++ b/fs/nfs/callback.c
@@ -19,6 +19,7 @@
 
 #include "nfs4_fs.h"
 #include "callback.h"
+#include "internal.h"
 
 #define NFSDBG_FACILITY NFSDBG_CALLBACK
 
@@ -166,15 +167,15 @@ void nfs_callback_down(void)
 
 static int nfs_callback_authenticate(struct svc_rqst *rqstp)
 {
-	struct in_addr *addr = &rqstp->rq_addr.sin_addr;
+	struct sockaddr_in *addr = &rqstp->rq_addr;
 	struct nfs_client *clp;
 
 	/* Don't talk to strangers */
-	clp = nfs4_find_client(addr);
+	clp = nfs_find_client(addr, 4);
 	if (clp == NULL)
 		return SVC_DROP;
-	dprintk("%s: %u.%u.%u.%u NFSv4 callback!\n", __FUNCTION__, NIPQUAD(addr));
-	nfs4_put_client(clp);
+	dprintk("%s: %u.%u.%u.%u NFSv4 callback!\n", __FUNCTION__, NIPQUAD(addr->sin_addr));
+	nfs_put_client(clp);
 	switch (rqstp->rq_authop->flavour) {
 		case RPC_AUTH_NULL:
 			if (rqstp->rq_proc != CB_NULL)
diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c
index 55d6e2ec157f..97cf8f71451f 100644
--- a/fs/nfs/callback_proc.c
+++ b/fs/nfs/callback_proc.c
@@ -10,6 +10,7 @@
 #include "nfs4_fs.h"
 #include "callback.h"
 #include "delegation.h"
+#include "internal.h"
 
 #define NFSDBG_FACILITY NFSDBG_CALLBACK
  
@@ -22,7 +23,7 @@ unsigned nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres
 	
 	res->bitmap[0] = res->bitmap[1] = 0;
 	res->status = htonl(NFS4ERR_BADHANDLE);
-	clp = nfs4_find_client(&args->addr->sin_addr);
+	clp = nfs_find_client(args->addr, 4);
 	if (clp == NULL)
 		goto out;
 	inode = nfs_delegation_find_inode(clp, &args->fh);
@@ -48,7 +49,7 @@ out_iput:
 	up_read(&nfsi->rwsem);
 	iput(inode);
 out_putclient:
-	nfs4_put_client(clp);
+	nfs_put_client(clp);
 out:
 	dprintk("%s: exit with status = %d\n", __FUNCTION__, ntohl(res->status));
 	return res->status;
@@ -61,7 +62,7 @@ unsigned nfs4_callback_recall(struct cb_recallargs *args, void *dummy)
 	unsigned res;
 	
 	res = htonl(NFS4ERR_BADHANDLE);
-	clp = nfs4_find_client(&args->addr->sin_addr);
+	clp = nfs_find_client(args->addr, 4);
 	if (clp == NULL)
 		goto out;
 	inode = nfs_delegation_find_inode(clp, &args->fh);
@@ -80,7 +81,7 @@ unsigned nfs4_callback_recall(struct cb_recallargs *args, void *dummy)
 	}
 	iput(inode);
 out_putclient:
-	nfs4_put_client(clp);
+	nfs_put_client(clp);
 out:
 	dprintk("%s: exit with status = %d\n", __FUNCTION__, ntohl(res));
 	return res;
diff --git a/fs/nfs/client.c b/fs/nfs/client.c
new file mode 100644
index 000000000000..cb5e92463bdb
--- /dev/null
+++ b/fs/nfs/client.c
@@ -0,0 +1,312 @@
+/* client.c: NFS client sharing and management code
+ *
+ * Copyright (C) 2006 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells@redhat.com)
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version
+ * 2 of the License, or (at your option) any later version.
+ */
+
+
+#include <linux/config.h>
+#include <linux/module.h>
+#include <linux/init.h>
+
+#include <linux/time.h>
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/string.h>
+#include <linux/stat.h>
+#include <linux/errno.h>
+#include <linux/unistd.h>
+#include <linux/sunrpc/clnt.h>
+#include <linux/sunrpc/stats.h>
+#include <linux/sunrpc/metrics.h>
+#include <linux/nfs_fs.h>
+#include <linux/nfs_mount.h>
+#include <linux/nfs4_mount.h>
+#include <linux/lockd/bind.h>
+#include <linux/smp_lock.h>
+#include <linux/seq_file.h>
+#include <linux/mount.h>
+#include <linux/nfs_idmap.h>
+#include <linux/vfs.h>
+#include <linux/inet.h>
+#include <linux/nfs_xdr.h>
+
+#include <asm/system.h>
+
+#include "nfs4_fs.h"
+#include "callback.h"
+#include "delegation.h"
+#include "iostat.h"
+#include "internal.h"
+
+#define NFSDBG_FACILITY		NFSDBG_CLIENT
+
+static DEFINE_SPINLOCK(nfs_client_lock);
+static LIST_HEAD(nfs_client_list);
+static DECLARE_WAIT_QUEUE_HEAD(nfs_client_active_wq);
+
+/*
+ * Allocate a shared client record
+ *
+ * Since these are allocated/deallocated very rarely, we don't
+ * bother putting them in a slab cache...
+ */
+static struct nfs_client *nfs_alloc_client(const char *hostname,
+					   const struct sockaddr_in *addr,
+					   int nfsversion)
+{
+	struct nfs_client *clp;
+	int error;
+
+	if ((clp = kzalloc(sizeof(*clp), GFP_KERNEL)) == NULL)
+		goto error_0;
+
+	error = rpciod_up();
+	if (error < 0) {
+		dprintk("%s: couldn't start rpciod! Error = %d\n",
+				__FUNCTION__, error);
+		__set_bit(NFS_CS_RPCIOD, &clp->cl_res_state);
+		goto error_1;
+	}
+
+	if (nfsversion == 4) {
+		if (nfs_callback_up() < 0)
+			goto error_2;
+		__set_bit(NFS_CS_CALLBACK, &clp->cl_res_state);
+	}
+
+	atomic_set(&clp->cl_count, 1);
+	clp->cl_cons_state = NFS_CS_INITING;
+
+	clp->cl_nfsversion = nfsversion;
+	memcpy(&clp->cl_addr, addr, sizeof(clp->cl_addr));
+
+	if (hostname) {
+		clp->cl_hostname = kstrdup(hostname, GFP_KERNEL);
+		if (!clp->cl_hostname)
+			goto error_3;
+	}
+
+	INIT_LIST_HEAD(&clp->cl_superblocks);
+	clp->cl_rpcclient = ERR_PTR(-EINVAL);
+
+#ifdef CONFIG_NFS_V4
+	init_rwsem(&clp->cl_sem);
+	INIT_LIST_HEAD(&clp->cl_delegations);
+	INIT_LIST_HEAD(&clp->cl_state_owners);
+	INIT_LIST_HEAD(&clp->cl_unused);
+	spin_lock_init(&clp->cl_lock);
+	INIT_WORK(&clp->cl_renewd, nfs4_renew_state, clp);
+	rpc_init_wait_queue(&clp->cl_rpcwaitq, "NFS client");
+	clp->cl_boot_time = CURRENT_TIME;
+	clp->cl_state = 1 << NFS4CLNT_LEASE_EXPIRED;
+#endif
+
+	return clp;
+
+error_3:
+	nfs_callback_down();
+	__clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state);
+error_2:
+	rpciod_down();
+	__clear_bit(NFS_CS_RPCIOD, &clp->cl_res_state);
+error_1:
+	kfree(clp);
+error_0:
+	return NULL;
+}
+
+/*
+ * Destroy a shared client record
+ */
+static void nfs_free_client(struct nfs_client *clp)
+{
+	dprintk("--> nfs_free_client(%d)\n", clp->cl_nfsversion);
+
+#ifdef CONFIG_NFS_V4
+	if (__test_and_clear_bit(NFS_CS_IDMAP, &clp->cl_res_state)) {
+		while (!list_empty(&clp->cl_unused)) {
+			struct nfs4_state_owner *sp;
+
+			sp = list_entry(clp->cl_unused.next,
+					struct nfs4_state_owner,
+					so_list);
+			list_del(&sp->so_list);
+			kfree(sp);
+		}
+		BUG_ON(!list_empty(&clp->cl_state_owners));
+		nfs_idmap_delete(clp);
+	}
+#endif
+
+	/* -EIO all pending I/O */
+	if (!IS_ERR(clp->cl_rpcclient))
+		rpc_shutdown_client(clp->cl_rpcclient);
+
+	if (__test_and_clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state))
+		nfs_callback_down();
+
+	if (__test_and_clear_bit(NFS_CS_RPCIOD, &clp->cl_res_state))
+	rpciod_down();
+
+	kfree(clp->cl_hostname);
+	kfree(clp);
+
+	dprintk("<-- nfs_free_client()\n");
+}
+
+/*
+ * Release a reference to a shared client record
+ */
+void nfs_put_client(struct nfs_client *clp)
+{
+	dprintk("--> nfs_put_client({%d})\n", atomic_read(&clp->cl_count));
+
+	if (atomic_dec_and_lock(&clp->cl_count, &nfs_client_lock)) {
+		list_del(&clp->cl_share_link);
+		spin_unlock(&nfs_client_lock);
+
+		BUG_ON(!list_empty(&clp->cl_superblocks));
+
+		nfs_free_client(clp);
+	}
+}
+
+/*
+ * Find a client by address
+ * - caller must hold nfs_client_lock
+ */
+static struct nfs_client *__nfs_find_client(const struct sockaddr_in *addr, int nfsversion)
+{
+	struct nfs_client *clp;
+
+	list_for_each_entry(clp, &nfs_client_list, cl_share_link) {
+		/* Different NFS versions cannot share the same nfs_client */
+		if (clp->cl_nfsversion != nfsversion)
+			continue;
+
+		if (memcmp(&clp->cl_addr.sin_addr, &addr->sin_addr,
+			   sizeof(clp->cl_addr.sin_addr)) != 0)
+			continue;
+
+		if (clp->cl_addr.sin_port == addr->sin_port)
+			goto found;
+	}
+
+	return NULL;
+
+found:
+	atomic_inc(&clp->cl_count);
+	return clp;
+}
+
+/*
+ * Find a client by IP address and protocol version
+ * - returns NULL if no such client
+ */
+struct nfs_client *nfs_find_client(const struct sockaddr_in *addr, int nfsversion)
+{
+	struct nfs_client *clp;
+
+	spin_lock(&nfs_client_lock);
+	clp = __nfs_find_client(addr, nfsversion);
+	spin_unlock(&nfs_client_lock);
+
+	BUG_ON(clp->cl_cons_state == 0);
+
+	return clp;
+}
+
+/*
+ * Look up a client by IP address and protocol version
+ * - creates a new record if one doesn't yet exist
+ */
+struct nfs_client *nfs_get_client(const char *hostname,
+				  const struct sockaddr_in *addr,
+				  int nfsversion)
+{
+	struct nfs_client *clp, *new = NULL;
+	int error;
+
+	dprintk("--> nfs_get_client(%s,"NIPQUAD_FMT":%d,%d)\n",
+		hostname ?: "", NIPQUAD(addr->sin_addr),
+		addr->sin_port, nfsversion);
+
+	/* see if the client already exists */
+	do {
+		spin_lock(&nfs_client_lock);
+
+		clp = __nfs_find_client(addr, nfsversion);
+		if (clp)
+			goto found_client;
+		if (new)
+			goto install_client;
+
+		spin_unlock(&nfs_client_lock);
+
+		new = nfs_alloc_client(hostname, addr, nfsversion);
+	} while (new);
+
+	return ERR_PTR(-ENOMEM);
+
+	/* install a new client and return with it unready */
+install_client:
+	clp = new;
+	list_add(&clp->cl_share_link, &nfs_client_list);
+	spin_unlock(&nfs_client_lock);
+	dprintk("--> nfs_get_client() = %p [new]\n", clp);
+	return clp;
+
+	/* found an existing client
+	 * - make sure it's ready before returning
+	 */
+found_client:
+	spin_unlock(&nfs_client_lock);
+
+	if (new)
+		nfs_free_client(new);
+
+	if (clp->cl_cons_state == NFS_CS_INITING) {
+		DECLARE_WAITQUEUE(myself, current);
+
+		add_wait_queue(&nfs_client_active_wq, &myself);
+
+		for (;;) {
+			set_current_state(TASK_INTERRUPTIBLE);
+			if (signal_pending(current) ||
+			    clp->cl_cons_state > NFS_CS_READY)
+				break;
+			schedule();
+		}
+
+		remove_wait_queue(&nfs_client_active_wq, &myself);
+
+		if (signal_pending(current)) {
+			nfs_put_client(clp);
+			return ERR_PTR(-ERESTARTSYS);
+		}
+	}
+
+	if (clp->cl_cons_state < NFS_CS_READY) {
+		error = clp->cl_cons_state;
+		nfs_put_client(clp);
+		return ERR_PTR(error);
+	}
+
+	dprintk("--> nfs_get_client() = %p [share]\n", clp);
+	return clp;
+}
+
+/*
+ * Mark a server as ready or failed
+ */
+void nfs_mark_client_ready(struct nfs_client *clp, int state)
+{
+	clp->cl_cons_state = state;
+	wake_up_all(&nfs_client_active_wq);
+}
diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c
index cfe239736ac0..57133678db16 100644
--- a/fs/nfs/delegation.c
+++ b/fs/nfs/delegation.c
@@ -18,6 +18,7 @@
 
 #include "nfs4_fs.h"
 #include "delegation.h"
+#include "internal.h"
 
 static struct nfs_delegation *nfs_alloc_delegation(void)
 {
@@ -145,7 +146,7 @@ int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct
 					sizeof(delegation->stateid)) != 0 ||
 				delegation->type != nfsi->delegation->type) {
 			printk("%s: server %u.%u.%u.%u, handed out a duplicate delegation!\n",
-					__FUNCTION__, NIPQUAD(clp->cl_addr));
+					__FUNCTION__, NIPQUAD(clp->cl_addr.sin_addr));
 			status = -EIO;
 		}
 	}
@@ -254,7 +255,7 @@ restart:
 	}
 out:
 	spin_unlock(&clp->cl_lock);
-	nfs4_put_client(clp);
+	nfs_put_client(clp);
 	module_put_and_exit(0);
 }
 
@@ -266,10 +267,10 @@ void nfs_expire_all_delegations(struct nfs_client *clp)
 	atomic_inc(&clp->cl_count);
 	task = kthread_run(nfs_do_expire_all_delegations, clp,
 			"%u.%u.%u.%u-delegreturn",
-			NIPQUAD(clp->cl_addr));
+			NIPQUAD(clp->cl_addr.sin_addr));
 	if (!IS_ERR(task))
 		return;
-	nfs4_put_client(clp);
+	nfs_put_client(clp);
 	module_put(THIS_MODULE);
 }
 
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index 4802157963f8..ac370d5d4494 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -15,6 +15,12 @@ struct nfs_clone_mount {
 	rpc_authflavor_t authflavor;
 };
 
+/* client.c */
+extern void nfs_put_client(struct nfs_client *);
+extern struct nfs_client *nfs_find_client(const struct sockaddr_in *, int);
+extern struct nfs_client *nfs_get_client(const char *, const struct sockaddr_in *, int);
+extern void nfs_mark_client_ready(struct nfs_client *, int);
+
 /* nfs4namespace.c */
 #ifdef CONFIG_NFS_V4
 extern struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentry *dentry);
diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h
index 4e334cb48498..e7879245361e 100644
--- a/fs/nfs/nfs4_fs.h
+++ b/fs/nfs/nfs4_fs.h
@@ -42,55 +42,6 @@ enum nfs4_client_state {
 	NFS4CLNT_LEASE_EXPIRED,
 };
 
-/*
- * The nfs_client identifies our client state to the server.
- */
-struct nfs_client {
-	struct list_head	cl_servers;	/* Global list of servers */
-	struct in_addr		cl_addr;	/* Server identifier */
-	u64			cl_clientid;	/* constant */
-	nfs4_verifier		cl_confirm;
-	unsigned long		cl_state;
-
-	u32			cl_lockowner_id;
-
-	/*
-	 * The following rwsem ensures exclusive access to the server
-	 * while we recover the state following a lease expiration.
-	 */
-	struct rw_semaphore	cl_sem;
-
-	struct list_head	cl_delegations;
-	struct list_head	cl_state_owners;
-	struct list_head	cl_unused;
-	int			cl_nunused;
-	spinlock_t		cl_lock;
-	atomic_t		cl_count;
-
-	struct rpc_clnt *	cl_rpcclient;
-
-	struct list_head	cl_superblocks;	/* List of nfs_server structs */
-
-	unsigned long		cl_lease_time;
-	unsigned long		cl_last_renewal;
-	struct work_struct	cl_renewd;
-	struct work_struct	cl_recoverd;
-
-	struct rpc_wait_queue	cl_rpcwaitq;
-
-	/* used for the setclientid verifier */
-	struct timespec		cl_boot_time;
-
-	/* idmapper */
-	struct idmap *		cl_idmap;
-
-	/* Our own IP address, as a null-terminated string.
-	 * This is used to generate the clientid, and the callback address.
-	 */
-	char			cl_ipaddr[16];
-	unsigned char		cl_id_uniquifier;
-};
-
 /*
  * struct rpc_sequence ensures that RPC calls are sent in the exact
  * order that they appear on the list.
@@ -239,9 +190,6 @@ extern void nfs4_renew_state(void *);
 /* nfs4state.c */
 extern void init_nfsv4_state(struct nfs_server *);
 extern void destroy_nfsv4_state(struct nfs_server *);
-extern struct nfs_client *nfs4_get_client(struct in_addr *);
-extern void nfs4_put_client(struct nfs_client *clp);
-extern struct nfs_client *nfs4_find_client(struct in_addr *);
 struct rpc_cred *nfs4_get_renew_cred(struct nfs_client *clp);
 extern u32 nfs4_alloc_lockowner_id(struct nfs_client *);
 
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 850f0851023a..803c31b88bb5 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -2968,7 +2968,7 @@ int nfs4_proc_setclientid(struct nfs_client *clp, u32 program, unsigned short po
 	for(;;) {
 		setclientid.sc_name_len = scnprintf(setclientid.sc_name,
 				sizeof(setclientid.sc_name), "%s/%u.%u.%u.%u %s %u",
-				clp->cl_ipaddr, NIPQUAD(clp->cl_addr.s_addr),
+				clp->cl_ipaddr, NIPQUAD(clp->cl_addr.sin_addr),
 				cred->cr_ops->cr_name,
 				clp->cl_id_uniquifier);
 		setclientid.sc_netid_len = scnprintf(setclientid.sc_netid,
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index fa51a7d4c022..058811e39555 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -50,12 +50,12 @@
 #include "nfs4_fs.h"
 #include "callback.h"
 #include "delegation.h"
+#include "internal.h"
 
 #define OPENOWNER_POOL_SIZE	8
 
 const nfs4_stateid zero_stateid;
 
-static DEFINE_SPINLOCK(state_spinlock);
 static LIST_HEAD(nfs4_clientid_list);
 
 void
@@ -71,127 +71,11 @@ destroy_nfsv4_state(struct nfs_server *server)
 	kfree(server->mnt_path);
 	server->mnt_path = NULL;
 	if (server->nfs_client) {
-		nfs4_put_client(server->nfs_client);
+		nfs_put_client(server->nfs_client);
 		server->nfs_client = NULL;
 	}
 }
 
-/*
- * nfs4_get_client(): returns an empty client structure
- * nfs4_put_client(): drops reference to client structure
- *
- * Since these are allocated/deallocated very rarely, we don't
- * bother putting them in a slab cache...
- */
-static struct nfs_client *
-nfs4_alloc_client(struct in_addr *addr)
-{
-	struct nfs_client *clp;
-
-	if (nfs_callback_up() < 0)
-		return NULL;
-	if ((clp = kzalloc(sizeof(*clp), GFP_KERNEL)) == NULL) {
-		nfs_callback_down();
-		return NULL;
-	}
-	memcpy(&clp->cl_addr, addr, sizeof(clp->cl_addr));
-	init_rwsem(&clp->cl_sem);
-	INIT_LIST_HEAD(&clp->cl_delegations);
-	INIT_LIST_HEAD(&clp->cl_state_owners);
-	INIT_LIST_HEAD(&clp->cl_unused);
-	spin_lock_init(&clp->cl_lock);
-	atomic_set(&clp->cl_count, 1);
-	INIT_WORK(&clp->cl_renewd, nfs4_renew_state, clp);
-	INIT_LIST_HEAD(&clp->cl_superblocks);
-	rpc_init_wait_queue(&clp->cl_rpcwaitq, "NFS4 client");
-	clp->cl_rpcclient = ERR_PTR(-EINVAL);
-	clp->cl_boot_time = CURRENT_TIME;
-	clp->cl_state = 1 << NFS4CLNT_LEASE_EXPIRED;
-	return clp;
-}
-
-static void
-nfs4_free_client(struct nfs_client *clp)
-{
-	struct nfs4_state_owner *sp;
-
-	while (!list_empty(&clp->cl_unused)) {
-		sp = list_entry(clp->cl_unused.next,
-				struct nfs4_state_owner,
-				so_list);
-		list_del(&sp->so_list);
-		kfree(sp);
-	}
-	BUG_ON(!list_empty(&clp->cl_state_owners));
-	nfs_idmap_delete(clp);
-	if (!IS_ERR(clp->cl_rpcclient))
-		rpc_shutdown_client(clp->cl_rpcclient);
-	kfree(clp);
-	nfs_callback_down();
-}
-
-static struct nfs_client *__nfs4_find_client(struct in_addr *addr)
-{
-	struct nfs_client *clp;
-	list_for_each_entry(clp, &nfs4_clientid_list, cl_servers) {
-		if (memcmp(&clp->cl_addr, addr, sizeof(clp->cl_addr)) == 0) {
-			atomic_inc(&clp->cl_count);
-			return clp;
-		}
-	}
-	return NULL;
-}
-
-struct nfs_client *nfs4_find_client(struct in_addr *addr)
-{
-	struct nfs_client *clp;
-	spin_lock(&state_spinlock);
-	clp = __nfs4_find_client(addr);
-	spin_unlock(&state_spinlock);
-	return clp;
-}
-
-struct nfs_client *
-nfs4_get_client(struct in_addr *addr)
-{
-	struct nfs_client *clp, *new = NULL;
-
-	spin_lock(&state_spinlock);
-	for (;;) {
-		clp = __nfs4_find_client(addr);
-		if (clp != NULL)
-			break;
-		clp = new;
-		if (clp != NULL) {
-			list_add(&clp->cl_servers, &nfs4_clientid_list);
-			new = NULL;
-			break;
-		}
-		spin_unlock(&state_spinlock);
-		new = nfs4_alloc_client(addr);
-		spin_lock(&state_spinlock);
-		if (new == NULL)
-			break;
-	}
-	spin_unlock(&state_spinlock);
-	if (new)
-		nfs4_free_client(new);
-	return clp;
-}
-
-void
-nfs4_put_client(struct nfs_client *clp)
-{
-	if (!atomic_dec_and_lock(&clp->cl_count, &state_spinlock))
-		return;
-	list_del(&clp->cl_servers);
-	spin_unlock(&state_spinlock);
-	BUG_ON(!list_empty(&clp->cl_superblocks));
-	rpc_wake_up(&clp->cl_rpcwaitq);
-	nfs4_kill_renewd(clp);
-	nfs4_free_client(clp);
-}
-
 static int nfs4_init_client(struct nfs_client *clp, struct rpc_cred *cred)
 {
 	int status = nfs4_proc_setclientid(clp, NFS4_CALLBACK,
@@ -771,11 +655,11 @@ static void nfs4_recover_state(struct nfs_client *clp)
 	__module_get(THIS_MODULE);
 	atomic_inc(&clp->cl_count);
 	task = kthread_run(reclaimer, clp, "%u.%u.%u.%u-reclaim",
-			NIPQUAD(clp->cl_addr));
+			NIPQUAD(clp->cl_addr.sin_addr));
 	if (!IS_ERR(task))
 		return;
 	nfs4_clear_recover_bit(clp);
-	nfs4_put_client(clp);
+	nfs_put_client(clp);
 	module_put(THIS_MODULE);
 }
 
@@ -970,12 +854,12 @@ out:
 	if (status == -NFS4ERR_CB_PATH_DOWN)
 		nfs_handle_cb_pathdown(clp);
 	nfs4_clear_recover_bit(clp);
-	nfs4_put_client(clp);
+	nfs_put_client(clp);
 	module_put_and_exit(0);
 	return 0;
 out_error:
 	printk(KERN_WARNING "Error: state recovery failed on NFSv4 server %u.%u.%u.%u with error %d\n",
-				NIPQUAD(clp->cl_addr.s_addr), -status);
+				NIPQUAD(clp->cl_addr.sin_addr), -status);
 	set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state);
 	goto out;
 }
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 3ee85c4e65d8..f97d7d9c5c32 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -1104,47 +1104,46 @@ static struct rpc_clnt *nfs4_create_client(struct nfs_server *server,
 	struct rpc_clnt *clnt = NULL;
 	int err = -EIO;
 
-	clp = nfs4_get_client(&server->addr.sin_addr);
+	clp = nfs_get_client(server->hostname, &server->addr, 4);
 	if (!clp) {
 		dprintk("%s: failed to create NFS4 client.\n", __FUNCTION__);
 		return ERR_PTR(err);
 	}
 
 	/* Now create transport and client */
-	down_write(&clp->cl_sem);
-	if (IS_ERR(clp->cl_rpcclient)) {
+	if (clp->cl_cons_state == NFS_CS_INITING) {
 		xprt = xprt_create_proto(proto, &server->addr, timeparms);
 		if (IS_ERR(xprt)) {
-			up_write(&clp->cl_sem);
 			err = PTR_ERR(xprt);
 			dprintk("%s: cannot create RPC transport. Error = %d\n",
 					__FUNCTION__, err);
-			goto out_fail;
+			goto client_init_error;
 		}
 		/* Bind to a reserved port! */
 		xprt->resvport = 1;
 		clnt = rpc_create_client(xprt, server->hostname, &nfs_program,
 				server->rpc_ops->version, flavor);
 		if (IS_ERR(clnt)) {
-			up_write(&clp->cl_sem);
 			err = PTR_ERR(clnt);
 			dprintk("%s: cannot create RPC client. Error = %d\n",
 					__FUNCTION__, err);
-			goto out_fail;
+			goto client_init_error;
 		}
 		clnt->cl_intr     = 1;
 		clnt->cl_softrtry = 1;
 		clp->cl_rpcclient = clnt;
 		memcpy(clp->cl_ipaddr, server->ip_addr, sizeof(clp->cl_ipaddr));
-		if (nfs_idmap_new(clp) < 0)
-			goto out_fail;
+		err = nfs_idmap_new(clp);
+		if (err < 0) {
+			dprintk("%s: failed to create idmapper.\n",
+				__FUNCTION__);
+			goto client_init_error;
+		}
+		__set_bit(NFS_CS_IDMAP, &clp->cl_res_state);
+		nfs_mark_client_ready(clp, 0);
 	}
-	list_add_tail(&server->nfs4_siblings, &clp->cl_superblocks);
+
 	clnt = rpc_clone_client(clp->cl_rpcclient);
-	if (!IS_ERR(clnt))
-		server->nfs_client = clp;
-	up_write(&clp->cl_sem);
-	clp = NULL;
 
 	if (IS_ERR(clnt)) {
 		dprintk("%s: cannot create RPC client. Error = %d\n",
@@ -1152,11 +1151,6 @@ static struct rpc_clnt *nfs4_create_client(struct nfs_server *server,
 		return clnt;
 	}
 
-	if (server->nfs_client->cl_idmap == NULL) {
-		dprintk("%s: failed to create idmapper.\n", __FUNCTION__);
-		return ERR_PTR(-ENOMEM);
-	}
-
 	if (clnt->cl_auth->au_flavor != flavor) {
 		struct rpc_auth *auth;
 
@@ -1166,11 +1160,16 @@ static struct rpc_clnt *nfs4_create_client(struct nfs_server *server,
 			return (struct rpc_clnt *)auth;
 		}
 	}
+
+	server->nfs_client = clp;
+	down_write(&clp->cl_sem);
+	list_add_tail(&server->nfs4_siblings, &clp->cl_superblocks);
+	up_write(&clp->cl_sem);
 	return clnt;
 
- out_fail:
-	if (clp)
-		nfs4_put_client(clp);
+client_init_error:
+	nfs_mark_client_ready(clp, err);
+	nfs_put_client(clp);
 	return ERR_PTR(err);
 }
 
@@ -1329,14 +1328,6 @@ static int nfs4_get_sb(struct file_system_type *fs_type,
 		goto out_free;
 	}
 
-	/* Fire up rpciod if not yet running */
-	error = rpciod_up();
-	if (error < 0) {
-		dprintk("%s: couldn't start rpciod! Error = %d\n",
-				__FUNCTION__, error);
-		goto out_free;
-	}
-
 	s = sget(fs_type, nfs4_compare_super, nfs_set_super, server);
 	if (IS_ERR(s)) {
 		error = PTR_ERR(s);
@@ -1383,8 +1374,6 @@ static void nfs4_kill_super(struct super_block *sb)
 
 	destroy_nfsv4_state(server);
 
-	rpciod_down();
-
 	nfs_free_iostats(server->io_stats);
 	kfree(server->hostname);
 	kfree(server);
diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h
index a36e01cd6321..70e1dc9162e2 100644
--- a/include/linux/nfs_fs.h
+++ b/include/linux/nfs_fs.h
@@ -586,6 +586,7 @@ extern void * nfs_root_data(void);
 #define NFSDBG_FILE		0x0040
 #define NFSDBG_ROOT		0x0080
 #define NFSDBG_CALLBACK		0x0100
+#define NFSDBG_CLIENT		0x0200
 #define NFSDBG_ALL		0xFFFF
 
 #ifdef __KERNEL__
diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index fc20d6b934fb..a727657e0ad3 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -6,6 +6,66 @@
 
 struct nfs_iostats;
 
+/*
+ * The nfs_client identifies our client state to the server.
+ */
+struct nfs_client {
+	atomic_t		cl_count;
+	int			cl_cons_state;	/* current construction state (-ve: init error) */
+#define NFS_CS_READY		0		/* ready to be used */
+#define NFS_CS_INITING		1		/* busy initialising */
+	int			cl_nfsversion;	/* NFS protocol version */
+	unsigned long		cl_res_state;	/* NFS resources state */
+#define NFS_CS_RPCIOD		0		/* - rpciod started */
+#define NFS_CS_CALLBACK		1		/* - callback started */
+#define NFS_CS_IDMAP		2		/* - idmap started */
+	struct sockaddr_in	cl_addr;	/* server identifier */
+	char *			cl_hostname;	/* hostname of server */
+	struct list_head	cl_share_link;	/* link in global client list */
+	struct list_head	cl_superblocks;	/* List of nfs_server structs */
+
+	struct rpc_clnt *	cl_rpcclient;
+
+#ifdef CONFIG_NFS_V4
+	u64			cl_clientid;	/* constant */
+	nfs4_verifier		cl_confirm;
+	unsigned long		cl_state;
+
+	u32			cl_lockowner_id;
+
+	/*
+	 * The following rwsem ensures exclusive access to the server
+	 * while we recover the state following a lease expiration.
+	 */
+	struct rw_semaphore	cl_sem;
+
+	struct list_head	cl_delegations;
+	struct list_head	cl_state_owners;
+	struct list_head	cl_unused;
+	int			cl_nunused;
+	spinlock_t		cl_lock;
+
+	unsigned long		cl_lease_time;
+	unsigned long		cl_last_renewal;
+	struct work_struct	cl_renewd;
+	struct work_struct	cl_recoverd;
+
+	struct rpc_wait_queue	cl_rpcwaitq;
+
+	/* used for the setclientid verifier */
+	struct timespec		cl_boot_time;
+
+	/* idmapper */
+	struct idmap *		cl_idmap;
+
+	/* Our own IP address, as a null-terminated string.
+	 * This is used to generate the clientid, and the callback address.
+	 */
+	char			cl_ipaddr[16];
+	unsigned char		cl_id_uniquifier;
+#endif
+};
+
 /*
  * NFS client parameters stored in the superblock.
  */
-- 
cgit v1.2.3


From 509de8111656a7d89b4a1a5f430f4460ce510f0f Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Tue, 22 Aug 2006 20:06:11 -0400
Subject: NFS: Add extra const qualifiers

Add some extra const qualifiers into NFS.

Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/namespace.c        |  3 ++-
 fs/nfs/nfs3proc.c         |  2 +-
 fs/nfs/nfs4namespace.c    |  8 ++++----
 fs/nfs/nfs4proc.c         |  2 +-
 fs/nfs/proc.c             |  2 +-
 fs/nfs/super.c            | 10 +++++-----
 include/linux/nfs_fs_sb.h |  2 +-
 include/linux/nfs_xdr.h   |  6 +++---
 8 files changed, 18 insertions(+), 17 deletions(-)

(limited to 'include')

diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c
index 86b3169c8cac..85d9ed1dcf42 100644
--- a/fs/nfs/namespace.c
+++ b/fs/nfs/namespace.c
@@ -172,7 +172,8 @@ void nfs_release_automount_timer(void)
 /*
  * Clone a mountpoint of the appropriate type
  */
-static struct vfsmount *nfs_do_clone_mount(struct nfs_server *server, char *devname,
+static struct vfsmount *nfs_do_clone_mount(struct nfs_server *server,
+					   const char *devname,
 					   struct nfs_clone_mount *mountdata)
 {
 #ifdef CONFIG_NFS_V4
diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c
index 7143b1f82cea..3e5371241cea 100644
--- a/fs/nfs/nfs3proc.c
+++ b/fs/nfs/nfs3proc.c
@@ -886,7 +886,7 @@ nfs3_proc_lock(struct file *filp, int cmd, struct file_lock *fl)
 	return nlmclnt_proc(filp->f_dentry->d_inode, cmd, fl);
 }
 
-struct nfs_rpc_ops	nfs_v3_clientops = {
+const struct nfs_rpc_ops nfs_v3_clientops = {
 	.version	= 3,			/* protocol version */
 	.dentry_ops	= &nfs_dentry_operations,
 	.dir_inode_ops	= &nfs3_dir_inode_operations,
diff --git a/fs/nfs/nfs4namespace.c b/fs/nfs/nfs4namespace.c
index ea38d27b74e6..faed9bcba50f 100644
--- a/fs/nfs/nfs4namespace.c
+++ b/fs/nfs/nfs4namespace.c
@@ -23,7 +23,7 @@
 /*
  * Check if fs_root is valid
  */
-static inline char *nfs4_pathname_string(struct nfs4_pathname *pathname,
+static inline char *nfs4_pathname_string(const struct nfs4_pathname *pathname,
 					 char *buffer, ssize_t buflen)
 {
 	char *end = buffer + buflen;
@@ -34,7 +34,7 @@ static inline char *nfs4_pathname_string(struct nfs4_pathname *pathname,
 
 	n = pathname->ncomponents;
 	while (--n >= 0) {
-		struct nfs4_string *component = &pathname->components[n];
+		const struct nfs4_string *component = &pathname->components[n];
 		buflen -= component->len + 1;
 		if (buflen < 0)
 			goto Elong;
@@ -60,7 +60,7 @@ Elong:
  */
 static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent,
 					    const struct dentry *dentry,
-					    struct nfs4_fs_locations *locations)
+					    const struct nfs4_fs_locations *locations)
 {
 	struct vfsmount *mnt = ERR_PTR(-ENOENT);
 	struct nfs_clone_mount mountdata = {
@@ -108,7 +108,7 @@ static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent,
 
 	loc = 0;
 	while (loc < locations->nlocations && IS_ERR(mnt)) {
-		struct nfs4_fs_location *location = &locations->locations[loc];
+		const struct nfs4_fs_location *location = &locations->locations[loc];
 		char *mnt_path;
 
 		if (location == NULL || location->nservers <= 0 ||
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 803c31b88bb5..061be713b206 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -3761,7 +3761,7 @@ static struct inode_operations nfs4_file_inode_operations = {
 	.listxattr	= nfs4_listxattr,
 };
 
-struct nfs_rpc_ops	nfs_v4_clientops = {
+const struct nfs_rpc_ops nfs_v4_clientops = {
 	.version	= 4,			/* protocol version */
 	.dentry_ops	= &nfs4_dentry_operations,
 	.dir_inode_ops	= &nfs4_dir_inode_operations,
diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c
index b3899ea3229e..77676903e0f5 100644
--- a/fs/nfs/proc.c
+++ b/fs/nfs/proc.c
@@ -671,7 +671,7 @@ nfs_proc_lock(struct file *filp, int cmd, struct file_lock *fl)
 }
 
 
-struct nfs_rpc_ops	nfs_v2_clientops = {
+const struct nfs_rpc_ops nfs_v2_clientops = {
 	.version	= 2,		       /* protocol version */
 	.dentry_ops	= &nfs_dentry_operations,
 	.dir_inode_ops	= &nfs_dir_inode_operations,
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index a41d516ed595..c97f30967955 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -329,10 +329,10 @@ static const char *nfs_pseudoflavour_to_name(rpc_authflavor_t flavour)
  */
 static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss, int showdefaults)
 {
-	static struct proc_nfs_info {
+	static const struct proc_nfs_info {
 		int flag;
-		char *str;
-		char *nostr;
+		const char *str;
+		const char *nostr;
 	} nfs_info[] = {
 		{ NFS_MOUNT_SOFT, ",soft", ",hard" },
 		{ NFS_MOUNT_INTR, ",intr", "" },
@@ -342,9 +342,9 @@ static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss,
 		{ NFS_MOUNT_NOACL, ",noacl", "" },
 		{ 0, NULL, NULL }
 	};
-	struct proc_nfs_info *nfs_infop;
+	const struct proc_nfs_info *nfs_infop;
 	char buf[12];
-	char *proto;
+	const char *proto;
 
 	seq_printf(m, ",vers=%d", nfss->rpc_ops->version);
 	seq_printf(m, ",rsize=%d", nfss->rsize);
diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index a727657e0ad3..95f32d5f6e9c 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -73,7 +73,7 @@ struct nfs_server {
 	struct rpc_clnt *	client;		/* RPC client handle */
 	struct rpc_clnt *	client_sys;	/* 2nd handle for FSINFO */
 	struct rpc_clnt *	client_acl;	/* ACL RPC client handle */
-	struct nfs_rpc_ops *	rpc_ops;	/* NFS protocol vector */
+	const struct nfs_rpc_ops *rpc_ops;	/* NFS protocol vector */
 	struct nfs_iostats *	io_stats;	/* I/O statistics */
 	struct backing_dev_info	backing_dev_info;
 	int			flags;		/* various flags */
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index dd9ae6761f71..2426b11b6cce 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -833,9 +833,9 @@ struct nfs_rpc_ops {
 /*
  * Function vectors etc. for the NFS client
  */
-extern struct nfs_rpc_ops	nfs_v2_clientops;
-extern struct nfs_rpc_ops	nfs_v3_clientops;
-extern struct nfs_rpc_ops	nfs_v4_clientops;
+extern const struct nfs_rpc_ops	nfs_v2_clientops;
+extern const struct nfs_rpc_ops	nfs_v3_clientops;
+extern const struct nfs_rpc_ops	nfs_v4_clientops;
 extern struct rpc_version	nfs_version2;
 extern struct rpc_version	nfs_version3;
 extern struct rpc_version	nfs_version4;
-- 
cgit v1.2.3


From 27951bd26031f6c27d38df9e94623bbe208a2464 Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Tue, 22 Aug 2006 20:06:11 -0400
Subject: NFS: Maintain a common server record for NFS2/3 as well as for NFS4

Maintain a common server record for NFS2/3 as well as for NFS4 so that common
stuff can be moved there from struct nfs_server.

Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/super.c            | 21 ++++++++++++++++++++-
 include/linux/nfs_fs_sb.h |  2 +-
 2 files changed, 21 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index c97f30967955..d1b4a5b36e33 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -658,11 +658,19 @@ static void nfs_init_timeout_values(struct rpc_timeout *to, int proto, unsigned
 static struct rpc_clnt *
 nfs_create_client(struct nfs_server *server, const struct nfs_mount_data *data)
 {
+	struct nfs_client	*clp;
 	struct rpc_timeout	timeparms;
 	struct rpc_xprt		*xprt = NULL;
 	struct rpc_clnt		*clnt = NULL;
 	int			proto = (data->flags & NFS_MOUNT_TCP) ? IPPROTO_TCP : IPPROTO_UDP;
 
+	clp = nfs_get_client(server->hostname, &server->addr,
+			     server->rpc_ops->version);
+	if (!clp) {
+		dprintk("%s: failed to create NFS4 client.\n", __FUNCTION__);
+		return ERR_PTR(PTR_ERR(clp));
+	}
+
 	nfs_init_timeout_values(&timeparms, proto, data->timeo, data->retrans);
 
 	server->retrans_timeo = timeparms.to_initval;
@@ -673,6 +681,8 @@ nfs_create_client(struct nfs_server *server, const struct nfs_mount_data *data)
 	if (IS_ERR(xprt)) {
 		dprintk("%s: cannot create RPC transport. Error = %ld\n",
 				__FUNCTION__, PTR_ERR(xprt));
+		nfs_mark_client_ready(clp, PTR_ERR(xprt));
+		nfs_put_client(clp);
 		return (struct rpc_clnt *)xprt;
 	}
 	clnt = rpc_create_client(xprt, server->hostname, &nfs_program,
@@ -686,9 +696,13 @@ nfs_create_client(struct nfs_server *server, const struct nfs_mount_data *data)
 	clnt->cl_intr     = 1;
 	clnt->cl_softrtry = 1;
 
+	nfs_mark_client_ready(clp, 0);
+	server->nfs_client = clp;
 	return clnt;
 
 out_fail:
+	nfs_mark_client_ready(clp, PTR_ERR(xprt));
+	nfs_put_client(clp);
 	return clnt;
 }
 
@@ -764,6 +778,7 @@ static int nfs_clone_generic_sb(struct nfs_clone_mount *data,
 	if (server == NULL)
 		goto out_err;
 	memcpy(server, parent, sizeof(*server));
+	atomic_inc(&server->nfs_client->cl_count);
 	hostname = (data->hostname != NULL) ? data->hostname : parent->hostname;
 	len = strlen(hostname) + 1;
 	server->hostname = kmalloc(len, GFP_KERNEL);
@@ -796,6 +811,7 @@ out_deactivate:
 out_rpciod_down:
 	rpciod_down();
 	kfree(server->hostname);
+	nfs_put_client(server->nfs_client);
 	kfree(server);
 	return simple_set_mnt(mnt, sb);
 kill_rpciod:
@@ -803,6 +819,7 @@ kill_rpciod:
 free_hostname:
 	kfree(server->hostname);
 free_server:
+	nfs_put_client(server->nfs_client);
 	kfree(server);
 out_err:
 	return error;
@@ -1071,6 +1088,7 @@ static void nfs_kill_super(struct super_block *s)
 
 	nfs_free_iostats(server->io_stats);
 	kfree(server->hostname);
+	nfs_put_client(server->nfs_client);
 	kfree(server);
 	nfs_release_automount_timer();
 }
@@ -1421,7 +1439,6 @@ static struct super_block *nfs4_clone_sb(struct nfs_server *server, struct nfs_c
 	nfs4_server_capabilities(server, &server->fh);
 
 	down_write(&clp->cl_sem);
-	atomic_inc(&clp->cl_count);
 	list_add_tail(&server->nfs4_siblings, &clp->cl_superblocks);
 	up_write(&clp->cl_sem);
 	return sb;
@@ -1476,6 +1493,8 @@ static struct nfs_server *nfs4_referral_server(struct super_block *sb, struct nf
 	retrans = 1;
 	nfs_init_timeout_values(&timeparms, proto, timeo, retrans);
 
+	nfs_put_client(server->nfs_client);
+	server->nfs_client = NULL;
 	server->client = nfs4_create_client(server, &timeparms, proto, data->authflavor);
 	if (IS_ERR((err = server->client)))
 		goto out_err;
diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index 95f32d5f6e9c..e7d7662f51fd 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -70,6 +70,7 @@ struct nfs_client {
  * NFS client parameters stored in the superblock.
  */
 struct nfs_server {
+	struct nfs_client *	nfs_client;	/* shared client and NFS4 state */
 	struct rpc_clnt *	client;		/* RPC client handle */
 	struct rpc_clnt *	client_sys;	/* 2nd handle for FSINFO */
 	struct rpc_clnt *	client_acl;	/* ACL RPC client handle */
@@ -103,7 +104,6 @@ struct nfs_server {
 	 */
 	char			ip_addr[16];
 	char *			mnt_path;
-	struct nfs_client *	nfs_client;	/* all NFSv4 state starts here */
 	struct list_head	nfs4_siblings;	/* List of other nfs_server structs
 						 * that share the same clientid
 						 */
-- 
cgit v1.2.3


From 8fa5c000d7f986ef9cdc6d95f9f7fcee20e0a7d6 Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Tue, 22 Aug 2006 20:06:12 -0400
Subject: NFS: Move rpc_ops from nfs_server to nfs_client

Move the rpc_ops from the nfs_server struct to the nfs_client struct as they're
common to all server records of a particular NFS protocol version.

Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/dir.c              |  2 +-
 fs/nfs/inode.c            |  4 ++--
 fs/nfs/namespace.c        |  6 +++--
 fs/nfs/nfs4proc.c         |  2 +-
 fs/nfs/super.c            | 59 ++++++++++++++++++++++++++---------------------
 include/linux/nfs_fs.h    |  2 +-
 include/linux/nfs_fs_sb.h |  2 +-
 7 files changed, 43 insertions(+), 34 deletions(-)

(limited to 'include')

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 067d144d141b..19362712452f 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -1147,7 +1147,7 @@ int nfs_instantiate(struct dentry *dentry, struct nfs_fh *fhandle,
 	}
 	if (!(fattr->valid & NFS_ATTR_FATTR)) {
 		struct nfs_server *server = NFS_SB(dentry->d_sb);
-		error = server->rpc_ops->getattr(server, fhandle, fattr);
+		error = server->nfs_client->rpc_ops->getattr(server, fhandle, fattr);
 		if (error < 0)
 			goto out_err;
 	}
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 6ed018c9aad2..771c3b833757 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -237,13 +237,13 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr)
 		/* Why so? Because we want revalidate for devices/FIFOs, and
 		 * that's precisely what we have in nfs_file_inode_operations.
 		 */
-		inode->i_op = NFS_SB(sb)->rpc_ops->file_inode_ops;
+		inode->i_op = NFS_SB(sb)->nfs_client->rpc_ops->file_inode_ops;
 		if (S_ISREG(inode->i_mode)) {
 			inode->i_fop = &nfs_file_operations;
 			inode->i_data.a_ops = &nfs_file_aops;
 			inode->i_data.backing_dev_info = &NFS_SB(sb)->backing_dev_info;
 		} else if (S_ISDIR(inode->i_mode)) {
-			inode->i_op = NFS_SB(sb)->rpc_ops->dir_inode_ops;
+			inode->i_op = NFS_SB(sb)->nfs_client->rpc_ops->dir_inode_ops;
 			inode->i_fop = &nfs_dir_operations;
 			if (nfs_server_capable(inode, NFS_CAP_READDIRPLUS)
 			    && fattr->size <= NFS_LIMIT_READDIRPLUS)
diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c
index 85d9ed1dcf42..d8b8d56266cb 100644
--- a/fs/nfs/namespace.c
+++ b/fs/nfs/namespace.c
@@ -104,7 +104,9 @@ static void * nfs_follow_mountpoint(struct dentry *dentry, struct nameidata *nd)
 		goto out_follow;
 	/* Look it up again */
 	parent = dget_parent(nd->dentry);
-	err = server->rpc_ops->lookup(parent->d_inode, &nd->dentry->d_name, &fh, &fattr);
+	err = server->nfs_client->rpc_ops->lookup(parent->d_inode,
+						  &nd->dentry->d_name,
+						  &fh, &fattr);
 	dput(parent);
 	if (err != 0)
 		goto out_err;
@@ -178,7 +180,7 @@ static struct vfsmount *nfs_do_clone_mount(struct nfs_server *server,
 {
 #ifdef CONFIG_NFS_V4
 	struct vfsmount *mnt = NULL;
-	switch (server->rpc_ops->version) {
+	switch (server->nfs_client->cl_nfsversion) {
 		case 2:
 		case 3:
 			mnt = vfs_kern_mount(&clone_nfs_fs_type, 0, devname, mountdata);
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index b731b1945270..1573eeb07ce1 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -758,7 +758,7 @@ static int _nfs4_proc_open(struct nfs4_opendata *data)
 	}
 	nfs_confirm_seqid(&data->owner->so_seqid, 0);
 	if (!(o_res->f_attr->valid & NFS_ATTR_FATTR))
-		return server->rpc_ops->getattr(server, &o_res->fh, o_res->f_attr);
+		return server->nfs_client->rpc_ops->getattr(server, &o_res->fh, o_res->f_attr);
 	return 0;
 }
 
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index d1b4a5b36e33..e1e5eab0259b 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -252,7 +252,7 @@ static int nfs_statfs(struct dentry *dentry, struct kstatfs *buf)
 
 	lock_kernel();
 
-	error = server->rpc_ops->statfs(server, fh, &res);
+	error = server->nfs_client->rpc_ops->statfs(server, fh, &res);
 	buf->f_type = NFS_SUPER_MAGIC;
 	if (error < 0)
 		goto out_err;
@@ -343,10 +343,11 @@ static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss,
 		{ 0, NULL, NULL }
 	};
 	const struct proc_nfs_info *nfs_infop;
+	struct nfs_client *clp = nfss->nfs_client;
 	char buf[12];
 	const char *proto;
 
-	seq_printf(m, ",vers=%d", nfss->rpc_ops->version);
+	seq_printf(m, ",vers=%d", clp->rpc_ops->version);
 	seq_printf(m, ",rsize=%d", nfss->rsize);
 	seq_printf(m, ",wsize=%d", nfss->wsize);
 	if (nfss->acregmin != 3*HZ || showdefaults)
@@ -427,7 +428,7 @@ static int nfs_show_stats(struct seq_file *m, struct vfsmount *mnt)
 	seq_printf(m, ",namelen=%d", nfss->namelen);
 
 #ifdef CONFIG_NFS_V4
-	if (nfss->rpc_ops->version == 4) {
+	if (nfss->nfs_client->cl_nfsversion == 4) {
 		seq_printf(m, "\n\tnfsv4:\t");
 		seq_printf(m, "bm0=0x%x", nfss->attr_bitmask[0]);
 		seq_printf(m, ",bm1=0x%x", nfss->attr_bitmask[1]);
@@ -503,7 +504,7 @@ nfs_get_root(struct super_block *sb, struct nfs_fh *rootfh, struct nfs_fsinfo *f
 	struct nfs_server	*server = NFS_SB(sb);
 	int			error;
 
-	error = server->rpc_ops->getroot(server, rootfh, fsinfo);
+	error = server->nfs_client->rpc_ops->getroot(server, rootfh, fsinfo);
 	if (error < 0) {
 		dprintk("nfs_get_root: getattr error = %d\n", -error);
 		return ERR_PTR(error);
@@ -553,14 +554,14 @@ nfs_sb_init(struct super_block *sb, rpc_authflavor_t authflavor)
 		no_root_error = -ENOMEM;
 		goto out_no_root;
 	}
-	sb->s_root->d_op = server->rpc_ops->dentry_ops;
+	sb->s_root->d_op = server->nfs_client->rpc_ops->dentry_ops;
 
 	/* mount time stamp, in seconds */
 	server->mount_time = jiffies;
 
 	/* Get some general file system info */
 	if (server->namelen == 0 &&
-	    server->rpc_ops->pathconf(server, &server->fh, &pathinfo) >= 0)
+	    server->nfs_client->rpc_ops->pathconf(server, &server->fh, &pathinfo) >= 0)
 		server->namelen = pathinfo.max_namelen;
 	/* Work out a lot of parameters */
 	if (server->rsize == 0)
@@ -663,9 +664,14 @@ nfs_create_client(struct nfs_server *server, const struct nfs_mount_data *data)
 	struct rpc_xprt		*xprt = NULL;
 	struct rpc_clnt		*clnt = NULL;
 	int			proto = (data->flags & NFS_MOUNT_TCP) ? IPPROTO_TCP : IPPROTO_UDP;
+	int			nfsversion = 2;
 
-	clp = nfs_get_client(server->hostname, &server->addr,
-			     server->rpc_ops->version);
+#ifdef CONFIG_NFS_V3
+	if (server->flags & NFS_MOUNT_VER3)
+		nfsversion = 3;
+#endif
+
+	clp = nfs_get_client(server->hostname, &server->addr, nfsversion);
 	if (!clp) {
 		dprintk("%s: failed to create NFS4 client.\n", __FUNCTION__);
 		return ERR_PTR(PTR_ERR(clp));
@@ -676,6 +682,19 @@ nfs_create_client(struct nfs_server *server, const struct nfs_mount_data *data)
 	server->retrans_timeo = timeparms.to_initval;
 	server->retrans_count = timeparms.to_retries;
 
+	/* Check NFS protocol revision and initialize RPC op vector
+	 * and file handle pool. */
+#ifdef CONFIG_NFS_V3
+	if (nfsversion == 3) {
+		clp->rpc_ops = &nfs_v3_clientops;
+		server->caps |= NFS_CAP_READDIRPLUS;
+	} else {
+		clp->rpc_ops = &nfs_v2_clientops;
+	}
+#else
+	clp->rpc_ops = &nfs_v2_clientops;
+#endif
+
 	/* create transport and client */
 	xprt = xprt_create_proto(proto, &server->addr, &timeparms);
 	if (IS_ERR(xprt)) {
@@ -686,7 +705,7 @@ nfs_create_client(struct nfs_server *server, const struct nfs_mount_data *data)
 		return (struct rpc_clnt *)xprt;
 	}
 	clnt = rpc_create_client(xprt, server->hostname, &nfs_program,
-				 server->rpc_ops->version, data->pseudoflavor);
+				 clp->cl_nfsversion, data->pseudoflavor);
 	if (IS_ERR(clnt)) {
 		dprintk("%s: cannot create RPC client. Error = %ld\n",
 				__FUNCTION__, PTR_ERR(xprt));
@@ -750,7 +769,7 @@ static struct nfs_server *nfs_clone_server(struct super_block *sb, struct nfs_cl
 	fsinfo.fattr = data->fattr;
 	if (NFS_PROTO(root_inode)->fsinfo(server, data->fh, &fsinfo) == 0)
 		nfs_super_set_maxbytes(sb, fsinfo.maxfilesize);
-	sb->s_root->d_op = server->rpc_ops->dentry_ops;
+	sb->s_root->d_op = server->nfs_client->rpc_ops->dentry_ops;
 	sb->s_flags |= MS_ACTIVE;
 	return server;
 out_put_root:
@@ -865,19 +884,6 @@ nfs_fill_super(struct super_block *sb, struct nfs_mount_data *data, int silent)
 		return -ENOMEM;
 	strcpy(server->hostname, data->hostname);
 
-	/* Check NFS protocol revision and initialize RPC op vector
-	 * and file handle pool. */
-#ifdef CONFIG_NFS_V3
-	if (server->flags & NFS_MOUNT_VER3) {
-		server->rpc_ops = &nfs_v3_clientops;
-		server->caps |= NFS_CAP_READDIRPLUS;
-	} else {
-		server->rpc_ops = &nfs_v2_clientops;
-	}
-#else
-	server->rpc_ops = &nfs_v2_clientops;
-#endif
-
 	/* Fill in pseudoflavor for mount version < 5 */
 	if (!(data->flags & NFS_MOUNT_SECFLAVOUR))
 		data->pseudoflavor = RPC_AUTH_UNIX;
@@ -888,6 +894,7 @@ nfs_fill_super(struct super_block *sb, struct nfs_mount_data *data, int silent)
 	server->client = nfs_create_client(server, data);
 	if (IS_ERR(server->client))
 		return PTR_ERR(server->client);
+
 	/* RFC 2623, sec 2.3.2 */
 	if (authflavor != RPC_AUTH_UNIX) {
 		struct rpc_auth *auth;
@@ -1129,6 +1136,8 @@ static struct rpc_clnt *nfs4_create_client(struct nfs_server *server,
 
 	/* Now create transport and client */
 	if (clp->cl_cons_state == NFS_CS_INITING) {
+		clp->rpc_ops = &nfs_v4_clientops;
+
 		xprt = xprt_create_proto(proto, &server->addr, timeparms);
 		if (IS_ERR(xprt)) {
 			err = PTR_ERR(xprt);
@@ -1139,7 +1148,7 @@ static struct rpc_clnt *nfs4_create_client(struct nfs_server *server,
 		/* Bind to a reserved port! */
 		xprt->resvport = 1;
 		clnt = rpc_create_client(xprt, server->hostname, &nfs_program,
-				server->rpc_ops->version, flavor);
+				clp->cl_nfsversion, flavor);
 		if (IS_ERR(clnt)) {
 			err = PTR_ERR(clnt);
 			dprintk("%s: cannot create RPC client. Error = %d\n",
@@ -1215,8 +1224,6 @@ static int nfs4_fill_super(struct super_block *sb, struct nfs4_mount_data *data,
 	server->acdirmin = data->acdirmin*HZ;
 	server->acdirmax = data->acdirmax*HZ;
 
-	server->rpc_ops = &nfs_v4_clientops;
-
 	nfs_init_timeout_values(&timeparms, data->proto, data->timeo, data->retrans);
 
 	server->retrans_timeo = timeparms.to_initval;
diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h
index 70e1dc9162e2..51e9bd90dedc 100644
--- a/include/linux/nfs_fs.h
+++ b/include/linux/nfs_fs.h
@@ -215,7 +215,7 @@ static inline struct nfs_inode *NFS_I(struct inode *inode)
 #define NFS_FH(inode)			(&NFS_I(inode)->fh)
 #define NFS_SERVER(inode)		(NFS_SB(inode->i_sb))
 #define NFS_CLIENT(inode)		(NFS_SERVER(inode)->client)
-#define NFS_PROTO(inode)		(NFS_SERVER(inode)->rpc_ops)
+#define NFS_PROTO(inode)		(NFS_SERVER(inode)->nfs_client->rpc_ops)
 #define NFS_ADDR(inode)			(RPC_PEERADDR(NFS_CLIENT(inode)))
 #define NFS_COOKIEVERF(inode)		(NFS_I(inode)->cookieverf)
 #define NFS_READTIME(inode)		(NFS_I(inode)->read_cache_jiffies)
diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index e7d7662f51fd..aae7c117597a 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -25,6 +25,7 @@ struct nfs_client {
 	struct list_head	cl_superblocks;	/* List of nfs_server structs */
 
 	struct rpc_clnt *	cl_rpcclient;
+	const struct nfs_rpc_ops *rpc_ops;	/* NFS protocol vector */
 
 #ifdef CONFIG_NFS_V4
 	u64			cl_clientid;	/* constant */
@@ -74,7 +75,6 @@ struct nfs_server {
 	struct rpc_clnt *	client;		/* RPC client handle */
 	struct rpc_clnt *	client_sys;	/* 2nd handle for FSINFO */
 	struct rpc_clnt *	client_acl;	/* ACL RPC client handle */
-	const struct nfs_rpc_ops *rpc_ops;	/* NFS protocol vector */
 	struct nfs_iostats *	io_stats;	/* I/O statistics */
 	struct backing_dev_info	backing_dev_info;
 	int			flags;		/* various flags */
-- 
cgit v1.2.3


From 5006a76cca8f86c6975c16fcf67e83b8b0eee2b6 Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Tue, 22 Aug 2006 20:06:12 -0400
Subject: NFS: Eliminate client_sys in favour of cl_rpcclient

Eliminate nfs_server::client_sys in favour of nfs_client::cl_rpcclient as we
only really need one per server that we're talking to since it doesn't have any
security on it.

The retransmission management variables are also moved to the common struct as
they're required to set up the cl_rpcclient connection.

The NFS2/3 client and client_acl connections are thenceforth derived by cloning
the cl_rpcclient connection and post-applying the authorisation flavour.

The code for setting up the initial common connection has been moved to
client.c as nfs_create_rpc_client().  All the NFS program definition tables are
also moved there as that's where they're now required rather than super.c.

Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/client.c           | 119 +++++++++++++++++++++++++
 fs/nfs/internal.h         |   2 +
 fs/nfs/nfs3proc.c         |   6 +-
 fs/nfs/proc.c             |   4 +-
 fs/nfs/super.c            | 222 +++++++++++-----------------------------------
 include/linux/nfs_fs_sb.h |   5 +-
 6 files changed, 179 insertions(+), 179 deletions(-)

(limited to 'include')

diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index cb5e92463bdb..c08cab935ad5 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -50,6 +50,48 @@ static DEFINE_SPINLOCK(nfs_client_lock);
 static LIST_HEAD(nfs_client_list);
 static DECLARE_WAIT_QUEUE_HEAD(nfs_client_active_wq);
 
+/*
+ * RPC cruft for NFS
+ */
+static struct rpc_version *nfs_version[5] = {
+	[2]			= &nfs_version2,
+#ifdef CONFIG_NFS_V3
+	[3]			= &nfs_version3,
+#endif
+#ifdef CONFIG_NFS_V4
+	[4]			= &nfs_version4,
+#endif
+};
+
+struct rpc_program nfs_program = {
+	.name			= "nfs",
+	.number			= NFS_PROGRAM,
+	.nrvers			= ARRAY_SIZE(nfs_version),
+	.version		= nfs_version,
+	.stats			= &nfs_rpcstat,
+	.pipe_dir_name		= "/nfs",
+};
+
+struct rpc_stat nfs_rpcstat = {
+	.program		= &nfs_program
+};
+
+
+#ifdef CONFIG_NFS_V3_ACL
+static struct rpc_stat		nfsacl_rpcstat = { &nfsacl_program };
+static struct rpc_version *	nfsacl_version[] = {
+	[3]			= &nfsacl_version3,
+};
+
+struct rpc_program		nfsacl_program = {
+	.name			= "nfsacl",
+	.number			= NFS_ACL_PROGRAM,
+	.nrvers			= ARRAY_SIZE(nfsacl_version),
+	.version		= nfsacl_version,
+	.stats			= &nfsacl_rpcstat,
+};
+#endif  /* CONFIG_NFS_V3_ACL */
+
 /*
  * Allocate a shared client record
  *
@@ -310,3 +352,80 @@ void nfs_mark_client_ready(struct nfs_client *clp, int state)
 	clp->cl_cons_state = state;
 	wake_up_all(&nfs_client_active_wq);
 }
+
+/*
+ * Initialise the timeout values for a connection
+ */
+static void nfs_init_timeout_values(struct rpc_timeout *to, int proto,
+				    unsigned int timeo, unsigned int retrans)
+{
+	to->to_initval = timeo * HZ / 10;
+	to->to_retries = retrans;
+	if (!to->to_retries)
+		to->to_retries = 2;
+
+	switch (proto) {
+	case IPPROTO_TCP:
+		if (!to->to_initval)
+			to->to_initval = 60 * HZ;
+		if (to->to_initval > NFS_MAX_TCP_TIMEOUT)
+			to->to_initval = NFS_MAX_TCP_TIMEOUT;
+		to->to_increment = to->to_initval;
+		to->to_maxval = to->to_initval + (to->to_increment * to->to_retries);
+		to->to_exponential = 0;
+		break;
+	case IPPROTO_UDP:
+	default:
+		if (!to->to_initval)
+			to->to_initval = 11 * HZ / 10;
+		if (to->to_initval > NFS_MAX_UDP_TIMEOUT)
+			to->to_initval = NFS_MAX_UDP_TIMEOUT;
+		to->to_maxval = NFS_MAX_UDP_TIMEOUT;
+		to->to_exponential = 1;
+		break;
+	}
+}
+
+/*
+ * Create an RPC client handle
+ */
+int nfs_create_rpc_client(struct nfs_client *clp, int proto,
+			  unsigned int timeo,
+			  unsigned int retrans,
+			  rpc_authflavor_t flavor)
+{
+	struct rpc_timeout	timeparms;
+	struct rpc_xprt		*xprt = NULL;
+	struct rpc_clnt		*clnt = NULL;
+
+	if (!IS_ERR(clp->cl_rpcclient))
+		return 0;
+
+	nfs_init_timeout_values(&timeparms, proto, timeo, retrans);
+	clp->retrans_timeo = timeparms.to_initval;
+	clp->retrans_count = timeparms.to_retries;
+
+	/* create transport and client */
+	xprt = xprt_create_proto(proto, &clp->cl_addr, &timeparms);
+	if (IS_ERR(xprt)) {
+		dprintk("%s: cannot create RPC transport. Error = %ld\n",
+				__FUNCTION__, PTR_ERR(xprt));
+		return PTR_ERR(xprt);
+	}
+
+	/* Bind to a reserved port! */
+	xprt->resvport = 1;
+	/* Create the client RPC handle */
+	clnt = rpc_create_client(xprt, clp->cl_hostname, &nfs_program,
+				 clp->rpc_ops->version, RPC_AUTH_UNIX);
+	if (IS_ERR(clnt)) {
+		dprintk("%s: cannot create RPC client. Error = %ld\n",
+				__FUNCTION__, PTR_ERR(clnt));
+		return PTR_ERR(clnt);
+	}
+
+	clnt->cl_intr     = 1;
+	clnt->cl_softrtry = 1;
+	clp->cl_rpcclient = clnt;
+	return 0;
+}
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index ac370d5d4494..2f3aa52fbefc 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -20,6 +20,8 @@ extern void nfs_put_client(struct nfs_client *);
 extern struct nfs_client *nfs_find_client(const struct sockaddr_in *, int);
 extern struct nfs_client *nfs_get_client(const char *, const struct sockaddr_in *, int);
 extern void nfs_mark_client_ready(struct nfs_client *, int);
+extern int nfs_create_rpc_client(struct nfs_client *, int, unsigned int,
+				 unsigned int, rpc_authflavor_t);
 
 /* nfs4namespace.c */
 #ifdef CONFIG_NFS_V4
diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c
index 3e5371241cea..0622af0122be 100644
--- a/fs/nfs/nfs3proc.c
+++ b/fs/nfs/nfs3proc.c
@@ -90,8 +90,8 @@ nfs3_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
 	int	status;
 
 	status = do_proc_get_root(server->client, fhandle, info);
-	if (status && server->client_sys != server->client)
-		status = do_proc_get_root(server->client_sys, fhandle, info);
+	if (status && server->nfs_client->cl_rpcclient != server->client)
+		status = do_proc_get_root(server->nfs_client->cl_rpcclient, fhandle, info);
 	return status;
 }
 
@@ -785,7 +785,7 @@ nfs3_proc_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle,
 
 	dprintk("NFS call  fsinfo\n");
 	nfs_fattr_init(info->fattr);
-	status = rpc_call_sync(server->client_sys, &msg, 0);
+	status = rpc_call_sync(server->nfs_client->cl_rpcclient, &msg, 0);
 	dprintk("NFS reply fsinfo: %d\n", status);
 	return status;
 }
diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c
index 77676903e0f5..5a8b9407ee9a 100644
--- a/fs/nfs/proc.c
+++ b/fs/nfs/proc.c
@@ -66,14 +66,14 @@ nfs_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
 
 	dprintk("%s: call getattr\n", __FUNCTION__);
 	nfs_fattr_init(fattr);
-	status = rpc_call_sync(server->client_sys, &msg, 0);
+	status = rpc_call_sync(server->nfs_client->cl_rpcclient, &msg, 0);
 	dprintk("%s: reply getattr: %d\n", __FUNCTION__, status);
 	if (status)
 		return status;
 	dprintk("%s: call statfs\n", __FUNCTION__);
 	msg.rpc_proc = &nfs_procedures[NFSPROC_STATFS];
 	msg.rpc_resp = &fsinfo;
-	status = rpc_call_sync(server->client_sys, &msg, 0);
+	status = rpc_call_sync(server->nfs_client->cl_rpcclient, &msg, 0);
 	dprintk("%s: reply statfs: %d\n", __FUNCTION__, status);
 	if (status)
 		return status;
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index e1e5eab0259b..85583414a3ca 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -60,52 +60,6 @@
  */
 #define NFS_MAX_READAHEAD	(RPC_DEF_SLOT_TABLE - 1)
 
-/*
- * RPC cruft for NFS
- */
-static struct rpc_version * nfs_version[] = {
-	NULL,
-	NULL,
-	&nfs_version2,
-#if defined(CONFIG_NFS_V3)
-	&nfs_version3,
-#elif defined(CONFIG_NFS_V4)
-	NULL,
-#endif
-#if defined(CONFIG_NFS_V4)
-	&nfs_version4,
-#endif
-};
-
-static struct rpc_program nfs_program = {
-	.name			= "nfs",
-	.number			= NFS_PROGRAM,
-	.nrvers			= ARRAY_SIZE(nfs_version),
-	.version		= nfs_version,
-	.stats			= &nfs_rpcstat,
-	.pipe_dir_name		= "/nfs",
-};
-
-struct rpc_stat nfs_rpcstat = {
-	.program		= &nfs_program
-};
-
-
-#ifdef CONFIG_NFS_V3_ACL
-static struct rpc_stat		nfsacl_rpcstat = { &nfsacl_program };
-static struct rpc_version *	nfsacl_version[] = {
-	[3]			= &nfsacl_version3,
-};
-
-struct rpc_program		nfsacl_program = {
-	.name =			"nfsacl",
-	.number =		NFS_ACL_PROGRAM,
-	.nrvers =		ARRAY_SIZE(nfsacl_version),
-	.version =		nfsacl_version,
-	.stats =		&nfsacl_rpcstat,
-};
-#endif  /* CONFIG_NFS_V3_ACL */
-
 static void nfs_umount_begin(struct vfsmount *, int);
 static int  nfs_statfs(struct dentry *, struct kstatfs *);
 static int  nfs_show_options(struct seq_file *, struct vfsmount *);
@@ -376,8 +330,8 @@ static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss,
 			proto = buf;
 	}
 	seq_printf(m, ",proto=%s", proto);
-	seq_printf(m, ",timeo=%lu", 10U * nfss->retrans_timeo / HZ);
-	seq_printf(m, ",retrans=%u", nfss->retrans_count);
+	seq_printf(m, ",timeo=%lu", 10U * clp->retrans_timeo / HZ);
+	seq_printf(m, ",retrans=%u", clp->retrans_count);
 	seq_printf(m, ",sec=%s", nfs_pseudoflavour_to_name(nfss->client->cl_auth->au_flavor));
 }
 
@@ -621,38 +575,6 @@ out_no_root:
 	return no_root_error;
 }
 
-/*
- * Initialise the timeout values for a connection
- */
-static void nfs_init_timeout_values(struct rpc_timeout *to, int proto, unsigned int timeo, unsigned int retrans)
-{
-	to->to_initval = timeo * HZ / 10;
-	to->to_retries = retrans;
-	if (!to->to_retries)
-		to->to_retries = 2;
-
-	switch (proto) {
-	case IPPROTO_TCP:
-		if (!to->to_initval)
-			to->to_initval = 60 * HZ;
-		if (to->to_initval > NFS_MAX_TCP_TIMEOUT)
-			to->to_initval = NFS_MAX_TCP_TIMEOUT;
-		to->to_increment = to->to_initval;
-		to->to_maxval = to->to_initval + (to->to_increment * to->to_retries);
-		to->to_exponential = 0;
-		break;
-	case IPPROTO_UDP:
-	default:
-		if (!to->to_initval)
-			to->to_initval = 11 * HZ / 10;
-		if (to->to_initval > NFS_MAX_UDP_TIMEOUT)
-			to->to_initval = NFS_MAX_UDP_TIMEOUT;
-		to->to_maxval = NFS_MAX_UDP_TIMEOUT;
-		to->to_exponential = 1;
-		break;
-	}
-}
-
 /*
  * Create an RPC client handle.
  */
@@ -660,11 +582,10 @@ static struct rpc_clnt *
 nfs_create_client(struct nfs_server *server, const struct nfs_mount_data *data)
 {
 	struct nfs_client	*clp;
-	struct rpc_timeout	timeparms;
-	struct rpc_xprt		*xprt = NULL;
-	struct rpc_clnt		*clnt = NULL;
+	struct rpc_clnt		*clnt;
 	int			proto = (data->flags & NFS_MOUNT_TCP) ? IPPROTO_TCP : IPPROTO_UDP;
 	int			nfsversion = 2;
+	int			err;
 
 #ifdef CONFIG_NFS_V3
 	if (server->flags & NFS_MOUNT_VER3)
@@ -677,52 +598,54 @@ nfs_create_client(struct nfs_server *server, const struct nfs_mount_data *data)
 		return ERR_PTR(PTR_ERR(clp));
 	}
 
-	nfs_init_timeout_values(&timeparms, proto, data->timeo, data->retrans);
-
-	server->retrans_timeo = timeparms.to_initval;
-	server->retrans_count = timeparms.to_retries;
-
-	/* Check NFS protocol revision and initialize RPC op vector
-	 * and file handle pool. */
+	if (clp->cl_cons_state == NFS_CS_INITING) {
+		/* Check NFS protocol revision and initialize RPC op
+		 * vector and file handle pool. */
 #ifdef CONFIG_NFS_V3
-	if (nfsversion == 3) {
-		clp->rpc_ops = &nfs_v3_clientops;
-		server->caps |= NFS_CAP_READDIRPLUS;
-	} else {
-		clp->rpc_ops = &nfs_v2_clientops;
-	}
+		if (nfsversion == 3) {
+			clp->rpc_ops = &nfs_v3_clientops;
+			server->caps |= NFS_CAP_READDIRPLUS;
+		} else {
+			clp->rpc_ops = &nfs_v2_clientops;
+		}
 #else
-	clp->rpc_ops = &nfs_v2_clientops;
+		clp->rpc_ops = &nfs_v2_clientops;
 #endif
 
-	/* create transport and client */
-	xprt = xprt_create_proto(proto, &server->addr, &timeparms);
-	if (IS_ERR(xprt)) {
-		dprintk("%s: cannot create RPC transport. Error = %ld\n",
-				__FUNCTION__, PTR_ERR(xprt));
-		nfs_mark_client_ready(clp, PTR_ERR(xprt));
-		nfs_put_client(clp);
-		return (struct rpc_clnt *)xprt;
+		/* create transport and client */
+		err = nfs_create_rpc_client(clp, proto, data->timeo,
+					    data->retrans, RPC_AUTH_UNIX);
+		if (err < 0)
+			goto client_init_error;
+
+		nfs_mark_client_ready(clp, 0);
 	}
-	clnt = rpc_create_client(xprt, server->hostname, &nfs_program,
-				 clp->cl_nfsversion, data->pseudoflavor);
+
+	/* create an nfs_server-specific client */
+	clnt = rpc_clone_client(clp->cl_rpcclient);
 	if (IS_ERR(clnt)) {
-		dprintk("%s: cannot create RPC client. Error = %ld\n",
-				__FUNCTION__, PTR_ERR(xprt));
-		goto out_fail;
+		dprintk("%s: couldn't create rpc_client!\n", __FUNCTION__);
+		nfs_put_client(clp);
+		return ERR_PTR(PTR_ERR(clnt));
 	}
 
-	clnt->cl_intr     = 1;
-	clnt->cl_softrtry = 1;
+	if (data->pseudoflavor != clp->cl_rpcclient->cl_auth->au_flavor) {
+		struct rpc_auth *auth;
+
+		auth = rpcauth_create(data->pseudoflavor, server->client);
+		if (IS_ERR(auth)) {
+			dprintk("%s: couldn't create credcache!\n", __FUNCTION__);
+			return ERR_PTR(PTR_ERR(auth));
+		}
+	}
 
-	nfs_mark_client_ready(clp, 0);
 	server->nfs_client = clp;
 	return clnt;
 
-out_fail:
-	nfs_mark_client_ready(clp, PTR_ERR(xprt));
+client_init_error:
+	nfs_mark_client_ready(clp, err);
 	nfs_put_client(clp);
-	return clnt;
+	return ERR_PTR(err);
 }
 
 /*
@@ -741,7 +664,7 @@ static struct nfs_server *nfs_clone_server(struct super_block *sb, struct nfs_cl
 	sb->s_blocksize_bits = data->sb->s_blocksize_bits;
 	sb->s_maxbytes = data->sb->s_maxbytes;
 
-	server->client_sys = server->client_acl = ERR_PTR(-EINVAL);
+	server->client_acl = ERR_PTR(-EINVAL);
 	server->io_stats = nfs_alloc_iostats();
 	if (server->io_stats == NULL)
 		goto out;
@@ -750,11 +673,6 @@ static struct nfs_server *nfs_clone_server(struct super_block *sb, struct nfs_cl
 	if (IS_ERR((err = server->client)))
 		goto out;
 
-	if (!IS_ERR(parent->client_sys)) {
-		server->client_sys = rpc_clone_client(parent->client_sys);
-		if (IS_ERR((err = server->client_sys)))
-			goto out;
-	}
 	if (!IS_ERR(parent->client_acl)) {
 		server->client_acl = rpc_clone_client(parent->client_acl);
 		if (IS_ERR((err = server->client_acl)))
@@ -813,7 +731,7 @@ static int nfs_clone_generic_sb(struct nfs_clone_mount *data,
 		error = PTR_ERR(sb);
 		goto kill_rpciod;
 	}
-		
+
 	if (sb->s_root)
 		goto out_rpciod_down;
 
@@ -896,19 +814,6 @@ nfs_fill_super(struct super_block *sb, struct nfs_mount_data *data, int silent)
 		return PTR_ERR(server->client);
 
 	/* RFC 2623, sec 2.3.2 */
-	if (authflavor != RPC_AUTH_UNIX) {
-		struct rpc_auth *auth;
-
-		server->client_sys = rpc_clone_client(server->client);
-		if (IS_ERR(server->client_sys))
-			return PTR_ERR(server->client_sys);
-		auth = rpcauth_create(RPC_AUTH_UNIX, server->client_sys);
-		if (IS_ERR(auth))
-			return PTR_ERR(auth);
-	} else {
-		atomic_inc(&server->client->cl_count);
-		server->client_sys = server->client;
-	}
 	if (server->flags & NFS_MOUNT_VER3) {
 #ifdef CONFIG_NFS_V3_ACL
 		if (!(server->flags & NFS_MOUNT_NOACL)) {
@@ -1012,7 +917,7 @@ static int nfs_get_sb(struct file_system_type *fs_type,
 		goto out_err_noserver;
 	/* Zero out the NFS state stuff */
 	init_nfsv4_state(server);
-	server->client = server->client_sys = server->client_acl = ERR_PTR(-EINVAL);
+	server->client = server->client_acl = ERR_PTR(-EINVAL);
 
 	root = &server->fh;
 	if (data->flags & NFS_MOUNT_VER3)
@@ -1083,8 +988,6 @@ static void nfs_kill_super(struct super_block *s)
 
 	if (!IS_ERR(server->client))
 		rpc_shutdown_client(server->client);
-	if (!IS_ERR(server->client_sys))
-		rpc_shutdown_client(server->client_sys);
 	if (!IS_ERR(server->client_acl))
 		rpc_shutdown_client(server->client_acl);
 
@@ -1121,10 +1024,9 @@ static int nfs_clone_nfs_sb(struct file_system_type *fs_type,
 
 #ifdef CONFIG_NFS_V4
 static struct rpc_clnt *nfs4_create_client(struct nfs_server *server,
-	struct rpc_timeout *timeparms, int proto, rpc_authflavor_t flavor)
+	int timeo, int retrans, int proto, rpc_authflavor_t flavor)
 {
 	struct nfs_client *clp;
-	struct rpc_xprt *xprt = NULL;
 	struct rpc_clnt *clnt = NULL;
 	int err = -EIO;
 
@@ -1138,26 +1040,10 @@ static struct rpc_clnt *nfs4_create_client(struct nfs_server *server,
 	if (clp->cl_cons_state == NFS_CS_INITING) {
 		clp->rpc_ops = &nfs_v4_clientops;
 
-		xprt = xprt_create_proto(proto, &server->addr, timeparms);
-		if (IS_ERR(xprt)) {
-			err = PTR_ERR(xprt);
-			dprintk("%s: cannot create RPC transport. Error = %d\n",
-					__FUNCTION__, err);
+		err = nfs_create_rpc_client(clp, proto, timeo, retrans, flavor);
+		if (err < 0)
 			goto client_init_error;
-		}
-		/* Bind to a reserved port! */
-		xprt->resvport = 1;
-		clnt = rpc_create_client(xprt, server->hostname, &nfs_program,
-				clp->cl_nfsversion, flavor);
-		if (IS_ERR(clnt)) {
-			err = PTR_ERR(clnt);
-			dprintk("%s: cannot create RPC client. Error = %d\n",
-					__FUNCTION__, err);
-			goto client_init_error;
-		}
-		clnt->cl_intr     = 1;
-		clnt->cl_softrtry = 1;
-		clp->cl_rpcclient = clnt;
+
 		memcpy(clp->cl_ipaddr, server->ip_addr, sizeof(clp->cl_ipaddr));
 		err = nfs_idmap_new(clp);
 		if (err < 0) {
@@ -1205,7 +1091,6 @@ client_init_error:
 static int nfs4_fill_super(struct super_block *sb, struct nfs4_mount_data *data, int silent)
 {
 	struct nfs_server *server;
-	struct rpc_timeout timeparms;
 	rpc_authflavor_t authflavour;
 	int err = -EIO;
 
@@ -1224,11 +1109,6 @@ static int nfs4_fill_super(struct super_block *sb, struct nfs4_mount_data *data,
 	server->acdirmin = data->acdirmin*HZ;
 	server->acdirmax = data->acdirmax*HZ;
 
-	nfs_init_timeout_values(&timeparms, data->proto, data->timeo, data->retrans);
-
-	server->retrans_timeo = timeparms.to_initval;
-	server->retrans_count = timeparms.to_retries;
-
 	/* Now create transport and client */
 	authflavour = RPC_AUTH_UNIX;
 	if (data->auth_flavourlen != 0) {
@@ -1244,7 +1124,8 @@ static int nfs4_fill_super(struct super_block *sb, struct nfs4_mount_data *data,
 		}
 	}
 
-	server->client = nfs4_create_client(server, &timeparms, data->proto, authflavour);
+	server->client = nfs4_create_client(server, data->timeo, data->retrans,
+					    data->proto, authflavour);
 	if (IS_ERR(server->client)) {
 		err = PTR_ERR(server->client);
 			dprintk("%s: cannot create RPC client. Error = %d\n",
@@ -1318,7 +1199,7 @@ static int nfs4_get_sb(struct file_system_type *fs_type,
 		return -ENOMEM;
 	/* Zero out the NFS state stuff */
 	init_nfsv4_state(server);
-	server->client = server->client_sys = server->client_acl = ERR_PTR(-EINVAL);
+	server->client = server->client_acl = ERR_PTR(-EINVAL);
 
 	p = nfs_copy_user_string(NULL, &data->hostname, 256);
 	if (IS_ERR(p))
@@ -1489,7 +1370,6 @@ err:
 static struct nfs_server *nfs4_referral_server(struct super_block *sb, struct nfs_clone_mount *data)
 {
 	struct nfs_server *server = NFS_SB(sb);
-	struct rpc_timeout timeparms;
 	int proto, timeo, retrans;
 	void *err;
 
@@ -1498,11 +1378,11 @@ static struct nfs_server *nfs4_referral_server(struct super_block *sb, struct nf
 	   set the timeouts and retries to low values */
 	timeo = 2;
 	retrans = 1;
-	nfs_init_timeout_values(&timeparms, proto, timeo, retrans);
 
 	nfs_put_client(server->nfs_client);
 	server->nfs_client = NULL;
-	server->client = nfs4_create_client(server, &timeparms, proto, data->authflavor);
+	server->client = nfs4_create_client(server, timeo, retrans, proto,
+					    data->authflavor);
 	if (IS_ERR((err = server->client)))
 		goto out_err;
 
diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index aae7c117597a..d404ceca9168 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -26,6 +26,8 @@ struct nfs_client {
 
 	struct rpc_clnt *	cl_rpcclient;
 	const struct nfs_rpc_ops *rpc_ops;	/* NFS protocol vector */
+	unsigned long		retrans_timeo;	/* retransmit timeout */
+	unsigned int		retrans_count;	/* number of retransmit tries */
 
 #ifdef CONFIG_NFS_V4
 	u64			cl_clientid;	/* constant */
@@ -73,7 +75,6 @@ struct nfs_client {
 struct nfs_server {
 	struct nfs_client *	nfs_client;	/* shared client and NFS4 state */
 	struct rpc_clnt *	client;		/* RPC client handle */
-	struct rpc_clnt *	client_sys;	/* 2nd handle for FSINFO */
 	struct rpc_clnt *	client_acl;	/* ACL RPC client handle */
 	struct nfs_iostats *	io_stats;	/* I/O statistics */
 	struct backing_dev_info	backing_dev_info;
@@ -90,8 +91,6 @@ struct nfs_server {
 	unsigned int		acregmax;
 	unsigned int		acdirmin;
 	unsigned int		acdirmax;
-	unsigned long		retrans_timeo;	/* retransmit timeout */
-	unsigned int		retrans_count;	/* number of retransmit tries */
 	unsigned int		namelen;
 	char *			hostname;	/* remote hostname */
 	struct nfs_fh		fh;
-- 
cgit v1.2.3


From 54ceac4515986030c2502960be620198dd8fe25b Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Tue, 22 Aug 2006 20:06:13 -0400
Subject: NFS: Share NFS superblocks per-protocol per-server per-FSID

The attached patch makes NFS share superblocks between mounts from the same
server and FSID over the same protocol.

It does this by creating each superblock with a false root and returning the
real root dentry in the vfsmount presented by get_sb(). The root dentry set
starts off as an anonymous dentry if we don't already have the dentry for its
inode, otherwise it simply returns the dentry we already have.

We may thus end up with several trees of dentries in the superblock, and if at
some later point one of anonymous tree roots is discovered by normal filesystem
activity to be located in another tree within the superblock, the anonymous
root is named and materialises attached to the second tree at the appropriate
point.

Why do it this way? Why not pass an extra argument to the mount() syscall to
indicate the subpath and then pathwalk from the server root to the desired
directory? You can't guarantee this will work for two reasons:

 (1) The root and intervening nodes may not be accessible to the client.

     With NFS2 and NFS3, for instance, mountd is called on the server to get
     the filehandle for the tip of a path. mountd won't give us handles for
     anything we don't have permission to access, and so we can't set up NFS
     inodes for such nodes, and so can't easily set up dentries (we'd have to
     have ghost inodes or something).

     With this patch we don't actually create dentries until we get handles
     from the server that we can use to set up their inodes, and we don't
     actually bind them into the tree until we know for sure where they go.

 (2) Inaccessible symbolic links.

     If we're asked to mount two exports from the server, eg:

	mount warthog:/warthog/aaa/xxx /mmm
	mount warthog:/warthog/bbb/yyy /nnn

     We may not be able to access anything nearer the root than xxx and yyy,
     but we may find out later that /mmm/www/yyy, say, is actually the same
     directory as the one mounted on /nnn. What we might then find out, for
     example, is that /warthog/bbb was actually a symbolic link to
     /warthog/aaa/xxx/www, but we can't actually determine that by talking to
     the server until /warthog is made available by NFS.

     This would lead to having constructed an errneous dentry tree which we
     can't easily fix. We can end up with a dentry marked as a directory when
     it should actually be a symlink, or we could end up with an apparently
     hardlinked directory.

     With this patch we need not make assumptions about the type of a dentry
     for which we can't retrieve information, nor need we assume we know its
     place in the grand scheme of things until we actually see that place.

This patch reduces the possibility of aliasing in the inode and page caches for
inodes that may be accessed by more than one NFS export. It also reduces the
number of superblocks required for NFS where there are many NFS exports being
used from a server (home directory server + autofs for example).

This in turn makes it simpler to do local caching of network filesystems, as it
can then be guaranteed that there won't be links from multiple inodes in
separate superblocks to the same cache file.

Obviously, cache aliasing between different levels of NFS protocol could still
be a problem, but at least that gives us another key to use when indexing the
cache.

This patch makes the following changes:

 (1) The server record construction/destruction has been abstracted out into
     its own set of functions to make things easier to get right.  These have
     been moved into fs/nfs/client.c.

     All the code in fs/nfs/client.c has to do with the management of
     connections to servers, and doesn't touch superblocks in any way; the
     remaining code in fs/nfs/super.c has to do with VFS superblock management.

 (2) The sequence of events undertaken by NFS mount is now reordered:

     (a) A volume representation (struct nfs_server) is allocated.

     (b) A server representation (struct nfs_client) is acquired.  This may be
     	 allocated or shared, and is keyed on server address, port and NFS
     	 version.

     (c) If allocated, the client representation is initialised.  The state
     	 member variable of nfs_client is used to prevent a race during
     	 initialisation from two mounts.

     (d) For NFS4 a simple pathwalk is performed, walking from FH to FH to find
     	 the root filehandle for the mount (fs/nfs/getroot.c).  For NFS2/3 we
     	 are given the root FH in advance.

     (e) The volume FSID is probed for on the root FH.

     (f) The volume representation is initialised from the FSINFO record
     	 retrieved on the root FH.

     (g) sget() is called to acquire a superblock.  This may be allocated or
     	 shared, keyed on client pointer and FSID.

     (h) If allocated, the superblock is initialised.

     (i) If the superblock is shared, then the new nfs_server record is
     	 discarded.

     (j) The root dentry for this mount is looked up from the root FH.

     (k) The root dentry for this mount is assigned to the vfsmount.

 (3) nfs_readdir_lookup() creates dentries for each of the entries readdir()
     returns; this function now attaches disconnected trees from alternate
     roots that happen to be discovered attached to a directory being read (in
     the same way nfs_lookup() is made to do for lookup ops).

     The new d_materialise_unique() function is now used to do this, thus
     permitting the whole thing to be done under one set of locks, and thus
     avoiding any race between mount and lookup operations on the same
     directory.

 (4) The client management code uses a new debug facility: NFSDBG_CLIENT which
     is set by echoing 1024 to /proc/net/sunrpc/nfs_debug.

 (5) Clone mounts are now called xdev mounts.

 (6) Use the dentry passed to the statfs() op as the handle for retrieving fs
     statistics rather than the root dentry of the superblock (which is now a
     dummy).

Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/Makefile           |    2 +-
 fs/nfs/client.c           |  735 ++++++++++++++++++++++++++-
 fs/nfs/dir.c              |   16 +-
 fs/nfs/getroot.c          |  306 ++++++++++++
 fs/nfs/idmap.c            |    3 +-
 fs/nfs/inode.c            |    2 +-
 fs/nfs/internal.h         |   82 +--
 fs/nfs/namespace.c        |   25 +-
 fs/nfs/nfs3proc.c         |    2 +-
 fs/nfs/nfs4_fs.h          |    6 -
 fs/nfs/nfs4namespace.c    |  110 ++++-
 fs/nfs/nfs4proc.c         |   59 +--
 fs/nfs/nfs4renewd.c       |   13 -
 fs/nfs/nfs4state.c        |   18 -
 fs/nfs/read.c             |    2 +-
 fs/nfs/super.c            | 1207 +++++++++++++++++----------------------------
 fs/nfs/write.c            |    2 +-
 include/linux/nfs_fs_sb.h |   21 +-
 18 files changed, 1655 insertions(+), 956 deletions(-)
 create mode 100644 fs/nfs/getroot.c

(limited to 'include')

diff --git a/fs/nfs/Makefile b/fs/nfs/Makefile
index 3b993a6f8163..f4580b44eef4 100644
--- a/fs/nfs/Makefile
+++ b/fs/nfs/Makefile
@@ -4,7 +4,7 @@
 
 obj-$(CONFIG_NFS_FS) += nfs.o
 
-nfs-y 			:= client.o dir.o file.o inode.o super.o nfs2xdr.o \
+nfs-y 			:= client.o dir.o file.o getroot.o inode.o super.o nfs2xdr.o \
 			   pagelist.o proc.o read.o symlink.o unlink.o \
 			   write.o namespace.o
 nfs-$(CONFIG_ROOT_NFS)	+= nfsroot.o mount_clnt.o      
diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index c08cab935ad5..dafba608c0a0 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -48,6 +48,7 @@
 
 static DEFINE_SPINLOCK(nfs_client_lock);
 static LIST_HEAD(nfs_client_list);
+static LIST_HEAD(nfs_volume_list);
 static DECLARE_WAIT_QUEUE_HEAD(nfs_client_active_wq);
 
 /*
@@ -268,9 +269,9 @@ struct nfs_client *nfs_find_client(const struct sockaddr_in *addr, int nfsversio
  * Look up a client by IP address and protocol version
  * - creates a new record if one doesn't yet exist
  */
-struct nfs_client *nfs_get_client(const char *hostname,
-				  const struct sockaddr_in *addr,
-				  int nfsversion)
+static struct nfs_client *nfs_get_client(const char *hostname,
+					 const struct sockaddr_in *addr,
+					 int nfsversion)
 {
 	struct nfs_client *clp, *new = NULL;
 	int error;
@@ -340,6 +341,8 @@ found_client:
 		return ERR_PTR(error);
 	}
 
+	BUG_ON(clp->cl_cons_state != NFS_CS_READY);
+
 	dprintk("--> nfs_get_client() = %p [share]\n", clp);
 	return clp;
 }
@@ -347,7 +350,7 @@ found_client:
 /*
  * Mark a server as ready or failed
  */
-void nfs_mark_client_ready(struct nfs_client *clp, int state)
+static void nfs_mark_client_ready(struct nfs_client *clp, int state)
 {
 	clp->cl_cons_state = state;
 	wake_up_all(&nfs_client_active_wq);
@@ -389,10 +392,10 @@ static void nfs_init_timeout_values(struct rpc_timeout *to, int proto,
 /*
  * Create an RPC client handle
  */
-int nfs_create_rpc_client(struct nfs_client *clp, int proto,
-			  unsigned int timeo,
-			  unsigned int retrans,
-			  rpc_authflavor_t flavor)
+static int nfs_create_rpc_client(struct nfs_client *clp, int proto,
+						unsigned int timeo,
+						unsigned int retrans,
+						rpc_authflavor_t flavor)
 {
 	struct rpc_timeout	timeparms;
 	struct rpc_xprt		*xprt = NULL;
@@ -429,3 +432,719 @@ int nfs_create_rpc_client(struct nfs_client *clp, int proto,
 	clp->cl_rpcclient = clnt;
 	return 0;
 }
+
+/*
+ * Version 2 or 3 client destruction
+ */
+static void nfs_destroy_server(struct nfs_server *server)
+{
+	if (!IS_ERR(server->client_acl))
+		rpc_shutdown_client(server->client_acl);
+
+	if (!(server->flags & NFS_MOUNT_NONLM))
+		lockd_down();	/* release rpc.lockd */
+}
+
+/*
+ * Version 2 or 3 lockd setup
+ */
+static int nfs_start_lockd(struct nfs_server *server)
+{
+	int error = 0;
+
+	if (server->nfs_client->cl_nfsversion > 3)
+		goto out;
+	if (server->flags & NFS_MOUNT_NONLM)
+		goto out;
+	error = lockd_up();
+	if (error < 0)
+		server->flags |= NFS_MOUNT_NONLM;
+	else
+		server->destroy = nfs_destroy_server;
+out:
+	return error;
+}
+
+/*
+ * Initialise an NFSv3 ACL client connection
+ */
+#ifdef CONFIG_NFS_V3_ACL
+static void nfs_init_server_aclclient(struct nfs_server *server)
+{
+	if (server->nfs_client->cl_nfsversion != 3)
+		goto out_noacl;
+	if (server->flags & NFS_MOUNT_NOACL)
+		goto out_noacl;
+
+	server->client_acl = rpc_bind_new_program(server->client, &nfsacl_program, 3);
+	if (IS_ERR(server->client_acl))
+		goto out_noacl;
+
+	/* No errors! Assume that Sun nfsacls are supported */
+	server->caps |= NFS_CAP_ACLS;
+	return;
+
+out_noacl:
+	server->caps &= ~NFS_CAP_ACLS;
+}
+#else
+static inline void nfs_init_server_aclclient(struct nfs_server *server)
+{
+	server->flags &= ~NFS_MOUNT_NOACL;
+	server->caps &= ~NFS_CAP_ACLS;
+}
+#endif
+
+/*
+ * Create a general RPC client
+ */
+static int nfs_init_server_rpcclient(struct nfs_server *server, rpc_authflavor_t pseudoflavour)
+{
+	struct nfs_client *clp = server->nfs_client;
+
+	server->client = rpc_clone_client(clp->cl_rpcclient);
+	if (IS_ERR(server->client)) {
+		dprintk("%s: couldn't create rpc_client!\n", __FUNCTION__);
+		return PTR_ERR(server->client);
+	}
+
+	if (pseudoflavour != clp->cl_rpcclient->cl_auth->au_flavor) {
+		struct rpc_auth *auth;
+
+		auth = rpcauth_create(pseudoflavour, server->client);
+		if (IS_ERR(auth)) {
+			dprintk("%s: couldn't create credcache!\n", __FUNCTION__);
+			return PTR_ERR(auth);
+		}
+	}
+	server->client->cl_softrtry = 0;
+	if (server->flags & NFS_MOUNT_SOFT)
+		server->client->cl_softrtry = 1;
+
+	server->client->cl_intr = 0;
+	if (server->flags & NFS4_MOUNT_INTR)
+		server->client->cl_intr = 1;
+
+	return 0;
+}
+
+/*
+ * Initialise an NFS2 or NFS3 client
+ */
+static int nfs_init_client(struct nfs_client *clp, const struct nfs_mount_data *data)
+{
+	int proto = (data->flags & NFS_MOUNT_TCP) ? IPPROTO_TCP : IPPROTO_UDP;
+	int error;
+
+	if (clp->cl_cons_state == NFS_CS_READY) {
+		/* the client is already initialised */
+		dprintk("<-- nfs_init_client() = 0 [already %p]\n", clp);
+		return 0;
+	}
+
+	/* Check NFS protocol revision and initialize RPC op vector */
+	clp->rpc_ops = &nfs_v2_clientops;
+#ifdef CONFIG_NFS_V3
+	if (clp->cl_nfsversion == 3)
+		clp->rpc_ops = &nfs_v3_clientops;
+#endif
+	/*
+	 * Create a client RPC handle for doing FSSTAT with UNIX auth only
+	 * - RFC 2623, sec 2.3.2
+	 */
+	error = nfs_create_rpc_client(clp, proto, data->timeo, data->retrans,
+			RPC_AUTH_UNIX);
+	if (error < 0)
+		goto error;
+	nfs_mark_client_ready(clp, NFS_CS_READY);
+	return 0;
+
+error:
+	nfs_mark_client_ready(clp, error);
+	dprintk("<-- nfs_init_client() = xerror %d\n", error);
+	return error;
+}
+
+/*
+ * Create a version 2 or 3 client
+ */
+static int nfs_init_server(struct nfs_server *server, const struct nfs_mount_data *data)
+{
+	struct nfs_client *clp;
+	int error, nfsvers = 2;
+
+	dprintk("--> nfs_init_server()\n");
+
+#ifdef CONFIG_NFS_V3
+	if (data->flags & NFS_MOUNT_VER3)
+		nfsvers = 3;
+#endif
+
+	/* Allocate or find a client reference we can use */
+	clp = nfs_get_client(data->hostname, &data->addr, nfsvers);
+	if (IS_ERR(clp)) {
+		dprintk("<-- nfs_init_server() = error %ld\n", PTR_ERR(clp));
+		return PTR_ERR(clp);
+	}
+
+	error = nfs_init_client(clp, data);
+	if (error < 0)
+		goto error;
+
+	server->nfs_client = clp;
+
+	/* Initialise the client representation from the mount data */
+	server->flags = data->flags & NFS_MOUNT_FLAGMASK;
+
+	if (data->rsize)
+		server->rsize = nfs_block_size(data->rsize, NULL);
+	if (data->wsize)
+		server->wsize = nfs_block_size(data->wsize, NULL);
+
+	server->acregmin = data->acregmin * HZ;
+	server->acregmax = data->acregmax * HZ;
+	server->acdirmin = data->acdirmin * HZ;
+	server->acdirmax = data->acdirmax * HZ;
+
+	/* Start lockd here, before we might error out */
+	error = nfs_start_lockd(server);
+	if (error < 0)
+		goto error;
+
+	error = nfs_init_server_rpcclient(server, data->pseudoflavor);
+	if (error < 0)
+		goto error;
+
+	server->namelen  = data->namlen;
+	/* Create a client RPC handle for the NFSv3 ACL management interface */
+	nfs_init_server_aclclient(server);
+	if (clp->cl_nfsversion == 3) {
+		if (server->namelen == 0 || server->namelen > NFS3_MAXNAMLEN)
+			server->namelen = NFS3_MAXNAMLEN;
+		server->caps |= NFS_CAP_READDIRPLUS;
+	} else {
+		if (server->namelen == 0 || server->namelen > NFS2_MAXNAMLEN)
+			server->namelen = NFS2_MAXNAMLEN;
+	}
+
+	dprintk("<-- nfs_init_server() = 0 [new %p]\n", clp);
+	return 0;
+
+error:
+	server->nfs_client = NULL;
+	nfs_put_client(clp);
+	dprintk("<-- nfs_init_server() = xerror %d\n", error);
+	return error;
+}
+
+/*
+ * Load up the server record from information gained in an fsinfo record
+ */
+static void nfs_server_set_fsinfo(struct nfs_server *server, struct nfs_fsinfo *fsinfo)
+{
+	unsigned long max_rpc_payload;
+
+	/* Work out a lot of parameters */
+	if (server->rsize == 0)
+		server->rsize = nfs_block_size(fsinfo->rtpref, NULL);
+	if (server->wsize == 0)
+		server->wsize = nfs_block_size(fsinfo->wtpref, NULL);
+
+	if (fsinfo->rtmax >= 512 && server->rsize > fsinfo->rtmax)
+		server->rsize = nfs_block_size(fsinfo->rtmax, NULL);
+	if (fsinfo->wtmax >= 512 && server->wsize > fsinfo->wtmax)
+		server->wsize = nfs_block_size(fsinfo->wtmax, NULL);
+
+	max_rpc_payload = nfs_block_size(rpc_max_payload(server->client), NULL);
+	if (server->rsize > max_rpc_payload)
+		server->rsize = max_rpc_payload;
+	if (server->rsize > NFS_MAX_FILE_IO_SIZE)
+		server->rsize = NFS_MAX_FILE_IO_SIZE;
+	server->rpages = (server->rsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
+	server->backing_dev_info.ra_pages = server->rpages * NFS_MAX_READAHEAD;
+
+	if (server->wsize > max_rpc_payload)
+		server->wsize = max_rpc_payload;
+	if (server->wsize > NFS_MAX_FILE_IO_SIZE)
+		server->wsize = NFS_MAX_FILE_IO_SIZE;
+	server->wpages = (server->wsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
+	server->wtmult = nfs_block_bits(fsinfo->wtmult, NULL);
+
+	server->dtsize = nfs_block_size(fsinfo->dtpref, NULL);
+	if (server->dtsize > PAGE_CACHE_SIZE)
+		server->dtsize = PAGE_CACHE_SIZE;
+	if (server->dtsize > server->rsize)
+		server->dtsize = server->rsize;
+
+	if (server->flags & NFS_MOUNT_NOAC) {
+		server->acregmin = server->acregmax = 0;
+		server->acdirmin = server->acdirmax = 0;
+	}
+
+	server->maxfilesize = fsinfo->maxfilesize;
+
+	/* We're airborne Set socket buffersize */
+	rpc_setbufsize(server->client, server->wsize + 100, server->rsize + 100);
+}
+
+/*
+ * Probe filesystem information, including the FSID on v2/v3
+ */
+static int nfs_probe_fsinfo(struct nfs_server *server, struct nfs_fh *mntfh, struct nfs_fattr *fattr)
+{
+	struct nfs_fsinfo fsinfo;
+	struct nfs_client *clp = server->nfs_client;
+	int error;
+
+	dprintk("--> nfs_probe_fsinfo()\n");
+
+	if (clp->rpc_ops->set_capabilities != NULL) {
+		error = clp->rpc_ops->set_capabilities(server, mntfh);
+		if (error < 0)
+			goto out_error;
+	}
+
+	fsinfo.fattr = fattr;
+	nfs_fattr_init(fattr);
+	error = clp->rpc_ops->fsinfo(server, mntfh, &fsinfo);
+	if (error < 0)
+		goto out_error;
+
+	nfs_server_set_fsinfo(server, &fsinfo);
+
+	/* Get some general file system info */
+	if (server->namelen == 0) {
+		struct nfs_pathconf pathinfo;
+
+		pathinfo.fattr = fattr;
+		nfs_fattr_init(fattr);
+
+		if (clp->rpc_ops->pathconf(server, mntfh, &pathinfo) >= 0)
+			server->namelen = pathinfo.max_namelen;
+	}
+
+	dprintk("<-- nfs_probe_fsinfo() = 0\n");
+	return 0;
+
+out_error:
+	dprintk("nfs_probe_fsinfo: error = %d\n", -error);
+	return error;
+}
+
+/*
+ * Copy useful information when duplicating a server record
+ */
+static void nfs_server_copy_userdata(struct nfs_server *target, struct nfs_server *source)
+{
+	target->flags = source->flags;
+	target->acregmin = source->acregmin;
+	target->acregmax = source->acregmax;
+	target->acdirmin = source->acdirmin;
+	target->acdirmax = source->acdirmax;
+	target->caps = source->caps;
+}
+
+/*
+ * Allocate and initialise a server record
+ */
+static struct nfs_server *nfs_alloc_server(void)
+{
+	struct nfs_server *server;
+
+	server = kzalloc(sizeof(struct nfs_server), GFP_KERNEL);
+	if (!server)
+		return NULL;
+
+	server->client = server->client_acl = ERR_PTR(-EINVAL);
+
+	/* Zero out the NFS state stuff */
+	INIT_LIST_HEAD(&server->client_link);
+	INIT_LIST_HEAD(&server->master_link);
+
+	server->io_stats = nfs_alloc_iostats();
+	if (!server->io_stats) {
+		kfree(server);
+		return NULL;
+	}
+
+	return server;
+}
+
+/*
+ * Free up a server record
+ */
+void nfs_free_server(struct nfs_server *server)
+{
+	dprintk("--> nfs_free_server()\n");
+
+	spin_lock(&nfs_client_lock);
+	list_del(&server->client_link);
+	list_del(&server->master_link);
+	spin_unlock(&nfs_client_lock);
+
+	if (server->destroy != NULL)
+		server->destroy(server);
+	if (!IS_ERR(server->client))
+		rpc_shutdown_client(server->client);
+
+	nfs_put_client(server->nfs_client);
+
+	nfs_free_iostats(server->io_stats);
+	kfree(server);
+	nfs_release_automount_timer();
+	dprintk("<-- nfs_free_server()\n");
+}
+
+/*
+ * Create a version 2 or 3 volume record
+ * - keyed on server and FSID
+ */
+struct nfs_server *nfs_create_server(const struct nfs_mount_data *data,
+				     struct nfs_fh *mntfh)
+{
+	struct nfs_server *server;
+	struct nfs_fattr fattr;
+	int error;
+
+	server = nfs_alloc_server();
+	if (!server)
+		return ERR_PTR(-ENOMEM);
+
+	/* Get a client representation */
+	error = nfs_init_server(server, data);
+	if (error < 0)
+		goto error;
+
+	BUG_ON(!server->nfs_client);
+	BUG_ON(!server->nfs_client->rpc_ops);
+	BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops);
+
+	/* Probe the root fh to retrieve its FSID */
+	error = nfs_probe_fsinfo(server, mntfh, &fattr);
+	if (error < 0)
+		goto error;
+	if (!(fattr.valid & NFS_ATTR_FATTR)) {
+		error = server->nfs_client->rpc_ops->getattr(server, mntfh, &fattr);
+		if (error < 0) {
+			dprintk("nfs_create_server: getattr error = %d\n", -error);
+			goto error;
+		}
+	}
+	memcpy(&server->fsid, &fattr.fsid, sizeof(server->fsid));
+
+	dprintk("Server FSID: %llx:%llx\n", server->fsid.major, server->fsid.minor);
+
+	BUG_ON(!server->nfs_client);
+	BUG_ON(!server->nfs_client->rpc_ops);
+	BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops);
+
+	spin_lock(&nfs_client_lock);
+	list_add_tail(&server->client_link, &server->nfs_client->cl_superblocks);
+	list_add_tail(&server->master_link, &nfs_volume_list);
+	spin_unlock(&nfs_client_lock);
+
+	server->mount_time = jiffies;
+	return server;
+
+error:
+	nfs_free_server(server);
+	return ERR_PTR(error);
+}
+
+#ifdef CONFIG_NFS_V4
+/*
+ * Initialise an NFS4 client record
+ */
+static int nfs4_init_client(struct nfs_client *clp,
+		int proto, int timeo, int retrans,
+		rpc_authflavor_t authflavour)
+{
+	int error;
+
+	if (clp->cl_cons_state == NFS_CS_READY) {
+		/* the client is initialised already */
+		dprintk("<-- nfs4_init_client() = 0 [already %p]\n", clp);
+		return 0;
+	}
+
+	/* Check NFS protocol revision and initialize RPC op vector */
+	clp->rpc_ops = &nfs_v4_clientops;
+
+	error = nfs_create_rpc_client(clp, proto, timeo, retrans, authflavour);
+	if (error < 0)
+		goto error;
+
+	error = nfs_idmap_new(clp);
+	if (error < 0) {
+		dprintk("%s: failed to create idmapper. Error = %d\n",
+			__FUNCTION__, error);
+		__set_bit(NFS_CS_IDMAP, &clp->cl_res_state);
+		goto error;
+	}
+
+	nfs_mark_client_ready(clp, NFS_CS_READY);
+	return 0;
+
+error:
+	nfs_mark_client_ready(clp, error);
+	dprintk("<-- nfs4_init_client() = xerror %d\n", error);
+	return error;
+}
+
+/*
+ * Set up an NFS4 client
+ */
+static int nfs4_set_client(struct nfs_server *server,
+		const char *hostname, const struct sockaddr_in *addr,
+		rpc_authflavor_t authflavour,
+		int proto, int timeo, int retrans)
+{
+	struct nfs_client *clp;
+	int error;
+
+	dprintk("--> nfs4_set_client()\n");
+
+	/* Allocate or find a client reference we can use */
+	clp = nfs_get_client(hostname, addr, 4);
+	if (IS_ERR(clp)) {
+		error = PTR_ERR(clp);
+		goto error;
+	}
+	error = nfs4_init_client(clp, proto, timeo, retrans, authflavour);
+	if (error < 0)
+		goto error_put;
+
+	server->nfs_client = clp;
+	dprintk("<-- nfs4_set_client() = 0 [new %p]\n", clp);
+	return 0;
+
+error_put:
+	nfs_put_client(clp);
+error:
+	dprintk("<-- nfs4_set_client() = xerror %d\n", error);
+	return error;
+}
+
+/*
+ * Create a version 4 volume record
+ */
+static int nfs4_init_server(struct nfs_server *server,
+		const struct nfs4_mount_data *data, rpc_authflavor_t authflavour)
+{
+	int error;
+
+	dprintk("--> nfs4_init_server()\n");
+
+	/* Initialise the client representation from the mount data */
+	server->flags = data->flags & NFS_MOUNT_FLAGMASK;
+	server->caps |= NFS_CAP_ATOMIC_OPEN;
+
+	if (data->rsize)
+		server->rsize = nfs_block_size(data->rsize, NULL);
+	if (data->wsize)
+		server->wsize = nfs_block_size(data->wsize, NULL);
+
+	server->acregmin = data->acregmin * HZ;
+	server->acregmax = data->acregmax * HZ;
+	server->acdirmin = data->acdirmin * HZ;
+	server->acdirmax = data->acdirmax * HZ;
+
+	error = nfs_init_server_rpcclient(server, authflavour);
+
+	/* Done */
+	dprintk("<-- nfs4_init_server() = %d\n", error);
+	return error;
+}
+
+/*
+ * Create a version 4 volume record
+ * - keyed on server and FSID
+ */
+struct nfs_server *nfs4_create_server(const struct nfs4_mount_data *data,
+				      const char *hostname,
+				      const struct sockaddr_in *addr,
+				      const char *mntpath,
+				      const char *ip_addr,
+				      rpc_authflavor_t authflavour,
+				      struct nfs_fh *mntfh)
+{
+	struct nfs_fattr fattr;
+	struct nfs_server *server;
+	int error;
+
+	dprintk("--> nfs4_create_server()\n");
+
+	server = nfs_alloc_server();
+	if (!server)
+		return ERR_PTR(-ENOMEM);
+
+	/* Get a client record */
+	error = nfs4_set_client(server, hostname, addr, authflavour,
+			data->proto, data->timeo, data->retrans);
+	if (error < 0)
+		goto error;
+
+	/* set up the general RPC client */
+	error = nfs4_init_server(server, data, authflavour);
+	if (error < 0)
+		goto error;
+
+	BUG_ON(!server->nfs_client);
+	BUG_ON(!server->nfs_client->rpc_ops);
+	BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops);
+
+	/* Probe the root fh to retrieve its FSID */
+	error = nfs4_path_walk(server, mntfh, mntpath);
+	if (error < 0)
+		goto error;
+
+	dprintk("Server FSID: %llx:%llx\n", server->fsid.major, server->fsid.minor);
+	dprintk("Mount FH: %d\n", mntfh->size);
+
+	error = nfs_probe_fsinfo(server, mntfh, &fattr);
+	if (error < 0)
+		goto error;
+
+	BUG_ON(!server->nfs_client);
+	BUG_ON(!server->nfs_client->rpc_ops);
+	BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops);
+
+	spin_lock(&nfs_client_lock);
+	list_add_tail(&server->client_link, &server->nfs_client->cl_superblocks);
+	list_add_tail(&server->master_link, &nfs_volume_list);
+	spin_unlock(&nfs_client_lock);
+
+	server->mount_time = jiffies;
+	dprintk("<-- nfs4_create_server() = %p\n", server);
+	return server;
+
+error:
+	nfs_free_server(server);
+	dprintk("<-- nfs4_create_server() = error %d\n", error);
+	return ERR_PTR(error);
+}
+
+/*
+ * Create an NFS4 referral server record
+ */
+struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *data,
+					       struct nfs_fh *fh)
+{
+	struct nfs_client *parent_client;
+	struct nfs_server *server, *parent_server;
+	struct nfs_fattr fattr;
+	int error;
+
+	dprintk("--> nfs4_create_referral_server()\n");
+
+	server = nfs_alloc_server();
+	if (!server)
+		return ERR_PTR(-ENOMEM);
+
+	parent_server = NFS_SB(data->sb);
+	parent_client = parent_server->nfs_client;
+
+	/* Get a client representation.
+	 * Note: NFSv4 always uses TCP, */
+	error = nfs4_set_client(server, data->hostname, data->addr,
+			data->authflavor,
+			parent_server->client->cl_xprt->prot,
+			parent_client->retrans_timeo,
+			parent_client->retrans_count);
+
+	/* Initialise the client representation from the parent server */
+	nfs_server_copy_userdata(server, parent_server);
+	server->caps |= NFS_CAP_ATOMIC_OPEN;
+
+	error = nfs_init_server_rpcclient(server, data->authflavor);
+	if (error < 0)
+		goto error;
+
+	BUG_ON(!server->nfs_client);
+	BUG_ON(!server->nfs_client->rpc_ops);
+	BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops);
+
+	/* probe the filesystem info for this server filesystem */
+	error = nfs_probe_fsinfo(server, fh, &fattr);
+	if (error < 0)
+		goto error;
+
+	dprintk("Referral FSID: %llx:%llx\n",
+		server->fsid.major, server->fsid.minor);
+
+	spin_lock(&nfs_client_lock);
+	list_add_tail(&server->client_link, &server->nfs_client->cl_superblocks);
+	list_add_tail(&server->master_link, &nfs_volume_list);
+	spin_unlock(&nfs_client_lock);
+
+	server->mount_time = jiffies;
+
+	dprintk("<-- nfs_create_referral_server() = %p\n", server);
+	return server;
+
+error:
+	nfs_free_server(server);
+	dprintk("<-- nfs4_create_referral_server() = error %d\n", error);
+	return ERR_PTR(error);
+}
+
+#endif /* CONFIG_NFS_V4 */
+
+/*
+ * Clone an NFS2, NFS3 or NFS4 server record
+ */
+struct nfs_server *nfs_clone_server(struct nfs_server *source,
+				    struct nfs_fh *fh,
+				    struct nfs_fattr *fattr)
+{
+	struct nfs_server *server;
+	struct nfs_fattr fattr_fsinfo;
+	int error;
+
+	dprintk("--> nfs_clone_server(,%llx:%llx,)\n",
+		fattr->fsid.major, fattr->fsid.minor);
+
+	server = nfs_alloc_server();
+	if (!server)
+		return ERR_PTR(-ENOMEM);
+
+	/* Copy data from the source */
+	server->nfs_client = source->nfs_client;
+	atomic_inc(&server->nfs_client->cl_count);
+	nfs_server_copy_userdata(server, source);
+
+	server->fsid = fattr->fsid;
+
+	error = nfs_init_server_rpcclient(server, source->client->cl_auth->au_flavor);
+	if (error < 0)
+		goto out_free_server;
+	if (!IS_ERR(source->client_acl))
+		nfs_init_server_aclclient(server);
+
+	/* probe the filesystem info for this server filesystem */
+	error = nfs_probe_fsinfo(server, fh, &fattr_fsinfo);
+	if (error < 0)
+		goto out_free_server;
+
+	dprintk("Cloned FSID: %llx:%llx\n",
+		server->fsid.major, server->fsid.minor);
+
+	error = nfs_start_lockd(server);
+	if (error < 0)
+		goto out_free_server;
+
+	spin_lock(&nfs_client_lock);
+	list_add_tail(&server->client_link, &server->nfs_client->cl_superblocks);
+	list_add_tail(&server->master_link, &nfs_volume_list);
+	spin_unlock(&nfs_client_lock);
+
+	server->mount_time = jiffies;
+
+	dprintk("<-- nfs_clone_server() = %p\n", server);
+	return server;
+
+out_free_server:
+	nfs_free_server(server);
+	dprintk("<-- nfs_clone_server() = error %d\n", error);
+	return ERR_PTR(error);
+}
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 19362712452f..9b496ef4abea 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -31,6 +31,7 @@
 #include <linux/pagemap.h>
 #include <linux/smp_lock.h>
 #include <linux/namei.h>
+#include <linux/mount.h>
 
 #include "nfs4_fs.h"
 #include "delegation.h"
@@ -870,14 +871,14 @@ int nfs_is_exclusive_create(struct inode *dir, struct nameidata *nd)
 	return (nd->intent.open.flags & O_EXCL) != 0;
 }
 
-static inline int nfs_reval_fsid(struct inode *dir,
-		struct nfs_fh *fh, struct nfs_fattr *fattr)
+static inline int nfs_reval_fsid(struct vfsmount *mnt, struct inode *dir,
+				 struct nfs_fh *fh, struct nfs_fattr *fattr)
 {
 	struct nfs_server *server = NFS_SERVER(dir);
 
 	if (!nfs_fsid_equal(&server->fsid, &fattr->fsid))
 		/* Revalidate fsid on root dir */
-		return __nfs_revalidate_inode(server, dir->i_sb->s_root->d_inode);
+		return __nfs_revalidate_inode(server, mnt->mnt_root->d_inode);
 	return 0;
 }
 
@@ -913,7 +914,7 @@ static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, stru
 		res = ERR_PTR(error);
 		goto out_unlock;
 	}
-	error = nfs_reval_fsid(dir, &fhandle, &fattr);
+	error = nfs_reval_fsid(nd->mnt, dir, &fhandle, &fattr);
 	if (error < 0) {
 		res = ERR_PTR(error);
 		goto out_unlock;
@@ -922,8 +923,9 @@ static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, stru
 	res = (struct dentry *)inode;
 	if (IS_ERR(res))
 		goto out_unlock;
+
 no_entry:
-	res = d_add_unique(dentry, inode);
+	res = d_materialise_unique(dentry, inode);
 	if (res != NULL)
 		dentry = res;
 	nfs_renew_times(dentry);
@@ -1117,11 +1119,13 @@ static struct dentry *nfs_readdir_lookup(nfs_readdir_descriptor_t *desc)
 		dput(dentry);
 		return NULL;
 	}
-	alias = d_add_unique(dentry, inode);
+
+	alias = d_materialise_unique(dentry, inode);
 	if (alias != NULL) {
 		dput(dentry);
 		dentry = alias;
 	}
+
 	nfs_renew_times(dentry);
 	nfs_set_verifier(dentry, nfs_save_change_attribute(dir));
 	return dentry;
diff --git a/fs/nfs/getroot.c b/fs/nfs/getroot.c
new file mode 100644
index 000000000000..977e59088eeb
--- /dev/null
+++ b/fs/nfs/getroot.c
@@ -0,0 +1,306 @@
+/* getroot.c: get the root dentry for an NFS mount
+ *
+ * Copyright (C) 2006 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells@redhat.com)
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version
+ * 2 of the License, or (at your option) any later version.
+ */
+
+#include <linux/config.h>
+#include <linux/module.h>
+#include <linux/init.h>
+
+#include <linux/time.h>
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/string.h>
+#include <linux/stat.h>
+#include <linux/errno.h>
+#include <linux/unistd.h>
+#include <linux/sunrpc/clnt.h>
+#include <linux/sunrpc/stats.h>
+#include <linux/nfs_fs.h>
+#include <linux/nfs_mount.h>
+#include <linux/nfs4_mount.h>
+#include <linux/lockd/bind.h>
+#include <linux/smp_lock.h>
+#include <linux/seq_file.h>
+#include <linux/mount.h>
+#include <linux/nfs_idmap.h>
+#include <linux/vfs.h>
+#include <linux/namei.h>
+#include <linux/namespace.h>
+
+#include <asm/system.h>
+#include <asm/uaccess.h>
+
+#include "nfs4_fs.h"
+#include "delegation.h"
+#include "internal.h"
+
+#define NFSDBG_FACILITY		NFSDBG_CLIENT
+#define NFS_PARANOIA 1
+
+/*
+ * get an NFS2/NFS3 root dentry from the root filehandle
+ */
+struct dentry *nfs_get_root(struct super_block *sb, struct nfs_fh *mntfh)
+{
+	struct nfs_server *server = NFS_SB(sb);
+	struct nfs_fsinfo fsinfo;
+	struct nfs_fattr fattr;
+	struct dentry *mntroot;
+	struct inode *inode;
+	int error;
+
+	/* create a dummy root dentry with dummy inode for this superblock */
+	if (!sb->s_root) {
+		struct nfs_fh dummyfh;
+		struct dentry *root;
+		struct inode *iroot;
+
+		memset(&dummyfh, 0, sizeof(dummyfh));
+		memset(&fattr, 0, sizeof(fattr));
+		nfs_fattr_init(&fattr);
+		fattr.valid = NFS_ATTR_FATTR;
+		fattr.type = NFDIR;
+		fattr.mode = S_IFDIR | S_IRUSR | S_IWUSR;
+		fattr.nlink = 2;
+
+		iroot = nfs_fhget(sb, &dummyfh, &fattr);
+		if (IS_ERR(iroot))
+			return ERR_PTR(PTR_ERR(iroot));
+
+		root = d_alloc_root(iroot);
+		if (!root) {
+			iput(iroot);
+			return ERR_PTR(-ENOMEM);
+		}
+
+		sb->s_root = root;
+	}
+
+	/* get the actual root for this mount */
+	fsinfo.fattr = &fattr;
+
+	error = server->nfs_client->rpc_ops->getroot(server, mntfh, &fsinfo);
+	if (error < 0) {
+		dprintk("nfs_get_root: getattr error = %d\n", -error);
+		return ERR_PTR(error);
+	}
+
+	inode = nfs_fhget(sb, mntfh, fsinfo.fattr);
+	if (IS_ERR(inode)) {
+		dprintk("nfs_get_root: get root inode failed\n");
+		return ERR_PTR(PTR_ERR(inode));
+	}
+
+	/* root dentries normally start off anonymous and get spliced in later
+	 * if the dentry tree reaches them; however if the dentry already
+	 * exists, we'll pick it up at this point and use it as the root
+	 */
+	mntroot = d_alloc_anon(inode);
+	if (!mntroot) {
+		iput(inode);
+		dprintk("nfs_get_root: get root dentry failed\n");
+		return ERR_PTR(-ENOMEM);
+	}
+
+	if (!mntroot->d_op)
+		mntroot->d_op = server->nfs_client->rpc_ops->dentry_ops;
+
+	return mntroot;
+}
+
+#ifdef CONFIG_NFS_V4
+
+/*
+ * Do a simple pathwalk from the root FH of the server to the nominated target
+ * of the mountpoint
+ * - give error on symlinks
+ * - give error on ".." occurring in the path
+ * - follow traversals
+ */
+int nfs4_path_walk(struct nfs_server *server,
+		   struct nfs_fh *mntfh,
+		   const char *path)
+{
+	struct nfs_fsinfo fsinfo;
+	struct nfs_fattr fattr;
+	struct nfs_fh lastfh;
+	struct qstr name;
+	int ret;
+	//int referral_count = 0;
+
+	dprintk("--> nfs4_path_walk(,,%s)\n", path);
+
+	fsinfo.fattr = &fattr;
+	nfs_fattr_init(&fattr);
+
+	if (*path++ != '/') {
+		dprintk("nfs4_get_root: Path does not begin with a slash\n");
+		return -EINVAL;
+	}
+
+	/* Start by getting the root filehandle from the server */
+	ret = server->nfs_client->rpc_ops->getroot(server, mntfh, &fsinfo);
+	if (ret < 0) {
+		dprintk("nfs4_get_root: getroot error = %d\n", -ret);
+		return ret;
+	}
+
+	if (fattr.type != NFDIR) {
+		printk(KERN_ERR "nfs4_get_root:"
+		       " getroot encountered non-directory\n");
+		return -ENOTDIR;
+	}
+
+	if (fattr.valid & NFS_ATTR_FATTR_V4_REFERRAL) {
+		printk(KERN_ERR "nfs4_get_root:"
+		       " getroot obtained referral\n");
+		return -EREMOTE;
+	}
+
+next_component:
+	dprintk("Next: %s\n", path);
+
+	/* extract the next bit of the path */
+	if (!*path)
+		goto path_walk_complete;
+
+	name.name = path;
+	while (*path && *path != '/')
+		path++;
+	name.len = path - (const char *) name.name;
+
+eat_dot_dir:
+	while (*path == '/')
+		path++;
+
+	if (path[0] == '.' && (path[1] == '/' || !path[1])) {
+		path += 2;
+		goto eat_dot_dir;
+	}
+
+	if (path[0] == '.' && path[1] == '.' && (path[2] == '/' || !path[2])
+	    ) {
+		printk(KERN_ERR "nfs4_get_root:"
+		       " Mount path contains reference to \"..\"\n");
+		return -EINVAL;
+	}
+
+	/* lookup the next FH in the sequence */
+	memcpy(&lastfh, mntfh, sizeof(lastfh));
+
+	dprintk("LookupFH: %*.*s [%s]\n", name.len, name.len, name.name, path);
+
+	ret = server->nfs_client->rpc_ops->lookupfh(server, &lastfh, &name,
+						    mntfh, &fattr);
+	if (ret < 0) {
+		dprintk("nfs4_get_root: getroot error = %d\n", -ret);
+		return ret;
+	}
+
+	if (fattr.type != NFDIR) {
+		printk(KERN_ERR "nfs4_get_root:"
+		       " lookupfh encountered non-directory\n");
+		return -ENOTDIR;
+	}
+
+	if (fattr.valid & NFS_ATTR_FATTR_V4_REFERRAL) {
+		printk(KERN_ERR "nfs4_get_root:"
+		       " lookupfh obtained referral\n");
+		return -EREMOTE;
+	}
+
+	goto next_component;
+
+path_walk_complete:
+	memcpy(&server->fsid, &fattr.fsid, sizeof(server->fsid));
+	dprintk("<-- nfs4_path_walk() = 0\n");
+	return 0;
+}
+
+/*
+ * get an NFS4 root dentry from the root filehandle
+ */
+struct dentry *nfs4_get_root(struct super_block *sb, struct nfs_fh *mntfh)
+{
+	struct nfs_server *server = NFS_SB(sb);
+	struct nfs_fattr fattr;
+	struct dentry *mntroot;
+	struct inode *inode;
+	int error;
+
+	dprintk("--> nfs4_get_root()\n");
+
+	/* create a dummy root dentry with dummy inode for this superblock */
+	if (!sb->s_root) {
+		struct nfs_fh dummyfh;
+		struct dentry *root;
+		struct inode *iroot;
+
+		memset(&dummyfh, 0, sizeof(dummyfh));
+		memset(&fattr, 0, sizeof(fattr));
+		nfs_fattr_init(&fattr);
+		fattr.valid = NFS_ATTR_FATTR;
+		fattr.type = NFDIR;
+		fattr.mode = S_IFDIR | S_IRUSR | S_IWUSR;
+		fattr.nlink = 2;
+
+		iroot = nfs_fhget(sb, &dummyfh, &fattr);
+		if (IS_ERR(iroot))
+			return ERR_PTR(PTR_ERR(iroot));
+
+		root = d_alloc_root(iroot);
+		if (!root) {
+			iput(iroot);
+			return ERR_PTR(-ENOMEM);
+		}
+
+		sb->s_root = root;
+	}
+
+	/* get the info about the server and filesystem */
+	error = nfs4_server_capabilities(server, mntfh);
+	if (error < 0) {
+		dprintk("nfs_get_root: getcaps error = %d\n",
+			-error);
+		return ERR_PTR(error);
+	}
+
+	/* get the actual root for this mount */
+	error = server->nfs_client->rpc_ops->getattr(server, mntfh, &fattr);
+	if (error < 0) {
+		dprintk("nfs_get_root: getattr error = %d\n", -error);
+		return ERR_PTR(error);
+	}
+
+	inode = nfs_fhget(sb, mntfh, &fattr);
+	if (IS_ERR(inode)) {
+		dprintk("nfs_get_root: get root inode failed\n");
+		return ERR_PTR(PTR_ERR(inode));
+	}
+
+	/* root dentries normally start off anonymous and get spliced in later
+	 * if the dentry tree reaches them; however if the dentry already
+	 * exists, we'll pick it up at this point and use it as the root
+	 */
+	mntroot = d_alloc_anon(inode);
+	if (!mntroot) {
+		iput(inode);
+		dprintk("nfs_get_root: get root dentry failed\n");
+		return ERR_PTR(-ENOMEM);
+	}
+
+	if (!mntroot->d_op)
+		mntroot->d_op = server->nfs_client->rpc_ops->dentry_ops;
+
+	dprintk("<-- nfs4_get_root()\n");
+	return mntroot;
+}
+
+#endif /* CONFIG_NFS_V4 */
diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c
index 231c20ffc0ff..f96dfac7dc9a 100644
--- a/fs/nfs/idmap.c
+++ b/fs/nfs/idmap.c
@@ -114,8 +114,7 @@ nfs_idmap_new(struct nfs_client *clp)
 	struct idmap *idmap;
 	int error;
 
-	if (clp->cl_idmap != NULL)
-		return 0;
+	BUG_ON(clp->cl_idmap != NULL);
 
         if ((idmap = kzalloc(sizeof(*idmap), GFP_KERNEL)) == NULL)
                 return -ENOMEM;
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 771c3b833757..a547c58a83e6 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -1020,7 +1020,7 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr)
  out_fileid:
 	printk(KERN_ERR "NFS: server %s error: fileid changed\n"
 		"fsid %s: expected fileid 0x%Lx, got 0x%Lx\n",
-		NFS_SERVER(inode)->hostname, inode->i_sb->s_id,
+		NFS_SERVER(inode)->nfs_client->cl_hostname, inode->i_sb->s_id,
 		(long long)nfsi->fileid, (long long)fattr->fileid);
 	goto out_err;
 }
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index 2f3aa52fbefc..e73ba4f1052a 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -4,6 +4,18 @@
 
 #include <linux/mount.h>
 
+struct nfs_string;
+struct nfs_mount_data;
+struct nfs4_mount_data;
+
+/* Maximum number of readahead requests
+ * FIXME: this should really be a sysctl so that users may tune it to suit
+ *        their needs. People that do NFS over a slow network, might for
+ *        instance want to reduce it to something closer to 1 for improved
+ *        interactive response.
+ */
+#define NFS_MAX_READAHEAD	(RPC_DEF_SLOT_TABLE - 1)
+
 struct nfs_clone_mount {
 	const struct super_block *sb;
 	const struct dentry *dentry;
@@ -16,12 +28,25 @@ struct nfs_clone_mount {
 };
 
 /* client.c */
+extern struct rpc_program nfs_program;
+
 extern void nfs_put_client(struct nfs_client *);
 extern struct nfs_client *nfs_find_client(const struct sockaddr_in *, int);
-extern struct nfs_client *nfs_get_client(const char *, const struct sockaddr_in *, int);
-extern void nfs_mark_client_ready(struct nfs_client *, int);
-extern int nfs_create_rpc_client(struct nfs_client *, int, unsigned int,
-				 unsigned int, rpc_authflavor_t);
+extern struct nfs_server *nfs_create_server(const struct nfs_mount_data *,
+					    struct nfs_fh *);
+extern struct nfs_server *nfs4_create_server(const struct nfs4_mount_data *,
+					     const char *,
+					     const struct sockaddr_in *,
+					     const char *,
+					     const char *,
+					     rpc_authflavor_t,
+					     struct nfs_fh *);
+extern struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *,
+						      struct nfs_fh *);
+extern void nfs_free_server(struct nfs_server *server);
+extern struct nfs_server *nfs_clone_server(struct nfs_server *,
+					   struct nfs_fh *,
+					   struct nfs_fattr *);
 
 /* nfs4namespace.c */
 #ifdef CONFIG_NFS_V4
@@ -89,10 +114,10 @@ extern void nfs4_clear_inode(struct inode *);
 #endif
 
 /* super.c */
-extern struct file_system_type nfs_referral_nfs4_fs_type;
-extern struct file_system_type clone_nfs_fs_type;
+extern struct file_system_type nfs_xdev_fs_type;
 #ifdef CONFIG_NFS_V4
-extern struct file_system_type clone_nfs4_fs_type;
+extern struct file_system_type nfs4_xdev_fs_type;
+extern struct file_system_type nfs4_referral_fs_type;
 #endif
 
 extern struct rpc_stat nfs_rpcstat;
@@ -101,28 +126,30 @@ extern int __init register_nfs_fs(void);
 extern void __exit unregister_nfs_fs(void);
 
 /* namespace.c */
-extern char *nfs_path(const char *base, const struct dentry *dentry,
+extern char *nfs_path(const char *base,
+		      const struct dentry *droot,
+		      const struct dentry *dentry,
 		      char *buffer, ssize_t buflen);
 
-/*
- * Determine the mount path as a string
- */
+/* getroot.c */
+extern struct dentry *nfs_get_root(struct super_block *, struct nfs_fh *);
 #ifdef CONFIG_NFS_V4
-static inline char *
-nfs4_path(const struct dentry *dentry, char *buffer, ssize_t buflen)
-{
-	return nfs_path(NFS_SB(dentry->d_sb)->mnt_path, dentry, buffer, buflen);
-}
+extern struct dentry *nfs4_get_root(struct super_block *, struct nfs_fh *);
+
+extern int nfs4_path_walk(struct nfs_server *server,
+			  struct nfs_fh *mntfh,
+			  const char *path);
 #endif
 
 /*
  * Determine the device name as a string
  */
 static inline char *nfs_devname(const struct vfsmount *mnt_parent,
-			 const struct dentry *dentry,
-			 char *buffer, ssize_t buflen)
+				const struct dentry *dentry,
+				char *buffer, ssize_t buflen)
 {
-	return nfs_path(mnt_parent->mnt_devname, dentry, buffer, buflen);
+	return nfs_path(mnt_parent->mnt_devname, mnt_parent->mnt_root,
+			dentry, buffer, buflen);
 }
 
 /*
@@ -178,20 +205,3 @@ void nfs_super_set_maxbytes(struct super_block *sb, __u64 maxfilesize)
 	if (sb->s_maxbytes > MAX_LFS_FILESIZE || sb->s_maxbytes <= 0)
 		sb->s_maxbytes = MAX_LFS_FILESIZE;
 }
-
-/*
- * Check if the string represents a "valid" IPv4 address
- */
-static inline int valid_ipaddr4(const char *buf)
-{
-	int rc, count, in[4];
-
-	rc = sscanf(buf, "%d.%d.%d.%d", &in[0], &in[1], &in[2], &in[3]);
-	if (rc != 4)
-		return -EINVAL;
-	for (count = 0; count < 4; count++) {
-		if (in[count] > 255)
-			return -EINVAL;
-	}
-	return 0;
-}
diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c
index d8b8d56266cb..77b00684894d 100644
--- a/fs/nfs/namespace.c
+++ b/fs/nfs/namespace.c
@@ -2,6 +2,7 @@
  * linux/fs/nfs/namespace.c
  *
  * Copyright (C) 2005 Trond Myklebust <Trond.Myklebust@netapp.com>
+ * - Modified by David Howells <dhowells@redhat.com>
  *
  * NFS namespace
  */
@@ -28,6 +29,7 @@ int nfs_mountpoint_expiry_timeout = 500 * HZ;
 /*
  * nfs_path - reconstruct the path given an arbitrary dentry
  * @base - arbitrary string to prepend to the path
+ * @droot - pointer to root dentry for mountpoint
  * @dentry - pointer to dentry
  * @buffer - result buffer
  * @buflen - length of buffer
@@ -38,7 +40,9 @@ int nfs_mountpoint_expiry_timeout = 500 * HZ;
  * This is mainly for use in figuring out the path on the
  * server side when automounting on top of an existing partition.
  */
-char *nfs_path(const char *base, const struct dentry *dentry,
+char *nfs_path(const char *base,
+	       const struct dentry *droot,
+	       const struct dentry *dentry,
 	       char *buffer, ssize_t buflen)
 {
 	char *end = buffer+buflen;
@@ -47,7 +51,7 @@ char *nfs_path(const char *base, const struct dentry *dentry,
 	*--end = '\0';
 	buflen--;
 	spin_lock(&dcache_lock);
-	while (!IS_ROOT(dentry)) {
+	while (!IS_ROOT(dentry) && dentry != droot) {
 		namelen = dentry->d_name.len;
 		buflen -= namelen + 1;
 		if (buflen < 0)
@@ -96,12 +100,13 @@ static void * nfs_follow_mountpoint(struct dentry *dentry, struct nameidata *nd)
 	struct nfs_fattr fattr;
 	int err;
 
+	dprintk("--> nfs_follow_mountpoint()\n");
+
 	BUG_ON(IS_ROOT(dentry));
 	dprintk("%s: enter\n", __FUNCTION__);
 	dput(nd->dentry);
 	nd->dentry = dget(dentry);
-	if (d_mountpoint(nd->dentry))
-		goto out_follow;
+
 	/* Look it up again */
 	parent = dget_parent(nd->dentry);
 	err = server->nfs_client->rpc_ops->lookup(parent->d_inode,
@@ -134,6 +139,8 @@ static void * nfs_follow_mountpoint(struct dentry *dentry, struct nameidata *nd)
 	schedule_delayed_work(&nfs_automount_task, nfs_mountpoint_expiry_timeout);
 out:
 	dprintk("%s: done, returned %d\n", __FUNCTION__, err);
+
+	dprintk("<-- nfs_follow_mountpoint() = %d\n", err);
 	return ERR_PTR(err);
 out_err:
 	path_release(nd);
@@ -183,14 +190,14 @@ static struct vfsmount *nfs_do_clone_mount(struct nfs_server *server,
 	switch (server->nfs_client->cl_nfsversion) {
 		case 2:
 		case 3:
-			mnt = vfs_kern_mount(&clone_nfs_fs_type, 0, devname, mountdata);
+			mnt = vfs_kern_mount(&nfs_xdev_fs_type, 0, devname, mountdata);
 			break;
 		case 4:
-			mnt = vfs_kern_mount(&clone_nfs4_fs_type, 0, devname, mountdata);
+			mnt = vfs_kern_mount(&nfs4_xdev_fs_type, 0, devname, mountdata);
 	}
 	return mnt;
 #else
-	return vfs_kern_mount(&clone_nfs_fs_type, 0, devname, mountdata);
+	return vfs_kern_mount(&nfs_xdev_fs_type, 0, devname, mountdata);
 #endif
 }
 
@@ -216,6 +223,8 @@ struct vfsmount *nfs_do_submount(const struct vfsmount *mnt_parent,
 	char *page = (char *) __get_free_page(GFP_USER);
 	char *devname;
 
+	dprintk("--> nfs_do_submount()\n");
+
 	dprintk("%s: submounting on %s/%s\n", __FUNCTION__,
 			dentry->d_parent->d_name.name,
 			dentry->d_name.name);
@@ -230,5 +239,7 @@ free_page:
 	free_page((unsigned long)page);
 out:
 	dprintk("%s: done\n", __FUNCTION__);
+
+	dprintk("<-- nfs_do_submount() = %p\n", mnt);
 	return mnt;
 }
diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c
index 0622af0122be..9e8258ece6fd 100644
--- a/fs/nfs/nfs3proc.c
+++ b/fs/nfs/nfs3proc.c
@@ -81,7 +81,7 @@ do_proc_get_root(struct rpc_clnt *client, struct nfs_fh *fhandle,
 }
 
 /*
- * Bare-bones access to getattr: this is for nfs_read_super.
+ * Bare-bones access to getattr: this is for nfs_get_root/nfs_get_sb
  */
 static int
 nfs3_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h
index e7879245361e..61095fe4b5ca 100644
--- a/fs/nfs/nfs4_fs.h
+++ b/fs/nfs/nfs4_fs.h
@@ -188,8 +188,6 @@ extern void nfs4_kill_renewd(struct nfs_client *);
 extern void nfs4_renew_state(void *);
 
 /* nfs4state.c */
-extern void init_nfsv4_state(struct nfs_server *);
-extern void destroy_nfsv4_state(struct nfs_server *);
 struct rpc_cred *nfs4_get_renew_cred(struct nfs_client *clp);
 extern u32 nfs4_alloc_lockowner_id(struct nfs_client *);
 
@@ -224,10 +222,6 @@ extern struct svc_version nfs4_callback_version1;
 
 #else
 
-#define init_nfsv4_state(server)  do { } while (0)
-#define destroy_nfsv4_state(server)       do { } while (0)
-#define nfs4_put_state_owner(inode, owner) do { } while (0)
-#define nfs4_put_open_state(state) do { } while (0)
 #define nfs4_close_state(a, b) do { } while (0)
 
 #endif /* CONFIG_NFS_V4 */
diff --git a/fs/nfs/nfs4namespace.c b/fs/nfs/nfs4namespace.c
index faed9bcba50f..24e47f3bbd17 100644
--- a/fs/nfs/nfs4namespace.c
+++ b/fs/nfs/nfs4namespace.c
@@ -2,6 +2,7 @@
  * linux/fs/nfs/nfs4namespace.c
  *
  * Copyright (C) 2005 Trond Myklebust <Trond.Myklebust@netapp.com>
+ * - Modified by David Howells <dhowells@redhat.com>
  *
  * NFSv4 namespace
  */
@@ -47,6 +48,68 @@ Elong:
 	return ERR_PTR(-ENAMETOOLONG);
 }
 
+/*
+ * Determine the mount path as a string
+ */
+static char *nfs4_path(const struct vfsmount *mnt_parent,
+		       const struct dentry *dentry,
+		       char *buffer, ssize_t buflen)
+{
+	const char *srvpath;
+
+	srvpath = strchr(mnt_parent->mnt_devname, ':');
+	if (srvpath)
+		srvpath++;
+	else
+		srvpath = mnt_parent->mnt_devname;
+
+	return nfs_path(srvpath, mnt_parent->mnt_root, dentry, buffer, buflen);
+}
+
+/*
+ * Check that fs_locations::fs_root [RFC3530 6.3] is a prefix for what we
+ * believe to be the server path to this dentry
+ */
+static int nfs4_validate_fspath(const struct vfsmount *mnt_parent,
+				const struct dentry *dentry,
+				const struct nfs4_fs_locations *locations,
+				char *page, char *page2)
+{
+	const char *path, *fs_path;
+
+	path = nfs4_path(mnt_parent, dentry, page, PAGE_SIZE);
+	if (IS_ERR(path))
+		return PTR_ERR(path);
+
+	fs_path = nfs4_pathname_string(&locations->fs_path, page2, PAGE_SIZE);
+	if (IS_ERR(fs_path))
+		return PTR_ERR(fs_path);
+
+	if (strncmp(path, fs_path, strlen(fs_path)) != 0) {
+		dprintk("%s: path %s does not begin with fsroot %s\n",
+			__FUNCTION__, path, fs_path);
+		return -ENOENT;
+	}
+
+	return 0;
+}
+
+/*
+ * Check if the string represents a "valid" IPv4 address
+ */
+static inline int valid_ipaddr4(const char *buf)
+{
+	int rc, count, in[4];
+
+	rc = sscanf(buf, "%d.%d.%d.%d", &in[0], &in[1], &in[2], &in[3]);
+	if (rc != 4)
+		return -EINVAL;
+	for (count = 0; count < 4; count++) {
+		if (in[count] > 255)
+			return -EINVAL;
+	}
+	return 0;
+}
 
 /**
  * nfs_follow_referral - set up mountpoint when hitting a referral on moved error
@@ -68,10 +131,9 @@ static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent,
 		.dentry = dentry,
 		.authflavor = NFS_SB(mnt_parent->mnt_sb)->client->cl_auth->au_flavor,
 	};
-	char *page, *page2;
-	char *path, *fs_path;
+	char *page = NULL, *page2 = NULL;
 	char *devname;
-	int loc, s;
+	int loc, s, error;
 
 	if (locations == NULL || locations->nlocations <= 0)
 		goto out;
@@ -79,31 +141,25 @@ static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent,
 	dprintk("%s: referral at %s/%s\n", __FUNCTION__,
 		dentry->d_parent->d_name.name, dentry->d_name.name);
 
-	/* Ensure fs path is a prefix of current dentry path */
 	page = (char *) __get_free_page(GFP_USER);
-	if (page == NULL)
+	if (!page)
 		goto out;
+
 	page2 = (char *) __get_free_page(GFP_USER);
-	if (page2 == NULL)
+	if (!page2)
 		goto out;
 
-	path = nfs4_path(dentry, page, PAGE_SIZE);
-	if (IS_ERR(path))
-		goto out_free;
-
-	fs_path = nfs4_pathname_string(&locations->fs_path, page2, PAGE_SIZE);
-	if (IS_ERR(fs_path))
-		goto out_free;
-
-	if (strncmp(path, fs_path, strlen(fs_path)) != 0) {
-		dprintk("%s: path %s does not begin with fsroot %s\n", __FUNCTION__, path, fs_path);
-		goto out_free;
+	/* Ensure fs path is a prefix of current dentry path */
+	error = nfs4_validate_fspath(mnt_parent, dentry, locations, page, page2);
+	if (error < 0) {
+		mnt = ERR_PTR(error);
+		goto out;
 	}
 
 	devname = nfs_devname(mnt_parent, dentry, page, PAGE_SIZE);
 	if (IS_ERR(devname)) {
 		mnt = (struct vfsmount *)devname;
-		goto out_free;
+		goto out;
 	}
 
 	loc = 0;
@@ -140,7 +196,7 @@ static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent,
 			addr.sin_port = htons(NFS_PORT);
 			mountdata.addr = &addr;
 
-			mnt = vfs_kern_mount(&nfs_referral_nfs4_fs_type, 0, devname, &mountdata);
+			mnt = vfs_kern_mount(&nfs4_referral_fs_type, 0, devname, &mountdata);
 			if (!IS_ERR(mnt)) {
 				break;
 			}
@@ -149,10 +205,9 @@ static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent,
 		loc++;
 	}
 
-out_free:
-	free_page((unsigned long)page);
-	free_page((unsigned long)page2);
 out:
+	free_page((unsigned long) page);
+	free_page((unsigned long) page2);
 	dprintk("%s: done\n", __FUNCTION__);
 	return mnt;
 }
@@ -165,7 +220,7 @@ out:
  */
 struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentry *dentry)
 {
-	struct vfsmount *mnt = ERR_PTR(-ENOENT);
+	struct vfsmount *mnt = ERR_PTR(-ENOMEM);
 	struct dentry *parent;
 	struct nfs4_fs_locations *fs_locations = NULL;
 	struct page *page;
@@ -183,11 +238,16 @@ struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentr
 		goto out_free;
 
 	/* Get locations */
+	mnt = ERR_PTR(-ENOENT);
+
 	parent = dget_parent(dentry);
-	dprintk("%s: getting locations for %s/%s\n", __FUNCTION__, parent->d_name.name, dentry->d_name.name);
+	dprintk("%s: getting locations for %s/%s\n",
+		__FUNCTION__, parent->d_name.name, dentry->d_name.name);
+
 	err = nfs4_proc_fs_locations(parent->d_inode, dentry, fs_locations, page);
 	dput(parent);
-	if (err != 0 || fs_locations->nlocations <= 0 ||
+	if (err != 0 ||
+	    fs_locations->nlocations <= 0 ||
 	    fs_locations->fs_path.ncomponents <= 0)
 		goto out_free;
 
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 1573eeb07ce1..a825547e8214 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -1393,70 +1393,19 @@ static int nfs4_lookup_root(struct nfs_server *server, struct nfs_fh *fhandle,
 	return err;
 }
 
+/*
+ * get the file handle for the "/" directory on the server
+ */
 static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
-		struct nfs_fsinfo *info)
+			      struct nfs_fsinfo *info)
 {
-	struct nfs_fattr *	fattr = info->fattr;
-	unsigned char *		p;
-	struct qstr		q;
-	struct nfs4_lookup_arg args = {
-		.dir_fh = fhandle,
-		.name = &q,
-		.bitmask = nfs4_fattr_bitmap,
-	};
-	struct nfs4_lookup_res res = {
-		.server = server,
-		.fattr = fattr,
-		.fh = fhandle,
-	};
-	struct rpc_message msg = {
-		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LOOKUP],
-		.rpc_argp = &args,
-		.rpc_resp = &res,
-	};
 	int status;
 
-	/*
-	 * Now we do a separate LOOKUP for each component of the mount path.
-	 * The LOOKUPs are done separately so that we can conveniently
-	 * catch an ERR_WRONGSEC if it occurs along the way...
-	 */
 	status = nfs4_lookup_root(server, fhandle, info);
-	if (status)
-		goto out;
-
-	p = server->mnt_path;
-	for (;;) {
-		struct nfs4_exception exception = { };
-
-		while (*p == '/')
-			p++;
-		if (!*p)
-			break;
-		q.name = p;
-		while (*p && (*p != '/'))
-			p++;
-		q.len = p - q.name;
-
-		do {
-			nfs_fattr_init(fattr);
-			status = nfs4_handle_exception(server,
-					rpc_call_sync(server->client, &msg, 0),
-					&exception);
-		} while (exception.retry);
-		if (status == 0)
-			continue;
-		if (status == -ENOENT) {
-			printk(KERN_NOTICE "NFS: mount path %s does not exist!\n", server->mnt_path);
-			printk(KERN_NOTICE "NFS: suggestion: try mounting '/' instead.\n");
-		}
-		break;
-	}
 	if (status == 0)
 		status = nfs4_server_capabilities(server, fhandle);
 	if (status == 0)
 		status = nfs4_do_fsinfo(server, fhandle, info);
-out:
 	return nfs4_map_errors(status);
 }
 
diff --git a/fs/nfs/nfs4renewd.c b/fs/nfs/nfs4renewd.c
index ff947ecb8b81..f2c893690ac4 100644
--- a/fs/nfs/nfs4renewd.c
+++ b/fs/nfs/nfs4renewd.c
@@ -127,26 +127,13 @@ nfs4_schedule_state_renewal(struct nfs_client *clp)
 void
 nfs4_renewd_prepare_shutdown(struct nfs_server *server)
 {
-	struct nfs_client *clp = server->nfs_client;
-
-	if (!clp)
-		return;
 	flush_scheduled_work();
-	down_write(&clp->cl_sem);
-	if (!list_empty(&server->nfs4_siblings))
-		list_del_init(&server->nfs4_siblings);
-	up_write(&clp->cl_sem);
 }
 
-/* Must be called with clp->cl_sem locked for writes */
 void
 nfs4_kill_renewd(struct nfs_client *clp)
 {
 	down_read(&clp->cl_sem);
-	if (!list_empty(&clp->cl_superblocks)) {
-		up_read(&clp->cl_sem);
-		return;
-	}
 	cancel_delayed_work(&clp->cl_renewd);
 	up_read(&clp->cl_sem);
 	flush_scheduled_work();
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index 058811e39555..5fffbdfa971f 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -58,24 +58,6 @@ const nfs4_stateid zero_stateid;
 
 static LIST_HEAD(nfs4_clientid_list);
 
-void
-init_nfsv4_state(struct nfs_server *server)
-{
-	server->nfs_client = NULL;
-	INIT_LIST_HEAD(&server->nfs4_siblings);
-}
-
-void
-destroy_nfsv4_state(struct nfs_server *server)
-{
-	kfree(server->mnt_path);
-	server->mnt_path = NULL;
-	if (server->nfs_client) {
-		nfs_put_client(server->nfs_client);
-		server->nfs_client = NULL;
-	}
-}
-
 static int nfs4_init_client(struct nfs_client *clp, struct rpc_cred *cred)
 {
 	int status = nfs4_proc_setclientid(clp, NFS4_CALLBACK,
diff --git a/fs/nfs/read.c b/fs/nfs/read.c
index f0aff824a291..dae33c1e8a77 100644
--- a/fs/nfs/read.c
+++ b/fs/nfs/read.c
@@ -171,7 +171,7 @@ static int nfs_readpage_sync(struct nfs_open_context *ctx, struct inode *inode,
 		rdata->args.offset = page_offset(page) + rdata->args.pgbase;
 
 		dprintk("NFS: nfs_proc_read(%s, (%s/%Ld), %Lu, %u)\n",
-			NFS_SERVER(inode)->hostname,
+			NFS_SERVER(inode)->nfs_client->cl_hostname,
 			inode->i_sb->s_id,
 			(long long)NFS_FILEID(inode),
 			(unsigned long long)rdata->args.pgbase,
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 5842d510d732..867b5dcd3a40 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -13,6 +13,11 @@
  *
  *  Split from inode.c by David Howells <dhowells@redhat.com>
  *
+ * - superblocks are indexed on server only - all inodes, dentries, etc. associated with a
+ *   particular server are held in the same superblock
+ * - NFS superblocks can have several effective roots to the dentry tree
+ * - directory type roots are spliced into the tree when a path from one root reaches the root
+ *   of another (see nfs_lookup())
  */
 
 #include <linux/config.h>
@@ -52,20 +57,12 @@
 
 #define NFSDBG_FACILITY		NFSDBG_VFS
 
-/* Maximum number of readahead requests
- * FIXME: this should really be a sysctl so that users may tune it to suit
- *        their needs. People that do NFS over a slow network, might for
- *        instance want to reduce it to something closer to 1 for improved
- *        interactive response.
- */
-#define NFS_MAX_READAHEAD	(RPC_DEF_SLOT_TABLE - 1)
-
 static void nfs_umount_begin(struct vfsmount *, int);
 static int  nfs_statfs(struct dentry *, struct kstatfs *);
 static int  nfs_show_options(struct seq_file *, struct vfsmount *);
 static int  nfs_show_stats(struct seq_file *, struct vfsmount *);
 static int nfs_get_sb(struct file_system_type *, int, const char *, void *, struct vfsmount *);
-static int nfs_clone_nfs_sb(struct file_system_type *fs_type,
+static int nfs_xdev_get_sb(struct file_system_type *fs_type,
 		int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
 static void nfs_kill_super(struct super_block *);
 
@@ -77,10 +74,10 @@ static struct file_system_type nfs_fs_type = {
 	.fs_flags	= FS_ODD_RENAME|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
 };
 
-struct file_system_type clone_nfs_fs_type = {
+struct file_system_type nfs_xdev_fs_type = {
 	.owner		= THIS_MODULE,
 	.name		= "nfs",
-	.get_sb		= nfs_clone_nfs_sb,
+	.get_sb		= nfs_xdev_get_sb,
 	.kill_sb	= nfs_kill_super,
 	.fs_flags	= FS_ODD_RENAME|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
 };
@@ -99,10 +96,10 @@ static struct super_operations nfs_sops = {
 #ifdef CONFIG_NFS_V4
 static int nfs4_get_sb(struct file_system_type *fs_type,
 	int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
-static int nfs_clone_nfs4_sb(struct file_system_type *fs_type,
-		int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
-static int nfs_referral_nfs4_sb(struct file_system_type *fs_type,
-		int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
+static int nfs4_xdev_get_sb(struct file_system_type *fs_type,
+	int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
+static int nfs4_referral_get_sb(struct file_system_type *fs_type,
+	int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
 static void nfs4_kill_super(struct super_block *sb);
 
 static struct file_system_type nfs4_fs_type = {
@@ -113,18 +110,18 @@ static struct file_system_type nfs4_fs_type = {
 	.fs_flags	= FS_ODD_RENAME|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
 };
 
-struct file_system_type clone_nfs4_fs_type = {
+struct file_system_type nfs4_xdev_fs_type = {
 	.owner		= THIS_MODULE,
 	.name		= "nfs4",
-	.get_sb		= nfs_clone_nfs4_sb,
+	.get_sb		= nfs4_xdev_get_sb,
 	.kill_sb	= nfs4_kill_super,
 	.fs_flags	= FS_ODD_RENAME|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
 };
 
-struct file_system_type nfs_referral_nfs4_fs_type = {
+struct file_system_type nfs4_referral_fs_type = {
 	.owner		= THIS_MODULE,
 	.name		= "nfs4",
-	.get_sb		= nfs_referral_nfs4_sb,
+	.get_sb		= nfs4_referral_get_sb,
 	.kill_sb	= nfs4_kill_super,
 	.fs_flags	= FS_ODD_RENAME|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
 };
@@ -345,7 +342,7 @@ static int nfs_show_options(struct seq_file *m, struct vfsmount *mnt)
 	nfs_show_mount_options(m, nfss, 0);
 
 	seq_puts(m, ",addr=");
-	seq_escape(m, nfss->hostname, " \t\n\\");
+	seq_escape(m, nfss->nfs_client->cl_hostname, " \t\n\\");
 
 	return 0;
 }
@@ -429,714 +426,351 @@ static int nfs_show_stats(struct seq_file *m, struct vfsmount *mnt)
 
 /*
  * Begin unmount by attempting to remove all automounted mountpoints we added
- * in response to traversals
+ * in response to xdev traversals and referrals
  */
 static void nfs_umount_begin(struct vfsmount *vfsmnt, int flags)
 {
-	struct nfs_server *server;
-	struct rpc_clnt	*rpc;
-
 	shrink_submounts(vfsmnt, &nfs_automount_list);
-	if (!(flags & MNT_FORCE))
-		return;
-	/* -EIO all pending I/O */
-	server = NFS_SB(vfsmnt->mnt_sb);
-	rpc = server->client;
-	if (!IS_ERR(rpc))
-		rpc_killall_tasks(rpc);
-	rpc = server->client_acl;
-	if (!IS_ERR(rpc))
-		rpc_killall_tasks(rpc);
 }
 
 /*
- * Obtain the root inode of the file system.
+ * Validate the NFS2/NFS3 mount data
+ * - fills in the mount root filehandle
  */
-static struct inode *
-nfs_get_root(struct super_block *sb, struct nfs_fh *rootfh, struct nfs_fsinfo *fsinfo)
+static int nfs_validate_mount_data(struct nfs_mount_data *data,
+				   struct nfs_fh *mntfh)
 {
-	struct nfs_server	*server = NFS_SB(sb);
-	int			error;
-
-	error = server->nfs_client->rpc_ops->getroot(server, rootfh, fsinfo);
-	if (error < 0) {
-		dprintk("nfs_get_root: getattr error = %d\n", -error);
-		return ERR_PTR(error);
+	if (data == NULL) {
+		dprintk("%s: missing data argument\n", __FUNCTION__);
+		return -EINVAL;
 	}
 
-	server->fsid = fsinfo->fattr->fsid;
-	return nfs_fhget(sb, rootfh, fsinfo->fattr);
-}
-
-/*
- * Do NFS version-independent mount processing, and sanity checking
- */
-static int
-nfs_sb_init(struct super_block *sb, rpc_authflavor_t authflavor)
-{
-	struct nfs_server	*server;
-	struct inode		*root_inode;
-	struct nfs_fattr	fattr;
-	struct nfs_fsinfo	fsinfo = {
-					.fattr = &fattr,
-				};
-	struct nfs_pathconf pathinfo = {
-			.fattr = &fattr,
-	};
-	int no_root_error = 0;
-	unsigned long max_rpc_payload;
-
-	/* We probably want something more informative here */
-	snprintf(sb->s_id, sizeof(sb->s_id), "%x:%x", MAJOR(sb->s_dev), MINOR(sb->s_dev));
-
-	server = NFS_SB(sb);
-
-	sb->s_magic      = NFS_SUPER_MAGIC;
-
-	server->io_stats = nfs_alloc_iostats();
-	if (server->io_stats == NULL)
-		return -ENOMEM;
+	if (data->version <= 0 || data->version > NFS_MOUNT_VERSION) {
+		dprintk("%s: bad mount version\n", __FUNCTION__);
+		return -EINVAL;
+	}
 
-	root_inode = nfs_get_root(sb, &server->fh, &fsinfo);
-	/* Did getting the root inode fail? */
-	if (IS_ERR(root_inode)) {
-		no_root_error = PTR_ERR(root_inode);
-		goto out_no_root;
+	switch (data->version) {
+		case 1:
+			data->namlen = 0;
+		case 2:
+			data->bsize  = 0;
+		case 3:
+			if (data->flags & NFS_MOUNT_VER3) {
+				dprintk("%s: mount structure version %d does not support NFSv3\n",
+						__FUNCTION__,
+						data->version);
+				return -EINVAL;
+			}
+			data->root.size = NFS2_FHSIZE;
+			memcpy(data->root.data, data->old_root.data, NFS2_FHSIZE);
+		case 4:
+			if (data->flags & NFS_MOUNT_SECFLAVOUR) {
+				dprintk("%s: mount structure version %d does not support strong security\n",
+						__FUNCTION__,
+						data->version);
+				return -EINVAL;
+			}
+			/* Fill in pseudoflavor for mount version < 5 */
+			data->pseudoflavor = RPC_AUTH_UNIX;
+		case 5:
+			memset(data->context, 0, sizeof(data->context));
 	}
-	sb->s_root = d_alloc_root(root_inode);
-	if (!sb->s_root) {
-		no_root_error = -ENOMEM;
-		goto out_no_root;
+
+#ifndef CONFIG_NFS_V3
+	/* If NFSv3 is not compiled in, return -EPROTONOSUPPORT */
+	if (data->flags & NFS_MOUNT_VER3) {
+		dprintk("%s: NFSv3 not compiled into kernel\n", __FUNCTION__);
+		return -EPROTONOSUPPORT;
 	}
-	sb->s_root->d_op = server->nfs_client->rpc_ops->dentry_ops;
-
-	/* mount time stamp, in seconds */
-	server->mount_time = jiffies;
-
-	/* Get some general file system info */
-	if (server->namelen == 0 &&
-	    server->nfs_client->rpc_ops->pathconf(server, &server->fh, &pathinfo) >= 0)
-		server->namelen = pathinfo.max_namelen;
-	/* Work out a lot of parameters */
-	if (server->rsize == 0)
-		server->rsize = nfs_block_size(fsinfo.rtpref, NULL);
-	if (server->wsize == 0)
-		server->wsize = nfs_block_size(fsinfo.wtpref, NULL);
-
-	if (fsinfo.rtmax >= 512 && server->rsize > fsinfo.rtmax)
-		server->rsize = nfs_block_size(fsinfo.rtmax, NULL);
-	if (fsinfo.wtmax >= 512 && server->wsize > fsinfo.wtmax)
-		server->wsize = nfs_block_size(fsinfo.wtmax, NULL);
-
-	max_rpc_payload = nfs_block_size(rpc_max_payload(server->client), NULL);
-	if (server->rsize > max_rpc_payload)
-		server->rsize = max_rpc_payload;
-	if (server->rsize > NFS_MAX_FILE_IO_SIZE)
-		server->rsize = NFS_MAX_FILE_IO_SIZE;
-	server->rpages = (server->rsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
-
-	if (server->wsize > max_rpc_payload)
-		server->wsize = max_rpc_payload;
-	if (server->wsize > NFS_MAX_FILE_IO_SIZE)
-		server->wsize = NFS_MAX_FILE_IO_SIZE;
-	server->wpages = (server->wsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
+#endif /* CONFIG_NFS_V3 */
 
-	if (sb->s_blocksize == 0)
-		sb->s_blocksize = nfs_block_bits(server->wsize,
-							 &sb->s_blocksize_bits);
-	server->wtmult = nfs_block_bits(fsinfo.wtmult, NULL);
-
-	server->dtsize = nfs_block_size(fsinfo.dtpref, NULL);
-	if (server->dtsize > PAGE_CACHE_SIZE)
-		server->dtsize = PAGE_CACHE_SIZE;
-	if (server->dtsize > server->rsize)
-		server->dtsize = server->rsize;
-
-	if (server->flags & NFS_MOUNT_NOAC) {
-		server->acregmin = server->acregmax = 0;
-		server->acdirmin = server->acdirmax = 0;
-		sb->s_flags |= MS_SYNCHRONOUS;
+	/* We now require that the mount process passes the remote address */
+	if (data->addr.sin_addr.s_addr == INADDR_ANY) {
+		dprintk("%s: mount program didn't pass remote address!\n",
+			__FUNCTION__);
+		return -EINVAL;
 	}
-	server->backing_dev_info.ra_pages = server->rpages * NFS_MAX_READAHEAD;
 
-	nfs_super_set_maxbytes(sb, fsinfo.maxfilesize);
+	/* Prepare the root filehandle */
+	if (data->flags & NFS_MOUNT_VER3)
+		mntfh->size = data->root.size;
+	else
+		mntfh->size = NFS2_FHSIZE;
 
-	server->client->cl_intr = (server->flags & NFS_MOUNT_INTR) ? 1 : 0;
-	server->client->cl_softrtry = (server->flags & NFS_MOUNT_SOFT) ? 1 : 0;
+	if (mntfh->size > sizeof(mntfh->data)) {
+		dprintk("%s: invalid root filehandle\n", __FUNCTION__);
+		return -EINVAL;
+	}
+
+	memcpy(mntfh->data, data->root.data, mntfh->size);
+	if (mntfh->size < sizeof(mntfh->data))
+		memset(mntfh->data + mntfh->size, 0,
+		       sizeof(mntfh->data) - mntfh->size);
 
-	/* We're airborne Set socket buffersize */
-	rpc_setbufsize(server->client, server->wsize + 100, server->rsize + 100);
 	return 0;
-	/* Yargs. It didn't work out. */
-out_no_root:
-	dprintk("nfs_sb_init: get root inode failed: errno %d\n", -no_root_error);
-	if (!IS_ERR(root_inode))
-		iput(root_inode);
-	return no_root_error;
 }
 
 /*
- * Create an RPC client handle.
+ * Initialise the common bits of the superblock
  */
-static struct rpc_clnt *
-nfs_create_client(struct nfs_server *server, const struct nfs_mount_data *data)
+static inline void nfs_initialise_sb(struct super_block *sb)
 {
-	struct nfs_client	*clp;
-	struct rpc_clnt		*clnt;
-	int			proto = (data->flags & NFS_MOUNT_TCP) ? IPPROTO_TCP : IPPROTO_UDP;
-	int			nfsversion = 2;
-	int			err;
-
-#ifdef CONFIG_NFS_V3
-	if (server->flags & NFS_MOUNT_VER3)
-		nfsversion = 3;
-#endif
-
-	clp = nfs_get_client(server->hostname, &server->addr, nfsversion);
-	if (!clp) {
-		dprintk("%s: failed to create NFS4 client.\n", __FUNCTION__);
-		return ERR_PTR(PTR_ERR(clp));
-	}
-
-	if (clp->cl_cons_state == NFS_CS_INITING) {
-		/* Check NFS protocol revision and initialize RPC op
-		 * vector and file handle pool. */
-#ifdef CONFIG_NFS_V3
-		if (nfsversion == 3) {
-			clp->rpc_ops = &nfs_v3_clientops;
-			server->caps |= NFS_CAP_READDIRPLUS;
-		} else {
-			clp->rpc_ops = &nfs_v2_clientops;
-		}
-#else
-		clp->rpc_ops = &nfs_v2_clientops;
-#endif
-
-		/* create transport and client */
-		err = nfs_create_rpc_client(clp, proto, data->timeo,
-					    data->retrans, RPC_AUTH_UNIX);
-		if (err < 0)
-			goto client_init_error;
-
-		nfs_mark_client_ready(clp, 0);
-	}
+	struct nfs_server *server = NFS_SB(sb);
 
-	/* create an nfs_server-specific client */
-	clnt = rpc_clone_client(clp->cl_rpcclient);
-	if (IS_ERR(clnt)) {
-		dprintk("%s: couldn't create rpc_client!\n", __FUNCTION__);
-		nfs_put_client(clp);
-		return ERR_PTR(PTR_ERR(clnt));
-	}
+	sb->s_magic = NFS_SUPER_MAGIC;
 
-	if (data->pseudoflavor != clp->cl_rpcclient->cl_auth->au_flavor) {
-		struct rpc_auth *auth;
+	/* We probably want something more informative here */
+	snprintf(sb->s_id, sizeof(sb->s_id),
+		 "%x:%x", MAJOR(sb->s_dev), MINOR(sb->s_dev));
 
-		auth = rpcauth_create(data->pseudoflavor, server->client);
-		if (IS_ERR(auth)) {
-			dprintk("%s: couldn't create credcache!\n", __FUNCTION__);
-			return ERR_PTR(PTR_ERR(auth));
-		}
-	}
+	if (sb->s_blocksize == 0)
+		sb->s_blocksize = nfs_block_bits(server->wsize,
+						 &sb->s_blocksize_bits);
 
-	server->nfs_client = clp;
-	return clnt;
+	if (server->flags & NFS_MOUNT_NOAC)
+		sb->s_flags |= MS_SYNCHRONOUS;
 
-client_init_error:
-	nfs_mark_client_ready(clp, err);
-	nfs_put_client(clp);
-	return ERR_PTR(err);
+	nfs_super_set_maxbytes(sb, server->maxfilesize);
 }
 
 /*
- * Clone a server record
+ * Finish setting up an NFS2/3 superblock
  */
-static struct nfs_server *nfs_clone_server(struct super_block *sb, struct nfs_clone_mount *data)
+static void nfs_fill_super(struct super_block *sb, struct nfs_mount_data *data)
 {
 	struct nfs_server *server = NFS_SB(sb);
-	struct nfs_server *parent = NFS_SB(data->sb);
-	struct inode *root_inode;
-	struct nfs_fsinfo fsinfo;
-	void *err = ERR_PTR(-ENOMEM);
-
-	sb->s_op = data->sb->s_op;
-	sb->s_blocksize = data->sb->s_blocksize;
-	sb->s_blocksize_bits = data->sb->s_blocksize_bits;
-	sb->s_maxbytes = data->sb->s_maxbytes;
-
-	server->client_acl = ERR_PTR(-EINVAL);
-	server->io_stats = nfs_alloc_iostats();
-	if (server->io_stats == NULL)
-		goto out;
-
-	server->client = rpc_clone_client(parent->client);
-	if (IS_ERR((err = server->client)))
-		goto out;
-
-	if (!IS_ERR(parent->client_acl)) {
-		server->client_acl = rpc_clone_client(parent->client_acl);
-		if (IS_ERR((err = server->client_acl)))
-			goto out;
-	}
-	root_inode = nfs_fhget(sb, data->fh, data->fattr);
-	if (!root_inode)
-		goto out;
-	sb->s_root = d_alloc_root(root_inode);
-	if (!sb->s_root)
-		goto out_put_root;
-	fsinfo.fattr = data->fattr;
-	if (NFS_PROTO(root_inode)->fsinfo(server, data->fh, &fsinfo) == 0)
-		nfs_super_set_maxbytes(sb, fsinfo.maxfilesize);
-	sb->s_root->d_op = server->nfs_client->rpc_ops->dentry_ops;
-	sb->s_flags |= MS_ACTIVE;
-	return server;
-out_put_root:
-	iput(root_inode);
-out:
-	return err;
-}
-
-/*
- * Copy an existing superblock and attach revised data
- */
-static int nfs_clone_generic_sb(struct nfs_clone_mount *data,
-		struct super_block *(*fill_sb)(struct nfs_server *, struct nfs_clone_mount *),
-		struct nfs_server *(*fill_server)(struct super_block *, struct nfs_clone_mount *),
-		struct vfsmount *mnt)
-{
-	struct nfs_server *server;
-	struct nfs_server *parent = NFS_SB(data->sb);
-	struct super_block *sb = ERR_PTR(-EINVAL);
-	char *hostname;
-	int error = -ENOMEM;
-	int len;
-
-	server = kmalloc(sizeof(struct nfs_server), GFP_KERNEL);
-	if (server == NULL)
-		goto out_err;
-	memcpy(server, parent, sizeof(*server));
-	atomic_inc(&server->nfs_client->cl_count);
-	hostname = (data->hostname != NULL) ? data->hostname : parent->hostname;
-	len = strlen(hostname) + 1;
-	server->hostname = kmalloc(len, GFP_KERNEL);
-	if (server->hostname == NULL)
-		goto free_server;
-	memcpy(server->hostname, hostname, len);
-
-	sb = fill_sb(server, data);
-	if (IS_ERR(sb)) {
-		error = PTR_ERR(sb);
-		goto free_hostname;
-	}
 
-	if (sb->s_root)
-		goto out_share;
+	sb->s_blocksize_bits = 0;
+	sb->s_blocksize = 0;
+	if (data->bsize)
+		sb->s_blocksize = nfs_block_size(data->bsize, &sb->s_blocksize_bits);
 
-	server = fill_server(sb, data);
-	if (IS_ERR(server)) {
-		error = PTR_ERR(server);
-		goto out_deactivate;
+	if (server->flags & NFS_MOUNT_VER3) {
+		/* The VFS shouldn't apply the umask to mode bits. We will do
+		 * so ourselves when necessary.
+		 */
+		sb->s_flags |= MS_POSIXACL;
+		sb->s_time_gran = 1;
 	}
-	return simple_set_mnt(mnt, sb);
-out_deactivate:
-	up_write(&sb->s_umount);
-	deactivate_super(sb);
-	return error;
-out_share:
-	kfree(server->hostname);
-	nfs_put_client(server->nfs_client);
-	kfree(server);
-	return simple_set_mnt(mnt, sb);
-free_hostname:
-	kfree(server->hostname);
-free_server:
-	nfs_put_client(server->nfs_client);
-	kfree(server);
-out_err:
-	return error;
+
+	sb->s_op = &nfs_sops;
+ 	nfs_initialise_sb(sb);
 }
 
 /*
- * Set up an NFS2/3 superblock
- *
- * The way this works is that the mount process passes a structure
- * in the data argument which contains the server's IP address
- * and the root file handle obtained from the server's mount
- * daemon. We stash these away in the private superblock fields.
+ * Finish setting up a cloned NFS2/3 superblock
  */
-static int
-nfs_fill_super(struct super_block *sb, struct nfs_mount_data *data, int silent)
+static void nfs_clone_super(struct super_block *sb,
+			    const struct super_block *old_sb)
 {
-	struct nfs_server	*server;
-	rpc_authflavor_t	authflavor;
+	struct nfs_server *server = NFS_SB(sb);
+
+	sb->s_blocksize_bits = old_sb->s_blocksize_bits;
+	sb->s_blocksize = old_sb->s_blocksize;
+	sb->s_maxbytes = old_sb->s_maxbytes;
 
-	server           = NFS_SB(sb);
-	sb->s_blocksize_bits = 0;
-	sb->s_blocksize = 0;
-	if (data->bsize)
-		sb->s_blocksize = nfs_block_size(data->bsize, &sb->s_blocksize_bits);
-	if (data->rsize)
-		server->rsize = nfs_block_size(data->rsize, NULL);
-	if (data->wsize)
-		server->wsize = nfs_block_size(data->wsize, NULL);
-	server->flags    = data->flags & NFS_MOUNT_FLAGMASK;
-
-	server->acregmin = data->acregmin*HZ;
-	server->acregmax = data->acregmax*HZ;
-	server->acdirmin = data->acdirmin*HZ;
-	server->acdirmax = data->acdirmax*HZ;
-
-	/* Start lockd here, before we might error out */
-	if (!(server->flags & NFS_MOUNT_NONLM))
-		lockd_up();
-
-	server->namelen  = data->namlen;
-	server->hostname = kmalloc(strlen(data->hostname) + 1, GFP_KERNEL);
-	if (!server->hostname)
-		return -ENOMEM;
-	strcpy(server->hostname, data->hostname);
-
-	/* Fill in pseudoflavor for mount version < 5 */
-	if (!(data->flags & NFS_MOUNT_SECFLAVOUR))
-		data->pseudoflavor = RPC_AUTH_UNIX;
-	authflavor = data->pseudoflavor;	/* save for sb_init() */
-	/* XXX maybe we want to add a server->pseudoflavor field */
-
-	/* Create RPC client handles */
-	server->client = nfs_create_client(server, data);
-	if (IS_ERR(server->client))
-		return PTR_ERR(server->client);
-
-	/* RFC 2623, sec 2.3.2 */
 	if (server->flags & NFS_MOUNT_VER3) {
-#ifdef CONFIG_NFS_V3_ACL
-		if (!(server->flags & NFS_MOUNT_NOACL)) {
-			server->client_acl = rpc_bind_new_program(server->client, &nfsacl_program, 3);
-			/* No errors! Assume that Sun nfsacls are supported */
-			if (!IS_ERR(server->client_acl))
-				server->caps |= NFS_CAP_ACLS;
-		}
-#else
-		server->flags &= ~NFS_MOUNT_NOACL;
-#endif /* CONFIG_NFS_V3_ACL */
-		/*
-		 * The VFS shouldn't apply the umask to mode bits. We will
-		 * do so ourselves when necessary.
+		/* The VFS shouldn't apply the umask to mode bits. We will do
+		 * so ourselves when necessary.
 		 */
 		sb->s_flags |= MS_POSIXACL;
-		if (server->namelen == 0 || server->namelen > NFS3_MAXNAMLEN)
-			server->namelen = NFS3_MAXNAMLEN;
 		sb->s_time_gran = 1;
-	} else {
-		if (server->namelen == 0 || server->namelen > NFS2_MAXNAMLEN)
-			server->namelen = NFS2_MAXNAMLEN;
 	}
 
-	sb->s_op = &nfs_sops;
-	return nfs_sb_init(sb, authflavor);
+	sb->s_op = old_sb->s_op;
+ 	nfs_initialise_sb(sb);
 }
 
-static int nfs_set_super(struct super_block *s, void *data)
+static int nfs_set_super(struct super_block *s, void *_server)
 {
-	s->s_fs_info = data;
-	return set_anon_super(s, data);
+	struct nfs_server *server = _server;
+	int ret;
+
+	s->s_fs_info = server;
+	ret = set_anon_super(s, server);
+	if (ret == 0)
+		server->s_dev = s->s_dev;
+	return ret;
 }
 
 static int nfs_compare_super(struct super_block *sb, void *data)
 {
-	struct nfs_server *server = data;
-	struct nfs_server *old = NFS_SB(sb);
+	struct nfs_server *server = data, *old = NFS_SB(sb);
 
-	if (old->addr.sin_addr.s_addr != server->addr.sin_addr.s_addr)
+	if (old->nfs_client != server->nfs_client)
 		return 0;
-	if (old->addr.sin_port != server->addr.sin_port)
+	if (memcmp(&old->fsid, &server->fsid, sizeof(old->fsid)) != 0)
 		return 0;
-	return !nfs_compare_fh(&old->fh, &server->fh);
+	return 1;
 }
 
 static int nfs_get_sb(struct file_system_type *fs_type,
 	int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt)
 {
-	int error;
 	struct nfs_server *server = NULL;
 	struct super_block *s;
-	struct nfs_fh *root;
+	struct nfs_fh mntfh;
 	struct nfs_mount_data *data = raw_data;
+	struct dentry *mntroot;
+	int error;
 
-	error = -EINVAL;
-	if (data == NULL) {
-		dprintk("%s: missing data argument\n", __FUNCTION__);
-		goto out_err_noserver;
-	}
-	if (data->version <= 0 || data->version > NFS_MOUNT_VERSION) {
-		dprintk("%s: bad mount version\n", __FUNCTION__);
-		goto out_err_noserver;
-	}
-	switch (data->version) {
-		case 1:
-			data->namlen = 0;
-		case 2:
-			data->bsize  = 0;
-		case 3:
-			if (data->flags & NFS_MOUNT_VER3) {
-				dprintk("%s: mount structure version %d does not support NFSv3\n",
-						__FUNCTION__,
-						data->version);
-				goto out_err_noserver;
-			}
-			data->root.size = NFS2_FHSIZE;
-			memcpy(data->root.data, data->old_root.data, NFS2_FHSIZE);
-		case 4:
-			if (data->flags & NFS_MOUNT_SECFLAVOUR) {
-				dprintk("%s: mount structure version %d does not support strong security\n",
-						__FUNCTION__,
-						data->version);
-				goto out_err_noserver;
-			}
-		case 5:
-			memset(data->context, 0, sizeof(data->context));
-	}
-#ifndef CONFIG_NFS_V3
-	/* If NFSv3 is not compiled in, return -EPROTONOSUPPORT */
-	error = -EPROTONOSUPPORT;
-	if (data->flags & NFS_MOUNT_VER3) {
-		dprintk("%s: NFSv3 not compiled into kernel\n", __FUNCTION__);
-		goto out_err_noserver;
-	}
-#endif /* CONFIG_NFS_V3 */
+	/* Validate the mount data */
+	error = nfs_validate_mount_data(data, &mntfh);
+	if (error < 0)
+		return error;
 
-	error = -ENOMEM;
-	server = kzalloc(sizeof(struct nfs_server), GFP_KERNEL);
-	if (!server)
+	/* Get a volume representation */
+	server = nfs_create_server(data, &mntfh);
+	if (IS_ERR(server)) {
+		error = PTR_ERR(server);
 		goto out_err_noserver;
-	/* Zero out the NFS state stuff */
-	init_nfsv4_state(server);
-	server->client = server->client_acl = ERR_PTR(-EINVAL);
-
-	root = &server->fh;
-	if (data->flags & NFS_MOUNT_VER3)
-		root->size = data->root.size;
-	else
-		root->size = NFS2_FHSIZE;
-	error = -EINVAL;
-	if (root->size > sizeof(root->data)) {
-		dprintk("%s: invalid root filehandle\n", __FUNCTION__);
-		goto out_err;
-	}
-	memcpy(root->data, data->root.data, root->size);
-
-	/* We now require that the mount process passes the remote address */
-	memcpy(&server->addr, &data->addr, sizeof(server->addr));
-	if (server->addr.sin_addr.s_addr == INADDR_ANY) {
-		dprintk("%s: mount program didn't pass remote address!\n",
-				__FUNCTION__);
-		goto out_err;
 	}
 
+	/* Get a superblock - note that we may end up sharing one that already exists */
 	s = sget(fs_type, nfs_compare_super, nfs_set_super, server);
 	if (IS_ERR(s)) {
 		error = PTR_ERR(s);
-		goto out_err;
+		goto out_err_nosb;
 	}
 
-	if (s->s_root)
-		goto out_share;
+	if (s->s_fs_info != server) {
+		nfs_free_server(server);
+		server = NULL;
+	}
 
-	s->s_flags = flags;
+	if (!s->s_root) {
+		/* initial superblock/root creation */
+		s->s_flags = flags;
+		nfs_fill_super(s, data);
+	}
 
-	error = nfs_fill_super(s, data, flags & MS_SILENT ? 1 : 0);
-	if (error) {
-		up_write(&s->s_umount);
-		deactivate_super(s);
-		return error;
+	mntroot = nfs_get_root(s, &mntfh);
+	if (IS_ERR(mntroot)) {
+		error = PTR_ERR(mntroot);
+		goto error_splat_super;
 	}
-	s->s_flags |= MS_ACTIVE;
-	return simple_set_mnt(mnt, s);
 
-out_share:
-	kfree(server);
-	return simple_set_mnt(mnt, s);
+	s->s_flags |= MS_ACTIVE;
+	mnt->mnt_sb = s;
+	mnt->mnt_root = mntroot;
+	return 0;
 
-out_err:
-	kfree(server);
+out_err_nosb:
+	nfs_free_server(server);
 out_err_noserver:
 	return error;
+
+error_splat_super:
+	up_write(&s->s_umount);
+	deactivate_super(s);
+	return error;
 }
 
+/*
+ * Destroy an NFS2/3 superblock
+ */
 static void nfs_kill_super(struct super_block *s)
 {
 	struct nfs_server *server = NFS_SB(s);
 
 	kill_anon_super(s);
-
-	if (!IS_ERR(server->client))
-		rpc_shutdown_client(server->client);
-	if (!IS_ERR(server->client_acl))
-		rpc_shutdown_client(server->client_acl);
-
-	if (!(server->flags & NFS_MOUNT_NONLM))
-		lockd_down();	/* release rpc.lockd */
-
-	nfs_free_iostats(server->io_stats);
-	kfree(server->hostname);
-	nfs_put_client(server->nfs_client);
-	kfree(server);
-	nfs_release_automount_timer();
+	nfs_free_server(server);
 }
 
-static struct super_block *nfs_clone_sb(struct nfs_server *server, struct nfs_clone_mount *data)
-{
-	struct super_block *sb;
-
-	server->fsid = data->fattr->fsid;
-	nfs_copy_fh(&server->fh, data->fh);
-	sb = sget(&nfs_fs_type, nfs_compare_super, nfs_set_super, server);
-	if (!IS_ERR(sb) && sb->s_root == NULL && !(server->flags & NFS_MOUNT_NONLM))
-		lockd_up();
-	return sb;
-}
-
-static int nfs_clone_nfs_sb(struct file_system_type *fs_type,
-		int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt)
+/*
+ * Clone an NFS2/3 server record on xdev traversal (FSID-change)
+ */
+static int nfs_xdev_get_sb(struct file_system_type *fs_type, int flags,
+			   const char *dev_name, void *raw_data,
+			   struct vfsmount *mnt)
 {
 	struct nfs_clone_mount *data = raw_data;
-	return nfs_clone_generic_sb(data, nfs_clone_sb, nfs_clone_server, mnt);
-}
+	struct super_block *s;
+	struct nfs_server *server;
+	struct dentry *mntroot;
+	int error;
 
-#ifdef CONFIG_NFS_V4
-static struct rpc_clnt *nfs4_create_client(struct nfs_server *server,
-	int timeo, int retrans, int proto, rpc_authflavor_t flavor)
-{
-	struct nfs_client *clp;
-	struct rpc_clnt *clnt = NULL;
-	int err = -EIO;
-
-	clp = nfs_get_client(server->hostname, &server->addr, 4);
-	if (!clp) {
-		dprintk("%s: failed to create NFS4 client.\n", __FUNCTION__);
-		return ERR_PTR(err);
-	}
+	dprintk("--> nfs_xdev_get_sb()\n");
 
-	/* Now create transport and client */
-	if (clp->cl_cons_state == NFS_CS_INITING) {
-		clp->rpc_ops = &nfs_v4_clientops;
+	/* create a new volume representation */
+	server = nfs_clone_server(NFS_SB(data->sb), data->fh, data->fattr);
+	if (IS_ERR(server)) {
+		error = PTR_ERR(server);
+		goto out_err_noserver;
+	}
 
-		err = nfs_create_rpc_client(clp, proto, timeo, retrans, flavor);
-		if (err < 0)
-			goto client_init_error;
+	/* Get a superblock - note that we may end up sharing one that already exists */
+	s = sget(&nfs_fs_type, nfs_compare_super, nfs_set_super, server);
+	if (IS_ERR(s)) {
+		error = PTR_ERR(s);
+		goto out_err_nosb;
+	}
 
-		memcpy(clp->cl_ipaddr, server->ip_addr, sizeof(clp->cl_ipaddr));
-		err = nfs_idmap_new(clp);
-		if (err < 0) {
-			dprintk("%s: failed to create idmapper.\n",
-				__FUNCTION__);
-			goto client_init_error;
-		}
-		__set_bit(NFS_CS_IDMAP, &clp->cl_res_state);
-		nfs_mark_client_ready(clp, 0);
+	if (s->s_fs_info != server) {
+		nfs_free_server(server);
+		server = NULL;
 	}
 
-	clnt = rpc_clone_client(clp->cl_rpcclient);
+	if (!s->s_root) {
+		/* initial superblock/root creation */
+		s->s_flags = flags;
+		nfs_clone_super(s, data->sb);
+	}
 
-	if (IS_ERR(clnt)) {
-		dprintk("%s: cannot create RPC client. Error = %d\n",
-				__FUNCTION__, err);
-		return clnt;
+	mntroot = nfs_get_root(s, data->fh);
+	if (IS_ERR(mntroot)) {
+		error = PTR_ERR(mntroot);
+		goto error_splat_super;
 	}
 
-	if (clnt->cl_auth->au_flavor != flavor) {
-		struct rpc_auth *auth;
+	s->s_flags |= MS_ACTIVE;
+	mnt->mnt_sb = s;
+	mnt->mnt_root = mntroot;
 
-		auth = rpcauth_create(flavor, clnt);
-		if (IS_ERR(auth)) {
-			dprintk("%s: couldn't create credcache!\n", __FUNCTION__);
-			return (struct rpc_clnt *)auth;
-		}
-	}
+	dprintk("<-- nfs_xdev_get_sb() = 0\n");
+	return 0;
 
-	server->nfs_client = clp;
-	down_write(&clp->cl_sem);
-	list_add_tail(&server->nfs4_siblings, &clp->cl_superblocks);
-	up_write(&clp->cl_sem);
-	return clnt;
+out_err_nosb:
+	nfs_free_server(server);
+out_err_noserver:
+	dprintk("<-- nfs_xdev_get_sb() = %d [error]\n", error);
+	return error;
 
-client_init_error:
-	nfs_mark_client_ready(clp, err);
-	nfs_put_client(clp);
-	return ERR_PTR(err);
+error_splat_super:
+	up_write(&s->s_umount);
+	deactivate_super(s);
+	dprintk("<-- nfs_xdev_get_sb() = %d [splat]\n", error);
+	return error;
 }
 
+#ifdef CONFIG_NFS_V4
+
 /*
- * Set up an NFS4 superblock
+ * Finish setting up a cloned NFS4 superblock
  */
-static int nfs4_fill_super(struct super_block *sb, struct nfs4_mount_data *data, int silent)
+static void nfs4_clone_super(struct super_block *sb,
+			    const struct super_block *old_sb)
 {
-	struct nfs_server *server;
-	rpc_authflavor_t authflavour;
-	int err = -EIO;
-
-	sb->s_blocksize_bits = 0;
-	sb->s_blocksize = 0;
-	server = NFS_SB(sb);
-	if (data->rsize != 0)
-		server->rsize = nfs_block_size(data->rsize, NULL);
-	if (data->wsize != 0)
-		server->wsize = nfs_block_size(data->wsize, NULL);
-	server->flags = data->flags & NFS_MOUNT_FLAGMASK;
-	server->caps = NFS_CAP_ATOMIC_OPEN;
-
-	server->acregmin = data->acregmin*HZ;
-	server->acregmax = data->acregmax*HZ;
-	server->acdirmin = data->acdirmin*HZ;
-	server->acdirmax = data->acdirmax*HZ;
-
-	/* Now create transport and client */
-	authflavour = RPC_AUTH_UNIX;
-	if (data->auth_flavourlen != 0) {
-		if (data->auth_flavourlen != 1) {
-			dprintk("%s: Invalid number of RPC auth flavours %d.\n",
-					__FUNCTION__, data->auth_flavourlen);
-			err = -EINVAL;
-			goto out_fail;
-		}
-		if (copy_from_user(&authflavour, data->auth_flavours, sizeof(authflavour))) {
-			err = -EFAULT;
-			goto out_fail;
-		}
-	}
-
-	server->client = nfs4_create_client(server, data->timeo, data->retrans,
-					    data->proto, authflavour);
-	if (IS_ERR(server->client)) {
-		err = PTR_ERR(server->client);
-			dprintk("%s: cannot create RPC client. Error = %d\n",
-					__FUNCTION__, err);
-			goto out_fail;
-	}
-
+	sb->s_blocksize_bits = old_sb->s_blocksize_bits;
+	sb->s_blocksize = old_sb->s_blocksize;
+	sb->s_maxbytes = old_sb->s_maxbytes;
 	sb->s_time_gran = 1;
-
-	sb->s_op = &nfs4_sops;
-	err = nfs_sb_init(sb, authflavour);
-
- out_fail:
-	return err;
+	sb->s_op = old_sb->s_op;
+ 	nfs_initialise_sb(sb);
 }
 
-static int nfs4_compare_super(struct super_block *sb, void *data)
+/*
+ * Set up an NFS4 superblock
+ */
+static void nfs4_fill_super(struct super_block *sb)
 {
-	struct nfs_server *server = data;
-	struct nfs_server *old = NFS_SB(sb);
-
-	if (strcmp(server->hostname, old->hostname) != 0)
-		return 0;
-	if (strcmp(server->mnt_path, old->mnt_path) != 0)
-		return 0;
-	return 1;
+	sb->s_time_gran = 1;
+	sb->s_op = &nfs4_sops;
+	nfs_initialise_sb(sb);
 }
 
-static void *
-nfs_copy_user_string(char *dst, struct nfs_string *src, int maxlen)
+static void *nfs_copy_user_string(char *dst, struct nfs_string *src, int maxlen)
 {
 	void *p = NULL;
 
@@ -1157,14 +791,22 @@ nfs_copy_user_string(char *dst, struct nfs_string *src, int maxlen)
 	return dst;
 }
 
+/*
+ * Get the superblock for an NFS4 mountpoint
+ */
 static int nfs4_get_sb(struct file_system_type *fs_type,
 	int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt)
 {
-	int error;
-	struct nfs_server *server;
-	struct super_block *s;
 	struct nfs4_mount_data *data = raw_data;
+	struct super_block *s;
+	struct nfs_server *server;
+	struct sockaddr_in addr;
+	rpc_authflavor_t authflavour;
+	struct nfs_fh mntfh;
+	struct dentry *mntroot;
+	char *mntpath = NULL, *hostname = NULL, ip_addr[16];
 	void *p;
+	int error;
 
 	if (data == NULL) {
 		dprintk("%s: missing data argument\n", __FUNCTION__);
@@ -1175,75 +817,107 @@ static int nfs4_get_sb(struct file_system_type *fs_type,
 		return -EINVAL;
 	}
 
-	server = kzalloc(sizeof(struct nfs_server), GFP_KERNEL);
-	if (!server)
-		return -ENOMEM;
-	/* Zero out the NFS state stuff */
-	init_nfsv4_state(server);
-	server->client = server->client_acl = ERR_PTR(-EINVAL);
+	/* We now require that the mount process passes the remote address */
+	if (data->host_addrlen != sizeof(addr))
+		return -EINVAL;
+
+	if (copy_from_user(&addr, data->host_addr, sizeof(addr)))
+		return -EFAULT;
+
+	if (addr.sin_family != AF_INET ||
+	    addr.sin_addr.s_addr == INADDR_ANY
+	    ) {
+		dprintk("%s: mount program didn't pass remote IP address!\n",
+				__FUNCTION__);
+		return -EINVAL;
+	}
+
+	/* Grab the authentication type */
+	authflavour = RPC_AUTH_UNIX;
+	if (data->auth_flavourlen != 0) {
+		if (data->auth_flavourlen != 1) {
+			dprintk("%s: Invalid number of RPC auth flavours %d.\n",
+					__FUNCTION__, data->auth_flavourlen);
+			error = -EINVAL;
+			goto out_err_noserver;
+		}
+
+		if (copy_from_user(&authflavour, data->auth_flavours,
+				   sizeof(authflavour))) {
+			error = -EFAULT;
+			goto out_err_noserver;
+		}
+	}
 
 	p = nfs_copy_user_string(NULL, &data->hostname, 256);
 	if (IS_ERR(p))
 		goto out_err;
-	server->hostname = p;
+	hostname = p;
 
 	p = nfs_copy_user_string(NULL, &data->mnt_path, 1024);
 	if (IS_ERR(p))
 		goto out_err;
-	server->mnt_path = p;
+	mntpath = p;
 
-	p = nfs_copy_user_string(server->ip_addr, &data->client_addr,
-			sizeof(server->ip_addr) - 1);
+	dprintk("MNTPATH: %s\n", mntpath);
+
+	p = nfs_copy_user_string(ip_addr, &data->client_addr,
+				 sizeof(ip_addr) - 1);
 	if (IS_ERR(p))
 		goto out_err;
 
-	/* We now require that the mount process passes the remote address */
-	if (data->host_addrlen != sizeof(server->addr)) {
-		error = -EINVAL;
-		goto out_free;
-	}
-	if (copy_from_user(&server->addr, data->host_addr, sizeof(server->addr))) {
-		error = -EFAULT;
-		goto out_free;
-	}
-	if (server->addr.sin_family != AF_INET ||
-	    server->addr.sin_addr.s_addr == INADDR_ANY) {
-		dprintk("%s: mount program didn't pass remote IP address!\n",
-				__FUNCTION__);
-		error = -EINVAL;
-		goto out_free;
+	/* Get a volume representation */
+	server = nfs4_create_server(data, hostname, &addr, mntpath, ip_addr,
+				    authflavour, &mntfh);
+	if (IS_ERR(server)) {
+		error = PTR_ERR(server);
+		goto out_err_noserver;
 	}
 
-	s = sget(fs_type, nfs4_compare_super, nfs_set_super, server);
+	/* Get a superblock - note that we may end up sharing one that already exists */
+	s = sget(fs_type, nfs_compare_super, nfs_set_super, server);
 	if (IS_ERR(s)) {
 		error = PTR_ERR(s);
 		goto out_free;
 	}
 
-	if (s->s_root) {
-		kfree(server->mnt_path);
-		kfree(server->hostname);
-		kfree(server);
-		return simple_set_mnt(mnt, s);
-	}
+	if (!s->s_root) {
+		/* initial superblock/root creation */
+		s->s_flags = flags;
 
-	s->s_flags = flags;
+		nfs4_fill_super(s);
+	} else {
+		nfs_free_server(server);
+	}
 
-	error = nfs4_fill_super(s, data, flags & MS_SILENT ? 1 : 0);
-	if (error) {
-		up_write(&s->s_umount);
-		deactivate_super(s);
-		return error;
+	mntroot = nfs4_get_root(s, &mntfh);
+	if (IS_ERR(mntroot)) {
+		error = PTR_ERR(mntroot);
+		goto error_splat_super;
 	}
+
 	s->s_flags |= MS_ACTIVE;
-	return simple_set_mnt(mnt, s);
+	mnt->mnt_sb = s;
+	mnt->mnt_root = mntroot;
+	kfree(mntpath);
+	kfree(hostname);
+	return 0;
+
 out_err:
 	error = PTR_ERR(p);
+	goto out_err_noserver;
+
 out_free:
-	kfree(server->mnt_path);
-	kfree(server->hostname);
-	kfree(server);
+	nfs_free_server(server);
+out_err_noserver:
+	kfree(mntpath);
+	kfree(hostname);
 	return error;
+
+error_splat_super:
+	up_write(&s->s_umount);
+	deactivate_super(s);
+	goto out_err_noserver;
 }
 
 static void nfs4_kill_super(struct super_block *sb)
@@ -1254,133 +928,140 @@ static void nfs4_kill_super(struct super_block *sb)
 	kill_anon_super(sb);
 
 	nfs4_renewd_prepare_shutdown(server);
-
-	if (server->client != NULL && !IS_ERR(server->client))
-		rpc_shutdown_client(server->client);
-
-	destroy_nfsv4_state(server);
-
-	nfs_free_iostats(server->io_stats);
-	kfree(server->hostname);
-	kfree(server);
-	nfs_release_automount_timer();
+	nfs_free_server(server);
 }
 
 /*
- * Constructs the SERVER-side path
+ * Clone an NFS4 server record on xdev traversal (FSID-change)
  */
-static inline char *nfs4_dup_path(const struct dentry *dentry)
+static int nfs4_xdev_get_sb(struct file_system_type *fs_type, int flags,
+			    const char *dev_name, void *raw_data,
+			    struct vfsmount *mnt)
 {
-	char *page = (char *) __get_free_page(GFP_USER);
-	char *path;
+	struct nfs_clone_mount *data = raw_data;
+	struct super_block *s;
+	struct nfs_server *server;
+	struct dentry *mntroot;
+	int error;
 
-	path = nfs4_path(dentry, page, PAGE_SIZE);
-	if (!IS_ERR(path)) {
-		int len = PAGE_SIZE + page - path;
-		char *tmp = path;
+	dprintk("--> nfs4_xdev_get_sb()\n");
 
-		path = kmalloc(len, GFP_KERNEL);
-		if (path)
-			memcpy(path, tmp, len);
-		else
-			path = ERR_PTR(-ENOMEM);
+	/* create a new volume representation */
+	server = nfs_clone_server(NFS_SB(data->sb), data->fh, data->fattr);
+	if (IS_ERR(server)) {
+		error = PTR_ERR(server);
+		goto out_err_noserver;
 	}
-	free_page((unsigned long)page);
-	return path;
-}
 
-static struct super_block *nfs4_clone_sb(struct nfs_server *server, struct nfs_clone_mount *data)
-{
-	const struct dentry *dentry = data->dentry;
-	struct nfs_client *clp = server->nfs_client;
-	struct super_block *sb;
-
-	server->fsid = data->fattr->fsid;
-	nfs_copy_fh(&server->fh, data->fh);
-	server->mnt_path = nfs4_dup_path(dentry);
-	if (IS_ERR(server->mnt_path)) {
-		sb = (struct super_block *)server->mnt_path;
-		goto err;
+	/* Get a superblock - note that we may end up sharing one that already exists */
+	s = sget(&nfs_fs_type, nfs_compare_super, nfs_set_super, server);
+	if (IS_ERR(s)) {
+		error = PTR_ERR(s);
+		goto out_err_nosb;
 	}
-	sb = sget(&nfs4_fs_type, nfs4_compare_super, nfs_set_super, server);
-	if (IS_ERR(sb) || sb->s_root)
-		goto free_path;
-	nfs4_server_capabilities(server, &server->fh);
-
-	down_write(&clp->cl_sem);
-	list_add_tail(&server->nfs4_siblings, &clp->cl_superblocks);
-	up_write(&clp->cl_sem);
-	return sb;
-free_path:
-	kfree(server->mnt_path);
-err:
-	server->mnt_path = NULL;
-	return sb;
-}
 
-static int nfs_clone_nfs4_sb(struct file_system_type *fs_type,
-		int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt)
-{
-	struct nfs_clone_mount *data = raw_data;
-	return nfs_clone_generic_sb(data, nfs4_clone_sb, nfs_clone_server, mnt);
-}
+	if (s->s_fs_info != server) {
+		nfs_free_server(server);
+		server = NULL;
+	}
 
-static struct super_block *nfs4_referral_sb(struct nfs_server *server, struct nfs_clone_mount *data)
-{
-	struct super_block *sb = ERR_PTR(-ENOMEM);
-	int len;
-
-	len = strlen(data->mnt_path) + 1;
-	server->mnt_path = kmalloc(len, GFP_KERNEL);
-	if (server->mnt_path == NULL)
-		goto err;
-	memcpy(server->mnt_path, data->mnt_path, len);
-	memcpy(&server->addr, data->addr, sizeof(struct sockaddr_in));
-
-	sb = sget(&nfs4_fs_type, nfs4_compare_super, nfs_set_super, server);
-	if (IS_ERR(sb) || sb->s_root)
-		goto free_path;
-	return sb;
-free_path:
-	kfree(server->mnt_path);
-err:
-	server->mnt_path = NULL;
-	return sb;
-}
+	if (!s->s_root) {
+		/* initial superblock/root creation */
+		s->s_flags = flags;
+		nfs4_clone_super(s, data->sb);
+	}
 
-static struct nfs_server *nfs4_referral_server(struct super_block *sb, struct nfs_clone_mount *data)
-{
-	struct nfs_server *server = NFS_SB(sb);
-	int proto, timeo, retrans;
-	void *err;
-
-	proto = IPPROTO_TCP;
-	/* Since we are following a referral and there may be alternatives,
-	   set the timeouts and retries to low values */
-	timeo = 2;
-	retrans = 1;
-
-	nfs_put_client(server->nfs_client);
-	server->nfs_client = NULL;
-	server->client = nfs4_create_client(server, timeo, retrans, proto,
-					    data->authflavor);
-	if (IS_ERR((err = server->client)))
-		goto out_err;
+	mntroot = nfs4_get_root(s, data->fh);
+	if (IS_ERR(mntroot)) {
+		error = PTR_ERR(mntroot);
+		goto error_splat_super;
+	}
 
-	sb->s_time_gran = 1;
-	sb->s_op = &nfs4_sops;
-	err = ERR_PTR(nfs_sb_init(sb, data->authflavor));
-	if (!IS_ERR(err))
-		return server;
-out_err:
-	return (struct nfs_server *)err;
+	s->s_flags |= MS_ACTIVE;
+	mnt->mnt_sb = s;
+	mnt->mnt_root = mntroot;
+
+	dprintk("<-- nfs4_xdev_get_sb() = 0\n");
+	return 0;
+
+out_err_nosb:
+	nfs_free_server(server);
+out_err_noserver:
+	dprintk("<-- nfs4_xdev_get_sb() = %d [error]\n", error);
+	return error;
+
+error_splat_super:
+	up_write(&s->s_umount);
+	deactivate_super(s);
+	dprintk("<-- nfs4_xdev_get_sb() = %d [splat]\n", error);
+	return error;
 }
 
-static int nfs_referral_nfs4_sb(struct file_system_type *fs_type,
-		int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt)
+/*
+ * Create an NFS4 server record on referral traversal
+ */
+static int nfs4_referral_get_sb(struct file_system_type *fs_type, int flags,
+				const char *dev_name, void *raw_data,
+				struct vfsmount *mnt)
 {
 	struct nfs_clone_mount *data = raw_data;
-	return nfs_clone_generic_sb(data, nfs4_referral_sb, nfs4_referral_server, mnt);
+	struct super_block *s;
+	struct nfs_server *server;
+	struct dentry *mntroot;
+	struct nfs_fh mntfh;
+	int error;
+
+	dprintk("--> nfs4_referral_get_sb()\n");
+
+	/* create a new volume representation */
+	server = nfs4_create_referral_server(data, &mntfh);
+	if (IS_ERR(server)) {
+		error = PTR_ERR(server);
+		goto out_err_noserver;
+	}
+
+	/* Get a superblock - note that we may end up sharing one that already exists */
+	s = sget(&nfs_fs_type, nfs_compare_super, nfs_set_super, server);
+	if (IS_ERR(s)) {
+		error = PTR_ERR(s);
+		goto out_err_nosb;
+	}
+
+	if (s->s_fs_info != server) {
+		nfs_free_server(server);
+		server = NULL;
+	}
+
+	if (!s->s_root) {
+		/* initial superblock/root creation */
+		s->s_flags = flags;
+		nfs4_fill_super(s);
+	}
+
+	mntroot = nfs4_get_root(s, data->fh);
+	if (IS_ERR(mntroot)) {
+		error = PTR_ERR(mntroot);
+		goto error_splat_super;
+	}
+
+	s->s_flags |= MS_ACTIVE;
+	mnt->mnt_sb = s;
+	mnt->mnt_root = mntroot;
+
+	dprintk("<-- nfs4_referral_get_sb() = 0\n");
+	return 0;
+
+out_err_nosb:
+	nfs_free_server(server);
+out_err_noserver:
+	dprintk("<-- nfs4_referral_get_sb() = %d [error]\n", error);
+	return error;
+
+error_splat_super:
+	up_write(&s->s_umount);
+	deactivate_super(s);
+	dprintk("<-- nfs4_referral_get_sb() = %d [splat]\n", error);
+	return error;
 }
 
-#endif
+#endif /* CONFIG_NFS_V4 */
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index 7084ac9a6455..453d44666ea5 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -1273,7 +1273,7 @@ int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data)
 		if (time_before(complain, jiffies)) {
 			dprintk("NFS: faulty NFS server %s:"
 				" (committed = %d) != (stable = %d)\n",
-				NFS_SERVER(data->inode)->hostname,
+				NFS_SERVER(data->inode)->nfs_client->cl_hostname,
 				resp->verf->committed, argp->stable);
 			complain = jiffies + 300 * HZ;
 		}
diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index d404ceca9168..6d0be0efd1b5 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -51,7 +51,6 @@ struct nfs_client {
 	unsigned long		cl_lease_time;
 	unsigned long		cl_last_renewal;
 	struct work_struct	cl_renewd;
-	struct work_struct	cl_recoverd;
 
 	struct rpc_wait_queue	cl_rpcwaitq;
 
@@ -74,6 +73,10 @@ struct nfs_client {
  */
 struct nfs_server {
 	struct nfs_client *	nfs_client;	/* shared client and NFS4 state */
+	struct list_head	client_link;	/* List of other nfs_server structs
+						 * that share the same client
+						 */
+	struct list_head	master_link;	/* link in master servers list */
 	struct rpc_clnt *	client;		/* RPC client handle */
 	struct rpc_clnt *	client_acl;	/* ACL RPC client handle */
 	struct nfs_iostats *	io_stats;	/* I/O statistics */
@@ -92,20 +95,13 @@ struct nfs_server {
 	unsigned int		acdirmin;
 	unsigned int		acdirmax;
 	unsigned int		namelen;
-	char *			hostname;	/* remote hostname */
-	struct nfs_fh		fh;
-	struct sockaddr_in	addr;
+
 	struct nfs_fsid		fsid;
+	__u64			maxfilesize;	/* maximum file size */
 	unsigned long		mount_time;	/* when this fs was mounted */
+	dev_t			s_dev;		/* superblock dev numbers */
+
 #ifdef CONFIG_NFS_V4
-	/* Our own IP address, as a null-terminated string.
-	 * This is used to generate the clientid, and the callback address.
-	 */
-	char			ip_addr[16];
-	char *			mnt_path;
-	struct list_head	nfs4_siblings;	/* List of other nfs_server structs
-						 * that share the same clientid
-						 */
 	u32			attr_bitmask[2];/* V4 bitmask representing the set
 						   of attributes supported on this
 						   filesystem */
@@ -113,6 +109,7 @@ struct nfs_server {
 						   that are supported on this
 						   filesystem */
 #endif
+	void (*destroy)(struct nfs_server *);
 };
 
 /* Server capabilities */
-- 
cgit v1.2.3


From ec739ef03dc926d05051c8c5838971445504470a Mon Sep 17 00:00:00 2001
From: Chuck Lever <chuck.lever@oracle.com>
Date: Tue, 22 Aug 2006 20:06:15 -0400
Subject: SUNRPC: Create a helper to tell whether a transport is bound

Hide the contents and format of xprt->addr by eliminating direct uses
of the xprt->addr.sin_port field.  This change is required to support
alternate RPC host address formats (eg IPv6).

Test-plan:
Destructive testing (unplugging the network temporarily).  Repeated runs of
Connectathon locking suite with UDP and TCP.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 include/linux/sunrpc/xprt.h | 16 ++++++++++++++++
 net/sunrpc/clnt.c           | 10 +++++-----
 net/sunrpc/pmap_clnt.c      |  5 ++++-
 net/sunrpc/xprt.c           |  2 +-
 net/sunrpc/xprtsock.c       | 14 ++++++++++----
 5 files changed, 36 insertions(+), 11 deletions(-)

(limited to 'include')

diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h
index 3a0cca255b76..a71106723d71 100644
--- a/include/linux/sunrpc/xprt.h
+++ b/include/linux/sunrpc/xprt.h
@@ -269,6 +269,7 @@ int			xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to);
 #define XPRT_CONNECTED		(1)
 #define XPRT_CONNECTING		(2)
 #define XPRT_CLOSE_WAIT		(3)
+#define XPRT_BOUND		(4)
 
 static inline void xprt_set_connected(struct rpc_xprt *xprt)
 {
@@ -312,6 +313,21 @@ static inline int xprt_test_and_set_connecting(struct rpc_xprt *xprt)
 	return test_and_set_bit(XPRT_CONNECTING, &xprt->state);
 }
 
+static inline void xprt_set_bound(struct rpc_xprt *xprt)
+{
+	test_and_set_bit(XPRT_BOUND, &xprt->state);
+}
+
+static inline int xprt_bound(struct rpc_xprt *xprt)
+{
+	return test_bit(XPRT_BOUND, &xprt->state);
+}
+
+static inline void xprt_clear_bound(struct rpc_xprt *xprt)
+{
+	clear_bit(XPRT_BOUND, &xprt->state);
+}
+
 #endif /* __KERNEL__*/
 
 #endif /* _LINUX_SUNRPC_XPRT_H */
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index 3e19d321067a..0b8d03d08561 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -148,7 +148,6 @@ rpc_new_client(struct rpc_xprt *xprt, char *servname,
 	clnt->cl_maxproc  = version->nrprocs;
 	clnt->cl_protname = program->name;
 	clnt->cl_pmap	  = &clnt->cl_pmap_default;
-	clnt->cl_port     = xprt->addr.sin_port;
 	clnt->cl_prog     = program->number;
 	clnt->cl_vers     = version->number;
 	clnt->cl_prot     = xprt->prot;
@@ -156,7 +155,7 @@ rpc_new_client(struct rpc_xprt *xprt, char *servname,
 	clnt->cl_metrics  = rpc_alloc_iostats(clnt);
 	rpc_init_wait_queue(&clnt->cl_pmap_default.pm_bindwait, "bindwait");
 
-	if (!clnt->cl_port)
+	if (!xprt_bound(clnt->cl_xprt))
 		clnt->cl_autobind = 1;
 
 	clnt->cl_rtt = &clnt->cl_rtt_default;
@@ -570,7 +569,7 @@ EXPORT_SYMBOL(rpc_max_payload);
 void rpc_force_rebind(struct rpc_clnt *clnt)
 {
 	if (clnt->cl_autobind)
-		clnt->cl_port = 0;
+		xprt_clear_bound(clnt->cl_xprt);
 }
 EXPORT_SYMBOL(rpc_force_rebind);
 
@@ -782,14 +781,15 @@ static void
 call_bind(struct rpc_task *task)
 {
 	struct rpc_clnt	*clnt = task->tk_client;
+	struct rpc_xprt *xprt = task->tk_xprt;
 
 	dprintk("RPC: %4d call_bind (status %d)\n",
 				task->tk_pid, task->tk_status);
 
 	task->tk_action = call_connect;
-	if (!clnt->cl_port) {
+	if (!xprt_bound(xprt)) {
 		task->tk_action = call_bind_status;
-		task->tk_timeout = task->tk_xprt->bind_timeout;
+		task->tk_timeout = xprt->bind_timeout;
 		rpc_getport(task, clnt);
 	}
 }
diff --git a/net/sunrpc/pmap_clnt.c b/net/sunrpc/pmap_clnt.c
index 623180f224c9..209ffdfee10b 100644
--- a/net/sunrpc/pmap_clnt.c
+++ b/net/sunrpc/pmap_clnt.c
@@ -142,15 +142,17 @@ pmap_getport_done(struct rpc_task *task)
 	dprintk("RPC: %4d pmap_getport_done(status %d, port %d)\n",
 			task->tk_pid, task->tk_status, clnt->cl_port);
 
-	xprt->ops->set_port(xprt, 0);
 	if (task->tk_status < 0) {
 		/* Make the calling task exit with an error */
+		xprt->ops->set_port(xprt, 0);
 		task->tk_action = rpc_exit_task;
 	} else if (clnt->cl_port == 0) {
 		/* Program not registered */
+		xprt->ops->set_port(xprt, 0);
 		rpc_exit(task, -EACCES);
 	} else {
 		xprt->ops->set_port(xprt, clnt->cl_port);
+		xprt_set_bound(xprt);
 		clnt->cl_port = htons(clnt->cl_port);
 	}
 	spin_lock(&pmap_lock);
@@ -218,6 +220,7 @@ pmap_create(char *hostname, struct sockaddr_in *srvaddr, int proto, int privileg
 	if (IS_ERR(xprt))
 		return (struct rpc_clnt *)xprt;
 	xprt->ops->set_port(xprt, RPC_PMAP_PORT);
+	xprt_set_bound(xprt);
 	if (!privileged)
 		xprt->resvport = 0;
 
diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index e8c2bc4977f3..e239ef985ef7 100644
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -534,7 +534,7 @@ void xprt_connect(struct rpc_task *task)
 	dprintk("RPC: %4d xprt_connect xprt %p %s connected\n", task->tk_pid,
 			xprt, (xprt_connected(xprt) ? "is" : "is not"));
 
-	if (!xprt->addr.sin_port) {
+	if (!xprt_bound(xprt)) {
 		task->tk_status = -EIO;
 		return;
 	}
diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c
index 441bd53f5eca..123ac1e5ba15 100644
--- a/net/sunrpc/xprtsock.c
+++ b/net/sunrpc/xprtsock.c
@@ -1016,7 +1016,7 @@ static void xs_udp_connect_worker(void *args)
 	struct socket *sock = xprt->sock;
 	int err, status = -EIO;
 
-	if (xprt->shutdown || xprt->addr.sin_port == 0)
+	if (xprt->shutdown || !xprt_bound(xprt))
 		goto out;
 
 	dprintk("RPC:      xs_udp_connect_worker for xprt %p\n", xprt);
@@ -1099,7 +1099,7 @@ static void xs_tcp_connect_worker(void *args)
 	struct socket *sock = xprt->sock;
 	int err, status = -EIO;
 
-	if (xprt->shutdown || xprt->addr.sin_port == 0)
+	if (xprt->shutdown || !xprt_bound(xprt))
 		goto out;
 
 	dprintk("RPC:      xs_tcp_connect_worker for xprt %p\n", xprt);
@@ -1307,8 +1307,11 @@ int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to)
 	if (xprt->slot == NULL)
 		return -ENOMEM;
 
-	xprt->prot = IPPROTO_UDP;
+	if (ntohs(xprt->addr.sin_port) != 0)
+		xprt_set_bound(xprt);
 	xprt->port = xs_get_random_port();
+
+	xprt->prot = IPPROTO_UDP;
 	xprt->tsh_size = 0;
 	xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0;
 	/* XXX: header size can vary due to auth type, IPv6, etc. */
@@ -1348,8 +1351,11 @@ int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to)
 	if (xprt->slot == NULL)
 		return -ENOMEM;
 
-	xprt->prot = IPPROTO_TCP;
+	if (ntohs(xprt->addr.sin_port) != 0)
+		xprt_set_bound(xprt);
 	xprt->port = xs_get_random_port();
+
+	xprt->prot = IPPROTO_TCP;
 	xprt->tsh_size = sizeof(rpc_fraghdr) / sizeof(u32);
 	xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0;
 	xprt->max_payload = RPC_MAX_FRAGMENT_SIZE;
-- 
cgit v1.2.3


From 4a68179d38874c37be2802442a71b847f5d1a2a9 Mon Sep 17 00:00:00 2001
From: Chuck Lever <chuck.lever@oracle.com>
Date: Tue, 22 Aug 2006 20:06:15 -0400
Subject: SUNRPC: Make RPC portmapper use per-transport storage

Move connection and bind state that was maintained in the rpc_clnt
structure to the rpc_xprt structure.  This will allow the creation of
a clean API for plugging in different types of bind mechanisms.

This brings improvements such as the elimination of a single spin lock to
control serialization for all in-kernel RPC binding.  A set of per-xprt
bitops is used to serialize tasks during RPC binding, just like it now
works for making RPC transport connections.

Test-plan:
Destructive testing (unplugging the network temporarily).  Connectathon
with UDP and TCP.  NFSv2/3 and NFSv4 mounting should be carefully checked.
Probably need to rig a server where certain services aren't running, or
that returns an error for some typical operation.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 include/linux/sunrpc/clnt.h |  23 +------
 include/linux/sunrpc/xprt.h |  14 ++++
 net/sunrpc/clnt.c           |   8 +--
 net/sunrpc/pmap_clnt.c      | 158 +++++++++++++++++++++++++++++---------------
 net/sunrpc/xprt.c           |   1 +
 5 files changed, 123 insertions(+), 81 deletions(-)

(limited to 'include')

diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h
index 8fe9f35eba31..00e9dbaec9c5 100644
--- a/include/linux/sunrpc/clnt.h
+++ b/include/linux/sunrpc/clnt.h
@@ -18,18 +18,6 @@
 #include <linux/sunrpc/timer.h>
 #include <asm/signal.h>
 
-/*
- * This defines an RPC port mapping
- */
-struct rpc_portmap {
-	__u32			pm_prog;
-	__u32			pm_vers;
-	__u32			pm_prot;
-	__u16			pm_port;
-	unsigned char		pm_binding : 1;	/* doing a getport() */
-	struct rpc_wait_queue	pm_bindwait;	/* waiting on getport() */
-};
-
 struct rpc_inode;
 
 /*
@@ -40,7 +28,9 @@ struct rpc_clnt {
 	atomic_t		cl_users;	/* number of references */
 	struct rpc_xprt *	cl_xprt;	/* transport */
 	struct rpc_procinfo *	cl_procinfo;	/* procedure info */
-	u32			cl_maxproc;	/* max procedure number */
+	u32			cl_prog,	/* RPC program number */
+				cl_vers,	/* RPC version number */
+				cl_maxproc;	/* max procedure number */
 
 	char *			cl_server;	/* server machine name */
 	char *			cl_protname;	/* protocol name */
@@ -55,7 +45,6 @@ struct rpc_clnt {
 				cl_dead     : 1;/* abandoned */
 
 	struct rpc_rtt *	cl_rtt;		/* RTO estimator data */
-	struct rpc_portmap *	cl_pmap;	/* port mapping */
 
 	int			cl_nodelen;	/* nodename length */
 	char 			cl_nodename[UNX_MAXNODENAME];
@@ -64,14 +53,8 @@ struct rpc_clnt {
 	struct dentry *		cl_dentry;	/* inode */
 	struct rpc_clnt *	cl_parent;	/* Points to parent of clones */
 	struct rpc_rtt		cl_rtt_default;
-	struct rpc_portmap	cl_pmap_default;
 	char			cl_inline_name[32];
 };
-#define cl_timeout		cl_xprt->timeout
-#define cl_prog			cl_pmap->pm_prog
-#define cl_vers			cl_pmap->pm_vers
-#define cl_port			cl_pmap->pm_port
-#define cl_prot			cl_pmap->pm_prot
 
 /*
  * General RPC program info
diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h
index a71106723d71..4ce82616873d 100644
--- a/include/linux/sunrpc/xprt.h
+++ b/include/linux/sunrpc/xprt.h
@@ -138,6 +138,7 @@ struct rpc_xprt {
 	unsigned int		tsh_size;	/* size of transport specific
 						   header */
 
+	struct rpc_wait_queue	binding;	/* requests waiting on rpcbind */
 	struct rpc_wait_queue	sending;	/* requests waiting to send */
 	struct rpc_wait_queue	resend;		/* requests waiting to resend */
 	struct rpc_wait_queue	pending;	/* requests in flight */
@@ -270,6 +271,7 @@ int			xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to);
 #define XPRT_CONNECTING		(2)
 #define XPRT_CLOSE_WAIT		(3)
 #define XPRT_BOUND		(4)
+#define XPRT_BINDING		(5)
 
 static inline void xprt_set_connected(struct rpc_xprt *xprt)
 {
@@ -328,6 +330,18 @@ static inline void xprt_clear_bound(struct rpc_xprt *xprt)
 	clear_bit(XPRT_BOUND, &xprt->state);
 }
 
+static inline void xprt_clear_binding(struct rpc_xprt *xprt)
+{
+	smp_mb__before_clear_bit();
+	clear_bit(XPRT_BINDING, &xprt->state);
+	smp_mb__after_clear_bit();
+}
+
+static inline int xprt_test_and_set_binding(struct rpc_xprt *xprt)
+{
+	return test_and_set_bit(XPRT_BINDING, &xprt->state);
+}
+
 #endif /* __KERNEL__*/
 
 #endif /* _LINUX_SUNRPC_XPRT_H */
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index 0b8d03d08561..cee504162a3f 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -147,13 +147,10 @@ rpc_new_client(struct rpc_xprt *xprt, char *servname,
 	clnt->cl_procinfo = version->procs;
 	clnt->cl_maxproc  = version->nrprocs;
 	clnt->cl_protname = program->name;
-	clnt->cl_pmap	  = &clnt->cl_pmap_default;
 	clnt->cl_prog     = program->number;
 	clnt->cl_vers     = version->number;
-	clnt->cl_prot     = xprt->prot;
 	clnt->cl_stats    = program->stats;
 	clnt->cl_metrics  = rpc_alloc_iostats(clnt);
-	rpc_init_wait_queue(&clnt->cl_pmap_default.pm_bindwait, "bindwait");
 
 	if (!xprt_bound(clnt->cl_xprt))
 		clnt->cl_autobind = 1;
@@ -243,8 +240,6 @@ rpc_clone_client(struct rpc_clnt *clnt)
 	atomic_set(&new->cl_users, 0);
 	new->cl_parent = clnt;
 	atomic_inc(&clnt->cl_count);
-	/* Duplicate portmapper */
-	rpc_init_wait_queue(&new->cl_pmap_default.pm_bindwait, "bindwait");
 	/* Turn off autobind on clones */
 	new->cl_autobind = 0;
 	new->cl_oneshot = 0;
@@ -254,8 +249,7 @@ rpc_clone_client(struct rpc_clnt *clnt)
 	rpc_init_rtt(&new->cl_rtt_default, clnt->cl_xprt->timeout.to_initval);
 	if (new->cl_auth)
 		atomic_inc(&new->cl_auth->au_count);
-	new->cl_pmap		= &new->cl_pmap_default;
-	new->cl_metrics         = rpc_alloc_iostats(clnt);
+	new->cl_metrics = rpc_alloc_iostats(clnt);
 	return new;
 out_no_clnt:
 	printk(KERN_INFO "RPC: out of memory in %s\n", __FUNCTION__);
diff --git a/net/sunrpc/pmap_clnt.c b/net/sunrpc/pmap_clnt.c
index 209ffdfee10b..59d542436ca9 100644
--- a/net/sunrpc/pmap_clnt.c
+++ b/net/sunrpc/pmap_clnt.c
@@ -24,11 +24,57 @@
 #define PMAP_UNSET		2
 #define PMAP_GETPORT		3
 
+struct portmap_args {
+	u32			pm_prog;
+	u32			pm_vers;
+	u32			pm_prot;
+	unsigned short		pm_port;
+	struct rpc_task *	pm_task;
+};
+
 static struct rpc_procinfo	pmap_procedures[];
 static struct rpc_clnt *	pmap_create(char *, struct sockaddr_in *, int, int);
-static void			pmap_getport_done(struct rpc_task *);
+static void			pmap_getport_done(struct rpc_task *, void *);
 static struct rpc_program	pmap_program;
-static DEFINE_SPINLOCK(pmap_lock);
+
+static void pmap_getport_prepare(struct rpc_task *task, void *calldata)
+{
+	struct portmap_args *map = calldata;
+	struct rpc_message msg = {
+		.rpc_proc	= &pmap_procedures[PMAP_GETPORT],
+		.rpc_argp	= map,
+		.rpc_resp	= &map->pm_port,
+	};
+
+	rpc_call_setup(task, &msg, 0);
+}
+
+static inline struct portmap_args *pmap_map_alloc(void)
+{
+	return kmalloc(sizeof(struct portmap_args), GFP_NOFS);
+}
+
+static inline void pmap_map_free(struct portmap_args *map)
+{
+	kfree(map);
+}
+
+static void pmap_map_release(void *data)
+{
+	pmap_map_free(data);
+}
+
+static const struct rpc_call_ops pmap_getport_ops = {
+	.rpc_call_prepare	= pmap_getport_prepare,
+	.rpc_call_done		= pmap_getport_done,
+	.rpc_release		= pmap_map_release,
+};
+
+static inline void pmap_wake_portmap_waiters(struct rpc_xprt *xprt)
+{
+	xprt_clear_binding(xprt);
+	rpc_wake_up(&xprt->binding);
+}
 
 /*
  * Obtain the port for a given RPC service on a given host. This one can
@@ -37,67 +83,71 @@ static DEFINE_SPINLOCK(pmap_lock);
 void
 rpc_getport(struct rpc_task *task, struct rpc_clnt *clnt)
 {
-	struct rpc_portmap *map = clnt->cl_pmap;
-	struct sockaddr_in *sap = &clnt->cl_xprt->addr;
-	struct rpc_message msg = {
-		.rpc_proc	= &pmap_procedures[PMAP_GETPORT],
-		.rpc_argp	= map,
-		.rpc_resp	= &clnt->cl_port,
-		.rpc_cred	= NULL
-	};
+	struct rpc_xprt *xprt = task->tk_xprt;
+	struct sockaddr_in *sap = &xprt->addr;
+	struct portmap_args *map;
 	struct rpc_clnt	*pmap_clnt;
-	struct rpc_task	*child;
+	struct rpc_task *child;
 
-	dprintk("RPC: %4d rpc_getport(%s, %d, %d, %d)\n",
+	dprintk("RPC: %4d rpc_getport(%s, %u, %u, %d)\n",
 			task->tk_pid, clnt->cl_server,
-			map->pm_prog, map->pm_vers, map->pm_prot);
+			clnt->cl_prog, clnt->cl_vers, xprt->prot);
 
 	/* Autobind on cloned rpc clients is discouraged */
 	BUG_ON(clnt->cl_parent != clnt);
 
-	spin_lock(&pmap_lock);
-	if (map->pm_binding) {
-		rpc_sleep_on(&map->pm_bindwait, task, NULL, NULL);
-		spin_unlock(&pmap_lock);
+	if (xprt_test_and_set_binding(xprt)) {
+		task->tk_status = -EACCES;	/* tell caller to check again */
+		rpc_sleep_on(&xprt->binding, task, NULL, NULL);
 		return;
 	}
-	map->pm_binding = 1;
-	spin_unlock(&pmap_lock);
+
+	/* Someone else may have bound if we slept */
+	if (xprt_bound(xprt)) {
+		task->tk_status = 0;
+		goto bailout_nofree;
+	}
+
+	map = pmap_map_alloc();
+	if (!map) {
+		task->tk_status = -ENOMEM;
+		goto bailout_nofree;
+	}
+	map->pm_prog = clnt->cl_prog;
+	map->pm_vers = clnt->cl_vers;
+	map->pm_prot = xprt->prot;
+	map->pm_port = 0;
+	map->pm_task = task;
 
 	pmap_clnt = pmap_create(clnt->cl_server, sap, map->pm_prot, 0);
 	if (IS_ERR(pmap_clnt)) {
 		task->tk_status = PTR_ERR(pmap_clnt);
 		goto bailout;
 	}
-	task->tk_status = 0;
 
-	/*
-	 * Note: rpc_new_child will release client after a failure.
-	 */
-	if (!(child = rpc_new_child(pmap_clnt, task)))
+	child = rpc_run_task(pmap_clnt, RPC_TASK_ASYNC, &pmap_getport_ops, map);
+	if (IS_ERR(child)) {
+		task->tk_status = -EIO;
 		goto bailout;
+	}
+	rpc_release_task(child);
 
-	/* Setup the call info struct */
-	rpc_call_setup(child, &msg, 0);
+	rpc_sleep_on(&xprt->binding, task, NULL, NULL);
 
-	/* ... and run the child task */
 	task->tk_xprt->stat.bind_count++;
-	rpc_run_child(task, child, pmap_getport_done);
 	return;
 
 bailout:
-	spin_lock(&pmap_lock);
-	map->pm_binding = 0;
-	rpc_wake_up(&map->pm_bindwait);
-	spin_unlock(&pmap_lock);
-	rpc_exit(task, -EIO);
+	pmap_map_free(map);
+bailout_nofree:
+	pmap_wake_portmap_waiters(xprt);
 }
 
 #ifdef CONFIG_ROOT_NFS
 int
 rpc_getport_external(struct sockaddr_in *sin, __u32 prog, __u32 vers, int prot)
 {
-	struct rpc_portmap map = {
+	struct portmap_args map = {
 		.pm_prog	= prog,
 		.pm_vers	= vers,
 		.pm_prot	= prot,
@@ -133,32 +183,32 @@ rpc_getport_external(struct sockaddr_in *sin, __u32 prog, __u32 vers, int prot)
 #endif
 
 static void
-pmap_getport_done(struct rpc_task *task)
+pmap_getport_done(struct rpc_task *child, void *data)
 {
-	struct rpc_clnt	*clnt = task->tk_client;
+	struct portmap_args *map = data;
+	struct rpc_task *task = map->pm_task;
 	struct rpc_xprt *xprt = task->tk_xprt;
-	struct rpc_portmap *map = clnt->cl_pmap;
-
-	dprintk("RPC: %4d pmap_getport_done(status %d, port %d)\n",
-			task->tk_pid, task->tk_status, clnt->cl_port);
+	int status = child->tk_status;
 
-	if (task->tk_status < 0) {
-		/* Make the calling task exit with an error */
+	if (status < 0) {
+		/* Portmapper not available */
 		xprt->ops->set_port(xprt, 0);
-		task->tk_action = rpc_exit_task;
-	} else if (clnt->cl_port == 0) {
-		/* Program not registered */
+		task->tk_status = status;
+	} else if (map->pm_port == 0) {
+		/* Requested RPC service wasn't registered */
 		xprt->ops->set_port(xprt, 0);
-		rpc_exit(task, -EACCES);
+		task->tk_status = -EACCES;
 	} else {
-		xprt->ops->set_port(xprt, clnt->cl_port);
+		/* Succeeded */
+		xprt->ops->set_port(xprt, map->pm_port);
 		xprt_set_bound(xprt);
-		clnt->cl_port = htons(clnt->cl_port);
+		task->tk_status = 0;
 	}
-	spin_lock(&pmap_lock);
-	map->pm_binding = 0;
-	rpc_wake_up(&map->pm_bindwait);
-	spin_unlock(&pmap_lock);
+
+	dprintk("RPC: %4d pmap_getport_done(status %d, port %u)\n",
+			child->tk_pid, child->tk_status, map->pm_port);
+
+	pmap_wake_portmap_waiters(xprt);
 }
 
 /*
@@ -172,7 +222,7 @@ rpc_register(u32 prog, u32 vers, int prot, unsigned short port, int *okay)
 		.sin_family	= AF_INET,
 		.sin_addr.s_addr = htonl(INADDR_LOOPBACK),
 	};
-	struct rpc_portmap	map = {
+	struct portmap_args	map = {
 		.pm_prog	= prog,
 		.pm_vers	= vers,
 		.pm_prot	= prot,
@@ -239,7 +289,7 @@ pmap_create(char *hostname, struct sockaddr_in *srvaddr, int proto, int privileg
  * XDR encode/decode functions for PMAP
  */
 static int
-xdr_encode_mapping(struct rpc_rqst *req, u32 *p, struct rpc_portmap *map)
+xdr_encode_mapping(struct rpc_rqst *req, u32 *p, struct portmap_args *map)
 {
 	dprintk("RPC: xdr_encode_mapping(%d, %d, %d, %d)\n",
 		map->pm_prog, map->pm_vers, map->pm_prot, map->pm_port);
diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index e239ef985ef7..b45abd0743cb 100644
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -928,6 +928,7 @@ static struct rpc_xprt *xprt_setup(int proto, struct sockaddr_in *ap, struct rpc
 	xprt->last_used = jiffies;
 	xprt->cwnd = RPC_INITCWND;
 
+	rpc_init_wait_queue(&xprt->binding, "xprt_binding");
 	rpc_init_wait_queue(&xprt->pending, "xprt_pending");
 	rpc_init_wait_queue(&xprt->sending, "xprt_sending");
 	rpc_init_wait_queue(&xprt->resend, "xprt_resend");
-- 
cgit v1.2.3


From 5b1eacbcd78930d976eb50a93f1779d311b553d1 Mon Sep 17 00:00:00 2001
From: Chuck Lever <chuck.lever@oracle.com>
Date: Tue, 22 Aug 2006 20:06:16 -0400
Subject: SUNRPC: Support for RPC child tasks no longer needed

The previous patches removed the last user of RPC child tasks, so we can
remove support for child tasks from net/sunrpc/sched.c now.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 include/linux/sunrpc/sched.h |  5 ---
 net/sunrpc/sched.c           | 82 --------------------------------------------
 2 files changed, 87 deletions(-)

(limited to 'include')

diff --git a/include/linux/sunrpc/sched.h b/include/linux/sunrpc/sched.h
index 82a91bb22362..f399c138f79d 100644
--- a/include/linux/sunrpc/sched.h
+++ b/include/linux/sunrpc/sched.h
@@ -127,7 +127,6 @@ struct rpc_call_ops {
  */
 #define RPC_TASK_ASYNC		0x0001		/* is an async task */
 #define RPC_TASK_SWAPPER	0x0002		/* is swapping in/out */
-#define RPC_TASK_CHILD		0x0008		/* is child of other task */
 #define RPC_CALL_MAJORSEEN	0x0020		/* major timeout seen */
 #define RPC_TASK_ROOTCREDS	0x0040		/* force root creds */
 #define RPC_TASK_DYNAMIC	0x0080		/* task was kmalloc'ed */
@@ -136,7 +135,6 @@ struct rpc_call_ops {
 #define RPC_TASK_NOINTR		0x0400		/* uninterruptible task */
 
 #define RPC_IS_ASYNC(t)		((t)->tk_flags & RPC_TASK_ASYNC)
-#define RPC_IS_CHILD(t)		((t)->tk_flags & RPC_TASK_CHILD)
 #define RPC_IS_SWAPPER(t)	((t)->tk_flags & RPC_TASK_SWAPPER)
 #define RPC_DO_ROOTOVERRIDE(t)	((t)->tk_flags & RPC_TASK_ROOTCREDS)
 #define RPC_ASSASSINATED(t)	((t)->tk_flags & RPC_TASK_KILLED)
@@ -253,7 +251,6 @@ struct rpc_task *rpc_new_task(struct rpc_clnt *, int flags,
 				const struct rpc_call_ops *ops, void *data);
 struct rpc_task *rpc_run_task(struct rpc_clnt *clnt, int flags,
 				const struct rpc_call_ops *ops, void *data);
-struct rpc_task *rpc_new_child(struct rpc_clnt *, struct rpc_task *parent);
 void		rpc_init_task(struct rpc_task *task, struct rpc_clnt *clnt,
 				int flags, const struct rpc_call_ops *ops,
 				void *data);
@@ -261,8 +258,6 @@ void		rpc_release_task(struct rpc_task *);
 void		rpc_exit_task(struct rpc_task *);
 void		rpc_killall_tasks(struct rpc_clnt *);
 int		rpc_execute(struct rpc_task *);
-void		rpc_run_child(struct rpc_task *parent, struct rpc_task *child,
-					rpc_action action);
 void		rpc_init_priority_wait_queue(struct rpc_wait_queue *, const char *);
 void		rpc_init_wait_queue(struct rpc_wait_queue *, const char *);
 void		rpc_sleep_on(struct rpc_wait_queue *, struct rpc_task *,
diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c
index 5c3eee768504..015ffe423a2f 100644
--- a/net/sunrpc/sched.c
+++ b/net/sunrpc/sched.c
@@ -44,12 +44,6 @@ static void			__rpc_default_timer(struct rpc_task *task);
 static void			rpciod_killall(void);
 static void			rpc_async_schedule(void *);
 
-/*
- * RPC tasks that create another task (e.g. for contacting the portmapper)
- * will wait on this queue for their child's completion
- */
-static RPC_WAITQ(childq, "childq");
-
 /*
  * RPC tasks sit here while waiting for conditions to improve.
  */
@@ -323,16 +317,6 @@ static void rpc_make_runnable(struct rpc_task *task)
 		wake_up_bit(&task->tk_runstate, RPC_TASK_QUEUED);
 }
 
-/*
- * Place a newly initialized task on the workqueue.
- */
-static inline void
-rpc_schedule_run(struct rpc_task *task)
-{
-	rpc_set_active(task);
-	rpc_make_runnable(task);
-}
-
 /*
  * Prepare for sleeping on a wait queue.
  * By always appending tasks to the list we ensure FIFO behavior.
@@ -933,72 +917,6 @@ struct rpc_task *rpc_run_task(struct rpc_clnt *clnt, int flags,
 }
 EXPORT_SYMBOL(rpc_run_task);
 
-/**
- * rpc_find_parent - find the parent of a child task.
- * @child: child task
- * @parent: parent task
- *
- * Checks that the parent task is still sleeping on the
- * queue 'childq'. If so returns a pointer to the parent.
- * Upon failure returns NULL.
- *
- * Caller must hold childq.lock
- */
-static inline struct rpc_task *rpc_find_parent(struct rpc_task *child, struct rpc_task *parent)
-{
-	struct rpc_task	*task;
-	struct list_head *le;
-
-	task_for_each(task, le, &childq.tasks[0])
-		if (task == parent)
-			return parent;
-
-	return NULL;
-}
-
-static void rpc_child_exit(struct rpc_task *child, void *calldata)
-{
-	struct rpc_task	*parent;
-
-	spin_lock_bh(&childq.lock);
-	if ((parent = rpc_find_parent(child, calldata)) != NULL) {
-		parent->tk_status = child->tk_status;
-		__rpc_wake_up_task(parent);
-	}
-	spin_unlock_bh(&childq.lock);
-}
-
-static const struct rpc_call_ops rpc_child_ops = {
-	.rpc_call_done = rpc_child_exit,
-};
-
-/*
- * Note: rpc_new_task releases the client after a failure.
- */
-struct rpc_task *
-rpc_new_child(struct rpc_clnt *clnt, struct rpc_task *parent)
-{
-	struct rpc_task	*task;
-
-	task = rpc_new_task(clnt, RPC_TASK_ASYNC | RPC_TASK_CHILD, &rpc_child_ops, parent);
-	if (!task)
-		goto fail;
-	return task;
-
-fail:
-	parent->tk_status = -ENOMEM;
-	return NULL;
-}
-
-void rpc_run_child(struct rpc_task *task, struct rpc_task *child, rpc_action func)
-{
-	spin_lock_bh(&childq.lock);
-	/* N.B. Is it possible for the child to have already finished? */
-	__rpc_sleep_on(&childq, task, func, NULL);
-	rpc_schedule_run(child);
-	spin_unlock_bh(&childq.lock);
-}
-
 /*
  * Kill all tasks for the given client.
  * XXX: kill their descendants as well?
-- 
cgit v1.2.3


From bbf7c1dd2ae2b4040b41b1065ee9b1b6905b1605 Mon Sep 17 00:00:00 2001
From: Chuck Lever <chuck.lever@oracle.com>
Date: Tue, 22 Aug 2006 20:06:16 -0400
Subject: SUNRPC: Introduce transport switch callout for pluggable rpcbind

Introduce a clean transport switch API for plugging in different types of
rpcbind mechanisms.  For instance, rpcbind can cleanly replace the
existing portmapper client, or a transport can choose to implement RPC
binding any way it likes.

Test plan:
Destructive testing (unplugging the network temporarily).  Connectathon
with UDP and TCP.  NFSv2/3 and NFSv4 mounting should be carefully checked.
Probably need to rig a server where certain services aren't running, or
that returns an error for some typical operation.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 include/linux/sunrpc/clnt.h | 2 +-
 include/linux/sunrpc/xprt.h | 1 +
 net/sunrpc/clnt.c           | 3 +--
 net/sunrpc/pmap_clnt.c      | 4 ++--
 net/sunrpc/xprtsock.c       | 2 ++
 5 files changed, 7 insertions(+), 5 deletions(-)

(limited to 'include')

diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h
index 00e9dbaec9c5..2e68ac0aa022 100644
--- a/include/linux/sunrpc/clnt.h
+++ b/include/linux/sunrpc/clnt.h
@@ -106,7 +106,7 @@ struct rpc_clnt *rpc_clone_client(struct rpc_clnt *);
 int		rpc_shutdown_client(struct rpc_clnt *);
 int		rpc_destroy_client(struct rpc_clnt *);
 void		rpc_release_client(struct rpc_clnt *);
-void		rpc_getport(struct rpc_task *, struct rpc_clnt *);
+void		rpc_getport(struct rpc_task *);
 int		rpc_register(u32, u32, int, unsigned short, int *);
 
 void		rpc_call_setup(struct rpc_task *, struct rpc_message *, int);
diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h
index 4ce82616873d..84122559fa17 100644
--- a/include/linux/sunrpc/xprt.h
+++ b/include/linux/sunrpc/xprt.h
@@ -105,6 +105,7 @@ struct rpc_xprt_ops {
 	void		(*set_buffer_size)(struct rpc_xprt *xprt, size_t sndsize, size_t rcvsize);
 	int		(*reserve_xprt)(struct rpc_task *task);
 	void		(*release_xprt)(struct rpc_xprt *xprt, struct rpc_task *task);
+	void		(*rpcbind)(struct rpc_task *task);
 	void		(*set_port)(struct rpc_xprt *xprt, unsigned short port);
 	void		(*connect)(struct rpc_task *task);
 	void *		(*buf_alloc)(struct rpc_task *task, size_t size);
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index cee504162a3f..d003c2f5688f 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -774,7 +774,6 @@ call_encode(struct rpc_task *task)
 static void
 call_bind(struct rpc_task *task)
 {
-	struct rpc_clnt	*clnt = task->tk_client;
 	struct rpc_xprt *xprt = task->tk_xprt;
 
 	dprintk("RPC: %4d call_bind (status %d)\n",
@@ -784,7 +783,7 @@ call_bind(struct rpc_task *task)
 	if (!xprt_bound(xprt)) {
 		task->tk_action = call_bind_status;
 		task->tk_timeout = xprt->bind_timeout;
-		rpc_getport(task, clnt);
+		xprt->ops->rpcbind(task);
 	}
 }
 
diff --git a/net/sunrpc/pmap_clnt.c b/net/sunrpc/pmap_clnt.c
index 0efcbf1302a2..f7b279a63baa 100644
--- a/net/sunrpc/pmap_clnt.c
+++ b/net/sunrpc/pmap_clnt.c
@@ -81,13 +81,13 @@ static inline void pmap_wake_portmap_waiters(struct rpc_xprt *xprt)
 /**
  * rpc_getport - obtain the port for a given RPC service on a given host
  * @task: task that is waiting for portmapper request
- * @clnt: controlling rpc_clnt
  *
  * This one can be called for an ongoing RPC request, and can be used in
  * an async (rpciod) context.
  */
-void rpc_getport(struct rpc_task *task, struct rpc_clnt *clnt)
+void rpc_getport(struct rpc_task *task)
 {
+	struct rpc_clnt *clnt = task->tk_client;
 	struct rpc_xprt *xprt = task->tk_xprt;
 	struct sockaddr_in *sap = &xprt->addr;
 	struct portmap_args *map;
diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c
index 123ac1e5ba15..4c98b89a5b48 100644
--- a/net/sunrpc/xprtsock.c
+++ b/net/sunrpc/xprtsock.c
@@ -1262,6 +1262,7 @@ static struct rpc_xprt_ops xs_udp_ops = {
 	.set_buffer_size	= xs_udp_set_buffer_size,
 	.reserve_xprt		= xprt_reserve_xprt_cong,
 	.release_xprt		= xprt_release_xprt_cong,
+	.rpcbind		= rpc_getport,
 	.set_port		= xs_set_port,
 	.connect		= xs_connect,
 	.buf_alloc		= rpc_malloc,
@@ -1278,6 +1279,7 @@ static struct rpc_xprt_ops xs_udp_ops = {
 static struct rpc_xprt_ops xs_tcp_ops = {
 	.reserve_xprt		= xprt_reserve_xprt,
 	.release_xprt		= xs_tcp_release_xprt,
+	.rpcbind		= rpc_getport,
 	.set_port		= xs_set_port,
 	.connect		= xs_connect,
 	.buf_alloc		= rpc_malloc,
-- 
cgit v1.2.3


From ed39440a2573abc926f230267000f21fa5a87822 Mon Sep 17 00:00:00 2001
From: Chuck Lever <chuck.lever@oracle.com>
Date: Tue, 22 Aug 2006 20:06:17 -0400
Subject: SUNRPC: create API for getting remote peer address

Provide an API for retrieving the remote peer address without allowing
direct access to the rpc_xprt struct.

Test-plan:
Compile kernel with CONFIG_NFS enabled.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 include/linux/sunrpc/clnt.h |  1 +
 net/sunrpc/clnt.c           | 21 +++++++++++++++++++++
 2 files changed, 22 insertions(+)

(limited to 'include')

diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h
index 2e68ac0aa022..65196b03f0ab 100644
--- a/include/linux/sunrpc/clnt.h
+++ b/include/linux/sunrpc/clnt.h
@@ -123,6 +123,7 @@ void		rpc_setbufsize(struct rpc_clnt *, unsigned int, unsigned int);
 size_t		rpc_max_payload(struct rpc_clnt *);
 void		rpc_force_rebind(struct rpc_clnt *);
 int		rpc_ping(struct rpc_clnt *clnt, int flags);
+size_t		rpc_peeraddr(struct rpc_clnt *, struct sockaddr *, size_t);
 
 /*
  * Helper function for NFSroot support
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index d003c2f5688f..94768cf5fd5b 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -533,6 +533,27 @@ rpc_call_setup(struct rpc_task *task, struct rpc_message *msg, int flags)
 		task->tk_action = rpc_exit_task;
 }
 
+/**
+ * rpc_peeraddr - extract remote peer address from clnt's xprt
+ * @clnt: RPC client structure
+ * @buf: target buffer
+ * @size: length of target buffer
+ *
+ * Returns the number of bytes that are actually in the stored address.
+ */
+size_t rpc_peeraddr(struct rpc_clnt *clnt, struct sockaddr *buf, size_t bufsize)
+{
+	size_t bytes;
+	struct rpc_xprt *xprt = clnt->cl_xprt;
+
+	bytes = sizeof(xprt->addr);
+	if (bytes > bufsize)
+		bytes = bufsize;
+	memcpy(buf, &clnt->cl_xprt->addr, bytes);
+	return sizeof(xprt->addr);
+}
+EXPORT_SYMBOL(rpc_peeraddr);
+
 void
 rpc_setbufsize(struct rpc_clnt *clnt, unsigned int sndsize, unsigned int rcvsize)
 {
-- 
cgit v1.2.3


From 39d7bbcb5ba5e9d8d658b70903dd7939400e57db Mon Sep 17 00:00:00 2001
From: Chuck Lever <chuck.lever@oracle.com>
Date: Tue, 22 Aug 2006 20:06:18 -0400
Subject: SUNRPC: remove extraneous header inclusions

include/linux/sunrpc/clnt.h already includes include/linux/sunrpc/xprt.h.
We can remove xprt.h from source files that already include clnt.h.
Likewise include/linux/sunrpc/timer.h.

Test plan:
Compile kernel with CONFIG_NFS enabled.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/mount_clnt.c     | 1 -
 include/linux/nfs_xdr.h | 1 -
 net/sunrpc/pmap_clnt.c  | 1 -
 net/sunrpc/sched.c      | 1 -
 net/sunrpc/timer.c      | 2 --
 5 files changed, 6 deletions(-)

(limited to 'include')

diff --git a/fs/nfs/mount_clnt.c b/fs/nfs/mount_clnt.c
index 445abb4d4214..41274874b9a5 100644
--- a/fs/nfs/mount_clnt.c
+++ b/fs/nfs/mount_clnt.c
@@ -14,7 +14,6 @@
 #include <linux/net.h>
 #include <linux/in.h>
 #include <linux/sunrpc/clnt.h>
-#include <linux/sunrpc/xprt.h>
 #include <linux/sunrpc/sched.h>
 #include <linux/nfs_fs.h>
 
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index 2426b11b6cce..0f33e621892f 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -1,7 +1,6 @@
 #ifndef _LINUX_NFS_XDR_H
 #define _LINUX_NFS_XDR_H
 
-#include <linux/sunrpc/xprt.h>
 #include <linux/nfsacl.h>
 
 /*
diff --git a/net/sunrpc/pmap_clnt.c b/net/sunrpc/pmap_clnt.c
index 3eee8e907275..523f0e825dea 100644
--- a/net/sunrpc/pmap_clnt.c
+++ b/net/sunrpc/pmap_clnt.c
@@ -15,7 +15,6 @@
 #include <linux/uio.h>
 #include <linux/in.h>
 #include <linux/sunrpc/clnt.h>
-#include <linux/sunrpc/xprt.h>
 #include <linux/sunrpc/sched.h>
 
 #ifdef RPC_DEBUG
diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c
index 015ffe423a2f..ecf366351bf7 100644
--- a/net/sunrpc/sched.c
+++ b/net/sunrpc/sched.c
@@ -21,7 +21,6 @@
 #include <linux/mutex.h>
 
 #include <linux/sunrpc/clnt.h>
-#include <linux/sunrpc/xprt.h>
 
 #ifdef RPC_DEBUG
 #define RPCDBG_FACILITY		RPCDBG_SCHED
diff --git a/net/sunrpc/timer.c b/net/sunrpc/timer.c
index bcbdf6430d5c..8142fdb8a930 100644
--- a/net/sunrpc/timer.c
+++ b/net/sunrpc/timer.c
@@ -19,8 +19,6 @@
 #include <linux/unistd.h>
 
 #include <linux/sunrpc/clnt.h>
-#include <linux/sunrpc/xprt.h>
-#include <linux/sunrpc/timer.h>
 
 #define RPC_RTO_MAX (60*HZ)
 #define RPC_RTO_INIT (HZ/5)
-- 
cgit v1.2.3


From edb267a688fcee5335d596752f117a30c7152e44 Mon Sep 17 00:00:00 2001
From: Chuck Lever <chuck.lever@oracle.com>
Date: Tue, 22 Aug 2006 20:06:18 -0400
Subject: SUNRPC: add xprt switch API for printing formatted remote peer
 addresses

Add a new method to the transport switch API to provide a way to convert
the opaque contents of xprt->addr to a human-readable string.

Test plan:
Compile kernel with CONFIG_NFS enabled.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 include/linux/sunrpc/xprt.h | 11 +++++++
 net/sunrpc/xprtsock.c       | 79 ++++++++++++++++++++++++++++++++++++++++-----
 2 files changed, 82 insertions(+), 8 deletions(-)

(limited to 'include')

diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h
index 84122559fa17..8372ab8fc9b5 100644
--- a/include/linux/sunrpc/xprt.h
+++ b/include/linux/sunrpc/xprt.h
@@ -51,6 +51,14 @@ struct rpc_timeout {
 	unsigned char		to_exponential;
 };
 
+enum rpc_display_format_t {
+	RPC_DISPLAY_ADDR = 0,
+	RPC_DISPLAY_PORT,
+	RPC_DISPLAY_PROTO,
+	RPC_DISPLAY_ALL,
+	RPC_DISPLAY_MAX,
+};
+
 struct rpc_task;
 struct rpc_xprt;
 struct seq_file;
@@ -103,6 +111,7 @@ struct rpc_rqst {
 
 struct rpc_xprt_ops {
 	void		(*set_buffer_size)(struct rpc_xprt *xprt, size_t sndsize, size_t rcvsize);
+	char *		(*print_addr)(struct rpc_xprt *xprt, enum rpc_display_format_t format);
 	int		(*reserve_xprt)(struct rpc_task *task);
 	void		(*release_xprt)(struct rpc_xprt *xprt, struct rpc_task *task);
 	void		(*rpcbind)(struct rpc_task *task);
@@ -207,6 +216,8 @@ struct rpc_xprt {
 	void			(*old_data_ready)(struct sock *, int);
 	void			(*old_state_change)(struct sock *);
 	void			(*old_write_space)(struct sock *);
+
+	char *			address_strings[RPC_DISPLAY_MAX];
 };
 
 #define XPRT_LAST_FRAG		(1 << 0)
diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c
index 4c98b89a5b48..cb8e6c34e12f 100644
--- a/net/sunrpc/xprtsock.c
+++ b/net/sunrpc/xprtsock.c
@@ -125,6 +125,47 @@ static inline void xs_pktdump(char *msg, u32 *packet, unsigned int count)
 }
 #endif
 
+static void xs_format_peer_addresses(struct rpc_xprt *xprt)
+{
+	struct sockaddr_in *addr = (struct sockaddr_in *) &xprt->addr;
+	char *buf;
+
+	buf = kzalloc(20, GFP_KERNEL);
+	if (buf) {
+		snprintf(buf, 20, "%u.%u.%u.%u",
+				NIPQUAD(addr->sin_addr.s_addr));
+	}
+	xprt->address_strings[RPC_DISPLAY_ADDR] = buf;
+
+	buf = kzalloc(8, GFP_KERNEL);
+	if (buf) {
+		snprintf(buf, 8, "%u",
+				ntohs(addr->sin_port));
+	}
+	xprt->address_strings[RPC_DISPLAY_PORT] = buf;
+
+	if (xprt->prot == IPPROTO_UDP)
+		xprt->address_strings[RPC_DISPLAY_PROTO] = "udp";
+	else
+		xprt->address_strings[RPC_DISPLAY_PROTO] = "tcp";
+
+	buf = kzalloc(48, GFP_KERNEL);
+	if (buf) {
+		snprintf(buf, 48, "addr=%u.%u.%u.%u port=%u proto=%s",
+			NIPQUAD(addr->sin_addr.s_addr),
+			ntohs(addr->sin_port),
+			xprt->prot == IPPROTO_UDP ? "udp" : "tcp");
+	}
+	xprt->address_strings[RPC_DISPLAY_ALL] = buf;
+}
+
+static void xs_free_peer_addresses(struct rpc_xprt *xprt)
+{
+	kfree(xprt->address_strings[RPC_DISPLAY_ADDR]);
+	kfree(xprt->address_strings[RPC_DISPLAY_PORT]);
+	kfree(xprt->address_strings[RPC_DISPLAY_ALL]);
+}
+
 #define XS_SENDMSG_FLAGS	(MSG_DONTWAIT | MSG_NOSIGNAL)
 
 static inline int xs_send_head(struct socket *sock, struct sockaddr *addr, int addrlen, struct xdr_buf *xdr, unsigned int base, unsigned int len)
@@ -490,6 +531,7 @@ static void xs_destroy(struct rpc_xprt *xprt)
 
 	xprt_disconnect(xprt);
 	xs_close(xprt);
+	xs_free_peer_addresses(xprt);
 	kfree(xprt->slot);
 }
 
@@ -964,6 +1006,19 @@ static unsigned short xs_get_random_port(void)
 	return rand + xprt_min_resvport;
 }
 
+/**
+ * xs_print_peer_address - format an IPv4 address for printing
+ * @xprt: generic transport
+ * @format: flags field indicating which parts of the address to render
+ */
+static char *xs_print_peer_address(struct rpc_xprt *xprt, enum rpc_display_format_t format)
+{
+	if (xprt->address_strings[format] != NULL)
+		return xprt->address_strings[format];
+	else
+		return "unprintable";
+}
+
 /**
  * xs_set_port - reset the port number in the remote endpoint address
  * @xprt: generic transport
@@ -1019,8 +1074,6 @@ static void xs_udp_connect_worker(void *args)
 	if (xprt->shutdown || !xprt_bound(xprt))
 		goto out;
 
-	dprintk("RPC:      xs_udp_connect_worker for xprt %p\n", xprt);
-
 	/* Start by resetting any existing state */
 	xs_close(xprt);
 
@@ -1034,6 +1087,9 @@ static void xs_udp_connect_worker(void *args)
 		goto out;
 	}
 
+	dprintk("RPC:      worker connecting xprt %p to address: %s\n",
+			xprt, xs_print_peer_address(xprt, RPC_DISPLAY_ALL));
+
 	if (!xprt->inet) {
 		struct sock *sk = sock->sk;
 
@@ -1102,8 +1158,6 @@ static void xs_tcp_connect_worker(void *args)
 	if (xprt->shutdown || !xprt_bound(xprt))
 		goto out;
 
-	dprintk("RPC:      xs_tcp_connect_worker for xprt %p\n", xprt);
-
 	if (!xprt->sock) {
 		/* start from scratch */
 		if ((err = sock_create_kern(PF_INET, SOCK_STREAM, IPPROTO_TCP, &sock)) < 0) {
@@ -1119,6 +1173,9 @@ static void xs_tcp_connect_worker(void *args)
 		/* "close" the socket, preserving the local port */
 		xs_tcp_reuse_connection(xprt);
 
+	dprintk("RPC:      worker connecting xprt %p to address: %s\n",
+			xprt, xs_print_peer_address(xprt, RPC_DISPLAY_ALL));
+
 	if (!xprt->inet) {
 		struct sock *sk = sock->sk;
 
@@ -1260,6 +1317,7 @@ static void xs_tcp_print_stats(struct rpc_xprt *xprt, struct seq_file *seq)
 
 static struct rpc_xprt_ops xs_udp_ops = {
 	.set_buffer_size	= xs_udp_set_buffer_size,
+	.print_addr		= xs_print_peer_address,
 	.reserve_xprt		= xprt_reserve_xprt_cong,
 	.release_xprt		= xprt_release_xprt_cong,
 	.rpcbind		= rpc_getport,
@@ -1277,6 +1335,7 @@ static struct rpc_xprt_ops xs_udp_ops = {
 };
 
 static struct rpc_xprt_ops xs_tcp_ops = {
+	.print_addr		= xs_print_peer_address,
 	.reserve_xprt		= xprt_reserve_xprt,
 	.release_xprt		= xs_tcp_release_xprt,
 	.rpcbind		= rpc_getport,
@@ -1301,8 +1360,6 @@ int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to)
 {
 	size_t slot_table_size;
 
-	dprintk("RPC:      setting up udp-ipv4 transport...\n");
-
 	xprt->max_reqs = xprt_udp_slot_table_entries;
 	slot_table_size = xprt->max_reqs * sizeof(xprt->slot[0]);
 	xprt->slot = kzalloc(slot_table_size, GFP_KERNEL);
@@ -1332,6 +1389,10 @@ int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to)
 	else
 		xprt_set_timeout(&xprt->timeout, 5, 5 * HZ);
 
+	xs_format_peer_addresses(xprt);
+	dprintk("RPC:      set up transport to address %s\n",
+			xs_print_peer_address(xprt, RPC_DISPLAY_ALL));
+
 	return 0;
 }
 
@@ -1345,8 +1406,6 @@ int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to)
 {
 	size_t slot_table_size;
 
-	dprintk("RPC:      setting up tcp-ipv4 transport...\n");
-
 	xprt->max_reqs = xprt_tcp_slot_table_entries;
 	slot_table_size = xprt->max_reqs * sizeof(xprt->slot[0]);
 	xprt->slot = kzalloc(slot_table_size, GFP_KERNEL);
@@ -1375,5 +1434,9 @@ int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to)
 	else
 		xprt_set_timeout(&xprt->timeout, 2, 60 * HZ);
 
+	xs_format_peer_addresses(xprt);
+	dprintk("RPC:      set up transport to address %s\n",
+			xs_print_peer_address(xprt, RPC_DISPLAY_ALL));
+
 	return 0;
 }
-- 
cgit v1.2.3


From f425eba437f0051bde979ea2eef8bc875a77cd00 Mon Sep 17 00:00:00 2001
From: Chuck Lever <chuck.lever@oracle.com>
Date: Tue, 22 Aug 2006 20:06:18 -0400
Subject: SUNRPC: Create API for displaying remote peer address

Provide an API for formatting the remote peer address for printing without
exposing its internal structure.  The address could be dynamic, so we
support a function call to get the address rather than reading it straight
out of a structure.

Test-plan:
Destructive testing (unplugging the network temporarily).  Probably need
to rig a server where certain services aren't running, or that returns an
error for some typical operation.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 include/linux/sunrpc/clnt.h |  1 +
 net/sunrpc/clnt.c           | 13 +++++++++++++
 2 files changed, 14 insertions(+)

(limited to 'include')

diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h
index 65196b03f0ab..b7d47f018353 100644
--- a/include/linux/sunrpc/clnt.h
+++ b/include/linux/sunrpc/clnt.h
@@ -124,6 +124,7 @@ size_t		rpc_max_payload(struct rpc_clnt *);
 void		rpc_force_rebind(struct rpc_clnt *);
 int		rpc_ping(struct rpc_clnt *clnt, int flags);
 size_t		rpc_peeraddr(struct rpc_clnt *, struct sockaddr *, size_t);
+char *		rpc_peeraddr2str(struct rpc_clnt *, enum rpc_display_format_t);
 
 /*
  * Helper function for NFSroot support
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index 94768cf5fd5b..e5b19e348d88 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -554,6 +554,19 @@ size_t rpc_peeraddr(struct rpc_clnt *clnt, struct sockaddr *buf, size_t bufsize)
 }
 EXPORT_SYMBOL(rpc_peeraddr);
 
+/**
+ * rpc_peeraddr2str - return remote peer address in printable format
+ * @clnt: RPC client structure
+ * @format: address format
+ *
+ */
+char *rpc_peeraddr2str(struct rpc_clnt *clnt, enum rpc_display_format_t format)
+{
+	struct rpc_xprt *xprt = clnt->cl_xprt;
+	return xprt->ops->print_addr(xprt, format);
+}
+EXPORT_SYMBOL(rpc_peeraddr2str);
+
 void
 rpc_setbufsize(struct rpc_clnt *clnt, unsigned int sndsize, unsigned int rcvsize)
 {
-- 
cgit v1.2.3


From c4efcb1d3e0bc76aeb9ca6301d19a5079893c6c9 Mon Sep 17 00:00:00 2001
From: Chuck Lever <chuck.lever@oracle.com>
Date: Tue, 22 Aug 2006 20:06:19 -0400
Subject: SUNRPC: Use "sockaddr_storage" for storing RPC client's remote peer
 address

IPv6 addresses are big (128 bytes).  Now that no RPC client consumers treat
the addr field in rpc_xprt structs as an opaque, and access it only via the
API calls, we can safely widen the field in the rpc_xprt struct to
accomodate larger addresses.

Test plan:
Compile kernel with CONFIG_NFS enabled.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 include/linux/sunrpc/xprt.h |  3 ++-
 net/sunrpc/clnt.c           |  2 +-
 net/sunrpc/xprt.c           |  3 ++-
 net/sunrpc/xprtsock.c       | 15 ++++++++++-----
 4 files changed, 15 insertions(+), 8 deletions(-)

(limited to 'include')

diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h
index 8372ab8fc9b5..fc05cfbd5805 100644
--- a/include/linux/sunrpc/xprt.h
+++ b/include/linux/sunrpc/xprt.h
@@ -134,7 +134,8 @@ struct rpc_xprt {
 	struct sock *		inet;		/* INET layer */
 
 	struct rpc_timeout	timeout;	/* timeout parms */
-	struct sockaddr_in	addr;		/* server address */
+	struct sockaddr_storage	addr;		/* server address */
+	size_t			addrlen;	/* size of server address */
 	int			prot;		/* IP protocol */
 
 	unsigned long		cong;		/* current congestion */
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index e5b19e348d88..ff1e90fd81ab 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -550,7 +550,7 @@ size_t rpc_peeraddr(struct rpc_clnt *clnt, struct sockaddr *buf, size_t bufsize)
 	if (bytes > bufsize)
 		bytes = bufsize;
 	memcpy(buf, &clnt->cl_xprt->addr, bytes);
-	return sizeof(xprt->addr);
+	return xprt->addrlen;
 }
 EXPORT_SYMBOL(rpc_peeraddr);
 
diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index b45abd0743cb..4987517cc74b 100644
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -896,7 +896,8 @@ static struct rpc_xprt *xprt_setup(int proto, struct sockaddr_in *ap, struct rpc
 	if ((xprt = kzalloc(sizeof(struct rpc_xprt), GFP_KERNEL)) == NULL)
 		return ERR_PTR(-ENOMEM);
 
-	xprt->addr = *ap;
+	memcpy(&xprt->addr, ap, sizeof(*ap));
+	xprt->addrlen = sizeof(*ap);
 
 	switch (proto) {
 	case IPPROTO_UDP:
diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c
index cb8e6c34e12f..17179aa4c207 100644
--- a/net/sunrpc/xprtsock.c
+++ b/net/sunrpc/xprtsock.c
@@ -341,7 +341,7 @@ static int xs_udp_send_request(struct rpc_task *task)
 
 	req->rq_xtime = jiffies;
 	status = xs_sendpages(xprt->sock, (struct sockaddr *) &xprt->addr,
-				sizeof(xprt->addr), xdr, req->rq_bytes_sent);
+				xprt->addrlen, xdr, req->rq_bytes_sent);
 
 	dprintk("RPC:      xs_udp_send_request(%u) = %d\n",
 			xdr->len - req->rq_bytes_sent, status);
@@ -1027,8 +1027,11 @@ static char *xs_print_peer_address(struct rpc_xprt *xprt, enum rpc_display_forma
  */
 static void xs_set_port(struct rpc_xprt *xprt, unsigned short port)
 {
+	struct sockaddr_in *sap = (struct sockaddr_in *) &xprt->addr;
+
 	dprintk("RPC:      setting port for xprt %p to %u\n", xprt, port);
-	xprt->addr.sin_port = htons(port);
+
+	sap->sin_port = htons(port);
 }
 
 static int xs_bindresvport(struct rpc_xprt *xprt, struct socket *sock)
@@ -1209,7 +1212,7 @@ static void xs_tcp_connect_worker(void *args)
 	xprt->stat.connect_count++;
 	xprt->stat.connect_start = jiffies;
 	status = sock->ops->connect(sock, (struct sockaddr *) &xprt->addr,
-			sizeof(xprt->addr), O_NONBLOCK);
+			xprt->addrlen, O_NONBLOCK);
 	dprintk("RPC: %p  connect status %d connected %d sock state %d\n",
 			xprt, -status, xprt_connected(xprt), sock->sk->sk_state);
 	if (status < 0) {
@@ -1359,6 +1362,7 @@ static struct rpc_xprt_ops xs_tcp_ops = {
 int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to)
 {
 	size_t slot_table_size;
+	struct sockaddr_in *addr = (struct sockaddr_in *) &xprt->addr;
 
 	xprt->max_reqs = xprt_udp_slot_table_entries;
 	slot_table_size = xprt->max_reqs * sizeof(xprt->slot[0]);
@@ -1366,7 +1370,7 @@ int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to)
 	if (xprt->slot == NULL)
 		return -ENOMEM;
 
-	if (ntohs(xprt->addr.sin_port) != 0)
+	if (ntohs(addr->sin_port != 0))
 		xprt_set_bound(xprt);
 	xprt->port = xs_get_random_port();
 
@@ -1405,6 +1409,7 @@ int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to)
 int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to)
 {
 	size_t slot_table_size;
+	struct sockaddr_in *addr = (struct sockaddr_in *) &xprt->addr;
 
 	xprt->max_reqs = xprt_tcp_slot_table_entries;
 	slot_table_size = xprt->max_reqs * sizeof(xprt->slot[0]);
@@ -1412,7 +1417,7 @@ int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to)
 	if (xprt->slot == NULL)
 		return -ENOMEM;
 
-	if (ntohs(xprt->addr.sin_port) != 0)
+	if (ntohs(addr->sin_port) != 0)
 		xprt_set_bound(xprt);
 	xprt->port = xs_get_random_port();
 
-- 
cgit v1.2.3


From 6ca948238724c945bd353f51d54ae7d285f3889f Mon Sep 17 00:00:00 2001
From: Chuck Lever <chuck.lever@oracle.com>
Date: Tue, 22 Aug 2006 20:06:19 -0400
Subject: SUNRPC: Clean-up after previous patches.

Remove some unused macros related to accessing an RPC peer address

Test plan:
Compile kernel with CONFIG_NFS option enabled.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/lockd/host.c             | 1 -
 include/linux/nfs_fs.h      | 1 -
 include/linux/sunrpc/clnt.h | 3 ---
 3 files changed, 5 deletions(-)

(limited to 'include')

diff --git a/fs/lockd/host.c b/fs/lockd/host.c
index 38b0e8a1aec0..a516a01561b8 100644
--- a/fs/lockd/host.c
+++ b/fs/lockd/host.c
@@ -26,7 +26,6 @@
 #define NLM_HOST_REBIND		(60 * HZ)
 #define NLM_HOST_EXPIRE		((nrhosts > NLM_HOST_MAX)? 300 * HZ : 120 * HZ)
 #define NLM_HOST_COLLECT	((nrhosts > NLM_HOST_MAX)? 120 * HZ :  60 * HZ)
-#define NLM_HOST_ADDR(sv)	(&(sv)->s_nlmclnt->cl_xprt->addr)
 
 static struct nlm_host *	nlm_hosts[NLM_HOST_NRHASH];
 static unsigned long		next_gc;
diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h
index 51e9bd90dedc..3b5b04193fee 100644
--- a/include/linux/nfs_fs.h
+++ b/include/linux/nfs_fs.h
@@ -216,7 +216,6 @@ static inline struct nfs_inode *NFS_I(struct inode *inode)
 #define NFS_SERVER(inode)		(NFS_SB(inode->i_sb))
 #define NFS_CLIENT(inode)		(NFS_SERVER(inode)->client)
 #define NFS_PROTO(inode)		(NFS_SERVER(inode)->nfs_client->rpc_ops)
-#define NFS_ADDR(inode)			(RPC_PEERADDR(NFS_CLIENT(inode)))
 #define NFS_COOKIEVERF(inode)		(NFS_I(inode)->cookieverf)
 #define NFS_READTIME(inode)		(NFS_I(inode)->read_cache_jiffies)
 #define NFS_CHANGE_ATTR(inode)		(NFS_I(inode)->change_attr)
diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h
index b7d47f018353..a26d69583c7a 100644
--- a/include/linux/sunrpc/clnt.h
+++ b/include/linux/sunrpc/clnt.h
@@ -89,9 +89,6 @@ struct rpc_procinfo {
 	char *			p_name;		/* name of procedure */
 };
 
-#define RPC_CONGESTED(clnt)	(RPCXPRT_CONGESTED((clnt)->cl_xprt))
-#define RPC_PEERADDR(clnt)	(&(clnt)->cl_xprt->addr)
-
 #ifdef __KERNEL__
 
 struct rpc_clnt *rpc_create_client(struct rpc_xprt *xprt, char *servname,
-- 
cgit v1.2.3


From c2866763b4029411d166040306691773c12d4caf Mon Sep 17 00:00:00 2001
From: Chuck Lever <chuck.lever@oracle.com>
Date: Tue, 22 Aug 2006 20:06:20 -0400
Subject: SUNRPC: use sockaddr + size when creating remote transport endpoints

Prepare for more generic transport endpoint handling needed by transports
that might use different forms of addressing, such as IPv6.

Introduce a single function call to replace the two-call
xprt_create_proto/rpc_create_client API.  Define a new rpc_create_args
structure that allows callers to pass in remote endpoint addresses of
varying length.

Test-plan:
Compile kernel with CONFIG_NFS enabled.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 include/linux/sunrpc/clnt.h | 22 +++++++++++++
 include/linux/sunrpc/xprt.h |  1 +
 net/sunrpc/clnt.c           | 61 ++++++++++++++++++++++++++++++++++++
 net/sunrpc/xprt.c           | 75 +++++++++++++++++++++++++++++++++++++++++++++
 4 files changed, 159 insertions(+)

(limited to 'include')

diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h
index a26d69583c7a..7817ba82f1b2 100644
--- a/include/linux/sunrpc/clnt.h
+++ b/include/linux/sunrpc/clnt.h
@@ -97,6 +97,28 @@ struct rpc_clnt *rpc_create_client(struct rpc_xprt *xprt, char *servname,
 struct rpc_clnt *rpc_new_client(struct rpc_xprt *xprt, char *servname,
 				struct rpc_program *info,
 				u32 version, rpc_authflavor_t authflavor);
+
+struct rpc_create_args {
+	int			protocol;
+	struct sockaddr		*address;
+	size_t			addrsize;
+	struct rpc_timeout	*timeout;
+	char			*servername;
+	struct rpc_program	*program;
+	u32			version;
+	rpc_authflavor_t	authflavor;
+	unsigned long		flags;
+};
+
+/* Values for "flags" field */
+#define RPC_CLNT_CREATE_HARDRTRY	(1UL << 0)
+#define RPC_CLNT_CREATE_INTR		(1UL << 1)
+#define RPC_CLNT_CREATE_AUTOBIND	(1UL << 2)
+#define RPC_CLNT_CREATE_ONESHOT		(1UL << 3)
+#define RPC_CLNT_CREATE_NONPRIVPORT	(1UL << 4)
+#define RPC_CLNT_CREATE_NOPING		(1UL << 5)
+
+struct rpc_clnt *rpc_create(struct rpc_create_args *args);
 struct rpc_clnt	*rpc_bind_new_program(struct rpc_clnt *,
 				struct rpc_program *, int);
 struct rpc_clnt *rpc_clone_client(struct rpc_clnt *);
diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h
index fc05cfbd5805..bc80fcfdd892 100644
--- a/include/linux/sunrpc/xprt.h
+++ b/include/linux/sunrpc/xprt.h
@@ -237,6 +237,7 @@ void			xprt_set_timeout(struct rpc_timeout *to, unsigned int retr, unsigned long
 /*
  * Generic internal transport functions
  */
+struct rpc_xprt *	xprt_create_transport(int proto, struct sockaddr *addr, size_t size, struct rpc_timeout *toparms);
 void			xprt_connect(struct rpc_task *task);
 void			xprt_reserve(struct rpc_task *task);
 int			xprt_reserve_xprt(struct rpc_task *task);
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index ff1e90fd81ab..dbb93bdf6cc9 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -192,6 +192,67 @@ out_no_xprt:
 	return ERR_PTR(err);
 }
 
+/*
+ * rpc_create - create an RPC client and transport with one call
+ * @args: rpc_clnt create argument structure
+ *
+ * Creates and initializes an RPC transport and an RPC client.
+ *
+ * It can ping the server in order to determine if it is up, and to see if
+ * it supports this program and version.  RPC_CLNT_CREATE_NOPING disables
+ * this behavior so asynchronous tasks can also use rpc_create.
+ */
+struct rpc_clnt *rpc_create(struct rpc_create_args *args)
+{
+	struct rpc_xprt *xprt;
+	struct rpc_clnt *clnt;
+
+	xprt = xprt_create_transport(args->protocol, args->address,
+					args->addrsize, args->timeout);
+	if (IS_ERR(xprt))
+		return (struct rpc_clnt *)xprt;
+
+	/*
+	 * By default, kernel RPC client connects from a reserved port.
+	 * CAP_NET_BIND_SERVICE will not be set for unprivileged requesters,
+	 * but it is always enabled for rpciod, which handles the connect
+	 * operation.
+	 */
+	xprt->resvport = 1;
+	if (args->flags & RPC_CLNT_CREATE_NONPRIVPORT)
+		xprt->resvport = 0;
+
+	dprintk("RPC:       creating %s client for %s (xprt %p)\n",
+		args->program->name, args->servername, xprt);
+
+	clnt = rpc_new_client(xprt, args->servername, args->program,
+				args->version, args->authflavor);
+	if (IS_ERR(clnt))
+		return clnt;
+
+	if (!(args->flags & RPC_CLNT_CREATE_NOPING)) {
+		int err = rpc_ping(clnt, RPC_TASK_SOFT|RPC_TASK_NOINTR);
+		if (err != 0) {
+			rpc_shutdown_client(clnt);
+			return ERR_PTR(err);
+		}
+	}
+
+	clnt->cl_softrtry = 1;
+	if (args->flags & RPC_CLNT_CREATE_HARDRTRY)
+		clnt->cl_softrtry = 0;
+
+	if (args->flags & RPC_CLNT_CREATE_INTR)
+		clnt->cl_intr = 1;
+	if (args->flags & RPC_CLNT_CREATE_AUTOBIND)
+		clnt->cl_autobind = 1;
+	if (args->flags & RPC_CLNT_CREATE_ONESHOT)
+		clnt->cl_oneshot = 1;
+
+	return clnt;
+}
+EXPORT_SYMBOL(rpc_create);
+
 /**
  * Create an RPC client
  * @xprt - pointer to xprt struct
diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index 4987517cc74b..17f56cfe2412 100644
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -887,6 +887,81 @@ void xprt_set_timeout(struct rpc_timeout *to, unsigned int retr, unsigned long i
 	to->to_exponential = 0;
 }
 
+/**
+ * xprt_create_transport - create an RPC transport
+ * @proto: requested transport protocol
+ * @ap: remote peer address
+ * @size: length of address
+ * @to: timeout parameters
+ *
+ */
+struct rpc_xprt *xprt_create_transport(int proto, struct sockaddr *ap, size_t size, struct rpc_timeout *to)
+{
+	int result;
+	struct rpc_xprt	*xprt;
+	struct rpc_rqst	*req;
+
+	if ((xprt = kzalloc(sizeof(struct rpc_xprt), GFP_KERNEL)) == NULL) {
+		dprintk("RPC:      xprt_create_transport: no memory\n");
+		return ERR_PTR(-ENOMEM);
+	}
+	if (size <= sizeof(xprt->addr)) {
+		memcpy(&xprt->addr, ap, size);
+		xprt->addrlen = size;
+	} else {
+		kfree(xprt);
+		dprintk("RPC:      xprt_create_transport: address too large\n");
+		return ERR_PTR(-EBADF);
+	}
+
+	switch (proto) {
+	case IPPROTO_UDP:
+		result = xs_setup_udp(xprt, to);
+		break;
+	case IPPROTO_TCP:
+		result = xs_setup_tcp(xprt, to);
+		break;
+	default:
+		printk(KERN_ERR "RPC: unrecognized transport protocol: %d\n",
+				proto);
+		return ERR_PTR(-EIO);
+	}
+	if (result) {
+		kfree(xprt);
+		dprintk("RPC:      xprt_create_transport: failed, %d\n", result);
+		return ERR_PTR(result);
+	}
+
+	spin_lock_init(&xprt->transport_lock);
+	spin_lock_init(&xprt->reserve_lock);
+
+	INIT_LIST_HEAD(&xprt->free);
+	INIT_LIST_HEAD(&xprt->recv);
+	INIT_WORK(&xprt->task_cleanup, xprt_autoclose, xprt);
+	init_timer(&xprt->timer);
+	xprt->timer.function = xprt_init_autodisconnect;
+	xprt->timer.data = (unsigned long) xprt;
+	xprt->last_used = jiffies;
+	xprt->cwnd = RPC_INITCWND;
+
+	rpc_init_wait_queue(&xprt->binding, "xprt_binding");
+	rpc_init_wait_queue(&xprt->pending, "xprt_pending");
+	rpc_init_wait_queue(&xprt->sending, "xprt_sending");
+	rpc_init_wait_queue(&xprt->resend, "xprt_resend");
+	rpc_init_priority_wait_queue(&xprt->backlog, "xprt_backlog");
+
+	/* initialize free list */
+	for (req = &xprt->slot[xprt->max_reqs-1]; req >= &xprt->slot[0]; req--)
+		list_add(&req->rq_list, &xprt->free);
+
+	xprt_init_xid(xprt);
+
+	dprintk("RPC:      created transport %p with %u slots\n", xprt,
+			xprt->max_reqs);
+
+	return xprt;
+}
+
 static struct rpc_xprt *xprt_setup(int proto, struct sockaddr_in *ap, struct rpc_timeout *to)
 {
 	int result;
-- 
cgit v1.2.3


From ff9aa5e56df60cc8565a93cc868fe25ae3f20e49 Mon Sep 17 00:00:00 2001
From: Chuck Lever <chuck.lever@oracle.com>
Date: Tue, 22 Aug 2006 20:06:21 -0400
Subject: SUNRPC: Eliminate xprt_create_proto and rpc_create_client

The two function call API for creating a new RPC client is now obsolete.
Remove it.

Also, remove an unnecessary check to see whether the caller is capable of
using privileged network services.  The kernel RPC client always uses a
privileged ephemeral port by default; callers are responsible for checking
the authority of users to make use of any RPC service, or for specifying
that a nonprivileged port is acceptable.

Test plan:
Repeated runs of Connectathon locking suite.  Check network trace to ensure
correctness of NLM requests and replies.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 include/linux/sunrpc/clnt.h |  7 ----
 include/linux/sunrpc/xprt.h |  1 -
 net/sunrpc/clnt.c           | 42 +-----------------------
 net/sunrpc/sunrpc_syms.c    |  3 --
 net/sunrpc/xprt.c           | 79 ---------------------------------------------
 net/sunrpc/xprtsock.c       |  2 --
 6 files changed, 1 insertion(+), 133 deletions(-)

(limited to 'include')

diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h
index 7817ba82f1b2..f6d1d646ce05 100644
--- a/include/linux/sunrpc/clnt.h
+++ b/include/linux/sunrpc/clnt.h
@@ -91,13 +91,6 @@ struct rpc_procinfo {
 
 #ifdef __KERNEL__
 
-struct rpc_clnt *rpc_create_client(struct rpc_xprt *xprt, char *servname,
-				struct rpc_program *info,
-				u32 version, rpc_authflavor_t authflavor);
-struct rpc_clnt *rpc_new_client(struct rpc_xprt *xprt, char *servname,
-				struct rpc_program *info,
-				u32 version, rpc_authflavor_t authflavor);
-
 struct rpc_create_args {
 	int			protocol;
 	struct sockaddr		*address;
diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h
index bc80fcfdd892..de4efea7c856 100644
--- a/include/linux/sunrpc/xprt.h
+++ b/include/linux/sunrpc/xprt.h
@@ -231,7 +231,6 @@ struct rpc_xprt {
 /*
  * Transport operations used by ULPs
  */
-struct rpc_xprt *	xprt_create_proto(int proto, struct sockaddr_in *addr, struct rpc_timeout *to);
 void			xprt_set_timeout(struct rpc_timeout *to, unsigned int retr, unsigned long incr);
 
 /*
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index dbb93bdf6cc9..428704dd5b3e 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -97,17 +97,7 @@ rpc_setup_pipedir(struct rpc_clnt *clnt, char *dir_name)
 	}
 }
 
-/*
- * Create an RPC client
- * FIXME: This should also take a flags argument (as in task->tk_flags).
- * It's called (among others) from pmap_create_client, which may in
- * turn be called by an async task. In this case, rpciod should not be
- * made to sleep too long.
- */
-struct rpc_clnt *
-rpc_new_client(struct rpc_xprt *xprt, char *servname,
-		  struct rpc_program *program, u32 vers,
-		  rpc_authflavor_t flavor)
+static struct rpc_clnt * rpc_new_client(struct rpc_xprt *xprt, char *servname, struct rpc_program *program, u32 vers, rpc_authflavor_t flavor)
 {
 	struct rpc_version	*version;
 	struct rpc_clnt		*clnt = NULL;
@@ -253,36 +243,6 @@ struct rpc_clnt *rpc_create(struct rpc_create_args *args)
 }
 EXPORT_SYMBOL(rpc_create);
 
-/**
- * Create an RPC client
- * @xprt - pointer to xprt struct
- * @servname - name of server
- * @info - rpc_program
- * @version - rpc_program version
- * @authflavor - rpc_auth flavour to use
- *
- * Creates an RPC client structure, then pings the server in order to
- * determine if it is up, and if it supports this program and version.
- *
- * This function should never be called by asynchronous tasks such as
- * the portmapper.
- */
-struct rpc_clnt *rpc_create_client(struct rpc_xprt *xprt, char *servname,
-		struct rpc_program *info, u32 version, rpc_authflavor_t authflavor)
-{
-	struct rpc_clnt *clnt;
-	int err;
-	
-	clnt = rpc_new_client(xprt, servname, info, version, authflavor);
-	if (IS_ERR(clnt))
-		return clnt;
-	err = rpc_ping(clnt, RPC_TASK_SOFT|RPC_TASK_NOINTR);
-	if (err == 0)
-		return clnt;
-	rpc_shutdown_client(clnt);
-	return ERR_PTR(err);
-}
-
 /*
  * This function clones the RPC client structure. It allows us to share the
  * same transport while varying parameters such as the authentication
diff --git a/net/sunrpc/sunrpc_syms.c b/net/sunrpc/sunrpc_syms.c
index f38f939ce95f..26c0531d7e25 100644
--- a/net/sunrpc/sunrpc_syms.c
+++ b/net/sunrpc/sunrpc_syms.c
@@ -36,8 +36,6 @@ EXPORT_SYMBOL(rpc_wake_up_status);
 EXPORT_SYMBOL(rpc_release_task);
 
 /* RPC client functions */
-EXPORT_SYMBOL(rpc_create_client);
-EXPORT_SYMBOL(rpc_new_client);
 EXPORT_SYMBOL(rpc_clone_client);
 EXPORT_SYMBOL(rpc_bind_new_program);
 EXPORT_SYMBOL(rpc_destroy_client);
@@ -57,7 +55,6 @@ EXPORT_SYMBOL(rpc_queue_upcall);
 EXPORT_SYMBOL(rpc_mkpipe);
 
 /* Client transport */
-EXPORT_SYMBOL(xprt_create_proto);
 EXPORT_SYMBOL(xprt_set_timeout);
 
 /* Client credential cache */
diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index 17f56cfe2412..e4f64fb58ff2 100644
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -962,85 +962,6 @@ struct rpc_xprt *xprt_create_transport(int proto, struct sockaddr *ap, size_t si
 	return xprt;
 }
 
-static struct rpc_xprt *xprt_setup(int proto, struct sockaddr_in *ap, struct rpc_timeout *to)
-{
-	int result;
-	struct rpc_xprt	*xprt;
-	struct rpc_rqst	*req;
-
-	if ((xprt = kzalloc(sizeof(struct rpc_xprt), GFP_KERNEL)) == NULL)
-		return ERR_PTR(-ENOMEM);
-
-	memcpy(&xprt->addr, ap, sizeof(*ap));
-	xprt->addrlen = sizeof(*ap);
-
-	switch (proto) {
-	case IPPROTO_UDP:
-		result = xs_setup_udp(xprt, to);
-		break;
-	case IPPROTO_TCP:
-		result = xs_setup_tcp(xprt, to);
-		break;
-	default:
-		printk(KERN_ERR "RPC: unrecognized transport protocol: %d\n",
-				proto);
-		result = -EIO;
-		break;
-	}
-	if (result) {
-		kfree(xprt);
-		return ERR_PTR(result);
-	}
-
-	spin_lock_init(&xprt->transport_lock);
-	spin_lock_init(&xprt->reserve_lock);
-
-	INIT_LIST_HEAD(&xprt->free);
-	INIT_LIST_HEAD(&xprt->recv);
-	INIT_WORK(&xprt->task_cleanup, xprt_autoclose, xprt);
-	init_timer(&xprt->timer);
-	xprt->timer.function = xprt_init_autodisconnect;
-	xprt->timer.data = (unsigned long) xprt;
-	xprt->last_used = jiffies;
-	xprt->cwnd = RPC_INITCWND;
-
-	rpc_init_wait_queue(&xprt->binding, "xprt_binding");
-	rpc_init_wait_queue(&xprt->pending, "xprt_pending");
-	rpc_init_wait_queue(&xprt->sending, "xprt_sending");
-	rpc_init_wait_queue(&xprt->resend, "xprt_resend");
-	rpc_init_priority_wait_queue(&xprt->backlog, "xprt_backlog");
-
-	/* initialize free list */
-	for (req = &xprt->slot[xprt->max_reqs-1]; req >= &xprt->slot[0]; req--)
-		list_add(&req->rq_list, &xprt->free);
-
-	xprt_init_xid(xprt);
-
-	dprintk("RPC:      created transport %p with %u slots\n", xprt,
-			xprt->max_reqs);
-	
-	return xprt;
-}
-
-/**
- * xprt_create_proto - create an RPC client transport
- * @proto: requested transport protocol
- * @sap: remote peer's address
- * @to: timeout parameters for new transport
- *
- */
-struct rpc_xprt *xprt_create_proto(int proto, struct sockaddr_in *sap, struct rpc_timeout *to)
-{
-	struct rpc_xprt	*xprt;
-
-	xprt = xprt_setup(proto, sap, to);
-	if (IS_ERR(xprt))
-		dprintk("RPC:      xprt_create_proto failed\n");
-	else
-		dprintk("RPC:      xprt_create_proto created xprt %p\n", xprt);
-	return xprt;
-}
-
 /**
  * xprt_destroy - destroy an RPC transport, killing off all requests.
  * @xprt: transport to destroy
diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c
index 17179aa4c207..0b84fab68d7e 100644
--- a/net/sunrpc/xprtsock.c
+++ b/net/sunrpc/xprtsock.c
@@ -1376,7 +1376,6 @@ int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to)
 
 	xprt->prot = IPPROTO_UDP;
 	xprt->tsh_size = 0;
-	xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0;
 	/* XXX: header size can vary due to auth type, IPv6, etc. */
 	xprt->max_payload = (1U << 16) - (MAX_HEADER << 3);
 
@@ -1423,7 +1422,6 @@ int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to)
 
 	xprt->prot = IPPROTO_TCP;
 	xprt->tsh_size = sizeof(rpc_fraghdr) / sizeof(u32);
-	xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0;
 	xprt->max_payload = RPC_MAX_FRAGMENT_SIZE;
 
 	INIT_WORK(&xprt->connect_worker, xs_tcp_connect_worker, xprt);
-- 
cgit v1.2.3


From 4f390c152bc87165da4b1f5b7d870b46fb106d4e Mon Sep 17 00:00:00 2001
From: Chuck Lever <chuck.lever@oracle.com>
Date: Tue, 22 Aug 2006 20:06:22 -0400
Subject: NFS: Fix double d_drop in nfs_instantiate() error path

If the LOOKUP or GETATTR in nfs_instantiate fail, nfs_instantiate will do a
d_drop before returning.  But some callers already do a d_drop in the case
of an error return.  Make certain we do only one d_drop in all error paths.

This issue was introduced because over time, the symlink proc API diverged
slightly from the create/mkdir/mknod proc API.  To prevent other coding
mistakes of this type, change the symlink proc API to be more like
create/mkdir/mknod and move the nfs_instantiate call into the symlink proc
routines so it is used in exactly the same way for create, mkdir, mknod,
and symlink.

Test plan:
Connectathon, all versions of NFS.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/dir.c            | 16 ++++------------
 fs/nfs/nfs3proc.c       | 26 ++++++++++++++++----------
 fs/nfs/nfs4proc.c       | 31 ++++++++++++++++---------------
 fs/nfs/proc.c           | 29 +++++++++++++++++++++--------
 include/linux/nfs_xdr.h |  5 ++---
 5 files changed, 59 insertions(+), 48 deletions(-)

(limited to 'include')

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 084e8cb41c84..affd3ae52e55 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -1147,23 +1147,20 @@ int nfs_instantiate(struct dentry *dentry, struct nfs_fh *fhandle,
 		struct inode *dir = dentry->d_parent->d_inode;
 		error = NFS_PROTO(dir)->lookup(dir, &dentry->d_name, fhandle, fattr);
 		if (error)
-			goto out_err;
+			return error;
 	}
 	if (!(fattr->valid & NFS_ATTR_FATTR)) {
 		struct nfs_server *server = NFS_SB(dentry->d_sb);
 		error = server->nfs_client->rpc_ops->getattr(server, fhandle, fattr);
 		if (error < 0)
-			goto out_err;
+			return error;
 	}
 	inode = nfs_fhget(dentry->d_sb, fhandle, fattr);
 	error = PTR_ERR(inode);
 	if (IS_ERR(inode))
-		goto out_err;
+		return error;
 	d_instantiate(dentry, inode);
 	return 0;
-out_err:
-	d_drop(dentry);
-	return error;
 }
 
 /*
@@ -1448,8 +1445,6 @@ static int
 nfs_symlink(struct inode *dir, struct dentry *dentry, const char *symname)
 {
 	struct iattr attr;
-	struct nfs_fattr sym_attr;
-	struct nfs_fh sym_fh;
 	struct qstr qsymname;
 	int error;
 
@@ -1473,12 +1468,9 @@ dentry->d_parent->d_name.name, dentry->d_name.name);
 
 	lock_kernel();
 	nfs_begin_data_update(dir);
-	error = NFS_PROTO(dir)->symlink(dir, &dentry->d_name, &qsymname,
-					  &attr, &sym_fh, &sym_attr);
+	error = NFS_PROTO(dir)->symlink(dir, dentry, &qsymname, &attr);
 	nfs_end_data_update(dir);
 	if (!error)
-		error = nfs_instantiate(dentry, &sym_fh, &sym_attr);
-	else
 		d_drop(dentry);
 	unlock_kernel();
 	return error;
diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c
index 9e8258ece6fd..d85ac427c326 100644
--- a/fs/nfs/nfs3proc.c
+++ b/fs/nfs/nfs3proc.c
@@ -544,23 +544,23 @@ nfs3_proc_link(struct inode *inode, struct inode *dir, struct qstr *name)
 }
 
 static int
-nfs3_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path,
-		  struct iattr *sattr, struct nfs_fh *fhandle,
-		  struct nfs_fattr *fattr)
+nfs3_proc_symlink(struct inode *dir, struct dentry *dentry, struct qstr *path,
+		  struct iattr *sattr)
 {
-	struct nfs_fattr	dir_attr;
+	struct nfs_fh fhandle;
+	struct nfs_fattr fattr, dir_attr;
 	struct nfs3_symlinkargs	arg = {
 		.fromfh		= NFS_FH(dir),
-		.fromname	= name->name,
-		.fromlen	= name->len,
+		.fromname	= dentry->d_name.name,
+		.fromlen	= dentry->d_name.len,
 		.topath		= path->name,
 		.tolen		= path->len,
 		.sattr		= sattr
 	};
 	struct nfs3_diropres	res = {
 		.dir_attr	= &dir_attr,
-		.fh		= fhandle,
-		.fattr		= fattr
+		.fh		= &fhandle,
+		.fattr		= &fattr
 	};
 	struct rpc_message msg = {
 		.rpc_proc	= &nfs3_procedures[NFS3PROC_SYMLINK],
@@ -571,11 +571,17 @@ nfs3_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path,
 
 	if (path->len > NFS3_MAXPATHLEN)
 		return -ENAMETOOLONG;
-	dprintk("NFS call  symlink %s -> %s\n", name->name, path->name);
+
+	dprintk("NFS call  symlink %s -> %s\n", dentry->d_name.name,
+			path->name);
 	nfs_fattr_init(&dir_attr);
-	nfs_fattr_init(fattr);
+	nfs_fattr_init(&fattr);
 	status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0);
 	nfs_post_op_update_inode(dir, &dir_attr);
+	if (status != 0)
+		goto out;
+	status = nfs_instantiate(dentry, &fhandle, &fattr);
+out:
 	dprintk("NFS reply symlink: %d\n", status);
 	return status;
 }
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index a825547e8214..2d18eac6bee5 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -2084,24 +2084,24 @@ static int nfs4_proc_link(struct inode *inode, struct inode *dir, struct qstr *n
 	return err;
 }
 
-static int _nfs4_proc_symlink(struct inode *dir, struct qstr *name,
-		struct qstr *path, struct iattr *sattr, struct nfs_fh *fhandle,
-		struct nfs_fattr *fattr)
+static int _nfs4_proc_symlink(struct inode *dir, struct dentry *dentry,
+		struct qstr *path, struct iattr *sattr)
 {
 	struct nfs_server *server = NFS_SERVER(dir);
-	struct nfs_fattr dir_fattr;
+	struct nfs_fh fhandle;
+	struct nfs_fattr fattr, dir_fattr;
 	struct nfs4_create_arg arg = {
 		.dir_fh = NFS_FH(dir),
 		.server = server,
-		.name = name,
+		.name = &dentry->d_name,
 		.attrs = sattr,
 		.ftype = NF4LNK,
 		.bitmask = server->attr_bitmask,
 	};
 	struct nfs4_create_res res = {
 		.server = server,
-		.fh = fhandle,
-		.fattr = fattr,
+		.fh = &fhandle,
+		.fattr = &fattr,
 		.dir_fattr = &dir_fattr,
 	};
 	struct rpc_message msg = {
@@ -2113,27 +2113,28 @@ static int _nfs4_proc_symlink(struct inode *dir, struct qstr *name,
 
 	if (path->len > NFS4_MAXPATHLEN)
 		return -ENAMETOOLONG;
+
 	arg.u.symlink = path;
-	nfs_fattr_init(fattr);
+	nfs_fattr_init(&fattr);
 	nfs_fattr_init(&dir_fattr);
 	
 	status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0);
-	if (!status)
+	if (!status) {
 		update_changeattr(dir, &res.dir_cinfo);
-	nfs_post_op_update_inode(dir, res.dir_fattr);
+		nfs_post_op_update_inode(dir, res.dir_fattr);
+		status = nfs_instantiate(dentry, &fhandle, &fattr);
+	}
 	return status;
 }
 
-static int nfs4_proc_symlink(struct inode *dir, struct qstr *name,
-		struct qstr *path, struct iattr *sattr, struct nfs_fh *fhandle,
-		struct nfs_fattr *fattr)
+static int nfs4_proc_symlink(struct inode *dir, struct dentry *dentry,
+		struct qstr *path, struct iattr *sattr)
 {
 	struct nfs4_exception exception = { };
 	int err;
 	do {
 		err = nfs4_handle_exception(NFS_SERVER(dir),
-				_nfs4_proc_symlink(dir, name, path, sattr,
-					fhandle, fattr),
+				_nfs4_proc_symlink(dir, dentry, path, sattr),
 				&exception);
 	} while (exception.retry);
 	return err;
diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c
index 5a8b9407ee9a..0b507bf0f330 100644
--- a/fs/nfs/proc.c
+++ b/fs/nfs/proc.c
@@ -425,14 +425,15 @@ nfs_proc_link(struct inode *inode, struct inode *dir, struct qstr *name)
 }
 
 static int
-nfs_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path,
-		 struct iattr *sattr, struct nfs_fh *fhandle,
-		 struct nfs_fattr *fattr)
+nfs_proc_symlink(struct inode *dir, struct dentry *dentry, struct qstr *path,
+		 struct iattr *sattr)
 {
+	struct nfs_fh fhandle;
+	struct nfs_fattr fattr;
 	struct nfs_symlinkargs	arg = {
 		.fromfh		= NFS_FH(dir),
-		.fromname	= name->name,
-		.fromlen	= name->len,
+		.fromname	= dentry->d_name.name,
+		.fromlen	= dentry->d_name.len,
 		.topath		= path->name,
 		.tolen		= path->len,
 		.sattr		= sattr
@@ -445,11 +446,23 @@ nfs_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path,
 
 	if (path->len > NFS2_MAXPATHLEN)
 		return -ENAMETOOLONG;
-	dprintk("NFS call  symlink %s -> %s\n", name->name, path->name);
-	nfs_fattr_init(fattr);
-	fhandle->size = 0;
+
+	dprintk("NFS call  symlink %s -> %s\n", dentry->d_name.name,
+			path->name);
 	status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0);
 	nfs_mark_for_revalidate(dir);
+
+	/*
+	 * V2 SYMLINK requests don't return any attributes.  Setting the
+	 * filehandle size to zero indicates to nfs_instantiate that it
+	 * should fill in the data with a LOOKUP call on the wire.
+	 */
+	if (status == 0) {
+		nfs_fattr_init(&fattr);
+		fhandle.size = 0;
+		status = nfs_instantiate(dentry, &fhandle, &fattr);
+	}
+
 	dprintk("NFS reply symlink: %d\n", status);
 	return status;
 }
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index 0f33e621892f..ddf5d75e97a2 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -793,9 +793,8 @@ struct nfs_rpc_ops {
 	int	(*rename)  (struct inode *, struct qstr *,
 			    struct inode *, struct qstr *);
 	int	(*link)    (struct inode *, struct inode *, struct qstr *);
-	int	(*symlink) (struct inode *, struct qstr *, struct qstr *,
-			    struct iattr *, struct nfs_fh *,
-			    struct nfs_fattr *);
+	int	(*symlink) (struct inode *, struct dentry *, struct qstr *,
+			    struct iattr *);
 	int	(*mkdir)   (struct inode *, struct dentry *, struct iattr *);
 	int	(*rmdir)   (struct inode *, struct qstr *);
 	int	(*readdir) (struct dentry *, struct rpc_cred *,
-- 
cgit v1.2.3


From 94a6d75320b3681e6e728b70e18bd186cb55e682 Mon Sep 17 00:00:00 2001
From: Chuck Lever <chuck.lever@oracle.com>
Date: Tue, 22 Aug 2006 20:06:23 -0400
Subject: NFS: Use cached page as buffer for NFS symlink requests

Now that we have a copy of the symlink path in the page cache, we can pass
a struct page down to the XDR routines instead of a string buffer.

Test plan:
Connectathon, all NFS versions.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/dir.c            |  8 +-------
 fs/nfs/nfs2xdr.c        | 21 ++++++++++++++++++---
 fs/nfs/nfs3proc.c       | 14 +++++++-------
 fs/nfs/nfs3xdr.c        |  7 +++++--
 fs/nfs/nfs4proc.c       | 12 +++++++-----
 fs/nfs/nfs4xdr.c        |  8 ++++----
 fs/nfs/proc.c           | 14 +++++++-------
 include/linux/nfs_xdr.h | 17 ++++++++++-------
 8 files changed, 59 insertions(+), 42 deletions(-)

(limited to 'include')

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index b483e5d206cb..51328ae640dd 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -1464,10 +1464,6 @@ static int nfs_symlink(struct inode *dir, struct dentry *dentry, const char *sym
 	char *kaddr;
 	struct iattr attr;
 	unsigned int pathlen = strlen(symname);
-	struct qstr qsymname = {
-		.name	= symname,
-		.len	= pathlen,
-	};
 	int error;
 
 	dfprintk(VFS, "NFS: symlink(%s/%ld, %s, %s)\n", dir->i_sb->s_id,
@@ -1493,10 +1489,8 @@ static int nfs_symlink(struct inode *dir, struct dentry *dentry, const char *sym
 		memset(kaddr + pathlen, 0, PAGE_SIZE - pathlen);
 	kunmap_atomic(kaddr, KM_USER0);
 
-	/* XXX: eventually this will pass in {page, pathlen},
-	 *	instead of qsymname; need XDR changes for that */
 	nfs_begin_data_update(dir);
-	error = NFS_PROTO(dir)->symlink(dir, dentry, &qsymname, &attr);
+	error = NFS_PROTO(dir)->symlink(dir, dentry, page, pathlen, &attr);
 	nfs_end_data_update(dir);
 	if (error != 0) {
 		dfprintk(VFS, "NFS: symlink(%s/%ld, %s, %s) error %d\n",
diff --git a/fs/nfs/nfs2xdr.c b/fs/nfs/nfs2xdr.c
index 67391eef6b93..b49501fc0a79 100644
--- a/fs/nfs/nfs2xdr.c
+++ b/fs/nfs/nfs2xdr.c
@@ -51,7 +51,7 @@
 #define NFS_createargs_sz	(NFS_diropargs_sz+NFS_sattr_sz)
 #define NFS_renameargs_sz	(NFS_diropargs_sz+NFS_diropargs_sz)
 #define NFS_linkargs_sz		(NFS_fhandle_sz+NFS_diropargs_sz)
-#define NFS_symlinkargs_sz	(NFS_diropargs_sz+NFS_path_sz+NFS_sattr_sz)
+#define NFS_symlinkargs_sz	(NFS_diropargs_sz+1+NFS_sattr_sz)
 #define NFS_readdirargs_sz	(NFS_fhandle_sz+2)
 
 #define NFS_attrstat_sz		(1+NFS_fattr_sz)
@@ -351,11 +351,26 @@ nfs_xdr_linkargs(struct rpc_rqst *req, u32 *p, struct nfs_linkargs *args)
 static int
 nfs_xdr_symlinkargs(struct rpc_rqst *req, u32 *p, struct nfs_symlinkargs *args)
 {
+	struct xdr_buf *sndbuf = &req->rq_snd_buf;
+	size_t pad;
+
 	p = xdr_encode_fhandle(p, args->fromfh);
 	p = xdr_encode_array(p, args->fromname, args->fromlen);
-	p = xdr_encode_array(p, args->topath, args->tolen);
+	*p++ = htonl(args->pathlen);
+	sndbuf->len = xdr_adjust_iovec(sndbuf->head, p);
+
+	xdr_encode_pages(sndbuf, args->pages, 0, args->pathlen);
+
+	/*
+	 * xdr_encode_pages may have added a few bytes to ensure the
+	 * pathname ends on a 4-byte boundary.  Start encoding the
+	 * attributes after the pad bytes.
+	 */
+	pad = sndbuf->tail->iov_len;
+	if (pad > 0)
+		p++;
 	p = xdr_encode_sattr(p, args->sattr);
-	req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
+	sndbuf->len += xdr_adjust_iovec(sndbuf->tail, p) - pad;
 	return 0;
 }
 
diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c
index d85ac427c326..f8688eaa0001 100644
--- a/fs/nfs/nfs3proc.c
+++ b/fs/nfs/nfs3proc.c
@@ -544,8 +544,8 @@ nfs3_proc_link(struct inode *inode, struct inode *dir, struct qstr *name)
 }
 
 static int
-nfs3_proc_symlink(struct inode *dir, struct dentry *dentry, struct qstr *path,
-		  struct iattr *sattr)
+nfs3_proc_symlink(struct inode *dir, struct dentry *dentry, struct page *page,
+		  unsigned int len, struct iattr *sattr)
 {
 	struct nfs_fh fhandle;
 	struct nfs_fattr fattr, dir_attr;
@@ -553,8 +553,8 @@ nfs3_proc_symlink(struct inode *dir, struct dentry *dentry, struct qstr *path,
 		.fromfh		= NFS_FH(dir),
 		.fromname	= dentry->d_name.name,
 		.fromlen	= dentry->d_name.len,
-		.topath		= path->name,
-		.tolen		= path->len,
+		.pages		= &page,
+		.pathlen	= len,
 		.sattr		= sattr
 	};
 	struct nfs3_diropres	res = {
@@ -569,11 +569,11 @@ nfs3_proc_symlink(struct inode *dir, struct dentry *dentry, struct qstr *path,
 	};
 	int			status;
 
-	if (path->len > NFS3_MAXPATHLEN)
+	if (len > NFS3_MAXPATHLEN)
 		return -ENAMETOOLONG;
 
-	dprintk("NFS call  symlink %s -> %s\n", dentry->d_name.name,
-			path->name);
+	dprintk("NFS call  symlink %s\n", dentry->d_name.name);
+
 	nfs_fattr_init(&dir_attr);
 	nfs_fattr_init(&fattr);
 	status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0);
diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c
index 0250269e9753..16556fa4effb 100644
--- a/fs/nfs/nfs3xdr.c
+++ b/fs/nfs/nfs3xdr.c
@@ -56,7 +56,7 @@
 #define NFS3_writeargs_sz	(NFS3_fh_sz+5)
 #define NFS3_createargs_sz	(NFS3_diropargs_sz+NFS3_sattr_sz)
 #define NFS3_mkdirargs_sz	(NFS3_diropargs_sz+NFS3_sattr_sz)
-#define NFS3_symlinkargs_sz	(NFS3_diropargs_sz+NFS3_path_sz+NFS3_sattr_sz)
+#define NFS3_symlinkargs_sz	(NFS3_diropargs_sz+1+NFS3_sattr_sz)
 #define NFS3_mknodargs_sz	(NFS3_diropargs_sz+2+NFS3_sattr_sz)
 #define NFS3_renameargs_sz	(NFS3_diropargs_sz+NFS3_diropargs_sz)
 #define NFS3_linkargs_sz		(NFS3_fh_sz+NFS3_diropargs_sz)
@@ -398,8 +398,11 @@ nfs3_xdr_symlinkargs(struct rpc_rqst *req, u32 *p, struct nfs3_symlinkargs *args
 	p = xdr_encode_fhandle(p, args->fromfh);
 	p = xdr_encode_array(p, args->fromname, args->fromlen);
 	p = xdr_encode_sattr(p, args->sattr);
-	p = xdr_encode_array(p, args->topath, args->tolen);
+	*p++ = htonl(args->pathlen);
 	req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
+
+	/* Copy the page */
+	xdr_encode_pages(&req->rq_snd_buf, args->pages, 0, args->pathlen);
 	return 0;
 }
 
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 2d18eac6bee5..7f60beb40df3 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -2085,7 +2085,7 @@ static int nfs4_proc_link(struct inode *inode, struct inode *dir, struct qstr *n
 }
 
 static int _nfs4_proc_symlink(struct inode *dir, struct dentry *dentry,
-		struct qstr *path, struct iattr *sattr)
+		struct page *page, unsigned int len, struct iattr *sattr)
 {
 	struct nfs_server *server = NFS_SERVER(dir);
 	struct nfs_fh fhandle;
@@ -2111,10 +2111,11 @@ static int _nfs4_proc_symlink(struct inode *dir, struct dentry *dentry,
 	};
 	int			status;
 
-	if (path->len > NFS4_MAXPATHLEN)
+	if (len > NFS4_MAXPATHLEN)
 		return -ENAMETOOLONG;
 
-	arg.u.symlink = path;
+	arg.u.symlink.pages = &page;
+	arg.u.symlink.len = len;
 	nfs_fattr_init(&fattr);
 	nfs_fattr_init(&dir_fattr);
 	
@@ -2128,13 +2129,14 @@ static int _nfs4_proc_symlink(struct inode *dir, struct dentry *dentry,
 }
 
 static int nfs4_proc_symlink(struct inode *dir, struct dentry *dentry,
-		struct qstr *path, struct iattr *sattr)
+		struct page *page, unsigned int len, struct iattr *sattr)
 {
 	struct nfs4_exception exception = { };
 	int err;
 	do {
 		err = nfs4_handle_exception(NFS_SERVER(dir),
-				_nfs4_proc_symlink(dir, dentry, path, sattr),
+				_nfs4_proc_symlink(dir, dentry, page,
+							len, sattr),
 				&exception);
 	} while (exception.retry);
 	return err;
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index 99926067eca4..3dd413f52da1 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -128,7 +128,7 @@ static int nfs4_stat_to_errno(int);
 #define decode_link_maxsz	(op_decode_hdr_maxsz + 5)
 #define encode_symlink_maxsz	(op_encode_hdr_maxsz + \
 				1 + nfs4_name_maxsz + \
-				nfs4_path_maxsz + \
+				1 + \
 				nfs4_fattr_maxsz)
 #define decode_symlink_maxsz	(op_decode_hdr_maxsz + 8)
 #define encode_create_maxsz	(op_encode_hdr_maxsz + \
@@ -673,9 +673,9 @@ static int encode_create(struct xdr_stream *xdr, const struct nfs4_create_arg *c
 
 	switch (create->ftype) {
 	case NF4LNK:
-		RESERVE_SPACE(4 + create->u.symlink->len);
-		WRITE32(create->u.symlink->len);
-		WRITEMEM(create->u.symlink->name, create->u.symlink->len);
+		RESERVE_SPACE(4);
+		WRITE32(create->u.symlink.len);
+		xdr_write_pages(xdr, create->u.symlink.pages, 0, create->u.symlink.len);
 		break;
 
 	case NF4BLK: case NF4CHR:
diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c
index 0b507bf0f330..630e50647bbb 100644
--- a/fs/nfs/proc.c
+++ b/fs/nfs/proc.c
@@ -425,8 +425,8 @@ nfs_proc_link(struct inode *inode, struct inode *dir, struct qstr *name)
 }
 
 static int
-nfs_proc_symlink(struct inode *dir, struct dentry *dentry, struct qstr *path,
-		 struct iattr *sattr)
+nfs_proc_symlink(struct inode *dir, struct dentry *dentry, struct page *page,
+		 unsigned int len, struct iattr *sattr)
 {
 	struct nfs_fh fhandle;
 	struct nfs_fattr fattr;
@@ -434,8 +434,8 @@ nfs_proc_symlink(struct inode *dir, struct dentry *dentry, struct qstr *path,
 		.fromfh		= NFS_FH(dir),
 		.fromname	= dentry->d_name.name,
 		.fromlen	= dentry->d_name.len,
-		.topath		= path->name,
-		.tolen		= path->len,
+		.pages		= &page,
+		.pathlen	= len,
 		.sattr		= sattr
 	};
 	struct rpc_message msg = {
@@ -444,11 +444,11 @@ nfs_proc_symlink(struct inode *dir, struct dentry *dentry, struct qstr *path,
 	};
 	int			status;
 
-	if (path->len > NFS2_MAXPATHLEN)
+	if (len > NFS2_MAXPATHLEN)
 		return -ENAMETOOLONG;
 
-	dprintk("NFS call  symlink %s -> %s\n", dentry->d_name.name,
-			path->name);
+	dprintk("NFS call  symlink %s\n", dentry->d_name.name);
+
 	status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0);
 	nfs_mark_for_revalidate(dir);
 
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index ddf5d75e97a2..dc5397d9d23c 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -358,8 +358,8 @@ struct nfs_symlinkargs {
 	struct nfs_fh *		fromfh;
 	const char *		fromname;
 	unsigned int		fromlen;
-	const char *		topath;
-	unsigned int		tolen;
+	struct page **		pages;
+	unsigned int		pathlen;
 	struct iattr *		sattr;
 };
 
@@ -434,8 +434,8 @@ struct nfs3_symlinkargs {
 	struct nfs_fh *		fromfh;
 	const char *		fromname;
 	unsigned int		fromlen;
-	const char *		topath;
-	unsigned int		tolen;
+	struct page **		pages;
+	unsigned int		pathlen;
 	struct iattr *		sattr;
 };
 
@@ -533,7 +533,10 @@ struct nfs4_accessres {
 struct nfs4_create_arg {
 	u32				ftype;
 	union {
-		struct qstr *		symlink;    /* NF4LNK */
+		struct {
+			struct page **	pages;
+			unsigned int	len;
+		} symlink;   /* NF4LNK */
 		struct {
 			u32		specdata1;
 			u32		specdata2;
@@ -793,8 +796,8 @@ struct nfs_rpc_ops {
 	int	(*rename)  (struct inode *, struct qstr *,
 			    struct inode *, struct qstr *);
 	int	(*link)    (struct inode *, struct inode *, struct qstr *);
-	int	(*symlink) (struct inode *, struct dentry *, struct qstr *,
-			    struct iattr *);
+	int	(*symlink) (struct inode *, struct dentry *, struct page *,
+			    unsigned int, struct iattr *);
 	int	(*mkdir)   (struct inode *, struct dentry *, struct iattr *);
 	int	(*rmdir)   (struct inode *, struct qstr *);
 	int	(*readdir) (struct dentry *, struct rpc_cred *,
-- 
cgit v1.2.3


From 275a082fe9308e710324e26ccb5363c53d8fd45f Mon Sep 17 00:00:00 2001
From: Trond Myklebust <Trond.Myklebust@netapp.com>
Date: Tue, 22 Aug 2006 20:06:24 -0400
Subject: Add a real API for dealing with blk_congestion_wait()

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 block/ll_rw_blk.c         | 12 ++++++++++++
 fs/nfs/write.c            |  1 +
 include/linux/blkdev.h    |  1 +
 include/linux/writeback.h |  1 +
 mm/page-writeback.c       |  9 +++++++++
 5 files changed, 24 insertions(+)

(limited to 'include')

diff --git a/block/ll_rw_blk.c b/block/ll_rw_blk.c
index ddd9253f9d55..dcbd6ff1fa33 100644
--- a/block/ll_rw_blk.c
+++ b/block/ll_rw_blk.c
@@ -2734,6 +2734,18 @@ long blk_congestion_wait(int rw, long timeout)
 
 EXPORT_SYMBOL(blk_congestion_wait);
 
+/**
+ * blk_congestion_end - wake up sleepers on a congestion queue
+ * @rw: READ or WRITE
+ */
+void blk_congestion_end(int rw)
+{
+	wait_queue_head_t *wqh = &congestion_wqh[rw];
+
+	if (waitqueue_active(wqh))
+		wake_up(wqh);
+}
+
 /*
  * Has to be called with the request spinlock acquired
  */
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index 453d44666ea5..38ba5c09af08 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -396,6 +396,7 @@ int nfs_writepages(struct address_space *mapping, struct writeback_control *wbc)
 out:
 	clear_bit(BDI_write_congested, &bdi->state);
 	wake_up_all(&nfs_write_congestion);
+	writeback_congestion_end();
 	return err;
 }
 
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index aafe82788b4e..96c9040c00a8 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -746,6 +746,7 @@ extern void blk_queue_free_tags(request_queue_t *);
 extern int blk_queue_resize_tags(request_queue_t *, int);
 extern void blk_queue_invalidate_tags(request_queue_t *);
 extern long blk_congestion_wait(int rw, long timeout);
+extern void blk_congestion_end(int rw);
 
 extern void blk_rq_bio_prep(request_queue_t *, struct request *, struct bio *);
 extern int blkdev_issue_flush(struct block_device *, sector_t *);
diff --git a/include/linux/writeback.h b/include/linux/writeback.h
index 9e38b566d0e7..0422036af4eb 100644
--- a/include/linux/writeback.h
+++ b/include/linux/writeback.h
@@ -85,6 +85,7 @@ int wakeup_pdflush(long nr_pages);
 void laptop_io_completion(void);
 void laptop_sync_completion(void);
 void throttle_vm_writeout(void);
+void writeback_congestion_end(void);
 
 /* These are exported to sysctl. */
 extern int dirty_background_ratio;
diff --git a/mm/page-writeback.c b/mm/page-writeback.c
index e630188ccc40..77a0bc4e261a 100644
--- a/mm/page-writeback.c
+++ b/mm/page-writeback.c
@@ -802,6 +802,15 @@ int test_set_page_writeback(struct page *page)
 }
 EXPORT_SYMBOL(test_set_page_writeback);
 
+/*
+ * Wakes up tasks that are being throttled due to writeback congestion
+ */
+void writeback_congestion_end(void)
+{
+	blk_congestion_end(WRITE);
+}
+EXPORT_SYMBOL(writeback_congestion_end);
+
 /*
  * Return true if any of the pages in the mapping are marged with the
  * passed tag.
-- 
cgit v1.2.3


From 5dd3177ae5012c1e2ad7a9ffdbd0e0d0de2f60e4 Mon Sep 17 00:00:00 2001
From: Trond Myklebust <Trond.Myklebust@netapp.com>
Date: Thu, 24 Aug 2006 01:03:05 -0400
Subject: NFSv4: Fix a use-after-free issue with the nfs server.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/client.c           | 36 +++++++++++++++++++++---------------
 fs/nfs/nfs4renewd.c       |  1 +
 fs/nfs/super.c            |  8 +++++---
 include/linux/nfs_fs_sb.h |  1 +
 4 files changed, 28 insertions(+), 18 deletions(-)

(limited to 'include')

diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index 12941a8a6d75..f1ff2aec2ca5 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -164,6 +164,26 @@ error_0:
 	return NULL;
 }
 
+static void nfs4_shutdown_client(struct nfs_client *clp)
+{
+#ifdef CONFIG_NFS_V4
+	if (__test_and_clear_bit(NFS_CS_RENEWD, &clp->cl_res_state))
+		nfs4_kill_renewd(clp);
+	while (!list_empty(&clp->cl_unused)) {
+		struct nfs4_state_owner *sp;
+
+		sp = list_entry(clp->cl_unused.next,
+				struct nfs4_state_owner,
+				so_list);
+		list_del(&sp->so_list);
+		kfree(sp);
+	}
+	BUG_ON(!list_empty(&clp->cl_state_owners));
+	if (__test_and_clear_bit(NFS_CS_IDMAP, &clp->cl_res_state))
+		nfs_idmap_delete(clp);
+#endif
+}
+
 /*
  * Destroy a shared client record
  */
@@ -171,21 +191,7 @@ static void nfs_free_client(struct nfs_client *clp)
 {
 	dprintk("--> nfs_free_client(%d)\n", clp->cl_nfsversion);
 
-#ifdef CONFIG_NFS_V4
-	if (__test_and_clear_bit(NFS_CS_IDMAP, &clp->cl_res_state)) {
-		while (!list_empty(&clp->cl_unused)) {
-			struct nfs4_state_owner *sp;
-
-			sp = list_entry(clp->cl_unused.next,
-					struct nfs4_state_owner,
-					so_list);
-			list_del(&sp->so_list);
-			kfree(sp);
-		}
-		BUG_ON(!list_empty(&clp->cl_state_owners));
-		nfs_idmap_delete(clp);
-	}
-#endif
+	nfs4_shutdown_client(clp);
 
 	/* -EIO all pending I/O */
 	if (!IS_ERR(clp->cl_rpcclient))
diff --git a/fs/nfs/nfs4renewd.c b/fs/nfs/nfs4renewd.c
index f2c893690ac4..7b6df1852e75 100644
--- a/fs/nfs/nfs4renewd.c
+++ b/fs/nfs/nfs4renewd.c
@@ -121,6 +121,7 @@ nfs4_schedule_state_renewal(struct nfs_client *clp)
 			__FUNCTION__, (timeout + HZ - 1) / HZ);
 	cancel_delayed_work(&clp->cl_renewd);
 	schedule_delayed_work(&clp->cl_renewd, timeout);
+	set_bit(NFS_CS_RENEWD, &clp->cl_res_state);
 	spin_unlock(&clp->cl_lock);
 }
 
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 97cfb143e09f..665949d27798 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -883,13 +883,15 @@ static int nfs4_get_sb(struct file_system_type *fs_type,
 		goto out_free;
 	}
 
+	if (s->s_fs_info != server) {
+		nfs_free_server(server);
+		server = NULL;
+	}
+
 	if (!s->s_root) {
 		/* initial superblock/root creation */
 		s->s_flags = flags;
-
 		nfs4_fill_super(s);
-	} else {
-		nfs_free_server(server);
 	}
 
 	mntroot = nfs4_get_root(s, &mntfh);
diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index 6d0be0efd1b5..7ccfc7ef0a83 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -19,6 +19,7 @@ struct nfs_client {
 #define NFS_CS_RPCIOD		0		/* - rpciod started */
 #define NFS_CS_CALLBACK		1		/* - callback started */
 #define NFS_CS_IDMAP		2		/* - idmap started */
+#define NFS_CS_RENEWD		3		/* - renewd started */
 	struct sockaddr_in	cl_addr;	/* server identifier */
 	char *			cl_hostname;	/* hostname of server */
 	struct list_head	cl_share_link;	/* link in global client list */
-- 
cgit v1.2.3


From 158998b6fe36f6acef087f574c96d44713499cc9 Mon Sep 17 00:00:00 2001
From: Trond Myklebust <Trond.Myklebust@netapp.com>
Date: Thu, 24 Aug 2006 01:03:17 -0400
Subject: SUNRPC: Make rpc_mkpipe() take the parent dentry as an argument

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/idmap.c                     |  6 +-----
 include/linux/sunrpc/rpc_pipe_fs.h |  2 +-
 net/sunrpc/auth_gss/auth_gss.c     |  7 ++-----
 net/sunrpc/rpc_pipe.c              | 38 +++++++++++++++++++++++---------------
 4 files changed, 27 insertions(+), 26 deletions(-)

(limited to 'include')

diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c
index f96dfac7dc9a..82ad7110a1c0 100644
--- a/fs/nfs/idmap.c
+++ b/fs/nfs/idmap.c
@@ -84,7 +84,6 @@ struct idmap_hashtable {
 };
 
 struct idmap {
-	char                  idmap_path[48];
 	struct dentry        *idmap_dentry;
 	wait_queue_head_t     idmap_wq;
 	struct idmap_msg      idmap_im;
@@ -119,10 +118,7 @@ nfs_idmap_new(struct nfs_client *clp)
         if ((idmap = kzalloc(sizeof(*idmap), GFP_KERNEL)) == NULL)
                 return -ENOMEM;
 
-	snprintf(idmap->idmap_path, sizeof(idmap->idmap_path),
-	    "%s/idmap", clp->cl_rpcclient->cl_pathname);
-
-        idmap->idmap_dentry = rpc_mkpipe(idmap->idmap_path,
+        idmap->idmap_dentry = rpc_mkpipe(clp->cl_rpcclient->cl_dentry, "idmap",
 	    idmap, &idmap_upcall_ops, 0);
         if (IS_ERR(idmap->idmap_dentry)) {
 		error = PTR_ERR(idmap->idmap_dentry);
diff --git a/include/linux/sunrpc/rpc_pipe_fs.h b/include/linux/sunrpc/rpc_pipe_fs.h
index a481472c9484..a2eb9b4a9de3 100644
--- a/include/linux/sunrpc/rpc_pipe_fs.h
+++ b/include/linux/sunrpc/rpc_pipe_fs.h
@@ -43,7 +43,7 @@ extern int rpc_queue_upcall(struct inode *, struct rpc_pipe_msg *);
 
 extern struct dentry *rpc_mkdir(char *, struct rpc_clnt *);
 extern int rpc_rmdir(struct dentry *);
-extern struct dentry *rpc_mkpipe(char *, void *, struct rpc_pipe_ops *, int flags);
+extern struct dentry *rpc_mkpipe(struct dentry *, const char *, void *, struct rpc_pipe_ops *, int flags);
 extern int rpc_unlink(struct dentry *);
 extern struct vfsmount *rpc_get_mount(void);
 extern void rpc_put_mount(void);
diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c
index ef1cf5b476c8..6eed3e166ba3 100644
--- a/net/sunrpc/auth_gss/auth_gss.c
+++ b/net/sunrpc/auth_gss/auth_gss.c
@@ -88,7 +88,6 @@ struct gss_auth {
 	struct list_head upcalls;
 	struct rpc_clnt *client;
 	struct dentry *dentry;
-	char path[48];
 	spinlock_t lock;
 };
 
@@ -690,10 +689,8 @@ gss_create(struct rpc_clnt *clnt, rpc_authflavor_t flavor)
 	if (err)
 		goto err_put_mech;
 
-	snprintf(gss_auth->path, sizeof(gss_auth->path), "%s/%s",
-			clnt->cl_pathname,
-			gss_auth->mech->gm_name);
-	gss_auth->dentry = rpc_mkpipe(gss_auth->path, clnt, &gss_upcall_ops, RPC_PIPE_WAIT_FOR_OPEN);
+	gss_auth->dentry = rpc_mkpipe(clnt->cl_dentry, gss_auth->mech->gm_name,
+			clnt, &gss_upcall_ops, RPC_PIPE_WAIT_FOR_OPEN);
 	if (IS_ERR(gss_auth->dentry)) {
 		err = PTR_ERR(gss_auth->dentry);
 		goto err_put_mech;
diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c
index c21dc07f2a8c..11ec12a09d70 100644
--- a/net/sunrpc/rpc_pipe.c
+++ b/net/sunrpc/rpc_pipe.c
@@ -621,17 +621,13 @@ __rpc_rmdir(struct inode *dir, struct dentry *dentry)
 }
 
 static struct dentry *
-rpc_lookup_negative(char *path, struct nameidata *nd)
+rpc_lookup_create(struct dentry *parent, const char *name, int len)
 {
+	struct inode *dir = parent->d_inode;
 	struct dentry *dentry;
-	struct inode *dir;
-	int error;
 
-	if ((error = rpc_lookup_parent(path, nd)) != 0)
-		return ERR_PTR(error);
-	dir = nd->dentry->d_inode;
 	mutex_lock_nested(&dir->i_mutex, I_MUTEX_PARENT);
-	dentry = lookup_one_len(nd->last.name, nd->dentry, nd->last.len);
+	dentry = lookup_one_len(name, parent, len);
 	if (IS_ERR(dentry))
 		goto out_err;
 	if (dentry->d_inode) {
@@ -642,7 +638,20 @@ rpc_lookup_negative(char *path, struct nameidata *nd)
 	return dentry;
 out_err:
 	mutex_unlock(&dir->i_mutex);
-	rpc_release_path(nd);
+	return dentry;
+}
+
+static struct dentry *
+rpc_lookup_negative(char *path, struct nameidata *nd)
+{
+	struct dentry *dentry;
+	int error;
+
+	if ((error = rpc_lookup_parent(path, nd)) != 0)
+		return ERR_PTR(error);
+	dentry = rpc_lookup_create(nd->dentry, nd->last.name, nd->last.len);
+	if (IS_ERR(dentry))
+		rpc_release_path(nd);
 	return dentry;
 }
 
@@ -701,17 +710,16 @@ rpc_rmdir(struct dentry *dentry)
 }
 
 struct dentry *
-rpc_mkpipe(char *path, void *private, struct rpc_pipe_ops *ops, int flags)
+rpc_mkpipe(struct dentry *parent, const char *name, void *private, struct rpc_pipe_ops *ops, int flags)
 {
-	struct nameidata nd;
 	struct dentry *dentry;
 	struct inode *dir, *inode;
 	struct rpc_inode *rpci;
 
-	dentry = rpc_lookup_negative(path, &nd);
+	dentry = rpc_lookup_create(parent, name, strlen(name));
 	if (IS_ERR(dentry))
 		return dentry;
-	dir = nd.dentry->d_inode;
+	dir = parent->d_inode;
 	inode = rpc_get_inode(dir->i_sb, S_IFSOCK | S_IRUSR | S_IWUSR);
 	if (!inode)
 		goto err_dput;
@@ -726,13 +734,13 @@ rpc_mkpipe(char *path, void *private, struct rpc_pipe_ops *ops, int flags)
 	dget(dentry);
 out:
 	mutex_unlock(&dir->i_mutex);
-	rpc_release_path(&nd);
 	return dentry;
 err_dput:
 	dput(dentry);
 	dentry = ERR_PTR(-ENOMEM);
-	printk(KERN_WARNING "%s: %s() failed to create pipe %s (errno = %d)\n",
-			__FILE__, __FUNCTION__, path, -ENOMEM);
+	printk(KERN_WARNING "%s: %s() failed to create pipe %s/%s (errno = %d)\n",
+			__FILE__, __FUNCTION__, parent->d_name.name, name,
+			-ENOMEM);
 	goto out;
 }
 
-- 
cgit v1.2.3


From 6b6ca86b77b62b798cf9ca2599036420abce7796 Mon Sep 17 00:00:00 2001
From: Trond Myklebust <Trond.Myklebust@netapp.com>
Date: Tue, 5 Sep 2006 12:55:57 -0400
Subject: SUNRPC: Add refcounting to the struct rpc_xprt

In a subsequent patch, this will allow the portmapper to take a reference
to the rpc_xprt for which it is updating the port number, fixing an Oops.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 include/linux/sunrpc/xprt.h |  5 ++++-
 net/sunrpc/clnt.c           |  8 +++-----
 net/sunrpc/xprt.c           | 28 +++++++++++++++++++++++++---
 3 files changed, 32 insertions(+), 9 deletions(-)

(limited to 'include')

diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h
index de4efea7c856..bdeba8538c71 100644
--- a/include/linux/sunrpc/xprt.h
+++ b/include/linux/sunrpc/xprt.h
@@ -12,6 +12,7 @@
 #include <linux/uio.h>
 #include <linux/socket.h>
 #include <linux/in.h>
+#include <linux/kref.h>
 #include <linux/sunrpc/sched.h>
 #include <linux/sunrpc/xdr.h>
 
@@ -129,6 +130,7 @@ struct rpc_xprt_ops {
 };
 
 struct rpc_xprt {
+	struct kref		kref;		/* Reference count */
 	struct rpc_xprt_ops *	ops;		/* transport methods */
 	struct socket *		sock;		/* BSD socket layer */
 	struct sock *		inet;		/* INET layer */
@@ -248,7 +250,8 @@ int			xprt_adjust_timeout(struct rpc_rqst *req);
 void			xprt_release_xprt(struct rpc_xprt *xprt, struct rpc_task *task);
 void			xprt_release_xprt_cong(struct rpc_xprt *xprt, struct rpc_task *task);
 void			xprt_release(struct rpc_task *task);
-int			xprt_destroy(struct rpc_xprt *xprt);
+struct rpc_xprt *	xprt_get(struct rpc_xprt *xprt);
+void			xprt_put(struct rpc_xprt *xprt);
 
 static inline u32 *xprt_skip_transport_header(struct rpc_xprt *xprt, u32 *p)
 {
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index ceadb728f0da..084a0ad5c64e 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -177,7 +177,7 @@ out_no_path:
 		kfree(clnt->cl_server);
 	kfree(clnt);
 out_err:
-	xprt_destroy(xprt);
+	xprt_put(xprt);
 out_no_xprt:
 	return ERR_PTR(err);
 }
@@ -261,6 +261,7 @@ rpc_clone_client(struct rpc_clnt *clnt)
 	atomic_set(&new->cl_users, 0);
 	new->cl_parent = clnt;
 	atomic_inc(&clnt->cl_count);
+	new->cl_xprt = xprt_get(clnt->cl_xprt);
 	/* Turn off autobind on clones */
 	new->cl_autobind = 0;
 	new->cl_oneshot = 0;
@@ -337,15 +338,12 @@ rpc_destroy_client(struct rpc_clnt *clnt)
 		rpc_rmdir(clnt->cl_dentry);
 		rpc_put_mount();
 	}
-	if (clnt->cl_xprt) {
-		xprt_destroy(clnt->cl_xprt);
-		clnt->cl_xprt = NULL;
-	}
 	if (clnt->cl_server != clnt->cl_inline_name)
 		kfree(clnt->cl_server);
 out_free:
 	rpc_free_iostats(clnt->cl_metrics);
 	clnt->cl_metrics = NULL;
+	xprt_put(clnt->cl_xprt);
 	kfree(clnt);
 	return 0;
 }
diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index a85f82baefc1..1f786f68729d 100644
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -926,6 +926,7 @@ struct rpc_xprt *xprt_create_transport(int proto, struct sockaddr *ap, size_t si
 		return ERR_PTR(result);
 	}
 
+	kref_init(&xprt->kref);
 	spin_lock_init(&xprt->transport_lock);
 	spin_lock_init(&xprt->reserve_lock);
 
@@ -958,16 +959,37 @@ struct rpc_xprt *xprt_create_transport(int proto, struct sockaddr *ap, size_t si
 
 /**
  * xprt_destroy - destroy an RPC transport, killing off all requests.
- * @xprt: transport to destroy
+ * @kref: kref for the transport to destroy
  *
  */
-int xprt_destroy(struct rpc_xprt *xprt)
+static void xprt_destroy(struct kref *kref)
 {
+	struct rpc_xprt *xprt = container_of(kref, struct rpc_xprt, kref);
+
 	dprintk("RPC:      destroying transport %p\n", xprt);
 	xprt->shutdown = 1;
 	del_timer_sync(&xprt->timer);
 	xprt->ops->destroy(xprt);
 	kfree(xprt);
+}
 
-	return 0;
+/**
+ * xprt_put - release a reference to an RPC transport.
+ * @xprt: pointer to the transport
+ *
+ */
+void xprt_put(struct rpc_xprt *xprt)
+{
+	kref_put(&xprt->kref, xprt_destroy);
+}
+
+/**
+ * xprt_get - return a reference to an RPC transport.
+ * @xprt: pointer to the transport
+ *
+ */
+struct rpc_xprt *xprt_get(struct rpc_xprt *xprt)
+{
+	kref_get(&xprt->kref);
+	return xprt;
 }
-- 
cgit v1.2.3


From 42750b04c5baa7c5ffdf0a8be2b9b320efdf069f Mon Sep 17 00:00:00 2001
From: Jaroslav Kysela <perex@suse.cz>
Date: Thu, 1 Jun 2006 18:34:01 +0200
Subject: [ALSA] Control API - TLV implementation for additional information
 like dB scale

This patch implements a TLV mechanism to transfer an additional information
like dB scale to the user space. The types might be extended in future.
Acked-by: Takashi Iwai <tiwai@suse.de>

Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 include/sound/asound.h          |  9 ++++++++-
 include/sound/control.h         |  2 ++
 include/sound/tlv.h             | 43 +++++++++++++++++++++++++++++++++++++++++
 sound/core/control.c            | 41 +++++++++++++++++++++++++++++++++++++--
 sound/pci/ca0106/ca0106_mixer.c |  4 ++++
 5 files changed, 96 insertions(+), 3 deletions(-)
 create mode 100644 include/sound/tlv.h

(limited to 'include')

diff --git a/include/sound/asound.h b/include/sound/asound.h
index 41885f48ad91..76a20406bd18 100644
--- a/include/sound/asound.h
+++ b/include/sound/asound.h
@@ -688,7 +688,7 @@ struct snd_timer_tread {
  *                                                                          *
  ****************************************************************************/
 
-#define SNDRV_CTL_VERSION		SNDRV_PROTOCOL_VERSION(2, 0, 3)
+#define SNDRV_CTL_VERSION		SNDRV_PROTOCOL_VERSION(2, 0, 4)
 
 struct snd_ctl_card_info {
 	int card;			/* card number */
@@ -818,6 +818,12 @@ struct snd_ctl_elem_value {
         unsigned char reserved[128-sizeof(struct timespec)];
 };
 
+struct snd_ctl_tlv {
+        unsigned int numid;	/* control element numeric identification */
+        unsigned int length;	/* in bytes aligned to 4 */
+        unsigned int tlv[0];	/* first TLV */
+};
+
 enum {
 	SNDRV_CTL_IOCTL_PVERSION = _IOR('U', 0x00, int),
 	SNDRV_CTL_IOCTL_CARD_INFO = _IOR('U', 0x01, struct snd_ctl_card_info),
@@ -831,6 +837,7 @@ enum {
 	SNDRV_CTL_IOCTL_ELEM_ADD = _IOWR('U', 0x17, struct snd_ctl_elem_info),
 	SNDRV_CTL_IOCTL_ELEM_REPLACE = _IOWR('U', 0x18, struct snd_ctl_elem_info),
 	SNDRV_CTL_IOCTL_ELEM_REMOVE = _IOWR('U', 0x19, struct snd_ctl_elem_id),
+	SNDRV_CTL_IOCTL_TLV_READ = _IOWR('U', 0x1a, struct snd_ctl_tlv),
 	SNDRV_CTL_IOCTL_HWDEP_NEXT_DEVICE = _IOWR('U', 0x20, int),
 	SNDRV_CTL_IOCTL_HWDEP_INFO = _IOR('U', 0x21, struct snd_hwdep_info),
 	SNDRV_CTL_IOCTL_PCM_NEXT_DEVICE = _IOR('U', 0x30, int),
diff --git a/include/sound/control.h b/include/sound/control.h
index 2489b1eb0110..a93a58d0e688 100644
--- a/include/sound/control.h
+++ b/include/sound/control.h
@@ -42,6 +42,7 @@ struct snd_kcontrol_new {
 	snd_kcontrol_info_t *info;
 	snd_kcontrol_get_t *get;
 	snd_kcontrol_put_t *put;
+	unsigned int *tlv;
 	unsigned long private_value;
 };
 
@@ -58,6 +59,7 @@ struct snd_kcontrol {
 	snd_kcontrol_info_t *info;
 	snd_kcontrol_get_t *get;
 	snd_kcontrol_put_t *put;
+	unsigned int *tlv;
 	unsigned long private_value;
 	void *private_data;
 	void (*private_free)(struct snd_kcontrol *kcontrol);
diff --git a/include/sound/tlv.h b/include/sound/tlv.h
new file mode 100644
index 000000000000..b826e1df1da6
--- /dev/null
+++ b/include/sound/tlv.h
@@ -0,0 +1,43 @@
+#ifndef __SOUND_TLV_H
+#define __SOUND_TLV_H
+
+/*
+ *  Advanced Linux Sound Architecture - ALSA - Driver
+ *  Copyright (c) 2006 by Jaroslav Kysela <perex@suse.cz>
+ *
+ *
+ *   This program is free software; you can redistribute it and/or modify
+ *   it under the terms of the GNU General Public License as published by
+ *   the Free Software Foundation; either version 2 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program is distributed in the hope that it will be useful,
+ *   but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *   GNU General Public License for more details.
+ *
+ *   You should have received a copy of the GNU General Public License
+ *   along with this program; if not, write to the Free Software
+ *   Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
+ *
+ */
+
+/*
+ * TLV structure is right behind the struct snd_ctl_tlv:
+ *   unsigned int type  	- see SNDRV_CTL_TLVT_*
+ *   unsigned int length
+ *   .... data aligned to sizeof(unsigned int), use
+ *        block_length = (length + (sizeof(unsigned int) - 1)) &
+ *                       ~(sizeof(unsigned int) - 1)) ....
+ */
+
+#define SNDRV_CTL_TLVT_CONTAINER 0	/* one level down - group of TLVs */
+#define SNDRV_CTL_TLVT_DB_SCALE	1       /* dB scale */
+
+#define DECLARE_TLV_DB_SCALE(name, min, step, mute) \
+unsigned int name[] = { \
+        SNDRV_CTL_TLVT_DB_SCALE, 2 * sizeof(unsigned int), \
+        (min), ((step) & 0xffff) | ((mute) ? 0x10000 : 0) \
+}
+
+#endif /* __SOUND_TLV_H */
diff --git a/sound/core/control.c b/sound/core/control.c
index bb397eaa7187..e9c8854d2f7b 100644
--- a/sound/core/control.c
+++ b/sound/core/control.c
@@ -241,6 +241,7 @@ struct snd_kcontrol *snd_ctl_new1(const struct snd_kcontrol_new *ncontrol,
 	kctl.info = ncontrol->info;
 	kctl.get = ncontrol->get;
 	kctl.put = ncontrol->put;
+	kctl.tlv = ncontrol->tlv;
 	kctl.private_value = ncontrol->private_value;
 	kctl.private_data = private_data;
 	return snd_ctl_new(&kctl, access);
@@ -1067,6 +1068,40 @@ static int snd_ctl_subscribe_events(struct snd_ctl_file *file, int __user *ptr)
 	return 0;
 }
 
+static int snd_ctl_tlv_read(struct snd_card *card,
+                            struct snd_ctl_tlv __user *_tlv)
+{
+	struct snd_ctl_tlv tlv;
+	struct snd_kcontrol *kctl;
+	unsigned int len;
+	int err = 0;
+
+	if (copy_from_user(&tlv, _tlv, sizeof(tlv)))
+		return -EFAULT;
+        if (tlv.length < sizeof(unsigned int) * 3)
+                return -EINVAL;
+        down_read(&card->controls_rwsem);
+        kctl = snd_ctl_find_numid(card, tlv.numid);
+        if (kctl == NULL) {
+                err = -ENOENT;
+                goto __kctl_end;
+        }
+        if (kctl->tlv == NULL) {
+                err = -ENXIO;
+                goto __kctl_end;
+        }
+        len = kctl->tlv[1] + 2 * sizeof(unsigned int);
+        if (tlv.length < len) {
+                err = -ENOMEM;
+                goto __kctl_end;
+        }
+        if (copy_to_user(_tlv->tlv, kctl->tlv, len))
+        	err = -EFAULT;
+      __kctl_end:
+        up_read(&card->controls_rwsem);
+        return err;
+}
+
 static long snd_ctl_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 {
 	struct snd_ctl_file *ctl;
@@ -1086,11 +1121,11 @@ static long snd_ctl_ioctl(struct file *file, unsigned int cmd, unsigned long arg
 	case SNDRV_CTL_IOCTL_CARD_INFO:
 		return snd_ctl_card_info(card, ctl, cmd, argp);
 	case SNDRV_CTL_IOCTL_ELEM_LIST:
-		return snd_ctl_elem_list(ctl->card, argp);
+		return snd_ctl_elem_list(card, argp);
 	case SNDRV_CTL_IOCTL_ELEM_INFO:
 		return snd_ctl_elem_info_user(ctl, argp);
 	case SNDRV_CTL_IOCTL_ELEM_READ:
-		return snd_ctl_elem_read_user(ctl->card, argp);
+		return snd_ctl_elem_read_user(card, argp);
 	case SNDRV_CTL_IOCTL_ELEM_WRITE:
 		return snd_ctl_elem_write_user(ctl, argp);
 	case SNDRV_CTL_IOCTL_ELEM_LOCK:
@@ -1105,6 +1140,8 @@ static long snd_ctl_ioctl(struct file *file, unsigned int cmd, unsigned long arg
 		return snd_ctl_elem_remove(ctl, argp);
 	case SNDRV_CTL_IOCTL_SUBSCRIBE_EVENTS:
 		return snd_ctl_subscribe_events(ctl, ip);
+        case SNDRV_CTL_IOCTL_TLV_READ:
+                return snd_ctl_tlv_read(card, argp);
 	case SNDRV_CTL_IOCTL_POWER:
 		return -ENOPROTOOPT;
 	case SNDRV_CTL_IOCTL_POWER_STATE:
diff --git a/sound/pci/ca0106/ca0106_mixer.c b/sound/pci/ca0106/ca0106_mixer.c
index 146eed70dce6..35309b3ed8c0 100644
--- a/sound/pci/ca0106/ca0106_mixer.c
+++ b/sound/pci/ca0106/ca0106_mixer.c
@@ -70,9 +70,12 @@
 #include <sound/pcm.h>
 #include <sound/ac97_codec.h>
 #include <sound/info.h>
+#include <sound/tlv.h>
 
 #include "ca0106.h"
 
+static DECLARE_TLV_DB_SCALE(snd_ca0106_db_scale, -5150, 75, 1);
+
 static int snd_ca0106_shared_spdif_info(struct snd_kcontrol *kcontrol,
 					struct snd_ctl_elem_info *uinfo)
 {
@@ -472,6 +475,7 @@ static int snd_ca0106_i2c_volume_put(struct snd_kcontrol *kcontrol,
 	.info =	 snd_ca0106_volume_info,			\
 	.get =   snd_ca0106_volume_get,				\
 	.put =   snd_ca0106_volume_put,				\
+	.tlv =	 snd_ca0106_db_scale,				\
 	.private_value = ((chid) << 8) | (reg)			\
 }
 
-- 
cgit v1.2.3


From 746d4a02e68499fc6c1f8d0c43d2271853ade181 Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Fri, 23 Jun 2006 14:37:59 +0200
Subject: [ALSA] Fix disconnection of proc interface

- Add the linked list to each proc entry to enable a single-shot
  disconnection (unregister)
- Deprecate snd_info_unregister(), use snd_info_free_entry()
- Removed NULL checks of snd_info_free_entry()

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 include/sound/info.h                |   7 +--
 sound/core/hwdep.c                  |   2 +-
 sound/core/info.c                   | 102 ++++++++++++++++++------------------
 sound/core/info_oss.c               |   6 +--
 sound/core/init.c                   |   7 +--
 sound/core/oss/mixer_oss.c          |   6 +--
 sound/core/oss/pcm_oss.c            |   8 ++-
 sound/core/pcm.c                    |  53 +++++++------------
 sound/core/pcm_memory.c             |   2 +-
 sound/core/rawmidi.c                |   2 +-
 sound/core/seq/oss/seq_oss.c        |   3 +-
 sound/core/seq/seq_device.c         |   2 +-
 sound/core/seq/seq_info.c           |   6 +--
 sound/core/sound.c                  |   3 +-
 sound/core/sound_oss.c              |   3 +-
 sound/core/timer.c                  |   2 +-
 sound/drivers/opl4/opl4_proc.c      |   3 +-
 sound/pci/ac97/ac97_proc.c          |  18 +++----
 sound/pci/cs46xx/dsp_spos.c         |  52 +++++++-----------
 sound/pci/cs46xx/dsp_spos_scb_lib.c |   2 +-
 sound/synth/emux/emux_proc.c        |   6 +--
 21 files changed, 123 insertions(+), 172 deletions(-)

(limited to 'include')

diff --git a/include/sound/info.h b/include/sound/info.h
index 74f6996769c7..97ffc4fb9969 100644
--- a/include/sound/info.h
+++ b/include/sound/info.h
@@ -71,7 +71,6 @@ struct snd_info_entry {
 	mode_t mode;
 	long size;
 	unsigned short content;
-	unsigned short disconnected: 1;
 	union {
 		struct snd_info_entry_text text;
 		struct snd_info_entry_ops *ops;
@@ -83,6 +82,8 @@ struct snd_info_entry {
 	void (*private_free)(struct snd_info_entry *entry);
 	struct proc_dir_entry *p;
 	struct mutex access;
+	struct list_head children;
+	struct list_head list;
 };
 
 #if defined(CONFIG_SND_OSSEMUL) && defined(CONFIG_PROC_FS)
@@ -122,8 +123,8 @@ int snd_info_restore_text(struct snd_info_entry * entry);
 int snd_info_card_create(struct snd_card * card);
 int snd_info_card_register(struct snd_card * card);
 int snd_info_card_free(struct snd_card * card);
+void snd_info_card_disconnect(struct snd_card * card);
 int snd_info_register(struct snd_info_entry * entry);
-int snd_info_unregister(struct snd_info_entry * entry);
 
 /* for card drivers */
 int snd_card_proc_new(struct snd_card *card, const char *name, struct snd_info_entry **entryp);
@@ -156,8 +157,8 @@ static inline void snd_info_free_entry(struct snd_info_entry * entry) { ; }
 static inline int snd_info_card_create(struct snd_card * card) { return 0; }
 static inline int snd_info_card_register(struct snd_card * card) { return 0; }
 static inline int snd_info_card_free(struct snd_card * card) { return 0; }
+static inline void snd_info_card_disconnect(struct snd_card * card) { }
 static inline int snd_info_register(struct snd_info_entry * entry) { return 0; }
-static inline int snd_info_unregister(struct snd_info_entry * entry) { return 0; }
 
 static inline int snd_card_proc_new(struct snd_card *card, const char *name,
 				    struct snd_info_entry **entryp) { return -EINVAL; }
diff --git a/sound/core/hwdep.c b/sound/core/hwdep.c
index 8bd0dcc93eba..cbd8a63282b6 100644
--- a/sound/core/hwdep.c
+++ b/sound/core/hwdep.c
@@ -497,7 +497,7 @@ static void __init snd_hwdep_proc_init(void)
 
 static void __exit snd_hwdep_proc_done(void)
 {
-	snd_info_unregister(snd_hwdep_proc_entry);
+	snd_info_free_entry(snd_hwdep_proc_entry);
 }
 #else /* !CONFIG_PROC_FS */
 #define snd_hwdep_proc_init()
diff --git a/sound/core/info.c b/sound/core/info.c
index 340332c6d973..9663b6be9c3a 100644
--- a/sound/core/info.c
+++ b/sound/core/info.c
@@ -78,6 +78,7 @@ struct snd_info_private_data {
 
 static int snd_info_version_init(void);
 static int snd_info_version_done(void);
+static void snd_info_disconnect(struct snd_info_entry *entry);
 
 
 /* resize the proc r/w buffer */
@@ -304,7 +305,7 @@ static int snd_info_entry_open(struct inode *inode, struct file *file)
 	mutex_lock(&info_mutex);
 	p = PDE(inode);
 	entry = p == NULL ? NULL : (struct snd_info_entry *)p->data;
-	if (entry == NULL || entry->disconnected) {
+	if (entry == NULL || ! entry->p) {
 		mutex_unlock(&info_mutex);
 		return -ENODEV;
 	}
@@ -586,10 +587,10 @@ int __exit snd_info_done(void)
 	snd_info_version_done();
 	if (snd_proc_root) {
 #if defined(CONFIG_SND_SEQUENCER) || defined(CONFIG_SND_SEQUENCER_MODULE)
-		snd_info_unregister(snd_seq_root);
+		snd_info_free_entry(snd_seq_root);
 #endif
 #ifdef CONFIG_SND_OSSEMUL
-		snd_info_unregister(snd_oss_root);
+		snd_info_free_entry(snd_oss_root);
 #endif
 		snd_remove_proc_entry(&proc_root, snd_proc_root);
 	}
@@ -648,17 +649,28 @@ int snd_info_card_register(struct snd_card *card)
  * de-register the card proc file
  * called from init.c
  */
-int snd_info_card_free(struct snd_card *card)
+void snd_info_card_disconnect(struct snd_card *card)
 {
-	snd_assert(card != NULL, return -ENXIO);
+	snd_assert(card != NULL, return);
+	mutex_lock(&info_mutex);
 	if (card->proc_root_link) {
 		snd_remove_proc_entry(snd_proc_root, card->proc_root_link);
 		card->proc_root_link = NULL;
 	}
-	if (card->proc_root) {
-		snd_info_unregister(card->proc_root);
-		card->proc_root = NULL;
-	}
+	if (card->proc_root)
+		snd_info_disconnect(card->proc_root);
+	mutex_unlock(&info_mutex);
+}
+
+/*
+ * release the card proc file resources
+ * called from init.c
+ */
+int snd_info_card_free(struct snd_card *card)
+{
+	snd_assert(card != NULL, return -ENXIO);
+	snd_info_free_entry(card->proc_root);
+	card->proc_root = NULL;
 	return 0;
 }
 
@@ -767,6 +779,8 @@ static struct snd_info_entry *snd_info_create_entry(const char *name)
 	entry->mode = S_IFREG | S_IRUGO;
 	entry->content = SNDRV_INFO_CONTENT_TEXT;
 	mutex_init(&entry->access);
+	INIT_LIST_HEAD(&entry->children);
+	INIT_LIST_HEAD(&entry->list);
 	return entry;
 }
 
@@ -819,30 +833,35 @@ struct snd_info_entry *snd_info_create_card_entry(struct snd_card *card,
 
 EXPORT_SYMBOL(snd_info_create_card_entry);
 
-static int snd_info_dev_free_entry(struct snd_device *device)
+static void snd_info_disconnect(struct snd_info_entry *entry)
 {
-	struct snd_info_entry *entry = device->device_data;
-	snd_info_free_entry(entry);
-	return 0;
-}
+	struct list_head *p, *n;
+	struct proc_dir_entry *root;
 
-static int snd_info_dev_register_entry(struct snd_device *device)
-{
-	struct snd_info_entry *entry = device->device_data;
-	return snd_info_register(entry);
+	list_for_each_safe(p, n, &entry->children) {
+		snd_info_disconnect(list_entry(p, struct snd_info_entry, list));
+	}
+
+	if (! entry->p)
+		return;
+	list_del_init(&entry->list);
+	root = entry->parent == NULL ? snd_proc_root : entry->parent->p;
+	snd_assert(root, return);
+	snd_remove_proc_entry(root, entry->p);
+	entry->p = NULL;
 }
 
-static int snd_info_dev_disconnect_entry(struct snd_device *device)
+static int snd_info_dev_free_entry(struct snd_device *device)
 {
 	struct snd_info_entry *entry = device->device_data;
-	entry->disconnected = 1;
+	snd_info_free_entry(entry);
 	return 0;
 }
 
-static int snd_info_dev_unregister_entry(struct snd_device *device)
+static int snd_info_dev_register_entry(struct snd_device *device)
 {
 	struct snd_info_entry *entry = device->device_data;
-	return snd_info_unregister(entry);
+	return snd_info_register(entry);
 }
 
 /**
@@ -871,8 +890,7 @@ int snd_card_proc_new(struct snd_card *card, const char *name,
 	static struct snd_device_ops ops = {
 		.dev_free = snd_info_dev_free_entry,
 		.dev_register =	snd_info_dev_register_entry,
-		.dev_disconnect = snd_info_dev_disconnect_entry,
-		.dev_unregister = snd_info_dev_unregister_entry
+		/* disconnect is done via snd_info_card_disconnect() */
 	};
 	struct snd_info_entry *entry;
 	int err;
@@ -901,6 +919,11 @@ void snd_info_free_entry(struct snd_info_entry * entry)
 {
 	if (entry == NULL)
 		return;
+	if (entry->p) {
+		mutex_lock(&info_mutex);
+		snd_info_disconnect(entry);
+		mutex_unlock(&info_mutex);
+	}
 	kfree(entry->name);
 	if (entry->private_free)
 		entry->private_free(entry);
@@ -935,38 +958,14 @@ int snd_info_register(struct snd_info_entry * entry)
 	p->size = entry->size;
 	p->data = entry;
 	entry->p = p;
+	if (entry->parent)
+		list_add_tail(&entry->list, &entry->parent->children);
 	mutex_unlock(&info_mutex);
 	return 0;
 }
 
 EXPORT_SYMBOL(snd_info_register);
 
-/**
- * snd_info_unregister - de-register the info entry
- * @entry: the info entry
- *
- * De-registers the info entry and releases the instance.
- *
- * Returns zero if successful, or a negative error code on failure.
- */
-int snd_info_unregister(struct snd_info_entry * entry)
-{
-	struct proc_dir_entry *root;
-
-	if (! entry)
-		return 0;
-	snd_assert(entry->p != NULL, return -ENXIO);
-	root = entry->parent == NULL ? snd_proc_root : entry->parent->p;
-	snd_assert(root, return -ENXIO);
-	mutex_lock(&info_mutex);
-	snd_remove_proc_entry(root, entry->p);
-	mutex_unlock(&info_mutex);
-	snd_info_free_entry(entry);
-	return 0;
-}
-
-EXPORT_SYMBOL(snd_info_unregister);
-
 /*
 
  */
@@ -999,8 +998,7 @@ static int __init snd_info_version_init(void)
 
 static int __exit snd_info_version_done(void)
 {
-	if (snd_info_version_entry)
-		snd_info_unregister(snd_info_version_entry);
+	snd_info_free_entry(snd_info_version_entry);
 	return 0;
 }
 
diff --git a/sound/core/info_oss.c b/sound/core/info_oss.c
index bb2c40d0ab66..3ebc34919c76 100644
--- a/sound/core/info_oss.c
+++ b/sound/core/info_oss.c
@@ -131,10 +131,8 @@ int snd_info_minor_register(void)
 
 int snd_info_minor_unregister(void)
 {
-	if (snd_sndstat_proc_entry) {
-		snd_info_unregister(snd_sndstat_proc_entry);
-		snd_sndstat_proc_entry = NULL;
-	}
+	snd_info_free_entry(snd_sndstat_proc_entry);
+	snd_sndstat_proc_entry = NULL;
 	return 0;
 }
 
diff --git a/sound/core/init.c b/sound/core/init.c
index 4d9258884e44..1ecb029ff4c9 100644
--- a/sound/core/init.c
+++ b/sound/core/init.c
@@ -310,6 +310,7 @@ int snd_card_disconnect(struct snd_card *card)
 	if (err < 0)
 		snd_printk(KERN_ERR "not all devices for card %i can be disconnected\n", card->number);
 
+	snd_info_card_disconnect(card);
 	return 0;	
 }
 
@@ -360,7 +361,7 @@ int snd_card_free(struct snd_card *card)
 	}
 	if (card->private_free)
 		card->private_free(card);
-	snd_info_unregister(card->proc_id);
+	snd_info_free_entry(card->proc_id);
 	if (snd_info_card_free(card) < 0) {
 		snd_printk(KERN_WARNING "unable to free card info\n");
 		/* Not fatal error */
@@ -625,9 +626,9 @@ int __init snd_card_info_init(void)
 
 int __exit snd_card_info_done(void)
 {
-	snd_info_unregister(snd_card_info_entry);
+	snd_info_free_entry(snd_card_info_entry);
 #ifdef MODULE
-	snd_info_unregister(snd_card_module_info_entry);
+	snd_info_free_entry(snd_card_module_info_entry);
 #endif
 	return 0;
 }
diff --git a/sound/core/oss/mixer_oss.c b/sound/core/oss/mixer_oss.c
index 75a9505c7445..00c95def95aa 100644
--- a/sound/core/oss/mixer_oss.c
+++ b/sound/core/oss/mixer_oss.c
@@ -1193,10 +1193,8 @@ static void snd_mixer_oss_proc_init(struct snd_mixer_oss *mixer)
 
 static void snd_mixer_oss_proc_done(struct snd_mixer_oss *mixer)
 {
-	if (mixer->proc_entry) {
-		snd_info_unregister(mixer->proc_entry);
-		mixer->proc_entry = NULL;
-	}
+	snd_info_free_entry(mixer->proc_entry);
+	mixer->proc_entry = NULL;
 }
 #else /* !CONFIG_PROC_FS */
 #define snd_mixer_oss_proc_init(mix)
diff --git a/sound/core/oss/pcm_oss.c b/sound/core/oss/pcm_oss.c
index 472fce0ee0e8..a92b93e5ebd5 100644
--- a/sound/core/oss/pcm_oss.c
+++ b/sound/core/oss/pcm_oss.c
@@ -2846,11 +2846,9 @@ static void snd_pcm_oss_proc_done(struct snd_pcm *pcm)
 	int stream;
 	for (stream = 0; stream < 2; ++stream) {
 		struct snd_pcm_str *pstr = &pcm->streams[stream];
-		if (pstr->oss.proc_entry) {
-			snd_info_unregister(pstr->oss.proc_entry);
-			pstr->oss.proc_entry = NULL;
-			snd_pcm_oss_proc_free_setup_list(pstr);
-		}
+		snd_info_free_entry(pstr->oss.proc_entry);
+		pstr->oss.proc_entry = NULL;
+		snd_pcm_oss_proc_free_setup_list(pstr);
 	}
 }
 #else /* !CONFIG_SND_VERBOSE_PROCFS */
diff --git a/sound/core/pcm.c b/sound/core/pcm.c
index 7581edd7b9ff..b8602471f7e5 100644
--- a/sound/core/pcm.c
+++ b/sound/core/pcm.c
@@ -494,19 +494,13 @@ static int snd_pcm_stream_proc_init(struct snd_pcm_str *pstr)
 static int snd_pcm_stream_proc_done(struct snd_pcm_str *pstr)
 {
 #ifdef CONFIG_SND_PCM_XRUN_DEBUG
-	if (pstr->proc_xrun_debug_entry) {
-		snd_info_unregister(pstr->proc_xrun_debug_entry);
-		pstr->proc_xrun_debug_entry = NULL;
-	}
+	snd_info_free_entry(pstr->proc_xrun_debug_entry);
+	pstr->proc_xrun_debug_entry = NULL;
 #endif
-	if (pstr->proc_info_entry) {
-		snd_info_unregister(pstr->proc_info_entry);
-		pstr->proc_info_entry = NULL;
-	}
-	if (pstr->proc_root) {
-		snd_info_unregister(pstr->proc_root);
-		pstr->proc_root = NULL;
-	}
+	snd_info_free_entry(pstr->proc_info_entry);
+	pstr->proc_info_entry = NULL;
+	snd_info_free_entry(pstr->proc_root);
+	pstr->proc_root = NULL;
 	return 0;
 }
 
@@ -570,29 +564,19 @@ static int snd_pcm_substream_proc_init(struct snd_pcm_substream *substream)
 
 	return 0;
 }
-		
+
 static int snd_pcm_substream_proc_done(struct snd_pcm_substream *substream)
 {
-	if (substream->proc_info_entry) {
-		snd_info_unregister(substream->proc_info_entry);
-		substream->proc_info_entry = NULL;
-	}
-	if (substream->proc_hw_params_entry) {
-		snd_info_unregister(substream->proc_hw_params_entry);
-		substream->proc_hw_params_entry = NULL;
-	}
-	if (substream->proc_sw_params_entry) {
-		snd_info_unregister(substream->proc_sw_params_entry);
-		substream->proc_sw_params_entry = NULL;
-	}
-	if (substream->proc_status_entry) {
-		snd_info_unregister(substream->proc_status_entry);
-		substream->proc_status_entry = NULL;
-	}
-	if (substream->proc_root) {
-		snd_info_unregister(substream->proc_root);
-		substream->proc_root = NULL;
-	}
+	snd_info_free_entry(substream->proc_info_entry);
+	substream->proc_info_entry = NULL;
+	snd_info_free_entry(substream->proc_hw_params_entry);
+	substream->proc_hw_params_entry = NULL;
+	snd_info_free_entry(substream->proc_sw_params_entry);
+	substream->proc_sw_params_entry = NULL;
+	snd_info_free_entry(substream->proc_status_entry);
+	substream->proc_status_entry = NULL;
+	snd_info_free_entry(substream->proc_root);
+	substream->proc_root = NULL;
 	return 0;
 }
 #else /* !CONFIG_SND_VERBOSE_PROCFS */
@@ -1090,8 +1074,7 @@ static void snd_pcm_proc_init(void)
 
 static void snd_pcm_proc_done(void)
 {
-	if (snd_pcm_proc_entry)
-		snd_info_unregister(snd_pcm_proc_entry);
+	snd_info_free_entry(snd_pcm_proc_entry);
 }
 
 #else /* !CONFIG_PROC_FS */
diff --git a/sound/core/pcm_memory.c b/sound/core/pcm_memory.c
index 067d2056db9a..be030cb4d373 100644
--- a/sound/core/pcm_memory.c
+++ b/sound/core/pcm_memory.c
@@ -101,7 +101,7 @@ int snd_pcm_lib_preallocate_free(struct snd_pcm_substream *substream)
 {
 	snd_pcm_lib_preallocate_dma_free(substream);
 #ifdef CONFIG_SND_VERBOSE_PROCFS
-	snd_info_unregister(substream->proc_prealloc_entry);
+	snd_info_free_entry(substream->proc_prealloc_entry);
 	substream->proc_prealloc_entry = NULL;
 #endif
 	return 0;
diff --git a/sound/core/rawmidi.c b/sound/core/rawmidi.c
index 8c15c66eb4aa..51577c22f8ce 100644
--- a/sound/core/rawmidi.c
+++ b/sound/core/rawmidi.c
@@ -1599,7 +1599,7 @@ static int snd_rawmidi_dev_unregister(struct snd_device *device)
 	mutex_lock(&register_mutex);
 	list_del(&rmidi->list);
 	if (rmidi->proc_entry) {
-		snd_info_unregister(rmidi->proc_entry);
+		snd_info_free_entry(rmidi->proc_entry);
 		rmidi->proc_entry = NULL;
 	}
 #ifdef CONFIG_SND_OSSEMUL
diff --git a/sound/core/seq/oss/seq_oss.c b/sound/core/seq/oss/seq_oss.c
index e7234135641c..92858cf8b6eb 100644
--- a/sound/core/seq/oss/seq_oss.c
+++ b/sound/core/seq/oss/seq_oss.c
@@ -303,8 +303,7 @@ register_proc(void)
 static void
 unregister_proc(void)
 {
-	if (info_entry)
-		snd_info_unregister(info_entry);
+	snd_info_free_entry(info_entry);
 	info_entry = NULL;
 }
 #endif /* CONFIG_PROC_FS */
diff --git a/sound/core/seq/seq_device.c b/sound/core/seq/seq_device.c
index 102ff548ce69..b85954e956d4 100644
--- a/sound/core/seq/seq_device.c
+++ b/sound/core/seq/seq_device.c
@@ -573,7 +573,7 @@ static void __exit alsa_seq_device_exit(void)
 {
 	remove_drivers();
 #ifdef CONFIG_PROC_FS
-	snd_info_unregister(info_entry);
+	snd_info_free_entry(info_entry);
 #endif
 	if (num_ops)
 		snd_printk(KERN_ERR "drivers not released (%d)\n", num_ops);
diff --git a/sound/core/seq/seq_info.c b/sound/core/seq/seq_info.c
index 142e9e6882c9..8a7fe5cca1c9 100644
--- a/sound/core/seq/seq_info.c
+++ b/sound/core/seq/seq_info.c
@@ -64,9 +64,9 @@ int __init snd_seq_info_init(void)
 
 int __exit snd_seq_info_done(void)
 {
-	snd_info_unregister(queues_entry);
-	snd_info_unregister(clients_entry);
-	snd_info_unregister(timer_entry);
+	snd_info_free_entry(queues_entry);
+	snd_info_free_entry(clients_entry);
+	snd_info_free_entry(timer_entry);
 	return 0;
 }
 #endif
diff --git a/sound/core/sound.c b/sound/core/sound.c
index 7edd1fc58b17..b4430db3fa4c 100644
--- a/sound/core/sound.c
+++ b/sound/core/sound.c
@@ -387,8 +387,7 @@ int __init snd_minor_info_init(void)
 
 int __exit snd_minor_info_done(void)
 {
-	if (snd_minor_info_entry)
-		snd_info_unregister(snd_minor_info_entry);
+	snd_info_free_entry(snd_minor_info_entry);
 	return 0;
 }
 #endif /* CONFIG_PROC_FS */
diff --git a/sound/core/sound_oss.c b/sound/core/sound_oss.c
index 74f0fe5a1ba0..b2fc40aa520b 100644
--- a/sound/core/sound_oss.c
+++ b/sound/core/sound_oss.c
@@ -270,8 +270,7 @@ int __init snd_minor_info_oss_init(void)
 
 int __exit snd_minor_info_oss_done(void)
 {
-	if (snd_minor_info_oss_entry)
-		snd_info_unregister(snd_minor_info_oss_entry);
+	snd_info_free_entry(snd_minor_info_oss_entry);
 	return 0;
 }
 #endif /* CONFIG_PROC_FS */
diff --git a/sound/core/timer.c b/sound/core/timer.c
index 0a984e881c10..52ecbe1e9abb 100644
--- a/sound/core/timer.c
+++ b/sound/core/timer.c
@@ -1126,7 +1126,7 @@ static void __init snd_timer_proc_init(void)
 
 static void __exit snd_timer_proc_done(void)
 {
-	snd_info_unregister(snd_timer_proc_entry);
+	snd_info_free_entry(snd_timer_proc_entry);
 }
 #else /* !CONFIG_PROC_FS */
 #define snd_timer_proc_init()
diff --git a/sound/drivers/opl4/opl4_proc.c b/sound/drivers/opl4/opl4_proc.c
index e552ec34166f..11dd811771a4 100644
--- a/sound/drivers/opl4/opl4_proc.c
+++ b/sound/drivers/opl4/opl4_proc.c
@@ -159,8 +159,7 @@ int snd_opl4_create_proc(struct snd_opl4 *opl4)
 
 void snd_opl4_free_proc(struct snd_opl4 *opl4)
 {
-	if (opl4->proc_entry)
-		snd_info_unregister(opl4->proc_entry);
+	snd_info_free_entry(opl4->proc_entry);
 }
 
 #endif /* CONFIG_PROC_FS */
diff --git a/sound/pci/ac97/ac97_proc.c b/sound/pci/ac97/ac97_proc.c
index 2118df50b9d6..a3fdd7da911c 100644
--- a/sound/pci/ac97/ac97_proc.c
+++ b/sound/pci/ac97/ac97_proc.c
@@ -457,14 +457,10 @@ void snd_ac97_proc_init(struct snd_ac97 * ac97)
 
 void snd_ac97_proc_done(struct snd_ac97 * ac97)
 {
-	if (ac97->proc_regs) {
-		snd_info_unregister(ac97->proc_regs);
-		ac97->proc_regs = NULL;
-	}
-	if (ac97->proc) {
-		snd_info_unregister(ac97->proc);
-		ac97->proc = NULL;
-	}
+	snd_info_free_entry(ac97->proc_regs);
+	ac97->proc_regs = NULL;
+	snd_info_free_entry(ac97->proc);
+	ac97->proc = NULL;
 }
 
 void snd_ac97_bus_proc_init(struct snd_ac97_bus * bus)
@@ -485,8 +481,6 @@ void snd_ac97_bus_proc_init(struct snd_ac97_bus * bus)
 
 void snd_ac97_bus_proc_done(struct snd_ac97_bus * bus)
 {
-	if (bus->proc) {
-		snd_info_unregister(bus->proc);
-		bus->proc = NULL;
-	}
+	snd_info_free_entry(bus->proc);
+	bus->proc = NULL;
 }
diff --git a/sound/pci/cs46xx/dsp_spos.c b/sound/pci/cs46xx/dsp_spos.c
index 5c9711c0265c..89c402770a1d 100644
--- a/sound/pci/cs46xx/dsp_spos.c
+++ b/sound/pci/cs46xx/dsp_spos.c
@@ -868,35 +868,23 @@ int cs46xx_dsp_proc_done (struct snd_cs46xx *chip)
 	struct dsp_spos_instance * ins = chip->dsp_spos_instance;
 	int i;
 
-	if (ins->proc_sym_info_entry) {
-		snd_info_unregister(ins->proc_sym_info_entry);
-		ins->proc_sym_info_entry = NULL;
-	}
-  
-	if (ins->proc_modules_info_entry) {
-		snd_info_unregister(ins->proc_modules_info_entry);
-		ins->proc_modules_info_entry = NULL;
-	}
- 
-	if (ins->proc_parameter_dump_info_entry) {
-		snd_info_unregister(ins->proc_parameter_dump_info_entry);
-		ins->proc_parameter_dump_info_entry = NULL;
-	}
-  
-	if (ins->proc_sample_dump_info_entry) {
-		snd_info_unregister(ins->proc_sample_dump_info_entry);
-		ins->proc_sample_dump_info_entry = NULL;
-	}
-  
-	if (ins->proc_scb_info_entry) {
-		snd_info_unregister(ins->proc_scb_info_entry);
-		ins->proc_scb_info_entry = NULL;
-	}
-  
-	if (ins->proc_task_info_entry) {
-		snd_info_unregister(ins->proc_task_info_entry);
-		ins->proc_task_info_entry = NULL;
-	}
+	snd_info_free_entry(ins->proc_sym_info_entry);
+	ins->proc_sym_info_entry = NULL;
+
+	snd_info_free_entry(ins->proc_modules_info_entry);
+	ins->proc_modules_info_entry = NULL;
+
+	snd_info_free_entry(ins->proc_parameter_dump_info_entry);
+	ins->proc_parameter_dump_info_entry = NULL;
+
+	snd_info_free_entry(ins->proc_sample_dump_info_entry);
+	ins->proc_sample_dump_info_entry = NULL;
+
+	snd_info_free_entry(ins->proc_scb_info_entry);
+	ins->proc_scb_info_entry = NULL;
+
+	snd_info_free_entry(ins->proc_task_info_entry);
+	ins->proc_task_info_entry = NULL;
 
 	mutex_lock(&chip->spos_mutex);
 	for (i = 0; i < ins->nscb; ++i) {
@@ -905,10 +893,8 @@ int cs46xx_dsp_proc_done (struct snd_cs46xx *chip)
 	}
 	mutex_unlock(&chip->spos_mutex);
 
-	if (ins->proc_dsp_dir) {
-		snd_info_unregister (ins->proc_dsp_dir);
-		ins->proc_dsp_dir = NULL;
-	}
+	snd_info_free_entry(ins->proc_dsp_dir);
+	ins->proc_dsp_dir = NULL;
 
 	return 0;
 }
diff --git a/sound/pci/cs46xx/dsp_spos_scb_lib.c b/sound/pci/cs46xx/dsp_spos_scb_lib.c
index 232b337852ff..343f51d5311b 100644
--- a/sound/pci/cs46xx/dsp_spos_scb_lib.c
+++ b/sound/pci/cs46xx/dsp_spos_scb_lib.c
@@ -233,7 +233,7 @@ void cs46xx_dsp_proc_free_scb_desc (struct dsp_scb_descriptor * scb)
 
 		snd_printdd("cs46xx_dsp_proc_free_scb_desc: freeing %s\n",scb->scb_name);
 
-		snd_info_unregister(scb->proc_info);
+		snd_info_free_entry(scb->proc_info);
 		scb->proc_info = NULL;
 
 		snd_assert (scb_info != NULL, return);
diff --git a/sound/synth/emux/emux_proc.c b/sound/synth/emux/emux_proc.c
index 58b9601f3ad0..59144ec026e4 100644
--- a/sound/synth/emux/emux_proc.c
+++ b/sound/synth/emux/emux_proc.c
@@ -128,10 +128,8 @@ void snd_emux_proc_init(struct snd_emux *emu, struct snd_card *card, int device)
 
 void snd_emux_proc_free(struct snd_emux *emu)
 {
-	if (emu->proc) {
-		snd_info_unregister(emu->proc);
-		emu->proc = NULL;
-	}
+	snd_info_free_entry(emu->proc);
+	emu->proc = NULL;
 }
 
 #endif /* CONFIG_PROC_FS */
-- 
cgit v1.2.3


From c461482c8072bb073e6146db320d3da85cdc89ad Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Fri, 23 Jun 2006 14:38:23 +0200
Subject: [ALSA] Unregister device files at disconnection

Orignally proposed by Sam Revitch <sam.revitch@gmail.com>.
Unregister device files at disconnection to avoid the futher accesses.
Also, the dev_unregister callback is removed and replaced with the
combination of disconnect + free.
A new function snd_card_free_when_closed() is introduced, which is
used in USB disconnect callback.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 include/sound/core.h        |  3 +-
 include/sound/timer.h       |  1 -
 sound/core/control.c        | 27 ++++++------------
 sound/core/device.c         | 20 ++++++-------
 sound/core/hwdep.c          | 10 +++----
 sound/core/init.c           | 69 +++++++++++++++++++++++++++++++++++----------
 sound/core/oss/mixer_oss.c  | 16 +++++------
 sound/core/oss/pcm_oss.c    | 16 +++++------
 sound/core/pcm.c            | 48 ++++++++++++-------------------
 sound/core/rawmidi.c        | 35 ++++++-----------------
 sound/core/rtctimer.c       |  2 +-
 sound/core/seq/seq_device.c | 11 --------
 sound/core/timer.c          | 52 +++++++++++++++-------------------
 sound/pci/ac97/ac97_codec.c |  8 +++---
 sound/usb/usbaudio.c        |  2 +-
 15 files changed, 148 insertions(+), 172 deletions(-)

(limited to 'include')

diff --git a/include/sound/core.h b/include/sound/core.h
index bab3ff457e40..cf4001cf6248 100644
--- a/include/sound/core.h
+++ b/include/sound/core.h
@@ -71,7 +71,6 @@ struct snd_device_ops {
 	int (*dev_free)(struct snd_device *dev);
 	int (*dev_register)(struct snd_device *dev);
 	int (*dev_disconnect)(struct snd_device *dev);
-	int (*dev_unregister)(struct snd_device *dev);
 };
 
 struct snd_device {
@@ -131,6 +130,7 @@ struct snd_card {
 								state */
 	spinlock_t files_lock;		/* lock the files for this card */
 	int shutdown;			/* this card is going down */
+	int free_on_last_close;		/* free in context of file_release */
 	wait_queue_head_t shutdown_sleep;
 	struct work_struct free_workq;	/* for free in workqueue */
 	struct device *dev;
@@ -244,6 +244,7 @@ struct snd_card *snd_card_new(int idx, const char *id,
 			 struct module *module, int extra_size);
 int snd_card_disconnect(struct snd_card *card);
 int snd_card_free(struct snd_card *card);
+int snd_card_free_when_closed(struct snd_card *card);
 int snd_card_free_in_thread(struct snd_card *card);
 int snd_card_register(struct snd_card *card);
 int snd_card_info_init(void);
diff --git a/include/sound/timer.h b/include/sound/timer.h
index 5ece2bf541dc..d42c083db1da 100644
--- a/include/sound/timer.h
+++ b/include/sound/timer.h
@@ -129,7 +129,6 @@ void snd_timer_notify(struct snd_timer *timer, int event, struct timespec *tstam
 int snd_timer_global_new(char *id, int device, struct snd_timer **rtimer);
 int snd_timer_global_free(struct snd_timer *timer);
 int snd_timer_global_register(struct snd_timer *timer);
-int snd_timer_global_unregister(struct snd_timer *timer);
 
 int snd_timer_open(struct snd_timer_instance **ti, char *owner, struct snd_timer_id *tid, unsigned int slave_id);
 int snd_timer_close(struct snd_timer_instance *timeri);
diff --git a/sound/core/control.c b/sound/core/control.c
index e9c8854d2f7b..f0c7272a2d48 100644
--- a/sound/core/control.c
+++ b/sound/core/control.c
@@ -1375,6 +1375,11 @@ static int snd_ctl_dev_disconnect(struct snd_device *device)
 	struct snd_card *card = device->device_data;
 	struct list_head *flist;
 	struct snd_ctl_file *ctl;
+	int err, cardnum;
+
+	snd_assert(card != NULL, return -ENXIO);
+	cardnum = card->number;
+	snd_assert(cardnum >= 0 && cardnum < SNDRV_CARDS, return -ENXIO);
 
 	down_read(&card->controls_rwsem);
 	list_for_each(flist, &card->ctl_files) {
@@ -1383,6 +1388,10 @@ static int snd_ctl_dev_disconnect(struct snd_device *device)
 		kill_fasync(&ctl->fasync, SIGIO, POLL_ERR);
 	}
 	up_read(&card->controls_rwsem);
+
+	if ((err = snd_unregister_device(SNDRV_DEVICE_TYPE_CONTROL,
+					 card, -1)) < 0)
+		return err;
 	return 0;
 }
 
@@ -1403,23 +1412,6 @@ static int snd_ctl_dev_free(struct snd_device *device)
 	return 0;
 }
 
-/*
- * de-registration of the control device
- */
-static int snd_ctl_dev_unregister(struct snd_device *device)
-{
-	struct snd_card *card = device->device_data;
-	int err, cardnum;
-
-	snd_assert(card != NULL, return -ENXIO);
-	cardnum = card->number;
-	snd_assert(cardnum >= 0 && cardnum < SNDRV_CARDS, return -ENXIO);
-	if ((err = snd_unregister_device(SNDRV_DEVICE_TYPE_CONTROL,
-					 card, -1)) < 0)
-		return err;
-	return snd_ctl_dev_free(device);
-}
-
 /*
  * create control core:
  * called from init.c
@@ -1430,7 +1422,6 @@ int snd_ctl_create(struct snd_card *card)
 		.dev_free = snd_ctl_dev_free,
 		.dev_register =	snd_ctl_dev_register,
 		.dev_disconnect = snd_ctl_dev_disconnect,
-		.dev_unregister = snd_ctl_dev_unregister
 	};
 
 	snd_assert(card != NULL, return -ENXIO);
diff --git a/sound/core/device.c b/sound/core/device.c
index 6ce4da4a1081..ccb25816ac9e 100644
--- a/sound/core/device.c
+++ b/sound/core/device.c
@@ -71,7 +71,7 @@ EXPORT_SYMBOL(snd_device_new);
  * @device_data: the data pointer to release
  *
  * Removes the device from the list on the card and invokes the
- * callback, dev_unregister or dev_free, corresponding to the state.
+ * callbacks, dev_disconnect and dev_free, corresponding to the state.
  * Then release the device.
  *
  * Returns zero if successful, or a negative error code on failure or if the
@@ -90,16 +90,14 @@ int snd_device_free(struct snd_card *card, void *device_data)
 			continue;
 		/* unlink */
 		list_del(&dev->list);
-		if ((dev->state == SNDRV_DEV_REGISTERED ||
-		     dev->state == SNDRV_DEV_DISCONNECTED) &&
-		    dev->ops->dev_unregister) {
-			if (dev->ops->dev_unregister(dev))
-				snd_printk(KERN_ERR "device unregister failure\n");
-		} else {
-			if (dev->ops->dev_free) {
-				if (dev->ops->dev_free(dev))
-					snd_printk(KERN_ERR "device free failure\n");
-			}
+		if (dev->state == SNDRV_DEV_REGISTERED &&
+		    dev->ops->dev_disconnect)
+			if (dev->ops->dev_disconnect(dev))
+				snd_printk(KERN_ERR
+					   "device disconnect failure\n");
+		if (dev->ops->dev_free) {
+			if (dev->ops->dev_free(dev))
+				snd_printk(KERN_ERR "device free failure\n");
 		}
 		kfree(dev);
 		return 0;
diff --git a/sound/core/hwdep.c b/sound/core/hwdep.c
index cbd8a63282b6..9aa9d94891f0 100644
--- a/sound/core/hwdep.c
+++ b/sound/core/hwdep.c
@@ -42,7 +42,7 @@ static DEFINE_MUTEX(register_mutex);
 static int snd_hwdep_free(struct snd_hwdep *hwdep);
 static int snd_hwdep_dev_free(struct snd_device *device);
 static int snd_hwdep_dev_register(struct snd_device *device);
-static int snd_hwdep_dev_unregister(struct snd_device *device);
+static int snd_hwdep_dev_disconnect(struct snd_device *device);
 
 
 static struct snd_hwdep *snd_hwdep_search(struct snd_card *card, int device)
@@ -353,7 +353,7 @@ int snd_hwdep_new(struct snd_card *card, char *id, int device,
 	static struct snd_device_ops ops = {
 		.dev_free = snd_hwdep_dev_free,
 		.dev_register = snd_hwdep_dev_register,
-		.dev_unregister = snd_hwdep_dev_unregister
+		.dev_disconnect = snd_hwdep_dev_disconnect,
 	};
 
 	snd_assert(rhwdep != NULL, return -EINVAL);
@@ -439,7 +439,7 @@ static int snd_hwdep_dev_register(struct snd_device *device)
 	return 0;
 }
 
-static int snd_hwdep_dev_unregister(struct snd_device *device)
+static int snd_hwdep_dev_disconnect(struct snd_device *device)
 {
 	struct snd_hwdep *hwdep = device->device_data;
 
@@ -454,9 +454,9 @@ static int snd_hwdep_dev_unregister(struct snd_device *device)
 		snd_unregister_oss_device(hwdep->oss_type, hwdep->card, hwdep->device);
 #endif
 	snd_unregister_device(SNDRV_DEVICE_TYPE_HWDEP, hwdep->card, hwdep->device);
-	list_del(&hwdep->list);
+	list_del_init(&hwdep->list);
 	mutex_unlock(&register_mutex);
-	return snd_hwdep_free(hwdep);
+	return 0;
 }
 
 #ifdef CONFIG_PROC_FS
diff --git a/sound/core/init.c b/sound/core/init.c
index 1ecb029ff4c9..5850d99d21e3 100644
--- a/sound/core/init.c
+++ b/sound/core/init.c
@@ -327,22 +327,10 @@ EXPORT_SYMBOL(snd_card_disconnect);
  *  Returns zero. Frees all associated devices and frees the control
  *  interface associated to given soundcard.
  */
-int snd_card_free(struct snd_card *card)
+static int snd_card_do_free(struct snd_card *card)
 {
 	struct snd_shutdown_f_ops *s_f_ops;
 
-	if (card == NULL)
-		return -EINVAL;
-	mutex_lock(&snd_card_mutex);
-	snd_cards[card->number] = NULL;
-	mutex_unlock(&snd_card_mutex);
-
-#ifdef CONFIG_PM
-	wake_up(&card->power_sleep);
-#endif
-	/* wait, until all devices are ready for the free operation */
-	wait_event(card->shutdown_sleep, card->files == NULL);
-
 #if defined(CONFIG_SND_MIXER_OSS) || defined(CONFIG_SND_MIXER_OSS_MODULE)
 	if (snd_mixer_oss_notify_callback)
 		snd_mixer_oss_notify_callback(card, SND_MIXER_OSS_NOTIFY_FREE);
@@ -371,10 +359,55 @@ int snd_card_free(struct snd_card *card)
 		card->s_f_ops = s_f_ops->next;
 		kfree(s_f_ops);
 	}
+	kfree(card);
+	return 0;
+}
+
+static int snd_card_free_prepare(struct snd_card *card)
+{
+	if (card == NULL)
+		return -EINVAL;
+	(void) snd_card_disconnect(card);
 	mutex_lock(&snd_card_mutex);
+	snd_cards[card->number] = NULL;
 	snd_cards_lock &= ~(1 << card->number);
 	mutex_unlock(&snd_card_mutex);
-	kfree(card);
+#ifdef CONFIG_PM
+	wake_up(&card->power_sleep);
+#endif
+	return 0;
+}
+
+int snd_card_free_when_closed(struct snd_card *card)
+{
+	int free_now = 0;
+	int ret = snd_card_free_prepare(card);
+	if (ret)
+		return ret;
+
+	spin_lock(&card->files_lock);
+	if (card->files == NULL)
+		free_now = 1;
+	else
+		card->free_on_last_close = 1;
+	spin_unlock(&card->files_lock);
+
+	if (free_now)
+		snd_card_do_free(card);
+	return 0;
+}
+
+EXPORT_SYMBOL(snd_card_free_when_closed);
+
+int snd_card_free(struct snd_card *card)
+{
+	int ret = snd_card_free_prepare(card);
+	if (ret)
+		return ret;
+
+	/* wait, until all devices are ready for the free operation */
+	wait_event(card->shutdown_sleep, card->files == NULL);
+	snd_card_do_free(card);
 	return 0;
 }
 
@@ -718,6 +751,7 @@ EXPORT_SYMBOL(snd_card_file_add);
 int snd_card_file_remove(struct snd_card *card, struct file *file)
 {
 	struct snd_monitor_file *mfile, *pfile = NULL;
+	int last_close = 0;
 
 	spin_lock(&card->files_lock);
 	mfile = card->files;
@@ -732,9 +766,14 @@ int snd_card_file_remove(struct snd_card *card, struct file *file)
 		pfile = mfile;
 		mfile = mfile->next;
 	}
-	spin_unlock(&card->files_lock);
 	if (card->files == NULL)
+		last_close = 1;
+	spin_unlock(&card->files_lock);
+	if (last_close) {
 		wake_up(&card->shutdown_sleep);
+		if (card->free_on_last_close)
+			snd_card_do_free(card);
+	}
 	if (!mfile) {
 		snd_printk(KERN_ERR "ALSA card file remove problem (%p)\n", file);
 		return -ENOENT;
diff --git a/sound/core/oss/mixer_oss.c b/sound/core/oss/mixer_oss.c
index 00c95def95aa..f4c67042e3ac 100644
--- a/sound/core/oss/mixer_oss.c
+++ b/sound/core/oss/mixer_oss.c
@@ -1310,21 +1310,19 @@ static int snd_mixer_oss_notify_handler(struct snd_card *card, int cmd)
 		card->mixer_oss = mixer;
 		snd_mixer_oss_build(mixer);
 		snd_mixer_oss_proc_init(mixer);
-	} else if (cmd == SND_MIXER_OSS_NOTIFY_DISCONNECT) {
-		mixer = card->mixer_oss;
-		if (mixer == NULL || !mixer->oss_dev_alloc)
-			return 0;
-		snd_unregister_oss_device(SNDRV_OSS_DEVICE_TYPE_MIXER, mixer->card, 0);
-		mixer->oss_dev_alloc = 0;
-	} else {		/* free */
+	} else {
 		mixer = card->mixer_oss;
 		if (mixer == NULL)
 			return 0;
+		if (mixer->oss_dev_alloc) {
 #ifdef SNDRV_OSS_INFO_DEV_MIXERS
-		snd_oss_info_unregister(SNDRV_OSS_INFO_DEV_MIXERS, mixer->card->number);
+			snd_oss_info_unregister(SNDRV_OSS_INFO_DEV_MIXERS, mixer->card->number);
 #endif
-		if (mixer->oss_dev_alloc)
 			snd_unregister_oss_device(SNDRV_OSS_DEVICE_TYPE_MIXER, mixer->card, 0);
+			mixer->oss_dev_alloc = 0;
+		}
+		if (cmd == SND_MIXER_OSS_NOTIFY_DISCONNECT)
+			return 0;
 		snd_mixer_oss_proc_done(mixer);
 		return snd_mixer_oss_free1(mixer);
 	}
diff --git a/sound/core/oss/pcm_oss.c b/sound/core/oss/pcm_oss.c
index a92b93e5ebd5..505b23ec4058 100644
--- a/sound/core/oss/pcm_oss.c
+++ b/sound/core/oss/pcm_oss.c
@@ -2929,25 +2929,23 @@ static int snd_pcm_oss_disconnect_minor(struct snd_pcm *pcm)
 			snd_unregister_oss_device(SNDRV_OSS_DEVICE_TYPE_PCM,
 						  pcm->card, 1);
 		}
-	}
-	return 0;
-}
-
-static int snd_pcm_oss_unregister_minor(struct snd_pcm *pcm)
-{
-	snd_pcm_oss_disconnect_minor(pcm);
-	if (pcm->oss.reg) {
 		if (dsp_map[pcm->card->number] == (int)pcm->device) {
 #ifdef SNDRV_OSS_INFO_DEV_AUDIO
 			snd_oss_info_unregister(SNDRV_OSS_INFO_DEV_AUDIO, pcm->card->number);
 #endif
 		}
 		pcm->oss.reg = 0;
-		snd_pcm_oss_proc_done(pcm);
 	}
 	return 0;
 }
 
+static int snd_pcm_oss_unregister_minor(struct snd_pcm *pcm)
+{
+	snd_pcm_oss_disconnect_minor(pcm);
+	snd_pcm_oss_proc_done(pcm);
+	return 0;
+}
+
 static struct snd_pcm_notify snd_pcm_oss_notify =
 {
 	.n_register =	snd_pcm_oss_register_minor,
diff --git a/sound/core/pcm.c b/sound/core/pcm.c
index b8602471f7e5..f52178abf120 100644
--- a/sound/core/pcm.c
+++ b/sound/core/pcm.c
@@ -42,7 +42,6 @@ static int snd_pcm_free(struct snd_pcm *pcm);
 static int snd_pcm_dev_free(struct snd_device *device);
 static int snd_pcm_dev_register(struct snd_device *device);
 static int snd_pcm_dev_disconnect(struct snd_device *device);
-static int snd_pcm_dev_unregister(struct snd_device *device);
 
 static struct snd_pcm *snd_pcm_search(struct snd_card *card, int device)
 {
@@ -680,7 +679,6 @@ int snd_pcm_new(struct snd_card *card, char *id, int device,
 		.dev_free = snd_pcm_dev_free,
 		.dev_register =	snd_pcm_dev_register,
 		.dev_disconnect = snd_pcm_dev_disconnect,
-		.dev_unregister = snd_pcm_dev_unregister
 	};
 
 	snd_assert(rpcm != NULL, return -EINVAL);
@@ -724,6 +722,7 @@ static void snd_pcm_free_stream(struct snd_pcm_str * pstr)
 	substream = pstr->substream;
 	while (substream) {
 		substream_next = substream->next;
+		snd_pcm_timer_done(substream);
 		snd_pcm_substream_proc_done(substream);
 		kfree(substream);
 		substream = substream_next;
@@ -740,7 +739,12 @@ static void snd_pcm_free_stream(struct snd_pcm_str * pstr)
 
 static int snd_pcm_free(struct snd_pcm *pcm)
 {
+	struct snd_pcm_notify *notify;
+
 	snd_assert(pcm != NULL, return -ENXIO);
+	list_for_each_entry(notify, &snd_pcm_notify_list, list) {
+		notify->n_unregister(pcm);
+	}
 	if (pcm->private_free)
 		pcm->private_free(pcm);
 	snd_pcm_lib_preallocate_free_for_all(pcm);
@@ -955,35 +959,22 @@ static int snd_pcm_dev_register(struct snd_device *device)
 static int snd_pcm_dev_disconnect(struct snd_device *device)
 {
 	struct snd_pcm *pcm = device->device_data;
-	struct list_head *list;
+	struct snd_pcm_notify *notify;
 	struct snd_pcm_substream *substream;
-	int cidx;
+	int cidx, devtype;
 
 	mutex_lock(&register_mutex);
+	if (list_empty(&pcm->list))
+		goto unlock;
+
 	list_del_init(&pcm->list);
 	for (cidx = 0; cidx < 2; cidx++)
 		for (substream = pcm->streams[cidx].substream; substream; substream = substream->next)
 			if (substream->runtime)
 				substream->runtime->status->state = SNDRV_PCM_STATE_DISCONNECTED;
-	list_for_each(list, &snd_pcm_notify_list) {
-		struct snd_pcm_notify *notify;
-		notify = list_entry(list, struct snd_pcm_notify, list);
+	list_for_each_entry(notify, &snd_pcm_notify_list, list) {
 		notify->n_disconnect(pcm);
 	}
-	mutex_unlock(&register_mutex);
-	return 0;
-}
-
-static int snd_pcm_dev_unregister(struct snd_device *device)
-{
-	int cidx, devtype;
-	struct snd_pcm_substream *substream;
-	struct list_head *list;
-	struct snd_pcm *pcm = device->device_data;
-
-	snd_assert(pcm != NULL, return -ENXIO);
-	mutex_lock(&register_mutex);
-	list_del(&pcm->list);
 	for (cidx = 0; cidx < 2; cidx++) {
 		devtype = -1;
 		switch (cidx) {
@@ -995,23 +986,20 @@ static int snd_pcm_dev_unregister(struct snd_device *device)
 			break;
 		}
 		snd_unregister_device(devtype, pcm->card, pcm->device);
-		for (substream = pcm->streams[cidx].substream; substream; substream = substream->next)
-			snd_pcm_timer_done(substream);
-	}
-	list_for_each(list, &snd_pcm_notify_list) {
-		struct snd_pcm_notify *notify;
-		notify = list_entry(list, struct snd_pcm_notify, list);
-		notify->n_unregister(pcm);
 	}
+ unlock:
 	mutex_unlock(&register_mutex);
-	return snd_pcm_free(pcm);
+	return 0;
 }
 
 int snd_pcm_notify(struct snd_pcm_notify *notify, int nfree)
 {
 	struct list_head *p;
 
-	snd_assert(notify != NULL && notify->n_register != NULL && notify->n_unregister != NULL, return -EINVAL);
+	snd_assert(notify != NULL &&
+		   notify->n_register != NULL &&
+		   notify->n_unregister != NULL &&
+		   notify->n_disconnect, return -EINVAL);
 	mutex_lock(&register_mutex);
 	if (nfree) {
 		list_del(&notify->list);
diff --git a/sound/core/rawmidi.c b/sound/core/rawmidi.c
index 51577c22f8ce..8a2bdfae63e3 100644
--- a/sound/core/rawmidi.c
+++ b/sound/core/rawmidi.c
@@ -55,7 +55,6 @@ static int snd_rawmidi_free(struct snd_rawmidi *rawmidi);
 static int snd_rawmidi_dev_free(struct snd_device *device);
 static int snd_rawmidi_dev_register(struct snd_device *device);
 static int snd_rawmidi_dev_disconnect(struct snd_device *device);
-static int snd_rawmidi_dev_unregister(struct snd_device *device);
 
 static LIST_HEAD(snd_rawmidi_devices);
 static DEFINE_MUTEX(register_mutex);
@@ -1426,7 +1425,6 @@ int snd_rawmidi_new(struct snd_card *card, char *id, int device,
 		.dev_free = snd_rawmidi_dev_free,
 		.dev_register = snd_rawmidi_dev_register,
 		.dev_disconnect = snd_rawmidi_dev_disconnect,
-		.dev_unregister = snd_rawmidi_dev_unregister
 	};
 
 	snd_assert(rrawmidi != NULL, return -EINVAL);
@@ -1479,6 +1477,14 @@ static void snd_rawmidi_free_substreams(struct snd_rawmidi_str *stream)
 static int snd_rawmidi_free(struct snd_rawmidi *rmidi)
 {
 	snd_assert(rmidi != NULL, return -ENXIO);	
+
+	snd_info_free_entry(rmidi->proc_entry);
+	rmidi->proc_entry = NULL;
+	mutex_lock(&register_mutex);
+	if (rmidi->ops && rmidi->ops->dev_unregister)
+		rmidi->ops->dev_unregister(rmidi);
+	mutex_unlock(&register_mutex);
+
 	snd_rawmidi_free_substreams(&rmidi->streams[SNDRV_RAWMIDI_STREAM_INPUT]);
 	snd_rawmidi_free_substreams(&rmidi->streams[SNDRV_RAWMIDI_STREAM_OUTPUT]);
 	if (rmidi->private_free)
@@ -1587,21 +1593,6 @@ static int snd_rawmidi_dev_disconnect(struct snd_device *device)
 
 	mutex_lock(&register_mutex);
 	list_del_init(&rmidi->list);
-	mutex_unlock(&register_mutex);
-	return 0;
-}
-
-static int snd_rawmidi_dev_unregister(struct snd_device *device)
-{
-	struct snd_rawmidi *rmidi = device->device_data;
-
-	snd_assert(rmidi != NULL, return -ENXIO);
-	mutex_lock(&register_mutex);
-	list_del(&rmidi->list);
-	if (rmidi->proc_entry) {
-		snd_info_free_entry(rmidi->proc_entry);
-		rmidi->proc_entry = NULL;
-	}
 #ifdef CONFIG_SND_OSSEMUL
 	if (rmidi->ossreg) {
 		if ((int)rmidi->device == midi_map[rmidi->card->number]) {
@@ -1615,17 +1606,9 @@ static int snd_rawmidi_dev_unregister(struct snd_device *device)
 		rmidi->ossreg = 0;
 	}
 #endif /* CONFIG_SND_OSSEMUL */
-	if (rmidi->ops && rmidi->ops->dev_unregister)
-		rmidi->ops->dev_unregister(rmidi);
 	snd_unregister_device(SNDRV_DEVICE_TYPE_RAWMIDI, rmidi->card, rmidi->device);
 	mutex_unlock(&register_mutex);
-#if defined(CONFIG_SND_SEQUENCER) || (defined(MODULE) && defined(CONFIG_SND_SEQUENCER_MODULE))
-	if (rmidi->seq_dev) {
-		snd_device_free(rmidi->card, rmidi->seq_dev);
-		rmidi->seq_dev = NULL;
-	}
-#endif
-	return snd_rawmidi_free(rmidi);
+	return 0;
 }
 
 /**
diff --git a/sound/core/rtctimer.c b/sound/core/rtctimer.c
index 84704ccb1829..412dd62b654e 100644
--- a/sound/core/rtctimer.c
+++ b/sound/core/rtctimer.c
@@ -156,7 +156,7 @@ static int __init rtctimer_init(void)
 static void __exit rtctimer_exit(void)
 {
 	if (rtctimer) {
-		snd_timer_global_unregister(rtctimer);
+		snd_timer_global_free(rtctimer);
 		rtctimer = NULL;
 	}
 }
diff --git a/sound/core/seq/seq_device.c b/sound/core/seq/seq_device.c
index b85954e956d4..b79d011813c0 100644
--- a/sound/core/seq/seq_device.c
+++ b/sound/core/seq/seq_device.c
@@ -90,7 +90,6 @@ static int snd_seq_device_free(struct snd_seq_device *dev);
 static int snd_seq_device_dev_free(struct snd_device *device);
 static int snd_seq_device_dev_register(struct snd_device *device);
 static int snd_seq_device_dev_disconnect(struct snd_device *device);
-static int snd_seq_device_dev_unregister(struct snd_device *device);
 
 static int init_device(struct snd_seq_device *dev, struct ops_list *ops);
 static int free_device(struct snd_seq_device *dev, struct ops_list *ops);
@@ -189,7 +188,6 @@ int snd_seq_device_new(struct snd_card *card, int device, char *id, int argsize,
 		.dev_free = snd_seq_device_dev_free,
 		.dev_register = snd_seq_device_dev_register,
 		.dev_disconnect = snd_seq_device_dev_disconnect,
-		.dev_unregister = snd_seq_device_dev_unregister
 	};
 
 	if (result)
@@ -308,15 +306,6 @@ static int snd_seq_device_dev_disconnect(struct snd_device *device)
 	return 0;
 }
 
-/*
- * unregister the existing device
- */
-static int snd_seq_device_dev_unregister(struct snd_device *device)
-{
-	struct snd_seq_device *dev = device->device_data;
-	return snd_seq_device_free(dev);
-}
-
 /*
  * register device driver
  * id = driver id
diff --git a/sound/core/timer.c b/sound/core/timer.c
index 52ecbe1e9abb..7e5e562fe356 100644
--- a/sound/core/timer.c
+++ b/sound/core/timer.c
@@ -88,7 +88,7 @@ static DEFINE_MUTEX(register_mutex);
 static int snd_timer_free(struct snd_timer *timer);
 static int snd_timer_dev_free(struct snd_device *device);
 static int snd_timer_dev_register(struct snd_device *device);
-static int snd_timer_dev_unregister(struct snd_device *device);
+static int snd_timer_dev_disconnect(struct snd_device *device);
 
 static void snd_timer_reschedule(struct snd_timer * timer, unsigned long ticks_left);
 
@@ -773,7 +773,7 @@ int snd_timer_new(struct snd_card *card, char *id, struct snd_timer_id *tid,
 	static struct snd_device_ops ops = {
 		.dev_free = snd_timer_dev_free,
 		.dev_register = snd_timer_dev_register,
-		.dev_unregister = snd_timer_dev_unregister
+		.dev_disconnect = snd_timer_dev_disconnect,
 	};
 
 	snd_assert(tid != NULL, return -EINVAL);
@@ -813,6 +813,21 @@ int snd_timer_new(struct snd_card *card, char *id, struct snd_timer_id *tid,
 static int snd_timer_free(struct snd_timer *timer)
 {
 	snd_assert(timer != NULL, return -ENXIO);
+
+	mutex_lock(&register_mutex);
+	if (! list_empty(&timer->open_list_head)) {
+		struct list_head *p, *n;
+		struct snd_timer_instance *ti;
+		snd_printk(KERN_WARNING "timer %p is busy?\n", timer);
+		list_for_each_safe(p, n, &timer->open_list_head) {
+			list_del_init(p);
+			ti = list_entry(p, struct snd_timer_instance, open_list);
+			ti->timer = NULL;
+		}
+	}
+	list_del(&timer->device_list);
+	mutex_unlock(&register_mutex);
+
 	if (timer->private_free)
 		timer->private_free(timer);
 	kfree(timer);
@@ -867,30 +882,13 @@ static int snd_timer_dev_register(struct snd_device *dev)
 	return 0;
 }
 
-static int snd_timer_unregister(struct snd_timer *timer)
+static int snd_timer_dev_disconnect(struct snd_device *device)
 {
-	struct list_head *p, *n;
-	struct snd_timer_instance *ti;
-
-	snd_assert(timer != NULL, return -ENXIO);
+	struct snd_timer *timer = device->device_data;
 	mutex_lock(&register_mutex);
-	if (! list_empty(&timer->open_list_head)) {
-		snd_printk(KERN_WARNING "timer 0x%lx is busy?\n", (long)timer);
-		list_for_each_safe(p, n, &timer->open_list_head) {
-			list_del_init(p);
-			ti = list_entry(p, struct snd_timer_instance, open_list);
-			ti->timer = NULL;
-		}
-	}
-	list_del(&timer->device_list);
+	list_del_init(&timer->device_list);
 	mutex_unlock(&register_mutex);
-	return snd_timer_free(timer);
-}
-
-static int snd_timer_dev_unregister(struct snd_device *device)
-{
-	struct snd_timer *timer = device->device_data;
-	return snd_timer_unregister(timer);
+	return 0;
 }
 
 void snd_timer_notify(struct snd_timer *timer, int event, struct timespec *tstamp)
@@ -955,11 +953,6 @@ int snd_timer_global_register(struct snd_timer *timer)
 	return snd_timer_dev_register(&dev);
 }
 
-int snd_timer_global_unregister(struct snd_timer *timer)
-{
-	return snd_timer_unregister(timer);
-}
-
 /*
  *  System timer
  */
@@ -1982,7 +1975,7 @@ static void __exit alsa_timer_exit(void)
 	/* unregister the system timer */
 	list_for_each_safe(p, n, &snd_timer_list) {
 		struct snd_timer *timer = list_entry(p, struct snd_timer, device_list);
-		snd_timer_unregister(timer);
+		snd_timer_free(timer);
 	}
 	snd_timer_proc_done();
 #ifdef SNDRV_OSS_INFO_DEV_TIMERS
@@ -2005,5 +1998,4 @@ EXPORT_SYMBOL(snd_timer_notify);
 EXPORT_SYMBOL(snd_timer_global_new);
 EXPORT_SYMBOL(snd_timer_global_free);
 EXPORT_SYMBOL(snd_timer_global_register);
-EXPORT_SYMBOL(snd_timer_global_unregister);
 EXPORT_SYMBOL(snd_timer_interrupt);
diff --git a/sound/pci/ac97/ac97_codec.c b/sound/pci/ac97/ac97_codec.c
index 51e83d7a839a..b35280ca2465 100644
--- a/sound/pci/ac97/ac97_codec.c
+++ b/sound/pci/ac97/ac97_codec.c
@@ -1817,13 +1817,13 @@ static int snd_ac97_dev_register(struct snd_device *device)
 	return 0;
 }
 
-/* unregister ac97 codec */
-static int snd_ac97_dev_unregister(struct snd_device *device)
+/* disconnect ac97 codec */
+static int snd_ac97_dev_disconnect(struct snd_device *device)
 {
 	struct snd_ac97 *ac97 = device->device_data;
 	if (ac97->dev.bus)
 		device_unregister(&ac97->dev);
-	return snd_ac97_free(ac97);
+	return 0;
 }
 
 /* build_ops to do nothing */
@@ -1860,7 +1860,7 @@ int snd_ac97_mixer(struct snd_ac97_bus *bus, struct snd_ac97_template *template,
 	static struct snd_device_ops ops = {
 		.dev_free =	snd_ac97_dev_free,
 		.dev_register =	snd_ac97_dev_register,
-		.dev_unregister =	snd_ac97_dev_unregister,
+		.dev_disconnect =	snd_ac97_dev_disconnect,
 	};
 
 	snd_assert(rac97 != NULL, return -EINVAL);
diff --git a/sound/usb/usbaudio.c b/sound/usb/usbaudio.c
index 1b7f499c549d..314431385913 100644
--- a/sound/usb/usbaudio.c
+++ b/sound/usb/usbaudio.c
@@ -3499,7 +3499,7 @@ static void snd_usb_audio_disconnect(struct usb_device *dev, void *ptr)
 		}
 		usb_chip[chip->index] = NULL;
 		mutex_unlock(&register_mutex);
-		snd_card_free(card);
+		snd_card_free_when_closed(card);
 	} else {
 		mutex_unlock(&register_mutex);
 	}
-- 
cgit v1.2.3


From 2b29b13c5794f648cd5e839796496704d787f5a6 Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Fri, 23 Jun 2006 14:38:26 +0200
Subject: [ALSA] Deprecate snd_card_free_in_thread()

Deprecated snd_card_free_in_thread(), replaced with
snd_card_free_when_closed().

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 .../sound/alsa/DocBook/writing-an-alsa-driver.tmpl |  5 +-
 include/sound/core.h                               |  3 --
 sound/core/init.c                                  | 56 ++--------------------
 sound/drivers/mpu401/mpu401.c                      |  2 +-
 sound/pcmcia/pdaudiocf/pdaudiocf.c                 |  2 +-
 sound/pcmcia/vx/vxpocket.c                         |  4 +-
 6 files changed, 9 insertions(+), 63 deletions(-)

(limited to 'include')

diff --git a/Documentation/sound/alsa/DocBook/writing-an-alsa-driver.tmpl b/Documentation/sound/alsa/DocBook/writing-an-alsa-driver.tmpl
index b8dc51ca776c..4807ef79a94d 100644
--- a/Documentation/sound/alsa/DocBook/writing-an-alsa-driver.tmpl
+++ b/Documentation/sound/alsa/DocBook/writing-an-alsa-driver.tmpl
@@ -1054,9 +1054,8 @@
 
       <para>
       For a device which allows hotplugging, you can use
-      <function>snd_card_free_in_thread</function>.  This one will
-      postpone the destruction and wait in a kernel-thread until all
-      devices are closed.
+      <function>snd_card_free_when_closed</function>.  This one will
+      postpone the destruction until all devices are closed.
       </para>
 
     </section>
diff --git a/include/sound/core.h b/include/sound/core.h
index cf4001cf6248..1359c532b68e 100644
--- a/include/sound/core.h
+++ b/include/sound/core.h
@@ -25,7 +25,6 @@
 #include <linux/sched.h>		/* wake_up() */
 #include <linux/mutex.h>		/* struct mutex */
 #include <linux/rwsem.h>		/* struct rw_semaphore */
-#include <linux/workqueue.h>		/* struct workqueue_struct */
 #include <linux/pm.h>			/* pm_message_t */
 
 /* forward declarations */
@@ -132,7 +131,6 @@ struct snd_card {
 	int shutdown;			/* this card is going down */
 	int free_on_last_close;		/* free in context of file_release */
 	wait_queue_head_t shutdown_sleep;
-	struct work_struct free_workq;	/* for free in workqueue */
 	struct device *dev;
 
 #ifdef CONFIG_PM
@@ -245,7 +243,6 @@ struct snd_card *snd_card_new(int idx, const char *id,
 int snd_card_disconnect(struct snd_card *card);
 int snd_card_free(struct snd_card *card);
 int snd_card_free_when_closed(struct snd_card *card);
-int snd_card_free_in_thread(struct snd_card *card);
 int snd_card_register(struct snd_card *card);
 int snd_card_info_init(void);
 int snd_card_info_done(void);
diff --git a/sound/core/init.c b/sound/core/init.c
index 5850d99d21e3..d7607a25acdf 100644
--- a/sound/core/init.c
+++ b/sound/core/init.c
@@ -81,8 +81,6 @@ static inline int init_info_for_card(struct snd_card *card)
 #define init_info_for_card(card)
 #endif
 
-static void snd_card_free_thread(void * __card);
-
 /**
  *  snd_card_new - create and initialize a soundcard structure
  *  @idx: card index (address) [0 ... (SNDRV_CARDS-1)]
@@ -145,7 +143,6 @@ struct snd_card *snd_card_new(int idx, const char *xid,
 	INIT_LIST_HEAD(&card->ctl_files);
 	spin_lock_init(&card->files_lock);
 	init_waitqueue_head(&card->shutdown_sleep);
-	INIT_WORK(&card->free_workq, snd_card_free_thread, card);
 #ifdef CONFIG_PM
 	mutex_init(&card->power_lock);
 	init_waitqueue_head(&card->power_sleep);
@@ -413,53 +410,6 @@ int snd_card_free(struct snd_card *card)
 
 EXPORT_SYMBOL(snd_card_free);
 
-static void snd_card_free_thread(void * __card)
-{
-	struct snd_card *card = __card;
-	struct module * module = card->module;
-
-	if (!try_module_get(module)) {
-		snd_printk(KERN_ERR "unable to lock toplevel module for card %i in free thread\n", card->number);
-		module = NULL;
-	}
-
-	snd_card_free(card);
-
-	module_put(module);
-}
-
-/**
- *  snd_card_free_in_thread - call snd_card_free() in thread
- *  @card: soundcard structure
- *
- *  This function schedules the call of snd_card_free() function in a
- *  work queue.  When all devices are released (non-busy), the work
- *  is woken up and calls snd_card_free().
- *
- *  When a card can be disconnected at any time by hotplug service,
- *  this function should be used in disconnect (or detach) callback
- *  instead of calling snd_card_free() directly.
- *  
- *  Returns - zero otherwise a negative error code if the start of thread failed.
- */
-int snd_card_free_in_thread(struct snd_card *card)
-{
-	if (card->files == NULL) {
-		snd_card_free(card);
-		return 0;
-	}
-
-	if (schedule_work(&card->free_workq))
-		return 0;
-
-	snd_printk(KERN_ERR "schedule_work() failed in snd_card_free_in_thread for card %i\n", card->number);
-	/* try to free the structure immediately */
-	snd_card_free(card);
-	return -EFAULT;
-}
-
-EXPORT_SYMBOL(snd_card_free_in_thread);
-
 static void choose_default_id(struct snd_card *card)
 {
 	int i, len, idx_flag = 0, loops = SNDRV_CARDS;
@@ -742,9 +692,9 @@ EXPORT_SYMBOL(snd_card_file_add);
  *
  *  This function removes the file formerly added to the card via
  *  snd_card_file_add() function.
- *  If all files are removed and the release of the card is
- *  scheduled, it will wake up the the thread to call snd_card_free()
- *  (see snd_card_free_in_thread() function).
+ *  If all files are removed and snd_card_free_when_closed() was
+ *  called beforehand, it processes the pending release of
+ *  resources.
  *
  *  Returns zero or a negative error code.
  */
diff --git a/sound/drivers/mpu401/mpu401.c b/sound/drivers/mpu401/mpu401.c
index 17cc105b26fc..2de181ad0b05 100644
--- a/sound/drivers/mpu401/mpu401.c
+++ b/sound/drivers/mpu401/mpu401.c
@@ -211,7 +211,7 @@ static void __devexit snd_mpu401_pnp_remove(struct pnp_dev *dev)
 	struct snd_card *card = (struct snd_card *) pnp_get_drvdata(dev);
 
 	snd_card_disconnect(card);
-	snd_card_free_in_thread(card);
+	snd_card_free_when_closed(card);
 }
 
 static struct pnp_driver snd_mpu401_pnp_driver = {
diff --git a/sound/pcmcia/pdaudiocf/pdaudiocf.c b/sound/pcmcia/pdaudiocf/pdaudiocf.c
index 1c09e5f49da8..fd3590fcaedb 100644
--- a/sound/pcmcia/pdaudiocf/pdaudiocf.c
+++ b/sound/pcmcia/pdaudiocf/pdaudiocf.c
@@ -206,7 +206,7 @@ static void snd_pdacf_detach(struct pcmcia_device *link)
 		snd_pdacf_powerdown(chip);
 	chip->chip_status |= PDAUDIOCF_STAT_IS_STALE; /* to be sure */
 	snd_card_disconnect(chip->card);
-	snd_card_free_in_thread(chip->card);
+	snd_card_free_when_closed(chip->card);
 }
 
 /*
diff --git a/sound/pcmcia/vx/vxpocket.c b/sound/pcmcia/vx/vxpocket.c
index cafe6640cc1a..76c85cffb40e 100644
--- a/sound/pcmcia/vx/vxpocket.c
+++ b/sound/pcmcia/vx/vxpocket.c
@@ -65,7 +65,7 @@ static void vxpocket_release(struct pcmcia_device *link)
 }
 
 /*
- * destructor, called from snd_card_free_in_thread()
+ * destructor, called from snd_card_free_when_closed()
  */
 static int snd_vxpocket_dev_free(struct snd_device *device)
 {
@@ -363,7 +363,7 @@ static void vxpocket_detach(struct pcmcia_device *link)
 	chip->chip_status |= VX_STAT_IS_STALE; /* to be sure */
 	snd_card_disconnect(chip->card);
 	vxpocket_release(link);
-	snd_card_free_in_thread(chip->card);
+	snd_card_free_when_closed(chip->card);
 }
 
 /*
-- 
cgit v1.2.3


From 6dbe662874ba08585eaf732d126762c25ac8e3f7 Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Tue, 27 Jun 2006 18:28:53 +0200
Subject: [ALSA] Add experimental support of aggressive AC97 power-saving mode

Added CONFIG_SND_AC97_POWER_SAVE kernel config to enable the support
of aggressive AC97 power-saving mode.  In this mode, the AC97
powerdown register bits are dynamically controlled at each open/close
of PCM streams.
The mode is activated via power_save option for snd-ac97-codec
driver.  As default it's off.  It can be turned on/off on the fly
via sysfs, too.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 include/sound/ac97_codec.h  |  32 ++++++
 sound/drivers/Kconfig       |  13 +++
 sound/pci/ac97/ac97_codec.c | 264 +++++++++++++++++++++++++++++++++++++-------
 sound/pci/ac97/ac97_pcm.c   |  18 +++
 sound/pci/intel8x0.c        |  14 ++-
 sound/pci/via82xx.c         |  13 ++-
 6 files changed, 315 insertions(+), 39 deletions(-)

(limited to 'include')

diff --git a/include/sound/ac97_codec.h b/include/sound/ac97_codec.h
index 758f8bf133c7..4c43521cc493 100644
--- a/include/sound/ac97_codec.h
+++ b/include/sound/ac97_codec.h
@@ -27,6 +27,7 @@
 
 #include <linux/bitops.h>
 #include <linux/device.h>
+#include <linux/workqueue.h>
 #include "pcm.h"
 #include "control.h"
 #include "info.h"
@@ -140,6 +141,20 @@
 #define AC97_GP_DRSS_1011	0x0000	/* LR(C) 10+11(+12) */
 #define AC97_GP_DRSS_78		0x0400	/* LR 7+8 */
 
+/* powerdown bits */
+#define AC97_PD_ADC_STATUS	0x0001	/* ADC status (RO) */
+#define AC97_PD_DAC_STATUS	0x0002	/* DAC status (RO) */
+#define AC97_PD_MIXER_STATUS	0x0004	/* Analog mixer status (RO) */
+#define AC97_PD_VREF_STATUS	0x0008	/* Vref status (RO) */
+#define AC97_PD_PR0		0x0100	/* Power down PCM ADCs and input MUX */
+#define AC97_PD_PR1		0x0200	/* Power down PCM front DAC */
+#define AC97_PD_PR2		0x0400	/* Power down Mixer (Vref still on) */
+#define AC97_PD_PR3		0x0800	/* Power down Mixer (Vref off) */
+#define AC97_PD_PR4		0x1000	/* Power down AC-Link */
+#define AC97_PD_PR5		0x2000	/* Disable internal clock usage */
+#define AC97_PD_PR6		0x4000	/* Headphone amplifier */
+#define AC97_PD_EAPD		0x8000	/* External Amplifer Power Down (EAPD) */
+
 /* extended audio ID bit defines */
 #define AC97_EI_VRA		0x0001	/* Variable bit rate supported */
 #define AC97_EI_DRA		0x0002	/* Double rate supported */
@@ -359,6 +374,7 @@
 #define AC97_SCAP_INV_EAPD	(1<<7)	/* inverted EAPD */
 #define AC97_SCAP_DETECT_BY_VENDOR (1<<8) /* use vendor registers for read tests */
 #define AC97_SCAP_NO_SPDIF	(1<<9)	/* don't build SPDIF controls */
+#define AC97_SCAP_EAPD_LED	(1<<10)	/* EAPD as mute LED */
 
 /* ac97->flags */
 #define AC97_HAS_PC_BEEP	(1<<0)	/* force PC Speaker usage */
@@ -491,6 +507,12 @@ struct snd_ac97 {
 	/* jack-sharing info */
 	unsigned char indep_surround;
 	unsigned char channel_mode;
+
+#ifdef CONFIG_SND_AC97_POWER_SAVE
+	unsigned int power_up;	/* power states */
+	struct workqueue_struct *power_workq;
+	struct work_struct power_work;
+#endif
 	struct device dev;
 };
 
@@ -532,6 +554,15 @@ unsigned short snd_ac97_read(struct snd_ac97 *ac97, unsigned short reg);
 void snd_ac97_write_cache(struct snd_ac97 *ac97, unsigned short reg, unsigned short value);
 int snd_ac97_update(struct snd_ac97 *ac97, unsigned short reg, unsigned short value);
 int snd_ac97_update_bits(struct snd_ac97 *ac97, unsigned short reg, unsigned short mask, unsigned short value);
+#ifdef CONFIG_SND_AC97_POWER_SAVE
+int snd_ac97_update_power(struct snd_ac97 *ac97, int reg, int powerup);
+#else
+static inline int snd_ac97_update_power(struct snd_ac97 *ac97, int reg,
+					int powerup)
+{
+	return 0;
+}
+#endif
 #ifdef CONFIG_PM
 void snd_ac97_suspend(struct snd_ac97 *ac97);
 void snd_ac97_resume(struct snd_ac97 *ac97);
@@ -583,6 +614,7 @@ struct ac97_pcm {
 		     copy_flag: 1,	   /* lowlevel driver must fill all entries */
 		     spdif: 1;		   /* spdif pcm */
 	unsigned short aslots;		   /* active slots */
+	unsigned short cur_dbl;		   /* current double-rate state */
 	unsigned int rates;		   /* available rates */
 	struct {
 		unsigned short slots;	   /* driver input: requested AC97 slot numbers */
diff --git a/sound/drivers/Kconfig b/sound/drivers/Kconfig
index 395c4ef52ac9..897dc2dfd7dd 100644
--- a/sound/drivers/Kconfig
+++ b/sound/drivers/Kconfig
@@ -100,4 +100,17 @@ config SND_MPU401
 	  To compile this driver as a module, choose M here: the module
 	  will be called snd-mpu401.
 
+config SND_AC97_POWER_SAVE
+	bool "AC97 Power-Saving Mode"
+	depends on SND_AC97_CODEC && EXPERIMENTAL
+	default n
+	help
+	  Say Y here to enable the aggressive power-saving support of
+	  AC97 codecs.  In this mode, the power-mode is dynamically
+	  controlled at each open/close.
+
+	  The mode is activated by passing power_save=1 option to
+	  snd-ac97-codec driver.  You can toggle it dynamically over
+	  sysfs, too.
+
 endmenu
diff --git a/sound/pci/ac97/ac97_codec.c b/sound/pci/ac97/ac97_codec.c
index b35280ca2465..f82c636e99a9 100644
--- a/sound/pci/ac97/ac97_codec.c
+++ b/sound/pci/ac97/ac97_codec.c
@@ -47,6 +47,11 @@ static int enable_loopback;
 module_param(enable_loopback, bool, 0444);
 MODULE_PARM_DESC(enable_loopback, "Enable AC97 ADC/DAC Loopback Control");
 
+#ifdef CONFIG_SND_AC97_POWER_SAVE
+static int power_save;
+module_param(power_save, bool, 0644);
+MODULE_PARM_DESC(power_save, "Enable AC97 power-saving control");
+#endif
 /*
 
  */
@@ -187,6 +192,8 @@ static const struct ac97_codec_id snd_ac97_codec_ids[] = {
 };
 
 
+static void update_power_regs(struct snd_ac97 *ac97);
+
 /*
  *  I/O routines
  */
@@ -554,6 +561,18 @@ int snd_ac97_put_volsw(struct snd_kcontrol *kcontrol, struct snd_ctl_elem_value
 	}
 	err = snd_ac97_update_bits(ac97, reg, val_mask, val);
 	snd_ac97_page_restore(ac97, page_save);
+#ifdef CONFIG_SND_AC97_POWER_SAVE
+	/* check analog mixer power-down */
+	if ((val_mask & 0x8000) &&
+	    (kcontrol->private_value & (1<<30))) {
+		if (val & 0x8000)
+			ac97->power_up &= ~(1 << (reg>>1));
+		else
+			ac97->power_up |= 1 << (reg>>1);
+		if (power_save)
+			update_power_regs(ac97);
+	}
+#endif
 	return err;
 }
 
@@ -962,6 +981,10 @@ static int snd_ac97_bus_dev_free(struct snd_device *device)
 static int snd_ac97_free(struct snd_ac97 *ac97)
 {
 	if (ac97) {
+#ifdef CONFIG_SND_AC97_POWER_SAVE
+		if (ac97->power_workq)
+			destroy_workqueue(ac97->power_workq);
+#endif
 		snd_ac97_proc_done(ac97);
 		if (ac97->bus)
 			ac97->bus->codec[ac97->num] = NULL;
@@ -1117,7 +1140,9 @@ struct snd_kcontrol *snd_ac97_cnew(const struct snd_kcontrol_new *_template, str
 /*
  * create mute switch(es) for normal stereo controls
  */
-static int snd_ac97_cmute_new_stereo(struct snd_card *card, char *name, int reg, int check_stereo, struct snd_ac97 *ac97)
+static int snd_ac97_cmute_new_stereo(struct snd_card *card, char *name, int reg,
+				     int check_stereo, int check_amix,
+				     struct snd_ac97 *ac97)
 {
 	struct snd_kcontrol *kctl;
 	int err;
@@ -1137,10 +1162,14 @@ static int snd_ac97_cmute_new_stereo(struct snd_card *card, char *name, int reg,
 	}
 	if (mute_mask == 0x8080) {
 		struct snd_kcontrol_new tmp = AC97_DOUBLE(name, reg, 15, 7, 1, 1);
+		if (check_amix)
+			tmp.private_value |= (1 << 30);
 		tmp.index = ac97->num;
 		kctl = snd_ctl_new1(&tmp, ac97);
 	} else {
 		struct snd_kcontrol_new tmp = AC97_SINGLE(name, reg, 15, 1, 1);
+		if (check_amix)
+			tmp.private_value |= (1 << 30);
 		tmp.index = ac97->num;
 		kctl = snd_ctl_new1(&tmp, ac97);
 	}
@@ -1186,7 +1215,9 @@ static int snd_ac97_cvol_new(struct snd_card *card, char *name, int reg, unsigne
 /*
  * create a mute-switch and a volume for normal stereo/mono controls
  */
-static int snd_ac97_cmix_new_stereo(struct snd_card *card, const char *pfx, int reg, int check_stereo, struct snd_ac97 *ac97)
+static int snd_ac97_cmix_new_stereo(struct snd_card *card, const char *pfx,
+				    int reg, int check_stereo, int check_amix,
+				    struct snd_ac97 *ac97)
 {
 	int err;
 	char name[44];
@@ -1197,7 +1228,9 @@ static int snd_ac97_cmix_new_stereo(struct snd_card *card, const char *pfx, int
 
 	if (snd_ac97_try_bit(ac97, reg, 15)) {
 		sprintf(name, "%s Switch", pfx);
-		if ((err = snd_ac97_cmute_new_stereo(card, name, reg, check_stereo, ac97)) < 0)
+		if ((err = snd_ac97_cmute_new_stereo(card, name, reg,
+						     check_stereo, check_amix,
+						     ac97)) < 0)
 			return err;
 	}
 	check_volume_resolution(ac97, reg, &lo_max, &hi_max);
@@ -1209,8 +1242,10 @@ static int snd_ac97_cmix_new_stereo(struct snd_card *card, const char *pfx, int
 	return 0;
 }
 
-#define snd_ac97_cmix_new(card, pfx, reg, ac97)	snd_ac97_cmix_new_stereo(card, pfx, reg, 0, ac97)
-#define snd_ac97_cmute_new(card, name, reg, ac97)	snd_ac97_cmute_new_stereo(card, name, reg, 0, ac97)
+#define snd_ac97_cmix_new(card, pfx, reg, acheck, ac97) \
+	snd_ac97_cmix_new_stereo(card, pfx, reg, 0, acheck, ac97)
+#define snd_ac97_cmute_new(card, name, reg, acheck, ac97) \
+	snd_ac97_cmute_new_stereo(card, name, reg, 0, acheck, ac97)
 
 static unsigned int snd_ac97_determine_spdif_rates(struct snd_ac97 *ac97);
 
@@ -1226,9 +1261,11 @@ static int snd_ac97_mixer_build(struct snd_ac97 * ac97)
 	/* AD claims to remove this control from AD1887, although spec v2.2 does not allow this */
 	if (snd_ac97_try_volume_mix(ac97, AC97_MASTER)) {
 		if (ac97->flags & AC97_HAS_NO_MASTER_VOL)
-			err = snd_ac97_cmute_new(card, "Master Playback Switch", AC97_MASTER, ac97);
+			err = snd_ac97_cmute_new(card, "Master Playback Switch",
+						 AC97_MASTER, 0, ac97);
 		else
-			err = snd_ac97_cmix_new(card, "Master Playback", AC97_MASTER, ac97);
+			err = snd_ac97_cmix_new(card, "Master Playback",
+						AC97_MASTER, 0, ac97);
 		if (err < 0)
 			return err;
 	}
@@ -1265,19 +1302,23 @@ static int snd_ac97_mixer_build(struct snd_ac97 * ac97)
 	if ((snd_ac97_try_volume_mix(ac97, AC97_SURROUND_MASTER)) 
 		&& !(ac97->flags & AC97_AD_MULTI)) {
 		/* Surround Master (0x38) is with stereo mutes */
-		if ((err = snd_ac97_cmix_new_stereo(card, "Surround Playback", AC97_SURROUND_MASTER, 1, ac97)) < 0)
+		if ((err = snd_ac97_cmix_new_stereo(card, "Surround Playback",
+						    AC97_SURROUND_MASTER, 1, 0,
+						    ac97)) < 0)
 			return err;
 	}
 
 	/* build headphone controls */
 	if (snd_ac97_try_volume_mix(ac97, AC97_HEADPHONE)) {
-		if ((err = snd_ac97_cmix_new(card, "Headphone Playback", AC97_HEADPHONE, ac97)) < 0)
+		if ((err = snd_ac97_cmix_new(card, "Headphone Playback",
+					     AC97_HEADPHONE, 0, ac97)) < 0)
 			return err;
 	}
 	
 	/* build master mono controls */
 	if (snd_ac97_try_volume_mix(ac97, AC97_MASTER_MONO)) {
-		if ((err = snd_ac97_cmix_new(card, "Master Mono Playback", AC97_MASTER_MONO, ac97)) < 0)
+		if ((err = snd_ac97_cmix_new(card, "Master Mono Playback",
+					     AC97_MASTER_MONO, 0, ac97)) < 0)
 			return err;
 	}
 	
@@ -1310,7 +1351,8 @@ static int snd_ac97_mixer_build(struct snd_ac97 * ac97)
 	/* build Phone controls */
 	if (!(ac97->flags & AC97_HAS_NO_PHONE)) {
 		if (snd_ac97_try_volume_mix(ac97, AC97_PHONE)) {
-			if ((err = snd_ac97_cmix_new(card, "Phone Playback", AC97_PHONE, ac97)) < 0)
+			if ((err = snd_ac97_cmix_new(card, "Phone Playback",
+						     AC97_PHONE, 1, ac97)) < 0)
 				return err;
 		}
 	}
@@ -1318,7 +1360,8 @@ static int snd_ac97_mixer_build(struct snd_ac97 * ac97)
 	/* build MIC controls */
 	if (!(ac97->flags & AC97_HAS_NO_MIC)) {
 		if (snd_ac97_try_volume_mix(ac97, AC97_MIC)) {
-			if ((err = snd_ac97_cmix_new(card, "Mic Playback", AC97_MIC, ac97)) < 0)
+			if ((err = snd_ac97_cmix_new(card, "Mic Playback",
+						     AC97_MIC, 1, ac97)) < 0)
 				return err;
 			if ((err = snd_ctl_add(card, snd_ac97_cnew(&snd_ac97_controls_mic_boost, ac97))) < 0)
 				return err;
@@ -1327,14 +1370,16 @@ static int snd_ac97_mixer_build(struct snd_ac97 * ac97)
 
 	/* build Line controls */
 	if (snd_ac97_try_volume_mix(ac97, AC97_LINE)) {
-		if ((err = snd_ac97_cmix_new(card, "Line Playback", AC97_LINE, ac97)) < 0)
+		if ((err = snd_ac97_cmix_new(card, "Line Playback",
+					     AC97_LINE, 1, ac97)) < 0)
 			return err;
 	}
 	
 	/* build CD controls */
 	if (!(ac97->flags & AC97_HAS_NO_CD)) {
 		if (snd_ac97_try_volume_mix(ac97, AC97_CD)) {
-			if ((err = snd_ac97_cmix_new(card, "CD Playback", AC97_CD, ac97)) < 0)
+			if ((err = snd_ac97_cmix_new(card, "CD Playback",
+						     AC97_CD, 1, ac97)) < 0)
 				return err;
 		}
 	}
@@ -1342,7 +1387,8 @@ static int snd_ac97_mixer_build(struct snd_ac97 * ac97)
 	/* build Video controls */
 	if (!(ac97->flags & AC97_HAS_NO_VIDEO)) {
 		if (snd_ac97_try_volume_mix(ac97, AC97_VIDEO)) {
-			if ((err = snd_ac97_cmix_new(card, "Video Playback", AC97_VIDEO, ac97)) < 0)
+			if ((err = snd_ac97_cmix_new(card, "Video Playback",
+						     AC97_VIDEO, 1, ac97)) < 0)
 				return err;
 		}
 	}
@@ -1350,7 +1396,8 @@ static int snd_ac97_mixer_build(struct snd_ac97 * ac97)
 	/* build Aux controls */
 	if (!(ac97->flags & AC97_HAS_NO_AUX)) {
 		if (snd_ac97_try_volume_mix(ac97, AC97_AUX)) {
-			if ((err = snd_ac97_cmix_new(card, "Aux Playback", AC97_AUX, ac97)) < 0)
+			if ((err = snd_ac97_cmix_new(card, "Aux Playback",
+						     AC97_AUX, 1, ac97)) < 0)
 				return err;
 		}
 	}
@@ -1385,9 +1432,12 @@ static int snd_ac97_mixer_build(struct snd_ac97 * ac97)
 	} else {
 		if (!(ac97->flags & AC97_HAS_NO_STD_PCM)) {
 			if (ac97->flags & AC97_HAS_NO_PCM_VOL)
-				err = snd_ac97_cmute_new(card, "PCM Playback Switch", AC97_PCM, ac97);
+				err = snd_ac97_cmute_new(card,
+							 "PCM Playback Switch",
+							 AC97_PCM, 0, ac97);
 			else
-				err = snd_ac97_cmix_new(card, "PCM Playback", AC97_PCM, ac97);
+				err = snd_ac97_cmix_new(card, "PCM Playback",
+							AC97_PCM, 0, ac97);
 			if (err < 0)
 				return err;
 		}
@@ -1398,7 +1448,9 @@ static int snd_ac97_mixer_build(struct snd_ac97 * ac97)
 		if ((err = snd_ctl_add(card, snd_ac97_cnew(&snd_ac97_control_capture_src, ac97))) < 0)
 			return err;
 		if (snd_ac97_try_bit(ac97, AC97_REC_GAIN, 15)) {
-			if ((err = snd_ac97_cmute_new(card, "Capture Switch", AC97_REC_GAIN, ac97)) < 0)
+			err = snd_ac97_cmute_new(card, "Capture Switch",
+						 AC97_REC_GAIN, 0, ac97);
+			if (err < 0)
 				return err;
 		}
 		if ((err = snd_ctl_add(card, snd_ac97_cnew(&snd_ac97_control_capture_vol, ac97))) < 0)
@@ -1829,6 +1881,13 @@ static int snd_ac97_dev_disconnect(struct snd_device *device)
 /* build_ops to do nothing */
 static struct snd_ac97_build_ops null_build_ops;
 
+#ifdef CONFIG_SND_AC97_POWER_SAVE
+static void do_update_power(void *data)
+{
+	update_power_regs(data);
+}
+#endif
+
 /**
  * snd_ac97_mixer - create an Codec97 component
  * @bus: the AC97 bus which codec is attached to
@@ -1883,6 +1942,10 @@ int snd_ac97_mixer(struct snd_ac97_bus *bus, struct snd_ac97_template *template,
 	bus->codec[ac97->num] = ac97;
 	mutex_init(&ac97->reg_mutex);
 	mutex_init(&ac97->page_mutex);
+#ifdef CONFIG_SND_AC97_POWER_SAVE
+	ac97->power_workq = create_workqueue("ac97");
+	INIT_WORK(&ac97->power_work, do_update_power, ac97);
+#endif
 
 #ifdef CONFIG_PCI
 	if (ac97->pci) {
@@ -2117,15 +2180,8 @@ int snd_ac97_mixer(struct snd_ac97_bus *bus, struct snd_ac97_template *template,
 			return -ENOMEM;
 		}
 	}
-	/* make sure the proper powerdown bits are cleared */
-	if (ac97->scaps && ac97_is_audio(ac97)) {
-		reg = snd_ac97_read(ac97, AC97_EXTENDED_STATUS);
-		if (ac97->scaps & AC97_SCAP_SURROUND_DAC) 
-			reg &= ~AC97_EA_PRJ;
-		if (ac97->scaps & AC97_SCAP_CENTER_LFE_DAC) 
-			reg &= ~(AC97_EA_PRI | AC97_EA_PRK);
-		snd_ac97_write_cache(ac97, AC97_EXTENDED_STATUS, reg);
-	}
+	if (ac97_is_audio(ac97))
+		update_power_regs(ac97);
 	snd_ac97_proc_init(ac97);
 	if ((err = snd_device_new(card, SNDRV_DEV_CODEC, ac97, &ops)) < 0) {
 		snd_ac97_free(ac97);
@@ -2153,22 +2209,155 @@ static void snd_ac97_powerdown(struct snd_ac97 *ac97)
 		snd_ac97_write(ac97, AC97_HEADPHONE, 0x9f9f);
 	}
 
-	power = ac97->regs[AC97_POWERDOWN] | 0x8000;	/* EAPD */
-	power |= 0x4000;	/* Headphone amplifier powerdown */
-	power |= 0x0300;	/* ADC & DAC powerdown */
+	/* surround, CLFE, mic powerdown */
+	power = ac97->regs[AC97_EXTENDED_STATUS];
+	if (ac97->scaps & AC97_SCAP_SURROUND_DAC)
+		power |= AC97_EA_PRJ;
+	if (ac97->scaps & AC97_SCAP_CENTER_LFE_DAC)
+		power |= AC97_EA_PRI | AC97_EA_PRK;
+	power |= AC97_EA_PRL;
+	snd_ac97_write(ac97, AC97_EXTENDED_STATUS, power);
+
+	/* powerdown external amplifier */
+	if (ac97->scaps & AC97_SCAP_INV_EAPD)
+		power = ac97->regs[AC97_POWERDOWN] & ~AC97_PD_EAPD;
+	else if (! (ac97->scaps & AC97_SCAP_EAPD_LED))
+		power = ac97->regs[AC97_POWERDOWN] | AC97_PD_EAPD;
+	power |= AC97_PD_PR6;	/* Headphone amplifier powerdown */
+	power |= AC97_PD_PR0 | AC97_PD_PR1;	/* ADC & DAC powerdown */
 	snd_ac97_write(ac97, AC97_POWERDOWN, power);
 	udelay(100);
-	power |= 0x0400;	/* Analog Mixer powerdown (Vref on) */
-	snd_ac97_write(ac97, AC97_POWERDOWN, power);
-	udelay(100);
-#if 0
-	/* FIXME: this causes click noises on some boards at resume */
-	power |= 0x3800;	/* AC-link powerdown, internal Clk disable */
+	power |= AC97_PD_PR2 | AC97_PD_PR3;	/* Analog Mixer powerdown */
 	snd_ac97_write(ac97, AC97_POWERDOWN, power);
+#ifdef CONFIG_SND_AC97_POWER_SAVE
+	if (power_save) {
+		udelay(100);
+		/* AC-link powerdown, internal Clk disable */
+		/* FIXME: this may cause click noises on some boards */
+		power |= AC97_PD_PR4 | AC97_PD_PR5;
+		snd_ac97_write(ac97, AC97_POWERDOWN, power);
+	}
 #endif
 }
 
 
+struct ac97_power_reg {
+	unsigned short reg;
+	unsigned short power_reg;
+	unsigned short mask;
+};
+
+enum { PWIDX_ADC, PWIDX_FRONT, PWIDX_CLFE, PWIDX_SURR, PWIDX_MIC, PWIDX_SIZE };
+
+static struct ac97_power_reg power_regs[PWIDX_SIZE] = {
+	[PWIDX_ADC] = { AC97_PCM_LR_ADC_RATE, AC97_POWERDOWN, AC97_PD_PR0},
+	[PWIDX_FRONT] = { AC97_PCM_FRONT_DAC_RATE, AC97_POWERDOWN, AC97_PD_PR1},
+	[PWIDX_CLFE] = { AC97_PCM_LFE_DAC_RATE, AC97_EXTENDED_STATUS,
+			 AC97_EA_PRI | AC97_EA_PRK},
+	[PWIDX_SURR] = { AC97_PCM_SURR_DAC_RATE, AC97_EXTENDED_STATUS,
+			 AC97_EA_PRJ},
+	[PWIDX_MIC] = { AC97_PCM_MIC_ADC_RATE, AC97_EXTENDED_STATUS,
+			AC97_EA_PRL},
+};
+
+#ifdef CONFIG_SND_AC97_POWER_SAVE
+/**
+ * snd_ac97_update_power - update the powerdown register
+ * @ac97: the codec instance
+ * @reg: the rate register, e.g. AC97_PCM_FRONT_DAC_RATE
+ * @powerup: non-zero when power up the part
+ *
+ * Update the AC97 powerdown register bits of the given part.
+ */
+int snd_ac97_update_power(struct snd_ac97 *ac97, int reg, int powerup)
+{
+	int i;
+
+	if (! ac97)
+		return 0;
+
+	if (reg) {
+		/* SPDIF requires DAC power, too */
+		if (reg == AC97_SPDIF)
+			reg = AC97_PCM_FRONT_DAC_RATE;
+		for (i = 0; i < PWIDX_SIZE; i++) {
+			if (power_regs[i].reg == reg) {
+				if (powerup)
+					ac97->power_up |= (1 << i);
+				else
+					ac97->power_up &= ~(1 << i);
+				break;
+			}
+		}
+	}
+
+	if (! power_save)
+		return 0;
+
+	if (! powerup && ac97->power_workq)
+		/* adjust power-down bits after two seconds delay
+		 * (for avoiding loud click noises for many (OSS) apps
+		 *  that open/close frequently)
+		 */
+		queue_delayed_work(ac97->power_workq, &ac97->power_work, HZ*2);
+	else
+		update_power_regs(ac97);
+
+	return 0;
+}
+
+EXPORT_SYMBOL(snd_ac97_update_power);
+#endif /* CONFIG_SND_AC97_POWER_SAVE */
+
+static void update_power_regs(struct snd_ac97 *ac97)
+{
+	unsigned int power_up, bits;
+	int i;
+
+#ifdef CONFIG_SND_AC97_POWER_SAVE
+	if (power_save)
+		power_up = ac97->power_up;
+	else {
+#endif
+		power_up = (1 << PWIDX_FRONT) | (1 << PWIDX_ADC);
+		power_up |= (1 << PWIDX_MIC);
+		if (ac97->scaps & AC97_SCAP_SURROUND_DAC)
+			power_up |= (1 << PWIDX_SURR);
+		if (ac97->scaps & AC97_SCAP_CENTER_LFE_DAC)
+			power_up |= (1 << PWIDX_CLFE);
+#ifdef CONFIG_SND_AC97_POWER_SAVE
+	}
+#endif
+	if (power_up) {
+		if (ac97->regs[AC97_POWERDOWN] & AC97_PD_PR2) {
+			/* needs power-up analog mix and vref */
+			snd_ac97_update_bits(ac97, AC97_POWERDOWN,
+					     AC97_PD_PR3, 0);
+			msleep(1);
+			snd_ac97_update_bits(ac97, AC97_POWERDOWN,
+					     AC97_PD_PR2, 0);
+		}
+	}
+	for (i = 0; i < PWIDX_SIZE; i++) {
+		if (power_up & (1 << i))
+			bits = 0;
+		else
+			bits = power_regs[i].mask;
+		snd_ac97_update_bits(ac97, power_regs[i].power_reg,
+				     power_regs[i].mask, bits);
+	}
+	if (! power_up) {
+		if (! (ac97->regs[AC97_POWERDOWN] & AC97_PD_PR2)) {
+			/* power down analog mix and vref */
+			snd_ac97_update_bits(ac97, AC97_POWERDOWN,
+					     AC97_PD_PR2, AC97_PD_PR2);
+			snd_ac97_update_bits(ac97, AC97_POWERDOWN,
+					     AC97_PD_PR3, AC97_PD_PR3);
+		}
+	}
+}
+
+
 #ifdef CONFIG_PM
 /**
  * snd_ac97_suspend - General suspend function for AC97 codec
@@ -2484,6 +2673,7 @@ static int tune_mute_led(struct snd_ac97 *ac97)
 	msw->put = master_mute_sw_put;
 	snd_ac97_remove_ctl(ac97, "External Amplifier", NULL);
 	snd_ac97_update_bits(ac97, AC97_POWERDOWN, 0x8000, 0x8000); /* mute LED on */
+	ac97->scaps |= AC97_SCAP_EAPD_LED;
 	return 0;
 }
 
diff --git a/sound/pci/ac97/ac97_pcm.c b/sound/pci/ac97/ac97_pcm.c
index f684aa2c0067..3758d07182f8 100644
--- a/sound/pci/ac97/ac97_pcm.c
+++ b/sound/pci/ac97/ac97_pcm.c
@@ -269,6 +269,7 @@ int snd_ac97_set_rate(struct snd_ac97 *ac97, int reg, unsigned int rate)
 			return -EINVAL;
 	}
 
+	snd_ac97_update_power(ac97, reg, 1);
 	switch (reg) {
 	case AC97_PCM_MIC_ADC_RATE:
 		if ((ac97->regs[AC97_EXTENDED_STATUS] & AC97_EA_VRM) == 0)	/* MIC VRA */
@@ -606,6 +607,7 @@ int snd_ac97_pcm_open(struct ac97_pcm *pcm, unsigned int rate,
 			goto error;
 		}
 	}
+	pcm->cur_dbl = r;
 	spin_unlock_irq(&pcm->bus->bus_lock);
 	for (i = 3; i < 12; i++) {
 		if (!(slots & (1 << i)))
@@ -651,6 +653,21 @@ int snd_ac97_pcm_close(struct ac97_pcm *pcm)
 	unsigned short slots = pcm->aslots;
 	int i, cidx;
 
+#ifdef CONFIG_SND_AC97_POWER_SAVE
+	int r = pcm->cur_dbl;
+	for (i = 3; i < 12; i++) {
+		if (!(slots & (1 << i)))
+			continue;
+		for (cidx = 0; cidx < 4; cidx++) {
+			if (pcm->r[r].rslots[cidx] & (1 << i)) {
+				int reg = get_slot_reg(pcm, cidx, i, r);
+				snd_ac97_update_power(pcm->r[r].codec[cidx],
+						      reg, 0);
+			}
+		}
+	}
+#endif
+
 	bus = pcm->bus;
 	spin_lock_irq(&pcm->bus->bus_lock);
 	for (i = 3; i < 12; i++) {
@@ -660,6 +677,7 @@ int snd_ac97_pcm_close(struct ac97_pcm *pcm)
 			bus->used_slots[pcm->stream][cidx] &= ~(1 << i);
 	}
 	pcm->aslots = 0;
+	pcm->cur_dbl = 0;
 	spin_unlock_irq(&pcm->bus->bus_lock);
 	return 0;
 }
diff --git a/sound/pci/intel8x0.c b/sound/pci/intel8x0.c
index 6874263f1681..72dbaedcbdf5 100644
--- a/sound/pci/intel8x0.c
+++ b/sound/pci/intel8x0.c
@@ -2251,6 +2251,16 @@ static int snd_intel8x0_ich_chip_init(struct intel8x0 *chip, int probing)
 	/* ACLink on, 2 channels */
 	cnt = igetdword(chip, ICHREG(GLOB_CNT));
 	cnt &= ~(ICH_ACLINK | ICH_PCM_246_MASK);
+#ifdef CONFIG_SND_AC97_POWER_SAVE
+	/* do cold reset - the full ac97 powerdown may leave the controller
+	 * in a warm state but actually it cannot communicate with the codec.
+	 */
+	iputdword(chip, ICHREG(GLOB_CNT), cnt & ~ICH_AC97COLD);
+	cnt = igetdword(chip, ICHREG(GLOB_CNT));
+	udelay(10);
+	iputdword(chip, ICHREG(GLOB_CNT), cnt | ICH_AC97COLD);
+	msleep(1);
+#else
 	/* finish cold or do warm reset */
 	cnt |= (cnt & ICH_AC97COLD) == 0 ? ICH_AC97COLD : ICH_AC97WARM;
 	iputdword(chip, ICHREG(GLOB_CNT), cnt);
@@ -2265,6 +2275,7 @@ static int snd_intel8x0_ich_chip_init(struct intel8x0 *chip, int probing)
 	return -EIO;
 
       __ok:
+#endif
 	if (probing) {
 		/* wait for any codec ready status.
 		 * Once it becomes ready it should remain ready
@@ -2485,7 +2496,7 @@ static int intel8x0_resume(struct pci_dev *pci)
 		    card->shortname, chip);
 	chip->irq = pci->irq;
 	synchronize_irq(chip->irq);
-	snd_intel8x0_chip_init(chip, 1);
+	snd_intel8x0_chip_init(chip, 0);
 
 	/* re-initialize mixer stuff */
 	if (chip->device_type == DEVICE_INTEL_ICH4) {
@@ -2615,6 +2626,7 @@ static void __devinit intel8x0_measure_ac97_clock(struct intel8x0 *chip)
 		/* not 48000Hz, tuning the clock.. */
 		chip->ac97_bus->clock = (chip->ac97_bus->clock * 48000) / pos;
 	printk(KERN_INFO "intel8x0: clocking to %d\n", chip->ac97_bus->clock);
+	snd_ac97_update_power(chip->ac97[0], AC97_PCM_FRONT_DAC_RATE, 0);
 }
 
 #ifdef CONFIG_PROC_FS
diff --git a/sound/pci/via82xx.c b/sound/pci/via82xx.c
index 08da9234efb3..2c23a665c3e3 100644
--- a/sound/pci/via82xx.c
+++ b/sound/pci/via82xx.c
@@ -1277,7 +1277,18 @@ static int snd_via82xx_pcm_close(struct snd_pcm_substream *substream)
 	if (! ratep->used)
 		ratep->rate = 0;
 	spin_unlock_irq(&ratep->lock);
-
+	if (! ratep->rate) {
+		if (! viadev->direction) {
+			snd_ac97_update_power(chip->ac97,
+					      AC97_PCM_FRONT_DAC_RATE, 0);
+			snd_ac97_update_power(chip->ac97,
+					      AC97_PCM_SURR_DAC_RATE, 0);
+			snd_ac97_update_power(chip->ac97,
+					      AC97_PCM_LFE_DAC_RATE, 0);
+		} else
+			snd_ac97_update_power(chip->ac97,
+					      AC97_PCM_LR_ADC_RATE, 0);
+	}
 	viadev->substream = NULL;
 	return 0;
 }
-- 
cgit v1.2.3


From 8aa9b586e42099817163aba01d925c2660c4dbbe Mon Sep 17 00:00:00 2001
From: Jaroslav Kysela <perex@suse.cz>
Date: Wed, 5 Jul 2006 17:34:51 +0200
Subject: [ALSA] Control API - more robust TLV implementation

- added callback option
- added READ/WRITE/COMMAND flags to access member
- added WRITE/COMMAND ioctls
- added SNDRV_CTL_EVENT_MASK_TLV for TLV change notifications
- added TLV support to ELEM_ADD ioctl

Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 include/sound/asound.h  |  10 +++-
 include/sound/control.h |  16 +++++-
 sound/core/control.c    | 139 +++++++++++++++++++++++++++++++++++++-----------
 3 files changed, 132 insertions(+), 33 deletions(-)

(limited to 'include')

diff --git a/include/sound/asound.h b/include/sound/asound.h
index 76a20406bd18..c1621c650a9a 100644
--- a/include/sound/asound.h
+++ b/include/sound/asound.h
@@ -727,10 +727,15 @@ typedef int __bitwise snd_ctl_elem_iface_t;
 #define SNDRV_CTL_ELEM_ACCESS_WRITE		(1<<1)
 #define SNDRV_CTL_ELEM_ACCESS_READWRITE		(SNDRV_CTL_ELEM_ACCESS_READ|SNDRV_CTL_ELEM_ACCESS_WRITE)
 #define SNDRV_CTL_ELEM_ACCESS_VOLATILE		(1<<2)	/* control value may be changed without a notification */
-#define SNDRV_CTL_ELEM_ACCESS_TIMESTAMP		(1<<2)	/* when was control changed */
+#define SNDRV_CTL_ELEM_ACCESS_TIMESTAMP		(1<<3)	/* when was control changed */
+#define SNDRV_CTL_ELEM_ACCESS_TLV_READ		(1<<4)	/* TLV read is possible */
+#define SNDRV_CTL_ELEM_ACCESS_TLV_WRITE		(1<<5)	/* TLV write is possible */
+#define SNDRV_CTL_ELEM_ACCESS_TLV_READWRITE	(SNDRV_CTL_ELEM_ACCESS_TLV_READ|SNDRV_CTL_ELEM_ACCESS_TLV_WRITE)
+#define SNDRV_CTL_ELEM_ACCESS_TLV_COMMAND	(1<<6)	/* TLV command is possible */
 #define SNDRV_CTL_ELEM_ACCESS_INACTIVE		(1<<8)	/* control does actually nothing, but may be updated */
 #define SNDRV_CTL_ELEM_ACCESS_LOCK		(1<<9)	/* write lock */
 #define SNDRV_CTL_ELEM_ACCESS_OWNER		(1<<10)	/* write lock owner */
+#define SNDRV_CTL_ELEM_ACCESS_TLV_CALLBACK	(1<<28)	/* kernel use a TLV callback */ 
 #define SNDRV_CTL_ELEM_ACCESS_USER		(1<<29) /* user space element */
 #define SNDRV_CTL_ELEM_ACCESS_DINDIRECT		(1<<30)	/* indirect access for matrix dimensions in the info structure */
 #define SNDRV_CTL_ELEM_ACCESS_INDIRECT		(1<<31)	/* indirect access for element value in the value structure */
@@ -838,6 +843,8 @@ enum {
 	SNDRV_CTL_IOCTL_ELEM_REPLACE = _IOWR('U', 0x18, struct snd_ctl_elem_info),
 	SNDRV_CTL_IOCTL_ELEM_REMOVE = _IOWR('U', 0x19, struct snd_ctl_elem_id),
 	SNDRV_CTL_IOCTL_TLV_READ = _IOWR('U', 0x1a, struct snd_ctl_tlv),
+	SNDRV_CTL_IOCTL_TLV_WRITE = _IOWR('U', 0x1b, struct snd_ctl_tlv),
+	SNDRV_CTL_IOCTL_TLV_COMMAND = _IOWR('U', 0x1c, struct snd_ctl_tlv),
 	SNDRV_CTL_IOCTL_HWDEP_NEXT_DEVICE = _IOWR('U', 0x20, int),
 	SNDRV_CTL_IOCTL_HWDEP_INFO = _IOR('U', 0x21, struct snd_hwdep_info),
 	SNDRV_CTL_IOCTL_PCM_NEXT_DEVICE = _IOR('U', 0x30, int),
@@ -862,6 +869,7 @@ enum sndrv_ctl_event_type {
 #define SNDRV_CTL_EVENT_MASK_VALUE	(1<<0)	/* element value was changed */
 #define SNDRV_CTL_EVENT_MASK_INFO	(1<<1)	/* element info was changed */
 #define SNDRV_CTL_EVENT_MASK_ADD	(1<<2)	/* element was added */
+#define SNDRV_CTL_EVENT_MASK_TLV	(1<<3)	/* element TLV tree was changed */
 #define SNDRV_CTL_EVENT_MASK_REMOVE	(~0U)	/* element was removed */
 
 struct snd_ctl_event {
diff --git a/include/sound/control.h b/include/sound/control.h
index a93a58d0e688..e3905c5a0950 100644
--- a/include/sound/control.h
+++ b/include/sound/control.h
@@ -30,6 +30,11 @@ struct snd_kcontrol;
 typedef int (snd_kcontrol_info_t) (struct snd_kcontrol * kcontrol, struct snd_ctl_elem_info * uinfo);
 typedef int (snd_kcontrol_get_t) (struct snd_kcontrol * kcontrol, struct snd_ctl_elem_value * ucontrol);
 typedef int (snd_kcontrol_put_t) (struct snd_kcontrol * kcontrol, struct snd_ctl_elem_value * ucontrol);
+typedef int (snd_kcontrol_tlv_rw_t)(struct snd_kcontrol *kcontrol,
+				    int op_flag, /* 0=read,1=write,-1=command */
+				    unsigned int size,
+				    unsigned int __user *tlv);
+
 
 struct snd_kcontrol_new {
 	snd_ctl_elem_iface_t iface;	/* interface identifier */
@@ -42,7 +47,10 @@ struct snd_kcontrol_new {
 	snd_kcontrol_info_t *info;
 	snd_kcontrol_get_t *get;
 	snd_kcontrol_put_t *put;
-	unsigned int *tlv;
+	union {
+		snd_kcontrol_tlv_rw_t *c;
+		unsigned int *p;
+	} tlv;
 	unsigned long private_value;
 };
 
@@ -59,7 +67,11 @@ struct snd_kcontrol {
 	snd_kcontrol_info_t *info;
 	snd_kcontrol_get_t *get;
 	snd_kcontrol_put_t *put;
-	unsigned int *tlv;
+	snd_kcontrol_tlv_rw_t *tlv_rw;
+	union {
+		snd_kcontrol_tlv_rw_t *c;
+		unsigned int *p;
+	} tlv;
 	unsigned long private_value;
 	void *private_data;
 	void (*private_free)(struct snd_kcontrol *kcontrol);
diff --git a/sound/core/control.c b/sound/core/control.c
index f0c7272a2d48..31ad58154c06 100644
--- a/sound/core/control.c
+++ b/sound/core/control.c
@@ -236,12 +236,16 @@ struct snd_kcontrol *snd_ctl_new1(const struct snd_kcontrol_new *ncontrol,
 	kctl.id.index = ncontrol->index;
 	kctl.count = ncontrol->count ? ncontrol->count : 1;
 	access = ncontrol->access == 0 ? SNDRV_CTL_ELEM_ACCESS_READWRITE :
-		 (ncontrol->access & (SNDRV_CTL_ELEM_ACCESS_READWRITE|SNDRV_CTL_ELEM_ACCESS_INACTIVE|
-		 		      SNDRV_CTL_ELEM_ACCESS_DINDIRECT|SNDRV_CTL_ELEM_ACCESS_INDIRECT));
+		 (ncontrol->access & (SNDRV_CTL_ELEM_ACCESS_READWRITE|
+				      SNDRV_CTL_ELEM_ACCESS_INACTIVE|
+		 		      SNDRV_CTL_ELEM_ACCESS_DINDIRECT|
+		 		      SNDRV_CTL_ELEM_ACCESS_INDIRECT|
+		 		      SNDRV_CTL_ELEM_ACCESS_TLV_READWRITE|
+		 		      SNDRV_CTL_ELEM_ACCESS_TLV_CALLBACK));
 	kctl.info = ncontrol->info;
 	kctl.get = ncontrol->get;
 	kctl.put = ncontrol->put;
-	kctl.tlv = ncontrol->tlv;
+	kctl.tlv.p = ncontrol->tlv.p;
 	kctl.private_value = ncontrol->private_value;
 	kctl.private_data = private_data;
 	return snd_ctl_new(&kctl, access);
@@ -883,6 +887,8 @@ struct user_element {
 	struct snd_ctl_elem_info info;
 	void *elem_data;		/* element data */
 	unsigned long elem_data_size;	/* size of element data in bytes */
+	void *tlv_data;			/* TLV data */
+	unsigned long tlv_data_size;	/* TLV data size */
 	void *priv_data;		/* private data (like strings for enumerated type) */
 	unsigned long priv_data_size;	/* size of private data in bytes */
 };
@@ -917,9 +923,46 @@ static int snd_ctl_elem_user_put(struct snd_kcontrol *kcontrol,
 	return change;
 }
 
+static int snd_ctl_elem_user_tlv(struct snd_kcontrol *kcontrol,
+				 int op_flag,
+				 unsigned int size,
+				 unsigned int __user *tlv)
+{
+	struct user_element *ue = kcontrol->private_data;
+	int change = 0;
+	void *new_data;
+
+	if (op_flag > 0) {
+		if (size > 1024 * 128)	/* sane value */
+			return -EINVAL;
+		new_data = kmalloc(size, GFP_KERNEL);
+		if (new_data == NULL)
+			return -ENOMEM;
+		if (copy_from_user(new_data, tlv, size)) {
+			kfree(new_data);
+			return -EFAULT;
+		}
+		change = ue->tlv_data_size != size;
+		if (!change)
+			change = memcmp(ue->tlv_data, new_data, size);
+		kfree(ue->tlv_data);
+		ue->tlv_data = new_data;
+		ue->tlv_data_size = size;
+	} else {
+		if (size < ue->tlv_data_size)
+			return -ENOSPC;
+		if (copy_to_user(tlv, ue->tlv_data, ue->tlv_data_size))
+			return -EFAULT;
+	}
+	return change;
+}
+
 static void snd_ctl_elem_user_free(struct snd_kcontrol *kcontrol)
 {
-	kfree(kcontrol->private_data);
+	struct user_element *ue = kcontrol->private_data;
+	if (ue->tlv_data)
+		kfree(ue->tlv_data);
+	kfree(ue);
 }
 
 static int snd_ctl_elem_add(struct snd_ctl_file *file,
@@ -938,7 +981,8 @@ static int snd_ctl_elem_add(struct snd_ctl_file *file,
 		return -EINVAL;
 	access = info->access == 0 ? SNDRV_CTL_ELEM_ACCESS_READWRITE :
 		(info->access & (SNDRV_CTL_ELEM_ACCESS_READWRITE|
-				 SNDRV_CTL_ELEM_ACCESS_INACTIVE));
+				 SNDRV_CTL_ELEM_ACCESS_INACTIVE|
+				 SNDRV_CTL_ELEM_ACCESS_TLV_READWRITE));
 	info->id.numid = 0;
 	memset(&kctl, 0, sizeof(kctl));
 	down_write(&card->controls_rwsem);
@@ -964,6 +1008,10 @@ static int snd_ctl_elem_add(struct snd_ctl_file *file,
 		kctl.get = snd_ctl_elem_user_get;
 	if (access & SNDRV_CTL_ELEM_ACCESS_WRITE)
 		kctl.put = snd_ctl_elem_user_put;
+	if (access & SNDRV_CTL_ELEM_ACCESS_TLV_READWRITE) {
+		kctl.tlv.c = snd_ctl_elem_user_tlv;
+		access |= SNDRV_CTL_ELEM_ACCESS_TLV_CALLBACK;
+	}
 	switch (info->type) {
 	case SNDRV_CTL_ELEM_TYPE_BOOLEAN:
 		private_size = sizeof(char);
@@ -1068,38 +1116,65 @@ static int snd_ctl_subscribe_events(struct snd_ctl_file *file, int __user *ptr)
 	return 0;
 }
 
-static int snd_ctl_tlv_read(struct snd_card *card,
-                            struct snd_ctl_tlv __user *_tlv)
+static int snd_ctl_tlv_ioctl(struct snd_ctl_file *file,
+                             struct snd_ctl_tlv __user *_tlv,
+                             int op_flag)
 {
+	struct snd_card *card = file->card;
 	struct snd_ctl_tlv tlv;
 	struct snd_kcontrol *kctl;
+	struct snd_kcontrol_volatile *vd;
 	unsigned int len;
 	int err = 0;
 
 	if (copy_from_user(&tlv, _tlv, sizeof(tlv)))
 		return -EFAULT;
-        if (tlv.length < sizeof(unsigned int) * 3)
-                return -EINVAL;
-        down_read(&card->controls_rwsem);
-        kctl = snd_ctl_find_numid(card, tlv.numid);
-        if (kctl == NULL) {
-                err = -ENOENT;
-                goto __kctl_end;
-        }
-        if (kctl->tlv == NULL) {
-                err = -ENXIO;
-                goto __kctl_end;
-        }
-        len = kctl->tlv[1] + 2 * sizeof(unsigned int);
-        if (tlv.length < len) {
-                err = -ENOMEM;
-                goto __kctl_end;
-        }
-        if (copy_to_user(_tlv->tlv, kctl->tlv, len))
-        	err = -EFAULT;
+	if (tlv.length < sizeof(unsigned int) * 3)
+		return -EINVAL;
+	down_read(&card->controls_rwsem);
+	kctl = snd_ctl_find_numid(card, tlv.numid);
+	if (kctl == NULL) {
+		err = -ENOENT;
+		goto __kctl_end;
+	}
+	if (kctl->tlv.p == NULL) {
+		err = -ENXIO;
+		goto __kctl_end;
+	}
+	vd = &kctl->vd[tlv.numid - kctl->id.numid];
+	if ((op_flag == 0 && (vd->access & SNDRV_CTL_ELEM_ACCESS_TLV_READ) == 0) ||
+	    (op_flag > 0 && (vd->access & SNDRV_CTL_ELEM_ACCESS_TLV_WRITE) == 0) ||
+	    (op_flag < 0 && (vd->access & SNDRV_CTL_ELEM_ACCESS_TLV_COMMAND) == 0)) {
+	    	err = -ENXIO;
+	    	goto __kctl_end;
+	}
+	if (vd->access & SNDRV_CTL_ELEM_ACCESS_TLV_CALLBACK) {
+		if (file && vd->owner != NULL && vd->owner != file) {
+			err = -EPERM;
+			goto __kctl_end;
+		}
+		err = kctl->tlv.c(kctl, op_flag, tlv.length, _tlv->tlv); 
+		if (err > 0) {
+			up_read(&card->controls_rwsem);
+			snd_ctl_notify(card, SNDRV_CTL_EVENT_MASK_TLV, &kctl->id);
+			return 0;
+		}
+	} else {
+		if (op_flag) {
+			err = -ENXIO;
+			goto __kctl_end;
+		}
+		len = kctl->tlv.p[1] + 2 * sizeof(unsigned int);
+		if (tlv.length < len) {
+			err = -ENOMEM;
+			goto __kctl_end;
+		}
+		if (copy_to_user(_tlv->tlv, kctl->tlv.p, len))
+			err = -EFAULT;
+	}
       __kctl_end:
-        up_read(&card->controls_rwsem);
-        return err;
+	up_read(&card->controls_rwsem);
+	return err;
 }
 
 static long snd_ctl_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
@@ -1140,8 +1215,12 @@ static long snd_ctl_ioctl(struct file *file, unsigned int cmd, unsigned long arg
 		return snd_ctl_elem_remove(ctl, argp);
 	case SNDRV_CTL_IOCTL_SUBSCRIBE_EVENTS:
 		return snd_ctl_subscribe_events(ctl, ip);
-        case SNDRV_CTL_IOCTL_TLV_READ:
-                return snd_ctl_tlv_read(card, argp);
+	case SNDRV_CTL_IOCTL_TLV_READ:
+		return snd_ctl_tlv_ioctl(ctl, argp, 0);
+	case SNDRV_CTL_IOCTL_TLV_WRITE:
+		return snd_ctl_tlv_ioctl(ctl, argp, 1);
+	case SNDRV_CTL_IOCTL_TLV_COMMAND:
+		return snd_ctl_tlv_ioctl(ctl, argp, -1);
 	case SNDRV_CTL_IOCTL_POWER:
 		return -ENOPROTOOPT;
 	case SNDRV_CTL_IOCTL_POWER_STATE:
-- 
cgit v1.2.3


From 6a65d793b0a82c7e190d9fd92a479401b6a127ca Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Fri, 14 Jul 2006 14:39:34 +0200
Subject: [ALSA] Remove unused tlv_rw field from struct snd_kcontrol

Remove unused tlv_rw field from struct snd_kcontrol.  The callback is
set in tlv.c field, instead.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 include/sound/control.h | 1 -
 1 file changed, 1 deletion(-)

(limited to 'include')

diff --git a/include/sound/control.h b/include/sound/control.h
index e3905c5a0950..1de148b0fd94 100644
--- a/include/sound/control.h
+++ b/include/sound/control.h
@@ -67,7 +67,6 @@ struct snd_kcontrol {
 	snd_kcontrol_info_t *info;
 	snd_kcontrol_get_t *get;
 	snd_kcontrol_put_t *put;
-	snd_kcontrol_tlv_rw_t *tlv_rw;
 	union {
 		snd_kcontrol_tlv_rw_t *c;
 		unsigned int *p;
-- 
cgit v1.2.3


From 31508f83f591dc8764427b6321c89f8f9e84bad2 Mon Sep 17 00:00:00 2001
From: James Courtier-Dutton <James@superbug.co.uk>
Date: Sat, 22 Jul 2006 17:02:10 +0100
Subject: [ALSA] snd-emu10k1: Implement dB gain infomation.

Signed-off-by: James Courtier-Dutton <James@superbug.co.uk>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 include/sound/emu10k1.h   |  4 ++++
 sound/pci/emu10k1/emufx.c | 11 +++++++++++
 sound/pci/emu10k1/p16v.c  |  5 +++++
 3 files changed, 20 insertions(+)

(limited to 'include')

diff --git a/include/sound/emu10k1.h b/include/sound/emu10k1.h
index 884bbf54cd36..892e310c504d 100644
--- a/include/sound/emu10k1.h
+++ b/include/sound/emu10k1.h
@@ -1524,6 +1524,10 @@ struct snd_emu10k1_fx8010_control_gpr {
 	unsigned int value[32];		/* initial values */
 	unsigned int min;		/* minimum range */
 	unsigned int max;		/* maximum range */
+	union {
+		snd_kcontrol_tlv_rw_t *c;
+		unsigned int *p;
+	} tlv;
 	unsigned int translation;	/* translation type (EMU10K1_GPR_TRANSLATION*) */
 };
 
diff --git a/sound/pci/emu10k1/emufx.c b/sound/pci/emu10k1/emufx.c
index dfba00230d4d..00fc904c251d 100644
--- a/sound/pci/emu10k1/emufx.c
+++ b/sound/pci/emu10k1/emufx.c
@@ -35,6 +35,7 @@
 #include <linux/mutex.h>
 
 #include <sound/core.h>
+#include <sound/tlv.h>
 #include <sound/emu10k1.h>
 
 #if 0		/* for testing purposes - digital out -> capture */
@@ -290,6 +291,9 @@ static const u32 db_table[101] = {
 	0x7fffffff,
 };
 
+/* EMU10k1/EMU10k2 DSP control db gain */
+static DECLARE_TLV_DB_SCALE(snd_emu10k1_db_scale1, -4000, 40, 1);
+
 static const u32 onoff_table[2] = {
 	0x00000000, 0x00000001
 };
@@ -755,6 +759,11 @@ static int snd_emu10k1_add_controls(struct snd_emu10k1 *emu,
 		knew.device = gctl->id.device;
 		knew.subdevice = gctl->id.subdevice;
 		knew.info = snd_emu10k1_gpr_ctl_info;
+		if (gctl->tlv.p) {
+			knew.tlv.p = gctl->tlv.p;
+			knew.access = SNDRV_CTL_ELEM_ACCESS_READWRITE |
+				SNDRV_CTL_ELEM_ACCESS_TLV_READ;
+		} 
 		knew.get = snd_emu10k1_gpr_ctl_get;
 		knew.put = snd_emu10k1_gpr_ctl_put;
 		memset(nctl, 0, sizeof(*nctl));
@@ -1013,6 +1022,7 @@ snd_emu10k1_init_mono_control(struct snd_emu10k1_fx8010_control_gpr *ctl,
 	ctl->gpr[0] = gpr + 0; ctl->value[0] = defval;
 	ctl->min = 0;
 	ctl->max = 100;
+	ctl->tlv.p = snd_emu10k1_db_scale1;
 	ctl->translation = EMU10K1_GPR_TRANSLATION_TABLE100;	
 }
 
@@ -1027,6 +1037,7 @@ snd_emu10k1_init_stereo_control(struct snd_emu10k1_fx8010_control_gpr *ctl,
 	ctl->gpr[1] = gpr + 1; ctl->value[1] = defval;
 	ctl->min = 0;
 	ctl->max = 100;
+	ctl->tlv.p = snd_emu10k1_db_scale1;
 	ctl->translation = EMU10K1_GPR_TRANSLATION_TABLE100;
 }
 
diff --git a/sound/pci/emu10k1/p16v.c b/sound/pci/emu10k1/p16v.c
index 9905651935fb..1e44714b8623 100644
--- a/sound/pci/emu10k1/p16v.c
+++ b/sound/pci/emu10k1/p16v.c
@@ -100,6 +100,7 @@
 #include <sound/pcm.h>
 #include <sound/ac97_codec.h>
 #include <sound/info.h>
+#include <sound/tlv.h>
 #include <sound/emu10k1.h>
 #include "p16v.h"
 
@@ -784,12 +785,16 @@ static int snd_p16v_capture_channel_put(struct snd_kcontrol *kcontrol,
 	}
         return change;
 }
+static DECLARE_TLV_DB_SCALE(snd_p16v_db_scale1, -5175, 25, 1);
 
 #define P16V_VOL(xname,xreg,xhl) { \
 	.iface = SNDRV_CTL_ELEM_IFACE_MIXER, .name = xname, \
+        .access = SNDRV_CTL_ELEM_ACCESS_READWRITE |             \
+                  SNDRV_CTL_ELEM_ACCESS_TLV_READ,               \
 	.info = snd_p16v_volume_info, \
 	.get = snd_p16v_volume_get, \
 	.put = snd_p16v_volume_put, \
+	.tlv.p = snd_p16v_db_scale1, \
 	.private_value = ((xreg) | ((xhl) << 8)) \
 }
 
-- 
cgit v1.2.3


From 548a648b98318e4b843b636dd2c7f42377e19a00 Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Mon, 31 Jul 2006 16:51:51 +0200
Subject: [ALSA] Fix control/status mmap with shared PCM substream

The flag to avoid 32bit-incompatible mmap for control/status records
should be outside the pcm substream instance since a substream can be
shared among multiple opens.  Now it's flagged in pcm_file list that
is directly assigned to file->private_data.
Also, removed snd_pcm_add_file() and remove_file() functions and
substream.files field that are not really used in the code.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 include/sound/pcm.h     |  4 +---
 sound/core/pcm_compat.c |  2 +-
 sound/core/pcm_native.c | 49 ++++++++++---------------------------------------
 3 files changed, 12 insertions(+), 43 deletions(-)

(limited to 'include')

diff --git a/include/sound/pcm.h b/include/sound/pcm.h
index f84d84993a31..60d40b34efc0 100644
--- a/include/sound/pcm.h
+++ b/include/sound/pcm.h
@@ -190,7 +190,7 @@ struct snd_pcm_ops {
 
 struct snd_pcm_file {
 	struct snd_pcm_substream *substream;
-	struct snd_pcm_file *next;
+	int no_compat_mmap;
 };
 
 struct snd_pcm_hw_rule;
@@ -384,7 +384,6 @@ struct snd_pcm_substream {
 	struct snd_info_entry *proc_prealloc_entry;
 #endif
 	/* misc flags */
-	unsigned int no_mmap_ctrl: 1;
 	unsigned int hw_opened: 1;
 };
 
@@ -402,7 +401,6 @@ struct snd_pcm_str {
 	/* -- OSS things -- */
 	struct snd_pcm_oss_stream oss;
 #endif
-	struct snd_pcm_file *files;
 #ifdef CONFIG_SND_VERBOSE_PROCFS
 	struct snd_info_entry *proc_root;
 	struct snd_info_entry *proc_info_entry;
diff --git a/sound/core/pcm_compat.c b/sound/core/pcm_compat.c
index 2b8aab6fd6cd..2b539799d23b 100644
--- a/sound/core/pcm_compat.c
+++ b/sound/core/pcm_compat.c
@@ -478,7 +478,7 @@ static long snd_pcm_ioctl_compat(struct file *file, unsigned int cmd, unsigned l
 	 * mmap of PCM status/control records because of the size
 	 * incompatibility.
 	 */
-	substream->no_mmap_ctrl = 1;
+	pcm_file->no_compat_mmap = 1;
 
 	switch (cmd) {
 	case SNDRV_PCM_IOCTL_PVERSION:
diff --git a/sound/core/pcm_native.c b/sound/core/pcm_native.c
index 439f047929e1..0224c70414f5 100644
--- a/sound/core/pcm_native.c
+++ b/sound/core/pcm_native.c
@@ -1992,35 +1992,9 @@ int snd_pcm_hw_constraints_complete(struct snd_pcm_substream *substream)
 	return 0;
 }
 
-static void snd_pcm_add_file(struct snd_pcm_str *str,
-			     struct snd_pcm_file *pcm_file)
-{
-	pcm_file->next = str->files;
-	str->files = pcm_file;
-}
-
-static void snd_pcm_remove_file(struct snd_pcm_str *str,
-				struct snd_pcm_file *pcm_file)
-{
-	struct snd_pcm_file * pcm_file1;
-	if (str->files == pcm_file) {
-		str->files = pcm_file->next;
-	} else {
-		pcm_file1 = str->files;
-		while (pcm_file1 && pcm_file1->next != pcm_file)
-			pcm_file1 = pcm_file1->next;
-		if (pcm_file1 != NULL)
-			pcm_file1->next = pcm_file->next;
-	}
-}
-
 static void pcm_release_private(struct snd_pcm_substream *substream)
 {
-	struct snd_pcm_file *pcm_file = substream->file;
-
 	snd_pcm_unlink(substream);
-	snd_pcm_remove_file(substream->pstr, pcm_file);
-	kfree(pcm_file);
 }
 
 void snd_pcm_release_substream(struct snd_pcm_substream *substream)
@@ -2060,7 +2034,6 @@ int snd_pcm_open_substream(struct snd_pcm *pcm, int stream,
 		return 0;
 	}
 
-	substream->no_mmap_ctrl = 0;
 	err = snd_pcm_hw_constraints_init(substream);
 	if (err < 0) {
 		snd_printd("snd_pcm_hw_constraints_init failed\n");
@@ -2105,19 +2078,16 @@ static int snd_pcm_open_file(struct file *file,
 	if (err < 0)
 		return err;
 
-	if (substream->ref_count > 1)
-		pcm_file = substream->file;
-	else {
-		pcm_file = kzalloc(sizeof(*pcm_file), GFP_KERNEL);
-		if (pcm_file == NULL) {
-			snd_pcm_release_substream(substream);
-			return -ENOMEM;
-		}
+	pcm_file = kzalloc(sizeof(*pcm_file), GFP_KERNEL);
+	if (pcm_file == NULL) {
+		snd_pcm_release_substream(substream);
+		return -ENOMEM;
+	}
+	pcm_file->substream = substream;
+	if (substream->ref_count == 1) {
 		str = substream->pstr;
 		substream->file = pcm_file;
 		substream->pcm_release = pcm_release_private;
-		pcm_file->substream = substream;
-		snd_pcm_add_file(str, pcm_file);
 	}
 	file->private_data = pcm_file;
 	*rpcm_file = pcm_file;
@@ -2209,6 +2179,7 @@ static int snd_pcm_release(struct inode *inode, struct file *file)
 	fasync_helper(-1, file, 0, &substream->runtime->fasync);
 	mutex_lock(&pcm->open_mutex);
 	snd_pcm_release_substream(substream);
+	kfree(pcm_file);
 	mutex_unlock(&pcm->open_mutex);
 	wake_up(&pcm->open_wait);
 	module_put(pcm->card->module);
@@ -3270,11 +3241,11 @@ static int snd_pcm_mmap(struct file *file, struct vm_area_struct *area)
 	offset = area->vm_pgoff << PAGE_SHIFT;
 	switch (offset) {
 	case SNDRV_PCM_MMAP_OFFSET_STATUS:
-		if (substream->no_mmap_ctrl)
+		if (pcm_file->no_compat_mmap)
 			return -ENXIO;
 		return snd_pcm_mmap_status(substream, file, area);
 	case SNDRV_PCM_MMAP_OFFSET_CONTROL:
-		if (substream->no_mmap_ctrl)
+		if (pcm_file->no_compat_mmap)
 			return -ENXIO;
 		return snd_pcm_mmap_control(substream, file, area);
 	default:
-- 
cgit v1.2.3


From 683fe1537e660c322c8af953773921e814791193 Mon Sep 17 00:00:00 2001
From: Jochen Voss <voss@seehuhn.de>
Date: Tue, 8 Aug 2006 21:12:44 +0200
Subject: [ALSA] Revolution 5.1 - add AK5365 ADC support

Add support for the AK5365 ADC.

Signed-off-by: Jochen Voss <voss@seehuhn.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 include/sound/ak4xxx-adda.h   |  3 ++-
 sound/i2c/other/ak4xxx-adda.c | 25 +++++++++++++++++++++++++
 2 files changed, 27 insertions(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/sound/ak4xxx-adda.h b/include/sound/ak4xxx-adda.h
index 3d9888492026..65ddfa3cac1f 100644
--- a/include/sound/ak4xxx-adda.h
+++ b/include/sound/ak4xxx-adda.h
@@ -53,7 +53,8 @@ struct snd_akm4xxx {
 	unsigned int idx_offset;		/* control index offset */
 	enum {
 		SND_AK4524, SND_AK4528, SND_AK4529,
-		SND_AK4355, SND_AK4358, SND_AK4381
+		SND_AK4355, SND_AK4358, SND_AK4381,
+		SND_AK5365
 	} type;
 	unsigned int *num_stereo;	/* array of combined counts
 					 * for the mixer
diff --git a/sound/i2c/other/ak4xxx-adda.c b/sound/i2c/other/ak4xxx-adda.c
index dc7cc2001b74..7d562f084207 100644
--- a/sound/i2c/other/ak4xxx-adda.c
+++ b/sound/i2c/other/ak4xxx-adda.c
@@ -598,6 +598,31 @@ int snd_akm4xxx_build_controls(struct snd_akm4xxx *ak)
 		if (err < 0)
 			goto __error;
 	}
+
+	if (ak->type == SND_AK5365) {
+		memset(ctl, 0, sizeof(*ctl));
+		if (ak->channel_names == NULL)
+			strcpy(ctl->id.name, "Capture Volume");
+		else
+			strcpy(ctl->id.name, ak->channel_names[0]);
+		ctl->id.index = ak->idx_offset * 2;
+		ctl->id.iface = SNDRV_CTL_ELEM_IFACE_MIXER;
+		ctl->count = 1;
+		ctl->info = snd_akm4xxx_stereo_volume_info;
+		ctl->get = snd_akm4xxx_stereo_volume_get;
+		ctl->put = snd_akm4xxx_stereo_volume_put;
+		/* Registers 4 & 5 (see AK5365 data sheet, pages 34 and 35):
+		 * valid values are from 0x00 (mute) to 0x98 (+12dB).  */
+		ctl->private_value =
+			AK_COMPOSE(0, 4, 0, 0x98);
+		ctl->private_data = ak;
+		err = snd_ctl_add(ak->card,
+				  snd_ctl_new(ctl, SNDRV_CTL_ELEM_ACCESS_READ|
+					      SNDRV_CTL_ELEM_ACCESS_WRITE));
+		if (err < 0)
+			goto __error;
+	}
+
 	if (ak->type == SND_AK4355 || ak->type == SND_AK4358)
 		num_emphs = 1;
 	else
-- 
cgit v1.2.3


From eac06a10d2b814dfacc36a8fff35ef07bf4eec8e Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Tue, 22 Aug 2006 13:16:25 +0200
Subject: [ALSA] Add dB scale information to ad1848 driver

Added the dB scale information to ad1848 driver.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 include/sound/ad1848.h        | 22 +++++++++++++------
 sound/isa/ad1848/ad1848_lib.c | 49 ++++++++++++++++++++++++++++++-------------
 2 files changed, 50 insertions(+), 21 deletions(-)

(limited to 'include')

diff --git a/include/sound/ad1848.h b/include/sound/ad1848.h
index 57af1fe7b309..c8de6f83338f 100644
--- a/include/sound/ad1848.h
+++ b/include/sound/ad1848.h
@@ -179,14 +179,13 @@ enum { AD1848_MIX_SINGLE, AD1848_MIX_DOUBLE, AD1848_MIX_CAPTURE };
 #define AD1848_MIXVAL_DOUBLE(left_reg, right_reg, shift_left, shift_right, mask, invert) \
 	((left_reg) | ((right_reg) << 8) | ((shift_left) << 16) | ((shift_right) << 19) | ((mask) << 24) | ((invert) << 22))
 
-int snd_ad1848_add_ctl(struct snd_ad1848 *chip, const char *name, int index, int type, unsigned long value);
-
 /* for ease of use */
 struct ad1848_mix_elem {
 	const char *name;
 	int index;
 	int type;
 	unsigned long private_value;
+	unsigned int *tlv;
 };
 
 #define AD1848_SINGLE(xname, xindex, reg, shift, mask, invert) \
@@ -195,15 +194,26 @@ struct ad1848_mix_elem {
   .type = AD1848_MIX_SINGLE, \
   .private_value = AD1848_MIXVAL_SINGLE(reg, shift, mask, invert) }
 
+#define AD1848_SINGLE_TLV(xname, xindex, reg, shift, mask, invert, xtlv) \
+{ .name = xname, \
+  .index = xindex, \
+  .type = AD1848_MIX_SINGLE, \
+  .private_value = AD1848_MIXVAL_SINGLE(reg, shift, mask, invert), \
+  .tlv = xtlv }
+
 #define AD1848_DOUBLE(xname, xindex, left_reg, right_reg, shift_left, shift_right, mask, invert) \
 { .name = xname, \
   .index = xindex, \
   .type = AD1848_MIX_DOUBLE, \
   .private_value = AD1848_MIXVAL_DOUBLE(left_reg, right_reg, shift_left, shift_right, mask, invert) }
 
-static inline int snd_ad1848_add_ctl_elem(struct snd_ad1848 *chip, const struct ad1848_mix_elem *c)
-{
-	return snd_ad1848_add_ctl(chip, c->name, c->index, c->type, c->private_value);
-}
+#define AD1848_DOUBLE_TLV(xname, xindex, left_reg, right_reg, shift_left, shift_right, mask, invert, xtlv) \
+{ .name = xname, \
+  .index = xindex, \
+  .type = AD1848_MIX_DOUBLE, \
+  .private_value = AD1848_MIXVAL_DOUBLE(left_reg, right_reg, shift_left, shift_right, mask, invert), \
+  .tlv = xtlv }
+
+int snd_ad1848_add_ctl_elem(struct snd_ad1848 *chip, const struct ad1848_mix_elem *c);
 
 #endif /* __SOUND_AD1848_H */
diff --git a/sound/isa/ad1848/ad1848_lib.c b/sound/isa/ad1848/ad1848_lib.c
index e711f87d5fd1..a6fbd5d1d62f 100644
--- a/sound/isa/ad1848/ad1848_lib.c
+++ b/sound/isa/ad1848/ad1848_lib.c
@@ -29,6 +29,7 @@
 #include <sound/core.h>
 #include <sound/ad1848.h>
 #include <sound/control.h>
+#include <sound/tlv.h>
 #include <sound/pcm_params.h>
 
 #include <asm/io.h>
@@ -118,6 +119,8 @@ void snd_ad1848_out(struct snd_ad1848 *chip,
 #endif
 }
 
+EXPORT_SYMBOL(snd_ad1848_out);
+
 static void snd_ad1848_dout(struct snd_ad1848 *chip,
 			    unsigned char reg, unsigned char value)
 {
@@ -941,6 +944,8 @@ int snd_ad1848_create(struct snd_card *card,
 	return 0;
 }
 
+EXPORT_SYMBOL(snd_ad1848_create);
+
 static struct snd_pcm_ops snd_ad1848_playback_ops = {
 	.open =		snd_ad1848_playback_open,
 	.close =	snd_ad1848_playback_close,
@@ -988,12 +993,16 @@ int snd_ad1848_pcm(struct snd_ad1848 *chip, int device, struct snd_pcm **rpcm)
 	return 0;
 }
 
+EXPORT_SYMBOL(snd_ad1848_pcm);
+
 const struct snd_pcm_ops *snd_ad1848_get_pcm_ops(int direction)
 {
 	return direction == SNDRV_PCM_STREAM_PLAYBACK ?
 		&snd_ad1848_playback_ops : &snd_ad1848_capture_ops;
 }
 
+EXPORT_SYMBOL(snd_ad1848_get_pcm_ops);
+
 /*
  *  MIXER part
  */
@@ -1171,7 +1180,8 @@ static int snd_ad1848_put_double(struct snd_kcontrol *kcontrol, struct snd_ctl_e
 
 /*
  */
-int snd_ad1848_add_ctl(struct snd_ad1848 *chip, const char *name, int index, int type, unsigned long value)
+int snd_ad1848_add_ctl_elem(struct snd_ad1848 *chip,
+			    const struct ad1848_mix_elem *c)
 {
 	static struct snd_kcontrol_new newctls[] = {
 		[AD1848_MIX_SINGLE] = {
@@ -1196,32 +1206,46 @@ int snd_ad1848_add_ctl(struct snd_ad1848 *chip, const char *name, int index, int
 	struct snd_kcontrol *ctl;
 	int err;
 
-	ctl = snd_ctl_new1(&newctls[type], chip);
+	ctl = snd_ctl_new1(&newctls[c->type], chip);
 	if (! ctl)
 		return -ENOMEM;
-	strlcpy(ctl->id.name, name, sizeof(ctl->id.name));
-	ctl->id.index = index;
-	ctl->private_value = value;
+	strlcpy(ctl->id.name, c->name, sizeof(ctl->id.name));
+	ctl->id.index = c->index;
+	ctl->private_value = c->private_value;
+	if (c->tlv) {
+		ctl->vd[0].access |= SNDRV_CTL_ELEM_ACCESS_TLV_READ;
+		ctl->tlv.p = c->tlv;
+	}
 	if ((err = snd_ctl_add(chip->card, ctl)) < 0)
 		return err;
 	return 0;
 }
 
+EXPORT_SYMBOL(snd_ad1848_add_ctl_elem);
+
+static DECLARE_TLV_DB_SCALE(db_scale_6bit, -9450, 150, 0);
+static DECLARE_TLV_DB_SCALE(db_scale_5bit_12db_max, -3450, 150, 0);
+static DECLARE_TLV_DB_SCALE(db_scale_rec_gain, 0, 150, 0);
 
 static struct ad1848_mix_elem snd_ad1848_controls[] = {
 AD1848_DOUBLE("PCM Playback Switch", 0, AD1848_LEFT_OUTPUT, AD1848_RIGHT_OUTPUT, 7, 7, 1, 1),
-AD1848_DOUBLE("PCM Playback Volume", 0, AD1848_LEFT_OUTPUT, AD1848_RIGHT_OUTPUT, 0, 0, 63, 1),
+AD1848_DOUBLE_TLV("PCM Playback Volume", 0, AD1848_LEFT_OUTPUT, AD1848_RIGHT_OUTPUT, 0, 0, 63, 1,
+		  db_scale_6bit),
 AD1848_DOUBLE("Aux Playback Switch", 0, AD1848_AUX1_LEFT_INPUT, AD1848_AUX1_RIGHT_INPUT, 7, 7, 1, 1),
-AD1848_DOUBLE("Aux Playback Volume", 0, AD1848_AUX1_LEFT_INPUT, AD1848_AUX1_RIGHT_INPUT, 0, 0, 31, 1),
+AD1848_DOUBLE_TLV("Aux Playback Volume", 0, AD1848_AUX1_LEFT_INPUT, AD1848_AUX1_RIGHT_INPUT, 0, 0, 31, 1,
+		  db_scale_5bit_12db_max),
 AD1848_DOUBLE("Aux Playback Switch", 1, AD1848_AUX2_LEFT_INPUT, AD1848_AUX2_RIGHT_INPUT, 7, 7, 1, 1),
-AD1848_DOUBLE("Aux Playback Volume", 1, AD1848_AUX2_LEFT_INPUT, AD1848_AUX2_RIGHT_INPUT, 0, 0, 31, 1),
-AD1848_DOUBLE("Capture Volume", 0, AD1848_LEFT_INPUT, AD1848_RIGHT_INPUT, 0, 0, 15, 0),
+AD1848_DOUBLE_TLV("Aux Playback Volume", 1, AD1848_AUX2_LEFT_INPUT, AD1848_AUX2_RIGHT_INPUT, 0, 0, 31, 1,
+		  db_scale_5bit_12db_max),
+AD1848_DOUBLE_TLV("Capture Volume", 0, AD1848_LEFT_INPUT, AD1848_RIGHT_INPUT, 0, 0, 15, 0,
+		  db_scale_rec_gain),
 {
 	.name = "Capture Source",
 	.type = AD1848_MIX_CAPTURE,
 },
 AD1848_SINGLE("Loopback Capture Switch", 0, AD1848_LOOPBACK, 0, 1, 0),
-AD1848_SINGLE("Loopback Capture Volume", 0, AD1848_LOOPBACK, 1, 63, 0)
+AD1848_SINGLE_TLV("Loopback Capture Volume", 0, AD1848_LOOPBACK, 1, 63, 0,
+		  db_scale_6bit),
 };
                                         
 int snd_ad1848_mixer(struct snd_ad1848 *chip)
@@ -1245,12 +1269,7 @@ int snd_ad1848_mixer(struct snd_ad1848 *chip)
 	return 0;
 }
 
-EXPORT_SYMBOL(snd_ad1848_out);
-EXPORT_SYMBOL(snd_ad1848_create);
-EXPORT_SYMBOL(snd_ad1848_pcm);
-EXPORT_SYMBOL(snd_ad1848_get_pcm_ops);
 EXPORT_SYMBOL(snd_ad1848_mixer);
-EXPORT_SYMBOL(snd_ad1848_add_ctl);
 
 /*
  *  INIT part
-- 
cgit v1.2.3


From 1186ed8c7dc9c0185e783beddf241509cc224f1a Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Wed, 23 Aug 2006 19:53:28 +0200
Subject: [ALSA] Add dB scale information to vxpocket and vx222 drivers

Added the dB scale information to vxpocket and vx222 drivers.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 include/sound/vx_core.h     |  1 +
 sound/drivers/vx/vx_mixer.c | 17 +++++++++++++++--
 sound/pci/vx222/vx222.c     |  7 +++++++
 sound/pci/vx222/vx222_ops.c |  9 +++++++++
 sound/pcmcia/vx/vxpocket.c  |  5 +++++
 5 files changed, 37 insertions(+), 2 deletions(-)

(limited to 'include')

diff --git a/include/sound/vx_core.h b/include/sound/vx_core.h
index 9821a6194caa..dbca14170615 100644
--- a/include/sound/vx_core.h
+++ b/include/sound/vx_core.h
@@ -128,6 +128,7 @@ struct snd_vx_hardware {
 	unsigned int num_ins;
 	unsigned int num_outs;
 	unsigned int output_level_max;
+	unsigned int *output_level_db_scale;
 };
 
 /* hwdep id string */
diff --git a/sound/drivers/vx/vx_mixer.c b/sound/drivers/vx/vx_mixer.c
index c1d7fcdd1973..1613ed844ac6 100644
--- a/sound/drivers/vx/vx_mixer.c
+++ b/sound/drivers/vx/vx_mixer.c
@@ -23,6 +23,7 @@
 #include <sound/driver.h>
 #include <sound/core.h>
 #include <sound/control.h>
+#include <sound/tlv.h>
 #include <sound/vx_core.h>
 #include "vx_cmd.h"
 
@@ -455,10 +456,13 @@ static int vx_output_level_put(struct snd_kcontrol *kcontrol, struct snd_ctl_ele
 
 static struct snd_kcontrol_new vx_control_output_level = {
 	.iface =	SNDRV_CTL_ELEM_IFACE_MIXER,
+	.access =	(SNDRV_CTL_ELEM_ACCESS_READWRITE |
+			 SNDRV_CTL_ELEM_ACCESS_TLV_READ),
 	.name =		"Master Playback Volume",
 	.info =		vx_output_level_info,
 	.get =		vx_output_level_get,
 	.put =		vx_output_level_put,
+	/* tlv will be filled later */
 };
 
 /*
@@ -712,12 +716,17 @@ static int vx_monitor_sw_put(struct snd_kcontrol *kcontrol, struct snd_ctl_elem_
 	return 0;
 }
 
+static DECLARE_TLV_DB_SCALE(db_scale_audio_gain, -10975, 25, 0);
+
 static struct snd_kcontrol_new vx_control_audio_gain = {
 	.iface =	SNDRV_CTL_ELEM_IFACE_MIXER,
+	.access =	(SNDRV_CTL_ELEM_ACCESS_READWRITE |
+			 SNDRV_CTL_ELEM_ACCESS_TLV_READ),
 	/* name will be filled later */
 	.info =         vx_audio_gain_info,
 	.get =          vx_audio_gain_get,
-	.put =          vx_audio_gain_put
+	.put =          vx_audio_gain_put,
+	.tlv = { .p = db_scale_audio_gain },
 };
 static struct snd_kcontrol_new vx_control_output_switch = {
 	.iface =	SNDRV_CTL_ELEM_IFACE_MIXER,
@@ -729,9 +738,12 @@ static struct snd_kcontrol_new vx_control_output_switch = {
 static struct snd_kcontrol_new vx_control_monitor_gain = {
 	.iface =	SNDRV_CTL_ELEM_IFACE_MIXER,
 	.name =         "Monitoring Volume",
+	.access =	(SNDRV_CTL_ELEM_ACCESS_READWRITE |
+			 SNDRV_CTL_ELEM_ACCESS_TLV_READ),
 	.info =         vx_audio_gain_info,	/* shared */
 	.get =          vx_audio_monitor_get,
-	.put =          vx_audio_monitor_put
+	.put =          vx_audio_monitor_put,
+	.tlv = { .p = db_scale_audio_gain },
 };
 static struct snd_kcontrol_new vx_control_monitor_switch = {
 	.iface =	SNDRV_CTL_ELEM_IFACE_MIXER,
@@ -918,6 +930,7 @@ int snd_vx_mixer_new(struct vx_core *chip)
 	for (i = 0; i < chip->hw->num_outs; i++) {
 		temp = vx_control_output_level;
 		temp.index = i;
+		temp.tlv.p = chip->hw->output_level_db_scale;
 		if ((err = snd_ctl_add(card, snd_ctl_new1(&temp, chip))) < 0)
 			return err;
 	}
diff --git a/sound/pci/vx222/vx222.c b/sound/pci/vx222/vx222.c
index 9c03c6b4e490..e7cd8acab59a 100644
--- a/sound/pci/vx222/vx222.c
+++ b/sound/pci/vx222/vx222.c
@@ -26,6 +26,7 @@
 #include <linux/moduleparam.h>
 #include <sound/core.h>
 #include <sound/initval.h>
+#include <sound/tlv.h>
 #include "vx222.h"
 
 #define CARD_NAME "VX222"
@@ -72,6 +73,9 @@ MODULE_DEVICE_TABLE(pci, snd_vx222_ids);
 /*
  */
 
+static DECLARE_TLV_DB_SCALE(db_scale_old_vol, -11350, 50, 0);
+static DECLARE_TLV_DB_SCALE(db_scale_akm, -7350, 50, 0);
+
 static struct snd_vx_hardware vx222_old_hw = {
 
 	.name = "VX222/Old",
@@ -81,6 +85,7 @@ static struct snd_vx_hardware vx222_old_hw = {
 	.num_ins = 1,
 	.num_outs = 1,
 	.output_level_max = VX_ANALOG_OUT_LEVEL_MAX,
+	.output_level_db_scale = db_scale_old_vol,
 };
 
 static struct snd_vx_hardware vx222_v2_hw = {
@@ -92,6 +97,7 @@ static struct snd_vx_hardware vx222_v2_hw = {
 	.num_ins = 1,
 	.num_outs = 1,
 	.output_level_max = VX2_AKM_LEVEL_MAX,
+	.output_level_db_scale = db_scale_akm,
 };
 
 static struct snd_vx_hardware vx222_mic_hw = {
@@ -103,6 +109,7 @@ static struct snd_vx_hardware vx222_mic_hw = {
 	.num_ins = 1,
 	.num_outs = 1,
 	.output_level_max = VX2_AKM_LEVEL_MAX,
+	.output_level_db_scale = db_scale_akm,
 };
 
 
diff --git a/sound/pci/vx222/vx222_ops.c b/sound/pci/vx222/vx222_ops.c
index 9b6d345b83a6..5e51950e05f9 100644
--- a/sound/pci/vx222/vx222_ops.c
+++ b/sound/pci/vx222/vx222_ops.c
@@ -28,6 +28,7 @@
 
 #include <sound/core.h>
 #include <sound/control.h>
+#include <sound/tlv.h>
 #include <asm/io.h>
 #include "vx222.h"
 
@@ -845,6 +846,8 @@ static void vx2_set_input_level(struct snd_vx222 *chip)
 
 #define MIC_LEVEL_MAX	0xff
 
+static DECLARE_TLV_DB_SCALE(db_scale_mic, -6450, 50, 0);
+
 /*
  * controls API for input levels
  */
@@ -922,18 +925,24 @@ static int vx_mic_level_put(struct snd_kcontrol *kcontrol, struct snd_ctl_elem_v
 
 static struct snd_kcontrol_new vx_control_input_level = {
 	.iface =	SNDRV_CTL_ELEM_IFACE_MIXER,
+	.access =	(SNDRV_CTL_ELEM_ACCESS_READWRITE |
+			 SNDRV_CTL_ELEM_ACCESS_TLV_READ),
 	.name =		"Capture Volume",
 	.info =		vx_input_level_info,
 	.get =		vx_input_level_get,
 	.put =		vx_input_level_put,
+	.tlv = { .p = db_scale_mic },
 };
 
 static struct snd_kcontrol_new vx_control_mic_level = {
 	.iface =	SNDRV_CTL_ELEM_IFACE_MIXER,
+	.access =	(SNDRV_CTL_ELEM_ACCESS_READWRITE |
+			 SNDRV_CTL_ELEM_ACCESS_TLV_READ),
 	.name =		"Mic Capture Volume",
 	.info =		vx_mic_level_info,
 	.get =		vx_mic_level_get,
 	.put =		vx_mic_level_put,
+	.tlv = { .p = db_scale_mic },
 };
 
 /*
diff --git a/sound/pcmcia/vx/vxpocket.c b/sound/pcmcia/vx/vxpocket.c
index 76c85cffb40e..3089fcca800e 100644
--- a/sound/pcmcia/vx/vxpocket.c
+++ b/sound/pcmcia/vx/vxpocket.c
@@ -27,6 +27,7 @@
 #include <pcmcia/ciscode.h>
 #include <pcmcia/cisreg.h>
 #include <sound/initval.h>
+#include <sound/tlv.h>
 
 /*
  */
@@ -90,6 +91,8 @@ static int snd_vxpocket_dev_free(struct snd_device *device)
  * Only output levels can be modified
  */
 
+static DECLARE_TLV_DB_SCALE(db_scale_old_vol, -11350, 50, 0);
+
 static struct snd_vx_hardware vxpocket_hw = {
 	.name = "VXPocket",
 	.type = VX_TYPE_VXPOCKET,
@@ -99,6 +102,7 @@ static struct snd_vx_hardware vxpocket_hw = {
 	.num_ins = 1,
 	.num_outs = 1,
 	.output_level_max = VX_ANALOG_OUT_LEVEL_MAX,
+	.output_level_db_scale = db_scale_old_vol,
 };	
 
 /* VX-pocket 440
@@ -120,6 +124,7 @@ static struct snd_vx_hardware vxp440_hw = {
 	.num_ins = 2,
 	.num_outs = 2,
 	.output_level_max = VX_ANALOG_OUT_LEVEL_MAX,
+	.output_level_db_scale = db_scale_old_vol,
 };	
 
 
-- 
cgit v1.2.3


From 063a40d9111ce7558f2fdfa4f85acfc47eb27353 Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Mon, 28 Aug 2006 13:20:13 +0200
Subject: [ALSA] Add the definition of linear volume TLV

Added the definition of linear volume TLV type.
Some DSP chips and codecs (e.g. AK codec) use linear volume control.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 include/sound/tlv.h | 10 ++++++++++
 1 file changed, 10 insertions(+)

(limited to 'include')

diff --git a/include/sound/tlv.h b/include/sound/tlv.h
index b826e1df1da6..7905841643df 100644
--- a/include/sound/tlv.h
+++ b/include/sound/tlv.h
@@ -33,6 +33,7 @@
 
 #define SNDRV_CTL_TLVT_CONTAINER 0	/* one level down - group of TLVs */
 #define SNDRV_CTL_TLVT_DB_SCALE	1       /* dB scale */
+#define SNDRV_CTL_TLVT_DB_LINEAR 2	/* linear volume */
 
 #define DECLARE_TLV_DB_SCALE(name, min, step, mute) \
 unsigned int name[] = { \
@@ -40,4 +41,13 @@ unsigned int name[] = { \
         (min), ((step) & 0xffff) | ((mute) ? 0x10000 : 0) \
 }
 
+/* linear volume between min_dB and max_dB (.01dB unit) */
+#define DECLARE_TLV_DB_LINEAR(name, min_dB, max_dB)	\
+unsigned int name[] = { \
+        SNDRV_CTL_TLVT_DB_LINEAR, 2 * sizeof(unsigned int), \
+        (min_dB), (max_dB)				\
+}
+
+#define TLV_DB_GAIN_MUTE	-9999999
+
 #endif /* __SOUND_TLV_H */
-- 
cgit v1.2.3


From 723b2b0d36fa7cea81a962af2d40d88520d5a5f1 Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Wed, 30 Aug 2006 16:49:54 +0200
Subject: [ALSA] Clean up and add TLV support to AK4xxx i2c driver

- Clean up the code in AK4xxx-ADDA i2c code.
- Fix capture gain controls for AK5365
- Changed the static table for DAC/ADC mixer labels to use
  structs
- Implemented TLV entries for each AK codec
  The volumes in AK4524, AK4528 and AK5365 are corrected with
  a table to be suitable for dB conversion.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 include/sound/ak4xxx-adda.h   |  40 +++-
 sound/i2c/other/ak4xxx-adda.c | 523 ++++++++++++++++++++++++------------------
 sound/pci/ice1712/revo.c      |  47 ++--
 3 files changed, 354 insertions(+), 256 deletions(-)

(limited to 'include')

diff --git a/include/sound/ak4xxx-adda.h b/include/sound/ak4xxx-adda.h
index 65ddfa3cac1f..026e4072a9a1 100644
--- a/include/sound/ak4xxx-adda.h
+++ b/include/sound/ak4xxx-adda.h
@@ -39,14 +39,26 @@ struct snd_ak4xxx_ops {
 
 #define AK4XXX_IMAGE_SIZE	(AK4XXX_MAX_CHIPS * 16)	/* 64 bytes */
 
+/* DAC label and channels */
+struct snd_akm4xxx_dac_channel {
+	char *name;		/* mixer volume name */
+	unsigned int num_channels;
+};
+
+/* ADC labels and channels */
+struct snd_akm4xxx_adc_channel {
+	char *name;		/* capture gain volume label */
+	char *gain_name;	/* IPGA */
+	char *switch_name;	/* capture switch */
+	unsigned int num_channels;
+};
+
 struct snd_akm4xxx {
 	struct snd_card *card;
 	unsigned int num_adcs;			/* AK4524 or AK4528 ADCs */
 	unsigned int num_dacs;			/* AK4524 or AK4528 DACs */
 	unsigned char images[AK4XXX_IMAGE_SIZE]; /* saved register image */
-	unsigned char ipga_gain[AK4XXX_MAX_CHIPS][2]; /* saved register image
-						       * for IPGA (AK4528)
-						       */
+	unsigned char volumes[AK4XXX_IMAGE_SIZE]; /* saved volume values */
 	unsigned long private_value[AK4XXX_MAX_CHIPS];	/* helper for driver */
 	void *private_data[AK4XXX_MAX_CHIPS];		/* helper for driver */
 	/* template should fill the following fields */
@@ -56,10 +68,11 @@ struct snd_akm4xxx {
 		SND_AK4355, SND_AK4358, SND_AK4381,
 		SND_AK5365
 	} type;
-	unsigned int *num_stereo;	/* array of combined counts
-					 * for the mixer
-					 */
-	char **channel_names;		/* array of mixer channel names */
+
+	/* (array) information of combined codecs */
+	struct snd_akm4xxx_dac_channel *dac_info;
+	struct snd_akm4xxx_adc_channel *adc_info;
+
 	struct snd_ak4xxx_ops ops;
 };
 
@@ -73,9 +86,18 @@ int snd_akm4xxx_build_controls(struct snd_akm4xxx *ak);
 	(ak)->images[(chip) * 16 + (reg)]
 #define snd_akm4xxx_set(ak,chip,reg,val) \
 	((ak)->images[(chip) * 16 + (reg)] = (val))
+#define snd_akm4xxx_get_vol(ak,chip,reg) \
+	(ak)->volumes[(chip) * 16 + (reg)]
+#define snd_akm4xxx_set_vol(ak,chip,reg,val) \
+	((ak)->volumes[(chip) * 16 + (reg)] = (val))
+
+/* Warning: IPGA is tricky - we assume the addr + 4 is unused
+ *   so far, it's OK for all AK codecs with IPGA:
+ *   AK4524, AK4528 and EK5365
+ */
 #define snd_akm4xxx_get_ipga(ak,chip,reg) \
-	(ak)->ipga_gain[chip][(reg)-4]
+	snd_akm4xxx_get_vol(ak, chip, (reg) + 4)
 #define snd_akm4xxx_set_ipga(ak,chip,reg,val) \
-	((ak)->ipga_gain[chip][(reg)-4] = (val))
+	snd_akm4xxx_set_vol(ak, chip, (reg) + 4, val)
 
 #endif /* __SOUND_AK4XXX_ADDA_H */
diff --git a/sound/i2c/other/ak4xxx-adda.c b/sound/i2c/other/ak4xxx-adda.c
index 89fc3cbc2356..c34cb4684607 100644
--- a/sound/i2c/other/ak4xxx-adda.c
+++ b/sound/i2c/other/ak4xxx-adda.c
@@ -28,12 +28,14 @@
 #include <linux/init.h>
 #include <sound/core.h>
 #include <sound/control.h>
+#include <sound/tlv.h>
 #include <sound/ak4xxx-adda.h>
 
 MODULE_AUTHOR("Jaroslav Kysela <perex@suse.cz>, Takashi Iwai <tiwai@suse.de>");
 MODULE_DESCRIPTION("Routines for control of AK452x / AK43xx  AD/DA converters");
 MODULE_LICENSE("GPL");
 
+/* write the given register and save the data to the cache */
 void snd_akm4xxx_write(struct snd_akm4xxx *ak, int chip, unsigned char reg,
 		       unsigned char val)
 {
@@ -41,15 +43,10 @@ void snd_akm4xxx_write(struct snd_akm4xxx *ak, int chip, unsigned char reg,
 	ak->ops.write(ak, chip, reg, val);
 
 	/* save the data */
-	if (ak->type == SND_AK4524 || ak->type == SND_AK4528) {
-		if ((reg != 0x04 && reg != 0x05) || (val & 0x80) == 0)
-			snd_akm4xxx_set(ak, chip, reg, val);
-		else
-			snd_akm4xxx_set_ipga(ak, chip, reg, val);
-	} else {
-		/* AK4529, or else */
+	/* don't overwrite with IPGA data */
+	if ((ak->type != SND_AK4524 && ak->type != SND_AK5365) ||
+	    (reg != 0x04 && reg != 0x05) || (val & 0x80) == 0)
 		snd_akm4xxx_set(ak, chip, reg, val);
-	}
 	ak->ops.unlock(ak, chip);
 }
 
@@ -78,7 +75,7 @@ static void ak4524_reset(struct snd_akm4xxx *ak, int state)
 		/* IPGA */
 		for (reg = 0x04; reg < 0x06; reg++)
 			snd_akm4xxx_write(ak, chip, reg,
-					  snd_akm4xxx_get_ipga(ak, chip, reg));
+					  snd_akm4xxx_get_ipga(ak, chip, reg) | 0x80);
 	}
 }
 
@@ -144,6 +141,42 @@ void snd_akm4xxx_reset(struct snd_akm4xxx *ak, int state)
 
 EXPORT_SYMBOL(snd_akm4xxx_reset);
 
+
+/*
+ * Volume conversion table for non-linear volumes
+ * from -63.5dB (mute) to 0dB step 0.5dB
+ *
+ * Used for AK4524 input/ouput attenuation, AK4528, and
+ * AK5365 input attenuation
+ */
+static unsigned char vol_cvt_datt[128] = {
+	0x00, 0x01, 0x01, 0x02, 0x02, 0x03, 0x03, 0x04,
+	0x04, 0x04, 0x04, 0x05, 0x05, 0x05, 0x06, 0x06,
+	0x06, 0x07, 0x07, 0x08, 0x08, 0x08, 0x09, 0x0a,
+	0x0a, 0x0b, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f, 0x0f,
+	0x10, 0x10, 0x11, 0x12, 0x12, 0x13, 0x13, 0x14,
+	0x15, 0x16, 0x17, 0x17, 0x18, 0x19, 0x1a, 0x1c,
+	0x1d, 0x1e, 0x1f, 0x20, 0x21, 0x22, 0x23, 0x23,
+	0x24, 0x25, 0x26, 0x28, 0x29, 0x2a, 0x2b, 0x2d,
+	0x2e, 0x30, 0x30, 0x31, 0x32, 0x33, 0x34, 0x35,
+	0x37, 0x38, 0x39, 0x3b, 0x3c, 0x3e, 0x3f, 0x40,
+	0x41, 0x42, 0x43, 0x44, 0x46, 0x47, 0x48, 0x4a,
+	0x4b, 0x4d, 0x4e, 0x50, 0x51, 0x52, 0x53, 0x54,
+	0x55, 0x56, 0x58, 0x59, 0x5b, 0x5c, 0x5e, 0x5f,
+	0x60, 0x61, 0x62, 0x64, 0x65, 0x66, 0x67, 0x69,
+	0x6a, 0x6c, 0x6d, 0x6f, 0x70, 0x71, 0x72, 0x73,
+	0x75, 0x76, 0x77, 0x79, 0x7a, 0x7c, 0x7d, 0x7f,
+};
+
+/*
+ * dB tables
+ */
+static DECLARE_TLV_DB_SCALE(db_scale_vol_datt, -6350, 50, 1);
+static DECLARE_TLV_DB_SCALE(db_scale_8bit, -12750, 50, 1);
+static DECLARE_TLV_DB_SCALE(db_scale_7bit, -6350, 50, 1);
+static DECLARE_TLV_DB_LINEAR(db_scale_linear, TLV_DB_GAIN_MUTE, 0);
+static DECLARE_TLV_DB_SCALE(db_scale_ipga, 0, 50, 0);
+
 /*
  * initialize all the ak4xxx chips
  */
@@ -240,6 +273,9 @@ void snd_akm4xxx_init(struct snd_akm4xxx *ak)
 	int chip, num_chips;
 	unsigned char *ptr, reg, data, *inits;
 
+	memset(ak->images, 0, sizeof(ak->images));
+	memset(ak->volumes, 0, sizeof(ak->volumes));
+
 	switch (ak->type) {
 	case SND_AK4524:
 		inits = inits_ak4524;
@@ -265,6 +301,9 @@ void snd_akm4xxx_init(struct snd_akm4xxx *ak)
 		inits = inits_ak4381;
 		num_chips = ak->num_dacs / 2;
 		break;
+	case SND_AK5365:
+		/* FIXME: any init sequence? */
+		return;
 	default:
 		snd_BUG();
 		return;
@@ -282,16 +321,21 @@ void snd_akm4xxx_init(struct snd_akm4xxx *ak)
 
 EXPORT_SYMBOL(snd_akm4xxx_init);
 
+/*
+ * Mixer callbacks
+ */
+#define AK_VOL_CVT 			(1<<21)	/* need dB conversion */
+#define AK_NEEDSMSB 			(1<<22)	/* need MSB update bit */
+#define AK_INVERT 			(1<<23)	/* data is inverted */
 #define AK_GET_CHIP(val)		(((val) >> 8) & 0xff)
 #define AK_GET_ADDR(val)		((val) & 0xff)
-#define AK_GET_SHIFT(val)		(((val) >> 16) & 0x3f)
+#define AK_GET_SHIFT(val)		(((val) >> 16) & 0x1f)
+#define AK_GET_VOL_CVT(val)		(((val) >> 21) & 1)
 #define AK_GET_NEEDSMSB(val)		(((val) >> 22) & 1)
 #define AK_GET_INVERT(val)		(((val) >> 23) & 1)
 #define AK_GET_MASK(val)		(((val) >> 24) & 0xff)
 #define AK_COMPOSE(chip,addr,shift,mask) \
 	(((chip) << 8) | (addr) | ((shift) << 16) | ((mask) << 24))
-#define AK_NEEDSMSB 			(1<<22)
-#define AK_INVERT 			(1<<23)
 
 static int snd_akm4xxx_volume_info(struct snd_kcontrol *kcontrol,
 				   struct snd_ctl_elem_info *uinfo)
@@ -311,37 +355,37 @@ static int snd_akm4xxx_volume_get(struct snd_kcontrol *kcontrol,
 	struct snd_akm4xxx *ak = snd_kcontrol_chip(kcontrol);
 	int chip = AK_GET_CHIP(kcontrol->private_value);
 	int addr = AK_GET_ADDR(kcontrol->private_value);
-	int needsmsb = AK_GET_NEEDSMSB(kcontrol->private_value);
-	int invert = AK_GET_INVERT(kcontrol->private_value);
-	unsigned int mask = AK_GET_MASK(kcontrol->private_value);
-	unsigned char val = snd_akm4xxx_get(ak, chip, addr);
 
-	if (needsmsb)
-		val &= 0x7f;
-	ucontrol->value.integer.value[0] = invert ? mask - val : val;
+	ucontrol->value.integer.value[0] = snd_akm4xxx_get_vol(ak, chip, addr);
 	return 0;
 }
 
-static int snd_akm4xxx_volume_put(struct snd_kcontrol *kcontrol,
-				  struct snd_ctl_elem_value *ucontrol)
+static int put_ak_reg(struct snd_kcontrol *kcontrol, int addr,
+		      unsigned char nval)
 {
 	struct snd_akm4xxx *ak = snd_kcontrol_chip(kcontrol);
-	int chip = AK_GET_CHIP(kcontrol->private_value);
-	int addr = AK_GET_ADDR(kcontrol->private_value);
-	int needsmsb = AK_GET_NEEDSMSB(kcontrol->private_value);
-	int invert = AK_GET_INVERT(kcontrol->private_value);
 	unsigned int mask = AK_GET_MASK(kcontrol->private_value);
-	unsigned char nval = ucontrol->value.integer.value[0] % (mask+1);
-	int change;
+	int chip = AK_GET_CHIP(kcontrol->private_value);
 
-	if (invert)
+	if (snd_akm4xxx_get_vol(ak, chip, addr) == nval)
+		return 0;
+
+	snd_akm4xxx_set_vol(ak, chip, addr, nval);
+	if (AK_GET_VOL_CVT(kcontrol->private_value))
+		nval = vol_cvt_datt[nval];
+	if (AK_GET_INVERT(kcontrol->private_value))
 		nval = mask - nval;
-	if (needsmsb)
+	if (AK_GET_NEEDSMSB(kcontrol->private_value))
 		nval |= 0x80;
-	change = snd_akm4xxx_get(ak, chip, addr) != nval;
-	if (change)
-		snd_akm4xxx_write(ak, chip, addr, nval);
-	return change;
+	snd_akm4xxx_write(ak, chip, addr, nval);
+	return 1;
+}
+
+static int snd_akm4xxx_volume_put(struct snd_kcontrol *kcontrol,
+				  struct snd_ctl_elem_value *ucontrol)
+{
+	return put_ak_reg(kcontrol, AK_GET_ADDR(kcontrol->private_value),
+			  ucontrol->value.integer.value[0]);
 }
 
 static int snd_akm4xxx_stereo_volume_info(struct snd_kcontrol *kcontrol,
@@ -362,66 +406,25 @@ static int snd_akm4xxx_stereo_volume_get(struct snd_kcontrol *kcontrol,
 	struct snd_akm4xxx *ak = snd_kcontrol_chip(kcontrol);
 	int chip = AK_GET_CHIP(kcontrol->private_value);
 	int addr = AK_GET_ADDR(kcontrol->private_value);
-	int needsmsb = AK_GET_NEEDSMSB(kcontrol->private_value);
-	int invert = AK_GET_INVERT(kcontrol->private_value);
-	unsigned int mask = AK_GET_MASK(kcontrol->private_value);
-	unsigned char val;
-
-	val = snd_akm4xxx_get(ak, chip, addr);
-	if (needsmsb)
-		val &= 0x7f;
-	ucontrol->value.integer.value[0] = invert ? mask - val : val;
-
-	val = snd_akm4xxx_get(ak, chip, addr+1);
-	if (needsmsb)
-		val &= 0x7f;
-	ucontrol->value.integer.value[1] = invert ? mask - val : val;
 
+	ucontrol->value.integer.value[0] = snd_akm4xxx_get_vol(ak, chip, addr);
+	ucontrol->value.integer.value[1] = snd_akm4xxx_get_vol(ak, chip, addr+1);
 	return 0;
 }
 
 static int snd_akm4xxx_stereo_volume_put(struct snd_kcontrol *kcontrol,
 					 struct snd_ctl_elem_value *ucontrol)
 {
-	struct snd_akm4xxx *ak = snd_kcontrol_chip(kcontrol);
-	int chip = AK_GET_CHIP(kcontrol->private_value);
 	int addr = AK_GET_ADDR(kcontrol->private_value);
-	int needsmsb = AK_GET_NEEDSMSB(kcontrol->private_value);
-	int invert = AK_GET_INVERT(kcontrol->private_value);
-	unsigned int mask = AK_GET_MASK(kcontrol->private_value);
-	unsigned char nval = ucontrol->value.integer.value[0] % (mask+1);
-	int change0, change1;
-
-	if (invert)
-		nval = mask - nval;
-	if (needsmsb)
-		nval |= 0x80;
-	change0 = snd_akm4xxx_get(ak, chip, addr) != nval;
-	if (change0)
-		snd_akm4xxx_write(ak, chip, addr, nval);
-
-	nval = ucontrol->value.integer.value[1] % (mask+1);
-	if (invert)
-		nval = mask - nval;
-	if (needsmsb)
-		nval |= 0x80;
-	change1 = snd_akm4xxx_get(ak, chip, addr+1) != nval;
-	if (change1)
-		snd_akm4xxx_write(ak, chip, addr+1, nval);
-
+	int change;
 
-	return change0 || change1;
+	change = put_ak_reg(kcontrol, addr, ucontrol->value.integer.value[0]);
+	change |= put_ak_reg(kcontrol, addr + 1,
+			     ucontrol->value.integer.value[1]);
+	return change;
 }
 
-static int snd_akm4xxx_ipga_gain_info(struct snd_kcontrol *kcontrol,
-				      struct snd_ctl_elem_info *uinfo)
-{
-	uinfo->type = SNDRV_CTL_ELEM_TYPE_INTEGER;
-	uinfo->count = 1;
-	uinfo->value.integer.min = 0;
-	uinfo->value.integer.max = 36;
-	return 0;
-}
+#define snd_akm4xxx_ipga_gain_info	snd_akm4xxx_volume_info
 
 static int snd_akm4xxx_ipga_gain_get(struct snd_kcontrol *kcontrol,
 				     struct snd_ctl_elem_value *ucontrol)
@@ -429,21 +432,57 @@ static int snd_akm4xxx_ipga_gain_get(struct snd_kcontrol *kcontrol,
 	struct snd_akm4xxx *ak = snd_kcontrol_chip(kcontrol);
 	int chip = AK_GET_CHIP(kcontrol->private_value);
 	int addr = AK_GET_ADDR(kcontrol->private_value);
+
 	ucontrol->value.integer.value[0] =
-		snd_akm4xxx_get_ipga(ak, chip, addr) & 0x7f;
+		snd_akm4xxx_get_ipga(ak, chip, addr);
 	return 0;
 }
 
+static int put_ak_ipga(struct snd_kcontrol *kcontrol, int addr,
+		       unsigned char nval)
+{
+	struct snd_akm4xxx *ak = snd_kcontrol_chip(kcontrol);
+	int chip = AK_GET_CHIP(kcontrol->private_value);
+
+	if (snd_akm4xxx_get_ipga(ak, chip, addr) == nval)
+		return 0;
+	snd_akm4xxx_set_ipga(ak, chip, addr, nval);
+	snd_akm4xxx_write(ak, chip, addr, nval | 0x80); /* need MSB */
+	return 1;
+}
+
 static int snd_akm4xxx_ipga_gain_put(struct snd_kcontrol *kcontrol,
 				     struct snd_ctl_elem_value *ucontrol)
+{
+	return put_ak_ipga(kcontrol, AK_GET_ADDR(kcontrol->private_value),
+			   ucontrol->value.integer.value[0]);
+}
+
+#define snd_akm4xxx_stereo_gain_info	snd_akm4xxx_stereo_volume_info
+
+static int snd_akm4xxx_stereo_gain_get(struct snd_kcontrol *kcontrol,
+				       struct snd_ctl_elem_value *ucontrol)
 {
 	struct snd_akm4xxx *ak = snd_kcontrol_chip(kcontrol);
 	int chip = AK_GET_CHIP(kcontrol->private_value);
 	int addr = AK_GET_ADDR(kcontrol->private_value);
-	unsigned char nval = (ucontrol->value.integer.value[0] % 37) | 0x80;
-	int change = snd_akm4xxx_get_ipga(ak, chip, addr) != nval;
-	if (change)
-		snd_akm4xxx_write(ak, chip, addr, nval);
+
+	ucontrol->value.integer.value[0] =
+		snd_akm4xxx_get_ipga(ak, chip, addr);
+	ucontrol->value.integer.value[1] =
+		snd_akm4xxx_get_ipga(ak, chip, addr + 1);
+	return 0;
+}
+
+static int snd_akm4xxx_stereo_gain_put(struct snd_kcontrol *kcontrol,
+				       struct snd_ctl_elem_value *ucontrol)
+{
+	int addr = AK_GET_ADDR(kcontrol->private_value);
+	int change;
+
+	change = put_ak_ipga(kcontrol, addr, ucontrol->value.integer.value[0]);
+	change |= put_ak_ipga(kcontrol, addr + 1,
+			      ucontrol->value.integer.value[1]);
 	return change;
 }
 
@@ -548,221 +587,247 @@ static int ak4xxx_switch_put(struct snd_kcontrol *kcontrol,
  * build AK4xxx controls
  */
 
-int snd_akm4xxx_build_controls(struct snd_akm4xxx *ak)
+static int build_dac_controls(struct snd_akm4xxx *ak)
 {
-	unsigned int idx, num_emphs;
-	struct snd_kcontrol *ctl;
-	int err;
-	int mixer_ch = 0;
-	int num_stereo;
-
-	ctl = kmalloc(sizeof(*ctl), GFP_KERNEL);
-	if (! ctl)
-		return -ENOMEM;
+	int idx, err, mixer_ch, num_stereo;
+	struct snd_kcontrol_new knew;
 
+	mixer_ch = 0;
 	for (idx = 0; idx < ak->num_dacs; ) {
-		memset(ctl, 0, sizeof(*ctl));
-		if (ak->channel_names == NULL) {
-			strcpy(ctl->id.name, "DAC Volume");
+		memset(&knew, 0, sizeof(knew));
+		if (! ak->dac_info || ! ak->dac_info[mixer_ch].name) {
+			knew.name = "DAC Volume";
+			knew.index = mixer_ch + ak->idx_offset * 2;
 			num_stereo = 1;
-			ctl->id.index = mixer_ch + ak->idx_offset * 2;
 		} else {
-			strcpy(ctl->id.name, ak->channel_names[mixer_ch]);
-			num_stereo = ak->num_stereo[mixer_ch];
-			ctl->id.index = 0;
+			knew.name = ak->dac_info[mixer_ch].name;
+			num_stereo = ak->dac_info[mixer_ch].num_channels;
 		}
-		ctl->id.iface = SNDRV_CTL_ELEM_IFACE_MIXER;
-		ctl->count = 1;
+		knew.iface = SNDRV_CTL_ELEM_IFACE_MIXER;
+		knew.count = 1;
+		knew.access = SNDRV_CTL_ELEM_ACCESS_READWRITE |
+			SNDRV_CTL_ELEM_ACCESS_TLV_READ;
 		if (num_stereo == 2) {
-			ctl->info = snd_akm4xxx_stereo_volume_info;
-			ctl->get = snd_akm4xxx_stereo_volume_get;
-			ctl->put = snd_akm4xxx_stereo_volume_put;
+			knew.info = snd_akm4xxx_stereo_volume_info;
+			knew.get = snd_akm4xxx_stereo_volume_get;
+			knew.put = snd_akm4xxx_stereo_volume_put;
 		} else {
-			ctl->info = snd_akm4xxx_volume_info;
-			ctl->get = snd_akm4xxx_volume_get;
-			ctl->put = snd_akm4xxx_volume_put;
+			knew.info = snd_akm4xxx_volume_info;
+			knew.get = snd_akm4xxx_volume_get;
+			knew.put = snd_akm4xxx_volume_put;
 		}
 		switch (ak->type) {
 		case SND_AK4524:
 			/* register 6 & 7 */
-			ctl->private_value =
-				AK_COMPOSE(idx/2, (idx%2) + 6, 0, 127);
+			knew.private_value =
+				AK_COMPOSE(idx/2, (idx%2) + 6, 0, 127) |
+				AK_VOL_CVT;
+			knew.tlv.p = db_scale_vol_datt;
 			break;
 		case SND_AK4528:
 			/* register 4 & 5 */
-			ctl->private_value =
-				AK_COMPOSE(idx/2, (idx%2) + 4, 0, 127);
+			knew.private_value =
+				AK_COMPOSE(idx/2, (idx%2) + 4, 0, 127) |
+				AK_VOL_CVT;
+			knew.tlv.p = db_scale_vol_datt;
 			break;
 		case SND_AK4529: {
 			/* registers 2-7 and b,c */
 			int val = idx < 6 ? idx + 2 : (idx - 6) + 0xb;
-			ctl->private_value =
+			knew.private_value =
 				AK_COMPOSE(0, val, 0, 255) | AK_INVERT;
+			knew.tlv.p = db_scale_8bit;
 			break;
 		}
 		case SND_AK4355:
 			/* register 4-9, chip #0 only */
-			ctl->private_value = AK_COMPOSE(0, idx + 4, 0, 255);
+			knew.private_value = AK_COMPOSE(0, idx + 4, 0, 255);
+			knew.tlv.p = db_scale_8bit;
 			break;
 		case SND_AK4358: {
 			/* register 4-9 and 11-12, chip #0 only */
 			int  addr = idx < 6 ? idx + 4 : idx + 5;
-			ctl->private_value =
+			knew.private_value =
 				AK_COMPOSE(0, addr, 0, 127) | AK_NEEDSMSB;
+			knew.tlv.p = db_scale_7bit;
 			break;
 		}
 		case SND_AK4381:
 			/* register 3 & 4 */
-			ctl->private_value =
+			knew.private_value =
 				AK_COMPOSE(idx/2, (idx%2) + 3, 0, 255);
+			knew.tlv.p = db_scale_linear;
 			break;
 		default:
-			err = -EINVAL;
-			goto __error;
+			return -EINVAL;
 		}
 
-		ctl->private_data = ak;
-		err = snd_ctl_add(ak->card,
-				  snd_ctl_new(ctl, SNDRV_CTL_ELEM_ACCESS_READ|
-					      SNDRV_CTL_ELEM_ACCESS_WRITE));
+		err = snd_ctl_add(ak->card, snd_ctl_new1(&knew, ak));
 		if (err < 0)
-			goto __error;
+			return err;
 
 		idx += num_stereo;
 		mixer_ch++;
 	}
-	for (idx = 0; idx < ak->num_adcs && ak->type == SND_AK4524; ++idx) {
-		memset(ctl, 0, sizeof(*ctl));
-		strcpy(ctl->id.name, "ADC Volume");
-		ctl->id.index = idx + ak->idx_offset * 2;
-		ctl->id.iface = SNDRV_CTL_ELEM_IFACE_MIXER;
-		ctl->count = 1;
-		ctl->info = snd_akm4xxx_volume_info;
-		ctl->get = snd_akm4xxx_volume_get;
-		ctl->put = snd_akm4xxx_volume_put;
-		/* register 4 & 5 */
-		ctl->private_value =
-			AK_COMPOSE(idx/2, (idx%2) + 4, 0, 127);
-		ctl->private_data = ak;
-		err = snd_ctl_add(ak->card,
-				  snd_ctl_new(ctl, SNDRV_CTL_ELEM_ACCESS_READ|
-					      SNDRV_CTL_ELEM_ACCESS_WRITE));
-		if (err < 0)
-			goto __error;
-
-		memset(ctl, 0, sizeof(*ctl));
-		strcpy(ctl->id.name, "IPGA Analog Capture Volume");
-		ctl->id.index = idx + ak->idx_offset * 2;
-		ctl->id.iface = SNDRV_CTL_ELEM_IFACE_MIXER;
-		ctl->count = 1;
-		ctl->info = snd_akm4xxx_ipga_gain_info;
-		ctl->get = snd_akm4xxx_ipga_gain_get;
-		ctl->put = snd_akm4xxx_ipga_gain_put;
+	return 0;
+}
+
+static int build_adc_controls(struct snd_akm4xxx *ak)
+{
+	int idx, err, mixer_ch, num_stereo;
+	struct snd_kcontrol_new knew;
+
+	mixer_ch = 0;
+	for (idx = 0; idx < ak->num_adcs;) {
+		memset(&knew, 0, sizeof(knew));
+		if (! ak->adc_info || ! ak->adc_info[mixer_ch].name) {
+			knew.name = "ADC Volume";
+			knew.index = mixer_ch + ak->idx_offset * 2;
+			num_stereo = 1;
+		} else {
+			knew.name = ak->adc_info[mixer_ch].name;
+			num_stereo = ak->adc_info[mixer_ch].num_channels;
+		}
+		knew.iface = SNDRV_CTL_ELEM_IFACE_MIXER;
+		knew.count = 1;
+		knew.access = SNDRV_CTL_ELEM_ACCESS_READWRITE |
+			SNDRV_CTL_ELEM_ACCESS_TLV_READ;
+		if (num_stereo == 2) {
+			knew.info = snd_akm4xxx_stereo_volume_info;
+			knew.get = snd_akm4xxx_stereo_volume_get;
+			knew.put = snd_akm4xxx_stereo_volume_put;
+		} else {
+			knew.info = snd_akm4xxx_volume_info;
+			knew.get = snd_akm4xxx_volume_get;
+			knew.put = snd_akm4xxx_volume_put;
+		}
 		/* register 4 & 5 */
-		ctl->private_value = AK_COMPOSE(idx/2, (idx%2) + 4, 0, 0);
-		ctl->private_data = ak;
-		err = snd_ctl_add(ak->card,
-				  snd_ctl_new(ctl, SNDRV_CTL_ELEM_ACCESS_READ|
-					      SNDRV_CTL_ELEM_ACCESS_WRITE));
+		knew.private_value =
+			AK_COMPOSE(idx/2, (idx%2) + 4, 0, 127) |
+			AK_VOL_CVT;
+		knew.tlv.p = db_scale_vol_datt;
+		err = snd_ctl_add(ak->card, snd_ctl_new1(&knew, ak));
 		if (err < 0)
-			goto __error;
-	}
+			return err;
 
-	if (ak->type == SND_AK5365) {
-		memset(ctl, 0, sizeof(*ctl));
-		if (ak->channel_names == NULL)
-			strcpy(ctl->id.name, "Capture Volume");
+		if (! ak->adc_info || ! ak->adc_info[mixer_ch].gain_name)
+			knew.name = "IPGA Analog Capture Volume";
 		else
-			strcpy(ctl->id.name, ak->channel_names[0]);
-		ctl->id.index = ak->idx_offset * 2;
-		ctl->id.iface = SNDRV_CTL_ELEM_IFACE_MIXER;
-		ctl->count = 1;
-		ctl->info = snd_akm4xxx_stereo_volume_info;
-		ctl->get = snd_akm4xxx_stereo_volume_get;
-		ctl->put = snd_akm4xxx_stereo_volume_put;
-		/* Registers 4 & 5 (see AK5365 data sheet, pages 34 and 35):
-		 * valid values are from 0x00 (mute) to 0x98 (+12dB).  */
-		ctl->private_value =
-			AK_COMPOSE(0, 4, 0, 0x98);
-		ctl->private_data = ak;
-		err = snd_ctl_add(ak->card,
-				  snd_ctl_new(ctl, SNDRV_CTL_ELEM_ACCESS_READ|
-					      SNDRV_CTL_ELEM_ACCESS_WRITE));
+			knew.name = ak->adc_info[mixer_ch].gain_name;
+		if (num_stereo == 2) {
+			knew.info = snd_akm4xxx_stereo_gain_info;
+			knew.get = snd_akm4xxx_stereo_gain_get;
+			knew.put = snd_akm4xxx_stereo_gain_put;
+		} else {
+			knew.info = snd_akm4xxx_ipga_gain_info;
+			knew.get = snd_akm4xxx_ipga_gain_get;
+			knew.put = snd_akm4xxx_ipga_gain_put;
+		}
+		/* register 4 & 5 */
+		if (ak->type == SND_AK4524)
+			knew.private_value = AK_COMPOSE(idx/2, (idx%2) + 4, 0,
+							24);
+		else /* AK5365 */
+			knew.private_value = AK_COMPOSE(idx/2, (idx%2) + 4, 0,
+							36);
+		knew.tlv.p = db_scale_ipga;
+		err = snd_ctl_add(ak->card, snd_ctl_new1(&knew, ak));
 		if (err < 0)
-			goto __error;
+			return err;
+
+		if (ak->type == SND_AK5365 && (idx % 2) == 0) {
+			if (! ak->adc_info || 
+			    ! ak->adc_info[mixer_ch].switch_name)
+				knew.name = "Capture Switch";
+			else
+				knew.name = ak->adc_info[mixer_ch].switch_name;
+			knew.info = ak4xxx_switch_info;
+			knew.get = ak4xxx_switch_get;
+			knew.put = ak4xxx_switch_put;
+			knew.access = 0;
+			/* register 2, bit 0 (SMUTE): 0 = normal operation,
+			   1 = mute */
+			knew.private_value =
+				AK_COMPOSE(idx/2, 2, 0, 0) | AK_INVERT;
+			err = snd_ctl_add(ak->card, snd_ctl_new1(&knew, ak));
+			if (err < 0)
+				return err;
+		}
 
-		memset(ctl, 0, sizeof(*ctl));
-		if (ak->channel_names == NULL)
-			strcpy(ctl->id.name, "Capture Switch");
-		else
-			strcpy(ctl->id.name, ak->channel_names[1]);
-		ctl->id.index = ak->idx_offset * 2;
-		ctl->id.iface = SNDRV_CTL_ELEM_IFACE_MIXER;
-		ctl->count = 1;
-		ctl->info = ak4xxx_switch_info;
-		ctl->get = ak4xxx_switch_get;
-		ctl->put = ak4xxx_switch_put;
-		/* register 2, bit 0 (SMUTE): 0 = normal operation, 1 = mute */
-		ctl->private_value =
-			AK_COMPOSE(0, 2, 0, 0) | AK_INVERT;
-		ctl->private_data = ak;
-		err = snd_ctl_add(ak->card,
-				  snd_ctl_new(ctl, SNDRV_CTL_ELEM_ACCESS_READ|
-					      SNDRV_CTL_ELEM_ACCESS_WRITE));
-		if (err < 0)
-			goto __error;
+		idx += num_stereo;
+		mixer_ch++;
 	}
+	return 0;
+}
+
+static int build_deemphasis(struct snd_akm4xxx *ak, int num_emphs)
+{
+	int idx, err;
+	struct snd_kcontrol_new knew;
 
-	if (ak->type == SND_AK4355 || ak->type == SND_AK4358)
-		num_emphs = 1;
-	else
-		num_emphs = ak->num_dacs / 2;
 	for (idx = 0; idx < num_emphs; idx++) {
-		memset(ctl, 0, sizeof(*ctl));
-		strcpy(ctl->id.name, "Deemphasis");
-		ctl->id.index = idx + ak->idx_offset;
-		ctl->id.iface = SNDRV_CTL_ELEM_IFACE_MIXER;
-		ctl->count = 1;
-		ctl->info = snd_akm4xxx_deemphasis_info;
-		ctl->get = snd_akm4xxx_deemphasis_get;
-		ctl->put = snd_akm4xxx_deemphasis_put;
+		memset(&knew, 0, sizeof(knew));
+		knew.name = "Deemphasis";
+		knew.index = idx + ak->idx_offset;
+		knew.iface = SNDRV_CTL_ELEM_IFACE_MIXER;
+		knew.count = 1;
+		knew.info = snd_akm4xxx_deemphasis_info;
+		knew.get = snd_akm4xxx_deemphasis_get;
+		knew.put = snd_akm4xxx_deemphasis_put;
 		switch (ak->type) {
 		case SND_AK4524:
 		case SND_AK4528:
 			/* register 3 */
-			ctl->private_value = AK_COMPOSE(idx, 3, 0, 0);
+			knew.private_value = AK_COMPOSE(idx, 3, 0, 0);
 			break;
 		case SND_AK4529: {
 			int shift = idx == 3 ? 6 : (2 - idx) * 2;
 			/* register 8 with shift */
-			ctl->private_value = AK_COMPOSE(0, 8, shift, 0);
+			knew.private_value = AK_COMPOSE(0, 8, shift, 0);
 			break;
 		}
 		case SND_AK4355:
 		case SND_AK4358:
-			ctl->private_value = AK_COMPOSE(idx, 3, 0, 0);
+			knew.private_value = AK_COMPOSE(idx, 3, 0, 0);
 			break;
 		case SND_AK4381:
-			ctl->private_value = AK_COMPOSE(idx, 1, 1, 0);
+			knew.private_value = AK_COMPOSE(idx, 1, 1, 0);
 			break;
 		default:
-			err = -EINVAL;
-			goto __error;
+			return -EINVAL;
 		}
-		ctl->private_data = ak;
-		err = snd_ctl_add(ak->card,
-				  snd_ctl_new(ctl, SNDRV_CTL_ELEM_ACCESS_READ|
-					      SNDRV_CTL_ELEM_ACCESS_WRITE));
+		err = snd_ctl_add(ak->card, snd_ctl_new1(&knew, ak));
 		if (err < 0)
-			goto __error;
+			return err;
 	}
-	err = 0;
-
- __error:
-	kfree(ctl);
-	return err;
+	return 0;
 }
 
+int snd_akm4xxx_build_controls(struct snd_akm4xxx *ak)
+{
+	int err, num_emphs;
+
+	err = build_dac_controls(ak);
+	if (err < 0)
+		return err;
+
+	if (ak->type == SND_AK4524 || ak->type == SND_AK5365) {
+		err = build_adc_controls(ak);
+		if (err < 0)
+			return err;
+	}
+
+	if (ak->type == SND_AK4355 || ak->type == SND_AK4358)
+		num_emphs = 1;
+	else
+		num_emphs = ak->num_dacs / 2;
+	err = build_deemphasis(ak, num_emphs);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+	
 EXPORT_SYMBOL(snd_akm4xxx_build_controls);
 
 static int __init alsa_akm4xxx_module_init(void)
diff --git a/sound/pci/ice1712/revo.c b/sound/pci/ice1712/revo.c
index 1134a57f9e65..c9eefa9bbfff 100644
--- a/sound/pci/ice1712/revo.c
+++ b/sound/pci/ice1712/revo.c
@@ -87,19 +87,34 @@ static void revo_set_rate_val(struct snd_akm4xxx *ak, unsigned int rate)
  * initialize the chips on M-Audio Revolution cards
  */
 
-static unsigned int revo71_num_stereo_front[] = {2};
-static char *revo71_channel_names_front[] = {"PCM Playback Volume"};
+#define AK_DAC(xname,xch) { .name = xname, .num_channels = xch }
 
-static unsigned int revo71_num_stereo_surround[] = {1, 1, 2, 2};
-static char *revo71_channel_names_surround[] = {"PCM Center Playback Volume", "PCM LFE Playback Volume",
-						"PCM Side Playback Volume", "PCM Rear Playback Volume"};
+static struct snd_akm4xxx_dac_channel revo71_front[] = {
+	AK_DAC("PCM Playback Volume", 2)
+};
+
+static struct snd_akm4xxx_dac_channel revo71_surround[] = {
+	AK_DAC("PCM Center Playback Volume", 1),
+	AK_DAC("PCM LFE Playback Volume", 1),
+	AK_DAC("PCM Side Playback Volume", 2),
+	AK_DAC("PCM Rear Playback Volume", 2),
+};
 
-static unsigned int revo51_num_stereo[] = {2, 1, 1, 2};
-static char *revo51_channel_names[] = {"PCM Playback Volume", "PCM Center Playback Volume",
-					"PCM LFE Playback Volume", "PCM Rear Playback Volume"};
+static struct snd_akm4xxx_dac_channel revo51_dac[] = {
+	AK_DAC("PCM Playback Volume", 2),
+	AK_DAC("PCM Center Playback Volume", 1),
+	AK_DAC("PCM LFE Playback Volume", 1),
+	AK_DAC("PCM Rear Playback Volume", 2),
+};
 
-static unsigned int revo51_adc_num_stereo[] = {2};
-static char *revo51_adc_channel_names[] = {"PCM Capture Volume","PCM Capture Switch"};
+static struct snd_akm4xxx_adc_channel revo51_adc[] = {
+	{
+		.name = "PCM Capture Volume",
+		.gain_name = "PCM Capture Gain Volume",
+		.switch_name = "PCM Capture Switch",
+		.num_channels = 2
+	},
+};
 
 static struct snd_akm4xxx akm_revo_front __devinitdata = {
 	.type = SND_AK4381,
@@ -107,8 +122,7 @@ static struct snd_akm4xxx akm_revo_front __devinitdata = {
 	.ops = {
 		.set_rate_val = revo_set_rate_val
 	},
-	.num_stereo = revo71_num_stereo_front,
-	.channel_names = revo71_channel_names_front
+	.dac_info = revo71_front,
 };
 
 static struct snd_ak4xxx_private akm_revo_front_priv __devinitdata = {
@@ -130,8 +144,7 @@ static struct snd_akm4xxx akm_revo_surround __devinitdata = {
 	.ops = {
 		.set_rate_val = revo_set_rate_val
 	},
-	.num_stereo = revo71_num_stereo_surround,
-	.channel_names = revo71_channel_names_surround
+	.dac_info = revo71_surround,
 };
 
 static struct snd_ak4xxx_private akm_revo_surround_priv __devinitdata = {
@@ -152,8 +165,7 @@ static struct snd_akm4xxx akm_revo51 __devinitdata = {
 	.ops = {
 		.set_rate_val = revo_set_rate_val
 	},
-	.num_stereo = revo51_num_stereo,
-	.channel_names = revo51_channel_names
+	.dac_info = revo51_dac,
 };
 
 static struct snd_ak4xxx_private akm_revo51_priv __devinitdata = {
@@ -171,8 +183,7 @@ static struct snd_ak4xxx_private akm_revo51_priv __devinitdata = {
 static struct snd_akm4xxx akm_revo51_adc __devinitdata = {
 	.type = SND_AK5365,
 	.num_adcs = 2,
-	.num_stereo = revo51_adc_num_stereo,
-	.channel_names = revo51_adc_channel_names
+	.adc_info = revo51_adc,
 };
 
 static struct snd_ak4xxx_private akm_revo51_adc_priv __devinitdata = {
-- 
cgit v1.2.3


From 55a29af5ed5d914f017e6a7c613a4d7cc34f82d9 Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Wed, 6 Sep 2006 12:15:34 +0200
Subject: [ALSA] Add definition of TLV dB range compound

Added the definition of TLV dB range compound.  It contains one or
more dB-range or linear-volume TLV entries with min/max ranges.
Used for volume controls with non-linear curves.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 include/sound/tlv.h | 23 +++++++++++++++--------
 1 file changed, 15 insertions(+), 8 deletions(-)

(limited to 'include')

diff --git a/include/sound/tlv.h b/include/sound/tlv.h
index 7905841643df..d93a96b91875 100644
--- a/include/sound/tlv.h
+++ b/include/sound/tlv.h
@@ -34,19 +34,26 @@
 #define SNDRV_CTL_TLVT_CONTAINER 0	/* one level down - group of TLVs */
 #define SNDRV_CTL_TLVT_DB_SCALE	1       /* dB scale */
 #define SNDRV_CTL_TLVT_DB_LINEAR 2	/* linear volume */
+#define SNDRV_CTL_TLVT_DB_RANGE 3	/* dB range container */
 
+#define TLV_DB_SCALE_ITEM(min, step, mute)			\
+	SNDRV_CTL_TLVT_DB_SCALE, 2 * sizeof(unsigned int),	\
+	(min), ((step) & 0xffff) | ((mute) ? 0x10000 : 0)
 #define DECLARE_TLV_DB_SCALE(name, min, step, mute) \
-unsigned int name[] = { \
-        SNDRV_CTL_TLVT_DB_SCALE, 2 * sizeof(unsigned int), \
-        (min), ((step) & 0xffff) | ((mute) ? 0x10000 : 0) \
-}
+	unsigned int name[] = { TLV_DB_SCALE_ITEM(min, step, mute) }
 
 /* linear volume between min_dB and max_dB (.01dB unit) */
+#define TLV_DB_LINEAR_ITEM(min_dB, max_dB)		    \
+	SNDRV_CTL_TLVT_DB_LINEAR, 2 * sizeof(unsigned int), \
+	(min_dB), (max_dB)
 #define DECLARE_TLV_DB_LINEAR(name, min_dB, max_dB)	\
-unsigned int name[] = { \
-        SNDRV_CTL_TLVT_DB_LINEAR, 2 * sizeof(unsigned int), \
-        (min_dB), (max_dB)				\
-}
+	unsigned int name[] = { TLV_DB_LINEAR_ITEM(min_dB, max_dB) }
+
+/* dB range container */
+/* Each item is: <min> <max> <TLV> */
+/* The below assumes that each item TLV is 4 words like DB_SCALE or LINEAR */
+#define TLV_DB_RANGE_HEAD(num)			\
+	SNDRV_CTL_TLVT_DB_RANGE, 6 * (num) * sizeof(unsigned int)
 
 #define TLV_DB_GAIN_MUTE	-9999999
 
-- 
cgit v1.2.3


From 9d19f48cfe2570562c2c6226780a7ca627b0f1f1 Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Wed, 6 Sep 2006 14:27:46 +0200
Subject: [ALSA] Add pcm_class attribute to PCM sysfs entry

This patch adds a new attribute, pcm_class, to each PCM sysfs entry.
It's useful to detect what kind of PCM stream is, for example, HAL
can check whether it's a modem or not.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 include/sound/core.h |  4 ++++
 sound/core/pcm.c     | 24 ++++++++++++++++++++++
 sound/core/sound.c   | 56 +++++++++++++++++++++++++++++++++++++++++-----------
 3 files changed, 72 insertions(+), 12 deletions(-)

(limited to 'include')

diff --git a/include/sound/core.h b/include/sound/core.h
index 1359c532b68e..b056ea925ecf 100644
--- a/include/sound/core.h
+++ b/include/sound/core.h
@@ -26,6 +26,7 @@
 #include <linux/mutex.h>		/* struct mutex */
 #include <linux/rwsem.h>		/* struct rw_semaphore */
 #include <linux/pm.h>			/* pm_message_t */
+#include <linux/device.h>
 
 /* forward declarations */
 #ifdef CONFIG_PCI
@@ -186,6 +187,7 @@ struct snd_minor {
 	int device;			/* device number */
 	const struct file_operations *f_ops;	/* file operations */
 	void *private_data;		/* private data for f_ops->open */
+	struct class_device *class_dev;	/* class device for sysfs */
 };
 
 /* sound.c */
@@ -200,6 +202,8 @@ int snd_register_device(int type, struct snd_card *card, int dev,
 			const char *name);
 int snd_unregister_device(int type, struct snd_card *card, int dev);
 void *snd_lookup_minor_data(unsigned int minor, int type);
+int snd_add_device_sysfs_file(int type, struct snd_card *card, int dev,
+			      const struct class_device_attribute *attr);
 
 #ifdef CONFIG_SND_OSSEMUL
 int snd_register_oss_device(int type, struct snd_card *card, int dev,
diff --git a/sound/core/pcm.c b/sound/core/pcm.c
index ed3b09469560..bf8f412988b8 100644
--- a/sound/core/pcm.c
+++ b/sound/core/pcm.c
@@ -907,6 +907,28 @@ void snd_pcm_detach_substream(struct snd_pcm_substream *substream)
 	substream->pstr->substream_opened--;
 }
 
+static ssize_t show_pcm_class(struct class_device *class_device, char *buf)
+{
+	struct snd_pcm *pcm;
+	const char *str;
+	static const char *strs[SNDRV_PCM_CLASS_LAST + 1] = {
+		[SNDRV_PCM_CLASS_GENERIC] = "generic",
+		[SNDRV_PCM_CLASS_MULTI] = "multi",
+		[SNDRV_PCM_CLASS_MODEM] = "modem",
+		[SNDRV_PCM_CLASS_DIGITIZER] = "digitizer",
+	};
+
+	if (! (pcm = class_get_devdata(class_device)) ||
+	    pcm->dev_class > SNDRV_PCM_CLASS_LAST)
+		str = "none";
+	else
+		str = strs[pcm->dev_class];
+        return snprintf(buf, PAGE_SIZE, "%s\n", str);
+}
+
+static struct class_device_attribute pcm_attrs =
+	__ATTR(pcm_class, S_IRUGO, show_pcm_class, NULL);
+
 static int snd_pcm_dev_register(struct snd_device *device)
 {
 	int cidx, err;
@@ -945,6 +967,8 @@ static int snd_pcm_dev_register(struct snd_device *device)
 			mutex_unlock(&register_mutex);
 			return err;
 		}
+		snd_add_device_sysfs_file(devtype, pcm->card, pcm->device,
+					  &pcm_attrs);
 		for (substream = pcm->streams[cidx].substream; substream; substream = substream->next)
 			snd_pcm_timer_init(substream);
 	}
diff --git a/sound/core/sound.c b/sound/core/sound.c
index b4430db3fa4c..efa476c5210a 100644
--- a/sound/core/sound.c
+++ b/sound/core/sound.c
@@ -268,7 +268,11 @@ int snd_register_device(int type, struct snd_card *card, int dev,
 	snd_minors[minor] = preg;
 	if (card)
 		device = card->dev;
-	class_device_create(sound_class, NULL, MKDEV(major, minor), device, "%s", name);
+	preg->class_dev = class_device_create(sound_class, NULL,
+					      MKDEV(major, minor),
+					      device, "%s", name);
+	if (preg->class_dev)
+		class_set_devdata(preg->class_dev, private_data);
 
 	mutex_unlock(&sound_mutex);
 	return 0;
@@ -276,6 +280,24 @@ int snd_register_device(int type, struct snd_card *card, int dev,
 
 EXPORT_SYMBOL(snd_register_device);
 
+/* find the matching minor record
+ * return the index of snd_minor, or -1 if not found
+ */
+static int find_snd_minor(int type, struct snd_card *card, int dev)
+{
+	int cardnum, minor;
+	struct snd_minor *mptr;
+
+	cardnum = card ? card->number : -1;
+	for (minor = 0; minor < ARRAY_SIZE(snd_minors); ++minor)
+		if ((mptr = snd_minors[minor]) != NULL &&
+		    mptr->type == type &&
+		    mptr->card == cardnum &&
+		    mptr->device == dev)
+			return minor;
+	return -1;
+}
+
 /**
  * snd_unregister_device - unregister the device on the given card
  * @type: the device type, SNDRV_DEVICE_TYPE_XXX
@@ -289,32 +311,42 @@ EXPORT_SYMBOL(snd_register_device);
  */
 int snd_unregister_device(int type, struct snd_card *card, int dev)
 {
-	int cardnum, minor;
-	struct snd_minor *mptr;
+	int minor;
 
-	cardnum = card ? card->number : -1;
 	mutex_lock(&sound_mutex);
-	for (minor = 0; minor < ARRAY_SIZE(snd_minors); ++minor)
-		if ((mptr = snd_minors[minor]) != NULL &&
-		    mptr->type == type &&
-		    mptr->card == cardnum &&
-		    mptr->device == dev)
-			break;
-	if (minor == ARRAY_SIZE(snd_minors)) {
+	minor = find_snd_minor(type, card, dev);
+	if (minor < 0) {
 		mutex_unlock(&sound_mutex);
 		return -EINVAL;
 	}
 
 	class_device_destroy(sound_class, MKDEV(major, minor));
 
+	kfree(snd_minors[minor]);
 	snd_minors[minor] = NULL;
 	mutex_unlock(&sound_mutex);
-	kfree(mptr);
 	return 0;
 }
 
 EXPORT_SYMBOL(snd_unregister_device);
 
+int snd_add_device_sysfs_file(int type, struct snd_card *card, int dev,
+			      const struct class_device_attribute *attr)
+{
+	int minor, ret = -EINVAL;
+	struct class_device *cdev;
+
+	mutex_lock(&sound_mutex);
+	minor = find_snd_minor(type, card, dev);
+	if (minor >= 0 && (cdev = snd_minors[minor]->class_dev) != NULL)
+		ret = class_device_create_file(cdev, attr);
+	mutex_unlock(&sound_mutex);
+	return ret;
+
+}
+
+EXPORT_SYMBOL(snd_add_device_sysfs_file);
+
 #ifdef CONFIG_PROC_FS
 /*
  *  INFO PART
-- 
cgit v1.2.3


From 854b66e44260320c21ebe4b8a18e189f2e45b5be Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Fri, 8 Sep 2006 12:27:38 +0200
Subject: [ALSA] ak4xxx - Remove bogus IPGA controls

Remove IPGA volume controls and merge the IPGA range to ADC volume
controls.  These two volumes are not really independent but connected
simply in different ranges 0-0x7f and 0x80-max.  It doesn't make sense
to provide two controls.
Since both 0x7f and 0x80 specify 0dB, a hack is needed for IPGA range
to skip 0x80 (increment one) for such controls.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
---
 include/sound/ak4xxx-adda.h   |  10 ----
 sound/i2c/other/ak4xxx-adda.c | 128 ++++++------------------------------------
 sound/pci/ice1712/revo.c      |   1 -
 3 files changed, 18 insertions(+), 121 deletions(-)

(limited to 'include')

diff --git a/include/sound/ak4xxx-adda.h b/include/sound/ak4xxx-adda.h
index 026e4072a9a1..d0deca669b92 100644
--- a/include/sound/ak4xxx-adda.h
+++ b/include/sound/ak4xxx-adda.h
@@ -48,7 +48,6 @@ struct snd_akm4xxx_dac_channel {
 /* ADC labels and channels */
 struct snd_akm4xxx_adc_channel {
 	char *name;		/* capture gain volume label */
-	char *gain_name;	/* IPGA */
 	char *switch_name;	/* capture switch */
 	unsigned int num_channels;
 };
@@ -91,13 +90,4 @@ int snd_akm4xxx_build_controls(struct snd_akm4xxx *ak);
 #define snd_akm4xxx_set_vol(ak,chip,reg,val) \
 	((ak)->volumes[(chip) * 16 + (reg)] = (val))
 
-/* Warning: IPGA is tricky - we assume the addr + 4 is unused
- *   so far, it's OK for all AK codecs with IPGA:
- *   AK4524, AK4528 and EK5365
- */
-#define snd_akm4xxx_get_ipga(ak,chip,reg) \
-	snd_akm4xxx_get_vol(ak, chip, (reg) + 4)
-#define snd_akm4xxx_set_ipga(ak,chip,reg,val) \
-	snd_akm4xxx_set_vol(ak, chip, (reg) + 4, val)
-
 #endif /* __SOUND_AK4XXX_ADDA_H */
diff --git a/sound/i2c/other/ak4xxx-adda.c b/sound/i2c/other/ak4xxx-adda.c
index c34cb4684607..5da49e2eb350 100644
--- a/sound/i2c/other/ak4xxx-adda.c
+++ b/sound/i2c/other/ak4xxx-adda.c
@@ -43,10 +43,7 @@ void snd_akm4xxx_write(struct snd_akm4xxx *ak, int chip, unsigned char reg,
 	ak->ops.write(ak, chip, reg, val);
 
 	/* save the data */
-	/* don't overwrite with IPGA data */
-	if ((ak->type != SND_AK4524 && ak->type != SND_AK5365) ||
-	    (reg != 0x04 && reg != 0x05) || (val & 0x80) == 0)
-		snd_akm4xxx_set(ak, chip, reg, val);
+	snd_akm4xxx_set(ak, chip, reg, val);
 	ak->ops.unlock(ak, chip);
 }
 
@@ -70,12 +67,6 @@ static void ak4524_reset(struct snd_akm4xxx *ak, int state)
 		for (reg = 0x04; reg < maxreg; reg++)
 			snd_akm4xxx_write(ak, chip, reg,
 					  snd_akm4xxx_get(ak, chip, reg));
-		if (ak->type == SND_AK4528)
-			continue;
-		/* IPGA */
-		for (reg = 0x04; reg < 0x06; reg++)
-			snd_akm4xxx_write(ak, chip, reg,
-					  snd_akm4xxx_get_ipga(ak, chip, reg) | 0x80);
 	}
 }
 
@@ -175,7 +166,6 @@ static DECLARE_TLV_DB_SCALE(db_scale_vol_datt, -6350, 50, 1);
 static DECLARE_TLV_DB_SCALE(db_scale_8bit, -12750, 50, 1);
 static DECLARE_TLV_DB_SCALE(db_scale_7bit, -6350, 50, 1);
 static DECLARE_TLV_DB_LINEAR(db_scale_linear, TLV_DB_GAIN_MUTE, 0);
-static DECLARE_TLV_DB_SCALE(db_scale_ipga, 0, 50, 0);
 
 /*
  * initialize all the ak4xxx chips
@@ -190,8 +180,6 @@ void snd_akm4xxx_init(struct snd_akm4xxx *ak)
 		0x01, 0x03, /* 1: ADC/DAC enable */
 		0x04, 0x00, /* 4: ADC left muted */
 		0x05, 0x00, /* 5: ADC right muted */
-		0x04, 0x80, /* 4: ADC IPGA gain 0dB */
-		0x05, 0x80, /* 5: ADC IPGA gain 0dB */
 		0x06, 0x00, /* 6: DAC left muted */
 		0x07, 0x00, /* 7: DAC right muted */
 		0xff, 0xff
@@ -324,13 +312,15 @@ EXPORT_SYMBOL(snd_akm4xxx_init);
 /*
  * Mixer callbacks
  */
+#define AK_IPGA 			(1<<20)	/* including IPGA */
 #define AK_VOL_CVT 			(1<<21)	/* need dB conversion */
 #define AK_NEEDSMSB 			(1<<22)	/* need MSB update bit */
 #define AK_INVERT 			(1<<23)	/* data is inverted */
 #define AK_GET_CHIP(val)		(((val) >> 8) & 0xff)
 #define AK_GET_ADDR(val)		((val) & 0xff)
-#define AK_GET_SHIFT(val)		(((val) >> 16) & 0x1f)
+#define AK_GET_SHIFT(val)		(((val) >> 16) & 0x0f)
 #define AK_GET_VOL_CVT(val)		(((val) >> 21) & 1)
+#define AK_GET_IPGA(val)		(((val) >> 20) & 1)
 #define AK_GET_NEEDSMSB(val)		(((val) >> 22) & 1)
 #define AK_GET_INVERT(val)		(((val) >> 23) & 1)
 #define AK_GET_MASK(val)		(((val) >> 24) & 0xff)
@@ -371,8 +361,10 @@ static int put_ak_reg(struct snd_kcontrol *kcontrol, int addr,
 		return 0;
 
 	snd_akm4xxx_set_vol(ak, chip, addr, nval);
-	if (AK_GET_VOL_CVT(kcontrol->private_value))
+	if (AK_GET_VOL_CVT(kcontrol->private_value) && nval < 128)
 		nval = vol_cvt_datt[nval];
+	if (AK_GET_IPGA(kcontrol->private_value) && nval >= 128)
+		nval++; /* need to correct + 1 since both 127 and 128 are 0dB */
 	if (AK_GET_INVERT(kcontrol->private_value))
 		nval = mask - nval;
 	if (AK_GET_NEEDSMSB(kcontrol->private_value))
@@ -424,68 +416,6 @@ static int snd_akm4xxx_stereo_volume_put(struct snd_kcontrol *kcontrol,
 	return change;
 }
 
-#define snd_akm4xxx_ipga_gain_info	snd_akm4xxx_volume_info
-
-static int snd_akm4xxx_ipga_gain_get(struct snd_kcontrol *kcontrol,
-				     struct snd_ctl_elem_value *ucontrol)
-{
-	struct snd_akm4xxx *ak = snd_kcontrol_chip(kcontrol);
-	int chip = AK_GET_CHIP(kcontrol->private_value);
-	int addr = AK_GET_ADDR(kcontrol->private_value);
-
-	ucontrol->value.integer.value[0] =
-		snd_akm4xxx_get_ipga(ak, chip, addr);
-	return 0;
-}
-
-static int put_ak_ipga(struct snd_kcontrol *kcontrol, int addr,
-		       unsigned char nval)
-{
-	struct snd_akm4xxx *ak = snd_kcontrol_chip(kcontrol);
-	int chip = AK_GET_CHIP(kcontrol->private_value);
-
-	if (snd_akm4xxx_get_ipga(ak, chip, addr) == nval)
-		return 0;
-	snd_akm4xxx_set_ipga(ak, chip, addr, nval);
-	snd_akm4xxx_write(ak, chip, addr, nval | 0x80); /* need MSB */
-	return 1;
-}
-
-static int snd_akm4xxx_ipga_gain_put(struct snd_kcontrol *kcontrol,
-				     struct snd_ctl_elem_value *ucontrol)
-{
-	return put_ak_ipga(kcontrol, AK_GET_ADDR(kcontrol->private_value),
-			   ucontrol->value.integer.value[0]);
-}
-
-#define snd_akm4xxx_stereo_gain_info	snd_akm4xxx_stereo_volume_info
-
-static int snd_akm4xxx_stereo_gain_get(struct snd_kcontrol *kcontrol,
-				       struct snd_ctl_elem_value *ucontrol)
-{
-	struct snd_akm4xxx *ak = snd_kcontrol_chip(kcontrol);
-	int chip = AK_GET_CHIP(kcontrol->private_value);
-	int addr = AK_GET_ADDR(kcontrol->private_value);
-
-	ucontrol->value.integer.value[0] =
-		snd_akm4xxx_get_ipga(ak, chip, addr);
-	ucontrol->value.integer.value[1] =
-		snd_akm4xxx_get_ipga(ak, chip, addr + 1);
-	return 0;
-}
-
-static int snd_akm4xxx_stereo_gain_put(struct snd_kcontrol *kcontrol,
-				       struct snd_ctl_elem_value *ucontrol)
-{
-	int addr = AK_GET_ADDR(kcontrol->private_value);
-	int change;
-
-	change = put_ak_ipga(kcontrol, addr, ucontrol->value.integer.value[0]);
-	change |= put_ak_ipga(kcontrol, addr + 1,
-			      ucontrol->value.integer.value[1]);
-	return change;
-}
-
 static int snd_akm4xxx_deemphasis_info(struct snd_kcontrol *kcontrol,
 				       struct snd_ctl_elem_info *uinfo)
 {
@@ -702,35 +632,15 @@ static int build_adc_controls(struct snd_akm4xxx *ak)
 			knew.put = snd_akm4xxx_volume_put;
 		}
 		/* register 4 & 5 */
-		knew.private_value =
-			AK_COMPOSE(idx/2, (idx%2) + 4, 0, 127) |
-			AK_VOL_CVT;
-		knew.tlv.p = db_scale_vol_datt;
-		err = snd_ctl_add(ak->card, snd_ctl_new1(&knew, ak));
-		if (err < 0)
-			return err;
-
-		if (! ak->adc_info || ! ak->adc_info[mixer_ch].gain_name)
-			knew.name = "IPGA Analog Capture Volume";
+		if (ak->type == SND_AK5365)
+			knew.private_value =
+				AK_COMPOSE(idx/2, (idx%2) + 4, 0, 151) |
+				AK_VOL_CVT | AK_IPGA;
 		else
-			knew.name = ak->adc_info[mixer_ch].gain_name;
-		if (num_stereo == 2) {
-			knew.info = snd_akm4xxx_stereo_gain_info;
-			knew.get = snd_akm4xxx_stereo_gain_get;
-			knew.put = snd_akm4xxx_stereo_gain_put;
-		} else {
-			knew.info = snd_akm4xxx_ipga_gain_info;
-			knew.get = snd_akm4xxx_ipga_gain_get;
-			knew.put = snd_akm4xxx_ipga_gain_put;
-		}
-		/* register 4 & 5 */
-		if (ak->type == SND_AK4524)
-			knew.private_value = AK_COMPOSE(idx/2, (idx%2) + 4, 0,
-							24);
-		else /* AK5365 */
-			knew.private_value = AK_COMPOSE(idx/2, (idx%2) + 4, 0,
-							36);
-		knew.tlv.p = db_scale_ipga;
+			knew.private_value =
+				AK_COMPOSE(idx/2, (idx%2) + 4, 0, 163) |
+				AK_VOL_CVT | AK_IPGA;
+		knew.tlv.p = db_scale_vol_datt;
 		err = snd_ctl_add(ak->card, snd_ctl_new1(&knew, ak));
 		if (err < 0)
 			return err;
@@ -811,11 +721,9 @@ int snd_akm4xxx_build_controls(struct snd_akm4xxx *ak)
 	if (err < 0)
 		return err;
 
-	if (ak->type == SND_AK4524 || ak->type == SND_AK5365) {
-		err = build_adc_controls(ak);
-		if (err < 0)
-			return err;
-	}
+	err = build_adc_controls(ak);
+	if (err < 0)
+		return err;
 
 	if (ak->type == SND_AK4355 || ak->type == SND_AK4358)
 		num_emphs = 1;
diff --git a/sound/pci/ice1712/revo.c b/sound/pci/ice1712/revo.c
index c9eefa9bbfff..bf98ea34feb0 100644
--- a/sound/pci/ice1712/revo.c
+++ b/sound/pci/ice1712/revo.c
@@ -110,7 +110,6 @@ static struct snd_akm4xxx_dac_channel revo51_dac[] = {
 static struct snd_akm4xxx_adc_channel revo51_adc[] = {
 	{
 		.name = "PCM Capture Volume",
-		.gain_name = "PCM Capture Gain Volume",
 		.switch_name = "PCM Capture Switch",
 		.num_channels = 2
 	},
-- 
cgit v1.2.3


From 4c8bd7eeee4c8f157fb61fb64b57500990b42e0e Mon Sep 17 00:00:00 2001
From: David Miller <davem@davemloft.net>
Date: Fri, 22 Sep 2006 22:31:36 -0700
Subject: [KERNEL] Do not truncate to 'int' in ALIGN() macro.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
---
 include/linux/kernel.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/linux/kernel.h b/include/linux/kernel.h
index 851aa1bcfc1a..2b2ae4fdce8b 100644
--- a/include/linux/kernel.h
+++ b/include/linux/kernel.h
@@ -31,7 +31,7 @@ extern const char linux_banner[];
 #define STACK_MAGIC	0xdeadbeef
 
 #define ARRAY_SIZE(x) (sizeof(x) / sizeof((x)[0]))
-#define ALIGN(x,a) (((x)+(a)-1)&~((a)-1))
+#define ALIGN(x,a) (((x)+(a)-1UL)&~((a)-1UL))
 #define FIELD_SIZEOF(t, f) (sizeof(((t*)0)->f))
 #define roundup(x, y) ((((x) + ((y) - 1)) / (y)) * (y))
 
-- 
cgit v1.2.3


From 2efc80cb8ddc341d81de996920e3b2ad8a12b1f7 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@ftp.linux.org.uk>
Date: Sat, 23 Sep 2006 16:45:55 +0100
Subject: [PATCH] #elif that should've been #elif defined

 #elif CONFIG_44x
in ibm4xx.h should've been
 #elif defined(CONFIG_44x)

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
---
 include/asm-ppc/ibm4xx.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include')

diff --git a/include/asm-ppc/ibm4xx.h b/include/asm-ppc/ibm4xx.h
index cf62b69cb69a..499c14691c71 100644
--- a/include/asm-ppc/ibm4xx.h
+++ b/include/asm-ppc/ibm4xx.h
@@ -86,7 +86,7 @@ void ppc4xx_init(unsigned long r3, unsigned long r4, unsigned long r5,
 #define PCI_DRAM_OFFSET	0
 #endif
 
-#elif CONFIG_44x
+#elif defined(CONFIG_44x)
 
 #if defined(CONFIG_BAMBOO)
 #include <platforms/4xx/bamboo.h>
-- 
cgit v1.2.3


From dfdc58ba354adb80d67c99f7be84f95a8e02e466 Mon Sep 17 00:00:00 2001
From: James Bottomley <James.Bottomley@steeleye.com>
Date: Wed, 20 Sep 2006 12:00:18 -0400
Subject: [SCSI] SPI transport class: misc DV fixes

Key more of the domain validation settings off the inquiry data from
the disk (in particular, don't try IU or DT unless the disk claims to
support them.

Also add a new dv_in_progress flag to prevent recursive DV.

Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
---
 drivers/scsi/scsi_transport_spi.c | 30 +++++++++++++++++++++++-------
 include/scsi/scsi_transport_spi.h |  3 ++-
 2 files changed, 25 insertions(+), 8 deletions(-)

(limited to 'include')

diff --git a/drivers/scsi/scsi_transport_spi.c b/drivers/scsi/scsi_transport_spi.c
index 29a9a53cdd1a..9f070f0d0f2b 100644
--- a/drivers/scsi/scsi_transport_spi.c
+++ b/drivers/scsi/scsi_transport_spi.c
@@ -47,6 +47,7 @@
 
 /* Private data accessors (keep these out of the header file) */
 #define spi_dv_pending(x) (((struct spi_transport_attrs *)&(x)->starget_data)->dv_pending)
+#define spi_dv_in_progress(x) (((struct spi_transport_attrs *)&(x)->starget_data)->dv_in_progress)
 #define spi_dv_mutex(x) (((struct spi_transport_attrs *)&(x)->starget_data)->dv_mutex)
 
 struct spi_internal {
@@ -240,6 +241,7 @@ static int spi_setup_transport_attrs(struct transport_container *tc,
 	spi_pcomp_en(starget) = 0;
 	spi_hold_mcs(starget) = 0;
 	spi_dv_pending(starget) = 0;
+	spi_dv_in_progress(starget) = 0;
 	spi_initial_dv(starget) = 0;
 	mutex_init(&spi_dv_mutex(starget));
 
@@ -830,28 +832,37 @@ spi_dv_device_internal(struct scsi_device *sdev, u8 *buffer)
 	DV_SET(period, spi_min_period(starget));
 	/* try QAS requests; this should be harmless to set if the
 	 * target supports it */
-	if (scsi_device_qas(sdev))
+	if (scsi_device_qas(sdev)) {
 		DV_SET(qas, 1);
-	/* Also try IU transfers */
-	if (scsi_device_ius(sdev))
+	} else {
+		DV_SET(qas, 0);
+	}
+
+	if (scsi_device_ius(sdev) && spi_min_period(starget) < 9) {
+		/* This u320 (or u640). Set IU transfers */
 		DV_SET(iu, 1);
-	if (spi_min_period(starget) < 9) {
-		/* This u320 (or u640). Ignore the coupled parameters
-		 * like DT and IU, but set the optional ones */
+		/* Then set the optional parameters */
 		DV_SET(rd_strm, 1);
 		DV_SET(wr_flow, 1);
 		DV_SET(rti, 1);
 		if (spi_min_period(starget) == 8)
 			DV_SET(pcomp_en, 1);
+	} else {
+		DV_SET(iu, 0);
 	}
+
 	/* now that we've done all this, actually check the bus
 	 * signal type (if known).  Some devices are stupid on
 	 * a SE bus and still claim they can try LVD only settings */
 	if (i->f->get_signalling)
 		i->f->get_signalling(shost);
 	if (spi_signalling(shost) == SPI_SIGNAL_SE ||
-	    spi_signalling(shost) == SPI_SIGNAL_HVD)
+	    spi_signalling(shost) == SPI_SIGNAL_HVD ||
+	    !scsi_device_dt(sdev)) {
 		DV_SET(dt, 0);
+	} else {
+		DV_SET(dt, 1);
+	}
 	/* Do the read only INQUIRY tests */
 	spi_dv_retrain(sdev, buffer, buffer + sdev->inquiry_len,
 		       spi_dv_device_compare_inquiry);
@@ -907,6 +918,10 @@ spi_dv_device(struct scsi_device *sdev)
 	if (unlikely(scsi_device_get(sdev)))
 		return;
 
+	if (unlikely(spi_dv_in_progress(starget)))
+		return;
+	spi_dv_in_progress(starget) = 1;
+
 	buffer = kzalloc(len, GFP_KERNEL);
 
 	if (unlikely(!buffer))
@@ -938,6 +953,7 @@ spi_dv_device(struct scsi_device *sdev)
  out_free:
 	kfree(buffer);
  out_put:
+	spi_dv_in_progress(starget) = 0;
 	scsi_device_put(sdev);
 }
 EXPORT_SYMBOL(spi_dv_device);
diff --git a/include/scsi/scsi_transport_spi.h b/include/scsi/scsi_transport_spi.h
index 302680c0c0de..da180f738477 100644
--- a/include/scsi/scsi_transport_spi.h
+++ b/include/scsi/scsi_transport_spi.h
@@ -53,7 +53,8 @@ struct spi_transport_attrs {
 	unsigned int support_ius; /* support Information Units */
 	unsigned int support_qas; /* supports quick arbitration and selection */
 	/* Private Fields */
-	unsigned int dv_pending:1; /* Internal flag */
+	unsigned int dv_pending:1; /* Internal flag: DV Requested */
+	unsigned int dv_in_progress:1;	/* Internal: DV started */
 	struct mutex dv_mutex; /* semaphore to serialise dv */
 };
 
-- 
cgit v1.2.3


From e18fa700c9a31360bc8f193aa543b7ef7b39a06b Mon Sep 17 00:00:00 2001
From: Jeff Garzik <jeff@garzik.org>
Date: Sun, 24 Sep 2006 11:13:19 -0400
Subject: Move several *_SUPER_MAGIC symbols to include/linux/magic.h.

Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 fs/affs/affs.h               |  1 -
 fs/affs/super.c              |  1 +
 fs/autofs/autofs_i.h         |  2 --
 fs/autofs/inode.c            |  1 +
 fs/autofs4/autofs_i.h        |  2 --
 fs/autofs4/inode.c           |  1 +
 fs/hpfs/hpfs_fn.h            |  1 -
 fs/hpfs/super.c              |  1 +
 fs/openpromfs/inode.c        |  2 +-
 include/linux/Kbuild         |  4 +---
 include/linux/adfs_fs.h      |  2 +-
 include/linux/affs_fs.h      |  7 -------
 include/linux/coda_psdev.h   |  4 ++--
 include/linux/efs_fs_sb.h    |  3 +--
 include/linux/ext2_fs.h      |  6 +-----
 include/linux/ext3_fs.h      |  6 +-----
 include/linux/hpfs_fs.h      |  8 --------
 include/linux/iso_fs.h       |  6 +++---
 include/linux/jffs2.h        |  4 ++--
 include/linux/magic.h        | 37 +++++++++++++++++++++++++++++++++++++
 include/linux/minix_fs.h     |  6 ++----
 include/linux/msdos_fs.h     |  4 ++--
 include/linux/ncp_fs.h       |  5 +----
 include/linux/nfs_fs.h       |  7 ++-----
 include/linux/openprom_fs.h  | 10 ----------
 include/linux/proc_fs.h      |  3 +--
 include/linux/qnx4_fs.h      |  2 +-
 include/linux/reiserfs_fs.h  | 10 ++--------
 include/linux/smb.h          |  3 +--
 include/linux/usbdevice_fs.h |  3 +--
 30 files changed, 67 insertions(+), 85 deletions(-)
 delete mode 100644 include/linux/affs_fs.h
 delete mode 100644 include/linux/hpfs_fs.h
 create mode 100644 include/linux/magic.h
 delete mode 100644 include/linux/openprom_fs.h

(limited to 'include')

diff --git a/fs/affs/affs.h b/fs/affs/affs.h
index 0ddd4cc0d1a0..1dc8438ef389 100644
--- a/fs/affs/affs.h
+++ b/fs/affs/affs.h
@@ -1,7 +1,6 @@
 #include <linux/types.h>
 #include <linux/fs.h>
 #include <linux/buffer_head.h>
-#include <linux/affs_fs.h>
 #include <linux/amigaffs.h>
 
 /* AmigaOS allows file names with up to 30 characters length.
diff --git a/fs/affs/super.c b/fs/affs/super.c
index 5200f4938df0..17352011ab67 100644
--- a/fs/affs/super.c
+++ b/fs/affs/super.c
@@ -14,6 +14,7 @@
 #include <linux/init.h>
 #include <linux/statfs.h>
 #include <linux/parser.h>
+#include <linux/magic.h>
 #include "affs.h"
 
 extern struct timezone sys_tz;
diff --git a/fs/autofs/autofs_i.h b/fs/autofs/autofs_i.h
index a62327f1bdff..c7700d9b3f96 100644
--- a/fs/autofs/autofs_i.h
+++ b/fs/autofs/autofs_i.h
@@ -37,8 +37,6 @@
 #define DPRINTK(D) ((void)0)
 #endif
 
-#define AUTOFS_SUPER_MAGIC 0x0187
-
 /*
  * If the daemon returns a negative response (AUTOFS_IOC_FAIL) then the
  * kernel will keep the negative response cached for up to the time given
diff --git a/fs/autofs/inode.c b/fs/autofs/inode.c
index 65e5ed42190e..af2efbbb5d76 100644
--- a/fs/autofs/inode.c
+++ b/fs/autofs/inode.c
@@ -16,6 +16,7 @@
 #include <linux/file.h>
 #include <linux/parser.h>
 #include <linux/bitops.h>
+#include <linux/magic.h>
 #include "autofs_i.h"
 #include <linux/module.h>
 
diff --git a/fs/autofs4/autofs_i.h b/fs/autofs4/autofs_i.h
index d6603d02304c..480ab178cba5 100644
--- a/fs/autofs4/autofs_i.h
+++ b/fs/autofs4/autofs_i.h
@@ -40,8 +40,6 @@
 #define DPRINTK(fmt,args...) do {} while(0)
 #endif
 
-#define AUTOFS_SUPER_MAGIC 0x0187
-
 /* Unified info structure.  This is pointed to by both the dentry and
    inode structures.  Each file in the filesystem has an instance of this
    structure.  It holds a reference to the dentry, so dentries are never
diff --git a/fs/autofs4/inode.c b/fs/autofs4/inode.c
index fde78b110ddd..11a6a9ae51b7 100644
--- a/fs/autofs4/inode.c
+++ b/fs/autofs4/inode.c
@@ -19,6 +19,7 @@
 #include <linux/parser.h>
 #include <linux/bitops.h>
 #include <linux/smp_lock.h>
+#include <linux/magic.h>
 #include "autofs_i.h"
 #include <linux/module.h>
 
diff --git a/fs/hpfs/hpfs_fn.h b/fs/hpfs/hpfs_fn.h
index f687d54ed442..32ab51e42b96 100644
--- a/fs/hpfs/hpfs_fn.h
+++ b/fs/hpfs/hpfs_fn.h
@@ -12,7 +12,6 @@
 #include <linux/mutex.h>
 #include <linux/pagemap.h>
 #include <linux/buffer_head.h>
-#include <linux/hpfs_fs.h>
 #include <linux/slab.h>
 #include <linux/smp_lock.h>
 
diff --git a/fs/hpfs/super.c b/fs/hpfs/super.c
index f798480a363f..8fe51c343786 100644
--- a/fs/hpfs/super.c
+++ b/fs/hpfs/super.c
@@ -11,6 +11,7 @@
 #include <linux/parser.h>
 #include <linux/init.h>
 #include <linux/statfs.h>
+#include <linux/magic.h>
 
 /* Mark the filesystem dirty, so that chkdsk checks it when os/2 booted */
 
diff --git a/fs/openpromfs/inode.c b/fs/openpromfs/inode.c
index 93a56bd4a2b7..592a6402e851 100644
--- a/fs/openpromfs/inode.c
+++ b/fs/openpromfs/inode.c
@@ -8,10 +8,10 @@
 #include <linux/types.h>
 #include <linux/string.h>
 #include <linux/fs.h>
-#include <linux/openprom_fs.h>
 #include <linux/init.h>
 #include <linux/slab.h>
 #include <linux/seq_file.h>
+#include <linux/magic.h>
 
 #include <asm/openprom.h>
 #include <asm/oplib.h>
diff --git a/include/linux/Kbuild b/include/linux/Kbuild
index 7d076d97b2f7..67383605f2e5 100644
--- a/include/linux/Kbuild
+++ b/include/linux/Kbuild
@@ -12,7 +12,6 @@ header-y += netfilter_bridge/
 header-y += netfilter_ipv4/
 header-y += netfilter_ipv6/
 
-header-y += affs_fs.h
 header-y += affs_hardblocks.h
 header-y += aio_abi.h
 header-y += a.out.h
@@ -67,7 +66,6 @@ header-y += genetlink.h
 header-y += gen_stats.h
 header-y += gigaset_dev.h
 header-y += hdsmart.h
-header-y += hpfs_fs.h
 header-y += hysdn_if.h
 header-y += i2c-dev.h
 header-y += i8k.h
@@ -103,6 +101,7 @@ header-y += ixjuser.h
 header-y += jffs2.h
 header-y += keyctl.h
 header-y += limits.h
+header-y += magic.h
 header-y += major.h
 header-y += matroxfb.h
 header-y += meye.h
@@ -116,7 +115,6 @@ header-y += netrom.h
 header-y += nfs2.h
 header-y += nfs4_mount.h
 header-y += nfs_mount.h
-header-y += openprom_fs.h
 header-y += param.h
 header-y += pci_ids.h
 header-y += pci_regs.h
diff --git a/include/linux/adfs_fs.h b/include/linux/adfs_fs.h
index 4a5d50c2bdbf..ef788c2085a1 100644
--- a/include/linux/adfs_fs.h
+++ b/include/linux/adfs_fs.h
@@ -2,6 +2,7 @@
 #define _ADFS_FS_H
 
 #include <linux/types.h>
+#include <linux/magic.h>
 
 /*
  * Disc Record at disc address 0xc00
@@ -38,7 +39,6 @@ struct adfs_discrecord {
 #define ADFS_DR_OFFSET		(0x1c0)
 #define ADFS_DR_SIZE		 60
 #define ADFS_DR_SIZE_BITS	(ADFS_DR_SIZE << 3)
-#define ADFS_SUPER_MAGIC	 0xadf5
 
 #ifdef __KERNEL__
 #include <linux/adfs_fs_i.h>
diff --git a/include/linux/affs_fs.h b/include/linux/affs_fs.h
deleted file mode 100644
index c57b5ee87d55..000000000000
--- a/include/linux/affs_fs.h
+++ /dev/null
@@ -1,7 +0,0 @@
-#ifndef _AFFS_FS_H
-#define _AFFS_FS_H
-/*
- * The affs filesystem constants/structures
- */
-#define AFFS_SUPER_MAGIC 0xadff
-#endif
diff --git a/include/linux/coda_psdev.h b/include/linux/coda_psdev.h
index 98f6c52c152b..b541bb3d1f4b 100644
--- a/include/linux/coda_psdev.h
+++ b/include/linux/coda_psdev.h
@@ -1,11 +1,11 @@
 #ifndef __CODA_PSDEV_H
 #define __CODA_PSDEV_H
 
+#include <linux/magic.h>
+
 #define CODA_PSDEV_MAJOR 67
 #define MAX_CODADEVS  5	   /* how many do we allow */
 
-#define CODA_SUPER_MAGIC	0x73757245
-
 struct kstatfs;
 
 struct coda_sb_info
diff --git a/include/linux/efs_fs_sb.h b/include/linux/efs_fs_sb.h
index c76088baef28..ff1945e37790 100644
--- a/include/linux/efs_fs_sb.h
+++ b/include/linux/efs_fs_sb.h
@@ -9,8 +9,7 @@
 #ifndef __EFS_FS_SB_H__
 #define __EFS_FS_SB_H__
 
-/* statfs() magic number for EFS */
-#define EFS_SUPER_MAGIC	0x414A53
+#include <linux/magic.h>
 
 /* EFS superblock magic numbers */
 #define EFS_MAGIC	0x072959
diff --git a/include/linux/ext2_fs.h b/include/linux/ext2_fs.h
index facf34e98954..33a1aa107329 100644
--- a/include/linux/ext2_fs.h
+++ b/include/linux/ext2_fs.h
@@ -17,6 +17,7 @@
 #define _LINUX_EXT2_FS_H
 
 #include <linux/types.h>
+#include <linux/magic.h>
 
 /*
  * The second extended filesystem constants/structures
@@ -63,11 +64,6 @@
 /* First non-reserved inode for old ext2 filesystems */
 #define EXT2_GOOD_OLD_FIRST_INO	11
 
-/*
- * The second extended file system magic number
- */
-#define EXT2_SUPER_MAGIC	0xEF53
-
 #ifdef __KERNEL__
 #include <linux/ext2_fs_sb.h>
 static inline struct ext2_sb_info *EXT2_SB(struct super_block *sb)
diff --git a/include/linux/ext3_fs.h b/include/linux/ext3_fs.h
index 9f9cce7bd86d..0eed918b3816 100644
--- a/include/linux/ext3_fs.h
+++ b/include/linux/ext3_fs.h
@@ -17,6 +17,7 @@
 #define _LINUX_EXT3_FS_H
 
 #include <linux/types.h>
+#include <linux/magic.h>
 
 /*
  * The second extended filesystem constants/structures
@@ -66,11 +67,6 @@
 /* First non-reserved inode for old ext3 filesystems */
 #define EXT3_GOOD_OLD_FIRST_INO	11
 
-/*
- * The second extended file system magic number
- */
-#define EXT3_SUPER_MAGIC	0xEF53
-
 /*
  * Maximal count of links to a file
  */
diff --git a/include/linux/hpfs_fs.h b/include/linux/hpfs_fs.h
deleted file mode 100644
index a5028dd94d31..000000000000
--- a/include/linux/hpfs_fs.h
+++ /dev/null
@@ -1,8 +0,0 @@
-#ifndef _LINUX_HPFS_FS_H
-#define _LINUX_HPFS_FS_H
-
-/* HPFS magic number (word 0 of block 16) */
-
-#define HPFS_SUPER_MAGIC 0xf995e849
-
-#endif
diff --git a/include/linux/iso_fs.h b/include/linux/iso_fs.h
index 47967878bfef..4688ac4284e2 100644
--- a/include/linux/iso_fs.h
+++ b/include/linux/iso_fs.h
@@ -2,6 +2,8 @@
 #define _ISOFS_FS_H
 
 #include <linux/types.h>
+#include <linux/magic.h>
+
 /*
  * The isofs filesystem constants/structures
  */
@@ -160,6 +162,4 @@ struct iso_directory_record {
 #define ISOFS_BUFFER_SIZE(INODE) ((INODE)->i_sb->s_blocksize)
 #define ISOFS_BUFFER_BITS(INODE) ((INODE)->i_sb->s_blocksize_bits)
 
-#define ISOFS_SUPER_MAGIC 0x9660
-
-#endif
+#endif /* _ISOFS_FS_H */
diff --git a/include/linux/jffs2.h b/include/linux/jffs2.h
index c9c760700bc3..840631fa5ff1 100644
--- a/include/linux/jffs2.h
+++ b/include/linux/jffs2.h
@@ -15,12 +15,12 @@
 #ifndef __LINUX_JFFS2_H__
 #define __LINUX_JFFS2_H__
 
+#include <linux/magic.h>
+
 /* You must include something which defines the C99 uintXX_t types. 
    We don't do it from here because this file is used in too many
    different environments. */
 
-#define JFFS2_SUPER_MAGIC 0x72b6
-
 /* Values we may expect to find in the 'magic' field */
 #define JFFS2_OLD_MAGIC_BITMASK 0x1984
 #define JFFS2_MAGIC_BITMASK 0x1985
diff --git a/include/linux/magic.h b/include/linux/magic.h
new file mode 100644
index 000000000000..22036dd2ba36
--- /dev/null
+++ b/include/linux/magic.h
@@ -0,0 +1,37 @@
+#ifndef __LINUX_MAGIC_H__
+#define __LINUX_MAGIC_H__
+
+#define ADFS_SUPER_MAGIC	0xadf5
+#define AFFS_SUPER_MAGIC	0xadff
+#define AUTOFS_SUPER_MAGIC	0x0187
+#define CODA_SUPER_MAGIC	0x73757245
+#define EFS_SUPER_MAGIC		0x414A53
+#define EXT2_SUPER_MAGIC	0xEF53
+#define EXT3_SUPER_MAGIC	0xEF53
+#define HPFS_SUPER_MAGIC	0xf995e849
+#define ISOFS_SUPER_MAGIC	0x9660
+#define JFFS2_SUPER_MAGIC	0x72b6
+
+#define MINIX_SUPER_MAGIC	0x137F		/* original minix fs */
+#define MINIX_SUPER_MAGIC2	0x138F		/* minix fs, 30 char names */
+#define MINIX2_SUPER_MAGIC	0x2468		/* minix V2 fs */
+#define MINIX2_SUPER_MAGIC2	0x2478		/* minix V2 fs, 30 char names */
+
+#define MSDOS_SUPER_MAGIC	0x4d44		/* MD */
+#define NCP_SUPER_MAGIC		0x564c		/* Guess, what 0x564c is :-) */
+#define NFS_SUPER_MAGIC		0x6969
+#define OPENPROM_SUPER_MAGIC	0x9fa1
+#define PROC_SUPER_MAGIC	0x9fa0
+#define QNX4_SUPER_MAGIC	0x002f		/* qnx4 fs detection */
+
+#define REISERFS_SUPER_MAGIC	0x52654973	/* used by gcc */
+					/* used by file system utilities that
+	                                   look at the superblock, etc.  */
+#define REISERFS_SUPER_MAGIC_STRING	"ReIsErFs"
+#define REISER2FS_SUPER_MAGIC_STRING	"ReIsEr2Fs"
+#define REISER2FS_JR_SUPER_MAGIC_STRING	"ReIsEr3Fs"
+
+#define SMB_SUPER_MAGIC		0x517B
+#define USBDEVICE_SUPER_MAGIC	0x9fa2
+
+#endif /* __LINUX_MAGIC_H__ */
diff --git a/include/linux/minix_fs.h b/include/linux/minix_fs.h
index 1ecc3cc8cef5..916e8f72c63d 100644
--- a/include/linux/minix_fs.h
+++ b/include/linux/minix_fs.h
@@ -1,6 +1,8 @@
 #ifndef _LINUX_MINIX_FS_H
 #define _LINUX_MINIX_FS_H
 
+#include <linux/magic.h>
+
 /*
  * The minix filesystem constants/structures
  */
@@ -19,10 +21,6 @@
 
 #define MINIX_I_MAP_SLOTS	8
 #define MINIX_Z_MAP_SLOTS	64
-#define MINIX_SUPER_MAGIC	0x137F		/* original minix fs */
-#define MINIX_SUPER_MAGIC2	0x138F		/* minix fs, 30 char names */
-#define MINIX2_SUPER_MAGIC	0x2468		/* minix V2 fs */
-#define MINIX2_SUPER_MAGIC2	0x2478		/* minix V2 fs, 30 char names */
 #define MINIX_VALID_FS		0x0001		/* Clean fs. */
 #define MINIX_ERROR_FS		0x0002		/* fs has errors. */
 
diff --git a/include/linux/msdos_fs.h b/include/linux/msdos_fs.h
index d9035c73e5d1..bae62d62dc3e 100644
--- a/include/linux/msdos_fs.h
+++ b/include/linux/msdos_fs.h
@@ -1,6 +1,8 @@
 #ifndef _LINUX_MSDOS_FS_H
 #define _LINUX_MSDOS_FS_H
 
+#include <linux/magic.h>
+
 /*
  * The MS-DOS filesystem constants/structures
  */
@@ -18,8 +20,6 @@
 #define CT_LE_L(v)	cpu_to_le32(v)
 
 
-#define MSDOS_SUPER_MAGIC 0x4d44 /* MD */
-
 #define MSDOS_ROOT_INO	1	/* == MINIX_ROOT_INO */
 #define MSDOS_DIR_BITS	5	/* log2(sizeof(struct msdos_dir_entry)) */
 
diff --git a/include/linux/ncp_fs.h b/include/linux/ncp_fs.h
index b208f0cd556b..02e352be717e 100644
--- a/include/linux/ncp_fs.h
+++ b/include/linux/ncp_fs.h
@@ -11,6 +11,7 @@
 #include <linux/fs.h>
 #include <linux/in.h>
 #include <linux/types.h>
+#include <linux/magic.h>
 
 #include <linux/ipx.h>
 #include <linux/ncp_no.h>
@@ -185,10 +186,6 @@ struct ncp_entry_info {
 	__u8			file_handle[6];
 };
 
-/* Guess, what 0x564c is :-) */
-#define NCP_SUPER_MAGIC  0x564c
-
-
 static inline struct ncp_server *NCP_SBP(struct super_block *sb)
 {
 	return sb->s_fs_info;
diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h
index 3b5b04193fee..36f5bcf513b0 100644
--- a/include/linux/nfs_fs.h
+++ b/include/linux/nfs_fs.h
@@ -9,6 +9,8 @@
 #ifndef _LINUX_NFS_FS_H
 #define _LINUX_NFS_FS_H
 
+#include <linux/magic.h>
+
 /*
  * Enable debugging support for nfs client.
  * Requires RPC_DEBUG.
@@ -21,11 +23,6 @@
 #define NFS_MAX_UDP_TIMEOUT	(60*HZ)
 #define NFS_MAX_TCP_TIMEOUT	(600*HZ)
 
-/*
- * superblock magic number for NFS
- */
-#define NFS_SUPER_MAGIC			0x6969
-
 /*
  * When flushing a cluster of dirty pages, there can be different
  * strategies:
diff --git a/include/linux/openprom_fs.h b/include/linux/openprom_fs.h
deleted file mode 100644
index a837aab8217e..000000000000
--- a/include/linux/openprom_fs.h
+++ /dev/null
@@ -1,10 +0,0 @@
-#ifndef _LINUX_OPENPROM_FS_H
-#define _LINUX_OPENPROM_FS_H
-
-/*
- * The openprom filesystem constants/structures
- */
-
-#define OPENPROM_SUPER_MAGIC 0x9fa1
-
-#endif /* _LINUX_OPENPROM_FS_H */
diff --git a/include/linux/proc_fs.h b/include/linux/proc_fs.h
index 17e75783e3a5..3435ca38dd14 100644
--- a/include/linux/proc_fs.h
+++ b/include/linux/proc_fs.h
@@ -4,6 +4,7 @@
 #include <linux/slab.h>
 #include <linux/fs.h>
 #include <linux/spinlock.h>
+#include <linux/magic.h>
 #include <asm/atomic.h>
 
 /*
@@ -24,8 +25,6 @@ enum {
 	PROC_ROOT_INO = 1,
 };
 
-#define PROC_SUPER_MAGIC 0x9fa0
-
 /*
  * This is not completely implemented yet. The idea is to
  * create an in-memory tree (like the actual /proc filesystem
diff --git a/include/linux/qnx4_fs.h b/include/linux/qnx4_fs.h
index 27f49c85d5d6..0c7ac444fd35 100644
--- a/include/linux/qnx4_fs.h
+++ b/include/linux/qnx4_fs.h
@@ -11,6 +11,7 @@
 #define _LINUX_QNX4_FS_H
 
 #include <linux/qnxtypes.h>
+#include <linux/magic.h>
 
 #define QNX4_ROOT_INO 1
 
@@ -25,7 +26,6 @@
 
 #define QNX4_I_MAP_SLOTS	8
 #define QNX4_Z_MAP_SLOTS	64
-#define QNX4_SUPER_MAGIC	0x002f	/* qnx4 fs detection */
 #define QNX4_VALID_FS		0x0001	/* Clean fs. */
 #define QNX4_ERROR_FS		0x0002	/* fs has errors. */
 #define QNX4_BLOCK_SIZE         0x200	/* blocksize of 512 bytes */
diff --git a/include/linux/reiserfs_fs.h b/include/linux/reiserfs_fs.h
index daa2d83cefe8..28493ffaafe7 100644
--- a/include/linux/reiserfs_fs.h
+++ b/include/linux/reiserfs_fs.h
@@ -12,6 +12,8 @@
 #define _LINUX_REISER_FS_H
 
 #include <linux/types.h>
+#include <linux/magic.h>
+
 #ifdef __KERNEL__
 #include <linux/slab.h>
 #include <linux/interrupt.h>
@@ -227,14 +229,6 @@ struct reiserfs_super_block {
          ((!is_reiserfs_jr(SB_DISK_SUPER_BLOCK(s)) ? \
          SB_ONDISK_JOURNAL_SIZE(s) + 1 : SB_ONDISK_RESERVED_FOR_JOURNAL(s)))
 
-				/* used by gcc */
-#define REISERFS_SUPER_MAGIC 0x52654973
-				/* used by file system utilities that
-				   look at the superblock, etc. */
-#define REISERFS_SUPER_MAGIC_STRING "ReIsErFs"
-#define REISER2FS_SUPER_MAGIC_STRING "ReIsEr2Fs"
-#define REISER2FS_JR_SUPER_MAGIC_STRING "ReIsEr3Fs"
-
 int is_reiserfs_3_5(struct reiserfs_super_block *rs);
 int is_reiserfs_3_6(struct reiserfs_super_block *rs);
 int is_reiserfs_jr(struct reiserfs_super_block *rs);
diff --git a/include/linux/smb.h b/include/linux/smb.h
index b0162208c963..6df3b1501559 100644
--- a/include/linux/smb.h
+++ b/include/linux/smb.h
@@ -10,6 +10,7 @@
 #define _LINUX_SMB_H
 
 #include <linux/types.h>
+#include <linux/magic.h>
 
 enum smb_protocol { 
 	SMB_PROTOCOL_NONE, 
@@ -101,8 +102,6 @@ enum smb_conn_state {
 	CONN_RETRYING		/* Currently trying to reconnect */
 };
 
-#define SMB_SUPER_MAGIC               0x517B
-
 #define SMB_HEADER_LEN   37     /* includes everything up to, but not
                                  * including smb_bcc */
 
diff --git a/include/linux/usbdevice_fs.h b/include/linux/usbdevice_fs.h
index 7b7aadb69092..617d8a1c59ae 100644
--- a/include/linux/usbdevice_fs.h
+++ b/include/linux/usbdevice_fs.h
@@ -32,11 +32,10 @@
 #define _LINUX_USBDEVICE_FS_H
 
 #include <linux/types.h>
+#include <linux/magic.h>
 
 /* --------------------------------------------------------------------- */
 
-#define USBDEVICE_SUPER_MAGIC 0x9fa2
-
 /* usbdevfs ioctl codes */
 
 struct usbdevfs_ctrltransfer {
-- 
cgit v1.2.3