Merge tag 'v3.16-rc7' into perf/core, to merge in the latest fixes before applying new changes

Signed-off-by: Ingo Molnar <mingo@kernel.org>
author: Ingo Molnar <mingo@kernel.org> 2014-07-28 10:00:33 +0200
committer: Ingo Molnar <mingo@kernel.org> 2014-07-28 10:00:33 +0200
commit: 5030c69755416d19516c0a61cd988a0e0062e041 (patch)
tree: ca4d3088428cf033420f9df9111d796650bc59e5 /arch
parent: 2336ebc32676df5b794acfe0c980583ec6c05f34 (diff)
parent: 64aa90f26c06e1cb2aacfb98a7d0eccfbd6c1a91 (diff)
download: blackbird-op-linux-5030c69755416d19516c0a61cd988a0e0062e041.tar.gz
blackbird-op-linux-5030c69755416d19516c0a61cd988a0e0062e041.zip
258 files changed, 1980 insertions, 1805 deletions
diff --git a/arch/arc/include/asm/cache.h b/arch/arc/include/asm/cache.h
index c1d3d2da1191..b3c750979aa1 100644
--- a/arch/arc/include/asm/cache.h
+++ b/arch/arc/include/asm/cache.h
@@ -60,7 +60,7 @@ extern void read_decode_cache_bcr(void);
 #define ARC_REG_IC_IVIC		0x10
 #define ARC_REG_IC_CTRL		0x11
 #define ARC_REG_IC_IVIL		0x19
-#if defined(CONFIG_ARC_MMU_V3) || defined (CONFIG_ARC_MMU_V4)
+#if defined(CONFIG_ARC_MMU_V3)
 #define ARC_REG_IC_PTAG		0x1E
 #endif
 
@@ -74,7 +74,7 @@ extern void read_decode_cache_bcr(void);
 #define ARC_REG_DC_IVDL		0x4A
 #define ARC_REG_DC_FLSH		0x4B
 #define ARC_REG_DC_FLDL		0x4C
-#if defined(CONFIG_ARC_MMU_V3) || defined (CONFIG_ARC_MMU_V4)
+#if defined(CONFIG_ARC_MMU_V3)
 #define ARC_REG_DC_PTAG		0x5C
 #endif
 
diff --git a/arch/arc/include/uapi/asm/ptrace.h b/arch/arc/include/uapi/asm/ptrace.h
index 2618cc13ba75..76a7739aab1c 100644
--- a/arch/arc/include/uapi/asm/ptrace.h
+++ b/arch/arc/include/uapi/asm/ptrace.h
@@ -11,6 +11,7 @@
 #ifndef _UAPI__ASM_ARC_PTRACE_H
 #define _UAPI__ASM_ARC_PTRACE_H
 
+#define PTRACE_GET_THREAD_AREA	25
 
 #ifndef __ASSEMBLY__
 /*
diff --git a/arch/arc/kernel/ctx_sw_asm.S b/arch/arc/kernel/ctx_sw_asm.S
index 2ff0347a2fd7..e248594097e7 100644
--- a/arch/arc/kernel/ctx_sw_asm.S
+++ b/arch/arc/kernel/ctx_sw_asm.S
@@ -10,9 +10,9 @@
  *  -This is the more "natural" hand written assembler
  */
 
+#include <linux/linkage.h>
 #include <asm/entry.h>       /* For the SAVE_* macros */
 #include <asm/asm-offsets.h>
-#include <asm/linkage.h>
 
 #define KSP_WORD_OFF 	((TASK_THREAD + THREAD_KSP) / 4)
 
diff --git a/arch/arc/kernel/devtree.c b/arch/arc/kernel/devtree.c
index 0b3ef4025d89..fffdb5e41b20 100644
--- a/arch/arc/kernel/devtree.c
+++ b/arch/arc/kernel/devtree.c
@@ -41,7 +41,7 @@ const struct machine_desc * __init setup_machine_fdt(void *dt)
 {
 	const struct machine_desc *mdesc;
 	unsigned long dt_root;
-	void *clk;
+	const void *clk;
 	int len;
 
 	if (!early_init_dt_scan(dt))
diff --git a/arch/arc/kernel/head.S b/arch/arc/kernel/head.S
index 07a58f2d3077..4d2481bd8b98 100644
--- a/arch/arc/kernel/head.S
+++ b/arch/arc/kernel/head.S
@@ -77,10 +77,11 @@ stext:
 	; Clear BSS before updating any globals
 	; XXX: use ZOL here
 	mov	r5, __bss_start
-	mov	r6, __bss_stop
+	sub	r6, __bss_stop, r5
+	lsr.f	lp_count, r6, 2
+	lpnz	1f
+	st.ab   0, [r5, 4]
 1:
-	st.ab   0, [r5,4]
-	brlt    r5, r6, 1b
 
 	; Uboot - kernel ABI
 	;    r0 = [0] No uboot interaction, [1] cmdline in r2, [2] DTB in r2
diff --git a/arch/arc/kernel/ptrace.c b/arch/arc/kernel/ptrace.c
index 5d76706139dd..13b3ffb27a38 100644
--- a/arch/arc/kernel/ptrace.c
+++ b/arch/arc/kernel/ptrace.c
@@ -146,6 +146,10 @@ long arch_ptrace(struct task_struct *child, long request,
 	pr_debug("REQ=%ld: ADDR =0x%lx, DATA=0x%lx)\n", request, addr, data);
 
 	switch (request) {
+	case PTRACE_GET_THREAD_AREA:
+		ret = put_user(task_thread_info(child)->thr_ptr,
+			       (unsigned long __user *)data);
+		break;
 	default:
 		ret = ptrace_request(child, request, addr, data);
 		break;
diff --git a/arch/arc/kernel/smp.c b/arch/arc/kernel/smp.c
index cf90b6f4d3e0..c802bb500602 100644
--- a/arch/arc/kernel/smp.c
+++ b/arch/arc/kernel/smp.c
@@ -337,8 +337,19 @@ irqreturn_t do_IPI(int irq, void *dev_id)
  * API called by platform code to hookup arch-common ISR to their IPI IRQ
  */
 static DEFINE_PER_CPU(int, ipi_dev);
+
+static struct irqaction arc_ipi_irq = {
+        .name    = "IPI Interrupt",
+        .flags   = IRQF_PERCPU,
+        .handler = do_IPI,
+};
+
 int smp_ipi_irq_setup(int cpu, int irq)
 {
-	int *dev_id = &per_cpu(ipi_dev, smp_processor_id());
-	return request_percpu_irq(irq, do_IPI, "IPI Interrupt", dev_id);
+	if (!cpu)
+		return setup_irq(irq, &arc_ipi_irq);
+	else
+		arch_unmask_irq(irq);
+
+	return 0;
 }
diff --git a/arch/arc/kernel/vmlinux.lds.S b/arch/arc/kernel/vmlinux.lds.S
index 2555f5886af6..dd35bde39f69 100644
--- a/arch/arc/kernel/vmlinux.lds.S
+++ b/arch/arc/kernel/vmlinux.lds.S
@@ -116,7 +116,7 @@ SECTIONS
 
 	_edata = .;
 
-	BSS_SECTION(0, 0, 0)
+	BSS_SECTION(4, 4, 4)
 
 #ifdef CONFIG_ARC_DW2_UNWIND
 	. = ALIGN(PAGE_SIZE);
diff --git a/arch/arc/mm/cache_arc700.c b/arch/arc/mm/cache_arc700.c
index 1f676c4794e0..353b202c37c9 100644
--- a/arch/arc/mm/cache_arc700.c
+++ b/arch/arc/mm/cache_arc700.c
@@ -389,7 +389,7 @@ static inline void __dc_line_op(unsigned long paddr, unsigned long vaddr,
 /***********************************************************
  * Machine specific helper for per line I-Cache invalidate.
  */
-static void __ic_line_inv_vaddr(unsigned long paddr, unsigned long vaddr,
+static void __ic_line_inv_vaddr_local(unsigned long paddr, unsigned long vaddr,
 				unsigned long sz)
 {
 	unsigned long flags;
@@ -405,6 +405,23 @@ static inline void __ic_entire_inv(void)
 	read_aux_reg(ARC_REG_IC_CTRL);	/* blocks */
 }
 
+struct ic_line_inv_vaddr_ipi {
+	unsigned long paddr, vaddr;
+	int sz;
+};
+
+static void __ic_line_inv_vaddr_helper(void *info)
+{
+        struct ic_line_inv_vaddr_ipi *ic_inv = (struct ic_line_inv_vaddr_ipi*) info;
+        __ic_line_inv_vaddr_local(ic_inv->paddr, ic_inv->vaddr, ic_inv->sz);
+}
+
+static void __ic_line_inv_vaddr(unsigned long paddr, unsigned long vaddr,
+				unsigned long sz)
+{
+	struct ic_line_inv_vaddr_ipi ic_inv = { paddr, vaddr , sz};
+	on_each_cpu(__ic_line_inv_vaddr_helper, &ic_inv, 1);
+}
 #else
 
 #define __ic_entire_inv()
@@ -553,12 +570,8 @@ void flush_icache_range(unsigned long kstart, unsigned long kend)
  */
 void __sync_icache_dcache(unsigned long paddr, unsigned long vaddr, int len)
 {
-	unsigned long flags;
-
-	local_irq_save(flags);
-	__ic_line_inv_vaddr(paddr, vaddr, len);
 	__dc_line_op(paddr, vaddr, len, OP_FLUSH_N_INV);
-	local_irq_restore(flags);
+	__ic_line_inv_vaddr(paddr, vaddr, len);
 }
 
 /* wrapper to compile time eliminate alignment checks in flush loop */
diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index 245058b3b0ef..88acf8bc1490 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -6,6 +6,7 @@ config ARM
 	select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
 	select ARCH_HAVE_CUSTOM_GPIO_H
 	select ARCH_MIGHT_HAVE_PC_PARPORT
+	select ARCH_SUPPORTS_ATOMIC_RMW
 	select ARCH_USE_BUILTIN_BSWAP
 	select ARCH_USE_CMPXCHG_LOCKREF
 	select ARCH_WANT_IPC_PARSE_VERSION
diff --git a/arch/arm/boot/dts/Makefile b/arch/arm/boot/dts/Makefile
index 5986ff63b901..adb5ed9e269e 100644
--- a/arch/arm/boot/dts/Makefile
+++ b/arch/arm/boot/dts/Makefile
@@ -357,7 +357,7 @@ dtb-$(CONFIG_ARCH_STI)+= stih407-b2120.dtb \
 	stih415-b2020.dtb \
 	stih416-b2000.dtb \
 	stih416-b2020.dtb \
-	stih416-b2020-revE.dtb
+	stih416-b2020e.dtb
 dtb-$(CONFIG_MACH_SUN4I) += \
 	sun4i-a10-a1000.dtb \
 	sun4i-a10-cubieboard.dtb \
diff --git a/arch/arm/boot/dts/am335x-evm.dts b/arch/arm/boot/dts/am335x-evm.dts
index ecb267767cf5..e2156a583de7 100644
--- a/arch/arm/boot/dts/am335x-evm.dts
+++ b/arch/arm/boot/dts/am335x-evm.dts
@@ -529,8 +529,8 @@
 		serial-dir = <  /* 0: INACTIVE, 1: TX, 2: RX */
 			0 0 1 2
 		>;
-		tx-num-evt = <1>;
-		rx-num-evt = <1>;
+		tx-num-evt = <32>;
+		rx-num-evt = <32>;
 };
 
 &tps {
diff --git a/arch/arm/boot/dts/am335x-evmsk.dts b/arch/arm/boot/dts/am335x-evmsk.dts
index ab9a34ce524c..80a3b215e7d6 100644
--- a/arch/arm/boot/dts/am335x-evmsk.dts
+++ b/arch/arm/boot/dts/am335x-evmsk.dts
@@ -560,8 +560,8 @@
 		serial-dir = <  /* 0: INACTIVE, 1: TX, 2: RX */
 			0 0 1 2
 		>;
-		tx-num-evt = <1>;
-		rx-num-evt = <1>;
+		tx-num-evt = <32>;
+		rx-num-evt = <32>;
 };
 
 &tscadc {
diff --git a/arch/arm/boot/dts/am335x-igep0033.dtsi b/arch/arm/boot/dts/am335x-igep0033.dtsi
index 8a0a72dc7dd7..a1a0cc5eb35c 100644
--- a/arch/arm/boot/dts/am335x-igep0033.dtsi
+++ b/arch/arm/boot/dts/am335x-igep0033.dtsi
@@ -105,10 +105,16 @@
 
 &cpsw_emac0 {
 	phy_id = <&davinci_mdio>, <0>;
+	phy-mode = "rmii";
 };
 
 &cpsw_emac1 {
 	phy_id = <&davinci_mdio>, <1>;
+	phy-mode = "rmii";
+};
+
+&phy_sel {
+	rmii-clock-ext;
 };
 
 &elm {
diff --git a/arch/arm/boot/dts/am43x-epos-evm.dts b/arch/arm/boot/dts/am43x-epos-evm.dts
index 19f1f7e87597..90098f98a5c8 100644
--- a/arch/arm/boot/dts/am43x-epos-evm.dts
+++ b/arch/arm/boot/dts/am43x-epos-evm.dts
@@ -319,6 +319,10 @@
 	phy-mode = "rmii";
 };
 
+&phy_sel {
+	rmii-clock-ext;
+};
+
 &i2c0 {
 	status = "okay";
 	pinctrl-names = "default";
diff --git a/arch/arm/boot/dts/armada-380.dtsi b/arch/arm/boot/dts/armada-380.dtsi
index e69bc6759c39..4173a8ab34e7 100644
--- a/arch/arm/boot/dts/armada-380.dtsi
+++ b/arch/arm/boot/dts/armada-380.dtsi
@@ -16,7 +16,7 @@
 
 / {
 	model = "Marvell Armada 380 family SoC";
-	compatible = "marvell,armada380", "marvell,armada38x";
+	compatible = "marvell,armada380";
 
 	cpus {
 		#address-cells = <1>;
diff --git a/arch/arm/boot/dts/armada-385-db.dts b/arch/arm/boot/dts/armada-385-db.dts
index 5bae4731828b..1af886f1e486 100644
--- a/arch/arm/boot/dts/armada-385-db.dts
+++ b/arch/arm/boot/dts/armada-385-db.dts
@@ -16,7 +16,7 @@
 
 / {
 	model = "Marvell Armada 385 Development Board";
-	compatible = "marvell,a385-db", "marvell,armada385", "marvell,armada38x";
+	compatible = "marvell,a385-db", "marvell,armada385", "marvell,armada380";
 
 	chosen {
 		bootargs = "console=ttyS0,115200 earlyprintk";
diff --git a/arch/arm/boot/dts/armada-385-rd.dts b/arch/arm/boot/dts/armada-385-rd.dts
index 40893255a3f0..aaca2861dc87 100644
--- a/arch/arm/boot/dts/armada-385-rd.dts
+++ b/arch/arm/boot/dts/armada-385-rd.dts
@@ -17,7 +17,7 @@
 
 / {
 	model = "Marvell Armada 385 Reference Design";
-	compatible = "marvell,a385-rd", "marvell,armada385", "marvell,armada38x";
+	compatible = "marvell,a385-rd", "marvell,armada385", "marvell,armada380";
 
 	chosen {
 		bootargs = "console=ttyS0,115200 earlyprintk";
diff --git a/arch/arm/boot/dts/armada-385.dtsi b/arch/arm/boot/dts/armada-385.dtsi
index f011009bf4cf..6283d7912f71 100644
--- a/arch/arm/boot/dts/armada-385.dtsi
+++ b/arch/arm/boot/dts/armada-385.dtsi
@@ -16,7 +16,7 @@
 
 / {
 	model = "Marvell Armada 385 family SoC";
-	compatible = "marvell,armada385", "marvell,armada38x";
+	compatible = "marvell,armada385", "marvell,armada380";
 
 	cpus {
 		#address-cells = <1>;
diff --git a/arch/arm/boot/dts/armada-38x.dtsi b/arch/arm/boot/dts/armada-38x.dtsi
index 3de364e81b52..689fa1a46728 100644
--- a/arch/arm/boot/dts/armada-38x.dtsi
+++ b/arch/arm/boot/dts/armada-38x.dtsi
@@ -20,7 +20,7 @@
 
 / {
 	model = "Marvell Armada 38x family SoC";
-	compatible = "marvell,armada38x";
+	compatible = "marvell,armada380";
 
 	aliases {
 		gpio0 = &gpio0;
diff --git a/arch/arm/boot/dts/at91sam9261.dtsi b/arch/arm/boot/dts/at91sam9261.dtsi
index b309c1c6e848..04927db1d6bf 100644
--- a/arch/arm/boot/dts/at91sam9261.dtsi
+++ b/arch/arm/boot/dts/at91sam9261.dtsi
@@ -568,24 +568,17 @@
 				#size-cells = <0>;
 				#interrupt-cells = <1>;
 
-				slow_rc_osc: slow_rc_osc {
-					compatible = "fixed-clock";
+				main_osc: main_osc {
+					compatible = "atmel,at91rm9200-clk-main-osc";
 					#clock-cells = <0>;
-					clock-frequency = <32768>;
-					clock-accuracy = <50000000>;
-				};
-
-				clk32k: slck {
-					compatible = "atmel,at91sam9260-clk-slow";
-					#clock-cells = <0>;
-					clocks = <&slow_rc_osc &slow_xtal>;
+					interrupts-extended = <&pmc AT91_PMC_MOSCS>;
+					clocks = <&main_xtal>;
 				};
 
 				main: mainck {
 					compatible = "atmel,at91rm9200-clk-main";
 					#clock-cells = <0>;
-					interrupts-extended = <&pmc AT91_PMC_MOSCS>;
-					clocks = <&main_xtal>;
+					clocks = <&main_osc>;
 				};
 
 				plla: pllack {
@@ -615,7 +608,7 @@
 					compatible = "atmel,at91rm9200-clk-master";
 					#clock-cells = <0>;
 					interrupts-extended = <&pmc AT91_PMC_MCKRDY>;
-					clocks = <&clk32k>, <&main>, <&plla>, <&pllb>;
+					clocks = <&slow_xtal>, <&main>, <&plla>, <&pllb>;
 					atmel,clk-output-range = <0 94000000>;
 					atmel,clk-divisors = <1 2 4 0>;
 				};
@@ -632,7 +625,7 @@
 					#address-cells = <1>;
 					#size-cells = <0>;
 					interrupt-parent = <&pmc>;
-					clocks = <&clk32k>, <&main>, <&plla>, <&pllb>;
+					clocks = <&slow_xtal>, <&main>, <&plla>, <&pllb>;
 
 					prog0: prog0 {
 						#clock-cells = <0>;
diff --git a/arch/arm/boot/dts/at91sam9261ek.dts b/arch/arm/boot/dts/at91sam9261ek.dts
index c6683ea8b743..aa35a7aec9a8 100644
--- a/arch/arm/boot/dts/at91sam9261ek.dts
+++ b/arch/arm/boot/dts/at91sam9261ek.dts
@@ -20,6 +20,10 @@
 		reg = <0x20000000 0x4000000>;
 	};
 
+	slow_xtal {
+		clock-frequency = <32768>;
+	};
+
 	main_xtal {
 		clock-frequency = <18432000>;
 	};
diff --git a/arch/arm/boot/dts/at91sam9n12.dtsi b/arch/arm/boot/dts/at91sam9n12.dtsi
index d1b82e6635d5..b84bac5bada4 100644
--- a/arch/arm/boot/dts/at91sam9n12.dtsi
+++ b/arch/arm/boot/dts/at91sam9n12.dtsi
@@ -132,8 +132,8 @@
 								      <595000000 650000000 3 0>,
 								      <545000000 600000000 0 1>,
 								      <495000000 555000000 1 1>,
-								      <445000000 500000000 1 2>,
-								      <400000000 450000000 1 3>;
+								      <445000000 500000000 2 1>,
+								      <400000000 450000000 3 1>;
 				};
 
 				plladiv: plladivck {
@@ -925,7 +925,7 @@
 			compatible = "atmel,at91rm9200-ohci", "usb-ohci";
 			reg = <0x00500000 0x00100000>;
 			interrupts = <22 IRQ_TYPE_LEVEL_HIGH 2>;
-			clocks = <&usb>, <&uhphs_clk>, <&udphs_clk>,
+			clocks = <&usb>, <&uhphs_clk>, <&uhphs_clk>,
 				 <&uhpck>;
 			clock-names = "usb_clk", "ohci_clk", "hclk", "uhpck";
 			status = "disabled";
diff --git a/arch/arm/boot/dts/at91sam9x5.dtsi b/arch/arm/boot/dts/at91sam9x5.dtsi
index 1a57298636a5..2c0d6ea3ab41 100644
--- a/arch/arm/boot/dts/at91sam9x5.dtsi
+++ b/arch/arm/boot/dts/at91sam9x5.dtsi
@@ -140,8 +140,8 @@
 								       595000000 650000000 3 0
 								       545000000 600000000 0 1
 								       495000000 555000000 1 1
-								       445000000 500000000 1 2
-								       400000000 450000000 1 3>;
+								       445000000 500000000 2 1
+								       400000000 450000000 3 1>;
 				};
 
 				plladiv: plladivck {
@@ -1045,6 +1045,8 @@
 				reg = <0x00500000 0x80000
 				       0xf803c000 0x400>;
 				interrupts = <23 IRQ_TYPE_LEVEL_HIGH 0>;
+				clocks = <&usb>, <&udphs_clk>;
+				clock-names = "hclk", "pclk";
 				status = "disabled";
 
 				ep0 {
@@ -1122,6 +1124,7 @@
 				compatible = "atmel,at91sam9rl-pwm";
 				reg = <0xf8034000 0x300>;
 				interrupts = <18 IRQ_TYPE_LEVEL_HIGH 4>;
+				clocks = <&pwm_clk>;
 				#pwm-cells = <3>;
 				status = "disabled";
 			};
@@ -1153,8 +1156,7 @@
 			compatible = "atmel,at91rm9200-ohci", "usb-ohci";
 			reg = <0x00600000 0x100000>;
 			interrupts = <22 IRQ_TYPE_LEVEL_HIGH 2>;
-			clocks = <&usb>, <&uhphs_clk>, <&udphs_clk>,
-				 <&uhpck>;
+			clocks = <&usb>, <&uhphs_clk>, <&uhphs_clk>, <&uhpck>;
 			clock-names = "usb_clk", "ohci_clk", "hclk", "uhpck";
 			status = "disabled";
 		};
diff --git a/arch/arm/boot/dts/dra7-evm.dts b/arch/arm/boot/dts/dra7-evm.dts
index 4adc28039c30..83089540e324 100644
--- a/arch/arm/boot/dts/dra7-evm.dts
+++ b/arch/arm/boot/dts/dra7-evm.dts
@@ -240,6 +240,7 @@
 					regulator-name = "ldo3";
 					regulator-min-microvolt = <1800000>;
 					regulator-max-microvolt = <1800000>;
+					regulator-always-on;
 					regulator-boot-on;
 				};
 
diff --git a/arch/arm/boot/dts/dra7.dtsi b/arch/arm/boot/dts/dra7.dtsi
index c29945e07c5a..80127638b379 100644
--- a/arch/arm/boot/dts/dra7.dtsi
+++ b/arch/arm/boot/dts/dra7.dtsi
@@ -773,7 +773,6 @@
 			clocks = <&qspi_gfclk_div>;
 			clock-names = "fck";
 			num-cs = <4>;
-			interrupts = <0 343 0x4>;
 			status = "disabled";
 		};
 
@@ -984,6 +983,17 @@
 			#size-cells = <1>;
 			status = "disabled";
 		};
+
+		atl: atl@4843c000 {
+			compatible = "ti,dra7-atl";
+			reg = <0x4843c000 0x3ff>;
+			ti,hwmods = "atl";
+			ti,provided-clocks = <&atl_clkin0_ck>, <&atl_clkin1_ck>,
+					     <&atl_clkin2_ck>, <&atl_clkin3_ck>;
+			clocks = <&atl_gfclk_mux>;
+			clock-names = "fck";
+			status = "disabled";
+		};
 	};
 };
 
diff --git a/arch/arm/boot/dts/dra7xx-clocks.dtsi b/arch/arm/boot/dts/dra7xx-clocks.dtsi
index b03cfe49d22b..dc7a292fe939 100644
--- a/arch/arm/boot/dts/dra7xx-clocks.dtsi
+++ b/arch/arm/boot/dts/dra7xx-clocks.dtsi
@@ -10,26 +10,26 @@
 &cm_core_aon_clocks {
 	atl_clkin0_ck: atl_clkin0_ck {
 		#clock-cells = <0>;
-		compatible = "fixed-clock";
-		clock-frequency = <0>;
+		compatible = "ti,dra7-atl-clock";
+		clocks = <&atl_gfclk_mux>;
 	};
 
 	atl_clkin1_ck: atl_clkin1_ck {
 		#clock-cells = <0>;
-		compatible = "fixed-clock";
-		clock-frequency = <0>;
+		compatible = "ti,dra7-atl-clock";
+		clocks = <&atl_gfclk_mux>;
 	};
 
 	atl_clkin2_ck: atl_clkin2_ck {
 		#clock-cells = <0>;
-		compatible = "fixed-clock";
-		clock-frequency = <0>;
+		compatible = "ti,dra7-atl-clock";
+		clocks = <&atl_gfclk_mux>;
 	};
 
 	atl_clkin3_ck: atl_clkin3_ck {
 		#clock-cells = <0>;
-		compatible = "fixed-clock";
-		clock-frequency = <0>;
+		compatible = "ti,dra7-atl-clock";
+		clocks = <&atl_gfclk_mux>;
 	};
 
 	hdmi_clkin_ck: hdmi_clkin_ck {
@@ -673,10 +673,12 @@
 
 	l3_iclk_div: l3_iclk_div {
 		#clock-cells = <0>;
-		compatible = "fixed-factor-clock";
+		compatible = "ti,divider-clock";
+		ti,max-div = <2>;
+		ti,bit-shift = <4>;
+		reg = <0x0100>;
 		clocks = <&dpll_core_h12x2_ck>;
-		clock-mult = <1>;
-		clock-div = <1>;
+		ti,index-power-of-two;
 	};
 
 	l4_root_clk_div: l4_root_clk_div {
@@ -684,7 +686,7 @@
 		compatible = "fixed-factor-clock";
 		clocks = <&l3_iclk_div>;
 		clock-mult = <1>;
-		clock-div = <1>;
+		clock-div = <2>;
 	};
 
 	video1_clk2_div: video1_clk2_div {
diff --git a/arch/arm/boot/dts/exynos4.dtsi b/arch/arm/boot/dts/exynos4.dtsi
index b8ece4be41ca..17b22e9cc2aa 100644
--- a/arch/arm/boot/dts/exynos4.dtsi
+++ b/arch/arm/boot/dts/exynos4.dtsi
@@ -113,7 +113,7 @@
 		compatible = "arm,cortex-a9-gic";
 		#interrupt-cells = <3>;
 		interrupt-controller;
-		reg = <0x10490000 0x1000>, <0x10480000 0x100>;
+		reg = <0x10490000 0x10000>, <0x10480000 0x10000>;
 	};
 
 	combiner: interrupt-controller@10440000 {
@@ -554,7 +554,7 @@
 		interrupts = <0 37 0>, <0 38 0>, <0 39 0>, <0 40 0>, <0 41 0>;
 		clocks = <&clock CLK_PWM>;
 		clock-names = "timers";
-		#pwm-cells = <2>;
+		#pwm-cells = <3>;
 		status = "disabled";
 	};
 
diff --git a/arch/arm/boot/dts/exynos5420.dtsi b/arch/arm/boot/dts/exynos5420.dtsi
index e38532271ef9..15957227ffda 100644
--- a/arch/arm/boot/dts/exynos5420.dtsi
+++ b/arch/arm/boot/dts/exynos5420.dtsi
@@ -167,7 +167,7 @@
 		compatible = "samsung,exynos5420-audss-clock";
 		reg = <0x03810000 0x0C>;
 		#clock-cells = <1>;
-		clocks = <&clock CLK_FIN_PLL>, <&clock CLK_FOUT_EPLL>,
+		clocks = <&clock CLK_FIN_PLL>, <&clock CLK_MAU_EPLL>,
 			 <&clock CLK_SCLK_MAUDIO0>, <&clock CLK_SCLK_MAUPCM0>;
 		clock-names = "pll_ref", "pll_in", "sclk_audio", "sclk_pcm_in";
 	};
@@ -260,6 +260,9 @@
 	mfc_pd: power-domain@10044060 {
 		compatible = "samsung,exynos4210-pd";
 		reg = <0x10044060 0x20>;
+		clocks = <&clock CLK_FIN_PLL>, <&clock CLK_MOUT_SW_ACLK333>,
+			<&clock CLK_MOUT_USER_ACLK333>;
+		clock-names = "oscclk", "pclk0", "clk0";
 	};
 
 	disp_pd: power-domain@100440C0 {
diff --git a/arch/arm/boot/dts/imx51-babbage.dts b/arch/arm/boot/dts/imx51-babbage.dts
index 6bc3243a80d3..181d77fa2fa6 100644
--- a/arch/arm/boot/dts/imx51-babbage.dts
+++ b/arch/arm/boot/dts/imx51-babbage.dts
@@ -315,15 +315,15 @@
 &esdhc1 {
 	pinctrl-names = "default";
 	pinctrl-0 = <&pinctrl_esdhc1>;
-	fsl,cd-controller;
-	fsl,wp-controller;
+	cd-gpios = <&gpio1 0 GPIO_ACTIVE_LOW>;
+	wp-gpios = <&gpio1 1 GPIO_ACTIVE_HIGH>;
 	status = "okay";
 };
 
 &esdhc2 {
 	pinctrl-names = "default";
 	pinctrl-0 = <&pinctrl_esdhc2>;
-	cd-gpios = <&gpio1 6 GPIO_ACTIVE_HIGH>;
+	cd-gpios = <&gpio1 6 GPIO_ACTIVE_LOW>;
 	wp-gpios = <&gpio1 5 GPIO_ACTIVE_HIGH>;
 	status = "okay";
 };
@@ -468,8 +468,8 @@
 				MX51_PAD_SD1_DATA1__SD1_DATA1		0x20d5
 				MX51_PAD_SD1_DATA2__SD1_DATA2		0x20d5
 				MX51_PAD_SD1_DATA3__SD1_DATA3		0x20d5
-				MX51_PAD_GPIO1_0__SD1_CD		0x20d5
-				MX51_PAD_GPIO1_1__SD1_WP		0x20d5
+				MX51_PAD_GPIO1_0__GPIO1_0		0x100
+				MX51_PAD_GPIO1_1__GPIO1_1		0x100
 			>;
 		};
 
diff --git a/arch/arm/boot/dts/imx51-eukrea-mbimxsd51-baseboard.dts b/arch/arm/boot/dts/imx51-eukrea-mbimxsd51-baseboard.dts
index 75e66c9c6144..31cfb7f2b02e 100644
--- a/arch/arm/boot/dts/imx51-eukrea-mbimxsd51-baseboard.dts
+++ b/arch/arm/boot/dts/imx51-eukrea-mbimxsd51-baseboard.dts
@@ -107,7 +107,7 @@
 &esdhc1 {
 	pinctrl-names = "default";
 	pinctrl-0 = <&pinctrl_esdhc1 &pinctrl_esdhc1_cd>;
-	fsl,cd-controller;
+	cd-gpios = <&gpio1 0 GPIO_ACTIVE_LOW>;
 	status = "okay";
 };
 
@@ -206,7 +206,7 @@
 
 		pinctrl_esdhc1_cd: esdhc1_cd {
 			fsl,pins = <
-				MX51_PAD_GPIO1_0__SD1_CD 0x20d5
+				MX51_PAD_GPIO1_0__GPIO1_0 0xd5
 			>;
 		};
 
diff --git a/arch/arm/boot/dts/imx53-m53evk.dts b/arch/arm/boot/dts/imx53-m53evk.dts
index d5d146a8b149..c4956b0ffb35 100644
--- a/arch/arm/boot/dts/imx53-m53evk.dts
+++ b/arch/arm/boot/dts/imx53-m53evk.dts
@@ -21,27 +21,25 @@
 		      <0xb0000000 0x20000000>;
 	};
 
-	soc {
-		display1: display@di1 {
-			compatible = "fsl,imx-parallel-display";
-			interface-pix-fmt = "bgr666";
-			pinctrl-names = "default";
-			pinctrl-0 = <&pinctrl_ipu_disp1>;
-
-			display-timings {
-				800x480p60 {
-					native-mode;
-					clock-frequency = <31500000>;
-					hactive = <800>;
-					vactive = <480>;
-					hfront-porch = <40>;
-					hback-porch = <88>;
-					hsync-len = <128>;
-					vback-porch = <33>;
-					vfront-porch = <9>;
-					vsync-len = <3>;
-					vsync-active = <1>;
-				};
+	display1: display@di1 {
+		compatible = "fsl,imx-parallel-display";
+		interface-pix-fmt = "bgr666";
+		pinctrl-names = "default";
+		pinctrl-0 = <&pinctrl_ipu_disp1>;
+
+		display-timings {
+			800x480p60 {
+				native-mode;
+				clock-frequency = <31500000>;
+				hactive = <800>;
+				vactive = <480>;
+				hfront-porch = <40>;
+				hback-porch = <88>;
+				hsync-len = <128>;
+				vback-porch = <33>;
+				vfront-porch = <9>;
+				vsync-len = <3>;
+				vsync-active = <1>;
 			};
 		};
 
diff --git a/arch/arm/boot/dts/imx6dl-hummingboard.dts b/arch/arm/boot/dts/imx6dl-hummingboard.dts
index 5373a5f2782b..c8e51dd41b8f 100644
--- a/arch/arm/boot/dts/imx6dl-hummingboard.dts
+++ b/arch/arm/boot/dts/imx6dl-hummingboard.dts
@@ -143,6 +143,14 @@
 			fsl,pins = <MX6QDL_PAD_GPIO_0__GPIO1_IO00 0x1b0b0>;
 		};
 
+		pinctrl_hummingboard_usbotg_id: hummingboard-usbotg-id {
+			/*
+			 * Similar to pinctrl_usbotg_2, but we want it
+			 * pulled down for a fixed host connection.
+			 */
+			fsl,pins = <MX6QDL_PAD_GPIO_1__USB_OTG_ID 0x13059>;
+		};
+
 		pinctrl_hummingboard_usbotg_vbus: hummingboard-usbotg-vbus {
 			fsl,pins = <MX6QDL_PAD_EIM_D22__GPIO3_IO22 0x1b0b0>;
 		};
@@ -178,6 +186,8 @@
 };
 
 &usbotg {
+	pinctrl-names = "default";
+	pinctrl-0 = <&pinctrl_hummingboard_usbotg_id>;
 	vbus-supply = <&reg_usbotg_vbus>;
 	status = "okay";
 };
diff --git a/arch/arm/boot/dts/imx6q-gw51xx.dts b/arch/arm/boot/dts/imx6q-gw51xx.dts
index af4929aee075..0e1406e58eff 100644
--- a/arch/arm/boot/dts/imx6q-gw51xx.dts
+++ b/arch/arm/boot/dts/imx6q-gw51xx.dts
@@ -11,7 +11,7 @@
 
 /dts-v1/;
 #include "imx6q.dtsi"
-#include "imx6qdl-gw54xx.dtsi"
+#include "imx6qdl-gw51xx.dtsi"
 
 / {
 	model = "Gateworks Ventana i.MX6 Quad GW51XX";
diff --git a/arch/arm/boot/dts/imx6qdl-cubox-i.dtsi b/arch/arm/boot/dts/imx6qdl-cubox-i.dtsi
index 25da82a03110..e8e781656b3f 100644
--- a/arch/arm/boot/dts/imx6qdl-cubox-i.dtsi
+++ b/arch/arm/boot/dts/imx6qdl-cubox-i.dtsi
@@ -12,6 +12,19 @@
 		pinctrl-0 = <&pinctrl_cubox_i_ir>;
 	};
 
+	pwmleds {
+		compatible = "pwm-leds";
+		pinctrl-names = "default";
+		pinctrl-0 = <&pinctrl_cubox_i_pwm1>;
+
+		front {
+			active-low;
+			label = "imx6:red:front";
+			max-brightness = <248>;
+			pwms = <&pwm1 0 50000>;
+		};
+	};
+
 	regulators {
 		compatible = "simple-bus";
 
@@ -109,6 +122,10 @@
 			>;
 		};
 
+		pinctrl_cubox_i_pwm1: cubox-i-pwm1-front-led {
+			fsl,pins = <MX6QDL_PAD_DISP0_DAT8__PWM1_OUT 0x1b0b0>;
+		};
+
 		pinctrl_cubox_i_spdif: cubox-i-spdif {
 			fsl,pins = <MX6QDL_PAD_GPIO_17__SPDIF_OUT 0x13091>;
 		};
@@ -117,6 +134,14 @@
 			fsl,pins = <MX6QDL_PAD_GPIO_0__GPIO1_IO00 0x4001b0b0>;
 		};
 
+		pinctrl_cubox_i_usbotg_id: cubox-i-usbotg-id {
+			/*
+			 * The Cubox-i pulls this low, but as it's pointless
+			 * leaving it as a pull-up, even if it is just 10uA.
+			 */
+			fsl,pins = <MX6QDL_PAD_GPIO_1__USB_OTG_ID 0x13059>;
+		};
+
 		pinctrl_cubox_i_usbotg_vbus: cubox-i-usbotg-vbus {
 			fsl,pins = <MX6QDL_PAD_EIM_D22__GPIO3_IO22 0x4001b0b0>;
 		};
@@ -153,6 +178,8 @@
 };
 
 &usbotg {
+	pinctrl-names = "default";
+	pinctrl-0 = <&pinctrl_cubox_i_usbotg_id>;
 	vbus-supply = <&reg_usbotg_vbus>;
 	status = "okay";
 };
diff --git a/arch/arm/boot/dts/imx6qdl-gw51xx.dtsi b/arch/arm/boot/dts/imx6qdl-gw51xx.dtsi
index 31665adcbf39..0db15af41cb1 100644
--- a/arch/arm/boot/dts/imx6qdl-gw51xx.dtsi
+++ b/arch/arm/boot/dts/imx6qdl-gw51xx.dtsi
@@ -161,7 +161,7 @@
 	status = "okay";
 
 	pmic: ltc3676@3c {
-		compatible = "ltc,ltc3676";
+		compatible = "lltc,ltc3676";
 		reg = <0x3c>;
 
 		regulators {
diff --git a/arch/arm/boot/dts/imx6qdl-gw52xx.dtsi b/arch/arm/boot/dts/imx6qdl-gw52xx.dtsi
index 367af3ec9435..744c8a2d81f6 100644
--- a/arch/arm/boot/dts/imx6qdl-gw52xx.dtsi
+++ b/arch/arm/boot/dts/imx6qdl-gw52xx.dtsi
@@ -220,7 +220,7 @@
 	};
 
 	pmic: ltc3676@3c {
-		compatible = "ltc,ltc3676";
+		compatible = "lltc,ltc3676";
 		reg = <0x3c>;
 
 		regulators {
@@ -288,7 +288,7 @@
 	codec: sgtl5000@0a {
 		compatible = "fsl,sgtl5000";
 		reg = <0x0a>;
-		clocks = <&clks 169>;
+		clocks = <&clks 201>;
 		VDDA-supply = <&reg_1p8v>;
 		VDDIO-supply = <&reg_3p3v>;
 	};
diff --git a/arch/arm/boot/dts/imx6qdl-gw53xx.dtsi b/arch/arm/boot/dts/imx6qdl-gw53xx.dtsi
index c91b5a6c769b..adf150c1be90 100644
--- a/arch/arm/boot/dts/imx6qdl-gw53xx.dtsi
+++ b/arch/arm/boot/dts/imx6qdl-gw53xx.dtsi
@@ -234,7 +234,7 @@
 	};
 
 	pmic: ltc3676@3c {
-		compatible = "ltc,ltc3676";
+		compatible = "lltc,ltc3676";
 		reg = <0x3c>;
 
 		regulators {
diff --git a/arch/arm/boot/dts/imx6qdl-microsom.dtsi b/arch/arm/boot/dts/imx6qdl-microsom.dtsi
index d729d0b15f25..79eac6849d4c 100644
--- a/arch/arm/boot/dts/imx6qdl-microsom.dtsi
+++ b/arch/arm/boot/dts/imx6qdl-microsom.dtsi
@@ -10,14 +10,6 @@
 				MX6QDL_PAD_CSI0_DAT11__UART1_RX_DATA	0x1b0b1
 			>;
 		};
-
-		pinctrl_microsom_usbotg: microsom-usbotg {
-			/*
-			 * Similar to pinctrl_usbotg_2, but we want it
-			 * pulled down for a fixed host connection.
-			 */
-			fsl,pins = <MX6QDL_PAD_GPIO_1__USB_OTG_ID 0x13059>;
-		};
 	};
 };
 
@@ -26,8 +18,3 @@
 	pinctrl-0 = <&pinctrl_microsom_uart1>;
 	status = "okay";
 };
-
-&usbotg {
-	pinctrl-names = "default";
-	pinctrl-0 = <&pinctrl_microsom_usbotg>;
-};
diff --git a/arch/arm/boot/dts/imx6sl.dtsi b/arch/arm/boot/dts/imx6sl.dtsi
index 2d4e5285f3f3..57d4abe03a94 100644
--- a/arch/arm/boot/dts/imx6sl.dtsi
+++ b/arch/arm/boot/dts/imx6sl.dtsi
@@ -686,7 +686,7 @@
 				compatible = "fsl,imx6sl-fec", "fsl,imx25-fec";
 				reg = <0x02188000 0x4000>;
 				interrupts = <0 114 IRQ_TYPE_LEVEL_HIGH>;
-				clocks = <&clks IMX6SL_CLK_ENET_REF>,
+				clocks = <&clks IMX6SL_CLK_ENET>,
 					 <&clks IMX6SL_CLK_ENET_REF>;
 				clock-names = "ipg", "ahb";
 				status = "disabled";
diff --git a/arch/arm/boot/dts/kirkwood-guruplug-server-plus.dts b/arch/arm/boot/dts/kirkwood-guruplug-server-plus.dts
index c5a1fc75c7a3..b2d9834bf458 100644
--- a/arch/arm/boot/dts/kirkwood-guruplug-server-plus.dts
+++ b/arch/arm/boot/dts/kirkwood-guruplug-server-plus.dts
@@ -105,7 +105,6 @@
 		compatible = "ethernet-phy-id0141.0cb0",
 		             "ethernet-phy-ieee802.3-c22";
 		reg = <0>;
-		phy-connection-type = "rgmii-id";
 	};
 
 	ethphy1: ethernet-phy@1 {
@@ -113,7 +112,6 @@
 		compatible = "ethernet-phy-id0141.0cb0",
 		             "ethernet-phy-ieee802.3-c22";
 		reg = <1>;
-		phy-connection-type = "rgmii-id";
 	};
 };
 
@@ -121,6 +119,7 @@
 	status = "okay";
 	ethernet0-port@0 {
 		phy-handle = <&ethphy0>;
+		phy-connection-type = "rgmii-id";
 	};
 };
 
@@ -128,5 +127,6 @@
 	status = "okay";
 	ethernet1-port@0 {
 		phy-handle = <&ethphy1>;
+		phy-connection-type = "rgmii-id";
 	};
 };
diff --git a/arch/arm/boot/dts/omap3-beagle-xm.dts b/arch/arm/boot/dts/omap3-beagle-xm.dts
index cf0be662297e..1becefce821b 100644
--- a/arch/arm/boot/dts/omap3-beagle-xm.dts
+++ b/arch/arm/boot/dts/omap3-beagle-xm.dts
@@ -251,6 +251,11 @@
 			codec {
 			};
 		};
+
+		twl_power: power {
+			compatible = "ti,twl4030-power-beagleboard-xm", "ti,twl4030-power-idle-osc-off";
+			ti,use_poweroff;
+		};
 	};
 };
 
@@ -301,6 +306,7 @@
 };
 
 &uart3 {
+	interrupts-extended = <&intc 74 &omap3_pmx_core OMAP3_UART3_RX>;
 	pinctrl-names = "default";
 	pinctrl-0 = <&uart3_pins>;
 };
diff --git a/arch/arm/boot/dts/omap3-evm-common.dtsi b/arch/arm/boot/dts/omap3-evm-common.dtsi
index 8ae8f007c8ad..c8747c7f1cc8 100644
--- a/arch/arm/boot/dts/omap3-evm-common.dtsi
+++ b/arch/arm/boot/dts/omap3-evm-common.dtsi
@@ -50,6 +50,13 @@
 	gpios = <&twl_gpio 18 GPIO_ACTIVE_LOW>;
 };
 
+&twl {
+	twl_power: power {
+		compatible = "ti,twl4030-power-omap3-evm", "ti,twl4030-power-idle";
+		ti,use_poweroff;
+	};
+};
+
 &i2c2 {
 	clock-frequency = <400000>;
 };
diff --git a/arch/arm/boot/dts/omap3-n900.dts b/arch/arm/boot/dts/omap3-n900.dts
index ae8ae3f4f9bf..1fe45d1f75ec 100644
--- a/arch/arm/boot/dts/omap3-n900.dts
+++ b/arch/arm/boot/dts/omap3-n900.dts
@@ -351,6 +351,11 @@
 		compatible = "ti,twl4030-audio";
 		ti,enable-vibra = <1>;
 	};
+
+	twl_power: power {
+		compatible = "ti,twl4030-power-n900", "ti,twl4030-power-idle-osc-off";
+		ti,use_poweroff;
+	};
 };
 
 &twl_keypad {
diff --git a/arch/arm/boot/dts/omap5.dtsi b/arch/arm/boot/dts/omap5.dtsi
index 3bfda16c8b52..a4ed54988866 100644
--- a/arch/arm/boot/dts/omap5.dtsi
+++ b/arch/arm/boot/dts/omap5.dtsi
@@ -45,7 +45,6 @@
 
 			operating-points = <
 				/* kHz    uV */
-				500000  880000
 				1000000 1060000
 				1500000 1250000
 			>;
diff --git a/arch/arm/boot/dts/stih415.dtsi b/arch/arm/boot/dts/stih415.dtsi
index d6f254f302fe..a0f6f75fe3b5 100644
--- a/arch/arm/boot/dts/stih415.dtsi
+++ b/arch/arm/boot/dts/stih415.dtsi
@@ -169,8 +169,8 @@
 
 			pinctrl-names 	= "default";
 			pinctrl-0	= <&pinctrl_mii0>;
-			clock-names	= "stmmaceth";
-			clocks		= <&clk_s_a1_ls CLK_GMAC0_PHY>;
+			clock-names	= "stmmaceth", "sti-ethclk";
+			clocks		= <&clk_s_a1_ls CLK_ICN_IF_2>, <&clk_s_a1_ls CLK_GMAC0_PHY>;
 		};
 
 		ethernet1: dwmac@fef08000 {
@@ -192,8 +192,8 @@
 			reset-names		= "stmmaceth";
 			pinctrl-names 	= "default";
 			pinctrl-0	= <&pinctrl_mii1>;
-			clock-names	= "stmmaceth";
-			clocks		= <&clk_s_a0_ls CLK_ETH1_PHY>;
+			clock-names	= "stmmaceth", "sti-ethclk";
+			clocks		= <&clk_s_a0_ls CLK_ICN_REG>, <&clk_s_a0_ls CLK_ETH1_PHY>;
 		};
 
 		rc: rc@fe518000 {
diff --git a/arch/arm/boot/dts/stih416-b2020-revE.dts b/arch/arm/boot/dts/stih416-b2020e.dts
index ba0fa2caaf18..ba0fa2caaf18 100644
--- a/arch/arm/boot/dts/stih416-b2020-revE.dts
+++ b/arch/arm/boot/dts/stih416-b2020e.dts
diff --git a/arch/arm/boot/dts/stih416.dtsi b/arch/arm/boot/dts/stih416.dtsi
index 06473c5d9ea9..84758d76d064 100644
--- a/arch/arm/boot/dts/stih416.dtsi
+++ b/arch/arm/boot/dts/stih416.dtsi
@@ -175,8 +175,8 @@
 			reset-names		= "stmmaceth";
 			pinctrl-names 	= "default";
 			pinctrl-0	= <&pinctrl_mii0>;
-			clock-names	= "stmmaceth";
-			clocks		= <&clk_s_a1_ls CLK_GMAC0_PHY>;
+			clock-names	= "stmmaceth", "sti-ethclk";
+			clocks		= <&clk_s_a1_ls CLK_ICN_IF_2>, <&clk_s_a1_ls CLK_GMAC0_PHY>;
 		};
 
 		ethernet1: dwmac@fef08000 {
@@ -197,8 +197,8 @@
 			reset-names	= "stmmaceth";
 			pinctrl-names 	= "default";
 			pinctrl-0	= <&pinctrl_mii1>;
-			clock-names	= "stmmaceth";
-			clocks		= <&clk_s_a0_ls CLK_ETH1_PHY>;
+			clock-names	= "stmmaceth", "sti-ethclk";
+			clocks		= <&clk_s_a0_ls CLK_ICN_REG>, <&clk_s_a0_ls CLK_ETH1_PHY>;
 		};
 
 		rc: rc@fe518000 {
diff --git a/arch/arm/common/scoop.c b/arch/arm/common/scoop.c
index 6ef146edd0cd..a20fa80776d3 100644
--- a/arch/arm/common/scoop.c
+++ b/arch/arm/common/scoop.c
@@ -182,7 +182,6 @@ static int scoop_probe(struct platform_device *pdev)
 	struct scoop_config *inf;
 	struct resource *mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	int ret;
-	int temp;
 
 	if (!mem)
 		return -EINVAL;
diff --git a/arch/arm/configs/bcm_defconfig b/arch/arm/configs/bcm_defconfig
index 9d13dae99125..4bf72264b175 100644
--- a/arch/arm/configs/bcm_defconfig
+++ b/arch/arm/configs/bcm_defconfig
@@ -94,10 +94,10 @@ CONFIG_BACKLIGHT_CLASS_DEVICE=y
 CONFIG_BACKLIGHT_PWM=y
 # CONFIG_USB_SUPPORT is not set
 CONFIG_MMC=y
-CONFIG_MMC_UNSAFE_RESUME=y
 CONFIG_MMC_BLOCK_MINORS=32
 CONFIG_MMC_TEST=y
 CONFIG_MMC_SDHCI=y
+CONFIG_MMC_SDHCI_PLTFM=y
 CONFIG_MMC_SDHCI_BCM_KONA=y
 CONFIG_NEW_LEDS=y
 CONFIG_LEDS_CLASS=y
diff --git a/arch/arm/configs/imx_v6_v7_defconfig b/arch/arm/configs/imx_v6_v7_defconfig
index ef8815327e5b..59b7e45142d8 100644
--- a/arch/arm/configs/imx_v6_v7_defconfig
+++ b/arch/arm/configs/imx_v6_v7_defconfig
@@ -186,6 +186,7 @@ CONFIG_VIDEO_MX3=y
 CONFIG_V4L_MEM2MEM_DRIVERS=y
 CONFIG_VIDEO_CODA=y
 CONFIG_SOC_CAMERA_OV2640=y
+CONFIG_IMX_IPUV3_CORE=y
 CONFIG_DRM=y
 CONFIG_DRM_PANEL_SIMPLE=y
 CONFIG_BACKLIGHT_LCD_SUPPORT=y
diff --git a/arch/arm/configs/multi_v7_defconfig b/arch/arm/configs/multi_v7_defconfig
index 17d9462b9fb9..534836497998 100644
--- a/arch/arm/configs/multi_v7_defconfig
+++ b/arch/arm/configs/multi_v7_defconfig
@@ -223,12 +223,12 @@ CONFIG_POWER_RESET_GPIO=y
 CONFIG_POWER_RESET_SUN6I=y
 CONFIG_SENSORS_LM90=y
 CONFIG_THERMAL=y
-CONFIG_DOVE_THERMAL=y
 CONFIG_ARMADA_THERMAL=y
 CONFIG_WATCHDOG=y
 CONFIG_ORION_WATCHDOG=y
 CONFIG_SUNXI_WATCHDOG=y
 CONFIG_MFD_AS3722=y
+CONFIG_MFD_BCM590XX=y
 CONFIG_MFD_CROS_EC=y
 CONFIG_MFD_CROS_EC_SPI=y
 CONFIG_MFD_MAX8907=y
@@ -240,6 +240,7 @@ CONFIG_MFD_TPS65910=y
 CONFIG_REGULATOR_VIRTUAL_CONSUMER=y
 CONFIG_REGULATOR_AB8500=y
 CONFIG_REGULATOR_AS3722=y
+CONFIG_REGULATOR_BCM590XX=y
 CONFIG_REGULATOR_GPIO=y
 CONFIG_REGULATOR_MAX8907=y
 CONFIG_REGULATOR_PALMAS=y
@@ -353,6 +354,7 @@ CONFIG_MFD_NVEC=y
 CONFIG_KEYBOARD_NVEC=y
 CONFIG_SERIO_NVEC_PS2=y
 CONFIG_NVEC_POWER=y
+CONFIG_QCOM_GSBI=y
 CONFIG_COMMON_CLK_QCOM=y
 CONFIG_MSM_GCC_8660=y
 CONFIG_MSM_MMCC_8960=y
diff --git a/arch/arm/configs/mvebu_v7_defconfig b/arch/arm/configs/mvebu_v7_defconfig
index e11170e37442..b0bfefa23902 100644
--- a/arch/arm/configs/mvebu_v7_defconfig
+++ b/arch/arm/configs/mvebu_v7_defconfig
@@ -14,6 +14,7 @@ CONFIG_MACH_ARMADA_370=y
 CONFIG_MACH_ARMADA_375=y
 CONFIG_MACH_ARMADA_38X=y
 CONFIG_MACH_ARMADA_XP=y
+CONFIG_MACH_DOVE=y
 CONFIG_NEON=y
 # CONFIG_CACHE_L2X0 is not set
 # CONFIG_SWP_EMULATE is not set
@@ -52,6 +53,7 @@ CONFIG_INPUT_EVDEV=y
 CONFIG_KEYBOARD_GPIO=y
 CONFIG_SERIAL_8250=y
 CONFIG_SERIAL_8250_CONSOLE=y
+CONFIG_SERIAL_OF_PLATFORM=y
 CONFIG_I2C=y
 CONFIG_SPI=y
 CONFIG_SPI_ORION=y
diff --git a/arch/arm/include/asm/mcpm.h b/arch/arm/include/asm/mcpm.h
index d9702eb0b02b..94060adba174 100644
--- a/arch/arm/include/asm/mcpm.h
+++ b/arch/arm/include/asm/mcpm.h
@@ -208,8 +208,6 @@ struct sync_struct {
 	struct mcpm_sync_struct clusters[MAX_NR_CLUSTERS];
 };
 
-extern unsigned long sync_phys;	/* physical address of *mcpm_sync */
-
 void __mcpm_cpu_going_down(unsigned int cpu, unsigned int cluster);
 void __mcpm_cpu_down(unsigned int cpu, unsigned int cluster);
 void __mcpm_outbound_leave_critical(unsigned int cluster, int state);
diff --git a/arch/arm/include/asm/thread_info.h b/arch/arm/include/asm/thread_info.h
index f989d7c22dc5..e4e4208a9130 100644
--- a/arch/arm/include/asm/thread_info.h
+++ b/arch/arm/include/asm/thread_info.h
@@ -114,8 +114,14 @@ static inline struct thread_info *current_thread_info(void)
 	((unsigned long)(task_thread_info(tsk)->cpu_context.pc))
 #define thread_saved_sp(tsk)	\
 	((unsigned long)(task_thread_info(tsk)->cpu_context.sp))
+
+#ifndef CONFIG_THUMB2_KERNEL
 #define thread_saved_fp(tsk)	\
 	((unsigned long)(task_thread_info(tsk)->cpu_context.fp))
+#else
+#define thread_saved_fp(tsk)	\
+	((unsigned long)(task_thread_info(tsk)->cpu_context.r7))
+#endif
 
 extern void crunch_task_disable(struct thread_info *);
 extern void crunch_task_copy(struct thread_info *, void *);
diff --git a/arch/arm/kernel/kprobes-test-arm.c b/arch/arm/kernel/kprobes-test-arm.c
index 9db4b659d03e..cb1424240ff6 100644
--- a/arch/arm/kernel/kprobes-test-arm.c
+++ b/arch/arm/kernel/kprobes-test-arm.c
@@ -74,8 +74,6 @@ void kprobe_arm_test_cases(void)
 	TEST_RRR( op "lt" s "	r11, r",11,VAL1,", r",14,N(val),", asr r",7, 6,"")\
 	TEST_RR(  op "gt" s "	r12, r13"       ", r",14,val, ", ror r",14,7,"")\
 	TEST_RR(  op "le" s "	r14, r",0, val, ", r13"       ", lsl r",14,8,"")\
-	TEST_RR(  op s "	r12, pc"        ", r",14,val, ", ror r",14,7,"")\
-	TEST_RR(  op s "	r14, r",0, val, ", pc"        ", lsl r",14,8,"")\
 	TEST_R(   op "eq" s "	r0,  r",11,VAL1,", #0xf5")			\
 	TEST_R(   op "ne" s "	r11, r",0, VAL1,", #0xf5000000")		\
 	TEST_R(   op s "	r7,  r",8, VAL2,", #0x000af000")		\
@@ -103,8 +101,6 @@ void kprobe_arm_test_cases(void)
 	TEST_RRR( op "ge	r",11,VAL1,", r",14,N(val),", asr r",7, 6,"")	\
 	TEST_RR(  op "le	r13"       ", r",14,val, ", ror r",14,7,"")	\
 	TEST_RR(  op "gt	r",0, val, ", r13"       ", lsl r",14,8,"")	\
-	TEST_RR(  op "	pc"        ", r",14,val, ", ror r",14,7,"")		\
-	TEST_RR(  op "	r",0, val, ", pc"        ", lsl r",14,8,"")		\
 	TEST_R(   op "eq	r",11,VAL1,", #0xf5")				\
 	TEST_R(   op "ne	r",0, VAL1,", #0xf5000000")			\
 	TEST_R(   op "	r",8, VAL2,", #0x000af000")
@@ -125,7 +121,6 @@ void kprobe_arm_test_cases(void)
 	TEST_RR(  op "ge" s "	r11, r",11,N(val),", asr r",7, 6,"")	\
 	TEST_RR(  op "lt" s "	r12, r",11,val, ", ror r",14,7,"")	\
 	TEST_R(   op "gt" s "	r14, r13"       ", lsl r",14,8,"")	\
-	TEST_R(   op "le" s "	r14, pc"        ", lsl r",14,8,"")	\
 	TEST(     op "eq" s "	r0,  #0xf5")				\
 	TEST(     op "ne" s "	r11, #0xf5000000")			\
 	TEST(     op s "	r7,  #0x000af000")			\
@@ -159,12 +154,19 @@ void kprobe_arm_test_cases(void)
 	TEST_SUPPORTED("cmp	pc, #0x1000");
 	TEST_SUPPORTED("cmp	sp, #0x1000");
 
-	/* Data-processing with PC as shift*/
+	/* Data-processing with PC and a shift count in a register */
 	TEST_UNSUPPORTED(__inst_arm(0xe15c0f1e) "	@ cmp	r12, r14, asl pc")
 	TEST_UNSUPPORTED(__inst_arm(0xe1a0cf1e) "	@ mov	r12, r14, asl pc")
 	TEST_UNSUPPORTED(__inst_arm(0xe08caf1e) "	@ add	r10, r12, r14, asl pc")
-
-	/* Data-processing with PC as shift*/
+	TEST_UNSUPPORTED(__inst_arm(0xe151021f) "	@ cmp	r1, pc, lsl r2")
+	TEST_UNSUPPORTED(__inst_arm(0xe17f0211) "	@ cmn	pc, r1, lsl r2")
+	TEST_UNSUPPORTED(__inst_arm(0xe1a0121f) "	@ mov	r1, pc, lsl r2")
+	TEST_UNSUPPORTED(__inst_arm(0xe1a0f211) "	@ mov	pc, r1, lsl r2")
+	TEST_UNSUPPORTED(__inst_arm(0xe042131f) "	@ sub	r1, r2, pc, lsl r3")
+	TEST_UNSUPPORTED(__inst_arm(0xe1cf1312) "	@ bic	r1, pc, r2, lsl r3")
+	TEST_UNSUPPORTED(__inst_arm(0xe081f312) "	@ add	pc, r1, r2, lsl r3")
+
+	/* Data-processing with PC as a target and status registers updated */
 	TEST_UNSUPPORTED("movs	pc, r1")
 	TEST_UNSUPPORTED("movs	pc, r1, lsl r2")
 	TEST_UNSUPPORTED("movs	pc, #0x10000")
@@ -187,14 +189,14 @@ void kprobe_arm_test_cases(void)
 	TEST_BF_R ("add	pc, pc, r",14,2f-1f-8,"")
 	TEST_BF_R ("add	pc, r",14,2f-1f-8,", pc")
 	TEST_BF_R ("mov	pc, r",0,2f,"")
-	TEST_BF_RR("mov	pc, r",0,2f,", asl r",1,0,"")
+	TEST_BF_R ("add	pc, pc, r",14,(2f-1f-8)*2,", asr #1")
 	TEST_BB(   "sub	pc, pc, #1b-2b+8")
 #if __LINUX_ARM_ARCH__ == 6 && !defined(CONFIG_CPU_V7)
 	TEST_BB(   "sub	pc, pc, #1b-2b+8-2") /* UNPREDICTABLE before and after ARMv6 */
 #endif
 	TEST_BB_R( "sub	pc, pc, r",14, 1f-2f+8,"")
 	TEST_BB_R( "rsb	pc, r",14,1f-2f+8,", pc")
-	TEST_RR(   "add	pc, pc, r",10,-2,", asl r",11,1,"")
+	TEST_R(    "add	pc, pc, r",10,-2,", asl #1")
 #ifdef CONFIG_THUMB2_KERNEL
 	TEST_ARM_TO_THUMB_INTERWORK_R("add	pc, pc, r",0,3f-1f-8+1,"")
 	TEST_ARM_TO_THUMB_INTERWORK_R("sub	pc, r",0,3f+8+1,", #8")
@@ -216,6 +218,7 @@ void kprobe_arm_test_cases(void)
 	TEST_BB_R("bx	r",7,2f,"")
 	TEST_BF_R("bxeq	r",14,2f,"")
 
+#if __LINUX_ARM_ARCH__ >= 5
 	TEST_R("clz	r0, r",0, 0x0,"")
 	TEST_R("clzeq	r7, r",14,0x1,"")
 	TEST_R("clz	lr, r",7, 0xffffffff,"")
@@ -337,6 +340,7 @@ void kprobe_arm_test_cases(void)
 	TEST_UNSUPPORTED(__inst_arm(0xe16f02e1) " @ smultt pc, r1, r2")
 	TEST_UNSUPPORTED(__inst_arm(0xe16002ef) " @ smultt r0, pc, r2")
 	TEST_UNSUPPORTED(__inst_arm(0xe1600fe1) " @ smultt r0, r1, pc")
+#endif
 
 	TEST_GROUP("Multiply and multiply-accumulate")
 
@@ -559,6 +563,7 @@ void kprobe_arm_test_cases(void)
 	TEST_UNSUPPORTED("ldrsht	r1, [r2], #48")
 #endif
 
+#if __LINUX_ARM_ARCH__ >= 5
 	TEST_RPR(  "strd	r",0, VAL1,", [r",1, 48,", -r",2,24,"]")
 	TEST_RPR(  "strccd	r",8, VAL2,", [r",13,0, ", r",12,48,"]")
 	TEST_RPR(  "strd	r",4, VAL1,", [r",2, 24,", r",3, 48,"]!")
@@ -595,6 +600,7 @@ void kprobe_arm_test_cases(void)
 	TEST_UNSUPPORTED(__inst_arm(0xe1efc3d0) "	@ ldrd r12, [pc, #48]!")
 	TEST_UNSUPPORTED(__inst_arm(0xe0c9f3d0) "	@ ldrd pc, [r9], #48")
 	TEST_UNSUPPORTED(__inst_arm(0xe0c9e3d0) "	@ ldrd lr, [r9], #48")
+#endif
 
 	TEST_GROUP("Miscellaneous")
 
@@ -1227,7 +1233,9 @@ void kprobe_arm_test_cases(void)
 	TEST_COPROCESSOR( "mrc"two"	0, 0, r0, cr0, cr0, 0")
 
 	COPROCESSOR_INSTRUCTIONS_ST_LD("",e)
+#if __LINUX_ARM_ARCH__ >= 5
 	COPROCESSOR_INSTRUCTIONS_MC_MR("",e)
+#endif
 	TEST_UNSUPPORTED("svc	0")
 	TEST_UNSUPPORTED("svc	0xffffff")
 
@@ -1287,7 +1295,9 @@ void kprobe_arm_test_cases(void)
 	TEST(	"blx	__dummy_thumb_subroutine_odd")
 #endif /* __LINUX_ARM_ARCH__ >= 6 */
 
+#if __LINUX_ARM_ARCH__ >= 5
 	COPROCESSOR_INSTRUCTIONS_ST_LD("2",f)
+#endif
 #if __LINUX_ARM_ARCH__ >= 6
 	COPROCESSOR_INSTRUCTIONS_MC_MR("2",f)
 #endif
diff --git a/arch/arm/kernel/kprobes-test.c b/arch/arm/kernel/kprobes-test.c
index 379639998d5a..08d731294bcd 100644
--- a/arch/arm/kernel/kprobes-test.c
+++ b/arch/arm/kernel/kprobes-test.c
@@ -225,6 +225,7 @@ static int pre_handler_called;
 static int post_handler_called;
 static int jprobe_func_called;
 static int kretprobe_handler_called;
+static int tests_failed;
 
 #define FUNC_ARG1 0x12345678
 #define FUNC_ARG2 0xabcdef
@@ -461,6 +462,13 @@ static int run_api_tests(long (*func)(long, long))
 
 	pr_info("    jprobe\n");
 	ret = test_jprobe(func);
+#if defined(CONFIG_THUMB2_KERNEL) && !defined(MODULE)
+	if (ret == -EINVAL) {
+		pr_err("FAIL: Known longtime bug with jprobe on Thumb kernels\n");
+		tests_failed = ret;
+		ret = 0;
+	}
+#endif
 	if (ret < 0)
 		return ret;
 
@@ -1672,6 +1680,8 @@ static int __init run_all_tests(void)
 
 out:
 	if (ret == 0)
+		ret = tests_failed;
+	if (ret == 0)
 		pr_info("Finished kprobe tests OK\n");
 	else
 		pr_err("kprobe tests failed\n");
diff --git a/arch/arm/kernel/perf_event_v7.c b/arch/arm/kernel/perf_event_v7.c
index 2037f7205987..1d37568c547a 100644
--- a/arch/arm/kernel/perf_event_v7.c
+++ b/arch/arm/kernel/perf_event_v7.c
@@ -1924,7 +1924,7 @@ static int krait_pmu_get_event_idx(struct pmu_hw_events *cpuc,
 				   struct perf_event *event)
 {
 	int idx;
-	int bit;
+	int bit = -1;
 	unsigned int prefix;
 	unsigned int region;
 	unsigned int code;
@@ -1953,7 +1953,7 @@ static int krait_pmu_get_event_idx(struct pmu_hw_events *cpuc,
 	}
 
 	idx = armv7pmu_get_event_idx(cpuc, event);
-	if (idx < 0 && krait_event)
+	if (idx < 0 && bit >= 0)
 		clear_bit(bit, cpuc->used_mask);
 
 	return idx;
diff --git a/arch/arm/kernel/probes-arm.c b/arch/arm/kernel/probes-arm.c
index 51a13a027989..8eaef81d8344 100644
--- a/arch/arm/kernel/probes-arm.c
+++ b/arch/arm/kernel/probes-arm.c
@@ -341,12 +341,12 @@ static const union decode_item arm_cccc_000x_table[] = {
 	/* CMP (reg-shift reg)	cccc 0001 0101 xxxx xxxx xxxx 0xx1 xxxx */
 	/* CMN (reg-shift reg)	cccc 0001 0111 xxxx xxxx xxxx 0xx1 xxxx */
 	DECODE_EMULATEX	(0x0f900090, 0x01100010, PROBES_DATA_PROCESSING_REG,
-						 REGS(ANY, 0, NOPC, 0, ANY)),
+						 REGS(NOPC, 0, NOPC, 0, NOPC)),
 
 	/* MOV (reg-shift reg)	cccc 0001 101x xxxx xxxx xxxx 0xx1 xxxx */
 	/* MVN (reg-shift reg)	cccc 0001 111x xxxx xxxx xxxx 0xx1 xxxx */
 	DECODE_EMULATEX	(0x0fa00090, 0x01a00010, PROBES_DATA_PROCESSING_REG,
-						 REGS(0, ANY, NOPC, 0, ANY)),
+						 REGS(0, NOPC, NOPC, 0, NOPC)),
 
 	/* AND (reg-shift reg)	cccc 0000 000x xxxx xxxx xxxx 0xx1 xxxx */
 	/* EOR (reg-shift reg)	cccc 0000 001x xxxx xxxx xxxx 0xx1 xxxx */
@@ -359,7 +359,7 @@ static const union decode_item arm_cccc_000x_table[] = {
 	/* ORR (reg-shift reg)	cccc 0001 100x xxxx xxxx xxxx 0xx1 xxxx */
 	/* BIC (reg-shift reg)	cccc 0001 110x xxxx xxxx xxxx 0xx1 xxxx */
 	DECODE_EMULATEX	(0x0e000090, 0x00000010, PROBES_DATA_PROCESSING_REG,
-						 REGS(ANY, ANY, NOPC, 0, ANY)),
+						 REGS(NOPC, NOPC, NOPC, 0, NOPC)),
 
 	DECODE_END
 };
diff --git a/arch/arm/kernel/ptrace.c b/arch/arm/kernel/ptrace.c
index 0dd3b79b15c3..0c27ed6f3f23 100644
--- a/arch/arm/kernel/ptrace.c
+++ b/arch/arm/kernel/ptrace.c
@@ -908,7 +908,7 @@ enum ptrace_syscall_dir {
 	PTRACE_SYSCALL_EXIT,
 };
 
-static int tracehook_report_syscall(struct pt_regs *regs,
+static void tracehook_report_syscall(struct pt_regs *regs,
 				    enum ptrace_syscall_dir dir)
 {
 	unsigned long ip;
@@ -926,7 +926,6 @@ static int tracehook_report_syscall(struct pt_regs *regs,
 		current_thread_info()->syscall = -1;
 
 	regs->ARM_ip = ip;
-	return current_thread_info()->syscall;
 }
 
 asmlinkage int syscall_trace_enter(struct pt_regs *regs, int scno)
@@ -938,7 +937,9 @@ asmlinkage int syscall_trace_enter(struct pt_regs *regs, int scno)
 		return -1;
 
 	if (test_thread_flag(TIF_SYSCALL_TRACE))
-		scno = tracehook_report_syscall(regs, PTRACE_SYSCALL_ENTER);
+		tracehook_report_syscall(regs, PTRACE_SYSCALL_ENTER);
+
+	scno = current_thread_info()->syscall;
 
 	if (test_thread_flag(TIF_SYSCALL_TRACEPOINT))
 		trace_sys_enter(regs, scno);
diff --git a/arch/arm/kernel/topology.c b/arch/arm/kernel/topology.c
index 9d853189028b..e35d880f9773 100644
--- a/arch/arm/kernel/topology.c
+++ b/arch/arm/kernel/topology.c
@@ -275,7 +275,7 @@ void store_cpu_topology(unsigned int cpuid)
 		cpu_topology[cpuid].socket_id, mpidr);
 }
 
-static inline const int cpu_corepower_flags(void)
+static inline int cpu_corepower_flags(void)
 {
 	return SD_SHARE_PKG_RESOURCES  | SD_SHARE_POWERDOMAIN;
 }
diff --git a/arch/arm/mach-exynos/exynos.c b/arch/arm/mach-exynos/exynos.c
index f38cf7c110cc..46d893fcbe85 100644
--- a/arch/arm/mach-exynos/exynos.c
+++ b/arch/arm/mach-exynos/exynos.c
@@ -173,10 +173,8 @@ static struct platform_device exynos_cpuidle = {
 
 void __init exynos_cpuidle_init(void)
 {
-	if (soc_is_exynos5440())
-		return;
-
-	platform_device_register(&exynos_cpuidle);
+	if (soc_is_exynos4210() || soc_is_exynos5250())
+		platform_device_register(&exynos_cpuidle);
 }
 
 void __init exynos_cpufreq_init(void)
@@ -297,7 +295,7 @@ static void __init exynos_dt_machine_init(void)
 	 * This is called from smp_prepare_cpus if we've built for SMP, but
 	 * we still need to set it up for PM and firmware ops if not.
 	 */
-	if (!IS_ENABLED(SMP))
+	if (!IS_ENABLED(CONFIG_SMP))
 		exynos_sysram_init();
 
 	exynos_cpuidle_init();
diff --git a/arch/arm/mach-exynos/firmware.c b/arch/arm/mach-exynos/firmware.c
index eb91d2350f8c..e8797bb78871 100644
--- a/arch/arm/mach-exynos/firmware.c
+++ b/arch/arm/mach-exynos/firmware.c
@@ -57,8 +57,13 @@ static int exynos_set_cpu_boot_addr(int cpu, unsigned long boot_addr)
 
 	boot_reg = sysram_ns_base_addr + 0x1c;
 
-	if (!soc_is_exynos4212() && !soc_is_exynos3250())
-		boot_reg += 4*cpu;
+	/*
+	 * Almost all Exynos-series of SoCs that run in secure mode don't need
+	 * additional offset for every CPU, with Exynos4412 being the only
+	 * exception.
+	 */
+	if (soc_is_exynos4412())
+		boot_reg += 4 * cpu;
 
 	__raw_writel(boot_addr, boot_reg);
 	return 0;
diff --git a/arch/arm/mach-exynos/hotplug.c b/arch/arm/mach-exynos/hotplug.c
index 69fa48397394..920a4baa53cd 100644
--- a/arch/arm/mach-exynos/hotplug.c
+++ b/arch/arm/mach-exynos/hotplug.c
@@ -40,21 +40,17 @@ static inline void cpu_leave_lowpower(void)
 
 static inline void platform_do_lowpower(unsigned int cpu, int *spurious)
 {
+	u32 mpidr = cpu_logical_map(cpu);
+	u32 core_id = MPIDR_AFFINITY_LEVEL(mpidr, 0);
+
 	for (;;) {
 
-		/* make cpu1 to be turned off at next WFI command */
-		if (cpu == 1)
-			exynos_cpu_power_down(cpu);
+		/* Turn the CPU off on next WFI instruction. */
+		exynos_cpu_power_down(core_id);
 
-		/*
-		 * here's the WFI
-		 */
-		asm(".word	0xe320f003\n"
-		    :
-		    :
-		    : "memory", "cc");
+		wfi();
 
-		if (pen_release == cpu_logical_map(cpu)) {
+		if (pen_release == core_id) {
 			/*
 			 * OK, proper wakeup, we're done
 			 */
diff --git a/arch/arm/mach-exynos/mcpm-exynos.c b/arch/arm/mach-exynos/mcpm-exynos.c
index 0498d0b887ef..ace0ed617476 100644
--- a/arch/arm/mach-exynos/mcpm-exynos.c
+++ b/arch/arm/mach-exynos/mcpm-exynos.c
@@ -25,7 +25,6 @@
 
 #define EXYNOS5420_CPUS_PER_CLUSTER	4
 #define EXYNOS5420_NR_CLUSTERS		2
-#define MCPM_BOOT_ADDR_OFFSET		0x1c
 
 /*
  * The common v7_exit_coherency_flush API could not be used because of the
@@ -343,11 +342,13 @@ static int __init exynos_mcpm_init(void)
 	pr_info("Exynos MCPM support installed\n");
 
 	/*
-	 * Future entries into the kernel can now go
-	 * through the cluster entry vectors.
+	 * U-Boot SPL is hardcoded to jump to the start of ns_sram_base_addr
+	 * as part of secondary_cpu_start().  Let's redirect it to the
+	 * mcpm_entry_point().
 	 */
-	__raw_writel(virt_to_phys(mcpm_entry_point),
-			ns_sram_base_addr + MCPM_BOOT_ADDR_OFFSET);
+	__raw_writel(0xe59f0000, ns_sram_base_addr);     /* ldr r0, [pc, #0] */
+	__raw_writel(0xe12fff10, ns_sram_base_addr + 4); /* bx  r0 */
+	__raw_writel(virt_to_phys(mcpm_entry_point), ns_sram_base_addr + 8);
 
 	iounmap(ns_sram_base_addr);
 
diff --git a/arch/arm/mach-exynos/platsmp.c b/arch/arm/mach-exynos/platsmp.c
index 1c8d31e39520..50b9aad5e27b 100644
--- a/arch/arm/mach-exynos/platsmp.c
+++ b/arch/arm/mach-exynos/platsmp.c
@@ -90,7 +90,8 @@ static void exynos_secondary_init(unsigned int cpu)
 static int exynos_boot_secondary(unsigned int cpu, struct task_struct *idle)
 {
 	unsigned long timeout;
-	unsigned long phys_cpu = cpu_logical_map(cpu);
+	u32 mpidr = cpu_logical_map(cpu);
+	u32 core_id = MPIDR_AFFINITY_LEVEL(mpidr, 0);
 	int ret = -ENOSYS;
 
 	/*
@@ -104,17 +105,18 @@ static int exynos_boot_secondary(unsigned int cpu, struct task_struct *idle)
 	 * the holding pen - release it, then wait for it to flag
 	 * that it has been released by resetting pen_release.
 	 *
-	 * Note that "pen_release" is the hardware CPU ID, whereas
+	 * Note that "pen_release" is the hardware CPU core ID, whereas
 	 * "cpu" is Linux's internal ID.
 	 */
-	write_pen_release(phys_cpu);
+	write_pen_release(core_id);
 
-	if (!exynos_cpu_power_state(cpu)) {
-		exynos_cpu_power_up(cpu);
+	if (!exynos_cpu_power_state(core_id)) {
+		exynos_cpu_power_up(core_id);
 		timeout = 10;
 
 		/* wait max 10 ms until cpu1 is on */
-		while (exynos_cpu_power_state(cpu) != S5P_CORE_LOCAL_PWR_EN) {
+		while (exynos_cpu_power_state(core_id)
+		       != S5P_CORE_LOCAL_PWR_EN) {
 			if (timeout-- == 0)
 				break;
 
@@ -145,20 +147,20 @@ static int exynos_boot_secondary(unsigned int cpu, struct task_struct *idle)
 		 * Try to set boot address using firmware first
 		 * and fall back to boot register if it fails.
 		 */
-		ret = call_firmware_op(set_cpu_boot_addr, phys_cpu, boot_addr);
+		ret = call_firmware_op(set_cpu_boot_addr, core_id, boot_addr);
 		if (ret && ret != -ENOSYS)
 			goto fail;
 		if (ret == -ENOSYS) {
-			void __iomem *boot_reg = cpu_boot_reg(phys_cpu);
+			void __iomem *boot_reg = cpu_boot_reg(core_id);
 
 			if (IS_ERR(boot_reg)) {
 				ret = PTR_ERR(boot_reg);
 				goto fail;
 			}
-			__raw_writel(boot_addr, cpu_boot_reg(phys_cpu));
+			__raw_writel(boot_addr, cpu_boot_reg(core_id));
 		}
 
-		call_firmware_op(cpu_boot, phys_cpu);
+		call_firmware_op(cpu_boot, core_id);
 
 		arch_send_wakeup_ipi_mask(cpumask_of(cpu));
 
@@ -227,22 +229,24 @@ static void __init exynos_smp_prepare_cpus(unsigned int max_cpus)
 	 * boot register if it fails.
 	 */
 	for (i = 1; i < max_cpus; ++i) {
-		unsigned long phys_cpu;
 		unsigned long boot_addr;
+		u32 mpidr;
+		u32 core_id;
 		int ret;
 
-		phys_cpu = cpu_logical_map(i);
+		mpidr = cpu_logical_map(i);
+		core_id = MPIDR_AFFINITY_LEVEL(mpidr, 0);
 		boot_addr = virt_to_phys(exynos4_secondary_startup);
 
-		ret = call_firmware_op(set_cpu_boot_addr, phys_cpu, boot_addr);
+		ret = call_firmware_op(set_cpu_boot_addr, core_id, boot_addr);
 		if (ret && ret != -ENOSYS)
 			break;
 		if (ret == -ENOSYS) {
-			void __iomem *boot_reg = cpu_boot_reg(phys_cpu);
+			void __iomem *boot_reg = cpu_boot_reg(core_id);
 
 			if (IS_ERR(boot_reg))
 				break;
-			__raw_writel(boot_addr, cpu_boot_reg(phys_cpu));
+			__raw_writel(boot_addr, cpu_boot_reg(core_id));
 		}
 	}
 }
diff --git a/arch/arm/mach-exynos/pm.c b/arch/arm/mach-exynos/pm.c
index 87c0d34c7fba..202ca73e49c4 100644
--- a/arch/arm/mach-exynos/pm.c
+++ b/arch/arm/mach-exynos/pm.c
@@ -300,7 +300,7 @@ static int exynos_pm_suspend(void)
 	tmp = (S5P_USE_STANDBY_WFI0 | S5P_USE_STANDBY_WFE0);
 	__raw_writel(tmp, S5P_CENTRAL_SEQ_OPTION);
 
-	if (!soc_is_exynos5250())
+	if (read_cpuid_part_number() == ARM_CPU_PART_CORTEX_A9)
 		exynos_cpu_save_register();
 
 	return 0;
@@ -334,7 +334,7 @@ static void exynos_pm_resume(void)
 	if (exynos_pm_central_resume())
 		goto early_wakeup;
 
-	if (!soc_is_exynos5250())
+	if (read_cpuid_part_number() == ARM_CPU_PART_CORTEX_A9)
 		exynos_cpu_restore_register();
 
 	/* For release retention */
@@ -353,7 +353,7 @@ static void exynos_pm_resume(void)
 
 	s3c_pm_do_restore_core(exynos_core_save, ARRAY_SIZE(exynos_core_save));
 
-	if (!soc_is_exynos5250())
+	if (read_cpuid_part_number() == ARM_CPU_PART_CORTEX_A9)
 		scu_enable(S5P_VA_SCU);
 
 early_wakeup:
@@ -440,15 +440,18 @@ static int exynos_cpu_pm_notifier(struct notifier_block *self,
 	case CPU_PM_ENTER:
 		if (cpu == 0) {
 			exynos_pm_central_suspend();
-			exynos_cpu_save_register();
+			if (read_cpuid_part_number() == ARM_CPU_PART_CORTEX_A9)
+				exynos_cpu_save_register();
 		}
 		break;
 
 	case CPU_PM_EXIT:
 		if (cpu == 0) {
-			if (!soc_is_exynos5250())
+			if (read_cpuid_part_number() ==
+					ARM_CPU_PART_CORTEX_A9) {
 				scu_enable(S5P_VA_SCU);
-			exynos_cpu_restore_register();
+				exynos_cpu_restore_register();
+			}
 			exynos_pm_central_resume();
 		}
 		break;
diff --git a/arch/arm/mach-exynos/pm_domains.c b/arch/arm/mach-exynos/pm_domains.c
index fe6570ebbdde..797cb134bfff 100644
--- a/arch/arm/mach-exynos/pm_domains.c
+++ b/arch/arm/mach-exynos/pm_domains.c
@@ -17,6 +17,7 @@
 #include <linux/err.h>
 #include <linux/slab.h>
 #include <linux/pm_domain.h>
+#include <linux/clk.h>
 #include <linux/delay.h>
 #include <linux/of_address.h>
 #include <linux/of_platform.h>
@@ -24,6 +25,8 @@
 
 #include "regs-pmu.h"
 
+#define MAX_CLK_PER_DOMAIN	4
+
 /*
  * Exynos specific wrapper around the generic power domain
  */
@@ -32,6 +35,9 @@ struct exynos_pm_domain {
 	char const *name;
 	bool is_off;
 	struct generic_pm_domain pd;
+	struct clk *oscclk;
+	struct clk *clk[MAX_CLK_PER_DOMAIN];
+	struct clk *pclk[MAX_CLK_PER_DOMAIN];
 };
 
 static int exynos_pd_power(struct generic_pm_domain *domain, bool power_on)
@@ -44,6 +50,19 @@ static int exynos_pd_power(struct generic_pm_domain *domain, bool power_on)
 	pd = container_of(domain, struct exynos_pm_domain, pd);
 	base = pd->base;
 
+	/* Set oscclk before powering off a domain*/
+	if (!power_on) {
+		int i;
+
+		for (i = 0; i < MAX_CLK_PER_DOMAIN; i++) {
+			if (IS_ERR(pd->clk[i]))
+				break;
+			if (clk_set_parent(pd->clk[i], pd->oscclk))
+				pr_err("%s: error setting oscclk as parent to clock %d\n",
+						pd->name, i);
+		}
+	}
+
 	pwr = power_on ? S5P_INT_LOCAL_PWR_EN : 0;
 	__raw_writel(pwr, base);
 
@@ -60,6 +79,20 @@ static int exynos_pd_power(struct generic_pm_domain *domain, bool power_on)
 		cpu_relax();
 		usleep_range(80, 100);
 	}
+
+	/* Restore clocks after powering on a domain*/
+	if (power_on) {
+		int i;
+
+		for (i = 0; i < MAX_CLK_PER_DOMAIN; i++) {
+			if (IS_ERR(pd->clk[i]))
+				break;
+			if (clk_set_parent(pd->clk[i], pd->pclk[i]))
+				pr_err("%s: error setting parent to clock%d\n",
+						pd->name, i);
+		}
+	}
+
 	return 0;
 }
 
@@ -152,9 +185,11 @@ static __init int exynos4_pm_init_power_domain(void)
 
 	for_each_compatible_node(np, NULL, "samsung,exynos4210-pd") {
 		struct exynos_pm_domain *pd;
-		int on;
+		int on, i;
+		struct device *dev;
 
 		pdev = of_find_device_by_node(np);
+		dev = &pdev->dev;
 
 		pd = kzalloc(sizeof(*pd), GFP_KERNEL);
 		if (!pd) {
@@ -170,6 +205,30 @@ static __init int exynos4_pm_init_power_domain(void)
 		pd->pd.power_on = exynos_pd_power_on;
 		pd->pd.of_node = np;
 
+		pd->oscclk = clk_get(dev, "oscclk");
+		if (IS_ERR(pd->oscclk))
+			goto no_clk;
+
+		for (i = 0; i < MAX_CLK_PER_DOMAIN; i++) {
+			char clk_name[8];
+
+			snprintf(clk_name, sizeof(clk_name), "clk%d", i);
+			pd->clk[i] = clk_get(dev, clk_name);
+			if (IS_ERR(pd->clk[i]))
+				break;
+			snprintf(clk_name, sizeof(clk_name), "pclk%d", i);
+			pd->pclk[i] = clk_get(dev, clk_name);
+			if (IS_ERR(pd->pclk[i])) {
+				clk_put(pd->clk[i]);
+				pd->clk[i] = ERR_PTR(-EINVAL);
+				break;
+			}
+		}
+
+		if (IS_ERR(pd->clk[0]))
+			clk_put(pd->oscclk);
+
+no_clk:
 		platform_set_drvdata(pdev, pd);
 
 		on = __raw_readl(pd->base + 0x4) & S5P_INT_LOCAL_PWR_EN;
diff --git a/arch/arm/mach-imx/Kconfig b/arch/arm/mach-imx/Kconfig
index 28fa2fa49e5d..4b5185748f74 100644
--- a/arch/arm/mach-imx/Kconfig
+++ b/arch/arm/mach-imx/Kconfig
@@ -734,9 +734,9 @@ config SOC_IMX6
 	select HAVE_IMX_MMDC
 	select HAVE_IMX_SRC
 	select MFD_SYSCON
-	select PL310_ERRATA_588369 if CACHE_PL310
-	select PL310_ERRATA_727915 if CACHE_PL310
-	select PL310_ERRATA_769419 if CACHE_PL310
+	select PL310_ERRATA_588369 if CACHE_L2X0
+	select PL310_ERRATA_727915 if CACHE_L2X0
+	select PL310_ERRATA_769419 if CACHE_L2X0
 
 config SOC_IMX6Q
 	bool "i.MX6 Quad/DualLite support"
@@ -771,9 +771,9 @@ config SOC_VF610
 	select ARM_GIC
 	select PINCTRL_VF610
 	select VF_PIT_TIMER
-	select PL310_ERRATA_588369 if CACHE_PL310
-	select PL310_ERRATA_727915 if CACHE_PL310
-	select PL310_ERRATA_769419 if CACHE_PL310
+	select PL310_ERRATA_588369 if CACHE_L2X0
+	select PL310_ERRATA_727915 if CACHE_L2X0
+	select PL310_ERRATA_769419 if CACHE_L2X0
 
 	help
 	  This enable support for Freescale Vybrid VF610 processor.
diff --git a/arch/arm/mach-imx/clk-gate2.c b/arch/arm/mach-imx/clk-gate2.c
index 4ba587da89d2..84acdfd1d715 100644
--- a/arch/arm/mach-imx/clk-gate2.c
+++ b/arch/arm/mach-imx/clk-gate2.c
@@ -67,8 +67,12 @@ static void clk_gate2_disable(struct clk_hw *hw)
 
 	spin_lock_irqsave(gate->lock, flags);
 
-	if (gate->share_count && --(*gate->share_count) > 0)
-		goto out;
+	if (gate->share_count) {
+		if (WARN_ON(*gate->share_count == 0))
+			goto out;
+		else if (--(*gate->share_count) > 0)
+			goto out;
+	}
 
 	reg = readl(gate->reg);
 	reg &= ~(3 << gate->bit_idx);
@@ -78,19 +82,26 @@ out:
 	spin_unlock_irqrestore(gate->lock, flags);
 }
 
-static int clk_gate2_is_enabled(struct clk_hw *hw)
+static int clk_gate2_reg_is_enabled(void __iomem *reg, u8 bit_idx)
 {
-	u32 reg;
-	struct clk_gate2 *gate = to_clk_gate2(hw);
+	u32 val = readl(reg);
 
-	reg = readl(gate->reg);
-
-	if (((reg >> gate->bit_idx) & 1) == 1)
+	if (((val >> bit_idx) & 1) == 1)
 		return 1;
 
 	return 0;
 }
 
+static int clk_gate2_is_enabled(struct clk_hw *hw)
+{
+	struct clk_gate2 *gate = to_clk_gate2(hw);
+
+	if (gate->share_count)
+		return !!(*gate->share_count);
+	else
+		return clk_gate2_reg_is_enabled(gate->reg, gate->bit_idx);
+}
+
 static struct clk_ops clk_gate2_ops = {
 	.enable = clk_gate2_enable,
 	.disable = clk_gate2_disable,
@@ -116,6 +127,10 @@ struct clk *clk_register_gate2(struct device *dev, const char *name,
 	gate->bit_idx = bit_idx;
 	gate->flags = clk_gate2_flags;
 	gate->lock = lock;
+
+	/* Initialize share_count per hardware state */
+	if (share_count)
+		*share_count = clk_gate2_reg_is_enabled(reg, bit_idx) ? 1 : 0;
 	gate->share_count = share_count;
 
 	init.name = name;
diff --git a/arch/arm/mach-imx/clk-imx6q.c b/arch/arm/mach-imx/clk-imx6q.c
index 8e795dea02ec..8556c787e59c 100644
--- a/arch/arm/mach-imx/clk-imx6q.c
+++ b/arch/arm/mach-imx/clk-imx6q.c
@@ -70,7 +70,7 @@ static const char *cko_sels[] = { "cko1", "cko2", };
 static const char *lvds_sels[] = {
 	"dummy", "dummy", "dummy", "dummy", "dummy", "dummy",
 	"pll4_audio", "pll5_video", "pll8_mlb", "enet_ref",
-	"pcie_ref", "sata_ref",
+	"pcie_ref_125m", "sata_ref_100m",
 };
 
 enum mx6q_clks {
@@ -491,7 +491,7 @@ static void __init imx6q_clocks_init(struct device_node *ccm_node)
 
 	/* All existing boards with PCIe use LVDS1 */
 	if (IS_ENABLED(CONFIG_PCI_IMX6))
-		clk_set_parent(clk[lvds1_sel], clk[sata_ref]);
+		clk_set_parent(clk[lvds1_sel], clk[sata_ref_100m]);
 
 	/* Set initial power mode */
 	imx6q_set_lpm(WAIT_CLOCKED);
diff --git a/arch/arm/mach-imx/clk-imx6sl.c b/arch/arm/mach-imx/clk-imx6sl.c
index 21cf06cebade..5408ca70c8d6 100644
--- a/arch/arm/mach-imx/clk-imx6sl.c
+++ b/arch/arm/mach-imx/clk-imx6sl.c
@@ -312,6 +312,7 @@ static void __init imx6sl_clocks_init(struct device_node *ccm_node)
 	clks[IMX6SL_CLK_ECSPI2]       = imx_clk_gate2("ecspi2",       "ecspi_root",        base + 0x6c, 2);
 	clks[IMX6SL_CLK_ECSPI3]       = imx_clk_gate2("ecspi3",       "ecspi_root",        base + 0x6c, 4);
 	clks[IMX6SL_CLK_ECSPI4]       = imx_clk_gate2("ecspi4",       "ecspi_root",        base + 0x6c, 6);
+	clks[IMX6SL_CLK_ENET]         = imx_clk_gate2("enet",         "ipg",               base + 0x6c, 10);
 	clks[IMX6SL_CLK_EPIT1]        = imx_clk_gate2("epit1",        "perclk",            base + 0x6c, 12);
 	clks[IMX6SL_CLK_EPIT2]        = imx_clk_gate2("epit2",        "perclk",            base + 0x6c, 14);
 	clks[IMX6SL_CLK_EXTERN_AUDIO] = imx_clk_gate2("extern_audio", "extern_audio_podf", base + 0x6c, 16);
diff --git a/arch/arm/mach-integrator/integrator_ap.c b/arch/arm/mach-integrator/integrator_ap.c
index dd0cc677d596..660ca6feff40 100644
--- a/arch/arm/mach-integrator/integrator_ap.c
+++ b/arch/arm/mach-integrator/integrator_ap.c
@@ -480,25 +480,18 @@ static const struct of_device_id ebi_match[] = {
 static void __init ap_init_of(void)
 {
 	unsigned long sc_dec;
-	struct device_node *root;
 	struct device_node *syscon;
 	struct device_node *ebi;
 	struct device *parent;
 	struct soc_device *soc_dev;
 	struct soc_device_attribute *soc_dev_attr;
 	u32 ap_sc_id;
-	int err;
 	int i;
 
-	/* Here we create an SoC device for the root node */
-	root = of_find_node_by_path("/");
-	if (!root)
-		return;
-
-	syscon = of_find_matching_node(root, ap_syscon_match);
+	syscon = of_find_matching_node(NULL, ap_syscon_match);
 	if (!syscon)
 		return;
-	ebi = of_find_matching_node(root, ebi_match);
+	ebi = of_find_matching_node(NULL, ebi_match);
 	if (!ebi)
 		return;
 
@@ -509,19 +502,17 @@ static void __init ap_init_of(void)
 	if (!ebi_base)
 		return;
 
+	of_platform_populate(NULL, of_default_bus_match_table,
+			ap_auxdata_lookup, NULL);
+
 	ap_sc_id = readl(ap_syscon_base);
 
 	soc_dev_attr = kzalloc(sizeof(*soc_dev_attr), GFP_KERNEL);
 	if (!soc_dev_attr)
 		return;
 
-	err = of_property_read_string(root, "compatible",
-				      &soc_dev_attr->soc_id);
-	if (err)
-		return;
-	err = of_property_read_string(root, "model", &soc_dev_attr->machine);
-	if (err)
-		return;
+	soc_dev_attr->soc_id = "XVC";
+	soc_dev_attr->machine = "Integrator/AP";
 	soc_dev_attr->family = "Integrator";
 	soc_dev_attr->revision = kasprintf(GFP_KERNEL, "%c",
 					   'A' + (ap_sc_id & 0x0f));
@@ -536,9 +527,6 @@ static void __init ap_init_of(void)
 	parent = soc_device_to_device(soc_dev);
 	integrator_init_sysfs(parent, ap_sc_id);
 
-	of_platform_populate(root, of_default_bus_match_table,
-			ap_auxdata_lookup, parent);
-
 	sc_dec = readl(ap_syscon_base + INTEGRATOR_SC_DEC_OFFSET);
 	for (i = 0; i < 4; i++) {
 		struct lm_device *lmdev;
diff --git a/arch/arm/mach-integrator/integrator_cp.c b/arch/arm/mach-integrator/integrator_cp.c
index a938242b0c95..0e57f8f820a5 100644
--- a/arch/arm/mach-integrator/integrator_cp.c
+++ b/arch/arm/mach-integrator/integrator_cp.c
@@ -279,20 +279,13 @@ static const struct of_device_id intcp_syscon_match[] = {
 
 static void __init intcp_init_of(void)
 {
-	struct device_node *root;
 	struct device_node *cpcon;
 	struct device *parent;
 	struct soc_device *soc_dev;
 	struct soc_device_attribute *soc_dev_attr;
 	u32 intcp_sc_id;
-	int err;
 
-	/* Here we create an SoC device for the root node */
-	root = of_find_node_by_path("/");
-	if (!root)
-		return;
-
-	cpcon = of_find_matching_node(root, intcp_syscon_match);
+	cpcon = of_find_matching_node(NULL, intcp_syscon_match);
 	if (!cpcon)
 		return;
 
@@ -300,19 +293,17 @@ static void __init intcp_init_of(void)
 	if (!intcp_con_base)
 		return;
 
+	of_platform_populate(NULL, of_default_bus_match_table,
+			     intcp_auxdata_lookup, NULL);
+
 	intcp_sc_id = readl(intcp_con_base);
 
 	soc_dev_attr = kzalloc(sizeof(*soc_dev_attr), GFP_KERNEL);
 	if (!soc_dev_attr)
 		return;
 
-	err = of_property_read_string(root, "compatible",
-				      &soc_dev_attr->soc_id);
-	if (err)
-		return;
-	err = of_property_read_string(root, "model", &soc_dev_attr->machine);
-	if (err)
-		return;
+	soc_dev_attr->soc_id = "XCV";
+	soc_dev_attr->machine = "Integrator/CP";
 	soc_dev_attr->family = "Integrator";
 	soc_dev_attr->revision = kasprintf(GFP_KERNEL, "%c",
 					   'A' + (intcp_sc_id & 0x0f));
@@ -326,8 +317,6 @@ static void __init intcp_init_of(void)
 
 	parent = soc_device_to_device(soc_dev);
 	integrator_init_sysfs(parent, intcp_sc_id);
-	of_platform_populate(root, of_default_bus_match_table,
-			intcp_auxdata_lookup, parent);
 }
 
 static const char * intcp_dt_board_compat[] = {
diff --git a/arch/arm/mach-mvebu/Kconfig b/arch/arm/mach-mvebu/Kconfig
index 4a7c250c9a30..b9bc599a5fd0 100644
--- a/arch/arm/mach-mvebu/Kconfig
+++ b/arch/arm/mach-mvebu/Kconfig
@@ -10,6 +10,7 @@ menuconfig ARCH_MVEBU
 	select ZONE_DMA if ARM_LPAE
 	select ARCH_REQUIRE_GPIOLIB
 	select PCI_QUIRKS if PCI
+	select OF_ADDRESS_PCI
 
 if ARCH_MVEBU
 
@@ -17,6 +18,7 @@ config MACH_MVEBU_V7
 	bool
 	select ARMADA_370_XP_TIMER
 	select CACHE_L2X0
+	select ARM_CPU_SUSPEND
 
 config MACH_ARMADA_370
 	bool "Marvell Armada 370 boards" if ARCH_MULTI_V7
diff --git a/arch/arm/mach-mvebu/Makefile b/arch/arm/mach-mvebu/Makefile
index 2ecb828e4a8b..1636cdbef01a 100644
--- a/arch/arm/mach-mvebu/Makefile
+++ b/arch/arm/mach-mvebu/Makefile
@@ -7,7 +7,7 @@ CFLAGS_pmsu.o			:= -march=armv7-a
 obj-y				 += system-controller.o mvebu-soc-id.o
 
 ifeq ($(CONFIG_MACH_MVEBU_V7),y)
-obj-y				 += cpu-reset.o board-v7.o coherency.o coherency_ll.o pmsu.o
+obj-y				 += cpu-reset.o board-v7.o coherency.o coherency_ll.o pmsu.o pmsu_ll.o
 obj-$(CONFIG_SMP)		 += platsmp.o headsmp.o platsmp-a9.o headsmp-a9.o
 obj-$(CONFIG_HOTPLUG_CPU)	 += hotplug.o
 endif
diff --git a/arch/arm/mach-mvebu/board-v7.c b/arch/arm/mach-mvebu/board-v7.c
index 8bb742fdf5ca..b2524d689f21 100644
--- a/arch/arm/mach-mvebu/board-v7.c
+++ b/arch/arm/mach-mvebu/board-v7.c
@@ -23,6 +23,7 @@
 #include <linux/mbus.h>
 #include <linux/signal.h>
 #include <linux/slab.h>
+#include <linux/irqchip.h>
 #include <asm/hardware/cache-l2x0.h>
 #include <asm/mach/arch.h>
 #include <asm/mach/map.h>
@@ -71,17 +72,23 @@ static int armada_375_external_abort_wa(unsigned long addr, unsigned int fsr,
 	return 1;
 }
 
-static void __init mvebu_timer_and_clk_init(void)
+static void __init mvebu_init_irq(void)
 {
-	of_clk_init(NULL);
-	clocksource_of_init();
+	irqchip_init();
 	mvebu_scu_enable();
 	coherency_init();
 	BUG_ON(mvebu_mbus_dt_init(coherency_available()));
+}
+
+static void __init external_abort_quirk(void)
+{
+	u32 dev, rev;
 
-	if (of_machine_is_compatible("marvell,armada375"))
-		hook_fault_code(16 + 6, armada_375_external_abort_wa, SIGBUS, 0,
-				"imprecise external abort");
+	if (mvebu_get_soc_id(&dev, &rev) == 0 && rev > ARMADA_375_Z1_REV)
+		return;
+
+	hook_fault_code(16 + 6, armada_375_external_abort_wa, SIGBUS, 0,
+			"imprecise external abort");
 }
 
 static void __init i2c_quirk(void)
@@ -169,8 +176,10 @@ static void __init mvebu_dt_init(void)
 {
 	if (of_machine_is_compatible("plathome,openblocks-ax3-4"))
 		i2c_quirk();
-	if (of_machine_is_compatible("marvell,a375-db"))
+	if (of_machine_is_compatible("marvell,a375-db")) {
+		external_abort_quirk();
 		thermal_quirk();
+	}
 
 	of_platform_populate(NULL, of_default_bus_match_table, NULL, NULL);
 }
@@ -185,7 +194,7 @@ DT_MACHINE_START(ARMADA_370_XP_DT, "Marvell Armada 370/XP (Device Tree)")
 	.l2c_aux_mask	= ~0,
 	.smp		= smp_ops(armada_xp_smp_ops),
 	.init_machine	= mvebu_dt_init,
-	.init_time	= mvebu_timer_and_clk_init,
+	.init_irq       = mvebu_init_irq,
 	.restart	= mvebu_restart,
 	.dt_compat	= armada_370_xp_dt_compat,
 MACHINE_END
@@ -198,7 +207,7 @@ static const char * const armada_375_dt_compat[] = {
 DT_MACHINE_START(ARMADA_375_DT, "Marvell Armada 375 (Device Tree)")
 	.l2c_aux_val	= 0,
 	.l2c_aux_mask	= ~0,
-	.init_time	= mvebu_timer_and_clk_init,
+	.init_irq       = mvebu_init_irq,
 	.init_machine	= mvebu_dt_init,
 	.restart	= mvebu_restart,
 	.dt_compat	= armada_375_dt_compat,
@@ -213,7 +222,7 @@ static const char * const armada_38x_dt_compat[] = {
 DT_MACHINE_START(ARMADA_38X_DT, "Marvell Armada 380/385 (Device Tree)")
 	.l2c_aux_val	= 0,
 	.l2c_aux_mask	= ~0,
-	.init_time	= mvebu_timer_and_clk_init,
+	.init_irq       = mvebu_init_irq,
 	.restart	= mvebu_restart,
 	.dt_compat	= armada_38x_dt_compat,
 MACHINE_END
diff --git a/arch/arm/mach-mvebu/coherency.c b/arch/arm/mach-mvebu/coherency.c
index 477202fd39cc..2bdc3233abe2 100644
--- a/arch/arm/mach-mvebu/coherency.c
+++ b/arch/arm/mach-mvebu/coherency.c
@@ -292,6 +292,10 @@ static struct notifier_block mvebu_hwcc_nb = {
 	.notifier_call = mvebu_hwcc_notifier,
 };
 
+static struct notifier_block mvebu_hwcc_pci_nb = {
+	.notifier_call = mvebu_hwcc_notifier,
+};
+
 static void __init armada_370_coherency_init(struct device_node *np)
 {
 	struct resource res;
@@ -427,7 +431,7 @@ static int __init coherency_pci_init(void)
 {
 	if (coherency_available())
 		bus_register_notifier(&pci_bus_type,
-				       &mvebu_hwcc_nb);
+				       &mvebu_hwcc_pci_nb);
 	return 0;
 }
 
diff --git a/arch/arm/mach-mvebu/headsmp-a9.S b/arch/arm/mach-mvebu/headsmp-a9.S
index 5925366bc03c..da5bb292b91c 100644
--- a/arch/arm/mach-mvebu/headsmp-a9.S
+++ b/arch/arm/mach-mvebu/headsmp-a9.S
@@ -15,6 +15,8 @@
 #include <linux/linkage.h>
 #include <linux/init.h>
 
+#include <asm/assembler.h>
+
 	__CPUINIT
 #define CPU_RESUME_ADDR_REG 0xf10182d4
 
@@ -22,13 +24,18 @@
 .global armada_375_smp_cpu1_enable_code_end
 
 armada_375_smp_cpu1_enable_code_start:
-	ldr     r0, [pc, #4]
+ARM_BE8(setend	be)
+	adr     r0, 1f
+	ldr	r0, [r0]
 	ldr     r1, [r0]
+ARM_BE8(rev	r1, r1)
 	mov     pc, r1
+1:
 	.word   CPU_RESUME_ADDR_REG
 armada_375_smp_cpu1_enable_code_end:
 
 ENTRY(mvebu_cortex_a9_secondary_startup)
+ARM_BE8(setend	be)
 	bl      v7_invalidate_l1
 	b	secondary_startup
 ENDPROC(mvebu_cortex_a9_secondary_startup)
diff --git a/arch/arm/mach-mvebu/pmsu.c b/arch/arm/mach-mvebu/pmsu.c
index 53a55c8520bf..25aa8237d668 100644
--- a/arch/arm/mach-mvebu/pmsu.c
+++ b/arch/arm/mach-mvebu/pmsu.c
@@ -66,6 +66,8 @@ static void __iomem *pmsu_mp_base;
 extern void ll_disable_coherency(void);
 extern void ll_enable_coherency(void);
 
+extern void armada_370_xp_cpu_resume(void);
+
 static struct platform_device armada_xp_cpuidle_device = {
 	.name = "cpuidle-armada-370-xp",
 };
@@ -140,13 +142,6 @@ static void armada_370_xp_pmsu_enable_l2_powerdown_onidle(void)
 	writel(reg, pmsu_mp_base + L2C_NFABRIC_PM_CTL);
 }
 
-static void armada_370_xp_cpu_resume(void)
-{
-	asm volatile("bl    ll_add_cpu_to_smp_group\n\t"
-		     "bl    ll_enable_coherency\n\t"
-		     "b	    cpu_resume\n\t");
-}
-
 /* No locking is needed because we only access per-CPU registers */
 void armada_370_xp_pmsu_idle_prepare(bool deepidle)
 {
@@ -206,12 +201,12 @@ static noinline int do_armada_370_xp_cpu_suspend(unsigned long deepidle)
 
 	/* Test the CR_C bit and set it if it was cleared */
 	asm volatile(
-	"mrc	p15, 0, %0, c1, c0, 0 \n\t"
-	"tst	%0, #(1 << 2) \n\t"
-	"orreq	%0, %0, #(1 << 2) \n\t"
-	"mcreq	p15, 0, %0, c1, c0, 0 \n\t"
+	"mrc	p15, 0, r0, c1, c0, 0 \n\t"
+	"tst	r0, #(1 << 2) \n\t"
+	"orreq	r0, r0, #(1 << 2) \n\t"
+	"mcreq	p15, 0, r0, c1, c0, 0 \n\t"
 	"isb	"
-	: : "r" (0));
+	: : : "r0");
 
 	pr_warn("Failed to suspend the system\n");
 
diff --git a/arch/arm/mach-mvebu/pmsu_ll.S b/arch/arm/mach-mvebu/pmsu_ll.S
new file mode 100644
index 000000000000..fc3de68d8c54
--- /dev/null
+++ b/arch/arm/mach-mvebu/pmsu_ll.S
@@ -0,0 +1,25 @@
+/*
+ * Copyright (C) 2014 Marvell
+ *
+ * Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
+ * Gregory Clement <gregory.clement@free-electrons.com>
+ *
+ * This file is licensed under the terms of the GNU General Public
+ * License version 2.  This program is licensed "as is" without any
+ * warranty of any kind, whether express or implied.
+ */
+
+#include <linux/linkage.h>
+#include <asm/assembler.h>
+
+/*
+ * This is the entry point through which CPUs exiting cpuidle deep
+ * idle state are going.
+ */
+ENTRY(armada_370_xp_cpu_resume)
+ARM_BE8(setend	be )			@ go BE8 if entered LE
+	bl	ll_add_cpu_to_smp_group
+	bl	ll_enable_coherency
+	b	cpu_resume
+ENDPROC(armada_370_xp_cpu_resume)
+
diff --git a/arch/arm/mach-omap2/Kconfig b/arch/arm/mach-omap2/Kconfig
index 062505345c95..1c1ed737f7ab 100644
--- a/arch/arm/mach-omap2/Kconfig
+++ b/arch/arm/mach-omap2/Kconfig
@@ -34,8 +34,8 @@ config ARCH_OMAP4
 	select HAVE_ARM_SCU if SMP
 	select HAVE_ARM_TWD if SMP
 	select OMAP_INTERCONNECT
-	select PL310_ERRATA_588369
-	select PL310_ERRATA_727915
+	select PL310_ERRATA_588369 if CACHE_L2X0
+	select PL310_ERRATA_727915 if CACHE_L2X0
 	select PM_OPP if PM
 	select PM_RUNTIME if CPU_IDLE
 	select ARM_ERRATA_754322
diff --git a/arch/arm/mach-omap2/Makefile b/arch/arm/mach-omap2/Makefile
index 8421f38cf445..8ca99e9321e3 100644
--- a/arch/arm/mach-omap2/Makefile
+++ b/arch/arm/mach-omap2/Makefile
@@ -110,14 +110,16 @@ obj-y					+= prm_common.o cm_common.o
 obj-$(CONFIG_ARCH_OMAP2)		+= prm2xxx_3xxx.o prm2xxx.o cm2xxx.o
 obj-$(CONFIG_ARCH_OMAP3)		+= prm2xxx_3xxx.o prm3xxx.o cm3xxx.o
 obj-$(CONFIG_ARCH_OMAP3)		+= vc3xxx_data.o vp3xxx_data.o
-obj-$(CONFIG_SOC_AM33XX)		+= prm33xx.o cm33xx.o
 omap-prcm-4-5-common			=  cminst44xx.o cm44xx.o prm44xx.o \
 					   prcm_mpu44xx.o prminst44xx.o \
 					   vc44xx_data.o vp44xx_data.o
 obj-$(CONFIG_ARCH_OMAP4)		+= $(omap-prcm-4-5-common)
 obj-$(CONFIG_SOC_OMAP5)			+= $(omap-prcm-4-5-common)
 obj-$(CONFIG_SOC_DRA7XX)		+= $(omap-prcm-4-5-common)
-obj-$(CONFIG_SOC_AM43XX)		+= $(omap-prcm-4-5-common)
+am33xx-43xx-prcm-common			+= prm33xx.o cm33xx.o
+obj-$(CONFIG_SOC_AM33XX)		+= $(am33xx-43xx-prcm-common)
+obj-$(CONFIG_SOC_AM43XX)		+= $(omap-prcm-4-5-common) \
+					   $(am33xx-43xx-prcm-common)
 
 # OMAP voltage domains
 voltagedomain-common			:= voltage.o vc.o vp.o
diff --git a/arch/arm/mach-omap2/clkt_dpll.c b/arch/arm/mach-omap2/clkt_dpll.c
index 332af927f4d3..67fd26a18441 100644
--- a/arch/arm/mach-omap2/clkt_dpll.c
+++ b/arch/arm/mach-omap2/clkt_dpll.c
@@ -76,7 +76,7 @@
  * (assuming that it is counting N upwards), or -2 if the enclosing loop
  * should skip to the next iteration (again assuming N is increasing).
  */
-static int _dpll_test_fint(struct clk_hw_omap *clk, u8 n)
+static int _dpll_test_fint(struct clk_hw_omap *clk, unsigned int n)
 {
 	struct dpll_data *dd;
 	long fint, fint_min, fint_max;
diff --git a/arch/arm/mach-omap2/cm-regbits-34xx.h b/arch/arm/mach-omap2/cm-regbits-34xx.h
index 04dab2fcf862..ee6c784cd6b7 100644
--- a/arch/arm/mach-omap2/cm-regbits-34xx.h
+++ b/arch/arm/mach-omap2/cm-regbits-34xx.h
@@ -26,11 +26,14 @@
 #define OMAP3430_EN_WDT3_SHIFT				12
 #define OMAP3430_CM_FCLKEN_IVA2_EN_IVA2_MASK		(1 << 0)
 #define OMAP3430_CM_FCLKEN_IVA2_EN_IVA2_SHIFT		0
+#define OMAP3430_IVA2_DPLL_FREQSEL_SHIFT		4
 #define OMAP3430_IVA2_DPLL_FREQSEL_MASK			(0xf << 4)
 #define OMAP3430_EN_IVA2_DPLL_DRIFTGUARD_SHIFT		3
+#define OMAP3430_EN_IVA2_DPLL_SHIFT			0
 #define OMAP3430_EN_IVA2_DPLL_MASK			(0x7 << 0)
 #define OMAP3430_ST_IVA2_SHIFT				0
 #define OMAP3430_ST_IVA2_CLK_MASK			(1 << 0)
+#define OMAP3430_AUTO_IVA2_DPLL_SHIFT			0
 #define OMAP3430_AUTO_IVA2_DPLL_MASK			(0x7 << 0)
 #define OMAP3430_IVA2_CLK_SRC_SHIFT			19
 #define OMAP3430_IVA2_CLK_SRC_WIDTH			3
diff --git a/arch/arm/mach-omap2/cm33xx.h b/arch/arm/mach-omap2/cm33xx.h
index 15a778ce7707..bd2441790779 100644
--- a/arch/arm/mach-omap2/cm33xx.h
+++ b/arch/arm/mach-omap2/cm33xx.h
@@ -380,7 +380,7 @@ void am33xx_cm_clkdm_disable_hwsup(u16 inst, u16 cdoffs);
 void am33xx_cm_clkdm_force_sleep(u16 inst, u16 cdoffs);
 void am33xx_cm_clkdm_force_wakeup(u16 inst, u16 cdoffs);
 
-#ifdef CONFIG_SOC_AM33XX
+#if defined(CONFIG_SOC_AM33XX) || defined(CONFIG_SOC_AM43XX)
 extern int am33xx_cm_wait_module_idle(u16 inst, s16 cdoffs,
 					u16 clkctrl_offs);
 extern void am33xx_cm_module_enable(u8 mode, u16 inst, s16 cdoffs,
diff --git a/arch/arm/mach-omap2/common.h b/arch/arm/mach-omap2/common.h
index a373d508799a..dc571f1d3b8a 100644
--- a/arch/arm/mach-omap2/common.h
+++ b/arch/arm/mach-omap2/common.h
@@ -162,7 +162,8 @@ static inline void omap3xxx_restart(enum reboot_mode mode, const char *cmd)
 }
 #endif
 
-#if defined(CONFIG_ARCH_OMAP4) || defined(CONFIG_SOC_OMAP5)
+#if defined(CONFIG_ARCH_OMAP4) || defined(CONFIG_SOC_OMAP5) || \
+	defined(CONFIG_SOC_DRA7XX) || defined(CONFIG_SOC_AM43XX)
 void omap44xx_restart(enum reboot_mode mode, const char *cmd);
 #else
 static inline void omap44xx_restart(enum reboot_mode mode, const char *cmd)
@@ -248,7 +249,6 @@ static inline void __iomem *omap4_get_scu_base(void)
 }
 #endif
 
-extern void __init gic_init_irq(void);
 extern void gic_dist_disable(void);
 extern void gic_dist_enable(void);
 extern bool gic_dist_disabled(void);
diff --git a/arch/arm/mach-omap2/devices.c b/arch/arm/mach-omap2/devices.c
index 592ba0a0ecf3..b6f8f348296e 100644
--- a/arch/arm/mach-omap2/devices.c
+++ b/arch/arm/mach-omap2/devices.c
@@ -297,33 +297,6 @@ static void omap_init_audio(void)
 static inline void omap_init_audio(void) {}
 #endif
 
-#if defined(CONFIG_SND_OMAP_SOC_OMAP_HDMI) || \
-		defined(CONFIG_SND_OMAP_SOC_OMAP_HDMI_MODULE)
-
-static struct platform_device omap_hdmi_audio = {
-	.name	= "omap-hdmi-audio",
-	.id	= -1,
-};
-
-static void __init omap_init_hdmi_audio(void)
-{
-	struct omap_hwmod *oh;
-	struct platform_device *pdev;
-
-	oh = omap_hwmod_lookup("dss_hdmi");
-	if (!oh)
-		return;
-
-	pdev = omap_device_build("omap-hdmi-audio-dai", -1, oh, NULL, 0);
-	WARN(IS_ERR(pdev),
-	     "Can't build omap_device for omap-hdmi-audio-dai.\n");
-
-	platform_device_register(&omap_hdmi_audio);
-}
-#else
-static inline void omap_init_hdmi_audio(void) {}
-#endif
-
 #if defined(CONFIG_SPI_OMAP24XX) || defined(CONFIG_SPI_OMAP24XX_MODULE)
 
 #include <linux/platform_data/spi-omap2-mcspi.h>
@@ -459,7 +432,6 @@ static int __init omap2_init_devices(void)
 	 */
 	omap_init_audio();
 	omap_init_camera();
-	omap_init_hdmi_audio();
 	omap_init_mbox();
 	/* If dtb is there, the devices will be created dynamically */
 	if (!of_have_populated_dt()) {
diff --git a/arch/arm/mach-omap2/dsp.c b/arch/arm/mach-omap2/dsp.c
index b8208b4b1bd9..f7492df1cbba 100644
--- a/arch/arm/mach-omap2/dsp.c
+++ b/arch/arm/mach-omap2/dsp.c
@@ -29,6 +29,7 @@
 #ifdef CONFIG_TIDSPBRIDGE_DVFS
 #include "omap-pm.h"
 #endif
+#include "soc.h"
 
 #include <linux/platform_data/dsp-omap.h>
 
@@ -59,6 +60,9 @@ void __init omap_dsp_reserve_sdram_memblock(void)
 	phys_addr_t size = CONFIG_TIDSPBRIDGE_MEMPOOL_SIZE;
 	phys_addr_t paddr;
 
+	if (!cpu_is_omap34xx())
+		return;
+
 	if (!size)
 		return;
 
@@ -83,6 +87,9 @@ static int __init omap_dsp_init(void)
 	int err = -ENOMEM;
 	struct omap_dsp_platform_data *pdata = &omap_dsp_pdata;
 
+	if (!cpu_is_omap34xx())
+		return 0;
+
 	pdata->phys_mempool_base = omap_dsp_get_mempool_base();
 
 	if (pdata->phys_mempool_base) {
@@ -115,6 +122,9 @@ module_init(omap_dsp_init);
 
 static void __exit omap_dsp_exit(void)
 {
+	if (!cpu_is_omap34xx())
+		return;
+
 	platform_device_unregister(omap_dsp_pdev);
 }
 module_exit(omap_dsp_exit);
diff --git a/arch/arm/mach-omap2/gpmc.c b/arch/arm/mach-omap2/gpmc.c
index 2c0c2816900f..8bc13380f0a0 100644
--- a/arch/arm/mach-omap2/gpmc.c
+++ b/arch/arm/mach-omap2/gpmc.c
@@ -1615,7 +1615,7 @@ static int gpmc_probe_dt(struct platform_device *pdev)
 		return ret;
 	}
 
-	for_each_child_of_node(pdev->dev.of_node, child) {
+	for_each_available_child_of_node(pdev->dev.of_node, child) {
 
 		if (!child->name)
 			continue;
diff --git a/arch/arm/mach-omap2/id.c b/arch/arm/mach-omap2/id.c
index 43969da5d50b..d42022f2a71e 100644
--- a/arch/arm/mach-omap2/id.c
+++ b/arch/arm/mach-omap2/id.c
@@ -649,6 +649,18 @@ void __init dra7xxx_check_revision(void)
 		}
 		break;
 
+	case 0xb9bc:
+		switch (rev) {
+		case 0:
+			omap_revision = DRA722_REV_ES1_0;
+			break;
+		default:
+			/* If we have no new revisions */
+			omap_revision = DRA722_REV_ES1_0;
+			break;
+		}
+		break;
+
 	default:
 		/* Unknown default to latest silicon rev as default*/
 		pr_warn("%s: unknown idcode=0x%08x (hawkeye=0x%08x,rev=0x%d)\n",
diff --git a/arch/arm/mach-omap2/mux.c b/arch/arm/mach-omap2/mux.c
index fd88edeb027f..f62f7537d899 100644
--- a/arch/arm/mach-omap2/mux.c
+++ b/arch/arm/mach-omap2/mux.c
@@ -183,8 +183,10 @@ static int __init _omap_mux_get_by_name(struct omap_mux_partition *partition,
 		m0_entry = mux->muxnames[0];
 
 		/* First check for full name in mode0.muxmode format */
-		if (mode0_len && strncmp(muxname, m0_entry, mode0_len))
-			continue;
+		if (mode0_len)
+			if (strncmp(muxname, m0_entry, mode0_len) ||
+			    (strlen(m0_entry) != mode0_len))
+				continue;
 
 		/* Then check for muxmode only */
 		for (i = 0; i < OMAP_MUX_NR_MODES; i++) {
diff --git a/arch/arm/mach-omap2/omap4-common.c b/arch/arm/mach-omap2/omap4-common.c
index 326cd982a3cb..539e8106eb96 100644
--- a/arch/arm/mach-omap2/omap4-common.c
+++ b/arch/arm/mach-omap2/omap4-common.c
@@ -102,26 +102,6 @@ void __init omap_barriers_init(void)
 {}
 #endif
 
-void __init gic_init_irq(void)
-{
-	void __iomem *omap_irq_base;
-
-	/* Static mapping, never released */
-	gic_dist_base_addr = ioremap(OMAP44XX_GIC_DIST_BASE, SZ_4K);
-	BUG_ON(!gic_dist_base_addr);
-
-	twd_base = ioremap(OMAP44XX_LOCAL_TWD_BASE, SZ_4K);
-	BUG_ON(!twd_base);
-
-	/* Static mapping, never released */
-	omap_irq_base = ioremap(OMAP44XX_GIC_CPU_BASE, SZ_512);
-	BUG_ON(!omap_irq_base);
-
-	omap_wakeupgen_init();
-
-	gic_init(0, 29, gic_dist_base_addr, omap_irq_base);
-}
-
 void gic_dist_disable(void)
 {
 	if (gic_dist_base_addr)
diff --git a/arch/arm/mach-omap2/omap_hwmod.c b/arch/arm/mach-omap2/omap_hwmod.c
index f7bb435bb543..6c074f37cdd2 100644
--- a/arch/arm/mach-omap2/omap_hwmod.c
+++ b/arch/arm/mach-omap2/omap_hwmod.c
@@ -4251,9 +4251,9 @@ void __init omap_hwmod_init(void)
 		soc_ops.enable_module = _omap4_enable_module;
 		soc_ops.disable_module = _omap4_disable_module;
 		soc_ops.wait_target_ready = _omap4_wait_target_ready;
-		soc_ops.assert_hardreset = _omap4_assert_hardreset;
-		soc_ops.deassert_hardreset = _omap4_deassert_hardreset;
-		soc_ops.is_hardreset_asserted = _omap4_is_hardreset_asserted;
+		soc_ops.assert_hardreset = _am33xx_assert_hardreset;
+		soc_ops.deassert_hardreset = _am33xx_deassert_hardreset;
+		soc_ops.is_hardreset_asserted = _am33xx_is_hardreset_asserted;
 		soc_ops.init_clkdm = _init_clkdm;
 	} else if (soc_is_am33xx()) {
 		soc_ops.enable_module = _am33xx_enable_module;
diff --git a/arch/arm/mach-omap2/omap_hwmod_54xx_data.c b/arch/arm/mach-omap2/omap_hwmod_54xx_data.c
index 290213f2cbe3..1103aa0e0d29 100644
--- a/arch/arm/mach-omap2/omap_hwmod_54xx_data.c
+++ b/arch/arm/mach-omap2/omap_hwmod_54xx_data.c
@@ -2020,6 +2020,77 @@ static struct omap_hwmod omap54xx_wd_timer2_hwmod = {
 	},
 };
 
+/*
+ * 'ocp2scp' class
+ * bridge to transform ocp interface protocol to scp (serial control port)
+ * protocol
+ */
+/* ocp2scp3 */
+static struct omap_hwmod omap54xx_ocp2scp3_hwmod;
+/* l4_cfg -> ocp2scp3 */
+static struct omap_hwmod_ocp_if omap54xx_l4_cfg__ocp2scp3 = {
+	.master		= &omap54xx_l4_cfg_hwmod,
+	.slave		= &omap54xx_ocp2scp3_hwmod,
+	.clk		= "l4_root_clk_div",
+	.user		= OCP_USER_MPU | OCP_USER_SDMA,
+};
+
+static struct omap_hwmod omap54xx_ocp2scp3_hwmod = {
+	.name		= "ocp2scp3",
+	.class		= &omap54xx_ocp2scp_hwmod_class,
+	.clkdm_name	= "l3init_clkdm",
+	.prcm = {
+		.omap4 = {
+			.clkctrl_offs = OMAP54XX_CM_L3INIT_OCP2SCP3_CLKCTRL_OFFSET,
+			.context_offs = OMAP54XX_RM_L3INIT_OCP2SCP3_CONTEXT_OFFSET,
+			.modulemode   = MODULEMODE_HWCTRL,
+		},
+	},
+};
+
+/*
+ * 'sata' class
+ * sata:  serial ata interface  gen2 compliant   ( 1 rx/ 1 tx)
+ */
+
+static struct omap_hwmod_class_sysconfig omap54xx_sata_sysc = {
+	.sysc_offs	= 0x0000,
+	.sysc_flags	= (SYSC_HAS_MIDLEMODE | SYSC_HAS_SIDLEMODE),
+	.idlemodes	= (SIDLE_FORCE | SIDLE_NO | SIDLE_SMART |
+			   SIDLE_SMART_WKUP | MSTANDBY_FORCE | MSTANDBY_NO |
+			   MSTANDBY_SMART | MSTANDBY_SMART_WKUP),
+	.sysc_fields	= &omap_hwmod_sysc_type2,
+};
+
+static struct omap_hwmod_class omap54xx_sata_hwmod_class = {
+	.name	= "sata",
+	.sysc	= &omap54xx_sata_sysc,
+};
+
+/* sata */
+static struct omap_hwmod omap54xx_sata_hwmod = {
+	.name		= "sata",
+	.class		= &omap54xx_sata_hwmod_class,
+	.clkdm_name	= "l3init_clkdm",
+	.flags		= HWMOD_SWSUP_SIDLE | HWMOD_SWSUP_MSTANDBY,
+	.main_clk	= "func_48m_fclk",
+	.mpu_rt_idx	= 1,
+	.prcm = {
+		.omap4 = {
+			.clkctrl_offs = OMAP54XX_CM_L3INIT_SATA_CLKCTRL_OFFSET,
+			.context_offs = OMAP54XX_RM_L3INIT_SATA_CONTEXT_OFFSET,
+			.modulemode   = MODULEMODE_SWCTRL,
+		},
+	},
+};
+
+/* l4_cfg -> sata */
+static struct omap_hwmod_ocp_if omap54xx_l4_cfg__sata = {
+	.master		= &omap54xx_l4_cfg_hwmod,
+	.slave		= &omap54xx_sata_hwmod,
+	.clk		= "l3_iclk_div",
+	.user		= OCP_USER_MPU | OCP_USER_SDMA,
+};
 
 /*
  * Interfaces
@@ -2765,6 +2836,8 @@ static struct omap_hwmod_ocp_if *omap54xx_hwmod_ocp_ifs[] __initdata = {
 	&omap54xx_l4_cfg__usb_tll_hs,
 	&omap54xx_l4_cfg__usb_otg_ss,
 	&omap54xx_l4_wkup__wd_timer2,
+	&omap54xx_l4_cfg__ocp2scp3,
+	&omap54xx_l4_cfg__sata,
 	NULL,
 };
 
diff --git a/arch/arm/mach-omap2/omap_hwmod_7xx_data.c b/arch/arm/mach-omap2/omap_hwmod_7xx_data.c
index 20b4398cec05..284324f2b98a 100644
--- a/arch/arm/mach-omap2/omap_hwmod_7xx_data.c
+++ b/arch/arm/mach-omap2/omap_hwmod_7xx_data.c
@@ -1268,9 +1268,6 @@ static struct omap_hwmod_class dra7xx_sata_hwmod_class = {
 };
 
 /* sata */
-static struct omap_hwmod_opt_clk sata_opt_clks[] = {
-	{ .role = "ref_clk", .clk = "sata_ref_clk" },
-};
 
 static struct omap_hwmod dra7xx_sata_hwmod = {
 	.name		= "sata",
@@ -1278,6 +1275,7 @@ static struct omap_hwmod dra7xx_sata_hwmod = {
 	.clkdm_name	= "l3init_clkdm",
 	.flags		= HWMOD_SWSUP_SIDLE | HWMOD_SWSUP_MSTANDBY,
 	.main_clk	= "func_48m_fclk",
+	.mpu_rt_idx	= 1,
 	.prcm = {
 		.omap4 = {
 			.clkctrl_offs = DRA7XX_CM_L3INIT_SATA_CLKCTRL_OFFSET,
@@ -1285,8 +1283,6 @@ static struct omap_hwmod dra7xx_sata_hwmod = {
 			.modulemode   = MODULEMODE_SWCTRL,
 		},
 	},
-	.opt_clks	= sata_opt_clks,
-	.opt_clks_cnt	= ARRAY_SIZE(sata_opt_clks),
 };
 
 /*
@@ -1731,8 +1727,20 @@ static struct omap_hwmod dra7xx_uart6_hwmod = {
  *
  */
 
+static struct omap_hwmod_class_sysconfig dra7xx_usb_otg_ss_sysc = {
+	.rev_offs	= 0x0000,
+	.sysc_offs	= 0x0010,
+	.sysc_flags	= (SYSC_HAS_DMADISABLE | SYSC_HAS_MIDLEMODE |
+			   SYSC_HAS_SIDLEMODE),
+	.idlemodes	= (SIDLE_FORCE | SIDLE_NO | SIDLE_SMART |
+			   SIDLE_SMART_WKUP | MSTANDBY_FORCE | MSTANDBY_NO |
+			   MSTANDBY_SMART | MSTANDBY_SMART_WKUP),
+	.sysc_fields	= &omap_hwmod_sysc_type2,
+};
+
 static struct omap_hwmod_class dra7xx_usb_otg_ss_hwmod_class = {
 	.name	= "usb_otg_ss",
+	.sysc	= &dra7xx_usb_otg_ss_sysc,
 };
 
 /* usb_otg_ss1 */
diff --git a/arch/arm/mach-omap2/prm-regbits-34xx.h b/arch/arm/mach-omap2/prm-regbits-34xx.h
index 106132db532b..cbefbd7cfdb5 100644
--- a/arch/arm/mach-omap2/prm-regbits-34xx.h
+++ b/arch/arm/mach-omap2/prm-regbits-34xx.h
@@ -35,6 +35,8 @@
 #define OMAP3430_LOGICSTATEST_MASK			(1 << 2)
 #define OMAP3430_LASTLOGICSTATEENTERED_MASK		(1 << 2)
 #define OMAP3430_LASTPOWERSTATEENTERED_MASK		(0x3 << 0)
+#define OMAP3430_GRPSEL_MCBSP5_MASK			(1 << 10)
+#define OMAP3430_GRPSEL_MCBSP1_MASK			(1 << 9)
 #define OMAP3630_GRPSEL_UART4_MASK			(1 << 18)
 #define OMAP3430_GRPSEL_GPIO6_MASK			(1 << 17)
 #define OMAP3430_GRPSEL_GPIO5_MASK			(1 << 16)
@@ -42,6 +44,10 @@
 #define OMAP3430_GRPSEL_GPIO3_MASK			(1 << 14)
 #define OMAP3430_GRPSEL_GPIO2_MASK			(1 << 13)
 #define OMAP3430_GRPSEL_UART3_MASK			(1 << 11)
+#define OMAP3430_GRPSEL_GPT8_MASK			(1 << 9)
+#define OMAP3430_GRPSEL_GPT7_MASK			(1 << 8)
+#define OMAP3430_GRPSEL_GPT6_MASK			(1 << 7)
+#define OMAP3430_GRPSEL_GPT5_MASK			(1 << 6)
 #define OMAP3430_GRPSEL_MCBSP4_MASK			(1 << 2)
 #define OMAP3430_GRPSEL_MCBSP3_MASK			(1 << 1)
 #define OMAP3430_GRPSEL_MCBSP2_MASK			(1 << 0)
diff --git a/arch/arm/mach-omap2/soc.h b/arch/arm/mach-omap2/soc.h
index de2a34c423a7..01ca8086fb6c 100644
--- a/arch/arm/mach-omap2/soc.h
+++ b/arch/arm/mach-omap2/soc.h
@@ -462,6 +462,7 @@ IS_OMAP_TYPE(3430, 0x3430)
 #define DRA7XX_CLASS		0x07000000
 #define DRA752_REV_ES1_0	(DRA7XX_CLASS | (0x52 << 16) | (0x10 << 8))
 #define DRA752_REV_ES1_1	(DRA7XX_CLASS | (0x52 << 16) | (0x11 << 8))
+#define DRA722_REV_ES1_0	(DRA7XX_CLASS | (0x22 << 16) | (0x10 << 8))
 
 void omap2xxx_check_revision(void);
 void omap3xxx_check_revision(void);
diff --git a/arch/arm/mach-sa1100/collie.c b/arch/arm/mach-sa1100/collie.c
index f9874ba60cc8..108939f8d053 100644
--- a/arch/arm/mach-sa1100/collie.c
+++ b/arch/arm/mach-sa1100/collie.c
@@ -329,6 +329,11 @@ static struct mtd_partition collie_partitions[] = {
 		.name		= "rootfs",
 		.offset 	= MTDPART_OFS_APPEND,
 		.size		= 0x00e20000,
+	}, {
+		.name		= "bootblock",
+		.offset		= MTDPART_OFS_APPEND,
+		.size		= 0x00020000,
+		.mask_flags	= MTD_WRITEABLE
 	}
 };
 
@@ -356,7 +361,7 @@ static void collie_flash_exit(void)
 }
 
 static struct flash_platform_data collie_flash_data = {
-	.map_name	= "jedec_probe",
+	.map_name	= "cfi_probe",
 	.init		= collie_flash_init,
 	.set_vpp	= collie_set_vpp,
 	.exit		= collie_flash_exit,
diff --git a/arch/arm/mach-sti/Kconfig b/arch/arm/mach-sti/Kconfig
index 7e33e9d2c42e..878e9ec97d0f 100644
--- a/arch/arm/mach-sti/Kconfig
+++ b/arch/arm/mach-sti/Kconfig
@@ -11,8 +11,8 @@ menuconfig ARCH_STI
 	select ARM_ERRATA_754322
 	select ARM_ERRATA_764369 if SMP
 	select ARM_ERRATA_775420
-	select PL310_ERRATA_753970 if CACHE_PL310
-	select PL310_ERRATA_769419 if CACHE_PL310
+	select PL310_ERRATA_753970 if CACHE_L2X0
+	select PL310_ERRATA_769419 if CACHE_L2X0
 	help
 	  Include support for STiH41x SOCs like STiH415/416 using the device tree
 	  for discovery
diff --git a/arch/arm/mach-sunxi/sunxi.c b/arch/arm/mach-sunxi/sunxi.c
index 3f9587bb51f6..b6085084e0ff 100644
--- a/arch/arm/mach-sunxi/sunxi.c
+++ b/arch/arm/mach-sunxi/sunxi.c
@@ -12,8 +12,81 @@
 
 #include <linux/clk-provider.h>
 #include <linux/clocksource.h>
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/of_platform.h>
+#include <linux/io.h>
+#include <linux/reboot.h>
 
 #include <asm/mach/arch.h>
+#include <asm/mach/map.h>
+#include <asm/system_misc.h>
+
+#define SUN4I_WATCHDOG_CTRL_REG		0x00
+#define SUN4I_WATCHDOG_CTRL_RESTART		BIT(0)
+#define SUN4I_WATCHDOG_MODE_REG		0x04
+#define SUN4I_WATCHDOG_MODE_ENABLE		BIT(0)
+#define SUN4I_WATCHDOG_MODE_RESET_ENABLE	BIT(1)
+
+#define SUN6I_WATCHDOG1_IRQ_REG		0x00
+#define SUN6I_WATCHDOG1_CTRL_REG	0x10
+#define SUN6I_WATCHDOG1_CTRL_RESTART		BIT(0)
+#define SUN6I_WATCHDOG1_CONFIG_REG	0x14
+#define SUN6I_WATCHDOG1_CONFIG_RESTART		BIT(0)
+#define SUN6I_WATCHDOG1_CONFIG_IRQ		BIT(1)
+#define SUN6I_WATCHDOG1_MODE_REG	0x18
+#define SUN6I_WATCHDOG1_MODE_ENABLE		BIT(0)
+
+static void __iomem *wdt_base;
+
+static void sun4i_restart(enum reboot_mode mode, const char *cmd)
+{
+	if (!wdt_base)
+		return;
+
+	/* Enable timer and set reset bit in the watchdog */
+	writel(SUN4I_WATCHDOG_MODE_ENABLE | SUN4I_WATCHDOG_MODE_RESET_ENABLE,
+	       wdt_base + SUN4I_WATCHDOG_MODE_REG);
+
+	/*
+	 * Restart the watchdog. The default (and lowest) interval
+	 * value for the watchdog is 0.5s.
+	 */
+	writel(SUN4I_WATCHDOG_CTRL_RESTART, wdt_base + SUN4I_WATCHDOG_CTRL_REG);
+
+	while (1) {
+		mdelay(5);
+		writel(SUN4I_WATCHDOG_MODE_ENABLE | SUN4I_WATCHDOG_MODE_RESET_ENABLE,
+		       wdt_base + SUN4I_WATCHDOG_MODE_REG);
+	}
+}
+
+static struct of_device_id sunxi_restart_ids[] = {
+	{ .compatible = "allwinner,sun4i-a10-wdt" },
+	{ /*sentinel*/ }
+};
+
+static void sunxi_setup_restart(void)
+{
+	struct device_node *np;
+
+	np = of_find_matching_node(NULL, sunxi_restart_ids);
+	if (WARN(!np, "unable to setup watchdog restart"))
+		return;
+
+	wdt_base = of_iomap(np, 0);
+	WARN(!wdt_base, "failed to map watchdog base address");
+}
+
+static void __init sunxi_dt_init(void)
+{
+	sunxi_setup_restart();
+
+	of_platform_populate(NULL, of_default_bus_match_table, NULL, NULL);
+}
 
 static const char * const sunxi_board_dt_compat[] = {
 	"allwinner,sun4i-a10",
@@ -23,7 +96,9 @@ static const char * const sunxi_board_dt_compat[] = {
 };
 
 DT_MACHINE_START(SUNXI_DT, "Allwinner A1X (Device Tree)")
+	.init_machine	= sunxi_dt_init,
 	.dt_compat	= sunxi_board_dt_compat,
+	.restart	= sun4i_restart,
 MACHINE_END
 
 static const char * const sun6i_board_dt_compat[] = {
@@ -51,5 +126,7 @@ static const char * const sun7i_board_dt_compat[] = {
 };
 
 DT_MACHINE_START(SUN7I_DT, "Allwinner sun7i (A20) Family")
+	.init_machine	= sunxi_dt_init,
 	.dt_compat	= sun7i_board_dt_compat,
+	.restart	= sun4i_restart,
 MACHINE_END
diff --git a/arch/arm/mach-ux500/Kconfig b/arch/arm/mach-ux500/Kconfig
index 5be7c4583a93..699e8601dbf0 100644
--- a/arch/arm/mach-ux500/Kconfig
+++ b/arch/arm/mach-ux500/Kconfig
@@ -15,7 +15,7 @@ menuconfig ARCH_U8500
 	select PINCTRL
 	select PINCTRL_ABX500
 	select PINCTRL_NOMADIK
-	select PL310_ERRATA_753970 if CACHE_PL310
+	select PL310_ERRATA_753970 if CACHE_L2X0
 	help
 	  Support for ST-Ericsson's Ux500 architecture
 
diff --git a/arch/arm/mach-vexpress/Kconfig b/arch/arm/mach-vexpress/Kconfig
index 99c1f151c403..d8b9330f896a 100644
--- a/arch/arm/mach-vexpress/Kconfig
+++ b/arch/arm/mach-vexpress/Kconfig
@@ -43,7 +43,7 @@ config ARCH_VEXPRESS_CORTEX_A5_A9_ERRATA
 	bool "Enable A5 and A9 only errata work-arounds"
 	default y
 	select ARM_ERRATA_720789
-	select PL310_ERRATA_753970 if CACHE_PL310
+	select PL310_ERRATA_753970 if CACHE_L2X0
 	help
 	  Provides common dependencies for Versatile Express platforms
 	  based on Cortex-A5 and Cortex-A9 processors. In order to
diff --git a/arch/arm/mm/Kconfig b/arch/arm/mm/Kconfig
index eda0dd0ab97b..c348eaee7ee2 100644
--- a/arch/arm/mm/Kconfig
+++ b/arch/arm/mm/Kconfig
@@ -889,9 +889,10 @@ config CACHE_L2X0
 	help
 	  This option enables the L2x0 PrimeCell.
 
+if CACHE_L2X0
+
 config CACHE_PL310
 	bool
-	depends on CACHE_L2X0
 	default y if CPU_V7 && !(CPU_V6 || CPU_V6K)
 	help
 	  This option enables optimisations for the PL310 cache
@@ -899,7 +900,6 @@ config CACHE_PL310
 
 config PL310_ERRATA_588369
 	bool "PL310 errata: Clean & Invalidate maintenance operations do not invalidate clean lines"
-	depends on CACHE_L2X0
 	help
 	   The PL310 L2 cache controller implements three types of Clean &
 	   Invalidate maintenance operations: by Physical Address
@@ -912,7 +912,6 @@ config PL310_ERRATA_588369
 
 config PL310_ERRATA_727915
 	bool "PL310 errata: Background Clean & Invalidate by Way operation can cause data corruption"
-	depends on CACHE_L2X0
 	help
 	  PL310 implements the Clean & Invalidate by Way L2 cache maintenance
 	  operation (offset 0x7FC). This operation runs in background so that
@@ -923,7 +922,6 @@ config PL310_ERRATA_727915
 
 config PL310_ERRATA_753970
 	bool "PL310 errata: cache sync operation may be faulty"
-	depends on CACHE_PL310
 	help
 	  This option enables the workaround for the 753970 PL310 (r3p0) erratum.
 
@@ -938,7 +936,6 @@ config PL310_ERRATA_753970
 
 config PL310_ERRATA_769419
 	bool "PL310 errata: no automatic Store Buffer drain"
-	depends on CACHE_L2X0
 	help
 	  On revisions of the PL310 prior to r3p2, the Store Buffer does
 	  not automatically drain. This can cause normal, non-cacheable
@@ -948,6 +945,8 @@ config PL310_ERRATA_769419
 	  on systems with an outer cache, the store buffer is drained
 	  explicitly.
 
+endif
+
 config CACHE_TAUROS2
 	bool "Enable the Tauros2 L2 cache controller"
 	depends on (ARCH_DOVE || ARCH_MMP || CPU_PJ4)
diff --git a/arch/arm/mm/cache-l2x0.c b/arch/arm/mm/cache-l2x0.c
index efc5cabf70e0..7c3fb41a462e 100644
--- a/arch/arm/mm/cache-l2x0.c
+++ b/arch/arm/mm/cache-l2x0.c
@@ -664,7 +664,7 @@ static int l2c310_cpu_enable_flz(struct notifier_block *nb, unsigned long act, v
 
 static void __init l2c310_enable(void __iomem *base, u32 aux, unsigned num_lock)
 {
-	unsigned rev = readl_relaxed(base + L2X0_CACHE_ID) & L2X0_CACHE_ID_PART_MASK;
+	unsigned rev = readl_relaxed(base + L2X0_CACHE_ID) & L2X0_CACHE_ID_RTL_MASK;
 	bool cortex_a9 = read_cpuid_part_number() == ARM_CPU_PART_CORTEX_A9;
 
 	if (rev >= L310_CACHE_ID_RTL_R2P0) {
@@ -1069,6 +1069,33 @@ static const struct l2c_init_data of_l2c310_data __initconst = {
 };
 
 /*
+ * This is a variant of the of_l2c310_data with .sync set to
+ * NULL. Outer sync operations are not needed when the system is I/O
+ * coherent, and potentially harmful in certain situations (PCIe/PL310
+ * deadlock on Armada 375/38x due to hardware I/O coherency). The
+ * other operations are kept because they are infrequent (therefore do
+ * not cause the deadlock in practice) and needed for secondary CPU
+ * boot and other power management activities.
+ */
+static const struct l2c_init_data of_l2c310_coherent_data __initconst = {
+	.type = "L2C-310 Coherent",
+	.way_size_0 = SZ_8K,
+	.num_lock = 8,
+	.of_parse = l2c310_of_parse,
+	.enable = l2c310_enable,
+	.fixup = l2c310_fixup,
+	.save  = l2c310_save,
+	.outer_cache = {
+		.inv_range   = l2c210_inv_range,
+		.clean_range = l2c210_clean_range,
+		.flush_range = l2c210_flush_range,
+		.flush_all   = l2c210_flush_all,
+		.disable     = l2c310_disable,
+		.resume      = l2c310_resume,
+	},
+};
+
+/*
  * Note that the end addresses passed to Linux primitives are
  * noninclusive, while the hardware cache range operations use
  * inclusive start and end addresses.
@@ -1487,6 +1514,10 @@ int __init l2x0_of_init(u32 aux_val, u32 aux_mask)
 
 	data = of_match_node(l2x0_ids, np)->data;
 
+	if (of_device_is_compatible(np, "arm,pl310-cache") &&
+	    of_property_read_bool(np, "arm,io-coherent"))
+		data = &of_l2c310_coherent_data;
+
 	old_aux = readl_relaxed(l2x0_base + L2X0_AUX_CTRL);
 	if (old_aux != ((old_aux & aux_mask) | aux_val)) {
 		pr_warn("L2C: platform modifies aux control register: 0x%08x -> 0x%08x\n",
diff --git a/arch/arm/mm/nommu.c b/arch/arm/mm/nommu.c
index da1874f9f8cf..a014dfacd5ca 100644
--- a/arch/arm/mm/nommu.c
+++ b/arch/arm/mm/nommu.c
@@ -300,6 +300,7 @@ void __init sanity_check_meminfo(void)
 	sanity_check_meminfo_mpu();
 	end = memblock_end_of_DRAM();
 	high_memory = __va(end - 1) + 1;
+	memblock_set_current_limit(end);
 }
 
 /*
diff --git a/arch/arm/mm/proc-arm925.S b/arch/arm/mm/proc-arm925.S
index 97448c3acf38..ba0d58e1a2a2 100644
--- a/arch/arm/mm/proc-arm925.S
+++ b/arch/arm/mm/proc-arm925.S
@@ -502,6 +502,7 @@ __\name\()_proc_info:
 	.long	\cpu_val
 	.long	\cpu_mask
 	.long   PMD_TYPE_SECT | \
+		PMD_SECT_CACHEABLE | \
 		PMD_BIT4 | \
 		PMD_SECT_AP_WRITE | \
 		PMD_SECT_AP_READ
diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig
index a474de346be6..839f48c26ef0 100644
--- a/arch/arm64/Kconfig
+++ b/arch/arm64/Kconfig
@@ -4,6 +4,7 @@ config ARM64
 	select ARCH_HAS_OPP
 	select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
 	select ARCH_USE_CMPXCHG_LOCKREF
+	select ARCH_SUPPORTS_ATOMIC_RMW
 	select ARCH_WANT_OPTIONAL_GPIOLIB
 	select ARCH_WANT_COMPAT_IPC_PARSE_VERSION
 	select ARCH_WANT_FRAME_POINTERS
diff --git a/arch/arm64/include/asm/memory.h b/arch/arm64/include/asm/memory.h
index 993bce527b85..902eb708804a 100644
--- a/arch/arm64/include/asm/memory.h
+++ b/arch/arm64/include/asm/memory.h
@@ -56,6 +56,8 @@
 #define TASK_SIZE_32		UL(0x100000000)
 #define TASK_SIZE		(test_thread_flag(TIF_32BIT) ? \
 				TASK_SIZE_32 : TASK_SIZE_64)
+#define TASK_SIZE_OF(tsk)	(test_tsk_thread_flag(tsk, TIF_32BIT) ? \
+				TASK_SIZE_32 : TASK_SIZE_64)
 #else
 #define TASK_SIZE		TASK_SIZE_64
 #endif /* CONFIG_COMPAT */
diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h
index 579702086488..e0ccceb317d9 100644
--- a/arch/arm64/include/asm/pgtable.h
+++ b/arch/arm64/include/asm/pgtable.h
@@ -292,7 +292,7 @@ extern pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
 #define pmd_sect(pmd)		((pmd_val(pmd) & PMD_TYPE_MASK) == \
 				 PMD_TYPE_SECT)
 
-#ifdef ARM64_64K_PAGES
+#ifdef CONFIG_ARM64_64K_PAGES
 #define pud_sect(pud)		(0)
 #else
 #define pud_sect(pud)		((pud_val(pud) & PUD_TYPE_MASK) == \
diff --git a/arch/arm64/include/asm/ptrace.h b/arch/arm64/include/asm/ptrace.h
index a429b5940be2..501000fadb6f 100644
--- a/arch/arm64/include/asm/ptrace.h
+++ b/arch/arm64/include/asm/ptrace.h
@@ -21,6 +21,10 @@
 
 #include <uapi/asm/ptrace.h>
 
+/* Current Exception Level values, as contained in CurrentEL */
+#define CurrentEL_EL1		(1 << 2)
+#define CurrentEL_EL2		(2 << 2)
+
 /* AArch32-specific ptrace requests */
 #define COMPAT_PTRACE_GETREGS		12
 #define COMPAT_PTRACE_SETREGS		13
diff --git a/arch/arm64/kernel/efi-entry.S b/arch/arm64/kernel/efi-entry.S
index 66716c9b9e5f..619b1dd7bcde 100644
--- a/arch/arm64/kernel/efi-entry.S
+++ b/arch/arm64/kernel/efi-entry.S
@@ -78,8 +78,7 @@ ENTRY(efi_stub_entry)
 
 	/* Turn off Dcache and MMU */
 	mrs	x0, CurrentEL
-	cmp	x0, #PSR_MODE_EL2t
-	ccmp	x0, #PSR_MODE_EL2h, #0x4, ne
+	cmp	x0, #CurrentEL_EL2
 	b.ne	1f
 	mrs	x0, sctlr_el2
 	bic	x0, x0, #1 << 0	// clear SCTLR.M
diff --git a/arch/arm64/kernel/efi-stub.c b/arch/arm64/kernel/efi-stub.c
index 60e98a639ac5..e786e6cdc400 100644
--- a/arch/arm64/kernel/efi-stub.c
+++ b/arch/arm64/kernel/efi-stub.c
@@ -12,8 +12,6 @@
 #include <linux/efi.h>
 #include <linux/libfdt.h>
 #include <asm/sections.h>
-#include <generated/compile.h>
-#include <generated/utsrelease.h>
 
 /*
  * AArch64 requires the DTB to be 8-byte aligned in the first 512MiB from
diff --git a/arch/arm64/kernel/head.S b/arch/arm64/kernel/head.S
index a96d3a6a63f6..a2c1195abb7f 100644
--- a/arch/arm64/kernel/head.S
+++ b/arch/arm64/kernel/head.S
@@ -270,8 +270,7 @@ ENDPROC(stext)
  */
 ENTRY(el2_setup)
 	mrs	x0, CurrentEL
-	cmp	x0, #PSR_MODE_EL2t
-	ccmp	x0, #PSR_MODE_EL2h, #0x4, ne
+	cmp	x0, #CurrentEL_EL2
 	b.ne	1f
 	mrs	x0, sctlr_el2
 CPU_BE(	orr	x0, x0, #(1 << 25)	)	// Set the EE bit for EL2
diff --git a/arch/arm64/mm/copypage.c b/arch/arm64/mm/copypage.c
index 9aecbace4128..13bbc3be6f5a 100644
--- a/arch/arm64/mm/copypage.c
+++ b/arch/arm64/mm/copypage.c
@@ -27,8 +27,10 @@ void __cpu_copy_user_page(void *kto, const void *kfrom, unsigned long vaddr)
 	copy_page(kto, kfrom);
 	__flush_dcache_area(kto, PAGE_SIZE);
 }
+EXPORT_SYMBOL_GPL(__cpu_copy_user_page);
 
 void __cpu_clear_user_page(void *kaddr, unsigned long vaddr)
 {
 	clear_page(kaddr);
 }
+EXPORT_SYMBOL_GPL(__cpu_clear_user_page);
diff --git a/arch/arm64/mm/flush.c b/arch/arm64/mm/flush.c
index e4193e3adc7f..0d64089d28b5 100644
--- a/arch/arm64/mm/flush.c
+++ b/arch/arm64/mm/flush.c
@@ -79,7 +79,8 @@ void __sync_icache_dcache(pte_t pte, unsigned long addr)
 		return;
 
 	if (!test_and_set_bit(PG_dcache_clean, &page->flags)) {
-		__flush_dcache_area(page_address(page), PAGE_SIZE);
+		__flush_dcache_area(page_address(page),
+				PAGE_SIZE << compound_order(page));
 		__flush_icache_all();
 	} else if (icache_is_aivivt()) {
 		__flush_icache_all();
diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
index f43db8a69262..e90c5426fe14 100644
--- a/arch/arm64/mm/init.c
+++ b/arch/arm64/mm/init.c
@@ -60,6 +60,17 @@ static int __init early_initrd(char *p)
 early_param("initrd", early_initrd);
 #endif
 
+/*
+ * Return the maximum physical address for ZONE_DMA (DMA_BIT_MASK(32)). It
+ * currently assumes that for memory starting above 4G, 32-bit devices will
+ * use a DMA offset.
+ */
+static phys_addr_t max_zone_dma_phys(void)
+{
+	phys_addr_t offset = memblock_start_of_DRAM() & GENMASK_ULL(63, 32);
+	return min(offset + (1ULL << 32), memblock_end_of_DRAM());
+}
+
 static void __init zone_sizes_init(unsigned long min, unsigned long max)
 {
 	struct memblock_region *reg;
@@ -70,9 +81,7 @@ static void __init zone_sizes_init(unsigned long min, unsigned long max)
 
 	/* 4GB maximum for 32-bit only capable devices */
 	if (IS_ENABLED(CONFIG_ZONE_DMA)) {
-		unsigned long max_dma_phys =
-			(unsigned long)(dma_to_phys(NULL, DMA_BIT_MASK(32)) + 1);
-		max_dma = max(min, min(max, max_dma_phys >> PAGE_SHIFT));
+		max_dma = PFN_DOWN(max_zone_dma_phys());
 		zone_size[ZONE_DMA] = max_dma - min;
 	}
 	zone_size[ZONE_NORMAL] = max - max_dma;
@@ -146,7 +155,7 @@ void __init arm64_memblock_init(void)
 
 	/* 4GB maximum for 32-bit only capable devices */
 	if (IS_ENABLED(CONFIG_ZONE_DMA))
-		dma_phys_limit = dma_to_phys(NULL, DMA_BIT_MASK(32)) + 1;
+		dma_phys_limit = max_zone_dma_phys();
 	dma_contiguous_reserve(dma_phys_limit);
 
 	memblock_allow_resize();
diff --git a/arch/blackfin/configs/BF609-EZKIT_defconfig b/arch/blackfin/configs/BF609-EZKIT_defconfig
index a7e9bfd84183..fcec5ce71392 100644
--- a/arch/blackfin/configs/BF609-EZKIT_defconfig
+++ b/arch/blackfin/configs/BF609-EZKIT_defconfig
@@ -102,7 +102,7 @@ CONFIG_I2C_CHARDEV=y
 CONFIG_I2C_BLACKFIN_TWI=y
 CONFIG_I2C_BLACKFIN_TWI_CLK_KHZ=100
 CONFIG_SPI=y
-CONFIG_SPI_BFIN_V3=y
+CONFIG_SPI_ADI_V3=y
 CONFIG_GPIOLIB=y
 CONFIG_GPIO_SYSFS=y
 # CONFIG_HWMON is not set
diff --git a/arch/blackfin/kernel/vmlinux.lds.S b/arch/blackfin/kernel/vmlinux.lds.S
index ba35864b2b74..c9eec84aa258 100644
--- a/arch/blackfin/kernel/vmlinux.lds.S
+++ b/arch/blackfin/kernel/vmlinux.lds.S
@@ -145,7 +145,7 @@ SECTIONS
 
 	.text_l1 L1_CODE_START : AT(LOADADDR(.exit.data) + SIZEOF(.exit.data))
 #else
-	.init.data : AT(__data_lma + __data_len)
+	.init.data : AT(__data_lma + __data_len + 32)
 	{
 		__sinitdata = .;
 		INIT_DATA
diff --git a/arch/blackfin/mach-bf533/boards/blackstamp.c b/arch/blackfin/mach-bf533/boards/blackstamp.c
index 63b0e4fe760c..0ccf0cf4daaf 100644
--- a/arch/blackfin/mach-bf533/boards/blackstamp.c
+++ b/arch/blackfin/mach-bf533/boards/blackstamp.c
@@ -20,6 +20,7 @@
 #include <linux/spi/spi.h>
 #include <linux/spi/flash.h>
 #include <linux/irq.h>
+#include <linux/gpio.h>
 #include <linux/i2c.h>
 #include <asm/dma.h>
 #include <asm/bfin5xx_spi.h>
diff --git a/arch/blackfin/mach-bf537/boards/cm_bf537e.c b/arch/blackfin/mach-bf537/boards/cm_bf537e.c
index c65c6dbda3da..1e7290ef3525 100644
--- a/arch/blackfin/mach-bf537/boards/cm_bf537e.c
+++ b/arch/blackfin/mach-bf537/boards/cm_bf537e.c
@@ -21,6 +21,7 @@
 #endif
 #include <linux/ata_platform.h>
 #include <linux/irq.h>
+#include <linux/gpio.h>
 #include <asm/dma.h>
 #include <asm/bfin5xx_spi.h>
 #include <asm/portmux.h>
diff --git a/arch/blackfin/mach-bf537/boards/cm_bf537u.c b/arch/blackfin/mach-bf537/boards/cm_bf537u.c
index af58454b4bff..c7495dc74690 100644
--- a/arch/blackfin/mach-bf537/boards/cm_bf537u.c
+++ b/arch/blackfin/mach-bf537/boards/cm_bf537u.c
@@ -21,6 +21,7 @@
 #endif
 #include <linux/ata_platform.h>
 #include <linux/irq.h>
+#include <linux/gpio.h>
 #include <asm/dma.h>
 #include <asm/bfin5xx_spi.h>
 #include <asm/portmux.h>
diff --git a/arch/blackfin/mach-bf537/boards/tcm_bf537.c b/arch/blackfin/mach-bf537/boards/tcm_bf537.c
index a0211225748d..6b988ad653d8 100644
--- a/arch/blackfin/mach-bf537/boards/tcm_bf537.c
+++ b/arch/blackfin/mach-bf537/boards/tcm_bf537.c
@@ -21,6 +21,7 @@
 #endif
 #include <linux/ata_platform.h>
 #include <linux/irq.h>
+#include <linux/gpio.h>
 #include <asm/dma.h>
 #include <asm/bfin5xx_spi.h>
 #include <asm/portmux.h>
diff --git a/arch/blackfin/mach-bf548/boards/ezkit.c b/arch/blackfin/mach-bf548/boards/ezkit.c
index 90138e6112c1..1fe7ff286619 100644
--- a/arch/blackfin/mach-bf548/boards/ezkit.c
+++ b/arch/blackfin/mach-bf548/boards/ezkit.c
@@ -2118,7 +2118,7 @@ static struct pinctrl_map __initdata bfin_pinmux_map[] = {
 	PIN_MAP_MUX_GROUP_DEFAULT("bfin-rotary",  "pinctrl-adi2.0", NULL, "rotary"),
 	PIN_MAP_MUX_GROUP_DEFAULT("bfin_can.0",  "pinctrl-adi2.0", NULL, "can0"),
 	PIN_MAP_MUX_GROUP_DEFAULT("bfin_can.1",  "pinctrl-adi2.0", NULL, "can1"),
-	PIN_MAP_MUX_GROUP_DEFAULT("bf54x-lq043",  "pinctrl-adi2.0", NULL, "ppi0_24b"),
+	PIN_MAP_MUX_GROUP_DEFAULT("bf54x-lq043",  "pinctrl-adi2.0", "ppi0_24bgrp", "ppi0"),
 	PIN_MAP_MUX_GROUP_DEFAULT("bfin-i2s.0",  "pinctrl-adi2.0", NULL, "sport0"),
 	PIN_MAP_MUX_GROUP_DEFAULT("bfin-tdm.0",  "pinctrl-adi2.0", NULL, "sport0"),
 	PIN_MAP_MUX_GROUP_DEFAULT("bfin-ac97.0",  "pinctrl-adi2.0", NULL, "sport0"),
@@ -2140,7 +2140,9 @@ static struct pinctrl_map __initdata bfin_pinmux_map[] = {
 	PIN_MAP_MUX_GROUP_DEFAULT("pata-bf54x",  "pinctrl-adi2.0", NULL, "atapi_alter"),
 #endif
 	PIN_MAP_MUX_GROUP_DEFAULT("bf5xx-nand.0",  "pinctrl-adi2.0", NULL, "nfc0"),
-	PIN_MAP_MUX_GROUP_DEFAULT("bf54x-keys",  "pinctrl-adi2.0", NULL, "keys_4x4"),
+	PIN_MAP_MUX_GROUP_DEFAULT("bf54x-keys",  "pinctrl-adi2.0", "keys_4x4grp", "keys"),
+	PIN_MAP_MUX_GROUP("bf54x-keys", "4bit",  "pinctrl-adi2.0", "keys_4x4grp", "keys"),
+	PIN_MAP_MUX_GROUP("bf54x-keys", "8bit",  "pinctrl-adi2.0", "keys_8x8grp", "keys"),
 };
 
 static int __init ezkit_init(void)
diff --git a/arch/blackfin/mach-bf561/boards/acvilon.c b/arch/blackfin/mach-bf561/boards/acvilon.c
index 430b16d5ccb1..6ab951534d79 100644
--- a/arch/blackfin/mach-bf561/boards/acvilon.c
+++ b/arch/blackfin/mach-bf561/boards/acvilon.c
@@ -44,6 +44,7 @@
 #include <linux/spi/flash.h>
 #include <linux/irq.h>
 #include <linux/interrupt.h>
+#include <linux/gpio.h>
 #include <linux/jiffies.h>
 #include <linux/i2c-pca-platform.h>
 #include <linux/delay.h>
diff --git a/arch/blackfin/mach-bf561/boards/cm_bf561.c b/arch/blackfin/mach-bf561/boards/cm_bf561.c
index 9f777df4cacc..e862f7823e68 100644
--- a/arch/blackfin/mach-bf561/boards/cm_bf561.c
+++ b/arch/blackfin/mach-bf561/boards/cm_bf561.c
@@ -18,6 +18,7 @@
 #endif
 #include <linux/ata_platform.h>
 #include <linux/irq.h>
+#include <linux/gpio.h>
 #include <asm/dma.h>
 #include <asm/bfin5xx_spi.h>
 #include <asm/portmux.h>
diff --git a/arch/blackfin/mach-bf561/boards/ezkit.c b/arch/blackfin/mach-bf561/boards/ezkit.c
index 88dee43e7abe..2de71e8c104b 100644
--- a/arch/blackfin/mach-bf561/boards/ezkit.c
+++ b/arch/blackfin/mach-bf561/boards/ezkit.c
@@ -14,6 +14,7 @@
 #include <linux/spi/spi.h>
 #include <linux/irq.h>
 #include <linux/interrupt.h>
+#include <linux/gpio.h>
 #include <linux/delay.h>
 #include <asm/dma.h>
 #include <asm/bfin5xx_spi.h>
diff --git a/arch/blackfin/mach-bf609/boards/ezkit.c b/arch/blackfin/mach-bf609/boards/ezkit.c
index 1ba4600de69f..e2c0b024ce88 100644
--- a/arch/blackfin/mach-bf609/boards/ezkit.c
+++ b/arch/blackfin/mach-bf609/boards/ezkit.c
@@ -698,8 +698,6 @@ int bf609_nor_flash_init(struct platform_device *pdev)
 {
 #define CONFIG_SMC_GCTL_VAL     0x00000010
 
-	if (!devm_pinctrl_get_select_default(&pdev->dev))
-		return -EBUSY;
 	bfin_write32(SMC_GCTL, CONFIG_SMC_GCTL_VAL);
 	bfin_write32(SMC_B0CTL, 0x01002011);
 	bfin_write32(SMC_B0TIM, 0x08170977);
@@ -709,7 +707,6 @@ int bf609_nor_flash_init(struct platform_device *pdev)
 
 void bf609_nor_flash_exit(struct platform_device *pdev)
 {
-	devm_pinctrl_put(pdev->dev.pins->p);
 	bfin_write32(SMC_GCTL, 0);
 }
 
@@ -2058,15 +2055,14 @@ static struct pinctrl_map __initdata bfin_pinmux_map[] = {
 	PIN_MAP_MUX_GROUP_DEFAULT("bfin-rotary",  "pinctrl-adi2.0", NULL, "rotary"),
 	PIN_MAP_MUX_GROUP_DEFAULT("bfin_can.0",  "pinctrl-adi2.0", NULL, "can0"),
 	PIN_MAP_MUX_GROUP_DEFAULT("physmap-flash.0",  "pinctrl-adi2.0", NULL, "smc0"),
-	PIN_MAP_MUX_GROUP_DEFAULT("bf609_nl8048.2",  "pinctrl-adi2.0", NULL, "ppi2_16b"),
-	PIN_MAP_MUX_GROUP_DEFAULT("bfin_display.0",  "pinctrl-adi2.0", NULL, "ppi0_16b"),
-#if IS_ENABLED(CONFIG_VIDEO_MT9M114)
-	PIN_MAP_MUX_GROUP_DEFAULT("bfin_capture.0",  "pinctrl-adi2.0", NULL, "ppi0_8b"),
-#elif IS_ENABLED(CONFIG_VIDEO_VS6624)
-	PIN_MAP_MUX_GROUP_DEFAULT("bfin_capture.0",  "pinctrl-adi2.0", NULL, "ppi0_16b"),
-#else
-	PIN_MAP_MUX_GROUP_DEFAULT("bfin_capture.0",  "pinctrl-adi2.0", NULL, "ppi0_24b"),
-#endif
+	PIN_MAP_MUX_GROUP_DEFAULT("bf609_nl8048.2",  "pinctrl-adi2.0", "ppi2_16bgrp", "ppi2"),
+	PIN_MAP_MUX_GROUP("bfin_display.0", "8bit",  "pinctrl-adi2.0", "ppi2_8bgrp", "ppi2"),
+	PIN_MAP_MUX_GROUP_DEFAULT("bfin_display.0",  "pinctrl-adi2.0", "ppi2_16bgrp", "ppi2"),
+	PIN_MAP_MUX_GROUP("bfin_display.0", "16bit",  "pinctrl-adi2.0", "ppi2_16bgrp", "ppi2"),
+	PIN_MAP_MUX_GROUP("bfin_capture.0", "8bit",  "pinctrl-adi2.0", "ppi0_8bgrp", "ppi0"),
+	PIN_MAP_MUX_GROUP_DEFAULT("bfin_capture.0",  "pinctrl-adi2.0", "ppi0_16bgrp", "ppi0"),
+	PIN_MAP_MUX_GROUP("bfin_capture.0", "16bit",  "pinctrl-adi2.0", "ppi0_16bgrp", "ppi0"),
+	PIN_MAP_MUX_GROUP("bfin_capture.0", "24bit",  "pinctrl-adi2.0", "ppi0_24bgrp", "ppi0"),
 	PIN_MAP_MUX_GROUP_DEFAULT("bfin-i2s.0",  "pinctrl-adi2.0", NULL, "sport0"),
 	PIN_MAP_MUX_GROUP_DEFAULT("bfin-tdm.0",  "pinctrl-adi2.0", NULL, "sport0"),
 	PIN_MAP_MUX_GROUP_DEFAULT("bfin-i2s.1",  "pinctrl-adi2.0", NULL, "sport1"),
diff --git a/arch/blackfin/mach-bf609/include/mach/pm.h b/arch/blackfin/mach-bf609/include/mach/pm.h
index 3ca0fb965636..a1efd936dd30 100644
--- a/arch/blackfin/mach-bf609/include/mach/pm.h
+++ b/arch/blackfin/mach-bf609/include/mach/pm.h
@@ -10,6 +10,7 @@
 #define __MACH_BF609_PM_H__
 
 #include <linux/suspend.h>
+#include <linux/platform_device.h>
 
 extern int bfin609_pm_enter(suspend_state_t state);
 extern int bf609_pm_prepare(void);
@@ -19,6 +20,6 @@ void bf609_hibernate(void);
 void bfin_sec_raise_irq(unsigned int sid);
 void coreb_enable(void);
 
-int bf609_nor_flash_init(void);
-void bf609_nor_flash_exit(void);
+int bf609_nor_flash_init(struct platform_device *pdev);
+void bf609_nor_flash_exit(struct platform_device *pdev);
 #endif
diff --git a/arch/blackfin/mach-bf609/pm.c b/arch/blackfin/mach-bf609/pm.c
index 0cdd6955c7be..b1bfcf434d16 100644
--- a/arch/blackfin/mach-bf609/pm.c
+++ b/arch/blackfin/mach-bf609/pm.c
@@ -291,13 +291,13 @@ static struct bfin_cpu_pm_fns bf609_cpu_pm = {
 #if defined(CONFIG_MTD_PHYSMAP) || defined(CONFIG_MTD_PHYSMAP_MODULE)
 static int smc_pm_syscore_suspend(void)
 {
-	bf609_nor_flash_exit();
+	bf609_nor_flash_exit(NULL);
 	return 0;
 }
 
 static void smc_pm_syscore_resume(void)
 {
-	bf609_nor_flash_init();
+	bf609_nor_flash_init(NULL);
 }
 
 static struct syscore_ops smc_pm_syscore_ops = {
diff --git a/arch/blackfin/mach-common/ints-priority.c b/arch/blackfin/mach-common/ints-priority.c
index 867b7cef204c..1f94784eab6d 100644
--- a/arch/blackfin/mach-common/ints-priority.c
+++ b/arch/blackfin/mach-common/ints-priority.c
@@ -1208,8 +1208,6 @@ int __init init_arch_irq(void)
 
 	bfin_sec_set_priority(CONFIG_SEC_IRQ_PRIORITY_LEVELS, sec_int_priority);
 
-	bfin_sec_set_priority(CONFIG_SEC_IRQ_PRIORITY_LEVELS, sec_int_priority);
-
 	/* Enable interrupts IVG7-15 */
 	bfin_irq_flags |= IMASK_IVG15 |
 	    IMASK_IVG14 | IMASK_IVG13 | IMASK_IVG12 | IMASK_IVG11 |
diff --git a/arch/ia64/include/uapi/asm/fcntl.h b/arch/ia64/include/uapi/asm/fcntl.h
index 1dd275dc8f65..7b485876cad4 100644
--- a/arch/ia64/include/uapi/asm/fcntl.h
+++ b/arch/ia64/include/uapi/asm/fcntl.h
@@ -8,6 +8,7 @@
 #define force_o_largefile()	\
 		(personality(current->personality) != PER_LINUX32)
 
+#include <linux/personality.h>
 #include <asm-generic/fcntl.h>
 
 #endif /* _ASM_IA64_FCNTL_H */
diff --git a/arch/m68k/kernel/head.S b/arch/m68k/kernel/head.S
index dbb118e1a4e0..a54788458ca3 100644
--- a/arch/m68k/kernel/head.S
+++ b/arch/m68k/kernel/head.S
@@ -921,7 +921,8 @@ L(nocon):
 	jls	1f
 	lsrl	#1,%d1
 1:
-	movel	%d1,m68k_init_mapped_size
+	lea	%pc@(m68k_init_mapped_size),%a0
+	movel	%d1,%a0@
 	mmu_map	#PAGE_OFFSET,%pc@(L(phys_kernel_start)),%d1,\
 		%pc@(m68k_supervisor_cachemode)
 
diff --git a/arch/m68k/kernel/time.c b/arch/m68k/kernel/time.c
index 958f1adb9d0c..3857737e3958 100644
--- a/arch/m68k/kernel/time.c
+++ b/arch/m68k/kernel/time.c
@@ -11,6 +11,7 @@
  */
 
 #include <linux/errno.h>
+#include <linux/export.h>
 #include <linux/module.h>
 #include <linux/sched.h>
 #include <linux/kernel.h>
@@ -30,6 +31,7 @@
 
 
 unsigned long (*mach_random_get_entropy)(void);
+EXPORT_SYMBOL_GPL(mach_random_get_entropy);
 
 
 /*
diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig
index 7a469acee33c..4e238e6e661c 100644
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -269,6 +269,7 @@ config LANTIQ
 config LASAT
 	bool "LASAT Networks platforms"
 	select CEVT_R4K
+	select CRC32
 	select CSRC_R4K
 	select DMA_NONCOHERENT
 	select SYS_HAS_EARLY_PRINTK
diff --git a/arch/mips/include/asm/sigcontext.h b/arch/mips/include/asm/sigcontext.h
index f54bdbe85c0d..eeeb0f48c767 100644
--- a/arch/mips/include/asm/sigcontext.h
+++ b/arch/mips/include/asm/sigcontext.h
@@ -32,8 +32,6 @@ struct sigcontext32 {
 	__u32		sc_lo2;
 	__u32		sc_hi3;
 	__u32		sc_lo3;
-	__u64		sc_msaregs[32];	/* Most significant 64 bits */
-	__u32		sc_msa_csr;
 };
 #endif /* _MIPS_SIM == _MIPS_SIM_ABI64 || _MIPS_SIM == _MIPS_SIM_NABI32 */
 #endif /* _ASM_SIGCONTEXT_H */
diff --git a/arch/mips/include/asm/uasm.h b/arch/mips/include/asm/uasm.h
index f8d63b3b40b4..708c5d414905 100644
--- a/arch/mips/include/asm/uasm.h
+++ b/arch/mips/include/asm/uasm.h
@@ -67,6 +67,9 @@ void ISAOPC(op)(u32 **buf, unsigned int a, unsigned int b, signed int c)
 #define Ip_u2s3u1(op)							\
 void ISAOPC(op)(u32 **buf, unsigned int a, signed int b, unsigned int c)
 
+#define Ip_s3s1s2(op)							\
+void ISAOPC(op)(u32 **buf, int a, int b, int c)
+
 #define Ip_u2u1s3(op)							\
 void ISAOPC(op)(u32 **buf, unsigned int a, unsigned int b, signed int c)
 
@@ -147,6 +150,7 @@ Ip_u2s3u1(_scd);
 Ip_u2s3u1(_sd);
 Ip_u2u1u3(_sll);
 Ip_u3u2u1(_sllv);
+Ip_s3s1s2(_slt);
 Ip_u2u1s3(_sltiu);
 Ip_u3u1u2(_sltu);
 Ip_u2u1u3(_sra);
diff --git a/arch/mips/include/uapi/asm/inst.h b/arch/mips/include/uapi/asm/inst.h
index 4b7160259292..4bfdb9d4c186 100644
--- a/arch/mips/include/uapi/asm/inst.h
+++ b/arch/mips/include/uapi/asm/inst.h
@@ -273,6 +273,7 @@ enum mm_32a_minor_op {
 	mm_and_op = 0x250,
 	mm_or32_op = 0x290,
 	mm_xor32_op = 0x310,
+	mm_slt_op = 0x350,
 	mm_sltu_op = 0x390,
 };
 
diff --git a/arch/mips/include/uapi/asm/sigcontext.h b/arch/mips/include/uapi/asm/sigcontext.h
index 681c17603a48..6c9906f59c6e 100644
--- a/arch/mips/include/uapi/asm/sigcontext.h
+++ b/arch/mips/include/uapi/asm/sigcontext.h
@@ -12,10 +12,6 @@
 #include <linux/types.h>
 #include <asm/sgidefs.h>
 
-/* Bits which may be set in sc_used_math */
-#define USEDMATH_FP	(1 << 0)
-#define USEDMATH_MSA	(1 << 1)
-
 #if _MIPS_SIM == _MIPS_SIM_ABI32
 
 /*
@@ -41,8 +37,6 @@ struct sigcontext {
 	unsigned long		sc_lo2;
 	unsigned long		sc_hi3;
 	unsigned long		sc_lo3;
-	unsigned long long	sc_msaregs[32];	/* Most significant 64 bits */
-	unsigned long		sc_msa_csr;
 };
 
 #endif /* _MIPS_SIM == _MIPS_SIM_ABI32 */
@@ -76,8 +70,6 @@ struct sigcontext {
 	__u32	sc_used_math;
 	__u32	sc_dsp;
 	__u32	sc_reserved;
-	__u64	sc_msaregs[32];
-	__u32	sc_msa_csr;
 };
 
 
diff --git a/arch/mips/kernel/asm-offsets.c b/arch/mips/kernel/asm-offsets.c
index 02f075df8f2e..4bb5107511e2 100644
--- a/arch/mips/kernel/asm-offsets.c
+++ b/arch/mips/kernel/asm-offsets.c
@@ -293,7 +293,6 @@ void output_sc_defines(void)
 	OFFSET(SC_LO2, sigcontext, sc_lo2);
 	OFFSET(SC_HI3, sigcontext, sc_hi3);
 	OFFSET(SC_LO3, sigcontext, sc_lo3);
-	OFFSET(SC_MSAREGS, sigcontext, sc_msaregs);
 	BLANK();
 }
 #endif
@@ -308,7 +307,6 @@ void output_sc_defines(void)
 	OFFSET(SC_MDLO, sigcontext, sc_mdlo);
 	OFFSET(SC_PC, sigcontext, sc_pc);
 	OFFSET(SC_FPC_CSR, sigcontext, sc_fpc_csr);
-	OFFSET(SC_MSAREGS, sigcontext, sc_msaregs);
 	BLANK();
 }
 #endif
@@ -320,7 +318,6 @@ void output_sc32_defines(void)
 	OFFSET(SC32_FPREGS, sigcontext32, sc_fpregs);
 	OFFSET(SC32_FPC_CSR, sigcontext32, sc_fpc_csr);
 	OFFSET(SC32_FPC_EIR, sigcontext32, sc_fpc_eir);
-	OFFSET(SC32_MSAREGS, sigcontext32, sc_msaregs);
 	BLANK();
 }
 #endif
diff --git a/arch/mips/kernel/irq-msc01.c b/arch/mips/kernel/irq-msc01.c
index 4858642d543d..a734b2c2f9ea 100644
--- a/arch/mips/kernel/irq-msc01.c
+++ b/arch/mips/kernel/irq-msc01.c
@@ -126,7 +126,7 @@ void __init init_msc_irqs(unsigned long icubase, unsigned int irqbase, msc_irqma
 
 	board_bind_eic_interrupt = &msc_bind_eic_interrupt;
 
-	for (; nirq >= 0; nirq--, imp++) {
+	for (; nirq > 0; nirq--, imp++) {
 		int n = imp->im_irq;
 
 		switch (imp->im_type) {
diff --git a/arch/mips/kernel/pm-cps.c b/arch/mips/kernel/pm-cps.c
index 5aa4c6f8cf83..c4c2069d3a20 100644
--- a/arch/mips/kernel/pm-cps.c
+++ b/arch/mips/kernel/pm-cps.c
@@ -101,7 +101,7 @@ static void coupled_barrier(atomic_t *a, unsigned online)
 	if (!coupled_coherence)
 		return;
 
-	smp_mb__before_atomic_inc();
+	smp_mb__before_atomic();
 	atomic_inc(a);
 
 	while (atomic_read(a) < online)
@@ -158,7 +158,7 @@ int cps_pm_enter_state(enum cps_pm_state state)
 
 	/* Indicate that this CPU might not be coherent */
 	cpumask_clear_cpu(cpu, &cpu_coherent_mask);
-	smp_mb__after_clear_bit();
+	smp_mb__after_atomic();
 
 	/* Create a non-coherent mapping of the core ready_count */
 	core_ready_count = per_cpu(ready_count, core);
diff --git a/arch/mips/kernel/r4k_fpu.S b/arch/mips/kernel/r4k_fpu.S
index 71814272d148..8352523568e6 100644
--- a/arch/mips/kernel/r4k_fpu.S
+++ b/arch/mips/kernel/r4k_fpu.S
@@ -13,7 +13,6 @@
  * Copyright (C) 1999, 2001 Silicon Graphics, Inc.
  */
 #include <asm/asm.h>
-#include <asm/asmmacro.h>
 #include <asm/errno.h>
 #include <asm/fpregdef.h>
 #include <asm/mipsregs.h>
@@ -246,218 +245,6 @@ LEAF(_restore_fp_context32)
 	END(_restore_fp_context32)
 #endif
 
-#ifdef CONFIG_CPU_HAS_MSA
-
-	.macro	save_sc_msareg	wr, off, sc, tmp
-#ifdef CONFIG_64BIT
-	copy_u_d \tmp, \wr, 1
-	EX sd	\tmp, (\off+(\wr*8))(\sc)
-#elif defined(CONFIG_CPU_LITTLE_ENDIAN)
-	copy_u_w \tmp, \wr, 2
-	EX sw	\tmp, (\off+(\wr*8)+0)(\sc)
-	copy_u_w \tmp, \wr, 3
-	EX sw	\tmp, (\off+(\wr*8)+4)(\sc)
-#else /* CONFIG_CPU_BIG_ENDIAN */
-	copy_u_w \tmp, \wr, 2
-	EX sw	\tmp, (\off+(\wr*8)+4)(\sc)
-	copy_u_w \tmp, \wr, 3
-	EX sw	\tmp, (\off+(\wr*8)+0)(\sc)
-#endif
-	.endm
-
-/*
- * int _save_msa_context(struct sigcontext *sc)
- *
- * Save the upper 64 bits of each vector register along with the MSA_CSR
- * register into sc. Returns zero on success, else non-zero.
- */
-LEAF(_save_msa_context)
-	save_sc_msareg	0, SC_MSAREGS, a0, t0
-	save_sc_msareg	1, SC_MSAREGS, a0, t0
-	save_sc_msareg	2, SC_MSAREGS, a0, t0
-	save_sc_msareg	3, SC_MSAREGS, a0, t0
-	save_sc_msareg	4, SC_MSAREGS, a0, t0
-	save_sc_msareg	5, SC_MSAREGS, a0, t0
-	save_sc_msareg	6, SC_MSAREGS, a0, t0
-	save_sc_msareg	7, SC_MSAREGS, a0, t0
-	save_sc_msareg	8, SC_MSAREGS, a0, t0
-	save_sc_msareg	9, SC_MSAREGS, a0, t0
-	save_sc_msareg	10, SC_MSAREGS, a0, t0
-	save_sc_msareg	11, SC_MSAREGS, a0, t0
-	save_sc_msareg	12, SC_MSAREGS, a0, t0
-	save_sc_msareg	13, SC_MSAREGS, a0, t0
-	save_sc_msareg	14, SC_MSAREGS, a0, t0
-	save_sc_msareg	15, SC_MSAREGS, a0, t0
-	save_sc_msareg	16, SC_MSAREGS, a0, t0
-	save_sc_msareg	17, SC_MSAREGS, a0, t0
-	save_sc_msareg	18, SC_MSAREGS, a0, t0
-	save_sc_msareg	19, SC_MSAREGS, a0, t0
-	save_sc_msareg	20, SC_MSAREGS, a0, t0
-	save_sc_msareg	21, SC_MSAREGS, a0, t0
-	save_sc_msareg	22, SC_MSAREGS, a0, t0
-	save_sc_msareg	23, SC_MSAREGS, a0, t0
-	save_sc_msareg	24, SC_MSAREGS, a0, t0
-	save_sc_msareg	25, SC_MSAREGS, a0, t0
-	save_sc_msareg	26, SC_MSAREGS, a0, t0
-	save_sc_msareg	27, SC_MSAREGS, a0, t0
-	save_sc_msareg	28, SC_MSAREGS, a0, t0
-	save_sc_msareg	29, SC_MSAREGS, a0, t0
-	save_sc_msareg	30, SC_MSAREGS, a0, t0
-	save_sc_msareg	31, SC_MSAREGS, a0, t0
-	jr	ra
-	 li	v0, 0
-	END(_save_msa_context)
-
-#ifdef CONFIG_MIPS32_COMPAT
-
-/*
- * int _save_msa_context32(struct sigcontext32 *sc)
- *
- * Save the upper 64 bits of each vector register along with the MSA_CSR
- * register into sc. Returns zero on success, else non-zero.
- */
-LEAF(_save_msa_context32)
-	save_sc_msareg	0, SC32_MSAREGS, a0, t0
-	save_sc_msareg	1, SC32_MSAREGS, a0, t0
-	save_sc_msareg	2, SC32_MSAREGS, a0, t0
-	save_sc_msareg	3, SC32_MSAREGS, a0, t0
-	save_sc_msareg	4, SC32_MSAREGS, a0, t0
-	save_sc_msareg	5, SC32_MSAREGS, a0, t0
-	save_sc_msareg	6, SC32_MSAREGS, a0, t0
-	save_sc_msareg	7, SC32_MSAREGS, a0, t0
-	save_sc_msareg	8, SC32_MSAREGS, a0, t0
-	save_sc_msareg	9, SC32_MSAREGS, a0, t0
-	save_sc_msareg	10, SC32_MSAREGS, a0, t0
-	save_sc_msareg	11, SC32_MSAREGS, a0, t0
-	save_sc_msareg	12, SC32_MSAREGS, a0, t0
-	save_sc_msareg	13, SC32_MSAREGS, a0, t0
-	save_sc_msareg	14, SC32_MSAREGS, a0, t0
-	save_sc_msareg	15, SC32_MSAREGS, a0, t0
-	save_sc_msareg	16, SC32_MSAREGS, a0, t0
-	save_sc_msareg	17, SC32_MSAREGS, a0, t0
-	save_sc_msareg	18, SC32_MSAREGS, a0, t0
-	save_sc_msareg	19, SC32_MSAREGS, a0, t0
-	save_sc_msareg	20, SC32_MSAREGS, a0, t0
-	save_sc_msareg	21, SC32_MSAREGS, a0, t0
-	save_sc_msareg	22, SC32_MSAREGS, a0, t0
-	save_sc_msareg	23, SC32_MSAREGS, a0, t0
-	save_sc_msareg	24, SC32_MSAREGS, a0, t0
-	save_sc_msareg	25, SC32_MSAREGS, a0, t0
-	save_sc_msareg	26, SC32_MSAREGS, a0, t0
-	save_sc_msareg	27, SC32_MSAREGS, a0, t0
-	save_sc_msareg	28, SC32_MSAREGS, a0, t0
-	save_sc_msareg	29, SC32_MSAREGS, a0, t0
-	save_sc_msareg	30, SC32_MSAREGS, a0, t0
-	save_sc_msareg	31, SC32_MSAREGS, a0, t0
-	jr	ra
-	 li	v0, 0
-	END(_save_msa_context32)
-
-#endif /* CONFIG_MIPS32_COMPAT */
-
-	.macro restore_sc_msareg	wr, off, sc, tmp
-#ifdef CONFIG_64BIT
-	EX ld	\tmp, (\off+(\wr*8))(\sc)
-	insert_d \wr, 1, \tmp
-#elif defined(CONFIG_CPU_LITTLE_ENDIAN)
-	EX lw	\tmp, (\off+(\wr*8)+0)(\sc)
-	insert_w \wr, 2, \tmp
-	EX lw	\tmp, (\off+(\wr*8)+4)(\sc)
-	insert_w \wr, 3, \tmp
-#else /* CONFIG_CPU_BIG_ENDIAN */
-	EX lw	\tmp, (\off+(\wr*8)+4)(\sc)
-	insert_w \wr, 2, \tmp
-	EX lw	\tmp, (\off+(\wr*8)+0)(\sc)
-	insert_w \wr, 3, \tmp
-#endif
-	.endm
-
-/*
- * int _restore_msa_context(struct sigcontext *sc)
- */
-LEAF(_restore_msa_context)
-	restore_sc_msareg	0, SC_MSAREGS, a0, t0
-	restore_sc_msareg	1, SC_MSAREGS, a0, t0
-	restore_sc_msareg	2, SC_MSAREGS, a0, t0
-	restore_sc_msareg	3, SC_MSAREGS, a0, t0
-	restore_sc_msareg	4, SC_MSAREGS, a0, t0
-	restore_sc_msareg	5, SC_MSAREGS, a0, t0
-	restore_sc_msareg	6, SC_MSAREGS, a0, t0
-	restore_sc_msareg	7, SC_MSAREGS, a0, t0
-	restore_sc_msareg	8, SC_MSAREGS, a0, t0
-	restore_sc_msareg	9, SC_MSAREGS, a0, t0
-	restore_sc_msareg	10, SC_MSAREGS, a0, t0
-	restore_sc_msareg	11, SC_MSAREGS, a0, t0
-	restore_sc_msareg	12, SC_MSAREGS, a0, t0
-	restore_sc_msareg	13, SC_MSAREGS, a0, t0
-	restore_sc_msareg	14, SC_MSAREGS, a0, t0
-	restore_sc_msareg	15, SC_MSAREGS, a0, t0
-	restore_sc_msareg	16, SC_MSAREGS, a0, t0
-	restore_sc_msareg	17, SC_MSAREGS, a0, t0
-	restore_sc_msareg	18, SC_MSAREGS, a0, t0
-	restore_sc_msareg	19, SC_MSAREGS, a0, t0
-	restore_sc_msareg	20, SC_MSAREGS, a0, t0
-	restore_sc_msareg	21, SC_MSAREGS, a0, t0
-	restore_sc_msareg	22, SC_MSAREGS, a0, t0
-	restore_sc_msareg	23, SC_MSAREGS, a0, t0
-	restore_sc_msareg	24, SC_MSAREGS, a0, t0
-	restore_sc_msareg	25, SC_MSAREGS, a0, t0
-	restore_sc_msareg	26, SC_MSAREGS, a0, t0
-	restore_sc_msareg	27, SC_MSAREGS, a0, t0
-	restore_sc_msareg	28, SC_MSAREGS, a0, t0
-	restore_sc_msareg	29, SC_MSAREGS, a0, t0
-	restore_sc_msareg	30, SC_MSAREGS, a0, t0
-	restore_sc_msareg	31, SC_MSAREGS, a0, t0
-	jr	ra
-	 li	v0, 0
-	END(_restore_msa_context)
-
-#ifdef CONFIG_MIPS32_COMPAT
-
-/*
- * int _restore_msa_context32(struct sigcontext32 *sc)
- */
-LEAF(_restore_msa_context32)
-	restore_sc_msareg	0, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	1, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	2, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	3, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	4, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	5, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	6, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	7, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	8, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	9, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	10, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	11, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	12, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	13, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	14, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	15, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	16, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	17, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	18, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	19, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	20, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	21, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	22, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	23, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	24, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	25, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	26, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	27, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	28, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	29, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	30, SC32_MSAREGS, a0, t0
-	restore_sc_msareg	31, SC32_MSAREGS, a0, t0
-	jr	ra
-	 li	v0, 0
-	END(_restore_msa_context32)
-
-#endif /* CONFIG_MIPS32_COMPAT */
-
-#endif /* CONFIG_CPU_HAS_MSA */
-
 	.set	reorder
 
 	.type	fault@function
diff --git a/arch/mips/kernel/signal.c b/arch/mips/kernel/signal.c
index 33133d3df3e5..9e60d117e41e 100644
--- a/arch/mips/kernel/signal.c
+++ b/arch/mips/kernel/signal.c
@@ -31,7 +31,6 @@
 #include <linux/bitops.h>
 #include <asm/cacheflush.h>
 #include <asm/fpu.h>
-#include <asm/msa.h>
 #include <asm/sim.h>
 #include <asm/ucontext.h>
 #include <asm/cpu-features.h>
@@ -48,9 +47,6 @@ static int (*restore_fp_context)(struct sigcontext __user *sc);
 extern asmlinkage int _save_fp_context(struct sigcontext __user *sc);
 extern asmlinkage int _restore_fp_context(struct sigcontext __user *sc);
 
-extern asmlinkage int _save_msa_context(struct sigcontext __user *sc);
-extern asmlinkage int _restore_msa_context(struct sigcontext __user *sc);
-
 struct sigframe {
 	u32 sf_ass[4];		/* argument save space for o32 */
 	u32 sf_pad[2];		/* Was: signal trampoline */
@@ -100,60 +96,20 @@ static int copy_fp_from_sigcontext(struct sigcontext __user *sc)
 }
 
 /*
- * These functions will save only the upper 64 bits of the vector registers,
- * since the lower 64 bits have already been saved as the scalar FP context.
- */
-static int copy_msa_to_sigcontext(struct sigcontext __user *sc)
-{
-	int i;
-	int err = 0;
-
-	for (i = 0; i < NUM_FPU_REGS; i++) {
-		err |=
-		    __put_user(get_fpr64(&current->thread.fpu.fpr[i], 1),
-			       &sc->sc_msaregs[i]);
-	}
-	err |= __put_user(current->thread.fpu.msacsr, &sc->sc_msa_csr);
-
-	return err;
-}
-
-static int copy_msa_from_sigcontext(struct sigcontext __user *sc)
-{
-	int i;
-	int err = 0;
-	u64 val;
-
-	for (i = 0; i < NUM_FPU_REGS; i++) {
-		err |= __get_user(val, &sc->sc_msaregs[i]);
-		set_fpr64(&current->thread.fpu.fpr[i], 1, val);
-	}
-	err |= __get_user(current->thread.fpu.msacsr, &sc->sc_msa_csr);
-
-	return err;
-}
-
-/*
  * Helper routines
  */
-static int protected_save_fp_context(struct sigcontext __user *sc,
-				     unsigned used_math)
+static int protected_save_fp_context(struct sigcontext __user *sc)
 {
 	int err;
-	bool save_msa = cpu_has_msa && (used_math & USEDMATH_MSA);
 #ifndef CONFIG_EVA
 	while (1) {
 		lock_fpu_owner();
 		if (is_fpu_owner()) {
 			err = save_fp_context(sc);
-			if (save_msa && !err)
-				err = _save_msa_context(sc);
 			unlock_fpu_owner();
 		} else {
 			unlock_fpu_owner();
 			err = copy_fp_to_sigcontext(sc);
-			if (save_msa && !err)
-				err = copy_msa_to_sigcontext(sc);
 		}
 		if (likely(!err))
 			break;
@@ -169,38 +125,24 @@ static int protected_save_fp_context(struct sigcontext __user *sc,
 	 * EVA does not have FPU EVA instructions so saving fpu context directly
 	 * does not work.
 	 */
-	disable_msa();
 	lose_fpu(1);
 	err = save_fp_context(sc); /* this might fail */
-	if (save_msa && !err)
-		err = copy_msa_to_sigcontext(sc);
 #endif
 	return err;
 }
 
-static int protected_restore_fp_context(struct sigcontext __user *sc,
-					unsigned used_math)
+static int protected_restore_fp_context(struct sigcontext __user *sc)
 {
 	int err, tmp __maybe_unused;
-	bool restore_msa = cpu_has_msa && (used_math & USEDMATH_MSA);
 #ifndef CONFIG_EVA
 	while (1) {
 		lock_fpu_owner();
 		if (is_fpu_owner()) {
 			err = restore_fp_context(sc);
-			if (restore_msa && !err) {
-				enable_msa();
-				err = _restore_msa_context(sc);
-			} else {
-				/* signal handler may have used MSA */
-				disable_msa();
-			}
 			unlock_fpu_owner();
 		} else {
 			unlock_fpu_owner();
 			err = copy_fp_from_sigcontext(sc);
-			if (!err && (used_math & USEDMATH_MSA))
-				err = copy_msa_from_sigcontext(sc);
 		}
 		if (likely(!err))
 			break;
@@ -216,11 +158,8 @@ static int protected_restore_fp_context(struct sigcontext __user *sc,
 	 * EVA does not have FPU EVA instructions so restoring fpu context
 	 * directly does not work.
 	 */
-	enable_msa();
 	lose_fpu(0);
 	err = restore_fp_context(sc); /* this might fail */
-	if (restore_msa && !err)
-		err = copy_msa_from_sigcontext(sc);
 #endif
 	return err;
 }
@@ -252,8 +191,7 @@ int setup_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc)
 		err |= __put_user(rddsp(DSP_MASK), &sc->sc_dsp);
 	}
 
-	used_math = used_math() ? USEDMATH_FP : 0;
-	used_math |= thread_msa_context_live() ? USEDMATH_MSA : 0;
+	used_math = !!used_math();
 	err |= __put_user(used_math, &sc->sc_used_math);
 
 	if (used_math) {
@@ -261,7 +199,7 @@ int setup_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc)
 		 * Save FPU state to signal context. Signal handler
 		 * will "inherit" current FPU state.
 		 */
-		err |= protected_save_fp_context(sc, used_math);
+		err |= protected_save_fp_context(sc);
 	}
 	return err;
 }
@@ -286,14 +224,14 @@ int fpcsr_pending(unsigned int __user *fpcsr)
 }
 
 static int
-check_and_restore_fp_context(struct sigcontext __user *sc, unsigned used_math)
+check_and_restore_fp_context(struct sigcontext __user *sc)
 {
 	int err, sig;
 
 	err = sig = fpcsr_pending(&sc->sc_fpc_csr);
 	if (err > 0)
 		err = 0;
-	err |= protected_restore_fp_context(sc, used_math);
+	err |= protected_restore_fp_context(sc);
 	return err ?: sig;
 }
 
@@ -333,10 +271,9 @@ int restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc)
 	if (used_math) {
 		/* restore fpu context if we have used it before */
 		if (!err)
-			err = check_and_restore_fp_context(sc, used_math);
+			err = check_and_restore_fp_context(sc);
 	} else {
-		/* signal handler may have used FPU or MSA. Disable them. */
-		disable_msa();
+		/* signal handler may have used FPU.  Give it up. */
 		lose_fpu(0);
 	}
 
diff --git a/arch/mips/kernel/signal32.c b/arch/mips/kernel/signal32.c
index 299f956e4db3..bae2e6ee2109 100644
--- a/arch/mips/kernel/signal32.c
+++ b/arch/mips/kernel/signal32.c
@@ -30,7 +30,6 @@
 #include <asm/sim.h>
 #include <asm/ucontext.h>
 #include <asm/fpu.h>
-#include <asm/msa.h>
 #include <asm/war.h>
 #include <asm/vdso.h>
 #include <asm/dsp.h>
@@ -43,9 +42,6 @@ static int (*restore_fp_context32)(struct sigcontext32 __user *sc);
 extern asmlinkage int _save_fp_context32(struct sigcontext32 __user *sc);
 extern asmlinkage int _restore_fp_context32(struct sigcontext32 __user *sc);
 
-extern asmlinkage int _save_msa_context32(struct sigcontext32 __user *sc);
-extern asmlinkage int _restore_msa_context32(struct sigcontext32 __user *sc);
-
 /*
  * Including <asm/unistd.h> would give use the 64-bit syscall numbers ...
  */
@@ -115,59 +111,19 @@ static int copy_fp_from_sigcontext32(struct sigcontext32 __user *sc)
 }
 
 /*
- * These functions will save only the upper 64 bits of the vector registers,
- * since the lower 64 bits have already been saved as the scalar FP context.
- */
-static int copy_msa_to_sigcontext32(struct sigcontext32 __user *sc)
-{
-	int i;
-	int err = 0;
-
-	for (i = 0; i < NUM_FPU_REGS; i++) {
-		err |=
-		    __put_user(get_fpr64(&current->thread.fpu.fpr[i], 1),
-			       &sc->sc_msaregs[i]);
-	}
-	err |= __put_user(current->thread.fpu.msacsr, &sc->sc_msa_csr);
-
-	return err;
-}
-
-static int copy_msa_from_sigcontext32(struct sigcontext32 __user *sc)
-{
-	int i;
-	int err = 0;
-	u64 val;
-
-	for (i = 0; i < NUM_FPU_REGS; i++) {
-		err |= __get_user(val, &sc->sc_msaregs[i]);
-		set_fpr64(&current->thread.fpu.fpr[i], 1, val);
-	}
-	err |= __get_user(current->thread.fpu.msacsr, &sc->sc_msa_csr);
-
-	return err;
-}
-
-/*
  * sigcontext handlers
  */
-static int protected_save_fp_context32(struct sigcontext32 __user *sc,
-				       unsigned used_math)
+static int protected_save_fp_context32(struct sigcontext32 __user *sc)
 {
 	int err;
-	bool save_msa = cpu_has_msa && (used_math & USEDMATH_MSA);
 	while (1) {
 		lock_fpu_owner();
 		if (is_fpu_owner()) {
 			err = save_fp_context32(sc);
-			if (save_msa && !err)
-				err = _save_msa_context32(sc);
 			unlock_fpu_owner();
 		} else {
 			unlock_fpu_owner();
 			err = copy_fp_to_sigcontext32(sc);
-			if (save_msa && !err)
-				err = copy_msa_to_sigcontext32(sc);
 		}
 		if (likely(!err))
 			break;
@@ -181,28 +137,17 @@ static int protected_save_fp_context32(struct sigcontext32 __user *sc,
 	return err;
 }
 
-static int protected_restore_fp_context32(struct sigcontext32 __user *sc,
-					  unsigned used_math)
+static int protected_restore_fp_context32(struct sigcontext32 __user *sc)
 {
 	int err, tmp __maybe_unused;
-	bool restore_msa = cpu_has_msa && (used_math & USEDMATH_MSA);
 	while (1) {
 		lock_fpu_owner();
 		if (is_fpu_owner()) {
 			err = restore_fp_context32(sc);
-			if (restore_msa && !err) {
-				enable_msa();
-				err = _restore_msa_context32(sc);
-			} else {
-				/* signal handler may have used MSA */
-				disable_msa();
-			}
 			unlock_fpu_owner();
 		} else {
 			unlock_fpu_owner();
 			err = copy_fp_from_sigcontext32(sc);
-			if (restore_msa && !err)
-				err = copy_msa_from_sigcontext32(sc);
 		}
 		if (likely(!err))
 			break;
@@ -241,8 +186,7 @@ static int setup_sigcontext32(struct pt_regs *regs,
 		err |= __put_user(mflo3(), &sc->sc_lo3);
 	}
 
-	used_math = used_math() ? USEDMATH_FP : 0;
-	used_math |= thread_msa_context_live() ? USEDMATH_MSA : 0;
+	used_math = !!used_math();
 	err |= __put_user(used_math, &sc->sc_used_math);
 
 	if (used_math) {
@@ -250,21 +194,20 @@ static int setup_sigcontext32(struct pt_regs *regs,
 		 * Save FPU state to signal context.  Signal handler
 		 * will "inherit" current FPU state.
 		 */
-		err |= protected_save_fp_context32(sc, used_math);
+		err |= protected_save_fp_context32(sc);
 	}
 	return err;
 }
 
 static int
-check_and_restore_fp_context32(struct sigcontext32 __user *sc,
-			       unsigned used_math)
+check_and_restore_fp_context32(struct sigcontext32 __user *sc)
 {
 	int err, sig;
 
 	err = sig = fpcsr_pending(&sc->sc_fpc_csr);
 	if (err > 0)
 		err = 0;
-	err |= protected_restore_fp_context32(sc, used_math);
+	err |= protected_restore_fp_context32(sc);
 	return err ?: sig;
 }
 
@@ -301,10 +244,9 @@ static int restore_sigcontext32(struct pt_regs *regs,
 	if (used_math) {
 		/* restore fpu context if we have used it before */
 		if (!err)
-			err = check_and_restore_fp_context32(sc, used_math);
+			err = check_and_restore_fp_context32(sc);
 	} else {
-		/* signal handler may have used FPU or MSA. Disable them. */
-		disable_msa();
+		/* signal handler may have used FPU.  Give it up. */
 		lose_fpu(0);
 	}
 
diff --git a/arch/mips/kernel/smp-cps.c b/arch/mips/kernel/smp-cps.c
index df0598d9bfdd..949f2c6827a0 100644
--- a/arch/mips/kernel/smp-cps.c
+++ b/arch/mips/kernel/smp-cps.c
@@ -301,7 +301,7 @@ static int cps_cpu_disable(void)
 
 	core_cfg = &mips_cps_core_bootcfg[current_cpu_data.core];
 	atomic_sub(1 << cpu_vpe_id(&current_cpu_data), &core_cfg->vpe_mask);
-	smp_mb__after_atomic_dec();
+	smp_mb__after_atomic();
 	set_cpu_online(cpu, false);
 	cpu_clear(cpu, cpu_callin_map);
 
diff --git a/arch/mips/kvm/kvm_mips.c b/arch/mips/kvm/kvm_mips.c
index cd5e4f568439..f3c56a182fd8 100644
--- a/arch/mips/kvm/kvm_mips.c
+++ b/arch/mips/kvm/kvm_mips.c
@@ -384,6 +384,7 @@ void kvm_arch_vcpu_free(struct kvm_vcpu *vcpu)
 
 	kfree(vcpu->arch.guest_ebase);
 	kfree(vcpu->arch.kseg0_commpage);
+	kfree(vcpu);
 }
 
 void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
diff --git a/arch/mips/math-emu/ieee754.c b/arch/mips/math-emu/ieee754.c
index 53f1d2287084..8e97acbbe22c 100644
--- a/arch/mips/math-emu/ieee754.c
+++ b/arch/mips/math-emu/ieee754.c
@@ -34,13 +34,22 @@
  * Special constants
  */
 
-#define DPCNST(s, b, m)							\
+/*
+ * Older GCC requires the inner braces for initialization of union ieee754dp's
+ * anonymous struct member.  Without an error will result.
+ */
+#define xPCNST(s, b, m, ebias)						\
 {									\
-	.sign	= (s),							\
-	.bexp	= (b) + DP_EBIAS,					\
-	.mant	= (m)							\
+	{								\
+		.sign	= (s),						\
+		.bexp	= (b) + ebias,					\
+		.mant	= (m)						\
+	}								\
 }
 
+#define DPCNST(s, b, m)							\
+	xPCNST(s, b, m, DP_EBIAS)
+
 const union ieee754dp __ieee754dp_spcvals[] = {
 	DPCNST(0, DP_EMIN - 1, 0x0000000000000ULL),	/* + zero   */
 	DPCNST(1, DP_EMIN - 1, 0x0000000000000ULL),	/* - zero   */
@@ -62,11 +71,7 @@ const union ieee754dp __ieee754dp_spcvals[] = {
 };
 
 #define SPCNST(s, b, m)							\
-{									\
-	.sign	= (s),							\
-	.bexp	= (b) + SP_EBIAS,					\
-	.mant	= (m)							\
-}
+	xPCNST(s, b, m, SP_EBIAS)
 
 const union ieee754sp __ieee754sp_spcvals[] = {
 	SPCNST(0, SP_EMIN - 1, 0x000000),	/* + zero   */
diff --git a/arch/mips/mm/uasm-micromips.c b/arch/mips/mm/uasm-micromips.c
index 775c2800cba2..8399ddf03a02 100644
--- a/arch/mips/mm/uasm-micromips.c
+++ b/arch/mips/mm/uasm-micromips.c
@@ -102,6 +102,7 @@ static struct insn insn_table_MM[] = {
 	{ insn_sd, 0, 0 },
 	{ insn_sll, M(mm_pool32a_op, 0, 0, 0, 0, mm_sll32_op), RT | RS | RD },
 	{ insn_sllv, M(mm_pool32a_op, 0, 0, 0, 0, mm_sllv32_op), RT | RS | RD },
+	{ insn_slt, M(mm_pool32a_op, 0, 0, 0, 0, mm_slt_op), RT | RS | RD },
 	{ insn_sltiu, M(mm_sltiu32_op, 0, 0, 0, 0, 0), RT | RS | SIMM },
 	{ insn_sltu, M(mm_pool32a_op, 0, 0, 0, 0, mm_sltu_op), RT | RS | RD },
 	{ insn_sra, M(mm_pool32a_op, 0, 0, 0, 0, mm_sra_op), RT | RS | RD },
diff --git a/arch/mips/mm/uasm-mips.c b/arch/mips/mm/uasm-mips.c
index 38792c2364f5..6708a2dbf934 100644
--- a/arch/mips/mm/uasm-mips.c
+++ b/arch/mips/mm/uasm-mips.c
@@ -89,7 +89,7 @@ static struct insn insn_table[] = {
 	{ insn_lb, M(lb_op, 0, 0, 0, 0, 0), RS | RT | SIMM },
 	{ insn_ld,  M(ld_op, 0, 0, 0, 0, 0),  RS | RT | SIMM },
 	{ insn_ldx, M(spec3_op, 0, 0, 0, ldx_op, lx_op), RS | RT | RD },
-	{ insn_lh,  M(lw_op, 0, 0, 0, 0, 0),  RS | RT | SIMM },
+	{ insn_lh,  M(lh_op, 0, 0, 0, 0, 0),  RS | RT | SIMM },
 	{ insn_lld,  M(lld_op, 0, 0, 0, 0, 0),	RS | RT | SIMM },
 	{ insn_ll,  M(ll_op, 0, 0, 0, 0, 0),  RS | RT | SIMM },
 	{ insn_lui,  M(lui_op, 0, 0, 0, 0, 0),	RT | SIMM },
@@ -110,6 +110,7 @@ static struct insn insn_table[] = {
 	{ insn_sd,  M(sd_op, 0, 0, 0, 0, 0),  RS | RT | SIMM },
 	{ insn_sll,  M(spec_op, 0, 0, 0, 0, sll_op),  RT | RD | RE },
 	{ insn_sllv,  M(spec_op, 0, 0, 0, 0, sllv_op),  RS | RT | RD },
+	{ insn_slt,  M(spec_op, 0, 0, 0, 0, slt_op),  RS | RT | RD },
 	{ insn_sltiu, M(sltiu_op, 0, 0, 0, 0, 0), RS | RT | SIMM },
 	{ insn_sltu, M(spec_op, 0, 0, 0, 0, sltu_op), RS | RT | RD },
 	{ insn_sra,  M(spec_op, 0, 0, 0, 0, sra_op),  RT | RD | RE },
diff --git a/arch/mips/mm/uasm.c b/arch/mips/mm/uasm.c
index 00515805fe41..a01b0d6cedd2 100644
--- a/arch/mips/mm/uasm.c
+++ b/arch/mips/mm/uasm.c
@@ -53,7 +53,7 @@ enum opcode {
 	insn_ld, insn_ldx, insn_lh, insn_ll, insn_lld, insn_lui, insn_lw,
 	insn_lwx, insn_mfc0, insn_mfhi, insn_mflo, insn_mtc0, insn_mul,
 	insn_or, insn_ori, insn_pref, insn_rfe, insn_rotr, insn_sc, insn_scd,
-	insn_sd, insn_sll, insn_sllv, insn_sltiu, insn_sltu, insn_sra,
+	insn_sd, insn_sll, insn_sllv, insn_slt, insn_sltiu, insn_sltu, insn_sra,
 	insn_srl, insn_srlv, insn_subu, insn_sw, insn_sync, insn_syscall,
 	insn_tlbp, insn_tlbr, insn_tlbwi, insn_tlbwr, insn_wait, insn_wsbh,
 	insn_xor, insn_xori, insn_yield,
@@ -139,6 +139,13 @@ Ip_u1u2u3(op)						\
 }							\
 UASM_EXPORT_SYMBOL(uasm_i##op);
 
+#define I_s3s1s2(op)					\
+Ip_s3s1s2(op)						\
+{							\
+	build_insn(buf, insn##op, b, c, a);		\
+}							\
+UASM_EXPORT_SYMBOL(uasm_i##op);
+
 #define I_u2u1u3(op)					\
 Ip_u2u1u3(op)						\
 {							\
@@ -289,6 +296,7 @@ I_u2s3u1(_scd)
 I_u2s3u1(_sd)
 I_u2u1u3(_sll)
 I_u3u2u1(_sllv)
+I_s3s1s2(_slt)
 I_u2u1s3(_sltiu)
 I_u3u1u2(_sltu)
 I_u2u1u3(_sra)
diff --git a/arch/mips/net/bpf_jit.c b/arch/mips/net/bpf_jit.c
index a67b9753330b..b87390a56a2f 100644
--- a/arch/mips/net/bpf_jit.c
+++ b/arch/mips/net/bpf_jit.c
@@ -119,8 +119,6 @@
 /* Arguments used by JIT */
 #define ARGS_USED_BY_JIT	2 /* only applicable to 64-bit */
 
-#define FLAG_NEED_X_RESET	(1 << 0)
-
 #define SBIT(x)			(1 << (x)) /* Signed version of BIT() */
 
 /**
@@ -153,6 +151,8 @@ static inline int optimize_div(u32 *k)
 	return 0;
 }
 
+static inline void emit_jit_reg_move(ptr dst, ptr src, struct jit_ctx *ctx);
+
 /* Simply emit the instruction if the JIT memory space has been allocated */
 #define emit_instr(ctx, func, ...)			\
 do {							\
@@ -166,9 +166,7 @@ do {							\
 /* Determine if immediate is within the 16-bit signed range */
 static inline bool is_range16(s32 imm)
 {
-	if (imm >= SBIT(15) || imm < -SBIT(15))
-		return true;
-	return false;
+	return !(imm >= SBIT(15) || imm < -SBIT(15));
 }
 
 static inline void emit_addu(unsigned int dst, unsigned int src1,
@@ -187,7 +185,7 @@ static inline void emit_load_imm(unsigned int dst, u32 imm, struct jit_ctx *ctx)
 {
 	if (ctx->target != NULL) {
 		/* addiu can only handle s16 */
-		if (is_range16(imm)) {
+		if (!is_range16(imm)) {
 			u32 *p = &ctx->target[ctx->idx];
 			uasm_i_lui(&p, r_tmp_imm, (s32)imm >> 16);
 			p = &ctx->target[ctx->idx + 1];
@@ -199,7 +197,7 @@ static inline void emit_load_imm(unsigned int dst, u32 imm, struct jit_ctx *ctx)
 	}
 	ctx->idx++;
 
-	if (is_range16(imm))
+	if (!is_range16(imm))
 		ctx->idx++;
 }
 
@@ -240,7 +238,7 @@ static inline void emit_daddiu(unsigned int dst, unsigned int src,
 static inline void emit_addiu(unsigned int dst, unsigned int src,
 			      u32 imm, struct jit_ctx *ctx)
 {
-	if (is_range16(imm)) {
+	if (!is_range16(imm)) {
 		emit_load_imm(r_tmp, imm, ctx);
 		emit_addu(dst, r_tmp, src, ctx);
 	} else {
@@ -313,8 +311,11 @@ static inline void emit_sll(unsigned int dst, unsigned int src,
 			    unsigned int sa, struct jit_ctx *ctx)
 {
 	/* sa is 5-bits long */
-	BUG_ON(sa >= BIT(5));
-	emit_instr(ctx, sll, dst, src, sa);
+	if (sa >= BIT(5))
+		/* Shifting >= 32 results in zero */
+		emit_jit_reg_move(dst, r_zero, ctx);
+	else
+		emit_instr(ctx, sll, dst, src, sa);
 }
 
 static inline void emit_srlv(unsigned int dst, unsigned int src,
@@ -327,8 +328,17 @@ static inline void emit_srl(unsigned int dst, unsigned int src,
 			    unsigned int sa, struct jit_ctx *ctx)
 {
 	/* sa is 5-bits long */
-	BUG_ON(sa >= BIT(5));
-	emit_instr(ctx, srl, dst, src, sa);
+	if (sa >= BIT(5))
+		/* Shifting >= 32 results in zero */
+		emit_jit_reg_move(dst, r_zero, ctx);
+	else
+		emit_instr(ctx, srl, dst, src, sa);
+}
+
+static inline void emit_slt(unsigned int dst, unsigned int src1,
+			    unsigned int src2, struct jit_ctx *ctx)
+{
+	emit_instr(ctx, slt, dst, src1, src2);
 }
 
 static inline void emit_sltu(unsigned int dst, unsigned int src1,
@@ -341,7 +351,7 @@ static inline void emit_sltiu(unsigned dst, unsigned int src,
 			      unsigned int imm, struct jit_ctx *ctx)
 {
 	/* 16 bit immediate */
-	if (is_range16((s32)imm)) {
+	if (!is_range16((s32)imm)) {
 		emit_load_imm(r_tmp, imm, ctx);
 		emit_sltu(dst, src, r_tmp, ctx);
 	} else {
@@ -408,7 +418,7 @@ static inline void emit_div(unsigned int dst, unsigned int src,
 		u32 *p = &ctx->target[ctx->idx];
 		uasm_i_divu(&p, dst, src);
 		p = &ctx->target[ctx->idx + 1];
-		uasm_i_mfhi(&p, dst);
+		uasm_i_mflo(&p, dst);
 	}
 	ctx->idx += 2; /* 2 insts */
 }
@@ -443,6 +453,17 @@ static inline void emit_wsbh(unsigned int dst, unsigned int src,
 	emit_instr(ctx, wsbh, dst, src);
 }
 
+/* load pointer to register */
+static inline void emit_load_ptr(unsigned int dst, unsigned int src,
+				     int imm, struct jit_ctx *ctx)
+{
+	/* src contains the base addr of the 32/64-pointer */
+	if (config_enabled(CONFIG_64BIT))
+		emit_instr(ctx, ld, dst, imm, src);
+	else
+		emit_instr(ctx, lw, dst, imm, src);
+}
+
 /* load a function pointer to register */
 static inline void emit_load_func(unsigned int reg, ptr imm,
 				  struct jit_ctx *ctx)
@@ -545,29 +566,13 @@ static inline u16 align_sp(unsigned int num)
 	return num;
 }
 
-static inline void update_on_xread(struct jit_ctx *ctx)
-{
-	if (!(ctx->flags & SEEN_X))
-		ctx->flags |= FLAG_NEED_X_RESET;
-
-	ctx->flags |= SEEN_X;
-}
-
 static bool is_load_to_a(u16 inst)
 {
 	switch (inst) {
-	case BPF_S_LD_W_LEN:
-	case BPF_S_LD_W_ABS:
-	case BPF_S_LD_H_ABS:
-	case BPF_S_LD_B_ABS:
-	case BPF_S_ANC_CPU:
-	case BPF_S_ANC_IFINDEX:
-	case BPF_S_ANC_MARK:
-	case BPF_S_ANC_PROTOCOL:
-	case BPF_S_ANC_RXHASH:
-	case BPF_S_ANC_VLAN_TAG:
-	case BPF_S_ANC_VLAN_TAG_PRESENT:
-	case BPF_S_ANC_QUEUE:
+	case BPF_LD | BPF_W | BPF_LEN:
+	case BPF_LD | BPF_W | BPF_ABS:
+	case BPF_LD | BPF_H | BPF_ABS:
+	case BPF_LD | BPF_B | BPF_ABS:
 		return true;
 	default:
 		return false;
@@ -618,7 +623,10 @@ static void save_bpf_jit_regs(struct jit_ctx *ctx, unsigned offset)
 	if (ctx->flags & SEEN_MEM) {
 		if (real_off % (RSIZE * 2))
 			real_off += RSIZE;
-		emit_addiu(r_M, r_sp, real_off, ctx);
+		if (config_enabled(CONFIG_64BIT))
+			emit_daddiu(r_M, r_sp, real_off, ctx);
+		else
+			emit_addiu(r_M, r_sp, real_off, ctx);
 	}
 }
 
@@ -705,11 +713,11 @@ static void build_prologue(struct jit_ctx *ctx)
 	if (ctx->flags & SEEN_SKB)
 		emit_reg_move(r_skb, MIPS_R_A0, ctx);
 
-	if (ctx->flags & FLAG_NEED_X_RESET)
+	if (ctx->flags & SEEN_X)
 		emit_jit_reg_move(r_X, r_zero, ctx);
 
 	/* Do not leak kernel data to userspace */
-	if ((first_inst != BPF_S_RET_K) && !(is_load_to_a(first_inst)))
+	if ((first_inst != (BPF_RET | BPF_K)) && !(is_load_to_a(first_inst)))
 		emit_jit_reg_move(r_A, r_zero, ctx);
 }
 
@@ -757,13 +765,17 @@ static u64 jit_get_skb_w(struct sk_buff *skb, unsigned offset)
 	return (u64)err << 32 | ntohl(ret);
 }
 
-#define PKT_TYPE_MAX 7
+#ifdef __BIG_ENDIAN_BITFIELD
+#define PKT_TYPE_MAX	(7 << 5)
+#else
+#define PKT_TYPE_MAX	7
+#endif
 static int pkt_type_offset(void)
 {
 	struct sk_buff skb_probe = {
 		.pkt_type = ~0,
 	};
-	char *ct = (char *)&skb_probe;
+	u8 *ct = (u8 *)&skb_probe;
 	unsigned int off;
 
 	for (off = 0; off < sizeof(struct sk_buff); off++) {
@@ -783,46 +795,62 @@ static int build_body(struct jit_ctx *ctx)
 	u32 k, b_off __maybe_unused;
 
 	for (i = 0; i < prog->len; i++) {
+		u16 code;
+
 		inst = &(prog->insns[i]);
 		pr_debug("%s: code->0x%02x, jt->0x%x, jf->0x%x, k->0x%x\n",
 			 __func__, inst->code, inst->jt, inst->jf, inst->k);
 		k = inst->k;
+		code = bpf_anc_helper(inst);
 
 		if (ctx->target == NULL)
 			ctx->offsets[i] = ctx->idx * 4;
 
-		switch (inst->code) {
-		case BPF_S_LD_IMM:
+		switch (code) {
+		case BPF_LD | BPF_IMM:
 			/* A <- k ==> li r_A, k */
 			ctx->flags |= SEEN_A;
 			emit_load_imm(r_A, k, ctx);
 			break;
-		case BPF_S_LD_W_LEN:
+		case BPF_LD | BPF_W | BPF_LEN:
 			BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, len) != 4);
 			/* A <- len ==> lw r_A, offset(skb) */
 			ctx->flags |= SEEN_SKB | SEEN_A;
 			off = offsetof(struct sk_buff, len);
 			emit_load(r_A, r_skb, off, ctx);
 			break;
-		case BPF_S_LD_MEM:
+		case BPF_LD | BPF_MEM:
 			/* A <- M[k] ==> lw r_A, offset(M) */
 			ctx->flags |= SEEN_MEM | SEEN_A;
 			emit_load(r_A, r_M, SCRATCH_OFF(k), ctx);
 			break;
-		case BPF_S_LD_W_ABS:
+		case BPF_LD | BPF_W | BPF_ABS:
 			/* A <- P[k:4] */
 			load_order = 2;
 			goto load;
-		case BPF_S_LD_H_ABS:
+		case BPF_LD | BPF_H | BPF_ABS:
 			/* A <- P[k:2] */
 			load_order = 1;
 			goto load;
-		case BPF_S_LD_B_ABS:
+		case BPF_LD | BPF_B | BPF_ABS:
 			/* A <- P[k:1] */
 			load_order = 0;
 load:
+			/* the interpreter will deal with the negative K */
+			if ((int)k < 0)
+				return -ENOTSUPP;
+
 			emit_load_imm(r_off, k, ctx);
 load_common:
+			/*
+			 * We may got here from the indirect loads so
+			 * return if offset is negative.
+			 */
+			emit_slt(r_s0, r_off, r_zero, ctx);
+			emit_bcond(MIPS_COND_NE, r_s0, r_zero,
+				   b_imm(prog->len, ctx), ctx);
+			emit_reg_move(r_ret, r_zero, ctx);
+
 			ctx->flags |= SEEN_CALL | SEEN_OFF | SEEN_S0 |
 				SEEN_SKB | SEEN_A;
 
@@ -852,39 +880,42 @@ load_common:
 			emit_b(b_imm(prog->len, ctx), ctx);
 			emit_reg_move(r_ret, r_zero, ctx);
 			break;
-		case BPF_S_LD_W_IND:
+		case BPF_LD | BPF_W | BPF_IND:
 			/* A <- P[X + k:4] */
 			load_order = 2;
 			goto load_ind;
-		case BPF_S_LD_H_IND:
+		case BPF_LD | BPF_H | BPF_IND:
 			/* A <- P[X + k:2] */
 			load_order = 1;
 			goto load_ind;
-		case BPF_S_LD_B_IND:
+		case BPF_LD | BPF_B | BPF_IND:
 			/* A <- P[X + k:1] */
 			load_order = 0;
 load_ind:
-			update_on_xread(ctx);
 			ctx->flags |= SEEN_OFF | SEEN_X;
 			emit_addiu(r_off, r_X, k, ctx);
 			goto load_common;
-		case BPF_S_LDX_IMM:
+		case BPF_LDX | BPF_IMM:
 			/* X <- k */
 			ctx->flags |= SEEN_X;
 			emit_load_imm(r_X, k, ctx);
 			break;
-		case BPF_S_LDX_MEM:
+		case BPF_LDX | BPF_MEM:
 			/* X <- M[k] */
 			ctx->flags |= SEEN_X | SEEN_MEM;
 			emit_load(r_X, r_M, SCRATCH_OFF(k), ctx);
 			break;
-		case BPF_S_LDX_W_LEN:
+		case BPF_LDX | BPF_W | BPF_LEN:
 			/* X <- len */
 			ctx->flags |= SEEN_X | SEEN_SKB;
 			off = offsetof(struct sk_buff, len);
 			emit_load(r_X, r_skb, off, ctx);
 			break;
-		case BPF_S_LDX_B_MSH:
+		case BPF_LDX | BPF_B | BPF_MSH:
+			/* the interpreter will deal with the negative K */
+			if ((int)k < 0)
+				return -ENOTSUPP;
+
 			/* X <- 4 * (P[k:1] & 0xf) */
 			ctx->flags |= SEEN_X | SEEN_CALL | SEEN_S0 | SEEN_SKB;
 			/* Load offset to a1 */
@@ -917,50 +948,49 @@ load_ind:
 			emit_b(b_imm(prog->len, ctx), ctx);
 			emit_load_imm(r_ret, 0, ctx); /* delay slot */
 			break;
-		case BPF_S_ST:
+		case BPF_ST:
 			/* M[k] <- A */
 			ctx->flags |= SEEN_MEM | SEEN_A;
 			emit_store(r_A, r_M, SCRATCH_OFF(k), ctx);
 			break;
-		case BPF_S_STX:
+		case BPF_STX:
 			/* M[k] <- X */
 			ctx->flags |= SEEN_MEM | SEEN_X;
 			emit_store(r_X, r_M, SCRATCH_OFF(k), ctx);
 			break;
-		case BPF_S_ALU_ADD_K:
+		case BPF_ALU | BPF_ADD | BPF_K:
 			/* A += K */
 			ctx->flags |= SEEN_A;
 			emit_addiu(r_A, r_A, k, ctx);
 			break;
-		case BPF_S_ALU_ADD_X:
+		case BPF_ALU | BPF_ADD | BPF_X:
 			/* A += X */
 			ctx->flags |= SEEN_A | SEEN_X;
 			emit_addu(r_A, r_A, r_X, ctx);
 			break;
-		case BPF_S_ALU_SUB_K:
+		case BPF_ALU | BPF_SUB | BPF_K:
 			/* A -= K */
 			ctx->flags |= SEEN_A;
 			emit_addiu(r_A, r_A, -k, ctx);
 			break;
-		case BPF_S_ALU_SUB_X:
+		case BPF_ALU | BPF_SUB | BPF_X:
 			/* A -= X */
 			ctx->flags |= SEEN_A | SEEN_X;
 			emit_subu(r_A, r_A, r_X, ctx);
 			break;
-		case BPF_S_ALU_MUL_K:
+		case BPF_ALU | BPF_MUL | BPF_K:
 			/* A *= K */
 			/* Load K to scratch register before MUL */
 			ctx->flags |= SEEN_A | SEEN_S0;
 			emit_load_imm(r_s0, k, ctx);
 			emit_mul(r_A, r_A, r_s0, ctx);
 			break;
-		case BPF_S_ALU_MUL_X:
+		case BPF_ALU | BPF_MUL | BPF_X:
 			/* A *= X */
-			update_on_xread(ctx);
 			ctx->flags |= SEEN_A | SEEN_X;
 			emit_mul(r_A, r_A, r_X, ctx);
 			break;
-		case BPF_S_ALU_DIV_K:
+		case BPF_ALU | BPF_DIV | BPF_K:
 			/* A /= k */
 			if (k == 1)
 				break;
@@ -973,7 +1003,7 @@ load_ind:
 			emit_load_imm(r_s0, k, ctx);
 			emit_div(r_A, r_s0, ctx);
 			break;
-		case BPF_S_ALU_MOD_K:
+		case BPF_ALU | BPF_MOD | BPF_K:
 			/* A %= k */
 			if (k == 1 || optimize_div(&k)) {
 				ctx->flags |= SEEN_A;
@@ -984,9 +1014,8 @@ load_ind:
 				emit_mod(r_A, r_s0, ctx);
 			}
 			break;
-		case BPF_S_ALU_DIV_X:
+		case BPF_ALU | BPF_DIV | BPF_X:
 			/* A /= X */
-			update_on_xread(ctx);
 			ctx->flags |= SEEN_X | SEEN_A;
 			/* Check if r_X is zero */
 			emit_bcond(MIPS_COND_EQ, r_X, r_zero,
@@ -994,9 +1023,8 @@ load_ind:
 			emit_load_imm(r_val, 0, ctx); /* delay slot */
 			emit_div(r_A, r_X, ctx);
 			break;
-		case BPF_S_ALU_MOD_X:
+		case BPF_ALU | BPF_MOD | BPF_X:
 			/* A %= X */
-			update_on_xread(ctx);
 			ctx->flags |= SEEN_X | SEEN_A;
 			/* Check if r_X is zero */
 			emit_bcond(MIPS_COND_EQ, r_X, r_zero,
@@ -1004,94 +1032,89 @@ load_ind:
 			emit_load_imm(r_val, 0, ctx); /* delay slot */
 			emit_mod(r_A, r_X, ctx);
 			break;
-		case BPF_S_ALU_OR_K:
+		case BPF_ALU | BPF_OR | BPF_K:
 			/* A |= K */
 			ctx->flags |= SEEN_A;
 			emit_ori(r_A, r_A, k, ctx);
 			break;
-		case BPF_S_ALU_OR_X:
+		case BPF_ALU | BPF_OR | BPF_X:
 			/* A |= X */
-			update_on_xread(ctx);
 			ctx->flags |= SEEN_A;
 			emit_ori(r_A, r_A, r_X, ctx);
 			break;
-		case BPF_S_ALU_XOR_K:
+		case BPF_ALU | BPF_XOR | BPF_K:
 			/* A ^= k */
 			ctx->flags |= SEEN_A;
 			emit_xori(r_A, r_A, k, ctx);
 			break;
-		case BPF_S_ANC_ALU_XOR_X:
-		case BPF_S_ALU_XOR_X:
+		case BPF_ANC | SKF_AD_ALU_XOR_X:
+		case BPF_ALU | BPF_XOR | BPF_X:
 			/* A ^= X */
-			update_on_xread(ctx);
 			ctx->flags |= SEEN_A;
 			emit_xor(r_A, r_A, r_X, ctx);
 			break;
-		case BPF_S_ALU_AND_K:
+		case BPF_ALU | BPF_AND | BPF_K:
 			/* A &= K */
 			ctx->flags |= SEEN_A;
 			emit_andi(r_A, r_A, k, ctx);
 			break;
-		case BPF_S_ALU_AND_X:
+		case BPF_ALU | BPF_AND | BPF_X:
 			/* A &= X */
-			update_on_xread(ctx);
 			ctx->flags |= SEEN_A | SEEN_X;
 			emit_and(r_A, r_A, r_X, ctx);
 			break;
-		case BPF_S_ALU_LSH_K:
+		case BPF_ALU | BPF_LSH | BPF_K:
 			/* A <<= K */
 			ctx->flags |= SEEN_A;
 			emit_sll(r_A, r_A, k, ctx);
 			break;
-		case BPF_S_ALU_LSH_X:
+		case BPF_ALU | BPF_LSH | BPF_X:
 			/* A <<= X */
 			ctx->flags |= SEEN_A | SEEN_X;
-			update_on_xread(ctx);
 			emit_sllv(r_A, r_A, r_X, ctx);
 			break;
-		case BPF_S_ALU_RSH_K:
+		case BPF_ALU | BPF_RSH | BPF_K:
 			/* A >>= K */
 			ctx->flags |= SEEN_A;
 			emit_srl(r_A, r_A, k, ctx);
 			break;
-		case BPF_S_ALU_RSH_X:
+		case BPF_ALU | BPF_RSH | BPF_X:
 			ctx->flags |= SEEN_A | SEEN_X;
-			update_on_xread(ctx);
 			emit_srlv(r_A, r_A, r_X, ctx);
 			break;
-		case BPF_S_ALU_NEG:
+		case BPF_ALU | BPF_NEG:
 			/* A = -A */
 			ctx->flags |= SEEN_A;
 			emit_neg(r_A, ctx);
 			break;
-		case BPF_S_JMP_JA:
+		case BPF_JMP | BPF_JA:
 			/* pc += K */
 			emit_b(b_imm(i + k + 1, ctx), ctx);
 			emit_nop(ctx);
 			break;
-		case BPF_S_JMP_JEQ_K:
+		case BPF_JMP | BPF_JEQ | BPF_K:
 			/* pc += ( A == K ) ? pc->jt : pc->jf */
 			condt = MIPS_COND_EQ | MIPS_COND_K;
 			goto jmp_cmp;
-		case BPF_S_JMP_JEQ_X:
+		case BPF_JMP | BPF_JEQ | BPF_X:
 			ctx->flags |= SEEN_X;
 			/* pc += ( A == X ) ? pc->jt : pc->jf */
 			condt = MIPS_COND_EQ | MIPS_COND_X;
 			goto jmp_cmp;
-		case BPF_S_JMP_JGE_K:
+		case BPF_JMP | BPF_JGE | BPF_K:
 			/* pc += ( A >= K ) ? pc->jt : pc->jf */
 			condt = MIPS_COND_GE | MIPS_COND_K;
 			goto jmp_cmp;
-		case BPF_S_JMP_JGE_X:
+		case BPF_JMP | BPF_JGE | BPF_X:
 			ctx->flags |= SEEN_X;
 			/* pc += ( A >= X ) ? pc->jt : pc->jf */
 			condt = MIPS_COND_GE | MIPS_COND_X;
 			goto jmp_cmp;
-		case BPF_S_JMP_JGT_K:
+		case BPF_JMP | BPF_JGT | BPF_K:
 			/* pc += ( A > K ) ? pc->jt : pc->jf */
 			condt = MIPS_COND_GT | MIPS_COND_K;
 			goto jmp_cmp;
-		case BPF_S_JMP_JGT_X:
+		case BPF_JMP | BPF_JGT | BPF_X:
 			ctx->flags |= SEEN_X;
 			/* pc += ( A > X ) ? pc->jt : pc->jf */
 			condt = MIPS_COND_GT | MIPS_COND_X;
@@ -1109,7 +1132,7 @@ jmp_cmp:
 				}
 				/* A < (K|X) ? r_scrach = 1 */
 				b_off = b_imm(i + inst->jf + 1, ctx);
-				emit_bcond(MIPS_COND_GT, r_s0, r_zero, b_off,
+				emit_bcond(MIPS_COND_NE, r_s0, r_zero, b_off,
 					   ctx);
 				emit_nop(ctx);
 				/* A > (K|X) ? scratch = 0 */
@@ -1167,7 +1190,7 @@ jmp_cmp:
 				}
 			}
 			break;
-		case BPF_S_JMP_JSET_K:
+		case BPF_JMP | BPF_JSET | BPF_K:
 			ctx->flags |= SEEN_S0 | SEEN_S1 | SEEN_A;
 			/* pc += (A & K) ? pc -> jt : pc -> jf */
 			emit_load_imm(r_s1, k, ctx);
@@ -1181,7 +1204,7 @@ jmp_cmp:
 			emit_b(b_off, ctx);
 			emit_nop(ctx);
 			break;
-		case BPF_S_JMP_JSET_X:
+		case BPF_JMP | BPF_JSET | BPF_X:
 			ctx->flags |= SEEN_S0 | SEEN_X | SEEN_A;
 			/* pc += (A & X) ? pc -> jt : pc -> jf */
 			emit_and(r_s0, r_A, r_X, ctx);
@@ -1194,7 +1217,7 @@ jmp_cmp:
 			emit_b(b_off, ctx);
 			emit_nop(ctx);
 			break;
-		case BPF_S_RET_A:
+		case BPF_RET | BPF_A:
 			ctx->flags |= SEEN_A;
 			if (i != prog->len - 1)
 				/*
@@ -1204,7 +1227,7 @@ jmp_cmp:
 				emit_b(b_imm(prog->len, ctx), ctx);
 			emit_reg_move(r_ret, r_A, ctx); /* delay slot */
 			break;
-		case BPF_S_RET_K:
+		case BPF_RET | BPF_K:
 			/*
 			 * It can emit two instructions so it does not fit on
 			 * the delay slot.
@@ -1219,19 +1242,18 @@ jmp_cmp:
 				emit_nop(ctx);
 			}
 			break;
-		case BPF_S_MISC_TAX:
+		case BPF_MISC | BPF_TAX:
 			/* X = A */
 			ctx->flags |= SEEN_X | SEEN_A;
 			emit_jit_reg_move(r_X, r_A, ctx);
 			break;
-		case BPF_S_MISC_TXA:
+		case BPF_MISC | BPF_TXA:
 			/* A = X */
 			ctx->flags |= SEEN_A | SEEN_X;
-			update_on_xread(ctx);
 			emit_jit_reg_move(r_A, r_X, ctx);
 			break;
 		/* AUX */
-		case BPF_S_ANC_PROTOCOL:
+		case BPF_ANC | SKF_AD_PROTOCOL:
 			/* A = ntohs(skb->protocol */
 			ctx->flags |= SEEN_SKB | SEEN_OFF | SEEN_A;
 			BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff,
@@ -1256,7 +1278,7 @@ jmp_cmp:
 			}
 #endif
 			break;
-		case BPF_S_ANC_CPU:
+		case BPF_ANC | SKF_AD_CPU:
 			ctx->flags |= SEEN_A | SEEN_OFF;
 			/* A = current_thread_info()->cpu */
 			BUILD_BUG_ON(FIELD_SIZEOF(struct thread_info,
@@ -1265,11 +1287,12 @@ jmp_cmp:
 			/* $28/gp points to the thread_info struct */
 			emit_load(r_A, 28, off, ctx);
 			break;
-		case BPF_S_ANC_IFINDEX:
+		case BPF_ANC | SKF_AD_IFINDEX:
 			/* A = skb->dev->ifindex */
 			ctx->flags |= SEEN_SKB | SEEN_A | SEEN_S0;
 			off = offsetof(struct sk_buff, dev);
-			emit_load(r_s0, r_skb, off, ctx);
+			/* Load *dev pointer */
+			emit_load_ptr(r_s0, r_skb, off, ctx);
 			/* error (0) in the delay slot */
 			emit_bcond(MIPS_COND_EQ, r_s0, r_zero,
 				   b_imm(prog->len, ctx), ctx);
@@ -1279,31 +1302,36 @@ jmp_cmp:
 			off = offsetof(struct net_device, ifindex);
 			emit_load(r_A, r_s0, off, ctx);
 			break;
-		case BPF_S_ANC_MARK:
+		case BPF_ANC | SKF_AD_MARK:
 			ctx->flags |= SEEN_SKB | SEEN_A;
 			BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, mark) != 4);
 			off = offsetof(struct sk_buff, mark);
 			emit_load(r_A, r_skb, off, ctx);
 			break;
-		case BPF_S_ANC_RXHASH:
+		case BPF_ANC | SKF_AD_RXHASH:
 			ctx->flags |= SEEN_SKB | SEEN_A;
 			BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, hash) != 4);
 			off = offsetof(struct sk_buff, hash);
 			emit_load(r_A, r_skb, off, ctx);
 			break;
-		case BPF_S_ANC_VLAN_TAG:
-		case BPF_S_ANC_VLAN_TAG_PRESENT:
+		case BPF_ANC | SKF_AD_VLAN_TAG:
+		case BPF_ANC | SKF_AD_VLAN_TAG_PRESENT:
 			ctx->flags |= SEEN_SKB | SEEN_S0 | SEEN_A;
 			BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff,
 						  vlan_tci) != 2);
 			off = offsetof(struct sk_buff, vlan_tci);
 			emit_half_load(r_s0, r_skb, off, ctx);
-			if (inst->code == BPF_S_ANC_VLAN_TAG)
-				emit_and(r_A, r_s0, VLAN_VID_MASK, ctx);
-			else
-				emit_and(r_A, r_s0, VLAN_TAG_PRESENT, ctx);
+			if (code == (BPF_ANC | SKF_AD_VLAN_TAG)) {
+				emit_andi(r_A, r_s0, (u16)~VLAN_TAG_PRESENT, ctx);
+			} else {
+				emit_andi(r_A, r_s0, VLAN_TAG_PRESENT, ctx);
+				/* return 1 if present */
+				emit_sltu(r_A, r_zero, r_A, ctx);
+			}
 			break;
-		case BPF_S_ANC_PKTTYPE:
+		case BPF_ANC | SKF_AD_PKTTYPE:
+			ctx->flags |= SEEN_SKB;
+
 			off = pkt_type_offset();
 
 			if (off < 0)
@@ -1311,8 +1339,12 @@ jmp_cmp:
 			emit_load_byte(r_tmp, r_skb, off, ctx);
 			/* Keep only the last 3 bits */
 			emit_andi(r_A, r_tmp, PKT_TYPE_MAX, ctx);
+#ifdef __BIG_ENDIAN_BITFIELD
+			/* Get the actual packet type to the lower 3 bits */
+			emit_srl(r_A, r_A, 5, ctx);
+#endif
 			break;
-		case BPF_S_ANC_QUEUE:
+		case BPF_ANC | SKF_AD_QUEUE:
 			ctx->flags |= SEEN_SKB | SEEN_A;
 			BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff,
 						  queue_mapping) != 2);
@@ -1322,8 +1354,8 @@ jmp_cmp:
 			emit_half_load(r_A, r_skb, off, ctx);
 			break;
 		default:
-			pr_warn("%s: Unhandled opcode: 0x%02x\n", __FILE__,
-				inst->code);
+			pr_debug("%s: Unhandled opcode: 0x%02x\n", __FILE__,
+				 inst->code);
 			return -1;
 		}
 	}
diff --git a/arch/parisc/include/uapi/asm/signal.h b/arch/parisc/include/uapi/asm/signal.h
index a2fa297196bc..f5645d6a89f2 100644
--- a/arch/parisc/include/uapi/asm/signal.h
+++ b/arch/parisc/include/uapi/asm/signal.h
@@ -69,8 +69,6 @@
 #define SA_NOMASK	SA_NODEFER
 #define SA_ONESHOT	SA_RESETHAND
 
-#define SA_RESTORER	0x04000000 /* obsolete -- ignored */
-
 #define MINSIGSTKSZ	2048
 #define SIGSTKSZ	8192
 
diff --git a/arch/parisc/kernel/hardware.c b/arch/parisc/kernel/hardware.c
index 608716f8496b..af3bc359dc70 100644
--- a/arch/parisc/kernel/hardware.c
+++ b/arch/parisc/kernel/hardware.c
@@ -1210,7 +1210,8 @@ static struct hp_hardware hp_hardware_list[] = {
 	{HPHW_FIO, 0x004, 0x00320, 0x0, "Metheus Frame Buffer"}, 
 	{HPHW_FIO, 0x004, 0x00340, 0x0, "BARCO CX4500 VME Grphx Cnsl"}, 
 	{HPHW_FIO, 0x004, 0x00360, 0x0, "Hughes TOG VME FDDI"}, 
-	{HPHW_FIO, 0x076, 0x000AD, 0x00, "Crestone Peak RS-232"},
+	{HPHW_FIO, 0x076, 0x000AD, 0x0, "Crestone Peak Core RS-232"},
+	{HPHW_FIO, 0x077, 0x000AD, 0x0, "Crestone Peak Fast? Core RS-232"},
 	{HPHW_IOA, 0x185, 0x0000B, 0x00, "Java BC Summit Port"}, 
 	{HPHW_IOA, 0x1FF, 0x0000B, 0x00, "Hitachi Ghostview Summit Port"}, 
 	{HPHW_IOA, 0x580, 0x0000B, 0x10, "U2-IOA BC Runway Port"}, 
diff --git a/arch/parisc/kernel/sys_parisc32.c b/arch/parisc/kernel/sys_parisc32.c
index bb9f3b64de55..93c1963d76fe 100644
--- a/arch/parisc/kernel/sys_parisc32.c
+++ b/arch/parisc/kernel/sys_parisc32.c
@@ -4,6 +4,7 @@
  * Copyright (C) 2000-2001 Hewlett Packard Company
  * Copyright (C) 2000 John Marvin
  * Copyright (C) 2001 Matthew Wilcox
+ * Copyright (C) 2014 Helge Deller <deller@gmx.de>
  *
  * These routines maintain argument size conversion between 32bit and 64bit
  * environment. Based heavily on sys_ia32.c and sys_sparc32.c.
@@ -11,44 +12,8 @@
 
 #include <linux/compat.h>
 #include <linux/kernel.h>
-#include <linux/sched.h>
-#include <linux/fs.h> 
-#include <linux/mm.h> 
-#include <linux/file.h> 
-#include <linux/signal.h>
-#include <linux/resource.h>
-#include <linux/times.h>
-#include <linux/time.h>
-#include <linux/smp.h>
-#include <linux/sem.h>
-#include <linux/shm.h>
-#include <linux/slab.h>
-#include <linux/uio.h>
-#include <linux/ncp_fs.h>
-#include <linux/poll.h>
-#include <linux/personality.h>
-#include <linux/stat.h>
-#include <linux/highmem.h>
-#include <linux/highuid.h>
-#include <linux/mman.h>
-#include <linux/binfmts.h>
-#include <linux/namei.h>
-#include <linux/vfs.h>
-#include <linux/ptrace.h>
-#include <linux/swap.h>
 #include <linux/syscalls.h>
 
-#include <asm/types.h>
-#include <asm/uaccess.h>
-#include <asm/mmu_context.h>
-
-#undef DEBUG
-
-#ifdef DEBUG
-#define DBG(x)	printk x
-#else
-#define DBG(x)
-#endif
 
 asmlinkage long sys32_unimplemented(int r26, int r25, int r24, int r23,
 	int r22, int r21, int r20)
@@ -57,3 +22,12 @@ asmlinkage long sys32_unimplemented(int r26, int r25, int r24, int r23,
     	current->comm, current->pid, r20);
     return -ENOSYS;
 }
+
+asmlinkage long sys32_fanotify_mark(compat_int_t fanotify_fd, compat_uint_t flags,
+	compat_uint_t mask0, compat_uint_t mask1, compat_int_t dfd,
+	const char  __user * pathname)
+{
+	return sys_fanotify_mark(fanotify_fd, flags,
+			((__u64)mask1 << 32) | mask0,
+			 dfd, pathname);
+}
diff --git a/arch/parisc/kernel/syscall_table.S b/arch/parisc/kernel/syscall_table.S
index c5fa7a697fba..84c5d3a58fa1 100644
--- a/arch/parisc/kernel/syscall_table.S
+++ b/arch/parisc/kernel/syscall_table.S
@@ -418,7 +418,7 @@
 	ENTRY_SAME(accept4)		/* 320 */
 	ENTRY_SAME(prlimit64)
 	ENTRY_SAME(fanotify_init)
-	ENTRY_COMP(fanotify_mark)
+	ENTRY_DIFF(fanotify_mark)
 	ENTRY_COMP(clock_adjtime)
 	ENTRY_SAME(name_to_handle_at)	/* 325 */
 	ENTRY_COMP(open_by_handle_at)
diff --git a/arch/parisc/mm/init.c b/arch/parisc/mm/init.c
index ae085ad0fba0..0bef864264c0 100644
--- a/arch/parisc/mm/init.c
+++ b/arch/parisc/mm/init.c
@@ -728,7 +728,6 @@ static void __init pagetable_init(void)
 #endif
 
 	empty_zero_page = alloc_bootmem_pages(PAGE_SIZE);
-	memset(empty_zero_page, 0, PAGE_SIZE);
 }
 
 static void __init gateway_init(void)
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
index bd6dd6ed3a9f..80b94b0add1f 100644
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -145,6 +145,7 @@ config PPC
 	select HAVE_IRQ_EXIT_ON_IRQ_STACK
 	select ARCH_USE_CMPXCHG_LOCKREF if PPC64
 	select HAVE_ARCH_AUDITSYSCALL
+	select ARCH_SUPPORTS_ATOMIC_RMW
 
 config GENERIC_CSUM
 	def_bool CPU_LITTLE_ENDIAN
@@ -414,7 +415,7 @@ config KEXEC
 config CRASH_DUMP
 	bool "Build a kdump crash kernel"
 	depends on PPC64 || 6xx || FSL_BOOKE || (44x && !SMP)
-	select RELOCATABLE if PPC64 || 44x || FSL_BOOKE
+	select RELOCATABLE if (PPC64 && !COMPILE_TEST) || 44x || FSL_BOOKE
 	help
 	  Build a kernel suitable for use as a kdump capture kernel.
 	  The same kernel binary can be used as production kernel and dump
@@ -1017,6 +1018,7 @@ endmenu
 if PPC64
 config RELOCATABLE
 	bool "Build a relocatable kernel"
+	depends on !COMPILE_TEST
 	select NONSTATIC_KERNEL
 	help
 	  This builds a kernel image that is capable of running anywhere
diff --git a/arch/powerpc/Kconfig.debug b/arch/powerpc/Kconfig.debug
index 790352f93700..35d16bd2760b 100644
--- a/arch/powerpc/Kconfig.debug
+++ b/arch/powerpc/Kconfig.debug
@@ -303,7 +303,6 @@ config PPC_EARLY_DEBUG_OPAL_VTERMNO
 	  This correspond to which /dev/hvcN you want to use for early
 	  debug.
 
-	  On OPAL v1 (takeover) this should always be 0
 	  On OPAL v2, this will be 0 for network console and 1 or 2 for
 	  the machine built-in serial ports.
 
diff --git a/arch/powerpc/include/asm/code-patching.h b/arch/powerpc/include/asm/code-patching.h
index 37991e154ef8..840a5509b3f1 100644
--- a/arch/powerpc/include/asm/code-patching.h
+++ b/arch/powerpc/include/asm/code-patching.h
@@ -88,4 +88,15 @@ static inline unsigned long ppc_function_entry(void *func)
 #endif
 }
 
+static inline unsigned long ppc_global_function_entry(void *func)
+{
+#if defined(CONFIG_PPC64) && defined(_CALL_ELF) && _CALL_ELF == 2
+	/* PPC64 ABIv2 the global entry point is at the address */
+	return (unsigned long)func;
+#else
+	/* All other cases there is no change vs ppc_function_entry() */
+	return ppc_function_entry(func);
+#endif
+}
+
 #endif /* _ASM_POWERPC_CODE_PATCHING_H */
diff --git a/arch/powerpc/include/asm/cputable.h b/arch/powerpc/include/asm/cputable.h
index bc2347774f0a..0fdd7eece6d9 100644
--- a/arch/powerpc/include/asm/cputable.h
+++ b/arch/powerpc/include/asm/cputable.h
@@ -447,6 +447,7 @@ extern const char *powerpc_base_platform;
 	    CPU_FTR_DBELL | CPU_FTR_HAS_PPR | CPU_FTR_DAWR | \
 	    CPU_FTR_ARCH_207S | CPU_FTR_TM_COMP)
 #define CPU_FTRS_POWER8E (CPU_FTRS_POWER8 | CPU_FTR_PMAO_BUG)
+#define CPU_FTRS_POWER8_DD1 (CPU_FTRS_POWER8 & ~CPU_FTR_DBELL)
 #define CPU_FTRS_CELL	(CPU_FTR_USE_TB | CPU_FTR_LWSYNC | \
 	    CPU_FTR_PPCAS_ARCH_V2 | CPU_FTR_CTRL | \
 	    CPU_FTR_ALTIVEC_COMP | CPU_FTR_MMCRA | CPU_FTR_SMT | \
diff --git a/arch/powerpc/include/asm/kvm_book3s_64.h b/arch/powerpc/include/asm/kvm_book3s_64.h
index fddb72b48ce9..d645428a65a4 100644
--- a/arch/powerpc/include/asm/kvm_book3s_64.h
+++ b/arch/powerpc/include/asm/kvm_book3s_64.h
@@ -198,8 +198,10 @@ static inline unsigned long compute_tlbie_rb(unsigned long v, unsigned long r,
 	return rb;
 }
 
-static inline unsigned long hpte_page_size(unsigned long h, unsigned long l)
+static inline unsigned long __hpte_page_size(unsigned long h, unsigned long l,
+					     bool is_base_size)
 {
+
 	int size, a_psize;
 	/* Look at the 8 bit LP value */
 	unsigned int lp = (l >> LP_SHIFT) & ((1 << LP_BITS) - 1);
@@ -214,14 +216,27 @@ static inline unsigned long hpte_page_size(unsigned long h, unsigned long l)
 				continue;
 
 			a_psize = __hpte_actual_psize(lp, size);
-			if (a_psize != -1)
+			if (a_psize != -1) {
+				if (is_base_size)
+					return 1ul << mmu_psize_defs[size].shift;
 				return 1ul << mmu_psize_defs[a_psize].shift;
+			}
 		}
 
 	}
 	return 0;
 }
 
+static inline unsigned long hpte_page_size(unsigned long h, unsigned long l)
+{
+	return __hpte_page_size(h, l, 0);
+}
+
+static inline unsigned long hpte_base_page_size(unsigned long h, unsigned long l)
+{
+	return __hpte_page_size(h, l, 1);
+}
+
 static inline unsigned long hpte_rpn(unsigned long ptel, unsigned long psize)
 {
 	return ((ptel & HPTE_R_RPN) & ~(psize - 1)) >> PAGE_SHIFT;
diff --git a/arch/powerpc/include/asm/mmu-hash64.h b/arch/powerpc/include/asm/mmu-hash64.h
index 807014dde821..c2b4dcf23d03 100644
--- a/arch/powerpc/include/asm/mmu-hash64.h
+++ b/arch/powerpc/include/asm/mmu-hash64.h
@@ -22,6 +22,7 @@
  */
 #include <asm/pgtable-ppc64.h>
 #include <asm/bug.h>
+#include <asm/processor.h>
 
 /*
  * Segment table
@@ -496,7 +497,7 @@ extern void slb_set_size(u16 size);
  */
 struct subpage_prot_table {
 	unsigned long maxaddr;	/* only addresses < this are protected */
-	unsigned int **protptrs[2];
+	unsigned int **protptrs[(TASK_SIZE_USER64 >> 43)];
 	unsigned int *low_prot[4];
 };
 
diff --git a/arch/powerpc/include/asm/mmu.h b/arch/powerpc/include/asm/mmu.h
index f8d1d6dcf7db..e61f24ed4e65 100644
--- a/arch/powerpc/include/asm/mmu.h
+++ b/arch/powerpc/include/asm/mmu.h
@@ -19,8 +19,7 @@
 #define MMU_FTR_TYPE_40x		ASM_CONST(0x00000004)
 #define MMU_FTR_TYPE_44x		ASM_CONST(0x00000008)
 #define MMU_FTR_TYPE_FSL_E		ASM_CONST(0x00000010)
-#define MMU_FTR_TYPE_3E			ASM_CONST(0x00000020)
-#define MMU_FTR_TYPE_47x		ASM_CONST(0x00000040)
+#define MMU_FTR_TYPE_47x		ASM_CONST(0x00000020)
 
 /*
  * This is individual features
@@ -106,13 +105,6 @@
 				MMU_FTR_CI_LARGE_PAGE
 #define MMU_FTRS_PA6T		MMU_FTRS_DEFAULT_HPTE_ARCH_V2 | \
 				MMU_FTR_CI_LARGE_PAGE | MMU_FTR_NO_SLBIE_B
-#define MMU_FTRS_A2		MMU_FTR_TYPE_3E | MMU_FTR_USE_TLBILX | \
-				MMU_FTR_USE_TLBIVAX_BCAST | \
-				MMU_FTR_LOCK_BCAST_INVAL | \
-				MMU_FTR_USE_TLBRSRV | \
-				MMU_FTR_USE_PAIRED_MAS | \
-				MMU_FTR_TLBIEL | \
-				MMU_FTR_16M_PAGE
 #ifndef __ASSEMBLY__
 #include <asm/cputable.h>
 
diff --git a/arch/powerpc/include/asm/opal.h b/arch/powerpc/include/asm/opal.h
index 460018889ba9..0da1dbd42e02 100644
--- a/arch/powerpc/include/asm/opal.h
+++ b/arch/powerpc/include/asm/opal.h
@@ -12,27 +12,7 @@
 #ifndef __OPAL_H
 #define __OPAL_H
 
-/****** Takeover interface ********/
-
-/* PAPR H-Call used to querty the HAL existence and/or instanciate
- * it from within pHyp (tech preview only).
- *
- * This is exclusively used in prom_init.c
- */
-
 #ifndef __ASSEMBLY__
-
-struct opal_takeover_args {
-	u64	k_image;		/* r4 */
-	u64	k_size;			/* r5 */
-	u64	k_entry;		/* r6 */
-	u64	k_entry2;		/* r7 */
-	u64	hal_addr;		/* r8 */
-	u64	rd_image;		/* r9 */
-	u64	rd_size;		/* r10 */
-	u64	rd_loc;			/* r11 */
-};
-
 /*
  * SG entry
  *
@@ -55,15 +35,6 @@ struct opal_sg_list {
 /* We calculate number of sg entries based on PAGE_SIZE */
 #define SG_ENTRIES_PER_NODE ((PAGE_SIZE - 16) / sizeof(struct opal_sg_entry))
 
-extern long opal_query_takeover(u64 *hal_size, u64 *hal_align);
-
-extern long opal_do_takeover(struct opal_takeover_args *args);
-
-struct rtas_args;
-extern int opal_enter_rtas(struct rtas_args *args,
-			   unsigned long data,
-			   unsigned long entry);
-
 #endif /* __ASSEMBLY__ */
 
 /****** OPAL APIs ******/
diff --git a/arch/powerpc/include/asm/perf_event_server.h b/arch/powerpc/include/asm/perf_event_server.h
index 9ed737146dbb..b3e936027b26 100644
--- a/arch/powerpc/include/asm/perf_event_server.h
+++ b/arch/powerpc/include/asm/perf_event_server.h
@@ -61,8 +61,7 @@ struct power_pmu {
 #define PPMU_SIAR_VALID		0x00000010 /* Processor has SIAR Valid bit */
 #define PPMU_HAS_SSLOT		0x00000020 /* Has sampled slot in MMCRA */
 #define PPMU_HAS_SIER		0x00000040 /* Has SIER */
-#define PPMU_BHRB		0x00000080 /* has BHRB feature enabled */
-#define PPMU_EBB		0x00000100 /* supports event based branch */
+#define PPMU_ARCH_207S		0x00000080 /* PMC is architecture v2.07S */
 
 /*
  * Values for flags to get_alternatives()
diff --git a/arch/powerpc/include/asm/ppc_asm.h b/arch/powerpc/include/asm/ppc_asm.h
index 9ea266eae33e..7e4612528546 100644
--- a/arch/powerpc/include/asm/ppc_asm.h
+++ b/arch/powerpc/include/asm/ppc_asm.h
@@ -277,6 +277,8 @@ n:
 	.globl n;	\
 n:
 
+#define _GLOBAL_TOC(name) _GLOBAL(name)
+
 #define _KPROBE(n)	\
 	.section ".kprobes.text","a";	\
 	.globl	n;	\
diff --git a/arch/powerpc/include/asm/swab.h b/arch/powerpc/include/asm/swab.h
index b9bd1ca944d0..96f59de61855 100644
--- a/arch/powerpc/include/asm/swab.h
+++ b/arch/powerpc/include/asm/swab.h
@@ -9,10 +9,6 @@
 
 #include <uapi/asm/swab.h>
 
-#ifdef __GNUC__
-#ifndef __powerpc64__
-#endif /* __powerpc64__ */
-
 static __inline__ __u16 ld_le16(const volatile __u16 *addr)
 {
 	__u16 val;
@@ -20,19 +16,12 @@ static __inline__ __u16 ld_le16(const volatile __u16 *addr)
 	__asm__ __volatile__ ("lhbrx %0,0,%1" : "=r" (val) : "r" (addr), "m" (*addr));
 	return val;
 }
-#define __arch_swab16p ld_le16
 
 static __inline__ void st_le16(volatile __u16 *addr, const __u16 val)
 {
 	__asm__ __volatile__ ("sthbrx %1,0,%2" : "=m" (*addr) : "r" (val), "r" (addr));
 }
 
-static inline void __arch_swab16s(__u16 *addr)
-{
-	st_le16(addr, *addr);
-}
-#define __arch_swab16s __arch_swab16s
-
 static __inline__ __u32 ld_le32(const volatile __u32 *addr)
 {
 	__u32 val;
@@ -40,42 +29,10 @@ static __inline__ __u32 ld_le32(const volatile __u32 *addr)
 	__asm__ __volatile__ ("lwbrx %0,0,%1" : "=r" (val) : "r" (addr), "m" (*addr));
 	return val;
 }
-#define __arch_swab32p ld_le32
 
 static __inline__ void st_le32(volatile __u32 *addr, const __u32 val)
 {
 	__asm__ __volatile__ ("stwbrx %1,0,%2" : "=m" (*addr) : "r" (val), "r" (addr));
 }
 
-static inline void __arch_swab32s(__u32 *addr)
-{
-	st_le32(addr, *addr);
-}
-#define __arch_swab32s __arch_swab32s
-
-static inline __attribute_const__ __u16 __arch_swab16(__u16 value)
-{
-	__u16 result;
-
-	__asm__("rlwimi %0,%1,8,16,23"
-	    : "=r" (result)
-	    : "r" (value), "0" (value >> 8));
-	return result;
-}
-#define __arch_swab16 __arch_swab16
-
-static inline __attribute_const__ __u32 __arch_swab32(__u32 value)
-{
-	__u32 result;
-
-	__asm__("rlwimi %0,%1,24,16,23\n\t"
-	    "rlwimi %0,%1,8,8,15\n\t"
-	    "rlwimi %0,%1,24,0,7"
-	    : "=r" (result)
-	    : "r" (value), "0" (value >> 24));
-	return result;
-}
-#define __arch_swab32 __arch_swab32
-
-#endif /* __GNUC__ */
 #endif /* _ASM_POWERPC_SWAB_H */
diff --git a/arch/powerpc/kernel/cputable.c b/arch/powerpc/kernel/cputable.c
index 965291b4c2fa..0c157642c2a1 100644
--- a/arch/powerpc/kernel/cputable.c
+++ b/arch/powerpc/kernel/cputable.c
@@ -527,6 +527,26 @@ static struct cpu_spec __initdata cpu_specs[] = {
 		.machine_check_early	= __machine_check_early_realmode_p8,
 		.platform		= "power8",
 	},
+	{	/* Power8 DD1: Does not support doorbell IPIs */
+		.pvr_mask		= 0xffffff00,
+		.pvr_value		= 0x004d0100,
+		.cpu_name		= "POWER8 (raw)",
+		.cpu_features		= CPU_FTRS_POWER8_DD1,
+		.cpu_user_features	= COMMON_USER_POWER8,
+		.cpu_user_features2	= COMMON_USER2_POWER8,
+		.mmu_features		= MMU_FTRS_POWER8,
+		.icache_bsize		= 128,
+		.dcache_bsize		= 128,
+		.num_pmcs		= 6,
+		.pmc_type		= PPC_PMC_IBM,
+		.oprofile_cpu_type	= "ppc64/power8",
+		.oprofile_type		= PPC_OPROFILE_INVALID,
+		.cpu_setup		= __setup_cpu_power8,
+		.cpu_restore		= __restore_cpu_power8,
+		.flush_tlb		= __flush_tlb_power8,
+		.machine_check_early	= __machine_check_early_realmode_p8,
+		.platform		= "power8",
+	},
 	{	/* Power8 */
 		.pvr_mask		= 0xffff0000,
 		.pvr_value		= 0x004d0000,
diff --git a/arch/powerpc/kernel/ftrace.c b/arch/powerpc/kernel/ftrace.c
index f202d0731b06..d178834fe508 100644
--- a/arch/powerpc/kernel/ftrace.c
+++ b/arch/powerpc/kernel/ftrace.c
@@ -10,6 +10,8 @@
  *
  */
 
+#define pr_fmt(fmt) "ftrace-powerpc: " fmt
+
 #include <linux/spinlock.h>
 #include <linux/hardirq.h>
 #include <linux/uaccess.h>
@@ -105,7 +107,7 @@ __ftrace_make_nop(struct module *mod,
 		  struct dyn_ftrace *rec, unsigned long addr)
 {
 	unsigned int op;
-	unsigned long ptr;
+	unsigned long entry, ptr;
 	unsigned long ip = rec->ip;
 	void *tramp;
 
@@ -115,7 +117,7 @@ __ftrace_make_nop(struct module *mod,
 
 	/* Make sure that that this is still a 24bit jump */
 	if (!is_bl_op(op)) {
-		printk(KERN_ERR "Not expected bl: opcode is %x\n", op);
+		pr_err("Not expected bl: opcode is %x\n", op);
 		return -EINVAL;
 	}
 
@@ -125,21 +127,21 @@ __ftrace_make_nop(struct module *mod,
 	pr_devel("ip:%lx jumps to %p", ip, tramp);
 
 	if (!is_module_trampoline(tramp)) {
-		printk(KERN_ERR "Not a trampoline\n");
+		pr_err("Not a trampoline\n");
 		return -EINVAL;
 	}
 
 	if (module_trampoline_target(mod, tramp, &ptr)) {
-		printk(KERN_ERR "Failed to get trampoline target\n");
+		pr_err("Failed to get trampoline target\n");
 		return -EFAULT;
 	}
 
 	pr_devel("trampoline target %lx", ptr);
 
+	entry = ppc_global_function_entry((void *)addr);
 	/* This should match what was called */
-	if (ptr != ppc_function_entry((void *)addr)) {
-		printk(KERN_ERR "addr %lx does not match expected %lx\n",
-			ptr, ppc_function_entry((void *)addr));
+	if (ptr != entry) {
+		pr_err("addr %lx does not match expected %lx\n", ptr, entry);
 		return -EINVAL;
 	}
 
@@ -179,7 +181,7 @@ __ftrace_make_nop(struct module *mod,
 
 	/* Make sure that that this is still a 24bit jump */
 	if (!is_bl_op(op)) {
-		printk(KERN_ERR "Not expected bl: opcode is %x\n", op);
+		pr_err("Not expected bl: opcode is %x\n", op);
 		return -EINVAL;
 	}
 
@@ -198,7 +200,7 @@ __ftrace_make_nop(struct module *mod,
 
 	/* Find where the trampoline jumps to */
 	if (probe_kernel_read(jmp, (void *)tramp, sizeof(jmp))) {
-		printk(KERN_ERR "Failed to read %lx\n", tramp);
+		pr_err("Failed to read %lx\n", tramp);
 		return -EFAULT;
 	}
 
@@ -209,7 +211,7 @@ __ftrace_make_nop(struct module *mod,
 	    ((jmp[1] & 0xffff0000) != 0x398c0000) ||
 	    (jmp[2] != 0x7d8903a6) ||
 	    (jmp[3] != 0x4e800420)) {
-		printk(KERN_ERR "Not a trampoline\n");
+		pr_err("Not a trampoline\n");
 		return -EINVAL;
 	}
 
@@ -221,8 +223,7 @@ __ftrace_make_nop(struct module *mod,
 	pr_devel(" %lx ", tramp);
 
 	if (tramp != addr) {
-		printk(KERN_ERR
-		       "Trampoline location %08lx does not match addr\n",
+		pr_err("Trampoline location %08lx does not match addr\n",
 		       tramp);
 		return -EINVAL;
 	}
@@ -263,15 +264,13 @@ int ftrace_make_nop(struct module *mod,
 	 */
 	if (!rec->arch.mod) {
 		if (!mod) {
-			printk(KERN_ERR "No module loaded addr=%lx\n",
-			       addr);
+			pr_err("No module loaded addr=%lx\n", addr);
 			return -EFAULT;
 		}
 		rec->arch.mod = mod;
 	} else if (mod) {
 		if (mod != rec->arch.mod) {
-			printk(KERN_ERR
-			       "Record mod %p not equal to passed in mod %p\n",
+			pr_err("Record mod %p not equal to passed in mod %p\n",
 			       rec->arch.mod, mod);
 			return -EINVAL;
 		}
@@ -307,26 +306,25 @@ __ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 	 * The load offset is different depending on the ABI. For simplicity
 	 * just mask it out when doing the compare.
 	 */
-	if ((op[0] != 0x48000008) || ((op[1] & 0xffff00000) != 0xe8410000)) {
-		printk(KERN_ERR "Unexpected call sequence: %x %x\n",
-			op[0], op[1]);
+	if ((op[0] != 0x48000008) || ((op[1] & 0xffff0000) != 0xe8410000)) {
+		pr_err("Unexpected call sequence: %x %x\n", op[0], op[1]);
 		return -EINVAL;
 	}
 
 	/* If we never set up a trampoline to ftrace_caller, then bail */
 	if (!rec->arch.mod->arch.tramp) {
-		printk(KERN_ERR "No ftrace trampoline\n");
+		pr_err("No ftrace trampoline\n");
 		return -EINVAL;
 	}
 
 	/* Ensure branch is within 24 bits */
-	if (create_branch(ip, rec->arch.mod->arch.tramp, BRANCH_SET_LINK)) {
-		printk(KERN_ERR "Branch out of range");
+	if (!create_branch(ip, rec->arch.mod->arch.tramp, BRANCH_SET_LINK)) {
+		pr_err("Branch out of range\n");
 		return -EINVAL;
 	}
 
 	if (patch_branch(ip, rec->arch.mod->arch.tramp, BRANCH_SET_LINK)) {
-		printk(KERN_ERR "REL24 out of range!\n");
+		pr_err("REL24 out of range!\n");
 		return -EINVAL;
 	}
 
@@ -345,13 +343,13 @@ __ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 
 	/* It should be pointing to a nop */
 	if (op != PPC_INST_NOP) {
-		printk(KERN_ERR "Expected NOP but have %x\n", op);
+		pr_err("Expected NOP but have %x\n", op);
 		return -EINVAL;
 	}
 
 	/* If we never set up a trampoline to ftrace_caller, then bail */
 	if (!rec->arch.mod->arch.tramp) {
-		printk(KERN_ERR "No ftrace trampoline\n");
+		pr_err("No ftrace trampoline\n");
 		return -EINVAL;
 	}
 
@@ -359,7 +357,7 @@ __ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 	op = create_branch((unsigned int *)ip,
 			   rec->arch.mod->arch.tramp, BRANCH_SET_LINK);
 	if (!op) {
-		printk(KERN_ERR "REL24 out of range!\n");
+		pr_err("REL24 out of range!\n");
 		return -EINVAL;
 	}
 
@@ -397,7 +395,7 @@ int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
 	 * already have a module defined.
 	 */
 	if (!rec->arch.mod) {
-		printk(KERN_ERR "No module loaded\n");
+		pr_err("No module loaded\n");
 		return -EINVAL;
 	}
 
diff --git a/arch/powerpc/kernel/idle_power7.S b/arch/powerpc/kernel/idle_power7.S
index 2480256272d4..5cf3d367190d 100644
--- a/arch/powerpc/kernel/idle_power7.S
+++ b/arch/powerpc/kernel/idle_power7.S
@@ -131,7 +131,7 @@ _GLOBAL(power7_nap)
 
 _GLOBAL(power7_sleep)
 	li	r3,1
-	li	r4,0
+	li	r4,1
 	b	power7_powersave_common
 	/* No return */
 
diff --git a/arch/powerpc/kernel/iomap.c b/arch/powerpc/kernel/iomap.c
index b82227e7e21b..12e48d56f771 100644
--- a/arch/powerpc/kernel/iomap.c
+++ b/arch/powerpc/kernel/iomap.c
@@ -23,7 +23,7 @@ unsigned int ioread16(void __iomem *addr)
 }
 unsigned int ioread16be(void __iomem *addr)
 {
-	return in_be16(addr);
+	return readw_be(addr);
 }
 unsigned int ioread32(void __iomem *addr)
 {
@@ -31,7 +31,7 @@ unsigned int ioread32(void __iomem *addr)
 }
 unsigned int ioread32be(void __iomem *addr)
 {
-	return in_be32(addr);
+	return readl_be(addr);
 }
 EXPORT_SYMBOL(ioread8);
 EXPORT_SYMBOL(ioread16);
@@ -49,7 +49,7 @@ void iowrite16(u16 val, void __iomem *addr)
 }
 void iowrite16be(u16 val, void __iomem *addr)
 {
-	out_be16(addr, val);
+	writew_be(val, addr);
 }
 void iowrite32(u32 val, void __iomem *addr)
 {
@@ -57,7 +57,7 @@ void iowrite32(u32 val, void __iomem *addr)
 }
 void iowrite32be(u32 val, void __iomem *addr)
 {
-	out_be32(addr, val);
+	writel_be(val, addr);
 }
 EXPORT_SYMBOL(iowrite8);
 EXPORT_SYMBOL(iowrite16);
@@ -75,15 +75,15 @@ EXPORT_SYMBOL(iowrite32be);
  */
 void ioread8_rep(void __iomem *addr, void *dst, unsigned long count)
 {
-	_insb((u8 __iomem *) addr, dst, count);
+	readsb(addr, dst, count);
 }
 void ioread16_rep(void __iomem *addr, void *dst, unsigned long count)
 {
-	_insw_ns((u16 __iomem *) addr, dst, count);
+	readsw(addr, dst, count);
 }
 void ioread32_rep(void __iomem *addr, void *dst, unsigned long count)
 {
-	_insl_ns((u32 __iomem *) addr, dst, count);
+	readsl(addr, dst, count);
 }
 EXPORT_SYMBOL(ioread8_rep);
 EXPORT_SYMBOL(ioread16_rep);
@@ -91,15 +91,15 @@ EXPORT_SYMBOL(ioread32_rep);
 
 void iowrite8_rep(void __iomem *addr, const void *src, unsigned long count)
 {
-	_outsb((u8 __iomem *) addr, src, count);
+	writesb(addr, src, count);
 }
 void iowrite16_rep(void __iomem *addr, const void *src, unsigned long count)
 {
-	_outsw_ns((u16 __iomem *) addr, src, count);
+	writesw(addr, src, count);
 }
 void iowrite32_rep(void __iomem *addr, const void *src, unsigned long count)
 {
-	_outsl_ns((u32 __iomem *) addr, src, count);
+	writesl(addr, src, count);
 }
 EXPORT_SYMBOL(iowrite8_rep);
 EXPORT_SYMBOL(iowrite16_rep);
diff --git a/arch/powerpc/kernel/kprobes.c b/arch/powerpc/kernel/kprobes.c
index 90fab64d911d..2f72af82513c 100644
--- a/arch/powerpc/kernel/kprobes.c
+++ b/arch/powerpc/kernel/kprobes.c
@@ -32,6 +32,7 @@
 #include <linux/module.h>
 #include <linux/kdebug.h>
 #include <linux/slab.h>
+#include <asm/code-patching.h>
 #include <asm/cacheflush.h>
 #include <asm/sstep.h>
 #include <asm/uaccess.h>
@@ -491,12 +492,10 @@ int __kprobes kprobe_exceptions_notify(struct notifier_block *self,
 	return ret;
 }
 
-#ifdef CONFIG_PPC64
 unsigned long arch_deref_entry_point(void *entry)
 {
-	return ((func_descr_t *)entry)->entry;
+	return ppc_global_function_entry(entry);
 }
-#endif
 
 int __kprobes setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs)
 {
@@ -508,8 +507,12 @@ int __kprobes setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs)
 	/* setup return addr to the jprobe handler routine */
 	regs->nip = arch_deref_entry_point(jp->entry);
 #ifdef CONFIG_PPC64
+#if defined(_CALL_ELF) && _CALL_ELF == 2
+	regs->gpr[12] = (unsigned long)jp->entry;
+#else
 	regs->gpr[2] = (unsigned long)(((func_descr_t *)jp->entry)->toc);
 #endif
+#endif
 
 	return 1;
 }
diff --git a/arch/powerpc/kernel/module_64.c b/arch/powerpc/kernel/module_64.c
index 077d2ce6c5a7..d807ee626af9 100644
--- a/arch/powerpc/kernel/module_64.c
+++ b/arch/powerpc/kernel/module_64.c
@@ -315,8 +315,17 @@ static void dedotify_versions(struct modversion_info *vers,
 	struct modversion_info *end;
 
 	for (end = (void *)vers + size; vers < end; vers++)
-		if (vers->name[0] == '.')
+		if (vers->name[0] == '.') {
 			memmove(vers->name, vers->name+1, strlen(vers->name));
+#ifdef ARCH_RELOCATES_KCRCTAB
+			/* The TOC symbol has no CRC computed. To avoid CRC
+			 * check failing, we must force it to the expected
+			 * value (see CRC check in module.c).
+			 */
+			if (!strcmp(vers->name, "TOC."))
+				vers->crc = -(unsigned long)reloc_start;
+#endif
+		}
 }
 
 /* Undefined symbols which refer to .funcname, hack to funcname (or .TOC.) */
diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c
index 613a860a203c..b694b0730971 100644
--- a/arch/powerpc/kernel/prom.c
+++ b/arch/powerpc/kernel/prom.c
@@ -662,13 +662,6 @@ void __init early_init_devtree(void *params)
 	of_scan_flat_dt(early_init_dt_scan_fw_dump, NULL);
 #endif
 
-	/* Pre-initialize the cmd_line with the content of boot_commmand_line,
-	 * which will be empty except when the content of the variable has
-	 * been overriden by a bootloading mechanism. This happens typically
-	 * with HAL takeover
-	 */
-	strlcpy(cmd_line, boot_command_line, COMMAND_LINE_SIZE);
-
 	/* Retrieve various informations from the /chosen node of the
 	 * device-tree, including the platform type, initrd location and
 	 * size, TCE reserve, and more ...
diff --git a/arch/powerpc/kernel/prom_init.c b/arch/powerpc/kernel/prom_init.c
index 078145acf7fb..1a85d8f96739 100644
--- a/arch/powerpc/kernel/prom_init.c
+++ b/arch/powerpc/kernel/prom_init.c
@@ -1268,201 +1268,6 @@ static u64 __initdata prom_opal_base;
 static u64 __initdata prom_opal_entry;
 #endif
 
-#ifdef __BIG_ENDIAN__
-/* XXX Don't change this structure without updating opal-takeover.S */
-static struct opal_secondary_data {
-	s64				ack;	/*  0 */
-	u64				go;	/*  8 */
-	struct opal_takeover_args	args;	/* 16 */
-} opal_secondary_data;
-
-static u64 __initdata prom_opal_align;
-static u64 __initdata prom_opal_size;
-static int __initdata prom_rtas_start_cpu;
-static u64 __initdata prom_rtas_data;
-static u64 __initdata prom_rtas_entry;
-
-extern char opal_secondary_entry;
-
-static void __init prom_query_opal(void)
-{
-	long rc;
-
-	/* We must not query for OPAL presence on a machine that
-	 * supports TNK takeover (970 blades), as this uses the same
-	 * h-call with different arguments and will crash
-	 */
-	if (PHANDLE_VALID(call_prom("finddevice", 1, 1,
-				    ADDR("/tnk-memory-map")))) {
-		prom_printf("TNK takeover detected, skipping OPAL check\n");
-		return;
-	}
-
-	prom_printf("Querying for OPAL presence... ");
-
-	rc = opal_query_takeover(&prom_opal_size,
-				 &prom_opal_align);
-	prom_debug("(rc = %ld) ", rc);
-	if (rc != 0) {
-		prom_printf("not there.\n");
-		return;
-	}
-	of_platform = PLATFORM_OPAL;
-	prom_printf(" there !\n");
-	prom_debug("  opal_size  = 0x%lx\n", prom_opal_size);
-	prom_debug("  opal_align = 0x%lx\n", prom_opal_align);
-	if (prom_opal_align < 0x10000)
-		prom_opal_align = 0x10000;
-}
-
-static int __init prom_rtas_call(int token, int nargs, int nret,
-				 int *outputs, ...)
-{
-	struct rtas_args rtas_args;
-	va_list list;
-	int i;
-
-	rtas_args.token = token;
-	rtas_args.nargs = nargs;
-	rtas_args.nret  = nret;
-	rtas_args.rets  = (rtas_arg_t *)&(rtas_args.args[nargs]);
-	va_start(list, outputs);
-	for (i = 0; i < nargs; ++i)
-		rtas_args.args[i] = va_arg(list, rtas_arg_t);
-	va_end(list);
-
-	for (i = 0; i < nret; ++i)
-		rtas_args.rets[i] = 0;
-
-	opal_enter_rtas(&rtas_args, prom_rtas_data,
-			prom_rtas_entry);
-
-	if (nret > 1 && outputs != NULL)
-		for (i = 0; i < nret-1; ++i)
-			outputs[i] = rtas_args.rets[i+1];
-	return (nret > 0)? rtas_args.rets[0]: 0;
-}
-
-static void __init prom_opal_hold_cpus(void)
-{
-	int i, cnt, cpu, rc;
-	long j;
-	phandle node;
-	char type[64];
-	u32 servers[8];
-	void *entry = (unsigned long *)&opal_secondary_entry;
-	struct opal_secondary_data *data = &opal_secondary_data;
-
-	prom_debug("prom_opal_hold_cpus: start...\n");
-	prom_debug("    - entry       = 0x%x\n", entry);
-	prom_debug("    - data        = 0x%x\n", data);
-
-	data->ack = -1;
-	data->go = 0;
-
-	/* look for cpus */
-	for (node = 0; prom_next_node(&node); ) {
-		type[0] = 0;
-		prom_getprop(node, "device_type", type, sizeof(type));
-		if (strcmp(type, "cpu") != 0)
-			continue;
-
-		/* Skip non-configured cpus. */
-		if (prom_getprop(node, "status", type, sizeof(type)) > 0)
-			if (strcmp(type, "okay") != 0)
-				continue;
-
-		cnt = prom_getprop(node, "ibm,ppc-interrupt-server#s", servers,
-			     sizeof(servers));
-		if (cnt == PROM_ERROR)
-			break;
-		cnt >>= 2;
-		for (i = 0; i < cnt; i++) {
-			cpu = servers[i];
-			prom_debug("CPU %d ... ", cpu);
-			if (cpu == prom.cpu) {
-				prom_debug("booted !\n");
-				continue;
-			}
-			prom_debug("starting ... ");
-
-			/* Init the acknowledge var which will be reset by
-			 * the secondary cpu when it awakens from its OF
-			 * spinloop.
-			 */
-			data->ack = -1;
-			rc = prom_rtas_call(prom_rtas_start_cpu, 3, 1,
-					    NULL, cpu, entry, data);
-			prom_debug("rtas rc=%d ...", rc);
-
-			for (j = 0; j < 100000000 && data->ack == -1; j++) {
-				HMT_low();
-				mb();
-			}
-			HMT_medium();
-			if (data->ack != -1)
-				prom_debug("done, PIR=0x%x\n", data->ack);
-			else
-				prom_debug("timeout !\n");
-		}
-	}
-	prom_debug("prom_opal_hold_cpus: end...\n");
-}
-
-static void __init prom_opal_takeover(void)
-{
-	struct opal_secondary_data *data = &opal_secondary_data;
-	struct opal_takeover_args *args = &data->args;
-	u64 align = prom_opal_align;
-	u64 top_addr, opal_addr;
-
-	args->k_image	= (u64)_stext;
-	args->k_size	= _end - _stext;
-	args->k_entry	= 0;
-	args->k_entry2	= 0x60;
-
-	top_addr = _ALIGN_UP(args->k_size, align);
-
-	if (prom_initrd_start != 0) {
-		args->rd_image = prom_initrd_start;
-		args->rd_size = prom_initrd_end - args->rd_image;
-		args->rd_loc = top_addr;
-		top_addr = _ALIGN_UP(args->rd_loc + args->rd_size, align);
-	}
-
-	/* Pickup an address for the HAL. We want to go really high
-	 * up to avoid problem with future kexecs. On the other hand
-	 * we don't want to be all over the TCEs on P5IOC2 machines
-	 * which are going to be up there too. We assume the machine
-	 * has plenty of memory, and we ask for the HAL for now to
-	 * be just below the 1G point, or above the initrd
-	 */
-	opal_addr = _ALIGN_DOWN(0x40000000 - prom_opal_size, align);
-	if (opal_addr < top_addr)
-		opal_addr = top_addr;
-	args->hal_addr = opal_addr;
-
-	/* Copy the command line to the kernel image */
-	strlcpy(boot_command_line, prom_cmd_line,
-		COMMAND_LINE_SIZE);
-
-	prom_debug("  k_image    = 0x%lx\n", args->k_image);
-	prom_debug("  k_size     = 0x%lx\n", args->k_size);
-	prom_debug("  k_entry    = 0x%lx\n", args->k_entry);
-	prom_debug("  k_entry2   = 0x%lx\n", args->k_entry2);
-	prom_debug("  hal_addr   = 0x%lx\n", args->hal_addr);
-	prom_debug("  rd_image   = 0x%lx\n", args->rd_image);
-	prom_debug("  rd_size    = 0x%lx\n", args->rd_size);
-	prom_debug("  rd_loc     = 0x%lx\n", args->rd_loc);
-	prom_printf("Performing OPAL takeover,this can take a few minutes..\n");
-	prom_close_stdin();
-	mb();
-	data->go = 1;
-	for (;;)
-		opal_do_takeover(args);
-}
-#endif /* __BIG_ENDIAN__ */
-
 /*
  * Allocate room for and instantiate OPAL
  */
@@ -1597,12 +1402,6 @@ static void __init prom_instantiate_rtas(void)
 			 &val, sizeof(val)) != PROM_ERROR)
 		rtas_has_query_cpu_stopped = true;
 
-#if defined(CONFIG_PPC_POWERNV) && defined(__BIG_ENDIAN__)
-	/* PowerVN takeover hack */
-	prom_rtas_data = base;
-	prom_rtas_entry = entry;
-	prom_getprop(rtas_node, "start-cpu", &prom_rtas_start_cpu, 4);
-#endif
 	prom_debug("rtas base     = 0x%x\n", base);
 	prom_debug("rtas entry    = 0x%x\n", entry);
 	prom_debug("rtas size     = 0x%x\n", (long)size);
@@ -3027,16 +2826,6 @@ unsigned long __init prom_init(unsigned long r3, unsigned long r4,
 		prom_instantiate_rtas();
 
 #ifdef CONFIG_PPC_POWERNV
-#ifdef __BIG_ENDIAN__
-	/* Detect HAL and try instanciating it & doing takeover */
-	if (of_platform == PLATFORM_PSERIES_LPAR) {
-		prom_query_opal();
-		if (of_platform == PLATFORM_OPAL) {
-			prom_opal_hold_cpus();
-			prom_opal_takeover();
-		}
-	} else
-#endif /* __BIG_ENDIAN__ */
 	if (of_platform == PLATFORM_OPAL)
 		prom_instantiate_opal();
 #endif /* CONFIG_PPC_POWERNV */
diff --git a/arch/powerpc/kernel/prom_init_check.sh b/arch/powerpc/kernel/prom_init_check.sh
index 77aa1e95e904..fe8e54b9ef7d 100644
--- a/arch/powerpc/kernel/prom_init_check.sh
+++ b/arch/powerpc/kernel/prom_init_check.sh
@@ -21,9 +21,7 @@ _end enter_prom memcpy memset reloc_offset __secondary_hold
 __secondary_hold_acknowledge __secondary_hold_spinloop __start
 strcmp strcpy strlcpy strlen strncmp strstr logo_linux_clut224
 reloc_got2 kernstart_addr memstart_addr linux_banner _stext
-opal_query_takeover opal_do_takeover opal_enter_rtas opal_secondary_entry
-boot_command_line __prom_init_toc_start __prom_init_toc_end
-btext_setup_display TOC."
+__prom_init_toc_start __prom_init_toc_end btext_setup_display TOC."
 
 NM="$1"
 OBJ="$2"
diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c
index e239df3768ac..e5b022c55ccd 100644
--- a/arch/powerpc/kernel/setup-common.c
+++ b/arch/powerpc/kernel/setup-common.c
@@ -469,9 +469,17 @@ void __init smp_setup_cpu_maps(void)
 		}
 
 		for (j = 0; j < nthreads && cpu < nr_cpu_ids; j++) {
+			bool avail;
+
 			DBG("    thread %d -> cpu %d (hard id %d)\n",
 			    j, cpu, be32_to_cpu(intserv[j]));
-			set_cpu_present(cpu, of_device_is_available(dn));
+
+			avail = of_device_is_available(dn);
+			if (!avail)
+				avail = !of_property_match_string(dn,
+						"enable-method", "spin-table");
+
+			set_cpu_present(cpu, avail);
 			set_hard_smp_processor_id(cpu, be32_to_cpu(intserv[j]));
 			set_cpu_possible(cpu, true);
 			cpu++;
diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c
index 4e47db686b5d..1bc5a1755ed4 100644
--- a/arch/powerpc/kernel/signal_32.c
+++ b/arch/powerpc/kernel/signal_32.c
@@ -54,7 +54,6 @@
 
 #include "signal.h"
 
-#undef DEBUG_SIG
 
 #ifdef CONFIG_PPC64
 #define sys_rt_sigreturn	compat_sys_rt_sigreturn
@@ -1063,10 +1062,6 @@ int handle_rt_signal32(unsigned long sig, struct k_sigaction *ka,
 	return 1;
 
 badframe:
-#ifdef DEBUG_SIG
-	printk("badframe in handle_rt_signal, regs=%p frame=%p newsp=%lx\n",
-	       regs, frame, newsp);
-#endif
 	if (show_unhandled_signals)
 		printk_ratelimited(KERN_INFO
 				   "%s[%d]: bad frame in handle_rt_signal32: "
@@ -1484,10 +1479,6 @@ int handle_signal32(unsigned long sig, struct k_sigaction *ka,
 	return 1;
 
 badframe:
-#ifdef DEBUG_SIG
-	printk("badframe in handle_signal, regs=%p frame=%p newsp=%lx\n",
-	       regs, frame, newsp);
-#endif
 	if (show_unhandled_signals)
 		printk_ratelimited(KERN_INFO
 				   "%s[%d]: bad frame in handle_signal32: "
diff --git a/arch/powerpc/kernel/signal_64.c b/arch/powerpc/kernel/signal_64.c
index d501dc4dc3e6..97c1e4b683fc 100644
--- a/arch/powerpc/kernel/signal_64.c
+++ b/arch/powerpc/kernel/signal_64.c
@@ -38,7 +38,6 @@
 
 #include "signal.h"
 
-#define DEBUG_SIG 0
 
 #define GP_REGS_SIZE	min(sizeof(elf_gregset_t), sizeof(struct pt_regs))
 #define FP_REGS_SIZE	sizeof(elf_fpregset_t)
@@ -700,10 +699,6 @@ int sys_rt_sigreturn(unsigned long r3, unsigned long r4, unsigned long r5,
 	return 0;
 
 badframe:
-#if DEBUG_SIG
-	printk("badframe in sys_rt_sigreturn, regs=%p uc=%p &uc->uc_mcontext=%p\n",
-	       regs, uc, &uc->uc_mcontext);
-#endif
 	if (show_unhandled_signals)
 		printk_ratelimited(regs->msr & MSR_64BIT ? fmt64 : fmt32,
 				   current->comm, current->pid, "rt_sigreturn",
@@ -809,10 +804,6 @@ int handle_rt_signal64(int signr, struct k_sigaction *ka, siginfo_t *info,
 	return 1;
 
 badframe:
-#if DEBUG_SIG
-	printk("badframe in setup_rt_frame, regs=%p frame=%p newsp=%lx\n",
-	       regs, frame, newsp);
-#endif
 	if (show_unhandled_signals)
 		printk_ratelimited(regs->msr & MSR_64BIT ? fmt64 : fmt32,
 				   current->comm, current->pid, "setup_rt_frame",
diff --git a/arch/powerpc/kernel/smp.c b/arch/powerpc/kernel/smp.c
index 51a3ff78838a..1007fb802e6b 100644
--- a/arch/powerpc/kernel/smp.c
+++ b/arch/powerpc/kernel/smp.c
@@ -747,7 +747,7 @@ int setup_profiling_timer(unsigned int multiplier)
 
 #ifdef CONFIG_SCHED_SMT
 /* cpumask of CPUs with asymetric SMT dependancy */
-static const int powerpc_smt_flags(void)
+static int powerpc_smt_flags(void)
 {
 	int flags = SD_SHARE_CPUCAPACITY | SD_SHARE_PKG_RESOURCES;
 
diff --git a/arch/powerpc/kvm/book3s_64_mmu_hv.c b/arch/powerpc/kvm/book3s_64_mmu_hv.c
index 80561074078d..68468d695f12 100644
--- a/arch/powerpc/kvm/book3s_64_mmu_hv.c
+++ b/arch/powerpc/kvm/book3s_64_mmu_hv.c
@@ -1562,7 +1562,7 @@ static ssize_t kvm_htab_write(struct file *file, const char __user *buf,
 				goto out;
 			}
 			if (!rma_setup && is_vrma_hpte(v)) {
-				unsigned long psize = hpte_page_size(v, r);
+				unsigned long psize = hpte_base_page_size(v, r);
 				unsigned long senc = slb_pgsize_encoding(psize);
 				unsigned long lpcr;
 
diff --git a/arch/powerpc/kvm/book3s_hv_interrupts.S b/arch/powerpc/kvm/book3s_hv_interrupts.S
index 8c86422a1e37..731be7478b27 100644
--- a/arch/powerpc/kvm/book3s_hv_interrupts.S
+++ b/arch/powerpc/kvm/book3s_hv_interrupts.S
@@ -127,11 +127,6 @@ BEGIN_FTR_SECTION
 	stw	r10, HSTATE_PMC + 24(r13)
 	stw	r11, HSTATE_PMC + 28(r13)
 END_FTR_SECTION_IFSET(CPU_FTR_ARCH_201)
-BEGIN_FTR_SECTION
-	mfspr	r9, SPRN_SIER
-	std	r8, HSTATE_MMCR + 40(r13)
-	std	r9, HSTATE_MMCR + 48(r13)
-END_FTR_SECTION_IFSET(CPU_FTR_ARCH_207S)
 31:
 
 	/*
diff --git a/arch/powerpc/kvm/book3s_hv_rm_mmu.c b/arch/powerpc/kvm/book3s_hv_rm_mmu.c
index 6e6224318c36..5a24d3c2b6b8 100644
--- a/arch/powerpc/kvm/book3s_hv_rm_mmu.c
+++ b/arch/powerpc/kvm/book3s_hv_rm_mmu.c
@@ -814,13 +814,10 @@ long kvmppc_hv_find_lock_hpte(struct kvm *kvm, gva_t eaddr, unsigned long slb_v,
 			r = hpte[i+1];
 
 			/*
-			 * Check the HPTE again, including large page size
-			 * Since we don't currently allow any MPSS (mixed
-			 * page-size segment) page sizes, it is sufficient
-			 * to check against the actual page size.
+			 * Check the HPTE again, including base page size
 			 */
 			if ((v & valid) && (v & mask) == val &&
-			    hpte_page_size(v, r) == (1ul << pshift))
+			    hpte_base_page_size(v, r) == (1ul << pshift))
 				/* Return with the HPTE still locked */
 				return (hash << 3) + (i >> 1);
 
diff --git a/arch/powerpc/kvm/book3s_hv_rmhandlers.S b/arch/powerpc/kvm/book3s_hv_rmhandlers.S
index 868347ef09fd..558a67df8126 100644
--- a/arch/powerpc/kvm/book3s_hv_rmhandlers.S
+++ b/arch/powerpc/kvm/book3s_hv_rmhandlers.S
@@ -48,7 +48,7 @@
  *
  * LR = return address to continue at after eventually re-enabling MMU
  */
-_GLOBAL(kvmppc_hv_entry_trampoline)
+_GLOBAL_TOC(kvmppc_hv_entry_trampoline)
 	mflr	r0
 	std	r0, PPC_LR_STKOFF(r1)
 	stdu	r1, -112(r1)
diff --git a/arch/powerpc/kvm/book3s_interrupts.S b/arch/powerpc/kvm/book3s_interrupts.S
index e2c29e381dc7..d044b8b7c69d 100644
--- a/arch/powerpc/kvm/book3s_interrupts.S
+++ b/arch/powerpc/kvm/book3s_interrupts.S
@@ -25,7 +25,11 @@
 #include <asm/exception-64s.h>
 
 #if defined(CONFIG_PPC_BOOK3S_64)
+#if defined(_CALL_ELF) && _CALL_ELF == 2
+#define FUNC(name) 		name
+#else
 #define FUNC(name) 		GLUE(.,name)
+#endif
 #define GET_SHADOW_VCPU(reg)    addi	reg, r13, PACA_SVCPU
 
 #elif defined(CONFIG_PPC_BOOK3S_32)
diff --git a/arch/powerpc/kvm/book3s_rmhandlers.S b/arch/powerpc/kvm/book3s_rmhandlers.S
index 9eec675220e6..16c4d88ba27d 100644
--- a/arch/powerpc/kvm/book3s_rmhandlers.S
+++ b/arch/powerpc/kvm/book3s_rmhandlers.S
@@ -36,7 +36,11 @@
 
 #if defined(CONFIG_PPC_BOOK3S_64)
 
+#if defined(_CALL_ELF) && _CALL_ELF == 2
+#define FUNC(name) 		name
+#else
 #define FUNC(name) 		GLUE(.,name)
+#endif
 
 #elif defined(CONFIG_PPC_BOOK3S_32)
 
@@ -146,7 +150,7 @@ kvmppc_handler_skip_ins:
  * On entry, r4 contains the guest shadow MSR
  * MSR.EE has to be 0 when calling this function
  */
-_GLOBAL(kvmppc_entry_trampoline)
+_GLOBAL_TOC(kvmppc_entry_trampoline)
 	mfmsr	r5
 	LOAD_REG_ADDR(r7, kvmppc_handler_trampoline_enter)
 	toreal(r7)
diff --git a/arch/powerpc/kvm/book3s_rtas.c b/arch/powerpc/kvm/book3s_rtas.c
index edb14ba992b3..ef27fbd5d9c5 100644
--- a/arch/powerpc/kvm/book3s_rtas.c
+++ b/arch/powerpc/kvm/book3s_rtas.c
@@ -23,20 +23,20 @@ static void kvm_rtas_set_xive(struct kvm_vcpu *vcpu, struct rtas_args *args)
 	u32 irq, server, priority;
 	int rc;
 
-	if (args->nargs != 3 || args->nret != 1) {
+	if (be32_to_cpu(args->nargs) != 3 || be32_to_cpu(args->nret) != 1) {
 		rc = -3;
 		goto out;
 	}
 
-	irq = args->args[0];
-	server = args->args[1];
-	priority = args->args[2];
+	irq = be32_to_cpu(args->args[0]);
+	server = be32_to_cpu(args->args[1]);
+	priority = be32_to_cpu(args->args[2]);
 
 	rc = kvmppc_xics_set_xive(vcpu->kvm, irq, server, priority);
 	if (rc)
 		rc = -3;
 out:
-	args->rets[0] = rc;
+	args->rets[0] = cpu_to_be32(rc);
 }
 
 static void kvm_rtas_get_xive(struct kvm_vcpu *vcpu, struct rtas_args *args)
@@ -44,12 +44,12 @@ static void kvm_rtas_get_xive(struct kvm_vcpu *vcpu, struct rtas_args *args)
 	u32 irq, server, priority;
 	int rc;
 
-	if (args->nargs != 1 || args->nret != 3) {
+	if (be32_to_cpu(args->nargs) != 1 || be32_to_cpu(args->nret) != 3) {
 		rc = -3;
 		goto out;
 	}
 
-	irq = args->args[0];
+	irq = be32_to_cpu(args->args[0]);
 
 	server = priority = 0;
 	rc = kvmppc_xics_get_xive(vcpu->kvm, irq, &server, &priority);
@@ -58,10 +58,10 @@ static void kvm_rtas_get_xive(struct kvm_vcpu *vcpu, struct rtas_args *args)
 		goto out;
 	}
 
-	args->rets[1] = server;
-	args->rets[2] = priority;
+	args->rets[1] = cpu_to_be32(server);
+	args->rets[2] = cpu_to_be32(priority);
 out:
-	args->rets[0] = rc;
+	args->rets[0] = cpu_to_be32(rc);
 }
 
 static void kvm_rtas_int_off(struct kvm_vcpu *vcpu, struct rtas_args *args)
@@ -69,18 +69,18 @@ static void kvm_rtas_int_off(struct kvm_vcpu *vcpu, struct rtas_args *args)
 	u32 irq;
 	int rc;
 
-	if (args->nargs != 1 || args->nret != 1) {
+	if (be32_to_cpu(args->nargs) != 1 || be32_to_cpu(args->nret) != 1) {
 		rc = -3;
 		goto out;
 	}
 
-	irq = args->args[0];
+	irq = be32_to_cpu(args->args[0]);
 
 	rc = kvmppc_xics_int_off(vcpu->kvm, irq);
 	if (rc)
 		rc = -3;
 out:
-	args->rets[0] = rc;
+	args->rets[0] = cpu_to_be32(rc);
 }
 
 static void kvm_rtas_int_on(struct kvm_vcpu *vcpu, struct rtas_args *args)
@@ -88,18 +88,18 @@ static void kvm_rtas_int_on(struct kvm_vcpu *vcpu, struct rtas_args *args)
 	u32 irq;
 	int rc;
 
-	if (args->nargs != 1 || args->nret != 1) {
+	if (be32_to_cpu(args->nargs) != 1 || be32_to_cpu(args->nret) != 1) {
 		rc = -3;
 		goto out;
 	}
 
-	irq = args->args[0];
+	irq = be32_to_cpu(args->args[0]);
 
 	rc = kvmppc_xics_int_on(vcpu->kvm, irq);
 	if (rc)
 		rc = -3;
 out:
-	args->rets[0] = rc;
+	args->rets[0] = cpu_to_be32(rc);
 }
 #endif /* CONFIG_KVM_XICS */
 
@@ -205,32 +205,6 @@ int kvm_vm_ioctl_rtas_define_token(struct kvm *kvm, void __user *argp)
 	return rc;
 }
 
-static void kvmppc_rtas_swap_endian_in(struct rtas_args *args)
-{
-#ifdef __LITTLE_ENDIAN__
-	int i;
-
-	args->token = be32_to_cpu(args->token);
-	args->nargs = be32_to_cpu(args->nargs);
-	args->nret = be32_to_cpu(args->nret);
-	for (i = 0; i < args->nargs; i++)
-		args->args[i] = be32_to_cpu(args->args[i]);
-#endif
-}
-
-static void kvmppc_rtas_swap_endian_out(struct rtas_args *args)
-{
-#ifdef __LITTLE_ENDIAN__
-	int i;
-
-	for (i = 0; i < args->nret; i++)
-		args->args[i] = cpu_to_be32(args->args[i]);
-	args->token = cpu_to_be32(args->token);
-	args->nargs = cpu_to_be32(args->nargs);
-	args->nret = cpu_to_be32(args->nret);
-#endif
-}
-
 int kvmppc_rtas_hcall(struct kvm_vcpu *vcpu)
 {
 	struct rtas_token_definition *d;
@@ -249,8 +223,6 @@ int kvmppc_rtas_hcall(struct kvm_vcpu *vcpu)
 	if (rc)
 		goto fail;
 
-	kvmppc_rtas_swap_endian_in(&args);
-
 	/*
 	 * args->rets is a pointer into args->args. Now that we've
 	 * copied args we need to fix it up to point into our copy,
@@ -258,13 +230,13 @@ int kvmppc_rtas_hcall(struct kvm_vcpu *vcpu)
 	 * value so we can restore it on the way out.
 	 */
 	orig_rets = args.rets;
-	args.rets = &args.args[args.nargs];
+	args.rets = &args.args[be32_to_cpu(args.nargs)];
 
 	mutex_lock(&vcpu->kvm->lock);
 
 	rc = -ENOENT;
 	list_for_each_entry(d, &vcpu->kvm->arch.rtas_tokens, list) {
-		if (d->token == args.token) {
+		if (d->token == be32_to_cpu(args.token)) {
 			d->handler->handler(vcpu, &args);
 			rc = 0;
 			break;
@@ -275,7 +247,6 @@ int kvmppc_rtas_hcall(struct kvm_vcpu *vcpu)
 
 	if (rc == 0) {
 		args.rets = orig_rets;
-		kvmppc_rtas_swap_endian_out(&args);
 		rc = kvm_write_guest(vcpu->kvm, args_phys, &args, sizeof(args));
 		if (rc)
 			goto fail;
diff --git a/arch/powerpc/kvm/e500_mmu_host.c b/arch/powerpc/kvm/e500_mmu_host.c
index dd2cc03f406f..86903d3f5a03 100644
--- a/arch/powerpc/kvm/e500_mmu_host.c
+++ b/arch/powerpc/kvm/e500_mmu_host.c
@@ -473,7 +473,8 @@ static inline int kvmppc_e500_shadow_map(struct kvmppc_vcpu_e500 *vcpu_e500,
 		if (printk_ratelimit())
 			pr_err("%s: pte not present: gfn %lx, pfn %lx\n",
 				__func__, (long)gfn, pfn);
-		return -EINVAL;
+		ret = -EINVAL;
+		goto out;
 	}
 	kvmppc_e500_ref_setup(ref, gtlbe, pfn, wimg);
 
diff --git a/arch/powerpc/lib/mem_64.S b/arch/powerpc/lib/mem_64.S
index 0738f96befbf..43435c6892fb 100644
--- a/arch/powerpc/lib/mem_64.S
+++ b/arch/powerpc/lib/mem_64.S
@@ -77,7 +77,7 @@ _GLOBAL(memset)
 	stb	r4,0(r6)
 	blr
 
-_GLOBAL(memmove)
+_GLOBAL_TOC(memmove)
 	cmplw	0,r3,r4
 	bgt	backwards_memcpy
 	b	memcpy
diff --git a/arch/powerpc/lib/sstep.c b/arch/powerpc/lib/sstep.c
index 412dd46dd0b7..5c09f365c842 100644
--- a/arch/powerpc/lib/sstep.c
+++ b/arch/powerpc/lib/sstep.c
@@ -1198,7 +1198,7 @@ int __kprobes emulate_step(struct pt_regs *regs, unsigned int instr)
 			sh = regs->gpr[rb] & 0x3f;
 			ival = (signed int) regs->gpr[rd];
 			regs->gpr[ra] = ival >> (sh < 32 ? sh : 31);
-			if (ival < 0 && (sh >= 32 || (ival & ((1 << sh) - 1)) != 0))
+			if (ival < 0 && (sh >= 32 || (ival & ((1ul << sh) - 1)) != 0))
 				regs->xer |= XER_CA;
 			else
 				regs->xer &= ~XER_CA;
@@ -1208,7 +1208,7 @@ int __kprobes emulate_step(struct pt_regs *regs, unsigned int instr)
 			sh = rb;
 			ival = (signed int) regs->gpr[rd];
 			regs->gpr[ra] = ival >> sh;
-			if (ival < 0 && (ival & ((1 << sh) - 1)) != 0)
+			if (ival < 0 && (ival & ((1ul << sh) - 1)) != 0)
 				regs->xer |= XER_CA;
 			else
 				regs->xer &= ~XER_CA;
@@ -1216,7 +1216,7 @@ int __kprobes emulate_step(struct pt_regs *regs, unsigned int instr)
 
 #ifdef __powerpc64__
 		case 27:	/* sld */
-			sh = regs->gpr[rd] & 0x7f;
+			sh = regs->gpr[rb] & 0x7f;
 			if (sh < 64)
 				regs->gpr[ra] = regs->gpr[rd] << sh;
 			else
@@ -1235,7 +1235,7 @@ int __kprobes emulate_step(struct pt_regs *regs, unsigned int instr)
 			sh = regs->gpr[rb] & 0x7f;
 			ival = (signed long int) regs->gpr[rd];
 			regs->gpr[ra] = ival >> (sh < 64 ? sh : 63);
-			if (ival < 0 && (sh >= 64 || (ival & ((1 << sh) - 1)) != 0))
+			if (ival < 0 && (sh >= 64 || (ival & ((1ul << sh) - 1)) != 0))
 				regs->xer |= XER_CA;
 			else
 				regs->xer &= ~XER_CA;
@@ -1246,7 +1246,7 @@ int __kprobes emulate_step(struct pt_regs *regs, unsigned int instr)
 			sh = rb | ((instr & 2) << 4);
 			ival = (signed long int) regs->gpr[rd];
 			regs->gpr[ra] = ival >> sh;
-			if (ival < 0 && (ival & ((1 << sh) - 1)) != 0)
+			if (ival < 0 && (ival & ((1ul << sh) - 1)) != 0)
 				regs->xer |= XER_CA;
 			else
 				regs->xer &= ~XER_CA;
diff --git a/arch/powerpc/mm/mmu_context_nohash.c b/arch/powerpc/mm/mmu_context_nohash.c
index af3d78e19302..928ebe79668b 100644
--- a/arch/powerpc/mm/mmu_context_nohash.c
+++ b/arch/powerpc/mm/mmu_context_nohash.c
@@ -410,17 +410,7 @@ void __init mmu_context_init(void)
 	} else if (mmu_has_feature(MMU_FTR_TYPE_47x)) {
 		first_context = 1;
 		last_context = 65535;
-	} else
-#ifdef CONFIG_PPC_BOOK3E_MMU
-	if (mmu_has_feature(MMU_FTR_TYPE_3E)) {
-		u32 mmucfg = mfspr(SPRN_MMUCFG);
-		u32 pid_bits = (mmucfg & MMUCFG_PIDSIZE_MASK)
-				>> MMUCFG_PIDSIZE_SHIFT;
-		first_context = 1;
-		last_context = (1UL << (pid_bits + 1)) - 1;
-	} else
-#endif
-	{
+	} else {
 		first_context = 1;
 		last_context = 255;
 	}
diff --git a/arch/powerpc/net/bpf_jit_comp.c b/arch/powerpc/net/bpf_jit_comp.c
index 6dcdadefd8d0..82e82cadcde5 100644
--- a/arch/powerpc/net/bpf_jit_comp.c
+++ b/arch/powerpc/net/bpf_jit_comp.c
@@ -390,12 +390,16 @@ static int bpf_jit_build_body(struct sk_filter *fp, u32 *image,
 		case BPF_ANC | SKF_AD_VLAN_TAG:
 		case BPF_ANC | SKF_AD_VLAN_TAG_PRESENT:
 			BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, vlan_tci) != 2);
+			BUILD_BUG_ON(VLAN_TAG_PRESENT != 0x1000);
+
 			PPC_LHZ_OFFS(r_A, r_skb, offsetof(struct sk_buff,
 							  vlan_tci));
-			if (code == (BPF_ANC | SKF_AD_VLAN_TAG))
-				PPC_ANDI(r_A, r_A, VLAN_VID_MASK);
-			else
+			if (code == (BPF_ANC | SKF_AD_VLAN_TAG)) {
+				PPC_ANDI(r_A, r_A, ~VLAN_TAG_PRESENT);
+			} else {
 				PPC_ANDI(r_A, r_A, VLAN_TAG_PRESENT);
+				PPC_SRWI(r_A, r_A, 12);
+			}
 			break;
 		case BPF_ANC | SKF_AD_QUEUE:
 			BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff,
diff --git a/arch/powerpc/perf/core-book3s.c b/arch/powerpc/perf/core-book3s.c
index 4520c9356b54..6b0641c3f03f 100644
--- a/arch/powerpc/perf/core-book3s.c
+++ b/arch/powerpc/perf/core-book3s.c
@@ -485,7 +485,7 @@ static bool is_ebb_event(struct perf_event *event)
 	 * check that the PMU supports EBB, meaning those that don't can still
 	 * use bit 63 of the event code for something else if they wish.
 	 */
-	return (ppmu->flags & PPMU_EBB) &&
+	return (ppmu->flags & PPMU_ARCH_207S) &&
 	       ((event->attr.config >> PERF_EVENT_CONFIG_EBB_SHIFT) & 1);
 }
 
@@ -777,7 +777,7 @@ void perf_event_print_debug(void)
 	if (ppmu->flags & PPMU_HAS_SIER)
 		sier = mfspr(SPRN_SIER);
 
-	if (ppmu->flags & PPMU_EBB) {
+	if (ppmu->flags & PPMU_ARCH_207S) {
 		pr_info("MMCR2: %016lx EBBHR: %016lx\n",
 			mfspr(SPRN_MMCR2), mfspr(SPRN_EBBHR));
 		pr_info("EBBRR: %016lx BESCR: %016lx\n",
@@ -996,7 +996,22 @@ static void power_pmu_read(struct perf_event *event)
 	} while (local64_cmpxchg(&event->hw.prev_count, prev, val) != prev);
 
 	local64_add(delta, &event->count);
-	local64_sub(delta, &event->hw.period_left);
+
+	/*
+	 * A number of places program the PMC with (0x80000000 - period_left).
+	 * We never want period_left to be less than 1 because we will program
+	 * the PMC with a value >= 0x800000000 and an edge detected PMC will
+	 * roll around to 0 before taking an exception. We have seen this
+	 * on POWER8.
+	 *
+	 * To fix this, clamp the minimum value of period_left to 1.
+	 */
+	do {
+		prev = local64_read(&event->hw.period_left);
+		val = prev - delta;
+		if (val < 1)
+			val = 1;
+	} while (local64_cmpxchg(&event->hw.period_left, prev, val) != prev);
 }
 
 /*
@@ -1300,6 +1315,9 @@ static void power_pmu_enable(struct pmu *pmu)
 
 	write_mmcr0(cpuhw, mmcr0);
 
+	if (ppmu->flags & PPMU_ARCH_207S)
+		mtspr(SPRN_MMCR2, 0);
+
 	/*
 	 * Enable instruction sampling if necessary
 	 */
@@ -1696,7 +1714,7 @@ static int power_pmu_event_init(struct perf_event *event)
 
 	if (has_branch_stack(event)) {
 	        /* PMU has BHRB enabled */
-		if (!(ppmu->flags & PPMU_BHRB))
+		if (!(ppmu->flags & PPMU_ARCH_207S))
 			return -EOPNOTSUPP;
 	}
 
diff --git a/arch/powerpc/perf/power8-pmu.c b/arch/powerpc/perf/power8-pmu.c
index fe2763b6e039..639cd9156585 100644
--- a/arch/powerpc/perf/power8-pmu.c
+++ b/arch/powerpc/perf/power8-pmu.c
@@ -792,7 +792,7 @@ static struct power_pmu power8_pmu = {
 	.get_constraint		= power8_get_constraint,
 	.get_alternatives	= power8_get_alternatives,
 	.disable_pmc		= power8_disable_pmc,
-	.flags			= PPMU_HAS_SSLOT | PPMU_HAS_SIER | PPMU_BHRB | PPMU_EBB,
+	.flags			= PPMU_HAS_SSLOT | PPMU_HAS_SIER | PPMU_ARCH_207S,
 	.n_generic		= ARRAY_SIZE(power8_generic_events),
 	.generic_events		= power8_generic_events,
 	.cache_events		= &power8_cache_events,
diff --git a/arch/powerpc/platforms/cell/cbe_thermal.c b/arch/powerpc/platforms/cell/cbe_thermal.c
index 94560db788bf..2c15ff094483 100644
--- a/arch/powerpc/platforms/cell/cbe_thermal.c
+++ b/arch/powerpc/platforms/cell/cbe_thermal.c
@@ -125,7 +125,7 @@ static ssize_t show_throttle(struct cbe_pmd_regs __iomem *pmd_regs, char *buf, i
 static ssize_t store_throttle(struct cbe_pmd_regs __iomem *pmd_regs, const char *buf, size_t size, int pos)
 {
 	u64 reg_value;
-	int temp;
+	unsigned int temp;
 	u64 new_value;
 	int ret;
 
diff --git a/arch/powerpc/platforms/cell/spu_syscalls.c b/arch/powerpc/platforms/cell/spu_syscalls.c
index 38e0a1a5cec3..5e6e0bad6db6 100644
--- a/arch/powerpc/platforms/cell/spu_syscalls.c
+++ b/arch/powerpc/platforms/cell/spu_syscalls.c
@@ -111,6 +111,7 @@ asmlinkage long sys_spu_run(int fd, __u32 __user *unpc, __u32 __user *ustatus)
 	return ret;
 }
 
+#ifdef CONFIG_COREDUMP
 int elf_coredump_extra_notes_size(void)
 {
 	struct spufs_calls *calls;
@@ -142,6 +143,7 @@ int elf_coredump_extra_notes_write(struct coredump_params *cprm)
 
 	return ret;
 }
+#endif
 
 void notify_spus_active(void)
 {
diff --git a/arch/powerpc/platforms/cell/spufs/Makefile b/arch/powerpc/platforms/cell/spufs/Makefile
index b9d5d678aa44..52a7d2596d30 100644
--- a/arch/powerpc/platforms/cell/spufs/Makefile
+++ b/arch/powerpc/platforms/cell/spufs/Makefile
@@ -1,8 +1,9 @@
 
 obj-$(CONFIG_SPU_FS) += spufs.o
-spufs-y += inode.o file.o context.o syscalls.o coredump.o
+spufs-y += inode.o file.o context.o syscalls.o
 spufs-y += sched.o backing_ops.o hw_ops.o run.o gang.o
 spufs-y += switch.o fault.o lscsa_alloc.o
+spufs-$(CONFIG_COREDUMP) += coredump.o
 
 # magic for the trace events
 CFLAGS_sched.o := -I$(src)
diff --git a/arch/powerpc/platforms/cell/spufs/syscalls.c b/arch/powerpc/platforms/cell/spufs/syscalls.c
index b045fdda4845..a87200a535fa 100644
--- a/arch/powerpc/platforms/cell/spufs/syscalls.c
+++ b/arch/powerpc/platforms/cell/spufs/syscalls.c
@@ -79,8 +79,10 @@ static long do_spu_create(const char __user *pathname, unsigned int flags,
 struct spufs_calls spufs_calls = {
 	.create_thread = do_spu_create,
 	.spu_run = do_spu_run,
-	.coredump_extra_notes_size = spufs_coredump_extra_notes_size,
-	.coredump_extra_notes_write = spufs_coredump_extra_notes_write,
 	.notify_spus_active = do_notify_spus_active,
 	.owner = THIS_MODULE,
+#ifdef CONFIG_COREDUMP
+	.coredump_extra_notes_size = spufs_coredump_extra_notes_size,
+	.coredump_extra_notes_write = spufs_coredump_extra_notes_write,
+#endif
 };
diff --git a/arch/powerpc/platforms/powernv/Makefile b/arch/powerpc/platforms/powernv/Makefile
index d55891f89a2c..4ad227d04c1a 100644
--- a/arch/powerpc/platforms/powernv/Makefile
+++ b/arch/powerpc/platforms/powernv/Makefile
@@ -1,4 +1,4 @@
-obj-y			+= setup.o opal-takeover.o opal-wrappers.o opal.o opal-async.o
+obj-y			+= setup.o opal-wrappers.o opal.o opal-async.o
 obj-y			+= opal-rtc.o opal-nvram.o opal-lpc.o opal-flash.o
 obj-y			+= rng.o opal-elog.o opal-dump.o opal-sysparam.o opal-sensor.o
 obj-y			+= opal-msglog.o
diff --git a/arch/powerpc/platforms/powernv/opal-takeover.S b/arch/powerpc/platforms/powernv/opal-takeover.S
deleted file mode 100644
index 11a3169ee583..000000000000
--- a/arch/powerpc/platforms/powernv/opal-takeover.S
+++ /dev/null
@@ -1,140 +0,0 @@
-/*
- * PowerNV OPAL takeover assembly code, for use by prom_init.c
- *
- * Copyright 2011 IBM Corp.
- *
- * 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 <asm/ppc_asm.h>
-#include <asm/hvcall.h>
-#include <asm/asm-offsets.h>
-#include <asm/opal.h>
-
-#define H_HAL_TAKEOVER			0x5124
-#define H_HAL_TAKEOVER_QUERY_MAGIC	-1
-
-	.text
-_GLOBAL(opal_query_takeover)
-	mfcr	r0
-	stw	r0,8(r1)
-	stdu	r1,-STACKFRAMESIZE(r1)
-	std	r3,STK_PARAM(R3)(r1)
-	std	r4,STK_PARAM(R4)(r1)
-	li	r3,H_HAL_TAKEOVER
-	li	r4,H_HAL_TAKEOVER_QUERY_MAGIC
-	HVSC
-	addi	r1,r1,STACKFRAMESIZE
-	ld	r10,STK_PARAM(R3)(r1)
-	std	r4,0(r10)
-	ld	r10,STK_PARAM(R4)(r1)
-	std	r5,0(r10)
-	lwz	r0,8(r1)
-	mtcrf	0xff,r0
-	blr
-
-_GLOBAL(opal_do_takeover)
-	mfcr	r0
-	stw	r0,8(r1)
-	mflr	r0
-	std	r0,16(r1)
-	bl	__opal_do_takeover
-	ld	r0,16(r1)
-	mtlr	r0
-	lwz	r0,8(r1)
-	mtcrf	0xff,r0
-	blr
-
-__opal_do_takeover:
-	ld	r4,0(r3)
-	ld	r5,0x8(r3)
-	ld	r6,0x10(r3)
-	ld	r7,0x18(r3)
-	ld	r8,0x20(r3)
-	ld	r9,0x28(r3)
-	ld	r10,0x30(r3)
-	ld	r11,0x38(r3)
-	li	r3,H_HAL_TAKEOVER
-	HVSC
-	blr
-
-	.globl opal_secondary_entry
-opal_secondary_entry:
-	mr	r31,r3
-	mfmsr	r11
-	li	r12,(MSR_SF | MSR_ISF)@highest
-	sldi	r12,r12,48
-	or	r11,r11,r12
-	mtmsrd	r11
-	isync
-	mfspr	r4,SPRN_PIR
-	std	r4,0(r3)
-1:	HMT_LOW
-	ld	r4,8(r3)
-	cmpli	cr0,r4,0
-	beq	1b
-	HMT_MEDIUM
-1:	addi	r3,r31,16
-	bl	__opal_do_takeover
-	b	1b
-
-_GLOBAL(opal_enter_rtas)
-	mflr	r0
-	std	r0,16(r1)
-        stdu	r1,-PROM_FRAME_SIZE(r1)	/* Save SP and create stack space */
-
-	/* Because PROM is running in 32b mode, it clobbers the high order half
-	 * of all registers that it saves.  We therefore save those registers
-	 * PROM might touch to the stack.  (r0, r3-r13 are caller saved)
-	*/
-	SAVE_GPR(2, r1)
-	SAVE_GPR(13, r1)
-	SAVE_8GPRS(14, r1)
-	SAVE_10GPRS(22, r1)
-	mfcr	r10
-	mfmsr	r11
-	std	r10,_CCR(r1)
-	std	r11,_MSR(r1)
-
-	/* Get the PROM entrypoint */
-	mtlr	r5
-
-	/* Switch MSR to 32 bits mode
-	 */
-        li      r12,1
-        rldicr  r12,r12,MSR_SF_LG,(63-MSR_SF_LG)
-        andc    r11,r11,r12
-        li      r12,1
-        rldicr  r12,r12,MSR_ISF_LG,(63-MSR_ISF_LG)
-        andc    r11,r11,r12
-        mtmsrd  r11
-        isync
-
-	/* Enter RTAS here... */
-	blrl
-
-	/* Just make sure that r1 top 32 bits didn't get
-	 * corrupt by OF
-	 */
-	rldicl	r1,r1,0,32
-
-	/* Restore the MSR (back to 64 bits) */
-	ld	r0,_MSR(r1)
-	MTMSRD(r0)
-        isync
-
-	/* Restore other registers */
-	REST_GPR(2, r1)
-	REST_GPR(13, r1)
-	REST_8GPRS(14, r1)
-	REST_10GPRS(22, r1)
-	ld	r4,_CCR(r1)
-	mtcr	r4
-
-        addi	r1,r1,PROM_FRAME_SIZE
-	ld	r0,16(r1)
-	mtlr    r0
-	blr
diff --git a/arch/powerpc/platforms/pseries/dlpar.c b/arch/powerpc/platforms/pseries/dlpar.c
index 022b38e6a80b..2d0b4d68a40a 100644
--- a/arch/powerpc/platforms/pseries/dlpar.c
+++ b/arch/powerpc/platforms/pseries/dlpar.c
@@ -86,6 +86,7 @@ static struct device_node *dlpar_parse_cc_node(struct cc_workarea *ccwa,
 	}
 
 	of_node_set_flag(dn, OF_DYNAMIC);
+	of_node_init(dn);
 
 	return dn;
 }
diff --git a/arch/powerpc/platforms/pseries/reconfig.c b/arch/powerpc/platforms/pseries/reconfig.c
index 0435bb65d0aa..1c0a60d98867 100644
--- a/arch/powerpc/platforms/pseries/reconfig.c
+++ b/arch/powerpc/platforms/pseries/reconfig.c
@@ -69,6 +69,7 @@ static int pSeries_reconfig_add_node(const char *path, struct property *proplist
 
 	np->properties = proplist;
 	of_node_set_flag(np, OF_DYNAMIC);
+	of_node_init(np);
 
 	np->parent = derive_parent(path);
 	if (IS_ERR(np->parent)) {
diff --git a/arch/powerpc/sysdev/dart_iommu.c b/arch/powerpc/sysdev/dart_iommu.c
index 62c47bb76517..9e5353ff6d1b 100644
--- a/arch/powerpc/sysdev/dart_iommu.c
+++ b/arch/powerpc/sysdev/dart_iommu.c
@@ -476,6 +476,11 @@ void __init alloc_dart_table(void)
 	 */
 	dart_tablebase = (unsigned long)
 		__va(memblock_alloc_base(1UL<<24, 1UL<<24, 0x80000000L));
+	/*
+	 * The DART space is later unmapped from the kernel linear mapping and
+	 * accessing dart_tablebase during kmemleak scanning will fault.
+	 */
+	kmemleak_no_scan((void *)dart_tablebase);
 
 	printk(KERN_INFO "DART table allocated at: %lx\n", dart_tablebase);
 }
diff --git a/arch/s390/include/asm/switch_to.h b/arch/s390/include/asm/switch_to.h
index df38c70cd59e..18ea9e3f8142 100644
--- a/arch/s390/include/asm/switch_to.h
+++ b/arch/s390/include/asm/switch_to.h
@@ -51,8 +51,8 @@ static inline int restore_fp_ctl(u32 *fpc)
 		return 0;
 
 	asm volatile(
-		"0:	lfpc    %1\n"
-		"	la	%0,0\n"
+		"	lfpc    %1\n"
+		"0:	la	%0,0\n"
 		"1:\n"
 		EX_TABLE(0b,1b)
 		: "=d" (rc) : "Q" (*fpc), "0" (-EINVAL));
diff --git a/arch/s390/include/uapi/asm/Kbuild b/arch/s390/include/uapi/asm/Kbuild
index 0e2b54db82bc..08fe6dad9026 100644
--- a/arch/s390/include/uapi/asm/Kbuild
+++ b/arch/s390/include/uapi/asm/Kbuild
@@ -37,6 +37,7 @@ header-y += signal.h
 header-y += socket.h
 header-y += sockios.h
 header-y += sclp_ctl.h
+header-y += sie.h
 header-y += stat.h
 header-y += statfs.h
 header-y += swab.h
diff --git a/arch/s390/include/uapi/asm/sie.h b/arch/s390/include/uapi/asm/sie.h
index 3d97f610198d..5d9cc19462c4 100644
--- a/arch/s390/include/uapi/asm/sie.h
+++ b/arch/s390/include/uapi/asm/sie.h
@@ -1,8 +1,6 @@
 #ifndef _UAPI_ASM_S390_SIE_H
 #define _UAPI_ASM_S390_SIE_H
 
-#include <asm/sigp.h>
-
 #define diagnose_codes						\
 	{ 0x10, "DIAG (0x10) release pages" },			\
 	{ 0x44, "DIAG (0x44) time slice end" },			\
@@ -13,18 +11,18 @@
 	{ 0x500, "DIAG (0x500) KVM virtio functions" },		\
 	{ 0x501, "DIAG (0x501) KVM breakpoint" }
 
-#define sigp_order_codes						\
-	{ SIGP_SENSE, "SIGP sense" },					\
-	{ SIGP_EXTERNAL_CALL, "SIGP external call" },			\
-	{ SIGP_EMERGENCY_SIGNAL, "SIGP emergency signal" },		\
-	{ SIGP_STOP, "SIGP stop" },					\
-	{ SIGP_STOP_AND_STORE_STATUS, "SIGP stop and store status" },	\
-	{ SIGP_SET_ARCHITECTURE, "SIGP set architecture" },		\
-	{ SIGP_SET_PREFIX, "SIGP set prefix" },				\
-	{ SIGP_SENSE_RUNNING, "SIGP sense running" },			\
-	{ SIGP_RESTART, "SIGP restart" },				\
-	{ SIGP_INITIAL_CPU_RESET, "SIGP initial cpu reset" },		\
-	{ SIGP_STORE_STATUS_AT_ADDRESS, "SIGP store status at address" }
+#define sigp_order_codes					\
+	{ 0x01, "SIGP sense" },					\
+	{ 0x02, "SIGP external call" },				\
+	{ 0x03, "SIGP emergency signal" },			\
+	{ 0x05, "SIGP stop" },					\
+	{ 0x06, "SIGP restart" },				\
+	{ 0x09, "SIGP stop and store status" },			\
+	{ 0x0b, "SIGP initial cpu reset" },			\
+	{ 0x0d, "SIGP set prefix" },				\
+	{ 0x0e, "SIGP store status at address" },		\
+	{ 0x12, "SIGP set architecture" },			\
+	{ 0x15, "SIGP sense running" }
 
 #define icpt_prog_codes						\
 	{ 0x0001, "Prog Operation" },				\
diff --git a/arch/s390/kernel/head.S b/arch/s390/kernel/head.S
index 7ba7d6784510..e88d35d74950 100644
--- a/arch/s390/kernel/head.S
+++ b/arch/s390/kernel/head.S
@@ -437,11 +437,11 @@ ENTRY(startup_kdump)
 
 #if defined(CONFIG_64BIT)
 #if defined(CONFIG_MARCH_ZEC12)
-	.long 3, 0xc100efea, 0xf46ce800, 0x00400000
+	.long 3, 0xc100eff2, 0xf46ce800, 0x00400000
 #elif defined(CONFIG_MARCH_Z196)
-	.long 2, 0xc100efea, 0xf46c0000
+	.long 2, 0xc100eff2, 0xf46c0000
 #elif defined(CONFIG_MARCH_Z10)
-	.long 2, 0xc100efea, 0xf0680000
+	.long 2, 0xc100eff2, 0xf0680000
 #elif defined(CONFIG_MARCH_Z9_109)
 	.long 1, 0xc100efc2
 #elif defined(CONFIG_MARCH_Z990)
diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c
index 2d716734b5b1..5dc7ad9e2fbf 100644
--- a/arch/s390/kernel/ptrace.c
+++ b/arch/s390/kernel/ptrace.c
@@ -334,9 +334,14 @@ static int __poke_user(struct task_struct *child, addr_t addr, addr_t data)
 			unsigned long mask = PSW_MASK_USER;
 
 			mask |= is_ri_task(child) ? PSW_MASK_RI : 0;
-			if ((data & ~mask) != PSW_USER_BITS)
+			if ((data ^ PSW_USER_BITS) & ~mask)
+				/* Invalid psw mask. */
+				return -EINVAL;
+			if ((data & PSW_MASK_ASC) == PSW_ASC_HOME)
+				/* Invalid address-space-control bits */
 				return -EINVAL;
 			if ((data & PSW_MASK_EA) && !(data & PSW_MASK_BA))
+				/* Invalid addressing mode bits */
 				return -EINVAL;
 		}
 		*(addr_t *)((addr_t) &task_pt_regs(child)->psw + addr) = data;
@@ -672,9 +677,12 @@ static int __poke_user_compat(struct task_struct *child,
 
 			mask |= is_ri_task(child) ? PSW32_MASK_RI : 0;
 			/* Build a 64 bit psw mask from 31 bit mask. */
-			if ((tmp & ~mask) != PSW32_USER_BITS)
+			if ((tmp ^ PSW32_USER_BITS) & ~mask)
 				/* Invalid psw mask. */
 				return -EINVAL;
+			if ((data & PSW32_MASK_ASC) == PSW32_ASC_HOME)
+				/* Invalid address-space-control bits */
+				return -EINVAL;
 			regs->psw.mask = (regs->psw.mask & ~PSW_MASK_USER) |
 				(regs->psw.mask & PSW_MASK_BA) |
 				(__u64)(tmp & mask) << 32;
diff --git a/arch/s390/pci/pci.c b/arch/s390/pci/pci.c
index 9ddc51eeb8d6..30de42730b2f 100644
--- a/arch/s390/pci/pci.c
+++ b/arch/s390/pci/pci.c
@@ -48,13 +48,10 @@
 static LIST_HEAD(zpci_list);
 static DEFINE_SPINLOCK(zpci_list_lock);
 
-static void zpci_enable_irq(struct irq_data *data);
-static void zpci_disable_irq(struct irq_data *data);
-
 static struct irq_chip zpci_irq_chip = {
 	.name = "zPCI",
-	.irq_unmask = zpci_enable_irq,
-	.irq_mask = zpci_disable_irq,
+	.irq_unmask = unmask_msi_irq,
+	.irq_mask = mask_msi_irq,
 };
 
 static DECLARE_BITMAP(zpci_domain, ZPCI_NR_DEVICES);
@@ -244,43 +241,6 @@ static int zpci_cfg_store(struct zpci_dev *zdev, int offset, u32 val, u8 len)
 	return rc;
 }
 
-static int zpci_msi_set_mask_bits(struct msi_desc *msi, u32 mask, u32 flag)
-{
-	int offset, pos;
-	u32 mask_bits;
-
-	if (msi->msi_attrib.is_msix) {
-		offset = msi->msi_attrib.entry_nr * PCI_MSIX_ENTRY_SIZE +
-			PCI_MSIX_ENTRY_VECTOR_CTRL;
-		msi->masked = readl(msi->mask_base + offset);
-		writel(flag, msi->mask_base + offset);
-	} else if (msi->msi_attrib.maskbit) {
-		pos = (long) msi->mask_base;
-		pci_read_config_dword(msi->dev, pos, &mask_bits);
-		mask_bits &= ~(mask);
-		mask_bits |= flag & mask;
-		pci_write_config_dword(msi->dev, pos, mask_bits);
-	} else
-		return 0;
-
-	msi->msi_attrib.maskbit = !!flag;
-	return 1;
-}
-
-static void zpci_enable_irq(struct irq_data *data)
-{
-	struct msi_desc *msi = irq_get_msi_desc(data->irq);
-
-	zpci_msi_set_mask_bits(msi, 1, 0);
-}
-
-static void zpci_disable_irq(struct irq_data *data)
-{
-	struct msi_desc *msi = irq_get_msi_desc(data->irq);
-
-	zpci_msi_set_mask_bits(msi, 1, 1);
-}
-
 void pcibios_fixup_bus(struct pci_bus *bus)
 {
 }
@@ -487,7 +447,10 @@ void arch_teardown_msi_irqs(struct pci_dev *pdev)
 
 	/* Release MSI interrupts */
 	list_for_each_entry(msi, &pdev->msi_list, list) {
-		zpci_msi_set_mask_bits(msi, 1, 1);
+		if (msi->msi_attrib.is_msix)
+			default_msix_mask_irq(msi, 1);
+		else
+			default_msi_mask_irq(msi, 1, 1);
 		irq_set_msi_desc(msi->irq, NULL);
 		irq_free_desc(msi->irq);
 		msi->msg.address_lo = 0;
diff --git a/arch/sh/Makefile b/arch/sh/Makefile
index d4d16e4be07c..bf5b3f5f4962 100644
--- a/arch/sh/Makefile
+++ b/arch/sh/Makefile
@@ -32,7 +32,8 @@ endif
 
 cflags-$(CONFIG_CPU_SH2)		:= $(call cc-option,-m2,)
 cflags-$(CONFIG_CPU_SH2A)		+= $(call cc-option,-m2a,) \
-					   $(call cc-option,-m2a-nofpu,)
+					   $(call cc-option,-m2a-nofpu,) \
+					   $(call cc-option,-m4-nofpu,)
 cflags-$(CONFIG_CPU_SH3)		:= $(call cc-option,-m3,)
 cflags-$(CONFIG_CPU_SH4)		:= $(call cc-option,-m4,) \
 	$(call cc-option,-mno-implicit-fp,-m4-nofpu)
diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig
index 29f2e988c56a..407c87d9879a 100644
--- a/arch/sparc/Kconfig
+++ b/arch/sparc/Kconfig
@@ -78,6 +78,7 @@ config SPARC64
 	select HAVE_C_RECORDMCOUNT
 	select NO_BOOTMEM
 	select HAVE_ARCH_AUDITSYSCALL
+	select ARCH_SUPPORTS_ATOMIC_RMW
 
 config ARCH_DEFCONFIG
 	string
diff --git a/arch/sparc/include/asm/irq_64.h b/arch/sparc/include/asm/irq_64.h
index 375cffcf7dbd..91d219381306 100644
--- a/arch/sparc/include/asm/irq_64.h
+++ b/arch/sparc/include/asm/irq_64.h
@@ -89,7 +89,7 @@ static inline unsigned long get_softint(void)
 	return retval;
 }
 
-void arch_trigger_all_cpu_backtrace(void);
+void arch_trigger_all_cpu_backtrace(bool);
 #define arch_trigger_all_cpu_backtrace arch_trigger_all_cpu_backtrace
 
 extern void *hardirq_stack[NR_CPUS];
diff --git a/arch/sparc/include/uapi/asm/unistd.h b/arch/sparc/include/uapi/asm/unistd.h
index b73274fb961a..42f2bca1d338 100644
--- a/arch/sparc/include/uapi/asm/unistd.h
+++ b/arch/sparc/include/uapi/asm/unistd.h
@@ -410,8 +410,9 @@
 #define __NR_finit_module	342
 #define __NR_sched_setattr	343
 #define __NR_sched_getattr	344
+#define __NR_renameat2		345
 
-#define NR_syscalls		345
+#define NR_syscalls		346
 
 /* Bitmask values returned from kern_features system call.  */
 #define KERN_FEATURE_MIXED_MODE_STACK	0x00000001
diff --git a/arch/sparc/kernel/process_64.c b/arch/sparc/kernel/process_64.c
index b2988f25e230..027e09986194 100644
--- a/arch/sparc/kernel/process_64.c
+++ b/arch/sparc/kernel/process_64.c
@@ -239,7 +239,7 @@ static void __global_reg_poll(struct global_reg_snapshot *gp)
 	}
 }
 
-void arch_trigger_all_cpu_backtrace(void)
+void arch_trigger_all_cpu_backtrace(bool include_self)
 {
 	struct thread_info *tp = current_thread_info();
 	struct pt_regs *regs = get_irq_regs();
@@ -251,16 +251,22 @@ void arch_trigger_all_cpu_backtrace(void)
 
 	spin_lock_irqsave(&global_cpu_snapshot_lock, flags);
 
-	memset(global_cpu_snapshot, 0, sizeof(global_cpu_snapshot));
-
 	this_cpu = raw_smp_processor_id();
 
-	__global_reg_self(tp, regs, this_cpu);
+	memset(global_cpu_snapshot, 0, sizeof(global_cpu_snapshot));
+
+	if (include_self)
+		__global_reg_self(tp, regs, this_cpu);
 
 	smp_fetch_global_regs();
 
 	for_each_online_cpu(cpu) {
-		struct global_reg_snapshot *gp = &global_cpu_snapshot[cpu].reg;
+		struct global_reg_snapshot *gp;
+
+		if (!include_self && cpu == this_cpu)
+			continue;
+
+		gp = &global_cpu_snapshot[cpu].reg;
 
 		__global_reg_poll(gp);
 
@@ -292,7 +298,7 @@ void arch_trigger_all_cpu_backtrace(void)
 
 static void sysrq_handle_globreg(int key)
 {
-	arch_trigger_all_cpu_backtrace();
+	arch_trigger_all_cpu_backtrace(true);
 }
 
 static struct sysrq_key_op sparc_globalreg_op = {
diff --git a/arch/sparc/kernel/sys32.S b/arch/sparc/kernel/sys32.S
index d066eb18650c..f834224208ed 100644
--- a/arch/sparc/kernel/sys32.S
+++ b/arch/sparc/kernel/sys32.S
@@ -48,6 +48,7 @@ SIGN1(sys32_futex, compat_sys_futex, %o1)
 SIGN1(sys32_recvfrom, compat_sys_recvfrom, %o0)
 SIGN1(sys32_recvmsg, compat_sys_recvmsg, %o0)
 SIGN1(sys32_sendmsg, compat_sys_sendmsg, %o0)
+SIGN2(sys32_renameat2, sys_renameat2, %o0, %o2)
 
 	.globl		sys32_mmap2
 sys32_mmap2:
diff --git a/arch/sparc/kernel/systbls_32.S b/arch/sparc/kernel/systbls_32.S
index 151ace8766cc..85fe9b1087cd 100644
--- a/arch/sparc/kernel/systbls_32.S
+++ b/arch/sparc/kernel/systbls_32.S
@@ -86,3 +86,4 @@ sys_call_table:
 /*330*/	.long sys_fanotify_mark, sys_prlimit64, sys_name_to_handle_at, sys_open_by_handle_at, sys_clock_adjtime
 /*335*/	.long sys_syncfs, sys_sendmmsg, sys_setns, sys_process_vm_readv, sys_process_vm_writev
 /*340*/	.long sys_ni_syscall, sys_kcmp, sys_finit_module, sys_sched_setattr, sys_sched_getattr
+/*345*/	.long sys_renameat2
diff --git a/arch/sparc/kernel/systbls_64.S b/arch/sparc/kernel/systbls_64.S
index 4bd4e2bb26cf..33ecba2826ea 100644
--- a/arch/sparc/kernel/systbls_64.S
+++ b/arch/sparc/kernel/systbls_64.S
@@ -87,6 +87,7 @@ sys_call_table32:
 /*330*/	.word compat_sys_fanotify_mark, sys_prlimit64, sys_name_to_handle_at, compat_sys_open_by_handle_at, compat_sys_clock_adjtime
 	.word sys_syncfs, compat_sys_sendmmsg, sys_setns, compat_sys_process_vm_readv, compat_sys_process_vm_writev
 /*340*/	.word sys_kern_features, sys_kcmp, sys_finit_module, sys_sched_setattr, sys_sched_getattr
+	.word sys32_renameat2
 
 #endif /* CONFIG_COMPAT */
 
@@ -165,3 +166,4 @@ sys_call_table:
 /*330*/	.word sys_fanotify_mark, sys_prlimit64, sys_name_to_handle_at, sys_open_by_handle_at, sys_clock_adjtime
 	.word sys_syncfs, sys_sendmmsg, sys_setns, sys_process_vm_readv, sys_process_vm_writev
 /*340*/	.word sys_kern_features, sys_kcmp, sys_finit_module, sys_sched_setattr, sys_sched_getattr
+	.word sys_renameat2
diff --git a/arch/um/kernel/tlb.c b/arch/um/kernel/tlb.c
index 9472079471bb..f1b3eb14b855 100644
--- a/arch/um/kernel/tlb.c
+++ b/arch/um/kernel/tlb.c
@@ -12,6 +12,7 @@
 #include <mem_user.h>
 #include <os.h>
 #include <skas.h>
+#include <kern_util.h>
 
 struct host_vm_change {
 	struct host_vm_op {
@@ -124,6 +125,9 @@ static int add_munmap(unsigned long addr, unsigned long len,
 	struct host_vm_op *last;
 	int ret = 0;
 
+	if ((addr >= STUB_START) && (addr < STUB_END))
+		return -EINVAL;
+
 	if (hvc->index != 0) {
 		last = &hvc->ops[hvc->index - 1];
 		if ((last->type == MUNMAP) &&
@@ -283,8 +287,11 @@ void fix_range_common(struct mm_struct *mm, unsigned long start_addr,
 	/* This is not an else because ret is modified above */
 	if (ret) {
 		printk(KERN_ERR "fix_range_common: failed, killing current "
-		       "process\n");
+		       "process: %d\n", task_tgid_vnr(current));
+		/* We are under mmap_sem, release it such that current can terminate */
+		up_write(&current->mm->mmap_sem);
 		force_sig(SIGKILL, current);
+		do_signal();
 	}
 }
 
diff --git a/arch/um/kernel/trap.c b/arch/um/kernel/trap.c
index 974b87474a99..5678c3571e7c 100644
--- a/arch/um/kernel/trap.c
+++ b/arch/um/kernel/trap.c
@@ -206,7 +206,7 @@ unsigned long segv(struct faultinfo fi, unsigned long ip, int is_user,
 	int is_write = FAULT_WRITE(fi);
 	unsigned long address = FAULT_ADDRESS(fi);
 
-	if (regs)
+	if (!is_user && regs)
 		current->thread.segv_regs = container_of(regs, struct pt_regs, regs);
 
 	if (!is_user && (address >= start_vm) && (address < end_vm)) {
diff --git a/arch/um/os-Linux/skas/process.c b/arch/um/os-Linux/skas/process.c
index d531879a4617..908579f2b0ab 100644
--- a/arch/um/os-Linux/skas/process.c
+++ b/arch/um/os-Linux/skas/process.c
@@ -54,7 +54,7 @@ static int ptrace_dump_regs(int pid)
 
 void wait_stub_done(int pid)
 {
-	int n, status, err, bad_stop = 0;
+	int n, status, err;
 
 	while (1) {
 		CATCH_EINTR(n = waitpid(pid, &status, WUNTRACED | __WALL));
@@ -74,8 +74,6 @@ void wait_stub_done(int pid)
 
 	if (((1 << WSTOPSIG(status)) & STUB_DONE_MASK) != 0)
 		return;
-	else
-		bad_stop = 1;
 
 bad_wait:
 	err = ptrace_dump_regs(pid);
@@ -85,10 +83,7 @@ bad_wait:
 	printk(UM_KERN_ERR "wait_stub_done : failed to wait for SIGTRAP, "
 	       "pid = %d, n = %d, errno = %d, status = 0x%x\n", pid, n, errno,
 	       status);
-	if (bad_stop)
-		kill(pid, SIGKILL);
-	else
-		fatal_sigsegv();
+	fatal_sigsegv();
 }
 
 extern unsigned long current_stub_stack(void);
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index a8f749ef0fdc..d24887b645dc 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -131,6 +131,7 @@ config X86
 	select HAVE_CC_STACKPROTECTOR
 	select GENERIC_CPU_AUTOPROBE
 	select HAVE_ARCH_AUDITSYSCALL
+	select ARCH_SUPPORTS_ATOMIC_RMW
 
 config INSTRUCTION_DECODER
 	def_bool y
diff --git a/arch/x86/boot/header.S b/arch/x86/boot/header.S
index 84c223479e3c..7a6d43a554d7 100644
--- a/arch/x86/boot/header.S
+++ b/arch/x86/boot/header.S
@@ -91,10 +91,9 @@ bs_die:
 
 	.section ".bsdata", "a"
 bugger_off_msg:
-	.ascii	"Direct floppy boot is not supported. "
-	.ascii	"Use a boot loader program instead.\r\n"
+	.ascii	"Use a boot loader.\r\n"
 	.ascii	"\n"
-	.ascii	"Remove disk and press any key to reboot ...\r\n"
+	.ascii	"Remove disk and press any key to reboot...\r\n"
 	.byte	0
 
 #ifdef CONFIG_EFI_STUB
@@ -108,7 +107,7 @@ coff_header:
 #else
 	.word	0x8664				# x86-64
 #endif
-	.word	3				# nr_sections
+	.word	4				# nr_sections
 	.long	0 				# TimeDateStamp
 	.long	0				# PointerToSymbolTable
 	.long	1				# NumberOfSymbols
@@ -250,6 +249,25 @@ section_table:
 	.word	0				# NumberOfLineNumbers
 	.long	0x60500020			# Characteristics (section flags)
 
+	#
+	# The offset & size fields are filled in by build.c.
+	#
+	.ascii	".bss"
+	.byte	0
+	.byte	0
+	.byte	0
+	.byte	0
+	.long	0
+	.long	0x0
+	.long	0				# Size of initialized data
+						# on disk
+	.long	0x0
+	.long	0				# PointerToRelocations
+	.long	0				# PointerToLineNumbers
+	.word	0				# NumberOfRelocations
+	.word	0				# NumberOfLineNumbers
+	.long	0xc8000080			# Characteristics (section flags)
+
 #endif /* CONFIG_EFI_STUB */
 
 	# Kernel attributes; used by setup.  This is part 1 of the
diff --git a/arch/x86/boot/tools/build.c b/arch/x86/boot/tools/build.c
index 1a2f2121cada..a7661c430cd9 100644
--- a/arch/x86/boot/tools/build.c
+++ b/arch/x86/boot/tools/build.c
@@ -143,7 +143,7 @@ static void usage(void)
 
 #ifdef CONFIG_EFI_STUB
 
-static void update_pecoff_section_header(char *section_name, u32 offset, u32 size)
+static void update_pecoff_section_header_fields(char *section_name, u32 vma, u32 size, u32 datasz, u32 offset)
 {
 	unsigned int pe_header;
 	unsigned short num_sections;
@@ -164,10 +164,10 @@ static void update_pecoff_section_header(char *section_name, u32 offset, u32 siz
 			put_unaligned_le32(size, section + 0x8);
 
 			/* section header vma field */
-			put_unaligned_le32(offset, section + 0xc);
+			put_unaligned_le32(vma, section + 0xc);
 
 			/* section header 'size of initialised data' field */
-			put_unaligned_le32(size, section + 0x10);
+			put_unaligned_le32(datasz, section + 0x10);
 
 			/* section header 'file offset' field */
 			put_unaligned_le32(offset, section + 0x14);
@@ -179,6 +179,11 @@ static void update_pecoff_section_header(char *section_name, u32 offset, u32 siz
 	}
 }
 
+static void update_pecoff_section_header(char *section_name, u32 offset, u32 size)
+{
+	update_pecoff_section_header_fields(section_name, offset, size, size, offset);
+}
+
 static void update_pecoff_setup_and_reloc(unsigned int size)
 {
 	u32 setup_offset = 0x200;
@@ -203,9 +208,6 @@ static void update_pecoff_text(unsigned int text_start, unsigned int file_sz)
 
 	pe_header = get_unaligned_le32(&buf[0x3c]);
 
-	/* Size of image */
-	put_unaligned_le32(file_sz, &buf[pe_header + 0x50]);
-
 	/*
 	 * Size of code: Subtract the size of the first sector (512 bytes)
 	 * which includes the header.
@@ -220,6 +222,22 @@ static void update_pecoff_text(unsigned int text_start, unsigned int file_sz)
 	update_pecoff_section_header(".text", text_start, text_sz);
 }
 
+static void update_pecoff_bss(unsigned int file_sz, unsigned int init_sz)
+{
+	unsigned int pe_header;
+	unsigned int bss_sz = init_sz - file_sz;
+
+	pe_header = get_unaligned_le32(&buf[0x3c]);
+
+	/* Size of uninitialized data */
+	put_unaligned_le32(bss_sz, &buf[pe_header + 0x24]);
+
+	/* Size of image */
+	put_unaligned_le32(init_sz, &buf[pe_header + 0x50]);
+
+	update_pecoff_section_header_fields(".bss", file_sz, bss_sz, 0, 0);
+}
+
 static int reserve_pecoff_reloc_section(int c)
 {
 	/* Reserve 0x20 bytes for .reloc section */
@@ -259,6 +277,8 @@ static void efi_stub_entry_update(void)
 static inline void update_pecoff_setup_and_reloc(unsigned int size) {}
 static inline void update_pecoff_text(unsigned int text_start,
 				      unsigned int file_sz) {}
+static inline void update_pecoff_bss(unsigned int file_sz,
+				     unsigned int init_sz) {}
 static inline void efi_stub_defaults(void) {}
 static inline void efi_stub_entry_update(void) {}
 
@@ -310,7 +330,7 @@ static void parse_zoffset(char *fname)
 
 int main(int argc, char ** argv)
 {
-	unsigned int i, sz, setup_sectors;
+	unsigned int i, sz, setup_sectors, init_sz;
 	int c;
 	u32 sys_size;
 	struct stat sb;
@@ -376,7 +396,9 @@ int main(int argc, char ** argv)
 	buf[0x1f1] = setup_sectors-1;
 	put_unaligned_le32(sys_size, &buf[0x1f4]);
 
-	update_pecoff_text(setup_sectors * 512, sz + i + ((sys_size * 16) - sz));
+	update_pecoff_text(setup_sectors * 512, i + (sys_size * 16));
+	init_sz = get_unaligned_le32(&buf[0x260]);
+	update_pecoff_bss(i + (sys_size * 16), init_sz);
 
 	efi_stub_entry_update();
 
diff --git a/arch/x86/crypto/sha512_ssse3_glue.c b/arch/x86/crypto/sha512_ssse3_glue.c
index f30cd10293f0..8626b03e83b7 100644
--- a/arch/x86/crypto/sha512_ssse3_glue.c
+++ b/arch/x86/crypto/sha512_ssse3_glue.c
@@ -141,7 +141,7 @@ static int sha512_ssse3_final(struct shash_desc *desc, u8 *out)
 
 	/* save number of bits */
 	bits[1] = cpu_to_be64(sctx->count[0] << 3);
-	bits[0] = cpu_to_be64(sctx->count[1] << 3) | sctx->count[0] >> 61;
+	bits[0] = cpu_to_be64(sctx->count[1] << 3 | sctx->count[0] >> 61);
 
 	/* Pad out to 112 mod 128 and append length */
 	index = sctx->count[0] & 0x7f;
diff --git a/arch/x86/include/asm/irq.h b/arch/x86/include/asm/irq.h
index cb6cfcd034cf..a80cbb88ea91 100644
--- a/arch/x86/include/asm/irq.h
+++ b/arch/x86/include/asm/irq.h
@@ -43,7 +43,7 @@ extern int vector_used_by_percpu_irq(unsigned int vector);
 extern void init_ISA_irqs(void);
 
 #ifdef CONFIG_X86_LOCAL_APIC
-void arch_trigger_all_cpu_backtrace(void);
+void arch_trigger_all_cpu_backtrace(bool);
 #define arch_trigger_all_cpu_backtrace arch_trigger_all_cpu_backtrace
 #endif
 
diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index 49314155b66c..49205d01b9ad 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -95,7 +95,7 @@ static inline gfn_t gfn_to_index(gfn_t gfn, gfn_t base_gfn, int level)
 #define KVM_REFILL_PAGES 25
 #define KVM_MAX_CPUID_ENTRIES 80
 #define KVM_NR_FIXED_MTRR_REGION 88
-#define KVM_NR_VAR_MTRR 8
+#define KVM_NR_VAR_MTRR 10
 
 #define ASYNC_PF_PER_VCPU 64
 
@@ -461,7 +461,7 @@ struct kvm_vcpu_arch {
 	bool nmi_injected;    /* Trying to inject an NMI this entry */
 
 	struct mtrr_state_type mtrr_state;
-	u32 pat;
+	u64 pat;
 
 	unsigned switch_db_regs;
 	unsigned long db[KVM_NR_DB_REGS];
diff --git a/arch/x86/include/asm/ptrace.h b/arch/x86/include/asm/ptrace.h
index 14fd6fd75a19..6205f0c434db 100644
--- a/arch/x86/include/asm/ptrace.h
+++ b/arch/x86/include/asm/ptrace.h
@@ -231,6 +231,22 @@ static inline unsigned long regs_get_kernel_stack_nth(struct pt_regs *regs,
 
 #define ARCH_HAS_USER_SINGLE_STEP_INFO
 
+/*
+ * When hitting ptrace_stop(), we cannot return using SYSRET because
+ * that does not restore the full CPU state, only a minimal set.  The
+ * ptracer can change arbitrary register values, which is usually okay
+ * because the usual ptrace stops run off the signal delivery path which
+ * forces IRET; however, ptrace_event() stops happen in arbitrary places
+ * in the kernel and don't force IRET path.
+ *
+ * So force IRET path after a ptrace stop.
+ */
+#define arch_ptrace_stop_needed(code, info)				\
+({									\
+	set_thread_flag(TIF_NOTIFY_RESUME);				\
+	false;								\
+})
+
 struct user_desc;
 extern int do_get_thread_area(struct task_struct *p, int idx,
 			      struct user_desc __user *info);
diff --git a/arch/x86/kernel/apic/hw_nmi.c b/arch/x86/kernel/apic/hw_nmi.c
index c3fcb5de5083..6a1e71bde323 100644
--- a/arch/x86/kernel/apic/hw_nmi.c
+++ b/arch/x86/kernel/apic/hw_nmi.c
@@ -33,31 +33,41 @@ static DECLARE_BITMAP(backtrace_mask, NR_CPUS) __read_mostly;
 /* "in progress" flag of arch_trigger_all_cpu_backtrace */
 static unsigned long backtrace_flag;
 
-void arch_trigger_all_cpu_backtrace(void)
+void arch_trigger_all_cpu_backtrace(bool include_self)
 {
 	int i;
+	int cpu = get_cpu();
 
-	if (test_and_set_bit(0, &backtrace_flag))
+	if (test_and_set_bit(0, &backtrace_flag)) {
 		/*
 		 * If there is already a trigger_all_cpu_backtrace() in progress
 		 * (backtrace_flag == 1), don't output double cpu dump infos.
 		 */
+		put_cpu();
 		return;
+	}
 
 	cpumask_copy(to_cpumask(backtrace_mask), cpu_online_mask);
+	if (!include_self)
+		cpumask_clear_cpu(cpu, to_cpumask(backtrace_mask));
 
-	printk(KERN_INFO "sending NMI to all CPUs:\n");
-	apic->send_IPI_all(NMI_VECTOR);
+	if (!cpumask_empty(to_cpumask(backtrace_mask))) {
+		pr_info("sending NMI to %s CPUs:\n",
+			(include_self ? "all" : "other"));
+		apic->send_IPI_mask(to_cpumask(backtrace_mask), NMI_VECTOR);
+	}
 
 	/* Wait for up to 10 seconds for all CPUs to do the backtrace */
 	for (i = 0; i < 10 * 1000; i++) {
 		if (cpumask_empty(to_cpumask(backtrace_mask)))
 			break;
 		mdelay(1);
+		touch_softlockup_watchdog();
 	}
 
 	clear_bit(0, &backtrace_flag);
 	smp_mb__after_atomic();
+	put_cpu();
 }
 
 static int
diff --git a/arch/x86/kernel/apm_32.c b/arch/x86/kernel/apm_32.c
index f3a1f04ed4cb..584874451414 100644
--- a/arch/x86/kernel/apm_32.c
+++ b/arch/x86/kernel/apm_32.c
@@ -841,7 +841,6 @@ static int apm_do_idle(void)
 	u32 eax;
 	u8 ret = 0;
 	int idled = 0;
-	int polling;
 	int err = 0;
 
 	if (!need_resched()) {
diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c
index a80029035bf2..f9e4fdd3b877 100644
--- a/arch/x86/kernel/cpu/intel.c
+++ b/arch/x86/kernel/cpu/intel.c
@@ -370,6 +370,17 @@ static void init_intel(struct cpuinfo_x86 *c)
 	 */
 	detect_extended_topology(c);
 
+	if (!cpu_has(c, X86_FEATURE_XTOPOLOGY)) {
+		/*
+		 * let's use the legacy cpuid vector 0x1 and 0x4 for topology
+		 * detection.
+		 */
+		c->x86_max_cores = intel_num_cpu_cores(c);
+#ifdef CONFIG_X86_32
+		detect_ht(c);
+#endif
+	}
+
 	l2 = init_intel_cacheinfo(c);
 	if (c->cpuid_level > 9) {
 		unsigned eax = cpuid_eax(10);
@@ -438,17 +449,6 @@ static void init_intel(struct cpuinfo_x86 *c)
 		set_cpu_cap(c, X86_FEATURE_P3);
 #endif
 
-	if (!cpu_has(c, X86_FEATURE_XTOPOLOGY)) {
-		/*
-		 * let's use the legacy cpuid vector 0x1 and 0x4 for topology
-		 * detection.
-		 */
-		c->x86_max_cores = intel_num_cpu_cores(c);
-#ifdef CONFIG_X86_32
-		detect_ht(c);
-#endif
-	}
-
 	/* Work around errata */
 	srat_detect_node(c);
 
diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c
index a952e9c85b6f..9c8f7394c612 100644
--- a/arch/x86/kernel/cpu/intel_cacheinfo.c
+++ b/arch/x86/kernel/cpu/intel_cacheinfo.c
@@ -730,6 +730,18 @@ unsigned int init_intel_cacheinfo(struct cpuinfo_x86 *c)
 #endif
 	}
 
+#ifdef CONFIG_X86_HT
+	/*
+	 * If cpu_llc_id is not yet set, this means cpuid_level < 4 which in
+	 * turns means that the only possibility is SMT (as indicated in
+	 * cpuid1). Since cpuid2 doesn't specify shared caches, and we know
+	 * that SMT shares all caches, we can unconditionally set cpu_llc_id to
+	 * c->phys_proc_id.
+	 */
+	if (per_cpu(cpu_llc_id, cpu) == BAD_APICID)
+		per_cpu(cpu_llc_id, cpu) = c->phys_proc_id;
+#endif
+
 	c->x86_cache_size = l3 ? l3 : (l2 ? l2 : (l1i+l1d));
 
 	return l2;
diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c
index bb92f38153b2..9a79c8dbd8e8 100644
--- a/arch/x86/kernel/cpu/mcheck/mce.c
+++ b/arch/x86/kernel/cpu/mcheck/mce.c
@@ -2451,6 +2451,12 @@ static __init int mcheck_init_device(void)
 	for_each_online_cpu(i) {
 		err = mce_device_create(i);
 		if (err) {
+			/*
+			 * Register notifier anyway (and do not unreg it) so
+			 * that we don't leave undeleted timers, see notifier
+			 * callback above.
+			 */
+			__register_hotcpu_notifier(&mce_cpu_notifier);
 			cpu_notifier_register_done();
 			goto err_device_create;
 		}
@@ -2471,10 +2477,6 @@ static __init int mcheck_init_device(void)
 err_register:
 	unregister_syscore_ops(&mce_syscore_ops);
 
-	cpu_notifier_register_begin();
-	__unregister_hotcpu_notifier(&mce_cpu_notifier);
-	cpu_notifier_register_done();
-
 err_device_create:
 	/*
 	 * We didn't keep track of which devices were created above, but
diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c
index 2bdfbff8a4f6..2879ecdaac43 100644
--- a/arch/x86/kernel/cpu/perf_event.c
+++ b/arch/x86/kernel/cpu/perf_event.c
@@ -118,6 +118,9 @@ static int x86_pmu_extra_regs(u64 config, struct perf_event *event)
 			continue;
 		if (event->attr.config1 & ~er->valid_mask)
 			return -EINVAL;
+		/* Check if the extra msrs can be safely accessed*/
+		if (!er->extra_msr_access)
+			return -ENXIO;
 
 		reg->idx = er->idx;
 		reg->config = event->attr.config1;
diff --git a/arch/x86/kernel/cpu/perf_event.h b/arch/x86/kernel/cpu/perf_event.h
index 3b2f9bdd974b..8ade93111e03 100644
--- a/arch/x86/kernel/cpu/perf_event.h
+++ b/arch/x86/kernel/cpu/perf_event.h
@@ -295,14 +295,16 @@ struct extra_reg {
 	u64			config_mask;
 	u64			valid_mask;
 	int			idx;  /* per_xxx->regs[] reg index */
+	bool			extra_msr_access;
 };
 
 #define EVENT_EXTRA_REG(e, ms, m, vm, i) {	\
-	.event = (e),		\
-	.msr = (ms),		\
-	.config_mask = (m),	\
-	.valid_mask = (vm),	\
-	.idx = EXTRA_REG_##i,	\
+	.event = (e),			\
+	.msr = (ms),			\
+	.config_mask = (m),		\
+	.valid_mask = (vm),		\
+	.idx = EXTRA_REG_##i,		\
+	.extra_msr_access = true,	\
 	}
 
 #define INTEL_EVENT_EXTRA_REG(event, msr, vm, idx)	\
diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c
index adb02aa62af5..2502d0d9d246 100644
--- a/arch/x86/kernel/cpu/perf_event_intel.c
+++ b/arch/x86/kernel/cpu/perf_event_intel.c
@@ -1382,6 +1382,15 @@ again:
 	intel_pmu_lbr_read();
 
 	/*
+	 * CondChgd bit 63 doesn't mean any overflow status. Ignore
+	 * and clear the bit.
+	 */
+	if (__test_and_clear_bit(63, (unsigned long *)&status)) {
+		if (!status)
+			goto done;
+	}
+
+	/*
 	 * PEBS overflow sets bit 62 in the global status register
 	 */
 	if (__test_and_clear_bit(62, (unsigned long *)&status)) {
@@ -2173,6 +2182,41 @@ static void intel_snb_check_microcode(void)
 	}
 }
 
+/*
+ * Under certain circumstances, access certain MSR may cause #GP.
+ * The function tests if the input MSR can be safely accessed.
+ */
+static bool check_msr(unsigned long msr, u64 mask)
+{
+	u64 val_old, val_new, val_tmp;
+
+	/*
+	 * Read the current value, change it and read it back to see if it
+	 * matches, this is needed to detect certain hardware emulators
+	 * (qemu/kvm) that don't trap on the MSR access and always return 0s.
+	 */
+	if (rdmsrl_safe(msr, &val_old))
+		return false;
+
+	/*
+	 * Only change the bits which can be updated by wrmsrl.
+	 */
+	val_tmp = val_old ^ mask;
+	if (wrmsrl_safe(msr, val_tmp) ||
+	    rdmsrl_safe(msr, &val_new))
+		return false;
+
+	if (val_new != val_tmp)
+		return false;
+
+	/* Here it's sure that the MSR can be safely accessed.
+	 * Restore the old value and return.
+	 */
+	wrmsrl(msr, val_old);
+
+	return true;
+}
+
 static __init void intel_sandybridge_quirk(void)
 {
 	x86_pmu.check_microcode = intel_snb_check_microcode;
@@ -2262,7 +2306,8 @@ __init int intel_pmu_init(void)
 	union cpuid10_ebx ebx;
 	struct event_constraint *c;
 	unsigned int unused;
-	int version;
+	struct extra_reg *er;
+	int version, i;
 
 	if (!cpu_has(&boot_cpu_data, X86_FEATURE_ARCH_PERFMON)) {
 		switch (boot_cpu_data.x86) {
@@ -2465,6 +2510,9 @@ __init int intel_pmu_init(void)
 	case 62: /* IvyBridge EP */
 		memcpy(hw_cache_event_ids, snb_hw_cache_event_ids,
 		       sizeof(hw_cache_event_ids));
+		/* dTLB-load-misses on IVB is different than SNB */
+		hw_cache_event_ids[C(DTLB)][C(OP_READ)][C(RESULT_MISS)] = 0x8108; /* DTLB_LOAD_MISSES.DEMAND_LD_MISS_CAUSES_A_WALK */
+
 		memcpy(hw_cache_extra_regs, snb_hw_cache_extra_regs,
 		       sizeof(hw_cache_extra_regs));
 
@@ -2565,6 +2613,34 @@ __init int intel_pmu_init(void)
 		}
 	}
 
+	/*
+	 * Access LBR MSR may cause #GP under certain circumstances.
+	 * E.g. KVM doesn't support LBR MSR
+	 * Check all LBT MSR here.
+	 * Disable LBR access if any LBR MSRs can not be accessed.
+	 */
+	if (x86_pmu.lbr_nr && !check_msr(x86_pmu.lbr_tos, 0x3UL))
+		x86_pmu.lbr_nr = 0;
+	for (i = 0; i < x86_pmu.lbr_nr; i++) {
+		if (!(check_msr(x86_pmu.lbr_from + i, 0xffffUL) &&
+		      check_msr(x86_pmu.lbr_to + i, 0xffffUL)))
+			x86_pmu.lbr_nr = 0;
+	}
+
+	/*
+	 * Access extra MSR may cause #GP under certain circumstances.
+	 * E.g. KVM doesn't support offcore event
+	 * Check all extra_regs here.
+	 */
+	if (x86_pmu.extra_regs) {
+		for (er = x86_pmu.extra_regs; er->msr; er++) {
+			er->extra_msr_access = check_msr(er->msr, 0x1ffUL);
+			/* Disable LBR select mapping */
+			if ((er->idx == EXTRA_REG_LBR) && !er->extra_msr_access)
+				x86_pmu.lbr_sel_map = NULL;
+		}
+	}
+
 	/* Support full width counters using alternative MSR range */
 	if (x86_pmu.intel_cap.full_width_write) {
 		x86_pmu.max_period = x86_pmu.cntval_mask;
diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c
index 980970cb744d..696ade311ded 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_ds.c
+++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c
@@ -311,9 +311,11 @@ static int alloc_bts_buffer(int cpu)
 	if (!x86_pmu.bts)
 		return 0;
 
-	buffer = kzalloc_node(BTS_BUFFER_SIZE, GFP_KERNEL, node);
-	if (unlikely(!buffer))
+	buffer = kzalloc_node(BTS_BUFFER_SIZE, GFP_KERNEL | __GFP_NOWARN, node);
+	if (unlikely(!buffer)) {
+		WARN_ONCE(1, "%s: BTS buffer allocation failure\n", __func__);
 		return -ENOMEM;
+	}
 
 	max = BTS_BUFFER_SIZE / BTS_RECORD_SIZE;
 	thresh = max / 16;
diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore.c b/arch/x86/kernel/cpu/perf_event_intel_uncore.c
index e009f3cedf89..cfc6f9dfcd90 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_uncore.c
+++ b/arch/x86/kernel/cpu/perf_event_intel_uncore.c
@@ -550,16 +550,16 @@ static struct extra_reg snbep_uncore_cbox_extra_regs[] = {
 	SNBEP_CBO_EVENT_EXTRA_REG(0x4134, 0xffff, 0x6),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x0135, 0xffff, 0x8),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x0335, 0xffff, 0x8),
-	SNBEP_CBO_EVENT_EXTRA_REG(0x4135, 0xffff, 0xc),
-	SNBEP_CBO_EVENT_EXTRA_REG(0x4335, 0xffff, 0xc),
+	SNBEP_CBO_EVENT_EXTRA_REG(0x4135, 0xffff, 0xa),
+	SNBEP_CBO_EVENT_EXTRA_REG(0x4335, 0xffff, 0xa),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x4435, 0xffff, 0x2),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x4835, 0xffff, 0x2),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x4a35, 0xffff, 0x2),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x5035, 0xffff, 0x2),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x0136, 0xffff, 0x8),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x0336, 0xffff, 0x8),
-	SNBEP_CBO_EVENT_EXTRA_REG(0x4136, 0xffff, 0xc),
-	SNBEP_CBO_EVENT_EXTRA_REG(0x4336, 0xffff, 0xc),
+	SNBEP_CBO_EVENT_EXTRA_REG(0x4136, 0xffff, 0xa),
+	SNBEP_CBO_EVENT_EXTRA_REG(0x4336, 0xffff, 0xa),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x4436, 0xffff, 0x2),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x4836, 0xffff, 0x2),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x4a36, 0xffff, 0x2),
@@ -1222,6 +1222,7 @@ static struct extra_reg ivt_uncore_cbox_extra_regs[] = {
 	SNBEP_CBO_EVENT_EXTRA_REG(SNBEP_CBO_PMON_CTL_TID_EN,
 				  SNBEP_CBO_PMON_CTL_TID_EN, 0x1),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x1031, 0x10ff, 0x2),
+
 	SNBEP_CBO_EVENT_EXTRA_REG(0x1134, 0xffff, 0x4),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x4134, 0xffff, 0xc),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x5134, 0xffff, 0xc),
@@ -1245,7 +1246,7 @@ static struct extra_reg ivt_uncore_cbox_extra_regs[] = {
 	SNBEP_CBO_EVENT_EXTRA_REG(0x8335, 0xffff, 0x10),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x0136, 0xffff, 0x10),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x0336, 0xffff, 0x10),
-	SNBEP_CBO_EVENT_EXTRA_REG(0x2336, 0xffff, 0x10),
+	SNBEP_CBO_EVENT_EXTRA_REG(0x2136, 0xffff, 0x10),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x2336, 0xffff, 0x10),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x4136, 0xffff, 0x18),
 	SNBEP_CBO_EVENT_EXTRA_REG(0x4336, 0xffff, 0x18),
diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S
index f0da82b8e634..0d0c9d4ab6d5 100644
--- a/arch/x86/kernel/entry_32.S
+++ b/arch/x86/kernel/entry_32.S
@@ -423,8 +423,9 @@ sysenter_past_esp:
 	jnz sysenter_audit
 sysenter_do_call:
 	cmpl $(NR_syscalls), %eax
-	jae syscall_badsys
+	jae sysenter_badsys
 	call *sys_call_table(,%eax,4)
+sysenter_after_call:
 	movl %eax,PT_EAX(%esp)
 	LOCKDEP_SYS_EXIT
 	DISABLE_INTERRUPTS(CLBR_ANY)
@@ -501,6 +502,7 @@ ENTRY(system_call)
 	jae syscall_badsys
 syscall_call:
 	call *sys_call_table(,%eax,4)
+syscall_after_call:
 	movl %eax,PT_EAX(%esp)		# store the return value
 syscall_exit:
 	LOCKDEP_SYS_EXIT
@@ -674,8 +676,13 @@ syscall_fault:
 END(syscall_fault)
 
 syscall_badsys:
-	movl $-ENOSYS,PT_EAX(%esp)
-	jmp resume_userspace
+	movl $-ENOSYS,%eax
+	jmp syscall_after_call
+END(syscall_badsys)
+
+sysenter_badsys:
+	movl $-ENOSYS,%eax
+	jmp sysenter_after_call
 END(syscall_badsys)
 	CFI_ENDPROC
 
diff --git a/arch/x86/kernel/espfix_64.c b/arch/x86/kernel/espfix_64.c
index 6afbb16e9b79..94d857fb1033 100644
--- a/arch/x86/kernel/espfix_64.c
+++ b/arch/x86/kernel/espfix_64.c
@@ -175,7 +175,7 @@ void init_espfix_ap(void)
 	if (!pud_present(pud)) {
 		pmd_p = (pmd_t *)__get_free_page(PGALLOC_GFP);
 		pud = __pud(__pa(pmd_p) | (PGTABLE_PROT & ptemask));
-		paravirt_alloc_pud(&init_mm, __pa(pmd_p) >> PAGE_SHIFT);
+		paravirt_alloc_pmd(&init_mm, __pa(pmd_p) >> PAGE_SHIFT);
 		for (n = 0; n < ESPFIX_PUD_CLONES; n++)
 			set_pud(&pud_p[n], pud);
 	}
@@ -185,7 +185,7 @@ void init_espfix_ap(void)
 	if (!pmd_present(pmd)) {
 		pte_p = (pte_t *)__get_free_page(PGALLOC_GFP);
 		pmd = __pmd(__pa(pte_p) | (PGTABLE_PROT & ptemask));
-		paravirt_alloc_pmd(&init_mm, __pa(pte_p) >> PAGE_SHIFT);
+		paravirt_alloc_pte(&init_mm, __pa(pte_p) >> PAGE_SHIFT);
 		for (n = 0; n < ESPFIX_PMD_CLONES; n++)
 			set_pmd(&pmd_p[n], pmd);
 	}
@@ -193,7 +193,6 @@ void init_espfix_ap(void)
 	pte_p = pte_offset_kernel(&pmd, addr);
 	stack_page = (void *)__get_free_page(GFP_KERNEL);
 	pte = __pte(__pa(stack_page) | (__PAGE_KERNEL_RO & ptemask));
-	paravirt_alloc_pte(&init_mm, __pa(stack_page) >> PAGE_SHIFT);
 	for (n = 0; n < ESPFIX_PTE_CLONES; n++)
 		set_pte(&pte_p[n*PTE_STRIDE], pte);
 
diff --git a/arch/x86/kernel/kprobes/core.c b/arch/x86/kernel/kprobes/core.c
index 7596df664901..67e6d19ef1be 100644
--- a/arch/x86/kernel/kprobes/core.c
+++ b/arch/x86/kernel/kprobes/core.c
@@ -574,6 +574,9 @@ int kprobe_int3_handler(struct pt_regs *regs)
 	struct kprobe *p;
 	struct kprobe_ctlblk *kcb;
 
+	if (user_mode_vm(regs))
+		return 0;
+
 	addr = (kprobe_opcode_t *)(regs->ip - sizeof(kprobe_opcode_t));
 	/*
 	 * We don't want to be preempted for the entire
diff --git a/arch/x86/kernel/signal.c b/arch/x86/kernel/signal.c
index a0da58db43a8..2851d63c1202 100644
--- a/arch/x86/kernel/signal.c
+++ b/arch/x86/kernel/signal.c
@@ -363,7 +363,7 @@ static int __setup_rt_frame(int sig, struct ksignal *ksig,
 
 		/* Set up to return from userspace.  */
 		restorer = current->mm->context.vdso +
-			selected_vdso32->sym___kernel_sigreturn;
+			selected_vdso32->sym___kernel_rt_sigreturn;
 		if (ksig->ka.sa.sa_flags & SA_RESTORER)
 			restorer = ksig->ka.sa.sa_restorer;
 		put_user_ex(restorer, &frame->pretcode);
diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c
index 57e5ce126d5a..ea030319b321 100644
--- a/arch/x86/kernel/tsc.c
+++ b/arch/x86/kernel/tsc.c
@@ -920,9 +920,9 @@ static int time_cpufreq_notifier(struct notifier_block *nb, unsigned long val,
 		tsc_khz = cpufreq_scale(tsc_khz_ref, ref_freq, freq->new);
 		if (!(freq->flags & CPUFREQ_CONST_LOOPS))
 			mark_tsc_unstable("cpufreq changes");
-	}
 
-	set_cyc2ns_scale(tsc_khz, freq->cpu);
+		set_cyc2ns_scale(tsc_khz, freq->cpu);
+	}
 
 	return 0;
 }
diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
index ec8366c5cfea..b5e994ad0135 100644
--- a/arch/x86/kvm/svm.c
+++ b/arch/x86/kvm/svm.c
@@ -1462,6 +1462,7 @@ static void svm_get_segment(struct kvm_vcpu *vcpu,
 		 */
 		if (var->unusable)
 			var->db = 0;
+		var->dpl = to_svm(vcpu)->vmcb->save.cpl;
 		break;
 	}
 }
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index f32a02578c0d..ef432f891d30 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -1898,7 +1898,7 @@ static int set_msr_hyperv_pw(struct kvm_vcpu *vcpu, u32 msr, u64 data)
 		if (!(data & HV_X64_MSR_TSC_REFERENCE_ENABLE))
 			break;
 		gfn = data >> HV_X64_MSR_TSC_REFERENCE_ADDRESS_SHIFT;
-		if (kvm_write_guest(kvm, data,
+		if (kvm_write_guest(kvm, gfn << HV_X64_MSR_TSC_REFERENCE_ADDRESS_SHIFT,
 			&tsc_ref, sizeof(tsc_ref)))
 			return 1;
 		mark_page_dirty(kvm, gfn);
@@ -5887,6 +5887,18 @@ static int inject_pending_event(struct kvm_vcpu *vcpu, bool req_int_win)
 			kvm_x86_ops->set_nmi(vcpu);
 		}
 	} else if (kvm_cpu_has_injectable_intr(vcpu)) {
+		/*
+		 * Because interrupts can be injected asynchronously, we are
+		 * calling check_nested_events again here to avoid a race condition.
+		 * See https://lkml.org/lkml/2014/7/2/60 for discussion about this
+		 * proposal and current concerns.  Perhaps we should be setting
+		 * KVM_REQ_EVENT only on certain events and not unconditionally?
+		 */
+		if (is_guest_mode(vcpu) && kvm_x86_ops->check_nested_events) {
+			r = kvm_x86_ops->check_nested_events(vcpu, req_int_win);
+			if (r != 0)
+				return r;
+		}
 		if (kvm_x86_ops->interrupt_allowed(vcpu)) {
 			kvm_queue_interrupt(vcpu, kvm_cpu_get_interrupt(vcpu),
 					    false);
diff --git a/arch/x86/vdso/Makefile b/arch/x86/vdso/Makefile
index 3c0809a0631f..61b04fe36e66 100644
--- a/arch/x86/vdso/Makefile
+++ b/arch/x86/vdso/Makefile
@@ -11,7 +11,6 @@ VDSO32-$(CONFIG_COMPAT)		:= y
 
 # files to link into the vdso
 vobjs-y := vdso-note.o vclock_gettime.o vgetcpu.o vdso-fakesections.o
-vobjs-nox32 := vdso-fakesections.o
 
 # files to link into kernel
 obj-y				+= vma.o
@@ -67,7 +66,8 @@ $(obj)/vdso-image-%.c: $(obj)/vdso%.so.dbg $(obj)/vdso2c FORCE
 #
 CFL := $(PROFILING) -mcmodel=small -fPIC -O2 -fasynchronous-unwind-tables -m64 \
        $(filter -g%,$(KBUILD_CFLAGS)) $(call cc-option, -fno-stack-protector) \
-       -fno-omit-frame-pointer -foptimize-sibling-calls
+       -fno-omit-frame-pointer -foptimize-sibling-calls \
+       -DDISABLE_BRANCH_PROFILING
 
 $(vobjs): KBUILD_CFLAGS += $(CFL)
 
@@ -134,7 +134,7 @@ override obj-dirs = $(dir $(obj)) $(obj)/vdso32/
 
 targets += vdso32/vdso32.lds
 targets += vdso32/note.o vdso32/vclock_gettime.o $(vdso32.so-y:%=vdso32/%.o)
-targets += vdso32/vclock_gettime.o
+targets += vdso32/vclock_gettime.o vdso32/vdso-fakesections.o
 
 $(obj)/vdso32.o: $(vdso32-images:%=$(obj)/%)
 
@@ -150,11 +150,13 @@ KBUILD_CFLAGS_32 += -m32 -msoft-float -mregparm=0 -fpic
 KBUILD_CFLAGS_32 += $(call cc-option, -fno-stack-protector)
 KBUILD_CFLAGS_32 += $(call cc-option, -foptimize-sibling-calls)
 KBUILD_CFLAGS_32 += -fno-omit-frame-pointer
+KBUILD_CFLAGS_32 += -DDISABLE_BRANCH_PROFILING
 $(vdso32-images:%=$(obj)/%.dbg): KBUILD_CFLAGS = $(KBUILD_CFLAGS_32)
 
 $(vdso32-images:%=$(obj)/%.dbg): $(obj)/vdso32-%.so.dbg: FORCE \
 				 $(obj)/vdso32/vdso32.lds \
 				 $(obj)/vdso32/vclock_gettime.o \
+				 $(obj)/vdso32/vdso-fakesections.o \
 				 $(obj)/vdso32/note.o \
 				 $(obj)/vdso32/%.o
 	$(call if_changed,vdso)
@@ -169,14 +171,24 @@ quiet_cmd_vdso = VDSO    $@
 		 sh $(srctree)/$(src)/checkundef.sh '$(NM)' '$@'
 
 VDSO_LDFLAGS = -fPIC -shared $(call cc-ldoption, -Wl$(comma)--hash-style=sysv) \
-	-Wl,-Bsymbolic $(LTO_CFLAGS)
+	$(call cc-ldoption, -Wl$(comma)--build-id) -Wl,-Bsymbolic $(LTO_CFLAGS)
 GCOV_PROFILE := n
 
 #
-# Install the unstripped copies of vdso*.so.
+# Install the unstripped copies of vdso*.so.  If our toolchain supports
+# build-id, install .build-id links as well.
 #
 quiet_cmd_vdso_install = INSTALL $(@:install_%=%)
-      cmd_vdso_install = cp $< $(MODLIB)/vdso/$(@:install_%=%)
+define cmd_vdso_install
+	cp $< "$(MODLIB)/vdso/$(@:install_%=%)"; \
+	if readelf -n $< |grep -q 'Build ID'; then \
+	  buildid=`readelf -n $< |grep 'Build ID' |sed -e 's/^.*Build ID: \(.*\)$$/\1/'`; \
+	  first=`echo $$buildid | cut -b-2`; \
+	  last=`echo $$buildid | cut -b3-`; \
+	  mkdir -p "$(MODLIB)/vdso/.build-id/$$first"; \
+	  ln -sf "../../$(@:install_%=%)" "$(MODLIB)/vdso/.build-id/$$first/$$last.debug"; \
+	fi
+endef
 
 vdso_img_insttargets := $(vdso_img_sodbg:%.dbg=install_%)
 
diff --git a/arch/x86/vdso/vclock_gettime.c b/arch/x86/vdso/vclock_gettime.c
index b2e4f493e5b0..9793322751e0 100644
--- a/arch/x86/vdso/vclock_gettime.c
+++ b/arch/x86/vdso/vclock_gettime.c
@@ -11,9 +11,6 @@
  * Check with readelf after changing.
  */
 
-/* Disable profiling for userspace code: */
-#define DISABLE_BRANCH_PROFILING
-
 #include <uapi/linux/time.h>
 #include <asm/vgtod.h>
 #include <asm/hpet.h>
diff --git a/arch/x86/vdso/vdso-fakesections.c b/arch/x86/vdso/vdso-fakesections.c
index cb8a8d72c24b..aa5fbfab20a5 100644
--- a/arch/x86/vdso/vdso-fakesections.c
+++ b/arch/x86/vdso/vdso-fakesections.c
@@ -2,31 +2,20 @@
  * Copyright 2014 Andy Lutomirski
  * Subject to the GNU Public License, v.2
  *
- * Hack to keep broken Go programs working.
- *
- * The Go runtime had a couple of bugs: it would read the section table to try
- * to figure out how many dynamic symbols there were (it shouldn't have looked
- * at the section table at all) and, if there were no SHT_SYNDYM section table
- * entry, it would use an uninitialized value for the number of symbols.  As a
- * workaround, we supply a minimal section table.  vdso2c will adjust the
- * in-memory image so that "vdso_fake_sections" becomes the section table.
- *
- * The bug was introduced by:
- * https://code.google.com/p/go/source/detail?r=56ea40aac72b (2012-08-31)
- * and is being addressed in the Go runtime in this issue:
- * https://code.google.com/p/go/issues/detail?id=8197
+ * String table for loadable section headers.  See vdso2c.h for why
+ * this exists.
  */
 
-#ifndef __x86_64__
-#error This hack is specific to the 64-bit vDSO
-#endif
-
-#include <linux/elf.h>
-
-extern const __visible struct elf64_shdr vdso_fake_sections[];
-const __visible struct elf64_shdr vdso_fake_sections[] = {
-	{
-		.sh_type = SHT_DYNSYM,
-		.sh_entsize = sizeof(Elf64_Sym),
-	}
-};
+const char fake_shstrtab[] __attribute__((section(".fake_shstrtab"))) =
+	".hash\0"
+	".dynsym\0"
+	".dynstr\0"
+	".gnu.version\0"
+	".gnu.version_d\0"
+	".dynamic\0"
+	".rodata\0"
+	".fake_shstrtab\0"  /* Yay, self-referential code. */
+	".note\0"
+	".eh_frame_hdr\0"
+	".eh_frame\0"
+	".text";
diff --git a/arch/x86/vdso/vdso-layout.lds.S b/arch/x86/vdso/vdso-layout.lds.S
index 2ec72f651ebf..9197544eea9a 100644
--- a/arch/x86/vdso/vdso-layout.lds.S
+++ b/arch/x86/vdso/vdso-layout.lds.S
@@ -6,6 +6,16 @@
  * This script controls its layout.
  */
 
+#if defined(BUILD_VDSO64)
+# define SHDR_SIZE 64
+#elif defined(BUILD_VDSO32) || defined(BUILD_VDSOX32)
+# define SHDR_SIZE 40
+#else
+# error unknown VDSO target
+#endif
+
+#define NUM_FAKE_SHDRS 13
+
 SECTIONS
 {
 	. = SIZEOF_HEADERS;
@@ -18,36 +28,53 @@ SECTIONS
 	.gnu.version_d	: { *(.gnu.version_d) }
 	.gnu.version_r	: { *(.gnu.version_r) }
 
+	.dynamic	: { *(.dynamic) }		:text	:dynamic
+
+	.rodata		: {
+		*(.rodata*)
+		*(.data*)
+		*(.sdata*)
+		*(.got.plt) *(.got)
+		*(.gnu.linkonce.d.*)
+		*(.bss*)
+		*(.dynbss*)
+		*(.gnu.linkonce.b.*)
+
+		/*
+		 * Ideally this would live in a C file, but that won't
+		 * work cleanly for x32 until we start building the x32
+		 * C code using an x32 toolchain.
+		 */
+		VDSO_FAKE_SECTION_TABLE_START = .;
+		. = . + NUM_FAKE_SHDRS * SHDR_SIZE;
+		VDSO_FAKE_SECTION_TABLE_END = .;
+	}						:text
+
+	.fake_shstrtab	: { *(.fake_shstrtab) }		:text
+
+
 	.note		: { *(.note.*) }		:text	:note
 
 	.eh_frame_hdr	: { *(.eh_frame_hdr) }		:text	:eh_frame_hdr
 	.eh_frame	: { KEEP (*(.eh_frame)) }	:text
 
-	.dynamic	: { *(.dynamic) }		:text	:dynamic
-
-	.rodata		: { *(.rodata*) }		:text
-	.data		: {
-	      *(.data*)
-	      *(.sdata*)
-	      *(.got.plt) *(.got)
-	      *(.gnu.linkonce.d.*)
-	      *(.bss*)
-	      *(.dynbss*)
-	      *(.gnu.linkonce.b.*)
-	}
-
-	.altinstructions	: { *(.altinstructions) }
-	.altinstr_replacement	: { *(.altinstr_replacement) }
 
 	/*
-	 * Align the actual code well away from the non-instruction data.
-	 * This is the best thing for the I-cache.
+	 * Text is well-separated from actual data: there's plenty of
+	 * stuff that isn't used at runtime in between.
 	 */
-	. = ALIGN(0x100);
 
 	.text		: { *(.text*) }			:text	=0x90909090,
 
 	/*
+	 * At the end so that eu-elflint stays happy when vdso2c strips
+	 * these.  A better implementation would avoid allocating space
+	 * for these.
+	 */
+	.altinstructions	: { *(.altinstructions) }	:text
+	.altinstr_replacement	: { *(.altinstr_replacement) }	:text
+
+	/*
 	 * The remainder of the vDSO consists of special pages that are
 	 * shared between the kernel and userspace.  It needs to be at the
 	 * end so that it doesn't overlap the mapping of the actual
@@ -75,6 +102,7 @@ SECTIONS
 	/DISCARD/ : {
 		*(.discard)
 		*(.discard.*)
+		*(__bug_table)
 	}
 }
 
diff --git a/arch/x86/vdso/vdso.lds.S b/arch/x86/vdso/vdso.lds.S
index 75e3404c83b1..6807932643c2 100644
--- a/arch/x86/vdso/vdso.lds.S
+++ b/arch/x86/vdso/vdso.lds.S
@@ -6,6 +6,8 @@
  * the DSO.
  */
 
+#define BUILD_VDSO64
+
 #include "vdso-layout.lds.S"
 
 /*
diff --git a/arch/x86/vdso/vdso2c.c b/arch/x86/vdso/vdso2c.c
index 7a6bf50f9165..238dbe82776e 100644
--- a/arch/x86/vdso/vdso2c.c
+++ b/arch/x86/vdso/vdso2c.c
@@ -23,6 +23,8 @@ enum {
 	sym_vvar_page,
 	sym_hpet_page,
 	sym_end_mapping,
+	sym_VDSO_FAKE_SECTION_TABLE_START,
+	sym_VDSO_FAKE_SECTION_TABLE_END,
 };
 
 const int special_pages[] = {
@@ -30,15 +32,26 @@ const int special_pages[] = {
 	sym_hpet_page,
 };
 
-char const * const required_syms[] = {
-	[sym_vvar_page] = "vvar_page",
-	[sym_hpet_page] = "hpet_page",
-	[sym_end_mapping] = "end_mapping",
-	"VDSO32_NOTE_MASK",
-	"VDSO32_SYSENTER_RETURN",
-	"__kernel_vsyscall",
-	"__kernel_sigreturn",
-	"__kernel_rt_sigreturn",
+struct vdso_sym {
+	const char *name;
+	bool export;
+};
+
+struct vdso_sym required_syms[] = {
+	[sym_vvar_page] = {"vvar_page", true},
+	[sym_hpet_page] = {"hpet_page", true},
+	[sym_end_mapping] = {"end_mapping", true},
+	[sym_VDSO_FAKE_SECTION_TABLE_START] = {
+		"VDSO_FAKE_SECTION_TABLE_START", false
+	},
+	[sym_VDSO_FAKE_SECTION_TABLE_END] = {
+		"VDSO_FAKE_SECTION_TABLE_END", false
+	},
+	{"VDSO32_NOTE_MASK", true},
+	{"VDSO32_SYSENTER_RETURN", true},
+	{"__kernel_vsyscall", true},
+	{"__kernel_sigreturn", true},
+	{"__kernel_rt_sigreturn", true},
 };
 
 __attribute__((format(printf, 1, 2))) __attribute__((noreturn))
@@ -83,37 +96,21 @@ extern void bad_put_le(void);
 
 #define NSYMS (sizeof(required_syms) / sizeof(required_syms[0]))
 
-#define BITS 64
-#define GOFUNC go64
-#define Elf_Ehdr Elf64_Ehdr
-#define Elf_Shdr Elf64_Shdr
-#define Elf_Phdr Elf64_Phdr
-#define Elf_Sym Elf64_Sym
-#define Elf_Dyn Elf64_Dyn
+#define BITSFUNC3(name, bits) name##bits
+#define BITSFUNC2(name, bits) BITSFUNC3(name, bits)
+#define BITSFUNC(name) BITSFUNC2(name, ELF_BITS)
+
+#define ELF_BITS_XFORM2(bits, x) Elf##bits##_##x
+#define ELF_BITS_XFORM(bits, x) ELF_BITS_XFORM2(bits, x)
+#define ELF(x) ELF_BITS_XFORM(ELF_BITS, x)
+
+#define ELF_BITS 64
 #include "vdso2c.h"
-#undef BITS
-#undef GOFUNC
-#undef Elf_Ehdr
-#undef Elf_Shdr
-#undef Elf_Phdr
-#undef Elf_Sym
-#undef Elf_Dyn
-
-#define BITS 32
-#define GOFUNC go32
-#define Elf_Ehdr Elf32_Ehdr
-#define Elf_Shdr Elf32_Shdr
-#define Elf_Phdr Elf32_Phdr
-#define Elf_Sym Elf32_Sym
-#define Elf_Dyn Elf32_Dyn
+#undef ELF_BITS
+
+#define ELF_BITS 32
 #include "vdso2c.h"
-#undef BITS
-#undef GOFUNC
-#undef Elf_Ehdr
-#undef Elf_Shdr
-#undef Elf_Phdr
-#undef Elf_Sym
-#undef Elf_Dyn
+#undef ELF_BITS
 
 static void go(void *addr, size_t len, FILE *outfile, const char *name)
 {
diff --git a/arch/x86/vdso/vdso2c.h b/arch/x86/vdso/vdso2c.h
index c6eefaf389b9..11b65d4f9414 100644
--- a/arch/x86/vdso/vdso2c.h
+++ b/arch/x86/vdso/vdso2c.h
@@ -4,23 +4,139 @@
  * are built for 32-bit userspace.
  */
 
-static void GOFUNC(void *addr, size_t len, FILE *outfile, const char *name)
+/*
+ * We're writing a section table for a few reasons:
+ *
+ * The Go runtime had a couple of bugs: it would read the section
+ * table to try to figure out how many dynamic symbols there were (it
+ * shouldn't have looked at the section table at all) and, if there
+ * were no SHT_SYNDYM section table entry, it would use an
+ * uninitialized value for the number of symbols.  An empty DYNSYM
+ * table would work, but I see no reason not to write a valid one (and
+ * keep full performance for old Go programs).  This hack is only
+ * needed on x86_64.
+ *
+ * The bug was introduced on 2012-08-31 by:
+ * https://code.google.com/p/go/source/detail?r=56ea40aac72b
+ * and was fixed on 2014-06-13 by:
+ * https://code.google.com/p/go/source/detail?r=fc1cd5e12595
+ *
+ * Binutils has issues debugging the vDSO: it reads the section table to
+ * find SHT_NOTE; it won't look at PT_NOTE for the in-memory vDSO, which
+ * would break build-id if we removed the section table.  Binutils
+ * also requires that shstrndx != 0.  See:
+ * https://sourceware.org/bugzilla/show_bug.cgi?id=17064
+ *
+ * elfutils might not look for PT_NOTE if there is a section table at
+ * all.  I don't know whether this matters for any practical purpose.
+ *
+ * For simplicity, rather than hacking up a partial section table, we
+ * just write a mostly complete one.  We omit non-dynamic symbols,
+ * though, since they're rather large.
+ *
+ * Once binutils gets fixed, we might be able to drop this for all but
+ * the 64-bit vdso, since build-id only works in kernel RPMs, and
+ * systems that update to new enough kernel RPMs will likely update
+ * binutils in sync.  build-id has never worked for home-built kernel
+ * RPMs without manual symlinking, and I suspect that no one ever does
+ * that.
+ */
+struct BITSFUNC(fake_sections)
+{
+	ELF(Shdr) *table;
+	unsigned long table_offset;
+	int count, max_count;
+
+	int in_shstrndx;
+	unsigned long shstr_offset;
+	const char *shstrtab;
+	size_t shstrtab_len;
+
+	int out_shstrndx;
+};
+
+static unsigned int BITSFUNC(find_shname)(struct BITSFUNC(fake_sections) *out,
+					  const char *name)
+{
+	const char *outname = out->shstrtab;
+	while (outname - out->shstrtab < out->shstrtab_len) {
+		if (!strcmp(name, outname))
+			return (outname - out->shstrtab) + out->shstr_offset;
+		outname += strlen(outname) + 1;
+	}
+
+	if (*name)
+		printf("Warning: could not find output name \"%s\"\n", name);
+	return out->shstr_offset + out->shstrtab_len - 1;  /* Use a null. */
+}
+
+static void BITSFUNC(init_sections)(struct BITSFUNC(fake_sections) *out)
+{
+	if (!out->in_shstrndx)
+		fail("didn't find the fake shstrndx\n");
+
+	memset(out->table, 0, out->max_count * sizeof(ELF(Shdr)));
+
+	if (out->max_count < 1)
+		fail("we need at least two fake output sections\n");
+
+	PUT_LE(&out->table[0].sh_type, SHT_NULL);
+	PUT_LE(&out->table[0].sh_name, BITSFUNC(find_shname)(out, ""));
+
+	out->count = 1;
+}
+
+static void BITSFUNC(copy_section)(struct BITSFUNC(fake_sections) *out,
+				   int in_idx, const ELF(Shdr) *in,
+				   const char *name)
+{
+	uint64_t flags = GET_LE(&in->sh_flags);
+
+	bool copy = flags & SHF_ALLOC &&
+		(GET_LE(&in->sh_size) ||
+		 (GET_LE(&in->sh_type) != SHT_RELA &&
+		  GET_LE(&in->sh_type) != SHT_REL)) &&
+		strcmp(name, ".altinstructions") &&
+		strcmp(name, ".altinstr_replacement");
+
+	if (!copy)
+		return;
+
+	if (out->count >= out->max_count)
+		fail("too many copied sections (max = %d)\n", out->max_count);
+
+	if (in_idx == out->in_shstrndx)
+		out->out_shstrndx = out->count;
+
+	out->table[out->count] = *in;
+	PUT_LE(&out->table[out->count].sh_name,
+	       BITSFUNC(find_shname)(out, name));
+
+	/* elfutils requires that a strtab have the correct type. */
+	if (!strcmp(name, ".fake_shstrtab"))
+		PUT_LE(&out->table[out->count].sh_type, SHT_STRTAB);
+
+	out->count++;
+}
+
+static void BITSFUNC(go)(void *addr, size_t len,
+			 FILE *outfile, const char *name)
 {
 	int found_load = 0;
 	unsigned long load_size = -1;  /* Work around bogus warning */
 	unsigned long data_size;
-	Elf_Ehdr *hdr = (Elf_Ehdr *)addr;
+	ELF(Ehdr) *hdr = (ELF(Ehdr) *)addr;
 	int i;
 	unsigned long j;
-	Elf_Shdr *symtab_hdr = NULL, *strtab_hdr, *secstrings_hdr,
+	ELF(Shdr) *symtab_hdr = NULL, *strtab_hdr, *secstrings_hdr,
 		*alt_sec = NULL;
-	Elf_Dyn *dyn = 0, *dyn_end = 0;
+	ELF(Dyn) *dyn = 0, *dyn_end = 0;
 	const char *secstrings;
 	uint64_t syms[NSYMS] = {};
 
-	uint64_t fake_sections_value = 0, fake_sections_size = 0;
+	struct BITSFUNC(fake_sections) fake_sections = {};
 
-	Elf_Phdr *pt = (Elf_Phdr *)(addr + GET_LE(&hdr->e_phoff));
+	ELF(Phdr) *pt = (ELF(Phdr) *)(addr + GET_LE(&hdr->e_phoff));
 
 	/* Walk the segment table. */
 	for (i = 0; i < GET_LE(&hdr->e_phnum); i++) {
@@ -51,7 +167,7 @@ static void GOFUNC(void *addr, size_t len, FILE *outfile, const char *name)
 	for (i = 0; dyn + i < dyn_end &&
 		     GET_LE(&dyn[i].d_tag) != DT_NULL; i++) {
 		typeof(dyn[i].d_tag) tag = GET_LE(&dyn[i].d_tag);
-		if (tag == DT_REL || tag == DT_RELSZ ||
+		if (tag == DT_REL || tag == DT_RELSZ || tag == DT_RELA ||
 		    tag == DT_RELENT || tag == DT_TEXTREL)
 			fail("vdso image contains dynamic relocations\n");
 	}
@@ -61,7 +177,7 @@ static void GOFUNC(void *addr, size_t len, FILE *outfile, const char *name)
 		GET_LE(&hdr->e_shentsize)*GET_LE(&hdr->e_shstrndx);
 	secstrings = addr + GET_LE(&secstrings_hdr->sh_offset);
 	for (i = 0; i < GET_LE(&hdr->e_shnum); i++) {
-		Elf_Shdr *sh = addr + GET_LE(&hdr->e_shoff) +
+		ELF(Shdr) *sh = addr + GET_LE(&hdr->e_shoff) +
 			GET_LE(&hdr->e_shentsize) * i;
 		if (GET_LE(&sh->sh_type) == SHT_SYMTAB)
 			symtab_hdr = sh;
@@ -82,29 +198,63 @@ static void GOFUNC(void *addr, size_t len, FILE *outfile, const char *name)
 	     i < GET_LE(&symtab_hdr->sh_size) / GET_LE(&symtab_hdr->sh_entsize);
 	     i++) {
 		int k;
-		Elf_Sym *sym = addr + GET_LE(&symtab_hdr->sh_offset) +
+		ELF(Sym) *sym = addr + GET_LE(&symtab_hdr->sh_offset) +
 			GET_LE(&symtab_hdr->sh_entsize) * i;
 		const char *name = addr + GET_LE(&strtab_hdr->sh_offset) +
 			GET_LE(&sym->st_name);
 
 		for (k = 0; k < NSYMS; k++) {
-			if (!strcmp(name, required_syms[k])) {
+			if (!strcmp(name, required_syms[k].name)) {
 				if (syms[k]) {
 					fail("duplicate symbol %s\n",
-					     required_syms[k]);
+					     required_syms[k].name);
 				}
 				syms[k] = GET_LE(&sym->st_value);
 			}
 		}
 
-		if (!strcmp(name, "vdso_fake_sections")) {
-			if (fake_sections_value)
-				fail("duplicate vdso_fake_sections\n");
-			fake_sections_value = GET_LE(&sym->st_value);
-			fake_sections_size = GET_LE(&sym->st_size);
+		if (!strcmp(name, "fake_shstrtab")) {
+			ELF(Shdr) *sh;
+
+			fake_sections.in_shstrndx = GET_LE(&sym->st_shndx);
+			fake_sections.shstrtab = addr + GET_LE(&sym->st_value);
+			fake_sections.shstrtab_len = GET_LE(&sym->st_size);
+			sh = addr + GET_LE(&hdr->e_shoff) +
+				GET_LE(&hdr->e_shentsize) *
+				fake_sections.in_shstrndx;
+			fake_sections.shstr_offset = GET_LE(&sym->st_value) -
+				GET_LE(&sh->sh_addr);
 		}
 	}
 
+	/* Build the output section table. */
+	if (!syms[sym_VDSO_FAKE_SECTION_TABLE_START] ||
+	    !syms[sym_VDSO_FAKE_SECTION_TABLE_END])
+		fail("couldn't find fake section table\n");
+	if ((syms[sym_VDSO_FAKE_SECTION_TABLE_END] -
+	     syms[sym_VDSO_FAKE_SECTION_TABLE_START]) % sizeof(ELF(Shdr)))
+		fail("fake section table size isn't a multiple of sizeof(Shdr)\n");
+	fake_sections.table = addr + syms[sym_VDSO_FAKE_SECTION_TABLE_START];
+	fake_sections.table_offset = syms[sym_VDSO_FAKE_SECTION_TABLE_START];
+	fake_sections.max_count = (syms[sym_VDSO_FAKE_SECTION_TABLE_END] -
+				   syms[sym_VDSO_FAKE_SECTION_TABLE_START]) /
+		sizeof(ELF(Shdr));
+
+	BITSFUNC(init_sections)(&fake_sections);
+	for (i = 0; i < GET_LE(&hdr->e_shnum); i++) {
+		ELF(Shdr) *sh = addr + GET_LE(&hdr->e_shoff) +
+			GET_LE(&hdr->e_shentsize) * i;
+		BITSFUNC(copy_section)(&fake_sections, i, sh,
+				       secstrings + GET_LE(&sh->sh_name));
+	}
+	if (!fake_sections.out_shstrndx)
+		fail("didn't generate shstrndx?!?\n");
+
+	PUT_LE(&hdr->e_shoff, fake_sections.table_offset);
+	PUT_LE(&hdr->e_shentsize, sizeof(ELF(Shdr)));
+	PUT_LE(&hdr->e_shnum, fake_sections.count);
+	PUT_LE(&hdr->e_shstrndx, fake_sections.out_shstrndx);
+
 	/* Validate mapping addresses. */
 	for (i = 0; i < sizeof(special_pages) / sizeof(special_pages[0]); i++) {
 		if (!syms[i])
@@ -112,25 +262,17 @@ static void GOFUNC(void *addr, size_t len, FILE *outfile, const char *name)
 
 		if (syms[i] % 4096)
 			fail("%s must be a multiple of 4096\n",
-			     required_syms[i]);
+			     required_syms[i].name);
 		if (syms[i] < data_size)
 			fail("%s must be after the text mapping\n",
-			     required_syms[i]);
+			     required_syms[i].name);
 		if (syms[sym_end_mapping] < syms[i] + 4096)
-			fail("%s overruns end_mapping\n", required_syms[i]);
+			fail("%s overruns end_mapping\n",
+			     required_syms[i].name);
 	}
 	if (syms[sym_end_mapping] % 4096)
 		fail("end_mapping must be a multiple of 4096\n");
 
-	/* Remove sections or use fakes */
-	if (fake_sections_size % sizeof(Elf_Shdr))
-		fail("vdso_fake_sections size is not a multiple of %ld\n",
-		     (long)sizeof(Elf_Shdr));
-	PUT_LE(&hdr->e_shoff, fake_sections_value);
-	PUT_LE(&hdr->e_shentsize, fake_sections_value ? sizeof(Elf_Shdr) : 0);
-	PUT_LE(&hdr->e_shnum, fake_sections_size / sizeof(Elf_Shdr));
-	PUT_LE(&hdr->e_shstrndx, SHN_UNDEF);
-
 	if (!name) {
 		fwrite(addr, load_size, 1, outfile);
 		return;
@@ -168,9 +310,9 @@ static void GOFUNC(void *addr, size_t len, FILE *outfile, const char *name)
 			(unsigned long)GET_LE(&alt_sec->sh_size));
 	}
 	for (i = 0; i < NSYMS; i++) {
-		if (syms[i])
+		if (required_syms[i].export && syms[i])
 			fprintf(outfile, "\t.sym_%s = 0x%" PRIx64 ",\n",
-				required_syms[i], syms[i]);
+				required_syms[i].name, syms[i]);
 	}
 	fprintf(outfile, "};\n");
 }
diff --git a/arch/x86/vdso/vdso32/vdso-fakesections.c b/arch/x86/vdso/vdso32/vdso-fakesections.c
new file mode 100644
index 000000000000..541468e25265
--- /dev/null
+++ b/arch/x86/vdso/vdso32/vdso-fakesections.c
@@ -0,0 +1 @@
+#include "../vdso-fakesections.c"
diff --git a/arch/x86/vdso/vdsox32.lds.S b/arch/x86/vdso/vdsox32.lds.S
index 46b991b578a8..697c11ece90c 100644
--- a/arch/x86/vdso/vdsox32.lds.S
+++ b/arch/x86/vdso/vdsox32.lds.S
@@ -6,6 +6,8 @@
  * the DSO.
  */
 
+#define BUILD_VDSOX32
+
 #include "vdso-layout.lds.S"
 
 /*
diff --git a/arch/x86/vdso/vma.c b/arch/x86/vdso/vma.c
index e1513c47872a..5a5176de8d0a 100644
--- a/arch/x86/vdso/vma.c
+++ b/arch/x86/vdso/vma.c
@@ -62,6 +62,9 @@ struct linux_binprm;
    Only used for the 64-bit and x32 vdsos. */
 static unsigned long vdso_addr(unsigned long start, unsigned len)
 {
+#ifdef CONFIG_X86_32
+	return 0;
+#else
 	unsigned long addr, end;
 	unsigned offset;
 	end = (start + PMD_SIZE - 1) & PMD_MASK;
@@ -83,6 +86,7 @@ static unsigned long vdso_addr(unsigned long start, unsigned len)
 	addr = align_vdso_addr(addr);
 
 	return addr;
+#endif
 }
 
 static int map_vdso(const struct vdso_image *image, bool calculate_addr)
diff --git a/arch/xtensa/kernel/vectors.S b/arch/xtensa/kernel/vectors.S
index f9e1ec346e35..8453e6e39895 100644
--- a/arch/xtensa/kernel/vectors.S
+++ b/arch/xtensa/kernel/vectors.S
@@ -376,38 +376,42 @@ _DoubleExceptionVector_WindowOverflow:
 	beqz	a2, 1f		# if at start of vector, don't restore
 
 	addi	a0, a0, -128
-	bbsi	a0, 8, 1f	# don't restore except for overflow 8 and 12
-	bbsi	a0, 7, 2f
+	bbsi.l	a0, 8, 1f	# don't restore except for overflow 8 and 12
+
+	/*
+	 * This fixup handler is for the extremely unlikely case where the
+	 * overflow handler's reference thru a0 gets a hardware TLB refill
+	 * that bumps out the (distinct, aliasing) TLB entry that mapped its
+	 * prior references thru a9/a13, and where our reference now thru
+	 * a9/a13 gets a 2nd-level miss exception (not hardware TLB refill).
+	 */
+	movi	a2, window_overflow_restore_a0_fixup
+	s32i	a2, a3, EXC_TABLE_FIXUP
+	l32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+	xsr	a3, excsave1
+
+	bbsi.l	a0, 7, 2f
 
 	/*
 	 * Restore a0 as saved by _WindowOverflow8().
-	 *
-	 * FIXME:  we really need a fixup handler for this L32E,
-	 * for the extremely unlikely case where the overflow handler's
-	 * reference thru a0 gets a hardware TLB refill that bumps out
-	 * the (distinct, aliasing) TLB entry that mapped its prior
-	 * references thru a9, and where our reference now thru a9
-	 * gets a 2nd-level miss exception (not hardware TLB refill).
 	 */
 
-	l32e	a2, a9, -16
-	wsr	a2, depc	# replace the saved a0
-	j	1f
+	l32e	a0, a9, -16
+	wsr	a0, depc	# replace the saved a0
+	j	3f
 
 2:
 	/*
 	 * Restore a0 as saved by _WindowOverflow12().
-	 *
-	 * FIXME:  we really need a fixup handler for this L32E,
-	 * for the extremely unlikely case where the overflow handler's
-	 * reference thru a0 gets a hardware TLB refill that bumps out
-	 * the (distinct, aliasing) TLB entry that mapped its prior
-	 * references thru a13, and where our reference now thru a13
-	 * gets a 2nd-level miss exception (not hardware TLB refill).
 	 */
 
-	l32e	a2, a13, -16
-	wsr	a2, depc	# replace the saved a0
+	l32e	a0, a13, -16
+	wsr	a0, depc	# replace the saved a0
+3:
+	xsr	a3, excsave1
+	movi	a0, 0
+	s32i	a0, a3, EXC_TABLE_FIXUP
+	s32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
 1:
 	/*
 	 * Restore WindowBase while leaving all address registers restored.
@@ -449,6 +453,7 @@ _DoubleExceptionVector_WindowOverflow:
 
 	s32i	a0, a2, PT_DEPC
 
+_DoubleExceptionVector_handle_exception:
 	addx4	a0, a0, a3
 	l32i	a0, a0, EXC_TABLE_FAST_USER
 	xsr	a3, excsave1
@@ -464,11 +469,120 @@ _DoubleExceptionVector_WindowOverflow:
 	rotw	-3
 	j	1b
 
-	.end literal_prefix
 
 ENDPROC(_DoubleExceptionVector)
 
 /*
+ * Fixup handler for TLB miss in double exception handler for window owerflow.
+ * We get here with windowbase set to the window that was being spilled and
+ * a0 trashed. a0 bit 7 determines if this is a call8 (bit clear) or call12
+ * (bit set) window.
+ *
+ * We do the following here:
+ * - go to the original window retaining a0 value;
+ * - set up exception stack to return back to appropriate a0 restore code
+ *   (we'll need to rotate window back and there's no place to save this
+ *    information, use different return address for that);
+ * - handle the exception;
+ * - go to the window that was being spilled;
+ * - set up window_overflow_restore_a0_fixup as a fixup routine;
+ * - reload a0;
+ * - restore the original window;
+ * - reset the default fixup routine;
+ * - return to user. By the time we get to this fixup handler all information
+ *   about the conditions of the original double exception that happened in
+ *   the window overflow handler is lost, so we just return to userspace to
+ *   retry overflow from start.
+ *
+ * a0: value of depc, original value in depc
+ * a2: trashed, original value in EXC_TABLE_DOUBLE_SAVE
+ * a3: exctable, original value in excsave1
+ */
+
+ENTRY(window_overflow_restore_a0_fixup)
+
+	rsr	a0, ps
+	extui	a0, a0, PS_OWB_SHIFT, PS_OWB_WIDTH
+	rsr	a2, windowbase
+	sub	a0, a2, a0
+	extui	a0, a0, 0, 3
+	l32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+	xsr	a3, excsave1
+
+	_beqi	a0, 1, .Lhandle_1
+	_beqi	a0, 3, .Lhandle_3
+
+	.macro	overflow_fixup_handle_exception_pane n
+
+	rsr	a0, depc
+	rotw	-\n
+
+	xsr	a3, excsave1
+	wsr	a2, depc
+	l32i	a2, a3, EXC_TABLE_KSTK
+	s32i	a0, a2, PT_AREG0
+
+	movi	a0, .Lrestore_\n
+	s32i	a0, a2, PT_DEPC
+	rsr	a0, exccause
+	j	_DoubleExceptionVector_handle_exception
+
+	.endm
+
+	overflow_fixup_handle_exception_pane 2
+.Lhandle_1:
+	overflow_fixup_handle_exception_pane 1
+.Lhandle_3:
+	overflow_fixup_handle_exception_pane 3
+
+	.macro	overflow_fixup_restore_a0_pane n
+
+	rotw	\n
+	/* Need to preserve a0 value here to be able to handle exception
+	 * that may occur on a0 reload from stack. It may occur because
+	 * TLB miss handler may not be atomic and pointer to page table
+	 * may be lost before we get here. There are no free registers,
+	 * so we need to use EXC_TABLE_DOUBLE_SAVE area.
+	 */
+	xsr	a3, excsave1
+	s32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+	movi	a2, window_overflow_restore_a0_fixup
+	s32i	a2, a3, EXC_TABLE_FIXUP
+	l32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+	xsr	a3, excsave1
+	bbsi.l	a0, 7, 1f
+	l32e	a0, a9, -16
+	j	2f
+1:
+	l32e	a0, a13, -16
+2:
+	rotw	-\n
+
+	.endm
+
+.Lrestore_2:
+	overflow_fixup_restore_a0_pane 2
+
+.Lset_default_fixup:
+	xsr	a3, excsave1
+	s32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+	movi	a2, 0
+	s32i	a2, a3, EXC_TABLE_FIXUP
+	l32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+	xsr	a3, excsave1
+	rfe
+
+.Lrestore_1:
+	overflow_fixup_restore_a0_pane 1
+	j	.Lset_default_fixup
+.Lrestore_3:
+	overflow_fixup_restore_a0_pane 3
+	j	.Lset_default_fixup
+
+ENDPROC(window_overflow_restore_a0_fixup)
+
+	.end literal_prefix
+/*
  * Debug interrupt vector
  *
  * There is not much space here, so simply jump to another handler.
diff --git a/arch/xtensa/kernel/vmlinux.lds.S b/arch/xtensa/kernel/vmlinux.lds.S
index ee32c0085dff..d16db6df86f8 100644
--- a/arch/xtensa/kernel/vmlinux.lds.S
+++ b/arch/xtensa/kernel/vmlinux.lds.S
@@ -269,13 +269,13 @@ SECTIONS
 		  .UserExceptionVector.literal)
   SECTION_VECTOR (_DoubleExceptionVector_literal,
 		  .DoubleExceptionVector.literal,
-		  DOUBLEEXC_VECTOR_VADDR - 16,
+		  DOUBLEEXC_VECTOR_VADDR - 40,
 		  SIZEOF(.UserExceptionVector.text),
 		  .UserExceptionVector.text)
   SECTION_VECTOR (_DoubleExceptionVector_text,
 		  .DoubleExceptionVector.text,
 		  DOUBLEEXC_VECTOR_VADDR,
-		  32,
+		  40,
 		  .DoubleExceptionVector.literal)
 
   . = (LOADADDR( .DoubleExceptionVector.text ) + SIZEOF( .DoubleExceptionVector.text ) + 3) & ~ 3;
diff --git a/arch/xtensa/mm/init.c b/arch/xtensa/mm/init.c
index 4224256bb215..77ed20209ca5 100644
--- a/arch/xtensa/mm/init.c
+++ b/arch/xtensa/mm/init.c
@@ -191,7 +191,7 @@ int __init mem_reserve(unsigned long start, unsigned long end, int must_exist)
 		return -EINVAL;
 	}
 
-	if (it && start - it->start < bank_sz) {
+	if (it && start - it->start <= bank_sz) {
 		if (start == it->start) {
 			if (end - it->start < bank_sz) {
 				it->start = end;
author	Ingo Molnar <mingo@kernel.org>	2014-07-28 10:00:33 +0200
committer	Ingo Molnar <mingo@kernel.org>	2014-07-28 10:00:33 +0200
commit	5030c69755416d19516c0a61cd988a0e0062e041 (patch)
tree	ca4d3088428cf033420f9df9111d796650bc59e5 /arch
parent	2336ebc32676df5b794acfe0c980583ec6c05f34 (diff)
parent	64aa90f26c06e1cb2aacfb98a7d0eccfbd6c1a91 (diff)
download	blackbird-op-linux-5030c69755416d19516c0a61cd988a0e0062e041.tar.gz blackbird-op-linux-5030c69755416d19516c0a61cd988a0e0062e041.zip