diff options
author | David S. Miller <davem@davemloft.net> | 2009-03-17 15:01:30 -0700 |
---|---|---|
committer | David S. Miller <davem@davemloft.net> | 2009-03-17 15:01:30 -0700 |
commit | 2d6a5e9500103680464a723a4564961675652680 (patch) | |
tree | d18903333aae8a4415b179d6e7d38f203724892c | |
parent | bd257ed9f1d129b4e881f513a406b435c8852565 (diff) | |
parent | f10023a4ef3f5cc05457b059c6880bc447adfa1f (diff) | |
download | blackbird-op-linux-2d6a5e9500103680464a723a4564961675652680.tar.gz blackbird-op-linux-2d6a5e9500103680464a723a4564961675652680.zip |
Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6
Conflicts:
drivers/net/igb/igb_main.c
drivers/net/qlge/qlge_main.c
drivers/net/wireless/ath9k/ath9k.h
drivers/net/wireless/ath9k/core.h
drivers/net/wireless/ath9k/hw.c
40 files changed, 5785 insertions, 231 deletions
diff --git a/MAINTAINERS b/MAINTAINERS index 0d9ea2a4e497..27114518eb05 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -3888,6 +3888,15 @@ L: linux-ide@vger.kernel.org T: git kernel.org:/pub/scm/linux/kernel/git/jgarzik/libata-dev.git S: Supported +SERVER ENGINES 10Gbps NIC - BladeEngine 2 DRIVER +P: Sathya Perla +M: sathyap@serverengines.com +P: Subbu Seetharaman +M: subbus@serverengines.com +L: netdev@vger.kernel.org +W: http://www.serverengines.com +S: Supported + SFC NETWORK DRIVER P: Steve Hodgson P: Ben Hutchings diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig index 45403e67e351..6b1af549b2ff 100644 --- a/drivers/net/Kconfig +++ b/drivers/net/Kconfig @@ -1040,6 +1040,17 @@ config NI65 To compile this driver as a module, choose M here. The module will be called ni65. +config DNET + tristate "Dave ethernet support (DNET)" + depends on NET_ETHERNET + select PHYLIB + help + The Dave ethernet interface (DNET) is found on Qong Board FPGA. + Say Y to include support for the DNET chip. + + To compile this driver as a module, choose M here: the module + will be called dnet. + source "drivers/net/tulip/Kconfig" config AT1700 @@ -2618,6 +2629,8 @@ config QLGE source "drivers/net/sfc/Kconfig" +source "drivers/net/benet/Kconfig" + endif # NETDEV_10000 source "drivers/net/tokenring/Kconfig" diff --git a/drivers/net/Makefile b/drivers/net/Makefile index 3f8cb311e077..758ecdf4c820 100644 --- a/drivers/net/Makefile +++ b/drivers/net/Makefile @@ -22,6 +22,7 @@ obj-$(CONFIG_GIANFAR) += gianfar_driver.o obj-$(CONFIG_TEHUTI) += tehuti.o obj-$(CONFIG_ENIC) += enic/ obj-$(CONFIG_JME) += jme.o +obj-$(CONFIG_BE2NET) += benet/ gianfar_driver-objs := gianfar.o \ gianfar_ethtool.o \ @@ -232,6 +233,7 @@ obj-$(CONFIG_ENC28J60) += enc28j60.o obj-$(CONFIG_XTENSA_XT2000_SONIC) += xtsonic.o +obj-$(CONFIG_DNET) += dnet.o obj-$(CONFIG_MACB) += macb.o obj-$(CONFIG_ARM) += arm/ diff --git a/drivers/net/benet/Kconfig b/drivers/net/benet/Kconfig new file mode 100644 index 000000000000..c6934f179c09 --- /dev/null +++ b/drivers/net/benet/Kconfig @@ -0,0 +1,7 @@ +config BE2NET + tristate "ServerEngines' 10Gbps NIC - BladeEngine 2" + depends on PCI && INET + select INET_LRO + help + This driver implements the NIC functionality for ServerEngines' + 10Gbps network adapter - BladeEngine 2. diff --git a/drivers/net/benet/Makefile b/drivers/net/benet/Makefile new file mode 100644 index 000000000000..a60cd8051135 --- /dev/null +++ b/drivers/net/benet/Makefile @@ -0,0 +1,7 @@ +# +# Makefile to build the network driver for ServerEngine's BladeEngine. +# + +obj-$(CONFIG_BE2NET) += be2net.o + +be2net-y := be_main.o be_cmds.o be_ethtool.o diff --git a/drivers/net/benet/be.h b/drivers/net/benet/be.h new file mode 100644 index 000000000000..63d593d53153 --- /dev/null +++ b/drivers/net/benet/be.h @@ -0,0 +1,327 @@ +/* + * Copyright (C) 2005 - 2009 ServerEngines + * All rights reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. The full GNU General + * Public License is included in this distribution in the file called COPYING. + * + * Contact Information: + * linux-drivers@serverengines.com + * + * ServerEngines + * 209 N. Fair Oaks Ave + * Sunnyvale, CA 94085 + */ + +#ifndef BE_H +#define BE_H + +#include <linux/pci.h> +#include <linux/etherdevice.h> +#include <linux/version.h> +#include <linux/delay.h> +#include <net/tcp.h> +#include <net/ip.h> +#include <net/ipv6.h> +#include <linux/if_vlan.h> +#include <linux/workqueue.h> +#include <linux/interrupt.h> +#include <linux/inet_lro.h> + +#include "be_hw.h" + +#define DRV_VER "2.0.348" +#define DRV_NAME "be2net" +#define BE_NAME "ServerEngines BladeEngine2 10Gbps NIC" +#define DRV_DESC BE_NAME "Driver" + +/* Number of bytes of an RX frame that are copied to skb->data */ +#define BE_HDR_LEN 64 +#define BE_MAX_JUMBO_FRAME_SIZE 9018 +#define BE_MIN_MTU 256 + +#define BE_NUM_VLANS_SUPPORTED 64 +#define BE_MAX_EQD 96 +#define BE_MAX_TX_FRAG_COUNT 30 + +#define EVNT_Q_LEN 1024 +#define TX_Q_LEN 2048 +#define TX_CQ_LEN 1024 +#define RX_Q_LEN 1024 /* Does not support any other value */ +#define RX_CQ_LEN 1024 +#define MCC_Q_LEN 64 /* total size not to exceed 8 pages */ +#define MCC_CQ_LEN 256 + +#define BE_NAPI_WEIGHT 64 +#define MAX_RX_POST BE_NAPI_WEIGHT /* Frags posted at a time */ +#define RX_FRAGS_REFILL_WM (RX_Q_LEN - MAX_RX_POST) + +#define BE_MAX_LRO_DESCRIPTORS 16 +#define BE_MAX_FRAGS_PER_FRAME 16 + +struct be_dma_mem { + void *va; + dma_addr_t dma; + u32 size; +}; + +struct be_queue_info { + struct be_dma_mem dma_mem; + u16 len; + u16 entry_size; /* Size of an element in the queue */ + u16 id; + u16 tail, head; + bool created; + atomic_t used; /* Number of valid elements in the queue */ +}; + +struct be_ctrl_info { + u8 __iomem *csr; + u8 __iomem *db; /* Door Bell */ + u8 __iomem *pcicfg; /* PCI config space */ + int pci_func; + + /* Mbox used for cmd request/response */ + spinlock_t cmd_lock; /* For serializing cmds to BE card */ + struct be_dma_mem mbox_mem; + /* Mbox mem is adjusted to align to 16 bytes. The allocated addr + * is stored for freeing purpose */ + struct be_dma_mem mbox_mem_alloced; +}; + +#include "be_cmds.h" + +struct be_drvr_stats { + u32 be_tx_reqs; /* number of TX requests initiated */ + u32 be_tx_stops; /* number of times TX Q was stopped */ + u32 be_fwd_reqs; /* number of send reqs through forwarding i/f */ + u32 be_tx_wrbs; /* number of tx WRBs used */ + u32 be_tx_events; /* number of tx completion events */ + u32 be_tx_compl; /* number of tx completion entries processed */ + u64 be_tx_jiffies; + ulong be_tx_bytes; + ulong be_tx_bytes_prev; + u32 be_tx_rate; + + u32 cache_barrier[16]; + + u32 be_ethrx_post_fail;/* number of ethrx buffer alloc failures */ + u32 be_polls; /* number of times NAPI called poll function */ + u32 be_rx_events; /* number of ucast rx completion events */ + u32 be_rx_compl; /* number of rx completion entries processed */ + u32 be_lro_hgram_data[8]; /* histogram of LRO data packets */ + u32 be_lro_hgram_ack[8]; /* histogram of LRO ACKs */ + u64 be_rx_jiffies; + ulong be_rx_bytes; + ulong be_rx_bytes_prev; + u32 be_rx_rate; + /* number of non ether type II frames dropped where + * frame len > length field of Mac Hdr */ + u32 be_802_3_dropped_frames; + /* number of non ether type II frames malformed where + * in frame len < length field of Mac Hdr */ + u32 be_802_3_malformed_frames; + u32 be_rxcp_err; /* Num rx completion entries w/ err set. */ + ulong rx_fps_jiffies; /* jiffies at last FPS calc */ + u32 be_rx_frags; + u32 be_prev_rx_frags; + u32 be_rx_fps; /* Rx frags per second */ +}; + +struct be_stats_obj { + struct be_drvr_stats drvr_stats; + struct net_device_stats net_stats; + struct be_dma_mem cmd; +}; + +struct be_eq_obj { + struct be_queue_info q; + char desc[32]; + + /* Adaptive interrupt coalescing (AIC) info */ + bool enable_aic; + u16 min_eqd; /* in usecs */ + u16 max_eqd; /* in usecs */ + u16 cur_eqd; /* in usecs */ + + struct napi_struct napi; +}; + +struct be_tx_obj { + struct be_queue_info q; + struct be_queue_info cq; + /* Remember the skbs that were transmitted */ + struct sk_buff *sent_skb_list[TX_Q_LEN]; +}; + +/* Struct to remember the pages posted for rx frags */ +struct be_rx_page_info { + struct page *page; + dma_addr_t bus; + u16 page_offset; + bool last_page_user; +}; + +struct be_rx_obj { + struct be_queue_info q; + struct be_queue_info cq; + struct be_rx_page_info page_info_tbl[RX_Q_LEN]; + struct net_lro_mgr lro_mgr; + struct net_lro_desc lro_desc[BE_MAX_LRO_DESCRIPTORS]; +}; + +#define BE_NUM_MSIX_VECTORS 2 /* 1 each for Tx and Rx */ +struct be_adapter { + struct pci_dev *pdev; + struct net_device *netdev; + + /* Mbox, pci config, csr address information */ + struct be_ctrl_info ctrl; + + struct msix_entry msix_entries[BE_NUM_MSIX_VECTORS]; + bool msix_enabled; + bool isr_registered; + + /* TX Rings */ + struct be_eq_obj tx_eq; + struct be_tx_obj tx_obj; + + u32 cache_line_break[8]; + + /* Rx rings */ + struct be_eq_obj rx_eq; + struct be_rx_obj rx_obj; + u32 big_page_size; /* Compounded page size shared by rx wrbs */ + + struct vlan_group *vlan_grp; + u16 num_vlans; + u8 vlan_tag[VLAN_GROUP_ARRAY_LEN]; + + struct be_stats_obj stats; + /* Work queue used to perform periodic tasks like getting statistics */ + struct delayed_work work; + + /* Ethtool knobs and info */ + bool rx_csum; /* BE card must perform rx-checksumming */ + u32 max_rx_coal; + char fw_ver[FW_VER_LEN]; + u32 if_handle; /* Used to configure filtering */ + u32 pmac_id; /* MAC addr handle used by BE card */ + + struct be_link_info link; + u32 port_num; +}; + +extern struct ethtool_ops be_ethtool_ops; + +#define drvr_stats(adapter) (&adapter->stats.drvr_stats) + +#define BE_SET_NETDEV_OPS(netdev, ops) (netdev->netdev_ops = ops) + +static inline u32 MODULO(u16 val, u16 limit) +{ + BUG_ON(limit & (limit - 1)); + return val & (limit - 1); +} + +static inline void index_adv(u16 *index, u16 val, u16 limit) +{ + *index = MODULO((*index + val), limit); +} + +static inline void index_inc(u16 *index, u16 limit) +{ + *index = MODULO((*index + 1), limit); +} + +#define PAGE_SHIFT_4K 12 +#define PAGE_SIZE_4K (1 << PAGE_SHIFT_4K) + +/* Returns number of pages spanned by the data starting at the given addr */ +#define PAGES_4K_SPANNED(_address, size) \ + ((u32)((((size_t)(_address) & (PAGE_SIZE_4K - 1)) + \ + (size) + (PAGE_SIZE_4K - 1)) >> PAGE_SHIFT_4K)) + +/* Byte offset into the page corresponding to given address */ +#define OFFSET_IN_PAGE(addr) \ + ((size_t)(addr) & (PAGE_SIZE_4K-1)) + +/* Returns bit offset within a DWORD of a bitfield */ +#define AMAP_BIT_OFFSET(_struct, field) \ + (((size_t)&(((_struct *)0)->field))%32) + +/* Returns the bit mask of the field that is NOT shifted into location. */ +static inline u32 amap_mask(u32 bitsize) +{ + return (bitsize == 32 ? 0xFFFFFFFF : (1 << bitsize) - 1); +} + +static inline void +amap_set(void *ptr, u32 dw_offset, u32 mask, u32 offset, u32 value) +{ + u32 *dw = (u32 *) ptr + dw_offset; + *dw &= ~(mask << offset); + *dw |= (mask & value) << offset; +} + +#define AMAP_SET_BITS(_struct, field, ptr, val) \ + amap_set(ptr, \ + offsetof(_struct, field)/32, \ + amap_mask(sizeof(((_struct *)0)->field)), \ + AMAP_BIT_OFFSET(_struct, field), \ + val) + +static inline u32 amap_get(void *ptr, u32 dw_offset, u32 mask, u32 offset) +{ + u32 *dw = (u32 *) ptr; + return mask & (*(dw + dw_offset) >> offset); +} + +#define AMAP_GET_BITS(_struct, field, ptr) \ + amap_get(ptr, \ + offsetof(_struct, field)/32, \ + amap_mask(sizeof(((_struct *)0)->field)), \ + AMAP_BIT_OFFSET(_struct, field)) + +#define be_dws_cpu_to_le(wrb, len) swap_dws(wrb, len) +#define be_dws_le_to_cpu(wrb, len) swap_dws(wrb, len) +static inline void swap_dws(void *wrb, int len) +{ +#ifdef __BIG_ENDIAN + u32 *dw = wrb; + BUG_ON(len % 4); + do { + *dw = cpu_to_le32(*dw); + dw++; + len -= 4; + } while (len); +#endif /* __BIG_ENDIAN */ +} + +static inline u8 is_tcp_pkt(struct sk_buff *skb) +{ + u8 val = 0; + + if (ip_hdr(skb)->version == 4) + val = (ip_hdr(skb)->protocol == IPPROTO_TCP); + else if (ip_hdr(skb)->version == 6) + val = (ipv6_hdr(skb)->nexthdr == NEXTHDR_TCP); + + return val; +} + +static inline u8 is_udp_pkt(struct sk_buff *skb) +{ + u8 val = 0; + + if (ip_hdr(skb)->version == 4) + val = (ip_hdr(skb)->protocol == IPPROTO_UDP); + else if (ip_hdr(skb)->version == 6) + val = (ipv6_hdr(skb)->nexthdr == NEXTHDR_UDP); + + return val; +} + +#endif /* BE_H */ diff --git a/drivers/net/benet/be_cmds.c b/drivers/net/benet/be_cmds.c new file mode 100644 index 000000000000..d444aed962bc --- /dev/null +++ b/drivers/net/benet/be_cmds.c @@ -0,0 +1,861 @@ +/* + * Copyright (C) 2005 - 2009 ServerEngines + * All rights reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. The full GNU General + * Public License is included in this distribution in the file called COPYING. + * + * Contact Information: + * linux-drivers@serverengines.com + * + * ServerEngines + * 209 N. Fair Oaks Ave + * Sunnyvale, CA 94085 + */ + +#include "be.h" + +static int be_mbox_db_ready_wait(void __iomem *db) +{ + int cnt = 0, wait = 5; + u32 ready; + + do { + ready = ioread32(db) & MPU_MAILBOX_DB_RDY_MASK; + if (ready) + break; + + if (cnt > 200000) { + printk(KERN_WARNING DRV_NAME + ": mbox_db poll timed out\n"); + return -1; + } + + if (cnt > 50) + wait = 200; + cnt += wait; + udelay(wait); + } while (true); + + return 0; +} + +/* + * Insert the mailbox address into the doorbell in two steps + */ +static int be_mbox_db_ring(struct be_ctrl_info *ctrl) +{ + int status; + u16 compl_status, extd_status; + u32 val = 0; + void __iomem *db = ctrl->db + MPU_MAILBOX_DB_OFFSET; + struct be_dma_mem *mbox_mem = &ctrl->mbox_mem; + struct be_mcc_mailbox *mbox = mbox_mem->va; + struct be_mcc_cq_entry *cqe = &mbox->cqe; + + memset(cqe, 0, sizeof(*cqe)); + + val &= ~MPU_MAILBOX_DB_RDY_MASK; + val |= MPU_MAILBOX_DB_HI_MASK; + /* at bits 2 - 31 place mbox dma addr msb bits 34 - 63 */ + val |= (upper_32_bits(mbox_mem->dma) >> 2) << 2; + iowrite32(val, db); + + /* wait for ready to be set */ + status = be_mbox_db_ready_wait(db); + if (status != 0) + return status; + + val = 0; + val &= ~MPU_MAILBOX_DB_RDY_MASK; + val &= ~MPU_MAILBOX_DB_HI_MASK; + /* at bits 2 - 31 place mbox dma addr lsb bits 4 - 33 */ + val |= (u32)(mbox_mem->dma >> 4) << 2; + iowrite32(val, db); + + status = be_mbox_db_ready_wait(db); + if (status != 0) + return status; + + /* compl entry has been made now */ + be_dws_le_to_cpu(cqe, sizeof(*cqe)); + if (!(cqe->flags & CQE_FLAGS_VALID_MASK)) { + printk(KERN_WARNING DRV_NAME ": ERROR invalid mbox compl\n"); + return -1; + } + + compl_status = (cqe->status >> CQE_STATUS_COMPL_SHIFT) & + CQE_STATUS_COMPL_MASK; + if (compl_status != MCC_STATUS_SUCCESS) { + extd_status = (cqe->status >> CQE_STATUS_EXTD_SHIFT) & + CQE_STATUS_EXTD_MASK; + printk(KERN_WARNING DRV_NAME + ": ERROR in cmd compl. status(compl/extd)=%d/%d\n", + compl_status, extd_status); + } + + return compl_status; +} + +static int be_POST_stage_get(struct be_ctrl_info *ctrl, u16 *stage) +{ + u32 sem = ioread32(ctrl->csr + MPU_EP_SEMAPHORE_OFFSET); + + *stage = sem & EP_SEMAPHORE_POST_STAGE_MASK; + if ((sem >> EP_SEMAPHORE_POST_ERR_SHIFT) & EP_SEMAPHORE_POST_ERR_MASK) + return -1; + else + return 0; +} + +static int be_POST_stage_poll(struct be_ctrl_info *ctrl, u16 poll_stage) +{ + u16 stage, cnt, error; + for (cnt = 0; cnt < 5000; cnt++) { + error = be_POST_stage_get(ctrl, &stage); + if (error) + return -1; + + if (stage == poll_stage) + break; + udelay(1000); + } + if (stage != poll_stage) + return -1; + return 0; +} + + +int be_cmd_POST(struct be_ctrl_info *ctrl) +{ + u16 stage, error; + + error = be_POST_stage_get(ctrl, &stage); + if (error) + goto err; + + if (stage == POST_STAGE_ARMFW_RDY) + return 0; + + if (stage != POST_STAGE_AWAITING_HOST_RDY) + goto err; + + /* On awaiting host rdy, reset and again poll on awaiting host rdy */ + iowrite32(POST_STAGE_BE_RESET, ctrl->csr + MPU_EP_SEMAPHORE_OFFSET); + error = be_POST_stage_poll(ctrl, POST_STAGE_AWAITING_HOST_RDY); + if (error) + goto err; + + /* Now kickoff POST and poll on armfw ready */ + iowrite32(POST_STAGE_HOST_RDY, ctrl->csr + MPU_EP_SEMAPHORE_OFFSET); + error = be_POST_stage_poll(ctrl, POST_STAGE_ARMFW_RDY); + if (error) + goto err; + + return 0; +err: + printk(KERN_WARNING DRV_NAME ": ERROR, stage=%d\n", stage); + return -1; +} + +static inline void *embedded_payload(struct be_mcc_wrb *wrb) +{ + return wrb->payload.embedded_payload; +} + +static inline struct be_sge *nonembedded_sgl(struct be_mcc_wrb *wrb) +{ + return &wrb->payload.sgl[0]; +} + +/* Don't touch the hdr after it's prepared */ +static void be_wrb_hdr_prepare(struct be_mcc_wrb *wrb, int payload_len, + bool embedded, u8 sge_cnt) +{ + if (embedded) + wrb->embedded |= MCC_WRB_EMBEDDED_MASK; + else + wrb->embedded |= (sge_cnt & MCC_WRB_SGE_CNT_MASK) << + MCC_WRB_SGE_CNT_SHIFT; + wrb->payload_length = payload_len; + be_dws_cpu_to_le(wrb, 20); +} + +/* Don't touch the hdr after it's prepared */ +static void be_cmd_hdr_prepare(struct be_cmd_req_hdr *req_hdr, + u8 subsystem, u8 opcode, int cmd_len) +{ + req_hdr->opcode = opcode; + req_hdr->subsystem = subsystem; + req_hdr->request_length = cpu_to_le32(cmd_len - sizeof(*req_hdr)); +} + +static void be_cmd_page_addrs_prepare(struct phys_addr *pages, u32 max_pages, + struct be_dma_mem *mem) +{ + int i, buf_pages = min(PAGES_4K_SPANNED(mem->va, mem->size), max_pages); + u64 dma = (u64)mem->dma; + + for (i = 0; i < buf_pages; i++) { + pages[i].lo = cpu_to_le32(dma & 0xFFFFFFFF); + pages[i].hi = cpu_to_le32(upper_32_bits(dma)); + dma += PAGE_SIZE_4K; + } +} + +/* Converts interrupt delay in microseconds to multiplier value */ +static u32 eq_delay_to_mult(u32 usec_delay) +{ +#define MAX_INTR_RATE 651042 + const u32 round = 10; + u32 multiplier; + + if (usec_delay == 0) + multiplier = 0; + else { + u32 interrupt_rate = 1000000 / usec_delay; + /* Max delay, corresponding to the lowest interrupt rate */ + if (interrupt_rate == 0) + multiplier = 1023; + else { + multiplier = (MAX_INTR_RATE - interrupt_rate) * round; + multiplier /= interrupt_rate; + /* Round the multiplier to the closest value.*/ + multiplier = (multiplier + round/2) / round; + multiplier = min(multiplier, (u32)1023); + } + } + return multiplier; +} + +static inline struct be_mcc_wrb *wrb_from_mbox(struct be_dma_mem *mbox_mem) +{ + return &((struct be_mcc_mailbox *)(mbox_mem->va))->wrb; +} + +int be_cmd_eq_create(struct be_ctrl_info *ctrl, + struct be_queue_info *eq, int eq_delay) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_eq_create *req = embedded_payload(wrb); + struct be_cmd_resp_eq_create *resp = embedded_payload(wrb); + struct be_dma_mem *q_mem = &eq->dma_mem; + int status; + + spin_lock(&ctrl->cmd_lock); + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON, + OPCODE_COMMON_EQ_CREATE, sizeof(*req)); + + req->num_pages = cpu_to_le16(PAGES_4K_SPANNED(q_mem->va, q_mem->size)); + + AMAP_SET_BITS(struct amap_eq_context, func, req->context, + ctrl->pci_func); + AMAP_SET_BITS(struct amap_eq_context, valid, req->context, 1); + /* 4byte eqe*/ + AMAP_SET_BITS(struct amap_eq_context, size, req->context, 0); + AMAP_SET_BITS(struct amap_eq_context, count, req->context, + __ilog2_u32(eq->len/256)); + AMAP_SET_BITS(struct amap_eq_context, delaymult, req->context, + eq_delay_to_mult(eq_delay)); + be_dws_cpu_to_le(req->context, sizeof(req->context)); + + be_cmd_page_addrs_prepare(req->pages, ARRAY_SIZE(req->pages), q_mem); + + status = be_mbox_db_ring(ctrl); + if (!status) { + eq->id = le16_to_cpu(resp->eq_id); + eq->created = true; + } + spin_unlock(&ctrl->cmd_lock); + return status; +} + +int be_cmd_mac_addr_query(struct be_ctrl_info *ctrl, u8 *mac_addr, + u8 type, bool permanent, u32 if_handle) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_mac_query *req = embedded_payload(wrb); + struct be_cmd_resp_mac_query *resp = embedded_payload(wrb); + int status; + + spin_lock(&ctrl->cmd_lock); + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON, + OPCODE_COMMON_NTWK_MAC_QUERY, sizeof(*req)); + + req->type = type; + if (permanent) { + req->permanent = 1; + } else { + req->if_id = cpu_to_le16((u16)if_handle); + req->permanent = 0; + } + + status = be_mbox_db_ring(ctrl); + if (!status) + memcpy(mac_addr, resp->mac.addr, ETH_ALEN); + + spin_unlock(&ctrl->cmd_lock); + return status; +} + +int be_cmd_pmac_add(struct be_ctrl_info *ctrl, u8 *mac_addr, + u32 if_id, u32 *pmac_id) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_pmac_add *req = embedded_payload(wrb); + int status; + + spin_lock(&ctrl->cmd_lock); + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON, + OPCODE_COMMON_NTWK_PMAC_ADD, sizeof(*req)); + + req->if_id = cpu_to_le32(if_id); + memcpy(req->mac_address, mac_addr, ETH_ALEN); + + status = be_mbox_db_ring(ctrl); + if (!status) { + struct be_cmd_resp_pmac_add *resp = embedded_payload(wrb); + *pmac_id = le32_to_cpu(resp->pmac_id); + } + + spin_unlock(&ctrl->cmd_lock); + return status; +} + +int be_cmd_pmac_del(struct be_ctrl_info *ctrl, u32 if_id, u32 pmac_id) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_pmac_del *req = embedded_payload(wrb); + int status; + + spin_lock(&ctrl->cmd_lock); + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON, + OPCODE_COMMON_NTWK_PMAC_DEL, sizeof(*req)); + + req->if_id = cpu_to_le32(if_id); + req->pmac_id = cpu_to_le32(pmac_id); + + status = be_mbox_db_ring(ctrl); + spin_unlock(&ctrl->cmd_lock); + + return status; +} + +int be_cmd_cq_create(struct be_ctrl_info *ctrl, + struct be_queue_info *cq, struct be_queue_info *eq, + bool sol_evts, bool no_delay, int coalesce_wm) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_cq_create *req = embedded_payload(wrb); + struct be_cmd_resp_cq_create *resp = embedded_payload(wrb); + struct be_dma_mem *q_mem = &cq->dma_mem; + void *ctxt = &req->context; + int status; + + spin_lock(&ctrl->cmd_lock); + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON, + OPCODE_COMMON_CQ_CREATE, sizeof(*req)); + + req->num_pages = cpu_to_le16(PAGES_4K_SPANNED(q_mem->va, q_mem->size)); + + AMAP_SET_BITS(struct amap_cq_context, coalescwm, ctxt, coalesce_wm); + AMAP_SET_BITS(struct amap_cq_context, nodelay, ctxt, no_delay); + AMAP_SET_BITS(struct amap_cq_context, count, ctxt, + __ilog2_u32(cq->len/256)); + AMAP_SET_BITS(struct amap_cq_context, valid, ctxt, 1); + AMAP_SET_BITS(struct amap_cq_context, solevent, ctxt, sol_evts); + AMAP_SET_BITS(struct amap_cq_context, eventable, ctxt, 1); + AMAP_SET_BITS(struct amap_cq_context, eqid, ctxt, eq->id); + AMAP_SET_BITS(struct amap_cq_context, armed, ctxt, 0); + AMAP_SET_BITS(struct amap_cq_context, func, ctxt, ctrl->pci_func); + be_dws_cpu_to_le(ctxt, sizeof(req->context)); + + be_cmd_page_addrs_prepare(req->pages, ARRAY_SIZE(req->pages), q_mem); + + status = be_mbox_db_ring(ctrl); + if (!status) { + cq->id = le16_to_cpu(resp->cq_id); + cq->created = true; + } + spin_unlock(&ctrl->cmd_lock); + + return status; +} + +int be_cmd_txq_create(struct be_ctrl_info *ctrl, + struct be_queue_info *txq, + struct be_queue_info *cq) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_eth_tx_create *req = embedded_payload(wrb); + struct be_dma_mem *q_mem = &txq->dma_mem; + void *ctxt = &req->context; + int status; + u32 len_encoded; + + spin_lock(&ctrl->cmd_lock); + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_ETH, OPCODE_ETH_TX_CREATE, + sizeof(*req)); + + req->num_pages = PAGES_4K_SPANNED(q_mem->va, q_mem->size); + req->ulp_num = BE_ULP1_NUM; + req->type = BE_ETH_TX_RING_TYPE_STANDARD; + + len_encoded = fls(txq->len); /* log2(len) + 1 */ + if (len_encoded == 16) + len_encoded = 0; + AMAP_SET_BITS(struct amap_tx_context, tx_ring_size, ctxt, len_encoded); + AMAP_SET_BITS(struct amap_tx_context, pci_func_id, ctxt, + ctrl->pci_func); + AMAP_SET_BITS(struct amap_tx_context, ctx_valid, ctxt, 1); + AMAP_SET_BITS(struct amap_tx_context, cq_id_send, ctxt, cq->id); + + be_dws_cpu_to_le(ctxt, sizeof(req->context)); + + be_cmd_page_addrs_prepare(req->pages, ARRAY_SIZE(req->pages), q_mem); + + status = be_mbox_db_ring(ctrl); + if (!status) { + struct be_cmd_resp_eth_tx_create *resp = embedded_payload(wrb); + txq->id = le16_to_cpu(resp->cid); + txq->created = true; + } + spin_unlock(&ctrl->cmd_lock); + + return status; +} + +int be_cmd_rxq_create(struct be_ctrl_info *ctrl, + struct be_queue_info *rxq, u16 cq_id, u16 frag_size, + u16 max_frame_size, u32 if_id, u32 rss) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_eth_rx_create *req = embedded_payload(wrb); + struct be_dma_mem *q_mem = &rxq->dma_mem; + int status; + + spin_lock(&ctrl->cmd_lock); + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_ETH, OPCODE_ETH_RX_CREATE, + sizeof(*req)); + + req->cq_id = cpu_to_le16(cq_id); + req->frag_size = fls(frag_size) - 1; + req->num_pages = 2; + be_cmd_page_addrs_prepare(req->pages, ARRAY_SIZE(req->pages), q_mem); + req->interface_id = cpu_to_le32(if_id); + req->max_frame_size = cpu_to_le16(max_frame_size); + req->rss_queue = cpu_to_le32(rss); + + status = be_mbox_db_ring(ctrl); + if (!status) { + struct be_cmd_resp_eth_rx_create *resp = embedded_payload(wrb); + rxq->id = le16_to_cpu(resp->id); + rxq->created = true; + } + spin_unlock(&ctrl->cmd_lock); + + return status; +} + +/* Generic destroyer function for all types of queues */ +int be_cmd_q_destroy(struct be_ctrl_info *ctrl, struct be_queue_info *q, + int queue_type) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_q_destroy *req = embedded_payload(wrb); + u8 subsys = 0, opcode = 0; + int status; + + spin_lock(&ctrl->cmd_lock); + + memset(wrb, 0, sizeof(*wrb)); + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + switch (queue_type) { + case QTYPE_EQ: + subsys = CMD_SUBSYSTEM_COMMON; + opcode = OPCODE_COMMON_EQ_DESTROY; + break; + case QTYPE_CQ: + subsys = CMD_SUBSYSTEM_COMMON; + opcode = OPCODE_COMMON_CQ_DESTROY; + break; + case QTYPE_TXQ: + subsys = CMD_SUBSYSTEM_ETH; + opcode = OPCODE_ETH_TX_DESTROY; + break; + case QTYPE_RXQ: + subsys = CMD_SUBSYSTEM_ETH; + opcode = OPCODE_ETH_RX_DESTROY; + break; + default: + printk(KERN_WARNING DRV_NAME ":bad Q type in Q destroy cmd\n"); + status = -1; + goto err; + } + be_cmd_hdr_prepare(&req->hdr, subsys, opcode, sizeof(*req)); + req->id = cpu_to_le16(q->id); + + status = be_mbox_db_ring(ctrl); +err: + spin_unlock(&ctrl->cmd_lock); + + return status; +} + +/* Create an rx filtering policy configuration on an i/f */ +int be_cmd_if_create(struct be_ctrl_info *ctrl, u32 flags, u8 *mac, + bool pmac_invalid, u32 *if_handle, u32 *pmac_id) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_if_create *req = embedded_payload(wrb); + int status; + + spin_lock(&ctrl->cmd_lock); + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON, + OPCODE_COMMON_NTWK_INTERFACE_CREATE, sizeof(*req)); + + req->capability_flags = cpu_to_le32(flags); + req->enable_flags = cpu_to_le32(flags); + if (!pmac_invalid) + memcpy(req->mac_addr, mac, ETH_ALEN); + + status = be_mbox_db_ring(ctrl); + if (!status) { + struct be_cmd_resp_if_create *resp = embedded_payload(wrb); + *if_handle = le32_to_cpu(resp->interface_id); + if (!pmac_invalid) + *pmac_id = le32_to_cpu(resp->pmac_id); + } + + spin_unlock(&ctrl->cmd_lock); + return status; +} + +int be_cmd_if_destroy(struct be_ctrl_info *ctrl, u32 interface_id) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_if_destroy *req = embedded_payload(wrb); + int status; + + spin_lock(&ctrl->cmd_lock); + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON, + OPCODE_COMMON_NTWK_INTERFACE_DESTROY, sizeof(*req)); + + req->interface_id = cpu_to_le32(interface_id); + status = be_mbox_db_ring(ctrl); + + spin_unlock(&ctrl->cmd_lock); + + return status; +} + +/* Get stats is a non embedded command: the request is not embedded inside + * WRB but is a separate dma memory block + */ +int be_cmd_get_stats(struct be_ctrl_info *ctrl, struct be_dma_mem *nonemb_cmd) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_get_stats *req = nonemb_cmd->va; + struct be_sge *sge = nonembedded_sgl(wrb); + int status; + + spin_lock(&ctrl->cmd_lock); + memset(wrb, 0, sizeof(*wrb)); + + memset(req, 0, sizeof(*req)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), false, 1); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_ETH, + OPCODE_ETH_GET_STATISTICS, sizeof(*req)); + sge->pa_hi = cpu_to_le32(upper_32_bits(nonemb_cmd->dma)); + sge->pa_lo = cpu_to_le32(nonemb_cmd->dma & 0xFFFFFFFF); + sge->len = cpu_to_le32(nonemb_cmd->size); + + status = be_mbox_db_ring(ctrl); + if (!status) { + struct be_cmd_resp_get_stats *resp = nonemb_cmd->va; + be_dws_le_to_cpu(&resp->hw_stats, sizeof(resp->hw_stats)); + } + + spin_unlock(&ctrl->cmd_lock); + return status; +} + +int be_cmd_link_status_query(struct be_ctrl_info *ctrl, + struct be_link_info *link) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_link_status *req = embedded_payload(wrb); + int status; + + spin_lock(&ctrl->cmd_lock); + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON, + OPCODE_COMMON_NTWK_LINK_STATUS_QUERY, sizeof(*req)); + + status = be_mbox_db_ring(ctrl); + if (!status) { + struct be_cmd_resp_link_status *resp = embedded_payload(wrb); + link->speed = resp->mac_speed; + link->duplex = resp->mac_duplex; + link->fault = resp->mac_fault; + } else { + link->speed = PHY_LINK_SPEED_ZERO; + } + + spin_unlock(&ctrl->cmd_lock); + return status; +} + +int be_cmd_get_fw_ver(struct be_ctrl_info *ctrl, char *fw_ver) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_get_fw_version *req = embedded_payload(wrb); + int status; + + spin_lock(&ctrl->cmd_lock); + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON, + OPCODE_COMMON_GET_FW_VERSION, sizeof(*req)); + + status = be_mbox_db_ring(ctrl); + if (!status) { + struct be_cmd_resp_get_fw_version *resp = embedded_payload(wrb); + strncpy(fw_ver, resp->firmware_version_string, FW_VER_LEN); + } + + spin_unlock(&ctrl->cmd_lock); + return status; +} + +/* set the EQ delay interval of an EQ to specified value */ +int be_cmd_modify_eqd(struct be_ctrl_info *ctrl, u32 eq_id, u32 eqd) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_modify_eq_delay *req = embedded_payload(wrb); + int status; + + spin_lock(&ctrl->cmd_lock); + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON, + OPCODE_COMMON_MODIFY_EQ_DELAY, sizeof(*req)); + + req->num_eq = cpu_to_le32(1); + req->delay[0].eq_id = cpu_to_le32(eq_id); + req->delay[0].phase = 0; + req->delay[0].delay_multiplier = cpu_to_le32(eqd); + + status = be_mbox_db_ring(ctrl); + + spin_unlock(&ctrl->cmd_lock); + return status; +} + +int be_cmd_vlan_config(struct be_ctrl_info *ctrl, u32 if_id, u16 *vtag_array, + u32 num, bool untagged, bool promiscuous) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_vlan_config *req = embedded_payload(wrb); + int status; + + spin_lock(&ctrl->cmd_lock); + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON, + OPCODE_COMMON_NTWK_VLAN_CONFIG, sizeof(*req)); + + req->interface_id = if_id; + req->promiscuous = promiscuous; + req->untagged = untagged; + req->num_vlan = num; + if (!promiscuous) { + memcpy(req->normal_vlan, vtag_array, + req->num_vlan * sizeof(vtag_array[0])); + } + + status = be_mbox_db_ring(ctrl); + + spin_unlock(&ctrl->cmd_lock); + return status; +} + +int be_cmd_promiscuous_config(struct be_ctrl_info *ctrl, u8 port_num, bool en) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_promiscuous_config *req = embedded_payload(wrb); + int status; + + spin_lock(&ctrl->cmd_lock); + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_ETH, + OPCODE_ETH_PROMISCUOUS, sizeof(*req)); + + if (port_num) + req->port1_promiscuous = en; + else + req->port0_promiscuous = en; + + status = be_mbox_db_ring(ctrl); + + spin_unlock(&ctrl->cmd_lock); + return status; +} + +int be_cmd_mcast_mac_set(struct be_ctrl_info *ctrl, u32 if_id, u8 *mac_table, + u32 num, bool promiscuous) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_mcast_mac_config *req = embedded_payload(wrb); + int status; + + spin_lock(&ctrl->cmd_lock); + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON, + OPCODE_COMMON_NTWK_MULTICAST_SET, sizeof(*req)); + + req->interface_id = if_id; + req->promiscuous = promiscuous; + if (!promiscuous) { + req->num_mac = cpu_to_le16(num); + if (num) + memcpy(req->mac, mac_table, ETH_ALEN * num); + } + + status = be_mbox_db_ring(ctrl); + + spin_unlock(&ctrl->cmd_lock); + return status; +} + +int be_cmd_set_flow_control(struct be_ctrl_info *ctrl, u32 tx_fc, u32 rx_fc) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_set_flow_control *req = embedded_payload(wrb); + int status; + + spin_lock(&ctrl->cmd_lock); + + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON, + OPCODE_COMMON_SET_FLOW_CONTROL, sizeof(*req)); + + req->tx_flow_control = cpu_to_le16((u16)tx_fc); + req->rx_flow_control = cpu_to_le16((u16)rx_fc); + + status = be_mbox_db_ring(ctrl); + + spin_unlock(&ctrl->cmd_lock); + return status; +} + +int be_cmd_get_flow_control(struct be_ctrl_info *ctrl, u32 *tx_fc, u32 *rx_fc) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_get_flow_control *req = embedded_payload(wrb); + int status; + + spin_lock(&ctrl->cmd_lock); + + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON, + OPCODE_COMMON_GET_FLOW_CONTROL, sizeof(*req)); + + status = be_mbox_db_ring(ctrl); + if (!status) { + struct be_cmd_resp_get_flow_control *resp = + embedded_payload(wrb); + *tx_fc = le16_to_cpu(resp->tx_flow_control); + *rx_fc = le16_to_cpu(resp->rx_flow_control); + } + + spin_unlock(&ctrl->cmd_lock); + return status; +} + +int be_cmd_query_fw_cfg(struct be_ctrl_info *ctrl, u32 *port_num) +{ + struct be_mcc_wrb *wrb = wrb_from_mbox(&ctrl->mbox_mem); + struct be_cmd_req_query_fw_cfg *req = embedded_payload(wrb); + int status; + + spin_lock(&ctrl->cmd_lock); + + memset(wrb, 0, sizeof(*wrb)); + + be_wrb_hdr_prepare(wrb, sizeof(*req), true, 0); + + be_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON, + OPCODE_COMMON_QUERY_FIRMWARE_CONFIG, sizeof(*req)); + + status = be_mbox_db_ring(ctrl); + if (!status) { + struct be_cmd_resp_query_fw_cfg *resp = embedded_payload(wrb); + *port_num = le32_to_cpu(resp->phys_port); + } + + spin_unlock(&ctrl->cmd_lock); + return status; +} diff --git a/drivers/net/benet/be_cmds.h b/drivers/net/benet/be_cmds.h new file mode 100644 index 000000000000..e499e2d5b8c3 --- /dev/null +++ b/drivers/net/benet/be_cmds.h @@ -0,0 +1,688 @@ +/* + * Copyright (C) 2005 - 2009 ServerEngines + * All rights reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. The full GNU General + * Public License is included in this distribution in the file called COPYING. + * + * Contact Information: + * linux-drivers@serverengines.com + * + * ServerEngines + * 209 N. Fair Oaks Ave + * Sunnyvale, CA 94085 + */ + +/* + * The driver sends configuration and managements command requests to the + * firmware in the BE. These requests are communicated to the processor + * using Work Request Blocks (WRBs) submitted to the MCC-WRB ring or via one + * WRB inside a MAILBOX. + * The commands are serviced by the ARM processor in the BladeEngine's MPU. + */ + +struct be_sge { + u32 pa_lo; + u32 pa_hi; + u32 len; +}; + +#define MCC_WRB_EMBEDDED_MASK 1 /* bit 0 of dword 0*/ +#define MCC_WRB_SGE_CNT_SHIFT 3 /* bits 3 - 7 of dword 0 */ +#define MCC_WRB_SGE_CNT_MASK 0x1F /* bits 3 - 7 of dword 0 */ +struct be_mcc_wrb { + u32 embedded; /* dword 0 */ + u32 payload_length; /* dword 1 */ + u32 tag0; /* dword 2 */ + u32 tag1; /* dword 3 */ + u32 rsvd; /* dword 4 */ + union { + u8 embedded_payload[236]; /* used by embedded cmds */ + struct be_sge sgl[19]; /* used by non-embedded cmds */ + } payload; +}; + +#define CQE_FLAGS_VALID_MASK (1 << 31) +#define CQE_FLAGS_ASYNC_MASK (1 << 30) +#define CQE_FLAGS_COMPLETED_MASK (1 << 28) +#define CQE_FLAGS_CONSUMED_MASK (1 << 27) + +/* Completion Status */ +enum { + MCC_STATUS_SUCCESS = 0x0, +/* The client does not have sufficient privileges to execute the command */ + MCC_STATUS_INSUFFICIENT_PRIVILEGES = 0x1, +/* A parameter in the command was invalid. */ + MCC_STATUS_INVALID_PARAMETER = 0x2, +/* There are insufficient chip resources to execute the command */ + MCC_STATUS_INSUFFICIENT_RESOURCES = 0x3, +/* The command is completing because the queue was getting flushed */ + MCC_STATUS_QUEUE_FLUSHING = 0x4, +/* The command is completing with a DMA error */ + MCC_STATUS_DMA_FAILED = 0x5 +}; + +#define CQE_STATUS_COMPL_MASK 0xFFFF +#define CQE_STATUS_COMPL_SHIFT 0 /* bits 0 - 15 */ +#define CQE_STATUS_EXTD_MASK 0xFFFF +#define CQE_STATUS_EXTD_SHIFT 0 /* bits 0 - 15 */ + +struct be_mcc_cq_entry { + u32 status; /* dword 0 */ + u32 tag0; /* dword 1 */ + u32 tag1; /* dword 2 */ + u32 flags; /* dword 3 */ +}; + +struct be_mcc_mailbox { + struct be_mcc_wrb wrb; + struct be_mcc_cq_entry cqe; +}; + +#define CMD_SUBSYSTEM_COMMON 0x1 +#define CMD_SUBSYSTEM_ETH 0x3 + +#define OPCODE_COMMON_NTWK_MAC_QUERY 1 +#define OPCODE_COMMON_NTWK_MAC_SET 2 +#define OPCODE_COMMON_NTWK_MULTICAST_SET 3 +#define OPCODE_COMMON_NTWK_VLAN_CONFIG 4 +#define OPCODE_COMMON_NTWK_LINK_STATUS_QUERY 5 +#define OPCODE_COMMON_CQ_CREATE 12 +#define OPCODE_COMMON_EQ_CREATE 13 +#define OPCODE_COMMON_MCC_CREATE 21 +#define OPCODE_COMMON_NTWK_RX_FILTER 34 +#define OPCODE_COMMON_GET_FW_VERSION 35 +#define OPCODE_COMMON_SET_FLOW_CONTROL 36 +#define OPCODE_COMMON_GET_FLOW_CONTROL 37 +#define OPCODE_COMMON_SET_FRAME_SIZE 39 +#define OPCODE_COMMON_MODIFY_EQ_DELAY 41 +#define OPCODE_COMMON_FIRMWARE_CONFIG 42 +#define OPCODE_COMMON_NTWK_INTERFACE_CREATE 50 +#define OPCODE_COMMON_NTWK_INTERFACE_DESTROY 51 +#define OPCODE_COMMON_CQ_DESTROY 54 +#define OPCODE_COMMON_EQ_DESTROY 55 +#define OPCODE_COMMON_QUERY_FIRMWARE_CONFIG 58 +#define OPCODE_COMMON_NTWK_PMAC_ADD 59 +#define OPCODE_COMMON_NTWK_PMAC_DEL 60 + +#define OPCODE_ETH_ACPI_CONFIG 2 +#define OPCODE_ETH_PROMISCUOUS 3 +#define OPCODE_ETH_GET_STATISTICS 4 +#define OPCODE_ETH_TX_CREATE 7 +#define OPCODE_ETH_RX_CREATE 8 +#define OPCODE_ETH_TX_DESTROY 9 +#define OPCODE_ETH_RX_DESTROY 10 + +struct be_cmd_req_hdr { + u8 opcode; /* dword 0 */ + u8 subsystem; /* dword 0 */ + u8 port_number; /* dword 0 */ + u8 domain; /* dword 0 */ + u32 timeout; /* dword 1 */ + u32 request_length; /* dword 2 */ + u32 rsvd; /* dword 3 */ +}; + +#define RESP_HDR_INFO_OPCODE_SHIFT 0 /* bits 0 - 7 */ +#define RESP_HDR_INFO_SUBSYS_SHIFT 8 /* bits 8 - 15 */ +struct be_cmd_resp_hdr { + u32 info; /* dword 0 */ + u32 status; /* dword 1 */ + u32 response_length; /* dword 2 */ + u32 actual_resp_len; /* dword 3 */ +}; + +struct phys_addr { + u32 lo; + u32 hi; +}; + +/************************** + * BE Command definitions * + **************************/ + +/* Pseudo amap definition in which each bit of the actual structure is defined + * as a byte: used to calculate offset/shift/mask of each field */ +struct amap_eq_context { + u8 cidx[13]; /* dword 0*/ + u8 rsvd0[3]; /* dword 0*/ + u8 epidx[13]; /* dword 0*/ + u8 valid; /* dword 0*/ + u8 rsvd1; /* dword 0*/ + u8 size; /* dword 0*/ + u8 pidx[13]; /* dword 1*/ + u8 rsvd2[3]; /* dword 1*/ + u8 pd[10]; /* dword 1*/ + u8 count[3]; /* dword 1*/ + u8 solevent; /* dword 1*/ + u8 stalled; /* dword 1*/ + u8 armed; /* dword 1*/ + u8 rsvd3[4]; /* dword 2*/ + u8 func[8]; /* dword 2*/ + u8 rsvd4; /* dword 2*/ + u8 delaymult[10]; /* dword 2*/ + u8 rsvd5[2]; /* dword 2*/ + u8 phase[2]; /* dword 2*/ + u8 nodelay; /* dword 2*/ + u8 rsvd6[4]; /* dword 2*/ + u8 rsvd7[32]; /* dword 3*/ +} __packed; + +struct be_cmd_req_eq_create { + struct be_cmd_req_hdr hdr; + u16 num_pages; /* sword */ + u16 rsvd0; /* sword */ + u8 context[sizeof(struct amap_eq_context) / 8]; + struct phys_addr pages[8]; +} __packed; + +struct be_cmd_resp_eq_create { + struct be_cmd_resp_hdr resp_hdr; + u16 eq_id; /* sword */ + u16 rsvd0; /* sword */ +} __packed; + +/******************** Mac query ***************************/ +enum { + MAC_ADDRESS_TYPE_STORAGE = 0x0, + MAC_ADDRESS_TYPE_NETWORK = 0x1, + MAC_ADDRESS_TYPE_PD = 0x2, + MAC_ADDRESS_TYPE_MANAGEMENT = 0x3 +}; + +struct mac_addr { + u16 size_of_struct; + u8 addr[ETH_ALEN]; +} __packed; + +struct be_cmd_req_mac_query { + struct be_cmd_req_hdr hdr; + u8 type; + u8 permanent; + u16 if_id; +} __packed; + +struct be_cmd_resp_mac_query { + struct be_cmd_resp_hdr hdr; + struct mac_addr mac; +}; + +/******************** PMac Add ***************************/ +struct be_cmd_req_pmac_add { + struct be_cmd_req_hdr hdr; + u32 if_id; + u8 mac_address[ETH_ALEN]; + u8 rsvd0[2]; +} __packed; + +struct be_cmd_resp_pmac_add { + struct be_cmd_resp_hdr hdr; + u32 pmac_id; +}; + +/******************** PMac Del ***************************/ +struct be_cmd_req_pmac_del { + struct be_cmd_req_hdr hdr; + u32 if_id; + u32 pmac_id; +}; + +/******************** Create CQ ***************************/ +/* Pseudo amap definition in which each bit of the actual structure is defined + * as a byte: used to calculate offset/shift/mask of each field */ +struct amap_cq_context { + u8 cidx[11]; /* dword 0*/ + u8 rsvd0; /* dword 0*/ + u8 coalescwm[2]; /* dword 0*/ + u8 nodelay; /* dword 0*/ + u8 epidx[11]; /* dword 0*/ + u8 rsvd1; /* dword 0*/ + u8 count[2]; /* dword 0*/ + u8 valid; /* dword 0*/ + u8 solevent; /* dword 0*/ + u8 eventable; /* dword 0*/ + u8 pidx[11]; /* dword 1*/ + u8 rsvd2; /* dword 1*/ + u8 pd[10]; /* dword 1*/ + u8 eqid[8]; /* dword 1*/ + u8 stalled; /* dword 1*/ + u8 armed; /* dword 1*/ + u8 rsvd3[4]; /* dword 2*/ + u8 func[8]; /* dword 2*/ + u8 rsvd4[20]; /* dword 2*/ + u8 rsvd5[32]; /* dword 3*/ +} __packed; + +struct be_cmd_req_cq_create { + struct be_cmd_req_hdr hdr; + u16 num_pages; + u16 rsvd0; + u8 context[sizeof(struct amap_cq_context) / 8]; + struct phys_addr pages[8]; +} __packed; + +struct be_cmd_resp_cq_create { + struct be_cmd_resp_hdr hdr; + u16 cq_id; + u16 rsvd0; +} __packed; + +/******************** Create TxQ ***************************/ +#define BE_ETH_TX_RING_TYPE_STANDARD 2 +#define BE_ULP1_NUM 1 + +/* Pseudo amap definition in which each bit of the actual structure is defined + * as a byte: used to calculate offset/shift/mask of each field */ +struct amap_tx_context { + u8 rsvd0[16]; /* dword 0 */ + u8 tx_ring_size[4]; /* dword 0 */ + u8 rsvd1[26]; /* dword 0 */ + u8 pci_func_id[8]; /* dword 1 */ + u8 rsvd2[9]; /* dword 1 */ + u8 ctx_valid; /* dword 1 */ + u8 cq_id_send[16]; /* dword 2 */ + u8 rsvd3[16]; /* dword 2 */ + u8 rsvd4[32]; /* dword 3 */ + u8 rsvd5[32]; /* dword 4 */ + u8 rsvd6[32]; /* dword 5 */ + u8 rsvd7[32]; /* dword 6 */ + u8 rsvd8[32]; /* dword 7 */ + u8 rsvd9[32]; /* dword 8 */ + u8 rsvd10[32]; /* dword 9 */ + u8 rsvd11[32]; /* dword 10 */ + u8 rsvd12[32]; /* dword 11 */ + u8 rsvd13[32]; /* dword 12 */ + u8 rsvd14[32]; /* dword 13 */ + u8 rsvd15[32]; /* dword 14 */ + u8 rsvd16[32]; /* dword 15 */ +} __packed; + +struct be_cmd_req_eth_tx_create { + struct be_cmd_req_hdr hdr; + u8 num_pages; + u8 ulp_num; + u8 type; + u8 bound_port; + u8 context[sizeof(struct amap_tx_context) / 8]; + struct phys_addr pages[8]; +} __packed; + +struct be_cmd_resp_eth_tx_create { + struct be_cmd_resp_hdr hdr; + u16 cid; + u16 rsvd0; +} __packed; + +/******************** Create RxQ ***************************/ +struct be_cmd_req_eth_rx_create { + struct be_cmd_req_hdr hdr; + u16 cq_id; + u8 frag_size; + u8 num_pages; + struct phys_addr pages[2]; + u32 interface_id; + u16 max_frame_size; + u16 rsvd0; + u32 rss_queue; +} __packed; + +struct be_cmd_resp_eth_rx_create { + struct be_cmd_resp_hdr hdr; + u16 id; + u8 cpu_id; + u8 rsvd0; +} __packed; + +/******************** Q Destroy ***************************/ +/* Type of Queue to be destroyed */ +enum { + QTYPE_EQ = 1, + QTYPE_CQ, + QTYPE_TXQ, + QTYPE_RXQ +}; + +struct be_cmd_req_q_destroy { + struct be_cmd_req_hdr hdr; + u16 id; + u16 bypass_flush; /* valid only for rx q destroy */ +} __packed; + +/************ I/f Create (it's actually I/f Config Create)**********/ + +/* Capability flags for the i/f */ +enum be_if_flags { + BE_IF_FLAGS_RSS = 0x4, + BE_IF_FLAGS_PROMISCUOUS = 0x8, + BE_IF_FLAGS_BROADCAST = 0x10, + BE_IF_FLAGS_UNTAGGED = 0x20, + BE_IF_FLAGS_ULP = 0x40, + BE_IF_FLAGS_VLAN_PROMISCUOUS = 0x80, + BE_IF_FLAGS_VLAN = 0x100, + BE_IF_FLAGS_MCAST_PROMISCUOUS = 0x200, + BE_IF_FLAGS_PASS_L2_ERRORS = 0x400, + BE_IF_FLAGS_PASS_L3L4_ERRORS = 0x800 +}; + +/* An RX interface is an object with one or more MAC addresses and + * filtering capabilities. */ +struct be_cmd_req_if_create { + struct be_cmd_req_hdr hdr; + u32 version; /* ignore currntly */ + u32 capability_flags; + u32 enable_flags; + u8 mac_addr[ETH_ALEN]; + u8 rsvd0; + u8 pmac_invalid; /* if set, don't attach the mac addr to the i/f */ + u32 vlan_tag; /* not used currently */ +} __packed; + +struct be_cmd_resp_if_create { + struct be_cmd_resp_hdr hdr; + u32 interface_id; + u32 pmac_id; +}; + +/****** I/f Destroy(it's actually I/f Config Destroy )**********/ +struct be_cmd_req_if_destroy { + struct be_cmd_req_hdr hdr; + u32 interface_id; +}; + +/*************** HW Stats Get **********************************/ +struct be_port_rxf_stats { + u32 rx_bytes_lsd; /* dword 0*/ + u32 rx_bytes_msd; /* dword 1*/ + u32 rx_total_frames; /* dword 2*/ + u32 rx_unicast_frames; /* dword 3*/ + u32 rx_multicast_frames; /* dword 4*/ + u32 rx_broadcast_frames; /* dword 5*/ + u32 rx_crc_errors; /* dword 6*/ + u32 rx_alignment_symbol_errors; /* dword 7*/ + u32 rx_pause_frames; /* dword 8*/ + u32 rx_control_frames; /* dword 9*/ + u32 rx_in_range_errors; /* dword 10*/ + u32 rx_out_range_errors; /* dword 11*/ + u32 rx_frame_too_long; /* dword 12*/ + u32 rx_address_match_errors; /* dword 13*/ + u32 rx_vlan_mismatch; /* dword 14*/ + u32 rx_dropped_too_small; /* dword 15*/ + u32 rx_dropped_too_short; /* dword 16*/ + u32 rx_dropped_header_too_small; /* dword 17*/ + u32 rx_dropped_tcp_length; /* dword 18*/ + u32 rx_dropped_runt; /* dword 19*/ + u32 rx_64_byte_packets; /* dword 20*/ + u32 rx_65_127_byte_packets; /* dword 21*/ + u32 rx_128_256_byte_packets; /* dword 22*/ + u32 rx_256_511_byte_packets; /* dword 23*/ + u32 rx_512_1023_byte_packets; /* dword 24*/ + u32 rx_1024_1518_byte_packets; /* dword 25*/ + u32 rx_1519_2047_byte_packets; /* dword 26*/ + u32 rx_2048_4095_byte_packets; /* dword 27*/ + u32 rx_4096_8191_byte_packets; /* dword 28*/ + u32 rx_8192_9216_byte_packets; /* dword 29*/ + u32 rx_ip_checksum_errs; /* dword 30*/ + u32 rx_tcp_checksum_errs; /* dword 31*/ + u32 rx_udp_checksum_errs; /* dword 32*/ + u32 rx_non_rss_packets; /* dword 33*/ + u32 rx_ipv4_packets; /* dword 34*/ + u32 rx_ipv6_packets; /* dword 35*/ + u32 rx_ipv4_bytes_lsd; /* dword 36*/ + u32 rx_ipv4_bytes_msd; /* dword 37*/ + u32 rx_ipv6_bytes_lsd; /* dword 38*/ + u32 rx_ipv6_bytes_msd; /* dword 39*/ + u32 rx_chute1_packets; /* dword 40*/ + u32 rx_chute2_packets; /* dword 41*/ + u32 rx_chute3_packets; /* dword 42*/ + u32 rx_management_packets; /* dword 43*/ + u32 rx_switched_unicast_packets; /* dword 44*/ + u32 rx_switched_multicast_packets; /* dword 45*/ + u32 rx_switched_broadcast_packets; /* dword 46*/ + u32 tx_bytes_lsd; /* dword 47*/ + u32 tx_bytes_msd; /* dword 48*/ + u32 tx_unicastframes; /* dword 49*/ + u32 tx_multicastframes; /* dword 50*/ + u32 tx_broadcastframes; /* dword 51*/ + u32 tx_pauseframes; /* dword 52*/ + u32 tx_controlframes; /* dword 53*/ + u32 tx_64_byte_packets; /* dword 54*/ + u32 tx_65_127_byte_packets; /* dword 55*/ + u32 tx_128_256_byte_packets; /* dword 56*/ + u32 tx_256_511_byte_packets; /* dword 57*/ + u32 tx_512_1023_byte_packets; /* dword 58*/ + u32 tx_1024_1518_byte_packets; /* dword 59*/ + u32 tx_1519_2047_byte_packets; /* dword 60*/ + u32 tx_2048_4095_byte_packets; /* dword 61*/ + u32 tx_4096_8191_byte_packets; /* dword 62*/ + u32 tx_8192_9216_byte_packets; /* dword 63*/ + u32 rx_fifo_overflow; /* dword 64*/ + u32 rx_input_fifo_overflow; /* dword 65*/ +}; + +struct be_rxf_stats { + struct be_port_rxf_stats port[2]; + u32 rx_drops_no_pbuf; /* dword 132*/ + u32 rx_drops_no_txpb; /* dword 133*/ + u32 rx_drops_no_erx_descr; /* dword 134*/ + u32 rx_drops_no_tpre_descr; /* dword 135*/ + u32 management_rx_port_packets; /* dword 136*/ + u32 management_rx_port_bytes; /* dword 137*/ + u32 management_rx_port_pause_frames; /* dword 138*/ + u32 management_rx_port_errors; /* dword 139*/ + u32 management_tx_port_packets; /* dword 140*/ + u32 management_tx_port_bytes; /* dword 141*/ + u32 management_tx_port_pause; /* dword 142*/ + u32 management_rx_port_rxfifo_overflow; /* dword 143*/ + u32 rx_drops_too_many_frags; /* dword 144*/ + u32 rx_drops_invalid_ring; /* dword 145*/ + u32 forwarded_packets; /* dword 146*/ + u32 rx_drops_mtu; /* dword 147*/ + u32 rsvd0[15]; +}; + +struct be_erx_stats { + u32 rx_drops_no_fragments[44]; /* dwordS 0 to 43*/ + u32 debug_wdma_sent_hold; /* dword 44*/ + u32 debug_wdma_pbfree_sent_hold; /* dword 45*/ + u32 debug_wdma_zerobyte_pbfree_sent_hold; /* dword 46*/ + u32 debug_pmem_pbuf_dealloc; /* dword 47*/ +}; + +struct be_hw_stats { + struct be_rxf_stats rxf; + u32 rsvd[48]; + struct be_erx_stats erx; +}; + +struct be_cmd_req_get_stats { + struct be_cmd_req_hdr hdr; + u8 rsvd[sizeof(struct be_hw_stats)]; +}; + +struct be_cmd_resp_get_stats { + struct be_cmd_resp_hdr hdr; + struct be_hw_stats hw_stats; +}; + +struct be_cmd_req_vlan_config { + struct be_cmd_req_hdr hdr; + u8 interface_id; + u8 promiscuous; + u8 untagged; + u8 num_vlan; + u16 normal_vlan[64]; +} __packed; + +struct be_cmd_req_promiscuous_config { + struct be_cmd_req_hdr hdr; + u8 port0_promiscuous; + u8 port1_promiscuous; + u16 rsvd0; +} __packed; + +struct macaddr { + u8 byte[ETH_ALEN]; +}; + +struct be_cmd_req_mcast_mac_config { + struct be_cmd_req_hdr hdr; + u16 num_mac; + u8 promiscuous; + u8 interface_id; + struct macaddr mac[32]; +} __packed; + +static inline struct be_hw_stats * +hw_stats_from_cmd(struct be_cmd_resp_get_stats *cmd) +{ + return &cmd->hw_stats; +} + +/******************** Link Status Query *******************/ +struct be_cmd_req_link_status { + struct be_cmd_req_hdr hdr; + u32 rsvd; +}; + +struct be_link_info { + u8 duplex; + u8 speed; + u8 fault; +}; + +enum { + PHY_LINK_DUPLEX_NONE = 0x0, + PHY_LINK_DUPLEX_HALF = 0x1, + PHY_LINK_DUPLEX_FULL = 0x2 +}; + +enum { + PHY_LINK_SPEED_ZERO = 0x0, /* => No link */ + PHY_LINK_SPEED_10MBPS = 0x1, + PHY_LINK_SPEED_100MBPS = 0x2, + PHY_LINK_SPEED_1GBPS = 0x3, + PHY_LINK_SPEED_10GBPS = 0x4 +}; + +struct be_cmd_resp_link_status { + struct be_cmd_resp_hdr hdr; + u8 physical_port; + u8 mac_duplex; + u8 mac_speed; + u8 mac_fault; + u8 mgmt_mac_duplex; + u8 mgmt_mac_speed; + u16 rsvd0; +} __packed; + +/******************** Get FW Version *******************/ +#define FW_VER_LEN 32 +struct be_cmd_req_get_fw_version { + struct be_cmd_req_hdr hdr; + u8 rsvd0[FW_VER_LEN]; + u8 rsvd1[FW_VER_LEN]; +} __packed; + +struct be_cmd_resp_get_fw_version { + struct be_cmd_resp_hdr hdr; + u8 firmware_version_string[FW_VER_LEN]; + u8 fw_on_flash_version_string[FW_VER_LEN]; +} __packed; + +/******************** Set Flow Contrl *******************/ +struct be_cmd_req_set_flow_control { + struct be_cmd_req_hdr hdr; + u16 tx_flow_control; + u16 rx_flow_control; +} __packed; + +/******************** Get Flow Contrl *******************/ +struct be_cmd_req_get_flow_control { + struct be_cmd_req_hdr hdr; + u32 rsvd; +}; + +struct be_cmd_resp_get_flow_control { + struct be_cmd_resp_hdr hdr; + u16 tx_flow_control; + u16 rx_flow_control; +} __packed; + +/******************** Modify EQ Delay *******************/ +struct be_cmd_req_modify_eq_delay { + struct be_cmd_req_hdr hdr; + u32 num_eq; + struct { + u32 eq_id; + u32 phase; + u32 delay_multiplier; + } delay[8]; +} __packed; + +struct be_cmd_resp_modify_eq_delay { + struct be_cmd_resp_hdr hdr; + u32 rsvd0; +} __packed; + +/******************** Get FW Config *******************/ +struct be_cmd_req_query_fw_cfg { + struct be_cmd_req_hdr hdr; + u32 rsvd[30]; +}; + +struct be_cmd_resp_query_fw_cfg { + struct be_cmd_resp_hdr hdr; + u32 be_config_number; + u32 asic_revision; + u32 phys_port; + u32 function_mode; + u32 rsvd[26]; +}; + +extern int be_pci_fnum_get(struct be_ctrl_info *ctrl); +extern int be_cmd_POST(struct be_ctrl_info *ctrl); +extern int be_cmd_mac_addr_query(struct be_ctrl_info *ctrl, u8 *mac_addr, + u8 type, bool permanent, u32 if_handle); +extern int be_cmd_pmac_add(struct be_ctrl_info *ctrl, u8 *mac_addr, + u32 if_id, u32 *pmac_id); +extern int be_cmd_pmac_del(struct be_ctrl_info *ctrl, u32 if_id, u32 pmac_id); +extern int be_cmd_if_create(struct be_ctrl_info *ctrl, u32 if_flags, u8 *mac, + bool pmac_invalid, u32 *if_handle, u32 *pmac_id); +extern int be_cmd_if_destroy(struct be_ctrl_info *ctrl, u32 if_handle); +extern int be_cmd_eq_create(struct be_ctrl_info *ctrl, + struct be_queue_info *eq, int eq_delay); +extern int be_cmd_cq_create(struct be_ctrl_info *ctrl, + struct be_queue_info *cq, struct be_queue_info *eq, + bool sol_evts, bool no_delay, + int num_cqe_dma_coalesce); +extern int be_cmd_txq_create(struct be_ctrl_info *ctrl, + struct be_queue_info *txq, + struct be_queue_info *cq); +extern int be_cmd_rxq_create(struct be_ctrl_info *ctrl, + struct be_queue_info *rxq, u16 cq_id, + u16 frag_size, u16 max_frame_size, u32 if_id, + u32 rss); +extern int be_cmd_q_destroy(struct be_ctrl_info *ctrl, struct be_queue_info *q, + int type); +extern int be_cmd_link_status_query(struct be_ctrl_info *ctrl, + struct be_link_info *link); +extern int be_cmd_reset(struct be_ctrl_info *ctrl); +extern int be_cmd_get_stats(struct be_ctrl_info *ctrl, + struct be_dma_mem *nonemb_cmd); +extern int be_cmd_get_fw_ver(struct be_ctrl_info *ctrl, char *fw_ver); + +extern int be_cmd_modify_eqd(struct be_ctrl_info *ctrl, u32 eq_id, u32 eqd); +extern int be_cmd_vlan_config(struct be_ctrl_info *ctrl, u32 if_id, + u16 *vtag_array, u32 num, bool untagged, + bool promiscuous); +extern int be_cmd_promiscuous_config(struct be_ctrl_info *ctrl, + u8 port_num, bool en); +extern int be_cmd_mcast_mac_set(struct be_ctrl_info *ctrl, u32 if_id, + u8 *mac_table, u32 num, bool promiscuous); +extern int be_cmd_set_flow_control(struct be_ctrl_info *ctrl, + u32 tx_fc, u32 rx_fc); +extern int be_cmd_get_flow_control(struct be_ctrl_info *ctrl, + u32 *tx_fc, u32 *rx_fc); +extern int be_cmd_query_fw_cfg(struct be_ctrl_info *ctrl, u32 *port_num); diff --git a/drivers/net/benet/be_ethtool.c b/drivers/net/benet/be_ethtool.c new file mode 100644 index 000000000000..04f4b73fa8d8 --- /dev/null +++ b/drivers/net/benet/be_ethtool.c @@ -0,0 +1,362 @@ +/* + * Copyright (C) 2005 - 2009 ServerEngines + * All rights reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. The full GNU General + * Public License is included in this distribution in the file called COPYING. + * + * Contact Information: + * linux-drivers@serverengines.com + * + * ServerEngines + * 209 N. Fair Oaks Ave + * Sunnyvale, CA 94085 + */ + +#include "be.h" +#include <linux/ethtool.h> + +struct be_ethtool_stat { + char desc[ETH_GSTRING_LEN]; + int type; + int size; + int offset; +}; + +enum {NETSTAT, PORTSTAT, MISCSTAT, DRVSTAT, ERXSTAT}; +#define FIELDINFO(_struct, field) FIELD_SIZEOF(_struct, field), \ + offsetof(_struct, field) +#define NETSTAT_INFO(field) #field, NETSTAT,\ + FIELDINFO(struct net_device_stats,\ + field) +#define DRVSTAT_INFO(field) #field, DRVSTAT,\ + FIELDINFO(struct be_drvr_stats, field) +#define MISCSTAT_INFO(field) #field, MISCSTAT,\ + FIELDINFO(struct be_rxf_stats, field) +#define PORTSTAT_INFO(field) #field, PORTSTAT,\ + FIELDINFO(struct be_port_rxf_stats, \ + field) +#define ERXSTAT_INFO(field) #field, ERXSTAT,\ + FIELDINFO(struct be_erx_stats, field) + +static const struct be_ethtool_stat et_stats[] = { + {NETSTAT_INFO(rx_packets)}, + {NETSTAT_INFO(tx_packets)}, + {NETSTAT_INFO(rx_bytes)}, + {NETSTAT_INFO(tx_bytes)}, + {NETSTAT_INFO(rx_errors)}, + {NETSTAT_INFO(tx_errors)}, + {NETSTAT_INFO(rx_dropped)}, + {NETSTAT_INFO(tx_dropped)}, + {DRVSTAT_INFO(be_tx_reqs)}, + {DRVSTAT_INFO(be_tx_stops)}, + {DRVSTAT_INFO(be_fwd_reqs)}, + {DRVSTAT_INFO(be_tx_wrbs)}, + {DRVSTAT_INFO(be_polls)}, + {DRVSTAT_INFO(be_tx_events)}, + {DRVSTAT_INFO(be_rx_events)}, + {DRVSTAT_INFO(be_tx_compl)}, + {DRVSTAT_INFO(be_rx_compl)}, + {DRVSTAT_INFO(be_ethrx_post_fail)}, + {DRVSTAT_INFO(be_802_3_dropped_frames)}, + {DRVSTAT_INFO(be_802_3_malformed_frames)}, + {DRVSTAT_INFO(be_tx_rate)}, + {DRVSTAT_INFO(be_rx_rate)}, + {PORTSTAT_INFO(rx_unicast_frames)}, + {PORTSTAT_INFO(rx_multicast_frames)}, + {PORTSTAT_INFO(rx_broadcast_frames)}, + {PORTSTAT_INFO(rx_crc_errors)}, + {PORTSTAT_INFO(rx_alignment_symbol_errors)}, + {PORTSTAT_INFO(rx_pause_frames)}, + {PORTSTAT_INFO(rx_control_frames)}, + {PORTSTAT_INFO(rx_in_range_errors)}, + {PORTSTAT_INFO(rx_out_range_errors)}, + {PORTSTAT_INFO(rx_frame_too_long)}, + {PORTSTAT_INFO(rx_address_match_errors)}, + {PORTSTAT_INFO(rx_vlan_mismatch)}, + {PORTSTAT_INFO(rx_dropped_too_small)}, + {PORTSTAT_INFO(rx_dropped_too_short)}, + {PORTSTAT_INFO(rx_dropped_header_too_small)}, + {PORTSTAT_INFO(rx_dropped_tcp_length)}, + {PORTSTAT_INFO(rx_dropped_runt)}, + {PORTSTAT_INFO(rx_fifo_overflow)}, + {PORTSTAT_INFO(rx_input_fifo_overflow)}, + {PORTSTAT_INFO(rx_ip_checksum_errs)}, + {PORTSTAT_INFO(rx_tcp_checksum_errs)}, + {PORTSTAT_INFO(rx_udp_checksum_errs)}, + {PORTSTAT_INFO(rx_non_rss_packets)}, + {PORTSTAT_INFO(rx_ipv4_packets)}, + {PORTSTAT_INFO(rx_ipv6_packets)}, + {PORTSTAT_INFO(tx_unicastframes)}, + {PORTSTAT_INFO(tx_multicastframes)}, + {PORTSTAT_INFO(tx_broadcastframes)}, + {PORTSTAT_INFO(tx_pauseframes)}, + {PORTSTAT_INFO(tx_controlframes)}, + {MISCSTAT_INFO(rx_drops_no_pbuf)}, + {MISCSTAT_INFO(rx_drops_no_txpb)}, + {MISCSTAT_INFO(rx_drops_no_erx_descr)}, + {MISCSTAT_INFO(rx_drops_no_tpre_descr)}, + {MISCSTAT_INFO(rx_drops_too_many_frags)}, + {MISCSTAT_INFO(rx_drops_invalid_ring)}, + {MISCSTAT_INFO(forwarded_packets)}, + {MISCSTAT_INFO(rx_drops_mtu)}, + {ERXSTAT_INFO(rx_drops_no_fragments)}, +}; +#define ETHTOOL_STATS_NUM ARRAY_SIZE(et_stats) + +static void +be_get_drvinfo(struct net_device *netdev, struct ethtool_drvinfo *drvinfo) +{ + struct be_adapter *adapter = netdev_priv(netdev); + + strcpy(drvinfo->driver, DRV_NAME); + strcpy(drvinfo->version, DRV_VER); + strncpy(drvinfo->fw_version, adapter->fw_ver, FW_VER_LEN); + strcpy(drvinfo->bus_info, pci_name(adapter->pdev)); + drvinfo->testinfo_len = 0; + drvinfo->regdump_len = 0; + drvinfo->eedump_len = 0; +} + +static int +be_get_coalesce(struct net_device *netdev, struct ethtool_coalesce *coalesce) +{ + struct be_adapter *adapter = netdev_priv(netdev); + struct be_eq_obj *rx_eq = &adapter->rx_eq; + struct be_eq_obj *tx_eq = &adapter->tx_eq; + + coalesce->rx_max_coalesced_frames = adapter->max_rx_coal; + + coalesce->rx_coalesce_usecs = rx_eq->cur_eqd; + coalesce->rx_coalesce_usecs_high = rx_eq->max_eqd; + coalesce->rx_coalesce_usecs_low = rx_eq->min_eqd; + + coalesce->tx_coalesce_usecs = tx_eq->cur_eqd; + coalesce->tx_coalesce_usecs_high = tx_eq->max_eqd; + coalesce->tx_coalesce_usecs_low = tx_eq->min_eqd; + + coalesce->use_adaptive_rx_coalesce = rx_eq->enable_aic; + coalesce->use_adaptive_tx_coalesce = tx_eq->enable_aic; + + return 0; +} + +/* + * This routine is used to set interrup coalescing delay *as well as* + * the number of pkts to coalesce for LRO. + */ +static int +be_set_coalesce(struct net_device *netdev, struct ethtool_coalesce *coalesce) +{ + struct be_adapter *adapter = netdev_priv(netdev); + struct be_ctrl_info *ctrl = &adapter->ctrl; + struct be_eq_obj *rx_eq = &adapter->rx_eq; + struct be_eq_obj *tx_eq = &adapter->tx_eq; + u32 tx_max, tx_min, tx_cur; + u32 rx_max, rx_min, rx_cur; + int status = 0; + + if (coalesce->use_adaptive_tx_coalesce == 1) + return -EINVAL; + + adapter->max_rx_coal = coalesce->rx_max_coalesced_frames; + if (adapter->max_rx_coal > MAX_SKB_FRAGS) + adapter->max_rx_coal = MAX_SKB_FRAGS - 1; + + /* if AIC is being turned on now, start with an EQD of 0 */ + if (rx_eq->enable_aic == 0 && + coalesce->use_adaptive_rx_coalesce == 1) { + rx_eq->cur_eqd = 0; + } + rx_eq->enable_aic = coalesce->use_adaptive_rx_coalesce; + + rx_max = coalesce->rx_coalesce_usecs_high; + rx_min = coalesce->rx_coalesce_usecs_low; + rx_cur = coalesce->rx_coalesce_usecs; + + tx_max = coalesce->tx_coalesce_usecs_high; + tx_min = coalesce->tx_coalesce_usecs_low; + tx_cur = coalesce->tx_coalesce_usecs; + + if (tx_cur > BE_MAX_EQD) + tx_cur = BE_MAX_EQD; + if (tx_eq->cur_eqd != tx_cur) { + status = be_cmd_modify_eqd(ctrl, tx_eq->q.id, tx_cur); + if (!status) + tx_eq->cur_eqd = tx_cur; + } + + if (rx_eq->enable_aic) { + if (rx_max > BE_MAX_EQD) + rx_max = BE_MAX_EQD; + if (rx_min > rx_max) + rx_min = rx_max; + rx_eq->max_eqd = rx_max; + rx_eq->min_eqd = rx_min; + if (rx_eq->cur_eqd > rx_max) + rx_eq->cur_eqd = rx_max; + if (rx_eq->cur_eqd < rx_min) + rx_eq->cur_eqd = rx_min; + } else { + if (rx_cur > BE_MAX_EQD) + rx_cur = BE_MAX_EQD; + if (rx_eq->cur_eqd != rx_cur) { + status = be_cmd_modify_eqd(ctrl, rx_eq->q.id, rx_cur); + if (!status) + rx_eq->cur_eqd = rx_cur; + } + } + return 0; +} + +static u32 be_get_rx_csum(struct net_device *netdev) +{ + struct be_adapter *adapter = netdev_priv(netdev); + + return adapter->rx_csum; +} + +static int be_set_rx_csum(struct net_device *netdev, uint32_t data) +{ + struct be_adapter *adapter = netdev_priv(netdev); + + if (data) + adapter->rx_csum = true; + else + adapter->rx_csum = false; + + return 0; +} + +static void +be_get_ethtool_stats(struct net_device *netdev, + struct ethtool_stats *stats, uint64_t *data) +{ + struct be_adapter *adapter = netdev_priv(netdev); + struct be_drvr_stats *drvr_stats = &adapter->stats.drvr_stats; + struct be_hw_stats *hw_stats = hw_stats_from_cmd(adapter->stats.cmd.va); + struct be_rxf_stats *rxf_stats = &hw_stats->rxf; + struct be_port_rxf_stats *port_stats = + &rxf_stats->port[adapter->port_num]; + struct net_device_stats *net_stats = &adapter->stats.net_stats; + struct be_erx_stats *erx_stats = &hw_stats->erx; + void *p = NULL; + int i; + + for (i = 0; i < ETHTOOL_STATS_NUM; i++) { + switch (et_stats[i].type) { + case NETSTAT: + p = net_stats; + break; + case DRVSTAT: + p = drvr_stats; + break; + case PORTSTAT: + p = port_stats; + break; + case MISCSTAT: + p = rxf_stats; + break; + case ERXSTAT: /* Currently only one ERX stat is provided */ + p = (u32 *)erx_stats + adapter->rx_obj.q.id; + break; + } + + p = (u8 *)p + et_stats[i].offset; + data[i] = (et_stats[i].size == sizeof(u64)) ? + *(u64 *)p: *(u32 *)p; + } + + return; +} + +static void +be_get_stat_strings(struct net_device *netdev, uint32_t stringset, + uint8_t *data) +{ + int i; + switch (stringset) { + case ETH_SS_STATS: + for (i = 0; i < ETHTOOL_STATS_NUM; i++) { + memcpy(data, et_stats[i].desc, ETH_GSTRING_LEN); + data += ETH_GSTRING_LEN; + } + break; + } +} + +static int be_get_stats_count(struct net_device *netdev) +{ + return ETHTOOL_STATS_NUM; +} + +static int be_get_settings(struct net_device *netdev, struct ethtool_cmd *ecmd) +{ + ecmd->speed = SPEED_10000; + ecmd->duplex = DUPLEX_FULL; + ecmd->autoneg = AUTONEG_DISABLE; + return 0; +} + +static void +be_get_ringparam(struct net_device *netdev, struct ethtool_ringparam *ring) +{ + struct be_adapter *adapter = netdev_priv(netdev); + + ring->rx_max_pending = adapter->rx_obj.q.len; + ring->tx_max_pending = adapter->tx_obj.q.len; + + ring->rx_pending = atomic_read(&adapter->rx_obj.q.used); + ring->tx_pending = atomic_read(&adapter->tx_obj.q.used); +} + +static void +be_get_pauseparam(struct net_device *netdev, struct ethtool_pauseparam *ecmd) +{ + struct be_adapter *adapter = netdev_priv(netdev); + + be_cmd_get_flow_control(&adapter->ctrl, &ecmd->tx_pause, + &ecmd->rx_pause); + ecmd->autoneg = AUTONEG_ENABLE; +} + +static int +be_set_pauseparam(struct net_device *netdev, struct ethtool_pauseparam *ecmd) +{ + struct be_adapter *adapter = netdev_priv(netdev); + int status; + + if (ecmd->autoneg != AUTONEG_ENABLE) + return -EINVAL; + + status = be_cmd_set_flow_control(&adapter->ctrl, ecmd->tx_pause, + ecmd->rx_pause); + if (!status) + dev_warn(&adapter->pdev->dev, "Pause param set failed.\n"); + + return status; +} + +struct ethtool_ops be_ethtool_ops = { + .get_settings = be_get_settings, + .get_drvinfo = be_get_drvinfo, + .get_link = ethtool_op_get_link, + .get_coalesce = be_get_coalesce, + .set_coalesce = be_set_coalesce, + .get_ringparam = be_get_ringparam, + .get_pauseparam = be_get_pauseparam, + .set_pauseparam = be_set_pauseparam, + .get_rx_csum = be_get_rx_csum, + .set_rx_csum = be_set_rx_csum, + .get_tx_csum = ethtool_op_get_tx_csum, + .set_tx_csum = ethtool_op_set_tx_csum, + .get_sg = ethtool_op_get_sg, + .set_sg = ethtool_op_set_sg, + .get_tso = ethtool_op_get_tso, + .set_tso = ethtool_op_set_tso, + .get_strings = be_get_stat_strings, + .get_stats_count = be_get_stats_count, + .get_ethtool_stats = be_get_ethtool_stats, +}; diff --git a/drivers/net/benet/be_hw.h b/drivers/net/benet/be_hw.h new file mode 100644 index 000000000000..b132aa4893ca --- /dev/null +++ b/drivers/net/benet/be_hw.h @@ -0,0 +1,211 @@ +/* + * Copyright (C) 2005 - 2009 ServerEngines + * All rights reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. The full GNU General + * Public License is included in this distribution in the file called COPYING. + * + * Contact Information: + * linux-drivers@serverengines.com + * + * ServerEngines + * 209 N. Fair Oaks Ave + * Sunnyvale, CA 94085 + */ + +/********* Mailbox door bell *************/ +/* Used for driver communication with the FW. + * The software must write this register twice to post any command. First, + * it writes the register with hi=1 and the upper bits of the physical address + * for the MAILBOX structure. Software must poll the ready bit until this + * is acknowledged. Then, sotware writes the register with hi=0 with the lower + * bits in the address. It must poll the ready bit until the command is + * complete. Upon completion, the MAILBOX will contain a valid completion + * queue entry. + */ +#define MPU_MAILBOX_DB_OFFSET 0x160 +#define MPU_MAILBOX_DB_RDY_MASK 0x1 /* bit 0 */ +#define MPU_MAILBOX_DB_HI_MASK 0x2 /* bit 1 */ + +#define MPU_EP_CONTROL 0 + +/********** MPU semphore ******************/ +#define MPU_EP_SEMAPHORE_OFFSET 0xac +#define EP_SEMAPHORE_POST_STAGE_MASK 0x0000FFFF +#define EP_SEMAPHORE_POST_ERR_MASK 0x1 +#define EP_SEMAPHORE_POST_ERR_SHIFT 31 +/* MPU semphore POST stage values */ +#define POST_STAGE_AWAITING_HOST_RDY 0x1 /* FW awaiting goahead from host */ +#define POST_STAGE_HOST_RDY 0x2 /* Host has given go-ahed to FW */ +#define POST_STAGE_BE_RESET 0x3 /* Host wants to reset chip */ +#define POST_STAGE_ARMFW_RDY 0xc000 /* FW is done with POST */ + +/********* Memory BAR register ************/ +#define PCICFG_MEMBAR_CTRL_INT_CTRL_OFFSET 0xfc +/* Host Interrupt Enable, if set interrupts are enabled although "PCI Interrupt + * Disable" may still globally block interrupts in addition to individual + * interrupt masks; a mechanism for the device driver to block all interrupts + * atomically without having to arbitrate for the PCI Interrupt Disable bit + * with the OS. + */ +#define MEMBAR_CTRL_INT_CTRL_HOSTINTR_MASK (1 << 29) /* bit 29 */ +/* PCI physical function number */ +#define MEMBAR_CTRL_INT_CTRL_PFUNC_MASK 0x7 /* bits 26 - 28 */ +#define MEMBAR_CTRL_INT_CTRL_PFUNC_SHIFT 26 + +/********* Event Q door bell *************/ +#define DB_EQ_OFFSET DB_CQ_OFFSET +#define DB_EQ_RING_ID_MASK 0x1FF /* bits 0 - 8 */ +/* Clear the interrupt for this eq */ +#define DB_EQ_CLR_SHIFT (9) /* bit 9 */ +/* Must be 1 */ +#define DB_EQ_EVNT_SHIFT (10) /* bit 10 */ +/* Number of event entries processed */ +#define DB_EQ_NUM_POPPED_SHIFT (16) /* bits 16 - 28 */ +/* Rearm bit */ +#define DB_EQ_REARM_SHIFT (29) /* bit 29 */ + +/********* Compl Q door bell *************/ +#define DB_CQ_OFFSET 0x120 +#define DB_CQ_RING_ID_MASK 0x3FF /* bits 0 - 9 */ +/* Number of event entries processed */ +#define DB_CQ_NUM_POPPED_SHIFT (16) /* bits 16 - 28 */ +/* Rearm bit */ +#define DB_CQ_REARM_SHIFT (29) /* bit 29 */ + +/********** TX ULP door bell *************/ +#define DB_TXULP1_OFFSET 0x60 +#define DB_TXULP_RING_ID_MASK 0x7FF /* bits 0 - 10 */ +/* Number of tx entries posted */ +#define DB_TXULP_NUM_POSTED_SHIFT (16) /* bits 16 - 29 */ +#define DB_TXULP_NUM_POSTED_MASK 0x3FFF /* bits 16 - 29 */ + +/********** RQ(erx) door bell ************/ +#define DB_RQ_OFFSET 0x100 +#define DB_RQ_RING_ID_MASK 0x3FF /* bits 0 - 9 */ +/* Number of rx frags posted */ +#define DB_RQ_NUM_POSTED_SHIFT (24) /* bits 24 - 31 */ + +/* + * BE descriptors: host memory data structures whose formats + * are hardwired in BE silicon. + */ +/* Event Queue Descriptor */ +#define EQ_ENTRY_VALID_MASK 0x1 /* bit 0 */ +#define EQ_ENTRY_RES_ID_MASK 0xFFFF /* bits 16 - 31 */ +#define EQ_ENTRY_RES_ID_SHIFT 16 +struct be_eq_entry { + u32 evt; +}; + +/* TX Queue Descriptor */ +#define ETH_WRB_FRAG_LEN_MASK 0xFFFF +struct be_eth_wrb { + u32 frag_pa_hi; /* dword 0 */ + u32 frag_pa_lo; /* dword 1 */ + u32 rsvd0; /* dword 2 */ + u32 frag_len; /* dword 3: bits 0 - 15 */ +} __packed; + +/* Pseudo amap definition for eth_hdr_wrb in which each bit of the + * actual structure is defined as a byte : used to calculate + * offset/shift/mask of each field */ +struct amap_eth_hdr_wrb { + u8 rsvd0[32]; /* dword 0 */ + u8 rsvd1[32]; /* dword 1 */ + u8 complete; /* dword 2 */ + u8 event; + u8 crc; + u8 forward; + u8 ipsec; + u8 mgmt; + u8 ipcs; + u8 udpcs; + u8 tcpcs; + u8 lso; + u8 vlan; + u8 gso[2]; + u8 num_wrb[5]; + u8 lso_mss[14]; + u8 len[16]; /* dword 3 */ + u8 vlan_tag[16]; +} __packed; + +struct be_eth_hdr_wrb { + u32 dw[4]; +}; + +/* TX Compl Queue Descriptor */ + +/* Pseudo amap definition for eth_tx_compl in which each bit of the + * actual structure is defined as a byte: used to calculate + * offset/shift/mask of each field */ +struct amap_eth_tx_compl { + u8 wrb_index[16]; /* dword 0 */ + u8 ct[2]; /* dword 0 */ + u8 port[2]; /* dword 0 */ + u8 rsvd0[8]; /* dword 0 */ + u8 status[4]; /* dword 0 */ + u8 user_bytes[16]; /* dword 1 */ + u8 nwh_bytes[8]; /* dword 1 */ + u8 lso; /* dword 1 */ + u8 cast_enc[2]; /* dword 1 */ + u8 rsvd1[5]; /* dword 1 */ + u8 rsvd2[32]; /* dword 2 */ + u8 pkts[16]; /* dword 3 */ + u8 ringid[11]; /* dword 3 */ + u8 hash_val[4]; /* dword 3 */ + u8 valid; /* dword 3 */ +} __packed; + +struct be_eth_tx_compl { + u32 dw[4]; +}; + +/* RX Queue Descriptor */ +struct be_eth_rx_d { + u32 fragpa_hi; + u32 fragpa_lo; +}; + +/* RX Compl Queue Descriptor */ + +/* Pseudo amap definition for eth_rx_compl in which each bit of the + * actual structure is defined as a byte: used to calculate + * offset/shift/mask of each field */ +struct amap_eth_rx_compl { + u8 vlan_tag[16]; /* dword 0 */ + u8 pktsize[14]; /* dword 0 */ + u8 port; /* dword 0 */ + u8 ip_opt; /* dword 0 */ + u8 err; /* dword 1 */ + u8 rsshp; /* dword 1 */ + u8 ipf; /* dword 1 */ + u8 tcpf; /* dword 1 */ + u8 udpf; /* dword 1 */ + u8 ipcksm; /* dword 1 */ + u8 l4_cksm; /* dword 1 */ + u8 ip_version; /* dword 1 */ + u8 macdst[6]; /* dword 1 */ + u8 vtp; /* dword 1 */ + u8 rsvd0; /* dword 1 */ + u8 fragndx[10]; /* dword 1 */ + u8 ct[2]; /* dword 1 */ + u8 sw; /* dword 1 */ + u8 numfrags[3]; /* dword 1 */ + u8 rss_flush; /* dword 2 */ + u8 cast_enc[2]; /* dword 2 */ + u8 qnq; /* dword 2 */ + u8 rss_bank; /* dword 2 */ + u8 rsvd1[23]; /* dword 2 */ + u8 lro_pkt; /* dword 2 */ + u8 rsvd2[2]; /* dword 2 */ + u8 valid; /* dword 2 */ + u8 rsshash[32]; /* dword 3 */ +} __packed; + +struct be_eth_rx_compl { + u32 dw[4]; +}; diff --git a/drivers/net/benet/be_main.c b/drivers/net/benet/be_main.c new file mode 100644 index 000000000000..897a63de5bdb --- /dev/null +++ b/drivers/net/benet/be_main.c @@ -0,0 +1,1903 @@ +/* + * Copyright (C) 2005 - 2009 ServerEngines + * All rights reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation. The full GNU General + * Public License is included in this distribution in the file called COPYING. + * + * Contact Information: + * linux-drivers@serverengines.com + * + * ServerEngines + * 209 N. Fair Oaks Ave + * Sunnyvale, CA 94085 + */ + +#include "be.h" + +MODULE_VERSION(DRV_VER); +MODULE_DEVICE_TABLE(pci, be_dev_ids); +MODULE_DESCRIPTION(DRV_DESC " " DRV_VER); +MODULE_AUTHOR("ServerEngines Corporation"); +MODULE_LICENSE("GPL"); + +static unsigned int rx_frag_size = 2048; +module_param(rx_frag_size, uint, S_IRUGO); +MODULE_PARM_DESC(rx_frag_size, "Size of a fragment that holds rcvd data."); + +#define BE_VENDOR_ID 0x19a2 +#define BE2_DEVICE_ID_1 0x0211 +static DEFINE_PCI_DEVICE_TABLE(be_dev_ids) = { + { PCI_DEVICE(BE_VENDOR_ID, BE2_DEVICE_ID_1) }, + { 0 } +}; +MODULE_DEVICE_TABLE(pci, be_dev_ids); + +static void be_queue_free(struct be_adapter *adapter, struct be_queue_info *q) +{ + struct be_dma_mem *mem = &q->dma_mem; + if (mem->va) + pci_free_consistent(adapter->pdev, mem->size, + mem->va, mem->dma); +} + +static int be_queue_alloc(struct be_adapter *adapter, struct be_queue_info *q, + u16 len, u16 entry_size) +{ + struct be_dma_mem *mem = &q->dma_mem; + + memset(q, 0, sizeof(*q)); + q->len = len; + q->entry_size = entry_size; + mem->size = len * entry_size; + mem->va = pci_alloc_consistent(adapter->pdev, mem->size, &mem->dma); + if (!mem->va) + return -1; + memset(mem->va, 0, mem->size); + return 0; +} + +static inline void *queue_head_node(struct be_queue_info *q) +{ + return q->dma_mem.va + q->head * q->entry_size; +} + +static inline void *queue_tail_node(struct be_queue_info *q) +{ + return q->dma_mem.va + q->tail * q->entry_size; +} + +static inline void queue_head_inc(struct be_queue_info *q) +{ + index_inc(&q->head, q->len); +} + +static inline void queue_tail_inc(struct be_queue_info *q) +{ + index_inc(&q->tail, q->len); +} + +static void be_intr_set(struct be_ctrl_info *ctrl, bool enable) +{ + u8 __iomem *addr = ctrl->pcicfg + PCICFG_MEMBAR_CTRL_INT_CTRL_OFFSET; + u32 reg = ioread32(addr); + u32 enabled = reg & MEMBAR_CTRL_INT_CTRL_HOSTINTR_MASK; + if (!enabled && enable) { + reg |= MEMBAR_CTRL_INT_CTRL_HOSTINTR_MASK; + } else if (enabled && !enable) { + reg &= ~MEMBAR_CTRL_INT_CTRL_HOSTINTR_MASK; + } else { + printk(KERN_WARNING DRV_NAME + ": bad value in membar_int_ctrl reg=0x%x\n", reg); + return; + } + iowrite32(reg, addr); +} + +static void be_rxq_notify(struct be_ctrl_info *ctrl, u16 qid, u16 posted) +{ + u32 val = 0; + val |= qid & DB_RQ_RING_ID_MASK; + val |= posted << DB_RQ_NUM_POSTED_SHIFT; + iowrite32(val, ctrl->db + DB_RQ_OFFSET); +} + +static void be_txq_notify(struct be_ctrl_info *ctrl, u16 qid, u16 posted) +{ + u32 val = 0; + val |= qid & DB_TXULP_RING_ID_MASK; + val |= (posted & DB_TXULP_NUM_POSTED_MASK) << DB_TXULP_NUM_POSTED_SHIFT; + iowrite32(val, ctrl->db + DB_TXULP1_OFFSET); +} + +static void be_eq_notify(struct be_ctrl_info *ctrl, u16 qid, + bool arm, bool clear_int, u16 num_popped) +{ + u32 val = 0; + val |= qid & DB_EQ_RING_ID_MASK; + if (arm) + val |= 1 << DB_EQ_REARM_SHIFT; + if (clear_int) + val |= 1 << DB_EQ_CLR_SHIFT; + val |= 1 << DB_EQ_EVNT_SHIFT; + val |= num_popped << DB_EQ_NUM_POPPED_SHIFT; + iowrite32(val, ctrl->db + DB_EQ_OFFSET); +} + +static void be_cq_notify(struct be_ctrl_info *ctrl, u16 qid, + bool arm, u16 num_popped) +{ + u32 val = 0; + val |= qid & DB_CQ_RING_ID_MASK; + if (arm) + val |= 1 << DB_CQ_REARM_SHIFT; + val |= num_popped << DB_CQ_NUM_POPPED_SHIFT; + iowrite32(val, ctrl->db + DB_CQ_OFFSET); +} + + +static int be_mac_addr_set(struct net_device *netdev, void *p) +{ + struct be_adapter *adapter = netdev_priv(netdev); + struct sockaddr *addr = p; + int status = 0; + + if (netif_running(netdev)) { + status = be_cmd_pmac_del(&adapter->ctrl, adapter->if_handle, + adapter->pmac_id); + if (status) + return status; + + status = be_cmd_pmac_add(&adapter->ctrl, (u8 *)addr->sa_data, + adapter->if_handle, &adapter->pmac_id); + } + + if (!status) + memcpy(netdev->dev_addr, addr->sa_data, netdev->addr_len); + + return status; +} + +static void netdev_stats_update(struct be_adapter *adapter) +{ + struct be_hw_stats *hw_stats = hw_stats_from_cmd(adapter->stats.cmd.va); + struct be_rxf_stats *rxf_stats = &hw_stats->rxf; + struct be_port_rxf_stats *port_stats = + &rxf_stats->port[adapter->port_num]; + struct net_device_stats *dev_stats = &adapter->stats.net_stats; + + dev_stats->rx_packets = port_stats->rx_total_frames; + dev_stats->tx_packets = port_stats->tx_unicastframes + + port_stats->tx_multicastframes + port_stats->tx_broadcastframes; + dev_stats->rx_bytes = (u64) port_stats->rx_bytes_msd << 32 | + (u64) port_stats->rx_bytes_lsd; + dev_stats->tx_bytes = (u64) port_stats->tx_bytes_msd << 32 | + (u64) port_stats->tx_bytes_lsd; + + /* bad pkts received */ + dev_stats->rx_errors = port_stats->rx_crc_errors + + port_stats->rx_alignment_symbol_errors + + port_stats->rx_in_range_errors + + port_stats->rx_out_range_errors + port_stats->rx_frame_too_long; + + /* packet transmit problems */ + dev_stats->tx_errors = 0; + + /* no space in linux buffers */ + dev_stats->rx_dropped = 0; + + /* no space available in linux */ + dev_stats->tx_dropped = 0; + + dev_stats->multicast = port_stats->tx_multicastframes; + dev_stats->collisions = 0; + + /* detailed rx errors */ + dev_stats->rx_length_errors = port_stats->rx_in_range_errors + + port_stats->rx_out_range_errors + port_stats->rx_frame_too_long; + /* receive ring buffer overflow */ + dev_stats->rx_over_errors = 0; + dev_stats->rx_crc_errors = port_stats->rx_crc_errors; + + /* frame alignment errors */ + dev_stats->rx_frame_errors = port_stats->rx_alignment_symbol_errors; + /* receiver fifo overrun */ + /* drops_no_pbuf is no per i/f, it's per BE card */ + dev_stats->rx_fifo_errors = port_stats->rx_fifo_overflow + + port_stats->rx_input_fifo_overflow + + rxf_stats->rx_drops_no_pbuf; + /* receiver missed packetd */ + dev_stats->rx_missed_errors = 0; + /* detailed tx_errors */ + dev_stats->tx_aborted_errors = 0; + dev_stats->tx_carrier_errors = 0; + dev_stats->tx_fifo_errors = 0; + dev_stats->tx_heartbeat_errors = 0; + dev_stats->tx_window_errors = 0; +} + +static void be_link_status_update(struct be_adapter *adapter) +{ + struct be_link_info *prev = &adapter->link; + struct be_link_info now = { 0 }; + struct net_device *netdev = adapter->netdev; + + be_cmd_link_status_query(&adapter->ctrl, &now); + + /* If link came up or went down */ + if (now.speed != prev->speed && (now.speed == PHY_LINK_SPEED_ZERO || + prev->speed == PHY_LINK_SPEED_ZERO)) { + if (now.speed == PHY_LINK_SPEED_ZERO) { + netif_stop_queue(netdev); + netif_carrier_off(netdev); + printk(KERN_INFO "%s: Link down\n", netdev->name); + } else { + netif_start_queue(netdev); + netif_carrier_on(netdev); + printk(KERN_INFO "%s: Link up\n", netdev->name); + } + } + *prev = now; +} + +/* Update the EQ delay n BE based on the RX frags consumed / sec */ +static void be_rx_eqd_update(struct be_adapter *adapter) +{ + u32 eqd; + struct be_ctrl_info *ctrl = &adapter->ctrl; + struct be_eq_obj *rx_eq = &adapter->rx_eq; + struct be_drvr_stats *stats = &adapter->stats.drvr_stats; + + /* Update once a second */ + if (((jiffies - stats->rx_fps_jiffies) < HZ) || rx_eq->enable_aic == 0) + return; + + stats->be_rx_fps = (stats->be_rx_frags - stats->be_prev_rx_frags) / + ((jiffies - stats->rx_fps_jiffies) / HZ); + + stats->rx_fps_jiffies = jiffies; + stats->be_prev_rx_frags = stats->be_rx_frags; + eqd = stats->be_rx_fps / 110000; + eqd = eqd << 3; + if (eqd > rx_eq->max_eqd) + eqd = rx_eq->max_eqd; + if (eqd < rx_eq->min_eqd) + eqd = rx_eq->min_eqd; + if (eqd < 10) + eqd = 0; + if (eqd != rx_eq->cur_eqd) + be_cmd_modify_eqd(ctrl, rx_eq->q.id, eqd); + + rx_eq->cur_eqd = eqd; +} + +static void be_worker(struct work_struct *work) +{ + struct be_adapter *adapter = + container_of(work, struct be_adapter, work.work); + int status; + + /* Check link */ + be_link_status_update(adapter); + + /* Get Stats */ + status = be_cmd_get_stats(&adapter->ctrl, &adapter->stats.cmd); + if (!status) + netdev_stats_update(adapter); + + /* Set EQ delay */ + be_rx_eqd_update(adapter); + + schedule_delayed_work(&adapter->work, msecs_to_jiffies(1000)); +} + +static struct net_device_stats *be_get_stats(struct net_device *dev) +{ + struct be_adapter *adapter = netdev_priv(dev); + + return &adapter->stats.net_stats; +} + +static void be_tx_stats_update(struct be_adapter *adapter, + u32 wrb_cnt, u32 copied, bool stopped) +{ + struct be_drvr_stats *stats = &adapter->stats.drvr_stats; + stats->be_tx_reqs++; + stats->be_tx_wrbs += wrb_cnt; + stats->be_tx_bytes += copied; + if (stopped) + stats->be_tx_stops++; + + /* Update tx rate once in two seconds */ + if ((jiffies - stats->be_tx_jiffies) > 2 * HZ) { + u32 r; + r = (stats->be_tx_bytes - stats->be_tx_bytes_prev) / + ((u32) (jiffies - stats->be_tx_jiffies) / HZ); + r = (r / 1000000); /* M bytes/s */ + stats->be_tx_rate = (r * 8); /* M bits/s */ + stats->be_tx_jiffies = jiffies; + stats->be_tx_bytes_prev = stats->be_tx_bytes; + } +} + +/* Determine number of WRB entries needed to xmit data in an skb */ +static u32 wrb_cnt_for_skb(struct sk_buff *skb, bool *dummy) +{ + int cnt = 0; + while (skb) { + if (skb->len > skb->data_len) + cnt++; + cnt += skb_shinfo(skb)->nr_frags; + skb = skb_shinfo(skb)->frag_list; + } + /* to account for hdr wrb */ + cnt++; + if (cnt & 1) { + /* add a dummy to make it an even num */ + cnt++; + *dummy = true; + } else + *dummy = false; + BUG_ON(cnt > BE_MAX_TX_FRAG_COUNT); + return cnt; +} + +static inline void wrb_fill(struct be_eth_wrb *wrb, u64 addr, int len) +{ + wrb->frag_pa_hi = upper_32_bits(addr); + wrb->frag_pa_lo = addr & 0xFFFFFFFF; + wrb->frag_len = len & ETH_WRB_FRAG_LEN_MASK; +} + +static void wrb_fill_hdr(struct be_eth_hdr_wrb *hdr, struct sk_buff *skb, + bool vlan, u32 wrb_cnt, u32 len) +{ + memset(hdr, 0, sizeof(*hdr)); + + AMAP_SET_BITS(struct amap_eth_hdr_wrb, crc, hdr, 1); + + if (skb_shinfo(skb)->gso_segs > 1 && skb_shinfo(skb)->gso_size) { + AMAP_SET_BITS(struct amap_eth_hdr_wrb, lso, hdr, 1); + AMAP_SET_BITS(struct amap_eth_hdr_wrb, lso_mss, + hdr, skb_shinfo(skb)->gso_size); + } else if (skb->ip_summed == CHECKSUM_PARTIAL) { + if (is_tcp_pkt(skb)) + AMAP_SET_BITS(struct amap_eth_hdr_wrb, tcpcs, hdr, 1); + else if (is_udp_pkt(skb)) + AMAP_SET_BITS(struct amap_eth_hdr_wrb, udpcs, hdr, 1); + } + + if (vlan && vlan_tx_tag_present(skb)) { + AMAP_SET_BITS(struct amap_eth_hdr_wrb, vlan, hdr, 1); + AMAP_SET_BITS(struct amap_eth_hdr_wrb, vlan_tag, + hdr, vlan_tx_tag_get(skb)); + } + + AMAP_SET_BITS(struct amap_eth_hdr_wrb, event, hdr, 1); + AMAP_SET_BITS(struct amap_eth_hdr_wrb, complete, hdr, 1); + AMAP_SET_BITS(struct amap_eth_hdr_wrb, num_wrb, hdr, wrb_cnt); + AMAP_SET_BITS(struct amap_eth_hdr_wrb, len, hdr, len); +} + + +static int make_tx_wrbs(struct be_adapter *adapter, + struct sk_buff *skb, u32 wrb_cnt, bool dummy_wrb) +{ + u64 busaddr; + u32 i, copied = 0; + struct pci_dev *pdev = adapter->pdev; + struct sk_buff *first_skb = skb; + struct be_queue_info *txq = &adapter->tx_obj.q; + struct be_eth_wrb *wrb; + struct be_eth_hdr_wrb *hdr; + + atomic_add(wrb_cnt, &txq->used); + hdr = queue_head_node(txq); + queue_head_inc(txq); + + while (skb) { + if (skb->len > skb->data_len) { + int len = skb->len - skb->data_len; + busaddr = pci_map_single(pdev, skb->data, len, + PCI_DMA_TODEVICE); + wrb = queue_head_node(txq); + wrb_fill(wrb, busaddr, len); + be_dws_cpu_to_le(wrb, sizeof(*wrb)); + queue_head_inc(txq); + copied += len; + } + + for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) { + struct skb_frag_struct *frag = + &skb_shinfo(skb)->frags[i]; + busaddr = pci_map_page(pdev, frag->page, + frag->page_offset, + frag->size, PCI_DMA_TODEVICE); + wrb = queue_head_node(txq); + wrb_fill(wrb, busaddr, frag->size); + be_dws_cpu_to_le(wrb, sizeof(*wrb)); + queue_head_inc(txq); + copied += frag->size; + } + skb = skb_shinfo(skb)->frag_list; + } + + if (dummy_wrb) { + wrb = queue_head_node(txq); + wrb_fill(wrb, 0, 0); + be_dws_cpu_to_le(wrb, sizeof(*wrb)); + queue_head_inc(txq); + } + + wrb_fill_hdr(hdr, first_skb, adapter->vlan_grp ? true : false, + wrb_cnt, copied); + be_dws_cpu_to_le(hdr, sizeof(*hdr)); + + return copied; +} + +static int be_xmit(struct sk_buff *skb, struct net_device *netdev) +{ + struct be_adapter *adapter = netdev_priv(netdev); + struct be_tx_obj *tx_obj = &adapter->tx_obj; + struct be_queue_info *txq = &tx_obj->q; + u32 wrb_cnt = 0, copied = 0; + u32 start = txq->head; + bool dummy_wrb, stopped = false; + + wrb_cnt = wrb_cnt_for_skb(skb, &dummy_wrb); + + copied = make_tx_wrbs(adapter, skb, wrb_cnt, dummy_wrb); + + /* record the sent skb in the sent_skb table */ + BUG_ON(tx_obj->sent_skb_list[start]); + tx_obj->sent_skb_list[start] = skb; + + /* Ensure that txq has space for the next skb; Else stop the queue + * *BEFORE* ringing the tx doorbell, so that we serialze the + * tx compls of the current transmit which'll wake up the queue + */ + if ((BE_MAX_TX_FRAG_COUNT + atomic_read(&txq->used)) >= txq->len) { + netif_stop_queue(netdev); + stopped = true; + } + + be_txq_notify(&adapter->ctrl, txq->id, wrb_cnt); + + netdev->trans_start = jiffies; + + be_tx_stats_update(adapter, wrb_cnt, copied, stopped); + return NETDEV_TX_OK; +} + +static int be_change_mtu(struct net_device *netdev, int new_mtu) +{ + struct be_adapter *adapter = netdev_priv(netdev); + if (new_mtu < BE_MIN_MTU || + new_mtu > BE_MAX_JUMBO_FRAME_SIZE) { + dev_info(&adapter->pdev->dev, + "MTU must be between %d and %d bytes\n", + BE_MIN_MTU, BE_MAX_JUMBO_FRAME_SIZE); + return -EINVAL; + } + dev_info(&adapter->pdev->dev, "MTU changed from %d to %d bytes\n", + netdev->mtu, new_mtu); + netdev->mtu = new_mtu; + return 0; +} + +/* + * if there are BE_NUM_VLANS_SUPPORTED or lesser number of VLANS configured, + * program them in BE. If more than BE_NUM_VLANS_SUPPORTED are configured, + * set the BE in promiscuous VLAN mode. + */ +static void be_vids_config(struct net_device *netdev) +{ + struct be_adapter *adapter = netdev_priv(netdev); + u16 vtag[BE_NUM_VLANS_SUPPORTED]; + u16 ntags = 0, i; + + if (adapter->num_vlans <= BE_NUM_VLANS_SUPPORTED) { + /* Construct VLAN Table to give to HW */ + for (i = 0; i < VLAN_GROUP_ARRAY_LEN; i++) { + if (adapter->vlan_tag[i]) { + vtag[ntags] = cpu_to_le16(i); + ntags++; + } + } + be_cmd_vlan_config(&adapter->ctrl, adapter->if_handle, + vtag, ntags, 1, 0); + } else { + be_cmd_vlan_config(&adapter->ctrl, adapter->if_handle, + NULL, 0, 1, 1); + } +} + +static void be_vlan_register(struct net_device *netdev, struct vlan_group *grp) +{ + struct be_adapter *adapter = netdev_priv(netdev); + struct be_eq_obj *rx_eq = &adapter->rx_eq; + struct be_eq_obj *tx_eq = &adapter->tx_eq; + struct be_ctrl_info *ctrl = &adapter->ctrl; + + be_eq_notify(ctrl, rx_eq->q.id, false, false, 0); + be_eq_notify(ctrl, tx_eq->q.id, false, false, 0); + adapter->vlan_grp = grp; + be_eq_notify(ctrl, rx_eq->q.id, true, false, 0); + be_eq_notify(ctrl, tx_eq->q.id, true, false, 0); +} + +static void be_vlan_add_vid(struct net_device *netdev, u16 vid) +{ + struct be_adapter *adapter = netdev_priv(netdev); + + adapter->num_vlans++; + adapter->vlan_tag[vid] = 1; + + be_vids_config(netdev); +} + +static void be_vlan_rem_vid(struct net_device *netdev, u16 vid) +{ + struct be_adapter *adapter = netdev_priv(netdev); + + adapter->num_vlans--; + adapter->vlan_tag[vid] = 0; + + vlan_group_set_device(adapter->vlan_grp, vid, NULL); + be_vids_config(netdev); +} + +static void be_set_multicast_filter(struct net_device *netdev) +{ + struct be_adapter *adapter = netdev_priv(netdev); + struct dev_mc_list *mc_ptr; + u8 mac_addr[32][ETH_ALEN]; + int i = 0; + + if (netdev->flags & IFF_ALLMULTI) { + /* set BE in Multicast promiscuous */ + be_cmd_mcast_mac_set(&adapter->ctrl, + adapter->if_handle, NULL, 0, true); + return; + } + + for (mc_ptr = netdev->mc_list; mc_ptr; mc_ptr = mc_ptr->next) { + memcpy(&mac_addr[i][0], mc_ptr->dmi_addr, ETH_ALEN); + if (++i >= 32) { + be_cmd_mcast_mac_set(&adapter->ctrl, + adapter->if_handle, &mac_addr[0][0], i, false); + i = 0; + } + + } + + if (i) { + /* reset the promiscuous mode also. */ + be_cmd_mcast_mac_set(&adapter->ctrl, + adapter->if_handle, &mac_addr[0][0], i, false); + } +} + +static void be_set_multicast_list(struct net_device *netdev) +{ + struct be_adapter *adapter = netdev_priv(netdev); + + if (netdev->flags & IFF_PROMISC) { + be_cmd_promiscuous_config(&adapter->ctrl, adapter->port_num, 1); + } else { + be_cmd_promiscuous_config(&adapter->ctrl, adapter->port_num, 0); + be_set_multicast_filter(netdev); + } +} + +static void be_rx_rate_update(struct be_adapter *adapter, u32 pktsize, + u16 numfrags) +{ + struct be_drvr_stats *stats = &adapter->stats.drvr_stats; + u32 rate; + + stats->be_rx_compl++; + stats->be_rx_frags += numfrags; + stats->be_rx_bytes += pktsize; + + /* Update the rate once in two seconds */ + if ((jiffies - stats->be_rx_jiffies) < 2 * HZ) + return; + + rate = (stats->be_rx_bytes - stats->be_rx_bytes_prev) / + ((u32) (jiffies - stats->be_rx_jiffies) / HZ); + rate = (rate / 1000000); /* MB/Sec */ + stats->be_rx_rate = (rate * 8); /* Mega Bits/Sec */ + stats->be_rx_jiffies = jiffies; + stats->be_rx_bytes_prev = stats->be_rx_bytes; +} + +static struct be_rx_page_info * +get_rx_page_info(struct be_adapter *adapter, u16 frag_idx) +{ + struct be_rx_page_info *rx_page_info; + struct be_queue_info *rxq = &adapter->rx_obj.q; + + rx_page_info = &adapter->rx_obj.page_info_tbl[frag_idx]; + BUG_ON(!rx_page_info->page); + + if (rx_page_info->last_page_user) + pci_unmap_page(adapter->pdev, pci_unmap_addr(rx_page_info, bus), + adapter->big_page_size, PCI_DMA_FROMDEVICE); + + atomic_dec(&rxq->used); + return rx_page_info; +} + +/* Throwaway the data in the Rx completion */ +static void be_rx_compl_discard(struct be_adapter *adapter, + struct be_eth_rx_compl *rxcp) +{ + struct be_queue_info *rxq = &adapter->rx_obj.q; + struct be_rx_page_info *page_info; + u16 rxq_idx, i, num_rcvd; + + rxq_idx = AMAP_GET_BITS(struct amap_eth_rx_compl, fragndx, rxcp); + num_rcvd = AMAP_GET_BITS(struct amap_eth_rx_compl, numfrags, rxcp); + + for (i = 0; i < num_rcvd; i++) { + page_info = get_rx_page_info(adapter, rxq_idx); + put_page(page_info->page); + memset(page_info, 0, sizeof(*page_info)); + index_inc(&rxq_idx, rxq->len); + } +} + +/* + * skb_fill_rx_data forms a complete skb for an ether frame + * indicated by rxcp. + */ +static void skb_fill_rx_data(struct be_adapter *adapter, + struct sk_buff *skb, struct be_eth_rx_compl *rxcp) +{ + struct be_queue_info *rxq = &adapter->rx_obj.q; + struct be_rx_page_info *page_info; + u16 rxq_idx, i, num_rcvd; + u32 pktsize, hdr_len, curr_frag_len; + u8 *start; + + rxq_idx = AMAP_GET_BITS(struct amap_eth_rx_compl, fragndx, rxcp); + pktsize = AMAP_GET_BITS(struct amap_eth_rx_compl, pktsize, rxcp); + num_rcvd = AMAP_GET_BITS(struct amap_eth_rx_compl, numfrags, rxcp); + + page_info = get_rx_page_info(adapter, rxq_idx); + + start = page_address(page_info->page) + page_info->page_offset; + prefetch(start); + + /* Copy data in the first descriptor of this completion */ + curr_frag_len = min(pktsize, rx_frag_size); + + /* Copy the header portion into skb_data */ + hdr_len = min((u32)BE_HDR_LEN, curr_frag_len); + memcpy(skb->data, start, hdr_len); + skb->len = curr_frag_len; + if (curr_frag_len <= BE_HDR_LEN) { /* tiny packet */ + /* Complete packet has now been moved to data */ + put_page(page_info->page); + skb->data_len = 0; + skb->tail += curr_frag_len; + } else { + skb_shinfo(skb)->nr_frags = 1; + skb_shinfo(skb)->frags[0].page = page_info->page; + skb_shinfo(skb)->frags[0].page_offset = + page_info->page_offset + hdr_len; + skb_shinfo(skb)->frags[0].size = curr_frag_len - hdr_len; + skb->data_len = curr_frag_len - hdr_len; + skb->tail += hdr_len; + } + memset(page_info, 0, sizeof(*page_info)); + + if (pktsize <= rx_frag_size) { + BUG_ON(num_rcvd != 1); + return; + } + + /* More frags present for this completion */ + pktsize -= curr_frag_len; /* account for above copied frag */ + for (i = 1; i < num_rcvd; i++) { + index_inc(&rxq_idx, rxq->len); + page_info = get_rx_page_info(adapter, rxq_idx); + + curr_frag_len = min(pktsize, rx_frag_size); + + skb_shinfo(skb)->frags[i].page = page_info->page; + skb_shinfo(skb)->frags[i].page_offset = page_info->page_offset; + skb_shinfo(skb)->frags[i].size = curr_frag_len; + skb->len += curr_frag_len; + skb->data_len += curr_frag_len; + skb_shinfo(skb)->nr_frags++; + pktsize -= curr_frag_len; + + memset(page_info, 0, sizeof(*page_info)); + } + + be_rx_rate_update(adapter, pktsize, num_rcvd); + return; +} + +/* Process the RX completion indicated by rxcp when LRO is disabled */ +static void be_rx_compl_process(struct be_adapter *adapter, + struct be_eth_rx_compl *rxcp) +{ + struct sk_buff *skb; + u32 vtp, vid; + int l4_cksm; + + l4_cksm = AMAP_GET_BITS(struct amap_eth_rx_compl, l4_cksm, rxcp); + vtp = AMAP_GET_BITS(struct amap_eth_rx_compl, vtp, rxcp); + + skb = netdev_alloc_skb(adapter->netdev, BE_HDR_LEN + NET_IP_ALIGN); + if (!skb) { + if (net_ratelimit()) + dev_warn(&adapter->pdev->dev, "skb alloc failed\n"); + be_rx_compl_discard(adapter, rxcp); + return; + } + + skb_reserve(skb, NET_IP_ALIGN); + + skb_fill_rx_data(adapter, skb, rxcp); + + if (l4_cksm && adapter->rx_csum) + skb->ip_summed = CHECKSUM_UNNECESSARY; + else + skb->ip_summed = CHECKSUM_NONE; + + skb->truesize = skb->len + sizeof(struct sk_buff); + skb->protocol = eth_type_trans(skb, adapter->netdev); + skb->dev = adapter->netdev; + + if (vtp) { + if (!adapter->vlan_grp || adapter->num_vlans == 0) { + kfree_skb(skb); + return; + } + vid = AMAP_GET_BITS(struct amap_eth_rx_compl, vlan_tag, rxcp); + vid = be16_to_cpu(vid); + vlan_hwaccel_receive_skb(skb, adapter->vlan_grp, vid); + } else { + netif_receive_skb(skb); + } + + adapter->netdev->last_rx = jiffies; + + return; +} + +/* Process the RX completion indicated by rxcp when LRO is enabled */ +static void be_rx_compl_process_lro(struct be_adapter *adapter, + struct be_eth_rx_compl *rxcp) +{ + struct be_rx_page_info *page_info; + struct skb_frag_struct rx_frags[BE_MAX_FRAGS_PER_FRAME]; + struct be_queue_info *rxq = &adapter->rx_obj.q; + u32 num_rcvd, pkt_size, remaining, vlanf, curr_frag_len; + u16 i, rxq_idx = 0, vid; + + num_rcvd = AMAP_GET_BITS(struct amap_eth_rx_compl, numfrags, rxcp); + pkt_size = AMAP_GET_BITS(struct amap_eth_rx_compl, pktsize, rxcp); + vlanf = AMAP_GET_BITS(struct amap_eth_rx_compl, vtp, rxcp); + rxq_idx = AMAP_GET_BITS(struct amap_eth_rx_compl, fragndx, rxcp); + + remaining = pkt_size; + for (i = 0; i < num_rcvd; i++) { + page_info = get_rx_page_info(adapter, rxq_idx); + + curr_frag_len = min(remaining, rx_frag_size); + + rx_frags[i].page = page_info->page; + rx_frags[i].page_offset = page_info->page_offset; + rx_frags[i].size = curr_frag_len; + remaining -= curr_frag_len; + + index_inc(&rxq_idx, rxq->len); + + memset(page_info, 0, sizeof(*page_info)); + } + + if (likely(!vlanf)) { + lro_receive_frags(&adapter->rx_obj.lro_mgr, rx_frags, pkt_size, + pkt_size, NULL, 0); + } else { + vid = AMAP_GET_BITS(struct amap_eth_rx_compl, vlan_tag, rxcp); + vid = be16_to_cpu(vid); + + if (!adapter->vlan_grp || adapter->num_vlans == 0) + return; + + lro_vlan_hwaccel_receive_frags(&adapter->rx_obj.lro_mgr, + rx_frags, pkt_size, pkt_size, adapter->vlan_grp, + vid, NULL, 0); + } + + be_rx_rate_update(adapter, pkt_size, num_rcvd); + return; +} + +static struct be_eth_rx_compl *be_rx_compl_get(struct be_adapter *adapter) +{ + struct be_eth_rx_compl *rxcp = queue_tail_node(&adapter->rx_obj.cq); + + if (rxcp->dw[offsetof(struct amap_eth_rx_compl, valid) / 32] == 0) + return NULL; + + be_dws_le_to_cpu(rxcp, sizeof(*rxcp)); + + rxcp->dw[offsetof(struct amap_eth_rx_compl, valid) / 32] = 0; + + queue_tail_inc(&adapter->rx_obj.cq); + return rxcp; +} + +static inline struct page *be_alloc_pages(u32 size) +{ + gfp_t alloc_flags = GFP_ATOMIC; + u32 order = get_order(size); + if (order > 0) + alloc_flags |= __GFP_COMP; + return alloc_pages(alloc_flags, order); +} + +/* + * Allocate a page, split it to fragments of size rx_frag_size and post as + * receive buffers to BE + */ +static void be_post_rx_frags(struct be_adapter *adapter) +{ + struct be_rx_page_info *page_info_tbl = adapter->rx_obj.page_info_tbl; + struct be_rx_page_info *page_info = NULL; + struct be_queue_info *rxq = &adapter->rx_obj.q; + struct page *pagep = NULL; + struct be_eth_rx_d *rxd; + u64 page_dmaaddr = 0, frag_dmaaddr; + u32 posted, page_offset = 0; + + + page_info = &page_info_tbl[rxq->head]; + for (posted = 0; posted < MAX_RX_POST && !page_info->page; posted++) { + if (!pagep) { + pagep = be_alloc_pages(adapter->big_page_size); + if (unlikely(!pagep)) { + drvr_stats(adapter)->be_ethrx_post_fail++; + break; + } + page_dmaaddr = pci_map_page(adapter->pdev, pagep, 0, + adapter->big_page_size, + PCI_DMA_FROMDEVICE); + page_info->page_offset = 0; + } else { + get_page(pagep); + page_info->page_offset = page_offset + rx_frag_size; + } + page_offset = page_info->page_offset; + page_info->page = pagep; + pci_unmap_addr_set(page_info, bus, page_dmaaddr); + frag_dmaaddr = page_dmaaddr + page_info->page_offset; + + rxd = queue_head_node(rxq); + rxd->fragpa_lo = cpu_to_le32(frag_dmaaddr & 0xFFFFFFFF); + rxd->fragpa_hi = cpu_to_le32(upper_32_bits(frag_dmaaddr)); + queue_head_inc(rxq); + + /* Any space left in the current big page for another frag? */ + if ((page_offset + rx_frag_size + rx_frag_size) > + adapter->big_page_size) { + pagep = NULL; + page_info->last_page_user = true; + } + page_info = &page_info_tbl[rxq->head]; + } + if (pagep) + page_info->last_page_user = true; + + if (posted) { + be_rxq_notify(&adapter->ctrl, rxq->id, posted); + atomic_add(posted, &rxq->used); + } + + return; +} + +static struct be_eth_tx_compl * +be_tx_compl_get(struct be_adapter *adapter) +{ + struct be_queue_info *tx_cq = &adapter->tx_obj.cq; + struct be_eth_tx_compl *txcp = queue_tail_node(tx_cq); + + if (txcp->dw[offsetof(struct amap_eth_tx_compl, valid) / 32] == 0) + return NULL; + + be_dws_le_to_cpu(txcp, sizeof(*txcp)); + + txcp->dw[offsetof(struct amap_eth_tx_compl, valid) / 32] = 0; + + queue_tail_inc(tx_cq); + return txcp; +} + +static void be_tx_compl_process(struct be_adapter *adapter, u16 last_index) +{ + struct be_queue_info *txq = &adapter->tx_obj.q; + struct be_eth_wrb *wrb; + struct sk_buff **sent_skbs = adapter->tx_obj.sent_skb_list; + struct sk_buff *sent_skb; + u64 busaddr; + u16 cur_index, num_wrbs = 0; + + cur_index = txq->tail; + sent_skb = sent_skbs[cur_index]; + BUG_ON(!sent_skb); + sent_skbs[cur_index] = NULL; + + do { + cur_index = txq->tail; + wrb = queue_tail_node(txq); + be_dws_le_to_cpu(wrb, sizeof(*wrb)); + busaddr = ((u64)wrb->frag_pa_hi << 32) | (u64)wrb->frag_pa_lo; + if (busaddr != 0) { + pci_unmap_single(adapter->pdev, busaddr, + wrb->frag_len, PCI_DMA_TODEVICE); + } + num_wrbs++; + queue_tail_inc(txq); + } while (cur_index != last_index); + + atomic_sub(num_wrbs, &txq->used); + + kfree_skb(sent_skb); +} + +static void be_rx_q_clean(struct be_adapter *adapter) +{ + struct be_rx_page_info *page_info; + struct be_queue_info *rxq = &adapter->rx_obj.q; + struct be_queue_info *rx_cq = &adapter->rx_obj.cq; + struct be_eth_rx_compl *rxcp; + u16 tail; + + /* First cleanup pending rx completions */ + while ((rxcp = be_rx_compl_get(adapter)) != NULL) { + be_rx_compl_discard(adapter, rxcp); + be_cq_notify(&adapter->ctrl, rx_cq->id, true, 1); + } + + /* Then free posted rx buffer that were not used */ + tail = (rxq->head + rxq->len - atomic_read(&rxq->used)) % rxq->len; + for (; tail != rxq->head; index_inc(&tail, rxq->len)) { + page_info = get_rx_page_info(adapter, tail); + put_page(page_info->page); + memset(page_info, 0, sizeof(*page_info)); + } + BUG_ON(atomic_read(&rxq->used)); +} + +static void be_tx_q_clean(struct be_adapter *adapter) +{ + struct sk_buff **sent_skbs = adapter->tx_obj.sent_skb_list; + struct sk_buff *sent_skb; + struct be_queue_info *txq = &adapter->tx_obj.q; + u16 last_index; + bool dummy_wrb; + + while (atomic_read(&txq->used)) { + sent_skb = sent_skbs[txq->tail]; + last_index = txq->tail; + index_adv(&last_index, + wrb_cnt_for_skb(sent_skb, &dummy_wrb) - 1, txq->len); + be_tx_compl_process(adapter, last_index); + } +} + +static void be_tx_queues_destroy(struct be_adapter *adapter) +{ + struct be_queue_info *q; + + q = &adapter->tx_obj.q; + if (q->created) + be_cmd_q_destroy(&adapter->ctrl, q, QTYPE_TXQ); + be_queue_free(adapter, q); + + q = &adapter->tx_obj.cq; + if (q->created) + be_cmd_q_destroy(&adapter->ctrl, q, QTYPE_CQ); + be_queue_free(adapter, q); + + /* No more tx completions can be rcvd now; clean up if there are + * any pending completions or pending tx requests */ + be_tx_q_clean(adapter); + + q = &adapter->tx_eq.q; + if (q->created) + be_cmd_q_destroy(&adapter->ctrl, q, QTYPE_EQ); + be_queue_free(adapter, q); +} + +static int be_tx_queues_create(struct be_adapter *adapter) +{ + struct be_queue_info *eq, *q, *cq; + + adapter->tx_eq.max_eqd = 0; + adapter->tx_eq.min_eqd = 0; + adapter->tx_eq.cur_eqd = 96; + adapter->tx_eq.enable_aic = false; + /* Alloc Tx Event queue */ + eq = &adapter->tx_eq.q; + if (be_queue_alloc(adapter, eq, EVNT_Q_LEN, sizeof(struct be_eq_entry))) + return -1; + + /* Ask BE to create Tx Event queue */ + if (be_cmd_eq_create(&adapter->ctrl, eq, adapter->tx_eq.cur_eqd)) + goto tx_eq_free; + /* Alloc TX eth compl queue */ + cq = &adapter->tx_obj.cq; + if (be_queue_alloc(adapter, cq, TX_CQ_LEN, + sizeof(struct be_eth_tx_compl))) + goto tx_eq_destroy; + + /* Ask BE to create Tx eth compl queue */ + if (be_cmd_cq_create(&adapter->ctrl, cq, eq, false, false, 3)) + goto tx_cq_free; + + /* Alloc TX eth queue */ + q = &adapter->tx_obj.q; + if (be_queue_alloc(adapter, q, TX_Q_LEN, sizeof(struct be_eth_wrb))) + goto tx_cq_destroy; + + /* Ask BE to create Tx eth queue */ + if (be_cmd_txq_create(&adapter->ctrl, q, cq)) + goto tx_q_free; + return 0; + +tx_q_free: + be_queue_free(adapter, q); +tx_cq_destroy: + be_cmd_q_destroy(&adapter->ctrl, cq, QTYPE_CQ); +tx_cq_free: + be_queue_free(adapter, cq); +tx_eq_destroy: + be_cmd_q_destroy(&adapter->ctrl, eq, QTYPE_EQ); +tx_eq_free: + be_queue_free(adapter, eq); + return -1; +} + +static void be_rx_queues_destroy(struct be_adapter *adapter) +{ + struct be_queue_info *q; + + q = &adapter->rx_obj.q; + if (q->created) { + be_cmd_q_destroy(&adapter->ctrl, q, QTYPE_RXQ); + be_rx_q_clean(adapter); + } + be_queue_free(adapter, q); + + q = &adapter->rx_obj.cq; + if (q->created) + be_cmd_q_destroy(&adapter->ctrl, q, QTYPE_CQ); + be_queue_free(adapter, q); + + q = &adapter->rx_eq.q; + if (q->created) + be_cmd_q_destroy(&adapter->ctrl, q, QTYPE_EQ); + be_queue_free(adapter, q); +} + +static int be_rx_queues_create(struct be_adapter *adapter) +{ + struct be_queue_info *eq, *q, *cq; + int rc; + + adapter->max_rx_coal = BE_MAX_FRAGS_PER_FRAME; + adapter->big_page_size = (1 << get_order(rx_frag_size)) * PAGE_SIZE; + adapter->rx_eq.max_eqd = BE_MAX_EQD; + adapter->rx_eq.min_eqd = 0; + adapter->rx_eq.cur_eqd = 0; + adapter->rx_eq.enable_aic = true; + + /* Alloc Rx Event queue */ + eq = &adapter->rx_eq.q; + rc = be_queue_alloc(adapter, eq, EVNT_Q_LEN, + sizeof(struct be_eq_entry)); + if (rc) + return rc; + + /* Ask BE to create Rx Event queue */ + rc = be_cmd_eq_create(&adapter->ctrl, eq, adapter->rx_eq.cur_eqd); + if (rc) + goto rx_eq_free; + + /* Alloc RX eth compl queue */ + cq = &adapter->rx_obj.cq; + rc = be_queue_alloc(adapter, cq, RX_CQ_LEN, + sizeof(struct be_eth_rx_compl)); + if (rc) + goto rx_eq_destroy; + + /* Ask BE to create Rx eth compl queue */ + rc = be_cmd_cq_create(&adapter->ctrl, cq, eq, false, false, 3); + if (rc) + goto rx_cq_free; + + /* Alloc RX eth queue */ + q = &adapter->rx_obj.q; + rc = be_queue_alloc(adapter, q, RX_Q_LEN, sizeof(struct be_eth_rx_d)); + if (rc) + goto rx_cq_destroy; + + /* Ask BE to create Rx eth queue */ + rc = be_cmd_rxq_create(&adapter->ctrl, q, cq->id, rx_frag_size, + BE_MAX_JUMBO_FRAME_SIZE, adapter->if_handle, false); + if (rc) + goto rx_q_free; + + return 0; +rx_q_free: + be_queue_free(adapter, q); +rx_cq_destroy: + be_cmd_q_destroy(&adapter->ctrl, cq, QTYPE_CQ); +rx_cq_free: + be_queue_free(adapter, cq); +rx_eq_destroy: + be_cmd_q_destroy(&adapter->ctrl, eq, QTYPE_EQ); +rx_eq_free: + be_queue_free(adapter, eq); + return rc; +} +static bool event_get(struct be_eq_obj *eq_obj, u16 *rid) +{ + struct be_eq_entry *entry = queue_tail_node(&eq_obj->q); + u32 evt = entry->evt; + + if (!evt) + return false; + + evt = le32_to_cpu(evt); + *rid = (evt >> EQ_ENTRY_RES_ID_SHIFT) & EQ_ENTRY_RES_ID_MASK; + entry->evt = 0; + queue_tail_inc(&eq_obj->q); + return true; +} + +static int event_handle(struct be_ctrl_info *ctrl, + struct be_eq_obj *eq_obj) +{ + u16 rid = 0, num = 0; + + while (event_get(eq_obj, &rid)) + num++; + + /* We can see an interrupt and no event */ + be_eq_notify(ctrl, eq_obj->q.id, true, true, num); + if (num) + napi_schedule(&eq_obj->napi); + + return num; +} + +static irqreturn_t be_intx(int irq, void *dev) +{ + struct be_adapter *adapter = dev; + struct be_ctrl_info *ctrl = &adapter->ctrl; + int rx, tx; + + tx = event_handle(ctrl, &adapter->tx_eq); + rx = event_handle(ctrl, &adapter->rx_eq); + + if (rx || tx) + return IRQ_HANDLED; + else + return IRQ_NONE; +} + +static irqreturn_t be_msix_rx(int irq, void *dev) +{ + struct be_adapter *adapter = dev; + + event_handle(&adapter->ctrl, &adapter->rx_eq); + + return IRQ_HANDLED; +} + +static irqreturn_t be_msix_tx(int irq, void *dev) +{ + struct be_adapter *adapter = dev; + + event_handle(&adapter->ctrl, &adapter->tx_eq); + + return IRQ_HANDLED; +} + +static inline bool do_lro(struct be_adapter *adapter, + struct be_eth_rx_compl *rxcp) +{ + int err = AMAP_GET_BITS(struct amap_eth_rx_compl, err, rxcp); + int tcp_frame = AMAP_GET_BITS(struct amap_eth_rx_compl, tcpf, rxcp); + + if (err) + drvr_stats(adapter)->be_rxcp_err++; + + return (!tcp_frame || err || (adapter->max_rx_coal <= 1)) ? + false : true; +} + +int be_poll_rx(struct napi_struct *napi, int budget) +{ + struct be_eq_obj *rx_eq = container_of(napi, struct be_eq_obj, napi); + struct be_adapter *adapter = + container_of(rx_eq, struct be_adapter, rx_eq); + struct be_queue_info *rx_cq = &adapter->rx_obj.cq; + struct be_eth_rx_compl *rxcp; + u32 work_done; + + for (work_done = 0; work_done < budget; work_done++) { + rxcp = be_rx_compl_get(adapter); + if (!rxcp) + break; + + if (do_lro(adapter, rxcp)) + be_rx_compl_process_lro(adapter, rxcp); + else + be_rx_compl_process(adapter, rxcp); + } + + lro_flush_all(&adapter->rx_obj.lro_mgr); + + /* Refill the queue */ + if (atomic_read(&adapter->rx_obj.q.used) < RX_FRAGS_REFILL_WM) + be_post_rx_frags(adapter); + + /* All consumed */ + if (work_done < budget) { + napi_complete(napi); + be_cq_notify(&adapter->ctrl, rx_cq->id, true, work_done); + } else { + /* More to be consumed; continue with interrupts disabled */ + be_cq_notify(&adapter->ctrl, rx_cq->id, false, work_done); + } + return work_done; +} + +/* For TX we don't honour budget; consume everything */ +int be_poll_tx(struct napi_struct *napi, int budget) +{ + struct be_eq_obj *tx_eq = container_of(napi, struct be_eq_obj, napi); + struct be_adapter *adapter = + container_of(tx_eq, struct be_adapter, tx_eq); + struct be_tx_obj *tx_obj = &adapter->tx_obj; + struct be_queue_info *tx_cq = &tx_obj->cq; + struct be_queue_info *txq = &tx_obj->q; + struct be_eth_tx_compl *txcp; + u32 num_cmpl = 0; + u16 end_idx; + + while ((txcp = be_tx_compl_get(adapter))) { + end_idx = AMAP_GET_BITS(struct amap_eth_tx_compl, + wrb_index, txcp); + be_tx_compl_process(adapter, end_idx); + num_cmpl++; + } + + /* As Tx wrbs have been freed up, wake up netdev queue if + * it was stopped due to lack of tx wrbs. + */ + if (netif_queue_stopped(adapter->netdev) && + atomic_read(&txq->used) < txq->len / 2) { + netif_wake_queue(adapter->netdev); + } + + napi_complete(napi); + + be_cq_notify(&adapter->ctrl, tx_cq->id, true, num_cmpl); + + drvr_stats(adapter)->be_tx_events++; + drvr_stats(adapter)->be_tx_compl += num_cmpl; + + return 1; +} + +static void be_msix_enable(struct be_adapter *adapter) +{ + int i, status; + + for (i = 0; i < BE_NUM_MSIX_VECTORS; i++) + adapter->msix_entries[i].entry = i; + + status = pci_enable_msix(adapter->pdev, adapter->msix_entries, + BE_NUM_MSIX_VECTORS); + if (status == 0) + adapter->msix_enabled = true; + return; +} + +static inline int be_msix_vec_get(struct be_adapter *adapter, u32 eq_id) +{ + return adapter->msix_entries[eq_id - + 8 * adapter->ctrl.pci_func].vector; +} + +static int be_msix_register(struct be_adapter *adapter) +{ + struct net_device *netdev = adapter->netdev; + struct be_eq_obj *tx_eq = &adapter->tx_eq; + struct be_eq_obj *rx_eq = &adapter->rx_eq; + int status, vec; + + sprintf(tx_eq->desc, "%s-tx", netdev->name); + vec = be_msix_vec_get(adapter, tx_eq->q.id); + status = request_irq(vec, be_msix_tx, 0, tx_eq->desc, adapter); + if (status) + goto err; + + sprintf(rx_eq->desc, "%s-rx", netdev->name); + vec = be_msix_vec_get(adapter, rx_eq->q.id); + status = request_irq(vec, be_msix_rx, 0, rx_eq->desc, adapter); + if (status) { /* Free TX IRQ */ + vec = be_msix_vec_get(adapter, tx_eq->q.id); + free_irq(vec, adapter); + goto err; + } + return 0; +err: + dev_warn(&adapter->pdev->dev, + "MSIX Request IRQ failed - err %d\n", status); + pci_disable_msix(adapter->pdev); + adapter->msix_enabled = false; + return status; +} + +static int be_irq_register(struct be_adapter *adapter) +{ + struct net_device *netdev = adapter->netdev; + int status; + + if (adapter->msix_enabled) { + status = be_msix_register(adapter); + if (status == 0) + goto done; + } + + /* INTx */ + netdev->irq = adapter->pdev->irq; + status = request_irq(netdev->irq, be_intx, IRQF_SHARED, netdev->name, + adapter); + if (status) { + dev_err(&adapter->pdev->dev, + "INTx request IRQ failed - err %d\n", status); + return status; + } +done: + adapter->isr_registered = true; + return 0; +} + +static void be_irq_unregister(struct be_adapter *adapter) +{ + struct net_device *netdev = adapter->netdev; + int vec; + + if (!adapter->isr_registered) + return; + + /* INTx */ + if (!adapter->msix_enabled) { + free_irq(netdev->irq, adapter); + goto done; + } + + /* MSIx */ + vec = be_msix_vec_get(adapter, adapter->tx_eq.q.id); + free_irq(vec, adapter); + vec = be_msix_vec_get(adapter, adapter->rx_eq.q.id); + free_irq(vec, adapter); +done: + adapter->isr_registered = false; + return; +} + +static int be_open(struct net_device *netdev) +{ + struct be_adapter *adapter = netdev_priv(netdev); + struct be_ctrl_info *ctrl = &adapter->ctrl; + struct be_eq_obj *rx_eq = &adapter->rx_eq; + struct be_eq_obj *tx_eq = &adapter->tx_eq; + u32 if_flags; + int status; + + if_flags = BE_IF_FLAGS_BROADCAST | BE_IF_FLAGS_PROMISCUOUS | + BE_IF_FLAGS_MCAST_PROMISCUOUS | BE_IF_FLAGS_UNTAGGED | + BE_IF_FLAGS_PASS_L3L4_ERRORS; + status = be_cmd_if_create(ctrl, if_flags, netdev->dev_addr, + false/* pmac_invalid */, &adapter->if_handle, + &adapter->pmac_id); + if (status != 0) + goto do_none; + + status = be_cmd_set_flow_control(ctrl, true, true); + if (status != 0) + goto if_destroy; + + status = be_tx_queues_create(adapter); + if (status != 0) + goto if_destroy; + + status = be_rx_queues_create(adapter); + if (status != 0) + goto tx_qs_destroy; + + /* First time posting */ + be_post_rx_frags(adapter); + + napi_enable(&rx_eq->napi); + napi_enable(&tx_eq->napi); + + be_irq_register(adapter); + + be_intr_set(ctrl, true); + + /* The evt queues are created in the unarmed state; arm them */ + be_eq_notify(ctrl, rx_eq->q.id, true, false, 0); + be_eq_notify(ctrl, tx_eq->q.id, true, false, 0); + + /* The compl queues are created in the unarmed state; arm them */ + be_cq_notify(ctrl, adapter->rx_obj.cq.id, true, 0); + be_cq_notify(ctrl, adapter->tx_obj.cq.id, true, 0); + + be_link_status_update(adapter); + + schedule_delayed_work(&adapter->work, msecs_to_jiffies(100)); + return 0; + +tx_qs_destroy: + be_tx_queues_destroy(adapter); +if_destroy: + be_cmd_if_destroy(ctrl, adapter->if_handle); +do_none: + return status; +} + +static int be_close(struct net_device *netdev) +{ + struct be_adapter *adapter = netdev_priv(netdev); + struct be_ctrl_info *ctrl = &adapter->ctrl; + struct be_eq_obj *rx_eq = &adapter->rx_eq; + struct be_eq_obj *tx_eq = &adapter->tx_eq; + int vec; + + cancel_delayed_work(&adapter->work); + + netif_stop_queue(netdev); + netif_carrier_off(netdev); + adapter->link.speed = PHY_LINK_SPEED_ZERO; + + be_intr_set(ctrl, false); + + if (adapter->msix_enabled) { + vec = be_msix_vec_get(adapter, tx_eq->q.id); + synchronize_irq(vec); + vec = be_msix_vec_get(adapter, rx_eq->q.id); + synchronize_irq(vec); + } else { + synchronize_irq(netdev->irq); + } + be_irq_unregister(adapter); + + napi_disable(&rx_eq->napi); + napi_disable(&tx_eq->napi); + + be_rx_queues_destroy(adapter); + be_tx_queues_destroy(adapter); + + be_cmd_if_destroy(ctrl, adapter->if_handle); + return 0; +} + +static int be_get_frag_header(struct skb_frag_struct *frag, void **mac_hdr, + void **ip_hdr, void **tcpudp_hdr, + u64 *hdr_flags, void *priv) +{ + struct ethhdr *eh; + struct vlan_ethhdr *veh; + struct iphdr *iph; + u8 *va = page_address(frag->page) + frag->page_offset; + unsigned long ll_hlen; + + prefetch(va); + eh = (struct ethhdr *)va; + *mac_hdr = eh; + ll_hlen = ETH_HLEN; + if (eh->h_proto != htons(ETH_P_IP)) { + if (eh->h_proto == htons(ETH_P_8021Q)) { + veh = (struct vlan_ethhdr *)va; + if (veh->h_vlan_encapsulated_proto != htons(ETH_P_IP)) + return -1; + + ll_hlen += VLAN_HLEN; + } else { + return -1; + } + } + *hdr_flags = LRO_IPV4; + iph = (struct iphdr *)(va + ll_hlen); + *ip_hdr = iph; + if (iph->protocol != IPPROTO_TCP) + return -1; + *hdr_flags |= LRO_TCP; + *tcpudp_hdr = (u8 *) (*ip_hdr) + (iph->ihl << 2); + + return 0; +} + +static void be_lro_init(struct be_adapter *adapter, struct net_device *netdev) +{ + struct net_lro_mgr *lro_mgr; + + lro_mgr = &adapter->rx_obj.lro_mgr; + lro_mgr->dev = netdev; + lro_mgr->features = LRO_F_NAPI; + lro_mgr->ip_summed = CHECKSUM_UNNECESSARY; + lro_mgr->ip_summed_aggr = CHECKSUM_UNNECESSARY; + lro_mgr->max_desc = BE_MAX_LRO_DESCRIPTORS; + lro_mgr->lro_arr = adapter->rx_obj.lro_desc; + lro_mgr->get_frag_header = be_get_frag_header; + lro_mgr->max_aggr = BE_MAX_FRAGS_PER_FRAME; +} + +static struct net_device_ops be_netdev_ops = { + .ndo_open = be_open, + .ndo_stop = be_close, + .ndo_start_xmit = be_xmit, + .ndo_get_stats = be_get_stats, + .ndo_set_rx_mode = be_set_multicast_list, + .ndo_set_mac_address = be_mac_addr_set, + .ndo_change_mtu = be_change_mtu, + .ndo_validate_addr = eth_validate_addr, + .ndo_vlan_rx_register = be_vlan_register, + .ndo_vlan_rx_add_vid = be_vlan_add_vid, + .ndo_vlan_rx_kill_vid = be_vlan_rem_vid, +}; + +static void be_netdev_init(struct net_device *netdev) +{ + struct be_adapter *adapter = netdev_priv(netdev); + + netdev->features |= NETIF_F_SG | NETIF_F_HW_VLAN_RX | NETIF_F_TSO | + NETIF_F_HW_VLAN_TX | NETIF_F_HW_VLAN_FILTER | NETIF_F_IP_CSUM | + NETIF_F_IPV6_CSUM | NETIF_F_TSO6; + + netdev->flags |= IFF_MULTICAST; + + BE_SET_NETDEV_OPS(netdev, &be_netdev_ops); + + SET_ETHTOOL_OPS(netdev, &be_ethtool_ops); + + be_lro_init(adapter, netdev); + + netif_napi_add(netdev, &adapter->rx_eq.napi, be_poll_rx, + BE_NAPI_WEIGHT); + netif_napi_add(netdev, &adapter->tx_eq.napi, be_poll_tx, + BE_NAPI_WEIGHT); + + netif_carrier_off(netdev); + netif_stop_queue(netdev); +} + +static void be_unmap_pci_bars(struct be_adapter *adapter) +{ + struct be_ctrl_info *ctrl = &adapter->ctrl; + if (ctrl->csr) + iounmap(ctrl->csr); + if (ctrl->db) + iounmap(ctrl->db); + if (ctrl->pcicfg) + iounmap(ctrl->pcicfg); +} + +static int be_map_pci_bars(struct be_adapter *adapter) +{ + u8 __iomem *addr; + + addr = ioremap_nocache(pci_resource_start(adapter->pdev, 2), + pci_resource_len(adapter->pdev, 2)); + if (addr == NULL) + return -ENOMEM; + adapter->ctrl.csr = addr; + + addr = ioremap_nocache(pci_resource_start(adapter->pdev, 4), + 128 * 1024); + if (addr == NULL) + goto pci_map_err; + adapter->ctrl.db = addr; + + addr = ioremap_nocache(pci_resource_start(adapter->pdev, 1), + pci_resource_len(adapter->pdev, 1)); + if (addr == NULL) + goto pci_map_err; + adapter->ctrl.pcicfg = addr; + + return 0; +pci_map_err: + be_unmap_pci_bars(adapter); + return -ENOMEM; +} + + +static void be_ctrl_cleanup(struct be_adapter *adapter) +{ + struct be_dma_mem *mem = &adapter->ctrl.mbox_mem_alloced; + + be_unmap_pci_bars(adapter); + + if (mem->va) + pci_free_consistent(adapter->pdev, mem->size, + mem->va, mem->dma); +} + +/* Initialize the mbox required to send cmds to BE */ +static int be_ctrl_init(struct be_adapter *adapter) +{ + struct be_ctrl_info *ctrl = &adapter->ctrl; + struct be_dma_mem *mbox_mem_alloc = &ctrl->mbox_mem_alloced; + struct be_dma_mem *mbox_mem_align = &ctrl->mbox_mem; + int status; + u32 val; + + status = be_map_pci_bars(adapter); + if (status) + return status; + + mbox_mem_alloc->size = sizeof(struct be_mcc_mailbox) + 16; + mbox_mem_alloc->va = pci_alloc_consistent(adapter->pdev, + mbox_mem_alloc->size, &mbox_mem_alloc->dma); + if (!mbox_mem_alloc->va) { + be_unmap_pci_bars(adapter); + return -1; + } + mbox_mem_align->size = sizeof(struct be_mcc_mailbox); + mbox_mem_align->va = PTR_ALIGN(mbox_mem_alloc->va, 16); + mbox_mem_align->dma = PTR_ALIGN(mbox_mem_alloc->dma, 16); + memset(mbox_mem_align->va, 0, sizeof(struct be_mcc_mailbox)); + spin_lock_init(&ctrl->cmd_lock); + + val = ioread32(ctrl->pcicfg + PCICFG_MEMBAR_CTRL_INT_CTRL_OFFSET); + ctrl->pci_func = (val >> MEMBAR_CTRL_INT_CTRL_PFUNC_SHIFT) & + MEMBAR_CTRL_INT_CTRL_PFUNC_MASK; + return 0; +} + +static void be_stats_cleanup(struct be_adapter *adapter) +{ + struct be_stats_obj *stats = &adapter->stats; + struct be_dma_mem *cmd = &stats->cmd; + + if (cmd->va) + pci_free_consistent(adapter->pdev, cmd->size, + cmd->va, cmd->dma); +} + +static int be_stats_init(struct be_adapter *adapter) +{ + struct be_stats_obj *stats = &adapter->stats; + struct be_dma_mem *cmd = &stats->cmd; + + cmd->size = sizeof(struct be_cmd_req_get_stats); + cmd->va = pci_alloc_consistent(adapter->pdev, cmd->size, &cmd->dma); + if (cmd->va == NULL) + return -1; + return 0; +} + +static void __devexit be_remove(struct pci_dev *pdev) +{ + struct be_adapter *adapter = pci_get_drvdata(pdev); + if (!adapter) + return; + + unregister_netdev(adapter->netdev); + + be_stats_cleanup(adapter); + + be_ctrl_cleanup(adapter); + + if (adapter->msix_enabled) { + pci_disable_msix(adapter->pdev); + adapter->msix_enabled = false; + } + + pci_set_drvdata(pdev, NULL); + pci_release_regions(pdev); + pci_disable_device(pdev); + + free_netdev(adapter->netdev); +} + +static int be_hw_up(struct be_adapter *adapter) +{ + struct be_ctrl_info *ctrl = &adapter->ctrl; + int status; + + status = be_cmd_POST(ctrl); + if (status) + return status; + + status = be_cmd_get_fw_ver(ctrl, adapter->fw_ver); + if (status) + return status; + + status = be_cmd_query_fw_cfg(ctrl, &adapter->port_num); + return status; +} + +static int __devinit be_probe(struct pci_dev *pdev, + const struct pci_device_id *pdev_id) +{ + int status = 0; + struct be_adapter *adapter; + struct net_device *netdev; + struct be_ctrl_info *ctrl; + u8 mac[ETH_ALEN]; + + status = pci_enable_device(pdev); + if (status) + goto do_none; + + status = pci_request_regions(pdev, DRV_NAME); + if (status) + goto disable_dev; + pci_set_master(pdev); + + netdev = alloc_etherdev(sizeof(struct be_adapter)); + if (netdev == NULL) { + status = -ENOMEM; + goto rel_reg; + } + adapter = netdev_priv(netdev); + adapter->pdev = pdev; + pci_set_drvdata(pdev, adapter); + adapter->netdev = netdev; + + be_msix_enable(adapter); + + status = pci_set_dma_mask(pdev, DMA_64BIT_MASK); + if (!status) { + netdev->features |= NETIF_F_HIGHDMA; + } else { + status = pci_set_dma_mask(pdev, DMA_32BIT_MASK); + if (status) { + dev_err(&pdev->dev, "Could not set PCI DMA Mask\n"); + goto free_netdev; + } + } + + ctrl = &adapter->ctrl; + status = be_ctrl_init(adapter); + if (status) + goto free_netdev; + + status = be_stats_init(adapter); + if (status) + goto ctrl_clean; + + status = be_hw_up(adapter); + if (status) + goto stats_clean; + + status = be_cmd_mac_addr_query(ctrl, mac, MAC_ADDRESS_TYPE_NETWORK, + true /* permanent */, 0); + if (status) + goto stats_clean; + memcpy(netdev->dev_addr, mac, ETH_ALEN); + + INIT_DELAYED_WORK(&adapter->work, be_worker); + be_netdev_init(netdev); + SET_NETDEV_DEV(netdev, &adapter->pdev->dev); + + status = register_netdev(netdev); + if (status != 0) + goto stats_clean; + + dev_info(&pdev->dev, BE_NAME " port %d\n", adapter->port_num); + return 0; + +stats_clean: + be_stats_cleanup(adapter); +ctrl_clean: + be_ctrl_cleanup(adapter); +free_netdev: + free_netdev(adapter->netdev); +rel_reg: + pci_release_regions(pdev); +disable_dev: + pci_disable_device(pdev); +do_none: + dev_warn(&pdev->dev, BE_NAME " initialization failed\n"); + return status; +} + +static int be_suspend(struct pci_dev *pdev, pm_message_t state) +{ + struct be_adapter *adapter = pci_get_drvdata(pdev); + struct net_device *netdev = adapter->netdev; + + netif_device_detach(netdev); + if (netif_running(netdev)) { + rtnl_lock(); + be_close(netdev); + rtnl_unlock(); + } + + pci_save_state(pdev); + pci_disable_device(pdev); + pci_set_power_state(pdev, pci_choose_state(pdev, state)); + return 0; +} + +static int be_resume(struct pci_dev *pdev) +{ + int status = 0; + struct be_adapter *adapter = pci_get_drvdata(pdev); + struct net_device *netdev = adapter->netdev; + + netif_device_detach(netdev); + + status = pci_enable_device(pdev); + if (status) + return status; + + pci_set_power_state(pdev, 0); + pci_restore_state(pdev); + + be_vids_config(netdev); + + if (netif_running(netdev)) { + rtnl_lock(); + be_open(netdev); + rtnl_unlock(); + } + netif_device_attach(netdev); + return 0; +} + +static struct pci_driver be_driver = { + .name = DRV_NAME, + .id_table = be_dev_ids, + .probe = be_probe, + .remove = be_remove, + .suspend = be_suspend, + .resume = be_resume +}; + +static int __init be_init_module(void) +{ + if (rx_frag_size != 8192 && rx_frag_size != 4096 + && rx_frag_size != 2048) { + printk(KERN_WARNING DRV_NAME + " : Module param rx_frag_size must be 2048/4096/8192." + " Using 2048\n"); + rx_frag_size = 2048; + } + /* Ensure rx_frag_size is aligned to chache line */ + if (SKB_DATA_ALIGN(rx_frag_size) != rx_frag_size) { + printk(KERN_WARNING DRV_NAME + " : Bad module param rx_frag_size. Using 2048\n"); + rx_frag_size = 2048; + } + + return pci_register_driver(&be_driver); +} +module_init(be_init_module); + +static void __exit be_exit_module(void) +{ + pci_unregister_driver(&be_driver); +} +module_exit(be_exit_module); diff --git a/drivers/net/dnet.c b/drivers/net/dnet.c new file mode 100644 index 000000000000..5c347f70cb67 --- /dev/null +++ b/drivers/net/dnet.c @@ -0,0 +1,994 @@ +/* + * Dave DNET Ethernet Controller driver + * + * Copyright (C) 2008 Dave S.r.l. <www.dave.eu> + * Copyright (C) 2009 Ilya Yanok, Emcraft Systems Ltd, <yanok@emcraft.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ +#include <linux/version.h> +#include <linux/module.h> +#include <linux/moduleparam.h> +#include <linux/kernel.h> +#include <linux/types.h> +#include <linux/slab.h> +#include <linux/delay.h> +#include <linux/init.h> +#include <linux/netdevice.h> +#include <linux/etherdevice.h> +#include <linux/dma-mapping.h> +#include <linux/platform_device.h> +#include <linux/phy.h> +#include <linux/platform_device.h> + +#include "dnet.h" + +#undef DEBUG + +/* function for reading internal MAC register */ +u16 dnet_readw_mac(struct dnet *bp, u16 reg) +{ + u16 data_read; + + /* issue a read */ + dnet_writel(bp, reg, MACREG_ADDR); + + /* since a read/write op to the MAC is very slow, + * we must wait before reading the data */ + ndelay(500); + + /* read data read from the MAC register */ + data_read = dnet_readl(bp, MACREG_DATA); + + /* all done */ + return data_read; +} + +/* function for writing internal MAC register */ +void dnet_writew_mac(struct dnet *bp, u16 reg, u16 val) +{ + /* load data to write */ + dnet_writel(bp, val, MACREG_DATA); + + /* issue a write */ + dnet_writel(bp, reg | DNET_INTERNAL_WRITE, MACREG_ADDR); + + /* since a read/write op to the MAC is very slow, + * we must wait before exiting */ + ndelay(500); +} + +static void __dnet_set_hwaddr(struct dnet *bp) +{ + u16 tmp; + + tmp = cpu_to_be16(*((u16 *) bp->dev->dev_addr)); + dnet_writew_mac(bp, DNET_INTERNAL_MAC_ADDR_0_REG, tmp); + tmp = cpu_to_be16(*((u16 *) (bp->dev->dev_addr + 2))); + dnet_writew_mac(bp, DNET_INTERNAL_MAC_ADDR_1_REG, tmp); + tmp = cpu_to_be16(*((u16 *) (bp->dev->dev_addr + 4))); + dnet_writew_mac(bp, DNET_INTERNAL_MAC_ADDR_2_REG, tmp); +} + +static void __devinit dnet_get_hwaddr(struct dnet *bp) +{ + u16 tmp; + u8 addr[6]; + + /* + * from MAC docs: + * "Note that the MAC address is stored in the registers in Hexadecimal + * form. For example, to set the MAC Address to: AC-DE-48-00-00-80 + * would require writing 0xAC (octet 0) to address 0x0B (high byte of + * Mac_addr[15:0]), 0xDE (octet 1) to address 0x0A (Low byte of + * Mac_addr[15:0]), 0x48 (octet 2) to address 0x0D (high byte of + * Mac_addr[15:0]), 0x00 (octet 3) to address 0x0C (Low byte of + * Mac_addr[15:0]), 0x00 (octet 4) to address 0x0F (high byte of + * Mac_addr[15:0]), and 0x80 (octet 5) to address * 0x0E (Low byte of + * Mac_addr[15:0]). + */ + tmp = dnet_readw_mac(bp, DNET_INTERNAL_MAC_ADDR_0_REG); + *((u16 *) addr) = be16_to_cpu(tmp); + tmp = dnet_readw_mac(bp, DNET_INTERNAL_MAC_ADDR_1_REG); + *((u16 *) (addr + 2)) = be16_to_cpu(tmp); + tmp = dnet_readw_mac(bp, DNET_INTERNAL_MAC_ADDR_2_REG); + *((u16 *) (addr + 4)) = be16_to_cpu(tmp); + + if (is_valid_ether_addr(addr)) + memcpy(bp->dev->dev_addr, addr, sizeof(addr)); +} + +static int dnet_mdio_read(struct mii_bus *bus, int mii_id, int regnum) +{ + struct dnet *bp = bus->priv; + u16 value; + + while (!(dnet_readw_mac(bp, DNET_INTERNAL_GMII_MNG_CTL_REG) + & DNET_INTERNAL_GMII_MNG_CMD_FIN)) + cpu_relax(); + + /* only 5 bits allowed for phy-addr and reg_offset */ + mii_id &= 0x1f; + regnum &= 0x1f; + + /* prepare reg_value for a read */ + value = (mii_id << 8); + value |= regnum; + + /* write control word */ + dnet_writew_mac(bp, DNET_INTERNAL_GMII_MNG_CTL_REG, value); + + /* wait for end of transfer */ + while (!(dnet_readw_mac(bp, DNET_INTERNAL_GMII_MNG_CTL_REG) + & DNET_INTERNAL_GMII_MNG_CMD_FIN)) + cpu_relax(); + + value = dnet_readw_mac(bp, DNET_INTERNAL_GMII_MNG_DAT_REG); + + pr_debug("mdio_read %02x:%02x <- %04x\n", mii_id, regnum, value); + + return value; +} + +static int dnet_mdio_write(struct mii_bus *bus, int mii_id, int regnum, + u16 value) +{ + struct dnet *bp = bus->priv; + u16 tmp; + + pr_debug("mdio_write %02x:%02x <- %04x\n", mii_id, regnum, value); + + while (!(dnet_readw_mac(bp, DNET_INTERNAL_GMII_MNG_CTL_REG) + & DNET_INTERNAL_GMII_MNG_CMD_FIN)) + cpu_relax(); + + /* prepare for a write operation */ + tmp = (1 << 13); + + /* only 5 bits allowed for phy-addr and reg_offset */ + mii_id &= 0x1f; + regnum &= 0x1f; + + /* only 16 bits on data */ + value &= 0xffff; + + /* prepare reg_value for a write */ + tmp |= (mii_id << 8); + tmp |= regnum; + + /* write data to write first */ + dnet_writew_mac(bp, DNET_INTERNAL_GMII_MNG_DAT_REG, value); + + /* write control word */ + dnet_writew_mac(bp, DNET_INTERNAL_GMII_MNG_CTL_REG, tmp); + + while (!(dnet_readw_mac(bp, DNET_INTERNAL_GMII_MNG_CTL_REG) + & DNET_INTERNAL_GMII_MNG_CMD_FIN)) + cpu_relax(); + + return 0; +} + +static int dnet_mdio_reset(struct mii_bus *bus) +{ + return 0; +} + +static void dnet_handle_link_change(struct net_device *dev) +{ + struct dnet *bp = netdev_priv(dev); + struct phy_device *phydev = bp->phy_dev; + unsigned long flags; + u32 mode_reg, ctl_reg; + + int status_change = 0; + + spin_lock_irqsave(&bp->lock, flags); + + mode_reg = dnet_readw_mac(bp, DNET_INTERNAL_MODE_REG); + ctl_reg = dnet_readw_mac(bp, DNET_INTERNAL_RXTX_CONTROL_REG); + + if (phydev->link) { + if (bp->duplex != phydev->duplex) { + if (phydev->duplex) + ctl_reg &= + ~(DNET_INTERNAL_RXTX_CONTROL_ENABLEHALFDUP); + else + ctl_reg |= + DNET_INTERNAL_RXTX_CONTROL_ENABLEHALFDUP; + + bp->duplex = phydev->duplex; + status_change = 1; + } + + if (bp->speed != phydev->speed) { + status_change = 1; + switch (phydev->speed) { + case 1000: + mode_reg |= DNET_INTERNAL_MODE_GBITEN; + break; + case 100: + case 10: + mode_reg &= ~DNET_INTERNAL_MODE_GBITEN; + break; + default: + printk(KERN_WARNING + "%s: Ack! Speed (%d) is not " + "10/100/1000!\n", dev->name, + phydev->speed); + break; + } + bp->speed = phydev->speed; + } + } + + if (phydev->link != bp->link) { + if (phydev->link) { + mode_reg |= + (DNET_INTERNAL_MODE_RXEN | DNET_INTERNAL_MODE_TXEN); + } else { + mode_reg &= + ~(DNET_INTERNAL_MODE_RXEN | + DNET_INTERNAL_MODE_TXEN); + bp->speed = 0; + bp->duplex = -1; + } + bp->link = phydev->link; + + status_change = 1; + } + + if (status_change) { + dnet_writew_mac(bp, DNET_INTERNAL_RXTX_CONTROL_REG, ctl_reg); + dnet_writew_mac(bp, DNET_INTERNAL_MODE_REG, mode_reg); + } + + spin_unlock_irqrestore(&bp->lock, flags); + + if (status_change) { + if (phydev->link) + printk(KERN_INFO "%s: link up (%d/%s)\n", + dev->name, phydev->speed, + DUPLEX_FULL == phydev->duplex ? "Full" : "Half"); + else + printk(KERN_INFO "%s: link down\n", dev->name); + } +} + +static int dnet_mii_probe(struct net_device *dev) +{ + struct dnet *bp = netdev_priv(dev); + struct phy_device *phydev = NULL; + int phy_addr; + + /* find the first phy */ + for (phy_addr = 0; phy_addr < PHY_MAX_ADDR; phy_addr++) { + if (bp->mii_bus->phy_map[phy_addr]) { + phydev = bp->mii_bus->phy_map[phy_addr]; + break; + } + } + + if (!phydev) { + printk(KERN_ERR "%s: no PHY found\n", dev->name); + return -ENODEV; + } + + /* TODO : add pin_irq */ + + /* attach the mac to the phy */ + if (bp->capabilities & DNET_HAS_RMII) { + phydev = phy_connect(dev, phydev->dev.bus_id, + &dnet_handle_link_change, 0, + PHY_INTERFACE_MODE_RMII); + } else { + phydev = phy_connect(dev, phydev->dev.bus_id, + &dnet_handle_link_change, 0, + PHY_INTERFACE_MODE_MII); + } + + if (IS_ERR(phydev)) { + printk(KERN_ERR "%s: Could not attach to PHY\n", dev->name); + return PTR_ERR(phydev); + } + + /* mask with MAC supported features */ + if (bp->capabilities & DNET_HAS_GIGABIT) + phydev->supported &= PHY_GBIT_FEATURES; + else + phydev->supported &= PHY_BASIC_FEATURES; + + phydev->supported |= SUPPORTED_Asym_Pause | SUPPORTED_Pause; + + phydev->advertising = phydev->supported; + + bp->link = 0; + bp->speed = 0; + bp->duplex = -1; + bp->phy_dev = phydev; + + return 0; +} + +static int dnet_mii_init(struct dnet *bp) +{ + int err, i; + + bp->mii_bus = mdiobus_alloc(); + if (bp->mii_bus == NULL) + return -ENOMEM; + + bp->mii_bus->name = "dnet_mii_bus"; + bp->mii_bus->read = &dnet_mdio_read; + bp->mii_bus->write = &dnet_mdio_write; + bp->mii_bus->reset = &dnet_mdio_reset; + + snprintf(bp->mii_bus->id, MII_BUS_ID_SIZE, "%x", 0); + + bp->mii_bus->priv = bp; + + bp->mii_bus->irq = kmalloc(sizeof(int) * PHY_MAX_ADDR, GFP_KERNEL); + if (!bp->mii_bus->irq) { + err = -ENOMEM; + goto err_out; + } + + for (i = 0; i < PHY_MAX_ADDR; i++) + bp->mii_bus->irq[i] = PHY_POLL; + + platform_set_drvdata(bp->dev, bp->mii_bus); + + if (mdiobus_register(bp->mii_bus)) { + err = -ENXIO; + goto err_out_free_mdio_irq; + } + + if (dnet_mii_probe(bp->dev) != 0) { + err = -ENXIO; + goto err_out_unregister_bus; + } + + return 0; + +err_out_unregister_bus: + mdiobus_unregister(bp->mii_bus); +err_out_free_mdio_irq: + kfree(bp->mii_bus->irq); +err_out: + mdiobus_free(bp->mii_bus); + return err; +} + +/* For Neptune board: LINK1000 as Link LED and TX as activity LED */ +int dnet_phy_marvell_fixup(struct phy_device *phydev) +{ + return phy_write(phydev, 0x18, 0x4148); +} + +static void dnet_update_stats(struct dnet *bp) +{ + u32 __iomem *reg = bp->regs + DNET_RX_PKT_IGNR_CNT; + u32 *p = &bp->hw_stats.rx_pkt_ignr; + u32 *end = &bp->hw_stats.rx_byte + 1; + + WARN_ON((unsigned long)(end - p - 1) != + (DNET_RX_BYTE_CNT - DNET_RX_PKT_IGNR_CNT) / 4); + + for (; p < end; p++, reg++) + *p += readl(reg); + + reg = bp->regs + DNET_TX_UNICAST_CNT; + p = &bp->hw_stats.tx_unicast; + end = &bp->hw_stats.tx_byte + 1; + + WARN_ON((unsigned long)(end - p - 1) != + (DNET_TX_BYTE_CNT - DNET_TX_UNICAST_CNT) / 4); + + for (; p < end; p++, reg++) + *p += readl(reg); +} + +static int dnet_poll(struct napi_struct *napi, int budget) +{ + struct dnet *bp = container_of(napi, struct dnet, napi); + struct net_device *dev = bp->dev; + int npackets = 0; + unsigned int pkt_len; + struct sk_buff *skb; + unsigned int *data_ptr; + u32 int_enable; + u32 cmd_word; + int i; + + while (npackets < budget) { + /* + * break out of while loop if there are no more + * packets waiting + */ + if (!(dnet_readl(bp, RX_FIFO_WCNT) >> 16)) { + napi_complete(napi); + int_enable = dnet_readl(bp, INTR_ENB); + int_enable |= DNET_INTR_SRC_RX_CMDFIFOAF; + dnet_writel(bp, int_enable, INTR_ENB); + return 0; + } + + cmd_word = dnet_readl(bp, RX_LEN_FIFO); + pkt_len = cmd_word & 0xFFFF; + + if (cmd_word & 0xDF180000) + printk(KERN_ERR "%s packet receive error %x\n", + __func__, cmd_word); + + skb = dev_alloc_skb(pkt_len + 5); + if (skb != NULL) { + /* Align IP on 16 byte boundaries */ + skb_reserve(skb, 2); + /* + * 'skb_put()' points to the start of sk_buff + * data area. + */ + data_ptr = (unsigned int *)skb_put(skb, pkt_len); + for (i = 0; i < (pkt_len + 3) >> 2; i++) + *data_ptr++ = dnet_readl(bp, RX_DATA_FIFO); + skb->protocol = eth_type_trans(skb, dev); + netif_receive_skb(skb); + npackets++; + } else + printk(KERN_NOTICE + "%s: No memory to allocate a sk_buff of " + "size %u.\n", dev->name, pkt_len); + } + + budget -= npackets; + + if (npackets < budget) { + /* We processed all packets available. Tell NAPI it can + * stop polling then re-enable rx interrupts */ + napi_complete(napi); + int_enable = dnet_readl(bp, INTR_ENB); + int_enable |= DNET_INTR_SRC_RX_CMDFIFOAF; + dnet_writel(bp, int_enable, INTR_ENB); + return 0; + } + + /* There are still packets waiting */ + return 1; +} + +static irqreturn_t dnet_interrupt(int irq, void *dev_id) +{ + struct net_device *dev = dev_id; + struct dnet *bp = netdev_priv(dev); + u32 int_src, int_enable, int_current; + unsigned long flags; + unsigned int handled = 0; + + spin_lock_irqsave(&bp->lock, flags); + + /* read and clear the DNET irq (clear on read) */ + int_src = dnet_readl(bp, INTR_SRC); + int_enable = dnet_readl(bp, INTR_ENB); + int_current = int_src & int_enable; + + /* restart the queue if we had stopped it for TX fifo almost full */ + if (int_current & DNET_INTR_SRC_TX_FIFOAE) { + int_enable = dnet_readl(bp, INTR_ENB); + int_enable &= ~DNET_INTR_ENB_TX_FIFOAE; + dnet_writel(bp, int_enable, INTR_ENB); + netif_wake_queue(dev); + handled = 1; + } + + /* RX FIFO error checking */ + if (int_current & + (DNET_INTR_SRC_RX_CMDFIFOFF | DNET_INTR_SRC_RX_DATAFIFOFF)) { + printk(KERN_ERR "%s: RX fifo error %x, irq %x\n", __func__, + dnet_readl(bp, RX_STATUS), int_current); + /* we can only flush the RX FIFOs */ + dnet_writel(bp, DNET_SYS_CTL_RXFIFOFLUSH, SYS_CTL); + ndelay(500); + dnet_writel(bp, 0, SYS_CTL); + handled = 1; + } + + /* TX FIFO error checking */ + if (int_current & + (DNET_INTR_SRC_TX_FIFOFULL | DNET_INTR_SRC_TX_DISCFRM)) { + printk(KERN_ERR "%s: TX fifo error %x, irq %x\n", __func__, + dnet_readl(bp, TX_STATUS), int_current); + /* we can only flush the TX FIFOs */ + dnet_writel(bp, DNET_SYS_CTL_TXFIFOFLUSH, SYS_CTL); + ndelay(500); + dnet_writel(bp, 0, SYS_CTL); + handled = 1; + } + + if (int_current & DNET_INTR_SRC_RX_CMDFIFOAF) { + if (napi_schedule_prep(&bp->napi)) { + /* + * There's no point taking any more interrupts + * until we have processed the buffers + */ + /* Disable Rx interrupts and schedule NAPI poll */ + int_enable = dnet_readl(bp, INTR_ENB); + int_enable &= ~DNET_INTR_SRC_RX_CMDFIFOAF; + dnet_writel(bp, int_enable, INTR_ENB); + __napi_schedule(&bp->napi); + } + handled = 1; + } + + if (!handled) + pr_debug("%s: irq %x remains\n", __func__, int_current); + + spin_unlock_irqrestore(&bp->lock, flags); + + return IRQ_RETVAL(handled); +} + +#ifdef DEBUG +static inline void dnet_print_skb(struct sk_buff *skb) +{ + int k; + printk(KERN_DEBUG PFX "data:"); + for (k = 0; k < skb->len; k++) + printk(" %02x", (unsigned int)skb->data[k]); + printk("\n"); +} +#else +#define dnet_print_skb(skb) do {} while (0) +#endif + +static int dnet_start_xmit(struct sk_buff *skb, struct net_device *dev) +{ + + struct dnet *bp = netdev_priv(dev); + u32 tx_status, irq_enable; + unsigned int len, i, tx_cmd, wrsz; + unsigned long flags; + unsigned int *bufp; + + tx_status = dnet_readl(bp, TX_STATUS); + + pr_debug("start_xmit: len %u head %p data %p\n", + skb->len, skb->head, skb->data); + dnet_print_skb(skb); + + /* frame size (words) */ + len = (skb->len + 3) >> 2; + + spin_lock_irqsave(&bp->lock, flags); + + tx_status = dnet_readl(bp, TX_STATUS); + + bufp = (unsigned int *)(((unsigned long) skb->data) & ~0x3UL); + wrsz = (u32) skb->len + 3; + wrsz += ((unsigned long) skb->data) & 0x3; + wrsz >>= 2; + tx_cmd = ((((unsigned long)(skb->data)) & 0x03) << 16) | (u32) skb->len; + + /* check if there is enough room for the current frame */ + if (wrsz < (DNET_FIFO_SIZE - dnet_readl(bp, TX_FIFO_WCNT))) { + for (i = 0; i < wrsz; i++) + dnet_writel(bp, *bufp++, TX_DATA_FIFO); + + /* + * inform MAC that a packet's written and ready to be + * shipped out + */ + dnet_writel(bp, tx_cmd, TX_LEN_FIFO); + } + + if (dnet_readl(bp, TX_FIFO_WCNT) > DNET_FIFO_TX_DATA_AF_TH) { + netif_stop_queue(dev); + tx_status = dnet_readl(bp, INTR_SRC); + irq_enable = dnet_readl(bp, INTR_ENB); + irq_enable |= DNET_INTR_ENB_TX_FIFOAE; + dnet_writel(bp, irq_enable, INTR_ENB); + } + + /* free the buffer */ + dev_kfree_skb(skb); + + spin_unlock_irqrestore(&bp->lock, flags); + + dev->trans_start = jiffies; + + return 0; +} + +static void dnet_reset_hw(struct dnet *bp) +{ + /* put ts_mac in IDLE state i.e. disable rx/tx */ + dnet_writew_mac(bp, DNET_INTERNAL_MODE_REG, DNET_INTERNAL_MODE_FCEN); + + /* + * RX FIFO almost full threshold: only cmd FIFO almost full is + * implemented for RX side + */ + dnet_writel(bp, DNET_FIFO_RX_CMD_AF_TH, RX_FIFO_TH); + /* + * TX FIFO almost empty threshold: only data FIFO almost empty + * is implemented for TX side + */ + dnet_writel(bp, DNET_FIFO_TX_DATA_AE_TH, TX_FIFO_TH); + + /* flush rx/tx fifos */ + dnet_writel(bp, DNET_SYS_CTL_RXFIFOFLUSH | DNET_SYS_CTL_TXFIFOFLUSH, + SYS_CTL); + msleep(1); + dnet_writel(bp, 0, SYS_CTL); +} + +static void dnet_init_hw(struct dnet *bp) +{ + u32 config; + + dnet_reset_hw(bp); + __dnet_set_hwaddr(bp); + + config = dnet_readw_mac(bp, DNET_INTERNAL_RXTX_CONTROL_REG); + + if (bp->dev->flags & IFF_PROMISC) + /* Copy All Frames */ + config |= DNET_INTERNAL_RXTX_CONTROL_ENPROMISC; + if (!(bp->dev->flags & IFF_BROADCAST)) + /* No BroadCast */ + config |= DNET_INTERNAL_RXTX_CONTROL_RXMULTICAST; + + config |= DNET_INTERNAL_RXTX_CONTROL_RXPAUSE | + DNET_INTERNAL_RXTX_CONTROL_RXBROADCAST | + DNET_INTERNAL_RXTX_CONTROL_DROPCONTROL | + DNET_INTERNAL_RXTX_CONTROL_DISCFXFCS; + + dnet_writew_mac(bp, DNET_INTERNAL_RXTX_CONTROL_REG, config); + + /* clear irq before enabling them */ + config = dnet_readl(bp, INTR_SRC); + + /* enable RX/TX interrupt, recv packet ready interrupt */ + dnet_writel(bp, DNET_INTR_ENB_GLOBAL_ENABLE | DNET_INTR_ENB_RX_SUMMARY | + DNET_INTR_ENB_TX_SUMMARY | DNET_INTR_ENB_RX_FIFOERR | + DNET_INTR_ENB_RX_ERROR | DNET_INTR_ENB_RX_FIFOFULL | + DNET_INTR_ENB_TX_FIFOFULL | DNET_INTR_ENB_TX_DISCFRM | + DNET_INTR_ENB_RX_PKTRDY, INTR_ENB); +} + +static int dnet_open(struct net_device *dev) +{ + struct dnet *bp = netdev_priv(dev); + + /* if the phy is not yet register, retry later */ + if (!bp->phy_dev) + return -EAGAIN; + + if (!is_valid_ether_addr(dev->dev_addr)) + return -EADDRNOTAVAIL; + + napi_enable(&bp->napi); + dnet_init_hw(bp); + + phy_start_aneg(bp->phy_dev); + + /* schedule a link state check */ + phy_start(bp->phy_dev); + + netif_start_queue(dev); + + return 0; +} + +static int dnet_close(struct net_device *dev) +{ + struct dnet *bp = netdev_priv(dev); + + netif_stop_queue(dev); + napi_disable(&bp->napi); + + if (bp->phy_dev) + phy_stop(bp->phy_dev); + + dnet_reset_hw(bp); + netif_carrier_off(dev); + + return 0; +} + +static inline void dnet_print_pretty_hwstats(struct dnet_stats *hwstat) +{ + pr_debug("%s\n", __func__); + pr_debug("----------------------------- RX statistics " + "-------------------------------\n"); + pr_debug("RX_PKT_IGNR_CNT %-8x\n", hwstat->rx_pkt_ignr); + pr_debug("RX_LEN_CHK_ERR_CNT %-8x\n", hwstat->rx_len_chk_err); + pr_debug("RX_LNG_FRM_CNT %-8x\n", hwstat->rx_lng_frm); + pr_debug("RX_SHRT_FRM_CNT %-8x\n", hwstat->rx_shrt_frm); + pr_debug("RX_IPG_VIOL_CNT %-8x\n", hwstat->rx_ipg_viol); + pr_debug("RX_CRC_ERR_CNT %-8x\n", hwstat->rx_crc_err); + pr_debug("RX_OK_PKT_CNT %-8x\n", hwstat->rx_ok_pkt); + pr_debug("RX_CTL_FRM_CNT %-8x\n", hwstat->rx_ctl_frm); + pr_debug("RX_PAUSE_FRM_CNT %-8x\n", hwstat->rx_pause_frm); + pr_debug("RX_MULTICAST_CNT %-8x\n", hwstat->rx_multicast); + pr_debug("RX_BROADCAST_CNT %-8x\n", hwstat->rx_broadcast); + pr_debug("RX_VLAN_TAG_CNT %-8x\n", hwstat->rx_vlan_tag); + pr_debug("RX_PRE_SHRINK_CNT %-8x\n", hwstat->rx_pre_shrink); + pr_debug("RX_DRIB_NIB_CNT %-8x\n", hwstat->rx_drib_nib); + pr_debug("RX_UNSUP_OPCD_CNT %-8x\n", hwstat->rx_unsup_opcd); + pr_debug("RX_BYTE_CNT %-8x\n", hwstat->rx_byte); + pr_debug("----------------------------- TX statistics " + "-------------------------------\n"); + pr_debug("TX_UNICAST_CNT %-8x\n", hwstat->tx_unicast); + pr_debug("TX_PAUSE_FRM_CNT %-8x\n", hwstat->tx_pause_frm); + pr_debug("TX_MULTICAST_CNT %-8x\n", hwstat->tx_multicast); + pr_debug("TX_BRDCAST_CNT %-8x\n", hwstat->tx_brdcast); + pr_debug("TX_VLAN_TAG_CNT %-8x\n", hwstat->tx_vlan_tag); + pr_debug("TX_BAD_FCS_CNT %-8x\n", hwstat->tx_bad_fcs); + pr_debug("TX_JUMBO_CNT %-8x\n", hwstat->tx_jumbo); + pr_debug("TX_BYTE_CNT %-8x\n", hwstat->tx_byte); +} + +static struct net_device_stats *dnet_get_stats(struct net_device *dev) +{ + + struct dnet *bp = netdev_priv(dev); + struct net_device_stats *nstat = &dev->stats; + struct dnet_stats *hwstat = &bp->hw_stats; + + /* read stats from hardware */ + dnet_update_stats(bp); + + /* Convert HW stats into netdevice stats */ + nstat->rx_errors = (hwstat->rx_len_chk_err + + hwstat->rx_lng_frm + hwstat->rx_shrt_frm + + /* ignore IGP violation error + hwstat->rx_ipg_viol + */ + hwstat->rx_crc_err + + hwstat->rx_pre_shrink + + hwstat->rx_drib_nib + hwstat->rx_unsup_opcd); + nstat->tx_errors = hwstat->tx_bad_fcs; + nstat->rx_length_errors = (hwstat->rx_len_chk_err + + hwstat->rx_lng_frm + + hwstat->rx_shrt_frm + hwstat->rx_pre_shrink); + nstat->rx_crc_errors = hwstat->rx_crc_err; + nstat->rx_frame_errors = hwstat->rx_pre_shrink + hwstat->rx_drib_nib; + nstat->rx_packets = hwstat->rx_ok_pkt; + nstat->tx_packets = (hwstat->tx_unicast + + hwstat->tx_multicast + hwstat->tx_brdcast); + nstat->rx_bytes = hwstat->rx_byte; + nstat->tx_bytes = hwstat->tx_byte; + nstat->multicast = hwstat->rx_multicast; + nstat->rx_missed_errors = hwstat->rx_pkt_ignr; + + dnet_print_pretty_hwstats(hwstat); + + return nstat; +} + +static int dnet_get_settings(struct net_device *dev, struct ethtool_cmd *cmd) +{ + struct dnet *bp = netdev_priv(dev); + struct phy_device *phydev = bp->phy_dev; + + if (!phydev) + return -ENODEV; + + return phy_ethtool_gset(phydev, cmd); +} + +static int dnet_set_settings(struct net_device *dev, struct ethtool_cmd *cmd) +{ + struct dnet *bp = netdev_priv(dev); + struct phy_device *phydev = bp->phy_dev; + + if (!phydev) + return -ENODEV; + + return phy_ethtool_sset(phydev, cmd); +} + +static int dnet_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) +{ + struct dnet *bp = netdev_priv(dev); + struct phy_device *phydev = bp->phy_dev; + + if (!netif_running(dev)) + return -EINVAL; + + if (!phydev) + return -ENODEV; + + return phy_mii_ioctl(phydev, if_mii(rq), cmd); +} + +static void dnet_get_drvinfo(struct net_device *dev, + struct ethtool_drvinfo *info) +{ + strcpy(info->driver, DRV_NAME); + strcpy(info->version, DRV_VERSION); + strcpy(info->bus_info, "0"); +} + +static const struct ethtool_ops dnet_ethtool_ops = { + .get_settings = dnet_get_settings, + .set_settings = dnet_set_settings, + .get_drvinfo = dnet_get_drvinfo, + .get_link = ethtool_op_get_link, +}; + +static const struct net_device_ops dnet_netdev_ops = { + .ndo_open = dnet_open, + .ndo_stop = dnet_close, + .ndo_get_stats = dnet_get_stats, + .ndo_start_xmit = dnet_start_xmit, + .ndo_do_ioctl = dnet_ioctl, + .ndo_set_mac_address = eth_mac_addr, + .ndo_validate_addr = eth_validate_addr, + .ndo_change_mtu = eth_change_mtu, +}; + +static int __devinit dnet_probe(struct platform_device *pdev) +{ + struct resource *res; + struct net_device *dev; + struct dnet *bp; + struct phy_device *phydev; + int err = -ENXIO; + unsigned int mem_base, mem_size, irq; + + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + if (!res) { + dev_err(&pdev->dev, "no mmio resource defined\n"); + goto err_out; + } + mem_base = res->start; + mem_size = resource_size(res); + irq = platform_get_irq(pdev, 0); + + if (!request_mem_region(mem_base, mem_size, DRV_NAME)) { + dev_err(&pdev->dev, "no memory region available\n"); + err = -EBUSY; + goto err_out; + } + + err = -ENOMEM; + dev = alloc_etherdev(sizeof(*bp)); + if (!dev) { + dev_err(&pdev->dev, "etherdev alloc failed, aborting.\n"); + goto err_out; + } + + /* TODO: Actually, we have some interesting features... */ + dev->features |= 0; + + bp = netdev_priv(dev); + bp->dev = dev; + + SET_NETDEV_DEV(dev, &pdev->dev); + + spin_lock_init(&bp->lock); + + bp->regs = ioremap(mem_base, mem_size); + if (!bp->regs) { + dev_err(&pdev->dev, "failed to map registers, aborting.\n"); + err = -ENOMEM; + goto err_out_free_dev; + } + + dev->irq = irq; + err = request_irq(dev->irq, dnet_interrupt, 0, DRV_NAME, dev); + if (err) { + dev_err(&pdev->dev, "Unable to request IRQ %d (error %d)\n", + irq, err); + goto err_out_iounmap; + } + + dev->netdev_ops = &dnet_netdev_ops; + netif_napi_add(dev, &bp->napi, dnet_poll, 64); + dev->ethtool_ops = &dnet_ethtool_ops; + + dev->base_addr = (unsigned long)bp->regs; + + bp->capabilities = dnet_readl(bp, VERCAPS) & DNET_CAPS_MASK; + + dnet_get_hwaddr(bp); + + if (!is_valid_ether_addr(dev->dev_addr)) { + /* choose a random ethernet address */ + random_ether_addr(dev->dev_addr); + __dnet_set_hwaddr(bp); + } + + err = register_netdev(dev); + if (err) { + dev_err(&pdev->dev, "Cannot register net device, aborting.\n"); + goto err_out_free_irq; + } + + /* register the PHY board fixup (for Marvell 88E1111) */ + err = phy_register_fixup_for_uid(0x01410cc0, 0xfffffff0, + dnet_phy_marvell_fixup); + /* we can live without it, so just issue a warning */ + if (err) + dev_warn(&pdev->dev, "Cannot register PHY board fixup.\n"); + + if (dnet_mii_init(bp) != 0) + goto err_out_unregister_netdev; + + dev_info(&pdev->dev, "Dave DNET at 0x%p (0x%08x) irq %d %pM\n", + bp->regs, mem_base, dev->irq, dev->dev_addr); + dev_info(&pdev->dev, "has %smdio, %sirq, %sgigabit, %sdma \n", + (bp->capabilities & DNET_HAS_MDIO) ? "" : "no ", + (bp->capabilities & DNET_HAS_IRQ) ? "" : "no ", + (bp->capabilities & DNET_HAS_GIGABIT) ? "" : "no ", + (bp->capabilities & DNET_HAS_DMA) ? "" : "no "); + phydev = bp->phy_dev; + dev_info(&pdev->dev, "attached PHY driver [%s] " + "(mii_bus:phy_addr=%s, irq=%d)\n", + phydev->drv->name, phydev->dev.bus_id, phydev->irq); + + return 0; + +err_out_unregister_netdev: + unregister_netdev(dev); +err_out_free_irq: + free_irq(dev->irq, dev); +err_out_iounmap: + iounmap(bp->regs); +err_out_free_dev: + free_netdev(dev); +err_out: + return err; +} + +static int __devexit dnet_remove(struct platform_device *pdev) +{ + + struct net_device *dev; + struct dnet *bp; + + dev = platform_get_drvdata(pdev); + + if (dev) { + bp = netdev_priv(dev); + if (bp->phy_dev) + phy_disconnect(bp->phy_dev); + mdiobus_unregister(bp->mii_bus); + kfree(bp->mii_bus->irq); + mdiobus_free(bp->mii_bus); + unregister_netdev(dev); + free_irq(dev->irq, dev); + iounmap(bp->regs); + free_netdev(dev); + } + + return 0; +} + +static struct platform_driver dnet_driver = { + .probe = dnet_probe, + .remove = __devexit_p(dnet_remove), + .driver = { + .name = "dnet", + }, +}; + +static int __init dnet_init(void) +{ + return platform_driver_register(&dnet_driver); +} + +static void __exit dnet_exit(void) +{ + platform_driver_unregister(&dnet_driver); +} + +module_init(dnet_init); +module_exit(dnet_exit); + +MODULE_LICENSE("GPL"); +MODULE_DESCRIPTION("Dave DNET Ethernet driver"); +MODULE_AUTHOR("Ilya Yanok <yanok@emcraft.com>, " + "Matteo Vit <matteo.vit@dave.eu>"); diff --git a/drivers/net/dnet.h b/drivers/net/dnet.h new file mode 100644 index 000000000000..37f5b30fa78b --- /dev/null +++ b/drivers/net/dnet.h @@ -0,0 +1,225 @@ +/* + * Dave DNET Ethernet Controller driver + * + * Copyright (C) 2008 Dave S.r.l. <www.dave.eu> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ +#ifndef _DNET_H +#define _DNET_H + +#define DRV_NAME "dnet" +#define DRV_VERSION "0.9.1" +#define PFX DRV_NAME ": " + +/* Register access macros */ +#define dnet_writel(port, value, reg) \ + writel((value), (port)->regs + DNET_##reg) +#define dnet_readl(port, reg) readl((port)->regs + DNET_##reg) + +/* ALL DNET FIFO REGISTERS */ +#define DNET_RX_LEN_FIFO 0x000 /* RX_LEN_FIFO */ +#define DNET_RX_DATA_FIFO 0x004 /* RX_DATA_FIFO */ +#define DNET_TX_LEN_FIFO 0x008 /* TX_LEN_FIFO */ +#define DNET_TX_DATA_FIFO 0x00C /* TX_DATA_FIFO */ + +/* ALL DNET CONTROL/STATUS REGISTERS OFFSETS */ +#define DNET_VERCAPS 0x100 /* VERCAPS */ +#define DNET_INTR_SRC 0x104 /* INTR_SRC */ +#define DNET_INTR_ENB 0x108 /* INTR_ENB */ +#define DNET_RX_STATUS 0x10C /* RX_STATUS */ +#define DNET_TX_STATUS 0x110 /* TX_STATUS */ +#define DNET_RX_FRAMES_CNT 0x114 /* RX_FRAMES_CNT */ +#define DNET_TX_FRAMES_CNT 0x118 /* TX_FRAMES_CNT */ +#define DNET_RX_FIFO_TH 0x11C /* RX_FIFO_TH */ +#define DNET_TX_FIFO_TH 0x120 /* TX_FIFO_TH */ +#define DNET_SYS_CTL 0x124 /* SYS_CTL */ +#define DNET_PAUSE_TMR 0x128 /* PAUSE_TMR */ +#define DNET_RX_FIFO_WCNT 0x12C /* RX_FIFO_WCNT */ +#define DNET_TX_FIFO_WCNT 0x130 /* TX_FIFO_WCNT */ + +/* ALL DNET MAC REGISTERS */ +#define DNET_MACREG_DATA 0x200 /* Mac-Reg Data */ +#define DNET_MACREG_ADDR 0x204 /* Mac-Reg Addr */ + +/* ALL DNET RX STATISTICS COUNTERS */ +#define DNET_RX_PKT_IGNR_CNT 0x300 +#define DNET_RX_LEN_CHK_ERR_CNT 0x304 +#define DNET_RX_LNG_FRM_CNT 0x308 +#define DNET_RX_SHRT_FRM_CNT 0x30C +#define DNET_RX_IPG_VIOL_CNT 0x310 +#define DNET_RX_CRC_ERR_CNT 0x314 +#define DNET_RX_OK_PKT_CNT 0x318 +#define DNET_RX_CTL_FRM_CNT 0x31C +#define DNET_RX_PAUSE_FRM_CNT 0x320 +#define DNET_RX_MULTICAST_CNT 0x324 +#define DNET_RX_BROADCAST_CNT 0x328 +#define DNET_RX_VLAN_TAG_CNT 0x32C +#define DNET_RX_PRE_SHRINK_CNT 0x330 +#define DNET_RX_DRIB_NIB_CNT 0x334 +#define DNET_RX_UNSUP_OPCD_CNT 0x338 +#define DNET_RX_BYTE_CNT 0x33C + +/* DNET TX STATISTICS COUNTERS */ +#define DNET_TX_UNICAST_CNT 0x400 +#define DNET_TX_PAUSE_FRM_CNT 0x404 +#define DNET_TX_MULTICAST_CNT 0x408 +#define DNET_TX_BRDCAST_CNT 0x40C +#define DNET_TX_VLAN_TAG_CNT 0x410 +#define DNET_TX_BAD_FCS_CNT 0x414 +#define DNET_TX_JUMBO_CNT 0x418 +#define DNET_TX_BYTE_CNT 0x41C + +/* SOME INTERNAL MAC-CORE REGISTER */ +#define DNET_INTERNAL_MODE_REG 0x0 +#define DNET_INTERNAL_RXTX_CONTROL_REG 0x2 +#define DNET_INTERNAL_MAX_PKT_SIZE_REG 0x4 +#define DNET_INTERNAL_IGP_REG 0x8 +#define DNET_INTERNAL_MAC_ADDR_0_REG 0xa +#define DNET_INTERNAL_MAC_ADDR_1_REG 0xc +#define DNET_INTERNAL_MAC_ADDR_2_REG 0xe +#define DNET_INTERNAL_TX_RX_STS_REG 0x12 +#define DNET_INTERNAL_GMII_MNG_CTL_REG 0x14 +#define DNET_INTERNAL_GMII_MNG_DAT_REG 0x16 + +#define DNET_INTERNAL_GMII_MNG_CMD_FIN (1 << 14) + +#define DNET_INTERNAL_WRITE (1 << 31) + +/* MAC-CORE REGISTER FIELDS */ + +/* MAC-CORE MODE REGISTER FIELDS */ +#define DNET_INTERNAL_MODE_GBITEN (1 << 0) +#define DNET_INTERNAL_MODE_FCEN (1 << 1) +#define DNET_INTERNAL_MODE_RXEN (1 << 2) +#define DNET_INTERNAL_MODE_TXEN (1 << 3) + +/* MAC-CORE RXTX CONTROL REGISTER FIELDS */ +#define DNET_INTERNAL_RXTX_CONTROL_RXSHORTFRAME (1 << 8) +#define DNET_INTERNAL_RXTX_CONTROL_RXBROADCAST (1 << 7) +#define DNET_INTERNAL_RXTX_CONTROL_RXMULTICAST (1 << 4) +#define DNET_INTERNAL_RXTX_CONTROL_RXPAUSE (1 << 3) +#define DNET_INTERNAL_RXTX_CONTROL_DISTXFCS (1 << 2) +#define DNET_INTERNAL_RXTX_CONTROL_DISCFXFCS (1 << 1) +#define DNET_INTERNAL_RXTX_CONTROL_ENPROMISC (1 << 0) +#define DNET_INTERNAL_RXTX_CONTROL_DROPCONTROL (1 << 6) +#define DNET_INTERNAL_RXTX_CONTROL_ENABLEHALFDUP (1 << 5) + +/* SYSTEM CONTROL REGISTER FIELDS */ +#define DNET_SYS_CTL_IGNORENEXTPKT (1 << 0) +#define DNET_SYS_CTL_SENDPAUSE (1 << 2) +#define DNET_SYS_CTL_RXFIFOFLUSH (1 << 3) +#define DNET_SYS_CTL_TXFIFOFLUSH (1 << 4) + +/* TX STATUS REGISTER FIELDS */ +#define DNET_TX_STATUS_FIFO_ALMOST_EMPTY (1 << 2) +#define DNET_TX_STATUS_FIFO_ALMOST_FULL (1 << 1) + +/* INTERRUPT SOURCE REGISTER FIELDS */ +#define DNET_INTR_SRC_TX_PKTSENT (1 << 0) +#define DNET_INTR_SRC_TX_FIFOAF (1 << 1) +#define DNET_INTR_SRC_TX_FIFOAE (1 << 2) +#define DNET_INTR_SRC_TX_DISCFRM (1 << 3) +#define DNET_INTR_SRC_TX_FIFOFULL (1 << 4) +#define DNET_INTR_SRC_RX_CMDFIFOAF (1 << 8) +#define DNET_INTR_SRC_RX_CMDFIFOFF (1 << 9) +#define DNET_INTR_SRC_RX_DATAFIFOFF (1 << 10) +#define DNET_INTR_SRC_TX_SUMMARY (1 << 16) +#define DNET_INTR_SRC_RX_SUMMARY (1 << 17) +#define DNET_INTR_SRC_PHY (1 << 19) + +/* INTERRUPT ENABLE REGISTER FIELDS */ +#define DNET_INTR_ENB_TX_PKTSENT (1 << 0) +#define DNET_INTR_ENB_TX_FIFOAF (1 << 1) +#define DNET_INTR_ENB_TX_FIFOAE (1 << 2) +#define DNET_INTR_ENB_TX_DISCFRM (1 << 3) +#define DNET_INTR_ENB_TX_FIFOFULL (1 << 4) +#define DNET_INTR_ENB_RX_PKTRDY (1 << 8) +#define DNET_INTR_ENB_RX_FIFOAF (1 << 9) +#define DNET_INTR_ENB_RX_FIFOERR (1 << 10) +#define DNET_INTR_ENB_RX_ERROR (1 << 11) +#define DNET_INTR_ENB_RX_FIFOFULL (1 << 12) +#define DNET_INTR_ENB_RX_FIFOAE (1 << 13) +#define DNET_INTR_ENB_TX_SUMMARY (1 << 16) +#define DNET_INTR_ENB_RX_SUMMARY (1 << 17) +#define DNET_INTR_ENB_GLOBAL_ENABLE (1 << 18) + +/* default values: + * almost empty = less than one full sized ethernet frame (no jumbo) inside + * the fifo almost full = can write less than one full sized ethernet frame + * (no jumbo) inside the fifo + */ +#define DNET_CFG_TX_FIFO_FULL_THRES 25 +#define DNET_CFG_RX_FIFO_FULL_THRES 20 + +/* + * Capabilities. Used by the driver to know the capabilities that the ethernet + * controller inside the FPGA have. + */ + +#define DNET_HAS_MDIO (1 << 0) +#define DNET_HAS_IRQ (1 << 1) +#define DNET_HAS_GIGABIT (1 << 2) +#define DNET_HAS_DMA (1 << 3) + +#define DNET_HAS_MII (1 << 4) /* or GMII */ +#define DNET_HAS_RMII (1 << 5) /* or RGMII */ + +#define DNET_CAPS_MASK 0xFFFF + +#define DNET_FIFO_SIZE 1024 /* 1K x 32 bit */ +#define DNET_FIFO_TX_DATA_AF_TH (DNET_FIFO_SIZE - 384) /* 384 = 1536 / 4 */ +#define DNET_FIFO_TX_DATA_AE_TH 384 + +#define DNET_FIFO_RX_CMD_AF_TH (1 << 16) /* just one frame inside the FIFO */ + +/* + * Hardware-collected statistics. + */ +struct dnet_stats { + u32 rx_pkt_ignr; + u32 rx_len_chk_err; + u32 rx_lng_frm; + u32 rx_shrt_frm; + u32 rx_ipg_viol; + u32 rx_crc_err; + u32 rx_ok_pkt; + u32 rx_ctl_frm; + u32 rx_pause_frm; + u32 rx_multicast; + u32 rx_broadcast; + u32 rx_vlan_tag; + u32 rx_pre_shrink; + u32 rx_drib_nib; + u32 rx_unsup_opcd; + u32 rx_byte; + u32 tx_unicast; + u32 tx_pause_frm; + u32 tx_multicast; + u32 tx_brdcast; + u32 tx_vlan_tag; + u32 tx_bad_fcs; + u32 tx_jumbo; + u32 tx_byte; +}; + +struct dnet { + void __iomem *regs; + spinlock_t lock; + struct platform_device *pdev; + struct net_device *dev; + struct dnet_stats hw_stats; + unsigned int capabilities; /* read from FPGA */ + struct napi_struct napi; + + /* PHY stuff */ + struct mii_bus *mii_bus; + struct phy_device *phy_dev; + unsigned int link; + unsigned int speed; + unsigned int duplex; +}; + +#endif /* _DNET_H */ diff --git a/drivers/net/ibm_newemac/core.c b/drivers/net/ibm_newemac/core.c index 87a706694fb3..6fd7aa61736e 100644 --- a/drivers/net/ibm_newemac/core.c +++ b/drivers/net/ibm_newemac/core.c @@ -2594,6 +2594,9 @@ static int __devinit emac_init_config(struct emac_instance *dev) if (of_device_is_compatible(np, "ibm,emac-460ex") || of_device_is_compatible(np, "ibm,emac-460gt")) dev->features |= EMAC_FTR_460EX_PHY_CLK_FIX; + if (of_device_is_compatible(np, "ibm,emac-405ex") || + of_device_is_compatible(np, "ibm,emac-405exr")) + dev->features |= EMAC_FTR_440EP_PHY_CLK_FIX; } else if (of_device_is_compatible(np, "ibm,emac4")) { dev->features |= EMAC_FTR_EMAC4; if (of_device_is_compatible(np, "ibm,emac-440gx")) diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c index 7124f59fb99f..7c4481b994ab 100644 --- a/drivers/net/igb/igb_main.c +++ b/drivers/net/igb/igb_main.c @@ -1128,11 +1128,10 @@ static int __devinit igb_probe(struct pci_dev *pdev, struct net_device *netdev; struct igb_adapter *adapter; struct e1000_hw *hw; - struct pci_dev *us_dev; const struct e1000_info *ei = igb_info_tbl[ent->driver_data]; unsigned long mmio_start, mmio_len; - int err, pci_using_dac, pos; - u16 eeprom_data = 0, state = 0; + int err, pci_using_dac; + u16 eeprom_data = 0; u16 eeprom_apme_mask = IGB_EEPROM_APME; u32 part_num; @@ -1158,27 +1157,6 @@ static int __devinit igb_probe(struct pci_dev *pdev, } } - /* 82575 requires that the pci-e link partner disable the L0s state */ - switch (pdev->device) { - case E1000_DEV_ID_82575EB_COPPER: - case E1000_DEV_ID_82575EB_FIBER_SERDES: - case E1000_DEV_ID_82575GB_QUAD_COPPER: - us_dev = pdev->bus->self; - pos = pci_find_capability(us_dev, PCI_CAP_ID_EXP); - if (pos) { - pci_read_config_word(us_dev, pos + PCI_EXP_LNKCTL, - &state); - state &= ~PCIE_LINK_STATE_L0S; - pci_write_config_word(us_dev, pos + PCI_EXP_LNKCTL, - state); - dev_info(&pdev->dev, - "Disabling ASPM L0s upstream switch port %s\n", - pci_name(us_dev)); - } - default: - break; - } - err = pci_request_selected_regions(pdev, pci_select_bars(pdev, IORESOURCE_MEM), igb_driver_name); diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c index 0b2b609fad85..335119a5687a 100644 --- a/drivers/net/ixgbe/ixgbe_main.c +++ b/drivers/net/ixgbe/ixgbe_main.c @@ -4402,6 +4402,7 @@ static const struct net_device_ops ixgbe_netdev_ops = { .ndo_stop = ixgbe_close, .ndo_start_xmit = ixgbe_xmit_frame, .ndo_get_stats = ixgbe_get_stats, + .ndo_set_rx_mode = ixgbe_set_rx_mode, .ndo_set_multicast_list = ixgbe_set_rx_mode, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = ixgbe_set_mac, diff --git a/drivers/net/mv643xx_eth.c b/drivers/net/mv643xx_eth.c index e1f7706c15cd..a56d9d2df73f 100644 --- a/drivers/net/mv643xx_eth.c +++ b/drivers/net/mv643xx_eth.c @@ -2289,11 +2289,6 @@ static void port_start(struct mv643xx_eth_private *mp) } /* - * Add configured unicast address to address filter table. - */ - mv643xx_eth_program_unicast_filter(mp->dev); - - /* * Receive all unmatched unicast, TCP, UDP, BPDU and broadcast * frames to RX queue #0, and include the pseudo-header when * calculating receive checksums. @@ -2306,6 +2301,11 @@ static void port_start(struct mv643xx_eth_private *mp) wrlp(mp, PORT_CONFIG_EXT, 0x00000000); /* + * Add configured unicast addresses to address filter table. + */ + mv643xx_eth_program_unicast_filter(mp->dev); + + /* * Enable the receive queues. */ for (i = 0; i < mp->rxq_count; i++) { diff --git a/drivers/net/netxen/netxen_nic.h b/drivers/net/netxen/netxen_nic.h index 78e6228937fe..c40815169f35 100644 --- a/drivers/net/netxen/netxen_nic.h +++ b/drivers/net/netxen/netxen_nic.h @@ -1590,7 +1590,6 @@ dma_watchdog_wakeup(struct netxen_adapter *adapter) } -int netxen_is_flash_supported(struct netxen_adapter *adapter); int netxen_get_flash_mac_addr(struct netxen_adapter *adapter, __le64 *mac); int netxen_p3_get_mac_addr(struct netxen_adapter *adapter, __le64 *mac); extern void netxen_change_ringparam(struct netxen_adapter *adapter); diff --git a/drivers/net/netxen/netxen_nic_hw.c b/drivers/net/netxen/netxen_nic_hw.c index b24cfddd6193..5026811c04ce 100644 --- a/drivers/net/netxen/netxen_nic_hw.c +++ b/drivers/net/netxen/netxen_nic_hw.c @@ -750,28 +750,6 @@ int netxen_nic_change_mtu(struct net_device *netdev, int mtu) return rc; } -int netxen_is_flash_supported(struct netxen_adapter *adapter) -{ - const int locs[] = { 0, 0x4, 0x100, 0x4000, 0x4128 }; - int addr, val01, val02, i, j; - - /* if the flash size less than 4Mb, make huge war cry and die */ - for (j = 1; j < 4; j++) { - addr = j * NETXEN_NIC_WINDOW_MARGIN; - for (i = 0; i < ARRAY_SIZE(locs); i++) { - if (netxen_rom_fast_read(adapter, locs[i], &val01) == 0 - && netxen_rom_fast_read(adapter, (addr + locs[i]), - &val02) == 0) { - if (val01 == val02) - return -1; - } else - return -1; - } - } - - return 0; -} - static int netxen_get_flash_block(struct netxen_adapter *adapter, int base, int size, __le32 * buf) { diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c index 1fb9bcf504e7..1af47257ba82 100644 --- a/drivers/net/netxen/netxen_nic_main.c +++ b/drivers/net/netxen/netxen_nic_main.c @@ -406,9 +406,6 @@ netxen_read_mac_addr(struct netxen_adapter *adapter) struct net_device *netdev = adapter->netdev; struct pci_dev *pdev = adapter->pdev; - if (netxen_is_flash_supported(adapter) != 0) - return -EIO; - if (NX_IS_REVISION_P3(adapter->ahw.revision_id)) { if (netxen_p3_get_mac_addr(adapter, &mac_addr) != 0) return -EIO; diff --git a/drivers/net/qlge/qlge_main.c b/drivers/net/qlge/qlge_main.c index b91b700a081b..170d3540f9c9 100644 --- a/drivers/net/qlge/qlge_main.c +++ b/drivers/net/qlge/qlge_main.c @@ -1533,7 +1533,6 @@ static void ql_process_mac_rx_intr(struct ql_adapter *qdev, QPRINTK(qdev, RX_STATUS, DEBUG, "Promiscuous Packet.\n"); } - skb->protocol = eth_type_trans(skb, ndev); skb->ip_summed = CHECKSUM_NONE; @@ -3248,6 +3247,7 @@ static int ql_adapter_down(struct ql_adapter *qdev) netif_napi_del(&qdev->rx_ring[i].napi); ql_free_rx_buffers(qdev); + spin_lock(&qdev->hw_lock); status = ql_adapter_reset(qdev); if (status) diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c index 7e4b586e306d..06c535222666 100644 --- a/drivers/net/r8169.c +++ b/drivers/net/r8169.c @@ -81,9 +81,9 @@ static const int multicast_filter_limit = 32; #define RTL8169_TX_TIMEOUT (6*HZ) #define RTL8169_PHY_TIMEOUT (10*HZ) -#define RTL_EEPROM_SIG 0x8129 +#define RTL_EEPROM_SIG cpu_to_le32(0x8129) +#define RTL_EEPROM_SIG_MASK cpu_to_le32(0xffff) #define RTL_EEPROM_SIG_ADDR 0x0000 -#define RTL_EEPROM_MAC_ADDR 0x0007 /* write/read MMIO register */ #define RTL_W8(reg, val8) writeb ((val8), ioaddr + (reg)) @@ -293,11 +293,6 @@ enum rtl_register_content { /* Cfg9346Bits */ Cfg9346_Lock = 0x00, Cfg9346_Unlock = 0xc0, - Cfg9346_Program = 0x80, /* Programming mode */ - Cfg9346_EECS = 0x08, /* Chip select */ - Cfg9346_EESK = 0x04, /* Serial data clock */ - Cfg9346_EEDI = 0x02, /* Data input */ - Cfg9346_EEDO = 0x01, /* Data output */ /* rx_mode_bits */ AcceptErr = 0x20, @@ -310,7 +305,6 @@ enum rtl_register_content { /* RxConfigBits */ RxCfgFIFOShift = 13, RxCfgDMAShift = 8, - RxCfg9356SEL = 6, /* EEPROM type: 0 = 9346, 1 = 9356 */ /* TxConfigBits */ TxInterFrameGapShift = 24, @@ -1969,108 +1963,6 @@ static const struct net_device_ops rtl8169_netdev_ops = { }; -/* Delay between EEPROM clock transitions. Force out buffered PCI writes. */ -#define RTL_EEPROM_DELAY() RTL_R8(Cfg9346) -#define RTL_EEPROM_READ_CMD 6 - -/* read 16bit word stored in EEPROM. EEPROM is addressed by words. */ -static u16 rtl_eeprom_read(void __iomem *ioaddr, int addr) -{ - u16 result = 0; - int cmd, cmd_len, i; - - /* check for EEPROM address size (in bits) */ - if (RTL_R32(RxConfig) & (1 << RxCfg9356SEL)) { - /* EEPROM is 93C56 */ - cmd_len = 3 + 8; /* 3 bits for command id and 8 for address */ - cmd = (RTL_EEPROM_READ_CMD << 8) | (addr & 0xff); - } else { - /* EEPROM is 93C46 */ - cmd_len = 3 + 6; /* 3 bits for command id and 6 for address */ - cmd = (RTL_EEPROM_READ_CMD << 6) | (addr & 0x3f); - } - - /* enter programming mode */ - RTL_W8(Cfg9346, Cfg9346_Program | Cfg9346_EECS); - RTL_EEPROM_DELAY(); - - /* write command and requested address */ - while (cmd_len--) { - u8 x = Cfg9346_Program | Cfg9346_EECS; - - x |= (cmd & (1 << cmd_len)) ? Cfg9346_EEDI : 0; - - /* write a bit */ - RTL_W8(Cfg9346, x); - RTL_EEPROM_DELAY(); - - /* raise clock */ - RTL_W8(Cfg9346, x | Cfg9346_EESK); - RTL_EEPROM_DELAY(); - } - - /* lower clock */ - RTL_W8(Cfg9346, Cfg9346_Program | Cfg9346_EECS); - RTL_EEPROM_DELAY(); - - /* read back 16bit value */ - for (i = 16; i > 0; i--) { - /* raise clock */ - RTL_W8(Cfg9346, Cfg9346_Program | Cfg9346_EECS | Cfg9346_EESK); - RTL_EEPROM_DELAY(); - - result <<= 1; - result |= (RTL_R8(Cfg9346) & Cfg9346_EEDO) ? 1 : 0; - - /* lower clock */ - RTL_W8(Cfg9346, Cfg9346_Program | Cfg9346_EECS); - RTL_EEPROM_DELAY(); - } - - RTL_W8(Cfg9346, Cfg9346_Program); - /* leave programming mode */ - RTL_W8(Cfg9346, Cfg9346_Lock); - - return result; -} - -static void rtl_init_mac_address(struct rtl8169_private *tp, - void __iomem *ioaddr) -{ - struct pci_dev *pdev = tp->pci_dev; - u16 x; - u8 mac[8]; - - /* read EEPROM signature */ - x = rtl_eeprom_read(ioaddr, RTL_EEPROM_SIG_ADDR); - - if (x != RTL_EEPROM_SIG) { - dev_info(&pdev->dev, "Missing EEPROM signature: %04x\n", x); - return; - } - - /* read MAC address */ - x = rtl_eeprom_read(ioaddr, RTL_EEPROM_MAC_ADDR); - mac[0] = x & 0xff; - mac[1] = x >> 8; - x = rtl_eeprom_read(ioaddr, RTL_EEPROM_MAC_ADDR + 1); - mac[2] = x & 0xff; - mac[3] = x >> 8; - x = rtl_eeprom_read(ioaddr, RTL_EEPROM_MAC_ADDR + 2); - mac[4] = x & 0xff; - mac[5] = x >> 8; - - if (netif_msg_probe(tp)) { - DECLARE_MAC_BUF(buf); - - dev_info(&pdev->dev, "MAC address found in EEPROM: %s\n", - print_mac(buf, mac)); - } - - if (is_valid_ether_addr(mac)) - rtl_rar_set(tp, mac); -} - static int __devinit rtl8169_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) { @@ -2249,8 +2141,6 @@ rtl8169_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) tp->mmio_addr = ioaddr; - rtl_init_mac_address(tp, ioaddr); - /* Get MAC address */ for (i = 0; i < MAC_ADDR_LEN; i++) dev->dev_addr[i] = RTL_R8(MAC0 + i); @@ -3363,13 +3253,6 @@ static int rtl8169_start_xmit(struct sk_buff *skb, struct net_device *dev) opts1 |= FirstFrag; } else { len = skb->len; - - if (unlikely(len < ETH_ZLEN)) { - if (skb_padto(skb, ETH_ZLEN)) - goto err_update_stats; - len = ETH_ZLEN; - } - opts1 |= FirstFrag | LastFrag; tp->tx_skb[entry].skb = skb; } @@ -3407,7 +3290,6 @@ out: err_stop: netif_stop_queue(dev); ret = NETDEV_TX_BUSY; -err_update_stats: dev->stats.tx_dropped++; goto out; } diff --git a/drivers/net/via-velocity.c b/drivers/net/via-velocity.c index c5691fdb7079..fb53ef872df3 100644 --- a/drivers/net/via-velocity.c +++ b/drivers/net/via-velocity.c @@ -1838,17 +1838,19 @@ static void velocity_free_tx_buf(struct velocity_info *vptr, struct velocity_td_ { struct sk_buff *skb = tdinfo->skb; int i; + int pktlen; /* * Don't unmap the pre-allocated tx_bufs */ if (tdinfo->skb_dma) { + pktlen = (skb->len > ETH_ZLEN ? : ETH_ZLEN); for (i = 0; i < tdinfo->nskb_dma; i++) { #ifdef VELOCITY_ZERO_COPY_SUPPORT pci_unmap_single(vptr->pdev, tdinfo->skb_dma[i], le16_to_cpu(td->tdesc1.len), PCI_DMA_TODEVICE); #else - pci_unmap_single(vptr->pdev, tdinfo->skb_dma[i], skb->len, PCI_DMA_TODEVICE); + pci_unmap_single(vptr->pdev, tdinfo->skb_dma[i], pktlen, PCI_DMA_TODEVICE); #endif tdinfo->skb_dma[i] = 0; } @@ -2080,17 +2082,14 @@ static int velocity_xmit(struct sk_buff *skb, struct net_device *dev) struct tx_desc *td_ptr; struct velocity_td_info *tdinfo; unsigned long flags; - int pktlen = skb->len; + int pktlen; __le16 len; int index; - - if (skb->len < ETH_ZLEN) { - if (skb_padto(skb, ETH_ZLEN)) - goto out; - pktlen = ETH_ZLEN; - } + if (skb_padto(skb, ETH_ZLEN)) + goto out; + pktlen = max_t(unsigned int, skb->len, ETH_ZLEN); len = cpu_to_le16(pktlen); diff --git a/drivers/net/wireless/ath9k/ath9k.h b/drivers/net/wireless/ath9k/ath9k.h index f0b105a11ae2..b64be8e9a690 100644 --- a/drivers/net/wireless/ath9k/ath9k.h +++ b/drivers/net/wireless/ath9k/ath9k.h @@ -579,6 +579,7 @@ struct ath_softc { void __iomem *mem; int irq; spinlock_t sc_resetlock; + spinlock_t sc_serial_rw; struct mutex mutex; u8 curbssid[ETH_ALEN]; @@ -724,4 +725,36 @@ void ath9k_wiphy_pause_all_forced(struct ath_softc *sc, bool ath9k_wiphy_scanning(struct ath_softc *sc); void ath9k_wiphy_work(struct work_struct *work); +/* + * Read and write, they both share the same lock. We do this to serialize + * reads and writes on Atheros 802.11n PCI devices only. This is required + * as the FIFO on these devices can only accept sanely 2 requests. After + * that the device goes bananas. Serializing the reads/writes prevents this + * from happening. + */ + +static inline void ath9k_iowrite32(struct ath_hw *ah, u32 reg_offset, u32 val) +{ + if (ah->config.serialize_regmode == SER_REG_MODE_ON) { + unsigned long flags; + spin_lock_irqsave(&ah->ah_sc->sc_serial_rw, flags); + iowrite32(val, ah->ah_sc->mem + reg_offset); + spin_unlock_irqrestore(&ah->ah_sc->sc_serial_rw, flags); + } else + iowrite32(val, ah->ah_sc->mem + reg_offset); +} + +static inline unsigned int ath9k_ioread32(struct ath_hw *ah, u32 reg_offset) +{ + u32 val; + if (ah->config.serialize_regmode == SER_REG_MODE_ON) { + unsigned long flags; + spin_lock_irqsave(&ah->ah_sc->sc_serial_rw, flags); + val = ioread32(ah->ah_sc->mem + reg_offset); + spin_unlock_irqrestore(&ah->ah_sc->sc_serial_rw, flags); + } else + val = ioread32(ah->ah_sc->mem + reg_offset); + return val; +} + #endif /* ATH9K_H */ diff --git a/drivers/net/wireless/ath9k/hw.c b/drivers/net/wireless/ath9k/hw.c index eb750a503999..60e55d8c510b 100644 --- a/drivers/net/wireless/ath9k/hw.c +++ b/drivers/net/wireless/ath9k/hw.c @@ -391,6 +391,25 @@ static void ath9k_hw_set_defaults(struct ath_hw *ah) } ah->config.intr_mitigation = 1; + + /* + * We need this for PCI devices only (Cardbus, PCI, miniPCI) + * _and_ if on non-uniprocessor systems (Multiprocessor/HT). + * This means we use it for all AR5416 devices, and the few + * minor PCI AR9280 devices out there. + * + * Serialization is required because these devices do not handle + * well the case of two concurrent reads/writes due to the latency + * involved. During one read/write another read/write can be issued + * on another CPU while the previous read/write may still be working + * on our hardware, if we hit this case the hardware poops in a loop. + * We prevent this by serializing reads and writes. + * + * This issue is not present on PCI-Express devices or pre-AR5416 + * devices (legacy, 802.11abg). + */ + if (num_possible_cpus() > 1) + ah->config.serialize_regmode = SER_REG_MODE_AUTO; } static struct ath_hw *ath9k_hw_newstate(u16 devid, struct ath_softc *sc, @@ -610,7 +629,8 @@ static struct ath_hw *ath9k_hw_do_attach(u16 devid, struct ath_softc *sc, } if (ah->config.serialize_regmode == SER_REG_MODE_AUTO) { - if (ah->hw_version.macVersion == AR_SREV_VERSION_5416_PCI) { + if (ah->hw_version.macVersion == AR_SREV_VERSION_5416_PCI || + (AR_SREV_9280(ah) && !ah->is_pciexpress)) { ah->config.serialize_regmode = SER_REG_MODE_ON; } else { diff --git a/drivers/net/wireless/ath9k/hw.h b/drivers/net/wireless/ath9k/hw.h index 89936a038da3..dc681f011fdf 100644 --- a/drivers/net/wireless/ath9k/hw.h +++ b/drivers/net/wireless/ath9k/hw.h @@ -42,8 +42,8 @@ #define AR5416_MAGIC 0x19641014 /* Register read/write primitives */ -#define REG_WRITE(_ah, _reg, _val) iowrite32(_val, _ah->ah_sc->mem + _reg) -#define REG_READ(_ah, _reg) ioread32(_ah->ah_sc->mem + _reg) +#define REG_WRITE(_ah, _reg, _val) ath9k_iowrite32((_ah), (_reg), (_val)) +#define REG_READ(_ah, _reg) ath9k_ioread32((_ah), (_reg)) #define SM(_v, _f) (((_v) << _f##_S) & _f) #define MS(_v, _f) (((_v) & _f) >> _f##_S) diff --git a/drivers/net/wireless/ath9k/main.c b/drivers/net/wireless/ath9k/main.c index f473fee72a2e..a9715f5b0af6 100644 --- a/drivers/net/wireless/ath9k/main.c +++ b/drivers/net/wireless/ath9k/main.c @@ -1370,6 +1370,7 @@ static int ath_init(u16 devid, struct ath_softc *sc) spin_lock_init(&sc->wiphy_lock); spin_lock_init(&sc->sc_resetlock); + spin_lock_init(&sc->sc_serial_rw); mutex_init(&sc->mutex); tasklet_init(&sc->intr_tq, ath9k_tasklet, (unsigned long)sc); tasklet_init(&sc->bcon_tasklet, ath_beacon_tasklet, diff --git a/drivers/net/wireless/zd1211rw/zd_mac.c b/drivers/net/wireless/zd1211rw/zd_mac.c index 0b01ff036aac..c3a51266de20 100644 --- a/drivers/net/wireless/zd1211rw/zd_mac.c +++ b/drivers/net/wireless/zd1211rw/zd_mac.c @@ -575,13 +575,17 @@ static int zd_op_tx(struct ieee80211_hw *hw, struct sk_buff *skb) r = fill_ctrlset(mac, skb); if (r) - return r; + goto fail; info->rate_driver_data[0] = hw; r = zd_usb_tx(&mac->chip.usb, skb); if (r) - return r; + goto fail; + return 0; + +fail: + dev_kfree_skb(skb); return 0; } diff --git a/include/net/netfilter/nf_conntrack_core.h b/include/net/netfilter/nf_conntrack_core.h index c25068e38516..5a449b44ba33 100644 --- a/include/net/netfilter/nf_conntrack_core.h +++ b/include/net/netfilter/nf_conntrack_core.h @@ -62,7 +62,8 @@ static inline int nf_conntrack_confirm(struct sk_buff *skb) if (ct && ct != &nf_conntrack_untracked) { if (!nf_ct_is_confirmed(ct) && !nf_ct_is_dying(ct)) ret = __nf_conntrack_confirm(skb); - nf_ct_deliver_cached_events(ct); + if (likely(ret == NF_ACCEPT)) + nf_ct_deliver_cached_events(ct); } return ret; } diff --git a/net/core/dev.c b/net/core/dev.c index 7bd3c29c5a78..c01303129119 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2627,9 +2627,9 @@ static int process_backlog(struct napi_struct *napi, int quota) local_irq_disable(); skb = __skb_dequeue(&queue->input_pkt_queue); if (!skb) { - __napi_complete(napi); local_irq_enable(); - break; + napi_complete(napi); + goto out; } local_irq_enable(); @@ -2638,6 +2638,7 @@ static int process_backlog(struct napi_struct *napi, int quota) napi_gro_flush(napi); +out: return work; } diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c index 3e2ddfaee81a..fbf533cc9dce 100644 --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -1202,6 +1202,9 @@ module_init(inet6_init); static void __exit inet6_exit(void) { + if (disable_ipv6) + return; + /* First of all disallow new sockets creation. */ sock_unregister(PF_INET6); /* Disallow any further netlink messages */ diff --git a/net/ipv6/netfilter/nf_conntrack_reasm.c b/net/ipv6/netfilter/nf_conntrack_reasm.c index ed4d79a9e4a6..058a5e4a60c3 100644 --- a/net/ipv6/netfilter/nf_conntrack_reasm.c +++ b/net/ipv6/netfilter/nf_conntrack_reasm.c @@ -528,14 +528,14 @@ find_prev_fhdr(struct sk_buff *skb, u8 *prevhdrp, int *prevhoff, int *fhoff) if (!ipv6_ext_hdr(nexthdr)) { return -1; } - if (len < (int)sizeof(struct ipv6_opt_hdr)) { - pr_debug("too short\n"); - return -1; - } if (nexthdr == NEXTHDR_NONE) { pr_debug("next header is none\n"); return -1; } + if (len < (int)sizeof(struct ipv6_opt_hdr)) { + pr_debug("too short\n"); + return -1; + } if (skb_copy_bits(skb, start, &hdr, sizeof(hdr))) BUG(); if (nexthdr == NEXTHDR_AUTH) diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c index c3f0e950125b..457238a2f3fc 100644 --- a/net/mac80211/tx.c +++ b/net/mac80211/tx.c @@ -784,6 +784,8 @@ ieee80211_tx_h_fragment(struct ieee80211_tx_data *tx) skb_copy_queue_mapping(frag, first); frag->do_not_encrypt = first->do_not_encrypt; + frag->dev = first->dev; + frag->iif = first->iif; pos += copylen; left -= copylen; diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c index 90ce9ddb9451..f4935e344b61 100644 --- a/net/netfilter/nf_conntrack_core.c +++ b/net/netfilter/nf_conntrack_core.c @@ -726,7 +726,7 @@ nf_conntrack_in(struct net *net, u_int8_t pf, unsigned int hooknum, NF_CT_ASSERT(skb->nfct); ret = l4proto->packet(ct, skb, dataoff, ctinfo, pf, hooknum); - if (ret < 0) { + if (ret <= 0) { /* Invalid: inverse of the return code tells * the netfilter core what to do */ pr_debug("nf_conntrack_in: Can't track with proto module\n"); diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c index cb78aa00399e..ed6d873ad384 100644 --- a/net/netfilter/nf_conntrack_netlink.c +++ b/net/netfilter/nf_conntrack_netlink.c @@ -1780,6 +1780,7 @@ ctnetlink_create_expect(struct nlattr *cda[], u_int8_t u3, u32 pid, int report) goto out; } + exp->class = 0; exp->expectfn = NULL; exp->flags = 0; exp->master = ct; diff --git a/net/netfilter/nf_conntrack_proto_tcp.c b/net/netfilter/nf_conntrack_proto_tcp.c index a1edb9c1adee..f3fd154d1ddd 100644 --- a/net/netfilter/nf_conntrack_proto_tcp.c +++ b/net/netfilter/nf_conntrack_proto_tcp.c @@ -859,7 +859,7 @@ static int tcp_packet(struct nf_conn *ct, */ if (nf_ct_kill(ct)) return -NF_REPEAT; - return -NF_DROP; + return NF_DROP; } /* Fall through */ case TCP_CONNTRACK_IGNORE: @@ -892,7 +892,7 @@ static int tcp_packet(struct nf_conn *ct, nf_log_packet(pf, 0, skb, NULL, NULL, NULL, "nf_ct_tcp: killing out of sync session "); nf_ct_kill(ct); - return -NF_DROP; + return NF_DROP; } ct->proto.tcp.last_index = index; ct->proto.tcp.last_dir = dir; diff --git a/net/wireless/Kconfig b/net/wireless/Kconfig index e28e2b8fa436..092ae6faccca 100644 --- a/net/wireless/Kconfig +++ b/net/wireless/Kconfig @@ -102,3 +102,13 @@ config LIB80211_CRYPT_CCMP config LIB80211_CRYPT_TKIP tristate + +config LIB80211_DEBUG + bool "lib80211 debugging messages" + depends on LIB80211 + default n + ---help--- + You can enable this if you want verbose debugging messages + from lib80211. + + If unsure, say N. diff --git a/net/wireless/lib80211_crypt_ccmp.c b/net/wireless/lib80211_crypt_ccmp.c index db428194c16a..2301dc1edc4c 100644 --- a/net/wireless/lib80211_crypt_ccmp.c +++ b/net/wireless/lib80211_crypt_ccmp.c @@ -337,6 +337,7 @@ static int lib80211_ccmp_decrypt(struct sk_buff *skb, int hdr_len, void *priv) pos += 8; if (ccmp_replay_check(pn, key->rx_pn)) { +#ifdef CONFIG_LIB80211_DEBUG if (net_ratelimit()) { printk(KERN_DEBUG "CCMP: replay detected: STA=%pM " "previous PN %02x%02x%02x%02x%02x%02x " @@ -346,6 +347,7 @@ static int lib80211_ccmp_decrypt(struct sk_buff *skb, int hdr_len, void *priv) key->rx_pn[3], key->rx_pn[4], key->rx_pn[5], pn[0], pn[1], pn[2], pn[3], pn[4], pn[5]); } +#endif key->dot11RSNAStatsCCMPReplays++; return -4; } diff --git a/net/wireless/lib80211_crypt_tkip.c b/net/wireless/lib80211_crypt_tkip.c index 7e8e22bfed90..c36287399d7e 100644 --- a/net/wireless/lib80211_crypt_tkip.c +++ b/net/wireless/lib80211_crypt_tkip.c @@ -465,12 +465,14 @@ static int lib80211_tkip_decrypt(struct sk_buff *skb, int hdr_len, void *priv) pos += 8; if (tkip_replay_check(iv32, iv16, tkey->rx_iv32, tkey->rx_iv16)) { +#ifdef CONFIG_LIB80211_DEBUG if (net_ratelimit()) { printk(KERN_DEBUG "TKIP: replay detected: STA=%pM" " previous TSC %08x%04x received TSC " "%08x%04x\n", hdr->addr2, tkey->rx_iv32, tkey->rx_iv16, iv32, iv16); } +#endif tkey->dot11RSNAStatsTKIPReplays++; return -4; } @@ -505,10 +507,12 @@ static int lib80211_tkip_decrypt(struct sk_buff *skb, int hdr_len, void *priv) * it needs to be recalculated for the next packet. */ tkey->rx_phase1_done = 0; } +#ifdef CONFIG_LIB80211_DEBUG if (net_ratelimit()) { printk(KERN_DEBUG "TKIP: ICV error detected: STA=" "%pM\n", hdr->addr2); } +#endif tkey->dot11RSNAStatsTKIPICVErrors++; return -5; } diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c index e25ff62ab2a6..62a5425cc6aa 100644 --- a/net/xfrm/xfrm_state.c +++ b/net/xfrm/xfrm_state.c @@ -748,12 +748,51 @@ static void xfrm_hash_grow_check(struct net *net, int have_hash_collision) schedule_work(&net->xfrm.state_hash_work); } +static void xfrm_state_look_at(struct xfrm_policy *pol, struct xfrm_state *x, + struct flowi *fl, unsigned short family, + xfrm_address_t *daddr, xfrm_address_t *saddr, + struct xfrm_state **best, int *acq_in_progress, + int *error) +{ + /* Resolution logic: + * 1. There is a valid state with matching selector. Done. + * 2. Valid state with inappropriate selector. Skip. + * + * Entering area of "sysdeps". + * + * 3. If state is not valid, selector is temporary, it selects + * only session which triggered previous resolution. Key + * manager will do something to install a state with proper + * selector. + */ + if (x->km.state == XFRM_STATE_VALID) { + if ((x->sel.family && + !xfrm_selector_match(&x->sel, fl, x->sel.family)) || + !security_xfrm_state_pol_flow_match(x, pol, fl)) + return; + + if (!*best || + (*best)->km.dying > x->km.dying || + ((*best)->km.dying == x->km.dying && + (*best)->curlft.add_time < x->curlft.add_time)) + *best = x; + } else if (x->km.state == XFRM_STATE_ACQ) { + *acq_in_progress = 1; + } else if (x->km.state == XFRM_STATE_ERROR || + x->km.state == XFRM_STATE_EXPIRED) { + if (xfrm_selector_match(&x->sel, fl, x->sel.family) && + security_xfrm_state_pol_flow_match(x, pol, fl)) + *error = -ESRCH; + } +} + struct xfrm_state * xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr, struct flowi *fl, struct xfrm_tmpl *tmpl, struct xfrm_policy *pol, int *err, unsigned short family) { + static xfrm_address_t saddr_wildcard = { }; struct net *net = xp_net(pol); unsigned int h; struct hlist_node *entry; @@ -773,40 +812,27 @@ xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr, xfrm_state_addr_check(x, daddr, saddr, family) && tmpl->mode == x->props.mode && tmpl->id.proto == x->id.proto && - (tmpl->id.spi == x->id.spi || !tmpl->id.spi)) { - /* Resolution logic: - 1. There is a valid state with matching selector. - Done. - 2. Valid state with inappropriate selector. Skip. - - Entering area of "sysdeps". - - 3. If state is not valid, selector is temporary, - it selects only session which triggered - previous resolution. Key manager will do - something to install a state with proper - selector. - */ - if (x->km.state == XFRM_STATE_VALID) { - if ((x->sel.family && !xfrm_selector_match(&x->sel, fl, x->sel.family)) || - !security_xfrm_state_pol_flow_match(x, pol, fl)) - continue; - if (!best || - best->km.dying > x->km.dying || - (best->km.dying == x->km.dying && - best->curlft.add_time < x->curlft.add_time)) - best = x; - } else if (x->km.state == XFRM_STATE_ACQ) { - acquire_in_progress = 1; - } else if (x->km.state == XFRM_STATE_ERROR || - x->km.state == XFRM_STATE_EXPIRED) { - if (xfrm_selector_match(&x->sel, fl, x->sel.family) && - security_xfrm_state_pol_flow_match(x, pol, fl)) - error = -ESRCH; - } - } + (tmpl->id.spi == x->id.spi || !tmpl->id.spi)) + xfrm_state_look_at(pol, x, fl, family, daddr, saddr, + &best, &acquire_in_progress, &error); + } + if (best) + goto found; + + h = xfrm_dst_hash(net, daddr, &saddr_wildcard, tmpl->reqid, family); + hlist_for_each_entry(x, entry, net->xfrm.state_bydst+h, bydst) { + if (x->props.family == family && + x->props.reqid == tmpl->reqid && + !(x->props.flags & XFRM_STATE_WILDRECV) && + xfrm_state_addr_check(x, daddr, saddr, family) && + tmpl->mode == x->props.mode && + tmpl->id.proto == x->id.proto && + (tmpl->id.spi == x->id.spi || !tmpl->id.spi)) + xfrm_state_look_at(pol, x, fl, family, daddr, saddr, + &best, &acquire_in_progress, &error); } +found: x = best; if (!x && !error && !acquire_in_progress) { if (tmpl->id.spi && |