iw_cxgb4 driver submission

new file mode 100644
index 0000000..f1d3c33
@@ -0,0 +1,742 @@
+/*
+ * Copyright (c) 2009-2010 Chelsio, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *	  copyright notice, this list of conditions and the following
+ *	  disclaimer.
+ *      - Redistributions in binary form must reproduce the above
+ *	  copyright notice, this list of conditions and the following
+ *	  disclaimer in the documentation and/or other materials
+ *	  provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#ifndef __IW_CXGB4_H__
+#define __IW_CXGB4_H__
+
+#include <linux/mutex.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/idr.h>
+#include <linux/workqueue.h>
+#include <linux/netdevice.h>
+#include <linux/sched.h>
+#include <linux/pci.h>
+#include <linux/dma-mapping.h>
+#include <linux/inet.h>
+#include <linux/wait.h>
+#include <linux/kref.h>
+#include <linux/timer.h>
+#include <linux/io.h>
+#include <linux/kfifo.h>
+
+#include <asm/byteorder.h>
+
+#include <net/net_namespace.h>
+
+#include <rdma/ib_verbs.h>
+#include <rdma/iw_cm.h>
+
+#include "cxgb4.h"
+#include "cxgb4_uld.h"
+#include "l2t.h"
+#include "user.h"
+
+#define DRV_NAME "iw_cxgb4"
+#define MOD DRV_NAME ":"
+
+extern int c4iw_debug;
+#define PDBG(fmt, args...) \
+do { \
+	if (c4iw_debug) \
+		printk(MOD fmt, ## args); \
+} while (0)
+
+#include "t4.h"
+
+#define PBL_OFF(rdev_p, a) ((a) - (rdev_p)->lldi.vr->pbl.start)
+#define RQT_OFF(rdev_p, a) ((a) - (rdev_p)->lldi.vr->rq.start)
+
+static inline void *cplhdr(struct sk_buff *skb)
+{
+	return skb->data;
+}
+
+#define C4IW_WR_TO (10*HZ)
+
+struct c4iw_wr_wait {
+	wait_queue_head_t wait;
+	int done;
+	int ret;
+};
+
+static inline void c4iw_init_wr_wait(struct c4iw_wr_wait *wr_waitp)
+{
+	wr_waitp->ret = 0;
+	wr_waitp->done = 0;
+	init_waitqueue_head(&wr_waitp->wait);
+}
+
+struct c4iw_resource {
+	struct kfifo tpt_fifo;
+	spinlock_t tpt_fifo_lock;
+	struct kfifo qid_fifo;
+	spinlock_t qid_fifo_lock;
+	struct kfifo pdid_fifo;
+	spinlock_t pdid_fifo_lock;
+};
+
+struct c4iw_qid_list {
+	struct list_head entry;
+	u32 qid;
+};
+
+struct c4iw_dev_ucontext {
+	struct list_head qpids;
+	struct list_head cqids;
+	struct mutex lock;
+};
+
+enum c4iw_rdev_flags {
+	T4_FATAL_ERROR = (1<<0),
+};
+
+struct c4iw_rdev {
+	struct c4iw_resource resource;
+	unsigned long qpshift;
+	u32 qpmask;
+	unsigned long cqshift;
+	u32 cqmask;
+	struct c4iw_dev_ucontext uctx;
+	struct gen_pool *pbl_pool;
+	struct gen_pool *rqt_pool;
+	u32 flags;
+	struct cxgb4_lld_info lldi;
+};
+
+static inline int c4iw_fatal_error(struct c4iw_rdev *rdev)
+{
+	return rdev->flags & T4_FATAL_ERROR;
+}
+
+static inline int c4iw_num_stags(struct c4iw_rdev *rdev)
+{
+	return min((int)T4_MAX_NUM_STAG, (int)(rdev->lldi.vr->stag.size >> 5));
+}
+
+struct c4iw_dev {
+	struct ib_device ibdev;
+	struct c4iw_rdev rdev;
+	u32 device_cap_flags;
+	struct idr cqidr;
+	struct idr qpidr;
+	struct idr mmidr;
+	spinlock_t lock;
+	struct list_head entry;
+	struct delayed_work db_drop_task;
+	struct dentry *debugfs_root;
+};
+
+static inline struct c4iw_dev *to_c4iw_dev(struct ib_device *ibdev)
+{
+	return container_of(ibdev, struct c4iw_dev, ibdev);
+}
+
+static inline struct c4iw_dev *rdev_to_c4iw_dev(struct c4iw_rdev *rdev)
+{
+	return container_of(rdev, struct c4iw_dev, rdev);
+}
+
+static inline struct c4iw_cq *get_chp(struct c4iw_dev *rhp, u32 cqid)
+{
+	return idr_find(&rhp->cqidr, cqid);
+}
+
+static inline struct c4iw_qp *get_qhp(struct c4iw_dev *rhp, u32 qpid)
+{
+	return idr_find(&rhp->qpidr, qpid);
+}
+
+static inline struct c4iw_mr *get_mhp(struct c4iw_dev *rhp, u32 mmid)
+{
+	return idr_find(&rhp->mmidr, mmid);
+}
+
+static inline int insert_handle(struct c4iw_dev *rhp, struct idr *idr,
+				void *handle, u32 id)
+{
+	int ret;
+	int newid;
+
+	do {
+		if (!idr_pre_get(idr, GFP_KERNEL))
+			return -ENOMEM;
+		spin_lock_irq(&rhp->lock);
+		ret = idr_get_new_above(idr, handle, id, &newid);
+		BUG_ON(newid != id);
+		spin_unlock_irq(&rhp->lock);
+	} while (ret == -EAGAIN);
+
+	return ret;
+}
+
+static inline void remove_handle(struct c4iw_dev *rhp, struct idr *idr, u32 id)
+{
+	spin_lock_irq(&rhp->lock);
+	idr_remove(idr, id);
+	spin_unlock_irq(&rhp->lock);
+}
+
+struct c4iw_pd {
+	struct ib_pd ibpd;
+	u32 pdid;
+	struct c4iw_dev *rhp;
+};
+
+static inline struct c4iw_pd *to_c4iw_pd(struct ib_pd *ibpd)
+{
+	return container_of(ibpd, struct c4iw_pd, ibpd);
+}
+
+struct tpt_attributes {
+	u64 len;
+	u64 va_fbo;
+	enum fw_ri_mem_perms perms;
+	u32 stag;
+	u32 pdid;
+	u32 qpid;
+	u32 pbl_addr;
+	u32 pbl_size;
+	u32 state:1;
+	u32 type:2;
+	u32 rsvd:1;
+	u32 remote_invaliate_disable:1;
+	u32 zbva:1;
+	u32 mw_bind_enable:1;
+	u32 page_size:5;
+};
+
+struct c4iw_mr {
+	struct ib_mr ibmr;
+	struct ib_umem *umem;
+	struct c4iw_dev *rhp;
+	u64 kva;
+	struct tpt_attributes attr;
+};
+
+static inline struct c4iw_mr *to_c4iw_mr(struct ib_mr *ibmr)
+{
+	return container_of(ibmr, struct c4iw_mr, ibmr);
+}
+
+struct c4iw_mw {
+	struct ib_mw ibmw;
+	struct c4iw_dev *rhp;
+	u64 kva;
+	struct tpt_attributes attr;
+};
+
+static inline struct c4iw_mw *to_c4iw_mw(struct ib_mw *ibmw)
+{
+	return container_of(ibmw, struct c4iw_mw, ibmw);
+}
+
+struct c4iw_fr_page_list {
+	struct ib_fast_reg_page_list ibpl;
+	DECLARE_PCI_UNMAP_ADDR(mapping);
+	dma_addr_t dma_addr;
+	struct c4iw_dev *dev;
+	int size;
+};
+
+static inline struct c4iw_fr_page_list *to_c4iw_fr_page_list(
+					struct ib_fast_reg_page_list *ibpl)
+{
+	return container_of(ibpl, struct c4iw_fr_page_list, ibpl);
+}
+
+struct c4iw_cq {
+	struct ib_cq ibcq;
+	struct c4iw_dev *rhp;
+	struct t4_cq cq;
+	spinlock_t lock;
+	atomic_t refcnt;
+	wait_queue_head_t wait;
+};
+
+static inline struct c4iw_cq *to_c4iw_cq(struct ib_cq *ibcq)
+{
+	return container_of(ibcq, struct c4iw_cq, ibcq);
+}
+
+struct c4iw_mpa_attributes {
+	u8 initiator;
+	u8 recv_marker_enabled;
+	u8 xmit_marker_enabled;
+	u8 crc_enabled;
+	u8 version;
+};
+
+struct c4iw_qp_attributes {
+	u32 scq;
+	u32 rcq;
+	u32 sq_num_entries;
+	u32 rq_num_entries;
+	u32 sq_max_sges;
+	u32 sq_max_sges_rdma_write;
+	u32 rq_max_sges;
+	u32 state;
+	u8 enable_rdma_read;
+	u8 enable_rdma_write;
+	u8 enable_bind;
+	u8 enable_mmid0_fastreg;
+	u32 max_ord;
+	u32 max_ird;
+	u32 pd;
+	u32 next_state;
+	char terminate_buffer[52];
+	u32 terminate_msg_len;
+	u8 is_terminate_local;
+	struct c4iw_mpa_attributes mpa_attr;
+	struct c4iw_ep *llp_stream_handle;
+};
+
+struct c4iw_qp {
+	struct ib_qp ibqp;
+	struct c4iw_dev *rhp;
+	struct c4iw_ep *ep;
+	struct c4iw_qp_attributes attr;
+	struct t4_wq wq;
+	spinlock_t lock;
+	atomic_t refcnt;
+	wait_queue_head_t wait;
+	struct timer_list timer;
+};
+
+static inline struct c4iw_qp *to_c4iw_qp(struct ib_qp *ibqp)
+{
+	return container_of(ibqp, struct c4iw_qp, ibqp);
+}
+
+struct c4iw_ucontext {
+	struct ib_ucontext ibucontext;
+	struct c4iw_dev_ucontext uctx;
+	u32 key;
+	spinlock_t mmap_lock;
+	struct list_head mmaps;
+};
+
+static inline struct c4iw_ucontext *to_c4iw_ucontext(struct ib_ucontext *c)
+{
+	return container_of(c, struct c4iw_ucontext, ibucontext);
+}
+
+struct c4iw_mm_entry {
+	struct list_head entry;
+	u64 addr;
+	u32 key;
+	unsigned len;
+};
+
+static inline struct c4iw_mm_entry *remove_mmap(struct c4iw_ucontext *ucontext,
+						u32 key, unsigned len)
+{
+	struct list_head *pos, *nxt;
+	struct c4iw_mm_entry *mm;
+
+	spin_lock(&ucontext->mmap_lock);
+	list_for_each_safe(pos, nxt, &ucontext->mmaps) {
+
+		mm = list_entry(pos, struct c4iw_mm_entry, entry);
+		if (mm->key == key && mm->len == len) {
+			list_del_init(&mm->entry);
+			spin_unlock(&ucontext->mmap_lock);
+			PDBG("%s key 0x%x addr 0x%llx len %d\n", __func__,
+			     key, (unsigned long long) mm->addr, mm->len);
+			return mm;
+		}
+	}
+	spin_unlock(&ucontext->mmap_lock);
+	return NULL;
+}
+
+static inline void insert_mmap(struct c4iw_ucontext *ucontext,
+			       struct c4iw_mm_entry *mm)
+{
+	spin_lock(&ucontext->mmap_lock);
+	PDBG("%s key 0x%x addr 0x%llx len %d\n", __func__,
+	     mm->key, (unsigned long long) mm->addr, mm->len);
+	list_add_tail(&mm->entry, &ucontext->mmaps);
+	spin_unlock(&ucontext->mmap_lock);
+}
+
+enum c4iw_qp_attr_mask {
+	C4IW_QP_ATTR_NEXT_STATE = 1 << 0,
+	C4IW_QP_ATTR_ENABLE_RDMA_READ = 1 << 7,
+	C4IW_QP_ATTR_ENABLE_RDMA_WRITE = 1 << 8,
+	C4IW_QP_ATTR_ENABLE_RDMA_BIND = 1 << 9,
+	C4IW_QP_ATTR_MAX_ORD = 1 << 11,
+	C4IW_QP_ATTR_MAX_IRD = 1 << 12,
+	C4IW_QP_ATTR_LLP_STREAM_HANDLE = 1 << 22,
+	C4IW_QP_ATTR_STREAM_MSG_BUFFER = 1 << 23,
+	C4IW_QP_ATTR_MPA_ATTR = 1 << 24,
+	C4IW_QP_ATTR_QP_CONTEXT_ACTIVATE = 1 << 25,
+	C4IW_QP_ATTR_VALID_MODIFY = (C4IW_QP_ATTR_ENABLE_RDMA_READ |
+				     C4IW_QP_ATTR_ENABLE_RDMA_WRITE |
+				     C4IW_QP_ATTR_MAX_ORD |
+				     C4IW_QP_ATTR_MAX_IRD |
+				     C4IW_QP_ATTR_LLP_STREAM_HANDLE |
+				     C4IW_QP_ATTR_STREAM_MSG_BUFFER |
+				     C4IW_QP_ATTR_MPA_ATTR |
+				     C4IW_QP_ATTR_QP_CONTEXT_ACTIVATE)
+};
+
+int c4iw_modify_qp(struct c4iw_dev *rhp,
+				struct c4iw_qp *qhp,
+				enum c4iw_qp_attr_mask mask,
+				struct c4iw_qp_attributes *attrs,
+				int internal);
+
+enum c4iw_qp_state {
+	C4IW_QP_STATE_IDLE,
+	C4IW_QP_STATE_RTS,
+	C4IW_QP_STATE_ERROR,
+	C4IW_QP_STATE_TERMINATE,
+	C4IW_QP_STATE_CLOSING,
+	C4IW_QP_STATE_TOT
+};
+
+static inline int c4iw_convert_state(enum ib_qp_state ib_state)
+{
+	switch (ib_state) {
+	case IB_QPS_RESET:
+	case IB_QPS_INIT:
+		return C4IW_QP_STATE_IDLE;
+	case IB_QPS_RTS:
+		return C4IW_QP_STATE_RTS;
+	case IB_QPS_SQD:
+		return C4IW_QP_STATE_CLOSING;
+	case IB_QPS_SQE:
+		return C4IW_QP_STATE_TERMINATE;
+	case IB_QPS_ERR:
+		return C4IW_QP_STATE_ERROR;
+	default:
+		return -1;
+	}
+}
+
+static inline u32 c4iw_ib_to_tpt_access(int a)
+{
+	return (a & IB_ACCESS_REMOTE_WRITE ? FW_RI_MEM_ACCESS_REM_WRITE : 0) |
+	       (a & IB_ACCESS_REMOTE_READ ? FW_RI_MEM_ACCESS_REM_READ : 0) |
+	       (a & IB_ACCESS_LOCAL_WRITE ? FW_RI_MEM_ACCESS_LOCAL_WRITE : 0) |
+	       FW_RI_MEM_ACCESS_LOCAL_READ;
+}
+
+static inline u32 c4iw_ib_to_tpt_bind_access(int acc)
+{
+	return (acc & IB_ACCESS_REMOTE_WRITE ? FW_RI_MEM_ACCESS_REM_WRITE : 0) |
+	       (acc & IB_ACCESS_REMOTE_READ ? FW_RI_MEM_ACCESS_REM_READ : 0);
+}
+
+enum c4iw_mmid_state {
+	C4IW_STAG_STATE_VALID,
+	C4IW_STAG_STATE_INVALID
+};
+
+#define C4IW_NODE_DESC "cxgb4 Chelsio Communications"
+
+#define MPA_KEY_REQ "MPA ID Req Frame"
+#define MPA_KEY_REP "MPA ID Rep Frame"
+
+#define MPA_MAX_PRIVATE_DATA	256
+#define MPA_REJECT		0x20
+#define MPA_CRC			0x40
+#define MPA_MARKERS		0x80
+#define MPA_FLAGS_MASK		0xE0
+
+#define c4iw_put_ep(ep) { \
+	PDBG("put_ep (via %s:%u) ep %p refcnt %d\n", __func__, __LINE__,  \
+	     ep, atomic_read(&((ep)->kref.refcount))); \
+	WARN_ON(atomic_read(&((ep)->kref.refcount)) < 1); \
+	kref_put(&((ep)->kref), _c4iw_free_ep); \
+}
+
+#define c4iw_get_ep(ep) { \
+	PDBG("get_ep (via %s:%u) ep %p, refcnt %d\n", __func__, __LINE__, \
+	     ep, atomic_read(&((ep)->kref.refcount))); \
+	kref_get(&((ep)->kref));  \
+}
+void _c4iw_free_ep(struct kref *kref);
+
+struct mpa_message {
+	u8 key[16];
+	u8 flags;
+	u8 revision;
+	__be16 private_data_size;
+	u8 private_data[0];
+};
+
+struct terminate_message {
+	u8 layer_etype;
+	u8 ecode;
+	__be16 hdrct_rsvd;
+	u8 len_hdrs[0];
+};
+
+#define TERM_MAX_LENGTH (sizeof(struct terminate_message) + 2 + 18 + 28)
+
+enum c4iw_layers_types {
+	LAYER_RDMAP		= 0x00,
+	LAYER_DDP		= 0x10,
+	LAYER_MPA		= 0x20,
+	RDMAP_LOCAL_CATA	= 0x00,
+	RDMAP_REMOTE_PROT	= 0x01,
+	RDMAP_REMOTE_OP		= 0x02,
+	DDP_LOCAL_CATA		= 0x00,
+	DDP_TAGGED_ERR		= 0x01,
+	DDP_UNTAGGED_ERR	= 0x02,
+	DDP_LLP			= 0x03
+};
+
+enum c4iw_rdma_ecodes {
+	RDMAP_INV_STAG		= 0x00,
+	RDMAP_BASE_BOUNDS	= 0x01,
+	RDMAP_ACC_VIOL		= 0x02,
+	RDMAP_STAG_NOT_ASSOC	= 0x03,
+	RDMAP_TO_WRAP		= 0x04,
+	RDMAP_INV_VERS		= 0x05,
+	RDMAP_INV_OPCODE	= 0x06,
+	RDMAP_STREAM_CATA	= 0x07,
+	RDMAP_GLOBAL_CATA	= 0x08,
+	RDMAP_CANT_INV_STAG	= 0x09,
+	RDMAP_UNSPECIFIED	= 0xff
+};
+
+enum c4iw_ddp_ecodes {
+	DDPT_INV_STAG		= 0x00,
+	DDPT_BASE_BOUNDS	= 0x01,
+	DDPT_STAG_NOT_ASSOC	= 0x02,
+	DDPT_TO_WRAP		= 0x03,
+	DDPT_INV_VERS		= 0x04,
+	DDPU_INV_QN		= 0x01,
+	DDPU_INV_MSN_NOBUF	= 0x02,
+	DDPU_INV_MSN_RANGE	= 0x03,
+	DDPU_INV_MO		= 0x04,
+	DDPU_MSG_TOOBIG		= 0x05,
+	DDPU_INV_VERS		= 0x06
+};
+
+enum c4iw_mpa_ecodes {
+	MPA_CRC_ERR		= 0x02,
+	MPA_MARKER_ERR		= 0x03
+};
+
+enum c4iw_ep_state {
+	IDLE = 0,
+	LISTEN,
+	CONNECTING,
+	MPA_REQ_WAIT,
+	MPA_REQ_SENT,
+	MPA_REQ_RCVD,
+	MPA_REP_SENT,
+	FPDU_MODE,
+	ABORTING,
+	CLOSING,
+	MORIBUND,
+	DEAD,
+};
+
+enum c4iw_ep_flags {
+	PEER_ABORT_IN_PROGRESS	= 0,
+	ABORT_REQ_IN_PROGRESS	= 1,
+	RELEASE_RESOURCES	= 2,
+	CLOSE_SENT		= 3,
+};
+
+struct c4iw_ep_common {
+	struct iw_cm_id *cm_id;
+	struct c4iw_qp *qp;
+	struct c4iw_dev *dev;
+	enum c4iw_ep_state state;
+	struct kref kref;
+	spinlock_t lock;
+	struct sockaddr_in local_addr;
+	struct sockaddr_in remote_addr;
+	wait_queue_head_t waitq;
+	int rpl_done;
+	int rpl_err;
+	unsigned long flags;
+};
+
+struct c4iw_listen_ep {
+	struct c4iw_ep_common com;
+	unsigned int stid;
+	int backlog;
+};
+
+struct c4iw_ep {
+	struct c4iw_ep_common com;
+	struct c4iw_ep *parent_ep;
+	struct timer_list timer;
+	unsigned int atid;
+	u32 hwtid;
+	u32 snd_seq;
+	u32 rcv_seq;
+	struct l2t_entry *l2t;
+	struct dst_entry *dst;
+	struct sk_buff *mpa_skb;
+	struct c4iw_mpa_attributes mpa_attr;
+	u8 mpa_pkt[sizeof(struct mpa_message) + MPA_MAX_PRIVATE_DATA];
+	unsigned int mpa_pkt_len;
+	u32 ird;
+	u32 ord;
+	u32 smac_idx;
+	u32 tx_chan;
+	u32 mtu;
+	u16 mss;
+	u16 emss;
+	u16 plen;
+	u16 rss_qid;
+	u16 txq_idx;
+	u8 tos;
+};
+
+static inline struct c4iw_ep *to_ep(struct iw_cm_id *cm_id)
+{
+	return cm_id->provider_data;
+}
+
+static inline struct c4iw_listen_ep *to_listen_ep(struct iw_cm_id *cm_id)
+{
+	return cm_id->provider_data;
+}
+
+static inline int compute_wscale(int win)
+{
+	int wscale = 0;
+
+	while (wscale < 14 && (65535<<wscale) < win)
+		wscale++;
+	return wscale;
+}
+
+typedef int (*c4iw_handler_func)(struct c4iw_dev *dev, struct sk_buff *skb);
+
+int c4iw_ep_redirect(void *ctx, struct dst_entry *old, struct dst_entry *new,
+		     struct l2t_entry *l2t);
+void c4iw_put_qpid(struct c4iw_rdev *rdev, u32 qpid,
+		   struct c4iw_dev_ucontext *uctx);
+u32 c4iw_get_resource(struct kfifo *fifo, spinlock_t *lock);
+void c4iw_put_resource(struct kfifo *fifo, u32 entry, spinlock_t *lock);
+int c4iw_init_resource(struct c4iw_rdev *rdev, u32 nr_tpt, u32 nr_pdid);
+int c4iw_init_ctrl_qp(struct c4iw_rdev *rdev);
+int c4iw_pblpool_create(struct c4iw_rdev *rdev);
+int c4iw_rqtpool_create(struct c4iw_rdev *rdev);
+void c4iw_pblpool_destroy(struct c4iw_rdev *rdev);
+void c4iw_rqtpool_destroy(struct c4iw_rdev *rdev);
+void c4iw_destroy_resource(struct c4iw_resource *rscp);
+int c4iw_destroy_ctrl_qp(struct c4iw_rdev *rdev);
+int c4iw_register_device(struct c4iw_dev *dev);
+void c4iw_unregister_device(struct c4iw_dev *dev);
+int __init c4iw_cm_init(void);
+void __exit c4iw_cm_term(void);
+void c4iw_release_dev_ucontext(struct c4iw_rdev *rdev,
+			       struct c4iw_dev_ucontext *uctx);
+void c4iw_init_dev_ucontext(struct c4iw_rdev *rdev,
+			    struct c4iw_dev_ucontext *uctx);
+int c4iw_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc);
+int c4iw_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
+		      struct ib_send_wr **bad_wr);
+int c4iw_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr,
+		      struct ib_recv_wr **bad_wr);
+int c4iw_bind_mw(struct ib_qp *qp, struct ib_mw *mw,
+		 struct ib_mw_bind *mw_bind);
+int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param);
+int c4iw_create_listen(struct iw_cm_id *cm_id, int backlog);
+int c4iw_destroy_listen(struct iw_cm_id *cm_id);
+int c4iw_accept_cr(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param);
+int c4iw_reject_cr(struct iw_cm_id *cm_id, const void *pdata, u8 pdata_len);
+void c4iw_qp_add_ref(struct ib_qp *qp);
+void c4iw_qp_rem_ref(struct ib_qp *qp);
+void c4iw_free_fastreg_pbl(struct ib_fast_reg_page_list *page_list);
+struct ib_fast_reg_page_list *c4iw_alloc_fastreg_pbl(
+					struct ib_device *device,
+					int page_list_len);
+struct ib_mr *c4iw_alloc_fast_reg_mr(struct ib_pd *pd, int pbl_depth);
+int c4iw_dealloc_mw(struct ib_mw *mw);
+struct ib_mw *c4iw_alloc_mw(struct ib_pd *pd);
+struct ib_mr *c4iw_reg_user_mr(struct ib_pd *pd, u64 start,
+					   u64 length, u64 virt, int acc,
+					   struct ib_udata *udata);
+struct ib_mr *c4iw_get_dma_mr(struct ib_pd *pd, int acc);
+struct ib_mr *c4iw_register_phys_mem(struct ib_pd *pd,
+					struct ib_phys_buf *buffer_list,
+					int num_phys_buf,
+					int acc,
+					u64 *iova_start);
+int c4iw_reregister_phys_mem(struct ib_mr *mr,
+				     int mr_rereg_mask,
+				     struct ib_pd *pd,
+				     struct ib_phys_buf *buffer_list,
+				     int num_phys_buf,
+				     int acc, u64 *iova_start);
+int c4iw_dereg_mr(struct ib_mr *ib_mr);
+int c4iw_destroy_cq(struct ib_cq *ib_cq);
+struct ib_cq *c4iw_create_cq(struct ib_device *ibdev, int entries,
+					int vector,
+					struct ib_ucontext *ib_context,
+					struct ib_udata *udata);
+int c4iw_resize_cq(struct ib_cq *cq, int cqe, struct ib_udata *udata);
+int c4iw_arm_cq(struct ib_cq *ibcq, enum ib_cq_notify_flags flags);
+int c4iw_destroy_qp(struct ib_qp *ib_qp);
+struct ib_qp *c4iw_create_qp(struct ib_pd *pd,
+			     struct ib_qp_init_attr *attrs,
+			     struct ib_udata *udata);
+int c4iw_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
+				 int attr_mask, struct ib_udata *udata);
+struct ib_qp *c4iw_get_qp(struct ib_device *dev, int qpn);
+u32 c4iw_rqtpool_alloc(struct c4iw_rdev *rdev, int size);
+void c4iw_rqtpool_free(struct c4iw_rdev *rdev, u32 addr, int size);
+u32 c4iw_pblpool_alloc(struct c4iw_rdev *rdev, int size);
+void c4iw_pblpool_free(struct c4iw_rdev *rdev, u32 addr, int size);
+int c4iw_ofld_send(struct c4iw_rdev *rdev, struct sk_buff *skb);
+void c4iw_flush_hw_cq(struct t4_cq *cq);
+void c4iw_count_rcqes(struct t4_cq *cq, struct t4_wq *wq, int *count);
+void c4iw_count_scqes(struct t4_cq *cq, struct t4_wq *wq, int *count);
+int c4iw_ep_disconnect(struct c4iw_ep *ep, int abrupt, gfp_t gfp);
+int c4iw_flush_rq(struct t4_wq *wq, struct t4_cq *cq, int count);
+int c4iw_flush_sq(struct t4_wq *wq, struct t4_cq *cq, int count);
+int c4iw_ev_handler(struct c4iw_dev *rnicp, u32 qid);
+u16 c4iw_rqes_posted(struct c4iw_qp *qhp);
+int c4iw_post_zb_read(struct c4iw_qp *qhp);
+int c4iw_post_terminate(struct c4iw_qp *qhp, struct t4_cqe *err_cqe);
+u32 c4iw_get_cqid(struct c4iw_rdev *rdev, struct c4iw_dev_ucontext *uctx);
+void c4iw_put_cqid(struct c4iw_rdev *rdev, u32 qid,
+		struct c4iw_dev_ucontext *uctx);
+u32 c4iw_get_qpid(struct c4iw_rdev *rdev, struct c4iw_dev_ucontext *uctx);
+void c4iw_put_qpid(struct c4iw_rdev *rdev, u32 qid,
+		struct c4iw_dev_ucontext *uctx);
+void c4iw_ev_dispatch(struct c4iw_dev *dev, struct t4_cqe *err_cqe);
+
+extern struct cxgb4_client t4c_client;
+extern c4iw_handler_func c4iw_handlers[NUM_CPL_CMDS];
+
+#endif

536 lines not on this page · raw message

798 lines not on this page · raw message

66 lines not on this page · raw message

new file mode 100644
index 0000000..be23b5e
@@ -0,0 +1,520 @@
+/*
+ * Copyright (c) 2009-2010 Chelsio, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *	  copyright notice, this list of conditions and the following
+ *	  disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *	  copyright notice, this list of conditions and the following
+ *	  disclaimer in the documentation and/or other materials
+ *	  provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/debugfs.h>
+
+#include <rdma/ib_verbs.h>
+
+#include "iw_cxgb4.h"
+
+#define DRV_VERSION "0.1"
+
+MODULE_AUTHOR("Steve Wise");
+MODULE_DESCRIPTION("Chelsio T4 RDMA Driver");
+MODULE_LICENSE("Dual BSD/GPL");
+MODULE_VERSION(DRV_VERSION);
+
+static LIST_HEAD(dev_list);
+static DEFINE_MUTEX(dev_mutex);
+
+static struct dentry *c4iw_debugfs_root;
+
+struct debugfs_qp_data {
+	struct c4iw_dev *devp;
+	char *buf;
+	int bufsize;
+	int pos;
+};
+
+static int count_qps(int id, void *p, void *data)
+{
+	struct c4iw_qp *qp = p;
+	int *countp = data;
+
+	if (id != qp->wq.sq.qid)
+		return 0;
+
+	*countp = *countp + 1;
+	return 0;
+}
+
+static int dump_qps(int id, void *p, void *data)
+{
+	struct c4iw_qp *qp = p;
+	struct debugfs_qp_data *qpd = data;
+	int space;
+	int cc;
+
+	if (id != qp->wq.sq.qid)
+		return 0;
+
+	space = qpd->bufsize - qpd->pos - 1;
+	if (space == 0)
+		return 1;
+
+	if (qp->ep)
+		cc = snprintf(qpd->buf + qpd->pos, space, "qp id %u state %u "
+			     "ep tid %u state %u %pI4:%u->%pI4:%u\n",
+			     qp->wq.sq.qid, (int)qp->attr.state,
+			     qp->ep->hwtid, (int)qp->ep->com.state,
+			     &qp->ep->com.local_addr.sin_addr.s_addr,
+			     ntohs(qp->ep->com.local_addr.sin_port),
+			     &qp->ep->com.remote_addr.sin_addr.s_addr,
+			     ntohs(qp->ep->com.remote_addr.sin_port));
+	else
+		cc = snprintf(qpd->buf + qpd->pos, space, "qp id %u state %u\n",
+			      qp->wq.sq.qid, (int)qp->attr.state);
+	if (cc < space)
+		qpd->pos += cc;
+	return 0;
+}
+
+static int qp_release(struct inode *inode, struct file *file)
+{
+	struct debugfs_qp_data *qpd = file->private_data;
+	if (!qpd) {
+		printk(KERN_INFO "%s null qpd?\n", __func__);
+		return 0;
+	}
+	kfree(qpd->buf);
+	kfree(qpd);
+	return 0;
+}
+
+static int qp_open(struct inode *inode, struct file *file)
+{
+	struct debugfs_qp_data *qpd;
+	int ret = 0;
+	int count = 1;
+
+	qpd = kmalloc(sizeof *qpd, GFP_KERNEL);
+	if (!qpd) {
+		ret = -ENOMEM;
+		goto out;
+	}
+	qpd->devp = inode->i_private;
+	qpd->pos = 0;
+
+	spin_lock_irq(&qpd->devp->lock);
+	idr_for_each(&qpd->devp->qpidr, count_qps, &count);
+	spin_unlock_irq(&qpd->devp->lock);
+
+	qpd->bufsize = count * 128;
+	qpd->buf = kmalloc(qpd->bufsize, GFP_KERNEL);
+	if (!qpd->buf) {
+		ret = -ENOMEM;
+		goto err1;
+	}
+
+	spin_lock_irq(&qpd->devp->lock);
+	idr_for_each(&qpd->devp->qpidr, dump_qps, qpd);
+	spin_unlock_irq(&qpd->devp->lock);
+
+	qpd->buf[qpd->pos++] = 0;
+	file->private_data = qpd;
+	goto out;
+err1:
+	kfree(qpd);
+out:
+	return ret;
+}
+
+static ssize_t qp_read(struct file *file, char __user *buf, size_t count,
+			loff_t *ppos)
+{
+	struct debugfs_qp_data *qpd = file->private_data;
+	loff_t pos = *ppos;
+	loff_t avail = qpd->pos;
+
+	if (pos < 0)
+		return -EINVAL;
+	if (pos >= avail)
+		return 0;
+	if (count > avail - pos)
+		count = avail - pos;
+
+	while (count) {
+		size_t len = 0;
+
+		len = min((int)count, (int)qpd->pos - (int)pos);
+		if (copy_to_user(buf, qpd->buf + pos, len))
+			return -EFAULT;
+		if (len == 0)
+			return -EINVAL;
+
+		buf += len;
+		pos += len;
+		count -= len;
+	}
+	count = pos - *ppos;
+	*ppos = pos;
+	return count;
+}
+
+static const struct file_operations qp_debugfs_fops = {
+	.owner   = THIS_MODULE,
+	.open    = qp_open,
+	.release = qp_release,
+	.read    = qp_read,
+};
+
+static int setup_debugfs(struct c4iw_dev *devp)
+{
+	struct dentry *de;
+
+	if (!devp->debugfs_root)
+		return -1;
+
+	de = debugfs_create_file("qps", S_IWUSR, devp->debugfs_root,
+				 (void *)devp, &qp_debugfs_fops);
+	if (de && de->d_inode)
+		de->d_inode->i_size = 4096;
+	return 0;
+}
+
+void c4iw_release_dev_ucontext(struct c4iw_rdev *rdev,
+			       struct c4iw_dev_ucontext *uctx)
+{
+	struct list_head *pos, *nxt;
+	struct c4iw_qid_list *entry;
+
+	mutex_lock(&uctx->lock);
+	list_for_each_safe(pos, nxt, &uctx->qpids) {
+		entry = list_entry(pos, struct c4iw_qid_list, entry);
+		list_del_init(&entry->entry);
+		if (!(entry->qid & rdev->qpmask))
+			c4iw_put_resource(&rdev->resource.qid_fifo, entry->qid,
+					  &rdev->resource.qid_fifo_lock);
+		kfree(entry);
+	}
+
+	list_for_each_safe(pos, nxt, &uctx->qpids) {
+		entry = list_entry(pos, struct c4iw_qid_list, entry);
+		list_del_init(&entry->entry);
+		kfree(entry);
+	}
+	mutex_unlock(&uctx->lock);
+}
+
+void c4iw_init_dev_ucontext(struct c4iw_rdev *rdev,
+			    struct c4iw_dev_ucontext *uctx)
+{
+	INIT_LIST_HEAD(&uctx->qpids);
+	INIT_LIST_HEAD(&uctx->cqids);
+	mutex_init(&uctx->lock);
+}
+
+/* Caller takes care of locking if needed */
+static int c4iw_rdev_open(struct c4iw_rdev *rdev)
+{
+	int err;
+
+	c4iw_init_dev_ucontext(rdev, &rdev->uctx);
+
+	/*
+	 * qpshift is the number of bits to shift the qpid left in order
+	 * to get the correct address of the doorbell for that qp.
+	 */
+	rdev->qpshift = PAGE_SHIFT - ilog2(rdev->lldi.udb_density);
+	rdev->qpmask = rdev->lldi.udb_density - 1;
+	rdev->cqshift = PAGE_SHIFT - ilog2(rdev->lldi.ucq_density);
+	rdev->cqmask = rdev->lldi.ucq_density - 1;
+	PDBG("%s dev %s stag start 0x%0x size 0x%0x num stags %d "
+	     "pbl start 0x%0x size 0x%0x rq start 0x%0x size 0x%0x\n",
+	     __func__, pci_name(rdev->lldi.pdev), rdev->lldi.vr->stag.start,
+	     rdev->lldi.vr->stag.size, c4iw_num_stags(rdev),
+	     rdev->lldi.vr->pbl.start,
+	     rdev->lldi.vr->pbl.size, rdev->lldi.vr->rq.start,
+	     rdev->lldi.vr->rq.size);
+	PDBG("udb len 0x%x udb base %p db_reg %p gts_reg %p qpshift %lu "
+	     "qpmask 0x%x cqshift %lu cqmask 0x%x\n",
+	     (unsigned)pci_resource_len(rdev->lldi.pdev, 2),
+	     (void *)pci_resource_start(rdev->lldi.pdev, 2),
+	     rdev->lldi.db_reg,
+	     rdev->lldi.gts_reg,
+	     rdev->qpshift, rdev->qpmask,
+	     rdev->cqshift, rdev->cqmask);
+
+	if (c4iw_num_stags(rdev) == 0) {
+		err = -EINVAL;
+		goto err1;
+	}
+
+	err = c4iw_init_resource(rdev, c4iw_num_stags(rdev), T4_MAX_NUM_PD);
+	if (err) {
+		printk(KERN_ERR MOD "error %d initializing resources\n", err);
+		goto err1;
+	}
+	err = c4iw_pblpool_create(rdev);
+	if (err) {
+		printk(KERN_ERR MOD "error %d initializing pbl pool\n", err);
+		goto err2;
+	}
+	err = c4iw_rqtpool_create(rdev);
+	if (err) {
+		printk(KERN_ERR MOD "error %d initializing rqt pool\n", err);
+		goto err3;
+	}
+	return 0;
+err3:
+	c4iw_pblpool_destroy(rdev);
+err2:
+	c4iw_destroy_resource(&rdev->resource);
+err1:
+	return err;
+}
+
+static void c4iw_rdev_close(struct c4iw_rdev *rdev)
+{
+	c4iw_pblpool_destroy(rdev);
+	c4iw_rqtpool_destroy(rdev);
+	c4iw_destroy_resource(&rdev->resource);
+}
+
+static void c4iw_remove(struct c4iw_dev *dev)
+{
+	PDBG("%s c4iw_dev %p\n", __func__,  dev);
+	cancel_delayed_work_sync(&dev->db_drop_task);
+	list_del(&dev->entry);
+	c4iw_unregister_device(dev);
+	c4iw_rdev_close(&dev->rdev);
+	idr_destroy(&dev->cqidr);
+	idr_destroy(&dev->qpidr);
+	idr_destroy(&dev->mmidr);
+	ib_dealloc_device(&dev->ibdev);
+}
+
+static struct c4iw_dev *c4iw_alloc(const struct cxgb4_lld_info *infop)
+{
+	struct c4iw_dev *devp;
+	int ret;
+
+	devp = (struct c4iw_dev *)ib_alloc_device(sizeof(*devp));
+	if (!devp) {
+		printk(KERN_ERR MOD "Cannot allocate ib device\n");
+		return NULL;
+	}
+	devp->rdev.lldi = *infop;
+
+	mutex_lock(&dev_mutex);
+
+	ret = c4iw_rdev_open(&devp->rdev);
+	if (ret) {
+		mutex_unlock(&dev_mutex);
+		printk(KERN_ERR MOD "Unable to open CXIO rdev err %d\n", ret);
+		ib_dealloc_device(&devp->ibdev);
+		return NULL;
+	}
+
+	idr_init(&devp->cqidr);
+	idr_init(&devp->qpidr);
+	idr_init(&devp->mmidr);
+	spin_lock_init(&devp->lock);
+	list_add_tail(&devp->entry, &dev_list);
+	mutex_unlock(&dev_mutex);
+
+	if (c4iw_register_device(devp)) {
+		printk(KERN_ERR MOD "Unable to register device\n");
+		mutex_lock(&dev_mutex);
+		c4iw_remove(devp);
+		mutex_unlock(&dev_mutex);
+	}
+	if (c4iw_debugfs_root) {
+		devp->debugfs_root = debugfs_create_dir(
+					pci_name(devp->rdev.lldi.pdev),
+					c4iw_debugfs_root);
+		setup_debugfs(devp);
+	}
+	return devp;
+}
+
+static void *c4iw_uld_add(const struct cxgb4_lld_info *infop)
+{
+	struct c4iw_dev *dev;
+	static int vers_printed;
+	int i;
+
+	if (!vers_printed++)
+		printk(KERN_INFO MOD "Chelsio T4 RDMA Driver - version %s\n",
+		       DRV_VERSION);
+
+	dev = c4iw_alloc(infop);
+	if (!dev)
+		goto out;
+
+	PDBG("%s found device %s nchan %u nrxq %u ntxq %u nports %u\n",
+	     __func__, pci_name(dev->rdev.lldi.pdev),
+	     dev->rdev.lldi.nchan, dev->rdev.lldi.nrxq,
+	     dev->rdev.lldi.ntxq, dev->rdev.lldi.nports);
+
+	for (i = 0; i < dev->rdev.lldi.nrxq; i++)
+		PDBG("rxqid[%u] %u\n", i, dev->rdev.lldi.rxq_ids[i]);
+
+	printk(KERN_INFO MOD "Initialized device %s\n",
+	       pci_name(dev->rdev.lldi.pdev));
+out:
+	return dev;
+}
+
+static struct sk_buff *t4_pktgl_to_skb(const struct pkt_gl *gl,
+				       unsigned int skb_len,
+				       unsigned int pull_len)
+{
+	struct sk_buff *skb;
+	struct skb_shared_info *ssi;
+
+	if (gl->tot_len <= 512) {
+		skb = alloc_skb(gl->tot_len, GFP_ATOMIC);
+		if (unlikely(!skb))
+			goto out;
+		__skb_put(skb, gl->tot_len);
+		skb_copy_to_linear_data(skb, gl->va, gl->tot_len);
+	} else {
+		skb = alloc_skb(skb_len, GFP_ATOMIC);
+		if (unlikely(!skb))
+			goto out;
+		__skb_put(skb, pull_len);
+		skb_copy_to_linear_data(skb, gl->va, pull_len);
+
+		ssi = skb_shinfo(skb);
+		ssi->frags[0].page = gl->frags[0].page;
+		ssi->frags[0].page_offset = gl->frags[0].page_offset + pull_len;
+		ssi->frags[0].size = gl->frags[0].size - pull_len;
+		if (gl->nfrags > 1)
+			memcpy(&ssi->frags[1], &gl->frags[1],
+			       (gl->nfrags - 1) * sizeof(skb_frag_t));
+		ssi->nr_frags = gl->nfrags;
+
+		skb->len = gl->tot_len;
+		skb->data_len = skb->len - pull_len;
+		skb->truesize += skb->data_len;
+
+		/* Get a reference for the last page, we don't own it */
+		get_page(gl->frags[gl->nfrags - 1].page);
+	}
+out:
+	return skb;
+}
+
+static int c4iw_uld_rx_handler(void *handle, const __be64 *rsp,
+			const struct pkt_gl *gl)
+{
+	struct c4iw_dev *dev = handle;
+	struct sk_buff *skb;
+	const struct cpl_act_establish *rpl;
+	unsigned int opcode;
+
+	if (gl == NULL) {
+		/* omit RSS and rsp_ctrl at end of descriptor */
+		unsigned int len = 64 - sizeof(struct rsp_ctrl) - 8;
+
+		skb = alloc_skb(256, GFP_ATOMIC);
+		if (!skb)
+			goto nomem;
+		__skb_put(skb, len);
+		skb_copy_to_linear_data(skb, &rsp[1], len);
+	} else if (gl == CXGB4_MSG_AN) {
+		const struct rsp_ctrl *rc = (void *)rsp;
+
+		u32 qid = be32_to_cpu(rc->pldbuflen_qid);
+		c4iw_ev_handler(dev, qid);
+		return 0;
+	} else {
+		skb = t4_pktgl_to_skb(gl, 128, 128);
+		if (unlikely(!skb))
+			goto nomem;
+	}
+
+	rpl = cplhdr(skb);
+	opcode = rpl->ot.opcode;
+
+	if (c4iw_handlers[opcode])
+		c4iw_handlers[opcode](dev, skb);
+	else
+		printk(KERN_INFO "%s no handler opcode 0x%x...\n", __func__,
+		       opcode);
+
+	return 0;
+nomem:
+	return -1;
+}
+
+static int c4iw_uld_state_change(void *handle, enum cxgb4_state new_state)
+{
+	PDBG("%s new_state %u\n", __func__, new_state);
+	return 0;
+}
+
+static struct cxgb4_uld_info c4iw_uld_info = {
+	.name = DRV_NAME,
+	.add = c4iw_uld_add,
+	.rx_handler = c4iw_uld_rx_handler,
+	.state_change = c4iw_uld_state_change,
+};
+
+static int __init c4iw_init_module(void)
+{
+	int err;
+
+	err = c4iw_cm_init();
+	if (err)
+		return err;
+
+	c4iw_debugfs_root = debugfs_create_dir(DRV_NAME, NULL);
+	if (!c4iw_debugfs_root)
+		printk(KERN_WARNING MOD
+		       "could not create debugfs entry, continuing\n");
+
+	cxgb4_register_uld(CXGB4_ULD_RDMA, &c4iw_uld_info);
+
+	return 0;
+}
+
+static void __exit c4iw_exit_module(void)
+{
+	struct c4iw_dev *dev, *tmp;
+
+	cxgb4_unregister_uld(CXGB4_ULD_RDMA);
+
+	mutex_lock(&dev_mutex);
+	list_for_each_entry_safe(dev, tmp, &dev_list, entry) {
+		c4iw_remove(dev);
+	}
+	mutex_unlock(&dev_mutex);
+
+	c4iw_cm_term();
+	debugfs_remove_recursive(c4iw_debugfs_root);
+}
+
+module_init(c4iw_init_module);
+module_exit(c4iw_exit_module);

new file mode 100644
index 0000000..3cb50af
@@ -0,0 +1,518 @@
+/*
+ * Copyright (c) 2009-2010 Chelsio, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/device.h>
+#include <linux/netdevice.h>
+#include <linux/etherdevice.h>
+#include <linux/delay.h>
+#include <linux/errno.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/ethtool.h>
+#include <linux/rtnetlink.h>
+#include <linux/inetdevice.h>
+#include <linux/io.h>
+
+#include <asm/irq.h>
+#include <asm/byteorder.h>
+
+#include <rdma/iw_cm.h>
+#include <rdma/ib_verbs.h>
+#include <rdma/ib_smi.h>
+#include <rdma/ib_umem.h>
+#include <rdma/ib_user_verbs.h>
+
+#include "iw_cxgb4.h"
+
+static int fastreg_support;
+module_param(fastreg_support, int, 0644);
+MODULE_PARM_DESC(fastreg_support, "Advertise fastreg support (default=0)");
+
+static int c4iw_modify_port(struct ib_device *ibdev,
+			    u8 port, int port_modify_mask,
+			    struct ib_port_modify *props)
+{
+	return -ENOSYS;
+}
+
+static struct ib_ah *c4iw_ah_create(struct ib_pd *pd,
+				    struct ib_ah_attr *ah_attr)
+{
+	return ERR_PTR(-ENOSYS);
+}
+
+static int c4iw_ah_destroy(struct ib_ah *ah)
+{
+	return -ENOSYS;
+}
+
+static int c4iw_multicast_attach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
+{
+	return -ENOSYS;
+}
+
+static int c4iw_multicast_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
+{
+	return -ENOSYS;
+}
+
+static int c4iw_process_mad(struct ib_device *ibdev, int mad_flags,
+			    u8 port_num, struct ib_wc *in_wc,
+			    struct ib_grh *in_grh, struct ib_mad *in_mad,
+			    struct ib_mad *out_mad)
+{
+	return -ENOSYS;
+}
+
+static int c4iw_dealloc_ucontext(struct ib_ucontext *context)
+{
+	struct c4iw_dev *rhp = to_c4iw_dev(context->device);
+	struct c4iw_ucontext *ucontext = to_c4iw_ucontext(context);
+	struct c4iw_mm_entry *mm, *tmp;
+
+	PDBG("%s context %p\n", __func__, context);
+	list_for_each_entry_safe(mm, tmp, &ucontext->mmaps, entry)
+		kfree(mm);
+	c4iw_release_dev_ucontext(&rhp->rdev, &ucontext->uctx);
+	kfree(ucontext);
+	return 0;
+}
+
+static struct ib_ucontext *c4iw_alloc_ucontext(struct ib_device *ibdev,
+					       struct ib_udata *udata)
+{
+	struct c4iw_ucontext *context;
+	struct c4iw_dev *rhp = to_c4iw_dev(ibdev);
+
+	PDBG("%s ibdev %p\n", __func__, ibdev);
+	context = kzalloc(sizeof(*context), GFP_KERNEL);
+	if (!context)
+		return ERR_PTR(-ENOMEM);
+	c4iw_init_dev_ucontext(&rhp->rdev, &context->uctx);
+	INIT_LIST_HEAD(&context->mmaps);
+	spin_lock_init(&context->mmap_lock);
+	return &context->ibucontext;
+}
+
+static int c4iw_mmap(struct ib_ucontext *context, struct vm_area_struct *vma)
+{
+	int len = vma->vm_end - vma->vm_start;
+	u32 key = vma->vm_pgoff << PAGE_SHIFT;
+	struct c4iw_rdev *rdev;
+	int ret = 0;
+	struct c4iw_mm_entry *mm;
+	struct c4iw_ucontext *ucontext;
+	u64 addr;
+
+	PDBG("%s pgoff 0x%lx key 0x%x len %d\n", __func__, vma->vm_pgoff,
+	     key, len);
+
+	if (vma->vm_start & (PAGE_SIZE-1))
+		return -EINVAL;
+
+	rdev = &(to_c4iw_dev(context->device)->rdev);
+	ucontext = to_c4iw_ucontext(context);
+
+	mm = remove_mmap(ucontext, key, len);
+	if (!mm)
+		return -EINVAL;
+	addr = mm->addr;
+	kfree(mm);
+
+	if ((addr >= pci_resource_start(rdev->lldi.pdev, 2)) &&
+	    (addr < (pci_resource_start(rdev->lldi.pdev, 2) +
+		       pci_resource_len(rdev->lldi.pdev, 2)))) {
+
+		/*
+		 * Map T4 DB register.
+		 */
+		if (vma->vm_flags & VM_READ)
+			return -EPERM;
+
+		vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
+		vma->vm_flags |= VM_DONTCOPY | VM_DONTEXPAND;
+		vma->vm_flags &= ~VM_MAYREAD;
+		ret = io_remap_pfn_range(vma, vma->vm_start,
+					 addr >> PAGE_SHIFT,
+					 len, vma->vm_page_prot);
+	} else {
+
+		/*
+		 * Map WQ or CQ contig dma memory...
+		 */
+		ret = remap_pfn_range(vma, vma->vm_start,
+				      addr >> PAGE_SHIFT,
+				      len, vma->vm_page_prot);
+	}
+
+	return ret;
+}
+
+static int c4iw_deallocate_pd(struct ib_pd *pd)
+{
+	struct c4iw_dev *rhp;
+	struct c4iw_pd *php;
+
+	php = to_c4iw_pd(pd);
+	rhp = php->rhp;
+	PDBG("%s ibpd %p pdid 0x%x\n", __func__, pd, php->pdid);
+	c4iw_put_resource(&rhp->rdev.resource.pdid_fifo, php->pdid,
+			  &rhp->rdev.resource.pdid_fifo_lock);
+	kfree(php);
+	return 0;
+}
+
+static struct ib_pd *c4iw_allocate_pd(struct ib_device *ibdev,
+				      struct ib_ucontext *context,
+				      struct ib_udata *udata)
+{
+	struct c4iw_pd *php;
+	u32 pdid;
+	struct c4iw_dev *rhp;
+
+	PDBG("%s ibdev %p\n", __func__, ibdev);
+	rhp = (struct c4iw_dev *) ibdev;
+	pdid =  c4iw_get_resource(&rhp->rdev.resource.pdid_fifo,
+				  &rhp->rdev.resource.pdid_fifo_lock);
+	if (!pdid)
+		return ERR_PTR(-EINVAL);
+	php = kzalloc(sizeof(*php), GFP_KERNEL);
+	if (!php) {
+		c4iw_put_resource(&rhp->rdev.resource.pdid_fifo, pdid,
+				  &rhp->rdev.resource.pdid_fifo_lock);
+		return ERR_PTR(-ENOMEM);
+	}
+	php->pdid = pdid;
+	php->rhp = rhp;
+	if (context) {
+		if (ib_copy_to_udata(udata, &php->pdid, sizeof(u32))) {
+			c4iw_deallocate_pd(&php->ibpd);
+			return ERR_PTR(-EFAULT);
+		}
+	}
+	PDBG("%s pdid 0x%0x ptr 0x%p\n", __func__, pdid, php);
+	return &php->ibpd;
+}
+
+static int c4iw_query_pkey(struct ib_device *ibdev, u8 port, u16 index,
+			   u16 *pkey)
+{
+	PDBG("%s ibdev %p\n", __func__, ibdev);
+	*pkey = 0;
+	return 0;
+}
+
+static int c4iw_query_gid(struct ib_device *ibdev, u8 port, int index,
+			  union ib_gid *gid)
+{
+	struct c4iw_dev *dev;
+
+	PDBG("%s ibdev %p, port %d, index %d, gid %p\n",
+	       __func__, ibdev, port, index, gid);
+	dev = to_c4iw_dev(ibdev);
+	BUG_ON(port == 0);
+	memset(&(gid->raw[0]), 0, sizeof(gid->raw));
+	memcpy(&(gid->raw[0]), dev->rdev.lldi.ports[port-1]->dev_addr, 6);
+	return 0;
+}
+
+static int c4iw_query_device(struct ib_device *ibdev,
+			     struct ib_device_attr *props)
+{
+
+	struct c4iw_dev *dev;
+	PDBG("%s ibdev %p\n", __func__, ibdev);
+
+	dev = to_c4iw_dev(ibdev);
+	memset(props, 0, sizeof *props);
+	memcpy(&props->sys_image_guid, dev->rdev.lldi.ports[0]->dev_addr, 6);
+	props->hw_ver = dev->rdev.lldi.adapter_type;
+	props->fw_ver = dev->rdev.lldi.fw_vers;
+	props->device_cap_flags = dev->device_cap_flags;
+	props->page_size_cap = T4_PAGESIZE_MASK;
+	props->vendor_id = (u32)dev->rdev.lldi.pdev->vendor;
+	props->vendor_part_id = (u32)dev->rdev.lldi.pdev->device;
+	props->max_mr_size = T4_MAX_MR_SIZE;
+	props->max_qp = T4_MAX_NUM_QP;
+	props->max_qp_wr = T4_MAX_QP_DEPTH;
+	props->max_sge = T4_MAX_RECV_SGE;
+	props->max_sge_rd = 1;
+	props->max_qp_rd_atom = T4_MAX_READ_DEPTH;
+	props->max_qp_init_rd_atom = T4_MAX_READ_DEPTH;
+	props->max_cq = T4_MAX_NUM_CQ;
+	props->max_cqe = T4_MAX_CQ_DEPTH;
+	props->max_mr = c4iw_num_stags(&dev->rdev);
+	props->max_pd = T4_MAX_NUM_PD;
+	props->local_ca_ack_delay = 0;
+	props->max_fast_reg_page_list_len = T4_MAX_FR_DEPTH;
+
+	return 0;
+}
+
+static int c4iw_query_port(struct ib_device *ibdev, u8 port,
+			   struct ib_port_attr *props)
+{
+	struct c4iw_dev *dev;
+	struct net_device *netdev;
+	struct in_device *inetdev;
+
+	PDBG("%s ibdev %p\n", __func__, ibdev);
+
+	dev = to_c4iw_dev(ibdev);
+	netdev = dev->rdev.lldi.ports[port-1];
+
+	memset(props, 0, sizeof(struct ib_port_attr));
+	props->max_mtu = IB_MTU_4096;
+	if (netdev->mtu >= 4096)
+		props->active_mtu = IB_MTU_4096;
+	else if (netdev->mtu >= 2048)
+		props->active_mtu = IB_MTU_2048;
+	else if (netdev->mtu >= 1024)
+		props->active_mtu = IB_MTU_1024;
+	else if (netdev->mtu >= 512)
+		props->active_mtu = IB_MTU_512;
+	else
+		props->active_mtu = IB_MTU_256;
+
+	if (!netif_carrier_ok(netdev))
+		props->state = IB_PORT_DOWN;
+	else {
+		inetdev = in_dev_get(netdev);
+		if (inetdev) {
+			if (inetdev->ifa_list)
+				props->state = IB_PORT_ACTIVE;
+			else
+				props->state = IB_PORT_INIT;
+			in_dev_put(inetdev);
+		} else
+			props->state = IB_PORT_INIT;
+	}
+
+	props->port_cap_flags =
+	    IB_PORT_CM_SUP |
+	    IB_PORT_SNMP_TUNNEL_SUP |
+	    IB_PORT_REINIT_SUP |
+	    IB_PORT_DEVICE_MGMT_SUP |
+	    IB_PORT_VENDOR_CLASS_SUP | IB_PORT_BOOT_MGMT_SUP;
+	props->gid_tbl_len = 1;
+	props->pkey_tbl_len = 1;
+	props->active_width = 2;
+	props->active_speed = 2;
+	props->max_msg_sz = -1;
+
+	return 0;
+}
+
+static ssize_t show_rev(struct device *dev, struct device_attribute *attr,
+			char *buf)
+{
+	struct c4iw_dev *c4iw_dev = container_of(dev, struct c4iw_dev,
+						 ibdev.dev);
+	PDBG("%s dev 0x%p\n", __func__, dev);
+	return sprintf(buf, "%d\n", c4iw_dev->rdev.lldi.adapter_type);
+}
+
+static ssize_t show_fw_ver(struct device *dev, struct device_attribute *attr,
+			   char *buf)
+{
+	struct c4iw_dev *c4iw_dev = container_of(dev, struct c4iw_dev,
+						 ibdev.dev);
+	PDBG("%s dev 0x%p\n", __func__, dev);
+
+	return sprintf(buf, "%u.%u.%u.%u\n",
+			FW_HDR_FW_VER_MAJOR_GET(c4iw_dev->rdev.lldi.fw_vers),
+			FW_HDR_FW_VER_MINOR_GET(c4iw_dev->rdev.lldi.fw_vers),
+			FW_HDR_FW_VER_MICRO_GET(c4iw_dev->rdev.lldi.fw_vers),
+			FW_HDR_FW_VER_BUILD_GET(c4iw_dev->rdev.lldi.fw_vers));
+}
+
+static ssize_t show_hca(struct device *dev, struct device_attribute *attr,
+			char *buf)
+{
+	struct c4iw_dev *c4iw_dev = container_of(dev, struct c4iw_dev,
+						 ibdev.dev);
+	struct ethtool_drvinfo info;
+	struct net_device *lldev = c4iw_dev->rdev.lldi.ports[0];
+
+	PDBG("%s dev 0x%p\n", __func__, dev);
+	lldev->ethtool_ops->get_drvinfo(lldev, &info);
+	return sprintf(buf, "%s\n", info.driver);
+}
+
+static ssize_t show_board(struct device *dev, struct device_attribute *attr,
+			  char *buf)
+{
+	struct c4iw_dev *c4iw_dev = container_of(dev, struct c4iw_dev,
+						 ibdev.dev);
+	PDBG("%s dev 0x%p\n", __func__, dev);
+	return sprintf(buf, "%x.%x\n", c4iw_dev->rdev.lldi.pdev->vendor,
+		       c4iw_dev->rdev.lldi.pdev->device);
+}
+
+static int c4iw_get_mib(struct ib_device *ibdev,
+			union rdma_protocol_stats *stats)
+{
+	return -ENOSYS;
+}
+
+static DEVICE_ATTR(hw_rev, S_IRUGO, show_rev, NULL);
+static DEVICE_ATTR(fw_ver, S_IRUGO, show_fw_ver, NULL);
+static DEVICE_ATTR(hca_type, S_IRUGO, show_hca, NULL);
+static DEVICE_ATTR(board_id, S_IRUGO, show_board, NULL);
+
+static struct device_attribute *c4iw_class_attributes[] = {
+	&dev_attr_hw_rev,
+	&dev_attr_fw_ver,
+	&dev_attr_hca_type,
+	&dev_attr_board_id,
+};
+
+int c4iw_register_device(struct c4iw_dev *dev)
+{
+	int ret;
+	int i;
+
+	PDBG("%s c4iw_dev %p\n", __func__, dev);
+	BUG_ON(!dev->rdev.lldi.ports[0]);
+	strlcpy(dev->ibdev.name, "cxgb4_%d", IB_DEVICE_NAME_MAX);
+	memset(&dev->ibdev.node_guid, 0, sizeof(dev->ibdev.node_guid));
+	memcpy(&dev->ibdev.node_guid, dev->rdev.lldi.ports[0]->dev_addr, 6);
+	dev->ibdev.owner = THIS_MODULE;
+	dev->device_cap_flags = IB_DEVICE_LOCAL_DMA_LKEY | IB_DEVICE_MEM_WINDOW;
+	if (fastreg_support)
+		dev->device_cap_flags |= IB_DEVICE_MEM_MGT_EXTENSIONS;
+	dev->ibdev.local_dma_lkey = 0;
+	dev->ibdev.uverbs_cmd_mask =
+	    (1ull << IB_USER_VERBS_CMD_GET_CONTEXT) |
+	    (1ull << IB_USER_VERBS_CMD_QUERY_DEVICE) |
+	    (1ull << IB_USER_VERBS_CMD_QUERY_PORT) |
+	    (1ull << IB_USER_VERBS_CMD_ALLOC_PD) |
+	    (1ull << IB_USER_VERBS_CMD_DEALLOC_PD) |
+	    (1ull << IB_USER_VERBS_CMD_REG_MR) |
+	    (1ull << IB_USER_VERBS_CMD_DEREG_MR) |
+	    (1ull << IB_USER_VERBS_CMD_CREATE_COMP_CHANNEL) |
+	    (1ull << IB_USER_VERBS_CMD_CREATE_CQ) |
+	    (1ull << IB_USER_VERBS_CMD_DESTROY_CQ) |
+	    (1ull << IB_USER_VERBS_CMD_REQ_NOTIFY_CQ) |
+	    (1ull << IB_USER_VERBS_CMD_CREATE_QP) |
+	    (1ull << IB_USER_VERBS_CMD_MODIFY_QP) |
+	    (1ull << IB_USER_VERBS_CMD_POLL_CQ) |
+	    (1ull << IB_USER_VERBS_CMD_DESTROY_QP) |
+	    (1ull << IB_USER_VERBS_CMD_POST_SEND) |
+	    (1ull << IB_USER_VERBS_CMD_POST_RECV);
+	dev->ibdev.node_type = RDMA_NODE_RNIC;
+	memcpy(dev->ibdev.node_desc, C4IW_NODE_DESC, sizeof(C4IW_NODE_DESC));
+	dev->ibdev.phys_port_cnt = dev->rdev.lldi.nports;
+	dev->ibdev.num_comp_vectors = 1;
+	dev->ibdev.dma_device = &(dev->rdev.lldi.pdev->dev);
+	dev->ibdev.query_device = c4iw_query_device;
+	dev->ibdev.query_port = c4iw_query_port;
+	dev->ibdev.modify_port = c4iw_modify_port;
+	dev->ibdev.query_pkey = c4iw_query_pkey;
+	dev->ibdev.query_gid = c4iw_query_gid;
+	dev->ibdev.alloc_ucontext = c4iw_alloc_ucontext;
+	dev->ibdev.dealloc_ucontext = c4iw_dealloc_ucontext;
+	dev->ibdev.mmap = c4iw_mmap;
+	dev->ibdev.alloc_pd = c4iw_allocate_pd;
+	dev->ibdev.dealloc_pd = c4iw_deallocate_pd;
+	dev->ibdev.create_ah = c4iw_ah_create;
+	dev->ibdev.destroy_ah = c4iw_ah_destroy;
+	dev->ibdev.create_qp = c4iw_create_qp;
+	dev->ibdev.modify_qp = c4iw_ib_modify_qp;
+	dev->ibdev.destroy_qp = c4iw_destroy_qp;
+	dev->ibdev.create_cq = c4iw_create_cq;
+	dev->ibdev.destroy_cq = c4iw_destroy_cq;
+	dev->ibdev.resize_cq = c4iw_resize_cq;
+	dev->ibdev.poll_cq = c4iw_poll_cq;
+	dev->ibdev.get_dma_mr = c4iw_get_dma_mr;
+	dev->ibdev.reg_phys_mr = c4iw_register_phys_mem;
+	dev->ibdev.rereg_phys_mr = c4iw_reregister_phys_mem;
+	dev->ibdev.reg_user_mr = c4iw_reg_user_mr;
+	dev->ibdev.dereg_mr = c4iw_dereg_mr;
+	dev->ibdev.alloc_mw = c4iw_alloc_mw;
+	dev->ibdev.bind_mw = c4iw_bind_mw;
+	dev->ibdev.dealloc_mw = c4iw_dealloc_mw;
+	dev->ibdev.alloc_fast_reg_mr = c4iw_alloc_fast_reg_mr;
+	dev->ibdev.alloc_fast_reg_page_list = c4iw_alloc_fastreg_pbl;
+	dev->ibdev.free_fast_reg_page_list = c4iw_free_fastreg_pbl;
+	dev->ibdev.attach_mcast = c4iw_multicast_attach;
+	dev->ibdev.detach_mcast = c4iw_multicast_detach;
+	dev->ibdev.process_mad = c4iw_process_mad;
+	dev->ibdev.req_notify_cq = c4iw_arm_cq;
+	dev->ibdev.post_send = c4iw_post_send;
+	dev->ibdev.post_recv = c4iw_post_receive;
+	dev->ibdev.get_protocol_stats = c4iw_get_mib;
+
+	dev->ibdev.iwcm = kmalloc(sizeof(struct iw_cm_verbs), GFP_KERNEL);
+	if (!dev->ibdev.iwcm)
+		return -ENOMEM;
+
+	dev->ibdev.iwcm->connect = c4iw_connect;
+	dev->ibdev.iwcm->accept = c4iw_accept_cr;
+	dev->ibdev.iwcm->reject = c4iw_reject_cr;
+	dev->ibdev.iwcm->create_listen = c4iw_create_listen;
+	dev->ibdev.iwcm->destroy_listen = c4iw_destroy_listen;
+	dev->ibdev.iwcm->add_ref = c4iw_qp_add_ref;
+	dev->ibdev.iwcm->rem_ref = c4iw_qp_rem_ref;
+	dev->ibdev.iwcm->get_qp = c4iw_get_qp;
+
+	ret = ib_register_device(&dev->ibdev);
+	if (ret)
+		goto bail1;
+
+	for (i = 0; i < ARRAY_SIZE(c4iw_class_attributes); ++i) {
+		ret = device_create_file(&dev->ibdev.dev,
+					 c4iw_class_attributes[i]);
+		if (ret)
+			goto bail2;
+	}
+	return 0;
+bail2:
+	ib_unregister_device(&dev->ibdev);
+bail1:
+	kfree(dev->ibdev.iwcm);
+	return ret;
+}
+
+void c4iw_unregister_device(struct c4iw_dev *dev)
+{
+	int i;
+
+	PDBG("%s c4iw_dev %p\n", __func__, dev);
+	for (i = 0; i < ARRAY_SIZE(c4iw_class_attributes); ++i)
+		device_remove_file(&dev->ibdev.dev,
+				   c4iw_class_attributes[i]);
+	ib_unregister_device(&dev->ibdev);
+	kfree(dev->ibdev.iwcm);
+	return;
+}

new file mode 100644
index 0000000..e1c1ee0
@@ -0,0 +1,2365 @@
+/*
+ * Copyright (c) 2009-2010 Chelsio, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *	  copyright notice, this list of conditions and the following
+ *	  disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *	  copyright notice, this list of conditions and the following
+ *	  disclaimer in the documentation and/or other materials
+ *	  provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include <linux/module.h>
+#include <linux/list.h>
+#include <linux/workqueue.h>
+#include <linux/skbuff.h>
+#include <linux/timer.h>
+#include <linux/notifier.h>
+#include <linux/inetdevice.h>
+#include <linux/ip.h>
+#include <linux/tcp.h>
+
+#include <net/neighbour.h>
+#include <net/netevent.h>
+#include <net/route.h>
+
+#include "iw_cxgb4.h"
+
+static char *states[] = {
+	"idle",
+	"listen",
+	"connecting",
+	"mpa_wait_req",
+	"mpa_req_sent",
+	"mpa_req_rcvd",
+	"mpa_rep_sent",
+	"fpdu_mode",
+	"aborting",
+	"closing",
+	"moribund",
+	"dead",
+	NULL,
+};
+
+static int enable_tcp_timestamps;
+module_param(enable_tcp_timestamps, int, 0644);
+MODULE_PARM_DESC(enable_tcp_timestamps, "Enable tcp timestamps (default=0)");
+
+static int enable_tcp_sack;
+module_param(enable_tcp_sack, int, 0644);
+MODULE_PARM_DESC(enable_tcp_sack, "Enable tcp SACK (default=0)");
+
+static int enable_tcp_window_scaling = 1;
+module_param(enable_tcp_window_scaling, int, 0644);
+MODULE_PARM_DESC(enable_tcp_window_scaling,
+		 "Enable tcp window scaling (default=1)");
+
+int c4iw_debug;
+module_param(c4iw_debug, int, 0644);
+MODULE_PARM_DESC(c4iw_debug, "Enable debug logging (default=0)");
+
+static int peer2peer;
+module_param(peer2peer, int, 0644);
+MODULE_PARM_DESC(peer2peer, "Support peer2peer ULPs (default=0)");
+
+static int ep_timeout_secs = 60;
+module_param(ep_timeout_secs, int, 0644);
+MODULE_PARM_DESC(ep_timeout_secs, "CM Endpoint operation timeout "
+				   "in seconds (default=60)");
+
+static int mpa_rev = 1;
+module_param(mpa_rev, int, 0644);
+MODULE_PARM_DESC(mpa_rev, "MPA Revision, 0 supports amso1100, "
+		 "1 is spec compliant. (default=1)");
+
+static int markers_enabled;
+module_param(markers_enabled, int, 0644);
+MODULE_PARM_DESC(markers_enabled, "Enable MPA MARKERS (default(0)=disabled)");
+
+static int crc_enabled = 1;
+module_param(crc_enabled, int, 0644);
+MODULE_PARM_DESC(crc_enabled, "Enable MPA CRC (default(1)=enabled)");
+
+static int rcv_win = 256 * 1024;
+module_param(rcv_win, int, 0644);
+MODULE_PARM_DESC(rcv_win, "TCP receive window in bytes (default=256KB)");
+
+static int snd_win = 32 * 1024;
+module_param(snd_win, int, 0644);
+MODULE_PARM_DESC(snd_win, "TCP send window in bytes (default=32KB)");
+
+static void process_work(struct work_struct *work);
+static struct workqueue_struct *workq;
+static DECLARE_WORK(skb_work, process_work);
+
+static struct sk_buff_head rxq;
+static c4iw_handler_func work_handlers[NUM_CPL_CMDS];
+c4iw_handler_func c4iw_handlers[NUM_CPL_CMDS];
+
+static struct sk_buff *get_skb(struct sk_buff *skb, int len, gfp_t gfp);
+static void ep_timeout(unsigned long arg);
+static void connect_reply_upcall(struct c4iw_ep *ep, int status);
+
+static void start_ep_timer(struct c4iw_ep *ep)
+{
+	PDBG("%s ep %p\n", __func__, ep);
+	if (timer_pending(&ep->timer)) {
+		PDBG("%s stopped / restarted timer ep %p\n", __func__, ep);
+		del_timer_sync(&ep->timer);
+	} else
+		c4iw_get_ep(&ep->com);
+	ep->timer.expires = jiffies + ep_timeout_secs * HZ;
+	ep->timer.data = (unsigned long)ep;
+	ep->timer.function = ep_timeout;
+	add_timer(&ep->timer);
+}
+
+static void stop_ep_timer(struct c4iw_ep *ep)
+{
+	PDBG("%s ep %p\n", __func__, ep);
+	if (!timer_pending(&ep->timer)) {
+		printk(KERN_ERR "%s timer stopped when its not running! "
+		       "ep %p state %u\n", __func__, ep, ep->com.state);
+		WARN_ON(1);
+		return;
+	}
+	del_timer_sync(&ep->timer);
+	c4iw_put_ep(&ep->com);
+}
+
+static int c4iw_l2t_send(struct c4iw_rdev *rdev, struct sk_buff *skb,
+		  struct l2t_entry *l2e)
+{
+	int	error = 0;
+
+	if (c4iw_fatal_error(rdev)) {
+		kfree_skb(skb);
+		PDBG("%s - device in error state - dropping\n", __func__);
+		return -EIO;
+	}
+	error = cxgb4_l2t_send(rdev->lldi.ports[0], skb, l2e);
+	if (error < 0)
+		kfree_skb(skb);
+	return error;
+}
+
+int c4iw_ofld_send(struct c4iw_rdev *rdev, struct sk_buff *skb)
+{
+	int	error = 0;
+
+	if (c4iw_fatal_error(rdev)) {
+		kfree_skb(skb);
+		PDBG("%s - device in error state - dropping\n", __func__);
+		return -EIO;
+	}
+	error = cxgb4_ofld_send(rdev->lldi.ports[0], skb);
+	if (error < 0)
+		kfree_skb(skb);
+	return error;
+}
+
+static void release_tid(struct c4iw_rdev *rdev, u32 hwtid, struct sk_buff *skb)
+{
+	struct cpl_tid_release *req;
+
+	skb = get_skb(skb, sizeof *req, GFP_KERNEL);
+	if (!skb)
+		return;
+	req = (struct cpl_tid_release *) skb_put(skb, sizeof(*req));
+	INIT_TP_WR(req, hwtid);
+	OPCODE_TID(req) = cpu_to_be32(MK_OPCODE_TID(CPL_TID_RELEASE, hwtid));
+	set_wr_txq(skb, CPL_PRIORITY_SETUP, 0);
+	c4iw_ofld_send(rdev, skb);
+	return;
+}
+
+static void set_emss(struct c4iw_ep *ep, u16 opt)
+{
+	ep->emss = ep->com.dev->rdev.lldi.mtus[GET_TCPOPT_MSS(opt)] - 40;
+	ep->mss = ep->emss;
+	if (GET_TCPOPT_TSTAMP(opt))
+		ep->emss -= 12;
+	if (ep->emss < 128)
+		ep->emss = 128;
+	PDBG("%s mss_idx %u mss %u emss=%u\n", __func__, GET_TCPOPT_MSS(opt),
+	     ep->mss, ep->emss);
+}
+
+static enum c4iw_ep_state state_read(struct c4iw_ep_common *epc)
+{
+	unsigned long flags;
+	enum c4iw_ep_state state;
+
+	spin_lock_irqsave(&epc->lock, flags);
+	state = epc->state;
+	spin_unlock_irqrestore(&epc->lock, flags);
+	return state;
+}
+
+static void __state_set(struct c4iw_ep_common *epc, enum c4iw_ep_state new)
+{
+	epc->state = new;
+}
+
+static void state_set(struct c4iw_ep_common *epc, enum c4iw_ep_state new)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&epc->lock, flags);
+	PDBG("%s - %s -> %s\n", __func__, states[epc->state], states[new]);
+	__state_set(epc, new);
+	spin_unlock_irqrestore(&epc->lock, flags);
+	return;
+}
+
+static void *alloc_ep(int size, gfp_t gfp)
+{
+	struct c4iw_ep_common *epc;
+
+	epc = kzalloc(size, gfp);
+	if (epc) {
+		kref_init(&epc->kref);
+		spin_lock_init(&epc->lock);
+		init_waitqueue_head(&epc->waitq);
+	}
+	PDBG("%s alloc ep %p\n", __func__, epc);
+	return epc;
+}
+
+void _c4iw_free_ep(struct kref *kref)
+{
+	struct c4iw_ep *ep;
+
+	ep = container_of(kref, struct c4iw_ep, com.kref);
+	PDBG("%s ep %p state %s\n", __func__, ep, states[state_read(&ep->com)]);
+	if (test_bit(RELEASE_RESOURCES, &ep->com.flags)) {
+		cxgb4_remove_tid(ep->com.dev->rdev.lldi.tids, 0, ep->hwtid);
+		dst_release(ep->dst);
+		cxgb4_l2t_release(ep->l2t);
+	}
+	kfree(ep);
+}
+
+static void release_ep_resources(struct c4iw_ep *ep)
+{
+	set_bit(RELEASE_RESOURCES, &ep->com.flags);
+	c4iw_put_ep(&ep->com);
+}
+
+static void process_work(struct work_struct *work)
+{
+	struct sk_buff *skb = NULL;
+	struct c4iw_dev *dev;
+	struct cpl_act_establish *rpl = cplhdr(skb);
+	unsigned int opcode;
+	int ret;
+
+	while ((skb = skb_dequeue(&rxq))) {
+		rpl = cplhdr(skb);
+		dev = *((struct c4iw_dev **) (skb->cb + sizeof(void *)));
+		opcode = rpl->ot.opcode;
+
+		BUG_ON(!work_handlers[opcode]);
+		ret = work_handlers[opcode](dev, skb);
+		if (!ret)
+			kfree_skb(skb);
+	}
+}
+
+static int status2errno(int status)
+{
+	switch (status) {
+	case CPL_ERR_NONE:
+		return 0;
+	case CPL_ERR_CONN_RESET:
+		return -ECONNRESET;
+	case CPL_ERR_ARP_MISS:
+		return -EHOSTUNREACH;
+	case CPL_ERR_CONN_TIMEDOUT:
+		return -ETIMEDOUT;
+	case CPL_ERR_TCAM_FULL:
+		return -ENOMEM;
+	case CPL_ERR_CONN_EXIST:
+		return -EADDRINUSE;
+	default:
+		return -EIO;
+	}
+}
+
+/*
+ * Try and reuse skbs already allocated...
+ */
+static struct sk_buff *get_skb(struct sk_buff *skb, int len, gfp_t gfp)
+{
+	if (skb && !skb_is_nonlinear(skb) && !skb_cloned(skb)) {
+		skb_trim(skb, 0);
+		skb_get(skb);
+		skb_reset_transport_header(skb);
+	} else {
+		skb = alloc_skb(len, gfp);
+	}
+	return skb;
+}
+
+static struct rtable *find_route(struct c4iw_dev *dev, __be32 local_ip,
+				 __be32 peer_ip, __be16 local_port,
+				 __be16 peer_port, u8 tos)
+{
+	struct rtable *rt;
+	struct flowi fl = {
+		.oif = 0,
+		.nl_u = {
+			 .ip4_u = {
+				   .daddr = peer_ip,
+				   .saddr = local_ip,
+				   .tos = tos}
+			 },
+		.proto = IPPROTO_TCP,
+		.uli_u = {
+			  .ports = {
+				    .sport = local_port,
+				    .dport = peer_port}
+			  }
+	};
+
+	if (ip_route_output_flow(&init_net, &rt, &fl, NULL, 0))
+		return NULL;
+	return rt;
+}
+
+static void arp_failure_discard(void *handle, struct sk_buff *skb)
+{
+	PDBG("%s c4iw_dev %p\n", __func__, handle);
+	kfree_skb(skb);
+}
+
+/*
+ * Handle an ARP failure for an active open.
+ */
+static void act_open_req_arp_failure(void *handle, struct sk_buff *skb)
+{
+	printk(KERN_ERR MOD "ARP failure duing connect\n");
+	kfree_skb(skb);
+}
+
+/*
+ * Handle an ARP failure for a CPL_ABORT_REQ.  Change it into a no RST variant
+ * and send it along.
+ */
+static void abort_arp_failure(void *handle, struct sk_buff *skb)
+{
+	struct c4iw_rdev *rdev = handle;
+	struct cpl_abort_req *req = cplhdr(skb);
+
+	PDBG("%s rdev %p\n", __func__, rdev);
+	req->cmd = CPL_ABORT_NO_RST;
+	c4iw_ofld_send(rdev, skb);
+}
+
+static void send_flowc(struct c4iw_ep *ep, struct sk_buff *skb)
+{
+	unsigned int flowclen = 80;
+	struct fw_flowc_wr *flowc;
+	int i;
+
+	skb = get_skb(skb, flowclen, GFP_KERNEL);
+	flowc = (struct fw_flowc_wr *)__skb_put(skb, flowclen);
+
+	flowc->op_to_nparams = cpu_to_be32(FW_WR_OP(FW_FLOWC_WR) |
+					   FW_FLOWC_WR_NPARAMS(8));
+	flowc->flowid_len16 = cpu_to_be32(FW_WR_LEN16(DIV_ROUND_UP(flowclen,
+					  16)) | FW_WR_FLOWID(ep->hwtid));
+
+	flowc->mnemval[0].mnemonic = FW_FLOWC_MNEM_PFNVFN;
+	flowc->mnemval[0].val = cpu_to_be32(0);
+	flowc->mnemval[1].mnemonic = FW_FLOWC_MNEM_CH;
+	flowc->mnemval[1].val = cpu_to_be32(ep->tx_chan);
+	flowc->mnemval[2].mnemonic = FW_FLOWC_MNEM_PORT;
+	flowc->mnemval[2].val = cpu_to_be32(ep->tx_chan);
+	flowc->mnemval[3].mnemonic = FW_FLOWC_MNEM_IQID;
+	flowc->mnemval[3].val = cpu_to_be32(ep->rss_qid);
+	flowc->mnemval[4].mnemonic = FW_FLOWC_MNEM_SNDNXT;
+	flowc->mnemval[4].val = cpu_to_be32(ep->snd_seq);
+	flowc->mnemval[5].mnemonic = FW_FLOWC_MNEM_RCVNXT;
+	flowc->mnemval[5].val = cpu_to_be32(ep->rcv_seq);
+	flowc->mnemval[6].mnemonic = FW_FLOWC_MNEM_SNDBUF;
+	flowc->mnemval[6].val = cpu_to_be32(snd_win);
+	flowc->mnemval[7].mnemonic = FW_FLOWC_MNEM_MSS;
+	flowc->mnemval[7].val = cpu_to_be32(ep->emss);
+	/* Pad WR to 16 byte boundary */
+	flowc->mnemval[8].mnemonic = 0;
+	flowc->mnemval[8].val = 0;
+	for (i = 0; i < 9; i++) {
+		flowc->mnemval[i].r4[0] = 0;
+		flowc->mnemval[i].r4[1] = 0;
+		flowc->mnemval[i].r4[2] = 0;
+	}
+
+	set_wr_txq(skb, CPL_PRIORITY_DATA, ep->txq_idx);
+	c4iw_ofld_send(&ep->com.dev->rdev, skb);
+}
+
+static int send_halfclose(struct c4iw_ep *ep, gfp_t gfp)
+{
+	struct cpl_close_con_req *req;
+	struct sk_buff *skb;
+	int wrlen = roundup(sizeof *req, 16);
+
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+	skb = get_skb(NULL, wrlen, gfp);
+	if (!skb) {
+		printk(KERN_ERR MOD "%s - failed to alloc skb\n", __func__);
+		return -ENOMEM;
+	}
+	set_wr_txq(skb, CPL_PRIORITY_DATA, ep->txq_idx);
+	t4_set_arp_err_handler(skb, NULL, arp_failure_discard);
+	req = (struct cpl_close_con_req *) skb_put(skb, wrlen);
+	memset(req, 0, wrlen);
+	INIT_TP_WR(req, ep->hwtid);
+	OPCODE_TID(req) = cpu_to_be32(MK_OPCODE_TID(CPL_CLOSE_CON_REQ,
+						    ep->hwtid));
+	return c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t);
+}
+
+static int send_abort(struct c4iw_ep *ep, struct sk_buff *skb, gfp_t gfp)
+{
+	struct cpl_abort_req *req;
+	int wrlen = roundup(sizeof *req, 16);
+
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+	skb = get_skb(skb, wrlen, gfp);
+	if (!skb) {
+		printk(KERN_ERR MOD "%s - failed to alloc skb.\n",
+		       __func__);
+		return -ENOMEM;
+	}
+	set_wr_txq(skb, CPL_PRIORITY_DATA, ep->txq_idx);
+	t4_set_arp_err_handler(skb, &ep->com.dev->rdev, abort_arp_failure);
+	req = (struct cpl_abort_req *) skb_put(skb, wrlen);
+	memset(req, 0, wrlen);
+	INIT_TP_WR(req, ep->hwtid);
+	OPCODE_TID(req) = cpu_to_be32(MK_OPCODE_TID(CPL_ABORT_REQ, ep->hwtid));
+	req->cmd = CPL_ABORT_SEND_RST;
+	return c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t);
+}
+
+static int send_connect(struct c4iw_ep *ep)
+{
+	struct cpl_act_open_req *req;
+	struct sk_buff *skb;
+	u64 opt0;
+	u32 opt2;
+	unsigned int mtu_idx;
+	int wscale;
+	int wrlen = roundup(sizeof *req, 16);
+
+	PDBG("%s ep %p atid %u\n", __func__, ep, ep->atid);
+
+	skb = get_skb(NULL, wrlen, GFP_KERNEL);
+	if (!skb) {
+		printk(KERN_ERR MOD "%s - failed to alloc skb.\n",
+		       __func__);
+		return -ENOMEM;
+	}
+	set_wr_txq(skb, CPL_PRIORITY_SETUP, ep->txq_idx);
+
+	cxgb4_best_mtu(ep->com.dev->rdev.lldi.mtus, ep->mtu, &mtu_idx);
+	wscale = compute_wscale(rcv_win);
+	opt0 = KEEP_ALIVE(1) |
+	       WND_SCALE(wscale) |
+	       MSS_IDX(mtu_idx) |
+	       L2T_IDX(ep->l2t->idx) |
+	       TX_CHAN(ep->tx_chan) |
+	       SMAC_SEL(ep->smac_idx) |
+	       DSCP(ep->tos) |
+	       RCV_BUFSIZ(rcv_win>>10);
+	opt2 = RX_CHANNEL(0) |
+	       RSS_QUEUE_VALID | RSS_QUEUE(ep->rss_qid);
+	if (enable_tcp_timestamps)
+		opt2 |= TSTAMPS_EN(1);
+	if (enable_tcp_sack)
+		opt2 |= SACK_EN(1);
+	if (wscale && enable_tcp_window_scaling)
+		opt2 |= WND_SCALE_EN(1);
+	t4_set_arp_err_handler(skb, NULL, act_open_req_arp_failure);
+
+	req = (struct cpl_act_open_req *) skb_put(skb, wrlen);
+	INIT_TP_WR(req, 0);
+	OPCODE_TID(req) = cpu_to_be32(
+		MK_OPCODE_TID(CPL_ACT_OPEN_REQ, ((ep->rss_qid<<14)|ep->atid)));
+	req->local_port = ep->com.local_addr.sin_port;
+	req->peer_port = ep->com.remote_addr.sin_port;
+	req->local_ip = ep->com.local_addr.sin_addr.s_addr;
+	req->peer_ip = ep->com.remote_addr.sin_addr.s_addr;
+	req->opt0 = cpu_to_be64(opt0);
+	req->params = 0;
+	req->opt2 = cpu_to_be32(opt2);
+	return c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t);
+}
+
+static void send_mpa_req(struct c4iw_ep *ep, struct sk_buff *skb)
+{
+	int mpalen, wrlen;
+	struct fw_ofld_tx_data_wr *req;
+	struct mpa_message *mpa;
+
+	PDBG("%s ep %p tid %u pd_len %d\n", __func__, ep, ep->hwtid, ep->plen);
+
+	BUG_ON(skb_cloned(skb));
+
+	mpalen = sizeof(*mpa) + ep->plen;
+	wrlen = roundup(mpalen + sizeof *req, 16);
+	skb = get_skb(skb, wrlen, GFP_KERNEL);
+	if (!skb) {
+		connect_reply_upcall(ep, -ENOMEM);
+		return;
+	}
+	set_wr_txq(skb, CPL_PRIORITY_DATA, ep->txq_idx);
+
+	req = (struct fw_ofld_tx_data_wr *)skb_put(skb, wrlen);
+	memset(req, 0, wrlen);
+	req->op_to_immdlen = cpu_to_be32(
+		FW_WR_OP(FW_OFLD_TX_DATA_WR) |
+		FW_WR_COMPL(1) |
+		FW_WR_IMMDLEN(mpalen));
+	req->flowid_len16 = cpu_to_be32(
+		FW_WR_FLOWID(ep->hwtid) |
+		FW_WR_LEN16(wrlen >> 4));
+	req->plen = cpu_to_be32(mpalen);
+	req->tunnel_to_proxy = cpu_to_be32(
+		FW_OFLD_TX_DATA_WR_FLUSH(1) |
+		FW_OFLD_TX_DATA_WR_SHOVE(1));
+
+	mpa = (struct mpa_message *)(req + 1);
+	memcpy(mpa->key, MPA_KEY_REQ, sizeof(mpa->key));
+	mpa->flags = (crc_enabled ? MPA_CRC : 0) |
+		     (markers_enabled ? MPA_MARKERS : 0);
+	mpa->private_data_size = htons(ep->plen);
+	mpa->revision = mpa_rev;
+
+	if (ep->plen)
+		memcpy(mpa->private_data, ep->mpa_pkt + sizeof(*mpa), ep->plen);
+
+	/*
+	 * Reference the mpa skb.  This ensures the data area
+	 * will remain in memory until the hw acks the tx.
+	 * Function fw4_ack() will deref it.
+	 */
+	skb_get(skb);
+	t4_set_arp_err_handler(skb, NULL, arp_failure_discard);
+	BUG_ON(ep->mpa_skb);
+	ep->mpa_skb = skb;
+	c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t);
+	start_ep_timer(ep);
+	state_set(&ep->com, MPA_REQ_SENT);
+	ep->mpa_attr.initiator = 1;
+	return;
+}
+
+static int send_mpa_reject(struct c4iw_ep *ep, const void *pdata, u8 plen)
+{
+	int mpalen, wrlen;
+	struct fw_ofld_tx_data_wr *req;
+	struct mpa_message *mpa;
+	struct sk_buff *skb;
+
+	PDBG("%s ep %p tid %u pd_len %d\n", __func__, ep, ep->hwtid, ep->plen);
+
+	mpalen = sizeof(*mpa) + plen;
+	wrlen = roundup(mpalen + sizeof *req, 16);
+
+	skb = get_skb(NULL, wrlen, GFP_KERNEL);
+	if (!skb) {
+		printk(KERN_ERR MOD "%s - cannot alloc skb!\n", __func__);
+		return -ENOMEM;
+	}
+	set_wr_txq(skb, CPL_PRIORITY_DATA, ep->txq_idx);
+
+	req = (struct fw_ofld_tx_data_wr *)skb_put(skb, wrlen);
+	memset(req, 0, wrlen);
+	req->op_to_immdlen = cpu_to_be32(
+		FW_WR_OP(FW_OFLD_TX_DATA_WR) |
+		FW_WR_COMPL(1) |
+		FW_WR_IMMDLEN(mpalen));
+	req->flowid_len16 = cpu_to_be32(
+		FW_WR_FLOWID(ep->hwtid) |
+		FW_WR_LEN16(wrlen >> 4));
+	req->plen = cpu_to_be32(mpalen);
+	req->tunnel_to_proxy = cpu_to_be32(
+		FW_OFLD_TX_DATA_WR_FLUSH(1) |
+		FW_OFLD_TX_DATA_WR_SHOVE(1));
+
+	mpa = (struct mpa_message *)(req + 1);
+	memset(mpa, 0, sizeof(*mpa));
+	memcpy(mpa->key, MPA_KEY_REP, sizeof(mpa->key));
+	mpa->flags = MPA_REJECT;
+	mpa->revision = mpa_rev;
+	mpa->private_data_size = htons(plen);
+	if (plen)
+		memcpy(mpa->private_data, pdata, plen);
+
+	/*
+	 * Reference the mpa skb again.  This ensures the data area
+	 * will remain in memory until the hw acks the tx.
+	 * Function fw4_ack() will deref it.
+	 */
+	skb_get(skb);
+	set_wr_txq(skb, CPL_PRIORITY_DATA, ep->txq_idx);
+	t4_set_arp_err_handler(skb, NULL, arp_failure_discard);
+	BUG_ON(ep->mpa_skb);
+	ep->mpa_skb = skb;
+	return c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t);
+}
+
+static int send_mpa_reply(struct c4iw_ep *ep, const void *pdata, u8 plen)
+{
+	int mpalen, wrlen;
+	struct fw_ofld_tx_data_wr *req;
+	struct mpa_message *mpa;
+	struct sk_buff *skb;
+
+	PDBG("%s ep %p tid %u pd_len %d\n", __func__, ep, ep->hwtid, ep->plen);
+
+	mpalen = sizeof(*mpa) + plen;
+	wrlen = roundup(mpalen + sizeof *req, 16);
+
+	skb = get_skb(NULL, wrlen, GFP_KERNEL);
+	if (!skb) {
+		printk(KERN_ERR MOD "%s - cannot alloc skb!\n", __func__);
+		return -ENOMEM;
+	}
+	set_wr_txq(skb, CPL_PRIORITY_DATA, ep->txq_idx);
+
+	req = (struct fw_ofld_tx_data_wr *) skb_put(skb, wrlen);
+	memset(req, 0, wrlen);
+	req->op_to_immdlen = cpu_to_be32(
+		FW_WR_OP(FW_OFLD_TX_DATA_WR) |
+		FW_WR_COMPL(1) |
+		FW_WR_IMMDLEN(mpalen));
+	req->flowid_len16 = cpu_to_be32(
+		FW_WR_FLOWID(ep->hwtid) |
+		FW_WR_LEN16(wrlen >> 4));
+	req->plen = cpu_to_be32(mpalen);
+	req->tunnel_to_proxy = cpu_to_be32(
+		FW_OFLD_TX_DATA_WR_FLUSH(1) |
+		FW_OFLD_TX_DATA_WR_SHOVE(1));
+
+	mpa = (struct mpa_message *)(req + 1);
+	memset(mpa, 0, sizeof(*mpa));
+	memcpy(mpa->key, MPA_KEY_REP, sizeof(mpa->key));
+	mpa->flags = (ep->mpa_attr.crc_enabled ? MPA_CRC : 0) |
+		     (markers_enabled ? MPA_MARKERS : 0);
+	mpa->revision = mpa_rev;
+	mpa->private_data_size = htons(plen);
+	if (plen)
+		memcpy(mpa->private_data, pdata, plen);
+
+	/*
+	 * Reference the mpa skb.  This ensures the data area
+	 * will remain in memory until the hw acks the tx.
+	 * Function fw4_ack() will deref it.
+	 */
+	skb_get(skb);
+	t4_set_arp_err_handler(skb, NULL, arp_failure_discard);
+	ep->mpa_skb = skb;
+	state_set(&ep->com, MPA_REP_SENT);
+	return c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t);
+}
+
+static int act_establish(struct c4iw_dev *dev, struct sk_buff *skb)
+{
+	struct c4iw_ep *ep;
+	struct cpl_act_establish *req = cplhdr(skb);
+	unsigned int tid = GET_TID(req);
+	unsigned int atid = GET_TID_TID(ntohl(req->tos_atid));
+	struct tid_info *t = dev->rdev.lldi.tids;
+
+	ep = lookup_atid(t, atid);
+
+	PDBG("%s ep %p tid %u snd_isn %u rcv_isn %u\n", __func__, ep, tid,
+	     be32_to_cpu(req->snd_isn), be32_to_cpu(req->rcv_isn));
+
+	dst_confirm(ep->dst);
+
+	/* setup the hwtid for this connection */
+	ep->hwtid = tid;
+	cxgb4_insert_tid(t, ep, tid);
+
+	ep->snd_seq = be32_to_cpu(req->snd_isn);
+	ep->rcv_seq = be32_to_cpu(req->rcv_isn);
+
+	set_emss(ep, ntohs(req->tcp_opt));
+
+	/* dealloc the atid */
+	cxgb4_free_atid(t, atid);
+
+	/* start MPA negotiation */
+	send_flowc(ep, NULL);
+	send_mpa_req(ep, skb);
+
+	return 0;
+}
+
+static void close_complete_upcall(struct c4iw_ep *ep)
+{
+	struct iw_cm_event event;
+
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+	memset(&event, 0, sizeof(event));
+	event.event = IW_CM_EVENT_CLOSE;
+	if (ep->com.cm_id) {
+		PDBG("close complete delivered ep %p cm_id %p tid %u\n",
+		     ep, ep->com.cm_id, ep->hwtid);
+		ep->com.cm_id->event_handler(ep->com.cm_id, &event);
+		ep->com.cm_id->rem_ref(ep->com.cm_id);
+		ep->com.cm_id = NULL;
+		ep->com.qp = NULL;
+	}
+}
+
+static int abort_connection(struct c4iw_ep *ep, struct sk_buff *skb, gfp_t gfp)
+{
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+	close_complete_upcall(ep);
+	state_set(&ep->com, ABORTING);
+	return send_abort(ep, skb, gfp);
+}
+
+static void peer_close_upcall(struct c4iw_ep *ep)
+{
+	struct iw_cm_event event;
+
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+	memset(&event, 0, sizeof(event));
+	event.event = IW_CM_EVENT_DISCONNECT;
+	if (ep->com.cm_id) {
+		PDBG("peer close delivered ep %p cm_id %p tid %u\n",
+		     ep, ep->com.cm_id, ep->hwtid);
+		ep->com.cm_id->event_handler(ep->com.cm_id, &event);
+	}
+}
+
+static void peer_abort_upcall(struct c4iw_ep *ep)
+{
+	struct iw_cm_event event;
+
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+	memset(&event, 0, sizeof(event));
+	event.event = IW_CM_EVENT_CLOSE;
+	event.status = -ECONNRESET;
+	if (ep->com.cm_id) {
+		PDBG("abort delivered ep %p cm_id %p tid %u\n", ep,
+		     ep->com.cm_id, ep->hwtid);
+		ep->com.cm_id->event_handler(ep->com.cm_id, &event);
+		ep->com.cm_id->rem_ref(ep->com.cm_id);
+		ep->com.cm_id = NULL;
+		ep->com.qp = NULL;
+	}
+}
+
+static void connect_reply_upcall(struct c4iw_ep *ep, int status)
+{
+	struct iw_cm_event event;
+
+	PDBG("%s ep %p tid %u status %d\n", __func__, ep, ep->hwtid, status);
+	memset(&event, 0, sizeof(event));
+	event.event = IW_CM_EVENT_CONNECT_REPLY;
+	event.status = status;
+	event.local_addr = ep->com.local_addr;
+	event.remote_addr = ep->com.remote_addr;
+
+	if ((status == 0) || (status == -ECONNREFUSED)) {
+		event.private_data_len = ep->plen;
+		event.private_data = ep->mpa_pkt + sizeof(struct mpa_message);
+	}
+	if (ep->com.cm_id) {
+		PDBG("%s ep %p tid %u status %d\n", __func__, ep,
+		     ep->hwtid, status);
+		ep->com.cm_id->event_handler(ep->com.cm_id, &event);
+	}
+	if (status < 0) {
+		ep->com.cm_id->rem_ref(ep->com.cm_id);
+		ep->com.cm_id = NULL;
+		ep->com.qp = NULL;
+	}
+}
+
+static void connect_request_upcall(struct c4iw_ep *ep)
+{
+	struct iw_cm_event event;
+
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+	memset(&event, 0, sizeof(event));
+	event.event = IW_CM_EVENT_CONNECT_REQUEST;
+	event.local_addr = ep->com.local_addr;
+	event.remote_addr = ep->com.remote_addr;
+	event.private_data_len = ep->plen;
+	event.private_data = ep->mpa_pkt + sizeof(struct mpa_message);
+	event.provider_data = ep;
+	if (state_read(&ep->parent_ep->com) != DEAD) {
+		c4iw_get_ep(&ep->com);
+		ep->parent_ep->com.cm_id->event_handler(
+						ep->parent_ep->com.cm_id,
+						&event);
+	}
+	c4iw_put_ep(&ep->parent_ep->com);
+	ep->parent_ep = NULL;
+}
+
+static void established_upcall(struct c4iw_ep *ep)
+{
+	struct iw_cm_event event;
+
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+	memset(&event, 0, sizeof(event));
+	event.event = IW_CM_EVENT_ESTABLISHED;
+	if (ep->com.cm_id) {
+		PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+		ep->com.cm_id->event_handler(ep->com.cm_id, &event);
+	}
+}
+
+static int update_rx_credits(struct c4iw_ep *ep, u32 credits)
+{
+	struct cpl_rx_data_ack *req;
+	struct sk_buff *skb;
+	int wrlen = roundup(sizeof *req, 16);
+
+	PDBG("%s ep %p tid %u credits %u\n", __func__, ep, ep->hwtid, credits);
+	skb = get_skb(NULL, wrlen, GFP_KERNEL);
+	if (!skb) {
+		printk(KERN_ERR MOD "update_rx_credits - cannot alloc skb!\n");
+		return 0;
+	}
+
+	req = (struct cpl_rx_data_ack *) skb_put(skb, wrlen);
+	memset(req, 0, wrlen);
+	INIT_TP_WR(req, ep->hwtid);
+	OPCODE_TID(req) = cpu_to_be32(MK_OPCODE_TID(CPL_RX_DATA_ACK,
+						    ep->hwtid));
+	req->credit_dack = cpu_to_be32(credits);
+	set_wr_txq(skb, CPL_PRIORITY_ACK, ep->txq_idx);
+	c4iw_ofld_send(&ep->com.dev->rdev, skb);
+	return credits;
+}
+
+static void process_mpa_reply(struct c4iw_ep *ep, struct sk_buff *skb)
+{
+	struct mpa_message *mpa;
+	u16 plen;
+	struct c4iw_qp_attributes attrs;
+	enum c4iw_qp_attr_mask mask;
+	int err;
+
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+
+	/*
+	 * Stop mpa timer.  If it expired, then the state has
+	 * changed and we bail since ep_timeout already aborted
+	 * the connection.
+	 */
+	stop_ep_timer(ep);
+	if (state_read(&ep->com) != MPA_REQ_SENT)
+		return;
+
+	/*
+	 * If we get more than the supported amount of private data
+	 * then we must fail this connection.
+	 */
+	if (ep->mpa_pkt_len + skb->len > sizeof(ep->mpa_pkt)) {
+		err = -EINVAL;
+		goto err;
+	}
+
+	/*
+	 * copy the new data into our accumulation buffer.
+	 */
+	skb_copy_from_linear_data(skb, &(ep->mpa_pkt[ep->mpa_pkt_len]),
+				  skb->len);
+	ep->mpa_pkt_len += skb->len;
+
+	/*
+	 * if we don't even have the mpa message, then bail.
+	 */
+	if (ep->mpa_pkt_len < sizeof(*mpa))
+		return;
+	mpa = (struct mpa_message *) ep->mpa_pkt;
+
+	/* Validate MPA header. */
+	if (mpa->revision != mpa_rev) {
+		err = -EPROTO;
+		goto err;
+	}
+	if (memcmp(mpa->key, MPA_KEY_REP, sizeof(mpa->key))) {
+		err = -EPROTO;
+		goto err;
+	}
+
+	plen = ntohs(mpa->private_data_size);
+
+	/*
+	 * Fail if there's too much private data.
+	 */
+	if (plen > MPA_MAX_PRIVATE_DATA) {
+		err = -EPROTO;
+		goto err;
+	}
+
+	/*
+	 * If plen does not account for pkt size
+	 */
+	if (ep->mpa_pkt_len > (sizeof(*mpa) + plen)) {
+		err = -EPROTO;
+		goto err;
+	}
+
+	ep->plen = (u8) plen;
+
+	/*
+	 * If we don't have all the pdata yet, then bail.
+	 * We'll continue process when more data arrives.
+	 */
+	if (ep->mpa_pkt_len < (sizeof(*mpa) + plen))
+		return;
+
+	if (mpa->flags & MPA_REJECT) {
+		err = -ECONNREFUSED;
+		goto err;
+	}
+
+	/*
+	 * If we get here we have accumulated the entire mpa
+	 * start reply message including private data. And
+	 * the MPA header is valid.
+	 */
+	state_set(&ep->com, FPDU_MODE);
+	ep->mpa_attr.crc_enabled = (mpa->flags & MPA_CRC) | crc_enabled ? 1 : 0;
+	ep->mpa_attr.recv_marker_enabled = markers_enabled;
+	ep->mpa_attr.xmit_marker_enabled = mpa->flags & MPA_MARKERS ? 1 : 0;
+	ep->mpa_attr.version = mpa_rev;
+	PDBG("%s - crc_enabled=%d, recv_marker_enabled=%d, "
+	     "xmit_marker_enabled=%d, version=%d\n", __func__,
+	     ep->mpa_attr.crc_enabled, ep->mpa_attr.recv_marker_enabled,
+	     ep->mpa_attr.xmit_marker_enabled, ep->mpa_attr.version);
+
+	attrs.mpa_attr = ep->mpa_attr;
+	attrs.max_ird = ep->ird;
+	attrs.max_ord = ep->ord;
+	attrs.llp_stream_handle = ep;
+	attrs.next_state = C4IW_QP_STATE_RTS;
+
+	mask = C4IW_QP_ATTR_NEXT_STATE |
+	    C4IW_QP_ATTR_LLP_STREAM_HANDLE | C4IW_QP_ATTR_MPA_ATTR |
+	    C4IW_QP_ATTR_MAX_IRD | C4IW_QP_ATTR_MAX_ORD;
+
+	/* bind QP and TID with INIT_WR */
+	err = c4iw_modify_qp(ep->com.qp->rhp,
+			     ep->com.qp, mask, &attrs, 1);
+	if (err)
+		goto err;
+
+	if (peer2peer && t4_rqes_posted(&ep->com.qp->wq) == 0)
+		c4iw_post_zb_read(ep->com.qp);
+
+	goto out;
+err:
+	abort_connection(ep, skb, GFP_KERNEL);
+out:
+	connect_reply_upcall(ep, err);
+	return;
+}
+
+static void process_mpa_request(struct c4iw_ep *ep, struct sk_buff *skb)
+{
+	struct mpa_message *mpa;
+	u16 plen;
+
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+
+	if (state_read(&ep->com) != MPA_REQ_WAIT)
+		return;
+
+	/*
+	 * If we get more than the supported amount of private data
+	 * then we must fail this connection.
+	 */
+	if (ep->mpa_pkt_len + skb->len > sizeof(ep->mpa_pkt)) {
+		stop_ep_timer(ep);
+		abort_connection(ep, skb, GFP_KERNEL);
+		return;
+	}
+
+	PDBG("%s enter (%s line %u)\n", __func__, __FILE__, __LINE__);
+
+	/*
+	 * Copy the new data into our accumulation buffer.
+	 */
+	skb_copy_from_linear_data(skb, &(ep->mpa_pkt[ep->mpa_pkt_len]),
+				  skb->len);
+	ep->mpa_pkt_len += skb->len;
+
+	/*
+	 * If we don't even have the mpa message, then bail.
+	 * We'll continue process when more data arrives.
+	 */
+	if (ep->mpa_pkt_len < sizeof(*mpa))
+		return;
+
+	PDBG("%s enter (%s line %u)\n", __func__, __FILE__, __LINE__);
+	stop_ep_timer(ep);
+	mpa = (struct mpa_message *) ep->mpa_pkt;
+
+	/*
+	 * Validate MPA Header.
+	 */
+	if (mpa->revision != mpa_rev) {
+		abort_connection(ep, skb, GFP_KERNEL);
+		return;
+	}
+
+	if (memcmp(mpa->key, MPA_KEY_REQ, sizeof(mpa->key))) {
+		abort_connection(ep, skb, GFP_KERNEL);
+		return;
+	}
+
+	plen = ntohs(mpa->private_data_size);
+
+	/*
+	 * Fail if there's too much private data.
+	 */
+	if (plen > MPA_MAX_PRIVATE_DATA) {
+		abort_connection(ep, skb, GFP_KERNEL);
+		return;
+	}
+
+	/*
+	 * If plen does not account for pkt size
+	 */
+	if (ep->mpa_pkt_len > (sizeof(*mpa) + plen)) {
+		abort_connection(ep, skb, GFP_KERNEL);
+		return;
+	}
+	ep->plen = (u8) plen;
+
+	/*
+	 * If we don't have all the pdata yet, then bail.
+	 */
+	if (ep->mpa_pkt_len < (sizeof(*mpa) + plen))
+		return;
+
+	/*
+	 * If we get here we have accumulated the entire mpa
+	 * start reply message including private data.
+	 */
+	ep->mpa_attr.initiator = 0;
+	ep->mpa_attr.crc_enabled = (mpa->flags & MPA_CRC) | crc_enabled ? 1 : 0;
+	ep->mpa_attr.recv_marker_enabled = markers_enabled;
+	ep->mpa_attr.xmit_marker_enabled = mpa->flags & MPA_MARKERS ? 1 : 0;
+	ep->mpa_attr.version = mpa_rev;
+	PDBG("%s - crc_enabled=%d, recv_marker_enabled=%d, "
+	     "xmit_marker_enabled=%d, version=%d\n", __func__,
+	     ep->mpa_attr.crc_enabled, ep->mpa_attr.recv_marker_enabled,
+	     ep->mpa_attr.xmit_marker_enabled, ep->mpa_attr.version);
+
+	state_set(&ep->com, MPA_REQ_RCVD);
+
+	/* drive upcall */
+	connect_request_upcall(ep);
+	return;
+}
+
+static int rx_data(struct c4iw_dev *dev, struct sk_buff *skb)
+{
+	struct c4iw_ep *ep;
+	struct cpl_rx_data *hdr = cplhdr(skb);
+	unsigned int dlen = ntohs(hdr->len);
+	unsigned int tid = GET_TID(hdr);
+	struct tid_info *t = dev->rdev.lldi.tids;
+
+	ep = lookup_tid(t, tid);
+	PDBG("%s ep %p tid %u dlen %u\n", __func__, ep, ep->hwtid, dlen);
+	skb_pull(skb, sizeof(*hdr));
+	skb_trim(skb, dlen);
+
+	ep->rcv_seq += dlen;
+	BUG_ON(ep->rcv_seq != (ntohl(hdr->seq) + dlen));
+
+	/* update RX credits */
+	update_rx_credits(ep, dlen);
+
+	switch (state_read(&ep->com)) {
+	case MPA_REQ_SENT:
+		process_mpa_reply(ep, skb);
+		break;
+	case MPA_REQ_WAIT:
+		process_mpa_request(ep, skb);
+		break;
+	case MPA_REP_SENT:
+		break;
+	default:
+		printk(KERN_ERR MOD "%s Unexpected streaming data."
+		       " ep %p state %d tid %u\n",
+		       __func__, ep, state_read(&ep->com), ep->hwtid);
+
+		/*
+		 * The ep will timeout and inform the ULP of the failure.
+		 * See ep_timeout().
+		 */
+		break;
+	}
+	return 0;
+}
+
+static int abort_rpl(struct c4iw_dev *dev, struct sk_buff *skb)
+{
+	struct c4iw_ep *ep;
+	struct cpl_abort_rpl_rss *rpl = cplhdr(skb);
+	unsigned long flags;
+	int release = 0;
+	unsigned int tid = GET_TID(rpl);
+	struct tid_info *t = dev->rdev.lldi.tids;
+
+	ep = lookup_tid(t, tid);
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+	BUG_ON(!ep);
+	spin_lock_irqsave(&ep->com.lock, flags);
+	switch (ep->com.state) {
+	case ABORTING:
+		__state_set(&ep->com, DEAD);
+		release = 1;
+		break;
+	default:
+		printk(KERN_ERR "%s ep %p state %d\n",
+		     __func__, ep, ep->com.state);
+		break;
+	}
+	spin_unlock_irqrestore(&ep->com.lock, flags);
+
+	if (release)
+		release_ep_resources(ep);
+	return 0;
+}
+
+/*
+ * Return whether a failed active open has allocated a TID
+ */
+static inline int act_open_has_tid(int status)
+{
+	return status != CPL_ERR_TCAM_FULL && status != CPL_ERR_CONN_EXIST &&
+	       status != CPL_ERR_ARP_MISS;
+}
+
+static int act_open_rpl(struct c4iw_dev *dev, struct sk_buff *skb)
+{
+	struct c4iw_ep *ep;
+	struct cpl_act_open_rpl *rpl = cplhdr(skb);
+	unsigned int atid = GET_TID_TID(GET_AOPEN_ATID(
+					ntohl(rpl->atid_status)));
+	struct tid_info *t = dev->rdev.lldi.tids;
+	int status = GET_AOPEN_STATUS(ntohl(rpl->atid_status));
+
+	ep = lookup_atid(t, atid);
+
+	PDBG("%s ep %p atid %u status %u errno %d\n", __func__, ep, atid,
+	     status, status2errno(status));
+
+	if (status == CPL_ERR_RTX_NEG_ADVICE) {
+		printk(KERN_WARNING MOD "Connection problems for atid %u\n",
+			atid);
+		return 0;
+	}
+
+	connect_reply_upcall(ep, status2errno(status));
+	state_set(&ep->com, DEAD);
+
+	if (status && act_open_has_tid(status))
+		cxgb4_remove_tid(ep->com.dev->rdev.lldi.tids, 0, GET_TID(rpl));
+
+	cxgb4_free_atid(t, atid);
+	dst_release(ep->dst);
+	cxgb4_l2t_release(ep->l2t);
+	c4iw_put_ep(&ep->com);
+
+	return 0;
+}
+
+static int pass_open_rpl(struct c4iw_dev *dev, struct sk_buff *skb)
+{
+	struct cpl_pass_open_rpl *rpl = cplhdr(skb);
+	struct tid_info *t = dev->rdev.lldi.tids;
+	unsigned int stid = GET_TID(rpl);
+	struct c4iw_listen_ep *ep = lookup_stid(t, stid);
+
+	if (!ep) {
+		printk(KERN_ERR MOD "stid %d lookup failure!\n", stid);
+		return 0;
+	}
+	PDBG("%s ep %p status %d error %d\n", __func__, ep,
+	     rpl->status, status2errno(rpl->status));
+	ep->com.rpl_err = status2errno(rpl->status);
+	ep->com.rpl_done = 1;
+	wake_up(&ep->com.waitq);
+
+	return 0;
+}
+
+static int listen_stop(struct c4iw_listen_ep *ep)
+{
+	struct sk_buff *skb;
+	struct cpl_close_listsvr_req *req;
+
+	PDBG("%s ep %p\n", __func__, ep);
+	skb = get_skb(NULL, sizeof(*req), GFP_KERNEL);
+	if (!skb) {
+		printk(KERN_ERR MOD "%s - failed to alloc skb\n", __func__);
+		return -ENOMEM;
+	}
+	req = (struct cpl_close_listsvr_req *) skb_put(skb, sizeof(*req));
+	INIT_TP_WR(req, 0);
+	OPCODE_TID(req) = cpu_to_be32(MK_OPCODE_TID(CPL_CLOSE_LISTSRV_REQ,
+						    ep->stid));
+	req->reply_ctrl = cpu_to_be16(
+			  QUEUENO(ep->com.dev->rdev.lldi.rxq_ids[0]));
+	set_wr_txq(skb, CPL_PRIORITY_SETUP, 0);
+	return c4iw_ofld_send(&ep->com.dev->rdev, skb);
+}
+
+static int close_listsrv_rpl(struct c4iw_dev *dev, struct sk_buff *skb)
+{
+	struct cpl_close_listsvr_rpl *rpl = cplhdr(skb);
+	struct tid_info *t = dev->rdev.lldi.tids;
+	unsigned int stid = GET_TID(rpl);
+	struct c4iw_listen_ep *ep = lookup_stid(t, stid);
+
+	PDBG("%s ep %p\n", __func__, ep);
+	ep->com.rpl_err = status2errno(rpl->status);
+	ep->com.rpl_done = 1;
+	wake_up(&ep->com.waitq);
+	return 0;
+}
+
+static void accept_cr(struct c4iw_ep *ep, __be32 peer_ip, struct sk_buff *skb,
+		      struct cpl_pass_accept_req *req)
+{
+	struct cpl_pass_accept_rpl *rpl;
+	unsigned int mtu_idx;
+	u64 opt0;
+	u32 opt2;
+	int wscale;
+
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+	BUG_ON(skb_cloned(skb));
+	skb_trim(skb, sizeof(*rpl));
+	skb_get(skb);
+	cxgb4_best_mtu(ep->com.dev->rdev.lldi.mtus, ep->mtu, &mtu_idx);
+	wscale = compute_wscale(rcv_win);
+	opt0 = KEEP_ALIVE(1) |
+	       WND_SCALE(wscale) |
+	       MSS_IDX(mtu_idx) |
+	       L2T_IDX(ep->l2t->idx) |
+	       TX_CHAN(ep->tx_chan) |
+	       SMAC_SEL(ep->smac_idx) |
+	       DSCP(ep->tos) |
+	       RCV_BUFSIZ(rcv_win>>10);
+	opt2 = RX_CHANNEL(0) |
+	       RSS_QUEUE_VALID | RSS_QUEUE(ep->rss_qid);
+
+	if (enable_tcp_timestamps && req->tcpopt.tstamp)
+		opt2 |= TSTAMPS_EN(1);
+	if (enable_tcp_sack && req->tcpopt.sack)
+		opt2 |= SACK_EN(1);
+	if (wscale && enable_tcp_window_scaling)
+		opt2 |= WND_SCALE_EN(1);
+
+	rpl = cplhdr(skb);
+	INIT_TP_WR(rpl, ep->hwtid);
+	OPCODE_TID(rpl) = cpu_to_be32(MK_OPCODE_TID(CPL_PASS_ACCEPT_RPL,
+				      ep->hwtid));
+	rpl->opt0 = cpu_to_be64(opt0);
+	rpl->opt2 = cpu_to_be32(opt2);
+	set_wr_txq(skb, CPL_PRIORITY_SETUP, ep->txq_idx);
+	c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t);
+
+	return;
+}
+
+static void reject_cr(struct c4iw_dev *dev, u32 hwtid, __be32 peer_ip,
+		      struct sk_buff *skb)
+{
+	PDBG("%s c4iw_dev %p tid %u peer_ip %x\n", __func__, dev, hwtid,
+	     peer_ip);
+	BUG_ON(skb_cloned(skb));
+	skb_trim(skb, sizeof(struct cpl_tid_release));
+	skb_get(skb);
+	release_tid(&dev->rdev, hwtid, skb);
+	return;
+}
+
+static void get_4tuple(struct cpl_pass_accept_req *req,
+		       __be32 *local_ip, __be32 *peer_ip,
+		       __be16 *local_port, __be16 *peer_port)
+{
+	int eth_len = G_ETH_HDR_LEN(be32_to_cpu(req->hdr_len));
+	int ip_len = G_IP_HDR_LEN(be32_to_cpu(req->hdr_len));
+	struct iphdr *ip = (struct iphdr *)((u8 *)(req + 1) + eth_len);
+	struct tcphdr *tcp = (struct tcphdr *)
+			     ((u8 *)(req + 1) + eth_len + ip_len);
+
+	PDBG("%s saddr 0x%x daddr 0x%x sport %u dport %u\n", __func__,
+	     ntohl(ip->saddr), ntohl(ip->daddr), ntohs(tcp->source),
+	     ntohs(tcp->dest));
+
+	*peer_ip = ip->saddr;
+	*local_ip = ip->daddr;
+	*peer_port = tcp->source;
+	*local_port = tcp->dest;
+
+	return;
+}
+
+static int pass_accept_req(struct c4iw_dev *dev, struct sk_buff *skb)
+{
+	struct c4iw_ep *child_ep, *parent_ep;
+	struct cpl_pass_accept_req *req = cplhdr(skb);
+	unsigned int stid = GET_POPEN_TID(ntohl(req->tos_stid));
+	struct tid_info *t = dev->rdev.lldi.tids;
+	unsigned int hwtid = GET_TID(req);
+	struct dst_entry *dst;
+	struct l2t_entry *l2t;
+	struct rtable *rt;
+	__be32 local_ip, peer_ip;
+	__be16 local_port, peer_port;
+	struct net_device *pdev;
+	u32 tx_chan, smac_idx;
+	u16 rss_qid;
+	u32 mtu;
+	int step;
+	int txq_idx;
+
+	parent_ep = lookup_stid(t, stid);
+	PDBG("%s parent ep %p tid %u\n", __func__, parent_ep, hwtid);
+
+	get_4tuple(req, &local_ip, &peer_ip, &local_port, &peer_port);
+
+	if (state_read(&parent_ep->com) != LISTEN) {
+		printk(KERN_ERR "%s - listening ep not in LISTEN\n",
+		       __func__);
+		goto reject;
+	}
+
+	/* Find output route */
+	rt = find_route(dev, local_ip, peer_ip, local_port, peer_port,
+			GET_POPEN_TOS(ntohl(req->tos_stid)));
+	if (!rt) {
+		printk(KERN_ERR MOD "%s - failed to find dst entry!\n",
+		       __func__);
+		goto reject;
+	}
+	dst = &rt->u.dst;
+	if (dst->neighbour->dev->flags & IFF_LOOPBACK) {
+		pdev = ip_dev_find(&init_net, peer_ip);
+		BUG_ON(!pdev);
+		l2t = cxgb4_l2t_get(dev->rdev.lldi.l2t, dst->neighbour,
+				    pdev, 0);
+		mtu = pdev->mtu;
+		tx_chan = cxgb4_port_chan(pdev);
+		smac_idx = tx_chan << 1;
+		step = dev->rdev.lldi.ntxq / dev->rdev.lldi.nchan;
+		txq_idx = cxgb4_port_idx(pdev) * step;
+		step = dev->rdev.lldi.nrxq / dev->rdev.lldi.nchan;
+		rss_qid = dev->rdev.lldi.rxq_ids[cxgb4_port_idx(pdev) * step];
+		dev_put(pdev);
+	} else {
+		l2t = cxgb4_l2t_get(dev->rdev.lldi.l2t, dst->neighbour,
+					dst->neighbour->dev, 0);
+		mtu = dst_mtu(dst);
+		tx_chan = cxgb4_port_chan(dst->neighbour->dev);
+		smac_idx = tx_chan << 1;
+		step = dev->rdev.lldi.ntxq / dev->rdev.lldi.nchan;
+		txq_idx = cxgb4_port_idx(dst->neighbour->dev) * step;
+		step = dev->rdev.lldi.nrxq / dev->rdev.lldi.nchan;
+		rss_qid = dev->rdev.lldi.rxq_ids[
+			  cxgb4_port_idx(dst->neighbour->dev) * step];
+	}
+	if (!l2t) {
+		printk(KERN_ERR MOD "%s - failed to allocate l2t entry!\n",
+		       __func__);
+		dst_release(dst);
+		goto reject;
+	}
+
+	child_ep = alloc_ep(sizeof(*child_ep), GFP_KERNEL);
+	if (!child_ep) {
+		printk(KERN_ERR MOD "%s - failed to allocate ep entry!\n",
+		       __func__);
+		cxgb4_l2t_release(l2t);
+		dst_release(dst);
+		goto reject;
+	}
+	state_set(&child_ep->com, CONNECTING);
+	child_ep->com.dev = dev;
+	child_ep->com.cm_id = NULL;
+	child_ep->com.local_addr.sin_family = PF_INET;
+	child_ep->com.local_addr.sin_port = local_port;
+	child_ep->com.local_addr.sin_addr.s_addr = local_ip;
+	child_ep->com.remote_addr.sin_family = PF_INET;
+	child_ep->com.remote_addr.sin_port = peer_port;
+	child_ep->com.remote_addr.sin_addr.s_addr = peer_ip;
+	c4iw_get_ep(&parent_ep->com);
+	child_ep->parent_ep = parent_ep;
+	child_ep->tos = GET_POPEN_TOS(ntohl(req->tos_stid));
+	child_ep->l2t = l2t;
+	child_ep->dst = dst;
+	child_ep->hwtid = hwtid;
+	child_ep->tx_chan = tx_chan;
+	child_ep->smac_idx = smac_idx;
+	child_ep->rss_qid = rss_qid;
+	child_ep->mtu = mtu;
+	child_ep->txq_idx = txq_idx;
+
+	PDBG("%s tx_chan %u smac_idx %u rss_qid %u\n", __func__,
+	     tx_chan, smac_idx, rss_qid);
+
+	init_timer(&child_ep->timer);
+	cxgb4_insert_tid(t, child_ep, hwtid);
+	accept_cr(child_ep, peer_ip, skb, req);
+	goto out;
+reject:
+	reject_cr(dev, hwtid, peer_ip, skb);
+out:
+	return 0;
+}
+
+static int pass_establish(struct c4iw_dev *dev, struct sk_buff *skb)
+{
+	struct c4iw_ep *ep;
+	struct cpl_pass_establish *req = cplhdr(skb);
+	struct tid_info *t = dev->rdev.lldi.tids;
+	unsigned int tid = GET_TID(req);
+
+	ep = lookup_tid(t, tid);
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+	ep->snd_seq = be32_to_cpu(req->snd_isn);
+	ep->rcv_seq = be32_to_cpu(req->rcv_isn);
+
+	set_emss(ep, ntohs(req->tcp_opt));
+
+	dst_confirm(ep->dst);
+	state_set(&ep->com, MPA_REQ_WAIT);
+	start_ep_timer(ep);
+	send_flowc(ep, skb);
+
+	return 0;
+}
+
+static int peer_close(struct c4iw_dev *dev, struct sk_buff *skb)
+{
+	struct cpl_peer_close *hdr = cplhdr(skb);
+	struct c4iw_ep *ep;
+	struct c4iw_qp_attributes attrs;
+	unsigned long flags;
+	int disconnect = 1;
+	int release = 0;
+	int closing = 0;
+	struct tid_info *t = dev->rdev.lldi.tids;
+	unsigned int tid = GET_TID(hdr);
+	int start_timer = 0;
+	int stop_timer = 0;
+
+	ep = lookup_tid(t, tid);
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+	dst_confirm(ep->dst);
+
+	spin_lock_irqsave(&ep->com.lock, flags);
+	switch (ep->com.state) {
+	case MPA_REQ_WAIT:
+		__state_set(&ep->com, CLOSING);
+		break;
+	case MPA_REQ_SENT:
+		__state_set(&ep->com, CLOSING);
+		connect_reply_upcall(ep, -ECONNRESET);
+		break;
+	case MPA_REQ_RCVD:
+
+		/*
+		 * We're gonna mark this puppy DEAD, but keep
+		 * the reference on it until the ULP accepts or
+		 * rejects the CR. Also wake up anyone waiting
+		 * in rdma connection migration (see c4iw_accept_cr()).
+		 */
+		__state_set(&ep->com, CLOSING);
+		ep->com.rpl_done = 1;
+		ep->com.rpl_err = -ECONNRESET;
+		PDBG("waking up ep %p tid %u\n", ep, ep->hwtid);
+		wake_up(&ep->com.waitq);
+		break;
+	case MPA_REP_SENT:
+		__state_set(&ep->com, CLOSING);
+		ep->com.rpl_done = 1;
+		ep->com.rpl_err = -ECONNRESET;
+		PDBG("waking up ep %p tid %u\n", ep, ep->hwtid);
+		wake_up(&ep->com.waitq);
+		break;
+	case FPDU_MODE:
+		start_timer = 1;
+		__state_set(&ep->com, CLOSING);
+		closing = 1;
+		peer_close_upcall(ep);
+		break;
+	case ABORTING:
+		disconnect = 0;
+		break;
+	case CLOSING:
+		__state_set(&ep->com, MORIBUND);
+		disconnect = 0;
+		break;
+	case MORIBUND:
+		stop_timer = 1;
+		if (ep->com.cm_id && ep->com.qp) {
+			attrs.next_state = C4IW_QP_STATE_IDLE;
+			c4iw_modify_qp(ep->com.qp->rhp, ep->com.qp,
+				       C4IW_QP_ATTR_NEXT_STATE, &attrs, 1);
+		}
+		close_complete_upcall(ep);
+		__state_set(&ep->com, DEAD);
+		release = 1;
+		disconnect = 0;
+		break;
+	case DEAD:
+		disconnect = 0;
+		break;
+	default:
+		BUG_ON(1);
+	}
+	spin_unlock_irqrestore(&ep->com.lock, flags);
+	if (closing) {
+		attrs.next_state = C4IW_QP_STATE_CLOSING;
+		c4iw_modify_qp(ep->com.qp->rhp, ep->com.qp,
+			       C4IW_QP_ATTR_NEXT_STATE, &attrs, 1);
+	}
+	if (start_timer)
+		start_ep_timer(ep);
+	if (stop_timer)
+		stop_ep_timer(ep);
+	if (disconnect)
+		c4iw_ep_disconnect(ep, 0, GFP_KERNEL);
+	if (release)
+		release_ep_resources(ep);
+	return 0;
+}
+
+/*
+ * Returns whether an ABORT_REQ_RSS message is a negative advice.
+ */
+static int is_neg_adv_abort(unsigned int status)
+{
+	return status == CPL_ERR_RTX_NEG_ADVICE ||
+	       status == CPL_ERR_PERSIST_NEG_ADVICE;
+}
+
+static int peer_abort(struct c4iw_dev *dev, struct sk_buff *skb)
+{
+	struct cpl_abort_req_rss *req = cplhdr(skb);
+	struct c4iw_ep *ep;
+	struct cpl_abort_rpl *rpl;
+	struct sk_buff *rpl_skb;
+	struct c4iw_qp_attributes attrs;
+	int ret;
+	int release = 0;
+	unsigned long flags;
+	struct tid_info *t = dev->rdev.lldi.tids;
+	unsigned int tid = GET_TID(req);
+	int stop_timer = 0;
+
+	ep = lookup_tid(t, tid);
+	if (is_neg_adv_abort(req->status)) {
+		PDBG("%s neg_adv_abort ep %p tid %u\n", __func__, ep,
+		     ep->hwtid);
+		return 0;
+	}
+	spin_lock_irqsave(&ep->com.lock, flags);
+	PDBG("%s ep %p tid %u state %u\n", __func__, ep, ep->hwtid,
+	     ep->com.state);
+	switch (ep->com.state) {
+	case CONNECTING:
+		break;
+	case MPA_REQ_WAIT:
+		stop_timer = 1;
+		break;
+	case MPA_REQ_SENT:
+		stop_timer = 1;
+		connect_reply_upcall(ep, -ECONNRESET);
+		break;
+	case MPA_REP_SENT:
+		ep->com.rpl_done = 1;
+		ep->com.rpl_err = -ECONNRESET;
+		PDBG("waking up ep %p\n", ep);
+		wake_up(&ep->com.waitq);
+		break;
+	case MPA_REQ_RCVD:
+
+		/*
+		 * We're gonna mark this puppy DEAD, but keep
+		 * the reference on it until the ULP accepts or
+		 * rejects the CR. Also wake up anyone waiting
+		 * in rdma connection migration (see c4iw_accept_cr()).
+		 */
+		ep->com.rpl_done = 1;
+		ep->com.rpl_err = -ECONNRESET;
+		PDBG("waking up ep %p tid %u\n", ep, ep->hwtid);
+		wake_up(&ep->com.waitq);
+		break;
+	case MORIBUND:
+	case CLOSING:
+		stop_timer = 1;
+		/*FALLTHROUGH*/
+	case FPDU_MODE:
+		if (ep->com.cm_id && ep->com.qp) {
+			attrs.next_state = C4IW_QP_STATE_ERROR;
+			ret = c4iw_modify_qp(ep->com.qp->rhp,
+				     ep->com.qp, C4IW_QP_ATTR_NEXT_STATE,
+				     &attrs, 1);
+			if (ret)
+				printk(KERN_ERR MOD
+				       "%s - qp <- error failed!\n",
+				       __func__);
+		}
+		peer_abort_upcall(ep);
+		break;
+	case ABORTING:
+		break;
+	case DEAD:
+		PDBG("%s PEER_ABORT IN DEAD STATE!!!!\n", __func__);
+		spin_unlock_irqrestore(&ep->com.lock, flags);
+		return 0;
+	default:
+		BUG_ON(1);
+		break;
+	}
+	dst_confirm(ep->dst);
+	if (ep->com.state != ABORTING) {
+		__state_set(&ep->com, DEAD);
+		release = 1;
+	}
+	spin_unlock_irqrestore(&ep->com.lock, flags);
+
+	rpl_skb = get_skb(skb, sizeof(*rpl), GFP_KERNEL);
+	if (!rpl_skb) {
+		printk(KERN_ERR MOD "%s - cannot allocate skb!\n",
+		       __func__);
+		release = 1;
+		goto out;
+	}
+	set_wr_txq(skb, CPL_PRIORITY_DATA, ep->txq_idx);
+	rpl = (struct cpl_abort_rpl *) skb_put(rpl_skb, sizeof(*rpl));
+	INIT_TP_WR(rpl, ep->hwtid);
+	OPCODE_TID(rpl) = cpu_to_be32(MK_OPCODE_TID(CPL_ABORT_RPL, ep->hwtid));
+	rpl->cmd = CPL_ABORT_NO_RST;
+	c4iw_ofld_send(&ep->com.dev->rdev, rpl_skb);
+out:
+	if (stop_timer)
+		stop_ep_timer(ep);
+	if (release)
+		release_ep_resources(ep);
+	return 0;
+}
+
+static int close_con_rpl(struct c4iw_dev *dev, struct sk_buff *skb)
+{
+	struct c4iw_ep *ep;
+	struct c4iw_qp_attributes attrs;
+	struct cpl_close_con_rpl *rpl = cplhdr(skb);
+	unsigned long flags;
+	int release = 0;
+	struct tid_info *t = dev->rdev.lldi.tids;
+	unsigned int tid = GET_TID(rpl);
+	int stop_timer = 0;
+
+	ep = lookup_tid(t, tid);
+
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+	BUG_ON(!ep);
+
+	/* The cm_id may be null if we failed to connect */
+	spin_lock_irqsave(&ep->com.lock, flags);
+	switch (ep->com.state) {
+	case CLOSING:
+		__state_set(&ep->com, MORIBUND);
+		break;
+	case MORIBUND:
+		stop_timer = 1;
+		if ((ep->com.cm_id) && (ep->com.qp)) {
+			attrs.next_state = C4IW_QP_STATE_IDLE;
+			c4iw_modify_qp(ep->com.qp->rhp,
+					     ep->com.qp,
+					     C4IW_QP_ATTR_NEXT_STATE,
+					     &attrs, 1);
+		}
+		close_complete_upcall(ep);
+		__state_set(&ep->com, DEAD);
+		release = 1;
+		break;
+	case ABORTING:
+	case DEAD:
+		break;
+	default:
+		BUG_ON(1);
+		break;
+	}
+	spin_unlock_irqrestore(&ep->com.lock, flags);
+	if (stop_timer)
+		stop_ep_timer(ep);
+	if (release)
+		release_ep_resources(ep);
+	return 0;
+}
+
+static int terminate(struct c4iw_dev *dev, struct sk_buff *skb)
+{
+	struct c4iw_ep *ep;
+	struct cpl_rdma_terminate *term = cplhdr(skb);
+	struct tid_info *t = dev->rdev.lldi.tids;
+	unsigned int tid = GET_TID(term);
+
+	ep = lookup_tid(t, tid);
+
+	if (state_read(&ep->com) != FPDU_MODE)
+		return 0;
+
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+	skb_pull(skb, sizeof *term);
+	PDBG("%s saving %d bytes of term msg\n", __func__, skb->len);
+	skb_copy_from_linear_data(skb, ep->com.qp->attr.terminate_buffer,
+				  skb->len);
+	ep->com.qp->attr.terminate_msg_len = skb->len;
+	ep->com.qp->attr.is_terminate_local = 0;
+	return 0;
+}
+
+/*
+ * Upcall from the adapter indicating data has been transmitted.
+ * For us its just the single MPA request or reply.  We can now free
+ * the skb holding the mpa message.
+ */
+static int fw4_ack(struct c4iw_dev *dev, struct sk_buff *skb)
+{
+	struct c4iw_ep *ep;
+	struct cpl_fw4_ack *hdr = cplhdr(skb);
+	u8 credits = hdr->credits;
+	unsigned int tid = GET_TID(hdr);
+	struct tid_info *t = dev->rdev.lldi.tids;
+
+
+	ep = lookup_tid(t, tid);
+	PDBG("%s ep %p tid %u credits %u\n", __func__, ep, ep->hwtid, credits);
+	if (credits == 0) {
+		PDBG(KERN_ERR "%s 0 credit ack ep %p tid %u state %u\n",
+			__func__, ep, ep->hwtid, state_read(&ep->com));
+		return 0;
+	}
+
+	dst_confirm(ep->dst);
+	if (!ep->mpa_skb) {
+		PDBG("%s rdma_init wr_ack ep %p tid %u state %u\n",
+			__func__, ep, ep->hwtid, state_read(&ep->com));
+		if (ep->mpa_attr.initiator) {
+			PDBG("%s initiator ep %p tid %u state %u\n",
+				__func__, ep, ep->hwtid, state_read(&ep->com));
+			if (peer2peer)
+				c4iw_post_zb_read(ep->com.qp);
+		} else {
+			PDBG("%s responder ep %p tid %u state %u\n",
+				__func__, ep, ep->hwtid, state_read(&ep->com));
+			ep->com.rpl_done = 1;
+			wake_up(&ep->com.waitq);
+		}
+	} else {
+		PDBG("%s last streaming msg ack ep %p tid %u state %u "
+		     "initiator %u freeing skb\n", __func__, ep, ep->hwtid,
+		     state_read(&ep->com), ep->mpa_attr.initiator ? 1 : 0);
+		kfree_skb(ep->mpa_skb);
+		ep->mpa_skb = NULL;
+	}
+	return 0;
+}
+
+static int fw6_msg(struct c4iw_dev *dev, struct sk_buff *skb)
+{
+	struct cpl_fw6_msg *rpl = cplhdr(skb);
+	struct c4iw_wr_wait *wr_waitp;
+	int ret;
+
+	PDBG("%s type %u\n", __func__, rpl->type);
+
+	switch (rpl->type) {
+	case 1:
+		ret = (int)((be64_to_cpu(rpl->data[0]) >> 8) & 0xff);
+		wr_waitp = (__force struct c4iw_wr_wait *)rpl->data[1];
+		PDBG("%s wr_waitp %p ret %u\n", __func__, wr_waitp, ret);
+		if (wr_waitp) {
+			wr_waitp->ret = ret;
+			wr_waitp->done = 1;
+			wake_up(&wr_waitp->wait);
+		}
+		break;
+	case 2:
+		c4iw_ev_dispatch(dev, (struct t4_cqe *)&rpl->data[0]);
+		break;
+	default:
+		printk(KERN_ERR MOD "%s unexpected fw6 msg type %u\n", __func__,
+		       rpl->type);
+		break;
+	}
+	return 0;
+}
+
+static void ep_timeout(unsigned long arg)
+{
+	struct c4iw_ep *ep = (struct c4iw_ep *)arg;
+	struct c4iw_qp_attributes attrs;
+	unsigned long flags;
+	int abort = 1;
+
+	spin_lock_irqsave(&ep->com.lock, flags);
+	PDBG("%s ep %p tid %u state %d\n", __func__, ep, ep->hwtid,
+	     ep->com.state);
+	switch (ep->com.state) {
+	case MPA_REQ_SENT:
+		__state_set(&ep->com, ABORTING);
+		connect_reply_upcall(ep, -ETIMEDOUT);
+		break;
+	case MPA_REQ_WAIT:
+		__state_set(&ep->com, ABORTING);
+		break;
+	case CLOSING:
+	case MORIBUND:
+		if (ep->com.cm_id && ep->com.qp) {
+			attrs.next_state = C4IW_QP_STATE_ERROR;
+			c4iw_modify_qp(ep->com.qp->rhp,
+				     ep->com.qp, C4IW_QP_ATTR_NEXT_STATE,
+				     &attrs, 1);
+		}
+		__state_set(&ep->com, ABORTING);
+		break;
+	default:
+		printk(KERN_ERR "%s unexpected state ep %p tid %u state %u\n",
+			__func__, ep, ep->hwtid, ep->com.state);
+		WARN_ON(1);
+		abort = 0;
+	}
+	spin_unlock_irqrestore(&ep->com.lock, flags);
+	if (abort)
+		abort_connection(ep, NULL, GFP_ATOMIC);
+	c4iw_put_ep(&ep->com);
+}
+
+int c4iw_reject_cr(struct iw_cm_id *cm_id, const void *pdata, u8 pdata_len)
+{
+	int err;
+	struct c4iw_ep *ep = to_ep(cm_id);
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+
+	if (state_read(&ep->com) == DEAD) {
+		c4iw_put_ep(&ep->com);
+		return -ECONNRESET;
+	}
+	BUG_ON(state_read(&ep->com) != MPA_REQ_RCVD);
+	if (mpa_rev == 0)
+		abort_connection(ep, NULL, GFP_KERNEL);
+	else {
+		err = send_mpa_reject(ep, pdata, pdata_len);
+		err = c4iw_ep_disconnect(ep, 0, GFP_KERNEL);
+	}
+	c4iw_put_ep(&ep->com);
+	return 0;
+}
+
+int c4iw_accept_cr(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
+{
+	int err;
+	struct c4iw_qp_attributes attrs;
+	enum c4iw_qp_attr_mask mask;
+	struct c4iw_ep *ep = to_ep(cm_id);
+	struct c4iw_dev *h = to_c4iw_dev(cm_id->device);
+	struct c4iw_qp *qp = get_qhp(h, conn_param->qpn);
+
+	PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid);
+	if (state_read(&ep->com) == DEAD) {
+		err = -ECONNRESET;
+		goto err;
+	}
+
+	BUG_ON(state_read(&ep->com) != MPA_REQ_RCVD);
+	BUG_ON(!qp);
+
+	if ((conn_param->ord > T4_MAX_READ_DEPTH) ||
+	    (conn_param->ird > T4_MAX_READ_DEPTH)) {
+		abort_connection(ep, NULL, GFP_KERNEL);
+		err = -EINVAL;
+		goto err;
+	}
+
+	cm_id->add_ref(cm_id);
+	ep->com.cm_id = cm_id;
+	ep->com.qp = qp;
+
+	ep->ird = conn_param->ird;
+	ep->ord = conn_param->ord;
+
+	if (peer2peer && ep->ird == 0)
+		ep->ird = 1;
+
+	PDBG("%s %d ird %d ord %d\n", __func__, __LINE__, ep->ird, ep->ord);
+
+	/* bind QP to EP and move to RTS */
+	attrs.mpa_attr = ep->mpa_attr;
+	attrs.max_ird = ep->ird;
+	attrs.max_ord = ep->ord;
+	attrs.llp_stream_handle = ep;
+	attrs.next_state = C4IW_QP_STATE_RTS;
+
+	/* bind QP and TID with INIT_WR */
+	mask = C4IW_QP_ATTR_NEXT_STATE |
+			     C4IW_QP_ATTR_LLP_STREAM_HANDLE |
+			     C4IW_QP_ATTR_MPA_ATTR |
+			     C4IW_QP_ATTR_MAX_IRD |
+			     C4IW_QP_ATTR_MAX_ORD;
+
+	err = c4iw_modify_qp(ep->com.qp->rhp,
+			     ep->com.qp, mask, &attrs, 1);
+	if (err)
+		goto err1;
+#if 0
+	/* if needed, wait for wr_ack */
+	if (t4_rqes_posted(&qp->wq)) {
+		wait_event(ep->com.waitq, ep->com.rpl_done);
+		err = ep->com.rpl_err;
+		if (err)
+			goto err1;
+	}
+#endif
+	err = send_mpa_reply(ep, conn_param->private_data,
+			     conn_param->private_data_len);
+	if (err)
+		goto err1;
+
+	state_set(&ep->com, FPDU_MODE);
+	established_upcall(ep);
+	c4iw_put_ep(&ep->com);
+	return 0;
+err1:
+	ep->com.cm_id = NULL;
+	ep->com.qp = NULL;
+	cm_id->rem_ref(cm_id);
+err:
+	c4iw_put_ep(&ep->com);
+	return err;
+}
+
+int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
+{
+	int err = 0;
+	struct c4iw_dev *dev = to_c4iw_dev(cm_id->device);
+	struct c4iw_ep *ep;
+	struct rtable *rt;
+	struct net_device *pdev;
+	int step;
+
+	ep = alloc_ep(sizeof(*ep), GFP_KERNEL);
+	if (!ep) {
+		printk(KERN_ERR MOD "%s - cannot alloc ep.\n", __func__);
+		err = -ENOMEM;
+		goto out;
+	}
+	init_timer(&ep->timer);
+	ep->plen = conn_param->private_data_len;
+	if (ep->plen)
+		memcpy(ep->mpa_pkt + sizeof(struct mpa_message),
+		       conn_param->private_data, ep->plen);
+	ep->ird = conn_param->ird;
+	ep->ord = conn_param->ord;
+
+	if (peer2peer && ep->ord == 0)
+		ep->ord = 1;
+
+	cm_id->add_ref(cm_id);
+	ep->com.dev = dev;
+	ep->com.cm_id = cm_id;
+	ep->com.qp = get_qhp(dev, conn_param->qpn);
+	BUG_ON(!ep->com.qp);
+	PDBG("%s qpn 0x%x qp %p cm_id %p\n", __func__, conn_param->qpn,
+	     ep->com.qp, cm_id);
+
+	/*
+	 * Allocate an active TID to initiate a TCP connection.
+	 */
+	ep->atid = cxgb4_alloc_atid(dev->rdev.lldi.tids, ep);
+	if (ep->atid == -1) {
+		printk(KERN_ERR MOD "%s - cannot alloc atid.\n", __func__);
+		err = -ENOMEM;
+		goto fail2;
+	}
+
+	PDBG("%s saddr 0x%x sport 0x%x raddr 0x%x rport 0x%x\n", __func__,
+	     ntohl(cm_id->local_addr.sin_addr.s_addr),
+	     ntohs(cm_id->local_addr.sin_port),
+	     ntohl(cm_id->remote_addr.sin_addr.s_addr),
+	     ntohs(cm_id->remote_addr.sin_port));
+
+	/* find a route */
+	rt = find_route(dev,
+			cm_id->local_addr.sin_addr.s_addr,
+			cm_id->remote_addr.sin_addr.s_addr,
+			cm_id->local_addr.sin_port,
+			cm_id->remote_addr.sin_port, 0);
+	if (!rt) {
+		printk(KERN_ERR MOD "%s - cannot find route.\n", __func__);
+		err = -EHOSTUNREACH;
+		goto fail3;
+	}
+	ep->dst = &rt->u.dst;
+
+	/* get a l2t entry */
+	if (ep->dst->neighbour->dev->flags & IFF_LOOPBACK) {
+		PDBG("%s LOOPBACK\n", __func__);
+		pdev = ip_dev_find(&init_net,
+				   cm_id->remote_addr.sin_addr.s_addr);
+		ep->l2t = cxgb4_l2t_get(ep->com.dev->rdev.lldi.l2t,
+					ep->dst->neighbour,
+					pdev, 0);
+		ep->mtu = pdev->mtu;
+		ep->tx_chan = cxgb4_port_chan(pdev);
+		ep->smac_idx = ep->tx_chan << 1;
+		step = ep->com.dev->rdev.lldi.ntxq /
+		       ep->com.dev->rdev.lldi.nchan;
+		ep->txq_idx = cxgb4_port_idx(pdev) * step;
+		step = ep->com.dev->rdev.lldi.nrxq /
+		       ep->com.dev->rdev.lldi.nchan;
+		ep->rss_qid = ep->com.dev->rdev.lldi.rxq_ids[
+			      cxgb4_port_idx(pdev) * step];
+		dev_put(pdev);
+	} else {
+		ep->l2t = cxgb4_l2t_get(ep->com.dev->rdev.lldi.l2t,
+					ep->dst->neighbour,
+					ep->dst->neighbour->dev, 0);
+		ep->mtu = dst_mtu(ep->dst);
+		ep->tx_chan = cxgb4_port_chan(ep->dst->neighbour->dev);
+		ep->smac_idx = ep->tx_chan << 1;
+		step = ep->com.dev->rdev.lldi.ntxq /
+		       ep->com.dev->rdev.lldi.nchan;
+		ep->txq_idx = cxgb4_port_idx(ep->dst->neighbour->dev) * step;
+		step = ep->com.dev->rdev.lldi.nrxq /
+		       ep->com.dev->rdev.lldi.nchan;
+		ep->rss_qid = ep->com.dev->rdev.lldi.rxq_ids[
+			      cxgb4_port_idx(ep->dst->neighbour->dev) * step];
+	}
+	if (!ep->l2t) {
+		printk(KERN_ERR MOD "%s - cannot alloc l2e.\n", __func__);
+		err = -ENOMEM;
+		goto fail4;
+	}
+
+	PDBG("%s txq_idx %u tx_chan %u smac_idx %u rss_qid %u l2t_idx %u\n",
+		__func__, ep->txq_idx, ep->tx_chan, ep->smac_idx, ep->rss_qid,
+		ep->l2t->idx);
+
+	state_set(&ep->com, CONNECTING);
+	ep->tos = 0;
+	ep->com.local_addr = cm_id->local_addr;
+	ep->com.remote_addr = cm_id->remote_addr;
+
+	/* send connect request to rnic */
+	err = send_connect(ep);
+	if (!err)
+		goto out;
+
+	cxgb4_l2t_release(ep->l2t);
+fail4:
+	dst_release(ep->dst);
+fail3:
+	cxgb4_free_atid(ep->com.dev->rdev.lldi.tids, ep->atid);
+fail2:
+	cm_id->rem_ref(cm_id);
+	c4iw_put_ep(&ep->com);
+out:
+	return err;
+}
+
+int c4iw_create_listen(struct iw_cm_id *cm_id, int backlog)
+{
+	int err = 0;
+	struct c4iw_dev *dev = to_c4iw_dev(cm_id->device);
+	struct c4iw_listen_ep *ep;
+
+
+	might_sleep();
+
+	ep = alloc_ep(sizeof(*ep), GFP_KERNEL);
+	if (!ep) {
+		printk(KERN_ERR MOD "%s - cannot alloc ep.\n", __func__);
+		err = -ENOMEM;
+		goto fail1;
+	}
+	PDBG("%s ep %p\n", __func__, ep);
+	cm_id->add_ref(cm_id);
+	ep->com.cm_id = cm_id;
+	ep->com.dev = dev;
+	ep->backlog = backlog;
+	ep->com.local_addr = cm_id->local_addr;
+
+	/*
+	 * Allocate a server TID.
+	 */
+	ep->stid = cxgb4_alloc_stid(dev->rdev.lldi.tids, PF_INET, ep);
+	if (ep->stid == -1) {
+		printk(KERN_ERR MOD "%s - cannot alloc atid.\n", __func__);
+		err = -ENOMEM;
+		goto fail2;
+	}
+
+	state_set(&ep->com, LISTEN);
+	err = cxgb4_create_server(ep->com.dev->rdev.lldi.ports[0], ep->stid,
+				  ep->com.local_addr.sin_addr.s_addr,
+				  ep->com.local_addr.sin_port,
+				  ep->com.dev->rdev.lldi.rxq_ids[0]);
+	if (err)
+		goto fail3;
+
+	/* wait for pass_open_rpl */
+	wait_event(ep->com.waitq, ep->com.rpl_done);
+	err = ep->com.rpl_err;
+	if (!err) {
+		cm_id->provider_data = ep;
+		goto out;
+	}
+fail3:
+	cxgb4_free_stid(ep->com.dev->rdev.lldi.tids, ep->stid, PF_INET);
+fail2:
+	cm_id->rem_ref(cm_id);
+	c4iw_put_ep(&ep->com);
+fail1:
+out:
+	return err;
+}
+
+int c4iw_destroy_listen(struct iw_cm_id *cm_id)
+{
+	int err;
+	struct c4iw_listen_ep *ep = to_listen_ep(cm_id);
+
+	PDBG("%s ep %p\n", __func__, ep);
+
+	might_sleep();
+	state_set(&ep->com, DEAD);
+	ep->com.rpl_done = 0;
+	ep->com.rpl_err = 0;
+	err = listen_stop(ep);
+	if (err)
+		goto done;
+	wait_event(ep->com.waitq, ep->com.rpl_done);
+	cxgb4_free_stid(ep->com.dev->rdev.lldi.tids, ep->stid, PF_INET);
+done:
+	err = ep->com.rpl_err;
+	cm_id->rem_ref(cm_id);
+	c4iw_put_ep(&ep->com);
+	return err;
+}
+
+int c4iw_ep_disconnect(struct c4iw_ep *ep, int abrupt, gfp_t gfp)
+{
+	int ret = 0;
+	unsigned long flags;
+	int close = 0;
+	int fatal = 0;
+	struct c4iw_rdev *rdev;
+	int start_timer = 0;
+	int stop_timer = 0;
+
+	spin_lock_irqsave(&ep->com.lock, flags);
+
+	PDBG("%s ep %p state %s, abrupt %d\n", __func__, ep,
+	     states[ep->com.state], abrupt);
+
+	rdev = &ep->com.dev->rdev;
+	if (c4iw_fatal_error(rdev)) {
+		fatal = 1;
+		close_complete_upcall(ep);
+		ep->com.state = DEAD;
+	}
+	switch (ep->com.state) {
+	case MPA_REQ_WAIT:
+	case MPA_REQ_SENT:
+	case MPA_REQ_RCVD:
+	case MPA_REP_SENT:
+	case FPDU_MODE:
+		close = 1;
+		if (abrupt)
+			ep->com.state = ABORTING;
+		else {
+			ep->com.state = CLOSING;
+			start_timer = 1;
+		}
+		set_bit(CLOSE_SENT, &ep->com.flags);
+		break;
+	case CLOSING:
+		if (!test_and_set_bit(CLOSE_SENT, &ep->com.flags)) {
+			close = 1;
+			if (abrupt) {
+				stop_timer = 1;
+				ep->com.state = ABORTING;
+			} else
+				ep->com.state = MORIBUND;
+		}
+		break;
+	case MORIBUND:
+	case ABORTING:
+	case DEAD:
+		PDBG("%s ignoring disconnect ep %p state %u\n",
+		     __func__, ep, ep->com.state);
+		break;
+	default:
+		BUG();
+		break;
+	}
+
+	spin_unlock_irqrestore(&ep->com.lock, flags);
+	if (start_timer)
+		start_ep_timer(ep);
+	if (stop_timer)
+		stop_ep_timer(ep);
+	if (close) {
+		if (abrupt)
+			ret = abort_connection(ep, NULL, gfp);
+		else
+			ret = send_halfclose(ep, gfp);
+		if (ret)
+			fatal = 1;
+	}
+	if (fatal)
+		release_ep_resources(ep);
+	return ret;
+}
+
+int c4iw_ep_redirect(void *ctx, struct dst_entry *old, struct dst_entry *new,
+		     struct l2t_entry *l2t)
+{
+	struct c4iw_ep *ep = ctx;
+
+	if (ep->dst != old)
+		return 0;
+
+	PDBG("%s ep %p redirect to dst %p l2t %p\n", __func__, ep, new,
+	     l2t);
+	dst_hold(new);
+	cxgb4_l2t_release(ep->l2t);
+	ep->l2t = l2t;
+	dst_release(old);
+	ep->dst = new;
+	return 1;
+}
+
+/*
+ * All the CM events are handled on a work queue to have a safe context.
+ */
+static int sched(struct c4iw_dev *dev, struct sk_buff *skb)
+{
+
+	/*
+	 * Save dev in the skb->cb area.
+	 */
+	*((struct c4iw_dev **) (skb->cb + sizeof(void *))) = dev;
+
+	/*
+	 * Queue the skb and schedule the worker thread.
+	 */
+	skb_queue_tail(&rxq, skb);
+	queue_work(workq, &skb_work);
+	return 0;
+}
+
+static int set_tcb_rpl(struct c4iw_dev *dev, struct sk_buff *skb)
+{
+	struct cpl_set_tcb_rpl *rpl = cplhdr(skb);
+
+	if (rpl->status != CPL_ERR_NONE) {
+		printk(KERN_ERR MOD "Unexpected SET_TCB_RPL status %u "
+		       "for tid %u\n", rpl->status, GET_TID(rpl));
+	}
+	return 0;
+}
+
+int __init c4iw_cm_init(void)
+{
+	skb_queue_head_init(&rxq);
+
+	workq = create_singlethread_workqueue("iw_cxgb4");
+	if (!workq)
+		return -ENOMEM;
+
+	/*
+	 * Most upcalls from the T4 Core go to sched() to
+	 * schedule the processing on a work queue.
+	 */
+	c4iw_handlers[CPL_ACT_ESTABLISH] = sched;
+	c4iw_handlers[CPL_ACT_OPEN_RPL] = sched;
+	c4iw_handlers[CPL_RX_DATA] = sched;
+	c4iw_handlers[CPL_ABORT_RPL_RSS] = sched;
+	c4iw_handlers[CPL_ABORT_RPL] = sched;
+	c4iw_handlers[CPL_PASS_OPEN_RPL] = sched;
+	c4iw_handlers[CPL_CLOSE_LISTSRV_RPL] = sched;
+	c4iw_handlers[CPL_PASS_ACCEPT_REQ] = sched;
+	c4iw_handlers[CPL_PASS_ESTABLISH] = sched;
+	c4iw_handlers[CPL_PEER_CLOSE] = sched;
+	c4iw_handlers[CPL_CLOSE_CON_RPL] = sched;
+	c4iw_handlers[CPL_ABORT_REQ_RSS] = sched;
+	c4iw_handlers[CPL_RDMA_TERMINATE] = sched;
+	c4iw_handlers[CPL_FW4_ACK] = sched;
+	c4iw_handlers[CPL_SET_TCB_RPL] = set_tcb_rpl;
+	c4iw_handlers[CPL_FW6_MSG] = fw6_msg;
+
+	/*
+	 * These are the real handlers that are called from a
+	 * work queue.
+	 */
+	work_handlers[CPL_ACT_ESTABLISH] = act_establish;
+	work_handlers[CPL_ACT_OPEN_RPL] = act_open_rpl;
+	work_handlers[CPL_RX_DATA] = rx_data;
+	work_handlers[CPL_ABORT_RPL_RSS] = abort_rpl;
+	work_handlers[CPL_ABORT_RPL] = abort_rpl;
+	work_handlers[CPL_PASS_OPEN_RPL] = pass_open_rpl;
+	work_handlers[CPL_CLOSE_LISTSRV_RPL] = close_listsrv_rpl;
+	work_handlers[CPL_PASS_ACCEPT_REQ] = pass_accept_req;
+	work_handlers[CPL_PASS_ESTABLISH] = pass_establish;
+	work_handlers[CPL_PEER_CLOSE] = peer_close;
+	work_handlers[CPL_ABORT_REQ_RSS] = peer_abort;
+	work_handlers[CPL_CLOSE_CON_RPL] = close_con_rpl;
+	work_handlers[CPL_RDMA_TERMINATE] = terminate;
+	work_handlers[CPL_FW4_ACK] = fw4_ack;
+	return 0;
+}
+
+void __exit c4iw_cm_term(void)
+{
+	flush_workqueue(workq);
+	destroy_workqueue(workq);
+}