Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
  RDMA/nes: Add a driver for NetEffect RNICs
  IB/mthca: Return proper error codes from mthca_fmr_alloc()
  IB: Avoid marking __devinitdata as const
  IB/mlx4: Actually print out the driver version
  IB/ib_mthca: Pre-link receive WQEs in Tavor mode
  IB/mthca: Remove checks for srq->first_free < 0
  IB/fmr_pool: Allocate page list for pool FMRs only when caching enabled
  IB/srp: Retry stale connections
  mlx4_core: Don't read reserved fields in mlx4_QUERY_ADAPTER()
  IB/mthca: Don't read reserved fields in mthca_QUERY_ADAPTER()
  IPoIB: Remove a misleading debug print
  IPoIB: Handle bonding failover race for connected neighbours too
  IB/mthca: Fix and simplify page size calculation in mthca_reg_phys_mr()
  IB/ehca: Add PMA support
  IB/ehca: Update sma_attr also in case of disruptive config change
  IB/ehca: Prevent sending UD packets to QP0
  IB/cm: Add interim support for routed paths
  mlx4_core: Fix more section mismatches

MAINTAINERS

@@ -2681,6 +2681,16 @@ M:	James.Bottomley@HansenPartnership.com
 L:	linux-scsi@vger.kernel.org
 S:	Maintained
 
+NETEFFECT IWARP RNIC DRIVER (IW_NES)
+P:	Faisal Latif
+M:	flatif@neteffect.com
+P:	Glenn Streiff
+M:	gstreiff@neteffect.com
+L:	general@lists.openfabrics.org
+W:	http://www.neteffect.com
+S:	Supported
+F:	drivers/infiniband/hw/nes/
+
 NETEM NETWORK EMULATOR
 P:	Stephen Hemminger
 M:	shemminger@linux-foundation.org

drivers/infiniband/Kconfig

@@ -44,8 +44,8 @@ source "drivers/infiniband/hw/ipath/Kconfig"
 source "drivers/infiniband/hw/ehca/Kconfig"
 source "drivers/infiniband/hw/amso1100/Kconfig"
 source "drivers/infiniband/hw/cxgb3/Kconfig"
-
 source "drivers/infiniband/hw/mlx4/Kconfig"
+source "drivers/infiniband/hw/nes/Kconfig"
 
 source "drivers/infiniband/ulp/ipoib/Kconfig"
 

drivers/infiniband/Makefile

@@ -5,6 +5,7 @@ obj-$(CONFIG_INFINIBAND_EHCA)		+= hw/ehca/
 obj-$(CONFIG_INFINIBAND_AMSO1100)	+= hw/amso1100/
 obj-$(CONFIG_INFINIBAND_CXGB3)		+= hw/cxgb3/
 obj-$(CONFIG_MLX4_INFINIBAND)		+= hw/mlx4/
+obj-$(CONFIG_INFINIBAND_NES)		+= hw/nes/
 obj-$(CONFIG_INFINIBAND_IPOIB)		+= ulp/ipoib/
 obj-$(CONFIG_INFINIBAND_SRP)		+= ulp/srp/
 obj-$(CONFIG_INFINIBAND_ISER)		+= ulp/iser/

drivers/infiniband/core/cm.c

@@ -974,6 +974,9 @@ static void cm_format_req(struct cm_req_msg *req_msg,
 			  struct cm_id_private *cm_id_priv,
 			  struct ib_cm_req_param *param)
 {
+	struct ib_sa_path_rec *pri_path = param->primary_path;
+	struct ib_sa_path_rec *alt_path = param->alternate_path;
+
 	cm_format_mad_hdr(&req_msg->hdr, CM_REQ_ATTR_ID,
 			  cm_form_tid(cm_id_priv, CM_MSG_SEQUENCE_REQ));
 
@@ -997,35 +1000,46 @@ static void cm_format_req(struct cm_req_msg *req_msg,
 	cm_req_set_max_cm_retries(req_msg, param->max_cm_retries);
 	cm_req_set_srq(req_msg, param->srq);
 
-	req_msg->primary_local_lid = param->primary_path->slid;
-	req_msg->primary_remote_lid = param->primary_path->dlid;
-	req_msg->primary_local_gid = param->primary_path->sgid;
-	req_msg->primary_remote_gid = param->primary_path->dgid;
-	cm_req_set_primary_flow_label(req_msg, param->primary_path->flow_label);
-	cm_req_set_primary_packet_rate(req_msg, param->primary_path->rate);
-	req_msg->primary_traffic_class = param->primary_path->traffic_class;
-	req_msg->primary_hop_limit = param->primary_path->hop_limit;
-	cm_req_set_primary_sl(req_msg, param->primary_path->sl);
-	cm_req_set_primary_subnet_local(req_msg, 1); /* local only... */
+	if (pri_path->hop_limit <= 1) {
+		req_msg->primary_local_lid = pri_path->slid;
+		req_msg->primary_remote_lid = pri_path->dlid;
+	} else {
+		/* Work-around until there's a way to obtain remote LID info */
+		req_msg->primary_local_lid = IB_LID_PERMISSIVE;
+		req_msg->primary_remote_lid = IB_LID_PERMISSIVE;
+	}
+	req_msg->primary_local_gid = pri_path->sgid;
+	req_msg->primary_remote_gid = pri_path->dgid;
+	cm_req_set_primary_flow_label(req_msg, pri_path->flow_label);
+	cm_req_set_primary_packet_rate(req_msg, pri_path->rate);
+	req_msg->primary_traffic_class = pri_path->traffic_class;
+	req_msg->primary_hop_limit = pri_path->hop_limit;
+	cm_req_set_primary_sl(req_msg, pri_path->sl);
+	cm_req_set_primary_subnet_local(req_msg, (pri_path->hop_limit <= 1));
 	cm_req_set_primary_local_ack_timeout(req_msg,
 		cm_ack_timeout(cm_id_priv->av.port->cm_dev->ack_delay,
-			       param->primary_path->packet_life_time));
+			       pri_path->packet_life_time));
 
-	if (param->alternate_path) {
-		req_msg->alt_local_lid = param->alternate_path->slid;
-		req_msg->alt_remote_lid = param->alternate_path->dlid;
-		req_msg->alt_local_gid = param->alternate_path->sgid;
-		req_msg->alt_remote_gid = param->alternate_path->dgid;
+	if (alt_path) {
+		if (alt_path->hop_limit <= 1) {
+			req_msg->alt_local_lid = alt_path->slid;
+			req_msg->alt_remote_lid = alt_path->dlid;
+		} else {
+			req_msg->alt_local_lid = IB_LID_PERMISSIVE;
+			req_msg->alt_remote_lid = IB_LID_PERMISSIVE;
+		}
+		req_msg->alt_local_gid = alt_path->sgid;
+		req_msg->alt_remote_gid = alt_path->dgid;
 		cm_req_set_alt_flow_label(req_msg,
-					  param->alternate_path->flow_label);
-		cm_req_set_alt_packet_rate(req_msg, param->alternate_path->rate);
-		req_msg->alt_traffic_class = param->alternate_path->traffic_class;
-		req_msg->alt_hop_limit = param->alternate_path->hop_limit;
-		cm_req_set_alt_sl(req_msg, param->alternate_path->sl);
-		cm_req_set_alt_subnet_local(req_msg, 1); /* local only... */
+					  alt_path->flow_label);
+		cm_req_set_alt_packet_rate(req_msg, alt_path->rate);
+		req_msg->alt_traffic_class = alt_path->traffic_class;
+		req_msg->alt_hop_limit = alt_path->hop_limit;
+		cm_req_set_alt_sl(req_msg, alt_path->sl);
+		cm_req_set_alt_subnet_local(req_msg, (alt_path->hop_limit <= 1));
 		cm_req_set_alt_local_ack_timeout(req_msg,
 			cm_ack_timeout(cm_id_priv->av.port->cm_dev->ack_delay,
-				       param->alternate_path->packet_life_time));
+				       alt_path->packet_life_time));
 	}
 
 	if (param->private_data && param->private_data_len)
@@ -1441,6 +1455,34 @@ static struct cm_id_private * cm_match_req(struct cm_work *work,
 	return listen_cm_id_priv;
 }
 
+/*
+ * Work-around for inter-subnet connections.  If the LIDs are permissive,
+ * we need to override the LID/SL data in the REQ with the LID information
+ * in the work completion.
+ */
+static void cm_process_routed_req(struct cm_req_msg *req_msg, struct ib_wc *wc)
+{
+	if (!cm_req_get_primary_subnet_local(req_msg)) {
+		if (req_msg->primary_local_lid == IB_LID_PERMISSIVE) {
+			req_msg->primary_local_lid = cpu_to_be16(wc->slid);
+			cm_req_set_primary_sl(req_msg, wc->sl);
+		}
+
+		if (req_msg->primary_remote_lid == IB_LID_PERMISSIVE)
+			req_msg->primary_remote_lid = cpu_to_be16(wc->dlid_path_bits);
+	}
+
+	if (!cm_req_get_alt_subnet_local(req_msg)) {
+		if (req_msg->alt_local_lid == IB_LID_PERMISSIVE) {
+			req_msg->alt_local_lid = cpu_to_be16(wc->slid);
+			cm_req_set_alt_sl(req_msg, wc->sl);
+		}
+
+		if (req_msg->alt_remote_lid == IB_LID_PERMISSIVE)
+			req_msg->alt_remote_lid = cpu_to_be16(wc->dlid_path_bits);
+	}
+}
+
 static int cm_req_handler(struct cm_work *work)
 {
 	struct ib_cm_id *cm_id;
@@ -1481,6 +1523,7 @@ static int cm_req_handler(struct cm_work *work)
 	cm_id_priv->id.service_id = req_msg->service_id;
 	cm_id_priv->id.service_mask = __constant_cpu_to_be64(~0ULL);
 
+	cm_process_routed_req(req_msg, work->mad_recv_wc->wc);
 	cm_format_paths_from_req(req_msg, &work->path[0], &work->path[1]);
 	ret = cm_init_av_by_path(&work->path[0], &cm_id_priv->av);
 	if (ret) {

drivers/infiniband/core/fmr_pool.c

@@ -320,10 +320,13 @@ struct ib_fmr_pool *ib_create_fmr_pool(struct ib_pd             *pd,
 			.max_maps   = pool->max_remaps,
 			.page_shift = params->page_shift
 		};
+		int bytes_per_fmr = sizeof *fmr;
+
+		if (pool->cache_bucket)
+			bytes_per_fmr += params->max_pages_per_fmr * sizeof (u64);
 
 		for (i = 0; i < params->pool_size; ++i) {
-			fmr = kmalloc(sizeof *fmr + params->max_pages_per_fmr * sizeof (u64),
-				      GFP_KERNEL);
+			fmr = kmalloc(bytes_per_fmr, GFP_KERNEL);
 			if (!fmr) {
 				printk(KERN_WARNING PFX "failed to allocate fmr "
 				       "struct for FMR %d\n", i);

drivers/infiniband/hw/ehca/ehca_classes.h

@@ -101,6 +101,7 @@ struct ehca_sport {
 	spinlock_t mod_sqp_lock;
 	enum ib_port_state port_state;
 	struct ehca_sma_attr saved_attr;
+	u32 pma_qp_nr;
 };
 
 #define HCA_CAP_MR_PGSIZE_4K  0x80000000

drivers/infiniband/hw/ehca/ehca_irq.c

@@ -403,6 +403,8 @@ static void parse_ec(struct ehca_shca *shca, u64 eqe)
 			sport->port_state = IB_PORT_ACTIVE;
 			dispatch_port_event(shca, port, IB_EVENT_PORT_ACTIVE,
 					    "is active");
+			ehca_query_sma_attr(shca, port,
+					    &sport->saved_attr);
 		} else
 			notify_port_conf_change(shca, port);
 		break;

drivers/infiniband/hw/ehca/ehca_iverbs.h

@@ -187,6 +187,11 @@ int ehca_dealloc_ucontext(struct ib_ucontext *context);
 
 int ehca_mmap(struct ib_ucontext *context, struct vm_area_struct *vma);
 
+int ehca_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);
+
 void ehca_poll_eqs(unsigned long data);
 
 int ehca_calc_ipd(struct ehca_shca *shca, int port,

drivers/infiniband/hw/ehca/ehca_main.c

@@ -472,7 +472,7 @@ int ehca_init_device(struct ehca_shca *shca)
 	shca->ib_device.dealloc_fmr	    = ehca_dealloc_fmr;
 	shca->ib_device.attach_mcast	    = ehca_attach_mcast;
 	shca->ib_device.detach_mcast	    = ehca_detach_mcast;
-	/* shca->ib_device.process_mad	    = ehca_process_mad;	    */
+	shca->ib_device.process_mad	    = ehca_process_mad;
 	shca->ib_device.mmap		    = ehca_mmap;
 
 	if (EHCA_BMASK_GET(HCA_CAP_SRQ, shca->hca_cap)) {

drivers/infiniband/hw/ehca/ehca_reqs.c

@@ -209,6 +209,10 @@ static inline int ehca_write_swqe(struct ehca_qp *qp,
 			ehca_gen_err("wr.ud.ah is NULL. qp=%p", qp);
 			return -EINVAL;
 		}
+		if (unlikely(send_wr->wr.ud.remote_qpn == 0)) {
+			ehca_gen_err("dest QP# is 0. qp=%x", qp->real_qp_num);
+			return -EINVAL;
+		}
 		my_av = container_of(send_wr->wr.ud.ah, struct ehca_av, ib_ah);
 		wqe_p->u.ud_av.ud_av = my_av->av;
 

drivers/infiniband/hw/ehca/ehca_sqp.c

@@ -39,12 +39,18 @@
  * POSSIBILITY OF SUCH DAMAGE.
  */
 
+#include <rdma/ib_mad.h>
 
 #include "ehca_classes.h"
 #include "ehca_tools.h"
 #include "ehca_iverbs.h"
 #include "hcp_if.h"
 
+#define IB_MAD_STATUS_REDIRECT		__constant_htons(0x0002)
+#define IB_MAD_STATUS_UNSUP_VERSION	__constant_htons(0x0004)
+#define IB_MAD_STATUS_UNSUP_METHOD	__constant_htons(0x0008)
+
+#define IB_PMA_CLASS_PORT_INFO		__constant_htons(0x0001)
 
 /**
  * ehca_define_sqp - Defines special queue pair 1 (GSI QP). When special queue
@@ -83,6 +89,9 @@ u64 ehca_define_sqp(struct ehca_shca *shca,
 				 port, ret);
 			return ret;
 		}
+		shca->sport[port - 1].pma_qp_nr = pma_qp_nr;
+		ehca_dbg(&shca->ib_device, "port=%x pma_qp_nr=%x",
+			 port, pma_qp_nr);
 		break;
 	default:
 		ehca_err(&shca->ib_device, "invalid qp_type=%x",
@@ -109,3 +118,85 @@ u64 ehca_define_sqp(struct ehca_shca *shca,
 
 	return H_SUCCESS;
 }
+
+struct ib_perf {
+	struct ib_mad_hdr mad_hdr;
+	u8 reserved[40];
+	u8 data[192];
+} __attribute__ ((packed));
+
+
+static int ehca_process_perf(struct ib_device *ibdev, u8 port_num,
+			     struct ib_mad *in_mad, struct ib_mad *out_mad)
+{
+	struct ib_perf *in_perf = (struct ib_perf *)in_mad;
+	struct ib_perf *out_perf = (struct ib_perf *)out_mad;
+	struct ib_class_port_info *poi =
+		(struct ib_class_port_info *)out_perf->data;
+	struct ehca_shca *shca =
+		container_of(ibdev, struct ehca_shca, ib_device);
+	struct ehca_sport *sport = &shca->sport[port_num - 1];
+
+	ehca_dbg(ibdev, "method=%x", in_perf->mad_hdr.method);
+
+	*out_mad = *in_mad;
+
+	if (in_perf->mad_hdr.class_version != 1) {
+		ehca_warn(ibdev, "Unsupported class_version=%x",
+			  in_perf->mad_hdr.class_version);
+		out_perf->mad_hdr.status = IB_MAD_STATUS_UNSUP_VERSION;
+		goto perf_reply;
+	}
+
+	switch (in_perf->mad_hdr.method) {
+	case IB_MGMT_METHOD_GET:
+	case IB_MGMT_METHOD_SET:
+		/* set class port info for redirection */
+		out_perf->mad_hdr.attr_id = IB_PMA_CLASS_PORT_INFO;
+		out_perf->mad_hdr.status = IB_MAD_STATUS_REDIRECT;
+		memset(poi, 0, sizeof(*poi));
+		poi->base_version = 1;
+		poi->class_version = 1;
+		poi->resp_time_value = 18;
+		poi->redirect_lid = sport->saved_attr.lid;
+		poi->redirect_qp = sport->pma_qp_nr;
+		poi->redirect_qkey = IB_QP1_QKEY;
+		poi->redirect_pkey = IB_DEFAULT_PKEY_FULL;
+
+		ehca_dbg(ibdev, "ehca_pma_lid=%x ehca_pma_qp=%x",
+			 sport->saved_attr.lid, sport->pma_qp_nr);
+		break;
+
+	case IB_MGMT_METHOD_GET_RESP:
+		return IB_MAD_RESULT_FAILURE;
+
+	default:
+		out_perf->mad_hdr.status = IB_MAD_STATUS_UNSUP_METHOD;
+		break;
+	}
+
+perf_reply:
+	out_perf->mad_hdr.method = IB_MGMT_METHOD_GET_RESP;
+
+	return IB_MAD_RESULT_SUCCESS | IB_MAD_RESULT_REPLY;
+}
+
+int ehca_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)
+{
+	int ret;
+
+	if (!port_num || port_num > ibdev->phys_port_cnt)
+		return IB_MAD_RESULT_FAILURE;
+
+	/* accept only pma request */
+	if (in_mad->mad_hdr.mgmt_class != IB_MGMT_CLASS_PERF_MGMT)
+		return IB_MAD_RESULT_SUCCESS;
+
+	ehca_dbg(ibdev, "port_num=%x src_qp=%x", port_num, in_wc->src_qp);
+	ret = ehca_process_perf(ibdev, port_num, in_mad, out_mad);
+
+	return ret;
+}

drivers/infiniband/hw/mlx4/main.c

@@ -52,7 +52,7 @@ MODULE_DESCRIPTION("Mellanox ConnectX HCA InfiniBand driver");
 MODULE_LICENSE("Dual BSD/GPL");
 MODULE_VERSION(DRV_VERSION);
 
-static const char mlx4_ib_version[] __devinitdata =
+static const char mlx4_ib_version[] =
 	DRV_NAME ": Mellanox ConnectX InfiniBand driver v"
 	DRV_VERSION " (" DRV_RELDATE ")\n";
 
@@ -468,6 +468,7 @@ static int init_node_data(struct mlx4_ib_dev *dev)
 	if (err)
 		goto out;
 
+	dev->dev->rev_id = be32_to_cpup((__be32 *) (out_mad->data + 32));
 	memcpy(&dev->ib_dev.node_guid, out_mad->data + 12, 8);
 
 out:
@@ -516,9 +517,16 @@ static struct class_device_attribute *mlx4_class_attributes[] = {
 
 static void *mlx4_ib_add(struct mlx4_dev *dev)
 {
+	static int mlx4_ib_version_printed;
 	struct mlx4_ib_dev *ibdev;
 	int i;
 
+
+	if (!mlx4_ib_version_printed) {
+		printk(KERN_INFO "%s", mlx4_ib_version);
+		++mlx4_ib_version_printed;
+	}
+
 	ibdev = (struct mlx4_ib_dev *) ib_alloc_device(sizeof *ibdev);
 	if (!ibdev) {
 		dev_err(&dev->pdev->dev, "Device struct alloc failed\n");

drivers/infiniband/hw/mthca/mthca_cmd.c

@@ -1255,9 +1255,14 @@ int mthca_QUERY_ADAPTER(struct mthca_dev *dev,
 	if (err)
 		goto out;
 
-	MTHCA_GET(adapter->vendor_id, outbox,   QUERY_ADAPTER_VENDOR_ID_OFFSET);
-	MTHCA_GET(adapter->device_id, outbox,   QUERY_ADAPTER_DEVICE_ID_OFFSET);
-	MTHCA_GET(adapter->revision_id, outbox, QUERY_ADAPTER_REVISION_ID_OFFSET);
+	if (!mthca_is_memfree(dev)) {
+		MTHCA_GET(adapter->vendor_id, outbox,
+			  QUERY_ADAPTER_VENDOR_ID_OFFSET);
+		MTHCA_GET(adapter->device_id, outbox,
+			  QUERY_ADAPTER_DEVICE_ID_OFFSET);
+		MTHCA_GET(adapter->revision_id, outbox,
+			  QUERY_ADAPTER_REVISION_ID_OFFSET);
+	}
 	MTHCA_GET(adapter->inta_pin, outbox,    QUERY_ADAPTER_INTA_PIN_OFFSET);
 
 	get_board_id(outbox + QUERY_ADAPTER_VSD_OFFSET / 4,

drivers/infiniband/hw/mthca/mthca_main.c

@@ -126,7 +126,7 @@ module_param_named(fmr_reserved_mtts, hca_profile.fmr_reserved_mtts, int, 0444);
 MODULE_PARM_DESC(fmr_reserved_mtts,
 		 "number of memory translation table segments reserved for FMR");
 
-static const char mthca_version[] __devinitdata =
+static char mthca_version[] __devinitdata =
 	DRV_NAME ": Mellanox InfiniBand HCA driver v"
 	DRV_VERSION " (" DRV_RELDATE ")\n";
 
@@ -735,7 +735,8 @@ static int mthca_init_hca(struct mthca_dev *mdev)
 	}
 
 	mdev->eq_table.inta_pin = adapter.inta_pin;
-	mdev->rev_id            = adapter.revision_id;
+	if (!mthca_is_memfree(mdev))
+		mdev->rev_id = adapter.revision_id;
 	memcpy(mdev->board_id, adapter.board_id, sizeof mdev->board_id);
 
 	return 0;

drivers/infiniband/hw/mthca/mthca_mr.c

@@ -613,8 +613,10 @@ int mthca_fmr_alloc(struct mthca_dev *dev, u32 pd,
 			sizeof *(mr->mem.tavor.mpt) * idx;
 
 	mr->mtt = __mthca_alloc_mtt(dev, list_len, dev->mr_table.fmr_mtt_buddy);
-	if (IS_ERR(mr->mtt))
+	if (IS_ERR(mr->mtt)) {
+		err = PTR_ERR(mr->mtt);
 		goto err_out_table;
+	}
 
 	mtt_seg = mr->mtt->first_seg * MTHCA_MTT_SEG_SIZE;
 
@@ -627,8 +629,10 @@ int mthca_fmr_alloc(struct mthca_dev *dev, u32 pd,
 		mr->mem.tavor.mtts = dev->mr_table.tavor_fmr.mtt_base + mtt_seg;
 
 	mailbox = mthca_alloc_mailbox(dev, GFP_KERNEL);
-	if (IS_ERR(mailbox))
+	if (IS_ERR(mailbox)) {
+		err = PTR_ERR(mailbox);
 		goto err_out_free_mtt;
+	}
 
 	mpt_entry = mailbox->buf;
 

drivers/infiniband/hw/mthca/mthca_provider.c

@@ -923,17 +923,13 @@ static struct ib_mr *mthca_reg_phys_mr(struct ib_pd       *pd,
 	struct mthca_mr *mr;
 	u64 *page_list;
 	u64 total_size;
-	u64 mask;
+	unsigned long mask;
 	int shift;
 	int npages;
 	int err;
 	int i, j, n;
 
-	/* First check that we have enough alignment */
-	if ((*iova_start & ~PAGE_MASK) != (buffer_list[0].addr & ~PAGE_MASK))
-		return ERR_PTR(-EINVAL);
-
-	mask = 0;
+	mask = buffer_list[0].addr ^ *iova_start;
 	total_size = 0;
 	for (i = 0; i < num_phys_buf; ++i) {
 		if (i != 0)
@@ -947,17 +943,7 @@ static struct ib_mr *mthca_reg_phys_mr(struct ib_pd       *pd,
 	if (mask & ~PAGE_MASK)
 		return ERR_PTR(-EINVAL);
 
-	/* Find largest page shift we can use to cover buffers */
-	for (shift = PAGE_SHIFT; shift < 31; ++shift)
-		if (num_phys_buf > 1) {
-			if ((1ULL << shift) & mask)
-				break;
-		} else {
-			if (1ULL << shift >=
-			    buffer_list[0].size +
-			    (buffer_list[0].addr & ((1ULL << shift) - 1)))
-				break;
-		}
+	shift = __ffs(mask | 1 << 31);
 
 	buffer_list[0].size += buffer_list[0].addr & ((1ULL << shift) - 1);
 	buffer_list[0].addr &= ~0ull << shift;
@@ -1270,6 +1256,8 @@ static int mthca_init_node_data(struct mthca_dev *dev)
 		goto out;
 	}
 
+	if (mthca_is_memfree(dev))
+		dev->rev_id = be32_to_cpup((__be32 *) (out_mad->data + 32));
 	memcpy(&dev->ib_dev.node_guid, out_mad->data + 12, 8);
 
 out:

drivers/infiniband/hw/mthca/mthca_qp.c

@@ -1175,6 +1175,7 @@ static int mthca_alloc_qp_common(struct mthca_dev *dev,
 {
 	int ret;
 	int i;
+	struct mthca_next_seg *next;
 
 	qp->refcount = 1;
 	init_waitqueue_head(&qp->wait);
@@ -1217,7 +1218,6 @@ static int mthca_alloc_qp_common(struct mthca_dev *dev,
 	}
 
 	if (mthca_is_memfree(dev)) {
-		struct mthca_next_seg *next;
 		struct mthca_data_seg *scatter;
 		int size = (sizeof (struct mthca_next_seg) +
 			    qp->rq.max_gs * sizeof (struct mthca_data_seg)) / 16;
@@ -1240,6 +1240,13 @@ static int mthca_alloc_qp_common(struct mthca_dev *dev,
 						    qp->sq.wqe_shift) +
 						   qp->send_wqe_offset);
 		}
+	} else {
+		for (i = 0; i < qp->rq.max; ++i) {
+			next = get_recv_wqe(qp, i);
+			next->nda_op = htonl((((i + 1) % qp->rq.max) <<
+					      qp->rq.wqe_shift) | 1);
+		}
+
 	}
 
 	qp->sq.last = get_send_wqe(qp, qp->sq.max - 1);
@@ -1863,7 +1870,6 @@ int mthca_tavor_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr,
 		prev_wqe = qp->rq.last;
 		qp->rq.last = wqe;
 
-		((struct mthca_next_seg *) wqe)->nda_op = 0;
 		((struct mthca_next_seg *) wqe)->ee_nds =
 			cpu_to_be32(MTHCA_NEXT_DBD);
 		((struct mthca_next_seg *) wqe)->flags = 0;
@@ -1885,9 +1891,6 @@ int mthca_tavor_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr,
 
 		qp->wrid[ind] = wr->wr_id;
 
-		((struct mthca_next_seg *) prev_wqe)->nda_op =
-			cpu_to_be32((ind << qp->rq.wqe_shift) | 1);
-		wmb();
 		((struct mthca_next_seg *) prev_wqe)->ee_nds =
 			cpu_to_be32(MTHCA_NEXT_DBD | size);
 

drivers/infiniband/hw/mthca/mthca_srq.c

@@ -175,9 +175,17 @@ static int mthca_alloc_srq_buf(struct mthca_dev *dev, struct mthca_pd *pd,
 	 * scatter list L_Keys to the sentry value of 0x100.
 	 */
 	for (i = 0; i < srq->max; ++i) {
-		wqe = get_wqe(srq, i);
+		struct mthca_next_seg *next;
 
-		*wqe_to_link(wqe) = i < srq->max - 1 ? i + 1 : -1;
+		next = wqe = get_wqe(srq, i);
+
+		if (i < srq->max - 1) {
+			*wqe_to_link(wqe) = i + 1;
+			next->nda_op = htonl(((i + 1) << srq->wqe_shift) | 1);
+		} else {
+			*wqe_to_link(wqe) = -1;
+			next->nda_op = 0;
+		}
 
 		for (scatter = wqe + sizeof (struct mthca_next_seg);
 		     (void *) scatter < wqe + (1 << srq->wqe_shift);
@@ -470,16 +478,15 @@ void mthca_srq_event(struct mthca_dev *dev, u32 srqn,
 void mthca_free_srq_wqe(struct mthca_srq *srq, u32 wqe_addr)
 {
 	int ind;
+	struct mthca_next_seg *last_free;
 
 	ind = wqe_addr >> srq->wqe_shift;
 
 	spin_lock(&srq->lock);
 
-	if (likely(srq->first_free >= 0))
-		*wqe_to_link(get_wqe(srq, srq->last_free)) = ind;
-	else
-		srq->first_free = ind;
-
+	last_free = get_wqe(srq, srq->last_free);
+	*wqe_to_link(last_free) = ind;
+	last_free->nda_op = htonl((ind << srq->wqe_shift) | 1);
 	*wqe_to_link(get_wqe(srq, ind)) = -1;
 	srq->last_free = ind;
 
@@ -506,15 +513,7 @@ int mthca_tavor_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr,
 	first_ind = srq->first_free;
 
 	for (nreq = 0; wr; wr = wr->next) {
-		ind = srq->first_free;
-
-		if (unlikely(ind < 0)) {
-			mthca_err(dev, "SRQ %06x full\n", srq->srqn);
-			err = -ENOMEM;
-			*bad_wr = wr;
-			break;
-		}
-
+		ind       = srq->first_free;
 		wqe       = get_wqe(srq, ind);
 		next_ind  = *wqe_to_link(wqe);
 
@@ -528,7 +527,6 @@ int mthca_tavor_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr,
 		prev_wqe  = srq->last;
 		srq->last = wqe;
 
-		((struct mthca_next_seg *) wqe)->nda_op = 0;
 		((struct mthca_next_seg *) wqe)->ee_nds = 0;
 		/* flags field will always remain 0 */
 
@@ -549,9 +547,6 @@ int mthca_tavor_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr,
 		if (i < srq->max_gs)
 			mthca_set_data_seg_inval(wqe);
 
-		((struct mthca_next_seg *) prev_wqe)->nda_op =
-			cpu_to_be32((ind << srq->wqe_shift) | 1);
-		wmb();
 		((struct mthca_next_seg *) prev_wqe)->ee_nds =
 			cpu_to_be32(MTHCA_NEXT_DBD);
 
@@ -614,15 +609,7 @@ int mthca_arbel_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr,
 	spin_lock_irqsave(&srq->lock, flags);
 
 	for (nreq = 0; wr; ++nreq, wr = wr->next) {
-		ind = srq->first_free;
-
-		if (unlikely(ind < 0)) {
-			mthca_err(dev, "SRQ %06x full\n", srq->srqn);
-			err = -ENOMEM;
-			*bad_wr = wr;
-			break;
-		}
-
+		ind       = srq->first_free;
 		wqe       = get_wqe(srq, ind);
 		next_ind  = *wqe_to_link(wqe);
 
@@ -633,8 +620,6 @@ int mthca_arbel_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr,
 			break;
 		}
 
-		((struct mthca_next_seg *) wqe)->nda_op =
-			cpu_to_be32((next_ind << srq->wqe_shift) | 1);
 		((struct mthca_next_seg *) wqe)->ee_nds = 0;
 		/* flags field will always remain 0 */
 

drivers/infiniband/hw/nes/Kconfig

+config INFINIBAND_NES
+	tristate "NetEffect RNIC Driver"
+	depends on PCI && INET && INFINIBAND
+	select LIBCRC32C
+	---help---
+	  This is a low-level driver for NetEffect RDMA enabled
+	  Network Interface Cards (RNIC).
+
+config INFINIBAND_NES_DEBUG
+	bool "Verbose debugging output"
+	depends on INFINIBAND_NES
+	default n
+	---help---
+	  This option causes the NetEffect RNIC driver to produce debug
+	  messages.  Select this if you are developing the driver
+	  or trying to diagnose a problem.

drivers/infiniband/hw/nes/Makefile

+obj-$(CONFIG_INFINIBAND_NES) += iw_nes.o
+
+iw_nes-objs := nes.o nes_hw.o nes_nic.o nes_utils.o nes_verbs.o nes_cm.o

drivers/infiniband/hw/nes/nes_context.h

+/*
+ * Copyright (c) 2006 - 2008 NetEffect, 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 NES_CONTEXT_H
+#define NES_CONTEXT_H
+
+struct nes_qp_context {
+	__le32   misc;
+	__le32   cqs;
+	__le32   sq_addr_low;
+	__le32   sq_addr_high;
+	__le32   rq_addr_low;
+	__le32   rq_addr_high;
+	__le32   misc2;
+	__le16   tcpPorts[2];
+	__le32   ip0;
+	__le32   ip1;
+	__le32   ip2;
+	__le32   ip3;
+	__le32   mss;
+	__le32   arp_index_vlan;
+	__le32   tcp_state_flow_label;
+	__le32   pd_index_wscale;
+	__le32   keepalive;
+	u32   ts_recent;
+	u32   ts_age;
+	__le32   snd_nxt;
+	__le32   snd_wnd;
+	__le32   rcv_nxt;
+	__le32   rcv_wnd;
+	__le32   snd_max;
+	__le32   snd_una;
+	u32   srtt;
+	__le32   rttvar;
+	__le32   ssthresh;
+	__le32   cwnd;
+	__le32   snd_wl1;
+	__le32   snd_wl2;
+	__le32   max_snd_wnd;
+	__le32   ts_val_delta;
+	u32   retransmit;
+	u32   probe_cnt;
+	u32   hte_index;
+	__le32   q2_addr_low;
+	__le32   q2_addr_high;
+	__le32   ird_index;
+	u32   Rsvd3;
+	__le32   ird_ord_sizes;
+	u32   mrkr_offset;
+	__le32   aeq_token_low;
+	__le32   aeq_token_high;
+};
+
+/* QP Context Misc Field */
+
+#define NES_QPCONTEXT_MISC_IWARP_VER_MASK    0x00000003
+#define NES_QPCONTEXT_MISC_IWARP_VER_SHIFT   0
+#define NES_QPCONTEXT_MISC_EFB_SIZE_MASK     0x000000C0
+#define NES_QPCONTEXT_MISC_EFB_SIZE_SHIFT    6
+#define NES_QPCONTEXT_MISC_RQ_SIZE_MASK      0x00000300
+#define NES_QPCONTEXT_MISC_RQ_SIZE_SHIFT     8
+#define NES_QPCONTEXT_MISC_SQ_SIZE_MASK      0x00000c00
+#define NES_QPCONTEXT_MISC_SQ_SIZE_SHIFT     10
+#define NES_QPCONTEXT_MISC_PCI_FCN_MASK      0x00007000
+#define NES_QPCONTEXT_MISC_PCI_FCN_SHIFT     12
+#define NES_QPCONTEXT_MISC_DUP_ACKS_MASK     0x00070000
+#define NES_QPCONTEXT_MISC_DUP_ACKS_SHIFT    16
+
+enum nes_qp_context_misc_bits {
+	NES_QPCONTEXT_MISC_RX_WQE_SIZE         = 0x00000004,
+	NES_QPCONTEXT_MISC_IPV4                = 0x00000008,
+	NES_QPCONTEXT_MISC_DO_NOT_FRAG         = 0x00000010,
+	NES_QPCONTEXT_MISC_INSERT_VLAN         = 0x00000020,
+	NES_QPCONTEXT_MISC_DROS                = 0x00008000,
+	NES_QPCONTEXT_MISC_WSCALE              = 0x00080000,
+	NES_QPCONTEXT_MISC_KEEPALIVE           = 0x00100000,
+	NES_QPCONTEXT_MISC_TIMESTAMP           = 0x00200000,
+	NES_QPCONTEXT_MISC_SACK                = 0x00400000,
+	NES_QPCONTEXT_MISC_RDMA_WRITE_EN       = 0x00800000,
+	NES_QPCONTEXT_MISC_RDMA_READ_EN        = 0x01000000,
+	NES_QPCONTEXT_MISC_WBIND_EN            = 0x10000000,
+	NES_QPCONTEXT_MISC_FAST_REGISTER_EN    = 0x20000000,
+	NES_QPCONTEXT_MISC_PRIV_EN             = 0x40000000,
+	NES_QPCONTEXT_MISC_NO_NAGLE            = 0x80000000
+};
+
+enum nes_qp_acc_wq_sizes {
+	HCONTEXT_TSA_WQ_SIZE_4 = 0,
+	HCONTEXT_TSA_WQ_SIZE_32 = 1,
+	HCONTEXT_TSA_WQ_SIZE_128 = 2,
+	HCONTEXT_TSA_WQ_SIZE_512 = 3
+};
+
+/* QP Context Misc2 Fields */
+#define NES_QPCONTEXT_MISC2_TTL_MASK            0x000000ff
+#define NES_QPCONTEXT_MISC2_TTL_SHIFT           0
+#define NES_QPCONTEXT_MISC2_HOP_LIMIT_MASK      0x000000ff
+#define NES_QPCONTEXT_MISC2_HOP_LIMIT_SHIFT     0
+#define NES_QPCONTEXT_MISC2_LIMIT_MASK          0x00000300
+#define NES_QPCONTEXT_MISC2_LIMIT_SHIFT         8
+#define NES_QPCONTEXT_MISC2_NIC_INDEX_MASK      0x0000fc00
+#define NES_QPCONTEXT_MISC2_NIC_INDEX_SHIFT     10
+#define NES_QPCONTEXT_MISC2_SRC_IP_MASK         0x001f0000
+#define NES_QPCONTEXT_MISC2_SRC_IP_SHIFT        16
+#define NES_QPCONTEXT_MISC2_TOS_MASK            0xff000000
+#define NES_QPCONTEXT_MISC2_TOS_SHIFT           24
+#define NES_QPCONTEXT_MISC2_TRAFFIC_CLASS_MASK  0xff000000
+#define NES_QPCONTEXT_MISC2_TRAFFIC_CLASS_SHIFT 24
+
+/* QP Context Tcp State/Flow Label Fields */
+#define NES_QPCONTEXT_TCPFLOW_FLOW_LABEL_MASK   0x000fffff
+#define NES_QPCONTEXT_TCPFLOW_FLOW_LABEL_SHIFT  0
+#define NES_QPCONTEXT_TCPFLOW_TCP_STATE_MASK    0xf0000000
+#define NES_QPCONTEXT_TCPFLOW_TCP_STATE_SHIFT   28
+
+enum nes_qp_tcp_state {
+	NES_QPCONTEXT_TCPSTATE_CLOSED = 1,
+	NES_QPCONTEXT_TCPSTATE_EST = 5,
+	NES_QPCONTEXT_TCPSTATE_TIME_WAIT = 11,
+};
+
+/* QP Context PD Index/wscale Fields */
+#define NES_QPCONTEXT_PDWSCALE_RCV_WSCALE_MASK  0x0000000f
+#define NES_QPCONTEXT_PDWSCALE_RCV_WSCALE_SHIFT 0
+#define NES_QPCONTEXT_PDWSCALE_SND_WSCALE_MASK  0x00000f00
+#define NES_QPCONTEXT_PDWSCALE_SND_WSCALE_SHIFT 8
+#define NES_QPCONTEXT_PDWSCALE_PDINDEX_MASK     0xffff0000
+#define NES_QPCONTEXT_PDWSCALE_PDINDEX_SHIFT    16
+
+/* QP Context Keepalive Fields */
+#define NES_QPCONTEXT_KEEPALIVE_DELTA_MASK      0x0000ffff
+#define NES_QPCONTEXT_KEEPALIVE_DELTA_SHIFT     0
+#define NES_QPCONTEXT_KEEPALIVE_PROBE_CNT_MASK  0x00ff0000
+#define NES_QPCONTEXT_KEEPALIVE_PROBE_CNT_SHIFT 16
+#define NES_QPCONTEXT_KEEPALIVE_INTV_MASK       0xff000000
+#define NES_QPCONTEXT_KEEPALIVE_INTV_SHIFT      24
+
+/* QP Context ORD/IRD Fields */
+#define NES_QPCONTEXT_ORDIRD_ORDSIZE_MASK       0x0000007f
+#define NES_QPCONTEXT_ORDIRD_ORDSIZE_SHIFT      0
+#define NES_QPCONTEXT_ORDIRD_IRDSIZE_MASK       0x00030000
+#define NES_QPCONTEXT_ORDIRD_IRDSIZE_SHIFT      16
+#define NES_QPCONTEXT_ORDIRD_IWARP_MODE_MASK    0x30000000
+#define NES_QPCONTEXT_ORDIRD_IWARP_MODE_SHIFT   28
+
+enum nes_ord_ird_bits {
+	NES_QPCONTEXT_ORDIRD_WRPDU                   = 0x02000000,
+	NES_QPCONTEXT_ORDIRD_LSMM_PRESENT            = 0x04000000,
+	NES_QPCONTEXT_ORDIRD_ALSMM                   = 0x08000000,
+	NES_QPCONTEXT_ORDIRD_AAH                     = 0x40000000,
+	NES_QPCONTEXT_ORDIRD_RNMC                    = 0x80000000
+};
+
+enum nes_iwarp_qp_state {
+	NES_QPCONTEXT_IWARP_STATE_NONEXIST  = 0,
+	NES_QPCONTEXT_IWARP_STATE_IDLE      = 1,
+	NES_QPCONTEXT_IWARP_STATE_RTS       = 2,
+	NES_QPCONTEXT_IWARP_STATE_CLOSING   = 3,
+	NES_QPCONTEXT_IWARP_STATE_TERMINATE = 5,
+	NES_QPCONTEXT_IWARP_STATE_ERROR     = 6
+};
+
+
+#endif		/* NES_CONTEXT_H */

drivers/infiniband/hw/nes/nes_user.h

+/*
+ * Copyright (c) 2006 - 2008 NetEffect.  All rights reserved.
+ * Copyright (c) 2005 Topspin Communications.  All rights reserved.
+ * Copyright (c) 2005 Cisco Systems.  All rights reserved.
+ * Copyright (c) 2005 Open Grid Computing, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+
+#ifndef NES_USER_H
+#define NES_USER_H
+
+#include <linux/types.h>
+
+#define NES_ABI_USERSPACE_VER 1
+#define NES_ABI_KERNEL_VER    1
+
+/*
+ * Make sure that all structs defined in this file remain laid out so
+ * that they pack the same way on 32-bit and 64-bit architectures (to
+ * avoid incompatibility between 32-bit userspace and 64-bit kernels).
+ * In particular do not use pointer types -- pass pointers in __u64
+ * instead.
+ */
+
+struct nes_alloc_ucontext_req {
+	__u32 reserved32;
+	__u8  userspace_ver;
+	__u8  reserved8[3];
+};
+
+struct nes_alloc_ucontext_resp {
+	__u32 max_pds; /* maximum pds allowed for this user process */
+	__u32 max_qps; /* maximum qps allowed for this user process */
+	__u32 wq_size; /* size of the WQs (sq+rq) allocated to the mmaped area */
+	__u8  virtwq;  /* flag to indicate if virtual WQ are to be used or not */
+	__u8  kernel_ver;
+	__u8  reserved[2];
+};
+
+struct nes_alloc_pd_resp {
+	__u32 pd_id;
+	__u32 mmap_db_index;
+};
+
+struct nes_create_cq_req {
+	__u64 user_cq_buffer;
+	__u32 mcrqf;
+	__u8 reserved[4];
+};
+
+struct nes_create_qp_req {
+	__u64 user_wqe_buffers;
+};
+
+enum iwnes_memreg_type {
+	IWNES_MEMREG_TYPE_MEM = 0x0000,
+	IWNES_MEMREG_TYPE_QP = 0x0001,
+	IWNES_MEMREG_TYPE_CQ = 0x0002,
+	IWNES_MEMREG_TYPE_MW = 0x0003,
+	IWNES_MEMREG_TYPE_FMR = 0x0004,
+};
+
+struct nes_mem_reg_req {
+	__u32 reg_type;	/* indicates if id is memory, QP or CQ */
+	__u32 reserved;
+};
+
+struct nes_create_cq_resp {
+	__u32 cq_id;
+	__u32 cq_size;
+	__u32 mmap_db_index;
+	__u32 reserved;
+};
+
+struct nes_create_qp_resp {
+	__u32 qp_id;
+	__u32 actual_sq_size;
+	__u32 actual_rq_size;
+	__u32 mmap_sq_db_index;
+	__u32 mmap_rq_db_index;
+	__u32 nes_drv_opt;
+};
+
+#endif				/* NES_USER_H */

drivers/infiniband/hw/nes/nes_verbs.h

+/*
+ * Copyright (c) 2006 - 2008 NetEffect, Inc. All rights reserved.
+ * Copyright (c) 2005 Open Grid Computing, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+
+#ifndef NES_VERBS_H
+#define NES_VERBS_H
+
+struct nes_device;
+
+#define NES_MAX_USER_DB_REGIONS  4096
+#define NES_MAX_USER_WQ_REGIONS  4096
+
+struct nes_ucontext {
+	struct ib_ucontext ibucontext;
+	struct nes_device  *nesdev;
+	unsigned long      mmap_wq_offset;
+	unsigned long      mmap_cq_offset; /* to be removed */
+	int                index;		/* rnic index (minor) */
+	unsigned long      allocated_doorbells[BITS_TO_LONGS(NES_MAX_USER_DB_REGIONS)];
+	u16                mmap_db_index[NES_MAX_USER_DB_REGIONS];
+	u16                first_free_db;
+	unsigned long      allocated_wqs[BITS_TO_LONGS(NES_MAX_USER_WQ_REGIONS)];
+	struct nes_qp      *mmap_nesqp[NES_MAX_USER_WQ_REGIONS];
+	u16                first_free_wq;
+	struct list_head   cq_reg_mem_list;
+	struct list_head   qp_reg_mem_list;
+	u32                mcrqf;
+	atomic_t	   usecnt;
+};
+
+struct nes_pd {
+	struct ib_pd ibpd;
+	u16          pd_id;
+	atomic_t     sqp_count;
+	u16          mmap_db_index;
+};
+
+struct nes_mr {
+	union {
+		struct ib_mr  ibmr;
+		struct ib_mw  ibmw;
+		struct ib_fmr ibfmr;
+	};
+	struct ib_umem    *region;
+	u16               pbls_used;
+	u8                mode;
+	u8                pbl_4k;
+};
+
+struct nes_hw_pb {
+	__le32 pa_low;
+	__le32 pa_high;
+};
+
+struct nes_vpbl {
+	dma_addr_t       pbl_pbase;
+	struct nes_hw_pb *pbl_vbase;
+};
+
+struct nes_root_vpbl {
+	dma_addr_t       pbl_pbase;
+	struct nes_hw_pb *pbl_vbase;
+	struct nes_vpbl  *leaf_vpbl;
+};
+
+struct nes_fmr {
+	struct nes_mr        nesmr;
+	u32                  leaf_pbl_cnt;
+	struct nes_root_vpbl root_vpbl;
+	struct ib_qp         *ib_qp;
+	int                  access_rights;
+	struct ib_fmr_attr   attr;
+};
+
+struct nes_av;
+
+struct nes_cq {
+	struct ib_cq     ibcq;
+	struct nes_hw_cq hw_cq;
+	u32              polled_completions;
+	u32              cq_mem_size;
+	spinlock_t       lock;
+	u8               virtual_cq;
+	u8               pad[3];
+};
+
+struct nes_wq {
+	spinlock_t lock;
+};
+
+struct iw_cm_id;
+struct ietf_mpa_frame;
+
+struct nes_qp {
+	struct ib_qp          ibqp;
+	void                  *allocated_buffer;
+	struct iw_cm_id       *cm_id;
+	struct workqueue_struct *wq;
+	struct work_struct    disconn_work;
+	struct nes_cq         *nesscq;
+	struct nes_cq         *nesrcq;
+	struct nes_pd         *nespd;
+	void *cm_node; /* handle of the node this QP is associated with */
+	struct ietf_mpa_frame *ietf_frame;
+	dma_addr_t            ietf_frame_pbase;
+	wait_queue_head_t     state_waitq;
+	unsigned long         socket;
+	struct nes_hw_qp      hwqp;
+	struct work_struct    work;
+	struct work_struct    ae_work;
+	enum ib_qp_state      ibqp_state;
+	u32                   iwarp_state;
+	u32                   hte_index;
+	u32                   last_aeq;
+	u32                   qp_mem_size;
+	atomic_t              refcount;
+	atomic_t              close_timer_started;
+	u32                   mmap_sq_db_index;
+	u32                   mmap_rq_db_index;
+	spinlock_t            lock;
+	struct nes_qp_context *nesqp_context;
+	dma_addr_t            nesqp_context_pbase;
+	void	              *pbl_vbase;
+	dma_addr_t            pbl_pbase;
+	struct page           *page;
+	wait_queue_head_t     kick_waitq;
+	u16                   in_disconnect;
+	u16                   private_data_len;
+	u8                    active_conn;
+	u8                    skip_lsmm;
+	u8                    user_mode;
+	u8                    hte_added;
+	u8                    hw_iwarp_state;
+	u8                    flush_issued;
+	u8                    hw_tcp_state;
+	u8                    disconn_pending;
+	u8                    destroyed;
+};
+#endif			/* NES_VERBS_H */

drivers/infiniband/ulp/ipoib/ipoib_main.c

@@ -680,12 +680,7 @@ static int ipoib_start_xmit(struct sk_buff *skb, struct net_device *dev)
 
 		neigh = *to_ipoib_neigh(skb->dst->neighbour);
 
-		if (ipoib_cm_get(neigh)) {
-			if (ipoib_cm_up(neigh)) {
-				ipoib_cm_send(dev, skb, ipoib_cm_get(neigh));
-				goto out;
-			}
-		} else if (neigh->ah) {
+		if (neigh->ah)
 			if (unlikely((memcmp(&neigh->dgid.raw,
 					    skb->dst->neighbour->ha + 4,
 					    sizeof(union ib_gid))) ||
@@ -706,6 +701,12 @@ static int ipoib_start_xmit(struct sk_buff *skb, struct net_device *dev)
 				goto out;
 			}
 
+		if (ipoib_cm_get(neigh)) {
+			if (ipoib_cm_up(neigh)) {
+				ipoib_cm_send(dev, skb, ipoib_cm_get(neigh));
+				goto out;
+			}
+		} else if (neigh->ah) {
 			ipoib_send(dev, skb, neigh->ah, IPOIB_QPN(skb->dst->neighbour->ha));
 			goto out;
 		}
@@ -813,11 +814,9 @@ static void ipoib_neigh_cleanup(struct neighbour *n)
 	struct ipoib_ah *ah = NULL;
 
 	neigh = *to_ipoib_neigh(n);
-	if (neigh) {
+	if (neigh)
 		priv = netdev_priv(neigh->dev);
-		ipoib_dbg(priv, "neigh_destructor for bonding device: %s\n",
-			  n->dev->name);
-	} else
+	else
 		return;
 	ipoib_dbg(priv,
 		  "neigh_cleanup for %06x " IPOIB_GID_FMT "\n",

drivers/infiniband/ulp/srp/ib_srp.c

@@ -204,6 +204,22 @@ static int srp_init_qp(struct srp_target_port *target,
 	return ret;
 }
 
+static int srp_new_cm_id(struct srp_target_port *target)
+{
+	struct ib_cm_id *new_cm_id;
+
+	new_cm_id = ib_create_cm_id(target->srp_host->dev->dev,
+				    srp_cm_handler, target);
+	if (IS_ERR(new_cm_id))
+		return PTR_ERR(new_cm_id);
+
+	if (target->cm_id)
+		ib_destroy_cm_id(target->cm_id);
+	target->cm_id = new_cm_id;
+
+	return 0;
+}
+
 static int srp_create_target_ib(struct srp_target_port *target)
 {
 	struct ib_qp_init_attr *init_attr;
@@ -436,6 +452,7 @@ static void srp_remove_work(struct work_struct *work)
 
 static int srp_connect_target(struct srp_target_port *target)
 {
+	int retries = 3;
 	int ret;
 
 	ret = srp_lookup_path(target);
@@ -468,6 +485,21 @@ static int srp_connect_target(struct srp_target_port *target)
 		case SRP_DLID_REDIRECT:
 			break;
 
+		case SRP_STALE_CONN:
+			/* Our current CM id was stale, and is now in timewait.
+			 * Try to reconnect with a new one.
+			 */
+			if (!retries-- || srp_new_cm_id(target)) {
+				shost_printk(KERN_ERR, target->scsi_host, PFX
+					     "giving up on stale connection\n");
+				target->status = -ECONNRESET;
+				return target->status;
+			}
+
+			shost_printk(KERN_ERR, target->scsi_host, PFX
+				     "retrying stale connection\n");
+			break;
+
 		default:
 			return target->status;
 		}
@@ -507,7 +539,6 @@ static void srp_reset_req(struct srp_target_port *target, struct srp_request *re
 
 static int srp_reconnect_target(struct srp_target_port *target)
 {
-	struct ib_cm_id *new_cm_id;
 	struct ib_qp_attr qp_attr;
 	struct srp_request *req, *tmp;
 	struct ib_wc wc;
@@ -526,14 +557,9 @@ static int srp_reconnect_target(struct srp_target_port *target)
 	 * Now get a new local CM ID so that we avoid confusing the
 	 * target in case things are really fouled up.
 	 */
-	new_cm_id = ib_create_cm_id(target->srp_host->dev->dev,
-				    srp_cm_handler, target);
-	if (IS_ERR(new_cm_id)) {
-		ret = PTR_ERR(new_cm_id);
+	ret = srp_new_cm_id(target);
+	if (ret)
 		goto err;
-	}
-	ib_destroy_cm_id(target->cm_id);
-	target->cm_id = new_cm_id;
 
 	qp_attr.qp_state = IB_QPS_RESET;
 	ret = ib_modify_qp(target->qp, &qp_attr, IB_QP_STATE);
@@ -1171,6 +1197,11 @@ static void srp_cm_rej_handler(struct ib_cm_id *cm_id,
 		target->status = -ECONNRESET;
 		break;
 
+	case IB_CM_REJ_STALE_CONN:
+		shost_printk(KERN_WARNING, shost, "  REJ reason: stale connection\n");
+		target->status = SRP_STALE_CONN;
+		break;
+
 	default:
 		shost_printk(KERN_WARNING, shost, "  REJ reason 0x%x\n",
 			     event->param.rej_rcvd.reason);
@@ -1862,11 +1893,9 @@ static ssize_t srp_create_target(struct class_device *class_dev,
 	if (ret)
 		goto err;
 
-	target->cm_id = ib_create_cm_id(host->dev->dev, srp_cm_handler, target);
-	if (IS_ERR(target->cm_id)) {
-		ret = PTR_ERR(target->cm_id);
+	ret = srp_new_cm_id(target);
+	if (ret)
 		goto err_free;
-	}
 
 	target->qp_in_error = 0;
 	ret = srp_connect_target(target);

drivers/infiniband/ulp/srp/ib_srp.h

@@ -54,6 +54,7 @@ enum {
 
 	SRP_PORT_REDIRECT	= 1,
 	SRP_DLID_REDIRECT	= 2,
+	SRP_STALE_CONN		= 3,
 
 	SRP_MAX_LUN		= 512,
 	SRP_DEF_SG_TABLESIZE	= 12,

drivers/net/mlx4/fw.c

@@ -617,9 +617,6 @@ int mlx4_QUERY_ADAPTER(struct mlx4_dev *dev, struct mlx4_adapter *adapter)
 	int err;
 
 #define QUERY_ADAPTER_OUT_SIZE             0x100
-#define QUERY_ADAPTER_VENDOR_ID_OFFSET     0x00
-#define QUERY_ADAPTER_DEVICE_ID_OFFSET     0x04
-#define QUERY_ADAPTER_REVISION_ID_OFFSET   0x08
 #define QUERY_ADAPTER_INTA_PIN_OFFSET      0x10
 #define QUERY_ADAPTER_VSD_OFFSET           0x20
 
@@ -633,9 +630,6 @@ int mlx4_QUERY_ADAPTER(struct mlx4_dev *dev, struct mlx4_adapter *adapter)
 	if (err)
 		goto out;
 
-	MLX4_GET(adapter->vendor_id, outbox,   QUERY_ADAPTER_VENDOR_ID_OFFSET);
-	MLX4_GET(adapter->device_id, outbox,   QUERY_ADAPTER_DEVICE_ID_OFFSET);
-	MLX4_GET(adapter->revision_id, outbox, QUERY_ADAPTER_REVISION_ID_OFFSET);
 	MLX4_GET(adapter->inta_pin, outbox,    QUERY_ADAPTER_INTA_PIN_OFFSET);
 
 	get_board_id(outbox + QUERY_ADAPTER_VSD_OFFSET / 4,

drivers/net/mlx4/fw.h

@@ -99,9 +99,6 @@ struct mlx4_dev_cap {
 };
 
 struct mlx4_adapter {
-	u32  vendor_id;
-	u32  device_id;
-	u32  revision_id;
 	char board_id[MLX4_BOARD_ID_LEN];
 	u8   inta_pin;
 };

drivers/net/mlx4/main.c

@@ -71,7 +71,7 @@ MODULE_PARM_DESC(msi_x, "attempt to use MSI-X if nonzero");
 
 #endif /* CONFIG_PCI_MSI */
 
-static const char mlx4_version[] __devinitdata =
+static char mlx4_version[] __devinitdata =
 	DRV_NAME ": Mellanox ConnectX core driver v"
 	DRV_VERSION " (" DRV_RELDATE ")\n";
 
@@ -163,7 +163,7 @@ static int mlx4_dev_cap(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap)
 	return 0;
 }
 
-static int __devinit mlx4_load_fw(struct mlx4_dev *dev)
+static int mlx4_load_fw(struct mlx4_dev *dev)
 {
 	struct mlx4_priv *priv = mlx4_priv(dev);
 	int err;
@@ -197,8 +197,8 @@ static int __devinit mlx4_load_fw(struct mlx4_dev *dev)
 	return err;
 }
 
-static int __devinit mlx4_init_cmpt_table(struct mlx4_dev *dev, u64 cmpt_base,
-					  int cmpt_entry_sz)
+static int mlx4_init_cmpt_table(struct mlx4_dev *dev, u64 cmpt_base,
+				int cmpt_entry_sz)
 {
 	struct mlx4_priv *priv = mlx4_priv(dev);
 	int err;
@@ -534,7 +534,6 @@ static int mlx4_init_hca(struct mlx4_dev *dev)
 	}
 
 	priv->eq_table.inta_pin = adapter.inta_pin;
-	dev->rev_id		= adapter.revision_id;
 	memcpy(dev->board_id, adapter.board_id, sizeof dev->board_id);
 
 	return 0;
@@ -688,7 +687,7 @@ static int mlx4_setup_hca(struct mlx4_dev *dev)
 	return err;
 }
 
-static void __devinit mlx4_enable_msi_x(struct mlx4_dev *dev)
+static void mlx4_enable_msi_x(struct mlx4_dev *dev)
 {
 	struct mlx4_priv *priv = mlx4_priv(dev);
 	struct msix_entry entries[MLX4_NUM_EQ];

drivers/net/mlx4/mr.c

@@ -122,7 +122,7 @@ static void mlx4_buddy_free(struct mlx4_buddy *buddy, u32 seg, int order)
 	spin_unlock(&buddy->lock);
 }
 
-static int __devinit mlx4_buddy_init(struct mlx4_buddy *buddy, int max_order)
+static int mlx4_buddy_init(struct mlx4_buddy *buddy, int max_order)
 {
 	int i, s;