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: (81 commits) RDMA/cxgb3: Fix the T3A workaround checks IB/ipath: Remove unnecessary cast IPoIB: Constify seq_operations function pointer tables RDMA/cxgb3: Mark QP as privileged based on user capabilities RDMA/cxgb3: Fix page shift calculation in build_phys_page_list() RDMA/cxgb3: Flush the receive queue when closing IB/ipath: Trivial simplification of ipath_make_ud_req() IB/mthca: Update latest "native Arbel" firmware revision IPoIB: Remove redundant check of netif_queue_stopped() in xmit handler IB/ipath: Add mappings from HW register to PortInfo port physical state IB/ipath: Changes to support PIO bandwidth check on IBA7220 IB/ipath: Minor cleanup of unused fields and chip-specific errors IB/ipath: New sysfs entries to control 7220 features IB/ipath: Add new chip-specific functions to older chips, consistent init IB/ipath: Remove unused MDIO interface code IB/ehca: Prevent RDMA-related connection failures on some eHCA2 hardware IB/ehca: Add "port connection autodetect mode" IB/ehca: Define array to store SMI/GSI QPs IB/ehca: Remove CQ-QP-link before destroying QP in error path of create_qp() IB/iser: Add change_queue_depth method ...
2026-05-01 15:00:59 -07:00 · 2008-01-25 14:41:24 -08:00
parent b31fde6db2 8176d297c7
commit 99f1c97dbd
75 changed files with 3127 additions and 1188 deletions
@@ -295,16 +295,6 @@ Who:  linuxppc-dev@ozlabs.org

 ---------------------------

-What:	mthca driver's MSI support
-When:	January 2008
-Files:	drivers/infiniband/hw/mthca/*.[ch]
-Why:	All mthca hardware also supports MSI-X, which provides
-	strictly more functionality than MSI.  So there is no point in
-	having both MSI-X and MSI support in the driver.
-Who:	Roland Dreier <rolandd@cisco.com>
-
---------------------------
-
 What:   sk98lin network driver
 When:   Feburary 2008
 Why:    In kernel tree version of driver is unmaintained. Sk98lin driver
@@ -488,7 +488,8 @@ void rdma_destroy_qp(struct rdma_cm_id *id)
 }
 EXPORT_SYMBOL(rdma_destroy_qp);

-static int cma_modify_qp_rtr(struct rdma_id_private *id_priv)
+static int cma_modify_qp_rtr(struct rdma_id_private *id_priv,
+			     struct rdma_conn_param *conn_param)
 {
 	struct ib_qp_attr qp_attr;
 	int qp_attr_mask, ret;
@@ -514,13 +515,16 @@ static int cma_modify_qp_rtr(struct rdma_id_private *id_priv)
 	if (ret)
 		goto out;

+	if (conn_param)
+		qp_attr.max_dest_rd_atomic = conn_param->responder_resources;
 	ret = ib_modify_qp(id_priv->id.qp, &qp_attr, qp_attr_mask);
 out:
 	mutex_unlock(&id_priv->qp_mutex);
 	return ret;
 }

-static int cma_modify_qp_rts(struct rdma_id_private *id_priv)
+static int cma_modify_qp_rts(struct rdma_id_private *id_priv,
+			     struct rdma_conn_param *conn_param)
 {
 	struct ib_qp_attr qp_attr;
 	int qp_attr_mask, ret;
@@ -536,6 +540,8 @@ static int cma_modify_qp_rts(struct rdma_id_private *id_priv)
 	if (ret)
 		goto out;

+	if (conn_param)
+		qp_attr.max_rd_atomic = conn_param->initiator_depth;
 	ret = ib_modify_qp(id_priv->id.qp, &qp_attr, qp_attr_mask);
 out:
 	mutex_unlock(&id_priv->qp_mutex);
@@ -866,11 +872,11 @@ static int cma_rep_recv(struct rdma_id_private *id_priv)
 {
 	int ret;

-	ret = cma_modify_qp_rtr(id_priv);
+	ret = cma_modify_qp_rtr(id_priv, NULL);
 	if (ret)
 		goto reject;

-	ret = cma_modify_qp_rts(id_priv);
+	ret = cma_modify_qp_rts(id_priv, NULL);
 	if (ret)
 		goto reject;

@@ -1122,8 +1128,10 @@ static int cma_req_handler(struct ib_cm_id *cm_id, struct ib_cm_event *ib_event)
 	cm_id->cm_handler = cma_ib_handler;

 	ret = conn_id->id.event_handler(&conn_id->id, &event);
-	if (!ret)
+	if (!ret) {
+		cma_enable_remove(conn_id);
 		goto out;
+	}

 	/* Destroy the CM ID by returning a non-zero value. */
 	conn_id->cm_id.ib = NULL;
@@ -1262,6 +1270,7 @@ static int iw_conn_req_handler(struct iw_cm_id *cm_id,
 	struct net_device *dev = NULL;
 	struct rdma_cm_event event;
 	int ret;
+	struct ib_device_attr attr;

 	listen_id = cm_id->context;
 	if (cma_disable_remove(listen_id, CMA_LISTEN))
@@ -1311,10 +1320,19 @@ static int iw_conn_req_handler(struct iw_cm_id *cm_id,
 	sin = (struct sockaddr_in *) &new_cm_id->route.addr.dst_addr;
 	*sin = iw_event->remote_addr;

+	ret = ib_query_device(conn_id->id.device, &attr);
+	if (ret) {
+		cma_enable_remove(conn_id);
+		rdma_destroy_id(new_cm_id);
+		goto out;
+	}
+
 	memset(&event, 0, sizeof event);
 	event.event = RDMA_CM_EVENT_CONNECT_REQUEST;
 	event.param.conn.private_data = iw_event->private_data;
 	event.param.conn.private_data_len = iw_event->private_data_len;
+	event.param.conn.initiator_depth = attr.max_qp_init_rd_atom;
+	event.param.conn.responder_resources = attr.max_qp_rd_atom;
 	ret = conn_id->id.event_handler(&conn_id->id, &event);
 	if (ret) {
 		/* User wants to destroy the CM ID */
@@ -2272,7 +2290,7 @@ static int cma_connect_iw(struct rdma_id_private *id_priv,
 	sin = (struct sockaddr_in*) &id_priv->id.route.addr.dst_addr;
 	cm_id->remote_addr = *sin;

-	ret = cma_modify_qp_rtr(id_priv);
+	ret = cma_modify_qp_rtr(id_priv, conn_param);
 	if (ret)
 		goto out;

@@ -2335,25 +2353,15 @@ static int cma_accept_ib(struct rdma_id_private *id_priv,
 			 struct rdma_conn_param *conn_param)
 {
 	struct ib_cm_rep_param rep;
-	struct ib_qp_attr qp_attr;
-	int qp_attr_mask, ret;
+	int ret;

-	if (id_priv->id.qp) {
-		ret = cma_modify_qp_rtr(id_priv);
-		if (ret)
-			goto out;
+	ret = cma_modify_qp_rtr(id_priv, conn_param);
+	if (ret)
+		goto out;

-		qp_attr.qp_state = IB_QPS_RTS;
-		ret = ib_cm_init_qp_attr(id_priv->cm_id.ib, &qp_attr,
-					 &qp_attr_mask);
-		if (ret)
-			goto out;
-
-		qp_attr.max_rd_atomic = conn_param->initiator_depth;
-		ret = ib_modify_qp(id_priv->id.qp, &qp_attr, qp_attr_mask);
-		if (ret)
-			goto out;
-	}
+	ret = cma_modify_qp_rts(id_priv, conn_param);
+	if (ret)
+		goto out;

 	memset(&rep, 0, sizeof rep);
 	rep.qp_num = id_priv->qp_num;
@@ -2378,7 +2386,7 @@ static int cma_accept_iw(struct rdma_id_private *id_priv,
 	struct iw_cm_conn_param iw_param;
 	int ret;

-	ret = cma_modify_qp_rtr(id_priv);
+	ret = cma_modify_qp_rtr(id_priv, conn_param);
 	if (ret)
 		return ret;

@@ -2598,11 +2606,9 @@ static void cma_set_mgid(struct rdma_id_private *id_priv,
 		/* IPv6 address is an SA assigned MGID. */
 		memcpy(mgid, &sin6->sin6_addr, sizeof *mgid);
 	} else {
-		ip_ib_mc_map(sin->sin_addr.s_addr, mc_map);
+		ip_ib_mc_map(sin->sin_addr.s_addr, dev_addr->broadcast, mc_map);
 		if (id_priv->id.ps == RDMA_PS_UDP)
 			mc_map[7] = 0x01;	/* Use RDMA CM signature */
-		mc_map[8] = ib_addr_get_pkey(dev_addr) >> 8;
-		mc_map[9] = (unsigned char) ib_addr_get_pkey(dev_addr);
 		*mgid = *(union ib_gid *) (mc_map + 4);
 	}
 }
@@ -139,7 +139,7 @@ static inline struct ib_pool_fmr *ib_fmr_cache_lookup(struct ib_fmr_pool *pool,
 static void ib_fmr_batch_release(struct ib_fmr_pool *pool)
 {
 	int                 ret;
-	struct ib_pool_fmr *fmr;
+	struct ib_pool_fmr *fmr, *next;
 	LIST_HEAD(unmap_list);
 	LIST_HEAD(fmr_list);

@@ -158,6 +158,20 @@ static void ib_fmr_batch_release(struct ib_fmr_pool *pool)
 #endif
 	}

+	/*
+	 * The free_list may hold FMRs that have been put there
+	 * because they haven't reached the max_remap count.
+	 * Invalidate their mapping as well.
+	 */
+	list_for_each_entry_safe(fmr, next, &pool->free_list, list) {
+		if (fmr->remap_count == 0)
+			continue;
+		hlist_del_init(&fmr->cache_node);
+		fmr->remap_count = 0;
+		list_add_tail(&fmr->fmr->list, &fmr_list);
+		list_move(&fmr->list, &unmap_list);
+	}
+
 	list_splice(&pool->dirty_list, &unmap_list);
 	INIT_LIST_HEAD(&pool->dirty_list);
 	pool->dirty_len = 0;
@@ -182,8 +196,7 @@ static int ib_fmr_cleanup_thread(void *pool_ptr)
 	struct ib_fmr_pool *pool = pool_ptr;

 	do {
-		if (pool->dirty_len >= pool->dirty_watermark ||
-		    atomic_read(&pool->flush_ser) - atomic_read(&pool->req_ser) < 0) {
+		if (atomic_read(&pool->flush_ser) - atomic_read(&pool->req_ser) < 0) {
 			ib_fmr_batch_release(pool);

 			atomic_inc(&pool->flush_ser);
@@ -194,8 +207,7 @@ static int ib_fmr_cleanup_thread(void *pool_ptr)
 		}

 		set_current_state(TASK_INTERRUPTIBLE);
-		if (pool->dirty_len < pool->dirty_watermark &&
-		    atomic_read(&pool->flush_ser) - atomic_read(&pool->req_ser) >= 0 &&
+		if (atomic_read(&pool->flush_ser) - atomic_read(&pool->req_ser) >= 0 &&
 		    !kthread_should_stop())
 			schedule();
 		__set_current_state(TASK_RUNNING);
@@ -369,11 +381,6 @@ void ib_destroy_fmr_pool(struct ib_fmr_pool *pool)

 	i = 0;
 	list_for_each_entry_safe(fmr, tmp, &pool->free_list, list) {
-		if (fmr->remap_count) {
-			INIT_LIST_HEAD(&fmr_list);
-			list_add_tail(&fmr->fmr->list, &fmr_list);
-			ib_unmap_fmr(&fmr_list);
-		}
 		ib_dealloc_fmr(fmr->fmr);
 		list_del(&fmr->list);
 		kfree(fmr);
@@ -511,8 +518,10 @@ int ib_fmr_pool_unmap(struct ib_pool_fmr *fmr)
 			list_add_tail(&fmr->list, &pool->free_list);
 		} else {
 			list_add_tail(&fmr->list, &pool->dirty_list);
-			++pool->dirty_len;
-			wake_up_process(pool->thread);
+			if (++pool->dirty_len >= pool->dirty_watermark) {
+				atomic_inc(&pool->req_ser);
+				wake_up_process(pool->thread);
+			}
 		}
 	}

@@ -701,7 +701,8 @@ static int handle_outgoing_dr_smp(struct ib_mad_agent_private *mad_agent_priv,
 	}

 	/* Check to post send on QP or process locally */
-	if (smi_check_local_smp(smp, device) == IB_SMI_DISCARD)
+	if (smi_check_local_smp(smp, device) == IB_SMI_DISCARD &&
+	    smi_check_local_returning_smp(smp, device) == IB_SMI_DISCARD)
 		goto out;

 	local = kmalloc(sizeof *local, GFP_ATOMIC);
@@ -752,8 +753,7 @@ static int handle_outgoing_dr_smp(struct ib_mad_agent_private *mad_agent_priv,
 		port_priv = ib_get_mad_port(mad_agent_priv->agent.device,
 					    mad_agent_priv->agent.port_num);
 		if (port_priv) {
-			mad_priv->mad.mad.mad_hdr.tid =
-				((struct ib_mad *)smp)->mad_hdr.tid;
+			memcpy(&mad_priv->mad.mad, smp, sizeof(struct ib_mad));
 			recv_mad_agent = find_mad_agent(port_priv,
 						        &mad_priv->mad.mad);
 		}
@@ -1100,7 +1100,9 @@ int ib_post_send_mad(struct ib_mad_send_buf *send_buf,
 		mad_send_wr->tid = ((struct ib_mad_hdr *) send_buf->mad)->tid;
 		/* Timeout will be updated after send completes */
 		mad_send_wr->timeout = msecs_to_jiffies(send_buf->timeout_ms);
-		mad_send_wr->retries = send_buf->retries;
+		mad_send_wr->max_retries = send_buf->retries;
+		mad_send_wr->retries_left = send_buf->retries;
+		send_buf->retries = 0;
 		/* Reference for work request to QP + response */
 		mad_send_wr->refcount = 1 + (mad_send_wr->timeout > 0);
 		mad_send_wr->status = IB_WC_SUCCESS;
@@ -1931,15 +1933,6 @@ local:
 	if (port_priv->device->process_mad) {
 		int ret;

-		if (!response) {
-			printk(KERN_ERR PFX "No memory for response MAD\n");
-			/*
-			 * Is it better to assume that
-			 * it wouldn't be processed ?
-			 */
-			goto out;
-		}
-
 		ret = port_priv->device->process_mad(port_priv->device, 0,
 						     port_priv->port_num,
 						     wc, &recv->grh,
@@ -2282,8 +2275,6 @@ static void cancel_mads(struct ib_mad_agent_private *mad_agent_priv)

 	/* Empty wait list to prevent receives from finding a request */
 	list_splice_init(&mad_agent_priv->wait_list, &cancel_list);
-	/* Empty local completion list as well */
-	list_splice_init(&mad_agent_priv->local_list, &cancel_list);
 	spin_unlock_irqrestore(&mad_agent_priv->lock, flags);

 	/* Report all cancelled requests */
@@ -2445,9 +2436,12 @@ static int retry_send(struct ib_mad_send_wr_private *mad_send_wr)
 {
 	int ret;

-	if (!mad_send_wr->retries--)
+	if (!mad_send_wr->retries_left)
 		return -ETIMEDOUT;

+	mad_send_wr->retries_left--;
+	mad_send_wr->send_buf.retries++;
+
 	mad_send_wr->timeout = msecs_to_jiffies(mad_send_wr->send_buf.timeout_ms);

 	if (mad_send_wr->mad_agent_priv->agent.rmpp_version) {
@@ -131,7 +131,8 @@ struct ib_mad_send_wr_private {
 	struct ib_sge sg_list[IB_MAD_SEND_REQ_MAX_SG];
 	__be64 tid;
 	unsigned long timeout;
-	int retries;
+	int max_retries;
+	int retries_left;
 	int retry;
 	int refcount;
 	enum ib_wc_status status;
@@ -684,7 +684,7 @@ static void process_rmpp_ack(struct ib_mad_agent_private *agent,

 	if (seg_num > mad_send_wr->last_ack) {
 		adjust_last_ack(mad_send_wr, seg_num);
-		mad_send_wr->retries = mad_send_wr->send_buf.retries;
+		mad_send_wr->retries_left = mad_send_wr->max_retries;
 	}
 	mad_send_wr->newwin = newwin;
 	if (mad_send_wr->last_ack == mad_send_wr->send_buf.seg_count) {
@@ -73,11 +73,20 @@ struct mcast_device {
 };

 enum mcast_state {
-	MCAST_IDLE,
 	MCAST_JOINING,
 	MCAST_MEMBER,
+	MCAST_ERROR,
+};
+
+enum mcast_group_state {
+	MCAST_IDLE,
 	MCAST_BUSY,
-	MCAST_ERROR
+	MCAST_GROUP_ERROR,
+	MCAST_PKEY_EVENT
+};
+
+enum {
+	MCAST_INVALID_PKEY_INDEX = 0xFFFF
 };

 struct mcast_member;
@@ -93,9 +102,10 @@ struct mcast_group {
 	struct mcast_member	*last_join;
 	int			members[3];
 	atomic_t		refcount;
-	enum mcast_state	state;
+	enum mcast_group_state	state;
 	struct ib_sa_query	*query;
 	int			query_id;
+	u16			pkey_index;
 };

 struct mcast_member {
@@ -378,9 +388,19 @@ static int fail_join(struct mcast_group *group, struct mcast_member *member,
 static void process_group_error(struct mcast_group *group)
 {
 	struct mcast_member *member;
-	int ret;
+	int ret = 0;
+	u16 pkey_index;
+
+	if (group->state == MCAST_PKEY_EVENT)
+		ret = ib_find_pkey(group->port->dev->device,
+				   group->port->port_num,
+				   be16_to_cpu(group->rec.pkey), &pkey_index);

 	spin_lock_irq(&group->lock);
+	if (group->state == MCAST_PKEY_EVENT && !ret &&
+	    group->pkey_index == pkey_index)
+		goto out;
+
 	while (!list_empty(&group->active_list)) {
 		member = list_entry(group->active_list.next,
 				    struct mcast_member, list);
@@ -399,6 +419,7 @@ static void process_group_error(struct mcast_group *group)
 	}

 	group->rec.join_state = 0;
+out:
 	group->state = MCAST_BUSY;
 	spin_unlock_irq(&group->lock);
 }
@@ -415,9 +436,9 @@ static void mcast_work_handler(struct work_struct *work)
 retest:
 	spin_lock_irq(&group->lock);
 	while (!list_empty(&group->pending_list) ||
-	       (group->state == MCAST_ERROR)) {
+	       (group->state != MCAST_BUSY)) {

-		if (group->state == MCAST_ERROR) {
+		if (group->state != MCAST_BUSY) {
 			spin_unlock_irq(&group->lock);
 			process_group_error(group);
 			goto retest;
@@ -494,12 +515,19 @@ static void join_handler(int status, struct ib_sa_mcmember_rec *rec,
 			 void *context)
 {
 	struct mcast_group *group = context;
+	u16 pkey_index = MCAST_INVALID_PKEY_INDEX;

 	if (status)
 		process_join_error(group, status);
 	else {
+		ib_find_pkey(group->port->dev->device, group->port->port_num,
+			     be16_to_cpu(rec->pkey), &pkey_index);
+
 		spin_lock_irq(&group->port->lock);
 		group->rec = *rec;
+		if (group->state == MCAST_BUSY &&
+		    group->pkey_index == MCAST_INVALID_PKEY_INDEX)
+			group->pkey_index = pkey_index;
 		if (!memcmp(&mgid0, &group->rec.mgid, sizeof mgid0)) {
 			rb_erase(&group->node, &group->port->table);
 			mcast_insert(group->port, group, 1);
@@ -539,6 +567,7 @@ static struct mcast_group *acquire_group(struct mcast_port *port,

 	group->port = port;
 	group->rec.mgid = *mgid;
+	group->pkey_index = MCAST_INVALID_PKEY_INDEX;
 	INIT_LIST_HEAD(&group->pending_list);
 	INIT_LIST_HEAD(&group->active_list);
 	INIT_WORK(&group->work, mcast_work_handler);
@@ -707,7 +736,8 @@ int ib_init_ah_from_mcmember(struct ib_device *device, u8 port_num,
 }
 EXPORT_SYMBOL(ib_init_ah_from_mcmember);

-static void mcast_groups_lost(struct mcast_port *port)
+static void mcast_groups_event(struct mcast_port *port,
+			       enum mcast_group_state state)
 {
 	struct mcast_group *group;
 	struct rb_node *node;
@@ -721,7 +751,8 @@ static void mcast_groups_lost(struct mcast_port *port)
 			atomic_inc(&group->refcount);
 			queue_work(mcast_wq, &group->work);
 		}
-		group->state = MCAST_ERROR;
+		if (group->state != MCAST_GROUP_ERROR)
+			group->state = state;
 		spin_unlock(&group->lock);
 	}
 	spin_unlock_irqrestore(&port->lock, flags);
@@ -731,16 +762,20 @@ static void mcast_event_handler(struct ib_event_handler *handler,
 				struct ib_event *event)
 {
 	struct mcast_device *dev;
+	int index;

 	dev = container_of(handler, struct mcast_device, event_handler);
+	index = event->element.port_num - dev->start_port;

 	switch (event->event) {
 	case IB_EVENT_PORT_ERR:
 	case IB_EVENT_LID_CHANGE:
 	case IB_EVENT_SM_CHANGE:
 	case IB_EVENT_CLIENT_REREGISTER:
-		mcast_groups_lost(&dev->port[event->element.port_num -
-					     dev->start_port]);
+		mcast_groups_event(&dev->port[index], MCAST_GROUP_ERROR);
+		break;
+	case IB_EVENT_PKEY_CHANGE:
+		mcast_groups_event(&dev->port[index], MCAST_PKEY_EVENT);
 		break;
 	default:
 		break;
@@ -59,7 +59,8 @@ extern enum smi_action smi_handle_dr_smp_send(struct ib_smp *smp,
 					      u8 node_type, int port_num);

 /*
- * Return 1 if the SMP should be handled by the local SMA/SM via process_mad
+ * Return IB_SMI_HANDLE if the SMP should be handled by the local SMA/SM
+ * via process_mad
 */
 static inline enum smi_action smi_check_local_smp(struct ib_smp *smp,
 						  struct ib_device *device)
@@ -71,4 +72,19 @@ static inline enum smi_action smi_check_local_smp(struct ib_smp *smp,
 		(smp->hop_ptr == smp->hop_cnt + 1)) ?
 		IB_SMI_HANDLE : IB_SMI_DISCARD);
 }
+
+/*
+ * Return IB_SMI_HANDLE if the SMP should be handled by the local SMA/SM
+ * via process_mad
+ */
+static inline enum smi_action smi_check_local_returning_smp(struct ib_smp *smp,
+						   struct ib_device *device)
+{
+	/* C14-13:3 -- We're at the end of the DR segment of path */
+	/* C14-13:4 -- Hop Pointer == 0 -> give to SM */
+	return ((device->process_mad &&
+		ib_get_smp_direction(smp) &&
+		!smp->hop_ptr) ? IB_SMI_HANDLE : IB_SMI_DISCARD);
+}
+
 #endif	/* __SMI_H_ */
@@ -106,6 +106,9 @@ enum {
 	IB_UCM_MAX_DEVICES = 32
 };

+/* ib_cm and ib_user_cm modules share /sys/class/infiniband_cm */
+extern struct class cm_class;
+
 #define IB_UCM_BASE_DEV MKDEV(IB_UCM_MAJOR, IB_UCM_BASE_MINOR)

 static void ib_ucm_add_one(struct ib_device *device);
@@ -1199,7 +1202,7 @@ static int ib_ucm_close(struct inode *inode, struct file *filp)
 	return 0;
 }

-static void ib_ucm_release_class_dev(struct class_device *class_dev)
+static void ucm_release_class_dev(struct class_device *class_dev)
 {
 	struct ib_ucm_device *dev;

@@ -1217,11 +1220,6 @@ static const struct file_operations ucm_fops = {
 	.poll    = ib_ucm_poll,
 };

-static struct class ucm_class = {
-	.name    = "infiniband_cm",
-	.release = ib_ucm_release_class_dev
-};
-
 static ssize_t show_ibdev(struct class_device *class_dev, char *buf)
 {
 	struct ib_ucm_device *dev;
@@ -1257,9 +1255,10 @@ static void ib_ucm_add_one(struct ib_device *device)
 	if (cdev_add(&ucm_dev->dev, IB_UCM_BASE_DEV + ucm_dev->devnum, 1))
 		goto err;

-	ucm_dev->class_dev.class = &ucm_class;
+	ucm_dev->class_dev.class = &cm_class;
 	ucm_dev->class_dev.dev = device->dma_device;
 	ucm_dev->class_dev.devt = ucm_dev->dev.dev;
+	ucm_dev->class_dev.release = ucm_release_class_dev;
 	snprintf(ucm_dev->class_dev.class_id, BUS_ID_SIZE, "ucm%d",
 		 ucm_dev->devnum);
 	if (class_device_register(&ucm_dev->class_dev))
@@ -1306,40 +1305,34 @@ static int __init ib_ucm_init(void)
 				     "infiniband_cm");
 	if (ret) {
 		printk(KERN_ERR "ucm: couldn't register device number\n");
-		goto err;
+		goto error1;
 	}

-	ret = class_register(&ucm_class);
-	if (ret) {
-		printk(KERN_ERR "ucm: couldn't create class infiniband_cm\n");
-		goto err_chrdev;
-	}
-
-	ret = class_create_file(&ucm_class, &class_attr_abi_version);
+	ret = class_create_file(&cm_class, &class_attr_abi_version);
 	if (ret) {
 		printk(KERN_ERR "ucm: couldn't create abi_version attribute\n");
-		goto err_class;
+		goto error2;
 	}

 	ret = ib_register_client(&ucm_client);
 	if (ret) {
 		printk(KERN_ERR "ucm: couldn't register client\n");
-		goto err_class;
+		goto error3;
 	}
 	return 0;

-err_class:
-	class_unregister(&ucm_class);
-err_chrdev:
+error3:
+	class_remove_file(&cm_class, &class_attr_abi_version);
+error2:
 	unregister_chrdev_region(IB_UCM_BASE_DEV, IB_UCM_MAX_DEVICES);
-err:
+error1:
 	return ret;
 }

 static void __exit ib_ucm_cleanup(void)
 {
 	ib_unregister_client(&ucm_client);
-	class_unregister(&ucm_class);
+	class_remove_file(&cm_class, &class_attr_abi_version);
 	unregister_chrdev_region(IB_UCM_BASE_DEV, IB_UCM_MAX_DEVICES);
 	idr_destroy(&ctx_id_table);
 }
@@ -31,6 +31,7 @@
 */

 #include <linux/completion.h>
+#include <linux/file.h>
 #include <linux/mutex.h>
 #include <linux/poll.h>
 #include <linux/idr.h>
@@ -991,6 +992,96 @@ out:
 	return ret;
 }

+static void ucma_lock_files(struct ucma_file *file1, struct ucma_file *file2)
+{
+	/* Acquire mutex's based on pointer comparison to prevent deadlock. */
+	if (file1 < file2) {
+		mutex_lock(&file1->mut);
+		mutex_lock(&file2->mut);
+	} else {
+		mutex_lock(&file2->mut);
+		mutex_lock(&file1->mut);
+	}
+}
+
+static void ucma_unlock_files(struct ucma_file *file1, struct ucma_file *file2)
+{
+	if (file1 < file2) {
+		mutex_unlock(&file2->mut);
+		mutex_unlock(&file1->mut);
+	} else {
+		mutex_unlock(&file1->mut);
+		mutex_unlock(&file2->mut);
+	}
+}
+
+static void ucma_move_events(struct ucma_context *ctx, struct ucma_file *file)
+{
+	struct ucma_event *uevent, *tmp;
+
+	list_for_each_entry_safe(uevent, tmp, &ctx->file->event_list, list)
+		if (uevent->ctx == ctx)
+			list_move_tail(&uevent->list, &file->event_list);
+}
+
+static ssize_t ucma_migrate_id(struct ucma_file *new_file,
+			       const char __user *inbuf,
+			       int in_len, int out_len)
+{
+	struct rdma_ucm_migrate_id cmd;
+	struct rdma_ucm_migrate_resp resp;
+	struct ucma_context *ctx;
+	struct file *filp;
+	struct ucma_file *cur_file;
+	int ret = 0;
+
+	if (copy_from_user(&cmd, inbuf, sizeof(cmd)))
+		return -EFAULT;
+
+	/* Get current fd to protect against it being closed */
+	filp = fget(cmd.fd);
+	if (!filp)
+		return -ENOENT;
+
+	/* Validate current fd and prevent destruction of id. */
+	ctx = ucma_get_ctx(filp->private_data, cmd.id);
+	if (IS_ERR(ctx)) {
+		ret = PTR_ERR(ctx);
+		goto file_put;
+	}
+
+	cur_file = ctx->file;
+	if (cur_file == new_file) {
+		resp.events_reported = ctx->events_reported;
+		goto response;
+	}
+
+	/*
+	 * Migrate events between fd's, maintaining order, and avoiding new
+	 * events being added before existing events.
+	 */
+	ucma_lock_files(cur_file, new_file);
+	mutex_lock(&mut);
+
+	list_move_tail(&ctx->list, &new_file->ctx_list);
+	ucma_move_events(ctx, new_file);
+	ctx->file = new_file;
+	resp.events_reported = ctx->events_reported;
+
+	mutex_unlock(&mut);
+	ucma_unlock_files(cur_file, new_file);
+
+response:
+	if (copy_to_user((void __user *)(unsigned long)cmd.response,
+			 &resp, sizeof(resp)))
+		ret = -EFAULT;
+
+	ucma_put_ctx(ctx);
+file_put:
+	fput(filp);
+	return ret;
+}
+
 static ssize_t (*ucma_cmd_table[])(struct ucma_file *file,
 				   const char __user *inbuf,
 				   int in_len, int out_len) = {
@@ -1012,6 +1103,7 @@ static ssize_t (*ucma_cmd_table[])(struct ucma_file *file,
 	[RDMA_USER_CM_CMD_NOTIFY]	= ucma_notify,
 	[RDMA_USER_CM_CMD_JOIN_MCAST]	= ucma_join_multicast,
 	[RDMA_USER_CM_CMD_LEAVE_MCAST]	= ucma_leave_multicast,
+	[RDMA_USER_CM_CMD_MIGRATE_ID]	= ucma_migrate_id
 };

 static ssize_t ucma_write(struct file *filp, const char __user *buf,
@@ -2,6 +2,7 @@
 * Copyright (c) 2004 Topspin Communications.  All rights reserved.
 * Copyright (c) 2005 Voltaire, Inc. All rights reserved.
 * Copyright (c) 2005 Sun Microsystems, Inc. All rights reserved.
+ * Copyright (c) 2008 Cisco. 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
@@ -42,7 +43,7 @@
 #include <linux/cdev.h>
 #include <linux/dma-mapping.h>
 #include <linux/poll.h>
-#include <linux/rwsem.h>
+#include <linux/mutex.h>
 #include <linux/kref.h>
 #include <linux/compat.h>

@@ -94,7 +95,7 @@ struct ib_umad_port {
 	struct class_device   *sm_class_dev;
 	struct semaphore       sm_sem;

-	struct rw_semaphore    mutex;
+	struct mutex	       file_mutex;
 	struct list_head       file_list;

 	struct ib_device      *ib_dev;
@@ -110,11 +111,11 @@ struct ib_umad_device {
 };

 struct ib_umad_file {
+	struct mutex		mutex;
 	struct ib_umad_port    *port;
 	struct list_head	recv_list;
 	struct list_head	send_list;
 	struct list_head	port_list;
-	spinlock_t		recv_lock;
 	spinlock_t		send_lock;
 	wait_queue_head_t	recv_wait;
 	struct ib_mad_agent    *agent[IB_UMAD_MAX_AGENTS];
@@ -156,7 +157,7 @@ static int hdr_size(struct ib_umad_file *file)
 		sizeof (struct ib_user_mad_hdr_old);
 }

-/* caller must hold port->mutex at least for reading */
+/* caller must hold file->mutex */
 static struct ib_mad_agent *__get_agent(struct ib_umad_file *file, int id)
 {
 	return file->agents_dead ? NULL : file->agent[id];
@@ -168,32 +169,30 @@ static int queue_packet(struct ib_umad_file *file,
 {
 	int ret = 1;

-	down_read(&file->port->mutex);
+	mutex_lock(&file->mutex);

 	for (packet->mad.hdr.id = 0;
 	     packet->mad.hdr.id < IB_UMAD_MAX_AGENTS;
 	     packet->mad.hdr.id++)
 		if (agent == __get_agent(file, packet->mad.hdr.id)) {
-			spin_lock_irq(&file->recv_lock);
 			list_add_tail(&packet->list, &file->recv_list);
-			spin_unlock_irq(&file->recv_lock);
 			wake_up_interruptible(&file->recv_wait);
 			ret = 0;
 			break;
 		}

-	up_read(&file->port->mutex);
+	mutex_unlock(&file->mutex);

 	return ret;
 }

 static void dequeue_send(struct ib_umad_file *file,
 			 struct ib_umad_packet *packet)
- {
+{
 	spin_lock_irq(&file->send_lock);
 	list_del(&packet->list);
 	spin_unlock_irq(&file->send_lock);
- }
+}

 static void send_handler(struct ib_mad_agent *agent,
 			 struct ib_mad_send_wc *send_wc)
@@ -341,10 +340,10 @@ static ssize_t ib_umad_read(struct file *filp, char __user *buf,
 	if (count < hdr_size(file))
 		return -EINVAL;

-	spin_lock_irq(&file->recv_lock);
+	mutex_lock(&file->mutex);

 	while (list_empty(&file->recv_list)) {
-		spin_unlock_irq(&file->recv_lock);
+		mutex_unlock(&file->mutex);

 		if (filp->f_flags & O_NONBLOCK)
 			return -EAGAIN;
@@ -353,13 +352,13 @@ static ssize_t ib_umad_read(struct file *filp, char __user *buf,
 					     !list_empty(&file->recv_list)))
 			return -ERESTARTSYS;

-		spin_lock_irq(&file->recv_lock);
+		mutex_lock(&file->mutex);
 	}

 	packet = list_entry(file->recv_list.next, struct ib_umad_packet, list);
 	list_del(&packet->list);

-	spin_unlock_irq(&file->recv_lock);
+	mutex_unlock(&file->mutex);

 	if (packet->recv_wc)
 		ret = copy_recv_mad(file, buf, packet, count);
@@ -368,9 +367,9 @@ static ssize_t ib_umad_read(struct file *filp, char __user *buf,

 	if (ret < 0) {
 		/* Requeue packet */
-		spin_lock_irq(&file->recv_lock);
+		mutex_lock(&file->mutex);
 		list_add(&packet->list, &file->recv_list);
-		spin_unlock_irq(&file->recv_lock);
+		mutex_unlock(&file->mutex);
 	} else {
 		if (packet->recv_wc)
 			ib_free_recv_mad(packet->recv_wc);
@@ -481,7 +480,7 @@ static ssize_t ib_umad_write(struct file *filp, const char __user *buf,
 		goto err;
 	}

-	down_read(&file->port->mutex);
+	mutex_lock(&file->mutex);

 	agent = __get_agent(file, packet->mad.hdr.id);
 	if (!agent) {
@@ -577,7 +576,7 @@ static ssize_t ib_umad_write(struct file *filp, const char __user *buf,
 	if (ret)
 		goto err_send;

-	up_read(&file->port->mutex);
+	mutex_unlock(&file->mutex);
 	return count;

 err_send:
@@ -587,7 +586,7 @@ err_msg:
 err_ah:
 	ib_destroy_ah(ah);
 err_up:
-	up_read(&file->port->mutex);
+	mutex_unlock(&file->mutex);
 err:
 	kfree(packet);
 	return ret;
@@ -613,11 +612,12 @@ static int ib_umad_reg_agent(struct ib_umad_file *file, void __user *arg,
 {
 	struct ib_user_mad_reg_req ureq;
 	struct ib_mad_reg_req req;
-	struct ib_mad_agent *agent;
+	struct ib_mad_agent *agent = NULL;
 	int agent_id;
 	int ret;

-	down_write(&file->port->mutex);
+	mutex_lock(&file->port->file_mutex);
+	mutex_lock(&file->mutex);

 	if (!file->port->ib_dev) {
 		ret = -EPIPE;
@@ -666,13 +666,13 @@ found:
 				      send_handler, recv_handler, file);
 	if (IS_ERR(agent)) {
 		ret = PTR_ERR(agent);
+		agent = NULL;
 		goto out;
 	}

 	if (put_user(agent_id,
 		     (u32 __user *) (arg + offsetof(struct ib_user_mad_reg_req, id)))) {
 		ret = -EFAULT;
-		ib_unregister_mad_agent(agent);
 		goto out;
 	}

@@ -690,7 +690,13 @@ found:
 	ret = 0;

 out:
-	up_write(&file->port->mutex);
+	mutex_unlock(&file->mutex);
+
+	if (ret && agent)
+		ib_unregister_mad_agent(agent);
+
+	mutex_unlock(&file->port->file_mutex);
+
 	return ret;
 }

@@ -703,7 +709,8 @@ static int ib_umad_unreg_agent(struct ib_umad_file *file, u32 __user *arg)
 	if (get_user(id, arg))
 		return -EFAULT;

-	down_write(&file->port->mutex);
+	mutex_lock(&file->port->file_mutex);
+	mutex_lock(&file->mutex);

 	if (id < 0 || id >= IB_UMAD_MAX_AGENTS || !__get_agent(file, id)) {
 		ret = -EINVAL;
@@ -714,11 +721,13 @@ static int ib_umad_unreg_agent(struct ib_umad_file *file, u32 __user *arg)
 	file->agent[id] = NULL;

 out:
-	up_write(&file->port->mutex);
+	mutex_unlock(&file->mutex);

 	if (agent)
 		ib_unregister_mad_agent(agent);

+	mutex_unlock(&file->port->file_mutex);
+
 	return ret;
 }

@@ -726,12 +735,12 @@ static long ib_umad_enable_pkey(struct ib_umad_file *file)
 {
 	int ret = 0;

-	down_write(&file->port->mutex);
+	mutex_lock(&file->mutex);
 	if (file->already_used)
 		ret = -EINVAL;
 	else
 		file->use_pkey_index = 1;
-	up_write(&file->port->mutex);
+	mutex_unlock(&file->mutex);

 	return ret;
 }
@@ -783,7 +792,7 @@ static int ib_umad_open(struct inode *inode, struct file *filp)
 	if (!port)
 		return -ENXIO;

-	down_write(&port->mutex);
+	mutex_lock(&port->file_mutex);

 	if (!port->ib_dev) {
 		ret = -ENXIO;
@@ -797,7 +806,7 @@ static int ib_umad_open(struct inode *inode, struct file *filp)
 		goto out;
 	}

-	spin_lock_init(&file->recv_lock);
+	mutex_init(&file->mutex);
 	spin_lock_init(&file->send_lock);
 	INIT_LIST_HEAD(&file->recv_list);
 	INIT_LIST_HEAD(&file->send_list);
@@ -809,7 +818,7 @@ static int ib_umad_open(struct inode *inode, struct file *filp)
 	list_add_tail(&file->port_list, &port->file_list);

 out:
-	up_write(&port->mutex);
+	mutex_unlock(&port->file_mutex);
 	return ret;
 }

@@ -821,7 +830,8 @@ static int ib_umad_close(struct inode *inode, struct file *filp)
 	int already_dead;
 	int i;

-	down_write(&file->port->mutex);
+	mutex_lock(&file->port->file_mutex);
+	mutex_lock(&file->mutex);

 	already_dead = file->agents_dead;
 	file->agents_dead = 1;
@@ -834,14 +844,14 @@ static int ib_umad_close(struct inode *inode, struct file *filp)

 	list_del(&file->port_list);

-	downgrade_write(&file->port->mutex);
+	mutex_unlock(&file->mutex);

 	if (!already_dead)
 		for (i = 0; i < IB_UMAD_MAX_AGENTS; ++i)
 			if (file->agent[i])
 				ib_unregister_mad_agent(file->agent[i]);

-	up_read(&file->port->mutex);
+	mutex_unlock(&file->port->file_mutex);

 	kfree(file);
 	kref_put(&dev->ref, ib_umad_release_dev);
@@ -914,10 +924,10 @@ static int ib_umad_sm_close(struct inode *inode, struct file *filp)
 	};
 	int ret = 0;

-	down_write(&port->mutex);
+	mutex_lock(&port->file_mutex);
 	if (port->ib_dev)
 		ret = ib_modify_port(port->ib_dev, port->port_num, 0, &props);
-	up_write(&port->mutex);
+	mutex_unlock(&port->file_mutex);

 	up(&port->sm_sem);

@@ -981,7 +991,7 @@ static int ib_umad_init_port(struct ib_device *device, int port_num,
 	port->ib_dev   = device;
 	port->port_num = port_num;
 	init_MUTEX(&port->sm_sem);
-	init_rwsem(&port->mutex);
+	mutex_init(&port->file_mutex);
 	INIT_LIST_HEAD(&port->file_list);

 	port->dev = cdev_alloc();
@@ -1052,6 +1062,7 @@ err_cdev:
 static void ib_umad_kill_port(struct ib_umad_port *port)
 {
 	struct ib_umad_file *file;
+	int already_dead;
 	int id;

 	class_set_devdata(port->class_dev,    NULL);
@@ -1067,42 +1078,22 @@ static void ib_umad_kill_port(struct ib_umad_port *port)
 	umad_port[port->dev_num] = NULL;
 	spin_unlock(&port_lock);

-	down_write(&port->mutex);
+	mutex_lock(&port->file_mutex);

 	port->ib_dev = NULL;

-	/*
-	 * Now go through the list of files attached to this port and
-	 * unregister all of their MAD agents.  We need to hold
-	 * port->mutex while doing this to avoid racing with
-	 * ib_umad_close(), but we can't hold the mutex for writing
-	 * while calling ib_unregister_mad_agent(), since that might
-	 * deadlock by calling back into queue_packet().  So we
-	 * downgrade our lock to a read lock, and then drop and
-	 * reacquire the write lock for the next iteration.
-	 *
-	 * We do list_del_init() on the file's list_head so that the
-	 * list_del in ib_umad_close() is still OK, even after the
-	 * file is removed from the list.
-	 */
-	while (!list_empty(&port->file_list)) {
-		file = list_entry(port->file_list.next, struct ib_umad_file,
-				  port_list);
-
+	list_for_each_entry(file, &port->file_list, port_list) {
+		mutex_lock(&file->mutex);
+		already_dead = file->agents_dead;
 		file->agents_dead = 1;
-		list_del_init(&file->port_list);
-
-		downgrade_write(&port->mutex);
+		mutex_unlock(&file->mutex);

 		for (id = 0; id < IB_UMAD_MAX_AGENTS; ++id)
 			if (file->agent[id])
 				ib_unregister_mad_agent(file->agent[id]);
-
-		up_read(&port->mutex);
-		down_write(&port->mutex);
 	}

-	up_write(&port->mutex);
+	mutex_unlock(&port->file_mutex);

 	clear_bit(port->dev_num, dev_map);
 }
@@ -179,7 +179,7 @@ int cxio_create_cq(struct cxio_rdev *rdev_p, struct t3_cq *cq)
 	setup.size = 1UL << cq->size_log2;
 	setup.credits = 65535;
 	setup.credit_thres = 1;
-	if (rdev_p->t3cdev_p->type == T3B)
+	if (rdev_p->t3cdev_p->type != T3A)
 		setup.ovfl_mode = 0;
 	else
 		setup.ovfl_mode = 1;
@@ -584,7 +584,7 @@ static int cxio_hal_ctrl_qp_write_mem(struct cxio_rdev *rdev_p, u32 addr,
 {
 	u32 i, nr_wqe, copy_len;
 	u8 *copy_data;
-	u8 wr_len, utx_len;	/* lenght in 8 byte flit */
+	u8 wr_len, utx_len;	/* length in 8 byte flit */
 	enum t3_wr_flags flag;
 	__be64 *wqe;
 	u64 utx_cmd;
@@ -315,7 +315,7 @@ struct t3_rdma_init_wr {
 	__be32 ird;
 	__be64 qp_dma_addr;	/* 7 */
 	__be32 qp_dma_size;	/* 8 */
-	u32 irs;
+	__be32 irs;
 };

 struct t3_genbit {
@@ -324,7 +324,8 @@ struct t3_genbit {
 };

 enum rdma_init_wr_flags {
-	RECVS_POSTED = 1,
+	RECVS_POSTED = (1<<0),
+	PRIV_QP = (1<<1),
 };

 union t3_wr {
@@ -1118,7 +1118,7 @@ static int act_open_rpl(struct t3cdev *tdev, struct sk_buff *skb, void *ctx)
 	     status2errno(rpl->status));
 	connect_reply_upcall(ep, status2errno(rpl->status));
 	state_set(&ep->com, DEAD);
-	if (ep->com.tdev->type == T3B && act_open_has_tid(rpl->status))
+	if (ep->com.tdev->type != T3A && act_open_has_tid(rpl->status))
 		release_tid(ep->com.tdev, GET_TID(rpl), NULL);
 	cxgb3_free_atid(ep->com.tdev, ep->atid);
 	dst_release(ep->dst);
@@ -1249,7 +1249,7 @@ static void reject_cr(struct t3cdev *tdev, u32 hwtid, __be32 peer_ip,
 	skb_trim(skb, sizeof(struct cpl_tid_release));
 	skb_get(skb);

-	if (tdev->type == T3B)
+	if (tdev->type != T3A)
 		release_tid(tdev, hwtid, skb);
 	else {
 		struct cpl_pass_accept_rpl *rpl;
@@ -122,6 +122,13 @@ int build_phys_page_list(struct ib_phys_buf *buffer_list,
 		*total_size += buffer_list[i].size;
 		if (i > 0)
 			mask |= buffer_list[i].addr;
+		else
+			mask |= buffer_list[i].addr & PAGE_MASK;
+		if (i != num_phys_buf - 1)
+			mask |= buffer_list[i].addr + buffer_list[i].size;
+		else
+			mask |= (buffer_list[i].addr + buffer_list[i].size +
+				PAGE_SIZE - 1) & PAGE_MASK;
 	}

 	if (*total_size > 0xFFFFFFFFULL)
@@ -39,6 +39,7 @@
 #include <linux/list.h>
 #include <linux/spinlock.h>
 #include <linux/ethtool.h>
+#include <linux/rtnetlink.h>

 #include <asm/io.h>
 #include <asm/irq.h>
@@ -645,7 +646,7 @@ static struct ib_mr *iwch_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 	if (err)
 		goto err;

-	if (udata && t3b_device(rhp)) {
+	if (udata && !t3a_device(rhp)) {
 		uresp.pbl_addr = (mhp->attr.pbl_addr -
 	                         rhp->rdev.rnic_info.pbl_base) >> 3;
 		PDBG("%s user resp pbl_addr 0x%x\n", __FUNCTION__,
@@ -1053,7 +1054,9 @@ static ssize_t show_fw_ver(struct class_device *cdev, char *buf)
 	struct net_device *lldev = dev->rdev.t3cdev_p->lldev;

 	PDBG("%s class dev 0x%p\n", __FUNCTION__, cdev);
+	rtnl_lock();
 	lldev->ethtool_ops->get_drvinfo(lldev, &info);
+	rtnl_unlock();
 	return sprintf(buf, "%s\n", info.fw_version);
 }

@@ -1065,7 +1068,9 @@ static ssize_t show_hca(struct class_device *cdev, char *buf)
 	struct net_device *lldev = dev->rdev.t3cdev_p->lldev;

 	PDBG("%s class dev 0x%p\n", __FUNCTION__, cdev);
+	rtnl_lock();
 	lldev->ethtool_ops->get_drvinfo(lldev, &info);
+	rtnl_unlock();
 	return sprintf(buf, "%s\n", info.driver);
 }

@@ -208,36 +208,19 @@ static int iwch_sgl2pbl_map(struct iwch_dev *rhp, struct ib_sge *sg_list,
 static int iwch_build_rdma_recv(struct iwch_dev *rhp, union t3_wr *wqe,
 				struct ib_recv_wr *wr)
 {
-	int i, err = 0;
-	u32 pbl_addr[4];
-	u8 page_size[4];
+	int i;
 	if (wr->num_sge > T3_MAX_SGE)
 		return -EINVAL;
-	err = iwch_sgl2pbl_map(rhp, wr->sg_list, wr->num_sge, pbl_addr,
-			       page_size);
-	if (err)
-		return err;
-	wqe->recv.pagesz[0] = page_size[0];
-	wqe->recv.pagesz[1] = page_size[1];
-	wqe->recv.pagesz[2] = page_size[2];
-	wqe->recv.pagesz[3] = page_size[3];
 	wqe->recv.num_sgle = cpu_to_be32(wr->num_sge);
 	for (i = 0; i < wr->num_sge; i++) {
 		wqe->recv.sgl[i].stag = cpu_to_be32(wr->sg_list[i].lkey);
 		wqe->recv.sgl[i].len = cpu_to_be32(wr->sg_list[i].length);
-
-		/* to in the WQE == the offset into the page */
-		wqe->recv.sgl[i].to = cpu_to_be64(((u32) wr->sg_list[i].addr) %
-				(1UL << (12 + page_size[i])));
-
-		/* pbl_addr is the adapters address in the PBL */
-		wqe->recv.pbl_addr[i] = cpu_to_be32(pbl_addr[i]);
+		wqe->recv.sgl[i].to = cpu_to_be64(wr->sg_list[i].addr);
 	}
 	for (; i < T3_MAX_SGE; i++) {
 		wqe->recv.sgl[i].stag = 0;
 		wqe->recv.sgl[i].len = 0;
 		wqe->recv.sgl[i].to = 0;
-		wqe->recv.pbl_addr[i] = 0;
 	}
 	return 0;
 }
@@ -659,6 +642,7 @@ static void __flush_qp(struct iwch_qp *qhp, unsigned long *flag)
 	cxio_flush_rq(&qhp->wq, &rchp->cq, count);
 	spin_unlock(&qhp->lock);
 	spin_unlock_irqrestore(&rchp->lock, *flag);
+	(*rchp->ibcq.comp_handler)(&rchp->ibcq, rchp->ibcq.cq_context);

 	/* locking heirarchy: cq lock first, then qp lock. */
 	spin_lock_irqsave(&schp->lock, *flag);
@@ -668,6 +652,7 @@ static void __flush_qp(struct iwch_qp *qhp, unsigned long *flag)
 	cxio_flush_sq(&qhp->wq, &schp->cq, count);
 	spin_unlock(&qhp->lock);
 	spin_unlock_irqrestore(&schp->lock, *flag);
+	(*schp->ibcq.comp_handler)(&schp->ibcq, schp->ibcq.cq_context);

 	/* deref */
 	if (atomic_dec_and_test(&qhp->refcnt))
@@ -678,7 +663,7 @@ static void __flush_qp(struct iwch_qp *qhp, unsigned long *flag)

 static void flush_qp(struct iwch_qp *qhp, unsigned long *flag)
 {
-	if (t3b_device(qhp->rhp))
+	if (qhp->ibqp.uobject)
 		cxio_set_wq_in_error(&qhp->wq);
 	else
 		__flush_qp(qhp, flag);
@@ -732,6 +717,7 @@ static int rdma_init(struct iwch_dev *rhp, struct iwch_qp *qhp,
 	init_attr.qp_dma_addr = qhp->wq.dma_addr;
 	init_attr.qp_dma_size = (1UL << qhp->wq.size_log2);
 	init_attr.flags = rqes_posted(qhp) ? RECVS_POSTED : 0;
+	init_attr.flags |= capable(CAP_NET_BIND_SERVICE) ? PRIV_QP : 0;
 	init_attr.irs = qhp->ep->rcv_seq;
 	PDBG("%s init_attr.rq_addr 0x%x init_attr.rq_size = %d "
 	     "flags 0x%x qpcaps 0x%x\n", __FUNCTION__,
@@ -847,10 +833,11 @@ int iwch_modify_qp(struct iwch_dev *rhp, struct iwch_qp *qhp,
 				disconnect = 1;
 				ep = qhp->ep;
 			}
+			flush_qp(qhp, &flag);
 			break;
 		case IWCH_QP_STATE_TERMINATE:
 			qhp->attr.state = IWCH_QP_STATE_TERMINATE;
-			if (t3b_device(qhp->rhp))
+			if (qhp->ibqp.uobject)
 				cxio_set_wq_in_error(&qhp->wq);
 			if (!internal)
 				terminate = 1;
@@ -1,7 +1,7 @@
 /*
 *  IBM eServer eHCA Infiniband device driver for Linux on POWER
 *
- *  adress vector functions
+ *  address vector functions
 *
 *  Authors: Hoang-Nam Nguyen <hnguyen@de.ibm.com>
 *           Khadija Souissi <souissik@de.ibm.com>
@@ -94,7 +94,11 @@ struct ehca_sma_attr {

 struct ehca_sport {
 	struct ib_cq *ibcq_aqp1;
-	struct ib_qp *ibqp_aqp1;
+	struct ib_qp *ibqp_sqp[2];
+	/* lock to serialze modify_qp() calls for sqp in normal
+	 * and irq path (when event PORT_ACTIVE is received first time)
+	 */
+	spinlock_t mod_sqp_lock;
 	enum ib_port_state port_state;
 	struct ehca_sma_attr saved_attr;
 };
@@ -141,6 +145,14 @@ enum ehca_ext_qp_type {
 	EQPT_SRQ       = 3,
 };

+/* struct to cache modify_qp()'s parms for GSI/SMI qp */
+struct ehca_mod_qp_parm {
+	int mask;
+	struct ib_qp_attr attr;
+};
+
+#define EHCA_MOD_QP_PARM_MAX 4
+
 struct ehca_qp {
 	union {
 		struct ib_qp ib_qp;
@@ -164,10 +176,18 @@ struct ehca_qp {
 	struct ehca_cq *recv_cq;
 	unsigned int sqerr_purgeflag;
 	struct hlist_node list_entries;
+	/* array to cache modify_qp()'s parms for GSI/SMI qp */
+	struct ehca_mod_qp_parm *mod_qp_parm;
+	int mod_qp_parm_idx;
 	/* mmap counter for resources mapped into user space */
 	u32 mm_count_squeue;
 	u32 mm_count_rqueue;
 	u32 mm_count_galpa;
+	/* unsolicited ack circumvention */
+	int unsol_ack_circ;
+	int mtu_shift;
+	u32 message_count;
+	u32 packet_count;
 };

 #define IS_SRQ(qp) (qp->ext_type == EQPT_SRQ)
@@ -323,6 +343,7 @@ extern int ehca_port_act_time;
 extern int ehca_use_hp_mr;
 extern int ehca_scaling_code;
 extern int ehca_lock_hcalls;
+extern int ehca_nr_ports;

 struct ipzu_queue_resp {
 	u32 qe_size;      /* queue entry size */
--- a/Show More
+++ b/Show More