From 52f993b8e89487ec9ee15a7fb4979e0f09a45b27 Mon Sep 17 00:00:00 2001
From: Yunhong Jiang <yunhong.jiang@linux.intel.com>
Date: Wed, 8 Mar 2017 23:13:28 -0800
Subject: Upgrade to 4.4.50-rt62

The current kernel is based on rt kernel v4.4.6-rt14. We will upgrade it
to 4.4.50-rt62.
The command to achieve it is:
a) Clone a git repo from
git://git.kernel.org/pub/scm/linux/kernel/git/rt/linux-stable-rt.git

b) Get the diff between this two changesets:
git diff 640eca2901f3435e616157b11379d3223a44b391
705619beeea1b0b48219a683fd1a901a86fdaf5e

where the two commits are:
[yjiang5@jnakajim-build linux-stable-rt]$ git show --oneline --name-only
640eca2901f3435e616157b11379d3223a44b391
640eca2901f3 v4.4.6-rt14
localversion-rt
[yjiang5@jnakajim-build linux-stable-rt]$ git show --oneline --name-only
705619beeea1b0b48219a683fd1a901a86fdaf5e
705619beeea1 Linux 4.4.50-rt62
localversion-rt

c) One patch has been backported thus revert the patch before applying.

 filterdiff -p1 -x scripts/package/Makefile
~/tmp/v4.4.6-rt14-4.4.50-rt62.diff  |patch -p1 --dry-run

Upstream status: backport

Change-Id: I244d57a32f6066e5a5b9915f9fbf99e7bbca6e01
Signed-off-by: Yunhong Jiang <yunhong.jiang@linux.intel.com>
---
 kernel/drivers/infiniband/core/cm.c                | 130 ++++++++++++++++++---
 kernel/drivers/infiniband/core/cma.c               |   3 +-
 kernel/drivers/infiniband/core/iwpm_util.c         |   1 +
 kernel/drivers/infiniband/core/mad.c               |   2 +-
 kernel/drivers/infiniband/core/multicast.c         |  20 ++--
 kernel/drivers/infiniband/core/sa_query.c          |   2 +-
 kernel/drivers/infiniband/core/ucm.c               |   4 +
 kernel/drivers/infiniband/core/ucma.c              |   3 +
 kernel/drivers/infiniband/core/umem.c              |   4 +-
 kernel/drivers/infiniband/core/uverbs.h            |   1 +
 kernel/drivers/infiniband/core/uverbs_main.c       |  49 +++++---
 kernel/drivers/infiniband/hw/cxgb4/cq.c            |   2 +-
 kernel/drivers/infiniband/hw/cxgb4/qp.c            |   6 +-
 kernel/drivers/infiniband/hw/mlx4/ah.c             |  13 ++-
 kernel/drivers/infiniband/hw/mlx4/cq.c             |   5 +-
 kernel/drivers/infiniband/hw/mlx4/mad.c            |  47 +++++++-
 kernel/drivers/infiniband/hw/mlx4/main.c           |  29 +++--
 kernel/drivers/infiniband/hw/mlx4/mcg.c            |  14 +--
 kernel/drivers/infiniband/hw/mlx4/mlx4_ib.h        |   2 +-
 kernel/drivers/infiniband/hw/mlx4/qp.c             |  46 ++++----
 kernel/drivers/infiniband/hw/mlx5/cq.c             |  15 ++-
 kernel/drivers/infiniband/hw/mlx5/main.c           |  18 +--
 kernel/drivers/infiniband/hw/mlx5/mr.c             |  28 +++++
 kernel/drivers/infiniband/hw/mlx5/qp.c             |  21 ++--
 kernel/drivers/infiniband/hw/qib/qib_file_ops.c    |   5 +
 kernel/drivers/infiniband/ulp/ipoib/ipoib.h        |  21 +++-
 kernel/drivers/infiniband/ulp/ipoib/ipoib_cm.c     |  33 ++++--
 kernel/drivers/infiniband/ulp/ipoib/ipoib_ib.c     |  23 ++--
 kernel/drivers/infiniband/ulp/ipoib/ipoib_main.c   |  60 ++++++----
 .../drivers/infiniband/ulp/ipoib/ipoib_multicast.c |  37 ++++--
 kernel/drivers/infiniband/ulp/isert/ib_isert.c     | 122 +++++++++----------
 kernel/drivers/infiniband/ulp/isert/ib_isert.h     |   2 +-
 kernel/drivers/infiniband/ulp/srp/ib_srp.c         |   2 +-
 kernel/drivers/infiniband/ulp/srpt/ib_srpt.c       |  59 +---------
 34 files changed, 523 insertions(+), 306 deletions(-)

(limited to 'kernel/drivers/infiniband')

diff --git a/kernel/drivers/infiniband/core/cm.c b/kernel/drivers/infiniband/core/cm.c
index d6d2b3582..941cd9b83 100644
--- a/kernel/drivers/infiniband/core/cm.c
+++ b/kernel/drivers/infiniband/core/cm.c
@@ -80,6 +80,8 @@ static struct ib_cm {
 	__be32 random_id_operand;
 	struct list_head timewait_list;
 	struct workqueue_struct *wq;
+	/* Sync on cm change port state */
+	spinlock_t state_lock;
 } cm;
 
 /* Counter indexes ordered by attribute ID */
@@ -161,6 +163,8 @@ struct cm_port {
 	struct ib_mad_agent *mad_agent;
 	struct kobject port_obj;
 	u8 port_num;
+	struct list_head cm_priv_prim_list;
+	struct list_head cm_priv_altr_list;
 	struct cm_counter_group counter_group[CM_COUNTER_GROUPS];
 };
 
@@ -241,6 +245,12 @@ struct cm_id_private {
 	u8 service_timeout;
 	u8 target_ack_delay;
 
+	struct list_head prim_list;
+	struct list_head altr_list;
+	/* Indicates that the send port mad is registered and av is set */
+	int prim_send_port_not_ready;
+	int altr_send_port_not_ready;
+
 	struct list_head work_list;
 	atomic_t work_count;
 };
@@ -259,20 +269,47 @@ static int cm_alloc_msg(struct cm_id_private *cm_id_priv,
 	struct ib_mad_agent *mad_agent;
 	struct ib_mad_send_buf *m;
 	struct ib_ah *ah;
+	struct cm_av *av;
+	unsigned long flags, flags2;
+	int ret = 0;
 
+	/* don't let the port to be released till the agent is down */
+	spin_lock_irqsave(&cm.state_lock, flags2);
+	spin_lock_irqsave(&cm.lock, flags);
+	if (!cm_id_priv->prim_send_port_not_ready)
+		av = &cm_id_priv->av;
+	else if (!cm_id_priv->altr_send_port_not_ready &&
+		 (cm_id_priv->alt_av.port))
+		av = &cm_id_priv->alt_av;
+	else {
+		pr_info("%s: not valid CM id\n", __func__);
+		ret = -ENODEV;
+		spin_unlock_irqrestore(&cm.lock, flags);
+		goto out;
+	}
+	spin_unlock_irqrestore(&cm.lock, flags);
+	/* Make sure the port haven't released the mad yet */
 	mad_agent = cm_id_priv->av.port->mad_agent;
-	ah = ib_create_ah(mad_agent->qp->pd, &cm_id_priv->av.ah_attr);
-	if (IS_ERR(ah))
-		return PTR_ERR(ah);
+	if (!mad_agent) {
+		pr_info("%s: not a valid MAD agent\n", __func__);
+		ret = -ENODEV;
+		goto out;
+	}
+	ah = ib_create_ah(mad_agent->qp->pd, &av->ah_attr);
+	if (IS_ERR(ah)) {
+		ret = PTR_ERR(ah);
+		goto out;
+	}
 
 	m = ib_create_send_mad(mad_agent, cm_id_priv->id.remote_cm_qpn,
-			       cm_id_priv->av.pkey_index,
+			       av->pkey_index,
 			       0, IB_MGMT_MAD_HDR, IB_MGMT_MAD_DATA,
 			       GFP_ATOMIC,
 			       IB_MGMT_BASE_VERSION);
 	if (IS_ERR(m)) {
 		ib_destroy_ah(ah);
-		return PTR_ERR(m);
+		ret = PTR_ERR(m);
+		goto out;
 	}
 
 	/* Timeout set by caller if response is expected. */
@@ -282,7 +319,10 @@ static int cm_alloc_msg(struct cm_id_private *cm_id_priv,
 	atomic_inc(&cm_id_priv->refcount);
 	m->context[0] = cm_id_priv;
 	*msg = m;
-	return 0;
+
+out:
+	spin_unlock_irqrestore(&cm.state_lock, flags2);
+	return ret;
 }
 
 static int cm_alloc_response_msg(struct cm_port *port,
@@ -352,7 +392,8 @@ static void cm_init_av_for_response(struct cm_port *port, struct ib_wc *wc,
 			   grh, &av->ah_attr);
 }
 
-static int cm_init_av_by_path(struct ib_sa_path_rec *path, struct cm_av *av)
+static int cm_init_av_by_path(struct ib_sa_path_rec *path, struct cm_av *av,
+			      struct cm_id_private *cm_id_priv)
 {
 	struct cm_device *cm_dev;
 	struct cm_port *port = NULL;
@@ -387,7 +428,17 @@ static int cm_init_av_by_path(struct ib_sa_path_rec *path, struct cm_av *av)
 			     &av->ah_attr);
 	av->timeout = path->packet_life_time + 1;
 
-	return 0;
+	spin_lock_irqsave(&cm.lock, flags);
+	if (&cm_id_priv->av == av)
+		list_add_tail(&cm_id_priv->prim_list, &port->cm_priv_prim_list);
+	else if (&cm_id_priv->alt_av == av)
+		list_add_tail(&cm_id_priv->altr_list, &port->cm_priv_altr_list);
+	else
+		ret = -EINVAL;
+
+	spin_unlock_irqrestore(&cm.lock, flags);
+
+	return ret;
 }
 
 static int cm_alloc_id(struct cm_id_private *cm_id_priv)
@@ -677,6 +728,8 @@ struct ib_cm_id *ib_create_cm_id(struct ib_device *device,
 	spin_lock_init(&cm_id_priv->lock);
 	init_completion(&cm_id_priv->comp);
 	INIT_LIST_HEAD(&cm_id_priv->work_list);
+	INIT_LIST_HEAD(&cm_id_priv->prim_list);
+	INIT_LIST_HEAD(&cm_id_priv->altr_list);
 	atomic_set(&cm_id_priv->work_count, -1);
 	atomic_set(&cm_id_priv->refcount, 1);
 	return &cm_id_priv->id;
@@ -892,6 +945,15 @@ retest:
 		break;
 	}
 
+	spin_lock_irq(&cm.lock);
+	if (!list_empty(&cm_id_priv->altr_list) &&
+	    (!cm_id_priv->altr_send_port_not_ready))
+		list_del(&cm_id_priv->altr_list);
+	if (!list_empty(&cm_id_priv->prim_list) &&
+	    (!cm_id_priv->prim_send_port_not_ready))
+		list_del(&cm_id_priv->prim_list);
+	spin_unlock_irq(&cm.lock);
+
 	cm_free_id(cm_id->local_id);
 	cm_deref_id(cm_id_priv);
 	wait_for_completion(&cm_id_priv->comp);
@@ -1192,12 +1254,13 @@ int ib_send_cm_req(struct ib_cm_id *cm_id,
 		goto out;
 	}
 
-	ret = cm_init_av_by_path(param->primary_path, &cm_id_priv->av);
+	ret = cm_init_av_by_path(param->primary_path, &cm_id_priv->av,
+				 cm_id_priv);
 	if (ret)
 		goto error1;
 	if (param->alternate_path) {
 		ret = cm_init_av_by_path(param->alternate_path,
-					 &cm_id_priv->alt_av);
+					 &cm_id_priv->alt_av, cm_id_priv);
 		if (ret)
 			goto error1;
 	}
@@ -1639,7 +1702,8 @@ static int cm_req_handler(struct cm_work *work)
 	cm_format_paths_from_req(req_msg, &work->path[0], &work->path[1]);
 
 	memcpy(work->path[0].dmac, cm_id_priv->av.ah_attr.dmac, ETH_ALEN);
-	ret = cm_init_av_by_path(&work->path[0], &cm_id_priv->av);
+	ret = cm_init_av_by_path(&work->path[0], &cm_id_priv->av,
+				 cm_id_priv);
 	if (ret) {
 		ib_get_cached_gid(work->port->cm_dev->ib_device,
 				  work->port->port_num, 0, &work->path[0].sgid,
@@ -1650,7 +1714,8 @@ static int cm_req_handler(struct cm_work *work)
 		goto rejected;
 	}
 	if (req_msg->alt_local_lid) {
-		ret = cm_init_av_by_path(&work->path[1], &cm_id_priv->alt_av);
+		ret = cm_init_av_by_path(&work->path[1], &cm_id_priv->alt_av,
+					 cm_id_priv);
 		if (ret) {
 			ib_send_cm_rej(cm_id, IB_CM_REJ_INVALID_ALT_GID,
 				       &work->path[0].sgid,
@@ -2705,7 +2770,8 @@ int ib_send_cm_lap(struct ib_cm_id *cm_id,
 		goto out;
 	}
 
-	ret = cm_init_av_by_path(alternate_path, &cm_id_priv->alt_av);
+	ret = cm_init_av_by_path(alternate_path, &cm_id_priv->alt_av,
+				 cm_id_priv);
 	if (ret)
 		goto out;
 	cm_id_priv->alt_av.timeout =
@@ -2817,7 +2883,8 @@ static int cm_lap_handler(struct cm_work *work)
 	cm_init_av_for_response(work->port, work->mad_recv_wc->wc,
 				work->mad_recv_wc->recv_buf.grh,
 				&cm_id_priv->av);
-	cm_init_av_by_path(param->alternate_path, &cm_id_priv->alt_av);
+	cm_init_av_by_path(param->alternate_path, &cm_id_priv->alt_av,
+			   cm_id_priv);
 	ret = atomic_inc_and_test(&cm_id_priv->work_count);
 	if (!ret)
 		list_add_tail(&work->list, &cm_id_priv->work_list);
@@ -3009,7 +3076,7 @@ int ib_send_cm_sidr_req(struct ib_cm_id *cm_id,
 		return -EINVAL;
 
 	cm_id_priv = container_of(cm_id, struct cm_id_private, id);
-	ret = cm_init_av_by_path(param->path, &cm_id_priv->av);
+	ret = cm_init_av_by_path(param->path, &cm_id_priv->av, cm_id_priv);
 	if (ret)
 		goto out;
 
@@ -3430,14 +3497,14 @@ static int cm_establish(struct ib_cm_id *cm_id)
 	work->cm_event.event = IB_CM_USER_ESTABLISHED;
 
 	/* Check if the device started its remove_one */
-	spin_lock_irq(&cm.lock);
+	spin_lock_irqsave(&cm.lock, flags);
 	if (!cm_dev->going_down) {
 		queue_delayed_work(cm.wq, &work->work, 0);
 	} else {
 		kfree(work);
 		ret = -ENODEV;
 	}
-	spin_unlock_irq(&cm.lock);
+	spin_unlock_irqrestore(&cm.lock, flags);
 
 out:
 	return ret;
@@ -3446,7 +3513,9 @@ out:
 static int cm_migrate(struct ib_cm_id *cm_id)
 {
 	struct cm_id_private *cm_id_priv;
+	struct cm_av tmp_av;
 	unsigned long flags;
+	int tmp_send_port_not_ready;
 	int ret = 0;
 
 	cm_id_priv = container_of(cm_id, struct cm_id_private, id);
@@ -3455,7 +3524,14 @@ static int cm_migrate(struct ib_cm_id *cm_id)
 	    (cm_id->lap_state == IB_CM_LAP_UNINIT ||
 	     cm_id->lap_state == IB_CM_LAP_IDLE)) {
 		cm_id->lap_state = IB_CM_LAP_IDLE;
+		/* Swap address vector */
+		tmp_av = cm_id_priv->av;
 		cm_id_priv->av = cm_id_priv->alt_av;
+		cm_id_priv->alt_av = tmp_av;
+		/* Swap port send ready state */
+		tmp_send_port_not_ready = cm_id_priv->prim_send_port_not_ready;
+		cm_id_priv->prim_send_port_not_ready = cm_id_priv->altr_send_port_not_ready;
+		cm_id_priv->altr_send_port_not_ready = tmp_send_port_not_ready;
 	} else
 		ret = -EINVAL;
 	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
@@ -3875,6 +3951,9 @@ static void cm_add_one(struct ib_device *ib_device)
 		port->cm_dev = cm_dev;
 		port->port_num = i;
 
+		INIT_LIST_HEAD(&port->cm_priv_prim_list);
+		INIT_LIST_HEAD(&port->cm_priv_altr_list);
+
 		ret = cm_create_port_fs(port);
 		if (ret)
 			goto error1;
@@ -3932,6 +4011,8 @@ static void cm_remove_one(struct ib_device *ib_device, void *client_data)
 {
 	struct cm_device *cm_dev = client_data;
 	struct cm_port *port;
+	struct cm_id_private *cm_id_priv;
+	struct ib_mad_agent *cur_mad_agent;
 	struct ib_port_modify port_modify = {
 		.clr_port_cap_mask = IB_PORT_CM_SUP
 	};
@@ -3955,15 +4036,27 @@ static void cm_remove_one(struct ib_device *ib_device, void *client_data)
 
 		port = cm_dev->port[i-1];
 		ib_modify_port(ib_device, port->port_num, 0, &port_modify);
+		/* Mark all the cm_id's as not valid */
+		spin_lock_irq(&cm.lock);
+		list_for_each_entry(cm_id_priv, &port->cm_priv_altr_list, altr_list)
+			cm_id_priv->altr_send_port_not_ready = 1;
+		list_for_each_entry(cm_id_priv, &port->cm_priv_prim_list, prim_list)
+			cm_id_priv->prim_send_port_not_ready = 1;
+		spin_unlock_irq(&cm.lock);
 		/*
 		 * We flush the queue here after the going_down set, this
 		 * verify that no new works will be queued in the recv handler,
 		 * after that we can call the unregister_mad_agent
 		 */
 		flush_workqueue(cm.wq);
-		ib_unregister_mad_agent(port->mad_agent);
+		spin_lock_irq(&cm.state_lock);
+		cur_mad_agent = port->mad_agent;
+		port->mad_agent = NULL;
+		spin_unlock_irq(&cm.state_lock);
+		ib_unregister_mad_agent(cur_mad_agent);
 		cm_remove_port_fs(port);
 	}
+
 	device_unregister(cm_dev->device);
 	kfree(cm_dev);
 }
@@ -3976,6 +4069,7 @@ static int __init ib_cm_init(void)
 	INIT_LIST_HEAD(&cm.device_list);
 	rwlock_init(&cm.device_lock);
 	spin_lock_init(&cm.lock);
+	spin_lock_init(&cm.state_lock);
 	cm.listen_service_table = RB_ROOT;
 	cm.listen_service_id = be64_to_cpu(IB_CM_ASSIGN_SERVICE_ID);
 	cm.remote_id_table = RB_ROOT;
diff --git a/kernel/drivers/infiniband/core/cma.c b/kernel/drivers/infiniband/core/cma.c
index 17a15c560..c9dcad6a5 100644
--- a/kernel/drivers/infiniband/core/cma.c
+++ b/kernel/drivers/infiniband/core/cma.c
@@ -2578,7 +2578,8 @@ static int cma_bind_addr(struct rdma_cm_id *id, struct sockaddr *src_addr,
 	if (!src_addr || !src_addr->sa_family) {
 		src_addr = (struct sockaddr *) &id->route.addr.src_addr;
 		src_addr->sa_family = dst_addr->sa_family;
-		if (dst_addr->sa_family == AF_INET6) {
+		if (IS_ENABLED(CONFIG_IPV6) &&
+		    dst_addr->sa_family == AF_INET6) {
 			struct sockaddr_in6 *src_addr6 = (struct sockaddr_in6 *) src_addr;
 			struct sockaddr_in6 *dst_addr6 = (struct sockaddr_in6 *) dst_addr;
 			src_addr6->sin6_scope_id = dst_addr6->sin6_scope_id;
diff --git a/kernel/drivers/infiniband/core/iwpm_util.c b/kernel/drivers/infiniband/core/iwpm_util.c
index 5fb089e91..fb43a2428 100644
--- a/kernel/drivers/infiniband/core/iwpm_util.c
+++ b/kernel/drivers/infiniband/core/iwpm_util.c
@@ -634,6 +634,7 @@ static int send_nlmsg_done(struct sk_buff *skb, u8 nl_client, int iwpm_pid)
 	if (!(ibnl_put_msg(skb, &nlh, 0, 0, nl_client,
 			   RDMA_NL_IWPM_MAPINFO, NLM_F_MULTI))) {
 		pr_warn("%s Unable to put NLMSG_DONE\n", __func__);
+		dev_kfree_skb(skb);
 		return -ENOMEM;
 	}
 	nlh->nlmsg_type = NLMSG_DONE;
diff --git a/kernel/drivers/infiniband/core/mad.c b/kernel/drivers/infiniband/core/mad.c
index 2281de122..8d84c563b 100644
--- a/kernel/drivers/infiniband/core/mad.c
+++ b/kernel/drivers/infiniband/core/mad.c
@@ -1745,7 +1745,7 @@ find_mad_agent(struct ib_mad_port_private *port_priv,
 			if (!class)
 				goto out;
 			if (convert_mgmt_class(mad_hdr->mgmt_class) >=
-			    IB_MGMT_MAX_METHODS)
+			    ARRAY_SIZE(class->method_table))
 				goto out;
 			method = class->method_table[convert_mgmt_class(
 							mad_hdr->mgmt_class)];
diff --git a/kernel/drivers/infiniband/core/multicast.c b/kernel/drivers/infiniband/core/multicast.c
index bb6685fb0..2cd97977b 100644
--- a/kernel/drivers/infiniband/core/multicast.c
+++ b/kernel/drivers/infiniband/core/multicast.c
@@ -106,7 +106,6 @@ struct mcast_group {
 	atomic_t		refcount;
 	enum mcast_group_state	state;
 	struct ib_sa_query	*query;
-	int			query_id;
 	u16			pkey_index;
 	u8			leave_state;
 	int			retries;
@@ -339,11 +338,7 @@ static int send_join(struct mcast_group *group, struct mcast_member *member)
 				       member->multicast.comp_mask,
 				       3000, GFP_KERNEL, join_handler, group,
 				       &group->query);
-	if (ret >= 0) {
-		group->query_id = ret;
-		ret = 0;
-	}
-	return ret;
+	return (ret > 0) ? 0 : ret;
 }
 
 static int send_leave(struct mcast_group *group, u8 leave_state)
@@ -363,11 +358,7 @@ static int send_leave(struct mcast_group *group, u8 leave_state)
 				       IB_SA_MCMEMBER_REC_JOIN_STATE,
 				       3000, GFP_KERNEL, leave_handler,
 				       group, &group->query);
-	if (ret >= 0) {
-		group->query_id = ret;
-		ret = 0;
-	}
-	return ret;
+	return (ret > 0) ? 0 : ret;
 }
 
 static void join_group(struct mcast_group *group, struct mcast_member *member,
@@ -526,8 +517,11 @@ static void join_handler(int status, struct ib_sa_mcmember_rec *rec,
 		process_join_error(group, status);
 	else {
 		int mgids_changed, is_mgid0;
-		ib_find_pkey(group->port->dev->device, group->port->port_num,
-			     be16_to_cpu(rec->pkey), &pkey_index);
+
+		if (ib_find_pkey(group->port->dev->device,
+				 group->port->port_num, be16_to_cpu(rec->pkey),
+				 &pkey_index))
+			pkey_index = MCAST_INVALID_PKEY_INDEX;
 
 		spin_lock_irq(&group->port->lock);
 		if (group->state == MCAST_BUSY &&
diff --git a/kernel/drivers/infiniband/core/sa_query.c b/kernel/drivers/infiniband/core/sa_query.c
index a95a32ba5..d3b7ecd10 100644
--- a/kernel/drivers/infiniband/core/sa_query.c
+++ b/kernel/drivers/infiniband/core/sa_query.c
@@ -534,7 +534,7 @@ static int ib_nl_send_msg(struct ib_sa_query *query, gfp_t gfp_mask)
 	data = ibnl_put_msg(skb, &nlh, query->seq, 0, RDMA_NL_LS,
 			    RDMA_NL_LS_OP_RESOLVE, NLM_F_REQUEST);
 	if (!data) {
-		kfree_skb(skb);
+		nlmsg_free(skb);
 		return -EMSGSIZE;
 	}
 
diff --git a/kernel/drivers/infiniband/core/ucm.c b/kernel/drivers/infiniband/core/ucm.c
index 6b4e8a008..564adf311 100644
--- a/kernel/drivers/infiniband/core/ucm.c
+++ b/kernel/drivers/infiniband/core/ucm.c
@@ -48,6 +48,7 @@
 
 #include <asm/uaccess.h>
 
+#include <rdma/ib.h>
 #include <rdma/ib_cm.h>
 #include <rdma/ib_user_cm.h>
 #include <rdma/ib_marshall.h>
@@ -1103,6 +1104,9 @@ static ssize_t ib_ucm_write(struct file *filp, const char __user *buf,
 	struct ib_ucm_cmd_hdr hdr;
 	ssize_t result;
 
+	if (WARN_ON_ONCE(!ib_safe_file_access(filp)))
+		return -EACCES;
+
 	if (len < sizeof(hdr))
 		return -EINVAL;
 
diff --git a/kernel/drivers/infiniband/core/ucma.c b/kernel/drivers/infiniband/core/ucma.c
index 8b5a934e1..886f61ea6 100644
--- a/kernel/drivers/infiniband/core/ucma.c
+++ b/kernel/drivers/infiniband/core/ucma.c
@@ -1574,6 +1574,9 @@ static ssize_t ucma_write(struct file *filp, const char __user *buf,
 	struct rdma_ucm_cmd_hdr hdr;
 	ssize_t ret;
 
+	if (WARN_ON_ONCE(!ib_safe_file_access(filp)))
+		return -EACCES;
+
 	if (len < sizeof(hdr))
 		return -EINVAL;
 
diff --git a/kernel/drivers/infiniband/core/umem.c b/kernel/drivers/infiniband/core/umem.c
index 38acb3cfc..0ae337bec 100644
--- a/kernel/drivers/infiniband/core/umem.c
+++ b/kernel/drivers/infiniband/core/umem.c
@@ -134,6 +134,7 @@ struct ib_umem *ib_umem_get(struct ib_ucontext *context, unsigned long addr,
 		 IB_ACCESS_REMOTE_ATOMIC | IB_ACCESS_MW_BIND));
 
 	if (access & IB_ACCESS_ON_DEMAND) {
+		put_pid(umem->pid);
 		ret = ib_umem_odp_get(context, umem);
 		if (ret) {
 			kfree(umem);
@@ -149,6 +150,7 @@ struct ib_umem *ib_umem_get(struct ib_ucontext *context, unsigned long addr,
 
 	page_list = (struct page **) __get_free_page(GFP_KERNEL);
 	if (!page_list) {
+		put_pid(umem->pid);
 		kfree(umem);
 		return ERR_PTR(-ENOMEM);
 	}
@@ -175,7 +177,7 @@ struct ib_umem *ib_umem_get(struct ib_ucontext *context, unsigned long addr,
 
 	cur_base = addr & PAGE_MASK;
 
-	if (npages == 0) {
+	if (npages == 0 || npages > UINT_MAX) {
 		ret = -EINVAL;
 		goto out;
 	}
diff --git a/kernel/drivers/infiniband/core/uverbs.h b/kernel/drivers/infiniband/core/uverbs.h
index 94bbd8c15..a2d19d136 100644
--- a/kernel/drivers/infiniband/core/uverbs.h
+++ b/kernel/drivers/infiniband/core/uverbs.h
@@ -116,6 +116,7 @@ struct ib_uverbs_event_file {
 struct ib_uverbs_file {
 	struct kref				ref;
 	struct mutex				mutex;
+	struct mutex                            cleanup_mutex; /* protect cleanup */
 	struct ib_uverbs_device		       *device;
 	struct ib_ucontext		       *ucontext;
 	struct ib_event_handler			event_handler;
diff --git a/kernel/drivers/infiniband/core/uverbs_main.c b/kernel/drivers/infiniband/core/uverbs_main.c
index e3ef28861..7becef27c 100644
--- a/kernel/drivers/infiniband/core/uverbs_main.c
+++ b/kernel/drivers/infiniband/core/uverbs_main.c
@@ -48,6 +48,8 @@
 
 #include <asm/uaccess.h>
 
+#include <rdma/ib.h>
+
 #include "uverbs.h"
 
 MODULE_AUTHOR("Roland Dreier");
@@ -242,12 +244,9 @@ static int ib_uverbs_cleanup_ucontext(struct ib_uverbs_file *file,
 			container_of(uobj, struct ib_uqp_object, uevent.uobject);
 
 		idr_remove_uobj(&ib_uverbs_qp_idr, uobj);
-		if (qp != qp->real_qp) {
-			ib_close_qp(qp);
-		} else {
+		if (qp == qp->real_qp)
 			ib_uverbs_detach_umcast(qp, uqp);
-			ib_destroy_qp(qp);
-		}
+		ib_destroy_qp(qp);
 		ib_uverbs_release_uevent(file, &uqp->uevent);
 		kfree(uqp);
 	}
@@ -682,6 +681,9 @@ static ssize_t ib_uverbs_write(struct file *filp, const char __user *buf,
 	int srcu_key;
 	ssize_t ret;
 
+	if (WARN_ON_ONCE(!ib_safe_file_access(filp)))
+		return -EACCES;
+
 	if (count < sizeof hdr)
 		return -EINVAL;
 
@@ -917,6 +919,7 @@ static int ib_uverbs_open(struct inode *inode, struct file *filp)
 	file->async_file = NULL;
 	kref_init(&file->ref);
 	mutex_init(&file->mutex);
+	mutex_init(&file->cleanup_mutex);
 
 	filp->private_data = file;
 	kobject_get(&dev->kobj);
@@ -942,18 +945,20 @@ static int ib_uverbs_close(struct inode *inode, struct file *filp)
 {
 	struct ib_uverbs_file *file = filp->private_data;
 	struct ib_uverbs_device *dev = file->device;
-	struct ib_ucontext *ucontext = NULL;
+
+	mutex_lock(&file->cleanup_mutex);
+	if (file->ucontext) {
+		ib_uverbs_cleanup_ucontext(file, file->ucontext);
+		file->ucontext = NULL;
+	}
+	mutex_unlock(&file->cleanup_mutex);
 
 	mutex_lock(&file->device->lists_mutex);
-	ucontext = file->ucontext;
-	file->ucontext = NULL;
 	if (!file->is_closed) {
 		list_del(&file->list);
 		file->is_closed = 1;
 	}
 	mutex_unlock(&file->device->lists_mutex);
-	if (ucontext)
-		ib_uverbs_cleanup_ucontext(file, ucontext);
 
 	if (file->async_file)
 		kref_put(&file->async_file->ref, ib_uverbs_release_event_file);
@@ -1167,22 +1172,30 @@ static void ib_uverbs_free_hw_resources(struct ib_uverbs_device *uverbs_dev,
 	mutex_lock(&uverbs_dev->lists_mutex);
 	while (!list_empty(&uverbs_dev->uverbs_file_list)) {
 		struct ib_ucontext *ucontext;
-
 		file = list_first_entry(&uverbs_dev->uverbs_file_list,
 					struct ib_uverbs_file, list);
 		file->is_closed = 1;
-		ucontext = file->ucontext;
 		list_del(&file->list);
-		file->ucontext = NULL;
 		kref_get(&file->ref);
 		mutex_unlock(&uverbs_dev->lists_mutex);
-		/* We must release the mutex before going ahead and calling
-		 * disassociate_ucontext. disassociate_ucontext might end up
-		 * indirectly calling uverbs_close, for example due to freeing
-		 * the resources (e.g mmput).
-		 */
+
 		ib_uverbs_event_handler(&file->event_handler, &event);
+
+		mutex_lock(&file->cleanup_mutex);
+		ucontext = file->ucontext;
+		file->ucontext = NULL;
+		mutex_unlock(&file->cleanup_mutex);
+
+		/* At this point ib_uverbs_close cannot be running
+		 * ib_uverbs_cleanup_ucontext
+		 */
 		if (ucontext) {
+			/* We must release the mutex before going ahead and
+			 * calling disassociate_ucontext. disassociate_ucontext
+			 * might end up indirectly calling uverbs_close,
+			 * for example due to freeing the resources
+			 * (e.g mmput).
+			 */
 			ib_dev->disassociate_ucontext(ucontext);
 			ib_uverbs_cleanup_ucontext(file, ucontext);
 		}
diff --git a/kernel/drivers/infiniband/hw/cxgb4/cq.c b/kernel/drivers/infiniband/hw/cxgb4/cq.c
index de9cd6901..bc147582b 100644
--- a/kernel/drivers/infiniband/hw/cxgb4/cq.c
+++ b/kernel/drivers/infiniband/hw/cxgb4/cq.c
@@ -162,7 +162,7 @@ static int create_cq(struct c4iw_rdev *rdev, struct t4_cq *cq,
 	cq->bar2_va = c4iw_bar2_addrs(rdev, cq->cqid, T4_BAR2_QTYPE_INGRESS,
 				      &cq->bar2_qid,
 				      user ? &cq->bar2_pa : NULL);
-	if (user && !cq->bar2_va) {
+	if (user && !cq->bar2_pa) {
 		pr_warn(MOD "%s: cqid %u not in BAR2 range.\n",
 			pci_name(rdev->lldi.pdev), cq->cqid);
 		ret = -EINVAL;
diff --git a/kernel/drivers/infiniband/hw/cxgb4/qp.c b/kernel/drivers/infiniband/hw/cxgb4/qp.c
index aa515afee..53aa7515f 100644
--- a/kernel/drivers/infiniband/hw/cxgb4/qp.c
+++ b/kernel/drivers/infiniband/hw/cxgb4/qp.c
@@ -185,6 +185,10 @@ void __iomem *c4iw_bar2_addrs(struct c4iw_rdev *rdev, unsigned int qid,
 
 	if (pbar2_pa)
 		*pbar2_pa = (rdev->bar2_pa + bar2_qoffset) & PAGE_MASK;
+
+	if (is_t4(rdev->lldi.adapter_type))
+		return NULL;
+
 	return rdev->bar2_kva + bar2_qoffset;
 }
 
@@ -270,7 +274,7 @@ static int create_qp(struct c4iw_rdev *rdev, struct t4_wq *wq,
 	/*
 	 * User mode must have bar2 access.
 	 */
-	if (user && (!wq->sq.bar2_va || !wq->rq.bar2_va)) {
+	if (user && (!wq->sq.bar2_pa || !wq->rq.bar2_pa)) {
 		pr_warn(MOD "%s: sqid %u or rqid %u not in BAR2 range.\n",
 			pci_name(rdev->lldi.pdev), wq->sq.qid, wq->rq.qid);
 		goto free_dma;
diff --git a/kernel/drivers/infiniband/hw/mlx4/ah.c b/kernel/drivers/infiniband/hw/mlx4/ah.c
index 86af71351..fc21bdbb8 100644
--- a/kernel/drivers/infiniband/hw/mlx4/ah.c
+++ b/kernel/drivers/infiniband/hw/mlx4/ah.c
@@ -47,6 +47,7 @@ static struct ib_ah *create_ib_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr,
 
 	ah->av.ib.port_pd = cpu_to_be32(to_mpd(pd)->pdn | (ah_attr->port_num << 24));
 	ah->av.ib.g_slid  = ah_attr->src_path_bits;
+	ah->av.ib.sl_tclass_flowlabel = cpu_to_be32(ah_attr->sl << 28);
 	if (ah_attr->ah_flags & IB_AH_GRH) {
 		ah->av.ib.g_slid   |= 0x80;
 		ah->av.ib.gid_index = ah_attr->grh.sgid_index;
@@ -64,7 +65,6 @@ static struct ib_ah *create_ib_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr,
 		       !(1 << ah->av.ib.stat_rate & dev->caps.stat_rate_support))
 			--ah->av.ib.stat_rate;
 	}
-	ah->av.ib.sl_tclass_flowlabel = cpu_to_be32(ah_attr->sl << 28);
 
 	return &ah->ibah;
 }
@@ -102,7 +102,10 @@ static struct ib_ah *create_iboe_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr
 	if (vlan_tag < 0x1000)
 		vlan_tag |= (ah_attr->sl & 7) << 13;
 	ah->av.eth.port_pd = cpu_to_be32(to_mpd(pd)->pdn | (ah_attr->port_num << 24));
-	ah->av.eth.gid_index = mlx4_ib_gid_index_to_real_index(ibdev, ah_attr->port_num, ah_attr->grh.sgid_index);
+	ret = mlx4_ib_gid_index_to_real_index(ibdev, ah_attr->port_num, ah_attr->grh.sgid_index);
+	if (ret < 0)
+		return ERR_PTR(ret);
+	ah->av.eth.gid_index = ret;
 	ah->av.eth.vlan = cpu_to_be16(vlan_tag);
 	if (ah_attr->static_rate) {
 		ah->av.eth.stat_rate = ah_attr->static_rate + MLX4_STAT_RATE_OFFSET;
@@ -110,7 +113,9 @@ static struct ib_ah *create_iboe_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr
 		       !(1 << ah->av.eth.stat_rate & dev->caps.stat_rate_support))
 			--ah->av.eth.stat_rate;
 	}
-
+	ah->av.eth.sl_tclass_flowlabel |=
+			cpu_to_be32((ah_attr->grh.traffic_class << 20) |
+				    ah_attr->grh.flow_label);
 	/*
 	 * HW requires multicast LID so we just choose one.
 	 */
@@ -118,7 +123,7 @@ static struct ib_ah *create_iboe_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr
 		ah->av.ib.dlid = cpu_to_be16(0xc000);
 
 	memcpy(ah->av.eth.dgid, ah_attr->grh.dgid.raw, 16);
-	ah->av.eth.sl_tclass_flowlabel = cpu_to_be32(ah_attr->sl << 29);
+	ah->av.eth.sl_tclass_flowlabel |= cpu_to_be32(ah_attr->sl << 29);
 
 	return &ah->ibah;
 }
diff --git a/kernel/drivers/infiniband/hw/mlx4/cq.c b/kernel/drivers/infiniband/hw/mlx4/cq.c
index b88fc8f5a..57e1a0892 100644
--- a/kernel/drivers/infiniband/hw/mlx4/cq.c
+++ b/kernel/drivers/infiniband/hw/mlx4/cq.c
@@ -253,11 +253,14 @@ struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev,
 	if (context)
 		if (ib_copy_to_udata(udata, &cq->mcq.cqn, sizeof (__u32))) {
 			err = -EFAULT;
-			goto err_dbmap;
+			goto err_cq_free;
 		}
 
 	return &cq->ibcq;
 
+err_cq_free:
+	mlx4_cq_free(dev->dev, &cq->mcq);
+
 err_dbmap:
 	if (context)
 		mlx4_ib_db_unmap_user(to_mucontext(context), &cq->db);
diff --git a/kernel/drivers/infiniband/hw/mlx4/mad.c b/kernel/drivers/infiniband/hw/mlx4/mad.c
index 870e56b6b..d862b9b79 100644
--- a/kernel/drivers/infiniband/hw/mlx4/mad.c
+++ b/kernel/drivers/infiniband/hw/mlx4/mad.c
@@ -526,7 +526,7 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port,
 		tun_tx_ix = (++tun_qp->tx_ix_head) & (MLX4_NUM_TUNNEL_BUFS - 1);
 	spin_unlock(&tun_qp->tx_lock);
 	if (ret)
-		goto out;
+		goto end;
 
 	tun_mad = (struct mlx4_rcv_tunnel_mad *) (tun_qp->tx_ring[tun_tx_ix].buf.addr);
 	if (tun_qp->tx_ring[tun_tx_ix].ah)
@@ -595,9 +595,15 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port,
 	wr.wr.send_flags = IB_SEND_SIGNALED;
 
 	ret = ib_post_send(src_qp, &wr.wr, &bad_wr);
-out:
-	if (ret)
-		ib_destroy_ah(ah);
+	if (!ret)
+		return 0;
+ out:
+	spin_lock(&tun_qp->tx_lock);
+	tun_qp->tx_ix_tail++;
+	spin_unlock(&tun_qp->tx_lock);
+	tun_qp->tx_ring[tun_tx_ix].ah = NULL;
+end:
+	ib_destroy_ah(ah);
 	return ret;
 }
 
@@ -1074,6 +1080,27 @@ void handle_port_mgmt_change_event(struct work_struct *work)
 
 		/* Generate GUID changed event */
 		if (changed_attr & MLX4_EQ_PORT_INFO_GID_PFX_CHANGE_MASK) {
+			if (mlx4_is_master(dev->dev)) {
+				union ib_gid gid;
+				int err = 0;
+
+				if (!eqe->event.port_mgmt_change.params.port_info.gid_prefix)
+					err = __mlx4_ib_query_gid(&dev->ib_dev, port, 0, &gid, 1);
+				else
+					gid.global.subnet_prefix =
+						eqe->event.port_mgmt_change.params.port_info.gid_prefix;
+				if (err) {
+					pr_warn("Could not change QP1 subnet prefix for port %d: query_gid error (%d)\n",
+						port, err);
+				} else {
+					pr_debug("Changing QP1 subnet prefix for port %d. old=0x%llx. new=0x%llx\n",
+						 port,
+						 (u64)atomic64_read(&dev->sriov.demux[port - 1].subnet_prefix),
+						 be64_to_cpu(gid.global.subnet_prefix));
+					atomic64_set(&dev->sriov.demux[port - 1].subnet_prefix,
+						     be64_to_cpu(gid.global.subnet_prefix));
+				}
+			}
 			mlx4_ib_dispatch_event(dev, port, IB_EVENT_GID_CHANGE);
 			/*if master, notify all slaves*/
 			if (mlx4_is_master(dev->dev))
@@ -1278,9 +1305,15 @@ int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port,
 
 
 	ret = ib_post_send(send_qp, &wr.wr, &bad_wr);
+	if (!ret)
+		return 0;
+
+	spin_lock(&sqp->tx_lock);
+	sqp->tx_ix_tail++;
+	spin_unlock(&sqp->tx_lock);
+	sqp->tx_ring[wire_tx_ix].ah = NULL;
 out:
-	if (ret)
-		ib_destroy_ah(ah);
+	ib_destroy_ah(ah);
 	return ret;
 }
 
@@ -2142,6 +2175,8 @@ int mlx4_ib_init_sriov(struct mlx4_ib_dev *dev)
 		if (err)
 			goto demux_err;
 		dev->sriov.demux[i].guid_cache[0] = gid.global.interface_id;
+		atomic64_set(&dev->sriov.demux[i].subnet_prefix,
+			     be64_to_cpu(gid.global.subnet_prefix));
 		err = alloc_pv_object(dev, mlx4_master_func_num(dev->dev), i + 1,
 				      &dev->sriov.sqps[i]);
 		if (err)
diff --git a/kernel/drivers/infiniband/hw/mlx4/main.c b/kernel/drivers/infiniband/hw/mlx4/main.c
index 97d6878f9..77ddf2fa8 100644
--- a/kernel/drivers/infiniband/hw/mlx4/main.c
+++ b/kernel/drivers/infiniband/hw/mlx4/main.c
@@ -630,9 +630,11 @@ static int eth_link_query_port(struct ib_device *ibdev, u8 port,
 	if (err)
 		goto out;
 
-	props->active_width	=  (((u8 *)mailbox->buf)[5] == 0x40) ?
-						IB_WIDTH_4X : IB_WIDTH_1X;
-	props->active_speed	= IB_SPEED_QDR;
+	props->active_width	=  (((u8 *)mailbox->buf)[5] == 0x40) ||
+				   (((u8 *)mailbox->buf)[5] == 0x20 /*56Gb*/) ?
+					   IB_WIDTH_4X : IB_WIDTH_1X;
+	props->active_speed	=  (((u8 *)mailbox->buf)[5] == 0x20 /*56Gb*/) ?
+					   IB_SPEED_FDR : IB_SPEED_QDR;
 	props->port_cap_flags	= IB_PORT_CM_SUP | IB_PORT_IP_BASED_GIDS;
 	props->gid_tbl_len	= mdev->dev->caps.gid_table_len[port];
 	props->max_msg_sz	= mdev->dev->caps.max_msg_sz;
@@ -2401,14 +2403,19 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 			goto err_steer_qp_release;
 		}
 
-		bitmap_zero(ibdev->ib_uc_qpns_bitmap, ibdev->steer_qpn_count);
-
-		err = mlx4_FLOW_STEERING_IB_UC_QP_RANGE(
-				dev, ibdev->steer_qpn_base,
-				ibdev->steer_qpn_base +
-				ibdev->steer_qpn_count - 1);
-		if (err)
-			goto err_steer_free_bitmap;
+		if (dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_DMFS_IPOIB) {
+			bitmap_zero(ibdev->ib_uc_qpns_bitmap,
+				    ibdev->steer_qpn_count);
+			err = mlx4_FLOW_STEERING_IB_UC_QP_RANGE(
+					dev, ibdev->steer_qpn_base,
+					ibdev->steer_qpn_base +
+					ibdev->steer_qpn_count - 1);
+			if (err)
+				goto err_steer_free_bitmap;
+		} else {
+			bitmap_fill(ibdev->ib_uc_qpns_bitmap,
+				    ibdev->steer_qpn_count);
+		}
 	}
 
 	for (j = 1; j <= ibdev->dev->caps.num_ports; j++)
diff --git a/kernel/drivers/infiniband/hw/mlx4/mcg.c b/kernel/drivers/infiniband/hw/mlx4/mcg.c
index 99451d887..36ec8aa04 100644
--- a/kernel/drivers/infiniband/hw/mlx4/mcg.c
+++ b/kernel/drivers/infiniband/hw/mlx4/mcg.c
@@ -489,7 +489,7 @@ static u8 get_leave_state(struct mcast_group *group)
 		if (!group->members[i])
 			leave_state |= (1 << i);
 
-	return leave_state & (group->rec.scope_join_state & 7);
+	return leave_state & (group->rec.scope_join_state & 0xf);
 }
 
 static int join_group(struct mcast_group *group, int slave, u8 join_mask)
@@ -564,8 +564,8 @@ static void mlx4_ib_mcg_timeout_handler(struct work_struct *work)
 		} else
 			mcg_warn_group(group, "DRIVER BUG\n");
 	} else if (group->state == MCAST_LEAVE_SENT) {
-		if (group->rec.scope_join_state & 7)
-			group->rec.scope_join_state &= 0xf8;
+		if (group->rec.scope_join_state & 0xf)
+			group->rec.scope_join_state &= 0xf0;
 		group->state = MCAST_IDLE;
 		mutex_unlock(&group->lock);
 		if (release_group(group, 1))
@@ -605,7 +605,7 @@ static int handle_leave_req(struct mcast_group *group, u8 leave_mask,
 static int handle_join_req(struct mcast_group *group, u8 join_mask,
 			   struct mcast_req *req)
 {
-	u8 group_join_state = group->rec.scope_join_state & 7;
+	u8 group_join_state = group->rec.scope_join_state & 0xf;
 	int ref = 0;
 	u16 status;
 	struct ib_sa_mcmember_data *sa_data = (struct ib_sa_mcmember_data *)req->sa_mad.data;
@@ -690,8 +690,8 @@ static void mlx4_ib_mcg_work_handler(struct work_struct *work)
 			u8 cur_join_state;
 
 			resp_join_state = ((struct ib_sa_mcmember_data *)
-						group->response_sa_mad.data)->scope_join_state & 7;
-			cur_join_state = group->rec.scope_join_state & 7;
+						group->response_sa_mad.data)->scope_join_state & 0xf;
+			cur_join_state = group->rec.scope_join_state & 0xf;
 
 			if (method == IB_MGMT_METHOD_GET_RESP) {
 				/* successfull join */
@@ -710,7 +710,7 @@ process_requests:
 		req = list_first_entry(&group->pending_list, struct mcast_req,
 				       group_list);
 		sa_data = (struct ib_sa_mcmember_data *)req->sa_mad.data;
-		req_join_state = sa_data->scope_join_state & 0x7;
+		req_join_state = sa_data->scope_join_state & 0xf;
 
 		/* For a leave request, we will immediately answer the VF, and
 		 * update our internal counters. The actual leave will be sent
diff --git a/kernel/drivers/infiniband/hw/mlx4/mlx4_ib.h b/kernel/drivers/infiniband/hw/mlx4/mlx4_ib.h
index 1caa11eda..78f29e916 100644
--- a/kernel/drivers/infiniband/hw/mlx4/mlx4_ib.h
+++ b/kernel/drivers/infiniband/hw/mlx4/mlx4_ib.h
@@ -441,7 +441,7 @@ struct mlx4_ib_demux_ctx {
 	struct workqueue_struct *wq;
 	struct workqueue_struct *ud_wq;
 	spinlock_t ud_lock;
-	__be64 subnet_prefix;
+	atomic64_t subnet_prefix;
 	__be64 guid_cache[128];
 	struct mlx4_ib_dev *dev;
 	/* the following lock protects both mcg_table and mcg_mgid0_list */
diff --git a/kernel/drivers/infiniband/hw/mlx4/qp.c b/kernel/drivers/infiniband/hw/mlx4/qp.c
index 13eaaf452..1c8b7c22c 100644
--- a/kernel/drivers/infiniband/hw/mlx4/qp.c
+++ b/kernel/drivers/infiniband/hw/mlx4/qp.c
@@ -357,7 +357,7 @@ static int send_wqe_overhead(enum mlx4_ib_qp_type type, u32 flags)
 			sizeof (struct mlx4_wqe_raddr_seg);
 	case MLX4_IB_QPT_RC:
 		return sizeof (struct mlx4_wqe_ctrl_seg) +
-			sizeof (struct mlx4_wqe_atomic_seg) +
+			sizeof (struct mlx4_wqe_masked_atomic_seg) +
 			sizeof (struct mlx4_wqe_raddr_seg);
 	case MLX4_IB_QPT_SMI:
 	case MLX4_IB_QPT_GSI:
@@ -1162,8 +1162,10 @@ struct ib_qp *mlx4_ib_create_qp(struct ib_pd *pd,
 	{
 		err = create_qp_common(to_mdev(pd->device), pd, init_attr,
 				       udata, 0, &qp, gfp);
-		if (err)
+		if (err) {
+			kfree(qp);
 			return ERR_PTR(err);
+		}
 
 		qp->ibqp.qp_num = qp->mqp.qpn;
 		qp->xrcdn = xrcdn;
@@ -1205,7 +1207,8 @@ int mlx4_ib_destroy_qp(struct ib_qp *qp)
 	if (is_qp0(dev, mqp))
 		mlx4_CLOSE_PORT(dev->dev, mqp->port);
 
-	if (dev->qp1_proxy[mqp->port - 1] == mqp) {
+	if (mqp->mlx4_ib_qp_type == MLX4_IB_QPT_PROXY_GSI &&
+	    dev->qp1_proxy[mqp->port - 1] == mqp) {
 		mutex_lock(&dev->qp1_proxy_lock[mqp->port - 1]);
 		dev->qp1_proxy[mqp->port - 1] = NULL;
 		mutex_unlock(&dev->qp1_proxy_lock[mqp->port - 1]);
@@ -2329,24 +2332,27 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr,
 		sqp->ud_header.grh.flow_label    =
 			ah->av.ib.sl_tclass_flowlabel & cpu_to_be32(0xfffff);
 		sqp->ud_header.grh.hop_limit     = ah->av.ib.hop_limit;
-		if (is_eth)
+		if (is_eth) {
 			memcpy(sqp->ud_header.grh.source_gid.raw, sgid.raw, 16);
-		else {
-		if (mlx4_is_mfunc(to_mdev(ib_dev)->dev)) {
-			/* When multi-function is enabled, the ib_core gid
-			 * indexes don't necessarily match the hw ones, so
-			 * we must use our own cache */
-			sqp->ud_header.grh.source_gid.global.subnet_prefix =
-				to_mdev(ib_dev)->sriov.demux[sqp->qp.port - 1].
-						       subnet_prefix;
-			sqp->ud_header.grh.source_gid.global.interface_id =
-				to_mdev(ib_dev)->sriov.demux[sqp->qp.port - 1].
-					       guid_cache[ah->av.ib.gid_index];
-		} else
-			ib_get_cached_gid(ib_dev,
-					  be32_to_cpu(ah->av.ib.port_pd) >> 24,
-					  ah->av.ib.gid_index,
-					  &sqp->ud_header.grh.source_gid, NULL);
+		} else {
+			if (mlx4_is_mfunc(to_mdev(ib_dev)->dev)) {
+				/* When multi-function is enabled, the ib_core gid
+				 * indexes don't necessarily match the hw ones, so
+				 * we must use our own cache
+				 */
+				sqp->ud_header.grh.source_gid.global.subnet_prefix =
+					cpu_to_be64(atomic64_read(&(to_mdev(ib_dev)->sriov.
+								    demux[sqp->qp.port - 1].
+								    subnet_prefix)));
+				sqp->ud_header.grh.source_gid.global.interface_id =
+					to_mdev(ib_dev)->sriov.demux[sqp->qp.port - 1].
+						       guid_cache[ah->av.ib.gid_index];
+			} else {
+				ib_get_cached_gid(ib_dev,
+						  be32_to_cpu(ah->av.ib.port_pd) >> 24,
+						  ah->av.ib.gid_index,
+						  &sqp->ud_header.grh.source_gid, NULL);
+			}
 		}
 		memcpy(sqp->ud_header.grh.destination_gid.raw,
 		       ah->av.ib.dgid, 16);
diff --git a/kernel/drivers/infiniband/hw/mlx5/cq.c b/kernel/drivers/infiniband/hw/mlx5/cq.c
index 92ddae101..02c8deab1 100644
--- a/kernel/drivers/infiniband/hw/mlx5/cq.c
+++ b/kernel/drivers/infiniband/hw/mlx5/cq.c
@@ -763,7 +763,8 @@ struct ib_cq *mlx5_ib_create_cq(struct ib_device *ibdev,
 	if (attr->flags)
 		return ERR_PTR(-EINVAL);
 
-	if (entries < 0)
+	if (entries < 0 ||
+	    (entries > (1 << MLX5_CAP_GEN(dev->mdev, log_max_cq_sz))))
 		return ERR_PTR(-EINVAL);
 
 	entries = roundup_pow_of_two(entries + 1);
@@ -786,8 +787,7 @@ struct ib_cq *mlx5_ib_create_cq(struct ib_device *ibdev,
 		if (err)
 			goto err_create;
 	} else {
-		/* for now choose 64 bytes till we have a proper interface */
-		cqe_size = 64;
+		cqe_size = cache_line_size() == 128 ? 128 : 64;
 		err = create_cq_kernel(dev, cq, entries, cqe_size, &cqb,
 				       &index, &inlen);
 		if (err)
@@ -1094,11 +1094,16 @@ int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata)
 		return -ENOSYS;
 	}
 
-	if (entries < 1)
+	if (entries < 1 ||
+	    entries > (1 << MLX5_CAP_GEN(dev->mdev, log_max_cq_sz))) {
+		mlx5_ib_warn(dev, "wrong entries number %d, max %d\n",
+			     entries,
+			     1 << MLX5_CAP_GEN(dev->mdev, log_max_cq_sz));
 		return -EINVAL;
+	}
 
 	entries = roundup_pow_of_two(entries + 1);
-	if (entries >  (1 << MLX5_CAP_GEN(dev->mdev, log_max_cq_sz)) + 1)
+	if (entries > (1 << MLX5_CAP_GEN(dev->mdev, log_max_cq_sz)) + 1)
 		return -EINVAL;
 
 	if (entries == ibcq->cqe + 1)
diff --git a/kernel/drivers/infiniband/hw/mlx5/main.c b/kernel/drivers/infiniband/hw/mlx5/main.c
index c4e091528..2a1fdcaa3 100644
--- a/kernel/drivers/infiniband/hw/mlx5/main.c
+++ b/kernel/drivers/infiniband/hw/mlx5/main.c
@@ -273,7 +273,7 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
 		     sizeof(struct mlx5_wqe_ctrl_seg)) /
 		     sizeof(struct mlx5_wqe_data_seg);
 	props->max_sge = min(max_rq_sg, max_sq_sg);
-	props->max_sge_rd = props->max_sge;
+	props->max_sge_rd	   = MLX5_MAX_SGE_RD;
 	props->max_cq		   = 1 << MLX5_CAP_GEN(mdev, log_max_cq);
 	props->max_cqe = (1 << MLX5_CAP_GEN(mdev, log_max_cq_sz)) - 1;
 	props->max_mr		   = 1 << MLX5_CAP_GEN(mdev, log_max_mkey);
@@ -405,8 +405,8 @@ static int mlx5_query_hca_port(struct ib_device *ibdev, u8 port,
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
 	struct mlx5_core_dev *mdev = dev->mdev;
 	struct mlx5_hca_vport_context *rep;
-	int max_mtu;
-	int oper_mtu;
+	u16 max_mtu;
+	u16 oper_mtu;
 	int err;
 	u8 ib_link_width_oper;
 	u8 vl_hw_cap;
@@ -947,13 +947,13 @@ static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context,
 {
 	struct mlx5_ib_dev *ibdev = (struct mlx5_ib_dev *)context;
 	struct ib_event ibev;
-
+	bool fatal = false;
 	u8 port = 0;
 
 	switch (event) {
 	case MLX5_DEV_EVENT_SYS_ERROR:
-		ibdev->ib_active = false;
 		ibev.event = IB_EVENT_DEVICE_FATAL;
+		fatal = true;
 		break;
 
 	case MLX5_DEV_EVENT_PORT_UP:
@@ -962,14 +962,11 @@ static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context,
 		break;
 
 	case MLX5_DEV_EVENT_PORT_DOWN:
+	case MLX5_DEV_EVENT_PORT_INITIALIZED:
 		ibev.event = IB_EVENT_PORT_ERR;
 		port = (u8)param;
 		break;
 
-	case MLX5_DEV_EVENT_PORT_INITIALIZED:
-		/* not used by ULPs */
-		return;
-
 	case MLX5_DEV_EVENT_LID_CHANGE:
 		ibev.event = IB_EVENT_LID_CHANGE;
 		port = (u8)param;
@@ -1001,6 +998,9 @@ static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context,
 
 	if (ibdev->ib_active)
 		ib_dispatch_event(&ibev);
+
+	if (fatal)
+		ibdev->ib_active = false;
 }
 
 static void get_ext_port_caps(struct mlx5_ib_dev *dev)
diff --git a/kernel/drivers/infiniband/hw/mlx5/mr.c b/kernel/drivers/infiniband/hw/mlx5/mr.c
index 6000f7aee..3399271c2 100644
--- a/kernel/drivers/infiniband/hw/mlx5/mr.c
+++ b/kernel/drivers/infiniband/hw/mlx5/mr.c
@@ -614,6 +614,33 @@ int mlx5_mr_cache_init(struct mlx5_ib_dev *dev)
 	return 0;
 }
 
+static void wait_for_async_commands(struct mlx5_ib_dev *dev)
+{
+	struct mlx5_mr_cache *cache = &dev->cache;
+	struct mlx5_cache_ent *ent;
+	int total = 0;
+	int i;
+	int j;
+
+	for (i = 0; i < MAX_MR_CACHE_ENTRIES; i++) {
+		ent = &cache->ent[i];
+		for (j = 0 ; j < 1000; j++) {
+			if (!ent->pending)
+				break;
+			msleep(50);
+		}
+	}
+	for (i = 0; i < MAX_MR_CACHE_ENTRIES; i++) {
+		ent = &cache->ent[i];
+		total += ent->pending;
+	}
+
+	if (total)
+		mlx5_ib_warn(dev, "aborted while there are %d pending mr requests\n", total);
+	else
+		mlx5_ib_warn(dev, "done with all pending requests\n");
+}
+
 int mlx5_mr_cache_cleanup(struct mlx5_ib_dev *dev)
 {
 	int i;
@@ -627,6 +654,7 @@ int mlx5_mr_cache_cleanup(struct mlx5_ib_dev *dev)
 		clean_keys(dev, i);
 
 	destroy_workqueue(dev->cache.wq);
+	wait_for_async_commands(dev);
 	del_timer_sync(&dev->delay_timer);
 
 	return 0;
diff --git a/kernel/drivers/infiniband/hw/mlx5/qp.c b/kernel/drivers/infiniband/hw/mlx5/qp.c
index 307bdbca8..cfcfbb6b8 100644
--- a/kernel/drivers/infiniband/hw/mlx5/qp.c
+++ b/kernel/drivers/infiniband/hw/mlx5/qp.c
@@ -226,6 +226,8 @@ static int set_rq_size(struct mlx5_ib_dev *dev, struct ib_qp_cap *cap,
 		qp->rq.max_gs = 0;
 		qp->rq.wqe_cnt = 0;
 		qp->rq.wqe_shift = 0;
+		cap->max_recv_wr = 0;
+		cap->max_recv_sge = 0;
 	} else {
 		if (ucmd) {
 			qp->rq.wqe_cnt = ucmd->rq_wqe_count;
@@ -2525,10 +2527,11 @@ static u8 get_fence(u8 fence, struct ib_send_wr *wr)
 			return MLX5_FENCE_MODE_SMALL_AND_FENCE;
 		else
 			return fence;
-
-	} else {
-		return 0;
+	} else if (unlikely(wr->send_flags & IB_SEND_FENCE)) {
+		return MLX5_FENCE_MODE_FENCE;
 	}
+
+	return 0;
 }
 
 static int begin_wqe(struct mlx5_ib_qp *qp, void **seg,
@@ -3092,17 +3095,19 @@ int mlx5_ib_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, int qp_attr
 	qp_attr->cap.max_recv_sge    = qp->rq.max_gs;
 
 	if (!ibqp->uobject) {
-		qp_attr->cap.max_send_wr  = qp->sq.wqe_cnt;
+		qp_attr->cap.max_send_wr  = qp->sq.max_post;
 		qp_attr->cap.max_send_sge = qp->sq.max_gs;
+		qp_init_attr->qp_context = ibqp->qp_context;
 	} else {
 		qp_attr->cap.max_send_wr  = 0;
 		qp_attr->cap.max_send_sge = 0;
 	}
 
-	/* We don't support inline sends for kernel QPs (yet), and we
-	 * don't know what userspace's value should be.
-	 */
-	qp_attr->cap.max_inline_data = 0;
+	qp_init_attr->qp_type = ibqp->qp_type;
+	qp_init_attr->recv_cq = ibqp->recv_cq;
+	qp_init_attr->send_cq = ibqp->send_cq;
+	qp_init_attr->srq = ibqp->srq;
+	qp_attr->cap.max_inline_data = qp->max_inline_data;
 
 	qp_init_attr->cap	     = qp_attr->cap;
 
diff --git a/kernel/drivers/infiniband/hw/qib/qib_file_ops.c b/kernel/drivers/infiniband/hw/qib/qib_file_ops.c
index e449e3949..24f4a782e 100644
--- a/kernel/drivers/infiniband/hw/qib/qib_file_ops.c
+++ b/kernel/drivers/infiniband/hw/qib/qib_file_ops.c
@@ -45,6 +45,8 @@
 #include <linux/export.h>
 #include <linux/uio.h>
 
+#include <rdma/ib.h>
+
 #include "qib.h"
 #include "qib_common.h"
 #include "qib_user_sdma.h"
@@ -2067,6 +2069,9 @@ static ssize_t qib_write(struct file *fp, const char __user *data,
 	ssize_t ret = 0;
 	void *dest;
 
+	if (WARN_ON_ONCE(!ib_safe_file_access(fp)))
+		return -EACCES;
+
 	if (count < sizeof(cmd.type)) {
 		ret = -EINVAL;
 		goto bail;
diff --git a/kernel/drivers/infiniband/ulp/ipoib/ipoib.h b/kernel/drivers/infiniband/ulp/ipoib/ipoib.h
index 3ede10309..07cfcc326 100644
--- a/kernel/drivers/infiniband/ulp/ipoib/ipoib.h
+++ b/kernel/drivers/infiniband/ulp/ipoib/ipoib.h
@@ -63,6 +63,8 @@ enum ipoib_flush_level {
 
 enum {
 	IPOIB_ENCAP_LEN		  = 4,
+	IPOIB_PSEUDO_LEN	  = 20,
+	IPOIB_HARD_LEN		  = IPOIB_ENCAP_LEN + IPOIB_PSEUDO_LEN,
 
 	IPOIB_UD_HEAD_SIZE	  = IB_GRH_BYTES + IPOIB_ENCAP_LEN,
 	IPOIB_UD_RX_SG		  = 2, /* max buffer needed for 4K mtu */
@@ -131,15 +133,21 @@ struct ipoib_header {
 	u16	reserved;
 };
 
-struct ipoib_cb {
-	struct qdisc_skb_cb	qdisc_cb;
-	u8			hwaddr[INFINIBAND_ALEN];
+struct ipoib_pseudo_header {
+	u8	hwaddr[INFINIBAND_ALEN];
 };
 
-static inline struct ipoib_cb *ipoib_skb_cb(const struct sk_buff *skb)
+static inline void skb_add_pseudo_hdr(struct sk_buff *skb)
 {
-	BUILD_BUG_ON(sizeof(skb->cb) < sizeof(struct ipoib_cb));
-	return (struct ipoib_cb *)skb->cb;
+	char *data = skb_push(skb, IPOIB_PSEUDO_LEN);
+
+	/*
+	 * only the ipoib header is present now, make room for a dummy
+	 * pseudo header and set skb field accordingly
+	 */
+	memset(data, 0, IPOIB_PSEUDO_LEN);
+	skb_reset_mac_header(skb);
+	skb_pull(skb, IPOIB_HARD_LEN);
 }
 
 /* Used for all multicast joins (broadcast, IPv4 mcast and IPv6 mcast) */
@@ -472,6 +480,7 @@ void ipoib_send(struct net_device *dev, struct sk_buff *skb,
 		struct ipoib_ah *address, u32 qpn);
 void ipoib_reap_ah(struct work_struct *work);
 
+struct ipoib_path *__path_find(struct net_device *dev, void *gid);
 void ipoib_mark_paths_invalid(struct net_device *dev);
 void ipoib_flush_paths(struct net_device *dev);
 struct ipoib_dev_priv *ipoib_intf_alloc(const char *format);
diff --git a/kernel/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/kernel/drivers/infiniband/ulp/ipoib/ipoib_cm.c
index 3ae9726ef..3ba7de5f9 100644
--- a/kernel/drivers/infiniband/ulp/ipoib/ipoib_cm.c
+++ b/kernel/drivers/infiniband/ulp/ipoib/ipoib_cm.c
@@ -63,6 +63,8 @@ MODULE_PARM_DESC(cm_data_debug_level,
 #define IPOIB_CM_RX_DELAY       (3 * 256 * HZ)
 #define IPOIB_CM_RX_UPDATE_MASK (0x3)
 
+#define IPOIB_CM_RX_RESERVE     (ALIGN(IPOIB_HARD_LEN, 16) - IPOIB_ENCAP_LEN)
+
 static struct ib_qp_attr ipoib_cm_err_attr = {
 	.qp_state = IB_QPS_ERR
 };
@@ -147,15 +149,15 @@ static struct sk_buff *ipoib_cm_alloc_rx_skb(struct net_device *dev,
 	struct sk_buff *skb;
 	int i;
 
-	skb = dev_alloc_skb(IPOIB_CM_HEAD_SIZE + 12);
+	skb = dev_alloc_skb(ALIGN(IPOIB_CM_HEAD_SIZE + IPOIB_PSEUDO_LEN, 16));
 	if (unlikely(!skb))
 		return NULL;
 
 	/*
-	 * IPoIB adds a 4 byte header. So we need 12 more bytes to align the
+	 * IPoIB adds a IPOIB_ENCAP_LEN byte header, this will align the
 	 * IP header to a multiple of 16.
 	 */
-	skb_reserve(skb, 12);
+	skb_reserve(skb, IPOIB_CM_RX_RESERVE);
 
 	mapping[0] = ib_dma_map_single(priv->ca, skb->data, IPOIB_CM_HEAD_SIZE,
 				       DMA_FROM_DEVICE);
@@ -624,9 +626,9 @@ void ipoib_cm_handle_rx_wc(struct net_device *dev, struct ib_wc *wc)
 	if (wc->byte_len < IPOIB_CM_COPYBREAK) {
 		int dlen = wc->byte_len;
 
-		small_skb = dev_alloc_skb(dlen + 12);
+		small_skb = dev_alloc_skb(dlen + IPOIB_CM_RX_RESERVE);
 		if (small_skb) {
-			skb_reserve(small_skb, 12);
+			skb_reserve(small_skb, IPOIB_CM_RX_RESERVE);
 			ib_dma_sync_single_for_cpu(priv->ca, rx_ring[wr_id].mapping[0],
 						   dlen, DMA_FROM_DEVICE);
 			skb_copy_from_linear_data(skb, small_skb->data, dlen);
@@ -663,8 +665,7 @@ void ipoib_cm_handle_rx_wc(struct net_device *dev, struct ib_wc *wc)
 
 copied:
 	skb->protocol = ((struct ipoib_header *) skb->data)->proto;
-	skb_reset_mac_header(skb);
-	skb_pull(skb, IPOIB_ENCAP_LEN);
+	skb_add_pseudo_hdr(skb);
 
 	++dev->stats.rx_packets;
 	dev->stats.rx_bytes += skb->len;
@@ -1035,8 +1036,6 @@ static struct ib_qp *ipoib_cm_create_tx_qp(struct net_device *dev, struct ipoib_
 
 	tx_qp = ib_create_qp(priv->pd, &attr);
 	if (PTR_ERR(tx_qp) == -EINVAL) {
-		ipoib_warn(priv, "can't use GFP_NOIO for QPs on device %s, using GFP_KERNEL\n",
-			   priv->ca->name);
 		attr.create_flags &= ~IB_QP_CREATE_USE_GFP_NOIO;
 		tx_qp = ib_create_qp(priv->pd, &attr);
 	}
@@ -1299,6 +1298,8 @@ void ipoib_cm_destroy_tx(struct ipoib_cm_tx *tx)
 	}
 }
 
+#define QPN_AND_OPTIONS_OFFSET	4
+
 static void ipoib_cm_tx_start(struct work_struct *work)
 {
 	struct ipoib_dev_priv *priv = container_of(work, struct ipoib_dev_priv,
@@ -1307,6 +1308,7 @@ static void ipoib_cm_tx_start(struct work_struct *work)
 	struct ipoib_neigh *neigh;
 	struct ipoib_cm_tx *p;
 	unsigned long flags;
+	struct ipoib_path *path;
 	int ret;
 
 	struct ib_sa_path_rec pathrec;
@@ -1319,7 +1321,19 @@ static void ipoib_cm_tx_start(struct work_struct *work)
 		p = list_entry(priv->cm.start_list.next, typeof(*p), list);
 		list_del_init(&p->list);
 		neigh = p->neigh;
+
 		qpn = IPOIB_QPN(neigh->daddr);
+		/*
+		 * As long as the search is with these 2 locks,
+		 * path existence indicates its validity.
+		 */
+		path = __path_find(dev, neigh->daddr + QPN_AND_OPTIONS_OFFSET);
+		if (!path) {
+			pr_info("%s ignore not valid path %pI6\n",
+				__func__,
+				neigh->daddr + QPN_AND_OPTIONS_OFFSET);
+			goto free_neigh;
+		}
 		memcpy(&pathrec, &p->path->pathrec, sizeof pathrec);
 
 		spin_unlock_irqrestore(&priv->lock, flags);
@@ -1331,6 +1345,7 @@ static void ipoib_cm_tx_start(struct work_struct *work)
 		spin_lock_irqsave(&priv->lock, flags);
 
 		if (ret) {
+free_neigh:
 			neigh = p->neigh;
 			if (neigh) {
 				neigh->cm = NULL;
diff --git a/kernel/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/kernel/drivers/infiniband/ulp/ipoib/ipoib_ib.c
index 5ea0c1407..8f8c3af9f 100644
--- a/kernel/drivers/infiniband/ulp/ipoib/ipoib_ib.c
+++ b/kernel/drivers/infiniband/ulp/ipoib/ipoib_ib.c
@@ -130,16 +130,15 @@ static struct sk_buff *ipoib_alloc_rx_skb(struct net_device *dev, int id)
 
 	buf_size = IPOIB_UD_BUF_SIZE(priv->max_ib_mtu);
 
-	skb = dev_alloc_skb(buf_size + IPOIB_ENCAP_LEN);
+	skb = dev_alloc_skb(buf_size + IPOIB_HARD_LEN);
 	if (unlikely(!skb))
 		return NULL;
 
 	/*
-	 * IB will leave a 40 byte gap for a GRH and IPoIB adds a 4 byte
-	 * header.  So we need 4 more bytes to get to 48 and align the
-	 * IP header to a multiple of 16.
+	 * the IP header will be at IPOIP_HARD_LEN + IB_GRH_BYTES, that is
+	 * 64 bytes aligned
 	 */
-	skb_reserve(skb, 4);
+	skb_reserve(skb, sizeof(struct ipoib_pseudo_header));
 
 	mapping = priv->rx_ring[id].mapping;
 	mapping[0] = ib_dma_map_single(priv->ca, skb->data, buf_size,
@@ -242,10 +241,7 @@ static void ipoib_ib_handle_rx_wc(struct net_device *dev, struct ib_wc *wc)
 	skb_pull(skb, IB_GRH_BYTES);
 
 	skb->protocol = ((struct ipoib_header *) skb->data)->proto;
-	skb_reset_mac_header(skb);
-	skb_pull(skb, IPOIB_ENCAP_LEN);
-
-	skb->truesize = SKB_TRUESIZE(skb->len);
+	skb_add_pseudo_hdr(skb);
 
 	++dev->stats.rx_packets;
 	dev->stats.rx_bytes += skb->len;
@@ -1030,8 +1026,17 @@ static void __ipoib_ib_dev_flush(struct ipoib_dev_priv *priv,
 	}
 
 	if (level == IPOIB_FLUSH_LIGHT) {
+		int oper_up;
 		ipoib_mark_paths_invalid(dev);
+		/* Set IPoIB operation as down to prevent races between:
+		 * the flush flow which leaves MCG and on the fly joins
+		 * which can happen during that time. mcast restart task
+		 * should deal with join requests we missed.
+		 */
+		oper_up = test_and_clear_bit(IPOIB_FLAG_OPER_UP, &priv->flags);
 		ipoib_mcast_dev_flush(dev);
+		if (oper_up)
+			set_bit(IPOIB_FLAG_OPER_UP, &priv->flags);
 		ipoib_flush_ah(dev);
 	}
 
diff --git a/kernel/drivers/infiniband/ulp/ipoib/ipoib_main.c b/kernel/drivers/infiniband/ulp/ipoib/ipoib_main.c
index 7d3281866..8a4d10452 100644
--- a/kernel/drivers/infiniband/ulp/ipoib/ipoib_main.c
+++ b/kernel/drivers/infiniband/ulp/ipoib/ipoib_main.c
@@ -481,7 +481,7 @@ int ipoib_set_mode(struct net_device *dev, const char *buf)
 	return -EINVAL;
 }
 
-static struct ipoib_path *__path_find(struct net_device *dev, void *gid)
+struct ipoib_path *__path_find(struct net_device *dev, void *gid)
 {
 	struct ipoib_dev_priv *priv = netdev_priv(dev);
 	struct rb_node *n = priv->path_tree.rb_node;
@@ -850,9 +850,12 @@ static void neigh_add_path(struct sk_buff *skb, u8 *daddr,
 				ipoib_neigh_free(neigh);
 				goto err_drop;
 			}
-			if (skb_queue_len(&neigh->queue) < IPOIB_MAX_PATH_REC_QUEUE)
+			if (skb_queue_len(&neigh->queue) <
+			    IPOIB_MAX_PATH_REC_QUEUE) {
+				/* put pseudoheader back on for next time */
+				skb_push(skb, IPOIB_PSEUDO_LEN);
 				__skb_queue_tail(&neigh->queue, skb);
-			else {
+			} else {
 				ipoib_warn(priv, "queue length limit %d. Packet drop.\n",
 					   skb_queue_len(&neigh->queue));
 				goto err_drop;
@@ -889,7 +892,7 @@ err_drop:
 }
 
 static void unicast_arp_send(struct sk_buff *skb, struct net_device *dev,
-			     struct ipoib_cb *cb)
+			     struct ipoib_pseudo_header *phdr)
 {
 	struct ipoib_dev_priv *priv = netdev_priv(dev);
 	struct ipoib_path *path;
@@ -897,16 +900,18 @@ static void unicast_arp_send(struct sk_buff *skb, struct net_device *dev,
 
 	spin_lock_irqsave(&priv->lock, flags);
 
-	path = __path_find(dev, cb->hwaddr + 4);
+	path = __path_find(dev, phdr->hwaddr + 4);
 	if (!path || !path->valid) {
 		int new_path = 0;
 
 		if (!path) {
-			path = path_rec_create(dev, cb->hwaddr + 4);
+			path = path_rec_create(dev, phdr->hwaddr + 4);
 			new_path = 1;
 		}
 		if (path) {
 			if (skb_queue_len(&path->queue) < IPOIB_MAX_PATH_REC_QUEUE) {
+				/* put pseudoheader back on for next time */
+				skb_push(skb, IPOIB_PSEUDO_LEN);
 				__skb_queue_tail(&path->queue, skb);
 			} else {
 				++dev->stats.tx_dropped;
@@ -934,10 +939,12 @@ static void unicast_arp_send(struct sk_buff *skb, struct net_device *dev,
 			  be16_to_cpu(path->pathrec.dlid));
 
 		spin_unlock_irqrestore(&priv->lock, flags);
-		ipoib_send(dev, skb, path->ah, IPOIB_QPN(cb->hwaddr));
+		ipoib_send(dev, skb, path->ah, IPOIB_QPN(phdr->hwaddr));
 		return;
 	} else if ((path->query || !path_rec_start(dev, path)) &&
 		   skb_queue_len(&path->queue) < IPOIB_MAX_PATH_REC_QUEUE) {
+		/* put pseudoheader back on for next time */
+		skb_push(skb, IPOIB_PSEUDO_LEN);
 		__skb_queue_tail(&path->queue, skb);
 	} else {
 		++dev->stats.tx_dropped;
@@ -951,13 +958,15 @@ static int ipoib_start_xmit(struct sk_buff *skb, struct net_device *dev)
 {
 	struct ipoib_dev_priv *priv = netdev_priv(dev);
 	struct ipoib_neigh *neigh;
-	struct ipoib_cb *cb = ipoib_skb_cb(skb);
+	struct ipoib_pseudo_header *phdr;
 	struct ipoib_header *header;
 	unsigned long flags;
 
+	phdr = (struct ipoib_pseudo_header *) skb->data;
+	skb_pull(skb, sizeof(*phdr));
 	header = (struct ipoib_header *) skb->data;
 
-	if (unlikely(cb->hwaddr[4] == 0xff)) {
+	if (unlikely(phdr->hwaddr[4] == 0xff)) {
 		/* multicast, arrange "if" according to probability */
 		if ((header->proto != htons(ETH_P_IP)) &&
 		    (header->proto != htons(ETH_P_IPV6)) &&
@@ -970,13 +979,13 @@ static int ipoib_start_xmit(struct sk_buff *skb, struct net_device *dev)
 			return NETDEV_TX_OK;
 		}
 		/* Add in the P_Key for multicast*/
-		cb->hwaddr[8] = (priv->pkey >> 8) & 0xff;
-		cb->hwaddr[9] = priv->pkey & 0xff;
+		phdr->hwaddr[8] = (priv->pkey >> 8) & 0xff;
+		phdr->hwaddr[9] = priv->pkey & 0xff;
 
-		neigh = ipoib_neigh_get(dev, cb->hwaddr);
+		neigh = ipoib_neigh_get(dev, phdr->hwaddr);
 		if (likely(neigh))
 			goto send_using_neigh;
-		ipoib_mcast_send(dev, cb->hwaddr, skb);
+		ipoib_mcast_send(dev, phdr->hwaddr, skb);
 		return NETDEV_TX_OK;
 	}
 
@@ -985,16 +994,16 @@ static int ipoib_start_xmit(struct sk_buff *skb, struct net_device *dev)
 	case htons(ETH_P_IP):
 	case htons(ETH_P_IPV6):
 	case htons(ETH_P_TIPC):
-		neigh = ipoib_neigh_get(dev, cb->hwaddr);
+		neigh = ipoib_neigh_get(dev, phdr->hwaddr);
 		if (unlikely(!neigh)) {
-			neigh_add_path(skb, cb->hwaddr, dev);
+			neigh_add_path(skb, phdr->hwaddr, dev);
 			return NETDEV_TX_OK;
 		}
 		break;
 	case htons(ETH_P_ARP):
 	case htons(ETH_P_RARP):
 		/* for unicast ARP and RARP should always perform path find */
-		unicast_arp_send(skb, dev, cb);
+		unicast_arp_send(skb, dev, phdr);
 		return NETDEV_TX_OK;
 	default:
 		/* ethertype not supported by IPoIB */
@@ -1011,11 +1020,13 @@ send_using_neigh:
 			goto unref;
 		}
 	} else if (neigh->ah) {
-		ipoib_send(dev, skb, neigh->ah, IPOIB_QPN(cb->hwaddr));
+		ipoib_send(dev, skb, neigh->ah, IPOIB_QPN(phdr->hwaddr));
 		goto unref;
 	}
 
 	if (skb_queue_len(&neigh->queue) < IPOIB_MAX_PATH_REC_QUEUE) {
+		/* put pseudoheader back on for next time */
+		skb_push(skb, sizeof(*phdr));
 		spin_lock_irqsave(&priv->lock, flags);
 		__skb_queue_tail(&neigh->queue, skb);
 		spin_unlock_irqrestore(&priv->lock, flags);
@@ -1047,8 +1058,8 @@ static int ipoib_hard_header(struct sk_buff *skb,
 			     unsigned short type,
 			     const void *daddr, const void *saddr, unsigned len)
 {
+	struct ipoib_pseudo_header *phdr;
 	struct ipoib_header *header;
-	struct ipoib_cb *cb = ipoib_skb_cb(skb);
 
 	header = (struct ipoib_header *) skb_push(skb, sizeof *header);
 
@@ -1057,12 +1068,13 @@ static int ipoib_hard_header(struct sk_buff *skb,
 
 	/*
 	 * we don't rely on dst_entry structure,  always stuff the
-	 * destination address into skb->cb so we can figure out where
+	 * destination address into skb hard header so we can figure out where
 	 * to send the packet later.
 	 */
-	memcpy(cb->hwaddr, daddr, INFINIBAND_ALEN);
+	phdr = (struct ipoib_pseudo_header *) skb_push(skb, sizeof(*phdr));
+	memcpy(phdr->hwaddr, daddr, INFINIBAND_ALEN);
 
-	return sizeof *header;
+	return IPOIB_HARD_LEN;
 }
 
 static void ipoib_set_mcast_list(struct net_device *dev)
@@ -1131,7 +1143,9 @@ struct ipoib_neigh *ipoib_neigh_get(struct net_device *dev, u8 *daddr)
 				neigh = NULL;
 				goto out_unlock;
 			}
-			neigh->alive = jiffies;
+
+			if (likely(skb_queue_len(&neigh->queue) < IPOIB_MAX_PATH_REC_QUEUE))
+				neigh->alive = jiffies;
 			goto out_unlock;
 		}
 	}
@@ -1636,7 +1650,7 @@ void ipoib_setup(struct net_device *dev)
 
 	dev->flags		|= IFF_BROADCAST | IFF_MULTICAST;
 
-	dev->hard_header_len	 = IPOIB_ENCAP_LEN;
+	dev->hard_header_len	 = IPOIB_HARD_LEN;
 	dev->addr_len		 = INFINIBAND_ALEN;
 	dev->type		 = ARPHRD_INFINIBAND;
 	dev->tx_queue_len	 = ipoib_sendq_size * 2;
diff --git a/kernel/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/kernel/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
index 6dcef673d..a123d0439 100644
--- a/kernel/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
+++ b/kernel/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
@@ -456,7 +456,10 @@ out_locked:
 	return status;
 }
 
-static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast)
+/*
+ * Caller must hold 'priv->lock'
+ */
+static int ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast)
 {
 	struct ipoib_dev_priv *priv = netdev_priv(dev);
 	struct ib_sa_multicast *multicast;
@@ -466,6 +469,10 @@ static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast)
 	ib_sa_comp_mask comp_mask;
 	int ret = 0;
 
+	if (!priv->broadcast ||
+	    !test_bit(IPOIB_FLAG_OPER_UP, &priv->flags))
+		return -EINVAL;
+
 	ipoib_dbg_mcast(priv, "joining MGID %pI6\n", mcast->mcmember.mgid.raw);
 
 	rec.mgid     = mcast->mcmember.mgid;
@@ -525,20 +532,23 @@ static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast)
 			rec.join_state = 4;
 #endif
 	}
+	spin_unlock_irq(&priv->lock);
 
 	multicast = ib_sa_join_multicast(&ipoib_sa_client, priv->ca, priv->port,
 					 &rec, comp_mask, GFP_KERNEL,
 					 ipoib_mcast_join_complete, mcast);
+	spin_lock_irq(&priv->lock);
 	if (IS_ERR(multicast)) {
 		ret = PTR_ERR(multicast);
 		ipoib_warn(priv, "ib_sa_join_multicast failed, status %d\n", ret);
-		spin_lock_irq(&priv->lock);
 		/* Requeue this join task with a backoff delay */
 		__ipoib_mcast_schedule_join_thread(priv, mcast, 1);
 		clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
 		spin_unlock_irq(&priv->lock);
 		complete(&mcast->done);
+		spin_lock_irq(&priv->lock);
 	}
+	return 0;
 }
 
 void ipoib_mcast_join_task(struct work_struct *work)
@@ -553,8 +563,11 @@ void ipoib_mcast_join_task(struct work_struct *work)
 	if (!test_bit(IPOIB_FLAG_OPER_UP, &priv->flags))
 		return;
 
-	if (ib_query_port(priv->ca, priv->port, &port_attr) ||
-	    port_attr.state != IB_PORT_ACTIVE) {
+	if (ib_query_port(priv->ca, priv->port, &port_attr)) {
+		ipoib_dbg(priv, "ib_query_port() failed\n");
+		return;
+	}
+	if (port_attr.state != IB_PORT_ACTIVE) {
 		ipoib_dbg(priv, "port state is not ACTIVE (state = %d) suspending join task\n",
 			  port_attr.state);
 		return;
@@ -620,9 +633,10 @@ void ipoib_mcast_join_task(struct work_struct *work)
 				/* Found the next unjoined group */
 				init_completion(&mcast->done);
 				set_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
-				spin_unlock_irq(&priv->lock);
-				ipoib_mcast_join(dev, mcast);
-				spin_lock_irq(&priv->lock);
+				if (ipoib_mcast_join(dev, mcast)) {
+					spin_unlock_irq(&priv->lock);
+					return;
+				}
 			} else if (!delay_until ||
 				 time_before(mcast->delay_until, delay_until))
 				delay_until = mcast->delay_until;
@@ -641,10 +655,9 @@ out:
 	if (mcast) {
 		init_completion(&mcast->done);
 		set_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
+		ipoib_mcast_join(dev, mcast);
 	}
 	spin_unlock_irq(&priv->lock);
-	if (mcast)
-		ipoib_mcast_join(dev, mcast);
 }
 
 int ipoib_mcast_start_thread(struct net_device *dev)
@@ -743,9 +756,11 @@ void ipoib_mcast_send(struct net_device *dev, u8 *daddr, struct sk_buff *skb)
 			__ipoib_mcast_add(dev, mcast);
 			list_add_tail(&mcast->list, &priv->multicast_list);
 		}
-		if (skb_queue_len(&mcast->pkt_queue) < IPOIB_MAX_MCAST_QUEUE)
+		if (skb_queue_len(&mcast->pkt_queue) < IPOIB_MAX_MCAST_QUEUE) {
+			/* put pseudoheader back on for next time */
+			skb_push(skb, sizeof(struct ipoib_pseudo_header));
 			skb_queue_tail(&mcast->pkt_queue, skb);
-		else {
+		} else {
 			++dev->stats.tx_dropped;
 			dev_kfree_skb_any(skb);
 		}
diff --git a/kernel/drivers/infiniband/ulp/isert/ib_isert.c b/kernel/drivers/infiniband/ulp/isert/ib_isert.c
index 8a51c3b5d..b0edb66a2 100644
--- a/kernel/drivers/infiniband/ulp/isert/ib_isert.c
+++ b/kernel/drivers/infiniband/ulp/isert/ib_isert.c
@@ -66,6 +66,7 @@ isert_rdma_accept(struct isert_conn *isert_conn);
 struct rdma_cm_id *isert_setup_id(struct isert_np *isert_np);
 
 static void isert_release_work(struct work_struct *work);
+static void isert_wait4flush(struct isert_conn *isert_conn);
 
 static inline bool
 isert_prot_cmd(struct isert_conn *conn, struct se_cmd *cmd)
@@ -815,12 +816,31 @@ isert_put_conn(struct isert_conn *isert_conn)
 	kref_put(&isert_conn->kref, isert_release_kref);
 }
 
+static void
+isert_handle_unbound_conn(struct isert_conn *isert_conn)
+{
+	struct isert_np *isert_np = isert_conn->cm_id->context;
+
+	mutex_lock(&isert_np->mutex);
+	if (!list_empty(&isert_conn->node)) {
+		/*
+		 * This means iscsi doesn't know this connection
+		 * so schedule a cleanup ourselves
+		 */
+		list_del_init(&isert_conn->node);
+		isert_put_conn(isert_conn);
+		complete(&isert_conn->wait);
+		queue_work(isert_release_wq, &isert_conn->release_work);
+	}
+	mutex_unlock(&isert_np->mutex);
+}
+
 /**
  * isert_conn_terminate() - Initiate connection termination
  * @isert_conn: isert connection struct
  *
  * Notes:
- * In case the connection state is FULL_FEATURE, move state
+ * In case the connection state is BOUND, move state
  * to TEMINATING and start teardown sequence (rdma_disconnect).
  * In case the connection state is UP, complete flush as well.
  *
@@ -832,23 +852,19 @@ isert_conn_terminate(struct isert_conn *isert_conn)
 {
 	int err;
 
-	switch (isert_conn->state) {
-	case ISER_CONN_TERMINATING:
-		break;
-	case ISER_CONN_UP:
-	case ISER_CONN_FULL_FEATURE: /* FALLTHRU */
-		isert_info("Terminating conn %p state %d\n",
-			   isert_conn, isert_conn->state);
-		isert_conn->state = ISER_CONN_TERMINATING;
-		err = rdma_disconnect(isert_conn->cm_id);
-		if (err)
-			isert_warn("Failed rdma_disconnect isert_conn %p\n",
-				   isert_conn);
-		break;
-	default:
-		isert_warn("conn %p teminating in state %d\n",
-			   isert_conn, isert_conn->state);
-	}
+	if (isert_conn->state >= ISER_CONN_TERMINATING)
+		return;
+
+	isert_info("Terminating conn %p state %d\n",
+		   isert_conn, isert_conn->state);
+	isert_conn->state = ISER_CONN_TERMINATING;
+	err = rdma_disconnect(isert_conn->cm_id);
+	if (err)
+		isert_warn("Failed rdma_disconnect isert_conn %p\n",
+			   isert_conn);
+
+	isert_info("conn %p completing wait\n", isert_conn);
+	complete(&isert_conn->wait);
 }
 
 static int
@@ -882,35 +898,27 @@ static int
 isert_disconnected_handler(struct rdma_cm_id *cma_id,
 			   enum rdma_cm_event_type event)
 {
-	struct isert_np *isert_np = cma_id->context;
-	struct isert_conn *isert_conn;
-	bool terminating = false;
-
-	if (isert_np->cm_id == cma_id)
-		return isert_np_cma_handler(cma_id->context, event);
-
-	isert_conn = cma_id->qp->qp_context;
+	struct isert_conn *isert_conn = cma_id->qp->qp_context;
 
 	mutex_lock(&isert_conn->mutex);
-	terminating = (isert_conn->state == ISER_CONN_TERMINATING);
-	isert_conn_terminate(isert_conn);
-	mutex_unlock(&isert_conn->mutex);
-
-	isert_info("conn %p completing wait\n", isert_conn);
-	complete(&isert_conn->wait);
-
-	if (terminating)
-		goto out;
-
-	mutex_lock(&isert_np->mutex);
-	if (!list_empty(&isert_conn->node)) {
-		list_del_init(&isert_conn->node);
-		isert_put_conn(isert_conn);
-		queue_work(isert_release_wq, &isert_conn->release_work);
+	switch (isert_conn->state) {
+	case ISER_CONN_TERMINATING:
+		break;
+	case ISER_CONN_UP:
+		isert_conn_terminate(isert_conn);
+		isert_wait4flush(isert_conn);
+		isert_handle_unbound_conn(isert_conn);
+		break;
+	case ISER_CONN_BOUND:
+	case ISER_CONN_FULL_FEATURE: /* FALLTHRU */
+		iscsit_cause_connection_reinstatement(isert_conn->conn, 0);
+		break;
+	default:
+		isert_warn("conn %p teminating in state %d\n",
+			   isert_conn, isert_conn->state);
 	}
-	mutex_unlock(&isert_np->mutex);
+	mutex_unlock(&isert_conn->mutex);
 
-out:
 	return 0;
 }
 
@@ -929,12 +937,16 @@ isert_connect_error(struct rdma_cm_id *cma_id)
 static int
 isert_cma_handler(struct rdma_cm_id *cma_id, struct rdma_cm_event *event)
 {
+	struct isert_np *isert_np = cma_id->context;
 	int ret = 0;
 
 	isert_info("%s (%d): status %d id %p np %p\n",
 		   rdma_event_msg(event->event), event->event,
 		   event->status, cma_id, cma_id->context);
 
+	if (isert_np->cm_id == cma_id)
+		return isert_np_cma_handler(cma_id->context, event->event);
+
 	switch (event->event) {
 	case RDMA_CM_EVENT_CONNECT_REQUEST:
 		ret = isert_connect_request(cma_id, event);
@@ -980,13 +992,10 @@ isert_post_recvm(struct isert_conn *isert_conn, u32 count)
 	rx_wr--;
 	rx_wr->next = NULL; /* mark end of work requests list */
 
-	isert_conn->post_recv_buf_count += count;
 	ret = ib_post_recv(isert_conn->qp, isert_conn->rx_wr,
 			   &rx_wr_failed);
-	if (ret) {
+	if (ret)
 		isert_err("ib_post_recv() failed with ret: %d\n", ret);
-		isert_conn->post_recv_buf_count -= count;
-	}
 
 	return ret;
 }
@@ -1002,12 +1011,9 @@ isert_post_recv(struct isert_conn *isert_conn, struct iser_rx_desc *rx_desc)
 	rx_wr.num_sge = 1;
 	rx_wr.next = NULL;
 
-	isert_conn->post_recv_buf_count++;
 	ret = ib_post_recv(isert_conn->qp, &rx_wr, &rx_wr_failed);
-	if (ret) {
+	if (ret)
 		isert_err("ib_post_recv() failed with ret: %d\n", ret);
-		isert_conn->post_recv_buf_count--;
-	}
 
 	return ret;
 }
@@ -1120,12 +1126,9 @@ isert_rdma_post_recvl(struct isert_conn *isert_conn)
 	rx_wr.sg_list = &sge;
 	rx_wr.num_sge = 1;
 
-	isert_conn->post_recv_buf_count++;
 	ret = ib_post_recv(isert_conn->qp, &rx_wr, &rx_wr_fail);
-	if (ret) {
+	if (ret)
 		isert_err("ib_post_recv() failed: %d\n", ret);
-		isert_conn->post_recv_buf_count--;
-	}
 
 	return ret;
 }
@@ -1620,7 +1623,6 @@ isert_rcv_completion(struct iser_rx_desc *desc,
 	ib_dma_sync_single_for_device(ib_dev, rx_dma, rx_buflen,
 				      DMA_FROM_DEVICE);
 
-	isert_conn->post_recv_buf_count--;
 }
 
 static int
@@ -2035,7 +2037,8 @@ is_isert_tx_desc(struct isert_conn *isert_conn, void *wr_id)
 	void *start = isert_conn->rx_descs;
 	int len = ISERT_QP_MAX_RECV_DTOS * sizeof(*isert_conn->rx_descs);
 
-	if (wr_id >= start && wr_id < start + len)
+	if ((wr_id >= start && wr_id < start + len) ||
+	    (wr_id == isert_conn->login_req_buf))
 		return false;
 
 	return true;
@@ -2059,10 +2062,6 @@ isert_cq_comp_err(struct isert_conn *isert_conn, struct ib_wc *wc)
 			isert_unmap_tx_desc(desc, ib_dev);
 		else
 			isert_completion_put(desc, isert_cmd, ib_dev, true);
-	} else {
-		isert_conn->post_recv_buf_count--;
-		if (!isert_conn->post_recv_buf_count)
-			iscsit_cause_connection_reinstatement(isert_conn->conn, 0);
 	}
 }
 
@@ -3193,6 +3192,7 @@ accept_wait:
 
 	conn->context = isert_conn;
 	isert_conn->conn = conn;
+	isert_conn->state = ISER_CONN_BOUND;
 
 	isert_set_conn_info(np, conn, isert_conn);
 
diff --git a/kernel/drivers/infiniband/ulp/isert/ib_isert.h b/kernel/drivers/infiniband/ulp/isert/ib_isert.h
index 3d7fbc47c..1874d21da 100644
--- a/kernel/drivers/infiniband/ulp/isert/ib_isert.h
+++ b/kernel/drivers/infiniband/ulp/isert/ib_isert.h
@@ -50,6 +50,7 @@ enum iser_ib_op_code {
 enum iser_conn_state {
 	ISER_CONN_INIT,
 	ISER_CONN_UP,
+	ISER_CONN_BOUND,
 	ISER_CONN_FULL_FEATURE,
 	ISER_CONN_TERMINATING,
 	ISER_CONN_DOWN,
@@ -144,7 +145,6 @@ struct isert_device;
 
 struct isert_conn {
 	enum iser_conn_state	state;
-	int			post_recv_buf_count;
 	u32			responder_resources;
 	u32			initiator_depth;
 	bool			pi_support;
diff --git a/kernel/drivers/infiniband/ulp/srp/ib_srp.c b/kernel/drivers/infiniband/ulp/srp/ib_srp.c
index 3db9a6597..5f0f4fc58 100644
--- a/kernel/drivers/infiniband/ulp/srp/ib_srp.c
+++ b/kernel/drivers/infiniband/ulp/srp/ib_srp.c
@@ -1519,7 +1519,7 @@ static int srp_map_idb(struct srp_rdma_ch *ch, struct srp_request *req,
 
 	if (dev->use_fast_reg) {
 		state.sg = idb_sg;
-		sg_set_buf(idb_sg, req->indirect_desc, idb_len);
+		sg_init_one(idb_sg, req->indirect_desc, idb_len);
 		idb_sg->dma_address = req->indirect_dma_addr; /* hack! */
 #ifdef CONFIG_NEED_SG_DMA_LENGTH
 		idb_sg->dma_length = idb_sg->length;	      /* hack^2 */
diff --git a/kernel/drivers/infiniband/ulp/srpt/ib_srpt.c b/kernel/drivers/infiniband/ulp/srpt/ib_srpt.c
index 2e2fe818c..eaabf3125 100644
--- a/kernel/drivers/infiniband/ulp/srpt/ib_srpt.c
+++ b/kernel/drivers/infiniband/ulp/srpt/ib_srpt.c
@@ -1737,47 +1737,6 @@ send_sense:
 	return -1;
 }
 
-/**
- * srpt_rx_mgmt_fn_tag() - Process a task management function by tag.
- * @ch: RDMA channel of the task management request.
- * @fn: Task management function to perform.
- * @req_tag: Tag of the SRP task management request.
- * @mgmt_ioctx: I/O context of the task management request.
- *
- * Returns zero if the target core will process the task management
- * request asynchronously.
- *
- * Note: It is assumed that the initiator serializes tag-based task management
- * requests.
- */
-static int srpt_rx_mgmt_fn_tag(struct srpt_send_ioctx *ioctx, u64 tag)
-{
-	struct srpt_device *sdev;
-	struct srpt_rdma_ch *ch;
-	struct srpt_send_ioctx *target;
-	int ret, i;
-
-	ret = -EINVAL;
-	ch = ioctx->ch;
-	BUG_ON(!ch);
-	BUG_ON(!ch->sport);
-	sdev = ch->sport->sdev;
-	BUG_ON(!sdev);
-	spin_lock_irq(&sdev->spinlock);
-	for (i = 0; i < ch->rq_size; ++i) {
-		target = ch->ioctx_ring[i];
-		if (target->cmd.se_lun == ioctx->cmd.se_lun &&
-		    target->cmd.tag == tag &&
-		    srpt_get_cmd_state(target) != SRPT_STATE_DONE) {
-			ret = 0;
-			/* now let the target core abort &target->cmd; */
-			break;
-		}
-	}
-	spin_unlock_irq(&sdev->spinlock);
-	return ret;
-}
-
 static int srp_tmr_to_tcm(int fn)
 {
 	switch (fn) {
@@ -1812,7 +1771,6 @@ static void srpt_handle_tsk_mgmt(struct srpt_rdma_ch *ch,
 	struct se_cmd *cmd;
 	struct se_session *sess = ch->sess;
 	uint64_t unpacked_lun;
-	uint32_t tag = 0;
 	int tcm_tmr;
 	int rc;
 
@@ -1828,25 +1786,10 @@ static void srpt_handle_tsk_mgmt(struct srpt_rdma_ch *ch,
 	srpt_set_cmd_state(send_ioctx, SRPT_STATE_MGMT);
 	send_ioctx->cmd.tag = srp_tsk->tag;
 	tcm_tmr = srp_tmr_to_tcm(srp_tsk->tsk_mgmt_func);
-	if (tcm_tmr < 0) {
-		send_ioctx->cmd.se_tmr_req->response =
-			TMR_TASK_MGMT_FUNCTION_NOT_SUPPORTED;
-		goto fail;
-	}
 	unpacked_lun = srpt_unpack_lun((uint8_t *)&srp_tsk->lun,
 				       sizeof(srp_tsk->lun));
-
-	if (srp_tsk->tsk_mgmt_func == SRP_TSK_ABORT_TASK) {
-		rc = srpt_rx_mgmt_fn_tag(send_ioctx, srp_tsk->task_tag);
-		if (rc < 0) {
-			send_ioctx->cmd.se_tmr_req->response =
-					TMR_TASK_DOES_NOT_EXIST;
-			goto fail;
-		}
-		tag = srp_tsk->task_tag;
-	}
 	rc = target_submit_tmr(&send_ioctx->cmd, sess, NULL, unpacked_lun,
-				srp_tsk, tcm_tmr, GFP_KERNEL, tag,
+				srp_tsk, tcm_tmr, GFP_KERNEL, srp_tsk->task_tag,
 				TARGET_SCF_ACK_KREF);
 	if (rc != 0) {
 		send_ioctx->cmd.se_tmr_req->response = TMR_FUNCTION_REJECTED;
-- 
cgit 1.2.3-korg