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

Pull infiniband/rdma fixes from Roland Dreier:
 "Last late set of InfiniBand/RDMA fixes for 3.17:

   - fixes for the new memory region re-registration support
   - iSER initiator error path fixes
   - grab bag of small fixes for the qib and ocrdma hardware drivers
   - larger set of fixes for mlx4, especially in RoCE mode"

* tag 'rdma-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (26 commits)
  IB/mlx4: Fix VF mac handling in RoCE
  IB/mlx4: Do not allow APM under RoCE
  IB/mlx4: Don't update QP1 in native mode
  IB/mlx4: Avoid accessing netdevice when building RoCE qp1 header
  mlx4: Fix mlx4 reg/unreg mac to work properly with 0-mac addresses
  IB/core: When marshaling uverbs path, clear unused fields
  IB/mlx4: Avoid executing gid task when device is being removed
  IB/mlx4: Fix lockdep splat for the iboe lock
  IB/mlx4: Get upper dev addresses as RoCE GIDs when port comes up
  IB/mlx4: Reorder steps in RoCE GID table initialization
  IB/mlx4: Don't duplicate the default RoCE GID
  IB/mlx4: Avoid null pointer dereference in mlx4_ib_scan_netdevs()
  IB/iser: Bump version to 1.4.1
  IB/iser: Allow bind only when connection state is UP
  IB/iser: Fix RX/TX CQ resource leak on error flow
  RDMA/ocrdma: Use right macro in query AH
  RDMA/ocrdma: Resolve L2 address when creating user AH
  mlx4: Correct error flows in rereg_mr
  IB/qib: Correct reference counting in debugfs qp_stats
  IPoIB: Remove unnecessary port query
  ...

drivers/infiniband/core/umem.c

@@ -105,6 +105,7 @@ struct ib_umem *ib_umem_get(struct ib_ucontext *context, unsigned long addr,
 	umem->length    = size;
 	umem->offset    = addr & ~PAGE_MASK;
 	umem->page_size = PAGE_SIZE;
+	umem->pid       = get_task_pid(current, PIDTYPE_PID);
 	/*
 	 * We ask for writable memory if any access flags other than
 	 * "remote read" are set.  "Local write" and "remote write"
@@ -198,6 +199,7 @@ struct ib_umem *ib_umem_get(struct ib_ucontext *context, unsigned long addr,
 	if (ret < 0) {
 		if (need_release)
 			__ib_umem_release(context->device, umem, 0);
+		put_pid(umem->pid);
 		kfree(umem);
 	} else
 		current->mm->pinned_vm = locked;
@@ -230,15 +232,19 @@ void ib_umem_release(struct ib_umem *umem)
 {
 	struct ib_ucontext *context = umem->context;
 	struct mm_struct *mm;
+	struct task_struct *task;
 	unsigned long diff;
 
 	__ib_umem_release(umem->context->device, umem, 1);
 
-	mm = get_task_mm(current);
-	if (!mm) {
-		kfree(umem);
-		return;
-	}
+	task = get_pid_task(umem->pid, PIDTYPE_PID);
+	put_pid(umem->pid);
+	if (!task)
+		goto out;
+	mm = get_task_mm(task);
+	put_task_struct(task);
+	if (!mm)
+		goto out;
 
 	diff = PAGE_ALIGN(umem->length + umem->offset) >> PAGE_SHIFT;
 
@@ -262,9 +268,10 @@ void ib_umem_release(struct ib_umem *umem)
 	} else
 		down_write(&mm->mmap_sem);
 
-	current->mm->pinned_vm -= diff;
+	mm->pinned_vm -= diff;
 	up_write(&mm->mmap_sem);
 	mmput(mm);
+out:
 	kfree(umem);
 }
 EXPORT_SYMBOL(ib_umem_release);

drivers/infiniband/core/uverbs_marshall.c

@@ -140,5 +140,9 @@ void ib_copy_path_rec_from_user(struct ib_sa_path_rec *dst,
 	dst->packet_life_time	= src->packet_life_time;
 	dst->preference		= src->preference;
 	dst->packet_life_time_selector = src->packet_life_time_selector;
+
+	memset(dst->smac, 0, sizeof(dst->smac));
+	memset(dst->dmac, 0, sizeof(dst->dmac));
+	dst->vlan_id = 0xffff;
 }
 EXPORT_SYMBOL(ib_copy_path_rec_from_user);

drivers/infiniband/hw/ipath/ipath_user_pages.c

@@ -54,7 +54,7 @@ static void __ipath_release_user_pages(struct page **p, size_t num_pages,
 
 /* call with current->mm->mmap_sem held */
 static int __ipath_get_user_pages(unsigned long start_page, size_t num_pages,
-				  struct page **p, struct vm_area_struct **vma)
+				  struct page **p)
 {
 	unsigned long lock_limit;
 	size_t got;
@@ -74,7 +74,7 @@ static int __ipath_get_user_pages(unsigned long start_page, size_t num_pages,
 		ret = get_user_pages(current, current->mm,
 				     start_page + got * PAGE_SIZE,
 				     num_pages - got, 1, 1,
-				     p + got, vma);
+				     p + got, NULL);
 		if (ret < 0)
 			goto bail_release;
 	}
@@ -165,7 +165,7 @@ int ipath_get_user_pages(unsigned long start_page, size_t num_pages,
 
 	down_write(&current->mm->mmap_sem);
 
-	ret = __ipath_get_user_pages(start_page, num_pages, p, NULL);
+	ret = __ipath_get_user_pages(start_page, num_pages, p);
 
 	up_write(&current->mm->mmap_sem);
 

drivers/infiniband/hw/mlx4/main.c

@@ -59,6 +59,7 @@
 
 #define MLX4_IB_FLOW_MAX_PRIO 0xFFF
 #define MLX4_IB_FLOW_QPN_MASK 0xFFFFFF
+#define MLX4_IB_CARD_REV_A0   0xA0
 
 MODULE_AUTHOR("Roland Dreier");
 MODULE_DESCRIPTION("Mellanox ConnectX HCA InfiniBand driver");
@@ -119,6 +120,17 @@ static int check_flow_steering_support(struct mlx4_dev *dev)
 	return dmfs;
 }
 
+static int num_ib_ports(struct mlx4_dev *dev)
+{
+	int ib_ports = 0;
+	int i;
+
+	mlx4_foreach_port(i, dev, MLX4_PORT_TYPE_IB)
+		ib_ports++;
+
+	return ib_ports;
+}
+
 static int mlx4_ib_query_device(struct ib_device *ibdev,
 				struct ib_device_attr *props)
 {
@@ -126,6 +138,7 @@ static int mlx4_ib_query_device(struct ib_device *ibdev,
 	struct ib_smp *in_mad  = NULL;
 	struct ib_smp *out_mad = NULL;
 	int err = -ENOMEM;
+	int have_ib_ports;
 
 	in_mad  = kzalloc(sizeof *in_mad, GFP_KERNEL);
 	out_mad = kmalloc(sizeof *out_mad, GFP_KERNEL);
@@ -142,6 +155,8 @@ static int mlx4_ib_query_device(struct ib_device *ibdev,
 
 	memset(props, 0, sizeof *props);
 
+	have_ib_ports = num_ib_ports(dev->dev);
+
 	props->fw_ver = dev->dev->caps.fw_ver;
 	props->device_cap_flags    = IB_DEVICE_CHANGE_PHY_PORT |
 		IB_DEVICE_PORT_ACTIVE_EVENT		|
@@ -152,13 +167,15 @@ static int mlx4_ib_query_device(struct ib_device *ibdev,
 		props->device_cap_flags |= IB_DEVICE_BAD_PKEY_CNTR;
 	if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_BAD_QKEY_CNTR)
 		props->device_cap_flags |= IB_DEVICE_BAD_QKEY_CNTR;
-	if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_APM)
+	if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_APM && have_ib_ports)
 		props->device_cap_flags |= IB_DEVICE_AUTO_PATH_MIG;
 	if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_UD_AV_PORT)
 		props->device_cap_flags |= IB_DEVICE_UD_AV_PORT_ENFORCE;
 	if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_IPOIB_CSUM)
 		props->device_cap_flags |= IB_DEVICE_UD_IP_CSUM;
-	if (dev->dev->caps.max_gso_sz && dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_BLH)
+	if (dev->dev->caps.max_gso_sz &&
+	    (dev->dev->rev_id != MLX4_IB_CARD_REV_A0) &&
+	    (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_BLH))
 		props->device_cap_flags |= IB_DEVICE_UD_TSO;
 	if (dev->dev->caps.bmme_flags & MLX4_BMME_FLAG_RESERVED_LKEY)
 		props->device_cap_flags |= IB_DEVICE_LOCAL_DMA_LKEY;
@@ -357,7 +374,7 @@ static int eth_link_query_port(struct ib_device *ibdev, u8 port,
 	props->state		= IB_PORT_DOWN;
 	props->phys_state	= state_to_phys_state(props->state);
 	props->active_mtu	= IB_MTU_256;
-	spin_lock(&iboe->lock);
+	spin_lock_bh(&iboe->lock);
 	ndev = iboe->netdevs[port - 1];
 	if (!ndev)
 		goto out_unlock;
@@ -369,7 +386,7 @@ static int eth_link_query_port(struct ib_device *ibdev, u8 port,
 					IB_PORT_ACTIVE : IB_PORT_DOWN;
 	props->phys_state	= state_to_phys_state(props->state);
 out_unlock:
-	spin_unlock(&iboe->lock);
+	spin_unlock_bh(&iboe->lock);
 out:
 	mlx4_free_cmd_mailbox(mdev->dev, mailbox);
 	return err;
@@ -811,11 +828,11 @@ int mlx4_ib_add_mc(struct mlx4_ib_dev *mdev, struct mlx4_ib_qp *mqp,
 	if (!mqp->port)
 		return 0;
 
-	spin_lock(&mdev->iboe.lock);
+	spin_lock_bh(&mdev->iboe.lock);
 	ndev = mdev->iboe.netdevs[mqp->port - 1];
 	if (ndev)
 		dev_hold(ndev);
-	spin_unlock(&mdev->iboe.lock);
+	spin_unlock_bh(&mdev->iboe.lock);
 
 	if (ndev) {
 		ret = 1;
@@ -1292,11 +1309,11 @@ static int mlx4_ib_mcg_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
 	mutex_lock(&mqp->mutex);
 	ge = find_gid_entry(mqp, gid->raw);
 	if (ge) {
-		spin_lock(&mdev->iboe.lock);
+		spin_lock_bh(&mdev->iboe.lock);
 		ndev = ge->added ? mdev->iboe.netdevs[ge->port - 1] : NULL;
 		if (ndev)
 			dev_hold(ndev);
-		spin_unlock(&mdev->iboe.lock);
+		spin_unlock_bh(&mdev->iboe.lock);
 		if (ndev)
 			dev_put(ndev);
 		list_del(&ge->list);
@@ -1417,6 +1434,9 @@ static void update_gids_task(struct work_struct *work)
 	int err;
 	struct mlx4_dev	*dev = gw->dev->dev;
 
+	if (!gw->dev->ib_active)
+		return;
+
 	mailbox = mlx4_alloc_cmd_mailbox(dev);
 	if (IS_ERR(mailbox)) {
 		pr_warn("update gid table failed %ld\n", PTR_ERR(mailbox));
@@ -1447,6 +1467,9 @@ static void reset_gids_task(struct work_struct *work)
 	int err;
 	struct mlx4_dev	*dev = gw->dev->dev;
 
+	if (!gw->dev->ib_active)
+		return;
+
 	mailbox = mlx4_alloc_cmd_mailbox(dev);
 	if (IS_ERR(mailbox)) {
 		pr_warn("reset gid table failed\n");
@@ -1581,7 +1604,7 @@ static int mlx4_ib_addr_event(int event, struct net_device *event_netdev,
 		return 0;
 
 	iboe = &ibdev->iboe;
-	spin_lock(&iboe->lock);
+	spin_lock_bh(&iboe->lock);
 
 	for (port = 1; port <= ibdev->dev->caps.num_ports; ++port)
 		if ((netif_is_bond_master(real_dev) &&
@@ -1591,7 +1614,7 @@ static int mlx4_ib_addr_event(int event, struct net_device *event_netdev,
 			update_gid_table(ibdev, port, gid,
 					 event == NETDEV_DOWN, 0);
 
-	spin_unlock(&iboe->lock);
+	spin_unlock_bh(&iboe->lock);
 	return 0;
 
 }
@@ -1664,13 +1687,21 @@ static void mlx4_ib_update_qps(struct mlx4_ib_dev *ibdev,
 	new_smac = mlx4_mac_to_u64(dev->dev_addr);
 	read_unlock(&dev_base_lock);
 
+	atomic64_set(&ibdev->iboe.mac[port - 1], new_smac);
+
+	/* no need for update QP1 and mac registration in non-SRIOV */
+	if (!mlx4_is_mfunc(ibdev->dev))
+		return;
+
 	mutex_lock(&ibdev->qp1_proxy_lock[port - 1]);
 	qp = ibdev->qp1_proxy[port - 1];
 	if (qp) {
 		int new_smac_index;
-		u64 old_smac = qp->pri.smac;
+		u64 old_smac;
 		struct mlx4_update_qp_params update_params;
 
+		mutex_lock(&qp->mutex);
+		old_smac = qp->pri.smac;
 		if (new_smac == old_smac)
 			goto unlock;
 
@@ -1685,17 +1716,20 @@ static void mlx4_ib_update_qps(struct mlx4_ib_dev *ibdev,
 			release_mac = new_smac;
 			goto unlock;
 		}
-
+		/* if old port was zero, no mac was yet registered for this QP */
+		if (qp->pri.smac_port)
+			release_mac = old_smac;
 		qp->pri.smac = new_smac;
+		qp->pri.smac_port = port;
 		qp->pri.smac_index = new_smac_index;
-
-		release_mac = old_smac;
 	}
 
 unlock:
-	mutex_unlock(&ibdev->qp1_proxy_lock[port - 1]);
 	if (release_mac != MLX4_IB_INVALID_MAC)
 		mlx4_unregister_mac(ibdev->dev, port, release_mac);
+	if (qp)
+		mutex_unlock(&qp->mutex);
+	mutex_unlock(&ibdev->qp1_proxy_lock[port - 1]);
 }
 
 static void mlx4_ib_get_dev_addr(struct net_device *dev,
@@ -1706,6 +1740,7 @@ static void mlx4_ib_get_dev_addr(struct net_device *dev,
 	struct inet6_dev *in6_dev;
 	union ib_gid  *pgid;
 	struct inet6_ifaddr *ifp;
+	union ib_gid default_gid;
 #endif
 	union ib_gid gid;
 
@@ -1726,12 +1761,15 @@ static void mlx4_ib_get_dev_addr(struct net_device *dev,
 		in_dev_put(in_dev);
 	}
 #if IS_ENABLED(CONFIG_IPV6)
+	mlx4_make_default_gid(dev, &default_gid);
 	/* IPv6 gids */
 	in6_dev = in6_dev_get(dev);
 	if (in6_dev) {
 		read_lock_bh(&in6_dev->lock);
 		list_for_each_entry(ifp, &in6_dev->addr_list, if_list) {
 			pgid = (union ib_gid *)&ifp->addr;
+			if (!memcmp(pgid, &default_gid, sizeof(*pgid)))
+				continue;
 			update_gid_table(ibdev, port, pgid, 0, 0);
 		}
 		read_unlock_bh(&in6_dev->lock);
@@ -1753,24 +1791,33 @@ static int mlx4_ib_init_gid_table(struct mlx4_ib_dev *ibdev)
 	struct	net_device *dev;
 	struct mlx4_ib_iboe *iboe = &ibdev->iboe;
 	int i;
+	int err = 0;
 
-	for (i = 1; i <= ibdev->num_ports; ++i)
-		if (reset_gid_table(ibdev, i))
-			return -1;
+	for (i = 1; i <= ibdev->num_ports; ++i) {
+		if (rdma_port_get_link_layer(&ibdev->ib_dev, i) ==
+		    IB_LINK_LAYER_ETHERNET) {
+			err = reset_gid_table(ibdev, i);
+			if (err)
+				goto out;
+		}
+	}
 
 	read_lock(&dev_base_lock);
-	spin_lock(&iboe->lock);
+	spin_lock_bh(&iboe->lock);
 
 	for_each_netdev(&init_net, dev) {
 		u8 port = mlx4_ib_get_dev_port(dev, ibdev);
-		if (port)
+		/* port will be non-zero only for ETH ports */
+		if (port) {
+			mlx4_ib_set_default_gid(ibdev, dev, port);
 			mlx4_ib_get_dev_addr(dev, ibdev, port);
+		}
 	}
 
-	spin_unlock(&iboe->lock);
+	spin_unlock_bh(&iboe->lock);
 	read_unlock(&dev_base_lock);
-
-	return 0;
+out:
+	return err;
 }
 
 static void mlx4_ib_scan_netdevs(struct mlx4_ib_dev *ibdev,
@@ -1784,7 +1831,7 @@ static void mlx4_ib_scan_netdevs(struct mlx4_ib_dev *ibdev,
 
 	iboe = &ibdev->iboe;
 
-	spin_lock(&iboe->lock);
+	spin_lock_bh(&iboe->lock);
 	mlx4_foreach_ib_transport_port(port, ibdev->dev) {
 		enum ib_port_state	port_state = IB_PORT_NOP;
 		struct net_device *old_master = iboe->masters[port - 1];
@@ -1816,35 +1863,47 @@ static void mlx4_ib_scan_netdevs(struct mlx4_ib_dev *ibdev,
 			port_state = (netif_running(curr_netdev) && netif_carrier_ok(curr_netdev)) ?
 						IB_PORT_ACTIVE : IB_PORT_DOWN;
 			mlx4_ib_set_default_gid(ibdev, curr_netdev, port);
-		} else {
-			reset_gid_table(ibdev, port);
-		}
-		/* if using bonding/team and a slave port is down, we don't the bond IP
-		 * based gids in the table since flows that select port by gid may get
-		 * the down port.
-		 */
-		if (curr_master && (port_state == IB_PORT_DOWN)) {
-			reset_gid_table(ibdev, port);
-			mlx4_ib_set_default_gid(ibdev, curr_netdev, port);
-		}
-		/* if bonding is used it is possible that we add it to masters
-		 * only after IP address is assigned to the net bonding
-		 * interface.
-		*/
-		if (curr_master && (old_master != curr_master)) {
-			reset_gid_table(ibdev, port);
-			mlx4_ib_set_default_gid(ibdev, curr_netdev, port);
-			mlx4_ib_get_dev_addr(curr_master, ibdev, port);
-		}
+			if (curr_master) {
+				/* if using bonding/team and a slave port is down, we
+				 * don't want the bond IP based gids in the table since
+				 * flows that select port by gid may get the down port.
+				*/
+				if (port_state == IB_PORT_DOWN) {
+					reset_gid_table(ibdev, port);
+					mlx4_ib_set_default_gid(ibdev,
+								curr_netdev,
+								port);
+				} else {
+					/* gids from the upper dev (bond/team)
+					 * should appear in port's gid table
+					*/
+					mlx4_ib_get_dev_addr(curr_master,
+							     ibdev, port);
+				}
+			}
+			/* if bonding is used it is possible that we add it to
+			 * masters only after IP address is assigned to the
+			 * net bonding interface.
+			*/
+			if (curr_master && (old_master != curr_master)) {
+				reset_gid_table(ibdev, port);
+				mlx4_ib_set_default_gid(ibdev,
+							curr_netdev, port);
+				mlx4_ib_get_dev_addr(curr_master, ibdev, port);
+			}
 
-		if (!curr_master && (old_master != curr_master)) {
+			if (!curr_master && (old_master != curr_master)) {
+				reset_gid_table(ibdev, port);
+				mlx4_ib_set_default_gid(ibdev,
+							curr_netdev, port);
+				mlx4_ib_get_dev_addr(curr_netdev, ibdev, port);
+			}
+		} else {
 			reset_gid_table(ibdev, port);
-			mlx4_ib_set_default_gid(ibdev, curr_netdev, port);
-			mlx4_ib_get_dev_addr(curr_netdev, ibdev, port);
 		}
 	}
 
-	spin_unlock(&iboe->lock);
+	spin_unlock_bh(&iboe->lock);
 
 	if (update_qps_port > 0)
 		mlx4_ib_update_qps(ibdev, dev, update_qps_port);
@@ -2186,6 +2245,9 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 			goto err_steer_free_bitmap;
 	}
 
+	for (j = 1; j <= ibdev->dev->caps.num_ports; j++)
+		atomic64_set(&iboe->mac[j - 1], ibdev->dev->caps.def_mac[j]);
+
 	if (ib_register_device(&ibdev->ib_dev, NULL))
 		goto err_steer_free_bitmap;
 
@@ -2222,12 +2284,8 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 			}
 		}
 #endif
-		for (i = 1 ; i <= ibdev->num_ports ; ++i)
-			reset_gid_table(ibdev, i);
-		rtnl_lock();
-		mlx4_ib_scan_netdevs(ibdev, NULL, 0);
-		rtnl_unlock();
-		mlx4_ib_init_gid_table(ibdev);
+		if (mlx4_ib_init_gid_table(ibdev))
+			goto err_notif;
 	}
 
 	for (j = 0; j < ARRAY_SIZE(mlx4_class_attributes); ++j) {
@@ -2375,6 +2433,9 @@ static void mlx4_ib_remove(struct mlx4_dev *dev, void *ibdev_ptr)
 	struct mlx4_ib_dev *ibdev = ibdev_ptr;
 	int p;
 
+	ibdev->ib_active = false;
+	flush_workqueue(wq);
+
 	mlx4_ib_close_sriov(ibdev);
 	mlx4_ib_mad_cleanup(ibdev);
 	ib_unregister_device(&ibdev->ib_dev);

drivers/infiniband/hw/mlx4/mlx4_ib.h

@@ -451,6 +451,7 @@ struct mlx4_ib_iboe {
 	spinlock_t		lock;
 	struct net_device      *netdevs[MLX4_MAX_PORTS];
 	struct net_device      *masters[MLX4_MAX_PORTS];
+	atomic64_t		mac[MLX4_MAX_PORTS];
 	struct notifier_block 	nb;
 	struct notifier_block	nb_inet;
 	struct notifier_block	nb_inet6;

drivers/infiniband/hw/mlx4/mr.c

@@ -234,14 +234,13 @@ int mlx4_ib_rereg_user_mr(struct ib_mr *mr, int flags,
 					0);
 		if (IS_ERR(mmr->umem)) {
 			err = PTR_ERR(mmr->umem);
+			/* Prevent mlx4_ib_dereg_mr from free'ing invalid pointer */
 			mmr->umem = NULL;
 			goto release_mpt_entry;
 		}
 		n = ib_umem_page_count(mmr->umem);
 		shift = ilog2(mmr->umem->page_size);
 
-		mmr->mmr.iova       = virt_addr;
-		mmr->mmr.size       = length;
 		err = mlx4_mr_rereg_mem_write(dev->dev, &mmr->mmr,
 					      virt_addr, length, n, shift,
 					      *pmpt_entry);
@@ -249,6 +248,8 @@ int mlx4_ib_rereg_user_mr(struct ib_mr *mr, int flags,
 			ib_umem_release(mmr->umem);
 			goto release_mpt_entry;
 		}
+		mmr->mmr.iova       = virt_addr;
+		mmr->mmr.size       = length;
 
 		err = mlx4_ib_umem_write_mtt(dev, &mmr->mmr.mtt, mmr->umem);
 		if (err) {
@@ -262,6 +263,8 @@ int mlx4_ib_rereg_user_mr(struct ib_mr *mr, int flags,
 	 * return a failure. But dereg_mr will free the resources.
 	 */
 	err = mlx4_mr_hw_write_mpt(dev->dev, &mmr->mmr, pmpt_entry);
+	if (!err && flags & IB_MR_REREG_ACCESS)
+		mmr->mmr.access = mr_access_flags;
 
 release_mpt_entry:
 	mlx4_mr_hw_put_mpt(dev->dev, pmpt_entry);

drivers/infiniband/hw/mlx4/qp.c

@@ -964,9 +964,10 @@ static void destroy_qp_common(struct mlx4_ib_dev *dev, struct mlx4_ib_qp *qp,
 				   MLX4_QP_STATE_RST, NULL, 0, 0, &qp->mqp))
 			pr_warn("modify QP %06x to RESET failed.\n",
 			       qp->mqp.qpn);
-		if (qp->pri.smac) {
+		if (qp->pri.smac || (!qp->pri.smac && qp->pri.smac_port)) {
 			mlx4_unregister_mac(dev->dev, qp->pri.smac_port, qp->pri.smac);
 			qp->pri.smac = 0;
+			qp->pri.smac_port = 0;
 		}
 		if (qp->alt.smac) {
 			mlx4_unregister_mac(dev->dev, qp->alt.smac_port, qp->alt.smac);
@@ -1325,7 +1326,8 @@ static int _mlx4_set_path(struct mlx4_ib_dev *dev, const struct ib_ah_attr *ah,
 		 * If one was already assigned, but the new mac differs,
 		 * unregister the old one and register the new one.
 		*/
-		if (!smac_info->smac || smac_info->smac != smac) {
+		if ((!smac_info->smac && !smac_info->smac_port) ||
+		    smac_info->smac != smac) {
 			/* register candidate now, unreg if needed, after success */
 			smac_index = mlx4_register_mac(dev->dev, port, smac);
 			if (smac_index >= 0) {
@@ -1390,21 +1392,13 @@ static void update_mcg_macs(struct mlx4_ib_dev *dev, struct mlx4_ib_qp *qp)
 static int handle_eth_ud_smac_index(struct mlx4_ib_dev *dev, struct mlx4_ib_qp *qp, u8 *smac,
 				    struct mlx4_qp_context *context)
 {
-	struct net_device *ndev;
 	u64 u64_mac;
 	int smac_index;
 
-
-	ndev = dev->iboe.netdevs[qp->port - 1];
-	if (ndev) {
-		smac = ndev->dev_addr;
-		u64_mac = mlx4_mac_to_u64(smac);
-	} else {
-		u64_mac = dev->dev->caps.def_mac[qp->port];
-	}
+	u64_mac = atomic64_read(&dev->iboe.mac[qp->port - 1]);
 
 	context->pri_path.sched_queue = MLX4_IB_DEFAULT_SCHED_QUEUE | ((qp->port - 1) << 6);
-	if (!qp->pri.smac) {
+	if (!qp->pri.smac && !qp->pri.smac_port) {
 		smac_index = mlx4_register_mac(dev->dev, qp->port, u64_mac);
 		if (smac_index >= 0) {
 			qp->pri.candidate_smac_index = smac_index;
@@ -1432,6 +1426,12 @@ static int __mlx4_ib_modify_qp(struct ib_qp *ibqp,
 	int steer_qp = 0;
 	int err = -EINVAL;
 
+	/* APM is not supported under RoCE */
+	if (attr_mask & IB_QP_ALT_PATH &&
+	    rdma_port_get_link_layer(&dev->ib_dev, qp->port) ==
+	    IB_LINK_LAYER_ETHERNET)
+		return -ENOTSUPP;
+
 	context = kzalloc(sizeof *context, GFP_KERNEL);
 	if (!context)
 		return -ENOMEM;
@@ -1786,9 +1786,10 @@ static int __mlx4_ib_modify_qp(struct ib_qp *ibqp,
 			if (qp->flags & MLX4_IB_QP_NETIF)
 				mlx4_ib_steer_qp_reg(dev, qp, 0);
 		}
-		if (qp->pri.smac) {
+		if (qp->pri.smac || (!qp->pri.smac && qp->pri.smac_port)) {
 			mlx4_unregister_mac(dev->dev, qp->pri.smac_port, qp->pri.smac);
 			qp->pri.smac = 0;
+			qp->pri.smac_port = 0;
 		}
 		if (qp->alt.smac) {
 			mlx4_unregister_mac(dev->dev, qp->alt.smac_port, qp->alt.smac);
@@ -1812,11 +1813,12 @@ static int __mlx4_ib_modify_qp(struct ib_qp *ibqp,
 	if (err && steer_qp)
 		mlx4_ib_steer_qp_reg(dev, qp, 0);
 	kfree(context);
-	if (qp->pri.candidate_smac) {
+	if (qp->pri.candidate_smac ||
+	    (!qp->pri.candidate_smac && qp->pri.candidate_smac_port)) {
 		if (err) {
 			mlx4_unregister_mac(dev->dev, qp->pri.candidate_smac_port, qp->pri.candidate_smac);
 		} else {
-			if (qp->pri.smac)
+			if (qp->pri.smac || (!qp->pri.smac && qp->pri.smac_port))
 				mlx4_unregister_mac(dev->dev, qp->pri.smac_port, qp->pri.smac);
 			qp->pri.smac = qp->pri.candidate_smac;
 			qp->pri.smac_index = qp->pri.candidate_smac_index;
@@ -2089,6 +2091,16 @@ static int build_sriov_qp0_header(struct mlx4_ib_sqp *sqp,
 	return 0;
 }
 
+static void mlx4_u64_to_smac(u8 *dst_mac, u64 src_mac)
+{
+	int i;
+
+	for (i = ETH_ALEN; i; i--) {
+		dst_mac[i - 1] = src_mac & 0xff;
+		src_mac >>= 8;
+	}
+}
+
 static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_send_wr *wr,
 			    void *wqe, unsigned *mlx_seg_len)
 {
@@ -2203,7 +2215,6 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_send_wr *wr,
 	}
 
 	if (is_eth) {
-		u8 *smac;
 		struct in6_addr in6;
 
 		u16 pcp = (be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) >> 29) << 13;
@@ -2216,12 +2227,17 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_send_wr *wr,
 		memcpy(&ctrl->imm, ah->av.eth.mac + 2, 4);
 		memcpy(&in6, sgid.raw, sizeof(in6));
 
-		if (!mlx4_is_mfunc(to_mdev(ib_dev)->dev))
-			smac = to_mdev(sqp->qp.ibqp.device)->
-				iboe.netdevs[sqp->qp.port - 1]->dev_addr;
-		else	/* use the src mac of the tunnel */
-			smac = ah->av.eth.s_mac;
-		memcpy(sqp->ud_header.eth.smac_h, smac, 6);
+		if (!mlx4_is_mfunc(to_mdev(ib_dev)->dev)) {
+			u64 mac = atomic64_read(&to_mdev(ib_dev)->iboe.mac[sqp->qp.port - 1]);
+			u8 smac[ETH_ALEN];
+
+			mlx4_u64_to_smac(smac, mac);
+			memcpy(sqp->ud_header.eth.smac_h, smac, ETH_ALEN);
+		} else {
+			/* use the src mac of the tunnel */
+			memcpy(sqp->ud_header.eth.smac_h, ah->av.eth.s_mac, ETH_ALEN);
+		}
+
 		if (!memcmp(sqp->ud_header.eth.smac_h, sqp->ud_header.eth.dmac_h, 6))
 			mlx->flags |= cpu_to_be32(MLX4_WQE_CTRL_FORCE_LOOPBACK);
 		if (!is_vlan) {

drivers/infiniband/hw/ocrdma/ocrdma_ah.c

@@ -38,7 +38,7 @@
 #define OCRDMA_VID_PCP_SHIFT	0xD
 
 static inline int set_av_attr(struct ocrdma_dev *dev, struct ocrdma_ah *ah,
-				struct ib_ah_attr *attr, int pdid)
+			struct ib_ah_attr *attr, union ib_gid *sgid, int pdid)
 {
 	int status = 0;
 	u16 vlan_tag; bool vlan_enabled = false;
@@ -49,8 +49,7 @@ static inline int set_av_attr(struct ocrdma_dev *dev, struct ocrdma_ah *ah,
 	memset(&eth, 0, sizeof(eth));
 	memset(&grh, 0, sizeof(grh));
 
-	ah->sgid_index = attr->grh.sgid_index;
-
+	/* VLAN */
 	vlan_tag = attr->vlan_id;
 	if (!vlan_tag || (vlan_tag > 0xFFF))
 		vlan_tag = dev->pvid;
@@ -65,15 +64,14 @@ static inline int set_av_attr(struct ocrdma_dev *dev, struct ocrdma_ah *ah,
 		eth.eth_type = cpu_to_be16(OCRDMA_ROCE_ETH_TYPE);
 		eth_sz = sizeof(struct ocrdma_eth_basic);
 	}
+	/* MAC */
 	memcpy(&eth.smac[0], &dev->nic_info.mac_addr[0], ETH_ALEN);
-	memcpy(&eth.dmac[0], attr->dmac, ETH_ALEN);
 	status = ocrdma_resolve_dmac(dev, attr, &eth.dmac[0]);
 	if (status)
 		return status;
-	status = ocrdma_query_gid(&dev->ibdev, 1, attr->grh.sgid_index,
-			(union ib_gid *)&grh.sgid[0]);
-	if (status)
-		return status;
+	ah->sgid_index = attr->grh.sgid_index;
+	memcpy(&grh.sgid[0], sgid->raw, sizeof(union ib_gid));
+	memcpy(&grh.dgid[0], attr->grh.dgid.raw, sizeof(attr->grh.dgid.raw));
 
 	grh.tclass_flow = cpu_to_be32((6 << 28) |
 			(attr->grh.traffic_class << 24) |
@@ -81,8 +79,7 @@ static inline int set_av_attr(struct ocrdma_dev *dev, struct ocrdma_ah *ah,
 	/* 0x1b is next header value in GRH */
 	grh.pdid_hoplimit = cpu_to_be32((pdid << 16) |
 			(0x1b << 8) | attr->grh.hop_limit);
-
-	memcpy(&grh.dgid[0], attr->grh.dgid.raw, sizeof(attr->grh.dgid.raw));
+	/* Eth HDR */
 	memcpy(&ah->av->eth_hdr, &eth, eth_sz);
 	memcpy((u8 *)ah->av + eth_sz, &grh, sizeof(struct ocrdma_grh));
 	if (vlan_enabled)
@@ -98,6 +95,8 @@ struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr)
 	struct ocrdma_ah *ah;
 	struct ocrdma_pd *pd = get_ocrdma_pd(ibpd);
 	struct ocrdma_dev *dev = get_ocrdma_dev(ibpd->device);
+	union ib_gid sgid;
+	u8 zmac[ETH_ALEN];
 
 	if (!(attr->ah_flags & IB_AH_GRH))
 		return ERR_PTR(-EINVAL);
@@ -111,7 +110,27 @@ struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr)
 	status = ocrdma_alloc_av(dev, ah);
 	if (status)
 		goto av_err;
-	status = set_av_attr(dev, ah, attr, pd->id);
+
+	status = ocrdma_query_gid(&dev->ibdev, 1, attr->grh.sgid_index, &sgid);
+	if (status) {
+		pr_err("%s(): Failed to query sgid, status = %d\n",
+		      __func__, status);
+		goto av_conf_err;
+	}
+
+	memset(&zmac, 0, ETH_ALEN);
+	if (pd->uctx &&
+	    memcmp(attr->dmac, &zmac, ETH_ALEN)) {
+		status = rdma_addr_find_dmac_by_grh(&sgid, &attr->grh.dgid,
+                                        attr->dmac, &attr->vlan_id);
+		if (status) {
+			pr_err("%s(): Failed to resolve dmac from gid." 
+				"status = %d\n", __func__, status);
+			goto av_conf_err;
+		}
+	}
+
+	status = set_av_attr(dev, ah, attr, &sgid, pd->id);
 	if (status)
 		goto av_conf_err;
 
@@ -145,7 +164,7 @@ int ocrdma_query_ah(struct ib_ah *ibah, struct ib_ah_attr *attr)
 	struct ocrdma_av *av = ah->av;
 	struct ocrdma_grh *grh;
 	attr->ah_flags |= IB_AH_GRH;
-	if (ah->av->valid & Bit(1)) {
+	if (ah->av->valid & OCRDMA_AV_VALID) {
 		grh = (struct ocrdma_grh *)((u8 *)ah->av +
 				sizeof(struct ocrdma_eth_vlan));
 		attr->sl = be16_to_cpu(av->eth_hdr.vlan_tag) >> 13;

drivers/infiniband/hw/ocrdma/ocrdma_verbs.c

@@ -101,7 +101,7 @@ int ocrdma_query_device(struct ib_device *ibdev, struct ib_device_attr *attr)
 	attr->max_srq_sge = dev->attr.max_srq_sge;
 	attr->max_srq_wr = dev->attr.max_rqe;
 	attr->local_ca_ack_delay = dev->attr.local_ca_ack_delay;
-	attr->max_fast_reg_page_list_len = 0;
+	attr->max_fast_reg_page_list_len = dev->attr.max_pages_per_frmr;
 	attr->max_pkeys = 1;
 	return 0;
 }
@@ -2846,11 +2846,9 @@ int ocrdma_arm_cq(struct ib_cq *ibcq, enum ib_cq_notify_flags cq_flags)
 	if (cq->first_arm) {
 		ocrdma_ring_cq_db(dev, cq_id, arm_needed, sol_needed, 0);
 		cq->first_arm = false;
-		goto skip_defer;
 	}
-	cq->deferred_arm = true;
 
-skip_defer:
+	cq->deferred_arm = true;
 	cq->deferred_sol = sol_needed;
 	spin_unlock_irqrestore(&cq->cq_lock, flags);
 

drivers/infiniband/hw/qib/qib_debugfs.c

@@ -193,6 +193,7 @@ static void *_qp_stats_seq_start(struct seq_file *s, loff_t *pos)
 	struct qib_qp_iter *iter;
 	loff_t n = *pos;
 
+	rcu_read_lock();
 	iter = qib_qp_iter_init(s->private);
 	if (!iter)
 		return NULL;
@@ -224,7 +225,7 @@ static void *_qp_stats_seq_next(struct seq_file *s, void *iter_ptr,
 
 static void _qp_stats_seq_stop(struct seq_file *s, void *iter_ptr)
 {
-	/* nothing for now */
+	rcu_read_unlock();
 }
 
 static int _qp_stats_seq_show(struct seq_file *s, void *iter_ptr)

drivers/infiniband/hw/qib/qib_qp.c

@@ -1325,7 +1325,6 @@ int qib_qp_iter_next(struct qib_qp_iter *iter)
 	struct qib_qp *pqp = iter->qp;
 	struct qib_qp *qp;
 
-	rcu_read_lock();
 	for (; n < dev->qp_table_size; n++) {
 		if (pqp)
 			qp = rcu_dereference(pqp->next);
@@ -1333,18 +1332,11 @@ int qib_qp_iter_next(struct qib_qp_iter *iter)
 			qp = rcu_dereference(dev->qp_table[n]);
 		pqp = qp;
 		if (qp) {
-			if (iter->qp)
-				atomic_dec(&iter->qp->refcount);
-			atomic_inc(&qp->refcount);
-			rcu_read_unlock();
 			iter->qp = qp;
 			iter->n = n;
 			return 0;
 		}
 	}
-	rcu_read_unlock();
-	if (iter->qp)
-		atomic_dec(&iter->qp->refcount);
 	return ret;
 }
 

drivers/infiniband/hw/qib/qib_user_pages.c

@@ -52,7 +52,7 @@ static void __qib_release_user_pages(struct page **p, size_t num_pages,
  * Call with current->mm->mmap_sem held.
  */
 static int __qib_get_user_pages(unsigned long start_page, size_t num_pages,
-				struct page **p, struct vm_area_struct **vma)
+				struct page **p)
 {
 	unsigned long lock_limit;
 	size_t got;
@@ -69,7 +69,7 @@ static int __qib_get_user_pages(unsigned long start_page, size_t num_pages,
 		ret = get_user_pages(current, current->mm,
 				     start_page + got * PAGE_SIZE,
 				     num_pages - got, 1, 1,
-				     p + got, vma);
+				     p + got, NULL);
 		if (ret < 0)
 			goto bail_release;
 	}
@@ -136,7 +136,7 @@ int qib_get_user_pages(unsigned long start_page, size_t num_pages,
 
 	down_write(&current->mm->mmap_sem);
 
-	ret = __qib_get_user_pages(start_page, num_pages, p, NULL);
+	ret = __qib_get_user_pages(start_page, num_pages, p);
 
 	up_write(&current->mm->mmap_sem);
 

drivers/infiniband/ulp/ipoib/ipoib_multicast.c

@@ -529,21 +529,13 @@ void ipoib_mcast_join_task(struct work_struct *work)
 			  port_attr.state);
 		return;
 	}
+	priv->local_lid = port_attr.lid;
 
 	if (ib_query_gid(priv->ca, priv->port, 0, &priv->local_gid))
 		ipoib_warn(priv, "ib_query_gid() failed\n");
 	else
 		memcpy(priv->dev->dev_addr + 4, priv->local_gid.raw, sizeof (union ib_gid));
 
-	{
-		struct ib_port_attr attr;
-
-		if (!ib_query_port(priv->ca, priv->port, &attr))
-			priv->local_lid = attr.lid;
-		else
-			ipoib_warn(priv, "ib_query_port failed\n");
-	}
-
 	if (!priv->broadcast) {
 		struct ipoib_mcast *broadcast;
 

drivers/infiniband/ulp/iser/iscsi_iser.c

@@ -344,7 +344,6 @@ iscsi_iser_conn_bind(struct iscsi_cls_session *cls_session,
 		     int is_leading)
 {
 	struct iscsi_conn *conn = cls_conn->dd_data;
-	struct iscsi_session *session;
 	struct iser_conn *ib_conn;
 	struct iscsi_endpoint *ep;
 	int error;
@@ -363,9 +362,17 @@ iscsi_iser_conn_bind(struct iscsi_cls_session *cls_session,
 	}
 	ib_conn = ep->dd_data;
 
-	session = conn->session;
-	if (iser_alloc_rx_descriptors(ib_conn, session))
-		return -ENOMEM;
+	mutex_lock(&ib_conn->state_mutex);
+	if (ib_conn->state != ISER_CONN_UP) {
+		error = -EINVAL;
+		iser_err("iser_conn %p state is %d, teardown started\n",
+			 ib_conn, ib_conn->state);
+		goto out;
+	}
+
+	error = iser_alloc_rx_descriptors(ib_conn, conn->session);
+	if (error)
+		goto out;
 
 	/* binds the iSER connection retrieved from the previously
 	 * connected ep_handle to the iSCSI layer connection. exchanges
@@ -375,7 +382,9 @@ iscsi_iser_conn_bind(struct iscsi_cls_session *cls_session,
 	conn->dd_data = ib_conn;
 	ib_conn->iscsi_conn = conn;
 
-	return 0;
+out:
+	mutex_unlock(&ib_conn->state_mutex);
+	return error;
 }
 
 static int

drivers/infiniband/ulp/iser/iscsi_iser.h

@@ -69,7 +69,7 @@
 
 #define DRV_NAME	"iser"
 #define PFX		DRV_NAME ": "
-#define DRV_VER		"1.4"
+#define DRV_VER		"1.4.1"
 
 #define iser_dbg(fmt, arg...)				\
 	do {						\

drivers/infiniband/ulp/iser/iser_verbs.c

@@ -73,7 +73,7 @@ static int iser_create_device_ib_res(struct iser_device *device)
 {
 	struct iser_cq_desc *cq_desc;
 	struct ib_device_attr *dev_attr = &device->dev_attr;
-	int ret, i, j;
+	int ret, i;
 
 	ret = ib_query_device(device->ib_device, dev_attr);
 	if (ret) {
@@ -125,16 +125,20 @@ static int iser_create_device_ib_res(struct iser_device *device)
 					  iser_cq_event_callback,
 					  (void *)&cq_desc[i],
 					  ISER_MAX_RX_CQ_LEN, i);
-		if (IS_ERR(device->rx_cq[i]))
+		if (IS_ERR(device->rx_cq[i])) {
+			device->rx_cq[i] = NULL;
 			goto cq_err;
+		}
 
 		device->tx_cq[i] = ib_create_cq(device->ib_device,
 					  NULL, iser_cq_event_callback,
 					  (void *)&cq_desc[i],
 					  ISER_MAX_TX_CQ_LEN, i);
 
-		if (IS_ERR(device->tx_cq[i]))
+		if (IS_ERR(device->tx_cq[i])) {
+			device->tx_cq[i] = NULL;
 			goto cq_err;
+		}
 
 		if (ib_req_notify_cq(device->rx_cq[i], IB_CQ_NEXT_COMP))
 			goto cq_err;
@@ -160,14 +164,14 @@ static int iser_create_device_ib_res(struct iser_device *device)
 handler_err:
 	ib_dereg_mr(device->mr);
 dma_mr_err:
-	for (j = 0; j < device->cqs_used; j++)
-		tasklet_kill(&device->cq_tasklet[j]);
+	for (i = 0; i < device->cqs_used; i++)
+		tasklet_kill(&device->cq_tasklet[i]);
 cq_err:
-	for (j = 0; j < i; j++) {
-		if (device->tx_cq[j])
-			ib_destroy_cq(device->tx_cq[j]);
-		if (device->rx_cq[j])
-			ib_destroy_cq(device->rx_cq[j]);
+	for (i = 0; i < device->cqs_used; i++) {
+		if (device->tx_cq[i])
+			ib_destroy_cq(device->tx_cq[i]);
+		if (device->rx_cq[i])
+			ib_destroy_cq(device->rx_cq[i]);
 	}
 	ib_dealloc_pd(device->pd);
 pd_err:

drivers/net/ethernet/mellanox/mlx4/mr.c

@@ -298,6 +298,7 @@ static int mlx4_HW2SW_MPT(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox
 			    MLX4_CMD_TIME_CLASS_B, MLX4_CMD_WRAPPED);
 }
 
+/* Must protect against concurrent access */
 int mlx4_mr_hw_get_mpt(struct mlx4_dev *dev, struct mlx4_mr *mmr,
 		       struct mlx4_mpt_entry ***mpt_entry)
 {
@@ -305,13 +306,10 @@ int mlx4_mr_hw_get_mpt(struct mlx4_dev *dev, struct mlx4_mr *mmr,
 	int key = key_to_hw_index(mmr->key) & (dev->caps.num_mpts - 1);
 	struct mlx4_cmd_mailbox *mailbox = NULL;
 
-	/* Make sure that at this point we have single-threaded access only */
-
 	if (mmr->enabled != MLX4_MPT_EN_HW)
 		return -EINVAL;
 
 	err = mlx4_HW2SW_MPT(dev, NULL, key);
-
 	if (err) {
 		mlx4_warn(dev, "HW2SW_MPT failed (%d).", err);
 		mlx4_warn(dev, "Most likely the MR has MWs bound to it.\n");
@@ -333,7 +331,6 @@ int mlx4_mr_hw_get_mpt(struct mlx4_dev *dev, struct mlx4_mr *mmr,
 				   0, MLX4_CMD_QUERY_MPT,
 				   MLX4_CMD_TIME_CLASS_B,
 				   MLX4_CMD_WRAPPED);
-
 		if (err)
 			goto free_mailbox;
 
@@ -378,9 +375,10 @@ int mlx4_mr_hw_write_mpt(struct mlx4_dev *dev, struct mlx4_mr *mmr,
 		err = mlx4_SW2HW_MPT(dev, mailbox, key);
 	}
 
-	mmr->pd = be32_to_cpu((*mpt_entry)->pd_flags) & MLX4_MPT_PD_MASK;
-	if (!err)
+	if (!err) {
+		mmr->pd = be32_to_cpu((*mpt_entry)->pd_flags) & MLX4_MPT_PD_MASK;
 		mmr->enabled = MLX4_MPT_EN_HW;
+	}
 	return err;
 }
 EXPORT_SYMBOL_GPL(mlx4_mr_hw_write_mpt);
@@ -400,11 +398,12 @@ EXPORT_SYMBOL_GPL(mlx4_mr_hw_put_mpt);
 int mlx4_mr_hw_change_pd(struct mlx4_dev *dev, struct mlx4_mpt_entry *mpt_entry,
 			 u32 pdn)
 {
-	u32 pd_flags = be32_to_cpu(mpt_entry->pd_flags);
+	u32 pd_flags = be32_to_cpu(mpt_entry->pd_flags) & ~MLX4_MPT_PD_MASK;
 	/* The wrapper function will put the slave's id here */
 	if (mlx4_is_mfunc(dev))
 		pd_flags &= ~MLX4_MPT_PD_VF_MASK;
-	mpt_entry->pd_flags = cpu_to_be32((pd_flags &  ~MLX4_MPT_PD_MASK) |
+
+	mpt_entry->pd_flags = cpu_to_be32(pd_flags |
 					  (pdn & MLX4_MPT_PD_MASK)
 					  | MLX4_MPT_PD_FLAG_EN_INV);
 	return 0;
@@ -600,14 +599,18 @@ int mlx4_mr_rereg_mem_write(struct mlx4_dev *dev, struct mlx4_mr *mr,
 {
 	int err;
 
-	mpt_entry->start       = cpu_to_be64(mr->iova);
-	mpt_entry->length      = cpu_to_be64(mr->size);
-	mpt_entry->entity_size = cpu_to_be32(mr->mtt.page_shift);
+	mpt_entry->start       = cpu_to_be64(iova);
+	mpt_entry->length      = cpu_to_be64(size);
+	mpt_entry->entity_size = cpu_to_be32(page_shift);
 
 	err = mlx4_mtt_init(dev, npages, page_shift, &mr->mtt);
 	if (err)
 		return err;
 
+	mpt_entry->pd_flags &= cpu_to_be32(MLX4_MPT_PD_MASK |
+					   MLX4_MPT_PD_FLAG_EN_INV);
+	mpt_entry->flags    &= cpu_to_be32(MLX4_MPT_FLAG_FREE |
+					   MLX4_MPT_FLAG_SW_OWNS);
 	if (mr->mtt.order < 0) {
 		mpt_entry->flags |= cpu_to_be32(MLX4_MPT_FLAG_PHYSICAL);
 		mpt_entry->mtt_addr = 0;
@@ -617,6 +620,14 @@ int mlx4_mr_rereg_mem_write(struct mlx4_dev *dev, struct mlx4_mr *mr,
 		if (mr->mtt.page_shift == 0)
 			mpt_entry->mtt_sz    = cpu_to_be32(1 << mr->mtt.order);
 	}
+	if (mr->mtt.order >= 0 && mr->mtt.page_shift == 0) {
+		/* fast register MR in free state */
+		mpt_entry->flags    |= cpu_to_be32(MLX4_MPT_FLAG_FREE);
+		mpt_entry->pd_flags |= cpu_to_be32(MLX4_MPT_PD_FLAG_FAST_REG |
+						   MLX4_MPT_PD_FLAG_RAE);
+	} else {
+		mpt_entry->flags    |= cpu_to_be32(MLX4_MPT_FLAG_SW_OWNS);
+	}
 	mr->enabled = MLX4_MPT_EN_SW;
 
 	return 0;

drivers/net/ethernet/mellanox/mlx4/port.c

@@ -103,7 +103,8 @@ static int find_index(struct mlx4_dev *dev,
 	int i;
 
 	for (i = 0; i < MLX4_MAX_MAC_NUM; i++) {
-		if ((mac & MLX4_MAC_MASK) ==
+		if (table->refs[i] &&
+		    (MLX4_MAC_MASK & mac) ==
 		    (MLX4_MAC_MASK & be64_to_cpu(table->entries[i])))
 			return i;
 	}
@@ -165,12 +166,14 @@ int __mlx4_register_mac(struct mlx4_dev *dev, u8 port, u64 mac)
 
 	mutex_lock(&table->mutex);
 	for (i = 0; i < MLX4_MAX_MAC_NUM; i++) {
-		if (free < 0 && !table->entries[i]) {
-			free = i;
+		if (!table->refs[i]) {
+			if (free < 0)
+				free = i;
 			continue;
 		}
 
-		if (mac == (MLX4_MAC_MASK & be64_to_cpu(table->entries[i]))) {
+		if ((MLX4_MAC_MASK & mac) ==
+		     (MLX4_MAC_MASK & be64_to_cpu(table->entries[i]))) {
 			/* MAC already registered, increment ref count */
 			err = i;
 			++table->refs[i];

include/rdma/ib_umem.h

@@ -47,6 +47,7 @@ struct ib_umem {
 	int                     writable;
 	int                     hugetlb;
 	struct work_struct	work;
+	struct pid             *pid;
 	struct mm_struct       *mm;
 	unsigned long		diff;
 	struct sg_table sg_head;