Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next

Pull networking changes from David Miller:
 "Noteworthy changes this time around:

   1) Multicast rejoin support for team driver, from Jiri Pirko.

   2) Centralize and simplify TCP RTT measurement handling in order to
      reduce the impact of bad RTO seeding from SYN/ACKs.  Also, when
      both timestamps and local RTT measurements are available prefer
      the later because there are broken middleware devices which
      scramble the timestamp.

      From Yuchung Cheng.

   3) Add TCP_NOTSENT_LOWAT socket option to limit the amount of kernel
      memory consumed to queue up unsend user data.  From Eric Dumazet.

   4) Add a "physical port ID" abstraction for network devices, from
      Jiri Pirko.

   5) Add a "suppress" operation to influence fib_rules lookups, from
      Stefan Tomanek.

   6) Add a networking development FAQ, from Paul Gortmaker.

   7) Extend the information provided by tcp_probe and add ipv6 support,
      from Daniel Borkmann.

   8) Use RCU locking more extensively in openvswitch data paths, from
      Pravin B Shelar.

   9) Add SCTP support to openvswitch, from Joe Stringer.

  10) Add EF10 chip support to SFC driver, from Ben Hutchings.

  11) Add new SYNPROXY netfilter target, from Patrick McHardy.

  12) Compute a rate approximation for sending in TCP sockets, and use
      this to more intelligently coalesce TSO frames.  Furthermore, add
      a new packet scheduler which takes advantage of this estimate when
      available.  From Eric Dumazet.

  13) Allow AF_PACKET fanouts with random selection, from Daniel
      Borkmann.

  14) Add ipv6 support to vxlan driver, from Cong Wang"

Resolved conflicts as per discussion.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next: (1218 commits)
  openvswitch: Fix alignment of struct sw_flow_key.
  netfilter: Fix build errors with xt_socket.c
  tcp: Add missing braces to do_tcp_setsockopt
  caif: Add missing braces to multiline if in cfctrl_linkup_request
  bnx2x: Add missing braces in bnx2x:bnx2x_link_initialize
  vxlan: Fix kernel panic on device delete.
  net: mvneta: implement ->ndo_do_ioctl() to support PHY ioctls
  net: mvneta: properly disable HW PHY polling and ensure adjust_link() works
  icplus: Use netif_running to determine device state
  ethernet/arc/arc_emac: Fix huge delays in large file copies
  tuntap: orphan frags before trying to set tx timestamp
  tuntap: purge socket error queue on detach
  qlcnic: use standard NAPI weights
  ipv6:introduce function to find route for redirect
  bnx2x: VF RSS support - VF side
  bnx2x: VF RSS support - PF side
  vxlan: Notify drivers for listening UDP port changes
  net: usbnet: update addr_assign_type if appropriate
  driver/net: enic: update enic maintainers and driver
  driver/net: enic: Exposing symbols for Cisco's low latency driver
  ...

9 files changed, 295 insertions, 92 deletions

diff --git a/drivers/net/ethernet/mellanox/mlx4/cmd.c b/drivers/net/ethernet/mellanox/mlx4/cmd.c
index 299d018..ea20182 100644
--- a/drivers/net/ethernet/mellanox/mlx4/cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx4/cmd.c
@@ -800,7 +800,16 @@ static int mlx4_MAD_IFC_wrapper(struct mlx4_dev *dev, int slave,
 				    vhcr->op, MLX4_CMD_TIME_CLASS_C, MLX4_CMD_NATIVE);
 }
 
-int MLX4_CMD_UPDATE_QP_wrapper(struct mlx4_dev *dev, int slave,
+static int MLX4_CMD_UPDATE_QP_wrapper(struct mlx4_dev *dev, int slave,
+		     struct mlx4_vhcr *vhcr,
+		     struct mlx4_cmd_mailbox *inbox,
+		     struct mlx4_cmd_mailbox *outbox,
+		     struct mlx4_cmd_info *cmd)
+{
+	return -EPERM;
+}
+
+static int MLX4_CMD_GET_OP_REQ_wrapper(struct mlx4_dev *dev, int slave,
 		     struct mlx4_vhcr *vhcr,
 		     struct mlx4_cmd_mailbox *inbox,
 		     struct mlx4_cmd_mailbox *outbox,
@@ -1252,6 +1261,15 @@ static struct mlx4_cmd_info cmd_info[] = {
 		.wrapper = MLX4_CMD_UPDATE_QP_wrapper
 	},
 	{
+		.opcode = MLX4_CMD_GET_OP_REQ,
+		.has_inbox = false,
+		.has_outbox = false,
+		.out_is_imm = false,
+		.encode_slave_id = false,
+		.verify = NULL,
+		.wrapper = MLX4_CMD_GET_OP_REQ_wrapper,
+	},
+	{
 		.opcode = MLX4_CMD_CONF_SPECIAL_QP,
 		.has_inbox = false,
 		.has_outbox = false,
@@ -1526,7 +1544,7 @@ static int calculate_transition(u16 oper_vlan, u16 admin_vlan)
 	return (2 * (oper_vlan == MLX4_VGT) + (admin_vlan == MLX4_VGT));
 }
 
-int mlx4_master_immediate_activate_vlan_qos(struct mlx4_priv *priv,
+static int mlx4_master_immediate_activate_vlan_qos(struct mlx4_priv *priv,
 					    int slave, int port)
 {
 	struct mlx4_vport_oper_state *vp_oper;
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_dcb_nl.c b/drivers/net/ethernet/mellanox/mlx4/en_dcb_nl.c
index 9d4a1ea..b4881b6 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_dcb_nl.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_dcb_nl.c
@@ -160,6 +160,7 @@ static int mlx4_en_dcbnl_ieee_setpfc(struct net_device *dev,
 		struct ieee_pfc *pfc)
 {
 	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_port_profile *prof = priv->prof;
 	struct mlx4_en_dev *mdev = priv->mdev;
 	int err;
 
@@ -169,15 +170,17 @@ static int mlx4_en_dcbnl_ieee_setpfc(struct net_device *dev,
 			pfc->mbc,
 			pfc->delay);
 
-	priv->prof->rx_pause = priv->prof->tx_pause = !!pfc->pfc_en;
-	priv->prof->rx_ppp = priv->prof->tx_ppp = pfc->pfc_en;
+	prof->rx_pause = !pfc->pfc_en;
+	prof->tx_pause = !pfc->pfc_en;
+	prof->rx_ppp = pfc->pfc_en;
+	prof->tx_ppp = pfc->pfc_en;
 
 	err = mlx4_SET_PORT_general(mdev->dev, priv->port,
 				    priv->rx_skb_size + ETH_FCS_LEN,
-				    priv->prof->tx_pause,
-				    priv->prof->tx_ppp,
-				    priv->prof->rx_pause,
-				    priv->prof->rx_ppp);
+				    prof->tx_pause,
+				    prof->tx_ppp,
+				    prof->rx_pause,
+				    prof->rx_ppp);
 	if (err)
 		en_err(priv, "Failed setting pause params\n");
 
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_tx.c b/drivers/net/ethernet/mellanox/mlx4/en_tx.c
index 7c49238..0698c82 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_tx.c
@@ -191,6 +191,39 @@ void mlx4_en_deactivate_tx_ring(struct mlx4_en_priv *priv,
 		       MLX4_QP_STATE_RST, NULL, 0, 0, &ring->qp);
 }
 
+static void mlx4_en_stamp_wqe(struct mlx4_en_priv *priv,
+			      struct mlx4_en_tx_ring *ring, int index,
+			      u8 owner)
+{
+	__be32 stamp = cpu_to_be32(STAMP_VAL | (!!owner << STAMP_SHIFT));
+	struct mlx4_en_tx_desc *tx_desc = ring->buf + index * TXBB_SIZE;
+	struct mlx4_en_tx_info *tx_info = &ring->tx_info[index];
+	void *end = ring->buf + ring->buf_size;
+	__be32 *ptr = (__be32 *)tx_desc;
+	int i;
+
+	/* Optimize the common case when there are no wraparounds */
+	if (likely((void *)tx_desc + tx_info->nr_txbb * TXBB_SIZE <= end)) {
+		/* Stamp the freed descriptor */
+		for (i = 0; i < tx_info->nr_txbb * TXBB_SIZE;
+		     i += STAMP_STRIDE) {
+			*ptr = stamp;
+			ptr += STAMP_DWORDS;
+		}
+	} else {
+		/* Stamp the freed descriptor */
+		for (i = 0; i < tx_info->nr_txbb * TXBB_SIZE;
+		     i += STAMP_STRIDE) {
+			*ptr = stamp;
+			ptr += STAMP_DWORDS;
+			if ((void *)ptr >= end) {
+				ptr = ring->buf;
+				stamp ^= cpu_to_be32(0x80000000);
+			}
+		}
+	}
+}
+
 
 static u32 mlx4_en_free_tx_desc(struct mlx4_en_priv *priv,
 				struct mlx4_en_tx_ring *ring,
@@ -205,8 +238,6 @@ static u32 mlx4_en_free_tx_desc(struct mlx4_en_priv *priv,
 	void *end = ring->buf + ring->buf_size;
 	int frags = skb_shinfo(skb)->nr_frags;
 	int i;
-	__be32 *ptr = (__be32 *)tx_desc;
-	__be32 stamp = cpu_to_be32(STAMP_VAL | (!!owner << STAMP_SHIFT));
 	struct skb_shared_hwtstamps hwts;
 
 	if (timestamp) {
@@ -232,12 +263,6 @@ static u32 mlx4_en_free_tx_desc(struct mlx4_en_priv *priv,
 					skb_frag_size(frag), PCI_DMA_TODEVICE);
 			}
 		}
-		/* Stamp the freed descriptor */
-		for (i = 0; i < tx_info->nr_txbb * TXBB_SIZE; i += STAMP_STRIDE) {
-			*ptr = stamp;
-			ptr += STAMP_DWORDS;
-		}
-
 	} else {
 		if (!tx_info->inl) {
 			if ((void *) data >= end) {
@@ -263,16 +288,6 @@ static u32 mlx4_en_free_tx_desc(struct mlx4_en_priv *priv,
 				++data;
 			}
 		}
-		/* Stamp the freed descriptor */
-		for (i = 0; i < tx_info->nr_txbb * TXBB_SIZE; i += STAMP_STRIDE) {
-			*ptr = stamp;
-			ptr += STAMP_DWORDS;
-			if ((void *) ptr >= end) {
-				ptr = ring->buf;
-				stamp ^= cpu_to_be32(0x80000000);
-			}
-		}
-
 	}
 	dev_kfree_skb_any(skb);
 	return tx_info->nr_txbb;
@@ -318,8 +333,9 @@ static void mlx4_en_process_tx_cq(struct net_device *dev, struct mlx4_en_cq *cq)
 	struct mlx4_en_tx_ring *ring = &priv->tx_ring[cq->ring];
 	struct mlx4_cqe *cqe;
 	u16 index;
-	u16 new_index, ring_index;
+	u16 new_index, ring_index, stamp_index;
 	u32 txbbs_skipped = 0;
+	u32 txbbs_stamp = 0;
 	u32 cons_index = mcq->cons_index;
 	int size = cq->size;
 	u32 size_mask = ring->size_mask;
@@ -335,6 +351,7 @@ static void mlx4_en_process_tx_cq(struct net_device *dev, struct mlx4_en_cq *cq)
 	index = cons_index & size_mask;
 	cqe = &buf[(index << factor) + factor];
 	ring_index = ring->cons & size_mask;
+	stamp_index = ring_index;
 
 	/* Process all completed CQEs */
 	while (XNOR(cqe->owner_sr_opcode & MLX4_CQE_OWNER_MASK,
@@ -345,6 +362,15 @@ static void mlx4_en_process_tx_cq(struct net_device *dev, struct mlx4_en_cq *cq)
 		 */
 		rmb();
 
+		if (unlikely((cqe->owner_sr_opcode & MLX4_CQE_OPCODE_MASK) ==
+			     MLX4_CQE_OPCODE_ERROR)) {
+			struct mlx4_err_cqe *cqe_err = (struct mlx4_err_cqe *)cqe;
+
+			en_err(priv, "CQE error - vendor syndrome: 0x%x syndrome: 0x%x\n",
+			       cqe_err->vendor_err_syndrome,
+			       cqe_err->syndrome);
+		}
+
 		/* Skip over last polled CQE */
 		new_index = be16_to_cpu(cqe->wqe_index) & size_mask;
 
@@ -359,6 +385,12 @@ static void mlx4_en_process_tx_cq(struct net_device *dev, struct mlx4_en_cq *cq)
 					priv, ring, ring_index,
 					!!((ring->cons + txbbs_skipped) &
 					ring->size), timestamp);
+
+			mlx4_en_stamp_wqe(priv, ring, stamp_index,
+					  !!((ring->cons + txbbs_stamp) &
+						ring->size));
+			stamp_index = ring_index;
+			txbbs_stamp = txbbs_skipped;
 			packets++;
 			bytes += ring->tx_info[ring_index].nr_bytes;
 		} while (ring_index != new_index);
@@ -556,17 +588,15 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev)
 {
 	struct mlx4_en_priv *priv = netdev_priv(dev);
 	struct mlx4_en_dev *mdev = priv->mdev;
+	struct device *ddev = priv->ddev;
 	struct mlx4_en_tx_ring *ring;
 	struct mlx4_en_tx_desc *tx_desc;
 	struct mlx4_wqe_data_seg *data;
-	struct skb_frag_struct *frag;
 	struct mlx4_en_tx_info *tx_info;
-	struct ethhdr *ethh;
 	int tx_ind = 0;
 	int nr_txbb;
 	int desc_size;
 	int real_size;
-	dma_addr_t dma;
 	u32 index, bf_index;
 	__be32 op_own;
 	u16 vlan_tag = 0;
@@ -642,6 +672,61 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev)
 	tx_info->skb = skb;
 	tx_info->nr_txbb = nr_txbb;
 
+	if (lso_header_size)
+		data = ((void *)&tx_desc->lso + ALIGN(lso_header_size + 4,
+						      DS_SIZE));
+	else
+		data = &tx_desc->data;
+
+	/* valid only for none inline segments */
+	tx_info->data_offset = (void *)data - (void *)tx_desc;
+
+	tx_info->linear = (lso_header_size < skb_headlen(skb) &&
+			   !is_inline(skb, NULL)) ? 1 : 0;
+
+	data += skb_shinfo(skb)->nr_frags + tx_info->linear - 1;
+
+	if (is_inline(skb, &fragptr)) {
+		tx_info->inl = 1;
+	} else {
+		/* Map fragments */
+		for (i = skb_shinfo(skb)->nr_frags - 1; i >= 0; i--) {
+			struct skb_frag_struct *frag;
+			dma_addr_t dma;
+
+			frag = &skb_shinfo(skb)->frags[i];
+			dma = skb_frag_dma_map(ddev, frag,
+					       0, skb_frag_size(frag),
+					       DMA_TO_DEVICE);
+			if (dma_mapping_error(ddev, dma))
+				goto tx_drop_unmap;
+
+			data->addr = cpu_to_be64(dma);
+			data->lkey = cpu_to_be32(mdev->mr.key);
+			wmb();
+			data->byte_count = cpu_to_be32(skb_frag_size(frag));
+			--data;
+		}
+
+		/* Map linear part */
+		if (tx_info->linear) {
+			u32 byte_count = skb_headlen(skb) - lso_header_size;
+			dma_addr_t dma;
+
+			dma = dma_map_single(ddev, skb->data +
+					     lso_header_size, byte_count,
+					     PCI_DMA_TODEVICE);
+			if (dma_mapping_error(ddev, dma))
+				goto tx_drop_unmap;
+
+			data->addr = cpu_to_be64(dma);
+			data->lkey = cpu_to_be32(mdev->mr.key);
+			wmb();
+			data->byte_count = cpu_to_be32(byte_count);
+		}
+		tx_info->inl = 0;
+	}
+
 	/*
 	 * For timestamping add flag to skb_shinfo and
 	 * set flag for further reference
@@ -666,6 +751,8 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev)
 	}
 
 	if (priv->flags & MLX4_EN_FLAG_ENABLE_HW_LOOPBACK) {
+		struct ethhdr *ethh;
+
 		/* Copy dst mac address to wqe. This allows loopback in eSwitch,
 		 * so that VFs and PF can communicate with each other
 		 */
@@ -688,8 +775,6 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev)
 		/* Copy headers;
 		 * note that we already verified that it is linear */
 		memcpy(tx_desc->lso.header, skb->data, lso_header_size);
-		data = ((void *) &tx_desc->lso +
-			ALIGN(lso_header_size + 4, DS_SIZE));
 
 		priv->port_stats.tso_packets++;
 		i = ((skb->len - lso_header_size) / skb_shinfo(skb)->gso_size) +
@@ -701,7 +786,6 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev)
 		op_own = cpu_to_be32(MLX4_OPCODE_SEND) |
 			((ring->prod & ring->size) ?
 			 cpu_to_be32(MLX4_EN_BIT_DESC_OWN) : 0);
-		data = &tx_desc->data;
 		tx_info->nr_bytes = max_t(unsigned int, skb->len, ETH_ZLEN);
 		ring->packets++;
 
@@ -710,38 +794,7 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev)
 	netdev_tx_sent_queue(ring->tx_queue, tx_info->nr_bytes);
 	AVG_PERF_COUNTER(priv->pstats.tx_pktsz_avg, skb->len);
 
-
-	/* valid only for none inline segments */
-	tx_info->data_offset = (void *) data - (void *) tx_desc;
-
-	tx_info->linear = (lso_header_size < skb_headlen(skb) && !is_inline(skb, NULL)) ? 1 : 0;
-	data += skb_shinfo(skb)->nr_frags + tx_info->linear - 1;
-
-	if (!is_inline(skb, &fragptr)) {
-		/* Map fragments */
-		for (i = skb_shinfo(skb)->nr_frags - 1; i >= 0; i--) {
-			frag = &skb_shinfo(skb)->frags[i];
-			dma = skb_frag_dma_map(priv->ddev, frag,
-					       0, skb_frag_size(frag),
-					       DMA_TO_DEVICE);
-			data->addr = cpu_to_be64(dma);
-			data->lkey = cpu_to_be32(mdev->mr.key);
-			wmb();
-			data->byte_count = cpu_to_be32(skb_frag_size(frag));
-			--data;
-		}
-
-		/* Map linear part */
-		if (tx_info->linear) {
-			dma = dma_map_single(priv->ddev, skb->data + lso_header_size,
-					     skb_headlen(skb) - lso_header_size, PCI_DMA_TODEVICE);
-			data->addr = cpu_to_be64(dma);
-			data->lkey = cpu_to_be32(mdev->mr.key);
-			wmb();
-			data->byte_count = cpu_to_be32(skb_headlen(skb) - lso_header_size);
-		}
-		tx_info->inl = 0;
-	} else {
+	if (tx_info->inl) {
 		build_inline_wqe(tx_desc, skb, real_size, &vlan_tag, tx_ind, fragptr);
 		tx_info->inl = 1;
 	}
@@ -781,6 +834,16 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev)
 
 	return NETDEV_TX_OK;
 
+tx_drop_unmap:
+	en_err(priv, "DMA mapping error\n");
+
+	for (i++; i < skb_shinfo(skb)->nr_frags; i++) {
+		data++;
+		dma_unmap_page(ddev, (dma_addr_t) be64_to_cpu(data->addr),
+			       be32_to_cpu(data->byte_count),
+			       PCI_DMA_TODEVICE);
+	}
+
 tx_drop:
 	dev_kfree_skb_any(skb);
 	priv->stats.tx_dropped++;
diff --git a/drivers/net/ethernet/mellanox/mlx4/eq.c b/drivers/net/ethernet/mellanox/mlx4/eq.c
index 7e04286..0416c5b 100644
--- a/drivers/net/ethernet/mellanox/mlx4/eq.c
+++ b/drivers/net/ethernet/mellanox/mlx4/eq.c
@@ -79,6 +79,7 @@ enum {
 			       (1ull << MLX4_EVENT_TYPE_SRQ_QP_LAST_WQE)    | \
 			       (1ull << MLX4_EVENT_TYPE_SRQ_LIMIT)	    | \
 			       (1ull << MLX4_EVENT_TYPE_CMD)		    | \
+			       (1ull << MLX4_EVENT_TYPE_OP_REQUIRED)	    | \
 			       (1ull << MLX4_EVENT_TYPE_COMM_CHANNEL)       | \
 			       (1ull << MLX4_EVENT_TYPE_FLR_EVENT)	    | \
 			       (1ull << MLX4_EVENT_TYPE_FATAL_WARNING))
@@ -629,6 +630,14 @@ static int mlx4_eq_int(struct mlx4_dev *dev, struct mlx4_eq *eq)
 			mlx4_warn(dev, "EQ overrun on EQN %d\n", eq->eqn);
 			break;
 
+		case MLX4_EVENT_TYPE_OP_REQUIRED:
+			atomic_inc(&priv->opreq_count);
+			/* FW commands can't be executed from interrupt context
+			 * working in deferred task
+			 */
+			queue_work(mlx4_wq, &priv->opreq_task);
+			break;
+
 		case MLX4_EVENT_TYPE_COMM_CHANNEL:
 			if (!mlx4_is_master(dev)) {
 				mlx4_warn(dev, "Received comm channel event "
diff --git a/drivers/net/ethernet/mellanox/mlx4/fw.c b/drivers/net/ethernet/mellanox/mlx4/fw.c
index 6fc6dab..0d63daa 100644
--- a/drivers/net/ethernet/mellanox/mlx4/fw.c
+++ b/drivers/net/ethernet/mellanox/mlx4/fw.c
@@ -1696,3 +1696,107 @@ int mlx4_wol_write(struct mlx4_dev *dev, u64 config, int port)
 			MLX4_CMD_TIME_CLASS_A, MLX4_CMD_NATIVE);
 }
 EXPORT_SYMBOL_GPL(mlx4_wol_write);
+
+enum {
+	ADD_TO_MCG = 0x26,
+};
+
+
+void mlx4_opreq_action(struct work_struct *work)
+{
+	struct mlx4_priv *priv = container_of(work, struct mlx4_priv,
+					      opreq_task);
+	struct mlx4_dev *dev = &priv->dev;
+	int num_tasks = atomic_read(&priv->opreq_count);
+	struct mlx4_cmd_mailbox *mailbox;
+	struct mlx4_mgm *mgm;
+	u32 *outbox;
+	u32 modifier;
+	u16 token;
+	u16 type_m;
+	u16 type;
+	int err;
+	u32 num_qps;
+	struct mlx4_qp qp;
+	int i;
+	u8 rem_mcg;
+	u8 prot;
+
+#define GET_OP_REQ_MODIFIER_OFFSET	0x08
+#define GET_OP_REQ_TOKEN_OFFSET		0x14
+#define GET_OP_REQ_TYPE_OFFSET		0x1a
+#define GET_OP_REQ_DATA_OFFSET		0x20
+
+	mailbox = mlx4_alloc_cmd_mailbox(dev);
+	if (IS_ERR(mailbox)) {
+		mlx4_err(dev, "Failed to allocate mailbox for GET_OP_REQ\n");
+		return;
+	}
+	outbox = mailbox->buf;
+
+	while (num_tasks) {
+		err = mlx4_cmd_box(dev, 0, mailbox->dma, 0, 0,
+				   MLX4_CMD_GET_OP_REQ, MLX4_CMD_TIME_CLASS_A,
+				   MLX4_CMD_NATIVE);
+		if (err) {
+			mlx4_err(dev, "Failed to retreive required operation: %d\n",
+				 err);
+			return;
+		}
+		MLX4_GET(modifier, outbox, GET_OP_REQ_MODIFIER_OFFSET);
+		MLX4_GET(token, outbox, GET_OP_REQ_TOKEN_OFFSET);
+		MLX4_GET(type, outbox, GET_OP_REQ_TYPE_OFFSET);
+		type_m = type >> 12;
+		type &= 0xfff;
+
+		switch (type) {
+		case ADD_TO_MCG:
+			if (dev->caps.steering_mode ==
+			    MLX4_STEERING_MODE_DEVICE_MANAGED) {
+				mlx4_warn(dev, "ADD MCG operation is not supported in DEVICE_MANAGED steering mode\n");
+				err = EPERM;
+				break;
+			}
+			mgm = (struct mlx4_mgm *)((u8 *)(outbox) +
+						  GET_OP_REQ_DATA_OFFSET);
+			num_qps = be32_to_cpu(mgm->members_count) &
+				  MGM_QPN_MASK;
+			rem_mcg = ((u8 *)(&mgm->members_count))[0] & 1;
+			prot = ((u8 *)(&mgm->members_count))[0] >> 6;
+
+			for (i = 0; i < num_qps; i++) {
+				qp.qpn = be32_to_cpu(mgm->qp[i]);
+				if (rem_mcg)
+					err = mlx4_multicast_detach(dev, &qp,
+								    mgm->gid,
+								    prot, 0);
+				else
+					err = mlx4_multicast_attach(dev, &qp,
+								    mgm->gid,
+								    mgm->gid[5]
+								    , 0, prot,
+								    NULL);
+				if (err)
+					break;
+			}
+			break;
+		default:
+			mlx4_warn(dev, "Bad type for required operation\n");
+			err = EINVAL;
+			break;
+		}
+		err = mlx4_cmd(dev, 0, ((u32) err | cpu_to_be32(token) << 16),
+			       1, MLX4_CMD_GET_OP_REQ, MLX4_CMD_TIME_CLASS_A,
+			       MLX4_CMD_NATIVE);
+		if (err) {
+			mlx4_err(dev, "Failed to acknowledge required request: %d\n",
+				 err);
+			goto out;
+		}
+		memset(outbox, 0, 0xffc);
+		num_tasks = atomic_dec_return(&priv->opreq_count);
+	}
+
+out:
+	mlx4_free_cmd_mailbox(dev, mailbox);
+}
diff --git a/drivers/net/ethernet/mellanox/mlx4/fw.h b/drivers/net/ethernet/mellanox/mlx4/fw.h
index fdf4166..a0a368b 100644
--- a/drivers/net/ethernet/mellanox/mlx4/fw.h
+++ b/drivers/net/ethernet/mellanox/mlx4/fw.h
@@ -220,5 +220,6 @@ int mlx4_MAP_ICM_AUX(struct mlx4_dev *dev, struct mlx4_icm *icm);
 int mlx4_UNMAP_ICM_AUX(struct mlx4_dev *dev);
 int mlx4_NOP(struct mlx4_dev *dev);
 int mlx4_MOD_STAT_CFG(struct mlx4_dev *dev, struct mlx4_mod_stat_cfg *cfg);
+void mlx4_opreq_action(struct work_struct *work);
 
 #endif /* MLX4_FW_H */
diff --git a/drivers/net/ethernet/mellanox/mlx4/main.c b/drivers/net/ethernet/mellanox/mlx4/main.c
index 36be320..60c9f4f 100644
--- a/drivers/net/ethernet/mellanox/mlx4/main.c
+++ b/drivers/net/ethernet/mellanox/mlx4/main.c
@@ -1692,11 +1692,19 @@ static int mlx4_setup_hca(struct mlx4_dev *dev)
 		goto err_xrcd_table_free;
 	}
 
+	if (!mlx4_is_slave(dev)) {
+		err = mlx4_init_mcg_table(dev);
+		if (err) {
+			mlx4_err(dev, "Failed to initialize multicast group table, aborting.\n");
+			goto err_mr_table_free;
+		}
+	}
+
 	err = mlx4_init_eq_table(dev);
 	if (err) {
 		mlx4_err(dev, "Failed to initialize "
 			 "event queue table, aborting.\n");
-		goto err_mr_table_free;
+		goto err_mcg_table_free;
 	}
 
 	err = mlx4_cmd_use_events(dev);
@@ -1746,19 +1754,10 @@ static int mlx4_setup_hca(struct mlx4_dev *dev)
 		goto err_srq_table_free;
 	}
 
-	if (!mlx4_is_slave(dev)) {
-		err = mlx4_init_mcg_table(dev);
-		if (err) {
-			mlx4_err(dev, "Failed to initialize "
-				 "multicast group table, aborting.\n");
-			goto err_qp_table_free;
-		}
-	}
-
 	err = mlx4_init_counters_table(dev);
 	if (err && err != -ENOENT) {
 		mlx4_err(dev, "Failed to initialize counters table, aborting.\n");
-		goto err_mcg_table_free;
+		goto err_qp_table_free;
 	}
 
 	if (!mlx4_is_slave(dev)) {
@@ -1803,9 +1802,6 @@ static int mlx4_setup_hca(struct mlx4_dev *dev)
 err_counters_table_free:
 	mlx4_cleanup_counters_table(dev);
 
-err_mcg_table_free:
-	mlx4_cleanup_mcg_table(dev);
-
 err_qp_table_free:
 	mlx4_cleanup_qp_table(dev);
 
@@ -1821,6 +1817,10 @@ err_cmd_poll:
 err_eq_table_free:
 	mlx4_cleanup_eq_table(dev);
 
+err_mcg_table_free:
+	if (!mlx4_is_slave(dev))
+		mlx4_cleanup_mcg_table(dev);
+
 err_mr_table_free:
 	mlx4_cleanup_mr_table(dev);
 
@@ -2197,6 +2197,9 @@ static int __mlx4_init_one(struct pci_dev *pdev, int pci_dev_data)
 			}
 		}
 
+		atomic_set(&priv->opreq_count, 0);
+		INIT_WORK(&priv->opreq_task, mlx4_opreq_action);
+
 		/*
 		 * Now reset the HCA before we touch the PCI capabilities or
 		 * attempt a firmware command, since a boot ROM may have left
@@ -2315,12 +2318,12 @@ err_port:
 		mlx4_cleanup_port_info(&priv->port[port]);
 
 	mlx4_cleanup_counters_table(dev);
-	mlx4_cleanup_mcg_table(dev);
 	mlx4_cleanup_qp_table(dev);
 	mlx4_cleanup_srq_table(dev);
 	mlx4_cleanup_cq_table(dev);
 	mlx4_cmd_use_polling(dev);
 	mlx4_cleanup_eq_table(dev);
+	mlx4_cleanup_mcg_table(dev);
 	mlx4_cleanup_mr_table(dev);
 	mlx4_cleanup_xrcd_table(dev);
 	mlx4_cleanup_pd_table(dev);
@@ -2403,12 +2406,12 @@ static void mlx4_remove_one(struct pci_dev *pdev)
 						   RES_TR_FREE_SLAVES_ONLY);
 
 		mlx4_cleanup_counters_table(dev);
-		mlx4_cleanup_mcg_table(dev);
 		mlx4_cleanup_qp_table(dev);
 		mlx4_cleanup_srq_table(dev);
 		mlx4_cleanup_cq_table(dev);
 		mlx4_cmd_use_polling(dev);
 		mlx4_cleanup_eq_table(dev);
+		mlx4_cleanup_mcg_table(dev);
 		mlx4_cleanup_mr_table(dev);
 		mlx4_cleanup_xrcd_table(dev);
 		mlx4_cleanup_pd_table(dev);
diff --git a/drivers/net/ethernet/mellanox/mlx4/mcg.c b/drivers/net/ethernet/mellanox/mlx4/mcg.c
index f3e804f..55f6245 100644
--- a/drivers/net/ethernet/mellanox/mlx4/mcg.c
+++ b/drivers/net/ethernet/mellanox/mlx4/mcg.c
@@ -39,19 +39,8 @@
 
 #include "mlx4.h"
 
-#define MGM_QPN_MASK       0x00FFFFFF
-#define MGM_BLCK_LB_BIT    30
-
 static const u8 zero_gid[16];	/* automatically initialized to 0 */
 
-struct mlx4_mgm {
-	__be32			next_gid_index;
-	__be32			members_count;
-	u32			reserved[2];
-	u8			gid[16];
-	__be32			qp[MLX4_MAX_QP_PER_MGM];
-};
-
 int mlx4_get_mgm_entry_size(struct mlx4_dev *dev)
 {
 	return 1 << dev->oper_log_mgm_entry_size;
diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4.h b/drivers/net/ethernet/mellanox/mlx4/mlx4.h
index 17d9277..348bb8c 100644
--- a/drivers/net/ethernet/mellanox/mlx4/mlx4.h
+++ b/drivers/net/ethernet/mellanox/mlx4/mlx4.h
@@ -554,6 +554,17 @@ struct mlx4_mfunc {
 	struct mlx4_mfunc_master_ctx	master;
 };
 
+#define MGM_QPN_MASK       0x00FFFFFF
+#define MGM_BLCK_LB_BIT    30
+
+struct mlx4_mgm {
+	__be32			next_gid_index;
+	__be32			members_count;
+	u32			reserved[2];
+	u8			gid[16];
+	__be32			qp[MLX4_MAX_QP_PER_MGM];
+};
+
 struct mlx4_cmd {
 	struct pci_pool	       *pool;
 	void __iomem	       *hcr;
@@ -802,6 +813,8 @@ struct mlx4_priv {
 	u8 virt2phys_pkey[MLX4_MFUNC_MAX][MLX4_MAX_PORTS][MLX4_MAX_PORT_PKEYS];
 	__be64			slave_node_guids[MLX4_MFUNC_MAX];
 
+	atomic_t		opreq_count;
+	struct work_struct	opreq_task;
 };
 
 static inline struct mlx4_priv *mlx4_priv(struct mlx4_dev *dev)