1 files changed, 115 insertions, 102 deletions

diff --git a/kernel/drivers/infiniband/hw/mlx4/mad.c b/kernel/drivers/infiniband/hw/mlx4/mad.c
index 9cd2b002d..870e56b6b 100644
--- a/kernel/drivers/infiniband/hw/mlx4/mad.c
+++ b/kernel/drivers/infiniband/hw/mlx4/mad.c
@@ -64,14 +64,6 @@ enum {
 #define GUID_TBL_BLK_NUM_ENTRIES 8
 #define GUID_TBL_BLK_SIZE (GUID_TBL_ENTRY_SIZE * GUID_TBL_BLK_NUM_ENTRIES)
 
-/* Counters should be saturate once they reach their maximum value */
-#define ASSIGN_32BIT_COUNTER(counter, value) do {\
-	if ((value) > U32_MAX)			 \
-		counter = cpu_to_be32(U32_MAX); \
-	else					 \
-		counter = cpu_to_be32(value);	 \
-} while (0)
-
 struct mlx4_mad_rcv_buf {
 	struct ib_grh grh;
 	u8 payload[256];
@@ -111,8 +103,9 @@ __be64 mlx4_ib_get_new_demux_tid(struct mlx4_ib_demux_ctx *ctx)
 }
 
 int mlx4_MAD_IFC(struct mlx4_ib_dev *dev, int mad_ifc_flags,
-		 int port, struct ib_wc *in_wc, struct ib_grh *in_grh,
-		 void *in_mad, void *response_mad)
+		 int port, const struct ib_wc *in_wc,
+		 const struct ib_grh *in_grh,
+		 const void *in_mad, void *response_mad)
 {
 	struct mlx4_cmd_mailbox *inmailbox, *outmailbox;
 	void *inbox;
@@ -220,7 +213,7 @@ static void update_sm_ah(struct mlx4_ib_dev *dev, u8 port_num, u16 lid, u8 sl)
  * Snoop SM MADs for port info, GUID info, and  P_Key table sets, so we can
  * synthesize LID change, Client-Rereg, GID change, and P_Key change events.
  */
-static void smp_snoop(struct ib_device *ibdev, u8 port_num, struct ib_mad *mad,
+static void smp_snoop(struct ib_device *ibdev, u8 port_num, const struct ib_mad *mad,
 		      u16 prev_lid)
 {
 	struct ib_port_info *pinfo;
@@ -356,7 +349,7 @@ static void node_desc_override(struct ib_device *dev,
 	}
 }
 
-static void forward_trap(struct mlx4_ib_dev *dev, u8 port_num, struct ib_mad *mad)
+static void forward_trap(struct mlx4_ib_dev *dev, u8 port_num, const struct ib_mad *mad)
 {
 	int qpn = mad->mad_hdr.mgmt_class != IB_MGMT_CLASS_SUBN_LID_ROUTED;
 	struct ib_mad_send_buf *send_buf;
@@ -366,7 +359,8 @@ static void forward_trap(struct mlx4_ib_dev *dev, u8 port_num, struct ib_mad *ma
 
 	if (agent) {
 		send_buf = ib_create_send_mad(agent, qpn, 0, 0, IB_MGMT_MAD_HDR,
-					      IB_MGMT_MAD_DATA, GFP_ATOMIC);
+					      IB_MGMT_MAD_DATA, GFP_ATOMIC,
+					      IB_MGMT_BASE_VERSION);
 		if (IS_ERR(send_buf))
 			return;
 		/*
@@ -463,7 +457,8 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port,
 			  struct ib_grh *grh, struct ib_mad *mad)
 {
 	struct ib_sge list;
-	struct ib_send_wr wr, *bad_wr;
+	struct ib_ud_wr wr;
+	struct ib_send_wr *bad_wr;
 	struct mlx4_ib_demux_pv_ctx *tun_ctx;
 	struct mlx4_ib_demux_pv_qp *tun_qp;
 	struct mlx4_rcv_tunnel_mad *tun_mad;
@@ -586,20 +581,20 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port,
 
 	list.addr = tun_qp->tx_ring[tun_tx_ix].buf.map;
 	list.length = sizeof (struct mlx4_rcv_tunnel_mad);
-	list.lkey = tun_ctx->mr->lkey;
-
-	wr.wr.ud.ah = ah;
-	wr.wr.ud.port_num = port;
-	wr.wr.ud.remote_qkey = IB_QP_SET_QKEY;
-	wr.wr.ud.remote_qpn = dqpn;
-	wr.next = NULL;
-	wr.wr_id = ((u64) tun_tx_ix) | MLX4_TUN_SET_WRID_QPN(dest_qpt);
-	wr.sg_list = &list;
-	wr.num_sge = 1;
-	wr.opcode = IB_WR_SEND;
-	wr.send_flags = IB_SEND_SIGNALED;
-
-	ret = ib_post_send(src_qp, &wr, &bad_wr);
+	list.lkey = tun_ctx->pd->local_dma_lkey;
+
+	wr.ah = ah;
+	wr.port_num = port;
+	wr.remote_qkey = IB_QP_SET_QKEY;
+	wr.remote_qpn = dqpn;
+	wr.wr.next = NULL;
+	wr.wr.wr_id = ((u64) tun_tx_ix) | MLX4_TUN_SET_WRID_QPN(dest_qpt);
+	wr.wr.sg_list = &list;
+	wr.wr.num_sge = 1;
+	wr.wr.opcode = IB_WR_SEND;
+	wr.wr.send_flags = IB_SEND_SIGNALED;
+
+	ret = ib_post_send(src_qp, &wr.wr, &bad_wr);
 out:
 	if (ret)
 		ib_destroy_ah(ah);
@@ -722,8 +717,8 @@ static int mlx4_ib_demux_mad(struct ib_device *ibdev, u8 port,
 }
 
 static int ib_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num,
-			struct ib_wc *in_wc, struct ib_grh *in_grh,
-			struct ib_mad *in_mad, struct ib_mad *out_mad)
+			const struct ib_wc *in_wc, const struct ib_grh *in_grh,
+			const struct ib_mad *in_mad, struct ib_mad *out_mad)
 {
 	u16 slid, prev_lid = 0;
 	int err;
@@ -825,34 +820,39 @@ static void edit_counter(struct mlx4_counter *cnt,
 }
 
 static int iboe_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num,
-			struct ib_wc *in_wc, struct ib_grh *in_grh,
-			struct ib_mad *in_mad, struct ib_mad *out_mad)
+			const struct ib_wc *in_wc, const struct ib_grh *in_grh,
+			const struct ib_mad *in_mad, struct ib_mad *out_mad)
 {
-	struct mlx4_cmd_mailbox *mailbox;
+	struct mlx4_counter counter_stats;
 	struct mlx4_ib_dev *dev = to_mdev(ibdev);
-	int err;
-	u32 inmod = dev->counters[port_num - 1] & 0xffff;
-	u8 mode;
+	struct counter_index *tmp_counter;
+	int err = IB_MAD_RESULT_FAILURE, stats_avail = 0;
 
 	if (in_mad->mad_hdr.mgmt_class != IB_MGMT_CLASS_PERF_MGMT)
 		return -EINVAL;
 
-	mailbox = mlx4_alloc_cmd_mailbox(dev->dev);
-	if (IS_ERR(mailbox))
-		return IB_MAD_RESULT_FAILURE;
-
-	err = mlx4_cmd_box(dev->dev, 0, mailbox->dma, inmod, 0,
-			   MLX4_CMD_QUERY_IF_STAT, MLX4_CMD_TIME_CLASS_C,
-			   MLX4_CMD_WRAPPED);
-	if (err)
-		err = IB_MAD_RESULT_FAILURE;
-	else {
+	memset(&counter_stats, 0, sizeof(counter_stats));
+	mutex_lock(&dev->counters_table[port_num - 1].mutex);
+	list_for_each_entry(tmp_counter,
+			    &dev->counters_table[port_num - 1].counters_list,
+			    list) {
+		err = mlx4_get_counter_stats(dev->dev,
+					     tmp_counter->index,
+					     &counter_stats, 0);
+		if (err) {
+			err = IB_MAD_RESULT_FAILURE;
+			stats_avail = 0;
+			break;
+		}
+		stats_avail = 1;
+	}
+	mutex_unlock(&dev->counters_table[port_num - 1].mutex);
+	if (stats_avail) {
 		memset(out_mad->data, 0, sizeof out_mad->data);
-		mode = ((struct mlx4_counter *)mailbox->buf)->counter_mode;
-		switch (mode & 0xf) {
+		switch (counter_stats.counter_mode & 0xf) {
 		case 0:
-			edit_counter(mailbox->buf,
-						(void *)(out_mad->data + 40));
+			edit_counter(&counter_stats,
+				     (void *)(out_mad->data + 40));
 			err = IB_MAD_RESULT_SUCCESS | IB_MAD_RESULT_REPLY;
 			break;
 		default:
@@ -860,25 +860,43 @@ static int iboe_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num,
 		}
 	}
 
-	mlx4_free_cmd_mailbox(dev->dev, mailbox);
-
 	return err;
 }
 
 int mlx4_ib_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num,
-			struct ib_wc *in_wc, struct ib_grh *in_grh,
-			struct ib_mad *in_mad, struct ib_mad *out_mad)
+			const struct ib_wc *in_wc, const struct ib_grh *in_grh,
+			const struct ib_mad_hdr *in, size_t in_mad_size,
+			struct ib_mad_hdr *out, size_t *out_mad_size,
+			u16 *out_mad_pkey_index)
 {
-	switch (rdma_port_get_link_layer(ibdev, port_num)) {
-	case IB_LINK_LAYER_INFINIBAND:
+	struct mlx4_ib_dev *dev = to_mdev(ibdev);
+	const struct ib_mad *in_mad = (const struct ib_mad *)in;
+	struct ib_mad *out_mad = (struct ib_mad *)out;
+	enum rdma_link_layer link = rdma_port_get_link_layer(ibdev, port_num);
+
+	if (WARN_ON_ONCE(in_mad_size != sizeof(*in_mad) ||
+			 *out_mad_size != sizeof(*out_mad)))
+		return IB_MAD_RESULT_FAILURE;
+
+	/* iboe_process_mad() which uses the HCA flow-counters to implement IB PMA
+	 * queries, should be called only by VFs and for that specific purpose
+	 */
+	if (link == IB_LINK_LAYER_INFINIBAND) {
+		if (mlx4_is_slave(dev->dev) &&
+		    in_mad->mad_hdr.mgmt_class == IB_MGMT_CLASS_PERF_MGMT &&
+		    in_mad->mad_hdr.attr_id == IB_PMA_PORT_COUNTERS)
+			return iboe_process_mad(ibdev, mad_flags, port_num, in_wc,
+						in_grh, in_mad, out_mad);
+
 		return ib_process_mad(ibdev, mad_flags, port_num, in_wc,
 				      in_grh, in_mad, out_mad);
-	case IB_LINK_LAYER_ETHERNET:
-		return iboe_process_mad(ibdev, mad_flags, port_num, in_wc,
-					  in_grh, in_mad, out_mad);
-	default:
-		return -EINVAL;
 	}
+
+	if (link == IB_LINK_LAYER_ETHERNET)
+		return iboe_process_mad(ibdev, mad_flags, port_num, in_wc,
+					in_grh, in_mad, out_mad);
+
+	return -EINVAL;
 }
 
 static void send_handler(struct ib_mad_agent *agent,
@@ -1127,7 +1145,7 @@ static int mlx4_ib_post_pv_qp_buf(struct mlx4_ib_demux_pv_ctx *ctx,
 
 	sg_list.addr = tun_qp->ring[index].map;
 	sg_list.length = size;
-	sg_list.lkey = ctx->mr->lkey;
+	sg_list.lkey = ctx->pd->local_dma_lkey;
 
 	recv_wr.next = NULL;
 	recv_wr.sg_list = &sg_list;
@@ -1166,10 +1184,11 @@ static int is_proxy_qp0(struct mlx4_ib_dev *dev, int qpn, int slave)
 int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port,
 			 enum ib_qp_type dest_qpt, u16 pkey_index,
 			 u32 remote_qpn, u32 qkey, struct ib_ah_attr *attr,
-			 u8 *s_mac, struct ib_mad *mad)
+			 u8 *s_mac, u16 vlan_id, struct ib_mad *mad)
 {
 	struct ib_sge list;
-	struct ib_send_wr wr, *bad_wr;
+	struct ib_ud_wr wr;
+	struct ib_send_wr *bad_wr;
 	struct mlx4_ib_demux_pv_ctx *sqp_ctx;
 	struct mlx4_ib_demux_pv_qp *sqp;
 	struct mlx4_mad_snd_buf *sqp_mad;
@@ -1238,24 +1257,27 @@ int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port,
 
 	list.addr = sqp->tx_ring[wire_tx_ix].buf.map;
 	list.length = sizeof (struct mlx4_mad_snd_buf);
-	list.lkey = sqp_ctx->mr->lkey;
-
-	wr.wr.ud.ah = ah;
-	wr.wr.ud.port_num = port;
-	wr.wr.ud.pkey_index = wire_pkey_ix;
-	wr.wr.ud.remote_qkey = qkey;
-	wr.wr.ud.remote_qpn = remote_qpn;
-	wr.next = NULL;
-	wr.wr_id = ((u64) wire_tx_ix) | MLX4_TUN_SET_WRID_QPN(src_qpnum);
-	wr.sg_list = &list;
-	wr.num_sge = 1;
-	wr.opcode = IB_WR_SEND;
-	wr.send_flags = IB_SEND_SIGNALED;
+	list.lkey = sqp_ctx->pd->local_dma_lkey;
+
+	wr.ah = ah;
+	wr.port_num = port;
+	wr.pkey_index = wire_pkey_ix;
+	wr.remote_qkey = qkey;
+	wr.remote_qpn = remote_qpn;
+	wr.wr.next = NULL;
+	wr.wr.wr_id = ((u64) wire_tx_ix) | MLX4_TUN_SET_WRID_QPN(src_qpnum);
+	wr.wr.sg_list = &list;
+	wr.wr.num_sge = 1;
+	wr.wr.opcode = IB_WR_SEND;
+	wr.wr.send_flags = IB_SEND_SIGNALED;
 	if (s_mac)
 		memcpy(to_mah(ah)->av.eth.s_mac, s_mac, 6);
+	if (vlan_id < 0x1000)
+		vlan_id |= (attr->sl & 7) << 13;
+	to_mah(ah)->av.eth.vlan = cpu_to_be16(vlan_id);
 
 
-	ret = ib_post_send(send_qp, &wr, &bad_wr);
+	ret = ib_post_send(send_qp, &wr.wr, &bad_wr);
 out:
 	if (ret)
 		ib_destroy_ah(ah);
@@ -1289,6 +1311,7 @@ static void mlx4_ib_multiplex_mad(struct mlx4_ib_demux_pv_ctx *ctx, struct ib_wc
 	u8 *slave_id;
 	int slave;
 	int port;
+	u16 vlan_id;
 
 	/* Get slave that sent this packet */
 	if (wc->src_qp < dev->dev->phys_caps.base_proxy_sqpn ||
@@ -1365,19 +1388,22 @@ static void mlx4_ib_multiplex_mad(struct mlx4_ib_demux_pv_ctx *ctx, struct ib_wc
 	 * stadard address handle by decoding the tunnelled mlx4_ah fields */
 	memcpy(&ah.av, &tunnel->hdr.av, sizeof (struct mlx4_av));
 	ah.ibah.device = ctx->ib_dev;
+
+	port = be32_to_cpu(ah.av.ib.port_pd) >> 24;
+	port = mlx4_slave_convert_port(dev->dev, slave, port);
+	if (port < 0)
+		return;
+	ah.av.ib.port_pd = cpu_to_be32(port << 24 | (be32_to_cpu(ah.av.ib.port_pd) & 0xffffff));
+
 	mlx4_ib_query_ah(&ah.ibah, &ah_attr);
 	if (ah_attr.ah_flags & IB_AH_GRH)
 		fill_in_real_sgid_index(dev, slave, ctx->port, &ah_attr);
 
-	port = mlx4_slave_convert_port(dev->dev, slave, ah_attr.port_num);
-	if (port < 0)
-		return;
-	ah_attr.port_num = port;
 	memcpy(ah_attr.dmac, tunnel->hdr.mac, 6);
-	ah_attr.vlan_id = be16_to_cpu(tunnel->hdr.vlan);
+	vlan_id = be16_to_cpu(tunnel->hdr.vlan);
 	/* if slave have default vlan use it */
 	mlx4_get_slave_default_vlan(dev->dev, ctx->port, slave,
-				    &ah_attr.vlan_id, &ah_attr.sl);
+				    &vlan_id, &ah_attr.sl);
 
 	mlx4_ib_send_to_wire(dev, slave, ctx->port,
 			     is_proxy_qp0(dev, wc->src_qp, slave) ?
@@ -1385,7 +1411,7 @@ static void mlx4_ib_multiplex_mad(struct mlx4_ib_demux_pv_ctx *ctx, struct ib_wc
 			     be16_to_cpu(tunnel->hdr.pkey_index),
 			     be32_to_cpu(tunnel->hdr.remote_qpn),
 			     be32_to_cpu(tunnel->hdr.qkey),
-			     &ah_attr, wc->smac, &tunnel->mad);
+			     &ah_attr, wc->smac, vlan_id, &tunnel->mad);
 }
 
 static int mlx4_ib_alloc_pv_bufs(struct mlx4_ib_demux_pv_ctx *ctx,
@@ -1773,6 +1799,7 @@ static int create_pv_resources(struct ib_device *ibdev, int slave, int port,
 			       int create_tun, struct mlx4_ib_demux_pv_ctx *ctx)
 {
 	int ret, cq_size;
+	struct ib_cq_init_attr cq_attr = {};
 
 	if (ctx->state != DEMUX_PV_STATE_DOWN)
 		return -EEXIST;
@@ -1801,8 +1828,9 @@ static int create_pv_resources(struct ib_device *ibdev, int slave, int port,
 	if (ctx->has_smi)
 		cq_size *= 2;
 
+	cq_attr.cqe = cq_size;
 	ctx->cq = ib_create_cq(ctx->ib_dev, mlx4_ib_tunnel_comp_handler,
-			       NULL, ctx, cq_size, 0);
+			       NULL, ctx, &cq_attr);
 	if (IS_ERR(ctx->cq)) {
 		ret = PTR_ERR(ctx->cq);
 		pr_err("Couldn't create tunnel CQ (%d)\n", ret);
@@ -1816,19 +1844,12 @@ static int create_pv_resources(struct ib_device *ibdev, int slave, int port,
 		goto err_cq;
 	}
 
-	ctx->mr = ib_get_dma_mr(ctx->pd, IB_ACCESS_LOCAL_WRITE);
-	if (IS_ERR(ctx->mr)) {
-		ret = PTR_ERR(ctx->mr);
-		pr_err("Couldn't get tunnel DMA MR (%d)\n", ret);
-		goto err_pd;
-	}
-
 	if (ctx->has_smi) {
 		ret = create_pv_sqp(ctx, IB_QPT_SMI, create_tun);
 		if (ret) {
 			pr_err("Couldn't create %s QP0 (%d)\n",
 			       create_tun ? "tunnel for" : "",  ret);
-			goto err_mr;
+			goto err_pd;
 		}
 	}
 
@@ -1865,10 +1886,6 @@ err_qp0:
 		ib_destroy_qp(ctx->qp[0].qp);
 	ctx->qp[0].qp = NULL;
 
-err_mr:
-	ib_dereg_mr(ctx->mr);
-	ctx->mr = NULL;
-
 err_pd:
 	ib_dealloc_pd(ctx->pd);
 	ctx->pd = NULL;
@@ -1905,8 +1922,6 @@ static void destroy_pv_resources(struct mlx4_ib_dev *dev, int slave, int port,
 		ib_destroy_qp(ctx->qp[1].qp);
 		ctx->qp[1].qp = NULL;
 		mlx4_ib_free_pv_qp_bufs(ctx, IB_QPT_GSI, 1);
-		ib_dereg_mr(ctx->mr);
-		ctx->mr = NULL;
 		ib_dealloc_pd(ctx->pd);
 		ctx->pd = NULL;
 		ib_destroy_cq(ctx->cq);
@@ -2039,8 +2054,6 @@ static void mlx4_ib_free_sqp_ctx(struct mlx4_ib_demux_pv_ctx *sqp_ctx)
 		ib_destroy_qp(sqp_ctx->qp[1].qp);
 		sqp_ctx->qp[1].qp = NULL;
 		mlx4_ib_free_pv_qp_bufs(sqp_ctx, IB_QPT_GSI, 0);
-		ib_dereg_mr(sqp_ctx->mr);
-		sqp_ctx->mr = NULL;
 		ib_dealloc_pd(sqp_ctx->pd);
 		sqp_ctx->pd = NULL;
 		ib_destroy_cq(sqp_ctx->cq);