Merge remote-tracking branch 'net-next/master' into mac80211-next

Merge back net-next to get wireless driver changes (from Kalle)
to be able to create the API change across all trees properly.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>

41 files changed, 1578 insertions, 752 deletions

diff --git a/drivers/infiniband/hw/cxgb4/cm.c b/drivers/infiniband/hw/cxgb4/cm.c
index 57176ddd4c50..3ad8dc798f52 100644
--- a/drivers/infiniband/hw/cxgb4/cm.c
+++ b/drivers/infiniband/hw/cxgb4/cm.c
@@ -583,6 +583,22 @@ static void c4iw_record_pm_msg(struct c4iw_ep *ep,
 		sizeof(ep->com.mapped_remote_addr));
 }
 
+static int get_remote_addr(struct c4iw_ep *parent_ep, struct c4iw_ep *child_ep)
+{
+	int ret;
+
+	print_addr(&parent_ep->com, __func__, "get_remote_addr parent_ep ");
+	print_addr(&child_ep->com, __func__, "get_remote_addr child_ep ");
+
+	ret = iwpm_get_remote_info(&parent_ep->com.mapped_local_addr,
+				   &child_ep->com.mapped_remote_addr,
+				   &child_ep->com.remote_addr, RDMA_NL_C4IW);
+	if (ret)
+		PDBG("Unable to find remote peer addr info - err %d\n", ret);
+
+	return ret;
+}
+
 static void best_mtu(const unsigned short *mtus, unsigned short mtu,
 		     unsigned int *idx, int use_ts, int ipv6)
 {
@@ -675,7 +691,7 @@ static int send_connect(struct c4iw_ep *ep)
 	if (is_t5(ep->com.dev->rdev.lldi.adapter_type)) {
 		opt2 |= T5_OPT_2_VALID_F;
 		opt2 |= CONG_CNTRL_V(CONG_ALG_TAHOE);
-		opt2 |= CONG_CNTRL_VALID; /* OPT_2_ISS for T5 */
+		opt2 |= T5_ISS_F;
 	}
 	t4_set_arp_err_handler(skb, ep, act_open_req_arp_failure);
 
@@ -2042,9 +2058,12 @@ static int act_open_rpl(struct c4iw_dev *dev, struct sk_buff *skb)
 	     status, status2errno(status));
 
 	if (is_neg_adv(status)) {
-		dev_warn(&dev->rdev.lldi.pdev->dev,
-			 "Connection problems for atid %u status %u (%s)\n",
-			 atid, status, neg_adv_str(status));
+		PDBG("%s Connection problems for atid %u status %u (%s)\n",
+		     __func__, atid, status, neg_adv_str(status));
+		ep->stats.connect_neg_adv++;
+		mutex_lock(&dev->rdev.stats.lock);
+		dev->rdev.stats.neg_adv++;
+		mutex_unlock(&dev->rdev.stats.lock);
 		return 0;
 	}
 
@@ -2214,7 +2233,7 @@ static void accept_cr(struct c4iw_ep *ep, struct sk_buff *skb,
 		u32 isn = (prandom_u32() & ~7UL) - 1;
 		opt2 |= T5_OPT_2_VALID_F;
 		opt2 |= CONG_CNTRL_V(CONG_ALG_TAHOE);
-		opt2 |= CONG_CNTRL_VALID; /* OPT_2_ISS for T5 */
+		opt2 |= T5_ISS_F;
 		rpl5 = (void *)rpl;
 		memset(&rpl5->iss, 0, roundup(sizeof(*rpl5)-sizeof(*rpl), 16));
 		if (peer2peer)
@@ -2352,27 +2371,57 @@ static int pass_accept_req(struct c4iw_dev *dev, struct sk_buff *skb)
 	state_set(&child_ep->com, CONNECTING);
 	child_ep->com.dev = dev;
 	child_ep->com.cm_id = NULL;
+
+	/*
+	 * The mapped_local and mapped_remote addresses get setup with
+	 * the actual 4-tuple.  The local address will be based on the
+	 * actual local address of the connection, but on the port number
+	 * of the parent listening endpoint.  The remote address is
+	 * setup based on a query to the IWPM since we don't know what it
+	 * originally was before mapping.  If no mapping was done, then
+	 * mapped_remote == remote, and mapped_local == local.
+	 */
 	if (iptype == 4) {
 		struct sockaddr_in *sin = (struct sockaddr_in *)
-			&child_ep->com.local_addr;
+			&child_ep->com.mapped_local_addr;
+
 		sin->sin_family = PF_INET;
 		sin->sin_port = local_port;
 		sin->sin_addr.s_addr = *(__be32 *)local_ip;
-		sin = (struct sockaddr_in *)&child_ep->com.remote_addr;
+
+		sin = (struct sockaddr_in *)&child_ep->com.local_addr;
+		sin->sin_family = PF_INET;
+		sin->sin_port = ((struct sockaddr_in *)
+				 &parent_ep->com.local_addr)->sin_port;
+		sin->sin_addr.s_addr = *(__be32 *)local_ip;
+
+		sin = (struct sockaddr_in *)&child_ep->com.mapped_remote_addr;
 		sin->sin_family = PF_INET;
 		sin->sin_port = peer_port;
 		sin->sin_addr.s_addr = *(__be32 *)peer_ip;
 	} else {
 		struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)
-			&child_ep->com.local_addr;
+			&child_ep->com.mapped_local_addr;
+
 		sin6->sin6_family = PF_INET6;
 		sin6->sin6_port = local_port;
 		memcpy(sin6->sin6_addr.s6_addr, local_ip, 16);
-		sin6 = (struct sockaddr_in6 *)&child_ep->com.remote_addr;
+
+		sin6 = (struct sockaddr_in6 *)&child_ep->com.local_addr;
+		sin6->sin6_family = PF_INET6;
+		sin6->sin6_port = ((struct sockaddr_in6 *)
+				   &parent_ep->com.local_addr)->sin6_port;
+		memcpy(sin6->sin6_addr.s6_addr, local_ip, 16);
+
+		sin6 = (struct sockaddr_in6 *)&child_ep->com.mapped_remote_addr;
 		sin6->sin6_family = PF_INET6;
 		sin6->sin6_port = peer_port;
 		memcpy(sin6->sin6_addr.s6_addr, peer_ip, 16);
 	}
+	memcpy(&child_ep->com.remote_addr, &child_ep->com.mapped_remote_addr,
+	       sizeof(child_ep->com.remote_addr));
+	get_remote_addr(parent_ep, child_ep);
+
 	c4iw_get_ep(&parent_ep->com);
 	child_ep->parent_ep = parent_ep;
 	child_ep->tos = PASS_OPEN_TOS_G(ntohl(req->tos_stid));
@@ -2520,9 +2569,13 @@ static int peer_abort(struct c4iw_dev *dev, struct sk_buff *skb)
 
 	ep = lookup_tid(t, tid);
 	if (is_neg_adv(req->status)) {
-		dev_warn(&dev->rdev.lldi.pdev->dev,
-			 "Negative advice on abort - tid %u status %d (%s)\n",
-			 ep->hwtid, req->status, neg_adv_str(req->status));
+		PDBG("%s Negative advice on abort- tid %u status %d (%s)\n",
+		     __func__, ep->hwtid, req->status,
+		     neg_adv_str(req->status));
+		ep->stats.abort_neg_adv++;
+		mutex_lock(&dev->rdev.stats.lock);
+		dev->rdev.stats.neg_adv++;
+		mutex_unlock(&dev->rdev.stats.lock);
 		return 0;
 	}
 	PDBG("%s ep %p tid %u state %u\n", __func__, ep, ep->hwtid,
@@ -3571,7 +3624,7 @@ static void send_fw_pass_open_req(struct c4iw_dev *dev, struct sk_buff *skb,
 	 * TP will ignore any value > 0 for MSS index.
 	 */
 	req->tcb.opt0 = cpu_to_be64(MSS_IDX_V(0xF));
-	req->cookie = (unsigned long)skb;
+	req->cookie = (uintptr_t)skb;
 
 	set_wr_txq(req_skb, CPL_PRIORITY_CONTROL, port_id);
 	ret = cxgb4_ofld_send(dev->rdev.lldi.ports[0], req_skb);
@@ -3931,9 +3984,11 @@ static int peer_abort_intr(struct c4iw_dev *dev, struct sk_buff *skb)
 		return 0;
 	}
 	if (is_neg_adv(req->status)) {
-		dev_warn(&dev->rdev.lldi.pdev->dev,
-			 "Negative advice on abort - tid %u status %d (%s)\n",
-			 ep->hwtid, req->status, neg_adv_str(req->status));
+		PDBG("%s Negative advice on abort- tid %u status %d (%s)\n",
+		     __func__, ep->hwtid, req->status,
+		     neg_adv_str(req->status));
+		ep->stats.abort_neg_adv++;
+		dev->rdev.stats.neg_adv++;
 		kfree_skb(skb);
 		return 0;
 	}
diff --git a/drivers/infiniband/hw/cxgb4/cq.c b/drivers/infiniband/hw/cxgb4/cq.c
index ab7692ac2044..68ddb3710215 100644
--- a/drivers/infiniband/hw/cxgb4/cq.c
+++ b/drivers/infiniband/hw/cxgb4/cq.c
@@ -55,7 +55,7 @@ static int destroy_cq(struct c4iw_rdev *rdev, struct t4_cq *cq,
 			FW_RI_RES_WR_NRES_V(1) |
 			FW_WR_COMPL_F);
 	res_wr->len16_pkd = cpu_to_be32(DIV_ROUND_UP(wr_len, 16));
-	res_wr->cookie = (unsigned long) &wr_wait;
+	res_wr->cookie = (uintptr_t)&wr_wait;
 	res = res_wr->res;
 	res->u.cq.restype = FW_RI_RES_TYPE_CQ;
 	res->u.cq.op = FW_RI_RES_OP_RESET;
@@ -125,7 +125,7 @@ static int create_cq(struct c4iw_rdev *rdev, struct t4_cq *cq,
 			FW_RI_RES_WR_NRES_V(1) |
 			FW_WR_COMPL_F);
 	res_wr->len16_pkd = cpu_to_be32(DIV_ROUND_UP(wr_len, 16));
-	res_wr->cookie = (unsigned long) &wr_wait;
+	res_wr->cookie = (uintptr_t)&wr_wait;
 	res = res_wr->res;
 	res->u.cq.restype = FW_RI_RES_TYPE_CQ;
 	res->u.cq.op = FW_RI_RES_OP_WRITE;
@@ -156,12 +156,19 @@ static int create_cq(struct c4iw_rdev *rdev, struct t4_cq *cq,
 		goto err4;
 
 	cq->gen = 1;
-	cq->gts = rdev->lldi.gts_reg;
 	cq->rdev = rdev;
 	if (user) {
-		cq->ugts = (u64)pci_resource_start(rdev->lldi.pdev, 2) +
-					(cq->cqid << rdev->cqshift);
-		cq->ugts &= PAGE_MASK;
+		u32 off = (cq->cqid << rdev->cqshift) & PAGE_MASK;
+
+		cq->ugts = (u64)rdev->bar2_pa + off;
+	} else if (is_t4(rdev->lldi.adapter_type)) {
+		cq->gts = rdev->lldi.gts_reg;
+		cq->qid_mask = -1U;
+	} else {
+		u32 off = ((cq->cqid << rdev->cqshift) & PAGE_MASK) + 12;
+
+		cq->gts = rdev->bar2_kva + off;
+		cq->qid_mask = rdev->qpmask;
 	}
 	return 0;
 err4:
@@ -970,8 +977,7 @@ struct ib_cq *c4iw_create_cq(struct ib_device *ibdev, int entries,
 	}
 	PDBG("%s cqid 0x%0x chp %p size %u memsize %zu, dma_addr 0x%0llx\n",
 	     __func__, chp->cq.cqid, chp, chp->cq.size,
-	     chp->cq.memsize,
-	     (unsigned long long) chp->cq.dma_addr);
+	     chp->cq.memsize, (unsigned long long) chp->cq.dma_addr);
 	return &chp->ibcq;
 err5:
 	kfree(mm2);
diff --git a/drivers/infiniband/hw/cxgb4/device.c b/drivers/infiniband/hw/cxgb4/device.c
index 8fb295e4a9ab..7e895d714b19 100644
--- a/drivers/infiniband/hw/cxgb4/device.c
+++ b/drivers/infiniband/hw/cxgb4/device.c
@@ -93,6 +93,7 @@ static struct ibnl_client_cbs c4iw_nl_cb_table[] = {
 	[RDMA_NL_IWPM_ADD_MAPPING] = {.dump = iwpm_add_mapping_cb},
 	[RDMA_NL_IWPM_QUERY_MAPPING] = {.dump = iwpm_add_and_query_mapping_cb},
 	[RDMA_NL_IWPM_HANDLE_ERR] = {.dump = iwpm_mapping_error_cb},
+	[RDMA_NL_IWPM_REMOTE_INFO] = {.dump = iwpm_remote_info_cb},
 	[RDMA_NL_IWPM_MAPINFO] = {.dump = iwpm_mapping_info_cb},
 	[RDMA_NL_IWPM_MAPINFO_NUM] = {.dump = iwpm_ack_mapping_info_cb}
 };
@@ -151,7 +152,7 @@ static int wr_log_show(struct seq_file *seq, void *v)
 	int prev_ts_set = 0;
 	int idx, end;
 
-#define ts2ns(ts) div64_ul((ts) * dev->rdev.lldi.cclk_ps, 1000)
+#define ts2ns(ts) div64_u64((ts) * dev->rdev.lldi.cclk_ps, 1000)
 
 	idx = atomic_read(&dev->rdev.wr_log_idx) &
 		(dev->rdev.wr_log_size - 1);
@@ -489,6 +490,7 @@ static int stats_show(struct seq_file *seq, void *v)
 		   dev->rdev.stats.act_ofld_conn_fails);
 	seq_printf(seq, "PAS_OFLD_CONN_FAILS: %10llu\n",
 		   dev->rdev.stats.pas_ofld_conn_fails);
+	seq_printf(seq, "NEG_ADV_RCVD: %10llu\n", dev->rdev.stats.neg_adv);
 	seq_printf(seq, "AVAILABLE IRD: %10u\n", dev->avail_ird);
 	return 0;
 }
@@ -560,10 +562,13 @@ static int dump_ep(int id, void *p, void *data)
 		cc = snprintf(epd->buf + epd->pos, space,
 			      "ep %p cm_id %p qp %p state %d flags 0x%lx "
 			      "history 0x%lx hwtid %d atid %d "
+			      "conn_na %u abort_na %u "
 			      "%pI4:%d/%d <-> %pI4:%d/%d\n",
 			      ep, ep->com.cm_id, ep->com.qp,
 			      (int)ep->com.state, ep->com.flags,
 			      ep->com.history, ep->hwtid, ep->atid,
+			      ep->stats.connect_neg_adv,
+			      ep->stats.abort_neg_adv,
 			      &lsin->sin_addr, ntohs(lsin->sin_port),
 			      ntohs(mapped_lsin->sin_port),
 			      &rsin->sin_addr, ntohs(rsin->sin_port),
@@ -581,10 +586,13 @@ static int dump_ep(int id, void *p, void *data)
 		cc = snprintf(epd->buf + epd->pos, space,
 			      "ep %p cm_id %p qp %p state %d flags 0x%lx "
 			      "history 0x%lx hwtid %d atid %d "
+			      "conn_na %u abort_na %u "
 			      "%pI6:%d/%d <-> %pI6:%d/%d\n",
 			      ep, ep->com.cm_id, ep->com.qp,
 			      (int)ep->com.state, ep->com.flags,
 			      ep->com.history, ep->hwtid, ep->atid,
+			      ep->stats.connect_neg_adv,
+			      ep->stats.abort_neg_adv,
 			      &lsin6->sin6_addr, ntohs(lsin6->sin6_port),
 			      ntohs(mapped_lsin6->sin6_port),
 			      &rsin6->sin6_addr, ntohs(rsin6->sin6_port),
@@ -765,6 +773,29 @@ static int c4iw_rdev_open(struct c4iw_rdev *rdev)
 	c4iw_init_dev_ucontext(rdev, &rdev->uctx);
 
 	/*
+	 * This implementation assumes udb_density == ucq_density!  Eventually
+	 * we might need to support this but for now fail the open. Also the
+	 * cqid and qpid range must match for now.
+	 */
+	if (rdev->lldi.udb_density != rdev->lldi.ucq_density) {
+		pr_err(MOD "%s: unsupported udb/ucq densities %u/%u\n",
+		       pci_name(rdev->lldi.pdev), rdev->lldi.udb_density,
+		       rdev->lldi.ucq_density);
+		err = -EINVAL;
+		goto err1;
+	}
+	if (rdev->lldi.vr->qp.start != rdev->lldi.vr->cq.start ||
+	    rdev->lldi.vr->qp.size != rdev->lldi.vr->cq.size) {
+		pr_err(MOD "%s: unsupported qp and cq id ranges "
+		       "qp start %u size %u cq start %u size %u\n",
+		       pci_name(rdev->lldi.pdev), rdev->lldi.vr->qp.start,
+		       rdev->lldi.vr->qp.size, rdev->lldi.vr->cq.size,
+		       rdev->lldi.vr->cq.size);
+		err = -EINVAL;
+		goto err1;
+	}
+
+	/*
 	 * qpshift is the number of bits to shift the qpid left in order
 	 * to get the correct address of the doorbell for that qp.
 	 */
@@ -784,10 +815,10 @@ static int c4iw_rdev_open(struct c4iw_rdev *rdev)
 	     rdev->lldi.vr->qp.size,
 	     rdev->lldi.vr->cq.start,
 	     rdev->lldi.vr->cq.size);
-	PDBG("udb len 0x%x udb base %llx db_reg %p gts_reg %p qpshift %lu "
+	PDBG("udb len 0x%x udb base %p db_reg %p gts_reg %p qpshift %lu "
 	     "qpmask 0x%x cqshift %lu cqmask 0x%x\n",
 	     (unsigned)pci_resource_len(rdev->lldi.pdev, 2),
-	     (u64)pci_resource_start(rdev->lldi.pdev, 2),
+	     (void *)pci_resource_start(rdev->lldi.pdev, 2),
 	     rdev->lldi.db_reg,
 	     rdev->lldi.gts_reg,
 	     rdev->qpshift, rdev->qpmask,
@@ -1355,7 +1386,7 @@ static void recover_lost_dbs(struct uld_ctx *ctx, struct qp_list *qp_list)
 					  t4_sq_host_wq_pidx(&qp->wq),
 					  t4_sq_wq_size(&qp->wq));
 		if (ret) {
-			pr_err(KERN_ERR MOD "%s: Fatal error - "
+			pr_err(MOD "%s: Fatal error - "
 			       "DB overflow recovery failed - "
 			       "error syncing SQ qid %u\n",
 			       pci_name(ctx->lldi.pdev), qp->wq.sq.qid);
@@ -1371,7 +1402,7 @@ static void recover_lost_dbs(struct uld_ctx *ctx, struct qp_list *qp_list)
 					  t4_rq_wq_size(&qp->wq));
 
 		if (ret) {
-			pr_err(KERN_ERR MOD "%s: Fatal error - "
+			pr_err(MOD "%s: Fatal error - "
 			       "DB overflow recovery failed - "
 			       "error syncing RQ qid %u\n",
 			       pci_name(ctx->lldi.pdev), qp->wq.rq.qid);
diff --git a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
index d87e1650f643..97bb5550a6cf 100644
--- a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
+++ b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
@@ -137,6 +137,7 @@ struct c4iw_stats {
 	u64  tcam_full;
 	u64  act_ofld_conn_fails;
 	u64  pas_ofld_conn_fails;
+	u64  neg_adv;
 };
 
 struct c4iw_hw_queue {
@@ -814,6 +815,11 @@ struct c4iw_listen_ep {
 	int backlog;
 };
 
+struct c4iw_ep_stats {
+	unsigned connect_neg_adv;
+	unsigned abort_neg_adv;
+};
+
 struct c4iw_ep {
 	struct c4iw_ep_common com;
 	struct c4iw_ep *parent_ep;
@@ -846,6 +852,7 @@ struct c4iw_ep {
 	unsigned int retry_count;
 	int snd_win;
 	int rcv_win;
+	struct c4iw_ep_stats stats;
 };
 
 static inline void print_addr(struct c4iw_ep_common *epc, const char *func,
diff --git a/drivers/infiniband/hw/cxgb4/mem.c b/drivers/infiniband/hw/cxgb4/mem.c
index 3ef0cf9f5c44..cff815b91707 100644
--- a/drivers/infiniband/hw/cxgb4/mem.c
+++ b/drivers/infiniband/hw/cxgb4/mem.c
@@ -144,7 +144,7 @@ static int _c4iw_write_mem_inline(struct c4iw_rdev *rdev, u32 addr, u32 len,
 		if (i == (num_wqe-1)) {
 			req->wr.wr_hi = cpu_to_be32(FW_WR_OP_V(FW_ULPTX_WR) |
 						    FW_WR_COMPL_F);
-			req->wr.wr_lo = (__force __be64)(unsigned long) &wr_wait;
+			req->wr.wr_lo = (__force __be64)&wr_wait;
 		} else
 			req->wr.wr_hi = cpu_to_be32(FW_WR_OP_V(FW_ULPTX_WR));
 		req->wr.wr_mid = cpu_to_be32(
@@ -676,12 +676,12 @@ struct ib_mr *c4iw_get_dma_mr(struct ib_pd *pd, int acc)
 	mhp->attr.zbva = 0;
 	mhp->attr.va_fbo = 0;
 	mhp->attr.page_size = 0;
-	mhp->attr.len = ~0UL;
+	mhp->attr.len = ~0ULL;
 	mhp->attr.pbl_size = 0;
 
 	ret = write_tpt_entry(&rhp->rdev, 0, &stag, 1, php->pdid,
 			      FW_RI_STAG_NSMR, mhp->attr.perms,
-			      mhp->attr.mw_bind_enable, 0, 0, ~0UL, 0, 0, 0);
+			      mhp->attr.mw_bind_enable, 0, 0, ~0ULL, 0, 0, 0);
 	if (ret)
 		goto err1;
 
diff --git a/drivers/infiniband/hw/cxgb4/provider.c b/drivers/infiniband/hw/cxgb4/provider.c
index 66bd6a2ad83b..d95a0c300b03 100644
--- a/drivers/infiniband/hw/cxgb4/provider.c
+++ b/drivers/infiniband/hw/cxgb4/provider.c
@@ -445,10 +445,10 @@ static int c4iw_get_mib(struct ib_device *ibdev,
 
 	cxgb4_get_tcp_stats(c4iw_dev->rdev.lldi.pdev, &v4, &v6);
 	memset(stats, 0, sizeof *stats);
-	stats->iw.tcpInSegs = v4.tcpInSegs + v6.tcpInSegs;
-	stats->iw.tcpOutSegs = v4.tcpOutSegs + v6.tcpOutSegs;
-	stats->iw.tcpRetransSegs = v4.tcpRetransSegs + v6.tcpRetransSegs;
-	stats->iw.tcpOutRsts = v4.tcpOutRsts + v6.tcpOutSegs;
+	stats->iw.tcpInSegs = v4.tcp_in_segs + v6.tcp_in_segs;
+	stats->iw.tcpOutSegs = v4.tcp_out_segs + v6.tcp_out_segs;
+	stats->iw.tcpRetransSegs = v4.tcp_retrans_segs + v6.tcp_retrans_segs;
+	stats->iw.tcpOutRsts = v4.tcp_out_rsts + v6.tcp_out_rsts;
 
 	return 0;
 }
diff --git a/drivers/infiniband/hw/cxgb4/qp.c b/drivers/infiniband/hw/cxgb4/qp.c
index 15cae5a31018..389ced335bc5 100644
--- a/drivers/infiniband/hw/cxgb4/qp.c
+++ b/drivers/infiniband/hw/cxgb4/qp.c
@@ -275,7 +275,7 @@ static int create_qp(struct c4iw_rdev *rdev, struct t4_wq *wq,
 			FW_RI_RES_WR_NRES_V(2) |
 			FW_WR_COMPL_F);
 	res_wr->len16_pkd = cpu_to_be32(DIV_ROUND_UP(wr_len, 16));
-	res_wr->cookie = (unsigned long) &wr_wait;
+	res_wr->cookie = (uintptr_t)&wr_wait;
 	res = res_wr->res;
 	res->u.sqrq.restype = FW_RI_RES_TYPE_SQ;
 	res->u.sqrq.op = FW_RI_RES_OP_WRITE;
@@ -1209,7 +1209,7 @@ static int rdma_fini(struct c4iw_dev *rhp, struct c4iw_qp *qhp,
 	wqe->flowid_len16 = cpu_to_be32(
 		FW_WR_FLOWID_V(ep->hwtid) |
 		FW_WR_LEN16_V(DIV_ROUND_UP(sizeof(*wqe), 16)));
-	wqe->cookie = (unsigned long) &ep->com.wr_wait;
+	wqe->cookie = (uintptr_t)&ep->com.wr_wait;
 
 	wqe->u.fini.type = FW_RI_TYPE_FINI;
 	ret = c4iw_ofld_send(&rhp->rdev, skb);
@@ -1279,7 +1279,7 @@ static int rdma_init(struct c4iw_dev *rhp, struct c4iw_qp *qhp)
 		FW_WR_FLOWID_V(qhp->ep->hwtid) |
 		FW_WR_LEN16_V(DIV_ROUND_UP(sizeof(*wqe), 16)));
 
-	wqe->cookie = (unsigned long) &qhp->ep->com.wr_wait;
+	wqe->cookie = (uintptr_t)&qhp->ep->com.wr_wait;
 
 	wqe->u.init.type = FW_RI_TYPE_INIT;
 	wqe->u.init.mpareqbit_p2ptype =
@@ -1766,11 +1766,11 @@ struct ib_qp *c4iw_create_qp(struct ib_pd *pd, struct ib_qp_init_attr *attrs,
 		mm2->len = PAGE_ALIGN(qhp->wq.rq.memsize);
 		insert_mmap(ucontext, mm2);
 		mm3->key = uresp.sq_db_gts_key;
-		mm3->addr = (__force unsigned long) qhp->wq.sq.udb;
+		mm3->addr = (__force unsigned long)qhp->wq.sq.udb;
 		mm3->len = PAGE_SIZE;
 		insert_mmap(ucontext, mm3);
 		mm4->key = uresp.rq_db_gts_key;
-		mm4->addr = (__force unsigned long) qhp->wq.rq.udb;
+		mm4->addr = (__force unsigned long)qhp->wq.rq.udb;
 		mm4->len = PAGE_SIZE;
 		insert_mmap(ucontext, mm4);
 		if (mm5) {
diff --git a/drivers/infiniband/hw/cxgb4/t4.h b/drivers/infiniband/hw/cxgb4/t4.h
index 871cdcac7be2..7f2a6c244d25 100644
--- a/drivers/infiniband/hw/cxgb4/t4.h
+++ b/drivers/infiniband/hw/cxgb4/t4.h
@@ -539,6 +539,7 @@ struct t4_cq {
 	size_t memsize;
 	__be64 bits_type_ts;
 	u32 cqid;
+	u32 qid_mask;
 	int vector;
 	u16 size; /* including status page */
 	u16 cidx;
@@ -563,12 +564,12 @@ static inline int t4_arm_cq(struct t4_cq *cq, int se)
 	set_bit(CQ_ARMED, &cq->flags);
 	while (cq->cidx_inc > CIDXINC_M) {
 		val = SEINTARM_V(0) | CIDXINC_V(CIDXINC_M) | TIMERREG_V(7) |
-		      INGRESSQID_V(cq->cqid);
+		      INGRESSQID_V(cq->cqid & cq->qid_mask);
 		writel(val, cq->gts);
 		cq->cidx_inc -= CIDXINC_M;
 	}
 	val = SEINTARM_V(se) | CIDXINC_V(cq->cidx_inc) | TIMERREG_V(6) |
-	      INGRESSQID_V(cq->cqid);
+	      INGRESSQID_V(cq->cqid & cq->qid_mask);
 	writel(val, cq->gts);
 	cq->cidx_inc = 0;
 	return 0;
@@ -601,7 +602,7 @@ static inline void t4_hwcq_consume(struct t4_cq *cq)
 		u32 val;
 
 		val = SEINTARM_V(0) | CIDXINC_V(cq->cidx_inc) | TIMERREG_V(7) |
-		      INGRESSQID_V(cq->cqid);
+		      INGRESSQID_V(cq->cqid & cq->qid_mask);
 		writel(val, cq->gts);
 		cq->cidx_inc = 0;
 	}
diff --git a/drivers/infiniband/hw/cxgb4/t4fw_ri_api.h b/drivers/infiniband/hw/cxgb4/t4fw_ri_api.h
index 5e53327fc647..343e8daf2270 100644
--- a/drivers/infiniband/hw/cxgb4/t4fw_ri_api.h
+++ b/drivers/infiniband/hw/cxgb4/t4fw_ri_api.h
@@ -848,6 +848,8 @@ enum {                     /* TCP congestion control algorithms */
 #define CONG_CNTRL_V(x) ((x) << CONG_CNTRL_S)
 #define CONG_CNTRL_G(x) (((x) >> CONG_CNTRL_S) & CONG_CNTRL_M)
 
-#define CONG_CNTRL_VALID   (1 << 18)
+#define T5_ISS_S    18
+#define T5_ISS_V(x) ((x) << T5_ISS_S)
+#define T5_ISS_F    T5_ISS_V(1U)
 
 #endif /* _T4FW_RI_API_H_ */
diff --git a/drivers/infiniband/hw/ehca/ehca_mcast.c b/drivers/infiniband/hw/ehca/ehca_mcast.c
index 120aedf9f989..cec181532924 100644
--- a/drivers/infiniband/hw/ehca/ehca_mcast.c
+++ b/drivers/infiniband/hw/ehca/ehca_mcast.c
@@ -77,7 +77,7 @@ int ehca_attach_mcast(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
 		return -EINVAL;
 	}
 
-	memcpy(&my_gid.raw, gid->raw, sizeof(union ib_gid));
+	memcpy(&my_gid, gid->raw, sizeof(union ib_gid));
 
 	subnet_prefix = be64_to_cpu(my_gid.global.subnet_prefix);
 	interface_id = be64_to_cpu(my_gid.global.interface_id);
@@ -114,7 +114,7 @@ int ehca_detach_mcast(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
 		return -EINVAL;
 	}
 
-	memcpy(&my_gid.raw, gid->raw, sizeof(union ib_gid));
+	memcpy(&my_gid, gid->raw, sizeof(union ib_gid));
 
 	subnet_prefix = be64_to_cpu(my_gid.global.subnet_prefix);
 	interface_id = be64_to_cpu(my_gid.global.interface_id);
diff --git a/drivers/infiniband/hw/ipath/ipath_fs.c b/drivers/infiniband/hw/ipath/ipath_fs.c
index 33c45dfcbd88..1ca8e32a9592 100644
--- a/drivers/infiniband/hw/ipath/ipath_fs.c
+++ b/drivers/infiniband/hw/ipath/ipath_fs.c
@@ -82,14 +82,14 @@ static int create_file(const char *name, umode_t mode,
 {
 	int error;
 
-	mutex_lock(&parent->d_inode->i_mutex);
+	mutex_lock(&d_inode(parent)->i_mutex);
 	*dentry = lookup_one_len(name, parent, strlen(name));
 	if (!IS_ERR(*dentry))
-		error = ipathfs_mknod(parent->d_inode, *dentry,
+		error = ipathfs_mknod(d_inode(parent), *dentry,
 				      mode, fops, data);
 	else
 		error = PTR_ERR(*dentry);
-	mutex_unlock(&parent->d_inode->i_mutex);
+	mutex_unlock(&d_inode(parent)->i_mutex);
 
 	return error;
 }
@@ -277,11 +277,11 @@ static int remove_file(struct dentry *parent, char *name)
 	}
 
 	spin_lock(&tmp->d_lock);
-	if (!d_unhashed(tmp) && tmp->d_inode) {
+	if (!d_unhashed(tmp) && d_really_is_positive(tmp)) {
 		dget_dlock(tmp);
 		__d_drop(tmp);
 		spin_unlock(&tmp->d_lock);
-		simple_unlink(parent->d_inode, tmp);
+		simple_unlink(d_inode(parent), tmp);
 	} else
 		spin_unlock(&tmp->d_lock);
 
@@ -302,7 +302,7 @@ static int remove_device_files(struct super_block *sb,
 	int ret;
 
 	root = dget(sb->s_root);
-	mutex_lock(&root->d_inode->i_mutex);
+	mutex_lock(&d_inode(root)->i_mutex);
 	snprintf(unit, sizeof unit, "%02d", dd->ipath_unit);
 	dir = lookup_one_len(unit, root, strlen(unit));
 
@@ -315,10 +315,10 @@ static int remove_device_files(struct super_block *sb,
 	remove_file(dir, "flash");
 	remove_file(dir, "atomic_counters");
 	d_delete(dir);
-	ret = simple_rmdir(root->d_inode, dir);
+	ret = simple_rmdir(d_inode(root), dir);
 
 bail:
-	mutex_unlock(&root->d_inode->i_mutex);
+	mutex_unlock(&d_inode(root)->i_mutex);
 	dput(root);
 	return ret;
 }
diff --git a/drivers/infiniband/hw/mlx4/alias_GUID.c b/drivers/infiniband/hw/mlx4/alias_GUID.c
index a31e031afd87..21cb41a60fe8 100644
--- a/drivers/infiniband/hw/mlx4/alias_GUID.c
+++ b/drivers/infiniband/hw/mlx4/alias_GUID.c
@@ -58,14 +58,19 @@ struct mlx4_alias_guid_work_context {
 	int			query_id;
 	struct list_head	list;
 	int			block_num;
+	ib_sa_comp_mask		guid_indexes;
+	u8			method;
 };
 
 struct mlx4_next_alias_guid_work {
 	u8 port;
 	u8 block_num;
+	u8 method;
 	struct mlx4_sriov_alias_guid_info_rec_det rec_det;
 };
 
+static int get_low_record_time_index(struct mlx4_ib_dev *dev, u8 port,
+				     int *resched_delay_sec);
 
 void mlx4_ib_update_cache_on_guid_change(struct mlx4_ib_dev *dev, int block_num,
 					 u8 port_num, u8 *p_data)
@@ -118,6 +123,57 @@ ib_sa_comp_mask mlx4_ib_get_aguid_comp_mask_from_ix(int index)
 	return IB_SA_COMP_MASK(4 + index);
 }
 
+void mlx4_ib_slave_alias_guid_event(struct mlx4_ib_dev *dev, int slave,
+				    int port,  int slave_init)
+{
+	__be64 curr_guid, required_guid;
+	int record_num = slave / 8;
+	int index = slave % 8;
+	int port_index = port - 1;
+	unsigned long flags;
+	int do_work = 0;
+
+	spin_lock_irqsave(&dev->sriov.alias_guid.ag_work_lock, flags);
+	if (dev->sriov.alias_guid.ports_guid[port_index].state_flags &
+	    GUID_STATE_NEED_PORT_INIT)
+		goto unlock;
+	if (!slave_init) {
+		curr_guid = *(__be64 *)&dev->sriov.
+			alias_guid.ports_guid[port_index].
+			all_rec_per_port[record_num].
+			all_recs[GUID_REC_SIZE * index];
+		if (curr_guid == cpu_to_be64(MLX4_GUID_FOR_DELETE_VAL) ||
+		    !curr_guid)
+			goto unlock;
+		required_guid = cpu_to_be64(MLX4_GUID_FOR_DELETE_VAL);
+	} else {
+		required_guid = mlx4_get_admin_guid(dev->dev, slave, port);
+		if (required_guid == cpu_to_be64(MLX4_GUID_FOR_DELETE_VAL))
+			goto unlock;
+	}
+	*(__be64 *)&dev->sriov.alias_guid.ports_guid[port_index].
+		all_rec_per_port[record_num].
+		all_recs[GUID_REC_SIZE * index] = required_guid;
+	dev->sriov.alias_guid.ports_guid[port_index].
+		all_rec_per_port[record_num].guid_indexes
+		|= mlx4_ib_get_aguid_comp_mask_from_ix(index);
+	dev->sriov.alias_guid.ports_guid[port_index].
+		all_rec_per_port[record_num].status
+		= MLX4_GUID_INFO_STATUS_IDLE;
+	/* set to run immediately */
+	dev->sriov.alias_guid.ports_guid[port_index].
+		all_rec_per_port[record_num].time_to_run = 0;
+	dev->sriov.alias_guid.ports_guid[port_index].
+		all_rec_per_port[record_num].
+		guids_retry_schedule[index] = 0;
+	do_work = 1;
+unlock:
+	spin_unlock_irqrestore(&dev->sriov.alias_guid.ag_work_lock, flags);
+
+	if (do_work)
+		mlx4_ib_init_alias_guid_work(dev, port_index);
+}
+
 /*
  * Whenever new GUID is set/unset (guid table change) create event and
  * notify the relevant slave (master also should be notified).
@@ -133,15 +189,20 @@ void mlx4_ib_notify_slaves_on_guid_change(struct mlx4_ib_dev *dev,
 {
 	int i;
 	u64 guid_indexes;
-	int slave_id;
+	int slave_id, slave_port;
 	enum slave_port_state new_state;
 	enum slave_port_state prev_state;
 	__be64 tmp_cur_ag, form_cache_ag;
 	enum slave_port_gen_event gen_event;
+	struct mlx4_sriov_alias_guid_info_rec_det *rec;
+	unsigned long flags;
+	__be64 required_value;
 
 	if (!mlx4_is_master(dev->dev))
 		return;
 
+	rec = &dev->sriov.alias_guid.ports_guid[port_num - 1].
+			all_rec_per_port[block_num];
 	guid_indexes = be64_to_cpu((__force __be64) dev->sriov.alias_guid.
 				   ports_guid[port_num - 1].
 				   all_rec_per_port[block_num].guid_indexes);
@@ -156,6 +217,11 @@ void mlx4_ib_notify_slaves_on_guid_change(struct mlx4_ib_dev *dev,
 		slave_id = (block_num * NUM_ALIAS_GUID_IN_REC) + i ;
 		if (slave_id >= dev->dev->persist->num_vfs + 1)
 			return;
+
+		slave_port = mlx4_phys_to_slave_port(dev->dev, slave_id, port_num);
+		if (slave_port < 0) /* this port isn't available for the VF */
+			continue;
+
 		tmp_cur_ag = *(__be64 *)&p_data[i * GUID_REC_SIZE];
 		form_cache_ag = get_cached_alias_guid(dev, port_num,
 					(NUM_ALIAS_GUID_IN_REC * block_num) + i);
@@ -166,8 +232,27 @@ void mlx4_ib_notify_slaves_on_guid_change(struct mlx4_ib_dev *dev,
 		 */
 		if (tmp_cur_ag != form_cache_ag)
 			continue;
-		mlx4_gen_guid_change_eqe(dev->dev, slave_id, port_num);
 
+		spin_lock_irqsave(&dev->sriov.alias_guid.ag_work_lock, flags);
+		required_value = *(__be64 *)&rec->all_recs[i * GUID_REC_SIZE];
+
+		if (required_value == cpu_to_be64(MLX4_GUID_FOR_DELETE_VAL))
+			required_value = 0;
+
+		if (tmp_cur_ag == required_value) {
+			rec->guid_indexes = rec->guid_indexes &
+			       ~mlx4_ib_get_aguid_comp_mask_from_ix(i);
+		} else {
+			/* may notify port down if value is 0 */
+			if (tmp_cur_ag != MLX4_NOT_SET_GUID) {
+				spin_unlock_irqrestore(&dev->sriov.
+					alias_guid.ag_work_lock, flags);
+				continue;
+			}
+		}
+		spin_unlock_irqrestore(&dev->sriov.alias_guid.ag_work_lock,
+				       flags);
+		mlx4_gen_guid_change_eqe(dev->dev, slave_id, port_num);
 		/*2 cases: Valid GUID, and Invalid Guid*/
 
 		if (tmp_cur_ag != MLX4_NOT_SET_GUID) { /*valid GUID*/
@@ -188,10 +273,14 @@ void mlx4_ib_notify_slaves_on_guid_change(struct mlx4_ib_dev *dev,
 			set_and_calc_slave_port_state(dev->dev, slave_id, port_num,
 						      MLX4_PORT_STATE_IB_EVENT_GID_INVALID,
 						      &gen_event);
-			pr_debug("sending PORT DOWN event to slave: %d, port: %d\n",
-				 slave_id, port_num);
-			mlx4_gen_port_state_change_eqe(dev->dev, slave_id, port_num,
-						       MLX4_PORT_CHANGE_SUBTYPE_DOWN);
+			if (gen_event == SLAVE_PORT_GEN_EVENT_DOWN) {
+				pr_debug("sending PORT DOWN event to slave: %d, port: %d\n",
+					 slave_id, port_num);
+				mlx4_gen_port_state_change_eqe(dev->dev,
+							       slave_id,
+							       port_num,
+							       MLX4_PORT_CHANGE_SUBTYPE_DOWN);
+			}
 		}
 	}
 }
@@ -206,6 +295,9 @@ static void aliasguid_query_handler(int status,
 	int i;
 	struct mlx4_sriov_alias_guid_info_rec_det *rec;
 	unsigned long flags, flags1;
+	ib_sa_comp_mask declined_guid_indexes = 0;
+	ib_sa_comp_mask applied_guid_indexes = 0;
+	unsigned int resched_delay_sec = 0;
 
 	if (!context)
 		return;
@@ -216,9 +308,9 @@ static void aliasguid_query_handler(int status,
 		all_rec_per_port[cb_ctx->block_num];
 
 	if (status) {
-		rec->status = MLX4_GUID_INFO_STATUS_IDLE;
 		pr_debug("(port: %d) failed: status = %d\n",
 			 cb_ctx->port, status);
+		rec->time_to_run = ktime_get_real_ns() + 1 * NSEC_PER_SEC;
 		goto out;
 	}
 
@@ -235,57 +327,101 @@ static void aliasguid_query_handler(int status,
 	rec = &dev->sriov.alias_guid.ports_guid[port_index].
 		all_rec_per_port[guid_rec->block_num];
 
-	rec->status = MLX4_GUID_INFO_STATUS_SET;
-	rec->method = MLX4_GUID_INFO_RECORD_SET;
-
+	spin_lock_irqsave(&dev->sriov.alias_guid.ag_work_lock, flags);
 	for (i = 0 ; i < NUM_ALIAS_GUID_IN_REC; i++) {
-		__be64 tmp_cur_ag;
-		tmp_cur_ag = *(__be64 *)&guid_rec->guid_info_list[i * GUID_REC_SIZE];
+		__be64 sm_response, required_val;
+
+		if (!(cb_ctx->guid_indexes &
+			mlx4_ib_get_aguid_comp_mask_from_ix(i)))
+			continue;
+		sm_response = *(__be64 *)&guid_rec->guid_info_list
+				[i * GUID_REC_SIZE];
+		required_val = *(__be64 *)&rec->all_recs[i * GUID_REC_SIZE];
+		if (cb_ctx->method == MLX4_GUID_INFO_RECORD_DELETE) {
+			if (required_val ==
+			    cpu_to_be64(MLX4_GUID_FOR_DELETE_VAL))
+				goto next_entry;
+
+			/* A new value was set till we got the response */
+			pr_debug("need to set new value %llx, record num %d, block_num:%d\n",
+				 be64_to_cpu(required_val),
+				 i, guid_rec->block_num);
+			goto entry_declined;
+		}
+
 		/* check if the SM didn't assign one of the records.
-		 * if it didn't, if it was not sysadmin request:
-		 * ask the SM to give a new GUID, (instead of the driver request).
+		 * if it didn't, re-ask for.
 		 */
-		if (tmp_cur_ag == MLX4_NOT_SET_GUID) {
-			mlx4_ib_warn(&dev->ib_dev, "%s:Record num %d in "
-				     "block_num: %d was declined by SM, "
-				     "ownership by %d (0 = driver, 1=sysAdmin,"
-				     " 2=None)\n", __func__, i,
-				     guid_rec->block_num, rec->ownership);
-			if (rec->ownership == MLX4_GUID_DRIVER_ASSIGN) {
-				/* if it is driver assign, asks for new GUID from SM*/
-				*(__be64 *)&rec->all_recs[i * GUID_REC_SIZE] =
-					MLX4_NOT_SET_GUID;
-
-				/* Mark the record as not assigned, and let it
-				 * be sent again in the next work sched.*/
-				rec->status = MLX4_GUID_INFO_STATUS_IDLE;
-				rec->guid_indexes |= mlx4_ib_get_aguid_comp_mask_from_ix(i);
-			}
+		if (sm_response == MLX4_NOT_SET_GUID) {
+			if (rec->guids_retry_schedule[i] == 0)
+				mlx4_ib_warn(&dev->ib_dev,
+					     "%s:Record num %d in  block_num: %d was declined by SM\n",
+					     __func__, i,
+					     guid_rec->block_num);
+			goto entry_declined;
 		} else {
 		       /* properly assigned record. */
 		       /* We save the GUID we just got from the SM in the
 			* admin_guid in order to be persistent, and in the
 			* request from the sm the process will ask for the same GUID */
-			if (rec->ownership == MLX4_GUID_SYSADMIN_ASSIGN &&
-			    tmp_cur_ag != *(__be64 *)&rec->all_recs[i * GUID_REC_SIZE]) {
-				/* the sysadmin assignment failed.*/
-				mlx4_ib_warn(&dev->ib_dev, "%s: Failed to set"
-					     " admin guid after SysAdmin "
-					     "configuration. "
-					     "Record num %d in block_num:%d "
-					     "was declined by SM, "
-					     "new val(0x%llx) was kept\n",
-					      __func__, i,
-					     guid_rec->block_num,
-					     be64_to_cpu(*(__be64 *) &
-							 rec->all_recs[i * GUID_REC_SIZE]));
+			if (required_val &&
+			    sm_response != required_val) {
+				/* Warn only on first retry */
+				if (rec->guids_retry_schedule[i] == 0)
+					mlx4_ib_warn(&dev->ib_dev, "%s: Failed to set"
+						     " admin guid after SysAdmin "
+						     "configuration. "
+						     "Record num %d in block_num:%d "
+						     "was declined by SM, "
+						     "new val(0x%llx) was kept, SM returned (0x%llx)\n",
+						      __func__, i,
+						     guid_rec->block_num,
+						     be64_to_cpu(required_val),
+						     be64_to_cpu(sm_response));
+				goto entry_declined;
 			} else {
-				memcpy(&rec->all_recs[i * GUID_REC_SIZE],
-				       &guid_rec->guid_info_list[i * GUID_REC_SIZE],
-				       GUID_REC_SIZE);
+				*(__be64 *)&rec->all_recs[i * GUID_REC_SIZE] =
+					sm_response;
+				if (required_val == 0)
+					mlx4_set_admin_guid(dev->dev,
+							    sm_response,
+							    (guid_rec->block_num
+							    * NUM_ALIAS_GUID_IN_REC) + i,
+							    cb_ctx->port);
+				goto next_entry;
 			}
 		}
+entry_declined:
+		declined_guid_indexes |= mlx4_ib_get_aguid_comp_mask_from_ix(i);
+		rec->guids_retry_schedule[i] =
+			(rec->guids_retry_schedule[i] == 0) ?  1 :
+			min((unsigned int)60,
+			    rec->guids_retry_schedule[i] * 2);
+		/* using the minimum value among all entries in that record */
+		resched_delay_sec = (resched_delay_sec == 0) ?
+				rec->guids_retry_schedule[i] :
+				min(resched_delay_sec,
+				    rec->guids_retry_schedule[i]);
+		continue;
+
+next_entry:
+		rec->guids_retry_schedule[i] = 0;
 	}
+
+	applied_guid_indexes =  cb_ctx->guid_indexes & ~declined_guid_indexes;
+	if (declined_guid_indexes ||
+	    rec->guid_indexes & ~(applied_guid_indexes)) {
+		pr_debug("record=%d wasn't fully set, guid_indexes=0x%llx applied_indexes=0x%llx, declined_indexes=0x%llx\n",
+			 guid_rec->block_num,
+			 be64_to_cpu((__force __be64)rec->guid_indexes),
+			 be64_to_cpu((__force __be64)applied_guid_indexes),
+			 be64_to_cpu((__force __be64)declined_guid_indexes));
+		rec->time_to_run = ktime_get_real_ns() +
+			resched_delay_sec * NSEC_PER_SEC;
+	} else {
+		rec->status = MLX4_GUID_INFO_STATUS_SET;
+	}
+	spin_unlock_irqrestore(&dev->sriov.alias_guid.ag_work_lock, flags);
 	/*
 	The func is call here to close the cases when the
 	sm doesn't send smp, so in the sa response the driver
@@ -297,10 +433,13 @@ static void aliasguid_query_handler(int status,
 out:
 	spin_lock_irqsave(&dev->sriov.going_down_lock, flags);
 	spin_lock_irqsave(&dev->sriov.alias_guid.ag_work_lock, flags1);
-	if (!dev->sriov.is_going_down)
+	if (!dev->sriov.is_going_down) {
+		get_low_record_time_index(dev, port_index, &resched_delay_sec);
 		queue_delayed_work(dev->sriov.alias_guid.ports_guid[port_index].wq,
 				   &dev->sriov.alias_guid.ports_guid[port_index].
-				   alias_guid_work, 0);
+				   alias_guid_work,
+				   msecs_to_jiffies(resched_delay_sec * 1000));
+	}
 	if (cb_ctx->sa_query) {
 		list_del(&cb_ctx->list);
 		kfree(cb_ctx);
@@ -317,9 +456,7 @@ static void invalidate_guid_record(struct mlx4_ib_dev *dev, u8 port, int index)
 	ib_sa_comp_mask comp_mask = 0;
 
 	dev->sriov.alias_guid.ports_guid[port - 1].all_rec_per_port[index].status
-		= MLX4_GUID_INFO_STATUS_IDLE;
-	dev->sriov.alias_guid.ports_guid[port - 1].all_rec_per_port[index].method
-		= MLX4_GUID_INFO_RECORD_SET;
+		= MLX4_GUID_INFO_STATUS_SET;
 
 	/* calculate the comp_mask for that record.*/
 	for (i = 0; i < NUM_ALIAS_GUID_IN_REC; i++) {
@@ -333,19 +470,21 @@ static void invalidate_guid_record(struct mlx4_ib_dev *dev, u8 port, int index)
 		need to assign GUIDs, then don't put it up for assignment.
 		*/
 		if (MLX4_GUID_FOR_DELETE_VAL == cur_admin_val ||
-		    (!index && !i) ||
-		    MLX4_GUID_NONE_ASSIGN == dev->sriov.alias_guid.
-		    ports_guid[port - 1].all_rec_per_port[index].ownership)
+		    (!index && !i))
 			continue;
 		comp_mask |= mlx4_ib_get_aguid_comp_mask_from_ix(i);
 	}
 	dev->sriov.alias_guid.ports_guid[port - 1].
-		all_rec_per_port[index].guid_indexes = comp_mask;
+		all_rec_per_port[index].guid_indexes |= comp_mask;
+	if (dev->sriov.alias_guid.ports_guid[port - 1].
+	    all_rec_per_port[index].guid_indexes)
+		dev->sriov.alias_guid.ports_guid[port - 1].
+		all_rec_per_port[index].status = MLX4_GUID_INFO_STATUS_IDLE;
+
 }
 
 static int set_guid_rec(struct ib_device *ibdev,
-			u8 port, int index,
-			struct mlx4_sriov_alias_guid_info_rec_det *rec_det)
+			struct mlx4_next_alias_guid_work *rec)
 {
 	int err;
 	struct mlx4_ib_dev *dev = to_mdev(ibdev);
@@ -354,6 +493,9 @@ static int set_guid_rec(struct ib_device *ibdev,
 	struct ib_port_attr attr;
 	struct mlx4_alias_guid_work_context *callback_context;
 	unsigned long resched_delay, flags, flags1;
+	u8 port = rec->port + 1;
+	int index = rec->block_num;
+	struct mlx4_sriov_alias_guid_info_rec_det *rec_det = &rec->rec_det;
 	struct list_head *head =
 		&dev->sriov.alias_guid.ports_guid[port - 1].cb_list;
 
@@ -380,6 +522,8 @@ static int set_guid_rec(struct ib_device *ibdev,
 	callback_context->port = port;
 	callback_context->dev = dev;
 	callback_context->block_num = index;
+	callback_context->guid_indexes = rec_det->guid_indexes;
+	callback_context->method = rec->method;
 
 	memset(&guid_info_rec, 0, sizeof (struct ib_sa_guidinfo_rec));
 
@@ -399,7 +543,7 @@ static int set_guid_rec(struct ib_device *ibdev,
 	callback_context->query_id =
 		ib_sa_guid_info_rec_query(dev->sriov.alias_guid.sa_client,
 					  ibdev, port, &guid_info_rec,
-					  comp_mask, rec_det->method, 1000,
+					  comp_mask, rec->method, 1000,
 					  GFP_KERNEL, aliasguid_query_handler,
 					  callback_context,
 					  &callback_context->sa_query);
@@ -434,6 +578,30 @@ out:
 	return err;
 }
 
+static void mlx4_ib_guid_port_init(struct mlx4_ib_dev *dev, int port)
+{
+	int j, k, entry;
+	__be64 guid;
+
+	/*Check if the SM doesn't need to assign the GUIDs*/
+	for (j = 0; j < NUM_ALIAS_GUID_REC_IN_PORT; j++) {
+		for (k = 0; k < NUM_ALIAS_GUID_IN_REC; k++) {
+			entry = j * NUM_ALIAS_GUID_IN_REC + k;
+			/* no request for the 0 entry (hw guid) */
+			if (!entry || entry > dev->dev->persist->num_vfs ||
+			    !mlx4_is_slave_active(dev->dev, entry))
+				continue;
+			guid = mlx4_get_admin_guid(dev->dev, entry, port);
+			*(__be64 *)&dev->sriov.alias_guid.ports_guid[port - 1].
+				all_rec_per_port[j].all_recs
+				[GUID_REC_SIZE * k] = guid;
+			pr_debug("guid was set, entry=%d, val=0x%llx, port=%d\n",
+				 entry,
+				 be64_to_cpu(guid),
+				 port);
+		}
+	}
+}
 void mlx4_ib_invalidate_all_guid_record(struct mlx4_ib_dev *dev, int port)
 {
 	int i;
@@ -443,6 +611,13 @@ void mlx4_ib_invalidate_all_guid_record(struct mlx4_ib_dev *dev, int port)
 
 	spin_lock_irqsave(&dev->sriov.going_down_lock, flags);
 	spin_lock_irqsave(&dev->sriov.alias_guid.ag_work_lock, flags1);
+
+	if (dev->sriov.alias_guid.ports_guid[port - 1].state_flags &
+		GUID_STATE_NEED_PORT_INIT) {
+		mlx4_ib_guid_port_init(dev, port);
+		dev->sriov.alias_guid.ports_guid[port - 1].state_flags &=
+			(~GUID_STATE_NEED_PORT_INIT);
+	}
 	for (i = 0; i < NUM_ALIAS_GUID_REC_IN_PORT; i++)
 		invalidate_guid_record(dev, port, i);
 
@@ -462,60 +637,107 @@ void mlx4_ib_invalidate_all_guid_record(struct mlx4_ib_dev *dev, int port)
 	spin_unlock_irqrestore(&dev->sriov.going_down_lock, flags);
 }
 
-/* The function returns the next record that was
- * not configured (or failed to be configured) */
-static int get_next_record_to_update(struct mlx4_ib_dev *dev, u8 port,
-				     struct mlx4_next_alias_guid_work *rec)
+static void set_required_record(struct mlx4_ib_dev *dev, u8 port,
+				struct mlx4_next_alias_guid_work *next_rec,
+				int record_index)
 {
-	int j;
-	unsigned long flags;
+	int i;
+	int lowset_time_entry = -1;
+	int lowest_time = 0;
+	ib_sa_comp_mask delete_guid_indexes = 0;
+	ib_sa_comp_mask set_guid_indexes = 0;
+	struct mlx4_sriov_alias_guid_info_rec_det *rec =
+			&dev->sriov.alias_guid.ports_guid[port].
+			all_rec_per_port[record_index];
 
-	for (j = 0; j < NUM_ALIAS_GUID_REC_IN_PORT; j++) {
-		spin_lock_irqsave(&dev->sriov.alias_guid.ag_work_lock, flags);
-		if (dev->sriov.alias_guid.ports_guid[port].all_rec_per_port[j].status ==
-		    MLX4_GUID_INFO_STATUS_IDLE) {
-			memcpy(&rec->rec_det,
-			       &dev->sriov.alias_guid.ports_guid[port].all_rec_per_port[j],
-			       sizeof (struct mlx4_sriov_alias_guid_info_rec_det));
-			rec->port = port;
-			rec->block_num = j;
-			dev->sriov.alias_guid.ports_guid[port].all_rec_per_port[j].status =
-				MLX4_GUID_INFO_STATUS_PENDING;
-			spin_unlock_irqrestore(&dev->sriov.alias_guid.ag_work_lock, flags);
-			return 0;
+	for (i = 0; i < NUM_ALIAS_GUID_IN_REC; i++) {
+		if (!(rec->guid_indexes &
+			mlx4_ib_get_aguid_comp_mask_from_ix(i)))
+			continue;
+
+		if (*(__be64 *)&rec->all_recs[i * GUID_REC_SIZE] ==
+				cpu_to_be64(MLX4_GUID_FOR_DELETE_VAL))
+			delete_guid_indexes |=
+				mlx4_ib_get_aguid_comp_mask_from_ix(i);
+		else
+			set_guid_indexes |=
+				mlx4_ib_get_aguid_comp_mask_from_ix(i);
+
+		if (lowset_time_entry == -1 || rec->guids_retry_schedule[i] <=
+			lowest_time) {
+			lowset_time_entry = i;
+			lowest_time = rec->guids_retry_schedule[i];
 		}
-		spin_unlock_irqrestore(&dev->sriov.alias_guid.ag_work_lock, flags);
 	}
-	return -ENOENT;
+
+	memcpy(&next_rec->rec_det, rec, sizeof(*rec));
+	next_rec->port = port;
+	next_rec->block_num = record_index;
+
+	if (*(__be64 *)&rec->all_recs[lowset_time_entry * GUID_REC_SIZE] ==
+				cpu_to_be64(MLX4_GUID_FOR_DELETE_VAL)) {
+		next_rec->rec_det.guid_indexes = delete_guid_indexes;
+		next_rec->method = MLX4_GUID_INFO_RECORD_DELETE;
+	} else {
+		next_rec->rec_det.guid_indexes = set_guid_indexes;
+		next_rec->method = MLX4_GUID_INFO_RECORD_SET;
+	}
 }
 
-static void set_administratively_guid_record(struct mlx4_ib_dev *dev, int port,
-					     int rec_index,
-					     struct mlx4_sriov_alias_guid_info_rec_det *rec_det)
+/* return index of record that should be updated based on lowest
+ * rescheduled time
+ */
+static int get_low_record_time_index(struct mlx4_ib_dev *dev, u8 port,
+				     int *resched_delay_sec)
 {
-	dev->sriov.alias_guid.ports_guid[port].all_rec_per_port[rec_index].guid_indexes =
-		rec_det->guid_indexes;
-	memcpy(dev->sriov.alias_guid.ports_guid[port].all_rec_per_port[rec_index].all_recs,
-	       rec_det->all_recs, NUM_ALIAS_GUID_IN_REC * GUID_REC_SIZE);
-	dev->sriov.alias_guid.ports_guid[port].all_rec_per_port[rec_index].status =
-		rec_det->status;
+	int record_index = -1;
+	u64 low_record_time = 0;
+	struct mlx4_sriov_alias_guid_info_rec_det rec;
+	int j;
+
+	for (j = 0; j < NUM_ALIAS_GUID_REC_IN_PORT; j++) {
+		rec = dev->sriov.alias_guid.ports_guid[port].
+			all_rec_per_port[j];
+		if (rec.status == MLX4_GUID_INFO_STATUS_IDLE &&
+		    rec.guid_indexes) {
+			if (record_index == -1 ||
+			    rec.time_to_run < low_record_time) {
+				record_index = j;
+				low_record_time = rec.time_to_run;
+			}
+		}
+	}
+	if (resched_delay_sec) {
+		u64 curr_time = ktime_get_real_ns();
+
+		*resched_delay_sec = (low_record_time < curr_time) ? 0 :
+			div_u64((low_record_time - curr_time), NSEC_PER_SEC);
+	}
+
+	return record_index;
 }
 
-static void set_all_slaves_guids(struct mlx4_ib_dev *dev, int port)
+/* The function returns the next record that was
+ * not configured (or failed to be configured) */
+static int get_next_record_to_update(struct mlx4_ib_dev *dev, u8 port,
+				     struct mlx4_next_alias_guid_work *rec)
 {
-	int j;
-	struct mlx4_sriov_alias_guid_info_rec_det rec_det ;
-
-	for (j = 0 ; j < NUM_ALIAS_GUID_REC_IN_PORT ; j++) {
-		memset(rec_det.all_recs, 0, NUM_ALIAS_GUID_IN_REC * GUID_REC_SIZE);
-		rec_det.guid_indexes = (!j ? 0 : IB_SA_GUIDINFO_REC_GID0) |
-			IB_SA_GUIDINFO_REC_GID1 | IB_SA_GUIDINFO_REC_GID2 |
-			IB_SA_GUIDINFO_REC_GID3 | IB_SA_GUIDINFO_REC_GID4 |
-			IB_SA_GUIDINFO_REC_GID5 | IB_SA_GUIDINFO_REC_GID6 |
-			IB_SA_GUIDINFO_REC_GID7;
-		rec_det.status = MLX4_GUID_INFO_STATUS_IDLE;
-		set_administratively_guid_record(dev, port, j, &rec_det);
+	unsigned long flags;
+	int record_index;
+	int ret = 0;
+
+	spin_lock_irqsave(&dev->sriov.alias_guid.ag_work_lock, flags);
+	record_index = get_low_record_time_index(dev, port, NULL);
+
+	if (record_index < 0) {
+		ret = -ENOENT;
+		goto out;
 	}
+
+	set_required_record(dev, port, rec, record_index);
+out:
+	spin_unlock_irqrestore(&dev->sriov.alias_guid.ag_work_lock, flags);
+	return ret;
 }
 
 static void alias_guid_work(struct work_struct *work)
@@ -545,9 +767,7 @@ static void alias_guid_work(struct work_struct *work)
 		goto out;
 	}
 
-	set_guid_rec(&dev->ib_dev, rec->port + 1, rec->block_num,
-		     &rec->rec_det);
-
+	set_guid_rec(&dev->ib_dev, rec);
 out:
 	kfree(rec);
 }
@@ -562,6 +782,12 @@ void mlx4_ib_init_alias_guid_work(struct mlx4_ib_dev *dev, int port)
 	spin_lock_irqsave(&dev->sriov.going_down_lock, flags);
 	spin_lock_irqsave(&dev->sriov.alias_guid.ag_work_lock, flags1);
 	if (!dev->sriov.is_going_down) {
+		/* If there is pending one should cancell then run, otherwise
+		  * won't run till previous one is ended as same work
+		  * struct is used.
+		  */
+		cancel_delayed_work(&dev->sriov.alias_guid.ports_guid[port].
+				    alias_guid_work);
 		queue_delayed_work(dev->sriov.alias_guid.ports_guid[port].wq,
 			   &dev->sriov.alias_guid.ports_guid[port].alias_guid_work, 0);
 	}
@@ -609,7 +835,7 @@ int mlx4_ib_init_alias_guid_service(struct mlx4_ib_dev *dev)
 {
 	char alias_wq_name[15];
 	int ret = 0;
-	int i, j, k;
+	int i, j;
 	union ib_gid gid;
 
 	if (!mlx4_is_master(dev->dev))
@@ -633,33 +859,25 @@ int mlx4_ib_init_alias_guid_service(struct mlx4_ib_dev *dev)
 	for (i = 0 ; i < dev->num_ports; i++) {
 		memset(&dev->sriov.alias_guid.ports_guid[i], 0,
 		       sizeof (struct mlx4_sriov_alias_guid_port_rec_det));
-		/*Check if the SM doesn't need to assign the GUIDs*/
+		dev->sriov.alias_guid.ports_guid[i].state_flags |=
+				GUID_STATE_NEED_PORT_INIT;
 		for (j = 0; j < NUM_ALIAS_GUID_REC_IN_PORT; j++) {
-			if (mlx4_ib_sm_guid_assign) {
-				dev->sriov.alias_guid.ports_guid[i].
-					all_rec_per_port[j].
-					ownership = MLX4_GUID_DRIVER_ASSIGN;
-				continue;
-			}
-			dev->sriov.alias_guid.ports_guid[i].all_rec_per_port[j].
-					ownership = MLX4_GUID_NONE_ASSIGN;
-			/*mark each val as it was deleted,
-			  till the sysAdmin will give it valid val*/
-			for (k = 0; k < NUM_ALIAS_GUID_IN_REC; k++) {
-				*(__be64 *)&dev->sriov.alias_guid.ports_guid[i].
-					all_rec_per_port[j].all_recs[GUID_REC_SIZE * k] =
-						cpu_to_be64(MLX4_GUID_FOR_DELETE_VAL);
-			}
+			/* mark each val as it was deleted */
+			memset(dev->sriov.alias_guid.ports_guid[i].
+				all_rec_per_port[j].all_recs, 0xFF,
+				sizeof(dev->sriov.alias_guid.ports_guid[i].
+				all_rec_per_port[j].all_recs));
 		}
 		INIT_LIST_HEAD(&dev->sriov.alias_guid.ports_guid[i].cb_list);
 		/*prepare the records, set them to be allocated by sm*/
+		if (mlx4_ib_sm_guid_assign)
+			for (j = 1; j < NUM_ALIAS_GUID_PER_PORT; j++)
+				mlx4_set_admin_guid(dev->dev, 0, j, i + 1);
 		for (j = 0 ; j < NUM_ALIAS_GUID_REC_IN_PORT; j++)
 			invalidate_guid_record(dev, i + 1, j);
 
 		dev->sriov.alias_guid.ports_guid[i].parent = &dev->sriov.alias_guid;
 		dev->sriov.alias_guid.ports_guid[i].port  = i;
-		if (mlx4_ib_sm_guid_assign)
-			set_all_slaves_guids(dev, i);
 
 		snprintf(alias_wq_name, sizeof alias_wq_name, "alias_guid%d", i);
 		dev->sriov.alias_guid.ports_guid[i].wq =
diff --git a/drivers/infiniband/hw/mlx4/mad.c b/drivers/infiniband/hw/mlx4/mad.c
index 59040265e361..ad6a8818608d 100644
--- a/drivers/infiniband/hw/mlx4/mad.c
+++ b/drivers/infiniband/hw/mlx4/mad.c
@@ -1365,14 +1365,17 @@ 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);
 	/* if slave have default vlan use it */
@@ -1430,6 +1433,10 @@ static int mlx4_ib_alloc_pv_bufs(struct mlx4_ib_demux_pv_ctx *ctx,
 							tun_qp->ring[i].addr,
 							rx_buf_size,
 							DMA_FROM_DEVICE);
+		if (ib_dma_mapping_error(ctx->ib_dev, tun_qp->ring[i].map)) {
+			kfree(tun_qp->ring[i].addr);
+			goto err;
+		}
 	}
 
 	for (i = 0; i < MLX4_NUM_TUNNEL_BUFS; i++) {
@@ -1442,6 +1449,11 @@ static int mlx4_ib_alloc_pv_bufs(struct mlx4_ib_demux_pv_ctx *ctx,
 					  tun_qp->tx_ring[i].buf.addr,
 					  tx_buf_size,
 					  DMA_TO_DEVICE);
+		if (ib_dma_mapping_error(ctx->ib_dev,
+					 tun_qp->tx_ring[i].buf.map)) {
+			kfree(tun_qp->tx_ring[i].buf.addr);
+			goto tx_err;
+		}
 		tun_qp->tx_ring[i].ah = NULL;
 	}
 	spin_lock_init(&tun_qp->tx_lock);
diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index 976bea794b5f..024b0f745035 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -66,9 +66,9 @@ MODULE_DESCRIPTION("Mellanox ConnectX HCA InfiniBand driver");
 MODULE_LICENSE("Dual BSD/GPL");
 MODULE_VERSION(DRV_VERSION);
 
-int mlx4_ib_sm_guid_assign = 1;
+int mlx4_ib_sm_guid_assign = 0;
 module_param_named(sm_guid_assign, mlx4_ib_sm_guid_assign, int, 0444);
-MODULE_PARM_DESC(sm_guid_assign, "Enable SM alias_GUID assignment if sm_guid_assign > 0 (Default: 1)");
+MODULE_PARM_DESC(sm_guid_assign, "Enable SM alias_GUID assignment if sm_guid_assign > 0 (Default: 0)");
 
 static const char mlx4_ib_version[] =
 	DRV_NAME ": Mellanox ConnectX InfiniBand driver v"
@@ -1090,7 +1090,7 @@ static int __mlx4_ib_create_flow(struct ib_qp *qp, struct ib_flow_attr *flow_att
 
 	ret = mlx4_cmd_imm(mdev->dev, mailbox->dma, reg_id, size >> 2, 0,
 			   MLX4_QP_FLOW_STEERING_ATTACH, MLX4_CMD_TIME_CLASS_A,
-			   MLX4_CMD_NATIVE);
+			   MLX4_CMD_WRAPPED);
 	if (ret == -ENOMEM)
 		pr_err("mcg table is full. Fail to register network rule.\n");
 	else if (ret == -ENXIO)
@@ -1107,7 +1107,7 @@ static int __mlx4_ib_destroy_flow(struct mlx4_dev *dev, u64 reg_id)
 	int err;
 	err = mlx4_cmd(dev, reg_id, 0, 0,
 		       MLX4_QP_FLOW_STEERING_DETACH, MLX4_CMD_TIME_CLASS_A,
-		       MLX4_CMD_NATIVE);
+		       MLX4_CMD_WRAPPED);
 	if (err)
 		pr_err("Fail to detach network rule. registration id = 0x%llx\n",
 		       reg_id);
@@ -1569,8 +1569,7 @@ static void reset_gids_task(struct work_struct *work)
 			       MLX4_CMD_TIME_CLASS_B,
 			       MLX4_CMD_WRAPPED);
 		if (err)
-			pr_warn(KERN_WARNING
-				"set port %d command failed\n", gw->port);
+			pr_warn("set port %d command failed\n", gw->port);
 	}
 
 	mlx4_free_cmd_mailbox(dev, mailbox);
@@ -2042,77 +2041,52 @@ static void init_pkeys(struct mlx4_ib_dev *ibdev)
 
 static void mlx4_ib_alloc_eqs(struct mlx4_dev *dev, struct mlx4_ib_dev *ibdev)
 {
-	char name[80];
-	int eq_per_port = 0;
-	int added_eqs = 0;
-	int total_eqs = 0;
-	int i, j, eq;
-
-	/* Legacy mode or comp_pool is not large enough */
-	if (dev->caps.comp_pool == 0 ||
-	    dev->caps.num_ports > dev->caps.comp_pool)
-		return;
-
-	eq_per_port = dev->caps.comp_pool / dev->caps.num_ports;
-
-	/* Init eq table */
-	added_eqs = 0;
-	mlx4_foreach_port(i, dev, MLX4_PORT_TYPE_IB)
-		added_eqs += eq_per_port;
-
-	total_eqs = dev->caps.num_comp_vectors + added_eqs;
+	int i, j, eq = 0, total_eqs = 0;
 
-	ibdev->eq_table = kzalloc(total_eqs * sizeof(int), GFP_KERNEL);
+	ibdev->eq_table = kcalloc(dev->caps.num_comp_vectors,
+				  sizeof(ibdev->eq_table[0]), GFP_KERNEL);
 	if (!ibdev->eq_table)
 		return;
 
-	ibdev->eq_added = added_eqs;
-
-	eq = 0;
-	mlx4_foreach_port(i, dev, MLX4_PORT_TYPE_IB) {
-		for (j = 0; j < eq_per_port; j++) {
-			snprintf(name, sizeof(name), "mlx4-ib-%d-%d@%s",
-				 i, j, dev->persist->pdev->bus->name);
-			/* Set IRQ for specific name (per ring) */
-			if (mlx4_assign_eq(dev, name, NULL,
-					   &ibdev->eq_table[eq])) {
-				/* Use legacy (same as mlx4_en driver) */
-				pr_warn("Can't allocate EQ %d; reverting to legacy\n", eq);
-				ibdev->eq_table[eq] =
-					(eq % dev->caps.num_comp_vectors);
-			}
-			eq++;
+	for (i = 1; i <= dev->caps.num_ports; i++) {
+		for (j = 0; j < mlx4_get_eqs_per_port(dev, i);
+		     j++, total_eqs++) {
+			if (i > 1 &&  mlx4_is_eq_shared(dev, total_eqs))
+				continue;
+			ibdev->eq_table[eq] = total_eqs;
+			if (!mlx4_assign_eq(dev, i,
+					    &ibdev->eq_table[eq]))
+				eq++;
+			else
+				ibdev->eq_table[eq] = -1;
 		}
 	}
 
-	/* Fill the reset of the vector with legacy EQ */
-	for (i = 0, eq = added_eqs; i < dev->caps.num_comp_vectors; i++)
-		ibdev->eq_table[eq++] = i;
+	for (i = eq; i < dev->caps.num_comp_vectors;
+	     ibdev->eq_table[i++] = -1)
+		;
 
 	/* Advertise the new number of EQs to clients */
-	ibdev->ib_dev.num_comp_vectors = total_eqs;
+	ibdev->ib_dev.num_comp_vectors = eq;
 }
 
 static void mlx4_ib_free_eqs(struct mlx4_dev *dev, struct mlx4_ib_dev *ibdev)
 {
 	int i;
+	int total_eqs = ibdev->ib_dev.num_comp_vectors;
 
-	/* no additional eqs were added */
+	/* no eqs were allocated */
 	if (!ibdev->eq_table)
 		return;
 
 	/* Reset the advertised EQ number */
-	ibdev->ib_dev.num_comp_vectors = dev->caps.num_comp_vectors;
+	ibdev->ib_dev.num_comp_vectors = 0;
 
-	/* Free only the added eqs */
-	for (i = 0; i < ibdev->eq_added; i++) {
-		/* Don't free legacy eqs if used */
-		if (ibdev->eq_table[i] <= dev->caps.num_comp_vectors)
-			continue;
+	for (i = 0; i < total_eqs; i++)
 		mlx4_release_eq(dev, ibdev->eq_table[i]);
-	}
 
 	kfree(ibdev->eq_table);
+	ibdev->eq_table = NULL;
 }
 
 static void *mlx4_ib_add(struct mlx4_dev *dev)
@@ -2791,9 +2765,31 @@ static void mlx4_ib_event(struct mlx4_dev *dev, void *ibdev_ptr,
 	case MLX4_DEV_EVENT_SLAVE_INIT:
 		/* here, p is the slave id */
 		do_slave_init(ibdev, p, 1);
+		if (mlx4_is_master(dev)) {
+			int i;
+
+			for (i = 1; i <= ibdev->num_ports; i++) {
+				if (rdma_port_get_link_layer(&ibdev->ib_dev, i)
+					== IB_LINK_LAYER_INFINIBAND)
+					mlx4_ib_slave_alias_guid_event(ibdev,
+								       p, i,
+								       1);
+			}
+		}
 		return;
 
 	case MLX4_DEV_EVENT_SLAVE_SHUTDOWN:
+		if (mlx4_is_master(dev)) {
+			int i;
+
+			for (i = 1; i <= ibdev->num_ports; i++) {
+				if (rdma_port_get_link_layer(&ibdev->ib_dev, i)
+					== IB_LINK_LAYER_INFINIBAND)
+					mlx4_ib_slave_alias_guid_event(ibdev,
+								       p, i,
+								       0);
+			}
+		}
 		/* here, p is the slave id */
 		do_slave_init(ibdev, p, 0);
 		return;
diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h
index f829fd935b79..ef80e6c99a68 100644
--- a/drivers/infiniband/hw/mlx4/mlx4_ib.h
+++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h
@@ -342,14 +342,9 @@ struct mlx4_ib_ah {
 enum mlx4_guid_alias_rec_status {
 	MLX4_GUID_INFO_STATUS_IDLE,
 	MLX4_GUID_INFO_STATUS_SET,
-	MLX4_GUID_INFO_STATUS_PENDING,
 };
 
-enum mlx4_guid_alias_rec_ownership {
-	MLX4_GUID_DRIVER_ASSIGN,
-	MLX4_GUID_SYSADMIN_ASSIGN,
-	MLX4_GUID_NONE_ASSIGN, /*init state of each record*/
-};
+#define GUID_STATE_NEED_PORT_INIT 0x01
 
 enum mlx4_guid_alias_rec_method {
 	MLX4_GUID_INFO_RECORD_SET	= IB_MGMT_METHOD_SET,
@@ -360,8 +355,8 @@ struct mlx4_sriov_alias_guid_info_rec_det {
 	u8 all_recs[GUID_REC_SIZE * NUM_ALIAS_GUID_IN_REC];
 	ib_sa_comp_mask guid_indexes; /*indicates what from the 8 records are valid*/
 	enum mlx4_guid_alias_rec_status status; /*indicates the administraively status of the record.*/
-	u8 method; /*set or delete*/
-	enum mlx4_guid_alias_rec_ownership ownership; /*indicates who assign that alias_guid record*/
+	unsigned int guids_retry_schedule[NUM_ALIAS_GUID_IN_REC];
+	u64 time_to_run;
 };
 
 struct mlx4_sriov_alias_guid_port_rec_det {
@@ -369,6 +364,7 @@ struct mlx4_sriov_alias_guid_port_rec_det {
 	struct workqueue_struct *wq;
 	struct delayed_work alias_guid_work;
 	u8 port;
+	u32 state_flags;
 	struct mlx4_sriov_alias_guid *parent;
 	struct list_head cb_list;
 };
@@ -527,7 +523,6 @@ struct mlx4_ib_dev {
 	struct mlx4_ib_iboe	iboe;
 	int			counters[MLX4_MAX_PORTS];
 	int		       *eq_table;
-	int			eq_added;
 	struct kobject	       *iov_parent;
 	struct kobject	       *ports_parent;
 	struct kobject	       *dev_ports_parent[MLX4_MFUNC_MAX];
@@ -802,6 +797,8 @@ int add_sysfs_port_mcg_attr(struct mlx4_ib_dev *device, int port_num,
 void del_sysfs_port_mcg_attr(struct mlx4_ib_dev *device, int port_num,
 			     struct attribute *attr);
 ib_sa_comp_mask mlx4_ib_get_aguid_comp_mask_from_ix(int index);
+void mlx4_ib_slave_alias_guid_event(struct mlx4_ib_dev *dev, int slave,
+				    int port, int slave_init);
 
 int mlx4_ib_device_register_sysfs(struct mlx4_ib_dev *device) ;
 
diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index ed2bd6701f9b..02fc91c68027 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -566,6 +566,10 @@ static int alloc_proxy_bufs(struct ib_device *dev, struct mlx4_ib_qp *qp)
 			ib_dma_map_single(dev, qp->sqp_proxy_rcv[i].addr,
 					  sizeof (struct mlx4_ib_proxy_sqp_hdr),
 					  DMA_FROM_DEVICE);
+		if (ib_dma_mapping_error(dev, qp->sqp_proxy_rcv[i].map)) {
+			kfree(qp->sqp_proxy_rcv[i].addr);
+			goto err;
+		}
 	}
 	return 0;
 
@@ -2605,8 +2609,7 @@ static int build_lso_seg(struct mlx4_wqe_lso_seg *wqe, struct ib_send_wr *wr,
 
 	memcpy(wqe->header, wr->wr.ud.header, wr->wr.ud.hlen);
 
-	*lso_hdr_sz  = cpu_to_be32((wr->wr.ud.mss - wr->wr.ud.hlen) << 16 |
-				   wr->wr.ud.hlen);
+	*lso_hdr_sz  = cpu_to_be32(wr->wr.ud.mss << 16 | wr->wr.ud.hlen);
 	*lso_seg_len = halign;
 	return 0;
 }
diff --git a/drivers/infiniband/hw/mlx4/sysfs.c b/drivers/infiniband/hw/mlx4/sysfs.c
index d10c2b8a5dad..6797108ce873 100644
--- a/drivers/infiniband/hw/mlx4/sysfs.c
+++ b/drivers/infiniband/hw/mlx4/sysfs.c
@@ -46,21 +46,17 @@
 static ssize_t show_admin_alias_guid(struct device *dev,
 			      struct device_attribute *attr, char *buf)
 {
-	int record_num;/*0-15*/
-	int guid_index_in_rec; /*0 - 7*/
 	struct mlx4_ib_iov_sysfs_attr *mlx4_ib_iov_dentry =
 		container_of(attr, struct mlx4_ib_iov_sysfs_attr, dentry);
 	struct mlx4_ib_iov_port *port = mlx4_ib_iov_dentry->ctx;
 	struct mlx4_ib_dev *mdev = port->dev;
+	__be64 sysadmin_ag_val;
 
-	record_num = mlx4_ib_iov_dentry->entry_num / 8 ;
-	guid_index_in_rec = mlx4_ib_iov_dentry->entry_num % 8 ;
+	sysadmin_ag_val = mlx4_get_admin_guid(mdev->dev,
+					      mlx4_ib_iov_dentry->entry_num,
+					      port->num);
 
-	return sprintf(buf, "%llx\n",
-		       be64_to_cpu(*(__be64 *)&mdev->sriov.alias_guid.
-				   ports_guid[port->num - 1].
-				   all_rec_per_port[record_num].
-				   all_recs[8 * guid_index_in_rec]));
+	return sprintf(buf, "%llx\n", be64_to_cpu(sysadmin_ag_val));
 }
 
 /* store_admin_alias_guid stores the (new) administratively assigned value of that GUID.
@@ -80,6 +76,7 @@ static ssize_t store_admin_alias_guid(struct device *dev,
 	struct mlx4_ib_iov_port *port = mlx4_ib_iov_dentry->ctx;
 	struct mlx4_ib_dev *mdev = port->dev;
 	u64 sysadmin_ag_val;
+	unsigned long flags;
 
 	record_num = mlx4_ib_iov_dentry->entry_num / 8;
 	guid_index_in_rec = mlx4_ib_iov_dentry->entry_num % 8;
@@ -87,6 +84,7 @@ static ssize_t store_admin_alias_guid(struct device *dev,
 		pr_err("GUID 0 block 0 is RO\n");
 		return count;
 	}
+	spin_lock_irqsave(&mdev->sriov.alias_guid.ag_work_lock, flags);
 	sscanf(buf, "%llx", &sysadmin_ag_val);
 	*(__be64 *)&mdev->sriov.alias_guid.ports_guid[port->num - 1].
 		all_rec_per_port[record_num].
@@ -96,33 +94,15 @@ static ssize_t store_admin_alias_guid(struct device *dev,
 	/* Change the state to be pending for update */
 	mdev->sriov.alias_guid.ports_guid[port->num - 1].all_rec_per_port[record_num].status
 		= MLX4_GUID_INFO_STATUS_IDLE ;
-
-	mdev->sriov.alias_guid.ports_guid[port->num - 1].all_rec_per_port[record_num].method
-		= MLX4_GUID_INFO_RECORD_SET;
-
-	switch (sysadmin_ag_val) {
-	case MLX4_GUID_FOR_DELETE_VAL:
-		mdev->sriov.alias_guid.ports_guid[port->num - 1].all_rec_per_port[record_num].method
-			= MLX4_GUID_INFO_RECORD_DELETE;
-		mdev->sriov.alias_guid.ports_guid[port->num - 1].all_rec_per_port[record_num].ownership
-			= MLX4_GUID_SYSADMIN_ASSIGN;
-		break;
-	/* The sysadmin requests the SM to re-assign */
-	case MLX4_NOT_SET_GUID:
-		mdev->sriov.alias_guid.ports_guid[port->num - 1].all_rec_per_port[record_num].ownership
-			= MLX4_GUID_DRIVER_ASSIGN;
-		break;
-	/* The sysadmin requests a specific value.*/
-	default:
-		mdev->sriov.alias_guid.ports_guid[port->num - 1].all_rec_per_port[record_num].ownership
-			= MLX4_GUID_SYSADMIN_ASSIGN;
-		break;
-	}
+	mlx4_set_admin_guid(mdev->dev, cpu_to_be64(sysadmin_ag_val),
+			    mlx4_ib_iov_dentry->entry_num,
+			    port->num);
 
 	/* set the record index */
 	mdev->sriov.alias_guid.ports_guid[port->num - 1].all_rec_per_port[record_num].guid_indexes
-		= mlx4_ib_get_aguid_comp_mask_from_ix(guid_index_in_rec);
+		|= mlx4_ib_get_aguid_comp_mask_from_ix(guid_index_in_rec);
 
+	spin_unlock_irqrestore(&mdev->sriov.alias_guid.ag_work_lock, flags);
 	mlx4_ib_init_alias_guid_work(mdev, port->num - 1);
 
 	return count;
diff --git a/drivers/infiniband/hw/mlx5/Kconfig b/drivers/infiniband/hw/mlx5/Kconfig
index 10df386c6344..bce263b92821 100644
--- a/drivers/infiniband/hw/mlx5/Kconfig
+++ b/drivers/infiniband/hw/mlx5/Kconfig
@@ -1,8 +1,6 @@
 config MLX5_INFINIBAND
 	tristate "Mellanox Connect-IB HCA support"
-	depends on NETDEVICES && ETHERNET && PCI
-	select NET_VENDOR_MELLANOX
-	select MLX5_CORE
+	depends on NETDEVICES && ETHERNET && PCI && MLX5_CORE
 	---help---
 	  This driver provides low-level InfiniBand support for
 	  Mellanox Connect-IB PCI Express host channel adapters (HCAs).
diff --git a/drivers/infiniband/hw/mlx5/cq.c b/drivers/infiniband/hw/mlx5/cq.c
index 2ee6b1051975..e2bea9ab93b3 100644
--- a/drivers/infiniband/hw/mlx5/cq.c
+++ b/drivers/infiniband/hw/mlx5/cq.c
@@ -590,8 +590,7 @@ static int alloc_cq_buf(struct mlx5_ib_dev *dev, struct mlx5_ib_cq_buf *buf,
 {
 	int err;
 
-	err = mlx5_buf_alloc(dev->mdev, nent * cqe_size,
-			     PAGE_SIZE * 2, &buf->buf);
+	err = mlx5_buf_alloc(dev->mdev, nent * cqe_size, &buf->buf);
 	if (err)
 		return err;
 
@@ -754,7 +753,7 @@ struct ib_cq *mlx5_ib_create_cq(struct ib_device *ibdev, int entries,
 		return ERR_PTR(-EINVAL);
 
 	entries = roundup_pow_of_two(entries + 1);
-	if (entries > dev->mdev->caps.gen.max_cqes)
+	if (entries > (1 << MLX5_CAP_GEN(dev->mdev, log_max_cq_sz)))
 		return ERR_PTR(-EINVAL);
 
 	cq = kzalloc(sizeof(*cq), GFP_KERNEL);
@@ -921,7 +920,7 @@ int mlx5_ib_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period)
 	int err;
 	u32 fsel;
 
-	if (!(dev->mdev->caps.gen.flags & MLX5_DEV_CAP_FLAG_CQ_MODER))
+	if (!MLX5_CAP_GEN(dev->mdev, cq_moderation))
 		return -ENOSYS;
 
 	in = kzalloc(sizeof(*in), GFP_KERNEL);
@@ -1076,7 +1075,7 @@ int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata)
 	int uninitialized_var(cqe_size);
 	unsigned long flags;
 
-	if (!(dev->mdev->caps.gen.flags & MLX5_DEV_CAP_FLAG_RESIZE_CQ)) {
+	if (!MLX5_CAP_GEN(dev->mdev, cq_resize)) {
 		pr_info("Firmware does not support resize CQ\n");
 		return -ENOSYS;
 	}
@@ -1085,7 +1084,7 @@ int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata)
 		return -EINVAL;
 
 	entries = roundup_pow_of_two(entries + 1);
-	if (entries > dev->mdev->caps.gen.max_cqes + 1)
+	if (entries >  (1 << MLX5_CAP_GEN(dev->mdev, log_max_cq_sz)) + 1)
 		return -EINVAL;
 
 	if (entries == ibcq->cqe + 1)
diff --git a/drivers/infiniband/hw/mlx5/mad.c b/drivers/infiniband/hw/mlx5/mad.c
index 9cf9a37bb5ff..a770490ebbf1 100644
--- a/drivers/infiniband/hw/mlx5/mad.c
+++ b/drivers/infiniband/hw/mlx5/mad.c
@@ -129,7 +129,7 @@ int mlx5_query_ext_port_caps(struct mlx5_ib_dev *dev, u8 port)
 
 	packet_error = be16_to_cpu(out_mad->status);
 
-	dev->mdev->caps.gen.ext_port_cap[port - 1] = (!err && !packet_error) ?
+	dev->mdev->port_caps[port - 1].ext_port_cap = (!err && !packet_error) ?
 		MLX_EXT_PORT_CAP_FLAG_EXTENDED_PORT_INFO : 0;
 
 out:
@@ -137,3 +137,300 @@ out:
 	kfree(out_mad);
 	return err;
 }
+
+int mlx5_query_mad_ifc_smp_attr_node_info(struct ib_device *ibdev,
+					  struct ib_smp *out_mad)
+{
+	struct ib_smp *in_mad = NULL;
+	int err = -ENOMEM;
+
+	in_mad = kzalloc(sizeof(*in_mad), GFP_KERNEL);
+	if (!in_mad)
+		return -ENOMEM;
+
+	init_query_mad(in_mad);
+	in_mad->attr_id = IB_SMP_ATTR_NODE_INFO;
+
+	err = mlx5_MAD_IFC(to_mdev(ibdev), 1, 1, 1, NULL, NULL, in_mad,
+			   out_mad);
+
+	kfree(in_mad);
+	return err;
+}
+
+int mlx5_query_mad_ifc_system_image_guid(struct ib_device *ibdev,
+					 __be64 *sys_image_guid)
+{
+	struct ib_smp *out_mad = NULL;
+	int err = -ENOMEM;
+
+	out_mad = kmalloc(sizeof(*out_mad), GFP_KERNEL);
+	if (!out_mad)
+		return -ENOMEM;
+
+	err = mlx5_query_mad_ifc_smp_attr_node_info(ibdev, out_mad);
+	if (err)
+		goto out;
+
+	memcpy(sys_image_guid, out_mad->data + 4, 8);
+
+out:
+	kfree(out_mad);
+
+	return err;
+}
+
+int mlx5_query_mad_ifc_max_pkeys(struct ib_device *ibdev,
+				 u16 *max_pkeys)
+{
+	struct ib_smp *out_mad = NULL;
+	int err = -ENOMEM;
+
+	out_mad = kmalloc(sizeof(*out_mad), GFP_KERNEL);
+	if (!out_mad)
+		return -ENOMEM;
+
+	err = mlx5_query_mad_ifc_smp_attr_node_info(ibdev, out_mad);
+	if (err)
+		goto out;
+
+	*max_pkeys = be16_to_cpup((__be16 *)(out_mad->data + 28));
+
+out:
+	kfree(out_mad);
+
+	return err;
+}
+
+int mlx5_query_mad_ifc_vendor_id(struct ib_device *ibdev,
+				 u32 *vendor_id)
+{
+	struct ib_smp *out_mad = NULL;
+	int err = -ENOMEM;
+
+	out_mad = kmalloc(sizeof(*out_mad), GFP_KERNEL);
+	if (!out_mad)
+		return -ENOMEM;
+
+	err = mlx5_query_mad_ifc_smp_attr_node_info(ibdev, out_mad);
+	if (err)
+		goto out;
+
+	*vendor_id = be32_to_cpup((__be32 *)(out_mad->data + 36)) & 0xffff;
+
+out:
+	kfree(out_mad);
+
+	return err;
+}
+
+int mlx5_query_mad_ifc_node_desc(struct mlx5_ib_dev *dev, char *node_desc)
+{
+	struct ib_smp *in_mad  = NULL;
+	struct ib_smp *out_mad = NULL;
+	int err = -ENOMEM;
+
+	in_mad  = kzalloc(sizeof(*in_mad), GFP_KERNEL);
+	out_mad = kmalloc(sizeof(*out_mad), GFP_KERNEL);
+	if (!in_mad || !out_mad)
+		goto out;
+
+	init_query_mad(in_mad);
+	in_mad->attr_id = IB_SMP_ATTR_NODE_DESC;
+
+	err = mlx5_MAD_IFC(dev, 1, 1, 1, NULL, NULL, in_mad, out_mad);
+	if (err)
+		goto out;
+
+	memcpy(node_desc, out_mad->data, 64);
+out:
+	kfree(in_mad);
+	kfree(out_mad);
+	return err;
+}
+
+int mlx5_query_mad_ifc_node_guid(struct mlx5_ib_dev *dev, __be64 *node_guid)
+{
+	struct ib_smp *in_mad  = NULL;
+	struct ib_smp *out_mad = NULL;
+	int err = -ENOMEM;
+
+	in_mad  = kzalloc(sizeof(*in_mad), GFP_KERNEL);
+	out_mad = kmalloc(sizeof(*out_mad), GFP_KERNEL);
+	if (!in_mad || !out_mad)
+		goto out;
+
+	init_query_mad(in_mad);
+	in_mad->attr_id = IB_SMP_ATTR_NODE_INFO;
+
+	err = mlx5_MAD_IFC(dev, 1, 1, 1, NULL, NULL, in_mad, out_mad);
+	if (err)
+		goto out;
+
+	memcpy(node_guid, out_mad->data + 12, 8);
+out:
+	kfree(in_mad);
+	kfree(out_mad);
+	return err;
+}
+
+int mlx5_query_mad_ifc_pkey(struct ib_device *ibdev, u8 port, u16 index,
+			    u16 *pkey)
+{
+	struct ib_smp *in_mad  = NULL;
+	struct ib_smp *out_mad = NULL;
+	int err = -ENOMEM;
+
+	in_mad  = kzalloc(sizeof(*in_mad), GFP_KERNEL);
+	out_mad = kmalloc(sizeof(*out_mad), GFP_KERNEL);
+	if (!in_mad || !out_mad)
+		goto out;
+
+	init_query_mad(in_mad);
+	in_mad->attr_id  = IB_SMP_ATTR_PKEY_TABLE;
+	in_mad->attr_mod = cpu_to_be32(index / 32);
+
+	err = mlx5_MAD_IFC(to_mdev(ibdev), 1, 1, port, NULL, NULL, in_mad,
+			   out_mad);
+	if (err)
+		goto out;
+
+	*pkey = be16_to_cpu(((__be16 *)out_mad->data)[index % 32]);
+
+out:
+	kfree(in_mad);
+	kfree(out_mad);
+	return err;
+}
+
+int mlx5_query_mad_ifc_gids(struct ib_device *ibdev, u8 port, int index,
+			    union ib_gid *gid)
+{
+	struct ib_smp *in_mad  = NULL;
+	struct ib_smp *out_mad = NULL;
+	int err = -ENOMEM;
+
+	in_mad  = kzalloc(sizeof(*in_mad), GFP_KERNEL);
+	out_mad = kmalloc(sizeof(*out_mad), GFP_KERNEL);
+	if (!in_mad || !out_mad)
+		goto out;
+
+	init_query_mad(in_mad);
+	in_mad->attr_id  = IB_SMP_ATTR_PORT_INFO;
+	in_mad->attr_mod = cpu_to_be32(port);
+
+	err = mlx5_MAD_IFC(to_mdev(ibdev), 1, 1, port, NULL, NULL, in_mad,
+			   out_mad);
+	if (err)
+		goto out;
+
+	memcpy(gid->raw, out_mad->data + 8, 8);
+
+	init_query_mad(in_mad);
+	in_mad->attr_id  = IB_SMP_ATTR_GUID_INFO;
+	in_mad->attr_mod = cpu_to_be32(index / 8);
+
+	err = mlx5_MAD_IFC(to_mdev(ibdev), 1, 1, port, NULL, NULL, in_mad,
+			   out_mad);
+	if (err)
+		goto out;
+
+	memcpy(gid->raw + 8, out_mad->data + (index % 8) * 8, 8);
+
+out:
+	kfree(in_mad);
+	kfree(out_mad);
+	return err;
+}
+
+int mlx5_query_mad_ifc_port(struct ib_device *ibdev, u8 port,
+			    struct ib_port_attr *props)
+{
+	struct mlx5_ib_dev *dev = to_mdev(ibdev);
+	struct mlx5_core_dev *mdev = dev->mdev;
+	struct ib_smp *in_mad  = NULL;
+	struct ib_smp *out_mad = NULL;
+	int ext_active_speed;
+	int err = -ENOMEM;
+
+	if (port < 1 || port > MLX5_CAP_GEN(mdev, num_ports)) {
+		mlx5_ib_warn(dev, "invalid port number %d\n", port);
+		return -EINVAL;
+	}
+
+	in_mad  = kzalloc(sizeof(*in_mad), GFP_KERNEL);
+	out_mad = kmalloc(sizeof(*out_mad), GFP_KERNEL);
+	if (!in_mad || !out_mad)
+		goto out;
+
+	memset(props, 0, sizeof(*props));
+
+	init_query_mad(in_mad);
+	in_mad->attr_id  = IB_SMP_ATTR_PORT_INFO;
+	in_mad->attr_mod = cpu_to_be32(port);
+
+	err = mlx5_MAD_IFC(dev, 1, 1, port, NULL, NULL, in_mad, out_mad);
+	if (err) {
+		mlx5_ib_warn(dev, "err %d\n", err);
+		goto out;
+	}
+
+	props->lid		= be16_to_cpup((__be16 *)(out_mad->data + 16));
+	props->lmc		= out_mad->data[34] & 0x7;
+	props->sm_lid		= be16_to_cpup((__be16 *)(out_mad->data + 18));
+	props->sm_sl		= out_mad->data[36] & 0xf;
+	props->state		= out_mad->data[32] & 0xf;
+	props->phys_state	= out_mad->data[33] >> 4;
+	props->port_cap_flags	= be32_to_cpup((__be32 *)(out_mad->data + 20));
+	props->gid_tbl_len	= out_mad->data[50];
+	props->max_msg_sz	= 1 << MLX5_CAP_GEN(mdev, log_max_msg);
+	props->pkey_tbl_len	= mdev->port_caps[port - 1].pkey_table_len;
+	props->bad_pkey_cntr	= be16_to_cpup((__be16 *)(out_mad->data + 46));
+	props->qkey_viol_cntr	= be16_to_cpup((__be16 *)(out_mad->data + 48));
+	props->active_width	= out_mad->data[31] & 0xf;
+	props->active_speed	= out_mad->data[35] >> 4;
+	props->max_mtu		= out_mad->data[41] & 0xf;
+	props->active_mtu	= out_mad->data[36] >> 4;
+	props->subnet_timeout	= out_mad->data[51] & 0x1f;
+	props->max_vl_num	= out_mad->data[37] >> 4;
+	props->init_type_reply	= out_mad->data[41] >> 4;
+
+	/* Check if extended speeds (EDR/FDR/...) are supported */
+	if (props->port_cap_flags & IB_PORT_EXTENDED_SPEEDS_SUP) {
+		ext_active_speed = out_mad->data[62] >> 4;
+
+		switch (ext_active_speed) {
+		case 1:
+			props->active_speed = 16; /* FDR */
+			break;
+		case 2:
+			props->active_speed = 32; /* EDR */
+			break;
+		}
+	}
+
+	/* If reported active speed is QDR, check if is FDR-10 */
+	if (props->active_speed == 4) {
+		if (mdev->port_caps[port - 1].ext_port_cap &
+		    MLX_EXT_PORT_CAP_FLAG_EXTENDED_PORT_INFO) {
+			init_query_mad(in_mad);
+			in_mad->attr_id = MLX5_ATTR_EXTENDED_PORT_INFO;
+			in_mad->attr_mod = cpu_to_be32(port);
+
+			err = mlx5_MAD_IFC(dev, 1, 1, port,
+					   NULL, NULL, in_mad, out_mad);
+			if (err)
+				goto out;
+
+			/* Checking LinkSpeedActive for FDR-10 */
+			if (out_mad->data[15] & 0x1)
+				props->active_speed = 8;
+		}
+	}
+
+out:
+	kfree(in_mad);
+	kfree(out_mad);
+
+	return err;
+}
diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 57c9809e8b87..d4dea86052d6 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -40,6 +40,7 @@
 #include <linux/io-mapping.h>
 #include <linux/sched.h>
 #include <rdma/ib_user_verbs.h>
+#include <linux/mlx5/vport.h>
 #include <rdma/ib_smi.h>
 #include <rdma/ib_umem.h>
 #include "user.h"
@@ -62,32 +63,168 @@ static char mlx5_version[] =
 	DRIVER_NAME ": Mellanox Connect-IB Infiniband driver v"
 	DRIVER_VERSION " (" DRIVER_RELDATE ")\n";
 
+static enum rdma_link_layer
+mlx5_ib_port_link_layer(struct ib_device *device)
+{
+	struct mlx5_ib_dev *dev = to_mdev(device);
+
+	switch (MLX5_CAP_GEN(dev->mdev, port_type)) {
+	case MLX5_CAP_PORT_TYPE_IB:
+		return IB_LINK_LAYER_INFINIBAND;
+	case MLX5_CAP_PORT_TYPE_ETH:
+		return IB_LINK_LAYER_ETHERNET;
+	default:
+		return IB_LINK_LAYER_UNSPECIFIED;
+	}
+}
+
+static int mlx5_use_mad_ifc(struct mlx5_ib_dev *dev)
+{
+	return !dev->mdev->issi;
+}
+
+enum {
+	MLX5_VPORT_ACCESS_METHOD_MAD,
+	MLX5_VPORT_ACCESS_METHOD_HCA,
+	MLX5_VPORT_ACCESS_METHOD_NIC,
+};
+
+static int mlx5_get_vport_access_method(struct ib_device *ibdev)
+{
+	if (mlx5_use_mad_ifc(to_mdev(ibdev)))
+		return MLX5_VPORT_ACCESS_METHOD_MAD;
+
+	if (mlx5_ib_port_link_layer(ibdev) ==
+	    IB_LINK_LAYER_ETHERNET)
+		return MLX5_VPORT_ACCESS_METHOD_NIC;
+
+	return MLX5_VPORT_ACCESS_METHOD_HCA;
+}
+
+static int mlx5_query_system_image_guid(struct ib_device *ibdev,
+					__be64 *sys_image_guid)
+{
+	struct mlx5_ib_dev *dev = to_mdev(ibdev);
+	struct mlx5_core_dev *mdev = dev->mdev;
+	u64 tmp;
+	int err;
+
+	switch (mlx5_get_vport_access_method(ibdev)) {
+	case MLX5_VPORT_ACCESS_METHOD_MAD:
+		return mlx5_query_mad_ifc_system_image_guid(ibdev,
+							    sys_image_guid);
+
+	case MLX5_VPORT_ACCESS_METHOD_HCA:
+		err = mlx5_query_hca_vport_system_image_guid(mdev, &tmp);
+		if (!err)
+			*sys_image_guid = cpu_to_be64(tmp);
+		return err;
+
+	default:
+		return -EINVAL;
+	}
+}
+
+static int mlx5_query_max_pkeys(struct ib_device *ibdev,
+				u16 *max_pkeys)
+{
+	struct mlx5_ib_dev *dev = to_mdev(ibdev);
+	struct mlx5_core_dev *mdev = dev->mdev;
+
+	switch (mlx5_get_vport_access_method(ibdev)) {
+	case MLX5_VPORT_ACCESS_METHOD_MAD:
+		return mlx5_query_mad_ifc_max_pkeys(ibdev, max_pkeys);
+
+	case MLX5_VPORT_ACCESS_METHOD_HCA:
+	case MLX5_VPORT_ACCESS_METHOD_NIC:
+		*max_pkeys = mlx5_to_sw_pkey_sz(MLX5_CAP_GEN(mdev,
+						pkey_table_size));
+		return 0;
+
+	default:
+		return -EINVAL;
+	}
+}
+
+static int mlx5_query_vendor_id(struct ib_device *ibdev,
+				u32 *vendor_id)
+{
+	struct mlx5_ib_dev *dev = to_mdev(ibdev);
+
+	switch (mlx5_get_vport_access_method(ibdev)) {
+	case MLX5_VPORT_ACCESS_METHOD_MAD:
+		return mlx5_query_mad_ifc_vendor_id(ibdev, vendor_id);
+
+	case MLX5_VPORT_ACCESS_METHOD_HCA:
+	case MLX5_VPORT_ACCESS_METHOD_NIC:
+		return mlx5_core_query_vendor_id(dev->mdev, vendor_id);
+
+	default:
+		return -EINVAL;
+	}
+}
+
+static int mlx5_query_node_guid(struct mlx5_ib_dev *dev,
+				__be64 *node_guid)
+{
+	u64 tmp;
+	int err;
+
+	switch (mlx5_get_vport_access_method(&dev->ib_dev)) {
+	case MLX5_VPORT_ACCESS_METHOD_MAD:
+		return mlx5_query_mad_ifc_node_guid(dev, node_guid);
+
+	case MLX5_VPORT_ACCESS_METHOD_HCA:
+		err = mlx5_query_hca_vport_node_guid(dev->mdev, &tmp);
+		if (!err)
+			*node_guid = cpu_to_be64(tmp);
+		return err;
+
+	default:
+		return -EINVAL;
+	}
+}
+
+struct mlx5_reg_node_desc {
+	u8	desc[64];
+};
+
+static int mlx5_query_node_desc(struct mlx5_ib_dev *dev, char *node_desc)
+{
+	struct mlx5_reg_node_desc in;
+
+	if (mlx5_use_mad_ifc(dev))
+		return mlx5_query_mad_ifc_node_desc(dev, node_desc);
+
+	memset(&in, 0, sizeof(in));
+
+	return mlx5_core_access_reg(dev->mdev, &in, sizeof(in), node_desc,
+				    sizeof(struct mlx5_reg_node_desc),
+				    MLX5_REG_NODE_DESC, 0, 0);
+}
+
 static int mlx5_ib_query_device(struct ib_device *ibdev,
 				struct ib_device_attr *props)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
-	struct ib_smp *in_mad  = NULL;
-	struct ib_smp *out_mad = NULL;
-	struct mlx5_general_caps *gen;
+	struct mlx5_core_dev *mdev = dev->mdev;
 	int err = -ENOMEM;
 	int max_rq_sg;
 	int max_sq_sg;
-	u64 flags;
 
-	gen = &dev->mdev->caps.gen;
-	in_mad  = kzalloc(sizeof(*in_mad), GFP_KERNEL);
-	out_mad = kmalloc(sizeof(*out_mad), GFP_KERNEL);
-	if (!in_mad || !out_mad)
-		goto out;
-
-	init_query_mad(in_mad);
-	in_mad->attr_id = IB_SMP_ATTR_NODE_INFO;
+	memset(props, 0, sizeof(*props));
+	err = mlx5_query_system_image_guid(ibdev,
+					   &props->sys_image_guid);
+	if (err)
+		return err;
 
-	err = mlx5_MAD_IFC(to_mdev(ibdev), 1, 1, 1, NULL, NULL, in_mad, out_mad);
+	err = mlx5_query_max_pkeys(ibdev, &props->max_pkeys);
 	if (err)
-		goto out;
+		return err;
 
-	memset(props, 0, sizeof(*props));
+	err = mlx5_query_vendor_id(ibdev, &props->vendor_id);
+	if (err)
+		return err;
 
 	props->fw_ver = ((u64)fw_rev_maj(dev->mdev) << 32) |
 		(fw_rev_min(dev->mdev) << 16) |
@@ -96,18 +233,18 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
 		IB_DEVICE_PORT_ACTIVE_EVENT		|
 		IB_DEVICE_SYS_IMAGE_GUID		|
 		IB_DEVICE_RC_RNR_NAK_GEN;
-	flags = gen->flags;
-	if (flags & MLX5_DEV_CAP_FLAG_BAD_PKEY_CNTR)
+
+	if (MLX5_CAP_GEN(mdev, pkv))
 		props->device_cap_flags |= IB_DEVICE_BAD_PKEY_CNTR;
-	if (flags & MLX5_DEV_CAP_FLAG_BAD_QKEY_CNTR)
+	if (MLX5_CAP_GEN(mdev, qkv))
 		props->device_cap_flags |= IB_DEVICE_BAD_QKEY_CNTR;
-	if (flags & MLX5_DEV_CAP_FLAG_APM)
+	if (MLX5_CAP_GEN(mdev, apm))
 		props->device_cap_flags |= IB_DEVICE_AUTO_PATH_MIG;
 	props->device_cap_flags |= IB_DEVICE_LOCAL_DMA_LKEY;
-	if (flags & MLX5_DEV_CAP_FLAG_XRC)
+	if (MLX5_CAP_GEN(mdev, xrc))
 		props->device_cap_flags |= IB_DEVICE_XRC;
 	props->device_cap_flags |= IB_DEVICE_MEM_MGT_EXTENSIONS;
-	if (flags & MLX5_DEV_CAP_FLAG_SIG_HAND_OVER) {
+	if (MLX5_CAP_GEN(mdev, sho)) {
 		props->device_cap_flags |= IB_DEVICE_SIGNATURE_HANDOVER;
 		/* At this stage no support for signature handover */
 		props->sig_prot_cap = IB_PROT_T10DIF_TYPE_1 |
@@ -116,221 +253,274 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
 		props->sig_guard_cap = IB_GUARD_T10DIF_CRC |
 				       IB_GUARD_T10DIF_CSUM;
 	}
-	if (flags & MLX5_DEV_CAP_FLAG_BLOCK_MCAST)
+	if (MLX5_CAP_GEN(mdev, block_lb_mc))
 		props->device_cap_flags |= IB_DEVICE_BLOCK_MULTICAST_LOOPBACK;
 
-	props->vendor_id	   = be32_to_cpup((__be32 *)(out_mad->data + 36)) &
-		0xffffff;
-	props->vendor_part_id	   = be16_to_cpup((__be16 *)(out_mad->data + 30));
-	props->hw_ver		   = be32_to_cpup((__be32 *)(out_mad->data + 32));
-	memcpy(&props->sys_image_guid, out_mad->data +	4, 8);
+	props->vendor_part_id	   = mdev->pdev->device;
+	props->hw_ver		   = mdev->pdev->revision;
 
 	props->max_mr_size	   = ~0ull;
-	props->page_size_cap	   = gen->min_page_sz;
-	props->max_qp		   = 1 << gen->log_max_qp;
-	props->max_qp_wr	   = gen->max_wqes;
-	max_rq_sg = gen->max_rq_desc_sz / sizeof(struct mlx5_wqe_data_seg);
-	max_sq_sg = (gen->max_sq_desc_sz - sizeof(struct mlx5_wqe_ctrl_seg)) /
-		sizeof(struct mlx5_wqe_data_seg);
+	props->page_size_cap	   = 1ull << MLX5_CAP_GEN(mdev, log_pg_sz);
+	props->max_qp		   = 1 << MLX5_CAP_GEN(mdev, log_max_qp);
+	props->max_qp_wr	   = 1 << MLX5_CAP_GEN(mdev, log_max_qp_sz);
+	max_rq_sg =  MLX5_CAP_GEN(mdev, max_wqe_sz_rq) /
+		     sizeof(struct mlx5_wqe_data_seg);
+	max_sq_sg = (MLX5_CAP_GEN(mdev, max_wqe_sz_sq) -
+		     sizeof(struct mlx5_wqe_ctrl_seg)) /
+		     sizeof(struct mlx5_wqe_data_seg);
 	props->max_sge = min(max_rq_sg, max_sq_sg);
-	props->max_cq		   = 1 << gen->log_max_cq;
-	props->max_cqe		   = gen->max_cqes - 1;
-	props->max_mr		   = 1 << gen->log_max_mkey;
-	props->max_pd		   = 1 << gen->log_max_pd;
-	props->max_qp_rd_atom	   = 1 << gen->log_max_ra_req_qp;
-	props->max_qp_init_rd_atom = 1 << gen->log_max_ra_res_qp;
-	props->max_srq		   = 1 << gen->log_max_srq;
-	props->max_srq_wr	   = gen->max_srq_wqes - 1;
-	props->local_ca_ack_delay  = gen->local_ca_ack_delay;
+	props->max_cq		   = 1 << MLX5_CAP_GEN(mdev, log_max_cq);
+	props->max_cqe = (1 << MLX5_CAP_GEN(mdev, log_max_eq_sz)) - 1;
+	props->max_mr		   = 1 << MLX5_CAP_GEN(mdev, log_max_mkey);
+	props->max_pd		   = 1 << MLX5_CAP_GEN(mdev, log_max_pd);
+	props->max_qp_rd_atom	   = 1 << MLX5_CAP_GEN(mdev, log_max_ra_req_qp);
+	props->max_qp_init_rd_atom = 1 << MLX5_CAP_GEN(mdev, log_max_ra_res_qp);
+	props->max_srq		   = 1 << MLX5_CAP_GEN(mdev, log_max_srq);
+	props->max_srq_wr = (1 << MLX5_CAP_GEN(mdev, log_max_srq_sz)) - 1;
+	props->local_ca_ack_delay  = MLX5_CAP_GEN(mdev, local_ca_ack_delay);
 	props->max_res_rd_atom	   = props->max_qp_rd_atom * props->max_qp;
 	props->max_srq_sge	   = max_rq_sg - 1;
 	props->max_fast_reg_page_list_len = (unsigned int)-1;
-	props->local_ca_ack_delay  = gen->local_ca_ack_delay;
 	props->atomic_cap	   = IB_ATOMIC_NONE;
 	props->masked_atomic_cap   = IB_ATOMIC_NONE;
-	props->max_pkeys	   = be16_to_cpup((__be16 *)(out_mad->data + 28));
-	props->max_mcast_grp	   = 1 << gen->log_max_mcg;
-	props->max_mcast_qp_attach = gen->max_qp_mcg;
+	props->max_mcast_grp	   = 1 << MLX5_CAP_GEN(mdev, log_max_mcg);
+	props->max_mcast_qp_attach = MLX5_CAP_GEN(mdev, max_qp_mcg);
 	props->max_total_mcast_qp_attach = props->max_mcast_qp_attach *
 					   props->max_mcast_grp;
 	props->max_map_per_fmr = INT_MAX; /* no limit in ConnectIB */
 
 #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
-	if (dev->mdev->caps.gen.flags & MLX5_DEV_CAP_FLAG_ON_DMND_PG)
+	if (MLX5_CAP_GEN(mdev, pg))
 		props->device_cap_flags |= IB_DEVICE_ON_DEMAND_PAGING;
 	props->odp_caps = dev->odp_caps;
 #endif
 
-out:
-	kfree(in_mad);
-	kfree(out_mad);
-
-	return err;
+	return 0;
 }
 
-int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
-		       struct ib_port_attr *props)
+enum mlx5_ib_width {
+	MLX5_IB_WIDTH_1X	= 1 << 0,
+	MLX5_IB_WIDTH_2X	= 1 << 1,
+	MLX5_IB_WIDTH_4X	= 1 << 2,
+	MLX5_IB_WIDTH_8X	= 1 << 3,
+	MLX5_IB_WIDTH_12X	= 1 << 4
+};
+
+static int translate_active_width(struct ib_device *ibdev, u8 active_width,
+				  u8 *ib_width)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
-	struct ib_smp *in_mad  = NULL;
-	struct ib_smp *out_mad = NULL;
-	struct mlx5_general_caps *gen;
-	int ext_active_speed;
-	int err = -ENOMEM;
-
-	gen = &dev->mdev->caps.gen;
-	if (port < 1 || port > gen->num_ports) {
-		mlx5_ib_warn(dev, "invalid port number %d\n", port);
-		return -EINVAL;
+	int err = 0;
+
+	if (active_width & MLX5_IB_WIDTH_1X) {
+		*ib_width = IB_WIDTH_1X;
+	} else if (active_width & MLX5_IB_WIDTH_2X) {
+		mlx5_ib_dbg(dev, "active_width %d is not supported by IB spec\n",
+			    (int)active_width);
+		err = -EINVAL;
+	} else if (active_width & MLX5_IB_WIDTH_4X) {
+		*ib_width = IB_WIDTH_4X;
+	} else if (active_width & MLX5_IB_WIDTH_8X) {
+		*ib_width = IB_WIDTH_8X;
+	} else if (active_width & MLX5_IB_WIDTH_12X) {
+		*ib_width = IB_WIDTH_12X;
+	} else {
+		mlx5_ib_dbg(dev, "Invalid active_width %d\n",
+			    (int)active_width);
+		err = -EINVAL;
 	}
 
-	in_mad  = kzalloc(sizeof(*in_mad), GFP_KERNEL);
-	out_mad = kmalloc(sizeof(*out_mad), GFP_KERNEL);
-	if (!in_mad || !out_mad)
-		goto out;
-
-	memset(props, 0, sizeof(*props));
-
-	init_query_mad(in_mad);
-	in_mad->attr_id  = IB_SMP_ATTR_PORT_INFO;
-	in_mad->attr_mod = cpu_to_be32(port);
+	return err;
+}
 
-	err = mlx5_MAD_IFC(dev, 1, 1, port, NULL, NULL, in_mad, out_mad);
-	if (err) {
-		mlx5_ib_warn(dev, "err %d\n", err);
-		goto out;
+static int mlx5_mtu_to_ib_mtu(int mtu)
+{
+	switch (mtu) {
+	case 256: return 1;
+	case 512: return 2;
+	case 1024: return 3;
+	case 2048: return 4;
+	case 4096: return 5;
+	default:
+		pr_warn("invalid mtu\n");
+		return -1;
 	}
+}
 
+enum ib_max_vl_num {
+	__IB_MAX_VL_0		= 1,
+	__IB_MAX_VL_0_1		= 2,
+	__IB_MAX_VL_0_3		= 3,
+	__IB_MAX_VL_0_7		= 4,
+	__IB_MAX_VL_0_14	= 5,
+};
 
-	props->lid		= be16_to_cpup((__be16 *)(out_mad->data + 16));
-	props->lmc		= out_mad->data[34] & 0x7;
-	props->sm_lid		= be16_to_cpup((__be16 *)(out_mad->data + 18));
-	props->sm_sl		= out_mad->data[36] & 0xf;
-	props->state		= out_mad->data[32] & 0xf;
-	props->phys_state	= out_mad->data[33] >> 4;
-	props->port_cap_flags	= be32_to_cpup((__be32 *)(out_mad->data + 20));
-	props->gid_tbl_len	= out_mad->data[50];
-	props->max_msg_sz	= 1 << gen->log_max_msg;
-	props->pkey_tbl_len	= gen->port[port - 1].pkey_table_len;
-	props->bad_pkey_cntr	= be16_to_cpup((__be16 *)(out_mad->data + 46));
-	props->qkey_viol_cntr	= be16_to_cpup((__be16 *)(out_mad->data + 48));
-	props->active_width	= out_mad->data[31] & 0xf;
-	props->active_speed	= out_mad->data[35] >> 4;
-	props->max_mtu		= out_mad->data[41] & 0xf;
-	props->active_mtu	= out_mad->data[36] >> 4;
-	props->subnet_timeout	= out_mad->data[51] & 0x1f;
-	props->max_vl_num	= out_mad->data[37] >> 4;
-	props->init_type_reply	= out_mad->data[41] >> 4;
-
-	/* Check if extended speeds (EDR/FDR/...) are supported */
-	if (props->port_cap_flags & IB_PORT_EXTENDED_SPEEDS_SUP) {
-		ext_active_speed = out_mad->data[62] >> 4;
-
-		switch (ext_active_speed) {
-		case 1:
-			props->active_speed = 16; /* FDR */
-			break;
-		case 2:
-			props->active_speed = 32; /* EDR */
-			break;
-		}
-	}
+enum mlx5_vl_hw_cap {
+	MLX5_VL_HW_0	= 1,
+	MLX5_VL_HW_0_1	= 2,
+	MLX5_VL_HW_0_2	= 3,
+	MLX5_VL_HW_0_3	= 4,
+	MLX5_VL_HW_0_4	= 5,
+	MLX5_VL_HW_0_5	= 6,
+	MLX5_VL_HW_0_6	= 7,
+	MLX5_VL_HW_0_7	= 8,
+	MLX5_VL_HW_0_14	= 15
+};
 
-	/* If reported active speed is QDR, check if is FDR-10 */
-	if (props->active_speed == 4) {
-		if (gen->ext_port_cap[port - 1] &
-		    MLX_EXT_PORT_CAP_FLAG_EXTENDED_PORT_INFO) {
-			init_query_mad(in_mad);
-			in_mad->attr_id = MLX5_ATTR_EXTENDED_PORT_INFO;
-			in_mad->attr_mod = cpu_to_be32(port);
-
-			err = mlx5_MAD_IFC(dev, 1, 1, port,
-					   NULL, NULL, in_mad, out_mad);
-			if (err)
-				goto out;
-
-			/* Checking LinkSpeedActive for FDR-10 */
-			if (out_mad->data[15] & 0x1)
-				props->active_speed = 8;
-		}
-	}
+static int translate_max_vl_num(struct ib_device *ibdev, u8 vl_hw_cap,
+				u8 *max_vl_num)
+{
+	switch (vl_hw_cap) {
+	case MLX5_VL_HW_0:
+		*max_vl_num = __IB_MAX_VL_0;
+		break;
+	case MLX5_VL_HW_0_1:
+		*max_vl_num = __IB_MAX_VL_0_1;
+		break;
+	case MLX5_VL_HW_0_3:
+		*max_vl_num = __IB_MAX_VL_0_3;
+		break;
+	case MLX5_VL_HW_0_7:
+		*max_vl_num = __IB_MAX_VL_0_7;
+		break;
+	case MLX5_VL_HW_0_14:
+		*max_vl_num = __IB_MAX_VL_0_14;
+		break;
 
-out:
-	kfree(in_mad);
-	kfree(out_mad);
+	default:
+		return -EINVAL;
+	}
 
-	return err;
+	return 0;
 }
 
-static int mlx5_ib_query_gid(struct ib_device *ibdev, u8 port, int index,
-			     union ib_gid *gid)
+static int mlx5_query_hca_port(struct ib_device *ibdev, u8 port,
+			       struct ib_port_attr *props)
 {
-	struct ib_smp *in_mad  = NULL;
-	struct ib_smp *out_mad = NULL;
-	int err = -ENOMEM;
+	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;
+	int err;
+	u8 ib_link_width_oper;
+	u8 vl_hw_cap;
 
-	in_mad  = kzalloc(sizeof(*in_mad), GFP_KERNEL);
-	out_mad = kmalloc(sizeof(*out_mad), GFP_KERNEL);
-	if (!in_mad || !out_mad)
+	rep = kzalloc(sizeof(*rep), GFP_KERNEL);
+	if (!rep) {
+		err = -ENOMEM;
 		goto out;
+	}
 
-	init_query_mad(in_mad);
-	in_mad->attr_id  = IB_SMP_ATTR_PORT_INFO;
-	in_mad->attr_mod = cpu_to_be32(port);
+	memset(props, 0, sizeof(*props));
+
+	err = mlx5_query_hca_vport_context(mdev, 0, port, 0, rep);
+	if (err)
+		goto out;
 
-	err = mlx5_MAD_IFC(to_mdev(ibdev), 1, 1, port, NULL, NULL, in_mad, out_mad);
+	props->lid		= rep->lid;
+	props->lmc		= rep->lmc;
+	props->sm_lid		= rep->sm_lid;
+	props->sm_sl		= rep->sm_sl;
+	props->state		= rep->vport_state;
+	props->phys_state	= rep->port_physical_state;
+	props->port_cap_flags	= rep->cap_mask1;
+	props->gid_tbl_len	= mlx5_get_gid_table_len(MLX5_CAP_GEN(mdev, gid_table_size));
+	props->max_msg_sz	= 1 << MLX5_CAP_GEN(mdev, log_max_msg);
+	props->pkey_tbl_len	= mlx5_to_sw_pkey_sz(MLX5_CAP_GEN(mdev, pkey_table_size));
+	props->bad_pkey_cntr	= rep->pkey_violation_counter;
+	props->qkey_viol_cntr	= rep->qkey_violation_counter;
+	props->subnet_timeout	= rep->subnet_timeout;
+	props->init_type_reply	= rep->init_type_reply;
+
+	err = mlx5_query_port_link_width_oper(mdev, &ib_link_width_oper, port);
 	if (err)
 		goto out;
 
-	memcpy(gid->raw, out_mad->data + 8, 8);
+	err = translate_active_width(ibdev, ib_link_width_oper,
+				     &props->active_width);
+	if (err)
+		goto out;
+	err = mlx5_query_port_proto_oper(mdev, &props->active_speed, MLX5_PTYS_IB,
+					 port);
+	if (err)
+		goto out;
+
+	err = mlx5_query_port_max_mtu(mdev, &max_mtu, port);
+	if (err)
+		goto out;
 
-	init_query_mad(in_mad);
-	in_mad->attr_id  = IB_SMP_ATTR_GUID_INFO;
-	in_mad->attr_mod = cpu_to_be32(index / 8);
+	props->max_mtu = mlx5_mtu_to_ib_mtu(max_mtu);
 
-	err = mlx5_MAD_IFC(to_mdev(ibdev), 1, 1, port, NULL, NULL, in_mad, out_mad);
+	err = mlx5_query_port_oper_mtu(mdev, &oper_mtu, port);
 	if (err)
 		goto out;
 
-	memcpy(gid->raw + 8, out_mad->data + (index % 8) * 8, 8);
+	props->active_mtu = mlx5_mtu_to_ib_mtu(oper_mtu);
+
+	err = mlx5_query_port_vl_hw_cap(mdev, &vl_hw_cap, port);
+	if (err)
+		goto out;
 
+	err = translate_max_vl_num(ibdev, vl_hw_cap,
+				   &props->max_vl_num);
 out:
-	kfree(in_mad);
-	kfree(out_mad);
+	kfree(rep);
 	return err;
 }
 
-static int mlx5_ib_query_pkey(struct ib_device *ibdev, u8 port, u16 index,
-			      u16 *pkey)
+int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
+		       struct ib_port_attr *props)
 {
-	struct ib_smp *in_mad  = NULL;
-	struct ib_smp *out_mad = NULL;
-	int err = -ENOMEM;
+	switch (mlx5_get_vport_access_method(ibdev)) {
+	case MLX5_VPORT_ACCESS_METHOD_MAD:
+		return mlx5_query_mad_ifc_port(ibdev, port, props);
 
-	in_mad  = kzalloc(sizeof(*in_mad), GFP_KERNEL);
-	out_mad = kmalloc(sizeof(*out_mad), GFP_KERNEL);
-	if (!in_mad || !out_mad)
-		goto out;
+	case MLX5_VPORT_ACCESS_METHOD_HCA:
+		return mlx5_query_hca_port(ibdev, port, props);
 
-	init_query_mad(in_mad);
-	in_mad->attr_id  = IB_SMP_ATTR_PKEY_TABLE;
-	in_mad->attr_mod = cpu_to_be32(index / 32);
+	default:
+		return -EINVAL;
+	}
+}
 
-	err = mlx5_MAD_IFC(to_mdev(ibdev), 1, 1, port, NULL, NULL, in_mad, out_mad);
-	if (err)
-		goto out;
+static int mlx5_ib_query_gid(struct ib_device *ibdev, u8 port, int index,
+			     union ib_gid *gid)
+{
+	struct mlx5_ib_dev *dev = to_mdev(ibdev);
+	struct mlx5_core_dev *mdev = dev->mdev;
 
-	*pkey = be16_to_cpu(((__be16 *)out_mad->data)[index % 32]);
+	switch (mlx5_get_vport_access_method(ibdev)) {
+	case MLX5_VPORT_ACCESS_METHOD_MAD:
+		return mlx5_query_mad_ifc_gids(ibdev, port, index, gid);
+
+	case MLX5_VPORT_ACCESS_METHOD_HCA:
+		return mlx5_query_hca_vport_gid(mdev, 0, port, 0, index, gid);
+
+	default:
+		return -EINVAL;
+	}
 
-out:
-	kfree(in_mad);
-	kfree(out_mad);
-	return err;
 }
 
-struct mlx5_reg_node_desc {
-	u8	desc[64];
-};
+static int mlx5_ib_query_pkey(struct ib_device *ibdev, u8 port, u16 index,
+			      u16 *pkey)
+{
+	struct mlx5_ib_dev *dev = to_mdev(ibdev);
+	struct mlx5_core_dev *mdev = dev->mdev;
+
+	switch (mlx5_get_vport_access_method(ibdev)) {
+	case MLX5_VPORT_ACCESS_METHOD_MAD:
+		return mlx5_query_mad_ifc_pkey(ibdev, port, index, pkey);
+
+	case MLX5_VPORT_ACCESS_METHOD_HCA:
+	case MLX5_VPORT_ACCESS_METHOD_NIC:
+		return mlx5_query_hca_vport_pkey(mdev, 0, port,  0, index,
+						 pkey);
+	default:
+		return -EINVAL;
+	}
+}
 
 static int mlx5_ib_modify_device(struct ib_device *ibdev, int mask,
 				 struct ib_device_modify *props)
@@ -392,7 +582,6 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
 	struct mlx5_ib_alloc_ucontext_req_v2 req;
 	struct mlx5_ib_alloc_ucontext_resp resp;
 	struct mlx5_ib_ucontext *context;
-	struct mlx5_general_caps *gen;
 	struct mlx5_uuar_info *uuari;
 	struct mlx5_uar *uars;
 	int gross_uuars;
@@ -403,7 +592,6 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
 	int i;
 	size_t reqlen;
 
-	gen = &dev->mdev->caps.gen;
 	if (!dev->ib_active)
 		return ERR_PTR(-EAGAIN);
 
@@ -436,14 +624,14 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
 
 	num_uars = req.total_num_uuars / MLX5_NON_FP_BF_REGS_PER_PAGE;
 	gross_uuars = num_uars * MLX5_BF_REGS_PER_PAGE;
-	resp.qp_tab_size      = 1 << gen->log_max_qp;
-	resp.bf_reg_size      = gen->bf_reg_size;
-	resp.cache_line_size  = L1_CACHE_BYTES;
-	resp.max_sq_desc_sz = gen->max_sq_desc_sz;
-	resp.max_rq_desc_sz = gen->max_rq_desc_sz;
-	resp.max_send_wqebb = gen->max_wqes;
-	resp.max_recv_wr = gen->max_wqes;
-	resp.max_srq_recv_wr = gen->max_srq_wqes;
+	resp.qp_tab_size = 1 << MLX5_CAP_GEN(dev->mdev, log_max_qp);
+	resp.bf_reg_size = 1 << MLX5_CAP_GEN(dev->mdev, log_bf_reg_size);
+	resp.cache_line_size = L1_CACHE_BYTES;
+	resp.max_sq_desc_sz = MLX5_CAP_GEN(dev->mdev, max_wqe_sz_sq);
+	resp.max_rq_desc_sz = MLX5_CAP_GEN(dev->mdev, max_wqe_sz_rq);
+	resp.max_send_wqebb = 1 << MLX5_CAP_GEN(dev->mdev, log_max_qp_sz);
+	resp.max_recv_wr = 1 << MLX5_CAP_GEN(dev->mdev, log_max_qp_sz);
+	resp.max_srq_recv_wr = 1 << MLX5_CAP_GEN(dev->mdev, log_max_srq_sz);
 
 	context = kzalloc(sizeof(*context), GFP_KERNEL);
 	if (!context)
@@ -493,7 +681,7 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
 	mutex_init(&context->db_page_mutex);
 
 	resp.tot_uuars = req.total_num_uuars;
-	resp.num_ports = gen->num_ports;
+	resp.num_ports = MLX5_CAP_GEN(dev->mdev, num_ports);
 	err = ib_copy_to_udata(udata, &resp,
 			       sizeof(resp) - sizeof(resp.reserved));
 	if (err)
@@ -731,37 +919,15 @@ static int mlx5_ib_mcg_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
 
 static int init_node_data(struct mlx5_ib_dev *dev)
 {
-	struct ib_smp *in_mad  = NULL;
-	struct ib_smp *out_mad = NULL;
-	int err = -ENOMEM;
-
-	in_mad  = kzalloc(sizeof(*in_mad), GFP_KERNEL);
-	out_mad = kmalloc(sizeof(*out_mad), GFP_KERNEL);
-	if (!in_mad || !out_mad)
-		goto out;
-
-	init_query_mad(in_mad);
-	in_mad->attr_id = IB_SMP_ATTR_NODE_DESC;
-
-	err = mlx5_MAD_IFC(dev, 1, 1, 1, NULL, NULL, in_mad, out_mad);
-	if (err)
-		goto out;
-
-	memcpy(dev->ib_dev.node_desc, out_mad->data, 64);
-
-	in_mad->attr_id = IB_SMP_ATTR_NODE_INFO;
+	int err;
 
-	err = mlx5_MAD_IFC(dev, 1, 1, 1, NULL, NULL, in_mad, out_mad);
+	err = mlx5_query_node_desc(dev, dev->ib_dev.node_desc);
 	if (err)
-		goto out;
+		return err;
 
-	dev->mdev->rev_id = be32_to_cpup((__be32 *)(out_mad->data + 32));
-	memcpy(&dev->ib_dev.node_guid, out_mad->data + 12, 8);
+	dev->mdev->rev_id = dev->mdev->pdev->revision;
 
-out:
-	kfree(in_mad);
-	kfree(out_mad);
-	return err;
+	return mlx5_query_node_guid(dev, &dev->ib_dev.node_guid);
 }
 
 static ssize_t show_fw_pages(struct device *device, struct device_attribute *attr,
@@ -895,11 +1061,9 @@ static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context,
 
 static void get_ext_port_caps(struct mlx5_ib_dev *dev)
 {
-	struct mlx5_general_caps *gen;
 	int port;
 
-	gen = &dev->mdev->caps.gen;
-	for (port = 1; port <= gen->num_ports; port++)
+	for (port = 1; port <= MLX5_CAP_GEN(dev->mdev, num_ports); port++)
 		mlx5_query_ext_port_caps(dev, port);
 }
 
@@ -907,11 +1071,9 @@ static int get_port_caps(struct mlx5_ib_dev *dev)
 {
 	struct ib_device_attr *dprops = NULL;
 	struct ib_port_attr *pprops = NULL;
-	struct mlx5_general_caps *gen;
 	int err = -ENOMEM;
 	int port;
 
-	gen = &dev->mdev->caps.gen;
 	pprops = kmalloc(sizeof(*pprops), GFP_KERNEL);
 	if (!pprops)
 		goto out;
@@ -926,14 +1088,17 @@ static int get_port_caps(struct mlx5_ib_dev *dev)
 		goto out;
 	}
 
-	for (port = 1; port <= gen->num_ports; port++) {
+	for (port = 1; port <= MLX5_CAP_GEN(dev->mdev, num_ports); port++) {
 		err = mlx5_ib_query_port(&dev->ib_dev, port, pprops);
 		if (err) {
-			mlx5_ib_warn(dev, "query_port %d failed %d\n", port, err);
+			mlx5_ib_warn(dev, "query_port %d failed %d\n",
+				     port, err);
 			break;
 		}
-		gen->port[port - 1].pkey_table_len = dprops->max_pkeys;
-		gen->port[port - 1].gid_table_len = pprops->gid_tbl_len;
+		dev->mdev->port_caps[port - 1].pkey_table_len =
+						dprops->max_pkeys;
+		dev->mdev->port_caps[port - 1].gid_table_len =
+						pprops->gid_tbl_len;
 		mlx5_ib_dbg(dev, "pkey_table_len %d, gid_table_len %d\n",
 			    dprops->max_pkeys, pprops->gid_tbl_len);
 	}
@@ -1159,8 +1324,29 @@ static int create_dev_resources(struct mlx5_ib_resources *devr)
 	atomic_inc(&devr->p0->usecnt);
 	atomic_set(&devr->s0->usecnt, 0);
 
+	memset(&attr, 0, sizeof(attr));
+	attr.attr.max_sge = 1;
+	attr.attr.max_wr = 1;
+	attr.srq_type = IB_SRQT_BASIC;
+	devr->s1 = mlx5_ib_create_srq(devr->p0, &attr, NULL);
+	if (IS_ERR(devr->s1)) {
+		ret = PTR_ERR(devr->s1);
+		goto error5;
+	}
+	devr->s1->device	= &dev->ib_dev;
+	devr->s1->pd		= devr->p0;
+	devr->s1->uobject       = NULL;
+	devr->s1->event_handler = NULL;
+	devr->s1->srq_context   = NULL;
+	devr->s1->srq_type      = IB_SRQT_BASIC;
+	devr->s1->ext.xrc.cq	= devr->c0;
+	atomic_inc(&devr->p0->usecnt);
+	atomic_set(&devr->s0->usecnt, 0);
+
 	return 0;
 
+error5:
+	mlx5_ib_destroy_srq(devr->s0);
 error4:
 	mlx5_ib_dealloc_xrcd(devr->x1);
 error3:
@@ -1175,6 +1361,7 @@ error0:
 
 static void destroy_dev_resources(struct mlx5_ib_resources *devr)
 {
+	mlx5_ib_destroy_srq(devr->s1);
 	mlx5_ib_destroy_srq(devr->s0);
 	mlx5_ib_dealloc_xrcd(devr->x0);
 	mlx5_ib_dealloc_xrcd(devr->x1);
@@ -1188,6 +1375,10 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
 	int err;
 	int i;
 
+	/* don't create IB instance over Eth ports, no RoCE yet! */
+	if (MLX5_CAP_GEN(mdev, port_type) == MLX5_CAP_PORT_TYPE_ETH)
+		return NULL;
+
 	printk_once(KERN_INFO "%s", mlx5_version);
 
 	dev = (struct mlx5_ib_dev *)ib_alloc_device(sizeof(*dev));
@@ -1200,15 +1391,16 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
 	if (err)
 		goto err_dealloc;
 
-	get_ext_port_caps(dev);
+	if (mlx5_use_mad_ifc(dev))
+		get_ext_port_caps(dev);
 
 	MLX5_INIT_DOORBELL_LOCK(&dev->uar_lock);
 
 	strlcpy(dev->ib_dev.name, "mlx5_%d", IB_DEVICE_NAME_MAX);
 	dev->ib_dev.owner		= THIS_MODULE;
 	dev->ib_dev.node_type		= RDMA_NODE_IB_CA;
-	dev->ib_dev.local_dma_lkey	= mdev->caps.gen.reserved_lkey;
-	dev->num_ports		= mdev->caps.gen.num_ports;
+	dev->ib_dev.local_dma_lkey	= 0 /* not supported for now */;
+	dev->num_ports		= MLX5_CAP_GEN(mdev, num_ports);
 	dev->ib_dev.phys_port_cnt     = dev->num_ports;
 	dev->ib_dev.num_comp_vectors    =
 		dev->mdev->priv.eq_table.num_comp_vectors;
@@ -1286,9 +1478,9 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
 	dev->ib_dev.free_fast_reg_page_list  = mlx5_ib_free_fast_reg_page_list;
 	dev->ib_dev.check_mr_status	= mlx5_ib_check_mr_status;
 
-	mlx5_ib_internal_query_odp_caps(dev);
+	mlx5_ib_internal_fill_odp_caps(dev);
 
-	if (mdev->caps.gen.flags & MLX5_DEV_CAP_FLAG_XRC) {
+	if (MLX5_CAP_GEN(mdev, xrc)) {
 		dev->ib_dev.alloc_xrcd = mlx5_ib_alloc_xrcd;
 		dev->ib_dev.dealloc_xrcd = mlx5_ib_dealloc_xrcd;
 		dev->ib_dev.uverbs_cmd_mask |=
diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index dff1cfcdf476..873dc354766a 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -415,6 +415,7 @@ struct mlx5_ib_resources {
 	struct ib_xrcd	*x1;
 	struct ib_pd	*p0;
 	struct ib_srq	*s0;
+	struct ib_srq	*s1;
 };
 
 struct mlx5_ib_dev {
@@ -594,6 +595,22 @@ struct ib_xrcd *mlx5_ib_alloc_xrcd(struct ib_device *ibdev,
 int mlx5_ib_dealloc_xrcd(struct ib_xrcd *xrcd);
 int mlx5_ib_get_buf_offset(u64 addr, int page_shift, u32 *offset);
 int mlx5_query_ext_port_caps(struct mlx5_ib_dev *dev, u8 port);
+int mlx5_query_mad_ifc_smp_attr_node_info(struct ib_device *ibdev,
+					  struct ib_smp *out_mad);
+int mlx5_query_mad_ifc_system_image_guid(struct ib_device *ibdev,
+					 __be64 *sys_image_guid);
+int mlx5_query_mad_ifc_max_pkeys(struct ib_device *ibdev,
+				 u16 *max_pkeys);
+int mlx5_query_mad_ifc_vendor_id(struct ib_device *ibdev,
+				 u32 *vendor_id);
+int mlx5_query_mad_ifc_node_desc(struct mlx5_ib_dev *dev, char *node_desc);
+int mlx5_query_mad_ifc_node_guid(struct mlx5_ib_dev *dev, __be64 *node_guid);
+int mlx5_query_mad_ifc_pkey(struct ib_device *ibdev, u8 port, u16 index,
+			    u16 *pkey);
+int mlx5_query_mad_ifc_gids(struct ib_device *ibdev, u8 port, int index,
+			    union ib_gid *gid);
+int mlx5_query_mad_ifc_port(struct ib_device *ibdev, u8 port,
+			    struct ib_port_attr *props);
 int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
 		       struct ib_port_attr *props);
 int mlx5_ib_init_fmr(struct mlx5_ib_dev *dev);
@@ -617,7 +634,7 @@ int mlx5_ib_check_mr_status(struct ib_mr *ibmr, u32 check_mask,
 #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
 extern struct workqueue_struct *mlx5_ib_page_fault_wq;
 
-int mlx5_ib_internal_query_odp_caps(struct mlx5_ib_dev *dev);
+void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev);
 void mlx5_ib_mr_pfault_handler(struct mlx5_ib_qp *qp,
 			       struct mlx5_ib_pfault *pfault);
 void mlx5_ib_odp_create_qp(struct mlx5_ib_qp *qp);
@@ -631,9 +648,9 @@ void mlx5_ib_invalidate_range(struct ib_umem *umem, unsigned long start,
 			      unsigned long end);
 
 #else /* CONFIG_INFINIBAND_ON_DEMAND_PAGING */
-static inline int mlx5_ib_internal_query_odp_caps(struct mlx5_ib_dev *dev)
+static inline void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev)
 {
-	return 0;
+	return;
 }
 
 static inline void mlx5_ib_odp_create_qp(struct mlx5_ib_qp *qp)		{}
diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c
index 71c593583864..bc9a0de897cb 100644
--- a/drivers/infiniband/hw/mlx5/mr.c
+++ b/drivers/infiniband/hw/mlx5/mr.c
@@ -975,8 +975,7 @@ static struct mlx5_ib_mr *reg_create(struct ib_pd *pd, u64 virt_addr,
 	struct mlx5_ib_mr *mr;
 	int inlen;
 	int err;
-	bool pg_cap = !!(dev->mdev->caps.gen.flags &
-			 MLX5_DEV_CAP_FLAG_ON_DMND_PG);
+	bool pg_cap = !!(MLX5_CAP_GEN(dev->mdev, pg));
 
 	mr = kzalloc(sizeof(*mr), GFP_KERNEL);
 	if (!mr)
diff --git a/drivers/infiniband/hw/mlx5/odp.c b/drivers/infiniband/hw/mlx5/odp.c
index 5099db08afd2..aa8391e75385 100644
--- a/drivers/infiniband/hw/mlx5/odp.c
+++ b/drivers/infiniband/hw/mlx5/odp.c
@@ -109,40 +109,33 @@ void mlx5_ib_invalidate_range(struct ib_umem *umem, unsigned long start,
 	ib_umem_odp_unmap_dma_pages(umem, start, end);
 }
 
-#define COPY_ODP_BIT_MLX_TO_IB(reg, ib_caps, field_name, bit_name) do {	\
-	if (be32_to_cpu(reg.field_name) & MLX5_ODP_SUPPORT_##bit_name)	\
-		ib_caps->field_name |= IB_ODP_SUPPORT_##bit_name;	\
-} while (0)
-
-int mlx5_ib_internal_query_odp_caps(struct mlx5_ib_dev *dev)
+void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev)
 {
-	int err;
-	struct mlx5_odp_caps hw_caps;
 	struct ib_odp_caps *caps = &dev->odp_caps;
 
 	memset(caps, 0, sizeof(*caps));
 
-	if (!(dev->mdev->caps.gen.flags & MLX5_DEV_CAP_FLAG_ON_DMND_PG))
-		return 0;
-
-	err = mlx5_query_odp_caps(dev->mdev, &hw_caps);
-	if (err)
-		goto out;
+	if (!MLX5_CAP_GEN(dev->mdev, pg))
+		return;
 
 	caps->general_caps = IB_ODP_SUPPORT;
-	COPY_ODP_BIT_MLX_TO_IB(hw_caps, caps, per_transport_caps.ud_odp_caps,
-			       SEND);
-	COPY_ODP_BIT_MLX_TO_IB(hw_caps, caps, per_transport_caps.rc_odp_caps,
-			       SEND);
-	COPY_ODP_BIT_MLX_TO_IB(hw_caps, caps, per_transport_caps.rc_odp_caps,
-			       RECV);
-	COPY_ODP_BIT_MLX_TO_IB(hw_caps, caps, per_transport_caps.rc_odp_caps,
-			       WRITE);
-	COPY_ODP_BIT_MLX_TO_IB(hw_caps, caps, per_transport_caps.rc_odp_caps,
-			       READ);
-
-out:
-	return err;
+
+	if (MLX5_CAP_ODP(dev->mdev, ud_odp_caps.send))
+		caps->per_transport_caps.ud_odp_caps |= IB_ODP_SUPPORT_SEND;
+
+	if (MLX5_CAP_ODP(dev->mdev, rc_odp_caps.send))
+		caps->per_transport_caps.rc_odp_caps |= IB_ODP_SUPPORT_SEND;
+
+	if (MLX5_CAP_ODP(dev->mdev, rc_odp_caps.receive))
+		caps->per_transport_caps.rc_odp_caps |= IB_ODP_SUPPORT_RECV;
+
+	if (MLX5_CAP_ODP(dev->mdev, rc_odp_caps.write))
+		caps->per_transport_caps.rc_odp_caps |= IB_ODP_SUPPORT_WRITE;
+
+	if (MLX5_CAP_ODP(dev->mdev, rc_odp_caps.read))
+		caps->per_transport_caps.rc_odp_caps |= IB_ODP_SUPPORT_READ;
+
+	return;
 }
 
 static struct mlx5_ib_mr *mlx5_ib_odp_find_mr_lkey(struct mlx5_ib_dev *dev,
diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index 4d7024b899cb..203c8a45e095 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -220,13 +220,11 @@ static void mlx5_ib_qp_event(struct mlx5_core_qp *qp, int type)
 static int set_rq_size(struct mlx5_ib_dev *dev, struct ib_qp_cap *cap,
 		       int has_rq, struct mlx5_ib_qp *qp, struct mlx5_ib_create_qp *ucmd)
 {
-	struct mlx5_general_caps *gen;
 	int wqe_size;
 	int wq_size;
 
-	gen = &dev->mdev->caps.gen;
 	/* Sanity check RQ size before proceeding */
-	if (cap->max_recv_wr  > gen->max_wqes)
+	if (cap->max_recv_wr > (1 << MLX5_CAP_GEN(dev->mdev, log_max_qp_sz)))
 		return -EINVAL;
 
 	if (!has_rq) {
@@ -246,10 +244,11 @@ static int set_rq_size(struct mlx5_ib_dev *dev, struct ib_qp_cap *cap,
 			wq_size = roundup_pow_of_two(cap->max_recv_wr) * wqe_size;
 			wq_size = max_t(int, wq_size, MLX5_SEND_WQE_BB);
 			qp->rq.wqe_cnt = wq_size / wqe_size;
-			if (wqe_size > gen->max_rq_desc_sz) {
+			if (wqe_size > MLX5_CAP_GEN(dev->mdev, max_wqe_sz_rq)) {
 				mlx5_ib_dbg(dev, "wqe_size %d, max %d\n",
 					    wqe_size,
-					    gen->max_rq_desc_sz);
+					    MLX5_CAP_GEN(dev->mdev,
+							 max_wqe_sz_rq));
 				return -EINVAL;
 			}
 			qp->rq.wqe_shift = ilog2(wqe_size);
@@ -330,11 +329,9 @@ static int calc_send_wqe(struct ib_qp_init_attr *attr)
 static int calc_sq_size(struct mlx5_ib_dev *dev, struct ib_qp_init_attr *attr,
 			struct mlx5_ib_qp *qp)
 {
-	struct mlx5_general_caps *gen;
 	int wqe_size;
 	int wq_size;
 
-	gen = &dev->mdev->caps.gen;
 	if (!attr->cap.max_send_wr)
 		return 0;
 
@@ -343,9 +340,9 @@ static int calc_sq_size(struct mlx5_ib_dev *dev, struct ib_qp_init_attr *attr,
 	if (wqe_size < 0)
 		return wqe_size;
 
-	if (wqe_size > gen->max_sq_desc_sz) {
+	if (wqe_size > MLX5_CAP_GEN(dev->mdev, max_wqe_sz_sq)) {
 		mlx5_ib_dbg(dev, "wqe_size(%d) > max_sq_desc_sz(%d)\n",
-			    wqe_size, gen->max_sq_desc_sz);
+			    wqe_size, MLX5_CAP_GEN(dev->mdev, max_wqe_sz_sq));
 		return -EINVAL;
 	}
 
@@ -358,9 +355,10 @@ static int calc_sq_size(struct mlx5_ib_dev *dev, struct ib_qp_init_attr *attr,
 
 	wq_size = roundup_pow_of_two(attr->cap.max_send_wr * wqe_size);
 	qp->sq.wqe_cnt = wq_size / MLX5_SEND_WQE_BB;
-	if (qp->sq.wqe_cnt > gen->max_wqes) {
+	if (qp->sq.wqe_cnt > (1 << MLX5_CAP_GEN(dev->mdev, log_max_qp_sz))) {
 		mlx5_ib_dbg(dev, "wqe count(%d) exceeds limits(%d)\n",
-			    qp->sq.wqe_cnt, gen->max_wqes);
+			    qp->sq.wqe_cnt,
+			    1 << MLX5_CAP_GEN(dev->mdev, log_max_qp_sz));
 		return -ENOMEM;
 	}
 	qp->sq.wqe_shift = ilog2(MLX5_SEND_WQE_BB);
@@ -375,13 +373,11 @@ static int set_user_buf_size(struct mlx5_ib_dev *dev,
 			    struct mlx5_ib_qp *qp,
 			    struct mlx5_ib_create_qp *ucmd)
 {
-	struct mlx5_general_caps *gen;
 	int desc_sz = 1 << qp->sq.wqe_shift;
 
-	gen = &dev->mdev->caps.gen;
-	if (desc_sz > gen->max_sq_desc_sz) {
+	if (desc_sz > MLX5_CAP_GEN(dev->mdev, max_wqe_sz_sq)) {
 		mlx5_ib_warn(dev, "desc_sz %d, max_sq_desc_sz %d\n",
-			     desc_sz, gen->max_sq_desc_sz);
+			     desc_sz, MLX5_CAP_GEN(dev->mdev, max_wqe_sz_sq));
 		return -EINVAL;
 	}
 
@@ -393,9 +389,10 @@ static int set_user_buf_size(struct mlx5_ib_dev *dev,
 
 	qp->sq.wqe_cnt = ucmd->sq_wqe_count;
 
-	if (qp->sq.wqe_cnt > gen->max_wqes) {
+	if (qp->sq.wqe_cnt > (1 << MLX5_CAP_GEN(dev->mdev, log_max_qp_sz))) {
 		mlx5_ib_warn(dev, "wqe_cnt %d, max_wqes %d\n",
-			     qp->sq.wqe_cnt, gen->max_wqes);
+			     qp->sq.wqe_cnt,
+			     1 << MLX5_CAP_GEN(dev->mdev, log_max_qp_sz));
 		return -EINVAL;
 	}
 
@@ -768,7 +765,7 @@ static int create_kernel_qp(struct mlx5_ib_dev *dev,
 	qp->sq.offset = qp->rq.wqe_cnt << qp->rq.wqe_shift;
 	qp->buf_size = err + (qp->rq.wqe_cnt << qp->rq.wqe_shift);
 
-	err = mlx5_buf_alloc(dev->mdev, qp->buf_size, PAGE_SIZE * 2, &qp->buf);
+	err = mlx5_buf_alloc(dev->mdev, qp->buf_size, &qp->buf);
 	if (err) {
 		mlx5_ib_dbg(dev, "err %d\n", err);
 		goto err_uuar;
@@ -866,22 +863,21 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd,
 			    struct ib_udata *udata, struct mlx5_ib_qp *qp)
 {
 	struct mlx5_ib_resources *devr = &dev->devr;
+	struct mlx5_core_dev *mdev = dev->mdev;
 	struct mlx5_ib_create_qp_resp resp;
 	struct mlx5_create_qp_mbox_in *in;
-	struct mlx5_general_caps *gen;
 	struct mlx5_ib_create_qp ucmd;
 	int inlen = sizeof(*in);
 	int err;
 
 	mlx5_ib_odp_create_qp(qp);
 
-	gen = &dev->mdev->caps.gen;
 	mutex_init(&qp->mutex);
 	spin_lock_init(&qp->sq.lock);
 	spin_lock_init(&qp->rq.lock);
 
 	if (init_attr->create_flags & IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK) {
-		if (!(gen->flags & MLX5_DEV_CAP_FLAG_BLOCK_MCAST)) {
+		if (!MLX5_CAP_GEN(mdev, block_lb_mc)) {
 			mlx5_ib_dbg(dev, "block multicast loopback isn't supported\n");
 			return -EINVAL;
 		} else {
@@ -914,15 +910,17 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd,
 
 	if (pd) {
 		if (pd->uobject) {
+			__u32 max_wqes =
+				1 << MLX5_CAP_GEN(mdev, log_max_qp_sz);
 			mlx5_ib_dbg(dev, "requested sq_wqe_count (%d)\n", ucmd.sq_wqe_count);
 			if (ucmd.rq_wqe_shift != qp->rq.wqe_shift ||
 			    ucmd.rq_wqe_count != qp->rq.wqe_cnt) {
 				mlx5_ib_dbg(dev, "invalid rq params\n");
 				return -EINVAL;
 			}
-			if (ucmd.sq_wqe_count > gen->max_wqes) {
+			if (ucmd.sq_wqe_count > max_wqes) {
 				mlx5_ib_dbg(dev, "requested sq_wqe_count (%d) > max allowed (%d)\n",
-					    ucmd.sq_wqe_count, gen->max_wqes);
+					    ucmd.sq_wqe_count, max_wqes);
 				return -EINVAL;
 			}
 			err = create_user_qp(dev, pd, qp, udata, &in, &resp, &inlen);
@@ -1014,7 +1012,8 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd,
 			in->ctx.rq_type_srqn |= cpu_to_be32(to_msrq(init_attr->srq)->msrq.srqn);
 		} else {
 			in->ctx.xrcd = cpu_to_be32(to_mxrcd(devr->x1)->xrcdn);
-			in->ctx.rq_type_srqn |= cpu_to_be32(to_msrq(devr->s0)->msrq.srqn);
+			in->ctx.rq_type_srqn |=
+				cpu_to_be32(to_msrq(devr->s1)->msrq.srqn);
 		}
 	}
 
@@ -1226,7 +1225,6 @@ struct ib_qp *mlx5_ib_create_qp(struct ib_pd *pd,
 				struct ib_qp_init_attr *init_attr,
 				struct ib_udata *udata)
 {
-	struct mlx5_general_caps *gen;
 	struct mlx5_ib_dev *dev;
 	struct mlx5_ib_qp *qp;
 	u16 xrcdn = 0;
@@ -1244,12 +1242,11 @@ struct ib_qp *mlx5_ib_create_qp(struct ib_pd *pd,
 		}
 		dev = to_mdev(to_mxrcd(init_attr->xrcd)->ibxrcd.device);
 	}
-	gen = &dev->mdev->caps.gen;
 
 	switch (init_attr->qp_type) {
 	case IB_QPT_XRC_TGT:
 	case IB_QPT_XRC_INI:
-		if (!(gen->flags & MLX5_DEV_CAP_FLAG_XRC)) {
+		if (!MLX5_CAP_GEN(dev->mdev, xrc)) {
 			mlx5_ib_dbg(dev, "XRC not supported\n");
 			return ERR_PTR(-ENOSYS);
 		}
@@ -1356,9 +1353,6 @@ enum {
 
 static int ib_rate_to_mlx5(struct mlx5_ib_dev *dev, u8 rate)
 {
-	struct mlx5_general_caps *gen;
-
-	gen = &dev->mdev->caps.gen;
 	if (rate == IB_RATE_PORT_CURRENT) {
 		return 0;
 	} else if (rate < IB_RATE_2_5_GBPS || rate > IB_RATE_300_GBPS) {
@@ -1366,7 +1360,7 @@ static int ib_rate_to_mlx5(struct mlx5_ib_dev *dev, u8 rate)
 	} else {
 		while (rate != IB_RATE_2_5_GBPS &&
 		       !(1 << (rate + MLX5_STAT_RATE_OFFSET) &
-			 gen->stat_rate_support))
+			 MLX5_CAP_GEN(dev->mdev, stat_rate_support)))
 			--rate;
 	}
 
@@ -1377,10 +1371,8 @@ static int mlx5_set_path(struct mlx5_ib_dev *dev, const struct ib_ah_attr *ah,
 			 struct mlx5_qp_path *path, u8 port, int attr_mask,
 			 u32 path_flags, const struct ib_qp_attr *attr)
 {
-	struct mlx5_general_caps *gen;
 	int err;
 
-	gen = &dev->mdev->caps.gen;
 	path->fl = (path_flags & MLX5_PATH_FLAG_FL) ? 0x80 : 0;
 	path->free_ar = (path_flags & MLX5_PATH_FLAG_FREE_AR) ? 0x80 : 0;
 
@@ -1391,9 +1383,11 @@ static int mlx5_set_path(struct mlx5_ib_dev *dev, const struct ib_ah_attr *ah,
 	path->rlid	= cpu_to_be16(ah->dlid);
 
 	if (ah->ah_flags & IB_AH_GRH) {
-		if (ah->grh.sgid_index >= gen->port[port - 1].gid_table_len) {
-			pr_err(KERN_ERR "sgid_index (%u) too large. max is %d\n",
-			       ah->grh.sgid_index, gen->port[port - 1].gid_table_len);
+		if (ah->grh.sgid_index >=
+		    dev->mdev->port_caps[port - 1].gid_table_len) {
+			pr_err("sgid_index (%u) too large. max is %d\n",
+			       ah->grh.sgid_index,
+			       dev->mdev->port_caps[port - 1].gid_table_len);
 			return -EINVAL;
 		}
 		path->grh_mlid |= 1 << 7;
@@ -1570,7 +1564,6 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
 	struct mlx5_ib_qp *qp = to_mqp(ibqp);
 	struct mlx5_ib_cq *send_cq, *recv_cq;
 	struct mlx5_qp_context *context;
-	struct mlx5_general_caps *gen;
 	struct mlx5_modify_qp_mbox_in *in;
 	struct mlx5_ib_pd *pd;
 	enum mlx5_qp_state mlx5_cur, mlx5_new;
@@ -1579,7 +1572,6 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
 	int mlx5_st;
 	int err;
 
-	gen = &dev->mdev->caps.gen;
 	in = kzalloc(sizeof(*in), GFP_KERNEL);
 	if (!in)
 		return -ENOMEM;
@@ -1619,7 +1611,8 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
 			err = -EINVAL;
 			goto out;
 		}
-		context->mtu_msgmax = (attr->path_mtu << 5) | gen->log_max_msg;
+		context->mtu_msgmax = (attr->path_mtu << 5) |
+				      (u8)MLX5_CAP_GEN(dev->mdev, log_max_msg);
 	}
 
 	if (attr_mask & IB_QP_DEST_QPN)
@@ -1777,11 +1770,9 @@ int mlx5_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 	struct mlx5_ib_dev *dev = to_mdev(ibqp->device);
 	struct mlx5_ib_qp *qp = to_mqp(ibqp);
 	enum ib_qp_state cur_state, new_state;
-	struct mlx5_general_caps *gen;
 	int err = -EINVAL;
 	int port;
 
-	gen = &dev->mdev->caps.gen;
 	mutex_lock(&qp->mutex);
 
 	cur_state = attr_mask & IB_QP_CUR_STATE ? attr->cur_qp_state : qp->state;
@@ -1793,21 +1784,25 @@ int mlx5_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 		goto out;
 
 	if ((attr_mask & IB_QP_PORT) &&
-	    (attr->port_num == 0 || attr->port_num > gen->num_ports))
+	    (attr->port_num == 0 ||
+	     attr->port_num > MLX5_CAP_GEN(dev->mdev, num_ports)))
 		goto out;
 
 	if (attr_mask & IB_QP_PKEY_INDEX) {
 		port = attr_mask & IB_QP_PORT ? attr->port_num : qp->port;
-		if (attr->pkey_index >= gen->port[port - 1].pkey_table_len)
+		if (attr->pkey_index >=
+		    dev->mdev->port_caps[port - 1].pkey_table_len)
 			goto out;
 	}
 
 	if (attr_mask & IB_QP_MAX_QP_RD_ATOMIC &&
-	    attr->max_rd_atomic > (1 << gen->log_max_ra_res_qp))
+	    attr->max_rd_atomic >
+	    (1 << MLX5_CAP_GEN(dev->mdev, log_max_ra_res_qp)))
 		goto out;
 
 	if (attr_mask & IB_QP_MAX_DEST_RD_ATOMIC &&
-	    attr->max_dest_rd_atomic > (1 << gen->log_max_ra_req_qp))
+	    attr->max_dest_rd_atomic >
+	    (1 << MLX5_CAP_GEN(dev->mdev, log_max_ra_req_qp)))
 		goto out;
 
 	if (cur_state == new_state && cur_state == IB_QPS_RESET) {
@@ -3009,7 +3004,7 @@ static void to_ib_ah_attr(struct mlx5_ib_dev *ibdev, struct ib_ah_attr *ib_ah_at
 	ib_ah_attr->port_num	  = path->port;
 
 	if (ib_ah_attr->port_num == 0 ||
-	    ib_ah_attr->port_num > dev->caps.gen.num_ports)
+	    ib_ah_attr->port_num > MLX5_CAP_GEN(dev, num_ports))
 		return;
 
 	ib_ah_attr->sl = path->sl & 0xf;
@@ -3135,12 +3130,10 @@ struct ib_xrcd *mlx5_ib_alloc_xrcd(struct ib_device *ibdev,
 					  struct ib_udata *udata)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
-	struct mlx5_general_caps *gen;
 	struct mlx5_ib_xrcd *xrcd;
 	int err;
 
-	gen = &dev->mdev->caps.gen;
-	if (!(gen->flags & MLX5_DEV_CAP_FLAG_XRC))
+	if (!MLX5_CAP_GEN(dev->mdev, xrc))
 		return ERR_PTR(-ENOSYS);
 
 	xrcd = kmalloc(sizeof(*xrcd), GFP_KERNEL);
diff --git a/drivers/infiniband/hw/mlx5/srq.c b/drivers/infiniband/hw/mlx5/srq.c
index 02d77a29764d..e008505e96e9 100644
--- a/drivers/infiniband/hw/mlx5/srq.c
+++ b/drivers/infiniband/hw/mlx5/srq.c
@@ -165,7 +165,7 @@ static int create_srq_kernel(struct mlx5_ib_dev *dev, struct mlx5_ib_srq *srq,
 		return err;
 	}
 
-	if (mlx5_buf_alloc(dev->mdev, buf_size, PAGE_SIZE * 2, &srq->buf)) {
+	if (mlx5_buf_alloc(dev->mdev, buf_size, &srq->buf)) {
 		mlx5_ib_dbg(dev, "buf alloc failed\n");
 		err = -ENOMEM;
 		goto err_db;
@@ -236,7 +236,6 @@ struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd,
 				  struct ib_udata *udata)
 {
 	struct mlx5_ib_dev *dev = to_mdev(pd->device);
-	struct mlx5_general_caps *gen;
 	struct mlx5_ib_srq *srq;
 	int desc_size;
 	int buf_size;
@@ -245,13 +244,13 @@ struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd,
 	int uninitialized_var(inlen);
 	int is_xrc;
 	u32 flgs, xrcdn;
+	__u32 max_srq_wqes = 1 << MLX5_CAP_GEN(dev->mdev, log_max_srq_sz);
 
-	gen = &dev->mdev->caps.gen;
 	/* Sanity check SRQ size before proceeding */
-	if (init_attr->attr.max_wr >= gen->max_srq_wqes) {
+	if (init_attr->attr.max_wr >= max_srq_wqes) {
 		mlx5_ib_dbg(dev, "max_wr %d, cap %d\n",
 			    init_attr->attr.max_wr,
-			    gen->max_srq_wqes);
+			    max_srq_wqes);
 		return ERR_PTR(-EINVAL);
 	}
 
@@ -303,7 +302,7 @@ struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd,
 
 	in->ctx.pd = cpu_to_be32(to_mpd(pd)->pdn);
 	in->ctx.db_record = cpu_to_be64(srq->db.dma);
-	err = mlx5_core_create_srq(dev->mdev, &srq->msrq, in, inlen);
+	err = mlx5_core_create_srq(dev->mdev, &srq->msrq, in, inlen, is_xrc);
 	kvfree(in);
 	if (err) {
 		mlx5_ib_dbg(dev, "create SRQ failed, err %d\n", err);
diff --git a/drivers/infiniband/hw/nes/nes.c b/drivers/infiniband/hw/nes/nes.c
index 3b2a6dc8ea99..9f9d5c563a61 100644
--- a/drivers/infiniband/hw/nes/nes.c
+++ b/drivers/infiniband/hw/nes/nes.c
@@ -116,6 +116,7 @@ static struct ibnl_client_cbs nes_nl_cb_table[] = {
 	[RDMA_NL_IWPM_REG_PID] = {.dump = iwpm_register_pid_cb},
 	[RDMA_NL_IWPM_ADD_MAPPING] = {.dump = iwpm_add_mapping_cb},
 	[RDMA_NL_IWPM_QUERY_MAPPING] = {.dump = iwpm_add_and_query_mapping_cb},
+	[RDMA_NL_IWPM_REMOTE_INFO] = {.dump = iwpm_remote_info_cb},
 	[RDMA_NL_IWPM_HANDLE_ERR] = {.dump = iwpm_mapping_error_cb},
 	[RDMA_NL_IWPM_MAPINFO] = {.dump = iwpm_mapping_info_cb},
 	[RDMA_NL_IWPM_MAPINFO_NUM] = {.dump = iwpm_ack_mapping_info_cb}
diff --git a/drivers/infiniband/hw/nes/nes_cm.c b/drivers/infiniband/hw/nes/nes_cm.c
index 6f09a72e78d7..72b43417cbe3 100644
--- a/drivers/infiniband/hw/nes/nes_cm.c
+++ b/drivers/infiniband/hw/nes/nes_cm.c
@@ -596,27 +596,52 @@ static void nes_form_reg_msg(struct nes_vnic *nesvnic,
 	memcpy(pm_msg->if_name, nesvnic->netdev->name, IWPM_IFNAME_SIZE);
 }
 
+static void record_sockaddr_info(struct sockaddr_storage *addr_info,
+					nes_addr_t *ip_addr, u16 *port_num)
+{
+	struct sockaddr_in *in_addr = (struct sockaddr_in *)addr_info;
+
+	if (in_addr->sin_family == AF_INET) {
+		*ip_addr = ntohl(in_addr->sin_addr.s_addr);
+		*port_num = ntohs(in_addr->sin_port);
+	}
+}
+
 /*
  * nes_record_pm_msg - Save the received mapping info
  */
 static void nes_record_pm_msg(struct nes_cm_info *cm_info,
 			struct iwpm_sa_data *pm_msg)
 {
-	struct sockaddr_in *mapped_loc_addr =
-			(struct sockaddr_in *)&pm_msg->mapped_loc_addr;
-	struct sockaddr_in *mapped_rem_addr =
-			(struct sockaddr_in *)&pm_msg->mapped_rem_addr;
-
-	if (mapped_loc_addr->sin_family == AF_INET) {
-		cm_info->mapped_loc_addr =
-			ntohl(mapped_loc_addr->sin_addr.s_addr);
-		cm_info->mapped_loc_port = ntohs(mapped_loc_addr->sin_port);
-	}
-	if (mapped_rem_addr->sin_family == AF_INET) {
-		cm_info->mapped_rem_addr =
-			ntohl(mapped_rem_addr->sin_addr.s_addr);
-		cm_info->mapped_rem_port = ntohs(mapped_rem_addr->sin_port);
-	}
+	record_sockaddr_info(&pm_msg->mapped_loc_addr,
+		&cm_info->mapped_loc_addr, &cm_info->mapped_loc_port);
+
+	record_sockaddr_info(&pm_msg->mapped_rem_addr,
+		&cm_info->mapped_rem_addr, &cm_info->mapped_rem_port);
+}
+
+/*
+ * nes_get_reminfo - Get the address info of the remote connecting peer
+ */
+static int nes_get_remote_addr(struct nes_cm_node *cm_node)
+{
+	struct sockaddr_storage mapped_loc_addr, mapped_rem_addr;
+	struct sockaddr_storage remote_addr;
+	int ret;
+
+	nes_create_sockaddr(htonl(cm_node->mapped_loc_addr),
+			htons(cm_node->mapped_loc_port), &mapped_loc_addr);
+	nes_create_sockaddr(htonl(cm_node->mapped_rem_addr),
+			htons(cm_node->mapped_rem_port), &mapped_rem_addr);
+
+	ret = iwpm_get_remote_info(&mapped_loc_addr, &mapped_rem_addr,
+				&remote_addr, RDMA_NL_NES);
+	if (ret)
+		nes_debug(NES_DBG_CM, "Unable to find remote peer address info\n");
+	else
+		record_sockaddr_info(&remote_addr, &cm_node->rem_addr,
+				&cm_node->rem_port);
+	return ret;
 }
 
 /**
@@ -1566,9 +1591,14 @@ static struct nes_cm_node *make_cm_node(struct nes_cm_core *cm_core,
 		return NULL;
 
 	/* set our node specific transport info */
-	cm_node->loc_addr = cm_info->loc_addr;
+	if (listener) {
+		cm_node->loc_addr = listener->loc_addr;
+		cm_node->loc_port = listener->loc_port;
+	} else {
+		cm_node->loc_addr = cm_info->loc_addr;
+		cm_node->loc_port = cm_info->loc_port;
+	}
 	cm_node->rem_addr = cm_info->rem_addr;
-	cm_node->loc_port = cm_info->loc_port;
 	cm_node->rem_port = cm_info->rem_port;
 
 	cm_node->mapped_loc_addr = cm_info->mapped_loc_addr;
@@ -2151,6 +2181,7 @@ static int handle_ack_pkt(struct nes_cm_node *cm_node, struct sk_buff *skb,
 		cm_node->state = NES_CM_STATE_ESTABLISHED;
 		if (datasize) {
 			cm_node->tcp_cntxt.rcv_nxt = inc_sequence + datasize;
+			nes_get_remote_addr(cm_node);
 			handle_rcv_mpa(cm_node, skb);
 		} else { /* rcvd ACK only */
 			dev_kfree_skb_any(skb);
diff --git a/drivers/infiniband/hw/ocrdma/ocrdma.h b/drivers/infiniband/hw/ocrdma/ocrdma.h
index c9780d919769..b396344fae16 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma.h
+++ b/drivers/infiniband/hw/ocrdma/ocrdma.h
@@ -40,7 +40,7 @@
 #include <be_roce.h>
 #include "ocrdma_sli.h"
 
-#define OCRDMA_ROCE_DRV_VERSION "10.4.205.0u"
+#define OCRDMA_ROCE_DRV_VERSION "10.6.0.0"
 
 #define OCRDMA_ROCE_DRV_DESC "Emulex OneConnect RoCE Driver"
 #define OCRDMA_NODE_DESC "Emulex OneConnect RoCE HCA"
@@ -515,6 +515,8 @@ static inline int ocrdma_resolve_dmac(struct ocrdma_dev *dev,
 	memcpy(&in6, ah_attr->grh.dgid.raw, sizeof(in6));
 	if (rdma_is_multicast_addr(&in6))
 		rdma_get_mcast_mac(&in6, mac_addr);
+	else if (rdma_link_local_addr(&in6))
+		rdma_get_ll_mac(&in6, mac_addr);
 	else
 		memcpy(mac_addr, ah_attr->dmac, ETH_ALEN);
 	return 0;
diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_ah.c b/drivers/infiniband/hw/ocrdma/ocrdma_ah.c
index d812904f3984..f5a5ea836dbd 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_ah.c
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_ah.c
@@ -56,7 +56,13 @@ static inline int set_av_attr(struct ocrdma_dev *dev, struct ocrdma_ah *ah,
 	vlan_tag = attr->vlan_id;
 	if (!vlan_tag || (vlan_tag > 0xFFF))
 		vlan_tag = dev->pvid;
-	if (vlan_tag && (vlan_tag < 0x1000)) {
+	if (vlan_tag || dev->pfc_state) {
+		if (!vlan_tag) {
+			pr_err("ocrdma%d:Using VLAN with PFC is recommended\n",
+				dev->id);
+			pr_err("ocrdma%d:Using VLAN 0 for this connection\n",
+				dev->id);
+		}
 		eth.eth_type = cpu_to_be16(0x8100);
 		eth.roce_eth_type = cpu_to_be16(OCRDMA_ROCE_ETH_TYPE);
 		vlan_tag |= (dev->sl & 0x07) << OCRDMA_VID_PCP_SHIFT;
@@ -121,7 +127,9 @@ struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr)
 		goto av_conf_err;
 	}
 
-	if (pd->uctx) {
+	if ((pd->uctx) &&
+	    (!rdma_is_multicast_addr((struct in6_addr *)attr->grh.dgid.raw)) &&
+	    (!rdma_link_local_addr((struct in6_addr *)attr->grh.dgid.raw))) {
 		status = rdma_addr_find_dmac_by_grh(&sgid, &attr->grh.dgid,
                                         attr->dmac, &attr->vlan_id);
 		if (status) {
diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_hw.c b/drivers/infiniband/hw/ocrdma/ocrdma_hw.c
index 0c9e95909a64..47615ff33bc6 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_hw.c
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_hw.c
@@ -933,12 +933,18 @@ static irqreturn_t ocrdma_irq_handler(int irq, void *handle)
 	struct ocrdma_eqe eqe;
 	struct ocrdma_eqe *ptr;
 	u16 cq_id;
+	u8 mcode;
 	int budget = eq->cq_cnt;
 
 	do {
 		ptr = ocrdma_get_eqe(eq);
 		eqe = *ptr;
 		ocrdma_le32_to_cpu(&eqe, sizeof(eqe));
+		mcode = (eqe.id_valid & OCRDMA_EQE_MAJOR_CODE_MASK)
+				>> OCRDMA_EQE_MAJOR_CODE_SHIFT;
+		if (mcode == OCRDMA_MAJOR_CODE_SENTINAL)
+			pr_err("EQ full on eqid = 0x%x, eqe = 0x%x\n",
+			       eq->q.id, eqe.id_valid);
 		if ((eqe.id_valid & OCRDMA_EQE_VALID_MASK) == 0)
 			break;
 
@@ -1434,27 +1440,30 @@ static int ocrdma_mbx_alloc_pd_range(struct ocrdma_dev *dev)
 	struct ocrdma_alloc_pd_range_rsp *rsp;
 
 	/* Pre allocate the DPP PDs */
-	cmd = ocrdma_init_emb_mqe(OCRDMA_CMD_ALLOC_PD_RANGE, sizeof(*cmd));
-	if (!cmd)
-		return -ENOMEM;
-	cmd->pd_count = dev->attr.max_dpp_pds;
-	cmd->enable_dpp_rsvd |= OCRDMA_ALLOC_PD_ENABLE_DPP;
-	status = ocrdma_mbx_cmd(dev, (struct ocrdma_mqe *)cmd);
-	if (status)
-		goto mbx_err;
-	rsp = (struct ocrdma_alloc_pd_range_rsp *)cmd;
-
-	if ((rsp->dpp_page_pdid & OCRDMA_ALLOC_PD_RSP_DPP) && rsp->pd_count) {
-		dev->pd_mgr->dpp_page_index = rsp->dpp_page_pdid >>
-				OCRDMA_ALLOC_PD_RSP_DPP_PAGE_SHIFT;
-		dev->pd_mgr->pd_dpp_start = rsp->dpp_page_pdid &
-				OCRDMA_ALLOC_PD_RNG_RSP_START_PDID_MASK;
-		dev->pd_mgr->max_dpp_pd = rsp->pd_count;
-		pd_bitmap_size = BITS_TO_LONGS(rsp->pd_count) * sizeof(long);
-		dev->pd_mgr->pd_dpp_bitmap = kzalloc(pd_bitmap_size,
-						     GFP_KERNEL);
+	if (dev->attr.max_dpp_pds) {
+		cmd = ocrdma_init_emb_mqe(OCRDMA_CMD_ALLOC_PD_RANGE,
+					  sizeof(*cmd));
+		if (!cmd)
+			return -ENOMEM;
+		cmd->pd_count = dev->attr.max_dpp_pds;
+		cmd->enable_dpp_rsvd |= OCRDMA_ALLOC_PD_ENABLE_DPP;
+		status = ocrdma_mbx_cmd(dev, (struct ocrdma_mqe *)cmd);
+		rsp = (struct ocrdma_alloc_pd_range_rsp *)cmd;
+
+		if (!status && (rsp->dpp_page_pdid & OCRDMA_ALLOC_PD_RSP_DPP) &&
+		    rsp->pd_count) {
+			dev->pd_mgr->dpp_page_index = rsp->dpp_page_pdid >>
+					OCRDMA_ALLOC_PD_RSP_DPP_PAGE_SHIFT;
+			dev->pd_mgr->pd_dpp_start = rsp->dpp_page_pdid &
+					OCRDMA_ALLOC_PD_RNG_RSP_START_PDID_MASK;
+			dev->pd_mgr->max_dpp_pd = rsp->pd_count;
+			pd_bitmap_size =
+				BITS_TO_LONGS(rsp->pd_count) * sizeof(long);
+			dev->pd_mgr->pd_dpp_bitmap = kzalloc(pd_bitmap_size,
+							     GFP_KERNEL);
+		}
+		kfree(cmd);
 	}
-	kfree(cmd);
 
 	cmd = ocrdma_init_emb_mqe(OCRDMA_CMD_ALLOC_PD_RANGE, sizeof(*cmd));
 	if (!cmd)
@@ -1462,10 +1471,8 @@ static int ocrdma_mbx_alloc_pd_range(struct ocrdma_dev *dev)
 
 	cmd->pd_count = dev->attr.max_pd - dev->attr.max_dpp_pds;
 	status = ocrdma_mbx_cmd(dev, (struct ocrdma_mqe *)cmd);
-	if (status)
-		goto mbx_err;
 	rsp = (struct ocrdma_alloc_pd_range_rsp *)cmd;
-	if (rsp->pd_count) {
+	if (!status && rsp->pd_count) {
 		dev->pd_mgr->pd_norm_start = rsp->dpp_page_pdid &
 					OCRDMA_ALLOC_PD_RNG_RSP_START_PDID_MASK;
 		dev->pd_mgr->max_normal_pd = rsp->pd_count;
@@ -1473,15 +1480,13 @@ static int ocrdma_mbx_alloc_pd_range(struct ocrdma_dev *dev)
 		dev->pd_mgr->pd_norm_bitmap = kzalloc(pd_bitmap_size,
 						      GFP_KERNEL);
 	}
+	kfree(cmd);
 
 	if (dev->pd_mgr->pd_norm_bitmap || dev->pd_mgr->pd_dpp_bitmap) {
 		/* Enable PD resource manager */
 		dev->pd_mgr->pd_prealloc_valid = true;
-	} else {
-		return -ENOMEM;
+		return 0;
 	}
-mbx_err:
-	kfree(cmd);
 	return status;
 }
 
@@ -2406,7 +2411,7 @@ int ocrdma_mbx_query_qp(struct ocrdma_dev *dev, struct ocrdma_qp *qp,
 	struct ocrdma_query_qp *cmd;
 	struct ocrdma_query_qp_rsp *rsp;
 
-	cmd = ocrdma_init_emb_mqe(OCRDMA_CMD_QUERY_QP, sizeof(*cmd));
+	cmd = ocrdma_init_emb_mqe(OCRDMA_CMD_QUERY_QP, sizeof(*rsp));
 	if (!cmd)
 		return status;
 	cmd->qp_id = qp->id;
@@ -2428,7 +2433,7 @@ static int ocrdma_set_av_params(struct ocrdma_qp *qp,
 	int status;
 	struct ib_ah_attr *ah_attr = &attrs->ah_attr;
 	union ib_gid sgid, zgid;
-	u32 vlan_id;
+	u32 vlan_id = 0xFFFF;
 	u8 mac_addr[6];
 	struct ocrdma_dev *dev = get_ocrdma_dev(qp->ibqp.device);
 
@@ -2468,12 +2473,22 @@ static int ocrdma_set_av_params(struct ocrdma_qp *qp,
 	cmd->params.vlan_dmac_b4_to_b5 = mac_addr[4] | (mac_addr[5] << 8);
 	if (attr_mask & IB_QP_VID) {
 		vlan_id = attrs->vlan_id;
+	} else if (dev->pfc_state) {
+		vlan_id = 0;
+		pr_err("ocrdma%d:Using VLAN with PFC is recommended\n",
+			dev->id);
+		pr_err("ocrdma%d:Using VLAN 0 for this connection\n",
+			dev->id);
+	}
+
+	if (vlan_id < 0x1000) {
 		cmd->params.vlan_dmac_b4_to_b5 |=
 		    vlan_id << OCRDMA_QP_PARAMS_VLAN_SHIFT;
 		cmd->flags |= OCRDMA_QP_PARA_VLAN_EN_VALID;
 		cmd->params.rnt_rc_sl_fl |=
 			(dev->sl & 0x07) << OCRDMA_QP_PARAMS_SL_SHIFT;
 	}
+
 	return 0;
 }
 
@@ -2519,8 +2534,10 @@ static int ocrdma_set_qp_params(struct ocrdma_qp *qp,
 		cmd->flags |= OCRDMA_QP_PARA_DST_QPN_VALID;
 	}
 	if (attr_mask & IB_QP_PATH_MTU) {
-		if (attrs->path_mtu < IB_MTU_256 ||
+		if (attrs->path_mtu < IB_MTU_512 ||
 		    attrs->path_mtu > IB_MTU_4096) {
+			pr_err("ocrdma%d: IB MTU %d is not supported\n",
+			       dev->id, ib_mtu_enum_to_int(attrs->path_mtu));
 			status = -EINVAL;
 			goto pmtu_err;
 		}
@@ -3147,9 +3164,9 @@ void ocrdma_cleanup_hw(struct ocrdma_dev *dev)
 	ocrdma_free_pd_pool(dev);
 	ocrdma_mbx_delete_ah_tbl(dev);
 
-	/* cleanup the eqs */
-	ocrdma_destroy_eqs(dev);
-
 	/* cleanup the control path */
 	ocrdma_destroy_mq(dev);
+
+	/* cleanup the eqs */
+	ocrdma_destroy_eqs(dev);
 }
diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_sli.h b/drivers/infiniband/hw/ocrdma/ocrdma_sli.h
index 243c87c8bd65..02ad0aee99af 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_sli.h
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_sli.h
@@ -1176,6 +1176,8 @@ struct ocrdma_query_qp_rsp {
 	struct ocrdma_mqe_hdr hdr;
 	struct ocrdma_mbx_rsp rsp;
 	struct ocrdma_qp_params params;
+	u32 dpp_credits_cqid;
+	u32 rbq_id;
 };
 
 enum {
@@ -1624,12 +1626,19 @@ struct ocrdma_delete_ah_tbl_rsp {
 enum {
 	OCRDMA_EQE_VALID_SHIFT		= 0,
 	OCRDMA_EQE_VALID_MASK		= BIT(0),
+	OCRDMA_EQE_MAJOR_CODE_MASK      = 0x0E,
+	OCRDMA_EQE_MAJOR_CODE_SHIFT     = 0x01,
 	OCRDMA_EQE_FOR_CQE_MASK		= 0xFFFE,
 	OCRDMA_EQE_RESOURCE_ID_SHIFT	= 16,
 	OCRDMA_EQE_RESOURCE_ID_MASK	= 0xFFFF <<
 				OCRDMA_EQE_RESOURCE_ID_SHIFT,
 };
 
+enum major_code {
+	OCRDMA_MAJOR_CODE_COMPLETION    = 0x00,
+	OCRDMA_MAJOR_CODE_SENTINAL      = 0x01
+};
+
 struct ocrdma_eqe {
 	u32 id_valid;
 };
diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c
index 877175563634..9dcb66077d6c 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c
@@ -365,7 +365,7 @@ static struct ocrdma_pd *_ocrdma_alloc_pd(struct ocrdma_dev *dev,
 	if (!pd)
 		return ERR_PTR(-ENOMEM);
 
-	if (udata && uctx) {
+	if (udata && uctx && dev->attr.max_dpp_pds) {
 		pd->dpp_enabled =
 			ocrdma_get_asic_type(dev) == OCRDMA_ASIC_GEN_SKH_R;
 		pd->num_dpp_qp =
@@ -1721,18 +1721,20 @@ int ocrdma_destroy_qp(struct ib_qp *ibqp)
 	struct ocrdma_qp *qp;
 	struct ocrdma_dev *dev;
 	struct ib_qp_attr attrs;
-	int attr_mask = IB_QP_STATE;
+	int attr_mask;
 	unsigned long flags;
 
 	qp = get_ocrdma_qp(ibqp);
 	dev = get_ocrdma_dev(ibqp->device);
 
-	attrs.qp_state = IB_QPS_ERR;
 	pd = qp->pd;
 
 	/* change the QP state to ERROR */
-	_ocrdma_modify_qp(ibqp, &attrs, attr_mask);
-
+	if (qp->state != OCRDMA_QPS_RST) {
+		attrs.qp_state = IB_QPS_ERR;
+		attr_mask = IB_QP_STATE;
+		_ocrdma_modify_qp(ibqp, &attrs, attr_mask);
+	}
 	/* ensure that CQEs for newly created QP (whose id may be same with
 	 * one which just getting destroyed are same), dont get
 	 * discarded until the old CQEs are discarded.
diff --git a/drivers/infiniband/hw/qib/qib.h b/drivers/infiniband/hw/qib/qib.h
index ffd48bfc4923..7df16f74bb45 100644
--- a/drivers/infiniband/hw/qib/qib.h
+++ b/drivers/infiniband/hw/qib/qib.h
@@ -903,7 +903,7 @@ struct qib_devdata {
 	/* PCI Device ID (here for NodeInfo) */
 	u16 deviceid;
 	/* for write combining settings */
-	unsigned long wc_cookie;
+	int wc_cookie;
 	unsigned long wc_base;
 	unsigned long wc_len;
 
@@ -1136,7 +1136,6 @@ extern struct qib_devdata *qib_lookup(int unit);
 extern u32 qib_cpulist_count;
 extern unsigned long *qib_cpulist;
 
-extern unsigned qib_wc_pat;
 extern unsigned qib_cc_table_size;
 int qib_init(struct qib_devdata *, int);
 int init_chip_wc_pat(struct qib_devdata *dd, u32);
diff --git a/drivers/infiniband/hw/qib/qib_file_ops.c b/drivers/infiniband/hw/qib/qib_file_ops.c
index 9ea6c440a00c..725881890c4a 100644
--- a/drivers/infiniband/hw/qib/qib_file_ops.c
+++ b/drivers/infiniband/hw/qib/qib_file_ops.c
@@ -835,7 +835,8 @@ static int mmap_piobufs(struct vm_area_struct *vma,
 	vma->vm_flags &= ~VM_MAYREAD;
 	vma->vm_flags |= VM_DONTCOPY | VM_DONTEXPAND;
 
-	if (qib_wc_pat)
+	/* We used PAT if wc_cookie == 0 */
+	if (!dd->wc_cookie)
 		vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot);
 
 	ret = io_remap_pfn_range(vma, vma->vm_start, phys >> PAGE_SHIFT,
diff --git a/drivers/infiniband/hw/qib/qib_fs.c b/drivers/infiniband/hw/qib/qib_fs.c
index 650897a8591e..bdd5d3857203 100644
--- a/drivers/infiniband/hw/qib/qib_fs.c
+++ b/drivers/infiniband/hw/qib/qib_fs.c
@@ -89,14 +89,14 @@ static int create_file(const char *name, umode_t mode,
 {
 	int error;
 
-	mutex_lock(&parent->d_inode->i_mutex);
+	mutex_lock(&d_inode(parent)->i_mutex);
 	*dentry = lookup_one_len(name, parent, strlen(name));
 	if (!IS_ERR(*dentry))
-		error = qibfs_mknod(parent->d_inode, *dentry,
+		error = qibfs_mknod(d_inode(parent), *dentry,
 				    mode, fops, data);
 	else
 		error = PTR_ERR(*dentry);
-	mutex_unlock(&parent->d_inode->i_mutex);
+	mutex_unlock(&d_inode(parent)->i_mutex);
 
 	return error;
 }
@@ -455,10 +455,10 @@ static int remove_file(struct dentry *parent, char *name)
 	}
 
 	spin_lock(&tmp->d_lock);
-	if (!d_unhashed(tmp) && tmp->d_inode) {
+	if (!d_unhashed(tmp) && d_really_is_positive(tmp)) {
 		__d_drop(tmp);
 		spin_unlock(&tmp->d_lock);
-		simple_unlink(parent->d_inode, tmp);
+		simple_unlink(d_inode(parent), tmp);
 	} else {
 		spin_unlock(&tmp->d_lock);
 	}
@@ -481,7 +481,7 @@ static int remove_device_files(struct super_block *sb,
 	int ret, i;
 
 	root = dget(sb->s_root);
-	mutex_lock(&root->d_inode->i_mutex);
+	mutex_lock(&d_inode(root)->i_mutex);
 	snprintf(unit, sizeof(unit), "%u", dd->unit);
 	dir = lookup_one_len(unit, root, strlen(unit));
 
@@ -491,7 +491,7 @@ static int remove_device_files(struct super_block *sb,
 		goto bail;
 	}
 
-	mutex_lock(&dir->d_inode->i_mutex);
+	mutex_lock(&d_inode(dir)->i_mutex);
 	remove_file(dir, "counters");
 	remove_file(dir, "counter_names");
 	remove_file(dir, "portcounter_names");
@@ -506,13 +506,13 @@ static int remove_device_files(struct super_block *sb,
 		}
 	}
 	remove_file(dir, "flash");
-	mutex_unlock(&dir->d_inode->i_mutex);
-	ret = simple_rmdir(root->d_inode, dir);
+	mutex_unlock(&d_inode(dir)->i_mutex);
+	ret = simple_rmdir(d_inode(root), dir);
 	d_delete(dir);
 	dput(dir);
 
 bail:
-	mutex_unlock(&root->d_inode->i_mutex);
+	mutex_unlock(&d_inode(root)->i_mutex);
 	dput(root);
 	return ret;
 }
diff --git a/drivers/infiniband/hw/qib/qib_iba6120.c b/drivers/infiniband/hw/qib/qib_iba6120.c
index 0d2ba59af30a..4b927809d1a1 100644
--- a/drivers/infiniband/hw/qib/qib_iba6120.c
+++ b/drivers/infiniband/hw/qib/qib_iba6120.c
@@ -3315,11 +3315,9 @@ static int init_6120_variables(struct qib_devdata *dd)
 	qib_6120_config_ctxts(dd);
 	qib_set_ctxtcnt(dd);
 
-	if (qib_wc_pat) {
-		ret = init_chip_wc_pat(dd, 0);
-		if (ret)
-			goto bail;
-	}
+	ret = init_chip_wc_pat(dd, 0);
+	if (ret)
+		goto bail;
 	set_6120_baseaddrs(dd); /* set chip access pointers now */
 
 	ret = 0;
diff --git a/drivers/infiniband/hw/qib/qib_iba7220.c b/drivers/infiniband/hw/qib/qib_iba7220.c
index 22affda8af88..00b2af211157 100644
--- a/drivers/infiniband/hw/qib/qib_iba7220.c
+++ b/drivers/infiniband/hw/qib/qib_iba7220.c
@@ -4126,11 +4126,9 @@ static int qib_init_7220_variables(struct qib_devdata *dd)
 	qib_7220_config_ctxts(dd);
 	qib_set_ctxtcnt(dd);  /* needed for PAT setup */
 
-	if (qib_wc_pat) {
-		ret = init_chip_wc_pat(dd, 0);
-		if (ret)
-			goto bail;
-	}
+	ret = init_chip_wc_pat(dd, 0);
+	if (ret)
+		goto bail;
 	set_7220_baseaddrs(dd); /* set chip access pointers now */
 
 	ret = 0;
diff --git a/drivers/infiniband/hw/qib/qib_iba7322.c b/drivers/infiniband/hw/qib/qib_iba7322.c
index ef97b71c8f7d..f32b4628e991 100644
--- a/drivers/infiniband/hw/qib/qib_iba7322.c
+++ b/drivers/infiniband/hw/qib/qib_iba7322.c
@@ -6429,6 +6429,7 @@ static int qib_init_7322_variables(struct qib_devdata *dd)
 	unsigned features, pidx, sbufcnt;
 	int ret, mtu;
 	u32 sbufs, updthresh;
+	resource_size_t vl15off;
 
 	/* pport structs are contiguous, allocated after devdata */
 	ppd = (struct qib_pportdata *)(dd + 1);
@@ -6677,29 +6678,27 @@ static int qib_init_7322_variables(struct qib_devdata *dd)
 	qib_7322_config_ctxts(dd);
 	qib_set_ctxtcnt(dd);
 
-	if (qib_wc_pat) {
-		resource_size_t vl15off;
-		/*
-		 * We do not set WC on the VL15 buffers to avoid
-		 * a rare problem with unaligned writes from
-		 * interrupt-flushed store buffers, so we need
-		 * to map those separately here.  We can't solve
-		 * this for the rarely used mtrr case.
-		 */
-		ret = init_chip_wc_pat(dd, 0);
-		if (ret)
-			goto bail;
+	/*
+	 * We do not set WC on the VL15 buffers to avoid
+	 * a rare problem with unaligned writes from
+	 * interrupt-flushed store buffers, so we need
+	 * to map those separately here.  We can't solve
+	 * this for the rarely used mtrr case.
+	 */
+	ret = init_chip_wc_pat(dd, 0);
+	if (ret)
+		goto bail;
 
-		/* vl15 buffers start just after the 4k buffers */
-		vl15off = dd->physaddr + (dd->piobufbase >> 32) +
-			dd->piobcnt4k * dd->align4k;
-		dd->piovl15base	= ioremap_nocache(vl15off,
-						  NUM_VL15_BUFS * dd->align4k);
-		if (!dd->piovl15base) {
-			ret = -ENOMEM;
-			goto bail;
-		}
+	/* vl15 buffers start just after the 4k buffers */
+	vl15off = dd->physaddr + (dd->piobufbase >> 32) +
+		  dd->piobcnt4k * dd->align4k;
+	dd->piovl15base	= ioremap_nocache(vl15off,
+					  NUM_VL15_BUFS * dd->align4k);
+	if (!dd->piovl15base) {
+		ret = -ENOMEM;
+		goto bail;
 	}
+
 	qib_7322_set_baseaddrs(dd); /* set chip access pointers now */
 
 	ret = 0;
diff --git a/drivers/infiniband/hw/qib/qib_init.c b/drivers/infiniband/hw/qib/qib_init.c
index 2ee36953e234..7e00470adc30 100644
--- a/drivers/infiniband/hw/qib/qib_init.c
+++ b/drivers/infiniband/hw/qib/qib_init.c
@@ -91,15 +91,6 @@ MODULE_PARM_DESC(krcvqs, "number of kernel receive queues per IB port");
 unsigned qib_cc_table_size;
 module_param_named(cc_table_size, qib_cc_table_size, uint, S_IRUGO);
 MODULE_PARM_DESC(cc_table_size, "Congestion control table entries 0 (CCA disabled - default), min = 128, max = 1984");
-/*
- * qib_wc_pat parameter:
- *      0 is WC via MTRR
- *      1 is WC via PAT
- *      If PAT initialization fails, code reverts back to MTRR
- */
-unsigned qib_wc_pat = 1; /* default (1) is to use PAT, not MTRR */
-module_param_named(wc_pat, qib_wc_pat, uint, S_IRUGO);
-MODULE_PARM_DESC(wc_pat, "enable write-combining via PAT mechanism");
 
 static void verify_interrupt(unsigned long);
 
@@ -1377,8 +1368,7 @@ static void cleanup_device_data(struct qib_devdata *dd)
 		spin_unlock(&dd->pport[pidx].cc_shadow_lock);
 	}
 
-	if (!qib_wc_pat)
-		qib_disable_wc(dd);
+	qib_disable_wc(dd);
 
 	if (dd->pioavailregs_dma) {
 		dma_free_coherent(&dd->pcidev->dev, PAGE_SIZE,
@@ -1547,14 +1537,12 @@ static int qib_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 		goto bail;
 	}
 
-	if (!qib_wc_pat) {
-		ret = qib_enable_wc(dd);
-		if (ret) {
-			qib_dev_err(dd,
-				"Write combining not enabled (err %d): performance may be poor\n",
-				-ret);
-			ret = 0;
-		}
+	ret = qib_enable_wc(dd);
+	if (ret) {
+		qib_dev_err(dd,
+			"Write combining not enabled (err %d): performance may be poor\n",
+			-ret);
+		ret = 0;
 	}
 
 	qib_verify_pioperf(dd);
diff --git a/drivers/infiniband/hw/qib/qib_wc_x86_64.c b/drivers/infiniband/hw/qib/qib_wc_x86_64.c
index 81b225f2300a..edd0ddbd4481 100644
--- a/drivers/infiniband/hw/qib/qib_wc_x86_64.c
+++ b/drivers/infiniband/hw/qib/qib_wc_x86_64.c
@@ -116,21 +116,10 @@ int qib_enable_wc(struct qib_devdata *dd)
 	}
 
 	if (!ret) {
-		int cookie;
-
-		cookie = mtrr_add(pioaddr, piolen, MTRR_TYPE_WRCOMB, 0);
-		if (cookie < 0) {
-			{
-				qib_devinfo(dd->pcidev,
-					 "mtrr_add()  WC for PIO bufs failed (%d)\n",
-					 cookie);
-				ret = -EINVAL;
-			}
-		} else {
-			dd->wc_cookie = cookie;
-			dd->wc_base = (unsigned long) pioaddr;
-			dd->wc_len = (unsigned long) piolen;
-		}
+		dd->wc_cookie = arch_phys_wc_add(pioaddr, piolen);
+		if (dd->wc_cookie < 0)
+			/* use error from routine */
+			ret = dd->wc_cookie;
 	}
 
 	return ret;
@@ -142,18 +131,7 @@ int qib_enable_wc(struct qib_devdata *dd)
  */
 void qib_disable_wc(struct qib_devdata *dd)
 {
-	if (dd->wc_cookie) {
-		int r;
-
-		r = mtrr_del(dd->wc_cookie, dd->wc_base,
-			     dd->wc_len);
-		if (r < 0)
-			qib_devinfo(dd->pcidev,
-				 "mtrr_del(%lx, %lx, %lx) failed: %d\n",
-				 dd->wc_cookie, dd->wc_base,
-				 dd->wc_len, r);
-		dd->wc_cookie = 0; /* even on failure */
-	}
+	arch_phys_wc_del(dd->wc_cookie);
 }
 
 /**