Merge tag 'for-linus-2' of git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma

Pull more rdma updates from Doug Ledford:
 "Minor updates for rxe driver"

[ Starting to do merge window pulls again - the current -git tree does
  appear to have some netfilter use-after-free issues, but I've sent
  off the report to the proper channels, and I don't want to delay merge
  window activity any more ]

* tag 'for-linus-2' of git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma:
  IB/rxe: improved debug prints & code cleanup
  rdma_rxe: Ensure rdma_rxe init occurs at correct time
  IB/rxe: Properly honor max IRD value for rd/atomic.
  IB/{rxe,core,rdmavt}: Fix kernel crash for reg MR
  IB/rxe: Fix sending out loopback packet on netdev interface.
  IB/rxe: Avoid scheduling tasklet for userspace QP

17 files changed, 190 insertions, 121 deletions

diff --git a/drivers/infiniband/sw/rdmavt/dma.c b/drivers/infiniband/sw/rdmavt/dma.c
index 33076a5eee2f..01f71caa3ac4 100644
--- a/drivers/infiniband/sw/rdmavt/dma.c
+++ b/drivers/infiniband/sw/rdmavt/dma.c
@@ -138,6 +138,21 @@ static void rvt_unmap_sg(struct ib_device *dev,
 	/* This is a stub, nothing to be done here */
 }
 
+static int rvt_map_sg_attrs(struct ib_device *dev, struct scatterlist *sgl,
+			    int nents, enum dma_data_direction direction,
+			    unsigned long attrs)
+{
+	return rvt_map_sg(dev, sgl, nents, direction);
+}
+
+static void rvt_unmap_sg_attrs(struct ib_device *dev,
+			       struct scatterlist *sg, int nents,
+			       enum dma_data_direction direction,
+			       unsigned long attrs)
+{
+	return rvt_unmap_sg(dev, sg, nents, direction);
+}
+
 static void rvt_sync_single_for_cpu(struct ib_device *dev, u64 addr,
 				    size_t size, enum dma_data_direction dir)
 {
@@ -177,6 +192,8 @@ struct ib_dma_mapping_ops rvt_default_dma_mapping_ops = {
 	.unmap_page = rvt_dma_unmap_page,
 	.map_sg = rvt_map_sg,
 	.unmap_sg = rvt_unmap_sg,
+	.map_sg_attrs = rvt_map_sg_attrs,
+	.unmap_sg_attrs = rvt_unmap_sg_attrs,
 	.sync_single_for_cpu = rvt_sync_single_for_cpu,
 	.sync_single_for_device = rvt_sync_single_for_device,
 	.alloc_coherent = rvt_dma_alloc_coherent,
diff --git a/drivers/infiniband/sw/rxe/rxe.c b/drivers/infiniband/sw/rxe/rxe.c
index ddd59270ff6d..ab6c3c25d7ff 100644
--- a/drivers/infiniband/sw/rxe/rxe.c
+++ b/drivers/infiniband/sw/rxe/rxe.c
@@ -358,38 +358,16 @@ static int __init rxe_module_init(void)
 	/* initialize slab caches for managed objects */
 	err = rxe_cache_init();
 	if (err) {
-		pr_err("rxe: unable to init object pools\n");
+		pr_err("unable to init object pools\n");
 		return err;
 	}
 
-	err = rxe_net_ipv4_init();
-	if (err) {
-		pr_err("rxe: unable to init ipv4 tunnel\n");
-		rxe_cache_exit();
-		goto exit;
-	}
-
-	err = rxe_net_ipv6_init();
-	if (err) {
-		pr_err("rxe: unable to init ipv6 tunnel\n");
-		rxe_cache_exit();
-		goto exit;
-	}
-
-	err = register_netdevice_notifier(&rxe_net_notifier);
-	if (err) {
-		pr_err("rxe: Failed to rigister netdev notifier\n");
-		goto exit;
-	}
-
-	pr_info("rxe: loaded\n");
+	err = rxe_net_init();
+	if (err)
+		return err;
 
+	pr_info("loaded\n");
 	return 0;
-
-exit:
-	rxe_release_udp_tunnel(recv_sockets.sk4);
-	rxe_release_udp_tunnel(recv_sockets.sk6);
-	return err;
 }
 
 static void __exit rxe_module_exit(void)
@@ -398,8 +376,8 @@ static void __exit rxe_module_exit(void)
 	rxe_net_exit();
 	rxe_cache_exit();
 
-	pr_info("rxe: unloaded\n");
+	pr_info("unloaded\n");
 }
 
-module_init(rxe_module_init);
+late_initcall(rxe_module_init);
 module_exit(rxe_module_exit);
diff --git a/drivers/infiniband/sw/rxe/rxe.h b/drivers/infiniband/sw/rxe/rxe.h
index 12c71c549f97..a696af81e4a5 100644
--- a/drivers/infiniband/sw/rxe/rxe.h
+++ b/drivers/infiniband/sw/rxe/rxe.h
@@ -34,6 +34,11 @@
 #ifndef RXE_H
 #define RXE_H
 
+#ifdef pr_fmt
+#undef pr_fmt
+#endif
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
 #include <linux/module.h>
 #include <linux/skbuff.h>
 #include <linux/crc32.h>
diff --git a/drivers/infiniband/sw/rxe/rxe_av.c b/drivers/infiniband/sw/rxe/rxe_av.c
index 5c9474212d4e..604f6fee96bd 100644
--- a/drivers/infiniband/sw/rxe/rxe_av.c
+++ b/drivers/infiniband/sw/rxe/rxe_av.c
@@ -39,7 +39,7 @@ int rxe_av_chk_attr(struct rxe_dev *rxe, struct ib_ah_attr *attr)
 	struct rxe_port *port;
 
 	if (attr->port_num != 1) {
-		pr_info("rxe: invalid port_num = %d\n", attr->port_num);
+		pr_info("invalid port_num = %d\n", attr->port_num);
 		return -EINVAL;
 	}
 
@@ -47,7 +47,7 @@ int rxe_av_chk_attr(struct rxe_dev *rxe, struct ib_ah_attr *attr)
 
 	if (attr->ah_flags & IB_AH_GRH) {
 		if (attr->grh.sgid_index > port->attr.gid_tbl_len) {
-			pr_info("rxe: invalid sgid index = %d\n",
+			pr_info("invalid sgid index = %d\n",
 				attr->grh.sgid_index);
 			return -EINVAL;
 		}
diff --git a/drivers/infiniband/sw/rxe/rxe_comp.c b/drivers/infiniband/sw/rxe/rxe_comp.c
index 1c59ef2c67aa..6c5e29db88e3 100644
--- a/drivers/infiniband/sw/rxe/rxe_comp.c
+++ b/drivers/infiniband/sw/rxe/rxe_comp.c
@@ -567,7 +567,8 @@ int rxe_completer(void *arg)
 	state = COMPST_GET_ACK;
 
 	while (1) {
-		pr_debug("state = %s\n", comp_state_name[state]);
+		pr_debug("qp#%d state = %s\n", qp_num(qp),
+			 comp_state_name[state]);
 		switch (state) {
 		case COMPST_GET_ACK:
 			skb = skb_dequeue(&qp->resp_pkts);
@@ -709,7 +710,8 @@ int rxe_completer(void *arg)
 					qp->comp.rnr_retry--;
 
 				qp->req.need_retry = 1;
-				pr_debug("set rnr nak timer\n");
+				pr_debug("qp#%d set rnr nak timer\n",
+					 qp_num(qp));
 				mod_timer(&qp->rnr_nak_timer,
 					  jiffies + rnrnak_jiffies(aeth_syn(pkt)
 						& ~AETH_TYPE_MASK));
diff --git a/drivers/infiniband/sw/rxe/rxe_dma.c b/drivers/infiniband/sw/rxe/rxe_dma.c
index 7634c1a81b2b..a0f8af5851ae 100644
--- a/drivers/infiniband/sw/rxe/rxe_dma.c
+++ b/drivers/infiniband/sw/rxe/rxe_dma.c
@@ -117,6 +117,21 @@ static void rxe_unmap_sg(struct ib_device *dev,
 	WARN_ON(!valid_dma_direction(direction));
 }
 
+static int rxe_map_sg_attrs(struct ib_device *dev, struct scatterlist *sgl,
+			    int nents, enum dma_data_direction direction,
+			    unsigned long attrs)
+{
+	return rxe_map_sg(dev, sgl, nents, direction);
+}
+
+static void rxe_unmap_sg_attrs(struct ib_device *dev,
+			       struct scatterlist *sg, int nents,
+			       enum dma_data_direction direction,
+			       unsigned long attrs)
+{
+	rxe_unmap_sg(dev, sg, nents, direction);
+}
+
 static void rxe_sync_single_for_cpu(struct ib_device *dev,
 				    u64 addr,
 				    size_t size, enum dma_data_direction dir)
@@ -159,6 +174,8 @@ struct ib_dma_mapping_ops rxe_dma_mapping_ops = {
 	.unmap_page		= rxe_dma_unmap_page,
 	.map_sg			= rxe_map_sg,
 	.unmap_sg		= rxe_unmap_sg,
+	.map_sg_attrs		= rxe_map_sg_attrs,
+	.unmap_sg_attrs		= rxe_unmap_sg_attrs,
 	.sync_single_for_cpu	= rxe_sync_single_for_cpu,
 	.sync_single_for_device	= rxe_sync_single_for_device,
 	.alloc_coherent		= rxe_dma_alloc_coherent,
diff --git a/drivers/infiniband/sw/rxe/rxe_loc.h b/drivers/infiniband/sw/rxe/rxe_loc.h
index 4a5484ef604f..73849a5a91b3 100644
--- a/drivers/infiniband/sw/rxe/rxe_loc.h
+++ b/drivers/infiniband/sw/rxe/rxe_loc.h
@@ -198,7 +198,7 @@ void free_rd_atomic_resource(struct rxe_qp *qp, struct resp_res *res);
 static inline void rxe_advance_resp_resource(struct rxe_qp *qp)
 {
 	qp->resp.res_head++;
-	if (unlikely(qp->resp.res_head == qp->attr.max_rd_atomic))
+	if (unlikely(qp->resp.res_head == qp->attr.max_dest_rd_atomic))
 		qp->resp.res_head = 0;
 }
 
diff --git a/drivers/infiniband/sw/rxe/rxe_mmap.c b/drivers/infiniband/sw/rxe/rxe_mmap.c
index 54b3c7c99eff..c572a4c09359 100644
--- a/drivers/infiniband/sw/rxe/rxe_mmap.c
+++ b/drivers/infiniband/sw/rxe/rxe_mmap.c
@@ -126,7 +126,7 @@ found_it:
 
 	ret = remap_vmalloc_range(vma, ip->obj, 0);
 	if (ret) {
-		pr_err("rxe: err %d from remap_vmalloc_range\n", ret);
+		pr_err("err %d from remap_vmalloc_range\n", ret);
 		goto done;
 	}
 
diff --git a/drivers/infiniband/sw/rxe/rxe_mr.c b/drivers/infiniband/sw/rxe/rxe_mr.c
index f3dab6574504..1869152f1d23 100644
--- a/drivers/infiniband/sw/rxe/rxe_mr.c
+++ b/drivers/infiniband/sw/rxe/rxe_mr.c
@@ -39,7 +39,7 @@
  */
 static u8 rxe_get_key(void)
 {
-	static unsigned key = 1;
+	static u32 key = 1;
 
 	key = key << 1;
 
diff --git a/drivers/infiniband/sw/rxe/rxe_net.c b/drivers/infiniband/sw/rxe/rxe_net.c
index eedf2f1cafdf..b8258e4f0aea 100644
--- a/drivers/infiniband/sw/rxe/rxe_net.c
+++ b/drivers/infiniband/sw/rxe/rxe_net.c
@@ -65,7 +65,7 @@ struct rxe_dev *net_to_rxe(struct net_device *ndev)
 	return found;
 }
 
-struct rxe_dev *get_rxe_by_name(const char* name)
+struct rxe_dev *get_rxe_by_name(const char *name)
 {
 	struct rxe_dev *rxe;
 	struct rxe_dev *found = NULL;
@@ -350,14 +350,14 @@ static void prepare_ipv6_hdr(struct dst_entry *dst, struct sk_buff *skb,
 	ip6h->payload_len = htons(skb->len - sizeof(*ip6h));
 }
 
-static int prepare4(struct rxe_dev *rxe, struct sk_buff *skb, struct rxe_av *av)
+static int prepare4(struct rxe_dev *rxe, struct rxe_pkt_info *pkt,
+		    struct sk_buff *skb, struct rxe_av *av)
 {
 	struct dst_entry *dst;
 	bool xnet = false;
 	__be16 df = htons(IP_DF);
 	struct in_addr *saddr = &av->sgid_addr._sockaddr_in.sin_addr;
 	struct in_addr *daddr = &av->dgid_addr._sockaddr_in.sin_addr;
-	struct rxe_pkt_info *pkt = SKB_TO_PKT(skb);
 
 	dst = rxe_find_route4(rxe->ndev, saddr, daddr);
 	if (!dst) {
@@ -376,12 +376,12 @@ static int prepare4(struct rxe_dev *rxe, struct sk_buff *skb, struct rxe_av *av)
 	return 0;
 }
 
-static int prepare6(struct rxe_dev *rxe, struct sk_buff *skb, struct rxe_av *av)
+static int prepare6(struct rxe_dev *rxe, struct rxe_pkt_info *pkt,
+		    struct sk_buff *skb, struct rxe_av *av)
 {
 	struct dst_entry *dst;
 	struct in6_addr *saddr = &av->sgid_addr._sockaddr_in6.sin6_addr;
 	struct in6_addr *daddr = &av->dgid_addr._sockaddr_in6.sin6_addr;
-	struct rxe_pkt_info *pkt = SKB_TO_PKT(skb);
 
 	dst = rxe_find_route6(rxe->ndev, saddr, daddr);
 	if (!dst) {
@@ -408,9 +408,9 @@ static int prepare(struct rxe_dev *rxe, struct rxe_pkt_info *pkt,
 	struct rxe_av *av = rxe_get_av(pkt);
 
 	if (av->network_type == RDMA_NETWORK_IPV4)
-		err = prepare4(rxe, skb, av);
+		err = prepare4(rxe, pkt, skb, av);
 	else if (av->network_type == RDMA_NETWORK_IPV6)
-		err = prepare6(rxe, skb, av);
+		err = prepare6(rxe, pkt, skb, av);
 
 	*crc = rxe_icrc_hdr(pkt, skb);
 
@@ -601,8 +601,7 @@ void rxe_port_up(struct rxe_dev *rxe)
 	port->attr.phys_state = IB_PHYS_STATE_LINK_UP;
 
 	rxe_port_event(rxe, IB_EVENT_PORT_ACTIVE);
-	pr_info("rxe: set %s active\n", rxe->ib_dev.name);
-	return;
+	pr_info("set %s active\n", rxe->ib_dev.name);
 }
 
 /* Caller must hold net_info_lock */
@@ -615,8 +614,7 @@ void rxe_port_down(struct rxe_dev *rxe)
 	port->attr.phys_state = IB_PHYS_STATE_LINK_DOWN;
 
 	rxe_port_event(rxe, IB_EVENT_PORT_ERR);
-	pr_info("rxe: set %s down\n", rxe->ib_dev.name);
-	return;
+	pr_info("set %s down\n", rxe->ib_dev.name);
 }
 
 static int rxe_notify(struct notifier_block *not_blk,
@@ -641,7 +639,7 @@ static int rxe_notify(struct notifier_block *not_blk,
 		rxe_port_down(rxe);
 		break;
 	case NETDEV_CHANGEMTU:
-		pr_info("rxe: %s changed mtu to %d\n", ndev->name, ndev->mtu);
+		pr_info("%s changed mtu to %d\n", ndev->name, ndev->mtu);
 		rxe_set_mtu(rxe, ndev->mtu);
 		break;
 	case NETDEV_REBOOT:
@@ -651,7 +649,7 @@ static int rxe_notify(struct notifier_block *not_blk,
 	case NETDEV_CHANGENAME:
 	case NETDEV_FEAT_CHANGE:
 	default:
-		pr_info("rxe: ignoring netdev event = %ld for %s\n",
+		pr_info("ignoring netdev event = %ld for %s\n",
 			event, ndev->name);
 		break;
 	}
@@ -671,7 +669,7 @@ int rxe_net_ipv4_init(void)
 				htons(ROCE_V2_UDP_DPORT), false);
 	if (IS_ERR(recv_sockets.sk4)) {
 		recv_sockets.sk4 = NULL;
-		pr_err("rxe: Failed to create IPv4 UDP tunnel\n");
+		pr_err("Failed to create IPv4 UDP tunnel\n");
 		return -1;
 	}
 
@@ -688,7 +686,7 @@ int rxe_net_ipv6_init(void)
 						htons(ROCE_V2_UDP_DPORT), true);
 	if (IS_ERR(recv_sockets.sk6)) {
 		recv_sockets.sk6 = NULL;
-		pr_err("rxe: Failed to create IPv6 UDP tunnel\n");
+		pr_err("Failed to create IPv6 UDP tunnel\n");
 		return -1;
 	}
 #endif
@@ -701,3 +699,26 @@ void rxe_net_exit(void)
 	rxe_release_udp_tunnel(recv_sockets.sk4);
 	unregister_netdevice_notifier(&rxe_net_notifier);
 }
+
+int rxe_net_init(void)
+{
+	int err;
+
+	recv_sockets.sk6 = NULL;
+
+	err = rxe_net_ipv4_init();
+	if (err)
+		return err;
+	err = rxe_net_ipv6_init();
+	if (err)
+		goto err_out;
+	err = register_netdevice_notifier(&rxe_net_notifier);
+	if (err) {
+		pr_err("Failed to register netdev notifier\n");
+		goto err_out;
+	}
+	return 0;
+err_out:
+	rxe_net_exit();
+	return err;
+}
diff --git a/drivers/infiniband/sw/rxe/rxe_net.h b/drivers/infiniband/sw/rxe/rxe_net.h
index 0daf7f09e5b5..1c06b3bfe1b6 100644
--- a/drivers/infiniband/sw/rxe/rxe_net.h
+++ b/drivers/infiniband/sw/rxe/rxe_net.h
@@ -49,8 +49,7 @@ void rxe_release_udp_tunnel(struct socket *sk);
 
 struct rxe_dev *rxe_net_add(struct net_device *ndev);
 
-int rxe_net_ipv4_init(void);
-int rxe_net_ipv6_init(void);
+int rxe_net_init(void);
 void rxe_net_exit(void);
 
 #endif /* RXE_NET_H */
diff --git a/drivers/infiniband/sw/rxe/rxe_qp.c b/drivers/infiniband/sw/rxe/rxe_qp.c
index 22ba24f2a2c1..b8036cfbce04 100644
--- a/drivers/infiniband/sw/rxe/rxe_qp.c
+++ b/drivers/infiniband/sw/rxe/rxe_qp.c
@@ -146,7 +146,7 @@ static void free_rd_atomic_resources(struct rxe_qp *qp)
 	if (qp->resp.resources) {
 		int i;
 
-		for (i = 0; i < qp->attr.max_rd_atomic; i++) {
+		for (i = 0; i < qp->attr.max_dest_rd_atomic; i++) {
 			struct resp_res *res = &qp->resp.resources[i];
 
 			free_rd_atomic_resource(qp, res);
@@ -174,7 +174,7 @@ static void cleanup_rd_atomic_resources(struct rxe_qp *qp)
 	struct resp_res *res;
 
 	if (qp->resp.resources) {
-		for (i = 0; i < qp->attr.max_rd_atomic; i++) {
+		for (i = 0; i < qp->attr.max_dest_rd_atomic; i++) {
 			res = &qp->resp.resources[i];
 			free_rd_atomic_resource(qp, res);
 		}
@@ -298,8 +298,8 @@ static int rxe_qp_init_resp(struct rxe_dev *rxe, struct rxe_qp *qp,
 
 		wqe_size = rcv_wqe_size(qp->rq.max_sge);
 
-		pr_debug("max_wr = %d, max_sge = %d, wqe_size = %d\n",
-			 qp->rq.max_wr, qp->rq.max_sge, wqe_size);
+		pr_debug("qp#%d max_wr = %d, max_sge = %d, wqe_size = %d\n",
+			 qp_num(qp), qp->rq.max_wr, qp->rq.max_sge, wqe_size);
 
 		qp->rq.queue = rxe_queue_init(rxe,
 					      &qp->rq.max_wr,
@@ -596,14 +596,21 @@ int rxe_qp_from_attr(struct rxe_qp *qp, struct ib_qp_attr *attr, int mask,
 	if (mask & IB_QP_MAX_QP_RD_ATOMIC) {
 		int max_rd_atomic = __roundup_pow_of_two(attr->max_rd_atomic);
 
+		qp->attr.max_rd_atomic = max_rd_atomic;
+		atomic_set(&qp->req.rd_atomic, max_rd_atomic);
+	}
+
+	if (mask & IB_QP_MAX_DEST_RD_ATOMIC) {
+		int max_dest_rd_atomic =
+			__roundup_pow_of_two(attr->max_dest_rd_atomic);
+
+		qp->attr.max_dest_rd_atomic = max_dest_rd_atomic;
+
 		free_rd_atomic_resources(qp);
 
-		err = alloc_rd_atomic_resources(qp, max_rd_atomic);
+		err = alloc_rd_atomic_resources(qp, max_dest_rd_atomic);
 		if (err)
 			return err;
-
-		qp->attr.max_rd_atomic = max_rd_atomic;
-		atomic_set(&qp->req.rd_atomic, max_rd_atomic);
 	}
 
 	if (mask & IB_QP_CUR_STATE)
@@ -673,24 +680,27 @@ int rxe_qp_from_attr(struct rxe_qp *qp, struct ib_qp_attr *attr, int mask,
 	if (mask & IB_QP_RETRY_CNT) {
 		qp->attr.retry_cnt = attr->retry_cnt;
 		qp->comp.retry_cnt = attr->retry_cnt;
-		pr_debug("set retry count = %d\n", attr->retry_cnt);
+		pr_debug("qp#%d set retry count = %d\n", qp_num(qp),
+			 attr->retry_cnt);
 	}
 
 	if (mask & IB_QP_RNR_RETRY) {
 		qp->attr.rnr_retry = attr->rnr_retry;
 		qp->comp.rnr_retry = attr->rnr_retry;
-		pr_debug("set rnr retry count = %d\n", attr->rnr_retry);
+		pr_debug("qp#%d set rnr retry count = %d\n", qp_num(qp),
+			 attr->rnr_retry);
 	}
 
 	if (mask & IB_QP_RQ_PSN) {
 		qp->attr.rq_psn = (attr->rq_psn & BTH_PSN_MASK);
 		qp->resp.psn = qp->attr.rq_psn;
-		pr_debug("set resp psn = 0x%x\n", qp->resp.psn);
+		pr_debug("qp#%d set resp psn = 0x%x\n", qp_num(qp),
+			 qp->resp.psn);
 	}
 
 	if (mask & IB_QP_MIN_RNR_TIMER) {
 		qp->attr.min_rnr_timer = attr->min_rnr_timer;
-		pr_debug("set min rnr timer = 0x%x\n",
+		pr_debug("qp#%d set min rnr timer = 0x%x\n", qp_num(qp),
 			 attr->min_rnr_timer);
 	}
 
@@ -698,12 +708,7 @@ int rxe_qp_from_attr(struct rxe_qp *qp, struct ib_qp_attr *attr, int mask,
 		qp->attr.sq_psn = (attr->sq_psn & BTH_PSN_MASK);
 		qp->req.psn = qp->attr.sq_psn;
 		qp->comp.psn = qp->attr.sq_psn;
-		pr_debug("set req psn = 0x%x\n", qp->req.psn);
-	}
-
-	if (mask & IB_QP_MAX_DEST_RD_ATOMIC) {
-		qp->attr.max_dest_rd_atomic =
-			__roundup_pow_of_two(attr->max_dest_rd_atomic);
+		pr_debug("qp#%d set req psn = 0x%x\n", qp_num(qp), qp->req.psn);
 	}
 
 	if (mask & IB_QP_PATH_MIG_STATE)
@@ -717,38 +722,38 @@ int rxe_qp_from_attr(struct rxe_qp *qp, struct ib_qp_attr *attr, int mask,
 
 		switch (attr->qp_state) {
 		case IB_QPS_RESET:
-			pr_debug("qp state -> RESET\n");
+			pr_debug("qp#%d state -> RESET\n", qp_num(qp));
 			rxe_qp_reset(qp);
 			break;
 
 		case IB_QPS_INIT:
-			pr_debug("qp state -> INIT\n");
+			pr_debug("qp#%d state -> INIT\n", qp_num(qp));
 			qp->req.state = QP_STATE_INIT;
 			qp->resp.state = QP_STATE_INIT;
 			break;
 
 		case IB_QPS_RTR:
-			pr_debug("qp state -> RTR\n");
+			pr_debug("qp#%d state -> RTR\n", qp_num(qp));
 			qp->resp.state = QP_STATE_READY;
 			break;
 
 		case IB_QPS_RTS:
-			pr_debug("qp state -> RTS\n");
+			pr_debug("qp#%d state -> RTS\n", qp_num(qp));
 			qp->req.state = QP_STATE_READY;
 			break;
 
 		case IB_QPS_SQD:
-			pr_debug("qp state -> SQD\n");
+			pr_debug("qp#%d state -> SQD\n", qp_num(qp));
 			rxe_qp_drain(qp);
 			break;
 
 		case IB_QPS_SQE:
-			pr_warn("qp state -> SQE !!?\n");
+			pr_warn("qp#%d state -> SQE !!?\n", qp_num(qp));
 			/* Not possible from modify_qp. */
 			break;
 
 		case IB_QPS_ERR:
-			pr_debug("qp state -> ERR\n");
+			pr_debug("qp#%d state -> ERR\n", qp_num(qp));
 			rxe_qp_error(qp);
 			break;
 		}
diff --git a/drivers/infiniband/sw/rxe/rxe_recv.c b/drivers/infiniband/sw/rxe/rxe_recv.c
index 144d2f129fcd..46f062842a9a 100644
--- a/drivers/infiniband/sw/rxe/rxe_recv.c
+++ b/drivers/infiniband/sw/rxe/rxe_recv.c
@@ -387,7 +387,8 @@ int rxe_rcv(struct sk_buff *skb)
 	pack_icrc = be32_to_cpu(*icrcp);
 
 	calc_icrc = rxe_icrc_hdr(pkt, skb);
-	calc_icrc = crc32_le(calc_icrc, (u8 *)payload_addr(pkt), payload_size(pkt));
+	calc_icrc = crc32_le(calc_icrc, (u8 *)payload_addr(pkt),
+			     payload_size(pkt));
 	calc_icrc = cpu_to_be32(~calc_icrc);
 	if (unlikely(calc_icrc != pack_icrc)) {
 		char saddr[sizeof(struct in6_addr)];
diff --git a/drivers/infiniband/sw/rxe/rxe_req.c b/drivers/infiniband/sw/rxe/rxe_req.c
index 13a848a518e8..832846b73ea0 100644
--- a/drivers/infiniband/sw/rxe/rxe_req.c
+++ b/drivers/infiniband/sw/rxe/rxe_req.c
@@ -38,7 +38,7 @@
 #include "rxe_queue.h"
 
 static int next_opcode(struct rxe_qp *qp, struct rxe_send_wqe *wqe,
-		       unsigned opcode);
+		       u32 opcode);
 
 static inline void retry_first_write_send(struct rxe_qp *qp,
 					  struct rxe_send_wqe *wqe,
@@ -121,7 +121,7 @@ void rnr_nak_timer(unsigned long data)
 {
 	struct rxe_qp *qp = (struct rxe_qp *)data;
 
-	pr_debug("rnr nak timer fired\n");
+	pr_debug("qp#%d rnr nak timer fired\n", qp_num(qp));
 	rxe_run_task(&qp->req.task, 1);
 }
 
@@ -187,7 +187,7 @@ static struct rxe_send_wqe *req_next_wqe(struct rxe_qp *qp)
 	return wqe;
 }
 
-static int next_opcode_rc(struct rxe_qp *qp, unsigned opcode, int fits)
+static int next_opcode_rc(struct rxe_qp *qp, u32 opcode, int fits)
 {
 	switch (opcode) {
 	case IB_WR_RDMA_WRITE:
@@ -259,7 +259,7 @@ static int next_opcode_rc(struct rxe_qp *qp, unsigned opcode, int fits)
 	return -EINVAL;
 }
 
-static int next_opcode_uc(struct rxe_qp *qp, unsigned opcode, int fits)
+static int next_opcode_uc(struct rxe_qp *qp, u32 opcode, int fits)
 {
 	switch (opcode) {
 	case IB_WR_RDMA_WRITE:
@@ -311,7 +311,7 @@ static int next_opcode_uc(struct rxe_qp *qp, unsigned opcode, int fits)
 }
 
 static int next_opcode(struct rxe_qp *qp, struct rxe_send_wqe *wqe,
-		       unsigned opcode)
+		       u32 opcode)
 {
 	int fits = (wqe->dma.resid <= qp->mtu);
 
@@ -588,7 +588,7 @@ int rxe_requester(void *arg)
 	struct rxe_pkt_info pkt;
 	struct sk_buff *skb;
 	struct rxe_send_wqe *wqe;
-	unsigned mask;
+	enum rxe_hdr_mask mask;
 	int payload;
 	int mtu;
 	int opcode;
@@ -626,7 +626,8 @@ next_wqe:
 			rmr = rxe_pool_get_index(&rxe->mr_pool,
 						 wqe->wr.ex.invalidate_rkey >> 8);
 			if (!rmr) {
-				pr_err("No mr for key %#x\n", wqe->wr.ex.invalidate_rkey);
+				pr_err("No mr for key %#x\n",
+				       wqe->wr.ex.invalidate_rkey);
 				wqe->state = wqe_state_error;
 				wqe->status = IB_WC_MW_BIND_ERR;
 				goto exit;
@@ -702,12 +703,12 @@ next_wqe:
 
 	skb = init_req_packet(qp, wqe, opcode, payload, &pkt);
 	if (unlikely(!skb)) {
-		pr_err("Failed allocating skb\n");
+		pr_err("qp#%d Failed allocating skb\n", qp_num(qp));
 		goto err;
 	}
 
 	if (fill_packet(qp, wqe, &pkt, skb, payload)) {
-		pr_debug("Error during fill packet\n");
+		pr_debug("qp#%d Error during fill packet\n", qp_num(qp));
 		goto err;
 	}
 
diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c
index 3e0f0f2baace..dd3d88adc003 100644
--- a/drivers/infiniband/sw/rxe/rxe_resp.c
+++ b/drivers/infiniband/sw/rxe/rxe_resp.c
@@ -383,7 +383,7 @@ static enum resp_states check_resource(struct rxe_qp *qp,
 		 * too many read/atomic ops, we just
 		 * recycle the responder resource queue
 		 */
-		if (likely(qp->attr.max_rd_atomic > 0))
+		if (likely(qp->attr.max_dest_rd_atomic > 0))
 			return RESPST_CHK_LENGTH;
 		else
 			return RESPST_ERR_TOO_MANY_RDMA_ATM_REQ;
@@ -749,6 +749,18 @@ static enum resp_states read_reply(struct rxe_qp *qp,
 	return state;
 }
 
+static void build_rdma_network_hdr(union rdma_network_hdr *hdr,
+				   struct rxe_pkt_info *pkt)
+{
+	struct sk_buff *skb = PKT_TO_SKB(pkt);
+
+	memset(hdr, 0, sizeof(*hdr));
+	if (skb->protocol == htons(ETH_P_IP))
+		memcpy(&hdr->roce4grh, ip_hdr(skb), sizeof(hdr->roce4grh));
+	else if (skb->protocol == htons(ETH_P_IPV6))
+		memcpy(&hdr->ibgrh, ipv6_hdr(skb), sizeof(hdr->ibgrh));
+}
+
 /* Executes a new request. A retried request never reach that function (send
  * and writes are discarded, and reads and atomics are retried elsewhere.
  */
@@ -761,13 +773,8 @@ static enum resp_states execute(struct rxe_qp *qp, struct rxe_pkt_info *pkt)
 		    qp_type(qp) == IB_QPT_SMI ||
 		    qp_type(qp) == IB_QPT_GSI) {
 			union rdma_network_hdr hdr;
-			struct sk_buff *skb = PKT_TO_SKB(pkt);
 
-			memset(&hdr, 0, sizeof(hdr));
-			if (skb->protocol == htons(ETH_P_IP))
-				memcpy(&hdr.roce4grh, ip_hdr(skb), sizeof(hdr.roce4grh));
-			else if (skb->protocol == htons(ETH_P_IPV6))
-				memcpy(&hdr.ibgrh, ipv6_hdr(skb), sizeof(hdr.ibgrh));
+			build_rdma_network_hdr(&hdr, pkt);
 
 			err = send_data_in(qp, &hdr, sizeof(hdr));
 			if (err)
@@ -881,7 +888,8 @@ static enum resp_states do_complete(struct rxe_qp *qp,
 				rmr = rxe_pool_get_index(&rxe->mr_pool,
 							 wc->ex.invalidate_rkey >> 8);
 				if (unlikely(!rmr)) {
-					pr_err("Bad rkey %#x invalidation\n", wc->ex.invalidate_rkey);
+					pr_err("Bad rkey %#x invalidation\n",
+					       wc->ex.invalidate_rkey);
 					return RESPST_ERROR;
 				}
 				rmr->state = RXE_MEM_STATE_FREE;
@@ -1208,7 +1216,8 @@ int rxe_responder(void *arg)
 	}
 
 	while (1) {
-		pr_debug("state = %s\n", resp_state_name[state]);
+		pr_debug("qp#%d state = %s\n", qp_num(qp),
+			 resp_state_name[state]);
 		switch (state) {
 		case RESPST_GET_REQ:
 			state = get_req(qp, &pkt);
diff --git a/drivers/infiniband/sw/rxe/rxe_sysfs.c b/drivers/infiniband/sw/rxe/rxe_sysfs.c
index cf8e77800046..d5ed7571128f 100644
--- a/drivers/infiniband/sw/rxe/rxe_sysfs.c
+++ b/drivers/infiniband/sw/rxe/rxe_sysfs.c
@@ -79,7 +79,7 @@ static int rxe_param_set_add(const char *val, const struct kernel_param *kp)
 
 	len = sanitize_arg(val, intf, sizeof(intf));
 	if (!len) {
-		pr_err("rxe: add: invalid interface name\n");
+		pr_err("add: invalid interface name\n");
 		err = -EINVAL;
 		goto err;
 	}
@@ -92,20 +92,20 @@ static int rxe_param_set_add(const char *val, const struct kernel_param *kp)
 	}
 
 	if (net_to_rxe(ndev)) {
-		pr_err("rxe: already configured on %s\n", intf);
+		pr_err("already configured on %s\n", intf);
 		err = -EINVAL;
 		goto err;
 	}
 
 	rxe = rxe_net_add(ndev);
 	if (!rxe) {
-		pr_err("rxe: failed to add %s\n", intf);
+		pr_err("failed to add %s\n", intf);
 		err = -EINVAL;
 		goto err;
 	}
 
 	rxe_set_port_state(ndev);
-	pr_info("rxe: added %s to %s\n", rxe->ib_dev.name, intf);
+	pr_info("added %s to %s\n", rxe->ib_dev.name, intf);
 err:
 	if (ndev)
 		dev_put(ndev);
@@ -120,7 +120,7 @@ static int rxe_param_set_remove(const char *val, const struct kernel_param *kp)
 
 	len = sanitize_arg(val, intf, sizeof(intf));
 	if (!len) {
-		pr_err("rxe: add: invalid interface name\n");
+		pr_err("add: invalid interface name\n");
 		return -EINVAL;
 	}
 
@@ -133,7 +133,7 @@ static int rxe_param_set_remove(const char *val, const struct kernel_param *kp)
 	rxe = get_rxe_by_name(intf);
 
 	if (!rxe) {
-		pr_err("rxe: not configured on %s\n", intf);
+		pr_err("not configured on %s\n", intf);
 		return -EINVAL;
 	}
 
diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.c b/drivers/infiniband/sw/rxe/rxe_verbs.c
index 4552be960c6a..19841c863daf 100644
--- a/drivers/infiniband/sw/rxe/rxe_verbs.c
+++ b/drivers/infiniband/sw/rxe/rxe_verbs.c
@@ -100,10 +100,12 @@ static int rxe_query_port(struct ib_device *dev,
 		rxe->ndev->ethtool_ops->get_settings(rxe->ndev, &cmd);
 		speed = cmd.speed;
 	} else {
-		pr_warn("%s speed is unknown, defaulting to 1000\n", rxe->ndev->name);
+		pr_warn("%s speed is unknown, defaulting to 1000\n",
+			rxe->ndev->name);
 		speed = 1000;
 	}
-	rxe_eth_speed_to_ib_speed(speed, &attr->active_speed, &attr->active_width);
+	rxe_eth_speed_to_ib_speed(speed, &attr->active_speed,
+				  &attr->active_width);
 	mutex_unlock(&rxe->usdev_lock);
 
 	return 0;
@@ -761,7 +763,7 @@ static int init_send_wqe(struct rxe_qp *qp, struct ib_send_wr *ibwr,
 }
 
 static int post_one_send(struct rxe_qp *qp, struct ib_send_wr *ibwr,
-			 unsigned mask, u32 length)
+			 unsigned int mask, u32 length)
 {
 	int err;
 	struct rxe_sq *sq = &qp->sq;
@@ -801,26 +803,15 @@ err1:
 	return err;
 }
 
-static int rxe_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
-			 struct ib_send_wr **bad_wr)
+static int rxe_post_send_kernel(struct rxe_qp *qp, struct ib_send_wr *wr,
+				struct ib_send_wr **bad_wr)
 {
 	int err = 0;
-	struct rxe_qp *qp = to_rqp(ibqp);
 	unsigned int mask;
 	unsigned int length = 0;
 	int i;
 	int must_sched;
 
-	if (unlikely(!qp->valid)) {
-		*bad_wr = wr;
-		return -EINVAL;
-	}
-
-	if (unlikely(qp->req.state < QP_STATE_READY)) {
-		*bad_wr = wr;
-		return -EINVAL;
-	}
-
 	while (wr) {
 		mask = wr_opcode_mask(wr->opcode, qp);
 		if (unlikely(!mask)) {
@@ -861,6 +852,29 @@ static int rxe_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 	return err;
 }
 
+static int rxe_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
+			 struct ib_send_wr **bad_wr)
+{
+	struct rxe_qp *qp = to_rqp(ibqp);
+
+	if (unlikely(!qp->valid)) {
+		*bad_wr = wr;
+		return -EINVAL;
+	}
+
+	if (unlikely(qp->req.state < QP_STATE_READY)) {
+		*bad_wr = wr;
+		return -EINVAL;
+	}
+
+	if (qp->is_user) {
+		/* Utilize process context to do protocol processing */
+		rxe_run_task(&qp->req.task, 0);
+		return 0;
+	} else
+		return rxe_post_send_kernel(qp, wr, bad_wr);
+}
+
 static int rxe_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr,
 			 struct ib_recv_wr **bad_wr)
 {
@@ -1133,8 +1147,8 @@ static int rxe_set_page(struct ib_mr *ibmr, u64 addr)
 	return 0;
 }
 
-static int rxe_map_mr_sg(struct ib_mr *ibmr, struct scatterlist *sg, int sg_nents,
-			 unsigned int *sg_offset)
+static int rxe_map_mr_sg(struct ib_mr *ibmr, struct scatterlist *sg,
+			 int sg_nents, unsigned int *sg_offset)
 {
 	struct rxe_mem *mr = to_rmr(ibmr);
 	int n;