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

Pull rdma updates from Doug Ledford:
 "Initial roundup of 4.6 merge window patches.

  This is the first of two pull requests.  It is the smaller request,
  but touches for more different things (this is everything but what is
  in or going into staging).  The pull request for the code in
  staging/rdma is on hold until after we decide what to do on the
  write/writev API issue and may be partially deferred until 4.7 as a
  result.

  Summary:

   - cxgb4 updates
   - nes updates
   - unification of iwarp portmapper code to core
   - add drain_cq API
   - various ib_core updates
   - minor ipoib updates
   - minor mlx4 updates
   - more significant mlx5 updates (including a minor merge conflict
     with net-next tree...merge is simple to resolve and Stephen's
     resolution was confirmed by Mellanox)
   - trivial net/9p rdma conversion
   - ocrdma RoCEv2 update
   - srpt updates"

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma: (85 commits)
  iwpm: crash fix for large connections test
  iw_cxgb3: support for iWARP port mapping
  iw_cxgb4: remove port mapper related code
  iw_nes: remove port mapper related code
  iwcm: common code for port mapper
  net/9p: convert to new CQ API
  IB/mlx5: Add support for don't trap rules
  net/mlx5_core: Introduce forward to next priority action
  net/mlx5_core: Create anchor of last flow table
  iser: Accept arbitrary sg lists mapping if the device supports it
  mlx5: Add arbitrary sg list support
  IB/core: Add arbitrary sg_list support
  IB/mlx5: Expose correct max_fast_reg_page_list_len
  IB/mlx5: Make coding style more consistent
  IB/mlx5: Convert UMR CQ to new CQ API
  IB/ocrdma: Skip using unneeded intermediate variable
  IB/ocrdma: Skip using unneeded intermediate variable
  IB/ocrdma: Delete unnecessary variable initialisations in 11 functions
  IB/core: Documentation fix in the MAD header file
  IB/core: trivial prink cleanup.
  ...

66 files changed, 3116 insertions, 1838 deletions

diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c
index 53343ffbff7a..cb00d59da456 100644
--- a/drivers/infiniband/core/cache.c
+++ b/drivers/infiniband/core/cache.c
@@ -1043,8 +1043,8 @@ static void ib_cache_update(struct ib_device *device,
 
 	ret = ib_query_port(device, port, tprops);
 	if (ret) {
-		printk(KERN_WARNING "ib_query_port failed (%d) for %s\n",
-		       ret, device->name);
+		pr_warn("ib_query_port failed (%d) for %s\n",
+			ret, device->name);
 		goto err;
 	}
 
@@ -1067,8 +1067,8 @@ static void ib_cache_update(struct ib_device *device,
 	for (i = 0; i < pkey_cache->table_len; ++i) {
 		ret = ib_query_pkey(device, port, i, pkey_cache->table + i);
 		if (ret) {
-			printk(KERN_WARNING "ib_query_pkey failed (%d) for %s (index %d)\n",
-			       ret, device->name, i);
+			pr_warn("ib_query_pkey failed (%d) for %s (index %d)\n",
+				ret, device->name, i);
 			goto err;
 		}
 	}
@@ -1078,8 +1078,8 @@ static void ib_cache_update(struct ib_device *device,
 			ret = ib_query_gid(device, port, i,
 					   gid_cache->table + i, NULL);
 			if (ret) {
-				printk(KERN_WARNING "ib_query_gid failed (%d) for %s (index %d)\n",
-				       ret, device->name, i);
+				pr_warn("ib_query_gid failed (%d) for %s (index %d)\n",
+					ret, device->name, i);
 				goto err;
 			}
 		}
@@ -1161,8 +1161,7 @@ int ib_cache_setup_one(struct ib_device *device)
 					  GFP_KERNEL);
 	if (!device->cache.pkey_cache ||
 	    !device->cache.lmc_cache) {
-		printk(KERN_WARNING "Couldn't allocate cache "
-		       "for %s\n", device->name);
+		pr_warn("Couldn't allocate cache for %s\n", device->name);
 		return -ENOMEM;
 	}
 
diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c
index 9729639df407..93ab0ae97208 100644
--- a/drivers/infiniband/core/cma.c
+++ b/drivers/infiniband/core/cma.c
@@ -1206,6 +1206,10 @@ static int cma_save_req_info(const struct ib_cm_event *ib_event,
 		req->has_gid	= true;
 		req->service_id	= req_param->primary_path->service_id;
 		req->pkey	= be16_to_cpu(req_param->primary_path->pkey);
+		if (req->pkey != req_param->bth_pkey)
+			pr_warn_ratelimited("RDMA CMA: got different BTH P_Key (0x%x) and primary path P_Key (0x%x)\n"
+					    "RDMA CMA: in the future this may cause the request to be dropped\n",
+					    req_param->bth_pkey, req->pkey);
 		break;
 	case IB_CM_SIDR_REQ_RECEIVED:
 		req->device	= sidr_param->listen_id->device;
@@ -1213,6 +1217,10 @@ static int cma_save_req_info(const struct ib_cm_event *ib_event,
 		req->has_gid	= false;
 		req->service_id	= sidr_param->service_id;
 		req->pkey	= sidr_param->pkey;
+		if (req->pkey != sidr_param->bth_pkey)
+			pr_warn_ratelimited("RDMA CMA: got different BTH P_Key (0x%x) and SIDR request payload P_Key (0x%x)\n"
+					    "RDMA CMA: in the future this may cause the request to be dropped\n",
+					    sidr_param->bth_pkey, req->pkey);
 		break;
 	default:
 		return -EINVAL;
@@ -1713,7 +1721,7 @@ static int cma_ib_handler(struct ib_cm_id *cm_id, struct ib_cm_event *ib_event)
 		event.param.conn.private_data_len = IB_CM_REJ_PRIVATE_DATA_SIZE;
 		break;
 	default:
-		printk(KERN_ERR "RDMA CMA: unexpected IB CM event: %d\n",
+		pr_err("RDMA CMA: unexpected IB CM event: %d\n",
 		       ib_event->event);
 		goto out;
 	}
@@ -2186,8 +2194,8 @@ static void cma_listen_on_dev(struct rdma_id_private *id_priv,
 
 	ret = rdma_listen(id, id_priv->backlog);
 	if (ret)
-		printk(KERN_WARNING "RDMA CMA: cma_listen_on_dev, error %d, "
-		       "listening on device %s\n", ret, cma_dev->device->name);
+		pr_warn("RDMA CMA: cma_listen_on_dev, error %d, listening on device %s\n",
+			ret, cma_dev->device->name);
 }
 
 static void cma_listen_on_all(struct rdma_id_private *id_priv)
@@ -3239,7 +3247,7 @@ static int cma_sidr_rep_handler(struct ib_cm_id *cm_id,
 		event.status = 0;
 		break;
 	default:
-		printk(KERN_ERR "RDMA CMA: unexpected IB CM event: %d\n",
+		pr_err("RDMA CMA: unexpected IB CM event: %d\n",
 		       ib_event->event);
 		goto out;
 	}
@@ -4003,8 +4011,8 @@ static int cma_netdev_change(struct net_device *ndev, struct rdma_id_private *id
 	if ((dev_addr->bound_dev_if == ndev->ifindex) &&
 	    (net_eq(dev_net(ndev), dev_addr->net)) &&
 	    memcmp(dev_addr->src_dev_addr, ndev->dev_addr, ndev->addr_len)) {
-		printk(KERN_INFO "RDMA CM addr change for ndev %s used by id %p\n",
-		       ndev->name, &id_priv->id);
+		pr_info("RDMA CM addr change for ndev %s used by id %p\n",
+			ndev->name, &id_priv->id);
 		work = kzalloc(sizeof *work, GFP_KERNEL);
 		if (!work)
 			return -ENOMEM;
@@ -4287,7 +4295,7 @@ static int __init cma_init(void)
 		goto err;
 
 	if (ibnl_add_client(RDMA_NL_RDMA_CM, RDMA_NL_RDMA_CM_NUM_OPS, cma_cb_table))
-		printk(KERN_WARNING "RDMA CMA: failed to add netlink callback\n");
+		pr_warn("RDMA CMA: failed to add netlink callback\n");
 	cma_configfs_init();
 
 	return 0;
diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c
index 94b80a51ab68..270c7ff6cba7 100644
--- a/drivers/infiniband/core/device.c
+++ b/drivers/infiniband/core/device.c
@@ -115,8 +115,8 @@ static int ib_device_check_mandatory(struct ib_device *device)
 
 	for (i = 0; i < ARRAY_SIZE(mandatory_table); ++i) {
 		if (!*(void **) ((void *) device + mandatory_table[i].offset)) {
-			printk(KERN_WARNING "Device %s is missing mandatory function %s\n",
-			       device->name, mandatory_table[i].name);
+			pr_warn("Device %s is missing mandatory function %s\n",
+				device->name, mandatory_table[i].name);
 			return -EINVAL;
 		}
 	}
@@ -255,8 +255,8 @@ static int add_client_context(struct ib_device *device, struct ib_client *client
 
 	context = kmalloc(sizeof *context, GFP_KERNEL);
 	if (!context) {
-		printk(KERN_WARNING "Couldn't allocate client context for %s/%s\n",
-		       device->name, client->name);
+		pr_warn("Couldn't allocate client context for %s/%s\n",
+			device->name, client->name);
 		return -ENOMEM;
 	}
 
@@ -343,29 +343,29 @@ int ib_register_device(struct ib_device *device,
 
 	ret = read_port_immutable(device);
 	if (ret) {
-		printk(KERN_WARNING "Couldn't create per port immutable data %s\n",
-		       device->name);
+		pr_warn("Couldn't create per port immutable data %s\n",
+			device->name);
 		goto out;
 	}
 
 	ret = ib_cache_setup_one(device);
 	if (ret) {
-		printk(KERN_WARNING "Couldn't set up InfiniBand P_Key/GID cache\n");
+		pr_warn("Couldn't set up InfiniBand P_Key/GID cache\n");
 		goto out;
 	}
 
 	memset(&device->attrs, 0, sizeof(device->attrs));
 	ret = device->query_device(device, &device->attrs, &uhw);
 	if (ret) {
-		printk(KERN_WARNING "Couldn't query the device attributes\n");
+		pr_warn("Couldn't query the device attributes\n");
 		ib_cache_cleanup_one(device);
 		goto out;
 	}
 
 	ret = ib_device_register_sysfs(device, port_callback);
 	if (ret) {
-		printk(KERN_WARNING "Couldn't register device %s with driver model\n",
-		       device->name);
+		pr_warn("Couldn't register device %s with driver model\n",
+			device->name);
 		ib_cache_cleanup_one(device);
 		goto out;
 	}
@@ -566,8 +566,8 @@ void ib_set_client_data(struct ib_device *device, struct ib_client *client,
 			goto out;
 		}
 
-	printk(KERN_WARNING "No client context found for %s/%s\n",
-	       device->name, client->name);
+	pr_warn("No client context found for %s/%s\n",
+		device->name, client->name);
 
 out:
 	spin_unlock_irqrestore(&device->client_data_lock, flags);
@@ -960,13 +960,13 @@ static int __init ib_core_init(void)
 
 	ret = class_register(&ib_class);
 	if (ret) {
-		printk(KERN_WARNING "Couldn't create InfiniBand device class\n");
+		pr_warn("Couldn't create InfiniBand device class\n");
 		goto err_comp;
 	}
 
 	ret = ibnl_init();
 	if (ret) {
-		printk(KERN_WARNING "Couldn't init IB netlink interface\n");
+		pr_warn("Couldn't init IB netlink interface\n");
 		goto err_sysfs;
 	}
 
diff --git a/drivers/infiniband/core/fmr_pool.c b/drivers/infiniband/core/fmr_pool.c
index 6ac3683c144b..cdbb1f1a6d97 100644
--- a/drivers/infiniband/core/fmr_pool.c
+++ b/drivers/infiniband/core/fmr_pool.c
@@ -150,8 +150,8 @@ static void ib_fmr_batch_release(struct ib_fmr_pool *pool)
 
 #ifdef DEBUG
 		if (fmr->ref_count !=0) {
-			printk(KERN_WARNING PFX "Unmapping FMR 0x%08x with ref count %d\n",
-			       fmr, fmr->ref_count);
+			pr_warn(PFX "Unmapping FMR 0x%08x with ref count %d\n",
+				fmr, fmr->ref_count);
 		}
 #endif
 	}
@@ -167,7 +167,7 @@ static void ib_fmr_batch_release(struct ib_fmr_pool *pool)
 
 	ret = ib_unmap_fmr(&fmr_list);
 	if (ret)
-		printk(KERN_WARNING PFX "ib_unmap_fmr returned %d\n", ret);
+		pr_warn(PFX "ib_unmap_fmr returned %d\n", ret);
 
 	spin_lock_irq(&pool->pool_lock);
 	list_splice(&unmap_list, &pool->free_list);
@@ -222,8 +222,7 @@ struct ib_fmr_pool *ib_create_fmr_pool(struct ib_pd             *pd,
 	device = pd->device;
 	if (!device->alloc_fmr    || !device->dealloc_fmr  ||
 	    !device->map_phys_fmr || !device->unmap_fmr) {
-		printk(KERN_INFO PFX "Device %s does not support FMRs\n",
-		       device->name);
+		pr_info(PFX "Device %s does not support FMRs\n", device->name);
 		return ERR_PTR(-ENOSYS);
 	}
 
@@ -233,13 +232,10 @@ struct ib_fmr_pool *ib_create_fmr_pool(struct ib_pd             *pd,
 		max_remaps = device->attrs.max_map_per_fmr;
 
 	pool = kmalloc(sizeof *pool, GFP_KERNEL);
-	if (!pool) {
-		printk(KERN_WARNING PFX "couldn't allocate pool struct\n");
+	if (!pool)
 		return ERR_PTR(-ENOMEM);
-	}
 
 	pool->cache_bucket   = NULL;
-
 	pool->flush_function = params->flush_function;
 	pool->flush_arg      = params->flush_arg;
 
@@ -251,7 +247,7 @@ struct ib_fmr_pool *ib_create_fmr_pool(struct ib_pd             *pd,
 			kmalloc(IB_FMR_HASH_SIZE * sizeof *pool->cache_bucket,
 				GFP_KERNEL);
 		if (!pool->cache_bucket) {
-			printk(KERN_WARNING PFX "Failed to allocate cache in pool\n");
+			pr_warn(PFX "Failed to allocate cache in pool\n");
 			ret = -ENOMEM;
 			goto out_free_pool;
 		}
@@ -275,7 +271,7 @@ struct ib_fmr_pool *ib_create_fmr_pool(struct ib_pd             *pd,
 				   "ib_fmr(%s)",
 				   device->name);
 	if (IS_ERR(pool->thread)) {
-		printk(KERN_WARNING PFX "couldn't start cleanup thread\n");
+		pr_warn(PFX "couldn't start cleanup thread\n");
 		ret = PTR_ERR(pool->thread);
 		goto out_free_pool;
 	}
@@ -294,11 +290,8 @@ struct ib_fmr_pool *ib_create_fmr_pool(struct ib_pd             *pd,
 
 		for (i = 0; i < params->pool_size; ++i) {
 			fmr = kmalloc(bytes_per_fmr, GFP_KERNEL);
-			if (!fmr) {
-				printk(KERN_WARNING PFX "failed to allocate fmr "
-				       "struct for FMR %d\n", i);
+			if (!fmr)
 				goto out_fail;
-			}
 
 			fmr->pool             = pool;
 			fmr->remap_count      = 0;
@@ -307,8 +300,8 @@ struct ib_fmr_pool *ib_create_fmr_pool(struct ib_pd             *pd,
 
 			fmr->fmr = ib_alloc_fmr(pd, params->access, &fmr_attr);
 			if (IS_ERR(fmr->fmr)) {
-				printk(KERN_WARNING PFX "fmr_create failed "
-				       "for FMR %d\n", i);
+				pr_warn(PFX "fmr_create failed for FMR %d\n",
+					i);
 				kfree(fmr);
 				goto out_fail;
 			}
@@ -363,8 +356,8 @@ void ib_destroy_fmr_pool(struct ib_fmr_pool *pool)
 	}
 
 	if (i < pool->pool_size)
-		printk(KERN_WARNING PFX "pool still has %d regions registered\n",
-		       pool->pool_size - i);
+		pr_warn(PFX "pool still has %d regions registered\n",
+			pool->pool_size - i);
 
 	kfree(pool->cache_bucket);
 	kfree(pool);
@@ -463,7 +456,7 @@ struct ib_pool_fmr *ib_fmr_pool_map_phys(struct ib_fmr_pool *pool_handle,
 		list_add(&fmr->list, &pool->free_list);
 		spin_unlock_irqrestore(&pool->pool_lock, flags);
 
-		printk(KERN_WARNING PFX "fmr_map returns %d\n", result);
+		pr_warn(PFX "fmr_map returns %d\n", result);
 
 		return ERR_PTR(result);
 	}
@@ -517,8 +510,8 @@ int ib_fmr_pool_unmap(struct ib_pool_fmr *fmr)
 
 #ifdef DEBUG
 	if (fmr->ref_count < 0)
-		printk(KERN_WARNING PFX "FMR %p has ref count %d < 0\n",
-		       fmr, fmr->ref_count);
+		pr_warn(PFX "FMR %p has ref count %d < 0\n",
+			fmr, fmr->ref_count);
 #endif
 
 	spin_unlock_irqrestore(&pool->pool_lock, flags);
diff --git a/drivers/infiniband/core/iwcm.c b/drivers/infiniband/core/iwcm.c
index ff9163dc1596..e28a160cdab0 100644
--- a/drivers/infiniband/core/iwcm.c
+++ b/drivers/infiniband/core/iwcm.c
@@ -50,6 +50,8 @@
 
 #include <rdma/iw_cm.h>
 #include <rdma/ib_addr.h>
+#include <rdma/iw_portmap.h>
+#include <rdma/rdma_netlink.h>
 
 #include "iwcm.h"
 
@@ -57,6 +59,16 @@ MODULE_AUTHOR("Tom Tucker");
 MODULE_DESCRIPTION("iWARP CM");
 MODULE_LICENSE("Dual BSD/GPL");
 
+static struct ibnl_client_cbs iwcm_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}
+};
+
 static struct workqueue_struct *iwcm_wq;
 struct iwcm_work {
 	struct work_struct work;
@@ -402,6 +414,11 @@ static void destroy_cm_id(struct iw_cm_id *cm_id)
 	}
 	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
 
+	if (cm_id->mapped) {
+		iwpm_remove_mapinfo(&cm_id->local_addr, &cm_id->m_local_addr);
+		iwpm_remove_mapping(&cm_id->local_addr, RDMA_NL_IWCM);
+	}
+
 	(void)iwcm_deref_id(cm_id_priv);
 }
 
@@ -426,6 +443,97 @@ void iw_destroy_cm_id(struct iw_cm_id *cm_id)
 }
 EXPORT_SYMBOL(iw_destroy_cm_id);
 
+/**
+ * iw_cm_check_wildcard - If IP address is 0 then use original
+ * @pm_addr: sockaddr containing the ip to check for wildcard
+ * @cm_addr: sockaddr containing the actual IP address
+ * @cm_outaddr: sockaddr to set IP addr which leaving port
+ *
+ *  Checks the pm_addr for wildcard and then sets cm_outaddr's
+ *  IP to the actual (cm_addr).
+ */
+static void iw_cm_check_wildcard(struct sockaddr_storage *pm_addr,
+				 struct sockaddr_storage *cm_addr,
+				 struct sockaddr_storage *cm_outaddr)
+{
+	if (pm_addr->ss_family == AF_INET) {
+		struct sockaddr_in *pm4_addr = (struct sockaddr_in *)pm_addr;
+
+		if (pm4_addr->sin_addr.s_addr == INADDR_ANY) {
+			struct sockaddr_in *cm4_addr =
+				(struct sockaddr_in *)cm_addr;
+			struct sockaddr_in *cm4_outaddr =
+				(struct sockaddr_in *)cm_outaddr;
+
+			cm4_outaddr->sin_addr = cm4_addr->sin_addr;
+		}
+	} else {
+		struct sockaddr_in6 *pm6_addr = (struct sockaddr_in6 *)pm_addr;
+
+		if (ipv6_addr_type(&pm6_addr->sin6_addr) == IPV6_ADDR_ANY) {
+			struct sockaddr_in6 *cm6_addr =
+				(struct sockaddr_in6 *)cm_addr;
+			struct sockaddr_in6 *cm6_outaddr =
+				(struct sockaddr_in6 *)cm_outaddr;
+
+			cm6_outaddr->sin6_addr = cm6_addr->sin6_addr;
+		}
+	}
+}
+
+/**
+ * iw_cm_map - Use portmapper to map the ports
+ * @cm_id: connection manager pointer
+ * @active: Indicates the active side when true
+ * returns nonzero for error only if iwpm_create_mapinfo() fails
+ *
+ * Tries to add a mapping for a port using the Portmapper. If
+ * successful in mapping the IP/Port it will check the remote
+ * mapped IP address for a wildcard IP address and replace the
+ * zero IP address with the remote_addr.
+ */
+static int iw_cm_map(struct iw_cm_id *cm_id, bool active)
+{
+	struct iwpm_dev_data pm_reg_msg;
+	struct iwpm_sa_data pm_msg;
+	int status;
+
+	cm_id->m_local_addr = cm_id->local_addr;
+	cm_id->m_remote_addr = cm_id->remote_addr;
+
+	memcpy(pm_reg_msg.dev_name, cm_id->device->name,
+	       sizeof(pm_reg_msg.dev_name));
+	memcpy(pm_reg_msg.if_name, cm_id->device->iwcm->ifname,
+	       sizeof(pm_reg_msg.if_name));
+
+	if (iwpm_register_pid(&pm_reg_msg, RDMA_NL_IWCM) ||
+	    !iwpm_valid_pid())
+		return 0;
+
+	cm_id->mapped = true;
+	pm_msg.loc_addr = cm_id->local_addr;
+	pm_msg.rem_addr = cm_id->remote_addr;
+	if (active)
+		status = iwpm_add_and_query_mapping(&pm_msg,
+						    RDMA_NL_IWCM);
+	else
+		status = iwpm_add_mapping(&pm_msg, RDMA_NL_IWCM);
+
+	if (!status) {
+		cm_id->m_local_addr = pm_msg.mapped_loc_addr;
+		if (active) {
+			cm_id->m_remote_addr = pm_msg.mapped_rem_addr;
+			iw_cm_check_wildcard(&pm_msg.mapped_rem_addr,
+					     &cm_id->remote_addr,
+					     &cm_id->m_remote_addr);
+		}
+	}
+
+	return iwpm_create_mapinfo(&cm_id->local_addr,
+				   &cm_id->m_local_addr,
+				   RDMA_NL_IWCM);
+}
+
 /*
  * CM_ID <-- LISTEN
  *
@@ -452,7 +560,9 @@ int iw_cm_listen(struct iw_cm_id *cm_id, int backlog)
 	case IW_CM_STATE_IDLE:
 		cm_id_priv->state = IW_CM_STATE_LISTEN;
 		spin_unlock_irqrestore(&cm_id_priv->lock, flags);
-		ret = cm_id->device->iwcm->create_listen(cm_id, backlog);
+		ret = iw_cm_map(cm_id, false);
+		if (!ret)
+			ret = cm_id->device->iwcm->create_listen(cm_id, backlog);
 		if (ret)
 			cm_id_priv->state = IW_CM_STATE_IDLE;