Merge branches 'i40iw', 'sriov' and 'hfi1' into k.o/for-4.6

Documentation/infiniband/sysfs.txt

@@ -78,9 +78,10 @@ HFI1
    chip_reset - diagnostic (root only)
    boardversion - board version
    ports/1/
-          CMgtA/
+          CCMgtA/
                cc_settings_bin - CCA tables used by PSM2
                cc_table_bin
+               cc_prescan - enable prescaning for faster BECN response
           sc2v/ - 32 files (0 - 31) used to translate sl->vl
           sl2sc/ - 32 files (0 - 31) used to translate sl->sc
           vl2mtu/ - 16 (0 - 15) files used to determine MTU for vl

MAINTAINERS

@@ -9085,6 +9085,12 @@ L:	rds-devel@oss.oracle.com (moderated for non-subscribers)
 S:	Supported
 F:	net/rds/
 
+RDMAVT - RDMA verbs software
+M:	Dennis Dalessandro <dennis.dalessandro@intel.com>
+L:	linux-rdma@vger.kernel.org
+S:	Supported
+F:	drivers/infiniband/sw/rdmavt
+
 READ-COPY UPDATE (RCU)
 M:	"Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
 M:	Josh Triplett <josh@joshtriplett.org>

drivers/infiniband/Kconfig

@@ -83,4 +83,6 @@ source "drivers/infiniband/ulp/srpt/Kconfig"
 source "drivers/infiniband/ulp/iser/Kconfig"
 source "drivers/infiniband/ulp/isert/Kconfig"
 
+source "drivers/infiniband/sw/rdmavt/Kconfig"
+
 endif # INFINIBAND

drivers/infiniband/Makefile

@@ -1,3 +1,4 @@
 obj-$(CONFIG_INFINIBAND)		+= core/
 obj-$(CONFIG_INFINIBAND)		+= hw/
 obj-$(CONFIG_INFINIBAND)		+= ulp/
+obj-$(CONFIG_INFINIBAND)		+= sw/

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",
+		pr_warn("ib_query_port failed (%d) for %s\n",
-		       ret, device->name);
+			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",
+			pr_warn("ib_query_pkey failed (%d) for %s (index %d)\n",
-			       ret, device->name, i);
+				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",
+				pr_warn("ib_query_gid failed (%d) for %s (index %d)\n",
-				       ret, device->name, i);
+					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 "
+		pr_warn("Couldn't allocate cache for %s\n", device->name);
-		       "for %s\n", device->name);
 		return -ENOMEM;
 	}
 

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, "
+		pr_warn("RDMA CMA: cma_listen_on_dev, error %d, listening on device %s\n",
-		       "listening on device %s\n", ret, cma_dev->device->name);
+			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",
+		pr_info("RDMA CM addr change for ndev %s used by id %p\n",
-		       ndev->name, &id_priv->id);
+			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;

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",
+			pr_warn("Device %s is missing mandatory function %s\n",
-			       device->name, mandatory_table[i].name);
+				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",
+		pr_warn("Couldn't allocate client context for %s/%s\n",
-		       device->name, client->name);
+			device->name, client->name);
 		return -ENOMEM;
 	}
 
@@ -343,28 +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",
+		pr_warn("Couldn't create per port immutable data %s\n",
-		       device->name);
+			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",
+		pr_warn("Couldn't register device %s with driver model\n",
-		       device->name);
+			device->name);
 		ib_cache_cleanup_one(device);
 		goto out;
 	}
@@ -565,8 +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",
+	pr_warn("No client context found for %s/%s\n",
-	       device->name, client->name);
+		device->name, client->name);
 
 out:
 	spin_unlock_irqrestore(&device->client_data_lock, flags);
@@ -649,10 +650,23 @@ int ib_query_port(struct ib_device *device,
 		  u8 port_num,
 		  struct ib_port_attr *port_attr)
 {
+	union ib_gid gid;
+	int err;
+
 	if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device))
 		return -EINVAL;
 
-	return device->query_port(device, port_num, port_attr);
+	memset(port_attr, 0, sizeof(*port_attr));
+	err = device->query_port(device, port_num, port_attr);
+	if (err || port_attr->subnet_prefix)
+		return err;
+
+	err = ib_query_gid(device, port_num, 0, &gid, NULL);
+	if (err)
+		return err;
+
+	port_attr->subnet_prefix = be64_to_cpu(gid.global.subnet_prefix);
+	return 0;
 }
 EXPORT_SYMBOL(ib_query_port);
 
@@ -959,13 +973,13 @@ 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;
 	}
 

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",
+			pr_warn(PFX "Unmapping FMR 0x%08x with ref count %d\n",
-			       fmr, fmr->ref_count);
+				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",
+		pr_info(PFX "Device %s does not support FMRs\n", device->name);
-		       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) {
+	if (!pool)
-		printk(KERN_WARNING PFX "couldn't allocate pool struct\n");
 		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) {
+			if (!fmr)
-				printk(KERN_WARNING PFX "failed to allocate fmr "
-				       "struct for FMR %d\n", i);
 				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 "
+				pr_warn(PFX "fmr_create failed for FMR %d\n",
-				       "for FMR %d\n", i);
+					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",
+		pr_warn(PFX "pool still has %d regions registered\n",
-		       pool->pool_size - i);
+			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",
+		pr_warn(PFX "FMR %p has ref count %d < 0\n",
-		       fmr, fmr->ref_count);
+			fmr, fmr->ref_count);
 #endif
 
 	spin_unlock_irqrestore(&pool->pool_lock, flags);

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;
 		spin_lock_irqsave(&cm_id_priv->lock, flags);
@@ -582,39 +692,37 @@ int iw_cm_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *iw_param)
 	spin_lock_irqsave(&cm_id_priv->lock, flags);
 
 	if (cm_id_priv->state != IW_CM_STATE_IDLE) {
-		spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+		ret = -EINVAL;
-		clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
+		goto err;
-		wake_up_all(&cm_id_priv->connect_wait);
-		return -EINVAL;
 	}
 
 	/* Get the ib_qp given the QPN */
 	qp = cm_id->device->iwcm->get_qp(cm_id->device, iw_param->qpn);
 	if (!qp) {
-		spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+		ret = -EINVAL;
-		clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
+		goto err;
-		wake_up_all(&cm_id_priv->connect_wait);
-		return -EINVAL;
 	}
 	cm_id->device->iwcm->add_ref(qp);
 	cm_id_priv->qp = qp;
 	cm_id_priv->state = IW_CM_STATE_CONN_SENT;
 	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
 
-	ret = cm_id->device->iwcm->connect(cm_id, iw_param);
+	ret = iw_cm_map(cm_id, true);
-	if (ret) {
+	if (!ret)
-		spin_lock_irqsave(&cm_id_priv->lock, flags);
+		ret = cm_id->device->iwcm->connect(cm_id, iw_param);
-		if (cm_id_priv->qp) {
+	if (!ret)
-			cm_id->device->iwcm->rem_ref(qp);
+		return 0;	/* success */
-			cm_id_priv->qp = NULL;
-		}
-		spin_unlock_irqrestore(&cm_id_priv->lock, flags);
-		BUG_ON(cm_id_priv->state != IW_CM_STATE_CONN_SENT);
-		cm_id_priv->state = IW_CM_STATE_IDLE;
-		clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
-		wake_up_all(&cm_id_priv->connect_wait);
-	}
 
+	spin_lock_irqsave(&cm_id_priv->lock, flags);
+	if (cm_id_priv->qp) {
+		cm_id->device->iwcm->rem_ref(qp);
+		cm_id_priv->qp = NULL;
+	}
+	cm_id_priv->state = IW_CM_STATE_IDLE;
+err:
+	spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+	clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
+	wake_up_all(&cm_id_priv->connect_wait);
 	return ret;
 }
 EXPORT_SYMBOL(iw_cm_connect);
@@ -656,8 +764,23 @@ static void cm_conn_req_handler(struct iwcm_id_private *listen_id_priv,
 		goto out;
 
 	cm_id->provider_data = iw_event->provider_data;
-	cm_id->local_addr = iw_event->local_addr;
+	cm_id->m_local_addr = iw_event->local_addr;
-	cm_id->remote_addr = iw_event->remote_addr;
+	cm_id->m_remote_addr = iw_event->remote_addr;
+	cm_id->local_addr = listen_id_priv->id.local_addr;
+
+	ret = iwpm_get_remote_info(&listen_id_priv->id.m_local_addr,
+				   &iw_event->remote_addr,
+				   &cm_id->remote_addr,
+				   RDMA_NL_IWCM);
+	if (ret) {
+		cm_id->remote_addr = iw_event->remote_addr;
+	} else {
+		iw_cm_check_wildcard(&listen_id_priv->id.m_local_addr,
+				     &iw_event->local_addr,
+				     &cm_id->local_addr);
+		iw_event->local_addr = cm_id->local_addr;
+		iw_event->remote_addr = cm_id->remote_addr;
+	}
 
 	cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
 	cm_id_priv->state = IW_CM_STATE_CONN_RECV;
@@ -753,8 +876,10 @@ static int cm_conn_rep_handler(struct iwcm_id_private *cm_id_priv,
 	clear_bit(IWCM_F_CONNECT_WAIT, &cm_id_priv->flags);
 	BUG_ON(cm_id_priv->state != IW_CM_STATE_CONN_SENT);
 	if (iw_event->status == 0) {
-		cm_id_priv->id.local_addr = iw_event->local_addr;
+		cm_id_priv->id.m_local_addr = iw_event->local_addr;
-		cm_id_priv->id.remote_addr = iw_event->remote_addr;
+		cm_id_priv->id.m_remote_addr = iw_event->remote_addr;
+		iw_event->local_addr = cm_id_priv->id.local_addr;
+		iw_event->remote_addr = cm_id_priv->id.remote_addr;
 		cm_id_priv->state = IW_CM_STATE_ESTABLISHED;
 	} else {
 		/* REJECTED or RESET */
@@ -1044,6 +1169,17 @@ EXPORT_SYMBOL(iw_cm_init_qp_attr);
 
 static int __init iw_cm_init(void)
 {
+	int ret;
+
+	ret = iwpm_init(RDMA_NL_IWCM);
+	if (ret)
+		pr_err("iw_cm: couldn't init iwpm\n");
+
+	ret = ibnl_add_client(RDMA_NL_IWCM, RDMA_NL_IWPM_NUM_OPS,
+			      iwcm_nl_cb_table);
+	if (ret)
+		pr_err("iw_cm: couldn't register netlink callbacks\n");
+
 	iwcm_wq = create_singlethread_workqueue("iw_cm_wq");
 	if (!iwcm_wq)
 		return -ENOMEM;
@@ -1063,6 +1199,8 @@ static void __exit iw_cm_cleanup(void)
 {
 	unregister_net_sysctl_table(iwcm_ctl_table_hdr);
 	destroy_workqueue(iwcm_wq);
+	ibnl_remove_client(RDMA_NL_IWCM);
+	iwpm_exit(RDMA_NL_IWCM);
 }
 
 module_init(iw_cm_init);

drivers/infiniband/core/iwpm_msg.c

@@ -88,8 +88,8 @@ int iwpm_register_pid(struct iwpm_dev_data *pm_msg, u8 nl_client)
 	ret = ibnl_put_attr(skb, nlh, sizeof(u32), &msg_seq, IWPM_NLA_REG_PID_SEQ);
 	if (ret)
 		goto pid_query_error;
-	ret = ibnl_put_attr(skb, nlh, IWPM_IFNAME_SIZE,
+	ret = ibnl_put_attr(skb, nlh, IFNAMSIZ,
-				pm_msg->if_name, IWPM_NLA_REG_IF_NAME);
+			    pm_msg->if_name, IWPM_NLA_REG_IF_NAME);
 	if (ret)
 		goto pid_query_error;
 	ret = ibnl_put_attr(skb, nlh, IWPM_DEVNAME_SIZE,
@@ -394,7 +394,7 @@ register_pid_response_exit:
 	/* always for found nlmsg_request */
 	kref_put(&nlmsg_request->kref, iwpm_free_nlmsg_request);
 	barrier();
-	wake_up(&nlmsg_request->waitq);
+	up(&nlmsg_request->sem);
 	return 0;
 }
 EXPORT_SYMBOL(iwpm_register_pid_cb);
@@ -463,7 +463,7 @@ add_mapping_response_exit:
 	/* always for found request */
 	kref_put(&nlmsg_request->kref, iwpm_free_nlmsg_request);
 	barrier();
-	wake_up(&nlmsg_request->waitq);
+	up(&nlmsg_request->sem);
 	return 0;
 }
 EXPORT_SYMBOL(iwpm_add_mapping_cb);
@@ -555,7 +555,7 @@ query_mapping_response_exit:
 	/* always for found request */
 	kref_put(&nlmsg_request->kref, iwpm_free_nlmsg_request);
 	barrier();
-	wake_up(&nlmsg_request->waitq);
+	up(&nlmsg_request->sem);
 	return 0;
 }
 EXPORT_SYMBOL(iwpm_add_and_query_mapping_cb);
@@ -749,7 +749,7 @@ int iwpm_mapping_error_cb(struct sk_buff *skb, struct netlink_callback *cb)
 	/* always for found request */
 	kref_put(&nlmsg_request->kref, iwpm_free_nlmsg_request);
 	barrier();
-	wake_up(&nlmsg_request->waitq);
+	up(&nlmsg_request->sem);
 	return 0;
 }
 EXPORT_SYMBOL(iwpm_mapping_error_cb);

drivers/infiniband/core/iwpm_util.c

@@ -254,9 +254,9 @@ void iwpm_add_remote_info(struct iwpm_remote_info *rem_info)
 }
 
 int iwpm_get_remote_info(struct sockaddr_storage *mapped_loc_addr,
-				struct sockaddr_storage *mapped_rem_addr,
+			 struct sockaddr_storage *mapped_rem_addr,
-				struct sockaddr_storage *remote_addr,
+			 struct sockaddr_storage *remote_addr,
-				u8 nl_client)
+			 u8 nl_client)
 {
 	struct hlist_node *tmp_hlist_node;
 	struct hlist_head *hash_bucket_head;
@@ -322,6 +322,8 @@ struct iwpm_nlmsg_request *iwpm_get_nlmsg_request(__u32 nlmsg_seq,
 	nlmsg_request->nl_client = nl_client;
 	nlmsg_request->request_done = 0;
 	nlmsg_request->err_code = 0;
+	sema_init(&nlmsg_request->sem, 1);
+	down(&nlmsg_request->sem);
 	return nlmsg_request;
 }
 
@@ -364,11 +366,9 @@ struct iwpm_nlmsg_request *iwpm_find_nlmsg_request(__u32 echo_seq)
 int iwpm_wait_complete_req(struct iwpm_nlmsg_request *nlmsg_request)
 {
 	int ret;
-	init_waitqueue_head(&nlmsg_request->waitq);
 
-	ret = wait_event_timeout(nlmsg_request->waitq,
+	ret = down_timeout(&nlmsg_request->sem, IWPM_NL_TIMEOUT);
-			(nlmsg_request->request_done != 0), IWPM_NL_TIMEOUT);
+	if (ret) {
-	if (!ret) {
 		ret = -EINVAL;
 		pr_info("%s: Timeout %d sec for netlink request (seq = %u)\n",
 			__func__, (IWPM_NL_TIMEOUT/HZ), nlmsg_request->nlmsg_seq);

drivers/infiniband/core/iwpm_util.h

@@ -69,7 +69,7 @@ struct iwpm_nlmsg_request {
 	u8	            nl_client;
 	u8                  request_done;
 	u16                 err_code;
-	wait_queue_head_t   waitq;
+	struct semaphore    sem;
 	struct kref         kref;
 };
 

drivers/infiniband/core/packer.c

@@ -44,7 +44,7 @@ static u64 value_read(int offset, int size, void *structure)
 	case 4: return be32_to_cpup((__be32 *) (structure + offset));
 	case 8: return be64_to_cpup((__be64 *) (structure + offset));
 	default:
-		printk(KERN_WARNING "Field size %d bits not handled\n", size * 8);
+		pr_warn("Field size %d bits not handled\n", size * 8);
 		return 0;
 	}
 }
@@ -104,9 +104,8 @@ void ib_pack(const struct ib_field        *desc,
 		} else {
 			if (desc[i].offset_bits % 8 ||
 			    desc[i].size_bits   % 8) {
-				printk(KERN_WARNING "Structure field %s of size %d "
+				pr_warn("Structure field %s of size %d bits is not byte-aligned\n",
-				       "bits is not byte-aligned\n",
+					desc[i].field_name, desc[i].size_bits);
-				       desc[i].field_name, desc[i].size_bits);
 			}
 
 			if (desc[i].struct_size_bytes)
@@ -132,7 +131,7 @@ static void value_write(int offset, int size, u64 val, void *structure)
 	case 32: *(__be32 *) (structure + offset) = cpu_to_be32(val); break;
 	case 64: *(__be64 *) (structure + offset) = cpu_to_be64(val); break;
 	default:
-		printk(KERN_WARNING "Field size %d bits not handled\n", size * 8);
+		pr_warn("Field size %d bits not handled\n", size * 8);
 	}
 }
 
@@ -188,9 +187,8 @@ void ib_unpack(const struct ib_field        *desc,
 		} else {
 			if (desc[i].offset_bits % 8 ||
 			    desc[i].size_bits   % 8) {
-				printk(KERN_WARNING "Structure field %s of size %d "
+				pr_warn("Structure field %s of size %d bits is not byte-aligned\n",
-				       "bits is not byte-aligned\n",
+					desc[i].field_name, desc[i].size_bits);
-				       desc[i].field_name, desc[i].size_bits);
 			}
 
 			memcpy(structure + desc[i].struct_offset_bytes,

drivers/infiniband/core/sa_query.c

@@ -864,13 +864,12 @@ static void update_sm_ah(struct work_struct *work)
 	struct ib_ah_attr   ah_attr;
 
 	if (ib_query_port(port->agent->device, port->port_num, &port_attr)) {
-		printk(KERN_WARNING "Couldn't query port\n");
+		pr_warn("Couldn't query port\n");
 		return;
 	}
 
 	new_ah = kmalloc(sizeof *new_ah, GFP_KERNEL);
 	if (!new_ah) {
-		printk(KERN_WARNING "Couldn't allocate new SM AH\n");
 		return;
 	}
 
@@ -880,16 +879,21 @@ static void update_sm_ah(struct work_struct *work)
 	new_ah->pkey_index = 0;
 	if (ib_find_pkey(port->agent->device, port->port_num,
 			 IB_DEFAULT_PKEY_FULL, &new_ah->pkey_index))
-		printk(KERN_ERR "Couldn't find index for default PKey\n");
+		pr_err("Couldn't find index for default PKey\n");
 
 	memset(&ah_attr, 0, sizeof ah_attr);
 	ah_attr.dlid     = port_attr.sm_lid;
 	ah_attr.sl       = port_attr.sm_sl;
 	ah_attr.port_num = port->port_num;
+	if (port_attr.grh_required) {
+		ah_attr.ah_flags = IB_AH_GRH;
+		ah_attr.grh.dgid.global.subnet_prefix = cpu_to_be64(port_attr.subnet_prefix);
+		ah_attr.grh.dgid.global.interface_id = cpu_to_be64(IB_SA_WELL_KNOWN_GUID);
+	}
 
 	new_ah->ah = ib_create_ah(port->agent->qp->pd, &ah_attr);
 	if (IS_ERR(new_ah->ah)) {
-		printk(KERN_WARNING "Couldn't create new SM AH\n");
+		pr_warn("Couldn't create new SM AH\n");
 		kfree(new_ah);
 		return;
 	}
@@ -1221,7 +1225,7 @@ static void ib_sa_path_rec_callback(struct ib_sa_query *sa_query,
 		rec.net = NULL;
 		rec.ifindex = 0;
 		rec.gid_type = IB_GID_TYPE_IB;
-		memset(rec.dmac, 0, ETH_ALEN);
+		eth_zero_addr(rec.dmac);
 		query->callback(status, &rec, query->context);
 	} else
 		query->callback(status, NULL, query->context);
@@ -1800,13 +1804,13 @@ static int __init ib_sa_init(void)
 
 	ret = ib_register_client(&sa_client);
 	if (ret) {
-		printk(KERN_ERR "Couldn't register ib_sa client\n");
+		pr_err("Couldn't register ib_sa client\n");
 		goto err1;
 	}
 
 	ret = mcast_init();
 	if (ret) {
-		printk(KERN_ERR "Couldn't initialize multicast handling\n");
+		pr_err("Couldn't initialize multicast handling\n");
 		goto err2;
 	}
 

drivers/infiniband/core/ucm.c

@@ -1234,7 +1234,7 @@ static int find_overflow_devnum(void)
 		ret = alloc_chrdev_region(&overflow_maj, 0, IB_UCM_MAX_DEVICES,
 					  "infiniband_cm");
 		if (ret) {
-			printk(KERN_ERR "ucm: couldn't register dynamic device number\n");
+			pr_err("ucm: couldn't register dynamic device number\n");
 			return ret;
 		}
 	}
@@ -1329,19 +1329,19 @@ static int __init ib_ucm_init(void)
 	ret = register_chrdev_region(IB_UCM_BASE_DEV, IB_UCM_MAX_DEVICES,
 				     "infiniband_cm");
 	if (ret) {
-		printk(KERN_ERR "ucm: couldn't register device number\n");
+		pr_err("ucm: couldn't register device number\n");
 		goto error1;
 	}
 
 	ret = class_create_file(&cm_class, &class_attr_abi_version.attr);
 	if (ret) {
-		printk(KERN_ERR "ucm: couldn't create abi_version attribute\n");
+		pr_err("ucm: couldn't create abi_version attribute\n");
 		goto error2;
 	}
 
 	ret = ib_register_client(&ucm_client);
 	if (ret) {
-		printk(KERN_ERR "ucm: couldn't register client\n");
+		pr_err("ucm: couldn't register client\n");
 		goto error3;
 	}
 	return 0;

drivers/infiniband/core/ucma.c

@@ -314,7 +314,7 @@ static void ucma_removal_event_handler(struct rdma_cm_id *cm_id)
 		}
 	}
 	if (!event_found)
-		printk(KERN_ERR "ucma_removal_event_handler: warning: connect request event wasn't found\n");
+		pr_err("ucma_removal_event_handler: warning: connect request event wasn't found\n");
 }
 
 static int ucma_event_handler(struct rdma_cm_id *cm_id,
@@ -1716,13 +1716,13 @@ static int __init ucma_init(void)
 
 	ret = device_create_file(ucma_misc.this_device, &dev_attr_abi_version);
 	if (ret) {
-		printk(KERN_ERR "rdma_ucm: couldn't create abi_version attr\n");
+		pr_err("rdma_ucm: couldn't create abi_version attr\n");
 		goto err1;
 	}
 
 	ucma_ctl_table_hdr = register_net_sysctl(&init_net, "net/rdma_ucm", ucma_ctl_table);
 	if (!ucma_ctl_table_hdr) {
-		printk(KERN_ERR "rdma_ucm: couldn't register sysctl paths\n");
+		pr_err("rdma_ucm: couldn't register sysctl paths\n");
 		ret = -ENOMEM;
 		goto err2;
 	}

drivers/infiniband/core/ud_header.c

@@ -479,8 +479,8 @@ int ib_ud_header_unpack(void                *buf,
 	buf += IB_LRH_BYTES;
 
 	if (header->lrh.link_version != 0) {
-		printk(KERN_WARNING "Invalid LRH.link_version %d\n",
+		pr_warn("Invalid LRH.link_version %d\n",
-		       header->lrh.link_version);
+			header->lrh.link_version);
 		return -EINVAL;
 	}
 
@@ -496,20 +496,20 @@ int ib_ud_header_unpack(void                *buf,
 		buf += IB_GRH_BYTES;
 
 		if (header->grh.ip_version != 6) {
-			printk(KERN_WARNING "Invalid GRH.ip_version %d\n",
+			pr_warn("Invalid GRH.ip_version %d\n",
-			       header->grh.ip_version);
+				header->grh.ip_version);
 			return -EINVAL;
 		}
 		if (header->grh.next_header != 0x1b) {
-			printk(KERN_WARNING "Invalid GRH.next_header 0x%02x\n",
+			pr_warn("Invalid GRH.next_header 0x%02x\n",
-			       header->grh.next_header);
+				header->grh.next_header);
 			return -EINVAL;
 		}
 		break;
 
 	default:
-		printk(KERN_WARNING "Invalid LRH.link_next_header %d\n",
+		pr_warn("Invalid LRH.link_next_header %d\n",
-		       header->lrh.link_next_header);
+			header->lrh.link_next_header);
 		return -EINVAL;
 	}
 
@@ -525,14 +525,13 @@ int ib_ud_header_unpack(void                *buf,
 		header->immediate_present = 1;
 		break;
 	default:
-		printk(KERN_WARNING "Invalid BTH.opcode 0x%02x\n",
+		pr_warn("Invalid BTH.opcode 0x%02x\n", header->bth.opcode);
-		       header->bth.opcode);
 		return -EINVAL;
 	}
 
 	if (header->bth.transport_header_version != 0) {
-		printk(KERN_WARNING "Invalid BTH.transport_header_version %d\n",
+		pr_warn("Invalid BTH.transport_header_version %d\n",
-		       header->bth.transport_header_version);
+			header->bth.transport_header_version);
 		return -EINVAL;
 	}
 

drivers/infiniband/core/uverbs_cmd.c

@@ -402,7 +402,7 @@ static void copy_query_dev_fields(struct ib_uverbs_file *file,
 	resp->hw_ver		= attr->hw_ver;
 	resp->max_qp		= attr->max_qp;
 	resp->max_qp_wr		= attr->max_qp_wr;
-	resp->device_cap_flags	= attr->device_cap_flags;
+	resp->device_cap_flags	= lower_32_bits(attr->device_cap_flags);
 	resp->max_sge		= attr->max_sge;
 	resp->max_sge_rd	= attr->max_sge_rd;
 	resp->max_cq		= attr->max_cq;
@@ -1174,6 +1174,7 @@ ssize_t ib_uverbs_alloc_mw(struct ib_uverbs_file *file,
 	struct ib_uobject             *uobj;
 	struct ib_pd                  *pd;
 	struct ib_mw                  *mw;
+	struct ib_udata		       udata;
 	int                            ret;
 
 	if (out_len < sizeof(resp))
@@ -1195,7 +1196,12 @@ ssize_t ib_uverbs_alloc_mw(struct ib_uverbs_file *file,
 		goto err_free;
 	}
 
-	mw = pd->device->alloc_mw(pd, cmd.mw_type);
+	INIT_UDATA(&udata, buf + sizeof(cmd),
+		   (unsigned long)cmd.response + sizeof(resp),
+		   in_len - sizeof(cmd) - sizeof(struct ib_uverbs_cmd_hdr),
+		   out_len - sizeof(resp));
+
+	mw = pd->device->alloc_mw(pd, cmd.mw_type, &udata);
 	if (IS_ERR(mw)) {
 		ret = PTR_ERR(mw);
 		goto err_put;
@@ -1970,7 +1976,8 @@ ssize_t ib_uverbs_create_qp(struct ib_uverbs_file *file,
 		   resp_size);
 	INIT_UDATA(&uhw, buf + sizeof(cmd),
 		   (unsigned long)cmd.response + resp_size,
-		   in_len - sizeof(cmd), out_len - resp_size);
+		   in_len - sizeof(cmd) - sizeof(struct ib_uverbs_cmd_hdr),
+		   out_len - resp_size);
 
 	memset(&cmd_ex, 0, sizeof(cmd_ex));
 	cmd_ex.user_handle = cmd.user_handle;
@@ -3085,6 +3092,14 @@ int ib_uverbs_ex_create_flow(struct ib_uverbs_file *file,
 	     !capable(CAP_NET_ADMIN)) || !capable(CAP_NET_RAW))
 		return -EPERM;
 
+	if (cmd.flow_attr.flags >= IB_FLOW_ATTR_FLAGS_RESERVED)
+		return -EINVAL;
+
+	if ((cmd.flow_attr.flags & IB_FLOW_ATTR_FLAGS_DONT_TRAP) &&
+	    ((cmd.flow_attr.type == IB_FLOW_ATTR_ALL_DEFAULT) ||
+	     (cmd.flow_attr.type == IB_FLOW_ATTR_MC_DEFAULT)))
+		return -EINVAL;
+
 	if (cmd.flow_attr.num_of_specs > IB_FLOW_SPEC_SUPPORT_LAYERS)
 		return -EINVAL;
 
@@ -3413,7 +3428,8 @@ ssize_t ib_uverbs_create_srq(struct ib_uverbs_file *file,
 
 	INIT_UDATA(&udata, buf + sizeof cmd,
 		   (unsigned long) cmd.response + sizeof resp,
-		   in_len - sizeof cmd, out_len - sizeof resp);
+		   in_len - sizeof cmd - sizeof(struct ib_uverbs_cmd_hdr),
+		   out_len - sizeof resp);
 
 	ret = __uverbs_create_xsrq(file, ib_dev, &xcmd, &udata);
 	if (ret)
@@ -3439,7 +3455,8 @@ ssize_t ib_uverbs_create_xsrq(struct ib_uverbs_file *file,
 
 	INIT_UDATA(&udata, buf + sizeof cmd,
 		   (unsigned long) cmd.response + sizeof resp,
-		   in_len - sizeof cmd, out_len - sizeof resp);
+		   in_len - sizeof cmd - sizeof(struct ib_uverbs_cmd_hdr),
+		   out_len - sizeof resp);
 
 	ret = __uverbs_create_xsrq(file, ib_dev, &cmd, &udata);
 	if (ret)
@@ -3583,9 +3600,9 @@ int ib_uverbs_ex_query_device(struct ib_uverbs_file *file,
 			      struct ib_udata *ucore,
 			      struct ib_udata *uhw)
 {
-	struct ib_uverbs_ex_query_device_resp resp;
+	struct ib_uverbs_ex_query_device_resp resp = { {0} };
 	struct ib_uverbs_ex_query_device  cmd;
-	struct ib_device_attr attr;
+	struct ib_device_attr attr = {0};
 	int err;
 
 	if (ucore->inlen < sizeof(cmd))
@@ -3606,14 +3623,11 @@ int ib_uverbs_ex_query_device(struct ib_uverbs_file *file,
 	if (ucore->outlen < resp.response_length)
 		return -ENOSPC;
 
-	memset(&attr, 0, sizeof(attr));
-
 	err = ib_dev->query_device(ib_dev, &attr, uhw);
 	if (err)
 		return err;
 
 	copy_query_dev_fields(file, ib_dev, &resp.base, &attr);
-	resp.comp_mask = 0;
 
 	if (ucore->outlen < resp.response_length + sizeof(resp.odp_caps))
 		goto end;
@@ -3626,9 +3640,6 @@ int ib_uverbs_ex_query_device(struct ib_uverbs_file *file,
 		attr.odp_caps.per_transport_caps.uc_odp_caps;
 	resp.odp_caps.per_transport_caps.ud_odp_caps =
 		attr.odp_caps.per_transport_caps.ud_odp_caps;
-	resp.odp_caps.reserved = 0;
-#else
-	memset(&resp.odp_caps, 0, sizeof(resp.odp_caps));
 #endif
 	resp.response_length += sizeof(resp.odp_caps);
 
@@ -3646,8 +3657,5 @@ int ib_uverbs_ex_query_device(struct ib_uverbs_file *file,
 
 end:
 	err = ib_copy_to_udata(ucore, &resp, resp.response_length);
-	if (err)
+	return err;
-		return err;
-
-	return 0;
 }

drivers/infiniband/core/uverbs_main.c

@@ -683,12 +683,28 @@ out:
 	return ev_file;
 }
 
+static int verify_command_mask(struct ib_device *ib_dev, __u32 command)
+{
+	u64 mask;
+
+	if (command <= IB_USER_VERBS_CMD_OPEN_QP)
+		mask = ib_dev->uverbs_cmd_mask;
+	else
+		mask = ib_dev->uverbs_ex_cmd_mask;
+
+	if (mask & ((u64)1 << command))
+		return 0;
+
+	return -1;
+}
+
 static ssize_t ib_uverbs_write(struct file *filp, const char __user *buf,
 			     size_t count, loff_t *pos)
 {
 	struct ib_uverbs_file *file = filp->private_data;
 	struct ib_device *ib_dev;
 	struct ib_uverbs_cmd_hdr hdr;
+	__u32 command;
 	__u32 flags;
 	int srcu_key;
 	ssize_t ret;
@@ -707,37 +723,34 @@ static ssize_t ib_uverbs_write(struct file *filp, const char __user *buf,
 		goto out;
 	}
 
+	if (hdr.command & ~(__u32)(IB_USER_VERBS_CMD_FLAGS_MASK |
+				   IB_USER_VERBS_CMD_COMMAND_MASK)) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	command = hdr.command & IB_USER_VERBS_CMD_COMMAND_MASK;
+	if (verify_command_mask(ib_dev, command)) {
+		ret = -EOPNOTSUPP;
+		goto out;
+	}
+
+	if (!file->ucontext &&
+	    command != IB_USER_VERBS_CMD_GET_CONTEXT) {
+		ret = -EINVAL;
+		goto out;
+	}
+
 	flags = (hdr.command &
 		 IB_USER_VERBS_CMD_FLAGS_MASK) >> IB_USER_VERBS_CMD_FLAGS_SHIFT;
 
 	if (!flags) {
-		__u32 command;
-
-		if (hdr.command & ~(__u32)(IB_USER_VERBS_CMD_FLAGS_MASK |
-					   IB_USER_VERBS_CMD_COMMAND_MASK)) {
-			ret = -EINVAL;
-			goto out;
-		}
-
-		command = hdr.command & IB_USER_VERBS_CMD_COMMAND_MASK;
-
 		if (command >= ARRAY_SIZE(uverbs_cmd_table) ||
 		    !uverbs_cmd_table[command]) {
 			ret = -EINVAL;
 			goto out;
 		}
 
-		if (!file->ucontext &&
-		    command != IB_USER_VERBS_CMD_GET_CONTEXT) {
-			ret = -EINVAL;
-			goto out;
-		}
-
-		if (!(ib_dev->uverbs_cmd_mask & (1ull << command))) {
-			ret = -ENOSYS;
-			goto out;
-		}
-
 		if (hdr.in_words * 4 != count) {
 			ret = -EINVAL;
 			goto out;
@@ -749,21 +762,11 @@ static ssize_t ib_uverbs_write(struct file *filp, const char __user *buf,
 						 hdr.out_words * 4);
 
 	} else if (flags == IB_USER_VERBS_CMD_FLAG_EXTENDED) {
-		__u32 command;
-
 		struct ib_uverbs_ex_cmd_hdr ex_hdr;
 		struct ib_udata ucore;
 		struct ib_udata uhw;
 		size_t written_count = count;
 
-		if (hdr.command & ~(__u32)(IB_USER_VERBS_CMD_FLAGS_MASK |
-					   IB_USER_VERBS_CMD_COMMAND_MASK)) {
-			ret = -EINVAL;
-			goto out;
-		}
-
-		command = hdr.command & IB_USER_VERBS_CMD_COMMAND_MASK;
-
 		if (command >= ARRAY_SIZE(uverbs_ex_cmd_table) ||
 		    !uverbs_ex_cmd_table[command]) {
 			ret = -ENOSYS;
@@ -775,11 +778,6 @@ static ssize_t ib_uverbs_write(struct file *filp, const char __user *buf,
 			goto out;
 		}
 
-		if (!(ib_dev->uverbs_ex_cmd_mask & (1ull << command))) {
-			ret = -ENOSYS;
-			goto out;
-		}
-
 		if (count < (sizeof(hdr) + sizeof(ex_hdr))) {
 			ret = -EINVAL;
 			goto out;
@@ -1058,7 +1056,7 @@ static int find_overflow_devnum(void)
 		ret = alloc_chrdev_region(&overflow_maj, 0, IB_UVERBS_MAX_DEVICES,
 					  "infiniband_verbs");
 		if (ret) {
-			printk(KERN_ERR "user_verbs: couldn't register dynamic device number\n");
+			pr_err("user_verbs: couldn't register dynamic device number\n");
 			return ret;
 		}
 	}
@@ -1279,14 +1277,14 @@ static int __init ib_uverbs_init(void)
 	ret = register_chrdev_region(IB_UVERBS_BASE_DEV, IB_UVERBS_MAX_DEVICES,
 				     "infiniband_verbs");
 	if (ret) {
-		printk(KERN_ERR "user_verbs: couldn't register device number\n");
+		pr_err("user_verbs: couldn't register device number\n");
 		goto out;
 	}
 
 	uverbs_class = class_create(THIS_MODULE, "infiniband_verbs");
 	if (IS_ERR(uverbs_class)) {
 		ret = PTR_ERR(uverbs_class);
-		printk(KERN_ERR "user_verbs: couldn't create class infiniband_verbs\n");
+		pr_err("user_verbs: couldn't create class infiniband_verbs\n");
 		goto out_chrdev;
 	}
 
@@ -1294,13 +1292,13 @@ static int __init ib_uverbs_init(void)
 
 	ret = class_create_file(uverbs_class, &class_attr_abi_version.attr);
 	if (ret) {
-		printk(KERN_ERR "user_verbs: couldn't create abi_version attribute\n");
+		pr_err("user_verbs: couldn't create abi_version attribute\n");
 		goto out_class;
 	}
 
 	ret = ib_register_client(&uverbs_client);
 	if (ret) {
-		printk(KERN_ERR "user_verbs: couldn't register client\n");
+		pr_err("user_verbs: couldn't register client\n");
 		goto out_class;
 	}
 

drivers/infiniband/core/verbs.c

@@ -1551,6 +1551,46 @@ int ib_check_mr_status(struct ib_mr *mr, u32 check_mask,
 }
 EXPORT_SYMBOL(ib_check_mr_status);
 
+int ib_set_vf_link_state(struct ib_device *device, int vf, u8 port,
+			 int state)
+{
+	if (!device->set_vf_link_state)
+		return -ENOSYS;
+
+	return device->set_vf_link_state(device, vf, port, state);
+}
+EXPORT_SYMBOL(ib_set_vf_link_state);
+
+int ib_get_vf_config(struct ib_device *device, int vf, u8 port,
+		     struct ifla_vf_info *info)
+{
+	if (!device->get_vf_config)
+		return -ENOSYS;
+
+	return device->get_vf_config(device, vf, port, info);
+}
+EXPORT_SYMBOL(ib_get_vf_config);
+
+int ib_get_vf_stats(struct ib_device *device, int vf, u8 port,
+		    struct ifla_vf_stats *stats)
+{
+	if (!device->get_vf_stats)
+		return -ENOSYS;
+
+	return device->get_vf_stats(device, vf, port, stats);
+}
+EXPORT_SYMBOL(ib_get_vf_stats);
+
+int ib_set_vf_guid(struct ib_device *device, int vf, u8 port, u64 guid,
+		   int type)
+{
+	if (!device->set_vf_guid)
+		return -ENOSYS;
+
+	return device->set_vf_guid(device, vf, port, guid, type);
+}
+EXPORT_SYMBOL(ib_set_vf_guid);
+
 /**
  * ib_map_mr_sg() - Map the largest prefix of a dma mapped SG list
  *     and set it the memory region.
@@ -1567,6 +1607,8 @@ EXPORT_SYMBOL(ib_check_mr_status);
  * - The last sg element is allowed to have length less than page_size.
  * - If sg_nents total byte length exceeds the mr max_num_sge * page_size
  *   then only max_num_sg entries will be mapped.
+ * - If the MR was allocated with type IB_MR_TYPE_SG_GAPS_REG, non of these
+ *   constraints holds and the page_size argument is ignored.
  *
  * Returns the number of sg elements that were mapped to the memory region.
  *
@@ -1657,3 +1699,167 @@ next_page:
 	return i;
 }
 EXPORT_SYMBOL(ib_sg_to_pages);
+
+struct ib_drain_cqe {
+	struct ib_cqe cqe;
+	struct completion done;
+};
+
+static void ib_drain_qp_done(struct ib_cq *cq, struct ib_wc *wc)
+{
+	struct ib_drain_cqe *cqe = container_of(wc->wr_cqe, struct ib_drain_cqe,
+						cqe);
+
+	complete(&cqe->done);
+}
+
+/*
+ * Post a WR and block until its completion is reaped for the SQ.
+ */
+static void __ib_drain_sq(struct ib_qp *qp)
+{
+	struct ib_qp_attr attr = { .qp_state = IB_QPS_ERR };
+	struct ib_drain_cqe sdrain;
+	struct ib_send_wr swr = {}, *bad_swr;
+	int ret;
+
+	if (qp->send_cq->poll_ctx == IB_POLL_DIRECT) {
+		WARN_ONCE(qp->send_cq->poll_ctx == IB_POLL_DIRECT,
+			  "IB_POLL_DIRECT poll_ctx not supported for drain\n");
+		return;
+	}
+
+	swr.wr_cqe = &sdrain.cqe;
+	sdrain.cqe.done = ib_drain_qp_done;
+	init_completion(&sdrain.done);
+
+	ret = ib_modify_qp(qp, &attr, IB_QP_STATE);
+	if (ret) {
+		WARN_ONCE(ret, "failed to drain send queue: %d\n", ret);
+		return;
+	}
+
+	ret = ib_post_send(qp, &swr, &bad_swr);
+	if (ret) {
+		WARN_ONCE(ret, "failed to drain send queue: %d\n", ret);
+		return;
+	}
+
+	wait_for_completion(&sdrain.done);
+}
+
+/*
+ * Post a WR and block until its completion is reaped for the RQ.
+ */
+static void __ib_drain_rq(struct ib_qp *qp)
+{
+	struct ib_qp_attr attr = { .qp_state = IB_QPS_ERR };
+	struct ib_drain_cqe rdrain;
+	struct ib_recv_wr rwr = {}, *bad_rwr;
+	int ret;
+
+	if (qp->recv_cq->poll_ctx == IB_POLL_DIRECT) {
+		WARN_ONCE(qp->recv_cq->poll_ctx == IB_POLL_DIRECT,
+			  "IB_POLL_DIRECT poll_ctx not supported for drain\n");
+		return;
+	}
+
+	rwr.wr_cqe = &rdrain.cqe;
+	rdrain.cqe.done = ib_drain_qp_done;
+	init_completion(&rdrain.done);
+
+	ret = ib_modify_qp(qp, &attr, IB_QP_STATE);
+	if (ret) {
+		WARN_ONCE(ret, "failed to drain recv queue: %d\n", ret);
+		return;
+	}
+
+	ret = ib_post_recv(qp, &rwr, &bad_rwr);
+	if (ret) {
+		WARN_ONCE(ret, "failed to drain recv queue: %d\n", ret);
+		return;
+	}
+
+	wait_for_completion(&rdrain.done);
+}
+
+/**
+ * ib_drain_sq() - Block until all SQ CQEs have been consumed by the
+ *		   application.
+ * @qp:            queue pair to drain
+ *
+ * If the device has a provider-specific drain function, then
+ * call that.  Otherwise call the generic drain function
+ * __ib_drain_sq().
+ *
+ * The caller must:
+ *
+ * ensure there is room in the CQ and SQ for the drain work request and
+ * completion.
+ *
+ * allocate the CQ using ib_alloc_cq() and the CQ poll context cannot be
+ * IB_POLL_DIRECT.
+ *
+ * ensure that there are no other contexts that are posting WRs concurrently.
+ * Otherwise the drain is not guaranteed.
+ */
+void ib_drain_sq(struct ib_qp *qp)
+{
+	if (qp->device->drain_sq)
+		qp->device->drain_sq(qp);
+	else
+		__ib_drain_sq(qp);
+}
+EXPORT_SYMBOL(ib_drain_sq);
+
+/**
+ * ib_drain_rq() - Block until all RQ CQEs have been consumed by the
+ *		   application.
+ * @qp:            queue pair to drain
+ *
+ * If the device has a provider-specific drain function, then
+ * call that.  Otherwise call the generic drain function
+ * __ib_drain_rq().
+ *
+ * The caller must:
+ *
+ * ensure there is room in the CQ and RQ for the drain work request and
+ * completion.
+ *
+ * allocate the CQ using ib_alloc_cq() and the CQ poll context cannot be
+ * IB_POLL_DIRECT.
+ *
+ * ensure that there are no other contexts that are posting WRs concurrently.
+ * Otherwise the drain is not guaranteed.
+ */
+void ib_drain_rq(struct ib_qp *qp)
+{
+	if (qp->device->drain_rq)
+		qp->device->drain_rq(qp);
+	else
+		__ib_drain_rq(qp);
+}
+EXPORT_SYMBOL(ib_drain_rq);
+
+/**
+ * ib_drain_qp() - Block until all CQEs have been consumed by the
+ *		   application on both the RQ and SQ.
+ * @qp:            queue pair to drain
+ *
+ * The caller must:
+ *
+ * ensure there is room in the CQ(s), SQ, and RQ for drain work requests
+ * and completions.
+ *
+ * allocate the CQs using ib_alloc_cq() and the CQ poll context cannot be
+ * IB_POLL_DIRECT.
+ *
+ * ensure that there are no other contexts that are posting WRs concurrently.
+ * Otherwise the drain is not guaranteed.
+ */
+void ib_drain_qp(struct ib_qp *qp)
+{
+	ib_drain_sq(qp);
+	ib_drain_rq(qp);
+}
+EXPORT_SYMBOL(ib_drain_qp);