Merge branch 'for-linus' of master.kernel.org:/pub/scm/linux/kernel/git/roland/infiniband

* 'for-linus' of master.kernel.org:/pub/scm/linux/kernel/git/roland/infiniband:
  IB/core: Set static rate in ib_init_ah_from_path()
  IB/ipath: Make ipath_map_sg() static
  IB/core: Fix sparse warnings about shadowed declarations
  RDMA/cma: Add multicast communication support
  IB/sa: Track multicast join/leave requests
  IPoIB: CM error handling thinko fix
  RDMA/cxgb3: Remove Open Grid Computing copyrights in iw_cxgb3 driver
  RDMA/cxgb3: Fail posts synchronously when in TERMINATE state
  RDMA/iwcm: iw_cm_id destruction race fixes
  IB/ehca: Change query_port() to return LINK_UP instead UNKNOWN
  IB/ehca: Allow en/disabling scaling code via module parameter
  IB/ehca: Fix race condition/locking issues in scaling code
  IB/ehca: Rework irq handler
  IPoIB: Only allow root to change between datagram and connected mode
  IB/mthca: Fix allocation of ICM chunks in coherent memory
  IB/mthca: Allow the QP state transition RESET->RESET

drivers/infiniband/core/Makefile

@@ -12,7 +12,7 @@ ib_core-y :=			packer.o ud_header.o verbs.o sysfs.o \
 
 ib_mad-y :=			mad.o smi.o agent.o mad_rmpp.o
 
-ib_sa-y :=			sa_query.o
+ib_sa-y :=			sa_query.o multicast.o
 
 ib_cm-y :=			cm.o
 

drivers/infiniband/core/fmr_pool.c

@@ -301,7 +301,7 @@ struct ib_fmr_pool *ib_create_fmr_pool(struct ib_pd             *pd,
 
 	{
 		struct ib_pool_fmr *fmr;
-		struct ib_fmr_attr attr = {
+		struct ib_fmr_attr fmr_attr = {
 			.max_pages  = params->max_pages_per_fmr,
 			.max_maps   = pool->max_remaps,
 			.page_shift = params->page_shift
@@ -321,7 +321,7 @@ struct ib_fmr_pool *ib_create_fmr_pool(struct ib_pd             *pd,
 			fmr->ref_count        = 0;
 			INIT_HLIST_NODE(&fmr->cache_node);
 
-			fmr->fmr = ib_alloc_fmr(pd, params->access, &attr);
+			fmr->fmr = ib_alloc_fmr(pd, params->access, &fmr_attr);
 			if (IS_ERR(fmr->fmr)) {
 				printk(KERN_WARNING "fmr_create failed for FMR %d", i);
 				kfree(fmr);

drivers/infiniband/core/iwcm.c

@@ -146,6 +146,12 @@ static int copy_private_data(struct iw_cm_event *event)
 	return 0;
 }
 
+static void free_cm_id(struct iwcm_id_private *cm_id_priv)
+{
+	dealloc_work_entries(cm_id_priv);
+	kfree(cm_id_priv);
+}
+
 /*
  * Release a reference on cm_id. If the last reference is being
  * released, enable the waiting thread (in iw_destroy_cm_id) to
@@ -153,21 +159,14 @@ static int copy_private_data(struct iw_cm_event *event)
  */
 static int iwcm_deref_id(struct iwcm_id_private *cm_id_priv)
 {
-	int ret = 0;
-
 	BUG_ON(atomic_read(&cm_id_priv->refcount)==0);
 	if (atomic_dec_and_test(&cm_id_priv->refcount)) {
 		BUG_ON(!list_empty(&cm_id_priv->work_list));
-		if (waitqueue_active(&cm_id_priv->destroy_comp.wait)) {
-			BUG_ON(cm_id_priv->state != IW_CM_STATE_DESTROYING);
-			BUG_ON(test_bit(IWCM_F_CALLBACK_DESTROY,
-					&cm_id_priv->flags));
-			ret = 1;
-		}
 		complete(&cm_id_priv->destroy_comp);
+		return 1;
 	}
 
-	return ret;
+	return 0;
 }
 
 static void add_ref(struct iw_cm_id *cm_id)
@@ -181,7 +180,11 @@ static void rem_ref(struct iw_cm_id *cm_id)
 {
 	struct iwcm_id_private *cm_id_priv;
 	cm_id_priv = container_of(cm_id, struct iwcm_id_private, id);
-	iwcm_deref_id(cm_id_priv);
+	if (iwcm_deref_id(cm_id_priv) &&
+	    test_bit(IWCM_F_CALLBACK_DESTROY, &cm_id_priv->flags)) {
+		BUG_ON(!list_empty(&cm_id_priv->work_list));
+		free_cm_id(cm_id_priv);
+	}
 }
 
 static int cm_event_handler(struct iw_cm_id *cm_id, struct iw_cm_event *event);
@@ -355,7 +358,9 @@ static void destroy_cm_id(struct iw_cm_id *cm_id)
 	case IW_CM_STATE_CONN_RECV:
 		/*
 		 * App called destroy before/without calling accept after
-		 * receiving connection request event notification.
+		 * receiving connection request event notification or
+		 * returned non zero from the event callback function.
+		 * In either case, must tell the provider to reject.
 		 */
 		cm_id_priv->state = IW_CM_STATE_DESTROYING;
 		break;
@@ -391,9 +396,7 @@ void iw_destroy_cm_id(struct iw_cm_id *cm_id)
 
 	wait_for_completion(&cm_id_priv->destroy_comp);
 
-	dealloc_work_entries(cm_id_priv);
-
-	kfree(cm_id_priv);
+	free_cm_id(cm_id_priv);
 }
 EXPORT_SYMBOL(iw_destroy_cm_id);
 
@@ -647,10 +650,11 @@ static void cm_conn_req_handler(struct iwcm_id_private *listen_id_priv,
 	/* Call the client CM handler */
 	ret = cm_id->cm_handler(cm_id, iw_event);
 	if (ret) {
+		iw_cm_reject(cm_id, NULL, 0);
 		set_bit(IWCM_F_CALLBACK_DESTROY, &cm_id_priv->flags);
 		destroy_cm_id(cm_id);
 		if (atomic_read(&cm_id_priv->refcount)==0)
-			kfree(cm_id);
+			free_cm_id(cm_id_priv);
 	}
 
 out:
@@ -854,13 +858,12 @@ static void cm_work_handler(struct work_struct *_work)
 			destroy_cm_id(&cm_id_priv->id);
 		}
 		BUG_ON(atomic_read(&cm_id_priv->refcount)==0);
-		if (iwcm_deref_id(cm_id_priv))
-			return;
-
-		if (atomic_read(&cm_id_priv->refcount)==0 &&
-		    test_bit(IWCM_F_CALLBACK_DESTROY, &cm_id_priv->flags)) {
-			dealloc_work_entries(cm_id_priv);
-			kfree(cm_id_priv);
+		if (iwcm_deref_id(cm_id_priv)) {
+			if (test_bit(IWCM_F_CALLBACK_DESTROY,
+				     &cm_id_priv->flags)) {
+				BUG_ON(!list_empty(&cm_id_priv->work_list));
+				free_cm_id(cm_id_priv);
+			}
 			return;
 		}
 		spin_lock_irqsave(&cm_id_priv->lock, flags);

drivers/infiniband/core/sa.h

+/*
+ * Copyright (c) 2004 Topspin Communications.  All rights reserved.
+ * Copyright (c) 2005 Voltaire, Inc.  All rights reserved.
+ * Copyright (c) 2006 Intel Corporation.  All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef SA_H
+#define SA_H
+
+#include <rdma/ib_sa.h>
+
+static inline void ib_sa_client_get(struct ib_sa_client *client)
+{
+	atomic_inc(&client->users);
+}
+
+static inline void ib_sa_client_put(struct ib_sa_client *client)
+{
+	if (atomic_dec_and_test(&client->users))
+		complete(&client->comp);
+}
+
+int ib_sa_mcmember_rec_query(struct ib_sa_client *client,
+			     struct ib_device *device, u8 port_num,
+			     u8 method,
+			     struct ib_sa_mcmember_rec *rec,
+			     ib_sa_comp_mask comp_mask,
+			     int timeout_ms, gfp_t gfp_mask,
+			     void (*callback)(int status,
+					      struct ib_sa_mcmember_rec *resp,
+					      void *context),
+			     void *context,
+			     struct ib_sa_query **sa_query);
+
+int mcast_init(void);
+void mcast_cleanup(void);
+
+#endif /* SA_H */

drivers/infiniband/core/sa_query.c

@@ -47,8 +47,8 @@
 #include <linux/workqueue.h>
 
 #include <rdma/ib_pack.h>
-#include <rdma/ib_sa.h>
 #include <rdma/ib_cache.h>
+#include "sa.h"
 
 MODULE_AUTHOR("Roland Dreier");
 MODULE_DESCRIPTION("InfiniBand subnet administration query support");
@@ -425,17 +425,6 @@ void ib_sa_register_client(struct ib_sa_client *client)
 }
 EXPORT_SYMBOL(ib_sa_register_client);
 
-static inline void ib_sa_client_get(struct ib_sa_client *client)
-{
-	atomic_inc(&client->users);
-}
-
-static inline void ib_sa_client_put(struct ib_sa_client *client)
-{
-	if (atomic_dec_and_test(&client->users))
-		complete(&client->comp);
-}
-
 void ib_sa_unregister_client(struct ib_sa_client *client)
 {
 	ib_sa_client_put(client);
@@ -482,6 +471,7 @@ int ib_init_ah_from_path(struct ib_device *device, u8 port_num,
 	ah_attr->sl = rec->sl;
 	ah_attr->src_path_bits = be16_to_cpu(rec->slid) & 0x7f;
 	ah_attr->port_num = port_num;
+	ah_attr->static_rate = rec->rate;
 
 	if (rec->hop_limit > 1) {
 		ah_attr->ah_flags = IB_AH_GRH;
@@ -901,7 +891,6 @@ int ib_sa_mcmember_rec_query(struct ib_sa_client *client,
 	kfree(query);
 	return ret;
 }
-EXPORT_SYMBOL(ib_sa_mcmember_rec_query);
 
 static void send_handler(struct ib_mad_agent *agent,
 			 struct ib_mad_send_wc *mad_send_wc)
@@ -1053,14 +1042,27 @@ static int __init ib_sa_init(void)
 	get_random_bytes(&tid, sizeof tid);
 
 	ret = ib_register_client(&sa_client);
-	if (ret)
+	if (ret) {
 		printk(KERN_ERR "Couldn't register ib_sa client\n");
+		goto err1;
+	}
+
+	ret = mcast_init();
+	if (ret) {
+		printk(KERN_ERR "Couldn't initialize multicast handling\n");
+		goto err2;
+	}
 
+	return 0;
+err2:
+	ib_unregister_client(&sa_client);
+err1:
 	return ret;
 }
 
 static void __exit ib_sa_cleanup(void)
 {
+	mcast_cleanup();
 	ib_unregister_client(&sa_client);
 	idr_destroy(&query_idr);
 }

drivers/infiniband/core/sysfs.c

@@ -714,8 +714,6 @@ int ib_device_register_sysfs(struct ib_device *device)
 		if (ret)
 			goto err_put;
 	} else {
-		int i;
-
 		for (i = 1; i <= device->phys_port_cnt; ++i) {
 			ret = add_port(device, i);
 			if (ret)

drivers/infiniband/core/ucma.c

@@ -70,10 +70,24 @@ struct ucma_context {
 	u64			uid;
 
 	struct list_head	list;
+	struct list_head	mc_list;
+};
+
+struct ucma_multicast {
+	struct ucma_context	*ctx;
+	int			id;
+	int			events_reported;
+
+	u64			uid;
+	struct list_head	list;
+	struct sockaddr		addr;
+	u8			pad[sizeof(struct sockaddr_in6) -
+				    sizeof(struct sockaddr)];
 };
 
 struct ucma_event {
 	struct ucma_context	*ctx;
+	struct ucma_multicast	*mc;
 	struct list_head	list;
 	struct rdma_cm_id	*cm_id;
 	struct rdma_ucm_event_resp resp;
@@ -81,6 +95,7 @@ struct ucma_event {
 
 static DEFINE_MUTEX(mut);
 static DEFINE_IDR(ctx_idr);
+static DEFINE_IDR(multicast_idr);
 
 static inline struct ucma_context *_ucma_find_context(int id,
 						      struct ucma_file *file)
@@ -124,6 +139,7 @@ static struct ucma_context *ucma_alloc_ctx(struct ucma_file *file)
 
 	atomic_set(&ctx->ref, 1);
 	init_completion(&ctx->comp);
+	INIT_LIST_HEAD(&ctx->mc_list);
 	ctx->file = file;
 
 	do {
@@ -147,6 +163,37 @@ static struct ucma_context *ucma_alloc_ctx(struct ucma_file *file)
 	return NULL;
 }
 
+static struct ucma_multicast* ucma_alloc_multicast(struct ucma_context *ctx)
+{
+	struct ucma_multicast *mc;
+	int ret;
+
+	mc = kzalloc(sizeof(*mc), GFP_KERNEL);
+	if (!mc)
+		return NULL;
+
+	do {
+		ret = idr_pre_get(&multicast_idr, GFP_KERNEL);
+		if (!ret)
+			goto error;
+
+		mutex_lock(&mut);
+		ret = idr_get_new(&multicast_idr, mc, &mc->id);
+		mutex_unlock(&mut);
+	} while (ret == -EAGAIN);
+
+	if (ret)
+		goto error;
+
+	mc->ctx = ctx;
+	list_add_tail(&mc->list, &ctx->mc_list);
+	return mc;
+
+error:
+	kfree(mc);
+	return NULL;
+}
+
 static void ucma_copy_conn_event(struct rdma_ucm_conn_param *dst,
 				 struct rdma_conn_param *src)
 {
@@ -180,8 +227,19 @@ static void ucma_set_event_context(struct ucma_context *ctx,
 				   struct ucma_event *uevent)
 {
 	uevent->ctx = ctx;
+	switch (event->event) {
+	case RDMA_CM_EVENT_MULTICAST_JOIN:
+	case RDMA_CM_EVENT_MULTICAST_ERROR:
+		uevent->mc = (struct ucma_multicast *)
+			     event->param.ud.private_data;
+		uevent->resp.uid = uevent->mc->uid;
+		uevent->resp.id = uevent->mc->id;
+		break;
+	default:
 		uevent->resp.uid = ctx->uid;
 		uevent->resp.id = ctx->id;
+		break;
+	}
 }
 
 static int ucma_event_handler(struct rdma_cm_id *cm_id,
@@ -199,7 +257,7 @@ static int ucma_event_handler(struct rdma_cm_id *cm_id,
 	ucma_set_event_context(ctx, event, uevent);
 	uevent->resp.event = event->event;
 	uevent->resp.status = event->status;
-	if (cm_id->ps == RDMA_PS_UDP)
+	if (cm_id->ps == RDMA_PS_UDP || cm_id->ps == RDMA_PS_IPOIB)
 		ucma_copy_ud_event(&uevent->resp.param.ud, &event->param.ud);
 	else
 		ucma_copy_conn_event(&uevent->resp.param.conn,
@@ -290,6 +348,8 @@ static ssize_t ucma_get_event(struct ucma_file *file, const char __user *inbuf,
 
 	list_del(&uevent->list);
 	uevent->ctx->events_reported++;
+	if (uevent->mc)
+		uevent->mc->events_reported++;
 	kfree(uevent);
 done:
 	mutex_unlock(&file->mut);
@@ -342,6 +402,19 @@ static ssize_t ucma_create_id(struct ucma_file *file,
 	return ret;
 }
 
+static void ucma_cleanup_multicast(struct ucma_context *ctx)
+{
+	struct ucma_multicast *mc, *tmp;
+
+	mutex_lock(&mut);
+	list_for_each_entry_safe(mc, tmp, &ctx->mc_list, list) {
+		list_del(&mc->list);
+		idr_remove(&multicast_idr, mc->id);
+		kfree(mc);
+	}
+	mutex_unlock(&mut);
+}
+
 static void ucma_cleanup_events(struct ucma_context *ctx)
 {
 	struct ucma_event *uevent, *tmp;
@@ -360,6 +433,19 @@ static void ucma_cleanup_events(struct ucma_context *ctx)
 	}
 }
 
+static void ucma_cleanup_mc_events(struct ucma_multicast *mc)
+{
+	struct ucma_event *uevent, *tmp;
+
+	list_for_each_entry_safe(uevent, tmp, &mc->ctx->file->event_list, list) {
+		if (uevent->mc != mc)
+			continue;
+
+		list_del(&uevent->list);
+		kfree(uevent);
+	}
+}
+
 static int ucma_free_ctx(struct ucma_context *ctx)
 {
 	int events_reported;
@@ -367,6 +453,8 @@ static int ucma_free_ctx(struct ucma_context *ctx)
 	/* No new events will be generated after destroying the id. */
 	rdma_destroy_id(ctx->cm_id);
 
+	ucma_cleanup_multicast(ctx);
+
 	/* Cleanup events not yet reported to the user. */
 	mutex_lock(&ctx->file->mut);
 	ucma_cleanup_events(ctx);
@@ -731,6 +819,114 @@ static ssize_t ucma_notify(struct ucma_file *file, const char __user *inbuf,
 	return ret;
 }
 
+static ssize_t ucma_join_multicast(struct ucma_file *file,
+				   const char __user *inbuf,
+				   int in_len, int out_len)
+{
+	struct rdma_ucm_join_mcast cmd;
+	struct rdma_ucm_create_id_resp resp;
+	struct ucma_context *ctx;
+	struct ucma_multicast *mc;
+	int ret;
+
+	if (out_len < sizeof(resp))
+		return -ENOSPC;
+
+	if (copy_from_user(&cmd, inbuf, sizeof(cmd)))
+		return -EFAULT;
+
+	ctx = ucma_get_ctx(file, cmd.id);
+	if (IS_ERR(ctx))
+		return PTR_ERR(ctx);
+
+	mutex_lock(&file->mut);
+	mc = ucma_alloc_multicast(ctx);
+	if (IS_ERR(mc)) {
+		ret = PTR_ERR(mc);
+		goto err1;
+	}
+
+	mc->uid = cmd.uid;
+	memcpy(&mc->addr, &cmd.addr, sizeof cmd.addr);
+	ret = rdma_join_multicast(ctx->cm_id, &mc->addr, mc);
+	if (ret)
+		goto err2;
+
+	resp.id = mc->id;
+	if (copy_to_user((void __user *)(unsigned long)cmd.response,
+			 &resp, sizeof(resp))) {
+		ret = -EFAULT;
+		goto err3;
+	}
+
+	mutex_unlock(&file->mut);
+	ucma_put_ctx(ctx);
+	return 0;
+
+err3:
+	rdma_leave_multicast(ctx->cm_id, &mc->addr);
+	ucma_cleanup_mc_events(mc);
+err2:
+	mutex_lock(&mut);
+	idr_remove(&multicast_idr, mc->id);
+	mutex_unlock(&mut);
+	list_del(&mc->list);
+	kfree(mc);
+err1:
+	mutex_unlock(&file->mut);
+	ucma_put_ctx(ctx);
+	return ret;
+}
+
+static ssize_t ucma_leave_multicast(struct ucma_file *file,
+				    const char __user *inbuf,
+				    int in_len, int out_len)
+{
+	struct rdma_ucm_destroy_id cmd;
+	struct rdma_ucm_destroy_id_resp resp;
+	struct ucma_multicast *mc;
+	int ret = 0;
+
+	if (out_len < sizeof(resp))
+		return -ENOSPC;
+
+	if (copy_from_user(&cmd, inbuf, sizeof(cmd)))
+		return -EFAULT;
+
+	mutex_lock(&mut);
+	mc = idr_find(&multicast_idr, cmd.id);
+	if (!mc)
+		mc = ERR_PTR(-ENOENT);
+	else if (mc->ctx->file != file)
+		mc = ERR_PTR(-EINVAL);
+	else {
+		idr_remove(&multicast_idr, mc->id);
+		atomic_inc(&mc->ctx->ref);
+	}
+	mutex_unlock(&mut);
+
+	if (IS_ERR(mc)) {
+		ret = PTR_ERR(mc);
+		goto out;
+	}
+
+	rdma_leave_multicast(mc->ctx->cm_id, &mc->addr);
+	mutex_lock(&mc->ctx->file->mut);
+	ucma_cleanup_mc_events(mc);
+	list_del(&mc->list);
+	mutex_unlock(&mc->ctx->file->mut);
+
+	ucma_put_ctx(mc->ctx);
+	resp.events_reported = mc->events_reported;
+	kfree(mc);
+
+	if (copy_to_user((void __user *)(unsigned long)cmd.response,
+			 &resp, sizeof(resp)))
+		ret = -EFAULT;
+out:
+	return ret;
+}
+
 static ssize_t (*ucma_cmd_table[])(struct ucma_file *file,
 				   const char __user *inbuf,
 				   int in_len, int out_len) = {
@@ -750,6 +946,8 @@ static ssize_t (*ucma_cmd_table[])(struct ucma_file *file,
 	[RDMA_USER_CM_CMD_GET_OPTION]	= NULL,
 	[RDMA_USER_CM_CMD_SET_OPTION]	= NULL,
 	[RDMA_USER_CM_CMD_NOTIFY]	= ucma_notify,
+	[RDMA_USER_CM_CMD_JOIN_MCAST]	= ucma_join_multicast,
+	[RDMA_USER_CM_CMD_LEAVE_MCAST]	= ucma_leave_multicast,
 };
 
 static ssize_t ucma_write(struct file *filp, const char __user *buf,

drivers/infiniband/hw/cxgb3/cxio_dbg.c

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU

drivers/infiniband/hw/cxgb3/cxio_hal.c

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU

drivers/infiniband/hw/cxgb3/cxio_hal.h

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU

drivers/infiniband/hw/cxgb3/cxio_resource.c

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU

drivers/infiniband/hw/cxgb3/cxio_resource.h

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU

drivers/infiniband/hw/cxgb3/cxio_wr.h

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU

drivers/infiniband/hw/cxgb3/iwch.c

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU

drivers/infiniband/hw/cxgb3/iwch.h

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU

drivers/infiniband/hw/cxgb3/iwch_cm.c

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU

drivers/infiniband/hw/cxgb3/iwch_cm.h

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU

drivers/infiniband/hw/cxgb3/iwch_cq.c

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU

drivers/infiniband/hw/cxgb3/iwch_ev.c

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU

drivers/infiniband/hw/cxgb3/iwch_mem.c

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU

drivers/infiniband/hw/cxgb3/iwch_provider.c

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU

drivers/infiniband/hw/cxgb3/iwch_provider.h

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU

drivers/infiniband/hw/cxgb3/iwch_qp.c

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU
@@ -846,6 +845,8 @@ int iwch_modify_qp(struct iwch_dev *rhp, struct iwch_qp *qhp,
 			break;
 		case IWCH_QP_STATE_TERMINATE:
 			qhp->attr.state = IWCH_QP_STATE_TERMINATE;
+			if (t3b_device(qhp->rhp))
+				cxio_set_wq_in_error(&qhp->wq);
 			if (!internal)
 				terminate = 1;
 			break;

drivers/infiniband/hw/cxgb3/iwch_user.h

 /*
  * Copyright (c) 2006 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU

drivers/infiniband/hw/ehca/Kconfig

@@ -7,11 +7,3 @@ config INFINIBAND_EHCA
 	To compile the driver as a module, choose M here. The module
 	will be called ib_ehca.
 
-config INFINIBAND_EHCA_SCALING
-	bool "Scaling support (EXPERIMENTAL)"
-	depends on IBMEBUS && INFINIBAND_EHCA && HOTPLUG_CPU && EXPERIMENTAL
-	default y
-	---help---
-	eHCA scaling support schedules the CQ callbacks to different CPUs.
-
-	To enable this feature choose Y here.

drivers/infiniband/hw/ehca/ehca_classes.h

@@ -42,8 +42,6 @@
 #ifndef __EHCA_CLASSES_H__
 #define __EHCA_CLASSES_H__
 
-#include "ehca_classes.h"
-#include "ipz_pt_fn.h"
 
 struct ehca_module;
 struct ehca_qp;
@@ -54,14 +52,22 @@ struct ehca_mw;
 struct ehca_pd;
 struct ehca_av;
 
+#include <rdma/ib_verbs.h>
+#include <rdma/ib_user_verbs.h>
+
 #ifdef CONFIG_PPC64
 #include "ehca_classes_pSeries.h"
 #endif
+#include "ipz_pt_fn.h"
+#include "ehca_qes.h"
+#include "ehca_irq.h"
 
-#include <rdma/ib_verbs.h>
-#include <rdma/ib_user_verbs.h>
+#define EHCA_EQE_CACHE_SIZE 20
 
-#include "ehca_irq.h"
+struct ehca_eqe_cache_entry {
+	struct ehca_eqe *eqe;
+	struct ehca_cq *cq;
+};
 
 struct ehca_eq {
 	u32 length;
@@ -74,6 +80,8 @@ struct ehca_eq {
 	spinlock_t spinlock;
 	struct tasklet_struct interrupt_task;
 	u32 ist;
+	spinlock_t irq_spinlock;
+	struct ehca_eqe_cache_entry eqe_cache[EHCA_EQE_CACHE_SIZE];
 };
 
 struct ehca_sport {
@@ -269,6 +277,7 @@ extern struct idr ehca_cq_idr;
 extern int ehca_static_rate;
 extern int ehca_port_act_time;
 extern int ehca_use_hp_mr;
+extern int ehca_scaling_code;
 
 struct ipzu_queue_resp {
 	u32 qe_size;      /* queue entry size */

drivers/infiniband/hw/ehca/ehca_eq.c

@@ -61,6 +61,7 @@ int ehca_create_eq(struct ehca_shca *shca,
 	struct ib_device *ib_dev = &shca->ib_device;
 
 	spin_lock_init(&eq->spinlock);
+	spin_lock_init(&eq->irq_spinlock);
 	eq->is_initialized = 0;
 
 	if (type != EHCA_EQ && type != EHCA_NEQ) {

drivers/infiniband/hw/ehca/ehca_hca.c

@@ -162,6 +162,9 @@ int ehca_query_port(struct ib_device *ibdev,
 	props->active_width    = IB_WIDTH_12X;
 	props->active_speed    = 0x1;
 
+	/* at the moment (logical) link state is always LINK_UP */
+	props->phys_state      = 0x5;
+
 query_port1:
 	ehca_free_fw_ctrlblock(rblock);
 

drivers/infiniband/hw/ehca/ehca_irq.h

@@ -56,6 +56,7 @@ void ehca_tasklet_neq(unsigned long data);
 
 irqreturn_t ehca_interrupt_eq(int irq, void *dev_id);
 void ehca_tasklet_eq(unsigned long data);
+void ehca_process_eq(struct ehca_shca *shca, int is_irq);
 
 struct ehca_cpu_comp_task {
 	wait_queue_head_t wait_queue;

drivers/infiniband/hw/ehca/ehca_main.c

@@ -52,7 +52,7 @@
 MODULE_LICENSE("Dual BSD/GPL");
 MODULE_AUTHOR("Christoph Raisch <raisch@de.ibm.com>");
 MODULE_DESCRIPTION("IBM eServer HCA InfiniBand Device Driver");
-MODULE_VERSION("SVNEHCA_0020");
+MODULE_VERSION("SVNEHCA_0021");
 
 int ehca_open_aqp1     = 0;
 int ehca_debug_level   = 0;
@@ -62,6 +62,7 @@ int ehca_use_hp_mr     = 0;
 int ehca_port_act_time = 30;
 int ehca_poll_all_eqs  = 1;
 int ehca_static_rate   = -1;
+int ehca_scaling_code  = 1;
 
 module_param_named(open_aqp1,     ehca_open_aqp1,     int, 0);
 module_param_named(debug_level,   ehca_debug_level,   int, 0);
@@ -71,6 +72,7 @@ module_param_named(use_hp_mr,     ehca_use_hp_mr,     int, 0);
 module_param_named(port_act_time, ehca_port_act_time, int, 0);
 module_param_named(poll_all_eqs,  ehca_poll_all_eqs,  int, 0);
 module_param_named(static_rate,   ehca_static_rate,   int, 0);
+module_param_named(scaling_code,   ehca_scaling_code,   int, 0);
 
 MODULE_PARM_DESC(open_aqp1,
 		 "AQP1 on startup (0: no (default), 1: yes)");
@@ -91,6 +93,8 @@ MODULE_PARM_DESC(poll_all_eqs,
 		 " (0: no, 1: yes (default))");
 MODULE_PARM_DESC(static_rate,
 		 "set permanent static rate (default: disabled)");
+MODULE_PARM_DESC(scaling_code,
+		 "set scaling code (0: disabled, 1: enabled/default)");
 
 spinlock_t ehca_qp_idr_lock;
 spinlock_t ehca_cq_idr_lock;
@@ -778,8 +782,24 @@ void ehca_poll_eqs(unsigned long data)
 
 	spin_lock(&shca_list_lock);
 	list_for_each_entry(shca, &shca_list, shca_list) {
-		if (shca->eq.is_initialized)
-			ehca_tasklet_eq((unsigned long)(void*)shca);
+		if (shca->eq.is_initialized) {
+			/* call deadman proc only if eq ptr does not change */
+			struct ehca_eq *eq = &shca->eq;
+			int max = 3;
+			volatile u64 q_ofs, q_ofs2;
+			u64 flags;
+			spin_lock_irqsave(&eq->spinlock, flags);
+			q_ofs = eq->ipz_queue.current_q_offset;
+			spin_unlock_irqrestore(&eq->spinlock, flags);
+			do {
+				spin_lock_irqsave(&eq->spinlock, flags);
+				q_ofs2 = eq->ipz_queue.current_q_offset;
+				spin_unlock_irqrestore(&eq->spinlock, flags);
+				max--;
+			} while (q_ofs == q_ofs2 && max > 0);
+			if (q_ofs == q_ofs2)
+				ehca_process_eq(shca, 0);
+		}
 	}
 	mod_timer(&poll_eqs_timer, jiffies + HZ);
 	spin_unlock(&shca_list_lock);
@@ -790,7 +810,7 @@ int __init ehca_module_init(void)
 	int ret;
 
 	printk(KERN_INFO "eHCA Infiniband Device Driver "
-	                 "(Rel.: SVNEHCA_0020)\n");
+	       "(Rel.: SVNEHCA_0021)\n");
 	idr_init(&ehca_qp_idr);
 	idr_init(&ehca_cq_idr);
 	spin_lock_init(&ehca_qp_idr_lock);

drivers/infiniband/hw/ehca/ipz_pt_fn.h

@@ -247,6 +247,15 @@ static inline void *ipz_eqit_eq_get_inc_valid(struct ipz_queue *queue)
 	return ret;
 }
 
+static inline void *ipz_eqit_eq_peek_valid(struct ipz_queue *queue)
+{
+	void *ret = ipz_qeit_get(queue);
+	u32 qe = *(u8 *) ret;
+	if ((qe >> 7) != (queue->toggle_state & 1))
+		return NULL;
+	return ret;
+}
+
 /* returns address (GX) of first queue entry */
 static inline u64 ipz_qpt_get_firstpage(struct ipz_qpt *qpt)
 {

drivers/infiniband/hw/ipath/ipath_dma.c

@@ -96,7 +96,7 @@ static void ipath_dma_unmap_page(struct ib_device *dev,
 	BUG_ON(!valid_dma_direction(direction));
 }
 
-int ipath_map_sg(struct ib_device *dev, struct scatterlist *sg, int nents,
+static int ipath_map_sg(struct ib_device *dev, struct scatterlist *sg, int nents,
 			enum dma_data_direction direction)
 {
 	u64 addr;

drivers/infiniband/hw/mthca/mthca_memfree.c

@@ -175,7 +175,9 @@ struct mthca_icm *mthca_alloc_icm(struct mthca_dev *dev, int npages,
 		if (!ret) {
 			++chunk->npages;
 
-			if (!coherent && chunk->npages == MTHCA_ICM_CHUNK_LEN) {
+			if (coherent)
+				++chunk->nsg;
+			else if (chunk->npages == MTHCA_ICM_CHUNK_LEN) {
 				chunk->nsg = pci_map_sg(dev->pdev, chunk->mem,
 							chunk->npages,
 							PCI_DMA_BIDIRECTIONAL);

drivers/infiniband/hw/mthca/mthca_qp.c

@@ -573,6 +573,11 @@ int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask,
 		goto out;
 	}
 
+	if (cur_state == new_state && cur_state == IB_QPS_RESET) {
+		err = 0;
+		goto out;
+	}
+
 	if ((attr_mask & IB_QP_PKEY_INDEX) &&
 	     attr->pkey_index >= dev->limits.pkey_table_len) {
 		mthca_dbg(dev, "P_Key index (%u) too large. max is %d\n",

drivers/infiniband/ulp/ipoib/ipoib_cm.c

@@ -145,7 +145,7 @@ static int ipoib_cm_alloc_rx_skb(struct net_device *dev, int id,
 	for (; i >= 0; --i)
 		ib_dma_unmap_single(priv->ca, mapping[i + 1], PAGE_SIZE, DMA_FROM_DEVICE);
 
-	kfree_skb(skb);
+	dev_kfree_skb_any(skb);
 	return -ENOMEM;
 }
 
@@ -1138,7 +1138,7 @@ static ssize_t set_mode(struct device *d, struct device_attribute *attr,
 	return -EINVAL;
 }
 
-static DEVICE_ATTR(mode, S_IWUGO | S_IRUGO, show_mode, set_mode);
+static DEVICE_ATTR(mode, S_IWUSR | S_IRUGO, show_mode, set_mode);
 
 int ipoib_cm_add_mode_attr(struct net_device *dev)
 {

drivers/infiniband/ulp/ipoib/ipoib_multicast.c

@@ -60,14 +60,11 @@ static DEFINE_MUTEX(mcast_mutex);
 /* Used for all multicast joins (broadcast, IPv4 mcast and IPv6 mcast) */
 struct ipoib_mcast {
 	struct ib_sa_mcmember_rec mcmember;
+	struct ib_sa_multicast	 *mc;
 	struct ipoib_ah          *ah;
 
 	struct rb_node    rb_node;
 	struct list_head  list;
-	struct completion done;
-
-	int                 query_id;
-	struct ib_sa_query *query;
 
 	unsigned long created;
 	unsigned long backoff;
@@ -299,18 +296,22 @@ static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast,
 	return 0;
 }
 
-static void
+static int
 ipoib_mcast_sendonly_join_complete(int status,
-				   struct ib_sa_mcmember_rec *mcmember,
-				   void *mcast_ptr)
+				   struct ib_sa_multicast *multicast)
 {
-	struct ipoib_mcast *mcast = mcast_ptr;
+	struct ipoib_mcast *mcast = multicast->context;
 	struct net_device *dev = mcast->dev;
 	struct ipoib_dev_priv *priv = netdev_priv(dev);
 
+	/* We trap for port events ourselves. */
+	if (status == -ENETRESET)
+		return 0;
+
 	if (!status)
-		ipoib_mcast_join_finish(mcast, mcmember);
-	else {
+		status = ipoib_mcast_join_finish(mcast, &multicast->rec);
+
+	if (status) {
 		if (mcast->logcount++ < 20)
 			ipoib_dbg_mcast(netdev_priv(dev), "multicast join failed for "
 					IPOIB_GID_FMT ", status %d\n",
@@ -325,11 +326,10 @@ ipoib_mcast_sendonly_join_complete(int status,
 		spin_unlock_irq(&priv->tx_lock);
 
 		/* Clear the busy flag so we try again */
-		clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
-		mcast->query = NULL;
+		status = test_and_clear_bit(IPOIB_MCAST_FLAG_BUSY,
+					    &mcast->flags);
 	}
-
-	complete(&mcast->done);
+	return status;
 }
 
 static int ipoib_mcast_sendonly_join(struct ipoib_mcast *mcast)
@@ -359,35 +359,33 @@ static int ipoib_mcast_sendonly_join(struct ipoib_mcast *mcast)
 	rec.port_gid = priv->local_gid;
 	rec.pkey     = cpu_to_be16(priv->pkey);
 
-	init_completion(&mcast->done);
-
-	ret = ib_sa_mcmember_rec_set(&ipoib_sa_client, priv->ca, priv->port, &rec,
+	mcast->mc = ib_sa_join_multicast(&ipoib_sa_client, priv->ca,
+					 priv->port, &rec,
 					 IB_SA_MCMEMBER_REC_MGID	|
 					 IB_SA_MCMEMBER_REC_PORT_GID	|
 					 IB_SA_MCMEMBER_REC_PKEY	|
 					 IB_SA_MCMEMBER_REC_JOIN_STATE,
-				     1000, GFP_ATOMIC,
+					 GFP_ATOMIC,
 					 ipoib_mcast_sendonly_join_complete,
-				     mcast, &mcast->query);
-	if (ret < 0) {
-		ipoib_warn(priv, "ib_sa_mcmember_rec_set failed (ret = %d)\n",
+					 mcast);
+	if (IS_ERR(mcast->mc)) {
+		ret = PTR_ERR(mcast->mc);
+		clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
+		ipoib_warn(priv, "ib_sa_join_multicast failed (ret = %d)\n",
 			   ret);
 	} else {
 		ipoib_dbg_mcast(priv, "no multicast record for " IPOIB_GID_FMT
 				", starting join\n",
 				IPOIB_GID_ARG(mcast->mcmember.mgid));
-
-		mcast->query_id = ret;
 	}
 
 	return ret;
 }
 
-static void ipoib_mcast_join_complete(int status,
-				      struct ib_sa_mcmember_rec *mcmember,
-				      void *mcast_ptr)
+static int ipoib_mcast_join_complete(int status,
+				     struct ib_sa_multicast *multicast)
 {
-	struct ipoib_mcast *mcast = mcast_ptr;
+	struct ipoib_mcast *mcast = multicast->context;
 	struct net_device *dev = mcast->dev;
 	struct ipoib_dev_priv *priv = netdev_priv(dev);
 
@@ -395,24 +393,25 @@ static void ipoib_mcast_join_complete(int status,
 			" (status %d)\n",
 			IPOIB_GID_ARG(mcast->mcmember.mgid), status);
 
-	if (!status && !ipoib_mcast_join_finish(mcast, mcmember)) {
+	/* We trap for port events ourselves. */
+	if (status == -ENETRESET)
+		return 0;
+
+	if (!status)
+		status = ipoib_mcast_join_finish(mcast, &multicast->rec);
+
+	if (!status) {
 		mcast->backoff = 1;
 		mutex_lock(&mcast_mutex);
 		if (test_bit(IPOIB_MCAST_RUN, &priv->flags))
 			queue_delayed_work(ipoib_workqueue,
 					   &priv->mcast_task, 0);
 		mutex_unlock(&mcast_mutex);
-		complete(&mcast->done);
-		return;
-	}
-
-	if (status == -EINTR) {
-		complete(&mcast->done);
-		return;
+		return 0;
 	}
 
-	if (status && mcast->logcount++ < 20) {
-		if (status == -ETIMEDOUT || status == -EINTR) {
+	if (mcast->logcount++ < 20) {
+		if (status == -ETIMEDOUT) {
 			ipoib_dbg_mcast(priv, "multicast join failed for " IPOIB_GID_FMT
 					", status %d\n",
 					IPOIB_GID_ARG(mcast->mcmember.mgid),
@@ -429,24 +428,18 @@ static void ipoib_mcast_join_complete(int status,
 	if (mcast->backoff > IPOIB_MAX_BACKOFF_SECONDS)
 		mcast->backoff = IPOIB_MAX_BACKOFF_SECONDS;
 
-	mutex_lock(&mcast_mutex);
+	/* Clear the busy flag so we try again */
+	status = test_and_clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
 
+	mutex_lock(&mcast_mutex);
 	spin_lock_irq(&priv->lock);
-	mcast->query = NULL;
-
-	if (test_bit(IPOIB_MCAST_RUN, &priv->flags)) {
-		if (status == -ETIMEDOUT)
-			queue_delayed_work(ipoib_workqueue, &priv->mcast_task,
-					   0);
-		else
+	if (test_bit(IPOIB_MCAST_RUN, &priv->flags))
 		queue_delayed_work(ipoib_workqueue, &priv->mcast_task,
 				   mcast->backoff * HZ);
-	} else
-		complete(&mcast->done);
 	spin_unlock_irq(&priv->lock);
 	mutex_unlock(&mcast_mutex);
 
-	return;
+	return status;
 }
 
 static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast,
@@ -495,15 +488,14 @@ static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast,
 		rec.hop_limit	  = priv->broadcast->mcmember.hop_limit;
 	}
 
-	init_completion(&mcast->done);
-
-	ret = ib_sa_mcmember_rec_set(&ipoib_sa_client, priv->ca, priv->port,
-				     &rec, comp_mask, mcast->backoff * 1000,
-				     GFP_ATOMIC, ipoib_mcast_join_complete,
-				     mcast, &mcast->query);
-
-	if (ret < 0) {
-		ipoib_warn(priv, "ib_sa_mcmember_rec_set failed, status %d\n", ret);
+	set_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
+	mcast->mc = ib_sa_join_multicast(&ipoib_sa_client, priv->ca, priv->port,
+					 &rec, comp_mask, GFP_KERNEL,
+					 ipoib_mcast_join_complete, mcast);
+	if (IS_ERR(mcast->mc)) {
+		clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags);
+		ret = PTR_ERR(mcast->mc);
+		ipoib_warn(priv, "ib_sa_join_multicast failed, status %d\n", ret);
 
 		mcast->backoff *= 2;
 		if (mcast->backoff > IPOIB_MAX_BACKOFF_SECONDS)
@@ -515,8 +507,7 @@ static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast,
 					   &priv->mcast_task,
 					   mcast->backoff * HZ);
 		mutex_unlock(&mcast_mutex);
-	} else
-		mcast->query_id = ret;
+	}
 }
 
 void ipoib_mcast_join_task(struct work_struct *work)
@@ -568,6 +559,7 @@ void ipoib_mcast_join_task(struct work_struct *work)
 	}
 
 	if (!test_bit(IPOIB_MCAST_FLAG_ATTACHED, &priv->broadcast->flags)) {
+		if (!test_bit(IPOIB_MCAST_FLAG_BUSY, &priv->broadcast->flags))
 			ipoib_mcast_join(dev, priv->broadcast, 0);
 		return;
 	}
@@ -625,26 +617,9 @@ int ipoib_mcast_start_thread(struct net_device *dev)
 	return 0;
 }
 
-static void wait_for_mcast_join(struct ipoib_dev_priv *priv,
-				struct ipoib_mcast *mcast)
-{
-	spin_lock_irq(&priv->lock);
-	if (mcast && mcast->query) {
-		ib_sa_cancel_query(mcast->query_id, mcast->query);
-		mcast->query = NULL;
-		spin_unlock_irq(&priv->lock);
-		ipoib_dbg_mcast(priv, "waiting for MGID " IPOIB_GID_FMT "\n",
-				IPOIB_GID_ARG(mcast->mcmember.mgid));
-		wait_for_completion(&mcast->done);
-	}
-	else
-		spin_unlock_irq(&priv->lock);
-}
-
 int ipoib_mcast_stop_thread(struct net_device *dev, int flush)
 {
 	struct ipoib_dev_priv *priv = netdev_priv(dev);
-	struct ipoib_mcast *mcast;
 
 	ipoib_dbg_mcast(priv, "stopping multicast thread\n");
 
@@ -660,52 +635,27 @@ int ipoib_mcast_stop_thread(struct net_device *dev, int flush)
 	if (flush)
 		flush_workqueue(ipoib_workqueue);
 
-	wait_for_mcast_join(priv, priv->broadcast);
-
-	list_for_each_entry(mcast, &priv->multicast_list, list)
-		wait_for_mcast_join(priv, mcast);
-
 	return 0;
 }
 
 static int ipoib_mcast_leave(struct net_device *dev, struct ipoib_mcast *mcast)
 {
 	struct ipoib_dev_priv *priv = netdev_priv(dev);
-	struct ib_sa_mcmember_rec rec = {
-		.join_state = 1
-	};
 	int ret = 0;
 
-	if (!test_and_clear_bit(IPOIB_MCAST_FLAG_ATTACHED, &mcast->flags))
-		return 0;
-
+	if (test_and_clear_bit(IPOIB_MCAST_FLAG_ATTACHED, &mcast->flags)) {
 		ipoib_dbg_mcast(priv, "leaving MGID " IPOIB_GID_FMT "\n",
 				IPOIB_GID_ARG(mcast->mcmember.mgid));
 
-	rec.mgid     = mcast->mcmember.mgid;
-	rec.port_gid = priv->local_gid;
-	rec.pkey     = cpu_to_be16(priv->pkey);
-
 		/* Remove ourselves from the multicast group */
 		ret = ipoib_mcast_detach(dev, be16_to_cpu(mcast->mcmember.mlid),
 					 &mcast->mcmember.mgid);
 		if (ret)
 			ipoib_warn(priv, "ipoib_mcast_detach failed (result = %d)\n", ret);
+	}
 
-	/*
-	 * Just make one shot at leaving and don't wait for a reply;
-	 * if we fail, too bad.
-	 */
-	ret = ib_sa_mcmember_rec_delete(&ipoib_sa_client, priv->ca, priv->port, &rec,
-					IB_SA_MCMEMBER_REC_MGID		|
-					IB_SA_MCMEMBER_REC_PORT_GID	|
-					IB_SA_MCMEMBER_REC_PKEY		|
-					IB_SA_MCMEMBER_REC_JOIN_STATE,
-					0, GFP_ATOMIC, NULL,
-					mcast, &mcast->query);
-	if (ret < 0)
-		ipoib_warn(priv, "ib_sa_mcmember_rec_delete failed "
-			   "for leave (result = %d)\n", ret);
+	if (test_and_clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags))
+		ib_sa_free_multicast(mcast->mc);
 
 	return 0;
 }
@@ -758,7 +708,7 @@ void ipoib_mcast_send(struct net_device *dev, void *mgid, struct sk_buff *skb)
 			dev_kfree_skb_any(skb);
 		}
 
-		if (mcast->query)
+		if (test_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags))
 			ipoib_dbg_mcast(priv, "no address vector, "
 					"but multicast join already started\n");
 		else if (test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags))
@@ -916,7 +866,6 @@ void ipoib_mcast_restart_task(struct work_struct *work)
 
 	/* We have to cancel outside of the spinlock */
 	list_for_each_entry_safe(mcast, tmcast, &remove_list, list) {
-		wait_for_mcast_join(priv, mcast);
 		ipoib_mcast_leave(mcast->dev, mcast);
 		ipoib_mcast_free(mcast);
 	}

include/rdma/ib_addr.h

@@ -110,6 +110,12 @@ static inline void ib_addr_set_pkey(struct rdma_dev_addr *dev_addr, u16 pkey)
 	dev_addr->broadcast[9] = (unsigned char) pkey;
 }
 
+static inline void ib_addr_get_mgid(struct rdma_dev_addr *dev_addr,
+				    union ib_gid *gid)
+{
+	memcpy(gid, dev_addr->broadcast + 4, sizeof *gid);
+}
+
 static inline void ib_addr_get_sgid(struct rdma_dev_addr *dev_addr,
 				    union ib_gid *gid)
 {

include/rdma/ib_sa.h

@@ -285,18 +285,6 @@ int ib_sa_path_rec_get(struct ib_sa_client *client,
 		       void *context,
 		       struct ib_sa_query **query);
 
-int ib_sa_mcmember_rec_query(struct ib_sa_client *client,
-			     struct ib_device *device, u8 port_num,
-			     u8 method,
-			     struct ib_sa_mcmember_rec *rec,
-			     ib_sa_comp_mask comp_mask,
-			     int timeout_ms, gfp_t gfp_mask,
-			     void (*callback)(int status,
-					      struct ib_sa_mcmember_rec *resp,
-					      void *context),
-			     void *context,
-			     struct ib_sa_query **query);
-
 int ib_sa_service_rec_query(struct ib_sa_client *client,
 			 struct ib_device *device, u8 port_num,
 			 u8 method,
@@ -309,93 +297,82 @@ int ib_sa_service_rec_query(struct ib_sa_client *client,
 			 void *context,
 			 struct ib_sa_query **sa_query);
 
+struct ib_sa_multicast {
+	struct ib_sa_mcmember_rec rec;
+	ib_sa_comp_mask		comp_mask;
+	int			(*callback)(int status,
+					    struct ib_sa_multicast *multicast);
+	void			*context;
+};
+
 /**
- * ib_sa_mcmember_rec_set - Start an MCMember set query
- * @client:SA client
- * @device:device to send query on
- * @port_num: port number to send query on
- * @rec:MCMember Record to send in query
- * @comp_mask:component mask to send in query
- * @timeout_ms:time to wait for response
- * @gfp_mask:GFP mask to use for internal allocations
- * @callback:function called when query completes, times out or is
- * canceled
- * @context:opaque user context passed to callback
- * @sa_query:query context, used to cancel query
+ * ib_sa_join_multicast - Initiates a join request to the specified multicast
+ *   group.
+ * @client: SA client
+ * @device: Device associated with the multicast group.
+ * @port_num: Port on the specified device to associate with the multicast
+ *   group.
+ * @rec: SA multicast member record specifying group attributes.
+ * @comp_mask: Component mask indicating which group attributes of %rec are
+ *   valid.
+ * @gfp_mask: GFP mask for memory allocations.
+ * @callback: User callback invoked once the join operation completes.
+ * @context: User specified context stored with the ib_sa_multicast structure.
  *
- * Send an MCMember Set query to the SA (eg to join a multicast
- * group).  The callback function will be called when the query
- * completes (or fails); status is 0 for a successful response, -EINTR
- * if the query is canceled, -ETIMEDOUT is the query timed out, or
- * -EIO if an error occurred sending the query.  The resp parameter of
- * the callback is only valid if status is 0.
+ * This call initiates a multicast join request with the SA for the specified
+ * multicast group.  If the join operation is started successfully, it returns
+ * an ib_sa_multicast structure that is used to track the multicast operation.
+ * Users must free this structure by calling ib_free_multicast, even if the
+ * join operation later fails.  (The callback status is non-zero.)
  *
- * If the return value of ib_sa_mcmember_rec_set() is negative, it is
- * an error code.  Otherwise it is a query ID that can be used to
- * cancel the query.
+ * If the join operation fails; status will be non-zero, with the following
+ * failures possible:
+ * -ETIMEDOUT: The request timed out.
+ * -EIO: An error occurred sending the query.
+ * -EINVAL: The MCMemberRecord values differed from the existing group's.
+ * -ENETRESET: Indicates that an fatal error has occurred on the multicast
+ *   group, and the user must rejoin the group to continue using it.
  */
-static inline int
-ib_sa_mcmember_rec_set(struct ib_sa_client *client,
+struct ib_sa_multicast *ib_sa_join_multicast(struct ib_sa_client *client,
 					     struct ib_device *device, u8 port_num,
 					     struct ib_sa_mcmember_rec *rec,
-		       ib_sa_comp_mask comp_mask,
-		       int timeout_ms, gfp_t gfp_mask,
-		       void (*callback)(int status,
-					struct ib_sa_mcmember_rec *resp,
-					void *context),
-		       void *context,
-		       struct ib_sa_query **query)
-{
-	return ib_sa_mcmember_rec_query(client, device, port_num,
-					IB_MGMT_METHOD_SET,
-					rec, comp_mask,
-					timeout_ms, gfp_mask, callback,
-					context, query);
-}
+					     ib_sa_comp_mask comp_mask, gfp_t gfp_mask,
+					     int (*callback)(int status,
+							     struct ib_sa_multicast
+								    *multicast),
+					     void *context);
 
 /**
- * ib_sa_mcmember_rec_delete - Start an MCMember delete query
- * @client:SA client
- * @device:device to send query on
- * @port_num: port number to send query on
- * @rec:MCMember Record to send in query
- * @comp_mask:component mask to send in query
- * @timeout_ms:time to wait for response
- * @gfp_mask:GFP mask to use for internal allocations
- * @callback:function called when query completes, times out or is
- * canceled
- * @context:opaque user context passed to callback
- * @sa_query:query context, used to cancel query
- *
- * Send an MCMember Delete query to the SA (eg to leave a multicast
- * group).  The callback function will be called when the query
- * completes (or fails); status is 0 for a successful response, -EINTR
- * if the query is canceled, -ETIMEDOUT is the query timed out, or
- * -EIO if an error occurred sending the query.  The resp parameter of
- * the callback is only valid if status is 0.
+ * ib_free_multicast - Frees the multicast tracking structure, and releases
+ *    any reference on the multicast group.
+ * @multicast: Multicast tracking structure allocated by ib_join_multicast.
  *
- * If the return value of ib_sa_mcmember_rec_delete() is negative, it
- * is an error code.  Otherwise it is a query ID that can be used to
- * cancel the query.
+ * This call blocks until the multicast identifier is destroyed.  It may
+ * not be called from within the multicast callback; however, returning a non-
+ * zero value from the callback will result in destroying the multicast
+ * tracking structure.
  */
-static inline int
-ib_sa_mcmember_rec_delete(struct ib_sa_client *client,
-			  struct ib_device *device, u8 port_num,
+void ib_sa_free_multicast(struct ib_sa_multicast *multicast);
+
+/**
+ * ib_get_mcmember_rec - Looks up a multicast member record by its MGID and
+ *   returns it if found.
+ * @device: Device associated with the multicast group.
+ * @port_num: Port on the specified device to associate with the multicast
+ *   group.
+ * @mgid: MGID of multicast group.
+ * @rec: Location to copy SA multicast member record.
+ */
+int ib_sa_get_mcmember_rec(struct ib_device *device, u8 port_num,
+			   union ib_gid *mgid, struct ib_sa_mcmember_rec *rec);
+
+/**
+ * ib_init_ah_from_mcmember - Initialize address handle attributes based on
+ * an SA multicast member record.
+ */
+int ib_init_ah_from_mcmember(struct ib_device *device, u8 port_num,
 			     struct ib_sa_mcmember_rec *rec,
-			  ib_sa_comp_mask comp_mask,
-			  int timeout_ms, gfp_t gfp_mask,
-			  void (*callback)(int status,
-					   struct ib_sa_mcmember_rec *resp,
-					   void *context),
-			  void *context,
-			  struct ib_sa_query **query)
-{
-	return ib_sa_mcmember_rec_query(client, device, port_num,
-					IB_SA_METHOD_DELETE,
-					rec, comp_mask,
-					timeout_ms, gfp_mask, callback,
-					context, query);
-}
+			     struct ib_ah_attr *ah_attr);
 
 /**
  * ib_init_ah_from_path - Initialize address handle attributes based on an SA

include/rdma/rdma_cm.h

@@ -52,10 +52,13 @@ enum rdma_cm_event_type {
 	RDMA_CM_EVENT_ESTABLISHED,
 	RDMA_CM_EVENT_DISCONNECTED,
 	RDMA_CM_EVENT_DEVICE_REMOVAL,
+	RDMA_CM_EVENT_MULTICAST_JOIN,
+	RDMA_CM_EVENT_MULTICAST_ERROR
 };
 
 enum rdma_port_space {
 	RDMA_PS_SDP  = 0x0001,
+	RDMA_PS_IPOIB= 0x0002,
 	RDMA_PS_TCP  = 0x0106,
 	RDMA_PS_UDP  = 0x0111,
 	RDMA_PS_SCTP = 0x0183
@@ -294,5 +297,21 @@ int rdma_reject(struct rdma_cm_id *id, const void *private_data,
  */
 int rdma_disconnect(struct rdma_cm_id *id);
 
-#endif /* RDMA_CM_H */
+/**
+ * rdma_join_multicast - Join the multicast group specified by the given
+ *   address.
+ * @id: Communication identifier associated with the request.
+ * @addr: Multicast address identifying the group to join.
+ * @context: User-defined context associated with the join request, returned
+ * to the user through the private_data pointer in multicast events.
+ */
+int rdma_join_multicast(struct rdma_cm_id *id, struct sockaddr *addr,
+			void *context);
 
+/**
+ * rdma_leave_multicast - Leave the multicast group specified by the given
+ *   address.
+ */
+void rdma_leave_multicast(struct rdma_cm_id *id, struct sockaddr *addr);
+
+#endif /* RDMA_CM_H */

include/rdma/rdma_cm_ib.h

@@ -44,7 +44,7 @@
 int rdma_set_ib_paths(struct rdma_cm_id *id,
 		      struct ib_sa_path_rec *path_rec, int num_paths);
 
-/* Global qkey for UD QPs and multicast groups. */
-#define RDMA_UD_QKEY 0x01234567
+/* Global qkey for UDP QPs and multicast groups. */
+#define RDMA_UDP_QKEY 0x01234567
 
 #endif /* RDMA_CM_IB_H */

include/rdma/rdma_user_cm.h

@@ -38,7 +38,7 @@
 #include <rdma/ib_user_verbs.h>
 #include <rdma/ib_user_sa.h>
 
-#define RDMA_USER_CM_ABI_VERSION	3
+#define RDMA_USER_CM_ABI_VERSION	4
 
 #define RDMA_MAX_PRIVATE_DATA		256
 
@@ -58,7 +58,9 @@ enum {
 	RDMA_USER_CM_CMD_GET_EVENT,
 	RDMA_USER_CM_CMD_GET_OPTION,
 	RDMA_USER_CM_CMD_SET_OPTION,
-	RDMA_USER_CM_CMD_NOTIFY
+	RDMA_USER_CM_CMD_NOTIFY,
+	RDMA_USER_CM_CMD_JOIN_MCAST,
+	RDMA_USER_CM_CMD_LEAVE_MCAST
 };
 
 /*
@@ -188,6 +190,13 @@ struct rdma_ucm_notify {
 	__u32 event;
 };
 
+struct rdma_ucm_join_mcast {
+	__u64 response;		/* rdma_ucm_create_id_resp */
+	__u64 uid;
+	struct sockaddr_in6 addr;
+	__u32 id;
+};
+
 struct rdma_ucm_get_event {
 	__u64 response;
 };