[PATCH V1 for-next 0/2] Support for Send Only Full Member multicast join

[PATCH V1 for-next 0/2] Support for Send Only Full Member multicast join

[Leon Romanovsky]

Added UCMA and CMA support for multicast join flags. Flags are
passed using UCMA CM join command previously reserved fields.
Currently supporting two join flags indicating two different
multicast JoinStates:

1. Full Member:
   The initiator creates the Multicast group(MCG) if it wasn't
   previously created, can send Multicast messages to the group
   and receive messages from the MCG.

2. Send Only Full Member:
   The initiator creates the Multicast group(MCG) if it wasn't
   previously created, can send Multicast messages to the group
   but doesn't receive any messages from the MCG.

Since Send Only Full Member requires SM support a new cached
SA port attribute containing SM ClassPortInfo fields was added
(ClassPortInfo fields: Table 126 IB Spec 1.3.). The attribute
is cached to avoid resending queries, caching is done when a
successful ClassPortInfo reply is received on the port.
Invalidation of the attribute is done on SM change events,
SM re-registration events, and SM LID change events. The fields
in ClassPortInfo should not change during SM runtime without an
event.

The changes compared to v0 of this patch series are:
 * Use multicast.c join state enum instead of the new added bits
   defined by IB_SA_MCMEMBER_REC_JOIN_STATE in ib_sa.h
 * Remove ib_sa_port struct alignment spaces

Alex Vesker (2):
  IB/sa: Add cached attribute containing SM information to SA port
  IB/core: Support for CMA multicast join flags

 drivers/infiniband/core/cma.c       | 98 ++++++++++++++++++++++++++++++++++---
 drivers/infiniband/core/multicast.c | 12 -----
 drivers/infiniband/core/sa_query.c  | 41 ++++++++++++++++
 drivers/infiniband/core/ucma.c      | 18 +++++--
 include/rdma/ib_sa.h                | 13 +++++
 include/rdma/rdma_cm.h              |  4 +-
 include/uapi/rdma/rdma_user_cm.h    |  9 +++-
 7 files changed, 171 insertions(+), 24 deletions(-)

-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH V1 for-next 1/2] IB/sa: Add cached attribute containing SM information to SA port

[Leon Romanovsky]

From: Alex Vesker <valex-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>

Added a new SA port attribute containing SM ClassPortInfo fields,
(ClassPortInfo fields: Table 126 IB Spec 1.3.). This is useful for
checking SM support for specific features. The attribute is cached
to avoid resending queries, caching is done when a successful
ClassPortInfo reply is received on the port. Invalidation of the
attribute is done on SM change events, SM re-registration events,
and SM LID change events. The fields in ClassPortInfo should not
change during SM runtime without an event.

Signed-off-by: Alex Vesker <valex-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
Reviewed by: Hal Rosenstock <hal-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
Signed-off-by: Leon Romanovsky <leon-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
---
 drivers/infiniband/core/sa_query.c | 41 ++++++++++++++++++++++++++++++++++++++
 1 file changed, 41 insertions(+)

diff --git a/drivers/infiniband/core/sa_query.c b/drivers/infiniband/core/sa_query.c
index e955386..b9bf7aa 100644
--- a/drivers/infiniband/core/sa_query.c
+++ b/drivers/infiniband/core/sa_query.c
@@ -65,10 +65,17 @@ struct ib_sa_sm_ah {
 	u8		     src_path_mask;
 };
 
+struct ib_sa_classport_cache {
+	bool valid;
+	struct ib_class_port_info data;
+};
+
 struct ib_sa_port {
 	struct ib_mad_agent *agent;
 	struct ib_sa_sm_ah  *sm_ah;
 	struct work_struct   update_task;
+	struct ib_sa_classport_cache classport_info;
+	spinlock_t                   classport_lock; /* protects class port info set */
 	spinlock_t           ah_lock;
 	u8                   port_num;
 };
@@ -998,6 +1005,13 @@ static void ib_sa_event(struct ib_event_handler *handler, struct ib_event *event
 		port->sm_ah = NULL;
 		spin_unlock_irqrestore(&port->ah_lock, flags);
 
+		if (event->event == IB_EVENT_SM_CHANGE ||
+		    event->event == IB_EVENT_CLIENT_REREGISTER ||
+		    event->event == IB_EVENT_LID_CHANGE) {
+			spin_lock_irqsave(&port->classport_lock, flags);
+			port->classport_info.valid = false;
+			spin_unlock_irqrestore(&port->classport_lock, flags);
+		}
 		queue_work(ib_wq, &sa_dev->port[event->element.port_num -
 					    sa_dev->start_port].update_task);
 	}
@@ -1719,6 +1733,7 @@ static void ib_sa_classport_info_rec_callback(struct ib_sa_query *sa_query,
 					      int status,
 					      struct ib_sa_mad *mad)
 {
+	unsigned long flags;
 	struct ib_sa_classport_info_query *query =
 		container_of(sa_query, struct ib_sa_classport_info_query, sa_query);
 
@@ -1728,6 +1743,16 @@ static void ib_sa_classport_info_rec_callback(struct ib_sa_query *sa_query,
 		ib_unpack(classport_info_rec_table,
 			  ARRAY_SIZE(classport_info_rec_table),
 			  mad->data, &rec);
+
+		spin_lock_irqsave(&sa_query->port->classport_lock, flags);
+		if (!status && !sa_query->port->classport_info.valid) {
+			memcpy(&sa_query->port->classport_info.data, &rec,
+			       sizeof(sa_query->port->classport_info.data));
+
+			sa_query->port->classport_info.valid = true;
+		}
+		spin_unlock_irqrestore(&sa_query->port->classport_lock, flags);
+
 		query->callback(status, &rec, query->context);
 	} else {
 		query->callback(status, NULL, query->context);
@@ -1754,7 +1779,9 @@ int ib_sa_classport_info_rec_query(struct ib_sa_client *client,
 	struct ib_sa_port *port;
 	struct ib_mad_agent *agent;
 	struct ib_sa_mad *mad;
+	struct ib_class_port_info cached_class_port_info;
 	int ret;
+	unsigned long flags;
 
 	if (!sa_dev)
 		return -ENODEV;
@@ -1762,6 +1789,17 @@ int ib_sa_classport_info_rec_query(struct ib_sa_client *client,
 	port  = &sa_dev->port[port_num - sa_dev->start_port];
 	agent = port->agent;
 
+	/* Use cached ClassPortInfo attribute if valid instead of sending mad */
+	spin_lock_irqsave(&port->classport_lock, flags);
+	if (port->classport_info.valid && callback) {
+		memcpy(&cached_class_port_info, &port->classport_info.data,
+		       sizeof(cached_class_port_info));
+		spin_unlock_irqrestore(&port->classport_lock, flags);
+		callback(0, &cached_class_port_info, context);
+		return 0;
+	}
+	spin_unlock_irqrestore(&port->classport_lock, flags);
+
 	query = kzalloc(sizeof(*query), gfp_mask);
 	if (!query)
 		return -ENOMEM;
@@ -1885,6 +1923,9 @@ static void ib_sa_add_one(struct ib_device *device)
 		sa_dev->port[i].sm_ah    = NULL;
 		sa_dev->port[i].port_num = i + s;
 
+		spin_lock_init(&sa_dev->port[i].classport_lock);
+		sa_dev->port[i].classport_info.valid = false;
+
 		sa_dev->port[i].agent =
 			ib_register_mad_agent(device, i + s, IB_QPT_GSI,
 					      NULL, 0, send_handler,
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH V1 for-next 2/2] IB/core: Support for CMA multicast join flags

[Leon Romanovsky]

From: Alex Vesker <valex-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>

Added UCMA and CMA support for multicast join flags. Flags are
passed using UCMA CM join command previously reserved fields.
Currently supporting two join flags indicating two different
multicast JoinStates:

1. Full Member:
   The initiator creates the Multicast group(MCG) if it wasn't
   previously created, can send Multicast messages to the group
   and receive messages from the MCG.

2. Send Only Full Member:
   The initiator creates the Multicast group(MCG) if it wasn't
   previously created, can send Multicast messages to the group
   but doesn't receive any messages from the MCG.

   IB: Send Only Full Member requires a query of ClassPortInfo
       to determine if SM/SA supports this option. If SM/SA
       doesn't support Send-Only there will be no join request
       sent and an error will be returned.

   ETH: When Send Only Full Member is requested no IGMP join
	will be sent.

Signed-off-by: Alex Vesker <valex-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
Reviewed by: Hal Rosenstock <hal-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
Signed-off-by: Leon Romanovsky <leon-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
---
 drivers/infiniband/core/cma.c       | 98 ++++++++++++++++++++++++++++++++++---
 drivers/infiniband/core/multicast.c | 12 -----
 drivers/infiniband/core/ucma.c      | 18 +++++--
 include/rdma/ib_sa.h                | 13 +++++
 include/rdma/rdma_cm.h              |  4 +-
 include/uapi/rdma/rdma_user_cm.h    |  9 +++-
 6 files changed, 130 insertions(+), 24 deletions(-)

diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c
index ad1b1ad..e6dfa1b 100644
--- a/drivers/infiniband/core/cma.c
+++ b/drivers/infiniband/core/cma.c
@@ -68,6 +68,7 @@ MODULE_DESCRIPTION("Generic RDMA CM Agent");
 MODULE_LICENSE("Dual BSD/GPL");
 
 #define CMA_CM_RESPONSE_TIMEOUT 20
+#define CMA_QUERY_CLASSPORT_INFO_TIMEOUT 3000
 #define CMA_MAX_CM_RETRIES 15
 #define CMA_CM_MRA_SETTING (IB_CM_MRA_FLAG_DELAY | 24)
 #define CMA_IBOE_PACKET_LIFETIME 18
@@ -162,6 +163,14 @@ struct rdma_bind_list {
 	unsigned short		port;
 };
 
+struct class_port_info_context {
+	struct ib_class_port_info	*class_port_info;
+	struct ib_device		*device;
+	struct completion		done;
+	struct ib_sa_query		*sa_query;
+	u8				port_num;
+};
+
 static int cma_ps_alloc(struct net *net, enum rdma_port_space ps,
 			struct rdma_bind_list *bind_list, int snum)
 {
@@ -306,6 +315,7 @@ struct cma_multicast {
 	struct sockaddr_storage	addr;
 	struct kref		mcref;
 	bool			igmp_joined;
+	u8			join_state;
 };
 
 struct cma_work {
@@ -3752,10 +3762,63 @@ static void cma_set_mgid(struct rdma_id_private *id_priv,
 	}
 }
 
+static void cma_query_sa_classport_info_cb(int status,
+					   struct ib_class_port_info *rec,
+					   void *context)
+{
+	struct class_port_info_context *cb_ctx = context;
+
+	WARN_ON(!context);
+
+	if (status || !rec) {
+		pr_debug("RDMA CM: %s port %u failed query ClassPortInfo status: %d\n",
+			 cb_ctx->device->name, cb_ctx->port_num, status);
+		goto out;
+	}
+
+	memcpy(cb_ctx->class_port_info, rec, sizeof(struct ib_class_port_info));
+
+out:
+	complete(&cb_ctx->done);
+}
+
+static int cma_query_sa_classport_info(struct ib_device *device, u8 port_num,
+				       struct ib_class_port_info *class_port_info)
+{
+	struct class_port_info_context *cb_ctx;
+	int ret;
+
+	cb_ctx = kmalloc(sizeof(*cb_ctx), GFP_KERNEL);
+	if (!cb_ctx)
+		return -ENOMEM;
+
+	cb_ctx->device = device;
+	cb_ctx->class_port_info = class_port_info;
+	cb_ctx->port_num = port_num;
+	init_completion(&cb_ctx->done);
+
+	ret = ib_sa_classport_info_rec_query(&sa_client, device, port_num,
+					     CMA_QUERY_CLASSPORT_INFO_TIMEOUT,
+					     GFP_KERNEL, cma_query_sa_classport_info_cb,
+					     cb_ctx, &cb_ctx->sa_query);
+	if (ret < 0) {
+		pr_err("RDMA CM: %s port %u failed to send ClassPortInfo query, ret: %d\n",
+		       device->name, port_num, ret);
+		goto out;
+	}
+
+	wait_for_completion(&cb_ctx->done);
+
+out:
+	kfree(cb_ctx);
+	return ret;
+}
+
 static int cma_join_ib_multicast(struct rdma_id_private *id_priv,
 				 struct cma_multicast *mc)
 {
 	struct ib_sa_mcmember_rec rec;
+	struct ib_class_port_info class_port_info;
 	struct rdma_dev_addr *dev_addr = &id_priv->id.route.addr.dev_addr;
 	ib_sa_comp_mask comp_mask;
 	int ret;
@@ -3774,7 +3837,24 @@ static int cma_join_ib_multicast(struct rdma_id_private *id_priv,
 	rec.qkey = cpu_to_be32(id_priv->qkey);
 	rdma_addr_get_sgid(dev_addr, &rec.port_gid);
 	rec.pkey = cpu_to_be16(ib_addr_get_pkey(dev_addr));
-	rec.join_state = 1;
+	rec.join_state = mc->join_state;
+
+	if (rec.join_state == BIT(SENDONLY_FULLMEMBER_JOIN)) {
+		ret = cma_query_sa_classport_info(id_priv->id.device,
+						  id_priv->id.port_num,
+						  &class_port_info);
+
+		if (ret)
+			return ret;
+
+		if (!(ib_get_cpi_capmask2(&class_port_info) &
+		      IB_SA_CAP_MASK2_SENDONLY_FULL_MEM_SUPPORT)) {
+			pr_warn("RDMA CM: %s port %u Unable to multicast join\n"
+				"RDMA CM: SM doesn't support Send Only Full Member option\n",
+				id_priv->id.device->name, id_priv->id.port_num);
+			return -EOPNOTSUPP;
+		}
+	}
 
 	comp_mask = IB_SA_MCMEMBER_REC_MGID | IB_SA_MCMEMBER_REC_PORT_GID |
 		    IB_SA_MCMEMBER_REC_PKEY | IB_SA_MCMEMBER_REC_JOIN_STATE |
@@ -3843,6 +3923,9 @@ static int cma_iboe_join_multicast(struct rdma_id_private *id_priv,
 	struct sockaddr *addr = (struct sockaddr *)&mc->addr;
 	struct net_device *ndev = NULL;
 	enum ib_gid_type gid_type;
+	bool send_only;
+
+	send_only = mc->join_state == BIT(SENDONLY_FULLMEMBER_JOIN);
 
 	if (cma_zero_addr((struct sockaddr *)&mc->addr))
 		return -EINVAL;
@@ -3878,10 +3961,12 @@ static int cma_iboe_join_multicast(struct rdma_id_private *id_priv,
 	if (addr->sa_family == AF_INET) {
 		if (gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP) {
 			mc->multicast.ib->rec.hop_limit = IPV6_DEFAULT_HOPLIMIT;
-			err = cma_igmp_send(ndev, &mc->multicast.ib->rec.mgid,
-					    true);
-			if (!err)
-				mc->igmp_joined = true;
+			if (!send_only) {
+				err = cma_igmp_send(ndev, &mc->multicast.ib->rec.mgid,
+						    true);
+				if (!err)
+					mc->igmp_joined = true;
+			}
 		}
 	} else {
 		if (gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP)
@@ -3911,7 +3996,7 @@ out1:
 }
 
 int rdma_join_multicast(struct rdma_cm_id *id, struct sockaddr *addr,
-			void *context)
+			u8 join_state, void *context)
 {
 	struct rdma_id_private *id_priv;
 	struct cma_multicast *mc;
@@ -3930,6 +4015,7 @@ int rdma_join_multicast(struct rdma_cm_id *id, struct sockaddr *addr,
 	mc->context = context;
 	mc->id_priv = id_priv;
 	mc->igmp_joined = false;
+	mc->join_state = join_state;
 	spin_lock(&id_priv->lock);
 	list_add(&mc->list, &id_priv->mc_list);
 	spin_unlock(&id_priv->lock);
diff --git a/drivers/infiniband/core/multicast.c b/drivers/infiniband/core/multicast.c
index a83ec28..3a3c5d7 100644
--- a/drivers/infiniband/core/multicast.c
+++ b/drivers/infiniband/core/multicast.c
@@ -93,18 +93,6 @@ enum {
 
 struct mcast_member;
 
-/*
-* There are 4 types of join states:
-* FullMember, NonMember, SendOnlyNonMember, SendOnlyFullMember.
-*/
-enum {
-	FULLMEMBER_JOIN,
-	NONMEMBER_JOIN,
-	SENDONLY_NONMEBER_JOIN,
-	SENDONLY_FULLMEMBER_JOIN,
-	NUM_JOIN_MEMBERSHIP_TYPES,
-};
-
 struct mcast_group {
 	struct ib_sa_mcmember_rec rec;
 	struct rb_node		node;
diff --git a/drivers/infiniband/core/ucma.c b/drivers/infiniband/core/ucma.c
index c0f3826..2825ece 100644
--- a/drivers/infiniband/core/ucma.c
+++ b/drivers/infiniband/core/ucma.c
@@ -106,6 +106,7 @@ struct ucma_multicast {
 	int			events_reported;
 
 	u64			uid;
+	u8			join_state;
 	struct list_head	list;
 	struct sockaddr_storage	addr;
 };
@@ -1317,12 +1318,20 @@ static ssize_t ucma_process_join(struct ucma_file *file,
 	struct ucma_multicast *mc;
 	struct sockaddr *addr;
 	int ret;
+	u8 join_state;
 
 	if (out_len < sizeof(resp))
 		return -ENOSPC;
 
 	addr = (struct sockaddr *) &cmd->addr;
-	if (cmd->reserved || !cmd->addr_size || (cmd->addr_size != rdma_addr_size(addr)))
+	if (!cmd->addr_size || (cmd->addr_size != rdma_addr_size(addr)))
+		return -EINVAL;
+
+	if (cmd->join_flags == RDMA_MC_JOIN_FLAG_FULLMEMBER)
+		join_state = BIT(FULLMEMBER_JOIN);
+	else if (cmd->join_flags == RDMA_MC_JOIN_FLAG_SENDONLY_FULLMEMBER)
+		join_state = BIT(SENDONLY_FULLMEMBER_JOIN);
+	else
 		return -EINVAL;
 
 	ctx = ucma_get_ctx(file, cmd->id);
@@ -1335,10 +1344,11 @@ static ssize_t ucma_process_join(struct ucma_file *file,
 		ret = -ENOMEM;
 		goto err1;
 	}
-
+	mc->join_state = join_state;
 	mc->uid = cmd->uid;
 	memcpy(&mc->addr, addr, cmd->addr_size);
-	ret = rdma_join_multicast(ctx->cm_id, (struct sockaddr *) &mc->addr, mc);
+	ret = rdma_join_multicast(ctx->cm_id, (struct sockaddr *)&mc->addr,
+				  join_state, mc);
 	if (ret)
 		goto err2;
 
@@ -1382,7 +1392,7 @@ static ssize_t ucma_join_ip_multicast(struct ucma_file *file,
 	join_cmd.uid = cmd.uid;
 	join_cmd.id = cmd.id;
 	join_cmd.addr_size = rdma_addr_size((struct sockaddr *) &cmd.addr);
-	join_cmd.reserved = 0;
+	join_cmd.join_flags = RDMA_MC_JOIN_FLAG_FULLMEMBER;
 	memcpy(&join_cmd.addr, &cmd.addr, join_cmd.addr_size);
 
 	return ucma_process_join(file, &join_cmd, out_len);
diff --git a/include/rdma/ib_sa.h b/include/rdma/ib_sa.h
index 3840416..5ee7aab 100644
--- a/include/rdma/ib_sa.h
+++ b/include/rdma/ib_sa.h
@@ -94,6 +94,19 @@ enum ib_sa_selector {
 	IB_SA_BEST = 3
 };
 
+/*
+ * There are 4 types of join states:
+ * FullMember, NonMember, SendOnlyNonMember, SendOnlyFullMember.
+ * The order corresponds to JoinState bits in MCMemberRecord.
+ */
+enum ib_sa_mc_join_states {
+	FULLMEMBER_JOIN,
+	NONMEMBER_JOIN,
+	SENDONLY_NONMEBER_JOIN,
+	SENDONLY_FULLMEMBER_JOIN,
+	NUM_JOIN_MEMBERSHIP_TYPES,
+};
+
 #define IB_SA_CAP_MASK2_SENDONLY_FULL_MEM_SUPPORT	BIT(12)
 
 /*
diff --git a/include/rdma/rdma_cm.h b/include/rdma/rdma_cm.h
index afe44fd..81fb1d1 100644
--- a/include/rdma/rdma_cm.h
+++ b/include/rdma/rdma_cm.h
@@ -333,11 +333,13 @@ int rdma_disconnect(struct rdma_cm_id *id);
  *   address.
  * @id: Communication identifier associated with the request.
  * @addr: Multicast address identifying the group to join.
+ * @join_state: Multicast JoinState bitmap requested by port.
+ *		Bitmap is based on IB_SA_MCMEMBER_REC_JOIN_STATE bits.
  * @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);
+			u8 join_state, void *context);
 
 /**
  * rdma_leave_multicast - Leave the multicast group specified by the given
diff --git a/include/uapi/rdma/rdma_user_cm.h b/include/uapi/rdma/rdma_user_cm.h
index 3066718..01923d4 100644
--- a/include/uapi/rdma/rdma_user_cm.h
+++ b/include/uapi/rdma/rdma_user_cm.h
@@ -244,12 +244,19 @@ struct rdma_ucm_join_ip_mcast {
 	__u32 id;
 };
 
+/* Multicast join flags */
+enum {
+	RDMA_MC_JOIN_FLAG_FULLMEMBER,
+	RDMA_MC_JOIN_FLAG_SENDONLY_FULLMEMBER,
+	RDMA_MC_JOIN_FLAG_RESERVED,
+};
+
 struct rdma_ucm_join_mcast {
 	__u64 response;		/* rdma_ucma_create_id_resp */
 	__u64 uid;
 	__u32 id;
 	__u16 addr_size;
-	__u16 reserved;
+	__u16 join_flags;
 	struct sockaddr_storage addr;
 };
 
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH V1 for-next 0/2] Support for Send Only Full Member multicast join

[Matan Barak]

On Wed, Jul 6, 2016 at 4:36 PM, Leon Romanovsky <leon-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org> wrote:
> Added UCMA and CMA support for multicast join flags. Flags are
> passed using UCMA CM join command previously reserved fields.
> Currently supporting two join flags indicating two different
> multicast JoinStates:
>
> 1. Full Member:
>    The initiator creates the Multicast group(MCG) if it wasn't
>    previously created, can send Multicast messages to the group
>    and receive messages from the MCG.
>
> 2. Send Only Full Member:
>    The initiator creates the Multicast group(MCG) if it wasn't
>    previously created, can send Multicast messages to the group
>    but doesn't receive any messages from the MCG.
>
> Since Send Only Full Member requires SM support a new cached
> SA port attribute containing SM ClassPortInfo fields was added
> (ClassPortInfo fields: Table 126 IB Spec 1.3.). The attribute
> is cached to avoid resending queries, caching is done when a
> successful ClassPortInfo reply is received on the port.
> Invalidation of the attribute is done on SM change events,
> SM re-registration events, and SM LID change events. The fields
> in ClassPortInfo should not change during SM runtime without an
> event.
>
> The changes compared to v0 of this patch series are:
>  * Use multicast.c join state enum instead of the new added bits
>    defined by IB_SA_MCMEMBER_REC_JOIN_STATE in ib_sa.h
>  * Remove ib_sa_port struct alignment spaces
>
> Alex Vesker (2):
>   IB/sa: Add cached attribute containing SM information to SA port
>   IB/core: Support for CMA multicast join flags
>

Hi Doug,

This series conflicts with c65f6c5a3650 ('IB/core: Fix RoCE v1
multicast join logic issue') which was merged to your k.o/for-4.7-rc
branch.
The solution for this conflict requires to take the hunk for cma.c
from this series.

If it's required to respin this series above c65f6c5a3650, please just say so.

Regards,
Alex and Matan

>  drivers/infiniband/core/cma.c       | 98 ++++++++++++++++++++++++++++++++++---
>  drivers/infiniband/core/multicast.c | 12 -----
>  drivers/infiniband/core/sa_query.c  | 41 ++++++++++++++++
>  drivers/infiniband/core/ucma.c      | 18 +++++--
>  include/rdma/ib_sa.h                | 13 +++++
>  include/rdma/rdma_cm.h              |  4 +-
>  include/uapi/rdma/rdma_user_cm.h    |  9 +++-
>  7 files changed, 171 insertions(+), 24 deletions(-)
>
> --
> 2.1.4
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH V1 for-next 0/2] Support for Send Only Full Member multicast join

[Doug Ledford]

[-- Attachment #1: Type: text/plain, Size: 2474 bytes --]

On Sun, 2016-07-24 at 16:11 +0300, Matan Barak wrote:
> On Wed, Jul 6, 2016 at 4:36 PM, Leon Romanovsky <leon-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
> wrote:
> > 
> > Added UCMA and CMA support for multicast join flags. Flags are
> > passed using UCMA CM join command previously reserved fields.
> > Currently supporting two join flags indicating two different
> > multicast JoinStates:
> > 
> > 1. Full Member:
> >    The initiator creates the Multicast group(MCG) if it wasn't
> >    previously created, can send Multicast messages to the group
> >    and receive messages from the MCG.
> > 
> > 2. Send Only Full Member:
> >    The initiator creates the Multicast group(MCG) if it wasn't
> >    previously created, can send Multicast messages to the group
> >    but doesn't receive any messages from the MCG.
> > 
> > Since Send Only Full Member requires SM support a new cached
> > SA port attribute containing SM ClassPortInfo fields was added
> > (ClassPortInfo fields: Table 126 IB Spec 1.3.). The attribute
> > is cached to avoid resending queries, caching is done when a
> > successful ClassPortInfo reply is received on the port.
> > Invalidation of the attribute is done on SM change events,
> > SM re-registration events, and SM LID change events. The fields
> > in ClassPortInfo should not change during SM runtime without an
> > event.
> > 
> > The changes compared to v0 of this patch series are:
> >  * Use multicast.c join state enum instead of the new added bits
> >    defined by IB_SA_MCMEMBER_REC_JOIN_STATE in ib_sa.h
> >  * Remove ib_sa_port struct alignment spaces
> > 
> > Alex Vesker (2):
> >   IB/sa: Add cached attribute containing SM information to SA port
> >   IB/core: Support for CMA multicast join flags
> > 
> 
> Hi Doug,
> 
> This series conflicts with c65f6c5a3650 ('IB/core: Fix RoCE v1
> multicast join logic issue') which was merged to your k.o/for-4.7-rc
> branch.
> The solution for this conflict requires to take the hunk for cma.c
> from this series.
> 
> If it's required to respin this series above c65f6c5a3650, please
> just say so.
> 

I'm not entirely sure what you are referring to.  The series applied
cleanly for me.  However, it's in my misc branch (to be pushed shortly)
, please double check that there isn't something wrong.

-- 
Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
              GPG KeyID: 0E572FDD

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [PATCH V1 for-next 0/2] Support for Send Only Full Member multicast join

[Leon Romanovsky]

[-- Attachment #1: Type: text/plain, Size: 2859 bytes --]

On Tue, Aug 02, 2016 at 10:37:52PM -0400, Doug Ledford wrote:
> On Sun, 2016-07-24 at 16:11 +0300, Matan Barak wrote:
> > On Wed, Jul 6, 2016 at 4:36 PM, Leon Romanovsky <leon-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
> > wrote:
> > > 
> > > Added UCMA and CMA support for multicast join flags. Flags are
> > > passed using UCMA CM join command previously reserved fields.
> > > Currently supporting two join flags indicating two different
> > > multicast JoinStates:
> > > 
> > > 1. Full Member:
> > >    The initiator creates the Multicast group(MCG) if it wasn't
> > >    previously created, can send Multicast messages to the group
> > >    and receive messages from the MCG.
> > > 
> > > 2. Send Only Full Member:
> > >    The initiator creates the Multicast group(MCG) if it wasn't
> > >    previously created, can send Multicast messages to the group
> > >    but doesn't receive any messages from the MCG.
> > > 
> > > Since Send Only Full Member requires SM support a new cached
> > > SA port attribute containing SM ClassPortInfo fields was added
> > > (ClassPortInfo fields: Table 126 IB Spec 1.3.). The attribute
> > > is cached to avoid resending queries, caching is done when a
> > > successful ClassPortInfo reply is received on the port.
> > > Invalidation of the attribute is done on SM change events,
> > > SM re-registration events, and SM LID change events. The fields
> > > in ClassPortInfo should not change during SM runtime without an
> > > event.
> > > 
> > > The changes compared to v0 of this patch series are:
> > >  * Use multicast.c join state enum instead of the new added bits
> > >    defined by IB_SA_MCMEMBER_REC_JOIN_STATE in ib_sa.h
> > >  * Remove ib_sa_port struct alignment spaces
> > > 
> > > Alex Vesker (2):
> > >   IB/sa: Add cached attribute containing SM information to SA port
> > >   IB/core: Support for CMA multicast join flags
> > > 
> > 
> > Hi Doug,
> > 
> > This series conflicts with c65f6c5a3650 ('IB/core: Fix RoCE v1
> > multicast join logic issue') which was merged to your k.o/for-4.7-rc
> > branch.
> > The solution for this conflict requires to take the hunk for cma.c
> > from this series.
> > 
> > If it's required to respin this series above c65f6c5a3650, please
> > just say so.
> > 
> 
> I'm not entirely sure what you are referring to.  The series applied
> cleanly for me.  However, it's in my misc branch (to be pushed shortly)
> , please double check that there isn't something wrong.

I checked it and everything looks correct.

Such false alarm came from our internal merge test of net-next which
was based on v4.7-rc5 stage and your rdma-next (+ our patches) tree
which was based on v4.7-rc4.

Thanks.

> 
> -- 
> Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
>               GPG KeyID: 0E572FDD



[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]