svn commit: r326703 - stable/11/sys/ofed/drivers/infiniband/core

[Hans Petter Selasky]

Author: hselasky
Date: Fri Dec  8 19:19:47 2017
New Revision: 326703
URL: https://svnweb.freebsd.org/changeset/base/326703

Log:
  Add support for IPv6 based addresses as part of the TCP unify portspace feature
  in ibcore. This resolves an interopability issue when using both iWarp(T6) and
  RDMA(CX-4 and CX-5) devices at the same time.
  
  The problem is IPv4 based sockets cannot be bound to an IPv6 based address
  causing sobind() to fail preventing all use of IPv6 based addresses with RDMA
  when an iWarp device is present.
  
  This is a direct commit.
  
  Tested by:	KrishnamRaju ErapaRaju <Krishna2 at chelsio.com>
  Sponsored by:	Mellanox Technologies

Modified:
  stable/11/sys/ofed/drivers/infiniband/core/cma.c

Modified: stable/11/sys/ofed/drivers/infiniband/core/cma.c
==============================================================================
--- stable/11/sys/ofed/drivers/infiniband/core/cma.c	Fri Dec  8 18:43:31 2017	(r326702)
+++ stable/11/sys/ofed/drivers/infiniband/core/cma.c	Fri Dec  8 19:19:47 2017	(r326703)
@@ -2569,32 +2569,37 @@ static int cma_get_tcp_port(struct rdma_id_private *id
 	int ret;
 	int size;
 	struct socket *sock;
+	struct sockaddr *src_addr = (struct sockaddr *)&id_priv->id.route.addr.src_addr;
 
-	ret = sock_create_kern(AF_INET, SOCK_STREAM, IPPROTO_TCP, &sock);
+	switch (src_addr->sa_family) {
+	case AF_INET:
+	case AF_INET6:
+		break;
+	default:
+		/* other address families are not handled by iWarp */
+		id_priv->unify_ps_tcp = 0;
+		return (0);
+	}
+
+	ret = sock_create_kern(src_addr->sa_family, SOCK_STREAM, IPPROTO_TCP, &sock);
 	if (ret)
 		return ret;
 #ifdef __linux__
-	ret = sock->ops->bind(sock,
-			(struct sockaddr *) &id_priv->id.route.addr.src_addr,
-			ip_addr_size((struct sockaddr *) &id_priv->id.route.addr.src_addr));
+	ret = sock->ops->bind(sock, src_addr, ip_addr_size(src_addr));
 #else
 	SOCK_LOCK(sock);
 	sock->so_options |= SO_REUSEADDR;
 	SOCK_UNLOCK(sock);
 
-	ret = -sobind(sock,
-			(struct sockaddr *)&id_priv->id.route.addr.src_addr,
-			curthread);
+	ret = -sobind(sock, src_addr, curthread);
 #endif
 	if (ret) {
 		sock_release(sock);
 		return ret;
 	}
 
-	size = ip_addr_size((struct sockaddr *) &id_priv->id.route.addr.src_addr);
-	ret = sock_getname(sock,
-			(struct sockaddr *) &id_priv->id.route.addr.src_addr,
-			&size, 0);
+	size = ip_addr_size(src_addr);
+	ret = sock_getname(sock, src_addr, &size, 0);
 	if (ret) {
 		sock_release(sock);
 		return ret;