xprtrdma: kmalloc rpcrdma_ep separate from rpcrdma_xprt

Change the rpcrdma_xprt_disconnect() function so that it no longer
waits for the DISCONNECTED event.  This prevents blocking if the
remote is unresponsive.

In rpcrdma_xprt_disconnect(), the transport's rpcrdma_ep is
detached. Upon return from rpcrdma_xprt_disconnect(), the transport
(r_xprt) is ready immediately for a new connection.

The RDMA_CM_DEVICE_REMOVAL and RDMA_CM_DISCONNECTED events are now
handled almost identically.

However, because the lifetimes of rpcrdma_xprt structures and
rpcrdma_ep structures are now independent, creating an rpcrdma_ep
needs to take a module ref count. The ep now owns most of the
hardware resources for a transport.

Also, a kref is needed to ensure that rpcrdma_ep sticks around
long enough for the cm_event_handler to finish.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>

1 files changed, 9 insertions, 8 deletions

diff --git a/net/sunrpc/xprtrdma/rpc_rdma.c b/net/sunrpc/xprtrdma/rpc_rdma.c
index ad7e6b0187bd..d1af48e0139c 100644
--- a/net/sunrpc/xprtrdma/rpc_rdma.c
+++ b/net/sunrpc/xprtrdma/rpc_rdma.c
@@ -131,9 +131,10 @@ static bool rpcrdma_args_inline(struct rpcrdma_xprt *r_xprt,
 				struct rpc_rqst *rqst)
 {
 	struct xdr_buf *xdr = &rqst->rq_snd_buf;
+	struct rpcrdma_ep *ep = r_xprt->rx_ep;
 	unsigned int count, remaining, offset;
 
-	if (xdr->len > r_xprt->rx_ep.re_max_inline_send)
+	if (xdr->len > ep->re_max_inline_send)
 		return false;
 
 	if (xdr->page_len) {
@@ -144,7 +145,7 @@ static bool rpcrdma_args_inline(struct rpcrdma_xprt *r_xprt,
 			remaining -= min_t(unsigned int,
 					   PAGE_SIZE - offset, remaining);
 			offset = 0;
-			if (++count > r_xprt->rx_ep.re_attr.cap.max_send_sge)
+			if (++count > ep->re_attr.cap.max_send_sge)
 				return false;
 		}
 	}
@@ -161,7 +162,7 @@ static bool rpcrdma_args_inline(struct rpcrdma_xprt *r_xprt,
 static bool rpcrdma_results_inline(struct rpcrdma_xprt *r_xprt,
 				   struct rpc_rqst *rqst)
 {
-	return rqst->rq_rcv_buf.buflen <= r_xprt->rx_ep.re_max_inline_recv;
+	return rqst->rq_rcv_buf.buflen <= r_xprt->rx_ep->re_max_inline_recv;
 }
 
 /* The client is required to provide a Reply chunk if the maximum
@@ -175,7 +176,7 @@ rpcrdma_nonpayload_inline(const struct rpcrdma_xprt *r_xprt,
 	const struct xdr_buf *buf = &rqst->rq_rcv_buf;
 
 	return (buf->head[0].iov_len + buf->tail[0].iov_len) <
-		r_xprt->rx_ep.re_max_inline_recv;
+		r_xprt->rx_ep->re_max_inline_recv;
 }
 
 /* Split @vec on page boundaries into SGEs. FMR registers pages, not
@@ -254,7 +255,7 @@ rpcrdma_convert_iovs(struct rpcrdma_xprt *r_xprt, struct xdr_buf *xdrbuf,
 	/* When encoding a Read chunk, the tail iovec contains an
 	 * XDR pad and may be omitted.
 	 */
-	if (type == rpcrdma_readch && r_xprt->rx_ep.re_implicit_roundup)
+	if (type == rpcrdma_readch && r_xprt->rx_ep->re_implicit_roundup)
 		goto out;
 
 	/* When encoding a Write chunk, some servers need to see an
@@ -262,7 +263,7 @@ rpcrdma_convert_iovs(struct rpcrdma_xprt *r_xprt, struct xdr_buf *xdrbuf,
 	 * layer provides space in the tail iovec that may be used
 	 * for this purpose.
 	 */
-	if (type == rpcrdma_writech && r_xprt->rx_ep.re_implicit_roundup)
+	if (type == rpcrdma_writech && r_xprt->rx_ep->re_implicit_roundup)
 		goto out;
 
 	if (xdrbuf->tail[0].iov_len)
@@ -1475,8 +1476,8 @@ void rpcrdma_reply_handler(struct rpcrdma_rep *rep)
 
 	if (credits == 0)
 		credits = 1;	/* don't deadlock */
-	else if (credits > r_xprt->rx_ep.re_max_requests)
-		credits = r_xprt->rx_ep.re_max_requests;
+	else if (credits > r_xprt->rx_ep->re_max_requests)
+		credits = r_xprt->rx_ep->re_max_requests;
 	if (buf->rb_credits != credits)
 		rpcrdma_update_cwnd(r_xprt, credits);
 	rpcrdma_post_recvs(r_xprt, false);