[PATCH 3.16 252/305] SUNRPC: Fix a potential race in xprt_connect()

[Ben Hutchings]

3.16.63-rc1 review patch.  If anyone has any objections, please let me know.

------------------

From: Trond Myklebust <trond.mykleb...@hammerspace.com>

commit 0a9a4304f3614e25d9de9b63502ca633c01c0d70 upstream.

If an asynchronous connection attempt completes while another task is
in xprt_connect(), then the call to rpc_sleep_on() could end up
racing with the call to xprt_wake_pending_tasks().
So add a second test of the connection state after we've put the
task to sleep and set the XPRT_CONNECTING flag, when we know that there
can be no asynchronous connection attempts still in progress.

Fixes: 0b9e79431377d ("SUNRPC: Move the test for XPRT_CONNECTING into...")
Signed-off-by: Trond Myklebust <trond.mykleb...@hammerspace.com>
Signed-off-by: Ben Hutchings <b...@decadent.org.uk>
---
 net/sunrpc/xprt.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -721,8 +721,15 @@ void xprt_connect(struct rpc_task *task)
                        return;
                if (xprt_test_and_set_connecting(xprt))
                        return;
-               xprt->stat.connect_start = jiffies;
-               xprt->ops->connect(xprt, task);
+               /* Race breaker */
+               if (!xprt_connected(xprt)) {
+                       xprt->stat.connect_start = jiffies;
+                       xprt->ops->connect(xprt, task);
+               } else {
+                       xprt_clear_connecting(xprt);
+                       task->tk_status = 0;
+                       rpc_wake_up_queued_task(&xprt->pending, task);
+               }
        }
 }