SUNRPC/xprtrdma: Fix reconnection locking

[ Upstream commit f99fa50880f5300fbbb3c0754ddc7f8738d24fe7 ]

The xprtrdma client code currently relies on the task that initiated the
connect to hold the XPRT_LOCK for the duration of the connection
attempt. If the task is woken early, due to some other event, then that
lock could get released early.
Avoid races by using the same mechanism that the socket code uses of
transferring lock ownership to the RDMA connect worker itself. That
frees us to call rpcrdma_xprt_disconnect() directly since we're now
guaranteed exclusion w.r.t. other callers.

Fixes: 4cf44be6f1e8 ("xprtrdma: Fix recursion into rpcrdma_xprt_disconnect()")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index ccfa85e995fd23f88c14e7f0b13f8ef9037dd8aa..8201531ce5d9771a6af1c032938f8718c5aa0cd8 100644 (file)
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -844,6 +844,7 @@ out:
        spin_unlock(&xprt->transport_lock);
        return ret;
 }
+EXPORT_SYMBOL_GPL(xprt_lock_connect);
 
 void xprt_unlock_connect(struct rpc_xprt *xprt, void *cookie)
 {
@@ -860,6 +861,7 @@ out:
        spin_unlock(&xprt->transport_lock);
        wake_up_bit(&xprt->state, XPRT_LOCKED);
 }
+EXPORT_SYMBOL_GPL(xprt_unlock_connect);
 
 /**
  * xprt_connect - schedule a transport connect operation

diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c
index c26db0a37996754ad7e371456e97f622ab167bd4..8e2368a0c2a2970bfd079afc8bd46ab381146fe4 100644 (file)
--- a/net/sunrpc/xprtrdma/transport.c
+++ b/net/sunrpc/xprtrdma/transport.c
@@ -249,12 +249,9 @@ xprt_rdma_connect_worker(struct work_struct *work)
                                           xprt->stat.connect_start;
                xprt_set_connected(xprt);
                rc = -EAGAIN;
-       } else {
-               /* Force a call to xprt_rdma_close to clean up */
-               spin_lock(&xprt->transport_lock);
-               set_bit(XPRT_CLOSE_WAIT, &xprt->state);
-               spin_unlock(&xprt->transport_lock);
-       }
+       } else
+               rpcrdma_xprt_disconnect(r_xprt);
+       xprt_unlock_connect(xprt, r_xprt);
        xprt_wake_pending_tasks(xprt, rc);
 }
 
@@ -487,6 +484,8 @@ xprt_rdma_connect(struct rpc_xprt *xprt, struct rpc_task *task)
        struct rpcrdma_ep *ep = r_xprt->rx_ep;
        unsigned long delay;
 
+       WARN_ON_ONCE(!xprt_lock_connect(xprt, task, r_xprt));
+
        delay = 0;
        if (ep && ep->re_connect_status != 0) {
                delay = xprt_reconnect_delay(xprt);