SUNRPC: Ensure client closes the socket when server initiates a close

 If the server decides to close the RPC socket, we currently don't actually
 respond until either another RPC call is scheduled, or until xprt_autoclose()
 gets called by the socket expiry timer (which may be up to 5 minutes
 later).

 This patch ensures that xprt_autoclose() is called much sooner if the
 server closes the socket.

 Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>

3 files changed, 27 insertions, 19 deletions

diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h
index dd860128ced..6ef99b14ff0 100644
--- a/include/linux/sunrpc/xprt.h
+++ b/include/linux/sunrpc/xprt.h
@@ -254,6 +254,7 @@ int			xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to);
 #define XPRT_LOCKED		(0)
 #define XPRT_CONNECTED		(1)
 #define XPRT_CONNECTING		(2)
+#define XPRT_CLOSE_WAIT		(3)
 
 static inline void xprt_set_connected(struct rpc_xprt *xprt)
 {
diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index 069a6cbd49e..8bc0d5acf0d 100644
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -119,6 +119,17 @@ out_sleep:
 	return 0;
 }
 
+static void xprt_clear_locked(struct rpc_xprt *xprt)
+{
+	xprt->snd_task = NULL;
+	if (!test_bit(XPRT_CLOSE_WAIT, &xprt->state) || xprt->shutdown) {
+		smp_mb__before_clear_bit();
+		clear_bit(XPRT_LOCKED, &xprt->state);
+		smp_mb__after_clear_bit();
+	} else
+		schedule_work(&xprt->task_cleanup);
+}
+
 /*
  * xprt_reserve_xprt_cong - serialize write access to transports
  * @task: task that is requesting access to the transport
@@ -145,9 +156,7 @@ int xprt_reserve_xprt_cong(struct rpc_task *task)
 		}
 		return 1;
 	}
-	smp_mb__before_clear_bit();
-	clear_bit(XPRT_LOCKED, &xprt->state);
-	smp_mb__after_clear_bit();
+	xprt_clear_locked(xprt);
 out_sleep:
 	dprintk("RPC: %4d failed to lock transport %p\n", task->tk_pid, xprt);
 	task->tk_timeout = 0;
@@ -193,9 +202,7 @@ static void __xprt_lock_write_next(struct rpc_xprt *xprt)
 	return;
 
 out_unlock:
-	smp_mb__before_clear_bit();
-	clear_bit(XPRT_LOCKED, &xprt->state);
-	smp_mb__after_clear_bit();
+	xprt_clear_locked(xprt);
 }
 
 static void __xprt_lock_write_next_cong(struct rpc_xprt *xprt)
@@ -222,9 +229,7 @@ static void __xprt_lock_write_next_cong(struct rpc_xprt *xprt)
 		return;
 	}
 out_unlock:
-	smp_mb__before_clear_bit();
-	clear_bit(XPRT_LOCKED, &xprt->state);
-	smp_mb__after_clear_bit();
+	xprt_clear_locked(xprt);
 }
 
 /**
@@ -237,10 +242,7 @@ out_unlock:
 void xprt_release_xprt(struct rpc_xprt *xprt, struct rpc_task *task)
 {
 	if (xprt->snd_task == task) {
-		xprt->snd_task = NULL;
-		smp_mb__before_clear_bit();
-		clear_bit(XPRT_LOCKED, &xprt->state);
-		smp_mb__after_clear_bit();
+		xprt_clear_locked(xprt);
 		__xprt_lock_write_next(xprt);
 	}
 }
@@ -256,10 +258,7 @@ void xprt_release_xprt(struct rpc_xprt *xprt, struct rpc_task *task)
 void xprt_release_xprt_cong(struct rpc_xprt *xprt, struct rpc_task *task)
 {
 	if (xprt->snd_task == task) {
-		xprt->snd_task = NULL;
-		smp_mb__before_clear_bit();
-		clear_bit(XPRT_LOCKED, &xprt->state);
-		smp_mb__after_clear_bit();
+		xprt_clear_locked(xprt);
 		__xprt_lock_write_next_cong(xprt);
 	}
 }
diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c
index 3e889300147..c458f8d1d6d 100644
--- a/net/sunrpc/xprtsock.c
+++ b/net/sunrpc/xprtsock.c
@@ -425,7 +425,7 @@ static void xs_close(struct rpc_xprt *xprt)
 	struct sock *sk = xprt->inet;
 
 	if (!sk)
-		return;
+		goto clear_close_wait;
 
 	dprintk("RPC:      xs_close xprt %p\n", xprt);
 
@@ -442,6 +442,10 @@ static void xs_close(struct rpc_xprt *xprt)
 	sk->sk_no_check = 0;
 
 	sock_release(sock);
+clear_close_wait:
+	smp_mb__before_clear_bit();
+	clear_bit(XPRT_CLOSE_WAIT, &xprt->state);
+	smp_mb__after_clear_bit();
 }
 
 /**
@@ -801,9 +805,13 @@ static void xs_tcp_state_change(struct sock *sk)
 	case TCP_SYN_SENT:
 	case TCP_SYN_RECV:
 		break;
+	case TCP_CLOSE_WAIT:
+		/* Try to schedule an autoclose RPC calls */
+		set_bit(XPRT_CLOSE_WAIT, &xprt->state);
+		if (test_and_set_bit(XPRT_LOCKED, &xprt->state) == 0)
+			schedule_work(&xprt->task_cleanup);
 	default:
 		xprt_disconnect(xprt);
-		break;
 	}
  out:
 	read_unlock(&sk->sk_callback_lock);