tipc: avoid to unnecessary process switch under non-block mode
authorYing Xue <ying.xue@windriver.com>
Thu, 6 Mar 2014 13:40:18 +0000 (14:40 +0100)
committerDavid S. Miller <davem@davemloft.net>
Thu, 6 Mar 2014 19:46:23 +0000 (14:46 -0500)
When messages are received via tipc socket under non-block mode,
schedule_timeout() is called in tipc_wait_for_rcvmsg(), that is,
the process of receiving messages will be scheduled once although
timeout value passed to schedule_timeout() is 0. The same issue
exists in accept()/wait_for_accept(). To avoid this unnecessary
process switch, we only call schedule_timeout() if the timeout
value is non-zero.

Signed-off-by: Ying Xue <ying.xue@windriver.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/tipc/socket.c

index a4cf274455aa42922c3e27947e0788da8e3ac794..0ed0eaa62f29e7148052907892ff90ebd20d2fed 100644 (file)
@@ -997,7 +997,7 @@ static int tipc_wait_for_rcvmsg(struct socket *sock, long timeo)
 
        for (;;) {
                prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE);
-               if (skb_queue_empty(&sk->sk_receive_queue)) {
+               if (timeo && skb_queue_empty(&sk->sk_receive_queue)) {
                        if (sock->state == SS_DISCONNECTING) {
                                err = -ENOTCONN;
                                break;
@@ -1623,7 +1623,7 @@ static int tipc_wait_for_accept(struct socket *sock, long timeo)
        for (;;) {
                prepare_to_wait_exclusive(sk_sleep(sk), &wait,
                                          TASK_INTERRUPTIBLE);
-               if (skb_queue_empty(&sk->sk_receive_queue)) {
+               if (timeo && skb_queue_empty(&sk->sk_receive_queue)) {
                        release_sock(sk);
                        timeo = schedule_timeout(timeo);
                        lock_sock(sk);