netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] ipv4: don't call upper-layer disconnect function if not connected
@ 2006-08-01 19:48 Brian Haley
  2006-08-01 22:07 ` David Miller
  0 siblings, 1 reply; 3+ messages in thread
From: Brian Haley @ 2006-08-01 19:48 UTC (permalink / raw)
  To: David Miller; +Cc: netdev

[-- Attachment #1: Type: text/plain, Size: 183 bytes --]

Calling connect() with AF_UNSPEC will disconnect a socket, but we don't 
need to do any work if the socket isn't currently connected.

Signed-off-by: Brian Haley <brian.haley@hp.com>

[-- Attachment #2: disconn.patch --]
[-- Type: text/x-patch, Size: 1133 bytes --]

diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index c84a320..b294b92 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -480,12 +480,16 @@ int inet_dgram_connect(struct socket *so
 {
 	struct sock *sk = sock->sk;
 
-	if (uaddr->sa_family == AF_UNSPEC)
-		return sk->sk_prot->disconnect(sk, flags);
+	if (uaddr->sa_family == AF_UNSPEC) {
+		if (sock->state != SS_UNCONNECTED)
+			return sk->sk_prot->disconnect(sk, flags);
+		else
+			return 0;
+	}
 
 	if (!inet_sk(sk)->num && inet_autobind(sk))
 		return -EAGAIN;
-	return sk->sk_prot->connect(sk, (struct sockaddr *)uaddr, addr_len);
+	return sk->sk_prot->connect(sk, uaddr, addr_len);
 }
 
 static long inet_wait_for_connect(struct sock *sk, long timeo)
@@ -525,8 +529,11 @@ int inet_stream_connect(struct socket *s
 	lock_sock(sk);
 
 	if (uaddr->sa_family == AF_UNSPEC) {
-		err = sk->sk_prot->disconnect(sk, flags);
-		sock->state = err ? SS_DISCONNECTING : SS_UNCONNECTED;
+		if (sock->state != SS_UNCONNECTED) {
+			err = sk->sk_prot->disconnect(sk, flags);
+			sock->state = err ? SS_DISCONNECTING : SS_UNCONNECTED;
+		} else
+			err = 0;
 		goto out;
 	}
 

^ permalink raw reply related	[flat|nested] 3+ messages in thread

* Re: [PATCH] ipv4: don't call upper-layer disconnect function if not connected
  2006-08-01 19:48 [PATCH] ipv4: don't call upper-layer disconnect function if not connected Brian Haley
@ 2006-08-01 22:07 ` David Miller
  2006-08-02 16:04   ` Brian Haley
  0 siblings, 1 reply; 3+ messages in thread
From: David Miller @ 2006-08-01 22:07 UTC (permalink / raw)
  To: brian.haley; +Cc: netdev

From: Brian Haley <brian.haley@hp.com>
Date: Tue, 01 Aug 2006 15:48:54 -0400

> Calling connect() with AF_UNSPEC will disconnect a socket, but we don't 
> need to do any work if the socket isn't currently connected.
> 
> Signed-off-by: Brian Haley <brian.haley@hp.com>

The socket could have been bind()'d to, in which case it will
not move to connected state and we still need to invoke
the disconnect methods such as udp_disconnect() to clear out
that binding.

You seem to be groveling in random areas of the ipv4 and ipv6 stack,
what are you working on?

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] ipv4: don't call upper-layer disconnect function if not connected
  2006-08-01 22:07 ` David Miller
@ 2006-08-02 16:04   ` Brian Haley
  0 siblings, 0 replies; 3+ messages in thread
From: Brian Haley @ 2006-08-02 16:04 UTC (permalink / raw)
  To: David Miller; +Cc: netdev


> The socket could have been bind()'d to, in which case it will
> not move to connected state and we still need to invoke
> the disconnect methods such as udp_disconnect() to clear out
> that binding.

Ok.

> You seem to be groveling in random areas of the ipv4 and ipv6 stack,
> what are you working on?

Was looking into a customer-reported memory leak that seemed to be in 
this code path.  It wasn't, but this tweak seemed sane at the time.

-Brian

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2006-08-02 16:04 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2006-08-01 19:48 [PATCH] ipv4: don't call upper-layer disconnect function if not connected Brian Haley
2006-08-01 22:07 ` David Miller
2006-08-02 16:04   ` Brian Haley

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).