tcp: reset sk_rx_dst in tcp_disconnect()

We have to reset the sk->sk_rx_dst when we disconnect a TCP connection, because otherwise when we re-connect it this dst reference is simply overridden in tcp_finish_connect(). This fixes a dst leak which leads to a loopback dev refcnt leak. It is a long-standing bug, Kevin reported a very similar (if not same) bug before. Thanks to Andrei for providing such a reliable reproducer which greatly narrows down the problem. Fixes: 41063e9 ("ipv4: Early TCP socket demux.") Reported-by: Andrei Vagin <[email protected]> Reported-by: Kevin Xu <[email protected]> Signed-off-by: Cong Wang <[email protected]> Signed-off-by: David S. Miller <[email protected]>
torvalds · Jun 25, 2017 · d747a7a · d747a7a
1 parent 85cb73f
commit d747a7a
Showing 1 changed file with 2 additions and 0 deletions.
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
@@ -2330,6 +2330,8 @@ int tcp_disconnect(struct sock *sk, int flags)
 	tcp_init_send_head(sk);
 	memset(&tp->rx_opt, 0, sizeof(tp->rx_opt));
 	__sk_dst_reset(sk);
+	dst_release(sk->sk_rx_dst);
+	sk->sk_rx_dst = NULL;
 	tcp_saved_syn_free(tp);
 
 	/* Clean up fastopen related fields */