l2tp: don't BUG_ON sk_socket being NULL
Tom Parkin [Tue, 19 Mar 2013 06:11:18 +0000 (06:11 +0000)]
It is valid for an existing struct sock object to have a NULL sk_socket
pointer, so don't BUG_ON in l2tp_tunnel_del_work if that should occur.

Signed-off-by: Tom Parkin <tparkin@katalix.com>
Signed-off-by: James Chapman <jchapman@katalix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/l2tp/l2tp_core.c

index 45373fe..e841ef2 100644 (file)
@@ -1412,19 +1412,21 @@ static void l2tp_tunnel_del_work(struct work_struct *work)
                return;
 
        sock = sk->sk_socket;
-       BUG_ON(!sock);
 
-       /* If the tunnel socket was created directly by the kernel, use the
-        * sk_* API to release the socket now.  Otherwise go through the
-        * inet_* layer to shut the socket down, and let userspace close it.
+       /* If the tunnel socket was created by userspace, then go through the
+        * inet layer to shut the socket down, and let userspace close it.
+        * Otherwise, if we created the socket directly within the kernel, use
+        * the sk API to release it here.
         * In either case the tunnel resources are freed in the socket
         * destructor when the tunnel socket goes away.
         */
-       if (sock->file == NULL) {
-               kernel_sock_shutdown(sock, SHUT_RDWR);
-               sk_release_kernel(sk);
+       if (tunnel->fd >= 0) {
+               if (sock)
+                       inet_shutdown(sock, 2);
        } else {
-               inet_shutdown(sock, 2);
+               if (sock)
+                       kernel_sock_shutdown(sock, SHUT_RDWR);
+               sk_release_kernel(sk);
        }
 
        l2tp_tunnel_sock_put(sk);