vxlan: fix a regression of igmp join

This is a regression introduced by:

	commit 3fc2de2faba387218bdf9dbc6b13f513ac3b060a
	Author: stephen hemminger <stephen@networkplumber.org>
	Date:   Thu Jul 18 08:40:15 2013 -0700

	    vxlan: fix igmp races

Before this commit, the old code was:

       if (vxlan_group_used(vn, vxlan->default_dst.remote_ip))
               ip_mc_join_group(sk, &mreq);
       else
               ip_mc_leave_group(sk, &mreq);

therefore we shoud check vxlan_group_used(), not its opposite,
for igmp_join.

Cc: Stephen Hemminger <stephen@networkplumber.org>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: Cong Wang <amwang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c
index f4c6db419ddb3b56a9ecede0b62988c33e1d7c29..11a6c03d202a9d8a98d35f21f8f03c8f63cac82d 100644 (file)
--- a/drivers/net/vxlan.c
+++ b/drivers/net/vxlan.c
@@ -1386,7 +1386,7 @@ static int vxlan_open(struct net_device *dev)
                return -ENOTCONN;
 
        if (IN_MULTICAST(ntohl(vxlan->default_dst.remote_ip)) &&
-           ! vxlan_group_used(vn, vxlan->default_dst.remote_ip)) {
+           vxlan_group_used(vn, vxlan->default_dst.remote_ip)) {
                vxlan_sock_hold(vs);
                dev_hold(dev);
                queue_work(vxlan_wq, &vxlan->igmp_join);