IPoIB: Rejoin all multicast groups after a port event
authorEli Cohen <eli@mellanox.co.il>
Fri, 22 Sep 2006 22:22:56 +0000 (15:22 -0700)
committerRoland Dreier <rolandd@cisco.com>
Fri, 22 Sep 2006 22:22:56 +0000 (15:22 -0700)
When ipoib_ib_dev_flush() is called because of a port event, the
driver needs to rejoin all multicast groups, since the flush will call
ipoib_mcast_dev_flush() (via ipoib_ib_dev_down()).  Otherwise no
(non-broadcast) multicast groups will be rejoined until the networking
core calls ->set_multicast_list again, and so multicast reception will
be broken for potentially a long time.

Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
drivers/infiniband/ulp/ipoib/ipoib_ib.c

index 722177ea069b78a334579e6e4037d2e76a1bd4a0..240befdf90dc96e5d84519aa8902fc848e04d7ab 100644 (file)
@@ -631,8 +631,10 @@ void ipoib_ib_dev_flush(void *_dev)
         * The device could have been brought down between the start and when
         * we get here, don't bring it back up if it's not configured up
         */
-       if (test_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags))
+       if (test_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags)) {
                ipoib_ib_dev_up(dev);
+               ipoib_mcast_restart_task(dev);
+       }
 
        mutex_lock(&priv->vlan_mutex);