drbd: Consider that the no-data-condition could be in connected state
authorPhilipp Reisner <philipp.reisner@linbit.com>
Fri, 15 Jul 2011 16:44:26 +0000 (18:44 +0200)
committerPhilipp Reisner <philipp.reisner@linbit.com>
Wed, 9 May 2012 13:15:42 +0000 (15:15 +0200)
...when the peer has inconsistent data. In that case we failed to
clear the susp_nod flag. When the local disk was attached again

Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
drivers/block/drbd/drbd_main.c

index b62e3559b50c55cfaa6b6a79cc3731a0628a98f2..0ee05315f9ac66b6e31cfdedb2339bd858b73754 100644 (file)
@@ -1412,7 +1412,8 @@ static void after_state_ch(struct drbd_conf *mdev, union drbd_state os,
                if (os.conn < C_CONNECTED && ns.conn >= C_CONNECTED)
                        what = resend;
 
-               if (os.disk == D_ATTACHING && ns.disk > D_ATTACHING)
+               if ((os.disk == D_ATTACHING || os.disk == D_NEGOTIATING) &&
+                   ns.disk > D_NEGOTIATING)
                        what = restart_frozen_disk_io;
 
                if (what != nothing)