md: Remove recent change which allows devices to skip recovery.
authorNeilBrown <neilb@suse.de>
Wed, 17 Jul 2013 04:55:31 +0000 (14:55 +1000)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Sun, 4 Aug 2013 08:50:53 +0000 (16:50 +0800)
commit 5024c298311f3b97c85cb034f9edaa333fdb9338 upstream.

commit 7ceb17e87bde79d285a8b988cfed9eaeebe60b86
    md: Allow devices to be re-added to a read-only array.

allowed a bit more than just that.  It also allows devices to be added
to a read-write array and to end up skipping recovery.

This patch removes the offending piece of code pending a rewrite for a
subsequent release.

More specifically:
 If the array has a bitmap, then the device will still need a bitmap
 based resync ('saved_raid_disk' is set under different conditions
 is a bitmap is present).
 If the array doesn't have a bitmap, then this is correct as long as
 nothing has been written to the array since the metadata was checked
 by ->validate_super.  However there is no locking to ensure that there
 was no write.

Bug was introduced in 3.10 and causes data corruption so
patch is suitable for 3.10-stable.

Reported-by: Joe Lawrence <joe.lawrence@stratus.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/md/md.c

index 9b82377..51f0345 100644 (file)
@@ -7697,20 +7697,6 @@ static int remove_and_add_spares(struct mddev *mddev,
                        continue;
 
                rdev->recovery_offset = 0;
-               if (rdev->saved_raid_disk >= 0 && mddev->in_sync) {
-                       spin_lock_irq(&mddev->write_lock);
-                       if (mddev->in_sync)
-                               /* OK, this device, which is in_sync,
-                                * will definitely be noticed before
-                                * the next write, so recovery isn't
-                                * needed.
-                                */
-                               rdev->recovery_offset = mddev->recovery_cp;
-                       spin_unlock_irq(&mddev->write_lock);
-               }
-               if (mddev->ro && rdev->recovery_offset != MaxSector)
-                       /* not safe to add this disk now */
-                       continue;
                if (mddev->pers->
                    hot_add_disk(mddev, rdev) == 0) {
                        if (sysfs_link_rdev(mddev, rdev))