Btrfs: fix race in run_clustered_refs
authorArne Jansen <sensille@gmx.net>
Thu, 9 Aug 2012 06:16:53 +0000 (00:16 -0600)
committerChris Mason <chris.mason@oracle.com>
Tue, 28 Aug 2012 20:53:35 +0000 (16:53 -0400)
commit22cd2e7de7b0bd68fb668d23e1564707ca689510
tree115a0c4217ed00d1d879925928dd4e7395a599b2
parentb12a3b1ea209d9dec02731fba58c3dbe7d31cfd8
Btrfs: fix race in run_clustered_refs

With commit

commit d1270cd91f308c9d22b2804720c36ccd32dbc35e
Author: Arne Jansen <sensille@gmx.net>
Date:   Tue Sep 13 15:16:43 2011 +0200

     Btrfs: put back delayed refs that are too new

I added a window where the delayed_ref's head->ref_mod code can diverge
from the sum of the remaining refs, because we release the head->mutex
in the middle. This leads to btrfs_lookup_extent_info returning wrong
numbers. This patch fixes this by adjusting the head's ref_mod with each
delayed ref we run.

Signed-off-by: Arne Jansen <sensille@gmx.net>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
fs/btrfs/extent-tree.c