[PATCH] btrfs: avoid double put of block group when emptying cluster

[Josef Bacik]

In __btrfs_return_cluster_to_free_space we will bail doing the cleanup
of the cluster if the block group we passed in doesn't match the block
group on the cluster.  However we drop a reference to block_group, as
the cluster holds a reference to the block group while it's attached to
the cluster.  If cluster->block_group != block_group however then this
is an extra put, which means we'll go negative and free this block group
down the line, leading to a UAF.

Fix this by simply bailing if the block group we passed in does not
match the block group on the cluster.

CC: sta...@vger.kernel.org
Fixes: fa9c0d795f7b ("Btrfs: rework allocation clustering")
Signed-off-by: Josef Bacik <jo...@toxicpanda.com>
---
 fs/btrfs/free-space-cache.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c
index 0d6dcb5ff963..8be36cc6cbd8 100644
--- a/fs/btrfs/free-space-cache.c
+++ b/fs/btrfs/free-space-cache.c
@@ -2711,8 +2711,10 @@ static void __btrfs_return_cluster_to_free_space(
        struct rb_node *node;
 
        spin_lock(&cluster->lock);
-       if (cluster->block_group != block_group)
-               goto out;
+       if (cluster->block_group != block_group) {
+               spin_unlock(&cluster->lock);
+               return;
+       }
 
        cluster->block_group = NULL;
        cluster->window_start = 0;
@@ -2750,8 +2752,6 @@ static void __btrfs_return_cluster_to_free_space(
                                   entry->offset, &entry->offset_index, bitmap);
        }
        cluster->root = RB_ROOT;
-
-out:
        spin_unlock(&cluster->lock);
        btrfs_put_block_group(block_group);
 }
-- 
2.26.2