btrfs: fix race in reada

When inserting into the radix tree returns EEXIST, get the existing
entry without giving up the spinlock in between.
There was a race for both the zones trees and the extent tree.

Signed-off-by: Arne Jansen <sensille@gmx.net>

1 files changed, 7 insertions, 1 deletions

diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index a682c267576d..98ee5a51aa29 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -4332,7 +4332,13 @@ static int btrfs_real_readdir(struct file *filp, void *dirent,
                         }
 no_dentry:
                         /* is this a reference to our own snapshot? If so
-                         * skip it
+                         * skip it.
+                         *
+                         * In contrast to old kernels, we insert the snapshot's
+                         * dir item and dir index after it has been created, so
+                         * we won't find a reference to our own snapshot. We
+                         * still keep the following code for backward
+                         * compatibility.
                          */
                         if (location.type == BTRFS_ROOT_ITEM_KEY &&
                             location.objectid == root->root_key.objectid) {