Btrfs: fix race in sync and freeze again

I screwed this up, there is a race between checking if there is a running
transaction and actually starting a transaction in sync where we could race
with a freezer and get ourselves into trouble.  To fix this we need to make
a new join type to only do the try lock on the freeze stuff.  If it fails
we'll return EPERM and just return from sync.  This fixes a hang Liu Bo
reported when running xfstest 68 in a loop.  Thanks,

Reported-by: Liu Bo <bo.li.liu@oracle.com>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>

1 files changed, 11 insertions, 1 deletions

diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index c01dec70c960..e4bfac8d54b8 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -272,6 +272,7 @@ enum btrfs_trans_type {
         TRANS_JOIN,
         TRANS_USERSPACE,
         TRANS_JOIN_NOLOCK,
+        TRANS_JOIN_FREEZE,
 };
 
 static int may_wait_transaction(struct btrfs_root *root, int type)
@@ -341,7 +342,11 @@ again:
         if (!h)
                 return ERR_PTR(-ENOMEM);
 
-        sb_start_intwrite(root->fs_info->sb);
+        if (!__sb_start_write(root->fs_info->sb, SB_FREEZE_FS, false)) {
+                if (type == TRANS_JOIN_FREEZE)
+                        return ERR_PTR(-EPERM);
+                sb_start_intwrite(root->fs_info->sb);
+        }
 
         if (may_wait_transaction(root, type))
                 wait_current_trans(root);
@@ -424,6 +429,11 @@ struct btrfs_trans_handle *btrfs_start_ioctl_transaction(struct btrfs_root *root
         return start_transaction(root, 0, TRANS_USERSPACE, 0);
 }
 
+struct btrfs_trans_handle *btrfs_join_transaction_freeze(struct btrfs_root *root)
+{
+        return start_transaction(root, 0, TRANS_JOIN_FREEZE, 0);
+}
+
 /* wait for a transaction commit to be fully complete */
 static noinline void wait_for_commit(struct btrfs_root *root,
                                     struct btrfs_transaction *commit)