[PATCH] Btrfs: don't bug_on when we fail when cleaning up transactions

[PATCH] Btrfs: don't bug_on when we fail when cleaning up transactions

[Josef Bacik]

There is no reason for this sort of jackassery.  Thanks,

Signed-off-by: Josef Bacik <jbacik@fusionio.com>
---
 fs/btrfs/transaction.c |    3 +--
 1 files changed, 1 insertions(+), 2 deletions(-)

diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index 18f7e71..ea8d522 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -1942,6 +1942,5 @@ int btrfs_clean_one_deleted_snapshot(struct btrfs_root *root)
 	 * If we encounter a transaction abort during snapshot cleaning, we
 	 * don't want to crash here
 	 */
-	BUG_ON(ret < 0 && ret != -EAGAIN && ret != -EROFS);
-	return 1;
+	return (ret < 0) ? 0 : 1;
 }
-- 
1.7.7.6

Re: [PATCH] Btrfs: don't bug_on when we fail when cleaning up transactions

[David Sterba]

On Wed, Jul 31, 2013 at 10:29:23AM -0400, Josef Bacik wrote:
> There is no reason for this sort of jackassery.  Thanks,
> 
> Signed-off-by: Josef Bacik <jbacik@fusionio.com>
> ---
>  fs/btrfs/transaction.c |    3 +--
>  1 files changed, 1 insertions(+), 2 deletions(-)
> 
> diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
> index 18f7e71..ea8d522 100644
> --- a/fs/btrfs/transaction.c
> +++ b/fs/btrfs/transaction.c
> @@ -1942,6 +1942,5 @@ int btrfs_clean_one_deleted_snapshot(struct btrfs_root *root)
>  	 * If we encounter a transaction abort during snapshot cleaning, we
>  	 * don't want to crash here
>  	 */
> -	BUG_ON(ret < 0 && ret != -EAGAIN && ret != -EROFS);
> -	return 1;
> +	return (ret < 0) ? 0 : 1;

The reason was to let known harmless error codes pass and catch the
unexpected ones, but the whitelist was incomplete and I guess you've seen
something like ENOMEM or EIO. As the cleaning process is restartable I'm
ok with removing the bugon.

Reviewed-by: David Sterba <dsterba@suse.cz>