Re: [PATCH 2/4] xfs: recovery of XLOG_UNMOUNT_TRANS leaks memory

[Brian Foster <bfoster@redhat.com>]

On Tue, Aug 26, 2014 at 11:21:39AM +1000, Dave Chinner wrote:
> From: Dave Chinner <dchinner@redhat.com>
> 
> It aborts recovery without freeing the current trans structure that
> we are decoding.
> 

What do you mean by "aborts recovery?" I don't see anything in the code
that reflects that behavior. Do you mean it's an on-disk marker for
completion?

> Signed-off-by: Dave Chinner <dchinner@redhat.com>
> ---
>  fs/xfs/xfs_log_recover.c | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
> index 1970732f..460cf98 100644
> --- a/fs/xfs/xfs_log_recover.c
> +++ b/fs/xfs/xfs_log_recover.c
> @@ -3587,8 +3587,9 @@ xlog_recovery_process_ophdr(
>  	/* unexpected flag values */
>  	case XLOG_UNMOUNT_TRANS:
>  		xfs_warn(log->l_mp, "%s: Unmount LR", __func__);
> -		error = 0;
> -		break;
> +		xlog_recover_free_trans(trans);
> +		return 0;
> +

The change to return here seems superfluous. It's fine, but just to
check, were you intending to alter behavior in some way (e.g., return
from xlog_recover_process_data())?

Brian

>  	case XLOG_START_TRANS:
>  		xfs_warn(log->l_mp, "%s: bad transaction 0x%x", __func__, tid);
>  		ASSERT(0);
> -- 
> 2.0.0
> 
> _______________________________________________
> xfs mailing list
> xfs@oss.sgi.com
> http://oss.sgi.com/mailman/listinfo/xfs

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs