On Wed, Apr 06, 2016 at 07:22:51PM +1000, Dave Chinner wrote:
> From: Dave Chinner <dchinner@xxxxxxxxxx>
>
> We don't write back stale inodes, so we should skip them in
> xfS_iflush_cluster, too.
xfs_iflush_cluster
>
> Signed-off-by: Dave Chinner <dchinner@xxxxxxxxxx>
> ---
Reviewed-by: Brian Foster <bfoster@xxxxxxxxxx>
> fs/xfs/xfs_inode.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
> index 2718d10..6598104 100644
> --- a/fs/xfs/xfs_inode.c
> +++ b/fs/xfs/xfs_inode.c
> @@ -3190,10 +3190,11 @@ xfs_iflush_cluster(
> * because this is an RCU protected lookup, we could find a
> * recently freed or even reallocated inode during the lookup.
> * We need to check under the i_flags_lock for a valid inode
> - * here. Skip it if it is not valid or the wrong inode.
> + * here. Skip it if it is not valid, stale or the wrong inode.
> */
> spin_lock(&cip->i_flags_lock);
> if (!cip->i_ino ||
> + __xfs_iflags_test(ip, XFS_ISTALE) ||
> (XFS_INO_TO_AGINO(mp, cip->i_ino) & mask) != first_index) {
> spin_unlock(&cip->i_flags_lock);
> continue;
> --
> 2.7.0
>
> _______________________________________________
> xfs mailing list
> xfs@xxxxxxxxxxx
> http://oss.sgi.com/mailman/listinfo/xfs
|