Re: [PATCH 4/8] xfs: allow extent free intents to be retried

[Dave Chinner <david@fromorbit.com>]

On Wed, Jun 28, 2023 at 10:48:36AM -0700, Darrick J. Wong wrote:
> On Wed, Jun 28, 2023 at 08:44:08AM +1000, Dave Chinner wrote:
> > From: Dave Chinner <dchinner@redhat.com>
> > @@ -653,9 +700,29 @@ xfs_efi_item_recover(
> >  		fake.xefi_startblock = extp->ext_start;
> >  		fake.xefi_blockcount = extp->ext_len;
> >  
> > -		xfs_extent_free_get_group(mp, &fake);
> > -		error = xfs_trans_free_extent(tp, efdp, &fake);
> > -		xfs_extent_free_put_group(&fake);
> > +		if (!requeue_only) {
> > +			xfs_extent_free_get_group(mp, &fake);
> > +			error = xfs_trans_free_extent(tp, efdp, &fake);
> > +			xfs_extent_free_put_group(&fake);
> > +		}
> > +
> > +		/*
> > +		 * If we can't free the extent without potentially deadlocking,
> > +		 * requeue the rest of the extents to a new so that they get
> > +		 * run again later with a new transaction context.
> > +		 */
> > +		if (error == -EAGAIN || requeue_only) {
> > +			error = xfs_free_extent_later(tp, fake.xefi_startblock,
> > +					fake.xefi_blockcount,
> > +					&XFS_RMAP_OINFO_ANY_OWNER,
> > +					fake.xefi_type);
> > +			if (!error) {
> > +				requeue_only = true;
> > +				error = 0;
> 
> @error is already zero so this is redundant, right?

Yes. Forgot to remove it when I changed that code to continue and
have errors fall through to the common handling...

> If yes then I'll remove the line on commit,
> Reviewed-by: Darrick J. Wong <djwong@kernel.org>

Thanks.

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com