On Wed, Oct 17, 2018 at 01:40:02AM -0700, Christoph Hellwig wrote:
> > @@ -1415,11 +1419,17 @@ xfs_reflink_remap_range(
> >
> > trace_xfs_reflink_remap_range(src, pos_in, len, dest, pos_out);
> >
> > + if (len == 0) {
> > + ret = 0;
> > + goto out_unlock;
> > + }
> > +
>
> As pointed out last time this check is superflous, right above we have
> this check:
>
> if (ret < 0 || len == 0)
> return ret;
Oops, sorry I missed that, will fix now.
> > ret = xfs_reflink_remap_blocks(src, sfsbno, dest, dfsbno, fsblen,
> > - pos_out + len);
> > + &remappedfsb, pos_out + len);
> > + remapped_bytes = min_t(loff_t, len, XFS_FSB_TO_B(mp, remappedfsb));
>
> I still think returning the bytes from the function would be saner,
> but maybe that's just me.
Hmmm, this call site is getting messy; I'll tack on another patch to
clean that up too.
--D
