Re: [PATCH v2 02/18] xfs: convert iomap ops to ->iomap_next()

[Darrick J. Wong]

On Thu, Jul 02, 2026 at 04:03:01PM +0200, Christoph Hellwig wrote:
> > -	 * The retry mechanism is based on the ->iomap_begin method returning
> > +	 * The retry mechanism is based on the ->iomap_next method returning
> 
> Nit: Maybe the drop the "->" in preparation of the last patch that would
> make it slightly misleading with the ops struct removal?
> 
> > +static int
> > +xfs_direct_write_iomap_next(
> > +	const struct iomap_iter *iter,
> > +	struct iomap		*iomap,
> > +	struct iomap		*srcmap)
> > +{
> > +	return iomap_process(iter, iomap, srcmap, xfs_direct_write_iomap_begin,
> > +			NULL);
> > +}
> 
> I wonder if we want a "simple" version of iomap_process that only takes
> a begin callback, as most current ops seems to be that way.  Then again
> between needing at least one longer function name it's probably not
> helping much with readability.  So not sure, just dropping it here if
> anyone has a good idea or storng opinion.

You /could/ use a bunch of fugly macros to simulate a default NULL
iomap_end_fn parameter.

#define __default_iomap_end(a,b,...) b
#define default_iomap_end(...) __default_iomap_end(,##__VA_ARGS,NULL)

#define iomap_process(iter, iomap, srcmap, iomap_begin, ...) \
	__iomap_process((iter), (iomap), (srcmap), (iomap_begin),
			default_iomap_end(__VA_ARGS__))

--D

> 
> Otherwise looks good:
> 
> Reviewed-by: Christoph Hellwig <hch@xxxxxx>
>