Re: [PATCH] shmem: do not extend i_size when no data was copied

[Jan Kara]

On Tue 03-11-26 02:34:11, Hungyu Lin wrote:
> On Mon, Oct 26, 2026 at 06:18:39PM +0000, Kiryl Shutsemau wrote:
> > The per-filesystem alternative is a 'copied &&' in every ->write_end.
> 
> Would it make sense to keep this fix local to shmem for now?
> I'll check the folio handling before sending v2.

The trouble is that practically any filesystem is going to have these
issues with 0-length writes starting beyond i_size. So if we can fix it in
generic code, it would be certainly easier in the grand scheme of things.

								Honza
-- 
Jan Kara <jack@xxxxxxxx>
SUSE Labs, CR