Re: [PATCH v4 1/4] mm/gup: fix wrongly calculated returned value in fault_in_safe_writeable()

[Andreas Gruenbacher]

On Fri, Apr 11, 2025 at 7:32 AM Baoquan He <bhe@xxxxxxxxxx> wrote:
> On 04/10/25 at 08:43pm, Andrew Morton wrote:
> > On Thu, 10 Apr 2025 11:57:14 +0800 Baoquan He <bhe@xxxxxxxxxx> wrote:
> >
> > > Not like fault_in_readable() or fault_in_writeable(), in
> > > fault_in_safe_writeable() local variable 'start' is increased page
> > > by page to loop till the whole address range is handled. However,
> > > it mistakenly calcalates the size of handled range with 'uaddr - start'.
> >
> > What are the userspace-visible runtime effects of this change?
>
> I see it mainly affect gfs2_file_direct_read(). Not sure if GFS2 people
> can sense any exceptional behaviour caused by this code bug.

Thanks for the heads up.

In gfs2, fault_in_iov_iter_writeable() is used in
gfs2_file_direct_read() and gfs2_file_read_iter(), so this potentially
affects buffered as well as direct reads. This bug could cause those
gfs2 functions to spin in a loop.

Can this fix please be sent to Linus for inclusion into 6.15?

Thanks,
Andreas