Re: [PATCH v1] mm/mempolicy: fix migrate_to_node() assuming there is at least one VMA in a MM

From: Liam R. Howlett
Date: Fri Nov 22 2024 - 09:44:57 EST


* David Hildenbrand <david@xxxxxxxxxx> [241122 04:32]:
> On 22.11.24 07:19, Andrew Morton wrote:
> > On Wed, 20 Nov 2024 15:27:46 -0500 "Liam R. Howlett" <Liam.Howlett@xxxxxxxxxx> wrote:
> >
> > > I hate the extra check because syzbot can cause this as this should
> > > basically never happen in real life, but it seems we have to add it.
> >
> > So..
> >
> > --- a/mm/mempolicy.c~mm-mempolicy-fix-migrate_to_node-assuming-there-is-at-least-one-vma-in-a-mm-fix
> > +++ a/mm/mempolicy.c
> > @@ -1080,7 +1080,7 @@ static long migrate_to_node(struct mm_st
> > mmap_read_lock(mm);
> > vma = find_vma(mm, 0);
> > - if (!vma) {
> > + if (unlikely(!vma)) {
> > mmap_read_unlock(mm);
> > return 0;
> > }
> > _
> >
> > ?
>
> Why not, at least for documentation purposes. Because I don't think this is
> any fast-path we really care about, so expect the runtime effects to be
> mostly negligible. Thanks!

The next email we get about this will be a bot with a micro benchmark
performance drop.

Really though, I'm happy either way because of what David said.

Thanks,
Liam