Re: [syzbot] [mm?] WARNING in vma_modify (2)

[Lorenzo Stoakes]

On Mon, Nov 24, 2025 at 11:25:14AM +0100, Aleksandr Nogikh wrote:
> Hi Lorenzo,
>
> Thanks for taking a closer look!
>
> On Mon, Nov 24, 2025 at 10:53 AM 'Lorenzo Stoakes' via syzkaller-bugs
> <syzkaller-bugs@xxxxxxxxxxxxxxxx> wrote:
> >
> > On Sat, Nov 22, 2025 at 09:30:29PM -0800, syzbot wrote:
> > > Hello,
> > >
> > > syzbot found the following issue on:
> >
> > Hi, thanks for the report!
> >
> > >
> > > HEAD commit:    fe4d0dea039f Add linux-next specific files for 20251119
> >
> > Hm this is quite far behind!
> >
> > This is a known issue that has already been fixed in linux-next.
>
> Could you please point to the patch series that introduced the original issue?

It's two versions of the same:

A fix-patch sent for v3 introduced it on Mon 17th Nov -
https://lore.kernel.org/all/13ae542f-5c47-4511-b0ad-6d1e9dcbdc9d@lucifer.local/

v4 fixed it on Tues 18th Nov -
https://lore.kernel.org/all/cover.1763460113.git.lorenzo.stoakes@xxxxxxxxxx/

Unfortunately it seemed the former got sent to -next quicker than the latter
did.

So next-20251119 introduced the bug, next-20251120 fixes it.

It's all rebased so there's no meaningful commit to mention etc.

> I'd love to debug why https://ci.syzbot.org/ missed it.

It's not that it missed it I guess, it just worked through a backlog. But I'd
say it's not really useful at all to do so for prior versions of -next? You
should always do everything against the latest, since -next _always_ rebases
like this.

It's useful to look at older commits for Linus's tree, as we don't want bugs to
exist at any commit there. For -next it's not really helpful.

>
> >
> > It'll all be rebased so no syz command is going to be too useful here I think!
>
> Yes, we can only close it as invalid so that the report doesn't stay open:
> #syz invalid

Ah wasn't aware of that one. Maybe worth putting in footer?

>
> --
> Aleksandr

Cheers, Lorenzo