Re: [MAINTAINERS SUMMIT] The place of AI code review in the Linux Kernel process

[Jason Gunthorpe <jgg@nvidia.com>]

On Fri, Jul 17, 2026 at 04:17:06PM -0700, Linus Torvalds wrote:

> > > I've gone and fixed a bunch on my own, mostly out of fear that they
> > > will just disappear and be lost, but it is an annoying urgency. I'd
> > > rather have as a giant todo list (that maybe other people could help
> > > with too)
> > >
> > > Often alot of tokens are spent to find these things, it feels wrong
> > > that they are effectively lost in the endless stream of reviews..
> 
> I think this ends up being one of those things that easily frustrates
> maintainers: low-grade constant noise from an AI that adds extra
> overhead and another thing people feel they need to look at.

To be clear, here I am thinking of things that are not low grade. Eg
recently sashiko found a drive by in a review and then a few weeks
later the same issue was sent as a security@ email.

So in my mind this is more like syzkaller where it found something, it
does need fixing. I don't want to drop everything and fix it now, I'm
OK leaving it on the syzkaller dashboard for a month, while trying to
get someone to donate fixing it. I don't want to block anyone, but I
don't want to loose this valuable information.

Since I want to at least try to delegate it is helpful to have a
public dashboard instead of my own todo list.

> (Or feel they really don't have time or interest to look at, and then
> just the knowledge that there's that nagging thing that you don't
> really care about adds stress and frustration).

Right, I'd feel better if there was reliable tracking for these
discoveries. I don't want to make the unlucky sumbitter do anything
about them as a pre-condition to make progress on whatever they are
doing, unless the series is making it worse or doubling down on
wrongness.

Jason