Re: Stop false review statements

[Jani Nikula <jani.nikula@intel.com>]

On Sun, 17 May 2026, Jonathan Corbet <corbet@lwn.net> wrote:
> That said, I was certainly considering *human* reviewers at the time,
> and all of the people who agreed with the suggested policy were too.
> Adding bots seems like a stretch to me.
>
> I can't speak for subsystems that require Reviewed-by tags on their
> commits, but I'm not sure that their maintainers would accept an
> automated review as satisfying that requirement.

For the parts of the drm subsystem that have a committer/maintainer
model, one of the requirements for pushing is that at least two people
have been involved. Reviewed-by is one of the ways to record this is
indeed the case.

I can't speak for the entire subsystem either, but to me it was always
about people, trust, and the community.

We may add ways to record that an LLM has reviewed a patch, but I think
the fundamental requirement that two human beings have been involved is
going to remain.


BR,
Jani.


-- 
Jani Nikula, Intel