Re: Sashiko review emails to the list

[Marc Zyngier <maz@kernel.org>]

On Fri, 19 Jun 2026 17:45:25 +0100,
Oliver Upton <oupton@kernel.org> wrote:
> 
> Hey,
> 
> On Fri, Jun 19, 2026 at 03:19:05PM +0100, Fuad Tabba wrote:
> > Hi folks,
> > 
> > I really like Sashiko and find it very useful. It's flagged real bugs
> > in series I and others have posted to the list (e.g. [1][2][3]), and I
> > run it locally before sending, which has saved me a few respins.
> > 
> > That said, it's been posting a lot lately, so it seemed worth asking
> > how the review emails to the list are working out, and whether we
> > should change anything.
> 
> So this is entirely my fault since I added the email configuration for
> the kvmarm list. Sashiko has been finding some truly nasty bugs, posting
> on-list is the easiest way to get attention from the right folks to get
> things fixed.

I don't mind this. Having Sashiko to tell me that something is off is
great, but we collectively need to learn how to treat the feedback.

> With that being said, the signal to noise ratio hasn't been ideal.

There are a few things that did it for me:

- hallucinations that are close enough to reality that you spend
  significant amount of time debunking it. Being creative about
  specifications is particularly annoying.

- the whole "I found a preexisting issue" is great, but leads to an
  avalanche of patches that themselves trigger cross-reporting
  (patch fixing issue A triggers a report for issue B, and the
  reciprocal happens as well). This is overwhelming and we need to
  better manage this.

- Related to the above, I see a flurry of activity following a review
  from Sashiko, resulting in unrelated patches that distract from the
  issue at hand. I reckon that *we* could do a better job keeping the
  initial patch series focused, and once that is solved, work on the
  tangential stuff that Sashiko may have reported. At the moment, it's
  hard to see the wood for the tree.

- We also often end-up with point fixes match the reported issue, but
  that don't address more general issues. That's even more noise.

My conclusion is that while Shashiko reviews can often be flawed, most
of the problems are with the way we deal with the reports. No, I don't
have a magic solution, but that's something to think about.

>
> > These fixes will take a while to land, so the question is what to do
> > meanwhile. Some options, and surely others:
> > 
> >   - Leave it as-is while the fixes propagate.
> >   - Stop the emails but keep the reviews on sashiko.dev, so people can
> > look rather than have them pushed.
> >   - Disable the emails until the noise is down to a reasonable level,
> > then re-enable.
> 
> This sounds like the right approach. I'd like to re-enable emails once
> we're happy with the quality of reviews.

I'd like to see how the fixes improve the situation, and re-evaluate
in a couple of weeks.

Thanks,

	M.

-- 
Without deviation from the norm, progress is not possible.