Re: [RFC PATCH 6/6] selftests/damon/sysfs.py: test damon probes

SJ Park <[email protected]> Wed, 29 Jul 2026 17:07:19 -0700
Newsgroups dev.linux.lists.damon
Message-ID <[email protected]>
On Wed, 29 Jul 2026 15:19:33 +0000 [email protected] wrote:

> Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
> - [Medium] DamonProbe.stage() completely ignores the error returned when writing to the 'weight' sysfs file, masking potential write failures.
> --
> 
> --- Patch [6]: [RFC PATCH 6/6] selftests/damon/sysfs.py: test damon probes ---
> commit e47c18c7c1e4b19237b2738bc5c072c3c570dcc9
> Author: SJ Park <[email protected]>
> 
> selftests/damon/sysfs.py: test damon probes
> 
> This commit extends sysfs.py to test DAMON probes by committing them via
> sysfs and checking the in-kernel status using drgn.
> 
> > diff --git a/tools/testing/selftests/damon/sysfs.py b/tools/testing/selftests/damon/sysfs.py
> > index 8c2dfc5ed8e04..852b6887a75ce 100755
> > --- a/tools/testing/selftests/damon/sysfs.py
> > +++ b/tools/testing/selftests/damon/sysfs.py
> > @@ -319,7 +319,21 @@ def main():
> >                  intervals_goal=_damon_sysfs.IntervalsGoal(
> >                      access_bp@0, aggrs=3, min_sample_usP00,
> >                      max_sample_us000000),
> > -                update_us 00000),
> > +                update_us 00000,
> > +                probes=_damon_sysfs.DamonProbes(
> > +                    probes=[_damon_sysfs.DamonProbe(
> > +                        weightB,
> 
> [Severity: Medium]
> Will this properly catch setup failures when setting the probe weight?
> 
> When testing DamonProbes, DamonProbe.stage() writes the weight to sysfs
> but overwrites the returned error status without checking it:
> 
> tools/testing/selftests/damon/_damon_sysfs.py:DamonProbe.stage()
> 
>     def stage(self):
>         err = write_file(
>                 os.path.join(self.sysfs_dir(), 'weight'), '%d' % self.weight)
>         return self.filters.stage()
> 
> If the sysfs write fails, such as the kernel rejecting the weight value
> or the sysfs node being absent, the error is silently ignored. Could this
> cause the test suite to proceed as if setup succeeded, leading to false
> positive passes or confusing downstream errors?

Yes, I will fix it.

> 
> [ ... ]
> 
> -- 
> Sashiko AI review ยท https://sashiko.dev/#/patchset/[email protected]?part=6
> 


Thanks,
SJ