On Wed, Oct 23, 2024 at 10:27:06PM -0700, Christoph Hellwig wrote: > On Thu, Oct 24, 2024 at 03:54:54PM +1030, Qu Wenruo wrote: > > And generic/095 has a much higher chance to trigger it than the minimal one. > > The minimal one is only easier to debug and faster to finish one run. > > > > Do I still need to add the minimal one to fstests? > > If I have bug with a well known isolated reproducer I love > having it in xfstests. I don't know if everyone agrees, though. generic/095 isn't a specific test case for a known issue, it's a common test case. If you want to have a reproducer for a particular bug, feel free add it to xfstests, mark _fixed_by_xxxx and describe more details about that bug in the test case. If one day we update g/095, it might not cover the bug you need. Then a specific regression test case will help. Thanks, Zorro > >
