Re: [PATCH v7 04/10] selftests/harness: Fix interleaved scheduling leading to race conditions

[Mark Brown]

On Tue, Jun 04, 2024 at 06:06:48PM +0200, Mickaël Salaün wrote:

> Thanks for the heads up.  I warned about not being able to test
> everything when fixing kselftest last time, but nobody show up.  Is
> there an easy way to run most kselftests?  We really need a (more
> accessible) CI...

You can just invoke the top level kselftests Makefile but between things
being flaky and runtime requirements there's a bunch of noise there.
KernelCI covers a bunch of it and would be my go to, I've got a good
chunk of the selftests that actually build and run reliably in my
personal CI but it has no visible UI.  Part of the issue here might be
platform specifics, I'm seeing this on arm64.  

> > > FWIW I'm still seeing this on -rc2...

> > AFAICT this is due to the switch to using clone3() with CLONE_VFORK

> I guess it started with the previous vfork() that was later replaced
> with CLONE_VFORK.

Bisect did seem to point at this commit FWIW, I've not dug into any API
differences or anything here.  The immediate thing being replaced was a
plain fork() though I see it was vfork() at some point before that, and
I'd not have noticed if the individual testcases weren't hanging so the
timeout was needed.

> > I'm not clear what the original race being fixed here was but it seems
> > like we should revert this since the timeout functionality is pretty
> > important?

> It took me a while to fix all the previous issues and it would be much
> easier to just fix this issue too.

> I'm working on it.

Great, thanks.

Attachment: signature.asc
Description: PGP signature