Re: [PATCH 24/42] migration-test: Re-enable multifd_cancel test

[Daniel P . Berrangé]

On Fri, Jun 09, 2023 at 12:22:33PM +0200, Juan Quintela wrote:
> Daniel P. Berrangé <berrange@redhat.com> wrote:
> > On Fri, Jun 09, 2023 at 12:49:25AM +0200, Juan Quintela wrote:
> >
> > Please explain why this is considered ok, given the comment about
> > why it is disabled. ie if we fixed something, refrence the commit.
> 
> I did in the cover letter, will put that on the commit:
> 
> - We used to share dest_serial file for the two targets of migration (to
>   and to2), where we have a race.
> 
> - this series fixes the races in two ways:
> 
>   * we wait for "to" to finish before we launch "to2", so the race can't
>     happen.
> 
>   * One of the reasons why I created GuestState is that I needed a place
>     to store the serial file name (now I call it "target" and "target2").
> 
> - I put on the cover letter that this is not enough, we also need
>   Fabiano fix for the thread list.
> 
> - Peter Mayel was the most vocal about this particular failure, I cc'd
>   him and asked on the cover letter for the people to used to have
>   failures to test.
> 
> So how should I handled this to be clearer?

Details just need to be in the commit message, because cover letters
aren't visible when someone is looking back at 'git log' to find out
why the test was re-enabled.


With regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|