[Intel-gfx] [Regression report] Weekly regression report WW40

[Jani Nikula]

On Wed, 05 Oct 2016, "Argotti, Yann" <yann.argotti at intel.com> wrote:
>> > On Mon, 03 Oct 2016, Jairo Miramontes
>> > <jairo.daniel.miramontes.caton at linux.intel.com> wrote:
>> > > This week regressions
>> >
>> > In the past we used "regression", "bisect_pending", and "bisected" in
>> > the bugzilla "Keywords" field. Can we start using those again, please?
>> 
>> I think this is a very good idea Jani. So we can start to scrub current
>> regression (and then igt linked one) and update accordingly.
>> Yann
>
> Two additional thoughts are:
> - add "regression_pending" (vs "regression") as well as keyword to
> indicate where bug reporter has doubt on the fact it is or not a
> regression.

I think it should be enough, at least for a start (or should I say
re-start), to freely use "regression" when things worked in the past but
do not work anymore. We can then drop the keyword if it's not proven to
be a regression.

No strong feelings on "regression_pending", though. But I don't think
you can add keywords at will, I think the field only accepts a
predefined set of keywords. So you'd need to talk to Ander or Martin
(Cc'd) to get the new one to bugzilla.

> - either use of "regression" or "regression_pending" must be coming
> with good & bad commit when reporting the bug

Agreed.

BR,
Jani.

-- 
Jani Nikula, Intel Open Source Technology Center