[igt-dev] [PATCH i-g-t] tests/gem_ctx_isoation: Remove dead code

Petri Latvala petri.latvala at intel.com
Fri Sep 7 07:11:23 UTC 2018


On Thu, Sep 06, 2018 at 06:19:20PM +0100, Chris Wilson wrote:
> Quoting Antonio Argenziano (2018-09-06 18:05:48)
> > 
> > 
> > On 06/09/18 09:58, Chris Wilson wrote:
> > > Quoting Antonio Argenziano (2018-09-06 17:51:56)
> > >> Probably just leftover code that escaped review.
> > > 
> > > The '//' indicates a wish that igt_ci_fail_on() existed. The problem as
> > > it currently stands is that we don't get a big warning to update this
> > > test for future gen. iirc the intent was to try and reinforce that this
> > > was a test limitation for CI and shouldn't alarm any other user.
> > 
> > Change it to an igt_warn()? Not sure how CI deals with those but maybe 
> > we can ask to mark them in a special way.
> 
> Seems reasonable. I can't remember if igt_warn + SKIP == orange flag in
> CI. If it does, yes that seems fine. It doesn't really matter that much,
> I just don't like indicating a fail in the case test hasn't been run
> because it is out of date. (But that's probably too much hairsplitting
> on my part.)

For the record: Currently, no, igt_warn + SKIP == SKIP == grey.

Good news is that it's changeable easily as soon as igt_runner is used
in CI.


-- 
Petri Latvala


More information about the igt-dev mailing list