Quoting Petri Latvala (2017-06-21 01:37:21) > On 06/20/2017 08:59 PM, Dylan Baker wrote: > > Quoting Petri Latvala (2017-06-20 05:41:11) > >> On 04/13/2017 09:46 PM, Dylan Baker wrote: > >>> Quoting Brian Paul (2017-04-12 13:04:59) > >>>> On 04/12/2017 11:55 AM, Dylan Baker wrote: > >>>>> It doesn't makes sense to run if a user has removed all tests from a > >>>>> selected profile, and currently if all tests are removed, then an > >>>>> assertion will be hit in the backend that isn't extremely clear about > >>>>> what went wrong. This should be much easier to understand. > >>>>> > >>>>> Signed-off-by: Dylan Baker <[email protected]> > >>>>> --- > >>>>> framework/profile.py | 7 +++++++ > >>>>> 1 file changed, 7 insertions(+) > >>>>> > >>>>> diff --git a/framework/profile.py b/framework/profile.py > >>>>> index 4604367e1..ce0b24ce8 100644 > >>>>> --- a/framework/profile.py > >>>>> +++ b/framework/profile.py > >>>>> @@ -389,6 +389,13 @@ def run(profiles, logger, backend, concurrency): > >>>>> profiles = [(p, list(p.itertests())) for p in profiles] > >>>>> log = LogManager(logger, sum(len(l) for _, l in profiles)) > >>>>> > >>>>> + # check that after the filters are run there are actually tests to > >>>>> run > >>>>> + for p, l in profiles: > >>>>> + if not l: > >>>>> + raise exceptions.PiglitUserError( > >>>>> + 'After running filters there are no tests in ' > >>>>> + 'profile "{}"'.format(p.name)) > >> Bumped into an issue caused by this commit with IGT tests. > >> > >> When the last test of a run never finishes and you attempt to > >> > >> piglit resume -n test-results-path > >> > >> this exception is raised instead of the expected result of finishing up > >> the run. > >> > >> > >> -- > >> Petri Latvala > >> > > That is expected with the command line you've supplied. > > > > The -n/--no-retry option instructs piglit to not retry tests that started > > but > > didn't finish, if you get the same behavior without the -n option that is a > > bug. > > > > Dylan > > > What is the supported method then to loop resume until results.json.bz2 > gets generated? Previous behaviour was that piglit reported that there's > nothing more to do and generated that. > > > -- > Petri Latvala >
I would use `piglit summary aggregate` to build the results.json.foo file.
signature.asc
Description: signature
_______________________________________________ Piglit mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/piglit
