[igt-dev] [PATCH v21 6/6] test: perf_pmu: use the gem_engine_topology library

Tvrtko Ursulin tvrtko.ursulin at linux.intel.com
Wed Apr 17 13:42:31 UTC 2019


On 17/04/2019 00:05, Andi Shyti wrote:
>>> @@ -1046,14 +1057,18 @@ static void cpu_hotplug(int gem_fd)
>>>    	igt_require(cpu0_hotplug_support());
>>> +	ctx = gem_context_create(gem_fd);
>>
>> TODO left was to try without new contexts for a handful of tests where I
>> added them. Instead try I915_EXEC_DEFAULT when submitting and (ab)use the
>> fact RCS is at index zero for foreseeable future. When it fails we'll know
>> about it. This applies to hotplug, frequency, interrupts subtests AFAIR.
> 
> yes, I was also thinking about adding a few wrappers around.

I'm afraid I lost you here, I did not mention any wrappers here? I was 
thinking about the thing we talked about before, where I knee-jerk added 
context creation to some subtests and have later realized we could 
possibly do without that.

> I managed to build quite a hefty list of things to do in these 21
> versions (22 with the coming)

Next time you can occasionally send updated individual patches as 
--in-reply-to instead of resending the whole series which would keep the 
cover letter change log smaller. ;)

You only need to send the whole series when there are new or removed 
patches, if the time delay between posts is long, or if there was as 
rebase/conflict due base moving. Or if the reply chain gets too long. 
Those are my guidelines anyway, I don't claim they are somewhere 
documented or match other people opinions.

> But I would happily send them once this series gets accepted to
> not procrastinate it further.

In my view it is basically almost there and accepted in principle, just 
needs some final tweaks.

Regards,

Tvrtko


More information about the igt-dev mailing list