[igt-dev] [PATCH i-g-t 1/2] test/i915: gem_ctx_exec: Fix for using context engine map

[Tvrtko Ursulin]

On 19/11/2019 11:35, Chris Wilson wrote:
> Quoting Tvrtko Ursulin (2019-11-19 11:24:17)
>>
>> On 19/11/2019 11:08, Chris Wilson wrote:
>>> Quoting Venkata Sandeep Dhanalakota (2019-11-19 11:01:16)
>>>> Since engines are now looked up based on context engine map,
>>>> context should be setup with engines before using in execbuf.
>>>>
>>>> Fixes: 1a7f2e59 igt("gem_ctx_exec: use the gem_engine_topology"
>>>>
>>>> Cc: Andi Shyti <andi.shyti at intel.com>
>>>> Cc: Chris Wilson <chris at chris-wilson.co.uk>
>>>> Cc: Tvrtko Ursulin <tvrtko.ursulin at linux.intel.com>
>>>> Signed-off-by: Venkata Sandeep Dhanalakota <venkata.s.dhanalakota at intel.com>
>>>> ---
>>>>    tests/i915/gem_ctx_exec.c | 1 +
>>>>    1 file changed, 1 insertion(+)
>>>>
>>>> diff --git a/tests/i915/gem_ctx_exec.c b/tests/i915/gem_ctx_exec.c
>>>> index 6c2cd922..2bf47cf2 100644
>>>> --- a/tests/i915/gem_ctx_exec.c
>>>> +++ b/tests/i915/gem_ctx_exec.c
>>>> @@ -132,6 +132,7 @@ static void invalid_context(int fd, const struct intel_execution_engine2 *e,
>>>>           gem_execbuf(fd, &execbuf);
>>>>    
>>>>           ctx = gem_context_create(fd);
>>>> +       gem_context_set_all_engines(fd, ctx);
>>>
>>> Irrelevant.
>>
>> I guess it is a bit, but it completes the test conversion to be correct.
>> You would want to split the legacy-vs-new engine selection in this one?
> 
> No, there shouldn't be a per-engine component to this test at all.
> The way we have separated the context state is that it is akin to the
> process state. Now whether or not the context has an engine is another
> question (and answered elsewhere), the question is most certainly not
> whether an engine has a context as is being asked here.

Okay, option 3, drop the per-engine nature of this subtest altogether.

Regards,

Tvrtko