[Intel-gfx] [PATCH 1/5] drm/i915: Record the sseu configuration per-context & engine

Tvrtko Ursulin tvrtko.ursulin at linux.intel.com
Wed Jan 30 09:56:25 UTC 2019


On 30/01/2019 09:40, Chris Wilson wrote:
> Quoting Tvrtko Ursulin (2019-01-30 09:34:12)
>> diff --git a/drivers/gpu/drm/i915/i915_request.h b/drivers/gpu/drm/i915/i915_request.h
>> index 3cffb96203b9..ecef6256dd24 100644
>> --- a/drivers/gpu/drm/i915/i915_request.h
>> +++ b/drivers/gpu/drm/i915/i915_request.h
>> @@ -38,6 +38,16 @@ struct drm_i915_gem_object;
>>   struct i915_request;
>>   struct i915_timeline;
>>   
>> +/*
>> + * Powergating configuration for a particular (context,engine).
>> + */
>> +struct intel_sseu {
>> +       u8 slice_mask;
>> +       u8 subslice_mask;
>> +       u8 min_eus_per_subslice;
>> +       u8 max_eus_per_subslice;
>> +};
> 
> Was there a reason why it ended up here? I noticed when I splitting out
> the intel_context into its own header. It doesn't matter in this series,
> just wondering if I broke a good rationale :)

It sure sticks out enough to make oneself ask that question.

I can only assume there was something in i915_request.h at some point 
which used it and some include hell made it the easiest solution. But 
it's a pure guess at this point.

I've tried moving it to i915_gem_context.h and it all still compiles 
fine. I better change it then..

Regards,

Tvrtko


More information about the Intel-gfx mailing list