On Fri, Feb 09, 2018 at 09:16:52PM +0000, Chris Wilson wrote:
> diff --git a/drivers/gpu/drm/i915/intel_device_info.h
> b/drivers/gpu/drm/i915/intel_device_info.h
> index 8a570f75d67e..d5f34f124e34 100644
> --- a/drivers/gpu/drm/i915/intel_device_info.h
> +++ b/drivers/gpu/drm/i915/intel_device_info.h
> @@ -98,7 +98,6 @@ enum intel_platform {
> func(has_logical_ring_contexts); \
> func(has_logical_ring_preemption); \
> func(has_overlay); \
> - func(has_pooled_eu); \
> func(has_psr); \
> func(has_rc6); \
> func(has_rc6p); \
> @@ -123,6 +122,7 @@ struct sseu_dev_info {
> u8 has_slice_pg:1;
> u8 has_subslice_pg:1;
> u8 has_eu_pg:1;
> + u8 has_pooled_eu:1;
> };
>
> struct intel_device_static_info {
> @@ -139,8 +139,6 @@ struct intel_device_static_info {
> u32 display_mmio_offset;
>
> u8 num_pipes;
> - u8 num_sprites[I915_MAX_PIPES];
> - u8 num_scalers[I915_MAX_PIPES];
num_sprites and num_scalers are actually static information, so we could
consider removing the runtime initialization for them instead.
Not sure what to do with num_sprites actually. We might think about just
removing it entirely, although we'd still want something to figure out
how many planes to initialize. That something could be just a local
variable in the display init code though. Oh, and I suppose the same
could be done for num_scalers since we copy it over into
crtc->num_scalers anyway.
PS. the patch subject is misleading since you do a lot more than the
sseu stuff here.
--
Ville Syrjälä
Intel OTC
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
