On 6/10/19 9:16 AM, Chris Wilson wrote:
Quoting Tvrtko Ursulin (2019-06-10 16:54:13)
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_types.h b/drivers/gpu/drm/i915/gt/intel_engine_types.h
index 01223864237a..343c4459e8a3 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_types.h
+++ b/drivers/gpu/drm/i915/gt/intel_engine_types.h
@@ -34,6 +34,7 @@ struct drm_i915_reg_table;
struct i915_gem_context;
struct i915_request;
struct i915_sched_attr;
+struct intel_gt;
struct intel_uncore;
typedef u8 intel_engine_mask_t;
@@ -266,6 +267,7 @@ struct intel_engine_execlists {
struct intel_engine_cs {
struct drm_i915_private *i915;
+ struct intel_gt *gt;
I'd push for gt as being the backpointer, and i915 its distant grand
parent. Not sure how much pain that would bring just for the elimination
of one more drm_i915_private, but that's how I picture the
encapsulation.
Would it be worth moving some of the flags in the device_info structure
in a gt substructure, like we did for display, and get a pointer to that
in intel_gt? We could save some jumps back that way and be more coherent
in where we store the info.
Daniele
I'm sure I'm missing a link or two :)
-Chris
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
