[Intel-gfx] [PATCH v2 0/8] RPM Encapsulation

[Chris Wilson]

Quoting Daniele Ceraolo Spurio (2019-06-12 19:37:39)
> Rebased series after the display power code has been split out.
> Mostly the same as V1, just small fixes based on review comments
> and CI results. The last patch is new.
> 
> The series grows the code a bit, but I believe it is worth it for the
> extra logical encapsulation. With all debug options disabled:
> 
> add/remove: 0/0 grow/shrink: 92/14 up/down: 1225/-162 (1063)
> Total: Before=1266893, After=1267956, chg +0.08%
> 
> Daniele Ceraolo Spurio (8):
>   drm/i915: prefer i915_runtime_pm in intel_runtime function
>   drm/i915: Remove rpm asserts that use i915
>   drm/i915: make enable/disable rpm assert function use the rpm
>     structure
>   drm/i915: move and rename i915_runtime_pm
>   drm/i915: move a few more functions to accept the rpm structure
>   drm/i915: update rpm_get/put to use the rpm structure
>   drm/i915: update with_intel_runtime_pm to use the rpm structure
>   drm/i915: make intel_wakeref work on the rpm struct

Looks fine. I can't think of any neat way to avoid the extra offsets, so
we'll just have to think of a better way of dividing up the driver to
avoid loading unneeded code.

Preemptive Reviewed-by: Chris Wilson <chris at chris-wilson.co.uk>
on the series as nothing leapt out at me from the mechanical churn. 
Please cc Imre on the rebase in case he has a good reason to restructure
it differently.
-Chris