[Intel-gfx] [PATCH] drm/i915/audio: Hook up component bindings even if displays are disabled

[Chris Wilson]

Quoting Imre Deak (2018-08-20 14:23:13)
> On Fri, Aug 17, 2018 at 11:02:41AM +0100, Chris Wilson wrote:
> > If the display has been disabled by modparam, we still want to connect
> > together the HW bits and bobs with the associated drivers so that we can
> > continue to manage their runtime power gating.
> > 
> > Fixes: 108109444ff6 ("drm/i915: Check num_pipes before initializing audio component")
> > Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
> > Cc: Imre Deak <imre.deak at intel.com>
> > Cc: Takashi Iwai <tiwai at suse.de>
> > Cc: Jani Nikula <jani.nikula at linux.intel.com>
> > Cc: Elaine Wang <elaine.wang at intel.com>
> 
> Going through the hooks in i915_audio_component_ops, I haven't noticed
> anything that would actually go wrong with num_pipes=0. Besides some
> audio HW register programming from i915_audio_component_codec_wake_override(),
> we'd skip doing everything else due to not reporting any active encoders
> to the audio component (as we do during normal display modeset). So
> 
> Reviewed-by: Imre Deak <imre.deak at intel.com>

I'm confident in that we at least exercise the no-pipes case, less
confident that we test for fused-off hw. However, the audio component
simply should not find any output channels even in that case, so it
indeed should be safe.

Cross your fingers, thanks for the review,
-Chris