[Intel-gfx] [RFC v2 1/1] drm/i915: Power gating display wells during i915_pm_suspend

[Daniel Vetter]

On Fri, Jul 25, 2014 at 01:28:48PM +1000, Dave Airlie wrote:
> On 23 July 2014 15:11, Daniel Vetter <daniel at ffwll.ch> wrote:
> > On Sat, Jul 12, 2014 at 10:02:27AM +0530, sagar.a.kamble at intel.com wrote:
> >> From: Borun Fu <borun.fu at intel.com>
> >>
> >> On VLV, after i915_pm_suspend display power wells are staying
> >> power ungated. So, after initiating mem sleep "echo mem > /sys/power/state"
> >> Display is staing D0 State. There might be better way/place to power gate
> >> these wells. Also, we need to make sure that if wells are power gated due to
> >> DPMS OFF sequence, they need not be turned off by i915_pm_suspend again.
> >>
> >> v2: Extracted helper for intel_crtc_disable and power gating CRTC power wells.
> >> [Daniel]
> >>
> >> Cc: Imre Deak <imre.deak at intel.com>
> >> Cc: Paulo Zanoni <paulo.r.zanoni at intel.com>
> >> Cc: Daniel Vetter <daniel.vetter at ffwll.ch>
> >> Cc: Jani Nikula <jani.nikula at linux.intel.com>
> >> Change-Id: I34c80da66aa24c423a5576c68aa1f3a8d0f43848
> >
> > s-o-b from the original author (Borun Fu) missing. Added myself since we
> > all work for the same company, but please don't forget this. Every person
> > including the original author, who handles a patch must add their sob
> > line.
> > -Daniel
> 
> Is this queued or on its way, I was getting a warning on HSW about not
> entering pc8+
> due to display with MST enabled, and I thought it was MSTs fault, but I suspect
> its just this,
> 
> mode set turns the global resources power well on, but nothing ever
> turns it off.

mst doesn't factor into this. Might need to double-check but from what
I've seen mst was only wired into the system s/r paths, not the runtime
pm. They're unfortunately not sufficiently shared. So I'd suspect that.

The bug here happens when you move/update the cursor (or sprite/primary
plane) while the display is off. You'll get unclaimed register read/write
errors, no WARNINGs if you hit this one.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch