[Intel-gfx] 4.0 -> 4.1 regression : after resume from s2ram both internal and external display of a docked ThinkPad ate black

[Pavel Machek]

> >>> commit e7d6f7d708290da1b7c92f533444b042c79412e0
> >>> Author: Dave Airlie <airlied at redhat.com>
> >>> Date:   Mon Dec 8 13:23:37 2014 +1000
> >>>
> >>>     drm/i915: resume MST after reading back hw state
....
> >> Is there anything else what I can do ?
> >>
> >> Current kernels up to 4.2.3 and 4.3-rc3 (not hardened) shows this issue here at my system.
> > 
> > Yes. Now you ask Dave Airlie <airlied at redhat.com> to fix it. If that
> 
> Dear Dave,
> 
> please fix it.
> 
> Here's a work around which works for me since kernel 4.1.x :

Dave. You broke it. You fix it. Don't make me less polite?

Daniel? Jani? Can you apply the patch below, or comment whats wrong
with that?

This is a regression, so it should not require much thinking.

									Pavel

> diff --git a/drivers/gpu/drm/i915/i915_drv.c
> b/drivers/gpu/drm/i915/i915_drv.c
> index ab64d68..3aeead2 100644
> --- a/drivers/gpu/drm/i915/i915_drv.c
> +++ b/drivers/gpu/drm/i915/i915_drv.c
> @@ -740,6 +740,8 @@ static int i915_drm_resume(struct drm_device *dev)
>         if (dev_priv->display.hpd_irq_setup)
>                 dev_priv->display.hpd_irq_setup(dev);
>         spin_unlock_irq(&dev_priv->irq_lock);
> +
> +       intel_dp_mst_resume(dev);
> 
>         drm_modeset_lock_all(dev);
>         intel_display_resume(dev);
> 
> 
> > does not work, you ask him to fix it, in less polite words. If that
> > does not work, you verify that reverting
> > e7d6f7d708290da1b7c92f533444b042c79412e0 fixes it for you, then ask
> > Daniel Vetter and Jani Nikula to revert it. If they fail to do that,
> > you go all the way up to Linus.
> > 
> > Good luck ;-), 
> > 									Pavel
> > 
> 
> 

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html