[Intel-gfx] [PATCH 2/8] drm/i915/vlv/chv: Move resume_prepare() after uncore_early_sanitize()

[Damien Lespiau]

> > diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
> > index 5cc57f2..5a9399c 100644
> > --- a/drivers/gpu/drm/i915/i915_drv.c
> > +++ b/drivers/gpu/drm/i915/i915_drv.c
> > @@ -808,17 +808,17 @@ static int i915_drm_resume_early(struct drm_device *dev)
> >  
> >  	pci_set_master(dev->pdev);
> >  
> > -	if (IS_VALLEYVIEW(dev_priv))
> > -		ret = vlv_resume_prepare(dev_priv, false);
> > -	if (ret)
> > -		DRM_ERROR("Resume prepare failed: %d,Continuing resume\n", ret);
> > -
> >  	intel_uncore_early_sanitize(dev, true);
> >  
> > -	if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
> > -		hsw_disable_pc8(dev_priv);
> > -	else if (IS_SKYLAKE(dev_priv))
> > +	if (IS_SKYLAKE(dev_priv))
> >  		ret = skl_resume_prepare(dev_priv);
> > +	else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
> > +		hsw_disable_pc8(dev_priv);
> > +	else if (IS_VALLEYVIEW(dev_priv))
> > +		ret = vlv_resume_prepare(dev_priv, false);
> > +
> > +	if (ret)
> > +		DRM_ERROR("Resume prepare failed: %d,Continuing resume\n", ret);
> 
> vlv_resume_prepare() needs to run before intel_uncore_early_sanitize(),
> as the former includes the steps to wake the HW from RC6 for the first
> time after suspend. At the time intel_resume_prepare() was removed, I
> suggested to instead split out the part from vlv_resume_prepare() that
> needs to be done early into a separate function, so that would be one
> way to go about this.

Oh, that was a carefully crafted trap! Will see what I can do.
Thankfully the active ingredients of that series don't depend on that
refactoring.

-- 
Damien