[Intel-gfx] [PATCH 7/7] drm/i915/skl: Warn if the cached cdclk freq is not the requested one

[Ville Syrjälä]

On Fri, Jun 05, 2015 at 01:40:29PM +0100, Damien Lespiau wrote:
> On Fri, Jun 05, 2015 at 03:24:45PM +0300, Ville Syrjälä wrote:
> > On Thu, Jun 04, 2015 at 06:21:35PM +0100, Damien Lespiau wrote:
> > > Signed-off-by: Damien Lespiau <damien.lespiau at intel.com>
> > > ---
> > >  drivers/gpu/drm/i915/intel_display.c | 4 ++++
> > >  1 file changed, 4 insertions(+)
> > > 
> > > diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
> > > index a232dc9..a018465 100644
> > > --- a/drivers/gpu/drm/i915/intel_display.c
> > > +++ b/drivers/gpu/drm/i915/intel_display.c
> > > @@ -5627,6 +5627,10 @@ static void skl_set_cdclk(struct drm_i915_private *dev_priv, unsigned int freq)
> > >  	mutex_unlock(&dev_priv->rps.hw_lock);
> > >  
> > >  	intel_update_cdclk(dev);
> > > +
> > > +	WARN(freq != dev_priv->cdclk_freq,
> > > +	     "cdclk requested %d kHz but got %d kHz\n",
> > > +	     freq, dev_priv->cdclk_freq);
> > >  }
> > 
> > Could you add this to all the set_cdclk() functions? Maybe
> > intel_check_cdclk() or something.
> 
> I was thinking that we should probably introduce a low level
> set_core_display_clock() vfunc and a intel_set_core_display_clock()
> wrapper were we'd put the common code (updating the cached value, that
> WARN(), ...)
> 
> Thoughts?

Yeah wrapper around the vfunc sounds better than sprinkling the same
stuff into every vfunc.

-- 
Ville Syrjälä
Intel OTC