[Intel-gfx] [PATCH 2/4] drm/i915: Add comments about fixed pipe->transcoder/PLL mapping

[Ville Syrjälä]

On Thu, May 12, 2016 at 05:00:56PM +0300, Imre Deak wrote:
> On Thu, 2016-05-12 at 16:30 +0300, Ville Syrjälä wrote:
> > On Thu, May 12, 2016 at 04:18:50PM +0300, Imre Deak wrote:
> > > Code checkers may complain about the explicit casts between different
> > > enum types, so add comments for known-valid cases to help future
> > > triaging of such complaints.
> > > 
> > > Signed-off-by: Imre Deak <imre.deak at intel.com>
> > > ---
> > >  drivers/gpu/drm/i915/intel_display.c | 5 +++++
> > >  1 file changed, 5 insertions(+)
> > > 
> > > diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
> > > index b60d9b6..763726d 100644
> > > --- a/drivers/gpu/drm/i915/intel_display.c
> > > +++ b/drivers/gpu/drm/i915/intel_display.c
> > > @@ -9294,6 +9294,7 @@ static bool ironlake_get_pipe_config(struct intel_crtc *crtc,
> > >  		ironlake_get_fdi_m_n_config(crtc, pipe_config);
> > >  
> > >  		if (HAS_PCH_IBX(dev_priv)) {
> > > +			/* the pipe->pll mapping is fixed */
> > 
> > Should perhaps say 'pch transocder->pll mapping' here.
> 
> Ok, but then 'pipe->pch transcoder and pch transcoder->pll mapping is
> fixed'?

I guess you could say that. We don't have a separate enum for pch
transcoder since it's the same as the pipe always, and thus we usually
use 'pipe' to denote the pch transcoder as well. Well, except on
HSW/BDW which only have pch transcoder A, and sometimes on those
platforms we use enum transcoder also for the pch transcoder.
So it's a huge mess really, and I was considering adding a pch
transcoder enum at some point but so far I didn't get bored enough
to actually do it.

> 
> > >  			pll_id = (enum intel_dpll_id) crtc->pipe;
> > >  		} else {
> > >  			tmp = I915_READ(PCH_DPLL_SEL);
> > > @@ -9844,6 +9845,10 @@ static bool hsw_get_transcoder_state(struct intel_crtc *crtc,
> > >  	enum intel_display_power_domain power_domain;
> > >  	u32 tmp;
> > >  
> > > +	/*
> > > +	 * The transcoder->pipe mapping is fixed with the exception of the eDP
> > 
> > Saying 'pipe->transcoder mapping' would feel more natural to me
> > since that's the way the data flows. Also that's the way assignment in
> > the code happens.
> 
> Ok, will change that.
> 
> > But I can live with the current thing too, so
> > Reviewed-by: Ville Syrjälä <ville.syrjala at linux.intel.com>
> > 
> > > +	 * transcoder handled below.
> > > +	 */
> > >  	pipe_config->cpu_transcoder = (enum transcoder) crtc->pipe;
> > >  
> > >  	/*
> > > -- 
> > > 2.5.0
> > > 
> > > _______________________________________________
> > > Intel-gfx mailing list
> > > Intel-gfx at lists.freedesktop.org
> > > https://lists.freedesktop.org/mailman/listinfo/intel-gfx
> > 

-- 
Ville Syrjälä
Intel OTC