[Intel-gfx] [PATCH] drm/i915/glk: CDCLK calculation changes for glk

[Ander Conselvan De Oliveira]

On Thu, 2017-03-16 at 15:10 +0200, Jani Nikula wrote:
> On Thu, 16 Mar 2017, "Chauhan, Madhav" <madhav.chauhan at intel.com> wrote:
> > > -----Original Message-----
> > > From: Nikula, Jani
> > > Sent: Thursday, February 16, 2017 9:03 PM
> > > To: Chauhan, Madhav <madhav.chauhan at intel.com>; intel-
> > > gfx at lists.freedesktop.org
> > > Cc: Conselvan De Oliveira, Ander <ander.conselvan.de.oliveira at intel.com>;
> > > Shankar, Uma <uma.shankar at intel.com>; Mukherjee, Indranil
> > > <indranil.mukherjee at intel.com>; Sharma, Shashank
> > > <shashank.sharma at intel.com>; Chauhan, Madhav
> > > <madhav.chauhan at intel.com>; ville.syrjala at linux.intel.com
> > > Subject: Re: [PATCH] drm/i915/glk: CDCLK calculation changes for glk
> > > 
> > > On Thu, 16 Feb 2017, Madhav Chauhan <madhav.chauhan at intel.com>
> > > wrote:
> > > > As per BSPEC, valid cdclk values for glk are 79.2, 158.4, 316.8 Mhz.
> > > > Practically we can achive only 99% of these cdclk values(HW team
> > > > checking on this). So cdclk should be calculated for the given pixclk
> > > > as per that otherwise it may lead to screen corruption for some scenarios.
> > > > 
> > > > v2: Rebased to new CDLCK code framework
> > > > 
> > > > Signed-off-by: Madhav Chauhan <madhav.chauhan at intel.com>
> > > > ---
> > > >  drivers/gpu/drm/i915/intel_cdclk.c | 4 ++--
> > > >  1 file changed, 2 insertions(+), 2 deletions(-)
> > > > 
> > > > diff --git a/drivers/gpu/drm/i915/intel_cdclk.c
> > > > b/drivers/gpu/drm/i915/intel_cdclk.c
> > > > index d643c0c..834df68 100644
> > > > --- a/drivers/gpu/drm/i915/intel_cdclk.c
> > > > +++ b/drivers/gpu/drm/i915/intel_cdclk.c
> > > > @@ -1071,9 +1071,9 @@ static int bxt_calc_cdclk(int max_pixclk)
> > > > 
> > > >  static int glk_calc_cdclk(int max_pixclk)  {
> > > > -	if (max_pixclk > 2 * 158400)
> > > > +	if (max_pixclk > DIV_ROUND_UP(2 * 158400 * 99, 100))
> > > 
> > > Where do we ensure we don't use pixel clock 312841..316800? Clearly we
> > > shouldn't use that because we can't guarantee it works, right?
> > 
> > Why do we need to ensure that ?? Can you please elaborate more on this?  
> > Here we are finding one of  the defined CDCLK value for a pixel clock
> 
> I probably had some great idea a month ago when I wrote that, but I can
> no longer remember what it was. :(

I'm not sure if that is what you meant, but if the hardware can't handle it,
intel_compute_max_dotclk() needs to take the 99% limitation into account too.
I.e., max dot clock would be .99 * 2 *  316800 = 627264.

Ander

> 
> BR,
> Jani.
> 
> 
> > > 
> > > Before we get the spec update to confirm what to do, I think we need a
> > > comment here explaining what's going on.
> > 
> > Will add the following comment, if that's fine, will send the rebased patch:
> > "For GLK platform, only 99% of the defined CDCLK value can be achieved 
> >   So calculate pixel clock on that basis"
> > 
> > Regards,
> > Madhav
> > > 
> > > BR,
> > > Jani.
> > > 
> > > >  		return 316800;
> > > > -	else if (max_pixclk > 2 * 79200)
> > > > +	else if (max_pixclk > DIV_ROUND_UP(2 * 79200 * 99, 100))
> > > >  		return 158400;
> > > >  	else
> > > >  		return 79200;
> > > 
> > > --
> > > Jani Nikula, Intel Open Source Technology Center
> 
>