On Fri, Aug 04, 2023 at 09:42:34AM +0530, Nautiyal, Ankit K wrote:
> 
> On 8/2/2023 5:35 PM, Lisovskiy, Stanislav wrote:
> > On Fri, Jul 28, 2023 at 09:41:40AM +0530, Ankit Nautiyal wrote:
> > > Separate out functions for getting maximum and minimum input BPC based
> > > on platforms, when DSC is used.
> > > 
> > > Signed-off-by: Ankit Nautiyal <ankit.k.nauti...@intel.com>
> > > ---
> > >   drivers/gpu/drm/i915/display/intel_dp.c | 38 +++++++++++++++++++------
> > >   1 file changed, 30 insertions(+), 8 deletions(-)
> > > 
> > > diff --git a/drivers/gpu/drm/i915/display/intel_dp.c 
> > > b/drivers/gpu/drm/i915/display/intel_dp.c
> > > index 7ec8a478e000..f41de126a8d3 100644
> > > --- a/drivers/gpu/drm/i915/display/intel_dp.c
> > > +++ b/drivers/gpu/drm/i915/display/intel_dp.c
> > > @@ -1535,6 +1535,18 @@ intel_dp_compute_link_config_wide(struct intel_dp 
> > > *intel_dp,
> > >           return -EINVAL;
> > >   }
> > > +static
> > > +u8 intel_dp_dsc_max_src_input_bpc(struct drm_i915_private *i915)
> > > +{
> > > + /* Max DSC Input BPC for ICL is 10 and for TGL+ is 12 */
> > > + if (DISPLAY_VER(i915) >= 12)
> > > +         return 12;
> > > + if (DISPLAY_VER(i915) == 11)
> > > +         return 10;
> > > +
> > > + return 0;
> > > +}
> > > +
> > >   int intel_dp_dsc_compute_bpp(struct intel_dp *intel_dp, u8 max_req_bpc)
> > >   {
> > >           struct drm_i915_private *i915 = dp_to_i915(intel_dp);
> > > @@ -1542,11 +1554,12 @@ int intel_dp_dsc_compute_bpp(struct intel_dp 
> > > *intel_dp, u8 max_req_bpc)
> > >           u8 dsc_bpc[3] = {0};
> > >           u8 dsc_max_bpc;
> > > - /* Max DSC Input BPC for ICL is 10 and for TGL+ is 12 */
> > > - if (DISPLAY_VER(i915) >= 12)
> > > -         dsc_max_bpc = min_t(u8, 12, max_req_bpc);
> > > - else
> > > -         dsc_max_bpc = min_t(u8, 10, max_req_bpc);
> > > + dsc_max_bpc = intel_dp_dsc_max_src_input_bpc(i915);
> > > +
> > > + if (!dsc_max_bpc)
> > > +         return dsc_max_bpc;
> > > +
> > > + dsc_max_bpc = min_t(u8, dsc_max_bpc, max_req_bpc);
> > >           num_bpc = 
> > > drm_dp_dsc_sink_supported_input_bpcs(intel_dp->dsc_dpcd,
> > >                                                          dsc_bpc);
> > > @@ -1674,6 +1687,16 @@ static bool intel_dp_dsc_supports_format(struct 
> > > intel_dp *intel_dp,
> > >           return drm_dp_dsc_sink_supports_format(intel_dp->dsc_dpcd, 
> > > sink_dsc_format);
> > >   }
> > > +static
> > > +u8 intel_dp_dsc_min_src_input_bpc(struct drm_i915_private *i915)
> > > +{
> > > + /* Min DSC Input BPC for ICL+ is 8 */
> > > + if (DISPLAY_VER(i915) >= 11)
> > > +         return 8;
> > > +
> > > + return 0;
> > So does it mean that for anything below gen 11, there is no limit at all?
> > Also it means that the condition below will never be executed for gen <= 11.
> 
> Hmm. Bspec says min bpc is 8 for DSC, so idea is to have min bpc as 8 when
> DSC is used.
> 
> This is supposed to be called only if DSC is supported, so perhaps HAS_DSC
> can be used?
> 
> return HAS_DSC(dev_priv) ? 8 : 0;
> 
> 
> Regards,
> 
> Ankit

Yeah, I think that might be better, since we actually car about if we support 
DSC
or not. HAS_DSC should do all the magic, to determine if we support it or not..

Stan

> 
> > 
> > Stan
> > 
> > > +}
> > > +
> > >   int intel_dp_dsc_compute_config(struct intel_dp *intel_dp,
> > >                                   struct intel_crtc_state *pipe_config,
> > >                                   struct drm_connector_state *conn_state,
> > > @@ -1707,10 +1730,9 @@ int intel_dp_dsc_compute_config(struct intel_dp 
> > > *intel_dp,
> > >                   pipe_bpp = pipe_config->pipe_bpp;
> > >           }
> > > - /* Min Input BPC for ICL+ is 8 */
> > > - if (pipe_bpp < 8 * 3) {
> > > + if (pipe_bpp < intel_dp_dsc_min_src_input_bpc(dev_priv) * 3) {
> > >                   drm_dbg_kms(&dev_priv->drm,
> > > -                     "No DSC support for less than 8bpc\n");
> > > +                     "Computed BPC less than min supported by source for 
> > > DSC\n");
> > >                   return -EINVAL;
> > >           }
> > > -- 
> > > 2.40.1
> > > 

Reply via email to