On Fri, Mar 15, 2024 at 01:56:19PM +0200, Ville Syrjälä wrote:
> On Fri, Mar 15, 2024 at 12:33:52PM +0100, Sebastian Wick wrote:
> > On Fri, Mar 15, 2024 at 10:35:30AM +0530, Suraj Kandpal wrote:
> <snip>
> > > @@ -318,11 +346,24 @@ intel_dp_aux_hdr_setup_backlight(struct intel_connector *connector, enum pipe pi
> > > panel->backlight.min = 0;
> > > }
> > >
> > > + if (DISPLAY_VER(i915) < 11) {
> >
> > This should check for INTEL_EDP_HDR_TCON_SDP_COLORIMETRY_CAP instead!
> > There is no reason to bind this to any hardware version if the hardware
> > itself can tell you if it supports SDP signalling or needs to set it via
> > AUX.
>
> That should be OK for icl+.
>
> To extend this to pre-icl hardware I think we pretty much need both checks:
> - has_gamut_metadata_dip() -> can we actually transmit the SDP?
> - INTEL_EDP_HDR_TCON_SDP_COLORIMETRY_CAP -> can the tcon extract the data from the SDP?
> and if either is false then we fall back to AUX.
Right, that's something my patch is missing as well.
> We should also change intel_dp_add_properties() to check
> the tcon caps instead of relying on has_gamut_metadata_dip(),
> for eDP.
>
> --
> Ville Syrjälä
> Intel
>
