On Tue, Dec 03, 2019 at 03:11:54PM +0200, Ville Syrjälä wrote:
> On Tue, Dec 03, 2019 at 11:08:52AM +0200, Jani Nikula wrote:
> > On Mon, 02 Dec 2019, José Roberto de Souza <jose.souza@xxxxxxxxx> wrote:
> > > This is better than keep those values in the code that you always
> > > need to check the DP spec to know what level of HBR it is.
> > >
> > > Signed-off-by: José Roberto de Souza <jose.souza@xxxxxxxxx>
> > > ---
> > > drivers/gpu/drm/i915/display/intel_ddi.c | 6 +++++-
> > > 1 file changed, 5 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/drivers/gpu/drm/i915/display/intel_ddi.c b/drivers/gpu/drm/i915/display/intel_ddi.c
> > > index a976606d21c7..914f0cc4d237 100644
> > > --- a/drivers/gpu/drm/i915/display/intel_ddi.c
> > > +++ b/drivers/gpu/drm/i915/display/intel_ddi.c
> > > @@ -49,6 +49,10 @@
> > > #include "intel_tc.h"
> > > #include "intel_vdsc.h"
> > >
> > > +#define HBR_RATE 270000
> > > +#define HBR2_RATE 540000
> > > +#define HBR3_RATE 810000
> > > +
> > > struct ddi_buf_trans {
> > > u32 trans1; /* balance leg enable, de-emph level */
> > > u32 trans2; /* vref sel, vswing */
> > > @@ -888,7 +892,7 @@ icl_get_combo_buf_trans(struct drm_i915_private *dev_priv, int type, int rate,
> > > if (type == INTEL_OUTPUT_HDMI) {
> > > *n_entries = ARRAY_SIZE(icl_combo_phy_ddi_translations_hdmi);
> > > return icl_combo_phy_ddi_translations_hdmi;
> > > - } else if (rate > 540000 && type == INTEL_OUTPUT_EDP) {
> > > + } else if (rate > HBR2_RATE && type == INTEL_OUTPUT_EDP) {
> >
> > I don't want a patch switching some random place to using a
> > macro. Either we stick to numbers or switch all.
> >
> > And if switch all, add the rates to drm core, not locally to
> > intel_ddi.c. (And then wonder what to do with the intermediate rates in
> > intel_dp_set_source_rates()...)
>
> Yeah, we'll still end up with a mix of defines and raw numbers.
>
> >
> > Personally, HBR<N> is less useful to me in code, it's the actual rate
> > that helps me.
> >
> > But I'll trust Ville's judgement on this one.
>
> I tend to prefer raw numbers for this sort of stuff. If we didn't have
> the intermediate rates I might have a different opinion. The only thing
> I really worry about with raw numbers is the potential for typos.
Yes, especially due to the typos and possibilities of missing or adding an extra 0
makes me wonder that it could be a good idea to add all the #defines RBR - HBR3
in drm_dp_helper.h somewhere with proper comments on which of the spec added which rate etc.
Regards
Manasi
>
> The original problem of bspec talking about hbr2 in the bug trans
> tables we could probably solve with a comment.
>
> --
> Ville Syrjälä
> Intel
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
> https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
