[Mesa-dev] [PATCH 6/9] anv/gen7_pipeline: Set few missing fields in 3DSTATE_VS

[Jason Ekstrand]

On Sep 30, 2016 3:56 PM, "Jason Ekstrand" <jason at jlekstrand.net> wrote:
>
> Flyby comments...
>
> On Sep 30, 2016 3:34 PM, "Anuj Phogat" <anuj.phogat at gmail.com> wrote:
> >
> > Signed-off-by: Anuj Phogat <anuj.phogat at gmail.com>
> > ---
> >  src/intel/vulkan/gen7_pipeline.c | 6 ++++++
> >  1 file changed, 6 insertions(+)
> >
> > diff --git a/src/intel/vulkan/gen7_pipeline.c
b/src/intel/vulkan/gen7_pipeline.c
> > index 8d133b6..f98aef5 100644
> > --- a/src/intel/vulkan/gen7_pipeline.c
> > +++ b/src/intel/vulkan/gen7_pipeline.c
> > @@ -121,6 +121,12 @@ genX(graphics_pipeline_create)(
> >           };
> >           vs.PerThreadScratchSpace      =
scratch_space(&vs_prog_data->base.base);
> >
> > +         vs.FloatingPointMode          =
vs_prog_data->base.base.use_alt_mode ?
> > +                                         Alternate : IEEE754;
>
> I'm pretty sure we only use alt mode for ARB programs do it doesn't
really apply to Vulkan.  Explicitly setting it to IEEE seems reasonable
though.
>
> > +         vs.SamplerCount               = 0;
> > +         vs.BindingTableEntryCount     =
> > +            vs_prog_data->base.base.binding_table.size_bytes / 4;
>
> We don't set the binding table size field in the Vulkan driver so I don't
think this actually does anything.  Maybe we should actually set sampler
and surface state counts to something reasonable?  It might have a
performance impact but I don't know if we've ever measured it.

Thinking a bit more, I think we can get both the sampler count and the
binding table entry count from the anv_bind_map data structure associated
with the stage.  We should either use that directly or plumb it through
prog_data somewhere. (I don't believe we fill of those prog_data fields
today.)

> > +
> >           vs.DispatchGRFStartRegisterforURBData    =
> >              vs_prog_data->base.base.dispatch_grf_start_reg;
> >
> > --
> > 2.5.5
> >
> > _______________________________________________
> > mesa-dev mailing list
> > mesa-dev at lists.freedesktop.org
> > https://lists.freedesktop.org/mailman/listinfo/mesa-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/mesa-dev/attachments/20160930/e9515118/attachment.html>