[Mesa-dev] List of unsupported extensions per driver
Ian Romanick
idr at freedesktop.org
Wed Sep 30 18:33:42 PDT 2015
On 09/29/2015 07:48 AM, Romain Failliot wrote:
> What I don't understand is that all the lines starting with a "-" seems
> to be part of the GL_ARB_gpu_shader5 extension. See the line here:
> http://cgit.freedesktop.org/mesa/mesa/tree/docs/GL3.txt#n99
The two lines that just have () are only supported on i965, nvc0, and
radeonsi. It's not really that clear.
GL_ARB_gpu_shader5 DONE (i965)
- 'precise' qualifier DONE
- Dynamically uniform sampler array indices DONE (r600, softpipe)
- Dynamically uniform UBO array indices DONE (r600)
- Implicit signed -> unsigned conversions DONE
- Fused multiply-add DONE ()
- Packing/bitfield/conversion functions DONE (r600, softpipe)
- Enhanced textureGather DONE (r600, softpipe)
- Geometry shader instancing DONE (r600, llvmpipe, softpipe)
- Geometry shader multiple streams DONE ()
- Enhanced per-sample shading DONE (r600)
- Interpolation functions DONE (r600)
- New overload resolution rules DONE
>
> If I'm right, it means that, considering Ilia's web site,
> GL_ARB_gpu_shader5 is unsupported by R600, but everything in its sublist
> is supported. You see why it is confusing?
>
> Le 29 sept. 2015 8:06 AM, "Marek Olšák" <maraeo at gmail.com
> <mailto:maraeo at gmail.com>> a écrit :
>
> FMA isn't required really. R600 is mainly missing GS streams, which
> are complete on the mailing somewhere I think.
>
> Marek
>
> On Tue, Sep 29, 2015 at 7:32 AM, Romain Failliot
> <romain.failliot at foolstep.com <mailto:romain.failliot at foolstep.com>>
> wrote:
> > Hi!
> >
> > I'm diving into the unsupported extensions list and I'm wondering
> how is it
> > possible that GL_ARB_gpu_shader5 is unsupported for R600, but some
> of the
> > "sub-extensions" like "Dynamically uniform sampler array indices" are
> > supported nonetheless.
> >
> > That makes me wonder if "not done" sub-extensions, like "Fused
> > multiply-add", are really not done for R600 yet, or if they are indeed
> > unsupported as the parent extension status let us suppose.
> >
> > Thx
> > Romain
> >
> > 2015-07-31 23:42 GMT+02:00 Ilia Mirkin <imirkin at alum.mit.edu
> <mailto:imirkin at alum.mit.edu>>:
> >>
> >> OK, I believe I've fixed my list up. Note that you may have to
> >> shift-reload to get the updates, I think fd.o isn't setting the
> proper
> >> cache headers or something else is messed up.
> >
> >
>
>
>
> _______________________________________________
> mesa-dev mailing list
> mesa-dev at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/mesa-dev
>
More information about the mesa-dev
mailing list