Another Fun Day In Mesa Git: RADV, NVC0, RadeonSI

[pal666]

Really?
GL 4.4, GLSL 4.40 -- all DONE: i965/gen8+, nvc0, radeonsi

really, this line means "all lines below are done"

[pal666]

Question: What would (legally) happen if someone would just change the supported OpenGL version to 4.5 for radeonsi without contacting Khronos or going through with these conformance tests?

my guess is trademark infringement

[pal666]

anyway this version number anxiety is irrational, since reported can be bumped by setting one environment variable, the only thing what matters is real feature implementation status

[chrisr]

Yes, notice the "all DONE" part at the top.

mesamatrix.net isn't showing radeonsi as green, maybe a bug on the site?

More likely, because features.txt still contains:

GL_ARB_enhanced_layouts DONE (i965, nv50, llvmpipe, softpipe)
...
- specified vec4-slot component numbers               DONE (i965, nv50, llvmpipe, softpipe)

instead of

GL_ARB_enhanced_layouts DONE

[FLHerne]

your post has nothing to do with reality, since radeonsi is developed by amd and radhat, both of which has access to cts. the only problem is someone has to fix all test failures ffs

If the entire Mesa development community could reproduce and discuss CTS failures, there'd be a lot more 'someones' and it would go quicker. A lot of bugs are in the shared layers of Mesa; many people who work on that code aren't Intel/AMD/RH employees.

Not to mention, I don't see AMD or Intel sponsoring Nouveau any time soon. RedHat maybe, but do they care about Freedreno when that gets merged?

Khronos clearly benefit from having an open, highly-conformant implementation (even the much-loved Nvidia blob fails or crashes in many Piglit tests); granting the Mesa team membership in their own right would be a good way to repay that.

[karolherbst]

my guess is trademark infringement

exactly

[karolherbst]

More likely, because features.txt still contains:

GL_ARB_enhanced_layouts DONE (i965, nv50, llvmpipe, softpipe)
...
- specified vec4-slot component numbers DONE (i965, nv50, llvmpipe, softpipe)

instead of

GL_ARB_enhanced_layouts DONE

nv50 != nvc0

[chrisr]

nv50 != nvc0

Whatever. I am simply pointing out why the OpenGL Mesa Matrix still displays GL_ARB_enhanced_layouts as incomplete for radeonsi.

[karolherbst]

Whatever. I am simply pointing out why the OpenGL Mesa Matrix still displays GL_ARB_enhanced_layouts as incomplete for radeonsi.

well but your argument makes no sense in this, because it should detect this as done.

[pal666]

If the entire Mesa development community could reproduce and discuss CTS failures, there'd be a lot more 'someones' and it would go quicker.

lol, did you try to do some fact-checking before posts? entire mesa community around radeonsi is more than 90% amd + redhat. and no community likes to fix cts failures

A lot of bugs are in the shared layers of Mesa; many people who work on that code aren't Intel/AMD/RH employees.

you are certainly from some other planet. those two people a busy making nouveau barely usable

Not to mention, I don't see AMD or Intel sponsoring Nouveau any time soon. RedHat maybe, but do they care about Freedreno when that gets merged?

redhat employs developers of nouveau and freedreno. what planet are you from again?