[Piglit] [PATCH 1/2] arb_shader_atomic_counters: fix gs support check

[Ilia Mirkin]

On Thu, Jun 19, 2014 at 6:45 AM, Francisco Jerez <currojerez at riseup.net> wrote:
> Ilia Mirkin <imirkin at alum.mit.edu> writes:
>
>> ---
>>
>> Not actually an issue on the blob, since it supports ARB_geometry_shader4, but
>> Chris Forbes pointed it out. This also works.
>>
>>  tests/spec/arb_shader_atomic_counters/common.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/tests/spec/arb_shader_atomic_counters/common.c b/tests/spec/arb_shader_atomic_counters/common.c
>> index 7547d5d..95d809e 100644
>> --- a/tests/spec/arb_shader_atomic_counters/common.c
>> +++ b/tests/spec/arb_shader_atomic_counters/common.c
>> @@ -211,7 +211,7 @@ atomic_counters_supported(GLenum shader_stage)
>>                  return n;
>>
>>          case GL_GEOMETRY_SHADER:
>> -                if (piglit_is_extension_supported("GL_ARB_geometry_shader4"))
>> +                if (piglit_get_gl_version() >= 32)
>
> Wouldn't this do the wrong thing on e.g. an implementation of GL3.1 that
> exposes ARB_geometry_shader4?

Not really. The rest of the code expects the GL3.2 GS, not the
ARB_geometry_shader4 one. For example, I moved to using #version 150,
and there are no ARB_geometry_shader4 enables. This seemed like the
right move since mesa isn't going to get ARB_gs4 anytime soon, and all
reasonable impls that have atomic counters would also have GL3.2.

Such a hypothetical implementation could be supported by conditionally
using version 140/ARB_gs4. When one finds such an implementation
(GL3.1 only, ARB_gs4, AND atomic counters), and wants to run piglit
against it, they'll have to fix up the test.

  -ilia