[Piglit] [PATCH] piglit-util-gl-common: fix ES1 / ES2 build error

[Brian Paul]

On 10/10/2012 12:27 PM, Chad Versace wrote:
> On 10/08/2012 06:23 PM, Brian Paul wrote:
>> On Mon, Oct 8, 2012 at 6:17 PM, Jordan Justen<jordan.l.justen at intel.com>  wrote:
>>> Signed-off-by: Jordan Justen<jordan.l.justen at intel.com>
>>> Cc: Brian Paul<brianp at vmware.com>
>>> Cc: Chad Versace<chad.versace at linux.intel.com>
>>> ---
>>>   tests/util/piglit-util-gl-common.c |    7 +++++++
>>>   1 file changed, 7 insertions(+)
>>>
>>> diff --git a/tests/util/piglit-util-gl-common.c b/tests/util/piglit-util-gl-common.c
>>> index 62b5312..4f9fe5f 100644
>>> --- a/tests/util/piglit-util-gl-common.c
>>> +++ b/tests/util/piglit-util-gl-common.c
>>> @@ -447,8 +447,14 @@ piglit_get_compressed_block_size(GLenum format,
>>>                                   unsigned *bw, unsigned *bh, unsigned *bytes)
>>>   {
>>>          switch (format) {
>>> +#if defined(USE_OPENGL) || defined(USE_OPENGL_ES2)
>>>          case GL_COMPRESSED_RGB_S3TC_DXT1_EXT:
>>>          case GL_COMPRESSED_RGBA_S3TC_DXT1_EXT:
>>> +               *bw = *bh = 4;
>>> +               *bytes = 8;
>>> +               return true;
>>> +#endif
>>> +#if defined(USE_OPENGL)
>>>          case GL_COMPRESSED_SRGB_S3TC_DXT1_EXT:
>>>          case GL_COMPRESSED_SRGB_ALPHA_S3TC_DXT1_EXT:
>>>          case GL_COMPRESSED_RED_RGTC1:
>>> @@ -475,6 +481,7 @@ piglit_get_compressed_block_size(GLenum format,
>>>                  *bh = 4;
>>>                  *bytes = 16;
>>>                  return true;
>>> +#endif
>>>          default:
>>>                  /* return something rather than uninitialized values */
>>>                  *bw = *bh = *bytes = 1;
>>
>> Actually, it would probably be better to test the extension #define,
>> such as GL_EXT_texture_compression_s3tc because off-hand I don't know
>> which compressed formats are supported by which APIs.
>
>
> Brian,
>
> The GL headers on my system don't define the GL_EXT_texture_compression_s3tc
> feature macro.
>
> Since the Piglit build is currently broken, this patch fixes it, and the
> suggested alternative doesn't work on my system, I've gone ahead and committed
> the patch. If we later find a better method to filter the enums in this switch,
> we can apply the better method then.

That's fine but I think I'm missing something.

Aren't all the GL_(s3tc)_EXT tokens coming from the 
generated_dispatch.h file?  It looks like that header also has a whole 
mess of "#define GL_EXT_foo_bar" lines.  Shouldn't there be one for 
GL_EXT_texture_compression_s3tc also?

Wouldn't everyone have the s3tc-related tokens defined in 
generated_dispatch.h like I do?

AFAICT, generated_dispatch.[ch] are generated from the glapi.json 
file.  And glapi.json is generated from enumext.spec and that file 
defines the extension tokens above.  So I don't see where the 
variation in #defined tokens is coming from.

-Brian