[Mesa-dev] [PATCH] mesa: support GL_RGB for GL_EXT_texture_type_2_10_10_10_REV

[Tapani]

On 01/14/2015 03:16 PM, Iago Toral Quiroga wrote:
> Hi Tapani,
>
> thanks for looking into this one, I certainly missed that extension...
>
> On mié, 2015-01-14 at 14:04 +0200, Tapani Pälli wrote:
>> Commit 8ec6534 changed texture upload path and the way how texture
>> format is being checked, this commit adds support for GL_RGB with
>> GL_UNSIGNED_INT_2_10_10_10_REV as specified by the extension
>> EXT_texture_type_2_10_10_10_REV specification.
> The extension mentions that when this is used with RGB, the alpha
> channel assumes a value 1.0 when expanded...
>
>> This fixes regression in ES3 conformance test
>>     ES3-CTS.gtf.GL3Tests.packed_pixels.packed_pixels
>>
>> Signed-off-by: Tapani Pälli <tapani.palli at intel.com>
>> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=88385
>> ---
>>   src/mesa/main/glformats.c | 3 ++-
>>   1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/src/mesa/main/glformats.c b/src/mesa/main/glformats.c
>> index 3e913ed..e54bb90 100644
>> --- a/src/mesa/main/glformats.c
>> +++ b/src/mesa/main/glformats.c
>> @@ -2666,7 +2666,8 @@ _mesa_format_from_format_and_type(GLenum format, GLenum type)
>>            return MESA_FORMAT_A2R10G10B10_UINT;
>>         break;
>>      case GL_UNSIGNED_INT_2_10_10_10_REV:
>> -      if (format == GL_RGBA)
>> +      if (format == GL_RGB || /* GL_EXT_texture_type_2_10_10_10_REV */
>> +          format == GL_RGBA)
>>            return MESA_FORMAT_R10G10B10A2_UNORM;
> ...so I think this is not correct: with this we will lose track that
> this was RGB data and we will not ensure A=1.0 if this is stored as
> RGBA.
>
> I think what we want to do here is add a new mesa_format like
> MESA_FORMAT_R10G10B10X2_UNORM, so we can use auto-generated pack/unpack
> functions that take care of this. We would add that enum to formats.h,
> then add the type description to formats.csv and then return that enum
> here for the RGB case.

Ah right, seems fine for me. I just tried to do what was done 
previously, which seemed to just enable GL_RGB usage. I'll check if 
there exists a test for actually using the format and implement this.

> Jason, what do you think?
>
>>         else if (format == GL_RGBA_INTEGER)
>>            return MESA_FORMAT_R10G10B10A2_UINT;
>