On 01.02.2010 21:44, Brian Paul wrote:
> Christoph Bumiller wrote:
>> I just noticed that alienarena fails to create an FBO for depth
>> rendering
>> because on st_validate_framebuffer, the depth attachment contains a pt
>> in the stObj with format XR8G8B8_UNORM, which nv50 (contrary to sp)
>> reports as not supported in screen->is_format_supported if usage is ZS.
>>
>> This happens because the app seems to use a texture ID that has
>> already been used with mipmaps, and so st_TexImage doesn't create
>> a new pipe texture with the different format, because teximage_realloc
>> is false (NULL passed to glTexImage2D).
>>
>> I'm not sure if that is to be regarded as an application bug or if we
>> should
>> dump the other mipmap levels an reallocate nonetheless, so I'm asking
>> here
>> (I guess the nvidia blob seems to permit this behaviour, I already had
>> another issue with this game where it was too forgiving).
>
>
> Christoph, can you try the attached patch to the state tracker?
>
> -Brian
>
Yes, that would fix the issue, since the it's the last_level == level check
that sets the condition false, I figured that out already.

As to Ian's questions:
The app doesn't use GenTextures or DeleteTextures, except when I
close it.

The new texture gets GL_NEAREST as MIN / MAX filters.

Pixels is NULL, as I said.

I don't think I'm going to write a piglit test right now, it's late. Maybe
I'll think about it when I encounter the next issue I want to write a
test for ...

Thanks,
Christoph.

------------------------------------------------------------------------------
The Planet: dedicated and managed hosting, cloud storage, colocation
Stay online with enterprise data centers and the best network in the business
Choose flexible plans and management services without long-term contracts
Personal 24x7 support from experience hosting pros just a phone call away.
http://p.sf.net/sfu/theplanet-com
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

Reply via email to