Well, with Roland's fix to the draw module, llvmpipe works with my
tests as well. Is it likely that r600 will work as-is, or do you think
it'll require some fixes? I can try to find someone with the requisite
hw...
On Mon, Jun 23, 2014 at 4:57 PM, Marek Olšák <mar...@gmail.com> wrote:
> The cap can be removed later if needed.
>
> Marek
>
> On Mon, Jun 23, 2014 at 2:50 PM, Ilia Mirkin <imir...@alum.mit.edu> wrote:
>> On Mon, Jun 23, 2014 at 7:55 AM, Roland Scheidegger <srol...@vmware.com>
>> wrote:
>>> Am 23.06.2014 12:49, schrieb Roland Scheidegger:
>>>> Am 22.06.2014 17:10, schrieb Ilia Mirkin:
>>>>> Signed-off-by: Ilia Mirkin <imir...@alum.mit.edu>
>>>>> ---
>>>>> docs/GL3.txt | 2 +-
>>>>> docs/relnotes/10.3.html | 2 +-
>>>>> src/gallium/drivers/nouveau/nvc0/nvc0_program.c | 2 ++
>>>>> src/gallium/drivers/nouveau/nvc0/nvc0_screen.c | 2 +-
>>>>> 4 files changed, 5 insertions(+), 3 deletions(-)
>>>>>
>>>>> diff --git a/docs/GL3.txt b/docs/GL3.txt
>>>>> index c47a54b..8cfa247 100644
>>>>> --- a/docs/GL3.txt
>>>>> +++ b/docs/GL3.txt
>>>>> @@ -161,7 +161,7 @@ GL 4.3:
>>>>> GL_ARB_copy_image not started
>>>>> GL_KHR_debug DONE (all drivers)
>>>>> GL_ARB_explicit_uniform_location DONE (all drivers
>>>>> that support GLSL)
>>>>> - GL_ARB_fragment_layer_viewport DONE (nv50)
>>>>> + GL_ARB_fragment_layer_viewport DONE (nv50, nvc0)
>>>>> GL_ARB_framebuffer_no_attachments not started
>>>>> GL_ARB_internalformat_query2 not started
>>>>> GL_ARB_invalidate_subdata DONE (all drivers)
>>>>> diff --git a/docs/relnotes/10.3.html b/docs/relnotes/10.3.html
>>>>> index 5e29f2d..b65e347 100644
>>>>> --- a/docs/relnotes/10.3.html
>>>>> +++ b/docs/relnotes/10.3.html
>>>>> @@ -53,7 +53,7 @@ Note: some of the new features are only available with
>>>>> certain drivers.
>>>>> <li>GL_ARB_texture_query_levels on nv50, nvc0, llvmpipe, r600, radeonsi,
>>>>> softpipe</li>
>>>>> <li>GL_ARB_texture_query_lod on radeonsi</li>
>>>>> <li>GL_ARB_viewport_array on nvc0</li>
>>>>> -<li>GL_ARB_fragment_layer_viewport on nv50</li>
>>>>> +<li>GL_ARB_fragment_layer_viewport on nv50, nvc0</li>
>>>>> </ul>
>>>>>
>>>>>
>>>>> diff --git a/src/gallium/drivers/nouveau/nvc0/nvc0_program.c
>>>>> b/src/gallium/drivers/nouveau/nvc0/nvc0_program.c
>>>>> index 667fbc8..ad8d1b0 100644
>>>>> --- a/src/gallium/drivers/nouveau/nvc0/nvc0_program.c
>>>>> +++ b/src/gallium/drivers/nouveau/nvc0/nvc0_program.c
>>>>> @@ -36,6 +36,8 @@ nvc0_shader_input_address(unsigned sn, unsigned si,
>>>>> unsigned ubase)
>>>>> switch (sn) {
>>>>> case NV50_SEMANTIC_TESSFACTOR: return 0x000 + si * 0x4;
>>>>> case TGSI_SEMANTIC_PRIMID: return 0x060;
>>>>> + case TGSI_SEMANTIC_LAYER: return 0x064;
>>>>> + case TGSI_SEMANTIC_VIEWPORT_INDEX:return 0x068;
>>>>> case TGSI_SEMANTIC_PSIZE: return 0x06c;
>>>>> case TGSI_SEMANTIC_POSITION: return 0x070;
>>>>> case TGSI_SEMANTIC_GENERIC: return ubase + si * 0x10;
>>>>> diff --git a/src/gallium/drivers/nouveau/nvc0/nvc0_screen.c
>>>>> b/src/gallium/drivers/nouveau/nvc0/nvc0_screen.c
>>>>> index 2c8f1ca..c02e10a 100644
>>>>> --- a/src/gallium/drivers/nouveau/nvc0/nvc0_screen.c
>>>>> +++ b/src/gallium/drivers/nouveau/nvc0/nvc0_screen.c
>>>>> @@ -164,6 +164,7 @@ nvc0_screen_get_param(struct pipe_screen *pscreen,
>>>>> enum pipe_cap param)
>>>>> case PIPE_CAP_SAMPLE_SHADING:
>>>>> case PIPE_CAP_TEXTURE_GATHER_OFFSETS:
>>>>> case PIPE_CAP_TEXTURE_GATHER_SM5:
>>>>> + case PIPE_CAP_TGSI_FS_LAYER_VIEWPORT:
>>>>> return 1;
>>>>> case PIPE_CAP_SEAMLESS_CUBE_MAP_PER_TEXTURE:
>>>>> return (class_3d >= NVE4_3D_CLASS) ? 1 : 0;
>>>>> @@ -181,7 +182,6 @@ nvc0_screen_get_param(struct pipe_screen *pscreen,
>>>>> enum pipe_cap param)
>>>>> case PIPE_CAP_TGSI_VS_LAYER:
>>>>> case PIPE_CAP_FAKE_SW_MSAA:
>>>>> case PIPE_CAP_TGSI_VS_WINDOW_SPACE_POSITION:
>>>>> - case PIPE_CAP_TGSI_FS_LAYER_VIEWPORT:
>>>>> return 0;
>>>>> }
>>>>>
>>>>>
>>>>
>>>> Is that cap bit really required? Everybody who can do layered viewports
>>>> ought to be able to do this (as it's required by d3d10). Though I can't
>>>> say if right now all drivers which support layered viewports can already
>>>> do it (llvmpipe can), but it might be trivial to fix those up which can't.
>>>>
>>>
>>> (Of course that should have been layered fbs and multiple viewports,
>>> that short-notation I used there doesn't make much sense...)
>>
>> Well, I think right now the only drivers that support multiple
>> viewports are llvmpipe, nv50, nvc0, and r600. I can test out llvmpipe
>> but would need some help for r600. Would love not to have to add the
>> cap :)
>>
>> -ilia
>> _______________________________________________
>> mesa-dev mailing list
>> mesa-dev@lists.freedesktop.org
>> http://lists.freedesktop.org/mailman/listinfo/mesa-dev
_______________________________________________
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
