On Tue, Dec 18, 2012 at 2:43 PM, Brian Paul wrote:
> create mode 100644 tests/spec/arb_es2_compatibility/ge7.frag
> create mode 100644 tests/spec/arb_es2_compatibility/ge7.vert
[snip]
> + * The vertex/fragment shaders are captured from Google Earth v7.
> + * Simpler shaders don't always trigger
On 01/07/2013 06:22 PM, Paul Berry wrote:
This patch introduces new vs and fs tests to provoke a bug in Mesa
which caused loop bounds to be inferred incorrectly when the loop
condition was expressed in an unusual way (e.g. "4 > i" instead of "i
< 4").
---
.../execution/fs-loop-bounds-unrolled.s
On Mon, Jan 7, 2013 at 6:22 PM, Paul Berry wrote:
> This patch introduces new vs and fs tests to provoke a bug in Mesa
> which caused loop bounds to be inferred incorrectly when the loop
> condition was expressed in an unusual way (e.g. "4 > i" instead of "i
> < 4").
> ---
> .../execution/fs-loop
This patch introduces new vs and fs tests to provoke a bug in Mesa
which caused loop bounds to be inferred incorrectly when the loop
condition was expressed in an unusual way (e.g. "4 > i" instead of "i
< 4").
---
.../execution/fs-loop-bounds-unrolled.shader_test | 40 +
.../e
On Mon, Jan 7, 2013 at 1:21 PM, Chad Versace
wrote:
> On 12/18/2012 02:43 PM, Brian Paul wrote:
>> This tests exercises a Mesa crash found with Google Earth v7.
>> Specifically, Google Earth is calling glReleaseShaderCompiler() between
>> glCompileShader() and glLinkProgram().
>>
>> It appears tha
On 5 January 2013 01:07, Chris Forbes wrote:
> These cover most of the behavior specified by ARB_texture_multisample.
>
> ___
> Piglit mailing list
> Piglit@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/piglit
>
This series look
On 7 January 2013 14:21, Chris Forbes wrote:
> This test actually gets *simpler*, generalized to arbitrary sample
> counts. I'll cook something up tonight.
>
Ok, cool. While you're looking into that, there are several other ways I
think this test needs to be generalized:
- We need to test a pi
This test actually gets *simpler*, generalized to arbitrary sample
counts. I'll cook something up tonight.
-- Chris
On Tue, Jan 8, 2013 at 10:24 AM, Paul Berry wrote:
> On 5 January 2013 01:07, Chris Forbes wrote:
>>
>> Writes a test pattern to individual samples in a multisample texture,
>> an
On 5 January 2013 01:07, Chris Forbes wrote:
> RGBA32I, RGBA32UI, and RGBA32F are huge and
> unsupported for multisample textures on at least Gen6 i965.
>
> This shouldn't be committed as-is, but it allowed for testing of
> multisample samplers with the generic textureSize() test.
>
> Signed-off-
On 01/07/2013 01:04 PM, Tom Gall wrote:
> I had posted this patch back on December 28th to fix shader_runner to
> work for gles2.
>
> http://lists.freedesktop.org/archives/piglit/2012-December/004391.html
> If there's no feedback or issues, could it be merged please?
>
> Thanks!
Hey Tom,
I just
On 12/28/2012 02:26 PM, Tom Gall wrote:
> Fix shader_test.py so it uses shader_runner_gles2 when running
> a shader test with GL 2.00 es in the [required] section.
^^
It should be "GL 2.0 es". Though, shader_runner accepts "GL 2.00 es"
because it's sloppy.
Otherwise, t
On 5 January 2013 01:07, Chris Forbes wrote:
> Writes a test pattern to individual samples in a multisample texture,
> and tests texelFetch() reads the correct values for each sample.
>
> Signed-off-by: Chris Forbes
>
Unforunately, I think we need to generalize this test to test a variety of
sa
I do indeed.
On Tue, Jan 8, 2013 at 10:11 AM, Paul Berry wrote:
> On 5 January 2013 01:07, Chris Forbes wrote:
>>
>> This tests that the new teximage state added in ARB_texture_multisample
>> exists and has correct defaults for non-multisample textures:
>>
>> - GL_TEXTURE_SAMPLES = 0
>>
On 5 January 2013 01:07, Chris Forbes wrote:
> This tests that the new teximage state added in ARB_texture_multisample
> exists and has correct defaults for non-multisample textures:
>
> - GL_TEXTURE_SAMPLES = 0
> - GL_TEXTURE_FIXED_SAMPLE_LOCATIONS = false
>
I think you mean "true" here
On 5 January 2013 01:07, Chris Forbes wrote:
> This tests FBO setup with various combinations of multisample textures
> and `classic` multisample renderbuffers, and for each, checks:
>
> - That the renderbuffers or textures can be created
> - Completeness status
>
> If the configuration is expect
I had posted this patch back on December 28th to fix shader_runner to
work for gles2.
http://lists.freedesktop.org/archives/piglit/2012-December/004391.html
If there's no feedback or issues, could it be merged please?
Thanks!
--
Regards,
Tom
"Where's the kaboom!? There was supposed to be an ear
On 12/18/2012 02:43 PM, Brian Paul wrote:
> This tests exercises a Mesa crash found with Google Earth v7.
> Specifically, Google Earth is calling glReleaseShaderCompiler() between
> glCompileShader() and glLinkProgram().
>
> It appears that glReleaseShaderCompiler() is freeing some data
> structur
On Tue, Jan 8, 2013 at 8:43 AM, Paul Berry wrote:
> Window height/width less than 116 pixels cause problems when Piglit is run
> on windows (because Windows forcibly resizes the window, causing the Piglit
> framework to report the test result as "WARN" instead of "PASS"). Since
> you're not actua
On 5 January 2013 01:07, Chris Forbes wrote:
> Verifies new minimum maximums in ARB_texture_multisample:
>
> GL_MAX_COLOR_TEXTURE_SAMPLES >= 1
> GL_MAX_DEPTH_TEXTURE_SAMPLES >= 1
> GL_MAX_INTEGER_SAMPLES >= 1
> GL_MAX_SAMPLE_MASK_WORDS >= 1
>
> Signed-off-by: Chris Forbes
> ---
> tests/all.test
19 matches
Mail list logo
