On Fri, Sep 9, 2016 at 7:14 PM, Dylan Baker wrote:
> This allows extensions to specify which stages they support, allowing
> such extensions as ARB_shader_viewport_layer_array to be added.
The ext can be enabled anywhere... only has an effect in vs/tes though.
>
>
Signed-off-by: Dylan Baker
---
generated_tests/gen_extensions_defined.py | 1 +
1 file changed, 1 insertion(+)
diff --git a/generated_tests/gen_extensions_defined.py
b/generated_tests/gen_extensions_defined.py
index cbacc12..745eb4a 100644
---
This adds a couple of original compiler tests, and ports the tests for
amd_vertex_shader_layer and amd_vertex_shader_viewport_index to the ARB
extension.
These tests touch the VS, but not the tes eval.
Signed-off-by: Dylan Baker
---
tests/all.py
This allows extensions to specify which stages they support, allowing
such extensions as ARB_shader_viewport_layer_array to be added.
Signed-off-by: Dylan Baker
---
generated_tests/gen_extensions_defined.py | 176 --
1 file changed, 94
Signed-off-by: Dylan Baker
---
generated_tests/gen_extensions_defined.py | 101 +++---
1 file changed, 50 insertions(+), 51 deletions(-)
diff --git a/generated_tests/gen_extensions_defined.py
b/generated_tests/gen_extensions_defined.py
index
From: Marek Olk
---
tests/util/piglit-framework-gl/piglit_wfl_framework.c | 3 ++-
tests/util/piglit-util-gl.c | 2 +-
tests/util/piglit-util-gl.h | 2 +-
3 files changed, 4 insertions(+), 3 deletions(-)
diff --git
This adds a switch to to turn off process isolation (it is normally on),
for now this only affects shader_runner, but in the future could affect
more tests, like glslparsertest or dEQP.
Signed-off-by: Dylan Baker
---
framework/options.py | 1 +
From: Marek Olk
The performance is measured by running these:
spec/glsl-1.30/execution/built-in-functions/ (1420 tests)
Running piglit-run.py using the old way: 98 s
Running .../shader_runner *.shader_test: 8 s
It's more than 12x faster!!
The tests being run must be
This adds fast-skipping support to the MultiShaderTest class. Both
patches seemed significant enough to warrant splitting them in two.
Signed-off-by: Dylan Baker
---
framework/test/base.py| 8 +-
framework/test/shader_test.py | 49
From: Marek Olk
---
tests/util/piglit-util.c | 6 --
tests/util/piglit-util.h | 6 ++
2 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/tests/util/piglit-util.c b/tests/util/piglit-util.c
index 0a66a85..6ed9c9c 100644
--- a/tests/util/piglit-util.c
Bug fixes from me for the above patch.
I've split these out to make them easier to review. Before merging to
master this will be squashed into the previous patch.
Changes:
- Fix the assertion of subuniform_locations
- Add teardown function for UBOs that destroys objects and frees memory
-
From: Marek Olk
---
tests/util/piglit-framework-gl/piglit_glut_framework.c | 7 ---
tests/util/piglit-framework-gl/piglit_wfl_framework.c | 2 ++
2 files changed, 6 insertions(+), 3 deletions(-)
diff --git a/tests/util/piglit-framework-gl/piglit_glut_framework.c
This change makes the TestResult.pid field into a list of ints instead
of a list. This will be used by the multiple test per process tests to
ensure that all PIDs are recorded when the cherry features is invoked.
This is groundwork for later patches in the series.
Signed-off-by: Dylan Baker
This Mixin makes writing classes for handling tests that run multiple
tests in a single process simpler. It does this through the use of the
subtest feature. It makes it possible to implement two new methods, and
an aware interpret_result method and have support for this feature,
including a
From: Marek Olk
this no longer crashes:
- create context (calls atexit)
- teardown context
- create context 2 (calls atexit)
- exit()
---
tests/util/piglit-framework-gl.c | 4
1 file changed, 4 insertions(+), 0 deletions(-)
diff --git
From: Marek Olk
piglit_report_result prevents running multiple tests per process.
---
tests/shaders/shader_runner.c | 231 +--
1 file changed, 140 insertions(+), 91 deletions(-)
diff --git a/tests/shaders/shader_runner.c
This splits the FastSkipMixin into two classes. One that actually does
the fast skipping, and one that provides a mixin for Test.
This split will allow the class to be used in the traditional way, but
also in shader_test.MultiShaderTest.
Signed-off-by: Dylan Baker
---
This allows shader_runner to consume more than one test at a time. When
it does this it reports the results as subtests rather than as regular
tests. It also updates the python framework to handle and resume from
crashes in the tests. When this mode is enabled, shader_runner runs an
entire
This is both nicer since it provides better code quality (a separation
of roles), and because it's going to make writing a shader_test class
that can do N:1 tests much easier.
Signed-off-by: Dylan Baker
---
framework/test/shader_test.py | 106
On Fri, Sep 9, 2016 at 3:07 PM, Ian Romanick wrote:
>> + /* Use values equal to zero (because "less than" is a spec bug). */
>
> That is true... have you submitted a bug to Khronos? That should get fixed.
Is there a good way of doing that for non-Khronos members? At
On 08/29/2016 09:24 AM, Adam Jackson wrote:
> Takes over 40 minutes and 16GB on Skylake. Most of the blend
16 *gigabytes*? Why? That has to indicate a bug somewhere... possibly
a leak in the test?
> functionality in llvmpipe is adequately covered by Mesa's in-tree tests
> anyway.
>
>
On 09/08/2016 12:00 PM, Samuel Pitoiset wrote:
> Signed-off-by: Samuel Pitoiset
> ---
> tests/all.py | 8 +
> tests/spec/CMakeLists.txt | 1 +
> .../CMakeLists.gl.txt | 16
On Fri, Sep 9, 2016 at 2:30 PM, Kristian Høgsberg wrote:
> On Fri, Sep 9, 2016 at 10:44 AM Rob Clark wrote:
>>
>> diff --git a/tests/spec/ext_image_dma_buf_import/missing_attributes.c
>> b/tests/spec/ext_image_dma_buf_import/missing_attributes.c
>> index
Ok, so the basic problem with the YUV tests is that they currently
completely ignore driver/hw pitch requirements, since the code that
allocates the buffer doesn't know the pixel format, only the 'cpp'.
The yuv test creates a small 4x4 yuv eglimage. If, say, the hardware
requires the pitch to be
Quoting Marek Olšák (2016-09-09 06:10:56)
> From: Marek Olšák
>
> Older versions don't work and there is no interest in fixing the official
> gl45release CTS branch, so it's useless to add those forever-broken tests.
>
> Using '-t gl45-cts' is annoying, because adding '-t
On 09/09/2016 10:53 AM, Nicolai Hähnle wrote:
Hi Samuel,
Hi Nicolai,
Thank you for doing this!
There should also be a test that exercises a range of different work
group sizes with the same shader, up to the implementation maximums.
Okay, I can write such a test.
On 08.09.2016
Hi Samuel,
Thank you for doing this!
There should also be a test that exercises a range of different work
group sizes with the same shader, up to the implementation maximums.
On 08.09.2016 21:00, Samuel Pitoiset wrote:
Signed-off-by: Samuel Pitoiset
---
27 matches
Mail list logo