Re: [Piglit] [PATCH 1/7] framework_tests/summary.py: Tests Summary.py's assignment of statues

On Sunday, October 13, 2013 10:51:50 PM Kenneth Graunke wrote: > On 09/27/2013 02:39 PM, Dylan Baker wrote: > > This code tests the way Summary.py assigns statuses to tests. > > Specifically it tests regressions, fixes, changes, and problems, > > assuring that only the correct permutations of these

Re: [Piglit] [PATCH 3/7] framework/summary.py: Fix problems with status handling

On Sunday, October 13, 2013 11:08:27 PM Kenneth Graunke wrote: > On 09/27/2013 02:39 PM, Dylan Baker wrote: > > Fixes status hanlding (fixes, regressions, changes) to be compliant with > > the status handling in framework_tests/summary.py. Mainly this addresses > > notrun -> * and * -> notrun, as w

Re: [Piglit] [PATCH 4/7] status.py: Adds a new status class and subclasses

On Sunday, October 13, 2013 11:13:48 PM Kenneth Graunke wrote: > On 09/27/2013 02:39 PM, Dylan Baker wrote: > > These status classes are intended to replace the string statuses > > currently used by piglit summary. > > > > They give a couple of very nice advantages, first, they can be directly > >

Re: [Piglit] [PATCH 5/7] status: Make use of the status objects

On Sunday, October 13, 2013 11:29:02 PM Kenneth Graunke wrote: > On 09/27/2013 02:39 PM, Dylan Baker wrote: > > This adds code in framework/summary.py and framework/core.py to make use > > of the status classes in status.py. This makes comparisons between > > statuses much simpler and cleaner > >

Re: [Piglit] [PATCH 1/8] GL3.2 GL_ARB_sync: Test to check the correct error messages are returned for invalid inputs for ClientWaitSync

Thanks for the cleanups and fixes. I've committed patches 1,3-7 with minor fixes. I've added a v4 annotation for each patch. As for the patch subject, the convention is that we use a simpler prefix. If a test is for an extension that was later promoted to a real OpenGL version, like GL_ARB_syn

Re: [Piglit] [PATCH 8/8] GL3.2 GL_ARB_sync: Test that the correct error messages are returned from invalid input for WaitSync

On 10/07/2013 08:46 AM, Nicholas Mack wrote: v2: Fix comments, remove redundant code v3: Add to all.tests --- tests/all.tests | 1 + tests/spec/arb_sync/CMakeLists.gl.txt | 1 + tests/spec/arb_sync/WaitSync-errors.c | 82 +++ 3 files c

Re: [Piglit] [PATCH 0/5] Support creation of debug and forward-compatible contexts

On 10/13/2013 05:38 AM, Timothy Arceri wrote: > Hi Chad, > > I successfully tested the patches for creating a debug-context with an > Intel GPU but if I added config.require_debug_context = true; to > tests/spec/arb_debug_output/api_error.c and run on a PC running AMD > Catalyst drivers I get the

Re: [Piglit] [PATCH 1/4] GS: Test that geometry shader input/output layout qualifiers only compile if valid

On 30 September 2013 15:38, Nicholas Mack wrote: > v2: Tests check against list of valid layouts instead of invalid layouts > Last time I reviewed this patch I pointed out that it's not necessary to have vertex and fragment shaders in the test, since the error is supposed to be detected at compi

[Piglit] [PATCH] arb_texture_storage: add some cube map error tests

Test error checking for width != height and depth % 6 != 0 for cube arrays. --- tests/spec/arb_texture_storage/texture-storage.c | 66 ++ 1 file changed, 66 insertions(+) diff --git a/tests/spec/arb_texture_storage/texture-storage.c b/tests/spec/arb_texture_storage/texture-

Re: [Piglit] [PATCH 4/4 v2] GS: Test that max_vertices cannot be set to INT_MAX

On 30 September 2013 15:38, Nicholas Mack wrote: > v2: Add comment > I sent a comment on patch 1. Patches 2-4 are: Reviewed-by: Paul Berry > > Reviewed-by: Paul Berry > --- > .../compiler/layout-max-verts-limited.geom | 23 > ++ > 1 file changed, 23 insertions(

Re: [Piglit] [PATCH 1/3] sso: Test mixing separable and non-separable programs in various ways

On 1 October 2013 18:22, Ian Romanick wrote: > From: Gregory Hainaut > > the goal is to test the state mix of glUseProgram / > glBindProgramPipeline / glActiveProgram. > > Ian R. quote: > "In this case, either the UseProgram state or the > BindProgramPipeline state. If UseProgram sets a

[Piglit] [PATCH 2/6] ARB_sample_shading: Add test to verify the functionality of gl_NumSamples

Signed-off-by: Anuj Phogat --- tests/all.tests| 7 + .../arb_sample_shading/execution/CMakeLists.gl.txt | 1 + .../execution/builtin-gl-num-samples.cpp | 228 + 3 files changed, 236 insertions(+) create mode 100644 tests/spec

[Piglit] [PATCH 1/6] ARB_sample_shading: Add test to verify new APIs and enums

Signed-off-by: Anuj Phogat --- tests/all.tests| 5 ++ tests/spec/CMakeLists.txt | 1 + tests/spec/arb_sample_shading/CMakeLists.txt | 1 + .../arb_sample_shading/execution/CMakeLists.gl.txt | 14 + .../arb_sample_shading/ex

[Piglit] [PATCH 3/6] ARB_sample_shading: Add test to verify the functionality of gl_SampleID

Signed-off-by: Anuj Phogat --- tests/all.tests| 6 + .../arb_sample_shading/execution/CMakeLists.gl.txt | 1 + .../execution/builtin-gl-sample-id.cpp | 239 + 3 files changed, 246 insertions(+) create mode 100644 tests/spec

[Piglit] [PATCH 5/6] ARB_sample_shading: Add test to verify the functionality of gl_SamplePosition

Signed-off-by: Anuj Phogat --- tests/all.tests| 6 + .../arb_sample_shading/execution/CMakeLists.gl.txt | 1 + .../execution/builtin-gl-sample-position.cpp | 240 + 3 files changed, 247 insertions(+) create mode 100644 tests/spec

[Piglit] [PATCH 6/6] ARB_sample_shading: Add compiler tests for new builtins

Signed-off-by: Anuj Phogat --- tests/all.tests | 4 tests/spec/arb_sample_shading/compiler/gl_NumSamples.frag | 12 tests/spec/arb_sample_shading/compiler/gl_SampleID.frag | 12 tests/spec/arb_sample_shading/compile

[Piglit] [PATCH 4/6] ARB_sample_shading: Add test to verify the functionality of gl_SampleMask[]

Signed-off-by: Anuj Phogat --- tests/all.tests| 6 + .../arb_sample_shading/execution/CMakeLists.gl.txt | 1 + .../execution/builtin-gl-sample-mask.cpp | 245 + 3 files changed, 252 insertions(+) create mode 100644 tests/spec

Re: [Piglit] [PATCH 2/3] sso: Verify that rendezvous by location also works

On 1 October 2013 18:22, Ian Romanick wrote: > From: Ian Romanick > > Separate shader objects can either connect outputs to inputs by matching > names (rendezvous by name) or by assigning explicit locations via the > layout(location=...) (rendezvous by location). This is a simple test > that a

Re: [Piglit] [PATCH 1/7] framework_tests/summary.py: Tests Summary.py's assignment of statues

On 10/14/2013 06:48 AM, Dylan Baker wrote: > On Sunday, October 13, 2013 10:51:50 PM Kenneth Graunke wrote: >> On 09/27/2013 02:39 PM, Dylan Baker wrote: [snip] >> pass? Will it actually work if this doesn't get created? I'm guessing >> you need to abort the test (or just don't catch the exceptio

[Piglit] [PATCH] cmake: Detect if system has POSIX clocks

Set CMake cache var PIGLIT_HAS_POSIX_CLOCKS if if clock_gettime() is available. Tests that need to measure time, such as GL_ARB_sync tests, will use Posix clocks. CC: Nicholas Mack Signed-off-by: Chad Versace --- CMakeLists.txt | 11 +++ 1 file changed, 11 insertions(+) diff --git a/C

Re: [Piglit] [PATCH 3/3] sso: Combine 20 vertex shaders and 20 fragment shaders in various ways.

On 1 October 2013 18:22, Ian Romanick wrote: > From: Ian Romanick > > Verify that the right shaders are used in the right combinations several > ways. > > * The vertex shader has information baked-in that determines the X position > of the block on the screen. > > * The fragment shader has inf

Re: [Piglit] [PATCH 3/3] sso: Combine 20 vertex shaders and 20 fragment shaders in various ways.

On 10/14/2013 11:20 AM, Paul Berry wrote: > On 1 October 2013 18:22, Ian Romanick > wrote: > > From: Ian Romanick > > > Verify that the right shaders are used in the right combinations several > ways. > > * The v

Re: [Piglit] [PATCH 0/5] Support creation of debug and forward-compatible contexts

On 10/13/2013 08:29 PM, Timothy Arceri wrote: > Ok I randomly came across that fact this morning that > GLX_ARB_create_context was not supported until X.org server 1.13. I was > doing my AMD testing on a Ubuntu 12.04 machine so I guess thats probably > the issue. That should only matter for Mesa-b

Re: [Piglit] [PATCH] Change INVALID_VALUE to INVALID_OPERATION in ARB_vertex_array_bgra test.

Hi, I think the glean/vertArrayBGRA test also need to be updated, because it fails with this message: vertArrayBGRA: Error: glColorPointer(size=GL_BGRA, type=GL_FLOAT) did not generate expected error. since commit: 0e7a61a29f883c63a5439ac16eddeba3aaf4 mesa: Update the BGRA vertex array error

Re: [Piglit] [PATCH 2/6] ARB_sample_shading: Add test to verify the functionality of gl_NumSamples

The projection seems spurious. Just set gl_Position = vec4(pos, 0.0, 1.0) in your VS and lose the matrix and supporting machinery? On Tue, Oct 15, 2013 at 6:19 AM, Anuj Phogat wrote: > Signed-off-by: Anuj Phogat > --- > tests/all.tests| 7 + > .../arb_sampl

Re: [Piglit] [PATCH 1/3] sso: Test mixing separable and non-separable programs in various ways

On 10/14/2013 10:17 AM, Paul Berry wrote: > On 1 October 2013 18:22, Ian Romanick > wrote: > > From: Gregory Hainaut > > > the goal is to test the state mix of glUseProgram / > glBindProgramPipeline / glActiveProgram

Re: [Piglit] [PATCH] built-in-constants: don't try to create GS with version < 150.

On 10/12/2013 10:11 AM, Paul Berry wrote: > Previously, built-in-constants looked at required_glsl_version to > determine what to include in the "#version" directive, but it looked > at glsl_version (the maximum version supported by the implementation) > to decide whether to test geometry shaders.

Re: [Piglit] [PATCH 5/6] ARB_sample_shading: Add test to verify the functionality of gl_SamplePosition

I'm not convinced that this test actually proves what you need it to. What if you pass the array of expected sample positions as a uniform, and compare expected_positions[gl_SampleID] to gl_SamplePosition in the shader? On Tue, Oct 15, 2013 at 6:19 AM, Anuj Phogat wrote: > Signed-off-by: Anuj Ph

Re: [Piglit] [PATCH] glut framework: skip if implementation doesn't support required GL version.

On 10/08/2013 05:12 PM, Paul Berry wrote: In effect, we are already doing this when Waffle is in use, since Waffle uses a context creation mechanism that ensures that the context won't be created if the implementation doesn't support the GL version needed by the test. But for GLUT we need to do

Re: [Piglit] [PATCH] sso: Add compile tests for the location layout qualifier

On 2 October 2013 11:04, Ian Romanick wrote: > From: Ian Romanick > > NVIDIA (304.64 on GTX 260) fails all of the 1.10, 1.20, and 1.30 tests > because they fail to recognize the layout keyword at all. Paul has > suggested that adding > > #extension GL_ARB_fragment_coord_conventions: require

Re: [Piglit] [PATCH] built-in-constants: don't try to create GS with version < 150.

On 14 October 2013 12:44, Ian Romanick wrote: > On 10/12/2013 10:11 AM, Paul Berry wrote: > > Previously, built-in-constants looked at required_glsl_version to > > determine what to include in the "#version" directive, but it looked > > at glsl_version (the maximum version supported by the implem

Re: [Piglit] [PATCH 2/8] GL3.2 GL_ARB_sync: Test for valid return values from ClientWaitSync

On 10/07/2013 08:46 AM, Nicholas Mack wrote: v2: Fix comments, initialize variables. Still need to figure out if GPU commands needed before testing these. v3: Rewrite test cases, fix comment and config block and add to all.tests --- tests/all.tests | 1 +

Re: [Piglit] [PATCH 0/5] Support creation of debug and forward-compatible contexts

On 10/14/2013 09:21 AM, Ian Romanick wrote: On 10/13/2013 05:38 AM, Timothy Arceri wrote: Hi Chad, I successfully tested the patches for creating a debug-context with an Intel GPU but if I added config.require_debug_context = true; to tests/spec/arb_debug_output/api_error.c and run on a PC runn

Re: [Piglit] [PATCH] built-in-constants: don't try to create GS with version < 150.

On 10/14/2013 02:05 PM, Paul Berry wrote: > On 14 October 2013 12:44, Ian Romanick > wrote: > > On 10/12/2013 10:11 AM, Paul Berry wrote: > > Previously, built-in-constants looked at required_glsl_version to > > determine what to include in the "#version"

Re: [Piglit] [PATCH 1/7] cl-program-tester: Use more appropriate member names for struct test_arg

On Mon, Sep 30, 2013 at 9:47 AM, Tom Stellard wrote: > From: Tom Stellard > > cl_size has been renamed to vec_elements and cl_mem_size has been > renamed to vec_mem_elements. This use of size is confusing, because > size generally means number of bytes when used to describe a type, and > in this

Re: [Piglit] [PATCH 2/7] piglit-cl-util: Add a helper for getting the alignment of vector types

On Mon, Sep 30, 2013 at 9:47 AM, Tom Stellard wrote: > From: Tom Stellard > > --- > tests/cl/program/program-tester.c | 2 +- > tests/util/piglit-util-cl.c | 13 + > tests/util/piglit-util-cl.h | 6 ++ > 3 files changed, 20 insertions(+), 1 deletion(-) > > diff --gi

Re: [Piglit] [PATCH 3/7] cl-program-tester: Add a helper function for getting type sizes

On Mon, Sep 30, 2013 at 9:47 AM, Tom Stellard wrote: > From: Tom Stellard > > --- > tests/cl/program/program-tester.c | 99 > +-- > tests/util/piglit-util-cl.c | 20 > tests/util/piglit-util-cl.h | 20 > 3 files changed, 84 inser

Re: [Piglit] [PATCH 4/7] cl-program-tester: Rename test_arg::size to test_arg::size_in_mem

Reviewed-by: Aaron Watry On Mon, Sep 30, 2013 at 9:47 AM, Tom Stellard wrote: > From: Tom Stellard > > The value being stored in this field is actually the size of the argument if > it were to be stored in memory (for a type like int3, its size is 12 bytes, > but its size in memory is 16 bytes

Re: [Piglit] [PATCH 5/7] cl-progam-tester: Pass correct size to piglit_cl_set_kernel_arg()

On Mon, Sep 30, 2013 at 9:47 AM, Tom Stellard wrote: > From: Tom Stellard > > --- > tests/cl/program/program-tester.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/tests/cl/program/program-tester.c > b/tests/cl/program/program-tester.c > index ea16c81..557dc4b 100644

Re: [Piglit] [PATCH 4/8] util: Move EGL utilities from libpiglitutil_gl to libpiglitutil

I think this needs a bit more love in the CMakeLists. It can't find EGL/egl.h and friends when they're not installed in the default places. I sure wish cmake and pkgconfig were better friends. :( On 10/11/2013 07:09 PM, Chad Versace wrote: > Many EGL tests make no GL calls. This is one step towa

Re: [Piglit] [PATCH 4/8] util: Move EGL utilities from libpiglitutil_gl to libpiglitutil

On 10/14/2013 03:25 PM, Ian Romanick wrote: > I think this needs a bit more love in the CMakeLists. It can't find > EGL/egl.h and friends when they're not installed in the default places. > I sure wish cmake and pkgconfig were better friends. :( Actually, it's a similar commit in a different tes

[Piglit] [PATCH] gleantest.py: Return skip when appropriate

Currently glean can only return dmesg-warn, dmesg-fail, fail, and pass. Which means that skip will never be returned even if should be. This patch adds a catch to find a skip status and return it. Signed-off-by: Dylan Baker --- framework/gleantest.py | 4 +++- 1 file changed, 3 insertions(+), 1

[Piglit] [PATCH] ARB_texture_view: Add test for API coverage

Tests GL_ARB_texture_view and validity of input parameters for API calls. Uses both valid and invalid parameters. Ran on Nvidia Quadro 600, all subtests pass. --- tests/spec/arb_texture_view/params.c | 643 +++ 1 file changed, 643 insertions(+) create mode 1006

Re: [Piglit] [PATCH 5/6] ARB_sample_shading: Add test to verify the functionality of gl_SamplePosition

On Mon, Oct 14, 2013 at 12:57 PM, Chris Forbes wrote: > I'm not convinced that this test actually proves what you need it to. > Yes, I just realized that the expected color can be obtained by various values of sample positions. Correct values are just one of all possible cases. > What if you pass

Re: [Piglit] [PATCH 2/6] ARB_sample_shading: Add test to verify the functionality of gl_NumSamples

On Mon, Oct 14, 2013 at 12:35 PM, Chris Forbes wrote: > The projection seems spurious. Just set gl_Position = vec4(pos, 0.0, > 1.0) in your VS and lose the matrix and supporting machinery? I'll get rid of projection matrix. > > On Tue, Oct 15, 2013 at 6:19 AM, Anuj Phogat wrote: >> Signed-off-by

[Piglit] [PATCH] ARB_texture_view: Add query tests for gl state effected by this extension

Tests the queryable state from ARB_texture_view extension is correct. Includes GL_TEXTURE_IMMUTABLE_LEVELS,GL_TEXTURE_VIEW_MIN_LEVEL, GL_TEXTURE_VIEW_NUM_LEVELS, GL_TEXTURE_IMMUTABLE_FORMAT. Tested on Nvidia Quadro 600, all subtests pass. --- tests/all.tests | 1 +

Re: [Piglit] [PATCH 7/7] cl-program-tester: Add gentype option to config section

On Mon, Sep 30, 2013 at 9:47 AM, Tom Stellard wrote: > From: Tom Stellard > > The gentype option allows a test file to specify a list of types and > vector sizes to be used for each test. cl-program-tester will run each > test once for each comibnation of type and vector size. > > This makes it

[Piglit] [PATCH] ARB_texture_view: Add test for rendering with texture view.

Tests GL_ARB_texture_view rendering and lifetime of views. Rendering includes using various targets, layers and levels. Lifetime tests are combined with various formats and tex image data reads. Tested on Nvidia Quadro 600; all subtests pass. --- tests/all.tests |

Re: [Piglit] [PATCH] cmake: Detect if system has POSIX clocks

Chad, I wanted a similar feature, but I was also thinking of adding a piglit-util function (piglit_get_microseconds). I think it should be easy to port this function to windows. I also had the cmake parts a little different, but I don't really have a preference on that. http://cgit.freedesktop.o

[Piglit] [PATCH] CL: Add global memory read/write and barrier tests

Based on local-memory tests. --- tests/cl/program/execute/global-memory.cl | 83 +++ 1 file changed, 83 insertions(+) create mode 100644 tests/cl/program/execute/global-memory.cl diff --git a/tests/cl/program/execute/global-memory.cl b/tests/cl/program/execute/global