[Git][xorg-team/app/piglit][debian-unstable] 208 commits: arb_direct_state_access: fix internaformat query behavior
"Jordan Justen (@jljusten)" <[email protected]>
| Newsgroups | gmane.linux.debian.devel.x |
|---|---|
| Message-ID | <[email protected]> |
Jordan Justen pushed to branch debian-unstable at X Strike Force / app / piglit
Commits:
0ecdebb0 by Erik Faye-Lund at 2025-04-14T10:23:02+02:00
arb_direct_state_access: fix internaformat query behavior
The error condition and error value changes with and without
GL_ARB_internalformat_query. Even though this test are using the named
version of these APIs, they inherit their error-handling rules from the
non-named versions.
Before introducing that extension, we need to check against
GL_MAX_SAMPLES, which is what we did. But the reported error should be
GL_INVALID_VALUE, not GL_INVALID_OPERATION like we checked against.
To quote the OpenGL 4.1 spec, section 4.4.2 ("Attaching Images to
Framebuffer Objects"):
If either width or height is greater than the value of
MAX_RENDERBUFFER_SIZE, or if samples is greater than the value of
MAX_SAMPLES, then the error INVALID_VALUE is generated.
After introducing that extension, we need to check against the
per-internal-format max instead of the "global max", because this is
now the smallest value that is *guaranteed* to work instead.
To quote the OpenGL 4.1 spec, section 4.4.2 ("Attaching Images to
Framebuffer Objects"):
If either width or height is greater than the value of
MAX_RENDERBUFFER_SIZE, then the error INVALID_VALUE is generated.
If samples is greater than the maximum number of samples supported
for internalformat, then the error INVALID_OPERATION is generated
(see GetInternalformativ in section 6.1.15).
So, both the value to check against and the error code changed at the
same time. The GL_ARB_internalformat_query specification also contains
these changes. It seems we never tested something that was consistent
here, but was instead relying on the max samples for GL_RGBA being the
same as the formatless max samples, which is not guaranteed to be the
case.
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1003>
- - - - -
d06f7bac by Ashley Smith at 2025-04-15T13:18:28+00:00
tests: Add test for GL_EXT_shader_clock
Reviewed-by: Erik Faye-Lund <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1002>
- - - - -
c50d9aa5 by Vinson Lee at 2025-04-21T22:47:33-07:00
cmake: Fix mis-matching arguments warning
CMake Warning (dev) in generated_tests/CMakeLists.txt:
A logical block opening on the line
generated_tests/CMakeLists.txt:345 (if)
closes on the line
generated_tests/CMakeLists.txt:347 (endif)
with mis-matching arguments.
This warning is for project developers. Use -Wno-dev to suppress it.
Fixes: af62c0dea ("vulkan: Add cmat tests using constant as accumulator")
Signed-off-by: Vinson Lee <[email protected]>
Reviewed-by: Caio Oliveira <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1008>
- - - - -
1498c397 by Caterina Shablia at 2025-04-22T14:38:42+00:00
tests: test UBO and SSBO backed by the same buffer
In the past, panfrost could not handle UBO and SSBO bindings being
backed by the same buffer. See
https://gitlab.freedesktop.org/mesa/mesa/-/issues/11100 for details.
Until
https://gitlab.freedesktop.org/mesa/mesa/-/commit/59a3e12039cde5df1451193557512b38cea0039e
this test will reliably cause panfrost to submit the batch it's
in the middle of using and crash.
Reviewed-by: Erik Faye-Lund <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1006>
- - - - -
1767af74 by Timothy Arceri at 2025-05-08T01:02:47+00:00
gl: test push pop attribute with color material
Tests the issue:
https://gitlab.freedesktop.org/mesa/mesa/-/issues/7122
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1009>
- - - - -
c32e5cc7 by Timothy Arceri at 2025-05-23T12:07:17+10:00
GL_EXT_texture_integer: basic validation tests
Tests FBO rendering interactions with GL_EXT_texture_integer attachments
specifically making sure that color attachments are ignored when not
active draw buffers.
Mesa bugs:
https://gitlab.freedesktop.org/mesa/mesa/-/issues/13144
https://gitlab.freedesktop.org/mesa/mesa/-/issues/13168
Reviewed-by: Emma Anholt <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1010>
- - - - -
685ea49b by Erik Faye-Lund at 2025-05-28T13:28:52+02:00
khr_texture_compression_astc: normalize error-color
The spec says that when using linear RGB decode, the error color can
either be magenta or all-NaN. So let's make sure we allow either.
This is slightly more lenient than needed, because we use the same
shader for the sRGB and linear RGB tests. But the complexity of
duplicating the shaders etc just doesn't seem to be worth it, so
let's just be tad lax here.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1012>
- - - - -
cabe21dc by ngcortes at 2025-05-29T14:57:59-07:00
fix invalid typecast
- - - - -
acae8274 by Tapani Pälli at 2025-06-02T10:28:06+00:00
tests/util: provide simple array for cleanup handlers
This makes it possible to call piglit_set_destroy_func multiple times.
This is done so that tests can register cleanup handler as well and
avoid usage of atexit which can cause issues with frameworks such as the
vulkan validation layers.
Signed-off-by: Tapani Pälli <[email protected]>
Acked-by: Lionel Landwerlin <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1011>
- - - - -
dba66b25 by Tapani Pälli at 2025-06-02T10:28:06+00:00
ext_external_objects: use piglit_set_destroy_func instead of atexit
Usage of atexit causes crash with validation layers and we would like
to run validation on vulkan tests.
Signed-off-by: Tapani Pälli <[email protected]>
Acked-by: Lionel Landwerlin <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1011>
- - - - -
31ea5461 by Tapani Pälli at 2025-06-02T10:28:06+00:00
ext_external_objects: fix timeline semaphore validation error
VUID-VkSemaphoreTypeCreateInfo-timelineSemaphore-03252
This was hit by ext_external_objects-vk-ping-pong.
Signed-off-by: Tapani Pälli <[email protected]>
Acked-by: Lionel Landwerlin <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1011>
- - - - -
51528d06 by Tapani Pälli at 2025-06-02T10:28:06+00:00
ext_external_objects: fix validation error with barriers
VUID-vkCmdPipelineBarrier-srcStageMask-09634
"If either srcStageMask or dstStageMask includes VK_PIPELINE_STAGE_HOST_BIT,
for any element of pBufferMemoryBarriers, srcQueueFamilyIndex and
dstQueueFamilyIndex must be equal"
This was hit by ext_external_objects-vk-buf-exchange.
Signed-off-by: Tapani Pälli <[email protected]>
Acked-by: Lionel Landwerlin <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1011>
- - - - -
dfe7c6c5 by Tapani Pälli at 2025-06-02T10:28:06+00:00
ext_external_objects: fix validation error on cmdbuf usage
VUID-vkBeginCommandBuffer-commandBuffer-00049
This was hit by ext_external_objects-vk-image-display-overwrite.
Signed-off-by: Tapani Pälli <[email protected]>
Acked-by: Lionel Landwerlin <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1011>
- - - - -
38bfc3ca by Tapani Pälli at 2025-06-02T10:28:06+00:00
ext_external_objects: fix validation error releasing resources
VUID-vkDestroyImage-image-01000
VUID-vkDestroyRenderPass-renderPass-00873
VUID-vkDestroyPipeline-pipeline-00765
VUID-vkDestroySemaphore-semaphore-05149
VUID-vkResetCommandBuffer-commandBuffer-00045
VUID-vkFreeCommandBuffers-pCommandBuffers-00047
These affected misc tests destroying resources when they were still in use.
Signed-off-by: Tapani Pälli <[email protected]>
Acked-by: Lionel Landwerlin <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1011>
- - - - -
e68d812f by Tapani Pälli at 2025-06-02T10:28:06+00:00
ext_external_objects: fix validation error for cmdbuffer usage
VUID-vkBeginCommandBuffer-commandBuffer-00049
VUID-vkQueueSubmit-pCommandBuffers-00071
This was hit by ext_external_objects-vk-vert-buf-update-errors.
Signed-off-by: Tapani Pälli <[email protected]>
Acked-by: Lionel Landwerlin <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1011>
- - - - -
dbfe2587 by Tapani Pälli at 2025-06-02T10:28:06+00:00
ext_external_objects: fix validation error with semaphore value
VUID-VkSemaphoreSignalInfo-value-03258
This was hit by ext_external_objects-vk-ping-pong.
Signed-off-by: Tapani Pälli <[email protected]>
Acked-by: Lionel Landwerlin <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1011>
- - - - -
a0a27e52 by Ashley Smith at 2025-06-03T09:53:44+00:00
tests: Add test for GL_EXT_shader_realtime_clock
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1000>
- - - - -
3608f71e by Christoph Neuhauser at 2025-06-13T18:07:45+02:00
tests: Add tests for EGL_EXT_device_query_name and EGL_EXT_device_persistent_id
Signed-off-by: Christoph Neuhauser <[email protected]>
Reviewed-by: Tapani Pälli <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1018>
- - - - -
fbc45075 by Ryan Mckeever at 2025-06-24T05:43:51+00:00
ext_image_dma_buf_import: Add support for RGB888/BGR888 formats
Signed-off-by: Ryan Mckeever <[email protected]>
Reviewed-by: Erik Faye-Lund <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1013>
- - - - -
5262e83b by Ryan Mckeever at 2025-06-24T05:43:51+00:00
ext_image_dma_buf_import: Add test for RGB888/BGR888 formats
This test verifies that we can successfully texture from and render to an
imported dmabuf for RGB888 and BGR888 formats.
Signed-off-by: Ryan Mckeever <[email protected]>
Reviewed-by: Erik Faye-Lund <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1013>
- - - - -
55d5b292 by Daniel Schürmann at 2025-06-25T10:26:03+00:00
Revert "ssa: Add a test for an edge case in into-SSA translation"
This reverts commit f18b947d7c03b81b989ed47551dc24bcc29cc0fc.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1017>
- - - - -
70df12a9 by Charlotte Pabst at 2025-06-25T23:30:36+00:00
tests: Add test for recompiling ARB fragment program
- - - - -
873a55ba by Timothy Arceri at 2025-07-07T00:52:34+00:00
teximage: fix broken teximage-scale-bias test
Apparently this is how it landed in piglit 10 years ago. However
the alpha test always causes the test to fail, not sure if piglit
was less strict about this 10 years ago.
Reviewed-by: Emma Anholt <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1021>
- - - - -
7a6d1dfb by Timothy Arceri at 2025-07-07T00:52:34+00:00
teximage: actually test the alpha scale and bias
Previously this was just set to the default scale and bias values
so lets actually change the values to test them.
Reviewed-by: Emma Anholt <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1021>
- - - - -
1a9e51cd by Timothy Arceri at 2025-07-07T00:52:34+00:00
teximage: expand transfer op test
Here we also test to make sure the correct pixel unpacking offsets
are applied when using transfer ops
Tests for the issues reported here:
https://gitlab.freedesktop.org/mesa/mesa/-/issues/10314
Reviewed-by: Emma Anholt <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1021>
- - - - -
0980079d by Samuel Pitoiset at 2025-07-11T14:05:09+02:00
arb_shader_image_load_store: remove index bounds test
This test is UB and might cause GPU hangs with Zink. I don't think it
makes sense to keep it.
Signed-off-by: Samuel Pitoiset <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1023>
- - - - -
0def6751 by Tapani Pälli at 2025-07-23T15:52:33+00:00
ext_external_objects: remove unneeded glFlush calls
These are not necessary as glSignalSemaphore is flushing already.
>From GL_EXT_semaphore spec:
"SignalSemaphoreEXT will insert a semaphore signaling operation
in the GL command stream, and flush the command stream as if
/Flush/ were inserted immediately after the semaphore operation."
Signed-off-by: Tapani Pälli <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1019>
- - - - -
00814522 by Emma Anholt at 2025-07-23T16:20:25+00:00
fog-modes: Add a variant using ARB_fp.
Turns out we had no coverage of ARB_fp fog functionality, just
compiler tests.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/813>
- - - - -
9eb217dd by Lucas Stach at 2025-07-23T16:45:39+00:00
masked-clear: test depth/stencil clear cross contamination
Add a test to validate that glClear with only depth or stencil buffer
bits set only affect the respective portion of the ZS buffer. This
reproduces a bug in etnaviv, where such clears would affect the whole
buffer when fast-clear is in use.
Signed-off-by: Lucas Stach <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/902>
- - - - -
6ce7e57e by Pierre-Eric Pelloux-Prayer at 2025-07-23T16:55:13+00:00
ARB_pipeline_statistics_query: don't use DISPLAY
It's not defined anywhere.
Also don't use piglit_report_result directly, it's done upper in the stack.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1024>
- - - - -
0c731f2b by Pierre-Eric Pelloux-Prayer at 2025-07-23T16:55:13+00:00
ARB_pipeline_statistics_query: prevent flat shading
GL_FRAGMENT_SHADER_INVOCATIONS_ARB will produce a value lower than
the number of fragments with radeonsi if vrs2x2 can be enabled.
Prevent this by enabling some interpolation.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1024>
- - - - -
0bd6b7e2 by Corentin Noël at 2025-07-23T17:05:03+00:00
arb_shader_image_load_store: Make sure to only call supported indirect function
As we are using glMultiDrawArraysIndirect, we need to ensure that
GL_ARB_multi_draw_indirect is available and not only GL_ARB_draw_indirect.
Signed-off-by: Corentin Noël <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/952>
- - - - -
1f5e149b by sarbes at 2025-07-23T17:27:36+00:00
tests: add tests for constant texture bias/lod
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1007>
- - - - -
4d176543 by Mike Blumenkrantz at 2025-07-23T17:48:13+00:00
glsl-1.50/execution/variable-indexing: check GS max input components
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1025>
- - - - -
39a6d580 by Erik Faye-Lund at 2025-07-23T18:22:01+00:00
add quirk-tests for ARB_depth_clamp
This verifies a few quirks of ARB_depth_clamp.
Signed-off-by: Erik Faye-Lund <[email protected]>
Acked-by: Mike Blumenkrantz <[email protected]>
Reviewed-by: David Heidelberg <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/925>
- - - - -
b603283d by Erik Faye-Lund at 2025-07-23T18:22:01+00:00
also test clamping with a geometry-shader
Acked-by: Mike Blumenkrantz <[email protected]>
Reviewed-by: David Heidelberg <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/925>
- - - - -
9d3d2c7b by Erik Faye-Lund at 2025-07-23T18:49:47+00:00
arb_robustness: modernize test
This propagates errors a bit more properly, and drops some asserts
in place of proper error-checks.
Signed-off-by: Erik Faye-Lund <[email protected]>
Reviewed-by: Alejandro Piñeiro <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/926>
- - - - -
88bc8051 by Erik Faye-Lund at 2025-07-23T18:49:47+00:00
arb_robustness: add test for empty vbo
Signed-off-by: Erik Faye-Lund <[email protected]>
Reviewed-by: Alejandro Piñeiro <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/926>
- - - - -
9309c413 by Ilario Gelmetti at 2025-07-30T06:07:21+00:00
Update README Ubuntu dependencies
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/622>
- - - - -
307511f1 by Mike Blumenkrantz at 2025-07-30T12:18:24+00:00
ext_external_objects: add NV_timeline_support to ping-pong test
this causes the test to add timeline semaphore waits/signals for even
more ping-ponging if all extensions are supported
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1022>
- - - - -
1e0b4de8 by Jonathan at 2025-07-30T14:45:33+00:00
tests/glx: test for buffer invalidation / rebind bugs
See https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17338
See https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/17394
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/702>
- - - - -
c3a3e29d by Mike Blumenkrantz at 2025-07-30T18:01:22+00:00
glx: fix cleanup on some multithread tests with -auto
this avoids leaking the contexts
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1027>
- - - - -
ed23c9da by Emma Anholt at 2025-08-01T17:47:24+00:00
Sort the output static test lists to improve build reproducibility.
We shouldn't be depending on FS walking behavior for contents of our
build outputs.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/459>
- - - - -
d3b07b43 by Emma Anholt at 2025-08-01T17:47:24+00:00
Sort the list of tests being filtered.
This should fix build reproducibility issues in the quick shaders
list, since we were handing things to the filter based on python
dictionary order.
No noticeable change in the 286MB maximum RSS of "/usr/bin/time -v
python3 ./piglit-run.py -d tests/quick.py results/asdf".
Closes: #10
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/459>
- - - - -
d04bd216 by Mike Blumenkrantz at 2025-08-06T17:36:28-04:00
glx: add a new multi-context-ib test for catching multi-context buffer destroys
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1029>
- - - - -
4060cda0 by Antia Puentes at 2025-08-11T17:02:43+02:00
arb_gl_spirv: Added tests for glGetProgramiv()
The SPIR-V used does not contain names. As names are optional in
SPIR-V this is useful to test the behaviour of the name related
queries when no name reflection is available.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
2f92d5a9 by Antia Puentes at 2025-08-12T12:35:37+02:00
shader_runner: Add skeleton to support more query commands
Interfaces and pnames are reordered alphabetically.
Transform feedback interface is removed as there is currently no
support for xfb in the shader_runner.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
7617dd1b by Antia Puentes at 2025-08-12T12:35:46+02:00
shader_runner: Allow to query resources using no-name data
Tests glGetProgramResourceiv(.., index, ..) identifying the resource,
and hence getting its index, using no-name data.
In ARB_gl_spirv name-reflection information is not mandatory. In that
case resources can be identified using other data depending on the
interface being queried, like binding, offset, location, etc.
This new query command is aimed to be used for ARB_gl_spirv testing,
but can be used as well for GLSL shaders. However, support for
subroutines was left for later implementation (they are not allowed
for ARB_gl_spirv).
Transform feedback interface is not allowed either because the
shader_runner lacks of XFB support right now (it will be added later).
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
2ed1ecfa by Antia Puentes at 2025-08-12T12:35:46+02:00
shader_runner: Simplify the code to query resources by name
Tests glGetProgramResourceiv(.., index, ..) locating the resource
and its index using the resource name.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
629b250b by Antia Puentes at 2025-08-12T12:35:46+02:00
shader_runner: Add command to query program interfaces
Tests glGetProgramInterfaceiv().
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
c83a3da2 by Antia Puentes at 2025-08-12T12:35:46+02:00
shader_runner: Add command to query the resource name
Tests glGetProgramResourceName().
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
bf158bbc by Antia Puentes at 2025-08-12T12:35:46+02:00
shader_runner: Add commands to query index, loc, loc-index.
Tests glGetProgramResourceIndex(), glGetProgramLocation() and
glGetProgramLocationIndex().
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
36642eab by Antia Puentes at 2025-08-12T12:46:32+02:00
shader_runner: Check compatibility for resource name queries
For several queries defined at ARB_program_interface_query, OpenGL
already provided and API to query such information.
With this patch we test that the queries defined at
ARB_program_interface_query returns the same values that such API
already existing. On the patch, we refer the latter for "old APIs".
This helps us to test that the old APIs return the desired values when
no name reflection is available as defined in the ARB_gl_spirv
specification:
"22. How do the program interface query APIs work when no name reflection
information is available?
...
GetUniformLocation(.., name) -> -1
GetActiveUniformName(.., length, uniformName) -> length=0, uniformName = ""
GetUniformIndices(..,uniformNames, uniformIndices) -> uniformIndices=INVALID_INDEX
GetActiveUniform(..,length,.., name) -> length = 0, name = ""
GetActiveUniformsiv(..,pname=UNIFORM_NAME_LENGTH,..) -> params = 1
GetUniformBlockIndex(.., uniformBlockName) -> INVALID_INDEX
GetActiveUniformBlockName(..,length,uniformBlockName) -> length=0, uniformBlockName=""
GetActiveUniformBlockiv(.., pname=UNIFORM_BLOCK_NAME_LENGTH, ..) -> params = 1
GetActiveAttrib(..) -> length = 0, name = ""
GetAttribLocation(.., name) -> -1
GetFragDataLocation(.., name) -> -1
GetFragDataIndex(.., name) -> -1"
v2: tweak the commit message to explain better what "old APIs" means
on this patch (Alejandro Piñeiro)
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
3e69d7ae by Antia Puentes at 2025-08-12T12:46:49+02:00
shader_runner: add UNSIGNED_INT_ATOMIC_COUNTER as allowed type
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
48face19 by Antia Puentes at 2025-08-12T12:46:49+02:00
arb_gl_spirv: Add ARB_program_interface_query tests
Added tests to check that the ARB_program_interface_query queries
involving names behave as described in the ARB_gl_spirv extension.
The SPIR-V used do not contain names. As names are optional in SPIR-V
this is useful to test the behaviour of the name related queries when
no name reflection is available.
Adds also tests to check that the queries not involving names work as
normally, as defined in the ARB_gl_spirv_extension specification.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
f2280b2b by Antia Puentes at 2025-08-12T12:46:49+02:00
shader_runner: support XFB in program interface queries
Also added the compatibility checks for the XFB name queries, which
compare the results obtained with the ones got when using
glGetTransformFeedbackVarying. This is used to indirectly test the
ARB_gl_spirv, Issue 22:
22. How do the program interface query APIs work when no name
reflection information is available?
... RESOLVED: The following naturally follows from the
specification:
GetTransformFeedbackVarying(..) -> length = 0, name = ""
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
98271224 by Alejandro Piñeiro at 2025-08-12T12:46:49+02:00
arb_gl_spirv: add a xfb test that uses different buffers, mixing order
In order to check that GL_TRANSFORM_FEEDBACK_BUFFER_INDEX is properly
computed, as at some point, when mixing order it was assigned extra
buffer_index.
It is worth to note that although the original GLSL do that definition
out, glslang groups and reorder the definitions based on xfb_offset
and xfb_buffer, so we needed to tweak by hand the SPIR-V generated.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
f0072053 by Antia Puentes at 2025-08-12T12:46:49+02:00
arb_gl_spirv: Add TRANSFORM_FEEDBACK_VARYING_MAX_LENGTH test
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
7828053f by Antia Puentes at 2025-08-12T12:46:49+02:00
arb_gl_spirv: added query tests for XFB
Adds program interface query tests for the
TRANSFORM_FEEDBACK_BUFFER and TRANSFORM_FEEDBACK_VARYING interfaces.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
85569df2 by Alejandro Piñeiro at 2025-08-12T12:46:49+02:00
arb_gl_spirv: array of interface block test
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
5418ac96 by Alejandro Piñeiro at 2025-08-12T12:46:49+02:00
arb_gl_spirv: add a basic geometry shader
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
5e476883 by Eleni Maria Stea at 2025-08-12T12:46:49+02:00
arb_gl_spirv: added tests for functions and function calls
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
83afcc36 by Alejandro Piñeiro at 2025-08-12T12:46:49+02:00
arb_gl_spirv: adding basic vertex attrib 64-bit tests
For now just:
* Simple
* Simple with array.
In general ARB_gl_spirv implementation didn't need to add too many
specifics. Those two were enough to catch the minor things, and enough
to cover the barebone testing. For a more detailed testing it is more
practical to borrow the extensive testing done on the
arb_vertex_attrib_64bit generated tests.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
b41013e4 by Neil Roberts at 2025-08-12T12:46:49+02:00
arb_gl_spirv: Add a test for matching a block by location with SSOs
This is copied from
arb_enhanced_layouts/execution/block-member-locations
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
fc99f525 by Neil Roberts at 2025-08-12T12:46:49+02:00
shader_runner: Add support for indexed draw calls
There is now an [elements] section which can contain a list of indices
as well as a draw elements command to draw with it.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
1049b9b6 by Neil Roberts at 2025-08-12T12:46:49+02:00
shader_runner: Add a “draw arrays instanced base” command
This is just like “draw arrays instanced” but has an extra parameter
to set the baseinstance.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
85c045c5 by Neil Roberts at 2025-08-12T12:46:49+02:00
arb_gl_spirv: Add a test for gl_VertexID
This value is different from Vulkan because it should include the
firstvertex and basevertex parameters.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
14581b6a by Neil Roberts at 2025-08-12T12:46:49+02:00
arb_gl_spirv: Add a test for gl_InstanceID
This value is different from InstanceID in Vulkan because it shouldn’t
include the baseinstance.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
7f46c6f1 by Neil Roberts at 2025-08-12T12:46:49+02:00
arb_gl_spirv: Add a test for the OpFUnord* opcodes
These opcodes are not available via GLSL so it seems worth testing
them separately. This involves generating NaN values which apparently
isn’t required by the spec at least on GLSL. The tests therefore try
to calculate the square root of -1 and if check if it actually
generated a NaN with isnan. If it didn’t generate NaN then the test is
assumed to just pass.
As mentioned, those opcodes are not available on GLSL, so the test is
SPIRV ONLY.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
8bdc7926 by Neil Roberts at 2025-08-12T12:46:49+02:00
arb_gl_spirv: Add tests for the gl_FragCoord origin
Using OriginLowerLeft isn’t allowed on Vulkan but explicitly allowed
by ARB_gl_spirv so it’s probably worth testing.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
fffd6202 by Neil Roberts at 2025-08-12T12:46:49+02:00
arb_gl_spirv: Add a test for PixelCenterInteger
This is forbidden in Vulkan but it is available in GLSL and the SPIR-V
spec allows it. It isn’t mentioned in the ARB_gl_spirv spec but it
seems likely that it’s intended to work.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
da1416a6 by Neil Roberts at 2025-08-12T12:46:49+02:00
arb_gl_spirv: Add a test for the SamplePosition builtin
Tests whether setting OriginUpperLeft and OriginLowerLeft affects the
SamplePosition builtin. It draws a grid of rectangles into a
multi-sample framebuffer where each rectangle only covers the top half
of the pixel. The fragment shaders store the sample position in the
buffer. It then uses multisample texturing to read back the samples.
Any samples written by the fragment shader should have the y position
all greater than 0.5. It’s not clear whether changing the origin
should affect this and there is an open spec issue about it, but for
now this test assumes that it shouldn’t.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
db4140bb by Neil Roberts at 2025-08-12T12:46:49+02:00
arb_gl_spirv: Add a test for gl_PatchVerticesIn
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
2d2040b4 by Antia Puentes at 2025-08-12T12:46:50+02:00
arb_gl_spirv: add test for atomic images
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
249451ab by Antia Puentes at 2025-08-12T12:46:50+02:00
arb_gl_spirv: Add test for atomic memory ops (shared uint)
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
5309b6fb by Alejandro Piñeiro at 2025-08-12T12:46:50+02:00
arb_gl_spirv: added regression catching function call test
This test was crashing until the following glslang issue got solved:
https://github.com/KhronosGroup/glslang/issues/1285
Putting here the test mostly for regression catching.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
fde91326 by Arcady Goldmints-Orlov at 2025-08-12T12:46:50+02:00
arb_gl_spirv: Add test equivalent to CS basic-ssbo
The difference with
arb_compute_shader/execution/basic-ssbo.shader_test is that we
explicitly assign a binding to the SSBO. Without a binding number
specified, glslang implicitly assigns a binding number of 1 to the
SSBO, while Mesa assigns a binding number of 0. To ensure the test
runs consistently in both cases, assign the binding explicitly.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
95f191eb by Arcady Goldmints-Orlov at 2025-08-12T12:46:50+02:00
arb_gl_spirv: add tests equivalent to component-layout SSO
The difference is that we explicitly specify locations for inputs and outputs in
component-layout tests.
The current version of glslang doesn't handle a mixture of explicit and implicit
locations for inputs and outputs, so make all locations explicit.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
9f9f603c by Alejandro Piñeiro at 2025-08-12T12:46:50+02:00
arb_gl_spirv: fix wrong extension name for atomic-uint-fs test
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
51f2f4da by Eleni Maria Stea at 2025-08-12T12:46:50+02:00
arb_gl_spirv: add tests for different types of varyings and blocks
Signed-off-by: Eleni Maria Estea <[email protected]>
Signed-off-by: Alejandro Piñeiro <[email protected]>
v2: (changes made by Alejandro Piñeiro)
* Change several tests to not use addition of values, don't use
repeated values, and use more different and complex values, to
ensure that we are getting the correct values, and not the same
values by (bad) luck.
* Change expected value to be green instead of red on several
tests. There is a kind of informal (and not always applied) rule
on piglit tests about green==pass color, red==fail
color.
* Strip names. Names are optional. It should work without names, so
it is preferred to have the shader without names.
* Removed several individual scalar type tests, as all of them are
already tested on all-scalar test, and partially on
all-data-types.
* Removed all individual vecX tests, as all of them are already
tested with all-data-types test.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
701843aa by Nicolai Hähnle at 2025-08-12T12:46:50+02:00
util: add piglit_is_file_older_than helper
This will be used by shader_runner to check whether SPIRV object files
are up-to-date.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
21d9c40c by Neil Roberts at 2025-08-12T12:46:50+02:00
shader_runner/spirv: Load a replacement .shader_test
This adds a new new command line option "-spirv". If no SPIRV line is
found in the require section and "-spirv" is passed to the command
line, then it will try to load a replacement script by appending .spv
to the name of the file.
In that way, now "SPIRV YES" meaning changes slightly, from "Use by
default the SPIR-V on this shader test" to "We have a SPIR-V on this
shader, by default use it without looking for a external replacement
file".
We also added a "SPIRV NO" line that points that the shader is not
compatible with SPIR-V, so would be skipped even using -spirv.
"-spirv" is the counter-part command line option to "-glsl". The new
"-spirv" option is used to try to run tests on SPIR-V mode even if
they doesn't have a SPIR-V included, while "-glsl" is for tests that
even having SPIR-V included, we want them to run on GLSL mode.
Obviously it is a mistake to use "-spirv" and "-glsl" at the same
time.
v2: update on the code and commit message due the addition of the
"-glsl" command line option, plus command line options
Signed-off-by: Neil Roberts <[email protected]>
Signed-off-by: Alejandro Piñeiro <[email protected]>
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
b2a1a12e by Alejandro Piñeiro at 2025-08-12T12:46:50+02:00
framework: add --spirv option
So when executing shader tests, they will be executed with -spirv.
So for example:
./piglit run tests/shader.py -t ARB_explicit_uniform_location --spirv results/explicit_uniform_location-spirv
Will run all the shader tests for that extension using -spirv.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
e653ac2d by Neil Roberts at 2025-08-12T12:46:50+02:00
generated_tests: Add script to translate shader_tests to SPIR-V
Utility script which translates GLSL shader_tests to OpenGL SPIR-V
shader_tests (ARB_gl_spirv), using glslang and SPIR-V tools.
The translation to SPIR-V is done using glslangValidator with the
--auto-map-locations option.
Names for uniforms, UBOs and vertex attributes in the [test] and
[vertex data] sections are remapped. In order to do that, the
generated SPIR-V is parsed to get the locations, bindings, offsets,
matrix stride and row major, array indices and the tree of types.
--aml can’t handle allocating the same varying locations across
multiple stages because it only sees the shaders for one stage at a
time. To work around this the script reads overwrites the locations
for the inputs in the generated SPIR-V with the locations with the
same name in the outputs from the previous stage.
Based on Nicolai Hänle's work who wrote the script when glslang did
not provide the --auto-map-location option and the GLSL code was
parsed by the script. The script was modified to use such option,
remove the GLSL parsing and implement the remapping of uniforms, UBOs
and vertex attributes, add more command line options and other
improvements.
Signed-off-by: Nicolai Hänle <[email protected]>
Signed-off-by: Alejandro Piñeiro <[email protected]>
Signed-off-by: Neil Roberts <[email protected]>
Signed-off-by: Antia Puentes <[email protected]>
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
8f5de67d by Antia Puentes at 2025-08-12T12:47:00+02:00
arb_gl_spirv: add gl_spirv.exclude file
Contains the list GLSL shader_tests which can not be translated to
SPIR-V or we are not interested in running as part of the ARB_gl_spirv
testing.
The tests are grouped by reason of exclusion, which may be:
* The translation to SPIR-V failed or is incorrect. Due to a glslang
limitation or an error in the GLSL shader which is detected by
glslang. The later can happen for linking tests which expect a linking
error to happen as glslang is able to detect intrastage linking
errors.
* The translation to SPIR-V succeeded but what the test wants to check
does not make sense for ARB_gl_spirv. For example, if the test makes
checks based on names (names are optional in SPIR-V) or the
shader_test expects an interstage linking error (glslang does not cope
with that and in ARB_gl_spirv a GLSL-like validation is not needed).
Based on previous work done by Nicolai Hähnle, who wrote the first
version of the file.
Several people have contributed:
Signed-off-by: Nicolai Hänle <[email protected]>
Signed-off-by: Antia Puentes <[email protected]>
Signed-off-by: Alejandro Piñeiro <[email protected]>
Signed-off-by: Neil Roberts <[email protected]>
Signed-off-by: Eleni Maria Stea <[email protected]>
Signed-off-by: Arcady Goldmints-Orlov <[email protected]>
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
85c72d17 by Nicolai Hähnle at 2025-08-12T12:47:11+02:00
gitignore: add *.spv
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
e957e8f3 by Neil Roberts at 2025-08-12T12:47:11+02:00
generated_tests/builtin_function: Add templates for equivalent SPIR-V
The test signatures now have an additional member called
template_spirv. If this is not None then it is a template which can be
used to generate an equivalent line of SPIR-V assembly.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
79d7ca47 by Neil Roberts at 2025-08-12T13:04:19+02:00
generated_tests: Generate SPIR-V equivalents for builtin tests
Adds a script based on gen_builtin_uniform_tests.py which directly
generates the equivalent SPIR-V assembly to test with the ARB_gl_spirv
extension. The main reason to do this rather than converting the GLSL
tests with glslang is that not all of the GLSL builtins have direct
equivalents in SPIR-V. For example most of them only work when the
operands all have the same type. If we tested different combinations
of types we’d just be testing glslang’s ability to add the opcodes
needed to promote the arguments. Additionally it’s nice not to depend
on glslang which tends to change a lot between versions.
Note that the script has the code needed to generate tests for vertex,
geometry, tessellation, fragment and compute shaders but everything
but the fragment shaders is commented out. This reduces the number of
test cases to a fifth. The assumption is that there probably isn’t
that much difference between the codepaths taken for the builtins in
the driver for the different stages so it’s not really worth testing
all of them.
v2: stop to use OriginLowerLeft always, as it is only valid for
fragment shaders, and mesa has recently added an assert for it
(Alejandro)
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
80a1c162 by Alejandro Piñeiro at 2025-08-12T13:10:39+02:00
generated_tests: remove extra space before #extension
glslang doesn't like those spaces, failing to generate SPIR-V.
Additionally this improves readability of the generated shader
(although this is subjective).
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
095176eb by Alejandro Piñeiro at 2025-08-12T13:10:39+02:00
shader_runner/spirv: log skip reason on ARB_gl_spirv extension check
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
d91380ac by Alejandro Piñeiro at 2025-08-12T13:10:39+02:00
shader_runner/spirv: check if test really includes spirv
SPIRV YES/ONLY assumes that the test includes the SPIR-V shader, and
that you should use it. This commit checks that this really happens,
and prints a proper error message if that is not the case.
Without this commit, a SPIRV YES/ONLY test that by mistake didn't
include the [spirv xx] sections would be just skipped with the
following non-intuitive message:
"Function "glDeleteProgramsARB" not supported on this implementation"
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
bd9c4878 by Alejandro Piñeiro at 2025-08-12T13:10:39+02:00
shader_runner/spirv: add check for SPV extension support
This only affects the case where SPIR-V shaders are run, specially
when using tests coming from other extensions.
In several cases, OpenGL supports an extension with a given
functionality. And such functionality it is not supported on core
SPIR-V but on a equivalent SPV functionality.
But it is not mandatory to support such SPV extension even if you
support the equivalent OpenGL one.
So for that case, we need to check for the equivalent SPV extension
and check if it is supported.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
7f38d003 by Alejandro Piñeiro at 2025-08-12T13:10:39+02:00
shader_runner/spirv: skip for GL extensions without SPIR-V equivalent
There are some GL extensions/features, like shader subroutines, that
doesn't have their equivalent on SPIR-V, so those tests should be
skipped on SPIR-V mode. We reuse the code that checks if a given
SPIR-V extension equivalent to a GL extension is supported, by setting
the mapping to a string that will never be the name of a SPIR-V
extension.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
8c1d8ad5 by Alejandro Piñeiro at 2025-08-12T13:10:39+02:00
shader_runner/spirv: skip compat profile tests on SPIR-V mode
>From ARB_gl_spirv spec:
"Changes to Chapter 3 of the OpenGL Shading Language 4.50 Specification
(Basics)
After describing the compatibility profile rules, add:
"Compatibility-profile features are not available when
generating SPIR-V.""
With this commit, tests that require a compatibility profile are
skipped if trying to be run on SPIR-V mode. This includes if we
specify force spirv on the command line or if the test includes a
SPIRV YES. The latter can happen as the GLSL from those tests can be
compatible, so glslang (or any other GLSL->SPIR-V backend) could be
able to generate the SPIRV.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
e9f8f62a by Antia Puentes at 2025-08-12T13:10:39+02:00
gen_gl_spirv_tests: mark shader_draw_params builtins as good
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
14a746bc by Antia Puentes at 2025-08-12T13:10:39+02:00
shader_runner: add a multidrawarrays command
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
84fc76cc by Antia Puentes at 2025-08-12T13:10:39+02:00
arb_spirv_extensions: Add tests for shader_draw_parameters
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
0287f389 by Alejandro Piñeiro at 2025-08-12T13:10:39+02:00
arb_gl_spirv: add one geometry xfb shader test
For a barebone test, just one test would be enough. We have enough VS
tests to check the type combination when capturing. This test is
mostly to check that GS works, specially with more than one
invocation.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
308b37c8 by Antia Puentes at 2025-08-12T13:10:39+02:00
arb_spirv_extensions: add tests for post_depth_coverage
Adds tests for SPV_KHR_post_depth_coverage.
Based on the existing tests for ARB_post_depth_coverage but
using SPIR-V shaders instead.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
94004a6f by Antia Puentes at 2025-08-12T13:10:39+02:00
arb_gl_spirv: Added tests for glGetProgramiv()
The SPIR-V used does not contain names. As names are optional in
SPIR-V this is useful to test the behaviour of the name related
queries when no name reflection is available.
Acked-by: Marek Olšák <[email protected]>
Acked-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1>
- - - - -
d5dbbc12 by Guilherme Gallo at 2025-08-12T13:34:51+00:00
replay: Calculate the chunk size via etag suffix
On S3, a '-' in the ETag header indicates a multipart upload, with the
number on the right representing the number of chunks used for the ETag
calculation. Although we usually use a 5 MB chunk size, this may differ
depending on the upload process.
Let's calculate the chunk size using the metadata when available,
falling back to the default size if not specified.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1030>
- - - - -
28d13498 by Vinson Lee at 2025-08-18T09:58:43-07:00
util: Fix build on macOS
tests/util/piglit-util.c:554:13: error: no member named 'st_mtim' in 'struct stat'
554 | if (stat_a.st_mtim.tv_sec < stat_b.st_mtim.tv_sec ||
| ~~~~~~ ^
tests/util/piglit-util.c:554:37: error: no member named 'st_mtim' in 'struct stat'
554 | if (stat_a.st_mtim.tv_sec < stat_b.st_mtim.tv_sec ||
| ~~~~~~ ^
tests/util/piglit-util.c:555:14: error: no member named 'st_mtim' in 'struct stat'
555 | (stat_a.st_mtim.tv_sec == stat_b.st_mtim.tv_sec &&
| ~~~~~~ ^
tests/util/piglit-util.c:555:39: error: no member named 'st_mtim' in 'struct stat'
555 | (stat_a.st_mtim.tv_sec == stat_b.st_mtim.tv_sec &&
| ~~~~~~ ^
tests/util/piglit-util.c:556:14: error: no member named 'st_mtim' in 'struct stat'
556 | stat_a.st_mtim.tv_nsec < stat_b.st_mtim.tv_nsec))
| ~~~~~~ ^
tests/util/piglit-util.c:556:39: error: no member named 'st_mtim' in 'struct stat'
556 | stat_a.st_mtim.tv_nsec < stat_b.st_mtim.tv_nsec))
| ~~~~~~ ^
Fixes: 701843aa2 ("util: add piglit_is_file_older_than helper")
Signed-off-by: Vinson Lee <[email protected]>
Reviewed-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1031>
- - - - -
517270cc by Randolph Sapp at 2025-09-05T13:23:10-05:00
framework: add and use TestPlaceholder objects
The actual Test classes are not that big, but with deqp-vk creating
around 2800 instances currently, we quickly consume around 3GB of ram
before beginning any tests. Given that a lot of these objects may not
even be used, we can reduce memory overhead by using a placeholder
namedtuple that is expanded to a full Test instance when requested from
the TestDict class. This cuts the initial memory usage back to 1.6 GB.
This does add a small lookup penalty, but given the lookup penalty is
still smaller that the time it takes to create an instance of a Test, it
results in a net improvement for most subsets of deqp tests. Not to
mention this penalty can be split between threads, unlike previously
when it was occurring in the single-threadded profile creation step.
The following data was collected using a subset of 3151 deqp-vk tests:
Threads Test Execution Time (s) TestPlaceholder Execution Time (s)
------- ----------------------- ----------------------------------
1 2613.93 2599.95
8 413.31 400.01
32 299.75 286.73
Reviewed-by: Christian Gmeiner <[email protected]>
Signed-off-by: Randolph Sapp <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1032>
- - - - -
705526fd by Juan A. Suarez Romero at 2025-09-12T23:53:55+02:00
glsl-1.40: require GL 4.5 for the two-stages test
This test is requiring GL >= 3.1, but then it declares to use GL 4.5, so
that makes the test failing for drivers supporting GL 3.1 but not 4.5.
We rely on using interface blocks and bindings, so we can't just declare
a 3.1 version.
So instead, let's require GL 4.5 instead, so those drivers just skip
this instead of failing.
Fixes: 2cc580621 (glsl-1.40: add a UBO test that I don't remember why I needed it)
Reviewed-by: Marek Olšák <[email protected]>
Signed-off-by: Juan A. Suarez Romero <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1034>
- - - - -
a70c3304 by Timothy Arceri at 2025-09-18T23:46:29+00:00
glsl-1.50: test ubo inteface-name mismatch
This test is designed to make sure mesas varying optimisation pass
correctly handles this mismatch by triggering the bug described in
https://gitlab.freedesktop.org/mesa/mesa/-/issues/13254
Reviewed-by: Alejandro Piñeiro <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1037>
- - - - -
4147e9d7 by Vinson Lee at 2025-10-07T21:35:14-07:00
egl: Fix create_wayland_display memory leak
Fix defect reported by Coverity Scan.
Resource leak (RESOURCE_LEAK)
leaked_storage: Variable d going out of scope leaks the storage it
points to.
Fixes: 2576be194584 ("tests: add tests for fixed-rate compression extensions")
Signed-off-by: Vinson Lee <[email protected]>
Reviewed-by: Erik Faye-Lund <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1039>
- - - - -
024b3349 by Ashley Smith at 2025-10-09T14:04:42+01:00
ext_shader_realtime_clock: Fix version number
GL version numbers (1.3) are different to GLSL version numbers 1.30.
Fixes: a0a27e52 ("tests: Add test for GL_EXT_shader_realtime_clock")
Signed-off-by: Ashley Smith <[email protected]>
- - - - -
f09abb0c by Vinson Lee at 2025-10-18T10:48:15-07:00
egl: Check result of create_wayland_display calloc
Suggested-by: Erik Faye-Lund <[email protected]>
Signed-off-by: Vinson Lee <[email protected]>
Reviewed-by: Erik Faye-Lund <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1041>
- - - - -
aea61f38 by Vinson Lee at 2025-10-18T18:25:04+00:00
modifiers: Fix out-of-bounds write
Fix defect reported by Coverity Scan.
Out-of-bounds write (OVERRUN)
sscanf_overrun: sscanf will overrun its argument mod_str which can
accommodate 4096 bytes. The number of bytes written may be 4097 bytes,
including the terminating null. Corresponding format specifier is:
%4096s.
Fixes: 1437619bb4cc ("modifiers: Allow specifying format and modifier")
Signed-off-by: Vinson Lee <[email protected]>
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1038>
- - - - -
2ac68e5f by Vinson Lee at 2025-10-18T18:35:34+00:00
util: Remove piglit_half_from_float dead code
If new_exp < -14, then exp_val > 0.
Fixes defect reported by Coverity Scan.
Logically dead code (DEADCODE)
dead_error_begin: Execution cannot reach this statement: case 0U:.
Fixes: 1625291c3f7c ("draw-vertices-half-float: new test for GL_ARB_half_float_vertex")
Signed-off-by: Vinson Lee <[email protected]>
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1040>
- - - - -
a68d0fec by David Rosca at 2025-10-20T06:45:58+00:00
ext_external_objects: Fix ping-pong test with vk and gl timeline semaphores
Change initial vk signal value to 6, otherwise vk and gl would signal
the same value.
Don't signal any semaphores from last submit to keep the timeline values
correct for second test run. Add wait for gl frame done to last vk submit.
Fixes: 307511f18 ("ext_external_objects: add NV_timeline_support to ping-pong test")
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1035>
- - - - -
ec76cc7a by Alejandro Piñeiro at 2025-10-24T19:15:09+02:00
arb_gpu_shader_fp64: print why layout tests are skipped
Those tests are skipped if GL_MAX_VERTEX_SHADER_STORAGE_BLOCKS < 1,
but doesn't print any info about it.
Transform feedback tests for this extension has similar messages for
tessellation related limits.
Reviewed-by: Marek Ol¨ák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1043>
- - - - -
5309e340 by Erik Faye-Lund at 2025-11-04T09:25:30+01:00
ext_gpu_shader4: require ext_texture_buffer_object
texelFetchBuffer is only available in GLSL if *both* GL_EXT_gpu_shader4
*and* GL_EXT_texture_buffer_object are supported.
>From the GL_EXT_gpu_shader4 spec:
> Interactions with EXT_texture_buffer_object
>
> If EXT_texture_buffer_object is not supported, references to buffer
> textures, as well as the texelFetchBuffer and texelSizeBuffer lookup
> functions and samplerBuffer types, need to be removed.
Fixes: 346d9a7a9 ("ext_gpu_shader4: add some texture buffer object tests")
Reviewed-by: Daniel Stone <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1047>
- - - - -
265582c0 by Marek Olšák at 2025-11-08T20:23:05+00:00
registry: update gl.xml
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1044>
- - - - -
6c81236b by Marek Olšák at 2025-11-09T10:07:46-05:00
small-prim: new demo for manual testing of small prim culling
I used this for development of small prim culling code in AMD drivers.
It's not an automated test.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
22cdbdab by Marek Olšák at 2025-11-09T10:07:46-05:00
pixel-rate: minor changes for MSAA
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
b15c00d7 by Marek Olšák at 2025-11-09T10:07:46-05:00
pixel-rate: rewrite measuring to record queries first, then get results
This is 100x faster.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
a79eb8b2 by Marek Olšák at 2025-11-09T10:07:46-05:00
shader-io-rate: don't align vertex_size to vec4
it was an AMD-specific workaround skewing the results
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
9801a704 by Marek Olšák at 2025-11-09T10:07:46-05:00
shader-io-rate: test TCS patch output throughput
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
3749ae19 by Marek Olšák at 2025-11-09T10:07:46-05:00
shader-io-rate: add an option to print 1 result per line instead of 3 results
for easy pasting into spreadsheets
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
2592f724 by Marek Olšák at 2025-11-09T10:07:46-05:00
draw-prim-rate: restructure code to get ready for faster measurements
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
845083ad by Marek Olšák at 2025-11-09T10:07:46-05:00
draw-prim-rate: remove unused VBO duplication code
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
f7e92963 by Marek Olšák at 2025-11-09T10:07:46-05:00
draw-prim-rate: rewrite part 1 for faster measurements
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
2f5a7d63 by Marek Olšák at 2025-11-09T10:07:46-05:00
draw-prim-rate: rework quad_sizes_in_pixels
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
13c57fa6 by Marek Olšák at 2025-11-09T10:07:46-05:00
draw-prim-rate: reduce memory usage by not creating identical BOs
the buffer set index uniquely identifies buffer contents.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
f098a932 by Marek Olšák at 2025-11-09T10:07:46-05:00
draw-prim-rate: rewrite measuring to record queries first, then get results
this is 100x faster
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
a6076fba by Marek Olšák at 2025-11-09T10:07:46-05:00
draw-prim-rate: remove unnecessary global variables
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
e0e22661 by Marek Olšák at 2025-11-09T10:07:46-05:00
draw-prim-rate: restructure code to inline execute_test
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
9bb8d5c1 by Marek Olšák at 2025-11-09T10:07:46-05:00
draw-prim-rate: print useful things
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
3550216e by Marek Olšák at 2025-11-09T10:07:46-05:00
draw-prim-rate: sink the loop in run_draw
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
e0796af6 by Marek Olšák at 2025-11-09T10:07:46-05:00
draw-prim-rate: generate VS, FS code programatically
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
29f12610 by Marek Olšák at 2025-11-09T10:34:30-05:00
draw-prim-rate: test mesh shader perf
linting is disabled because it's difficult to use and isn't useful to
the project. We need better standards for what we put in the CI.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
a269c72f by Marek Olšák at 2025-11-09T10:35:06-05:00
draw-prim-rate: fix MSVC build by moving ceil+sqrt out of a const expression
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1045>
- - - - -
00f7ea59 by Marek Olšák at 2025-11-09T12:16:41-05:00
shader_runner: allow enabling GL_CULL_FACE from shader tests
This has been useful to me.
Reviewed-by: Alejandro Piñeiro <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1050>
- - - - -
b38d533e by Timothy Arceri at 2025-11-09T22:37:20+00:00
glx: test shared shaders uniforms are flushed
This tests that uniforms of shared shaders are flushed properly
when updated by multiple contexts.
Mesa issue: https://gitlab.freedesktop.org/mesa/mesa/-/issues/14129
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1048>
- - - - -
1c505d31 by Eric R. Smith at 2025-11-12T11:48:08+00:00
glslparser: added test for accidental token pasting
If after macro expansion two identifiers would appear one
after the other, there should be a space between them so
that they aren't mistaken for one identifier. That is,
token pasting should only happen with a deliberate `##`,
not by accident. Both clang and gcc insert a space in the
test here, but Mesa did not (pending a merge request to
fix this).
Reviewed-by: Erik Faye-Lund <[email protected]>
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1033>
- - - - -
5330c708 by Christian Gmeiner at 2025-11-14T11:22:37+01:00
vertex-attrib-pointer-type-size-match: check for GL_EXT_vertex_array_bgra
The test was unconditionally using GL_BGRA as a size parameter without
checking for GL_EXT_vertex_array_bgra extension support.
GL_ARB_vertex_type_2_10_10_10_rev introduces the packed vertex formats
(INT_2_10_10_10_REV and UNSIGNED_INT_2_10_10_10_REV), but the ability
to use GL_BGRA as a size parameter is a separate feature provided by
GL_EXT_vertex_array_bgra.
The ARB_vertex_type_2_10_10_10_rev spec explicitly states:
"If EXT_vertex_array_bgra is not supported, remove references to BGRA
as an allowed parameter for <size> in VertexAttrib."
This fix ensures GL_BGRA is only tested when the extension is actually
supported, preventing false failures on implementations that support
the packed formats but not the BGRA size extension.
Signed-off-by: Christian Gmeiner <[email protected]>
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1051>
- - - - -
4a606082 by Ryan Mckeever at 2025-11-17T18:13:04+00:00
tests: add EXT_shader_pixel_local_storage tests
This commit adds the following tests of pixel_local_storage features:
- check that EXT_shader_pixel_local_storage is disabled in initial state
- render to a pixel_local_storage variable when extension is disabled
- use pixel_local_storage in shaders without the required #extension
- use pixel_local_storage in a vertex shader
- valid shader with named pixel_local_storage block
- valid shader with anonymous pixel_local_storage block
- write to both outputs and pixel_local_storage variables
- write to outputs and read pixel_local_storage variables
- write to read-only pixel_local_storage variables
- read write-only pixel_local_storage variables
- declare an inout pixel_local_storage block with an input and output
variable
- declare a pixel_local_storage variable outside a block
- declare a pixel_local_storage variable inside function
- declare a shader with multiple input pixel_local_storage blocks
- declare a shader with multiple output pixel_local_storage blocks
- declare a shader with multiple inout pixel_local_storage blocks
- declare a shader with 1 input and 1 output pixel_local_storage block
- pixel_local_storage block with redundant qualifiers
- pixel_local_storage block with inconsistent qualifiers
- pixel_local_storage block with inconsistent formats
- pixel_local_storage variables override block layout
- check for required features of max size and max fast size
- declare a pixel_local_storage block of exactly
MAX_SHADER_PIXEL_LOCAL_STORAGE_SIZE_EXT bytes
- declare a pixel_local_storage block larger than
MAX_SHADER_PIXEL_LOCAL_STORAGE_SIZE_EXT bytes
- check for required features of using draw framebuffers with
pixel_local_storage
- enable pixel_local_storage when SAMPLE_BUFFERS == 1
- check if blending or dithering changes pixel_local_storage output
- use initializers in pixel_local_storage
- write to and read from pixel_local_storage variables with each of the
supported layout qualifiers and check precision/accuracy
- read from color buffer after clearing
Acked-by: Erik Faye-Lund <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1036>
- - - - -
877f4214 by Ryan Mckeever at 2025-11-18T16:58:24-08:00
tests: Skip EXT_shader_pixel_local_storage tests when not enabled
Fixes: 4a606082 ("tests: add EXT_shader_pixel_local_storage tests")
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1053>
- - - - -
50d66611 by Timothy Arceri at 2025-11-20T23:36:22+00:00
gl-1.0: test texture matrix push pop
Tests the bug in mesa bug report:
https://gitlab.freedesktop.org/mesa/mesa/-/issues/14286
Reviewed-by: Emma Anholt <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1052>
- - - - -
5698a63e by Ryan Mckeever at 2025-11-25T12:42:16-08:00
pixel local storage: update read_after_clear test so it passes with -fbo
This commit changes the pls variable to input-only and conditionally
sets the color of an output variable depending on the color of the pls
variable.
Fixes: 4a606082 ("tests: add EXT_shader_pixel_local_storage tests")
Reviewed-by: Erik Faye-Lund <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1054>
- - - - -
cddbaaea by Ryan Mckeever at 2025-11-25T12:42:33-08:00
pixel_local_storage: use unique names for inconsistent formats tests
Fixes: 4a606082 ("tests: add EXT_shader_pixel_local_storage tests")
Reviewed-by: Erik Faye-Lund <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1054>
- - - - -
be54f1af by Eric R. Smith at 2025-11-25T12:42:41-08:00
pixel local storage: make some format tests use unnamed structures
Pixel local storage variables can be declared in two different ways:
__pixel_localEXT mydata1 {
vec4 color1;
vec4 val1;
} pls1;
__pixel_localEXT mydata2 {
vec4 color2;
vec4 val2;
};
In the first way the variables are referred to in later code as
pls1.color1 and pls1.val1, but in the second way they are just color2
and val2. The final generated code at the device end should be the same
for both cases, but internally the compiler handles them quite differently
(the first is a single variable that is a struct with fields, the second
has two separate variables that are vec4). This commit makes sure that
both paths through the compiler are tested.
Reviewed-by: Erik Faye-Lund <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1054>
- - - - -
096719f3 by Tapani Pälli at 2025-11-27T18:29:27+00:00
glsl-1.50: add tests for invalid interface with bool type
Signed-off-by: Tapani Pälli <[email protected]>
Acked-by: Emma Anholt <[email protected]>
Reviewed-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1055>
- - - - -
62d17d67 by Ryan Mckeever at 2025-12-04T18:08:51-08:00
pixel local storage: change format/type of texture in run_check_format
Since this test reads outColor, which is a float, rather than pls
variables, we need to set up the texture attached to the framebuffer with
GL_RGBA format and GL_UNSIGNED_BYTE type to match the values that
piglit_probe_rect_rgb() uses.
Fixes: 4a606082 ("tests: add EXT_shader_pixel_local_storage tests")
Reviewed-by: Eric R. Smith <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1058>
- - - - -
3a2ac019 by Ryan Mckeever at 2025-12-04T18:09:58-08:00
pixel local storage: only bind the default framebuffer during test cleanup
This ensures that tests only use the framebuffers generated within the
test and the default framebuffer is restored after each test.
Fixes: 4a606082 ("tests: add EXT_shader_pixel_local_storage tests")
Reviewed-by: Eric R. Smith <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1058>
- - - - -
2842979e by Pierre-Eric Pelloux-Prayer at 2025-12-11T13:46:14+01:00
gl-1.0: fix texture-matrix-push-pop with "-fbo"
Use a texture id generated by glGenTextures to avoid bad
interactions with the FBO's texture.
While at it insert a call to piglit_present_results so the
draws effect are visible.
Fixes: 50d666112 ("gl-1.0: test texture matrix push pop")
Reviewed-by: Erik Faye-Lund <[email protected]>
Reviewed-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1059>
- - - - -
8d9e8231 by Omar Pakker at 2025-12-15T15:03:15+01:00
Fix snprintf call in pixel-local-tests.c
This resolves a format-security warning introduced by 4a606082,
which results in build errors on systems that default to -Werror=format-security.
Reviewed-by: Erik Faye-Lund <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1057>
- - - - -
62d499d6 by Pierre-Eric Pelloux-Prayer at 2025-12-19T15:11:05+01:00
arb_texture_view: add reproducer for #13527
Based on clear-into-view-2d-array.c and the reproducer attached
to the bug report.
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1060>
- - - - -
5b0642ac by Eric Engestrom at 2026-01-26T21:04:48+01:00
tox.ini: limit numpy to working versions
2.4 removed a function parameter that we use; someone else's MR will fix
that, for now we just prevent the breakage.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1066>
- - - - -
50c4d756 by Eric Engestrom at 2026-01-26T21:23:16+01:00
ci/debian: install all the python packages from pip and pin their version
Pinning them to a specific version is to ensure we can have a stable CI,
as it avoids everything breaking because someone regenerated the debian
image.
To note, numpy is pinned at the last version <2.4; see explanation in
the previous commit.
A few python packages had to be left in debian's care, namely packaging,
psutil and wheel, because thing were breaking too much.
A clean solution would be to setup a venv but I'll leave that as future
improvement, I'm focusing on just unbreaking CI right now 😅
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1066>
- - - - -
0d3dcb1c by Eric Engestrom at 2026-01-26T21:42:25+01:00
unittests/framework: autouse the setup fixture
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1066>
- - - - -
2477544a by Eric Engestrom at 2026-01-26T21:17:07+00:00
ci: avoid duplicate fork pipeline in MRs
Copied from Mesa CI.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1069>
- - - - -
fae2f6ce by Mel Henning at 2026-01-26T21:40:11+00:00
ext_msrtt: Add a reproducer for #14647
Reviewed-by: Mike Blumenkrantz <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1065>
- - - - -
50b8b67d by Eric Engestrom at 2026-01-29T12:45:53+01:00
tox.ini: simplify envlist & deps definition
These envs need to be created for every python version, the constraint
doesn't provide anything other than extra work every year when a new
python version comes out.
As for the deps, they apply to all python versions, so there is no point
listing them, same as all the other deps in this file.
The "special" case here is the `numpy2` factor which installs numpy>2
while `generator` installs numpy<2, but the constraint skipping
explicitly numpy2 on python 3.8 doesn't need to be repeated here, it's
already expressed in numpy itself and we won't hit it anyway as we don't
try to run the `numpy2` factor on python 3.8 in our CI, and a user
running this locally will get the nicer pip error about numpy>2 not
being available on python<3.9, instead of a runtime error in pytest when
importing numpy because its install was skipped.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1071>
- - - - -
095e16e9 by Eric Engestrom at 2026-01-29T12:51:28+00:00
tests/create-kernels-in-program: move free(kernels) after we're done reading kernels
Fixes: 13d744e7c3cf04f41529 ("cl: Add OpenCL tests")
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1070>
- - - - -
a79e32b7 by Eric Engestrom at 2026-01-29T13:32:12+00:00
tests/util-cl: fix free of incorrect pointer level
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1067>
- - - - -
89171ff6 by Christian Gmeiner at 2026-01-29T15:33:38+00:00
cmake: Add Python 3.14 to supported versions
And test it in CI.
Signed-off-by: Christian Gmeiner <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1049>
- - - - -
d1350bed by Eric Engestrom at 2026-01-29T19:44:12+01:00
ci: drop redundant py.test job
This is a subset of what the tox jobs already run, this adds no value, so let's drop it.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1074>
- - - - -
b6a6e188 by Eric Engestrom at 2026-01-29T19:59:39+01:00
ci: fix indentation
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1073>
- - - - -
701eb174 by Eric Engestrom at 2026-01-29T19:59:39+01:00
ci: be more explicit about the different pipelines
In particular, this does not include post-merge pipelines anymore
(previously the last rule created one) because it is not useful in this
project and doubles the CI resource usage with no benefit.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1073>
- - - - -
0458c8f8 by Eric Engestrom at 2026-01-29T20:05:40+01:00
ci: use priority tags to make Marge faster and be good citizens in fork pipelines
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1073>
- - - - -
a1a46f81 by Eric Engestrom at 2026-01-30T08:13:15+00:00
tests/ext_direct_state_access: fix pointer types
Fixes build failures on mingw32 that look like:
```
build/mingw32/tests/util/piglit-dispatch-gen.h:6685:31: error: initialization of 'void (*)(GLenum, GLuint, GLfloat *)' {aka 'void (*)(unsigned int, unsigned int, float *)'} from incompatible pointer type 'PFNGLGETFLOATINDEXEDVEXTPROC' {aka 'void (__attribute__((stdcall)) *)(unsigned int, unsigned int, float *)'} [-Wincompatible-pointer-types]
6685 | #define glGetFloatIndexedvEXT piglit_dispatch_glGetFloatIndexedvEXT
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
tests/spec/ext_direct_state_access/indexed-state-queries.c:330:17: note: in expansion of macro 'glGetFloatIndexedvEXT'
330 | glGetFloatIndexedvEXT, "GetFloatIndexedvEXT", 12
| ^~~~~~~~~~~~~~~~~~~~~
build/mingw32/tests/util/piglit-dispatch-gen.h:6685:31: note: (near initialization for 'get_float_12.get_float_fn')
6685 | #define glGetFloatIndexedvEXT piglit_dispatch_glGetFloatIndexedvEXT
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
tests/spec/ext_direct_state_access/indexed-state-queries.c:330:17: note: in expansion of macro 'glGetFloatIndexedvEXT'
330 | glGetFloatIndexedvEXT, "GetFloatIndexedvEXT", 12
| ^~~~~~~~~~~~~~~~~~~~~
```
(Repeated twice for each of the 3 pointers)
Fixes: 3cd7ce2b44e3a093912b ("ext_direct_state_access: glGet*IndexedvEXT")
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1072>
- - - - -
e929d198 by Eric Engestrom at 2026-01-30T08:28:43+00:00
tests/ext_image_dma_buf_import: initialise image pixels
We don't care about the pixels for these tests, but the compiler doesn't
know that and errors out.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1068>
- - - - -
7a05aa17 by Eric Engestrom at 2026-01-30T11:24:55+00:00
ci/tox: drop python version selection
We run in `image: python:${PY_MAJVER}.${PY_MINVER}`, there can only be
one python version installed in there.
Makes the commands much easier to read and understand.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1077>
- - - - -
7a776c24 by Eric Engestrom at 2026-01-30T11:24:55+00:00
ci/tox: merge into a single tox command
That way there only one summary to read, instead of having to scroll up
to find the first one.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1077>
- - - - -
16b96155 by Ryan Mckeever at 2026-01-31T01:37:51+00:00
pixel local storage: explicitly specify format indices when using full list
This will make things more straightforward later when we expand the layout
qualifier tests to use PLS blocks with multiple variables.
Reviewed-by: Eric R. Smith <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1064>
- - - - -
cd7c06f8 by Ryan Mckeever at 2026-01-31T01:37:51+00:00
pixel local storage: use strncmp instead of strstr in make_vars_str
Reviewed-by: Eric R. Smith <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1064>
- - - - -
190005a6 by Ryan Mckeever at 2026-01-31T01:37:51+00:00
pixel local storage: allow specifying per-variable data in make_vars_str
Also adds documentation about the parameters and allows for pls_name to be
NULL.
Reviewed-by: Eric R. Smith <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1064>
- - - - -
815d84f9 by Ryan Mckeever at 2026-01-31T01:37:51+00:00
pixel local storage: use multiple PLS variables in layout qualifier tests
Previously these tests only used PLS blocks with one variable. Now they
use four variables per PLS block and test valid combinations of the
requested formats.
Reviewed-by: Eric R. Smith <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1064>
- - - - -
c1dd58ff by Trevor Gamblin at 2026-02-05T13:22:24+00:00
generated_tests: use 'shape' in place of 'newshape' on numpy>=2.1
The 'newshape' parameter was deprecated upstream but kept for backwards
compatibility reasons. It was finally removed as of numpy 2.4.0, so
transition to using 'shape' in its place when available.
Signed-off-by: Trevor Gamblin <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1063>
- - - - -
483b011e by Trevor Gamblin at 2026-02-05T13:22:24+00:00
tox: test numpy>=2.4
Signed-off-by: Trevor Gamblin <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1063>
- - - - -
ef0ee668 by Eric Engestrom at 2026-02-05T20:38:37+01:00
ci/debian: use array instead of word-split string for packages list
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1076>
- - - - -
1939e32f by Eric Engestrom at 2026-02-05T20:38:37+01:00
ci/debian: update from bullseye to trixie
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1076>
- - - - -
81e6505f by Lakshman Chandu Kondreddy at 2026-02-05T20:40:08+00:00
tests/egl: Add test for fp32 DMA-BUF image stores
Add a new test which validates DMA-BUF imported EGL images with
high-precision float formats (R32F,RG32F,RGBA32F) to work
correctly with compute shader image store write only operation.
Signed-off-by: Lakshman Chandu Kondreddy <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1061>
- - - - -
cc887283 by Eric Engestrom at 2026-02-05T21:26:08+00:00
ci: drop codespell from linting job
This is not a useful check and causes a lot of annoyance for developers
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1075>
- - - - -
bb52bae4 by Eric Engestrom at 2026-02-05T21:26:08+00:00
framework/replay: fix type annotation not including the default value
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1075>
- - - - -
fe77555c by Eric Engestrom at 2026-02-05T21:26:08+00:00
ci: restore linting job
It had been disabled in 29f12610715c0b7cb9c7 ("draw-prim-rate: test mesh
shader perf") because of `codespell` causing unreasonable friction for
developers.
Now that `codespell` is gone, let's restore it.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1075>
- - - - -
8df27345 by GKraats at 2026-02-06T19:32:51+00:00
Test fence register shortage at (i915) driver
Test that glBlitFramebufferEXT with GL_NEAREST does
not use too many fence registers at the blit,
causing loss of current batchbuffer and missing data
on the screen or causing an abort at an
assert, if compiled with debug.
Verifies MR:
https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/26769
Signed-off-by: GKraats <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/936>
- - - - -
0d79fb4a by Pavel Ondračka at 2026-02-13T11:54:51+01:00
pbo-getteximage: check for GL_EXT_texture_array
Only run GL_TEXTURE_1D_ARRAY and GL_TEXTURE_2D_ARRAY subtests when
texture arrays are supported (GL 3.0+ or GL_EXT_texture_array).
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1082>
- - - - -
59111996 by Pavel Ondračka at 2026-02-14T10:18:31+00:00
ext_packed_depth_stencil: accept INVALID_ENUM for TexStorage
For GL_DEPTH_STENCIL with TexStorage, ARB_texture_storage says:
"If <internalformat> is one of the internal formats listed in table
3.11, an INVALID_ENUM error is generated."
The test previously expected only INVALID_VALUE for the cube-map
TexStorage path when depth cube textures are not supported.
Mesa may return INVALID_ENUM first, because unsized-internalformat
validation can trigger before capability-specific validation.
Accept INVALID_ENUM as an alternate expected error in that branch.
Spec references:
- GL_ARB_texture_storage.txt (TexStorage errors, table 3.11 unsized formats)
- OpenGL 2.1 spec, section 3.8.1 (Texture Image Specification)
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1081>
- - - - -
1a1ae6b6 by Juan A. Suarez Romero at 2026-02-16T13:50:28+01:00
egl_khr_gl_image: request OpenGL 3.0
This extension requires an OpenGL client supporting render buffers. This
were introduced in OpenGL 3.0 core spec (before they were available
through GL_ARB_framebuffer_object extension).
So relax the OpenGL context created to 3.0.
Also, if the driver does not support OpenGL 3.0, then just skip the test
instead of fail.
Signed-off-by: Juan A. Suarez Romero <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1084>
- - - - -
c859cec4 by Emma Anholt at 2026-02-18T03:49:02+00:00
glx@glx_ext_no_config_context@no fbconfig: Don't leak a window per visual.
zink was having trouble with the X server dying during this test due
to running out of fds. This probably won't fix that, but we at least
shouldn't be leaking over the course of the test.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1078>
- - - - -
8e2c8bc0 by Maaz Mombasawala at 2026-02-18T04:17:30+00:00
egl_khr_fence_sync: Fix use-after-free error with gl_extensions array.
Testing the svga driver I am running into occasional use-after-free error
in egl_khr_fence_sync. The memory region for gl_extensions array sometimes
overlaps with memory used by mesa's internal structs, which is freed at the
end of a subtest. So when the next subtest runs and accesses the gl_extensions
array, strcmp() on that free region crashes the process.
Invalidating the gl_extensions array on test_cleanup() forces the next subtest
running to reinitialize the array, which avoids this issue.
Fixes egl_khr_fence_sync on svga.
Signed-off-by: Maaz Mombasawala <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/993>
- - - - -
6721f002 by Pavel Ondračka at 2026-02-19T10:43:26+01:00
arb_framebuffer_object: gate depth-stencil cube attachment test
Depth cubemaps are only supported with desktop OpenGL version >= 3.0
or EXT_gpu_shader4.
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1083>
- - - - -
84706218 by Pavel Ondračka at 2026-02-19T10:43:39+01:00
fbo-luminance-alpha: skip when blend support is not advertised
Use GL_ARB_internalformat_query2/GL_FRAMEBUFFER_BLEND to detect
whether GL_LUMINANCE8_ALPHA8 supports blending and skip when it
does not.
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1083>
- - - - -
daf73aaa by Pavel Ondračka at 2026-02-19T10:43:49+01:00
fbo-mipmap-copypix: skip non-renderable formats
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1083>
- - - - -
3051e13a by Pavel Ondračka at 2026-02-19T10:43:56+01:00
ext_image_dma_buf_import: log invalid modifier only once
ext_image_dma_buf_import-tex-modifier probes 2 formats over 7x7 sizes,
so on drivers returning DRM_FORMAT_MOD_INVALID this warning can be
printed 98 times in one test run.
Example CI log with this spam:
https://gitlab.freedesktop.org/mesa/mesa/-/jobs/93288084
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1083>
- - - - -
4d658330 by Vinson Lee at 2026-02-27T17:50:49-08:00
tests/util: fix float-to-integer narrowing in ColorGradientSunburst
Newer versions of Apple Clang (shipped with Xcode on arm64 macOS)
treat -Wc++11-narrowing as an error by default, unlike GCC and
older Clang which only emit a warning. This causes a build failure
in draw_with_scale_and_offset() where the float parameter 'offset'
is used to brace-initialize int and unsigned int arrays, which is
a narrowing conversion forbidden by C++11.
Fix by explicitly casting offset to the target type before using
it in the initializer list.
Fixes: 9c1b2acb6 ("Make the functions in ext_framebuffer_multisample/common.cpp available globally")
Signed-off-by: Vinson Lee <[email protected]>
Reviewed-by: Anuj Phogat <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1088>
- - - - -
b62f933b by Vinson Lee at 2026-02-28T10:11:21-08:00
cmake: skip -ffloat-store on ARM architectures
-ffloat-store is an x86-specific GCC flag that forces floating-point
values to be stored to memory after each expression to avoid x87
80-bit extended precision non-determinism. ARM has no x87 FPU and
uses IEEE 754 precision-matching registers, so the flag is meaningless
and produces a -Wignored-optimization-argument warning on Apple Clang.
Add an elseif for ARM/AArch64 to skip the flag entirely.
Fixes: c42eee7a8 ("cmake: add fallback for non-SSE2 builds")
Signed-off-by: Vinson Lee <[email protected]>
Reviewed-by: David Heidelberg <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1090>
- - - - -
ec0fc2fe by Vinson Lee at 2026-03-02T10:32:44+00:00
fbo-blit-stretch: rename local lerp to linear_interp
Avoid conflict with C++20 std::lerp from <cmath>, which
causes redefinition and ambiguous call errors with newer
toolchains.
Signed-off-by: Vinson Lee <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1079>
- - - - -
3513ef37 by Vinson Lee at 2026-03-02T23:05:49-08:00
shader_runner: initialize variables in default switch cases
Fix the following compiler warnings:
shader_runner.c:3686:3: warning: variable 'old_got' is used
uninitialized whenever switch default is taken
[-Wsometimes-uninitialized]
shader_runner.c:3879:3: warning: variable 'old_length' is used
uninitialized whenever switch default is taken
[-Wsometimes-uninitialized]
shader_runner.c:3918:3: warning: variable 'old' is used uninitialized
whenever switch default is taken [-Wsometimes-uninitialized]
shader_runner.c:3960:3: warning: variable 'old' is used uninitialized
whenever switch default is taken [-Wsometimes-uninitialized]
The surrounding if-guards make the default cases unreachable, but the
compiler cannot prove this. Initialize the variables in the default
cases to silence the warnings.
Fixes: 36642eaba ("shader_runner: Check compatibility for resource name queries")
Signed-off-by: Vinson Lee <[email protected]>
Reviewed-by: Alejandro Piñeiro <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1094>
- - - - -
d7b6f4b5 by Pavel Ondračka at 2026-03-04T10:03:16+01:00
glsl-1.10: test multiple render target out of order alpha writes
Acked-by: Marek Olšák <[email protected]>
Reviewed-by: Emma Anholt <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1091>
- - - - -
b7099198 by Vinson Lee at 2026-03-04T21:09:14-08:00
s3tc-errors: initialize format and compressed_size
Fix the following compiler warnings:
s3tc-errors.c:201:6: warning: variable 'format' is used uninitialized
whenever 'if' condition is false [-Wsometimes-uninitialized]
s3tc-errors.c:201:6: warning: variable 'compressed_size' is used
uninitialized whenever 'if' condition is false
[-Wsometimes-uninitialized]
The warnings are false positives. compressed_size is used at line 238
under #ifdef PIGLIT_USE_OPENGL, and when PIGLIT_USE_OPENGL is defined,
piglit_is_gles() returns false so the if guard at line 201 always
enters and the variables are always initialized. The compiler cannot
reason across the mix of #ifdef and runtime checks. Initialize both
variables to 0 at declaration to silence the warnings.
Signed-off-by: Vinson Lee <[email protected]>
Reviewed-by: Erik Faye-Lund <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1095>
- - - - -
d071c520 by Rhys Perry at 2026-03-05T15:06:12+00:00
tests: add tests for unwritten dual-source blending outputs
Signed-off-by: Rhys Perry <[email protected]>
Acked-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1087>
- - - - -
1c4e0f4b by Pavel Ondračka at 2026-03-06T10:11:26+00:00
glsl-1.10/fragdata-mrt-alpha-target: add missing require
It needs two draw buffers.
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1099>
- - - - -
9b8ed8ab by Vinson Lee at 2026-03-06T20:04:26-08:00
cl: fix printf format for 64-bit OpenCL types in get-device-info
cl_device_fp_config and cl_device_affinity_domain are cl_ulong which is
unsigned long long, but were printed with %lx which expects unsigned
long. Even though both types are 64-bit on LP64 platforms, they are
distinct types in C, causing a -Wformat warning:
format specifies type 'unsigned long' but the argument has type 'cl_device_fp_config' (aka 'unsigned long long')
Replace %lx with PRIx64 from inttypes.h. PRIx64 expands to the correct
platform-specific format specifier for uint64_t, which matches cl_ulong
since both are guaranteed 64-bit unsigned integers.
Fixes: 4f2016b2d9f3 ("cl: clGetDeviceInfo check returned values")
Fixes: ef502923eb3b ("cl: Correct CL_DEVICE_DOUBLE_FP_CONFIG check for OpenCL > 1.2")
Signed-off-by: Vinson Lee <[email protected]>
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1097>
- - - - -
d82f836a by Vinson Lee at 2026-03-07T05:15:27+00:00
fbo-integer-attachments: remove erroneous argument to glEnd()
Fix the following compiler warning:
tests/fbo/fbo-integer-attachments.c:286:9: warning: passing arguments
to a function without a prototype is deprecated in all versions of C
and is not supported in C23 [-Wdeprecated-non-prototype]
glEnd() takes no arguments. The GL_LINES primitive type is already
specified via glBegin(GL_LINES) on the preceding line.
Fixes: c32e5cc71 ("GL_EXT_texture_integer: basic validation tests")
Signed-off-by: Vinson Lee <[email protected]>
Reviewed-by: Marek Olšák <[email protected]>
Reviewed-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1093>
- - - - -
a482ca5e by Vinson Lee at 2026-03-07T05:42:10+00:00
amd_vertex_shader_layer: remove unused color_index variable
Fix the following compiler warning:
layered-depth-texture-render.c:89:6: warning: variable 'color_index'
set but not used [-Wunused-but-set-variable]
The variable is initialized and incremented but never read.
Fixes: 118ebd283 ("AMD_vertex_shader_layer: Add depth render test")
Signed-off-by: Vinson Lee <[email protected]>
Reviewed-by: Marek Olšák <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1096>
- - - - -
d0a16eee by Ian Romanick at 2026-03-09T17:35:51+00:00
arb_gpu_shader5: Saturate min(NaN, 1.0)
A very, very old optimization in the Intel backend will incorrectly
optimize this to just saturate(NaN). That produces 0.0 instead of the
expected 1.0.
Ref: mesa!40284
Reviewed-by: Caio Oliveira <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1100>
- - - - -
af7d2bea by Alejandro Piñeiro at 2026-03-13T18:16:40+01:00
arb_explicit_attrib_location: add test for the use of same location but different index
This was used to test a recent fix on mesa, that were passing the
linkage of the test layout-13 in all cases, when it should fail.
Reviewed-by: Timothy Arceri <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/985>
- - - - -
040ed91f by Timothy Arceri at 2026-03-13T23:33:44+00:00
arb_sampler_objects: test multiple sampler views
Tests binding both samplers to different texture units on the same
texture.
Mesa bug report:
https://gitlab.freedesktop.org/mesa/mesa/-/issues/15045
Reviewed-by: Mike Blumenkrantz <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1101>
- - - - -
d32d31b6 by Marek Olšák at 2026-03-23T18:06:31+00:00
khr_shader_subgroup: add Bitonic Sort test using subgroup shuffles
Part-of: <https://gitlab.freedesktop.org/mesa/piglit/-/merge_requests/1102>
- - - - -
941b2491 by Jordan Justen at 2026-03-24T16:33:36+00:00
Merge remote-tracking branch 'origin/main' into debian-unstable
- - - - -
d5cd7b9c by Jordan Justen at 2026-03-24T16:33:41+00:00
d/changelog: Start 0.0~git20260323.d32d31b6c-1 release changelog
Signed-off-by: Jordan Justen <[email protected]>
- - - - -
7eccd0ca by Jordan Justen at 2026-03-24T09:37:59-07:00
d/control: Drop python3-six dependency (Closes: #1068256)
Signed-off-by: Jordan Justen <[email protected]>
- - - - -
55d8c7bd by Jordan Justen at 2026-03-26T00:31:03-07:00
d/patches: Drop patch for change that was made upstream
Fixes: upstream cabe21dc6 ("fix invalid typecast")
Signed-off-by: Jordan Justen <[email protected]>
- - - - -
98162c73 by Jordan Justen at 2026-03-26T01:00:20-07:00
d/changelog: Release 0.0~git20260323.d32d31b6c-1 to unstable
Signed-off-by: Jordan Justen <[email protected]>
- - - - -
265 changed files:
- .gitignore
- .gitlab-ci.yml
- .gitlab-ci/debian-install.sh
- CMakeLists.txt
- README.md
- debian/changelog
- debian/control
- − debian/patches/0001-ext_external_objects-Fix-two-null-handle-misassignme.patch
- − debian/patches/series
- framework/options.py
- framework/profile.py
- framework/programs/run.py
- framework/replay/backends/angle.py
- framework/replay/download_utils.py
- framework/test/base.py
- framework/test/deqp.py
- framework/test/shader_test.py
- generated_tests/CMakeLists.txt
- generated_tests/builtin_function.py
- generated_tests/builtin_function_fp64.py
- + generated_tests/gen_builtin_uniform_spirv_tests.py
- + generated_tests/gen_gl_spirv_tests.py
- generated_tests/templates/gen_extensions_defined/disabled.glsl.mako
- generated_tests/templates/gen_extensions_defined/enabled.glsl.mako
- generated_tests/templates/gen_extensions_defined/undefined-require.glsl.mako
- generated_tests/templates/gen_vs_in_fp64/columns.shader_test.mako
- generated_tests/templates/gen_vs_in_fp64/regular.shader_test.mako
- registry/gl.xml
- tests/cl/api/create-kernels-in-program.c
- tests/cl/api/get-device-info.c
- tests/egl/CMakeLists.gles2.txt
- + tests/egl/egl-image-external-compute-writeonly-float32.c
- tests/egl/egl-wayland.c
- tests/egl/spec/CMakeLists.txt
- + tests/egl/spec/egl_ext_device_persistent_id/CMakeLists.no_api.txt
- + tests/egl/spec/egl_ext_device_persistent_id/CMakeLists.txt
- + tests/egl/spec/egl_ext_device_persistent_id/egl_ext_device_persistent_id.c
- + tests/egl/spec/egl_ext_device_query_name/CMakeLists.no_api.txt
- + tests/egl/spec/egl_ext_device_query_name/CMakeLists.txt
- + tests/egl/spec/egl_ext_device_query_name/egl_ext_device_query_name.c
- tests/egl/spec/egl_khr_fence_sync/egl_khr_fence_sync.c
- tests/egl/spec/egl_khr_gl_image/egl_khr_gl_renderbuffer_image-clear-shared-image.c
- tests/fbo/CMakeLists.gl.txt
- + tests/fbo/fbo-blit-fence-reg.c
- tests/fbo/fbo-blit-stretch.cpp
- + tests/fbo/fbo-integer-attachments.c
- tests/fbo/fbo-luminance-alpha.c
- tests/fbo/fbo-mipmap-copypix.c
- tests/find_static_tests.py
- tests/general/CMakeLists.gl.txt
- + tests/general/depth_clamp_quirks.c
- tests/general/fog-modes.c
- tests/general/masked-clear.c
- tests/general/pbo-getteximage.c
- + tests/general/push-pop-color-material.c
- tests/general/teximage-scale-bias.c
- + tests/gl_spirv.exclude
- + tests/glslparsertest/shaders/preprocess11.frag
- tests/glx/CMakeLists.gl.txt
- tests/glx/glx-multi-context-ib-1.c
- + tests/glx/glx-multi-context-ib-2.c
- tests/glx/glx-multi-context-single-window.c
- tests/glx/glx-multi-window-single-context.c
- + tests/glx/glx-multithread-buffer-busy-tracking-bug.c
- tests/glx/glx-multithread-texture.c
- tests/glx/glx-multithread.c
- + tests/glx/glx-shader-sharing-uniforms.c
- tests/opengl.py
- tests/perf/CMakeLists.gl.txt
- tests/perf/common.c
- tests/perf/common.h
- tests/perf/draw-prim-rate.c
- tests/perf/pixel-rate.c
- tests/perf/shader-io-rate.c
- + tests/perf/small-prim-filter.c
- tests/shaders/CMakeLists.gles3.txt
- + tests/shaders/mesa-bug11100-cs.c
- + tests/shaders/mesa-bug11100-fs.c
- tests/shaders/shader_runner.c
- − tests/shaders/ssa/fs-if-def-else-break.shader_test
- tests/spec/CMakeLists.txt
- tests/spec/amd_vertex_shader_layer/layered-depth-texture-render.c
- tests/spec/arb_blend_func_extended/execution/CMakeLists.gl.txt
- + tests/spec/arb_blend_func_extended/execution/dual-src-blending-undef-output.c
- tests/spec/arb_direct_state_access/namedrenderbuffer.c
- + tests/spec/arb_explicit_attrib_location/1.10/compiler/layout-15.frag
- tests/spec/arb_fragment_program/CMakeLists.gl.txt
- + tests/spec/arb_fragment_program/recompile.c
- tests/spec/arb_framebuffer_object/same-attachment-tex2d-depth_stencil.c
- + tests/spec/arb_gl_spirv/CMakeLists.gl.txt
- + tests/spec/arb_gl_spirv/CMakeLists.txt
- + tests/spec/arb_gl_spirv/execution/array_of_input_output_block.shader_test
- + tests/spec/arb_gl_spirv/execution/atomic-image.shader_test
- + tests/spec/arb_gl_spirv/execution/component-layout-sso-vs-gs-fs-array-interleave.shader_test
- + tests/spec/arb_gl_spirv/execution/component-layout-vs-gs-fs-double.shader_test
- + tests/spec/arb_gl_spirv/execution/component-layout-vs-gs-fs.shader_test
- + tests/spec/arb_gl_spirv/execution/cs-basic-ssbo.shader_test
- + tests/spec/arb_gl_spirv/execution/function-call-glslang-issue-1285.shader_test
- + tests/spec/arb_gl_spirv/execution/origin-lower-left.shader_test
- + tests/spec/arb_gl_spirv/execution/origin-upper-left.shader_test
- + tests/spec/arb_gl_spirv/execution/pixel-center-half.shader_test
- + tests/spec/arb_gl_spirv/execution/pixel-center-integer.shader_test
- + tests/spec/arb_gl_spirv/execution/queries/program-interface/aggregate-types-name-queries.shader_test
- + tests/spec/arb_gl_spirv/execution/queries/program-interface/aggregate-types-other-queries.shader_test
- + tests/spec/arb_gl_spirv/execution/queries/program-interface/atomic-counter-other-queries.shader_test
- + tests/spec/arb_gl_spirv/execution/queries/program-interface/simple-types-name-queries.shader_test
- + tests/spec/arb_gl_spirv/execution/queries/program-interface/simple-types-other-queries.shader_test
- + tests/spec/arb_gl_spirv/execution/queries/program-interface/ubo-matrices-other-queries.shader_test
- + tests/spec/arb_gl_spirv/execution/queries/program-interface/ubo-nested-structs-other-queries.shader_test
- + tests/spec/arb_gl_spirv/execution/queries/program-interface/xfb-name-queries.shader_test
- + tests/spec/arb_gl_spirv/execution/queries/program-interface/xfb-simple-other-queries.shader_test
- + tests/spec/arb_gl_spirv/execution/queries/program/max-length-queries.shader_test
- + tests/spec/arb_gl_spirv/execution/queries/program/max-length-xfb-query.shader_test
- + tests/spec/arb_gl_spirv/execution/queries/program/other-queries.shader_test
- + tests/spec/arb_gl_spirv/execution/shared-uint-atomic-mem-functions.shader_test
- + tests/spec/arb_gl_spirv/execution/simple-function.shader_test
- + tests/spec/arb_gl_spirv/execution/simple-geometry.shader_test
- + tests/spec/arb_gl_spirv/execution/sso-block-member-layout-location.shader_test
- + tests/spec/arb_gl_spirv/execution/tcs-patch-vertices-in.shader_test
- tests/spec/arb_gl_spirv/execution/uniform/atomic-uint-fs.shader_runner → tests/spec/arb_gl_spirv/execution/uniform/atomic-uint-fs.shader_test
- + tests/spec/arb_gl_spirv/execution/unordered-comparison.shader_test
- + tests/spec/arb_gl_spirv/execution/va64/array.shader_test
- + tests/spec/arb_gl_spirv/execution/va64/simple.shader_test
- + tests/spec/arb_gl_spirv/execution/varying/all-array-types.shader_test
- + tests/spec/arb_gl_spirv/execution/varying/all-data-types.shader_test
- + tests/spec/arb_gl_spirv/execution/varying/all-scalar.shader_test
- + tests/spec/arb_gl_spirv/execution/varying/all-stages-test.shader_test
- + tests/spec/arb_gl_spirv/execution/varying/block.shader_test
- + tests/spec/arb_gl_spirv/execution/varying/mat/mat-arb.shader_test
- + tests/spec/arb_gl_spirv/execution/varying/mat/mat2.shader_test
- + tests/spec/arb_gl_spirv/execution/varying/mat/mat3.shader_test
- + tests/spec/arb_gl_spirv/execution/varying/mat/mat4.shader_test
- + tests/spec/arb_gl_spirv/execution/vs-instance-id.shader_test
- + tests/spec/arb_gl_spirv/execution/vs-vertex-id.shader_test
- + tests/spec/arb_gl_spirv/execution/xfb/gs_two_invocations.shader_test
- + tests/spec/arb_gl_spirv/execution/xfb/vs_two_sets_mixing_order.shader_test
- + tests/spec/arb_gl_spirv/linker/simple-function-call.shader_test
- + tests/spec/arb_gl_spirv/sample-position.c
- + tests/spec/arb_gpu_shader5/execution/precise/saturate-min-nan.shader_test
- tests/spec/arb_gpu_shader_fp64/shader_storage/layout-std140-fp64-mixed-shader.c
- tests/spec/arb_gpu_shader_fp64/shader_storage/layout-std140-fp64-shader.c
- tests/spec/arb_gpu_shader_fp64/shader_storage/layout-std430-fp64-mixed-shader.c
- tests/spec/arb_gpu_shader_fp64/shader_storage/layout-std430-fp64-shader.c
- tests/spec/arb_pipeline_statistics_query/pipeline_stats_clip.c
- tests/spec/arb_pipeline_statistics_query/pipeline_stats_extra_prims.c
- tests/spec/arb_pipeline_statistics_query/pipeline_stats_frag.c
- tests/spec/arb_pipeline_statistics_query/pipeline_stats_geom.c
- tests/spec/arb_pipeline_statistics_query/pipeline_stats_vert.c
- tests/spec/arb_pipeline_statistics_query/pipeline_stats_vert_adj.c
- tests/spec/arb_pipeline_statistics_query/pipestat_help.c
- tests/spec/arb_robustness/draw-vbo-bounds.c
- tests/spec/arb_sampler_objects/CMakeLists.gl.txt
- + tests/spec/arb_sampler_objects/srgb-decode-multiple-sampler-views.c
- tests/spec/arb_shader_image_load_store/host-mem-barrier.c
- tests/spec/arb_shader_image_load_store/invalid.c
- + tests/spec/arb_spirv_extensions/CMakeLists.txt
- + tests/spec/arb_spirv_extensions/execution/shader_draw_parameters/vs-baseinstance.shader_test
- + tests/spec/arb_spirv_extensions/execution/shader_draw_parameters/vs-basevertex.shader_test
- + tests/spec/arb_spirv_extensions/execution/shader_draw_parameters/vs-drawid.shader_test
- + tests/spec/arb_spirv_extensions/execution/storage_buffer_storage_class/aoa-2.shader_test
- + tests/spec/arb_spirv_extensions/execution/storage_buffer_storage_class/array-indirect.shader_test
- + tests/spec/arb_spirv_extensions/execution/storage_buffer_storage_class/array-inside-ssbo.shader_test
- + tests/spec/arb_spirv_extensions/execution/storage_buffer_storage_class/array-of-arrays-inside-ssbo.shader_test
- + tests/spec/arb_spirv_extensions/execution/storage_buffer_storage_class/array.shader_test
- + tests/spec/arb_spirv_extensions/execution/storage_buffer_storage_class/simple.shader_test
- + tests/spec/arb_spirv_extensions/execution/storage_buffer_storage_class/two-ssbo-different-layouts.shader_test
- + tests/spec/arb_spirv_extensions/execution/storage_buffer_storage_class/unsized-array.shader_test
- + tests/spec/arb_spirv_extensions/post_depth_coverage/CMakeLists.gl.txt
- + tests/spec/arb_spirv_extensions/post_depth_coverage/CMakeLists.txt
- + tests/spec/arb_spirv_extensions/post_depth_coverage/basic.c
- + tests/spec/arb_spirv_extensions/post_depth_coverage/common.c
- + tests/spec/arb_spirv_extensions/post_depth_coverage/common.h
- + tests/spec/arb_spirv_extensions/post_depth_coverage/multisampling.c
- + tests/spec/arb_spirv_extensions/post_depth_coverage/sample-shading.c
- + tests/spec/arb_spirv_extensions/post_depth_coverage/shader_source/basic-fs.shader_source
- + tests/spec/arb_spirv_extensions/post_depth_coverage/shader_source/fs.shader_source
- + tests/spec/arb_spirv_extensions/post_depth_coverage/shader_source/multisampling-fs-1.shader_source
- + tests/spec/arb_spirv_extensions/post_depth_coverage/shader_source/multisampling-fs-2.shader_source
- + tests/spec/arb_spirv_extensions/post_depth_coverage/shader_source/multisampling-fs-3.shader_source
- + tests/spec/arb_spirv_extensions/post_depth_coverage/shader_source/sample-shading-fs.shader_source
- + tests/spec/arb_spirv_extensions/post_depth_coverage/shader_source/vs.shader_source
- tests/spec/arb_texture_buffer_object/formats.c
- tests/spec/arb_texture_view/CMakeLists.gl.txt
- + tests/spec/arb_texture_view/blit-from-tex-view.c
- tests/spec/arb_vertex_type_2_10_10_10_rev/vertex-attrib-pointer-type-size-match.c
- tests/spec/ext_direct_state_access/indexed-state-queries.c
- tests/spec/ext_external_objects/interop.c
- tests/spec/ext_external_objects/interop.h
- tests/spec/ext_external_objects/vk.c
- tests/spec/ext_external_objects/vk.h
- tests/spec/ext_external_objects/vk_buf_exchange.c
- tests/spec/ext_external_objects/vk_depth_display.c
- tests/spec/ext_external_objects/vk_image_display.c
- tests/spec/ext_external_objects/vk_image_display_multiple_textures.c
- tests/spec/ext_external_objects/vk_image_display_overwrite.c
- tests/spec/ext_external_objects/vk_image_overwrite.c
- tests/spec/ext_external_objects/vk_ping_pong.c
- tests/spec/ext_external_objects/vk_pix_buf_update_errors.c
- tests/spec/ext_external_objects/vk_semaphores.c
- tests/spec/ext_external_objects/vk_semaphores2.c
- tests/spec/ext_external_objects/vk_stencil_display.c
- tests/spec/ext_external_objects/vk_vert_buf_reuse.c
- tests/spec/ext_external_objects/vk_vert_buf_update_errors.c
- tests/spec/ext_image_dma_buf_import/CMakeLists.gles2.txt
- tests/spec/ext_image_dma_buf_import/intel_external_sampler_only.c
- tests/spec/ext_image_dma_buf_import/invalid_attributes.c
- tests/spec/ext_image_dma_buf_import/invalid_hints.c
- tests/spec/ext_image_dma_buf_import/missing_attributes.c
- tests/spec/ext_image_dma_buf_import/modifiers.c
- tests/spec/ext_image_dma_buf_import/ownership_transfer.c
- + tests/spec/ext_image_dma_buf_import/render-rgb24.c
- tests/spec/ext_image_dma_buf_import/tex-modifier.c
- tests/spec/ext_image_dma_buf_import/unsupported_format.c
- + tests/spec/ext_multisampled_render_to_texture/CMakeLists.gles2.txt
- + tests/spec/ext_multisampled_render_to_texture/CMakeLists.txt
- + tests/spec/ext_multisampled_render_to_texture/clear_color_and_depth.c
- tests/spec/ext_packed_depth_stencil/depth-stencil-texture.c
- + tests/spec/ext_shader_clock/execution/clock2x32.shader_test
- + tests/spec/ext_shader_pixel_local_storage/CMakeLists.gles2.txt
- + tests/spec/ext_shader_pixel_local_storage/CMakeLists.txt
- + tests/spec/ext_shader_pixel_local_storage/pixel-local-tests.c
- + tests/spec/ext_shader_realtime_clock/execution/clock2x32.shader_test
- tests/spec/gl-1.0/CMakeLists.gl.txt
- + tests/spec/gl-1.0/texture-matrix-push-pop.c
- + tests/spec/glsl-1.10/execution/glsl-fs-fragdata-mrt-alpha-target.shader_test
- + tests/spec/glsl-1.10/execution/glsl-fs-texture2d-mipmap-const-bias-01.shader_test
- + tests/spec/glsl-1.10/execution/glsl-fs-texture2d-mipmap-const-bias-02.shader_test.shader_test
- + tests/spec/glsl-1.10/execution/glsl-fs-texture2d-mipmap-const-bias-03.shader_test
- + tests/spec/glsl-1.10/execution/glsl-fs-texture2d-mipmap-const-bias-04.shader_test
- + tests/spec/glsl-1.10/execution/glsl-fs-texture2d-mipmap-const-bias-05.shader_test
- + tests/spec/glsl-1.30/execution/glsl-fs-textureLod-mipmap-const-lod-01.shader_test
- + tests/spec/glsl-1.30/execution/glsl-fs-textureLod-mipmap-const-lod-02.shader_test
- + tests/spec/glsl-1.30/execution/glsl-fs-textureLod-mipmap-const-lod-03.shader_test
- + tests/spec/glsl-1.30/execution/glsl-fs-textureLod-mipmap-const-lod-04.shader_test
- + tests/spec/glsl-1.30/execution/glsl-fs-textureLod-mipmap-const-lod-05.shader_test
- + tests/spec/glsl-1.30/execution/glsl-fs-textureLod-mipmap-const-lod-06.shader_test
- + tests/spec/glsl-1.30/execution/glsl-fs-textureLod-mipmap-const-lod-07.shader_test
- tests/spec/glsl-1.40/uniform_buffer/two-stages.shader_test
- + tests/spec/glsl-1.50/compiler/fs-invalid-iface-bool-input.frag
- + tests/spec/glsl-1.50/compiler/vs-invalid-iface-bool-output.vert
- tests/spec/glsl-1.50/execution/variable-indexing/gs-input-array-vec2-index-rd.shader_test
- tests/spec/glsl-1.50/execution/variable-indexing/gs-input-array-vec3-index-rd.shader_test
- tests/spec/glsl-1.50/execution/variable-indexing/gs-input-array-vec4-index-rd.shader_test
- tests/spec/glsl-1.50/execution/variable-indexing/vs-output-array-vec2-index-wr-before-gs.shader_test
- tests/spec/glsl-1.50/execution/variable-indexing/vs-output-array-vec3-index-wr-before-gs.shader_test
- tests/spec/glsl-1.50/execution/variable-indexing/vs-output-array-vec4-index-wr-before-gs.shader_test
- + tests/spec/glsl-1.50/linker/instance-matching-uniform-blocks-member-name-mismatch-varying-opt.shader_test
- tests/spec/glx_arb_create_context/ext-no-config-context.c
- + tests/spec/khr_shader_subgroup/execution/bitonic_sort.shader_test
- tests/spec/khr_texture_compression_astc/khr_compressed_astc-miptree.c
- tests/texturing/s3tc-errors.c
- tests/util/piglit-framework-gl/piglit_drm_dma_buf.c
- tests/util/piglit-test-pattern.cpp
- tests/util/piglit-util-cl.c
- tests/util/piglit-util-gl.c
- tests/util/piglit-util.c
- tests/util/piglit-util.h
- tox.ini
- unittests/framework/replay/test_download_utils.py
- unittests/framework/test/test_deqp.py
- unittests/framework/test_dmesg.py
- unittests/framework/test_monitoring.py
- unittests/framework/test_profile.py
- unittests/framework/test_wflinfo.py
- unittests/generators/test_generators.py
The diff was not included because it is too large.
View it on GitLab: https://salsa.debian.org/xorg-team/app/piglit/-/compare/4842e5b1ec64485b9ec8017c098c19b47309a2a1...98162c730b8b7aa57ce9536d18d323933dd968e4
--
View it on GitLab: https://salsa.debian.org/xorg-team/app/piglit/-/compare/4842e5b1ec64485b9ec8017c098c19b47309a2a1...98162c730b8b7aa57ce9536d18d323933dd968e4
You're receiving this email because of your account on salsa.debian.org.