Tapani Pälli [Tue, 27 Oct 2020 10:27:37 +0000 (12:27 +0200)]
framework: fix deinitSession when using waiver xml
Take numWaived in to account when considering if run was ok, patch also
add result.numWaived to the total number of tests, otherwise we hit
following assert when running with waiver:
cts-runner: glcTestRunner.cpp:829: Unknown function:
Assertion `result.numExecuted == result.numPassed + result.numFailed +
result.numNotSupported + result.numWarnings' failed.
Components: Framework, OpenGL
VK-GL-CTS issue: 2631
Affects:
cts-runner
Change-Id: Iac776fd1174e5461c455282cecca208f0322358d
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Jianle Wang [Tue, 20 Oct 2020 02:31:48 +0000 (10:31 +0800)]
Reset texture data to 0 after each 16 iterations to avoid overflow
The default precision of int in the fragment shader is mediump. The fs_out_color
in GeometryShaderMaxCombinedTextureUnitsTest::m_fragment_shader_code_body will
overflow when the number of texture unit become large, since each texture data
is equal to index of the texture units.
So reset texture data to 0 after each 16 iterations to avoid the overflow.
Component: OpenGL
Affected tests: KHR-GLES31.core.geometry_shader.limits.max_combined_texture_units
VK-GL-CTS issue: 2622
Change-Id: Icc9f2ed77a5a1cf6f315605f5f1045a60a56396c
Mika Väinölä [Mon, 26 Oct 2020 21:19:59 +0000 (23:19 +0200)]
Update Amber to fix Clang 11 build
Building Amber with Clang 11 failed due to -Werror and several
-Wsuggest-destructor-override warnings.
Affects: dEQP-VK.*
VK-GL-CTS issue: 2628
Change-Id: Ibae8dbc955d32bb50eb6f5c26cd62579bc6967dd
Ricardo Garcia [Fri, 30 Oct 2020 10:36:16 +0000 (11:36 +0100)]
Fix multisample resolve compatibility tests
Two render passes cannot be compatible if their multisample and resolve
attachments differ in the image format. However, the existing
multisample resolve compatibility tests break this rule.
Render pass compatibility rules do specify the attachment reference can
be different, and in particular this means one of the attachments can be
unused in one render pass and used in another. This fix takes advantage
of that to test render pass compatiblity.
Affected and new tests:
dEQP-VK.renderpass*.suballocation.multisample_resolve.*.compatibility_*
Components: Vulkan
VK-GL-CTS issue: 2629
Change-Id: I9a54f9a8ffc79f89d4b2dc2179b22acdba16aab5
Ricardo Garcia [Fri, 23 Oct 2020 10:18:22 +0000 (12:18 +0200)]
Create debug report recorder earlier
Change debug report recorder creation so it happens before the device is
created, so any error messages reported during device creation can be
properly reported.
This commit doesn't affect normal test runs, but it could be beneficial
to detect system setup problems.
Components: Vulkan
VK-GL-CTS issue: 2627
Change-Id: Iefe7c46e2a90673e8625a541f35fbfb35c1afd8d
Mikko Tiusanen [Mon, 19 Oct 2020 15:03:50 +0000 (18:03 +0300)]
Add tests for oob access in unexecuted shader paths
Added tests for non-robust out of bounds buffer access in unexecuted
shader code paths.
New tests:
dEQP-VK.robustness.non_robust_buffer_access.*
Components: Vulkan
VK-GL-CTS issue: 2171
Change-Id: I43ec27225ed949964d1d9783745326ff7a4dd528
Ari Suonpaa [Mon, 26 Oct 2020 06:03:05 +0000 (08:03 +0200)]
Add subpass self dependency test with a single attachment
This test replicates an issue seen in an app that behaves incorrectly
on a specific driver implementation. It uses a single attachment
referenced as both an input and an output. There is also alpha blending
enabled even though it has no effect in this particular test case.
This was needed to reproduce the original bug.
VK-GL-CTS Issue: 2597
New tests:
dEQP-VK.renderpass*.suballocation.subpass_dependencies.single_attachment.*
Components: Vulkan
Change-Id: Iaf79b3937dec4eebae1b6ad7ec035c1bdf05e811
Alexander Galazin [Sat, 24 Oct 2020 08:37:17 +0000 (10:37 +0200)]
Re-enable cube.*.mag_filter.linear test
Should now work according to
https://gitlab.khronos.org/Tracker/vulkancts/-/issues/340#note_58553
Components: Vulkan
Affects: dEQP-VK.pipeline.sampler.view_type.cube.format.r5g6b5_unorm_pack16.mag_filter.linear
Change-Id: Iffca5cd79053d1e9874beb12c9fafd8edd1cb6a8
Yiwei Zhang [Tue, 9 Jun 2020 04:01:09 +0000 (21:01 -0700)]
Add VkDeviceMemory tests for the VK_EXT_device_memory_report extension
This test allocates and frees VkDeviceMemory explicitly and verifies the
allocate/free/allocation_failed callback events are emitted at the
proper timing as well as verifying the correctness of the callback data.
Components: Vulkan
VK-GL-CTS issue: 2266
New tests: dEQP-VK.memory.device_memory_report.vk_device_memory.*
Change-Id: I96c7973452f512bf4d6f1c819bcf53eb0101b474
Andrea Faulds [Thu, 5 Nov 2020 15:35:51 +0000 (16:35 +0100)]
Permit not flushing operands for FP16 DenormFlushToZero add, dot, vmulm
Where `d` is a denormal value such that `2 * d` is not denormal:
- The F16 OpFAdd tests assumed (d + d) is zero when DenormFlushToZero is
used. This is not true if the operands (d) are not flushed to zero by
the implementation, which is allowed by the Vulkan specification.
- The F16 OpDot test assumed dot(vec2(d, d), vec2(1, 1)) is zero. Since
that is equivalent to (d + d), it has the same problem as OpFAdd.
- The F16 OpVectorTimesMatrix test made the same assumption as the OpDot
test but for both columns of a 2-by-2 matrix, with the same problem.
These tests are changed to accept `2 * d` as a valid result alongside
zero, such that they will pass both on implementations which flush
operands to zero and on implementations which only flush results to
zero.
Modified tests:
dEQP-VK.spirv_assembly.instruction.*.float_controls.fp16.input_args.denorm_*flush_to_zero*
dEQP-VK.spirv_assembly.instruction.compute.float_controls.independence_settings.denorm_ind_*_fp16_flush_*
VK-GL-CTS Issue: 2644
Components: Vulkan
Change-Id: I0ef4a57915e3bdab100f55f52d21913f87af4927
Ricardo Garcia [Tue, 3 Nov 2020 13:55:51 +0000 (14:55 +0100)]
Avoid using intersection shaders with triangles
dEQP-VK.ray_tracing_pipeline.misc.mixedPrimTL used a single shader group to
handle both triangles and AABBs, which resulted in intersection shaders
being used for triangles, which is forbidden by the spec.
This fix creates two shader groups: one for AABBs and one for triangles, using
instanceShaderBindingTableRecordOffset to choose between both.
In addition, it fixes a validation error related to invalidating a mapped
memory range without using a multiple of nonCoherentAtomSize for the size.
Affected tests:
dEQP-VK.ray_tracing_pipeline.misc.*
Components: Vulkan
VK-GL-CTS issue: 2637
Change-Id: I9d181d0646cef6ab23062ebfa1e035affce05237
Jason Ekstrand [Thu, 29 Oct 2020 21:48:46 +0000 (16:48 -0500)]
Fix the winding order of triangle cubes in dEQP-VK.ray_tracing_pipeline.misc.*
This changes all the faces to have right-handed normals that point into
the cube. This fixes a couple of test failures on Intel.
Affected tests: dEQP-VK.ray_tracing_pipeline.misc.*
Components: Vulkan
VK-GL-CTS issue: 2636
Change-Id: I70028a236cbd46d63e236a2cc57657ee251ce288
Ricardo Garcia [Tue, 3 Nov 2020 08:17:17 +0000 (09:17 +0100)]
Avoid invalid hit kinds with OpReportIntersection
OpReportIntersectionKHR accepts values in the [0, 127] range for the hit kind,
while gl_HitKindFrontFacingTriangleEXT or gl_HitKindBackFacingTriangleEXT are
reserved values outside that range and should not be used with
reportIntersectionEXT() in GLSL.
Affected tests:
dEQP-VK.ray_tracing_pipeline.builtin.*
dEQP-VK.ray_tracing_pipeline.memguarantee.*
dEQP-VK.ray_tracing_pipeline.complexcontrolflow.*
Components: Vulkan
VK-GL-CTS issue: 2626
Change-Id: Ic69abbb53968036ab56ebad4a832485fbccd8fb5
Mika Väinölä [Tue, 25 Aug 2020 06:29:07 +0000 (09:29 +0300)]
ANGLE waiver for GLES XFB array element capture
Tests waived:
dEQP-GLES3.functional.transform_feedback.array_element.*
dEQP-GLES3.functional.transform_feedback.random.*
dEQP-GLES31.functional.program_interface_query.transform_feedback_varying.*_array_element
dEQP-GLES31.functional.program_interface_query.transform_feedback_varying.type.*.array.*
KHR-GLES31.core.program_interface_query.transform-feedback-types
Components: OpenGL ES
VK-GL-CTS issue: 2237
Change-Id: I72be7800ced02423a4ac120e5d9ce581636a6a4f
(cherry picked from commit
96fbe712c585b5955b0363cada87ab62a6e4a72b)
Ricardo Garcia [Tue, 13 Oct 2020 11:17:49 +0000 (13:17 +0200)]
Check images can use external host memory before creating them
Call vkGetPhysicalDeviceImageFormatProperties2 before creating an image
with external host memory, to verify it's safe and the implementation
does not return VK_ERROR_FORMAT_NOT_SUPPORTED.
Affected tests:
dEQP-VK.memory.external_memory_host.*
Components: Vulkan
VK-GL-CTS issue: 2620
Change-Id: Ied982d4a8b61a7ad97bac8a096f1a2827a0c0f49
Andrew Fobel [Wed, 28 Oct 2020 19:41:58 +0000 (15:41 -0400)]
Fix descriptorCount for YCbCr-related tests
Certain YCbCr tests did not behave consistently with the Vulkan spec
when creating descriptor pools.
Tests now properly query combinedImageSamplerDescriptorCount to
correctly set pool size in the event of multiplane images.
Affects:
dEQP-VK.protected_memory.interaction.ycbcr.*
dEQP-VK.ycbcr.format.*
dEQP-VK.ycbcr.plane_view.*
dEQP-VK.ycbcr.query.*
Components: Vulkan
VK-GL-CTS Issue: 2386
Change-Id: I8bb261a29c7e4ce82deced447767de915cee451a
Alexander Galazin [Mon, 2 Nov 2020 14:25:06 +0000 (15:25 +0100)]
Update Vulkan headers
Components: Vulkan
Change-Id: Ibcd4bef05a3eb6ed0c9de70765e3f50a1549ba69
Daniel Koch [Tue, 27 Oct 2020 22:56:01 +0000 (18:56 -0400)]
Update glslang and deps for latest rebase
Also includes support for new ignoreIntersectionEXT and terminateRayEXT terminators
and updates affected shaders.
Component: Framework, Vulkan
Affected tests:
dEQP-VK.ray_tracing_pipeline.*
dEQP-VK.ray_query.*
Change-Id: I9114e0637fa3d4bac3afc6af5d02f573270eb709
Ricardo Garcia [Mon, 19 Oct 2020 15:40:35 +0000 (17:40 +0200)]
Check large instanceCustomIndex values
Add a new set of tests that check the value of the
InstanceCustomIndexKHR builtin from the closest-hit, intersection and
any-hit shaders correctly matches the instanceCustomIndex value when
creating instances. The tests use large values for instanceCustomIndex,
making sure the most-significant bit when represented as a 24-bit value
is a 1 and it's not incorrectly sign-extended by the implementation when
presented as the InstanceCustomIndexKHR builtin in shaders.
This value is combined with an all-ones mask in the highest bits of the
32-bit dword.
New tests:
dEQP-VK.ray_tracing_pipeline.acceleration_structures.instance_index.*
Affected tests:
dEQP-VK.ray_tracing_pipeline.acceleration_structures.*
Components: Vulkan
VK-GL-CTS issue: 2586
Change-Id: I2af85c9e2a2c78075502ab42db451651c21e0f1a
Ricardo Garcia [Wed, 14 Oct 2020 10:48:48 +0000 (12:48 +0200)]
Test invalid pInheritanceInfo for primary command buffers
When beginning a primary command buffer, the pInheritanceInfo structure
member should be ignored according to the specification. This commit
introduces a few test that set pInheritanceInfo to different invalid
values to make sure they are being ignored.
New tests:
dEQP-VK.api.command_buffers.bad_inheritance_info_*
Components: Vulkan
VK-GL-CTS issue: 2616
Change-Id: Iabd6e82b080d8cc57476dc303f19374c1a833b04
Michal Pawlowski [Tue, 20 Oct 2020 05:53:04 +0000 (07:53 +0200)]
Add array of images copy tests
This commit adds image to image, image to buffer and buffer to image array of images copy tests.
New tests:
dEQP-VK.api.copy_and_blit.core.image_to_image.array.*
dEQP-VK.api.copy_and_blit.core.image_to_buffer.array
dEQP-VK.api.copy_and_blit.core.image_to_buffer.array_tightly_sized_buffer
dEQP-VK.api.copy_and_blit.core.buffer_to_image.array
dEQP-VK.api.copy_and_blit.core.buffer_to_image.array_tightly_sized_buffer
Components: Vulkan
VK-GL-CTS Issue: 2452
Change-Id: I1f9282af60c66e4e9cd1efe6c8c22b65acf9969d
Ricardo Garcia [Tue, 20 Oct 2020 12:52:17 +0000 (14:52 +0200)]
Test extended dynamic state vertex data offsets
The vkCmdBindVertexBuffers2EXT function allows API users to specify the
offset of vertex input data bindings inside vertex buffers as well as
vertex data sizes inside those buffers. Both parameters are related:
the sum of offset plus size must not be greater than the buffer size.
All extended dynamic state tests were using an offset of zero. This
commit adds new tests with nonzero offsets to check those parameters are
used correctly by the implementation.
New tests:
dEQP-VK.pipeline.extended_dynamic_state.*.stride_with_offset
Affected tests:
dEQP-VK.pipeline.extended_dynamic_state.*
Components: Vulkan
VK-GL-CTS issue: 2459
Change-Id: I6b00769395406ff03f01553c1ff1e48a8fc7d308
Michal Pawlowski [Wed, 5 Aug 2020 14:18:18 +0000 (16:18 +0200)]
Add push constant lifetime tests
This commit tries to clarify the push constant lifetime uncertainty.
Components: Vulkan
New tests:
dEQP-VK.pipeline.push_constant.lifetime.*
Affects:
dEQP-VK.pipeline.push_constant.graphics_pipeline.*
VK-GL-CTS issue: 2393
Change-Id: I21a4b46ae1ce00caa67d72ab593ca456f5fa5da1
Iago Toral Quiroga [Tue, 27 Oct 2020 11:24:52 +0000 (12:24 +0100)]
Fix alpha to one tests
The tests were not checking that alpha is actually set to one, which
apparently, would make the tests pass always, even if the driver
doesn't actually set alpha to one. I observed this behavior with two
different drivers.
Component: Vulkan
VK-GL-CTS Issue: 2630
Affects: dEQP-VK.pipeline.multisample.alpha_to_one.*
Change-Id: I9002464c4ae731b6c2855c76ae48329cd1acd43f
Alexander Galazin [Sun, 25 Oct 2020 17:01:49 +0000 (18:01 +0100)]
Merge vk-gl-cts/opengl-cts-4.6.0 into vk-gl-cts/master
Change-Id: Id1b01740cb67862f257cb668234cb760a15505df
Alexander Galazin [Sun, 25 Oct 2020 09:25:21 +0000 (10:25 +0100)]
Merge vk-gl-cts/opengl-es-cts-3.2.6 into vk-gl-cts/master
Change-Id: Ie3a7279296c9a1bec8e0711b0efcf0265ebad249
Alexander Galazin [Sat, 24 Oct 2020 08:27:56 +0000 (10:27 +0200)]
Notice of withdrawal of Vulkan CTS 1.1.6.x
Components: Vulkan
Change-Id: I6f5df9b2b72c4838e0721a137eb9d43d314b83fe
Alexander Galazin [Fri, 23 Oct 2020 18:38:25 +0000 (20:38 +0200)]
Merge vk-gl-cts/vulkan-cts-1.2.4 into vk-gl-cts/master
Change-Id: I559614f2f74c3975eba31caa8025205eea5d0816
Alexander Galazin [Fri, 23 Oct 2020 15:30:46 +0000 (17:30 +0200)]
Merge vk-gl-cts/vulkan-cts-1.2.3 into vk-gl-cts/vulkan-cts-1.2.4
Change-Id: I48eda80e9fcd03bb5d402e94b873cb536ffb53e7
Varvara Poskonina [Fri, 23 Oct 2020 12:23:44 +0000 (14:23 +0200)]
Fix validation of DRM modifier compatibility with image properties
Vulkan spec states that in case of image creation with list of
DRM modifiers (VkImageDrmFormatModifierListCreateInfoEXT), each
of them must be compatible with the parameters in
VkImageCreateInfo and its pNext chain.
This change adds a mandatory check for modifier compatibility with
the specified image properties by querying
vkGetPhysicalDeviceImageFormatProperties in a place where it was
missing prior to the actual image creation.
Components: Vulkan
Affects: dEQP-VK.drm_format_modifiers.create_list_modifiers.*
VK-GL-CTS Issue: 2611
Change-Id: Iab5deff6c8cbe6d2c510224322f7be91c7874acf
Alexander Galazin [Fri, 23 Oct 2020 09:30:12 +0000 (11:30 +0200)]
Update KC CTS reference
Components: Framework
Change-Id: I733864e2e41aa6ec10a444a282b105aa39925ffd
Mika Väinölä [Tue, 29 Sep 2020 06:58:03 +0000 (09:58 +0300)]
Modify/add GLES XFB tests without element capture
Some GLES implementations are unable to support array elements as
transform feedback varyings. Capture the full array when element
capture is not explicitly tested. Add new variants for tests where
different types of varyings are tested at the same time.
New tests:
dEQP-GLES3.functional.transform_feedback.random_full_array_capture.*
KHR-GLES31.core.program_interface_query.transform-feedback-types-full-array-capture
Affects:
dEQP-GLES3.functional.transform_feedback.random.*
KHR-GLES31.core.shader_image_size.*
KHR-GLES31.core.vertex_attrib_binding.basic-input*
Components: OpenGL ES
VK-GL-CTS issue: 2237
Change-Id: Ib6164d89bc99f8ca8287dad7cc1e0789916e3256
(cherry picked from commit
8404b33c25c7f7cf5cd49a7c18c3a63b0f1f70ae)
Mika Väinölä [Tue, 25 Aug 2020 06:29:07 +0000 (09:29 +0300)]
ANGLE waiver for GLES XFB array element capture
Tests waived:
dEQP-GLES3.functional.transform_feedback.array_element.*
dEQP-GLES3.functional.transform_feedback.random.*
dEQP-GLES31.functional.program_interface_query.transform_feedback_varying.*_array_element
dEQP-GLES31.functional.program_interface_query.transform_feedback_varying.type.*.array.*
KHR-GLES31.core.program_interface_query.transform-feedback-types
Components: OpenGL ES
VK-GL-CTS issue: 2237
Change-Id: I72be7800ced02423a4ac120e5d9ce581636a6a4f
Piotr Byszewski [Wed, 7 Oct 2020 09:35:55 +0000 (11:35 +0200)]
Waiver for wide-line attribute interpolation
Wide-line attribute interpolation tests recently added to
the mustpass list cannot be passed by any hardware that
Intel has ever shipped.
Components: OpenGL
Affects:
dEQP-GLES2.functional.rasterization.interpolation.basic.line_loop_wide
dEQP-GLES2.functional.rasterization.interpolation.basic.line_strip_wide
dEQP-GLES2.functional.rasterization.interpolation.basic.lines_wide
dEQP-GLES2.functional.rasterization.interpolation.projected.line_loop_wide
dEQP-GLES2.functional.rasterization.interpolation.projected.line_strip_wide
dEQP-GLES2.functional.rasterization.interpolation.projected.lines_wide
dEQP-GLES3.functional.rasterization.fbo.rbo_singlesample.interpolation.lines_wide
dEQP-GLES3.functional.rasterization.fbo.texture_2d.interpolation.lines_wide
dEQP-GLES3.functional.rasterization.interpolation.basic.line_loop_wide
dEQP-GLES3.functional.rasterization.interpolation.basic.line_strip_wide
dEQP-GLES3.functional.rasterization.interpolation.basic.lines_wide
dEQP-GLES3.functional.rasterization.interpolation.projected.line_loop_wide
dEQP-GLES3.functional.rasterization.interpolation.projected.line_strip_wide
dEQP-GLES3.functional.rasterization.interpolation.projected.lines_wide
VK-GL-CTS issue: 2601
Change-Id: I7508429613e0d8cde45d6e9d6b5d07f0b17f7753
Piotr Byszewski [Wed, 7 Oct 2020 14:27:02 +0000 (16:27 +0200)]
Add support for waiver mechanism to cts-runner
Affects: cts-runner
Change-Id: I147cbb49114386b939a6563a1b4b96f620310736
Components: Framework
Samuel Iglesias Gonsálvez [Fri, 9 Oct 2020 08:47:31 +0000 (10:47 +0200)]
Fix depth/stencil resolve compatibility tests
This commit fixes validation errors related to the depth/stencil resolve
attachment, making them follow validation rules. Depth/stencil resolve
attachments must match in bit size and format with their multisample
counterparts, but they can remove aspects (depth or stencil) in the
resolve operation.
Some of the existing tests did not match these rules.
Removed tests:
dEQP-VK.renderpass2.depth_stencil_resolve.image_2d_32_32.samples_2.d16_unorm.compatibility_*
dEQP-VK.renderpass2.depth_stencil_resolve.image_2d_32_32.samples_2.x8_d24_unorm_pack32.compatibility_*
dEQP-VK.renderpass2.depth_stencil_resolve.image_2d_32_32.samples_2.d32_sfloat.compatibility_*
dEQP-VK.renderpass2.depth_stencil_resolve.image_2d_32_32.samples_2.s8_uint.compatibility_*
Affected tests:
dEQP-VK.renderpass2.depth_stencil_resolve.image_2d_32_32.samples_2.*.compatibility_*
Components: Vulkan
VK-GL-CTS issue: 2596
Change-Id: Ifa8d3de9ca71d9c2889a86949014774d6879e537
ajerecze [Thu, 15 Oct 2020 09:56:47 +0000 (11:56 +0200)]
Fix command buffer usage flags in dEQP-VK.query_pool.performance_query tests
The test is resubmitting command buffer few times in a loop, however it
is using beginCommandBuffer utility function, which sets the command buffer usage flag to
VK_COMMAND_BUFFER_USAGE_ONE_TIME_SUBMIT_BIT by default. This commit changes that to
0 (no flags).
Moreover, VK_COMMAND_BUFFER_USAGE_SIMULTANEOUS_USE_BIT flag is removed, as it seems
The submits are serialized by a fence.
Additionally replaced raw vkCmdBeginCommandBuffr with beginCommandBuffer for clarity.
Components: Vulkan
VK-GL-CTS Issue: 2617
Affects:
dEQP-VK.query_pool.performance_query.query*
dEQP-VK.query_pool.performance_query.multiple_pools*
Change-Id: Iff25e4cc715582710f3ed23367a7fd8f9a23b29e
Ari Suonpaa [Wed, 9 Sep 2020 11:27:26 +0000 (14:27 +0300)]
Fix validation errors in AHB tests
AHB protected memory tests were also removed as the test
tried to use protected memory without creating a device
and queue that would support protected operations.
VK-GL-CTS Issue: 2508
Affects:
dEQP-VK.api.external.memory.android_hardware_buffer.*
Components: Vulkan
Change-Id: I307e7450d06d20b6497f2a529e5724ae0538e5a9
Yiwei Zhang [Mon, 25 May 2020 22:32:49 +0000 (15:32 -0700)]
Add create and destroy object tests for the VK_EXT_device_memory_report extension
This test creates and destroys various kinds of Vulkan objects and
verifies that all the emitted callback events are properly paired.
Components: Vulkan
VK-GL-CTS issue: 2266
New tests: dEQP-VK.memory.device_memory_report.create_and_destroy_object.*
Change-Id: I1a3e83e1bfe6f89fbcc511e9b43facbddda79b60
Mikko Tiusanen [Thu, 23 Jul 2020 15:18:21 +0000 (18:18 +0300)]
Add tests for clamping of depth values
Added tests for clamping of depth values.
Includes depth bias and VK_EXT_depth_range_unrestricted tests for formats
that can support it.
New tests:
dEQP-VK.draw.depth_clamp.*
Components: Vulkan
VK-GL-CTS issue: 2312
Change-Id: Id9b27146032d6e08cbc7362aff3fd64cc9ed2556
Joshua Ashton [Fri, 17 Apr 2020 04:10:55 +0000 (05:10 +0100)]
Implement custom border color tests
New Tests:
dEQP-VK.pipeline.sampler.*.all_mode_clamp_to_border_custom_*
Components: Vulkan
Vulkan issue: 2088
Change-Id: I5b8b73147aad8b3398c9ba55577397b2224d3940
Alexander Galazin [Thu, 22 Oct 2020 11:04:14 +0000 (13:04 +0200)]
Amend OpenGL/ES Conformance Submission Package Requirements
The current practice is that all bugfixes should be merged
into the CTS repo before being used in a submission.
This change promotes this practices to a requirement.
Components: OpenGL
Change-Id: Id5e5782c0418cbedb0ca23b0abd06051021b63ec
Alexander Galazin [Thu, 22 Oct 2020 10:50:23 +0000 (12:50 +0200)]
Amend Vulkan Conformance Submission Package Requirements
The current practice is that all bugfixes should be merged
into the CTS repo before being used in a submission.
This change promotes this practices to a requirement.
Compontents: Vulkan
Change-Id: I61eef5522aecf97a3d42c49777681eb671affe87
Alexander Galazin [Wed, 21 Oct 2020 19:39:37 +0000 (21:39 +0200)]
Merge vk-gl-cts/master into vk-gl-cts/vk_khr_ray_tracing
Change-Id: Ibedf80e7b25e227e5055531a308ef3a48095bb54
Matthew Netsch [Wed, 21 Oct 2020 19:46:27 +0000 (15:46 -0400)]
Splits long shader in memory model tests
Components: Vulkan
VK-GL-CTS Issue: 2625
Affects:
dEQP-VK.memory_model.*
Change-Id: I8478e268fa5988b015acd2dd70cd370bfc4c082c
Ilia Mirkin [Sat, 1 Feb 2020 07:34:30 +0000 (02:34 -0500)]
Fix tests to work with pre-GL4 drivers
Components: OpenGL
Affects:
KHR-GL33.cull_distance.*
KHR-GL33.pipeline_statistics_query_tests_ARB.*
KHR-GL33.gpu_shader5_gl.*
KHR-GL33.transform_feedback.api_errors_test
(cherry picked from commit
e14d294a7b773e90b350b41d4cc791ea01185b29)
Change-Id: Iab1400128c644214d224b6e15daf09dbfe88951f
Eric Werness [Fri, 16 Nov 2018 21:50:40 +0000 (13:50 -0800)]
Support Python3 as well in gen_release_info.py
The CMake file version check can end up picking python3 in some cases -
making the Python code compatible with python3 seems easier than
adjusting the version check.
Components: Frameworks
Change-Id: Iab37090ab813dcb7839490225ca2a4bf2f721c12
(cherry picked from commit
f1bcdd5e9df1f83cc9dc4ef7235d3da16e9f7fab)
(cherry picked from commit
506598e19e7154d58b258f61f3b8a4cf5c87d0c5)
Alexander Galazin [Wed, 7 Oct 2020 10:28:33 +0000 (12:28 +0200)]
Update vk.xml revision
Components: Vulkan
Affects: dEQP-VK.api.info.*extensions
Change-Id: I2f47f4dfdcb83a05712a268c1e1117dc8fb571f3
Lionel Landwerlin [Sat, 26 Sep 2020 12:27:56 +0000 (15:27 +0300)]
Replace incorrect undefined->transfer-dst-opt layout transfers
The additional transfer from the UNDEFINED layout right after a clear
command seem incorrect, as they can have the effect of discarding the
previous command's effect on some implementations.
Put barriers in front of the clears using the same source & destination
layouts.
Affects:
dEQP-VK.api.image_clearing.core.clear_color_image.*
Components: Vulkan
VK-GL-CTS issue: 2603
Change-Id: I57377f12968f6ffd38ac017f874324cd4bfdf3e4
Michal Pawlowski [Sun, 4 Oct 2020 04:47:07 +0000 (06:47 +0200)]
Add cube compat. to array image copy tests
This commit adds image to image cube compatible to 2d image array and vice versa copy tests.
VK-GL-CTS Issue: 2542
New tests:
dEQP-VK.api.copy_and_blit.core.image_to_image.cube.*
Change-Id: I37d191b05c70a74f5e659018aab4c73114fc250e
Components: Vulkan
Mikko Tiusanen [Fri, 21 Aug 2020 09:55:51 +0000 (12:55 +0300)]
Tests for pipeline laoyut early destroy
Added tests that use the created graphics pipeline after destroying
the pipeline layout used to create them.
New tests:
dEQP-VK.pipeline.early_destroy.no_cache_destroy_layout
dEQP-VK.pipeline.early_destroy.cache_destroy_layout
Affects:
dEQP-VK.pipeline.early_destroy.*
Components: Vulkan
VK-GL-CTS issue: 2362
Change-Id: Ib296b2834ea7c31891532e5c8e7102f30acc000f
Matthew Netsch [Wed, 14 Oct 2020 20:51:51 +0000 (16:51 -0400)]
Fixes cubic filter copy and blt tests
Only 2D is supported per spec
Components: Vulkan
VK-GL-CTS Issue: 2600
Affects:
dEQP-VK.api.copy_and_blit.*3d*
Change-Id: I145ec2bf638c497e08f85b328c3e39bf6ee5c636
Timothy Arceri [Thu, 30 Apr 2020 07:31:13 +0000 (17:31 +1000)]
Fix broken layout binding tests
Some of the shaders being generated contain a misplaced ;
An example of the code being gererated is as follows:
fragColor = vec4(float(atomicCounter(atomic0)), 1.0, 0.0, 1.0);
+vec4(float(atomicCounter(atomic1)), 1.0, 0.0, 1.0);
+vec4(float(atomicCounter(atomic2)), 1.0, 0.0, 1.0);
+vec4(float(atomicCounter(atomic3)), 1.0, 0.0, 1.0);
Affects:
KHR-GL46.layout_binding.atomic_uint_layout_binding_atomic_*
Components: OpenGL
VK-GL-CTS issue: 2348
Change-Id: Ie209de8954c8b8e3920c17be9190030946c93354
(cherry picked from commit
64d344b60bb7b4324525e08766fdeb14bf1f18c6)
Alexander Galazin [Fri, 9 Oct 2020 15:32:21 +0000 (17:32 +0200)]
Merge vk-gl-cts/vulkan-cts-1.2.4 into vk-gl-cts/master
Change-Id: Id859c0fcaa52f04154e5686ec8a9659c75635ad2
Timothy Arceri [Sun, 3 May 2020 05:57:02 +0000 (15:57 +1000)]
Fix some egl robustness tests
Because the values of color were initially undefined an optimised
compiler can simply set all components of the vector to the value
being assigned to color[u_index] and eventually eliminate color and
u_index completely.
To avoid this we initialise color.
Affects:
dEQP-EGL.functional.robustness.reset_context.shaders*
Components: OpenGL
VK-GL-CTS issue: 2349
Change-Id: I6eb929eec07b039afb247427af91aba65dbade9b
(cherry picked from commit
4392facc87c2da80b5065f8663b9303fb1113e16)
Alexander Galazin [Fri, 9 Oct 2020 10:17:53 +0000 (12:17 +0200)]
Allow Vulkan CTS 1.2.4.1
Components: Vulkan
Affects: dEQP-VK.api.driver_properties.properties
Change-Id: I2bef3f7c8fb106ce98d450383d3f9cec7facd800
Mika Väinölä [Tue, 29 Sep 2020 06:58:03 +0000 (09:58 +0300)]
Modify/add GLES XFB tests without element capture
Some GLES implementations are unable to support array elements as
transform feedback varyings. Capture the full array when element
capture is not explicitly tested. Add new variants for tests where
different types of varyings are tested at the same time.
New tests:
dEQP-GLES3.functional.transform_feedback.random_full_array_capture.*
KHR-GLES31.core.program_interface_query.transform-feedback-types-full-array-capture
Affects:
dEQP-GLES3.functional.transform_feedback.random.*
KHR-GLES31.core.shader_image_size.*
KHR-GLES31.core.vertex_attrib_binding.basic-input*
Components: OpenGL ES
VK-GL-CTS issue: 2237
Change-Id: Ib6164d89bc99f8ca8287dad7cc1e0789916e3256
Alastair Donaldson [Wed, 16 Sep 2020 20:37:51 +0000 (21:37 +0100)]
Add a batch of GraphicsFuzz tests
Components: Vulkan
New Tests:
dEQP-VK.graphicsfuzz.spv-double-branch-to-same-block
dEQP-VK.graphicsfuzz.spv-double-branch-to-same-block2
dEQP-VK.graphicsfuzz.spv-double-branch-to-same-block3
dEQP-VK.graphicsfuzz.spv-load-from-frag-color
Change-Id: I8da268adc32c75e2cc78feb4c23040264b503e66
Samuel Iglesias Gonsálvez [Tue, 22 Sep 2020 15:28:43 +0000 (17:28 +0200)]
Fix wrong access flag for depth/stencil resolve
According to the spec:
"End-of-subpass multisample resolves are treated as color attachment
writes for the purposes of synchronization."
Affects: dEQP-VK.renderpass2.depth_stencil_resolve.*
Component: Vulkan
VK-GL-CTS issue: 2593
Change-Id: I055049fffc0fd4bdbd162c4468cf80ff0c5ca373
Lionel Landwerlin [Wed, 30 Sep 2020 15:19:27 +0000 (18:19 +0300)]
Fix wsi tests using disallowed image usage
Even though STORAGE usage might be reported by
VkSurfaceCapabilitiesKHR::supportedUsageFlags, an application should
always verify whether for a given format the STORAGE usage is
supported through vkGetPhysicalDeviceImageFormatProperties().
This also fixes validation layer warnings
(VUID-VkSwapchainCreateInfoKHR-imageFormat-01778).
Affects:
dEQP-VK.wsi.*.swapchain.simulate_oom.image_usage
Components: Vulkan
VK-GL-CTS issue: 2602
Change-Id: I21a5faff6206bdeaf6ecd5d8c8024d17965db548
Samuel Iglesias Gonsálvez [Thu, 24 Sep 2020 10:15:23 +0000 (12:15 +0200)]
Output depth image in case of failure
Affects:
dEQP-VK.draw.inverted_depth_ranges.*
Components: Vulkan
VK-GL-CTS issue: 2446
Change-Id: Ic60c321d8889638af71e5ed4828bfcd007a7e8f3
Karolina Palka [Thu, 24 Sep 2020 14:49:29 +0000 (16:49 +0200)]
Check SparseResidency flag before OpImageSparseRead
This patch adds check if SpvCapabilitySparseResidency
is supported before doing SpvOpImageSparseRead
Affects:
dEQP-VK.image.mismatched_formats.sparse_image_read.*
Components: Vulkan
VK-GL-CTS issue: 2591
Change-Id: I98852842285d4df9008e7e2403a4680c81b6a9f1
Varvara Poskonina [Wed, 7 Oct 2020 13:33:57 +0000 (15:33 +0200)]
Fix validation of DRM modifier compatibility with image properties
Vulkan spec states that in case of image creation with list of
DRM modifiers (VkImageDrmFormatModifierListCreateInfoEXT), each
of them must be compatible with the parameters in
VkImageCreateInfo and its pNext chain.
This change adds a mandatory check for modifier compatibility with
the specified image properties by querying
vkGetPhysicalDeviceImageFormatProperties in a place where it was
missing prior to the actual image creation.
Components: Vulkan
Affects: dEQP-VK.drm_format_modifiers.create_list_modifiers.*
VK-GL-CTS Issue: 2611
Change-Id: Iab5deff6c8cbe6d2c510224322f7be91c7874acf
Ari Suonpaa [Tue, 18 Aug 2020 12:45:29 +0000 (15:45 +0300)]
Fix result checking of output location tests
Output location tests were expecting bit accurate results
when comparing to reference values. However some of the
tests use lowp or mediump precision. This change adds
some tolerance for the result checking. Also lowp variants
are removed as they acted identically to mediump ones.
VK-GL-CTS Issue: 2513
VK-GL-CTS Issue: 2510
VK-GL-CTS Issue: 2550
Affects:
dEQP-VK.draw.output_location.*
Components: Vulkan
Change-Id: Ic7b96a5b5be441cd78a0e50be5fcf86308c80652
Alexander Galazin [Mon, 5 Oct 2020 07:43:21 +0000 (09:43 +0200)]
Update Vulkan headers
Components: Vulkan
Change-Id: I476905a4d715e98b3499efe83fb0861f2f1e4a57
Alexander Galazin [Sat, 3 Oct 2020 08:45:05 +0000 (10:45 +0200)]
Notice of withdrawal of Vulkan CTS 1.2.0.x
Components: Vulkan
Change-Id: Iecc87a7a4ea482e7a0be102c1b11e72dce5f10d6
Alexander Galazin [Fri, 2 Oct 2020 16:32:58 +0000 (18:32 +0200)]
Merge vk-gl-cts/vulkan-cts-1.2.4 into vk-gl-cts/master
Change-Id: Icdd0f8c4935129168a8bf7d1607fc26a7fac9e32
Alexander Galazin [Fri, 2 Oct 2020 15:13:01 +0000 (17:13 +0200)]
Merge vk-gl-cts/vulkan-cts-1.2.3 into vk-gl-cts/vulkan-cts-1.2.4
Change-Id: I0260b62c60b55364b6b0777c70ce605d36bda861
Alexander Galazin [Fri, 2 Oct 2020 11:29:44 +0000 (13:29 +0200)]
Merge vk-gl-cts/vulkan-cts-1.2.2 into vk-gl-cts/vulkan-cts-1.2.3
Change-Id: Ie884684e771f567171331f327cd6fa13f15c1fb3
Ari Suonpaa [Tue, 18 Aug 2020 12:45:29 +0000 (15:45 +0300)]
Fix result checking of output location tests
Output location tests were expecting bit accurate results
when comparing to reference values. However some of the
tests use lowp or mediump precision. This change adds
some tolerance for the result checking. Also lowp variants
are removed as they acted identically to mediump ones.
VK-GL-CTS Issue: 2513
VK-GL-CTS Issue: 2510
VK-GL-CTS Issue: 2550
Affects:
dEQP-VK.draw.output_location.*
Components: Vulkan
Change-Id: I3d48832f3818431332421172ef97f4900e2679ed
Ralph Potter [Mon, 20 Jul 2020 10:07:48 +0000 (11:07 +0100)]
Experimental crash postmortem use-after-free test
Adds a tests to deliberately free a buffer before executing a compute
shader that accesses it for the purposes of deliberately triggering
device-loss.
This test is intended for experimental/investigative use by the crash
postmortem TSG, is expected to fail, and should not be included in
mustpass lists.
Added Tests:
dEQP-VK-experimental.postmortem.use_after_free.*
Components: Vulkan
Vulkan issue: 2255
Change-Id: Icb7f47a37e864932044343493057ef4882473b94
Alexander Galazin [Fri, 2 Oct 2020 10:35:42 +0000 (12:35 +0200)]
Merge vk-gl-cts/vulkan-cts-1.2.1 into vk-gl-cts/vulkan-cts-1.2.2
Change-Id: I2104e27164b15d2f66896b1c1c4f188fe8df83e4
Alexander Galazin [Fri, 2 Oct 2020 09:58:50 +0000 (11:58 +0200)]
Merge vk-gl-cts/vulkan-cts-1.2.0 into vk-gl-cts/vulkan-cts-1.2.1
Change-Id: Ia15f227ae78f00e0d142f0b2ce3d966d10a09cab
Alexander Galazin [Fri, 2 Oct 2020 09:15:55 +0000 (11:15 +0200)]
Merge vk-gl-cts/vulkan-cts-1.1.6 into vk-gl-cts/vulkan-cts-1.2.1
Change-Id: I4c86e4fe92acca4c4260ab43a492ace9f527f986
Alexander Galazin [Thu, 1 Oct 2020 11:24:44 +0000 (13:24 +0200)]
Merge vk-gl-cts/vulkan-cts-1.1.5 into vk-gl-cts/vulkan-cts-1.2.0
Change-Id: Id37d7b8198e8d5867e4b802e638eb1e0010c2351
Alexander Galazin [Thu, 1 Oct 2020 10:02:06 +0000 (12:02 +0200)]
Merge vk-gl-cts/vulkan-cts-1.1.5 into vk-gl-cts/vulkan-cts-1.1.6
Change-Id: I840b7caa71134a32ba49b484b716b97c31ace9ee
Alexander Galazin [Thu, 1 Oct 2020 08:57:05 +0000 (10:57 +0200)]
Merge vk-gl-cts/vulkan-cts-1.1.4 into vk-gl-cts/vulkan-cts-1.1.5
Change-Id: I04485faa2b85d5af689da6143b130e3c92bd9253
Alexander Galazin [Thu, 1 Oct 2020 07:43:39 +0000 (09:43 +0200)]
Merge vk-gl-cts/vulkan-cts-1.1.3 into vk-gl-cts/vulkan-cts-1.1.4
Change-Id: Ic81486b5b158f0edf964d29a6e084a9c6e2dee00
Ricardo Garcia [Mon, 14 Sep 2020 11:25:42 +0000 (13:25 +0200)]
Fix post-trace rgen image write barriers
Use VK_ACCESS_SHADER_WRITE_BIT when synchronizing image stores from the
rgen shader stage to the transfer stage before copying result image
contents to a buffer.
Some ray tracing pipeline tests are not affected but most of them are,
hence marking them all as affected.
Affected tests:
dEQP-VK.ray_tracing_pipeline.*
Components: Vulkan
VK-GL-CTS issue: 2579
Change-Id: Ifccd0b5d9a039f7cfef8ca09a9a397e99987b821
Dominik Witczak [Wed, 30 Sep 2020 16:38:38 +0000 (18:38 +0200)]
Add tests that trace rays recursively.
New tests:
dEQP-VK.ray_tracing_pipeline.misc.recursiveTraces_*
Components: Vulkan
VK-GL-CTS issue: 2480
Change-Id: I33cc90fb44fbef139de43d1df53634d8620ac2bc
Matthew Netsch [Wed, 30 Sep 2020 16:27:25 +0000 (12:27 -0400)]
Fixes multisample resolve sRGB cases
CTS expects downsampling in linear space,
where spec says implementation dependent
Components: Vulkan
VK-GL-CTS Issue: 2584
Affects:
dEQP-VK.renderpass.suballocation.multisample_resolve.*_srgb*
Change-Id: I8512ef058b03fe55c5429d3add5d2f4dc890c663
Graeme Leese [Thu, 10 Jan 2019 11:20:02 +0000 (11:20 +0000)]
Fix validation errors in subgroups tests
Some subgroups tests were attempting to create a descriptor pool of size
0, which is invalid. Instead, skip the whole construction of descriptor
sets since they aren't needed.
Components: Vulkan
Affects: dEQP-VK.subgroups.*.graphics.*
Change-Id: I5abedcc2360a0336c521aea855a9725b127c9a0f
(cherry picked from commit
a887be07dd7b63ce89794af9f70f893dc2de5955)
Alexander Galazin [Mon, 28 Sep 2020 11:07:26 +0000 (13:07 +0200)]
Update glslang, SPIR-V Tools and Headers
Change-Id: Ic18e8d56ea22256353bee952aa5aa081959600a7
Alexander Galazin [Mon, 28 Sep 2020 08:28:01 +0000 (10:28 +0200)]
Update Vulkan headers
Components: Vulkan
Change-Id: I5201a8ce0fc604bad99fdcc4c2a2aa945e66ec2b
Alexander Galazin [Fri, 25 Sep 2020 09:13:55 +0000 (11:13 +0200)]
Merge vk-gl-cts/github-master into vk-gl-cts/master
Change-Id: Id08c5fcbaea1f9cee9671fa40e5490d000c38711
Alexander Galazin [Fri, 25 Sep 2020 09:12:48 +0000 (11:12 +0200)]
Merge pull request #228 from robclark/fix-surfaceless-config
platform/surfaceless: pick config's better
Mikko Tiusanen [Fri, 11 Sep 2020 09:19:22 +0000 (12:19 +0300)]
Avoid gcc divide by zero in getEdgeVolumeClip
Adjusted getEdgeVolumeClip equation to avoid divide by zero without
causing a meaningful change in the calculation. Fixes divide by zero when
compiled using the gcc toolset.
Affects:
dEQP-GLES2.functional.clipping.triangle_vertex.clip_three.*
Components: OpenGL ES
VK-GL-CTS issue: 2035
Change-Id: I359f0486d98bfe4c642679d11bec4c8704c72943
Jeff Leger [Tue, 30 Jun 2020 01:04:14 +0000 (21:04 -0400)]
Add tests for KHR_copy_commands2
New Tests:
dEQP-VK.api.copy_and_blit.copy_commands2.*
Affects:
dEQP-VK.api.copy_and_blit.*
Component: Vulkan
VK-GL-CTS issue: 2430
Change-Id: I447f518248b43d4fbe2a84b5de6ee3f1e5e2eaf0
(cherry picked from commit
713d656677d147086e7814dd3ecfafa4f41071c6)
Karolina Palka [Wed, 2 Sep 2020 12:09:59 +0000 (14:09 +0200)]
Add tests for incompatible resolve attachments
Testing special case where resolve attachments (including depth/stencil) are completely different
New tests:
dEQP-VK.renderpass.suballocation.multisample_resolve.*.compatibility_*
dEQP-VK.renderpass2.depth_stencil_resolve.image_2d_32_32.samples_2.*.compatibility_*
Components: Vulkan
VK-GL-CTS issue: 2225
Change-Id: Ic442da66413714131d040891d56a8819d4cd4ff4
Dominik Witczak [Wed, 23 Sep 2020 11:12:18 +0000 (13:12 +0200)]
Add tests verifying shader record blocks work as expected.
New tests:
dEQP-VK.ray_tracing_pipeline.misc.shaderRecord*
Components: Vulkan
VK-GL-CTS: 2423
Change-Id: I15c59b0147c3779ca8c103bb2c2b27f3a30fb890
Ricardo Garcia [Fri, 18 Sep 2020 15:10:39 +0000 (17:10 +0200)]
Ray Tracing barrier tests
Tests exercising barriers of different types using ray tracing stages,
mixing them with other stages, different resource types, accesses and
pipeline types.
New tests:
dEQP-VK.ray_tracing_pipeline.barrier.*
Components: Vulkan
VK-GL-CTS issue: 2492
Change-Id: Iad910c4513362fabcc75b9ba4d8ca0291afa672d
Alexander Galazin [Fri, 11 Sep 2020 09:34:55 +0000 (11:34 +0200)]
Enable more lines tests
Components: Vulkan
Affects:
dEQP-VK.rasterization.interpolation.basic.lines_wide
dEQP-VK.rasterization.interpolation.projected.line_strip_wide
dEQP-VK.rasterization.flatshading.lines_wide
dEQP-VK.pipeline.multisample.*.primitive_point_1px
dEQP-VK.pipeline.multisample.*.primitive_point_1px_sparse
VK-GL-CTS issue: 1274
Change-Id: If6a3debf1bf1794ca9175fd2c5404df2891ffb85
Alexander Galazin [Wed, 16 Sep 2020 07:12:43 +0000 (09:12 +0200)]
Update Amber
Components: Framework
Affects: dEQP-VK.*
Change-Id: I0bb5016b2ec063396cd797060e14d6de474b96af
Piotr Byszewski [Tue, 22 Sep 2020 08:32:24 +0000 (10:32 +0200)]
Test passing RayQueryKHR object through function
Components: Vulkan
VK-GL-CTS issue: 2578
New tests:
dEQP-VK.ray_query.advanced.using_wrapper_function.*
Affects:
dEQP-VK.ray_query.advanced.*
dEQP-VK.ray_query.builtin.*
Change-Id: I8edfa27a481e7903580a4be6567824620397f1a0
Alexander Galazin [Mon, 21 Sep 2020 09:34:32 +0000 (11:34 +0200)]
Update headers
Components: Vulkan
Change-Id: Icfd9ec046c450a431e2723970b92a2a2a547dfd2
Alexander Galazin [Fri, 18 Sep 2020 09:07:54 +0000 (11:07 +0200)]
Merge vk-gl-cts/github-master into vk-gl-cts/master
Change-Id: I7dcae64a5618fbbed63da7bd8cd649add44d5e0a
Alexander Galazin [Fri, 18 Sep 2020 09:06:09 +0000 (11:06 +0200)]
Merge pull request #227 from kusma/nearest-edge-8-bit-precision
Lower precision required for texture-lookups
Michal Pawlowski [Fri, 4 Sep 2020 05:49:03 +0000 (07:49 +0200)]
Issues detected by synchronization validation layer fix
This commit fixes certain synchronization issues detected by synchronization validation layer.
Component: Vulkan
Affects:
dEQP-VK.api.image_clearing.core.clear_color_image.1d.optimal.remaining_array_layers_twostep.r4g4_unorm_pack8
dEQP-VK.pipeline.timestamp.transfer_tests.transfer_stage_with_copy_buffer_method
dEQP-VK.pipeline.dynamic_offset.compute.single_set.uniform_buffer.numcmdbuffers_1.sameorder.numdescriptorsetbindings_2.numdynamicbindings_1.numnondynamicbindings_0
dEQP-VK.spirv_assembly.instruction.graphics.cross_stage.basic_type.flat
dEQP-VK.glsl.builtin_var.fragdepth.point_list_d32_sfloat_multisample_2
dEQP-VK.synchronization.op.single_queue.event.write_fill_buffer_read_copy_buffer.buffer_16384
dEQP-VK.compute.basic.copy_ssbo_to_image_small
dEQP-VK.image.store.with_format.1d.r32g32b32a32_sfloat
dEQP-VK.image.qualifiers.coherent.1d.r32f
dEQP-VK.image.image_size.1d.readonly_32
dEQP-VK.multiview.depth.3_6_12_9_6_12_9_3_6_12_9_3
dEQP-VK.pipeline.push_descriptor.graphics.binding0_numcalls1_input_attachment
dEQP-VK.pipeline.cache.graphics_tests.vertex_stage_fragment_stage
dEQP-VK.ycbcr.format.g8_b8_r8_3plane_420_unorm.vertex_linear_disjoint
dEQP-VK.imageless_framebuffer.depth_stencil_resolve
dEQP-VK.renderpass.suballocation.attachment_allocation.input_output.*
Public issue: 217
Change-Id: Idaed4cad0a15e84ed3803566eb2f3e9493c2f7a1
Graeme Leese [Mon, 7 Sep 2020 14:00:08 +0000 (15:00 +0100)]
Fix memory barriers in image_subresrouce tests
Having written the data from the CPU we need to flush the allocation,
not invalidate it.
Components: Vulkan
VK-GL-CTS issue: 2445
Affects: dEQP-VK.image.subresource_layout.*
Change-Id: I45e891e8476f2d4088ca9f7bb0e1923490579f43
Slawomir Cygan [Fri, 6 Dec 2019 13:36:11 +0000 (14:36 +0100)]
Bump fence timeout in Amber test cases to infinity
Components: Vulkan
VK-GL-CTS Issue: 2119
Affects: dEQP-VK.* (only Amber test cases)
Change-Id: I5f363b7509a72e82dfcc4b3ad37891908a34a5cf