Samuel Iglesias Gonsálvez [Mon, 13 Jan 2020 09:37:56 +0000 (10:37 +0100)]
Add dual-source blending tests
Components: Vulkan, Framework
VK-GL-CTS issue: 2157
Added tests:
dEQP-VK.pipeline.blend.dual_source.*
Change-Id: I51f2660590615e4829ca45765d1f387ae575178b
Signed-off-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Alexander Galazin [Thu, 26 Mar 2020 09:09:37 +0000 (10:09 +0100)]
Merge vk-gl-cts/opengl-es-cts-3.2.6 into vk-gl-cts/master
Change-Id: I8808cedf8c77bde2e9290802edb09f4225b0f00d
Boris Zanin [Wed, 18 Mar 2020 14:54:06 +0000 (15:54 +0100)]
Squash ext-dep-target into gen-inl-files
The check_build_sanity.py target gen-inl-files unites
several generation scripts and gen_ext_deps.py should be
one of them instead of having separate target.
This will reduce CI server load.
Components: Vulkan
VK-GL-CTS Issue: 1340
Change-Id: If2153fcbc9cb779bdb86934c0a3c8a6b260e6a5f
Slawomir Cygan [Fri, 28 Feb 2020 12:52:01 +0000 (13:52 +0100)]
Add support for properly styling SPIR-V sources from XML logs
This fixes missing line endings of SPIR-V sources in XSLT-ized XML logs.
I have tested that by doing XSLT in recent Chrome version.
Components: Framework
VK-GL-CTS Issue: 2246
Change-Id: Ie9ef0b7cad1c5789c35cea599c145a524d123b33
James Hauxwell [Thu, 27 Feb 2020 14:23:12 +0000 (14:23 +0000)]
fix deqp reference texture lookup not adhearing to the ES2.0 spec language
"If the magnification filter is given by LINEAR and the minification
filter is given by NEAREST_MIPMAP_NEAREST or NEAREST_MIPMAP_LINEAR,
then c = 0.5. This is done to ensure that a minified texture does not
appear “sharper” than a magnified texture. Otherwise c = 0"
This was removed in ES3.0
The reference rasterizer is ES3, so it needs to get the version of the
device under test (DUT) and change its behaviour acordingly
Components: GLES2
VK-GL-CTS issue: #890
Affects:
dEQP-GLES2.functional.texture.filtering.2d.nearest_mipmap_nearest_linear_clamp_rgba8888
dEQP-GLES2.functional.texture.filtering.2d.nearest_mipmap_nearest_linear_repeat_rgba8888
dEQP-GLES2.functional.texture.filtering.2d.nearest_mipmap_nearest_linear_mirror_rgba8888
dEQP-GLES2.functional.texture.filtering.2d.nearest_mipmap_linear_linear_clamp_rgba8888
dEQP-GLES2.functional.texture.filtering.2d.nearest_mipmap_linear_linear_repeat_rgba8888
dEQP-GLES2.functional.texture.filtering.2d.nearest_mipmap_linear_linear_mirror_rgba8888
dEQP-GLES2.functional.texture.filtering.2d.nearest_mipmap_nearest_linear_repeat_etc1
dEQP-GLES2.functional.texture.filtering.2d.nearest_mipmap_nearest_linear_mirror_etc1
dEQP-GLES2.functional.texture.filtering.2d.nearest_mipmap_linear_linear_repeat_etc1
dEQP-GLES2.functional.texture.filtering.2d.nearest_mipmap_linear_linear_mirror_etc1
dEQP-GLES2.functional.texture.vertex.cube.filtering.nearest_mipmap_nearest_linear_clamp
dEQP-GLES2.functional.texture.vertex.cube.filtering.nearest_mipmap_nearest_linear_repeat
dEQP-GLES2.functional.texture.vertex.cube.filtering.nearest_mipmap_nearest_linear_mirror
dEQP-GLES2.functional.texture.vertex.cube.filtering.nearest_mipmap_linear_linear_clamp
dEQP-GLES2.functional.texture.vertex.cube.filtering.nearest_mipmap_linear_linear_repeat
dEQP-GLES2.functional.texture.vertex.cube.filtering.nearest_mipmap_linear_linear_mirror
Change-Id: I43024a13e0f63eb929da60d02c9722900c46506c
Graeme Leese [Thu, 20 Jun 2019 16:22:37 +0000 (17:22 +0100)]
Single random float generator for precision tests
Essentially the same random generator was written out three times, once
for double, once for float and once for fp16. The FloatFormat class
already takes care of the precision requirements, etc. so the same
function can be used for all three.
Components: Vulkan
Affects: dEQP-VK.glsl.builtin.precision*
Change-Id: I4c29eac8f8f71ddf20c2f7e12afc314eba71442d
Jussi Venho [Thu, 9 Jan 2020 11:02:38 +0000 (13:02 +0200)]
Add tests for scissor constraint
Test that scissor is correctly clamped to the framebuffer size.
New tests:
dEQP-VK.draw.scissor.static_scissor_framebuffer_border_in
dEQP-VK.draw.scissor.dynamic_scissor_framebuffer_border_in
dEQP-VK.draw.scissor.static_scissor_framebuffer_border_out
dEQP-VK.draw.scissor.dynamic_scissor_framebuffer_border_out
dEQP-VK.draw.scissor.static_scissor_oversized
dEQP-VK.draw.scissor.dynamic_scissor_oversized
Components: Vulkan
VK-GL-CTS issue: 2095
Change-Id: I7e15e4db12380c84cd7d7297de62bca0a7551f2e
Alexander Galazin [Wed, 25 Mar 2020 17:48:41 +0000 (18:48 +0100)]
Merge vk-gl-cts/vulkan-cts-1.2.2 into vk-gl-cts/master
Change-Id: I9ffc85a4101eba5770a31adde44c406686b76bb2
Alexander Galazin [Wed, 25 Mar 2020 17:48:39 +0000 (18:48 +0100)]
Merge vk-gl-cts/vulkan-cts-1.2.1 into vk-gl-cts/vulkan-cts-1.2.2
Change-Id: Ib397408e9bba043a187708d5d439507df3da6d14
Alexander Galazin [Wed, 25 Mar 2020 17:48:35 +0000 (18:48 +0100)]
Merge vk-gl-cts/vulkan-cts-1.2.0 into vk-gl-cts/vulkan-cts-1.2.1
Change-Id: If3617a4a3902a2343a8efee197e8bcd895c60f49
Alexander Galazin [Wed, 25 Mar 2020 11:19:50 +0000 (12:19 +0100)]
Merge vk-gl-cts/vulkan-cts-1.1.5 into vk-gl-cts/vulkan-cts-1.2.0
Change-Id: Iff8de9b31b571ac8606d6b6530bf19b21e3c63ac
Ari Suonpaa [Tue, 3 Mar 2020 13:03:58 +0000 (15:03 +0200)]
Added a second batch of coverage tests found by GraphicsFuzz
New tests:
dEQP-VK.graphicsfuzz.cov-basic-block-discard-in-function
dEQP-VK.graphicsfuzz.cov-const-folding-clamp-inside-while
dEQP-VK.graphicsfuzz.cov-const-folding-clamp-vs-original
dEQP-VK.graphicsfuzz.cov-const-folding-min-as-loop-range
dEQP-VK.graphicsfuzz.cov-constant-folding-atan-over-tanh
dEQP-VK.graphicsfuzz.cov-dead-branch-func-return-arg
dEQP-VK.graphicsfuzz.cov-fold-and-in-for-loop-range
dEQP-VK.graphicsfuzz.cov-folding-clamp-cmp-const-first
dEQP-VK.graphicsfuzz.cov-folding-rules-merge-div-mul
dEQP-VK.graphicsfuzz.cov-folding-rules-merge-divs
dEQP-VK.graphicsfuzz.cov-folding-rules-merge-mul-div
dEQP-VK.graphicsfuzz.cov-folding-rules-negate-div
dEQP-VK.graphicsfuzz.cov-inline-pass-return-in-loop
dEQP-VK.graphicsfuzz.cov-machinevaluetype-one-iter-loop
dEQP-VK.graphicsfuzz.cov-vector-log2-cosh
Components: Vulkan
Change-Id: Ifcb87053f5ff52ee5b0f7460f289e3e144e31a1a
Alexander Galazin [Wed, 25 Mar 2020 10:09:47 +0000 (11:09 +0100)]
Merge vk-gl-cts/vulkan-cts-1.2.1 into vk-gl-cts/vulkan-cts-1.2.2
Change-Id: I4e83df12355f93f835da9812a28665f434459a5d
Alexander Galazin [Wed, 25 Mar 2020 09:06:35 +0000 (10:06 +0100)]
Merge vk-gl-cts/vulkan-cts-1.1.6 into vk-gl-cts/vulkan-cts-1.2.1
Change-Id: I561526a8e0cc985c8183688ee4af633a6e91fa8b
Alexander Galazin [Wed, 25 Mar 2020 07:46:07 +0000 (08:46 +0100)]
Merge vk-gl-cts/vulkan-cts-1.1.5 into vk-gl-cts/vulkan-cts-1.1.6
Change-Id: I36c436cd72ec705b74e162b7bd11aff4a684f123
Alexander Galazin [Mon, 23 Mar 2020 11:35:35 +0000 (12:35 +0100)]
Merge vk-gl-cts/github-master into vk-gl-cts/master
Change-Id: I6aa744934228c4dd877a41a50e26247baede5053
Alexander Galazin [Mon, 23 Mar 2020 11:34:36 +0000 (12:34 +0100)]
Merge pull request #199 from aeubanks/cast
Cast pointer to deUintptr instead of long
Alexander Galazin [Mon, 23 Mar 2020 09:45:20 +0000 (10:45 +0100)]
Merge pull request #197 from baryluk/patch-1
Fix a typo for Linux run of deqp-vk
Matthew Netsch [Fri, 20 Mar 2020 15:16:23 +0000 (11:16 -0400)]
Fixes 16bit storage check in spirv_assembly tests
SPIRV shaders were still using 16bit storage extension
when API disabled it
Components: Vulkan
VK-GL-CTS Issue: 2272
Affects:
dEQP-VK.spirv_assembly.instruction.graphics.*convert*
Change-Id: I0135b7b6c4e84cbfc7afb050e5ecca4d5c4e82a2
Graeme Leese [Fri, 20 Mar 2020 13:30:31 +0000 (13:30 +0000)]
More vertex-invocation fixes for synchronisation tests
The fix for multiple vertex-shader instances was missed when the
timeline semaphore version of these tests was added. This is the fix
from https://gerrit.khronos.org/c/4451 applied to these new tests.
Components: Vulkan
VK-GL-CTS issue: 2253, 1958
Affects: dEQP-VK.synchronization.op.*.timeline_semaphore.write_indirect_buffer_draw*
Change-Id: I6013ef88a0acd362948c179416928c1d3fc064dd
Boris Zanin [Thu, 19 Mar 2020 13:55:22 +0000 (14:55 +0100)]
Integrate vulkan_core.h and vulkan_beta.h into master
Integrates changes from https://github.com/KhronosGroup/Vulkan-Headers/
commit
fb7f9c9bcd1d1544ea203a1f3d4253d0e90c5a90.
Components: Framework, Vulkan
VK-GL-CTS issue: 2270
Change-Id: I99990212504a64445aff5ef1023af4d1f66e4329
Alexander Galazin [Thu, 19 Mar 2020 10:20:22 +0000 (11:20 +0100)]
Merge vk-gl-cts/vulkan-cts-1.2.2 into vk-gl-cts/master
Change-Id: I7e5f20b43320020302f9735eeda726dcabcb8b3b
Alexander Galazin [Thu, 19 Mar 2020 08:54:16 +0000 (09:54 +0100)]
Merge vk-gl-cts/github-master into vk-gl-cts/master
Change-Id: I865b9f8d9808556dfab94bdbcac5c2f8287a6e9b
Ricardo Garcia [Wed, 4 Mar 2020 11:47:47 +0000 (12:47 +0100)]
Test nearest sampling edge cases for OpenGL(ES)
Add tests to sample textures close to texel edges in GL_NEAREST mode, in
order to verify texel coordinates are properly truncated.
New tests:
KHR-GL*.nearest_edge.*
Components: OpenGL
VK-GL-CTS issue: 2240
Change-Id: If9b32e425c68eb912dda64a7c559a5f5bcdff760
Alexander Galazin [Thu, 19 Mar 2020 08:51:38 +0000 (09:51 +0100)]
Merge pull request #196 from leix1234/astc_hdr_check
add a check about "astc_hdr"
sfricke_samsung [Sun, 8 Mar 2020 12:23:49 +0000 (05:23 -0700)]
Add check for YCbCr Chroma Reconstruction Explicit format support
If the format being queried supports
VK_FORMAT_FEATURE_SAMPLED_IMAGE_YCBCR_CONVERSION_CHROMA_RECONSTRUCTION_EXPLICIT_BIT
it must also support
VK_FORMAT_FEATURE_SAMPLED_IMAGE_YCBCR_CONVERSION_CHROMA_RECONSTRUCTION_EXPLICIT_FORCEABLE_BIT
Components: Vulkan
Affects: dEQP-VK.api.info.format_properties.*
VK-GL-CTS issue: 2247
Change-Id: Id771210bb7bb0445d2ec326a7cd706f1b053f96b
Ricardo Garcia [Fri, 6 Mar 2020 11:28:32 +0000 (12:28 +0100)]
Test multiple sets with dynamic offsets
This commit modifies the dynamic offset tests with two additional
strategies for grouping dynamic descriptors. Originally, these tests put
all descriptors in a single set.
Now, there is a second strategy in which every descriptor is put in a
separate set (at most 4 descriptors are used, so the tests never go over
maxBoundDescriptorSets in a conformant implementation).
A third strategy groups descriptors in arrays at binding zero, using one
set for dynamic descriptors and one set for nondynamic descriptors.
Affected and new tests:
dEQP-VK.pipeline.dynamic_offset.*
Components: Vulkan
VK-GL-CTS issue: 2248
Change-Id: I2e5d55b166f7e4424ba7ff7d6389d308da408eeb
Samuel Iglesias Gonsálvez [Fri, 6 Mar 2020 11:31:40 +0000 (12:31 +0100)]
Add tests for VK_AMD_shader_explicit_vertex_parameter
Components: Vulkan
VK-GL-CTS issue: 2254
Added tests:
dEQP-VK.draw.explicit_vertex_parameter.*
Signed-off-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Change-Id: I3e3183d15b607ec4c55fd21b3227818e610e15a3
Ricardo Garcia [Thu, 20 Feb 2020 15:30:51 +0000 (16:30 +0100)]
Test not/xor SPIR-V operation sequence
Add simple Amber test case to check the bitwise negation of a bitwise
xor result is properly calculated, as this combination appears to be
optimized by some implementations.
New test:
dEQP-VK.glsl.combined_operations.notxor
Components: Vulkan
VK-GL-CTS issue: 2233
Change-Id: I8cff49e47d9c8ce33174e2c0afcd3908198f3c17
Piotr Byszewski [Mon, 10 Feb 2020 11:50:21 +0000 (12:50 +0100)]
Test filtering of YCbCr formats
This change ads min/mag linear filtering tests for two YCbCr formats:
g8_b8_r8_3plane_420_unorm, g8_b8r8_2plane_420_unorm.
Components: Vulkan
VK-GL-CTS issue: 2185
Affects:
dEQP-VK.ycbcr.filtering.*
Change-Id: If690b66a6d0ed982c4ea78da5073b78f0ea75dc9
Piotr Byszewski [Wed, 15 Jan 2020 16:14:24 +0000 (17:14 +0100)]
Test winding order of primitive topologies
Components: Vulkan
VK-GL-CTS issue: 1713
Affects:
dEQP-VK.transform_feedback.simple.winding_*
Change-Id: Ia5878614dd3e380e6b220d2be777670de2898e58
Hyunjun Ko [Mon, 16 Mar 2020 08:16:23 +0000 (17:16 +0900)]
Skip dEQP-VK.transform_feedback.fuzz.*.geometry if geometry not supported.
Affects: dEQP-VK.transform_feedback.fuzz.*.geometry
Components: Vulkan
VK-GL-CTS issue: 2267
Change-Id: Ifb0d29c049ccaa686132e2bb9c24fe7c797f5f6c
Boris Zanin [Tue, 17 Mar 2020 08:28:55 +0000 (09:28 +0100)]
Fix verifyMultisampleLineGroupRasterization
Incorrect resolution has damaged verifcation function
verifyMultisampleLineGroupRasterization forcing it
to work always in strict mode causing tests to report
false failures.
Affected tests:
* dEQP-VK.rasterization.*
Components: Vulkan
VK-GL-CTS issue: 1024
Change-Id: Id079cd4d25f9b2f21a94d8443b9dce2679fcbc8c
Arthur Eubanks [Wed, 18 Mar 2020 16:15:26 +0000 (09:15 -0700)]
Cast to deUintptr instead
Arthur Eubanks [Wed, 18 Mar 2020 15:58:04 +0000 (08:58 -0700)]
Cast pointer to uintptr_t instead of long
long may not be able to fit a pointer, and C++ does not allow casting
from a pointer to a smaller int.
Alexander Galazin [Tue, 17 Mar 2020 09:31:59 +0000 (10:31 +0100)]
Merge vk-gl-cts/vulkan-cts-1.2.1 into vk-gl-cts/vulkan-cts-1.2.2
Change-Id: I5be459898168e3ac3c2365a970201ed738ef257a
Slawomir Cygan [Fri, 28 Feb 2020 14:36:58 +0000 (15:36 +0100)]
Allow not flusing denorms in operands
This fixes tests for denorm value flush to zero in Ceil(), Log(),
Log2(), Sqrt(), Sqrt2() operations. The test was assuming, that
operands of denormal value are flushed before the operation, however
the specification does not require that:
"If the entry point is declared with the DenormFlushToZero execution mode then for the
affected instuctions the denormalized result must be flushed to zero and the denormalized
operands **may** be flushed to zero."
VK-GL-CTS Issue: 2242
Affects: dEQP-VK.spirv_assembly.instruction.compute.float_controls.*.input_args.*_denorm_flush_to_zero
Components: Vulkan
Change-Id: Iedeb0d18fde3749a83be3262dba5efabf365aea2
Joerg Wagner [Sat, 14 Mar 2020 12:42:17 +0000 (13:42 +0100)]
Fix build for older GCCs
GCC 4.9 doesn't understand aggregate initialization
syntax introduced in 3 files:
- vktBindingDescriptorUpdateTests.cpp
- vktPipelineMultisampleShaderBuiltInTests.cpp
- vktPipelineSamplerTests.cpp
Align test case constructors with the rest of
the code base by avoiding the feature.
Affects:
dEQP-VK.binding_model.descriptor_update.*
dEQP-VK.pipeline.multisample_shader_builtin.*
dEQP-VK.pipeline.sampler.*
Components: Vulkan
VK-GL-CTS issue: 2265
Change-Id: I908baf308d62586e7d9baf8d389b2ce599e8912e
Alexander Galazin [Fri, 13 Mar 2020 12:43:27 +0000 (13:43 +0100)]
Whitelst Vulkan CTS 1.2.2.0
Affects: dEQP-VK.api.driver_properties.properties
Components: Vulkan
Change-Id: I395c3a1739e770b62c63d11bb8accc4535a6767c
Alexander Galazin [Fri, 13 Mar 2020 12:40:27 +0000 (13:40 +0100)]
Merge vk-gl-cts/vulkan-cts-1.2.1 into vk-gl-cts/vulkan-cts-1.2.2
Change-Id: Ib9d2d64c02625ed43632c2a6b35afdf8d71cafcf
Jeff Bolz [Mon, 24 Feb 2020 20:18:27 +0000 (14:18 -0600)]
Add more tests for VK_BUFFER_CREATE_DEVICE_ADDRESS_CAPTURE_REPLAY_BIT
Existing tests had all buffers use the same small size, and
missed bugs in our implementation. Add tests using larger,
random sizes.
New Tests: dEQP-VK.binding_model.buffer_device_address.capture_replay_stress.*
Component: Vulkan
Change-Id: I846bcac16a28502902a70886b9e0949c8c2f1575
Ari Suonpaa [Fri, 13 Mar 2020 11:58:02 +0000 (13:58 +0200)]
Removed a test using uninitialized variable
The use of uninitialized variable is being debated in
a spec issue: https://gitlab.khronos.org/spirv/SPIR-V/issues/490.
Removed dEQP-VK.graphicsfuzz.cov-reduce-load-array-replace-extract
because the spec is currently not clear if the test is correct
or not.
VK-GL-CTS issue: 2258
Components: Vulkan
Change-Id: I06736154745d73d6a914d2b814c1aa49e8bb3a86
Ricardo Garcia [Mon, 9 Mar 2020 15:13:59 +0000 (16:13 +0100)]
Require 16-bit storage in some 8-bit storage tests
This commit adds uniform 16-bit storage features to the list of
requested features for 8-bit storage tests that use the SPIR-V
StorageUniform16 capability.
Affected tests:
dEQP-VK.spirv_assembly.instruction.compute.8bit_storage.storagebuffer_16_to_8.*
dEQP-VK.spirv_assembly.instruction.compute.8bit_storage.uniform_8_to_16.*
dEQP-VK.spirv_assembly.instruction.compute.8bit_storage.push_constant_8_to_16.*
dEQP-VK.spirv_assembly.instruction.graphics.8bit_storage.storagebuffer_int_16_to_8.*
dEQP-VK.spirv_assembly.instruction.graphics.8bit_storage.uniform_int_8_to_16.*
dEQP-VK.spirv_assembly.instruction.graphics.8bit_storage.push_constant_int_8_to_16.*
Components: Vulkan
VK-GL-CTS issue: 2256
Change-Id: I12535b8529f139718d8e6a8add7e3f24231319f7
amhagan [Fri, 6 Mar 2020 14:08:25 +0000 (09:08 -0500)]
Fix barrier code in dEQP-VK.api.buffer_marker.* tests
Fix RAW barrier hazard in the buffer_marker tests.
Change the destination stage to TOP_OF_PIPE so that the buffer marker
is finished writing before being read.
Components: Vulkan
VK-GL-CTS issue: 2255
Modified tests:
dEQP-VK.api.buffer_marker.*.external_host_mem.bottom_of_pipe.*
Change-Id: I5f17199240f8a43337d9b6ea2fbaa0e540dcc3be
Boris Zanin [Mon, 8 Apr 2019 06:34:49 +0000 (08:34 +0200)]
Add tests for non strict lines
Tests for strict and non-strict lines rasterization:
* dEQP-VK.rasterization.*.strict_*
* dEQP-VK.rasterization.*.non_strict_*
Components: Vulkan
VK-GL-CTS issue: 1024
Change-Id: If2c4dbe359be0e2550b063216dbd427f80c980db
Venni Ihanakangas [Tue, 7 Jan 2020 08:43:27 +0000 (10:43 +0200)]
dEQP-VK.graphicsfuzz.vectors-and-discard-in-function
Components: Vulkan
New Tests:
dEQP-VK.graphicsfuzz.vectors-and-discard-in-function
Change-Id: I11e738b1d5e59c58ef83005a6ac9374b292d8af5
Slawomir Cygan [Mon, 2 Mar 2020 13:35:30 +0000 (14:35 +0100)]
Fix dEQP-VK.subgroups.ballot.framebuffer.* to use core API
This test group was mistakenly using VK_EXT_shader_subgroups_ballot
extension and int64 feature, effectively duplicating the
dEQP-VK.subgroups.ballot.ext_shader_subgroup_ballot.framebuffer test
group.
This fix makes this group use core API.
Components: Vulkan
VK-GL-CTS Issue: 2249
Affects: dEQP-VK.subgroups.ballot.framebuffer.*
Change-Id: Id884d6e5ea46fe1eae749e2aec123a5863b2cc81
Ricardo Garcia [Thu, 27 Feb 2020 16:30:52 +0000 (17:30 +0100)]
Test OpUndef with OpConstantComposite and OpSpecConstantComposite
This commit tests OpUndef can be used to create partially defined
constant composite structures by using OpUndev values when constructing
other values with OpConstantComposite and OpSpecConstantComposite.
New tests:
dEQP-VK.spirv_assembly.instruction.compute.opundef.undefined_*
Components: Vulkan
VK-GL-CTS issue: 2243
Change-Id: I518dd029b4e00a5d5d33603d4776ce643c2adba8
Piers Daniell [Wed, 4 Mar 2020 06:29:28 +0000 (23:29 -0700)]
OpenGL ES CTS fixes for passing on WGL
When running cts-runner --type=es32 on a WGL platform it will pick up several
pixel formats for the default framebuffer that aren't supported correctly
by the existing OpenGL ES 3.2 tests.
1. OpenGL ES 3.2 CTS expects the platform to expose rgb565, but that is not
available on WGL and should not be a requirement of OpenGL ES 3.2
conformance. This CL disables rgb565 is a requirement for all non-Android
platforms, which extends the change originally made for X11 in
https://gerrit.khronos.org/#/c/4615/ for issue VK-GL-CTS issue: 1474.
2. The 10-bit+ bits per pixel component WGL pixel formats are skipped since
the existing OpenGL ES 3.2 tests, which use pixel transfer functions
like glReadPixels, and assume an 8-bit type, don't work correctly since
glReadPixels is not defined to do format and type conversions like it
does with OpenGL.
3. Default framebuffers with sRGB color space are skipped because
under OpenGL ES sRGB encoded will default for on in this case and the
existing OpenGL ES tests do not currently support a non-linear
default framebuffer. This is not a problem with OpenGL since with
OpenGL sRGB encoding has to be enabled explicitly.
Affects: cts-runner --type=es32 on WGL
Components: OpenGL
VK-GL-CTS issue: 2257
Change-Id: If12aff8c10571b7c0f72da618b13753c340a60c4
Venni Ihanakangas [Mon, 13 Jan 2020 13:35:27 +0000 (15:35 +0200)]
dEQP-VK.graphicsfuzz.always-discarding-function
Components: Vulkan
New Tests:
dEQP-VK.graphicsfuzz.always-discarding-function
Change-Id: Ia921f7fdf9fe211787959af4d499eb20bfb5dbb5
Ricardo Garcia [Mon, 9 Mar 2020 15:45:16 +0000 (16:45 +0100)]
Update Android mustpass
Change-Id: I87ef66e9cd6aeae9b47360ab44f4cbe70334c3a2
Ari Suonpaa [Fri, 14 Feb 2020 07:51:28 +0000 (09:51 +0200)]
Make derivative tests to have dependency to both axes
Derivative tests had zero values for x when doing dfdy
and y when doing dfdx. Modified varyings and a texture
to vary in both x and y directions for both axes
to avoid these zeros.
Affects:
dEQP-GLES3.functional.shaders.derivate.*
Components: OpenGL ES
VK-GL-CTS issue: 1865
Change-Id: If36a4d857167ae7eede2b8f6b0c26308b7e5fae9
James Glanville [Fri, 17 Jan 2020 15:33:07 +0000 (15:33 +0000)]
Exclude configs for EGL_YUV_BUFFER_EXT color buffer
Test dEQP-EGL.functional.get_frame_timestamps.other fails due to
mismatch between the format of fragment shader output and framebuffer
for YUV configs with EGL_COLOR_BUFFER_TYPE == EGL_YUV_BUFFER_EXT.
Exclude from the other list.
Components: EGL
VK-GL-CTS issue: 2187
Change-Id: I08e3466bb0721e6b424a07a09587b56eb28d6cff
Graeme Leese [Fri, 17 Jan 2020 12:06:15 +0000 (12:06 +0000)]
Verify results in ballot bitcount tests
The old code only checked whether or not the results of these operations
were zero. The non-zero results were never checked for correctness. Add
checks that these results are actually correct.
Components: Vulkan
VK-GL-CTS issue: 2192
Affects: dEQP-VK.subgroups.ballot_other.*bitcount*
Change-Id: Icec4f558f556bc126b95bd925455620c5d8815f7
Jussi Venho [Mon, 3 Feb 2020 14:44:59 +0000 (16:44 +0200)]
Add depth and stencil formats to buffer copy tests
Test copy from buffer to image with offset, depth and stencil formats.
New tests:
dEQP-VK.api.copy_and_blit.*.buffer_to_depthstencil.buffer_offset_*
Components: Vulkan
VK-GL-CTS issue: 2140
Change-Id: I41bb79e0a52cc4ca0a32b67815df5f6c6b8ebd47
Samuel Iglesias Gonsálvez [Mon, 27 Jan 2020 08:56:14 +0000 (09:56 +0100)]
Add concurrent query pool tests
Components: Vulkan
VK-GL-CTS issue: 2195
Added tests:
dEQP-VK.query_pool.concurrent_queries.*
Change-Id: I6852ce9865f2bbceace5d0df648e050d68b88bbb
Witold Baryluk [Fri, 6 Mar 2020 17:41:49 +0000 (17:41 +0000)]
Fix a typo for Linux run of deqp-vk
Unrecognized command line option
'--deqp-vk-caselist-file=/home/user/VK-GL-CTS/external/vulkancts/mustpass/master/vk-default.txt'
It should be `--deqp-caselist-file`
Arkadiusz Sarwa [Thu, 27 Feb 2020 11:36:09 +0000 (12:36 +0100)]
Add VK_KHR_shader_non_semantic_info to ext list
VK-GL-CTS issue: 2241
Components: Vulkan, Framework
Affects:
dEQP-VK.info.device_extensions
Change-Id: I1303a3bc9fac2dd09e0902c20f378b13816d1695
(cherry picked from commit
3ea0200411dcc5a190d9f68bc06939efe29d2f55)
Piotr Byszewski [Thu, 14 Nov 2019 13:20:35 +0000 (14:20 +0100)]
Add tests for VK_KHR_shader_non_semantic_info
VK-GL-CTS issue: 2076
Components: Vulkan
Affects:
dEQP-VK.spirv_assembly.instruction.compute.non_semantic_info.*
Change-Id: I0554c1a3dea49a66d28d1058b3fa63cc4a378a1f
(cherry picked from commit
a1c717411059c366058008e86c94d9f8d9b01d1a)
Alexander Galazin [Thu, 5 Mar 2020 12:08:11 +0000 (13:08 +0100)]
Whitelist Vulkan CTS 1.2.1.1
Affects: dEQP-VK.api.driver_properties.properties
Components: Vulkan
Change-Id: I433638426b621635a2955e2f5fc40102a0938fa8
Alexander Galazin [Thu, 5 Mar 2020 11:48:26 +0000 (12:48 +0100)]
Update Android mustpass
Change-Id: I9f650303ef470f85bf0343f525c242db39fe6360
Ricardo Garcia [Thu, 13 Feb 2020 12:17:40 +0000 (13:17 +0100)]
Test SampledImage as function argument
Add test to check sampled images work properly when used as function
arguments.
New test:
dEQP-VK.spirv_assembly.instruction.function_params.sampler_param
Components: Vulkan
VK-GL-CTS issue: 2218
Change-Id: I4b760cf0e5cf8ad980fd33e9b237778a2451a243
Ari Suonpaa [Tue, 4 Feb 2020 12:21:53 +0000 (14:21 +0200)]
Added a batch of coverage tests found by GraphicsFuzz
New tests:
dEQP-VK.graphicsfuzz.cov-*
Components: Vulkan
Change-Id: I24ac5446fda48f9000b261e83e43dab44eb2c832
Matthew Netsch [Fri, 21 Feb 2020 21:12:45 +0000 (16:12 -0500)]
Add 8bit support checks to subgroup tests
Components: Vulkan
VK-GL-CTS Issue: 2082
Affects:
dEQP-VK.subgroups.*
Change-Id: I39367c6ee2d868330dac563a78309b18100ecf08
Alexander Galazin [Wed, 4 Mar 2020 10:06:53 +0000 (11:06 +0100)]
Merge vk-gl-cts/vulkan-cts-1.2.1 into vk-gl-cts/master
Change-Id: Iab703c04c7b00f7d51a15ebb708ef06c151e1448
Alexander Galazin [Wed, 4 Mar 2020 10:06:52 +0000 (11:06 +0100)]
Merge vk-gl-cts/vulkan-cts-1.2.0 into vk-gl-cts/vulkan-cts-1.2.1
Change-Id: Ia84f84f07054a5c6c175fdf152c669fb3a186690
Alexander Galazin [Wed, 4 Mar 2020 10:06:52 +0000 (11:06 +0100)]
Merge vk-gl-cts/vulkan-cts-1.1.6 into vk-gl-cts/vulkan-cts-1.2.1
Change-Id: I360bd965daf24ebd38a05a5205d3a85311a60bfd
Alexander Galazin [Wed, 4 Mar 2020 10:06:51 +0000 (11:06 +0100)]
Merge vk-gl-cts/vulkan-cts-1.1.5 into vk-gl-cts/vulkan-cts-1.1.6
Change-Id: Ifeab9cfd06891288494e9080d40819c62a2092a7
Alexander Galazin [Wed, 4 Mar 2020 09:09:29 +0000 (10:09 +0100)]
Merge vk-gl-cts/vulkan-cts-1.1.5 into vk-gl-cts/vulkan-cts-1.2.0
Change-Id: I710f52c5271173e7592c92db91a6debd1ebfff7c
Alexander Galazin [Wed, 4 Mar 2020 07:50:07 +0000 (08:50 +0100)]
Merge vk-gl-cts/vulkan-cts-1.1.4 into vk-gl-cts/vulkan-cts-1.1.5
Change-Id: I6221b31126a23415f4aceb05f91fc8c7b166c876
Piers Daniell [Fri, 21 Feb 2020 18:52:11 +0000 (11:52 -0700)]
Allow 1ULP of tolerance in ldexp precision tests
https://gerrit.khronos.org/#/c/4726/ modified how the expected values were
calculated for the ldexp() precision tests. This change caused previously
passing platforms to now fail for double precision, which were added
by https://gerrit.khronos.org/#/c/4355/.
This CL adds 1ULP to the expected value interval to allow for a possible
difference in rounding mode between the graphics device and host machine
ldexp implementations.
Components: Vulkan
Change-Id: If92047f0a421624ed853a9690f3e3980e5cd79ef
Affects: dEQP-VK.glsl.builtin.precision*.ldexp.*
Ricardo Garcia [Fri, 14 Feb 2020 12:37:43 +0000 (13:37 +0100)]
Fix 32-bit statistic query tests using 64 bits
This commit fixes the 32/64-bit split in the statistics query group that
tried to use vkGetQueryPoolResults with different bit widths but were
actually always using 64-bits.
Note: the intermediate reset results that are copied to a buffer in some
tests always use the 64-bit flag. This is expected.
Affected tests:
dEQP-VK.query_pool.statistics_query.*
Components: Vulkan
VK-GL-CTS issue: 2215
Change-Id: Ib73ae9fbd8fbaf1ec9195090e20ec4392dc62444
xiaolei [Mon, 2 Mar 2020 14:30:19 +0000 (22:30 +0800)]
add a check about "astc_hdr"
Piotr Byszewski [Wed, 12 Feb 2020 14:57:16 +0000 (15:57 +0100)]
Restore disabled texture function tests
Commit restores 12 tests. Only 2 of those tests were now failing.
This happened because of hitting cubemap edge and was resolved by
modifying MaxDx test parameter (to the same value used in other
cube texture format tests).
Components: Vulkan
VK-GL-CTS issue: 864
Affects:
dEQP-VK.glsl.texture_functions.texturegrad.samplercubeshadow_*
Change-Id: I09e012463fb0e633409d0aa58913d666f555998a
Alexander Galazin [Thu, 27 Feb 2020 11:20:44 +0000 (12:20 +0100)]
Merge vk-gl-cts/vulkan-cts-1.1.3 into vk-gl-cts/vulkan-cts-1.1.4
Change-Id: Ie332664c1049e2e1339f60aa7dd2615d1aa83105
Sai Kiran Korwar [Wed, 19 Feb 2020 08:53:39 +0000 (14:23 +0530)]
Check if display is available for Vulkan.
If the display is already being used by native windowing system,
then it may not be available to Vulkan, which may cause the test
to fail.
To account for this, check if any windowing system is already present
and running and then report the test to be unsupported.
Since the extension is meant to support direct rendering to display,
avoid running it when a windowing system is present.
Components: Vulkan
VK-GL-CTS issue: #2230
Affects:
dEQP-VK.wsi.display_control.swapchain_counter
Change-Id: I7fd00b417021e4a2ef017b742e529f5a3917a056
(cherry picked from commit
5318e8023d181def3d6affe446ab251f9898737e)
Sai Kiran Korwar [Wed, 19 Feb 2020 08:53:39 +0000 (14:23 +0530)]
Check if display is available for Vulkan.
If the display is already being used by native windowing system,
then it may not be available to Vulkan, which may cause the test
to fail.
To account for this, check if any windowing system is already present
and running and then report the test to be unsupported.
Since the extension is meant to support direct rendering to display,
avoid running it when a windowing system is present.
Components: Vulkan
VK-GL-CTS issue: #2230
Affects:
dEQP-VK.wsi.display_control.swapchain_counter
Change-Id: I7fd00b417021e4a2ef017b742e529f5a3917a056
Aleksy Sokolowski [Wed, 12 Feb 2020 11:17:49 +0000 (12:17 +0100)]
Fix interface mismatch in case "spirv_validation_capabilities_test"
User variable texCoord in location = 3 was only used in VS and FS,
lack of the variable in TCS, TES and GS is causing interface mismatch.
VK-GL-CTS Issue: 2219
Affects: KHR-GL46.gl_spirv.spirv_validation_capabilities_test
Components: OpenGL
Change-Id: I5899906f89efb380896a277939b532f393d00fce
Ari Suonpaa [Wed, 12 Feb 2020 11:28:47 +0000 (13:28 +0200)]
Make sure readpixels work with 16bit fixed point surfaces
Added FBO tests that use unsigned fixed point textures
as render target. Pixels are read from the FBO using
pixel type GL_UNSIGNED_SHORT since that was reported
to have problems on some implementations.
New tests:
dEQP-GLES31.functional.fbo.color.tex2d.r16
dEQP-GLES31.functional.fbo.color.tex2d.rg16
dEQP-GLES31.functional.fbo.color.tex2d.rgba16
Components: OpenGL ES
VK-GL-CTS issue: 2124
Change-Id: Icad477ae62ff6961f5989b097aa3ce73002b8dd8
James Hauxwell [Wed, 12 Feb 2020 10:55:24 +0000 (10:55 +0000)]
Fix EGL1.5 pixmap features being tested on a EGL1.4 device
createWindowSurface() uses the blockout function checkEGL15Support() to
stop EGL1.5 functions being run. This is missing from the
createPixmapSurface() variant.
Components: EGL
VK-GL-CTS issue: 2217
Affects:
dEQP-EGL.functional.create_surface.platform_pixmap.*
Change-Id: Ifb234e0c683eb755de20b2755be80716c0ffeaa9
Slawomir Cygan [Tue, 18 Feb 2020 16:18:01 +0000 (17:18 +0100)]
Fix subgroups tests using basePipelineHandle with index != -1
The value of basePipelineIndex must be -1 when basePipelineHandle
is not null in VkComputePipelineCreateInfo structure.
Components: Vulkan
VK-GL-CTS Issue: 2228
Change-Id: Iabbb98753179a0db6153853b9959eba6fee19ed1
Affects: dEQP-VK.subgroups.*compute*
Mika Väinölä [Mon, 16 Dec 2019 16:24:14 +0000 (18:24 +0200)]
Don't enable both buffer device address extensions
VK_KHR_buffer_device_address deprecates VK_EXT_buffer_device_address.
VkDeviceCreateInfo.ppEnabledExtensionNames must not contain both
extensions.
Components: Framework, Vulkan
Affects:
dEQP-VK.binding_model.buffer_device_address.*
dEQP-VK.memory_model.message_passing.*.physbuffer.*
dEQP-VK.ssbo.phys.layout.*
dEQP-VK.compute.cooperative_matrix.*.physical_buffer.*
VK-GL-CTS issue: 2148
Change-Id: I88dbb3c4b4deb6f38f37ec0a7f40b1660aeffad5
Igor Ostrowski [Mon, 10 Feb 2020 10:28:52 +0000 (11:28 +0100)]
Fix memory leak in pipeline.executable_properties tests
Affects:
dEQP-VK.pipeline.executable_properties.*
Components: Vulkan
VK-GL-CTS issue: 2220
Change-Id: I3f82e17be51a643cd64270e7f05fdf496e648f00
Alexander Galazin [Thu, 27 Feb 2020 09:45:22 +0000 (10:45 +0100)]
Merge vk-gl-cts/vulkan-cts-1.2.1 into vk-gl-cts/master
Change-Id: I91967444cfbf1b86a9c4473e2bbd83dbbd241876
Alexander Galazin [Wed, 26 Feb 2020 14:36:27 +0000 (15:36 +0100)]
Merge vk-gl-cts/vulkan-cts-1.2.0 into vk-gl-cts/vulkan-cts-1.2.1
Change-Id: Ia9d1bd6dd594c1a9082e84a98c3b5ec506cfb988
Alexander Galazin [Wed, 26 Feb 2020 14:36:26 +0000 (15:36 +0100)]
Merge vk-gl-cts/vulkan-cts-1.1.6 into vk-gl-cts/vulkan-cts-1.2.1
Change-Id: I0f3ab098b8831b1c421745daa491d323a533f2f4
Alexander Galazin [Tue, 25 Feb 2020 09:56:29 +0000 (10:56 +0100)]
Merge vk-gl-cts/github-master into vk-gl-cts/master
Change-Id: I7649e7f482a31edb244999a81593fe99f006ab3f
Alexander Galazin [Tue, 25 Feb 2020 07:54:10 +0000 (08:54 +0100)]
Merge pull request #189 from imirkin/master
Improve support for KHR-GL33.* tests
dzabaj01 [Mon, 24 Feb 2020 12:39:21 +0000 (13:39 +0100)]
Fix dEQP-VK.geometry.layered.2d_array.readback
The patch fixes several image memory barriers that
are missing the (color or depth/stencil) read access
mask. The read access mask is needed because the
attachments are setup with the VK_ATTACHMENT_LOAD_OP_LOAD
load operation.
Components: Vulkan
VK-GL-CTS Issue: 2227
Affects: dEQP-VK.geometry.layered.2d_array.readback
Change-Id: I22b696f449cb83d88f3ec15d928577d6b3d65b6f
Alexander Galazin [Mon, 24 Feb 2020 10:31:54 +0000 (11:31 +0100)]
Merge vk-gl-cts/vulkan-cts-1.2.1 into vk-gl-cts/master
Change-Id: I44bb78fa8e39d0a2dfd9a0d6d08d7163d6eba76a
Graeme Leese [Thu, 13 Feb 2020 14:23:57 +0000 (14:23 +0000)]
Don't remove 16-bit denorms from fp64 tests
The fixed values that are generated for double precision precision tests
were incorrectly skipping any that corresponds to a 16-bit denorm. At
the moment I don't think that there are any such values so this isn't
(currently) a practical problem.
Components: Vulkan
VK-GL-CTS Issue: 2231
Affects: dEQP-VK.glsl.builtin.precision_double.*
Change-Id: Ic7a27974fba09c94c5b4ab77b30b56f0b673aced
Graeme Leese [Thu, 16 Jan 2020 14:23:22 +0000 (14:23 +0000)]
Stricter codomain for atan tests
If the codomain is going to be enforced strictly then we should update
the precision tests to check it. The atan precision test looks to have
had an issue applying the codomain incorrectly which meant that large
values of the arguments were not tested at all.
Components: Vulkan
VK-GL-CTS issue: 479
Affects: dEQP-VK.glsl.builtin.precision*atan.*
Change-Id: I842adb035cb59fe8ff1423f11acdad0a90ee1549
Graeme Leese [Thu, 30 Jan 2020 10:35:17 +0000 (10:35 +0000)]
Use common code for SG arithmetic-like tests
Factor out the parts of arithmetic, clustered, and partitioned subgroup
tests so that common code can be written where possible.
Components: Vulkan
Affects: dEQP-VK.subgroups.arithmetic.*
dEQP-VK.subgroups.clustered.*
dEQP-VK.subgroups.partitioned.*
Change-Id: I8b580c6b4f90229f2cb5a6324dc0c338709c3921
Ricardo Garcia [Mon, 20 Jan 2020 09:02:04 +0000 (10:02 +0100)]
Test additional atomic image SPIR-V operations
This commit adds tests to check OpAtomicIIncrement, OpAtomicIDecrement
and OpAtomicISub with images.
New tests:
dEQP-VK.image.atomic_operations.sub.*
dEQP-VK.image.atomic_operations.inc.*
dEQP-VK.image.atomic_operations.dec.*
Components: Framework, Vulkan
VK-GL-CTS issue: 2186
Change-Id: Ib2c5d5075555ad0be025b0498960b646852fdcd4
Vihanakangas [Fri, 24 Jan 2020 08:33:22 +0000 (10:33 +0200)]
Add three new GraphicsFuzz tests
Components: Vulkan
New Tests:
dEQP-VK.graphicsfuzz.call-function-with-discard
dEQP-VK.graphicsfuzz.color-set-in-for-loop
dEQP-VK.graphicsfuzz.switch-inside-while-always-return
Change-Id: Ibfc1c236f94d774d3f70818e6675e90bf7829b3c
Alastair Donaldson [Thu, 13 Feb 2020 11:19:55 +0000 (11:19 +0000)]
Updates to Android test list generation
Fixes to reflect changes on the Android side.
Change-Id: I4eece1d987c35e27046061adc79f9ecde755a56a
Thomas Spurden [Wed, 29 Jan 2020 14:58:52 +0000 (14:58 +0000)]
Don't assume CMAKE_SOURCE_DIR is the root of the vk-gl-cts git repo
If vk-gl-cts is included into a build using cmake's add_subdirectory
then CMAKE_SOURCE_DIR will be the root of the parent cmake project. This
results in the release name/id being taken from the git repo of whatever
includes vk-gl-cts via add_subdirectory rather than the vk-gl-cts git.
Change-Id: Iafc0d2458b58413bc5c55a413e38bedbec165394
James Fitzpatrick [Mon, 20 Jan 2020 09:02:24 +0000 (09:02 +0000)]
Correctly handle normalized formats in FloatFormat
For the texture explict lod tests normalized values were treated as 8
bit floating point formats when calculating the reference interval.
This ignores the fact that normalized formats are encoded as 1 / N,
where N is an integer, this results in the rounding used in tests not
being able to accurately round normalized formats.
This incorrect rounding causes the acceptable ranges to be
significantly smaller than other formats and extremely asymetric
around the ideal filtered value for certain samples.
Affects: dEQP-VK.texture.explicit_lod.*
Components: Vulkan
VK-GL-CTS issue: 2189
Change-Id: Ia4f3adefab4bfb5d9c91bd7033199b0e107179d5
Alexander Galazin [Thu, 20 Feb 2020 14:41:31 +0000 (15:41 +0100)]
Merge vk-gl-cts/vulkan-cts-1.1.5 into vk-gl-cts/vulkan-cts-1.1.6
Change-Id: If35028985c57f66a6fc3f3d11ac1a9485493f7f2
Alexander Galazin [Thu, 20 Feb 2020 13:37:54 +0000 (14:37 +0100)]
Merge vk-gl-cts/vulkan-cts-1.1.5 into vk-gl-cts/vulkan-cts-1.2.0
Change-Id: Ic305f1a757dae469f6b0fe801e82640b83f989ed