Chris Forbes [Thu, 15 Feb 2018 00:25:39 +0000 (16:25 -0800)]
Merge remote-tracking branch 'aosp/master' into deqp-dev
Change-Id: Iee4ea748158c78112727bf21b905432b62b4fca9
Chris Forbes [Wed, 14 Feb 2018 17:22:18 +0000 (09:22 -0800)]
Merge remote-tracking branch 'khronos/master' into deqp-dev
Change-Id: Ib2085147e9679d6dd16109c660f9777d680d54fe
Juan A. Suarez Romero [Tue, 30 Jan 2018 09:55:40 +0000 (10:55 +0100)]
Add more cases for block/non-block uniforms match
Cover more situations as described in GLSL 3.20 spec, section 4.3.9.
- It is a link-error when using the same name inside a block with non
instance name and outside a block, but it is correct if the block has
instance name.
- It is a link-error when using the same name inside two different
blocks with non instance name.
Components: OpenGL
VK-GL-CTS issue: 972
Affects:
KHR-GLES3.shaders.uniform_block.common.name_matching
KHR-GL33.shaders.uniform_block.common.name_matching
Change-Id: Ie91300685e807774a9cc9bbb9139936133a8ae1c
Piotr Byszewski [Mon, 29 Jan 2018 08:58:50 +0000 (09:58 +0100)]
Fix texture lookups tests
Some shaders didn't compile because of use of reserved word
textureCubeLod. This patch fixes this by checking if
ES implementation is ES3 capable and adjusting the shaders
generation parameters accordingly.
Components: OpenGL
VK-GL-CTS issue: 283
Affects:
dEQP-GLES2.functional.shaders.random.texture.vertex.*
Change-Id: Ib1e9983a9503780f992dcd00e8dc380a4f01bef4
Piotr Byszewski [Tue, 30 Jan 2018 14:19:34 +0000 (15:19 +0100)]
Fix invalid_assign_to_1 fragdata tests
For ES3 capable hardware compilation of shaders in
invalid_assign_to_1 test succeeds resulting in test failure.
This patch resolves that by adding only_glsl_es_100_support
and exactly_one_draw_buffer flags that indicate that test
shouldn't be executed on ES3 capable hardwere or on a
implementation that supports more then one draw buffer.
This change also removes redundant CaseRequirement structure
and parses requirements directly to ShaderCaseSpecification.
Components: OpenGL
VK-GL-CTS issue: 282
Affects:
dEQP-GLES2.functional.shaders.fragdata.invalid_assign_to_1
dEQP-GLES2.functional.shaders.*
dEQP-GLES3.functional.shaders.fragdata.invalid_assign_to_1
dEQP-GLES3.functional.shaders.*
Change-Id: I44949fde7ac9bc724965ec166ed55433c91857c2
Piotr Byszewski [Thu, 1 Feb 2018 12:08:10 +0000 (13:08 +0100)]
Fix texturecubelod test
This patch fixes issue with validation of texturecubelod
test result. Issue happened when test was executed on ES3
capable hardware.
Components: OpenGL
VK-GL-CTS issue: 280
Affects:
dEQP-GLES2.functional.shaders.texture_functions.vertex.*
Change-Id: I92d81ca93cb84785ea806155d0436f167c4fbf31
Juan A. Suarez Romero [Mon, 15 Jan 2018 11:10:46 +0000 (11:10 +0000)]
Check for depth/stencil buffer in default FBO
The affected tests invoke gl.getNamedFramebufferAttachmentParameteriv()
with GL_FRAMEBUFFER_ATTACHMENT_DEPTH_SIZE to get the depth bits used in
the default FBO. And something similar is done for the stencil bits.
But wrongly it assumes that if the FBO does not have an associated depth
or stencil buffer, it would return 0. Rather, this function returns an
error if it is invoked in a FBO without a depth/stencil buffer attached.
So better check first if the default FBO has a depth/stencil buffer
attached, and if so then invoke the above functions to get the
depth/stencil bits.
Components: OpenGL
VK-GL-CTS issue: 969
Affects:
KHR-*.internalformat.renderbuffer.*
Change-Id: I74d28d5d089790fed4dd2728cb5143e7c03d7442
Jari Komppa [Wed, 3 Jan 2018 11:00:55 +0000 (13:00 +0200)]
Added image blit tests with multi-part image lock
Added versions of the image blitting tests that lock the image resource
in multiple parts to stress the VK_REMAINING_MIP_LEVELS and
VK_REMAINING_ARRAY_LAYERS functionality.
Affects:
dEQP-VK.api.copy_and_blit.core.blit_image.all_formats.generate_mipmaps.from_previous_level.*
Components: Vulkan
VK-GL-CTS issue: 901
Change-Id: Ia7f80a0ff1b432af8fe4e032f7e6524785c90bf8
Paavo Pessi [Fri, 26 Jan 2018 13:33:37 +0000 (15:33 +0200)]
Add API info tests for YCbCr formats
The API feature info tests were extended to query physical device's
image format capabilities using YCbCr formats.
New tests:
dEQP-VK.api.info.image_format_properties.2d.*
dEQP-VK.api.info.image_format_properties2.2d.*
Components: Vulkan
VK-GL-CTS issue: 961
Change-Id: I04b948ff3f3f2af809c87c5f6c251385e5bba24f
Chris Forbes [Tue, 13 Feb 2018 21:27:05 +0000 (21:27 +0000)]
Merge "Merge remote-tracking branch 'origin/vulkan-cts-1.0.2' into HEAD" into oreo-cts-dev am:
6d6ebc1b4b
am:
cc01ada806
Change-Id: I6827e2294199817a66666ed5db941491f5462f2a
Chris Forbes [Tue, 13 Feb 2018 21:20:03 +0000 (21:20 +0000)]
Merge "Merge remote-tracking branch 'origin/vulkan-cts-1.0.2' into HEAD" into oreo-cts-dev
am:
6d6ebc1b4b
Change-Id: I71f98787bf29843efbe66343187c3a67c7ca39f9
Treehugger Robot [Tue, 13 Feb 2018 21:08:38 +0000 (21:08 +0000)]
Merge "Merge remote-tracking branch 'origin/vulkan-cts-1.0.2' into HEAD" into oreo-cts-dev
Piotr Byszewski [Tue, 23 Jan 2018 09:43:37 +0000 (10:43 +0100)]
Fix TCS negative test
One of three tests that are checking incorrectly sized tessellation
control output array was actually correct. This is fixed by changing
array size to value that is small then layout declaration. Remaining
two tests check if compilation fails when value is greater then
layout declaration.
Components: OpenGL
VK-GL-CTS issue: 297
Affects:
dEQP-GLES31.functional.tessellation.user_defined_io.negative.es32.per_vertex_incorrect_control_explicit_output_array_size_1
Change-Id: I3f06e36ab8bab4eae541204bdab5db1f5d698ff2
Kantoch [Fri, 9 Feb 2018 19:05:40 +0000 (19:05 +0000)]
DO NOT MERGE: Sparse resources fixes am:
ae793c389b -s ours am:
ce25f76fbb -s ours
am:
fe93d4f83c -s ours
Change-Id: I14929e52bc985f4d152628b48b4a6a76577738a7
Kantoch [Fri, 9 Feb 2018 19:00:10 +0000 (19:00 +0000)]
DO NOT MERGE: Sparse resources fixes am:
ae793c389b -s ours
am:
ce25f76fbb -s ours
Change-Id: I8fa668d939b6c45df4590f823b029536895b4574
Kantoch [Fri, 9 Feb 2018 18:54:40 +0000 (18:54 +0000)]
DO NOT MERGE: Sparse resources fixes
am:
ae793c389b -s ours
Change-Id: Ib87ed9843313e5ba76ccae32161ed5719615fa55
Paavo Pessi [Mon, 22 Jan 2018 12:23:53 +0000 (14:23 +0200)]
Add descriptor set layout lifetime tests
Test was added to verify that implementation honors descriptor set
layout lifetime. Descriptor set layout is destroyed immediately
after creating the pipeline layout, prior to creating the pipeline.
New tests:
dEQP-VK.api.descriptor_set.descriptor_set_layout_lifetime.graphics
dEQP-VK.api.descriptor_set.descriptor_set_layout_lifetime.compute
Components: Vulkan
VK-GL-CTS issue: 856
Change-Id: I978675ddea30e3b312db4bc65da77c26149cc2df
Jari Komppa [Fri, 2 Feb 2018 13:00:22 +0000 (15:00 +0200)]
Added optimization avoiding shader disassembly
This change avoids disassembling shaders if the
--deqp-log-shader-sources=disable option is used.
The default functionality disassembles everything and discards the
result if the shader source logging is disabled. This change skips the
disassembly too.
The performance gain for this is approximately 10% of runtime.
The timings are based on assumption of debug build and having the
shader cache with zero cache misses.
Affects:
none
Components: Framework
Vk-GL-CTS isue: 899
Change-Id: I36eda000a94400d8f50a565c7ef30f4170d51c83
Hernan Liatis [Wed, 24 Jan 2018 01:47:07 +0000 (17:47 -0800)]
Create Compute tests for OpName
These tests makes sure that OpName has no impact on the semantics of
compute kernels.
New tests:
dEQP-VK.spirv_assembly.instruction.compute.opname.entry_main.*
dEQP-VK.spirv_assembly.instruction.compute.opname.entry_rdc.*
Components: Vulkan
VK-GL-CTS issue: 936
VK-GL-CTS public issue: 76
Change-Id: I1ba6b3e47fe838d9ecc721daa78a869bfcf8bc8f
Courtney Goeltzenleuchter [Wed, 7 Feb 2018 23:54:23 +0000 (23:54 +0000)]
Merge "Separate HDR metadata into two tests"
Kantoch [Tue, 19 Jul 2016 11:37:33 +0000 (13:37 +0200)]
DO NOT MERGE: Sparse resources fixes
1. Do not allow to request more queues than available
2. Allow to reuse queue if it supports more than one type of operations
(cherry picked from VK-GL-CTS commit
5f2cf2f7139c2f603a02fe3c284921dd9152dee3)
Bug: b/
72734354
Change-Id: Ia063df59c2220a082ea5a73a0fc968cfb97fda2c
Mark Adams [Wed, 7 Feb 2018 17:36:09 +0000 (17:36 +0000)]
Handle unsupported no_config contexts am:
b60ae978ad
am:
307c716b4e
Change-Id: I2f90861c7c32f75a1611cc13f0e892195daee747
Mark Adams [Wed, 7 Feb 2018 17:31:36 +0000 (17:31 +0000)]
Handle unsupported no_config contexts
am:
b60ae978ad
Change-Id: Idf51783fac99d14fa5ed28dbffd0db36b80e998b
Courtney Goeltzenleuchter [Wed, 7 Feb 2018 00:57:42 +0000 (16:57 -0800)]
Separate HDR metadata into two tests
Separated the HDR metadata test into two tests that
each test a single HDR metadata extension (SMPTE 2086 and CTA 861.3).
Also rename test group to be more specific to tests being added.
Change-Id: Icb98a5f47796c2c21faf0020221e9f5bb2969f9d
Paavo Pessi [Thu, 11 Jan 2018 09:52:45 +0000 (11:52 +0200)]
Test updating set with an empty binding
This test matches VkPositiveLayerTest.EmptyDescriptorUpdateTest and
updates the last descriptor in a set that includes an empty binding.
The test doesn't check any results and it passes as long as no crash
happens.
New tests:
dEQP-VK.binding_model.descriptor_update.empty_descriptor.uniform_buffer
Components: Vulkan
VK-GL-CTS issue: 831
Change-Id: I795e08fa516d42d980d7b3b8baba6a1ba4a29ec7
Paavo Pessi [Mon, 15 Jan 2018 12:40:38 +0000 (14:40 +0200)]
Test discontiguous descriptor sets
Test two descriptor sets bound to discontiguous set numbers with
different descriptor types and shader stages.
New tests:
dEQP-VK.binding_model.shader_access.*.multiple_discontiguous_descriptor_sets.*
Components: Vulkan
VK-GL-CTS issue: 902
Change-Id: I81b94e5a319738411958f93a02a27ba80b5dbbb8
Mika Väinölä [Thu, 18 Jan 2018 13:02:00 +0000 (15:02 +0200)]
Test creating device when out of host memory
This test matches Allocation.CreateInstanceDeviceIntentionalAllocFail
found in loader validation tests which fails on some shipping drivers.
New tests:
dEQP-VK.api.device_init.create_instance_device_intentional_alloc_fail
Components: Vulkan
VK-GL-CTS issue: 823
Change-Id: I8a9d3bce86beac81fc53f2d58a002dee2318610d
Marcin Rogucki [Thu, 25 Jan 2018 16:18:20 +0000 (17:18 +0100)]
UnpackHalf2x16: consistent subnormal 16bit values
Added subnormal values to the test.
3 subnormal special values added:
- mantissa LSB set
- mantissa MSB set
- all mantissa bits set
15 randomly generated subnormal values added.
Test passes iff:
- all normal values are bit by bit matching
and
- all subnormal values are bit by bit matching
or
- all subnormal values are flushed as zeros
Components: Vulkan
VK-GL-CTS issue: 798
Changes:
dEQP-VK.glsl.builtin.function.pack_unpack.unpackhalf2x16_*
Change-Id: Ib211ec70fac00f94d8dd3496c36a34f4ae17a593
Juan A. Suarez Romero [Thu, 25 Jan 2018 12:23:18 +0000 (12:23 +0000)]
Add operation tests for 64-bit integers
This create tests for SPIR-V arithmetic operations that deal with
integer types.
Right now it covers 32-bit and 64-bit signed and unsigned integers, but
it is designed to easily extend to other types, like 16-bit integers.
New tests:
dEQP-VK.spirv_assembly.type.*
VK-GL-CTS issue: 187
Components: Vulkan
Change-Id: Ib84024019d33bac5026cba5e39bc7247c15ba5bb
Chris Forbes [Tue, 6 Feb 2018 01:12:53 +0000 (17:12 -0800)]
Merge remote-tracking branch 'origin/vulkan-cts-1.0.2' into HEAD
Bug: b/
72972234
Change-Id: I537a4f3207c1f88d68a51a23f21a5fed0975b037
Chris Forbes [Mon, 5 Feb 2018 22:52:52 +0000 (14:52 -0800)]
Merge remote-tracking branch 'origin/master' into HEAD
Change-Id: I46bafca084cccb850bdef6657548fc34088ef0af
Chris Forbes [Mon, 5 Feb 2018 21:10:48 +0000 (13:10 -0800)]
Merge remote-tracking branch 'aosp/master' into HEAD
Change-Id: Iecc6b03291d94c485544ba6a270161f69fa267e3
Chris Forbes [Mon, 5 Feb 2018 20:21:03 +0000 (20:21 +0000)]
Merge "Fix issues in the multiview tests"
Sean Callanan [Mon, 29 Jan 2018 19:55:37 +0000 (11:55 -0800)]
Test invalid precision specifier in initializer
The expression
mediump vec2 v = mediump vec2(...);
is not legal GLSL because of the "mediump" in the initializer.
Many implementations accept it, though. This test verifies that they
catch the syntax error correctly.
Bug:
13881278
Test: Ran this on X11 GLX, saw the test case fail as desired
Change-Id: I12e59ce2d813e72f878ae5c8eaf384063b362cdf
Alexander Galazin [Mon, 5 Feb 2018 12:13:47 +0000 (13:13 +0100)]
Merge vk-gl-cts/vulkan-cts-1.0.3 into vk-gl-cts/master
Change-Id: I5197455fbb3dc596087f7e9ca344ef67e011455b
Alexander Galazin [Mon, 5 Feb 2018 12:13:42 +0000 (13:13 +0100)]
Merge vk-gl-cts/vulkan-cts-1.0.2 into vk-gl-cts/vulkan-cts-1.0.3
Change-Id: If7dc3d47395a03af3a34987f08c873c13effd366
Alexander Galazin [Mon, 5 Feb 2018 08:36:46 +0000 (09:36 +0100)]
Fix issues in the multiview tests
* Fix few asserts
* Add forgotten TestLog::EndSection
Components: AOSP
Affects: dEQP-GLES3.functional.mulitiview.*
Change-Id: Ieb7d09beace12ce7f00e6d9cddacea2465101cca
Piotr Byszewski [Fri, 19 Jan 2018 14:15:07 +0000 (15:15 +0100)]
Fix ES2 rbo clear tests for RGB5_A1
color_clear and stencil_clear tests for GL_RGB5_A1 format
should not perform validation for alpha channel as
test thresholds for this channel do not account for dithering.
Components: OpenGL
VK-GL-CTS issue: 348
Affects:
dEQP-GLES2.functional.fbo.render.color_clear.*
dEQP-GLES2.functional.fbo.render.stencil_clear.*
Change-Id: I72fb6ec98dacfe9b2ad8178d0a01f7bb0bee01fc
Daniel Koch [Wed, 31 Jan 2018 22:22:15 +0000 (17:22 -0500)]
Fix use of sparse binding queue on default context
Avoid trying to create two separate queues
when the queueIndex and sparseQueueIndex are the same,
as this violates valid usage which states:
"The queueFamilyIndex member of each element of pQueueCreateInfos
must be unique within pQueueCreateInfos"
Affects:
All vulkan tests
The forerunner change claimed it only affected the following specifically:
dEQP-VK.texture.shadow.*
dEQP-VK.renderpass.sparserendertarget.*
Component: Vuklan
VK-GL-CTS issue: 898, 964
Change-Id: I342e11f9a63eab875336b8078890e677a5d29a2d
Vamsidhar Gaddam [Wed, 31 Jan 2018 10:23:48 +0000 (11:23 +0100)]
Pass in proper pWaitDstStageMask in vktWsiSharedPresentableImageTests
One of the submitInfos for a queue submit had
a semaphore passed in without passing in any stage mask. This is a
spec violation.
Affects:
dEQP-VK.wsi.android.shared_presentable_image.*
Components: Vulkan
VK-GL-CTS issue: 974
Change-Id: Iee2ea838e971a3610f44b8c1f765869a8a21fedb
Panagiotis Apostolou [Wed, 17 Jan 2018 09:01:58 +0000 (10:01 +0100)]
Spec relaxation for YCbCr Implicit chroma reconstruction
Incorporates spec relaxation in the YCbCr tests which describes that when
doing implicit reconstruction with nearest chroma filtering, an
implementation may behave as if both xChromaOffset and yChromaOffset were
VK_CHROMA_LOCATION_MIDPOINT.
Tests affected:
dEQP-VK.ycbcr.conversion.*
Components: Vulkan
VK-GL-CTS issue: 979
Change-Id: I10e83d2f5845c3e74673aa9e42fc4f0950f9e6cd
Alexander Galazin [Thu, 1 Feb 2018 11:48:25 +0000 (12:48 +0100)]
Note that git/python/cmake must be in PATH
Components: Vulkan
VK-GL-CTS issue: 931
Change-Id: I92197ee4e6b03ff3e220936f73ec73e6808be792
Chris Forbes [Sat, 3 Feb 2018 01:14:59 +0000 (17:14 -0800)]
Don't try to pass metadata attribs to eglCreate*Surface
These attribs are only valid for eglSurfaceAttrib / eglQueryAttrib.
Bug: b/
72870242
Change-Id: I6aad5c19e24f775e0b8be6374716cb4ceef01491
Treehugger Robot [Fri, 2 Feb 2018 20:37:08 +0000 (20:37 +0000)]
Merge "Fixes pixelformat utils for RGB10_A2"
Alexander Galazin [Fri, 2 Feb 2018 15:37:21 +0000 (16:37 +0100)]
Merge vk-gl-cts/vulkan-cts-1.0.3 into vk-gl-cts/master
Change-Id: I0ec19f7ba49c9c543cc3a0268779af0768e06d49
Ari Suonpaa [Tue, 26 Sep 2017 07:51:57 +0000 (10:51 +0300)]
Add sparse image support for pipeline multisample tests.
Extended pipeline multisample tests to support source images that
are using sparse binding.
Affects:
dEQP-VK.pipeline.multisample.*
Components: Vulkan
VK-GL-CTS issue: 101
Change-Id: Ice9624de3e88d0a5daf6574f257714fa83db0877
Treehugger Robot [Fri, 2 Feb 2018 04:19:27 +0000 (04:19 +0000)]
Merge "GLES2/3: error on declarations with unspecified precision"
Treehugger Robot [Fri, 2 Feb 2018 01:14:33 +0000 (01:14 +0000)]
Merge "Relax ULP requirement for Log/Log2"
Chris Forbes [Thu, 1 Feb 2018 23:53:06 +0000 (15:53 -0800)]
Merge remote-tracking branch 'aosp/master' into HEAD
Change-Id: Ib3563123d2da1bd9fe6b81a9aa02cfadfad004c5
Juan A. Suarez Romero [Thu, 11 Jan 2018 12:52:42 +0000 (13:52 +0100)]
Return UNSUPPORTED if no compressed textures available
When testing compressed textures, return NOT_SUPPORTED if none of the
compressed textures under testing are available, instead of returning
FAIL.
Components: OpenGL
VK-GL-CTS issue: 943
Affects:
KHR-GLES2.texture_3d.compressed_texture.negative_compressed_*
Change-Id: I77d18fdbb86fef88c680f63f7fe522ed079887aa
Andres Gomez [Thu, 1 Feb 2018 10:03:21 +0000 (12:03 +0200)]
gl_spirv: Correct ARB_gl_spirv check function type
The check for GL_ARB_gl_spirv or GL 4.6 doesn't return anything but
just throws an exception. This corrects the type of the call.
Affects:
KHR-GL45.gl_spirv.*
Components: OpenGL
VK-GL-CTS issue: 981
Change-Id: Ic051d7949672b7a824ebe5c67761be866cd1877f
Piotr Byszewski [Thu, 18 Jan 2018 12:59:57 +0000 (13:59 +0100)]
Fix ES2 cube map GL_LINEAR_* filtering tests
When ES2 context is created on ES3 compatible hardware
seamless sampler should be used during cube map filtering
test result verification.
Components: OpenGL
VK-GL-CTS issue: 284
Affects:
dEQP-GLES2.functional.texture.mipmap.cube.*
Change-Id: I442429ca2e1c368bd2dba0948923034b8a3e54be
Piotr Byszewski [Wed, 17 Jan 2018 14:45:28 +0000 (15:45 +0100)]
Fixed ES2 fbo completeness tests
Twelve ES2 fbo completeness tests failed when they
were executed on hardware that also supports ES3.
This was fixed by detecting this case and verifying
tests results using ES3 rules.
Components: OpenGL
VK-GL-CTS issue: 288
Affects:
dEQP-GLES2.functional.fbo.completeness.renderable.*
Change-Id: I0377021a5a83cfb08c5222895e357182c2cafe67
Kai Ninomiya [Wed, 31 Jan 2018 23:42:43 +0000 (15:42 -0800)]
GLES2/3: error on declarations with unspecified precision
As specified:
The fragment language has no default precision qualifier for floating
point types. Hence for float, floating point vector and matrix
variable declarations, either the declaration must include a precision
qualifier or the default float precision must have been previously
declared.
[GLSL ES 1.00 r17, section 4.5.3 (p36)]
[GLSL ES 3.00 r6, section 4.5.4 (p56)]
Similarly, there is no default precision qualifier for the following
sampler types in either the vertex or fragment language: (...)
[GLSL ES 3.00 r6, section 4.5.4 (p56)]
New tests:
dEQP-GLES2.functional.shaders.declarations.unspecified_precision.*
dEQP-GLES3.functional.shaders.declarations.unspecified_precision.*
Components: AOSP
Change-Id: Ic649556c73c83ba5f31125d25baf4c1c44fcf93f
Jeff Leger [Fri, 12 Jan 2018 20:26:37 +0000 (15:26 -0500)]
Relax ULP requirement for Log/Log2
Relax ULP requirement from 3 to 4 for log/log2. Implementations
that require 4 ULPs will generate a quality warning.
Affects:
dEQP-GLES3.functional.shaders.builtin_functions.precision.log.highp_vertex.*
Components: Framework, AOSP
Khronos OpenGL/API issue: 57
Change-Id: Ibc910322058ee0d4f0a33e21ba256fffe200b3be
Alexander Galazin [Wed, 31 Jan 2018 10:09:13 +0000 (11:09 +0100)]
Merge vk-gl-cts/aosp-deqp-dev into vk-gl-cts/master
Change-Id: I105841cb1422bf42de35208d39195b9e2ddc1469
Paavo Pessi [Fri, 3 Nov 2017 11:49:27 +0000 (13:49 +0200)]
Test texture gather with separate sparse binding queue
Existing texture gather tests assume the binding queue is part of
graphics queue. Tests were modified to support both combined and
separate sparse binding queues.
Affects:
dEQP-VK.glsl.texture_gather.*.sparse_*
Components: Vulkan
VK-GL-CTS issue: 361
Change-Id: I0c2160649e6d0fd3e45401a43ab6b3960312c5e0
Ari Suonpaa [Mon, 16 Oct 2017 06:29:08 +0000 (09:29 +0300)]
Add tests for sparse compressed textures.
Extended compressed texture test to include sparse images used as
textures.
Affects:
dEQP-VK.texture.compressed.*
Components: Vulkan
VK-GL-CTS issue: 101
Change-Id: If317985fa8ff1c9a83040a77209c90172834968d
Piotr Byszewski [Tue, 16 Jan 2018 12:42:07 +0000 (13:42 +0100)]
Re-enable reflect and dot tests
Restored dot and reflect precision tests.
Components: OpenGL
VK-GL-CTS issue: 292
Affects:
dEQP-GLES31.functional.shaders.builtin_functions.precision.reflect.*
dEQP-GLES31.functional.shaders.builtin_functions.precision.dot.*
Change-Id: If3acfadc2e41c8c39e2b5d2f065582bdbfdc2e7a
Courtney Goeltzenleuchter [Sat, 27 Jan 2018 00:43:28 +0000 (16:43 -0800)]
Add new EGL tests for HDR
Add test that will create a surface with smpte & cta metadata,
verify the metadata with eglQuerySurface, modify the metadata
with eglSurfaceAttrib and verify that data via eglQuerySurface.
Unfortunately, not able to verify actual display behavior.
New Test: dEQP-EGL.functional.hdr_color.8888
Components: AOSP
Change-Id: I88bcd99e1c0ab5ab4236ddbcdd7ec3e7d658d710
Courtney Goeltzenleuchter [Sat, 27 Jan 2018 00:39:49 +0000 (16:39 -0800)]
Add support for HDR extension and colorspaces
Components: framework
Change-Id: I66ffcdb458fd6ccad768693d96a73c1d31785138
Kai Ninomiya [Sat, 27 Jan 2018 03:25:26 +0000 (19:25 -0800)]
GLES3: complex exprs, side effects, short circuits
Some of these 8 tests (the 4 that rely on short circuiting) fail on
Nexus6P/7.0 (they were fixed in 8.0 and pass on Pixel2016/8.1):
https://crbug.com/695685
This is a port from WebGL:
https://github.com/KhronosGroup/WebGL/blob/master/sdk/tests/conformance2/glsl3/array-in-complex-expression.html
https://github.com/KhronosGroup/WebGL/blob/master/sdk/tests/conformance2/glsl3/short-circuiting-in-loop-condition.html
All of those tests are ported just for completeness.
New tests:
dEQP-GLES3.functional.shaders.arrays.complex_expression.and_short_circuits_vertex
dEQP-GLES3.functional.shaders.arrays.complex_expression.and_short_circuits_fragment
dEQP-GLES3.functional.shaders.arrays.complex_expression.or_short_circuits_vertex
dEQP-GLES3.functional.shaders.arrays.complex_expression.or_short_circuits_fragment
dEQP-GLES3.functional.shaders.arrays.complex_expression.ternary_only_evaluates_one_operand_vertex
dEQP-GLES3.functional.shaders.arrays.complex_expression.ternary_only_evaluates_one_operand_fragment
dEQP-GLES3.functional.shaders.arrays.complex_expression.sequence_side_effects_affecting_compared_array_content_vertex
dEQP-GLES3.functional.shaders.arrays.complex_expression.sequence_side_effects_affecting_compared_array_content_fragment
dEQP-GLES3.functional.shaders.loops.custom.short_circuit_in_while_vertex
dEQP-GLES3.functional.shaders.loops.custom.short_circuit_in_while_fragment
dEQP-GLES3.functional.shaders.loops.custom.short_circuit_in_for_vertex
dEQP-GLES3.functional.shaders.loops.custom.short_circuit_in_for_fragment
dEQP-GLES3.functional.shaders.loops.custom.short_circuit_in_dowhile_vertex
dEQP-GLES3.functional.shaders.loops.custom.short_circuit_in_dowhile_fragment
dEQP-GLES3.functional.shaders.loops.custom.short_circuit_in_while_sequence_vertex
dEQP-GLES3.functional.shaders.loops.custom.short_circuit_in_while_sequence_fragment
Components: AOSP
Change-Id: Ifa5590f91a8b16e1e2ab5ea8531f0d783c61cda4
Treehugger Robot [Tue, 30 Jan 2018 20:22:15 +0000 (20:22 +0000)]
Merge "Fixes threshold in texture size tests"
Treehugger Robot [Tue, 30 Jan 2018 03:04:39 +0000 (03:04 +0000)]
Merge "GLES3: shaders.indexing.moredynamic"
Chris Forbes [Tue, 30 Jan 2018 01:46:30 +0000 (17:46 -0800)]
Merge remote-tracking branch 'aosp/master' into HEAD
Change-Id: Ifa613aba874e394049ebd9d9bc0c9730505e595a
Kai Ninomiya [Mon, 29 Jan 2018 18:55:22 +0000 (10:55 -0800)]
GLES2: fix typo in es2fShaderStructTests
I happened to notice this typo while editing the file. It could cause
some tests to be skipped when they should not be: it skips a case if
(1) the device has no vertex texture image units and
(2) the case uses any flags at all and
(3) the case is a vertex case.
There could be conformance impact from this change.
Affects:
dEQP-GLES2.functional.shaders.struct.uniform.array_member_dynamic_index_vertex
dEQP-GLES2.functional.shaders.struct.uniform.struct_array_dynamic_index_vertex
dEQP-GLES2.functional.shaders.struct.uniform.nested_struct_array_dynamic_index_vertex
dEQP-GLES2.functional.shaders.struct.uniform.dynamic_loop_struct_array_vertex
dEQP-GLES2.functional.shaders.struct.uniform.dynamic_loop_nested_struct_array_vertex
Components: AOSP
Change-Id: I5110d6c240938bc58664e7dd5ee2b43e26ceea6e
Matthew Netsch [Tue, 19 Dec 2017 20:45:33 +0000 (12:45 -0800)]
Fixes threshold in texture size tests
Threshold was computed directly from backbuffer format.
This makes the threshold too strict if the texture
sampled from has a lower bit-depth than the backbuffer.
This change takes the minimium bit depth between the
backbuffer and texture to compute the threshold.
Components: OpenGL ES
VK-GL-CTS issue: 163
Affects:
dEQP-GLES2.functional.texture.size.*
dEQP-GLES3.functional.texture.size.*
Change-Id: Id491ff4ef1b958459156fe83d9d5099cf81de81e
Kai Ninomiya [Sat, 27 Jan 2018 02:38:36 +0000 (18:38 -0800)]
GLES2: uniform struct{sampler} as a function arg
(Some of?) these tests failed on Pixel2016/7.0 (they were
fixed in 8.0 and pass on Pixel2016/8.1):
https://crbug.com/695680
This is a port from WebGL:
https://github.com/KhronosGroup/WebGL/blob/master/sdk/tests/conformance/glsl/bugs/sampler-struct-function-arg.html
New tests:
dEQP-GLES2.functional.shaders.struct.uniform.sampler_in_function_arg_vertex
dEQP-GLES2.functional.shaders.struct.uniform.sampler_in_function_arg_fragment
dEQP-GLES2.functional.shaders.struct.uniform.sampler_in_array_function_arg_vertex
dEQP-GLES2.functional.shaders.struct.uniform.sampler_in_array_function_arg_fragment
Components: AOSP
Change-Id: Ic71a7969c9bcf481c12b95a99e6379cfec659a8e
Kai Ninomiya [Sat, 27 Jan 2018 21:06:00 +0000 (13:06 -0800)]
GLES3: shaders.indexing.moredynamic
Many of these tests failed on Pixel2016/7.0 and Nexus6P/7.0. (They were
fixed in Pixel2016/8.0.):
https://crbug.com/695689
This is a port from WebGL:
https://github.com/KhronosGroup/WebGL/blob/master/sdk/tests/conformance2/glsl3/vector-dynamic-indexing.html
New tests: dEQP-GLES3.functional.shaders.indexing.moredynamic.*
Components: AOSP
Change-Id: I0b3d738e374c68eebb5cfca1cf9ce5d1c265237b
Kai Ninomiya [Sat, 27 Jan 2018 00:02:25 +0000 (16:02 -0800)]
GLES2: sequence operator (,) eval order and side effects
Ported from:
https://github.com/KhronosGroup/WebGL/blob/master/sdk/tests/conformance/glsl/bugs/sequence-operator-evaluation-order.html
Which was added in:
https://github.com/KhronosGroup/WebGL/pull/981
https://github.com/KhronosGroup/WebGL/pull/1932
These tests were known to fail in on Pixel 2016 / Android 7.1.1.
New tests:
dEQP-GLES2.functional.shaders.operator.sequence.side_effects.affect_ternary
dEQP-GLES2.functional.shaders.operator.sequence.side_effects.affect_and
dEQP-GLES2.functional.shaders.operator.sequence.side_effects.affect_or
Components: AOSP
Change-Id: I7bbcaecc08279df6f30b6855cc99b19b402892c7
Treehugger Robot [Fri, 26 Jan 2018 23:30:16 +0000 (23:30 +0000)]
Merge "Work around VC++ code-gen bug in shader matrix tests." into deqp-dev
Treehugger Robot [Fri, 26 Jan 2018 23:29:52 +0000 (23:29 +0000)]
Merge "Fix missing case for ANGLE_depth_texture" into deqp-dev
Jamie Madill [Fri, 26 Jan 2018 22:57:17 +0000 (17:57 -0500)]
Fix missing case for ANGLE_depth_texture
Happens when ANGLE_depth_texure is exposed and not OES_depth_texture.
Component: AOSP
Affects:
dEQP-GLES2.functional.fbo.completeness.renderable.texture.stencil.depth_stencil_unsigned_int_24_8
dEQP-GLES2.functional.fbo.completeness.renderable.texture.depth.depth_stencil_unsigned_int_24_8
dEQP-GLES3.functional.fbo.completeness.renderable.texture.stencil.depth_stencil_unsigned_int_24_8
dEQP-GLES3.functional.fbo.completeness.renderable.texture.depth.depth_stencil_unsigned_int_24_8
Jamie Madill [Fri, 26 Jan 2018 17:55:20 +0000 (12:55 -0500)]
Work around VC++ code-gen bug in shader matrix tests.
This code-gen bug was found in crbug.com/759402, and was causing test
runtime failures when run. Work around it by disabling optimization
for the specific compiler version.
Components: AOSP
Affects:
dEQP-GLES2.functional.shaders.matrix.*
dEQP-GLES3.functional.shaders.matrix.*
Chris Forbes [Fri, 26 Jan 2018 19:00:13 +0000 (19:00 +0000)]
Merge "GLES3 tests for equality of structs with array members"
Treehugger Robot [Fri, 26 Jan 2018 03:36:57 +0000 (03:36 +0000)]
Merge "Fix using representable texture channel ranges."
Chris Forbes [Fri, 26 Jan 2018 00:47:12 +0000 (16:47 -0800)]
Merge remote-tracking branch 'aosp/master' into deqp-dev
Change-Id: Ie7623be11190d678413fbe64da5d41d98e370b2c
Treehugger Robot [Thu, 25 Jan 2018 23:26:38 +0000 (23:26 +0000)]
Merge "Tolerate flush/finish test assumption failures."
Kai Ninomiya [Thu, 25 Jan 2018 02:41:55 +0000 (18:41 -0800)]
GLES3 tests for equality of structs with array members
The new test case adds regression coverage for the Qualcomm driver
shader compiler crash seen in Android N:
https://crbug.com/695687
It is a port of this WebGL test case:
https://github.com/KhronosGroup/WebGL/blob/master/sdk/tests/conformance2/glsl3/compare-structs-containing-arrays.html
New tests: dEQP-GLES3.functional.shaders.struct.local.array_member_equality_vertex, dEQP-GLES3.functional.shaders.struct.local.array_member_equality_fragment
Components: AOSP
Change-Id: Ie130439f80fddb081fb36db02c24a9e18f5c4bca
Nicolas Capens [Thu, 25 Jan 2018 16:58:26 +0000 (16:58 +0000)]
Tolerate flush/finish test assumption failures.
These tests assume that draw calls are queued up until they're forced
to be executed by a ReadPixels, Flush, or Finish call. The spec does
not require this, and implementations are free to eagerly start
execution of draw calls.
Thus this change reports a CompatibilityWarning if the assumption isn't
met, instead of a Failure.
Google bug: b/
71858368
Components: AOSP
Tests: dEQP-GLES2.functional.flush_finish,
dEQP-GLES3.functional.flush_finish
Treehugger Robot [Wed, 24 Jan 2018 13:29:22 +0000 (13:29 +0000)]
Merge "Test glRenderbufferStorage with unsized depth format"
Slawomir Cygan [Fri, 19 Jan 2018 13:20:47 +0000 (14:20 +0100)]
dEQP-VK.renderpass.sparserendertarget: bind metadata
Memory was not bound to METADATA aspect, when COLOR aspect
mip tail size was 0. METADATA requirements are not related
to mip tails in other aspects - memory must be always and
unconditionally bound to METADATA.
Additionally, change VK_SPARSE_IMAGE_FORMAT_SINGLE_MIPTAIL_BIT
condition to layerNdx == 0, as imageMipTailMemoryBinds vector
is now used both for metadata and other mip tails.
Affects: dEQP-VK.renderpass.sparserendertarget.*
Components: Vulkan
VK-GL-CTS issue: 958
Change-Id: I34661c82e0d9829448ac85546055c72151bfe85e
Kai Ninomiya [Tue, 23 Jan 2018 23:28:27 +0000 (15:28 -0800)]
Test that points are clamped to ALIASED_POINT_SIZE_RANGE
This test verifies that the point size written to gl_PointSize is
clipped to the range ALIASED_POINT_SIZE_RANGE before rasterization.
See also: http://crbug.com/740560
In particular this comment: http://crbug.com/740560#c27
New tests: dEQP-GLES2.functional.rasterization.limits.points
Components: AOSP
Change-Id: I98708ebece4be9c2bce3c7ba3b57454aec657cce
Nicolas Capens [Tue, 23 Jan 2018 23:16:39 +0000 (23:16 +0000)]
Test glRenderbufferStorage with unsized depth format
The OpenGL ES 2.0 specification only supports formats from Table 4.5,
which are all sized internal formats, while ES 3.0 explicitly states
that "<internalformat> must be a sized internal format".
A color format (GL_RGBA) was already being negatively tested for, but
not a depth format.
Mark Adams [Fri, 12 Jan 2018 21:21:00 +0000 (16:21 -0500)]
Handle unsupported no_config contexts
EGL_KHR_no_config_context doesn't mandate support for ES 1.1 or
for GL versions prior to 3.0. Update the test to handle
optional support, and also update the GL version to 3.0 in order
to test a case which is expected to work.
Bug: b/
72041864
Components: AOSP
Test: dEQP-EGL.functional.create_context.no_config
Change-Id: Ibe7856000c67326fe23cd024435290fba536d127
Alexander Galazin [Tue, 23 Jan 2018 08:01:54 +0000 (09:01 +0100)]
Merge vk-gl-cts/aosp-deqp-dev into vk-gl-cts/master
Change-Id: I39c67077ab1c9842ab43708f83335eba0d707890
Nicolas Capens [Mon, 8 Jan 2018 21:49:05 +0000 (16:49 -0500)]
Fix using representable texture channel ranges.
A value of
4294967295 (2^32 - 1) is not exactly representable in
IEEE-754 single-precision floating-point format, and so it gets rounded
to the next representable value, which is
4294967296. However, this
value can't be cast to an unsigned 32-bit integer without overflowing.
GLSL does not define what happens on overflow, and IEEE-754 defines it
as an exception but GLSL forbids exceptions. Hence some implementations
may produce unexpected results. dEQP assumed clamping to the largest
representable integer.
This change fixes that false assumption by reducing the range to values
representable in both float and integer formats.
Note that 32-bit integer formats can still hold values slightly larger
than these ranges. So while previously the floating-point ranges were
too large to represent integer values, they are now too small. This
can't be fixed without separating the integer format tests and only
using integer values to represent their ranges. This doesn't appear
necessary for the time being since the tests that use these floating-
point ranges have large 12/256 tolerances for the output color.
VK-GL-CTS issue: 937
Google bug:
70910885
Component: Framework
Change-Id: Ia2e8a187772bf37ed6ad35b13ed6f5071c46fb43
Adam Czupryna [Fri, 24 Nov 2017 13:59:39 +0000 (14:59 +0100)]
Add SpirvValidationCapabilitiesTest test case implementation
Add CTS_ARB_gl_spirv test implementation that contains:
SpirvValidationCapabilitiesTest
Affects:
KHR-GL45.gl_spirv.spirv_validation_capabilities_test
Components: OpenGL
VK-GL-CTS issue: 554
Change-Id: I0ba93f0ea0dd5c8c9c1348033448fe5ada82b99b
Alexander Galazin [Tue, 23 Jan 2018 01:56:18 +0000 (02:56 +0100)]
Freeze GL 4.6.0 mustpass, add 4.6.1 mustpass
Change-Id: Iae3345620e1dc59437c4b2910fcd08f66cb9eab4
Piotr Byszewski [Wed, 13 Dec 2017 09:48:54 +0000 (10:48 +0100)]
Test packed depth and stencil
GTFTestPackedDepthStencil tests were ported to glcts.
Packed format loop was extracted from tests.
Components: OpenGL
VK-GL-CTS issue: 815
Affects:
KHR-GLES3.packed_depth_stencil.*
KHR-GL*.packed_depth_stencil.*
Change-Id: Ia13d200909cfe097bfbafcb15d5482d30b4d8c5b
Paavo Pessi [Fri, 5 Jan 2018 13:57:10 +0000 (15:57 +0200)]
Test discontiguous descriptor binding values
New tests were added to test discontiguous and arbitrary descriptor
binding values with different descriptor types and shader stages.
New tests:
dEQP-VK.binding_model.shader_access.*.multiple_discontiguous_descriptors.*
dEQP-VK.binding_model.shader_access.*.multiple_arbitrary_descriptors.*
Components: Vulkan
VK-GL-CTS issue: 905
VK-GL-CTS public issue: 67
Change-Id: I00d1ec1e302a03e0f67ba967b6949578ec58c436
Samuel Iglesias Gonsálvez [Fri, 1 Dec 2017 08:36:57 +0000 (09:36 +0100)]
Add more component qualifier array tests for fragment outputs
Tests were added to verify that various combinations of component
qualifiers can be used with different data arrays (float, vec2, vec3)
in fragment shader outputs.
New tests:
dEQP-VK.glsl.440.linkage.varying.component.frag_out.two_vec4.*
Components: Vulkan
VK-GL-CTS issue: 896
Change-Id: If3ed04b88f4a496dcdada9c50a45bbf7580df88e
Alejandro Piñeiro [Tue, 16 Jan 2018 14:43:49 +0000 (15:43 +0100)]
gl_spirv: update ARB_gl_spirv availability check
All tests were checking for the availability of the extension in
order to return NotSupported but one, so it returned Failed.
Additionally we complete the check, so now we are checking for
GL_ARB_gl_spirv or GL 4.6, as other tests already do.
As this is common for all the 8 tests, we add a utility helper at
gl4cts::commonUtils.
Modules: OpenGL
VK-GL-CTS issue: 947
Affects:
KHR-GL45.gl_spirv.spirv_modules_shader_binary_multiple_shader_objects_test
KHR-GL46.gl_spirv.spirv_modules_shader_binary_multiple_shader_objects_test
Change-Id: I8a3be211363d287696ee86f7d6f503794b63710f
Ari Suonpaa [Wed, 13 Dec 2017 09:49:35 +0000 (11:49 +0200)]
Use default device for sparse binding tests.
The default device supports sparse binding when the underlying
physical device has support. There's no need to create a separate
device object. Test context creates a sparse queue whenever
sparse binding is supported by the physical device. This
change also adds support for having a separate sparse binding
queue.
Affects:
dEQP-VK.texture.shadow.*
dEQP-VK.renderpass.sparserendertarget.*
Components: Vulkan
VK-GL-CTS issue: 898
Change-Id: I2ca36d8e641fe3ef5694784e7bb3d8ebe6ed51bf
Treehugger Robot [Sat, 20 Jan 2018 03:01:01 +0000 (03:01 +0000)]
Merge "Port WebGL multisampled renderbuffer resize test"
Chris Forbes [Sat, 20 Jan 2018 00:46:18 +0000 (00:46 +0000)]
Suppress new warnings from gcc-6 am:
9a58d041fe
am:
fd34041ed0
Change-Id: I0b9eca4091dc271141027d68a6a46d58937e56ac
Chris Forbes [Sat, 20 Jan 2018 00:43:46 +0000 (00:43 +0000)]
Suppress new warnings from gcc-6
am:
9a58d041fe
Change-Id: I176dd24b363db336f4f42f8d8c2c5aebb2632dcb
Chris Forbes [Fri, 19 Jan 2018 18:06:53 +0000 (10:06 -0800)]
Suppress new warnings from gcc-6
check_build_sanity currently fails with the system gcc on rodete. Let's
just suppress this warning for now. Intend to back out the suppression
and fix the warnings in next release.
Bug: b/
72219443
Test: build
Change-Id: Ic842a3de3d247d53c49117562b51e8eece3752f9
Alexander Galazin [Fri, 19 Jan 2018 10:12:43 +0000 (11:12 +0100)]
Merge vk-gl-cts/vulkan-cts-1.0.3 into vk-gl-cts/master
Change-Id: I2de5195522bb42f25a2eb7181b366c8284ef7fec
Alexander Galazin [Fri, 19 Jan 2018 08:22:31 +0000 (09:22 +0100)]
Merge vk-gl-cts/vulkan-cts-1.0.2 into vk-gl-cts/vulkan-cts-1.0.3
Change-Id: Ie1eb632fc4601acfc34db84410e04e06cc2da243