Fix signed integer overflow in shader precision tests

32bit integer reference value calculation had possible
signed integer overflows which are undefined. Now the
references are calculated using 64bit integers.

VK-GL-CTS Issue: 3336

Affects:

dEQP-GLES3.functional.shaders.precision.*

Components: OpenGL ES
Change-Id: I682a8d790f1ceaaf5c2caf67a87655835a0aaef8

framework: unify TestPackage behaviour between ES 3.1 and 3.2

Main issue seen here was that ES 3.2 did not reset context state when
initializing TestCaseWrapper, this was causing different behaviour in
following tests:

  KHR-GLES31.core.draw_buffers_indexed.blending
  KHR-GLES32.core.draw_buffers_indexed.blending

Patch also fixes few other differences spotted.

Components: Framework, OpenGL

VK-GL-CTS issue: 3318

Affects:
   glcts
   cts-runner

Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Change-Id: Ie5848e41eba64c3b76b526738e8612c8bebdfbfc

Improve and fix quad randomization in vertex array test

Minimum quad size for fixed point format was larger
than the maximum value range, causing only extremely
large quads to be used. Also the quad randomization
relied on a loop with fixed number of attempts to
produce proper quads. Now the randomization ranges
are selected so that the quads can be accepted without
multiple attempts.

VK-GL-CTS Issue: 3376

Affects:

dEQP-GLES3.functional.vertex_arrays.*

Components: OpenGL ES
Change-Id: I648f2218aa9b661969c7b78c2f81b5e120ae9275

Avoid division by zero in cube map coordinates

Cube map coordinates are calculated from 3-dimensional uv coordinates by
scaling two of the coordinates by the third. In case of using 0,0,0 as
the coordinates, a division by zero occurs. The result is undefined
behavior.

This change detects the division by zero event and returns uv
coordinates as 0,0. Since the behavior is undefined, it doesn't matter
what we return, but after this change the result should at least be
consistent.

Affects:
dEQP-GL*cube*

Components: Framework
VK-GL-CTS issue: 3351

Change-Id: I094a98f790a28bdf7b26ccb170888f2ea9c1e64c

Avoid shifting by more than bit width

Half float mantissa calculation was sometimes using
shift larger than a bit width. This change sets the
mantissa to zero in those cases.

VK-GL-CTS Issue: 3346
VK-GL-CTS Issue: 3349

Affects:

KHR-GLES*.core.internalformat.texture2d.*half_float*
KHR-GLES3.packed_pixels.*16f

Components: OpenGL ES
Change-Id: Ib91082fab530f6ff9e7cd5b6346ad7d433543a47

Pad out glClearBufferuiv value

The test calls glClearBufferuiv with GL_COLOR and a pointer to a
single GLuint when four components are expected.

Affects: KHR-GLES31.core.framebuffer_no_attachments.render

Components: OpenGL

VK-GL-CTS issue: 3295

Change-Id: I0ce2512cc619410c91104b083ab7bb7b9d667c4c

Add missing extension requirement

This change adds the missing #extension line in the shader for the
fragdata.draw_buffers test if extension is supported.

Affects:
dEQP-GLES3.functional.shaders.fragdata.draw_buffers

Components: OpenGL
VK-GL-CTS issue: 3259

Change-Id: Id644e83b938fd3f2e22769faf727043bee88f62d

Fix signed integer overflow in vertex array tests

Random quad calculation in some of the vertex array tests were
calculating ranges for allowed quad sizes using signed integers,
and there was a chance for an overflow, which is undefined behavior.

Now the quad calculation randomizes only coordinates instead of
coordinates and the quad size. This lets us avoid the calculation
of the range.

Some bug fixes to the quad randomization code were also added.

VK-GL-CTS Issue: 3357

Affects:

dEQP-GLES3.functional.vertex_arrays.*

Components: OpenGL ES
Change-Id: I651ac6cef94e15d3fc0bc322ea8d65cbd96a4c2d

Fix signed integer overflow

VK-GL-CTS Issue: 3347

Affects:

KHR-GLES31.core.shader_bitfield_operation.bitfieldExtract.*

Components: OpenGL ES
Change-Id: Ifa35375cc20b54b9e2ff4950d519d0b0dad68812

Fix glcPackedPixelsTests to properly account for skip images

RectangleTest::makeGradient in glcPackedPixelsTests does not
properly account for skip images. This commit prevents it
writing over the end of an array.

Affects:

KHR-GL33.packed_pixels.*

Components: OpenGL ES

VK-GL-CTS Issue: 3275

Change-Id: I9d8a47b59c000512926c3bfc85b6c6dc66eee213

Fix out-of-bounds stack read

glBufferData was given a pointer to a C struct, but
a wrong struct type was used for sizeof.

VK-GL-CTS Issue: 3294

Affects:

KHR-GLES31.core.draw_indirect.negative-noV*arrays

Components: OpenGL
Change-Id: If02916683e0fee46b05dd4774b56f7fe335cc80f

Avoid OOB params access in call log wrapper

Some negative coverage tests call glTexParameter*v and
glGetSamplerParameter*v with GL_TEXTURE_BORDER_COLOR and a pointer to
a single GLint, GLuint or GLfloat. The call log wrapper functions
attempt to log all four parameters expected for the pname and we get
OOB access for params. Avoid this by using arrays of four instead of
single values for params.

Affects:
dEQP-GLES31.functional.debug.negative_coverage.*.texture.texparameter*
dEQP-GLES31.functional.debug.negative_coverage.*.shader.get_sampler_parameter*
dEQP-GLES31.functional.debug.error_filters.case_11
dEQP-GLES31.functional.debug.error_filters.case_18
dEQP-GLES31.functional.debug.error_filters.case_22
dEQP-GLES31.functional.debug.error_groups.case_11

Components: OpenGL

VK-GL-CTS issue: 3273

Change-Id: If244a008c9cfcc6cc443e8d5475f6ce4add1c838

Fix array iteration range

The function getCompressedTexSubImage2DFormat was using function
parameter to loop through an array defined inside the function, while
these two were unrelated. The result may either not loop through the
whole array or go past the end of the array.

Affects:
dEQP-GLES2.functional.negative_api.texture.compressedtexsubimage2d_*

Components: OpenGL
VK-GL-CTS issue: 3272

Change-Id: Ie88d9330e03f56b7afe858bf51e9bf0fa41e4822

Fix source data size in SSBO array length tests

The tests were calling glBufferData with a pointer to a single
deUint32 and a size of 2 * sizeof(deUint32). Change invalidData to
an array of two deUint32s.

Affects: dEQP-GLES31.functional.ssbo.array_length.*

Components: OpenGL

VK-GL-CTS issue: 3274

Change-Id: Id318c6541bbd8b61eec6d2ca460a3404c5c887e8

Remove few EGL robustness tests

Components: EGL

Affects:
dEQP-EGL.functional.robustness.reset_context.shaders.infinite_loop.*

VK-GL-CTS issue: 257

Change-Id: I9edb2e577fb14b98d83bb16dc31188d3a901606d

GCC 11 build fixes for opengl-es-cts

Add missing limits header to misc files to fix compilation.

+ changes to fix following compilation warnings:

"warning: expression does not compute the number of elements in this array;
element type is ‘std::string’ {aka ‘std::__cxx11::basic_string<char>’},
not ‘std::string [15]’ {aka ‘std::__cxx11::basic_string<char> [15]’}
[-Wsizeof-array-div]"

"warning: loop variable ‘ext’ creates a copy from type
‘const std::__cxx11::basic_string<char>’ [-Wrange-loop-construct]"

"warning: comparison of unsigned expression in ‘>= 0’ is always true
[-Wtype-limits]"

Components: OpenGL ES
VK-GL-CTS issue: 3245

Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Change-Id: If62a44d47afccf1118a6fb0986d0c6e715f78c48

x11_egl: only append library path if it's not empty

Pkg-config will strip out any -L if they match the default search
directory, so we ended up with a -L and no arguments to it, that would
consume whatever came next.

Components: EGL

VK-GL-CTS issue: 2726

Change-Id: Iba0f7d8e0a4ab66e5fd369f60e4d04ac5420df65

Use GL_FRAMEBUFFER instead of GL_[DRAW|READ]_FRAMEBUFFER with GLES2

Affects:
KHR-GLES2.texture_3d.copy_sub_image.negative
KHR-GLES2.texture_3d.copy_sub_image.rgba

Components: OpenGL
VK-GL-CTS issue: 3133

Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Change-Id: Ia32e16e6231b53c9f58e99d85e6178dca5b7f012

Fix vector access in EGL partial update tests

The vector of damage regions was being accessed even when the vector
was empty, which is invalid.

This changes the behaviour of the test, which used to invalidate the
whole framebuffer when it found no damage regions, but not invalidates
none of it. This seems like the correct behaviour.

Components: EGL
Affects: dEQP-EGL.functional.partial_update.*

Change-Id: I8c5500a85d000d16b240359de7eadc67144fb7d3

Check for integer texture support in negative teximage tests

Affects:
KHR-GLES2.texture_3d.filtering.combinations.negative

Components: OpenGL
VK-GL-CTS issue: 3134

Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Change-Id: I352d408a486fa12b406fadd6ec6b8d24e1cd211d

Add a glMemoryBarrier between buffer write and read

The tests' behavior is like this:
1. Use transform feedback to write to one buffer
2. Set this buffer to texture, and then to image
3. glDispatchCompute -> use ImageLoad to read from the image
We have to add a glMemoryBarrier between step 1 and 3.

Affects:
KHR-GLES3*.core.texture_buffer.texture_buffer_operations_transform_feedback

Components: OpenGL

VK-GL-CTS issue: 2936

Change-Id: If711469a6cbf4ae1df6851e4006d937760e37ca7

Merge vk-gl-cts/opengl-es-cts-3.2.5 into vk-gl-cts/opengl-es-cts-3.2.6

Change-Id: Ia93ae886128b6368736e4dc433bbc689d80fae23

Remove in-repo verify_submission scripts

These scripts are now replaced by the CTS Tools repo:
https://github.com/KhronosGroup/VK-GL-CTS-Tools

Components: Vulkan, OpenGL

Change-Id: If0f47b4186ed10f9946852e8d37956d7328625a8
(cherry picked from commit bdfb5d67b268ab2dba5178618cc1515f68087be3)

Update Vulkan/GL Readme

Update the Readme files to point to the new
submission verifcation script

Components: Vulkan, OpenGL

Change-Id: I6fb35a383fcc49e543ae67822be22d744c5fcd9c
(cherry picked from commit 412f5a3363eb228ca1a63be337de172a2e622571)
(cherry picked from commit cdf7014b4bcecc17be7abe71ab604cd82f830d19)

Revise the render_list string of angle vulkan waiver

Use 'ANGLE (*Vulkan *' to replace 'ANGLE (Vulkan*'.

Components: OpenGL ES
VK-GL-CTS issue: 2770

Change-Id: I8e5c8a86a81a538b829873a7c133a4c14b181087

Fix precision issue in NearestEdgeTests shaders

We should also use the calculated offset in shader.

Affects:
KHR-GLES32.core.nearest_edge.*

Components: OpenGL ES

VK-GL-CTS issue: 2651

Change-Id: I9427d3a95a98e3be58dd7ea16a2bf96822f33450
(cherry picked from commit b5328021844d8fcca5d0facc327f1e31785ed880)

Update Vulkan/GL Readme

Update the Readme files to point to the new
submission verifcation script

Components: Vulkan, OpenGL

Change-Id: I6fb35a383fcc49e543ae67822be22d744c5fcd9c
(cherry picked from commit 412f5a3363eb228ca1a63be337de172a2e622571)

Allow wildcards in waiver vendor strings

Use wildcard matching instead of exact string comparison for GL
waiver vendor attributes. Use "Google*" prefix for ANGLE XFB
waiver and remove a duplicate ANGLE waiver.

VK-GL-CTS issue: 2770

Components: Framework, OpenGL

Change-Id: I0ffc54af6aaa4c562bf1e8b5077e365da71f9a3e
(cherry picked from commit 074fc201162b347720c4ba3a013101837edf2a8d)

Disable dithering in nearest edge test cases

This commits disables dithering in the nearest edge test cases so the
tests can correctly work with less than 8 bits per pixel.

Affected tests:
KHR-GL*.nearest_edge.*

Components: OpenGL
VK-GL-CTS issue: 2748

Change-Id: I906d7973be1a77e598135b3ed0dde5d671ba0ba7
(cherry picked from commit 671dbeb2ca8d3793f06fb9fe7838e0f02951141c)

Return a valid format when GL_RGB8 is not treated as GL_RGBA8

There was a special case added to NearestEdgeTestCase::toTextureFormat()
in https://gerrit.khronos.org/c/vk-gl-cts/+/6634 to support implementations
that treat GLRGB8 internally as GL_RGBA8888. This code change had a coding
error is that for implementations that don't support GLRGB8 in this way
ended up aborting with a fatal error:
    "Unable to map pixel format to texture format"

This trivial fix corrects the coding bug to simply return the
unadjusted format.

Affects:

KHR-GLES3.core.nearest_edge.offset_*
KHR-GLES31.core.nearest_edge.offset_*

Components: OpenGL

VK-GL-CTS issue: 2758

Change-Id: I639aed8bfed70941688d30edb02a770be4b05b08

Unicode and Python3 fixes for verification scripts

Fix several issues related to running verification scripts under
different versions of Python, making sure the scripts are compatible
with both Python 2 and Python 3 and do not choke on invalid characters,
which are sometimes present in test shaders.

Components: OpenGL, Framework
VK-GL-CTS issue: 2756

Change-Id: Id472b842cb8fd6536e61489f3c01e28c462dbebd

Threshold Consistency for ES2 FBO Tests

ES3 and 3.1 tests both use the same fuzzycompare threshold value
for their FBO tests, whereas ES2 uses a smaller value despite
having the same coarse formats that require the high threshold.

This change adjusts the ES2 FBO Render Test threshold to match
the value set for ES3 and ES3.1

Affects:

dEQP-GLES2.functional.fbo.render.*

Components: GLES2

VK-GL-CTS Issue: 2715

Change-Id: I15203e1ecf45081ec27baee59b780c5bdf5aacaf
(cherry picked from commit 34d1a2e19b9e12b6413fe61d8dc4ee752c6053e4)

Fix ReadPixels format

Pass implementation format to ReadPixels when pixel format is 8,8,8,0
because some implementations treat RGB8 as RGBA8888 which causes these
tests to fail.

Affected tests:
KHR-GLES3.core.nearest_edge.offset_*
KHR-GLES31.core.nearest_edge.offset_*

Components: OpenGL ES
VK-GL-CTS issue: 2712

Change-Id: I6ce92f19d2e329c75ca5fc6800092af623e25829
(cherry picked from commit 780408248fd9e84162f7f6ccf94677952ab0d0da)

Merge vk-gl-cts/opengl-es-cts-3.2.5 into vk-gl-cts/opengl-es-cts-3.2.6

Change-Id: Ib50bbe4e7c82ff147376432d2bf474348e38d952

oglcts-runner: Include cstdlib in glcTestRunnerMain

To pick up the definitions of EXIT_SUCCESS/EXIT_FAILURE.

Fixes these errors building on FreeBSD 10.4 with llvm60:
/root/nvt/tests/buildvulkancts/vk-gl-cts/external/openglcts/modules/runner/glcTestRunnerMain.cpp:115:19: error: use of undeclared identifier 'EXIT_SUCCESS'
        int exitStatus = EXIT_SUCCESS;
                         ^
/root/nvt/tests/buildvulkancts/vk-gl-cts/external/openglcts/modules/runner/glcTestRunnerMain.cpp:136:19: error: use of undeclared identifier 'EXIT_FAILURE'
                                        exitStatus = EXIT_FAILURE;
                                                     ^

Components: OpenGL

VK-GL-CTS issue: 2696

Change-Id: I918f3de12dde8871df8905f5870c3d6c73ae8048

Use shader to calculate sample point

The test's sampling point is taking from interpolated varying, the inaccuracy
of interpolation may cause wrong sampling, use gl_FragCoord in fragment
shader to calcuate sample point directly.

Component: OpenGL

Affected tests: KHR-GLES32.core.nearest_edge.offset_*

VK-GL-CTS issue: 2651

Change-Id: Ib91595ca07a2b146059c51f2d68ebf086c10553c
(cherry picked from commit 3d55f6adbf38215ce3c4e2986a2a2e8a91086f08)

Merge vk-gl-cts/opengl-es-cts-3.2.5 into vk-gl-cts/opengl-es-cts-3.2.6

Change-Id: I3dc4a855b6a1e8e5c9b1794b528fa433c2e7d4ff

Merge vk-gl-cts/opengl-es-cts-3.2.4 into vk-gl-cts/opengl-es-cts-3.2.5

Change-Id: I4d5fd46d7f654ae13dcac771a16aca3683bfd735

Merge vk-gl-cts/opengl-es-cts-3.2.5 into vk-gl-cts/opengl-es-cts-3.2.6

Change-Id: Ie1cf28bfe3f8dcb9d0714472bb60dfc098811716

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
(cherry picked from commit ccb924af7b96ee3771ce25f6d31a0b243bbddee1)

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>

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)

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)

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

Add support for waiver mechanism to cts-runner

Affects: cts-runner

Change-Id: I147cbb49114386b939a6563a1b4b96f620310736
Components: Framework

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

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)

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)

Tweak matrix input values to avoid small determinants

Modify the matrix test input values to better avoid
zero and close to zero determinants, which can produce
invalid inverse matrices.

Affects:

dEQP-GLES3.functional.shaders.matrix.*

Components: OpenGL

VK-GL-CTS issue: 2559

Change-Id: I2015e0f60dbe6271bc540f95dc6631928d38c108

Add formats support in nearest_edge tests

Add RGBA_10_10_10_2 and RGBA_F16 support to pixelFormatMap

Affects:

KHR-GLES32.core.nearest_edge.*

Components: OpenGL

VK-GL-CTS issue: 2547

Change-Id: I6a96f4e5c87967b06bbfed1566df7e784009ad50
(cherry picked from commit dea39590d12bf91ac37e66460a0aa908c5816216)

Offset tris from fragments for triangle tests

At some render resolutions, these tests can rely on the outcome of a
rasterizer tie-break. Since such behavior is left unspecified, we
offset the generated triangles slightly to ensure that they reliably
hit (or miss) the fragment locations we expect them to.

Affects:
KHR-GLES31.core.draw_indirect.basic-mode-drawArrays-triangle_strip_adjacency
KHR-GLES31.core.draw_indirect.basic-mode-drawArrays-triangles_adjacency
KHR-GLES31.core.draw_indirect.basic-mode-drawElements-triangle_strip_adjacency
KHR-GLES31.core.draw_indirect.basic-mode-drawElements-triangles_adjacency

Components: OpenGL ES

VK-GL-CTS issue: 2522

Change-Id: I63ab56dba99b5e6f6369aa4fd8b4538c9d01d874

Account for aliasing exceptions in codegen

Fixes a problem where functions for which aliasing exceptions are in
place (at present this is just glRenderbufferStorageMultisampleEXT)
were not being treated properly in code generation -- their aliased
names were being erroneously used.

Component: Framework

Affects: dEQP-GLES*multisample*

VK-GL-CTS issue: 2291

Change-Id: I5e9caf8828b192d15bb3c91f53a9d09aa9442129

Fix pixel pack/unpack in nearest edge tests

With certain fbconfigs and width/height combos, this would
corrupt memory due to overwriting too many bytes by 1.

Components: OpenGL

VK-GL-CTS Issue: 2496

Affects: KHR-GL46.nearest_edge.offset_*
         KHR-GLES32.core.nearest_edge.offset*

Change-Id: I2b31c8791337555b0bb154cea1f676f0daf43c7e
(cherry picked from commit 9486b751ce73aa581af1c16494506569cf2e8364)

Stop 10bit formats requiring GL_SRGB encoding

The EGL_KHR_gl_colorspace spec does not require formats that
are not SRGB8 to respect the given EGL_GL_COLORSPACE_KHR.

Components: EGL

VK-GL-CTS issue: 2288

Affects:

dEQP-EGL.functional.wide_color.window_1010102_colorspace_srgb
dEQP-EGL.functional.wide_color.pbuffer_1010102_colorspace_srgb

Change-Id: I1d9c502ef2f7068cf13d13564abdfe2d3c75b93b

Check presence of WGL_EXT_colorspace before use

This CL fixes an issue where the WGL_COLORSPACE_EXT_attribute
pixel format query was used without checking the implementation
supports the WGL_EXT_colorspace extension.

Affects:

cts-runner initialization on WGL
glcts.exe initialization on WGL

Components: OpenGL

VK-GL-CTS issue: 2257

Change-Id: Ib4177b851c856570c4b125d537ffa476a02c04f3

Merge vk-gl-cts/opengl-es-cts-3.2.5 into vk-gl-cts/opengl-es-cts-3.2.6

Change-Id: Ib82d397035bcfd1903575e28aefd21addb29d69b

Fix the sorting order in eglChooseConfig test

EGL_COLOR_COMPONENT_TYPE_EXT should be before EGL_COLOR_BUFFER_TYPE
in sorting.

Components: AOSP

VK-GL-CTS issue: 2370

Affects:
dEQP-EGL.functional.choose_config.color_component_type_ext.dont_care

Change-Id: Ic502f17c800dac9f889931eb71971366a5cdb392
(cherry picked from commit 137766c7c9b4f7dac80b90ebf6f9a032188afe0f)

Merge vk-gl-cts/opengl-es-cts-3.2.5 into vk-gl-cts/opengl-es-cts-3.2.6

Change-Id: I73359af82cbc484aceffcf45628b50ee1d84f762

Add texture specification test for GL_OES_texture_half_float.

Google bug: b/74609191
Components: AOSP
Affects: dEQP-GLES2.functional.texture.specification.*

Change-Id: I85aec3469872f2733f528ef7902427e9e87935e7
(cherry picked from commit 19dd72470bdd1a772c08474535e40874624a93b1)

Fix uninitialized index array

The test is doing a glDrawRangeElements using uninitialized data for
the element array. ANGLE validates that the indices fall within the
specified range. Data needs to be initialized to ensure that happens.

The spec says:
It is an error for index values (other than the primitive restart index,
when primitive restart is enabled) to lie outside the range [start, end], but
implementations are not required to check for this. Such indices will cause
implementation-dependent behavior.

Either the test needs to allow for an error or initialize the data so
as to be valid for the call.

Affected Test:
    dEQP-GLES31.functional.debug.negative_coverage.get_error.vertex_array.draw_range_elements

Components: OpenGL ES
VK-GL-CTS issue: 2118

Change-Id: I51f8abb86f088b67990001454bb00d59313200c2
(cherry picked from commit bf7a86af5a959c3f4ad39deabe8b396d8581470a)

Limit dimensionality of arrays-of-arrays in random SSBO tests

One of these tests used a 5 dimensional array-of-arrays, resulting
in thousands of instructions that caused issues on devices with low
ram. This change limits the total dimensionality to prevent this.

Components: AOSP

VK-GL-CTS issue: 2355

Affects:

dEQP-GLES31.functional.ssbo.layout.random.nested_structs_arrays.15
dEQP-GLES31.functional.ssbo.layout.random.nested_structs_arrays_instance_arrays.23
dEQP-GLES31.functional.ssbo.layout.random.all_per_block_buffers.41
dEQP-GLES31.functional.ssbo.layout.random.all_shared_buffer.7
dEQP-GLES31.functional.ssbo.layout.random.all_shared_buffer.21
dEQP-GLES31.functional.ssbo.layout.random.all_shared_buffer.36
dEQP-GLES31.functional.ssbo.layout.random.all_shared_buffer.48
dEQP-GLES31.functional.ssbo.layout.random.all_shared_buffer.49

Change-Id: I9211e376130014e6c1d50f4a2d843033a5219b2d

Merge vk-gl-cts/opengl-es-cts-3.2.6 into vk-gl-cts/master

Change-Id: Ieeee1002be28bdaffed9177386edf13cd860d610

Merge vk-gl-cts/opengl-es-cts-3.2.5 into vk-gl-cts/opengl-es-cts-3.2.6

Change-Id: I9ff5ee4471a46bfc98a96186c79bbaa79aa5d37f

Merge vk-gl-cts/vulkan-cts-1.2.2 into vk-gl-cts/master

Change-Id: Ie87d541785c979466d4a45dbb0c3df8b8a0f6884

Update Vulkan headers

Components: Vulkan

Change-Id: I545a33bdcfbd332d0646e35d1815cc84f2fe645d

Whitelist Vulkan CTS 1.2.2.1

Components: Vulkan

Affects: dEQP-VK.api.driver_properties.properties

Change-Id: I8c1b46d210dec2ad6f85fbb2cd04c9969d350069

Merge vk-gl-cts/vulkan-cts-1.2.1 into vk-gl-cts/vulkan-cts-1.2.2

Change-Id: I8224a74ae64ae8aafa4e0edbaab5aa7c5f2b322d

Merge vk-gl-cts/vulkan-cts-1.2.0 into vk-gl-cts/vulkan-cts-1.2.1

Change-Id: I2594c8998a3e6dcc779d6fd84715ee26dddce726

Add missing barrier to HOST in memory_barrier test

The test was lacking a barrier between device write access and
validation of the data on host.

Components: Vulkan

VK-GL-CTS Issue: 2359

Affects: dEQP-VK.memory_model.message_passing.*

Change-Id: Ia75277aaa2c7e25bff807cd3c6ebe14513403bfc

Fix wide line rasterization tests

Components: Vulkan
VK-GL-CTS Issue: 2347

Affects:
dEQP-VK.rasterization.primitives*line*wide

Change-Id: Ia27d80bfb57b9a09a631580e37d1f1277e99d365

GLES 3.1 negative framebuffer texture2D test fails

Test is using wrong max size when checking cube map textures.

Affected Test:
   dEQP-GLES31.functional.debug.negative_coverage.get_error.buffer.framebuffer_texture2d

Components: OpenGL ES
VK-GL-CTS issue: 2141

Change-Id: I35bcdcfb8181a9eb062da10234f56267c56ac188
(cherry picked from commit 40d1312085353785186aae62ef098ae5c3a2a5c5)

OpenGL ES: Use shader storage alignment

Affected Test:
  KHR-GLES31.core.shader_storage_buffer_object.advanced-switchBuffers-vs

Components: OpenGL ES
VK-GL-CTS issue: 2151

Change-Id: If427cad30ac36cc96a29f936e2ef7a0c4a301667
(cherry picked from commit 5fd95f6e03e1a0ece7052c6d9931cac825909204)

Remove illegal parameter use in fbo.completeness

dEQP.GLES2.functional.fbo.completeness.renderable.texture
was calling glTexImage2D with a internalformat of GL_SRGB8
which is not valid for GLES 2.0 even with GL_EXT_sRGB.

Affects:
dEQP-GLES2.functional.fbo.completeness.renderable.texture.color0.srgb8
dEQP-GLES2.functional.fbo.completeness.renderable.texture.stencil.srgb8
dEQP-GLES2.functional.fbo.completeness.renderable.texture.depth.srgb8

Components: AOSP

VK-GL-CTS issue: 1906

Change-Id: I97003eda10600a0a33e7212b6b5de34c5b2c8848
(cherry picked from commit 1de1dae629667dd79c896906540b1360e78d001a)

Fix use of invalid HALF_FLOAT enum

GLES 2.0 supports GL_HALF_FLOAT_OES via the OES_texture_half_float
extension. When the enum GL_HALF_FLOAT was added in GLES 3.0,
unfortunately it did not use the same enum value as GL_HALF_FLOAT_OES.
Technically, GL_HALF_FLOAT is not supported on a GLES 2.0 implementation
even with OES_texture_half_float. Fix GLES 2.0 tests to always use
GL_HALF_FLOAT_OES to avoid the issue.

Affects:

KHR-GLES2.core.internalformat.texture2d.*_half_float_*
KHR-GLES3*.core.internalformat.texture2d.*_half_float_*

Components: OpenGL

VK-GL-CTS issue: 1923

Change-Id: I9b43091382341798a0d62b9a4ec09153e915dc5c
(cherry picked from commit f7ede20a9805615a3c371b786fa7407708cf612c)

Fix invalid combinations for texture/copy on ES2 driver

OES_packed_depth_stencil describes the valid combinations available.
Texturing is only valid from it when OES_depth_texture is also present.
ES3 added depth texturing, but used different tokens than the initial
extension.

Affects:

KHR-GLES2.core.internalformat.texture2d.
  depth_stencil_unsigned_int_24_8_depth24_stencil8
KHR-GLES*.core.internalformat.texture2d.
  depth_stencil_unsigned_int_24_8_depth_stencil

Components: OpenGL
VK-GL-CTS issue: 1261

Change-Id: I91262c2c7eb71c9d3ff1eb59931227132b1828f9
(cherry picked from commit 33af336e530859e099df76b4c7d4acc5add3c2f7)

Use unsized internal format in KHR-GLES2.texture_3d.[copy_sub_image|framebuffer_texture].rgba8

These tests are using GL_RGBA8 as an internal format for TexImage2D.
Sized internal formats are not allowed for GLES 2.0.

Affects:
KHR-GLES2.texture_3d.copy_sub_image.rgba
KHR-GLES2.texture_3d.framebuffer_texture.rgba8

VK-GL-CTS issue: 1960

Change-Id: I8272e4ca0e2a9eda1161a5135e11f67b4ea5af6d
(cherry picked from commit eec82ceff77877bdad7fc94f8ae793db09ffff46)

Fix invalid parameter for GLES 2.0

Test was creating a texture with format of GL_DEPTH_STENCIL and
internalformat of GL_DEPTH24_STENCIL8_OES. In ES 2.0 the format and
internalformat must match.
Added an extra case when running 3.0 or better that does allow different
values for format and internalformat.

Affects:

KHR-GLES2.core.internalformat.texture2d.depth_stencil_unsigned_int_24_8_depth24_stencil

Components: OpenGL

VK-GL-CTS issue: 1903

Change-Id: I710975a0ea7c1d551f412590f887d9412be6f897
(cherry picked from commit 0c9f0c072d530bf378bc50265284fbc95c450e69)

Fix invalid combinations for texture/copy on ES2 driver

The OES_rgb8_rgba8 extension specification only specifies
RenderbufferStorageOES for additional GL_RGB8/GL_RGBA8 tokens.  ES3 spec
made these mandatory formats for texturing/copying.

Affects:

KHR-GLES2.core.internalformat.texture2d.rgb_unsigned_byte_rgb8
KHR-GLES2.core.internalformat.texture2d.rgba_unsigned_byte_rgb8
KHR-GLES2.core.internalformat.copy_tex_image.rgba8
KHR-GLES2.core.internalformat.copy_tex_image.rgb8

Components: OpenGL
VK-GL-CTS issue: 1260

Change-Id: Icbbd8882c5e623f5ac4725eebf977f535cfbfa31
(cherry picked from commit 5f7e621d6065f843651f7ca8f98989d05ca0bd65)

Merge vk-gl-cts/opengl-es-cts-3.2.5 into vk-gl-cts/opengl-es-cts-3.2.6

Change-Id: Ic55e8e0a9f73adb4d1c3c9a9bbd4f15932511027

Migrate deRoundUp32 to deInt32.h

deRoundUp32 is used in a number of places, promote to utility function.
back ported https://gerrit.khronos.org/#/c/5197 work at 3.2.5.1 tag.

Components: OpenGL ES
VK-GL-CTS issue: 2151

Change-Id: I3aaec51e0eec0e8e31e2082443d0d332bb8c3a56

Merge vk-gl-cts/vulkan-cts-1.2.2 into vk-gl-cts/master

Change-Id: Ib55bcc487e332625eccada5ce31be1b56c73395e

Merge vk-gl-cts/vulkan-cts-1.2.1 into vk-gl-cts/vulkan-cts-1.2.2

Change-Id: I14bd589f1bd3adfabed0c7247fbef7deebe74c96

Non-strict line interpolation and rasterization

Firstly non-strict line rasterization (parallelogram decomposed to
triangles) is changed to use the TriangleInterpolator do determine
interpolated colours. Line endpoint attributes are duplicated to the
end vertices of parallelogram lines, which means that interpolation is
not along the normal of the line, but would be parallel to the major
axis.

For the interpolation to be calculated correctly by the
TriangleInterpolator, the triangle geometry needed to be produced with
the correct w values from the line endpoints, previously the triangle
was generated pre-projected with 1.0 for the W-coordinates. The
perspective interpolation previously was calculated separately using
the original line parameters passed to MultisampleLineInterpolator.

The second part of the change is to allow for a non-strict line 'GLES'
centric relaxation - GLES specified 1px non-antialiased lines would use
the 'diamond-exit' rule to produce bresenham lines. Vulkan only
specified parallelograms, but not all GLES implementations are able
to disable the 1px bresenham behaviour.

All changes are in verifyMultisameplLineGroupInterpolation and
verifyMultisameplLineGroupInterpolationInternal. Strict lines are not
affected.

This fixes Imaginations 1.2.2.x failures in:

dEQP-VK.rasterization.flatshading*line*
dEQP-VK.rasterization.interpolation*line*

Affects:

dEQP-VK.rasterization*line*
dEQP-GLES2.functional.rasterization*line*
dEQP-GLES3.functional.rasterization*line*

Components: Framework

VK-GL-CTS Issue: 2274

Change-Id: I0cfd496969512bdbf9d0156ea8f43e737c44512f

Suppress -Werror=array-bounds for latest GCC

VK-GL-CTS Issue: 2345

Change-Id: Id2a04a13dd3a13cc159919778315218e2f136fca

Extend Android Hardware Buffer external memory format tests

Add testing of depth and stencil Android Hardware Buffer formats.
These are added in Android API version 28 and present in the Format
Equivalence Table for VK_ANDROID_external_memory_android_hardware_buffer

Added tests:
dEQP-VK.api.external.memory.android_hardware_buffer.image_formats.d16_unorm
dEQP-VK.api.external.memory.android_hardware_buffer.image_formats.x8_d24_unorm_pack32
dEQP-VK.api.external.memory.android_hardware_buffer.image_formats.d24_unorm_s8_uint
dEQP-VK.api.external.memory.android_hardware_buffer.image_formats.d32_sfloat
dEQP-VK.api.external.memory.android_hardware_buffer.image_formats.d32_sfloat_s8_uint
dEQP-VK.api.external.memory.android_hardware_buffer.image_formats.s8_uint

Components: Vulkan

VK-GL-CTS issue: 2280

Change-Id: Ia90efc34f600eea6a1510276f3edab668c253459

Ensure alphabetical order of xml attributes

Python 3.8 removed artificial alphabetical ordering
of attributes in ElementTree. The recommended solution
to avoid changes in generated xml is to add attributes
in the alphabetical order directly in the source code.

See https://github.com/python/cpython/commit/63673916464bace8e2147357395fdf3497967ecb

Components: Framework

VK-GL-CTS issue: 2358

Change-Id: Ife4b93a01d76f1f99fcdf2c1d13291ae5c706d85
(cherry picked from commit b2d84418efc74c558be55c2934e14c2a93edefed)

Ensure alphabetical order of xml attributes

Python 3.8 removed artificial alphabetical ordering
of attributes in ElementTree. The recommended solution
to avoid changes in generated xml is to add attributes
in the alphabetical order directly in the source code.

See https://github.com/python/cpython/commit/63673916464bace8e2147357395fdf3497967ecb

Components: Framework

VK-GL-CTS issue: 2358

Change-Id: Ife4b93a01d76f1f99fcdf2c1d13291ae5c706d85

Merge vk-gl-cts/vulkan-cts-1.2.2 into vk-gl-cts/master

Change-Id: I4769eb2ad81f29030c947fb1507fd8d3b64d385b

Merge vk-gl-cts/vulkan-cts-1.2.1 into vk-gl-cts/vulkan-cts-1.2.2

Change-Id: I6f501d0a79d072f3c3178ceda69dff5dfa9f98a0

Merge vk-gl-cts/vulkan-cts-1.2.0 into vk-gl-cts/vulkan-cts-1.2.1

Change-Id: Ibebde574b312ceb0f9077bcbe13ef5594387b2eb

fix input dependency for renderpass depth/stencil format tests

This appears to be a copy/paste mistake. The exact stage/access bits was
taken from other dependencies with an input attachment which is a
depth/stencil attachment.

Affected tests:
dEQP-VK.renderpass.*.formats.d16_unorm.input.*
dEQP-VK.renderpass.*.formats.x8_d24_unorm_pack32.input.*
dEQP-VK.renderpass.*.formats.d32_sfloat.input.*
dEQP-VK.renderpass.*.formats.d24_unorm_s8_uint.input.*
dEQP-VK.renderpass.*.formats.d32_sfloat_s8_uint.input.*

Change-Id: Ib1cde2371781cc8337905af04e0b7732b74333af
Components: Vulkan
VK-GL-CTS issue: 2351

Avoid exceeding maxPerStageResources in descriptor_indexing tests

The 'calculateInLoop; (*_in_loop) variant of the test is using additional
uniform texel buffer descriptor binding in pipeline layout.

This binding was counted only against maxPerStageDescriptorSampledImages and
maxDescriptorSetSampledImages limits, but not maxPerStageResources.

The fix is to count it against all three limits - but do this only when needed (in
*_in_loop) tests.

Component: Vulkan

VK-GL-CTS Issue: 2333

Affects: dEQP-VK.descriptor_indexing.*

Change-Id: I1c3476cb8c50c3ed61e73ecd4c106e6e9b613022

Split subgroupBroadcast tests

Subgroup broadcast tests with required subgroup size were
iterating the range of supported sizes in a single test.
These are now split into own test cases to prevent long
execution times of these tests and to allow the test
group to run faster when running tests in parallel.

Affects:

dEQP-VK.subgroups.ballot_broadcast.compute.subgroupbroadcast*

Components: Vulkan

VK-GL-CTS issue: 2327

Change-Id: I5af34cebee0fd58e5e065dbdef882badcf04a90b

Test variableMultisampleRate with non-empty framebuffer

This commits adds new test variants to the existing variable rate
multisample tests to check a non-empty framebuffer does not affect the
results, both when the subpass is still completely empty and when the
subpass contains an VK_ATTACHMENT_UNUSED attachment reference.

New tests:
dEQP-VK.pipeline.multisample.variable_rate.*_fb_*

Components: Vulkan
VK-GL-CTS issue: 2152

Change-Id: Ic8aabad6a00ca29c4eb624222548c6c607f6dec6

Added a fourth batch of coverage tests found by GraphicsFuzz

New tests:

dEQP-VK.graphicsfuzz.cov-const-folding-ceil-vec4
dEQP-VK.graphicsfuzz.cov-const-folding-det-identity
dEQP-VK.graphicsfuzz.cov-const-folding-same-condition
dEQP-VK.graphicsfuzz.cov-const-folding-sinh-inf
dEQP-VK.graphicsfuzz.cov-dag-combiner-neg-div-pow2
dEQP-VK.graphicsfuzz.cov-dag-combiner-same-cond-nested
dEQP-VK.graphicsfuzz.cov-increment-inside-clamp
dEQP-VK.graphicsfuzz.cov-loop-findmsb-findlsb
dEQP-VK.graphicsfuzz.cov-max-clamp-same-minval
dEQP-VK.graphicsfuzz.cov-selection-dag-same-cond-twice
dEQP-VK.graphicsfuzz.cov-simplify-clamp-max-itself
dEQP-VK.graphicsfuzz.cov-simplify-component-uniform-idx
dEQP-VK.graphicsfuzz.cov-simplify-div-by-uint-one
dEQP-VK.graphicsfuzz.cov-simplify-ldexp-exponent-zero
dEQP-VK.graphicsfuzz.cov-simplify-mul-identity
dEQP-VK.graphicsfuzz.cov-simplify-not-less-than-neg
dEQP-VK.graphicsfuzz.cov-simplify-select-fragcoord
dEQP-VK.graphicsfuzz.cov-simplify-sign-cosh
dEQP-VK.graphicsfuzz.cov-simplify-smoothstep-undef
dEQP-VK.graphicsfuzz.cov-value-tracking-const-dfdy
dEQP-VK.graphicsfuzz.cov-value-tracking-inclusive-or
dEQP-VK.graphicsfuzz.cov-value-tracking-uniform-incident

Components: Vulkan
Change-Id: If6d0803f73424c2e1fcc922e738803fac8b7cf60

Merge vk-gl-cts/opengl-cts-4.6.0 into vk-gl-cts/master

Change-Id: I98e6a4466dad826eb55fb9cf99d81d217f610123

Merge vk-gl-cts/opengl-es-cts-3.2.6 into vk-gl-cts/master

Change-Id: I209deda6936efa25b41d98bbefcca889fa9f1906

Make ImageSamplingInstance loop over pNexts

We need this to potentially support multiple pNexts.

Affected tests:
dEQP-VK.pipeline.sampler.*

Components: Vulkan
VK-GL-CTS issue: 2331

Change-Id: Ia38f9a6f2a2ffafe1d71c8169bb8f20f8cde1da5

Merge vk-gl-cts/vulkan-cts-1.2.2 into vk-gl-cts/master

Change-Id: I0e426c55c733847803339048389c18163350ef4f

Merge vk-gl-cts/vulkan-cts-1.2.1 into vk-gl-cts/vulkan-cts-1.2.2

Change-Id: I37ceae996c069a640b7c9c67177bc6b0fe6ee5ee