stride3d
/
xkslang
зеркало из https://github.com/stride3d/xkslang.git
1
0
Форкнуть 0
Код Задачи Пакеты Проекты Релизы Вики Активность
xkslang/gtests/Spv.FromFile.cpp

441 строка
15 KiB
C++
Исходник Обычный вид История

GLSL: Add binding map tests This PR adds a GLSL equivalent to the HLSL binding mapping tests for offsets and auto-numbering. The shaders are as equivalent as possible. The bindings of the base results match exactly between the two.
2016-09-27 02:02:44 +03:00
//
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
// Copyright (C) 2016 Google, Inc.
//
// All rights reserved.
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions
// are met:
//
// Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
//
// Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following
// disclaimer in the documentation and/or other materials provided
// with the distribution.
//
// Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived
// from this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
// FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
// COPYRIGHT HOLDERS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
// INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
// BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
// LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
// CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
// LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
// ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
// POSSIBILITY OF SUCH DAMAGE.
#include <algorithm>
#include <gtest/gtest.h>
#include "TestFixture.h"
namespace glslangtest {
namespace {
HLSL: Add location offsets per resource type This PR adds the ability to offset sampler, texture, and UBO bindings from provided base bindings, and to auto-number bindings that are not provided with explicit register numbers. The mechanism works as follows: - Offsets may be given on the command line for all stages, or individually for one or more single stages, in which case the offset will be auto-selected according to the stage being compiled. There is also an API to set them. The new command line options are --shift-sampler-binding, --shift-texture-binding, and --shift-UBO-binding. - Uniforms which are not given explicit bindings in the source code are auto-numbered if and only if they are in live code as determined by the algorithm used to build the reflection database, and the --auto-map-bindings option is given. This auto-numbering avoids using any binding slots which were explicitly provided in the code, whether or not that explicit use was live. E.g, "uniform Texture1D foo : register(t3);" with --shift-texture-binding 10 will reserve binding 13, whether or not foo is used in live code. - Shorter synonyms for the command line options are available. See the --help output. The testing infrastructure is slightly extended to allow use of the binding offset API, and two new tests spv.register.(no)autoassign.frag are added for comparing the resulting SPIR-V.
2016-09-08 00:20:19 +03:00
struct IoMapData {
const char* fileName;
const char* entryPoint;
int baseSamplerBinding;
int baseTextureBinding;
Add UAV (image) binding offset and HLSL register support This PR adds: 1. The "u" register class for RW* objects. 2. --shift-image-bindings (== --sib), analogous to --shift-texture-bindings etc. 3. Case insensitive reg classes. 4. Tests for above.
2016-11-01 19:31:42 +03:00
int baseImageBinding;
HLSL: Add location offsets per resource type This PR adds the ability to offset sampler, texture, and UBO bindings from provided base bindings, and to auto-number bindings that are not provided with explicit register numbers. The mechanism works as follows: - Offsets may be given on the command line for all stages, or individually for one or more single stages, in which case the offset will be auto-selected according to the stage being compiled. There is also an API to set them. The new command line options are --shift-sampler-binding, --shift-texture-binding, and --shift-UBO-binding. - Uniforms which are not given explicit bindings in the source code are auto-numbered if and only if they are in live code as determined by the algorithm used to build the reflection database, and the --auto-map-bindings option is given. This auto-numbering avoids using any binding slots which were explicitly provided in the code, whether or not that explicit use was live. E.g, "uniform Texture1D foo : register(t3);" with --shift-texture-binding 10 will reserve binding 13, whether or not foo is used in live code. - Shorter synonyms for the command line options are available. See the --help output. The testing infrastructure is slightly extended to allow use of the binding offset API, and two new tests spv.register.(no)autoassign.frag are added for comparing the resulting SPIR-V.
2016-09-08 00:20:19 +03:00
int baseUboBinding;
Add reflection queries for thread local size and ssbo auto-binding New command line option --shift-ssbo-binding mirrors --shift-ubo-binding, etc. New reflection query getLocalSize(int dim) queries local size, e.g, CS threads.
2017-02-22 03:19:08 +03:00
int baseSsboBinding;
HLSL: Add location offsets per resource type This PR adds the ability to offset sampler, texture, and UBO bindings from provided base bindings, and to auto-number bindings that are not provided with explicit register numbers. The mechanism works as follows: - Offsets may be given on the command line for all stages, or individually for one or more single stages, in which case the offset will be auto-selected according to the stage being compiled. There is also an API to set them. The new command line options are --shift-sampler-binding, --shift-texture-binding, and --shift-UBO-binding. - Uniforms which are not given explicit bindings in the source code are auto-numbered if and only if they are in live code as determined by the algorithm used to build the reflection database, and the --auto-map-bindings option is given. This auto-numbering avoids using any binding slots which were explicitly provided in the code, whether or not that explicit use was live. E.g, "uniform Texture1D foo : register(t3);" with --shift-texture-binding 10 will reserve binding 13, whether or not foo is used in live code. - Shorter synonyms for the command line options are available. See the --help output. The testing infrastructure is slightly extended to allow use of the binding offset API, and two new tests spv.register.(no)autoassign.frag are added for comparing the resulting SPIR-V.
2016-09-08 00:20:19 +03:00
bool autoMapBindings;
Fix defects in uniform array flattening Fix for two defects as follows: - The IO mapping traverser was not setting inVisit, and would skip some AST nodes. Depending on the order of nodes, this could have prevented the binding from showing up in the generated SPIR-V. - If a uniform array was flattened, each of the flattened scalars from the array is still a (now-scalar) uniform. It was being converted to a temporary.
2016-09-22 23:35:23 +03:00
bool flattenUniforms;
HLSL: Add location offsets per resource type This PR adds the ability to offset sampler, texture, and UBO bindings from provided base bindings, and to auto-number bindings that are not provided with explicit register numbers. The mechanism works as follows: - Offsets may be given on the command line for all stages, or individually for one or more single stages, in which case the offset will be auto-selected according to the stage being compiled. There is also an API to set them. The new command line options are --shift-sampler-binding, --shift-texture-binding, and --shift-UBO-binding. - Uniforms which are not given explicit bindings in the source code are auto-numbered if and only if they are in live code as determined by the algorithm used to build the reflection database, and the --auto-map-bindings option is given. This auto-numbering avoids using any binding slots which were explicitly provided in the code, whether or not that explicit use was live. E.g, "uniform Texture1D foo : register(t3);" with --shift-texture-binding 10 will reserve binding 13, whether or not foo is used in live code. - Shorter synonyms for the command line options are available. See the --help output. The testing infrastructure is slightly extended to allow use of the binding offset API, and two new tests spv.register.(no)autoassign.frag are added for comparing the resulting SPIR-V.
2016-09-08 00:20:19 +03:00
};
std::string FileNameAsCustomTestSuffixIoMap(
const ::testing::TestParamInfo<IoMapData>& info) {
std::string name = info.param.fileName;
// A valid test case suffix cannot have '.' and '-' inside.
std::replace(name.begin(), name.end(), '.', '_');
std::replace(name.begin(), name.end(), '-', '_');
return name;
}
SPV: Turn on atomic-storage functionality for SPIR-V. This is used by OpenGL, but not Vulkan. Includes: - atomicCounter, atomicIncrement, atomicCounterDecrement - atomic_uint layout-offset checking - AtomicStorage capability
2016-07-07 22:20:00 +03:00
using CompileVulkanToSpirvTest = GlslangTest<::testing::TestWithParam<std::string>>;
using CompileOpenGLToSpirvTest = GlslangTest<::testing::TestWithParam<std::string>>;
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
using VulkanSemantics = GlslangTest<::testing::TestWithParam<std::string>>;
SPV: Further refine OpenGL vs. Vulkan differences for SPIR-V. Includes adding test cases to verify the differences.
2016-07-08 02:46:42 +03:00
using OpenGLSemantics = GlslangTest<::testing::TestWithParam<std::string>>;
Vulkan: Finish semantics for what creates spec-const-semantics. Note: This required adding a new test mode to see the AST for vulkan tests. This also required reworking some deeper parts of type creation, regarding when storage qualification and constness is deduced bottom-up or dictated top-down.
2016-05-24 01:07:07 +03:00
using VulkanAstSemantics = GlslangTest<::testing::TestWithParam<std::string>>;
GLSL: Add binding map tests This PR adds a GLSL equivalent to the HLSL binding mapping tests for offsets and auto-numbering. The shaders are as equivalent as possible. The bindings of the base results match exactly between the two.
2016-09-27 02:02:44 +03:00
using HlslIoMap = GlslangTest<::testing::TestWithParam<IoMapData>>;
using GlslIoMap = GlslangTest<::testing::TestWithParam<IoMapData>>;
Parser: Implement extension GL_AMD_gpu_shader_half_float. - Add built-in types: float16_t, f16vec, f16mat. - Add support of half float constant: hf, HF. - Extend built-in floating-point operators: +, -, *, /, ++, --, +=, -=, *=, /=, ==, !=, >=, <=, >, <. - Add support of type conversions: float16_t -> XXX, XXX -> float16_t. - Add new built-in functions.
2016-07-29 11:00:05 +03:00
#ifdef AMD_EXTENSIONS
using CompileVulkanToSpirvTestAMD = GlslangTest<::testing::TestWithParam<std::string>>;
#endif
support SPV_NV_viewport_array2 and SPV_NV_stereo_view_rendering
2017-01-13 12:10:53 +03:00
#ifdef NV_EXTENSIONS
using CompileVulkanToSpirvTestNV = GlslangTest<::testing::TestWithParam<std::string>>;
#endif
Pure Texture to Sampled Texture Transform Adds a transformation step to the post processing step. Two modes are available: 1) keep - Keeps samplers, textures and sampled textures as is 2) transform pure texture into sampled texture and remove pure samplers - removes all pure samplers - transforms all pure textures into its sampled counter part Change-Id: If54972e8052961db66c23f4b7e719d363cf6edbd
2017-04-26 00:31:03 +03:00
using CompileUpgradeTextureToSampledTextureAndDropSamplersTest = GlslangTest<::testing::TestWithParam<std::string>>;
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
// Compiling GLSL to SPIR-V under Vulkan semantics. Expected to successfully
// generate SPIR-V.
SPV: Turn on atomic-storage functionality for SPIR-V. This is used by OpenGL, but not Vulkan. Includes: - atomicCounter, atomicIncrement, atomicCounterDecrement - atomic_uint layout-offset checking - AtomicStorage capability
2016-07-07 22:20:00 +03:00
TEST_P(CompileVulkanToSpirvTest, FromFile)
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
{
Gtests can be run on another source tree The gtest executable accepts a --test-root option to specify a root directory for test files. It defaults to the Test directory in the source tree from which the executable is built. For example, this lets us run test exectuables built with MinGW on Linux on a Windows machine with its own copy of the source tree.
2016-10-05 17:25:09 +03:00
loadFileCompileAndCheck(GlobalTestSettings.testRoot, GetParam(),
Add support for testing file-based HLSL source code in GTest.
2016-05-16 19:50:30 +03:00
Source::GLSL, Semantics::Vulkan,
Target::Spv);
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
}
SPV: Turn on atomic-storage functionality for SPIR-V. This is used by OpenGL, but not Vulkan. Includes: - atomicCounter, atomicIncrement, atomicCounterDecrement - atomic_uint layout-offset checking - AtomicStorage capability
2016-07-07 22:20:00 +03:00
// Compiling GLSL to SPIR-V under OpenGL semantics. Expected to successfully
// generate SPIR-V.
TEST_P(CompileOpenGLToSpirvTest, FromFile)
{
Gtests can be run on another source tree The gtest executable accepts a --test-root option to specify a root directory for test files. It defaults to the Test directory in the source tree from which the executable is built. For example, this lets us run test exectuables built with MinGW on Linux on a Windows machine with its own copy of the source tree.
2016-10-05 17:25:09 +03:00
loadFileCompileAndCheck(GlobalTestSettings.testRoot, GetParam(),
SPV: Turn on atomic-storage functionality for SPIR-V. This is used by OpenGL, but not Vulkan. Includes: - atomicCounter, atomicIncrement, atomicCounterDecrement - atomic_uint layout-offset checking - AtomicStorage capability
2016-07-07 22:20:00 +03:00
Source::GLSL, Semantics::OpenGL,
Target::Spv);
}
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
// GLSL-level Vulkan semantics test. Expected to error out before generating
// SPIR-V.
TEST_P(VulkanSemantics, FromFile)
{
Gtests can be run on another source tree The gtest executable accepts a --test-root option to specify a root directory for test files. It defaults to the Test directory in the source tree from which the executable is built. For example, this lets us run test exectuables built with MinGW on Linux on a Windows machine with its own copy of the source tree.
2016-10-05 17:25:09 +03:00
loadFileCompileAndCheck(GlobalTestSettings.testRoot, GetParam(),
Add support for testing file-based HLSL source code in GTest.
2016-05-16 19:50:30 +03:00
Source::GLSL, Semantics::Vulkan,
Target::Spv);
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
}
SPV: Further refine OpenGL vs. Vulkan differences for SPIR-V. Includes adding test cases to verify the differences.
2016-07-08 02:46:42 +03:00
// GLSL-level Vulkan semantics test. Expected to error out before generating
// SPIR-V.
TEST_P(OpenGLSemantics, FromFile)
{
Gtests can be run on another source tree The gtest executable accepts a --test-root option to specify a root directory for test files. It defaults to the Test directory in the source tree from which the executable is built. For example, this lets us run test exectuables built with MinGW on Linux on a Windows machine with its own copy of the source tree.
2016-10-05 17:25:09 +03:00
loadFileCompileAndCheck(GlobalTestSettings.testRoot, GetParam(),
SPV: Further refine OpenGL vs. Vulkan differences for SPIR-V. Includes adding test cases to verify the differences.
2016-07-08 02:46:42 +03:00
Source::GLSL, Semantics::OpenGL,
Target::Spv);
}
Vulkan: Finish semantics for what creates spec-const-semantics. Note: This required adding a new test mode to see the AST for vulkan tests. This also required reworking some deeper parts of type creation, regarding when storage qualification and constness is deduced bottom-up or dictated top-down.
2016-05-24 01:07:07 +03:00
// GLSL-level Vulkan semantics test that need to see the AST for validation.
TEST_P(VulkanAstSemantics, FromFile)
{
Gtests can be run on another source tree The gtest executable accepts a --test-root option to specify a root directory for test files. It defaults to the Test directory in the source tree from which the executable is built. For example, this lets us run test exectuables built with MinGW on Linux on a Windows machine with its own copy of the source tree.
2016-10-05 17:25:09 +03:00
loadFileCompileAndCheck(GlobalTestSettings.testRoot, GetParam(),
Vulkan: Finish semantics for what creates spec-const-semantics. Note: This required adding a new test mode to see the AST for vulkan tests. This also required reworking some deeper parts of type creation, regarding when storage qualification and constness is deduced bottom-up or dictated top-down.
2016-05-24 01:07:07 +03:00
Source::GLSL, Semantics::Vulkan,
Target::AST);
}
HLSL: Add location offsets per resource type This PR adds the ability to offset sampler, texture, and UBO bindings from provided base bindings, and to auto-number bindings that are not provided with explicit register numbers. The mechanism works as follows: - Offsets may be given on the command line for all stages, or individually for one or more single stages, in which case the offset will be auto-selected according to the stage being compiled. There is also an API to set them. The new command line options are --shift-sampler-binding, --shift-texture-binding, and --shift-UBO-binding. - Uniforms which are not given explicit bindings in the source code are auto-numbered if and only if they are in live code as determined by the algorithm used to build the reflection database, and the --auto-map-bindings option is given. This auto-numbering avoids using any binding slots which were explicitly provided in the code, whether or not that explicit use was live. E.g, "uniform Texture1D foo : register(t3);" with --shift-texture-binding 10 will reserve binding 13, whether or not foo is used in live code. - Shorter synonyms for the command line options are available. See the --help output. The testing infrastructure is slightly extended to allow use of the binding offset API, and two new tests spv.register.(no)autoassign.frag are added for comparing the resulting SPIR-V.
2016-09-08 00:20:19 +03:00
// HLSL-level Vulkan semantics tests.
GLSL: Add binding map tests This PR adds a GLSL equivalent to the HLSL binding mapping tests for offsets and auto-numbering. The shaders are as equivalent as possible. The bindings of the base results match exactly between the two.
2016-09-27 02:02:44 +03:00
TEST_P(HlslIoMap, FromFile)
HLSL: Add location offsets per resource type This PR adds the ability to offset sampler, texture, and UBO bindings from provided base bindings, and to auto-number bindings that are not provided with explicit register numbers. The mechanism works as follows: - Offsets may be given on the command line for all stages, or individually for one or more single stages, in which case the offset will be auto-selected according to the stage being compiled. There is also an API to set them. The new command line options are --shift-sampler-binding, --shift-texture-binding, and --shift-UBO-binding. - Uniforms which are not given explicit bindings in the source code are auto-numbered if and only if they are in live code as determined by the algorithm used to build the reflection database, and the --auto-map-bindings option is given. This auto-numbering avoids using any binding slots which were explicitly provided in the code, whether or not that explicit use was live. E.g, "uniform Texture1D foo : register(t3);" with --shift-texture-binding 10 will reserve binding 13, whether or not foo is used in live code. - Shorter synonyms for the command line options are available. See the --help output. The testing infrastructure is slightly extended to allow use of the binding offset API, and two new tests spv.register.(no)autoassign.frag are added for comparing the resulting SPIR-V.
2016-09-08 00:20:19 +03:00
{
Gtests can be run on another source tree The gtest executable accepts a --test-root option to specify a root directory for test files. It defaults to the Test directory in the source tree from which the executable is built. For example, this lets us run test exectuables built with MinGW on Linux on a Windows machine with its own copy of the source tree.
2016-10-05 17:25:09 +03:00
loadFileCompileIoMapAndCheck(GlobalTestSettings.testRoot, GetParam().fileName,
HLSL: Add location offsets per resource type This PR adds the ability to offset sampler, texture, and UBO bindings from provided base bindings, and to auto-number bindings that are not provided with explicit register numbers. The mechanism works as follows: - Offsets may be given on the command line for all stages, or individually for one or more single stages, in which case the offset will be auto-selected according to the stage being compiled. There is also an API to set them. The new command line options are --shift-sampler-binding, --shift-texture-binding, and --shift-UBO-binding. - Uniforms which are not given explicit bindings in the source code are auto-numbered if and only if they are in live code as determined by the algorithm used to build the reflection database, and the --auto-map-bindings option is given. This auto-numbering avoids using any binding slots which were explicitly provided in the code, whether or not that explicit use was live. E.g, "uniform Texture1D foo : register(t3);" with --shift-texture-binding 10 will reserve binding 13, whether or not foo is used in live code. - Shorter synonyms for the command line options are available. See the --help output. The testing infrastructure is slightly extended to allow use of the binding offset API, and two new tests spv.register.(no)autoassign.frag are added for comparing the resulting SPIR-V.
2016-09-08 00:20:19 +03:00
Source::HLSL, Semantics::Vulkan,
Target::Spv, GetParam().entryPoint,
GetParam().baseSamplerBinding,
GetParam().baseTextureBinding,
Add UAV (image) binding offset and HLSL register support This PR adds: 1. The "u" register class for RW* objects. 2. --shift-image-bindings (== --sib), analogous to --shift-texture-bindings etc. 3. Case insensitive reg classes. 4. Tests for above.
2016-11-01 19:31:42 +03:00
GetParam().baseImageBinding,
HLSL: Add location offsets per resource type This PR adds the ability to offset sampler, texture, and UBO bindings from provided base bindings, and to auto-number bindings that are not provided with explicit register numbers. The mechanism works as follows: - Offsets may be given on the command line for all stages, or individually for one or more single stages, in which case the offset will be auto-selected according to the stage being compiled. There is also an API to set them. The new command line options are --shift-sampler-binding, --shift-texture-binding, and --shift-UBO-binding. - Uniforms which are not given explicit bindings in the source code are auto-numbered if and only if they are in live code as determined by the algorithm used to build the reflection database, and the --auto-map-bindings option is given. This auto-numbering avoids using any binding slots which were explicitly provided in the code, whether or not that explicit use was live. E.g, "uniform Texture1D foo : register(t3);" with --shift-texture-binding 10 will reserve binding 13, whether or not foo is used in live code. - Shorter synonyms for the command line options are available. See the --help output. The testing infrastructure is slightly extended to allow use of the binding offset API, and two new tests spv.register.(no)autoassign.frag are added for comparing the resulting SPIR-V.
2016-09-08 00:20:19 +03:00
GetParam().baseUboBinding,
Add reflection queries for thread local size and ssbo auto-binding New command line option --shift-ssbo-binding mirrors --shift-ubo-binding, etc. New reflection query getLocalSize(int dim) queries local size, e.g, CS threads.
2017-02-22 03:19:08 +03:00
GetParam().baseSsboBinding,
Fix defects in uniform array flattening Fix for two defects as follows: - The IO mapping traverser was not setting inVisit, and would skip some AST nodes. Depending on the order of nodes, this could have prevented the binding from showing up in the generated SPIR-V. - If a uniform array was flattened, each of the flattened scalars from the array is still a (now-scalar) uniform. It was being converted to a temporary.
2016-09-22 23:35:23 +03:00
GetParam().autoMapBindings,
GetParam().flattenUniforms);
HLSL: Add location offsets per resource type This PR adds the ability to offset sampler, texture, and UBO bindings from provided base bindings, and to auto-number bindings that are not provided with explicit register numbers. The mechanism works as follows: - Offsets may be given on the command line for all stages, or individually for one or more single stages, in which case the offset will be auto-selected according to the stage being compiled. There is also an API to set them. The new command line options are --shift-sampler-binding, --shift-texture-binding, and --shift-UBO-binding. - Uniforms which are not given explicit bindings in the source code are auto-numbered if and only if they are in live code as determined by the algorithm used to build the reflection database, and the --auto-map-bindings option is given. This auto-numbering avoids using any binding slots which were explicitly provided in the code, whether or not that explicit use was live. E.g, "uniform Texture1D foo : register(t3);" with --shift-texture-binding 10 will reserve binding 13, whether or not foo is used in live code. - Shorter synonyms for the command line options are available. See the --help output. The testing infrastructure is slightly extended to allow use of the binding offset API, and two new tests spv.register.(no)autoassign.frag are added for comparing the resulting SPIR-V.
2016-09-08 00:20:19 +03:00
}
GLSL: Add binding map tests This PR adds a GLSL equivalent to the HLSL binding mapping tests for offsets and auto-numbering. The shaders are as equivalent as possible. The bindings of the base results match exactly between the two.
2016-09-27 02:02:44 +03:00
// GLSL-level Vulkan semantics tests.
TEST_P(GlslIoMap, FromFile)
{
Gtests can be run on another source tree The gtest executable accepts a --test-root option to specify a root directory for test files. It defaults to the Test directory in the source tree from which the executable is built. For example, this lets us run test exectuables built with MinGW on Linux on a Windows machine with its own copy of the source tree.
2016-10-05 17:25:09 +03:00
loadFileCompileIoMapAndCheck(GlobalTestSettings.testRoot, GetParam().fileName,
GLSL: Add binding map tests This PR adds a GLSL equivalent to the HLSL binding mapping tests for offsets and auto-numbering. The shaders are as equivalent as possible. The bindings of the base results match exactly between the two.
2016-09-27 02:02:44 +03:00
Source::GLSL, Semantics::Vulkan,
Target::Spv, GetParam().entryPoint,
GetParam().baseSamplerBinding,
GetParam().baseTextureBinding,
Add UAV (image) binding offset and HLSL register support This PR adds: 1. The "u" register class for RW* objects. 2. --shift-image-bindings (== --sib), analogous to --shift-texture-bindings etc. 3. Case insensitive reg classes. 4. Tests for above.
2016-11-01 19:31:42 +03:00
GetParam().baseImageBinding,
GLSL: Add binding map tests This PR adds a GLSL equivalent to the HLSL binding mapping tests for offsets and auto-numbering. The shaders are as equivalent as possible. The bindings of the base results match exactly between the two.
2016-09-27 02:02:44 +03:00
GetParam().baseUboBinding,
Add reflection queries for thread local size and ssbo auto-binding New command line option --shift-ssbo-binding mirrors --shift-ubo-binding, etc. New reflection query getLocalSize(int dim) queries local size, e.g, CS threads.
2017-02-22 03:19:08 +03:00
GetParam().baseSsboBinding,
GLSL: Add binding map tests This PR adds a GLSL equivalent to the HLSL binding mapping tests for offsets and auto-numbering. The shaders are as equivalent as possible. The bindings of the base results match exactly between the two.
2016-09-27 02:02:44 +03:00
GetParam().autoMapBindings,
GetParam().flattenUniforms);
}
Parser: Implement extension GL_AMD_gpu_shader_half_float. - Add built-in types: float16_t, f16vec, f16mat. - Add support of half float constant: hf, HF. - Extend built-in floating-point operators: +, -, *, /, ++, --, +=, -=, *=, /=, ==, !=, >=, <=, >, <. - Add support of type conversions: float16_t -> XXX, XXX -> float16_t. - Add new built-in functions.
2016-07-29 11:00:05 +03:00
#ifdef AMD_EXTENSIONS
// Compiling GLSL to SPIR-V under Vulkan semantics (AMD extensions enabled).
// Expected to successfully generate SPIR-V.
TEST_P(CompileVulkanToSpirvTestAMD, FromFile)
{
Gtests can be run on another source tree The gtest executable accepts a --test-root option to specify a root directory for test files. It defaults to the Test directory in the source tree from which the executable is built. For example, this lets us run test exectuables built with MinGW on Linux on a Windows machine with its own copy of the source tree.
2016-10-05 17:25:09 +03:00
loadFileCompileAndCheck(GlobalTestSettings.testRoot, GetParam(),
Parser: Implement extension GL_AMD_gpu_shader_half_float. - Add built-in types: float16_t, f16vec, f16mat. - Add support of half float constant: hf, HF. - Extend built-in floating-point operators: +, -, *, /, ++, --, +=, -=, *=, /=, ==, !=, >=, <=, >, <. - Add support of type conversions: float16_t -> XXX, XXX -> float16_t. - Add new built-in functions.
2016-07-29 11:00:05 +03:00
Source::GLSL, Semantics::Vulkan,
Target::Spv);
}
#endif
support SPV_NV_viewport_array2 and SPV_NV_stereo_view_rendering
2017-01-13 12:10:53 +03:00
#ifdef NV_EXTENSIONS
// Compiling GLSL to SPIR-V under Vulkan semantics (AMD extensions enabled).
// Expected to successfully generate SPIR-V.
TEST_P(CompileVulkanToSpirvTestNV, FromFile)
{
loadFileCompileAndCheck(GlobalTestSettings.testRoot, GetParam(),
Source::GLSL, Semantics::Vulkan,
Target::Spv);
}
#endif
Pure Texture to Sampled Texture Transform Adds a transformation step to the post processing step. Two modes are available: 1) keep - Keeps samplers, textures and sampled textures as is 2) transform pure texture into sampled texture and remove pure samplers - removes all pure samplers - transforms all pure textures into its sampled counter part Change-Id: If54972e8052961db66c23f4b7e719d363cf6edbd
2017-04-26 00:31:03 +03:00
TEST_P(CompileUpgradeTextureToSampledTextureAndDropSamplersTest, FromFile)
{
loadCompileUpgradeTextureToSampledTextureAndDropSamplersAndCheck(GlobalTestSettings.testRoot,
GetParam(),
Source::GLSL,
Semantics::Vulkan,
Target::Spv);
}
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
// clang-format off
INSTANTIATE_TEST_CASE_P(
SPV: Turn on atomic-storage functionality for SPIR-V. This is used by OpenGL, but not Vulkan. Includes: - atomicCounter, atomicIncrement, atomicCounterDecrement - atomic_uint layout-offset checking - AtomicStorage capability
2016-07-07 22:20:00 +03:00
Glsl, CompileVulkanToSpirvTest,
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
::testing::ValuesIn(std::vector<std::string>({
// Test looping constructs.
// No tests yet for making sure break and continue from a nested loop
// goes to the innermost target.
"spv.do-simple.vert",
"spv.do-while-continue-break.vert",
"spv.for-complex-condition.vert",
"spv.for-continue-break.vert",
"spv.for-simple.vert",
"spv.for-notest.vert",
"spv.for-nobody.vert",
"spv.while-continue-break.vert",
"spv.while-simple.vert",
// vulkan-specific tests
"spv.set.vert",
"spv.double.comp",
"spv.100ops.frag",
"spv.130.frag",
"spv.140.frag",
"spv.150.geom",
"spv.150.vert",
"spv.300BuiltIns.vert",
"spv.300layout.frag",
"spv.300layout.vert",
"spv.300layoutp.vert",
"spv.310.comp",
Front-end: Make ES floatBitsToInt/intBitsToFloat/etc. all be highp.
2016-06-15 05:38:58 +03:00
"spv.310.bitcast.frag",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
"spv.330.geom",
"spv.400.frag",
"spv.400.tesc",
"spv.400.tese",
"spv.420.geom",
SPV: Fix issue #376; declare Geometry capability for fragment use of gl_Layer.
2016-07-08 23:05:15 +03:00
"spv.430.frag",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
"spv.430.vert",
SPV: Don't decorate locations within an array, it doesn't make sense. This fixes issue #360.
2016-07-01 06:47:35 +03:00
"spv.450.tesc",
Fix #790: Don't emit NV-specific interface members unless enabled by extension.
2017-03-25 03:38:16 +03:00
"spv.450.geom",
"spv.450.noRedecl.tesc",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
"spv.accessChain.frag",
"spv.aggOps.frag",
"spv.always-discard.frag",
"spv.always-discard2.frag",
Implement SPV_KHR_post_depth_coverage Added support for both extension GL_ARB_post_depth_coverage and GL_EXT_post_depth_coverage.
2017-07-01 03:14:30 +03:00
"spv.arbPostDepthCoverage.frag",
"spv.arbPostDepthCoverage_Error.frag",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
"spv.bitCast.frag",
"spv.bool.vert",
"spv.boolInBlock.frag",
"spv.branch-return.vert",
"spv.conditionalDiscard.frag",
"spv.conversion.frag",
"spv.dataOut.frag",
"spv.dataOutIndirect.frag",
"spv.dataOutIndirect.vert",
"spv.deepRvalue.frag",
"spv.depthOut.frag",
SPV: Implement new extensions GL_KHX_device_group and GL_KHX_multiview. These correspond to SPV_KHR_device_group and SPV_KHR_multiview. Also, bring tests up to date with Khronos internals, and some misc. related changes.
2017-02-27 11:20:51 +03:00
"spv.deviceGroup.frag",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
"spv.discard-dce.frag",
SPV: Implement new extensions GL_KHX_device_group and GL_KHX_multiview. These correspond to SPV_KHR_device_group and SPV_KHR_multiview. Also, bring tests up to date with Khronos internals, and some misc. related changes.
2017-02-27 11:20:51 +03:00
"spv.drawParams.vert",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
"spv.doWhileLoop.frag",
"spv.earlyReturnDiscard.frag",
Implement SPV_KHR_post_depth_coverage Added support for both extension GL_ARB_post_depth_coverage and GL_EXT_post_depth_coverage.
2017-07-01 03:14:30 +03:00
"spv.extPostDepthCoverage.frag",
"spv.extPostDepthCoverage_Error.frag",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
"spv.flowControl.frag",
"spv.forLoop.frag",
"spv.forwardFun.frag",
"spv.functionCall.frag",
SPV: Handle nested opaque types as function parameters.
2017-02-25 05:15:46 +03:00
"spv.functionNestedOpaque.vert",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
"spv.functionSemantics.frag",
SPV: Implement new extensions GL_KHX_device_group and GL_KHX_multiview. These correspond to SPV_KHR_device_group and SPV_KHR_multiview. Also, bring tests up to date with Khronos internals, and some misc. related changes.
2017-02-27 11:20:51 +03:00
"spv.GeometryShaderPassthrough.geom",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
"spv.interpOps.frag",
Implement the extension GL_ARB_gpu_shader_int64 - Add new keyword int64_t/uint64_t/i64vec/u64vec. - Support 64-bit integer literals (dec/hex/oct). - Support built-in operators for 64-bit integer type. - Add implicit and explicit type conversion for 64-bit integer type. - Add new built-in functions defined in this extension.
2016-04-22 11:51:45 +03:00
"spv.int64.frag",
SPV: Implement new extensions GL_KHX_device_group and GL_KHX_multiview. These correspond to SPV_KHR_device_group and SPV_KHR_multiview. Also, bring tests up to date with Khronos internals, and some misc. related changes.
2017-02-27 11:20:51 +03:00
"spv.intOps.vert",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
"spv.layoutNested.vert",
"spv.length.frag",
"spv.localAggregates.frag",
"spv.loops.frag",
"spv.loopsArtificial.frag",
"spv.matFun.vert",
"spv.matrix.frag",
"spv.matrix2.frag",
"spv.memoryQualifier.frag",
"spv.merge-unreachable.frag",
SPV: Flatten structs for copy when they are GLSL type aliases. Addresses issue #304 and issue #307 by replacing unmatched type OpStores with per-member copies. Covers assignment statements and most argument passing, but does not yet cover r-value-based argument passing.
2016-09-02 20:20:21 +03:00
"spv.multiStruct.comp",
"spv.multiStructFuncall.frag",
SPV: Implement new extensions GL_KHX_device_group and GL_KHX_multiview. These correspond to SPV_KHR_device_group and SPV_KHR_multiview. Also, bring tests up to date with Khronos internals, and some misc. related changes.
2017-02-27 11:20:51 +03:00
"spv.multiView.frag",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
"spv.newTexture.frag",
"spv.noDeadDecorations.vert",
"spv.nonSquare.vert",
SPV: Implement new extensions GL_KHX_device_group and GL_KHX_multiview. These correspond to SPV_KHR_device_group and SPV_KHR_multiview. Also, bring tests up to date with Khronos internals, and some misc. related changes.
2017-02-27 11:20:51 +03:00
"spv.noWorkgroup.comp",
SPV: Distinguish between SPV and non-SPV rules for member overlap.
2016-10-07 20:50:25 +03:00
"spv.offsets.frag",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
"spv.Operations.frag",
"spv.precision.frag",
"spv.prepost.frag",
"spv.qualifiers.vert",
SPV: Implement new extensions GL_KHX_device_group and GL_KHX_multiview. These correspond to SPV_KHR_device_group and SPV_KHR_multiview. Also, bring tests up to date with Khronos internals, and some misc. related changes.
2017-02-27 11:20:51 +03:00
"spv.sampleMaskOverrideCoverage.frag",
Implement the extension GL_ARB_shader_ballot Add new built-in variables and functions to the parser (SPIR-V tokens are missing).
2016-04-14 11:53:07 +03:00
"spv.shaderBallot.comp",
SPV: Implement extension SPV_KHR_shader_draw_parameters.
2016-07-22 13:15:31 +03:00
"spv.shaderDrawParams.vert",
Implement the extension GL_ARB_shader_group_vote.
2016-05-05 15:38:33 +03:00
"spv.shaderGroupVote.comp",
Implement extension GL_ARB_shader_stencil_export
2017-06-29 12:46:34 +03:00
"spv.shaderStencilExport.frag",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
"spv.shiftOps.frag",
"spv.simpleFunctionCall.frag",
"spv.simpleMat.vert",
"spv.sparseTexture.frag",
"spv.sparseTextureClamp.frag",
"spv.structAssignment.frag",
"spv.structDeref.frag",
"spv.structure.frag",
"spv.switch.frag",
"spv.swizzle.frag",
SPV: For AST interpolateAt* ops consuming a swizzle, invert the order. Apply the interpolation first, then apply the swizzle to the result, the inverse of the order requested by the AST. This fixes issue #411.
2016-07-26 21:50:38 +03:00
"spv.swizzleInversion.frag",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
"spv.test.frag",
"spv.test.vert",
"spv.texture.frag",
"spv.texture.vert",
SPV: Support texelFetch() on a textureBuffer (no sampler). Fixes Vulkan Khronos-internal issue 770.
2017-05-08 09:12:28 +03:00
"spv.textureBuffer.vert",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
"spv.image.frag",
"spv.types.frag",
"spv.uint.frag",
"spv.uniformArray.frag",
"spv.variableArrayIndex.frag",
"spv.varyingArray.frag",
"spv.varyingArrayIndirect.frag",
"spv.voidFunction.frag",
"spv.whileLoop.frag",
"spv.AofA.frag",
"spv.queryL.frag",
"spv.separate.frag",
"spv.shortCircuit.frag",
"spv.pushConstant.vert",
SPV: Allow push constant buffers to not have an instance name.
2017-03-06 18:51:35 +03:00
"spv.pushConstantAnon.vert",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
"spv.subpass.frag",
"spv.specConstant.vert",
"spv.specConstant.comp",
"spv.specConstantComposite.vert",
Remove duplicated tests from the runtests script.
2016-05-04 21:35:16 +03:00
"spv.specConstantOperations.vert",
SPV: Support test #pragma for generating the StorageBuffer storage class. Longer term, this storage class should be generated based on the mode of compilation.
2017-04-20 03:34:49 +03:00
"spv.storageBuffer.vert",
Remove duplicated tests from the runtests script.
2016-05-04 21:35:16 +03:00
"spv.precise.tese",
"spv.precise.tesc",
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
})),
Add support for testing file-based HLSL source code in GTest.
2016-05-16 19:50:30 +03:00
FileNameAsCustomTestSuffix
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
);
HLSL: Add location offsets per resource type This PR adds the ability to offset sampler, texture, and UBO bindings from provided base bindings, and to auto-number bindings that are not provided with explicit register numbers. The mechanism works as follows: - Offsets may be given on the command line for all stages, or individually for one or more single stages, in which case the offset will be auto-selected according to the stage being compiled. There is also an API to set them. The new command line options are --shift-sampler-binding, --shift-texture-binding, and --shift-UBO-binding. - Uniforms which are not given explicit bindings in the source code are auto-numbered if and only if they are in live code as determined by the algorithm used to build the reflection database, and the --auto-map-bindings option is given. This auto-numbering avoids using any binding slots which were explicitly provided in the code, whether or not that explicit use was live. E.g, "uniform Texture1D foo : register(t3);" with --shift-texture-binding 10 will reserve binding 13, whether or not foo is used in live code. - Shorter synonyms for the command line options are available. See the --help output. The testing infrastructure is slightly extended to allow use of the binding offset API, and two new tests spv.register.(no)autoassign.frag are added for comparing the resulting SPIR-V.
2016-09-08 00:20:19 +03:00
// clang-format off
INSTANTIATE_TEST_CASE_P(
GLSL: Add binding map tests This PR adds a GLSL equivalent to the HLSL binding mapping tests for offsets and auto-numbering. The shaders are as equivalent as possible. The bindings of the base results match exactly between the two.
2016-09-27 02:02:44 +03:00
Hlsl, HlslIoMap,
HLSL: Add location offsets per resource type This PR adds the ability to offset sampler, texture, and UBO bindings from provided base bindings, and to auto-number bindings that are not provided with explicit register numbers. The mechanism works as follows: - Offsets may be given on the command line for all stages, or individually for one or more single stages, in which case the offset will be auto-selected according to the stage being compiled. There is also an API to set them. The new command line options are --shift-sampler-binding, --shift-texture-binding, and --shift-UBO-binding. - Uniforms which are not given explicit bindings in the source code are auto-numbered if and only if they are in live code as determined by the algorithm used to build the reflection database, and the --auto-map-bindings option is given. This auto-numbering avoids using any binding slots which were explicitly provided in the code, whether or not that explicit use was live. E.g, "uniform Texture1D foo : register(t3);" with --shift-texture-binding 10 will reserve binding 13, whether or not foo is used in live code. - Shorter synonyms for the command line options are available. See the --help output. The testing infrastructure is slightly extended to allow use of the binding offset API, and two new tests spv.register.(no)autoassign.frag are added for comparing the resulting SPIR-V.
2016-09-08 00:20:19 +03:00
::testing::ValuesIn(std::vector<IoMapData>{
Add reflection queries for thread local size and ssbo auto-binding New command line option --shift-ssbo-binding mirrors --shift-ubo-binding, etc. New reflection query getLocalSize(int dim) queries local size, e.g, CS threads.
2017-02-22 03:19:08 +03:00
{ "spv.register.autoassign.frag", "main_ep", 5, 10, 0, 20, 30, true, false },
{ "spv.register.noautoassign.frag", "main_ep", 5, 10, 0, 15, 30, false, false },
{ "spv.register.autoassign-2.frag", "main", 5, 10, 0, 15, 30, true, true },
{ "spv.buffer.autoassign.frag", "main", 5, 10, 0, 15, 30, true, true },
{ "spv.ssbo.autoassign.frag", "main", 5, 10, 0, 15, 30, true, true },
HLSL: iomapper: Fix #914. Tolerate user aliasing of bindings. Because it is valid in HLSL to alias bindings: A) remove validation that aliasing is not done B) make the algorithms tolerate aliasing
2017-06-02 03:16:33 +03:00
{ "spv.ssboAlias.frag", "main", 0, 0, 0, 0, 83, true, false },
Add reflection queries for thread local size and ssbo auto-binding New command line option --shift-ssbo-binding mirrors --shift-ubo-binding, etc. New reflection query getLocalSize(int dim) queries local size, e.g, CS threads.
2017-02-22 03:19:08 +03:00
{ "spv.rw.autoassign.frag", "main", 5, 10, 20, 15, 30, true, true },
Implement extensions GL_OVR_multiview and GL_OVR_multiview2 They are almost identical to the already supported GL_EXT_multiview
2017-06-30 20:33:51 +03:00
{ "spv.register.autoassign.rangetest.frag", "main",
Check for out-of-range bindings during IO mapping.
2016-10-05 22:40:13 +03:00
glslang::TQualifier::layoutBindingEnd-2,
glslang::TQualifier::layoutBindingEnd+5,
Add reflection queries for thread local size and ssbo auto-binding New command line option --shift-ssbo-binding mirrors --shift-ubo-binding, etc. New reflection query getLocalSize(int dim) queries local size, e.g, CS threads.
2017-02-22 03:19:08 +03:00
20, 30, true, false },
HLSL: Add location offsets per resource type This PR adds the ability to offset sampler, texture, and UBO bindings from provided base bindings, and to auto-number bindings that are not provided with explicit register numbers. The mechanism works as follows: - Offsets may be given on the command line for all stages, or individually for one or more single stages, in which case the offset will be auto-selected according to the stage being compiled. There is also an API to set them. The new command line options are --shift-sampler-binding, --shift-texture-binding, and --shift-UBO-binding. - Uniforms which are not given explicit bindings in the source code are auto-numbered if and only if they are in live code as determined by the algorithm used to build the reflection database, and the --auto-map-bindings option is given. This auto-numbering avoids using any binding slots which were explicitly provided in the code, whether or not that explicit use was live. E.g, "uniform Texture1D foo : register(t3);" with --shift-texture-binding 10 will reserve binding 13, whether or not foo is used in live code. - Shorter synonyms for the command line options are available. See the --help output. The testing infrastructure is slightly extended to allow use of the binding offset API, and two new tests spv.register.(no)autoassign.frag are added for comparing the resulting SPIR-V.
2016-09-08 00:20:19 +03:00
}),
FileNameAsCustomTestSuffixIoMap
GLSL: Add binding map tests This PR adds a GLSL equivalent to the HLSL binding mapping tests for offsets and auto-numbering. The shaders are as equivalent as possible. The bindings of the base results match exactly between the two.
2016-09-27 02:02:44 +03:00
);
// clang-format off
INSTANTIATE_TEST_CASE_P(
Hlsl, GlslIoMap,
::testing::ValuesIn(std::vector<IoMapData>{
Add reflection queries for thread local size and ssbo auto-binding New command line option --shift-ssbo-binding mirrors --shift-ubo-binding, etc. New reflection query getLocalSize(int dim) queries local size, e.g, CS threads.
2017-02-22 03:19:08 +03:00
{ "spv.glsl.register.autoassign.frag", "main", 5, 10, 0, 20, 30, true, false },
{ "spv.glsl.register.noautoassign.frag", "main", 5, 10, 0, 15, 30, false, false },
GLSL: Add binding map tests This PR adds a GLSL equivalent to the HLSL binding mapping tests for offsets and auto-numbering. The shaders are as equivalent as possible. The bindings of the base results match exactly between the two.
2016-09-27 02:02:44 +03:00
}),
FileNameAsCustomTestSuffixIoMap
HLSL: Add location offsets per resource type This PR adds the ability to offset sampler, texture, and UBO bindings from provided base bindings, and to auto-number bindings that are not provided with explicit register numbers. The mechanism works as follows: - Offsets may be given on the command line for all stages, or individually for one or more single stages, in which case the offset will be auto-selected according to the stage being compiled. There is also an API to set them. The new command line options are --shift-sampler-binding, --shift-texture-binding, and --shift-UBO-binding. - Uniforms which are not given explicit bindings in the source code are auto-numbered if and only if they are in live code as determined by the algorithm used to build the reflection database, and the --auto-map-bindings option is given. This auto-numbering avoids using any binding slots which were explicitly provided in the code, whether or not that explicit use was live. E.g, "uniform Texture1D foo : register(t3);" with --shift-texture-binding 10 will reserve binding 13, whether or not foo is used in live code. - Shorter synonyms for the command line options are available. See the --help output. The testing infrastructure is slightly extended to allow use of the binding offset API, and two new tests spv.register.(no)autoassign.frag are added for comparing the resulting SPIR-V.
2016-09-08 00:20:19 +03:00
);
SPV: Turn on atomic-storage functionality for SPIR-V. This is used by OpenGL, but not Vulkan. Includes: - atomicCounter, atomicIncrement, atomicCounterDecrement - atomic_uint layout-offset checking - AtomicStorage capability
2016-07-07 22:20:00 +03:00
// clang-format off
INSTANTIATE_TEST_CASE_P(
Glsl, CompileOpenGLToSpirvTest,
::testing::ValuesIn(std::vector<std::string>({
"spv.atomic.comp",
SPV: Further refine OpenGL vs. Vulkan differences for SPIR-V. Includes adding test cases to verify the differences.
2016-07-08 02:46:42 +03:00
"spv.glFragColor.frag",
"spv.specConst.vert",
Implement extensions GL_OVR_multiview and GL_OVR_multiview2 They are almost identical to the already supported GL_EXT_multiview
2017-06-30 20:33:51 +03:00
"spv.OVR_multiview.vert",
SPV: Turn on atomic-storage functionality for SPIR-V. This is used by OpenGL, but not Vulkan. Includes: - atomicCounter, atomicIncrement, atomicCounterDecrement - atomic_uint layout-offset checking - AtomicStorage capability
2016-07-07 22:20:00 +03:00
})),
FileNameAsCustomTestSuffix
);
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
INSTANTIATE_TEST_CASE_P(
Glsl, VulkanSemantics,
::testing::ValuesIn(std::vector<std::string>({
"vulkan.frag",
"vulkan.vert",
"vulkan.comp",
})),
Add support for testing file-based HLSL source code in GTest.
2016-05-16 19:50:30 +03:00
FileNameAsCustomTestSuffix
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
);
Vulkan: Finish semantics for what creates spec-const-semantics. Note: This required adding a new test mode to see the AST for vulkan tests. This also required reworking some deeper parts of type creation, regarding when storage qualification and constness is deduced bottom-up or dictated top-down.
2016-05-24 01:07:07 +03:00
SPV: Further refine OpenGL vs. Vulkan differences for SPIR-V. Includes adding test cases to verify the differences.
2016-07-08 02:46:42 +03:00
INSTANTIATE_TEST_CASE_P(
Glsl, OpenGLSemantics,
::testing::ValuesIn(std::vector<std::string>({
"glspv.esversion.vert",
"glspv.version.frag",
"glspv.version.vert",
"glspv.frag",
"glspv.vert",
})),
FileNameAsCustomTestSuffix
);
Vulkan: Finish semantics for what creates spec-const-semantics. Note: This required adding a new test mode to see the AST for vulkan tests. This also required reworking some deeper parts of type creation, regarding when storage qualification and constness is deduced bottom-up or dictated top-down.
2016-05-24 01:07:07 +03:00
INSTANTIATE_TEST_CASE_P(
Glsl, VulkanAstSemantics,
::testing::ValuesIn(std::vector<std::string>({
"vulkan.ast.vert",
})),
FileNameAsCustomTestSuffix
);
Parser: Implement extension GL_AMD_gpu_shader_half_float. - Add built-in types: float16_t, f16vec, f16mat. - Add support of half float constant: hf, HF. - Extend built-in floating-point operators: +, -, *, /, ++, --, +=, -=, *=, /=, ==, !=, >=, <=, >, <. - Add support of type conversions: float16_t -> XXX, XXX -> float16_t. - Add new built-in functions.
2016-07-29 11:00:05 +03:00
#ifdef AMD_EXTENSIONS
INSTANTIATE_TEST_CASE_P(
Glsl, CompileVulkanToSpirvTestAMD,
::testing::ValuesIn(std::vector<std::string>({
"spv.float16.frag",
Implement extension GL_AMD_gpu_shader_int16 - Add int16 types (int16_t, uint16_t, i16vec, u16vec). - Add int16 support to GLSL operators. - Add int16 type conversions (to int16, from int16). - Add int16 built-in functions.
2017-03-24 08:41:14 +03:00
"spv.int16.frag",
Implement the extension GL_AMD_texture_gather_bias_lod
2016-11-17 12:47:59 +03:00
"spv.shaderBallotAMD.comp",
"spv.textureGatherBiasLod.frag"
Parser: Implement extension GL_AMD_gpu_shader_half_float. - Add built-in types: float16_t, f16vec, f16mat. - Add support of half float constant: hf, HF. - Extend built-in floating-point operators: +, -, *, /, ++, --, +=, -=, *=, /=, ==, !=, >=, <=, >, <. - Add support of type conversions: float16_t -> XXX, XXX -> float16_t. - Add new built-in functions.
2016-07-29 11:00:05 +03:00
})),
FileNameAsCustomTestSuffix
);
#endif
support SPV_NV_viewport_array2 and SPV_NV_stereo_view_rendering
2017-01-13 12:10:53 +03:00
#ifdef NV_EXTENSIONS
INSTANTIATE_TEST_CASE_P(
Glsl, CompileVulkanToSpirvTestNV,
::testing::ValuesIn(std::vector<std::string>({
"spv.sampleMaskOverrideCoverage.frag",
"spv.GeometryShaderPassthrough.geom",
"spv.viewportArray2.vert",
"spv.viewportArray2.tesc",
"spv.stereoViewRendering.vert",
"spv.stereoViewRendering.tesc",
Implement NVX_multiview_per_view_attributes
2017-02-15 01:52:34 +03:00
"spv.multiviewPerViewAttributes.vert",
"spv.multiviewPerViewAttributes.tesc",
support SPV_NV_viewport_array2 and SPV_NV_stereo_view_rendering
2017-01-13 12:10:53 +03:00
})),
FileNameAsCustomTestSuffix
);
#endif
Pure Texture to Sampled Texture Transform Adds a transformation step to the post processing step. Two modes are available: 1) keep - Keeps samplers, textures and sampled textures as is 2) transform pure texture into sampled texture and remove pure samplers - removes all pure samplers - transforms all pure textures into its sampled counter part Change-Id: If54972e8052961db66c23f4b7e719d363cf6edbd
2017-04-26 00:31:03 +03:00
INSTANTIATE_TEST_CASE_P(
Glsl, CompileUpgradeTextureToSampledTextureAndDropSamplersTest,
::testing::ValuesIn(std::vector<std::string>({
"spv.texture.sampler.transform.frag",
})),
FileNameAsCustomTestSuffix
);
Link in Google Test framework. The existing test harness is a homemade shell script. All the tests and the expected results are written in plain text files. The harness just reads in a test, invoke the glslangValidator binary on it, and compare the result with the golden file. All tests are kinda integration tests. This patch add Google Test as an external project, which provides a new harness for reading shader source files, compile to SPIR-V, and then compare with the expected output.
2016-03-05 00:22:34 +03:00
// clang-format on
} // anonymous namespace
} // namespace glslangtest