451 lines
17 KiB
C++
Raw Normal View History

2017-11-17 11:14:29 -05:00
/*
* SPIRVToMSLConverter.cpp
*
2020-01-07 16:47:29 -05:00
* Copyright (c) 2015-2020 The Brenwill Workshop Ltd. (http://www.brenwill.com)
2017-11-17 11:14:29 -05:00
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "SPIRVToMSLConverter.h"
#include "MVKCommonEnvironment.h"
#include "MVKStrings.h"
#include "FileSupport.h"
#include "SPIRVSupport.h"
#include <fstream>
2017-11-17 11:14:29 -05:00
using namespace mvk;
using namespace std;
using namespace SPIRV_CROSS_NAMESPACE;
2017-11-17 11:14:29 -05:00
#pragma mark -
#pragma mark SPIRVToMSLConversionConfiguration
2017-11-17 11:14:29 -05:00
// Returns whether the vector contains the value (using a matches(T&) comparison member function). */
template<class T>
bool containsMatching(const vector<T>& vec, const T& val) {
for (const T& vecVal : vec) { if (vecVal.matches(val)) { return true; } }
2017-11-17 11:14:29 -05:00
return false;
}
MVK_PUBLIC_SYMBOL bool SPIRVToMSLConversionOptions::matches(const SPIRVToMSLConversionOptions& other) const {
if (entryPointStage != other.entryPointStage) { return false; }
if (entryPointName != other.entryPointName) { return false; }
Add support for tessellation. At long last, tessellation comes to MoltenVK! With this change, clients will now be able to specify tessellation shaders when creating pipelines, and then draw tessellated patches with them. Unfortunately, there seem to be a few gotchas with tessellation in Metal. For one thing, tessellation pipelines in Metal are structured very differently from Vulkan. There is no tessellation control or even vertex stage. Instead, the tessellation evaluation shader takes the place of the vertex function as a "post-tessellation vertex function." The tessellation levels are supplied in a buffer to the tessellator, which you are expected to populate. The most common way to do this is by running a compute shader. MoltenVK thus runs the vertex shader and tessellation control shader by themselves; a single `VkPipeline` object then requires at least *three* `MTLPipelineState` objects. But wait, there's more! The tessellation-control-as-compute stage uses Metal's support for vertex-style stage input to a compute shader. But, this support requires one to declare indexing *ahead of time*, when the pipeline state is created. So a single `VkPipeline` object could have as many as *five* `MTLPipelineState` objects. Further, if there are more output than input control points for the tessellation control stage, then later invocations may end up fetching the wrong attributes! To get around this, this change uses index buffers to ensure that all tessellation control shaders see the correct input. Unfortunately, in the indexed draw case, this means that the incoming index buffer needs to be munged. Instancing is another pain point here. In Vulkan, as in OpenGL and Direct3D, instancing is done in the vertex shader; but in Metal, it is done at the tessellation evaluation stage. For this reason, only the vertex stage of a tessellated draw supports instancing. Additional memory is required to hold data for the extra vertices generated by instancing. This also requires still more munging of index buffers for indexed draws. Indirect draws are even more painful. Because the number of vertices and instances is unknown, storage for the maximum possible number of vertices must be allocated. This change imposes a totally arbitrary limit of 131072 vertices from a single draw, including all vertices generated by instancing. On a Mac, this requires about 194-256 MB of VRAM for all the temporary buffers. There are some possible optimizations here. If we could prove that the vertex shader's output doesn't depend on the instance ID, either directly or through a per-instance attribute, then we could avoid running the vertex and tess. control stages per instance, and take advantage of Metal's support for tess. eval instancing. If we could also prove that the vertex shader simply passes instance attributes through (similarly with the tess. control shader), we could do this for many more instanced draws as well. It should also be possible to cache the output from the tess. control stage; if the draw comes up again, we can then skip the vertex and tess. control stages entirely! Fixes #56 and #501.
2019-02-18 20:56:42 -06:00
if (tessPatchKind != other.tessPatchKind) { return false; }
if (numTessControlPoints != other.numTessControlPoints) { return false; }
if (shouldFlipVertexY != other.shouldFlipVertexY) { return false; }
if (memcmp(&mslOptions, &other.mslOptions, sizeof(mslOptions)) != 0) { return false; }
return true;
2017-11-17 11:14:29 -05:00
}
MVK_PUBLIC_SYMBOL std::string SPIRVToMSLConversionOptions::printMSLVersion(uint32_t mslVersion, bool includePatch) {
string verStr;
uint32_t major = mslVersion / 10000;
verStr += to_string(major);
uint32_t minor = (mslVersion - CompilerMSL::Options::make_msl_version(major)) / 100;
verStr += ".";
verStr += to_string(minor);
if (includePatch) {
uint32_t patch = mslVersion - CompilerMSL::Options::make_msl_version(major, minor);
verStr += ".";
verStr += to_string(patch);
}
return verStr;
}
MVK_PUBLIC_SYMBOL SPIRVToMSLConversionOptions::SPIRVToMSLConversionOptions() {
// Explicitly set mslOptions to defaults over cleared memory to ensure all instances
// have exactly the same memory layout when using memory comparison in matches().
memset(&mslOptions, 0, sizeof(mslOptions));
mslOptions = CompilerMSL::Options();
#if MVK_MACOS
mslOptions.platform = CompilerMSL::Options::macOS;
#endif
#if MVK_IOS
mslOptions.platform = CompilerMSL::Options::iOS;
#endif
#if MVK_TVOS
mslOptions.platform = CompilerMSL::Options::iOS;
#endif
mslOptions.pad_fragment_output_components = true;
}
MVK_PUBLIC_SYMBOL bool mvk::MSLShaderInput::matches(const mvk::MSLShaderInput& other) const {
if (shaderInput.location != other.shaderInput.location) { return false; }
if (shaderInput.format != other.shaderInput.format) { return false; }
if (shaderInput.builtin != other.shaderInput.builtin) { return false; }
if (shaderInput.vecsize != other.shaderInput.vecsize) { return false; }
if (binding != other.binding) { return false; }
return true;
2017-11-17 11:14:29 -05:00
}
MVK_PUBLIC_SYMBOL bool mvk::MSLResourceBinding::matches(const MSLResourceBinding& other) const {
if (resourceBinding.stage != other.resourceBinding.stage) { return false; }
if (resourceBinding.desc_set != other.resourceBinding.desc_set) { return false; }
if (resourceBinding.binding != other.resourceBinding.binding) { return false; }
if (resourceBinding.msl_buffer != other.resourceBinding.msl_buffer) { return false; }
if (resourceBinding.msl_texture != other.resourceBinding.msl_texture) { return false; }
if (resourceBinding.msl_sampler != other.resourceBinding.msl_sampler) { return false; }
if (requiresConstExprSampler != other.requiresConstExprSampler) { return false; }
// If requiresConstExprSampler is false, constExprSampler can be ignored
if (requiresConstExprSampler) {
if (constExprSampler.coord != other.constExprSampler.coord) { return false; }
if (constExprSampler.min_filter != other.constExprSampler.min_filter) { return false; }
if (constExprSampler.mag_filter != other.constExprSampler.mag_filter) { return false; }
if (constExprSampler.mip_filter != other.constExprSampler.mip_filter) { return false; }
if (constExprSampler.s_address != other.constExprSampler.s_address) { return false; }
if (constExprSampler.t_address != other.constExprSampler.t_address) { return false; }
if (constExprSampler.r_address != other.constExprSampler.r_address) { return false; }
if (constExprSampler.compare_func != other.constExprSampler.compare_func) { return false; }
if (constExprSampler.border_color != other.constExprSampler.border_color) { return false; }
if (constExprSampler.lod_clamp_min != other.constExprSampler.lod_clamp_min) { return false; }
if (constExprSampler.lod_clamp_max != other.constExprSampler.lod_clamp_max) { return false; }
if (constExprSampler.max_anisotropy != other.constExprSampler.max_anisotropy) { return false; }
if (constExprSampler.planes != other.constExprSampler.planes) { return false; }
if (constExprSampler.resolution != other.constExprSampler.resolution) { return false; }
if (constExprSampler.chroma_filter != other.constExprSampler.chroma_filter) { return false; }
if (constExprSampler.x_chroma_offset != other.constExprSampler.x_chroma_offset) { return false; }
if (constExprSampler.y_chroma_offset != other.constExprSampler.y_chroma_offset) { return false; }
for(uint32_t i = 0; i < 4; ++i)
if (constExprSampler.swizzle[i] != other.constExprSampler.swizzle[i]) { return false; }
if (constExprSampler.ycbcr_model != other.constExprSampler.ycbcr_model) { return false; }
if (constExprSampler.ycbcr_range != other.constExprSampler.ycbcr_range) { return false; }
if (constExprSampler.bpc != other.constExprSampler.bpc) { return false; }
if (constExprSampler.compare_enable != other.constExprSampler.compare_enable) { return false; }
if (constExprSampler.lod_clamp_enable != other.constExprSampler.lod_clamp_enable) { return false; }
if (constExprSampler.anisotropy_enable != other.constExprSampler.anisotropy_enable) { return false; }
if (constExprSampler.ycbcr_conversion_enable != other.constExprSampler.ycbcr_conversion_enable) { return false; }
}
return true;
2017-11-17 11:14:29 -05:00
}
MVK_PUBLIC_SYMBOL bool SPIRVToMSLConversionConfiguration::stageSupportsVertexAttributes() const {
return (options.entryPointStage == spv::ExecutionModelVertex ||
options.entryPointStage == spv::ExecutionModelTessellationControl ||
options.entryPointStage == spv::ExecutionModelTessellationEvaluation);
}
2017-11-17 11:14:29 -05:00
// Check them all in case inactive VA's duplicate locations used by active VA's.
MVK_PUBLIC_SYMBOL bool SPIRVToMSLConversionConfiguration::isShaderInputLocationUsed(uint32_t location) const {
for (auto& si : shaderInputs) {
if ((si.shaderInput.location == location) && si.isUsedByShader) { return true; }
2017-11-17 11:14:29 -05:00
}
return false;
}
MVK_PUBLIC_SYMBOL uint32_t SPIRVToMSLConversionConfiguration::countShaderInputsAt(uint32_t binding) const {
uint32_t siCnt = 0;
for (auto& si : shaderInputs) {
if ((si.binding == binding) && si.isUsedByShader) { siCnt++; }
}
return siCnt;
2017-11-17 11:14:29 -05:00
}
MVK_PUBLIC_SYMBOL void SPIRVToMSLConversionConfiguration::markAllInputsAndResourcesUsed() {
for (auto& si : shaderInputs) { si.isUsedByShader = true; }
for (auto& rb : resourceBindings) { rb.isUsedByShader = true; }
}
MVK_PUBLIC_SYMBOL bool SPIRVToMSLConversionConfiguration::matches(const SPIRVToMSLConversionConfiguration& other) const {
2017-11-17 11:14:29 -05:00
if ( !options.matches(other.options) ) { return false; }
for (const auto& si : shaderInputs) {
if (si.isUsedByShader && !containsMatching(other.shaderInputs, si)) { return false; }
}
2017-11-17 11:14:29 -05:00
for (const auto& rb : resourceBindings) {
if (rb.isUsedByShader && !containsMatching(other.resourceBindings, rb)) { return false; }
2017-11-17 11:14:29 -05:00
}
2017-11-17 11:14:29 -05:00
return true;
}
MVK_PUBLIC_SYMBOL void SPIRVToMSLConversionConfiguration::alignWith(const SPIRVToMSLConversionConfiguration& srcContext) {
2017-11-17 11:14:29 -05:00
for (auto& si : shaderInputs) {
si.isUsedByShader = false;
for (auto& srcSI : srcContext.shaderInputs) {
if (si.matches(srcSI)) { si.isUsedByShader = srcSI.isUsedByShader; }
}
}
2017-11-17 11:14:29 -05:00
for (auto& rb : resourceBindings) {
rb.isUsedByShader = false;
for (auto& srcRB : srcContext.resourceBindings) {
if (rb.matches(srcRB)) { rb.isUsedByShader = srcRB.isUsedByShader; }
}
}
}
#pragma mark -
#pragma mark SPIRVToMSLConverter
MVK_PUBLIC_SYMBOL void SPIRVToMSLConverter::setSPIRV(const uint32_t* spirvCode, size_t length) {
_spirv.clear(); // Clear for reuse
_spirv.reserve(length);
for (size_t i = 0; i < length; i++) {
_spirv.push_back(spirvCode[i]);
}
}
MVK_PUBLIC_SYMBOL bool SPIRVToMSLConverter::convert(SPIRVToMSLConversionConfiguration& context,
2017-11-17 11:14:29 -05:00
bool shouldLogSPIRV,
bool shouldLogMSL,
bool shouldLogGLSL) {
// Uncomment to write SPIR-V to file as a debugging aid
// ofstream spvFile("spirv.spv", ios::binary);
// spvFile.write((char*)_spirv.data(), _spirv.size() << 2);
// spvFile.close();
2017-11-17 11:14:29 -05:00
_wasConverted = true;
_resultLog.clear();
_msl.clear();
_shaderConversionResults.reset();
2017-11-17 11:14:29 -05:00
if (shouldLogSPIRV) { logSPIRV("Converting"); }
SPIRV_CROSS_NAMESPACE::CompilerMSL* pMSLCompiler = nullptr;
#ifndef SPIRV_CROSS_EXCEPTIONS_TO_ASSERTIONS
try {
#endif
pMSLCompiler = new SPIRV_CROSS_NAMESPACE::CompilerMSL(_spirv);
if (context.options.hasEntryPoint()) {
pMSLCompiler->set_entry_point(context.options.entryPointName, context.options.entryPointStage);
}
Add support for tessellation. At long last, tessellation comes to MoltenVK! With this change, clients will now be able to specify tessellation shaders when creating pipelines, and then draw tessellated patches with them. Unfortunately, there seem to be a few gotchas with tessellation in Metal. For one thing, tessellation pipelines in Metal are structured very differently from Vulkan. There is no tessellation control or even vertex stage. Instead, the tessellation evaluation shader takes the place of the vertex function as a "post-tessellation vertex function." The tessellation levels are supplied in a buffer to the tessellator, which you are expected to populate. The most common way to do this is by running a compute shader. MoltenVK thus runs the vertex shader and tessellation control shader by themselves; a single `VkPipeline` object then requires at least *three* `MTLPipelineState` objects. But wait, there's more! The tessellation-control-as-compute stage uses Metal's support for vertex-style stage input to a compute shader. But, this support requires one to declare indexing *ahead of time*, when the pipeline state is created. So a single `VkPipeline` object could have as many as *five* `MTLPipelineState` objects. Further, if there are more output than input control points for the tessellation control stage, then later invocations may end up fetching the wrong attributes! To get around this, this change uses index buffers to ensure that all tessellation control shaders see the correct input. Unfortunately, in the indexed draw case, this means that the incoming index buffer needs to be munged. Instancing is another pain point here. In Vulkan, as in OpenGL and Direct3D, instancing is done in the vertex shader; but in Metal, it is done at the tessellation evaluation stage. For this reason, only the vertex stage of a tessellated draw supports instancing. Additional memory is required to hold data for the extra vertices generated by instancing. This also requires still more munging of index buffers for indexed draws. Indirect draws are even more painful. Because the number of vertices and instances is unknown, storage for the maximum possible number of vertices must be allocated. This change imposes a totally arbitrary limit of 131072 vertices from a single draw, including all vertices generated by instancing. On a Mac, this requires about 194-256 MB of VRAM for all the temporary buffers. There are some possible optimizations here. If we could prove that the vertex shader's output doesn't depend on the instance ID, either directly or through a per-instance attribute, then we could avoid running the vertex and tess. control stages per instance, and take advantage of Metal's support for tess. eval instancing. If we could also prove that the vertex shader simply passes instance attributes through (similarly with the tess. control shader), we could do this for many more instanced draws as well. It should also be possible to cache the output from the tess. control stage; if the draw comes up again, we can then skip the vertex and tess. control stages entirely! Fixes #56 and #501.
2019-02-18 20:56:42 -06:00
// Set up tessellation parameters if needed.
if (context.options.entryPointStage == spv::ExecutionModelTessellationControl ||
context.options.entryPointStage == spv::ExecutionModelTessellationEvaluation) {
if (context.options.tessPatchKind != spv::ExecutionModeMax) {
pMSLCompiler->set_execution_mode(context.options.tessPatchKind);
}
if (context.options.numTessControlPoints != 0) {
pMSLCompiler->set_execution_mode(spv::ExecutionModeOutputVertices, context.options.numTessControlPoints);
}
}
// Establish the MSL options for the compiler
// This needs to be done in two steps...for CompilerMSL and its superclass.
pMSLCompiler->set_msl_options(context.options.mslOptions);
2017-11-17 11:14:29 -05:00
auto scOpts = pMSLCompiler->get_common_options();
scOpts.vertex.flip_vert_y = context.options.shouldFlipVertexY;
pMSLCompiler->set_common_options(scOpts);
2017-11-17 11:14:29 -05:00
// Add shader inputs
for (auto& si : context.shaderInputs) {
pMSLCompiler->add_msl_shader_input(si.shaderInput);
}
// Add resource bindings and hardcoded constexpr samplers
for (auto& rb : context.resourceBindings) {
auto& rbb = rb.resourceBinding;
pMSLCompiler->add_msl_resource_binding(rbb);
if (rb.requiresConstExprSampler) {
pMSLCompiler->remap_constexpr_sampler_by_binding(rbb.desc_set, rbb.binding, rb.constExprSampler);
}
}
_msl = pMSLCompiler->compile();
2017-11-17 11:14:29 -05:00
if (shouldLogMSL) { logSource(_msl, "MSL", "Converted"); }
#ifndef SPIRV_CROSS_EXCEPTIONS_TO_ASSERTIONS
} catch (SPIRV_CROSS_NAMESPACE::CompilerError& ex) {
2017-11-17 11:14:29 -05:00
string errMsg("MSL conversion error: ");
errMsg += ex.what();
logError(errMsg.data());
if (shouldLogMSL && pMSLCompiler) {
_msl = pMSLCompiler->get_partial_source();
2017-11-17 11:14:29 -05:00
logSource(_msl, "MSL", "Partially converted");
}
}
#endif
2017-11-17 11:14:29 -05:00
// Populate the shader conversion results with info from the compilation run,
// and mark which vertex attributes and resource bindings are used by the shader
populateEntryPoint(pMSLCompiler, context.options);
_shaderConversionResults.isRasterizationDisabled = pMSLCompiler && pMSLCompiler->get_is_rasterization_disabled();
_shaderConversionResults.needsSwizzleBuffer = pMSLCompiler && pMSLCompiler->needs_swizzle_buffer();
_shaderConversionResults.needsOutputBuffer = pMSLCompiler && pMSLCompiler->needs_output_buffer();
_shaderConversionResults.needsPatchOutputBuffer = pMSLCompiler && pMSLCompiler->needs_patch_output_buffer();
_shaderConversionResults.needsBufferSizeBuffer = pMSLCompiler && pMSLCompiler->needs_buffer_size_buffer();
_shaderConversionResults.needsInputThreadgroupMem = pMSLCompiler && pMSLCompiler->needs_input_threadgroup_mem();
_shaderConversionResults.needsDispatchBaseBuffer = pMSLCompiler && pMSLCompiler->needs_dispatch_base_buffer();
for (auto& ctxSI : context.shaderInputs) {
ctxSI.isUsedByShader = pMSLCompiler->is_msl_shader_input_used(ctxSI.shaderInput.location);
}
for (auto& ctxRB : context.resourceBindings) {
ctxRB.isUsedByShader = pMSLCompiler->is_msl_resource_binding_used(ctxRB.resourceBinding.stage,
ctxRB.resourceBinding.desc_set,
ctxRB.resourceBinding.binding);
}
2017-11-17 11:14:29 -05:00
delete pMSLCompiler;
2017-11-17 11:14:29 -05:00
// To check GLSL conversion
if (shouldLogGLSL) {
SPIRV_CROSS_NAMESPACE::CompilerGLSL* pGLSLCompiler = nullptr;
#ifndef SPIRV_CROSS_EXCEPTIONS_TO_ASSERTIONS
try {
#endif
pGLSLCompiler = new SPIRV_CROSS_NAMESPACE::CompilerGLSL(_spirv);
auto options = pGLSLCompiler->get_common_options();
options.vulkan_semantics = true;
options.separate_shader_objects = true;
pGLSLCompiler->set_common_options(options);
string glsl = pGLSLCompiler->compile();
2017-11-17 11:14:29 -05:00
logSource(glsl, "GLSL", "Estimated original");
#ifndef SPIRV_CROSS_EXCEPTIONS_TO_ASSERTIONS
} catch (SPIRV_CROSS_NAMESPACE::CompilerError& ex) {
2017-11-17 11:14:29 -05:00
string errMsg("Original GLSL extraction error: ");
errMsg += ex.what();
logMsg(errMsg.data());
if (pGLSLCompiler) {
string glsl = pGLSLCompiler->get_partial_source();
logSource(glsl, "GLSL", "Partially converted");
}
2017-11-17 11:14:29 -05:00
}
#endif
delete pGLSLCompiler;
2017-11-17 11:14:29 -05:00
}
return _wasConverted;
}
// Appends the message text to the result log.
2017-11-17 11:14:29 -05:00
void SPIRVToMSLConverter::logMsg(const char* logMsg) {
string trimMsg = trim(logMsg);
if ( !trimMsg.empty() ) {
_resultLog += trimMsg;
_resultLog += "\n\n";
}
}
// Appends the error text to the result log, sets the wasConverted property to false, and returns it.
2017-11-17 11:14:29 -05:00
bool SPIRVToMSLConverter::logError(const char* errMsg) {
logMsg(errMsg);
_wasConverted = false;
return _wasConverted;
}
// Appends the SPIR-V to the result log, indicating whether it is being converted or was converted.
2017-11-17 11:14:29 -05:00
void SPIRVToMSLConverter::logSPIRV(const char* opDesc) {
string spvLog;
mvk::logSPIRV(_spirv, spvLog);
_resultLog += opDesc;
_resultLog += " SPIR-V:\n";
_resultLog += spvLog;
_resultLog += "\nEnd SPIR-V\n\n";
// Uncomment one or both of the following lines to get additional debugging and tracability capabilities.
// The SPIR-V can be written in binary form to a file, and/or logged in human readable form to the console.
// These can be helpful if errors occur during conversion of SPIR-V to MSL.
// writeSPIRVToFile("spvout.spv");
// printf("\n%s\n", getResultLog().c_str());
}
// Writes the SPIR-V code to a file. This can be useful for debugging
// when the SPRIR-V did not originally come from a known file
void SPIRVToMSLConverter::writeSPIRVToFile(string spvFilepath) {
vector<char> fileContents;
spirvToBytes(_spirv, fileContents);
string errMsg;
if (writeFile(spvFilepath, fileContents, errMsg)) {
_resultLog += "Saved SPIR-V to file: " + absolutePath(spvFilepath) + "\n\n";
} else {
_resultLog += "Could not write SPIR-V file. " + errMsg + "\n\n";
}
2017-11-17 11:14:29 -05:00
}
// Validates that the SPIR-V code will disassemble during logging.
2017-11-17 11:14:29 -05:00
bool SPIRVToMSLConverter::validateSPIRV() {
if (_spirv.size() < 5) { return false; }
if (_spirv[0] != spv::MagicNumber) { return false; }
if (_spirv[4] != 0) { return false; }
return true;
}
// Appends the source to the result log, prepending with the operation.
2017-11-17 11:14:29 -05:00
void SPIRVToMSLConverter::logSource(string& src, const char* srcLang, const char* opDesc) {
_resultLog += opDesc;
_resultLog += " ";
_resultLog += srcLang;
_resultLog += ":\n";
_resultLog += src;
_resultLog += "\nEnd ";
_resultLog += srcLang;
_resultLog += "\n\n";
}
void SPIRVToMSLConverter::populateWorkgroupDimension(SPIRVWorkgroupSizeDimension& wgDim,
uint32_t size,
SPIRV_CROSS_NAMESPACE::SpecializationConstant& spvSpecConst) {
wgDim.size = max(size, 1u);
wgDim.isSpecialized = (uint32_t(spvSpecConst.id) != 0);
wgDim.specializationID = spvSpecConst.constant_id;
}
// Populates the entry point with info extracted from the SPRI-V compiler.
void SPIRVToMSLConverter::populateEntryPoint(SPIRV_CROSS_NAMESPACE::Compiler* pCompiler,
SPIRVToMSLConversionOptions& options) {
if ( !pCompiler ) { return; }
2017-11-17 11:14:29 -05:00
SPIRV_CROSS_NAMESPACE::SPIREntryPoint spvEP;
if (options.hasEntryPoint()) {
spvEP = pCompiler->get_entry_point(options.entryPointName, options.entryPointStage);
} else {
const auto& entryPoints = pCompiler->get_entry_points_and_stages();
if ( !entryPoints.empty() ) {
auto& ep = entryPoints[0];
spvEP = pCompiler->get_entry_point(ep.name, ep.execution_model);
}
}
auto& ep = _shaderConversionResults.entryPoint;
ep.mtlFunctionName = spvEP.name;
SPIRV_CROSS_NAMESPACE::SpecializationConstant widthSC, heightSC, depthSC;
pCompiler->get_work_group_size_specialization_constants(widthSC, heightSC, depthSC);
auto& wgSize = ep.workgroupSize;
populateWorkgroupDimension(wgSize.width, spvEP.workgroup_size.x, widthSC);
populateWorkgroupDimension(wgSize.height, spvEP.workgroup_size.y, heightSC);
populateWorkgroupDimension(wgSize.depth, spvEP.workgroup_size.z, depthSC);
2017-11-17 11:14:29 -05:00
}