sid/NazaraEngine
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Add initial support for compute pipelines

Browse Source
This commit is contained in:
SirLynix
2022-12-24 11:54:55 +01:00
committed by Jérôme Leclercq
parent e4064997d8
commit 9578ba3ef5
57 changed files with 915 additions and 182 deletions
Download Patch File Download Diff File Expand all files Collapse all files

71
src/Nazara/OpenGLRenderer/OpenGLDevice.cpp
View File

@@ -5,6 +5,7 @@
#include <Nazara/OpenGLRenderer/OpenGLDevice.hpp>
#include <Nazara/OpenGLRenderer/OpenGLBuffer.hpp>
#include <Nazara/OpenGLRenderer/OpenGLCommandPool.hpp>
#include <Nazara/OpenGLRenderer/OpenGLComputePipeline.hpp>
#include <Nazara/OpenGLRenderer/OpenGLFboFramebuffer.hpp>
#include <Nazara/OpenGLRenderer/OpenGLRenderPass.hpp>
#include <Nazara/OpenGLRenderer/OpenGLRenderPipeline.hpp>
@@ -62,47 +63,58 @@ namespace Nz
if (m_referenceContext->IsExtensionSupported(GL::Extension::TextureFilterAnisotropic))
m_deviceInfo.features.anisotropicFiltering = true;
if (m_referenceContext->IsExtensionSupported(GL::Extension::ComputeShader))
m_deviceInfo.features.computeShaders = true;
if (m_referenceContext->IsExtensionSupported(GL::Extension::DepthClamp))
m_deviceInfo.features.depthClamping = true;
if (m_referenceContext->glPolygonMode) //< not supported in core OpenGL ES, but supported in OpenGL or with GL_NV_polygon_mode extension
m_deviceInfo.features.nonSolidFaceFilling = true;
if (m_referenceContext->IsExtensionSupported(GL::Extension::StorageBuffers))
m_deviceInfo.features.storageBuffers = true;
if (m_referenceContext->glPolygonMode) //< not supported in core OpenGL ES, but supported in OpenGL or with GL_NV_polygon_mode extension
m_deviceInfo.features.nonSolidFaceFilling = true;
if (m_referenceContext->IsExtensionSupported(GL::Extension::ShaderImageLoadStore))
{
m_deviceInfo.features.textureRead = true;
m_deviceInfo.features.textureWrite = true;
m_deviceInfo.features.textureWriteWithoutFormat = true;
}
if (m_referenceContext->IsExtensionSupported(GL::Extension::ShaderImageLoadFormatted))
m_deviceInfo.features.textureReadWithoutFormat = true;
if (m_referenceContext->IsExtensionSupported(GL::Extension::TextureView))
m_deviceInfo.features.unrestrictedTextureViews = true;
// Limits
GLint minUboOffsetAlignment;
m_referenceContext->glGetIntegerv(GL_UNIFORM_BUFFER_OFFSET_ALIGNMENT, &minUboOffsetAlignment);
assert(minUboOffsetAlignment >= 1);
m_deviceInfo.limits.minUniformBufferOffsetAlignment = static_cast<UInt64>(minUboOffsetAlignment);
GLint maxUboBlockSize;
m_referenceContext->glGetIntegerv(GL_MAX_UNIFORM_BLOCK_SIZE, &maxUboBlockSize);
assert(maxUboBlockSize >= 1);
m_deviceInfo.limits.maxUniformBufferSize = static_cast<UInt64>(maxUboBlockSize);
m_deviceInfo.limits.maxUniformBufferSize = m_referenceContext->GetInteger<UInt64>(GL_MAX_UNIFORM_BLOCK_SIZE);
m_deviceInfo.limits.minUniformBufferOffsetAlignment = m_referenceContext->GetInteger<UInt64>(GL_UNIFORM_BUFFER_OFFSET_ALIGNMENT);
if (m_deviceInfo.features.storageBuffers)
{
GLint minStorageOffsetAlignment;
m_referenceContext->glGetIntegerv(GL_SHADER_STORAGE_BUFFER_OFFSET_ALIGNMENT, &minStorageOffsetAlignment);
assert(minStorageOffsetAlignment >= 1);
m_deviceInfo.limits.minStorageBufferOffsetAlignment = static_cast<UInt64>(minStorageOffsetAlignment);
GLint maxStorageBlockSize;
m_referenceContext->glGetIntegerv(GL_MAX_SHADER_STORAGE_BLOCK_SIZE, &maxStorageBlockSize);
assert(maxStorageBlockSize >= 1);
m_deviceInfo.limits.maxStorageBufferSize = static_cast<UInt64>(maxStorageBlockSize);
m_deviceInfo.limits.maxStorageBufferSize = m_referenceContext->GetInteger<UInt64>(GL_MAX_SHADER_STORAGE_BLOCK_SIZE);
m_deviceInfo.limits.minStorageBufferOffsetAlignment = m_referenceContext->GetInteger<UInt64>(GL_SHADER_STORAGE_BUFFER_OFFSET_ALIGNMENT);
}
if (m_deviceInfo.features.computeShaders)
{
m_deviceInfo.limits.maxComputeSharedMemorySize = m_referenceContext->GetInteger<UInt64>(GL_MAX_COMPUTE_SHARED_MEMORY_SIZE);
m_deviceInfo.limits.maxComputeWorkGroupInvocations = m_referenceContext->GetInteger<UInt32>(GL_MAX_COMPUTE_WORK_GROUP_INVOCATIONS);
m_deviceInfo.limits.maxComputeWorkGroupCount = {
m_referenceContext->GetInteger<UInt32>(GL_MAX_COMPUTE_WORK_GROUP_COUNT, 0),
m_referenceContext->GetInteger<UInt32>(GL_MAX_COMPUTE_WORK_GROUP_COUNT, 1),
m_referenceContext->GetInteger<UInt32>(GL_MAX_COMPUTE_WORK_GROUP_COUNT, 2)
};
m_deviceInfo.limits.maxComputeWorkGroupSize = {
m_referenceContext->GetInteger<UInt32>(GL_MAX_COMPUTE_WORK_GROUP_SIZE, 0),
m_referenceContext->GetInteger<UInt32>(GL_MAX_COMPUTE_WORK_GROUP_SIZE, 1),
m_referenceContext->GetInteger<UInt32>(GL_MAX_COMPUTE_WORK_GROUP_SIZE, 2)
};
}
else
m_deviceInfo.limits.maxStorageBufferSize = 0;
m_contexts.insert(m_referenceContext.get());
}
@@ -154,6 +166,11 @@ namespace Nz
return std::make_shared<OpenGLCommandPool>();
}
std::shared_ptr<ComputePipeline> OpenGLDevice::InstantiateComputePipeline(ComputePipelineInfo pipelineInfo)
{
return std::make_shared<OpenGLComputePipeline>(*this, std::move(pipelineInfo));
}
std::shared_ptr<Framebuffer> OpenGLDevice::InstantiateFramebuffer(unsigned int /*width*/, unsigned int /*height*/, const std::shared_ptr<RenderPass>& /*renderPass*/, const std::vector<std::shared_ptr<Texture>>& attachments)
{
return std::make_shared<OpenGLFboFramebuffer>(*this, attachments);
@@ -246,7 +263,7 @@ namespace Nz
case PixelFormat::RGBA32F:
case PixelFormat::RGBA32I:
case PixelFormat::RGBA32UI:
return usage == TextureUsage::ColorAttachment || usage == TextureUsage::InputAttachment || usage == TextureUsage::ShaderSampling || usage == TextureUsage::TransferDestination || usage == TextureUsage::TransferSource;
return usage == TextureUsage::ColorAttachment || usage == TextureUsage::InputAttachment || usage == TextureUsage::ShaderSampling || usage == TextureUsage::ShaderReadWrite || usage == TextureUsage::TransferDestination || usage == TextureUsage::TransferSource;
case PixelFormat::DXT1:
case PixelFormat::DXT3: