// Render the scene.
void D3D1211on12::OnRender()
{
PIXBeginEvent(m_commandQueue.Get(), 0, L"Render 3D");
// Record all the commands we need to render the scene into the command list.
PopulateCommandList();
// Execute the command list.
ID3D12CommandList* ppCommandLists[] = { m_commandList.Get() };
m_commandQueue->ExecuteCommandLists(_countof(ppCommandLists), ppCommandLists);
PIXEndEvent(m_commandQueue.Get());
PIXBeginEvent(m_commandQueue.Get(), 0, L"Render UI");
RenderUI();
PIXEndEvent(m_commandQueue.Get());
// Present the frame.
ThrowIfFailed(m_swapChain->Present(1, 0));
MoveToNextFrame();
}
// Fill the command list with all the render commands and dependent state.
void D3D12Fullscreen::PopulateCommandList()
{
// Command list allocators can only be reset when the associated
// command lists have finished execution on the GPU; apps should use
// fences to determine GPU execution progress.
ThrowIfFailed(m_commandAllocators[m_frameIndex]->Reset());
// However, when ExecuteCommandList() is called on a particular command
// list, that command list can then be reset at any time and must be before
// re-recording.
ThrowIfFailed(m_commandList->Reset(m_commandAllocators[m_frameIndex].Get(), m_pipelineState.Get()));
if (m_resizeResources)
{
// Reload resources that depend on the size of the swap chain.
LoadSizeDependentResources();
}
// Set necessary state.
m_commandList->SetGraphicsRootSignature(m_rootSignature.Get());
ID3D12DescriptorHeap* ppHeaps[] = { m_cbvHeap.Get() };
m_commandList->SetDescriptorHeaps(_countof(ppHeaps), ppHeaps);
CD3DX12_GPU_DESCRIPTOR_HANDLE cbvHandle(m_cbvHeap->GetGPUDescriptorHandleForHeapStart(), m_frameIndex, m_cbvDescriptorSize);
m_commandList->SetGraphicsRootDescriptorTable(0, cbvHandle);
m_commandList->IASetPrimitiveTopology(D3D_PRIMITIVE_TOPOLOGY_TRIANGLELIST);
m_commandList->RSSetViewports(1, &m_viewport);
m_commandList->RSSetScissorRects(1, &m_scissorRect);
// Indicate that the back buffer will be used as a render target.
m_commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_renderTargets[m_frameIndex].Get(), D3D12_RESOURCE_STATE_PRESENT, D3D12_RESOURCE_STATE_RENDER_TARGET));
CD3DX12_CPU_DESCRIPTOR_HANDLE rtvHandle(m_rtvHeap->GetCPUDescriptorHandleForHeapStart(), m_frameIndex, m_rtvDescriptorSize);
m_commandList->OMSetRenderTargets(1, &rtvHandle, FALSE, nullptr);
// Record commands.
const float clearColor[] = { 0.0f, 0.2f, 0.4f, 1.0f };
m_commandList->ClearRenderTargetView(rtvHandle, clearColor, 0, nullptr);
m_commandList->IASetPrimitiveTopology(D3D_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP);
m_commandList->IASetVertexBuffers(0, 1, &m_vertexBufferView);
PIXBeginEvent(m_commandList.Get(), 0, L"Draw a thin rectangle");
m_commandList->DrawInstanced(4, 1, 0, 0);
PIXEndEvent(m_commandList.Get());
// Indicate that the back buffer will now be used to present.
m_commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_renderTargets[m_frameIndex].Get(), D3D12_RESOURCE_STATE_RENDER_TARGET, D3D12_RESOURCE_STATE_PRESENT));
ThrowIfFailed(m_commandList->Close());
}
void FrameResource::PopulateCommandList(ID3D12GraphicsCommandList* pCommandList, ID3D12PipelineState* pPso1, ID3D12PipelineState* pPso2,
UINT frameResourceIndex, UINT numIndices, D3D12_INDEX_BUFFER_VIEW* pIndexBufferViewDesc, D3D12_VERTEX_BUFFER_VIEW* pVertexBufferViewDesc,
ID3D12DescriptorHeap* pCbvSrvDescriptorHeap, UINT cbvSrvDescriptorSize, ID3D12DescriptorHeap* pSamplerDescriptorHeap, ID3D12RootSignature* pRootSignature)
{
// If the root signature matches the root signature of the caller, then
// bindings are inherited, otherwise the bind space is reset.
pCommandList->SetGraphicsRootSignature(pRootSignature);
ID3D12DescriptorHeap* ppHeaps[] = { pCbvSrvDescriptorHeap, pSamplerDescriptorHeap };
pCommandList->SetDescriptorHeaps(_countof(ppHeaps), ppHeaps);
pCommandList->IASetPrimitiveTopology(D3D_PRIMITIVE_TOPOLOGY_TRIANGLELIST);
pCommandList->IASetIndexBuffer(pIndexBufferViewDesc);
pCommandList->IASetVertexBuffers(0, 1, pVertexBufferViewDesc);
pCommandList->SetGraphicsRootDescriptorTable(0, pCbvSrvDescriptorHeap->GetGPUDescriptorHandleForHeapStart());
pCommandList->SetGraphicsRootDescriptorTable(1, pSamplerDescriptorHeap->GetGPUDescriptorHandleForHeapStart());
// Calculate the descriptor offset due to multiple frame resources.
// 1 SRV + how many CBVs we have currently.
UINT frameResourceDescriptorOffset = 1 + (frameResourceIndex * m_cityRowCount * m_cityColumnCount);
CD3DX12_GPU_DESCRIPTOR_HANDLE cbvSrvHandle(pCbvSrvDescriptorHeap->GetGPUDescriptorHandleForHeapStart(), frameResourceDescriptorOffset, cbvSrvDescriptorSize);
PIXBeginEvent(pCommandList, 0, L"Draw cities");
BOOL usePso1 = TRUE;
for (UINT i = 0; i < m_cityRowCount; i++)
{
for (UINT j = 0; j < m_cityColumnCount; j++)
{
// Alternate which PSO to use; the pixel shader is different on
// each just as a PSO setting demonstration.
pCommandList->SetPipelineState(usePso1 ? pPso1 : pPso2);
usePso1 = !usePso1;
// Set this city's CBV table and move to the next descriptor.
pCommandList->SetGraphicsRootDescriptorTable(2, cbvSrvHandle);
cbvSrvHandle.Offset(cbvSrvDescriptorSize);
pCommandList->DrawIndexedInstanced(numIndices, 1, 0, 0, 0);
}
}
PIXEndEvent(pCommandList);
}
// Render the scene.
void D3D12PipelineStateCache::OnRender()
{
PIXBeginEvent(m_commandQueue.Get(), 0, L"Render");
// Record all the commands we need to render the scene into the command list.
PopulateCommandList();
// Execute the command list.
ID3D12CommandList* ppCommandLists[] = { m_commandList.Get() };
m_commandQueue->ExecuteCommandLists(_countof(ppCommandLists), ppCommandLists);
PIXEndEvent(m_commandQueue.Get());
// Present the frame.
ThrowIfFailed(m_swapChain->Present(1, 0));
m_drawIndex = 0;
m_psoLibrary.EndFrame();
MoveToNextFrame();
}
// Render the scene.
void D3D12Fullscreen::OnRender()
{
if (m_windowVisible)
{
PIXBeginEvent(m_commandQueue.Get(), 0, L"Render");
// Record all the commands we need to render the scene into the command list.
PopulateCommandList();
// Execute the command list.
ID3D12CommandList* ppCommandLists[] = { m_commandList.Get() };
m_commandQueue->ExecuteCommandLists(_countof(ppCommandLists), ppCommandLists);
PIXEndEvent(m_commandQueue.Get());
// Present the frame.
ThrowIfFailed(m_swapChain->Present(0, 0));
MoveToNextFrame();
}
}
void FrameResource::PopulateCommandList(ID3D12GraphicsCommandList* pCommandList, ID3D12PipelineState* pPso,
UINT frameResourceIndex, UINT numIndices, D3D12_INDEX_BUFFER_VIEW* pIndexBufferViewDesc, D3D12_VERTEX_BUFFER_VIEW* pVertexBufferViewDesc,
ID3D12DescriptorHeap* pCbvSrvDescriptorHeap, UINT cbvSrvDescriptorSize, ID3D12DescriptorHeap* pSamplerDescriptorHeap, ID3D12RootSignature* pRootSignature)
{
// If the root signature matches the root signature of the caller, then
// bindings are inherited, otherwise the bind space is reset.
pCommandList->SetGraphicsRootSignature(pRootSignature);
ID3D12DescriptorHeap* ppHeaps[] = { pCbvSrvDescriptorHeap, pSamplerDescriptorHeap };
pCommandList->SetDescriptorHeaps(_countof(ppHeaps), ppHeaps);
pCommandList->IASetPrimitiveTopology(D3D_PRIMITIVE_TOPOLOGY_TRIANGLELIST);
pCommandList->IASetIndexBuffer(pIndexBufferViewDesc);
pCommandList->IASetVertexBuffers(0, 1, pVertexBufferViewDesc);
pCommandList->SetGraphicsRootDescriptorTable(0, pCbvSrvDescriptorHeap->GetGPUDescriptorHandleForHeapStart());
pCommandList->SetGraphicsRootDescriptorTable(1, pSamplerDescriptorHeap->GetGPUDescriptorHandleForHeapStart());
// Calculate the descriptor offset due to multiple frame resources.
// (m_cityMaterialCount + 1) SRVs + how many CBVs we have currently.
UINT frameResourceDescriptorOffset = (m_cityMaterialCount + 1) + (frameResourceIndex * m_cityRowCount * m_cityColumnCount);
CD3DX12_GPU_DESCRIPTOR_HANDLE cbvSrvHandle(pCbvSrvDescriptorHeap->GetGPUDescriptorHandleForHeapStart(), frameResourceDescriptorOffset, cbvSrvDescriptorSize);
PIXBeginEvent(pCommandList, 0, L"Draw cities");
for (UINT i = 0; i < m_cityRowCount; i++)
{
for (UINT j = 0; j < m_cityColumnCount; j++)
{
pCommandList->SetPipelineState(pPso);
// Set the city's root constant for dynamically indexing into the material array.
pCommandList->SetGraphicsRoot32BitConstant(3, (i * m_cityColumnCount) + j, 0);
// Set this city's CBV table and move to the next descriptor.
pCommandList->SetGraphicsRootDescriptorTable(2, cbvSrvHandle);
cbvSrvHandle.Offset(cbvSrvDescriptorSize);
pCommandList->DrawIndexedInstanced(numIndices, 1, 0, 0, 0);
}
}
PIXEndEvent(pCommandList);
}
// Render the scene.
void D3D12DynamicIndexing::OnRender()
{
PIXBeginEvent(m_commandQueue.Get(), 0, L"Render");
// Record all the commands we need to render the scene into the command list.
PopulateCommandList(m_pCurrentFrameResource);
// Execute the command list.
ID3D12CommandList* ppCommandLists[] = { m_commandList.Get() };
m_commandQueue->ExecuteCommandLists(_countof(ppCommandLists), ppCommandLists);
PIXEndEvent(m_commandQueue.Get());
// Present and update the frame index for the next frame.
ThrowIfFailed(m_swapChain->Present(1, 0));
m_frameIndex = m_swapChain->GetCurrentBackBufferIndex();
// Signal and increment the fence value.
m_pCurrentFrameResource->m_fenceValue = m_fenceValue;
ThrowIfFailed(m_commandQueue->Signal(m_fence.Get(), m_fenceValue));
m_fenceValue++;
}
// Fill the command list with all the render commands and dependent state and
// submit it to the command queue.
void D3D12HDR::RenderScene()
{
// Command list allocators can only be reset when the associated
// command lists have finished execution on the GPU; apps should use
// fences to determine GPU execution progress.
ThrowIfFailed(m_commandAllocators[m_frameIndex]->Reset());
// However, when ExecuteCommandList() is called on a particular command
// list, that command list can then be reset at any time and must be before
// re-recording.
ThrowIfFailed(m_commandList->Reset(m_commandAllocators[m_frameIndex].Get(), m_pipelineStates[GradientPSO].Get()));
if (m_updateVertexBuffer)
{
UpdateVertexBuffer();
m_updateVertexBuffer = false;
}
// Set necessary state.
m_commandList->SetGraphicsRootSignature(m_rootSignature.Get());
ID3D12DescriptorHeap* ppHeaps[] = { m_srvHeap.Get() };
m_commandList->SetDescriptorHeaps(_countof(ppHeaps), ppHeaps);
m_commandList->IASetPrimitiveTopology(D3D_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP);
m_commandList->RSSetViewports(1, &m_viewport);
m_commandList->RSSetScissorRects(1, &m_scissorRect);
// Bind the root constants and the SRV table to the pipeline.
m_rootConstantsF[ReferenceWhiteNits] = m_referenceWhiteNits;
m_commandList->SetGraphicsRoot32BitConstants(0, RootConstantsCount, m_rootConstants, 0);
m_commandList->SetGraphicsRootDescriptorTable(1, m_srvHeap->GetGPUDescriptorHandleForHeapStart());
// Draw the scene into the intermediate render target.
{
PIXBeginEvent(m_commandList.Get(), 0, L"Draw scene content");
CD3DX12_CPU_DESCRIPTOR_HANDLE intermediateRtv(m_rtvHeap->GetCPUDescriptorHandleForHeapStart(), FrameCount, m_rtvDescriptorSize);
m_commandList->OMSetRenderTargets(1, &intermediateRtv, FALSE, nullptr);
const float clearColor[] = { 0.0f, 0.0f, 0.0f, 0.0f };
m_commandList->ClearRenderTargetView(intermediateRtv, clearColor, 0, nullptr);
m_commandList->IASetVertexBuffers(0, 1, &m_gradientVertexBufferView);
PIXBeginEvent(m_commandList.Get(), 0, L"Standard Gradient");
m_commandList->DrawInstanced(4, 1, 0, 0);
PIXEndEvent(m_commandList.Get());
PIXBeginEvent(m_commandList.Get(), 0, L"Bright Gradient");
m_commandList->DrawInstanced(4, 1, 4, 0);
PIXEndEvent(m_commandList.Get());
m_commandList->SetPipelineState(m_pipelineStates[PalettePSO].Get());
m_commandList->IASetVertexBuffers(0, 1, &m_trianglesVertexBufferView);
m_commandList->IASetPrimitiveTopology(D3D_PRIMITIVE_TOPOLOGY_TRIANGLELIST);
PIXBeginEvent(m_commandList.Get(), 0, L"Rec709 Triangles");
m_commandList->DrawInstanced(3, 1, 0, 0);
m_commandList->DrawInstanced(3, 1, 3, 0);
m_commandList->DrawInstanced(3, 1, 6, 0);
PIXEndEvent(m_commandList.Get());
m_commandList->SetPipelineState(m_pipelineStates[PalettePSO].Get());
PIXBeginEvent(m_commandList.Get(), 0, L"Rec2020 Triangles");
m_commandList->DrawInstanced(3, 1, 9, 0);
m_commandList->DrawInstanced(3, 1, 12, 0);
m_commandList->DrawInstanced(3, 1, 15, 0);
PIXEndEvent(m_commandList.Get());
PIXEndEvent(m_commandList.Get());
if (!m_enableUI)
{
intermediateRtv.Offset(1, m_rtvDescriptorSize);
m_commandList->OMSetRenderTargets(1, &intermediateRtv, FALSE, nullptr);
m_commandList->ClearRenderTargetView(intermediateRtv, clearColor, 0, nullptr);
}
}
// Indicate that the intermediates will be used as SRVs in the pixel shader
// and the back buffer will be used as a render target.
D3D12_RESOURCE_BARRIER barriers[] = {
CD3DX12_RESOURCE_BARRIER::Transition(m_intermediateRenderTarget.Get(), D3D12_RESOURCE_STATE_RENDER_TARGET, D3D12_RESOURCE_STATE_PIXEL_SHADER_RESOURCE),
CD3DX12_RESOURCE_BARRIER::Transition(m_renderTargets[m_frameIndex].Get(), D3D12_RESOURCE_STATE_PRESENT, D3D12_RESOURCE_STATE_RENDER_TARGET),
CD3DX12_RESOURCE_BARRIER::Transition(m_UIRenderTarget.Get(), D3D12_RESOURCE_STATE_RENDER_TARGET, D3D12_RESOURCE_STATE_PIXEL_SHADER_RESOURCE),
};
// Process the intermediate and draw into the swap chain render target.
{
PIXBeginEvent(m_commandList.Get(), 0, L"Apply HDR");
// If the UI is enabled, the 11on12 layer will do the state transition for us.
UINT barrierCount = m_enableUI ? 2 : _countof(barriers);
m_commandList->ResourceBarrier(barrierCount, barriers);
m_commandList->SetPipelineState(m_pipelineStates[Present8bitPSO + m_currentSwapChainBitDepth].Get());
CD3DX12_CPU_DESCRIPTOR_HANDLE rtvHandle(m_rtvHeap->GetCPUDescriptorHandleForHeapStart(), m_frameIndex, m_rtvDescriptorSize);
m_commandList->OMSetRenderTargets(1, &rtvHandle, FALSE, nullptr);
m_commandList->ClearRenderTargetView(rtvHandle, ClearColor, 0, nullptr);
m_commandList->IASetVertexBuffers(0, 1, &m_presentVertexBufferView);
m_commandList->DrawInstanced(3, 1, 0, 0);
PIXEndEvent(m_commandList.Get());
}
// Indicate that the intermediates will be used as render targets and the swap chain
// back buffer will be used for presentation.
barriers[0].Transition.StateBefore = D3D12_RESOURCE_STATE_PIXEL_SHADER_RESOURCE;
barriers[0].Transition.StateAfter = D3D12_RESOURCE_STATE_RENDER_TARGET;
barriers[1].Transition.StateBefore = D3D12_RESOURCE_STATE_RENDER_TARGET;
barriers[1].Transition.StateAfter = D3D12_RESOURCE_STATE_PRESENT;
barriers[2].Transition.StateBefore = D3D12_RESOURCE_STATE_PIXEL_SHADER_RESOURCE;
barriers[2].Transition.StateAfter = D3D12_RESOURCE_STATE_RENDER_TARGET;
m_commandList->ResourceBarrier(_countof(barriers), barriers);
ThrowIfFailed(m_commandList->Close());
// Execute the command list.
ID3D12CommandList* ppCommandLists[] = { m_commandList.Get() };
m_commandQueue->ExecuteCommandLists(_countof(ppCommandLists), ppCommandLists);
}
// Worker thread body. workerIndex is an integer from 0 to NumContexts
// describing the worker's thread index.
void D3D12Multithreading::WorkerThread(LPVOID workerIndex)
{
int threadIndex = reinterpret_cast<int>(workerIndex);
assert(threadIndex >= 0);
assert(threadIndex < NumContexts);
#if !SINGLETHREADED
while (threadIndex >= 0 && threadIndex < NumContexts)
{
// Wait for main thread to tell us to draw.
WaitForSingleObject(m_workerBeginRenderFrame[threadIndex], INFINITE);
#endif
ID3D12GraphicsCommandList* pShadowCommandList = m_pCurrentFrameResource->m_shadowCommandLists[threadIndex].Get();
ID3D12GraphicsCommandList* pSceneCommandList = m_pCurrentFrameResource->m_sceneCommandLists[threadIndex].Get();
//
// Shadow pass
//
// Populate the command list.
SetCommonPipelineState(pShadowCommandList);
m_pCurrentFrameResource->Bind(pShadowCommandList, FALSE, nullptr, nullptr); // No need to pass RTV or DSV descriptor heap.
// Set null SRVs for the diffuse/normal textures.
pShadowCommandList->SetGraphicsRootDescriptorTable(0, m_cbvSrvHeap->GetGPUDescriptorHandleForHeapStart());
// Distribute objects over threads by drawing only 1/NumContexts
// objects per worker (i.e. every object such that objectnum %
// NumContexts == threadIndex).
PIXBeginEvent(pShadowCommandList, 0, L"Worker drawing shadow pass...");
for (int j = threadIndex; j < _countof(SampleAssets::Draws); j += NumContexts)
{
SampleAssets::DrawParameters drawArgs = SampleAssets::Draws[j];
pShadowCommandList->DrawIndexedInstanced(drawArgs.IndexCount, 1, drawArgs.IndexStart, drawArgs.VertexBase, 0);
}
PIXEndEvent(pShadowCommandList);
ThrowIfFailed(pShadowCommandList->Close());
#if !SINGLETHREADED
// Submit shadow pass.
SetEvent(m_workerFinishShadowPass[threadIndex]);
#endif
//
// Scene pass
//
// Populate the command list. These can only be sent after the shadow
// passes for this frame have been submitted.
SetCommonPipelineState(pSceneCommandList);
CD3DX12_CPU_DESCRIPTOR_HANDLE rtvHandle(m_rtvHeap->GetCPUDescriptorHandleForHeapStart(), m_frameIndex, m_rtvDescriptorSize);
CD3DX12_CPU_DESCRIPTOR_HANDLE dsvHandle(m_dsvHeap->GetCPUDescriptorHandleForHeapStart());
m_pCurrentFrameResource->Bind(pSceneCommandList, TRUE, &rtvHandle, &dsvHandle);
PIXBeginEvent(pSceneCommandList, 0, L"Worker drawing scene pass...");
D3D12_GPU_DESCRIPTOR_HANDLE cbvSrvHeapStart = m_cbvSrvHeap->GetGPUDescriptorHandleForHeapStart();
const UINT cbvSrvDescriptorSize = m_device->GetDescriptorHandleIncrementSize(D3D12_DESCRIPTOR_HEAP_TYPE_CBV_SRV_UAV);
const UINT nullSrvCount = 2;
for (int j = threadIndex; j < _countof(SampleAssets::Draws); j += NumContexts)
{
SampleAssets::DrawParameters drawArgs = SampleAssets::Draws[j];
// Set the diffuse and normal textures for the current object.
CD3DX12_GPU_DESCRIPTOR_HANDLE cbvSrvHandle(cbvSrvHeapStart, nullSrvCount + drawArgs.DiffuseTextureIndex, cbvSrvDescriptorSize);
pSceneCommandList->SetGraphicsRootDescriptorTable(0, cbvSrvHandle);
pSceneCommandList->DrawIndexedInstanced(drawArgs.IndexCount, 1, drawArgs.IndexStart, drawArgs.VertexBase, 0);
}
PIXEndEvent(pSceneCommandList);
ThrowIfFailed(pSceneCommandList->Close());
#if !SINGLETHREADED
// Tell main thread that we are done.
SetEvent(m_workerFinishedRenderFrame[threadIndex]);
}
#endif
}
// Render the scene.
void D3D12Multithreading::OnRender()
{
BeginFrame();
#if SINGLETHREADED
for (int i = 0; i < NumContexts; i++)
{
WorkerThread(reinterpret_cast<LPVOID>(i));
}
MidFrame();
EndFrame();
m_commandQueue->ExecuteCommandLists(_countof(m_pCurrentFrameResource->m_batchSubmit), m_pCurrentFrameResource->m_batchSubmit);
#else
for (int i = 0; i < NumContexts; i++)
{
SetEvent(m_workerBeginRenderFrame[i]); // Tell each worker to start drawing.
}
MidFrame();
EndFrame();
WaitForMultipleObjects(NumContexts, m_workerFinishShadowPass, TRUE, INFINITE);
// You can execute command lists on any thread. Depending on the work
// load, apps can choose between using ExecuteCommandLists on one thread
// vs ExecuteCommandList from multiple threads.
m_commandQueue->ExecuteCommandLists(NumContexts + 2, m_pCurrentFrameResource->m_batchSubmit); // Submit PRE, MID and shadows.
WaitForMultipleObjects(NumContexts, m_workerFinishedRenderFrame, TRUE, INFINITE);
// Submit remaining command lists.
m_commandQueue->ExecuteCommandLists(_countof(m_pCurrentFrameResource->m_batchSubmit) - NumContexts - 2, m_pCurrentFrameResource->m_batchSubmit + NumContexts + 2);
#endif
m_cpuTimer.Tick(NULL);
if (m_titleCount == TitleThrottle)
{
WCHAR cpu[64];
swprintf_s(cpu, L"%.4f CPU", m_cpuTime / m_titleCount);
SetCustomWindowText(cpu);
m_titleCount = 0;
m_cpuTime = 0;
}
else
{
m_titleCount++;
m_cpuTime += m_cpuTimer.GetElapsedSeconds() * 1000;
m_cpuTimer.ResetElapsedTime();
}
// Present and update the frame index for the next frame.
PIXBeginEvent(m_commandQueue.Get(), 0, L"Presenting to screen");
ThrowIfFailed(m_swapChain->Present(0, 0));
PIXEndEvent(m_commandQueue.Get());
m_frameIndex = m_swapChain->GetCurrentBackBufferIndex();
// Signal and increment the fence value.
m_pCurrentFrameResource->m_fenceValue = m_fenceValue;
ThrowIfFailed(m_commandQueue->Signal(m_fence.Get(), m_fenceValue));
m_fenceValue++;
}
// Load the sample assets.
void D3D12Multithreading::LoadAssets()
{
// Create the root signature.
{
CD3DX12_DESCRIPTOR_RANGE ranges[4]; // Perfomance TIP: Order from most frequent to least frequent.
ranges[0].Init(D3D12_DESCRIPTOR_RANGE_TYPE_SRV, 2, 1); // 2 frequently changed diffuse + normal textures - using registers t1 and t2.
ranges[1].Init(D3D12_DESCRIPTOR_RANGE_TYPE_CBV, 1, 0); // 1 frequently changed constant buffer.
ranges[2].Init(D3D12_DESCRIPTOR_RANGE_TYPE_SRV, 1, 0); // 1 infrequently changed shadow texture - starting in register t0.
ranges[3].Init(D3D12_DESCRIPTOR_RANGE_TYPE_SAMPLER, 2, 0); // 2 static samplers.
CD3DX12_ROOT_PARAMETER rootParameters[4];
rootParameters[0].InitAsDescriptorTable(1, &ranges[0], D3D12_SHADER_VISIBILITY_PIXEL);
rootParameters[1].InitAsDescriptorTable(1, &ranges[1], D3D12_SHADER_VISIBILITY_ALL);
rootParameters[2].InitAsDescriptorTable(1, &ranges[2], D3D12_SHADER_VISIBILITY_PIXEL);
rootParameters[3].InitAsDescriptorTable(1, &ranges[3], D3D12_SHADER_VISIBILITY_PIXEL);
CD3DX12_ROOT_SIGNATURE_DESC rootSignatureDesc;
rootSignatureDesc.Init(_countof(rootParameters), rootParameters, 0, nullptr, D3D12_ROOT_SIGNATURE_FLAG_ALLOW_INPUT_ASSEMBLER_INPUT_LAYOUT);
ComPtr<ID3DBlob> signature;
ComPtr<ID3DBlob> error;
ThrowIfFailed(D3D12SerializeRootSignature(&rootSignatureDesc, D3D_ROOT_SIGNATURE_VERSION_1, &signature, &error));
ThrowIfFailed(m_device->CreateRootSignature(0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS(&m_rootSignature)));
}
// Create the pipeline state, which includes loading shaders.
{
ComPtr<ID3DBlob> vertexShader;
ComPtr<ID3DBlob> pixelShader;
#ifdef _DEBUG
// Enable better shader debugging with the graphics debugging tools.
UINT compileFlags = D3DCOMPILE_DEBUG | D3DCOMPILE_SKIP_OPTIMIZATION;
#else
UINT compileFlags = D3DCOMPILE_OPTIMIZATION_LEVEL3;
#endif
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"shaders.hlsl").c_str(), nullptr, nullptr, "VSMain", "vs_5_0", compileFlags, 0, &vertexShader, nullptr));
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"shaders.hlsl").c_str(), nullptr, nullptr, "PSMain", "ps_5_0", compileFlags, 0, &pixelShader, nullptr));
D3D12_INPUT_LAYOUT_DESC inputLayoutDesc;
inputLayoutDesc.pInputElementDescs = SampleAssets::StandardVertexDescription;
inputLayoutDesc.NumElements = _countof(SampleAssets::StandardVertexDescription);
CD3DX12_DEPTH_STENCIL_DESC depthStencilDesc(D3D12_DEFAULT);
depthStencilDesc.DepthEnable = true;
depthStencilDesc.DepthWriteMask = D3D12_DEPTH_WRITE_MASK_ALL;
depthStencilDesc.DepthFunc = D3D12_COMPARISON_FUNC_LESS_EQUAL;
depthStencilDesc.StencilEnable = FALSE;
// Describe and create the PSO for rendering the scene.
D3D12_GRAPHICS_PIPELINE_STATE_DESC psoDesc = {};
psoDesc.InputLayout = inputLayoutDesc;
psoDesc.pRootSignature = m_rootSignature.Get();
psoDesc.VS = { reinterpret_cast<UINT8*>(vertexShader->GetBufferPointer()), vertexShader->GetBufferSize() };
psoDesc.PS = { reinterpret_cast<UINT8*>(pixelShader->GetBufferPointer()), pixelShader->GetBufferSize() };
psoDesc.RasterizerState = CD3DX12_RASTERIZER_DESC(D3D12_DEFAULT);
psoDesc.BlendState = CD3DX12_BLEND_DESC(D3D12_DEFAULT);
psoDesc.DepthStencilState = depthStencilDesc;
psoDesc.SampleMask = UINT_MAX;
psoDesc.PrimitiveTopologyType = D3D12_PRIMITIVE_TOPOLOGY_TYPE_TRIANGLE;
psoDesc.NumRenderTargets = 1;
psoDesc.RTVFormats[0] = DXGI_FORMAT_R8G8B8A8_UNORM;
psoDesc.DSVFormat = DXGI_FORMAT_D32_FLOAT;
psoDesc.SampleDesc.Count = 1;
ThrowIfFailed(m_device->CreateGraphicsPipelineState(&psoDesc, IID_PPV_ARGS(&m_pipelineState)));
// Alter the description and create the PSO for rendering
// the shadow map. The shadow map does not use a pixel
// shader or render targets.
psoDesc.PS.pShaderBytecode = 0;
psoDesc.PS.BytecodeLength = 0;
psoDesc.RTVFormats[0] = DXGI_FORMAT_UNKNOWN;
psoDesc.NumRenderTargets = 0;
ThrowIfFailed(m_device->CreateGraphicsPipelineState(&psoDesc, IID_PPV_ARGS(&m_pipelineStateShadowMap)));
}
// Create temporary command list for initial GPU setup.
ComPtr<ID3D12GraphicsCommandList> commandList;
ThrowIfFailed(m_device->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_commandAllocator.Get(), m_pipelineState.Get(), IID_PPV_ARGS(&commandList)));
// Create render target views (RTVs).
CD3DX12_CPU_DESCRIPTOR_HANDLE rtvHandle(m_rtvHeap->GetCPUDescriptorHandleForHeapStart());
for (UINT i = 0; i < FrameCount; i++)
{
ThrowIfFailed(m_swapChain->GetBuffer(i, IID_PPV_ARGS(&m_renderTargets[i])));
m_device->CreateRenderTargetView(m_renderTargets[i].Get(), nullptr, rtvHandle);
rtvHandle.Offset(1, m_rtvDescriptorSize);
}
// Create the depth stencil.
{
CD3DX12_RESOURCE_DESC shadowTextureDesc(
D3D12_RESOURCE_DIMENSION_TEXTURE2D,
0,
static_cast<UINT>(m_viewport.Width),
static_cast<UINT>(m_viewport.Height),
1,
1,
DXGI_FORMAT_D32_FLOAT,
1,
0,
D3D12_TEXTURE_LAYOUT_UNKNOWN,
D3D12_RESOURCE_FLAG_ALLOW_DEPTH_STENCIL | D3D12_RESOURCE_FLAG_DENY_SHADER_RESOURCE);
D3D12_CLEAR_VALUE clearValue; // Performance tip: Tell the runtime at resource creation the desired clear value.
clearValue.Format = DXGI_FORMAT_D32_FLOAT;
clearValue.DepthStencil.Depth = 1.0f;
clearValue.DepthStencil.Stencil = 0;
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&shadowTextureDesc,
D3D12_RESOURCE_STATE_DEPTH_WRITE,
&clearValue,
IID_PPV_ARGS(&m_depthStencil)));
// Create the depth stencil view.
m_device->CreateDepthStencilView(m_depthStencil.Get(), nullptr, m_dsvHeap->GetCPUDescriptorHandleForHeapStart());
}
// Load scene assets.
UINT fileSize = 0;
UINT8* pAssetData;
ThrowIfFailed(ReadDataFromFile(GetAssetFullPath(SampleAssets::DataFileName).c_str(), &pAssetData, &fileSize));
// Create the vertex buffer.
{
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(SampleAssets::VertexDataSize),
D3D12_RESOURCE_STATE_COPY_DEST,
nullptr,
IID_PPV_ARGS(&m_vertexBuffer)));
{
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(SampleAssets::VertexDataSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_vertexBufferUpload)));
// Copy data to the upload heap and then schedule a copy
// from the upload heap to the vertex buffer.
D3D12_SUBRESOURCE_DATA vertexData = {};
vertexData.pData = pAssetData + SampleAssets::VertexDataOffset;
vertexData.RowPitch = SampleAssets::VertexDataSize;
vertexData.SlicePitch = vertexData.RowPitch;
PIXBeginEvent(commandList.Get(), 0, L"Copy vertex buffer data to default resource...");
UpdateSubresources<1>(commandList.Get(), m_vertexBuffer.Get(), m_vertexBufferUpload.Get(), 0, 0, 1, &vertexData);
commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_vertexBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_VERTEX_AND_CONSTANT_BUFFER));
PIXEndEvent(commandList.Get());
}
// Initialize the vertex buffer view.
m_vertexBufferView.BufferLocation = m_vertexBuffer->GetGPUVirtualAddress();
m_vertexBufferView.SizeInBytes = SampleAssets::VertexDataSize;
m_vertexBufferView.StrideInBytes = SampleAssets::StandardVertexStride;
}
// Create the index buffer.
{
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(SampleAssets::IndexDataSize),
D3D12_RESOURCE_STATE_COPY_DEST,
nullptr,
IID_PPV_ARGS(&m_indexBuffer)));
{
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(SampleAssets::IndexDataSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_indexBufferUpload)));
// Copy data to the upload heap and then schedule a copy
// from the upload heap to the index buffer.
D3D12_SUBRESOURCE_DATA indexData = {};
indexData.pData = pAssetData + SampleAssets::IndexDataOffset;
indexData.RowPitch = SampleAssets::IndexDataSize;
indexData.SlicePitch = indexData.RowPitch;
PIXBeginEvent(commandList.Get(), 0, L"Copy index buffer data to default resource...");
UpdateSubresources<1>(commandList.Get(), m_indexBuffer.Get(), m_indexBufferUpload.Get(), 0, 0, 1, &indexData);
commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_indexBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_INDEX_BUFFER));
PIXEndEvent(commandList.Get());
}
// Initialize the index buffer view.
m_indexBufferView.BufferLocation = m_indexBuffer->GetGPUVirtualAddress();
m_indexBufferView.SizeInBytes = SampleAssets::IndexDataSize;
m_indexBufferView.Format = SampleAssets::StandardIndexFormat;
}
// Create shader resources.
{
// Get the CBV SRV descriptor size for the current device.
const UINT cbvSrvDescriptorSize = m_device->GetDescriptorHandleIncrementSize(D3D12_DESCRIPTOR_HEAP_TYPE_CBV_SRV_UAV);
// Get a handle to the start of the descriptor heap.
CD3DX12_CPU_DESCRIPTOR_HANDLE cbvSrvHandle(m_cbvSrvHeap->GetCPUDescriptorHandleForHeapStart());
{
// Describe and create 2 null SRVs. Null descriptors are needed in order
// to achieve the effect of an "unbound" resource.
D3D12_SHADER_RESOURCE_VIEW_DESC nullSrvDesc = {};
nullSrvDesc.ViewDimension = D3D12_SRV_DIMENSION_TEXTURE2D;
nullSrvDesc.Shader4ComponentMapping = D3D12_DEFAULT_SHADER_4_COMPONENT_MAPPING;
nullSrvDesc.Format = DXGI_FORMAT_R8G8B8A8_UNORM;
nullSrvDesc.Texture2D.MipLevels = 1;
nullSrvDesc.Texture2D.MostDetailedMip = 0;
nullSrvDesc.Texture2D.ResourceMinLODClamp = 0.0f;
m_device->CreateShaderResourceView(nullptr, &nullSrvDesc, cbvSrvHandle);
cbvSrvHandle.Offset(cbvSrvDescriptorSize);
m_device->CreateShaderResourceView(nullptr, &nullSrvDesc, cbvSrvHandle);
cbvSrvHandle.Offset(cbvSrvDescriptorSize);
}
// Create each texture and SRV descriptor.
const UINT srvCount = _countof(SampleAssets::Textures);
PIXBeginEvent(commandList.Get(), 0, L"Copy diffuse and normal texture data to default resources...");
for (int i = 0; i < srvCount; i++)
{
// Describe and create a Texture2D.
const SampleAssets::TextureResource &tex = SampleAssets::Textures[i];
CD3DX12_RESOURCE_DESC texDesc(
D3D12_RESOURCE_DIMENSION_TEXTURE2D,
0,
tex.Width,
tex.Height,
1,
static_cast<UINT16>(tex.MipLevels),
tex.Format,
1,
0,
D3D12_TEXTURE_LAYOUT_UNKNOWN,
D3D12_RESOURCE_FLAG_NONE);
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&texDesc,
D3D12_RESOURCE_STATE_COPY_DEST,
nullptr,
IID_PPV_ARGS(&m_textures[i])));
{
const UINT subresourceCount = texDesc.DepthOrArraySize * texDesc.MipLevels;
UINT64 uploadBufferSize = GetRequiredIntermediateSize(m_textures[i].Get(), 0, subresourceCount);
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(uploadBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_textureUploads[i])));
// Copy data to the intermediate upload heap and then schedule a copy
// from the upload heap to the Texture2D.
D3D12_SUBRESOURCE_DATA textureData = {};
textureData.pData = pAssetData + tex.Data->Offset;
textureData.RowPitch = tex.Data->Pitch;
textureData.SlicePitch = tex.Data->Size;
UpdateSubresources(commandList.Get(), m_textures[i].Get(), m_textureUploads[i].Get(), 0, 0, subresourceCount, &textureData);
commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_textures[i].Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_PIXEL_SHADER_RESOURCE));
}
// Describe and create an SRV.
D3D12_SHADER_RESOURCE_VIEW_DESC srvDesc = {};
srvDesc.ViewDimension = D3D12_SRV_DIMENSION_TEXTURE2D;
srvDesc.Shader4ComponentMapping = D3D12_DEFAULT_SHADER_4_COMPONENT_MAPPING;
srvDesc.Format = tex.Format;
srvDesc.Texture2D.MipLevels = tex.MipLevels;
srvDesc.Texture2D.MostDetailedMip = 0;
srvDesc.Texture2D.ResourceMinLODClamp = 0.0f;
m_device->CreateShaderResourceView(m_textures[i].Get(), &srvDesc, cbvSrvHandle);
// Move to the next descriptor slot.
cbvSrvHandle.Offset(cbvSrvDescriptorSize);
}
PIXEndEvent(commandList.Get());
}
free(pAssetData);
// Create the samplers.
{
// Get the sampler descriptor size for the current device.
const UINT samplerDescriptorSize = m_device->GetDescriptorHandleIncrementSize(D3D12_DESCRIPTOR_HEAP_TYPE_SAMPLER);
// Get a handle to the start of the descriptor heap.
CD3DX12_CPU_DESCRIPTOR_HANDLE samplerHandle(m_samplerHeap->GetCPUDescriptorHandleForHeapStart());
// Describe and create the wrapping sampler, which is used for
// sampling diffuse/normal maps.
D3D12_SAMPLER_DESC wrapSamplerDesc = {};
wrapSamplerDesc.Filter = D3D12_FILTER_MIN_MAG_MIP_LINEAR;
wrapSamplerDesc.AddressU = D3D12_TEXTURE_ADDRESS_MODE_WRAP;
wrapSamplerDesc.AddressV = D3D12_TEXTURE_ADDRESS_MODE_WRAP;
wrapSamplerDesc.AddressW = D3D12_TEXTURE_ADDRESS_MODE_WRAP;
wrapSamplerDesc.MinLOD = 0;
wrapSamplerDesc.MaxLOD = D3D12_FLOAT32_MAX;
wrapSamplerDesc.MipLODBias = 0.0f;
wrapSamplerDesc.MaxAnisotropy = 1;
wrapSamplerDesc.ComparisonFunc = D3D12_COMPARISON_FUNC_ALWAYS;
wrapSamplerDesc.BorderColor[0] = wrapSamplerDesc.BorderColor[1] = wrapSamplerDesc.BorderColor[2] = wrapSamplerDesc.BorderColor[3] = 0;
m_device->CreateSampler(&wrapSamplerDesc, samplerHandle);
// Move the handle to the next slot in the descriptor heap.
samplerHandle.Offset(samplerDescriptorSize);
// Describe and create the point clamping sampler, which is
// used for the shadow map.
D3D12_SAMPLER_DESC clampSamplerDesc = {};
clampSamplerDesc.Filter = D3D12_FILTER_MIN_MAG_MIP_POINT;
clampSamplerDesc.AddressU = D3D12_TEXTURE_ADDRESS_MODE_CLAMP;
clampSamplerDesc.AddressV = D3D12_TEXTURE_ADDRESS_MODE_CLAMP;
clampSamplerDesc.AddressW = D3D12_TEXTURE_ADDRESS_MODE_CLAMP;
clampSamplerDesc.MipLODBias = 0.0f;
clampSamplerDesc.MaxAnisotropy = 1;
clampSamplerDesc.ComparisonFunc = D3D12_COMPARISON_FUNC_ALWAYS;
clampSamplerDesc.BorderColor[0] = clampSamplerDesc.BorderColor[1] = clampSamplerDesc.BorderColor[2] = clampSamplerDesc.BorderColor[3] = 0;
clampSamplerDesc.MinLOD = 0;
clampSamplerDesc.MaxLOD = D3D12_FLOAT32_MAX;
m_device->CreateSampler(&clampSamplerDesc, samplerHandle);
}
// Create lights.
for (int i = 0; i < NumLights; i++)
{
// Set up each of the light positions and directions (they all start
// in the same place).
m_lights[i].position = { 0.0f, 15.0f, -30.0f, 1.0f };
m_lights[i].direction = { 0.0, 0.0f, 1.0f, 0.0f };
m_lights[i].falloff = { 800.0f, 1.0f, 0.0f, 1.0f };
m_lights[i].color = { 0.7f, 0.7f, 0.7f, 1.0f };
XMVECTOR eye = XMLoadFloat4(&m_lights[i].position);
XMVECTOR at = XMVectorAdd(eye, XMLoadFloat4(&m_lights[i].direction));
XMVECTOR up = { 0, 1, 0 };
m_lightCameras[i].Set(eye, at, up);
}
// Close the command list and use it to execute the initial GPU setup.
ThrowIfFailed(commandList->Close());
ID3D12CommandList* ppCommandLists[] = { commandList.Get() };
m_commandQueue->ExecuteCommandLists(_countof(ppCommandLists), ppCommandLists);
// Create frame resources.
for (int i = 0; i < FrameCount; i++)
{
m_frameResources[i] = new FrameResource(m_device.Get(), m_pipelineState.Get(), m_pipelineStateShadowMap.Get(), m_dsvHeap.Get(), m_cbvSrvHeap.Get(), &m_viewport, i);
m_frameResources[i]->WriteConstantBuffers(&m_viewport, &m_camera, m_lightCameras, m_lights);
}
m_currentFrameResourceIndex = 0;
m_pCurrentFrameResource = m_frameResources[m_currentFrameResourceIndex];
// Create synchronization objects and wait until assets have been uploaded to the GPU.
{
ThrowIfFailed(m_device->CreateFence(m_fenceValue, D3D12_FENCE_FLAG_NONE, IID_PPV_ARGS(&m_fence)));
m_fenceValue++;
// Create an event handle to use for frame synchronization.
m_fenceEvent = CreateEventEx(nullptr, FALSE, FALSE, EVENT_ALL_ACCESS);
if (m_fenceEvent == nullptr)
{
ThrowIfFailed(HRESULT_FROM_WIN32(GetLastError()));
}
// Wait for the command list to execute; we are reusing the same command
// list in our main loop but for now, we just want to wait for setup to
// complete before continuing.
// Signal and increment the fence value.
const UINT64 fenceToWaitFor = m_fenceValue;
ThrowIfFailed(m_commandQueue->Signal(m_fence.Get(), fenceToWaitFor));
m_fenceValue++;
// Wait until the fence is completed.
ThrowIfFailed(m_fence->SetEventOnCompletion(fenceToWaitFor, m_fenceEvent));
WaitForSingleObject(m_fenceEvent, INFINITE);
}
}
// Fill the command list with all the render commands and dependent state.
void D3D12PredicationQueries::PopulateCommandList()
{
// Command list allocators can only be reset when the associated
// command lists have finished execution on the GPU; apps should use
// fences to determine GPU execution progress.
ThrowIfFailed(m_commandAllocators[m_frameIndex]->Reset());
// However, when ExecuteCommandList() is called on a particular command
// list, that command list can then be reset at any time and must be before
// re-recording.
ThrowIfFailed(m_commandList->Reset(m_commandAllocators[m_frameIndex].Get(), m_pipelineState.Get()));
// Set necessary state.
m_commandList->SetGraphicsRootSignature(m_rootSignature.Get());
ID3D12DescriptorHeap* ppHeaps[] = { m_cbvHeap.Get() };
m_commandList->SetDescriptorHeaps(_countof(ppHeaps), ppHeaps);
m_commandList->RSSetViewports(1, &m_viewport);
m_commandList->RSSetScissorRects(1, &m_scissorRect);
// Indicate that the back buffer will be used as a render target.
m_commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_renderTargets[m_frameIndex].Get(), D3D12_RESOURCE_STATE_PRESENT, D3D12_RESOURCE_STATE_RENDER_TARGET));
CD3DX12_CPU_DESCRIPTOR_HANDLE rtvHandle(m_rtvHeap->GetCPUDescriptorHandleForHeapStart(), m_frameIndex, m_rtvDescriptorSize);
CD3DX12_CPU_DESCRIPTOR_HANDLE dsvHandle(m_dsvHeap->GetCPUDescriptorHandleForHeapStart());
m_commandList->OMSetRenderTargets(1, &rtvHandle, FALSE, &dsvHandle);
// Record commands.
const float clearColor[] = { 0.0f, 0.2f, 0.4f, 1.0f };
m_commandList->ClearRenderTargetView(rtvHandle, clearColor, 0, nullptr);
m_commandList->ClearDepthStencilView(dsvHandle, D3D12_CLEAR_FLAG_DEPTH, 1.0f, 0, 0, nullptr);
// Draw the quads and perform the occlusion query.
{
CD3DX12_GPU_DESCRIPTOR_HANDLE cbvFarQuad(m_cbvHeap->GetGPUDescriptorHandleForHeapStart(), m_frameIndex * CbvCountPerFrame, m_cbvSrvDescriptorSize);
CD3DX12_GPU_DESCRIPTOR_HANDLE cbvNearQuad(cbvFarQuad, m_cbvSrvDescriptorSize);
m_commandList->IASetPrimitiveTopology(D3D_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP);
m_commandList->IASetVertexBuffers(0, 1, &m_vertexBufferView);
// Draw the far quad conditionally based on the result of the occlusion query
// from the previous frame.
PIXBeginEvent(m_commandList.Get(), 0, L"Draw potentially occluded geometry");
m_commandList->SetGraphicsRootDescriptorTable(0, cbvFarQuad);
m_commandList->SetPredication(m_queryResult.Get(), 0, D3D12_PREDICATION_OP_EQUAL_ZERO);
m_commandList->DrawInstanced(4, 1, 0, 0);
PIXEndEvent(m_commandList.Get());
// Disable predication and always draw the near quad.
PIXBeginEvent(m_commandList.Get(), 0, L"Draw animating geometry");
m_commandList->SetPredication(nullptr, 0, D3D12_PREDICATION_OP_EQUAL_ZERO);
m_commandList->SetGraphicsRootDescriptorTable(0, cbvNearQuad);
m_commandList->DrawInstanced(4, 1, 4, 0);
PIXEndEvent(m_commandList.Get());
// Run the occlusion query with the bounding box quad.
PIXBeginEvent(m_commandList.Get(), 0, L"Execute occlusion query");
m_commandList->SetGraphicsRootDescriptorTable(0, cbvFarQuad);
m_commandList->SetPipelineState(m_queryState.Get());
m_commandList->BeginQuery(m_queryHeap.Get(), D3D12_QUERY_TYPE_BINARY_OCCLUSION, 0);
m_commandList->DrawInstanced(4, 1, 8, 0);
m_commandList->EndQuery(m_queryHeap.Get(), D3D12_QUERY_TYPE_BINARY_OCCLUSION, 0);
PIXEndEvent(m_commandList.Get());
// Resolve the occlusion query and store the results in the query result buffer
// to be used on the subsequent frame.
m_commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_queryResult.Get(), D3D12_RESOURCE_STATE_PREDICATION, D3D12_RESOURCE_STATE_COPY_DEST));
m_commandList->ResolveQueryData(m_queryHeap.Get(), D3D12_QUERY_TYPE_BINARY_OCCLUSION, 0, 1, m_queryResult.Get(), 0);
m_commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_queryResult.Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_PREDICATION));
}
// Indicate that the back buffer will now be used to present.
m_commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_renderTargets[m_frameIndex].Get(), D3D12_RESOURCE_STATE_RENDER_TARGET, D3D12_RESOURCE_STATE_PRESENT));
ThrowIfFailed(m_commandList->Close());
}
// Fill the command list with all the render commands and dependent state.
void D3D12PipelineStateCache::PopulateCommandList()
{
// Command list allocators can only be reset when the associated
// command lists have finished execution on the GPU; apps should use
// fences to determine GPU execution progress.
ThrowIfFailed(m_commandAllocators[m_frameIndex]->Reset());
// However, when ExecuteCommandList() is called on a particular command
// list, that command list can then be reset at any time and must be before
// re-recording.
ThrowIfFailed(m_commandList->Reset(m_commandAllocators[m_frameIndex].Get(), nullptr));
// Set necessary state.
m_commandList->SetGraphicsRootSignature(m_rootSignature.Get());
ID3D12DescriptorHeap* ppHeaps[] = { m_srvHeap.Get() };
m_commandList->SetDescriptorHeaps(_countof(ppHeaps), ppHeaps);
m_commandList->IASetPrimitiveTopology(D3D_PRIMITIVE_TOPOLOGY_TRIANGLELIST);
m_commandList->RSSetViewports(1, &m_viewport);
m_commandList->RSSetScissorRects(1, &m_scissorRect);
CD3DX12_CPU_DESCRIPTOR_HANDLE rtvHandle(m_rtvHeap->GetCPUDescriptorHandleForHeapStart(), m_frameIndex, m_rtvDescriptorSize);
CD3DX12_CPU_DESCRIPTOR_HANDLE intermediateRtvHandle(m_rtvHeap->GetCPUDescriptorHandleForHeapStart(), FrameCount, m_rtvDescriptorSize);
CD3DX12_GPU_DESCRIPTOR_HANDLE srvHandle(m_srvHeap->GetGPUDescriptorHandleForHeapStart());
m_commandList->OMSetRenderTargets(1, &intermediateRtvHandle, FALSE, nullptr);
// Record commands.
m_commandList->ClearRenderTargetView(intermediateRtvHandle, IntermediateClearColor, 0, nullptr);
// Draw the scene as normal into the intermediate buffer.
PIXBeginEvent(m_commandList.Get(), 0, L"Draw cube");
{
static float rot = 0.0f;
DrawConstantBuffer* drawCB = (DrawConstantBuffer*)m_dynamicCB.GetMappedMemory(m_drawIndex, m_frameIndex);
drawCB->worldViewProjection = XMMatrixTranspose(XMMatrixRotationY(rot) * XMMatrixRotationX(-rot) * m_camera.GetViewMatrix() * m_projectionMatrix);
rot += 0.01f;
m_commandList->IASetVertexBuffers(0, 1, &m_cubeVbv);
m_commandList->IASetIndexBuffer(&m_cubeIbv);
m_psoLibrary.SetPipelineState(m_device.Get(), m_rootSignature.Get(), m_commandList.Get(), BaseNormal3DRender, m_frameIndex);
m_commandList->SetGraphicsRootConstantBufferView(RootParameterCB, m_dynamicCB.GetGpuVirtualAddress(m_drawIndex, m_frameIndex));
m_commandList->DrawIndexedInstanced(36, 1, 0, 0, 0);
m_drawIndex++;
}
PIXEndEvent(m_commandList.Get());
// Set up the state for a fullscreen quad.
m_commandList->IASetVertexBuffers(0, 1, &m_quadVbv);
m_commandList->IASetPrimitiveTopology(D3D_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP);
// Indicate that the back buffer will be used as a render target and the
// intermediate render target will be used as a SRV.
D3D12_RESOURCE_BARRIER barriers[] = {
CD3DX12_RESOURCE_BARRIER::Transition(m_renderTargets[m_frameIndex].Get(), D3D12_RESOURCE_STATE_PRESENT, D3D12_RESOURCE_STATE_RENDER_TARGET),
CD3DX12_RESOURCE_BARRIER::Transition(m_intermediateRenderTarget.Get(), D3D12_RESOURCE_STATE_RENDER_TARGET, D3D12_RESOURCE_STATE_PIXEL_SHADER_RESOURCE)
};
m_commandList->ResourceBarrier(_countof(barriers), barriers);
m_commandList->SetGraphicsRootDescriptorTable(RootParameterSRV, m_srvHeap->GetGPUDescriptorHandleForHeapStart());
const float black[] = { 0.0f, 0.0f, 0.0f, 0.0f };
m_commandList->ClearRenderTargetView(rtvHandle, black, 0, nullptr);
m_commandList->OMSetRenderTargets(1, &rtvHandle, FALSE, nullptr);
// Draw some quads using the rendered scene with some effect shaders.
PIXBeginEvent(m_commandList.Get(), 0, L"Post-processing");
{
UINT quadCount = 0;
static const UINT quadsX = 3;
static const UINT quadsY = 3;
// Cycle through all of the effects.
for (UINT i = PostBlit; i < EffectPipelineTypeCount; i++)
{
if (m_enabledEffects[i])
{
D3D12_VIEWPORT viewport = {};
viewport.TopLeftX = (quadCount % quadsX) * (m_viewport.Width / quadsX);
viewport.TopLeftY = (quadCount / quadsY) * (m_viewport.Height / quadsY);
viewport.Width = m_viewport.Width / quadsX;
viewport.Height = m_viewport.Height / quadsY;
viewport.MinDepth = 0.0f;
viewport.MaxDepth = 0.0f;
PIXBeginEvent(m_commandList.Get(), 0, g_cEffectNames[i]);
m_commandList->RSSetViewports(1, &viewport);
m_psoLibrary.SetPipelineState(m_device.Get(), m_rootSignature.Get(), m_commandList.Get(), static_cast<EffectPipelineType>(i), m_frameIndex);
m_commandList->DrawInstanced(4, 1, 0, 0);
PIXEndEvent(m_commandList.Get());
}
quadCount++;
}
}
PIXEndEvent(m_commandList.Get());
// Revert resource states back to original values.
barriers[0].Transition.StateBefore = D3D12_RESOURCE_STATE_RENDER_TARGET;
barriers[0].Transition.StateAfter = D3D12_RESOURCE_STATE_PRESENT;
barriers[1].Transition.StateBefore = D3D12_RESOURCE_STATE_PIXEL_SHADER_RESOURCE;
barriers[1].Transition.StateAfter = D3D12_RESOURCE_STATE_RENDER_TARGET;
m_commandList->ResourceBarrier(_countof(barriers), barriers);
ThrowIfFailed(m_commandList->Close());
}
// Fill the command list with all the render commands and dependent state.
void D3D12Fullscreen::PopulateCommandLists()
{
// Command list allocators can only be reset when the associated
// command lists have finished execution on the GPU; apps should use
// fences to determine GPU execution progress.
ThrowIfFailed(m_sceneCommandAllocators[m_frameIndex]->Reset());
ThrowIfFailed(m_postCommandAllocators[m_frameIndex]->Reset());
// However, when ExecuteCommandList() is called on a particular command
// list, that command list can then be reset at any time and must be before
// re-recording.
ThrowIfFailed(m_sceneCommandList->Reset(m_sceneCommandAllocators[m_frameIndex].Get(), m_scenePipelineState.Get()));
ThrowIfFailed(m_postCommandList->Reset(m_postCommandAllocators[m_frameIndex].Get(), m_postPipelineState.Get()));
// Populate m_sceneCommandList to render scene to intermediate render target.
{
// Set necessary state.
m_sceneCommandList->SetGraphicsRootSignature(m_sceneRootSignature.Get());
ID3D12DescriptorHeap* ppHeaps[] = { m_cbvSrvHeap.Get() };
m_sceneCommandList->SetDescriptorHeaps(_countof(ppHeaps), ppHeaps);
CD3DX12_GPU_DESCRIPTOR_HANDLE cbvHandle(m_cbvSrvHeap->GetGPUDescriptorHandleForHeapStart(), m_frameIndex + 1, m_cbvSrvDescriptorSize);
m_sceneCommandList->SetGraphicsRootDescriptorTable(0, cbvHandle);
m_sceneCommandList->IASetPrimitiveTopology(D3D_PRIMITIVE_TOPOLOGY_TRIANGLELIST);
m_sceneCommandList->RSSetViewports(1, &m_sceneViewport);
m_sceneCommandList->RSSetScissorRects(1, &m_sceneScissorRect);
CD3DX12_CPU_DESCRIPTOR_HANDLE rtvHandle(m_rtvHeap->GetCPUDescriptorHandleForHeapStart(), FrameCount, m_rtvDescriptorSize);
m_sceneCommandList->OMSetRenderTargets(1, &rtvHandle, FALSE, nullptr);
// Record commands.
m_sceneCommandList->ClearRenderTargetView(rtvHandle, ClearColor, 0, nullptr);
m_sceneCommandList->IASetPrimitiveTopology(D3D_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP);
m_sceneCommandList->IASetVertexBuffers(0, 1, &m_sceneVertexBufferView);
PIXBeginEvent(m_sceneCommandList.Get(), 0, L"Draw a thin rectangle");
m_sceneCommandList->DrawInstanced(4, 1, 0, 0);
PIXEndEvent(m_sceneCommandList.Get());
}
ThrowIfFailed(m_sceneCommandList->Close());
// Populate m_postCommandList to scale intermediate render target to screen.
{
// Set necessary state.
m_postCommandList->SetGraphicsRootSignature(m_postRootSignature.Get());
ID3D12DescriptorHeap* ppHeaps[] = { m_cbvSrvHeap.Get() };
m_postCommandList->SetDescriptorHeaps(_countof(ppHeaps), ppHeaps);
// Indicate that the back buffer will be used as a render target and the
// intermediate render target will be used as a SRV.
D3D12_RESOURCE_BARRIER barriers[] = {
CD3DX12_RESOURCE_BARRIER::Transition(m_renderTargets[m_frameIndex].Get(), D3D12_RESOURCE_STATE_PRESENT, D3D12_RESOURCE_STATE_RENDER_TARGET),
CD3DX12_RESOURCE_BARRIER::Transition(m_intermediateRenderTarget.Get(), D3D12_RESOURCE_STATE_RENDER_TARGET, D3D12_RESOURCE_STATE_PIXEL_SHADER_RESOURCE)
};
m_postCommandList->ResourceBarrier(_countof(barriers), barriers);
m_postCommandList->SetGraphicsRootDescriptorTable(0, m_cbvSrvHeap->GetGPUDescriptorHandleForHeapStart());
m_postCommandList->IASetPrimitiveTopology(D3D_PRIMITIVE_TOPOLOGY_TRIANGLELIST);
m_postCommandList->RSSetViewports(1, &m_postViewport);
m_postCommandList->RSSetScissorRects(1, &m_postScissorRect);
CD3DX12_CPU_DESCRIPTOR_HANDLE rtvHandle(m_rtvHeap->GetCPUDescriptorHandleForHeapStart(), m_frameIndex, m_rtvDescriptorSize);
m_postCommandList->OMSetRenderTargets(1, &rtvHandle, FALSE, nullptr);
// Record commands.
m_postCommandList->ClearRenderTargetView(rtvHandle, LetterboxColor, 0, nullptr);
m_postCommandList->IASetPrimitiveTopology(D3D_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP);
m_postCommandList->IASetVertexBuffers(0, 1, &m_postVertexBufferView);
PIXBeginEvent(m_postCommandList.Get(), 0, L"Draw texture to screen.");
m_postCommandList->DrawInstanced(4, 1, 0, 0);
PIXEndEvent(m_postCommandList.Get());
// Revert resource states back to original values.
barriers[0].Transition.StateBefore = D3D12_RESOURCE_STATE_RENDER_TARGET;
barriers[0].Transition.StateAfter = D3D12_RESOURCE_STATE_PRESENT;
barriers[1].Transition.StateBefore = D3D12_RESOURCE_STATE_PIXEL_SHADER_RESOURCE;
barriers[1].Transition.StateAfter = D3D12_RESOURCE_STATE_RENDER_TARGET;
m_postCommandList->ResourceBarrier(_countof(barriers), barriers);
}
ThrowIfFailed(m_postCommandList->Close());
}
// Fill the command list with all the render commands and dependent state.
void D3D12ExecuteIndirect::PopulateCommandLists()
{
// Command list allocators can only be reset when the associated
// command lists have finished execution on the GPU; apps should use
// fences to determine GPU execution progress.
ThrowIfFailed(m_computeCommandAllocators[m_frameIndex]->Reset());
ThrowIfFailed(m_commandAllocators[m_frameIndex]->Reset());
// However, when ExecuteCommandList() is called on a particular command
// list, that command list can then be reset at any time and must be before
// re-recording.
ThrowIfFailed(m_computeCommandList->Reset(m_computeCommandAllocators[m_frameIndex].Get(), m_computeState.Get()));
ThrowIfFailed(m_commandList->Reset(m_commandAllocators[m_frameIndex].Get(), m_pipelineState.Get()));
// Record the compute commands that will cull triangles and prevent them from being processed by the vertex shader.
if (m_enableCulling)
{
UINT frameDescriptorOffset = m_frameIndex * CbvSrvUavDescriptorCountPerFrame;
D3D12_GPU_DESCRIPTOR_HANDLE cbvSrvUavHandle = m_cbvSrvUavHeap->GetGPUDescriptorHandleForHeapStart();
m_computeCommandList->SetComputeRootSignature(m_computeRootSignature.Get());
ID3D12DescriptorHeap* ppHeaps[] = { m_cbvSrvUavHeap.Get() };
m_computeCommandList->SetDescriptorHeaps(_countof(ppHeaps), ppHeaps);
m_computeCommandList->SetComputeRootDescriptorTable(
SrvUavTable,
CD3DX12_GPU_DESCRIPTOR_HANDLE(cbvSrvUavHandle, CbvSrvOffset + frameDescriptorOffset, m_cbvSrvUavDescriptorSize));
m_computeCommandList->SetComputeRoot32BitConstants(RootConstants, 4, reinterpret_cast<void*>(&m_csRootConstants), 0);
// Reset the UAV counter for this frame.
m_computeCommandList->CopyBufferRegion(m_processedCommandBuffers[m_frameIndex].Get(), CommandBufferCounterOffset, m_processedCommandBufferCounterReset.Get(), 0, sizeof(UINT));
D3D12_RESOURCE_BARRIER barrier = CD3DX12_RESOURCE_BARRIER::Transition(m_processedCommandBuffers[m_frameIndex].Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_UNORDERED_ACCESS);
m_computeCommandList->ResourceBarrier(1, &barrier);
m_computeCommandList->Dispatch(static_cast<UINT>(ceil(TriangleCount / float(ComputeThreadBlockSize))), 1, 1);
}
ThrowIfFailed(m_computeCommandList->Close());
// Record the rendering commands.
{
// Set necessary state.
m_commandList->SetGraphicsRootSignature(m_rootSignature.Get());
ID3D12DescriptorHeap* ppHeaps[] = { m_cbvSrvUavHeap.Get() };
m_commandList->SetDescriptorHeaps(_countof(ppHeaps), ppHeaps);
m_commandList->RSSetViewports(1, &m_viewport);
m_commandList->RSSetScissorRects(1, m_enableCulling ? &m_cullingScissorRect : &m_scissorRect);
// Indicate that the command buffer will be used for indirect drawing
// and that the back buffer will be used as a render target.
D3D12_RESOURCE_BARRIER barriers[2] = {
CD3DX12_RESOURCE_BARRIER::Transition(
m_enableCulling ? m_processedCommandBuffers[m_frameIndex].Get() : m_commandBuffer.Get(),
m_enableCulling ? D3D12_RESOURCE_STATE_UNORDERED_ACCESS : D3D12_RESOURCE_STATE_NON_PIXEL_SHADER_RESOURCE,
D3D12_RESOURCE_STATE_INDIRECT_ARGUMENT),
CD3DX12_RESOURCE_BARRIER::Transition(
m_renderTargets[m_frameIndex].Get(),
D3D12_RESOURCE_STATE_PRESENT,
D3D12_RESOURCE_STATE_RENDER_TARGET)
};
m_commandList->ResourceBarrier(_countof(barriers), barriers);
CD3DX12_CPU_DESCRIPTOR_HANDLE rtvHandle(m_rtvHeap->GetCPUDescriptorHandleForHeapStart(), m_frameIndex, m_rtvDescriptorSize);
CD3DX12_CPU_DESCRIPTOR_HANDLE dsvHandle(m_dsvHeap->GetCPUDescriptorHandleForHeapStart());
m_commandList->OMSetRenderTargets(1, &rtvHandle, FALSE, &dsvHandle);
// Record commands.
const float clearColor[] = { 0.0f, 0.2f, 0.4f, 1.0f };
m_commandList->ClearRenderTargetView(rtvHandle, clearColor, 0, nullptr);
m_commandList->ClearDepthStencilView(dsvHandle, D3D12_CLEAR_FLAG_DEPTH, 1.0f, 0, 0, nullptr);
m_commandList->IASetPrimitiveTopology(D3D_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP);
m_commandList->IASetVertexBuffers(0, 1, &m_vertexBufferView);
if (m_enableCulling)
{
PIXBeginEvent(m_commandList.Get(), 0, L"Draw visible triangles");
// Draw the triangles that have not been culled.
m_commandList->ExecuteIndirect(
m_commandSignature.Get(),
TriangleCount,
m_processedCommandBuffers[m_frameIndex].Get(),
0,
m_processedCommandBuffers[m_frameIndex].Get(),
CommandBufferCounterOffset);
}
else
{
PIXBeginEvent(m_commandList.Get(), 0, L"Draw all triangles");
// Draw all of the triangles.
m_commandList->ExecuteIndirect(
m_commandSignature.Get(),
TriangleCount,
m_commandBuffer.Get(),
CommandSizePerFrame * m_frameIndex,
nullptr,
0);
}
PIXEndEvent(m_commandList.Get());
// Indicate that the command buffer may be used by the compute shader
// and that the back buffer will now be used to present.
barriers[0].Transition.StateBefore = D3D12_RESOURCE_STATE_INDIRECT_ARGUMENT;
barriers[0].Transition.StateAfter = m_enableCulling ? D3D12_RESOURCE_STATE_COPY_DEST : D3D12_RESOURCE_STATE_NON_PIXEL_SHADER_RESOURCE;
barriers[1].Transition.StateBefore = D3D12_RESOURCE_STATE_RENDER_TARGET;
barriers[1].Transition.StateAfter = D3D12_RESOURCE_STATE_PRESENT;
m_commandList->ResourceBarrier(_countof(barriers), barriers);
ThrowIfFailed(m_commandList->Close());
}
}
~_ActualBracket()
{
PIXEndEvent(copyPtr);
}