void VolumetricAnimation::PopulateComputeCommandList()
{
HRESULT hr;
V( m_computeCmdAllocator->Reset() );
V( m_computeCmdList->Reset( m_computeCmdAllocator.Get(), m_computeState.Get() ) );
m_computeCmdList->SetPipelineState( m_computeState.Get() );
m_computeCmdList->SetComputeRootSignature( m_computeRootSignature.Get() );
ID3D12DescriptorHeap* ppHeaps[] = { m_cbvsrvuavHeap.Get() };
m_computeCmdList->SetDescriptorHeaps( _countof( ppHeaps ), ppHeaps );
CD3DX12_GPU_DESCRIPTOR_HANDLE cbvHandle( m_cbvsrvuavHeap->GetGPUDescriptorHandleForHeapStart(), RootParameterCBV, m_cbvsrvuavDescriptorSize );
CD3DX12_GPU_DESCRIPTOR_HANDLE uavHandle( m_cbvsrvuavHeap->GetGPUDescriptorHandleForHeapStart(), RootParameterUAV, m_cbvsrvuavDescriptorSize );
m_computeCmdList->SetComputeRootDescriptorTable( RootParameterCBV, cbvHandle );
m_computeCmdList->SetComputeRootDescriptorTable( RootParameterUAV, uavHandle );
m_computeCmdList->Dispatch( m_volumeWidth / 8, m_volumeHeight/ 8, m_volumeDepth/ 8);
m_computeCmdList->Close();
}
// Load the assets.
HRESULT VolumetricAnimation::LoadAssets()
{
HRESULT hr;
// Create a root signature consisting of a descriptor table with a CBV SRV and a sampler.
{
CD3DX12_DESCRIPTOR_RANGE ranges[3];
CD3DX12_ROOT_PARAMETER rootParameters[3];
ranges[0].Init( D3D12_DESCRIPTOR_RANGE_TYPE_CBV, 1, 0 );
ranges[1].Init( D3D12_DESCRIPTOR_RANGE_TYPE_SRV, 1, 0 );
ranges[2].Init( D3D12_DESCRIPTOR_RANGE_TYPE_UAV, 1, 0 );
rootParameters[RootParameterCBV].InitAsDescriptorTable( 1, &ranges[0], D3D12_SHADER_VISIBILITY_ALL );
rootParameters[RootParameterSRV].InitAsDescriptorTable( 1, &ranges[1], D3D12_SHADER_VISIBILITY_PIXEL );
rootParameters[RootParameterUAV].InitAsDescriptorTable( 1, &ranges[2], D3D12_SHADER_VISIBILITY_ALL );
D3D12_STATIC_SAMPLER_DESC sampler = {};
sampler.Filter = D3D12_FILTER_MIN_MAG_MIP_POINT;
sampler.AddressU = D3D12_TEXTURE_ADDRESS_MODE_BORDER;
sampler.AddressV = D3D12_TEXTURE_ADDRESS_MODE_BORDER;
sampler.AddressW = D3D12_TEXTURE_ADDRESS_MODE_BORDER;
sampler.MipLODBias = 0;
sampler.MaxAnisotropy = 0;
sampler.ComparisonFunc = D3D12_COMPARISON_FUNC_NEVER;
sampler.BorderColor = D3D12_STATIC_BORDER_COLOR_TRANSPARENT_BLACK;
sampler.MinLOD = 0.0f;
sampler.MaxLOD = D3D12_FLOAT32_MAX;
sampler.ShaderRegister = 0;
sampler.RegisterSpace = 0;
sampler.ShaderVisibility = D3D12_SHADER_VISIBILITY_PIXEL;
// Allow input layout and deny unnecessary access to certain pipeline stages.
D3D12_ROOT_SIGNATURE_FLAGS rootSignatureFlags =
D3D12_ROOT_SIGNATURE_FLAG_ALLOW_INPUT_ASSEMBLER_INPUT_LAYOUT |
D3D12_ROOT_SIGNATURE_FLAG_DENY_HULL_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_DOMAIN_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_GEOMETRY_SHADER_ROOT_ACCESS;
CD3DX12_ROOT_SIGNATURE_DESC rootSignatureDesc;
rootSignatureDesc.Init( _countof(rootParameters), rootParameters, 1, &sampler, rootSignatureFlags );
ComPtr<ID3DBlob> signature;
ComPtr<ID3DBlob> error;
V( D3D12SerializeRootSignature( &rootSignatureDesc, D3D_ROOT_SIGNATURE_VERSION_1, &signature, &error ) );
if ( error ) PRINTERROR( reinterpret_cast< const char* >( error->GetBufferPointer() ) );
VRET( m_device->CreateRootSignature( 0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS( &m_graphicsRootSignature ) ) );
DXDebugName( m_graphicsRootSignature );
// Create compute signature. Must change visibility for the SRV.
rootParameters[RootParameterSRV].ShaderVisibility = D3D12_SHADER_VISIBILITY_ALL;
CD3DX12_ROOT_SIGNATURE_DESC computeRootSignatureDesc( _countof( rootParameters ), rootParameters, 0, nullptr );
VRET( D3D12SerializeRootSignature( &computeRootSignatureDesc, D3D_ROOT_SIGNATURE_VERSION_1, &signature, &error ) );
VRET( m_device->CreateRootSignature( 0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS( &m_computeRootSignature ) ) );
}
// Create the pipeline state, which includes compiling and loading shaders.
{
ComPtr<ID3DBlob> vertexShader;
ComPtr<ID3DBlob> pixelShader;
ComPtr<ID3DBlob> computeShader;
UINT compileFlags = 0;
VRET( CompileShaderFromFile( GetAssetFullPath( _T( "VolumetricAnimation_shader.hlsl" ) ).c_str(), nullptr, D3D_COMPILE_STANDARD_FILE_INCLUDE, "vsmain", "vs_5_0", compileFlags, 0, &vertexShader ) );
VRET( CompileShaderFromFile( GetAssetFullPath( _T( "VolumetricAnimation_shader.hlsl" ) ).c_str(), nullptr, D3D_COMPILE_STANDARD_FILE_INCLUDE, "psmain", "ps_5_0", compileFlags, 0, &pixelShader ) );
VRET( CompileShaderFromFile( GetAssetFullPath( _T( "VolumetricAnimation_shader.hlsl" ) ).c_str(), nullptr, D3D_COMPILE_STANDARD_FILE_INCLUDE, "csmain", "cs_5_0", compileFlags, 0, &computeShader ) );
// Define the vertex input layout.
D3D12_INPUT_ELEMENT_DESC inputElementDescs[] =
{
{ "POSITION", 0, DXGI_FORMAT_R32G32B32_FLOAT, 0, D3D12_APPEND_ALIGNED_ELEMENT, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 }
};
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 graphics pipeline state object (PSO).
D3D12_GRAPHICS_PIPELINE_STATE_DESC psoDesc = {};
psoDesc.InputLayout = { inputElementDescs, _countof( inputElementDescs ) };
psoDesc.pRootSignature = m_graphicsRootSignature.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;
VRET( m_device->CreateGraphicsPipelineState( &psoDesc, IID_PPV_ARGS( &m_pipelineState ) ) );
DXDebugName( m_pipelineState );
// Describe and create the compute pipeline state object (PSO).
D3D12_COMPUTE_PIPELINE_STATE_DESC computePsoDesc = {};
computePsoDesc.pRootSignature = m_computeRootSignature.Get();
computePsoDesc.CS = { reinterpret_cast< UINT8* >( computeShader->GetBufferPointer() ), computeShader->GetBufferSize() };
VRET( m_device->CreateComputePipelineState( &computePsoDesc, IID_PPV_ARGS( &m_computeState ) ) );
DXDebugName( m_computeState );
}
// Create the compute command list.
VRET( m_device->CreateCommandList( 0, D3D12_COMMAND_LIST_TYPE_COMPUTE, m_computeCmdAllocator.Get(),m_computeState.Get(), IID_PPV_ARGS( &m_computeCmdList ) ) );
DXDebugName( m_computeCmdList );
VRET( m_computeCmdList->Close() );
// Create the graphics command list.
VRET( m_device->CreateCommandList( 0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_graphicCmdAllocator.Get(), m_pipelineState.Get(), IID_PPV_ARGS( &m_graphicCmdList ) ) );
DXDebugName( m_graphicCmdList );
// Note: ComPtr's are CPU objects but this resource needs to stay in scope until
// the command list that references it has finished executing on the GPU.
// We will flush the GPU at the end of this method to ensure the resource is not
// prematurely destroyed.
ComPtr<ID3D12Resource> volumeBufferUploadHeap;
// Create the volumeBuffer.
{
UINT volumeBufferSize = m_volumeDepth*m_volumeHeight*m_volumeWidth * 4 * sizeof( UINT8 );
D3D12_RESOURCE_DESC bufferDesc = CD3DX12_RESOURCE_DESC::Buffer( volumeBufferSize, D3D12_RESOURCE_FLAG_ALLOW_UNORDERED_ACCESS );
D3D12_RESOURCE_DESC uploadBufferDesc = CD3DX12_RESOURCE_DESC::Buffer( volumeBufferSize );
VRET( m_device->CreateCommittedResource(&CD3DX12_HEAP_PROPERTIES( D3D12_HEAP_TYPE_DEFAULT ),D3D12_HEAP_FLAG_NONE,
&bufferDesc,D3D12_RESOURCE_STATE_COPY_DEST,nullptr,IID_PPV_ARGS( &m_volumeBuffer ) ) );
const UINT64 uploadBufferSize = GetRequiredIntermediateSize( m_volumeBuffer.Get(), 0, 1 );
// Create the GPU upload buffer.
VRET( m_device->CreateCommittedResource(&CD3DX12_HEAP_PROPERTIES( D3D12_HEAP_TYPE_UPLOAD ),D3D12_HEAP_FLAG_NONE,
&uploadBufferDesc,D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,IID_PPV_ARGS( &volumeBufferUploadHeap ) ) );
// Copy data to the intermediate upload heap and then schedule a copy
// from the upload heap to the Texture2D.
UINT8* volumeBuffer = ( UINT8* ) malloc( volumeBufferSize );
memset( volumeBuffer, 64, volumeBufferSize );
//float radius = m_volumeHeight / 2.f;
float a = m_volumeWidth / 2.f;
float b = m_volumeHeight / 2.f;
float c = m_volumeDepth / 2.f;
float radius = sqrt( a*a + b*b + c*c );
for ( UINT z = 0; z < m_volumeDepth; z++ )
for ( UINT y = 0; y < m_volumeHeight; y++ )
for ( UINT x = 0; x < m_volumeWidth; x++ )
{
float _x = x - m_volumeWidth / 2.f;
float _y = y - m_volumeHeight / 2.f;
float _z = z - m_volumeDepth / 2.f;
//float currentRaidus =abs(_x)+abs(_y)+abs(_z);
float currentRaidus = sqrt( _x*_x + _y*_y + _z*_z );
float scale = currentRaidus *3.f / radius;
UINT idx = 4 - (UINT)floor( scale );
UINT interm = ( UINT ) ( 192 * scale +0.5f );
UINT8 col = interm % 192+1;
volumeBuffer[( x + y*m_volumeWidth + z*m_volumeHeight*m_volumeWidth ) * 4 + 0] += col * m_constantBufferData.colVal[idx].x;
volumeBuffer[( x + y*m_volumeWidth + z*m_volumeHeight*m_volumeWidth ) * 4 + 1] += col * m_constantBufferData.colVal[idx].y;
volumeBuffer[( x + y*m_volumeWidth + z*m_volumeHeight*m_volumeWidth ) * 4 + 2] += col * m_constantBufferData.colVal[idx].z;
volumeBuffer[( x + y*m_volumeWidth + z*m_volumeHeight*m_volumeWidth ) * 4 + 3] = m_constantBufferData.colVal[idx].w;
}
D3D12_SUBRESOURCE_DATA volumeBufferData = {};
volumeBufferData.pData = &volumeBuffer[0];
volumeBufferData.RowPitch = volumeBufferSize;
volumeBufferData.SlicePitch = volumeBufferData.RowPitch;
UpdateSubresources( m_graphicCmdList.Get(), m_volumeBuffer.Get(), volumeBufferUploadHeap.Get(), 0, 0, 1, &volumeBufferData );
m_graphicCmdList->ResourceBarrier( 1, &CD3DX12_RESOURCE_BARRIER::Transition( m_volumeBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_UNORDERED_ACCESS ) );
// Describe and create a SRV for the volumeBuffer.
D3D12_SHADER_RESOURCE_VIEW_DESC srvDesc = {};
srvDesc.Shader4ComponentMapping = D3D12_DEFAULT_SHADER_4_COMPONENT_MAPPING;
srvDesc.Format = DXGI_FORMAT_UNKNOWN;
srvDesc.ViewDimension = D3D12_SRV_DIMENSION_BUFFER;
srvDesc.Buffer.FirstElement = 0;
srvDesc.Buffer.NumElements = m_volumeDepth*m_volumeHeight*m_volumeWidth;
srvDesc.Buffer.StructureByteStride = 4 * sizeof( UINT8 );
srvDesc.Buffer.Flags = D3D12_BUFFER_SRV_FLAG_NONE;
CD3DX12_CPU_DESCRIPTOR_HANDLE srvHandle( m_cbvsrvuavHeap->GetCPUDescriptorHandleForHeapStart(), RootParameterSRV, m_cbvsrvuavDescriptorSize );
m_device->CreateShaderResourceView( m_volumeBuffer.Get(), &srvDesc, srvHandle );
// Describe and create a UAV for the volumeBuffer.
D3D12_UNORDERED_ACCESS_VIEW_DESC uavDesc = {};
uavDesc.Format = DXGI_FORMAT_UNKNOWN;
uavDesc.ViewDimension = D3D12_UAV_DIMENSION_BUFFER;
uavDesc.Buffer.FirstElement = 0;
uavDesc.Buffer.NumElements = m_volumeWidth*m_volumeHeight*m_volumeDepth;
uavDesc.Buffer.StructureByteStride = 4 * sizeof( UINT8 );
uavDesc.Buffer.CounterOffsetInBytes = 0;
uavDesc.Buffer.Flags = D3D12_BUFFER_UAV_FLAG_NONE;
CD3DX12_CPU_DESCRIPTOR_HANDLE uavHandle( m_cbvsrvuavHeap->GetCPUDescriptorHandleForHeapStart(), RootParameterUAV, m_cbvsrvuavDescriptorSize );
m_device->CreateUnorderedAccessView( m_volumeBuffer.Get(), nullptr, &uavDesc, uavHandle );
free( volumeBuffer );
}
// Create the vertex buffer.
// Note: ComPtr's are CPU objects but this resource needs to stay in scope until
// the command list that references it has finished executing on the GPU.
// We will flush the GPU at the end of this method to ensure the resource is not
// prematurely destroyed.
ComPtr<ID3D12Resource> vertexBufferUpload;
{
// Define the geometry for a triangle.
Vertex cubeVertices[] =
{
{ XMFLOAT3( -128.f, -128.f, -128.f ) },
{ XMFLOAT3( -128.f, -128.f, 128.f ) },
{ XMFLOAT3( -128.f, 128.f, -128.f ) },
{ XMFLOAT3( -128.f, 128.f, 128.f ) },
{ XMFLOAT3( 128.f, -128.f, -128.f )},
{ XMFLOAT3( 128.f, -128.f, 128.f )},
{ XMFLOAT3( 128.f, 128.f, -128.f )},
{ XMFLOAT3( 128.f, 128.f, 128.f )},
};
const UINT vertexBufferSize = sizeof( cubeVertices );
VRET( m_device->CreateCommittedResource( &CD3DX12_HEAP_PROPERTIES( D3D12_HEAP_TYPE_UPLOAD ), D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer( vertexBufferSize ), D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr, IID_PPV_ARGS( &vertexBufferUpload ) ) );
VRET( m_device->CreateCommittedResource( &CD3DX12_HEAP_PROPERTIES( D3D12_HEAP_TYPE_DEFAULT ), D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer( vertexBufferSize ), D3D12_RESOURCE_STATE_COPY_DEST,
nullptr, IID_PPV_ARGS( &m_vertexBuffer ) ) );
DXDebugName( m_vertexBuffer );
D3D12_SUBRESOURCE_DATA vertexData = {};
vertexData.pData = reinterpret_cast< UINT8* >( cubeVertices );
vertexData.RowPitch = vertexBufferSize;
vertexData.SlicePitch = vertexBufferSize;
UpdateSubresources<1>( m_graphicCmdList.Get(), m_vertexBuffer.Get(), vertexBufferUpload.Get(), 0, 0, 1, &vertexData );
m_graphicCmdList->ResourceBarrier( 1, &CD3DX12_RESOURCE_BARRIER::Transition( m_vertexBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST,
D3D12_RESOURCE_STATE_VERTEX_AND_CONSTANT_BUFFER ));
// Initialize the vertex buffer view.
m_vertexBufferView.BufferLocation = m_vertexBuffer->GetGPUVirtualAddress();
m_vertexBufferView.StrideInBytes = sizeof( Vertex );
m_vertexBufferView.SizeInBytes = vertexBufferSize;
}
// Create the index buffer
// Note: ComPtr's are CPU objects but this resource needs to stay in scope until
// the command list that references it has finished executing on the GPU.
// We will flush the GPU at the end of this method to ensure the resource is not
// prematurely destroyed.
ComPtr<ID3D12Resource> indexBufferUpload;
{
uint16_t cubeIndices[] =
{
0,2,1, 1,2,3, 4,5,6, 5,7,6, 0,1,5, 0,5,4, 2,6,7, 2,7,3, 0,4,6, 0,6,2, 1,3,7, 1,7,5,
};
const UINT indexBufferSize = sizeof( cubeIndices );
VRET( m_device->CreateCommittedResource( &CD3DX12_HEAP_PROPERTIES( D3D12_HEAP_TYPE_UPLOAD ), D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer( indexBufferSize ), D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr, IID_PPV_ARGS( &indexBufferUpload ) ) );
VRET( m_device->CreateCommittedResource( &CD3DX12_HEAP_PROPERTIES( D3D12_HEAP_TYPE_DEFAULT ), D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer( indexBufferSize ), D3D12_RESOURCE_STATE_COPY_DEST,
nullptr, IID_PPV_ARGS( &m_indexBuffer ) ) );
DXDebugName( m_indexBuffer );
D3D12_SUBRESOURCE_DATA indexData = {};
indexData.pData = reinterpret_cast< UINT8* >( cubeIndices );
indexData.RowPitch = indexBufferSize;
indexData.SlicePitch = indexBufferSize;
UpdateSubresources<1>( m_graphicCmdList.Get(), m_indexBuffer.Get(), indexBufferUpload.Get(), 0, 0, 1, &indexData );
m_graphicCmdList->ResourceBarrier( 1, &CD3DX12_RESOURCE_BARRIER::Transition( m_indexBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST,
D3D12_RESOURCE_STATE_INDEX_BUFFER ) );
m_indexBufferView.BufferLocation = m_indexBuffer->GetGPUVirtualAddress();
m_indexBufferView.SizeInBytes = sizeof( cubeIndices );
m_indexBufferView.Format = DXGI_FORMAT_R16_UINT;
}
// Create the constant buffer
{
VRET( m_device->CreateCommittedResource( &CD3DX12_HEAP_PROPERTIES( D3D12_HEAP_TYPE_UPLOAD ), D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer( 1024 * 64 ), D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr, IID_PPV_ARGS( &m_constantBuffer ) ) );
DXDebugName( m_constantBuffer );
// Describe and create a constant buffer view.
D3D12_CONSTANT_BUFFER_VIEW_DESC cbvDesc = {};
cbvDesc.BufferLocation = m_constantBuffer->GetGPUVirtualAddress();
cbvDesc.SizeInBytes = ( sizeof( ConstantBuffer ) + 255 ) & ~255; // CB size is required to be 256-byte aligned.
m_device->CreateConstantBufferView( &cbvDesc, m_cbvsrvuavHeap->GetCPUDescriptorHandleForHeapStart() );
// Initialize and map the constant buffers. We don't unmap this until the
// app closes. Keeping things mapped for the lifetime of the resource is okay.
CD3DX12_RANGE readRange( 0, 0 ); // We do not intend to read from this resource on the CPU.
VRET( m_constantBuffer->Map( 0, &readRange, reinterpret_cast< void** >( &m_pCbvDataBegin ) ) );
memcpy( m_pCbvDataBegin, &m_constantBufferData, sizeof( m_constantBufferData ) );
}
// Close the command list and execute it to begin the initial GPU setup.
VRET( m_graphicCmdList->Close() );
ID3D12CommandList* ppCommandLists[] = { m_graphicCmdList.Get() };
m_graphicCmdQueue->ExecuteCommandLists( _countof( ppCommandLists ), ppCommandLists );
// Create synchronization objects and wait until assets have been uploaded to the GPU.
{
VRET( m_device->CreateFence( 0, D3D12_FENCE_FLAG_NONE, IID_PPV_ARGS( &m_fence ) ) );
DXDebugName( m_fence );
m_fenceValue = 1;
// Create an event handle to use for frame synchronization.
m_fenceEvent = CreateEvent( nullptr, FALSE, FALSE, nullptr );
if ( m_fenceEvent == nullptr )
{
VRET( 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.
WaitForGraphicsCmd();
}
XMVECTORF32 vecEye = { 500.0f, 500.0f, -500.0f };
XMVECTORF32 vecAt = { 0.0f, 0.0f, 0.0f };
m_camera.SetViewParams( vecEye, vecAt );
m_camera.SetEnablePositionMovement( true );
m_camera.SetButtonMasks( MOUSE_RIGHT_BUTTON, MOUSE_WHEEL, MOUSE_LEFT_BUTTON );
return S_OK;
}
/**
***************************************************************************************************
* DX12ImageRenderer::CaptureImage
*
* @brief
* Convert a DX12 resource to a CPU-visible linear buffer of pixels. The data is filled
* in a user-provided CpuImage struct.
*
* IMPORTANT: Memory inside pImgOut is allocated on behalf of the caller, so it is their
* responsibility to free it.
*
* @return
* S_OK if successful.
***************************************************************************************************
*/
HRESULT DX12ImageRenderer::CaptureImage(
ID3D12Resource* pRes,
D3D12_RESOURCE_STATES prevState,
UINT newWidth,
UINT newHeight,
CpuImage* pImgOut,
bool bFlipX,
bool bFlipY)
{
HRESULT result = E_FAIL;
if ((pRes != nullptr) && (pImgOut != nullptr) && (newWidth > 0) && (newHeight > 0))
{
// Create temp assets
result = CreateCaptureAssets(pRes, newWidth, newHeight);
if (result == S_OK)
{
result = m_pCmdAllocator->Reset();
}
if (result == S_OK)
{
result = m_pCmdList->Reset(m_pCmdAllocator, m_pPipelineStateGraphics);
}
// Render work
if (result == S_OK)
{
// Set root sig
m_pCmdList->SetGraphicsRootSignature(m_pRootSignatureGraphics);
// Set descriptors and tables
ID3D12DescriptorHeap* ppHeaps[] = { m_pSrvUavCbHeap };
m_pCmdList->SetDescriptorHeaps(_countof(ppHeaps), ppHeaps);
CD3DX12_GPU_DESCRIPTOR_HANDLE srvHandle(m_pSrvUavCbHeap->GetGPUDescriptorHandleForHeapStart(), RootParameterSRV, m_srvUavCbDescriptorSize);
CD3DX12_GPU_DESCRIPTOR_HANDLE uavHandle(m_pSrvUavCbHeap->GetGPUDescriptorHandleForHeapStart(), RootParameterUAV, m_srvUavCbDescriptorSize);
CD3DX12_GPU_DESCRIPTOR_HANDLE cbHandle(m_pSrvUavCbHeap->GetGPUDescriptorHandleForHeapStart(), RootParameterCBV, m_srvUavCbDescriptorSize);
m_pCmdList->SetGraphicsRootDescriptorTable(RootParameterSRV, srvHandle);
m_pCmdList->SetGraphicsRootDescriptorTable(RootParameterUAV, uavHandle);
m_pCmdList->SetGraphicsRootDescriptorTable(RootParameterCBV, cbHandle);
// Set viewport
D3D12_VIEWPORT viewport = {};
viewport.Width = static_cast<float>(newWidth);
viewport.Height = static_cast<float>(newHeight);
viewport.MaxDepth = 1.0f;
m_pCmdList->RSSetViewports(1, &viewport);
// Set scissor
D3D12_RECT scissorRect = {};
scissorRect.right = static_cast<LONG>(newWidth);
scissorRect.bottom = static_cast<LONG>(newHeight);
m_pCmdList->RSSetScissorRects(1, &scissorRect);
// Update const buf
m_constantBufferData.rtWidth = newWidth;
m_constantBufferData.flipX = bFlipX ? 1 : 0;
m_constantBufferData.flipY = bFlipY ? 1 : 0;
CD3DX12_RANGE readRange(0, 0);
result = m_pConstantBuffer->Map(0, &readRange, reinterpret_cast<void**>(&m_pCbvDataBegin));
memcpy(m_pCbvDataBegin, &m_constantBufferData, sizeof(m_constantBufferData));
m_pConstantBuffer->Unmap(0, nullptr);
// Set RT
CD3DX12_CPU_DESCRIPTOR_HANDLE internalRtvHandle(m_pInternalRtvHeap->GetCPUDescriptorHandleForHeapStart());
m_config.pDevice->CreateRenderTargetView(m_pInternalRT, nullptr, internalRtvHandle);
m_pCmdList->OMSetRenderTargets(1, &internalRtvHandle, FALSE, nullptr);
// Record commands
m_pCmdList->ClearRenderTargetView(internalRtvHandle, ClearColor, 0, nullptr);
m_pCmdList->IASetPrimitiveTopology(D3D_PRIMITIVE_TOPOLOGY_TRIANGLELIST);
CD3DX12_RESOURCE_BARRIER barrier = {};
// Render full screen quad and write out UAV
barrier = CD3DX12_RESOURCE_BARRIER::Transition(pRes, prevState, D3D12_RESOURCE_STATE_PIXEL_SHADER_RESOURCE);
m_pCmdList->ResourceBarrier(1, &barrier);
m_pCmdList->DrawInstanced(3, 1, 0, 0);
#if OVERWRITE_SRC_RES
barrier = CD3DX12_RESOURCE_BARRIER::Transition(pRes, D3D12_RESOURCE_STATE_PIXEL_SHADER_RESOURCE, D3D12_RESOURCE_STATE_COPY_DEST);
m_pCmdList->ResourceBarrier(1, &barrier);
barrier = CD3DX12_RESOURCE_BARRIER::Transition(m_pInternalRT, D3D12_RESOURCE_STATE_RENDER_TARGET, D3D12_RESOURCE_STATE_COPY_SOURCE);
m_pCmdList->ResourceBarrier(1, &barrier);
m_pCmdList->CopyResource(pRes, m_pInternalRT);
barrier = CD3DX12_RESOURCE_BARRIER::Transition(m_pInternalRT, D3D12_RESOURCE_STATE_COPY_SOURCE, D3D12_RESOURCE_STATE_RENDER_TARGET);
m_pCmdList->ResourceBarrier(1, &barrier);
barrier = CD3DX12_RESOURCE_BARRIER::Transition(pRes, D3D12_RESOURCE_STATE_COPY_DEST, prevState);
m_pCmdList->ResourceBarrier(1, &barrier);
#endif
// Copy UAV to CPU-visible buffer
barrier = CD3DX12_RESOURCE_BARRIER::Transition(m_pPSWriteBuf, D3D12_RESOURCE_STATE_UNORDERED_ACCESS, D3D12_RESOURCE_STATE_COPY_SOURCE);
m_pCmdList->ResourceBarrier(1, &barrier);
m_pCmdList->CopyResource(m_pPSWriteBufReadBack, m_pPSWriteBuf);
barrier = CD3DX12_RESOURCE_BARRIER::Transition(m_pPSWriteBuf, D3D12_RESOURCE_STATE_COPY_SOURCE, D3D12_RESOURCE_STATE_UNORDERED_ACCESS);
m_pCmdList->ResourceBarrier(1, &barrier);
// Execute the command list
result = m_pCmdList->Close();
ID3D12CommandList* ppCommandLists[] = { m_pCmdList };
m_config.pCmdQueue->ExecuteCommandLists(_countof(ppCommandLists), ppCommandLists);
WaitCmdListFinish();
// Read back UAV results
void* pUavData = nullptr;
result = m_pPSWriteBufReadBack->Map(0, &readRange, &pUavData);
if (result == S_OK)
{
const UINT totalBytes = newWidth * newHeight * BytesPerPixel;
pImgOut->pitch = newWidth * BytesPerPixel;
pImgOut->width = newWidth;
pImgOut->height = newHeight;
pImgOut->pData = new char[totalBytes];
memcpy(pImgOut->pData, pUavData, totalBytes);
m_pPSWriteBufReadBack->Unmap(0, nullptr);
}
}
// Free temp assets
FreeCaptureAssets();
}
return result;
}
