#include "../win/Win32Common.hpp" #include "boo/graphicsdev/D3D.hpp" #include "boo/graphicsdev/IGraphicsCommandQueue.hpp" #include "boo/IGraphicsContext.hpp" #include "lib/graphicsdev/Common.hpp" #include #include #include #include #include #include #include #include #include #include #undef min #undef max extern pD3DCompile D3DCompilePROC; extern pD3DPERF_BeginEvent D3DPERF_BeginEventPROC; extern pD3DPERF_EndEvent D3DPERF_EndEventPROC; constexpr char GammaVS[] = "struct VertData\n" "{\n" " float4 posIn : POSITION;\n" " float4 uvIn : UV;\n" "};\n" "\n" "struct VertToFrag\n" "{\n" " float4 pos : SV_Position;\n" " float2 uv : UV;\n" "};\n" "\n" "VertToFrag main(in VertData v)\n" "{\n" " VertToFrag vtf;\n" " vtf.uv = v.uvIn.xy;\n" " vtf.pos = v.posIn;\n" " return vtf;\n" "}\n"; constexpr char GammaFS[] = "struct VertToFrag\n" "{\n" " float4 pos : SV_Position;\n" " float2 uv : UV;\n" "};\n" "\n" "Texture2D screenTex : register(t0);\n" "Texture2D gammaLUT : register(t1);\n" "SamplerState samp : register(s3);\n" "float4 main(in VertToFrag vtf) : SV_Target0\n" "{\n" " int4 tex = int4(saturate(screenTex.Sample(samp, vtf.uv)) * float4(65535.0, 65535.0, 65535.0, 65535.0));\n" " float4 colorOut;\n" " for (int i=0 ; i<3 ; ++i)\n" " colorOut[i] = gammaLUT.Load(int3(tex[i] % 256, tex[i] / 256, 0)).r;\n" " return colorOut;\n" "}\n"; namespace boo { static logvisor::Module Log("boo::D3D11"); class D3D11DataFactory; static inline void ThrowIfFailed(HRESULT hr) { if (FAILED(hr)) { // Set a breakpoint on this line to catch Win32 API errors. #if !WINDOWS_STORE _com_error err(hr); #else _com_error err(hr, L"D3D11 fail"); #endif LPCTSTR errMsg = err.ErrorMessage(); Log.report(logvisor::Fatal, fmt(_SYS_STR("{}")), errMsg); } } constexpr std::array USE_TABLE{ D3D11_BIND_VERTEX_BUFFER, D3D11_BIND_VERTEX_BUFFER, D3D11_BIND_INDEX_BUFFER, D3D11_BIND_CONSTANT_BUFFER, }; class D3D11GraphicsBufferS : public GraphicsDataNode { friend class D3D11DataFactory::Context; friend struct D3D11CommandQueue; size_t m_sz; D3D11GraphicsBufferS(const boo::ObjToken& parent, BufferUse use, D3D11Context* ctx, const void* data, size_t stride, size_t count) : GraphicsDataNode(parent), m_sz(stride * count), m_stride(stride), m_count(count) { D3D11_SUBRESOURCE_DATA iData = {data}; CD3D11_BUFFER_DESC desc(m_sz, USE_TABLE[int(use)], D3D11_USAGE_IMMUTABLE); ThrowIfFailed(ctx->m_dev->CreateBuffer(&desc, &iData, &m_buf)); } public: size_t m_stride; size_t m_count; ComPtr m_buf; ~D3D11GraphicsBufferS() override = default; }; template class D3D11GraphicsBufferD : public GraphicsDataNode { friend class D3D11DataFactory::Context; friend class D3D11DataFactoryImpl; friend struct D3D11CommandQueue; D3D11CommandQueue* m_q; std::unique_ptr m_cpuBuf; size_t m_cpuSz; int m_validSlots = 0; D3D11GraphicsBufferD(const boo::ObjToken& parent, D3D11CommandQueue* q, BufferUse use, D3D11Context* ctx, size_t stride, size_t count) : GraphicsDataNode(parent), m_q(q), m_stride(stride), m_count(count) { m_cpuSz = stride * count; m_cpuBuf.reset(new uint8_t[m_cpuSz]); for (auto& buf : m_bufs) { const CD3D11_BUFFER_DESC desc(m_cpuSz, USE_TABLE[int(use)], D3D11_USAGE_DYNAMIC, D3D11_CPU_ACCESS_WRITE); ThrowIfFailed(ctx->m_dev->CreateBuffer(&desc, nullptr, &buf)); } } void update(ID3D11DeviceContext* ctx, int b); public: ~D3D11GraphicsBufferD() override = default; void load(const void* data, size_t sz) override; void* map(size_t sz) override; void unmap() override; size_t m_stride; size_t m_count; std::array, 3> m_bufs; }; class D3D11TextureS : public GraphicsDataNode { friend class D3D11DataFactory::Context; D3D11TextureS(const boo::ObjToken& parent, D3D11Context* ctx, size_t width, size_t height, size_t mips, TextureFormat fmt, const void* data, size_t sz) : GraphicsDataNode(parent) { DXGI_FORMAT pfmt = DXGI_FORMAT_UNKNOWN; int pxPitchNum = 1; int pxPitchDenom = 1; int pxTilePitch = 0; bool compressed = false; switch (fmt) { case TextureFormat::RGBA8: pfmt = DXGI_FORMAT_R8G8B8A8_UNORM; pxPitchNum = 4; break; case TextureFormat::I8: pfmt = DXGI_FORMAT_R8_UNORM; break; case TextureFormat::I16: pfmt = DXGI_FORMAT_R16_UNORM; pxPitchNum = 2; break; case TextureFormat::DXT1: pfmt = DXGI_FORMAT_BC1_UNORM; compressed = true; pxPitchNum = 1; pxPitchDenom = 2; pxTilePitch = 2; break; case TextureFormat::DXT3: pfmt = DXGI_FORMAT_BC2_UNORM; compressed = true; pxPitchNum = 1; pxPitchDenom = 1; pxTilePitch = 4; break; default: Log.report(logvisor::Fatal, fmt("unsupported tex format")); } const CD3D11_TEXTURE2D_DESC desc(pfmt, width, height, 1, mips, D3D11_BIND_SHADER_RESOURCE, D3D11_USAGE_IMMUTABLE); const uint8_t* dataIt = static_cast(data); std::array upData{}; for (size_t i = 0; i < mips && i < upData.size(); ++i) { upData[i].pSysMem = dataIt; upData[i].SysMemPitch = width * pxPitchNum / pxPitchDenom; upData[i].SysMemSlicePitch = upData[i].SysMemPitch * height; if (compressed) { upData[i].SysMemPitch = width * pxTilePitch; } dataIt += upData[i].SysMemSlicePitch; if (width > 1) { width /= 2; } if (height > 1) { height /= 2; } } ThrowIfFailed(ctx->m_dev->CreateTexture2D(&desc, upData.data(), &m_tex)); const CD3D11_SHADER_RESOURCE_VIEW_DESC srvDesc(D3D_SRV_DIMENSION_TEXTURE2D, pfmt, 0, mips); ThrowIfFailed(ctx->m_dev->CreateShaderResourceView(m_tex.Get(), &srvDesc, &m_srv)); } public: ComPtr m_tex; ComPtr m_srv; ~D3D11TextureS() override = default; }; class D3D11TextureSA : public GraphicsDataNode { friend class D3D11DataFactory::Context; D3D11TextureSA(const boo::ObjToken& parent, D3D11Context* ctx, size_t width, size_t height, size_t layers, size_t mips, TextureFormat fmt, const void* data, size_t sz) : GraphicsDataNode(parent) { size_t pixelPitch = 0; DXGI_FORMAT pixelFmt = DXGI_FORMAT_UNKNOWN; switch (fmt) { case TextureFormat::RGBA8: pixelPitch = 4; pixelFmt = DXGI_FORMAT_R8G8B8A8_UNORM; break; case TextureFormat::I8: pixelPitch = 1; pixelFmt = DXGI_FORMAT_R8_UNORM; break; case TextureFormat::I16: pixelPitch = 2; pixelFmt = DXGI_FORMAT_R16_UNORM; break; default: Log.report(logvisor::Fatal, fmt("unsupported tex format")); } CD3D11_TEXTURE2D_DESC desc(pixelFmt, width, height, layers, mips, D3D11_BIND_SHADER_RESOURCE, D3D11_USAGE_IMMUTABLE); const uint8_t* dataIt = static_cast(data); std::unique_ptr upData(new D3D11_SUBRESOURCE_DATA[layers * mips]); D3D11_SUBRESOURCE_DATA* outIt = upData.get(); for (size_t i = 0; i < mips; ++i) { for (size_t j = 0; j < layers; ++j) { outIt->pSysMem = dataIt; outIt->SysMemPitch = width * pixelPitch; outIt->SysMemSlicePitch = outIt->SysMemPitch * height; dataIt += outIt->SysMemSlicePitch; ++outIt; } if (width > 1) width /= 2; if (height > 1) height /= 2; } ThrowIfFailed(ctx->m_dev->CreateTexture2D(&desc, upData.get(), &m_tex)); CD3D11_SHADER_RESOURCE_VIEW_DESC srvDesc(D3D_SRV_DIMENSION_TEXTURE2DARRAY, pixelFmt, 0, mips, 0, layers); ThrowIfFailed(ctx->m_dev->CreateShaderResourceView(m_tex.Get(), &srvDesc, &m_srv)); } public: ComPtr m_tex; ComPtr m_srv; ~D3D11TextureSA() override = default; }; class D3D11TextureD : public GraphicsDataNode { friend class D3D11DataFactory::Context; friend struct D3D11CommandQueue; size_t m_width = 0; D3D11CommandQueue* m_q; std::unique_ptr m_cpuBuf; size_t m_cpuSz = 0; size_t m_pxPitch = 0; int m_validSlots = 0; D3D11TextureD(const boo::ObjToken& parent, D3D11CommandQueue* q, D3D11Context* ctx, size_t width, size_t height, TextureFormat fmt) : GraphicsDataNode(parent), m_width(width), m_q(q) { DXGI_FORMAT pixelFmt = DXGI_FORMAT_UNKNOWN; switch (fmt) { case TextureFormat::RGBA8: pixelFmt = DXGI_FORMAT_R8G8B8A8_UNORM; m_pxPitch = 4; break; case TextureFormat::I8: pixelFmt = DXGI_FORMAT_R8_UNORM; m_pxPitch = 1; break; case TextureFormat::I16: pixelFmt = DXGI_FORMAT_R16_UNORM; m_pxPitch = 2; break; default: Log.report(logvisor::Fatal, fmt("unsupported tex format")); } m_cpuSz = width * height * m_pxPitch; m_cpuBuf.reset(new uint8_t[m_cpuSz]); const CD3D11_TEXTURE2D_DESC desc(pixelFmt, width, height, 1, 1, D3D11_BIND_SHADER_RESOURCE, D3D11_USAGE_DYNAMIC, D3D11_CPU_ACCESS_WRITE); const CD3D11_SHADER_RESOURCE_VIEW_DESC srvDesc(D3D_SRV_DIMENSION_TEXTURE2D, pixelFmt, 0, 1); for (size_t i = 0; i < NumResources; ++i) { ThrowIfFailed(ctx->m_dev->CreateTexture2D(&desc, nullptr, &m_texs[i])); ThrowIfFailed(ctx->m_dev->CreateShaderResourceView(m_texs[i].Get(), &srvDesc, &m_srvs[i])); } } void update(ID3D11DeviceContext* ctx, int b); public: ~D3D11TextureD() override = default; void load(const void* data, size_t sz) override; void* map(size_t sz) override; void unmap() override; static constexpr size_t NumResources = 3; std::array, NumResources> m_texs; std::array, NumResources> m_srvs; }; class D3D11TextureR : public GraphicsDataNode { friend class D3D11DataFactory::Context; friend struct D3D11CommandQueue; size_t m_width = 0; size_t m_height = 0; size_t m_samples = 0; size_t m_colorBindCount; size_t m_depthBindCount; void Setup(D3D11Context* ctx) { const CD3D11_TEXTURE2D_DESC colorDesc(ctx->m_fbFormat, m_width, m_height, 1, 1, D3D11_BIND_RENDER_TARGET, D3D11_USAGE_DEFAULT, 0, m_samples); ThrowIfFailed(ctx->m_dev->CreateTexture2D(&colorDesc, nullptr, &m_colorTex)); const CD3D11_TEXTURE2D_DESC depthDesc(DXGI_FORMAT_D32_FLOAT, m_width, m_height, 1, 1, D3D11_BIND_DEPTH_STENCIL, D3D11_USAGE_DEFAULT, 0, m_samples); ThrowIfFailed(ctx->m_dev->CreateTexture2D(&depthDesc, nullptr, &m_depthTex)); D3D11_RTV_DIMENSION rtvDim; D3D11_DSV_DIMENSION dsvDim; if (m_samples > 1) { rtvDim = D3D11_RTV_DIMENSION_TEXTURE2DMS; dsvDim = D3D11_DSV_DIMENSION_TEXTURE2DMS; } else { rtvDim = D3D11_RTV_DIMENSION_TEXTURE2D; dsvDim = D3D11_DSV_DIMENSION_TEXTURE2D; } const CD3D11_RENDER_TARGET_VIEW_DESC rtvDesc(rtvDim, ctx->m_fbFormat); ThrowIfFailed(ctx->m_dev->CreateRenderTargetView(m_colorTex.Get(), &rtvDesc, &m_rtv)); const CD3D11_DEPTH_STENCIL_VIEW_DESC dsvDesc(dsvDim, DXGI_FORMAT_D32_FLOAT); ThrowIfFailed(ctx->m_dev->CreateDepthStencilView(m_depthTex.Get(), &dsvDesc, &m_dsv)); const CD3D11_TEXTURE2D_DESC colorBindDesc(ctx->m_fbFormat, m_width, m_height, 1, 1, D3D11_BIND_SHADER_RESOURCE, D3D11_USAGE_DEFAULT, 0, 1); const CD3D11_SHADER_RESOURCE_VIEW_DESC colorSrvDesc(D3D11_SRV_DIMENSION_TEXTURE2D, ctx->m_fbFormat, 0, 1); for (size_t i = 0; i < m_colorBindCount; ++i) { ThrowIfFailed(ctx->m_dev->CreateTexture2D(&colorBindDesc, nullptr, &m_colorBindTex[i])); ThrowIfFailed(ctx->m_dev->CreateShaderResourceView(m_colorBindTex[i].Get(), &colorSrvDesc, &m_colorSrv[i])); } const CD3D11_TEXTURE2D_DESC depthBindDesc(DXGI_FORMAT_R32_FLOAT, m_width, m_height, 1, 1, D3D11_BIND_SHADER_RESOURCE, D3D11_USAGE_DEFAULT, 0, 1); const CD3D11_SHADER_RESOURCE_VIEW_DESC depthSrvDesc(D3D11_SRV_DIMENSION_TEXTURE2D, DXGI_FORMAT_R32_FLOAT, 0, 1); for (size_t i = 0; i < m_depthBindCount; ++i) { ThrowIfFailed(ctx->m_dev->CreateTexture2D(&depthBindDesc, nullptr, &m_depthBindTex[i])); ThrowIfFailed(ctx->m_dev->CreateShaderResourceView(m_depthBindTex[i].Get(), &depthSrvDesc, &m_depthSrv[i])); } } D3D11TextureR(const boo::ObjToken& parent, D3D11Context* ctx, size_t width, size_t height, size_t samples, size_t colorBindCount, size_t depthBindCount) : GraphicsDataNode(parent) , m_width(width) , m_height(height) , m_samples(samples) , m_colorBindCount(colorBindCount) , m_depthBindCount(depthBindCount) { if (colorBindCount > m_colorBindTex.size()) { Log.report(logvisor::Fatal, fmt("too many color bindings for render texture")); } if (depthBindCount > m_depthBindTex.size()) { Log.report(logvisor::Fatal, fmt("too many depth bindings for render texture")); } if (samples == 0) { m_samples = 1; } Setup(ctx); } public: size_t samples() const { return m_samples; } static constexpr size_t MaxBindTexs = 4; ComPtr m_colorTex; ComPtr m_rtv; ComPtr m_depthTex; ComPtr m_dsv; std::array, MaxBindTexs> m_colorBindTex; std::array, MaxBindTexs> m_colorSrv; std::array, MaxBindTexs> m_depthBindTex; std::array, MaxBindTexs> m_depthSrv; ~D3D11TextureR() override = default; void resize(D3D11Context* ctx, size_t width, size_t height) { if (width < 1) width = 1; if (height < 1) height = 1; m_width = width; m_height = height; Setup(ctx); } }; class D3D11TextureCubeR : public GraphicsDataNode { friend class D3D11DataFactory::Context; friend struct D3D11CommandQueue; size_t m_width = 0; size_t m_mipCount = 0; void Setup(D3D11Context* ctx) { const CD3D11_TEXTURE2D_DESC colorDesc(ctx->m_fbFormat, m_width, m_width, 6, m_mipCount, D3D11_BIND_RENDER_TARGET | D3D11_BIND_SHADER_RESOURCE, D3D11_USAGE_DEFAULT, 0, 1, 0, D3D11_RESOURCE_MISC_TEXTURECUBE | D3D11_RESOURCE_MISC_GENERATE_MIPS); ThrowIfFailed(ctx->m_dev->CreateTexture2D(&colorDesc, nullptr, &m_colorTex)); const CD3D11_TEXTURE2D_DESC depthDesc(DXGI_FORMAT_D32_FLOAT, m_width, m_width, 6, 1, D3D11_BIND_DEPTH_STENCIL, D3D11_USAGE_DEFAULT, 0, 1, 0, D3D11_RESOURCE_MISC_TEXTURECUBE); ThrowIfFailed(ctx->m_dev->CreateTexture2D(&depthDesc, nullptr, &m_depthTex)); CD3D11_RENDER_TARGET_VIEW_DESC rtvDesc(D3D11_RTV_DIMENSION_TEXTURE2DARRAY, ctx->m_fbFormat, 0, 0, 1); CD3D11_DEPTH_STENCIL_VIEW_DESC dsvDesc(D3D11_DSV_DIMENSION_TEXTURE2DARRAY, DXGI_FORMAT_D32_FLOAT, 0, 0, 1); for (size_t i = 0; i < NumViews; ++i) { rtvDesc.Texture2DArray.FirstArraySlice = UINT(i); ThrowIfFailed(ctx->m_dev->CreateRenderTargetView(m_colorTex.Get(), &rtvDesc, &m_rtv[i])); dsvDesc.Texture2DArray.FirstArraySlice = UINT(i); ThrowIfFailed(ctx->m_dev->CreateDepthStencilView(m_depthTex.Get(), &dsvDesc, &m_dsv[i])); } const CD3D11_SHADER_RESOURCE_VIEW_DESC colorSrvDesc(D3D11_SRV_DIMENSION_TEXTURECUBE, ctx->m_fbFormat, 0, m_mipCount); ThrowIfFailed(ctx->m_dev->CreateShaderResourceView(m_colorTex.Get(), &colorSrvDesc, &m_colorSrv)); } D3D11TextureCubeR(const boo::ObjToken& parent, D3D11Context* ctx, size_t width, size_t mips) : GraphicsDataNode(parent), m_width(width), m_mipCount(mips) { Setup(ctx); } public: static constexpr size_t NumViews = 6; ComPtr m_colorTex; std::array, NumViews> m_rtv; ComPtr m_depthTex; std::array, NumViews> m_dsv; ComPtr m_colorSrv; ~D3D11TextureCubeR() override = default; void resize(D3D11Context* ctx, size_t width, size_t mips) { if (width < 1) width = 1; m_width = width; m_mipCount = mips; Setup(ctx); } }; constexpr std::array SEMANTIC_SIZE_TABLE{ 0, 12, 16, 12, 16, 16, 4, 8, 16, 16, 16, }; constexpr std::array SEMANTIC_NAME_TABLE{ nullptr, "POSITION", "POSITION", "NORMAL", "NORMAL", "COLOR", "COLOR", "UV", "UV", "WEIGHT", "MODELVIEW", }; constexpr std::array SEMANTIC_TYPE_TABLE{ DXGI_FORMAT_UNKNOWN, DXGI_FORMAT_R32G32B32_FLOAT, DXGI_FORMAT_R32G32B32A32_FLOAT, DXGI_FORMAT_R32G32B32_FLOAT, DXGI_FORMAT_R32G32B32A32_FLOAT, DXGI_FORMAT_R32G32B32A32_FLOAT, DXGI_FORMAT_R8G8B8A8_UNORM, DXGI_FORMAT_R32G32_FLOAT, DXGI_FORMAT_R32G32B32A32_FLOAT, DXGI_FORMAT_R32G32B32A32_FLOAT, DXGI_FORMAT_R32G32B32A32_FLOAT, }; constexpr std::array PRIMITIVE_TABLE{ D3D11_PRIMITIVE_TOPOLOGY_TRIANGLELIST, D3D11_PRIMITIVE_TOPOLOGY_TRIANGLESTRIP, D3D11_PRIMITIVE_TOPOLOGY_1_CONTROL_POINT_PATCHLIST, }; constexpr std::array BLEND_FACTOR_TABLE{ D3D11_BLEND_ZERO, D3D11_BLEND_ONE, D3D11_BLEND_SRC_COLOR, D3D11_BLEND_INV_SRC_COLOR, D3D11_BLEND_DEST_COLOR, D3D11_BLEND_INV_DEST_COLOR, D3D11_BLEND_SRC_ALPHA, D3D11_BLEND_INV_SRC_ALPHA, D3D11_BLEND_DEST_ALPHA, D3D11_BLEND_INV_DEST_ALPHA, D3D11_BLEND_SRC1_COLOR, D3D11_BLEND_INV_SRC1_COLOR, }; class D3D11ShaderStage : public GraphicsDataNode { friend class D3D11DataFactory; ComPtr m_shader; D3D11ShaderStage(const boo::ObjToken& parent, D3D11Context* ctx, const uint8_t* data, size_t size, PipelineStage stage) : GraphicsDataNode(parent) { switch (stage) { case PipelineStage::Vertex: { ThrowIfFailed(D3DCreateBlobPROC(size, &m_vtxBlob)); memcpy(m_vtxBlob->GetBufferPointer(), data, size); ComPtr vShader; ThrowIfFailed(ctx->m_dev->CreateVertexShader(data, size, nullptr, &vShader)); m_shader = vShader; break; } case PipelineStage::Fragment: { ComPtr pShader; ThrowIfFailed(ctx->m_dev->CreatePixelShader(data, size, nullptr, &pShader)); m_shader = pShader; break; } case PipelineStage::Geometry: { ComPtr gShader; ThrowIfFailed(ctx->m_dev->CreateGeometryShader(data, size, nullptr, &gShader)); m_shader = gShader; break; } case PipelineStage::Control: { ComPtr hShader; ThrowIfFailed(ctx->m_dev->CreateHullShader(data, size, nullptr, &hShader)); m_shader = hShader; break; } case PipelineStage::Evaluation: { ComPtr dShader; ThrowIfFailed(ctx->m_dev->CreateDomainShader(data, size, nullptr, &dShader)); m_shader = dShader; break; } default: break; } } public: ComPtr m_vtxBlob; template void shader(ComPtr& ret) const { m_shader.As(&ret); } }; class D3D11ShaderPipeline : public GraphicsDataNode { friend class D3D11DataFactory; friend struct D3D11ShaderDataBinding; D3D11ShaderPipeline(const boo::ObjToken& parent, D3D11Context* ctx, ObjToken vertex, ObjToken fragment, ObjToken geometry, ObjToken control, ObjToken evaluation, const VertexFormatInfo& vtxFmt, const AdditionalPipelineInfo& info) : GraphicsDataNode(parent), m_topology(PRIMITIVE_TABLE[int(info.prim)]) { if (auto* s = vertex.cast()) s->shader(m_vShader); if (auto* s = fragment.cast()) s->shader(m_pShader); if (auto* s = geometry.cast()) s->shader(m_gShader); if (auto* s = control.cast()) s->shader(m_hShader); if (auto* s = evaluation.cast()) s->shader(m_dShader); if (control && evaluation) m_topology = D3D11_PRIMITIVE_TOPOLOGY(D3D11_PRIMITIVE_TOPOLOGY_1_CONTROL_POINT_PATCHLIST + info.patchSize - 1); D3D11_CULL_MODE cullMode; switch (info.culling) { case CullMode::None: default: cullMode = D3D11_CULL_NONE; break; case CullMode::Backface: cullMode = D3D11_CULL_BACK; break; case CullMode::Frontface: cullMode = D3D11_CULL_FRONT; break; } CD3D11_RASTERIZER_DESC rasDesc(D3D11_FILL_SOLID, cullMode, true, D3D11_DEFAULT_DEPTH_BIAS, D3D11_DEFAULT_DEPTH_BIAS_CLAMP, D3D11_DEFAULT_SLOPE_SCALED_DEPTH_BIAS, true, true, false, false); ThrowIfFailed(ctx->m_dev->CreateRasterizerState(&rasDesc, &m_rasState)); CD3D11_DEPTH_STENCIL_DESC dsDesc(D3D11_DEFAULT); dsDesc.DepthEnable = info.depthTest != ZTest::None; dsDesc.DepthWriteMask = D3D11_DEPTH_WRITE_MASK(info.depthWrite); switch (info.depthTest) { case ZTest::None: default: dsDesc.DepthFunc = D3D11_COMPARISON_ALWAYS; break; case ZTest::LEqual: dsDesc.DepthFunc = D3D11_COMPARISON_GREATER_EQUAL; break; case ZTest::Greater: dsDesc.DepthFunc = D3D11_COMPARISON_LESS; break; case ZTest::GEqual: dsDesc.DepthFunc = D3D11_COMPARISON_LESS_EQUAL; break; case ZTest::Equal: dsDesc.DepthFunc = D3D11_COMPARISON_EQUAL; break; } ThrowIfFailed(ctx->m_dev->CreateDepthStencilState(&dsDesc, &m_dsState)); CD3D11_BLEND_DESC blDesc(D3D11_DEFAULT); blDesc.RenderTarget[0].BlendEnable = (info.dstFac != BlendFactor::Zero); if (info.srcFac == BlendFactor::Subtract || info.dstFac == BlendFactor::Subtract) { blDesc.RenderTarget[0].SrcBlend = D3D11_BLEND_SRC_ALPHA; blDesc.RenderTarget[0].DestBlend = D3D11_BLEND_ONE; blDesc.RenderTarget[0].BlendOp = D3D11_BLEND_OP_REV_SUBTRACT; if (info.overwriteAlpha) { blDesc.RenderTarget[0].SrcBlendAlpha = D3D11_BLEND_ONE; blDesc.RenderTarget[0].DestBlendAlpha = D3D11_BLEND_ZERO; blDesc.RenderTarget[0].BlendOpAlpha = D3D11_BLEND_OP_ADD; } else { blDesc.RenderTarget[0].SrcBlendAlpha = D3D11_BLEND_SRC_ALPHA; blDesc.RenderTarget[0].DestBlendAlpha = D3D11_BLEND_ONE; blDesc.RenderTarget[0].BlendOpAlpha = D3D11_BLEND_OP_REV_SUBTRACT; } } else { blDesc.RenderTarget[0].SrcBlend = BLEND_FACTOR_TABLE[int(info.srcFac)]; blDesc.RenderTarget[0].DestBlend = BLEND_FACTOR_TABLE[int(info.dstFac)]; blDesc.RenderTarget[0].BlendOp = D3D11_BLEND_OP_ADD; if (info.overwriteAlpha) { blDesc.RenderTarget[0].SrcBlendAlpha = D3D11_BLEND_ONE; blDesc.RenderTarget[0].DestBlendAlpha = D3D11_BLEND_ZERO; } else { blDesc.RenderTarget[0].SrcBlendAlpha = BLEND_FACTOR_TABLE[int(info.srcFac)]; blDesc.RenderTarget[0].DestBlendAlpha = BLEND_FACTOR_TABLE[int(info.dstFac)]; } blDesc.RenderTarget[0].BlendOpAlpha = D3D11_BLEND_OP_ADD; } blDesc.RenderTarget[0].RenderTargetWriteMask = (info.colorWrite ? (D3D11_COLOR_WRITE_ENABLE_RED | D3D11_COLOR_WRITE_ENABLE_GREEN | D3D11_COLOR_WRITE_ENABLE_BLUE) : 0) | (info.alphaWrite ? D3D11_COLOR_WRITE_ENABLE_ALPHA : 0); ThrowIfFailed(ctx->m_dev->CreateBlendState(&blDesc, &m_blState)); { std::vector elements(vtxFmt.elementCount); for (size_t i = 0; i < vtxFmt.elementCount; ++i) { const VertexElementDescriptor* elemin = &vtxFmt.elements[i]; D3D11_INPUT_ELEMENT_DESC& elem = elements[i]; int semantic = int(elemin->semantic & boo::VertexSemantic::SemanticMask); elem.SemanticName = SEMANTIC_NAME_TABLE[semantic]; elem.SemanticIndex = elemin->semanticIdx; elem.Format = SEMANTIC_TYPE_TABLE[semantic]; if ((elemin->semantic & boo::VertexSemantic::Instanced) != boo::VertexSemantic::None) { elem.InputSlotClass = D3D11_INPUT_PER_INSTANCE_DATA; elem.InstanceDataStepRate = 1; elem.InputSlot = 1; elem.AlignedByteOffset = m_instStride; m_instStride += SEMANTIC_SIZE_TABLE[semantic]; } else { elem.InputSlotClass = D3D11_INPUT_PER_VERTEX_DATA; elem.AlignedByteOffset = m_stride; m_stride += SEMANTIC_SIZE_TABLE[semantic]; } } const auto& vertBuf = vertex.cast()->m_vtxBlob; ThrowIfFailed(ctx->m_dev->CreateInputLayout(elements.data(), vtxFmt.elementCount, vertBuf->GetBufferPointer(), vertBuf->GetBufferSize(), &m_inLayout)); } } public: ComPtr m_vShader; ComPtr m_pShader; ComPtr m_gShader; ComPtr m_hShader; ComPtr m_dShader; ComPtr m_rasState; ComPtr m_dsState; ComPtr m_blState; ComPtr m_inLayout; D3D11_PRIMITIVE_TOPOLOGY m_topology; size_t m_stride = 0; size_t m_instStride = 0; ~D3D11ShaderPipeline() override = default; D3D11ShaderPipeline& operator=(const D3D11ShaderPipeline&) = delete; D3D11ShaderPipeline(const D3D11ShaderPipeline&) = delete; void bind(ID3D11DeviceContext* ctx) { ctx->VSSetShader(m_vShader.Get(), nullptr, 0); ctx->PSSetShader(m_pShader.Get(), nullptr, 0); ctx->GSSetShader(m_gShader.Get(), nullptr, 0); ctx->HSSetShader(m_hShader.Get(), nullptr, 0); ctx->DSSetShader(m_dShader.Get(), nullptr, 0); ctx->RSSetState(m_rasState.Get()); ctx->OMSetDepthStencilState(m_dsState.Get(), 0); ctx->OMSetBlendState(m_blState.Get(), nullptr, 0xffffffff); ctx->IASetInputLayout(m_inLayout.Get()); ctx->IASetPrimitiveTopology(m_topology); } bool isReady() const override { return true; } }; struct D3D11ShaderDataBinding : public GraphicsDataNode { boo::ObjToken m_pipeline; boo::ObjToken m_vbuf; boo::ObjToken m_instVbuf; boo::ObjToken m_ibuf; std::vector> m_ubufs; std::unique_ptr m_ubufFirstConsts; std::unique_ptr m_ubufNumConsts; std::unique_ptr m_pubufs; struct BoundTex { boo::ObjToken tex; int idx; bool depth; }; std::vector m_texs; std::array m_baseOffsets; D3D11ShaderDataBinding(const boo::ObjToken& d, D3D11Context* ctx, const boo::ObjToken& pipeline, const boo::ObjToken& vbuf, const boo::ObjToken& instVbuf, const boo::ObjToken& ibuf, size_t ubufCount, const boo::ObjToken* ubufs, const PipelineStage* ubufStages, const size_t* ubufOffs, const size_t* ubufSizes, size_t texCount, const boo::ObjToken* texs, const int* texBindIdx, const bool* depthBind, size_t baseVert, size_t baseInst) : GraphicsDataNode(d), m_pipeline(pipeline), m_vbuf(vbuf), m_instVbuf(instVbuf), m_ibuf(ibuf) { m_ubufs.reserve(ubufCount); m_texs.reserve(texCount); D3D11ShaderPipeline* cpipeline = m_pipeline.cast(); m_baseOffsets[0] = UINT(baseVert * cpipeline->m_stride); m_baseOffsets[1] = UINT(baseInst * cpipeline->m_instStride); if (ubufStages) { m_pubufs.reset(new bool[ubufCount]); for (size_t i = 0; i < ubufCount; ++i) m_pubufs[i] = ubufStages[i] == PipelineStage::Fragment; } if (ubufOffs && ubufSizes) { m_ubufFirstConsts.reset(new UINT[ubufCount]); m_ubufNumConsts.reset(new UINT[ubufCount]); for (size_t i = 0; i < ubufCount; ++i) { #ifndef NDEBUG if (ubufOffs[i] % 256) { Log.report(logvisor::Fatal, fmt("non-256-byte-aligned uniform-offset {} provided to newShaderDataBinding"), i); } #endif m_ubufFirstConsts[i] = ubufOffs[i] / 16; m_ubufNumConsts[i] = ((ubufSizes[i] + 255) & ~255) / 16; } } for (size_t i = 0; i < ubufCount; ++i) { #ifndef NDEBUG if (!ubufs[i]) { Log.report(logvisor::Fatal, fmt("null uniform-buffer {} provided to newShaderDataBinding"), i); } #endif m_ubufs.push_back(ubufs[i]); } for (size_t i = 0; i < texCount; ++i) { m_texs.push_back({texs[i], texBindIdx ? texBindIdx[i] : 0, depthBind ? depthBind[i] : false}); } } void bind(ID3D11DeviceContext1* ctx, int b) { m_pipeline.cast()->bind(ctx); std::array bufs{}; std::array strides{}; if (m_vbuf) { if (m_vbuf->dynamic()) { const auto* const cbuf = m_vbuf.cast>(); bufs[0] = cbuf->m_bufs[b].Get(); strides[0] = UINT(cbuf->m_stride); } else { const auto* const cbuf = m_vbuf.cast(); bufs[0] = cbuf->m_buf.Get(); strides[0] = UINT(cbuf->m_stride); } } if (m_instVbuf) { if (m_instVbuf->dynamic()) { const auto* const cbuf = m_instVbuf.cast>(); bufs[1] = cbuf->m_bufs[b].Get(); strides[1] = UINT(cbuf->m_stride); } else { const auto* const cbuf = m_instVbuf.cast(); bufs[1] = cbuf->m_buf.Get(); strides[1] = UINT(cbuf->m_stride); } } ctx->IASetVertexBuffers(0, UINT(bufs.size()), bufs.data(), strides.data(), m_baseOffsets.data()); if (m_ibuf) { if (m_ibuf->dynamic()) { const auto* const cbuf = m_ibuf.cast>(); ctx->IASetIndexBuffer(cbuf->m_bufs[b].Get(), DXGI_FORMAT_R32_UINT, 0); } else { const auto* const cbuf = m_ibuf.cast(); ctx->IASetIndexBuffer(cbuf->m_buf.Get(), DXGI_FORMAT_R32_UINT, 0); } } if (m_ubufs.size()) { if (m_ubufFirstConsts) { std::array constBufs{}; ctx->VSSetConstantBuffers(0, m_ubufs.size(), constBufs.data()); ctx->DSSetConstantBuffers(0, m_ubufs.size(), constBufs.data()); for (size_t i = 0; i < constBufs.size() && i < m_ubufs.size(); ++i) { if (m_pubufs && m_pubufs[i]) { continue; } if (m_ubufs[i]->dynamic()) { const auto* const cbuf = m_ubufs[i].cast>(); constBufs[i] = cbuf->m_bufs[b].Get(); } else { const auto* const cbuf = m_ubufs[i].cast(); constBufs[i] = cbuf->m_buf.Get(); } } ctx->VSSetConstantBuffers1(0, m_ubufs.size(), constBufs.data(), m_ubufFirstConsts.get(), m_ubufNumConsts.get()); ctx->DSSetConstantBuffers1(0, m_ubufs.size(), constBufs.data(), m_ubufFirstConsts.get(), m_ubufNumConsts.get()); if (m_pubufs) { std::array constBufs2{}; ctx->PSSetConstantBuffers(0, m_ubufs.size(), constBufs2.data()); for (size_t i = 0; i < constBufs2.size() && i < m_ubufs.size(); ++i) { if (!m_pubufs[i]) { continue; } if (m_ubufs[i]->dynamic()) { const auto* const cbuf = m_ubufs[i].cast>(); constBufs2[i] = cbuf->m_bufs[b].Get(); } else { const auto* const cbuf = m_ubufs[i].cast(); constBufs2[i] = cbuf->m_buf.Get(); } } ctx->PSSetConstantBuffers1(0, m_ubufs.size(), constBufs2.data(), m_ubufFirstConsts.get(), m_ubufNumConsts.get()); } } else { std::array constBufs{}; for (size_t i = 0; i < constBufs.size() && i < m_ubufs.size(); ++i) { if (m_pubufs && m_pubufs[i]) { continue; } if (m_ubufs[i]->dynamic()) { const auto* const cbuf = m_ubufs[i].cast>(); constBufs[i] = cbuf->m_bufs[b].Get(); } else { const auto* const cbuf = m_ubufs[i].cast(); constBufs[i] = cbuf->m_buf.Get(); } } ctx->VSSetConstantBuffers(0, m_ubufs.size(), constBufs.data()); ctx->DSSetConstantBuffers(0, m_ubufs.size(), constBufs.data()); if (m_pubufs) { std::array constBufs2{}; for (size_t i = 0; i < constBufs2.size() && i < m_ubufs.size(); ++i) { if (!m_pubufs[i]) { continue; } if (m_ubufs[i]->dynamic()) { const auto* const cbuf = m_ubufs[i].cast>(); constBufs2[i] = cbuf->m_bufs[b].Get(); } else { const auto* const cbuf = m_ubufs[i].cast(); constBufs2[i] = cbuf->m_buf.Get(); } } ctx->PSSetConstantBuffers(0, m_ubufs.size(), constBufs2.data()); } } } if (m_texs.size()) { std::array srvs{}; for (size_t i = 0; i < srvs.size() && i < m_texs.size(); ++i) { if (m_texs[i].tex) { switch (m_texs[i].tex->type()) { case TextureType::Dynamic: { const auto* const ctex = m_texs[i].tex.cast(); srvs[i] = ctex->m_srvs[b].Get(); break; } case TextureType::Static: { const auto* const ctex = m_texs[i].tex.cast(); srvs[i] = ctex->m_srv.Get(); break; } case TextureType::StaticArray: { const auto* const ctex = m_texs[i].tex.cast(); srvs[i] = ctex->m_srv.Get(); break; } case TextureType::Render: { const auto* const ctex = m_texs[i].tex.cast(); srvs[i] = m_texs[i].depth ? ctex->m_depthSrv[m_texs[i].idx].Get() : ctex->m_colorSrv[m_texs[i].idx].Get(); break; } case TextureType::CubeRender: { const auto* const ctex = m_texs[i].tex.cast(); srvs[i] = ctex->m_colorSrv.Get(); break; } } } } ctx->PSSetShaderResources(0, m_texs.size(), srvs.data()); ctx->DSSetShaderResources(0, m_texs.size(), srvs.data()); } } }; struct D3D11CommandQueue final : IGraphicsCommandQueue { Platform platform() const override { return IGraphicsDataFactory::Platform::D3D11; } const SystemChar* platformName() const override { return _SYS_STR("D3D11"); } D3D11Context* m_ctx; D3D11Context::Window* m_windowCtx; IGraphicsContext* m_parent; ComPtr m_deferredCtx; ComPtr m_deferredAnnot; size_t m_fillBuf = 0; size_t m_completeBuf = 0; size_t m_drawBuf = 0; bool m_running = true; std::mutex m_mt; std::condition_variable m_cv; std::mutex m_initmt; std::condition_variable m_initcv; std::unique_lock m_initlk; std::thread m_thr; struct CommandList { ComPtr list; std::vector> resTokens; boo::ObjToken workDoPresent; void reset() { list.Reset(); resTokens.clear(); workDoPresent.reset(); } }; std::array m_cmdLists; std::recursive_mutex m_dynamicLock; void ProcessDynamicLoads(ID3D11DeviceContext* ctx); static void RenderingWorker(D3D11CommandQueue* self); D3D11CommandQueue(D3D11Context* ctx, D3D11Context::Window* windowCtx, IGraphicsContext* parent) : m_ctx(ctx), m_windowCtx(windowCtx), m_parent(parent), m_initlk(m_initmt), m_thr(RenderingWorker, this) { m_initcv.wait(m_initlk); m_initlk.unlock(); ThrowIfFailed(ctx->m_dev->CreateDeferredContext1(0, &m_deferredCtx)); m_deferredCtx.As(&m_deferredAnnot); } void startRenderer() override; void stopRenderer() override { m_running = false; m_cv.notify_one(); m_thr.join(); } ~D3D11CommandQueue() override { if (m_running) stopRenderer(); } void setShaderDataBinding(const boo::ObjToken& binding) override { auto* const cbind = binding.cast(); cbind->bind(m_deferredCtx.Get(), m_fillBuf); m_cmdLists[m_fillBuf].resTokens.push_back(binding.get()); const std::array samp{ m_ctx->m_ss[0].Get(), m_ctx->m_ss[1].Get(), m_ctx->m_ss[2].Get(), m_ctx->m_ss[3].Get(), m_ctx->m_ss[4].Get(), }; m_deferredCtx->PSSetSamplers(0, UINT(samp.size()), samp.data()); m_deferredCtx->DSSetSamplers(0, UINT(samp.size()), samp.data()); } boo::ObjToken m_boundTarget; void setRenderTarget(const boo::ObjToken& target) override { auto* const ctarget = target.cast(); const std::array view{ctarget->m_rtv.Get()}; m_deferredCtx->OMSetRenderTargets(UINT(view.size()), view.data(), ctarget->m_dsv.Get()); m_boundTarget = target.get(); } int m_boundFace = 0; void setRenderTarget(const ObjToken& target, int face) override { static constexpr std::array cubeFaceRemap{0, 1, 3, 2, 4, 5}; face = cubeFaceRemap[face]; auto* const ctarget = target.cast(); const std::array view{ctarget->m_rtv[face].Get()}; m_deferredCtx->OMSetRenderTargets(UINT(view.size()), view.data(), ctarget->m_dsv[face].Get()); m_boundTarget = target.get(); m_boundFace = face; } void setViewport(const SWindowRect& rect, float znear, float zfar) override { if (!m_boundTarget) { return; } int boundHeight = 0; switch (m_boundTarget->type()) { case TextureType::Render: { const auto* const ctarget = m_boundTarget.cast(); boundHeight = ctarget->m_height; break; } case TextureType::CubeRender: { const auto* const ctarget = m_boundTarget.cast(); boundHeight = ctarget->m_width; break; } default: break; } const D3D11_VIEWPORT vp = {FLOAT(rect.location[0]), FLOAT(boundHeight - rect.location[1] - rect.size[1]), FLOAT(rect.size[0]), FLOAT(rect.size[1]), 1.f - zfar, 1.f - znear}; m_deferredCtx->RSSetViewports(1, &vp); } void setScissor(const SWindowRect& rect) override { if (!m_boundTarget) { return; } int boundHeight = 0; switch (m_boundTarget->type()) { case TextureType::Render: { const auto* const ctarget = m_boundTarget.cast(); boundHeight = ctarget->m_height; break; } case TextureType::CubeRender: { const auto* const ctarget = m_boundTarget.cast(); boundHeight = ctarget->m_width; break; } default: break; } const D3D11_RECT d3drect = {LONG(rect.location[0]), LONG(boundHeight - rect.location[1] - rect.size[1]), LONG(rect.location[0] + rect.size[0]), LONG(boundHeight - rect.location[1])}; m_deferredCtx->RSSetScissorRects(1, &d3drect); } std::unordered_map> m_texResizes; void resizeRenderTexture(const boo::ObjToken& tex, size_t width, size_t height) override { D3D11TextureR* ctex = tex.cast(); std::unique_lock lk(m_mt); m_texResizes[ctex] = std::make_pair(width, height); } std::unordered_map> m_cubeTexResizes; void resizeRenderTexture(const boo::ObjToken& tex, size_t width, size_t mips) override { D3D11TextureCubeR* ctex = tex.cast(); std::unique_lock lk(m_mt); m_cubeTexResizes[ctex] = std::make_pair(width, mips); } void generateMipmaps(const ObjToken& tex) override; void schedulePostFrameHandler(std::function&& func) override { func(); } std::array m_clearColor{0.0, 0.0, 0.0, 0.0}; void setClearColor(const float rgba[4]) override { m_clearColor[0] = rgba[0]; m_clearColor[1] = rgba[1]; m_clearColor[2] = rgba[2]; m_clearColor[3] = rgba[3]; } void clearTarget(bool render = true, bool depth = true) override { if (!m_boundTarget) { return; } switch (m_boundTarget->type()) { case TextureType::Render: { auto* const ctarget = m_boundTarget.cast(); if (render) { m_deferredCtx->ClearRenderTargetView(ctarget->m_rtv.Get(), m_clearColor.data()); } if (depth) { m_deferredCtx->ClearDepthStencilView(ctarget->m_dsv.Get(), D3D11_CLEAR_DEPTH, 0.0f, 0); } break; } case TextureType::CubeRender: { auto* const ctarget = m_boundTarget.cast(); if (render) { m_deferredCtx->ClearRenderTargetView(ctarget->m_rtv[m_boundFace].Get(), m_clearColor.data()); } if (depth) { m_deferredCtx->ClearDepthStencilView(ctarget->m_dsv[m_boundFace].Get(), D3D11_CLEAR_DEPTH, 0.0f, 0); } break; } default: break; } } void draw(size_t start, size_t count) override { m_deferredCtx->Draw(count, start); } void drawIndexed(size_t start, size_t count) override { m_deferredCtx->DrawIndexed(count, start, 0); } void drawInstances(size_t start, size_t count, size_t instCount, size_t startInst) override { m_deferredCtx->DrawInstanced(count, instCount, start, startInst); } void drawInstancesIndexed(size_t start, size_t count, size_t instCount, size_t startInst) override { m_deferredCtx->DrawIndexedInstanced(count, instCount, start, 0, startInst); } void _resolveBindTexture(ID3D11DeviceContext1* ctx, const D3D11TextureR* tex, const SWindowRect& rect, bool tlOrigin, int bindIdx, bool color, bool depth) { if (color && tex->m_colorBindCount) { if (tex->m_samples > 1) { ctx->ResolveSubresource(tex->m_colorBindTex[bindIdx].Get(), 0, tex->m_colorTex.Get(), 0, m_ctx->m_fbFormat); } else { SWindowRect intersectRect = rect.intersect(SWindowRect(0, 0, tex->m_width, tex->m_height)); int y = tlOrigin ? intersectRect.location[1] : (tex->m_height - intersectRect.size[1] - intersectRect.location[1]); D3D11_BOX box = { UINT(intersectRect.location[0]), UINT(y), 0, UINT(intersectRect.location[0] + intersectRect.size[0]), UINT(y + intersectRect.size[1]), 1}; ctx->CopySubresourceRegion1(tex->m_colorBindTex[bindIdx].Get(), 0, box.left, box.top, 0, tex->m_colorTex.Get(), 0, &box, D3D11_COPY_DISCARD); } } if (depth && tex->m_depthBindCount) { if (tex->m_samples > 1) { ctx->ResolveSubresource(tex->m_depthBindTex[bindIdx].Get(), 0, tex->m_depthTex.Get(), 0, DXGI_FORMAT_D32_FLOAT); } else { ctx->CopyResource(tex->m_depthBindTex[bindIdx].Get(), tex->m_depthTex.Get()); } } } void resolveBindTexture(const boo::ObjToken& texture, const SWindowRect& rect, bool tlOrigin, int bindIdx, bool color, bool depth, bool clearDepth) override { const D3D11TextureR* tex = texture.cast(); _resolveBindTexture(m_deferredCtx.Get(), tex, rect, tlOrigin, bindIdx, color, depth); if (clearDepth) m_deferredCtx->ClearDepthStencilView(tex->m_dsv.Get(), D3D11_CLEAR_DEPTH, 0.0f, 0); } boo::ObjToken m_doPresent; void resolveDisplay(const boo::ObjToken& source) override { m_doPresent = source; } void execute() override; #ifdef BOO_GRAPHICS_DEBUG_GROUPS void pushDebugGroup(const char* name, const std::array& color) override { if (m_deferredAnnot) m_deferredAnnot->BeginEvent(MBSTWCS(name).c_str()); } void popDebugGroup() override { if (m_deferredAnnot) m_deferredAnnot->EndEvent(); } #endif }; template void D3D11GraphicsBufferD::update(ID3D11DeviceContext* ctx, int b) { int slot = 1 << b; if ((slot & m_validSlots) == 0) { ID3D11Buffer* res = m_bufs[b].Get(); D3D11_MAPPED_SUBRESOURCE d; if (SUCCEEDED(ctx->Map(res, 0, D3D11_MAP_WRITE_DISCARD, 0, &d))) { memcpy(d.pData, m_cpuBuf.get(), m_cpuSz); ctx->Unmap(res, 0); } m_validSlots |= slot; } } template void D3D11GraphicsBufferD::load(const void* data, size_t sz) { std::unique_lock lk(m_q->m_dynamicLock); size_t bufSz = std::min(sz, m_cpuSz); memcpy(m_cpuBuf.get(), data, bufSz); m_validSlots = 0; } template void* D3D11GraphicsBufferD::map(size_t sz) { if (sz > m_cpuSz) return nullptr; m_q->m_dynamicLock.lock(); return m_cpuBuf.get(); } template void D3D11GraphicsBufferD::unmap() { m_validSlots = 0; m_q->m_dynamicLock.unlock(); } void D3D11TextureD::update(ID3D11DeviceContext* ctx, int b) { int slot = 1 << b; if ((slot & m_validSlots) == 0) { ID3D11Texture2D* res = m_texs[b].Get(); D3D11_MAPPED_SUBRESOURCE d; ctx->Map(res, 0, D3D11_MAP_WRITE_DISCARD, 0, &d); size_t rowSz = m_pxPitch * m_width; for (size_t i = 0; i < m_cpuSz; i += rowSz, reinterpret_cast(d.pData) += d.RowPitch) memmove(d.pData, m_cpuBuf.get() + i, rowSz); ctx->Unmap(res, 0); m_validSlots |= slot; } } void D3D11TextureD::load(const void* data, size_t sz) { std::unique_lock lk(m_q->m_dynamicLock); size_t bufSz = std::min(sz, m_cpuSz); memcpy(m_cpuBuf.get(), data, bufSz); m_validSlots = 0; } void* D3D11TextureD::map(size_t sz) { if (sz > m_cpuSz) return nullptr; m_q->m_dynamicLock.lock(); return m_cpuBuf.get(); } void D3D11TextureD::unmap() { m_validSlots = 0; m_q->m_dynamicLock.unlock(); } class D3D11DataFactoryImpl : public D3D11DataFactory, public GraphicsDataFactoryHead { friend struct D3D11CommandQueue; friend class D3D11DataFactory::Context; IGraphicsContext* m_parent; struct D3D11Context* m_ctx; float m_gamma = 1.f; ObjToken m_gammaShader; ObjToken m_gammaLUT; ObjToken m_gammaVBO; ObjToken m_gammaBinding; void SetupGammaResources() { commitTransaction([this](IGraphicsDataFactory::Context& ctx) { auto vertexHlsl = D3D11DataFactory::CompileHLSL(GammaVS, PipelineStage::Vertex); auto vertexShader = ctx.newShaderStage(vertexHlsl, PipelineStage::Vertex); auto fragmentHlsl = D3D11DataFactory::CompileHLSL(GammaFS, PipelineStage::Fragment); auto fragmentShader = ctx.newShaderStage(fragmentHlsl, PipelineStage::Fragment); const std::array vfmt{{{VertexSemantic::Position4}, {VertexSemantic::UV4}}}; AdditionalPipelineInfo info = { BlendFactor::One, BlendFactor::Zero, Primitive::TriStrips, ZTest::None, false, true, false, CullMode::None}; m_gammaShader = ctx.newShaderPipeline(vertexShader, fragmentShader, VertexFormatInfo(vfmt.size(), vfmt.data()), info); m_gammaLUT = ctx.newDynamicTexture(256, 256, TextureFormat::I16, TextureClampMode::ClampToEdge); setDisplayGamma(1.f); struct Vert { std::array pos; std::array uv; }; constexpr std::array verts{{ {{-1.f, 1.f, 0.f, 1.f}, {0.f, 0.f, 0.f, 0.f}}, {{1.f, 1.f, 0.f, 1.f}, {1.f, 0.f, 0.f, 0.f}}, {{-1.f, -1.f, 0.f, 1.f}, {0.f, 1.f, 0.f, 0.f}}, {{1.f, -1.f, 0.f, 1.f}, {1.f, 1.f, 0.f, 0.f}}, }}; m_gammaVBO = ctx.newStaticBuffer(BufferUse::Vertex, verts.data(), sizeof(Vert), verts.size()); const std::array, 2> texs{{ {}, m_gammaLUT.get(), }}; m_gammaBinding = ctx.newShaderDataBinding(m_gammaShader, m_gammaVBO.get(), {}, {}, 0, nullptr, nullptr, texs.size(), texs.data(), nullptr, nullptr); return true; } BooTrace); } public: D3D11DataFactoryImpl(IGraphicsContext* parent, D3D11Context* ctx) : m_parent(parent), m_ctx(ctx) { UINT qLevels; while (SUCCEEDED(ctx->m_dev->CheckMultisampleQualityLevels(m_ctx->m_fbFormat, m_ctx->m_sampleCount, &qLevels)) && !qLevels) m_ctx->m_sampleCount = flp2(m_ctx->m_sampleCount - 1); } boo::ObjToken newPoolBuffer(BufferUse use, size_t stride, size_t count __BooTraceArgs) override { D3D11CommandQueue* q = static_cast(m_parent->getCommandQueue()); boo::ObjToken pool(new BaseGraphicsPool(*this __BooTraceArgsUse)); return {new D3D11GraphicsBufferD(pool, q, use, m_ctx, stride, count)}; } void commitTransaction(const FactoryCommitFunc& trans __BooTraceArgs) override { D3D11DataFactory::Context ctx(*this __BooTraceArgsUse); trans(ctx); } void setDisplayGamma(float gamma) override { if (m_ctx->m_fbFormat == DXGI_FORMAT_R16G16B16A16_FLOAT) m_gamma = gamma * 2.2f; else m_gamma = gamma; if (m_gamma != 1.f) UpdateGammaLUT(m_gammaLUT.get(), m_gamma); } bool isTessellationSupported(uint32_t& maxPatchSizeOut) override { maxPatchSizeOut = 32; return true; } void waitUntilShadersReady() override {} bool areShadersReady() override { return true; } }; void D3D11CommandQueue::generateMipmaps(const ObjToken& tex) { D3D11TextureCubeR* ctex = tex.cast(); m_deferredCtx->GenerateMips(ctex->m_colorSrv.Get()); } D3D11DataFactory::Context::Context(D3D11DataFactory& parent __BooTraceArgs) : m_parent(parent), m_data(new BaseGraphicsData(static_cast(parent) __BooTraceArgsUse)) {} D3D11DataFactory::Context::~Context() {} boo::ObjToken D3D11DataFactory::Context::newStaticBuffer(BufferUse use, const void* data, size_t stride, size_t count) { D3D11DataFactoryImpl& factory = static_cast(m_parent); return {new D3D11GraphicsBufferS(m_data, use, factory.m_ctx, data, stride, count)}; } boo::ObjToken D3D11DataFactory::Context::newDynamicBuffer(BufferUse use, size_t stride, size_t count) { D3D11DataFactoryImpl& factory = static_cast(m_parent); D3D11CommandQueue* q = static_cast(factory.m_parent->getCommandQueue()); return {new D3D11GraphicsBufferD(m_data, q, use, factory.m_ctx, stride, count)}; } boo::ObjToken D3D11DataFactory::Context::newStaticTexture(size_t width, size_t height, size_t mips, TextureFormat fmt, TextureClampMode clampMode, const void* data, size_t sz) { D3D11DataFactoryImpl& factory = static_cast(m_parent); return {new D3D11TextureS(m_data, factory.m_ctx, width, height, mips, fmt, data, sz)}; } boo::ObjToken D3D11DataFactory::Context::newStaticArrayTexture(size_t width, size_t height, size_t layers, size_t mips, TextureFormat fmt, TextureClampMode clampMode, const void* data, size_t sz) { D3D11DataFactoryImpl& factory = static_cast(m_parent); return {new D3D11TextureSA(m_data, factory.m_ctx, width, height, layers, mips, fmt, data, sz)}; } boo::ObjToken D3D11DataFactory::Context::newDynamicTexture(size_t width, size_t height, TextureFormat fmt, TextureClampMode clampMode) { D3D11DataFactoryImpl& factory = static_cast(m_parent); D3D11CommandQueue* q = static_cast(factory.m_parent->getCommandQueue()); return {new D3D11TextureD(m_data, q, factory.m_ctx, width, height, fmt)}; } boo::ObjToken D3D11DataFactory::Context::newRenderTexture(size_t width, size_t height, TextureClampMode clampMode, size_t colorBindCount, size_t depthBindCount) { D3D11DataFactoryImpl& factory = static_cast(m_parent); return {new D3D11TextureR(m_data, factory.m_ctx, width, height, factory.m_ctx->m_sampleCount, colorBindCount, depthBindCount)}; } ObjToken D3D11DataFactory::Context::newCubeRenderTexture(size_t width, size_t mips) { D3D11DataFactoryImpl& factory = static_cast(m_parent); return {new D3D11TextureCubeR(m_data, factory.m_ctx, width, mips)}; } boo::ObjToken D3D11DataFactory::Context::newShaderStage(const uint8_t* data, size_t size, PipelineStage stage) { D3D11DataFactoryImpl& factory = static_cast(m_parent); return {new D3D11ShaderStage(m_data, factory.m_ctx, data, size, stage)}; } boo::ObjToken D3D11DataFactory::Context::newShaderPipeline(ObjToken vertex, ObjToken fragment, ObjToken geometry, ObjToken control, ObjToken evaluation, const VertexFormatInfo& vtxFmt, const AdditionalPipelineInfo& additionalInfo, bool asynchronous) { D3D11DataFactoryImpl& factory = static_cast(m_parent); struct D3D11Context* ctx = factory.m_ctx; return { new D3D11ShaderPipeline(m_data, ctx, vertex, fragment, geometry, control, evaluation, vtxFmt, additionalInfo)}; } boo::ObjToken D3D11DataFactory::Context::newShaderDataBinding( const boo::ObjToken& pipeline, const boo::ObjToken& vbuf, const boo::ObjToken& instVbo, const boo::ObjToken& ibuf, size_t ubufCount, const boo::ObjToken* ubufs, const PipelineStage* ubufStages, const size_t* ubufOffs, const size_t* ubufSizes, size_t texCount, const boo::ObjToken* texs, const int* texBindIdx, const bool* depthBind, size_t baseVert, size_t baseInst) { D3D11DataFactoryImpl& factory = static_cast(m_parent); return {new D3D11ShaderDataBinding(m_data, factory.m_ctx, pipeline, vbuf, instVbo, ibuf, ubufCount, ubufs, ubufStages, ubufOffs, ubufSizes, texCount, texs, texBindIdx, depthBind, baseVert, baseInst)}; } void D3D11CommandQueue::RenderingWorker(D3D11CommandQueue* self) { { std::unique_lock lk(self->m_initmt); } self->m_initcv.notify_one(); D3D11DataFactoryImpl* dataFactory = static_cast(self->m_parent->getDataFactory()); while (self->m_running) { { std::unique_lock lk(self->m_mt); self->m_cv.wait(lk); if (!self->m_running) break; self->m_drawBuf = self->m_completeBuf; auto& CmdList = self->m_cmdLists[self->m_drawBuf]; self->ProcessDynamicLoads(self->m_ctx->m_devCtx.Get()); bool doReset = false; if (self->m_texResizes.size()) { for (const auto& resize : self->m_texResizes) resize.first->resize(self->m_ctx, resize.second.first, resize.second.second); self->m_texResizes.clear(); doReset = true; } if (self->m_cubeTexResizes.size()) { for (const auto& resize : self->m_cubeTexResizes) resize.first->resize(self->m_ctx, resize.second.first, resize.second.second); self->m_cubeTexResizes.clear(); doReset = true; } if (doReset) { CmdList.reset(); continue; } if (self->m_windowCtx->m_needsFSTransition) { if (self->m_windowCtx->m_fs) { self->m_windowCtx->m_swapChain->SetFullscreenState(true, nullptr); self->m_windowCtx->m_swapChain->ResizeTarget(&self->m_windowCtx->m_fsdesc); } else self->m_windowCtx->m_swapChain->SetFullscreenState(false, nullptr); self->m_windowCtx->m_needsFSTransition = false; CmdList.reset(); continue; } if (self->m_windowCtx->m_needsResize) { self->m_windowCtx->clearRTV(); self->m_windowCtx->m_swapChain->ResizeBuffers(2, self->m_windowCtx->width, self->m_windowCtx->height, self->m_ctx->m_fbFormat, DXGI_SWAP_CHAIN_FLAG_ALLOW_MODE_SWITCH); self->m_windowCtx->setupRTV(self->m_windowCtx->m_swapChain, self->m_ctx->m_dev.Get()); self->m_windowCtx->m_needsResize = false; CmdList.reset(); continue; } } auto& CmdList = self->m_cmdLists[self->m_drawBuf]; ID3D11CommandList* list = CmdList.list.Get(); self->m_ctx->m_devCtx->ExecuteCommandList(list, false); if (D3D11TextureR* csource = CmdList.workDoPresent.cast()) { #ifndef NDEBUG if (!csource->m_colorBindCount) Log.report(logvisor::Fatal, fmt("texture provided to resolveDisplay() must have at least 1 color binding")); #endif if (dataFactory->m_gamma != 1.f) { SWindowRect rect(0, 0, csource->m_width, csource->m_height); self->_resolveBindTexture(self->m_ctx->m_devCtx.Get(), csource, rect, true, 0, true, false); ID3D11RenderTargetView* rtv = self->m_windowCtx->m_swapChainRTV.Get(); self->m_ctx->m_devCtx->OMSetRenderTargets(1, &rtv, nullptr); const D3D11_VIEWPORT vp = {0.f, 0.f, FLOAT(csource->m_width), FLOAT(csource->m_height), 0.f, 1.f}; self->m_ctx->m_devCtx->RSSetViewports(1, &vp); const D3D11_RECT d3drect = {0, 0, LONG(csource->m_width), LONG(csource->m_height)}; self->m_ctx->m_devCtx->RSSetScissorRects(1, &d3drect); const std::array samp{ self->m_ctx->m_ss[0].Get(), self->m_ctx->m_ss[1].Get(), self->m_ctx->m_ss[2].Get(), self->m_ctx->m_ss[3].Get(), self->m_ctx->m_ss[4].Get(), }; self->m_ctx->m_devCtx->PSSetSamplers(0, UINT(samp.size()), samp.data()); self->m_ctx->m_devCtx->DSSetSamplers(0, UINT(samp.size()), samp.data()); D3D11ShaderDataBinding* gammaBinding = dataFactory->m_gammaBinding.cast(); gammaBinding->m_texs[0].tex = CmdList.workDoPresent.get(); gammaBinding->bind(self->m_ctx->m_devCtx.Get(), self->m_drawBuf); self->m_ctx->m_devCtx->Draw(4, 0); gammaBinding->m_texs[0].tex.reset(); } else { ComPtr dest = self->m_windowCtx->m_swapChainTex; ID3D11Texture2D* src = csource->m_colorTex.Get(); if (csource->m_samples > 1) self->m_ctx->m_devCtx->ResolveSubresource(dest.Get(), 0, src, 0, self->m_ctx->m_fbFormat); else self->m_ctx->m_devCtx->CopyResource(dest.Get(), src); } self->m_windowCtx->m_swapChain->Present(1, 0); } CmdList.reset(); } } void D3D11CommandQueue::startRenderer() { static_cast(m_parent->getDataFactory())->SetupGammaResources(); } void D3D11CommandQueue::execute() { /* Finish command list */ auto& CmdList = m_cmdLists[m_fillBuf]; ThrowIfFailed(m_deferredCtx->FinishCommandList(false, &CmdList.list)); CmdList.workDoPresent = m_doPresent; m_doPresent = nullptr; /* Wait for worker thread to become ready */ std::unique_lock lk(m_mt); /* Ready for next frame */ m_completeBuf = m_fillBuf; for (size_t i = 0; i < m_cmdLists.size(); ++i) { if (i == m_completeBuf || i == m_drawBuf) { continue; } m_fillBuf = i; break; } /* Return control to worker thread */ lk.unlock(); m_cv.notify_one(); } void D3D11CommandQueue::ProcessDynamicLoads(ID3D11DeviceContext* ctx) { D3D11DataFactoryImpl* gfxF = static_cast(m_parent->getDataFactory()); std::unique_lock lk(m_dynamicLock); std::unique_lock datalk(gfxF->m_dataMutex); if (gfxF->m_dataHead) { for (BaseGraphicsData& d : *gfxF->m_dataHead) { if (d.m_DBufs) for (IGraphicsBufferD& b : *d.m_DBufs) static_cast&>(b).update(ctx, m_drawBuf); if (d.m_DTexs) for (ITextureD& t : *d.m_DTexs) static_cast(t).update(ctx, m_drawBuf); } } if (gfxF->m_poolHead) { for (BaseGraphicsPool& p : *gfxF->m_poolHead) { if (p.m_DBufs) for (IGraphicsBufferD& b : *p.m_DBufs) static_cast&>(b).update(ctx, m_drawBuf); } } } std::unique_ptr _NewD3D11CommandQueue(D3D11Context* ctx, D3D11Context::Window* windowCtx, IGraphicsContext* parent) { return std::make_unique(ctx, windowCtx, parent); } std::unique_ptr _NewD3D11DataFactory(D3D11Context* ctx, IGraphicsContext* parent) { return std::make_unique(parent, ctx); } constexpr std::array D3DShaderTypes{ nullptr, "vs_5_0", "ps_5_0", "gs_5_0", "hs_5_0", "ds_5_0", }; #if _DEBUG && 0 #define BOO_D3DCOMPILE_FLAG D3DCOMPILE_DEBUG | D3DCOMPILE_OPTIMIZATION_LEVEL0 #else #define BOO_D3DCOMPILE_FLAG D3DCOMPILE_OPTIMIZATION_LEVEL3 #endif std::vector D3D11DataFactory::CompileHLSL(const char* source, PipelineStage stage) { ComPtr errBlob; ComPtr blobOut; if (FAILED(D3DCompilePROC(source, strlen(source), "Boo HLSL Source", nullptr, nullptr, "main", D3DShaderTypes[int(stage)], BOO_D3DCOMPILE_FLAG, 0, &blobOut, &errBlob))) { fmt::print(fmt("{}\n"), source); Log.report(logvisor::Fatal, fmt("error compiling shader: {}"), errBlob->GetBufferPointer()); return {}; } std::vector ret(blobOut->GetBufferSize()); memcpy(ret.data(), blobOut->GetBufferPointer(), ret.size()); return ret; } } // namespace boo