perf: 特定情况下检查重复帧不执行边界检查

Author: Blinue

src/Magpie.Core/DuplicateFrameChecker.cpp

@@ -4,6 +4,7 @@
 #include "Logger.h"
 #include "ScalingWindow.h"
 #include "shaders/DuplicateFrameCS.h"
+#include "shaders/DuplicateFrameCS_NoBoundsChecking.h"
 
 namespace Magpie {
 
@@ -18,12 +19,16 @@ DuplicateFrameChecker::DuplicateFrameChecker() noexcept :
 // 1. D3D11 支持 IDXGIDevice::SetGPUThreadPriority，可以提高 GPU 优先级，
 // 而 D3D12 没有等价接口。
 // 2. 对于小任务 D3D11 启动渲染的耗时比 D3D12 短，差距可以达到 50us 以上。
+// 
+// 对于不支持脏矩形且捕获帧右下两边没有多余像素的捕获方式，可以禁用边界检查获得
+// 性能提升。
 bool DuplicateFrameChecker::Initialize(
 	ID3D11Device5* d3d11Device,
 	ID3D11DeviceContext4* d3d11DC,
 	const ColorInfo& colorInfo,
 	Size frameSize,
-	uint32_t frameCount
+	uint32_t frameCount,
+	bool disableBoundsChecking
 ) noexcept {
 	assert(ScalingWindow::Get().Options().duplicateFrameDetectionMode !=
 		DuplicateFrameDetectionMode::Never);
@@ -32,11 +37,18 @@ bool DuplicateFrameChecker::Initialize(
 	_deviceContext = d3d11DC;
 	_isScRGB = colorInfo.kind != winrt::AdvancedColorKind::StandardDynamicRange;
 	_frameSize = frameSize;
+#ifdef _DEBUG
+	_isBoundsCheckingDisabled = disableBoundsChecking;
+#endif
 
 	_frameSrvs.resize(frameCount);
 
 	HRESULT hr = d3d11Device->CreateComputeShader(
-		DuplicateFrameCS, sizeof(DuplicateFrameCS), nullptr, _dupFrameCS.put());
+		disableBoundsChecking ? DuplicateFrameCS_NoBoundsChecking : DuplicateFrameCS,
+		disableBoundsChecking ? sizeof(DuplicateFrameCS_NoBoundsChecking) : sizeof(DuplicateFrameCS),
+		nullptr,
+		_dupFrameCS.put()
+	);
 	if (FAILED(hr)) {
 		Logger::Get().ComError("CreateComputeShader 失败", hr);
 		return false;
@@ -134,6 +146,13 @@ HRESULT DuplicateFrameChecker::CheckFrame(
 		D3D11_TEXTURE2D_DESC desc;
 		frameResource->GetDesc(&desc);
 		assert(desc.Width == _frameSize.width && desc.Height == _frameSize.height);
+
+		if (_isBoundsCheckingDisabled) {
+			// 确保捕获帧右下两边没有多余像素
+			for (const Rect& rect : dirtyRects) {
+				assert(rect.right == desc.Width && rect.bottom == desc.Height);
+			}
+		}
 	}
 #endif
 

src/Magpie.Core/DuplicateFrameChecker.h

@@ -16,7 +16,8 @@ class DuplicateFrameChecker {
 		ID3D11DeviceContext4* d3d11DC,
 		const ColorInfo& colorInfo,
 		Size frameSize,
-		uint32_t frameCount
+		uint32_t frameCount,
+		bool disableBoundsChecking
 	) noexcept;
 
 	HRESULT CheckFrame(ID3D11Texture2D* frameResource, uint32_t frameIdx, SmallVectorImpl<Rect>& dirtyRects) noexcept;
@@ -49,6 +50,9 @@ class DuplicateFrameChecker {
 	uint16_t _framesLeft;
 
 	bool _isScRGB = false;
+#ifdef _DEBUG
+	bool _isBoundsCheckingDisabled = false;
+#endif
 	bool _isCheckingForDuplicateFrame = true;
 };
 

src/Magpie.Core/GraphicsCaptureFrameSource.cpp

@@ -261,8 +261,9 @@ bool GraphicsCaptureFrameSource::Initialize(
 
 	if (options.duplicateFrameDetectionMode != DuplicateFrameDetectionMode::Never) {
 		_duplicateFrameChecker = std::make_unique<DuplicateFrameChecker>();
+		// 不使用脏区域时可以跳过边界检查，因为捕获帧右下两边没有多余像素
 		if (!_duplicateFrameChecker->Initialize(_d3d11Device.get(), _d3d11DC.get(),
-			colorInfo, Size{ _frameBox.right, _frameBox.bottom }, frameCount)) {
+			colorInfo, Size{ _frameBox.right, _frameBox.bottom }, frameCount, !_isDirtyRegionSupported)) {
 			Logger::Get().Error("DuplicateFrameChecker::Initialize 失败");
 			return false;
 		}
@@ -1153,7 +1154,7 @@ HRESULT GraphicsCaptureFrameSource::_StartCapture() noexcept {
 	assert(!_captureFramePool && !_captureSession);
 
 	try {
-		// 创建帧缓冲池。帧的尺寸为包含源窗口的最小尺寸
+		// 创建帧缓冲池。帧的尺寸为包含捕获区域的最小尺寸，既能降低显存压力，又使得检查重复帧时无需边界检查
 		_captureFramePool = winrt::Direct3D11CaptureFramePool::CreateFreeThreaded(
 			_wrappedDevice,
 			_isScRGB ? winrt::DirectXPixelFormat::R16G16B16A16Float : winrt::DirectXPixelFormat::B8G8R8A8UIntNormalized,

src/Magpie.Core/Magpie.Core.vcxproj

@@ -156,6 +156,9 @@
     <FxCompile Include="shaders\DuplicateFrameCS.hlsl">
       <ShaderType>Compute</ShaderType>
     </FxCompile>
+    <FxCompile Include="shaders\DuplicateFrameCS_NoBoundsChecking.hlsl">
+      <ShaderType>Compute</ShaderType>
+    </FxCompile>
     <FxCompile Include="shaders\ImGuiImplPS.hlsl">
       <ShaderType>Pixel</ShaderType>
     </FxCompile>

src/Magpie.Core/Magpie.Core.vcxproj.filters

@@ -309,6 +309,9 @@
     <FxCompile Include="CatmullRomCS_sRGB.hlsl">
       <Filter>Shaders</Filter>
     </FxCompile>
+    <FxCompile Include="shaders\DuplicateFrameCS_NoBoundsChecking.hlsl">
+      <Filter>Shaders</Filter>
+    </FxCompile>
   </ItemGroup>
   <ItemGroup>
     <None Include="packages.config" />

src/Magpie.Core/shaders/DuplicateFrameCS.hlsl

@@ -20,12 +20,30 @@ void main(uint3 tid : SV_GroupThreadID, uint3 gid : SV_GroupID) {
 	}
 
 	const uint2 gxy = dirtyRect.xy + (gid.xy << 4) + (tid.xy << 1);
+	
+#ifndef MP_NO_BOUNDS_CHECKING
 	if (gxy.x >= dirtyRect.z || gxy.y >= dirtyRect.w) {
 		return;
 	}
+#endif
 
 	const float2 pos = (gxy + 1) * texPt;
+
+#ifdef MP_NO_BOUNDS_CHECKING
+	if (any(tex1.GatherRed(sam, pos) != tex2.GatherRed(sam, pos))) {
+        result[resultOffset] = target;
+		return;
+	}
+	
+	if (any(tex1.GatherGreen(sam, pos) != tex2.GatherGreen(sam, pos))) {
+        result[resultOffset] = target;
+		return;
+	}
 
+	if (any(tex1.GatherBlue(sam, pos) != tex2.GatherBlue(sam, pos))) {
+        result[resultOffset] = target;
+    }
+#else
 	// w z
 	// x y
 	float4 mask = 1.0f;
@@ -57,4 +75,5 @@ void main(uint3 tid : SV_GroupThreadID, uint3 gid : SV_GroupID) {
 		result[resultOffset] = target;
 		return;
 	}
+#endif
 }

src/Magpie.Core/shaders/DuplicateFrameCS_NoBoundsChecking.hlsl

@@ -0,0 +1,2 @@
+#define MP_NO_BOUNDS_CHECKING
+#include "DuplicateFrameCS.hlsl"