Platform toolset143 compatibility

Samples/BulkLoadDemo/BulkLoadDemo/BulkLoadDemo.vcxproj

@@ -23,12 +23,12 @@
     <PlatformToolset>v142</PlatformToolset>
     <MinimumVisualStudioVersion>16.0</MinimumVisualStudioVersion>
     <TargetRuntime>Native</TargetRuntime>
-    <WindowsTargetPlatformVersion>10.0.19041.0</WindowsTargetPlatformVersion>
+    <WindowsTargetPlatformVersion>10.0</WindowsTargetPlatformVersion>
   </PropertyGroup>
   <Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
   <PropertyGroup Label="Configuration">
     <ConfigurationType>Application</ConfigurationType>
-    <PlatformToolset>v142</PlatformToolset>
+    <PlatformToolset>v143</PlatformToolset>
     <CharacterSet>Unicode</CharacterSet>
     <EmbedManifest>false</EmbedManifest>
   </PropertyGroup>

Samples/BulkLoadDemo/Core/Core.vcxproj

@@ -20,13 +20,13 @@
     <DefaultLanguage>en-US</DefaultLanguage>
     <Keyword>Win32Proj</Keyword>
     <MinimumVisualStudioVersion>16.0</MinimumVisualStudioVersion>
-    <WindowsTargetPlatformVersion>10.0.19041.0</WindowsTargetPlatformVersion>
+    <WindowsTargetPlatformVersion>10.0</WindowsTargetPlatformVersion>
     <TargetRuntime>Native</TargetRuntime>
   </PropertyGroup>
   <Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
   <PropertyGroup Label="Configuration">
     <ConfigurationType>StaticLibrary</ConfigurationType>
-    <PlatformToolset>v142</PlatformToolset>
+    <PlatformToolset>v143</PlatformToolset>
     <CharacterSet>Unicode</CharacterSet>
     <EmbedManifest>false</EmbedManifest>
     <GenerateManifest>false</GenerateManifest>

Samples/BulkLoadDemo/Core/Shaders/BicubicFilterFunctions.hlsli

@@ -31,7 +31,7 @@ float4 GetBicubicFilterWeights(float offset, float A)
     //return ComputeWeights(offset, A);
 
     // Precompute weights for 16 discrete offsets
-    static const float4 FilterWeights[16] =
+    const float4 FilterWeights[16] =
     {
         ComputeWeights( 0.5 / 16.0, -0.5),
         ComputeWeights( 1.5 / 16.0, -0.5),

Samples/BulkLoadDemo/Core/Shaders/BicubicHorizontalUpsamplePS.hlsl

@@ -41,7 +41,7 @@ float3 GetColor(uint s, uint t)
 }
 
 [RootSignature(Present_RootSig)]
-float4 main(float4 position : SV_Position, float2 uv : TexCoord0) : SV_Target0
+float3 main(float4 position : SV_Position, float2 uv : TexCoord0) : SV_Target0
 {
     float2 t = uv * TextureSize + 0.5;
     float2 f = frac(t);
@@ -61,5 +61,5 @@ float4 main(float4 position : SV_Position, float2 uv : TexCoord0) : SV_Target0
         W.z * GetColor(s2, st.y) +
         W.w * GetColor(s3, st.y);
 
-    return float4(Color, 1);
+    return Color;
 }

Samples/BulkLoadDemo/Core/Shaders/BicubicUpsamplePS.hlsl

@@ -61,7 +61,7 @@ float3 GetColor(uint s, uint t)
 }
 
 [RootSignature(Present_RootSig)]
-float4 main(float4 position : SV_Position, float2 uv : TexCoord0) : SV_Target0
+float3 main(float4 position : SV_Position, float2 uv : TexCoord0) : SV_Target0
 {
     float2 t = uv * TextureSize + 0.5;
     float2 f = frac(t);
@@ -88,8 +88,8 @@ float4 main(float4 position : SV_Position, float2 uv : TexCoord0) : SV_Target0
     float3 Color = Cubic(GetWeights(f.y), c0, c1, c2, c3);
 
 #ifdef GAMMA_SPACE
-    return float4(Color, 1);
+    return Color;
 #else
-    return float4(ApplyDisplayProfile(Color, DISPLAY_PLANE_FORMAT), 1);
+    return ApplyDisplayProfile(Color, DISPLAY_PLANE_FORMAT);
 #endif
 }

Samples/BulkLoadDemo/Core/Shaders/BicubicVerticalUpsamplePS.hlsl

@@ -37,7 +37,7 @@ float3 GetColor(uint s, uint t)
 }
 
 [RootSignature(Present_RootSig)]
-float4 main(float4 position : SV_Position, float2 uv : TexCoord0) : SV_Target0
+float3 main(float4 position : SV_Position, float2 uv : TexCoord0) : SV_Target0
 {
     float2 t = uv * TextureSize + 0.5;
     float2 f = frac(t);
@@ -58,8 +58,8 @@ float4 main(float4 position : SV_Position, float2 uv : TexCoord0) : SV_Target0
         W.w * GetColor(st.x, t3);
 
 #ifdef GAMMA_SPACE
-        return float4(Color,1);
+        return Color;
 #else
-        return float4(ApplyDisplayProfile(Color, DISPLAY_PLANE_FORMAT), 1);
+        return ApplyDisplayProfile(Color, DISPLAY_PLANE_FORMAT);
 #endif
 }

Samples/BulkLoadDemo/Core/Shaders/BilinearUpsamplePS.hlsl

@@ -18,8 +18,8 @@ Texture2D<float3> ColorTex : register(t0);
 SamplerState BilinearFilter : register(s0);
 
 [RootSignature(Present_RootSig)]
-float4 main( float4 position : SV_Position, float2 uv : TexCoord0 ) : SV_Target0
+float3 main( float4 position : SV_Position, float2 uv : TexCoord0 ) : SV_Target0
 {
     float3 LinearRGB = RemoveDisplayProfile(ColorTex.SampleLevel(BilinearFilter, uv, 0), LDR_COLOR_FORMAT);
-    return float4(ApplyDisplayProfile(LinearRGB, DISPLAY_PLANE_FORMAT), 1);
+    return ApplyDisplayProfile(LinearRGB, DISPLAY_PLANE_FORMAT);
 }

Samples/BulkLoadDemo/Core/Shaders/BlurCS.hlsl

@@ -34,7 +34,7 @@ float3 BlurPixels( float3 a, float3 b, float3 c, float3 d, float3 e, float3 f, f
     return Weights[0]*e + Weights[1]*(d+f) + Weights[2]*(c+g) + Weights[3]*(b+h) + Weights[4]*(a+i);
 }
 
-// 16x16 pixels with an 8x8 center that we will be blurring writing out.  Each uint is two color channels packed together
+// 16x16 pixels with an 8x8 center that we will be blurring writing out. Each uint is two color channels packed together
 groupshared uint CacheR[128];
 groupshared uint CacheG[128];
 groupshared uint CacheB[128];
@@ -67,7 +67,7 @@ void Load1Pixel( uint index, out float3 pixel )
     pixel = asfloat( uint3(CacheR[index], CacheG[index], CacheB[index]) );
 }
 
-// Blur two pixels horizontally.  This reduces LDS reads and pixel unpacking.
+// Blur two pixels horizontally. This reduces LDS reads and pixel unpacking.
 void BlurHorizontally( uint outIndex, uint leftMostIndex )
 {
     float3 s0, s1, s2, s3, s4, s5, s6, s7, s8, s9;
@@ -77,6 +77,9 @@ void BlurHorizontally( uint outIndex, uint leftMostIndex )
     Load2Pixels( leftMostIndex + 3, s6, s7 );
     Load2Pixels( leftMostIndex + 4, s8, s9 );
 
+    // Be sure to finish loading values before we rewrite them.
+    GroupMemoryBarrierWithGroupSync();
+
     Store1Pixel(outIndex  , BlurPixels(s0, s1, s2, s3, s4, s5, s6, s7, s8));
     Store1Pixel(outIndex+1, BlurPixels(s1, s2, s3, s4, s5, s6, s7, s8, s9));
 }
@@ -104,8 +107,8 @@ void main( uint3 Gid : SV_GroupID, uint3 GTid : SV_GroupThreadID, uint3 DTid : S
     //
     // Load 4 pixels per thread into LDS
     //
-    int2 GroupUL = (Gid.xy << 3) - 4;                // Upper-left pixel coordinate of group read location
-    int2 ThreadUL = (GTid.xy << 1) + GroupUL;        // Upper-left pixel coordinate of quad that this thread will read
+    int2 GroupUL = (Gid.xy << 3) - 4;           // Upper-left pixel coordinate of group read location
+    int2 ThreadUL = (GTid.xy << 1) + GroupUL;   // Upper-left pixel coordinate of quad that this thread will read
 
     //
     // Store 4 unblurred pixels in LDS

Samples/BulkLoadDemo/Core/Shaders/ColorSpaceUtility.hlsli

@@ -37,37 +37,37 @@
 float3 ApplySRGBCurve( float3 x )
 {
     // Approximately pow(x, 1.0 / 2.2)
-    return x < 0.0031308 ? 12.92 * x : 1.055 * pow(x, 1.0 / 2.4) - 0.055;
+    return select(x < 0.0031308, 12.92 * x, 1.055 * pow(x, 1.0 / 2.4) - 0.055);
 }
 
 float3 RemoveSRGBCurve( float3 x )
 {
     // Approximately pow(x, 2.2)
-    return x < 0.04045 ? x / 12.92 : pow( (x + 0.055) / 1.055, 2.4 );
+    return select(x < 0.04045, x / 12.92, pow( (x + 0.055) / 1.055, 2.4 ));
 }
 
 // These functions avoid pow() to efficiently approximate sRGB with an error < 0.4%.
 float3 ApplySRGBCurve_Fast( float3 x )
 {
-    return x < 0.0031308 ? 12.92 * x : 1.13005 * sqrt(x - 0.00228) - 0.13448 * x + 0.005719;
+    return select(x < 0.0031308, 12.92 * x, 1.13005 * sqrt(x - 0.00228) - 0.13448 * x + 0.005719);
 }
 
 float3 RemoveSRGBCurve_Fast( float3 x )
 {
-    return x < 0.04045 ? x / 12.92 : -7.43605 * x - 31.24297 * sqrt(-0.53792 * x + 1.279924) + 35.34864;
+    return select(x < 0.04045, x / 12.92, -7.43605 * x - 31.24297 * sqrt(-0.53792 * x + 1.279924) + 35.34864);
 }
 
 // The OETF recommended for content shown on HDTVs.  This "gamma ramp" may increase contrast as
 // appropriate for viewing in a dark environment.  Always use this curve with Limited RGB as it is
 // used in conjunction with HDTVs.
 float3 ApplyREC709Curve( float3 x )
 {
-    return x < 0.0181 ? 4.5 * x : 1.0993 * pow(x, 0.45) - 0.0993;
+    return select(x < 0.0181, 4.5 * x, 1.0993 * pow(x, 0.45) - 0.0993);
 }
 
 float3 RemoveREC709Curve( float3 x )
 {
-    return x < 0.08145 ? x / 4.5 : pow((x + 0.0993) / 1.0993, 1.0 / 0.45);
+    return select(x < 0.08145, x / 4.5, pow((x + 0.0993) / 1.0993, 1.0 / 0.45));
 }
 
 // This is the new HDR transfer function, also called "PQ" for perceptual quantizer.  Note that REC2084

Samples/BulkLoadDemo/Core/Shaders/CompositeHDRPS.hlsl

@@ -24,7 +24,7 @@ cbuffer CB0 : register(b0)
 }
 
 [RootSignature(Present_RootSig)]
-float4 main( float4 position : SV_Position, float2 uv : TexCoord0 ) : SV_Target0
+float3 main( float4 position : SV_Position, float2 uv : TexCoord0 ) : SV_Target0
 {
     int2 ST = (int2)position.xy;
 
@@ -36,5 +36,5 @@ float4 main( float4 position : SV_Position, float2 uv : TexCoord0 ) : SV_Target0
     OverlayColor.rgb = REC709toREC2020(OverlayColor.rgb / (OverlayColor.a == 0.0 ? 1.0 : OverlayColor.a));
     OverlayColor.rgb = ApplyREC2084Curve(OverlayColor.rgb * PaperWhiteRatio);
 
-    return float4(lerp(MainColor, OverlayColor.rgb, OverlayColor.a), 1);
+    return lerp(MainColor, OverlayColor.rgb, OverlayColor.a);
 }

Samples/BulkLoadDemo/Core/Shaders/CompositeSDRPS.hlsl

@@ -18,9 +18,9 @@ Texture2D<float3> MainBuffer : register(t0);
 Texture2D<float4> OverlayBuffer : register(t1);
 
 [RootSignature(Present_RootSig)]
-float4 main( float4 position : SV_Position ) : SV_Target0
+float3 main( float4 position : SV_Position ) : SV_Target0
 {
     float3 MainColor = ApplyDisplayProfile(MainBuffer[(int2)position.xy], DISPLAY_PLANE_FORMAT);
     float4 OverlayColor = OverlayBuffer[(int2)position.xy];
-    return float4(OverlayColor.rgb + MainColor.rgb * (1.0 - OverlayColor.a), 1);
+    return OverlayColor.rgb + MainColor.rgb * (1.0 - OverlayColor.a);
 }

Samples/BulkLoadDemo/Core/Shaders/DoFPass1CS.hlsl

@@ -37,21 +37,24 @@ void main( uint3 Gid : SV_GroupID, uint GI : SV_GroupIndex, uint3 GTid : SV_Grou
     float TileMaxDepth = Max4(Depths);
     float TileMaxCoC = MaxCoC(Depths);
 
+    // Write and sync
+    gs_ClosestDepthSearch[GI] = TileMinDepth;
+    gs_FarthestDepthSearch[GI] = TileMaxDepth;
+    gs_MaximumCoC[GI] = TileMaxCoC;
+    GroupMemoryBarrierWithGroupSync();
+
     for (uint i = 32; i > 0; i >>= 1)
     {
-        // Write and sync
-        gs_ClosestDepthSearch[GI] = TileMinDepth;
-        gs_FarthestDepthSearch[GI] = TileMaxDepth;
-        gs_MaximumCoC[GI] = TileMaxCoC;
-        GroupMemoryBarrierWithGroupSync();
-
         // Read and sync
-        TileMinDepth = min(TileMinDepth, gs_ClosestDepthSearch[(GI + i) % 64]);
-        TileMaxDepth = max(TileMaxDepth, gs_FarthestDepthSearch[(GI + i) % 64]);
-        TileMaxCoC = max(TileMaxCoC, gs_MaximumCoC[(GI + i) % 64]);
+        if (GI < i)
+        {
+            gs_ClosestDepthSearch[i] = min(gs_ClosestDepthSearch[i], gs_ClosestDepthSearch[GI + i]);
+            gs_FarthestDepthSearch[i] = max(gs_FarthestDepthSearch[i], gs_FarthestDepthSearch[GI + i]);
+            gs_MaximumCoC[i] = max(gs_MaximumCoC[i], gs_MaximumCoC[GI + i]);
+        }
         GroupMemoryBarrierWithGroupSync();
     }
 
     if (GI == 0)
-        TileClass[Gid.xy] = float3(TileMaxCoC, TileMinDepth, TileMaxDepth);
+        TileClass[Gid.xy] = float3(gs_MaximumCoC[0], gs_FarthestDepthSearch[0], gs_FarthestDepthSearch[0]);
 }

Samples/BulkLoadDemo/Core/Shaders/GenerateMipsCS.hlsli

@@ -55,10 +55,10 @@ float4 LoadColor( uint Index )
 float3 ApplySRGBCurve(float3 x)
 {
     // This is exactly the sRGB curve
-    //return x < 0.0031308 ? 12.92 * x : 1.055 * pow(abs(x), 1.0 / 2.4) - 0.055;
+    //return select(x < 0.0031308, 12.92 * x, 1.055 * pow(abs(x), 1.0 / 2.4) - 0.055);
 
     // This is cheaper but nearly equivalent
-    return x < 0.0031308 ? 12.92 * x : 1.13005 * sqrt(abs(x - 0.00228)) - 0.13448 * x + 0.005719;
+    return select(x < 0.0031308, 12.92 * x, 1.13005 * sqrt(abs(x - 0.00228)) - 0.13448 * x + 0.005719);
 }
 
 float4 PackColor(float4 Linear)

Samples/BulkLoadDemo/Core/Shaders/LanczosFunctions.hlsli

@@ -41,7 +41,7 @@ float4 GetUpscaleFilterWeights(float offset)
     //return ComputeWeights(offset);
 
     // Precompute weights for 16 discrete offsets
-    static const float4 FilterWeights[16] =
+    const float4 FilterWeights[16] =
     {
         ComputeWeights( 0.5 / 16.0),
         ComputeWeights( 1.5 / 16.0),

Samples/BulkLoadDemo/Core/Shaders/LanczosHorizontalPS.hlsl

@@ -30,7 +30,7 @@ float4x3 LoadSamples(int2 ST, uint2 Stride)
 }
 
 [RootSignature(Present_RootSig)]
-float4 main(float4 Pos : SV_Position, float2 UV : TexCoord0) : SV_Target0
+float3 main(float4 Pos : SV_Position, float2 UV : TexCoord0) : SV_Target0
 {
     // We subtract 0.5 because that represents the center of the pixel.  We need to know where
     // we lie between two pixel centers, and we will use frac() for that.  We subtract another
@@ -53,5 +53,5 @@ float4 main(float4 Pos : SV_Position, float2 UV : TexCoord0) : SV_Target0
     Result = ApplyDisplayProfile(Result, DISPLAY_PLANE_FORMAT);
 #endif
 
-    return float4(Result, 1);
+    return Result;
 }

Samples/BulkLoadDemo/Core/Shaders/MagnifyPixelsPS.hlsl

@@ -23,8 +23,8 @@ cbuffer Constants : register(b0)
 }
 
 [RootSignature(Present_RootSig)]
-float4 main( float4 position : SV_Position, float2 uv : TexCoord0 ) : SV_Target0
+float3 main( float4 position : SV_Position, float2 uv : TexCoord0 ) : SV_Target0
 {
     float2 ScaledUV = ScaleFactor * (uv - 0.5) + 0.5;
-    return float4(ColorTex.SampleLevel(PointSampler, ScaledUV, 0), 1);
+    return ColorTex.SampleLevel(PointSampler, ScaledUV, 0);
 }

Samples/BulkLoadDemo/Core/Shaders/ParticleTileCullingCS.hlsl

@@ -106,7 +106,7 @@ void main( uint3 Gid : SV_GroupID, uint GI : SV_GroupIndex, uint3 GTid : SV_Grou
     ParticleCountInBin = min(MAX_PARTICLES_PER_BIN, ParticleCountInBin);
 
     // Compute the next power of two for the bitonic sort
-    uint NextPow2 = countbits(ParticleCountInBin) <= 1u ? ParticleCountInBin : (2u << firstbithigh(ParticleCountInBin));
+    uint NextPow2 = countbits(ParticleCountInBin) <= 1 ? ParticleCountInBin : (2u << firstbithigh(ParticleCountInBin));
 
     // Fill in the sort key array.  Each sort key has passenger data (in the least signficant
     // bits, so that as the sort keys are moved around, they retain a pointer to the particle

Samples/BulkLoadDemo/Core/Shaders/PresentHDRPS.hlsl

@@ -17,7 +17,7 @@
 Texture2D<float3> MainBuffer : register(t0);
 
 [RootSignature(Present_RootSig)]
-float4 main( float4 position : SV_Position, float2 uv : TexCoord0 ) : SV_Target0
+float3 main( float4 position : SV_Position, float2 uv : TexCoord0 ) : SV_Target0
 {
-    return float4(ApplyREC2084Curve(MainBuffer[(int2)position.xy] / 10000.0), 1);
+    return ApplyREC2084Curve(MainBuffer[(int2)position.xy] / 10000.0);
 }

Samples/BulkLoadDemo/Core/Shaders/PresentSDRPS.hlsl

@@ -17,8 +17,8 @@
 Texture2D<float3> ColorTex : register(t0);
 
 [RootSignature(Present_RootSig)]
-float4 main( float4 position : SV_Position ) : SV_Target0
+float3 main( float4 position : SV_Position ) : SV_Target0
 {
     float3 LinearRGB = ColorTex[(int2)position.xy];
-    return float4(ApplyDisplayProfile(LinearRGB, DISPLAY_PLANE_FORMAT), 1);
+    return ApplyDisplayProfile(LinearRGB, DISPLAY_PLANE_FORMAT);
 }

Samples/BulkLoadDemo/Core/Shaders/ScaleAndCompositeHDRPS.hlsl

@@ -42,14 +42,14 @@ float3 ScaleBuffer(float2 uv)
 }
 
 [RootSignature(Present_RootSig)]
-float4 main( float4 position : SV_Position, float2 uv : TexCoord0 ) : SV_Target0
+float3 main( float4 position : SV_Position, float2 uv : TexCoord0 ) : SV_Target0
 {
-    float3 MainColor = ApplyREC2084Curve(ScaleBuffer(uv) / 10000.0);
+    float3 MainColor = ApplyREC2084Curve( saturate(ScaleBuffer(uv) / 10000.0) );
 
     float4 OverlayColor = OverlayBuffer[(int2)position.xy];
     OverlayColor.rgb = RemoveSRGBCurve(OverlayColor.rgb);
     OverlayColor.rgb = REC709toREC2020(OverlayColor.rgb / (OverlayColor.a == 0.0 ? 1.0 : OverlayColor.a));
     OverlayColor.rgb = ApplyREC2084Curve(OverlayColor.rgb * PaperWhiteRatio);
 
-    return float4(lerp(MainColor, OverlayColor.rgb, OverlayColor.a), 1);
+    return lerp(MainColor, OverlayColor.rgb, OverlayColor.a);
 }

Samples/BulkLoadDemo/Core/Shaders/ScaleAndCompositeSDRPS.hlsl

@@ -40,9 +40,9 @@ float3 ScaleBuffer(float2 uv)
 }
 
 [RootSignature(Present_RootSig)]
-float4 main( float4 position : SV_Position, float2 uv : TexCoord0 ) : SV_Target0
+float3 main( float4 position : SV_Position, float2 uv : TexCoord0 ) : SV_Target0
 {
     float3 MainColor = ApplyDisplayProfile(ScaleBuffer(uv), DISPLAY_PLANE_FORMAT);
     float4 OverlayColor = OverlayBuffer[(int2)position.xy];
-    return float4(OverlayColor.rgb + MainColor.rgb * (1.0 - OverlayColor.a), 1);
+    return OverlayColor.rgb + MainColor.rgb * (1.0 - OverlayColor.a);
 }

Samples/BulkLoadDemo/Core/Shaders/SharpeningUpsamplePS.hlsl

@@ -37,15 +37,15 @@ float3 GetColor(float2 UV)
 }
 
 [RootSignature(Present_RootSig)]
-float4 main(float4 position : SV_Position, float2 uv : TexCoord0) : SV_Target0
+float3 main(float4 position : SV_Position, float2 uv : TexCoord0) : SV_Target0
 {
     float3 Color = WB * GetColor(uv) - WA * (
         GetColor(uv + UVOffset0) + GetColor(uv - UVOffset0) +
         GetColor(uv + UVOffset1) + GetColor(uv - UVOffset1));
 
 #ifdef GAMMA_SPACE
-    return float4(Color, 1);
+    return Color;
 #else
-    return float4(ApplyDisplayProfile(Color, DISPLAY_PLANE_FORMAT), 1);
+    return ApplyDisplayProfile(Color, DISPLAY_PLANE_FORMAT);
 #endif
 }

Samples/BulkLoadDemo/Core/Shaders/TemporalBlendCS.hlsl

@@ -194,8 +194,6 @@ void main(uint3 DTid : SV_DispatchThreadID, uint GI : SV_GroupIndex, uint3 GTid
     uint Idx0 = GTid.x * 2 + GTid.y * kLdsPitch + kLdsPitch + 1;
     uint Idx1 = Idx0 + 1;
 
-    GroupMemoryBarrierWithGroupSync();
-
     float3 BoxMin, BoxMax;
     GetBBoxForPair(Idx0, Idx1, BoxMin, BoxMax);
 

Samples/BulkLoadDemo/Core/Shaders/ToneMappingUtility.hlsli

@@ -30,7 +30,7 @@ float3 TM_Reinhard(float3 hdr, float k = 1.0)
 // The inverse of Reinhard
 float3 ITM_Reinhard(float3 sdr, float k = 1.0)
 {
-    return k * sdr / (k - sdr);
+    return k * sdr / (1.0 - sdr);
 }
 
 //