DirectXShaderCompiler/tools/clang/test/CodeGenHLSL/Samples/MiniEngine/GenerateHistogramCS.hlsl

// RUN: %dxc -E main -T cs_6_0 %s | FileCheck %s

// CHECK: flattenedThreadIdInGroup
// CHECK: threadId
// CHECK: barrier
// CHECK: textureLoad
// CHECK: AtomicAdd

//
// Copyright (c) Microsoft. All rights reserved.
// This code is licensed under the MIT License (MIT).
// THIS CODE IS PROVIDED *AS IS* WITHOUT WARRANTY OF
// ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING ANY
// IMPLIED WARRANTIES OF FITNESS FOR A PARTICULAR
// PURPOSE, MERCHANTABILITY, OR NON-INFRINGEMENT.
//
// Developed by Minigraph
//
// Author:  James Stanard 
//
// The group size is 16x16, but one group iterates over an entire 16-wide column of pixels (384 pixels tall)
// Assuming the total workspace is 640x384, there will be 40 thread groups computing the histogram in parallel.
// The histogram measures logarithmic luminance ranging from 2^-12 up to 2^4.  This should provide a nice window
// where the exposure would range from 2^-4 up to 2^4.

#include "PostEffectsRS.hlsli"

Texture2D<uint> LumaBuf : register( t0 );
RWByteAddressBuffer Histogram : register( u0 );

groupshared uint g_TileHistogram[256];

[RootSignature(PostEffects_RootSig)]
[numthreads( 16, 16, 1 )]
void main( uint GI : SV_GroupIndex, uint3 DTid : SV_DispatchThreadID )
{
	g_TileHistogram[GI] = 0;

	GroupMemoryBarrierWithGroupSync();

	// Loop 24 times until the entire column has been processed
	for (uint TopY = 0; TopY < 384; TopY += 16)
	{
		uint QuantizedLogLuma = LumaBuf[DTid.xy + uint2(0, TopY)];
		InterlockedAdd( g_TileHistogram[QuantizedLogLuma], 1 );
	}

	GroupMemoryBarrierWithGroupSync();

	Histogram.InterlockedAdd( GI * 4, g_TileHistogram[GI] );
}