Allocate buffer for gathered luma values

Author: nipunG314

26_Autoexposure/app_resources/common.hlsl

@@ -0,0 +1,13 @@
+// Copyright (C) 2018-2024 - DevSH Graphics Programming Sp. z O.O.
+// This file is part of the "Nabla Engine".
+// For conditions of distribution and use, see copyright notice in nabla.h
+
+#ifndef _AUTOEXPOSURE_COMMON_INCLUDED_
+#define _AUTOEXPOSURE_COMMON_INCLUDED_
+
+struct AutoexposurePushData
+{
+    uint32_t viewportSizeX, viewportSizeY;
+};
+
+#endif

26_Autoexposure/app_resources/luma_meter.comp.hlsl

@@ -0,0 +1,17 @@
+// Copyright (C) 2024-2024 - DevSH Graphics Programming Sp. z O.O.
+// This file is part of the "Nabla Engine".
+// For conditions of distribution and use, see copyright notice in nabla.h
+
+#include "app_resources/common.hlsl"
+
+[[vk::push_constant]] AutoexposurePushData pushData;
+
+uint32_t3 nbl::hlsl::glsl::gl_WorkGroupSize()
+{
+    return uint32_t3(WorkgroupSize, 1, 1);
+}
+
+[numthreads(SubgroupSize, SubgroupSize, 1)]
+void main(uint32_t3 ID : SV_GroupThreadID, uint32_t3 GroupID : SV_GroupID)
+{
+}

26_Autoexposure/app_resources/present.frag.hlsl

@@ -8,8 +8,6 @@
 #include <nbl/builtin/hlsl/ext/FullScreenTriangle/SVertexAttributes.hlsl>
 using namespace nbl::hlsl::ext::FullScreenTriangle;
 
-#include <nbl/builtin/hlsl/luma_meter/luma_meter.hlsl>
-
 [[vk::combinedImageSampler]] [[vk::binding(0, 3)]] Texture2D texture;
 [[vk::combinedImageSampler]] [[vk::binding(0, 3)]] SamplerState samplerState;
 

26_Autoexposure/app_resources/tonemap.comp.hlsl

@@ -0,0 +1,17 @@
+// Copyright (C) 2024-2024 - DevSH Graphics Programming Sp. z O.O.
+// This file is part of the "Nabla Engine".
+// For conditions of distribution and use, see copyright notice in nabla.h
+
+#include "app_resources/common.hlsl"
+
+[[vk::push_constant]] AutoexposurePushData pushData;
+
+uint32_t3 nbl::hlsl::glsl::gl_WorkGroupSize()
+{
+    return uint32_t3(WorkgroupSize, 1, 1);
+}
+
+[numthreads(SubgroupSize, SubgroupSize, 1)]
+void main(uint32_t3 ID : SV_GroupThreadID, uint32_t3 GroupID : SV_GroupID)
+{
+}

26_Autoexposure/main.cpp

@@ -227,6 +227,46 @@ class AutoexposureApp final : public examples::SimpleWindowedApplication, public
 			};
 		}
 
+		// Allocate and create buffer for Luma Gather
+		{
+			// Allocate memory
+			nbl::video::IDeviceMemoryAllocator::SAllocation allocation = {};
+			smart_refctd_ptr<IGPUBuffer> buffer;
+			//smart_refctd_ptr<nbl::video::IGPUDescriptorSet> ds;
+			{
+				auto build_buffer = [this](
+					smart_refctd_ptr<ILogicalDevice> m_device,
+					nbl::video::IDeviceMemoryAllocator::SAllocation* allocation,
+					smart_refctd_ptr<IGPUBuffer>& buffer,
+					size_t buffer_size,
+					const char* label)
+				{
+					IGPUBuffer::SCreationParams params;
+					params.size = buffer_size;
+					params.usage = IGPUBuffer::EUF_STORAGE_BUFFER_BIT | IGPUBuffer::EUF_SHADER_DEVICE_ADDRESS_BIT;
+					buffer = m_device->createBuffer(std::move(params));
+					if (!buffer)
+						return logFail("Failed to create GPU buffer of size %d!\n", buffer_size);
+
+					buffer->setObjectDebugName(label);
+
+					auto reqs = buffer->getMemoryReqs();
+					reqs.memoryTypeBits &= m_physicalDevice->getHostVisibleMemoryTypeBits();
+
+					*allocation = m_device->allocate(reqs, buffer.get(), IDeviceMemoryAllocation::EMAF_DEVICE_ADDRESS_BIT);
+					if (!allocation->isValid())
+						return logFail("Failed to allocate Device Memory compatible with our GPU Buffer!\n");
+
+					assert(allocation->memory.get() == buffer->getBoundMemory().memory);
+				};
+
+				auto x = m_physicalDevice->getLimits();
+
+				build_buffer(m_device, &allocation, buffer, m_physicalDevice->getLimits().maxSubgroupSize, "Luma Gather Buffer");
+			}
+			m_lumaGatherBDA = buffer->getDeviceAddress();
+		}
+
 		// Allocate and Leave 1/4 for image uploads, to test image copy with small memory remaining
 		{
 			uint32_t localOffset = video::StreamingTransientDataBufferMT<>::invalid_value;
@@ -347,8 +387,6 @@ class AutoexposureApp final : public examples::SimpleWindowedApplication, public
 			queue->endCapture();
 		}
 
-		m_computeSubgroupSize = m_physicalDevice->getLimits().maxComputeWorkgroupSubgroups;
-
 		return true;
 	}
 
@@ -487,8 +525,8 @@ class AutoexposureApp final : public examples::SimpleWindowedApplication, public
 	smart_refctd_ptr<IWindow> m_window;
 	smart_refctd_ptr<CSimpleResizeSurface<CDefaultSwapchainFramebuffers>> m_surface;
 
-	// constants
-	uint32_t m_computeSubgroupSize = 0;
+	// luma gather
+	uint64_t m_lumaGatherBDA;
 };
 
 NBL_MAIN_FUNC(AutoexposureApp)
@@ -531,20 +569,6 @@ int main()
 	if (!device)
 		return 1; // could not create selected driver.
 
-	QToQuitEventReceiver receiver;
-	device->setEventReceiver(&receiver);
-
-	IVideoDriver* driver = device->getVideoDriver();
-	
-	nbl::io::IFileSystem* filesystem = device->getFileSystem();
-	IAssetManager* am = device->getAssetManager();
-
-	IAssetLoader::SAssetLoadParams lp;
-	auto imageBundle = am->getAsset("../../media/noises/spp_benchmark_4k_512.exr", lp);
-
-	auto glslCompiler = am->getCompilerSet();
-	const auto inputColorSpace = std::make_tuple(inFormat,ECP_SRGB,EOTF_IDENTITY);
-
 	using LumaMeterClass = ext::LumaMeter::CLumaMeter;
 	constexpr auto MeterMode = LumaMeterClass::EMM_MEDIAN;
 	const float minLuma = 1.f/2048.f;