Bring heterogeneous-hello-world back up to date. (#1935)

* Bring heterogeneous-hello-world back up to date.

* Reintroduced heterogeneous-hello-world into the premake

* No longer uses compiled bytecode for entry point, instead a loadModule
call is hardocoded with the slang file name.

* Entry point is, similarly, hardcoded for now.

* Added a bypass to slang-legalize-types for an unneeded GPUForeach check

* Run premake and change to relative path

* Removed experimental and added README

Co-authored-by: Yong He <yonghe@outlook.com>

4 files changed, 0 insertions, 646 deletions

diff --git a/examples/experimental/heterogeneous-hello-world/README.md b/examples/experimental/heterogeneous-hello-world/README.md
deleted file mode 100644
index 709652922..000000000
--- a/examples/experimental/heterogeneous-hello-world/README.md
+++ /dev/null
@@ -1,4 +0,0 @@
-Slang "CPU Hello World Heterogeneous" Example
-===============================
-
-This example is a work-in-progress to illustrate how a heterogeneous programming example might work.  It should NOT be used as a reference for working Slang code yet.
\ No newline at end of file
diff --git a/examples/experimental/heterogeneous-hello-world/main.cpp b/examples/experimental/heterogeneous-hello-world/main.cpp
deleted file mode 100644
index 372fcd615..000000000
--- a/examples/experimental/heterogeneous-hello-world/main.cpp
+++ /dev/null
@@ -1,380 +0,0 @@
-// This example is out of date and currently disabled from build.
-// The `gfx` layer has been refactored with a new command list based
-// model. The example must be updated to use the new `gfx` interface
-// before it can be included in build.
-
-#if 0
-// main.cpp
-
-// This file implements an extremely simple example of loading and
-// executing a Slang shader program. This is primarily an example
-// of how to use Slang as a "drop-in" replacement for an existing
-// HLSL compiler like the `D3DCompile` API. More advanced usage
-// of advanced Slang language and API features is left to the
-// next example.
-//
-// The comments in the file will attempt to explain concepts as
-// they are introduced.
-//
-// Of course, in order to use the Slang API, we need to include
-// its header. We have set up the build options for this project
-// so that it is as simple as:
-//
-#include <slang.h>
-//
-// Other build setups are possible, and Slang doesn't assume that
-// its include directory must be added to your global include
-// path.
-
-// For the purposes of keeping the demo code as simple as possible,
-// while still retaining some level of portability, our examples
-// make use of a small platform and graphics API abstraction layer,
-// which is included in the Slang source distribution under the
-// `tools/` directory.
-//
-// Applications can of course use Slang without ever touching this
-// abstraction layer, so we will not focus on it when explaining
-// examples, except in places where best practices for interacting
-// with Slang may depend on an application/engine making certain
-// design choices in their abstraction layer.
-//
-#include "slang-com-ptr.h"
-#include "slang-gfx.h"
-#include "tools/graphics-app-framework/window.h"
-#include "../../prelude/slang-cpp-types.h"
-#include "source/core/slang-basic.h"
-
-using namespace gfx;
-
-// We create global ref pointers to avoid dereferencing values
-//
-ComPtr<gfx::IShaderProgram>         gShaderProgram;
-Slang::ComPtr<gfx::IRenderer>      gRenderer;
-
-ComPtr<gfx::IBufferResource> gStructuredBuffer;
-
-ComPtr<gfx::IPipelineLayout> gPipelineLayout;
-ComPtr<gfx::IPipelineState> gPipelineState;
-ComPtr<gfx::IDescriptorSetLayout> gDescriptorSetLayout;
-ComPtr<gfx::IDescriptorSet> gDescriptorSet;
-
-// Boilerplate types to help the slan-generated file
-//
-struct gfx_Window_0;
-struct gfx_Renderer_0;
-struct gfx_BufferResource_0;
-struct gfx_ShaderProgram_0;
-struct gfx_DescriptorSetLayout_0;
-struct gfx_PipelineLayout_0;
-struct gfx_DescriptorSet_0;
-struct gfx_PipelineState_0;
-
-bool executeComputation_0();
-extern unsigned char __computeMain[];
-extern size_t __computeMainSize;
-
-gfx::IShaderProgram* loadShaderProgram(gfx::IRenderer* renderer, unsigned char computeCode[], size_t computeCodeSize)
-{
-    // We extract the begin/end pointers to the output code buffers directly
-    //
-    char unsigned const* computeCodeEnd = computeCode + computeCodeSize;
-
-    // Now we use the operations of the example graphics API abstraction
-    // layer to load shader code into the underlying API.
-    //
-    // Reminder: this section does not involve the Slang API at all.
-    //
-
-    gfx::IShaderProgram::KernelDesc kernelDescs[] =
-    {
-        { gfx::StageType::Compute,    computeCode,     computeCodeEnd },
-    };
-
-    gfx::IShaderProgram::Desc programDesc = {};
-    programDesc.pipelineType = gfx::PipelineType::Compute;
-    programDesc.kernels = &kernelDescs[0];
-    programDesc.kernelCount = 1;
-
-    gShaderProgram = renderer->createProgram(programDesc);
-
-    return gShaderProgram;
-}
-
-// Now that we've covered the function that actually loads and
-// compiles our Slang shade code, we can go through the rest
-// of the application code without as much commentary.
-//
-gfx::Window* createWindow(int windowWidth, int windowHeight)
-{
-    // Create a window for our application to render into.
-    //
-    WindowDesc windowDesc;
-    windowDesc.title = "Hello, World!";
-    windowDesc.width = windowWidth;
-    windowDesc.height = windowHeight;
-    return createWindow(windowDesc);
-    //return globalWindow;
-}
-
-gfx::IRenderer* createRenderer(
-    int windowWidth,
-    int windowHeight,
-    gfx::Window* window)
-{
-    // Initialize the rendering layer.
-    //
-    // Note: for now we are hard-coding logic to use the
-    // Direct3D11 back-end for the graphics API abstraction.
-    // A future version of this example may support multiple
-    // platforms/APIs.
-    //
-    IRenderer::Desc rendererDesc = {};
-    rendererDesc.rendererType = gfx::RendererType::DirectX11;
-    Result res = gfxCreateRenderer(&rendererDesc, gRenderer.writeRef());
-
-    if (SLANG_FAILED(res)) return nullptr;
-    return gRenderer;
-}
-
-gfx::IBufferResource* createStructuredBuffer(gfx::IRenderer* renderer, float* initialArray)
-{
-    // Create a structured buffer for storing the data for computation
-    //
-    int structuredBufferSize = 4 * sizeof(float);
-
-    IBufferResource::Desc structuredBufferDesc;
-    structuredBufferDesc.init(structuredBufferSize);
-    structuredBufferDesc.setDefaults(IResource::Usage::UnorderedAccess);
-    structuredBufferDesc.elementSize = 4;
-    structuredBufferDesc.cpuAccessFlags = IResource::AccessFlag::Read;
-
-    gStructuredBuffer = renderer->createBufferResource(
-        IResource::Usage::UnorderedAccess,
-        structuredBufferDesc,
-        initialArray);
-    return gStructuredBuffer;
-}
-
-gfx::IDescriptorSetLayout* buildDescriptorSetLayout(gfx::IRenderer* renderer)
-{
-    // Our example graphics API usess a "modern" D3D12/Vulkan style
-    // of resource binding, so now we will dive into describing and
-    // allocating "descriptor sets."
-    //
-    // First, we need to construct a descriptor set *layout*.
-    //
-    IDescriptorSetLayout::SlotRangeDesc slotRanges[] =
-    {
-        IDescriptorSetLayout::SlotRangeDesc(DescriptorSlotType::StorageBuffer),
-    };
-    IDescriptorSetLayout::Desc descriptorSetLayoutDesc;
-    descriptorSetLayoutDesc.slotRangeCount = 1;
-    descriptorSetLayoutDesc.slotRanges = &slotRanges[0];
-    gDescriptorSetLayout = renderer->createDescriptorSetLayout(descriptorSetLayoutDesc);
-    return gDescriptorSetLayout;
-}
-
-gfx::IPipelineLayout* buildPipeline(gfx::IRenderer* renderer, gfx::IDescriptorSetLayout* descriptorSetLayout)
-{
-    // Next we will allocate a pipeline layout, which specifies
-    // that we will render with only a single descriptor set bound.
-    //
-
-    IPipelineLayout::DescriptorSetDesc descriptorSets[] =
-    {
-        IPipelineLayout::DescriptorSetDesc(descriptorSetLayout),
-    };
-    IPipelineLayout::Desc pipelineLayoutDesc;
-    pipelineLayoutDesc.renderTargetCount = 1;
-    pipelineLayoutDesc.descriptorSetCount = 1;
-    pipelineLayoutDesc.descriptorSets = &descriptorSets[0];
-    gPipelineLayout = renderer->createPipelineLayout(pipelineLayoutDesc);
-
-    return gPipelineLayout;
-}
-
-gfx::IDescriptorSet* buildDescriptorSet(
-    gfx::IRenderer* renderer,
-    gfx::IDescriptorSetLayout* descriptorSetLayout,
-    gfx::IBufferResource* structuredBuffer)
-{
-    // Once we have the descriptor set layout, we can allocate
-    // and fill in a descriptor set to hold our parameters.
-    //
-    gDescriptorSet = renderer->createDescriptorSet(descriptorSetLayout, gfx::IDescriptorSet::Flag::Transient);
-    if(!gDescriptorSet) return nullptr;
-
-    // Once we have the bufferResource created, we can fill in
-    // a descriptor set for creating a structured buffer
-    //
-    IResourceView::Desc resourceViewDesc;
-    resourceViewDesc.type = IResourceView::Type::UnorderedAccess;
-    auto resourceView = renderer->createBufferView(structuredBuffer, resourceViewDesc);
-    gDescriptorSet->setResource(0, 0, resourceView);
-
-    return gDescriptorSet;
-}
-
-gfx::IPipelineState* buildPipelineState(
-    gfx::IShaderProgram* shaderProgram,
-    gfx::IRenderer* renderer,
-    gfx::IPipelineLayout* pipelineLayout)
-{
-    // Following the D3D12/Vulkan style of API, we need a pipeline state object
-    // (PSO) to encapsulate the configuration of the overall graphics pipeline.
-    //
-    ComputePipelineStateDesc desc;
-    desc.pipelineLayout = pipelineLayout;
-    desc.program = shaderProgram;
-    gPipelineState = renderer->createComputePipelineState(desc);
-    return gPipelineState;
-}
-
-void printInitialValues(float* initialArray, int length)
-{
-    // Print out the values before the computation
-    printf("Before:\n");
-    for (int i = 0; i < length; i++)
-    {
-        printf("%f, ", initialArray[i]);
-    }
-    printf("\n");
-}
-
-void dispatchComputation(
-    gfx::ICommandQueue* gQueue,
-    gfx::IPipelineState* gPipelineState,
-    gfx::IPipelineLayout* gPipelineLayout,
-    gfx::IDescriptorSet* gDescriptorSet,
-    unsigned int gridDimsX,
-    unsigned int gridDimsY,
-    unsigned int gridDimsZ)
-{
-    auto cmdBuf = gQueue->createCommandBuffer();
-    auto encoder = cmdBuf->encodeComputeCommands();
-    encoder->setPipelineState(gPipelineState);
-    encoder->setDescriptorSet(PipelineType::Compute, gPipelineLayout, 0, gDescriptorSet);
-    encoder->dispatchCompute(gridDimsX, gridDimsY, gridDimsZ);
-    encoder->endEncoding();
-    gQueue->executeCommandBuffer(cmdBuf);
-}
-
-void print_output(
-    gfx::IRenderer* renderer,
-    gfx::IBufferResource* structuredBuffer,
-    int length)
-{
-    ComPtr<ISlangBlob> blob;
-    renderer->readBufferResource(structuredBuffer, 0, length * sizeof(float), blob.writeRef());
-    if (float* outputData = (float*)blob->getBufferPointer())
-    {
-        // Print out the values the the kernel produced
-        printf("After: \n");
-        for (int i = 0; i < 4; i++)
-        {
-            printf("%f, ", outputData[i]);
-        }
-        printf("\n");
-    }
-}
-
-// Boilerplate functions to help the slang-generated file and types
-gfx_Window_0* createWindow_0(int32_t _0, int32_t _1)
-{
-    return (gfx_Window_0*)createWindow(_0, _1);
-}
-
-gfx_Renderer_0* createRenderer_0(int32_t _0, int32_t _1, gfx_Window_0* _2)
-{
-    return (gfx_Renderer_0*)createRenderer(_0, _1, (gfx::Window*)_2);
-}
-
-gfx_BufferResource_0* createStructuredBuffer_0(gfx_Renderer_0* _0, FixedArray<float, 4> _1)
-{
-    return (gfx_BufferResource_0*)createStructuredBuffer((gfx::IRenderer*)_0, (float*)&_1);
-}
-
-gfx_ShaderProgram_0* loadShaderProgram_0(gfx_Renderer_0* _0, unsigned char _1[], size_t _2)
-{
-    return (gfx_ShaderProgram_0*)loadShaderProgram((gfx::IRenderer*)_0, _1, _2);
-}
-
-gfx_DescriptorSetLayout_0* buildDescriptorSetLayout_0(gfx_Renderer_0* _0)
-{
-    return (gfx_DescriptorSetLayout_0*)buildDescriptorSetLayout((gfx::IRenderer*)_0);
-}
-
-gfx_PipelineLayout_0* buildPipeline_0(gfx_Renderer_0* _0, gfx_DescriptorSetLayout_0* _1)
-{
-    return (gfx_PipelineLayout_0*)buildPipeline((gfx::IRenderer*)_0, (gfx::IDescriptorSetLayout*)_1);
-}
-
-gfx_DescriptorSet_0* buildDescriptorSet_0(gfx_Renderer_0* _0, gfx_DescriptorSetLayout_0* _1, gfx_BufferResource_0* _2)
-{
-    return (gfx_DescriptorSet_0*)buildDescriptorSet(
-        (gfx::IRenderer*)_0,
-        (gfx::IDescriptorSetLayout*)_1,
-        (gfx::IBufferResource*)_2);
-}
-
-gfx_PipelineState_0* buildPipelineState_0(gfx_ShaderProgram_0* _0, gfx_Renderer_0* _1, gfx_PipelineLayout_0* _2)
-{
-    return (gfx_PipelineState_0*)buildPipelineState(
-        (gfx::IShaderProgram*)_0, (gfx::IRenderer*)_1,
-        (gfx::IPipelineLayout*)_2);
-}
-
-void printInitialValues_0(FixedArray<float, 4> _0, int32_t _1)
-{
-    printInitialValues((float*)&_0, _1);
-}
-
-void dispatchComputation_0(gfx_CommandQueue_0* _0, gfx_PipelineState_0* _1, gfx_PipelineLayout_0* _2, gfx_DescriptorSet_0* _3, unsigned int gridDimsX, unsigned int gridDimsY, unsigned int gridDimsZ)
-{
-    dispatchComputation(
-        (gfx::ICommandQueue*)_0,
-        (gfx::IPipelineState*)_1,
-        (gfx::IPipelineLayout*)_2,
-        (gfx::IDescriptorSet*)_3,
-        gridDimsX,
-        gridDimsY,
-        gridDimsZ);
-}
-
-RWStructuredBuffer<float> convertBuffer_0(gfx_BufferResource_0* _0) {
-    RWStructuredBuffer<float> result;
-    result.data = (float*)_0;
-    return result;
-}
-
-gfx_BufferResource_0* unconvertBuffer_0(RWStructuredBuffer<float> _0) {
-    return (gfx_BufferResource_0*)(_0.data);
-}
-
-void print_output_0(gfx_CommandQueue_0* _0, gfx_BufferResource_0* _1, int32_t _2)
-{
-    print_output((gfx::ICommandQueue*)_0, (gfx::IBufferResource*)_1, _2);
-}
-
-// This "inner" main function is used by the platform abstraction
-// layer to deal with differences in how an entry point needs
-// to be defined for different platforms.
-//
-void innerMain(ApplicationContext* context)
-{
-    // We construct an instance of our example application
-    // `struct` type, and then walk through the lifecyle
-    // of the application.
-
-    if (!(executeComputation_0()))
-    {
-        return exitApplication(context, 1);
-    }
-}
-
-// This macro instantiates an appropriate main function to
-// invoke the `innerMain` above.
-//
-GFX_CONSOLE_MAIN(innerMain)
-
-#endif
diff --git a/examples/experimental/heterogeneous-hello-world/shader.cpp b/examples/experimental/heterogeneous-hello-world/shader.cpp
deleted file mode 100644
index 5a8dd7815..000000000
--- a/examples/experimental/heterogeneous-hello-world/shader.cpp
+++ /dev/null
@@ -1,197 +0,0 @@
-#if 0
-#include "../../prelude/slang-cpp-prelude.h"
-
-
-#ifdef SLANG_PRELUDE_NAMESPACE
-using namespace SLANG_PRELUDE_NAMESPACE;
-#endif
-
-Vector<uint32_t, 3> operator*(Vector<uint32_t, 3> a, Vector<uint32_t, 3> b)
-{
-    Vector<uint32_t, 3> r;
-    r.x = a.x * b.x;
-    r.y = a.y * b.y;
-    r.z = a.z * b.z;
-    return r;
-}
-
-Vector<uint32_t, 3> operator+(Vector<uint32_t, 3> a, Vector<uint32_t, 3> b)
-{
-    Vector<uint32_t, 3> r;
-    r.x = a.x + b.x;
-    r.y = a.y + b.y;
-    r.z = a.z + b.z;
-    return r;
-}
-
-Vector<uint32_t, 3> make_VecU3(uint32_t a, uint32_t b, uint32_t c)
-{
-    return Vector<uint32_t, 3>{ a, b, c};
-}
-
-size_t __computeMainSize = 668;
-unsigned char __computeMain[] = {68, 88, 66, 67, 87, 111, 81, 164, 2, 29, 72, 42, 151, 28, 13, 217, 55, 37, 7, 95, 1, 0, 0, 0, 156, 2, 0, 0, 5, 0, 0, 0, 52, 0, 0, 0, 8, 1, 0, 0, 24, 1, 0, 0, 40, 1, 0, 0, 32, 2, 0, 0, 82, 68, 69, 70, 204, 0, 0, 0, 1, 0, 0, 0, 88, 0, 0, 0, 1, 0, 0, 0, 28, 0, 0, 0, 0, 4, 83, 67, 0, 9, 16, 0, 164, 0, 0, 0, 60, 0, 0, 0, 6, 0, 0, 0, 6, 0, 0, 0, 1, 0, 0, 0, 4, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 101, 110, 116, 114, 121, 80, 111, 105, 110, 116, 80, 97, 114, 97, 109, 115, 95, 105, 111, 66, 117, 102, 102, 101, 114, 95, 48, 0, 60, 0, 0, 0, 1, 0, 0, 0, 112, 0, 0, 0, 4, 0, 0, 0, 0, 0, 0, 0, 3, 0, 0, 0, 136, 0, 0, 0, 0, 0, 0, 0, 4, 0, 0, 0, 2, 0, 0, 0, 148, 0, 0, 0, 0, 0, 0, 0, 36, 69, 108, 101, 109, 101, 110, 116, 0, 171, 171, 171, 0, 0, 3, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 77, 105, 99, 114, 111, 115, 111, 102, 116, 32, 40, 82, 41, 32, 72, 76, 83, 76, 32, 83, 104, 97, 100, 101, 114, 32, 67, 111, 109, 112, 105, 108, 101, 114, 32, 49, 48, 46, 49, 0, 73, 83, 71, 78, 8, 0, 0, 0, 0, 0, 0, 0, 8, 0, 0, 0, 79, 83, 71, 78, 8, 0, 0, 0, 0, 0, 0, 0, 8, 0, 0, 0, 83, 72, 69, 88, 240, 0, 0, 0, 64, 0, 5, 0, 60, 0, 0, 0, 106, 8, 0, 1, 158, 0, 0, 4, 0, 224, 17, 0, 0, 0, 0, 0, 4, 0, 0, 0, 95, 0, 0, 2, 18, 0, 2, 0, 104, 0, 0, 2, 1, 0, 0, 0, 155, 0, 0, 4, 4, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 167, 0, 0, 8, 18, 0, 16, 0, 0, 0, 0, 0, 10, 0, 2, 0, 1, 64, 0, 0, 0, 0, 0, 0, 6, 224, 17, 0, 0, 0, 0, 0, 49, 0, 0, 7, 34, 0, 16, 0, 0, 0, 0, 0, 10, 0, 16, 0, 0, 0, 0, 0, 1, 64, 0, 0, 0, 0, 0, 63, 0, 0, 0, 7, 66, 0, 16, 0, 0, 0, 0, 0, 10, 0, 16, 0, 0, 0, 0, 0, 10, 0, 16, 0, 0, 0, 0, 0, 75, 0, 0, 5, 18, 0, 16, 0, 0, 0, 0, 0, 10, 0, 16, 0, 0, 0, 0, 0, 55, 0, 0, 9, 18, 0, 16, 0, 0, 0, 0, 0, 26, 0, 16, 0, 0, 0, 0, 0, 42, 0, 16, 0, 0, 0, 0, 0, 10, 0, 16, 0, 0, 0, 0, 0, 168, 0, 0, 8, 18, 224, 17, 0, 0, 0, 0, 0, 10, 0, 2, 0, 1, 64, 0, 0, 0, 0, 0, 0, 10, 0, 16, 0, 0, 0, 0, 0, 62, 0, 0, 1, 83, 84, 65, 84, 116, 0, 0, 0, 7, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0};
-void computeMain_wrapper(gfx_Renderer_0* renderer, Vector<uint32_t, 3> gridDims, 
-	RWStructuredBuffer<float> buffer)
-{
-	gfx_ShaderProgram_0* shaderProgram = loadShaderProgram_0(renderer, __computeMain, __computeMainSize);
-	gfx_DescriptorSetLayout_0* setLayout = buildDescriptorSetLayout_0(renderer);
-	gfx_PipelineLayout_0* pipelineLayout = buildPipeline_0(renderer, setLayout);
-	gfx_DescriptorSet_0* descriptorSet = buildDescriptorSet_0(renderer, setLayout, unconvertBuffer_0(buffer));
-	gfx_PipelineState_0* pipelineState = buildPipelineState_0(shaderProgram, renderer, pipelineLayout);
-	dispatchComputation_0(renderer, pipelineState, pipelineLayout, descriptorSet, gridDims.x, gridDims.y, gridDims.z);
-}
-
-#line 7 "../../examples/heterogeneous-hello-world/shader.slang"
-struct EntryPointParams_0
-{
-    RWStructuredBuffer<float> ioBuffer_0;
-};
-
-struct KernelContext_0
-{
-};
-
-
-#line 21
-struct gfx_Window_0
-{
-};
-
-
-#line 22
-struct gfx_Renderer_0
-{
-};
-
-
-#line 23
-struct gfx_BufferResource_0
-{
-};
-
-
-#line 7
-void _computeMain(void* _S1, void* entryPointParams_0, void* _S2)
-{
-    ComputeThreadVaryingInput* _S3 = ((ComputeThreadVaryingInput*)(_S1));
-    KernelContext_0 kernelContext_0;
-
-#line 9
-    uint32_t tid_0 = (*(&_S3->groupID) * make_VecU3(4U, 1U, 1U) + *(&_S3->groupThreadID)).x;
-
-    float* _S4 = &(*(&((EntryPointParams_0*)(entryPointParams_0))->ioBuffer_0))[tid_0];
-
-#line 11
-    float i_0 = *_S4;
-    bool _S5 = i_0 < 0.50000000000000000000f;
-
-#line 12
-    float _S6 = i_0 + i_0;
-
-#line 12
-    float _S7 = (F32_sqrt((i_0)));
-
-#line 12
-    float o_0 = _S5 ? _S6 : _S7;
-
-    float* _S8 = &(*(&((EntryPointParams_0*)(entryPointParams_0))->ioBuffer_0))[tid_0];
-
-#line 14
-    *_S8 = o_0;
-
-#line 7
-    return;
-}
-
-
-#line 34
-gfx_Window_0* createWindow_0(int32_t _0, int32_t _1);
-
-
-#line 35
-gfx_Renderer_0* createRenderer_0(int32_t _0, int32_t _1, gfx_Window_0* _2);
-
-
-
-gfx_BufferResource_0* createStructuredBuffer_0(gfx_Renderer_0* _0, FixedArray<float, 4> _1);
-
-
-#line 4
-RWStructuredBuffer<float> convertBuffer_0(gfx_BufferResource_0* _0);
-
-
-#line 40
-void printInitialValues_0(FixedArray<float, 4> _0, int32_t _1);
-
-
-#line 41
-void print_output_0(gfx_Renderer_0* _0, gfx_BufferResource_0* _1, int32_t _2);
-
-
-
-
-bool executeComputation_0()
-{
-
-
-
-    FixedArray<float, 4> initialArray_0 = { 3.00000000000000000000f, -20.00000000000000000000f, -6.00000000000000000000f, 8.00000000000000000000f };
-
-
-    gfx_Window_0* _S9 = createWindow_0(int(1024), int(768));
-    gfx_Renderer_0* _S10 = createRenderer_0(int(1024), int(768), _S9);
-    gfx_CommandQueue_0* _
-    gfx_BufferResource_0* _S11 = createStructuredBuffer_0(_S10, initialArray_0);
-    Vector<uint32_t, 3> _S12 = make_VecU3(uint32_t(int(4)), uint32_t(int(1)), uint32_t(int(1)));
-    RWStructuredBuffer<float> _S13 = convertBuffer_0(_S11);
-
-#line 57
-    computeMain_wrapper(_S10, _S12, _S13);
-
-    printInitialValues_0(initialArray_0, int(4));
-    print_output_0(_S10, _S11, int(4));
-
-
-    return true;
-}
-
-// [numthreads(4, 1, 1)]
-SLANG_PRELUDE_EXPORT
-void computeMain_Thread(ComputeThreadVaryingInput* varyingInput, void* entryPointParams, void* globalParams)
-{
-    _computeMain(varyingInput, entryPointParams, globalParams);
-}
-// [numthreads(4, 1, 1)]
-SLANG_PRELUDE_EXPORT
-void computeMain_Group(ComputeVaryingInput* varyingInput, void* entryPointParams, void* globalParams)
-{
-    ComputeThreadVaryingInput threadInput = {};
-    threadInput.groupID = varyingInput->startGroupID;
-    for (uint32_t x = 0; x < 4; ++x)
-    {
-        threadInput.groupThreadID.x = x;
-        _computeMain(&threadInput, entryPointParams, globalParams);
-    }
-}
-// [numthreads(4, 1, 1)]
-SLANG_PRELUDE_EXPORT
-void computeMain(ComputeVaryingInput* varyingInput, void* entryPointParams, void* globalParams)
-{
-    ComputeVaryingInput vi = *varyingInput;
-    ComputeVaryingInput groupVaryingInput = {};
-    for (uint32_t z = vi.startGroupID.z; z < vi.endGroupID.z; ++z)
-    {
-        groupVaryingInput.startGroupID.z = z;
-        for (uint32_t y = vi.startGroupID.y; y < vi.endGroupID.y; ++y)
-        {
-            groupVaryingInput.startGroupID.y = y;
-            for (uint32_t x = vi.startGroupID.x; x < vi.endGroupID.x; ++x)
-            {
-                groupVaryingInput.startGroupID.x = x;
-                computeMain_Group(&groupVaryingInput, entryPointParams, globalParams);
-            }
-        }
-    }
-}
-#endif
diff --git a/examples/experimental/heterogeneous-hello-world/shader.slang b/examples/experimental/heterogeneous-hello-world/shader.slang
deleted file mode 100644
index 47c883b39..000000000
--- a/examples/experimental/heterogeneous-hello-world/shader.slang
+++ /dev/null
@@ -1,65 +0,0 @@
-// shader.slang
-
-//TEST_INPUT:ubuffer(random(float, 4096, -1.0, 1.0), stride=4):name=ioBuffer
-RWStructuredBuffer<float> convertBuffer(Ptr<gfx::BufferResource> x);
-
-[shader("compute")]
-[numthreads(4, 1, 1)]
-void computeMain(uniform RWStructuredBuffer<float> ioBuffer, uint3 dispatchThreadID : SV_DispatchThreadID)
-{
-    uint tid = dispatchThreadID.x;
-
-    float i = ioBuffer[tid];
-    float o = i < 0.5 ? (i + i) : sqrt(i);
-
-    ioBuffer[tid] = o;
-}
-
-// Forward declarations of gfx types
-//
-namespace gfx {
-    struct ApplicationContext{};
-    struct Window{};
-    struct Renderer{};
-    struct BufferResource{};
-    struct PipelineLayout{};
-    struct PipelineState{};
-    struct DescriptorSetLayout{};
-    struct DescriptorSet{};
-    struct ShaderProgram{};
-}
-
-// Forward declarations of cpp functions
-//
-Ptr<gfx::ShaderProgram> loadShaderProgram(Ptr<gfx::Renderer> renderer);
-Ptr<gfx::Window> createWindow(int gWindowWidth, int gWindowHeight);
-Ptr<gfx::Renderer> createRenderer(
-    int gWindowWidth,
-    int gWindowHeight,
-    Ptr<gfx::Window> gWindow);
-Ptr<gfx::BufferResource> createStructuredBuffer(Ptr<gfx::Renderer> gRenderer, float[4] initialArray);
-void printInitialValues(float[4] initialArray, int length);
-void print_output(
-    Ptr<gfx::Renderer> gRenderer,
-    Ptr<gfx::BufferResource> gStructuredBuffer,
-    int length);
-
-public bool executeComputation() {
-    // We will hard-code the size of our rendering window and initial array.
-    //
-    int     windowWidth = 1024;
-    int     windowHeight = 768;
-    float   initialArray[4] = { 3.0f, -20.0f, -6.0f, 8.0f };
-
-    // Declare functions
-    let window = createWindow(windowWidth, windowHeight);
-    let renderer = createRenderer(windowWidth, windowHeight, window);
-    let structuredBuffer = createStructuredBuffer(renderer, initialArray);
-    __GPU_FOREACH(renderer, uint3(4, 1, 1), LAMBDA(uint3 dispatchThreadID)
-        { computeMain(convertBuffer(structuredBuffer), dispatchThreadID) ; });
-    printInitialValues(initialArray, 4);
-    print_output(renderer, structuredBuffer, 4);
-
-
-    return true;
-}