GPU Foreach Parsing and Checking (#1482)

This PR introduces parsing and semantic checking for a GPU foreach loop
for heterogeneouis programming. A GPU foreach loop takes the form:
```
__GPU_FOREACH(renderer, gridDims, LAMBDA(uint3 dispatchThreadID) {
        kernelCall(args, ...); });
```

And will allow the host code to call into a kernel with the correct
renderer and grid dimensions. This commit also introduces a hack to
unify types in the heterogeneous hello world file, which will hopefully
be amended in the future.

Co-authored-by: Tim Foley <tfoleyNV@users.noreply.github.com>

2 files changed, 5 insertions, 3 deletions

diff --git a/examples/heterogeneous-hello-world/main.cpp b/examples/heterogeneous-hello-world/main.cpp
index 47df20dc5..6bb1bc071 100644
--- a/examples/heterogeneous-hello-world/main.cpp
+++ b/examples/heterogeneous-hello-world/main.cpp
@@ -35,7 +35,7 @@
 #include "gfx/render.h"
 #include "gfx/d3d11/render-d3d11.h"
 #include "gfx/window.h"
-#include "../../prelude/slang-cpp-types.h";
+#include "../../prelude/slang-cpp-types.h"
 using namespace gfx;
 
 // We create global ref pointers to avoid dereferencing values
diff --git a/examples/heterogeneous-hello-world/shader.slang b/examples/heterogeneous-hello-world/shader.slang
index a9ad66cc7..6b56c8700 100644
--- a/examples/heterogeneous-hello-world/shader.slang
+++ b/examples/heterogeneous-hello-world/shader.slang
@@ -1,10 +1,10 @@
 // shader.slang
 
 //TEST_INPUT:ubuffer(random(float, 4096, -1.0, 1.0), stride=4):name=ioBuffer
-RWStructuredBuffer<float> ioBuffer;
+RWStructuredBuffer<float> convertBuffer(Ptr<gfx::BufferResource> x);
 
 [numthreads(4, 1, 1)]
-void computeMain(uint3 dispatchThreadID : SV_DispatchThreadID)
+void computeMain(uniform RWStructuredBuffer<float> ioBuffer, uint3 dispatchThreadID : SV_DispatchThreadID)
 {
     uint tid = dispatchThreadID.x;
 
@@ -69,6 +69,8 @@ public bool executeComputation() {
     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) ; });
     let shaderProgram = loadShaderProgram(renderer);
     let descriptorSetLayout = buildDescriptorSetLayout(renderer);
     let pipelineLayout = buildPipeline(renderer, descriptorSetLayout);