Disable specializing function calls if they have a struct param, that contains an array (#1448)

* This code is disabled, it was part of the optimization `Specialize function calls involving array arguments. (#1389)` on github.
It is disabled here because it causes a problem when a struct is passed to a function that contains a structured buffer *and* an array. It is specialized on the struct type, and so those types become parameters to the function. If the struct contains a structured buffer this is a problem on GLSL/VK based targets because currently structured buffers cannot be function parameters.
The fix for now is to just disable this optimization.

* Fix typo in name of test expected values.

1 files changed, 41 insertions, 0 deletions

diff --git a/tests/bugs/specialize-function-array-args.slang b/tests/bugs/specialize-function-array-args.slang
new file mode 100644
index 000000000..6b7ec44b4
--- /dev/null
+++ b/tests/bugs/specialize-function-array-args.slang
@@ -0,0 +1,41 @@
+//TEST(compute):COMPARE_COMPUTE_EX:-slang -compute
+//TEST(compute,vulkan):COMPARE_COMPUTE_EX:-vk -slang -compute
+
+// When a function is passed a parameter that contains an array, it specialized it as a performance
+// improvement for VK. If the struct contained a structured buffer, though it meant that the 
+// function was specialized to have the structured buffer passed as a parameter. For GLSL/VK based
+// targets Slang cannot pass StructuredBuffer types and so would subsequently fail in 
+// GLSLSourceEmitter::emitSimpleTypeImpl when trying to output the StructuredBuffer type as a parameter
+//
+// This test, checks that when passing a struct with an array and structured buffer
+// code functions as expected. 
+
+struct Params
+{
+    RWStructuredBuffer<int> buffer;
+    int a[2];
+};
+
+//TEST_INPUT:ubuffer(data=[9 9 9 9], stride=4):out,name=outputBuffer
+RWStructuredBuffer<int> outputBuffer;
+
+//TEST_INPUT:ubuffer(data=[1 2 3 4], stride=4):name=anotherBuffer
+RWStructuredBuffer<int> anotherBuffer;
+// 
+int doSomething(Params params, int v)
+{
+    return params.a[v & 1] - v;
+}
+
+[numthreads(4, 1, 1)]
+void computeMain(uint3 dispatchThreadID : SV_DispatchThreadID)
+{
+    Params params;
+    params.buffer = anotherBuffer;
+    params.a[0] = 1;
+    params.a[1] = 2;
+
+	int tid = int(dispatchThreadID.x);
+
+	outputBuffer[tid] = doSomething(params, tid);
+}
\ No newline at end of file