Allow 1D SV_DispatchThreadID in CPU targets (#8612)

The varying param legalization pass didn't deal with this 1D form of
SV_DispatchThreadID for CPU targets:

```slang
void computeMain(int i : SV_DispatchThreadID)
```

Instead, it just overrode the type of `i` with a `uint3`, breaking lots
of code that attempted to use `i` for something, like a `switch`
statement for example.

I ran across this when going through `language-feature` tests for the
LLVM target, which will also use this legalization pass. I'm separately
submitting this now because this also fixes the existing CPU target. The
test I enable in this PR is one that was previously generating broken
code on CPU.

(somewhat related issue: #7468)

1 files changed, 5 insertions, 4 deletions

diff --git a/tests/cuda/dispatch-thread-id-extraction.slang b/tests/cuda/dispatch-thread-id-extraction.slang
index 5fc3c89a6..02705ff24 100644
--- a/tests/cuda/dispatch-thread-id-extraction.slang
+++ b/tests/cuda/dispatch-thread-id-extraction.slang
@@ -28,8 +28,9 @@ void computeMain3(int2 tid: SV_DispatchThreadID, StructuredBuffer<uint> src, RWS
 {
     dst[tid.x] = src[tid.x];
 }
-// CHECK: int _S3 = (slang_bit_cast<int2 >(uint2 {(blockIdx * blockDim + threadIdx).x, (blockIdx * blockDim + threadIdx).y})).x;
-
+// CHECK: uint2  _S3 = uint2 {(blockIdx * blockDim + threadIdx).x, (blockIdx * blockDim + threadIdx).y};
+// CHECK: int2  _S4 = make_int2 ((int)_S3.x, (int)_S3.y);
+// CHECK: int _S5 = _S4.x;
 
 [shader("compute")]
 [numthreads(1, 1, 1)]
@@ -37,7 +38,7 @@ void computeMain4(int tid: SV_DispatchThreadID, StructuredBuffer<uint> src, RWSt
 {
     dst[tid.x] = src[tid.x];
 }
-// CHECK: int _S4 = (slang_bit_cast<int>((blockIdx * blockDim + threadIdx).x));
+// CHECK: int _S6 = int((blockIdx * blockDim + threadIdx).x);
 
 [shader("compute")]
 [numthreads(1, 1, 1)]
@@ -45,4 +46,4 @@ void computeMain5(int tid: SV_GroupIndex, StructuredBuffer<uint> src, RWStructur
 {
     dst[tid.x] = src[tid.x];
 }
-// CHECK: int _S5 = (slang_bit_cast<int>((threadIdx.z * blockDim.y + threadIdx.y) * blockDim.x + threadIdx.x));
\ No newline at end of file
+// CHECK: int _S7 = int((threadIdx.z * blockDim.y + threadIdx.y) * blockDim.x + threadIdx.x);