[clang] [CIR][CUDA] Add semantic surface reference type (PR #223013)

[Konstantinos Parasyris via cfe-commits]

================
@@ -3855,6 +3855,9 @@ static void prepareTypeConverter(mlir::LLVMTypeConverter &converter,
     assert(!cir::MissingFeatures::addressSpace());
     return mlir::LLVM::LLVMPointerType::get(type.getContext());
   });
+  converter.addConversion([&](cir::CUDADeviceSurfaceType type) -> mlir::Type {
+    return mlir::IntegerType::get(type.getContext(), 64);
+  });
----------------
koparasy wrote:

NIT: Add the deleted comment that existed above:
```
// On the device side, surface reference is represented as an object handle
// in 64-bit integer.
```

https://github.com/llvm/llvm-project/pull/223013