================
@@ -2023,6 +2024,18 @@ static bool CheckAllArgsHaveFloatRepresentation(Sema *S,
CallExpr *TheCall) {
checkAllFloatTypes);
}
+static bool CheckUnsignedIntegerRepresentation(Sema *S, CallExpr *TheCall) {
+ auto checkUnsignedInteger = [](clang::QualType PassedType) -> bool {
+ clang::QualType BaseType =
+ PassedType->isVectorType()
+ ? PassedType->getAs<clang::VectorType>()->getElementType()
+ : PassedType;
+ return !BaseType->isUnsignedIntegerType();
----------------
Icohedron wrote:
According to what I see in the original issue and in the DXC implementation,
only unsigned integers are allowed (specifically, `uint2` and `uint4`).
`clang/test/SemaHLSL/BuiltIns/AddUIint64-errors.hlsl` has a test with `int2`
passed as arguments that is expected to fail.
As for checking if the integer is 32-bits, that is a good idea. The following
test is currently valid and does not emit an error.
```c++
uint2 test_16_bit_integer_type(uint16_t2 a, uint16_t2 b) {
return __builtin_hlsl_adduint64(a, b);
}
```
I am having trouble trying to implement a check on the integer bit count
however.
I do not know how to get access to `llvm::IntegerType::getIntegerBitWidth()`
from a `clang::QualType`.
https://github.com/llvm/llvm-project/pull/125319
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
