bogner wrote:
> Don't see a problem with this but could we elaborate on the motivation for
> this? Looks like this is required for #119041? Why is that?
>
> The original change this is based on is:
> [microsoft/DirectXShaderCompiler#6296](https://github.com/microsoft/DirectXShaderCompiler/pull/6296)?
> Which makes it sounds like it does affect the generated debug-info. I'm
> confused as to what's NFC and what's not
This PR is entirely NFC. #119041 additionally has a functional change - the
relevant part of which is the addition of an overload:
```c++
llvm::DIType *CGDebugInfo::CreateType(const HLSLAttributedResourceType *Ty,
llvm::DIFile *U) {
return getOrCreateType(Ty->getWrappedType(), U);
}
```
The reason this change is needed for that PR, is that this new overload needs
to be called whether or not we're dealing with a function specialization, but
since the code as is special-cases specializations we don't ever call the
overloaded CreateType.
https://github.com/llvm/llvm-project/pull/119445
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
