[libcxx-commits] [libcxx] [libc++] Protect the libc++ implementation from CUDA SDK's `__noinline__` macro (PR #73838)

[Artem Belevich via libcxx-commits]

Artem-B wrote:

> Should the fix be done by adding a `clang/lib/Headers/cuda_wrappers/__config` wrapper instead, so that all the CUDA `__noinline__` hacks are kept in one place?

If that's the only place where the issue pops up, then it's an option. So far that's what we did for libstdc++ workarounds.


https://github.com/llvm/llvm-project/pull/73838