[clang] [Clang] Do not emit intrinsic math functions on GPU targets (PR #98209)

[Artem Belevich via cfe-commits]

Artem-B wrote:

Given that the prevalent compilation for CUDA has no standard library whatsoever, preserving libcalls may break some existing users that may be relying on library call lowering to an intrinsic that *is* implemented by the back-end.

Perhaps this "no library call to intrinsic conversion"  mode could be limited to the stand-alone compilation only, at least until we have a more granular control over whether particular library call is save to lower to an intrinsic?



https://github.com/llvm/llvm-project/pull/98209