[CUDA] Unbreak CUDA compilation with -std=c++20

[CUDA] Unbreak CUDA compilation with -std=c++20

Standard libc++ headers in stdc++ mode include <new> which picks up
cuda_wrappers/new before any of the CUDA macros have been defined.

We can not include CUDA headers that early, so the work-around is to define
__device__ in the wrapper header itself.

Differential Revision: https://reviews.llvm.org/D91807