[clang] [llvm] [Offload][CUDA] Allow CUDA kernels to use LLVM/Offload (PR #94549)

[Artem Belevich via cfe-commits]

https://github.com/Artem-B approved this pull request.

LGTM in principle.

Will kernels in TUs compiled with `-foffload-via-llvm` be interoperable with code that wants to launch them from another TU compiled w/o `-foffload-via-llvm` ?

E.g.:
- a.cu: `__global__ void kernel() { ... }`
- b.cu: `extern __global__ void kernel(); void func() { kernel<<<1,1>>>();}`

This could use a test in the testsuite to actually check whether it works.

https://github.com/llvm/llvm-project/pull/94549