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

[LLVM Continuous Integration via cfe-commits]

llvm-ci wrote:

LLVM Buildbot has detected a new failure on builder `openmp-offload-libc-amdgpu-runtime` running on `omp-vega20-1` while building `clang,offload` at step 10 "Add check check-offload".

Full details are available at: https://lab.llvm.org/buildbot/#/builders/73/builds/3628

Here is the relevant piece of the build log for the reference:
```
Step 10 (Add check check-offload) failure: test (failure)
******************** TEST 'libomptarget :: amdgcn-amd-amdhsa :: offloading/CUDA/basic_launch_multi_arg.cu' FAILED ********************
Exit Code: 1

Command Output (stdout):
--
# RUN: at line 1
/home/ompworker/bbot/openmp-offload-libc-amdgpu-runtime/llvm.build/./bin/clang++ -foffload-via-llvm --offload-arch=native /home/ompworker/bbot/openmp-offload-libc-amdgpu-runtime/llvm.src/offload/test/offloading/CUDA/basic_launch_multi_arg.cu -o /home/ompworker/bbot/openmp-offload-libc-amdgpu-runtime/llvm.build/runtimes/runtimes-bins/offload/test/amdgcn-amd-amdhsa/offloading/CUDA/Output/basic_launch_multi_arg.cu.tmp
# executed command: /home/ompworker/bbot/openmp-offload-libc-amdgpu-runtime/llvm.build/./bin/clang++ -foffload-via-llvm --offload-arch=native /home/ompworker/bbot/openmp-offload-libc-amdgpu-runtime/llvm.src/offload/test/offloading/CUDA/basic_launch_multi_arg.cu -o /home/ompworker/bbot/openmp-offload-libc-amdgpu-runtime/llvm.build/runtimes/runtimes-bins/offload/test/amdgcn-amd-amdhsa/offloading/CUDA/Output/basic_launch_multi_arg.cu.tmp
# .---command stderr------------
# | /home/ompworker/bbot/openmp-offload-libc-amdgpu-runtime/llvm.build/bin/clang-linker-wrapper: error: unable to find library -lomptarget
# | clang++: error: linker command failed with exit code 1 (use -v to see invocation)
# `-----------------------------
# error: command failed with exit status: 1

--

********************


```

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