[clang] [Driver] Test ignored target-specific options for AMDGPU/NVPTX (PR #79222)

[Fangrui Song via cfe-commits]

================
@@ -0,0 +1,7 @@
+/// Some target-specific options are ignored for GPU, so %clang exits with code 0.
+// DEFINE: %{gpu_opts} = --cuda-gpu-arch=sm_60 --cuda-path=%S/Inputs/CUDA/usr/local/cuda --no-cuda-version-check
----------------
MaskRay wrote:

I think `gpu_opts` can be merged into `%{check}`. `%{check}` should be kept as it will become longer soon.

https://github.com/llvm/llvm-project/pull/79222