[clang] [llvm] [Offload][Driver] Split frontend CUDA/HIP language assumptions from backend device (PR #212373)

[Matt Arsenault via llvm-commits]

================
@@ -8313,11 +8339,11 @@ void Clang::ConstructJob(Compilation &C, const JobAction &JA,
   // Host-side offloading compilation receives all device-side outputs. Include
   // them in the host compilation depending on the target. If the host inputs
   // are not empty we use the new-driver scheme, otherwise use the old scheme.
-  if ((IsCuda || IsHIP) && CudaDeviceInput) {
+  if ((IsCuda || IsHIP) && !UsesLLVMOffloading && CudaDeviceInput) {
----------------
arsenm wrote:

This doesn't look like reducing the language mode dependence 

https://github.com/llvm/llvm-project/pull/212373