[clang] [AMDGPU] Use the AMDGPUToolChain when targeting C/C++ directly (PR #99687)

[Yaxun Liu via cfe-commits]

================
@@ -6405,7 +6424,12 @@ const ToolChain &Driver::getToolChain(const ArgList &Args,
       TC = std::make_unique<toolchains::NVPTXToolChain>(*this, Target, Args);
       break;
     case llvm::Triple::AMDHSA:
-      TC = std::make_unique<toolchains::ROCMToolChain>(*this, Target, Args);
+      TC =
+          llvm::any_of(Inputs,
+                       [](auto &Input) { return types::isOpenCL(Input.first); })
+              ? std::make_unique<toolchains::ROCMToolChain>(*this, Target, Args)
----------------
yxsamliu wrote:

If input files have mixed OpenCL and C/C++ programs, ROCMToolChain will be chosen. Is ROCmToolChain is able to compile C/C++ programs like AMDGPUToolChain, or it will cause errors? Should we emit diagnostic message that mixed OpenCL and C/C++ programs are not supported?

https://github.com/llvm/llvm-project/pull/99687