[llvm] [Offload][AMDGPU] accept generic target (PR #118919)

[via llvm-commits]

================
@@ -205,13 +206,19 @@ Expected<std::string> getTargetTripleAndFeatures(hsa_agent_t Agent) {
       return Status;
 
     llvm::StringRef TripleTarget(ISAName.begin(), Length);
-    if (TripleTarget.consume_front("amdgcn-amd-amdhsa"))
-      Target = TripleTarget.ltrim('-').rtrim('\0').str();
-    return HSA_STATUS_INFO_BREAK;
+    if (TripleTarget.consume_front("amdgcn-amd-amdhsa")) {
+      auto Target = TripleTarget.ltrim('-').rtrim('\0');
+      Targets.push_back(Target);
+      if (!Target.ends_with("generic"))
+        SpecificTarget = Target; // Expect one (and only one) to be found
----------------
hidekisaito wrote:

If desired, we could add an assertion for SpecificTarget still empty before assigning, which might help catch HSA behavior change.

https://github.com/llvm/llvm-project/pull/118919