[Lldb-commits] [lldb] [lldb] Use TargetParser for AMDGPU architectures (PR #223192)

[satyanarayana reddy janga via lldb-commits]

================
@@ -1452,6 +1041,16 @@ bool ArchSpec::IsMatch(const ArchSpec &rhs, MatchType match) const {
   const llvm::Triple &lhs_triple = GetTriple();
   const llvm::Triple &rhs_triple = rhs.GetTriple();
 
+  if (lhs_triple.getArch() == llvm::Triple::amdgpu &&
+      rhs_triple.getArch() == llvm::Triple::amdgpu) {
+    bool sub_arches_match =
+        match == ExactMatch
+            ? lhs_triple.getSubArch() == rhs_triple.getSubArch()
+            : llvm::AMDGPU::isSubArchCompatible(lhs_triple, rhs_triple);
+    if (!sub_arches_match)
+      return false;
----------------
satyajanga wrote:

The AMDGPU special casing is still necessary because now all the models now share `eCore_amd_gpu`

and isCompatibleWith treats 2 different AMD GPU Subarchs as the same. for ExactMatch we still need to compare the SubArch as well.

Let me know if I am misunderstanding your question.

https://github.com/llvm/llvm-project/pull/223192