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

[Matt Arsenault via lldb-commits]

================
@@ -1528,24 +1127,31 @@ bool ArchSpec::IsMatch(const ArchSpec &rhs, MatchType match) const {
 }
 
 void ArchSpec::UpdateCore() {
-  llvm::StringRef arch_name(m_triple.getArchName());
-  const CoreDefinition *core_def = FindCoreDefinition(arch_name);
+  const CoreDefinition *core_def;
+  if (m_triple.getArch() == llvm::Triple::amdgpu) {
+    core_def = FindCoreDefinition(eCore_amd_gpu);
+    if (m_triple.getSubArch() == llvm::Triple::NoSubArch) {
+      llvm::StringRef arch_name = m_triple.getArchName();
+      if (arch_name != "amdgpu" && arch_name != "amdgcn") {
----------------
arsenm wrote:

You shouldn't need to check against the name 

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