[clang] [clang] Avoid recalculating TBAA base type info (PR #73264)

[Eli Friedman via cfe-commits]

================
@@ -418,14 +418,20 @@ llvm::MDNode *CodeGenTBAA::getBaseTypeInfo(QualType QTy) {
     return nullptr;
 
   const Type *Ty = Context.getCanonicalType(QTy).getTypePtr();
-  if (llvm::MDNode *N = BaseTypeMetadataCache[Ty])
-    return N;
 
-  // Note that the following helper call is allowed to add new nodes to the
-  // cache, which invalidates all its previously obtained iterators. So we
-  // first generate the node for the type and then add that node to the cache.
+  // nullptr is a valid value in the cache, so use find rather than []
+  auto I = BaseTypeMetadataCache.find(Ty);
+  if (I != BaseTypeMetadataCache.end())
+    return I->second;
+
+  // First calculate the metadata, before recomputing the insertion point, as
+  // the helper can recursively call us.
   llvm::MDNode *TypeNode = getBaseTypeInfoHelper(Ty);
-  return BaseTypeMetadataCache[Ty] = TypeNode;
+  LLVM_ATTRIBUTE_UNUSED auto inserted =
+      BaseTypeMetadataCache.try_emplace(Ty, TypeNode);
----------------
efriedma-quic wrote:

Please just use "insert" when you're dealing with trivially constructible types.

https://github.com/llvm/llvm-project/pull/73264