[clang] Reduce memory usage in AST parent map generation by lazily checking if nodes have been seen (PR #129934)

[Erich Keane via cfe-commits]

================
@@ -70,16 +93,37 @@ class ParentMapContext::ParentMap {
         push_back(Value);
     }
     bool contains(const DynTypedNode &Value) {
-      return Seen.contains(Value);
+      assert(Value.getMemoizationData());
+      bool found = FragileLazySeenCache.contains(&Value);
+      while (!found && ItemsProcessed < Items.size()) {
+        found |= FragileLazySeenCache.insert(&Items[ItemsProcessed]).second;
+        ++ItemsProcessed;
+      }
+      return found;
     }
     void push_back(const DynTypedNode &Value) {
-      if (!Value.getMemoizationData() || Seen.insert(Value).second)
+      if (!Value.getMemoizationData() || !contains(Value)) {
+        const size_t OldCapacity = Items.capacity();
         Items.push_back(Value);
+        if (OldCapacity != Items.capacity()) {
+          // Pointers are invalidated; remove them.
+          ItemsProcessed = 0;
+          // Free memory to avoid doubling peak memory usage during rehashing
+          FragileLazySeenCache.clear();
+        }
+      }
     }
     llvm::ArrayRef<DynTypedNode> view() const { return Items; }
   private:
+    // BE CAREFUL. Pointers into this container are stored in the cache.
----------------
erichkeane wrote:

```suggestion
    // BE CAREFUL. Pointers into this container are stored in the 
`FragileLazySeenCache` set below.
```

https://github.com/llvm/llvm-project/pull/129934
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits