René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"LIST_APPEND and SET_ADD opcodes safety in Free Threading","articleBody":"Hi,\n\nWhile reviewing PR gh-152273, I moved `_Py_CRITICAL_SECTION_ASSERT_OBJECT_LOCKED(so);` check to `set_add_entry_takeref()` to check for bugs. I discovered that `SET_ADD` opcode calls `_PySet_AddTakeRef()` without getting the set critical section. That's surprising and IMO it would deserve adding a comment to explain why it's safe to omit locking here.\n\n`LIST_APPEND` has a similar design (also omit locking).\n\n---\n\nThe following code uses the evil `gc.get_objects()` function to call `set.add()` in a different thread while the main thread is building a set:\n\n```py\nimport dis\nimport gc\nimport threading\n\nGLOBAL_SET = None\n\ndef worker():\n    EVENT.wait()\n    obj = GLOBAL_SET\n    for i in range(100):\n        # Call set.add() which uses the critical section\n        obj.add(i)\n\n\nMARKER = b\"MARKER\".decode()\nEVENT = threading.Event()\n\nclass EvilHash:\n    def __init__(self, hash_value):\n        self.hash_value = hash_value\n\n    def __hash__(self):\n        global GLOBAL_SET, EVENT\n        if GLOBAL_SET is None:\n            # Invoke the evil gc.get_objects()!\n            for obj in gc.get_objects():\n                if isinstance(obj, set) and MARKER in obj:\n                    GLOBAL_SET = obj\n                    EVENT.set()\n        return self.hash_value\n\n    def __repr__(self):\n        return f\"EvilHash({self.hash_value})\"\n\ndef func():\n    build_set = {\n        # Use *list so following items are added by SET_ADD opcode\n        *[MARKER],\n        # Added by SET_ADD which calls _PySet_AddTakeRef()\n        # without the critical section\n        EvilHash(0), EvilHash(1), EvilHash(2), EvilHash(3), EvilHash(4),\n        EvilHash(5)}\n\n    print(build_set)\n    print(\"Length:\", len(build_set))\n    if len(build_set) != 107:\n        raise Exception(\"race condition!\")\n\nthread = threading.Thread(target=worker)\nthread.start()\nfunc()\nthread.join()\n#dis.dis(func)\n```\n\nI expected the code to fail randomly, but so far I failed to trigger a race condition on Free Threading. Running the code on Python built with `--with-thread-sanitizer` doesn't show any warning.","author":{"url":"https://github.com/vstinner","@type":"Person","name":"vstinner"},"datePublished":"2026-06-26T14:06:20.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":5},"url":"https://github.com/152288/cpython/issues/152288"}