René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"free-threading: PyImport_AddModuleRef race condition","articleBody":"# Bug report\n\n### Bug description:\n\nOn a NoGIL interpreter, if two threads call [PyImport_AddModuleRef](https://github.com/python/cpython/blob/a50822ff94ae0625f0b46480857fb141531c0688/Python/import.c#L339) for the same module name at the same time, and that module does not exist yet, it is possible that the two calls to the function return two different objects, only one of which is stored in `sys.modules`.\n\nThe race condition is actually in the helper function import_add_module:\nhttps://github.com/python/cpython/blob/a50822ff94ae0625f0b46480857fb141531c0688/Python/import.c#L319-L335\nif both threads call `PyMapping_GetOptionalItem` before either has called `PyObject_SetItem`, you will get two different modules.\n\n## Context\nEach SWIG module calls `PyImport_AddModuleRef` to retrieve or initialise a shared global state module with the hardcoded name `swig_runtime_data5`, and then stores its pointer in a module-local global variable.\n\nSWIG is actually immune from this race condition because the call to `PyImport_AddModuleRef` always happens inside `SWIG_init`, which is an alias for `PyInit_\u003cmodulename\u003e` which holds the GIL.\n\nThis race condition only affects lazy initialisation patterns, where `PyImport_AddModuleRef` is executed for the first time after module init.\n\n### CPython versions tested on:\n\nCPython main branch\n\n### Operating systems tested on:\n\n_No response_\n\n\u003c!-- gh-linked-prs --\u003e\n### Linked PRs\n* gh-141822\n* gh-141830\n\u003c!-- /gh-linked-prs --\u003e\n","author":{"url":"https://github.com/crusaderky","@type":"Person","name":"crusaderky"},"datePublished":"2025-08-05T15:24:12.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":4},"url":"https://github.com/137422/cpython/issues/137422"}