Feature or enhancement

Add the rtype_cache (which is a set that includes all leaked objects of a particular type) to the warnings.warn message in Lib/multiprocessing/resource_tracker.py when any leaked objects are found to make debugging easier:

Pitch

When the resource_tracker module (Lib/multiprocessing/resource_tracker.py) finds leaked objects, the finally block in the main function includes the type of leaked objects and the number of leaked objects, but does not actually include the objects that are leaked. Adding the set of rtype_cache to the warnings.warn message will make debugging much more useful, as the names of the leaked objects could help more quickly identify what was leaked and/or why the leaked object was not properly cleaned up.

The permalink attached above links directly to the relevant code as it is currently implemented, but I'm adding it again below with some surrounding code for reference here:

    finally:
        # all processes have terminated; cleanup any remaining resources
        for rtype, rtype_cache in cache.items():
            if rtype_cache:
                try:
                    warnings.warn('resource_tracker: There appear to be %d '
                                  'leaked %s objects to clean up at shutdown' %
                                  (len(rtype_cache), rtype))
                except Exception:
                    pass

Previous discussion

A recent example of an issue where the additional context would have been useful is #104090, in which the current warning message is /workspace/cpython/Lib/multiprocessing/resource_tracker.py:224: UserWarning: resource_tracker: There appear to be 6 leaked semaphore objects to clean up at shutdown.

Linked PRs

• gh-106739: Add rtype_cache to warnings.warn message when leaked objects found #106740