multiprocessing: shared manager Pool fails with AttributeError

[giampaolo]

BPO	35919
Nosy	@pitrou, @giampaolo, @applio

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields: ```python assignee = None closed_at = None created_at = labels = ['3.7', '3.8', 'type-bug', 'library', '3.9'] title = 'multiprocessing: shared manager Pool fails with AttributeError' updated_at = user = 'https://github.com/giampaolo' ``` bugs.python.org fields: ```python activity = actor = 'Felipe A. Hernandez' assignee = 'none' closed = False closed_date = None closer = None components = ['Library (Lib)'] creation = creator = 'giampaolo.rodola' dependencies = [] files = [] hgrepos = [] issue_num = 35919 keywords = [] message_count = 2.0 messages = ['334962', '373427'] nosy_count = 4.0 nosy_names = ['pitrou', 'giampaolo.rodola', 'davin', 'Felipe A. Hernandez'] pr_nums = [] priority = 'normal' resolution = None stage = None status = 'open' superseder = None type = 'behavior' url = 'https://bugs.python.org/issue35919' versions = ['Python 3.5', 'Python 3.6', 'Python 3.7', 'Python 3.8', 'Python 3.9'] ```

[giampaolo]

import multiprocessing
import multiprocessing.managers

def f(n):
    return n * n

def worker(pool):
    with pool:
        pool.apply_async(f, (10, ))

manager = multiprocessing.managers.SyncManager()
manager.start()
pool = manager.Pool(processes=4)
proc = multiprocessing.Process(target=worker, args=(pool, ))
proc.start()
proc.join()

This is related to BPO-35917 and it fails with:

Process Process-2:
Traceback (most recent call last):
  File "/home/giampaolo/cpython/Lib/multiprocessing/process.py", line 302, in _bootstrap
    self.run()
  File "/home/giampaolo/cpython/Lib/multiprocessing/process.py", line 99, in run
    self._target(*self._args, **self._kwargs)
  File "foo.py", line 54, in worker
    pool.apply_async(f, (10, ))
  File "<string>", line 2, in apply_async
  File "/home/giampaolo/cpython/Lib/multiprocessing/managers.py", line 802, in _callmethod
    proxytype = self._manager._registry[token.typeid][-1]
AttributeError: 'NoneType' object has no attribute '_registry'

[a644cf03-66e9-4772-bfe4-dbdc508bbd35]

import traceback
import multiprocessing.managers

class MyManager(multiprocessing.managers.SyncManager):
    pass

class DictList(multiprocessing.managers.BaseProxy):
    _method_to_typeid_ = {'__getitem__': 'dict'}

    def __getitem__(self, key):
        return self._callmethod('__getitem__', (key,))

MyManager.register('DictList', None, DictList)

with MyManager() as manager:

    nested = manager.DictList([{'hello': 'world'}])
    print(nested[0]['hello'])  # world

    proxy = manager.list([nested])
    try:
        print(proxy[0][0]['hello'])
    except AttributeError:
        traceback.print_exc()
        print("""
        Bug: AttributeError: ProxyBase._callmethod is None

---

    This error is raised because proxies returned as #RETURN messages
    have no manager reference, which is required to resolve typeids
    (using BaseManager._registry).

    Only proxies returned as #PROXY messages will be fully functional.

    This is an undocumented current implementation limitation.

    Fix (proposal)
    \--------------

    Include the manager class (not the instance) as part of the proxy
    serialization in BaseProxy.__reduce__, as BaseManager._registry is
    a class variable.

    Note: #PROXY message protocol can be also replaced by regular proxy
          serialization after this fix, resulting on simpler codebase.
    """)

[bryantChhun]

has this problem been solved? I am still seeing this in 3.9.18