`multiprocessing.Pool` is incompatible with `concurrent.interpreters`

[bluetech]

Bug report

Bug description:

Trying to create a mutliprocessing.Pool in a sub-interpreter fails due to the use of daemon threads.

How I ran into it: as an experiment, I was trying to run pytest in a sub-interpreter (on its own test suite). One of the tests is using multiprocessing.Pool and fails. (There are other failures, particularly faulthandler and readline incompatibilities, but those are unrelated to this issue).

If mutliprocessing.Pool and concurrent.interpreters are fundamentally incompatible for some reason, perhaps it can be documented.

from concurrent import interpreters

interp = interpreters.create()

interp.exec("""
import multiprocessing
pool = multiprocessing.Pool()
""")

Traceback:

Traceback (most recent call last):
  File "x.py", line 5, in <module>
    interp.exec("""
    ~~~~~~~~~~~^^^^
    import multiprocessing
    ^^^^^^^^^^^^^^^^^^^^^^
    pool = multiprocessing.Pool()
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    """)
    ^^^^
  File "/home/ran/.local/share/uv/python/cpython-3.14.0-linux-x86_64-gnu/lib/python3.14/concurrent/interpreters/__init__.py", line 212, in exec
    raise ExecutionFailed(excinfo)
concurrent.interpreters.ExecutionFailed: RuntimeError: daemon threads are disabled in this interpreter

Uncaught in the interpreter:

Traceback (most recent call last):
  File "<script>", line 3, in <module>
  File "/home/ran/.local/share/uv/python/cpython-3.14.0-linux-x86_64-gnu/lib/python3.14/multiprocessing/context.py", line 119, in Pool
    return Pool(processes, initializer, initargs, maxtasksperchild,
                context=self.get_context())
  File "/home/ran/.local/share/uv/python/cpython-3.14.0-linux-x86_64-gnu/lib/python3.14/multiprocessing/pool.py", line 233, in __init__
    self._worker_handler.daemon = True
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/ran/.local/share/uv/python/cpython-3.14.0-linux-x86_64-gnu/lib/python3.14/threading.py", line 1207, in daemon
    raise RuntimeError('daemon threads are disabled in this interpreter')
RuntimeError: daemon threads are disabled in this interpreter

Link to relevant code:

cpython/Lib/multiprocessing/pool.py

Lines 226 to 253 in 6481539

	self._worker_handler = threading.Thread(
	target=Pool._handle_workers,
	args=(self._cache, self._taskqueue, self._ctx, self.Process,
	self._processes, self._pool, self._inqueue, self._outqueue,
	self._initializer, self._initargs, self._maxtasksperchild,
	self._wrap_exception, sentinels, self._change_notifier)
	)
	self._worker_handler.daemon = True
	self._worker_handler._state = RUN
	self._worker_handler.start()
	self._task_handler = threading.Thread(
	target=Pool._handle_tasks,
	args=(self._taskqueue, self._quick_put, self._outqueue,
	self._pool, self._cache)
	)
	self._task_handler.daemon = True
	self._task_handler._state = RUN
	self._task_handler.start()
	self._result_handler = threading.Thread(
	target=Pool._handle_results,
	args=(self._outqueue, self._quick_get, self._cache)
	)
	self._result_handler.daemon = True
	self._result_handler._state = RUN
	self._result_handler.start()

I'd also note that multiprocessing.Queue uses daemon thread:

cpython/Lib/multiprocessing/queues.py

Lines 178 to 186 in 6481539

	self._thread = threading.Thread(
	target=Queue._feed,
	args=(self._buffer, self._notempty, self._send_bytes,
	self._wlock, self._reader.close, self._writer.close,
	self._ignore_epipe, self._on_queue_feeder_error,
	self._sem),
	name='QueueFeederThread',
	daemon=True,
	)

CPython versions tested on:

3.14

Operating systems tested on:

Linux

Linked PRs

• gh-140057: Allow daemon threads in subinterpreters #140238

[picnixz]

@ericsnowcurrently Was it intended for multiprocessing pools to be compatible with subinterpreters?

[ZeroIntensity]

Now that #128639 is fixed, we can probably enable daemon threads in subinterpreters.

[xuantengh]

Raised a PR for this.