Deadlock with writing to stderr from forked process

[ionelmc]

BPO	22697
Nosy	@nirs, @serhiy-storchaka
Superseder	bpo-6721: Locks in the standard library should be sanitized on fork

^{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:

assignee = None
closed_at = <Date 2014-11-02.15:44:19.804>
created_at = <Date 2014-10-22.13:59:27.083>
labels = ['interpreter-core', 'expert-IO']
title = 'Deadlock with writing to stderr from forked process'
updated_at = <Date 2014-11-02.15:45:41.537>
user = 'https://bugs.python.org/ionelmc'

bugs.python.org fields:

activity = <Date 2014-11-02.15:45:41.537>
actor = 'serhiy.storchaka'
assignee = 'none'
closed = True
closed_date = <Date 2014-11-02.15:44:19.804>
closer = 'neologix'
components = ['Interpreter Core', 'IO']
creation = <Date 2014-10-22.13:59:27.083>
creator = 'ionelmc'
dependencies = []
files = []
hgrepos = []
issue_num = 22697
keywords = []
message_count = 5.0
messages = ['229825', '230470', '230489', '230493', '230494']
nosy_count = 4.0
nosy_names = ['nirs', 'ionelmc', 'neologix', 'serhiy.storchaka']
pr_nums = []
priority = 'normal'
resolution = 'duplicate'
stage = 'resolved'
status = 'closed'
superseder = '6721'
type = None
url = 'https://bugs.python.org/issue22697'
versions = ['Python 3.3', 'Python 3.4']

[ionelmc]

Example code:

import os
import sys
import threading


def run():
    sys.stderr.write("in parent thread\n")

threading.Thread(target=run).start()
pid = os.fork()
if pid:
os.waitpid(pid, 0)
else:
sys.stderr.write("in child\n")

To run: while python3 deadlock.py; do; done

Note: does not reproduce if ran with python -u (unbuffered)

[nirs]

This is a duplicate of http://bugs.python.org/issue6721

[ionelmc]

Serhiy, I don't think this is a duplicate. Odd that you closed this without any explanation.

This happens in a internal lock in cpython's runtime, while the other bug is about locks used in the logging module (which are very different).

[neologix]

Maries Ionel Cristian added the comment:

Serhiy, I don't think this is a duplicate. Odd that you closed this without any explanation.

This happens in a internal lock in cpython's runtime, while the other bug is about locks used in the logging module (which are very different).

Yes, this is a duplicate. Whether the lock is an internal or a user-created one doesn't change anything, it's always the same problem of having a lock locked by another thread at the time of the fork().

[serhiy-storchaka]

Sorry. I thought bpo-6721 is more general, not just about the logging module.