concurrent.futures: ProcessPoolExecutor.shutdown(wait=True) should wait for the call queue thread

[vstinner]

BPO	31326
Nosy	@vstinner
PRs	bpo-31326: ProcessPoolExecutor waits for the call queue thread #3265 [3.6] bpo-31326: ProcessPoolExecutor waits for the call queue thread (#3265) #3309

^{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 2017-09-05.16:43:29.207>
created_at = <Date 2017-09-01.16:45:47.113>
labels = ['3.7', 'performance']
title = 'concurrent.futures: ProcessPoolExecutor.shutdown(wait=True) should wait for the call queue thread'
updated_at = <Date 2017-09-05.16:43:29.206>
user = 'https://github.com/vstinner'

bugs.python.org fields:

activity = <Date 2017-09-05.16:43:29.206>
actor = 'vstinner'
assignee = 'none'
closed = True
closed_date = <Date 2017-09-05.16:43:29.207>
closer = 'vstinner'
components = []
creation = <Date 2017-09-01.16:45:47.113>
creator = 'vstinner'
dependencies = []
files = []
hgrepos = []
issue_num = 31326
keywords = []
message_count = 4.0
messages = ['301135', '301139', '301154', '301332']
nosy_count = 1.0
nosy_names = ['vstinner']
pr_nums = ['3265', '3309']
priority = 'normal'
resolution = 'fixed'
stage = 'resolved'
status = 'closed'
superseder = None
type = 'resource usage'
url = 'https://bugs.python.org/issue31326'
versions = ['Python 3.7']

[vstinner]

concurrent.futures.ProcessPoolExecutor.shutdown(wait=True) doesn't wait for the call queue thread, and so test_concurrent_futures randomly emits warnings about dangling threads, especially with PR 3138 of bpo-31234.

Attached PR fixes the issue.

[vstinner]

See also bpo-31249 and commit bc61315: "Fix ref cycle in ThreadPoolExecutor".

[vstinner]

New changeset b713adf by Victor Stinner in branch 'master':
bpo-31326: ProcessPoolExecutor waits for the call queue thread (bpo-3265)
b713adf

[vstinner]

Antoine Pitrou: "I don't terribly like this being backported. It does not fix any user-visible problem AFAIK."

#3309 (comment)

Ok, let's close this issue.