Rewrite asyncio.to_thread tests to use IsolatedAsyncioTestCase

[asvetlov]

BPO	47104
Nosy	@asvetlov, @1st1, @miss-islington, @Fidget-Spinner
PRs	bpo-47104: Rewrite asyncio.to_thread tests to use IsolatedAsyncioTestCase #32086 [3.10] bpo-47104: Rewrite asyncio.to_thread tests to use IsolatedAsyncioTestCase (GH-32086) #32087 [3.9] bpo-47104: Rewrite asyncio.to_thread tests to use IsolatedAsyncioTestCase (GH-32086) #32088

^{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 = None
created_at = <Date 2022-03-23.20:42:40.140>
labels = ['3.11', 'tests', '3.10', 'expert-asyncio']
title = 'Rewrite asyncio.to_thread tests to use IsolatedAsyncioTestCase'
updated_at = <Date 2022-04-06.17:56:00.892>
user = 'https://github.com/asvetlov'

bugs.python.org fields:

activity = <Date 2022-04-06.17:56:00.892>
actor = 'kj'
assignee = 'none'
closed = False
closed_date = None
closer = None
components = ['Tests', 'asyncio']
creation = <Date 2022-03-23.20:42:40.140>
creator = 'asvetlov'
dependencies = []
files = []
hgrepos = []
issue_num = 47104
keywords = ['patch']
message_count = 4.0
messages = ['415913', '415914', '416890', '416893']
nosy_count = 4.0
nosy_names = ['asvetlov', 'yselivanov', 'miss-islington', 'kj']
pr_nums = ['32086', '32087', '32088']
priority = 'normal'
resolution = None
stage = 'resolved'
status = 'open'
superseder = None
type = None
url = 'https://bugs.python.org/issue47104'
versions = ['Python 3.10', 'Python 3.11']

[asvetlov]

New changeset ff619c7 by Andrew Svetlov in branch 'main':
bpo-47104: Rewrite asyncio.to_thread tests to use IsolatedAsyncioTestCase (GH-32086)
ff619c7

[miss-islington]

New changeset 9e1bfd8 by Miss Islington (bot) in branch '3.10':
bpo-47104: Rewrite asyncio.to_thread tests to use IsolatedAsyncioTestCase (GH-32086)
9e1bfd8

[Fidget-Spinner]

Andrew, it seems that the post-commit CI has started failing after this commit. Specifically it says "test_asyncio failed (env changed)"

From this page,
https://github.com/python/cpython/commits/3.10?after=3856b4995ec0e632d47b733cdecb5183ac830568+34&branch=3.10

the Ubuntu CI link for that commit:
https://github.com/python/cpython/runs/5667655099?check_suite_focus=true

Failing buildbots on 3.10:
https://buildbot.python.org/all/#/release_status

Sorry that I can't help more. I'm not an asyncio expert so I'm not sure why this is triggering.

[Fidget-Spinner]

I forgot to specify, that this is *only* on 3.10, not main.

[kumaraditya303]

The test is failing on GH Actions too on 3.10 branch, it should be reverted until a fix is found.

See #91667

[gpshead]

Using python -m test.bisect_cmd test_asyncio --fail-env-changed in a 3.10 build I got:

== Tests result: ENV CHANGED ==

1 test altered the execution environment:
    test_asyncio

Total duration: 2.7 sec
Tests result: ENV CHANGED
ran 2 tests/4
exit 3
Tests failed: continuing with this subtest

Tests (2):
* test.test_asyncio.test_threads.ToThreadTests.test_to_thread_concurrent
* test.test_asyncio.test_unix_events.SelectorEventLoopSignalTests.test_add_signal_handler

Bisection failed after 101 iterations and 0:04:02

[gpshead]

0:00:00 load avg: 0.93 [1/2/1] test_asyncio.test_threads failed (env changed)
Warning -- threading._dangling was modified by test_asyncio.test_threads
  Before: {<weakref at 0x7f3b2e77fa10; to '_MainThread' at 0x7f3b2f107490>}
  After:  {<weakref at 0x7f3b2e77ffb0; to '_MainThread' at 0x7f3b2f107490>}

[gpshead]

I don't think this PR is the cause, just a code tweak that causes it to be exposed more often. Other tests lead to

0:09:20 load avg: 1.49 [427/427/1] test_asyncio failed (env changed) (1 min 22 sec)
Warning -- threading_cleanup() failed to cleanup -1 threads (count: 0, dangling: 1)
Warning -- Dangling thread: <_MainThread(MainThread, started 140300013757632)>

which is an environment change warning codepath from test.support.threading_helper which the test_asyncio.test_threads code no longer calls. from CI run https://github.com/python/cpython/runs/6069450575?check_suite_focus=true

[gpshead]

tracked in #91676 now as this issue's PR is merely a tickler of a latent problem, not the cause.

[kumaraditya303]

Closing as it is now tracked in #91676