[3.10] gh-91118: Fix docstrings that do not honor --without-doc-strings (GH-31769)

[arhadthedev]

Co-authored-by: Éric [email protected]
Co-authored-by: Jelle Zijlstra [email protected]
(cherry picked from commit a573cb2)

Co-authored-by: Oleg Iarygin [email protected]

[arhadthedev]

A failed test has no relation with changed docstrings:

== Tests result: ENV CHANGED ==

417 tests OK.

10 slowest tests:
- test_concurrent_futures: 2 min 41 sec
- test_multiprocessing_spawn: 2 min 39 sec
- test_gdb: 2 min 1 sec
- test_multiprocessing_forkserver: 1 min 38 sec
- test_asyncio: 1 min 29 sec
- test_multiprocessing_fork: 1 min 21 sec
- test_peg_generator: 1 min 13 sec
- test_signal: 51.0 sec
- test_subprocess: 43.9 sec
- test_venv: 35.7 sec

1 test altered the execution environment:
    test_asyncio

9 tests skipped:
    test_devpoll test_ioctl test_kqueue test_msilib test_startfile
    test_winconsoleio test_winreg test_winsound test_zipfile64

Total duration: 9 min 15 sec
Tests result: ENV CHANGED
make: *** [Makefile:1255: buildbottest] Error 3
Error: Process completed with exit code 2.

[AlexWaygood]

A failed test has no relation with changed docstrings:

The test is known to be flaky on the backport branches, for some reason (I think @JelleZijlstra was having trouble with it recently). Not sure that anybody's worked out what the problem is, though.

[AlexWaygood]

A failed test has no relation with changed docstrings:

The test is known to be flaky on the backport branches, for some reason (I think @JelleZijlstra was having trouble with it recently). Not sure that anybody's worked out what the problem is, though.

This looks like it should fix test_asyncio on 3.10:

• [3.10] Revert "[3.10] bpo-47104: Rewrite asyncio.to_thread tests to use IsolatedAsyncioTestCase (GH-32086)" #91667

[arhadthedev]

@AlexWaygood A non-reverting fix (#91680) got merged so I've updated this PR. Update: Ubuntu tests now pass.