Repository navigation
Fix potential flakiness in test_run_until_complete_baseexception - #100148
kumaraditya303 merged 1 commit into
Conversation
Sorry, something went wrong.
Done & merged. |
Sorry, something went wrong.
|
This is indeed better than sleeping for a small delay. |
Sorry, something went wrong.
|
Yup, a very insightful solution. Thanks @fantix ! |
Sorry, something went wrong.
|
This PR appears to have resulted in an empty commit, likely because the PR branch was not fully up to date with main: a9bad4d :/ |
Sorry, something went wrong.
Very weird. Especially since when I check out main and pull to head, the change is there, and attributed to this PR. So maybe the problem is with the GitHub UI? I saw some mention of enpty branches on Discord but chose to ignore it. |
Sorry, something went wrong.
|
Oh, weird. Yes, I see the same. It must be something wrong with the GitHub UI. My sincere apologies for the false accusation, @fantix! |
Sorry, something went wrong.
|
No worries @AlexWaygood , thank you for being careful ❤️ |
Sorry, something went wrong.
|
Hm, there's definitely something weird though. I see three commits for this PR in the main branch: commit a9bad4d28413666edc57551dd439bca6a6a59dd9
Author: Fantix King <fantix.king@gmail.com>
Date: Sat Dec 10 02:05:24 2022
Fix potential flakiness in `test_run_until_complete_baseexception` (#100148)
commit f28e537125941d9f8894ec01b5e87252e824a688
Author: Fantix King <fantix.king@gmail.com>
Date: Sat Dec 10 02:04:50 2022
Fix potential flakiness in `test_run_until_complete_baseexception` (#100148)
commit 50b08d5b5fd3e280b99b97a42d07089a64bf12c6
Author: Fantix King <fantix.king@gmail.com>
Date: Sat Dec 10 02:04:22 2022
Fix potential flakiness in `test_run_until_complete_baseexception` (#100148)
diff --git a/Lib/test/test_asyncio/test_base_events.py b/Lib/test/test_asyncio/test_base_events.py
index 532a7e5385..3b4026cb73 100644
--- a/Lib/test/test_asyncio/test_base_events.py
+++ b/Lib/test/test_asyncio/test_base_events.py
@@ -868,7 +868,7 @@ def func():
self.loop.stop()
func.called = True
func.called = False
- self.loop.call_later(0.01, func)
+ self.loop.call_soon(self.loop.call_soon, func)
self.loop.run_forever()
self.assertTrue(func.called)
Two of those are empty. What kind of GitHub hiccup could have caused that??? |
Sorry, something went wrong.
|
Possibly we can report this to GitHub somehow or ask anyone from GitHub to look at this? I don't know how to do it though. |
Sorry, something went wrong.
Refs #24477, use a more reliable way to run the test.
If the covering issue happens again, the loop would be requested to stop in the first iteration and got stopped after the first iteration. In that case, our func() in the second iteration won't be called.