https://bugs.winehq.org/show_bug.cgi?id=52083
Bug ID: 52083 Summary: Some test_TimerResolution() tests fail randomly in ntdll:time Product: Wine Version: unspecified Hardware: x86-64 OS: Windows Status: NEW Severity: normal Priority: P2 Component: ntdll Assignee: wine-bugs@winehq.org Reporter: fgouget@codeweavers.com
Some test_TimerResolution() tests fail randomly in ntdll:time:
time.c:293: Test failed: expected requested timer resolution 156250, got 10006 time.c:298: Test failed: expected requested timer resolution 20000, got 10006 time.c:310: Test failed: expected new timer resolution 20000, got 10006 time.c:318: timer resolution: 5000(max) <= 5003(cur) <= 20000(prev) <= 156250(min) time.c:322: Test failed: expected new timer resolution 20000, got 5003 time.c:322: Test failed: NtQueryTimerResolution() expected timer resolution 5003, got 20000 time.c:328: Test failed: expected requested timer resolution 5003, got 75011
https://test.winehq.org/data/patterns.html#ntdll:time
The number of failures varies and six seems to be a maximum. This appears to only impact Windows 10 1709+ but impacts both the TestBot VMs and real test machines (cw-gtx560 and cw-rx460).
Based on the current failure pattern it looks like the failures started on 2021-09-30. That is, there was no failure for any of the 252 builds from 2021-09-21 to 2021-09-29 and there has been at least one failure for every build since.
Yet the failure is reproducible with the commit that introduced the tests (6015fee5b218), and also with the one that fixed some of the issues a few days later (b2c89b1a41fa):
commit 6015fee5b21870ddd57477843354173d2e7f9696 Author: François Gouget fgouget@codeweavers.com AuthorDate: Thu Jul 29 17:24:15 2021 +0200
ntdll/tests: Add some Nt{Query,Set}TimerResolution() tests.
Signed-off-by: Francois Gouget fgouget@codeweavers.com Signed-off-by: Alexandre Julliard julliard@winehq.org
https://bugs.winehq.org/show_bug.cgi?id=52083
François Gouget fgouget@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |regression, source, | |testcase Regression SHA1| |6015fee5b21870ddd5747784335 | |4173d2e7f9696
https://bugs.winehq.org/show_bug.cgi?id=52083
--- Comment #1 from François Gouget fgouget@codeweavers.com --- Created attachment 71097 --> https://bugs.winehq.org/attachment.cgi?id=71097 Extended ntdll:time failure pattern
So the ntdll:time failures during WineTest did start on 2021-09-30, long after the tests were added (see attached extended ntdll:time failure pattern).
The failures can still be reproduced with earlier builds but I suspect that's only after running other Wine tests. That's something I need to retest and, if confirmed, it could help to identify which other test is causing the failures.
https://bugs.winehq.org/show_bug.cgi?id=52083
François Gouget fgouget@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Resolution|--- |WORKSFORME Status|NEW |RESOLVED
--- Comment #2 from François Gouget fgouget@codeweavers.com --- This is fixed. The last random failures happened on 2022-10-10 despite there being no change to the test or the VMs on that date.
Now the test fails systematically on Windows 11 but that's a different failure mode.
https://bugs.winehq.org/show_bug.cgi?id=52083
Gijs Vermeulen gijsvrm@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #3 from Gijs Vermeulen gijsvrm@gmail.com --- Closing WORKSFORME.
-
WineHQ Bugzilla