[Bug 52994] New: mstask:task_trigger - test_GetNextRunTime() fails in Wine on date change

[WineHQ Bugzilla]

https://bugs.winehq.org/show_bug.cgi?id=52994

Bug ID: 52994
           Summary: mstask:task_trigger - test_GetNextRunTime() fails in
                    Wine on date change
           Product: Wine
           Version: unspecified
          Hardware: x86-64
                OS: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: mstask
          Assignee: [e-mail]
          Reporter: [e-mail]
      Distribution: ---

Created attachment 72360
  --> https://bugs.winehq.org/attachment.cgi?id=72360
Make the failure 100% reproducible

mstask:task_trigger - test_GetNextRunTime() fails in Wine on date change:

task_trigger.c:434: old: 5/5/2022 wday 0 23:58:00.000
task_trigger.c:443: new: 6/5/2022 wday 5 0:03:00.000
task_trigger.c:533: Test failed: got 0x41307
task_trigger.c:534: Test failed: got 0/0/0 wday 0 0:00:00

https://test.winehq.org/data/patterns.html#mstask:task_trigger

What seems to be happening is that ITask_GetNextRunTime() returns
SCHED_S_TASK_NO_VALID_TRIGGERS for TASK_TIME_TRIGGER_ONCE tasks where the
trigger is on the next day.

So I suspect the bug is in the ITask_GetNextRunTime() implementation rather
than in the test itself.

Note: Initially I was a bit alarmed by the week day change in the traces. But
in fact trigger_add_ms() just sets it to 0 so that's what we get on the "old:"
line. Then FileTimeToSystemTime() fully initializes it so we get the correct
value on the "new:" line. That's why it changes in surprising ways.

Luckily the test and Itask_GetNextRunTime() implementation were both added in
the commit below and I was also able to confirm that the failure was already
present:

commit ac322334fa4d6740e89c6f19361a66f063103a46
Author:     Dmitry Timoshkov 
AuthorDate: Mon May 21 18:17:47 2018 +0800

mstask: Implement ITask::GetNextRunTime() for ONCE trigger type.

Signed-off-by: Dmitry Timoshkov 
    Signed-off-by: Alexandre Julliard