RFR: 8305915: java/awt/Frame/FrameLocation/FrameLocation.java fails with "The frame location is wrong!" [v4]

[Alexey Ivanov]

On Fri, 26 Sep 2025 21:33:34 GMT, Damon Nguyen <dnguyen at openjdk.org> wrote:

>>> > > Error
>>> > > ⚠️ 8305915 is used in problem lists: [test/jdk/ProblemList.txt]
>>> > 
>>> > 
>>> > However, the problem list only contains another test under the 8305915 bug ID.
>>> > > java/awt/Frame/SizeMinimizedTest.java 8305915 linux-x64
>>> > 
>>> > 
>>> > And that is what we should pay attention to.
>>> 
>>> We should resolve this problem.
>>> 
>>> Tejesh problem-listed `SizeMinimizedTest.java` in #18448. The PR description states:
>>> 
>>> > Automated the test SizeMinimizedTest.java. Since the test is failing Linux arch (Maybe related to the bug - [JDK-8305915](https://bugs.openjdk.org/browse/JDK-8305915)), the test is problem listed for Linux arch until the issue is fixed/further investigation is done.
>>> 
>>> If `SizeMinimizedTest.java` still fails, it's time to file a new bug for it. A sub-task of the new bug should be used to fix up the bugid in the problem list.
>> 
>> I tested `SizeMinimizedTest.java` running in CI and no failure is observed. It's true that I had added the test for different bug ID long time which is wrong. Should we just move it to another JBS and PL it while removing 8305915 PL ? I guess we should handle both the issue separately with different BugID ?
>
>> I tested SizeMinimizedTest.java running in CI and no failure is observed. It's true that I had added the test for different bug ID long time which is wrong. Should we just move it to another JBS and PL it while removing 8305915 PL ? I guess we should handle both the issue separately with different BugID ?
> 
> At the very least, this info should be added to the current JBS issue. Maybe update the title and the summary of the issue to also include the removal of `SizeMinimizedTest.java` from the PL.

> > As @DamonGuy mentioned, a better way to deal with [JDK-8305915](https://bugs.openjdk.org/browse/JDK-8305915) could be closing it as _Cannot reproduce_ and submitting a new bug to update `FrameLocation.java` and remove `SizeMinimizedTest.java` from the problem list.
> 
> Probably I'll capture screenshot on failure and deproblemlist it. If I close the bug as _Cannot reproduce_ then it remains in problem listing, so better way is to de-problemlist it and add screen capture.

That's the point—you cannot reproduce the failure of `FrameLocation.java`, can you?

Thus, the correct resolution is *Cannot Reproduce*.

Adding a screenshot is not fixing the failure reported in [JDK-8305915](https://bugs.openjdk.org/browse/JDK-8305915), it's an improvement for the test to have the more information to analyse the failure if it occurs again. This is why it's a new issue.

8305915 has to be removed from the PL, submit a new bug, or rather create a subtask of JDK-8305915, to remove both `FrameLocation.java` and `SizeMinimizedTest.java` from the problem list.

I strongly believe this approach creates a clearer bug history in JBS and clearer changelog in the JDK. This also makes backporting easier as well as dealing with test failures in older releases.

Yes, submitting two new bugs creates more work but we're not in a rush here, and I think *it's worth the time*.

I can submit the bugs for you if you prefer so.

-------------

PR Comment: https://git.openjdk.org/jdk/pull/27366#issuecomment-3376307327