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

[Alexey Ivanov]

On Fri, 19 Sep 2025 08:05:19 GMT, Tejesh R <tr at openjdk.org> wrote:

> > What are the frame coordinates when the test fails? Do these coordinates correspond to those on a screenshot if you add taking one?
> 
> Don't know the reason for the failure since it's intermittent, not seen any failure while running in mach5. Frame coordinates during failure as per JBS - "The frame location is wrong! Current: 250, 287; expected: 250, 250" slight change in Y position.

It could be because of something on the screen, like a panel. Ubuntu has a menu panel at the top and a dock on the left, if you set the location of the frame to (0, 0), the frame will not be at (0, 0).

So, we should understand better what happens and why the frame was not positioned at the expected location.

> Not recent failures seen too, so stabilization would be better fix.

No, it wouldn't be because it could break the original intention of the test.

Since there have been no recent failures of the test, it's even more suspicious. Instead of moving to using `invokeAndWait`, I suggest *capturing a screenshot* before the frame is disposed of so that we could see what's on the screen when the test fails.

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

PR Comment: https://git.openjdk.org/jdk/pull/27366#issuecomment-3311596660