8158801: [TEST_BUG] Mixing tests fail because of focus workaround trick

[kboulanou]

Many Mixing tests failed because the work around click lands on the minimizing area in the window control and causes the tests to fail.

This fix changes the width of base frames which allows most of tests to pass.

---

Progress

• Change must be properly reviewed (1 review required, with at least 1 Reviewer)
• Change must not contain extraneous whitespace
• Commit message must refer to an issue

Issue

• JDK-8158801: [TEST_BUG] Mixing tests fail because of focus workaround trick (Bug - P4)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk.git pull/25971/head:pull/25971
$ git checkout pull/25971

Update a local copy of the PR:
$ git checkout pull/25971
$ git pull https://git.openjdk.org/jdk.git pull/25971/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 25971

View PR using the GUI difftool:
$ git pr show -t 25971

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk/pull/25971.diff

Using Webrev

Link to Webrev Comment

[bridgekeeper]

👋 Welcome back kboulanou! A progress list of the required criteria for merging this PR into master will be added to the body of your pull request. There are additional pull request commands available for use with this pull request.

[openjdk]

❗ This change is not yet ready to be integrated.
See the Progress checklist in the description for automated requirements.

[openjdk]

@kboulanou The following label will be automatically applied to this pull request:

• client

When this pull request is ready to be reviewed, an "RFR" email will be sent to the corresponding mailing list. If you would like to change these labels, use the /label pull request command.

[mlbridge]

Webrevs

• 26: Full - Incremental (a1788ae0)
• 25: Full - Incremental (7b46aef4)
• 24: Full - Incremental (9358a9d8)
• 23: Full - Incremental (449c397e)
• 22: Full - Incremental (ffbe6963)
• 21: Full - Incremental (3fac9c1e)
• 20: Full - Incremental (4df40993)
• 19: Full - Incremental (b39fc42d)
• 18: Full - Incremental (3f8742e7)
• 17: Full - Incremental (abc0797c)
• 16: Full - Incremental (fcde066a)
• 15: Full - Incremental (ce91a9d7)
• 14: Full - Incremental (890430eb)
• 13: Full - Incremental (d986747c)
• 12: Full - Incremental (3cae7357)
• 11: Full - Incremental (b558508f)
• 10: Full - Incremental (fdd052fc)
• 09: Full - Incremental (f99d5599)
• 08: Full - Incremental (f041ed29)
• 07: Full - Incremental (655487d0)
• 06: Full - Incremental (0b53b89e)
• 05: Full - Incremental (35cf3e5b)
• 04: Full - Incremental (3b0c2462)
• 03: Full - Incremental (6fe853a8)
• 02: Full - Incremental (9c7b4c70)
• 01: Full - Incremental (fa864b2f)
• 00: Full (f68973cc)

[prsadhuk]

What do you mean by "focus workaround trick" and " minimizing area in the window control"? Are you seeing click is happening on the minimize button on the frame and frame getting iconified?
In that case instead of changing the frame width/height, guess you can set with frame.setUndecorated to have undecorated windows..Also, I have seen this test renders the frame on the top-left corner, it will be good to keep it at the center of the frame using setLocationRelativeTo(null)

[kboulanou]

In Windows machines, the click lands on the reduce control of the window, and this causes the windows to be iconified and the test fails.

I have tested with undecorated frames, but the test continues to fail. The test needs to click on the middle of the Windows title bar (Outside of the controls) to pass.

I will set the frames (The background undecorated frame, and the heavyweight component frames) on center as you have suggested.

[prsadhuk]

In Windows machines, the click lands on the reduce control of the window, and this causes the windows to be iconified and the test fails.

WHich test in particular? or is it all tests have this behavior?

The test needs to click on the middle of the Windows title bar (Outside of the controls) to pass.

I tried JMenuBarOverlapping.java but I dont see any click on title bar..The cursor in on top-left of "Test Menu" menu..
If clicking on the title bar is needed to make it gain focus, probably you can try with calling frame.requestFocus and use CountDownLatch with timeout to wait till FOCUS_GAINED event is received and release the latch in focusGained() and then start the test

[myankelev]

Minor: could you please update the copyright years?

[kboulanou]

@prsadhuk I have reviewed the code again and I can point out where we do have the workaround click. It can be found in class SimpleOverlappingTestBase, in method performTest. Following is an excerpt of the source code:

/* this is a workaround for certain jtreg(?) focus issue:
           tests fail starting after failing mixing tests but always pass alone.
         */
        JFrame ancestor = (JFrame)(testedComponent.getTopLevelAncestor());
        if( ancestor != null ) {
            Point ancestorLoc = ancestor.getLocationOnScreen();
            ancestorLoc.translate(isOel7orLater() ? 5 :
                                             ancestor.getWidth() / 2 - 15, 2);
            robot.mouseMove(ancestorLoc.x, ancestorLoc.y);
            Util.waitForIdle(robot);
            robot.mousePress(InputEvent.BUTTON1_MASK);
            robot.delay(50);
            robot.mouseRelease(InputEvent.BUTTON1_MASK);
            Util.waitForIdle(robot);
        }

I will use requestFocus as you have suggested and see if I get tests to pass.

[kboulanou]

I will update copyright years for changed classes.

[kboulanou]

@prsadhuk after applying your suggested change, I can see that it has fixed the issue for most of the tests but still the following tests didn't pass on Windows compared with the original fix of changing frame width :

java/awt/Mixing/AWT_Mixing/JProgressBarInGlassPaneOverlapping.java
java/awt/Mixing/AWT_Mixing/JScrollBarInGlassPaneOverlapping.java
java/awt/Mixing/AWT_Mixing/JSliderInGlassPaneOverlapping.java
java/awt/Mixing/AWT_Mixing/JTextAreaInGlassPaneOverlapping.java
java/awt/Mixing/AWT_Mixing/JTextFieldInGlassPaneOverlapping.java
java/awt/Mixing/AWT_Mixing/JToggleButtonInGlassPaneOverlapping.java
java/awt/Mixing/AWT_Mixing/OpaqueOverlapping.java

I will check why these tests had failed.

[openjdk]

@kboulanou Please do not rebase or force-push to an active PR as it invalidates existing review comments. Note for future reference, the bots always squash all changes into a single commit automatically as part of the integration. See OpenJDK Developers’ Guide for more information.

[kboulanou]

@prsadhuk All tests passed after the applying both changes: Awaiting time set to one second and checking if component has focus first before requesting focus.

[aivanov-jdk]

Strangely enough, these changes make no difference for me when I run Mixing tests locally. The same number of tests fails with and without this fix. However, it seems to affect the CI.

Additionally, I still see the windows / frames are minimised on Windows. I don't think this is expected, is it? Thus, your previous suggestion for increasing the size of the windows looks reasonable to me.

[kboulanou]

@aivanov-jdk both increasing the width of the frames and the requesting focus in frames worked for me. The test creates many frames and once a test completed for an AWT component, the next frame is created and shown on top of previously created frame (Maybe we can improve this by disposing the frame that finished testing ...). For each AWT component, a mixing frame is created. I will try on Windows and see if I can catch anything.

[kboulanou]

@aivanov-jdk I have tested all tests with the new modifications and they passed.

[kboulanou]

@aivanov-jdk I have tested on a Windows 11 Pro, before and after the fix in this PR, and I can see only 3 tests failing after an initial 26 failures.

[kboulanou]

@prsadhuk I have updated this PR to center all possible tests in the middle of the screen.

[kboulanou]

Centered some missed frames. Tested all de-problem listed tests on all platforms and tests passed.

[manukumarvs]

Are we disposing this frame somewhere(it's applicable to all the test files)?

[kboulanou]

@manukumarvs in the original test, frames are not disposed. I will include disposing frame after testing the heavyweight component.

[kboulanou]

It seems there are two types of testing, some use one frame (GlassPane) and others uses multiple frames. I have updated code to dispose of multiple frames. Currently testing my changes.