JDK-8357082 : Stabilize and add debug logs to CopyAreaOOB.java

[honkar-jdk]

CopyAreaOOB.java was failing intermittently on some platforms on CI but recently it started to fail more frequently on macos-aarch64 when the entire test suite runs.

Test failure is not reproducible when the test is run individually (multiple times) on CI.

Rewritten the test and added debug logs that will be helpful to figure out the issue.

• Added code to capture screenshot (entire screen as well as the frame) in case of failure.
• Since it tests multiple regions added a StringBuffer to consolidate the error logs before throwing RuntimeException.

---

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-8357082: Stabilize and add debug logs to CopyAreaOOB.java (Sub-task - P4)

Reviewers

• Sergey Bylokhov (@mrserb - Reviewer)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk.git pull/25279/head:pull/25279
$ git checkout pull/25279

Update a local copy of the PR:
$ git checkout pull/25279
$ git pull https://git.openjdk.org/jdk.git pull/25279/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 25279

View PR using the GUI difftool:
$ git pr show -t 25279

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk/pull/25279.diff

Using Webrev

Link to Webrev Comment

[bridgekeeper]

👋 Welcome back honkar! 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]

@honkar-jdk This change now passes all automated pre-integration checks.

ℹ️ This project also has non-automated pre-integration requirements. Please see the file CONTRIBUTING.md for details.

After integration, the commit message for the final commit will be:

8357082: Stabilize and add debug logs to CopyAreaOOB.java

Reviewed-by: serb

You can use pull request commands such as /summary, /contributor and /issue to adjust it as needed.

At the time when this comment was updated there had been 324 new commits pushed to the master branch:

• 0418b32: 8357696: Enhance code consistency: java.desktop/unix
• 5243f38: 8357924: Remove runtime/ErrorHandling/CreateCoredumpOnCrash.java from problem list for macosx-x64
• 1373ceb: 8356698: JFR: @contextual
• ... and 321 more: https://git.openjdk.org/jdk/compare/75504a35ee663625982f8aed508e5ff562643334...master

As there are no conflicts, your changes will automatically be rebased on top of these commits when integrating. If you prefer to avoid this automatic rebasing, please check the documentation for the /integrate command for further details.

➡️ To integrate this PR with the above commit message to the master branch, type /integrate in a new comment.

[openjdk]

@honkar-jdk 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

• 06: Full - Incremental (f42399ca)
• 05: Full - Incremental (7c288c91)
• 04: Full - Incremental (e9c5b3ac)
• 03: Full - Incremental (73861a9c)
• 02: Full - Incremental (707ed6df)
• 01: Full - Incremental (8cc8fa24)
• 00: Full (ee307064)

[honkar-jdk]

@mrserb @aivanov-jdk

I got the test to fail after some iterations. The issue is with the mouse pointer being at the place where robot.getRGB() is called as seen below. The fix is to move the mouse pointer to a location away from the test UI before taking a screenshot.

This happens on macos-aarch64 and I think the mouse pointer remains at the location last moved to by previous test (It explains why it happens only when certain tests run before it during entire test suite run and not when the test is run individually).
I'm still not sure as to why the mouse pointer is captured during screencapture on macos-aarch64 as I don't see it happening on macosx-x64.

[mrserb]

The new version looks like a different test, previously we copyArea from the paint method for the awt frame, then call sync to flush all rendering, then check the rendering on screen, and all of that on "each render frame".

Now it will draw everything in paint then sometime later(or even after a few repaints) we check the actual rendering(waitForIdle/delay(100))....

[honkar-jdk]

@mrserb

Added back Toolkit.sync() and canvas capture to paint method. Although I'm not sure if I understood "each render frame" correctly. Can you please explain?

        BufferedImage capture = null;
        try {
            Thread.sleep(500);
            if (robot == null) robot = new Robot();
            Point pt1 = getLocationOnScreen();
            Rectangle rect = new Rectangle(pt1.x, pt1.y, 400, 400);
            capture = robot.createScreenCapture(rect);
        } catch (Exception e) {
            throw new RuntimeException("Problems handling Robot");
        }
        // Test pixels
        testRegion(capture, "green",          0,   0, 400,  10, 0xff00ff00);
        testRegion(capture, "original red",   0,  10,  50, 400, 0xffff0000);
        testRegion(capture, "background",    50,  10,  60, 400, 0xff000000);
        testRegion(capture, "in-between",    60,  10, 110,  20, 0xff000000);
        testRegion(capture, "copied red",    60,  20, 110, 400, 0xffff0000);
        testRegion(capture, "background",   110,  10, 400, 400, 0xff000000);

Isn't capture going to be the same image throughout? I'm not sure why it is passed as an extra argument to testRegion. It can be saved once as static BufferedImage var and the same can be used for testing or am I missing something?

[mrserb]

Added back Toolkit.sync() and canvas capture to paint method. Although I'm not sure if I understood "each render frame" correctly. Can you please explain?

The paint method may be invoked multiple times after the frame becomes visible, so the validation logic inside paint will be executed for each rendering data.

[honkar-jdk]

oh okay, although when I debug I don't see paint method being called multiple times. Probably the case when frame is resized or screen configuration changes?

[mrserb]

Yes, and validation of each frame will confirm no garbage on the screen.

[honkar-jdk]

@mrserb @aivanov-jdk Can you please re-review when you get a chance?
This issue fix is targeted for jdk-25 and I was planning on integrating it on Monday if the updated test looks good.

[honkar-jdk]

/integrate

[openjdk]

Going to push as commit e490b4f.
Since your change was applied there have been 324 commits pushed to the master branch:

• 0418b32: 8357696: Enhance code consistency: java.desktop/unix
• 5243f38: 8357924: Remove runtime/ErrorHandling/CreateCoredumpOnCrash.java from problem list for macosx-x64
• 1373ceb: 8356698: JFR: @contextual
• ... and 321 more: https://git.openjdk.org/jdk/compare/75504a35ee663625982f8aed508e5ff562643334...master

Your commit was automatically rebased without conflicts.

[openjdk]

@honkar-jdk Pushed as commit e490b4f.

💡 You may see a message that your pull request was closed with unmerged commits. This can be safely ignored.