8373102: com/sun/jdi/MethodInvokeWithTraceOnTest.java can fail with ObjectCollectedException when run with a small heap

[plummercj]

With JDI tracing enabled, an attempt to trace an event when using SUSPEND_NONE can result in an ObjectCollectedException because the thread is still running and objects that are part of the trace can be collected already. In this case were were dealing with trying to print the thread name for a VirtualThreadEndEvent, and the virtual thread has already exited and been gc'd.

Tested by running test case before change (verified failure) and after change (50 runs without a failure). Will also run tier1 and tier5 svc just to be sure.

---

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-8373102: com/sun/jdi/MethodInvokeWithTraceOnTest.java can fail with ObjectCollectedException when run with a small heap (Bug - P4)

Reviewers

• Serguei Spitsyn (@sspitsyn - Reviewer) Review applies to c46c099f
• Alex Menkov (@alexmenkov - Reviewer)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk.git pull/28666/head:pull/28666
$ git checkout pull/28666

Update a local copy of the PR:
$ git checkout pull/28666
$ git pull https://git.openjdk.org/jdk.git pull/28666/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 28666

View PR using the GUI difftool:
$ git pr show -t 28666

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk/pull/28666.diff

Using Webrev

Link to Webrev Comment

[bridgekeeper]

👋 Welcome back cjplummer! 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]

@plummercj 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:

8373102: com/sun/jdi/MethodInvokeWithTraceOnTest.java can fail with ObjectCollectedException when run with a small heap

Reviewed-by: amenkov, sspitsyn

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 30 new commits pushed to the master branch:

• 15f2538: 8373056: Shenandoah: Remove unnecessary use of ShenandoahAllocRequest.type()
• 13e32bf: 8372098: Move AccessFlags to InstanceKlass
• 6db1c4f: 8371409: Wrong lock ordering between FullGCALot_lock and ThreadsLockThrottle_lock/MethodCompileQueue_lock
• ... and 27 more: https://git.openjdk.org/jdk/compare/4856344668042fcbc4d15966519d27fb0a4f509f...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]

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

• serviceability

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

• 01: Full - Incremental (6e7ba1be)
• 00: Full (c46c099f)

[sspitsyn]

Looks good.

[alexmenkov]

I think better solution would be to catch ObjectCollectedException in EventSetImpl.toString implementations (and print al least event name).
But it might not be worth the effort, so it's up to you

[plummercj]

I think better solution would be to catch ObjectCollectedException in EventSetImpl.toString implementations (and print al least event name).
But it might not be worth the effort, so it's up to you

I think you're suggesting to implement it in all of the Event subclasses implemented as inner classes of EventSetImpl, right?

[plummercj]

I think you're suggesting to implement it in all of the Event subclasses implemented as inner classes of EventSetImpl, right?

ThreadedEventImpl has the following for toString():

        return eventName() + " in thread " + thread.name();

This is the one being triggerd by the test via the thread.name() call. LocatableEventImpl inherits from ThreadedEventImpl and has its own toString() with the same thread.name() reference, so it also needs to be fixed. I think that might be it.

[alexmenkov]

I think better solution would be to catch ObjectCollectedException in EventSetImpl.toString implementations (and print al least event name).
But it might not be worth the effort, so it's up to you

I think your suggesting to implement it in all of the Event subclasses implemented as inner classes of EventSetImpl, right?

It would be more complete solution, but even fixing only EventSetImpl.toString (for each event, printing event name (or class name of the event) + "" on the exception ) should be good enough

[plummercj]

It would be more complete solution, but even fixing only EventSetImpl.toString (for each event, printing event name (or class name of the event) + "" on the exception ) should be good enough

I don't think EventSetImpl.toString() is involved at all, although if called it can suffer the same ObjectCollectedException, which would be fixed if the issue is addressed in the toString() method of the Event sub-classes instead.

[plummercj]

I've updated the implementation so now the event type is printed even if there is an ObjectCollectedException. Turns out I was worng about this being triggered by a VirtualThreadEndEvent. It was actually a ClassPrepareEvent, but it was for the class VirtualThreadEndEvent. Here's all the output for the event set, including the handling of the ObjectCollectedException (2nd to last line).

[JDI: Receiving:                suspendPolicy(byte): 0]
[JDI: Receiving:                events(Events[]): ]
[JDI: Receiving:                    events[i](Events): ]
[JDI: Receiving:                    eventKind(byte): 8]
[JDI: Receiving:                        requestID(int): 2]
[JDI: Receiving:                        thread(ThreadReferenceImpl): ref=151]
[JDI: Receiving:                        refTypeTag(byte): 1]
[JDI: Receiving:                        typeID(long): ref=231]
[JDI: Receiving:                        signature(String): Ljdk/internal/event/VirtualThreadEndEvent;]
[JDI: Receiving:                        status(int): 3]
[JDI: EventSet: SUSPEND_NONE]
[JDI: Looking up Class, signature='Ljdk/internal/event/VirtualThreadEndEvent;', id=231]
[JDI: Caching new ReferenceType, sig=Ljdk/internal/event/VirtualThreadEndEvent;, id=231]
[JDI: Sending Command(id=438) JDWP.ThreadReference.Name]
[JDI: Sending:                 thread(ThreadReferenceImpl): ref=151]
[JDI: Event: ClassPrepareEvent in thread <thread collected>]
[JDI: Handled Prepare Event for jdk.internal.event.VirtualThreadEndEvent]```

[plummercj]

Thank you for the reviews Alex and Serguei!

/integrate

[openjdk]

Going to push as commit ee0b8a7.
Since your change was applied there have been 33 commits pushed to the master branch:

• c09167d: 8373113: Fix whitespace in RunTests.gmk
• 674cc3e: 8042054: JTree.updateUI uses out-of-date item size information
• 7e91d34: 8365732: RISC-V: implement AES CTR intrinsics
• ... and 30 more: https://git.openjdk.org/jdk/compare/4856344668042fcbc4d15966519d27fb0a4f509f...master

Your commit was automatically rebased without conflicts.

[openjdk]

@plummercj Pushed as commit ee0b8a7.

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