Clean monitor enter records during VM shutdown

[thallium]

No description provided.

[thallium]

@tajila can you launch tests again?

[tajila]

jenkins test sanity alinux64 jdk21,jdk25

[thallium]

@tajila I moved the clean-up code to protectedDestroyJavaVM(), can you launch the test again if the change looks good to you?

[tajila]

jenkins test sanity alinux64 jdk21,jdk25

[thallium]

My personal build https://hyc-runtimes-jenkins.swg-devops.com/view/OpenJ9%20-%20Personal/job/Pipeline-Build-Test-Personal/33821/ passed sanity.functional and sanity.openjdk, so I'm expecting the running https://openj9-jenkins.osuosl.org/job/Pipeline_Build_Test_JDK25_aarch64_linux/430/ to pass as well.

Just fixed compile issue on JDK21.

[tajila]

jenkins test sanity alinux64 jdk21,jdk25

[thallium]

testJITServer failure in https://openj9-jenkins.osuosl.org/job/Pipeline_Build_Test_JDK25_aarch64_linux/430/ seems to be known issue.

[thallium]

Also failed on cmdLineTester_jfr_0 Testing: runWorkload 2 - approx 2mins because of timeout:

***[TEST INFO 2026/06/09 17:08:22] ProcessKiller detected a timeout after 3000000 milliseconds!***

Might be the same as #23120.

[tajila]

jenkins test sanity alinux64 jdk21,jdk25

[tajila]

jenkins test sanity alinux64 jdk21,jdk25

[tajila]

failures in cmdLineTester_jfr_0

[thallium]

I think it was just a machine being slow or something? I'm rerunning with grinder https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/60906/ and it's passing so far but spent ~10mins on it. I think the extra virtual thread load makes it a lot heavier. When I was running the test with nightly build on my local VM, I even got native OOM for some reason.

[tajila]

jenkins test sanity alinux64 jdk21,jdk25