search

eclipse-openj9 / openj9

Eclipse OpenJ9: A Java Virtual Machine for OpenJDK that's optimized for small footprint, fast start-up, and high throughput. Builds on Eclipse OMR (https://github.com/eclipse/omr) and combines with the Extensions for OpenJDK for OpenJ9 repo.
Other
3.28k stars 721 forks source link

JDK21 serviceability_jvmti_j9_1_FAILED serviceability/jvmti/stress/StackTrace/NotSuspended/GetStackTraceNotSuspendedStressTest.java Fatal error: Incorrect frame count #18730

Open JasonFengJ9 opened 9 months ago

JasonFengJ9 commented 9 months ago

Failure link

From an internal build(ubu20le-svl-rt7-1):

java version "21.0.1" 2023-10-17 LTS
IBM Semeru Runtime Certified Edition 21.0.1.0 (build 21.0.1+12-LTS)
Eclipse OpenJ9 VM 21.0.1.0 (build openj9-0.42.0, JRE 21 Linux ppc64le-64-Bit Compressed References 20231017_77 (JIT enabled, AOT enabled)
OpenJ9   - 874af12ee
OMR      - 11700e64f
JCL      - eb47ed546b0 based on jdk-21.0.1+12)

Rerun in Grinder - Change TARGET to run only the failed test targets.

Optional info

Failure output (captured from console output)

[2024-01-09T22:51:14.769Z] variation: Mode650
[2024-01-09T22:51:14.769Z] JVM_OPTIONS:  -XX:-UseCompressedOops 

[2024-01-09T22:56:27.766Z] TEST: serviceability/jvmti/stress/StackTrace/NotSuspended/GetStackTraceNotSuspendedStressTest.java

[2024-01-09T22:56:27.766Z] STDERR:
[2024-01-09T22:56:27.766Z] 
[2024-01-09T22:56:27.766Z] Fatal error: Incorrect frame count

[2024-01-09T22:56:27.767Z] TEST RESULT: Failed. Unexpected exit from test [exit code: 87]
[2024-01-09T22:56:27.767Z] --------------------------------------------------
[2024-01-09T22:59:24.455Z] Test results: passed: 155; failed: 1

[2024-01-09T22:59:41.713Z] serviceability_jvmti_j9_1_FAILED

50x internal grinder - 1 failure reproduced, 4 timeout failures, and 3 failures as https://github.com/eclipse-openj9/openj9/issues/18738

pshipton commented 9 months ago

Maybe related to https://github.com/eclipse-openj9/openj9/issues/17871

pshipton commented 9 months ago

8 failures, but only 1 of this test in https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder_iteration_0/3530 on ubu20le-rtp-rt3-1 with Incorrect frame count

pshipton commented 9 months ago

Created https://github.com/eclipse-openj9/openj9/issues/18738 for the serviceability/jvmti/stress/StackTrace/Suspended/GetStackTraceSuspendedStressTest .java failure.

JasonFengJ9 commented 6 months ago

JDK21 aarch64_linux(rhel9-aarch64-3)

[2024-04-04T02:18:55.852Z] variation: Mode150
[2024-04-04T02:18:55.852Z] JVM_OPTIONS:  -XX:+UseCompressedOops -Xverbosegclog 

[2024-04-04T02:20:55.278Z] TEST: serviceability/jvmti/stress/StackTrace/NotSuspended/GetStackTraceNotSuspendedStressTest.java

[2024-04-04T02:20:55.278Z] STDERR:
[2024-04-04T02:20:55.278Z] 
[2024-04-04T02:20:55.278Z] Fatal error: Incorrect frame count

[2024-04-04T02:23:24.832Z] serviceability_jvmti_j9_0_FAILED
babsingh commented 4 months ago

This failure is also seen with the RI: https://bugs.openjdk.org/browse/JDK-8315980. The failure is highly intermittent and it shouldn't block the 0.46 release. The source of the issue is the RI's Virtual Thread code that OpenJ9 reuses. RI has targeted JDK24 to fix this issue. Moving forward to the 0.48 release and keeping the issue open until it is fixed in the upstream by the RI.

babsingh commented 1 month ago

re https://github.com/eclipse-openj9/openj9/issues/18730#issuecomment-2142864348: The failure is still highly intermittent and it also happens with the RI. This issue is not needed for the 0.48 release. The fix for the test will be delivered by the RI. The RI has targeted to fix the test in Java 24. Moving it to the Java 24 milestone to the match the RI.