Closed knn-k closed 2 years ago
Another issue with SCM23 on AArch64 macOS: #14621
Some timeouts in 20x Grinder job: https://openj9-jenkins.osuosl.org/job/Grinder_testList_0/126/
The timeouts in the Grinder job may be related to #15352.
Changes in the build where this was seen.
https://github.com/eclipse-openj9/openj9/compare/3d06b2f9c2c...7c5c4148460
"Unable to locate JIT stack map" was observed in a 20x Grinder run: https://openj9-jenkins.osuosl.org/job/Grinder/1025/ . The job uses the build from June 7 nightly build.
OpenJ9 - 0414be65436
OMR - c859f0d17d0
https://openj9-jenkins.osuosl.org/job/Test_openjdk18_j9_extended.system_aarch64_mac_Nightly_testList_0/112 SharedClasses.SCM23.MultiCL_0
MCL4 11:50:52 >> Total classes loaded = 20001
MCL4 stderr JVMCDRT000E Unable to locate JIT stack map - aborting VM
MCL4 stderr JVMCDRT001E Method: java/lang/ClassLoader.loadClass(Ljava/lang/String;)Ljava/lang/Class; (000000012811F788)
MCL4 stderr JVMCDRT002E Failing PC: 00000001389E81D5 (offset 0000000000000001), metaData = 00000001180DF6E8
STF 11:50:53.044 - **FAILED** Process MCL4 ended with exit code (255) and not the expected exit code/s (0)
"Unable to locate JIT stack map" seems to be thrown as an exception by jitWalkFrame()
in com/ibm/j9ddr/vm29/j9/stackwalker/JITStackWalker.java.
The method shown in the message is the same: java/lang/ClassLoader.loadClass(Ljava/lang/String;)Ljava/lang/Class;
https://openj9-jenkins.osuosl.org/job/Test_openjdk19_j9_sanity.openjdk_aarch64_mac_Nightly/8 java/lang/String/CompactString/CompareToIgnoreCase.java
21:34:54 STDERR:
21:34:54 STATUS:Passed.
21:34:54 JVMCDRT000E Unable to locate JIT stack map - aborting VM
21:34:54 JVMCDRT001E Method: java/lang/ClassLoader.loadClass(Ljava/lang/String;)Ljava/lang/Class; (0000000138DB9E40)
21:34:54 JVMCDRT002E Failing PC: 000000016F0463F5 (offset 0000000000000001), metaData = 00000002A8029F68
21:34:54 01:18:48.778 0x15083b900j9codertvm(j9ji.110 * ** ASSERTION FAILED ** at /Users/jenkins/workspace/Build_JDK19_aarch64_mac_Nightly@2/openj9/runtime/codert_vm/jswalk.c:538: ((0 ))
The symptom is different, but I guess it is related to the crash in Issue #15518 and some other macOS issues.
Unfortunately, javacore.20220915.111848.38544.0002.txt for the CompareToIgnoreCase test does not contain meaningful information.
There is a core file for this one, from the assert. It's not helpful?
PR #15907 should fix this issue.
https://openj9-jenkins.osuosl.org/job/Test_openjdk17_j9_extended.system_aarch64_mac_Nightly_testList_0/77/
OpenJ9 - 7c5c4148460 OMR - cf8ddbd1adc JCL - d7f7dff2a37 based on jdk-17.0.4+6)