JDK-8281647 : G1: Overflow during reference processing
  • Type: Bug
  • Component: hotspot
  • Sub-Component: gc
  • Affected Version: 19
  • Priority: P4
  • Status: Closed
  • Resolution: Duplicate
  • Submitted: 2022-02-11
  • Updated: 2022-02-14
  • Resolved: 2022-02-14
Related Reports
Duplicate :  
Relates :  
Relates :  
Relates :  
Description
When running github automatic test, Linux x86 (langtools/tier1) fails with:

#
# A fatal error has been detected by the Java Runtime Environment:
#
#  Internal Error (g1ConcurrentMark.cpp:1649), pid=29286, tid=29295
#  fatal error: Overflow during reference processing, can not continue. Current mark stack depth: 65472, MarkStackSize: 65536, MarkStackSizeMax: 4194304. Please increase MarkStackSize and/or MarkStackSizeMax and restart.
#
# JRE version: OpenJDK Runtime Environment (19.0) (build 19-internal+0-lkorinth-7d3e7a1b8ad0acd10bbae51aa037c95f8365c845)
# Java VM: OpenJDK Server VM (19-internal+0-lkorinth-7d3e7a1b8ad0acd10bbae51aa037c95f8365c845, mixed mode, tiered, g1 gc, linux-x86)
# Problematic frame:
# V  [libjvm.so+0x57265d]  G1ConcurrentMark::weak_refs_work()+0x31d
#
# CreateCoredumpOnCrash turned off, no core file dumped
#
# An error report file with more information is saved as:
# /home/runner/work/jdk/jdk/build/run-test-prebuilt/test-support/jtreg_test_langtools_tier1/scratch/hs_err_pid29286.log
#
# If you would like to submit a bug report, please visit:
#   https://bugreport.java.com/bugreport/crash.jsp
#

See: https://github.com/lkorinth/jdk/runs/5154094102?check_suite_focus=true

The code is based on code from February 10th, so JDK-8280029 is applied.
Comments
Changed JDK-8280087 to a bug and resolved this as a duplicate.
14-02-2022

OK, so this appears to be highly intermittent? Unfortunately, I temporarily lost the access to machine where JDK-8280029 used to reproduce, I can do longer runs once that machine available. At the same time, I believe the real fix is JDK-8280087. Bumping the stack size even further on x86_32 looks counter-productive to cover the highly-intermittent test failure.
14-02-2022

[~shade] FYI.
11-02-2022