-
Notifications
You must be signed in to change notification settings - Fork 737
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
J9vmTest_FAILED j9mm.316 ASSERTION FAILED at openj9/runtime/gc_vlhgc/GlobalMarkingScheme.cpp:957: ((uintptr_t)0x99669966 == (clazz)->eyecatcher) & Segmentation error vmState=0x00020002 #20150
Comments
It might be regression due #20132 |
Seems it is not a regression from #20132, at least it's GC part. This crash occur due heap corruption. An attempt to scan class object
Balanced has individual Memory Pool for each region, so there is no way to allocate object out-of-line across region border (and there were no recent changes in this area). Also (by the same reason) TLH memory can not cross region's border. |
Taking a look |
Looked into the core-dump from https://hyc-runtimes-jenkins.swg-devops.com/job/Test_openjdk23_j9_sanity.functional_s390x_linux_testList_1/21/ to see if I can figure out which JIT code created the the object in question. Given the internal grinder Jason has posted, has 100% failure rate, I replicated the test on my Linux on Z fyre image. I can reproduce the failure without JIT as well with same failing stack.
Loaded core-dump into jdmpview and ran GC Check, although bit different stack, I am seeing following corruption,
Given that this is reproducible with |
ok, thank you very much, it is puzzling now. I will take a look again. May I get system core from -Xint failure somehow? Ideally if you can upload it to |
I have uploaded the core-dump with the setup to reproduce this at |
I don't think the problem occurs in 0.48, moving the target milestone. |
I have a confirmation this is regression caused by #20132
back to:
it seems work again. So, new code is missing array classes somehow and it is the reason for crashes and assertions. I don't see the problem in a new code at the moment. I am looking to it now. I hope if I find the problem we don't need to revert entire PR again. |
- Do proper initialization of state machine before use it for walk. - Move check that list is switched to super Classloader to proper place to cancel walk on individual list, not for all lists. relates to eclipse-openj9#20150 Signed-off-by: Dmitri Pivkine <[email protected]>
Should be fixed by #20199 |
- Do proper initialization of state machine before use it for walk. - Move check that list is switched to super Classloader to proper place to cancel walk on individual list, not for all lists. relates to eclipse-openj9#20150 Signed-off-by: Dmitri Pivkine <[email protected]>
Failure link
From an internal build(
ubu22s390x-rt-1
):Rerun in Grinder - Change TARGET to run only the failed test targets.
Optional info
Failure output (captured from console output)
50x internal grinder - all failed
Also seen on JDK17 s390x_linux
JDK11 s390x_linux
The text was updated successfully, but these errors were encountered: