-
Notifications
You must be signed in to change notification settings - Fork 728
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
jdk19 OpenJDK java/lang/Thread/virtual/stress/TimedGet.java crash vmState=0x00000000 #16729
Comments
This test has just been re-enabled by adoptium/aqa-tests#4337. |
Grinder with |
@babsingh how is it a perf issue is causing a crash? |
The stress test exhausts sub-4G memory by allocating Java stack for 250K virtual threads. The |
Dont we throw OOM if we cant allocate the stack? |
Yes, OOM is thrown multiple times:
|
So what is the cause of the |
We reproduced the failure on @ehrenjulzert's machine last Friday but lldb didn't show line numbers for the source code. @ehrenjulzert can you try again to find the cause of the Unhandled exception? |
Sure, I can try taking a look again today |
I'm still not sure what's going on, but here's what we found so far: originally we were only getting this traceback in lldb that didn't tell us much:
we tried changing the code in
which gave us this much nicer traceback but it's still crashing in
We also tried removing the
When I ran this I just got the same crash as in the first change we made, however it's weirdly inconsistent. Sometimes it will crash in
Also, for some reason adding |
The return value of The correct syntax should be (apology for the confusion) VMINLINE VM_BytecodeAction
enterContinuationImpl(REGISTER_ARGS_LIST)
{
VM_BytecodeAction rc = EXECUTE_BYTECODE;
j9object_t continuationObject = *(j9object_t*)_sp;
buildInternalNativeStackFrame(REGISTER_ARGS);
updateVMStruct(REGISTER_ARGS);
/* Notify GC of Continuation stack swap */
_vm->memoryManagerFunctions->preMountContinuation(_currentThread, continuationObject);
if (enterContinuation(_currentThread, continuationObject)) {
_sendMethod = J9VMJDKINTERNALVMCONTINUATION_ENTER_METHOD(_currentThread->javaVM);
rc = GOTO_RUN_METHOD;
}
VMStructHasBeenUpdated(REGISTER_ARGS);
if (immediateAsyncPending()) {
rc = GOTO_ASYNC_CHECK;
} else if (VM_VMHelpers::exceptionPending(_currentThread)) {
rc = GOTO_THROW_CURRENT_EXCEPTION;
}
return rc;
} |
Ok with that change I'm still getting |
Excellent, that's the expected behaviour. @ehrenjulzert Can you create a PR for this fix? |
Fixes eclipse-openj9#16729 Signed-off-by: Ehren Julien-Neitzert <[email protected]>
PR: #16953 |
@ehrenjulzert Two more pending tasks:
|
Changed due to eclipse-openj9/openj9/issues/16729 being resolved Signed-off-by: Ehren Julien-Neitzert <[email protected]>
aqa-tests PR: adoptium/aqa-tests#4452 |
Fixes eclipse-openj9#16729 Signed-off-by: Ehren Julien-Neitzert <[email protected]>
0.37 PR: #16961 |
Closing. Pendings task in #16729 (comment) have been completed. |
Changed due to eclipse-openj9/openj9/issues/16729 being resolved Signed-off-by: Ehren Julien-Neitzert <[email protected]> Co-authored-by: Ehren Julien-Neitzert <[email protected]>
I'll reopen this until we unexclude the test. |
I was looking at an old exclude, the test is excluded against #15184 |
https://openj9-jenkins.osuosl.org/job/Test_openjdk19_j9_sanity.openjdk_x86-64_mac_Nightly/110
jdk_lang_0
java/lang/Thread/virtual/stress/TimedGet.java
@babsingh is this caused by #16678
@tajila @fengxue-IS fyi
The text was updated successfully, but these errors were encountered: