Non-Loaded/'Idle' system suffered segfault in java.lang.ProcessImpl.create #9134
Assignees
Labels
Comments
|
OK. It is happening when Windows has been left alone and has blanked/powered down the screen. As soon as I wake the screen, I see the segfault report, as below. In the same method as before, please note. This IS repeatable. Interestingly, the original jar from which the native application was compiled (with "-Dspring.aot.enabled=true") carried on happily and remained able to serve requests once the machine was 'awake' again. Note that the boot jar was executed with "-agentlib:native-image-agent=config-output-dir=.\temp,experimental-class-define-support" I found #1185 but that isn't quite it... |
|
Well! Maybe #1185 was it, after all. I added to my application's RuntimeHintsRegistrar:
No more kablooie on screen blank/wake :-) Hmmmmm... On one hand I am happy to have found this. On the other...this enhances my worry about GraalVM's"closed world" assumption: how many poeple are going to think of adding "test the application's response to screen blank" to their dev procedure? Not many, I'll wager. ..especially given that most/many are probably working on server-side applications... And what else is lurking behind some obscure corner? |
|
Oh well, spoke/closed too soon, so reopening. Latest segfault (pretty much as before, but it might help?) |
|
OK. Another "data point"... I decided to be brave and try with JDK 23: graalvm-jdk-23.0.0-ea.13_windows-x64_bin Process is working beautifully under normal conditions...I can do a quick test with: All goes well. THIS time, the process survived wake up long enough to handle a couple of requests and THEN gave the following: |
|
This may be related: #7599 marked as closed a week ago... |
|
Hello, is it possible to share a minimal reproducer that we can use to get the same error locally? |
|
I've been trying to make a minimal reproducer...no luck :-( sadly, I can't share the actual application. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Describe the issue
Segfault in "AOT compiled code java.lang.ProcessImpl.create(Native Method)"
Steps to reproduce the issue
I had just finished running 7500 (as 5 users * 1500 requests) identical GET requests to one page in my UI.
Just a simple smoke test to see what happened.
Left the system idle while off doing other things...
came back to a segfault.
Suffered by an @scheduled activity.
I THINK I saw memory stayikng a little high after the test (not being garbage collected) but I wasn't deliberately looking for this.
I haven't been able to reproduce this...yet.
But I thought it MIGHT be a useful "data point."
Describe GraalVM and your environment:
Windows 11
graalvm-jdk-22.0.1+8.1
Microsoft (R) C/C++ Optimizing Compiler Version 19.39.33523 for x64
Spring Boot 3.3.0
Tomcat 10.1.24
More details
The code is executed via spring boot's @scheduled feature:
The text was updated successfully, but these errors were encountered: