figuring out what happenend within the java.lang.Thread.<init>

[David Holmes - Sun Microsystems]

Peter,

Try -XX:+TraceExceptions on a non-product build.

David Holmes

Peter Helfer said the following on 24/11/07 01:43 AM:
> Ok, I'm running again into troubles..
> 
> This is what I'm seeing:
> 
> VM option '+UnlockDiagnosticVMOptions'
> VM option '-VerifyBeforeExit'
> PARENT[80D8800]: getThread: Linked threads 80D8800 <--> 80DA800
> SpecPool::createThread: EXCEPTION OCCURED IN PART1
>         at java.lang.Thread.<init>(Thread.java:463)
> PARENT[80D8800]: getThread: Linked threads 80D8800 <--> 80DC400
> SpecPool::createThread: EXCEPTION OCCURED IN PART3      at 
> java.lang.ThreadGroup.add(ThreadGroup.java:868)
> PARENT[80D8800]: getThread: Linked threads 80D8800 <--> 80DFC00
> SpecPool::createThread: EXCEPTION OCCURED IN PART1
>         at java.lang.Thread.<init>(Thread.java :463)
> PARENT[80D8800]: getThread: Linked threads 80D8800 <--> 80E1C00
> SpecPool::createThread: EXCEPTION OCCURED IN PART1
>         at java.lang.Thread.<init>(Thread.java:463)
> PARENT[806DC00]: getThread: Linked threads 806DC00 <--> 80E5400
> PARENT[806DC00]: getThread: Linked threads 806DC00 <--> 80E3800
> Error occurred during initialization of VM
> java.lang.ExceptionInInitializerError
>         at sun.misc.Launcher.<init>(Launcher.java:71)
>         at sun.misc.Launcher.<clinit>(Launcher.java:59)
>         at 
> java.lang.ClassLoader.initSystemClassLoader(ClassLoader.java:1322)
>         at 
> java.lang.ClassLoader.getSystemClassLoader(ClassLoader.java:1304)
> Caused by: java.lang.NullPointerException
>         at java.net.URLClassLoader.<clinit>(URLClassLoader.java:589)
>         at sun.misc.Launcher.<init>(Launcher.java:71)
>         at sun.misc.Launcher.<clinit>( Launcher.java:59)
>         at 
> java.lang.ClassLoader.initSystemClassLoader(ClassLoader.java:1322)
>         at java.lang.ClassLoader.getSystemClassLoader(ClassLoader.java:1304)
> 
> Now how can I track or figure out, what has happend within 
> java.lang.Thread.<init> ? - this is just the normal stack trace output, 
> are there any other means ? When using GDB (thx Tom for the script!), I 
> can only see backtraces, which end up in os::raise(), but I'm having 
> issues to set breakpoints, even when using "sharedlibrary libjvm"...
> 
> Regards, Peter