On Sat, 3 Nov 2001 21:19, Hendrik Schreiber wrote: > Pete, > > > Well it stalls the whole JVM with that message and I am forced to > > kill -9 it. > > I will try it again tomorrow and see if I can track down exactly where it > > occurs. > > If it does it again, can you try a sig quit instead of the -9 (in case you > haven't tried that) and mail the thread dump to me.
The problem is that I can only produce the error when the JVM is force shutdown (ie by ctrl-c ing the application) and thus it is already in a signal handler and wont get the sig quit. If I shutdown JVM gracefully (via the test web interface thingie) then it goes down fine and I dont get that wierd message. So I did a siguit just prior to shutdown. It seems that "get_self" is the name of the signal handler for SIGUSR2. It is located in libhpi.so (an IBM JVM library). Heres the relevent part of log SIGUSR2 : get_self (libhpi.so) Not sure if this makes it easier to understand what "get_self can't find pthread_self" means. I may have to go through and put lots of System.out.printlns() in jo to see where exactly the issue is. -- Cheers, Pete -------------------------------------------------- you've made a dangerous leap right over common sense, like some kind of metaphysical Evil Knievel -------------------------------------------------- -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
