Very Computer

Does anyone know what would cause a libthread panic?  I have
included the output from my program and the system message
that came with the core dump. The program runs for hours before
the panic arrives. The LWP number looks awful big
in the panic fault.  If you have seen this before or have any
ideas please send to:

Thanks,

Chuck Williams
--------------------------------------------------------------
*** Execution output
            .
            .
            .
10/27/96 15:14:50 ->About to send a stat message for Resource
SRQ_LectroZT36 and Measurement Standby Abort value = 0
10/27/96 15:14:50 ->DoThis returned thread 16
10/27/96 15:14:50 ->Suspending thread 16
10/27/96 15:14:51 ->ResumeThread(13)
10/27/96 15:14:51 ->ResumeThread(16)
signal fault in critical section
signal number: 11, signal code: 1,                  fault
address: 0xecc88000, pc: 0xee72231c, sp: 0xecc86b58
libthread panic: fault in libthread critical section (PID: 418
LWP 1481650998)
stacktrace:
        ee7dac14
        ee7e1f3c
        ede17c2c
        ede144c8
        9

***When I try to load core file with dbx
                    .
                    .
                    .
Reading symbolic information for io_snmp.dll
Reading symbolic information for nss_files.so.1
Reading symbolic information for cmm3.dll
detected a multithreaded program
dbx: warning: couldn't get LWP status information -- Bad file
number



libthread_db.so error

: Does anyone know what would cause a libthread panic?  I have

I have encountered many such panics. In my personal experience,
it was due to signal reception in a critical state ie: mutex_lock(),
or failure of the library to deliver a signal via lwp_kill to a
specific thread.

While this probably should not happen in the library, it may mean
a bug in your code. Check you mutex usage & signal handling.

/*  

 |
 | OBTEC Research, Inc.
*/