core file from 'sh' - received SIGSEGV

[j773303]

Does anyone has the experience about followng message? Due to I get a core file, but I don't know how to explain this message. Thanks.


/tmp#what core
core:
Version M-11/16/88f
$ B.11.23 Aug 7 2004 17:33:30 $
$ B11.23.0409LR Aug 16 2004 18:22:14 $

/tmp#file core
core: core file from 'sh' - received SIGSEGV

P.S. this thread has been moved from HP-UX > System Administration to HP-UX > languages - HP Forums Moderator

[Carsten Krege · Accepted solution]

Because the "sh" binary has its symbols stripped off, you would not get a stack trace if you run a debugger on it (like adb or gdb). THe best is to install the latest patch for the shell

PHCO_32444 11.23 sh-posix(1) cumulative patch

and see if the problem goes away. If not you should open a call with HP Support. They will be able to help.

Carsten

[Frank de Vries]

I have some experience, but this is a tough one. I am realy puzzled about the circumstances in your case , just
doing a file core.

Many signals are catchable, and this includes the famous SEGV signals.
In your caase your program has exited with a message such as 'Segmentation Violation - Core Dumped',

But that you get it from doing a file type
command on the core file is rather odd.

It literall means a SEGV signal was sent to your program due to accessing an illegal memory address and the signal was caught
by the shell sh.

Some part of your kernel memory appears
have gotten corrupted. Either by some
faulty program or the some hardware instability.

In an ideal world it would be nice to cleanup the corrupted part, but we need to
know which one it is.

Can you do a strings core | more ?
Do you have defunct processes or zombie
processes ?
What does the output of dmesg say ?
And give us an output of ipcs -a

DO you have any other symptons ?
What about /var/adm/syslog/syslog.log ?

If you can guess from your process list
ps -ef | more which are the hostile processes
try to kill them -15 and otherwise -9.

A reboot may solve it temporarily, until
the same corruption re-occurs.
Still it maybe interesting to know if you
get this error in single user mode too ,
so we can eliminate a lot of factors.

I just hope it is not your kernel.
Anyway let's wait and see.

Good luck troubleshooting:)

[Frank de Vries]

I saw Carstens reply,
and yes,
naturally bugfix patches are always a big possibility in such cases !!
(Sorry I forget to mention that earlier)