PC-Offset Stack Trace error

[Charlie Bulosan_1]

Hi Guys,

A workstation rebooted due to panic,I have this error in OLDsyslog.log below. shutdownlog says a reboot, dmesg has no error messages, tombstones says no chassis codes. Pls help to understand the problem.

=============================
Oct 10 08:15:21 JWONGCC0 vmunix: NFS server jwgrns88 not responding still trying
Oct 10 08:15:21 JWONGCC0 vmunix: NFS server jwgrns88 ok
Oct 10 08:15:21 JWONGCC0 vmunix: NFS server jwongs88 not responding still trying
Oct 10 08:15:21 JWONGCC0 vmunix: NFS server jwongs88 ok
Oct 10 08:15:21 JWONGCC0 vmunix: linkstamp: Sun Sep 14 19:02:50 BST 2003
Oct 10 08:15:21 JWONGCC0 vmunix: NFS server jwongs88 not responding still trying
Oct 10 08:15:21 JWONGCC0 above message repeats 2 times
Oct 10 08:15:21 JWONGCC0 vmunix: _release_version: @(#)B2352B/9245XB HP-UX (B.11.00) #1: Wed Nov 5 22:38
:19 PST 1997
Oct 10 08:15:21 JWONGCC0 vmunix: NFS server jwongs88 ok
Oct 10 08:15:21 JWONGCC0 above message repeats 2 times
Oct 10 08:15:21 JWONGCC0 vmunix: __kern_ci_revision: $Header: kern_sequence.c,v 1.6.106.512 97/11/05 18:01:
46 msosa Exp $
Oct 10 08:15:21 JWONGCC0 vmunix: panic: vn_rele
Oct 10 08:15:21 JWONGCC0 vmunix: PC-Offset Stack Trace (read across, top of stack is 1st):
Oct 10 08:15:21 JWONGCC0 vmunix: 0x002fc8bc 0x000ee104 0x000ee64c
Oct 10 08:15:21 JWONGCC0 vmunix: 0x000ee7a0 0x000ec570 0x000cebe0
Oct 10 08:15:21 JWONGCC0 vmunix: 0x00134804 0x00035944
Oct 10 08:15:21 JWONGCC0 vmunix:
Oct 10 08:15:21 JWONGCC0 above message repeats 3 times
Oct 10 08:15:21 JWONGCC0 vmunix: End Of Stack

Thanks in advance.

Kind Regards,
Charlie

[melvyn burnard · Accepted solution]

So what exactly is the string in /etc/shutdownlog?
If it is a panic, then htere should be a crash dump available in /var/adm/crash.
Check htere, and look at the INDEX file to confirm the dump matches up wiht your panic date, and what the panic string is in there as well.
The probablity is you will need to install the latest patch bundle.

[Dietmar Konermann]

Well, this is an explicit panic from vnrele()... maybe you need to open a call with your response center to get this anaylzed.

We can try to get an idea if especially this instance of the problem is known.

Post the result of the following command:

echo "
0x002fc8oc?i
0x000ee104?i
0x000ee64c?i
0x000ee7a0?i
0x000ec570?i
0x000cebe0?i
0x00134804?i
0x00035944?i
" | /usr/bin/adb /stand/vmunix


Best regards...
Dietmar.

[Charlie Bulosan_1]

Hi Melvyn,

It didn't show any message at all this 10th of October on the shutdownlog.
# more /var/adm/shutdownlog
18:54 Sun Sep 14, 2003. Reboot:
19:07 Sun Sep 14, 2003. Reboot:
19:16 Sun Sep 14, 2003. Reboot:
22:36 Thu Oct 2, 2003. Reboot:

But according to the user, he is logging in a citrix(windows terminal) within hp-ux then it suddenly rebooted. This problem occurred only once.

The workstations here are configured with no /var/adm/crash. This is their worldwide config. And they have the same built for all workstations. Other workstations don't encounter this.

I'm just wondering what's the meaning of this PC-offset stack trace error. Is there a hardware problem?


Thanks,
Charlie

[Kent Ostby]

Charlie --

panic: vn_rele is a system panic out of the routine vn_rele. vn_rele is short for "vnode release".

Some of the time that this has been seen on 11.00, it has been fixed by updating the NFS, ARPA, LAN, and STREAMS patches to a more current level.

Please post your patch list here as well. (swlist -l product output).

If you dont want to go through a lot of work on this call then install the NFS, ARPA, LAN, and STREAMS patches plus their dependencies (although you should NOT reinstall PHKL_18543).

Since this is not always a clear situation, you may want to involve the HP Response Center to troubleshoot it.

Best regards,

Kent M. Ostby

[Charlie Bulosan_1]

Hi Dietmar,

Here's the output.

# echo "
> 0x002fc8oc?i
> 0x000ee104?i
> 0x000ee64c?i
> 0x000ee7a0?i
> 0x000ec570?i
> 0x000cebe0?i
> 0x00134804?i
> 0x00035944?i
> " | /usr/bin/adb /stand/vmunix

bad modifier
bad command
locallookuppn+0xB2C: B locallookuppn+0x031c
lookuppn+104: COMDB,*<> r0,r3,lookuppn+0x0184
lookupname+40: COPY r28,r3
stat1+38: COMB,<> r28,r0,stat1+0x00d0
stat+20: ADDIL L%0xfff8d800,r27
syscall+28C: LDI 1,r10
$syscallrtn:
$syscallrtn: LDO -1232(r30),r19
$syscallrtn+4: LDIL L%0x40000000,r21

Thanks,
Charlie

[Dietmar Konermann]

Charlie,

sorry had a typo in one of my hex numbers. Nevertheless, the stack trace seems not to point directly to a specific subsystem. The panic happened during a stat() system call, obviously. Deeper dump analysis could easily find the corresponding file location (file system, NFS, etc.).
After scanning through known problems, I agree with Kent. Networking stuff, especially NFS, ist the most likely area where I would start patching.

Best regards...
Dietmar.

[melvyn burnard]

Without a panic dump, in-depth analysis is not possible. A panic in vn_rele() could eb a number of issues.
As has already been stated, you should now look at installing a patch bundle, and the minimum you should look at is to install the latest Networing/Streams/ARPA/NFS patches and all their dependencies.

[Charlie Bulosan_1]

Hi Guys,

Thank you so much for all your time and efforts.

I need to convince them that to solve the problem is to patch it. I have attached the output of "swlist -l product".


Kind Regards,
Charlie