Hi Alphafolks,
I hope someone can tell me what the following messages might mean in the
file /usr/adm/syslog.dated/DD-MMM-HH:mm/kern.log:
Feb 9 09:57:35 mach2 vmunix:
Feb 9 09:57:36 mach2 vmunix: trap: invalid memory ifetch access from kernel mode
Feb 9 09:57:36 mach2 vmunix:
Feb 9 09:57:36 mach2 vmunix: faulting virtual address: 0xffffffffa055fcb0
Feb 9 09:57:36 mach2 vmunix: pc of faulting instruction: 0xffffffffa055fcb0
Feb 9 09:57:36 mach2 vmunix: ra contents at time of fault: 0xfffffc00003c007c
Feb 9 09:57:36 mach2 vmunix: sp contents at time of fault: 0xffffffffa112db40
Feb 9 09:57:36 mach2 vmunix:
Feb 9 09:57:36 mach2 vmunix: panic: kernel memory fault
Feb 9 09:57:36 mach2 vmunix: syncing disks... done
Feb 9 09:57:36 mach2 vmunix: dumpsys: entered.
Feb 9 09:57:36 mach2 vmunix: secondary cache was auto sized to 64 pages
.... (boot procedure)
I have a 3000/500 workstation, still running OSF/1 v1.3 for various
reasons, which behaved strangely last week. It was working fine until
at one point, when a user was logging out of the DXsession from the
console, and the machine crashed. Unfortunately, other errors prevented
it from leaving a crash dump, such as the disks not being found. My
first few attempts at rebooting it resulted in a message on the console:
? 40 NOSUCHDEV
84 FAIL
when it was trying to reboot after hardware reset.
At the console, I was able to do SHOW DEV and SHOW CONFIG, and
everything looked reasonable; all devices showed up with proper IDs, and
all my memories were available. But the system will not boot.
I ran the console tests with the simple T command, and it gave a couple
of messages:
at the NVR test, it gave a message like "ALWAYS SET CLOCK ON RESTART
ERROR" but the test finished with an "NVR TEST OK" message.
then at the NI test, it stopped with a complaint about loop-back
testing, and what looks like a stack dump. It would hang at a
particular point, and I would have to use the halt button to get the
prompt back to do anything else.
After much fiddling and power on/off cycling, but no obvious service,
the machine finally rebooted, and all seemed fine. I brought it back
down to run those tests again, and they still gave the same results.
Yet when I rebooted again, the network seemed to work fine.
Just as a precaution, I had the local field service replace the I/O
module, and all the console tests ran with no errors, and again, the
system seemed fine. But I'm still getting the above error messages in
the kern.log file each time the system boots up.
Does anyone know what's going on, or what I can do to find out?
thanks in advance.
eyc
Received on Tue Feb 13 1996 - 20:35:12 NZDT