Kernel Panic Linux 2.4.3 RH7

Kernel Panic Linux 2.4.3 RH7

[lkern]

Machine has been locking up between 0-3 times a day
sporadically.  Nothing predictable about it. Hadn't locked up for 3
days, and locked 3x today, the last 2 times within 20 minutes of each 
other. Had run stable with 2.2.18, and was running fairly stable on 2.4.3
up until about last week.  (might be coincidence, or not, but seems to
happen when I am on IRC -- DOS?, nothing in log files or firewall logs
however)

Machine info:

RedHat 7, Intel Celeron 450, 256Meg ECC PC100 DRAM, 1 15G IDE (/usr, /tmp,
/home,
/var, swap ), 1 2G SCSI (/, /boot, swap) HDD, 2 identical tulip chipset
eth cards.

I just swapped the RAM out this last crash, seeing if that makes any
differences.

Kernel is configured as advanced router with packet filtering, ip
firewalling, NAT, IPTABLES and others enabled running a fairly
comprehensive iptables ruleset, nothing superhuge though.


Machine was recently upgraded to 2.4.3 kernel, from 2.2.18.
Dump message is partially as follows:

Mismatch in TCPACCEPT IN=ETH0 OUT= MAC= 00:00:e8:24:53

(some other stuff below this, but the machine then rolled to the below and
I couldn't finish copying down the above, remainder is as follows:

Kernel Panic:  Attempted to Kill init!
Unable to handle kernel paging request at virtual address e030a40c

printing eip:
c01194ae
*pid=00000000
Oops: 0000
CPU: 0
Eip: 0010 [<c01194ae7>]
EFLAGS: 0010087

(stack information skipped, if you need it let me know and I will write it
all down the next time)

Process: swapper (pid0, stack page c02dl000)

stack 00000000

Code 8b1186108b 590c85d274 088b410489 42048910

Kernel Panic: Aiee, killing interrupt handler
In interrupt handler - not syncing

Any ideas?

-buddy ellis

Re: Kernel Panic Linux 2.4.3 RH7

[Tim Haynes]

<lkern@mail.co.gilchrist.fl.us> writes:

> Machine has been locking up between 0-3 times a day sporadically. Nothing
> predictable about it. Hadn't locked up for 3 days, and locked 3x today,
> the last 2 times within 20 minutes of each other. Had run stable with
> 2.2.18, and was running fairly stable on 2.4.3 up until about last week.
> (might be coincidence, or not, but seems to happen when I am on IRC --
> DOS?, nothing in log files or firewall logs however)
> 
> Machine info:
> 
> RedHat 7, Intel Celeron 450, 256Meg ECC PC100 DRAM, 1 15G IDE (/usr,
> /tmp, /home, /var, swap ), 1 2G SCSI (/, /boot, swap) HDD, 2 identical
> tulip chipset eth cards.
[snip]
> 
> Mismatch in TCPACCEPT IN=ETH0 OUT= MAC= 00:00:e8:24:53
[snip]
> (stack information skipped, if you need it let me know and I will write it
> all down the next time)
> 
> Process: swapper (pid0, stack page c02dl000)
[snip]

Take your pick from one or more of the following:
    a) dodgy RAM - less likely if a different set of RAM has also had
       the problem;
    b) a faulty network card - a tulip (Netgear FA310TX) reliably caused me
       a hang ~6 months ago, so there is a precedent;
    c) overheating - see if, left to its own devices, the hangs become
       more frequent the longer the box is left on; also implement
       lm_sensors and keep an eye on the CPU temperature if possible;
    d) RH7 alert: did you use kgcc to build the kernel?
    e) Something totally else :8)

HTH,

~Tim
-- 
   10:32pm  up 3 days, 46 min, 10 users,  load average: 0.18, 0.10, 0.08
piglet@stirfried.vegetable.org.uk |Clouds cross the black moonlight,
http://piglet.is.dreaming.org     |Rushing on down to the sound 
                                  |of a turning world