I'd simply like to report that I've been having similar problems on a
box. Suspecting kernel problems, I upgraded from 2.2.12 to 2.2.14,
building my own kernel. Suspecting hardware problems, I swapped first
memory, then the entire system (hard drive excepted) for a new one. The
faults have persisted as recently as last week, and were first noticed
in January.
The only common elements between the original and current system are
the hard drive (now one of two, with a SCSI disk now on the system),
and an APC Back UPS 500 -- which I've heard has been attributed to
system instability due to poor power quality. Disconnecting the power
supply was briefly attempted, then abandoned, due to a tendency for power
interrupts to occur as furniture and my feet are moved at critical points
along the circuit (by others and myself, respectively). <g>
The typical problem manifests as a series of kernal Oopses, usually
involving rxvt (I run a lot of them), generally several hours before the
system finally crashes. Crashes themselves appear to be associated with
chron jobs -- I suspect a problem ultimately with memory or virtual
memory subsystems.
Freshmeat lists a number of memory testing utilities. You can also try
to isolate specific portions of RAM by using boot options to restrict
your system to less than the full physical RAM available. You may also
wish to run a badblocks test (read-only, or write (destructive!!))
across your disk partitions, in particular any swap partitions you may
have.
On Mon, Mar 27, 2000 at 12:14:31PM +0100, José Luis Gómez Dans wrote:
> Hi!
> I have installed potato recently (with the 2.2.8 install disks)
> on my Athlon system (someone might remember the problems I had compiling
> a new kernel for it; well that hasn't _yet_ happpened). This system has
> 2.2.14, and I am experiencing a rather lot of crashes. Most of the time,
> these are just pointer de-references (NULL pointers instead of
> meaningful stuff), and ensuing kernel panics. One that happened today:
>
> Mar 27 11:10:50 faust kernel: Call Trace: [sys_select+270/1264]
> [system_call+52/56]
> Mar 27 11:10:50 faust kernel: Code: 8b 01 89 03 85 c0 74 27 8b 53 04 85 d2 75
> 0e 89 19 89 c8 2b
> Mar 27 11:10:55 faust kernel: Unable to handle kernel NULL pointer
> dereference at virtual address 00000000
--
Karsten M. Self (kmself@ix.netcom.com)
What part of "Gestalt" don't you understand?
http://gestalt-system.sourceforge.net/
