2.5.14 -- dmesg corruption -- Error seeking in /dev/kmem, Symbol #floppy, value d990e000, Error adding kernel module table entry

2.5.14 -- dmesg corruption -- Error seeking in /dev/kmem, Symbol #floppy, value d990e000, Error adding kernel module table entry

[Miles Lane]

I have reported this at least twice before.  At least one
other person has reported reproducing this problem.
Would someone please care enough to help me track this
down?

All the kernel messages seem to get emitted to the
VT/debug serial port during the boot process, but
the ouput to /var/log/kernel is currupted.

May  6 01:21:19 turbulence syslogd 1.4.1: restart.
May  6 01:21:20 turbulence syslog: syslogd startup succeeded
May  6 01:21:20 turbulence kernel: klogd 1.4.1, log source = /proc/kmsg 
started.
May  6 01:21:20 turbulence kernel: Inspecting /boot/System.map-2.5.14
May  6 01:21:20 turbulence syslog: klogd startup succeeded
May  6 01:21:20 turbulence kernel: Loaded 20642 symbols from 
/boot/System.map-2.5.14.
May  6 01:21:20 turbulence kernel: Symbols match kernel version 2.5.14.
May  6 01:21:20 turbulence kernel: Error seeking in /dev/kmem
May  6 01:21:20 turbulence kernel: Symbol #floppy, value d990e000
May  6 01:21:20 turbulence kernel: Error adding kernel module table entry.

Here's the high-level options from .config
(I have seen this problem with many, many 2.5.x
kernels with varying options selected):

#
# Automatically generated by make menuconfig: don't edit
#
CONFIG_X86=y
CONFIG_ISA=y
CONFIG_UID16=y

#
# Code maturity level options
#
CONFIG_EXPERIMENTAL=y

#
# General setup
#
CONFIG_NET=y
CONFIG_SYSVIPC=y
CONFIG_BSD_PROCESS_ACCT=y
CONFIG_SYSCTL=y

#
# Loadable module support
#
CONFIG_MODULES=y
CONFIG_KMOD=y

#
# Processor type and features
#
CONFIG_MK7=y
CONFIG_X86_WP_WORKS_OK=y
CONFIG_X86_INVLPG=y
CONFIG_X86_CMPXCHG=y
CONFIG_X86_XADD=y
CONFIG_X86_BSWAP=y
CONFIG_X86_POPAD_OK=y
CONFIG_RWSEM_XCHGADD_ALGORITHM=y
CONFIG_X86_L1_CACHE_SHIFT=6
CONFIG_X86_TSC=y
CONFIG_X86_GOOD_APIC=y
CONFIG_X86_USE_3DNOW=y
CONFIG_X86_USE_PPRO_CHECKSUM=y
CONFIG_X86_MCE=y
CONFIG_X86_MCE_NONFATAL=y
CONFIG_X86_MCE_P4THERMAL=y
CONFIG_NOHIGHMEM=y
CONFIG_MTRR=y
CONFIG_PREEMPT=y
CONFIG_X86_UP_APIC=y
CONFIG_X86_LOCAL_APIC=y
CONFIG_HAVE_DEC_LOCK=y

#
# ACPI Support
#
CONFIG_ACPI=y
CONFIG_ACPI_BOOT=y
CONFIG_ACPI_BUS=y
CONFIG_ACPI_EC=y
CONFIG_ACPI_INTERPRETER=y
CONFIG_ACPI_PCI=y
CONFIG_ACPI_POWER=y
CONFIG_ACPI_SLEEP=y
CONFIG_ACPI_SYSTEM=y
CONFIG_ACPI_AC=m
CONFIG_ACPI_BATTERY=m
CONFIG_ACPI_BUTTON=m
CONFIG_ACPI_FAN=m
CONFIG_ACPI_PROCESSOR=m
CONFIG_ACPI_THERMAL=m
CONFIG_ACPI_DEBUG=y
CONFIG_PCI=y
CONFIG_PCI_GOANY=y
CONFIG_PCI_BIOS=y
CONFIG_PCI_DIRECT=y
CONFIG_PCI_NAMES=y
CONFIG_HOTPLUG=y

#
# PCMCIA/CardBus support
#
# CONFIG_PCMCIA is not set

#
# PCI Hotplug Support
#
CONFIG_KCORE_ELF=y
CONFIG_BINFMT_AOUT=y
CONFIG_BINFMT_ELF=y
CONFIG_BINFMT_MISC=y
CONFIG_PM=y

Re: 2.5.14 -- dmesg corruption -- Error seeking in /dev/kmem, Symbol #floppy, value d990e000, Error adding kernel module table entry

[Keith Owens]

On Mon, 06 May 2002 02:33:05 -0700, 
Miles Lane <miles@megapathdsl.net> wrote:
>I have reported this at least twice before.  At least one
>other person has reported reproducing this problem.
>Would someone please care enough to help me track this
>down?

It is the old klogd code that tries to decode oops and rarely gets it
right.  Start klogd as 'klogd -x'.