Red Hat Bugzilla – Bug 56498
cpqarray.o fails to load during boot with error: init_module: Input/Output error
Last modified: 2007-04-18 12:38:21 EDT
From Bugzilla Helper:
User-Agent: Mozilla/4.0 (compatible; MSIE 5.5; Windows NT 5.0)
Description of problem:
Installed both i386 and i686 kernel version 2.4.9-13 with the same
results. Server is a Compaq Proliant 5500 with a single P-III 500 Xeon
CPU, 3.5GB RAM, 3200 Compaq RAID Array.
Kernel 2.4.7 has no problems. Downloaded, compiled and tested kernel-
2.4.13-0.5 and encountered no problems. Recompiled both i686 and i386
versions of kernel-2.4.9-13 on this server while running kernel 2.4.7, but
recompiled kernel had the same problem.
We tried different types of RAID controllers (all Compaq, but different
model #'s). All produced the same error. kernel-2.4.9-13 on a non Xeon
system with similar RAID had no problems. I suspect it is something about
the Xeon processor, but I have no way to determine that, as the mainboard
prevents us from putting anything but a Xeon in it. kernel-2.4.9-13
worked fine on a Compaq Proliant 1850R (P-III 500 non-Xeon) with the same
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. Install stock RH7.2 (kernel-2.4.7-10)
2. Install i686 version of kernel-2.4.9-13
3. Choose kernel-2.4.9-13 during next boot and watch the kernel-panic
Actual Results: Tried installing the kernel-debug-2.4.9-13 but it
produced no more output on the insmod failure than the non-debug kernel
(which is to say it only output the init_module error). I'm not familiar
with the debugger, so I didn't get any farther than that.
I considered just applying cpqarray_2.4.20D_for_2.4.9.patch from Compaq,
but there were about 6 other patches that were already being applied that
effected the key files related to the cpqarray, so the patch from Compaq
wouldn't work...and I didn't want to create a patch to your patches...
Expected Results: Kernel should install with little error on all i386 and
better Intel architectures.
I'm going to experiment with removing the specific pieces of each of those
patches that deal with cpqarray.h, cpqarray.c, genhd.c, ida_ioctl.h and
smart1,2.h until I can successfully apply
cpqarray_2.4.20D_for_2.4.9.patch. I'm open for any useful suggestions or
assistance...perhaps a debugger command that might produce some useful
output after the panic?
can you attach "lspci" and "lspci -v" output ? (from the working kernel?)
Created attachment 40328 [details]
lspci -v output
Some added info on the kernel problems I'm experiencing. Aside from not being
able to get 2.4.9-13 to work at all, 2.4.7-10 that ships with RH7.2 doesn't
detect SMP. The BOOT kernel that the installer uses detects SMP to the point
that it knows to install the SMP kernel, but when the server boots after
install, the SMP kernel doesn't detect an SMP capable motherboard.
I'll attach the dmesg output of the boot.
Created attachment 40329 [details]
Thanks for the new kernel. It boots well and I no longer have the problem
loading cpqarray.o. However, even with adding "acpismp=force" (minus the
quotes, of course) to both the enterprise and smp kernel lines of grub.conf,
the system still fails to detect the 2nd CPU (still says that an SMP
motherboard isn't detected).
Is there any output from anything that I can give you to help me work this
problem out? Should we move this problem to a new bug report or continue to
work from here?
Finally met with some Compaq reps on this. There was a buried BIOS option for
APIC settings. Turns out the APIC was disabled on the motherboard. Once Full
Map was enabled, SMP was detected and the additional CPU's were detected.
Thanks for your time and assistance in this.