Skip to contentSkip to conversation

boot problem with kernel 2.4.19

DISCUSSION Christophe A. Gevaudan 1 participant 1 message Feb 1, 2003

lore
Share & tools
lore
#0 lore raw

Hi,
I'm new to this list and don't know if they're achives around to look for problems which occured in the past.

I used to run a 2.2.x SMP kernel for years with no problem so far. I installed a new distro from scratch a few weeks ago (SuSE 8.1) - I didn't want to upgrade the system because of the several upgrades I had done in the past.
Now, everything's ok so far, as long as I boot a single CPU kernel...

The SMP-kernel of the distro didn't work - I first thought it was a problem with my SCSI host adapter, but this was quickly fixed.
I compiled a new SMP-kernel where I stripped a lot of things away which I don't need (joystick, ISDN, hardware devices...).
The SMP-kernel still doesn't boot, and I'm bit confused about the message I get.

My system is based on a Gygabyte GA-6BXDU board, equipped with two Pentium III Coppermine "slot 1" CPUs clocked at 450 MHz (the board is clocked at 100MHz, the CPUs are 133MHz ones and run at 450 MHz when I setup the board for 600MHz CPU clock rate). The mainbord is based on the Intel 440BX chipset. The CPUs have both ID:0681 and Patch ID:000D. The SCSI chip on the mainboard is disabled, as it never worked correctly.
The system has 512 MB memory, two IDE disks, one IDE DVD drive and one SCSI CD-RW connected to an Adaptec AHA-2904 card - which runs an AHA-7850 chip (the aic78xxx_old driver works perfect).
The BIOS has the version v4.51PG, AGPBIOS for 6BXDU has version V.F1.

The "/" filesystem is on /dev/hdc1, the "/boot" filesystem is on /dev/hdc2. The boot options are "root=/dev/hdc1 vga=normal nmi_watchdog=1". Oh, I nearly forgot: I use "grub" for booting (not LILO).
The first thing I tried is to remove the "nmi_watchdog=1" statement from the boot options, but this switches the complete hardware off.

Now, here are the messages I noted from the boot phase until everything stops:
...
Total of 2 processors activated (1789.13 BogoMIPS).
ENABLING IO-APIC IRQs
... TIMER:vector=0x31 pin1=2 pin2=0
activating NMI watchdog ... OK.
testing the IO APIC .................................................... ....................................................................... done. Using local APIC timer interrupts.
calibrating APIC timer ...
...... CPU clock speed is 448.7852 MHz.
...... host bus clock speed is 99.7299 MHz.
cpu:0, clocks: 997299, slice: 332433 CPU0<T0:997296,T1:664818,D:15,S:332433,C:997299>
cpu:1, clocks: 997299, slice: 332433 CPU1<T0:997296,T1:332416,D:14,S:332433,C:997299>
checking TSC synchronization across CPUs: passed.
migration_task 0 on cpu=0
migration_task1 on cpu=1
mtrr: your CPUs had inconsistent fixed MTRR settings
mtrr: probably your BIOS does not setup all CPUs
ACPI: Subsystem revision 20020829
PCI: PCI BIOS revicion 2.10 entry at 0xfb150, last bus=1
PCI: Using configuration type 1

and then the system hangs...
I'm not sure if the PCI messages are relevant here, but I must admit that I have no clue what could happen here.
If I boot with the option "nmi_watchdog=1" replaced by "nosmp", the system boots without trouble (and without SMP) and runs normally.

Has somebody an idea what it could be?

Thanks for your help

Christophe