Bug#279370: does not boot: 'CLAIM FAILED' error on IBM 7043-260.
Package: kernel-image-2.6.8-power3
Version: 2.6.8-6
Severity: important
Hi,
i tried to install sarge with daily d-i on a ppc64 machine and failed in a
number of interesting ways :). i'm new to ppc hardware so be aware that i
might get the arch details wrong.
the machine is a standard IBM 7043-260 (CHRP) with 2 x 64bit POWER3
processors, 1 GiB of RAM and 2 x SCSI disks on the on-board symbios
controller up-to-date firmware.
the only unusual detail is the graphic card: a matrox g200 instead of the
standard gxt3000p (lucky me ;). all hardware appear to work properly
under AIX.
the short summary:
- the machine doesn't boot due to a 'CLAIM FAILED' error (solved thanks
to leighbb on #ppc64).
the long story:
after getting to the OF prompt i do a couple of timid tries to boot from
the cdrom following instructions found on the net, but fail (probably due to
my newbie status. people say the machine has a chance to boot with yaboot).
anyhow, i set up a netboot server following the instruction at:
http://master.penguinppc.org/~hollis/linux/rs6k-netboot.shtml
and using various kernel+initrd found at:
http://people.debian.org/~luther/d-i/images/*/power3/netboot/
i finally managed to start the boot process with:
> boot net:<server_ip>,,<client_ip>
after loading the 5+ MiB vmlinuz.initrd the OF throws a 'CLAIM FAILED' error.
i'm lost.
during my first irc session with leighbb on #ppc64 (or #mklinux) i learn
a nice black magic trick:
- with an hexeditor look for the values '00 0c 00 00' and '00 00 40 00'
in the first few bytes of the vmlinuz.initrd file and change both them of
them to 'ff ff ff ff'.
- into OF change the real-base value:
> printenv real-base
> setenv real-base 100000
> boot net:<server_ip>,,<client_ip>
wow! the penguin smiles on the screen (the unexpected matrox g200 saved me
from the serial console).
the black magic has to do with the huge size of the kernel+initrd and with
hardcoded values on where to expect the firmware (leighbb knows more than me)
the d-i process goes on nicely until i get to the storage hardware
configuration that appear to hang. but this is another story.
cheers,
alessandro
-- System Information:
Debian Release: 3.1
APT prefers testing
APT policy: (500, 'testing')
Architecture: i386 (i686)
Kernel: Linux 2.6.8-1-k7-smp
Locale: LANG=C, LC_CTYPE=C
Reply to: