[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#427514: linux-image-2.6.21-1-amd64: System does not boot with 2.6.21-4



On Mon, Jun 04, 2007 at 06:55:14PM +0200, Jack Malmostoso wrote:
> Hi there,
> 
> since I have installed 2.6.21-4 from unstable I cannot boot my system
> anymore.
> First my configuration: an Asus A8N-E (nforce4) motherboard with two
> SATA disks. 
> The disks are configured in a software RAID1 using mdadm:
> 
> /dev/md0 is my /
> /dev/md1 is my swap
> /dev/md2 is my /home

please post the output for check:
cat /etc/fstab
cat /proc/mdstat
cat /etc/mdadm/mdadm.conf
 
> If I start the system with 2.6.20 it boots fine. Here's a picture of the
> end of the initramfs and the start of the normal boot:
> 
> http://www.flickr.com/photos/90819597@N00/530049031/
> 
> The system reaches /sbin/init and it boots just fine.
> 
> Here is what happens with 2.6.21:
> 
> http://www.flickr.com/photos/90819597@N00/530049047/
> 
> Something goes wrong in the initramfs, and despite /dev/md0 is started,
> it can't be mounted.
> At this point I have tried to toy around in the initramfs, read
> the /scripts directory but couldn't understand how to make the system
> move along. (btw, I could really use a pager in there... I hate vi :D)
> It should be noted that I can start my other two RAID arrays without
> problems!
> 
> Please ignore the fact that my 2.6.21 is now a snapshot from the Debian
> Kernel Team (btw, I have tried also 2.6.22rc3 but the problem is still
> there) as the problem arised with the upgrade from 2.6.21-3 and -4 from
> unstable.
> 
> I don't recall any udev or initramfs-tools upgrade that could have
> broken the system, especially since 2.6.20 works just fine. Additionally
> the changelog between -3 and -4 is extremely short so no luck there.
> 
> Thanks for any help you might give me.

try to boot with break=mount and please check which devices are up
adding rootdelay=9 should shorten the time to land into console
also you have more as pager there in busybox ;)

check inside too
fstype /dev/md0

 
regards

--
maks



Reply to: