Kernel panic after yum update

cwatson · Post by **cwatson** » 2010/10/11 16:33:38

Hello,
I powered down my system Friday after the most recent yum update (I believe the kernel update was to 2.6.9-89). When I stopped by last night, the machine wouldn't boot up. The message I received was
[code]
Kernel panic: VFS: Unable to mount root fs on unknown-block(0,0)
[/code]

I hadn't changed anything else (except for some other random packages that were to be updated). This machine is running CentOS 4.8, and is 64-bit. I'm running memtest86+ on it currently, so I can't get other info for now. [As an aside, it's gone through 7 passes with no errors; can I conclude that it isn't a failure in a memory module? I believe there are 4 2Gb sticks in there.]

When I booted from a CentOS 5.2 LiveCD, it mounted the filesystems with no problems. I could see all the data (thankfully!), and was able to back everything up. I noticed that the boot partition was completely full, so I removed the oldest kernel. This didn't fix the problem. I am considering upgrading to CentOS 5 (as I mention in another forum post), but would much rather keep it how it is for now, maybe until CentOS 6 comes out.

Does anyone know how to solve this particular problem? If necessary, I will stop memtest and get information via the LiveCD.

Thanks,
Chris

pschaff · Post by **pschaff** » 2010/10/11 16:59:07

The latest CentOS 4 kernel is 2.6.9-89.29.1.EL. The original 4.8 kernel was 2.6.9-89.EL. Did you try booting the previous kernel 2.6.9-89.0.29.EL?

P.S. Odd how we went from a long series of 2.6.9-89.0.X.EL to 2.6.9-89.29.1.EL but that is the way it came down from upstream.

cwatson · Post by **cwatson** » 2010/10/11 17:01:17

I did try booting some of the earlier kernels (I believe the ones that ended with ELsmp; not sure if that was correct...). I saw the same error.

Chris

pschaff · Post by **pschaff** » 2010/10/11 17:08:49

May want to check the SMART status of the disk if it supports it. Should be able to do that from the LiveCD. I'd also have a careful look at /boot/grub/grub.conf - perhaps something got corrupted because of the full partition.

cwatson · Post by **cwatson** » 2010/10/11 17:24:04

Here's part of grub.conf:
[code]
# grub.conf generated by anaconda
#
# Note that you do not have to rerun grub after making changes to this file
# NOTICE: You have a /boot partition. This means that
# all kernel and initrd paths are relative to /boot/, eg.
# root (hd0,1)
# kernel /vmlinuz-version ro root=/dev/sda3
# initrd /initrd-version.img
#boot=/dev/sda
default=0
timeout=5
splashimage=(hd0,1)/grub/splash.xpm.gz
hiddenmenu
title CentOS (2.6.9-89.0.29.ELsmp)
root (hd0,1)
kernel /vmlinuz-2.6.9-89.0.29.ELsmp ro root=LABEL=/ rhgb quiet
initrd /initrd-2.6.9-89.0.29.ELsmp.img
title CentOS (2.6.9-89.0.29.EL)
root (hd0,1)
kernel /vmlinuz-2.6.9-89.0.29.EL ro root=LABEL=/ rhgb quiet
initrd /initrd-2.6.9-89.0.29.EL.img
title CentOS (2.6.9-89.0.25.ELsmp)
root (hd0,1)
kernel /vmlinuz-2.6.9-89.0.25.ELsmp ro root=LABEL=/ rhgb quiet
initrd /initrd-2.6.9-89.0.25.ELsmp.img
title CentOS (2.6.9-89.0.25.EL)
root (hd0,1)
kernel /vmlinuz-2.6.9-89.0.25.EL ro root=LABEL=/ rhgb quiet
initrd /initrd-2.6.9-89.0.25.EL.img
title CentOS (2.6.9-89.0.23.EL)
root (hd0,1)
kernel /vmlinuz-2.6.9-89.0.23.EL ro root=LABEL=/ rhgb quiet
initrd /initrd-2.6.9-89.0.23.EL.img
title CentOS (2.6.9-89.0.23.ELsmp)
root (hd0,1)
kernel /vmlinuz-2.6.9-89.0.23.ELsmp ro root=LABEL=/ rhgb quiet
initrd /initrd-2.6.9-89.0.23.ELsmp.img
title CentOS (2.6.9-89.0.20.EL)
root (hd0,1)
kernel /vmlinuz-2.6.9-89.0.20.EL ro root=LABEL=/ rhgb quiet
initrd /initrd-2.6.9-89.0.20.EL.img
title CentOS (2.6.9-89.0.18.ELsmp)
root (hd0,1)
kernel /vmlinuz-2.6.9-89.0.18.ELsmp ro root=LABEL=/ rhgb quiet
initrd /initrd-2.6.9-89.0.18.ELsmp.img
title CentOS (2.6.9-89.0.18.EL)
root (hd0,1)
kernel /vmlinuz-2.6.9-89.0.18.EL ro root=LABEL=/ rhgb quiet
initrd /initrd-2.6.9-89.0.18.EL.img
title CentOS (2.6.9-89.0.16.EL)
root (hd0,1)
kernel /vmlinuz-2.6.9-89.0.16.EL ro root=LABEL=/ rhgb quiet
initrd /initrd-2.6.9-89.0.16.EL.img
title CentOS (2.6.9-89.0.16.ELsmp)
root (hd0,1)
kernel /vmlinuz-2.6.9-89.0.16.ELsmp ro root=LABEL=/ rhgb quiet
initrd /initrd-2.6.9-89.0.16.ELsmp.img
title CentOS (2.6.9-89.ELsmp)
root (hd0,1)
kernel /vmlinuz-2.6.9-89.ELsmp ro root=LABEL=/ rhgb quiet
initrd /initrd-2.6.9-89.ELsmp.img
[/code]

Below that are only older RHEL kernels (e.g. 2.6.9-42) and "Other", which says "rootnoverify hd(0,0)" and "chainloader +1". Does this look normal? I notice that the line "boot=/dev/sda" is commented... Should I uncomment that?

cwatson · Post by **cwatson** » 2010/10/11 19:18:46

I forgot to mention that this is what appears before the "Kernel panic" line:
[code]
Decompressing Linux...done.
Booting the kernel.
i8042.c: Can't read CTR while initializing i8042.
Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0)
[/code]

I've tried booting with the previous 3 or 4 kernels; I get the same thing each time.

Chris

pschaff · Post by **pschaff** » 2010/10/11 21:15:30

A google on [url=http://www.google.com/search?q=i8042.c%3A+Can%27t+read+CTR+while+initializing+i8042.&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a#about:blank]i8042.c: Can't read CTR while initializing i8042[/url] finds several mentions of USB problems. Is there a USB drive attached that may not have been there before? Anything in the BIOS changed?

cwatson · Post by **cwatson** » 2010/10/11 21:24:35

No, there were no USB devices.
I think what I'm going to do is just upgrade to CentOS 5, once I'm satisfied that VirtualBox will properly run the old VMware machine.

Thanks,
Chris

CentOS