Lilo error after 10.2 upgrade

PurpleMotion · 10-12-2005, 04:35 AM

Hi

I have just upgraded installed packages on a slackware 10.1 install to slackware 10.2 and updated config files.

I am about to reboot, but when I run lilo to make sure the boot loader is installed, i get the following error:

Code:

Fatal: VolumeID read error: sector 0 of /dev/sdb not readable

Now this is strange, this device isn't even installed. My disks is a hardware raid5 array installed on sda.

Anyone know what could be wrong? Or might there not be anything wrong at all? And everything will probably be ok if I reboot?

Here's the installed partitions:

Code:

/dev/sda1        swap             swap        defaults         0   0
/dev/sda3        /                ext3        defaults         1   1
/dev/sda2        /boot            ext3        defaults         1   2
/dev/sda5        /var/log         ext3        rw,noexec,nosuid,nodev         1   2
/dev/sda6        /tmp             ext3        rw,noexec,nosuid,nodev
/dev/sda7        /home            ext3        rw,nodev         1   2
/dev/cdrom       /mnt/cdrom       iso9660     noauto,owner,ro  0   0
/dev/fd0         /mnt/floppy      auto        noauto,owner     0   0
devpts           /dev/pts         devpts      gid=5,mode=620   0   0
proc             /proc            proc        defaults         0   0

Here's the mounted volumes:

Code:

/dev/sda3 on / type ext3 (rw)
proc on /proc type proc (rw)
/dev/sda2 on /boot type ext3 (rw)
/dev/sda5 on /var/log type ext3 (rw,noexec,nosuid,nodev)
/dev/sda6 on /tmp type ext3 (rw,noexec,nosuid,nodev)
/dev/sda7 on /home type ext3 (rw,nodev)
devpts on /dev/pts type devpts (rw,gid=5,mode=620)
usbfs on /proc/bus/usb type usbfs (rw)

PurpleMotion · 10-12-2005, 05:00 AM

Sorry, forgot lilos verbose output:

Code:

LILO version 22.5.9, Copyright (C) 1992-1998 Werner Almesberger
Development beyond version 21 Copyright (C) 1999-2004 John Coffman
Released 08-Apr-2004 and compiled at 00:18:50 on May 21 2004.

Warning: LBA32 addressing assumed
Reading boot sector from /dev/sda
Fatal: VolumeID read error: sector 0 of /dev/sdb not readable

heltreko · 10-12-2005, 05:19 AM

Take a look in /etc/lilo.conf and make sure you are trying to install lilo to /dev/sda and not /dev/sdb.

PurpleMotion · 10-12-2005, 05:31 AM

Quote:

Originally posted by heltreko
Take a look in /etc/lilo.conf and make sure you are trying to install lilo to /dev/sda and not /dev/sdb.

lilo.conf is ok. it hasn't been modified since last kernel upgrade.

Alien Bob · 10-12-2005, 07:58 AM

What is the output of

Code:

fdisk -l

? I had a strange problem with lilo too in Slackware 10.2 where I had a microdrive (/dev/hde) inserted in the PCMCIA slot of my laptop, and did not have it mounted.

Lilo refused to install a new bootloader and gave weird errors, unless I mounted the microdrive... even though /dev/hde was nowhere mentioned in /etc/lilo.conf

Eric

cb951303 · 10-12-2005, 08:39 AM

the error should be in lilo.conf ....

PurpleMotion · 10-12-2005, 08:44 AM

Quote:

Originally posted by Alien Bob
What is the output of

Code:

fdisk -l

? I had a strange problem with lilo too in Slackware 10.2 where I had a microdrive (/dev/hde) inserted in the PCMCIA slot of my laptop, and did not have it mounted.

Lilo refused to install a new bootloader and gave weird errors, unless I mounted the microdrive... even though /dev/hde was nowhere mentioned in /etc/lilo.conf

Eric

fdisk -l output follows:

Code:

Disk /dev/sda: 72.8 GB, 72802631680 bytes
254 heads, 63 sectors/track, 8885 cylinders
Units = cylinders of 16002 * 512 = 8193024 bytes

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1               1         122      976090+  82  Linux swap
/dev/sda2   *         123         128       48006   83  Linux
/dev/sda3             129         494     2928366   83  Linux
/dev/sda4             495        8885    67136391    5  Extended
/dev/sda5             495         561      536035+  83  Linux
/dev/sda6             562         579      143986+  83  Linux
/dev/sda7             678        8885    65672176+  83  Linux

Here is direct output from /proc/partitions
Notice those sdb partitions listed. Which are unknown to me.

Code:

major minor  #blocks  name

   8     0   71096320 sda
   8     1     976090 sda1
   8     2      48006 sda2
   8     3    2928366 sda3
   8     4          1 sda4
   8     5     536035 sda5
   8     6     143986 sda6
   8     7   65672176 sda7
   8    16  156290904 sdb
   8    17  156288321 sdb1

I noticed dmesg output the following when lilo is run:

Code:

Device 08:10 not ready.
 I/O error: dev 08:10, sector 0

Might be hardware error. I will run checks later.

But it doesn't make sense to me that lilo reports sdb

PurpleMotion · 10-12-2005, 08:53 AM

Quote:

Originally posted by cb951303
the error should be in lilo.conf ....

That would make sense. But I can't find anything wrong with it:

Code:

# Start LILO global section
boot = /dev/sda
prompt
timeout = 150
# Override dangerous defaults that rewrite the partition table:
change-rules
  reset
# Normal VGA console
vga = normal
# End LILO global section

# Linux bootable partition config begins
image = /boot/bzImage2.4.31-ow1
  root = /dev/sda3
  label = Linux-2.4.31ow1
  read-only
# Linux booble partition config ends

I've doublechecked for possible nasty formatting errors. though that kind of error never happened to me. vim does its job

Alien Bob · 10-12-2005, 11:00 AM

Quote:

Device 08:10 not ready.
I/O error: dev 08:10, sector 0

That is definitely /dev/sdb you see complaints about in dmesg....

Eric

PurpleMotion · 10-12-2005, 11:44 AM

Indeed it was sdb, and somewhat similar to your microdrive issue. I sat down and checked some old logs for issues regarding sdb, and behold:

Code:

Jul 14 15:01:34 mx03 kernel: hub.c: new USB device 00:1d.7-2, assigned address 2
Jul 14 15:01:37 mx03 kernel: Initializing USB Mass Storage driver...
Jul 14 15:01:37 mx03 kernel: usb.c: registered new driver usb-storage
Jul 14 15:01:37 mx03 kernel: scsi2 : SCSI emulation for USB Mass Storage devices
Jul 14 15:01:37 mx03 kernel:  sdb: sdb1
Jul 14 15:01:37 mx03 kernel: USB Mass Storage support registered.
Jul 14 15:05:52 mx03 kernel: usb.c: USB disconnect on device 00:1d.7-2 address 2

Someone connected a usb drive (I did think the 156290904 block size in the partition table was rather large..) to the server without my knowledge. (the server is remotely hosted.)
The sdb device was still known in the active partition table and caused problems as it obviously ain't mounted. That left two solutions, mount it or remove it. The latter would be my choice.
I reloaded slackwares hotplug system which smoothly removed the sdb device from the partition table and lilo worked just fine and dandy.

heltreko · 10-13-2005, 12:28 AM

This was an interesting lilo issue well worth reading.

Is this a desired behavior of lilo or should it be viewed as a problem and reported to someone working with developing lilo?

Alien Bob · 10-13-2005, 03:22 AM

As far as I am concerned this is either a bug in lilo or an undesired feature. But I have not delved too deep into the problem after I found out I could get away with mounting the microdrive before running lilo. There might be something I have overlooked.
Comments, anyone?

Eric

PurpleMotion · 10-13-2005, 06:38 AM

It's maybe more likely to be a issue with the usb subsystem in the 2.4 kernel than lilo. It obviously didn't cleanly remove the drive, as it was left in /proc/partitions as a "zombie".

KMcD · 10-13-2005, 06:50 AM

this has often happened to me and also happens for installation of 2.6.x kernels so is a lilo/hotplug prob rather then kernel prob (I'd guess).

Damn annoying betimes, having to plug back in usb device to install lilo.

Alien Bob · 10-13-2005, 06:58 AM

Quote:

Originally posted by PurpleMotion
It's maybe more likely to be a issue with the usb subsystem in the 2.4 kernel than lilo. It obviously didn't cleanly remove the drive, as it was left in /proc/partitions as a "zombie".

Well.... I have only experienced this with 2.6 kernels. When booting the same hardware with a 2.4 kernel, I do not have this problem. In fact, before I discovered how I could circumvent the problem by mounting my microdrive, I rebooted to a 2.4 kernel just to run lilo!

Eric