Adam,

see below:

On Mon, 3 May 2004 09:57 pm, Adam Felix Bogacki wrote:
> Thanks Greg.
>
> >Adam,
> >
> >Have you got:
> >initrd=/initrd.img
> >
> >in your lilo.conf file?
>
> Yep.
>
> >If you can boot with another kernel why don't you just put in the old
> > 2.6.5 kernel (it is in /var/cache/apt/archives)
>
> Thanks, I found it. But how do I do that ?

dpkg -i kernel-image.........

>
> >your mouse problem with XFree86 and 2.6.X kernel is all over the net. Just
> >search in google.  The quick fix is something like this:
> >
> >edit /etc/X11/XF86Config-4
> >
> >find a line that has /dev/psaux or /dev/usb/mouse0 or similar
> >replace with /dev/input/mice and restart X.
>
> WhenI did that X would not start. I have two entries in separate
> paragraphs under 'pointing
> device' (or whatever), 'dev/psaux' and '/dev/input/mice' ... changing
> the first crashes X.
>
> ........<snip>....

Leave it as is if both are in there.  I have not installed a debian configured 
XFree86 for a while, as I have my own config, but as usual, they have taken 
care of every possibility for you.
> . 
>
> >Also I try and keep the last working kernel at /vmlinuz.old so I can
> > always boot from that if I get into trouble.

should be a symlink into /boot of the previous kernel (only if the kernel 
version has changed not the packaging version)
/vmlinuz.old -> /boot/vmlinuz-2.X.X


>
> ... is there a symlink from /vmlinuz.old to /var/cache/apt/archives ?
>
> ...but more importantly, I keep getting the message
>
> >"Error -3 While decompressing !
> >
> >> c0355f8c(1027641439)->d7c20000(4096)
> >> VFS: Cannot open root device "345" or unknown-block (3,69)
> >> Please append a correct "root=" boot option
> >> Kernel panic: VFS:Unable to mount root fs on unknown-block (3,69)"
>

what this means is it can not find the root device to mount.

There are several reasons for this, but since you are using a debian kernel, I 
am going to assume that things are OK (though I shouldn't assume, 'cause it 
don't appear that way.)

I suggest you remove that old 2.6.5 kernel and install the previous one as 
this is gonna be hard to figure out from here.

It boils down to the fact that the kernel can't see /dev/hdb5 for some reason, 
but it should be looking for the initrd.img, which it doesn't seem to be 
loading.  Something is just not right.

Debian kernels don't have drivers in them to boot many (if any) devices, but 
put them all in the /initrd.img file and load them once the kernel is in 
(read booted and mounted the RAM disk) the RAM disk.

the above usually happens when you haven't put initrd.img in lilo.

But.  There is the line "Error -3 While decompressing !".  This could mean the 
initrd.img is corrupt. Doh!

And that probably stems from your /boot space problem.

So how to fix it.

You are going to have to figure out if the initrd.img file is corrupt. If it 
is, reinstall the pkg.

Good luck,
Greg.

> and the boot process hangs.
>
> lilo.conf (attached) has
> boot=/dev/hda
> root/dev/hdb5
> but neither of them, or 'root=/dev/hda1'
> work in allowing the boot process to continue. There must be a simple
> solution.
>
> [Sorry about the graphic mess below. I transcribed it by hand from Linux
> to XP
> drives, as I had no cut & paste. It looked OK in Moz-Thunderbird]
>
> Any ideas ?
>
> Adam Bogacki,
> [EMAIL PROTECTED]
>
> >On Sun, 2 May 2004 08:40 pm, Adam Felix Bogacki wrote:
> >>> Oops, I forgot to run lilo ... but I then got the error message
> >>>
> >>> "/dev/mapper/control: open failed: No such device
> >>> Is device mapper driver missing from kernel ?
> >>> Incompatible libdevmapper 1.00.15-ioctl (2004-04-07)  (compat) and
> >>> kernel driver
> >>> Warning: The boot sector and map file are on different disks
> >>> Added Linux *
> >>> Skipping /vmlinuz.old
> >>> Added DOS"
> >>>
> >>> when I try to reboot  from C: using the newly installed & apt-upgraded
> >>> kernel I get
> >>>
> >>> "Error -3 While decompressing !
> >>> c0355f8c(1027641439)->d7c20000(4096)
> >>> VFS: Cannot open root device "345" or unknown-block (3,69)
> >>> Please append a correct "root=" boot option
> >>> Kernel panic: VFS:Unable to mount root fs on unknown-block (3,69)"
> >>>
> >>> Tux:~# df
> >>> Filesystem    1K-blocks     Used             Available    Use    
> >>> Mounted on /dev/hdb5      1369008        218325        1079997      
> >>> 17%     / tmpfs                191980                    0        
> >>> 191980 0%     /dev/shm
> >>> /dev/hda1    19991104     10655424      9335680       54%      /mnt/DOS
> >>> /dev/hdb2              7746             6970                376
> >>> 95%      /boot
> >>> /dev/hdb9     18951504       3723408    14265404       21%     /home
> >>> /dev/hdb6      12342396      1934112       9781352      17%      /usr
> >>> /dev/hdb7        1872960        435864        1341916      25%
> >>> /usr/local
> >>> /dev/hdb8        1824768        298408        1433588     18%      /var
> >>> /dev/hde5         7692876                   8         7302092       1%
> >>>    /home/cyberia
> >>> /dev/hde6         1344016                   8         1275736       1%
> >>>    /usr/cyberia
> >>>
> >>> ... previously /boot was showing 100% full.
> >>>
> >>> Tux:/boot# ls . -la
> >>> system.map-2.6.5-1-686
> >>> 787887 lost+found
> >>> 1024
> >>> boot.0300
> >>> 512
> >>> map
> >>> 73728
> >>> vmlinuz-2.6.5-1-686
> >>> 1125072
> >>> config-2.6.5-1-686
> >>> 52167
> >>> initrd.img-2.6.5-1-686
> >>> 4009984
> >>> and four symlinks
> >>> range: 31-34
> >>>
> >>> The boot error message makes reference to an unknown block. This may be
> >>> due to
> >>> my use of 'parted' to increase space in /usr which was previously 99%
> >>> full. This was successful, but my attempt to increase the size of 
> >>> /boot was not -
> >>>  I'm at a loss to understand why.
> >>>
> >>> "[EMAIL PROTECTED] /root %  parted  /dev/hdb
> >>> ----<snip>----
> >>> Information: The operating system thinks the geometry on
> >>> /dev/ide/host0/bus0/target1/lun0/disc is
> >>> 4865/255/63. Therefore, cylinder 1024 ends at 8032.499M
> >>>
> >>> (parted)  print
> >>> -----<snip>-----
> >>> Disk label type: msdos
> >>> Minor           Start              End
> >>> Type              Filesystem          Flags
> >>> 2                    0.031                   7.844
> >>> Primary            ext3                   boot
> >>> 1                    7.844         38162.219             extended
> >>> 5                956.997           2337.583
> >>> logical               ext3
> >>> 7              2384.679           4243.732
> >>> logical               ext3
> >>> 8              4290.829            6102.817
> >>> logical               ext3
> >>> 9              6196.979          32153.532            logical
> >>>    ext3
> >>> 6             32247.694         37244.443
> >>> logical               ext3
> >>> 10           37244.474         38162.219            logical
> >>>         ext3
> >>>
> >>> (parted)  resize  2  0.031    25.567       [trying to enlarge  /boot]
> >>>
> >>> Warning: You requested to resize the partition to 0.031-25.567Mb.
> >>> The closest Parted can manage is 0.031-7.844MB."
> >>>
> >>> I understand that Parted is choosing options consistent with
> >>> interactions between
> >>> firmware and software, but it has previously stated that
> >>> "...cylinder 1024 ends at 8032.499Mb" so that there should be plenty of
> >>> room to increase the size of /boot, given that it should be placed
> >>> lower than cylinder 1024.
> >>>
> >>> In any case, the size of  /boot may not be the reason for the boot
> >>> process failing.
> >>>
> >>> What would be a " correct "root=" boot option ? I have unsuccessfully
> >>> tried the option
> >>> "boot: /dev/hdb5" .
> >>>
> >>> What am I missing here ? I'm currently reduced to tweaking my DOS drive
> >>> ...
> >>>
> >>> Adam Bogacki,
> >>> [EMAIL PROTECTED]
-- 
SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/
Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html

Reply via email to