Re: I screwed up and cannot boot
On Sat, Mar 11, 2000 at 11:44:01PM -0800, email@example.com wrote:
> On Sat, Mar 11, 2000 at 03:45:25AM -0600, David J. Kanter wrote:
> > Well, I'm getting there. Yes, the new rescue and boot disks helped. I was
> > able to mount the partitions, run fsck on them (not while mounted, mind
> > you), and edit /etc/fstab and /etc/lilo.conf. Still there are problems.
Remember that you need to re-run LILO, not just change the config file.
This is because LILO runs when the partition is unmounted, and so the
config changes need to be 'hard coded' into the bootsector to have an
> > Lilo refuses to recognize that I've got any kernels; perhaps they are dead.
> > When I run lilo -D Testing (Testing is the label for my newest kernel) I
> > get that it's not recognized.
Please provide the error message via the gifts of cut & paste. An
additional comment on what you think's going on is helpful, but it
should always be in addition to, not instead of, the actual error
> > The only way I can get a reasonably functioning machine is if I use the new
> > rescue disk, and at the boot prompt type Linux single root=/dev/hda6. An
> > error comes up that I must type in the root password for maintenance, or
> > Control-D to go on.
Use the rescue and root disks to start the installation program.
Configure your keyboard, swap space and mount your previously
initialised partition (/dev/hda6). It will be mounted as "/target".
Then "Execute A Shell" and do :
# cd /target
# chroot .
You can then access all your existing binaries and libraries while using
the known good rescue kernel. This is enough to run lilo and compile a
kernel (although make xconfig is out for now ... :-)
As long as you pick the right options, compile, make lilo changes,
re-run lilo and reboot, you should get your old system back.
> This is standard for "single" mode boot.
> > If I type in the password, I've got access to my files
> > (I've pulled off my home directory), but I cannot startx.
> Whoa, Nellie! One thing at a time. Let's get your system up and
> running before we worry about X windows.
> Rather than starting X, you might try doing something useful like
> building yourself a working kernel -- install the kernel package from
> your installation media, do :
> make mrproper
> make xmenuconfig # Configures kernel
> make-kpkg --bzimage --revision=Custom_1.0 binary
> and install the resulting debs (man make-kpkg for more info).
> > If I type
> > Control-D, I end in a perpetual loop of cannot find module net-pf-1, at
> > which point I must reboot.
> Compiling and installing a kernel, with appropriate device support, will
> solve this issue.
from /usr/src/linux/Documentation/Configure.help :
Unix domain sockets
If you say Y here, you will include support for Unix domain sockets;
sockets are the standard Unix mechanism for establishing and
accessing network connections. Many commonly used programs such as
the X Window system and syslog use these sockets even if your
machine is not connected to any network. Unless you are working on
an embedded system or something similar, you therefore definitely
want to say Y here.
However, the socket support is also available as a module ( = code
which can be inserted in and removed from the running kernel
whenever you want). If you want to compile it as a module, say M
here and read Documentation/modules.txt. The module will be called
unix.o. If you try building this as a module and you have said Y to
"Kernel module loader support" above, be sure to add 'alias net-pf-1
unix' to your /etc/conf.modules file. Note that several important
services won't work correctly if you say M here and then neglect to
load the module.
Say Y unless you know what you are doing.
It sounds like this may apply.
If you compiled this as a module, try doing an 'insmod unix' after
booting to prevent the messages.
> > My latest kernel, which I've got on disk, doesn't work. I get kernel panic
> > messages that there is no initd.
> > Any ideas still? Potato will be out on CD soon so I could reinstall then. I
> > do have my /home. But I feel like I'm almost there.
It could be that you didn't select "Kernel support for ELF binaries"
(CONFIG_BINFMT_ELF). Make sure it's enabled.
> You're close. Try a kernel build. Let us know what happens.
> Karsten M. Self (firstname.lastname@example.org)
> What part of "Gestalt" don't you understand?