[SLE] Chos!!! (was: RE: [SLE] HELP (again :-/) ?hardware err???)
Well, it seems this si all just a lilo problem. Using /sbin/chos is somewhat more terse but /Works/! If anyone know why Lilo is croaking, I'd love to know. But after I've had a dead system for 3 weeks, I don't care. Now I need to determine how to make chos probe for a second enet card (the lilo append=0,0,eth1 option) Thanks... -mab At 3:48 PM -0600 3/13/00, Michael Bartosh wrote:
One update:
it appears there is something is wrong with my floppy drive.
Although I can boot from it, when I try to copy something to it (like bzimage) the /whole machine/ locks up so hard that it requires I throw the power switch on the back of the machine (the atx-esq soft switch does not work).
This of course keeps syslinux from working. I could create a loopback and dd ontothe drive (this worked a couple of days ago) but frankly I'm fed up with this to an extreme and might possibly pursue the dd thing tomorrow.
Is there any reason why syswlinux would work to boot a machine while Lilo would fail??? (even lilo on a floppy dr?)
-mab
At 1:29 PM -0600 3/13/00, Michael Bartosh wrote:
On Mon, 13 Mar 2000, Tim Duggan wrote:
Hi
-----Original Message----- From: Michael Bartosh [SMTP:bartosh@ocean.tamu.edu] Sent: Friday, March 10, 2000 9:32 PM
On Fri, 10 Mar 2000, Tim Duggan wrote:
-----Original Message----- From: Michael Bartosh [SMTP:bartosh@ocean.tamu.edu] Sent: Friday, March 10, 2000 4:55 AM
Well, turns out I an make it boot off of some floppies. Like a freesh
Suse one. Which wouldn't be bad for a temporary solution- after I boot into the installer I can chose 'bboot installed system, and most everything works-- except my 3com 509 for some reason (works ine in my notmal srtup) and that kernel doesn't have any IPMasq on.
So I tried to be smart.
dd if=/dev/fdo of=/tmp/file.img (switch disks) dd if=/tmp/file.img of=/dev/fd0
mount floppy, replace the old ram disk w/ the one I usually boot from and the kernel w/ my own home brewed (that works fine afaik) but poof- reboot, and it's
LI
again
IIRC LI is an indication that lilo cannot find the kernel image. If I understand lilo correctly, it looks for the kernel at a very specific position on the disk, if it has changed, lilo gets lost. (Or I could be wrong and its a different problem, no matter).
Oops, this is lilo not being able to execute the second stage boot loader. From /usr/doc/packages/lilo/README "This can either be caused by a geometry mismatch or by moving /boot/boot.b without running the map installer"
Try booting your system as you describe with the install floppys and use Yast to create a boot floppy. See if that floppy will boot.
I seem to remember that you can also recompile your kernel using # make bzdisk (instead of bzImage or bzlilo) I think that's right, double check the docs to be sure as I don't
have
access to a linux box right now (again I could be wrong).
I did make bzImage and it still croaks. The pattern is that any installer kernel boots the machine, and any live kernel- or lilo itself- does not. Suse, deb and redhat floppies all boot the machine. But any other kernel fails.
Just so I understand, you can build the kernel with '# make bzImage' but '# make bzlilo' and '# make bzdisk' fail. Or do they seem to work and just not boot?
It works but does not boot the machine.
If you have made a good bzImage you can try to move it directly to a floppy (make sure to use a high density floppy, 1.44MB and don't mount it) with # cp /usr/src/linux/arch/i386/boot/bzImage /dev/fd0 This should boot if the kernel is good (you could also try installing lilo on a floppy)
Lilo on a floppy installs correctly but sticks at LI. This is consistent with the observation that lilo fails but syslinux does not. I am in the process of trying to get syslinux to work w/ my current kernel.
I need a refresher, since the old CMM (cranial memory module) isn't what it used to be. Do you have a separate /boot partition or is it part of the / partition?
part of /
How big is your disk and what cylinders does /boot reside? The relavent portions of /etc/fstab and output from "# fdisk -l" will do.
heretic: ~ # fdisk -l
Disk /dev/hda: 255 heads, 63 sectors, 1027 cylinders Units = cylinders of 16065 * 512 bytes
Device Boot Start End Blocks Id System /dev/hda1 * 1 26 208813+ 83 Linux /dev/hda2 27 53 216877+ 82 Linux swap /dev/hda3 54 403 2811375 8e Unknown /dev/hda4 404 1027 5012280 5 Extended /dev/hda5 404 531 1028128+ 83 Linux /dev/hda6 532 1027 3984088+ 8e Unknown
Disk /dev/hdc: 16 heads, 63 secters, 26500 cylinders Units = cylinders of 1008 * 512 bytes
Device Boot Start End Blocks Id System /dev/hdc1 1 16254 8191984+ 83 Linux /dev/hdc2 16255 26500 5163984 5 Extended /dev/hdc5 16255 19302 1536160+ 83 Linux /dev/hdc6 19303 21334 1024096+ 8e Unknown
_____ Naturally:
heretic: ~ # df Filesystem 1k blocks Used Available Use% Mounted on /dev/hda1 202220 45343 146437 24% / /dev/hda5 1011928 21936 938588 2% /var /dev/hdc5 1511968 519948 915212 36% /usr /dev/local_vg/opt 297485 35671 246454 13% /opt /dev/local_vg/src 604736 90240 483776 16% /usr/src
hdc1 is a reiser partition that I cannot mount until I get my own kernel w/ the new reiser in it.
Thx for your help...
Tim
-- To unsubscribe send e-mail to suse-linux-e-unsubscribe@suse.com For additional commands send e-mail to suse-linux-e-help@suse.com Also check the FAQ at http://www.suse.com/Support/Doku/FAQ/
-- To unsubscribe send e-mail to suse-linux-e-unsubscribe@suse.com For additional commands send e-mail to suse-linux-e-help@suse.com Also check the FAQ at http://www.suse.com/Support/Doku/FAQ/
-- To unsubscribe send e-mail to suse-linux-e-unsubscribe@suse.com For additional commands send e-mail to suse-linux-e-help@suse.com Also check the FAQ at http://www.suse.com/Support/Doku/FAQ/
participants (1)
-
bartosh＠ocean.tamu.edu