Scenario: almost 6-year old HP Slimline with dual-core AMD Athlon II and Win 7 on internal hdd.
My second hdd is "external" only in the sense that it is lying on a shelf behind my computer; it is connected directly to the m/b through a shielded SATA cable - right next to my "internal" hdd and DVD drive. As far as the computer knows, the second (Linux) hdd is internal.

I installed Slack64 late last winter and managed to get it to a point where I was about to transfer the larger part of my work from Win 7 to Slack64. Then, about ten days ago, there was a sudden and unexpected power outage while I was using Slack64. When the power returned and I tried to boot back to Slack64, it started out as usual, but soon slowed down and got stuck while doing something concerning my mouse. It kept trying to the same thing repeatedly but was unable to succeed.

I went back to the installation DVD, deleted all my partitions, repartitioned the same way as before, reinstalled Slackware, reconfigured and rewrote lilo.conf, and I was back in business.

Today I switched to the generic kernel (3.10.17) and made initrd-generic-3.10.17.gz. I removed the symlinks from /config, /System.map and /vmlunuz, all of which pointed to the huge.s kernel; then created new symlinks to the -generic-3.10.17 versions of these files - including a symlink to the new initrd.gz.

I edited lilo.conf and ran lilo -v, then rebooted. I asked for the boot menu, intending to select the generic Linux kernel. Instead of the lilo boot menu, I got a black screen with a white upper case L in the top, left corner, followed by a long string of ninety-nines; like so:

L 99 99 99 99 ... on and on.

Then the computer froze. I gave the three-finger salute - which did work - and tried again with the same result.

So, back to the Slackware64 install DVD. After I logged in as root, I tried to run fdisk to check the partition table: "can't open /dev/sdb, no (? a bad time to run out of memory; it meant that fdisk could not find anything on the disk.)

I ran a program in Win 7 that provides a lot of system information, and that program showed that my second hard drive had the correct partitions of the correct size.

I suppose the MBR on that drive has been corrupted. Does any kind soul have a suggestion about how I can get out of this dead end?

Thanks in advance. You've helped me before; maybe one of these days I will be able to give answers rather than just to pose questions.

paul2

That sounds like a different issue other than just a typo since you didn't even get the LILO menu.

Can you get into Windows at least? If you can, try using EasyBCD to setup Grub2 possibly, and use the BCD Chainloader.

Thank you, ReaperX7, for your reply. Yes, I can boot Win 7 normally. Since Lilo has worked well for me, I know nothing about Grub2; so, I will have to do some more learning before I try your suggestion. Knoppix suggests re-installing the MBR; but, that presupposes that I had backed up the MBR in advance, which I did not have the foresight to do

In my OP, I suggested the possibility of a typo in lilo.config, because that was the last thing I did before rebooting and then did not get the lilo menu. When I had run lilo -v as the last thing in my previous session, the only warning was the standard one about LBA32, so I was not aware of any problem there.

Next (= right now), I will run a S.M.A.R.T. test on my Linux hdd. If any problem shows up, I'll let you know.

paul2

Try googling "invalid second stage index sector (LILO)"

FYI: I ran the S.M.A.R.T. test and a separate HP diagnostic test on my Linux hdd and it passed both tests. So, there is no evidence to support the idea of a failed hdd. And yes, I know": "Absence of evidence is not evidence of absence."

paulb2

Thank you, Richard. I have a pressing job to do right now; but, as soon as I return, I'll do that. I think this is the first time I have heard of that phrase. Thanks for the tip.

paulb2

I knew I'd had the L 99 99 99 99... problem sometime in the past, but had to trawl through my notebooks (a mess/mass of random scribblings) to find it. Fixed it by booting into Slackware using the install dvd boot command:
then deleting /boot/map, and running lilo.

Hello paulb2
According to these links, seems you must reinstall lilo.

https://www.google.com/search?q=L+99...EILKggTUtYHgDw
http://en.wikipedia.org/wiki/LILO_%2...29#Error_Codes
http://wiki.linuxquestions.org/wiki/LILO_Error_Codes

Windows always record at the MBR, so I boot my machine with a floppy to load Lilo (lilo -b /dev/fd0).
I think that way is less problematic, since I have to use (and reinstall) Windows sometimes.

I'm sorry for my extended absence, fellows; I ran into an unexpected squall of busy-ness.

I'm very happy to say that I have corrected my error (yes, it was _my_ error), but I am even more embarrassed to tell you how.

First, to bring you up to date, I followed brianL's suggestion and deleted </boot/map>, reran LILO, and re-booted Slackware64 - and got the same <L 99 ...> error.

Then I used the installation DVD to re-boot again and prepared to follow Paulo2's advice to uninstall LILO and reinstall it. But first, I decided to proofread lilo.conf one more time - and there it was, my mistake staring me in the face: I had typed <root=sdb1> so often, that I typed <boot=sdb1> without thinking. I deleted the misplaced <1>, reran LILO and re-booted. Worked perfectly!

I apologize for wasting your time with my dumb-headedness and I also thank you again, because I do appreciate the time and effort that all who replied to my plea for help expended.

BTW: Paulo2, I think you have a nice username.

paulb2

Richard's post was right, before my search
I didn't know about these Lilo errors, so I have learned something new.

Our usernames are alike
Sometimes I use 'PaulBruce' as username (Up The Irons!)