On Sat, 04 Sep 2010 15:18:40 -0400 (EDT), David Baron wrote: > > Using the Debian rescue CD, I installed the linux-image-2.6.32-5 and linux- > base from Sid. > > The install edited my fstab and lilo.conf files for me, putting in the UUID > numbers for everything except lilo.conf boot=. I left that as-is /dev/sda but > both variations, UUID and ata-ID had previously worked. > > Too bad the previous install did not fix this up for me. Might have saved a > lot of grief, huh?
Yes, that seems to be one weakness of linux-base. (I think that's the package that converts the files.) It doesn't handle the "boot" record of /etc/lilo.conf properly. I discovered that the hard way a few months ago. And that is why I document the stuff I do in my kernel-building web page. I thought about reporting a bug, but I didn't want to give the movers and shakers at Debian an excuse to pull lilo from the distribution. It isn't lilo's fault, but some people I know are looking for an excuse to pull it. > The initrd (needed most, dep failed) needed to go into "highest memory" with > scarey warnings but .... it worked. MODULES=dep usually works, but only if you don't "cross-build" your initial RAM file systems. In other words, if you're using MODULES=dep, and you're building an initial RAM file system image for the 2.6.32-5-amd64 kernel, you had better be *running* the 2.6.32-5-amd64 kernel at the time you do the build. If you're running any other kernel, there's a good chance that the right drivers won't be included. Do you have the "large-memory" option specified in /etc/lilo.conf? If not, you should. This will probably eliminate the warnings. > > Thanks for all the help. Debian folks are the best! You're welcome. I'm glad I could help. -- .''`. Stephen Powell : :' : `. `'` `- -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected] Archive: http://lists.debian.org/1554450231.612763.1283637242567.javamail.r...@md01.wow.synacor.com
