Slackware 12 boot error revisited
Hi,
Slackware 12.0 'Invalid module format' post I had a few problems with module symbols and 'Invalid module format'. I had blacklisted the symbol specific modules kernel errors. I attempted to blacklist the 'sunrpc.ko & lockd.ko' modules to no avail. So I decided to re-compile. A good practice, I kept my known working kernel & specifics intact in my 'lilo.conf'. I modified the '/usr/src/linux/.config' to suit my needs. I then modified the '/usr/src/linux/Makefile' > 'EXTRAVERSION = .51-smp' so as to create a unique compile. I then do a compile. Do the setup for the new kernel image. I rid myself of the problems, so it seems. I do most of my maintenance remotely (ssh), I went to the server room and decided to look at the new box at the console. Some of my devices were not recognized. Simple omission(s). At least I think so. So I reboot to start the old kernel, vmlinuz-generic-smp-2.6.21.5-smp. Well I got a kernel panic. No biggy! Just reboot the new one. More devices dropped, /dev/sda2. Created the initrd with; Code:
#mkinitrd -c -k 2.6.21.5-smp -m ext3 -f ext3 -r /dev/hdb3 Code:
Do a 'lilo -v -t -b /dev/hda' to test, no errors. The do a 'lilo -v -b /dev/hda3', writes 'MBR' I then went to the /etc/modprobe.d/blacklist an removed all the before added modules via commenting out; Code:
excerpt /etc/modprobe.d/blacklist; The system is fully functional at this moment. 'rworkman', any ideas? The module dependency's were updated. But what would cause the symbols not to invoke the kernel use error message since nothing was changed with the original modules and kernel. The system is back to the original known state, at least I think so. I revert to the original except for the 'mkinitrd' and lilo.conf update. Any help, suggestions? edit: correct some grammar & spell errors |
That's exactly the behavior that I would expect. :-)
To clarify, and maybe I wasn't clear originally, those invalid symbols and -EEXIST errors should only occur if you're using the huge{,smp} kernels due to the fact that the affected modules are actually built into those kernels. On the generic{,smp} kernels, the modules should (and do, according to your latest round) load without problems, as that's how they're intended to work. In short, if you were not using one of the huge* kernels when you had the previous errors, I have no idea - that's unexplainable in my mind :/ |
Hi,
Thanks for the response. I was using the smp kernels at all times. I really don't see why this occurred myself. I'm going to go through another install on this very machine to see if the problem comes up again. I wish to find out the problem and not pass on this one. |
All times are GMT -5. The time now is 09:38 AM. |