Article View: aus.computers.linux
Article #59979Re: 1.2.11 kernel crashes with /dev/fd0
From: Brett Gersekowski
Date: Sun, 30 Jul 1995 00:00
Date: Sun, 30 Jul 1995 00:00
120 lines
4753 bytes
4753 bytes
Rachel Polanskis (grove@zeta.org.au) wrote: : Hello, : I followed various pieces of advice and after RTFM, I decided to upgrade my : kernel from v 1.1.59 to 1.2.11 : Ok, so the kernel compiled without error, and I managed to free up space, : with the removal of redundant drivers. : Now, I did 'make zlilo', and rebooted, it works fine, : BUT: : Now I am having all kinds of weird problems. : First up, when booting, I see that Multi mode on my HD's are now Disabled. : Why? : WORSE!!! : I went to make a rescue disk, using the Bootkit package from sunsite, : It GPF'd and I had to restart. : Tried it again, same problem. (I always try to repeat an error) : OK, so I tried to do an 'fdformat /dev/fd0' Does it also fail if you explicitly specify /dev/fd0H1440?? I also run 1.2.11. fdformat worked fine. mind you - the drive has been mounted and unmounted during this session. I will try again later after reboot. : SO, restart and run setup - trying to make a boot disk (at least) : through the Slackware menu. : CRASH!!! : "Virtual Memory exhausted" : in the output of the GPF junk was: : "swap: weirdness" : Regular access to fd0 seems OK, I can mount and unmount. : It was fine before with 1.1.59. : (everything was good with 1.1.59) : 2x16MB swap partitions : 1 64MB swapfile Why so much swap?? I heard somewhere that linux screws up with huge amounts of swap - from memory these were >128M swap partitions, though. Try rebooting without the swap - or with only 1x16M partition. Granted - if it worked under 1.1.59 - you'd expect it still should now. WHEN YOU UPGRADE YOUR KERNEL... ALWAYS make sure you can still reboot your computer without the new kernel. I usually keep two kernel images on my root partition, namely /zImage and /zImage.old I also have two sections in my lilo.conf - labelled "linux" and "oldlinux" pointing to the two previous files respectively. This way - you know you have a working kernel. First - if you can, re-install 1.1.59 kernel which worked for you. If you can't boot off anything stably - try using the disk you originally booted from in order to install Linux. Mount your HDD on /mnt and do some surgery there. Basically, you'll need to look in /mnt/etc to check your lilo.conf out and then you'll need to do a /mnt/sbin/lilo -r /mnt. Hopefully, now you'll have a bootable HDD. Forget zlilo for now. Perhaps things will work better with a manual config of lilo. Re-build the kernel (1.2.11) but use make zImage instead of zlilo. (still check config and make dependancies) copy the kernel image to / - MAKING SURE THAT YOU HAVE YOUR OLD KERNEL IMAGE, AS WELL do something like ... mv /zImage /zImage.old cp /usr/src/linux/arch/i386/zImage /zImage Now, edit your /etc/lilo.conf. Make sure you have two linux bootable sections - ie. after all the crap at the top - something like :- # Linux bootable partition config begins image = /zImage root = /dev/hda2 label = Linux image = /zImage.old root = /dev/hda2 label = oldLinux You might want to read through /usr/doc/lilo/README if you're not familiar with some of the lilo configuration options. Again run lilo - it'll either be /sbin/lilo or try /etc/lilo/install (NOTE: if you use "install" you HAVE to include the path, even if just ./install) Hopefully it all worked. Try rebooting with the new kernel. If the new kernel is still problematic, reboot, but use "oldlinux" at the boot prompt, which should boot your old kernel - which SHOULD work... Any problems, gimme a yell... -- ******************************************************************************* *** Brett Gersekowski *** Experience is directly- *** *** bgerseko@enterprise.powerup.com.au *** proportional to the amount *** *** e1135350@watt.lab.eesse.bee.qut.edu.au *** of equipment ruined. *** *** n1135350@student.fit.qut.edu.au *** *** *******************************************************************************
Technical Details
Message-ID:
<3vfboq$f4g@enterprise.powerup.com.au>Path:
arc1-ca.pugleaf.net!unknown!not-for-mailReferences:
<3veo1l$k2s@godzilla.zeta.org.au>Show Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: 102983,7899b7e59e9f2d9e X-Google-Attributes: gid102983,public X-Google-Thread: f9b51,7899b7e59e9f2d9e X-Google-Attributes: gidf9b51,public From: bgerseko@powerup.com.au (Brett Gersekowski) Subject: Re: 1.2.11 kernel crashes with /dev/fd0 Date: 1995/07/30 Message-ID: <3vfboq$f4g@enterprise.powerup.com.au> X-Deja-AN: 107189713 references: <3veo1l$k2s@godzilla.zeta.org.au> followup-to: aus.computers.linux,comp.os.linux.setup organization: PowerUp reply-to: bgerseko@mail.powerup.com.au newsgroups: aus.computers.linux,comp.os.linux.setup