On Wed, Aug 31, 2005 at 04:14:58PM -0600, dann frazier wrote: | On Wed, 2005-08-31 at 16:39 -0500, Greg Edwards wrote: | > On Wed, Aug 31, 2005 at 03:30:09PM -0600, dann frazier wrote: | > | On Wed, 2005-08-31 at 09:14 -0500, Greg Edwards wrote: | > | > I'll try to dig into this today and see if I can isolate what's going | > | > on. | > | | > | Great, thanks. I've been given access to an Altix box as well (thanks | > | Ian), so I might be able to help. | > | > console output is stopping when busybox init is run. I added a set -x to | > sbin/init and saw: | > | > + uname -r | > + grep ^2.2. | > + [ = ] | > + mount /proc | > + umount initrd | > + true | > + mount -t tmpfs -o size=100M tmpfs /mnt | > + : | > + umount /proc | > + ls -1 / | > + grep -v \(dev\|lost+found\|mnt\|proc\) | > + cp -a bin etc floppy init initrd lib preseed.cfg sbin sys tmp usr var | > /mnt | > + cd /mnt | > + pivot_root . initrd | > + mkdir proc dev | > + [ -x /sbin/udevstart ] | > + /lib/debian-installer/init-udev-devices | > + mount /proc | > + exec /usr/sbin/chroot . /bin/busybox init | > [... no more ...] | > | > The box is still up at this point, as I was able to enter kdb (I had | > kdb built into a kernel I had shoehorned in). Interestingly, the kdb | > output showed on the screen, but I never got the kdb prompt, so something | > is messing with the console (ttySG0 is the console on Altix). | | Yes - this is exactly what I saw w/ 2.6.8. Is this w/ the latest d-i | build w/ a 2.6.12 kernel?
I couldn't get the d-i (from svn) to build on my box, so I pulled the initrd apart from the latest daily images http://cdimage.debian.org/pub/cdimage-testing/daily/ia64/20050820/ which I think was still 2.6.8-based. Anyway, I repopulated with a 2.6.13 + LSI SAS + kdb kernel and modules and saw the same thing. | > SuSE (and Red Hat too, I believe) have relocatable turned on all the ia64 | > boxes by default. It doesn't have an adverse effect on those platforms | > that don't require it. If you need to check for Altix, you could look | > for the existence of /proc/sgi_sn. This will be present on any | > Altix/Prism system. | | Well, I'm told its a problem for 2.4 kernels. Debian still includes a | 2.4.27 (though we'll drop it before etch). Ah, forgot about 2.4.x kernels. You're right, it would probably cause problems there. I'm leaving on vacation now for a couple weeks, but I'll look at this again when I get back. Greg -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
