Hehe.. I can see the scrunching on your face now.. It was actually simple to do, it should fix the problem with x86_64 bootstraps. I slimmed the livecd-mkinitrd.sh (mkinitrd) script WAY down in the process.
This doesn't mean much but I think we can probably do as you suggest and use /sbin/mkinitrd on the host. I'll experiment with this soon after I get done with a current tester on x86_64.
J. Hartline
Jeremy Katz wrote:
On Sun, 2006-08-27 at 17:57 -0500, Jasper O'neal Hartline wrote:
I've been working on an initramfs prototype for Kadischi and it seems I've got it under wraps. Could all testers of x86 and x86_64 Fedora Core repository users test these two files?
Ugh. Why not use the actual stock mkinitrd pieces with just an appropriate livecd-mkrootdev as I've outlined a few times previously and implemented in my hg repo... :/
There's a lot of logic in mkinitrd to handle a lot of cases and trying to replicate all of that in a live CD specific environment is just going to lead to problems that are very difficult to debug