In article <NBBBJHKIOKPKOGOEPEDPKELFEBAA.stuart@bh90210.net>, IT3 Stuart Blake Tener, USNR-R <stuart@bh90210.net> wrote: >Zygo: > > I am reading your post below with regard to making the linux crypto part of >the kernel, but I am curious how does this impact those people whom are >using devfs=mount with their kernels? If there was an in-kernel losetup, it would presumably use the raw device major/minor numbers just like the existing 'root=' kernel command-line parameter. So you'd say something like append locrypt=aes lokeysize=256 loroot=0302 lodev=7 root=0707 which would be equivalent to something like: losetup -e aes -k 256 /dev/loop7 /dev/hda2 mount /dev/hda2 /somewhere cd /somewhere pivot_root /somewhere /somewhere/else exec chroot /sbin/init Now interestingly enough, if you use devfs to do that actual losetup command, you get around the busy-device-inode problem that prevents you from dropping the init RAM disk (you get around it because the busy inode is on devfs, and devfs doesn't care about busy inodes when you umount it). Next time my laptop crashes I will have to try this. ;-) -- Zygo Blaxell (Laptop) <zblaxell@feedme.hungrycats.org> GPG = D13D 6651 F446 9787 600B AD1E CCF3 6F93 2823 44AD - Linux-crypto: cryptography in and on the Linux system Archive: http://mail.nl.linux.org/linux-crypto/
