Date: Sat, 26 Oct 2002 06:59:09 +1000 (EST) From: Bruce Evans <bde@zeta.org.au> To: Julian Elischer <julian@elischer.org> Cc: Mark Valentine <mark@thuvia.demon.co.uk>, Poul-Henning Kamp <phk@critter.freebsd.dk>, "M. Warner Losh" <imp@bsdimp.com>, <freebsd-arch@FreeBSD.ORG> Subject: Re: cvs commit: src/lib/libdisk Makefile chunk.c write_alpha_disk.c write_i386_disk.c write_pc98_disk.c Message-ID: <20021026064258.W5345-100000@gamplex.bde.org> In-Reply-To: <Pine.BSF.4.21.0210251230340.7147-100000@InterJet.elischer.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, 25 Oct 2002, Julian Elischer wrote: > On Fri, 25 Oct 2002, Mark Valentine wrote: > > > The GEOM naming scheme therefore removes my ability to specify the partition > > in the most natural way for this platform. > I dispute that. > The install code has been using ad0s1a for about 5 years I think. > Very few systems have ad0a in /etc/fstab as we specifically have been > telling people to not do that for ages.. I haven't been using the install code for about 10 years now :-). > Anyway I had a machine where the -current root partition was ad1s4e. > How does the old scheme help me? ad1s3 was also a BSD slice (FBSD3.x) Run fdisk to temporarily change ad1s3 to non-BSD. ad0[a-h] in /etc/fstab for the s4 slice then starts referring to that slice. I have used this to boot backups of partitions (cp /dev/ad0s3 /dev/ad0s4; this invalidates any s3 numbers in the copy of /etc/fstab on the s4 slice, so booting the s4 partition later doesn't work if there are any such numbers, but everything works perfectlyf ad[0-h] is used. At least with my versions of the boot blocks. The switch to using sN in fstab started with changing the boot blocks to pass sN in an incompatible way; I had already fixed the boot blocks in a compatible way and didn't like this or most subsequent changes in this area so I didn't use them. > One scheme I played with was: > /dev/ad0/whole > /dev/ad0/s1 > /dev/ad0/s1/a > /dev/ad0/bsd0 --> s1 > /dev/ad0/s1/whole > /dev/ad0/s2 > /dev/ad0/s2/whole > /dev/ad0/s3 > /dev/ad0/s3/whole > /dev/ad0/s4 > /dev/ad0/s4/whole > > so that /dev/ad0/bsd0/a would always be the root > but I don't know if the new devfs can do that.. > > This had a lot of advantages but blew POLA right out the window :-) Encoding the disk structure in the directory tree would be another mistake. Fotrunately this mistake has already been made, so we know not to repeat it. Bruce To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-arch" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20021026064258.W5345-100000>