On Sat, Feb 19, 2000 at 08:55:02PM -0800, Lars Kellogg-Stedman wrote:
> Or, to elaborate:
>
> Devfs provides several ways of managing your device name space, and
> there are alternatives to /dev/scsi/host0/.../lun0/part3.
>
> (1) The /dev/discs hiearchy contains symlinks to each of your disk
> devices as /dev/discs/disc0, /dev/discs/disc1, etc. These may
> point to either scsi or ide disks (and, presumably, to others, but I
> only have scsi and ide on my system). They are named in the order the
> devices were scanned by the kernel during the boot process, and
> correspond roughly to /dev/hda or /dev/sda, etc., under the older
> system. If you're not going to be shuffling disks around, this means
> you can use this:
>
> /dev/discs/disc0/part1
>
> Instead of (e.g.),
>
> /dev/scsi/host1/bus0/target1/lun0/part1
>
> Of course, this means you lose some of the advantages that devfs gives
> you.
Correct.
> (2) devfsd can do a number of nifty things. One of those things is
> automatically creating device aliases when new device nodes are added
> to devfs. Using the MKNEWCOMPAT option, for instance, scsi disks might
> look like:
>
> /dev/sd/c1b0t1u0
> /dev/sd/c1b0t1u0p7
Ok, these are the 'old-style' devfs names, but I think my idea was better,
cause all disks were in /dev/dsk/.
> [...]
>
> I think devfs is a wonderful thing, and I've been using it for a year
> or so, but I agree that the documentation could use some help. I've
> already submitted some man page patches to Mr. Gooch; while they
> haven't yet been acknowledged I hope that they do make it in and help
> make things clearer.
I use devfs on some machines quite along time now, too and I think
it's very usefull.
The only think I don't like is the new naming scheme for the mass storage
devices.
Christoph
-- Always remember that you are unique. Just like everyone else.- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Wed Feb 23 2000 - 21:00:25 EST