[RFC] change in /proc/devices

[david parsons]

Alexander Viro wrote:
> 
> 
> 
> On Thu, 27 Jan 2000, david parsons wrote:
> 
> >    I'd prefer a postfix notation that separates out major and minor
> >    numbers, like so:
> > 
> >    hda 03 00-39
> >    hdb 03 40-79
> >    hde 21 00-39
> >    hdf 21 00-40
> >    sda 08 00-09
> 
> Erm... It was in hex ;-)

    hda b 03 00-3f
    hdb b 03 40-7f

    Happy? ;-)


> >    Because that would explicitly show the major and minors, instead
> >    of having to rely on hidden magic to partition the majorspace.
> 
> Argh... I'ld _really_ like to get rid of major/minor stuff wherever
> possible. Besides, says who that we will not get a driver for terabyte
> hardware raid where one can expect more than 256 partitions?

     Nothing.  That's the advantage of separating out the major and
     minor numbers -- if you explicitly list the major and minor
     numbers, userland apps don't have to know how wide a major is
     to do something with it.

     0300[40] means that I have to know that a major is the first 8 bits.
     03 00-3f means that I need to know the major is the first field, and
     that the minors run from 0 to 3f.   This is MUCH better for userland,
     because that's one less stupid detail that apps have to remember.

> >    And, actually, while I'm dreaming, I'd love to have it be:

> >    device type major minors
> >    hda    b    03    00-39
> >    hdb    b    03    40-79
> >    hde    b    21    00-39
> >    hdf    b    21    00-40
> >    sda    b    08    00-09

> Maybe, but I'ld rather keep character ones away from that. If we will
> start touching them (and you will be the first to cry bloody murder) I'ld
> prefer to keep them in a separate file.

       They're not in a separate file right now, and putting them into
       the same file reduces the rooting around.

> >                  ____
> >    david parsons \bi/ It's not a devfs, but this is a format I can do
> >                   \/  something with either in userspace or in a lkm.
> 
> In lkm you will be able to do much better - another question being what
> would you do with the results...

        Need you ask?  If I wrote a lkm, I'd have it export this
        information as a filesystem so I could mount it on /dev like G-d
        himself intended, and then I could maybe wedge a 2.4 kernel
	onto a Mastodon boot floppy [*].

> 	ObBletch: All this stuff bears some uncanny resemblance to
> IP-space-dries-up fights. I just hope that we will not end up with an
> analog of NetSol<spit> on hands...

        HPA as SAIC and Linus as the Department of Defense?

        The mind, it boggles.

                      ____
	david parsons \bi/ [* Anybody know of a good way to trim 300k from a
                       \/                                       2.3.x zImage? ]

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo en vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/