Subject: Re: cdev vs. bdev (32 bit dev_t)
To: None <tech-kern@NetBSD.ORG>
From: Todd Vierling <tv@NetBSD.ORG>
List: tech-kern
Date: 01/14/1998 08:50:50
On Tue, 13 Jan 1998, Paul A Vixie wrote:

: > I think that, as long as we really only have one layer of abstraction in our
: > devsw, we really only need two parts (major and minor).  Option flags will
: > probabl;y vary in number and use from device to device, and I think they
: > should just be encoded into the minor number as bits.  ("It's just a little
: > bit of binary arithmetic...."  :)
: 
: not to "ls -l" it's not.

Hm, then I have two options, both of which sound good, but ONLY in the cases
of mknod(8) and ls(1)--everywhere else would still use major() and minor():

- hexadecimal representation of minor numbers in a `ls -l' output;
- split minor into 12/8 bits to keep them alinged on hex digits (this
  would, incidentally, let us see 16 bit devices at work--with major 0, the
  minor and subminor would be the major and minor of a 16 bit device!).

I'm favoring the latter now that I see a useful split point.

=====
===== Todd Vierling (Personal tv@pobox.com) =====
== "There's a myth that there is a scarcity of justice to go around, so
== that if we extend justice to 'those people,' it will somehow erode the
== quality of justice everyone else receives."  -- Maria Price