From owner-freebsd-hackers  Sat Oct 19 17:08:57 1996
Return-Path: owner-hackers
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id RAA22430
          for hackers-outgoing; Sat, 19 Oct 1996 17:08:57 -0700 (PDT)
Received: from darling.cs.umd.edu (10862@darling.cs.umd.edu [128.8.128.115])
          by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id RAA22422
          for <freebsd-hackers@freebsd.org>; Sat, 19 Oct 1996 17:08:55 -0700 (PDT)
Received: by darling.cs.umd.edu (8.7.6/UMIACS-0.9/04-05-88)
	id UAA26134; Sat, 19 Oct 1996 20:08:44 -0400 (EDT)
Message-Id: <199610200008.UAA26134@darling.cs.umd.edu>
To: joerg_wunsch@uriah.heep.sax.de (Joerg Wunsch)
cc: freebsd-hackers@freebsd.org (FreeBSD hackers)
Subject: Re: Conventions/Rules for adding Local ioctls 
In-reply-to: Your message of "Sun, 20 Oct 1996 00:40:33 +0200."
             <199610192240.AAA03744@uriah.heep.sax.de> 
Date: Sat, 19 Oct 1996 20:08:43 -0400
From: Rohit Dube <rohit@cs.umd.edu>
Sender: owner-hackers@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

On Sun, 20 Oct 1996 00:40:33 +0200 (MET DST) j@uriah.heep.sax.de writes:
=>As Bruce Evans wrote:
=>> Ioctls are per-major, so there is no chance of ones for the local major
=>> conflicting with ones for ttys.
=>
=>...unless your local driver actually _is_ a tty driver.  The
=>originator of the question didn't tell us.
=>
=>(I meant the generic tty ioctls, all that TIOxxx stuff.)
=>

Hi,

Sorry for not putting this on to my orignal post : I am using Major
(Character) device number 20 which is reserved for local use.

The pseudo driver (I call it /dev/cntrl0) is not a tty driver. I
am structuring it like 'bpf' I use it to control a bunch of 
other pseudo devices which sit on top of the 'de' ethernet driver.
Just like any usual network driver, these pseudo network device drivers
do not have any /dev entries.

My question applies to both /dev/cntrl0 and to the pseudo network devices.

>From what I make of the previous replies '/dev/cntrl0' is ok as it has
a unique Major device number which the kernel code switches on while
handling ioctls. For the network drivers, I guess I am going to have
to find a unique group my eliminating those already used by the
net and tty code.

I was hoping for a globally maintained file hidden somewhere which listed
at least the 'taken' groups. More comments??

--rohit.
(rohit@cs.umd.edu)