From owner-freebsd-current  Sun Feb  2 08:35:22 1997
Return-Path: <owner-current>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.5/8.8.5) id IAA24482
          for current-outgoing; Sun, 2 Feb 1997 08:35:22 -0800 (PST)
Received: from chai.plexuscom.com (chai.plexuscom.com [207.87.46.100])
          by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id IAA24475;
          Sun, 2 Feb 1997 08:35:19 -0800 (PST)
Received: from chai.plexuscom.com (localhost [127.0.0.1]) by chai.plexuscom.com (8.8.4/8.6.12) with ESMTP id LAA21764; Sun, 2 Feb 1997 11:35:43 -0500 (EST)
Message-Id: <199702021635.LAA21764@chai.plexuscom.com>
To: phk@freebsd.org
Cc: current@freebsd.org
Subject: Re: device driver open semantics... 
In-reply-to: Your message of "Sat, 01 Feb 1997 13:31:40 +0100."
             <3050.854800300@critter.dk.tfs.com> 
Date: Sun, 02 Feb 1997 11:35:43 -0500
From: Bakul Shah <bakul@torrentnet.com>
Sender: owner-current@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

> Comments ?
> 
> I realize that changing this behaviour in general would
> probably surprise most if not all of our device-drivers,
> So we're probably talking about a per-driver flag or possibly
> a different open/close entry point in the [cb]devsw structure.
> 
> Preferences ? 
> 
> Objections ?

Most device drivers depend on a close() being the final close.  Any
change in this semantics must provide a *significant* benefit to all
-- not just fix a few esoteric bugs.  Adding more entry points would
also further complicate the interface.

Perhaps the bugs Bruce mentioned (+ things that make you
dissatisfied with the way things work now) can be handled by passing
some more state between the two layers?  May be a callback function?

In any case, cases where the present *observed* behavior does
not cause faulty, unintended or inconsistent operation should remain
_invariant_ under any changes.  IMHO, of course!

-- bakul