From owner-cvs-src@FreeBSD.ORG  Fri Sep  5 15:15:23 2003
Return-Path: <owner-cvs-src@FreeBSD.ORG>
Delivered-To: cvs-src@freebsd.org
Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125])
	by hub.freebsd.org (Postfix) with ESMTP id 16C6A16A4C0
	for <cvs-src@FreeBSD.org>; Fri,  5 Sep 2003 15:15:23 -0700 (PDT)
Received: from rootlabs.com (root.org [67.118.192.226])
	by mx1.FreeBSD.org (Postfix) with SMTP id 98C3844001
	for <cvs-src@FreeBSD.org>; Fri,  5 Sep 2003 15:15:21 -0700 (PDT)
	(envelope-from nate@rootlabs.com)
Received: (qmail 72744 invoked by uid 1000); 5 Sep 2003 22:15:22 -0000
Date: Fri, 5 Sep 2003 15:15:22 -0700 (PDT)
From: Nate Lawson <nate@root.org>
To: "David G. Lawrence" <dg@dglawrence.com>
In-Reply-To: <20030905200502.GB61917@nexus.dglawrence.com>
Message-ID: <20030905150357.U72582@root.org>
References: <20030905182744.6630916A4E8@hub.freebsd.org>
	<20030905124936.F72174@root.org>
	<20030905200502.GB61917@nexus.dglawrence.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
cc: cvs-src@FreeBSD.org
cc: src-committers@FreeBSD.org
cc: cvs-all@FreeBSD.org
Subject: Re: cvs commit: src/sys/dev/ata ata-all.c ata-disk.c ata-disk.h
 atapi-all.c atapi-all.h
X-BeenThere: cvs-src@freebsd.org
X-Mailman-Version: 2.1.1
Precedence: list
List-Id: CVS commit messages for the src tree <cvs-src.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/listinfo/cvs-src>,
	<mailto:cvs-src-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/cvs-src>
List-Post: <mailto:cvs-src@freebsd.org>
List-Help: <mailto:cvs-src-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/cvs-src>,
	<mailto:cvs-src-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Fri, 05 Sep 2003 22:15:23 -0000

On Fri, 5 Sep 2003, David G. Lawrence wrote:
> Nate Lawson wrote:
> > Yikes, invasive.  Is there a way to do a "if (held(lock))" construct
> > instead?
>
>    What do you mean by invasive?

Changing all the callers to pass a flag whether or not they hold the lock.
I'm not experienced with what the best accepted approach to locking is.
However, it seems like locks should not require external information (i.e.
hints) or recursion generally.  I understand this is not an OS lock but an
ATA lock but it seems like the same principle would apply.  My question is
whether there is a way the call path can be changed so that the flag is
implicit or for the exception case there is a way to query the lock to see
if it's already held.  Would such an approach be feasible?

-Nate