From owner-cvs-src  Mon Mar 10 10:47:35 2003
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 2B2B337B401; Mon, 10 Mar 2003 10:47:34 -0800 (PST)
Received: from ebb.errno.com (ebb.errno.com [66.127.85.87])
	by mx1.FreeBSD.org (Postfix) with ESMTP
	id 4518243F93; Mon, 10 Mar 2003 10:47:33 -0800 (PST)
	(envelope-from sam@errno.com)
Received: from melange (melange.errno.com [66.127.85.82])
	(authenticated bits=0)
	by ebb.errno.com (8.12.8/8.12.6) with ESMTP id h2AIlUeV077442
	(version=TLSv1/SSLv3 cipher=RC4-MD5 bits=128 verify=NO);
	Mon, 10 Mar 2003 10:47:32 -0800 (PST)
	(envelope-from sam@errno.com)
Message-ID: <15bf01c2e735$825f1eb0$52557f42@errno.com>
From: "Sam Leffler" <sam@errno.com>
To: "John Baldwin" <jhb@FreeBSD.org>,
	"Poul-Henning Kamp" <phk@phk.freebsd.dk>
Cc: <src-committers@FreeBSD.org>, <cvs-src@FreeBSD.org>,
	"Nate Lawson" <nate@root.org>
References: <3039.1047321059@critter.freebsd.dk>
Subject: Re: cvs commit: src/sys/sys devicestat.h 
Date: Mon, 10 Mar 2003 10:47:30 -0800
Organization: Errno Consulting
MIME-Version: 1.0
Content-Type: text/plain;
	charset="Windows-1252"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.50.4807.1700
X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4807.1700
Sender: owner-cvs-src@FreeBSD.ORG
Precedence: bulk
List-ID: <cvs-src.FreeBSD.ORG>
List-Archive: <http://docs.freebsd.org/mail/> (Web Archive)
List-Help: <mailto:majordomo@FreeBSD.ORG?subject=help> (List Instructions)
List-Subscribe: <mailto:majordomo@FreeBSD.ORG?subject=subscribe%20cvs-src>
List-Unsubscribe: <mailto:majordomo@FreeBSD.ORG?subject=unsubscribe%20cvs-src>
X-Loop: FreeBSD.ORG

> This is disk statistics, it's not really worth that effort.  The
> data in the structure will not get corrupted, bue one of our
> readings may.  So what...

All the more reason why I question the complexity.  Are there really going
to be that many clients polling for statistics that using a system call is
bad?

FWIW I worked on a real-time event/statistics package for IRIX.  There were
per-cpu queues of events mapped into user-space.  Lots of atomic ops and
such to get a lock-free mechanism with guaranteed consistency of results
(e.g. this was supposedly used to monitor real-time data collection sites
like physical plants).  Aggregating the per-cpu streams was a pain.

For something like the above I could understand the added complexity.  For
disk stats for top, etc, no.

    Sam


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe cvs-src" in the body of the message