From owner-cvs-all@FreeBSD.ORG Thu Oct 6 12:15:56 2005 Return-Path: X-Original-To: cvs-all@FreeBSD.org Delivered-To: cvs-all@FreeBSD.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id B4C6F16A420 for ; Thu, 6 Oct 2005 12:15:56 +0000 (GMT) (envelope-from silby@silby.com) Received: from relay.pair.com (relay00.pair.com [209.68.5.9]) by mx1.FreeBSD.org (Postfix) with SMTP id 9685543D49 for ; Thu, 6 Oct 2005 12:15:55 +0000 (GMT) (envelope-from silby@silby.com) Received: (qmail 48963 invoked from network); 6 Oct 2005 12:15:54 -0000 Received: from unknown (HELO localhost) (unknown) by unknown with SMTP; 6 Oct 2005 12:15:54 -0000 X-pair-Authenticated: 209.68.2.70 Date: Thu, 6 Oct 2005 07:15:52 -0500 (CDT) From: Mike Silbersack To: Robert Watson In-Reply-To: <20051006124123.Y87201@fledge.watson.org> Message-ID: <20051006071440.F40058@odysseus.silby.com> References: <200510061028.j96ASVoL031977@repoman.freebsd.org> <20051006063636.S29769@odysseus.silby.com> <20051006124123.Y87201@fledge.watson.org> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Cc: cvs-src@FreeBSD.org, src-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/tools/tools/netrate/httpd httpd.c X-BeenThere: cvs-all@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: CVS commit messages for the entire tree List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 06 Oct 2005 12:15:56 -0000 On Thu, 6 Oct 2005, Robert Watson wrote: >> You can't use this to compare to 4.x then, FWIW. 4.x's sendfile always >> puts the headers in a separate packet, so in a large percentage of cases >> it's noticeably less efficient, network-traffic wise. > > It depends what you're trying to benchmark. If the goal is to illustrate the > performance changes as a result of on-going development, it's legitimate to > say that the changes in sendfile() are simply part of that process. I.e., > it's not cheating to have sendfile() improvements count towards overall > performance when evaluating overall performance. > > Robert N M Watson True, just don't start thinking that locking improvements are why you're seeing differences here. :) Mike "Silby" Silbersack