From owner-freebsd-arch@FreeBSD.ORG Sun Jan 21 22:21:44 2007 Return-Path: X-Original-To: freebsd-arch@freebsd.org Delivered-To: freebsd-arch@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id 249EF16A405 for ; Sun, 21 Jan 2007 22:21:44 +0000 (UTC) (envelope-from rizzo@icir.org) Received: from xorpc.icir.org (xorpc.icir.org [192.150.187.68]) by mx1.freebsd.org (Postfix) with ESMTP id 0E67E13C478 for ; Sun, 21 Jan 2007 22:21:44 +0000 (UTC) (envelope-from rizzo@icir.org) Received: from xorpc.icir.org (localhost [127.0.0.1]) by xorpc.icir.org (8.12.11/8.13.6) with ESMTP id l0LMLgnn083988; Sun, 21 Jan 2007 14:21:42 -0800 (PST) (envelope-from rizzo@xorpc.icir.org) Received: (from rizzo@localhost) by xorpc.icir.org (8.12.11/8.12.3/Submit) id l0LMLf0B083987; Sun, 21 Jan 2007 14:21:41 -0800 (PST) (envelope-from rizzo) Date: Sun, 21 Jan 2007 14:21:41 -0800 From: Luigi Rizzo To: Poul-Henning Kamp Message-ID: <20070121142141.A83926@xorpc.icir.org> References: <20070121140019.A83688@xorpc.icir.org> <18141.1169417423@critter.freebsd.dk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5.1i In-Reply-To: <18141.1169417423@critter.freebsd.dk>; from phk@phk.freebsd.dk on Sun, Jan 21, 2007 at 10:10:23PM +0000 Cc: Ivan Voras , freebsd-arch@freebsd.org Subject: Re: bzero & bcopy alignment X-BeenThere: freebsd-arch@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussion related to FreeBSD architecture List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 21 Jan 2007 22:21:44 -0000 On Sun, Jan 21, 2007 at 10:10:23PM +0000, Poul-Henning Kamp wrote: > In message <20070121140019.A83688@xorpc.icir.org>, Luigi Rizzo writes: > >On Sun, Jan 21, 2007 at 10:41:09PM +0100, Ivan Voras wrote: > >> Following recent discussion on alignment of bzero() and bcopy(), I've > >> added some statistics collecting code to bzero() and bcopy() for > >> practice (on a RELENG_6 box), and here are the cumulative results for > >> argument alignment: > > > >i think these profiles depend heavily on the hardware > >and usage patterns. > > Undoubtedly, but on the other hand, a lot of them operate on > structs and other data laid out by the compiler, so the > unaligned calls are special cases. and that is why i suggested to instrument the functions to record where in the source the unaligned functions are called, so we can figure out whether it is something unavoidable or something that can be fixed. cheers luigi