From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: with ECARTIS (v1.0.0; list xfs); Fri, 18 Apr 2008 06:34:42 -0700 (PDT) Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id m3IDYPD9005520 for ; Fri, 18 Apr 2008 06:34:28 -0700 Received: from one.firstfloor.org (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 70095158FBC2 for ; Fri, 18 Apr 2008 06:35:05 -0700 (PDT) Received: from one.firstfloor.org (one.firstfloor.org [213.235.205.2]) by cuda.sgi.com with ESMTP id spcN6v39T7bB05v5 for ; Fri, 18 Apr 2008 06:35:05 -0700 (PDT) Message-ID: <4808A383.5050907@firstfloor.org> Date: Fri, 18 Apr 2008 15:34:59 +0200 From: Andi Kleen MIME-Version: 1.0 Subject: Re: likely and unlikely was: Re: [PATCH] split xfs_ioc_xattr References: <20080319204014.GA23644@lst.de> <20080414032940.GA10579@lst.de> <20080416063712.GN108924158@sgi.com> <4805A589.7080906@sgi.com> <87ve2i5kbs.fsf@basil.nowhere.org> <4808488A.7010204@sgi.com> In-Reply-To: <4808488A.7010204@sgi.com> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: xfs-bounce@oss.sgi.com Errors-to: xfs-bounce@oss.sgi.com List-Id: xfs To: Timothy Shimmin Cc: David Chinner , Niv Sardi , Christoph Hellwig , xfs@oss.sgi.com Timothy Shimmin wrote: > Hi, > > Thanks for the explanation, Andi. > So I guess the upshot is, that it can make a difference but > in many cases (where the perf difference isn't an issue) > it is probably not worth the ugliness. > And in performance cases, it would be best to test the hypothesis > with the unlikely profiler patch > => it will be _unlikely_ we will bother ;-) > So I don't think I'll be bothering with them then unless > an issue comes up :) Ideal would be to not bother by default, but occasionally run oprofile with icache and branch misprediction profiling for macro benchmarks (significant user space code running) and if you see any icache miss/mispredict hot spots in your code add the annotations there and then double check with the unlikely profiler. -Andi