Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754931Ab1EQNzh (ORCPT ); Tue, 17 May 2011 09:55:37 -0400 Received: from smtp104.prem.mail.ac4.yahoo.com ([76.13.13.43]:39190 "HELO smtp104.prem.mail.ac4.yahoo.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1754697Ab1EQNzg (ORCPT ); Tue, 17 May 2011 09:55:36 -0400 X-Yahoo-SMTP: _Dag8S.swBC1p4FJKLCXbs8NQzyse1SYSgnAbY0- X-YMail-OSG: R1m6J4oVM1n8ibXH1DcgM6Z6dvgN78rkgVrdktSzgPYz7Io rI2mcqz9jduN5.aCla_gYT8QvL23hsOviU8F7yHcEvaTDT3LuBqyqBF8BunG tGEQC43m3JY6Sk9JV__sqvchWvh53_mIgyTMz4EJTFuzPfYo5oY3G_UxxWcI eHgjQjgeKog53DLPfHXgLLPHuiq0DgChZWwWs2wm.YY_bI2X4d1NQbnqXRYE MRL2py_szyRgsFAVqjU3kvujK6lKnDTcSGAv169iZKum.EOP2sj1UxkPKXzd G928f4axJDav030h9tUoIri8OopIS8YMLm9kMgBMh8yrs0DbW X-Yahoo-Newman-Property: ymail-3 Date: Tue, 17 May 2011 08:55:33 -0500 (CDT) From: Christoph Lameter X-X-Sender: cl@router.home To: Tejun Heo cc: Eric Dumazet , Shaohua Li , "linux-kernel@vger.kernel.org" , "akpm@linux-foundation.org" , "npiggin@kernel.dk" Subject: Re: [patch V3] percpu_counter: scalability works In-Reply-To: <20110517130414.GO20624@htj.dyndns.org> Message-ID: References: <1305593751.2375.69.camel@sli10-conroe> <1305608212.9466.45.camel@edumazet-laptop> <1305609768.2375.84.camel@sli10-conroe> <1305622861.2850.21.camel@edumazet-laptop> <20110517091102.GE20624@htj.dyndns.org> <1305625541.2850.29.camel@edumazet-laptop> <20110517095001.GF20624@htj.dyndns.org> <1305634807.2850.89.camel@edumazet-laptop> <20110517124528.GN20624@htj.dyndns.org> <1305637223.2850.95.camel@edumazet-laptop> <20110517130414.GO20624@htj.dyndns.org> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 947 Lines: 24 On Tue, 17 May 2011, Tejun Heo wrote: > Hello, > > On Tue, May 17, 2011 at 03:00:23PM +0200, Eric Dumazet wrote: > > Absolutely not. Its slow path yes, but not super-slow. > > I was speaking in relative terms. We're talking about local only fast > path vs. something which hits every percpu counter likely causing > cache misses in many of them. It's bound to be multiple orders of > magnitude heavier than fast path. Well lets just adopt the system that vm statistics use. Bound the error by time and batch and allow the user to change the batch if more accuracy is desired. The _sum function is optional and should it should be explained that the result *could* be better but dont count on it. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/