Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S264030AbTH1N6m (ORCPT ); Thu, 28 Aug 2003 09:58:42 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S264077AbTH1N6l (ORCPT ); Thu, 28 Aug 2003 09:58:41 -0400 Received: from dyn-ctb-210-9-246-251.webone.com.au ([210.9.246.251]:29192 "EHLO chimp.local.net") by vger.kernel.org with ESMTP id S264030AbTH1N6h (ORCPT ); Thu, 28 Aug 2003 09:58:37 -0400 Message-ID: <3F4E0A85.1040208@cyberone.com.au> Date: Thu, 28 Aug 2003 23:58:29 +1000 From: Nick Piggin User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4) Gecko/20030714 Debian/1.4-2 X-Accept-Language: en MIME-Version: 1.0 To: Guillaume Chazarain CC: LKML Subject: Re: [PATCH]O18.1int References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1642 Lines: 54 Guillaume Chazarain wrote: >28/08/03 14:34:15, Nick Piggin wrote: > >>Guillaume Chazarain wrote: >> >> >>>Hi Con (and linux-kernel), >>> >>>I noticed a regression wrt 2.6.0-test4 and 2.4.22 with this >>>big context-switcher: >>> >>> >>Hi Guillaume, >>If you get the time, would you be able to try my patch? Thanks. >> > >Here are the results for Nick's v8: > >top(1): > > 639 g 30 0 1336 260 1308 R 51.2 0.1 0:03.80 a.out > 638 g 22 0 1336 260 1308 S 47.3 0.1 0:03.39 a.out > >User time (seconds): 0.57 >System time (seconds): 2.72 >Elapsed (wall clock) time (h:mm:ss or m:ss): 0:06.85 >Minor (reclaiming a frame) page faults: 17 > > Thanks Guillaume, so not very good. Its interesting that there can be such a big difference in performance, but its a very simple app so makes a good test for the specific regression. In both Con's and my patches, the reader gets a bit more CPU. This might be due to it preempting the writer more often on wakeups, which would lead to more scheduling per work done and a regression. If this is the case, I'm not sure the behaviour is too undesirable though: its often very important for woken processes to be run quickly. Its not clear that this workload is something we would want to optimize for. Assuming the problem is what I guess. I will take a look into it further when I get time tomorrow. - 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/