Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S965140AbVIAOYU (ORCPT ); Thu, 1 Sep 2005 10:24:20 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S965141AbVIAOYU (ORCPT ); Thu, 1 Sep 2005 10:24:20 -0400 Received: from mail24.syd.optusnet.com.au ([211.29.133.165]:9652 "EHLO mail24.syd.optusnet.com.au") by vger.kernel.org with ESMTP id S965140AbVIAOYT (ORCPT ); Thu, 1 Sep 2005 10:24:19 -0400 From: Con Kolivas To: Anupama Chandwani Subject: Re: swappable scheduler in linux kernel Date: Fri, 2 Sep 2005 00:23:21 +1000 User-Agent: KMail/1.8.2 Cc: Ingo Molnar , Pavel Machek , linux kernel mailing list , Andrew Morton , Peter Williams , William Lee Irwin III , Alexander Nyberg , Nick Piggin , Linus Torvalds References: <689eb34705090107196f3ef558@mail.gmail.com> In-Reply-To: <689eb34705090107196f3ef558@mail.gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200509020023.22485.kernel@kolivas.org> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1497 Lines: 29 On Fri, 2 Sep 2005 00:19, Anupama Chandwani wrote: > Hello, > I am a final year coputre enginering student from Pune university,India. > Im interested in linux kernel & saw the plugsched echanism, how its been > made possible to change schedler policy at compile time. > > However, Im thinking of a scheduler that can dynamically change its policy > or soe part of code for that matter hot swap with a copletely different > scheduler at RUNTIME. something like IBM's K42. We can have scheduler modes > like realtime or interactive which can be changed at runtime. > Though its a pretty much completed problem to solve, there seem to be > little commertial need of it.. can you suggest of its application or maybe > your comments on this idea. Runtime modification has possible uses in the form of hot scheduler upgrades for systems in production for security upgrades or performance enhancements, with virtually nil downtime. This was one of my ultimate aims when pushing the original plugsched infrastructure. However both Linus and Ingo did not like the plugsched idea at all thinking it would fragment work effort put into the mainline scheduler and have lots of less good schedulers instead of one scheduler to rule them all. Cheers, Con - 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/