From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932142AbWCHTfu (ORCPT ); Wed, 8 Mar 2006 14:35:50 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932126AbWCHTfu (ORCPT ); Wed, 8 Mar 2006 14:35:50 -0500 Received: from mail.gmx.net ([213.165.64.20]:10376 "HELO mail.gmx.net") by vger.kernel.org with SMTP id S932159AbWCHTft (ORCPT ); Wed, 8 Mar 2006 14:35:49 -0500 X-Authenticated: #30760622 Subject: dynsched bug From: Gunter Fritz To: linux-kernel@vger.kernel.org Content-Type: text/plain Date: Wed, 08 Mar 2006 21:14:19 +0100 Message-Id: <1141848859.4446.12.camel@linux.site> Mime-Version: 1.0 X-Mailer: Evolution 2.0.1 Content-Transfer-Encoding: 7bit X-Y-GMX-Trusted: 0 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org hey there is a bug in the dynsched patch. (http://sourceforge.net/projects/dynsched) the recalculation of the nice prio must be done before the new scheduler activate the running tasks. if you had download it, change scheduler_switch in sched.c like following: //add the running processes to the new rq for_each_process(p){ nice = sched_drvp->prio_to_nice(p); p->static_prio = data->new_scheduler->nice_to_prio(nice); p->prio = p->static_prio; if(p->state == TASK_RUNNING){ data->new_scheduler->activate_task(p,rq, 1); count_rq_processes++; } }