From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1764894AbXGNUZY (ORCPT ); Sat, 14 Jul 2007 16:25:24 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1760256AbXGNUZM (ORCPT ); Sat, 14 Jul 2007 16:25:12 -0400 Received: from tomts10.bellnexxia.net ([209.226.175.54]:50145 "EHLO tomts10-srv.bellnexxia.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1760673AbXGNUZL (ORCPT ); Sat, 14 Jul 2007 16:25:11 -0400 Date: Sat, 14 Jul 2007 16:25:09 -0400 From: Mathieu Desnoyers To: Peter Zijlstra Cc: Oleg Nesterov , linux-kernel@vger.kernel.org, Ingo Molnar , Steven Rostedt Subject: Re: [RFC] Thread Migration Preemption - v4 Message-ID: <20070714202509.GR6975@Krystal> References: <20070706060257.GA188@tv-sign.ru> <20070706142339.GA32754@Krystal> <20070706145634.GA198@tv-sign.ru> <20070711044915.GA4025@Krystal> <20070711163648.GA232@tv-sign.ru> <20070714184237.GI6975@Krystal> <1184440451.5284.75.camel@lappy> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline In-Reply-To: <1184440451.5284.75.camel@lappy> X-Editor: vi X-Info: http://krystal.dyndns.org:8080 X-Operating-System: Linux/2.6.21.3-grsec (i686) X-Uptime: 16:24:37 up 7 days, 10:29, 2 users, load average: 0.15, 0.24, 0.26 User-Agent: Mutt/1.5.13 (2006-08-11) Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org * Peter Zijlstra (a.p.zijlstra@chello.nl) wrote: > Would have been nice if you'd CC'ed me :-/ > Sorry > On Sat, 2007-07-14 at 14:42 -0400, Mathieu Desnoyers wrote: > > > @@ -4891,10 +4948,42 @@ static int migration_thread(void *data) > > list_del_init(head->next); > > > > spin_unlock(&rq->lock); > > - __migrate_task(req->task, cpu, req->dest_cpu); > > + migrated = __migrate_task(req->task, cpu, req->dest_cpu); > > local_irq_enable(); > > - > > - complete(&req->done); > > + if (!migrated) { > > + /* > > + * If the process has not been migrated, let it run > > + * until it reaches a migration_check() so it can > > + * wake us up. > > + */ > > + spin_lock_irq(&rq->lock); > > + head = &rq->migration_queue; > > + list_add(&req->list, head); > > + if (req->task->se.on_rq > > + || !task_migrate_count(req->task)) { > > + /* > > + * The process is on the runqueue, it could > > + * exit its critical section at any moment, > > + * don't race with it and retry actively. > > + * Also, if the thread is not on the runqueue > > + * and has a zero migration count > > + * (__migrate_task failed because cpus allowed > > + * changed), just retry. > > + */ > > + spin_unlock_irq(&rq->lock); > > + continue; > > + } > > + set_tsk_thread_flag(req->task, TIF_NEED_MIGRATE); > > + set_current_state(TASK_INTERRUPTIBLE); > > + spin_unlock_irq(&rq->lock); > > + /* > > + * Wait for the process currently in its critical > > + * section. > > + */ > > + wake_up_process(req->task); > > + schedule(); > > + } else > > + complete(&req->done); > > } > > __set_current_state(TASK_RUNNING); > > return 0; > > I'm not really liking this, this way we hold up all other migration > requests. > True. You approach of delayed migration helps on this. -- Mathieu Desnoyers Computer Engineering Ph.D. Student, Ecole Polytechnique de Montreal OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F BA06 3F25 A8FE 3BAE 9A68