From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1759442AbXJKWNz (ORCPT ); Thu, 11 Oct 2007 18:13:55 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1756453AbXJKWNQ (ORCPT ); Thu, 11 Oct 2007 18:13:16 -0400 Received: from 75-130-111-13.dhcp.oxfr.ma.charter.com ([75.130.111.13]:45347 "EHLO novell1.haskins.net" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1756232AbXJKWNP (ORCPT ); Thu, 11 Oct 2007 18:13:15 -0400 From: Gregory Haskins Subject: [PATCH 2/7] RT: Peter Zijlstra's suggested improvements to rt-push patch To: Steven Rostedt , Peter Zijlstra Cc: linux-rt-users , LKML , Gregory Haskins Date: Thu, 11 Oct 2007 17:59:53 -0400 Message-ID: <20071011215953.27097.46608.stgit@novell1.haskins.net> In-Reply-To: <20071011213916.27097.42597.stgit@novell1.haskins.net> References: <20071011213916.27097.42597.stgit@novell1.haskins.net> User-Agent: StGIT/0.12.1 MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org This is my own interpretation of Peter's recommended changes Steven's push-rt patch. Just to be clear, Peter does not endorse this patch unless he himself specifically says so ;). Signed-off-by: Gregory Haskins --- kernel/sched.c | 12 ++++++------ 1 files changed, 6 insertions(+), 6 deletions(-) diff --git a/kernel/sched.c b/kernel/sched.c index 0f0af6d..24b38b4 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -1499,7 +1499,7 @@ static int push_rt_task(struct rq *this_rq) int dst_cpu = -1; int ret = 0; - BUG_ON(!spin_is_locked(&this_rq->lock)); + assert_spin_locked(&this_rq->lock); next_task = rt_next_highest_task(this_rq); if (!next_task) @@ -1553,7 +1553,8 @@ static int push_rt_task(struct rq *this_rq) set_task_cpu(next_task, dst_cpu); activate_task(lowest_rq, next_task, 0); - set_tsk_need_resched(lowest_rq->curr); + resched_task(lowest_rq->curr); + schedstat_inc(this_rq, rto_schedule); spin_unlock(&lowest_rq->lock); ret = 1; @@ -2290,10 +2291,9 @@ static inline void finish_task_switch(struct rq *rq, struct task_struct *prev) * then kick other CPUs, they might run it: */ rq->curr_prio = current->prio; - if (unlikely(rt_task(current) && push_rt_task(rq))) { - schedstat_inc(rq, rto_schedule); - smp_send_reschedule_allbutself_cpumask(current->cpus_allowed); - } + if (unlikely(rq->rt_nr_running > 1)) + push_rt_task(rq); + #endif prev_state = prev->state; _finish_arch_switch(prev);