From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752627AbdCAQox (ORCPT ); Wed, 1 Mar 2017 11:44:53 -0500 Received: from merlin.infradead.org ([205.233.59.134]:43378 "EHLO merlin.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752311AbdCAQou (ORCPT ); Wed, 1 Mar 2017 11:44:50 -0500 Date: Wed, 1 Mar 2017 17:03:52 +0100 From: Peter Zijlstra To: Steven Rostedt Cc: Pavan Kondeti , LKML , Ingo Molnar , Andrew Morton Subject: Re: [PATCH] sched: Optimize pick_next_task for idle_sched_class too Message-ID: <20170301160352.GO6515@twins.programming.kicks-ass.net> References: <20170119101703.2abeaeb6@gandalf.local.home> <20170119174408.GN6485@twins.programming.kicks-ass.net> <20170223135458.GC6515@twins.programming.kicks-ass.net> <20170301105303.00773560@gandalf.local.home> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20170301105303.00773560@gandalf.local.home> User-Agent: Mutt/1.5.23.1 (2014-03-12) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, Mar 01, 2017 at 10:53:03AM -0500, Steven Rostedt wrote: > Peter, do we have a solution for this yet? Are you going to add the one > with the linker magic? I queued the below earlier today. --- Subject: sched: Fix pick_next_task() for RT,DL From: Peter Zijlstra Date: Wed Mar 1 10:51:47 CET 2017 Pavan noticed that commit 49ee576809d8 ("sched/core: Optimize pick_next_task() for idle_sched_class") broke RT,DL balancing by robbing them of the opportinty to do new-'idle' balancing when their last runnable task (on that runqueue) goes away. Cc: Steven Rostedt Reported-by: Pavan Kondeti Fixes: 49ee576809d8 ("sched/core: Optimize pick_next_task() for idle_sched_class") Signed-off-by: Peter Zijlstra (Intel) --- --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -3273,10 +3273,15 @@ pick_next_task(struct rq *rq, struct tas struct task_struct *p; /* - * Optimization: we know that if all tasks are in - * the fair class we can call that function directly: + * Optimization: we know that if all tasks are in the fair class we can + * call that function directly, but only if the @pref task wasn't of a + * higher scheduling class, because otherwise those loose the + * opportunity to pull in more work from other CPUs. */ - if (likely(rq->nr_running == rq->cfs.h_nr_running)) { + if (likely((prev->sched_class == &idle_sched_class || + prev->sched_class == &fair_sched_class) && + rq->nr_running == rq->cfs.h_nr_running)) { + p = fair_sched_class.pick_next_task(rq, prev, rf); if (unlikely(p == RETRY_TASK)) goto again;