From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1758721Ab1FBLw7 (ORCPT ); Thu, 2 Jun 2011 07:52:59 -0400 Received: from casper.infradead.org ([85.118.1.10]:50795 "EHLO casper.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755712Ab1FBLw6 (ORCPT ); Thu, 2 Jun 2011 07:52:58 -0400 Subject: Re: [PATCH] sched: fix conflict of schedule domain balance in RT scheduling From: Peter Zijlstra To: Hillf Danton Cc: LKML , Steven Rostedt , Mike Galbraith , Yong Zhang , Ingo Molnar In-Reply-To: References: Content-Type: text/plain; charset="UTF-8" Date: Thu, 02 Jun 2011 13:56:29 +0200 Message-ID: <1307015789.2497.640.camel@laptop> Mime-Version: 1.0 X-Mailer: Evolution 2.30.3 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, 2011-06-01 at 21:58 +0800, Hillf Danton wrote: > SD_BALANCE_WAKE and SD_WAKE_AFFINE are defined to be mutually > exclusive Uhm, no they're not.. both affect placement of a woken task but they're complementary. WAKE_AFFINE is a check to see if it makes sense to run the woken task on the same cpu as the wakee, BALANCE_WAKE does a full load-balance pass, in case its combined with WAKE_AFFINE it does so in case that test is negative.