Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751653AbdFZOoE (ORCPT ); Mon, 26 Jun 2017 10:44:04 -0400 Received: from merlin.infradead.org ([205.233.59.134]:58126 "EHLO merlin.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751745AbdFZOnz (ORCPT ); Mon, 26 Jun 2017 10:43:55 -0400 Date: Mon, 26 Jun 2017 16:43:38 +0200 From: Peter Zijlstra To: riel@redhat.com Cc: linux-kernel@vger.kernel.org, jhladky@redhat.com, mingo@kernel.org, mgorman@suse.de Subject: Re: [PATCH 3/4] sched,numa: implement numa node level wake_affine Message-ID: <20170626144338.GA4941@worktop> References: <20170623165530.22514-1-riel@redhat.com> <20170623165530.22514-4-riel@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20170623165530.22514-4-riel@redhat.com> User-Agent: Mutt/1.5.22.1 (2013-10-16) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 492 Lines: 12 On Fri, Jun 23, 2017 at 12:55:29PM -0400, riel@redhat.com wrote: > From: Rik van Riel > > Since select_idle_sibling can place a task anywhere on a socket, > comparing loads between individual CPU cores makes no real sense > for deciding whether to do an affine wakeup across sockets, either. > > Instead, compare the load between the sockets in a similar way the > load balancer and the numa balancing code do. This seems to assume LLC == NUMA, which isn't strictly so.