From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754257AbcGZAL5 (ORCPT ); Mon, 25 Jul 2016 20:11:57 -0400 Received: from mail-qt0-f193.google.com ([209.85.216.193]:33714 "EHLO mail-qt0-f193.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753190AbcGZALy (ORCPT ); Mon, 25 Jul 2016 20:11:54 -0400 Date: Mon, 25 Jul 2016 20:11:51 -0400 From: Tejun Heo To: Andrew Morton Cc: Dou Liyang , cl@linux.com, mika.j.penttila@gmail.com, mingo@redhat.com, rjw@rjwysocki.net, hpa@zytor.com, yasu.isimatu@gmail.com, isimatu.yasuaki@jp.fujitsu.com, kamezawa.hiroyu@jp.fujitsu.com, izumi.taku@jp.fujitsu.com, gongzhaogang@inspur.com, len.brown@intel.com, lenb@kernel.org, tglx@linutronix.de, chen.tang@easystack.cn, rafael@kernel.org, x86@kernel.org, linux-acpi@vger.kernel.org, linux-kernel@vger.kernel.org, linux-mm@kvack.org Subject: Re: [PATCH v9 0/7] Make cpuid <-> nodeid mapping persistent Message-ID: <20160726001151.GN19588@mtj.duckdns.org> References: <1469435749-19582-1-git-send-email-douly.fnst@cn.fujitsu.com> <20160725162022.e90e9c6c74a5d147e39e5945@linux-foundation.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20160725162022.e90e9c6c74a5d147e39e5945@linux-foundation.org> User-Agent: Mutt/1.6.1 (2016-04-27) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hello, Andrew. On Mon, Jul 25, 2016 at 04:20:22PM -0700, Andrew Morton wrote: > > When a pool workqueue is initialized, if its cpumask belongs to a node, its > > pool->node will be mapped to that node. And memory used by this workqueue will > > also be allocated on that node. > > Plan B is to hunt down and fix up all the workqueue structures at > hotplug-time. Has that option been evaluated? > > Your fix is x86-only and this bug presumably affects other > architectures, yes? I think a "Plan B" would fix all architectures? Yeah, that was one of the early approaches. The issue isn't limited to wq. Any memory allocation can have similar issues of underlying node association changing and we don't have any synchronization mechanism around it. It doesn't make any sense to make NUMA association dynamic when the consumer surface is vastly larger and there's nothing inherently dynamic about the association itself. Thanks. -- tejun