From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756501Ab1I3J23 (ORCPT ); Fri, 30 Sep 2011 05:28:29 -0400 Received: from merlin.infradead.org ([205.233.59.134]:37558 "EHLO merlin.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754590Ab1I3J22 convert rfc822-to-8bit (ORCPT ); Fri, 30 Sep 2011 05:28:28 -0400 Subject: Re: [PATCH] sched/kthread: Complain loudly when others violate our flags From: Peter Zijlstra To: Tejun Heo Cc: Steven Rostedt , LKML , Thomas Gleixner Date: Fri, 30 Sep 2011 11:27:45 +0200 In-Reply-To: <20110930034815.GF10425@mtj.dyndns.org> References: <1317158254.26514.55.camel@gandalf.stny.rr.com> <20110930034815.GF10425@mtj.dyndns.org> Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 8BIT X-Mailer: Evolution 3.0.3- Message-ID: <1317374865.19415.15.camel@twins> Mime-Version: 1.0 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, 2011-09-29 at 20:48 -0700, Tejun Heo wrote: > IIRC, this was because there was no way to set PF_THREAD_BOUND once a > kthread starts to run and workers can stay active across CPU bring > down/up cycle. Per-cpu kthreads need PF_THREAD_BOUND to prevent cpu > affinity manipulation by third party for correctness. But that's the whole point isn't it. You mark threads that aren't strictly per-cpu with that. Aside from the unplug trainwreck, you also mark your unbound workers with that. There is no correctness issue what so ever with those, and userspace moving them about doesn't matter one whit.