From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754811Ab1JCBWz (ORCPT ); Sun, 2 Oct 2011 21:22:55 -0400 Received: from mail-iy0-f174.google.com ([209.85.210.174]:34474 "EHLO mail-iy0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751243Ab1JCBWs (ORCPT ); Sun, 2 Oct 2011 21:22:48 -0400 Date: Sun, 2 Oct 2011 18:22:44 -0700 From: Tejun Heo To: Peter Zijlstra Cc: Steven Rostedt , LKML , Thomas Gleixner Subject: Re: [PATCH] sched/kthread: Complain loudly when others violate our flags Message-ID: <20111003012244.GF31799@mtj.dyndns.org> References: <1317158254.26514.55.camel@gandalf.stny.rr.com> <20110930034815.GF10425@mtj.dyndns.org> <1317374865.19415.15.camel@twins> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1317374865.19415.15.camel@twins> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Sep 30, 2011 at 11:27:45AM +0200, Peter Zijlstra wrote: > There is no correctness issue what so ever with those, and userspace > moving them about doesn't matter one whit. I wanted to make it consistent and disallow diddling wq worker affinity from userland. As those workers are shared and managed dynamically, it didn't make sense to allow affinity manipulation. That said, it might make sense allowing control over affinity of unbound wq workers for things like cpu isolation. Userspace doing that per-worker doesn't make whole lot of sense. Allowing putting unbound workers automatically into a cpuset would work better, I think. Thanks. -- tejun