From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263555AbUGICwO (ORCPT ); Thu, 8 Jul 2004 22:52:14 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263664AbUGICwO (ORCPT ); Thu, 8 Jul 2004 22:52:14 -0400 Received: from holomorphy.com ([207.189.100.168]:21225 "EHLO holomorphy.com") by vger.kernel.org with ESMTP id S263555AbUGICwE (ORCPT ); Thu, 8 Jul 2004 22:52:04 -0400 Date: Thu, 8 Jul 2004 19:51:51 -0700 From: William Lee Irwin III To: Nick Piggin Cc: "David S. Miller" , Ingo Molnar , akpm@osdl.org, linux-kernel@vger.kernel.org Subject: Re: 2.6.7-mm6 Message-ID: <20040709025151.GV21066@holomorphy.com> Mail-Followup-To: William Lee Irwin III , Nick Piggin , "David S. Miller" , Ingo Molnar , akpm@osdl.org, linux-kernel@vger.kernel.org References: <20040705023120.34f7772b.akpm@osdl.org> <20040706125438.GS21066@holomorphy.com> <20040706233618.GW21066@holomorphy.com> <20040706170247.5bca760c.davem@redhat.com> <20040707073510.GA27609@elte.hu> <20040707140249.2bfe0a4b.davem@redhat.com> <40EE06B1.1090202@yahoo.com.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <40EE06B1.1090202@yahoo.com.au> User-Agent: Mutt/1.5.5.1+cvs20040105i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org David S. Miller wrote: >> The parent's regs (stored in current_thread_info() at trap time, >> and also needed by copy_thread() processing) will also be garbage >> since we're avoiding the fork syscall trap. >> In short, this won't work :) >> This is why I use kernel_thread(). Why is that so bad? On Fri, Jul 09, 2004 at 12:45:05PM +1000, Nick Piggin wrote: > We could make CLONE_IDLETASK clones not do the wakeup? > Ingo? I guess an alternative is to have the arch explicitly > make a call to dequeue it. This is all just context switching and bootstrap ordering, but I really have other vastly more urgent things to do at the moment than cleanups. Please present a self-contained fixed-up init_idle() cleanup for me to testboot. Even the one in -mm is not so, as it depends on later patches to even compile. -- wli