From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1423200AbWF1HyE (ORCPT ); Wed, 28 Jun 2006 03:54:04 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1030471AbWF1HyE (ORCPT ); Wed, 28 Jun 2006 03:54:04 -0400 Received: from mx2.suse.de ([195.135.220.15]:17623 "EHLO mx2.suse.de") by vger.kernel.org with ESMTP id S1030468AbWF1HyD (ORCPT ); Wed, 28 Jun 2006 03:54:03 -0400 Message-ID: <44A23598.4020108@suse.de> Date: Wed, 28 Jun 2006 09:54:00 +0200 From: Gerd Hoffmann User-Agent: Thunderbird 1.5 (X11/20060317) MIME-Version: 1.0 To: Dave Jones , Linux Kernel Mailing List , Gerd Hoffmann , Andi Kleen , Linus Torvalds Subject: Re: [PATCH] x86_64: x86_64 version of the smp alternative patch. References: <200606261900.k5QJ0k9J028243@hera.kernel.org> <20060627175741.GF1280@redhat.com> In-Reply-To: <20060627175741.GF1280@redhat.com> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org > This has one behaviour slightly different to the i386 version however. > If you boot an SMP machine it does this.. > > SMP: Allowing 4 CPUs, 2 hotplug CPUs > Initializing CPU#0 > CPU: Physical Processor ID: 0 > CPU: Processor Core ID: 0 > CPU0: Thermal monitoring enabled (TM1) > SMP alternatives: switching to UP code > .. > SMP alternatives: switching to SMP code > Booting processor 1/2 APIC 0x1 > Initializing CPU#1 i386 shows the same behavior btw, but with CPU_HOTPLUG=y only. "booting the second CPU" is just a special case of "plugging in a CPU" then, that is where the double-switch comes from. cheers, Gerd -- Gerd Hoffmann http://www.suse.de/~kraxel/julika-dora.jpeg