From mboxrd@z Thu Jan 1 00:00:00 1970 From: Gleb Natapov Subject: Re: [PATCH] cpu hotplug issue Date: Thu, 21 Jul 2011 15:40:31 +0300 Message-ID: <20110721124031.GH3044@redhat.com> References: <20110720083507.GS2400@redhat.com> <4E2819EF.1040301@siemens.com> <20110721122559.GG3044@redhat.com> <4E281CFD.2030805@siemens.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Vasilis Liaskovitis , "kvm@vger.kernel.org" To: Jan Kiszka Return-path: Received: from mx1.redhat.com ([209.132.183.28]:2070 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751888Ab1GUMkf (ORCPT ); Thu, 21 Jul 2011 08:40:35 -0400 Content-Disposition: inline In-Reply-To: <4E281CFD.2030805@siemens.com> Sender: kvm-owner@vger.kernel.org List-ID: On Thu, Jul 21, 2011 at 02:35:09PM +0200, Jan Kiszka wrote: > On 2011-07-21 14:25, Gleb Natapov wrote: > > On Thu, Jul 21, 2011 at 02:22:07PM +0200, Jan Kiszka wrote: > >> On 2011-07-21 13:06, Vasilis Liaskovitis wrote: > >>> More precisely: To enable cpu-hotplug at each bisect stage, I apply > >>> this patch derived from: > >>> http://lists.gnu.org/archive/html/qemu-devel/2010-08/msg00850.html > >>> > >>> diff --git a/hw/qdev.c b/hw/qdev.c > >>> index 1aa1ea0..aed48ce 100644 > >>> --- a/hw/qdev.c > >>> +++ b/hw/qdev.c > >>> @@ -327,6 +327,7 @@ BusState *sysbus_get_default(void) > >>> if (!main_system_bus) { > >>> main_system_bus = qbus_create(&system_bus_info, NULL, > >>> "main-system-bus"); > >>> + main_system_bus->allow_hotplug = 1; > >> > >> BTW, this reminds me why "fixing" CPU hotplug won't help a lot. We > > I do not see relation between qdev not allowing cpu hotplug bug > > It's not a bug, it's a feature: The system bus we attach APICs to so far > is a statically configured beast, intentionally hotplug-incapable. > Hacking this away in qemu-kvm will be a step towards where we came from > and will break sooner or later again. We need a CPU/APIC bus, also for > other reasons, that can then be made hotplug-capable. > Agree. I was calling the whole situation the "bug", not this missing line in qdev in particular. > > and cpu been > > created with incorrect state bug. Fixing of former will not magically > > make later disappear. Both should be solved. > > The issues hidden by the topology design problem should be solved > nevertheless, that's true. > > Jan > > -- > Siemens AG, Corporate Technology, CT T DE IT 1 > Corporate Competence Center Embedded Linux -- Gleb.