From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Michael S. Tsirkin" Subject: Re: About hotplug multifunction Date: Mon, 12 Sep 2011 15:24:59 +0300 Message-ID: <20110912122458.GC12294@redhat.com> References: <1749303074.590739.1315551830460.JavaMail.root@zmail05.collab.prod.int.phx2.redhat.com> <1466280941.590770.1315552101666.JavaMail.root@zmail05.collab.prod.int.phx2.redhat.com> <20110909073226.GB5593@redhat.com> <1315587902.2662.19.camel@x201.home> <20110909183426.GD22662@amt.cnet> <20110911092357.GC27096@redhat.com> <20110911150149.GA28477@amt.cnet> <20110911185106.GA4740@redhat.com> <20110912102148.GA12332@amt.cnet> <20110912105607.GA12849@amt.cnet> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Alex Williamson , Amos Kong , qemu-devel@nongnu.org, kvm@vger.kernel.org, yamahata@valinux.co.jp, blauwirbel@gmail.com To: Marcelo Tosatti Return-path: Received: from mx1.redhat.com ([209.132.183.28]:47335 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755240Ab1ILMYA (ORCPT ); Mon, 12 Sep 2011 08:24:00 -0400 Content-Disposition: inline In-Reply-To: <20110912105607.GA12849@amt.cnet> Sender: kvm-owner@vger.kernel.org List-ID: On Mon, Sep 12, 2011 at 07:56:07AM -0300, Marcelo Tosatti wrote: > On Mon, Sep 12, 2011 at 07:21:48AM -0300, Marcelo Tosatti wrote: > > > We could, for example, keep a stub function 0 around. > > > > I suppose the guest will remove all functions of a device once you > > attempt to hot-unplug a function. > > > > What is the problem with adding more PCI buses, instead of multifunction > > ? > > The advantage is that its not only possible to use virtio devices, but any > kind of PCI device. Yes. But it is a new feature. Creating multifunction devs is *already* possible. Hotplug is partially broken, so getting it into a consistent non buggy state has value imho. -- MST