From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:35968) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1ZkSEo-00077E-5l for qemu-devel@nongnu.org; Fri, 09 Oct 2015 03:37:11 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1ZkSEh-0006nn-5k for qemu-devel@nongnu.org; Fri, 09 Oct 2015 03:37:08 -0400 Received: from mx1.redhat.com ([209.132.183.28]:40031) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1ZkSEh-0006nX-0q for qemu-devel@nongnu.org; Fri, 09 Oct 2015 03:37:03 -0400 Date: Fri, 9 Oct 2015 10:36:56 +0300 From: "Michael S. Tsirkin" Message-ID: <20151009103639-mutt-send-email-mst@redhat.com> References: <20151005090710.20420.26941.stgit@bahia.huguette.org> <20151005090722.20420.71571.stgit@bahia.huguette.org> <20151006080914.GB19089@stefanha-thinkpad> <87pp0r12qz.fsf@linux.vnet.ibm.com> <20151009092349.16dc882c@bahia.local> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20151009092349.16dc882c@bahia.local> Subject: Re: [Qemu-devel] [PATCH 2/3] virtio-9p: add unrealize handler List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Greg Kurz Cc: Stefan Hajnoczi , "Aneesh Kumar K.V" , qemu-devel@nongnu.org On Fri, Oct 09, 2015 at 09:23:49AM +0200, Greg Kurz wrote: > On Wed, 07 Oct 2015 14:20:28 +0530 > "Aneesh Kumar K.V" wrote: > > > Stefan Hajnoczi writes: > > > > > On Mon, Oct 05, 2015 at 11:07:23AM +0200, Greg Kurz wrote: > > >> If the user tries to hot unplug a virtio-9p device, it seems to succeed but > > >> in fact: > > >> - virtio-9p coroutines thread pool and async queue are leaked > > >> - QEMU crashes in virtio_vmstate_change() if the user tries to live migrate > > >> > > >> This patch brings hot unplug support to virtio-9p-device. It fixes both > > >> above issues. > > >> > > >> Signed-off-by: Greg Kurz > > >> --- > > >> hw/9pfs/virtio-9p-device.c | 12 ++++++++++++ > > >> 1 file changed, 12 insertions(+) > > > > > > What happens to in-flight I/O requests? We cannot assume that the guest > > > driver quiesces the device. > > > > We enable migration blocker when we have an active mount. So if we get > > here, that should indicate no active 9p mounts. > > > > -aneesh > > Oops.. Stefan is talking about hot-unplug versus in-flight requests... not > about migration. And there is no such thing as a hot-unplug blocker... If unplug request fails, that should be enough. > -- > Greg