From mboxrd@z Thu Jan 1 00:00:00 1970 From: Rusty Russell Subject: Re: [PATCH] virtio_net: add link status handling Date: Fri, 12 Dec 2008 18:34:48 +1030 Message-ID: <200812121834.49464.rusty@rustcorp.com.au> References: <1228817973.26198.1.camel@blaa> <493F336B.5010402@codemonkey.ws> <1228934084.5384.66.camel@blaa> Mime-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit Cc: Anthony Liguori , netdev , kvm To: Mark McLoughlin Return-path: In-Reply-To: <1228934084.5384.66.camel@blaa> Content-Disposition: inline Sender: kvm-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On Thursday 11 December 2008 05:04:44 Mark McLoughlin wrote: > On Tue, 2008-12-09 at 21:11 -0600, Anthony Liguori wrote: > > Rusty Russell wrote: > > > On Wednesday 10 December 2008 08:02:14 Anthony Liguori wrote: > > >> It would be nice if the virtio-net card wrote some acknowledgement that > > >> it has received the link status down/up events. > > > > > > How about of every status change event? ie. a generic virtio_pci solution? > > > > A really simple way to do it would just be to have another status field > > that was the guest's status (verses the host requested status which the > > current field is). All config reads/writes result in exits so it's easy > > to track. > > > > Adding YA virtio event may be a little overkill. > > Sounds very reasonable; that and Rusty's "mask out unknown bits" > suggestion in the version below. Not quite what I was after. I've taken the original patch, added the masking change. I'll test here and feed to DaveM. As far as Anthony's live migrate race, I've decided to declare with Canutian arrogance that it Will Never Happen. Cheers, Rusty.