From mboxrd@z Thu Jan 1 00:00:00 1970 From: Markus Armbruster Subject: Re: virtio block drivers not working Date: Mon, 23 Mar 2009 08:48:04 +0100 Message-ID: <87y6uwvfwr.fsf@pike.pond.sub.org> References: <20090322213035.GA8400@infradead.org> <7EE71E85-ED07-4C3D-8747-092F8F478A8E@aei-tech.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: kvm@vger.kernel.org To: Caleb Tennis Return-path: Received: from mx2.redhat.com ([66.187.237.31]:55783 "EHLO mx2.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752043AbZCWHsK (ORCPT ); Mon, 23 Mar 2009 03:48:10 -0400 In-Reply-To: <7EE71E85-ED07-4C3D-8747-092F8F478A8E@aei-tech.com> (Caleb Tennis's message of "Sun\, 22 Mar 2009 18\:06\:17 -0400") Sender: kvm-owner@vger.kernel.org List-ID: Caleb Tennis writes: > On Mar 22, 2009, at 5:30 PM, Christoph Hellwig wrote: > >> I do you virtio block in a very similar setup to yours (fully static >> kernel, -kernel option to kvm/qemu) sucesfully for quite a a while. >> >> Can you post your kernel .config and the contents of /proc/devices >> and /proc/partitions to debug this further? >> > > I've found the problem. In my testing, I was changing my second drive > from ide to virtio, so it looked like this: > > -drive index=0,media=disk,if=ide,file=/root > -drive index=1,media=disk,if=virtio,file=/data > > This doesn't work though, even as vdb, as since the interface is the > first virtio, the index needs to be set back to 0. This gives me a / > dev/vda like I was expecting. > > Thanks, > Caleb Yes, the common loop to create virtio block devices (all three copies of it) stops at the first unused unit number / index. A rude trap for the unwary. The device configuration patch I'm working on upstream doesn't do that.