On Wed, Apr 13, 2016 at 09:04:54PM +0300, Roman Kagan wrote:
On Wed, Apr 13, 2016 at 06:39:55PM +0100, Richard W.M. Jones wrote:
> I think this is possible, although massively unlikely in practice. I
> think the different controllers would have different
> index="0"/index="1"/... attributes.
How the binding of drives to controllers is done, then? Is it specified
in drive's "address"?
Yes, apparently using <address/> under <disk/>.
This case is handled fine in the patch, isn't it?
has_virtio_scsi is a
local variable which is only consulted when populating per-drive
s_controller field: for drives with bus='scsi' it differentiates between
Source_virtio_SCSI and Source_SCSI.
Oh I see, target_bus is consulted for each drive. So this is right
(except for the scsi/virtio-scsi case, but who knows about that).
The place where we make up a single disk type is
v2v.ml:rcaps_from_source, but there I explicitly error out if the guest
has multiple disk types (which may be too harsh but simplifies things
for now).
Yup.
Rich.
--
Richard Jones, Virtualization Group, Red Hat
http://people.redhat.com/~rjones
Read my programming and virtualization blog:
http://rwmj.wordpress.com
virt-df lists disk usage of guests without needing to install any
software inside the virtual machine. Supports Linux and Windows.
http://people.redhat.com/~rjones/virt-df/