On 11/11/2015 10:09, Denis V. Lunev wrote:
QEMU accepts this option through HyperV set of CPU options.
Thus your proposal means that in this case we will have to add
this option to CPU part of the command line from devices
part.
Is this correct?
And, finally, 'no', this is not a device from QEMU perspective
of view.
That doesn't really matter. It's okay if libvirt and QEMU model feature
in a slightly different manner.
For an example, see how kvmclock is changed from a <timer> element to an
option of -cpu.
I think I now agree with the libvirt folks. The effects of hv_crash are
the same as the effects of <panic>, for example the VM will not anymore
obey the "reboot on BSoD" setting, so it should be added to <panic>.
Paolo