On 19 December 2012 20:00, Blue Swirl <blauwir...@gmail.com> wrote:
> On Wed, Dec 19, 2012 at 9:53 AM, <fred.kon...@greensocs.com> wrote:
>> +#ifdef __linux__
>> +#define DEFINE_VIRTIO_BLK_PROPERTIES(_state, _field) \
>> + DEFINE_BLOCK_PROPERTIES(_state, _field.conf), \
>> + DEFINE_BLOCK_CHS_PROPERTIES(_state, _field.conf), \
>> + DEFINE_PROP_STRING("serial", _state, _field.serial), \
>> + DEFINE_PROP_BIT("scsi", _state, _field.scsi, 0, true)
>> +#else
>> +#define DEFINE_VIRTIO_BLK_PROPERTIES(_state, _field) \
>> + DEFINE_BLOCK_PROPERTIES(_state, _field.conf), \
>> + DEFINE_BLOCK_CHS_PROPERTIES(_state, _field.conf), \
>> + DEFINE_PROP_STRING("serial", _state, _field.serial)
>
> Does the different property set mean that it would not be possible to
> migrate VMs which use virtio from non-Linux host to Linux and vice
> versa?
Probably, but note that this is just reflecting an existing ifdef in
the non-refactored implementation.
(I looked at that existing ifdef and thought "smells dubious", but
presumably there is a reason.)
-- PMM
