Re: [Qemu-devel] [PATCH v4 10/11] virtio: migration support for packed ring

[Jason Wang <jasowang@redhat.com>]

On 2019/2/19 下午7:00, Wei Xu wrote:
> On Tue, Feb 19, 2019 at 03:30:41PM +0800, Jason Wang wrote:
>> On 2019/2/14 下午12:26,wexu@redhat.com  wrote:
>>> From: Wei Xu<wexu@redhat.com>
>>>
>>> Both userspace and vhost-net/user are supported with this patch.
>>>
>>> A new subsection is introduced for packed ring, only 'last_avail_idx'
>>> and 'last_avail_wrap_counter' are saved/loaded presumably based on
>>> all the others relevant data(inuse, used/avail index and wrap count
>>> should be the same.
>> This is probably only true for net device, see comment in virtio_load():
>>
>>              /*
>>               * Some devices migrate VirtQueueElements that have been popped
>>               * from the avail ring but not yet returned to the used ring.
>>               * Since max ring size < UINT16_MAX it's safe to use modulo
>>               * UINT16_MAX + 1 subtraction.
>>               */
>>              vdev->vq[i].inuse = (uint16_t)(vdev->vq[i].last_avail_idx -
>>                                  vdev->vq[i].used_idx);
>>
>>
>> So you need to migrate used_idx and used_wrap_counter since we don't have
>> used idx.
> This is trying to align with vhost-net/user as we discussed, since all we
> have done is to support  virtio-net device for packed ring,


Well, what you did probably work for virtio-net. But again, virtio-net 
is not the only device we want to support.


>   maybe we can
> consider supporting other devices after we have got it verified.
>

Please don't and fix the issue, packed virtqueue is a generic layout 
changes that should work for all kinds of virtio devices. It should be 
specially optimized for net, but it doesn't mean it was only designed 
for net.

Thanks