Re: [PATCH 2/2] hyperv/synic: Allocate as ram_device

[Paolo Bonzini]

On 09/01/20 16:40, Vitaly Kuznetsov wrote:
>>>>> Do we want a new memory_region_init for that or just to be able to add
>>>>> a flag?
>>>>>
>>>> I think a flag API is preferable since this can apply to any kind of
>>>> region. But can go either way, Paolo's the maintainer there.
>>> (Copying Paolo in)
>>> So what exactly does this flag mean; to me it's 'no vhost' - but is it
>>> actually more general?
>> It has two more effects in addition to no vhost:
>>
>> 1) it is skipped when dumping the guest (is this a good or bad idea for
>> SynIC?)
> Imagine we have an not yet consumed message sitting in message page, or
> a signalled event, do I understand correctly that these are going to get
> lost upon migration? This may not work then -- unless we transfer
> in-QEMU synic state somehow separately.

(Thread necromancy).

This is just dumping (and it should probably be extended to core
dumping, i.e. MADV_DONTDUMP).  Migration is controlled separately with
vmstate_register_ram.

Paolo