[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] [PATCH RFC 1/1] virtio: migrate config_vector

From: Christian Borntraeger
Subject: Re: [Qemu-devel] [PATCH RFC 1/1] virtio: migrate config_vector
Date: Thu, 14 May 2015 12:38:03 +0200
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.6.0

Am 14.05.2015 um 12:09 schrieb Paolo Bonzini:
> On 14/05/2015 12:07, Michael S. Tsirkin wrote:
>>> If you really want that, the hook should be in the virtio device class.
>>> Keying on the machine type is almost never the answer, at least upstream.
>>> Distros can always override it if they know what they're getting into.
>> I beg to disagree.  Check out PC_COMPAT_ macros.
> I'm talking specifically about migration subsections.  None of them are
> keyed on the machine type downstream, as far as I know.
>> But if you are saying
>> migration with virtio-ccw is just too broken ATM, so we shouldn't care
>> about its migration compatibility upstream, then I can buy that.
> Yes, I'm also saying that (and Christian must agree or they would not
> have posted this patch).

So some background: Migration works most of the time but it will fail 
if config changes happen (like virtio-balloon). When trying to fix this Jason
came up with a patch that added qemu_put_be16(f, vdev->config_vector) in the 
virtio_ccw save_config. As this does not use vmstates and versions we were not
sure about how to ensure compatibility. So Conny came up with this proposal to
detect and reject migration from an old (broken) version.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]