[libvirt] [PATCH v2 2/4] conf: Parse virtio-crypto in the domain XML

[Martin Kletzander]

On Wed, Feb 08, 2017 at 11:25:21AM +0800, Longpeng (Mike) wrote:
>Hi Martin,
>
>On 2017/2/7 20:15, Martin Kletzander wrote:
>
>> On Wed, Jan 11, 2017 at 04:28:24PM +0800, Longpeng(Mike) wrote:
>>> This patch parse the domain XML with virtio-crypto
>>> support, the virtio-crypto XML looks like this:
>>>
>>>  <crypto model='virtio'>
>>>    <backend type='builtin' queues='1'/>
>>>  </crypto>
>>>
>>> Signed-off-by: Longpeng(Mike) <longpeng2 at huawei.com>
>>> ---
>>> src/conf/domain_conf.c         | 213 ++++++++++++++++++++++++++++++++++++++++-
>>> src/conf/domain_conf.h         |  32 +++++++
>>> src/libvirt_private.syms       |   2 +
>>> src/qemu/qemu_domain.c         |   2 +
>>> src/qemu/qemu_domain_address.c |   1 +
>>> src/qemu/qemu_driver.c         |   6 ++
>>> src/qemu/qemu_hotplug.c        |   1 +
>>> 7 files changed, 256 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c
>>> index 52aee2b..ef44930 100644
>>> --- a/src/conf/domain_conf.c
>>> +++ b/src/conf/domain_conf.c
>>> @@ -18967,6 +19096,25 @@ virDomainRNGDefCheckABIStability(virDomainRNGDefPtr src,
>>>
>>>
>>> static bool
>>> +virDomainCryptoDefCheckABIStability(virDomainCryptoDefPtr src,
>>> +                                    virDomainCryptoDefPtr dst)
>>> +{
>>> +    if (src->model != dst->model) {
>>> +        virReportError(VIR_ERR_CONFIG_UNSUPPORTED,
>>> +                       _("Target Crypto model '%s' does not match source '%s'"),
>>> +                       virDomainCryptoModelTypeToString(dst->model),
>>> +                       virDomainCryptoModelTypeToString(src->model));
>>> +        return false;
>>> +    }
>>> +
>>
>> The number of queues is not part of ABI?  That'd make sense, I'm just
>> making sure.
>
>
>Oh, yep! I think it's necessary to check 'queues' for future scalability,
>although QEMU cryptodev only support one queue currently.
>

Just try migrating (or restoring) a domain with one queue to a domain
with multiple queues.  If that works, there is no need to check for the
number of queues in this function.

>I will take all your other suggestions and rebase the patchset on master in V3.
>
>Thanks. :)
>
>>
>...
>>>     case VIR_DOMAIN_DEVICE_LAST:
>>>     case VIR_DOMAIN_DEVICE_NONE:
>>>         return 0;
>
>
>--
>Regards,
>Longpeng(Mike)
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 801 bytes
Desc: Digital signature
URL: <http://listman.redhat.com/archives/libvir-list/attachments/20170208/5fe3c2df/attachment-0001.sig>